-------
63      AZEOTROPIC DISTILLATION - A CONTINUING EVALUATION FOR THE
             DETERMINATION OF POLAR, WATER - SOLUBLE ORGANICS


        Paul H. Cramer and Jay Wilner, Midwest Research Institute
        James W. Eichelberger, USEPA, Environmental Monitoring Support - Cincinnati


        ABSTRACT.  The  determination of volatile organic compounds (VOCs) is an
        important step in the assessment of water quality. Methods for the determination
        of water-soluble compounds such as ketones, aldehydes, nitriles, and alcohols
        have not been developed for routine use because of the difficulties in removing and
        concentrating these compounds from the aqueous matrix.  Azeotropic distillation
        has been  evaluated  as  a  possible  method for determining the aqueous
        concentration of selected compounds from  the RCRA Appendix VIII, Michigan,
        and BDAT analytes lists.

        An  aqueous  binary azeotrope is a mixture of an organic compound and water
        which produces  a vapor with the same composition as the liquid when boiled. A
        minimum boiling azeotrope boils at a lower temperature than either the water or
        the organic compound, and as such, can be  removed from the aqueous sample by
        careful distillation. Thus, azeotropic distillation can be a  viable method for
        determining the concentration of those compounds which form binary azeotropes
        with water.

        The objectives of this continuing program were to: (1) determine the number of
        target analytes  that could be successfully concentrated by azeotropic distillation,
        (2)  determine the maximum number of analytes that could be chromatographed
        simultaneously  by  direct aqueous injection HRGC and (3) determine the overall
        method performance for each compound. Data from two distillation methods will
        be presented, namely, trap-to-trap distillation under low vacuum (<0.1 mm Hg)
        and atmospheric distillation using a modified Nielsen-Kryger distillation head.

        The anaytes were  tested individually for chromatographic performance on a
        selection of wide-bore fused  silica columns.   Direct aqeuous injection was
        performed since the sample would be in aqueous  solution after distillation. Gas
        chromatographic conditions were optimized to resolve the greatest number of
        analytes  simultaneously.    The   analytes  that  could  be  successfully
        chromatographed were tested for their ability to azeotropically distill. Recoveries
        were determined from  the distillation, investigations were made into suitable
        surrogates and internal standards, and precision and accuracy were collected at
        three concentrations levels.  Method performance and method detection limits
        were determined using "real-world"  samples.  Stability of the  analytes in
        chlorinated and dechlorinated waters was also determined over a 14-day holding
        time.
                                         n-92

-------
6 4           A METHOD FOR THE CONCENTRATION AND ANALYSIS
                               OF TRACE METHANOL
           IN WATER BY DISTILLATION AND GAS CHROMATOGRAPHY

                     Mark L. Bruce, Richard P. Lee, Marvin W. Stephens,
                           Wadsworth/ALERT Laboratories, Inc.,
                     4101 Shuffel Dr. N.W., North Canton, Ohio 44720.

     ABSTRACT

     Under the land disposal restriction (40 CFR part 268.41) for spent solvents,  methanol
     has a treatment standard of 0.25 mg/L for wastewaters containing spent solvents and 0.75
     mg/L for  all other spent solvent wastes in the waste extract using zero headspace
     extraction. This paper presents the  development of an aqueous sample concentration,
     cleanup and analysis method that surpasses these regulatory treatment standards for
     methanol.  The total sample handling time from the start of distillation to the completion
     of analysis is less than one hour.  The initial experimental parameters were derived from a
     method for the azeotropic distillation of water soluble volatile organic compounds (1,2).
     This method uses the principles of distillation and steam stripping.  Several modified
     Nielson   Kryger condenser designs were developed to enrich (relative to the original
     sample) the resulting distillate in methanol. The distillate is then analyzed  by gas
     chromatography using a DB-WAX capillary column and flame ionization detection.  For
     this application the instrument detection limit was 0.15 ng in a 2 |iL injection. With a 40
     mL sample,  the method detection limit was 0.026 mg/L for reagent water and 0.031
     mg/L for ZHE extract.

     INTRODUCTION

     The Hazardous and Solid Waste Amendments of 1984 amended RCRA by banning all
     land disposal of untreated hazardous waste within five and one-half years after passage
     (May 8, 1990).   The  basic purpose of the land disposal restrictions is to discourage
     activities that involve placing untreated wastes in or on the land when a better treatment
     or destruction alternative exists. Under the land disposal restrictions  (40 CFR  part
     268.41) for spent solvents,  methanol has a treatment standard of  0.25  mg/L for
     wastewaters containing spent solvents and 0.75 mg/L for all other spent solvent wastes in
     the waste extract using zero headspace extraction (ZHE).  The proposed  treatment
     standard for multi-source leachate wastewaters is 0.033 mg/L for methanol.  To date there
     are no EPA approved methods  for methanol that have detection limits below these
     treatment  standards.   The effect of this situation is that residues from the treatment of
     solvent wastes and multi-source leachate wastewaters cannot at present be certified to
     meet the corresponding treatment standards and thus be landfilled.

     This paper presents the  development of an aqueous  sample concentration, cleanup and
     analysis method with a detection limit lower than the spent solvent treatment standards
     for methanol. The method also shows promise for meeting the detection limit required
     for the proposed treatment standard for multi-source leachate wastewaters.  The total
     sample handling time from the start of distillation to the completion  of analysis is less
     than one hour. The initial experimental parameters were derived from a method for the
     azeotropic distillation of water soluble volatile organic compounds (1,2). This  method
                                          n-93

-------
uses the principles of distillation and steam stripping.  When distilling a 40 mL aqueous
sample or ZHE extract, actual distillation time from the start of visible boiling is 10
minutes or less.  GC run time is approximately 17 minutes.  A significant advantage is
that the distillate is free from nonvolatile organic and inorganic interferences. These
nonvolatile components may degrade chromatographic performance and shorten the life
of the GC column.

INSTRUMENTATION. EHTTTPMFNT AND SUPPLIES

  Gas Chromatograph/Data System
   Hewlett Packard 5890 equipped with a flame ionization detector,  Macintosh Ilci
     (Apple) with Lab View (National Instruments) and  GC Integrator & Workmate
     (WillS tein) software.
  Columns
   Quantitation: DB-Wax 30m X 0.53mm I.D. 1.0 micron film thickness
   Confirmation: DB-1 30m X 0.53mm I.D. 1.5 micron film thickness
  Glassware
   Modified Nielson Kryger condenser; Ace Catalog # 6555-07, Shamrock Catalog #
     6170
   Vigrex columns
     200 mm column (350 mm overall) 24/40 joint Shamrock Catalog # 23202-102
     130 mm column (200 mm overall) 14/20 joint Shamrock Catalog # 23242-106
   Round bottom flasks
     2L, 500 mL, 100 mL
   Glas-Col Heating Mantles
     2 L flask, 500 watts, Catalog # 04100
     500 mL flask, 270 watts Catalog # TM106
     100 mL flask, 230 watts Catalog # STM400
   Fisher burner
   Prototype Volatile Organic Compound Concentrating Condensers (VOC^)
     Shamrock Glass, Seaford, Delaware
  Methanol, B&J Brand Catalog # 232-1, purity 99.9%
  Reagent water, deionized, (methanol content must be less than practical quantitation
  limit)
  Glass beads, 5 mm diameter
  Boiling chips, VWR Scientific, Inc. Porous Boiling Chips Catalog # 26397-409
  Cold water supply, Neslab Coolflow CFT-25

DISTILLATION PRTNCTPT .F.S

Methanol is volatilized from the aqueous sample by boiling. The steam is enriched in
methanol relative to the  original sample. As the steam passes up through the Vigrex
fractionation column it is further enriched in methanol with each plate. The steam is then
condensed and collected in a small volume collection chamber in the concentrating
condenser. Figure 1 shows the overall distillation system. Once the collection chamber
is filled it overflows and methanol enriched water flows back through the lower part of
the condenser and the Vigrex column toward the flask.  This methanol enriched water
comes  in contact with additional rising steam.  The steam revolatilizes most of the
methanol and carries it back to the condenser. When equilibrium is reached most of the
methanol has been trapped in the collection chamber.  The methanol concentration
                                     n-94

-------
enhancement factor is equal to the ratio of the sample volume divided by the collection
chamber volume.  Figure 2 is a cross section of the main section of the condenser.
j




E
E
o
0
C\J
i t

E
o
LO
CD



H •








•
                                    -Condenser
                                    -Vigrex column
                                      2 L Flask
                                       Heating Mantle
                                  Figure 1
                 Nielson  Kryger Condenser/Vigrex/Flask/Mantle
                            53 mm
                                                 Cooling water

                                                 Condensed Steam

                                               Collection Chamber
                                                (Volume = 20 ml_)
                                                 Overflow tube
                                                   (4 mm ID)
                                                 Overflowing Water
                                                      Stopcock
                                  Figure 2
               Main section of Modified Nielson  Kryger Condenser
                                    n-95

-------
For efficient steam stripping of methanol the overflowing condensed water must flow in a
thin layer. If the overflow water layer is thick, methanol is not be able to diffuse to the
surface and be revolatilized efficiently by the steam. This situation will occur if large
drops form or a wave of water is released from the collection chamber. See Figures 4 and
5 for an illustration of this phenomenon. The overflow mechanism must account for this
or poor recoveries  will result. If the drop or wave is large enough it may actually travel
all the way down to the flask carrying much of the methanol with it.  In addition when the
condensed steam flow rate was high and distances between tube walls were small, water
bridges formed which also hindered the steam stripping process.
                                                 CH3OH
             Tube Wall
             Water (H2o)

 Methanol Liquid
         H2o  Steam
III   CH3OH   Methanol Gas
                                   Figure 3
                        The Steam Stripping of Methanol
                                          Collection Chamber

                                          Overflow Tube

                                          Water Bridge


                                          Water Drop
                                  Figure 4
                                  _ -^ ^.
        Modified Nielson  Kryger Condenser with Drops and Water Bridges
                                    11-96

-------
                            Collection Chamber

             Water Meniscus
                               Overflow Tube

                                 Scale 2 : 1
                                    Figure 5
                          Alternate Design with Waves
          {just before the meniscus breaks (A) and shortly after it breaks (B)}

It is desirable for the steam to come in close  contact with the overflowing water to
volatilize most of the methanol.  A narrow condenser neck and overflow tube can
accomplish this. But when the inside diameter of these tubes is too small flow problems
also result.  The size and frequency of the waves as shown in Figure 5 are determined by
the flow of condensed water  (which is derived from the sample boil/reflux rate), the
inside diameter of the overflow tube, the distance between the overflow tube and the
collection chamber outside wall and tube materials. If the overflow tube inside diameter
and/or the distance to the chamber wall is small the surface tension of water  will cause
the water level to rise above the top of the overflow tube.  Once there is sufficient water
pressure on this meniscus it will break and send a wave back to the flask. When the wave
is  thick poor methanol recoveries  will result.  The surface tension/wave effect is
exacerbated by using nonpolar materials, such  as Teflon, for the overflow tube.  The
surface  tension effect can be reduced by  notching or flaring the overflow tube.
Alternately a wick can be used. But in all cases the water overflow should be smooth and
in a thin layer.  The sample  must boil smoothly as well.  If the sample bumps  and
splatters the overflowing water will reflect this and yield poor recoveries. If the flow rate
of steam (sample boil/reflux rate)  is too high and the condenser neck is too narrow the
overflowing water will be blown back into the collection chamber rather than trickling
back through the Vigrex column to the flask.

Although this method is applicable to many water soluble volatile organic compounds it
was developed and optimized specifically for methanol. Since methanol does not form an
azeotrope with water this method is technically not azeotropic distillation even though the
basic principles  are the same.
                                      n-97

-------
DISCUSSION AND RESULTS

Many parameters were investigated.  There are physical parameters such as the sample
volume, it's boil/reflux rate  and the total distillation  time.  The physical design
characteristics of the condenser itself were examined  The most critical of these is the
collection chamber volume. Several condenser designs were examined; a commercial
modified Nielson-Kryger, and  two miniaturized Nielson-Kryger (Peters)  systems.
Several alternate overflow systems were studied; the straight tube, notched, flared, wick
and hoop systems.  Two chemical parameters were also studied; methanol concentration
and matrix. Table 1 lists the parameters that were studied.

                        Table 1    Distillation Parameters
         Physical
            sample volume
            boil/reflux rate
            evenness of boil
            distillation time
            cooling water temperature
40 to 1000 mL
2 to 7 mL/min
glass beads or boiling chips
5 to 120 minutes
7°C
         Physical design
            Vigrex column
            VOC3 design
            collection chamber volume
            condenser height
            overflow design
            overflow tube inside
            diameter
            overflow tube height
            overflow tube shape
            (alternate design only)
         Chemical
            analyte concentration
            matrix
with or without

1 to 20 mL
15 to 60 cm cooling coil, baffles
Peters/Dow and alternate
2 to 10 mm

2 to 35 mm
straight, notched, flared, side drain,
   wick, hoop	
detection limit to 10 mg/L
reagent water, ZHE extract
MODIFIED NIELSON - KRYGER

Initial experiments used a commercially available Modified Nielson  Kryger condenser
from Ace Glass.  It uses basically the design described by Peters (2) except that most
dimensions are much larger. It is shown as a cross section in Figures 1 and 2.  In addition
sample removal is through a stopcock rather than with a syringe as described by Peters
(2). The collection chamber volume was quite large, 20 mL. This necessitated the use of
large sample volumes, 200 to  1000 mL,  to achieve the desired concentration factor.
Factorial design experiments indicated that 70% recovery and estimated detection limits
in the mid ppb range could be  obtained with distillation times of one hour.   The
boil/reflux rate was 5 mL/min. Increasing the  sample volume should theoretically lower
detection limits but in actual practice the improvement was not as good as expected since
the methanol recovery was reduced.  Longer distillation times improved recoveries to a
point. However sample preparation times of 1  to 2 hours were not desirable.
                                      n-98

-------
Several other parameters were also examined. The necessity of smooth or even rolling
boil was also noted. Rough boiling or bumping was seen when glass beads were used;
but the boiling was much smoother with boiling chips.  Any irregularity in the boiling
rate was reflected in the steam flow rate which affected the water condensation rate.
When the water condensation rate changed rapidly large drops or waves of overflow
water were seen.  As  described above this leads to poor recoveries.  Acetate buffer
solutions were also studied as sample matrixes. No significant difference in recovery was
found between reagent water and the acetate buffer. The primary matrix of interest was
the ZHE extract .  Normally only  200 to 300  mL is available for all of the volatile
analyses. After examining all of these factors it was considered necessary to miniaturize
the condenser, particularly the collection chamber volume.  Miniaturizing allowed for
both smaller sample volumes and shorter distillation times.

PROTOTYPE #1

The  first miniaturized prototype, called a Volatile Organic Compound Concentrating
Condenser (VOC3) focused on reducing the collection chamber volume.  It is shown in
Figure 6.  This allowed the use of smaller sample volumes while still maintaining a
sufficient concentration factor. The collection chamber volume was reduced to 5 mL.
The condenser height was 550 mm. Full and fractional factorial design experiments were
used to study many of the parameters.  The boil/reflux rate was varied from 1 mL/min. to
5 mL/min.  The effect on recovery was very small  when distillation times where at least
15 minutes.  When sample volume was varied from 40 to 250 mL the recovery was better
at the smaller volumes by 20%. Varying the methanol concentration from 0.025 mg/L to
10 mg/L did not produce a significant effect.  Adding the ZHE acetate buffer matrix to
reagent water improved recoveries up to 10% under some non-optimum conditions but in
general had little effect.  Distillation times were shortened from 1 hour to 5-10 minutes
without sacrificing recovery when the sample volumes were decreased from 250 mL to
40 mL. With a 10 minute distillation time, 4 to 5 samples can be distilled per hour using
one distillation system. Experiments were run with and without the Vigrex fractionation
column.  Average recovery was 90% with and 35% without the column.  Thus the
fractionation provided by the Vigrex column is necessary.

ANALYSTS

EPA SW-846 Method 8015 was used for analyzing the  concentrated aqueous samples.
The analytical conditions are summarized in Table 2.

                          Table 2   Analysis Parameters
quantitation column
confirmation column
instrument calibration range
response factor %RSD
response factor %D
methanol retention time
injection volume
injection type
injection port temperature
temperature program
DB-Wax
DB-1
0.2 to 2000 ng
3.56 ± 0.07 min.
 2uL
splitless
180°C
external standardization
carrier gas
carrier gas flow
detector
detector temperature
hydrogen flow
air flow
make-up gas
make-up gas flow
helium
2.5 mL/min.
FID
230°C
37 mL/min.
426 mL/min.
nitrogen
41 mL/min.
35°C for 0.0 min., 5°C/min. to 100°C, hold for 2 min.
                                      n-99

-------
The instrument detection limit (IDL) was calculated to be 0.15 ng using ten  2  (J.L
injections of a 0.10 mg/L standard.
IDL = (tn-i,99%)(Std Dev) = (2.821)(0.0528) = 0.15 ng.

A special note of caution regarding the GC temperature program is in order. Even though
methanol elutes relatively early, the GC temperature must be ramped up high enough and
held there long enough to remove any water and other compounds from the capillary
column. Retention time shifts may result if the water is not completely eluted from  the
column.

Since methanol is a common laboratory solvent it is difficult to obtain methanol free
water.  One deionized lab water system contaminated reagent water with methanol.  Also
airborne methanol can be absorbed by water in open containers.  The concentrator unit
appears to be able to extract methanol  from the air and contaminate a  sample that is
undergoing distillation.

                            32 mm
                           24 mm
                            10 mm
                                           Cooling water
                                              Collection Chamber
                                               (Volume = 5 ml)
                                                 Overflow tube
                                                 " (4 mm OD)
                                                   Stopcock
                                   Figure 6
                     Main section of Prototype #1 Condenser

METHOD DETECTION T.TMTT

Method detection limit studies  for reagent water and zero headspace extract  were
performed.  The reagent water detection limit study is summarized in the Table 3.
                                     n-ioo

-------
                 Table 3    Reagent Water Detection Limit Study
Replicate
#
1
2
3
4
5
6
7
Average
Total
concentration
(measured)
mg/L
0.042
0.043
0.048
0.041
0.042
0.049
0.065
0.047
Spiked
concentration
(measured)
mg/L
0.023
0.024
0.029
0.022
0.023
0.030
0.046
0.028
S.D.=0.0084
%Recovery
92
96
116
88
92
120
184
112
The unspiked sample concentration was 0.019 mg/L. The matrix spike added was 0.025
mg/L. The %RSD for the matrix spikes was 31%.
The METHOD DETECTION LIMIT = (tn-i,99%)(Std Dev) = (3.143)(0.0084) = 0.026 mg/L.

The ZHE extract (real sample) method detection limit study is summarized in the Table 4.

                  Table 4   ZHE Extract Detection Limit Study
Replicate
#
1
2
3
4
5
6
7
Average
Total
concentration
(measured)
ug/mL
0.038
0.058
0.062
0.054
0.043
0.044
0.065
0.053
Spiked
concentration
(measured)
ug/mL
0.016
0.036
0.040
0.032
0.021
0.022
0.043
0.031
S.D.=0.010
%Recovery
64
144
160
128
84
88
172
120
The unspiked sample concentration was 0.022 mg/L. The matrix spike added was 0.025
mg/L. The %RSD for the matrix spikes was 35%.
The METHOD DETECTION LIMIT = (tn-i,99%)(Std Dev) - (3.143)(0.010) = 0.031 mg/L.

A ZHE extract from acid stabilized kiln dust sample was spiked with methanol and the
recovery calculated. The results are summarized in Table 5.
                                    n-ioi

-------
                       Table 5    Kiln Dust Matrix Spike
Spike
Added
ug/mL
1.0
Sample
Cone
ug/mL
1.6
MS
Cone.
ug/mL
2.32
MS
%Rec

72
MSD
Cone
ug/mL
2.34
MSD
%Rec

74

%RPD

1
Under normal operating conditions carryover from one sample to the next was minimal.
Normal cleaning was to rinse 3 times with 50 mL portions of reagent water. To test the
effectiveness of the rinse a system blank was distilled immediately after a 100 mg/L
sample.  This system blank had a methanol concentration near the method detection limit
(0.026 mg/L). This represents about 0.03% carryover from the previous sample.

OTHER PROTOTYPES

Further  miniaturization  has been pursued to design a  system that provides better
enrichment with  40  mL samples or permits  work with smaller sample volumes.  A
smaller glassware system (height of 500 mm) has been designed which is much easier to
handle than one that has an overall height of 1000 mm. Figures 7, 9, 10 and 11 show four
mini-systems. The condenser  height is 150 mm.  The first design uses the Peters (2)
overflow system while the others have been experiments  with other overflow designs.
When miniaturizing the following parameters must be taken into account;  water surface
tension effects, boil/reflux flow  rates and steam linear velocity.

                       32 mm
                       24 mm
                        8 mm
                                       Cooling water

                                        Cooling Coil

                                        Collection Chamber
                                          (Volume = 5 mL)
                                           Overflow tube
                                             (3 mm OD)

                                             -Stopcock
                                  Figure 7
                     Miniaturized Peter's Overflow System
                                   n-io2

-------
It is desirable to keep the boil/reflux rate high since the system will come to equilibrium
faster and thus keep the sample preparation time to a minimum. The chart below shows
that most of the methanol is distilled in the first few minutes. As the boil/reflux rate is
increased the methanol distills off in less time. This is shown in Figure 8.


"S °-7~

^ 0.6 -
-i— >
.'£. 0.5 -
Q
£Z 0.4 -
o
13 0.3 -
o
2 0.2 —
u_
0.1 -
(



I
1
\
1
1
1
*••-»
Ik
^"%;\
\%^s^_
1 I 1 1 f 1 1
D 2 4 6 8 10 12 14 1
Time (min.)


Boil/Reflu
Rate
mL/min

"- 1.5

« 3.5
• 7.4


6



X










                                     Figure 8
                     Fraction of Methanol Distilled Off vs Time

The boil/reflux rate must not be so high that  the collection chamber overflows in large
drops or waves. In extreme cases with narrow condenser necks and high steam linear
velocities the steam may actually prevent the water from returning to the flask. Thus all
condensed water will collect in the condenser  and the steam will simply blow bubbles in
it.  As the dimensions of a VOC3 are reduced  surface tension effects become more
pronounced.  The diameter of the overflow tube and it's distance to the  nearest tube wall
is important. If this distance is  too small a strong water meniscus bridges the gap and
disrupts the even flow of water (see Figure 4).

The Peters overflow  system is difficult to make in  small sizes so alternate overflow
systems were investigated.  The most successful have been the straight tube overflow,
wick and hoop systems.  High boil/reflux rates (4 to 7 mL/min) with  smooth overflow
characteristics have been sustained.  Figures  9, 10 and 11 show the alternate overflow
systems. It was necessary to have the cooling coils at a steep angle to prevent condensed
steam from dripping down the center of the condenser at high boil/reflux rates.
                                      n-103

-------
                 32 mm
                 24 mm
Overflow tube
 (10 mm OD)
                                Cooling water
                               Cooling Coil
                                 Collection Chamber
                                  (Volume = 1  ml_)
                                  Stopcock
                      Figure 9
            Straight Tube Overflow Condenser
                 32 mm
                 17 mm
Overflow tube
  (8 mm OD)
                                Cooling water
                                Cooling Coil
                                 Collection Chamber
                                   (Volume = 1 ml_)
                         Wick
                                  Stopcock
                      Figure 10
               Wick Overflow Condenser
                        n-i04

-------
                             32 mm
                             17 mm
           Overflow tube
            (8 mm OD)
                                             Cooling water


                                             Cooling Coil
                                              Collection Chamber
                                                (Volume = 1 ml_)
                                               •Stopcock
                                     Hoop
                                  Figure 11
                           Hoop Overflow Condenser
SUMMARY
The Volatile Organic Compound Concentrating Condenser (VOC3) has been used to
concentrate methanol in aqueous samples to achieve method detection limits well below
the current regulatory thresholds (0.75 and 0.25 mg/L) for zero headspace extracts.
Enrichment factors of at least 8 with recoveries in the 65-90% range have been routinely
achieved. Method detection limits in reagent water and THE extract are 0.026 and 0.031
mg/L, respectively, for 40 mL samples.  Typical distillation times are 10 minutes.  The
approximate sample load is 4 to  5 samples/hour for sample preparation per condenser
system including cleanup time.

REFERENCES

1) Report to EPA: Measurement of Polar, Water - soluble, Nonpurgeable VOCs in
Aqueous matrices by Azeotropic Distillation - Gas Chromatography / Mass Spectrometry,
Midwest Research Institute, September 30, 1989.

2) Peters, Anal. Chem., 1980, 52, 211  213, Steam Distillation Apparatus for
Concentration of Trace Water Soluble Organics.
                                    n-ios

-------
                          ADAPTATION OF SW-846 METHODOLOGY FOR THE
63                      ORGANIC ANALYSIS OF RADIOACTIVE MIXED WASTES

         W.  H. Griest, R. L. Schenley, B. A. Tomkins, J.  E. Caton, Jr., G. S. Fleming, S. H. Harmon, L. J.
         Wachter, M. D. Edwards1, and M. E. Garcia2, Organic  Chemistry Section,  Analytical Chemistry
         Division, Oak Ridge National Laboratory, Oak Ridge,* Tennessee 37831-6120

         ABSTRACT

         Modifications to SW-846 sample preparation methodology permit the organic analysis of radioactive
         mixed waste with minimum personnel radiation exposure and equipment contamination. This paper
         describes modifications to SW-846 methods 5030 and 3510-3550 for sample preparation in radiation-
         zoned facilities (hood, glove box, and  hot cell) and GC-MS analysis of the decontaminated organic
         extracts in a conventional laboratory  for volatile and semivolatile organics by methods  8240 and
         8270 (respectively).  Results will be presented from the analysis of nearly 70 nuclear waste storage
         tank liquids and 17 sludges.  Regulatory organics do not account for the organic matter  suggested
         to  be present by total organic carbon measurements.

         INTRODUCTION

         The  closure and  decommissioning of  nuclear waste storage tanks at  the  Oak Ridge  National
         Laboratory (ORNL) required the chemical, physical, and radiochemical analysis of highly radioactive
          liquids and sludges to determine their  regulatory classifications and to aid in selection of appropriate
          treatment  and disposal methods. A part of this  characterization was  the analysis of volatile and
          semivolatile organic compounds in the  liquids and  sludges.

          Approved methodologies for the determination of regulated volatile and semivolatile organic (and
          other) compounds in wastes are described in  the U.  S.  Environmental Protection Agency (EPA)
          Solid Waste Manual 846  (SW-846) (1).   However, these methods were designed for nonradioactive
          wastes, and their direct application to radioactive wastes would result in the exposure of laboratory
          personnel to high radiation fields, and  could contaminate personnel, equipment, and instruments
          with radionuclides.

          We  find  that  modifications can be  made to SW-846  methods  to limit  radiation  exposure and
          contamination in  keeping with "ALARA"  (As  Low  As Reasonably Achievable) policy, and yet
          achieve reasonable method performance.  Exposures are minimized  by traditional  radiochemical
          means of  shielding, minimizing the time of exposure to the sample,  and maximizing the distance
          from the sample that the operator conducts  the preparation.  In addition, sample amounts must be
          reduced for some preparations.

          'Now with the Tennessee Eastman Company, Kingsport, TN.
          :Now with the Environmental  Compliance Division, Oak Ridge  National Laboratory.


          "Operated by Martin Marietta Energy Systems,  Inc., under  U.S.  Department of Energy  contract
           DE-AC05-S40R21400.


                                          "The  submitted  manuscnpt has  been
                                          authored by  a contractor  of the U.S.
                                          Government  under   contract  No.   DE-
                                          ACOS-84OR21400. Accordngry.  the U.S.
                                          Government  retains  a  nonexclusive.
                                          royalty-tree license to publish or reproduce
                                          the  published  form of thra contribution, or
                                          snow others to do so. tor U.S. Government
                                          purposes."
                                                       n-io6

-------
Our approach has been to prepare the samples using modified SW-846 methods in radioactivity
zoned facilities such as radiochemical  hoods, glove boxes, or hot  cells.  The  organic extracts are
effectively decontaminated of the bulk of the radionuclides, and they can be analyzed by SW-846
GC-MS methods in a conventional laboratory.

EXPERIMENTAL

Radioactive Waste Samples and Characteristics
The nuclear waste  liquids  analyzed in this project were collected by suction from underground
storage tanks (2), some of which date  to  the 1940s.  Sludges were collected using a coring device
(2). Neither sample type was uniform in characteristics from layer to layer or tank to  tank, and
exhibited considerably diverse properties.  The gross alpha activities ranged from <1 to 8.3  x 103
becquerels/mL (Bq/mL) for liquids and to 6.5 x 10s Bq/g for  sludges.  Gross beta/gamma activities
were <20  to 3.6 x 106 Bq/mL for liquids and  up to  5.9 x 107 Bq/g  for sludges.  The major
radionuclides were 137Cs, ""Sr, and '"Co. Lesser activities of other radionuclides such as 3H and "•'U
also were present. The sludges were also  enriched in 2+4Cm, B8Pu, "'Pu, and 241Am.  The pH of the
liquids ranged from highly acidic (0.2)  to highly alkaline (12.7), and total solids ran from 0.3 to 170
mg/mL.  The dominant anions were sulfate (to 83,000 mg/L) and nitrate  (to 31,000 mg/L).   The
total organic carbon contents were high: liquids were 10 to 12,000 mg/L and sludges were 4,000 to
28,000 mg/kg.

Radiochemical Facilities
The radioactivity of the samples required  that they be prepared for organic analysis in radioactivity
zoned facilities.  Health  Physics  guidelines at ORNL (3) limit the total activity of "very  high"
radiotoxicity isotopes (such as ^Sr)  to  0.1 microcuries  (uCi) for monitored  benchtop sample
preparation, 10 uCi for radiochemical hood work, and 10 millicuries (mCi) for glove box operations.
Benchtop operations are performed in radioactive contamination-zoned laboratories with limited
 personnel access and periodic health physics monitoring for contamination.  Samples, equipment,
and staff  cannot  leave  the lab  without health physics  screening for contamination.    The
 radiochemical hood is located  in the contamination-zoned  laboratory, and consists of a normal
 stainless steel fume hood which exhausts through high efficiency particle filters and charcoal filters.
 Radioactive samples are contained in  lead "pigs" when not being processed.  The laboratory  room
 air pressure is kept at a slightly  lower level than that of surrounding halls or rooms  to prevent
 spread of any airborne contamination.  Laboratory room air exhausts through the hood.  Glove
 boxes  are located in another contamination-zoned laboratory and  they consist of a sealed stainless
 steel box which  is vented to an  air exhaust header  (also carefully filtered)  maintained at  lower
 pressure than the laboratory air.  Samples are bagged in and out  in such a manner that the  glove
box atmosphere never is in direct contact with or vents to the laboratory air. Sample manipulations
 are conducted with the protection of heavy rubber gloves protruding through one wall of the box
and below viewing windows.  Work with higher activities  (>10 mCi) must be performed in hot cells,
which have three foot thick concrete walls and remote manipulators.  Personnel using all types of
facilities wear  special coveralls  and shoes,  and carry  radiation  dosimeters  of  several  types.
 Cleanliness  and a deliberate,  unhurried, and  carefully considered work plan are essential  to
successful operations.  As noted  in the  Results and  Discussion section, sample preparation was
performed in all three types of facilities at the ORNL High Radiation Level Analytical Laboratory,
depending upon the activity of  the samples, and the operations performed.

Methods
The methodology described in this paper is based upon several SW-846 methods: 5030, 5040, and
8240 for volatile organic compounds  in aqueous  liquids; 3510 or 3550 and 8270 for semivolatile
organic compounds in aqueous liquids and sludges (respectively). These methods were approached
                                              H-107

-------
as prescribed  in  the  SW-846 manual (1),  and the  original  methods  are not  described here.
Modifications necessary for limiting radiation exposure and contamination are discussed in the text.
In some cases,  EPA Contract Laboratory Program surrogate standards or matrix spike mixtures were
used in method evaluation.  Also, a modification was made to method 8015 to permit the direct
aqueous injection GC analysis of mg/L concentrations of several alcohols and ketones.

RESULTS AND  DISCUSSION

Volatile Organics Analysis (VOA1  of  Aqueous Liquid Wastes
The approach  for the VGA entailed an off-line purge and trap (PAT) in a glove box located in a
radiochemical  laboratory.  A glove box operation was required because the whole sample container
(250 mL volume)  had to  be  handled  for the first opening for VGA. This step was followed by
thermal desorption and a  second purge and trap  and gas chromatography-mass spectrometry (GC-
MS) in a  separate, conventional GC-MS laboratory.  The procedure and apparatus are described
in detail elsewhere (4).  Briefly, 5 mL  of aqueous waste sample and CLP surrogate standards were
purged from a 40 mL VGA vial and a  specially designed P/T head into a method 624 triple sorbent
trap located outside of the glove box using method 5030 purging conditions.  We have not detected
the transfer into the trap  of any radioactivity from waste samples.  The internal standard was not
added at this step to allow the recoveries of the two P/T stages to be  differentiated. Matrix spiked
samples and blanks also were prepared with the samples.  The traps were screened by standard
smear and probe procedures for external contamination, and were then transferred to a conventional
GC-MS laboratory.  Analysis was conducted by thermally desorbing the traps  in a tube furnace at
 182°C with a helium flow  of 35 mL/min for 11 min.  The effluent was bubbled  through the Tekmar
purging vessel which contained 5 mL of laboratory distilled water and  CLP internal standards. This
is similar  to the desorption of volatile  organics sampling train traps in method 5040, except that the
internal standard  was added to the water,  and not swept into  the  triple  sorbent trap prior to
thermal desorption.  The  remainder of the analysis  was  conducted per method  8240 P/T GC-MS.

We find that this procedure performs quite well for the determination of volatile organic compounds
in radioactive aqueous liquids.  Keeping the same sample volume as stipulated in  SW-846 permits
the same reporting limits (i.e., 5-10/
-------
The method blanks were good, and showed only traces of methylene chloride and toluene (<20 pg/L
each) and acetone (<40 ^wg/L).  Other volatile organics were occasionally detected at levels of <20
fig/L. As reported elsewhere (4), the main volatile organics found in the samples were acetone (<5-
600 ng/L), methylene chloride (8-1,000 /*g/L), chloroform (3-400 /ig/L), and methyl isobutyl ketone
(< 5-3,000 ,wg/L).   Lesser  concentrations   of  benzene,  toluene,  xylenes, trichloroethene, and
tetrachloroethane also  were determined in some samples.

Major Volatile Organic Compounds in Aqueous Liquids
The VOA was supplemented by a direct aqueous injection GC procedure similar to SW-846 method
8015.  This analysis was conducted for two purposes: (a) to determine certain compounds which do
not purge and cannot be determined by the VOA, eg, methanol, and (b) to identify aqueous samples
which were  too concentrated in organic matter to P/T from a 5 mL volume.  The latter was  to
protect the  GC-MS from contamination from overloaded traps. A 1.5 mL volume of sample was
taken at the time the  P/T for VOA was conducted in the glove box, and a 3 ,«L aliquot of the
sample was  analyzed before the traps were taken to the GC-MS laboratory.  The direct aqueous
injection GC was performed using a GC located in a radioactivity contamination-zoned laboratory.
The instrument was equipped with the same column packing as the method 8240 GC-MS (i.e., 0.125
in. OD x 8  ft. stainless steel packed with 1% SP-100 on 60/80 mesh Carbopack B), and a  flame
ionization detector, but was temperature programmed differently (70°C for 2 min, then  program to
220°C at 16°C/min and hold at 220°C  for 16 min).  Three ;90% for  all
 analytes except for allyl alcohol (86%), because of peak tailing for the latter.

 Semivolatile Organics  Analysis (SVOA) of Aqueous Liquid Wastes
 This sample preparation was conducted in a radiochemical hood after the sample had been opened,
 the P/T for VOA  completed, and an analysis  of  gross alpha and beta/gamma activity had been
 conducted.  A 20 mL  volume of sample and a 40 mL  VOA vial were utilized instead of the 1  L
volume and separatory funnel stipulated by SW-846 method 3510 because the  radioactivity of the
latter was generally too great  for a hood,  and separatory funnel  extractions  in a hot cell were
considered  impractical for extensive numbers  of samples.  This reduced the reporting limits 50-
fold, to 500-2,500 uglL. The sample was spiked with CLP surrogate standards  (and matrix spikes,
when required) and three extractions were made with 5 mL of methylene chloride.  The initial pH
of the sample determined the pH of the first set of extractions. Three distinct cases were observed:
pH <2, pH ca. 6-9, and pH >10. When the pH was <2,  the sample was extracted as is to recover
an acid/neutral fraction, and then the pH  was adjusted to >10 with 1 M sodium hydroxide for
extraction of  the base fraction.  When the initial pH was  >10, the  base/neutral fraction was
extracted  first, before  pH adjustment  and recovery of the acid fraction.  For samples  with  pH 6-
9, the pH was first adjusted to  >10 and treated as noted above. The extractions were performed
by gently tumbling  the vial ca.  30 times. More vigorous  agitation  caused emulsion problems.  In
some cases, etflulsions formed in spite of the gentle agitation, and centrifugation was  required to
break the emulsion.  Other types of  complicating sample behavior observed with nuclear wastes
included  evolution  of oxides  of  nitrogen,  precipitation,  and  significant   buffering  capacity.
                                             H-109
-------
Generation of oxides of nitrogen during acidification suggests the presence of nitrite in the samples,
and  raises the possibility  of artifact formation.   In  a  few cases,  precipitates  formed  upon
acidification.   Also,  a  few  of the initially alkaline samples  required  up to ca.  7 mL of 12 M
hydrochloric acid, versus the 1-2 mL typically needed for adjustment to a pH of <2. The methylene
chloride layers were recovered with a  Pasteur pipette, and were separated from traces of water by
passing through a disposable 10 mL polypropylene syringe fitted with a 0.45 urn porosity Acrodisc
CR Teflon membrane filter. The fractions were combined, reduced to a 1 mL  volume under dry,
flowing nitrogen gas, and were then  transferred to  autosampler  vials and the CLP semivolatile
organic internal standard solution was added.

Two characterizations were conducted before the SVGA extracts  were transferred to the GC-MS
laboratory.  A GC equipped with an  autosampler and a  flame ionization detector,  located in the
contamination-zoned laboratory, was used to prescreen the samples to identify those which did not
require GC-MS analysis and also those which required dilution to  prevent overloading and organic
contamination of the GC-MS.  For the aqueous  nuclear wastes, the latter was not a problem, and
the main  use was to screen out samples.   The criteria used  here was  that if the sample did not
contain any analyte peaks (different from  those  in a blank sample) greater than  the responses of
4  mg/L (injected concentrations) of  CLP  Target  Compound List   (TCL) base/neutral  and acid
standards, TCL pesticide standards, or other selected Appendix VIII compounds (corresponding to
a  concentration in  the original aqueous  sample of 200 fig/L), then it  did  not  require  GC-MS
analysis.  In practice, less than 20% of the samples were rejected at this point.  The SVGA extract
also was sampled for gross alpha and beta/gamma activity determination. This characterization was
conducted to prevent  contamination of  the GC-MS  laboratory.   For  aqueous  samples, the
decontamination factors ranged from 2 to 4 orders of magnitude, and very  little radioactivity
typically carried over into the SVGA extract.  Typically, the samples contained  less than 0.01 uCi
of radioactivity.  In the very few cases where greater activity was observed, a 100 uL aliquot of the
SVGA extract or a 1 mL aliquot of a 1:10 dilution (the latter re-fortified with internal standard)
were sent to the  GC-MS  laboratory.   It should be  pointed out that  this  reported level of
decontamination must never be assumed. The presence of large amounts of chelators or extractants
in  wastes conceivably  could  enhance  carry-over of radionuclides.   The GC-MS analysis was
conducted as required by method 8270, except that it was not  possible for the base/neutral and acid
fractions  to be combined and  internal standard added at that point.

The modified procedure performed well in comparison with SW-846 QC Acceptance Limits for the
conventional procedures.   The recoveries  of surrogate  standards and  matrix spikes are listed  in
Tables 3  and 4.   The  average recoveries fall well  within the QC limits.   Although all of the
base/neutral  compound recoveries were used for the tabulation, some  of the acid compound data
were deleted because of possible  preparation or analysis problems.  It  is possible that the high
initial pH of some  samples caused the destruction of the phenolic surrogates and matrix spikes (a
matrix effect).  Quantitation of at least one phenol, pentachlorophenol  historically is difficult. The
method blanks were free of TCL compounds, except for traces of the  ubiquitous  phthalates, di-n-
butyl (110 figfL) and di-n-octyl (170/
-------
samples raises  the question of at lest some of these nitro-derivatives being artifactual.   Our
experiences with the SVOA of aqueous liquids are described in more detail elsewhere (5).

SVOA of Sludges
The SVOA of waste tank sludges followed methods 3550 and 8270 with the main exception being
the masses extracted.  The aliquots varied from ca. 2 to 20 g because of the limited amounts of
sample available.  The reporting limits accordingly ranged from 500 - 2,500 /
-------
SUMMARY

SW-846 sample methodology can be adapted to radiochemical  facility use for the preparation of
highly radioactive samples for VOA and SVOA.  Analyses of regulated volatile and semivolatile
organics can be conducted with minimal personnel radiation exposure and instrument or equipment
contamination and with acceptable method performance.

Improvements in  instrumental sensitivity are needed  to improve the detection limits for SVOA
where limited by sample amount.  Development of new extraction technology applicable to hot cell
or glove  box  use  with larger sample aliquots also would  improve SVOA sensitivity.  Sludges in
particular would benefit from extraction methodology including pH adjustment.
                                           n-112
-------
REFERENCES
1.      Test  Methods  for  Evaluating Solid Waste.   Volume IB, Laboratory Manual
       Physical/Chemical Methods; SW-846, Third Edition, United States Environmental
       Protection Agency,  Office of Solid Waste and Emergency Response, Washington,
       D.  C, November, 1986.

2.      J. W. Autrey, D. A. Costanzo, W. H. Griest, L. L. Kaiser, J. M. Keller, C. E. Nix,
       and B. A. Tomkins, Sampling and Analysis of the Inactive Waste Storage Tanks at
       ORNL, ORNL/RAP-53, Environmental and Health Protection Division, Oak Ridge
       National Laboratory, Oak Ridge, TN,  August, 1989.

3.      Procedures and Practices for Radiation Protection.   Health Physics  Manual,
       Appendix A-7, Oak Ridge National Laboratory, Oak Ridge TN, July 1, 1987.

4.     B.  A. Tomkins, J. E. Caton, Jr.,  M. D. Edwards, M. E. Garcia, R. L. Schenley, L.
       J. Wachter, and W. H. Griest, "Determination of Regulatory Organic Compounds
       in  Radioactive Waste Samples.   Volatile Organics in Aqueous  Liquids,"  Anal.
       Chem.. 61, 2751-2756 (1989).

5.     B.  A. Tomkins, J. E. Caton, Jr., G. S. Fleming, M. E. Garcia, S. H. Harmon, R. L.
        Schenley, C. A. Treese, and W. H. Griest, "Determination  of Regulatory Organic
        Compounds in Radioactive  Waste SAmples.  Semivolatile Organics in Aqueous
        Liquids,"  Anal. Chem.. 62, 253-257 (1990).

6.      R. L. Schenley and W. H. Griest, Investigation of the  Organic Matter in Inactive
        Nuclear Waste Tank Liquids, ORNL/ER-13, Oak Ridge National Laboratory, Oak
        Ridge, TN, in press.

7.      A. P.  Toste, T. J. Lechner-Fish,  D. J. Hendren, R. D. Scheele, and W. G.
        Richmond, "Analysis of Organics in Highly Radioactive Nuclear Wastes," J. Rad.
        Nucl. Chem..  123. 149-166 (1988).
                                            H-113
-------
        TABLE 1.   RECOVERIES OF VGA SURROGATE STANDARDS FROM
                     RADIOACTIVE AQUEOUS LIQUID WASTE
   Standard3
Toluene-d8
Bromofluorobenzene

l,2-Dichloroethane-d4
Recovery6, %
Wastes
89
59
81
(n - 65}
± 17
± 15
± 11
Blanks
91
61
86
(n
±
±
;*;
= 1}
11
12
13
QC
88
86
76
Limits
110
- 115
- 114
    a Spiked at concentration of 50
    b Average  ± standard deviations for samples and blanks, and ranges of QC Acceptance Limits.
              TABLE 2.  RECOVERIES OF VGA MATRIX SPIKES FROM
                          RADIOACTIVE AQUEOUS LIQUID WASTE
                                                        Recovery1".
    Spike3
 1,1-Dichloroethene

 Trichloroethene

 Benzene

 Toluene

 Chlorobenzene
Waste (n = 20Y
105 ± 18
87 ± 14
89 ± 15e
81 ± 20f
67 ± 13C
QC Limitsd
D - 234
71 - 157
37 - 151
47 - 150
37 - 160
3 Spiked at 50 f.ig/L Concentration.
^b Averages ± Standard deviations for waste and ranges for QC Acceptance Limits.
c n  = 20 except as noted.
d for 20 /
-------
        TABLE 3.       SVOA SURROGATE STANDARD RECOVERIES FROM
                         RADIOACTIVE AQUEOUS LIQUID WASTES
                                                        Recovery1*. %
   Standard3                     Wastes (n = 671       Blanks (n = 8)        QC Limits
Nitrobenzene-d5                      70 ±15              65 ± 18            35-114
2-Fluorobiphenyl                     64 ± 14              58 ± 13            43-116
Terphenyl-dM                         88 ±17              85 ± 18            33 - 141
Phenol-d5                            53 ± 12C             49 ±12            10 - 94
2-Fluorophenol                      48 ± 11°             42 ±  8            21   100
2,4,6-Tribromophenol                 71 ± 17°             69 ±14            10 - 123
3 Base/neutral and acid compounds spiked at 5 and 10 mg/L concentrations, respectively.
b Averages ± standard deviations for wastes and blanks and ranges for QC Acceptance Limits.
* n = 63.
               TABLE 4.    SVOA MATRIX SPIKE RECOVERIES FROM
                            RADIOACTIVE AQUEOUS LIQUID WASTE
                                                       Recovery6. %
 	Spike3	              Wastes (n = 14)              QC Limits'
 1,4-Dichlorobenzene                           50 ±10                   20 - 124
 N-nitroso-di-n-propylamine                     66 ± 10                    D   230
 1,2,4-Trichlorobenzene                         53 ± 12                   44  142
 Acenaphthene                                64 ±13                   47 - 145
 2,4-Dinitrotoluene                             71 ± 18                   39  139
 Pyrene                                       78 ± 12                   52  115
 Phenol                                       40 ± 9d                     5-112
 2-Chlorophenol                               43 ± 7d                    23 - 134
 4-Chloro-3-methylphenol                       55 ± 19d                   22  147
 4-Nitrophenol                                69 ± 31d                    D - 132
 Pentachlorophenol                            74 ± 32'                   14 - 176
 3 Base/neutral and acid compounds spiked at 5 and 10 mg/L concentrations, respectively.
 b Averages d^andard deviations for wastes and ranges for QC Acceptance Limits.
 c For 100 figjL spikes.
 d n = 10.
 en = 8.
                                          H-115
-------
          TABLE 5.     RECOVERIES OF SVGA SURROGATE STANDARDS
                         FROM RADIOACTIVE WASTE SLUDGES
   Standard2
Nitrobenzene-d;
2-Fluorobiphonyl
Terphenyl-d14
Phenol-d5
2-Fluorophenol
2,4,6-Tribromophenol

Sludges (n = 19)
18 ± 17
37 ± 19
66 ± 20
47 ± 20
35 ± 19
37 ± 18
Recovery5, %
Blank (n = 1)
33
35
99
46
39
92

QC Limit
23 120
30 - 115
18 137
24 ± 113
25 - 121
19 122
3 Base/neutral and acid compounds spiked at concentrations of 1 and 2 mg/kg (20 g sample) to 10
  and 20 mg/kg (2g sample), respectively.
b Averages ± standard deviations for sludges, one result for blank, and range for QC Acceptance
  Limits.
                                          n-116
-------
              A METHOD  TO  IMPROVE  THE  COLUMN  CLEANUP  EFFICIENCY
                                     AND
                  THROUGHPUT OF OILY WASTE  SAMPLE  EXTRACTS
                THROUGH A  NITROGEN PRESSURIZED ALUMINA COLUMN
                    (A MODIFICATION OF  SW-846  METHOD 3611).

                                     by

                     Robert Moul, Ty Lawson, Mark Dymerski
                           Laboratory  Supervisors
                  Enseco-Rocky Mountain Analytical  Laboratory
                                 INTRODUCTION

     The alumina column cleanup  technique that  is described  as  Method 3611  in
 Test Methods for Evaluating Solid Waste: Physical/Chemical Methods  (SW-846,
 3rd Ed., Vol.  IB,  1986) for the  separation of petroleum wastes  has  one
 drawback when  it is  applied to samples  that have very  high levels of non-
 target  hydrocarbons.   The flow of the column frequently becomes obstructed  by
 large molecular weight (non-gas  chromatographable)  aliphatic hydrocarbons.
 This obstruction results in very low flow through the  column and causes poor
 separation, breakthrough of non-target  aliphatic hydrocarbons,  and  sometimes
 even the complete  evaporation of the eluting solvent faster  than it drips from
 the column.  The time  required to perform this  cleanup can exceed twenty four
 hours.  A solution to  this problem  is the addition  of  nitrogen  pressure to  the
 head of the column to  maintain the  optimum flow through the  column.  By
 pressurizing the column, the time required to perform  the alumina column
 cleanup of petroleum wastes is dramatically reduced and the  cleanup efficiency
 is improved.   Also discussed will be the amount of  material  that may be loaded
 onto the column before cleanup efficiency become impaired.


                            EXPERIMENTAL SECTION

 Reagents;

     *    Methylene chloride, pesticide grade

     *    Hexanes,  pesticide grade

     *    Alumina,  Chromatographic  grade 80-325 mesh,  EmScience.  Activated at
          190 degrees  centigrade for four hours.

Apparatus;

     A  10 mm i.d. glass column with a 250 mL reservoir is attached to an
addition funnel assembly with a ball and socket joint.  The  addition funnel
assembly has a gas  inlet port through which nitrogen pressure is applied (See
Fig.  1).  This procedure employs glassware under pressure, so appropriate
safety practices should be followed.  These include the use  of  shields,
glassware that is in good condition, and regulating the pressure to less than
20 psi.
T:AT7                                 n-117
-------
Sample Description:

     A refinery  landfarm soil sample was chosen  because it  had proved
difficult to column clean during  its initial  analysis.   Also,  the sample
contained several target analytes as well  as  large  amounts  of  interfering
hydrocarbons.  A  residue determination was  performed  in order  to place a
200 mg equivalent of extractable  organics  onto the  columns.  An appropriate
amount of the sample was extracted so that  all of the experimental  cleanups
could be performed from a single  extract.   Surrogate  compounds were added to
the sample prior  to extraction to monitor  the preparation quality.   The sample
was extracted by  sonication and then exchanged to hexane using a Kuderna
Danish evaporative concentrator with a three  ball Snyder column.   Two aliquots
of the extract were cleaned with  gravity flow columns and two  aliquots were
cleaned using nitrogen pressurized columns.

     The columns  were  packed with 10 g. of alumina and pre-eluted  with
hexane.  The aliquots were applied to the  heads  of  the  columns.   The
base/neutral aliphatic fractions were eluted with 13 ml of  hexane and then
discarded.  The  base/neutral aromatic fractions  were eluted  with  100 ml of
methylene chloride.

     All of the  elution times were recorded.  The methylene  chloride fractions
were then concentrated and the target compounds  and surrogates were analyzed
according to SW-846 Method 8270.  An indication  of  remaining aliphatics was
shown by quantitating mass 57 for the entire chromatogram.


                           RESULTS AND DISCUSSION

     The elution  times of the two pressurized columns were  dramatically lower
than for the two  gravity columns, as shown  in Table 1.   The  total  elution time
was reduced from  more than 24 hours to less than two hours.  This is a
significant reduction of sample cleanup time and can markedly  increase the
laboratory's throughput.  The flow rate of 2 mL/min. that is specified in
Method 3611 can be achieved by regulating the pressure  at the  head  of the
column.

     Table 2 contains the surrogate and target compound results.  Analyte
recoveries for pressurized column cleaned extracts were essentially identical
to the recoveries obtained from gravity columns.  The surrogates  were
recovered within acceptable limits in all cases.  The total  area  for the mass
chromatogram of mass 57 shows a reduction of an  order of magnitude  in  the
amount of aliphatic hydrocarbon interference in  the pressure cleaned extracts
(see fig.  2).

     Some preliminary work has been performed to determine the amount  of
extractable material  that may be placed on the column before the  cleanup
efficiency becomes impaired.   Initial  results show that  the  column  may  be
loaded  at  up  to  two or three times the  amount recommended in Method  3611.
However,  difficulties  in  the extraction and partition processes may  prevent
the use  of these  larger sample amounts.
                                     n-ii8
-------
                                 CONCLUSION

     The environmental  analytical  market is requiring quicker turnaround times
and lower limits of detection.   These two requirements are frequently at odds.
But,  with the utilization of this  pressurized column technique,  the advantages
of column cleaning the  extracts of samples from difficult matrices  can be
enjoyed while still maintaining high laboratory productivity.

     This technique can also be expanded to include other preparative column
cleanup methods that are plagued by slow elution times such as SW-846 Method
3630, the silica gel column cleanup for polycyclic aromatic hydrocarbons.
                                     n-H9
-------
SoIvent
Addition
Funnel
     Ball and  socke
     joint
                               Teflon  threaded plug for
                               extract  introduction
                                    Inlet  for  regulated nitrao
                              10  mm i.d. X 300 mm column with
                              250  mL resevo ir
         PRESSURE COLUMN  ASSEMBLY
                Fig.  1
                   n-i20
-------
       MASS CIIROIWTMP.,11
       82'18'88 22.31iGO
                                                                SCflllS    I  10 2758
                       HI FRflCT.  502/1.DHL REF.  COL. aHO. GRftUlTY  II
WSS CMPOIIntOGRftU
Wxia'DG  QilBiQO
                Bll FRACT. COL. CLIE.  50Z'I.OIL PRESSURE II
139.3-j
                                                                                                          100008.         I06.2-,
                                                                                                           37.017
                                                                                                         i  0.500
                                                                                                                                                                          'JvXU.'d^J^uJ^^

                                                                                       i-X-A-
                                                                                                                                                                                                     ss.
                                                                                                    37.017
                                                                                                   t  8.588
         HISS ClKOMflTOCR«H
         Olvisxea 23i2iioo
                                                                  SCrtlS    1 TO 2750
                         PI FRflCT. COL. CUD. se/M.m. CRWITY 12
MASS CHROHflTOCRfUl
l)2'!9/88  1100:00
                                                                                                                                                                                         SCflllS    I TO 2758
                                                                                                                                                 I FRflCT. 50Z/l.etL COL. CLIO. PRESSURE 12
  I07.5-,
                                                                                                             57.817
                                                                                                           t  a.-ma
                                                                                                              SCftll
                                                                                                              inn:
                jUU
                6:15
luuu
12:30
                                                                                                                                                                                                     UjJJJjj-
                                                                                                    37.817
                                                                                                  i  6. M8
20UG
25: DO
SCfKI
TUE
                                                                                                           Fig.   2
-------
   TABLE 1

ELUTION TIMES
 in minutes
SAMPLE
PRESS.
#1
PRESS.
#2
GRAV.
#1
GRAV.
#2
HEXANE
PRE-ELUTION
2.5
3.0
20.0
15.0
EXTRACT HEXANE METHYLENE CHLORIDE
APPLICATION ELUTION ELUTION
1.5
1.5
65.0
57.0
8.0
10.0
245.0
228.0
90
95
18
HOURS
18
HOURS
TOTAL
TIME
104
113
24
HOURS
24
HOURS
    n-i22
-------
                                   TABLE 2

                      COMPARISON OF ANALYTICAL RESULTS
ANALYTE              GRAVITY #1     GRAVITY #2

Naphthalene            100,000         99,000
1-Methylnaphthalene    290,000        270,000
Phenanthrene           170,000        160,000
Anthracene               7,900          8,500
Fluoranthene             5,400          6,900
Pyrene                  39,000         36,000
Benzo(A)anthracene      25,000         24,000
Chrysene                37,000         33,000
Benzo(B)fluoranthene     9,900          9,200
Benzo(A)pyrene          11,000         11,000
Dibenz(A,H)anthracene    3,700          3,900

Analyte results in ug/kg
                               PRESSURE II

                                  100,000
                                  310,000
                                  180,000
                                    7,
                                    5,
                       ,400
                       ,500
                    47,000
                    26,000
                    36,000
                     9,500
                    11,000
                     3,500
                PRESSURE#2

                  110,000
                  310,000
                  170,000
                    8,300
                    5,300
                   47,000
                   25,000
                   37,000
                    9,800
                   11,000
                    4,200
SURROGATE

DS-Nitrobenzene
2-Fluorobiphenyl
D14-Terphenyl
GRAVITY fl

    77%
    83%
    78%
GRAVITY #2

    72%
    73%
    78%
PRESSURE #1

     76%
     91%
     89%
PRESSURES

    85%
    96%
    86%
COMPOUND NAME
GRAVITY #1
Total Area Mass 57   14,000,000
GRAVITY #2
               26,000,000
PRESSURE #1

2,800,000
PRESSURES

5,500,000
                                     n-i23
-------
67                            A NEW AND IMPROVED TECHNIQUE
            FOR SPECIATION AND QUANTITATION OF AROCLORS IN HAZARDOUS WASTES

         William M.  Draper,  Donald Wijekoon, Hazardous Materials  Laboratory,  State
         of California, Department of Health Services, 2151 Berkeley Way,  Berkeley,
         CA 94704

         ABSTRACT

         Skilled  chemists  can  usually  recognize  Aroclor  patterns  on visible
         inspection of packed column  chromatograms.   Identification is  difficult
         or impossible, however,   when two or more Aroclors are present, or the PCB
         residue  has been  modified  by  weathering,  metabolism  or  treatment.
         Quantitation also  presents  problems  for packed  column  GC  methods.   In
         particular, integrating  the total area  under  the pattern is  error prone
         because of interfering compounds in sample extracts.

         In  an effort to  improve  PCB  analysis  in support of  hazardous  waste
         regulations, our laboratory has  investigated high resolution GC and new
         approaches  to  interpreting high resolution   GC  data.   One  approach,
         measurement of selected  chlorobiphenyl  congeners,  has proven useful for
         reducing  data  from  a  60  meter DB-5  column.    Only  12 chlorobiphenyl
         congeners  must be determined,  International Union of  Pure  and Applied
         Chemistry  (IUPAC) Nos.  15,  18,  31,  87,  105, 110,  118, 170, 180,  183, 203
         and 206,  and from these  marker compounds Aroclors  1016, 1254,   1260 and
         1268  are  estimated.   The remaining regulated Aroclor mixtures  are also
         measured by this  technique  making it suitable for enforcement  of existing
         regulations.

         Our laboratory determines an  additional  34 chlorobiphenyl congeners which
         together   are   the  major   constituents   of  the   commercial  Aroclor
         formulations,  as well  as the predominant congeners in the environment.
         The sum of  PCB congener  concentrations,  "s-PCB",  is generally  about 70%
         of the total Aroclor content,  and may provide  information needed in the
         future for measuring residues after treatment.

         The use  of capillary column separations  and data  reduction procedures
         described has a number  of advantages over packed  column  GC procedures: 1)
         data  interpretation does  not  require  analyst  judgement  and can  be
         automated;  2) the techique makes  full  use  of  the  separation power of
         capillary columns to minimize interference by pesticides and other sample
         components; and  3)  the  technique  accurately  measures PCBs in samples with
         more than one Aroclor.

         The application of  this technique to the  determination of Aroclors in auto
         shredder waste  is  described and  results  compared to SW-846 method  8080.
                                           n-124
-------
INTRODUCTION

Much has been written about the problems of identification and measurement
of Arocior mixtures (1,2).   Packed column gas  chromatography (GC) with an
electron  capture  detector  (BCD)  has  been   the  preferred  analytical
technique for PCB analysis  for over 25  years.   Historically, quarter inch
diameter packed columns with chromatographic efficiencies of about 3,500
theoretical  plates  were  operated under  isothermal  conditions.    The
detectors with either a tritium or nickel source were readily fouled due
to column bleed,  and even so-called linearized detectors had linear ranges
of  less  than  two  decades.    This nonlinearity  frequently  necessitated
sample  dilution and  repeat  analysis  for  accurate  quantitation.   The
Arocior pattern was typically confirmed by chromatography on a second GC
column requiring further analysis time.  A major problem discovered early
on  was  that  of   interferences  including  pesticides,   i.e.,  DDT  (3),
phthalates,  and   other  common   environmental  contaminants.     These
interference problems were  usually solved by adsorption  chromatography on
silica (3), Florisil (4) or alumina (5).

The limitations of packed column techniques are most apparent when samples
contain more than one Arocior, or the Arocior  pattern has been altered by
environmental  weathering, metabolism  or treatment.  The most demanding
samples are usually referred  to  a senior chemist or  supervisor,  because
of the need for a  trained eye to properly "judge"  the  sample  contents.

Today,  capillary  GC is  a mature technology due  to  advancements  in
microprocessor   controls,   splitless   and   cold  on-column   injection
techniques,  and  computerized data  acquisition  and  processing.    The
modulated pulsed frequency ECDs have greatly extended the linear dynamic
range.   Fused  silica capillary  columns with  various bonded  and coated
stationary phases  are available  commercially,  and  these columns  deliver
efficiencies of 3,000   5,000  plates/meter.  Since  most  public health and
hazardous waste  testing laboratories  now have capillary instruments,  a
need has arisen to  develop appropriate tools for data interpretation.

Over  the  past  three years  our  laboratory has been  determining  PCBs  in
sportfish and  other marine  organisms   in order to  estimate  human health
risks to consumers  (6).  The  PCB residues in these organisms  may differ
from single Arocior standards due  to:  1)  weathering  in the water column
and marine sediments; 2) metabolism at various levels in the food chain;
and 3) the occurrance of more  than one Arocior mixture as  well as other
electron-capturing, halogenated pesticides and metabolites.

We have  found  that  the determination of Aroclors is   possible  in such
samples, but only  using capillary GC.   When  using GC  columns with over
200,000 plates, however, data reduction is  extremely important as it is
not uncommon to have more than 100 peaks in a chromatogram.  In the marine
studies seven  chlorobiphenyl markers were used to  estimate Arocior 1254
and Arocior 1260 (7), Aroclors for which carcinogenicity potency factors
have been estimated.  In further developing this  technique  we find that
                                   H-125
-------
all of the  regulated Aroclors can be determined in hazardous wastes using
data on 12  individual  isomers.

The  purpose  of  this  paper   is  to  describe  the  identification  and
quantitation of Aroclors using selected  congener  data.   The analysis is
simplified because all  estimates  are based on the  determination of 12
chlorobiphenyl  congeners.   Because  the  procedure  is  simple  and very
specific  it is  rugged,  accomodates  a range of operator  skills, and is
amenable'to automated data processing.   Application of the method to the
analysis of auto  shredder waste, a matrix which has proved difficult to
analyze by packed column procedures  (SW-846 method 8080)  is described.


METHODS AND MATERIALS

Chemicals.   Chlorobiphenyl isomer mixtures were obtained from the National
Research Council Canada (NRCC), Marine Analytical  Chemistry  Standards
Program  (Halifax, Nova  Scotia)    Individual  PCB  congeners  also  are
available  commercially  from several  suppliers.   Aroclor  standards were
provided by the  U. S  Environmental Protection Agency (Research Triangle
Park,  NC)     Aroclor  formulations  were  diluted  to  100  ng/mL  and
chlorobiphenyl isomers standards were  diluted 100-fold using isooctane in
both  cases.

Nomenclature.    The  International  Union  of Pure   and  Applied  Chemistry
 (IUPAC)  numbering system  developed by  Ballschmiter  and  Zell (8)  and
designating all  209  polychlorinated biphenyl isomers is used throughout
this  report.   The sum of all  identified chlorinated biphenyl isomers is
denoted s-PCB.   Similarly,  the sum of Aroclor concentrations  is referred
to as s-Aroclor.   To facilitate  the  interpretation  and  comparison of
congener  patterns,  PCB data is   normalized by  dividing the individual
isomer concentrations by s-PCB.

Quantitative Analysis  of  Chlorobiphenyls.   Details of the  determination
of PCB congeners are described elsewhere  (7)    Briefly,  compounds were
measured by gas-liquid chromatography on  an  instrument equipped with a  Ni
electron capture detector,  an  autosampler/injector and a chromatography
digitizer/computerized data system.  Purged splitless sample  introduction,
a  60  m X 0.32 mm 0.25 urn DB-5 capillary column (J  & W Scientific, Folsom,
CA)  and a  103  min oven temperature program  were used.

Extraction and Sample Cleanup.  Auto shredder waste samples were extracted
using the  EPA  slurry extraction procedure.  Air dried samples (ca. 500  g)
were  extracted with 3 X 2 L of hexane/acetone  on  a platform shaker.  The
extracts were  combined, exchanged to hexane, cleaned by partitioning with
sulfuric acid,  and finally  exchanged to  isooctane.

RESULTS AND DISCUSSION

Determination  of Chlorobiphenyl Congeners.   The  analysis  of individual
 chlorobiphenyl   isomers,   so-called   PCB  congeners,   by   gas-liquid
                                   H-126
-------
chromatography requires not only high resolution, but also a high degree
of reproducibility.    The  cleanup of  sample extracts  by partitioning,
treatment  with  acid,   and adsorption  chromatography  eliminates  many
interferences and improves  specificity.  However, congener identification
is based primarily on gas chromatography retention times  (tR) .

Much of the early work on capillary chromatography of Aroclors relied on
glass  columns  coated  with  SE-54,   a  5%   phenyl-95%  methyl  silicone
stationary phase (8,9).  Today fused silica columns  and bonded phases have
largely  replaced  coated  glass  capillary  columns  because  of  their
durability.   The  most  recent  PCB studies  (1,10) often  use  DB-5 columns
which have selectivity and retention charactistics equivalent to SE- 54.

The 60 m DB-5 widebore column (0.32 mm i.d., 0.25 micron film thickness)
used  in  this study  is very  efficient  with 3,500  theoretical  plates/m
according to manufacturer specifications.   A number of suppliers provide
bonded phase capillaries with  efficiencies  in the 3,000 to 5,000 plates/m
range, and comparable results are expected with any 5% phenyl-95% methyl
silicone  column with  over 200,000  theoretical  plates  including  HP-5,
RSL-200,  SPB-5 columns  and  others.  The highest efficiencies are obtained
with narrow  bore  columns  (0.2 mm i.d.), thinner stationary phases,  and
efficiency also is proportional  the  square  root  of the  column length.

The  tR precisions are  excellent with modern  gas  chromatographs.    For
example,  the standard deviation of the tR for  PCB-153 was  less than one sec
during a 45 h period  (6 replicates)(7).  With this precision the retention
time windows  for  each  congener are narrow,  ca.  +/- 1.2  sec  in a 103 min
chromatogram, and the ability to distinquish isomers is improved.

Using the  60  m  DB-5  column 46 of the polychlorinated biphenyl congeners
in the NRCC standard mixtures were resolved -- their tR were between 17 and
85 min (Table 1).   PCB-159/182/187 and PCB-171/202  which are also present
in the mixed standards  are not adequately separated with  this system.  The
ECD  response factors generally  increase  with  chlorine content  and the
responses are linear over a broad range (7).

Chlorobiphenyl  Content  of  Aroclor  Reference   Materials.     The  46
chlorobiphenyls  determined in  this  study  accounted for  a very  large
proportion of each of the regulated Aroclor  formulations  (Table 2.).  The
only  exception  was  the least  chlorinated mixture, Aroclor  1221,  where
2,2',5-trichlorobiphenyl (PCB-18), 4,4'-dichlorobiphenyl  (PCB-  15),  and
2,4',5-trichlorobiphenyl (PCB-31), were the only identified components.
For the most  widely used Aroclors, 1242,  1254,  and 1260,  between 73 and
85% of the mixtures were accounted for.   Thus,  for most of the regulated
Aroclors, s-PCB determined by GC-ECD with this  group of  standards  is a
good  approximation  of  the  Aroclor  content.   s-PCB,  of  course,  varies
depending on congeners  determined and for s-PCB  to approximate the Aroclor
content,  the most abundant congeners must be among the target compounds.
                                   n-127
-------
Chromatograms for Aroclors  1254,  1260,  1262 and 1268 had peaks  with the
same tR as PCB-159/182/187 and Aroclors 1260, 1262 and 1268 also contained
the  PCB-172/202  peak,  but  these isomers  were  not quantitated.    The
compound  coeluting  with  PCB-77  both  in  the  Aroclor  formulations  and
environmental  samples  has  been  unequivocally  identified  as  PCB-110.
Accordingly, the sample component is  quantified as  PCB-77 and corrected
for the differential in published detector  response  factors  (7).

Determination of Aroclors Using Selected Congeners  as Markers.   Without
question  capillary  GC  provides  a  great  deal  more  information  than
packed-column  GC.     The  problem  with  identification  of  Aroclors by
capillary GC is one of data  reduction.   As  with  any method for estimating
Aroclors, a number  of assumptions must be  made.   The present  procedure
assumes that the composition of each Aroclor does  not vary from batch to
batch.   In  other words,  all  lots have  the same congener distribution.
Monsanto Company was the sole manufacturer  of PCBs in North  America.  We
have not analyzed commercial  PCB  mixtures  manufactured or used  on  other
continents  and this  assumption may not be  universally applicable.   The
second assumption is  that the  congener  distribution does not  change  due
to  selective weathering.   In our experience in  the  analysis of  PCBs in
marine organisms  from California coastal  waters (specifically,  mussels
from inner harbor areas),  both assumptions  are realistic.  With hazardous
waste  matrices  where Aroclors  are often  at high concentration  (e.g.,
transformer oils) little degradation of the original Aroclor mixtures is
expected.

Selection of  Aroclor  Markers.    The  selection  of  marker  congeners is
straighforward.    Chlorobiphenyl  markers   should be  abundant  in  one
regulated Aroclor, but not another.  The more prominent the  congener  is,
the  lower the  detection limit for the Aroclor mixture  will  be.  In  the
case  of  Aroclor  1268 there  are  a number of  unique and  abundant  PCB
isomers.   In particular,  PCB-203, -208, -206 and -209  are  abundant  and
relatively  unique   to  Aroclor  1268   (Table   2)     For   this   study,
2 , 2 ' , 3 , 4 ,4' , 5 , 5 ' , 6  octach1orobipheny 1    (PCB-203)     and
2 , 2' , 3 , 3' ,4,4' ,5 , 5' ,6-nonachlorobiphenyl    (PCB-206)   were    selected.
Aroclors 1254 and 1260 were industrially important in the U.  S.  (11)  and
are  major  environmental   contaminants   and residues  in biota.     The
pentachlorobiphenyls  (PCB-87,   110,  118,  and   105)  are associated with
Aroclor 1254 and  the heptachloro compounds  (PCB-183,   180,  and   170)  are
abundant in Aroclors 1260 and 1262 (6)(Table 2)

Distinguishing the  least  chlorinated  Aroclors  (e.g.,  1221,  1232,  1016,
1242, and 1248)  is  the most  difficult because the  most abundant  congeners
are  the  same  in  each mixture  (Figure  1),  i.e.,  4,4'  dichlorobiphenyl
(PCB-15),   and  two   trichlorobiphenyls   (PCB-18  and   31)  are   major
constituents of  each.   Therefore, these   three  compounds  are  used as
surrogates for all of the low-chlorine-content  Aroclors which  cannot be
distinguished.

Composition  Factors.   Composition factors  define the  [Aroclor]:[marker
chlorobiphenyl]   ratio   and  are   essential  for   estimating   Aroclor
                                   n-i28
-------
concentrations.  Composition  factors  are  equivalent to 100/wt % for any
given congener-Aroclor combination.  Using the formulation data in Table
2, Aroclor 1016  composition factors of 100/15 =6.7,  100/43 =2.3, and
100/11  =  9.1  are  estimated  for  PCB-18,  -15  and   31,  respectively.
Composition factors for each of the four measured Aroclors were calculated
similarly and appear in Table 1.

Calculations.    Aroclor  concentrations are  estimated  in  a  three  step
process.   First,  the  concentrations  of  the 12  marker  compounds  are
determined in  the  sample.   Second, the concentration  of  each marker is
multiplied by the appropriate composition  factor  to  give a single Aroclor
estimate.  Third,  the mean Aroclor estimates are calculated  and reported.

If  the  congener  pattern in  the sample  matches the  Aroclor  reference
material, the relative standard deviation  (RSD) of Aroclor estimates will
be  low,  generally  well below 50%.   A graphic example  of  this  is again
provided by our  recent studies of marine  contamination  (7).   In Mussel
Watch samples  from throughout  California coastal  waters,  Aroclor  1254
congener profiles were very similar  to the U.  S. EPA reference material
with RSDs  of  28 +/-  15%.   Even  when there are multiple,  disperse PCB
sources,   and   residues    have   been   subjected   to   volatilization,
photochemistry,   biological  and   nonbiological   tranformations,   and
partitioning in the environmental  compartments,  the original pattern of
congeners  is  reasonably well retained.   Clearly,  the  Aroclor  mixtures
released must have been similar in composition.

Quantitation of the Other Aroclors.  The regulation of PCBs in wastes is
based on the sum of all commercial Aroclors. To enforce these regulations
all Aroclors must be detected and quantified reliably  Analytical methods
which overestimate (positive bias) the Aroclor residues are particularly
undesirable because enforcement  actions must  withstand legal challenge.
As the present procedure defines and measures  all PCB residues as the sum
of  4 Aroclor  mixtures, it  is  important to establish  how  the  technique
quantifies the other 5 regulated Aroclors.

The accuracy  of  the procedure  is acceptable  for  all of  the  regulated
Aroclor mixtures except Aroclor 1221 (Table 3).  Recoveries were between
61 and 135% for all PCB formulations with greater than 30% Cl by weight.
For example,  100 mg of Aroclor 1248 is quantified as  90 mg s-Aroclor which
is the sum of 55 mg of Aroclor 1016, 43 mg of Aroclor 1254 and 4.7 mg of
Aroclor 1260.   The  major components  of Aroclor  1221 are  one-  and  two-
chlorine  compounds  with  very  low  electron-capture  detector  response
factors.  The NRCC mixtures contain only  one  dichloro compound,  PCB-15,
and no  monochlorobiphenyls  resulting  in  a  sizable  underestimation  of
Aroclor  1221  by  either   s-PCB  (Table  2)  or  s-Aroclor  (Table  3).
Fortunately,  Aroclor  1221  is  one  of  the least  toxic and persistent
Aroclors,  was manufactured in comparatively low volume (11), and is rarely
detected in hazardous wastes.    In summary, the 12 chlorobiphenyl markers
give good estimates of all  of the important,  regulated Aroclors without
any judgement  on  the part of the  analyst as to the Aroclor type.  Because
the marker chlorobiphenyls  are not absolutely  specific, there may be some
                                   n-i29
-------
overestimation,  for example, of Aroclor  1254 and 1260.  Simple  formulas
are used to correct these overestimates  (7)

If the  chemist  has additional  information about the  sample,  i.e.,  the
sample  originates  from  electical equipment  known  to contain  a  given
Aroclor,  the  selected  congener  technique  is accurate  for all of  the
commercial  Aroclors.     The appropriate  composition  factors  must be
calculated using the formulation data in Table 2.  PCB-18,  -15,  and  31
are markers for  the low-chlorine-content  PCBs,  i.e.,  those  with 21 to 48%
Cl by weight.  Similarly,  PCB-183,  180 and  170  are  used as Aroclor 1262
markers with the appropriate composition factors.

PCBs in Auto  Shredder  Wastes.   Auto shredder  waste  is the residue  from
metal  recovery  operations  which  use discarded  automobiles as  a  metal
source.   Auto shredder wastes  are  non-RCRA hazardous  wastes  which  are
subject to  treatment  by  chemical  stabilization in  California prior to
disposal because of their heavy metal content  (e.g., Zn, Pb, Cd,  Cr,  Cu,
Hg, Ni)   The  Hazardous Materials  Laboratory has  had  extensive  experience
in the analysis  of autoshredder wastes,  not only for regulated inorganic
elements,   but also  for  PCB  contamination which  the  laboratory  first
reported.    Auto shredder  waste  extracts  were selected for  study  here
because they  contain more  than  one  Aroclor  type  and  are difficult to
analyze by packed column methods.

Two  auto  shredder waste  samples  were analyzed using  the conventional
packed  column procedure,  SW-846  Method 8080.   A  highly  trained  and
experienced chemist examining the packed column  chromatograms  identified
two  Aroclors,  1016 and 1260,  in  shredder waste  extracts.   Using  the
standard quantitation  technique (total area),  sample No.   1827 was  found
to contain 43 mg  Aroclor 1016/kg  and 14 mg  Aroclor  1260/kg (dry weight
basis)   A second  auto  shredder waste sample,  No.  1831, contained  73 mg
Aroclor 1016/kg and 20 mg Aroclor 1260/kg.

Almost all of the PCB congeners were detected  in capillary chromatograms
(Figure 2)    In sample  No   1827 32  chlorobiphenyls  were detected and in
No.  1831   34  of  the   congeners  were present  (Table  4)    The twelve
chlorobiphenyl markers  indicated  the presence of three Aroclors,  1016,
1254 and 1260 (Table 4)   Although small  amounts of Aroclor 1268  markers
were measured, Aroclor 1268 was not included in  s-Aroclor  because it did
not match  the reference material,  e.g.,  the  dispersion in Aroclor  1268
estimates  was too  great  with RSDs of 69  and  72%.   In contast the  three
lower chlorinated Aroclors detected in shredder waste extracts matched the
reference  materials well with RSDs between 11  and 36%.

s-Aroclor  concentrations estimated by either 8080 or the congener-  based
procedure  were similar, i.e.,  57 vs 99 mg/kg and 93 vs 145 mg/kg for No.
1827  and  No.  1831,  respectively    On  average  the  relative   percent
difference for  the two methods  is  less  than  50%.    However,   Aroclor
identifications   were   distinctly  different.    Aroclor   1254  was  not
detectable by Method  8080  at  the detection  limit  of 2 mg/kg,  yet  the
                                  n-iso
-------
concentrations of PCS-87,  110,  118, and  105 clearly indicated Aroclor
1254 concentrations of 35 (No. 1827) and 52 mg/kg (No. 1831).

Method  accuracy was  tested  by  reconstructing  the  expected  congener
profiles  (using  Table 2  data)  and  comparing  the predictions  with the
experimental data.   If sample  No.  1827  contained 43 mg/kg of Aroclor 1016
and  14  mg/kg Aroclor 1260,  a congener profile  plotted in  Figure  3  is
expected.    Similarly,   the  congener   profile   expected   for  a  sample
containing Aroclor 1016  (43 mg/kg),  Aroclor  1254 (35  mg/kg)  and Aroclor
1260 (21 mg/kg) are obtained.   The congener profiles clearly demonstrate
the presence of Aroclor  1254  in the sample,  a  fact that  is  not evident
from the packed column data.

This example demonstrates the utility of selected congener data in Aroclor
speciation.  The  capillary chromatograms actually provide a great deal
more information in the  form  of many additional Aroclor 1254-associated
peaks.   However, data reduction through the use of selected congeners is
a much more efficient means for evaluating the raw data.

CONCLUSION

All of the regulated Aroclor formulations can be determined based on the
concentrations of  12  chlorinated biphenyl congeners.    Three congeners,
PCB-18,   15, and -31,  are markers  for Aroclors with low chlorine content,
pentachlorobiphenyls are  markers  for Aroclor 1254,  heptachlorobiphenyls
are markers for Aroclor 1260,  and Aroclor 1268 is estimated from PCB-203
and -206.

The  advantages  of  the  use  of  high  resolution  chromatography  and
consideration of only  selected congeners are the following:  1) the chemist
is relieved from "eyeballing" the chromatogram and attempting  to interpret
patterns;  2)  the  technique makes  full use  of the separation  power  of
capillary chromatography and minimizes interferences caused  by chlorinated
pesticides,  pesticide  metabolites,  chlorinated  terphenyls,  phthalates,
sulfur  allotropes,  and  other  electron-capturing coextractives;  3)  the
technique readily accomodates  samples containing more than  one commercial
Aroclor;  4) multipoint  calibration with each  congener  is less  time
consuming than multipoint calibration with all 9 regulated Aroclors; and
5) the technique lends  itself well  to  automated interpretation and data
processing.

Finally,  with   little  additional  effort  the  number  of   PCB  congeners
determined can be increased  to over 45 using the NRCC mixtures, or similar
mixtures  obtained  commercially.   The  determination of  these  compounds
provides a second measure of PCB content,  s-PCB, which may be useful for
determination  of  PCB  residues  after treatment.   Some scientists  feel
strongly  that   the  only  valid means   for  determining PCBs in  highly
weathered or treated samples is by congener measurement.
                                  n-isi
-------
ACKNOWLEDGEMENTS

We gratefully acknowledge the staff of the Hazardous Materials  Laboratory
for their cooperation  and  assistance.    J   Chang,  Y.  Chew, and  S.  Gill
assisted with the analysis of  auto shredder waste samples.  H. Okamoto,
B. Simmons and R. Stephens provided valuable comments on  the manuscript.
This study was supported in part through California Department of Health
Services Contract 87-91874 and by the Superfund  Program Project  No. ES
04705 from the National Institute  of Environmental Health  Sciences of the
National Institutes of Health.

LITERATURE CITED

1.  Pellizzari,  E.  D.;  Moseley,  M.  A.;  Cooper,  S.  D. J.  Chromatog.   1985,
334, 277

2.  Alford-Stevens,  A.  L. Environ. Sci. Technol.  1986,  .20,  1194

3.  Armour,  J A.; Burke, J  A. J.  Assoc.  Off. Anal.  Chem.  1970, 553,  761.


4    Lopez-Avila, V.; Yeager,  S.   Proceedings  of  the  Fifth Annual  Waste
Testing and Quality Assurance Symposium 1989, Vol. II, pp.  102-114,  U. S.
Environmental Protection Agency, Washington, DC.

5.  Satsmadjis, J ;  Georgakopoulos-Gregoriades,  E.; Voutsinou- Taliadouri,
F  J. of Chromatogr. 1988,  437. 254.

6     Draper,  W   M.     External  Quality  Control  Program:  Chemical
Contamination  of Marine Fish,  Final  Report,  California  Department of
Health  Services  Contract   No.   87-91874,  California  Public   Health
Foundation,  Berkeley,  CA (1990)

7   Draper,  W. M.;  Koszdin,  S  , submitted for publication,  1990.

8.  Ballschmiter, K.;  Zell,  M. Fresenium Z. Anal. Chem.  1980,  302,  20.

9   Cooper,  S D.; Moseley, M. A.;  Pellizzari, E.  D.  Anal.  Chem. 1985,  57,
2469.

10   Norstrom,  R.  J ;   Simon,  M. ; Muir, D.  C. G.;  Schweinsburg, R. E.
Environ. Sci. Technol.  1988,  22, 1063.

11.  Nisbet, I  C.  T ;  Sarofim, A. F  Environ.  Health Perspect. 1972, 1,
21.
                                   n-132
-------
Table 1.  Typical retention times for chlorinated biphenyl congeners and composition factors for marker
chlorobiphenyls.
Chlorobiphenyl
Structure
Chlorobiphenyl
   IUPAC No.a
2,2',5-Tri
4,4'-Di
2,2',6,61-Tetra
2,4',5-Tri
2,2l,5,5'-Tetra
2,2',4,5'-Tetra
2,2',3,5'-Tetra
2,2',3,3'-Tetra
2,2',4,5',6-Penta
2,3',4,5',6-Penta
2,3,4,4'-Tetra
2,2',4,5,5'-Penta
2,2',3,4,5-Penta
2,2',3,4,5'-Penta
3,3',4,4'-Tetra
2,2',4,41,5,6'-Hexa
2,2',3,5,5',6-Hexa
2,3',4,4',5-Penta
2,2',3,4,5,6'-Hexa
2,3,4,4',5-Penta
2,21,4,4',5,5'-Hexa
2,3,3',4,4'-Penta
2,2',3,4,5,5'-Hexa
2,2',3,4,4',5-Hexa
2,2l,3,4,41,5'-Hexa
2,2',3,3',4,5-Hexa
2,2',3,4,4',5',6-Hepta
2,21,3,3',4,41-Hexa
2,2',3,4,5,5',6-Hepta
2,3,3',4,4',5-Hexa
2,2',3,3',4,5,6-Hepta
2,2',3,3l,4,5',6,6'-0cta
2,2',3,4,4',5,5'-Hepta
2,3,3',4,4',5',6-Hepta
2,21,3,31,4,4',5-Hepta
2,2l,3,3',4',5,51,6-0cta
2,2',3,4,4',5,5',6-Octa
2,2',3,3',4,4',5',6-Octa
2,3,3',4,4',5,5'-Hepta
2,2',3,3',4,5/51,6,6'-Nona
2,2',3,3',4,4',5,6-Octa
2,2',3,3',4,4',5,6,61-Nona
2,2',3,3',4,4',5,5l-0cta
2,3,3',4,4',5,5',6-Octa
2,2',3,3',4,4',5,5',6-Nona
Decachlorobiphenyl
 18
 15
 54
 31
 52
 49
 44
 40
103
121
 60
101
 86
 87
 77
154
151
118
143
114
153
105
141
137
138
129b
183
128
185b
156
173
200
180
191
170
201
203
196
189
208
195
207
194
205
206
209
                                           (min)
                       RRTC
 Composition Factor
(100/wt % in Aroclor)
17.87
18.06
20.59
21.73
25.80
26.38
28.64
31.74
31.84
35.65
37.53
38.57
41.58
42.13
43.36
43.64
45.07
46.78
47.65
48.09
49.61
49.92
51.07
51.86
52.78
53.71
55.56
56.11
56.90
59.49
60.16
60.44
62.17
63.25
67.09
68.89
69.86
69.99
72.93
76.03
76.18
77.71
78.88
79.34
82.06
84.69
0.421
0.423
0.486
0.513
0.609
0.622
0.676
0.749
0.751
0.841
0.885
0.910
0.981
0.994
1.023
1.029
1.063
1.103
1.124
1.134
1.170
1.177
1.205
1.223
1.245
1.267
1.310
1.323
1.342
1.403
1.419
1.425
1.466
1.492
1.582
1.625
1.648
1.651
1.720
1.793
1.797
1.833
1.860
1.871
1.935
1.997
6.7 (Aroclor 1016)
2.3 (Aroclor 1016)

9.1 (Aroclor 1016)









16 (Aroclor 1254)
6.3 (Aroclor 1254)d


9.1 (Aroclor 1254)



30 (Aroclor 1254)




18 (Aroclor 1260)





9.1 (Aroclor 1260)

19 (Aroclor 1260)

5.3 (Aroclor 1268)







2.0 (Aroclor 1268)

a  Structures and IUPAC numbering from Ballschmiter and Zetl  (8); b  The NRCC mixtures contain PCB-159,
-182, and -187 which elute between PCB-129 and -183, but these congeners  are  not  adequately resolved
by the 60 m  DB-5  capillary column,  similarly,  PCB-171  and -202 elute between PCB-185 and  PCB-156;  c
Retention time relative to 4,4'-DDE; d Composition factor for PCB-110.
                                                n-133
-------
Table 2.    PCB  Congener  Composition  of  U.  S.  Environmental  Protection  Agency  Aroclor Reference
Materials.


                                      Aroclor Mixture Composition (wt %)


IUPAC No.       1221       1232      1016      1242      1248      1254      1260      1262      1268
0.62
9.3
0.85







8.1
27
5.9
3.9
2.7
2.6
0.65
0.70
0.38
0.73
15
43
11
7.6
5.7
5.6
1.3



14
33
6.9
5.7
3.6
3.8
0.89
1.5
0.75
1.0
7.4
15
8.9
13
8.4
8.1
1.8
3.8
2.1
5.2
18
15
31
52                       3.9       7.6       5.7       13        7.8       1.3
49                       2.7       5.7       3.6       8.4       1.3
44                       2.6       5.6       3.8       8.1       2.5
40
101                      0.70                1.5       3.8       13        6.3       2.6
87                       0.38                0.75      2.1       6.2       0.86
110                      0.73                1.0       5.2       16        3.3       0.80
151                                                              1.5       5.2       6.0
118                      0.35                0.91      3.5       11
153                                          0.23      0.59      5.3       13        11
105                                          0.69      2.4       3.3
141                                                              1.5       3.3       3.0
137                                                              0.54
138                                          0.28      0.75      9.5       13        7.3
129                                                              0.67
183                                                              0.49      5.5       7.9
128                                                              1.6       0.77
185                                                                        0.81      1.6
156                                                              1.0
200
180
170
201
203                                                                                            19
208                                                                                            18
195                                                                        0.66      2.3
207                                                                                  0.50      5.6
194                                                    0.22                1.3
205                                                                                  0.34
2°6                                                                        0.78      3.4       49
209                                                                                            11

s-PCB          10.8      52.6      89.2      73.3      82.3      85.0      76.4      79.5      139


The composition data for Aroclor 1254 and 1260  (congeners  eluting  between  PCB-101  and PCB-205) were
published previously (7).   Aroclors  1254,  1260, 1262,  and 1268 contained congeners  coeluting with
PCB-159/182/187 which were not resolved on the 60 meter DB-5 column.   Aroclors  1260,  1262 and 1268 had
peaks corresponding to PCB-172/202, also not resolved.   PCB-110 was  determined  as PCB-77 and corrected
(7).

0.70
0.40


0.93
0.86

0.51
11
5.3
3.3
1.5
14
6.3
11
2.5
1.6

32
                                                n-134
-------
Table 3.  Estimated Aroclor Content  of  Nine Regulated Aroclors Based on Content of Selected Congeners
                              Estimated Concentrations of Aroclor 1016,  1254,  1260 and 1268
                                in  100 pg/uL of Commercial Aroclors (pg/uL)
Marker
Chlorobiphenyl    1221
1232
1016
                    1242
                    1248
                                        1254
1260
                                                            1262
1268
18
15
31
Aroclor 1016

4.2
21
7.7
11
(81)a
54
62
54
57
(8.1)
101
99
100
100
(1.0)
94
76
63
78
(20)
50
35
81
55
(43)
87
110
118
105
Aroclor 1254
183
180
170
Aroclor 1260
6.1
4.6
3.2

3.5
(74)
          12
          6.3
          8.3
          21
          12
         (54)
34
33
32
72
43
(45)
99
101
100
99
100
(0.96)
14
21


8.1
-------
Table 4.  Polychlorinated Biphenyls  and  Estimated  Aroclor  Concentrations  in Auto Shredder Waste.


                     mg/kg                  Normalized
  Estimated Aroclor
Concentration (mg/kg)a
IUPAC No.
18
15
31
Aroclor 1016
52
49
44
60
101
87
110
151
118
114
153
105
Aroclor 1254
141
137
138
129
183
128
185
156
180
191
170
Aroclor 1260
201
203
189
207
194
205
206
Aroclor 1268
209
s-PCB
s-Aroclor
No. 1827
5.5
20
4.9

4.8
3.1
3.6
1.6
4.9
1.8
3.7
0.83
4.5
0.16
2.3
1.6

1.2
0.55
5.1
0.60
0.89
1 .4
0.081
ND
3.2
0.39
0.98

1.9
0.90
0.016
0.12
0.96
ND
0.82

0.15
82.917

No. 1831
8.6
31
7.9

6.6
4.2
4.8
2.6
6.8
2.6
5.1
1.2
6.6
0.21
3.3
2.4

1 .1
0.48
6.6
0.37
0.92
1.4
0.059
0.94
3.5
0.013
1.6

2.2
0.99
0.048
0.14
1.4
0.035
0.92

0.20
116.25

No. 1827
6.6
24
5.9

5.8
3.7
4.3
1.9
5.9
2.2
4.5
1.0
5.4
0.19
2.8
1.9

1.5
0.66
6.2
0.72
1.1
1.7
0.098
ND
3.9
0.47
1.2

2.3
1.1
0.019
0.14
1.2
ND
0.99

0.18


No. 1831
7.4
27
6.8

5.7
3.6
4.1
2.2
5.8
2.2
4.4
1.0
5.6
0.18
2.8
2.1

0.94
0.41
5.7
0.32
0.79
1.2
0.051
0.080
3.0
0.011
1.4

1.9
0.85
0.041
0.12
1.2
0.030
0.79

0.17


No. 1827
37
46
45
43(11)b





29
23

41


48
35(31)




16



29

19
21(32)

4.8




1.6
3.2(72)


99.0°
No. 1831
58
71
72
67(12)





42
32

60


72
52(35)




15



32

30
26(36)

5.2




1.8
3.5(69)


145.0°
   congeners.   The  reported Aroclor concentrations is the mean.
b  The number  in  parentheses  is  the RSD for Aroclor estimates.
c  Because of  the poor  fit, i.e., RSD >50%, Aroclor 1268 was not  included  in s-Aroclor.
                                               n-136
-------
   100
PQ
O
ft
I
01
0

0\<>
w
0
Q
2

§
                                     52
                                         49
                        44
     Figure 1.   Chlorobiphenyl congener content of low-
     chlorine-content  Aroclors.   Percent of s-PCB for PCBs
     18, 15, 31,  52, 49,  and 44  are compared.
 tn
 ,X
 \
 tn
 O
 H
 EH
 s
 H
 U
 S
 O
 u
 w
 H
 ffl
 O
 O
40
     30
     20
     10
      0
                                        • Observed
                                          (mg/kg)

                                        fal Selected
                                          Congeners

                                        U Method 8O80
                15    31
                     87
110  118  105  183  18O   170
  Figure  3.   Predicted and observed congener profiles
  (mg/kg)  for auto shredder waste sample No. 1831.   The
  profiles are in good agreement except for the Aroclor
  1254-associated congeners.
                               H-137
-------
w
CO
w
o
EH
O
w
EH
W
Q


                                                          AUTO SHREDDER WASTE
                                                                                                00
                                                                                                m

                                                                                       L^

                                                                                      JL
                                        TIME
            Figure 2.  Chlorobiphenyl congeners in automobile shredder waste
            sample extract no. 1824.  The chromatogram shows compounds eluting
            between about 10 and 90 min.
-------
68 THE DETERMINATION OF PART PER TRILLION LEVELS OF NITROAROMATICS IN GROUND
        AND DRINKING  WATER BY WIDE-BORE CAPILLARY GAS  CHROMATOGRAPHY


        Michael  Hable,  Catherine  Stern, Kenneth Williams,  United
        States Army Environmental Hygiene  Agency,  Aberdeen Proving
        Grounds,  Maryland 21010


        ABSTRACT


        A method  has  been  developed  to  extract and  analyze  for
        selected nitroaromatics at part per trillion levels in ground
        and drinking  water.   The compounds included are Nitrobenzene,
        1,3-Dinitrobenzene,  2,4-Dinitrotoluene,  2,6-Dinitrotoluene,
        2,4,6-Trinitrotoluene  and   1,3,5-Trinitrobenzene.      The
        extraction is  accomplished  using  a  simple  liquid/liquid
        extraction with toluene.  A surrogate, 3,4-Dinitrotoluene, is
        added to each sample in order to track sample recoveries.  The
        toluene  extract is analyzed  via a  gas chromatograph equipped
        with  a 12 meter DB-210 wide-bore fused silica capillary column
        and an electron capture detector.  Method detection limits of
        10 ppt for nitrobenzene; 30 ppt  for 1,3-DNB and 2,4-DNT; 6 ppt
        for 2,6-DNT;  and  20 ppt for 2,4,6-TNT and 1,3,5-TNB have been
        attained using  this method.


        INTRODUCTION


        Monitoring for  nitroaromatics in ground and drinking water is
        a primary  concern of the United  States Army.   Areas  most
        likely  to be contaminated include ballistic test  ranges as
        well  as  munitions processing and storage  sites.  This method
        was   developed   in   response   to   environmental   and  state
        regulatory concerns about possible nitroaromatic contamination
        of ground and drinking water.


        Initial  testing performed by other laboratories  as part of  a
        preliminary study of  one  of these  areas  had indicated trace
        level contamination of both 2,4-DNT and 2,6-DNT in  the ground
        water.    The  results  however were  inconclusive because of
        possible interferences from  the analytical procedure  used.
        In response to  the preliminary analysis,  geological studies
        and   shallow  ground   water   dye  tracing  experiments  were
        performed.   The results of these  experiments  indicated that
        any possible  contamination had  most likely originated from  a
        former ordnance works.
                                    n-i39
-------
The method initially chosen for the analysis of these samples
was EPA Method  609 because of  its'  traditional use  for  the
analysis of 2,4-DNT, 2,6-DNT, nitrobenzene  and isophorone in
municipal  and  industrial  discharges.   The methods'  stated
detection  limits  are  14 ppb for  nitrobenzene, 0.01  ppb  for
2,6-DNT and 0.02 ppb for 2,4-DNT.   Sample  preparation for  EPA
Method 609 involves the extraction of approximately one liter
of  sample  water  with  methylene   chloride,   then  solvent
exchanging the methylene  chloride with hexane, then  finally
concentrating the sample to ten milliliters or less.  Problems
encountered using this  method included the loss of some of  the
more volatile nitroaromatics and a concentrating effect of  any
interferences present in the sample water.  In  addition,  EPA
Method 609 does not address the analysis  of 1,3-DNB,  2,4,6-
TNT or  1,3,5-TNB.   In contrast,  the  simple  liquid/liquid
toluene extraction described in this presentation  offers  the
advantages of minimal  solvent  use and no  heating,  solvent
exchange or blowdown.
                           n-i40
-------
6o                   A PERFORMANCE  EVALUATION OF THE  CLP
                    HIGH  CONCENTRATION ORGANIC PROTOCOL


     Harald G. Buhle.  Gary L. Robertson, Lockheed Engineering & Sciences
     Company, 1050 E.  Flamingo Rd,  Las Vegas,  NV 89119 under contract to
     U.S.  Environmental  Protection  Agency,  Environmental  Monitoring
     Systems  Laboratory,  Las Vegas,  Nevada.


     Abstract;

     The   United  States  Environmental  Protection   Agency  Contract

     Laboratory   Program   (CLP)  has  issued  contracts   for  the  gas

     chromatography/mass  spectroscopy  analysis of samples  containing

     high concentrations  of organic chemicals.  It  is essential that the

     data users  know the  quality  of data produced by the method.   As

     this is  a new protocol, many of the quality  assurance  parameters

     have only  recommended  control  limits.    The performance of  the

     method  on  actual  samples will  be discussed.  The  precision  and

     accuracy of the method  determined  from  sample  quality assurance

     data will be presented.  Data  will  be  presented on the following

     quality   control  parameters   with   the  intent   of   suggesting

     acceptability criteria  where  appropriate:   surrogate recovery,

     retention times, response factors, internal standard area response,

     control  matrix spike recovery  and method blanks.   The  laboratory

     results  obtained  on  performance evaluation  samples will  also be

     discussed.
     Notice; Although the  research  described in this article has been
     supported by  the Environmental Protection Agency under contract
     68-03-3249 with Lockheed Engineering & Sciences Company, it has  not
     been subjected to Agency review and therefore does not necessarily
     reflect the views of the Agency and no  official  endorsement should
     be inferred.
                                     n-141
-------
70           OXIDATION OF ACID SURROGATES AND TARGET ANALYTES
                          IN ENVIRONMENTAL WATER SAMPLES
                                  USING EPA METHOD 625

          Paul R Chen. Group Leader, William  A. Van Ausdale, Senior Associate Scientist, and
          Dwight  F.  Roberts, Manager, GC/MS Department,  Analytical Services Division,
          Environmental Science and Engineering, Inc., P.O. Box 1703, Gainesville, Florida 32602

          ABSTRACT

          Apart from potentially poor laboratory techniques and occasionally severe emulsion prob-
          lems during extraction, very low acid surrogate recoveries for environmental water samples
          using EPA Method 625 are usually caused by the oxidation of the acid surrogates during
          acid extraction.  These very low recoveries are usually accompanied by the presence of
          iodocyclohexanol in the chromatogram of the sample extracts.  lodocyclohexanol is pre-
          sumably a reaction product of iodine, generated in the sample from the oxidation of iodide,
          with cyclohexene which was added by the manufacturer to the methylene chloride as a
          preservative. Thus, the presence of iodocyclohexanol can be used as a marker for the exis-
          tence of oxidizing agents in the sample.  The oxidation products of the acid surrogates, i.e.,
          2-fluorophenol, phenol-d5, and 2,4,6-tribromophenol, were identified as fluorobenzo-
          quinone, benzoquinone-d4, and 2,6-dibromohydroquinone, respectively. The presence of
          these oxidation products can be used to confirm that the low acid surrogate recoveries for a
          specific  sample are due to the oxidation degradation of the acid surrogates.  The matrix ef-
          fects of samples with low acid surrogate recoveries on the recoveries of phenolic target ana-
          lytes of  CLP-HSL compounds were also investigated. The results showed that good re-
          coveries were obtained for phenols with strong electron-withdrawing substituents, while
          low recoveries were obtained for phenols with electron-donating substituents or with weak
          electron-withdrawing substituents. The problem of oxidation of acid surrogates and target
          phenolic analytes can be eliminated by adding a reducing agent such as sodium thiosulfate
          to the sample before extraction.  The problem also can be reduced by adjusting the pH for
          the extraction.  Thus, the sample can be initially extracted at pH 10 to remove 2,4-
          dimethylphenol, then at pH 7 to partially extract most of the other phenols and finally at pH
          2, to extract strong acids such as the nitrophenols and dinitrophenols.

          INTRODUCTION

          EPA Method 625(1) is widely used in environmental  laboratories for the analysis of
          semivolatile organic compounds by gas chromatography/mass spectrometry (GC/MS). In
          this method, a 1-L aliquot of water is extracted with methylene chloride at pH > 11 and
                                              n-142
-------
then at pH < 2. Before extraction, acid and base/neutral surrogate standards are spiked into
the water samples to monitor the recoveries of compounds in each sample. If the surrogate
recoveries are within established control limits, it is assumed that the method is in control.

In our analyses of thousands of water samples, it was not unusual that we obtained no or
low recoveries for acid surrogates (2-fluorophenol, phenol-d5, and 2,4,6-tribromophenol),
but good recoveries for base/neutral surrogates in the same sample. This no or low acid
surrogate recovery only occurs in the environmental water samples, not in the soil samples
nor in the method blanks using HPLC grade water. This suggests that the no or low acid
surrogate recoveries  observed in the samples were due to matrix effects of the envi-
ronmental water samples.

EXPERIMENTAL SECTION

Sample Preparation: Most samples were obtained from our clients and extracted at pH > 11
and then at pH < 2 with methylene chloride according to  EPA Method 625(1).  Before
extraction, each sample was spiked with 1.0 mL of surrogate  standard spiking solution
which contained 100 ug/mL each of acid surrogates (2-fluorophenol, phenol-d5, and
2,4,6-tribromophenol) and 50 ug/mL each of base/neutral surrogates (nitrobenzene-d5,
2-fluorobiphenyl, and terphenyl-d!4). To study the effects  of a  reducing agent and pH on
the recoveries of surrogates and target analytes, a composite sample was prepared in the
laboratory by combining field samples which gave no or low acid surrogate recoveries.
This composite sample was divided into four equivalent test samples which were all spiked
with the surrogates and EPA Contract Laboratory Program (CLP) hazardous substances list
(HSL) of 65  semivolatile organic target analytes (2).   A sufficient amount of sodium
thiosulfate was added to one of the test samples.  This sodium thiosulfate treated test
sample and an untreated test sample were extracted at pH 12 and then at pH 2.  The re-
maining two test samples were extracted at pH 10 then pH 2 and pH 12 then pH 7, respec-
tively. All methylene chloride extracts were concentrated to 1 ml with a Kuderna-Danish
(K-D) concentrator. For the samples obtained from our clients, the base extract and acid
extract of each sample are combined and analyzed by GC/MS. For the test samples the two
extracts were analyzed separately by GC/MS.

GC/MS Analysis:  Samples were analyzed on a HP 5988  GC/MS system.  The column
used was a 30m x 0.25 mm DB-5 fused silica capillary column  (J&W Scientific, Folsom,
CA).  The column  temperature was held isothermal at 40°C for 4 minutes and then pro-
grammed at 10°C per minute to 280°C, and held isothermal at this final temperature for 12
minutes.  The mass spectrometer was scanned from 35 to 500 amu per half second. CLP-
HSL compounds and the surrogate  compounds were quantified using multiple internal
                                     n-143
-------
standards. The area of the extracted ion current profile at the characteristic m/z of a given
analyte was used to calculate the concentration.

RESULTS AND DISCUSSION

 Most of the samples obtained from our clients had good acid and base/neutral surrogate
recoveries.  However, in some sets of samples, a significant number of samples had low
acid surrogate recoveries. A chromatogram of one of these samples is shown in Figure 1.



^
't/j
S 500000_
1z


0


3
\



x 6
N.

• — 	 	 .
' 1 '
4

4
V'



!e ' els
V '
\ '
r
jl
~^Jk_lN^^
6 8









1 '
1(








V.
1 1 '
) 1
1





2
/
jL..
2 14









1 ' 1
ie







5
L . x,
i , . | , 1 1
5 18 2









I ' I ' I '
0 22









2









1 I ' I
4 26







1
\l . ^ . ,1 . .
28 30 32
                                        Time (min)
    Figure 1. Total ion chromatogram of a water sample which has no acid surrogate
    recoveries. Peaks 1 and 2 are artifacts formed during extraction. Peaks 3, 4, and 5
    are oxidation products of surrogates.  The large unlabeled peaks are internal
    standards and base/neutral surrogates.
 These samples usually  contained iodocyclohexanol (peak 1), and to a lesser extent,
 chloroiodocyclohexane (peak 2). Iodocyclohexanol, in some cases bromocyclohexanol,
 and halogenated cyclohexanes are artifacts formed in acidic media during extraction (3,4).
 Iodocyclohexanol is presumably a reaction product of iodine (generated in the sample under
 acidic conditions) with cyclohexene. Cyclohexene is added by the manufacturer to the
 methylene chloride as a preservative and scavenger (3,4).  Thus, the presence of iodocyclo-
                                      n-144
-------
hexanol in samples suggests that the sample contains an oxidizing agent or agents which
oxidize iodide to iodine. Iodine then reacts with cyclohexene in acidic media to from
iodocyclohexanol. The presence of iodocyclohexanol can be used as a marker for the exis-
tence of oxidizing agents in the sample.  In addition to iodocyclohexanol, bromocyclohex-
anol and halogenated cyclohexanes were observed in some samples.  We found large
amounts of bromocyclohexanol, iodocyclohexanol, bromochlorocyclohexane, dibromo-
cyclohexane, and chloroiodocyclohexane in high salinity or brine samples.

Examinations of the mass spectra of the small peaks in the chromatograms of samples
yielding no or low acid surrogate recoveries resulted in the identification of three oxidation
products, one for each of the three acid surrogates. Peaks 3, 4, and 5 in Figure 1  are iden-
tified as benzoquinone-d4, fluorobenzoquinone, and 2,6-dibromohydroquinone which are
the oxidation products of phenol-d5, 2-fluorophenol, and 2,4,6-tribromophenol, respec-
tively.  The presence of these oxidation products can be used as a confirmation that the low
acid surrogate recoveries for the samples are due to the oxidation degradation of the acid
surrogates. The amount of the three identified oxidation products only accounts for a small
percentage of the amount of the acid surrogates spiked into the sample.  This may  be due to
the formation of other intermediate or final oxidation products which are soluble in acidic
media  and are not extractable by methylene chloride, or to the formation of nonvolatile
products, e.g., polymeric oxidation products (5), and are not amenable to analysis by
GC/MS.

For the test sample to which no reducing agent was added and was extracted under the
normal conditions of the method (pH 12 then pH 7) the recoveries were good only for phe-
nols with strong electron-withdrawing substituents or with strong acidity such as 2,4-dini-
trophenol, 2-nitrophenol, 4-nitrophenol, and 4,6-dinitro-2-methylphenol. Low recoveries
were obtained for phenols with electron-releasing substituents or with weaker  electron-
withdrawing substituents. This is in agreement with the results reported by Stone(5) that
phenols with electron-withdrawing substituents are more resistant to oxidation than phenols
with electron-donating substituents.  For the test sample to which sodium thiosulfate was
added, good recoveries were obtained for all the phenols because the oxidizing  power of
the sample had been suppressed. For the sample extracted at pH 7 instead of pH 2, the
phenols with medium acidity had reasonable recoveries while the phenols with very weak
acidity such  as 2,4-dimethylphenol and with strong acidity such as 2,4-dinitrophenol, 4-
nitrophenol and 4,6-dinitro-2-methylphenol had very low or no recoveries.  The low
recovery for 2,4-dimethylphenol can be explained by the fact that it is the weakest acid
among the phenols we studied and is  most susceptible to oxidation degradation even at pH
7. The reason that no recoveries were obtained for 4-nitrophenol and 2,4-dinitrophenol is
                                      n-145
-------
that their acidities are too strong for them to be extracted at pH 7. If the pH was lowered to
pH 2, good recoveries for these two compounds would be expected. If the sample was
initially extracted at pH  10 even in the presence of oxidizing agents a good recovery was
obtained for 2,4-dimethylphenol.  This is because at this high pH, the oxidation rate for
dimethylphenol proceeds slowly.  Our unpublished data show that at pH > 10, the recovery
for dimethylphenol is  lower.  This is because at this high pH most of dimethylphenol will
be in phenoxide ion form and will not be extracted by methylene chloride.  Thus,  the
sample can be initially extracted at pH 10 to remove 2,4-dimenthylphenol, then at pH 7 to
partially extract most of  the other phenols, and finally at pH 2 to extract strong acids such
as the nitrophenols and dinitrophenols.

ACKNOWLEDGEMENTS

We thank P. Dumas, K. Cunningham, M. Mignardi, and D. Schindler for laboratory assis-
tance.

REFERENCES

(1)    EPA 40 CFR Part 136, Fed. Regist. 1984. 49, No 209.
(2)    EPA Contract Laboratory Program, Statement of Work for Organic Analysis.
       Multi-Media Multi-Concentration. 1988.
(3)    Campbell,  J. A.; LaPack, M. A.; Peters,  T.L.; Smock, T. A. Environ. Sci.
       Technol. 1987. 21, 110-112.
(4)    Fayad, N. M. Environ. Sci. Technol. 1988. 22, 1347-1348.
(5)    Stone. A. L. Environ. Sci. Technol. 1987. 21. 979-988
                                     H-146
-------
71       Hans Cornet , Stephan Rose ,  Daan  Lammerts  van  Bueren
        Chrompack  Int.  B.V.,  P.O.Box  8033,  4330  EA  Middelburg,   The
        Netherlands

        The use  of high__gejrfQrmance  liquid chroroatography techniques  in
        the screening  for priority  PQ Ilu
        The pollution of  the  biosphere  is  becoming more  and  more  a  public
        and governmental  concern,  resulting in a  growing  need for  more
        sensitive  and  reliable monitoring procedures for a  wide  range  of
        compounds. Due to the complexity of  environmental samples and the
        growing  number of  contaminants to  be analyzed, the demands  on
        analytical procedures are becoming more and more demanding.  High
        performance  techniques,  both  for  analysis  as  well  as  sample
        preparation  are   required,  while  automation  is   desired  for
        obvious  reasons .
        This  paper investigates  the  use of various  HPLC procedures for
        sample preparation  as well as chromatographic analysis of  a  wide
        range of compounds .

        Firstly  a method  is shown  for  the  screening of  surface- and
        drinking water  for pesticides and their residues, combining  off-
        line preconcentration with gradient  reversed phase HPLC.
        For  the  analysis  a  silica based  reversed  phase  column  with
        polymeric  modification  is  used.  This method,  showed, detection
        limits  of  sub ppb  levels  in drinking water for substances  like
        ureapesticides , chlorophenoxy^icids ,  nitro- and  chlorophenols

        The second method describes the analysis  of volatile ketones and
        aldehydes  in air.  The off-line  sampling  procedure  incorporates
        pre-column  derivatisation  with 2,4  dinitrophenylhydrasine .  The
        trapped  DNPH-derivatives are extracted and analysed  using F.PL.C.
        Limits of  detection are  in the  ppb range.

        Thirdly  the possibilities of  HPLC  as sample preparation technique
        are evaluated.  Complex samples are  purified using miniature  HPLC
        columns, in multidimensional  chromatography approaches.
        In  the  analysis of pesticides and polychlorinated biphenyls  HPLC
        is  used  successfully to  separate the  analytes  from  the matrix.
        The  HPLC  column  eluate   is  collected  and analyzed using  high
        resolution capillary  GC.
        In the analysis of  poly  aromatic hydrocarbons  a preconcentration/
        sample clean up step  is  combined on-line with  gradient EPLC.
                                    n-147
-------
72      OPTIMIZATION OF CONTINUOUS LIQUID-LIQUID EXTRACTION PROCEDURES FOR
        SEMTVOLATILE AND PESTICIDE ANALYSIS

        John DeWald. Daniel TeUez, Mostafa Mayahi, S-CUBED, A Division of Maxwell Laboratories,
        Inc., 3398 Carmel Mountain Road, San Diego, California 92121-1095

        ABSTRACT. The authors have performed an extensive investigation into the performance and
        optimization of sample preparation techniques utilizing continuous liquid-liquid extraction for
        organochlorine pesticide and semivolatile compound analysis of water matrices. The extraction
        procedures used were based on the EPA  CLP Statement of Work 2/88 and SW-846 Method
        3520.  The analytical procedures used were based on EPA CLP Statement of Work 2/88 for
        pesticides and semivolatiles.  The investigation focused on four topics:  (1) optimization of
        extraction efficiency, (2) determination of the steps in which significant analyte loss may occur,
        (3) determination of errors and accidents which commonly cause failures, and (4) comparison of
        the continuous liquid-liquid extraction technique to the separatory funnel shake out technique.


        The goal of this investigation was to identify critical parameters in the extraction process and to
        optimize analyte recovery and process efficiency in a laboratory where over 75 continuous
        extractors may be in operation at once. This paper presents data from studies conducted by the
        authors which included the effects of extraction rate, extraction time, extractjdrying, and errors in
        Kuderna Danish concentration on analyte recovery. The  advantages of continuous liquid-liquid
        extraction over separatory funnel shake out extraction, such as increased acid fraction recoveries
        will also be discussed.  This  paper will present recommendations for the optimum performance
        of continuous liquid-liquid extraction in an environment requiring high sample throughput.
                                              n-148
-------
        EXTRACTION OF CONTAMINATED SOILS WITH COSOLVENTS

David C. Erickson. Research Associate, Raymond C. Loehr, Professor, and William F.
Lane, Graduate Research Assistant, Department of Civil Engineering, University of Texas
at Austin 78712.

ABSTRACT

The solubilization of hydrophobic compounds in cosolvents was investigated. Soil samples
contaminated with wood preserving or utility residues were extracted with aqueous dilutions
of methanol and 2-propanol. The resulting extracts were analyzed for polynuclear aromatic
hydrocarbons using high performance liquid chromatography. Results of the analyses
indicated that the solubility of the compounds tested increased with increasing volume
fraction of methanol or 2-propanol and that the increase was semi-logrithmic. The aqueous
solubilities of the compounds tested were less than that predicted from crystalline solubility
data.

INTRODUCTION

The University of Texas Environmental and Water Resources Engineering (UT-EWRE)
laboratory  is conducting research to investigate the solubility of polynuclear aromatic
hydrocarbons (PAHs) in solutions of contaminated soils and water, and cosolvents.

Polynuclear aromatic hydrocarbons may be found in soils and in industrial residues as the
by-product of  activities such as petroleum refining and combustion of fossil fuels. PAH
compounds are neutral nonpolar organics consisting of two or more benzene rings arranged
in various configurations. Many of these compounds are of concern because of their toxicity
or carcinogenicity. Consequently, sixteen of the compounds are regulated as hazardous (40
CFR 261.31-32) by the United States Environmental Protection Agency (US-EPA).

These compounds may be characterized by their low aqueous solubility and their strong
sorption to soils  and soil  organics  (Dzombak  1984). Consequently, the  solubility and
movement of  many of these compounds in ground water is  low.  Solubilization of the
compounds is enhanced, however, in  solutions containing miscible organics.  Chiou (1989)
indicated that the presence of a high molecular weight humic material in water can enhance
the apparent solubility of hydrophobic solutes. Cosolvents also may enhance the solubility
of hydrophobic molecules.  The term cosolvent refers to a solution containing a completely
miscible, or a partially miscible organic solvent and water. As the concentration of a miscible
solvent such as methanol or 2-propanol increases, the hydrophobicity of the solution
increases and so does the solubility of hydrophobic compounds in the  solution. Fu et al.
(1986) reported that hydrophobic aromatic solutes displayed a semi-logrithmic increase in
solubility with increasing volume fraction of solvent in cosolvent mixtures. Their results
were obtained using soils with low organic carbon content (0.5 to 2.85% by dry weight) and
added aromatic solutes. The sorptive capacity of soils decreases in the presence of cosolvents.
Rao et al. (1985)  presented isotherm data indicating that the sorption of anthracene with
uncontaminated soils decreased by 4 orders of magnitude as the volume fraction of methanol
in water increased from 0 to 100%.
                                       n-149
-------
Objectives In this study, the effect of cosolvent solutions on the solubility of PAHs present
in soils contaminated with wood preserving and manufactured gas plant (MGP) residues
was investigated. The objectives were to: (a) identify the solubilities of the PAHs in various
cosolvent fractions and (b) to use these enhanced solubilities to identify relationships which
can be used to predict released amounts of constituents if environmental factors and chemical
characteristics are known.

MATERIALS AND METHODS

Soils Contaminated soils containing wastes from different sources were used in this study.
Descriptions of these soils and wastes are provided below.

(i) Soils Containing MGP Wastes: These soil samples were collected at a site in the northeast
at which coal tar was buried in an unconfined disposal pit. The coal tar,  a by-product of the
MGP process,  had been in place for more than 20 years and  some migration of the
constituents had occurred. Soils from near the source (Site A) and soils  downgradient (Site
B) were  collected and used  in  this  study. The  site (A) samples were more  heavily
contaminated than were the site (B) samples.

(ii) Wood Preserving Wastes:  Samples of soil collected from this site (Site C) contained
residues of organic chemicals remaining from a midwestern land treatment operation. The
land treatment is part of an ongoing program to evaluate the biological treatment of wood
preserving wastes containing compounds such as creosote and pentachlorophenol.

(iii) Utility Wastes: This midwestern site contains significant levels of organic compounds
remaining from the disposal of MGP wastes. Site (D) samples were collected in an area of
actual waste disposal, while the site (E) samples were collected from an area downgradient
from the disposal area which was  less contaminated.

Desorption Method  Soils were extracted using water or cosolvent solutions following the
procedure in Table 1.   Methanol and 2-propanol, used as cosolvents, were supplied by
Fisher (Houston, TX) and were HPLC grade.  The cosolvents  were made by adding
deionized-distilled water (Millipore Milli-Q , Bedford, MA) on a volume-volume basis to
a sufficient volume of organic solvent. Typical cosolvent percentages (fc) used during the
desorption experiments included 0,10,20,30,40,50, and 75 percent.  All desorbing solutions
contained 0.01 N CaCLj.

Analytical Methods  High performance liquid chromatography (HPLC) was used to
quantify the PAH compounds in the solutions from the desorption studies. Analyses were
performed using a Waters HPLC system controlled by an NEC PC running the Maxima 820
software.  Operating conditions are shown in Table 2. Samples were injected using a WISP
autosampler, and a Waters variable wavelength  UV detector was used to detect eluting
compounds^ Compounds in the samples were quantified using a five-point standard curve
and standards were prepared from a commercial stock solution of the 16 PAH compounds
(610A, Supelco, Bellefonte.PA).
                                      n-iso
-------
                                    TABLE 1

                    PROTOCOL FOR BATCH DESORPTION STUDIES
(a) Sieve wet soil throughrlmm sieve. Air dried soil may be used, but the wet soil is preferred
to prevent loss of volatile compounds.

(b) Weigh a known amount of soil (approximately 9 grams dry weight) into a centrifuge
tube. Repeat for other tubes.

(c) Prepare solutions of the various concentrations of the cosolvent to be studied. For
example, 0,10,20,30,40,50  and  75 percent  solutions are suggested. All aqueous and
cosolvent solutions should contain 0.01 N CaCL^ to facilitate phase separation.

(d) Add a known volume of water or cosolvent solution (approximately 36 mL) into a teflon
centrifuge tube containing the soil sample. Prepare triplicate tubes for each concentration
of cosolvent.

(e) Place the tubes in a rotary mixer perpendicular to the axis of rotation, and rotate for 24
hours at 30 revolutions per minute.

(f) After mixing, centrifuge the tubes at 10,000 revolutions per minute.

(g) Using a Pasteur pipet, remove the centrifugate from a tube and add to a glass vial equipped
with a teflon lined cap. Repeat for the other tubes.

(h) Store the vials at 4° C until ready for HPLC analysis.
                                    TABLE 2
HPLC OPERATING CONDITIONS
Instrument
Initial Solvent Ratio
Gradient
Column
Detector
Waters Gradient HPLC
35% Acetonitrile and 65% Water
Linear gradient to 100% acetonitrile in 30 minutes
Supelco, Supelcosil LC-PAH reverse
mm
Variable wavelength, operated at 254
phase, 15cmx
nm
4.6

 Solid Phase Extraction. Quantification of the PAH compounds in the aqueous solutions
 (zero fc) was difficult because of their low solubilities. Consequently, in these samples, a
 solid phase extraction/concentration procedure was used  to  concentrate  the target
 compounds. The concentration was achieved by pumping, with a 10 mL glass syringe, 20
                                        n-151
-------
mL of the aqueous sample through a Waters C-18 Sep-Pak cartridge. The PAH compounds
retained on the cartridge were eluted with 2 mL of acetonitrile (Fisher Optima) which then
was analyzed using HPLC.

RESULTS AND DISCUSSION

The desorption studies were performed to  evaluate the solubility of PAH compounds in
cosolvent  solutions. Complete results of the desorption experiments with data for each
compound in each soil are not shown in this paper because the amount of data generated
was prohibitively large. Instead, typical results are presented that are representative of the
trends observed. Figures  1 and 2 graphically present the concentration of the respective
3-and 4-ring PAH compounds, anthracene and pyrene, present in the extracting solutions
of the MGP high level soil (A). Figures 3 and 4 present similar data for the 4 ring compounds,
fluoranthene and pyrene in extracts of the wood preserving soil (C).  In each Figure, the fc
values are plotted as the x-axis and the log PAH concentration are plotted as the y-axis.
Each point is the average of triplicate analyses, and the best-fit-line as determined by
regression analysis is shown in the Figures. The best-fit-line was calculated using:

                                     y=ae°*                                  (1)

                                      where:

                           y =PAH Concentration (mg/L),

                                  a=y intercept,

                      a = slope and x = fraction cosolvent (fc)

The plots  indicate that the concentration of compounds present in the cosolvent solutions
increased  semi-logrithmically as the fraction of organic solvent in solution was increased.
This observation is based on the apparent linearity of the data shown.  The linear regression
for the data presented in Figures 1 to 4 were calculated and equations for the best-fit-lines
are shown on the Figures. The r2 values for the lines were near one signalling close agreement
of the data to the regressed line. The slopes (a) of the regressed lines for these and other
desorption analyses (data not shown) are presented in Table 3.  The slopes for the data
showed a  possible trend of increased slope with increasing compound ring number.  This
trend was most evident for the 2 and 3 ring compounds.  This trend is similar to the data
presented  by Fu et  al. (1986) who found  that increases of solute solubility  in miscible
cosolvents were more pronounced for more hydrophobic solutes.

The respective 2-propanol a values in Table 3 are generally larger  than the a values for
methanol. This indicates that the 2-propanol cosolvents extracted higher concentrations of
compound than did the methanol. Decreasing the polarity of the cosolvent may enhance the
solubility of hydrophobic compounds, however, the limited miscibility of most hydrophobic
solvents would limit their effectiveness as cosolvents.
                                        n-i52
-------
FIGURE 1. EXTRACTION OF 3-RING PAHs USING
 METHANOL-WATER AND 2-PROPANOL-WATER
                   MGP Wastes - High Level, Site A
    1 ,OOO F-
     1OO —
       1 O
 i
o
Q_
        1  —
      0.1  -
     0.01
                     y — O.O28 exp O.136X
                         r2- 0.99
                           Anthracene  Anthracene
                           2-Propanol   Methanol
                     2O         4O         6O         8O
                     °/o Volume  Fraction Solvent
                                                                  100
 FIGURE 2. EXTRACTION OF 4-RING PAHs USING
  METHANOL-WATER AND 2-PROPANOL-WATER
                   MGP Waste - High Level, Site A
    1 ,OOO
      1OO
       10  -
 8
        1  -
      0.1  -
     0.01
                              Pyrene     Pyrene
                            2-Propanol  Methanol
                                          	A	
                     20         40         60         80
                     °/o Volume  Fraction Solvent
                                                                  1 OO
                                 H-153
-------
FIGURE 3.  EXTRACTION OF 4-RING PAHs USING
 METHANOL-WATER AND 2-PROPANOL-WATER
              Soils with Wood Treating Residues, Site C
   1 ,OOO  c-
CJ>
o
     100  —
      10 —
i   —
      O. 1
     O.O1
                         Fluoranthene   Fluoranthene
                           Methanol      2-Propanol
                    2O         -4O         6O          BO
                    %> Volume Fraction Solvent
                                                                  100
FIGURE 4. EXTRACTION OF 4-RING PAHs USING
 METHANOL-WATER AND 2-PROPANOL-WATER
              Soils with Wood Treating Residues, Site C
    1 ,OOO
     1OO —
O
"TO
                             Pyrene      Pyrene
                            Methanol   2-Propanol
    O.O1
                    20          40         60          80
                        Volume Fraction Solvent
                                                          100
                                 n-i54
-------
                                        TABLE 3

         SIGMA a VALUES FOR DESORPTION STUDIES, METHANOL COSOLVENTS
COMPOUNDS :

2-Rings
Naphthalene
3-Rings
Acenaphthylene
Acenaphthene
Fluorene
Phenanthrene
Anthracene
4-Rings
Fluoranthene
Pyrene
Benzo(a)anthracene
Chrysene
5-Rings
BenzofbAfluoranthene
Benzofkjfluoranthene
Benzo(a)pyrene
Dibenzo(ah)anthracene
6-Rings
Benzo(ghi)perylene
Indeno(123-cd)pyrene
SIXES.
(A)
MGP High

0.053

0.068
0.085
0.086
0.088

„
0.097
0.069

	
—
—
	

	
~
(B)
MGP Low

__(!)

0.053
0.084
0.097
0.110

„
0.088
0.11
0.062

	
	
	
	

	
—
WoodPr.

„

—
—
	
0.074

—
0.083
0.068
0.066

0.052
0.029
0.049


	
--

(D)
MGP High

0.036

—
—
0.082


0.090
0.077
0.062

	
—
—
	

	
—
MGP Low

0.066

0.092
0.037
0.01
0.002

0.011
0.009
0.001

	
	
—
	

	
—
(l)--Compound was below quantitation limit in cosolvent solutions.

                                      TABLE 3 cont.

         SIGMA a VALUES FOR DESORPTION STUDIES, 2-PROPANOL COSOL VENTS
                                                        SITES
                                 (A)
                              MGP High
                                            (B)
                                         MGP Low
                      Woo
}od Pr.    MGP Hi
                      MGP High
             (E)
          MGP Low
          2-Rings
Naphthalene
          3-Rings
Acenaphthylene
Acenaphthene
Fluorene
Phenanthrene
Anthracene
          4-Rings
Fluoranthene
Pyrene
B enzo(a)anthracene
Chrysene
          5-Rings
Benzo(b)fluoranthene
Benzofkjfluoranthene
Benzo(a)pyrene
Dibenzo(ah)anthracene
          6-Rings
Benzo(ghi)perylene
IndenpT123-cd)pyrene
0.059

 0.09

 OJ'l
 0.13
 0.14
                                 0.15
                                 0.12
0.053

 0~09
 0.14
 0.15
            0.14
            0.23
            0.13
                                                         0.14
            0.13
            0.12
            0.12

            0.13
            0.11
                                                                    0.044
                                                                    0.068
                                                                    0.11
 0.13
 0.11

0.094
                       0.076

                       0.102

                        0~05
                        0.01
                       0.013

                       0.014
                       0.012
                       0.005
                       0.002
                                                                                0.007
(l)--Compound was below quantitation limit in cosolvent solutions.
                                           H-155
-------
 Results similar to these have been described for the solubilization of drugs in cosolvents.
Yalkowsky et al. (1981) reported that the solubility of nonpolar compounds in cosolvents
(log SJ increased exponentially with increasing cosolvent composition. If a is defined as
the slope of the log Sm versus fc plot, then the solubility of hydrophobic compounds in
cosolvents can be estimated by:
                                             + afc                              (2)

                                      where:

                           Sm is the cosolvent solubility,

                             Sw aqueous solubility and

                             fc is the cosolvent fraction
 For the soils tested, the aqueous solubilities of the compounds at zero/c were less than would
 be predicted  based on crystalline solubility data. In Tables 4 and 5 the y-intercepts, which
 were calculated during the regression analyses, are shown. Also shown  are reported
 solubilities of the pure compounds dissolved in water (Mackay et al. 1977). In some of the
 soils, one or more of the 16 PAHs were not present in the sample and therefore would not
 be expected to be found in the zero/c solutions. To determine which compounds were present
 in the extracted soil, data from the 75 percent 2-propanol desorption solution were evaluated
 (Table 6). The  75 percent solution contained a significant fraction of the organic solvent
 and was  effective for extracting the  PAH compounds  as shown  in comparisons  with
 methylene chloride extracts (data not shown).  If a particular compound was not present in
 the 75 percent 2-propanol cosolvent, then it was considered to not be present in the soil and
 was so indicated in Tables 4 and 5. The quantitation limits (indicated with the less than
 symbol <) are shown in instances where the compound was present in the soil, but was not
 detected in a sufficient number of different cosolvent ratios to calculate a value for the y
 intercept. Compounds such as acenaphthene which were not observed in any of the five
 soils are not listed in Tables 4 and 5.

 The y intercept data, which correspond to the concentration of compound in  the zero fe
 cosolvent solutions, are estimates of the aqueous solubilities. Actual measured values using
 zero/c are shown in Table 7. which were obtained using a Sep-Pak extraction/concentration
 technique. In most instances  the  measured values  were  comparable to the calculated
 solubilites shown in Tables 4 and 5. Two exceptions were naphthalene and acenaphthylene
 which were present in the  Site (B) and in the Site (E) zero/c solution respectively, but not
 in the  higher (75% 2-propanol, Table 6) cosolvent solution. These compounds may have
 been detected in the zero fc because  of the good analytical sensitivity of the  Sep-Pak
 concentration method.
                                        n-ise
-------
                                    TABLE 4
   CALCULATED AQUEOUS SOLUBILITIES FOR METHANOL COSOLVENT DESORPTION
                                    CURVES

~~
Naphthalene
Acenaphthylene
Fluorene
Phenanthrene
Anthracene
Fluoranthene
Pyrene
Benzo(a)anthracene
Chrysene
Benzo(b)fluoranthene
Benzo(k)fluoranthene
Benzo(a)pyrene
SQLJUBPTY1
(m$L)

31
3.9
1.98
1.29
0.073
0.26
0.135
0.014
0.002
_J2>
—
-
SQLIB1LQ3BS (»JS&)JOR SOILS
M.OM STTES:
CAI '•
MCrP ;
Ki£h
2.94
1.0
0.45
0.25
0.053
<0.002
0.059
0.033
<0.002
<0.007
<0.004
<0.002
Cil
MGF
Low
N.R.
0.48
0.24
0.12
0.02
<0.002
0.02
0.007
0.009
<0.007
<0.004
<0.002
WoodPr.
N.R.
<.08
<.002
<0.004
0.015
0.098
0.13
0.03
0.02
0.02
0.019
0.017
MdP
High
13.8
N.R.
<0.002
0.009
<0.004
0.01
0.01
N.R.
0.001
N.R.
N.R.
N.R.
Mf
... .Low 	
0.07
N.R.
0.03
0.01
0.002
0.01
0.009
N.R.
0.001
N.R.
N.R.
N.R.
f NJR. Compound not present in soil.
'Crystalline solubilities (Mackay and Shiu 1977)
 Value not in reference.
                                    TABLE 5
   CALCULATED AQUEOUS SOLUBILITIES FOR 2-PROPANOL COSOLVENT DESORPTION
                                    CURVES


Naphthalene
Acenaphthylene
Fluorene
Phenanthrene
Anthracene
Fluoranthene
Pyrene
Benzo(a)anthracene
Chrysene
Benzo(b)fluoranthene
Benzo(k)fluoranthene
Benzo(a)pyrene
SOUJBHJTy*
{rog/t)

31
3.9
1.98
1.29
0.073
0.26
0.135
0.014
0.002
~(2)
__
~
souJBiLrrrEs (m»oFQR SOILS
FROM SITES;
Hieh
3.1
0.87
0.34
0.12
0.03
<.002
0.04
0.004
<.002
<0.007
<0.004
<0.002
^
Low
N.R.*
0.83
0.25
0.12
0.02
<0.002
0.04
<0.003
0.01
<0.007
<0.004
<0.002
(Q
WoodPr.
N.R.
<0.08
<.002
<0.002
0.01
0.06
0.1
0.02
0.01
0.006
0.002
0.009
A
High
12.0
N.R.
0.016
0.009
<0.004
0.008
0.008
N.R.
0.009
N.R.
N.R.
N.R.
Low
0.08
N.R.
0.05
0.01
0.01
0.01
0.01
N.R.
0.002
N.R.
N.R.
N.R.
fN.R. Compound not. present in soil.
-Crystalline solubilities (Mackay and Shiu 1977)
 Value not in reference.
                                       n-157
-------
                                     TABLE 6
                AVAILABLE COMPOUND IN DESORPTION SOLUTIONS™
                                      (mg/L)
COMPOUNDS

Naphthalene
Acenaphthylene
Acenaphthene
Fluorene
Phenanthrene
Anthracene
Fluoranthene
Pyrene
Benzo(a)anthracene
Chrysene
Benzo(b)fluoranthene
Benzofkmuoranthene
Benzo(a)pyrene
Dibenzo(ah)anthracene
B enzo(eh itoerylene
Indeno(12T-cd)pyrene

fAY
MGP Hiffb
61
71(2)
<8'8
76
2&
84
15
11
3
9
<0.14
<.07

(S)
MGP Low
<.2
16
<0.2
40
19
118
61
10
6
2
4.5
<07
1
SITES .
{C)
WobdPr.
19
<0.2
4
20
200
164
52
53
32
15
29
<0.14
<0.07
6

MGPHieh
78
<0.2
0.3
0.7
0.2
2
0.9
<0.06
<0.14
<0.08
<0.04
<0.14
<0.07
<0.06

MGr!
-£TW\
0.2
<3.2
0.4
0.2
0.04
0.2
0.1
<0.06
0.2
<0.14
<0.08
<0.04
<0.14
<0.07
<0.06
(1) Based upon concentrations recovered from the 75% 2-propanol cosolvent.
7 Quantitation limit.
(' Present, but not quantified because of coelutmg peaks.
                                     TABLE 7
                      MEASURED AQUEOUS CONCENTRATIONS
COMPOUNDS
RECOVERED

Naphthalene
Acenaphthylene
Fluorene
Phenanthrene
Anthracene
Fluoranthene
Pyrene
Benzo(a)anthracene
Chrysene
Benzo(b)fluoranthene
Benzo(k)fluoranthene
CONCBNTRA^gM g^ ^R SOILS
MGP^iett
2.3
0.9
0.3
0.11
0.03
<.002
0.06
0.01
<0.002
<0.007
<0.007
CP)
MOP tow
1.4
0.7
0.2
0.14
0.03
<0.002
0.09
<0.003
<0.002
<0.007
<0.007
Wooc??r.
N.R.
<0.08
<.002
<0.004
<0.004
0.07
0.06
0.01
0.01
0.01
<0.007
MGP^h
12.4
N.R.
<0.002
0.01
<0.004
0.01
0.01
N.R.
0.02
N.R.
N.R.
MOP LOW
0.07
0.1
0.03
0.01
<0.004
0.01
0.01
N.R.
<0.002
N.R.
N.R.
In nearly all cases, the calculated and measured solubilities were lower than that predicted
by crystalline solubilities. The decreased solubilities could be the result of unidentified
organic solutes from the soil present in solution. The following relation (equation 3) based
on Raoult's Law may explain the low solubilities of the compounds in the aqueous samples:

                                                                                (3)
                                        n-i58
-------
                                      where:

                        Sw is the measured aqueous solubility

                          X,- is the solute mole fraction and

                  5f is the ideal solute solubility (Yalkowsky 1980)

The ideal solubility 5, can be calculated from the following equation:

                           logS(0 = logSc + 0.01(MP - 25)                         (4)


                        where : Sc = Crystalline Solubility and

                              MP= Melting Point (C°)
Equation 3 predicts that the aqueous solubility of solutes dissolved in soil organic matter
will be influenced by the amount of other organic compounds present in the mixture. As
the mole fraction of the solute increases, aqueous solubility will increase to a maximum
equal to the crystalline solubility. In complex wastes, where the mole fraction of individual
compounds may be small, the aqueous solubilities of the compounds may be less than the
respective  crystalline solubilities.  Validation of this relationship for any soil requires
knowledge of the mole fraction of solutes of interest. This information is not readily available
because of the many compounds present in these types of wastes.  Work, however, is
continuing in this research to investigate the appropriateness of using Equation 3 to explain
the lowered solubilities.


The solubility data suggest that the aqueous solubilities of compounds emanating from
heavily contaminated soils may be less than their respective crystalline solubilities. This
may result in inaccurate modelling of pollutant fate and transport in soil water systems. In
addition, the aqueous concentrations of hydrophobic compounds,  whose source  is a
contaminated soil, can be difficult to determine because of the low concentrations involved.
An alternate approach  to quantifying these  compounds may involve cosolvents.  By
analyzing cosolvent extracts of the soil, and plotting these concentrations, the aqueous
concentration of material may be estimated by extrapolating to zero cosolvent.

Conclusions

PAH compounds in aseries of soils previously contaminated with coal tar or with creosote
were extracted with cosolvents containing methanol or 2-propanol.  Data from this work
indicate that the solubility of PAH compounds extracted from these soils increases with
increasing  volume  fraction of  methanol  or 2-propanol,  and that the  increase is
semi-logrithmic. In addition, the aqueous solubility of compounds extracted from heavily
contaminated soils may be lower than predicted from crystalline solubility data through the
influence of other organic compounds present in the soil.
                                        n-159
-------
Acknowledgements

This work was supported by Electric Power Research Institute Agreement No. RP2879-7.
We wish to acknowledge Nadine Gordon for her technical assistance with the analysis of
samples for this project.
References

Chiou, C.T. 1989. Theoretical Considerations of the Partition Uptake of Nonionic organic
compounds by Soil Organic  Matter, p.  1-29. In Reactions and Movement of Organic
Chemicals in Soils,  Soil Science Society of America, Special Publication no. 22.

Dzombak, D.A.,  and R.C. Luthy. 1984. Estimating Adsorption of Polycyclic  Aromatic
Hydrocarbons on Soils. Soil Sci. 137:292-307.

Fu, J.K., and R.C. Luthy . 1986. Effect of Organic Solvent on Sorption of Aromatic Solutes
onto Soils. J. Env. Eng.  112:346-366.

Mackay, D.  and  W.Y.  Shiu.  1977.  Aqueous Solubility of  Polynuclear  Aromatic
Hydrocarbons. J. Chem. Eng.  Data. 22:399-402.

Rao,  P.S.C., A.G.  Hornsby,  D.P.  Kilcrease, and P. Nkedi-Kizza. 1985. Sorption and
Transport of Toxic Organic Substances in Aqueous and Mixed Solvent Systems. J. Environ.
Qual. 14:376-383.

Yalkowsky, S.H. and S.C. Valvani. 1980.  Solubility and Partitioning I:  Solubility  of
Nonelectrolytes in Water. J. Pharm. Sci. 69:912-922.

Yalkowsky, S.H. And TJ.Roseman. 1981. Solubilizationof Drugs by Cosolvents. p91-134.
In S.H. Yalkowsky Ed. Drugs  and the Pharmaceutical Sciences, Volume 12, Techniques of
Solubih'zation of Drugs. Marcel Dekker, Inc. New York.
                                      n-160
-------
74                   HIGH-PERFORMANCE  THIN  LAYER  CHROMATOGRAPHIC ANALYSIS
                                   OF ANILINES AND PHENOLS

         Sam Ferro,  Chemist, Midwest  Research Institute, 425 Volker Boulevard,
         Kansas City, Missouri 64110

         Aromatic amines  and  phenols  are of environmental  interest  because  they  are
         commonly found at  Superfund  sites.  They are  also found  as contaminants in
         many  dyestuffs.    Several  of  these  compounds  are  known  to  be  toxic  or
         carcinogenic.

         A  rapid,  sensitive method  has  recently been developed  for  identification
         and  quantitation of  aromatic amines and  phenols.   Mixes  of  aniline  and
         phenol standards  have been separated,  identified, and  quantitated  by  high-
         performance thin  layer  chromatography  (HPTLC).   This technique can be used
         quantitatively  if  only a few components  are  to  be analyzed,  or it can be
         used  as  a  screening  device  for several target  analytes.  Fifteen  samples
         and  standards can  be  applied  to  a  single plate, which  allows  for high
         sample  throughput.   The  entire   analysis  and  detection  time is  usually
         about  one  hour per plate.   Samples and standards are applied to  the same
         plate  so  that  a  direct  comparison  can  be  made.   Approximate  on-plate
         detection  limits are  in the  100-nanogram  to 1-microgram  range.

         Individual  standards  and  the  mixes  are applied  to  Whatman silica  gel
         60 F25it  HPTLC  plates with a  Camag Linomat IV band applicator.  Since  the
         rate  of  sample  application  can  be varied,  samples  can  be  applied  from
         almost  any  solvent.    Volatile solvents  such  as methanol  or hexane  are
         preferred,  but  even  water  samples can  be directly applied  to the  plates.
         All  samples analyzed  for this work were dissolved and  applied  to  the  plate
         in methanol.  Typical volumes spotted were from 5 to 50  microliters.

         The  plates  were  developed in classic twin trough TLC chambers. The mobile
         phase  used for  separation  of the  anilines is methylene chloride:methanol
         (97:3).   A slightly  less polar  solvent  system,  methylene chloride:hexane
         (98:2),  is  used for  the  phenols  analysis.  The TLC tanks were allowed to
         equilibrate with the solvent vapors for  at least 30 min before the plates
         were  inserted.   The  solvent is  eluted up the  HPTLC  plate.  The development
         distance  (the distance  traveled   by the  solvent  front) was  approximately
         7 centimeters.

         Detection  and quantitation  are  obtained by  scanning  the  plates  with LFV
         light  using the Camag  TLC Scanner II  scanning densitometer.  The optimum
         detection  wavelengths were determined  to be 210  nanometers for phenols  and
         254 nanometers  for  anilines.   Figure 1  is  a  HPTLC  chromatogram of  six
         phenols  that was obtained using the analytical conditions  reported  above.

         Identification of  the individual  components in  a mix is  typically obtained
         by  comparing the  retention  distance  of  the separated  components to  the
         individual  standards.    UV-visible  spectra  of  individual  components  can
         also  be  obtained  with  the  scanner as  an additional  identification  tool.
         The  retention  characteristics of  several anilines and  phenols are compared
         in Tables  1 and  2.
                                             n-161
-------
100 mV -i
        0 urn
10
20
30
40
                                                             50
                                                       60 mm
         Figure  1:  Chromatogram of a Standards Mixture of Six Phenols
           Separated on an HPTLC Plate (210 nm wavelength detection)
           The mixture contained 1.25  pg nitrophenol;  2.5  yg phenol
             2.5 pg 2,4-dimethylphenol;  2.5  pg 2, A-dichlorophenol
                          1.25 pg  2,4-dinitrophenol;
                  and  1.25 yg of 2-methyl-4,6-dinitrophenol.
                                     fl-162
-------
Table 1:  Retention Characteristics of Selected Aromatic Amines
    Analysis  performed on HPTLC plates with a 97:3 methylene
                chloride:methanol  solvent system

                   Compound                         Rr3
           Aniline                                  0.59
           2-Methylaniline                          0.66
           3-Methylaniline                          0.60
           4-Methylaniline                          0.56
           2,4-Dimethylaniline                      0.62
           n-Ethylaniline                           0.76
           a-Phenylethylamine                       0.23
           p-Phenylenediamine                       0.03
           Tribenzylamine                           0.88
           p-Dimethylaminoazobenzene                0.87
           a   R£ = distance  traveled  by sample/distance
              traveled by solvent  front.
    Table 2:  Retention  Characteristics of Selected Phenols
    Analysis performed on HPTLC plates with a 98:2 methylene
                 chloride:hexane solvent system

                   Compound                        Rfa
           Phenol
           2-Chlorophenol
           2,4-Dichlorophenol
           2,4,6-Trichlorophenol
           Pentachlorophenol
           4-Chloro-3-methylphenol
           2,4-Dimethylphenol
           2-Methyl-4,6-dinitrophenol
           4-Nitrophenol
           2,4-Dinitrophenol
           a  R£ = distance traveled by sample/distance
              traveled by solvent front.
                              n-i63
-------
75       A INTERLABORATORY  COMPARISON  OF A  SW-846  METHOD  FOR THE
        ANALYSIS  OF THE  CHLORINATED PHENOXYACID HERBICIDES BY LC/MS,
        Tammy L.  Jones,  Chemist,  Quality Assurance Division, Leon D.
        Betowski,  Research Chemist,  Quality Assurance Division,
        U.S.    Environmental   Protection   Agency ,    Environmental
        Monitoring Systems  Laboratory,  P.O.  Box 93478,  Las Vegas,
        Nevada,  89193-3478;  Tom C.  Chiang,  Staff Scientist, Lockheed
        Environmental  Services Company,  1050 E. Flamingo Rd.,
        Suite 120,  Las Vegas,  Nevada,  89119.
        ABSTRACT

        Recently the U.S. Environmental Protection Agency's  (U.S.EPA)
        Environmental  Monitoring Systems Laboratory-Las Vegas
        (EMSL-LV),  working in  close  cooperation with  the  Office of
        Solid Waste (OSW),  completed  a interlaboratory evaluation of
        a liquid chromatography/mass  spectrometry (LC/MS) method for
        the analysis  of chlorinated  phenoxyacid herbicides.   This
        method uses neither hydrolysis nor esterification to prepare
        samples  for  analysis.    While method  performance  data was
        obtained in the study,  the  focus of this  evaluation was to
        test  the  intercomparability   of  LC/MS  data.    In  order  to
        minimize  interlaboratory  variability  due  to  sample  and
        standard preparation, the sample extracts and a  stock standard
        solution   for   calibration   were   prepared   by   Lockheed
        Environmental    Services  Company    (LESC)   and    sent   to
        participating   laboratories   for  LC/MS   analysis.    Another
        element of  this study was the  comparison of data obtained from
        both types of  LC/MS  interfaces  [i.e.  thermospray   (TSP) and
        particle beam   (PB) ] .    The   data  generated   by  the  study
        demonstrated that phenoxyacid herbicides can be diagnostic of
        instrument   performance  problems.     Some  characteristics
        important to instrument performance,  particularly  for particle
        beam,  were interface temperature and source cleanliness,  an
        increase in  thermal degradation  ions  was  observed  in the
        spectra of systems whose performance was not  optimum.   With
        thermospray the loss of sensitivity  due to high thermospray
        temperatures can be a diagnostic problem.  Those laboratories
        proficient  in  LC/MS analysis  demonstrated that the  limits of
        detection (LCD's) were  comparable between the thermospray and
        particle beam  interfaces  for compounds  with  high  molecular
           NOTICE:  Although the research described  in  this  article has
     been  supported  by the  United  States  Environmental  Protection
     Agency, it has not been subjected to Agency review and, therefore,
     does  not  necessarily  reflect the  views  of  the  Agency,  and no
     official endorsement  should  be inferred.  Mention  of  trade  names
     or  commercial  products  does   not   constitute   endorsement nor
     recommendation for use.
                                    n-164
-------
weight  and  thermal" stability.   However, the  low molecular
weight and thermally labile compounds  (e.g. dalapon, dinoseb)
were better observed using thermospray.

INTRODUCTION

Gas  chromatography/mass   spectrometry  (GC/MS)   is  widely
recognized  as  one  of  the  most  powerful  techniques  in
analytical chemistry.  Despite its' utility, the application
of GC/MS is limited to the  following categories of compounds:
those that are amenable to  solvent extraction, those that can
be isolated by purge and trap techniques, and those that are
sufficiently volatile  to pass  through the GC column and yet
are resistant to thermal degradation.  Unfortunately, A number
of compounds of environmental interest, including many on the
Resource Conservation  and  Recovery Act (RCRA)  Appendix VIII
list, are polar, non-volatile, and/or  thermally  labile.  One
important goal of  the U.S. EPA methods development efforts has
been  to develop  techniques  for  compounds  that can  not be
measured by conventional techniques (e.g. GC).

High performance liquid chromatography (HPLC) methods are not
subject to the above mentioned  analytical  limitations, making
them suitable for many of the analytes which lie outside the
scope of conventional  GC.  Gradient elution techniques provide
analysts with the  tools necessary to separate complex mixtures
of polar, ionic, and high molecular weight organic materials.
However,  due  to  limited   types  of   HPLC  detectors   (e.g.
Ultraviolet/Visible, fluorescent, electrochemical) and their
lack  of structure confirmation  capability,  HPLC sacrifices
many of the benefits of selective GC and GC/MS analysis.

The  logical solution  to  this dilemma is to  couple liquid
chromatography  and  mass  spectroscopy  to produce a sensitive
and reliable technique for  identifying and quantitating polar
and  thermally  labile compounds.    Examples  of environmental
organic  contaminants  which  could  be  analyzed by  LC/MS are
organophosphorus pesticides(1),  triazine herbicidesc2), and the
chlorinated phenoxyacid  herbicides.<3)   This latter group of
compounds can  be  analyzed  as the  free acid,  plus the ester
form, by LC/MS.   The  chlorinated herbicides include some of
the  most widely  used  herbicides  for systemic  control of
broadleaf  herbage  for  both  agricultural   and  residential
usages.

Currently SW-846  methods 8150 and  8151  are approved by the
U.S. EPA, for the analysis  of chlorinated herbicides  in solid
waste under the Resource Conservation and Recovery Act (RCRA).
These methods specify quantitation by GC with electron capture
detector (BCD) and optional GC/MS confirmation. These methods
require the use of hydrolysis and subsequent esterification
                            n-i65
-------
of the sample extracts  before  analysis.   Sample  hydrolysis  is
a time consuming process and may not always be  quantitative.
The usual esterification reagent,  diazomethane, is a both a
potential  carcinogen  and  explosive.    Therefore,  method
accuracy,  sample  throughput,   and  laboratory   safety,  are
improved by the elimination of these two steps.  The analysis
of  several   chlorinated   herbicides  (specifically  2,4-D,
dicamba, and MCPP)  by HPLC using reverse phase  column and UV
detector has  been  reported in  the  literature( }.   A multi-
laboratory collaborative study of the HPLC/UV method has been
conducted and published(5).   In addition,  a ACS  monograph and
a U.S.EPA  internal report have  been written on the  use of
LC/MS for the analysis of these herbicides( ' }.   The use of
LC/MS not  only eliminates the  need for the  hydrolysis and
esterification  steps,  but also provides  a method  for the
direct and specific analysis of  these compounds, using a mass
spectrometer.

Presented  in   this  manuscript  are  the  results   of  an
interlaboratory  evaluation of  the  analysis  of chlorinated
herbicides and  their  esters  by LC/MS,   using  two different
types of LC/MS interface devices  (TSP and PB).   Samples were
sent  to  thirteen  participating   laboratories  utilizing
thermospray and/or particle beam interfaces originating from
several  different  manufacturers.   Statistical  analyses were
performed on the data (only 7  laboratories returned data) for
individual analytes  using  Statistical Analysis  System  (SAS)
software.  The  performance of  the two interface devices was
evaluated in terms  of the applicability of each  to  quantitate
these herbicides, as  well as,  the intra- and  inter-  laboratory
precision and accuracy of these data.
EXPERIMENTAL

Since the purpose of this  study  was to compare and evaluate
the different LC/MS  interface devices  for their applicability
to  the  analysis of  a  particular  group of  compounds  (acid
herbicides and  their esters),  sample  extracts,  instead of
samples, were provided to all the participating laboratories.
This not only eliminated any discrepancies in results due to
variations   in   extraction   efficiencies   by   different
laboratories, but also reduced the amount of work required for
each laboratory.

Duplicate sample extracts,  consisting of the eleven analytes
in acetonitrile, at  four different  concentration levels,  were
sent to each of  the  thirteen  laboratories.  Each laboratory
was asked to perform triplicate analysis on the four duplicate
samples using the  equipment available  in  their laboratory.
The laboratories involved utilized approximately equal numbers
                            n-166
-------
of TSP-  and PB-LC/MS instruments,  and  the instruments were
manufactured by five different instrument companies.

A concentrated stock standard solution, containing the
eleven analytes in acetonitrile at  1000 /xg/mL, was sent with
each  sample  set,  for  instrument  calibration  and  analyte
quantitation.   The  same  stock standard  was  also  used for
preparation of sample extracts, so that all the chemicals used
in this study were traceable to a single original source.
It  is  necessary for  all  the  standards and  extracts  to be
prepared in  acetonitrile.   The use of methanol could result
in the methylation of the free acid herbicides.

A method blank extract was shipped with  each sample set.  The
blank was prepared with a  sample of tap  water,  using the same
procedure employed for sample extraction  (Method 8150).

For  calibration  standards a minimum  of three concentration
levels  were  specified.    The  recommended  low  calibration
standard  was to  be at  a  level  close to  the  instruments
detection limit.  The recommended medium and high calibration
standards were five times  and fifty times  higher  than the low
level calibration standard,  respectively-  Due to  the unstable
nature of some of the instruments, no acceptance criteria were
required for the calibration factors.

No   specific  parameters  were  given   to   the  participants
concerning   instrument   (interface   and    MS)   tuning   and
calibration.   The laboratories  were advised  to  follow the
instrument   manufacturer's   specifications   for   optimal
performance.

Separate HPLC conditions were recommended to laboratories with
different instrumentation.  Flow rates for those laboratories
with TSP interfaces  and post-column  0.1  M ammonium acetate
additions: flow rate  0.4 mL/min to 0.6 mL/min,  with 0.8 mL/min
post-column  flow.    Flow  rates  for  laboratories  with TSP
interfaces, but without post-column addition:  flow rate
1.0  mL/min to 1.2 mL/min.   Flow rates for those laboratories
with PB  interfaces: flow rate 0.4 to 0.6 mL/min.  Analytical
column:  15  cm x  2.1 mm  i.d.,  C-18,  reverse  phase,  0.5 jum
particle size.  Use of a guard column was recommended.
                            n-167
-------
HPLC gradient elution conditions:

Time
(min.)   1% acetic acid in Water     1% acetic acid in methanol

  0              50%                       50%
  2              50%                       50%
 12              40%                       60%
 18                0%                      100%
 28                0%                      100%
 33              50%                       50%
 38              50%                       50%
It is necessary to have  1%  acetic acid in the mobile phases
in order to keep the  acid analytes equilibrated in their acid
form.
RESULTS AND DISCUSSION

Thirteen  laboratories  were each provided  with eight  sample
extracts, one blank, and one concentrated  stock standard for
instrument calibration.  The eight sample  extracts  consisted
of four duplicate extracts each containing  eleven  analytes at
different concentration  levels.   In addition, the  duplicate
samples at each concentration level were treated as individual
samples.

Data  were   received  from  only   seven   of  the   thirteen
laboratories.  Data were collected by four  laboratories using
the particle beam interface (two different  manufacturers) and
by three laboratories using the thermospray interface,  (three
different manufacturers) .  These seven data packages were used
to evaluate the two main LC/MS interface devices (i.e. TSP and
PB)   for   their   applicability  and   performance   in  the
quantitation of the acid herbicides  and their esters.

A SAS software program was employed to perform the comparison
of  the  data  for  the  overall  performance between the two
interfaces.  A probability (P) value(7) was calculated in order
to determine  if  a  significant difference exists  between the
two groups  of data.    In  order for the two data  sets to be
significantly different with 95% confidence, the P value must
be  less than  0.05.    This comparison  was performed  on an
individual analyte basis at each  concentration level.  For
example, 2,4,5-T at the theoretical  concentration of
500  Mg/mL,  the  mean  analytical results  from the four PB
instruments  is  543.68 ±  76.24, and  the  corresponding mean
value from the three  TSP instruments  is 448.985 ± 102.57.  The
calculated P  value is  0.2169,  indicating  that there is no
                            n-168
-------
significant statistical  difference between the  two  sets of
data.

Another objective of this study was to compare the performance
of  the  two  LC/MS  interface  devices  made  by  different
instrument manufacturers.  The experimental design for making
such   an   evaluation  is   influenced  by  the   number  of
participating laboratories  and their instrumentations. Because
less than half  of the  expected data packages were returned,
some of the comparisons were impossible  to perform.  However,
four sets of PB  data, produced by using interfaces made by two
different  instrument  manufacturers,  were  received.    This
allows for  a  limited comparison of PB results obtained from
manufacturer  A  and B.   The P test was again  used  for this
comparison, by  combining the data from the two laboratories
using  PB  from  manufacturer  A  and comparing  it with the
combined data from two  laboratories using manufacturer B.  The
test was  performed on  an  individual  analyte  basis,  and was
done both by  considering concentration  as a variable (i.e.,
comparing  results  from  samples  at the  same  concentration
level), and by considering  the concentration as  a non-variable
(i.e.,  comparing  results  of samples  combined  from  all four
concentration levels).

When results  of samples  from  all  four  concentration levels
generated by instrument A were combined  and compared with the
combined results  generated by instrument  B,  the statistical
test indicated again no significant difference between the two
data sets.

Table 1 shows  the  overall precision and accuracy data from the
seven data packages  received.   Included in this table is the
designation  of   either particle  beam,  denoted  as  PB,  or
thermospray, TS,  for interface  device.

Initially,  the  duplicate extracts at the same concentration
level were evaluated separately in order to determine if there
was any statistical difference in the results between the two
samples.   SAS  analysis  of the 'relative  standard deviation
(RSD)  values  from the  two  duplicate  samples,  from  the same
laboratory, indicated that  there was no significant difference
in  the  results  between  the  two  samples.    This  was not
unexpected,   since   these   two  extracts  were  identically
prepared.   However, it  was  important  to  demonstrate that
statistically there was no  difference between the  samples from
duplicate extracts  so  that triplicate data  from each of the
two duplicates  could be combined for statistical analysis.
                            n-169
-------
SUMMARY

Although the data collected was from a limited data set (i.e.
seven laboratories, and one type of extract)  the  statistical
results showed some interesting and informative results,  from
which a few general and specific conclusions can be extracted.

Dalapon was not detected by PB, presumably because it  is too
volatile to be transmitted through  the PB interface. Dinoseb
is a phenol, which respond poorly to  PB  as a  class.  Even at
the highest concentration  level, 500  /ig/mL,  only  one  of the
four PB laboratories reported values for dinoseb. Dicamba also
did not  respond well by  PB,  especially  on  extracts  at the
lower concentration  levels.   None  of the four laboratories
using  PB detected  dicamba in  the low  concentration  level
extract, 5 /jg/mL.

With the exception of dalapon,  dinoseb,  and dicamba, PB tends
to give bias high results, at 500 jitg/mL,  (average percent mean
bias is +16%)  and TSP tends to  give  bias  low results (average
percent  mean  bias is  -10%)  when compared to the  true  value.
The tabulated results  for the precision data indicate that PB,
at 500 Mg/mL,  gives better precision (average %RSD is 7%)  than
TSP (average  %RSD is 22%).

At  the  medium concentration level,  50  |ig/mL,  there was no
clear difference between the results obtained from PB and TSP,
except for the previously mentioned compounds,  although  both
are biased low compared to the true value.

Only one of the  four  PB laboratories, could detect analytes
in  the  low  concentration level,  5  /ig/mL,  extract.    This
laboratory used a 20 /xL injection volume instead  of the  4 nL
as most other  laboratories used.  The results reported by this
laboratory  were  at least  twice as high  s  the  theoretical
value. This strongly indicates that the  detection limits for
these compounds by PB  is at least above 5 /xg/mL in the extract
(or above 20 ng at a 4 juL injection volume) .  Two of the three
laboratories  using  TSP,   reported  values  for   this  low
concentration  level  extract,   and   these  results are in
reasonable agreement, with  a very  low bias (average percent
bias is  4%)  and an average %RSD of  ±  19%,  see Table 1.   This
is  a strong indication  that  TSP provides better  sensitivity
in detecting  low  levels of these compounds than PB.

None of  the  participating  laboratories  reported any  value
above the detection limit on the blank sample,  indicating no
serious contamination problem throughout this  study.

PB generally gives better precision than TSP,  particularly at
the high concentration level (500 Mg/mL).  This is indicated
                            n-no
-------
by the lower %RSD values, shown in Table 1.  Since TSP  is more
sensitive  in  detecting these target  analytes,  the extracts
often had  to  be diluted prior to  injection,  in order to be
within  the   linear  response  calibration   range  of  the
instrument.   Therefore,  dilution  steps may have  contributed
partially to the higher %RSD  (poorer precision)  observed for
TSP.   However,  it can  be assumed  that the  difference in
precision in part is due to the fundamental differences in the
operating principles of the two interface systems.

Instrument  calibration is an important  factor in producing
accurate data.  A  uniform calibration method was  not  used by
all  of  the  participants,  due to  the instabilities  of the
instruments.   This  may have  contributed to  poor accuracy
observed in this  study.

As  far  as  the results of the comparison between the two PB
manufacturers  from  the  four  sets  of   data   a  couple  of
conclusions can be drawn.  For sample extracts  at 500 /xg/mL,
250  ng/mL,, and  50 ^g/mL  levels,  there is  no  significant
difference between the results obtained by using manufacturer
A  or B  on  all 11  analytes tested.   When results of  samples
from all four  concentration levels generated by instrument A
were combined and compared with the combined results generated
by  instrument B,  the statistical  test   indicated  again no
significant difference between the two data sets.

Based on the  results  obtained from this  study, LC/MS can be
used, and  should  be the choice of instrument in  the  future,
for  the analysis of chlorinated phenoxyacid herbicides. While
TSP  interfaces provide better sensitivity, PB  interfaces tend
to  provide  better precision  for  a majority of  the compounds
tested.   While PB has the advantage  of providing detailed
information  on the  structure of  the  analytes,  it  can not
detect  dalapon,  and it responds  very poorly to both  dicamba
and  dinoseb.

The  choice for which interface to use,  TSP or  PB,  will depend
on the type of analytes and the analytical requirements of the
data user. From  the  data  obtained in  this   study  one can
conclude  that  for the  analysis  of   low  level  samples,
thermospray,  in negative ionization mode, would be preferred
for  phenoxyacid herbicides.   For the analysis  of  high level
samples in  which  structural  confirmation of the  analytes is
essential, particle beam with electron impact ionization might
be preferred.
                            n-171
-------
ACKNOWLEDGEMENTS

The authors would like  to  thank all of the laboratories  who
participated in this  study.   They are not identified  as  all
laboratories were promised anonymity.
REFERENCES

1. Betowski,  L.D.,  Jones, T.L.,  Environ.  Sci. Tech.  (1988),
   22,  1430-1434.

2. Parker,  C.E., Haney, C.A, Harvan,  D.J.,  and Mass,  J.R., J^
   Chro.  (1982), 242,  77-96.

3. Jones,  T.L.,  Betowski, L.D.,  Yinon,  J., "The Analysis of
   Chlorinated   Herbicides   by   High   Performance   Liquid
   Chromatography/Mass    Spectrometry".        In     Liquid
   Chromatography/Mass	Spectrometry:	Applications	in
   Agricultural, Pharmaceutical,  and  Environmental Chemistry;
   Brown,  M.A.,  Ed.;  American Chemical  Society:  Washington
   D.C., 1990.

4. Grorud,  R.B., Forrette, J.E.,  J. Assoc.  Anal. Chem.  (1983)
   66,  1220-1225.
5. Grorud, R.B., Forrette, J.E., J. Assoc. Anal.  Chem.  (1984)
   67,  837-840.

6. Jones,  T.L.,  Betowski, L.D.,  "Liquid Chromatography/Mass
   Spectrometry   Performance   Evaluation    of    Chlorinated
   Phenoxyacid Herbicides and Their Esters", EPA/600/X-89/176,
   July,  1989.

7. SAS  Institute Inc. SAS/STAT™ Guide  for Personal Computers.
   Version 6 Edition. Gary, NC: SAS Institute Inc., 1987.  1028
   pp.
                            n-172
-------
Table I - Precision and Accuracy of Interlaboratory  Data

  Analytes (500 ug/mL)       Mean               % RSD
  2,4,5-T
  2,4,5-T,BUTOXY
  2,4-D
  2,4-DB
  DALAPON
  DICAMBA
  DICHLORPROP
  DINOSEB
  MCPA
  MCPP
  SILVEX

  Analytes (50 ug/mL)
  2,4,5-T
  2,4,5-T,BUTOXY
  2,4-D
  2,4-DB
  DALAPON
  DICAMBA
  DICHLORPROP
  DINOSEB
  MCPA
  MCPP
  SILVEX

  Analytes (5 uq/mL)
  2,4,5-T
  2,4,5-T,BUTOXY
  2,4-D
  2,4-DB
  DALAPON
  DICAMBA
  DICHLORPROP
  DINOSEB
  MCPA
  MCPP
  SILVEX
PB
544
675
556
602
•
473
552
317
553
533
609
Mean
PB
31.1
39.4
42.7
35.6
•
36.7
48.6
15.0
52.9
51.0
35.9
Mean
PB
11.2
12.0
13.5
10.3
m
*
16.2
•
14.0
14.5
10.8
TS
449
450
431
475
415
386
422
392
444
430
480
TS
31.1
42.6
31.9
51.8
47.0
45.2
48.2
36.5
47.8
38.2
32.4

TS
4.52
4.97
5.60
4.80
7.48
5.26
5.10
5.41
4.73
4.92
4.35
PB
6.27
11.9
9.06
4.08
•
10.7
5.73
14.4
5.35
7.46
5.78
%
PB
21.0
13.9
25.7
12.4
•
23.4
35.9
39.1
41.3
46.5
27.1
%
PB
3.66
19.3
32.6
14.8
•
•
34.5
•
20.8
18.0
10.7
TS
24.5
22.8
25.5
22.3
36.3
17.6
18.1
13.8
22.9
18.2
20.4
RSD
TS
34.5
14.5
26.8
18.3
22.6
16.6
12.5
11.8
5.93
43.1
29.8
RSD
TS
32.6
24.1
24.1
7.35
19.6
16.8
15.1
23.3
18.7
14.0
11.4
                          n-m
-------
       THE DISTRIBUTION OF TARGET  COMPOUND LIST
             ANALYTES IN SUPERFUND SAMPLES
     Y.  Joyce Lee, Gary Robertson, Jack Berges, Lockheed
     Engineering &  Sciences  Company, 1050  E.  Flamingo
     Road,  Las Vegas, Nevada  89119  under  contract to
     U.S. Environmental Protection Agency, Environmental
     Monitoring Systems  Laboratory,  Las Vegas, Nevada.
ABSTRACT:

The United  States  Environmental Protection  Agency has

established a database known as the Contract Laboratory

Program (CLP) Analytical Results Database (CARD) to store

all of the analytical data reported by CLP laboratories

on samples from Superfund sites.   This is the first time

that  all  the analytical  results and  their  associated

quality  assurance  data  has  been  assembled  in  one

database.   With this  enormous  database,  it is possible

to examine many  facets of program performance.  One area

of great  interest  is the identities  of  the  frequently

found compounds  at Superfund sites, their concentration

levels and their distribution trends.  This presentation

will   describe   the  frequency,   distribution,   and

concentration of the detected compounds both nationally

and regionally.   The  data  will  be examined  for both

geographical  trends   and   possible  relationships  of

compounds that are frequently found together, both within

and between analytical fractions.   The results will be

presented in tabular and graphical formats.
                          n-i74
-------
Notice; Although the  research  described in this article has been
supported by  the Environmental Protection  Agency under contract
68-03-3249 with Lockheed Engineering & Sciences Company,  it has not
been subjected to Agency review and therefore does not necessarily
reflect the views of the Agency and no official  endorsement should
be inferred.
                                n-175
-------
77      DUAL-COLUMN/DUAL-DETECTOR APPROACH TO GAS CHROMATOGRAPHT.C ANALYSIS
       OF  ENVIRONMENTAL POLLUTANTS
        _l,,  E.  Ualdin, J.  Benedicto,  and J.  Mi lanes,  Mid-
       Pacific  Environmental Laboratory, Mountain  view,  California,  and
       W.F.  Beckert,  Environmental  Monitoring Systems  Laboratory,  U.S.
       EPA,  Las Vegas,  Nevada.

       Dual-column/dual-detector gas  chrowatoqraphic procedures  using
       30-m  x 0.53-mm  ID  fused-silica  open  tubular columns  are  being
       developed  in  our  laboratory  for the analysis  of  compounds of
       environmental  significance. Two columns  of  different polarities,
       thus  different  selectiv it ier.  toward  the.  target  compounds,  are
       connected to an  injection  too. and identical detectors . This  allows
       the   primary   and   confirmatory   analyses   to    be  performed
       simultaneously.  The  target  compounds include  34 phenolic compounds
       (as  pentaf luorobenzyl  bromide  derivatives),  v*'l.  organochlorine
       pesticides,   42    organophosphorus   pesticides,   22  chlorinated
       hydrocarbons,  16 phthalate  esters, and .16 nit.roaromatic compounds.
       Retention times, relative  rotention times,  method reproducibility
       and   linearity , instrument  detection  .limits,  and  selection  of
       surrogate compounds  and internal  standards  will  b« discussed  1'or
       each group of  target compound:.;.
      NOTICE: Although the  research  described in this abstract has been
      supported by the United States Environmental Protection Agency, it
      has  not  been subjected  to Agency  review and  therefore  does not
      necessarily  reflect  the views  oi:  the Agency,  and  no  official
      endorsement  should  be  infer-real .  Mention  of  trade  names  or
      commercial   products   does   not   constitute   endorsement   or
      recommendation  for  use.
                                      n-176
-------
78   OFF-LINE  SUPERCRITICAL KI..U1I) KXTRACTION TECHNT.QUK KOK
    DIFFICULT ENVIRONMENTAL MATH.ICKS  CONTAMINATED WVTH  COMPOUNDS OF
    ENVIRONMENTAL SIGHTFTCANOK

    V.  Lopes-Avilci  and  N.S.  Dodhiwala,  Mid-Pacific  Environmental
    Laboratory,   Mountain   View,   California,   and   W-f.   Deckert,
    Environmental Monitoring Systems  Laboratory, U.S. EPA, Las Vegas,
    Nevada.


    Supercritical  fluids  have  unique   properties   (thair  solvent
    strengths approach those of 1 iquids, their viscosities are low, and
    solute diffusivities are  hi.qher  in supercritical  fluids  than in
     liquid solvents)  which make them  very suitable for use in sample
    preparation.  This,  and the availability  of  both  off-line and on-
     line   equipment   for  supercritical   fluid   extraction   (SFE),.
    constituted  the driving  force.- behind  the evaluation  of  the SFE
     technique as an alternative  .".ample preparation  technique  to tho
     time-consuming  SoxhLot  extraction   or   the  very  nonsclective
     sonication extraction.  We  are  in Uio process  ol: evaluating the SFE
     technique with difficult  matrices (primarily  standard reference
    materials such as marine sediment, incinerator fly ash,  coal tar
    contaminated soil)  and many cl oases oL' compounds of environmental
    significance  (  e.g.,   orqanophosphorus   pesticides,   phenolic
    compounds,   chlorinated borviones,  n.i troaromatic  compounds,  and
    haloethers).  The effee Us  of pressure, temperature,  sample size,
    analyte  concentration,  addition of: modifier  to the matrix or the
    carbon dioxide will  be  discussed. The. SFE  equipment that allows
    either single or multiple, extractions  to be performed  in unattended
    operation  was  made  available   to  this   study   by   .several
    manufacturers.
     NOTICE:  Although the research described in this abstract has been
     supported by the United States Environmental Protection Agency, it
     has  not been  subjected  to Agency  review  and therefore  does not
     necessarily  reflect  the  views  of  the:  Agency,  and  no  official
     endorsement  should  be  inferred.    Mention   of:  trade  names  or
     commercial   products   does   not:   constitute   endorsement   or
     recommendation for  use.
                                   n-i??
-------
   79
A RETENTION INDEX SYSTEM FOR IMPROVING THE RELIABILITY OF GC/MS TENTATIVE IDENTIFICATION

                            M. P. MASKARINEC, S. H. HARMON, AND G. S. FLEMING
                                     OAK RIDGE  NATIONAL LABORATORY

                 In many, if not all, samples analyzed by GC/MS for volatile and semivolatile organic compounds,

         significant concentrations of tentatively identified compounds (TICs) are found. A very large percentage

         (>90%) of these TICs are reported as unknowns. When decisions are made on remedial action steps or

         cleanup standards,  it is becoming increasingly more important to have reliable identification of non-target

         compounds.  Therefore, we have developed a retention  index system which can be searched using existing

         GC/MS data  systems to  provide additional information on the TICs.  The advantages of a  retention index

         are  the  ability  to  transfer  data  between laboratories,  as long as the stationary phase is identical, the

         independence of the retention index and the mass spectrum, and the relative ease  of computer searching

         for matches.

                 The retention indices were developed as follows.  A series of normal hydrocarbons were analyzed

         along with the internal standards appropriate for the method. The retention index of each internal standard

         was  calculated using the hydrocarbons  in the standard  fashion.  The hydrocarbon spectra were the first

         additions to the new library.  The internal standards were then used to calculate the retention index of non-

         target analytes. The mass spectra were added to a user-created mass spectral library, while the retention

         indices were added to a user-created retention library.  Therefore, when TICs are found in a sample, they

         can  be searched through both libraries, and a more positive identification can be made.

                 To date, over 100 compounds have been entered into the  new libraries. There are three criteria

         for selection of compounds to be added.  First, compounds which appear on regulatory lists other than the

         Contract Laboratory Program Target Compound  List were added,  e.g. Appendix IX compounds. Second,

         compounds which have had a high frequency  of occurrence in samples analyzed in the Contract Laboratory

         Program program  (as provided  by the database at the  Sample Management Office) were added.  Third,

         compounds which are assumed to be chromatographable under the standard method conditions were added.

         We  are continuing  to add compounds to the library at  this  time.
                                                        n-178
-------
Refer to paper  number 43, Vol. 1 - 297
                                     n-i79
-------
81         ON-LINE  SUPERCRITICAL  FLUID  EXTRACTION/GAS  CHROMATOGRAPHIC  (SFE/GC)
           METHOD SUITABLE FOR USE WITH MODIFIED CARBON DIOXIDE

           James H. Raymer, Ph.D., Linda S. Sheldon, Ph.D.,  and  George  R.  Velez

           Analytical and  Chemical  Sciences,  Research  Triangle   Institute, 3040
           Cornwall-is Road, Research Triangle Park, North Carolina   27709


           ABSTRACT

           An SFE/GC interface was constructed and  studied  both  with and  without an
           intermediate Tenax-GC  adsorption  step.    Semivolatile  pesticides and
           herbicides were used as  test  analytes.    The  addition of  the Tenax-GC
           step allowed for  larger  extraction  volumes  than   were possible  using
           SFE/GC  with  analyte  transfer  directly   into   the   GC  column.    The
           intermediate trapping step also  improved the chromatographic  efficiency
           relative to direct SFE/GC.    Replicate  analyses indicated variabilities
           less than 3% relative standard deviation.   This system  should allow for
           on-line analysis of extracts  obtained   using extraction fluids modified
           with  polar  solvents.    Experiments  to   test   this with  C02/methanol
           mixtures are in progress.


           INTRODUCTION

           In recent years, chemists have   begun  to   investigate the ways in  which
           supercritical fluids can be exploited to simplify and improve  analytical
           methods.    A   number   of  researchers   has published  on  the  use  of
           supercritical fluids,  such as C02  and  N20, to extract a wide  variety of
           organic compounds from  several different matrices.    Through  the use of
           SFE, pesticides, polynuclear  aromatic compounds (PNAs), polychlorinated
           biphenyls  (PCBs), polychlorinated  dibenzo-p-dioxins, dibenzofurans, and
           phenolic compounds have been recovered from matrices  such as soil  (1,2),
           urban dust  (3,4), sediment   (5),   fly  ash   (6), and  polymeric materials
           such as Tenax-GC (7),  polyimides  (8), and polyurethane foam (9).

           The  main  benefits  of   SFE  arise  from  the density-dependent solvation
           capability  that allows  some  degree  of  extraction selectivity, the ease
           with which  the  extraction  solvent can  be removed, and the  ability to
           minimize  the use  of   organic   solvents, such  as dichloromethane,  which
           themselves  contribute  to the current  waste and pollution problems.  In
           addition,   supercritical  fluids   have   lower   viscosities  and  higher
           diffusivitites  than  liquid solvents.     As  a result,  supercritical  fluid
           extractions can be completed  in   far   less time  than  is necessary for
           comparable  Soxhlet extractions.

           Extractions utilizing  supercritical fluids   can  be performed either off-
           line or on-line.   In off-line   methods,  the effluent  from the  extraction
           cell  is expanded into  a  small volume of  receiving solvent.  This results
                                            n-iso
-------
in extracts that are analyzed  in  the  same manner as extracts obtained
from Soxhlet extractions.   In  on-line  methods, the entire effluent is
directed into the chromatographic system (GC or HPLC) where the analytes
are refocused  prior  to  separation.    On-line  methods  offer several
advantages.    First,  better  method  limits  of  detection  (LODs) are
possible because all of the extract  is analyzed.  Thus, LODs achievable
with current methods are possible  using on-line SFE methods and smaller
samples.   Second,  the  chances  of  sample  contamination  are greatly
reduced because sample handling is  minimized.  Third, sample throughput
can be increased because extraction and analysis occur in the same step.
On-line methods requiring less than one  hour for both extraction and GC
analysis have been reported (4).

All of  the  directly  coupled  SFE/GC  methods  reported  to  date have
utilized pure C02 or N20 as  the  extraction fluid.  There are, however,
many instances where C02 alone is  inadequate to recover the analytes of
interest (1, 10, 11).   Although  this  might  be  due, in part, to poor
solubility of the analyte in the  fluid, there is sufficient evidence to
conclude that added modifiers  serve  to  displace solutes from sites on
the surface of the matrix   (12).     In  these cases, an organic compound
such as methanol is added to the sample in order to  improve the recovery
of the analyte.  Unfortunately, the direct  coupling of SFE to GC is not
straightforward when  modified  extraction  fluids   are  used.  Methanol
added to the extraction  fluid  will  accumulate  in  the GC column upon
expansion  of the supercritical fluid and cause poor  chromatography.  For
maximum analytical utility  and  flexibility,  a  coupled SFE/GC needs to
handle pure and modified supercritical fluids.

An additional  constraint   on  directly  coupled  SFE/GC  systems is the
extraction flow rate that can  be  tolerated.    If  the flow^rate is too
fast, poor refocusing of the analytes occurs upon fluid expansion within
the column and poor chromatographic  peak  shapes result (13).   If a flow
rate is used that results in  good   peak shape,  very long times might be
required to completely extract the analytes from the sample.  This would
be true, for example, if low levels  of analyte were  to be recovered from
a sample volume  of  several  milliliters.    Such   a situation could be
envisioned in the  SFE/GC   analysis  of  organic compounds from sorbents
 (Tenax-GC, XAD, PUF) used for  air   sampling.    Therefore, the research
described  in this paper  has been  directed towards the development  of an
SFE/GC  system  that  would provide  flexibility  with  regard  to both
extraction flow rate or  time and extraction fluid composition.

Earlier research  in  our  laboratory  indicated that semivolatile  compounds
could be recovered  easily   from  Tenax-GC using  relatively  small volumes
of pure supercritical C02   (7).     Because semivolatile compounds  should
have very  large retention volumes on  a Tenax-GC cartridge  at  ambient or
near-ambient conditions,  this   sorbent  could   be   used  to  retain  such
analytes after  expansion of the  supercritical  effluent from  an SFE cell
containing the  sample of interest.   The analytes would be  well-retained
on the  Tenax-GC   independent   of  extraction   flow   rate,   time or fluid
                                   n-isi
-------
composition.   This would  provide  a  great  deal  of flexibility  in the
extraction conditions.   The  analytes  could  then be recovered  from the
Tenax-GC using supercritical C02  and  the effluent from this  extraction
could be introduced into a GC column.   If only small volumes  of C02 are
needed  for  the  recovery  of   analytes  from  the  Tenax-GC,  optimal
chromatographic performance can be realized  in  this step  by  using slow
SFE flow rates over a  relatively  short time.  In addition, co-solvents
such as  methanol  will  not  be  well  retained  by  the Tenax-GC (14).
Consequently,   such   compounds   could   be   incorporated    into  the
supercritical fluid used to  extract  the  sample.   Upon expansion, the
analytes would be retained by  the Tenax-GC and the vapor-phase  modifier
would be poorly retained.    The  majority  of the solvent  would quickly
break through the cartridge and pass out  of the system.    The remaining
modifier would be swept from the sorbent with either C02 or helium prior
to the transfer of the analytes from the Tenax-GC to the GC column.  The
volume and dlow rate of  supercritical  CC>2  in  this last  step  would be
such that good chromatography would result.
 EXPERIMENTAL

 The  SFE/GC  interface developed during  this  work   is  shown  in  Figure  1.
 It is  shown  as  configured  for  analyte  deposition   onto  Tenax-GC for
 subsequent  extraction and  deposition  onto  the GC  column  (SFE/SFE/GC).
 For  direct  SFE/GC  operation,  the  extraction  cell   (EC) was  connected
 where  the Tenax-GC cartridge is shown  in  the  lower half of the  Figure.
 In either configuration, a  Valco  HPLC  injector   fitted  with a 500  nL
 rotor  (not  shown) was placed in-line and before the  0.41 ml  SFE cell (EC
 in Figure 1) which  was  filled  with  sea  sand  and   held  at  50*C in  a
 modified Lee Scientific Model 501 SFE/SFC system.   The valve allowed for
 the  reproducible  introduction  of  a  methanol  solution  of pesticides
 (approximately 100  ng  each)  into  the  pressurized   cell   so that SFE
 conditions  could  be  mimicked.    The  pesticides   used  were  molinate,
 propoxur, atrazine,  7-BHC,  trial late,  terbutyrn,  ethyl parathion,  7-
 chlordane,  and phosmet.  The outlet of the SFE  cell  was directed  through
 a multiport switching  valve  to  an  11  cm  x  15  pm id  fused silica
 restrictor  for deposition onto the Tenax-GC   (Ri) or the GC  column (R2).
 Restrictors RI and R2 had the same dimensions.

 In the SFE/SFE/GC configuration, the  system  was pressurized to  400 atm
 (C02 density = 0.928 g/cm3)  with  valves  1  and 3  open.  The  pesticide
 solution was introduced  to  the  extraction  cell,  migrated through the
 cell and expanded through RI onto  the  head  of a  column comprised of a
 steel  tube  (6 cm x 4 mm  id) with fritted column end-fittings and packed
 with 0.14 g of Tenax-GC.  The Tenax cell was  also held at 50'C.  Gaseous
 C02  passed  through valve  #3  to  vent.    The  C02 flow measured at the
 outlet of valve #3 was 81  mL/min.   The  length of  this first extraction
 (Ei) was varied  from  7  to  30  minutes  to   study possible changes  in
 recovery.   At the end of EI, valve #1 was closed and valves  5 and 7 were
 opened.  This introduced a flow  of  helium through a  "T" union near the
                                   n-182
-------
end of RI, past  the  end  of  RI  and  across the Tenax-GC cartridge to
remove any residual methanol,  if used.   After this purge, the multiport
switching valve was moved to its other position, valves 5, 7, and 3 were
closed, and valves 2 and 6  were  opened.   In this manner, the Tenax-GC
cartridge was back extracted with supercritical C02 and the effluent was
expanded through R2, which passed through  a 1/16" cross union, onto the
first few cm of the GC column (30  m x 0.32 mm id DB-5, J&W Scientific).
Some of the C02 gas passed  through  the  column and the rest was vented
through valve #6 (vent 2).  The  time of this second extraction (£2) was
varied between 4 and  15  minutes.    The column temperature during this
extraction was maintained at 20*C.   After £2 was complete, valve #2 was
closed, valve #3 was  opened  to  release  the  pressure in the Tenax-GC
cartridge, and valve  #5  was  opened  to  introduce  helium into the GC
column and to  sweep  residual  C02  from  the  cross  union.  After one
minute, valve #6 was closed to  purge  C02  from the column and the oven
temperature was raised to  180*C  for  1  minute  and then programmed to
300*C at 5 degrees  per  minute.    Flame ionization detection (FID) was
used.  During E2, the extraction cell could be cleaned by opening valves
1  and 4.

In SFE/GC operation,  the  supercritical  effluent  from the extractiuon
cell was expanded through R2 as in SFE/SFE/GC operation.  Because RI and
R2 were of the same dimensions,  the  flows were assumed to be the same.
The  flows were measured periodically  to  be  sure they had not changed.
Times of 7,  10, and  15  minutes  for this extraction step were studied.
Area ratios  relative to  7-BHC  were  used to gauge relative recoveries.
Separation   and  detection   were   performed   as  for  the  SFE/SFE/GC
experiment.  All data  were  collected  using a Nelson Analytical (model
4400, version 7.2)  data system.

For  comparison, the pesticide mixture was analyzed using 30 s splitless/
split  injection on  the same GC column.
 RESULTS

 The effect  of EI  extraction   time   on   the   areas  measured  in  the  SFE/GC
 experiment  are  shown   in  Table   1.     Each  entry   is   the   average  of
 triplicate  analyses.   As the   time of  extraction  increased, the  measured
 areas  decreased.   This was   especially  true   for terbutryn.  Since the
 extraction  flow rates  were the same for each time  period, the  losses are
 the result  of total  flow.   That   is,   it appears  as  if  the analytes are
 deposited into the column  then lost  from  the system.   A lower  trapping
 temperature might improve  this situation  but   was not tried  because the
 goal  here was to  determine   if  the  presence of  the secondary  Tenax-GC
 trapping step could minimize  the  problems  associated with longer  EI
 times.

 In the study of the time  of  E2   in   the  SFE/SFE/GC experiment,  the  GC
 peaks  became broader as the time  was progressively increased  from  4 to 7
                                 n-183
-------
to 11 and to 15 minutes.   A  carryover of 20% was observed for BHC at an
EZ time of 4  minutes  so  an  £2  time  of  7  minutes was used for the
remainder of this work.

Using an £2 time of  7  minutes,  the  EI  time had little effect on the
areas measured when it  was  varied  between  7 and 30 minutes, although
better recoveries were observed for phosmet using a 15 minute extraction
time.  These results indicate that the analytes do not break through the
Tenax-GC column, at least up to times  of 30 minutes.  This allows for a
great deal of flexibility in the selection of the EI time as long as the
analytes of interest are well-retained  by  the Tenax-GC.  For compounds
of relatively low volatility, this should not be a problem.

The SFE/SFE/GC chromatogram obtained using an  EI time of 15 minutes and
an £2 time of 7 minutes is shown in Figure 2.  A gas chromatogram of the
test compounds obtained using  conventional splitless/split injection is
shown in Figure 3 for comparative purposes.  Although some broadening of
the chromatographic peaks is seen  in  the case represented in Figure 2,
relative to Figure 3, the  efficiencies of the SFE/SFE/GC separation are
certainly adequate for most purposes  and  could probably be improved by
careful optimization of  flow  rate  and  column  temperature during E2-
Replicate analyses using SFE/SFE/GC provided area precision of less than
3% RSD.

Table 2 shows a comparison  of  the  chromatographic peak area to height
ratios  for each of the  analytes  for  both SFE/GC and SFE/SFE/GC.  This
ratio is  indicative  of  chromatographic  efficiency,  with  a low ratio
reflecting a higher efficiency.    The  extraction times for both SFE/GC
(Ei) and  SFE/SFE/GC  (£2)  were  7  minutes  at  the same extraction flow
rate, and thus  differences  reflect  the  effect  of the Tenax-GC on the
system  performance.  It can be  seen  that  in all cases the presence of
the Tenax-GC resulted in sharper peaks.    This is presumably due to the
accumulation of the  analytes  at  the  head  of the Tenax-GC cartridge.
When the  Tenax-GC  is  back  extracted,  the analytes are introduced into
the column as a tight  band.    In  SFE/GC, the analytes were introduced
into the  column over a  longer  time  because of the band spreading that
occurred  during the migration through the  sand. Such an effect would be
expected  to be  more pronounced  in an actual extraction where desorption
of the  analytes from the matrix might take place over a longer time with
the  result that they  would  be  spread  out  even  more.   The use of a
secondary sorbent  can minimize this effect.

A study of  the  performance  of  the  SFE/SFE/GC  system when methanol-
modified  C02 is used as the extraction fluid  is currently  in progress.

SUMMARY

The  use of the  secondary sorbent  Tenax-GC  in  an on-line SFE/GC  analysis
scheme  (SFE/SFE/GC) can provide greater flexibility  in the  extraction  of
the  sample than can direct  SFE/GC  for analysis of  semivolatile  organic
                                   II-184
-------
compounds.  The use of on-line SFE/GC methods has the potential to allow
for the collection  of  smaller  samples  (smaller air sampling volumes)
and, because the entire  extract  is  analyzed,  to  lower the limits of
detection of the analytical method.

ACKNOWLEDGEMENT

This work was funded by  an  Independent Research and Development  (IR&D)
grant from Research Triangle Institute.


REFERENCES

 1.  J.R.  Wheeler,  M.E.  McNally,   "Supercritical  Fluid   Extraction  and
     Chromatography of Representative  Agricultural  Products with  Capillary
     and Microbore Columns," J. Chromatog. Sci., 2_7, 534  (1989).

 2.  B.W.  Wright,  C.W.   Wright,    J.S.  Fruchten,  "Supercritical  Fluid
     Extraction of Coal Tar Contaminated Soil  Samples," Energy and Fuels  3,
     474  (1989).

 3.  J.M. Levy,   R.A.  Cavalier,  T.N.  Bosch,   A.F.  Rynaski,  W.E.  Huhak,
     "Multidimensional Supercritical  Fluid Chromatography and Supercritical
     Fluid Extraction," J. Chromatogr. Sci.,  2J_, 341  (1989).

 4.  S.B. Hawthorne, D.J.  Miller,   M.S.  Krieger,   "Coupled  SFE-GC: A Rapid
     and  Sample   Technique for   Extracting,   Identifying, and Quantitating
     Organic Analytes  from Solids and Sorbent Resins," J. Chromatogr. Sci.,
     27,  347  (1989).

 5.  F.I. Onuska,  K.A. Terry,  "Supercritical  Fluid  Extraction of  2,3,7,8-
     Tetrachlorodibenzo-p-dioxin  from Sediment  Samples,"   J.  High  Resol.
     Chromatogr.  12, 357  (1989).

 6.  N. Alexandrou, J. Pawliszyn,   "Supercritical   Fluid  Extraction for the
     Rapid   determination  of    Polychlorinated    Dibenzo-p-dioxins  and
     Dibenzofurans in  Municipal  Incinerator   Fly Ash," Anal. Chem, 61.  2770
     (1989).

 7.  J.H. Raymer  and   E.D.  Pellizzari,  "Toxic  Organic Compound  Recoveries
     from 2,6-Diphenyl-p-phenylene  Oxide Porous  Polymer Using Supercritical
     Carbon Dioxide and Thermal  Desorption Methods,"  Anal.  Chem., 59,  1043
     (1987).

 8.  J.H. Raymer,  E.D.  Pellizzari,  S.D. Cooper,"Desorption  Characteristics
     of Four Polymide  Sorbent  Materials Using Supercritical  Carbon Dioxide
     and  Thermal  Methods,"  Anal. Chem., 59,  2069  (1987).
                                   n-185
-------
 9.   S.B.   Hawthorne,   M.S.   Kriegen,  D.J.  Miller,   "Supercritical  Carbon
     Dioxide Extraction of  Polychlorinated  Biphenyls,  Polycyclic  Aromatic
     Hydrocarbons,  Heteroatom-Containing  Polycyclic Aromatic  Hydrocarbons,
     and n-Alkanes  from Polyurethane Foam  Sorbents,"    Anal.  Chem. 61., 736
     (1989).

10.   S.B.  Hawthorne and D.J.   Miller,  "Extraction  and  Recovery  of Organic
     Pollutants  from Environmental  Solids  and Tenax-GC  Using  Supercritical
     C02,"  J.  Chromatogr.  Sci. 24,  258-264 (1986).

11.   V.   Lopez-Avila,   W.F.   Beckert,  S.  Billets,  "The  Why and   How of
     Supercritical  Fluid Extraction  and  its  Application to  Environmental
     Analysis,"  Proceedings  of  the  5th  Annual  Water  Testing and Quality
     Assurance Symposium,  Washington, D.C., pII-73, July 1989.

12.   S.B.  Hawthorne, personal  communication.

13.   S.B.  Hawthorne,  D.J.  Miller,  "Directly  Coupled  Supercritical Fluid
     Extraction   -   Gas  Chromatographic  Analysis  of   Polycyclic  Aromatic
     Hydrocarbons and Polychlorinated Biphenyls from Environmental  Solids "
     J.  Chromatogr. 403, 63-76 (1987).

14.   K.J.  Krost,  E.D.  Pellizzari,  S.G.  Walburn, S.A. Hubbard, "Collection
     and Analysis of Hazardous Organic  Emissions,  "Anal. Chem. 54(4), 810-
     817 (1982).                                                 ~~
                                   n-186
-------
  TABLE  1.   EFFECT  OF  FIRST  EXTRACTION  TIME (El)  ON RECOVERIES IN SFE/GC
                             Chromatographic Peak Areas


Compounds
Molinate
Propoxur
Atrazine
7-BHC
Trial! ate
Terbutyrn
Ethyl parathion
7-Chlordane
Phosmet
EI = 7 min
Average
Area (n=3)
149527
123733
84996
73335
107195
31662
97223
92127
62853
EI = 10 min
Average
Area (n=3)
142925
117224
68710
70537
103578
12248
93410
87681
60652
EI = 15 min
Average
Area (n=3)
123094
101087
66362
60252
89790
17006
81241
77910
51310
         TABLE 2.  CHROMATOGRAPHIC EFFICENCY OF PESTICIDE SEPARATION
                     AS REFLECTED BY AREA/HEIGHT RATIOS
Compound
SFE/GCa (%RSD)
SFE/SFE/GCb (% RSD)
Molinate
Propoxur
Atrazine
7-BHC
Trial! ate
Terbutyrn
Ethyl parathion
7-Chlordane
Phosmet
7.4 (3.9)
7.6 (4.7)
8.2 (1.2)
8.4 (3.8)
7.4 (5.6)
8.5 (4.2)
8.3 (5.7)
8.9 (4.3)
12.8 (3.7)
5.4 (4.9)
5.4 (7.5)
5.8 (6.5)
6.2 (7.4)
5.7 (6.3)
5.8 (1.0)
6.0 (5.8)
6.3 (6.0)
8.8 (1.3)
a Extraction time of 7 minutes; triplicate analysis.
b Extraction of sand for 15 minutes followed by 7 minute extraction of
  Tenax-GC; triplicate analysis.
                                n-187
-------
A.
                           .	SFEolE-C with exp»n»Ion onto Tt««
                           _	i*n«x purge wtth holkim

                                    Extraction Oven	
                              r,             _,
                                                            Vent 2
                            	SFE of Tenax with expansion onto
                                   GC column
                            		Helium flow (or GC separation
                            	Extraction cell (E. C.) purge or cleanup
                                   option
                                    Extraction Oven
                                  Chromatographic Oven	|
         Figure  1.   SFE/SFE/GC  interface shown with flow paths  for A)
                    transfer of  extracted components from extraction
                    cell  to Tenax-GC,  and B)  transfer of  analytes from
                    Tenax-GC to  capillary GC  column.
                                     n-iss
-------
H
1-^
OO
          Temperature (°C)  30
                 Time (min)   0 -
                                Figure  2
SFE/SFE/GC chromalograin  of  pesticide  test invxture.  Compound Identi-
fications are (1)  molinate,  (2)  propoxur, (3) atrazine.  (4) v-BHC.
(5) tridllate.  (6)  terbutryn,  (7)  ethyl parathion.  (8) v-chlordane.
(9) phosmet.  Conditions  as  described  in the  te-'t
-------
UL
                           fjtji
Temperature (°C)  35
        70
120
170
220
270
                                                                                 300
Time (min)  0
                         10
                    20
            30
             40
             50
             60
  F-iqure 3.  GC chrornatogram of the pesticide test mixture after- sp 1 i t less/sp 1i t
            injection.  The Targe peak before molinate  is either an impurity or
            ti-.ermal decomposition product of propoxur.
-------
82       Refer  to paper number 55, Vol. II -  32




g3       Refer  to paper number 42, Vol. I - 286
                                            n-i9i
-------
84                    MULTILABORATORY VALIDATION STUDY OF PCBs IN SOILS
                       USING SOXTEC® EXTRACTION TECHNIQUE (METHOD 3541)
           Charles K. Bayne, Computing & Telecommunication Division; Joseph H. Stewart, Jr., Analytical
           Chemistry Division,  Oak Ridge National Laboratory,  Bldg. 6011 MS 6370,  P.O.  Box 2008, Oak
           Ridge, Tennessee  37909-6370.
          The submitted manuscript has been authored by a contractor of the U.S.  Government under
          contract No.  DE-AC05-840R21400.  Accordingly, the U.S.  Government retains  a nonexclusive
          royalty-free license to publish or reproduce the published form of this contribution, or allow others
          to do so, for U.S. Government purposes.


          Research sponsored by Office of Energy Research, U.S. Department of Energy under contract DE-
          AC05-840R21400 with  Martin  Marietta Energy Systems, Inc.
                                                   n-i92
-------
            MULTILABORATORY VALIDATION STUDY OF PCBs IN SOILS
             USING SOXTEC® EXTRACTION TECHNIQUE (METHOD 3541)
Charles K. Bayne. Computing & Telecommunication Division; Joseph H.  Stewart, Jr., Analytical
Chemistry Division, Oak Ridge National Laboratory, Bldg.  6011 MS 6370, P.O.  Box 2008, Oak
Ridge, Tennessee 37909-6370.
ABSTRACT
The Oak Ridge National Laboratory has evaluated a multilaboratory Soxtec® PCB extraction study
and has compiled the data from the eight participating laboratories. The experiment required each
laboratory to extract PCB  aroclor 1254 and 1260  from  10-g samples of Fuller's earth on 3 non-
consecutive days. Each sample was to be spiked  at a concentration level of 5 ppm or 50 ppm.
The average for  all PCB percent recovery values is 88%, with a 95% confidence interval of (82%,
93%).  The estimated standard deviation for a single measurement within a laboratory is 19%.
INTRODUCTION
In  1987  the  Oak Ridge National  Laboratory  (ORNL)  submitted  laboratory data  to  the
Environmental Protection Agency (EPA) supporting an alternative method to EPA 3540 and EPA
3550 for the rapid and quantitative extraction of polychlorinated biphenyls  (PCBs) in soils. The
new procedure for PCB extraction was developed for Tecator's Soxtec® System HT and reduced the
extraction time from  16-17 hours to  2  hours.  The laboratory data  for the Soxtec® extraction
procedure showed that about 82.0% of PCBs can be recovered in soil and clay for a concentration
range of 5 to 50  ppm.  The recovery  percentages may vary with concentration  level and day-to-
day operations.  For a single PCB extraction, the recovery percentage will vary about 13.0%, with
a 95% confidence interval on the variance of (9.93)2 <_ Variance <_ (16.29)2.
Subsequently, EPA gave  ORNL a single-site  approval to use the Soxtec® extraction procedure.
EPA indicated that a blanket approval would be given for all laboratories if at least six additional
facilities  could  successfully  use  the  detailed  procedure and  obtain PCB recovery percentages
equivalent to the existing  SW846 Method 3540. EPA said that  this validation could be done more
simply than  a formal petition for each site.


In 1988 ORNL sent EPA a statistically designed experiment to demonstrate the capability of a
laboratory to use  the  Soxtec®  extraction procedure.   This designed experiment required  each
laboratory to extract PCB aroclors 1254 and 1260 from 10-g samples of Fuller's earth on 3 non-
consecutive days.   Each sample was to be spiked at a concentration level of 5  ppm or 50 ppm.
The 12 samples extracted were then sent to  ORNL  for analysis by gas chromatography procedure
EPA Method in SW846.   The  analytical  data from the multilaboratory experiment were  sent to
EPA for  comparison with recovery percentages  equivalent  to  the existing SW846 Method 3540.
EPA has approved  the new Soxtec® extraction method and has integrated it into SW846 as Method
3541.
                                          n-193
-------
MATF.RTAT.S AND METHODS
Method 3541 was developed to extract PCBs [1] from soil, sediment, sludges, and waste solids.  The
method uses a commercially available, unique three-stage extraction system to achieve comparable
PCB recovery with  method  3540 but in a much shorter time.  The two differences between this
proposed  method and Method 3540 are stages (1) and (3).  In the initial  extraction stage, the
specimen-loaded extraction thimble is immersed into the boiling solvent.  This  stage ensures very
rapid intimate contact between the specimen and  solvent, with subsequent rapid recovery of the
PCB.  In  the second stage,  the thimble is elevated above the solvent and  is  rinse-extracted as in
Method 3540. In the third stage, the solvent is evaporated, as would occur in the Kuderna/Danish
(K/D) step in Method 3540.  The concentrated specimen is then ready for measurement of the PCB
concentrations, as in SW 846, Method  8080.
After air-drying  of the specimens, following EPA Method 600/4-81-055, interim methods for the
Sampling and Analysis of Priority Pollutants in Sediments and Fish Tissue (Step 3.1.3), the specimen
is  ground to  100-200 mesh (840 jum) using a Fisher Cyclotec 1093 grinder (or equivalent). The
powdered specimen is extracted using a 1:1 acetone/hexane mixture as the extraction solvent. The
extract is then concentrated and exchanged into pure hexane prior to final gas chromatographic PCB
measurement.
A Tecator's Soxtec* System HT extraction system with controlled heated oil bath and multiples of
six extraction modules was  used for this  work.   The  apparatus is  used  in  a hood.   PCB
contamination-free cellulose extraction thimbles (Fisher catalog 1522-0018 or equivalent) are used.
If the sample does not pass through  a 1-mm standard sieve or cannot be  extruded through a 1-
mm  opening, it is processed into a homogeneous sample  that meets these requirements.  Fisher
Cyclotec Model 1093, Fisher Mortar Model 155 Grinder, Fisher scientific Co., Catalog Number 8-
323,  or equivalent brands and models are recommended for sample processing.  These grinders will
handle most solid samples, except gummy, fibrous, or oily materials.  These types of specimens may
be mixed with anhydrous sodium sulfate  to improve grinding efficiency.


Sediment/soil samples were processed  as follows.  Any water layer is decanted and discarded. The
sample is thoroughly mixed, especially composite samples.  Any foreign objects, such as  sticks,
leaves, and rocks, are discarded.  The  specimen  is air-dried at room  temperature for 48 hours in a
glass tray or on hexane-cleaned aluminum foil.
A sufficient amount of dried specimen is ground to yield 20 g of powder, using the Cyclotec 1093
sample mill. The mesh size is typically 100-200 mesh (840 jUm).  Ten grams of specimen is weighed
into the extraction  thimbles.   The  thimbles  are placed  in  the device with  50 mL  solvent
(acetone/hexane, 1/1).


The extraction is carried out by  immersing the  sample thimble in  boiling solvent for 60 minutes.
The sample is  then raised above  the solvent and rinsed for an additional 60 minutes.  Finally, the
solvent is evaporated, and the extract volume is reduced.  The evaporated solvent is collected, and
the extraction  is stopped when  the  desired  concentration factor  (1-2  mL remaining  extract)  is
                                          n-194
-------
collected.  The extract can be analyzed directly, carried through a cleanup procedure, or solvent
exchanged, depending on the requirements of the measurement method.


For the analysis of PCBs, the extracted and hexane-exchanged specimens are prepared for Meihod
8080, Determination of Potychlorinated Biphenyls (PCB) by Gas Chromatography. If further cleanup
is necessary, the Florisil® and/or sulfur procedures may be employed.  In such cases Method 3620
is carried  out, followed by  Method 3660, if necessary, using the 10-mL  hexane extracts obtained.
Following cleanup, the extracts are analyzed by electron capture GC, as  described in the previous
paragraphs and in Method  8080.
MULTILABORATORY EXPERIMENT
Eight laboratories agreed to participate in a multilaboratory experiment to verify Method 3541 for
the PCB extraction procedure developed for Tecator's Soxtec* System HT.  The experiment for
each laboratory consisted of 12 samples of Fuller's earth spiked with either 5 ppm or 50 ppm and
either aroclors of 1254 or 1260.  The laboratories were responsible for spiking the samples. These
samples were to be extracted  by the proposed PCB extraction procedure over 3  nonconsecutive
days. The experimental factors and their levels are listed in Table 1.
                      Table 1. Experimental factors and their levels for
                       the PCB extraction multilaboratory experiment.
Factors
1. Aroclor
2. PCB Level in Fuller's earth (ppm)
3. Days

Low
1254
5
Day 1
Levels
Middle
Day 2

High
1260
50
Day 3
Each laboratory was to extract 4 Fuller's earth 10-g samples per day for 3 nonconsecutive days for
a total of 12 experimental runs. The experimental design randomly assigned the order of extracting
the PCB-spiked samples for each day. All 4 daily samples can be extracted concurrently by a multi-
unit Tecator's Soxtec® Systems (e.g., HT6, HT12).  For each day, the experimental design is a two-
level factorial (22) for the aroclor- and spike-level factors. The percent recovery was evaluated for
each sample and calculated by:
             Percent Recovery = 100% x (mg of PCB found)/(mg of PCB spiked).
Slight  deviations  from  the  planned experiment  occurred  which  influenced the results.   Two
laboratories, No. 2 and No. 5, didn't interpret our spiking instructions correctly and used 50 and
500 ppm, rather than 5 and 50 ppm.  Replicate samples for each extraction were sent to ORNL by
laboratories No. 2 and No. 4.  Replicate samples are one extraction  that was  divided into two vials.
Laboratory No. 6 repeated two conditions (day = 2, PCB  level = 50 ppm, aroclor = 1260 and
                                         H-195
-------
day = 3 pCB level = 50 ppm, aroclor = 1254) but didn't do two other conditions (day = 2, PCB
level =  50 ppm, aroclor - 1254, and day  = 3, PCB level  = 5 ppm, aroclor =  1254).


The PCB extracts were sent to ORNL for chemical analysis after the completion of the extraction
procedure.   A single Analytical Chemistry Division  staff member  analyzed the entire series of
submitted  PCB extracts, using a single electron-capture gas chromatograph to minimize personnel
and equipment variability.


RESULTS AND  DISCUSSION
The PCB  percent recovery  (%) results for the multilaboratory experiment are averaged over the
3 days and presented in Table 2.  The PCB percent recovery averaged  over all laboratories and
factor levels is 88%, with a  95%  confidence interval of (82%,  93%). The sources of experimental
variation were investigated  using the  method of analysis of variance  (ANOVA) [2].  Only data
from the five laboratories were included that followed the planned experiment  (No. 1, No. 3, No.
4, No. 7, and No. 8).  The average of  the replicate samples for laboratory No.  4 was used for the
ANOVA.   The estimated variance from different  sources is measured by the mean square. The
mean squares  are illustrated  in  Fig. 1 for the following sources  of variations: (Lab)  differences
among laboratories,  (Arcolor) differences between aroclor  types within each laboratory, (Level)
differences between spike levels  within each laboratory,  (A X L)  interaction differences between
aroclor types  and  spike  levels  within each  laboratory,  and  (Error)  experimental error due to
differences among days.
                                    ANOVA of % Recovery
                      5500
                              Lab     Aroclor     Level      A X L
                                        Variation Source
Error
           Fig.  1.  ANOVA mean squares for percent recovery sources of variation.
                                          H-196
-------
             Table 2.  Summary statistics for PCB percent recovery (%) extracted
                               by Tecator's Soxtec® System HT.
Aroclor


1254
PCB Level
Laboratory Statistics
No. 1 Number
Average
St Dev
No. 2 Number
Average
St Dev
No. 3 Number
Average
St Dev
No. 4 Number
Average
St Dev
No. 5 Number
Average
St Dev
No. 6 Number
Average
St Dev
No. 7 Number
Average
St Dev
No. 8 Number
Average
St Dev
All Laboratories Number
Average
St Dev
5
3
101
35


3
73
11
6
113
18


2
141
4
3
100
18
3
65
16
20
99
29
50 500
3
74
42
6 6
57 67
7 15
3
63
8
6
144
30
3 3
97 80
9 5
3
128
15
3
123
15
3
38
22
30 9
93 71
43 14

1260
PCB Level
5
3
84
7


3
71
3
6
100
13


3
138
15
3
82
8
3
93
37
21
96
25
50 500
3
79
8
6 6
70 57
15 10
3
57
6
6
85
4
3 3
80 77
3 9
4
106
8
3
94
5
3
52
13
31 9
79 75
18 10
All
Levels
12
84
26
24
67
13
12
66
9
24
111
29
12
84
10
12
125
18
12
100
19
12
62
29
120
88
30
The sources of variation are tested to be significantly different from zero by comparing the ratio
of the mean squares for the sources of variation with the mean square for the experimental error
using an  F-statistic.   This F-test  at  the 5%  significance level shows  significant variation  for
laboratories, aroclor type within laboratories, and spike level within laboratories, but not for the
interaction between aroclor  type  and  spike  level within  laboratories.   The estimated standard
deviation for a single measurement within a laboratory is 19%.

ORNL speculates that the major  source of variation of the  percent recovery among different
laboratories is caused  by the  shipping  method  for samples.   The instructions requested that the
                                           H-197
-------
concentrated  extract be  sealed in septum-capped gas chromatographic vials  and sent to  ORNL.
These vials were to contain the 10-mL hexane extracts; however, some vials had less  then 10 mL
when they arrived at ORNL. Hexane was added to those vials to bring the total volume to 10 mL
This volatilization may account for differences in ORNL results and those results measured at other
laboratories.  For example, ORNL found about a 67% recovery rate  for samples from laboratory
No. 2,  but this laboratory has  indicated that they found  about a 90% recovery rate for the  same
samples.  Other shipping methods were also used.  For example, laboratory No. 7 evaporated the
samples to  dryness for shipping, and 10 mL of hexane was added to the samples at  ORNL.  In
retrospect, this shipping  method would be the preferred  method.
NOTES


Soxtec® is a registered trademark of Tecator, Inc., Herndon, Virginia, and is distributed by Fisher
Scientific.
Florisil* is a registered trademark of U.S. Silica Co., Berkeley Springs, West Virginia.


ACKNOWLEDGEMENTS
We would like to express our appreciation to W. F. Rogers for chemically analyzing the submitted
PCB extracts and to R. L.  Holmes for performing PCB extraction studies.
REFERENCES
1.   J. H. Stewart, Jr., C.  K.  Bayne, R.  L. Holmes, W. F. Rogers, and M.  P. Maskarinec (1988).
    Evaluation  of a Rapid Quantitative  Organic  Extraction System  for  Determining  the
    Concentration of PCB in Soils, Proceedings of the United States Environmental Protection Agency
    Symposium on Waste Testing and Quality Assurance, Washington D.C., Volume II, pp G37 - G41.


2.   G. W. Snedecor and W. G. Cochran (1967). Statistical Methods, 6th Edition, Ames: Iowa State
    University Press.
                                          n-i98
-------
INORGANICS
-------
85              PRE-CONCENTRATION TECHNIQUES FOR TRACE METALS

     E.M. Heithmar. T.A. Hinners, U.S. Environmental Protection Agency,
     Environmental  Monitoring Systems Laboratory, P.O.  Box  93478,  Las
     Vegas,  NV  89193-3478.
     J.T.  Rowan, Lockheed  Engineering and  Sciences  Company,  1050  E.
     Flamingo Road,  Las Vegas,  NV  89119,
     J.M. Riviello,  Dionex  Corporation,  1228 Titan Way,  P.O. Box 3603,
     Sunnyvale,  CA   94088-3603.

         Virtually all trace metal  analyses of hazardous  wastes  are

     currently  performed  by   graphite  furnace  atomic   absorption

     spectrometry (GFAAS) or inductively coupled plasma-atomic emission,

     spectrometry (ICP-AES).   Recently, environmental applications  of

     inductively coupled plasma-mass spectrometry (ICP-MS) have appeared

     in  the  literature.     Each  of  these  techniques  has  certain

     limitations when applied to hazardous  waste  analysis.   GFAAS  and

     ICP-MS  provide very low detection  limits,  but they are  prone  to

     serious spectral  and  physical-chemical interferences  in complex

     matrices.   While ICP-AES is more robust, it often does not possess

     adequate detection limits.

         This presentation will demonstrate the  application  of trace

     element pre-concentration  (TEP)  to ICP-MS and ICP-AES,  as well  as

     to  ion chromatography (1C).   The  TEP methods are  based on  a

     commercially available device, with some modifications necessitated

     by the  specific measurement systems.   After  adjusting  the sample

     pH to  about 5.4,  trace metals  are pre-concentrated on  a column

     packed  with a macroporous iminodiacetate-functionalized  resin.  The

     alkali  and alkaline-earth  metals,  as  well as residual concomitant

     anions, such as  sulfate and phosphate,  are removed by flushing the

     column  with ammonium acetate.   The trace metals are then eluted
                                    n-199
-------
with  nitric  acid  directly  to  the   instrumentation  used  for


quantitation.

     TEP-ICP-MS  can be  applied  to  samples  such  as  brines  and

caustics, since  the trace metals are  separated from the  sample

components which cause interferences in conventional ICP-MS.   TEP

improves the  detection  limits of ICP-AES  for nine  analytes by a

factor  of  10  to 100,  depending on  the  volume  of sample pre-

concentrated.  Finally,  both  the sensitivity enhancement and  the

matrix elimination aspects of TEP allow the analysis of  hazardous
                                                                 4
waste  digests by  1C.    TEP-IC  instrumentation  is  rugged  and

potentially field-portable.

     The  analytical performance of  the  various  TEP-hyphenated

techniques   will   be   documented   by   specific   applications.

Limitations of the  methods will be discussed.   Finally, current

work on extending TEP to more analytes  will be described.

     NOTICE:  Although the research described  in  this abstract  has

been supported by the U. S. Environmental Protection Agency,  it  has

not been  subjected to  Agency review,  and,   therefore,  does  not

necessarily  reflect the  views   of  the Agency,  and no official

endorsement should be inferred.
                               n-200
-------
86                       ANALYSIS  OF WATER AND  WASTES  USING  ICP-AES
                                WITH ULTRASONIC NEBULIZATION

       Shi-Kit  Chan,   PhD  and  Cecilia A.  Zechmann,  CETAC  Technologies,  Inc.,   302   S
       36th Street,  Suite 101B,  Omaha,  Nebraska 68131;  Michael  M.  Yanak,  Imaging  and
       Sensing  Technology   Corporation,  Westinghouse  Circle,   Horseheads,  New York
       14845.

       ABSTRACT

       An  ultrasonic nebulizer was  used  to lower  the  detection limits   of  pollutant
       metals  in water and waste  samples.   Procedures  described in  the  USEPA   Method
       200.7  were followed to  demonstrate its  capability  for routine  ICP  analysis.
       Statistical   and  control data showed that  the  ultrasonic  nebulizer  provided
       comparable accuracy  and  better precision than the pneumatic nebulizer.

       INTRODUCTION

       The methods for  metal analysis in  water  and waste samples,  currently  approved
       by  the United States Environmental Protection  Agency (USEPA),  include  flame
       atomic   absorption    spectrometry,    graphite   furnace   atomic   absorption
       spectrometry,  and  inductively coupled   plasma-atomic   emission  spectrometry
       (ICP-AES).  Among the three techniques cited, the ICP-AES method  is  preferred
       because of its capability of  fast,  multielement  analysis.

       Most  commercial  ICP spectrometers utilize pneumatic nebulizers  such  as  the
       concentric, the  cross-flow, and the Babington-type  nebulizers.  In the past 20
       years,  laboratory-built ultrasonic nebulizers  have been used successfully  to
       enhance  the  detecting   power  of ICP-AES  [1].   Detection  limits  for most
       elements  in  aqueous solutions obtained with   an  ultrasonic  nebulizer  are
       generally 5 to 25 times  better than those achieved  with  a pneumatic nebulizer.

       Taylor  and   Floyd  at  the  USEPA [2] reported  the  use  of  a  water-cooled
       ultrasonic  nebulizer for  ICP analysis  of  environmental samples.   Pollutant
       metal analysis using the ultrasonic system  was not  affected by the presence of
       organic  pollutants   or  moderately high  salt  contents.   Correspondingly,  no
       nebulizer  plugging   problems  were  encountered,   no interferences  due   to
       desolvation were noticed, and no problems with  sample memory were observed.

       Despite  the  advantages  of  ultrasonic  nebulization, poor system   design  often
       led to short  transducer  life,  short-term instability, long-term signal   drift,
       and severe memory effect.   Recently,  a new  air-cooled ultrasonic  nebulizer [3]
       was  introduced   for routine  analysis  in ICP-AES.   This   system   overcame  the
       shortcomings  mentioned above.

       In  this study,  an air-cooled ultrasonic nebulizer  was used to aspirate  water
       and  waste  samples.   The ICP-AES  technique was  able  to determine  pollutant
       metals below  their regulated  levels  when an ultrasonic nebulizer  was utilized.
       Quality assurance and quality control  procedures described  in the USEPA  Method
       200.7  [4,5]  were followed  to demonstrate the capability of   this  methodology
       for routine analysis of  water and  wastes.
                                            H-201
-------
EXPERIMENTAL

A simultaneous and a sequential ICP spectrometer were  used in this study.   The
pneumatic  nebulizers  on  both  systems were   replaced  with  the  air-cooled
ultrasonic  nebulizers.   Table 1 shows the operating  conditions for  the   ICP
spectrometers  and the ultrasonic nebulizer.  A conventional torch and a  low-
gas-flow torch were used for the simultaneous and  the  sequential spectrometer,
respectively.  The ICP operating conditions for the  ultrasonic nebulizer  were
similar to those for the pneumatic nebulizer.

Multielement  solutions were prepared from their 1000  mg/L reference  standard
solutions.  Unless otherwise stated, all standard, blank,  and sample solutions
were  acidified with 5% HCl and 1% HNC>3.  Sample preparation and  experimental
procedures described in Method 200.7 [4,5] were followed  whenever possible.

RESULTS AND DISCUSSION

The  operating  principle  of the air-cooled  ultrasonic   nebulizer  has  been
described  elsewhere   [3]    Table  2 lists  the   analytical  wavelengths  and
instrumental  detection limits for the  ICP spectrometers.   Comparable  results
were  achieved with both systems, regardless of the  difference in torch type.
In  general, these detection limits are factors of 5 to 25 better  than those
obtained with a pneumatic nebulizer.

A  major problem in ICP analysis of wastes is spectral  interference.    Typical
spectral  interferences  on pollutant metals such  as As,  Pb,  Se,   and  Tl  are
caused  by the presence of large amounts of Al  and Fe  in  the  samples.   Since
the  ultrasonic  nebulizer  has  a higher  nebulization  efficiency  than  the
pneumatic  nebulizer   [1],   larger amounts of analytes  and  interferents  are
introduced into the plasma.  Unless the interferent  peak  is extremely close to
the  analyte peak and the resolution of the spectrometer  is relatively low,  a
lesser  extent  of  spectral  interference  is  expected   for  the  ultrasonic
nebulizer, as demonstrated by the results shown in Table  3.   The  interference
information is expressed as analyte concentration  equivalents arising from 100
mg/L of the interferent elements.  The  extent of spectral  interference greatly
depends  on the choice of analytical wavelengths and their corresponding  off-
line  background  correction positions.  Negative  values  are due  to  spectral
interferences  near  the background correction  positions  rather  than  at  the
analyte peaks.  It is important to note that some  of the  analytical lines  used
in  this study (Table 2) are different  from those  listed  in Method 200.7  [4].
In general,  the analyte concentration equivalents  obtained with the ultrasonic
nebulizer  in  this  study  are either  comparable  to   or   smaller  than those
obtained with the pneumatic nebulizer [4]

In  the  appendix  to  Method 200.7  [5],  a  preconcentration  procedure  was
described  to lower the method detection limits for  the analysis  of  drinking
water.    The modified procedure also provided improved accuracy and  precision
by concentrating the analytes four-fold prior to pneumatic aspiration into the
plasma.    Table  4  compares the ICP method  detection limits  obtained  with
ultrasonic  nebulization,  pneumatic nebulization,  and   pneumatic  nebulization
with  4-fold preconcentration.  Detection limits achieved with the  ultrasonic
nebulizer are much lower than those obtained with  the  pneumatic nebulizer, and
are well below the maximum contaminant  levels set  by the  National Primary  and
                                     n-202
-------
 Secondary   Drinking Water  Regulations  [6,7].   The  accuracy and precision  data
 for  As  and Pb  are  compared in Table  5.   As  clearly indicated,  the   accuracies
 of determination are comparable for all  systems.   Among  the procedures  listed,
 direct   analysis with ultrasonic nebulizer yielded the best precisions  at  the
 maximum  contaminant levels (MCL).

 Table  6  lists precision  and accuracy data  for  deionized  water   spiked  with
 contaminants at low concentration levels.   Superior precisions were  obtained
 with  the  ultrasonic nebulizer.   Except  for  Cr and Zn, the recoveries   of  all
 elements  are   acceptable   at such low concentrations.    The  relatively  pool-
 recoveries  for  Cr  and Zn at these  concentrations  were  attributed   to  the
 considerable amounts of impurities found in  the  nitric acid.

 Tap water  samples from Omaha,  Nebraska were  spiked with  contaminants at  their
 respective MCL  and  1/2 MCL,  Table 7.   Results  obtained with both  spike   levels
 are similar.  The standard deviations  for the  measurements are  less  than those
 obtained with a pneumatic  nebulizer  [5],  As compared to the results in  Table
 6,  the  recovery  of  Cr  in Table 7  has been  improved.    Because   the  spike
 concentration of Zn was quite high in the tap  water,  the level  of contaminant
 in  the  nitric  acid could not significantly  affect its  recovery.    Further
 investigation   revealed that  the signal suppression was  due  to   ionization
 interference [1]  caused  by the presence of Na  (about 100 mg/L)   in the  tap
 water.   The ionization effect can be corrected by buffering the sample,   by
 matrix matching,  or by standard addition procedures [4j.

 A major  concern  for  the analysis  of  water  and   waste  samples is   quality
 assurance  and quality control.   Preliminary  analysis of  a USEPA ICAP-23  water
 pollution   quality  control sample revealed that  only half of the  results  were
 within +/- 5% of the true  value listed,  Table  8.   A new  solution  was prepared
 and aspirated with  the pneumatic nebulizer.  Data  obtained with the  pneumatic
 nebulizer   are   primarily  the same, except that  the  standard  deviations  are
 higher.    The   relatively   poor recoveries for Al,  Ba, Ca,  K,   Na,   and  other
 elements  might  be due to .contamination from  the   glassware,   especially  the
 ampul which was used to store the concentrated ICAP-23 quality control  sample.
 It  is   also interesting to point out  that the ampul contained slightly  less
 than 20  mL of the concentrated sample  rather than  the stated amount  of  23 mL.

 Figures  1  to   3  show the control charts for  routine   monitoring  of  waste
 effluents.   As  illustrated in Figure 3, the  mean value,   the  upper   control
 limit,   and the lower control  limit (95% confidence interval)  are represented
 by horizontal grids on each control chart.   The  true value of concentration is
 shown in parentheses.   These data were obtained  through  an extended  period  of
 time  by a technician.   Figure 1 shows the control charts for As,  Cd, Cr,  Cu,
 Ni,  and  Zn in the USEPA WP-287 water pollution  quality  control   sample.
^Excellent  recoveries were  obtained for all elements including Cr  and Zn.  Most
 of the data are within the control limits of the true values.   Control   charts
 for the  USEPA ICAP-19 water pollution  quality  control sample are  compared  for
 the  ultrasonic  and  the  pneumatic nebulizers,  Figure 2.    The  dates   listed
 indicate  the change of sample introduction  system for routine  ICP  analysis.
 Obviously,   the  capability of ultrasonic nebulization for routine   water  and
 waste  analysis  is clearly documented.   Finally,  the advantage  of  low-level
 detection   is   demonstrated for the determination  of As  in  the  USEPA   WS-378
 (Cone.   18)  water supply quality control sample, Figure  3.   The mean value  of
                                     H-203
-------
0 053  mg/L  (Std  Dev  = 0.003) for 20  measurements   in  5  months  compares
favorably  with  the true value of 0.051 mg/L  (Std  Dev = 0.004).    These  data
collectively  suggest that the ICP-AES method with  ultrasonic  nebulization is
an excellent alternative for the determination of metals in water and wastes.

SUMMARY

In conclusion,  the air-cooled ultrasonic nebulizer  significantly  improved  the
detection  limits of pollutant metals in water and  waste samples.    Comparable
accuracy  and better precision were obtained when the  pneumatic   nebulizer of
the ICP-AES system was replaced with the ultrasonic nebulizer.  The ultrasonic
nebulization system was applicable to both the conventional  torch and the  low-
gas-flow  torch.   Prior  to  the use  of  ultrasonic   nebulizer,   trace-level
analysis  of  pollutant metals including As and Pb  had to be  performed  on  a
graphite  furnace  atomic absorption spectrometer.  The  ICP-AES   method  with
ultrasonic  nebulization  significantly improved  laboratory  productivity by
minimizing the usage of graphite furnace atomic absorption.

ACKNOWLEDGMENT

We  greatly appreciate Theodore D. Martin at the Environmental Monitoring  and
Support  Laboratory.  USEPA,  Cincinnati, Ohio for  his  helpful   suggestions.
Acknowledgment  is also made to Billy J.  Fairless and  Douglas  J.  Brune at  the
Environmental Services Division, USEPA, Kansas City, Kansas  for providing  some
of the water pollution quality control samples.

REFERENCES

1. V. A. Fassel and B.  R. Bear, Spectrochim. Acta 41B,  1089  (1986).
2. C. E. Taylor and T.  L. Floyd, Appl.  Spectrosc. 35,  408 (1981).
3. S.-K. Chan and J.-P. Zajac,  "Analytical Performance  of an Air-Cooled
   Ultrasonic Nebulizer for Inductively Coupled Plasma-Atomic  Emission
   Spectrometry",  presented at the 21st Annual Conference of the  Canadian
   Mineral Analysts, Timmins,  Ontario,  September 26-29  (1989).
4  T. D. Martin and J.  F. Kopp, "Inductively Coupled Plasma-Atomic  Emission
   Spectrometric Method for Trace Element Analysis  of  Water  and Wastes:  Method
   200.7",  Environmental Monitoring and Support Laboratory.  USEPA,  Cincinnati,
   Ohio (1984).
5. T  D. Martin,  E.  R.  Martin,  and G.  D.  McKee, "Inductively Coupled  Plasma
   Atomic Emission Analysis of Drinking Water:  Appendix to  Method  200.7,
   Revision 1.3",  Environmental Monitoring and Support  Laboratory,  USEPA,
   Cincinnati,  Ohio, March (1987).
6. National Primary Drinking Water Regulations, 40 CFR  141,  USEPA.
7. National Secondary Drinking Water Regulations, 40 CFR 143,  USEPA.
                                    H-204
-------
TABLE 1.   Experimental Facilities and Operating Conditions

Simultaneous ICP Spectrometer  (System I)
     Model
     Manufacturer
     ICP torch
     Operating conditions
ICAP 61
Thermo Jarrell Ash Corporation
Conventional torch
Forward power 1.1 kW
Outer gas flow rate 16 L/min
Intermediate gas flow rate 0.5 L/min
Injector gas flow rate 0.7 L/min
Observation height 15 mm above load coil
Integration time 4 s
Sequential ICP Spectrometer (System II)

     Model                     : 3410 ICP
     Manufacturer              : Applied Research Laboratories, Inc.
     ICP torch                 : Low-gas-flow, mini-torch
     Operating conditions      : Forward power 0.65 kW
                                Outer gas flow rate 7.5 L/min
                                Intermediate gas flow rate 0.8 L/min
                                Injector gas flow rate 0.8 L/min
                                Observation height 9 mm above load coil
                                Integration time 1 s
Ultrasonic Nebulizer (For Systems I and II)

     Model                     : U-5000
     Manufacturer              : CETAC Technologies, Inc.
     Transducer                : Air-cooled piezoelectric, 1.4 MHz
     Operating conditions      : Current 5 A
                                Heating temperature 140  C
                                Cooling temperature 5  C
                                Sample uptake rate 2.5 mL/min
                                     n-205
-------
Table 2.   Analytical Wavelengths  and Instrumental Detection Limits (ug/L)
System I ,
Wavelength
328.068
396.152
193.696
493.409
234.861
317.933
228.802*2
228.616
205.552*2
324.754
259.940
417.206
766.491
279.553
257.610
202.030
588.995
231.604*2
220.353
217.581
361.384
196.026
189.989
421.552
190.864*2
292.402
213.856
USN
IDL
0.07
0.2
1
0.2
0.03
0.3
0.1
0.3
0.5
0.06
0.2
0.4
10
0.03
0.03
0.3
0.4
0.8
1
3
0.02
2
2
0.1
3
0.1
0.07
Element

  Ag
  Al
  As
  Ba
  Be
  Ca
  Cd
  Co
  Cr
  Cu
  Fe
  Ga
  K
  Mg
  Mn
  Mo
  Na
  Ni
  Pb
  Sb
  Sc
  Se
  Sn
  Sr
  Tl
  V
  Zn
*2 means second order.
    System II,  USN
 Wavelength     IDL

 328.068        0.1
 396.152        0.5
 193.696        2

 234.861        0.01

 214.438        0.2
 237.862        0.4
 267.716        0.3
 324.754        0.5
 259.940        0.2
257.610         0.07
231.604         0.7
220.353         1
196.026
311.071        0.2
213.856        0.05
IDL=3*Std Dev, based on 10 measurements of a blank solution.
                                   H-206
-------
Table 3.   Analyte Concentration Equivalents  (mg/L) Arising from Interferents
          at the 100 mg/L Level

                                   Interferent
lalyte
Al
Sb
As
Ba
Be
Cd
Ca
Cr
Co
Cu
Fe
Pb
Mg
Mn
Mo
Ni
Se
Na
Tl
V
Zn
> Al

0.04
-0.09





--

0.03
-0.01
--





0.03


Ca Cr
- _

0.07
_ .

. .
0.12
. _
..

. .
0.05
0.01

0.01
_ .
__

0.11
0.01

Cu
0
-0.05 -0


0

0
0.02 0



0.03 0
--
0

0


0
0
1.18 0
Fe
.02
.03


.02

.02
.01
--


.02
--
.15

.01


.18
.03
.02
Mg
—
0.03 0




0
0


0.02 0

-1

0

0
0
0
-0

Mn
—
.03




.05
.01
--

.03

.03

.01
--
.04
.02
.27
.02



0
0


-0

0
0


0


-0





0
Ni

.10
.23


.01

.21
.10
--

.04


.01





.40

0
-0
-0



0
0
0

0

0


0

0
0
0

Ti
.02
.06
.03



.07
.03
.31

.03

.02
--

.01
--
.02
.03
.14

V
—
0.21
5.04


0.01
0.04
--


0.07
0.01

-0.01
-0.02

0.04

1.38


 ICP System I with USN was used.
Table 4.  Comparison of Maximum Contaminant Levels (mg/L) and ICP Method
          Detection Limits3 (mg/L) for National Primary and Secondary
          Drinking Water Regulations

                      Method Detection Limit

Element
Ag
As
Ba
Cd
Cr
Pb
Cu
Fe
Mn
Zn

MCL
0.05
0.05
1
0.01
0.05
0.05
1
0.3
0.05
5
USN
(System I)
0.0002
0.0017
0.0002
0.0002
0.0007
0.0009
0.0002
0.0004
0.00003
0.0001
PN
(Ref 5)
0.0028
0.0157
0.0013
0.0013
0.0031
0.0157
0.0028
0.0063
0.0003
0.0019
PN 4X
(Ref 5)
0,0013
0.0030
0.0004
0.0006
0.0006
0.0046
0.0007
0.0037
0.0002
0.0010
 MDL=3.143*Std Dev, based on 7 measurements of a low-level spiked solution.
                                    H-207
-------
Table 5.   Comparison of Precision and Accuracy Data   (mg/L) for As  and  Pb




                            Determined Accuracy & Precision
Method Element
USN
(System I)
PN
(Ref 5)
PN 4X
(Ref 5)
As
Pb
As
Pb
As
Pb
MCL
Spike
0.
0.
0.
0.
0.
0.
.05
05
.05
05
.05
05
Mean
0.052
0.053
0.050
0.050
0.051
0.050
Std Dev
0
0,
0
0,
0,
0.
.0008
.0002
.007
.006
.001
,002
95% Confidence
Interval
0.051
0.053
0.036
0.038
0.048
0.046
0.053
0.054
0.064
- 0.062
- 0.052
0.054
Percent Recovery
Range
101
106
84
88
98
96
- 106%
- 107%
108%
- 106%
102%
- 102%
Mean
104%
107%
99%
100%
101%
99%
 Based on 7 measurements of a single aliquot.
Table 6.   Precision and Accuracy Data  (mg/L)  for Deionized Water
                           USN (System I)
PN 4X (Ref 5)
. ement
Ag
As
Ba
Cd
Cr
Cu
Fe
Mn
Pb
Zn
Spike
0.0020
0.0100
0.0025
0.0025
0.0025
0.0020
0.0160
0.0025
0.0100
0.0040
Mean
0.0018
0.0100
0.0026
0.0024
0.0011
0.0021
0.0153
0.0023
0.0100
0.0033
Std Dev
0.0001
0.0001
0.0001
0.0001
0.0003
0.0000
0.0003
0.0000
0.0004
0.0001
Recovery
89%
100%
103%
97%
43%
105%
95%
92%
100%
84%
Mean
0.0021
0.0107
0.0028
0.0024
0.0027
0.0018
0.0170
0.0025
0.0097
0.0044
Std Dev
0.0002
0.0012
0.0002
0.0002
0.0002
0.0002
0.0006
0.0001
0.0013
0.0006
Recovery
105%
107%
108%
96%
108%
90%
107%
100%
97%
110%
 Based  on  7  measurements  of  a  single  aliquot.
                                     H-208
-------
Table 7.  Precision and Accuracy Data   (mg/L)  for Omaha Tap Water
MCL
Element
Ag
As
Ba
Cd
Cr
Cu
Fe
Mn
Pb,
Znb
Spike
0.
0.
1
0.
0.
1
0.
0.
0.
1
05
05

01
05

3
05
05

Average
Mean
0
0
0
0
0
0
0
0
0
0
.0490
.0497
.945
.0093
.0384
.941
.285
.0459
.0474
.901
Recovery
Percent
98%
99%
95%
93%
77%
94%
95%
92%
95%
90%
95% Confidence
Std Dev
0
0
0
0
0
0
0
0
0
0
.0003
.0011
.005
.0001
.0005
.005
.002
.0003
.0008
.004
0
0
0
0
0
0
0
0
0
0
Interval
.0484
.0475
.935
.0091
.0374
.931
.282
.0453
.0458
.894
- 0
- 0
- 0
- 0
- 0
0
0
0
- 0
- 0
.0496
.0519
.955
.0095
.0394
.951
.288
.0465
.0490
.909

Element
Ag
As
Ba
Cd
Cr
Cu
Fe
Mn
Pb
ZnC
1/2 MCL
Spike
0.025
0.025
0.5
0.005
0.025
0.5
0.15
0.025
0.025
0.5
Average
Mean
0.0245
0.0248
0.467
0.0047
0.0204
0.467
0.142
0.0230
0.0231
0.451
Recovery
Percent
98%
99%
93%
93%
82%
93%
94%
92%
92%
90%
95% Confidence
Std Dev
0.0001
0.0005
0.003
0.0001
0.0004
0.0025
0.001
0.0001
0.0007
0.004
Interval
0.0243
0.0238 -
0.460 -
0.0045 -
0.0196 -
0.462
0.139 -
0.0228
0.0217 -
0.444
0.0247
0.0258
0.474
0.0049
0.0212
0.472
0.144
0.0232
0.0245
0.458
 Based on measurements of 7 aliquots.
 ^Spike level for Zn is 1/5 of the MCL.
 :Spike level for Zn is 1/10 of the MCL.
                                     H-209
-------
Table 8.   Analysis  of USEPA ICAP-23 Water  Pollution Quality Control Sample
          (mg/L)

Element
Al
As
Ba
Be
Ca
Cd
Co
Cr
Cu
Fe
K
Mg
Mn
Mo
Na
Ni
Pb
Sb
Se
Tl
V
Zn
True
Value
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
10.0
1.0
1.0
1.2
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
USN
Mean
1.631
1.004
1.291
1.083
1.118
1.038
1.010
1.089
1.101
1.038
9.131
1.033
1.008
1.142
1.863
1.013
1.136
1.113
1.117
0.963
1.032
1.084
(System
Std Dev
0.009
0.002
0.008
0.005
0.006
0.003
0.008
0.009
0.004
0.008
0.054
0.007
0.007
0.009
0.014
0.007
0.004
0.003
0.004
0.016
0.007
0.006
I)
RSD
0.5%
0.2%
0.6%
0.5%
0.6%
0.3%
0.8%
0.8%
0.4%
0.7%
0.6%
0.7%
0.7%
0.8%
0.7%
0.7%
0.4%
0.3%
0.3%
1.7%
0.7%
0.6%
PN
Mean
1.532
0.936
1.289
1.065
1.089
0.958
1.009
1.033
1.073
1.041
9.416
1.016
0.993
1.140
1.783
1.018
1.027
1.026
0.981
1.051
1.024
1.027
(System I)
Std Dev
0.024
0.011
0.009
0.003
0.004
0.004
0.007
0.011
0.015
0.005
0.476
0.006
0.004
0.004
0.079
0.006
0.012
0.016
0.016
0.026
0.005
0.006

RSD
1.5%
l.U
0.7%
0.3%
0.2%
0.4%
0.7%
1.1%
1.4%
0.5%
5.1%
0.5%
0.4%
0.3%
4.4%
0.6%
1.1%
1.5%
1.6%
2.4%
0.5%
0.5%
Based on 7 measurements of a  single aliquot.
                                   n-2io
-------
H
to
                                     As  (O.1O  mg/L)
                                                                 CTJ.   (O.1O  mg/L)
0.103


0.063

                                                              ¥
                                                                                        Ni  (O.1O  mg/L)
                             1   1  !   1  1   1   1  1   1  1   1  1   1  1   1   1
                                     Cr  (0.10  mg/L)
                                                                          0.101
                                                                          0.089
                                                                                I   I  T  II   I  I   I   I  I   I  I   I  I   I  T

                                                                                        5            10            15
                                                                     DAY
                       Figure 1.   Control Charts for WP-28?  (USN,  ICP  System  II)
-------
                  PN  Cd   (1.0   mg/L)
                                                                  USN   Cd   (1.0   mg/L)
2
o
:   i  i   i   i  \   i—i—i—i—\—rn—i—i—i—i—r
       PN   Cr  (1.0   mg/L)
0.90

1.15

1.07
0.99

0.91

0.83
-i	1	1	1	1	1	1	i	r
     USN   Cr   (l.O   mg/L)
                                                                               i    i     i	1    i     i	1	1     r
                                                                             USN   Pt>   (1.0   mg/L)
     0.85
                                                                          i    i     i	1	1	1	1	1	1	1
                                                      0.89
                                                           DATE
     F i gu re  2.    Compar ison of Control  Data  (1 CAP-29)  for  Pneumatic  and  Ultrasonic Nebulizer
                    ( ICP  System  I I )
-------
                                              As    (O.O5O    mg/L)
¥
to
oc
h-
z
LU
o
•z.
o
o
                                                                                                          UCL
Mean
                                                                                                          LCL
                                              I    I    I    I    I    I   I    I    I    I    I    I    !    I    I

                             11/1011/11 11/1612/812/2212/29 1/5  1/12 1/19  1/26  2/6 2/10 2/15 2/22  3/12 3/19 3/22  3/30 4/11 4/12


                                                                 DATE


                        Figure  3.   Control  Chart for As  in WS-378  (USN,  ICP  System II)
-------
87                             A STUDY OF THE LINEAR RANGES
                           OF SEVERAL ACID DIGESTION PROCEDURES

         DAVID E KIMBROUGH. PUBLIC HEALTH CHEMIST, AND JANICE WAKAKUWA, SUPERVISING
         CHEMIST, CALIFORNIA DEPARTMENT OF  HEALTH  SERVICES,  SOUTHERN CALIFORNIA
         LABORATORY, 1449 W. TEMPLE STREET, LOS ANGELES, CALIFORNIA 90026-5698.

         ABSTRACT

         The  analysis of  solid  matrices  (sediments,  sludges,  soils,  and  solid
         wastes) for the presence of regulated elements is most commonly performed
         using  an acid digestion procedure. The acids destroy the matrix and react
         with the elements  of  interest  to  form water soluble  compounds.   When the
         digestion  is complete, it  is usual  to add  water,  forming a liquid which
         can then be  analyzed  by  a  variety of analytical  instruments.  Implicit in
         most published methods is the belief that elements will be solubilized in
         direct proportion to the concentration of the  element  in  the matrix at any
         concentration or in any molecular  form.  However,  previous studies1-2-3have
         shown  this is not  always the  case.


         The purpose  of  this study  is to examine some of  the  factors  that affect
         the  linear range  of  commonly regulated elements.  Five factors  will be
         examined:  1) the  total amount  in  micrograms of the  analyte in a two gram
         sample; 2) The  solubility  of  the  compound  in  which  the element is bound;
         3) The vigor of the acid cocktail; 4)  the effect  of  other target elements
         in  the sample  on  solubilization  and/or co-precipitation  of  the  element
         under  consideration;  5)  concentration of hydrochloric acid.

         Four methods, EPA 3050,  SCL, ASTM  9 3.4 (a nitric acid/hydrogen peroxide
         digestion  modified for  solids analysis), and the digestion described in
         EPA draft  Method  6020 will be  compared.

         INTRODUCTION

         The  determination of the  concentration of regulated elements  in  solid
         matrices   (sediments,  sludges,  soils,  and solid wastes  such as  spent
         catalysts, press cakes,  slags,  powders,  etc.)  is generally performed using
         acid  digestion procedures.  The  purpose  of  the  acid  digestion is to
         solubilize all the  elements  of  interest.   To do  this,  a  digestion
         procedure  must perform two  distinct  tasks:  1)  It must decompose the sample
         matrix to  expose the  entire mass  to the acid  cocktail.  2) It must react
         with the elements  of interest  to form  water  soluble  compounds.   If the
         elements are already  in a soluble state, then this task is not necessary.
         When  the   digestion  is  complete,  it  is  usual  to  add  water  to form a
         solution   that   is  suitable  for  analysis  by a  variety  of  analytical
         instruments (typically Flame Atomic  Absorption Spectrophotometry (FAA) and
         Inductively Coupled Plasma  Atomic  Emission  Spectroscopy (ICP-AES) but also
         Graphite   Furnace  Atomic   Absorption  Spectrophotometry   (GFAA)   and
         Inductively Coupled Plasma  Mass-Spectroscopy (ICP-MS)) .  If the digestion
         is successful,   the  amount  of  the  element in the  solution is  equal to the
                                           H-214
-------
amount  of the  element  in the  sample  matrix.    Ideally,  a digestion
procedure  should  be able  to  solubilize  any  amount of  the  elements of
interest in  the matrix,  irrespective of  the  molecular  state, so that a
graph of  concentration found versus concentration  of  analyte in matrix
would be linear.   Previous studies have shown  that this  is  not always the
case1-2-3-   There is a limit to the range of matrix concentrations that will
yield a linear relationship,  for a given digestion procedure.   This is
called the linear  range.  Above or below this range,  the  amount of analyte
in solution is significantly lower than the amount  in the  sample matrix.

There is no  data  available on  the linear ranges for most  published acid
digestion  procedures.  Therefore,  analysts and review personnel have no
guide for evaluating the data obtained from these methods.  Linear ranges
for target elements  vary from element  to element, and exhibit several
different types of relationships between the concentration in  solution and
the concentration  in the sample matrix.
This study will examine the factors that determine the  type  and  length of
the  linear  ranges  of  antimony,   arsenic,  barium,  beryllium,  cadmium,
chromium,  cobalt,  copper,  lead,  molybdenum, nickel,  selenium, silver,
thallium,  vanadium,   and zinc  (hereafter  referred to  as  the "target
elements")  by four published  methods.    These  include EPA  SW  846  (3rd
Edition) method 3050 (Version 1.) ,  an alternative  method from the Southern
California  Laboratory  (SCL)   of   the  California  Department of  Health
Services  (DOHS), hereafter  referred to  as the SCL method, the  EPA draft
method  6020  for ICP-MS,  and ASTM  method  9.3.4 modified  for solids,  was
also compared  (this method  is very similar  to EPA  SW 846 method 3050 for
GFAA).    This  list  of  regulated elements  is  based  on California
Administrative Code Title 224 which is  similar to Federal  regulations5-6.

Five factors that contribute  to  linear range will be examined:  1)  the
total amount in micrograms  of the  analyte  in a  two gram sample;  2)  The
solubility of the compound in which the  element is bound;  3)  The vigor of
the acid cocktail, i.e., the power of the cocktail to  oxidize the target
elements;  4)  the  effect  of  other target  elements  in  the sample on
solubilization and/or co-precripitation of the element under consideration;
5) the  amount of hydrochloric  acid present.

EXPERIMENTAL SECTION

A) Study Design.

The first factor is the amount of  analyte in a sample.  To  examine this,
all sixteen target elements were digested by all  four methods, EPA SW 846
method  3050, the SCL method, EPA draft method 6020, and ASTM 9.3.4 over  a
range of concentrations  from about 100  ug/g to  1,000,000 ug/g  for a two
gram sample size (with the exception of method 6020 where  one gram samples
were used).
                                  E-215
-------
Initially, three solid materials were spiked with all sixteen elements at
different' concentrations  from  100  ug/g  to 10,000  ug/g  (these  three
materials were analyzed by  10 different laboratories by the SCL method and
method 3050.  See reference 2).   Elements that showed linear  behavior in
this range were digested individually at concentrations  of  25%,  50%,  and
100% (by weight) using reagent grade compounds that were,  where possible,
water soluble.  Elements  that showed non-linear behavior were  then spiked
into  solid  matrices  at  concentrations  that  seemed  to  bracket  the
transition  region  from  linearity  to non-linearity.    This process  was
repeated using  less  soluble  compounds.   In  this  way it was  possible  to
separate the power of the acid cocktail  to  decompose a sample  matrix from
its ability to react with target elements to form water soluble compounds.
For each  element  and method linear  range  curves  were prepared  for  each
type of  compound.  Finally,  eleven  samples  from  industrial  sites were
analyzed by all four methods.

B) Digestion Procedures

(1) EPA  SW 846  method 30507  was  used  as  designated.  It  directs  that
1.00-2.00 grams of sample  be digested first with 10 mL 1:1 (v/v)  nitric
acid at 95°C for 15 minutes,  then 5 mL cone,   nitric acid is added and  is
refluxed for 30 minutes.  This step  is repeated until the sample no longer
changes in appearance.  The digestate is then concentrated  to 5  mL.  The
sample  is treated with  no more  than 10 mL of  30%  hydrogen  peroxide.
Finally,  5 mL  of cone.   hydrochloric acid  and 10 mL  of de-ionized water
are added and the sample is refluxed for 15 minutes.   The sample is  then
either filtered (Whatman  41 or equivalent) or centrifuged (2-3,000 rpm for
10  min)   and  the  filtrate  (or supernatant)  is  collected  in  100 mL
volumetric flask and analyzed by either FAA or ICP.

(2) The SCL method1'2 calls  for 1.00   4.00  grams  of sample to  be  digested
in a mixture of 10 mL cone, hydrochloric acid and  2.5 mL  of cone,  nitric
acid at ambient temperature.  The sample and reaction mixture are slowly
heated to 95°C to prevent an overly vigorous reaction.  The digestion  is
continued until the disappearance of N02 (reddish brown) fumes and no more
change in appearance.    The  digestate is  then filtered  (Whatman 41  or
equivalent) and collected in a 100  mL volumetric  flask.  The filter paper
is washed with no more than 5 mL hot  (95°C)  cone,   hydrochloric acid, and
then 20 mL hot de-ionized water,  all  of which is  collected into one flask
(this will be referred to as the primary filtrate). The  filter paper and
residue are placed back in the digestion vessel,  5 mL cone,  hydrochloric
acid are added and refluxed at 95°C until the filter  paper  disintegrates
(approx.   10-15  min.) The disintegrated   paper  is  then  washed with
de-ionized water and again filtered (this filtrate will be referred to  as
the secondary filtrate.)  The filtrates should be  allowed  to come to  room
temperature before bringing to volume.   These filtrates are  then analyzed
by either ICP or FAA.  The results are combined as  follows:
                                 H-216
-------
                                (EQUATION 1)

Xx    =     concentration of element x in primary filtrate
X2    =     concentration of element x in secondary filtrate
Xt    =     total concentration of element x  «-  Xj^ +  X2
V     =     volume of volumetric flask  used for both primary and secondary
            filtrate.
w     =     weight of sample taken
C     =     concentration of element x in sample.

If a precipitate  forms  on the  bottom of either the primary or secondary
filtrate flasks after the flasks have cooled, then add  up to 10 mLs of
cone. HC1 to the  flasks. The additional  acid will either 1) dissolve the
precipitate or 2) more  precipitate  will  form.   If the latter occurs: a)
Decant the liquid portion and  filter through a Whatman 41 or equivalent
and collect the filtrate in a new flask marked  "filtrate" (either primary
or secondary).  Place the original flask with the precipitate under this
same funnel and wash  the filter paper  with hot HC1 until all precipitate
in filter paper has dissolved.   Then add enough cone.  HCl to the flask as
needed to dissolve the remaining precipitate on the bottom.   This flask is
marked  as   either   primary  or  secondary   "residue".     The  primary
concentration is  then equal  to  the  filtrate plus the  residue, as is the
secondary  concentration.  (NOTE:  This  method  is  an  updated  version of
previously published versions)

3) A.S.T.M. Method 9.3.4 calls for 2.0  grams of sample  digested with 10 mL
of cone. HN03 and  swirl and allow it to  sit until any frothing subsides.
Heat to 95° C and  allow  refluxing  for 30  minutes.  The digestion cover is
then removed  and the  sample is  brought to  near dryness  (this  step is
modified from the published method that calls for  taking  the sample to
dryness).  5 mL of cone.  HN03 is then added.  30% Hydrogen peroxide is then
added drop-wise until the solution cease to change color.  The sample is
then evaporated to  about 3 mL and  filtered  through a Whatman 41 filter
paper and collected in  a 100 mL volumetric flask.

4) EPA Draft Method 6020  (CLP-M)  Version 3.4 SAS calls for 1.0 grams of
samples digested  with 2 mL  of  (2+3)  HN03 and  10 mL  of  (1+4)  HCl.   The
slurry is heated to  95°  C and refluxed for 30 minutes.  The  sample is then
filtered and brought  to volume with ASTM type I water to 200 mL.

C)  Instrumentation  and  Analysis.   The  digestates were  analyzed  on  a
Perkin-Elmer 5500 ICP-AES by EPA SW 846 method 6010  and in  a few cases on
a Varian Video 12 FAA by EPA SW 846 methods 7090, 7130, 7760.

D) Materials.  Three solid phase samples  were prepared for the  study.  The
solid phase samples  were designated  from  D to F.  Samples  D, E,  and F were
spiked samples with all the target elements  in  each  sample.  Each analyte
was spiked at three  different concentrations, designated  low, middle, and
                                  H-217
-------
high.  These materials were spiked and homogenized in the laboratory.  The
concentrations in these samples are referred to as  the  "true"  value (for
a more  detailed  description of  these  materials see  reference 1) .   All
materials with concentrations above 200,000 ug/g were neat  reagent grade
chemicals.

RESULTS

Table  I  lists the  linear  ranges for  all  the target  elements for  each
digestion procedure.   Of  the  four digestions,  the SCL  method had  the
fewest  limitations  on  its  linear ranges,  there were only  four elements
that could not be solubilized up  to  2  grams,  while method  3050 had six,
ASTM had seven,  and method 6020 had  10,  depending  on  the  state  of  the
element.

Table  II lists the  results from each  method  for  eleven different field
samples.  Sample A  is a spent catalyst, sample B is  spent catalyst mixed
with soil,  samples C through G are materials from battery recovery  plants,
H is a mixed non-petroleum waste from  an oil refinery possibly including
spent  catalyst,  I and  J are materials from a  solid waste  refinery,  and
sample  K  is  a press cake  mixed  with petroleum products.   All of these
samples were dried,  milled, and sieved  through a U.S. standard #10  sieve.

Table  III shows the  distribution of elements solubilized in  water  and the
amount  trapped in the filter paper and  residue  using the SCL method.   For
comparison, the  results from method  3050  are  listed next to the  primary
filtrate results.

Four types of relationships  were observed between the concentration of
target  elements  in  the sample matrix and the amount  of  target  element in
solution:

1)  Direct  Linear Relationship.   Most elements  showed  a direct  linear
relationship  between  the  concentration  in  the  solid matrix  and  the
concentration solubilized in the liquid from a concentration of 100 ug/g
to  1,000,000 ug/g for  all  four methods.   This means that if the  element
was  in a  water  soluble form,  all of  it  can  be held  in solution.   For
example, all  four methods completely solubilized, as little as 130  ug/g up
to 1,000,000  ug/g, of chromium, when  it was present in either  the  form of
potassium dichromate or chromium trioxide.

2) Asymptotic Relationship.  Several elements were solubilized in a direct
linear fashion at low concentrations but due to limited solubilities the
element either  did  not  solubilize  or was precipitated out  at  higher
concentrations   so   the   graph  curved  asymptotically  to   a   maximum
concentration. As concentration  in the sample  matrix increases above the
linear range,  the amount in solution did not proportionally increase.  A
good example  of this phenomena is the  analysis of lead  using
digestion method 3050.   Seven materials  were  prepared from a  local soil
and lead nitrate  to  yield matrix concentrations from 260  ug/g  to  120,000
                                  H-218
-------
ug/g.   These spiked soil  samples  were digested using  method 3050, the
results yielded part of Graph I:  At concentrations below 40,000 ug/g, all
the  lead  in  a  two   gram sample  will  be  solubilized.    Above  this
concentration,  the  number of micrograms  of lead in the  sample  will no
longer equal the number in solution, although it will increase up  to  about
60,000 ug/g.  The remaining  lead is insoluble and remains either at the
bottom of  the  digestion beaker or  trapped  in  the residue in the filter
paper (see table III).   Using other compounds of lead, metallic lead, lead
sub-oxide, lead  oxide,  and lead  dioxide,  the curve  was still asymptotic
but the point of deviation from linearity is dependent on the  chemistry of
the compound under consideration (see table II).

3) Sigmoidal Relationship. Antimony methods 3050 and 6020  and thallium by
ASTM  9.3.4  were  accurate  and  linear  at middle  concentrations  but
inaccurate  and non-linear at  low and  high concentrations.   This  is  a
result of two factors:  a) a fixed amount of the element  is trapped in the
filter  paper and residue.   As the matrix  concentration  increases, the
amount trapped is fixed and a greater proportion of the element present in
the matrix is solubilized. At low concentrations, the amount trapped can
be significantly more  than 25%  of  the total present and so  the curve is
inaccurate.    At higher  concentrations  this   fixed  amount  can become
insignificantly  small   and so  the  curve  is effectively  linear,  albeit
slightly displaced  on  the lower  side, b)  At even higher concentrations,
the digestate becomes  saturated and can solubilize no more and the  curve
becomes asymptotic.   This creates  a  curve  that is  somewhat S-shaped or
sigmoidal.   The  degree of  curvature can  depend  on which  compound is
present e.g., if elemental antimony is present the curve will be strongly
sigmoidal.  Using potassium antimony  tartrate  the curve is only slightly
sigmoidal at the low end  (Graph II).

4) Non-Linear Relationship.  The element is either simply not soluble or
only partially soluble.  Antimony digested by ASTM method  9.3.4 is a good
example (see Table II).

DISCUSSION

1) All of these method  can solubilize  up to 1,000,000 ug/g  of any  of  these
elements, provided  it  is in a soluble  form,  with only four exceptions.
Lead  cannot  be  solubilized by  hydrochloric acid above the low percent
range, from 6% to about 25%, depending on solubility  and reactivity of the
compound of  lead.   Chlorides of lead have  only a limited solubility in
HC1.

Antimony  and  its  salts  cannot  be solubilized by  nitric   acid  in any
significant proportion since nitric acid reacts with antimony to produce
Sb2058'10, which  is insoluble  in nitric acid.  Barium salts have extremely
variable solubilities  in nitric  acid while they will  react with HC1 to
form BaCl2  which is somewhat soluble in HC1.   Silver also has a limited
linear range for all the methods for similar reasons  to barium, AgCl  forms
easily with and is soluble in HC1.  Most of the AgCl  is  lost  when water is
                                  n-219
-------
added9.   For these elements there are inherent limitations  on the linear
range using HN03  or HC1 digestions.  These elements either do not go into
solution at all,  or  once in solution, they precipitate.  As a result, they
become  trapped  in   the  filter  paper  and  residue,  or  if  they  have
temperature dependent solubilities,  they settle to the bottom of the flask
(see Table III).

2) The solubility of the compound in which the element  is bound can have
an effect on the  linear ranges if the digestion procedure is not vigorous
enough.  This is  a problem  especially for  method  6020 but all the methods
have it to one degree or another.  This  is illustrated by the nickel data
for  method  6020  (Graph III).    Virtually  no   metallic nickel  can be
solubilized  by method 6020 but  up  to  1  gram of nickel in the  form of
Ni(N03)2-6H20.  In contrast, the  other three methods  were able to oxidize
metallic nickel  into  a soluble compound.   What  was  true for  nickel was
also true  for  antimony,  chromium,  molybdenum, and selenium.  Method 6020
was able to  oxidize up to 1 gram of elemental cobalt,  copper and zinc.

3)  The  vigor  of  the  acid  cocktail is  crucial   to a good linear  range.
Since  it  is impossible to predict  the  molecular  forms of  the  elements
present, we  cannot predict  the solubility  of the  elements in water.  Thus,
a digestion  should be vigorous enough to change the molecular form of the
element so that it can be solubilized in water.   By far the  most vigorous
digestion is the  SCL method since aqua regia has four  active species, HN03,
HC1, NOC1,  and C12.  As  a  result,  the  SCL method had the longest linear
ranges for the most elements.

4) Co-precipitation is mainly a problem  with large  quantities of lead and
to  a  far  lesser  extent,  silver.   When the  other  target elements  were
present in quantities in excess of their linear range, they became trapped
in  the filter paper  and  residue.    Lead  and   silver  have  temperature
dependent  solubilities  and  will  pass  through  filter paper  and  then
precipitates   in  the   flask  as   the  solution  cools.     During the
crystallization process, other elements  can become trapped in the lattice,
lowering the effective linear range of the  other  elements. This can be
seen on table  III.

5) The importance of  the concentration  of hydrochloric  acid varies with
the  element and  its  molecular  form.    For  some  elements  it makes no
difference how much HC1 is used,  such as,  arsenic, cadmium,  chromium,
cobalt,  copper,  nickel,  selenium,  and  zinc.   The results  for  these
elements were same for  the ASTM method, which employed no HC1, and the SCL
method, which  uses  mostly  HC1.   For a  number of elements,  however, the
amount of  HC1  used  was of critical importance.  These  include antimony,
barium,  lead,  molybdenum,  silver,  thallium, and  vanadium.   For  these
elements,  the  ASTM  method  performed the  most poorly and the  SCL method
performed  best,  having  the  longest  linear  ranges.    Method 3050  also
performed better than  the ASTM method but not as well  as the SCL method
due the much larger amount of HC1 used  in the SCL  method.
                                  n-220
-------
SUMMARY AND RECOMMENDATIONS

None of these acid digestion procedures have  completely  linear ranges for
all  the target  elements and  their various  compounds.   In  fact,  the
linearity not only varied from  method  to method and element to element,
but in some cases they also varied widely from compound to compound.  This
wide variability  does not even  begin to take into account the effect of
matrix  on  linear range.   This violates  one  of  the  basic  assumptions on
which most  acid  digestions  are based,   an  infinite  linear  range for all
elements and for all  of their compounds.  Without this  assumption, much of
inorganic  data   generated  by hazardous materials  laboratories  becomes
questionable.  Most  "high" results, especially those involving antimony,
barium,  molybdenum,   lead,   silver,  and vanadium  may  well  be  highly
inaccurate.

A  quality  control procedure  that might  take into account  the limited
linear ranges would be running duplicates of  different mass e.g., a 2.0 g
and a 1.0 g duplicates.  If the second  duplicate is significantly higher
than half of the  first, there is a strong indication that the linear range
has been exceeded.

The  data  above,   especially  the antimony  data,  also suggests  that the
existing methods  for determining "less  than" values are inadequate.  The
majority of these methods are based solely on instrument performance and
assume 100% extraction efficiency.   This assumption  is false since it has
been  shown  that the  filtration  step  actually  removes  analytes  from
solution.  Therefore, if small amounts  of a given analyte are present in
the sample matrix, they  may  not make it through the filtration step and
into the filtrate. No matter how sensitive the instrument, if the analyte
cannot  find  its way  into  solution,   the  instrument  cannot  see  it.
Furthermore, the common practice of increasing sample size to lower method
detection limits may have entirely  the opposite effect.  By increasing the
sample  size, one has increased the amount residue left by the digestion.
This residue can act like a  ion-exchange resin.   Thus  an increased sample
size can reduce  the  amount of analyte  in the  filtrate.

It  should  not be  assumed  by laboratories  and  regulators that  an acid
digestion procedure  that can solubilize an element at one concentration
can solubilize  that  element at all  concentrations  nor  in all matrices.
Laboratories must take into  consideration studies on the effective linear
range  for  methods  they are  using  and for  each element  of interest.
Published methods should list the approximate linear range for  the covered
elements.   At the very  minimum,  we recommend  that SW 846  method 3050
should be amended in the method performance  section to  note that it has
only a  limited linear range  for antimony,  barium,  lead, and silver.  We
recommend  that the  SCL method be  adopted  as  an accepted  alternative to
method 3050 for soils, sediments and sludges.   However, method  3050 should
not be used for  solid wastes  with high  concentrations  of target elements.
Method  6020  should definitely not be  used for solid wastes  due to the
large number of elements with compromised linear ranges.
                                  tt-221
-------
REFERENCES

1.    Kimbrough,  D.E.  and J.R. Wakakuwa,  "Acid  Digestion for Sediments,
     Sludges,  Soils,  and Solid Wastes.  A Proposed Alternative to EPA SW
     846 Method  3050."  ; Environmental Science and Technology, 23, pages
     898-900,  July 1989.

2.    Kimbrough,  D.E.  and J.R. Wakakuwa,  "Report of  an Interlaboratory
     Study of  an Interlaboratory  Study  Comparing EPA SW 846 Method 3050
     and an Alternative  Method from the California Department of Health
     Service"; Proceedings  of the Fifth Annual  USEPA Symposium on Solid
     Waste Testing and Quality Assurance, Washington,  D.C.  July  1989.

3.    Hinners,  T.A.,  et  al.;"Results of an Interlaboratory  Study of ICP
     Method 6010 Combined with Digestion Method 3050.";  Proceedings of the
     Third Annual USEPA  Symposium on  Solid  Waste Testing  and  Quality
     Assurance,  Washington, D.C.  July 1987.

4.    California  Administrative Code Title 22.   Social Security Division
     4  Environmental  Health Sect.   66699(b)  (Register 85,  No.1-12-85)  p
     1800.77

5.    Comprehensive Environmental Response Compensation and Liability  Act
     (CERCLA or  "Superfund")  sect.101 (14)d Title 40 CFR Part 261

6.    Federal Water  Pollution  Control Act,sect.  307(a)(1)   Title  40  CFR
     Sub-Chapter D

7.    Test Methods for Evaluating  Solid Wastes (EPA SW 846 Volume 1A)  3rd
     Edition, Method 3050.  Office of Solid Waste and Emergency Response,
     U.S.Environmental Protection Agency:Washington, D.C.,  November 1986

8.    Condensed Chemical Dictionary,  10th Ed.;  ed.  G.G.    Hawley,  Van
     Nostrand  Reinhold Company Inc.:New York  ,  1981; pg. 79-81.

9.    Cotton, F.A. & Wilkenson, G. , Advanced Inorganic Chemistry, 3rd Ed. ;
     Interscience Publishers:  New York,  1972; pg.  1048

10.   Kirk-Othmer  Encyclopedia   of  Chemical  Engineering,  3rd   Ed.;
     Wiley-Interscience: New  York,  1978;  Vol. 3, pg. 98  & 109
                                 H-222
-------
                                         TABLE I
                                  LINEAR RANGES in jzg/g

Element

Ag (AgN03)

Ba (BaOH2)

Cr (Elemental)

Mo (MoO3)

Ni (Elemental)

Pb (Elemental)

Pb (Pb(N03)2)

Pb (Pb20)

Pb (PbO)

Pb (Pb02)

Sb (Elemental)

Sb(K(SbO)Tartate)

Se (Elemental)

Tl CTI2S04)

V (NH4V03)

As203, BeSO4, Cd, Co, CoCI2.6H2O, CrO3> K2Cr207, Cu, CuSO4> (NH4)6Mo7O24, 4H2ONi(NO3)2.6H2O, H2Se03, and Zn
were all analyzed and found to be linear for all four methods.

L = 50- 1,000,000 Jig/g

NL = Not Linear
3050
50 - 150
50 700
L
50 60,000
L
50 200,000
50 - 40,000
50 - 250,000
50 - 60,000
50 - 60,000
5,000 - 20,000
5,000 - L
L
L
50 250,000
SCL
50 700
50 - 2500
L
L
L
50 - 50,000
50 120,000
50 120,000
50 120,000
50 250,000
50 - 50,000
L
L
L
L
6020
NL
NL
200 - 250,000
200 - 60,000
200 100,000
200 10,000
200 - 40,000
200 100,000
200 - 20,000
200 - 20,000
3,000 - 20,000
3,000 - L
50 - 20,000
L
200 - 250,000
ASTM 9.3.4
NL
NL
L
50 - 500
L
L
L
L
L
L
NL
NL
L
3,000 - 10,000
50-1,000
                                          H-223
-------
                                  TABLE II
                  RESULTS FROM FIELD SAMPLES in j/g/g
Element
Nl
V

BA
CU
PB
ZN

AS
BA
CR
CU
PB
SB
ZN

PB
SB

AS
CU
PB
SB

AG
CD
CU
Nl
PB
ZN
3050

31,000

270
28.0OO

340
1200
340,000
140

280
210
300
1300
280,000
12OO
120O

130,000
2600

330
800
220,000
23,000

<50
1600
6600
290
280,000
14OO
31,000

310
32,000

510
1100
110,000
93

400
310
260
1200
140,000
1300
1100

140,000
5300

330
730
110,000
97,000

230
2000
4400
370
110,000
1500
6020

29.0OO

210
28,000

<200
1500
140,000
<200

<200
<200
<200
1600
150,000
510
620

150,000
2400

<200
1300
90,000
13,000

<200
1800

390
99,000
16OO
ASTM 9.3.4

1 9.OOO

230
19,000

220
930
880,000
88

120
97
290
670
430,000
<50
940

380,000
<50

-------
                                    TABLE III
                    DISTRIBUTION OF ELEMENTS IN SCL & 3050
                  DIGESTION PROCEDUREIN TOTAL MICROGRAMS
Sample

 A

 B
              3050
              62,000
Primary
Rltrate

61,000
Primary
Residue

510
Secondary
Rltrate
Nl
V

BA
CU
PB
ZN
AS
BA
CR
CU
PB
SB
ZN

PB
SB

AS
CU
PB
SB

AG
CD
CU
Nl
PB
ZN

CO
CR
CU
MO
Nl
PB

AG
BA
CR
CU
N!
PB
ZN

AG
BA
CU
PB
ZN

AG
BA
CR
CU
MO
Nl
PB
ZN
540
56,000
680
2400
680,000
280
560
420
600
2600
560,000
2400
2400
260,000
5200
660
1600
440,000
46,000
<100
3200
13000
580
560,000
2800
1200
12,000
940
4600
26,000
380
<100
420
2000
33,000
1200
6200
8200
<100
340
10,000
38,000
88,000
170
2600
1500
3800
160
820
42,000
11,000
610
64,000
743
1700
57,000
180
800
370
520
2100
220,000
2600
2200
231,000
11,000
670
1450
149,000
180,000
460
4000
12500
750
170,000
3100
920
17,000
1400
6000
29,000
760
12,000
400
1500
33,000
1100
11,000
7900
18,000
420
7300
36,000
82,000
280
1500
2500
7300
460
1500
38,000
9600
<100
790
280
410
166,000
<100
<100
260
<100
390
53,000
130
<100
53,000
130
<100
<100
6200
6200
<100
<100
690
<100
230,000
<100
<100
350
<100
290
1300
<100
8500
<100
<100
240
<100
640
1500
3500
<100
250
640
1500
<100
1000
<100
7300
<100
<100
820
210















<100
<100
63,000
10,000






<100
350
<100
140
810
<100




















                                     H-225
-------
  1 20,000
  1 OOP0(
 5f 80,000
. .,-,.
 6

   RD non
   • •"" '"' f
   40.000'
   20fOOO
          0
                                         G rcj p i i  I
                             . i n e 1.1 r I \ c:i r i g e. f c > r  I..«.; a c 1  N i 1 r a i: e
"Hieore llr.al
 .3O50 A
 SOL O
 602.0 V
 A O'T f. ,1  ••'"••;
 .<'-\C.:' I  IVI  ••.,./
                               r
ill
                       i
                                                 4
  20,00(1   40,000   60,000
                   Pb Spiked  i.-i
                                                    80J)QO   1 00.000  1 20,000
-------
(=i
to
       1 .000.000
         1 00,000
       "g 10

       Q-:   1 ,000
       . o
       CO
9 9 i
   } A (
                                               Graph II

                              L i n e a r R ci 11 q e  f o r  EI e r n e n t a I  An t i mo n y
       Theor^'tlca I  L.

       ."5050 A
       8CL O

       6020 V
              ,,D"
                                       ,0*
                              \- ,,-J
                                                                	-:;t: ]
                                                                            .[."']'

                                                                        x<:

	-	H-	-			—I	
    1 0,000         1 00 000

Sb S[Diked ug/q
                                                                              ,000,000
-------
  ,000,000
o> BOO OOO
      f •— . —•
7 600,000-4
 > 400000
  200,000-
                                    Graph 111
                        Linear  Ranae for Elemental Nickel
               Theoneiioa
6020
ASTM
                                        V
    	I 	_._. ...			I	l__		I			
    2 0 0,0 0 0   4 0 0,0 0 0   6 0 0,00 0    8 0 0 T 0 0 C
                  Ni Spiked ug/q
                                                                1,000,00
-------
            THE "ART" OF SUCCESSFUL ANALYSES OF INORGANIC
                  CLP PERFORMANCE EVALUATION SAMPLES


Mark E.  Tatro, President, SPECTRA Spectroscopy & Chromatography
Specialists, Inc., P.O. Box 352, Pompton Lakes, New Jersey 07442


Most environmental laboratories view the passing of the EPA Contract
Laboratory Program [CLP] Performance Evaluation [PE] QA/QC sample
analysis as an almost impossible task. More laboratories today are
required to pass the CLP PE QA/QC analysis to bid on jobs submitted by
clients. There are strong indications that SW-846 regulations will
require CLP type QA/QC protocols to be followed for inorganics
analysis. Therefore,  instituting CLP QA/QC protocols for inorganics
analysis is of interest to non CLP environmental laboratories.

We have assisted numerous laboratories in instituting CLP QA/QC
protocols where we have stressed increased productivity through a
minimum of errors (1). This paper will detail our experience in
assisting laboratories to institute CLP QA/QC protocols for inorganics
analysis and discuss the common errors that lead to failure prior to
our on-site training support.

The analysis of the CLP PE samples for inorganics requires an ICP, a
Furnace AA and a cold vapor mercury analyzer. It will be assumed for
this paper that the methods for these systems have been developed
properly. This is not always the case as many laboratories approach
ICP from an AA background and fail to account for ICP interferences or
attempt to standardize the ICP with too many calibration standards. As
for Furnace AA analyses, many laboratories accept the method of
standard additions as a necessary evil and employ it routinely which
greatly reduces productivity. We always set up modern Furnace AA
systems to avoid the method of standard additions (2).

The first common error made is in sample preparation.  Consider that
the analyses of three CLP PE samples (blank, water and soil)  requires
the digestion of 26 samples using the hot plate methodology (Figure
1) .  Consider that the digested samples require pre-digestion spikes at
the ppb level and that the soil samples require separation of the
solids from the digestate liquid following digestion.  This requires a
level of organization and reduction of contamination that many
laboratories are not used to meeting. We will discuss the use of a
very simple acid washing protocol to reduce contamination and the use
of multi-element trace metal standards, designed by SPECTRA,  to easily
add the required pre-digestion spike standards with fixed volume
automatic pipets.

The most recent CLP inorganics statement of work issued in April 1990
(3)  has  approved closed vessel microwave digestion for the preparation
of waters and soils.  This has reduced the amount of digestions
required from 26 for  hot plate (Figure 1)  to 13 for microwave (Figure
2).  The  reason for this is that the hot plate technique requires a
separate digestion based on analyses by ICP (HNO3/H2O2/HC1)  or by
                                  n-229
-------
Furnace AA (HNO3/H202).  The microwave technique uses  only HNO^  and is
divided into Water and Soil categories since the Water  digestion
requires 5 ml HNO3 and the Soil digestion requires  10 ml  HN03.
Therefore, laboratories should convert from hot plate digestions  to
closed vessel microwave digestions.

The preparation of standards for CLP inorganic analyses is a major
source of lowered productivity. Different multi-element standard  mixes
are required for each CLP QA/QC requirement including ICP calibration,
Furnace AA calibration,  ICP Initial and Continuing  Calibration
Verification, Furnace AA Initial and Continuing Calibration
Verification, ICP Pre-Digestion Spikes, Furnace AA  Pre-Digestion
Spikes, Furnace AA Post-Digestion Spikes, ICP Interference Check, ICP
Inter-Element Correction and ICP 2xCRDL analyses.

SPECTRA developed the first set of CLP standards to meet  the entire
CLP QA/QC requirements for ICP and Furnace AA in 1986 (4).  Our  newest
standards designed for the 7/87 and 7/88 CLP Statement  of Work  can
also be used for the 4/90 ILM01.0 Statement of Work for both hot  plate
and microwave digestion. The standards are designed around the  use of
automatic pipets which greatly increases productivity and decreases
contamination errors. Figure 3 outlines the use of  the  SPECTRA  CLP
QA/QC multi-element standards for pre-digestion spiking of waters and
soils based on the 4/90 Statement of Work requirements  for hot  plate
and microwave digestion.

Examples of increasing productivity for both ICP and  Furnace AA
analyses of CLP PE samples will be discussed.


REFERENCES


(1) Tatro, M.E., Teaching the art of environmental  analysis,
    Environmental Laboratory, 1(5), 32 (1989).

(2) Tatro, M.E., Meeting Furnace AA CLP QA/QC Requirements For
    Inorganics Analyses, American Environmental Laboratory,  2(1), 44
    (1990) .

(3) USEPA Contract Laboratory Program Statement of  Work for Inorganic
    Analysis, Document Number ILM01.0  (April 27, 1990).

(4) Tatro, M.E., Methods Development Logic for CLP  Inorganics
    Analyses, American Laboratory. 20(12), 48 (1988).
                                  n-230
-------
    HOT PLATE DIGESTION SET UP FOR INORGANIC CLP P-E SAMPLES
                               ICP
         Water
         Water Duplicate
         Water + Pre-Digestion Spike
         Water Preparation Blank
         Aqueous Laboratory Control Sample

         Blank
         Blank Duplicate
         Blank + Pre-Digestion Spike

         Soil
         Soil Duplicate
         Soil + Pre-Digestion Spike
         Soil Preparation Blank
         Solid Laboratory Control Sample
                           FURNACE AA
         Water
         Water Duplicate
         Water + Pre-Digestion Spike
         Water Preparation Blank
         Aqueous Laboratory Control Sample

         Blank
         Blank Duplicate
         Blank + Pre-Digestion Spike

         Soil
         Soil Duplicate
         Soil + Pre-Digestion Spike
         Soil Preparation Blank
         Solid Laboratory Control Sample
         Total Digestions =26
FIGURE 1. Required preparations using hot plate digestion.
                             n-231
-------
    MICROWAVE DIGESTION SET UP FOR INORGANIC CLP P-E SAMPLES
                               WATER
         Water
         Water Duplicate
         Water + Pre-Digestion Spike
         Water Preparation Blank
         Aqueous Laboratory Control Sample

         Blank
         Blank Duplicate
         Blank + Pre-Digestion Spike
                              SOIL
         Soil
         Soil Duplicate
         Soil + Pre-Digestion Spike
         Soil Preparation Blank
         Solid Laboratory Control Sample
         Total Digestions = 13
FIGURE 2.  Required preparations using closed vessel microwave
          digestion.
                             H-232
-------
           USE OF SPECTRA CLP QA/QC MULTI-ELEMENT STANDARDS
           FOR HOT PLATE AND MICROWAVE PRE-DIGESTION SPIKES
                              HOT PLATE


       SPIKE REQUIREMENT                         ml ADDED


       ICP  - WATER                                 1.0

       ICP  - SOIL             -                     2.0

       FURNACE AA - WATER                          1.0

       FURNACE AA - SOIL                           2.0
                              MICROWAVE


        WATER                                      0. 5

        SOIL                                       1.0
FIGURE  3. Additions  of Pre-Digestion Spike Standard using the SPECTRA
         CLP  QA/QC  Multi-Element Kit for both hot plate and microwave
         digestion  procedures.
                                 H-233
-------
             STATE-OF-THE-ART SAMPLE  PREPARATION METHODS
                 FOR ENVIRONMENTAL  INORGANIC  ANALYSIS


Mark E. Tatro. President, SPECTRA Spectroscopy  & Chromatography
Specialists, Inc., P.O. Box  352, Pompton Lakes,  New Jersey 07442


The weak link in atomic spectroscopy  analysis continues to be sample
preparation. The EPA Contract Laboratory Program [CLP]  has recently
approved closed vessel microwave digestion  for  the  preparation of
waters and soils prior to ICP and Furnace AA  analysis (1).  However,
microwave digestion has not  yet been  approved by NPDES  and SW-846
regulations. Therefore, most environmental  laboratories still must
resort to antiquated hot plate methods for  the  preparation of samples
for inorganics analysis.

ICP analysis is far more productive than Furnace AA for trace metal
analysis. However, analysts  are forced to use Furnace AA to meet
required detection limits not attainable by ICP.  If samples were
concentrated prior to analysis to increase  the  analyte  concentrations,
then analysts could use the  more productive ICP  technique.  At this
time, EPA approves only evaporation of liquid samples to concentrate
samples. Evaporation techniques lead  to contamination from the
atmosphere and also concentrate ICP interfering  elements such as Na,
K, Ca and Mg.

The ICP and Furnace AA analysis of metals in  oil need not  be digested
if the oil is in a liquid state. A simple dilution  with xylene or
kerosene is that is needed to dissolve the  oil  sample prior to
analysis. This "dilute and shoot" technique is  favored  over digestion,
since it leads to a minimum  of contamination  and a  minimum of analyte
loss.

This paper will discuss three techniques of sample  preparation that
are state-of-the-art. The first technique,  closed vessel microwave
digestion, will be presented as adopted by  CLP.  The logic  of
developing a closed vessel microwave  digestion method will  be shown
from our development of a microwave digestion method for the
preparation of sewerage sludge prior  to ICP analysis (2).  The need to
characterize the sludge sample by hot plate digestion prior to
microwave digestion to obtain the "true" values  of  analytes extracted
by the EPA approved hot plate technique will  be  stressed.  The advent
of pressure control microwave digestion techniques  will also be
addressed.

The use of the Trace-Con automated extraction/concentration system
developed by Knapp (3)  will  be discussed as it  applies  to  the
concentration of waters prior to ICP  analysis. The  Trace-Con is a
computer controlled solid phase ion-exchange  resin  prepared from
either oxine/cellulose or EDTrA/cellulose that  is designed to
automatically concentrate transition  elements while allowing alkali
and alkaline-earth elements  to pass through the  ion exchange material.
                                  n-234
-------
Elution of the ion-exchange resin with a small volume of acid provides
a concentrated sample suitable for ICP analysis. The mechanism of the
Trace-Con as well as real world applications will be presented.

The "dilute and shoot" technique of inorganic analysis where sample
preparation is avoided will be discussed using an application
developed by SPECTRA for the Furnace AA analysis of copper in jet fuel
oils.  The method will demonstrate the need for computer graphics to
identify optimum conditions of analysis.


REFERENCES

(1) USEPA Contract Laboratory Program Statement of Work for Inorganic
    Analysis, Document Number ILM01.0 (April 27, 1990).

(2) Tatro, M.E., Sample Preparation: The Weak Link in Atomic
    Spectroscopy, Spectroscopy, 5(3), 14 (1990).

(3) Tatro, M.E., Sample Preparation Techniques for Inorganic
    Environmental Analyses, American Environmental Laboratory (in
    press).
                                  H-235
-------
90                    ANALYSIS OF ARSENIC, SELENIUM, AND MERCURY  IN
                 TCLP  EXTRACTS  OF  STABILIZED  HAZARDOUS WASTE BY HYDRIDE
               GENERATION/MULTI-ELEMENT  ICP OPTICAL EMISSION SPECTROSCOPY

         Dr.  Donald  R.  Hull,  Sr. Chemist, Rita A. Atwood,  Chemist,  and Dr.  Peter
         A.  Pospisil,  Manager,  Methods Development,  Chemical Waste  Management,
         Inc.,  150 West  137th  Street,  Riverdale,  Illinois  60627

         ABSTRACT

         Hydride  Generation-Inductively   Coupled  Plasma   (HG-ICP)   analysis  of
         Landban  TCLP  extracts  makes  possible  the  simultaneous  analysis  of
         arsenic,  selenium,  and mercury,  at  the ppb  levels,  with only  a  single
         sample preparation.   HG-ICP analysis streamlines  the overall  analytical
         procedure,  greatly  reduces the  time  required for analysis,  and reduces
         analytical  and  capital  equipment  costs.

         TCLP extracts  of  stabilized  hazardous wastes  are  an  excellent matrix for
         this technology  since they generally contain only low concentrations  of
         transition  metals,  which  can  cause  interferences  with  the  hydride
         generation  chemistry.   The  vapor generation  device  can be  quickly  and
         easily  connected  to  the  ICP  sample  introduction  system  and  fully
         automatic operation,  with  an  autosampler,  is  possible.

         INTRODUCTION

         The   analytical  sensitivity  requirements  for  arsenic,  selenium,  and
         mercury analysis,  as  set  forth by the Landban  legislation,  are given  in
         Tables I,  II,  and III.   Graphite Furnace Atomic Absorption  Spectroscopy
         (GFAA) is  usually the method  of  choice  for the analysis of  arsenic  and
         selenium,   while  Cold   Vapor  Generation   Atomic    Absorption   (CVAA)
         Spectroscopy  is  used  for  mercury.    These  methods  provide  adequate
         sensitivity,  accuracy,  and  precision foremost purposes,  but  are  very
         expensive  in  terms of  capital equipment  investments, labor  costs,  and
         analytical  time.

         Generation  of  volatile hydrides, by reaction with   sodium  tetraborate
         (NaBH4),  is commonly  used  for  sensitivity  enhancement  in   either  the
         atomic absorption  or  ICP  analysis of many metalloid  elements.  Hydrides
         of elements from  Groups IVA  (Ge,  Sn,  and Pb),  Group  VA (As,  Sb, Bi),  and
         VIB  (Se,  Te)  can  all  be produced under  appropriate conditions.   Mercury
         doesn't form  a  volatile hydride like the metalloids,  but can be reduced
         to the volatile metallic  state with NaBH4.

         Hydride  Generation Atomic Absorption  (HGAA)  methods  for  arsenic  and
         selenium  are  given  in  either  SW-8461  or  Standard  Methods  of  the
         Examination  of Water and  Wastewater ,   however  there are problems with
         these methods.   The SW-846  use  of metallic tin for  arsenic  or selenium
         hydride generation  makes  automation   virtually  impossible.   The Standard
                                           n-236
-------
Methods   required   sulfuric   acid   fuming   also   causes   selenium
volatilization losses.

The  use  of  NaBH4  instead  of metallic  tin,  to  produce arsenic  and
selenium hydrides  and to  reduce  mercury to  the metallic  state,  makes
automation of the hydride  generation  step easy.   Chemical interferences
in  the  hydride  generation step,  arising  from  high  concentrations  of
transition and noble  metals can be  a problem, but  careful  sample type
choice, and if necessary  the  use  of up to 50% HC1  solutions  to complex
the metals, can make this problem manageable.

Using a combined nitric and hydrochloric acids for digestion eliminates
the need for a separate mercury cLigestion and  analyzing arsenic from the
As1*"1"  state,   instead  of  the  As^^  state,  eliminates  the  need   for  a
sulfuric acid fuming.

Combining  hydride  generation  with   Multi-Element   Inductively  Coupled
Plasma-Optical Emission  Spectroscopic (HG-ICP)  provides  a means  for  a
more rapid analysis of these  three  elements  than with atomic absorption
instruments.    ICP  analysis, therefore, greatly  reduces  sample analysis
time, reduces the need for for atomic absorption equipment,  reduces the
number of QC samples that must be run, and reduces labor costs.

THEORY

Arsenic

Arsine (AsH3) is generated very rapidly from  arsenic  in  the  As3+ state
by reaction with NaBH^.   Borohydride  can also generate arsine from Asb+
by  first  reducing  it  to   As   ,  however this  reaction is much  slower,
generally  resulting  in  a  lower efficiency  of arsine production  and  a
loss  of  analytical  sensitivity.    Pre-reduction  of Asb+  to As   ,  by
reaction with Nal or KI,  can  be used  to  obtain maximum arsenic analysis
sensitivity,  however  iodide salts  interfere  with selenium  and  mercury
hydride reductions.  Arsine generation  interference  can  also  occur when
residual  NC^", from the  nitric acid  used  for sample  digestion,  reacts
with iodide to form N02,  which then reoxidizes Asd+ to As  .

Even though  arsine generation  from  As3+ is  more efficient,  there  are
advantages to performing  the  analysis with  arsenic in the Asb+  state.
Digestion of the samples with  nitric  acid conveniently insures that all
the arsenic present in the sample  solution  is present in  the plus five
oxidation state and the  elimination of the  iodide  reduction  simplifies
the sample  preparation  and  eliminates the  NC^  interference.   Arsenic
sensitivity can also  be  largely restored,  without  interfering  with the
production of hydrogen selenide and mercury vapor,  by  the use of a flow
injection  hydride  generator  with   a  reaction  coil   before   gas  liquid
separation.
                                   H-237
-------
Selenium

Hydrogen  selenide  production  is  only  possible  from  the  Se4+  state,
however selenium may exist  as  elemental  selenium (Seu), Se   , Se   , and
Se°+ in samples.   Satisfactory analysis of  selenium  can be  achieved by
first oxidizing of all the  selenium  in the samples to the Seb+  state by
using a nitric acid digestion  and then reducing  the  Seb+  to the Se4*
state  by  adding  hydrochloric  acid  and  boiling  gently  for  about  30
minutes.  Deterioration of  the  samples,  with slow oxidation  of  the Se4"1"
back to Se   ,  has  been  reported, however,  Se  standards have  been  found
to be stable for at  least  a week with  concentrations  of 10%  nitric and
10%, or more,  hydrochloric  acids.   Samples,  treated  as described  above
and with final  concentrations of 10% nitric and hydrochloric  acids, have
also been  found to  be  stable  for several  days,  however  storage time
should be  minimized after the hydrochloric acid reduction step.

Mercury

The high  vapor pressure  of mercury  first  lead to  its  determination  in
air by  Muller3 and Woodson4  in  1930 and  1939.   Poluektov  et al. ,  in
1964,  first  noted  an   enhancement  in  the  flame  atomic   absorption
determination  of mercury  in liquid  samples  when  stannous  chloride was
present.  This work, later,  lead to  the development of the mercury cold
vapor atomic absorption  analysis method, which uses stannous  chloride to
reduce Hg2+ to volatile  metallic mercury prior to its being swept  out of
the  solution  and   jnto  the  absorbance  cell  by  a   stream  of gas.
Schlesinger  et  al.   reported   that   sodium   borohydride  effected  the
reduction  of a variety of metaJ  ions,  including  mercury in 1953,  but it
wasn't until 1971  that  Braman'  reported using it  for the determination
of mercury in samples.

Interferences

The  hydride   generation  technique   is  prone  to   several   types  of
interferences  that  can  occur  in  either  the  hydride production  or
transportation stages.   Since the study, by Smith8, of  the effects  of 48
different  elements  on  the  hydride  determination of  antimony,  arsenic,
bismuth germanium,  selenium, and tin, a large number of  papers have been
published  concerning interferences and their control.
The most important interferences to  the determination of As,  Se,  and Hg
by hydride  generation are  those caused by transition  or  noble metals,
those  caused  by  volatile   nitrogen  oxides  or  chlorine,   and  physical
interactions   between   the    sample    vapors   and   the    apparatus.
Concentrations of  interfering  metals  above approximately  1  to 100 ppm,
depending  on the metal,  can  interfere in  the hydride  production  step,
causing a  lowering  of sensitivity.   If the metal  is in  sufficiently high
concentration,  it can even  be  reduced to  the  metallic  state which then
can react  with  mercury  vapor,  arsine,  or  hydrogen selenide  to form an
intermetallic  compound,   causing severe loss  of  sensitivity and  carry
over.   Metal interferences  can be largely controlled  by the use  of low
borohydride  concentrations,  high  concentrations  of  HC1,   and  masking
                                   H-238
-------
agents.  Unfortunately,  while a masking agent  such  as L-Cysteine works
well for arsenic,  it interferes  in  the determinations  of  selenium and
mercury.   The primary  means of  controlling metals  interferences are,
therefore,  to keep the metals concentrations as low as possible, use low
borohydride concentrations,  and use  higher  concentrations  of HC1   (up to
50%) to tie up the metals as chloro-complexes.

As described  above,  reactions between the  volatile  nitrogen oxide
and  I"  leads to the  production of  N02  which reoxidizes As3"1"  to As
which results in reduced sensitivity when the analysis is performed with
arsenic  in  the  AsJ+  state.   This problem  is  easily controlled  by
performing the analysis  with arsenic in the As5+ state.   N03"  is also
eliminated during  the selenium  reduction   step  when the  sample,  after
addition of  HC1,  is boiled.  The addition  of another reactant  such  as
sulfamic acidy  or  urea,  which react  with  NO^",  can also  be used  to
reduce or eliminate this interference.

If hydrogen peroxide, ^2,  is used  in the digestion, care must be taken
to completely destroy or eliminate all  traces of  this reagent  prior  to
adding HC1  and heating otherwise dissolved chlorine gas C12  is produced.
This C12 gas then  slowly  reoxidizes Se4+  to Seb+,  which  in unreactive
with borohydride.   This slow reoxidation  causes a  loss  of sensitivity
which increases with time.  This problem is most easily avoided by using
a   nitric   acid/hydrochloric  acid   digestion   instead   of  a   nitric
acid/hydrogen peroxide one.

As described  above,  when high concentrations of  interfering metals are
used, especially copper or noble metals, plating out of the metal  on the
reaction device  surfaces  can  occur,  producing  interferences  by  the
production of inter-metallic compounds.  Mercury  also  has  a tendency  to
adsorb or plate out on parts of the  reaction apparatus or the gas/liquid
separator,   especially   if   the   parts   are  dry,   when   the  Hg   vapor
concentration  is   high,   and    then   revolatilize   when   the   vapor
concentration is low.   This leads to severe  problems  of carry  over and
drift.  Hydrogen  selenide also seems to readily dissolve in this mercury
film, possibly being reduced to metallic Se at the same time, leading to
a very large  and irrecoverable  loss  of  sensitivity.   These problems are
controllable by making all  samples,  standards,  and the rinse solution a
mixture of nitric and hydrochloric acids and using the nebulizer  as the
gas liquid separator.  The HN03/HC1 mixture prevents any mercury plating
out in the reaction  portions of the hydride  generator and  the constant
spraying of  this mixture by the nebulizer  seems  to  prevent  any plating
out in the spray  chamber or torch.

EXPERIMENTAL SECTION

Apparatus

The  hydride  vapor  generator   can  be  constructed,  using  a 4   roller
peristaltic pump  and a manifold assembled according  to Figures 1  and 2,
or a commercial  system  such as the  Varian  VGA-76 may be  used.   If the
                                   H-239
-------
VGA-76  is  used the  sample  and  drain  pump tubes  should  be  placed on
inside  position  of rollers  and the  sample pump  tube  on  the  outside.
(Varian,  Instrument  Group,  220 Humboldt   Court,  Sunnyvale,  CA  94089,
(800)231-8134.)

A  Leeman   Labs  Plasma-Spec  I  sequential   Inductively  Coupled   Plasma,
equipped with a Thermo Jarrell Ash  fixed cross-flow nebulizer, was  used
for  all   sample  analysis  work  (Leeman Labs,  Inc.,  600  Suffolk  St.,
Lowell, MA  01854,  (508)454-4442 and  Thermo Jarrell Ash Corporation, 8
East Forge Parkway, Franklin, MA 02038-9101, (508)520-1880.)

Reagents

Reagent water, defined as water  in  which an interferant is not  observed
at, or  above, the  methods detection  limit  of the analyte(s) of  interest
was used throughout.   Typically this water was equivalent to ASTM  Type I
water.  All reagents  used were ACS reagent grade or better.

Sample Preparation

If the TCLP extract of the stabilized waste contains undissolved  solids,
the  extract  must   be  well  shaken  prior to  analysis.    A  50  mL  sample
aliquot is transferred to a 125 mL Erlenmeyer flask.

When all samples are  transferred, add 5 mL of concentrated  HNOo  and  5 mL
of concentrated HC1 to the  flasks,  hang a glass hook  on the Tip of the
flask to speed up  evaporation, and  cover them  with a reflux cap.  Place
the samples on a hot  plate and, with gentle refluxing, evaporate them to
near  dryness,  making  certain  that the  samples are  not allowed  to go
completely dry.  Cool  the flasks and,  if the digestion  is  not complete,
add additional 3-mL portions  of concentrated HNO^  and  HC1.  Return the
samples to the hot plate for continued refluxing and evaporation.

Continue the  heating,  with  additional  acids being  added  as necessary,
until the digestion is complete  (generally indicated when  the digestate
is  light  in  color or  does  not  change  in  appearance  with  continued
refluxing).  Again, evaporate  the  sample to near dryness and cool.  Add
5 mL of HCL, 10 mL of reagent  water,  and 1 g or urea.    Remove the glass
hook  and  heat, at a  gentle  simmer,  for  30  minutes   to  dissolve  any
precipitate  or  residue  resulting  from  evaporation and to  reduce  the
selenium to  the Se^+  state  for analysis.   Cool  and add additional  HN03
to bring the total  HN03 to 5 mL (about 2-3 mL).

Transfer the  digested sample  to a  50  mL  volumetric  flask.   Wash the
digestion flask and  reflux cap  carefully  into the  volumetric flask to
insure that  sample transfer  is  quantitative.   After dilution to  volume,
if necessary, filter the sample  (Whatman #1 filter paper or equivalent)
to remove any  undissolved  particles that  might  clog the ICP nebulizer.
The  sample  is now ready for  analysis.    Because  of  the  potential   for
degradation  of the   sample,   samples  should  be  analyzed  as  soon as
                                   n-240
-------
possible after  the final  heating  and dilution.   Maximum  storage time
before analysis should be less than 48 hours.

Connection of Hydride Generator to ICP

Minor modifications are made  to the  Thermo  Jarrell  Ash fixed cross flow
nebulizer and  to  the Leeman  ICP  spray chamber to  facilitate switching
between normal  sample  nebulization and hydride generation.   With these
modifications,  switching   between  the  sample  input  devices will  not
require  shutting   down  the   ICP  or  making  any  changes   in  operating
conditions.

A  section  of  Leeman  sample  pump  tubing .about  3/16"  long  is  used  to
attach about 8" of the  1/16"  0.  D. Teflonm  tubing  to the  sample inlet
of the nebulizer.   This tubing extends outside the torch box where it is
connected to either  the  Leeman sample pump tubing or  the  reaction coil
of the hydride  generator.   By turning off  the nebulizer  argon gas flow
for  only  a  few seconds  it  is possible  to  switch  between  the Leeman
sample pump tubing or the reaction coil of the hydride generator.

About 8"  of the 1/8" 0.  D.  Teflon™  tubing  is  attached to  the Leeman
spray chamber  drain  and  extended  outside  the torch  box  where  it  is
connected to either  the Leeman drain  pump  tubing or  the  hydride drain
pump tubing.

Operating Conditions

Hydride  Generation-Multi/element   Inductively  Coupled  Plasma-Optical
Emission Spectroscopy instrumentation and operating conditions are given
in Table IV.

Calibration

A  linear calibration  is  typically obtained with  four  standards  and the
range limited to a maximum of 100 ppb to insure greatest accuracy in the
low ppb range.  Typical calibration curves  are shown in Figure 3.  With
the use of quadric fitting, the calibration range  can be extended to 500
to 1000 ppb, however  carry over  between samples  or  standards becomes a
significant problem.

Analytical  Results for Undigested Standards

Leeman ICP Sample trays containing 40  alternating blanks  and standards,
containing 25 ppb  of As,  Se,  and  Hg and 2500 ppm of Ca to  matrix match
the TCLP extracts, were analyzed on 5  different days.   The  25 ppb level
was chosen  for  the standard to match  the lowest  regulation  levels for
TCLP extracts for Se and Hg.   The  daily  and overall  average results are
shown in Table V.   These  results  indicate  that the  Instrument Detection
Limits (IDL's)  for the three elements are between  2 and 2.5 ppb, the Per
Cent Relative Standard Deviation (%RSD) at  25 ppb is between 3% and 7%,
and that an  accuracy of +20% at 25 ppb is possible at the 50 ppb level.
                                   n-24i
-------
Analytical Results for Digested Standards

Twenty 50 mi aliquots of a mixed  standard,  containing 50 ppb of As,  Se,
and  Hg  and  2500  ppm  of  Ca  to  matrix  match  the TCLP  extracts, were
prepared  according   to  the  sample  preparation  method  given  in this
report.    These  20   standards  aliquots  were  then placed  in  a  Leeman
autosampler tray,  alternating with  20  blanks,  and analyzed.  Over  three
sets of  analysis  of these digested  standards  42  results  for As and 52
results  for Se and  Hg  were  obtained.  The  results,  shown at the  bottom
of Table  V,  are  similar to  those for the  undigested  standard and show
that the %RSD's range between about  3% and  7% with an accuracy of  about
20%.

CONCLUSIONS

1.    Hydride Generation/Multi-Element ICP Optical Emission Spectroscopy
can be shown to provide sufficient  sensitivity,  accuracy, and precision
to allow analysis of arsenic, selenium,  and mercury in TCLP extracts of
stabilized hazardous wastes  at the ppb levels.

2.    Multi-Element  Hydride  Generation/ICP Optical Emission Spectroscopy
is  much  more  rapid than  single  element  atomic  absorption  analysis,
requiring only 4  minutes of  instrument time instead of about 20 minutes.
3.    Additional  time  savings are
only be run once, not three times.
                                achieved  since all  QC samples  need
4.    Sample digestion with  combined nitric and  hydrochloric acids can
be  shown  to  produce  excellent  recoveries  of  arsenic,   selenium,  and
mercury.
5.    The
eliminated
digestion.
       need  of  a  separate  sample
        by  the  use  of  a  combined
digestion  for  mercury  can  be
nitric  and  hydrochloric acids
6.    By using existing ICP equipment, the purchase of additional atomic
absorption can be avoided, reducing capital equipment costs.

REFERENCES
1
3
4
Test Methods  for  Evaluating  Solid Waste,  Physical/Chemical  Methods,
2rd  ed.,  U.  S.   Environmental  Protection  Agency,  Office of  Solid
Waste  and  Emergency  Response.    U.  S.  Government  printing  Office,
Washington,  DC, 1982, SW-846.
Standard Methods  for the Examination  of Water and
Edition, American  Public Health  Association,  1015
NW, Washington, DC 20005.
K. Muller,  Z.  Phys., 65, 739 (1930).
T. T. Woodson, Rev. Sci. Instrum., 10, 308  (1939).
                                                        Wastewater,  17th
                                                        Fifteenth  Street
                                   n-242
-------
b  N. S.  Poluektov,  R.  A.  Vitkun,  and T.  V.  Zelyukova, Ah. Anal. Khim.,
   19, 873  (1964).
6  H.  I.  Schlesinger, H.  C.  Brown,  A. E. Finholt,  J.  R.  Gilbreath, H.
   R. Hoekstra,  and  E.  K.  Hyde,  J. Am. Chem.  Soc.,  75,  215 (1953).
;  R. S.  Braman,  Anal.  Chem.,  43,  1462 (1971).
°  A. E.  Smith,  Analyst  (London),  100, 300-306  (1975).
9  R.  M.  Brown,  Jr.,  R.  C.  Fry,  J.   L.  Moyers,  S. J. Northway,  M. B.
   Denton,  and  G.  S.  Wilson,  Anal. Chem., 53,  1560-1566 (1981).
                                   H-243
-------
                            TABLE I

                 ARSENIC HAZARDOUS WASTE CODES AND
                       TREATMENT STANDARDS
WASTE CODE
CALIF
D004
K031
K048
K049
K050
K051
K052
K084
K101
K102
LEACHATE
P010
P011
P012
P036
P038
U136
D004
K031
K084
K101
K101
K102
K102
LEACHATE
P010
P011
P012
P036
P038
U136
WW/ TREATMENT CCW/CCWE1 REGULATION
NWW2 STANDARD SOURCE^
LIQ 500. ppm CCW CALIF
NWW
NWW
NWW
NWW
NWW
NWW
NWW
NWW
NWW
NWW
NWW
NWW
NWW
NWW
NWW
NWW
NWW
WW
WW
WW
WW
WW
WW
WW
WW
WW
WW
WW
WW
WW
WW
5.6 ppm
5.6 ppm
0.004ppm
0.004ppm
0.004ppm
0.004ppm
0.004ppm
5.6 ppm
5.6 ppm
5.6 ppm
5.6 ppm
5.6 ppm
5.6 ppm
5.6 ppm
5.6 ppm
5.6 ppm
5.6 ppm
0.79 ppm
0.79 ppm
0.79 ppm
2. ppm
0.79 ppm
2. ppm
0.79 ppm
1.39 ppm
0.79 ppm
0.79 ppm
0.79 ppm
0.79 ppm
0.79 ppm
0.79 ppm
CCWE
CCWE
CCWE
CCWE
CCWE
CCWE
CCWE
CCWE
CCWE
CCWE
CCWE
CCWE
CCWE
CCWE
CCWE
CCWE
CCWE
CCW
CCW
CCW
CCW
CCW
CCW
CCW
CCW
CCW
CCW
CCW
CCW
CCW
CCW
3rd 3rd
3rd 3rd
1st 3rd
1st 3rd
1st 3rd
1st 3rd
1st 3rd
3rd 3rd
3rd 3rd
3rd 3rd
3rd 3rd
3rd 3rd
3rd 3rd
3rd 3rd
3rd 3rd
3rd 3rd
3rd 3rd
3rd 3rd
3rd 3rd
3rd 3rd
Ird 3rd
3rd 3rd
Ird 3rd
3rd 3rd
3rd 3rd
3rd 3rd
3rd 3rd
3rd 3rd
3rd 3rd
3rd 3rd
3rd 3rd
  Constituent Concentration in Waste  (CCW)  or Constituent
  Concentration  in  Waste Extract  (CCWE).
^ Waste Water  (WW)  or Non-Waste Water  (NWW).
  California Listed Waste (CALIF) or  1st,  2nd, or 3rd third
  of Landban.
                              H-244
-------
                           TABLE II

                SELENIUM HAZARDOUS WASTE CODES AND
                       TREATMENT  STANDARDS
WASTE CODE
CALIF
D010
K048
K049
K050
K051
K052
LEACHATE
P103
P104
U204
U205
D010
LEACHATE
P103
P114
U204
U205
WWA TREATMENT CCW/CCWE1
NWW2 STANDARD
LIQ
NWW
NWW
NWW
NWW
NWW
NWW
NWW
NWW
NWW
NWW
NWW
WW
WW
WW
WW
WW
WW
20. ppm CCW
5.6 ppm
0.025ppm
0.025ppm
0.025ppm
0.025ppm
0.025ppm
5.6 ppm
5.6 ppm
5.6 ppm
5.6 ppm
5.6 ppm
0.79 ppm
0.82 ppm
0.79 ppm
0.79 ppm
0.79 ppm
0.79 ppm

CCWE
CCWE
CCWE
CCWE
CCWE
CCWE
CCWE
CCWE
CCWE
CCWE
CCWE
CCW
CCW
CCW
CCW
CCW
CCW
REGULATION
SOURCE3
CALIF
3rd 3rd
1st 3rd
1st 3rd
1st 3rd
1st 3rd
1st 3rd
3rd 3rd
3rd 3rd
3rd 3rd
3rd 3rd
3rd 3rd
3rd 3rd
3rd 3rd
3rd 3rd
3rd 3rd
3rd 3rd
3rd 3rd
1 Constituent Concentration in Waste  (CCW)  or Constituent
  Concentration  in Waste Extract  (CCWE).
2 Waste Water  (WW)  or Non-Waste Water  (NWW).
3 California Listed Waste (CALIF) or 1st,  2nd, or 3rd third
  of Landban.
                              H-245
-------
                            TABLE III
                 MERCURY HAZARDOUS WASTE CODES AND
                       TREATMENT STANDARDS
WASTE CODE WWA TREATMENT CCW/CCWE1 REGULATION
NWW2 . STANDARD . SOURCE"3
D009
K071
K106
U151
NWW'*
NWW;
NWW;
NWW4
RORTIN0
RORT7
RORT
RORT
MOT6
MOT
MOT
MOT
3rd 3rd
3rd 3rd
3rd 3rd
3rd 3rd
CALIF
D009
K071
K071
K106
LEACHATE
U151
D009
K071
K101
K101
K102
K102
K106
LEACHATE
P065
P092
U151

P065

P092
          LIQ
          NWW
          NWW
          WW
          WW
          WW
          WW
          WW
          WW
          WW
          WW
          WW
          WW
          WW

          NWW8

          NWW10
20. ppm CCW
0.025ppm
0.025ppm
0.025ppm
0.025ppm
0.2 ppm
0.025ppm
0.03 ppm
0.03 ppm
0.027ppm
0.082ppm
0.027ppm
0.082ppm
0.03 ppm
0. 15 ppm
0.03 ppm
0.03 ppm
0.03 ppm

CCWE
CCWE
CCWE
CCWE
CCWE
CCWE
CCW
CCW
CCW
CCW
CCW
CCW
CCW
CCW
CCW
CCW
CCW
INRORT9
INRORT
CALIF
CCWE
CCWE
CCWE
CCWE
CCWE
CCWE
CCW
CCW
CCW
CCW
CCW
CCW
CCW
CCW
CCW
CCW
CCW
MOT
MOT
3rd 3rd
1st 3rd
3rd 3rd
3rd 3rd
3rd 3rd
3rd 3rd
3rd 3rd
1st 3rd
1st 3rd
3rd 3rd
1st 3rd
3rd 3rd
3rd 3rd
3rd 3rd
3rd 3rd
3rd 3rd
3rd 3rd
3rd 3rd
3rd 3rd
2
3

4
5

6
7
10
Constituent  Concentration in Waste (CCW) or Constituent
Concentration in Waste Extract  (CCWE).
Waste Water  (WW)  or Non-Waste Water (NWW).
California Listed Waste  (CALIF)  or 1st, 2nd, or  3rd third
of Landban.
High Mercury Subcategory
Roasting  or  retorting; or incineration followed  by
roasting  or  retorting as a method of treatment.
Method of Treatment (MOT).
Roasting  or  retorting as methods of treatment.
Mercury Fulminate.
Incineration followed by roasting or retorting as a
method of treatment.
Phenyl Mercury Acetate.
                              n-246
-------
                                TABLE IV
     Instrumentation
  Component
  Model/size .
ICP (Sequential)   Plasma-Spec I
Nebulizer
Fixed Cross-flow
Vapor Generator   VGA-76

or

Peristaltic Pump   Rabbit or Rabbit
                   Plus

Hydride Manifold   99-100399 (or lab.
                   constructed)
  Manufacturer
Leeman Labs, Inc.
Lowel1,  MA

Thermo Jarrell-Ash
Franklin, MA

     Varian Instruments
Sunnyvale, CA
                       Rannin Instrument
                       Woburn, MA

                       Varian Instruments
                       Sunnyvale, CA
     Emission Wavelengths
Leeman ICP Line
As3 (with Background Position Bl)
Sel (with Background Position Bl)
Hgl (with Background Position Bl)
Wavelength
234.984 nm
203.985 nm
253.652 nm
     Operational  Parameters
  Component
                         Setting
Power (@ 0.5 amps)
Outer Torch Flow
Intermediate Torch Flow
Sample Torch Flow (@ 37.5 psig)
Sodium Borohydride (1% w/V)
Sample (10% HNOo/10% HC1)
Wavelength Update (between each sample)
Rinse Time
Delay Time
Update Frequency (Ul and U2)
Analysis Times:
     As  and Hg
     Se
                       1000 Watts
                       12 L/minute
                       0 L/minute
                       0.5 L/minute
                       1 mL/minute
                       8 mL/minute
                         3  minutes
                       99 seconds
                       60 seconds
                       10 samples

                       3x3 seconds
                       3x6 seconds
                                   E-247
-------
                                 TABLE V



               ANALYTICAL  RESULTS FOR UNDIGESTED STANDARDS
CALIBRATION BLANKS



Date 	As(ppb)
Se(ppb)
3/13/90
3/14/90
3/15/90
3/16/90
3/19/90
Average
IDL (3 Std.
-3.16 + 0.43 -1.7 + 0.77 -5.55 + 1.16
-1.73 + 0.71 1.68 + 0.83 -2.50 + 0.93
-2.15 + 0.94 -0.11 + 0.56 -3.51 + 0.60
-0.53 + 1.20 -0.56 + 0.65 -0.01 + 0.30
-1.54 + 0.84 -0.12 ± 0.70 -2.31 ± 0.62
-1.54 0.84 -0.12 0.70 -2.31 0.62
Dev.) 2.5 ppb 2.1 ppb 1.9 ppb
UNDIGESTED 25 ppb MIXED STANDARD CONTAINING 2500 PPM Ca



Date               As(ppb)	Se(ppb)	Hq(ppb)
3/13/90
3/14/90
3/15/90
3/16/90
3/19/90
Averages
%RSD's
21.9 + 1.28
22.6 + 0.78
22.4 + 1.57
21.6 + 2.35
24.8 ± 1.11
22.7 1.42
6.3%
21.2 + 1.60
23.6 + 1.25
23.8 + 0.78
25.8 + 1.07
29.7 ± 0.89
24.8 1.12
4.7%
21.4 + 0.73
23.6 + 1.02
23.0 + 0.49
24.2 + 0.43
24.8 ± 0.31
23.40 0.60
2.6%
DIGESTED 50 ppb MIXED STANDARD CONTAINING 2500 PPM Ca



Date	As(ppb)	Se(ppb)
4/6/90
4/9/90
4/9/90
44.6 + 4.11
54.8 + 4.83
47.1 + 2.87
42.6 + 2.29
47.2 + 1.78
47.3 + 1.97
42.8 + 0.92
43.0 + 1.10
43.1 + 0.91
Averages        48.8 +  3.94    45.7 ± 2.01   43.0 + 0.98



%RSD's                 6.8%           4.4%          2.3%
                                 H-248
-------
" OD  x
Vlton
Tube
     0.030"
     Pump
     Tube
,042"OD
,022"ID     ->
Teflon Tube
                    " ID
V8"OD x 1/16" ID  X
Teflon Reaction  Coll
                                   -fr    To
                                      Nebulizer
               <- 0,081"  Pump Tube
                                          t
                                         0.110"
                                         Pump
                                         Tube
                                             <— Drain
           NaBH,     Sample
           1%   *
        FIGURE 1 .  HYDRIDE GENERATOR MANIFOLD
                  Pe r is ta 11ic

                     P u np
Borohydride Sanple

 1 nLXmin  8 n»L/nin Drain
                Bottle
   FIGURE 2.  HYDRIDE GENERATION - ICP  ASSEMBLY
                            H-249
-------
t-o
Ul
o
       D
       o
       z ^
       Q o
       UJ
                   o
                                HYDRIDE GENERATION - ICP ANALYSIS

                                    CALIBRATION CURVES FOR AS, SE, AND HG
25              50


         CONCENTRATION (ppb)

D  As      -I-   Se      O  Hg
75
1OO
                                            FIGURE 3 -
-------
91                     X-RAY FLUORESCENCE SPECTROSCOPY
              IN HAZARDOUS WASTE AND CONTAMINATED SOIL ANALYSES

        Douglas  S.   Kendall,   National  Enforcement  Investigations
        Center, U.S.  Environmental  Protection Agency,  Building  53,
        Box 25227,   Denver Federal Center,  Denver,  Colorado 80225

        ABSTRACT

        X-Ray  fluorescence  (XRF)  spectroscopy  has  a  number  of
        advantages  which recommend it for the determination of toxic
        metals  in   wastes and  soils.    Salient  features  of  the
        instrumentation   will   be   described.      Laboratory   XRF
        instruments can  make  precise,  rapid measurements,  and,  if
        properly calibrated,  accurate  ones.   Portable  instruments
        can find hot  spots at  waste sites  and  aid in  designing a
        sampling plan.  The determination of lead,  arsenic,  cadmium
        and other  metals in contaminated  soils will  be described.
        The   fundamental   parameters  approach  to   "standardless"
        analysis is very  promising  for  the  semiguantitative  (within
        50%)   determination of  metals  in   completely  unknown  and
        unspecified matrices.   The Jease  of sample  preparation  for
        XRF  spectroscopy  and   the   ability to  deal  with  problem
        matrices is illustrated by the analysis  of waste oil.

        INTRODUCTION

        X-ray fluorescence (XRF)  spectrometry has  an important role
        in the analysis of hazardous wastes.  Its potential is still
        being  developed,   and   there are  many  elemental  analysis
        problems   involving   hazardous   waste   in   which   more
        consideration should be given to XRF.

        XRF  spectrometry  has   long  had   an   important  niche  in
        industrial   chemical analysis.    Steel  mills,  foundries  and
        non-ferrous metal  smelters  use XRF for rapid,  precise  and
        unbiased analyses.   The mining  industry  uses  XRF  in  all
        aspects  of  its  work  from  prospecting  to  assaying  the
        finished products.    The  petroleum  industry  uses  XRF  to
        accurately  determine metals in petroleum products.

        XRF could be almost as useful to those doing hazardous waste
        analyses, if  it  were  properly understood  and  more widely
        available.   Environmental analyses require great sensitivity
        and XRF  does not always offer  sufficient  sensitivity,  as
        detection limits  are  in the 1  to  100 ppm range.   However,
        hazardous waste  analysis is very  different from measuring
        natural levels in the environment,  and  is  in many ways more
        akin  to  the analysis  of commercial materials.    For  those
        analyses in which significant concentrations of heavy metals
                                   n-25i
-------
or  other  target  elements  are present,   conventional XRF
instrumentation  very  often  has sufficiently  low  detection
limits.

Once it  is  established that XRF has sufficient  sensitivity
for  a   particular  hazardous  waste   analysis,   the  many
advantages of XRF  determinations can be appreciated.   Often
very little sample preparation  is needed.   Solids or liquids
can  simply  be   placed  in  disposable  cups  with  an  X-ray
transparent  window and  analyzed directly.   Viscous  waste
oils  and   paint  sludges  can be  characterized   without
problematic digestions such  as  those required by AA or ICP.
The easy sample preparation makes XRF spectrometry ideal for
rapid qualitative analysis.

XRF instrumentation is usually  quite trouble free and easily
maintained.   It  is very stable and holds  calibration  well.
The precision of repeated measurements  is usually very  good.
It  is  multielement,   in  either  a  simultaneous  or  rapid
sequential fashion.   Computer  controlled spectrometers with
sample changers  can analyze many samples while  unattended.
A wide variety of quantitative  methods  are  available.   Often
methods developed for industrial materials  can be adapted to
hazardous wastes.   The  above  factors  combine to make XRF
very  useful  for  efficiently  analyzing  a  large  number  of
similar samples,  such as determining toxic  metals in a  large
number of soil samples or used  oil  samples.

The remainder of the  paper  describes  some applications  of
XRF   spectrometry   to  hazardous   waste   analysis.    The
advantages  of XRF  and areas  in which it is particularly
useful are illustrated.

FUNDAMENTALS

X-rays are a form of radiation  more energetic and of shorter
wavelength  than   visible  and  ultraviolet   light.    The
wavelengths   of   X-rays   are   traditionally   measured  in
Angstroms  (1  Angstrom  equals  10~10  meters),  although the
more correct  metric unit  is  nanometers (10 Angstroms equals
1 nanometer).  The shorter the wavelength  of  an X-ray, the
higher  the   energy.     X-ray  energies   are  measured  in
kiloelectron volts  (keV).  Dividing 12.4 by an energy in keV
gives the wavelength in Angstroms.   Of  course, dividing 12.4
by a wavelength  in Angstroms gives the energy of the  X-ray
in keV-   Analytically useful  X-rays are in  the  range from
about 0.2 to  20  Angstroms or from  about  .6 to 60 keV.   For
example,   the  K  alpha  emission   of   sulfur  is  at  5.373
Angstroms (2.31 keV), the  K  alpha  emission of cadmium  is at
0.54 Angstroms (23.0 keV) and the L beta emission of lead is
at 0.98 Angstroms  (12.7 keV).
                            H-252
-------
The  X-ray emission  spectrum  of  an  element  is  relatively
simple compared to ultraviolet and visible emission spectra
(which are  used in  ICP).    A regular pattern  is followed.
Within a  given series,  such as  the  K  lines,  the emission
energy  increases   with   increasing   atomic   number    (the
wavelength gets shorter).
When  X-rays  irradiate  a  material,  they  can  be  either
absorbed, scattered or transmitted.  Absorbed X-rays lead to
fluorescence   or   emission   of   characteristic   X-rays.
Scattered X-rays  contribute to the background.   The amount
and type  of scattering  is  a source of  information about a
sample.   Absorption  and scattering  coefficients  are  well
known for almost  all  elements.   It is  possible to use these
values  in  calculations  of  matrix  effects  which  greatly
reduce  the   need  for  matrix  matching   of   samples  and
standards.    Such  calculations  are  known  as  fundamental
parameters methods  and  are  described more completely in the
section on standardless analysis.

XRF  spectrometry  is  an elemental technique.    The results
determined  by XRF  are  the  total amount  of  a particular
element  present   in  a sample, regardless of  what compound
contains  the element.    This is  an  advantage  when  total
amounts are  of  interest, as there is no digestion to cause
incomplete recoveries.   It  is a  disadvantage if  the concern
is with amount extractable by a certain digestion.

INSTRUMENTATION

There  are two  types of X-ray  fluorescence  spectrometers,
energy   dispersive   and   wavelength    dispersive.      The
classification is based on  the  method by which the X-rays
are detected.  Wavelength dispersive spectrometers have been
in analytical laboratories for a longer time, and use a high
powered X-ray tube  to irradiate a sample  with X-rays.   The
fluorescent X-rays, which are characteristic of the elements
present  in  the  sample,  are dispersed  by  the  analyzing
crystal,  where   Bragg  diffraction  separates   X-rays  by
wavelength.   In  a sequential spectrometer the detector and
analyzing  crystal  are  moved   in  concert  so   that  each
wavelength  is individually  focused  on  the detector.   The
detector, either  a  scintillation or  a proportional counter,
counts  all  the X-rays  which reach  it.    In  a simultaneous
spectrometer  there  is a  separate  crystal and  detector for
each element  of  interest.    A typical  wavelength dispersive
sequential  spectrometer  will cost  about $150,000  and  a
simultaneous spectrometer over $200,000.
                            H-253
-------
One  product  of  research  on  radioactivity  and  transuranic
elements  is  the  energy  dispersive  detector,   a  lithium
drifted  silicon semiconductor  (the  SiLi detector).    This
detector produces a current proportional  to  the  energy of an
X-ray which  strikes  it.    It  can  resolve the  characteristic
X-rays  of  the  elements  and  is  used in energy  dispersive
spectrometers  as  both  the   dispersive  element   and the
detector.    As  in  wavelength  dispersive  instruments, the
sample  is  irradiated by X-rays produced by an X-ray  tube.
The SiLi detector is placed close to the  sample  and receives
X-rays  of  all  energies.    X-rays  are  counted  and  sorted by
energy.  A multi-channel analyzer accumulates the  counts to
produce  a  spectrum.    An energy  dispersive  spectrometer is
naturally  multi-element,  although  to  collect  a  complete
spectrum several sets  of excitation conditions may be  used.
An   energy  dispersive   XRF   spectrometer   suitable  for
laboratory use costs $50,000 to $125,000.

Portable XRF spectrometers can be used in the  field at  waste
sites   and  to  survey   quite   large   areas   for   soil
contamination.   In their current  state of development,  they
have   much  poorer    resolution    and   considerably   less
sensitivity  than  laboratory  instruments.     Most  of  them
employ a rather crude energy dispersive detector.   They work
best  with  a  known,  unchanging  sample matrix.     Matrix
matching of  standards  is highly desirable,   perhaps obtained
by  using an  alternative  technique  to  analyze  preliminary
samples  from a site.   The  best  use  for  these  portable
instruments  is  for  finding  hot  spots and  for  providing a
large  number  of  semiquantitative  results  as  an  aid in
sampling.   It  should  be  remembered that the overall  error
from such field measurements  is no better than  the sampling
error.

A  few  generalizations which  compare  energy dispersive and
wavelength  dispersive  instruments  will be  given.    The
treatment  is far  from exhaustive,  but  serves  to  contrast
relative strengths and weakeners.  It should be noted that
there  is a considerable body of  knowledge  on XRF methods,
and the  literature should be consulted when  new  problems are
approached.

Wavelength dispersive spectrometers  are ideal  for  studies
which involve determining  5 to  10 elements in  a  large number
of samples.  Such an effort would justify the  preparation or
purchase of  a   suitable  number of  matrix matched  standards
and  careful  calibration  of  the  instrument.    A  typical
measurement  scheme  would  use  about one  minute per  element
per  sample  for trace  elements.   This  one  minute includes
both peak  and  background  measurement.   The calculation of
                            H-254
-------
the final  concentrations would be  almost  instantaneous and
would be available as soon as each sample was completed.

Since   energy   dispersive   instruments   are   inherently
multielement,  hey   are  ideal  for   the   qualitative  and
semiquantitative analysis of  complete  unknowns  and one of a
kind samples,  such as  might  be found at  an  abandoned drum
site.   Energy dispersive  instruments  are  also  well suited
for all types of quantitative analyses.  It may be necessary
to employ  several  excitation conditions to achieve optimum
results.   For instance, chromium,  lead and cadmium are far
enough  apart  in  the  periodic table to  usually  require
different operating  conditions  for  the instrument.  Because
the  deconvolution  of  complex  spectra is  often necessary,
data processing  can  consume a  significant portion  of the
analysis time.   All  in all,  energy dispersive spectrometers
can  make the same  quantitative measurements  as  the  more
expensive wavelength dispersive instruments.
SOILS CONTAMINATED WITH HEAVY METALS

The  contamination of  soils by  toxic  metals  is  a serious
environmental problem, which often involves large areas.  It
is  important that  the contaminated  area be  well defined.
After remedial   actions,  careful measurements  must be made
in  order  to certify  the  success of  the  abatement process.
X-ray fluorescence spectrometry  is  an effective way to meet
the  analytical  challenges  posed by  this type  of problem.
XRF  has several  advantages over  AA  or  ICP.    It  is less
operator   dependent   and  has   a   much  simpler   sample
preparation.

Prior to laboratory measurements, portable XRF equipment can
be used to perform a preliminary survey of a site.  The main
sampling can be better planned  and executed  if some basic
facts about the site are known.  A preliminary survey  with a
portable  XRF spectrometer  can  set the  boundaries  of  the
contaminated area, find hot spots,  and establish gradients.
This  type   of   information   will  greatly   increase  the
likelihood  that  the principal  sampling will  be proper and
valid.

Soil  or rock  samples can  be analyzed as ground powders,
after  drying and then  grinding them to  a   uniform size.
This, of  course,  should  be done no  matter what analytical
technique is applied.  A five to  ten gram portion  of the
sample  is  then placed  in a disposable plastic cup  with a
thin  mylar  or  polypropylene   window,   which  is  X-ray
transparent.  The XRF  analysis  is  done on this  sample, non-
destructively, and can easily be repeated.
                            H-255
-------
Obtaining suitable standards is the main  difficulty with XRF
measurements  of  contaminated  soils.    The  most  accurate
calibrations are done with standards  which  closely match the
samples.       Ideally,   the  standards   should   span   the
concentration range  of the analytes  and  the major elements
in the  standards should  match those in the  samples.    The
parameters  in  the calibration equations,  which may  contain
interelement  corrections,  are  adjusted  by  least  squares
techniques  to  fit the standards.    In order  to  accurately
determine   the   parameters,   there   should  be  many   more
standards than parameters.  Several standards should  be  left
out of the calibration and then used  to test the final fit.

One approach  is  to  take  samples from a  similar project or
from a  previous  sampling of the  target  site as  calibration
standards.   They can  be prepared and analyzed by  several
techniques  until  the results are self-consistent.   Another
possibility  is  to  add known  amounts of  an  element,  as a
solution, to  a base material.    This  is  followed  by drying
and mixing.  A third approach is to  use certified  reference
materials such as those  from NIST or the Canadian  Certified
Reference Material Project  (Canadian Centre for Mineral  and
Energy Technology, Ottawa, Ontario).

The need  for  matrix matching is  reduced  by using  scattered
radiation as an internal  standard.  As discussed previously,
the composition  of  the matrix determines the scattering of
X-rays.   By measuring the scattering for all standards  and
samples, and by correcting for differences  in  scattering, it
is  possible  to  significantly   improve  accuracy.     This
technique can compensate  for differences  between samples and
standards.   Andermann and  Kemp  (1958)  were  some  of   the
earliest  investigators to use scattered  X-rays as internal
standards.   Reynolds  (1963)  was  an  early  user of  Compton
scattering  as  an  internal  standard.   Since  absorption is
roughly   inversely   proportional   to   scattering,   these
investigators  found  that  by  ration  the  intensities of
characteristic  lines  to  scattered   intensities  they could
compensate  for varying absorption by different matrices.  A
common   choice   is  the   Compton  peak   of   one  of   the
characteristic lines of the  X-ray tube.   Scattering  is  used
as  an   equivalent  to  an  internal  standard,  attempting to
compensate for inadequacies in empirical  calibrations.   More
recent  applications  of this  technique have included trace
elements  in geological materials (Feather  and Willis,  1976
and Giauque et.al, 1977).

The needs of  the study,   the  availability of standards,  and
the time available  all   affect  the  calibration.    Usually,
environmental  studies  can  be   done  with   a  basic  set of
standards with scattering corrections used  to  compensate for
                            n-256
-------
matrix matching problems.   Care in XRF calibrations is well
worth the time and  effort.   XRF instruments are very stable
and  hold  calibration  well.    Once  calibrated,  analyses
proceed quickly.

One of the  few interference problems  of  any consequence in
XRF spectrometry is the almost exact coincidence of the lead
L  alpha  line and  the arsenic  K alpha  line.   Lead  can be
measured  just as   well  using  the  L  beta  line,  but  the
alternative  isn't  as  good  for  arsenic.    To  avoid  the
interference from lead, arsenic can be measured using the K
beta  line,  which is  not as  intense  as  the K  alpha line.
Using the K  beta line of  arsenic raises the detection limit
by a  factor of four  or  five.   An  alternative  suitable for
some  situations  is  to measure  the combined  line  from lead
and arsenic.   The  presence of either  or  both is a concern.
When assessing a clean-up, both must be removed and this can
be checked with the combined line.  A  potential disadvantage
can be turned to good use.
SEMIOUANTITATIVE ANALYSIS "WITHOUT" STANDARDS

An  energy  dispersive X-ray  fluorescence  spectrometer can
produce spectra covering all elements from sodium to uranium
in about  ten minutes.  This permits  a  rapid analysis of an
unknown with detection limits for most elements of less than
100 ppm.  A rapid,  multielement qualitative analysis can be
of  great  utility.     In  our   laboratory,   qualitative  XRF
analysis  is  used  as a screening technique.   Screening means
initial,  preliminary  analyses   which  serve   to  direct
subsequent  analyses.   If  no   toxic  or  priority pollutant
metals  are  found in a sample, then  it may be  possible to
eliminate  additional  metals  analyses.   Difficult  samples
such as  sludges,  paint wastes  and used  oils can be checked
for  the  presence  of toxic metals.    If  none  are  found,
further work is not  necessary.

The principal  difficulty with  the quantitative  analysis of
hazardous   wastes  by  XRF  spectrometry  is  calibration.
Although  the  best  XRF  work   is  done   with  an  empirical
calibration,  one  which uses  standards which  closely match
the samples, this type of approach is not possible with most
hazardous wastes.   A laboratory at a treatment and disposal
facility, an enforcement  laboratory or  a   laboratory  at a
remedial action site often does not know what will be in the
next  waste   it   encounters.     While   XRF  is  ideal  for
qualitatively  analyzing  such wastes,  as  described earlier,
quantitative analysis is not so straightforward.   However,
it is  now possible  to analyze samples  by  XRF spectrometry
with only minimal standardization.  A few standards are used
                            H-257
-------
to calibrate the  sensitivity of the spectrometer,  but  there
is no need for close matrix matching of  standards to unknown
wastes.

Although empirical  calibrations are still the most accurate
and precise,  great progress  is being  made in  standardless
analyses.     These   methods   offer   the   possibility  of
quantitative  analysis  of  hazardous wastes  with  acceptable
accuracy.   For  most wastes,  sampling is the  source of the
largest error.  This  is  particularly true for solids,  soils
and viscous liquids.   If the analytical error is  kept  under
20%,   then   the  sampling   error  will  be  the   dominant
contribution  to  the overall error.   Accuracies  of 20% or
better   are   possible    with   standardless    calibration
procedures.

Methods which rely  on few standards,  but instead  are  based
on the fundamentals of X-ray physics, are called fundamental
parameters  methods.    These  methods  take  advantage of the
well known  behavior of  X-rays and their interactions with
matter.   Using  a  few  parameters  such  as the X-ray tube
voltage,  current  and  anode material,  it  is  possible  to
predict the output  of  the tube.  For each X-ray photon of a
particular  energy which strikes a  particular  atom in the
sample,  the probabilities of  scattering or absorption are
well known.  The fluorescent yield of those  atoms  excited by
the incident  X-rays can also  be predicted.   Absorption or
scattering  of  the emitted X-rays must  also be  considered.
By  integrating the  above factors for  the whole sample, the
complete emission spectrum of  any material can  be predicted
by calculation.   It is more usual that the  emission spectrum
is  measured and  the  composition of  the sample is  sought.
This too  can be  done.   With  a  given  set  of  experimental
conditions  and an  observed spectrum,  the sample composition
can  be  varied  until  the  predicted  spectrum  matches the
observed spectrum.   Fundamental parameters  methods rely on
accurate  data  for  fluorescent  yields,  absorption cross-
sections, etc.  and on  only  a  few  experimental  parameters.
These  include geometric  factors for instrumental  design and
some measure of detector sensitivity.

The basis for these methods has  been  understood for a long
time.     Development  of practical  fundamental  parameters
procedures  has   required    the  availability  of  extensive
computer power and  the  talent  of  many researchers (Jenkins
et.al,  1981).  One of the more  influential  computer programs
implementing  the  fundamental  parameters   approach  was
developed  at  the  Naval  Research laboratory  (Criss et.al,
1978).    This  program  required a  mainframe  computer, but
similar programs  have  been  written  for personal  computers.
Programs   such   as   these  generally  require   that  one
                            H-258
-------
measurement  for  each  element  of  interest  be made  from
standards, which  may be  pure  elements or  compounds.   They
require that at least one emission line be measured for each
element  in  the  sample  (except  that  one  element can  be
determined  by  difference)   and  allow specification  of  a
particular compound  for a given element (such as an oxide).
Fundamental Parameters  programs have been shown  to work well
on such materials  as metal  alloys and the major elements in
rocks.

There is  one major limitation  preventing the application of
the  fundamental parameters  approach described  above  to the
XRF  analysis  of  hazardous  wastes.    Wastes  often contain
significant  amounts  of organic material,  which, because of
its  light  element  nature,  is  not  directly  observed  by
conventional  XRF  spectrometers.    The  composition of  the
light element part of the sample matrix must be  known if the
fundamental  parameters  approach to  quantification  is  to be
successful.    Fortunately  approaches  have been  developed
which  allow estimation of   the light  element  nature  of  a
sample.

When X-ray photons  impinge  upon  a  sample,  they are either
transmitted, absorbed  or scattered.   There are  two types of
scattered radiation.  Rayleigh  (or coherent) scattering does
not  cause a change  in  wavelength.   Compton (or incoherent)
scattering  produces  scattered  radiation which  is  of longer
wavelength or  less energy than  the incident radiation.  Each
element   is  characterized   by  a  different  ratio  between
scattering  and absorption and  by  a  different ratio between
Compton and Rayleigh scattering.  Light elements scatter the
most,  and heavier  elements absorb  a higher  proportion of
incident   X-rays.     Thus   the  scattered  X-rays  contain
information  about  the  composition  of the whole  matrix,
including the  light  element  content.

A  combination  of  the  above approaches promises to provide
standardless analysis  for any  sample.   Characteristic line
spectra  are  used  to  identify  and  quantify  the  heavier
elements  (sodium  and above), and the scattered  radiation is
used  to  determine the  nature   of  the light  elements.   The
light  elements  are  determined by  comparing  the  measured
scattered  radiation  with the calculated contribution to the
scattering   of  the   heavier   elements.      A  fundamental
parameters approach  is  used  to  adjust  the sample composition
until  the  calculated   spectrum  matches   the   experimental
spectrum,  including  the  scattered radiation.   Since basic
physical  constants  are  used   to  account   for  absorption,
scattering  and fluorescence of each  element,  there  is no
need for  standards  which closely  match the  sample.   A few
standards,  perhaps  pure   elements  or  standard  reference
                            n-259
-------
materials,  are  used  to  calibrate  geometric  factors  and
detector sensitivities.

Fundamental  parameters  programs  which  include  scattering
calculations  to account  for the  light  elements  have  been
implemented  by  several  investigators.   Nielson  (1977  and
1983)  used  the ratio  of  the  incoherent  scatter  to  the
coherent  scatter to characterize  the  light elements of  the
sample.   This is done by using  the scatter ratio to select
two light elements whose concentrations are adjusted to give
the appropriate  light  element atomic number,  absorption  and
scattering.    The  program  iteratively  adjusts  the sample
composition,  both   light and heavy  elements,  until  self-
consistency  is  achieved  and the calculated  spectrum matches
the observed  spectrum.  The spectrometer  was calibrated with
multielement  thin films.   The procedure was used  to analyze
coal,  NBS orchard  leaves,  a soil and  a  ground  rock with
excellent results.   Measured  values were  within  two  standard
deviations of  the reference values for most elements,  which
included  major  and  trace elements.   This type  of procedure
seems ideal  for hazardous waste  analysis  because it  does  not
require similar  standards.

A similar approach  has been  implemented  by Van Grieken  and
colleagues   (Van Grieken  et. al  1979;  Van  Dyck  and  Van
Grieken,  1980).    The  ratio   of coherent  to  incoherent
scattered  radiation  was  used  to  characterize  the  light
elements.  For  a number  of standard reference materials  the
absorption   coefficients  determined  by  this  method  were
within   5%   of  the   values  calculated  from  the   known
compositions.  Good results were obtained in the analysis of
several   reference   materials.      Calibration    of    the
spectrometer was done with commercially available  thin  metal
or  compound  films.    There was  no need  for  soil,  rock or
organic   calibration   standards   to   match  the  reference
materials analyzed.

The  preceding  section   has  outlined   the  potential  of
fundamental   parameters   methods   to   provide  virtually
standardless  analysis  of  a  wide variety  of  materials.  This
potential  has  yet  to  be   realized  in  the  analysis  of
hazardous  waste.    However,  the  basis   exists   for  the
development  of  XRF  methods  for the quantitative analysis of
hazardous  wastes.    In  addition  to  programs  developed by
independent  investigators,  such as those  mentioned above,
some  instrument vendors have  developed  similar programs.
With the  necessary  computer  programs becoming  increasingly
available, the tools for the  development  of standardless  XRF
analysis  of  hazardous  waste  are available.   It should soon
be  possible   to  place a  small  subsample  of  any hazardous
waste  in  an  XRF spectrometer and  within  a  few  minutes know
                            n-260
-------
its composition with an accuracy of 10 to 20% with limits of
detection of about 10 ppm.

USED OIL

A very large amount of used  oil  is produced every year from
vehicles and  industrial  sources.   Much  of it  is  burned as
fuel,  including  a   significant  amount  burned  in  small
commercial  and  residential boilers.   This large  amount of
used  oil in  commerce  and  its  use  as   a  fuel  make  it  a
convenient   receptacle   for  hazardous   waste.      Spent
halogenated solvents, often  from degreasing,  can be present
in used  oil.    Sometimes  this is  inadvertent  and sometimes
intentional.   The EPA  assumes that  if  more than  1000 ppm
chlorine is present  in used  oil,  then halogenated solvents
have  been  added  to  the  used  oil and  the  mixture  is  a
hazardous waste.  The mixture must  be treated as a hazardous
waste  fuel.    This   presumption  can be  rebutted   if  the
generator can show that the chlorine is not from chlorinated
solvents.   Another concern is the  presence of toxic metals
in used  oil.    Lead is  often found in used automotive oil,
although   the  amount has decreased  over  the  last several
years  as  the  amount  of  lead  in  regular  gasoline  has
decreased.  If  used  oil contains more than 100 ppm lead or
4000 ppm chlorine, it is  off specification used oil, and is
subject  to  regulation.    Burning oil  high  in  chlorine will
produce  hydrochloric   acid,   which  can  cause   corrosion  in
boilers.   Lead in  used  oil  which is  burned can  lead  to
hazardous emissions.

Thus there is a need to determine lead and total chlorine in
used  oil.   Measuring total  chlorine  should be much easier
than  measuring  individual   chlorinated  solvents  by  gas
chromatography/mass spectrometry-   The reference method for
determining chlorine  in oil  should be combustion of the oil
in an  oxygen  bomb followed by ion chromatographic analysis
of the chloride in  the combustate.   This technique is slow
and requires a  skilled operator.

XRF analysis of used  oil  for  lead  and chlorine has much to
recommend  it.    Little sample preparation  is  needed;  the
sample need only be  placed  in   a  disposable  plastic cup,
perhaps   preceded  by   dilution.      X-ray   fluorescence
spectrometry is an  ideal way to determine chlorine in new
oil,  and is perhaps  the  method of  choice because  of its
accuracy and precision.   Indeed, ASTM has  a  XRF method for
sulfur  in  oil   (ASTM D-2622),  and  sulfur  is similar  to
chlorine in XRF analyses.  Used  oils  present a much tougher
problem.    Two  principal   reasons  are  sediment and water.
Sediment is often present in  oil, as a product of combustion
or as  particulates from  machining.   Since almost  all XRF
                           n-26i
-------
spectrometers  irradiate  the bottom  of  the sample,  sediment
which settles  to the  bottom of the sample is a problem.   If
it  contains  the  analyte,   it  will  increase  the  reported
concentration.     In  any   event,  the   sediment  is   not
representative of the oil  matrix,  and it will absorb  X-rays
emitted by chlorine in the  oil matrix.   Water can also be a
problem.  Used oil  tanks often have considerable  amounts of
very dirty water on the bottom.  Also, some cutting  oils  are
water  based  with  water  miscible  oils.    Water  can  be a
problem because it presents  a substantially different  matrix
from  the oil matrix  used  in calibration  standards.   An
analyst  may  unknowingly attempt  to analyze  a  very dirty
water sample  or a two  phase  sample with water on the bottom.
Another problem is the often messy nature of used  oils.   The
viscosity may reach that of a  sludge.   Three phases  may be
present; oil, water and  sediment.  Large amounts of  sediment
may  be present,  including  some  that settles  very slowly.
All  these  factors  make  oil analysis  a problem.    However,
these  factors complicate  all  analytical  methods  for used
oil.  XRF, if used properly, still has much to recommend  it.

The  most  suitable  methodology  for the  determination  of
chlorine in  used  oil  by XRF is not  yet completely  defined,
but    the    following   describes   possible   techniques.
Calibration  should  be  done using  a  suitably  spiked   oil
matrix.  Mineral spirits, mineral  oil, or a similar  chlorine
free  material  should  be  used.     A  readily  available
chlorinated  hydrocarbon,  such  as  1-chlorodecane  (available
from  Aldrich)  is added.    In  the preparation  of standards
volatile  compounds  should  be   avoided  in order  to produce
stable  standards.   Standards  which include  the range from
1000  to  4000 mg/kg   should produce  a  linear  calibration
curve.   Samples  high  in chlorine should be  diluted  to  the
linear portion of the  curve.  Of course, the K alpha peak of
chlorine   would  be   measured  and   suitably   background
corrected.

The  main  area of  concern  is  whether  the  sample  matrix
matches the  matrix  of the  standards.  When it does  XRF will
give  excellent results.    The  analyst should  check this by
preparing  spiked  samples.     A   spiking  solution  can  be
prepared in  the  same  way as the  calibration standards.   If
spike recoveries are  not at least 85%,  then the results  for
that  sample  should  not  be  accepted.     It   is  sometimes
possible to  improve results  by  diluting samples with mineral
spirits.  Dilution  reduces  the influence of X-ray absorbing
sample constituents.   In our laboratory used oil  samples  are
often  diluted by  a   factor of five.  Until  experience  is
acquired, it  is best  to  treat  each sample or type of  sample
separately,  and spike  them all.
                            n-262
-------
XRF analysis is  even  more suitable for the determination of
lead.  This  is because the X-rays emitted  by lead are more
energetic and  penetrating than chlorine  K  alpha emissions,
and thus less  affected  by the sample matrix.   XRF is widely
used  to  determine  a  wide variety  of  metals  in petroleum
products.     XRF  spectroscopy   has  been   used  for  the
determination  of wear  metals  in used  lubricating  oil  (Liu
et. at 1986).  Determining lead  in all types of used oil is
a  similar  measurement.   The  determination  of  lead  can be
done  similarly  to  the XRF  analysis  of  oil  for chlorine.
Standards containing  lead  in oil are available from Conostan
(Conoco,  Ponca City,  Oklahoma).    Detection  limits  for lead
in oil are on  the order of lOppm, so at the regulatory limit
of  100  ppm  lead  can  be  accurately  determined.    Our
laboratory has had success using  XRF to determine  lead in
used oil.  Since preparation of oil samples for AA or ICP is
a  difficult  task,  additional  development of  XRF methods is
well warranted.

SUMMARY

The analysis of  hazardous wastes is a complex and difficult
task.    Hazardous wastes are of  widely varying  chemical
composition and  physical  characteristics.  Improved accuracy
and  increased  efficiency  in their  analyses   are  needed.
Elemental analyses  by  X-ray  fluorescence spectrometry have
much  to  offer and  should be  considered  as  an alternative.
XRF  spectroscopy is  an ideal  way  to determine toxic metals
in contaminated  soils,  particularly when there  are  a large
number   of    samples.       It    is    now    practical   to
semiquantitatively  (within 25%)   determine  metals  in waste
samples  with unknown or  unique  matrices.   The  theoretical
calibration  approach,  fundamental  parameters,  can  produce
such results with minimal use of standards.   Waste oils are
a  large  volume  waste  stream  particularly  amenable  to  XRF
analyses.
                         REFERENCES

Andermann,  G.  and Kemp, J.W.  (1958).    Scattered X-rays as
      Internal  Standards  in  X-ray Emission  Spectroscopy.
      Anal. Chem. 30: 1306-1309.

Bain,  D.C.,  Berrow,  M.L.,   McHardy,   W.J.,   Paterson,  E.,
      Russell, J.D.,  Sharp,  B.L., Ore,  A.M.  and West, T.S.
      (1986).   Optical, Electron and  X-ray Spectrometry in
      Soil Analysis.  Anal. Chim. Acta,  180: 163-185.
                            H-263
-------
Bertin,  E.P-   (1975) .    Principles  and  Practice  of  X-Ray
      Spectrometric Analysis.  Plenum  Press,  New York.

Clayton,  E.  and Wooller,  K.K.   (1985).   Sample  Preparation
      and   System   Calibration   for  Proton-Induced   X-ray
      Emission Analysis  of Hair from  Occupationally  Exposed
      Workers.  Anal. Chem. 57:  1075-1079.

Criss,  J.W.,   Birks,   L.S.  and  Gilfrich,   J.V.   (1978).
      Versatile X-ray Analysis Program Combining Fundamental
      Parameters and Empirical  Coefficients.   Anal.  Chem.
      50: 33-37.

Feather,  C.E.  and Willis,  J.P.  (1976).  A Simple Method  for
      Background and Matrix Correction of Spectral Peaks in
      Trace  Element  Determination  by  X-ray   Fluorescence
      Spectrometry.  X-ray  Spec.  5: 41-48.

Giauque,  R.D., Garrett,  R.B.  and Goda, L.Y.  (1979).  Energy
      Dispersive   X-ray    Fluorescence   Spectroscopy    for
      Determination  of  Twenty-Six  Trace  and  Two   Major
      Elements  in  Geochemical  Specimens.   Anal.  Chem.  49:
      62-67.

lida, A., Yoshinaga, A., Sakurai, K. and Gohshi, Y.  (1986).
      Synchrotron   Radiation  Excited  X-ray   Fluorescence
      Analysis  Using  Total Reflection  of  X-rays.    Anal.
      Chem. 58: 394-397.

Jenkins, R. (1988).  X-ray  Fluorescence Spectrometry.
      Wiley, New York.

Jenkins,   R.,   Gould,  R.W.    and   Gedcke,    D.    (1981).
      Quantitative  X-Ray Spectrometry.   Marcel Dekker,  New
      York.

Liu,   Y.,  Harding,   A.R.   and   Leyden,   D.E.    (1986).
      Determination  of  Wear  Metals  in  Oil  Using  Energy
      Dispersive X-ray Spectrometry.   Anal. Chim. Acta, 180:
      349-355.

Kendall,  D.S.,  Lowry,  J.H.,  Bour,   E.L.  and  Meszaros,  T.M.
      (1984).    Comparison   of Trace  Metal Determinations in
      Contaminated Soils by XRF  and  ICAP Spectroscopies.  In
      J.B.  Cohen,  J.C.  Russ, D.E.  Leyden,  C.S. Barrett  and
      P.K.  Predecki,  Eds., Advances  in  X-ray  Analysis,  V.
      27. Plenum Press, New York, pp.  467-473.

Matsumoto, K.  and Fuwa, K.   (1979).   Major and Trace Elements
      Determination in Geological and  Biological Samples  by
                            n-264
-------
      Energy-Dispersive X-ray Fluorescence Spectrometry-
      Anal. Chem. 51: 2355-58.

Nielson,  K.K.   (1977).     Matrix  Corrections  for  Energy
      Dispersive     X-ray     Fluorescence    Analysis    of
      Environmental    Samples    with    Coherent/Incoherent
      Scattered X-rays.  Anal. Chem. 49: 641-648.

Nielson,  K.K.  and  Sanders,  R.W.  (1983).    Multielement
      Analysis   of   Unweighed  Biological   and  Geological
      Samples Using  Backscatter  and Fundamental Parameters.
      In C.R. Hubbard,  C.S.  Barrett,  P.K.  Predecki and D.E.
      Leyden,  Eds.,  Advances   in  X-ray  Analysis,  V.26.
      Plenum Press, New York, pp. 385-390.

Paveley,   C.F.,   Davies,   B.E.    and   Jones,   K.    (1988).
      Comparison  of  Results Obtained  by  X-ray Fluorescence
      of   the  Total   Soil  and  the   Atomic  Absorption
      Spectrometry Assay  of an  Acid Digest  in the Routine
      Determination of Lead and Zinc in Soils.  Commun. Soil
      Sci.  Plant Anal.  19:107-116.Pella,  P.A.  and Dobbyn,
      R.C.  (1988).  Total Reflection Energy-Dispersive X-ray
      Fluorescence    Spectrometry    Using    Monochromatic
      Synchrotron  Radiation:   Application  to  Selenium  in
      Blood Serum.  Anal. Chem. 60: 684-687.

Reynolds, R.C.  (1963).  Matrix  Corrections in Trace Element
      Analysis  by X-ray  Fluorescence:   Estimation  of  the
      Mass  Absorption  Coefficient  by Compton Scattering.
      American Mineralogist  48:  1133-1143.

Van  Dyck,  P.M.  and  Van Grieken,  R.E. (1980).   Absorption
      Correction   via   Scattered   Radiation   in   Energy-
      Dispersive  X-ray  Fluorescence Analysis for Samples of
      Variable Composition  and  Thickness.   Anal.  Chem.  52:
      1859-1864.

Van Grieken, R., Van't Dack, L.,  Dantas, C.C. and Dantas, H.
      D.S.  (1979).     Soil   Analysis   by  Thin-film Energy-
      dispersive  X-ray  Fluorescence.   Anal.  Chim.  Acta 180:
      93-101.

Zsolnay, I. M. ,  Brauer, J.M.  and Sojka,  S.A. (1984).  X-ray
      Fluorescence Determination  of Trace Elements in Soil.
      Anal. Chim. Acta, 162: 423-426.
                            n-265
-------
92                  RECENT ADVANCES  IN MEASURING MERCURY
                    AT  TRACE  LEVELS IN THE ENVIRONMENT

       Peter  Stockwell,  P.S. Analytical, Arthur House,  B4,  Chaucer
       Business  Park,  Watery  Lane,  Kemsing,  Kent   England: Rich
       Comeau,  Questron  Corporation,  PO  Box  2387,  Princeton, NJ
       08543

       ABSTRACT

       The   environmental  significance  of  Mercury   and  numerous
       methods  for  measuring  mercury  have long been known.  This
       paper  will  review  and    compare  existing  methods,  then
       summarize  recent  advances  in  inehtodology  including  Cold
       Vapor  Atomic  Fluorescence.    Emphasis  will   be placed on
       accuracy,  sensitivity,  practicality  for  routine operation
       and level automated through-put.

       Methodology  for a complete system will be described based on
       selected   sample  preparation,  introduction  and detection
       techniques.    Sample  preparation  techniques include manual
       water  bath  digestions,  automated flow-through heating bath
       digestions   with   heating   coils   and   autoclaved batch
       digestions.      Each  is  considered  for  accuracy,  sample
       through-put,  sample  size,  reagent  conservation and labor
       requirements.     Cold  Vapor  Sample  Introduction techniques
       include   manual   batch   generation,  automated  continuous
       generation  and  continuous  generation  with  gold or silver
       amalgamation.     Mercury  detection  methods  include  various
       configurations of Atomic Absorption and Atomic Fluorescence.

       INTRODUCTION

       The  determination of ultratrace concentrations  of mercury by
       cold  vapor  atomic  absorption  and  most  recently,  atomic
       fluorescence,    has   become   widely   accepted.      Various
       techniques  for  both  the  continuous  generation of  mercury
       vapor  and  its  detection have been reported.   Nearly all of
       the  methods  are  based on the reduction of mercuric  ions to
       elemental   mercury   with  SnCl2   separation   of elemental
       mercury  from  solution  as  a "cold vapor" and  determination
       with  a mercury specific spectrometer.  There is still a need
       for  a method that readily lends itself to routine operation,
       (excessive  performance  for practical usage).   It would be a
       combination  of  techniques  that  are  widely   accepted  yet
       flexible  enough  to  incorporate  the  most recent advances.
       Routine   methods   should   be   EPA  approved,   simple  and
       automated.      These  traits  must  be  implemented   without
       affecting  sensitivity and accuracy.  A method that addresses
       itself to these questions is presented in this paper.
                                    H-266
-------
  APPARATUS

  A  system   was  assembled  with   four  common  components;  an
  autosampler,   vapor generator, dual  detector, and an industry
  standard   computer (Figure 1).  Automation is accomplished  by
  continuous   pumping  of  reagents   and  computer  instrument
  control.     A  flow  control  valve   alternates  the reaction
  mixture  between  SnCl2  plus  Blank  and  SnCl2 plus Sample.
  Mercury  is then removed from the reaction mixture as a  "cold
  vapor"  by  continuous  aeration  with  a  carrier  gas.  The
  detector   has  been  specifically designed  .for conveniently
  interchanging   atomic  absorption   and  atomic  fluorescence
  optics  modules,  both  techniques   are  readily available  to
  operate.     Data  acquisition  and   storage  are  handled   by
  "Touchstone"    (a   unique   software   package),   and   a   PC
  compatible computer.
                 Vapor Generator
                                           Dual Detector
Autosampler


Random Access
8 ml/sample
 40/hour
           Reagent   Peristaltic  Switching   Vapor
           Reservoirs   Pump    Valve   Separator
                                           Fluorescence
                                               Detector
                           Hg Lamp
                          Hg Lamp
                                       Detector
                                           Absorption
                                                PC Computer
 Data Acquisition


Automation Control
  DISCUSSION
  The   reaction
  detector
      mixture,   vapor-liquid  separator   and  dual
were   designed  for  optimal  performance.     The
  reaction  mixture  (SnCl2  plus  Blank   to SnCl2 plus Sample)
  forms  from   a   fixed  ratio  rates of  3 ml/min of SnCl2 to  8
  ml/min  of blank or sample.  The reaction mixing and aeration
  times  are   operator  selected  and valve controlled to allow
  for flexibility  of sample types and concentration ranges.
      ^  atomic    absorption   optics  module  was  specifically
  designed  and   optimized for cold vapor mercury determination
  as  described   by  EPA method 245.2.  The  atomic fluorescence
  module  reflects  the latest technique of mercury detection by
  EPA  method  245.2.    The atomic fluorescence module reflects
  the  latest technique of mercury detection by emission rather
  than   absorption.       Since  emission  is  inherently  more
  sensitive  than  absorption,  mercury detection at ppt levels
  becomes possible.
                                H-267
-------
Conclusion

For  those  operations  which  require  measurement under EPA
Method   245.2,  the  atomic  absorption  optics  provides  a
detection  limit  well  within the required minimum detection
level  of  0.2ppb.  The fluorescence detection with detection
limits  well below O.Olppb is then used to corroborate the AA
results.    Together  they provide verified results which can
withstand legal scrutiny.
References:

1)Hatch, W.R. and Ott, W.L., Analyt, Chem., 1968
2)Thompson K.C. and Reynolds G.D. Analyst, 1971
3)Winefordner    J.D.    and   Elser   R.C.   Analyst,   1971
                             H-268
-------
93          The USE of ION CHROMATOGRAPHY  in  SOLID  WASTE MATRICES;METHOD 3OO


                     John  D.  Pfaff and Carol A. Brockhoff





         Introduction:


              The Environmental Protection Agency  approved the use of method

            V
         3OO for  the analysis of  nitrate in drinking waters in  1964.  This


         constituted  the   first  approved  method   for  the   use  of   ion


         chrornatography  to analyze drinking waters and/or wastewaters. The


         Environmental Monitoring Systems Laboratory  (EMSL) in  Cincinnati,


         Ohio  has revised  method  300 to bring  current equipment  into the


         method  . Additionally  matrices  were investigated  to extend the


         methods  usefulness.  This  paper will  cover  the  findings  of the


         effort  to  use  this  method  in  matrices  of  interest  in  the solid


         waste  field. The analysis  of  solid  materials,  groundwaters and


         leachates  were investigated and single  laboratory precision and


         bias data produced.  The method has  been sent  to the Office of Solid


         Waste  with EMSL's  recommendations for  acceptance as  an approved


         method  to  be published  in  the  Test  Methods  for Evaluating Solid


         Waste, SW846.


         Experimental:


                In  general,  Method  3OO  is  used  to  analyze the  following


         common anions; fluoride,  chloride,  bromide, nitrate, nitrite,  ortho


         phosphate  and sulfate.  Since these anions are extremely  soluble  in


         water  it was felt that  they could easily be extracted front  solid


         materials  and then analyzed through the use of ion chromatography .


         If this  were shown  to  be the case the use  of Method  3OO  could  be


         extended into solid matrices.
                                     H-269
-------
       Groundwater  is  also of great interest in the  protection of




the  environment  from  solid  wastes.  Material  disposed  of on the




surface  or buried  in  the  earth  can  migrate by  percolation  into




aquifers  or through a leaching  technique by  surface waters  which




then,  in  turn, contaminate groundwaters.




       The method described  here  uses a  chemically suppressed ion




chromatograph with conductometric detection. The AG4A guard and the




AS4A   separator   columns   were   used   with   the  micro   membrane




suppressor. These items are produced by the Dionex  Corp.  The eluant




used  was  1.7  mM sodium   bicarbonate  (NaHC03)  and 1.8  mM sodium




carbonate (Na2C03).




       The author  contacted  the  producers of  standard  reference




materials but  found no solid materials  which had a  known anionic




content.  Consequently an  EPA quality  control  sample of  shale (WP




386) was  used. The material was  first extracted  with each technique




under  investigation to  produce  a background value.  To  produce a




solid  with all the  anions of  interest  a volume of  reagent  water




which  had been fortified  with  the anions of interest was added to




the  solid and thoroughly  mixed.  The slurry  was placed on a magnetic




stirrer/hot plate and  heated  at  a medium  heat  while stirring taking




care not to  let  the mixture  boil.




       This  was  continued until  little  water  was left.  Then the




slurry was  transferred  to an  oven,  heated  at  95°C  overnight and




stored in  a   desiccator.   Enough  of the  material  was  produced to




allow  all  analyses  needed in the study.  Separate  portions of the




well   mixed   material  was  used  to  investigate  recovery of  the




an ions.
                              H-270
-------
Solid Materials Extraction Technique:




       The  unfortified  shale  solid  was  used  to  evaluate  the




extraction  techniques.  Three techniques  were evaluated,  first;  a




volume of reagent water ten times the weight of solid material used




was added to  the  solid  .  This slurry was then mixed  on  a  magnetic




stirrer for ten minutes  then filtered and analyzed.




       Secondly;  again the  ten  times  ratio  of  reagent  water  to




solid was used. This time  the slurry  was  sonicated for ten minutes




then filtered  and analyzed.




       Thirdly; the same water ratio was used and heated to 6O  C




then placed in a  sonicator  while still  hot and sonicated for ten




minutes then  filtered  and analyzed.




       The  results of  these extractions  and  analyses are  shown  in




Table    1.    Three   anions    were   found   in    the    unspiked




shale;f1uoride,chloride  and  sulfate. Each technique was carried out




with four repetitions  and the results compared. One  gram  of solid




was  used  with   ten  milliliters  of   reagent  water.  The  large




concentration  of  sulfate  present  made comparison difficult and the




values are suspect. However the fluoride and chloride values showed




that  the first  extraction  technique  gave  not  only the  higher




recovery weights  but also the most consistent and lowest  standard




deviation  values.  Consequently,   this  extraction   technique  was




chosen and  used to produce the remaining values.




       The  amount of solid used  was investigated  using weights of




one, two and  five grams  of the spiked shale  solid.  The spike used




was  the  same  in  all   cases  and consisted  of  the anions  and
                            H-271
-------
concentrations shown in Table 2. Each weight used was run twice and




the average values compared. These values are  shown in Table 3. jt




can  be  seen  that for  weights of  1 or  2 grams the  value  remains




constant showing good extraction. Five grams of solid extracted may




have  overtaxed  the  volume  of water used  and gave  poor comparison




va 1 ues .




        The spike  recovery values were obtained  using weights of 2




and  5 grams of solid  material .  These values are shown  in Table 4.




It can  be  seen that values  for fluoride are  not usable. It is felt




that  two  conditions  contribute  to  these values.  First,  fluoride




elutes  in   an  area affected  by  the" water dip". This  area  is  the




result  of  the negative response to the reagent water which has less




conductance than  the  eluant.  Thus the positive  fluoride peak fall




in an  area of  negative  background.  Secondly,  all  retained  anions




elute  in  this  area and  add  to the  response of the  fluoride peak




making  quantitation extremely difficult. The high concentration of




fluoride in  the  blank makes the  differentiation of  1  mg/liter of




fluoride difficult  to detect.  Because of these  effects it  is felt




that  this method  cannot  at  this  time  be recommended for  the




analysis of fluoride.




        Although  the solid shale was fortified  with orto phosphate




and  sulfate   the  very  high  concentration   of   sulfate  caused  an




overlape  of   the  sulfate   and  ortho   phosphate   peaks  and  made




quantitation difficult.  Inorder to  quantitate  ortho phosphate and




sulfate the  concentration  of  sulfate  that  interfers  with ortho




phosphate  will have to  be determined.  EMSL  intends at a  later date




to  investigate other solid  materials which, hopefully, will have
                             H-272
-------
lower background  concentrations of sulfate and  will  allow further




investigation of  ortho  phosphate and sulfate.




       A chromatogram of the spiked  shale solid extract  is shown




in Figure 1.









Leachates:




       Inorder  to mimic the effects  of water percolating through




a solid waste  and then further  into  the ground EPA  developed the




Leachate technique.  This technique utilizes a  24  hour  extraction




with water and the possible addition of 0.5 N acetic acid to adjust




the ph. The  technique stipulates that no more than 4 ml_ of acetic




acid  be  used  for each gram  of solid  material  used.  After  the




extraction  is  carried out the  total  volume of  the liquid extract




should be adjusted to 2OOO mL.  If  1OO grams of solid  material is




used  ttiat would mean that a maximum  acetic acid used  would be 20




mL of 0.5 N  acetic acid per 1OO ml of water.




       This  same  volume of  acetic acid  was added  to reagent water




and  injected into the  ion  chromatograph.  The resulting off scale




peak  has been superimposed  onto a typical  chromatogram of a spiked




solid shale  extract  inorder to  show  which peaks would  be lost if




the maximum  acetic acid  is used  in  the leachate  procedure. This is




shown in figure 2. It can be seen that all  anions  from fluoride to




bromide could  not be detected.









Groundwaters:




       The  only  precaution  that  should   be  taken  prior  to the




analysis  of  a   groundwater   is  to   filter  the  sample   if  any
                            H-273
-------
particulate  material is  present. This  is  to protect  the  columns of




the chromatograph which cannot tolerate solids. With  this exception




a groumdwater would  be  the same  as any other  aqueous sample.










Summary:




        Ion  chromatography can be used to analyze anions in solids




if preceded  by  a simple extraction technique.  A solid sample, that




is  of  a  fine  consistency,  and  weighing  about  1  or 2  grams  is




extracted  with  a  volume of  reagent water ten  times  the weight  of




the solid  material used  and stirred  on a magnetic stirrer for ten




minutes. The resulting slurry is  filtered and injected into the ion




chromatograph.   The  resulting  analysis can  be used  for chloride,




nitrite,  bromide, nitrate, ortho phosphate  and sulfate.




        Leachates  that  have  not  used acetic acid  to  adjust the  pH




can be analyzes after filtration.  If acetic acid has  been used the




analyst  can  determine  if any anions  can  be  quantitated by putting




the same concentration in  reagent  water  and running  it to see what




area is  covered over  by  the acetic acid  peak.




        Groundwater should  not pose any problems for analysis by ion




chromatography  after filtration. This  method has been published  by




the Environmental  Monitoring  Systems Laboratory, 26 W. M.L.King




Drive,  Cincinnati, Ohio  45268 as method  300.O, Revised 12/89.
                             E-274
-------
       Table 1: COMPARISON OF THE EXTRACTION TECHNIQUES
Anion   Technique 1
       10 min stir
          mg/L
   Technique 2
10 min sonicate
     mg/L
 Technique 3
Heat-sonicate
   mg/L

F
Cl
SO

X
Sd
1
6.
1O.
977
2
3 6.4
6 9.6
956
3
6.5
9.8
1O14
F Cl
6
0
.4 10
.10 O.
.2
51
4
6.5
10.5
1012
SO
990
28.1
1
4.7
10.0
814
F
4.4
1 .54
2
4.8
1O.1
831
Cl
9.4
0.87
3 4
4.2 4.0
9-1 8.4
716 660
SO
756
161
1
1.9
7.4
440
F
4.1
1.54
2
4.3
9.4
674
Cl
8.
0.
3
5.1
8.9
750

6
87
4
5.2
8.7
806
SO
667
161
                  TABLE 2  SPIKE  USED
                  Anion
  Cone.(mg/L)
F
Cl
N02-N
Br
NO,-N
HP04-P
SO.
1
10
5
5
10
10
20
        TABLE 3  COMPARISON of CHLORIDE for VARIOUS
                       AMOUNTS of  SOLID
Anion

Cl


Wt. of Solid
grams
1
2
5
Recovery
mg ' s
10.1
10.2
8.5
Water Used
mL ' s
10
20
50
                                H-275
-------
        TABLE  4   RECOVERIES  of SPIKES

Anions    Solid Used   Spike   Am' t Found    Blank
            grams       mg/L      mg/L         mg/L
  F                      1          4.5         6.5      0
  Cl                     10        15.4        10.2     52
  N02                    5           3.7         —       74
  Br                     5          5.3         —      106
  N03                    10          7.5         —       75
               5
  F                      1          6.3        6.50
  Cl                     10        15.6        8.5      71
  N02                    5           2.8        —       56
  Br                     5          4.2        —      84
  NO,                    10          6.4        —       64
                    H-276
-------
                         Figure 1
            METHOD 300 EXTRACTION of  SPIKED  SHALE
157300
nS
-14300
      0.0       1.0       2.0        3.0       4.0        S.O
                                        Minutes

             Anion  Peak   Retention   7.  Rec.
                     Num.   Time(min.)

               F       1       1.1            0
               Cl      2       1.6            72
               N02     3        1.9           56
               Br      4       2.9            84
               NO,     5        3.2           64
               S04     6        5.4
6.0
7.0
                               tt-277
-------
                                       Figure  2
                              EFFECT of  ACETIC  ACID
157300
nS
 Area  of  20/100 ml_
   0.5 N  AcOH
-14300
        0.0
1.0
2.0
3.0
4.0
Minutes
5.0
6.0
7.0
                                               E-278
-------
94       THE DETERMINATION OF  THE  EFFECTS  OF  PRESERVATION ON  NITRITE  AND
          NITRATE IN THREE TYPES  OF WATER SAMPLES  UTILIZING  TRAACS 800
               AUTOANALYZER AND SINGLE  COLUMN  ION  CHROMATOGRAPHY

       Miriam  Roman,  Robert  Dovi,  Rhonda  Yoder,   Frank   Dias, Bruce
       Warden,  Waste  Management  Enviromental  Monitoring   Laboratory,
       Geneva, Illinois  60134

       ABSTRACT   Nitrite and nitrate are important  parameters  in ground
       water  analysis.  Excessive  amounts of  nitrates  has been shown to
       increase  the methemoglobin  in  the  blood  of   infants.  A 35-50%
       increase  causes  headaches,  a 70%  increase  is  lethal.  Nitrite in
       acidic   solutions  forms  nitrous acid  which   can   react   with
       secondary  amines to  form  nitrosoamines,   many of  which  are
       carcinogenic.

       The  1989 EPA Federal  Register states that  unpreserved samples to
       be  analyzed for  nitrite   and  nitrate  are  to  be   kept at 4°C and
       analyzed  within  48 hours.   However,  if samples   are  to  be tested
       after  48 hours,  the procedure states that  the  samples  should be
       preserved  with H-SO,  to  pH<2  and stored  at 4°C.   Under  these
       conditions,  the  samples have been allowed  a  28  day holding time.
       Some  states recommend acidification  while  others do  not.  Early
       indications  in our lab suggested  that  nitrite  was not  stable in
       acidic  solutions and  was,  in   fact, converted   to nitrate   over
       time .

       This  paper describes  the  effects  of  preservation on  nitrite and
       nitrate in ground, leachate  and  surface waters.

       INTRODUCTION  The stability  of   nitrite and nitrate   in  acidified
       and unacidified water  were determined using a  colorimetric method
       and   ion  exchange  chromatography.    The    colorimetric  method
       measures  nitrite, after reaction  to  form a color complex, at 520
       nm.   Nitrate is  reduced to  nitrite and the combined  nitrate  plus
       nitrite  is measured.  The concentration of nitrate is determined
       by  difference.   Nitrite and nitrate  separates on an  ion exchange
       column using 2.5mM lithium hydroxide  as the eluent. The  separated
       ions are measured by UV detection  at  214nm.
                 Tests   were  run  on unpreserved reagent  grade  water and
       reagent  grade  water  acidified  to  pH=2  and  pH<2  with  H^SO^.
       Acidified  and unacidified water were spiked   with nitrite at the
       following  levels: 0.5, 1.0, 2.0  and 5.0 mg/L.  Ground,  leachate
       and  surface water samples were  run unpreserved  and preserved, at
       pH=12,  with NaOH. The three types of  water  samples,  unpreserved
       and  preserved, were spiked  at the following   levels:  0.5 and 1.0
       mg/L  nitrite for analysis  on the TRAACS   800 and 10.0  and  50.0
       mg/L  nitrite for analysis using Waters Ion  Chromatograph. Prior
       to  analysis on the 1C, samples  were  diluted  10-fold  and filtered
       through  a  0.45  micron   filter.   Samples  for  the colorimeteric
       method  were  filtered  through  a 0.45 micron   filter  prior  to
       analysis.
                                   H-279
-------
Experimental:

Apparatus:   The colorimetric  instrumentation  used was a   Bran &
Luebbe  TRAACS 800  AutoAnalyzer   segmented  flow system with the
following components:

          .Random    Access    Autosampler     sampling   at    120
           samples/hour.
          .Multi-test cartridge for  nitrite.
          .Multi-test cartridge for  nitrate  + nitrite.
          .Reagent Sequencer.
          .IBM PC PS/2  30 data system.
          .UV/VIS Detector @ 520nm.

The  chromatographic  instrumentation   used  was  a  Waters   Single
Column Ion Chromatography system with  the  following components:

          .510 Pump, flow rate @ 1.2ml/min.
          .712 WISP Autosampler, lOOul  loop.
          .IC-Pak Anion Column
          .In-line Precolumn Filter
          .441 UV Detector @ 214nm
          .840 Data Handling System

In   the  colorimetric  reaction   nitrate,   which  has   undergone
reduction  to nitrite,  and nitrite originally present reacts with
sulfanilamide  to  form  a  diazonium   salt.   This   couples  with
N-(1-naphtha)-ethylenediamine  dihydrochloride to  form a   highly
colored azo dye which is measured  at 520nm.

In  the chromatographic system, 2.5mM  lithium hydroxideis used as
the eluent and the species monitored using UV detection at  214nm.


Reagents and Chemicals:

Colorimetric Method:
          .Cupric Sulfate  ( CuSO ,, . 5H-0 )
                                T    XLi

          .Brij-35, surfactant.

          .Hydrazine Sulfate  (N2H4.H2SC>4)

          .Hydrochloric Acid  (HCl)

          .N-1-Naphthy1ethylenediamine  dihydrochloride.
                            U-280
-------
         .Phosphoric Acid, cone. (H.,P04)

         .Potassium Nitrate (KNO-,), standard

         .Sodium Hydroxide (NaOH)

         .Sodium Nitrite (NaNO~), standard

         .Sulfanilamide (CgHgN-O-S)


Ion Chromatography Method:
         .Eluent
          - LiOH/H20
          - Boric Acid
          - D-Gluconic Acid
          - Glycerin
          - Acetonitrile
         .Wescan 200 ppm Nitrite Standard
         .Wescan 200 ppm Nitrate Standard

RESULTS AND DISCUSSION

         Reagent   grade  water   spikes  stored   at  4°C   and
unpreserved showed no deterioration of nitrite or nitrate during
a  thirty day test period.  Water samples (ground, leachate  and
surface)  stored at  4°C  and unpreserved showed  stable nitrite
concentrations  from 14-37 days.  The surface water  showed some
deterioration  of the nitrite after 14 days. This is believed to
be  caused  by  conversion  of  nitrite to  nitrate by  bacteria
present in the surface water.
(Fig.1-4)
         Reagent  grade water spikes stored at 4 C and acidified
with  H_SO, showed immediate deterioration of nitrite to nitrate
while  nitrate  remained  stable.  This  would explain  the poor
nitrite  spike recoveries experienced in the  lab during routine
analysis.  Total conversion of nitrite to nitrate in 14 days was
seen in water preserved to pH<2. (Fig.5-6)
         The  water samples referred to above  stored at 4°C and
preserved  with NaOH  to  a pH=12 showed  excellent stability of
nitrite  out to 37 days. No conversion of nitrite to nitrate was
seen  during this test period.It  is believed that at  this high
pH,  bacteria that effects  nitrite conversion was  not present.
Nitrate  was  seen  to  be  stable in  both acidic  and alkaline
conditions. (Fig.7-9)
                           H-281
-------
SUMMARY

         Past  practices  of  acid preservation will not allow for
accurate  nitrite determination due to conversion  of nitrite to
nitrate within a 24 hr. period.  Samples kept at 4 C with no acid
preservation  showed better  stability,  while samples that  were
base  preserved  showed   excellent stability.  Base preservation
appears  to  eliminate  nitrite conversion  caused by  bacterial
action.
         Base  preservation  seems to be the  method of choice to
obtain  accurate determination  of both nitrite and nitrate while
still   having   a   holding  time   of  more   than  28   days.
                           H-282
-------
     NITRITE/NITRATE   STABILITY   STUDY
                        Surface  Water Unpreserved
J. . 4U -

1.2U-

1,00-
0.80-


0.60-

0.40-

0.20-

n
u •
n.m-




I 	 ^~ ~ ~ - -*.
«*,



' * 	 4 	
"""



g 	 	 — P 	 g 	 B- 	

5K- -3K 1.0 N02 SPIKE
4 4 n ^ MAO GDTIfP
~ T U.D WU/ orlMli
D 	 D UNSPIKED

1
\
\
\
\

4-.... \
''•-.., \
^ -O \
'-.._\
-0 	 Q 	 ^|

                                             16
                     23
30
37
    FIG. 1
DAYS
     NIRTITE/NITRITE   STABILITY   STUDY
                       Surface Water Unpreserved
18
15-
    *	4 NITRATE
    D	0 NITRITE
-------
NITRITE/NITRATE   STABILITY   STUDY
                   Ground  Water  Unpreserved
1UU -
Q f!
OU -
* 60
\
& 40-
e
20-
0 -
-10-

i sn
J. . o U
1.50-

A
\
& 1.00-
6
CM
0 0.50-
^-i
-0 10 -


)K- -I 50 N02 SPIKE
• -- --4 10 MO9 9P T KF
T 1U N\J £ Orir\.Ij
Dn T1HQP T FFFi
U UWorilxDiJ
«/ 	 	 ^ 	 I 	 $ 	 1 * ft *" ~ "" ~ SK
A i A A 	 1 	 A 	 A 	 A A
T 	 f 	 f 	 T^ T T T T^ 	 f
nnnnnnnnn
LJLJUULJLJLJUU
2 3 4 5 8 15 23 30 36
FIG. 3 DAYS
NITRITE \NITRATE STABILITY STUDY
Leachate Water Unpreserved
)K- -)K 1.0 N02 SPIKE
4 	 4 O R Mf)9 ^PTIfF
T T U.3 fiUii orlMli
DR ITM^PTVFn
U UliLJiifvijlJ
¥- *- a JiL J*
r r 	 * 	 ft----_| 	 ^.----t
^ 	 A A A A A_ ...A
t 	 T 	 t 	 4 	 T- 	 *

1 2 3 4 8 16 23 30 37
FIG- 4 DAYS
D-284
-------
      NITRITE/NITRATE  STABILITY   STUDY
                    NITRITE SPIKED WATER ACIDIFIED  pH=2
      	4 NITRATE
     D	D NITRITE SPIKE
2.00-
1.50-
1.00-
0.50-
      NITRITE/NITRATE   STABILITY  STUDY
                    Nitrite  Spiked Water Acidifed pH<2
              NITRATE
        D	D NITRITE SPIKE
    FIG.  6
DAYS
  H-285
-------
         NITRITE/NITRATE  STABILITY   STUDY
                         Surface Water  Preserved pH=12
   100T
   80-
      I- -*' 50 N02 SPIKE
      +	+ 10 N02 SPIKE
      D	D UNSPIKED
    60
    40
    20-
   -10
                 -B-
                        -9-
                                                 15
                                                       23
-B-
30
36
       FIG.  7
                                   DAYS
1.80

1.50
         NITRITE/NITRATE   STABILITY  STUDY
                         Leachate  Water  Preserved pH=12
        - -*  1.0  N02 mg/L
        	4  0.5  N02 SPIKE
           0  UNSPIKED
 1.00-
  .50-
-0.10J
       FIG. 8
                                    DAYS
                                       n-286
-------
      NITRITE/NITRATE   STABILITY   STUDY
                        Ground Water Preserved pH=12
 80-
 40


 20-


  0

-10
$- -5K 50 N02  SPIKE
*	* 10 N02
D	D UNSPIKED
     FIG. 9
          -3-
-B-
                                   DAYS
-&-
                                                15
-B-
                                                   23
-B-
                                 30
                         36
                                     H-287
-------
95                  A STUDY OF THE EFFECTIVENESS OF SW 846 METHOD 9010
                    FOR THE DETERMINATION OF TOTAL  AND AMENABLE CYANIDE
                                IN HAZARDOUS WASTE  MATRICES

          Ronald  J.  Osborn.  Group  Leader Wet Chemistry,  Roger A. Kell,  Laboratory
          Manager,  Ramona  S.  Zully, Chemist  II  Advanced  Waste Research,  Chemical
          Waste   Management,   Inc.,  Technical  Center,  150  West   137th  Street,
          Riverdale, Illinois  60627

          ABSTRACT

          Recently  imposed regulations  on  the  hazardous  waste  industry include
          specific  levels  for  Total and  Amenable Cyanide.  Precise measurements  of
          these  parameters at  these regulatory levels  are  critical.    The  accepted
          method  for the determination of these analytes  is  EPA  SW 846  Method  9010.
          Cyanide amenable to  chlorination is  the result of the difference between
          two  total  cyanide measurements;  the first on the  sample "as is" and the
          second  on the same sample that has  been treated with hypochlorite.   Since
          chlorination  is  intended  to break  down cyanide, the second total cyanide
          level  should  always  be less than the first result (before  pretreatment).

          The  applicability of this method  for hazardous waste  matrices  is evaluated
          in this  paper  by:   1)   investigating the  factors  that  influence the
          performance of this method;  2)  examining cyanide amenable to  chlorination
          results that  are negative;  and 3)   proposing potential solutions to the
          problems  encountered.

          INTRODUCTION

          The  recently  mandated  Land  Disposal  Restrictions  as found in  40CFR Part
          268, impose regulatory levels  on  total and  amenable cyanide  concentrations
          for  land  disposal.    Some  of  the waste  codes  have  regulatory levels
          associated  with  them  that  are quite  low which  will  make accurate and
          precise measurements of  these components  critical.   The accepted method
          for  the measurement  of total  and amenable cyanide is SW-846 Method 9010
          2nd  Ed.   Although   this  method  may perform well  in  water  and related
          matrices,  its performance in hazardous  waste matrices is in  question. The
          purpose of this  paper  will  be to evaluate  applicability  of this method
          when used  to determine  total  and amenable  cyanide  concentration  in
          hazardous  waste  matrices,  and  to  be able  to  better understand the
          information provided by  the results  of Method 9010.

          CYANIDE AMENABLE TO  CHLORINATION

          A major part of  Method 9010 is devoted  to  determining  cyanide amenable  to
          chlorination.  Cyanide  amenable to  chlorination is  actually  the  difference
          between two total cyanide analyses.  A sample is split into  two equal test
          portions:  the first is used to determine total  cyanide concentration, and
          the  second test portion is subjected to an  alkaline chlorination procedure
          and  then analyzed under the  identical conditions as the first test portion
                                           H-288
-------
to  determine the  total  remaining  cyanide  concentration.    It  is  the
difference between these two "total" cyanide analyses that is defined as
the cyanide amenable to chlorination.  It is the intention  of the alkaline
chlorination to destroy the cyanide present in the waste material.  As a
result of this,  the cyanide remaining  after chlorination should always be
lower  than  before the  chlorination  step.    We have observed  that when
working with hazardous  waste matrices,  this  is frequently not the case.
It is  a very common  occurrence  for  the  cyanide amenable to chlorination
to have a negative result due to the fact there is a higher concentration
of cyanide remaining following chlorination than was present in the test
portion that was not chlorinated.

This situation presents  a  dilemma to  those disposing of cyanide-bearing
hazardous  wastes.   It  is entirely  possible  that the  total  cyanide
concentration  will  be below  the  appropriate regulatory  level   for  a
particular waste (when analyzed by Method 9010), but when  it is subjected
to alkaline chlorination the resultant cyanide  concentration is above the
regulatory  level.    In this case,  the  result  for  cyanide amenable  to
chlorination is negative.  Although in this scenario, the waste would be
acceptable  from  a  legal   standpoint,  the  chemistry  that supports  the
evaluation of this waste does not always work.  In  fact, the guidance from
the  EPA  in  reference  to  this  scenario is that  a  negative  result  for
cyanide amenable to  chlorination is to  be  considered "zero,"  which will
be below the regulatory level,  thus,  making  the material  acceptable for
disposal.  Obviously, there was more cyanide present than was originally
detected by the total cyanide analysis,  but because the higher result of
the cyanide analysis following chlorination is only used for calculating
purposes, the information  that this result provides is ignored.

One might propose many ideas as  to why more cyanide is detected following
chlorination than before.   A discussion of some of  these theories follows.
It must,  however, be understood  that the exact  cause of this phenomena is
not yet  known  or understood and that the  following  discussion explores
only possibilities that may influence this occurrence.

One idea is  that when  hypochlorite  is added  to the  sample to  break down
the  cyanide it  introduces an  interference  into the  distillate.   The
assumption is that upon  completion of the  chlorination  step  some of the
chloride remained  in  solution  and  distilled over with the HCN gas.   In
order  to  investigate this possibility,  several distillate  samples were
analyzed for chloride.   It  was determined that all  of the distillates that
were analyzed contained less than 10 ppm chloride,  and it was decided that
this possibility of interference could be disregarded.

One theory  as  to  the  cause of higher  results for total  cyanide after
chlorination than  before  is  due to  the  complexation  of cyanide with
various metals.    Although in  some  cases these  compounds   are  easily
dissociable and quantitated; some are not--such  as Fe[CN]x complexes.  This
theory assumes that  the cyanide complexes that are  dissociable  will  be
broken down  in  the  acid  reflux distillation  while the  nondissociable
                                  H-289
-------
cyanide complexes will  not.   Upon the addition of hypochlorite into these
materials which contain nondissociable cyanide  complexes  to oxidize the
cyanide, some of the  metals  are also oxidized thus weakening the complex.
When  these   treated  materials  are now  subjected  to  the acid  reflux
distillation the  cyanide  associated with  these complexes  will  also be
released, producing HCN gas.

To investigate  the relationship between metal  content  and cyanide amenable
to chlorination  results,  the metal content  for samples which  had both
higher  and   lower  cyanide  concentrations  following chlorination  were
reviewed.  Review of  the data did not seem to  provide  any definite answers
to the problem of higher cyanide concentrations following treatment with
hypochlorite.   However,  this information does  propose  some interesting
suggestions.  In looking at the matrices  presented in both cases, it was
noted that there are similar matrices in  both situations.   This suggests
that this problem is not matrix specific  unless all  hazardous wastes are
to be considered one  matrix.  Also,  there  does not  seem to be any definite
trend corresponding to one metal  affecting  a  recovery as there seem to be
high levels  of the same metals in both cases.

In considering the vast array of matrices  that qualify as hazardous waste,
Figures  1     6  show  the  concentration  of  cyanide before  and  after
chlorination in several different matrices.

It  can  be concluded  from the these graphs that in each  matrix studied,
with  the exception of  aqueous  and  cyanide  by-product matrices,  that the
problem  with  recoveries  before  and  after  chlorination  is not  matrix
specific.   Another  conclusion  that  can  be  drawn about  these  negative
cyanide  amenable results is that all of the data presented here indicate
that  much more  study needs  to  be  done  so  that this  problem  can  be
thoroughly understood.

TOTAL CYANIDE

In  order to  better understand the  problems  associated with the cyanide
amenable to chlorination  results,  a  greater  knowledge  of all  of  the
contributing factors  used  to determine cyanide concentration  by Method
9010 must be achieved.  There are several variables  inside of the method
that will drastically  affect the performance  of this method.   They are:
distillation time, test portion  size, particle size,  effectiveness of the
sulfuric acid,  and the efficiency  of  the  quantitation  procedure used.
Once  the effects  of  all  of the variables  have  been  taken  into account,
perhaps  it  will  be easier  to  evaluate the  information provided  by the
results  of this method.

DISTILLATION TIME

The section  of  Method  9010  which contains  the instructions that pertain
to the actual distillation step is Section 7.3.4  which gives  the following
instructions:
                                  n-290
-------
         "Heat the solution to  boiling,  taking  care to prevent
         the solution from backing up into and overflowing from
         the air inlet tube.  Reflux  for 1  hour.   Turn off the
         heat and continue the airflow for at least 15 min."

This step  incorporates many variables which are not  entertained  in the
evaluation of the data that this  method  produces.   The actual  length of
time that the sample  is  heated  is directly  proportional  to the recovery
of cyanide in the absorber  tube.   This  is  illustrated in  Figures  7   9.
These figures present the results of 3 actual hazardous waste samples of
varying cyanide  concentration  where all  parameters of the distillation
were held constant except for the length of time they were distilled.

Figure  7  illustrates the  results of an  incinerator  ash  sample  of low
cyanide concentration (<40  ppm),  where a lOg sample size  was used.  The
length  of  distillation time was varied  by one-half hour  intervals.   As
can be seen from this graph, the recovery of  cyanide tends to increase as
distillation time increases.

Figure  8  illustrates the  results  of  a  sample   of moderate  cyanide
concentration (500    4000  ppm), where a lOg sample size  was  used.  The
length  of  the distillation time was varied  by  one-half  hour intervals.
As  can  be seen  from  this  graph,   the  recovery  of  cyanide  increases as
distillation time increases.

Figure 9 illustrates the  results of a  copper  cyanide waste  sample of high
cyanide concentration (2-5%),  where a  Ig  sample  size was used.  As can be
seen from this graph,  the recovery of cyanide increased initially but then
leveled off as the distillation time increased.

Several conclusions  can  be drawn  about  the  dependency of the recovered
cyanide on the length of the distillation procedure.   It seems that when
working with materials  with a  cyanide  concentration   less  than  1%,  the
amount  of  cyanide which  is recovered is  directly proportional  to the
length  of the distillation  process.   This  particular  study only covered
a five-hour time span,  but it  is  believed  that  this  relationship would
continue.   It would be difficult  to  determine  when  all  of the cyanide
would be distilled because  these tests were performed  on actual hazardous
waste whose exact total  cyanide concentrations are unknown.

In  Figure  9,  when a material with a  high concentration  of cyanide was
distilled, the recovered  cyanide seemed to  level  off after  just one hour,
which  may indicate  that  only  a  certain  portion  of  cyanide will  be
distilled  regardless of  how  long the  distillation  period.    Another
possible  source  of  error  could  be that  the  sodium hydroxide  in  the
scrubber will only  trap  a  certain  amount  of cyanide  which this  sample
might have exceeded.   Both of  these  possibilities  will  require further
investigation.
                                  n-29i
-------
In looking at the data, one will also  notice  points  that seem to be out
of line with that particular graph's trend.  It is believed that this may
have been due to  inconsistency in the heating mantles, which may point out
another  potential  source  of  error in  this  procedure.    Perhaps  the
temperature  at which  the  samples reflux will  also effect the recovery.
This  point  would  be especially  significant  at  the  beginning  of  the
distillation  process  when  the procedure  calls  for heating  to boiling,
prior to the  reflux period.  A quicker heating  period may  have a differing
result  possibly  lower  than  a  slower  and  consequently  longer  heating
period.   The increased distillation time,  in turn,  would  increase the
amount of cyanide recovered as shown by the graphs in the above section.

SAMPLE SIZE

The only  guidance  in the  method  regarding the size  of  a  test  portion
states that a sample aliquot of 500 ml  or a sample volume diluted to 500
ml should be used.  This seems to imply that only liquids  can be analyzed
by this method.   Since  this  is the same method being  used  on hazardous
wastes, most  of  which  is  of  a solid  or semi-solid  type matrix, there
really is no  clear guidance as to  what  is  a suitable  test portion size.
When duplicate analyses were run  on some  samples, it was noticed that the
amount  of cyanide  that  was  recovered  differed   significantly.    Upon
investigation, it was observed that the only variable that was different
was  the  size  of the  test portion used for  each  distillation.   This
indicated that the recovery of the  cyanide was dependent  upon the size of
the test portion that was  analyzed.  The  relationship between sample size
(with  all  other  variables   held   constant)  and  cyanide  recovery  is
illustrated  in Figures  10   13.

Figure  10 illustrates the  effect of varying the test  portion from Ig to
lOOg on a material  of relatively low cyanide concentration (25 - 200 ppm).
As can be seen from this graph, the  amount of cyanide recovered decreases
as the test  portion  increases.

Figure  11  illustrates the  effect of varying the test  portion from Ig to
100 g on  a material  of a relatively moderate cyanide concentration (500  -
5000  ppm).    As  can  be seen  from this  graph,  the  amount of  cyanide
recovered decreases  as  test portion increases.

Figure  12 illustrates the  effect of varying the test  portion from 0.25g
to  50g  on a  sample  of relatively high  cyanide concentration  (2500 ppm  -
20%).   The  reason  that  the test  portion  range is  lower in this graph is
due  to the high level  of  cyanide  present  in  the  material.   Again,  the
amount  of cyanide recovered decreases as the test  portion increases.

Figure  13 illustrates the  effect of varying the test  portion from 0 25q
to  lOg  on a  sample of  approximately twice  the cyanide concentration as
Figure 6   Again  the reason for  the lowered test portion  range is due to
the high  level of cyanide present in the material.   Once again, the same
relatTonship,  although  not as dramatic,  seems  to be present with  the
                                  n-292
-------
amount of cyanide recovered decreasing as the test portion is increased.

The obvious  conclusion that can  be drawn from  these  data  is  that  the
sample size  is  inversely  proportional  to the amount of  cyanide that is
recovered from the distillation.   This  relationship, however, did seem to
decline  in  proportion when working  with high level cyanide containing
samples  (concentrations >2%).

EFFECTIVENESS OF THE SULFURIC ACID

The most  basic  assumption  of Method 9010 is that when sulfuric acid is
added  to a  sample it  will  liberate HCN gas.   The  gas  is  then bubbled
through  sodium  hydroxide to convert  the  HCN  to  NaCN, and it is the NaCN
which is subsequently quantitated.  The first comment when discussing the
performance  of  this  method always  seems to be  that  since  many  of  the
matrices  subjected  to this  analysis are of  a  caustic  nature  that  the
amount of acid  is not sufficient to lower  the  pH so as  to  provide  the
proper conditions  for HCN to  be  formed.  Figures  14     16  present  the
information gathered when investigating this potential  problem.

Figure  14 illustrates the  pH  in the  distillation  flask  following  the
addition  of 50 ml  of 50%  sulfuric acid.   The pH  is  compared  to  the
percentage of sodium hydroxide present in a lOg test portion.  It can be
seen from this graph  that  even  with  a sample of 100% sodium hydroxide, if
a  lOg  test  portion  is  used,  the pH  is  <0.4 which provides the proper
conditions for  the generation of HCN gas.

Figure 15 illustrates the change in  pH in the distilling  flask following
the  addition of  the sulfuric  acid using a  lOg test  portion  with  an
increasing percentage of sodium hydroxide in the test portion.

Figure 16 illustrates the  neutralization  capacity of the  sulfuric acid in
the distilling  flask  as  it corresponds  to  number  of  grams  of sodium
hydroxide present in  the flask.  It  can be seen from this graph that even
if  the  material  being analyzed  for  cyanide  is as  caustic  as  sodium
hydroxide, a test portion  of up to 25g  can be used and a sufficiently low
pH will  still be achieved.

The conclusion  that  can  be drawn from the above  information is that as
long as the  test portion used  is 25g or less, the amount of acid required
to be added to  the distillation flask should be sufficient to create the
conditions for  the generation of HCN gas.

PARTICLE SIZE

A common difficulty in all  analytical  chemistry techniques is the ability
to  ensure that a  complete reaction takes  place to  enable  the proper
quantitation of the desired analyte.  Obviously, if the reagents are not
allowed  to  fully  react with  the material  being tested,  an inaccurate
result will  be reached.  One factor that may inhibit a  complete reaction
                                  H-293
-------
 with reagents is particle size.  The effect of particle size on the amount
 of  cyanide  recovered  was  examined  below in  an F006 waste.  The material
 was  run  through a series of  sieves  of the following  pore sizes:   9roni)
 4.5mm, 2mm  and  subsequently analyzed with the results  being presented  in
 Figure 17.

 This  graph  illustrates the relationship  between particle  size and the
 amount  of  cyanide recovered.   Test  portions,  distillation  times, and
 aliquot  sizes were all kept constant.

 In  this  particular waste it  seems  particle size did  not affect the amount
 of  cyanide recovered.   This may  have been  due to  the  fact  that the
 material  used in this  study, regardless of the particle size, completely
 solubilized during the distillation  procedures.   It is also interesting
 to  note  that  the  test portions  taken  from  the material that was >9mm  in
 size  had the  most  consistent  recoveries.   It  is  possible that when this
 material  was  broken  into  smaller  pieces,  the  accurate representation  of
 the material was lost.  In any event, this graph is an excellent example
 of  the variability of  Method 9010.

 CONCENTRATION AS RELATED TO CYANIDE RECOVERY

 In  order to assess the effectiveness of the acid reflux distillation for
 the recovery  of cyanide,  several  concentrations  of cyanide ranging from
 120 ppm  to 10,000 ppm were prepared from  commercially  available  NaCN.
 Approximately lOg of  each NaCN solution was distilled  in  accordance with
 Method 9010 with a reflux/distillation period  of  2  hours.   The results  of
 this study  are  summarized in Figure 18.

 This  graph  illustrates  that  the percentage of  cyanide  recovered   is
 independent of the   initial  cyanide concentration.    On  the  average,
 approximately 90% of the cyanide was recovered regardless  of the initial
 cyanide  concentration.

 These results suggest  that the acid distillation method does not  allow for
 complete cyanide recovery from  aqueous  solutions of simple alkali  metal
 cyanides after  a two-hour period.   Since  the  2-hour reflux/distillation
 period resulted in  an incomplete  cyanide  recovery,  the  aqueous  NaCN
 solutions were further distilled for  periods of 4.75 and  5.75 hours.  The
 results of  this study  are summarized graphically in Figure  19.

This graph illustrates the increase in the amount of cyanide recovered  as
 the distillation time  increases for a  sample  of  aqueous  NaCN.   The NaCN
 stock solutions containing relatively high cyanide concentrations;  i.e.,
 4,000 ppm to 10,000 ppm,  showed only  a  slight increase in the amount  of
cyanide  recovered  after  lengthening  the distillation  period.   Complete
cyanide  recovery,  however,  was  achieved for  those  NaCN  solutions which
 had a low cyanide concentration.

 The results of  this  study on aqueous  NaCN  suggest that  cyanide  can   be
                                 E-294
-------
quantitatively  recovered  for  low  cyanide  concentrations  only  after  a
distillation  period  which  is  considerably  longer than  that which  is
prescribed by Method 9010.  The correlation that can be drawn from these
data to  that which was gathered  on actual  hazardous  waste is:    if  an
incomplete cyanide recovery is achieved on a simple sample matrix of low
concentration  such as  aqueous NaCN  in  a  specified  reflux period,  a
complete  recovery  in  complex  hazardous  waste  matrices  of  unknown
concentrations cannot be expected.

EFFICIENCY OF THE QUANTITATION STEP

The actual quantitation of the cyanide recovered from the material being
tested is performed  on  the distillate collected in the  scrubber system
which converts the HCN gas generated to  NaCN.   Since  this  is the actual
step  which   quantifies  cyanide,   the  effectiveness  of  the  titration
procedure must also be evaluated.   Since the compleximetric titration  is
the  generally accepted method of  determination  for  materials   with  a
concentration level greater than 1  ppm  and the  evaluation  of cyanide  in
hazardous waste  seldom  has  to  be  performed  below  that level,  it is the
titration procedure which is investigated here.

Several stock solutions  were prepared with  cyanide  concentrations ranging
from 100  ppm to  10,000  ppm using  fresh reagent  grade  NaCN.   The silver
nitrate solution  utilized  in the determination  was standardized  against
standard  NaCl  solution  using the  argentometric method  with  K2Cr04
indicator.    The  titrations  were  carried  out   in  accordance with  the
procedure in Method 9010.   The  results of  these  titrations can be seen in
Figure 20.

This graph illustrates the results  of titrations of NaCN which is exactly
what  is  done  in Method  9010.   In each  case, the  amount  of  cyanide
recovered was approximately  90%,   regardless   of  the concentration  of
cyanide involved  in the titration.

The obvious  conclusion reached here is that the quantitation step itself
is not completely effective.  Taking into  account that the compleximetric
titration does not accurately quantify the total  amount of cyanide present
in standard  NaCN  stock  solutions,  it  is  not surprising that the  cyanide
recoveries after  the  acid  reflux distillation are  not quantitative.  It
may be possible that the 10% discrepancy  in the  titration may account for
the  incomplete   cyanide   recovery  experienced   in   the   acid   reflux
distillation.   However,  it is interesting  to  note  that the  amount  of
cyanide recovered for the 120 ppm and  400 ppm aqueous  NaCN solutions
approached 100% after the extended distillation period (Figure 19).

CONCLUSION

The intent of this paper was to explore the effectiveness  of SW 846 Method
9010 for hazardous waste matrices.  The  information  presented has shown
that there  are many  factors  which can  influence  the results that are
                                  H-295
-------
 obtained  by  using this method.   This method does  not  produce a  result
 which can be accurately called "Total Cyanide," since varying  parameters
 within  the method  can  produce  drastically different results.   It  would,
 perhaps,  be  better to  refer to the results  achieved  through the  use  of
 this method as "Cyanide by Method  9010."   It  would also have  been better,
 if when regulatory levels were established, these variabilities  would have
 been taken into account.  These factors should have also been  considered
 when the regulatory level  was set  in terms of forcing this method down  to
 analyzing cyanide  levels  where even  a slight variability  could cause  a
 drastic mistake.

 In  order  to alleviate  some of the  inherent method  discrepancies, one
 suggestion is that the method be made more definite by mandating the use
 of  standardized  quantities  and times.  This will  enable any  lab  which
 analyzes  a  sample  to  arrive  at  the same  result   as  another  lab, not
 necessarily a  "total"  cyanide  but at least  there would  be  consistency.
 Another suggestion is that the  way that the results are reported needs to
 be  regulated,  especially  for cyanide  amenable to chlorination when the
 results are  negative because,  obviously,  there is  more cyanide  in the
 material than first thought.

 In  closing,  Method  9010  either  needs to be better  understood by all
concerned, or  another  way  of determining  cyanide  concentration  must be
developed.
                                H-296
-------
                                               FIGURE 1
                                            INCINERATOR ASH
                                                                                       FIGURE 2
                                                                                   METALS DERIVATIVES
                    P
                    P
                    M
120

100

 BO


 60

 40


 20


  0
                                     CYANIDE CONCENTRATION

                                    H TOTAL CN  •• REMAINING CN

                                                                                    P
                                                                                    P
                                                                                    M
                                                                                       1.5 h
                                                                                       0.5 h
                                                                                                    eH
                                                                              CYANIDE CONCENTRATION

                                                                             H TOTAL CN  •• REMAINING CN
H
N)
                   P
                   P
                   M
1400

1200

1000

 800

 600

 400

 200

   0
                                               FIGURE 3
                                             WASTE SLUDGE
                                     CYANIDE CONCENTRATION

                                    ;H TOTAL CN   • REMAINING CN
                                                                                       400
                                                                                       300
P
p  200
M
                                                                                       100
                                                                                        FIGURE 4
                                                                                  MISCELLANEOUS WATERS
                                                                              CYANIDE CONCENTRATION

                                                                             H TOTAL CN  •• REMAINING CN
-------
                     FIGURE 5
                  MISCELLANEOUS SOLIDS
            FIGURE e
        CYANIDE SOLUTIONS/COMPOUNDS
          3000

          2500

          2000 -
        P
        P 1500
        M
          1000 -

          500 -

           0
 40
 35
 30
 25
 20
 15
 10
  5
  0
                 L
                CYANIDE CONCENTRATION
                H TOTAL CN Hi REMAINING CN
       CYANIDE CONCENTRATION
       H TOTAL CN  IB REMAINING CN
VO
oo
                    FIGURE 7
                                                  FIGURE 8
          40
           l.llll.ll.
           0.5 1.0 1.5  2.0 2.5 3.0 3.5 4.0  4.5 5.0
                    TIME (hr)
 4000
 3500
 3000
 2500
P
p 2000
M 1500
 1000
  500
   0
  .l.llllll
0.5 1.0 1.5  2.0  2.5 3.0 3.5 4.0 4.5 5.0
                   10 g sample
                   I CONCENTRATION
            TIME (hr)
            10 g sample
           I CONCENTRATION
-------
                     FIGURE 9
                                                   FIGURE 10
           50
         P  40
           30
         1  20
         0  10
.Illllllll
0.5  1.0 1.5 2.0 2.5  3.0 3.5 4.0 4.5 5.0
         TIME (hr)
                                        200
                     1 g sample

                    I CONCENTRATION
                                   10   20   50   75

                                     SAMPLE SIZE (g)


                                     IB CONCENTRATION
                                                             100
H
to
VD
                     FIGURE 11
                                                   FIGURE 12
         1
         0 2
         0
         0
          1
                10   20   50   75

                   SAMPLE SIZE (g)


                  • CONCENTRATION
                               100
                               0.25  0.50  1    10   20

                                     SAMPLE SIZE (g)


                                    •I CONCENTRATION
                                                             50
-------
                          FIGURE 13
   60 r
          0.25         0.50           1
                      SAMPLE SIZE (g)

                     • CONCENTRATION
10
                          FIGURE 15
P
H
            20   30   40   50   60   70   80   90
               % SODIUM HYDROXIDE IN SAMPLE
                                                  100
                  AFTER ACID
                                I BEFORE ACID
                                                                N
                                                                A
                                                                L
                                                                P
                                                                H
                          10
20   30   40   50   60    70   80   90
  % SODIUM HYDROXIDE IN SAMPLE
                                               I pH
                                                                                           FIGURE 16
                  P
                  H
                                     10        25        50
                                     SODIUM HYDROXIDE (g)
                                   I AFTER ACID
                                                                                                  I BEFORE ACID
-------
p
o
                                          4.5-9        2-4.5
                                         PARTICLE SIZE (mm)
                                                                  <2
                           •1 ORIGINAL CONC.
                           CD TRIPLICATE CONC.
         DUPLICATE CONC.
                   100
                                            FIGURE 19

                      PERCENT CYANIDE RECOVERED (ppm)
                         120      400      727     4000     6000     10000
                              ACTUAL CYANIDE CONCENTRATION (ppm)
                                  i 2 HR
U.75HR   CD 5.75 HR
                                                                                  100r
                                                                                   80
                                                                                   60
                                                                                   40
                                        20
                                                                                                           FIGURE 18

                                                                                     PERCENT CYANIDE RECOVERED
                                                                                                        I       I
                                                                                         120        400       4000       6000      10000
                                                                                             ACTUAL CYANIDE CONCENTRATION (ppm)
                                                               FIGURE 20

                                          PERCENT CYANIDE RECOVERED
                                                                                  100
100          727         1000         10000
  ACTUAL CYANIDE CONCENTRATION (ppm)
-------
96                           MICROWAVE  DIGESTION
                      PRESSURE  CONTROL AND MONITORING

       Anqelo C.  Grille.  President,  Questron Corporation,  P.O.
       Box 2387,  Princeton,  New Jersey 08543-2387:  Terry  Floyd,
       President,   Floyd   Associates,  5440 Highway  55E, Lake Wylie,
       South Carolina  29710

       ABSTRACT

       Microwave   Digestion   is now an EPA-CLP approved  method  for
       waters  and  solids  digestion  before  analysis by GFAA  and
       ICP.     In  the effort  to recognize Microwave Digestion,  a
       great  deal  of study  was put into calibration of the oven
       power.     If one  were to establish a procedure based upon  %
       power  and  time   in  one oven,  how is the procedure going to
       be  repeated in  another oven?

       In   this  paper we explain true pressure control;  how it  can
       be   used as a much more  reproducible method  for establishing
       procedures,   and   how it can  enhance closed  vessel  microwave
       for many types  of  organic samples.

       Examples  are  offered  to  show  how  reactions in a closed
       vessel   can  be controlled   by small incremental  changes in
       pressure.     Comparisons are  also made between power control
       and pressure  controlled  methods.

       Finally,  we  conclude  that  calibration of  power  introduces
       too  many   variables   and  that the  more correct  method to
       establish   true reproducibility is by incremental  control of
      pressure.
                                  H-302
-------
97        COMPARISON  OF CHROWATOORAPHIC  AND  COT,ORTMETRTC TECTTNTQTJRP  FOR
         ANALYZING CHLORIDE AND SULFATE IN GROUND AND SURFACE WATERS


         Robert  T. Dovi,  Miriam Roman,  Bruce Warden,  Frank Dias,   WMI
         Environmental   Monitoring  Laboratory,  Inc.,  2100  Cleanwater
         Drive, Geneva, Illinois  60134

         ABSTRACT

         A  two part  study  was performed which  compared Suppressed  Ion
         Chromatography,    Non-suppressed   Ion   Chromatography,     and
         Conventional  Colorimetric instrumentation used in analyzing  for
         chloride  and sulfate in ground and surface  waters. Part I is a
         study  comparing Automated  Colorimetric  and Non-suppressed  Ion
         Chromatography  instrumentation. The intent of this  study is to
         show  comparability between EPA  approved methodology 375.2   and
         325.1  (Colorimetric) with EPA  accepted methodology 300.0  (Ion
         Chromatography) when analyzing for secondary pollutants. Part II
         is   a  study  comparing   Suppressed  and  Non-suppressed    Ion
         Chromatographic   instrumentation.   Several   1C  methods    are
         recommended for analyzing inorganic anions in ground and surface
         waters.   They are EPA Method  300.0, ASTM D4327-84, and  Method
         429  which is found in "Standard Methods  for Water and Wastes".
         These  methods are all written using "Suppressed" 1C techniques,
         but  each makes reference to  "Equivalent" instrumentation which
         can  be used  to  generate results.  It  is this instrumentation
         difference outlined in these 1C methods that forms the basis  for
         this comparison.

         INTRODUCTION

         The  Automated Colorimetric instrument used  in this study is  a
         TRAACS800 autoanalyzer.The autoanalyzer is  used to  liberate
         highly  colored ions that form in concentrations proportional to
         the  anions of interest.  In the case of analyzing for chloride,
         mercuric thiocyanate and ferric nitrate react with the sample to
         form  ferric thiocyanate and  mercuric chloride-  The  amount of
         ferric  thiocyanate read  at  480nm corresponds to  the  chloride
         concentration  in the sample.  In the  case of sulfate analysis,
         the  sample first passes through  an ion-exchange column to   get
         rid   of  cations  in  the  sample.   Sample  then  reacts  with
         barium/methylthymol  blue (MTB) in the presence  of acid to form
         barium  sulfate and MTB.  This  mixture then reacts with  sodium
         hydroxide  where excess barium  reacts with MTB  to form a  blue
         chelate.  The amount of uncomplexed MTB (gray in color)  measured
         at  460nm is proportional to the concentration of sulfate in  the
         sample.

         For  Colorimetric analyses,  very little  sample preparation  is
         required.   If  a  sample  is  out of  the linear  range of   the
         instrument,  a dilution is  made.  This occurs  in approximately
         10%   of   samples   tested.    Reagent   and   stock    solution
                                    n-303
-------
preparations  include  standards,  color reagents,  and numerous
reacting solutions.

Tests involving Non-suppressed Ion Chromatography (Waters) allow
for  the simultaneous conductimetric  detection of chloride  and
sulfate  in an eluent of low background conductivity.  Sample is
introduced  into  an  eluent stream  of borate/gluconate  and is
carried  to  the  anion  exchange  column.   Separation  of  the
individual  anions is effected by their  differing affinities to
the  exchange sites of the resin within the column.  Competition
between  the analyte anions and the anions present in the eluent
for  these exchange sites results in the  separation of the ions
of  interest into discrete bands which are carried by the eluent
through the column.  The separated ions are then introduced into
a temperature controlled conductivity detector which produces an
electrical  signal  for  each  of  the anionic  species that  is,
proportional  to the amount  present.  The electrical  signal is
documented  as a  peak  on a recording  device.  The anions  are
qualitatively  identified by the retention time  of the peak and
quantitation  is obtained  by the  measurement of  peak area  as
compared to known standards.

Analyses   performed  by  1C  include   two  preparation  steps:
l)samples  are first diluted then 2) passed through a 0.45micron
filter.  , Few reagents  are used  in testing  with this  system.
Only  borate/gluconate  eluent  and  calibration  standards  are
frequently prepared.

In   the  case  of  Suppressed  Ion  Chromatography  (Dionex)  a
suppressor  "membrane"  is  used to  exchange eluent  and sample
cations  for  hydronium  ions.  These  reactions neutralize  the
eluent  but  do  not  affect  the sample  anions.  The  exchange
reaction  results,in a reduction  of eluent conductivity and  an
increase  in the conductivity of  the sample anions.  As  in the
Non-suppressed  system, anions  are  qualitatively identified by
the  retention time of the peak and  quantitation is obtained by
the  measurement of  peak  area or height  as compared to  known
standards.

SUMMARY

Several  factors  were  used to  evaluate comparability  between
systems:

          1.  Strength of association between techniques.
          2.  Identify systematic bias (i.e.. results from one
                technique generally higher than the other.)
          3.  Proportion of variance in common between two
                variables   (i.e. overlap  of  one technique with
                the variance of the other).
                          n-304
-------
The  data set from Part I of this  study covered a dynamic range
in  analyte concentrations.   The  measured concentration ranges
were as follows:

          .Below detection limit to 1080mg/L for chloride
          .Below detection limit to  767mg/L for sulfate

Statistical  evaluation of this data was provided using a linear
regression  model, pearson correlation  coefficient, coefficient
of  determination, and  paired t-Tests.   The linear  regression
model  described  the  strength of  association between  the two
techniques;    correlation   coefficients   identified    shared
variances, and the paired t-Test was used to identify systematic
bias.

After performing this statistical evaluation, the two techniques
show  a  98.8%  shared variance  for chloride  and 99.0%  shared
variance  for sulfate.  There is no  significant bias of results
at   the  95%  confidence  level,  yet,   with  regards  to  low
concentrations  of  chloride  (around  5ppm),  the  autoanalyzer
produces  consistently lower  values  for chloride than  the 1C.
One  explanation for  this  might be that  in the reaction,  the
presence  of  chloride  is  necessary  for  the  dissociation of
mercuric  thiocyanate  into  mercury  and  thiocyanate.  If  the
chloride  level is low in the solution,  not all of the mercuric
thiocyanate  will  be  dissociated,  therefore,  the  amount  of
measured   ferric   thiocyanate   would  be   low.   Since   the
concentration  of  ferric  thiocyanate  is  proportional to  the
concentration  of chloride in solution, the chloride value would
be   low-    Although   this   difference   is  of   statistical
significance,  the magnitude of the  effect (i.e. bias) is  less
than  1/2  of  a standard  deviation unit  which is  negligible.
Overall,  these results indicate that the  agreement between the
two instruments is excellent for these analytes.

In  order to determine comparability between  systems in Part II
of  this  study,  a statistical  evaluation similar  to the  one
mentioned  in  Part  I was  used.  The  comparison showed  98.7%
shared  variance  for  sulfate  and  99.3%  shared  variance for
chloride.   However, a systematic difference  in chloride values
was    found   for   suppressed   chromatography   relative   to
non-suppressed  technique.  On average, the suppressed technique
yielded  11%  higher  values  for  chloride.   This  finding  is
consistent  with  results  obtained  on  four  runs  of a  lOppm
chloride standard included in the sample batch.  Chloride values
for  this standard averaged 10%  high.  It is possible  that the
increased   chloride  values  may  be  due   to  differences  in
instrumentation calibration.
                          E-305
-------
CONCLUSION

Results  obtained  in  Part  I  of  this  study   show   excellent
correlation between chloride and sulfate values when analyzed by
either  technique.   In Part II, no  difference was  observed  when
comparing sulfate values between 1C systems, yet  the question of
the  chloride discrepancy could warrant further investigation to
determine comparability between techniques.
                         H-306
-------
98                        METALS DIGESTION:  A COMPARATIVE STUDY

         Tim   Fletcher   Analytical  Chemist,  Gary   Wiggenhauser  Group   Leader
         Spectroscopy,  Roger  Kell  Laboratory Manager, Chemical Waste Management,
         Inc., Technical Center, 150 West 137th  Street,  Riverdale,  Illinois   60627

         ABSTRACT

         The  USEPA  SW 846  Method  3050,  2nd Edition "Acid Digestions of Sludges,"
         is a commonly  applied  technique to digest metals from heated samples  in
         an open  vessel using low pressure  and  an acid environment.  Currently,
         closed vessel  microwave digestion  is receiving much  attention as the new
         "high tech" means of metals digestion.   This  is indicated  by the numerous
         microwave  digestion  studies  being reported,  as well  as by  the evolution
         of microwave digestion  systems.   Because many of these studies indicate
         important  applications to our area of analytical  chemistry,  we propose  to
         provide  a  comparative  overview  of  the  various digestion  techniques
         available.  In this study, comparisons are made of the data  produced from
         the  analysis of hazardous  waste streams containing silver,  chromium,  lead,
         arsenic,  and  selenium  digestates utilizing:    1)   open vessel  hot  plate
         digestion;  2)   high  temperature/pressure Parr  bombs; and 3)  a  closed
         vessel Teflon-lined microwave digestion  vessel.  Utilizing  ICAP and GFAA
         methods  of  instrumental   analysis,  these  comparisons  demonstrate the
         precision,  accuracy, efficiency of performance,  and the  applications  of
         each technique as well as commentary on  the  results  obtained.

         INTRODUCTION

         The   analytical  division  of   Chemical   Waste   Management's  Riverdale,
         Illinois,  Technical  Center is  a high  performance analytical  laboratory
         which analyzes over  1,000 samples  per month  for  various parameters.  The
         results of these analyses are used  to determine  the  final disposition  of
         a  customer's  waste  stream and  support  the company's R&D and Methods
         Development efforts.  Common disposal decisions made on the data generated
         are  incineration, landfill ing,  fuels blending, and resource recovery.

         The  Spectroscopy Group at the CWM Technical Center utilizes a modification
         of SW 846  Method  3050  to prepare samples for certain  metals analysis  by
         inductively coupled argon plasma emission  Spectroscopy  (ICAP) and graphite
         furnace atomic absorption  (GFAA)  with Zeeman background correction.  Ag,
         Cr,  and Pb are analyzed by SW 846 Method 6010.   As and Se  are analyzed  by
         SW 846 Methods 7060  and 7740, respectively.

         The  laboratory goal  of  twelve day sample turnaround time includes  the long
         digestion  times needed  for certain samples by the traditional hot  plate
         technique  prescribed  by  Method  3050.   To circumvent  this  problem, ASTM
         Method  C  (E926-88)  "Bomb,  Acid  Digestion  Method"  (modified)  is  often
         employed.  This method  uses Parr  digestion bombs to quickly digest samples
         by increasing  the  temperature/pressure.  This  is an  adequate technique
         but  includes many drawbacks, not least  of which  is the bomb's cumbersome
                                           H-307
-------
nature and possible  contamination due to the breakdown of the bomb's metal
constituents.

In order to improve  costs,  turnaround time,  and handling  techniques while
maintaining QA/QC and  safety  performance,  it was decided to  investigate
microwave digestion  technology as  it applies to hazardous waste analysis.

To  accomplish  this  task,  five sample matrices were  selected.  Of these
five, two were of known values:   1)   an aqueous multi-element  standard;
and  2)   a NBS certified  solid  standard.    The other three samples  were
"typical" of the lab's sample  flow  and were  primarily selected because of
their physical composition:

     SAMPLE 3: an unknown sludge with free  liquid.
     SAMPLE 4: a homogenous metal  dust.
     SAMPLE 5: a mixed oily granulated clay.

To  further  enhance  the "real  world" nature  of the study, five  "problem"
metals  at the  following  wavelengths  were  analyzed  for: Ag  328.07,  Cr
267.72,  Pb  220.3, As  197.3, and Se 196.

The data was generated after digesting each  sample  in duplicate and spike
utilizing  hot  plate (Method 3050), Parr acid  bomb,  and microwave-aided
methodologies.

PROCEDURE/METHODOLOGY

General.  Each sample was  set up in duplicate and  spiked for  determination
of  precision and accuracy.   The spike contained an aqueous 5  ppm spike for
Ag,  Cr,  and  Pb, and  .1 ppm  spike for As and Se.

The sample  matrix was kept at 10% HN03 by  volume,  and  all  samples  were
filtered  through Whatman  42 paper.  Double  deionized water was used  to
bring  the samples  to final volume.   Appendix  I  lists  the  standards,
reagents, and  apparatus used in the study.

Method 3050 (modified). Approximately 2.000g of  test portion  was weighed
into a 125ml Erlenmeyer flask.   Five  (5) mis of HN03  were then added,  and
the  samples placed  on  a hot plate  to near  boiling.  They remained there
until no  NOx   fumes  were  observed.  The  samples were then removed  from
the hot plate and  cooled.   At  this  time, 3mls of  H202 were added, and  the
samples were returned  to  the  hot  plate to  start the peroxide  reaction.
This  peroxide  process was  repeated a  second time, and  the  digestion
continued until the  flask  became clear;  i.e., the production of NOx fumes
ceased.   The samples were  cooled,  diluted  to volume  in a 50ml  volumetric
flask,  filtered,  and analyzed.

Parr Bomb Digestion.   Approximately 0.5g of  test  portion  was weighed  into
the  125ml Teflon sample cups, and  2.5mls  of HN03 and 2.5mls  of H20  were
added.   The cups were then placed  into  the Parr  stainless steel bomb  and
                                  H-308
-------
the bomb assembled.

After bringing up the oven  temperature  to  130 degrees C,  the bombs were
placed inside and digested for 24 hours.  The bombs were then cooled and
disassembled.  The samples were diluted  to 25mls, filtered, and analyzed.

Microwave Digestion.  Approximately 0.5g of  test  portions  were weighed
into 60ml  narrow-mouth bottles.  2.5mls of HNO, were added,  and the samples
were left to react  under  the  fume hood for 4-5 minutes.  2.5mls  of  H20p
were added and after any effervescing  ceased,  the bottles were capped and
tightened.  The caps were then unscrewed 1/4 turn to prevent the bottles
from rupture due to the rising internal pressures.

Before placing the  60ml bottles  into  the Teflon digestion vessels, 2mls
of HN03 were added to the vessel itself in order to maintain the same vapor
pressure within and outside of the 60ml bottles.

The  vessels  were  then  capped and  torqued  to  pressure in  the capping
station,  placed   in  the  sample   turntable,   and   hooked  to  vapor
exhaust/condensate collection hoses, and the digestion was started.

The  following  4-stage  digestion  method  was  developed   by  visually
monitoring at  which power  setting  the internal vessel  pressure caused
venting and condensate to form in the hoses.

                      MICROWAVE  DIGESTION  PROGRAM

         STAGE                   MINUTES                 %POWER
           1                        10                     10
           2                        10                     30
           3                        40                     50*
           4                        60                     45
* Slight condensate formed  toward  end  of stage.   Power reduced, and the
  digestion continued smoothly.

The vessels  are  designed to vent  at lOOpsig, and  lacking  the optional
pressure controller, this method proves adequate.  The main intention is
to  slowly  raise  the  vessel  temperature/pressure to  optimize digestion
efficiency.  When digestion was complete, the vessels were disassembled,
the digestate collected,  diluted to a 25ml volume,  filtered, and analyzed.

RESULTS

The results of the  study  are contained  in  the  chart presented  in Appendix
II.  Each sample  type  is  listed individually.  The left side of the chart
lists the methods  used,  and the top of the chart tells  of  the metal of
interest, listing  the values  for  the  original  (0),  duplicate  (D),  and
spike (S).   Percent error (%E)  and percent recovery (%R) are  also shown.
                                  H-309
-------
Most notable when  reviewing the data are the problems with Ag and Se  spike
recoveries  in  complex  matrices;  also,  the  bomb  method  of  digestion
generally yields the highest concentration.

SAMPLE 3  had precision  and  accuracy problems that  were not  acceptable.
This was primarily due to sample inhomogeneity.   Arsenic errors in SAMPLES
2  and 4  were  due to a  high  aluminum  content.   In  all  samples,   iron
interfered  at   a  level  that  even  with  matrix modifiers  and Zeeman
background correction,  interferences could still not be corrected.

OBSERVATIONS/CONCLUSIONS

This  study  aimed  to compare  the  analytical   equivalence  in  terms of
precision  and accuracy of three methods of metal digestion  by  hotplate,
Parr  bomb, and  microwave.  Many observations can  be made from studying the
derived data.  Appendix  III  lists general advantages  and disadvantages of
each  method.

From   those  "real   world"   samples   in   complex  matrices,   analytical
reproducibility is only  as good as the test portions  and methods of sample
prep  themselves.     As  instrumentation  has  become more  accurate and
sophisticated,  sample  prep  methodologies  have   lagged  far  behind.   In
general,  we can observe (as  expected)  that when sample  homogeneity is
assured,  high temperature/pressure Parr bombs yield the best  recoveries,
and it  is  also apparent that microwave digestion is accomplished  faster.
The volatility of  arsenic and the many spectral and chemical  interferences
on  selenium  also  make firm  conclusions favoring one method over  another
difficult.

At  this  time,  Method 3050  in SW  846  will not  give an  answer that is
absolute.   With this in mind,  the  results  show  that for the  five metals
selected,  all  three  methods have  some  applicability  and  have the  same
probability  of  variability;  some have more severe  problems than others
that  in time must  be addressed as have other methods in the past.

In any case, the endpoint of digestion in  two of the three prep  methods
is a fixed time; only the hot plate method is indefinite.   As was revealed
in the data,  the bomb method appears to generally yield  the larger results
when compared to the others.   Method 3050 uses a reaction indicator  (NOx
evolution) which compensated  for  matrix variation; microwave methods  have
a set  time usually determined by standards or only a  few matrices and  does
not cover all sample variability.

Until  sample prep  methods  become more precise,  laboratories may be  at an
advantage  utilizing  all  three prep  methods (depending  on  sample  types
analyzed).  The validity of test portions and hence analytical  integrity
is a  major  consideration  and  is  guaranteed  by QA/QC   Policy.    Blind
duplicates,  ten    percent    duplicate   and   spike   ratios,   strict
acceptance/sample  reset  criteria,  instrument  performance  checks,  and
                                 n-3io
-------
continuing instrument calibration verification are all among the necessary
ingredients to ensure the production of defensible data.
                                  n-3ii
-------
                               APPENDIX I

                   STANDARDS,  REAGENTS, AND APPARATUS

STANDARDS.   Leeman  Labs  Plasma Pure multi-element  in  10% HN03 for  ICAP
calibration (Lowell, Ma.)-   Single element As and Se for Varian 400  GFAA
in 10% HN03 from EM  Science (Cherry Hill, NJ).  Spike standard  was  multi-
element in  10% HN03  from  Spex  Industries  (Edison,  NJ).

   REAGENT  WATER.   Double deionized from a Millipore system.

   HN03.  Baker Instra Analyzed 70.0%-71.0%.

   H202.  Baker Analyzed--30%.

   HOT PLATE.   Variable temperature control.

   FILTER  PAPER.   Whatman 42.

   FLASKS.   125ml  Erlenmeyer with Tuttle caps, 50ml  and 25 ml
   volumetrics.

   PARR BOMBS.  Model 4748 with 125ml  Teflon sample cups.

   BOMB OVEN.   Thermolyne mechanical  oven with temperature and
   timer settings.

   MICROWAVE DIGESTION SYSTEM.  CEM model  MDS-81D 600 watt microwave oven.
   120ml. Teflon sample vessels.   Turntable/vessel carrier.  Vent  tubing
   and vapor collection container.  Capping station.

   NARROW-MOUTH BOTTLES.   60ml. FEP Fluorocarbon.

   INSTRUMENTATION.   Leeman  Labs  Plasma Spec 2.5 and Varian 400 GFAA with
   Zeeman correction.
TF/gf
5/17/90
                                 n-3i2
-------
                                                                              SAMPLE 1 QC ,

METHOD
HOT PLATE
BOMB
MWD

! Ag
0
3.97
<.055
<.034

D
4.71
<.034
<.034

S
9.52
<.034
<.034

%E
8.52
23.5
0

%R
103.6
0.21
0

Cr
0
5.18
5.58
5.17

D
5.15
5.81
5.12

S
9.95
10.8
10.4

%E
0.29
2.02
.486

%R
95.7
102
105

Pb '
0
5.11
5.51
5.05

0
5.07
5.81
4.99

S
9.76
11.0
10.3

%E
0.39
2.65
.598

%R
93.4
107
106

As
0
.119
.130
.135

D
.121
.140
.138

S
.232
.262
.248

%E
0.8
3.7
1.1

%R
112
127
112

Se
0
.108
.131
.122

D
.107
.136
.122

S
.175
.239
.219


%E
0.5
1.9
0


%R
67.5
106
97

                                                                 SAMPLE 2 NBS CERTIFIED SOLID STANDARD

METHOD
HOT PLATE
BOMB
MWD

Ag
0
<.79
<1.7
<1.7

D
<.79
<1.7
<1.7

S
12.3
<1.7
<1.7

%E
0
0
0

%R
10.6
.044
0

Cr '
0
93.9
112
88.1

D
95
107
89.7

S
197
346
329

%E
.582
2.28
0.90

%R
88.5
96.4
97.9

Pb '
0
122
143
159

D
122
138
126

S
222
353
366

%E
0
1.78
11.5

%R
86.3
86.7
91.1

As
0
20.2
24.8
22.6

D
18.7
28.5
23.0

S
20.8
28.9
28.9

%E
3.9
6.9
0.9

%R
60
10.0
120

Se
0
.16
<.31
.38

D
.16
<.33
<.34

S
.44
2.40
3.14

%E
0
0
100

%R
19
49
64

                                                                         SAMPLE 3 SLUDGE/LIQUID

METHOD
HOT PLATE
BOMB
MWD

Ag
0
<.80
<1.6
<1.5

D
<.80
<1.6
<1.5

S
10.3
1.80
<1.5

%E
0
0
0

%R
8.78
0.734
0

Cr
0
107
173
290

D
424
132
83.5

S
205
343
299

%E
59.7
13.4
55.2

%R
51.5
77.7
49.3

Pb
0
214
304
318

D
147
229
99.5

S
180
368
486

%E
18.5
14
523

%R
40.4
41.4
121

As
0
7.44
9.82
1.83

D
6.38
6.79
1.81

S
7.89
12.4
5.82

%E
8.1
18.2
0.5

%R
42.0
72.0
85.0

Se
0
99.6
151
24.6

D
83.3
118
16.8

S
92.0
449
51.8

%E
8.9
12.3
18.8

%R
0
314
65

                                                                          SAMPLE 4 METAL DUST

METHOD
HOT PLATE
BOMB
MWD
Ag
0
<.84
<1.6
<1.6
D
<.84
<1.6
<1.6
S
37.1
2.89
<1.6
%E
0
0
0
%R
30.1
1.27
0
Cr
0
605
518
455
D
626
529
444
S
736
742
648
%E
1.7
1.05
1.22
%R
97.6
96.1
81
Pb
0
51.8
75.2
71.4
D
50.5
79.2
74.9
S
126
259
247
%E
1.27
2.59
2.89
%R
60.6
80
70.9
As
0
9.12
10.9
9.95
D
10.2
10.8
10.2
S
9.06
23.4
18.4
%E
5.6
0.5
2.3
%R
0
223
172
Se
0
<.17
<.32
<.32
D
<.17
<.34
<-33
S
1.44
3.36
2.84
%E
0
0
0
%R
59.0
74.0
58.0
                                                                           SAMPLE 5 OILY CLAY

METHOD
HOT PLATE
BOMB
MWD

Ag
0
<.82
<1.6
<1.7

D
<.82
<1.6
<1.7

S
81.2
<1.6
<1.7

%E
0
0
0

%R
68.2
0
0

Cr '
0
56.4
62.3
34.3

D
56.4
61.0
53.1

S
163.0
296
282

%E
0
1.05
21.5

%R
89.5
99.3
102

Pb '
0
2.75
2.89
2.35

D
2.31
1.90
3.97

S
99.5
212
209

%E
8.69
20.6
25.6

%R
81.4
88.8
88.9

As '
0
0.50
3.56
3.41

D
0.69
3.68
3.62

S
0.27
7.78
7.35

%E
16.0
1.7
3.0.

%R
0
86.0
83.0

Se
0
<.17
<-32
<.30

D
<.17
<.32
<.31

S
0.52
2.59
3.01

%E
0
0
0

%R
22
55
65

TF/gf
5/17/90
-------
                             APPENDIX  III

                  METHOD ADVANTAGES AND DISADVANTAGES
   METHOD
          ADVANTAGE
                                                  DISADVANTAGE
 HOT PLATE
1)
2)
 PARR BOMB
 MICROWAVE
1)

2)
3)
4)
5)

6)
1)
            2)

            3)
Analyst acceptance.
Method currently gener-
ates defensible data.
Higher temp./pressure
yields better recoveries,
No acid loss.
24-hour digestion time.
No loss of samples.
Equipment not exposed to
acid fumes.
Applicable to all
matrices.

Staged program gives
strict control over
temp./pressure.
Approx. 2-hour digestion
time.
Sample contamination
minimized.
1)  Digestion time varies
    1-10 days.
2)  Equipment in contact with
    acid fumes.
3)  Constant analyst observa-
    tion.
4)  Boil-overs greater  safety
    risk.
5)  Applicable to certain
    metals only.
6)  Sludge digestion proce-
    dure.

1)  Bombs are cumbersome.
2)  Bombs' metal constituents
    may break down causing
    contamination problem.
3)  Bombs costly.
4)  Small sample size accen-
    tuates inhomogeneous
    samples.
1)  Microwave digestion
    system is expensive.
2)  Labor intensive.
3)  Reduced sample size.
TF/gf
5/17/90
                                 n-314
-------
99    Using Flow Injection to Meet QA Criteria for ICPMS Method 6020






     D.J. Northington and Michael Shelton



     West Coast Analytical Service, Inc., 9840 Alburtis Ave, Santa Fe



     Springs, CA 90670








     In EPA Method 6020, Inductively Coupled Plasma-Mass Spectrometry



     (ICPMS) is used to determine trace metal analytes in sample



     digests.  However it has been reported that under the current



     protocol of conventional sample introduction using solution



     nebulization difficulties are encountered in meeting QC



     control limits such as continuing calibration and internal



     standard response drift (1,2).  Many of these problems can be



     caused by deposition of solids on the sampling and skimmer cones



     and detector fatigue form high analyte concentrations.  In this



     work a protocol was sought using flow injection of the sample



     digests which would minimize exposure of the equipment to the



     sample matrix.  By flow injecting a small aliquot of the digest



     and by limiting data aquisition times to correspond to the flow of



     the aliquots, one can reduce the amount of sample introduced from



     several milliliters to fractions of a milliliter.  The results of



     this study on continued calibration and internal standard response



     drift will be presented.  Additional benefits of increased sample



     through put and reduced memory effects will also be demonstrated.








     1.  Aleckson, K., et al., Fifth Annual Waste Testing and
                                   n-315
-------
Quality Assurance Symposium, July 24-28, 1989.



2.  Hinners,  T.A.,  et. al.,  Fifth Annual Waste Testing  and



Quality Assurance Symposium, July 24-28, 1989.
                             n-316
-------
10°      COMPARISON OF THE DETERMINATION OF HEXAVALENT CHROMIUM
                  BY ION CHROMATOGRAPHY COUPLED WITH ICP-MS
                                 OR WITH COLORIMETRY

         Raimund  Roehl. Ph.D.. and  Maricia  M.  Alforque,  California  Public  Health
         Foundation, California  Department  of Health  Services,  Hazardous  Materials
         Laboratory, 2151 Berkeley Way, Berkeley, California, 94704


         ABSTRACT

         A method for the  determination of hexavalent chromium  in aqueous samples or
         sample extracts using ion chromatography (1C) combined with inductively coupled
         plasma  mass spectrometry (ICP-MS)  has been  developed.  Its performance is
         compared to a method using the same ion chromatographic separation coupled with
         a post-column reactor and the colorimetric  detection of  a Cr(VI)  diphenylcar-
         bohydrazide complex. The observed detection limits and linear dynamic ranges are
         similar for both methods, i.e., about 1-2 ppb and 4 decades respectively. Compared
         to the colorimetric method, IC-ICP-MS has the advantage that  it  can also  be
         employed to  determine oxy-anions  of  other metals,  such as arsenic,  selenium,
         vanadium, molybdenum and tungsten. In  addition, ICP-MS without the  preceding 1C
         separation can be used to quantitate total metals. In the case of chromium, Cr(III)
         can then be determined  by difference.
         INTRODUCTION

         The determination  of  hexavalent chromium in addition  to  total and  trivalent
         chromium in environmental samples is important because of the large difference in
         toxicity  between Cr(III)  and Cr(VI).  Most  popular  methods for the  selective
         determination of hexavalent  chromium are based on the reaction of Cr(VI) with
         diphenylcarbohydrazide (DPC) in acidic solution, resulting in the formation of a
         complex which  has  an absorption maximum at 540 nm [1-3]. Analysis of bulk
         samples using this methodology is possible, but suffers from potential interferences
         by other colored species (e.g., Fe(III) or Cu(II) complexes)  or  species forming
         colored  reaction products with DPC such as vanadium, molybdenum and  mercury.
         Low results and  poor spike recoveries are observed when samples contain substances
         which reduce Cr(VI) in acidic solution [1,2]. Furthermore, the presence of chemical
         species  which  oxidize  Cr(III)  to  Cr(Vl),  e.g.,  free  chlorine, can give rise to
         erroneously high results [3].

         To  avoid the aforementioned interferences, methods have  been developed which
         isolate hexavalent chromium by ion chromatography before reacting it with DPC
         [4,5]. The chromatographic separation of Cr(VI) from other sample  constituents, and
         especially from other chemical species of the same element, also makes it possible
         to use element specific atomic spectrometry methods for its selective determination.
                                            n-317
-------
All of the major atomic spectrometry techniques employed in environmental analysis,
namely atomic absorption spectrometry (AAS), inductively coupled plasma atomic
emission spectrometry (ICP-AES), and inductively coupled plasma mass spectrometry
(ICP-MS) have been interfaced with liquid chromatography for chemical speciation
work Of those detection methods, ICP-MS is the most versatile and useful, because
it  can measure several elements simultaneously and with very high sensitivity.

One of the first published reports on interfacing ICP-MS with liquid chromatography
(LC) appeared in 1986 [6],  only three years  after the  first  commercial ICP-MS
instrument was introduced. Since then, a  number of research groups, including our
own, have explored the potential  of coupling ICP-MS with  LC for  studies involving
the speciation of metals [e.g., 7-10] as well as non-metals [10,11]. Although interest
in this area  has been growing steadily, progress is hampered somewhat by the fact
that specialized data transfer techniques are still  required to analyze  LC-1CP-MS
chromatograms,  i.e., to perform peak  detection and integration (cf. Experimental
section).

 In this report, a method for the determination  of Cr(VI) based on the combination
 of 1C  and  colorimetry  is compared  to  a  technique which  uses  ICP-MS for the
 detection of chromium in the chromatographic effluent. In  addition, applications  of
 ICP-MS to the selective detection of oxy-anions of vanadium, molybdenum, tungsten,
 arsenic and selenium  are presented.
 EXPERIMENTAL

 All ion chromatographic separations were performed  with  a Dionex 4000i  ion
 chromatograph, a Dionex AS4A anion exchange  column and an AG4A guard
 column. Typical operating conditions and the main  eluent system employed in  this
                                                    work are summarized in
 	  Table  1.  Other  eluent
                                                    systems  and   conditions
                                                    were also tested; they are
                                                    described in the Results
                                                    and  Discussion sections
                                                    where  appropriate.  The
Table 1:
1C operating conditions and eluent
system
    Injection volume

    Elution mode
    Eluent
    Flow rate
                     25 or 100 ul
                     isocratic
                     6 mM (NH4)2SO,, 10'5 M
                     HC1O4; pH adjusted to 9.0
                     with NH4OH

                     1.0 ml/mm
                                       ICP-MS instrumentation
                                       and operating parameters
                                       are listed in Table 2.

                                       For ICP-MS  work,  the
                                       effluent from  the analyti-
                                       cal column was passed to
                                       the ICP nebulizer through
                                       a  60   cm long  Tefzel
                                       transfer line (1.5 mm OD,
                                    n-318
-------
0.3 mm ID). At the nebulizer end, the Tefzel tubing was stretched and cut to form
a taper in order to reduce the dead volume  at this juncture.
   Table 2:
ICP-MS instrumentation and operat-
ing conditions
                                                    For   the    colorimetric
                                                    work, the 1C effluent was
                                                    mixed  with  the  DPC
                                                    reagent  in  a  T-fitting
                                                    (Dionex PN 024313) and
                                                    then  passed to  the UV
                                                    detector, using either a 60
                                                    cm Tefzel  (0.3  mm ID)
                                                    transfer line or a Dionex
                                                    packed-bead reaction coil
                                                    (PN 036036), which has a
                                                    length  of  122 cm. The
                                                    color  reagent  was pre-
                                                    pared  fresh daily by dis-
                                                    solving  50  mg  of  1,5-
                                                    diphenylcarbohydrazide in
                                                    10  ml methanol, adding
                                                    this solution to a mixture
                                                    of 80 ml deionized water
                                                    and 5  ml  concentrated
     ,,, and making up to a final volume of 100 ml. The DPC reagent was supplied
to the mixing "T" by a Gilson Minipuls II peristaltic pump using 0.76 mm ID pump
tubing. Absorbance of the effluent/reagent mixture at 540 nm was measured with
a Linear UVIS 203 detector equipped with a 6mm path length cell. The detector
signal was acquired, stored  and analyzed using a  Dynamic Solutions Maxima 820
chromatography data station and software.

In the case of ICP-MS detection, the time-dependent signals for metal ions were
acquired by the ICP-MS computer with  the Multiple Elements program. The data
in ASCII format was transferred to the  Maxima computer via serial interface and
translated  to data interchange format (DIP) using a Turbo Pascal program written
by one of the authors (RR). The DIP files were read and analyzed with the Maxima
software in the same fashion as the UV chromatograms.
   Spectrometer
   RF power
   Plasma gas flow
   Auxiliary gas flow
   Nebulizer gas flow
   Nebulizer type
   Spray chamber

   Interface cones
   Data acquisition mode
   Individual dwell time
   Total integration
   time per point
            Sciex ELAN 500
            1.25 kW
            14 1/min
            1.4 1/min
            1.24 1/min
            Meinhard
            Water cooled
            (20 °C)
            Platinum
            Multichannel
            50 ms

            1 or 2 s
RESULTS

Ion Chromatography.  The ion chromatographic conditions used in this study were
adopted after considerable experimentation with different columns, eluent systems,
and flow rate settings. Our primary requirements for the separation system were: (1)
an eluent  system which is compatible with the ICP-MS and colorimetric  detection
methods,  (2) a column/eluent combination resulting in good  chromatographic
                                   n-319
-------
 separation of chromate from other chemical species, and (3) short analysis times
 (preferably 10 min or less).

 Most of this  work was performed with  a Dionex AS4A anion exchange  column
 because that  column type had been  employed successfully in earlier IC-ICP-MS
 speciation studies  [10]. The utility of a Dionex OmniPac PAX-500 column was also
 explored  briefly. However, the necessity of maintaining a  low percentage of an
 organic solvent (e.g.,  1 % methanol or acetonitrile)  in the eluent resulted in
 excessively high background counts from  40Ar12C+ when 52Cr was monitored by ICP-
 MS.
 Eluent systems based  on
 slightly    alkaline
 ammonium sulfate  solu-
 tions had been used suc-
 cessfully for Cr(VI) deter-
 minations   with   other
 anion exchange columns
 and  colorimetric  detec-
 tion  [4,5]. We  observed
 some  peak   tailing for
 chromate when a pH  9
 buffered   ammonium
 sulfate eluent was tested
 in  conjunction  with  an
 AS4A column. This  peak
 tailing  could  be  mini-
 mized by adding  10"5  M
 perchlorate to the eluent.
  1.50
  0.90-
 030-
 -030-
 -0.90-
-1.50
          uv
                                      •25OO
                                      •2000
ICP-MS
                                       3000
                                       1500
                                       10OO
                                       500
                                             •a
                                             8
                                    c

                                    OJ
                  4      6

                  Tim© (min)
                            10
Figure 1:
Chromatograms obtained for a 10 ug/1 solution
of Cr(VI) with colorimetric detection and ICP-
MS detection of Cr-52
 Figure 1 shows chromatograms obtained for a 10 ug/1 Cr(VI) solution analyzed by
 IC-colorimetry and IC-ICP-MS  using the conditions given in the Experimental
 section. As expected,  the peaks  obtained with ICP-MS  detection were  somewhat
 wider because of additional band broadening in the ICP spray chamber.

 Detection Limits. Exploratory measurements had indicated that the detection limits
 for Cr(VI) with both detection methods and with both isotopes used in ICP-MS (52Cr
 and  Cr) were in the  range 1-3 ug/1. The actual  detection limits were determined
 by making nine repetitive injections of a 5 ug/1 Cr(VI) standard onto the 1C column
 [c. ^'J) and lettinS  the chromatography software find peak areas  and peak
 heights. The peak detection parameters were set using the software's automatic peak
 integration setup mode. Detection limits were taken to correspond to three times the
standard deviations of the peak areas or heights for a given set of conditions. All of
the results compiled in Table 3 were obtained using 100 ul injections and  the eluent
                                   H-320
-------
parameters  given  in Table 1. It was interesting to find  that significantly smaller
injections (25 ul) resulted in only slightly inferior detection limit values.
   Table 3:   Chromium (VI) Detection Limits in
              ug/r
   Colorimetry with PER2

   Colorimetry without PER

   ICP-MS using 52Cr

   ICP-MS using 53Cr
                                 Peak
                                 Area

                                   0.9
                                   2.2

                                   1.3

                                   1.9
Peak
Height

 0.8

 1.5
 1.1
 1.4
     Based on three standard deviations of integrated peaks

   2 PER = Packed-bead reactor
    -1.00
          B
                                           2.00
                                                      The   colorimetric
                                                      measurements  were re-
                                                      peated with  and without
                                                      the  packed-bead reactor
                                                      (PER) inserted between
                                                      the  mking  "T" and the
                                                      UV  detector. As can be
                                                      seen from Fig. 2, use of
                                                      the  PER   resulted  in
                                                      somewhat  larger  peaks
                                                      and  less  baseline  noise,
                                                      which  explains the  im-
                                                      provement of the detec-
                                                      tion  limits by a factor of
                                                      about 2.  The  chromato-
                                                      graphic traces  shown in
                                                      Fig.  2 were also analyzed
                                                      to   estimate  detection
                                                      limits based  on the  com-
                                                      monly  employed method
                                                      of   calculating    three
                                                      standard deviations of the
                                                      baseline noise. Values of
                                                      0.4  and   0.6  ug/1   were
                                                      obtained  for the experi-
                                                      ments  with   and  without
                                                      the  PER   respectively;
                                                      they are  comparable to
                                                      the detection limits given
                                                      in Table 3.

                                                      With ICP-MS, the signals
                                                      for  52Cr  and 53Cr  were
                                                      recorded  simultaneously,
                                                      i.e.,  the total acquisition
                                                      time was divided between
                                                      the  two  isotopes.  The
                                                      signal-to-noise ratios, and
                                                      therefore  the  detection
                                                      limits, could  probably be
improved by  a factor  of 1.4  by employing single  ion monitoring. A detailed
description of the advantages  and disadvantages of  using  either one of the two
chromium isotopes for ICP-MS measurements is given in the  Discussion section.
    1.00


    0.60


<   020-


»  -0.20-


   -0.6O
                                          M-60
                                          • 1.20   m
                                           OJBO
                                           0.40
                                           0.00
        18
              20
                     22
                            24
                                   26
                                         28
                      Time (mm)
   Figure 2:
            Four repetitive injections of a 5 ug/l Cr(VI)
            standard onto the 1C column using colorimetric
            detection: (A) with, (B) without the packed-
            bead reactor
                                    H-321
-------
Linear Dynamic Range.  The upper concentration limit of the colorunetnc method
was assumed to be reached when the measured absorbance exceeded 10 AU. With
our 1C system and UV detector, this was found to occur at about 10 mg/1 of Cr(VI).
Analysis of a series of standards ranging up to 10 mg/1 showed that there was no
significant deviation from linearity between the detection limit and 10 mg/1, resulting
in a linear dynamic range spanning at least four  decades.

For ICP-MS detection, the dynamic range was expected to be limited by channel
electron multiplier (CEM)  saturation,  which generally occurs when count rates
exceed 106 s'1 Under the experimental conditions used in this study, CEM saturation
should have been observed for 52Cr at  Cr(VI) concentrations  exceeding about 30
mg/1  (or 300 mg/1 for 53Cr). However, peak heights as well as peak areas for both
isotopes began to deviate from linearity at concentrations above 10 mg/1. Therefore,
the linear dynamic range of the IC-ICP-MS  combination is identical to that of the
colorimetric method.

Analysis of Field  Samples.   The  performance of the two  ion chromatography
methods with actual field samples containing low levels of Cr(VI)  was Devaluated by
analyzing a series of filters, probe washes and impinger solutions which had been
collected during recent emission tests at a cement plant burning hazardous waste as
a supplemental fuel. The samples were derived from special Cr(VI) sampling trains,
                                                     consisting    of  Teflon
                                                     coated  glass fiber  filters
                                                     and impingers  filled with
                                                     0.02 M NaHCO3. Sodium
                                                     bicarbonate solutions  of
                                                     the same  strength were
                                                     used  to rinse the sam-
                                                     pling probes  and to ex-
                                                     tract  the filters. Whereas
                                                     none  of  the  impinger
                                                     solutions had  detectable
                                                     concentrations  of  hexa-
                                                     valent   chromium,  two
                                                     probe washes  and filter
                                                     extracts  were  found  to
                                                     contain  Cr(VI)  at  low
                                                     ug/1  levels. The results
                                                     obtained  with the  two
detection  methods  are compared in  Table 4.  Considering  that  the measured
concentrations are very close to the detection limits of both methods, the results are
in excellent agreement.
Table 4:
Sample
Chromium (VI) concentrations in five
field samples as determined by IC-
colorimetry and ICP-MS using 52Cr.
The units are ug/1.
      IC-Colorimetry  IC-ICP-MS
Probe Wash 1
Probe Wash 2
Filter Extract 1
Filter Extract 2
Soil Extract
4.3
2.0
2.0
5.3
284
4.0
1.9
1.6
4.7
264
                                    n-322
-------
It may be noted that when the same samples were re-analyzed three months after
collection, essentially identical results were  obtained. This indicates that 0.02 M
NaHCO3 solution is a suitable storage medium for hexavalent chromium.

The performance of IC-colorimetry and IC-ICP-MS with a complex matrix containing
higher levels of Cr(VI) was tested by analyzing an alkaline extract of a soil sample
from the Stringfellow hazardous waste site in Riverside Co., California. This sample
contains a significant amount of humic material  and high  levels of chloride and
sulfate, in addition to a large number of other constituents. Neither method had any
problems with this matrix  and the  results for hexavalent chromium are  in good
agreement (Table  4).
DISCUSSION

This section addresses several issues which are relevant to the application of ICP-
MS as a detection method for chromium in general and to the combination of ICP-
MS with ion chromatography for the specific determination of Cr(VI) in particular.
It also discusses some applications of IC-ICP-MS to the speciation of other elements.

Detection of Chromium Isotopes by ICP-MS.  Chromium has four stable isotopes:
^Cr (4.31 % relative abundance), 52Cr (83.8 %), 53Cr  (9.6 %)  and 54Cr (2.38 %).
However, only the isotopes with masses of 52 and 53 are of analytical utility in ICP-
MS, because measurements of the 50Cr and 54Cr isotopes generally suffer from high
background counts due to 36Ar14N+ and 38Ar16O+ respectively.

When clean water or dilute nitric acid are analyzed by ICP-MS, the background at
mass 52 is mostly due to 36Ar16CT and is significantly higher than that at mass 53
(36Ar17O+, 36Ar18O1H+).  This largely offsets the abundance advantage  of the  52Cr
isotope for the determination of chromium, as illustrated by the detection limit
results given in Table 3. Additional interferences for 52Cr can arise when samples
contain high concentrations  of carbon (e.g., TCLP extracts) or  sulfur; the primary
interferant species in those cases are 40Ar12C+ and 36S16O+. The measurement of the
53Cr isotope can suffer significant interference from 37C116O+ when high concentra-
tions of chlorine are present.

Although the list of possible interferences in the determination of chromium by ICP-
MS may appear intimidating at first, it must be emphasized that those interferences
do not seriously affect the IC-ICP-MS method  for Cr(VI) presented in this report,
because  ion chromatography  effectively  separates  chromate from  potentially
interfering anionic species which may occur at high concentrations in environmental
samples (e.g., carbonate, sulfate and chloride). The level of sulfate in the 1C eluent
used in this work (6 mM) increased the background for 52Cr only to a small extent.
It should be noted that  an eluent containing 250 mM (NHJ2SO4, such as  is used
with the Dionex AS7 column [5], would probably preclude the use of  52Cr  for IC-
ICP-MS determinations of Cr(VI).
                                    H-323
-------
The  aforementioned interferences must  be considered when total  chromium is
determined by ICP-MS without a preceding 1C separation. In this case, the presence
of potentially interfering concentrations of carbon, sulfur or chlorine can be detected
by simultaneously  monitoring  for 13C+, 34S+ and 35C1+. The best isotope for  the
quantification  of chromium can then be selected  based on  the results of those
additional measurements. The extent to which molecular ion species, such as ArC+
or C1O+, affect ICP-MS  measurements can also be reduced by mixing  a small
amount of  nitrogen into the  argon supplying  the ICP  [Roehl and  Alforque,
unpublished work].

Application of IC-ICP-MS to Other Oxy-Anions.  Earlier work performed in this
laboratory had indicated that vanadate, molybdate, tungstate and chromate could be
separated with an AS4A column using a 2 mM Na2CO3  eluent [10]. However, even
at a flow rate of 2 ml/min, the retention time for chromate was rather long (20 min;
capacity factor k'  = 31). Oxy-anions of the different oxidation states of arsenic (III
and V) and selenium (IV and VI) were separated very well by this chromatographic
system, with retention times ranging from 0.9 to 13.8 min (k' = 0.16  17). It should
be noted that those earlier experiments were performed with an older AS4A column
than the one used  for the work presented here.

Switching to a new column and changing from a 2 mM Na2CO3 eluent at a flow rate
of 2 ml/min to a 6 mM  (NHJ2SO4 eluent at 1 ml/min resulted in a significantly
shorter retention time for chromate  (about 7.5 min); the capacity factor for Cr(VI)
was reduced from 31 to 3.8. Therefore, it was of interest to also explore the behavior
of other oxy-anions under the new 1C conditions.

Figure 3 shows a set of chromatograms obtained when a mixed standard of vanadate,
tungstate, molybdate and chromate was analyzed by ICP-MS. The concentration of
each metal was 10 ug/1. Vanadate exhibited pronounced peak tailing, which made
peak area determinations difficult. Tungstate, molybdate  and chromate, on the other
hand, produced sharp,  well resolved peaks. Compared  to our earlier experiments
with the sodium carbonate system, all four oxy-anions eluted more rapidly.

Similar results were obtained  for  the inorganic  arsenic  and  selenium  species.
Arsenite eluted at  1.8 min, i.e., just  after  the void volume, closely followed by
selenite and arsenate (both at 2.7 min). Eluting after 4.0 min, selenate  was the most
strongly retained species in this set of four analytes.  With the carbonate system, the
elution order had been arsenite < selenite < selenate < arsenate.

When an alkaline extract of a soil sample from  the Stringfellow site (cf.  Results
section) was analyzed for oxy-anions, no arsenic or selenium was detected.  However,
vanadate, tungstate and molybdate were present in addition to chromate (Fig.  4).
Based on peak height analysis, their concentrations were estimated to be  49 ug/1
(vanadate), 3.6 ug/1 (tungstate)  and 45 ug/1 (molybdate) respectively. The Cr(VI)
concentration (264 ug/1) was already given in Table 4.
                                   H-324
-------
             16OO
             12OO-
              80O-
              400-
                                     Molybdate
                       Tunastat©
                   Vanadate
                                  6       9

                                   Tim©  (min)
                                                  12
                                                     12


                                                     1O


                                                   -8


                                                   -6


                                                    4


                                                    2
                                                          15
        Figure 3:   Chromatograms of a standard containing 10 ug/1 each of V, W, Mo
                  and Cr in the form of their oxy-anions. V-51, W-182, Mo-98 and Cr-
                  52 were detected simultaneously by ICP-MS
     32OO



 _  2400

 £

 c  16OO-
              8OO-
                                                            1OO
                                                            80
                                                            •6O
                                                            •4O
                                                            •20
                                                  12
                                                          15
                                   Time  (min)
Figure 4:
                  Chromatograms of an alkaline soil extract from the Stringfellow site
                  showing the presence of vanadate, tungstate, molybdate and chromate.
                  See Fig. 3 for peak identification
SUMMARY

The method  comparison  in this  report  has shown that IC-ICP-MS is a viable
alternative to the more classical technique of IC-colorimetry for the determination
of hexavalent chromium in aqueous samples. The detection limits and dynamic ranges
of both methods are similar. Compared to colorimetry, ICP-MS is a much more expen-
sive technology  to implement in a laboratory, but if the instrumentation is already
available, IC-ICP-MS has the advantage that it can be used to speciate other elements
as well.
                                     n-325
-------
ACKNQWT .HDGEMENTS

The authors thank John Riviello and Robert Joyce of Dionex Corporation for suggesting
the addition of perchlorate to the 1C eluent and for several very helpful discussions.


REFERENCES

[I]   Standard Methods for the Examination of Water and Wastewater, 17th edition.
     American Public Health Association,  Washington, B.C. (1989)

[2]   SW-846 Method 7196

[3]   Water Analysis. Fresenius, W., K.E. Quentin and W. Schneider (eds.), Springer
     Verlag (1988)

[4]  Technical Note 24, Dionex Corporation, Sunnyvale, CA, May 1987

[5]  Joyce, R.J. and A. Schein, 1C: A powerful analytical technique for environmental
     laboratories. American Environmental Laboratory, 1, No. 2, 46  (1989)

[6]  Thompson, J.J. and R.S. Houk, Inductively Coupled Plasma Mass Spectrometric
     Detection for Multielement Flow Injection Analysis and Elemental Speciation
     by Reversed-Phase Liquid Chromatography. Anal.  Chem., 58, 2541 (1986)

[7]  Dean, J.R., S.  Munro, L. Ebdon, H.M. Crews  and R.C.  Massey, Studies of
     Metalloprotein  Species  by  Directly  Coupled  High-performance  Liquid
     Chromatography Inductively Coupled Plasma Mass  Spectrometry. J. Anal. At.
     Spectrom., 2, 607 (1987)

[8]  Heitkemper, D., J. Creed, J. Caruso and F.L. Fricke, Speciation of Arsenic in
     Urine Using High Performance Liquid Chromatography with  Plasma  Mass
     Spectrometric Detection. J. Anal. At. Spectrom., 4,  279 (1989)

[9]  Suyani, H., D. Heitkemper, J. Creed and J. Caruso, Inductively Coupled Plasma
     Mass Spectrometry as a  Detector for Micellar Liquid Chromatography: Speciation
     of Alkyltin Compounds. Appl. Spectrosc., 43, 962 (1989)

[10]  Roehl, R., New LC-ICP-MS Techniques,  In: Proceedings of the Fifth Annual
     Hazardous Materials Management Conference West. Tower Conference Management
     Co., Glen Ellyn, IL (1989)

[11]  Jiang, SJ. and R.S. Houk, Inductively Coupled Plasma Mass Spectrometric Detection
     for Phosphorus and Sulfur Compounds Separated by Liquid Chromatography.
     Spectrochimica Acta, 43B. 405 (1988)
                                   H-326
-------
MOBILITY
-------
          A Proposed Waste Component Mobility Scale

A. D. Sauter, J. Downs, A. D. Sauter Consulting, 2356 Aqua Vista
     Ave., Henderson, NV 89014-3636

     A universal scalling system would be of use in  studies  where
waste component mobility in solid/liquid systems is  of  interest.
A technically accurate scaling system would be useful to, for
example, compare waste component  ( organic, inorganic and other
species ) mobility between treated and untreated wastes to
demonstrate the effect of waste fixation technologies.
Alternatively, such a scale or meteric could be useful  in
comparing mobility of waste components between various  waste to
assess one aspect of a wastes potential to contaminate  ground
water. Also, such a scale would have utility in comparing waste
component mobility under varying analysis conditions including:
extraction time; liquid phase properties and solid phase
characteristics. In this paper, we assert that where biphasic
systems are accurate representations of the system under study  or
where solid/liquid partitioning information is of import in
describing, comparing or assessing waste component mobility  (as
it virtually always is) that elementary physical chemistry and
statistics has already given us the tools to propose such a
scale. In this paper we propose a mobility scale based  on
partition coefficients for organic compounds.
     To visualize this approach, we show below a comparison  of
partition coefficients for 12 volatile organic compounds for
three wastes. It is clear that for these waste and analytes  that
K values differ by about 3 orders of magnitude from  the top  to
the bottom graph. In our paper, we discuss our mobility K
Scale(s) applied to organic wastes, inorganic species and we
consider statistical and graphical interpretation of our results.
We propose that a such a scale be adopted for the classification
of waste component mobility in solid/liquid systems. Extension  to
gas/solid/liquid systems is also discussed.
                                        12 Aneiut™
                        K V.lue
A


e
;
;
;
-








|






I







	 *"




r__
;
;
. — :
                          e.i        i         10
                             Partition Coefficient
                        K V«iue
                                        12 An«lut"
                  ......
                                                                ,...._
                          0.1        1         10
                             Partition Coefficient
                               H-327
-------
102                     CONTAMINATED SOILS LEACHING PART I
                             MOBILITY OF SOLUBLE SPECIES

      G. Hansen, USEPA, GJ. DuBose, S. Hartwell, and J. Guterriez, SAIC

      Introduction

            The development of laboratory tests that may be used to predict whether a waste is
      hazardous has been and continues to be of paramount importance to EPA and the regulated
      community.  The development of these tests is a two step process.  The first step is used to
      establish what will happen to waste when it is disposed in a given environment. That is,
      what components from the waste will migrate and what will  their concentrations be in the
      migrating media (usually water). The second step  is  to devise a laboratory test that will
      model the migration potential of the waste. The laboratory  test is  then used by EPA and
      the regulated community to determine the risk posed  by the waste.

            Contaminated soils  have become a significant environmental medium over the past
      few years.  The EPA regional labs, as well as superfund remediation  contractors, need
      methods to establish whether soil that has been  contaminated as  the result of a spill is
      hazardous; and, if they remove to treat  contaminated soil,  what post-treated toxicant
      concentration would be considered protective.  The  question being asked is "What test can
      be used to best determine whether a site is safe?" This research was undertaken to address
      this question.

      Experimental

             Four soils were collected that represented a  range  of soil types and characteristics.
      Two soils were clay type, designated alfisol and ultisol,  one soil was a very dark organic rich
      soil, designated Mollisol, and one soil  was a sandy carbonaceous soil, designated aridisol.
      The soils were tested for alkalinity, cation exchange capacity, pH, and Total organic carbon.

             The soils were then spiked with inorganic and organic contaminants at concentrations
      that would cause them to be judged hazardous using the toxicity characteristic leaching
      procedure if only 2% of the spiked amount leached from the  soil. Soils were spiked in two
      groups.  One group was spiked with metals and semivolatile organic compounds. The other
      set was  spiked with volatile organic compounds and cyanide.  Water soluble chloride salts
      were  used for spiking  for  all  metals except lead,  which was  spiked  as the  nitrate.
      Semivolatile compound were spiked in methylene chloride and benzene solutions. Cyanide
      was spike in water as  sodium cyanide.  After spiking the soils were allowed to dry in the
      hood  to remove the solvent.  Volatile compound were spike directly onto  prechilled soil.

            Metals that were spiked included Cadmium, Chromium, Copper, Lead, and Mercury.
      Semivolatiles   included phenol,  paracresol,  nitrobenzene,  pentachlorophenol,  and
      hexachlorobenzene.   Volatiles included  chlorobenzene,  1,2-dichloroethane, Methylethyl
      ketone, tetrachloroethylene, and toluene.  Cyanide was also spiked.  After spiking each soil
      was analyzed in triplicate to establish the  spiking levels that were actually found in the soil.
                                             n-328
-------
      The soils were packed into laboratory scale lysimeters constructed from pyrex glass
to a depth of 2 feet. The soil compaction was adjusted to give the same compaction that
would be expected for undisturbed soil.  Two lysimeters of each spiked soil were prepared
for each soil batch to provide an estimate of lysimeter variability.  One unspiked soil
lysimeter was also prepared to serve as a control and to provide samples for laboratory
recovery experiments.  Thus for each soil batch there were 12 lysimeters (i.e. 4 soils x 2
spiked and 4 controls). A total of 24 lysimeters were prepared in all.  The lysimeters were
3 inches internal diameter and 5 feet long.  Each lysimeter was filled with a leaching fluid
to a height of 2 feet above the top of the soil in the lysimeter. Simulated acid rain leaching
fluid (unbuffered deionized water adjusted to pH 4.2 with 60:40 H2SO4:HNO3) was used for
the metal and semivolatile soil batch and deionized water was used for the volatile and
cyanide soil batch.

      The pore volume of each lysimeter was calculated by measuring the distance the
leaching fluid level fell until liquid exited the bottom of the lysimeter.  Samples were then
collected every pore volume until the concentration of at least 5 consecutive pore volumes
did not  change.

      Spiked soils were also tested using the EP Toxicity test  (Method 1310), the Toxicity
Characteristic Leaching Procedure (Method 1311), and the simulated acid raid test (Method
1312).
                                       H-329
-------
Results and Discussion
      The results  of  the lysimeter  studies  provided the  leaching  profiles  of  each
contaminant spiked onto the soil and their leaching rates.  With lead being the only
exception  the  leaching  curves all  represented a exponential decay.  The leaching of
components was complete within the first ten pore volumes, again with lead the exception.

      All of the  leaching curves derived from this lysimeter study can be described as
gaussian curves.  The curves aU follow the general formula for a bell shaped curve:
                            y=-
              where:

             a =  standard deviation of the curve and
             x =  displacement from the top of the peak, i.e. the mean
             u =  mean.

That  is, all of the  curves follow this  equation even  though  the top of the peaks may be
shifted due to a strong interaction with the soil.  Thus, some curves seem to follow a simple
exponential decay while others show,  what appears to be, a chromatographic profile. The
a value reflects how broad the curve is. If a curve peaks at u, then u-x describes the shape
of the curve in the positive and negative directions.
                                                                              (2)
              where:
                 C, = analyte concentration at t
                 nax = analyte concentration at tr
                  a - peak standard deviation
                  t  = time displacement from peak maxima
                 tr  = peak maxima.
                                      n-330
-------
      Equation 2 is the gaussian analog for compounds eluting under chromatographic
conditions. Here, however, the mean and displacement are replaced with time variables.

      Under a chromatographic system, a mobile phase percolates over the stationary
phase. In the case of the lysimeters, the homologs are the leaching media and the soil,
respectively. An analyte introduced into the mobile phase will eventually exit the column,
with the time of elution dependent on any interactions (and therefore retardation) with the
stationary phase. Thus, a strong interaction (physical or chemical) between the analyte and
the stationary phase (soil) will produce a long retention time tr

      The leaching curves show that all significant leaching (from the elution peak shape)
takes  place within the first 10 lysimeter pore volumes.  This was much faster than anyone
had expected.  The general view is that the soils would interact with the contaminants to a
greater degree and retain them.  Furthermore, almost all of the  curves show maximum
leaching in the first pore volume. Thus, the mechanism of the leaching, particularly for the
metals, probably depends on solubility and adsorption.

      The summary results of the lysimeter experiments for the  metals are provided in
Table 1 along with the results  of the leaching tests.
                                       n-33i
-------
                 Table 1
Percent Recovery of Metals from Lysimeters
         and Batch Leaching Tests
Soil
Cadmium
Alfisol
Aridisol
Mollisol
Ultisol
Copper
Alfisol
Aridisol
Mollisol
Ultisol
Mercury
Alfisol
Aridisol
Mollisol
Ultisol
Chromium
Alfisol
Aridisol
Mollisol
Ultisol
Lead
Alfisol
Aridisol
Mollisol
Ultisol
Column 1

65.4
33.2
61.5
64.8

74.5
0.6
8.9
81.7

96.4
89.4
44.2
88.8

49.4
0.0
0.1
57.9

58.2
3.8
13.0
88.8
Column 2

96.2
38.8
68.8
87.8

110.8
0.3
10.1
104.4

118.6
58
53
109

70.7
0.0
0.2
82.7

64.7
2.4
10.7
78.4
EP

96.4
73.7
73.8
84.2

114
13.3
16.1
92.7

90.2
94.4
42.5
100

52.6
<0.9
<1.1
61.5

97
22.2
24.8
95.4
TCLP

89.2
80.1
80.7
82.3

85.6
28.7
16
70.3

81.6
73.7
66.5
92.4

57.1
5.4
2.4
59.8

80.9
37.5
44.8
82.1
Acid Rain

94.2
47.7
82.1
95

101
<1.6
13.6
110

89
76.4
67.8
99.1

47.1
0.9
1.1
62.9

97.9
0.8
24.8
100.0
                  n-332
-------
      Experimental results for the organic compounds show that the higher the water
solubility of the compound, the faster it eluted from the lysimeters.  For the semivolatile
compounds, phenol is the most soluble and it eluted the most rapidly, followed by p-cresol,
and nitrobenzene.  Hexachlorobenzene and pentachlorophenol  did not  elute from  the
lysimeters.  Similar to the metals, the elution of the semivolatile compounds ended within
the first 10 pore volumes.

      Detectable  amounts of all of the volatile compounds eluted from the lysimeters.
Recoveries of the volatile compounds from the lysimeters were generally very low. This may
be due to evaporative losses or interaction with the soil. Volatile data was only obtained
for aridisol.  The other soils interacted so strongly with the volatile constituents that leaching
fluid would not pass through the lysimeters.

      Data for the organic compounds are contained in Tables 2 and 3.

                                      Table 2
                  Recovery of Semivolatile Organics from Lysimeters
                              and Batch Leaching Tests
Soil
Phenol
Alfisol
Aridisol
Mollisol
Ultisol
Cresol
Alfisol
Aridisol
Mollisol
Ultisol
Nitrobenzene
Alfisol
Aridisol
Mollisol
Ultisol
Column 1

39.5
29.6
71.8
71.1

5.5
7.1
36.4
50.2

3.7
3.9
6.0
20.3
Column 2

49.5
58.6
65.7
70.7

10.8
43.3
31.9
32.7

7.0
6.0
10.0
9.6
EP


27.1
40.3
13.7


1.2
14.2
1.3


22.3
23.4
5.5
TCLP

10.0
27.8
23.5
13.7

2.3
1.7
0.7
1.3

3.4
16.4
35.0
0.4
Acid Rain

10.7
34.2
24.3
16.1

ND
1.7
6.3
0.8

2.3
11.0
ND
ND
                                       H-333
-------
                                      Table 3
                   Recovery of Volatile Organics from Lysimeters
                             and Batch Leaching Tests
Soil/Analyte
Aridisol
DCE
MEK
TCE
Toluene
CL-Benzene
Column 1

27.0
7.5
0.3
34.3
12.6
Column 2

32.8
3.5
0.2
30.7
11.6
EP

20.8
12.3
11.1
83.0
43.2
      The soils were leached using batch extraction methods  1310,  1311, and 1312 to
determine which, if any, of these tests best simulate the column data.  Tables 1, 2, and 3
provide the recoveries for the inorganic and organic contaminants leached respectively by
these methods.  All three extraction methods proved to  be similar in terms of percent
analyte  extracted, with few variances observed.

      All inorganic analytes  extract well, dependent upon their  solubilities.   Organic
analytes similarly follow this solubility trend.

      In general, the batch leaching of the metal analytes with synthetic acid rain showed
good agreement with that of the acid rain lysimeters. In addition, the batch leaching data
using the EP procedure resembled that of the acid rain data.  The extraction data for the
organics seems to be less comparable with that of the columns.

      Again, both analyte leaching in the lysimeter and the extraction vessel seem wholly
dependent on the solubility of a chemical species. Anomalies to this behavior are copper,
chromium, and lead on aridisol and mollisol, showing depressed extraction levels under both
methods.

      Low level leaching of  chromium, copper,  and lead in aridisol could be  due to
formation of  metal carbonates or hydroxides.  The high alkalinity of aridisol supports this
conclusion.

      One can also note a higher or equivalent extraction efficiency for the batch method
than for the lysimeter for most of the analytes.  This is to be expected as the batch method
uses a 20 to 1 liquid to solid ratio whereas the lysimeters end up with a final liquid to solid
ratio between 0.1 and 0.2 at 10 pore volumes.  The constant agitation coupled to the high
liquid volume of the batch method favors  a high level  of extraction,  especially with the
solubility dependence of the contaminants involved.
                                       n-334
-------
      In summary, the column lysimeters and the batch extractions produced approximately
the same results with regards to percent toxicant recovered. The contaminants were leached
to a similar extent due to their solubility behavior in the media. The different liquid media
slightly affected the leaching efficiency, and of the three batch tests, the acid rain procedure
gave similar results to the lysimeter, as expected. The EP method, however, also produced
analogous data.
                                       H-335
-------
103                      CONTAMINATED SOILS LEACHING PART II
                    MOBILITY OF LEAD FROM CONTAMINATED SOILS

       G. Hansen, USEPA, G.J. DuBose, H. Huppert, S. Hartwell, P. White, J. Guterriez, SAIC

       Introduction

             Recently, EPA Region I and several states have expressed concern regarding the
       disposition of lead contaminated soils.  Their  concern is  founded on  the fear that a
       significant percentage of soil in urban areas may fail the Toxicity Characteristic Leaching
       Procedure (TCLP) for lead. Specifically, they are  concerned that lead contaminated soils
       may be classified as a hazardous waste if it is moved during excavation or landscaping.  Such
       a situation, if it could be enforced, would require that homeowners and building contractors
       comply with the hazardous waste regulations.

             Because of these concerns, EPA conducted two interlaboratory studies designed to
       compare the aggressiveness of several leaching media toward lead contaminated soils.  In
       one study, six different leaching media were used on five soils taken from Region I.  The
       leaching media varied in acid strength from simulated acid rain to the acetic acid buffer
       used in the TCLP. A microwave digestion  method was also evaluated for determining the
       concentration of lead in the soils  during  this study.  The  second interlaboratory study
       compared the leachabiliry of lead from five soils  taken around Baltimore, Maryland.  This
       second study utilized the EP, TCLP, and simulated acid rain as  leaching media.

             The questions we attempted to answer  involved whether there was a significant
       difference between the EP and TCLP, the soil concentration required to exceed 5 mg/L in
       the leachate,  the  affect of diluted  acetic  buffers, and whether  there was a significant
       difference between standard digestion methods and the microwave method.

       Experimental

             Five soil samples  were   collected  by EPA  Region I  personnel  from an  area
       surrounding a house in Boston, MA believed to contaminated with lead from leaded paint.
       These samples were used in the first interlaboratory study. Five soil samples were collected
       by personnel from the State of Maryland Department of the Environment.

             Each soil was sieved to remove debris (i.e. rocks and plant material) and tumbled to
       achieve a homogeneous sample.  The soils were then aliquoted and sent to the participating
       laboratories.

             Samples were  characterized by analyzing pH (Method 9045), alkalinity, total metals
       (Methods 3050, 3051, and 6010).  The boston soils were leached  using methods 1310 (EP),
       1311  (TCLP), 1312 (Synthetic acid rain), 1311 @ 10% buffer, 1311 @ 25%, and 1311 @
       :>0% buffer. The baltimore soils  were leached via methods  1310, 1311, and 1312.

       Results and Discussion
                                            E-336
-------
      The pH of the boston soils ranged from 5.7 to 6.5 with alkalinity from 150 to 300
mg/kg. The pH of the baltimore soils ranged from 6.6 to 7.9 with alkalinity going from non
detectable to 200 mg/kg.

Boston Soil Intel-laboratory Study

      Table 1 shows a comparison of the microwave digestion method and method 3050
for lead.  The lead concentrations determined with the microwave digestion were similar to
those determined using method 3050.  The precision for these measurements were also
about the same.  Generally, the microwave digestion gave equal or higher concentrations
and equal or better precision than the conventional digestion for all metals tested.

                                     Table 1
                Comparison of Digestion Methods for Lead ± % RSD

Soil Sample        Method 3050             Method 3051 (Microwave)

      1                   3700 ± 29               4299  ± 27
      3                   2214 ± 24               1510  ±  3
      4                   2750 ±  4                5690  ±  13
      5                   874 ±  6                 798 ±  4

      The apparent discrepancies in total lead values can be attributed to small paint chips
in the soil.  Although, the chips were not visible, upon addition of acid small paint chips
would float to the surface and effervesce. This observation agrees with the assumption that
the soil was contaminated by lead from paint chips since lead carbonate was the primary
pigment used in white house paint prior to 1950.  Thus, the differences in total lead shown
by both methods may be due to different amounts of paint  chips in each sample aliquot
taken rather than to the methods themselves.

      Table 2 shows the results form the interlaboratory leaching study of the soils. The
results show that Method 1311 leached the highest concentration of lead of any of the
leaching  media.  Furthermore, the concentration of lead  leached is proportional to the
amount of acid present in the leaching media.
                                      n-337
-------
                                                             Table 2
                                                Results of the Interlaboratory Study
                                        Average Lead Concentration in the Leachate (mg/L)
Soil Number

5
4
3
1
1311

0.8
10.5
2.1
4.75
1311 50%

0.36
6.1
0.86
2.7
1311 25%

0.17
2.8
0.41
1.12
1311 10%

0.125
0.78
0.23
0.375
1312

0.21
0.91
0.35
0.63
1310

0.05
0.5
0.07
0.17
-------
      The results of the interlaboratory study of the Baltimore soils were similar to the
Boston soils.  Method 1311 was clearly more aggressive than either method 1310 or 1312.
The results of the Baltimore soil interlaboratory study are summarized in Table 3.

                                      Table 3
                         Results of the Interlaboratory Study
                            Average Lead Concentration
                       Total and Leachate (mg/kg and mg/L)
Soil
Number

1
2
3
4
5
Total

 8253
 1533
19900
 1633
 1046
Methods
1311

29.0
1.3
50.8
3.0
1.6
1312

0.56
0.38
2.70
0.12
0.10
1310

0.79
0.09
3.03
0.61
0.07
Conclusions

      This study showed that soils containing lead at concentrations greater than about
2000 mg/kg would likely leach lead in excess of 5 mg/L when using method 1311.  When
using methods 1310 or 1312, however, lead  concentration in excess of  10,000 mg/kg are
required to produce leachates greater than 5 mg/L. The leaching power of each solution
was proportional to the acid (buffer) content.

      By using this approach, one could estimate the effect of the TCLP on the disposition
of urban soils, assuming one knew the distribution of lead in the soils.  Soils adjacent to
buildings that were painted with leaded paints will have a higher likelihood of being judged
hazardous.  Since this type  of paint was in  wide use prior to  1950, property with older
structures would be more likely to have elevated soil lead concentrations.
                                       n-339
-------
104                        MOBILITY OF CONTAMINANTS FROM
                           MUNICIPAL WASTE COMBUSTION ASH

       G. Hansen, USEPA, G. Polanski, G. J. DuBose, P. White, J. Guterriez, SAIC

       Introduction

             The health and environmental hazards posed by the disposal of municipal waste
       combustion (MWC) ash are largely unknown.  Several legislative proposals have been
       introduced in both houses of Congress dealing with regulating the disposal of MWC ash.
       EPA has conducted several studies over the past few years to determine whether MWC ash
       is a  RCRA  hazardous  waste and whether the ash would require  special management
       standards for safe disposal.  While these studies  have  contributed a great deal to  our
       understanding about the placement  of MWC ash in the environment, very little is known
       about the basic mechanism of release of hazardous materials from the ash over time.

             The Office of Solid waste has conducted a study to  establish the leaching behavior
       of MWC ash over time and to compare these results with laboratory batch leaching tests.
       The time dependent leaching behavior of MWC ash is important since it influences not only
       disposal design standards, but also our basic understanding regarding the hazardousness of
       MWC ash.

       Experimental

             Three municipal waste incinerators, representing the three basic types of combustors
       (i.e.,  mass burn, modular, and refuse derived fuel), were sampled.  Samples were taken of
       mixed ash (combined fly and bottom  ash) over a four  hour period and composited to
       provide  the  ash sample for testing.  During  sampling, large particles  were removed or
       broken to provide a reasonably homogenous (i.e. particle size of 1/2 in diameter, or less)
       ash sample.  A total of about 300 Ib of ash was collected at each incinerator.

             The ash was shipped to the laboratory and placed  in a series of lysimeters.  The
       lysimeters were constructed of pyrex glass and were 4 inches in  diameter  and 5 feet long.
       Ash was placed in each lysimeter to a height of 3 feet.  The lysimeters were tapped lightly
       to allow the  ash to settle. Each column typically contained between 15 and 20  Ibs of ash.

             There were a total of 6 lysimeters for each ash or 18 lysimeters in all.  For each ash
       two lysimeters were  leached  with distilled water, two with  simulated acid rain (60:40
       H2SO4:HNO3), and two with simulated landfill leachate.  The lysimeters were maintained
       under a head pressure of approximately 2 ft of leachate  during  the leaching experiments.
       The effluent  from the  lysimeters were collected every 1/2  lysimeter pore volume.  The pore
       volumes were established by measuring the  amount of leaching fluid needed to filling the
       lysimeters just to the top of the ash.

             Samples were analyzed for total dissolved solids, specific conductivity, pH, chloride,
       sulfate, and metals (As, Ba, Cd, Cr, Cu, Ni, Pb, and Zn).  Ash samples were also subjected
                                             n-340
-------
to the EP Extraction Procedure (Method 1310), the Toxicity Characteristic Leaching
Procedure (TCLP - Method 1311), and the Simulated Acid Rain Leaching Test (Method
1312).

Results and Discussion

      The leaching curves were established for each lysimeter for each analyte. Generally,
water soluble species (e.g. chloride and sulfate) leached from the lysimeters very rapidly.
These species were essentially washed from the lysimeters within the first 10 pore volumes
(See Table 1). All of these curves exhibited an exponential decay. Several of the metals
also followed this pattern. Figures 1 and 2 demonstrate the leaching of barium and lead
from two of the ashes using simulated acid rain. Barium and lead were the most mobile
metals from the lysimeters regardless of the leaching medium.
                             Acifl  Rain  Lysimeters
                 50


                 45


                 40


                 35


                 30


                 25


                 20


                 15


                 10


                 5


                 0
                                      ¥C Astl   BarlUH
                                    7    9    11

                                        Pore Volume
                                                  13
                                                       15
                                                            17
                                                                 19
                                                                      21
      Figure 1 - Barium leaching from WC Ash with Simulated Acid Rain
                                      n-34i
-------
      It was somewhat unexpected that barium and lead would be so mobile. This may be
due to the pH.  The pH of the leachates from the lysimeters were quite high due to lime
addition to the fly ash.  The fly and bottom ash are mixed to produce the ash that was
sampled.

      In this case the pH ranged from about 11.5 to 12.5.  It is possible that these metal
species are fairly soluble in this pH range and thus are washed from the lysimeters.
                            Acid  Rain  Lysimeters
                                      FL
                                             Leafl
                10
                                            11
                                       Pore Volnue
                                                 13
                                                      15
                                                           17
                                                               19
                                                                    21
      Figure 2  Leaching of Lead from FL Ash with Simulated Acid Rain
      Appreciable leaching was observed for lead, barium, zinc, and copper for the three
ashes. As with anions, most of the leaching for the metals occurred within the first 10 pore
volumes.

      Batch leaching was performed for each ash type using the EP, TCLP, and Deionized
Water (DI).  There was reasonably good agreement between the leaching tests and the
column results when pH of the leaching media is taken into account.  For example, when
using the EP or TCLP the final pH of the leaching solution was around 6.  At this pH little
or no barium and lead leached from the ash, however, zinc leaching is increased
dramatically.
                                     H-342
-------
Table 1 - Chloride mg/L
Pore Number
0.5
1.0
1.5
2.0
2.5
3.0
3.5
4.0
4.5
5.0
5.5
DI Water
7000
4950
3040
2010
1030
530
390
270
200
180
160
Acid Rain
7900
6000
4500
2720
1880
1200
700
460
450
350
200
Synthetic Leachate
6280
4560
2870
1970
2080
1610
1160
870
650
660
380
                          H-343
-------
105                    THE WASTE INTERFACE LEACHING  TEST:   A LONG-TERM

                   STATIC LEACHING METHOD FOR SOLIDIFIED/STABILIZED WASTE

       Dr,  Danny R.  Jackson,  Senior Staff Scientist; Ms. Debra L. Bisson, Staff
       Scientist;  and Mr. Kenneth R.  Williams,  Scientist.  Radian Corporation, P.O.
       Box 201088,  Austin,  TX  78720

                                          ABSTRACT

             The purpose of a solidification/stabilization process is to reduce the
       leachability of hazardous  constituents in treated wastes to a level required
       to protect groundwater resources if the  solidified waste is placed in a
       landfill.  Many stabilization processes  include the addition of cement or
       other pozzolanic materials to the wastes to produce a product which can be
       molded into monolithic shapes.   The Waste Interface Leaching Test (WILT) was
       developed to assess the leachability of  monolithic waste forms exposed to an
       aqueous environment.

             Features of the WILT include;  (1)  large monolithic samples up to 11 Kg
       in mass, (2) a low leachant to unit surface area ratio of less than 1.5:1, and
       (3) a sequential leaching period of up to 6 months.  The WILT was used to
       evaluate solidified waste produced by the Soliditech process as part of a
       U.S.EPA SITE demonstration project.   Solidified masses were placed in plastic
       tanks with sand packing and were sequentially with distilled water on a
       biweekly basis.  A diffusion coefficient and leachability index for several
       inorganic constituents were determined at the end of the 6-month leaching
       period.  Results from the WILT corroborated results from other leaching tests
       in demonstrating the effectiveness of the solidification process.

             The WILT was shown to be a leaching method that can provide long-term
       data on large monolithic forms to evaluate the efficacy of
       solidification/stabilization processes.
                                            H-344
-------
106               ALTERNATIVE METHODS FOR ESTIMATING LEACHING OF INORGANIC
                        CONSTITUENTS FROM COAL-COMBUSTION RESIDUES

         Ishwar P. Murarka, Senior Program Manager, Electric Power Research
         Institute, 3412 Hillview Avenue, Palo Alto, California 94304
         Calvin C. Ainsworth, Research Scientist, Earth and Environmental
         Sciences Center, Dhanpat Rai, Senior Staff Scientist, Earth and
         Environmental Sciences Center, Battelle, Pacific Northwest Laboratories,
         Battelle Boulevard, Richland, Washington 99352

         ABSTRACT

         In  this  paper,  we compare data on  leachates  obtained by the extraction
         procedure (EP), by the toxicity characteristic leaching procedure  (TCLP),
         in  field leachate samples,  and in  pore waters  for five coal-combustion
         disposal  sites.   The results for  the  EP  and the  TCLP  extracts and the
         concentrations in the other leachates are also compared with predictions
         made using the reaction-based FOWL™ computer code.  With the exception of
         Ba, whose concentration  is  typically higher in  the EP and TCLP extracts
         than  that observed in disposal  site pore waters  or  drainage leachate,
         leachate  concentrations  obtained by the EP and  TCLP methods are about  a
         factor of 10 lower than those observed in the  disposal site leachates.
         The EP  and TCLP methods  appear  to dilute the effect of highly soluble
         salts because of the 20:1 solution-to-solid ratio used and to solubilize
         large amounts of Ca from calcareous  samples because of the low pH of the
         extracts.  Concentrations for Ca, Ba, Sr,  and SO^ predicted by the FOWL™
         code were closer  to  the  concentrations  in the pore waters than were the
         concentrations in the EP and TCLP extracts, in several cases to within  a
         factor of 2.   These  comparisons  and other experiments have provided the
         additional information needed to improve the geochemistry incorporated in
         the FOWL™ code.  The results presented show that planned modifications to
         the FOWL™ code will improve its accuracy for predicting concentrations of
         several  of the  elements  discussed  in this paper and of several elements
         that are described only empirically  in the current version of FOWL™.

         INTRODUCTION

         Coal-combustion residues include fly ash, bottom ash,  and scrubber sludge.
         These solid wastes are disposed  of on  land in ponds and landfills.  The
         1980 Bevill Amendment to  the Resource Conservation and Recovery Act  (RCRA)
         exempted  these   high-volume   solid wastes  from  the  hazardous-waste
         provisions  of the law.   Therefore, the  use of  the  U.S.  Environmental
         Protection  Agency's  (EPA's)  extraction  procedure  (EP)  or  toxicity
         characteristic leaching procedure  (TCLP) is not required for determining
         whether  or not  these wastes  are  to  be  considered hazardous wastes.
         Nonetheless,  a  number of samples  of fly  ash, bottom ash,  and scrubber
         sludge have been analyzed for leachate composition using the EP,  the TCLP,
         and a water-extraction  procedure,  and  by  field sampling  of disposal
                                           H-345
-------
facilities.  In addition, the Electric Power Research Institute (EPRI) has
developed  a  reaction-based   computer  code,   FOWL™,   for  predicting
compositions and quantities of leachates generated at disposal sites.  The
current discussions on RCRA reauthorization are  likely to result in the
development and adoption of new methods  and approaches to determine the
leaching  characteristics  of nonhazardous  wastes.    This paper  is our
contribution to the scientific  debate.

The Electric Power Research Institute  has sponsored  a variety  of research
projects to develop the scientific understanding necessary for predicting
leaching  characteristics  of  coal-combustion  wastes  disposed  of   in
landfills and  ponds.   Both laboratory  and field studies have been carried
out to define  the leaching chemistry of inorganic constituents contained
in fly ashes,  bottom ashes, and scrubber sludges  (Ainsworth and Rai 1987;
Rai et al. 1987,   1989; Fruchter et al. 1988;  Mattigod et al.  1990; Eary
et al. 1990).   Extensive  laboratory and field  studies of numerous samples
have  resulted  in  identification  of  fundamental  chemical  reactions
responsible for   leaching  of several elements  in  the  coal-combustion
wastes.   As a result,  the reaction-based model  FOWL™  was developed and
released  for  general use in 1988  (Hostetler et al.  1988).   Additional
research, including sampling and analysis  of  waste  and leachates from a
variety of sites  where coal-combustion wastes have been disposed of, has
resulted  in  improvements  to  the  FOWL™   code.    The  enhanced  FOWL™
(Version 2.0)  will be released in late 1990.

In this  paper, we focus on  leachate  studies conducted with  five waste
samples  from  actual waste-disposal sites.   The  aqueous  concentrations
observed  in 1) pore waters extracted from the  moist field  samples  by
immiscible  displacement  (Kinniburgh  and Miles  1983) ,  2) leachates
collected from the field disposal sites, 3)  extracts obtained by the EP
and  TCLP tests,  and  4)  calculations made  with  the  FOWL™  code  were
compared.   It was  found that the  FOWL™-calculated  concentrations  for
several elements   were closer to  those observed in  the  pore  waters than
were either the leachates or  the  EP or TCLP extracts.

MATERIALS AND  METHODS

MATERIALS

Five samples of combustion wastes were collected from disposal areas at
five power  plants in Pennsylvania.   The  wastes  consisted of  fly ash,
bottom ash,  and scrubber sludge removed from  active disposal areas that
ranged in age  from 6 months  to 4 years.   The materials  were typically
collected by mixing several samples taken from depths of 3 to  7 feet from
two areas  in  each  landfill.   A brief  description  of  the  samples  is
presented in Table 1.
                                 H-346
-------
METHODS

The  pH  of  each  material  after  1 hour  was   determined  from  a  1:2
solid-to-solution  (mass/vol)  paste  using  deionized  water.    The total
chemical composition of the dried samples was determined by proton-induced
X-ray emission spectroscopy (PIXE).

The leachates for chemical analyses were generated by several  methods  (the
EP and  TCLP;  extraction of pore  water from the  moist field samples by
immiscible displacement with heavy liquid; and collection of  leachates in
the  field) .    The methods  to  obtain EP and TCLP extracts  have  been
described elsewhere  (EPA 1982  and 1986,  respectively).  Pore water from
the moist samples was extracted under a N2 atmosphere by displacement with
Freon®^1) (Kinniburgh and Miles 1983).  The field leachates were collected
and  analyzed by participating utilities.    It  should  be  noted that,
unfortunately, the pathways of some of the field  leachates  (corresponding
to samples 703 and 705)  intercepted springs and other offsite water and
that, therefore,  these  leachates  do not accurately represent what would
be leached from  coal-combustion wastes alone.

The  composition  of  the  extracts  was determined  by inductively  coupled
plasma (ICP) spectroscopy,  and anions were analyzed by ion chromatography
or ion-specific  electrode.

Geochemical Interpretations and Predictions

To determine  the aqueous speciation and  types of solid phases that may
control the aqueous concentrations of selected elements,  the  compositions
of pore-water extracts were modeled using the  geochemical  code MINTEQ
(Felmy et al. 1984).   In addition,  the chemical compositions  of the solid
samples,  along  with  pore-water pH values, were  used as  the  input for
calculating  the   leachate  compositions with  the  FOWL™  code  (Hostetler
et al. 1988).

RESULTS AND DISCUSSION

Several methods  (EP, TCLP,  FOWL™ calculations)  can be used to estimate
aqueous concentrations of different elements.  These estimates in turn can
be used to estimate the potential  impacts  of disposing of coal-combustion
wastes  on land  surfaces.   To check  the  validity  of  these  methods for
estimating  leachate   composition,  their  results  must be  compared with
leachates collected from actual disposal sites.   We obtained  actual field
leachates by two  methods:  1) extracting pore water from samples collected
at disposal sites  and 2) collecting the drainage leachate from the  same
disposal  sites.     If   the   sampling  site  is   chemically  relatively
homogeneous,  as  most coal-combustion disposal sites are, then the field
leachates and pore waters  should  have similar concentrations.  However,
a  comparison  of  the  pH,  which  generally  influences  the  leachate
                                  H-347
-------
composition  through  its  effect on  both  precipitation  and adsorption
reactions for most elements,  showed that for two of the samples  (samples
703 and 705) , the pH values were considerably different (Table 1) .  Such
a difference  indicates  that these field leachates may not  truly represent
the leachates that would  be  generated solely from  the  wastes that were
sampled.  Correspondingly, the chemical analysis  of field leachates for
sample 705,  for example,  showed concentrations of Al,  Fe, and Zn  of 123,
756, and 3.28 mg/L,  respectively,  and a pH of 3.31,  in contrast to the
concentrations  of these  elements  in  pore  waters  (0.29,  <0.01,  and
>0.05 mg/L)  and the pH  of  6.70.  Therefore,  the comparisons in this paper
will be based on the pore-water compositions.

In the EP and TCLP, the pH of the waste/water suspensions generally ends
up at about 5, and these methods use 1:20 solid-to-solution ratios.  Under
such conditions, 1) those  elements  that are  present  in the wastes  as very
soluble  salts  will be extracted  and proportionately  diluted,  2) those
elements  that  are  present   in   fairly   insoluble  solids   and  whose
solubilities  are  pH-dependent  will   occur  in  concentrations that  are
different  from those  that would  be found in  equilibrium  with field
leachates with different pH, 3) anions whose  concentrations are controlled
by adsorption will occur in reduced concentrations because of the acidic
pH  of  the  extractants,   and 4)  those  elements   that   are  solubility
controlled but have one  ion  that is  affected by dilution will  show
enhanced concentrations for the other  ion.   In addition, the pH values of
the EP and TCLP are artificially low and do not represent field conditions
associated with disposal of coal-combustion wastes.   Although acidic fly
ash  is  sometimes  produced, recent research would indicate  that the pH
increases to pH 7  (for example, see Roy et  al.  1984).   In addition,  the
pH  of  scrubber  sludge  and calcareous fly ashes would be  expected to be
closer  to 8.3, as dictated by CaCOj equilibrium with COo  in the air.   In
view  of these conditions, therefore,  the  EP and  TCLP results  may  not
reflect  the  concentrations actually to be expected  at  a given site.   As
a  matter of fact,  the Ca concentrations  observed in  the  EP and  TCLP
extracts of scrubber  sludge samples (sample  703, Table 2)  are higher than
those  in the  pore waters by about  a  factor  of  3,  because of CaCO-j
dissolution as a result of the  low pH levels  reached  in these extracts.
An  additional  problem  with these  extraction  tests  is  that  the  results
cannot  be  used to predict changing  leachate  concentrations  with time.
When EP and TCLP results are compared  for other selected elements  that are
present  in  measurable  concentrations and  important  in coal-combustion
wastes,  the  concentrations are about a factor of  10  lower  than those
observed in pore waters (except for Ba,  whose  concentrations are  about a
factor of 10 higher).  For many of the other trace elements,  such as Pb,
Cr, and  Cd,  no  valid comparisons  could be  made, because  these elements
were present in the  EP, TCLP,  and pore waters at or  near the detection
limits.
                                 H-348
-------
A fundamental approach that relates the aqueous  concentrations of elements
to specific  chemical  reactions that  may  occur in  different  wastes and
disposal environments  is currently being developed.  Under this approach,
information    regarding    the    precipitation/dissolution    and
adsorption/desorption reactions that  may  occur in the wastes is needed.
If it can be  shown for a given element that the  concentrations are limited
by  precipitation/dissolution  reactions  and  if   the  identity  of  the
solubility-limiting solid is known, then the concentrations of the element
are predictable from the thermochemical data for the appropriate solid and
aqueous  species.   Therefore,  as  a first  approximation  in  developing a
fundamental    approach,    emphasis    has    been    placed   on    the
precipitation/dissolution reactions.   Based on initial  studies  using a
large  number   of   coal-combustion  residues,   it   appeared   that  the
concentrations of several elements, such as Al, Ba, Ca, Sr,  SO^,  and Mo,
in  laboratory studies  involving  samples   obtained from dry  collection
systems  and  a dry  fly  ash  disposal  site  were  controlled  by  their
respective solid phases [AlOHSO^,  AlCOH)^, sulfate solids of Ca,  Ba, and
Sr,  and  CaMoO^].   This  solubility  information,  along with  empirical
observations of leachate quality for several other  elements (i.e., As, Cd,
Cu,  Mg,  K, Na,  Ni,  Se,  and  Zn), was  used to develop  the  FOWL™ code
(Hostetler et al.  1988)  for predicting aqueous  concentrations based on pH
and  the  total  chemical  composition of the wastes.   When the analytical
concentrations  in  pore waters  are modeled  using an  equilibrium  model
(MINTEQ;  Felmy et al.  1984) and the  resulting  activities of Ca,  Ba, and
Sr are plotted as a function of SO^ activity (Figure 1),  the Ca,  Sr, and
Ba  concentrations  in  the extracts  appear  to be  controlled by  their
respective sulfate  solids.   (In  the  case  of Ba,  it is  controlled by a
solid that is slightly more soluble than barite, as observed  in many other
samples in studies we have conducted.)  Given that  the FOWL™  code uses the
same solids as the basis of its calculations,  the FOWL™-calculated results
would be  expected to be similar to those obtained in the pore waters.

The FOWL™-predicted leachate concentrations of several elements associated
with the wastes are  given in Table 2.  The FOWL™-predicted concentrations
of  As  shown  in Table 2  are  typical  of  those of several  other  trace
elements  (e.g., Cd,  Cr,  Cu, Fe, Pb, and Zn) determined  empirically.  That
is,  the  pore-water concentrations of  these  elements are below  their
respective detection  limits,  and  the FOWL™-predicted values  also fall
below the detection limits of the analytical techniques used.

Although  the  FOWL™-predicted  concentrations for most  elements  that are
above detection limits are within a  factor of  3,  the model overpredicts
Ba and underpredicts SO^,  Ca, Mg,  Mo,  Na,  Ni, and  Sr.   Therefore, further
improvements are being carried out.  These  improvements include 1) use of
an equilibrium geochemical code,  instead of looking values up in tables,
2) better  estimation  of SO^  concentrations  by incorporating electrical
conductance and its relationship to sulfate, which would in turn improve
predictions   of   Ca,   Ba,  and   Sr,   3)  more  reliable  and  verified
                                  H-349
-------
thermochemical  data,   4)  incorporation  of  thermochemical   data  for
additional elements  (e.g.,  Cu,  Zn, and  Cr)  that have been  found to be
limited by solubility since the development of FOWL™, and 5) incorporation
of mechanistically determined masses  of  particular solid phases instead
of  empirically  estimated  leachable  fractions.     Several  of  these
improvements  have  already  been  incorporated  into  FOWL™  Version 2.0,
resulting  in  significantly   better  predictive  capabilities.     The
concentrations of Ca, S04, and Sr predicted by FOWL™ Version 2.0 (Table 2)
differ  in most  cases  from concentrations  in  actual  pore  waters  by
significantly less  than a factor  of 2.  Barium predictions have also been
improved significantly, although they are so far still unsatisfactory.

CONCLUSIONS

The EP and TCLP extracts  of  coal-combustion wastes  do not, in most cases,
reflect the  leachate composition found at disposal sites.  Often this is
a  result  of  either  dilution  or  adjustment  of  pH to artificially  low
levels.  Typically,  the EP and TCLP extracts underestimate concentrations
by as  much  as an  order of magnitude.   In contrast,  concentrations  of
selected  elements  predicted by  the   computer  code FOWL™  (Version 2.0)
differ from concentrations  observed in pore  waters from  coal-combustion
waste disposal sites by less than a factor of 2.  Further improvements to
FOWL™  based  on  thermochemical  data  collected from  sound  laboratory
experimentation and tested using  data  collected in  the field will provide
an even better predictive capability,  for an even larger suite of elements
associated with coal-combustion waste leachate.
                                 H-350
-------
                    Waste-Material Location,  Type, Age,  and pH

Sample                                     	EH	
Number      Material(a)      Age  (yr)      Pore Water      Field Leachate

                                 4              7.3               7.2
                                 2              6.8               7.2
                                 0.5            10.1               7.4
                                 0.5             6.5               7.0
                              1.5 -  2           6.7               3.3

(a)  FA = fly ash; BA = bottom  ash;  SS =  scrubber sludge.
701
702
703
704
705
FA
FA/BA
SS/FA
FA/BA
FA
                                  H-351
-------
TABLE  2.
Concentrations  of Selected  Elements  Extracted  from  Coal-
Combustion Wastes by EP and TCLP  and in Pore Waters and Field  Leachates
Associated with Actual Disposal Sites, and Values Predicted by  FOWL™
        Method
              Concentrations (mg/L) in Different  Samples
             701       702        703        704        705
As
  EP
  TCLP
  Pore Water
  Field Leachate
  FOWL™
             0.085
             0.20
            <0.16
            <0.01
             0.1
 0.011
 0.047
<0.16
<0.01
 0.01
 0.032
 0.038
<0.16
<0.01
 0.01
 0.24
 0.26
<0.16
<0.01
 0.1
 0.071
 0.20
<0.16
<0.01
 0.1
  EP
  TCLP
  Pore Water
  Field  Leachate
  FOWL™
             0.45
             1.7
            10.4
            21.6
 0.29
 1.7
 3.45
 3.3
10.7
 7.1
 0.19
 1.3
 1.11
<0.6
77.7
38.1
 0.45
<0.6
23.3
 1.6
Ba
  EP                      0.64      0.64
  TCLP                    0.58      0.44
  Pore Water              0.06      0.06
  Field  Leachate         <0.1      <0.1
  FOWL™                   0.25      0.25
  FOWL™  (Version 2.0)     0.012     0.012

Ca
  EP                     69        80
  TCLP                   77        83
  Pore Water            587       597
  Field  Leachate        410       339
  FOWL™                 394       394
  FOWL™  (Version 2.0)   611       611
                                   0.73
                                   0.53
                                   0.06
                                   0.27
                                   0.012
                               1,880
                               1,630
                                 642
                                 452
                                 408
                                 611
                        0.68
                        0.33
                        0.08
                        0.25
                        0.010
                       70
                      139
                      458
                      467
                      394
                      444
                      0.17
                      0.22
                      0.07
                      0.1
                      0.25
                      0.011
                     97
                     97
                   473
                   323
                   394
                   468
(a) FOWL™  (Version 2.0) is currently under development and will  be
    available in late 1990.  FOWL™ (Version 2.0) will differ  from
    Version 1.0 in that it will contain 1) a code for calculating
    geochemical equilibrium instead of looking it up in  tables   2)  an
    improved thermochemical database, 3) an improved database for
    empirical elements that is based on field and laboratory
    experiments, and 4) an improved ability for application to sluiced
    sites.
                                 H-352
-------
                                 (contd)

                          Concentrations  (mg/L)  in  Different  Samples
_ Method _    701       702       703        704         705

Sr
  EP                        1.10      2.04      3.96        1.22      1.19
  TCLP                      1.02      2.67      3.89        1.48      1.54
  Pore  Water              12.40    16.80      4.25       10.50      7.40
  Field Leachate           4.97      4.10      3.67        8.73      6.63
  FOWL™                    1.62      1.62      1.72        1.62      1.62
  FOWL™ (Version 2.0)      12.2     12.2     12.2         8.9        9.4
  EP                     160       190     2,400         190       260
  TCLP                   170       190     1,800         210       250
  Pore Water           1,850     1,748     1,284       4,488     4,333
  Field Leachate       1,800     2,140     1,230       3,940     4,260
  FOWL™                  945       945       945         945       945
  FOWL™ (Version 2.0)   1,478     1,478     1,478       3,793     2,914
                                  H-353
-------
ACKNOWLEDGMENTS

This research was funded by the Electric Power Research Institute,  Inc.
(EPRI), under contract RP 2485-08 titled "Leaching Chemistry Studies."
We thank Larry Holcolme of Radian Corporation for conducting the  EP and
TCLP leaching tests.  We also wish to thank the individuals from  the
power plants who participated in this study by collecting onsite
leachate and solid waste samples.

FOOTNOTES

(1)  Freon® is a trademark of E. I. Dupont de Nemours & Co. , Wilmington,
Delaware  19898.

REFERENCES

Ainsworth, C. C., and Dhanpat Rai.  1987.  Chemical Characterization of
Fossil Fuel Combustion Wastes.  Palo Alto, California:  Electric  Power
Research Institute.  EPRI EA-5321.

Eary,  L. E. , Dhanpat Rai, S. V. Mattigod, and C. C. Ainsworth.  1990.
"Geochemical Factors Controlling the Mobilization of Inorganic
Constituents from Fossil Fuel Combustion Residues.  II.  Review of  the
Minor  Elements."  J. Environ. Qual. 19:202-214.

Felmy, A. R. , D. C. Girvin, and E. A. Jenne.  1984.  MINTEQ- A Computer
Program for Calculating Aqueous Geochemical Equilibria.  Springfield,
Virginia:  National Technical Information Service.  EPA-600/3-84-032
(PB  84-157148).

Fruchter, J. S., Dhanpat Rai, J. M. Zachara, and R. L. Schmidt.   1988.
Leachate Chemistry  at the Montour Fly Ash Test Cell.  Palo Alto,
California:  Electric Power Research Institute.  EPRI EA-5922.

Hostetler, C. J., R. L. Erickson, and Dhanpat Rai.  1988.  User's Manual
for  the Fossil  Fuel Combustion Waste Leaching (FOWL™) Code.  Palo Alto,
California:  Electric Power Research Institute.  EPRI EA-5742-CCM.

Kinniburgh, D.  G.,  and D. L. Miles.  1983.  "Extraction and Chemical
Analysis of Interstitial Water from Soils and Rocks."  Environ. Sci.
Technol.  17:362-368.

Mattigod, S. V., Dhanpat Rai, L. E. Eary, and C. C. Ainsworth.  1990.
"Geochemical Factors Controlling the Mobilization of Inorganic
Constitutents from  Fossil Fuel Combustion Residues.  I.  Review of  the
Major  Elements."  J. Environ. Qual. 19:188-201.
                                   10
                                  H-354
-------
Rai, Dhanpat, J. M. Zachara, D. A. Moore, K. M. McFadden, and
C. T. Resch.  1989.  Field Investigation of a Flue Gas Desulfurization
(FCD) Sludge Disposal Site.  Palo Alto, California:  Electric Power
Research Institute.  EPRI EA-5923.

Rai, Dhanpat, C. C. Ainsworth, L. E. Eary, S. V. Mattigod, and
D. R. Jackson.  1987.  Inorganic and Organic Constituents in Fossil Fuel
Combustion Residues:  A Critical Review.  Palo Alto, California:
Electric Power Research Institute.  EPRI EA-5176, Vol. 1.

Roy, W. R.,  R. A. Griffin, D. R. Dickerson, and R. M. Schuller.  1984.
"Illinois Basin Coal Fly Ashes.  1.  Chemical Characterization and
Solubility."  Environ. Sci. Technol. 18:734-739.

U.S. Environmental Protection Agency (EPA).  1982.  Test Methods for
Evaluating Solid Wastes:  Physical/Chemical Methods.  Washington, D.C.:
U.S. Environmental Protection Agency.   SW-846.

U.S. Environmental Protection Agency (EPA).  1986.  "Hazardous Waste
Management I  - Toxicity Characteristic  Leaching Procedure."  Federal
Register 51(02-1766), January 14, 1986.
                                    11
                                  H-355
-------
FIGURE 1.   Activities of Ca,  Sr,  and Ba in Pore Waters as a Function of
Sulfate in Different Samples.   Solid lines represent ion activities
calculated to be in equilibrium with the solids identified in the
figure.
                                 H-356
-------
   -4.5-
-
   -6.5-
CO
CO

D)  -7 j
o -/.
   -8.5



   -3.5
   -4.5
CO

 0-5.5
   -6.5
AA   A
                 BaS04(c)
                             A    A
                                    SrS04 (c)
   -2.0-
 o
 E
O

 O)
 o
   -3.0 -
   -4.0
       -2.3
                                CaSO4-2H2O(c)
-2.2          -2.1         -2.0


        log SOf (mol/L)
                                            -1.9
                          n-357
-------
AIR/GROUNDWATER
-------
107    DETERMINATION OF TARGET ORGANIC? IN AIR USING ION TRAP MASS SPECTROMETRY


                  Marcus B. Wise, Ralph H. Ilgner,  Michelle V. Buchanan, and Michael R. Guerin
                           Analytical Chemistry Division, Oak Ridge National Laboratory,
                                    P.O. Box 2008,  Oak Ridge, TN  37831-6120

               As a result  of their  simplicity,  versatility, sensitivity, and ease of operation, ion  trap mass

       spectrometers are emerging as a  potentially  important new technology  for environmental monitoring

       applications.   In particular,  the tolerance of  these mass spectrometers  toward relatively high  operating


       pressures places fewer restraints on the interfacing of these devices with a variety of sample introduction

       systems.  For example, we have previously demonstrated the  ability to rapidly detect and quantify trace

       volatile organics in water, soil slurries, oil, and  other matrices by purging a sample directly into an ion trap

       through an open/split capillary interface.  Detection limits of  1 ppb or  less are possible with  no sample

       preconcentration and with splitting 90% or more of the sample to a vent.  As an extension of this work,


       we have been investigating the use of ion traps for the determination of trace organics in ambient air.

               The equipment used for this research consists of a Firmigan ITMS ion trap mass spectrometer which

       is equipped with a specially designed thermal desorption device as well as a direct air sampling probe.  The

       ITMS is equipped with the hardware and  software required  for chemical ionization,  selective ion storage,

       and collision induced dissociation (CID) tandem mass spectrometry (MS/MS).  Volatile organics  in air can

       be detected in real time at levels of  10-100 ppb using the direct sampling probe. Trace analysis is performed


       by preconcentration on resin  traps  followed  by rapid thermal desorption  into  the  ITMS through an


       open/split interface.   Using preconcentration and thermal desorption, low  pptr levels of semivolatile


       compounds  have been  successfully determined.   Methods  have  been developed and  evaluated  for  the


       determination of nicotine in environmental tobacco smoke, organophosphonate compounds in ambient air,


       and certain chemical warfare  agents using direct thermal desorption into the ITMS. The turn around time


       between samples is typically 2-3 minutes (not counting the collection of a sample on a resin trap). For real-


       time measurements, the  direct air sampling  probe is  being tested for its applicability  to monitoring


       constituents in environmental tobacco smoke, process streams, and headspace of various wastes.
       *Research  sponsored jointly  by the  U.S.  Army  Toxic and  Hazardous Materials Agency,  Interagency

       Agreement 1769-A073-A1, and the National Cancer Institute, Interagency Agreement 0485-0485-A1, under

       Martin Marietta Energy Systems, Inc., contract DE-AC05-84OR21400 with the U.S. Department of Energy.


           The  submitted  manuscript  has  been
           authored by  a  contractor of the U.S.
           Government  under contract  No. D£-
           AC05-84OR21400. Accordingly, the U.S.
           Government  retains  a  nonexclusive,
           royalty-free license to publish or reproduce
           the published  form of this contribution, or
           allow others to do so, for U.S.  Government
           purposes."
-------
108             Ion Chromatography for the Detection of Formic Acid



                in Incinerator Emissions and Ash from the Use of



                            Formic Acid as a POHC
                     Dr. Stan R. Spur1in, Ms. Patti M. Aim



                                Ms. Leigh Labor



                          Midwest Research Institute



                            Kansas City, MO 64110
     Formic acid has  many  advantages  as  a Principle Hazardous Organic



     Constituent (POHC) for hazardous waste incinerator trial burns. The



     compound is water soluble, available in bulk, inexpensive, and is



     high on the incinerability list.  One of the major drawbacks in the



     use of formic  acid  has  been  the  lack of analytical methodologies



     suitable for determining formic acid at levels sufficient to verify



     destruction efficiency-








     Ion chromatography has been previously applied to the  detection of



     the formate  ion in impinger  solutions  generated by  incineration



     trial burns.  However, limited success was obtained with limits of



     detection too high for verification  of destruction efficiency-  We



     have  modified  and  expanded   the  ion chromatographic method  for



     formate  analysis in  impingers  as well  as in  ash extracts  from



     incinerator trial burns.  Method  limits of detection are 30ppb for
                                    n-360
-------
impinger  solutions  and  lOOOppb  for  ash  samples.    A  caustic



extraction of the ash produced recoveries of better than 90% from



spiked ash samples.  Preliminary studies of the M5 train indicate



that no  carryover of  the  formic  acid  (formate)  is  experienced



between the first and second  caustic  impinger if  0.1N hydroxide is



used as  the  scrubber solution.   Over  75% of  the  fromic acid



(formate) is found in  the  condensate impingers in the trial burn



in which  formic  acid was an aqueous and solid feed POHC.    Data



will be presented  on this  revised  method as well as a discussion



of the application of  the  analytical methodology under a variety



of scenarios.
                               n-361
-------
                                           -1-

109                       Ambient  Air Monitoring for Benzene and
                             Ethylene Oxide at Texaco Conroe
                              Chemical Plant, Conroe, Texas


          P. Kittikul, Texaco Research and Development Department, Port Arthur
          Research Laboratories, P.O. Box 1608, Port Arthur, Texas 77641.
          ABSTRACT
           During  the month of April 1989, Texaco Conroe Chemical  Plant,  Conroe,
           Texas,  collected approximately three hundred samples of benzene  and
           ethylene  oxide  in adsorption tubes at the upwind fenceline,  the
           downwind  fenceline, the benzene processing unit, the ethylene  oxide
           processing unit, and outside the fenceline in a near-by commercial
           area.   All of the samples in this study were duplicated and  analyzed
           at  the  Southwest Research Institute Laboratories, San Antonio,
           Texas.  Analytical results show that the highest measured benzene and
           ethylene  oxide  concentrations were found at the process unit.  These
           measured  values of 246.6 and 429.4 micrograms per cubic meter  for
           benzene and  ethylene oxide are well below the OSHA 8-hour
           occupational Threshold Limit Value (TLV) of 3,187 and 1,798
           micrograms per  cubic meter, respectively.
           INTRODUCTION
           Toxic  chemicals  released from chemical plants  in  the United  States
           have become  an issue of national concern.  These  concerns  include
           both health  effects and the threat of continued degradation  of
           environmental quality^   .  The Superfund Amendments and
           Reauthorization  Act (SARA) was signed into law by President  Ronald
           Reagan on  October  17, 1986.  Section 313 of the SARA Title III
           requires all facilities that store, manufacture,  or process  hazardous
           chemicals  to report the type and quantity of toxic chemicals their
           operations release to the air, land, and water.   Title III is also
           known  as the Emergency Planning and Community  Right-to Know Act of
           1986.   It  is a free-standing statute to provide the public with
           information  about  release of toxic chemicals that results  from  the
                                           n-362
-------
                                 -2-

operations of facilities in their community.  The results of ambient
air monitoring for benzene and ethylene oxide from the Texaco Conroe
Chemical Plant can be used to address public exposure issues that may
be raised by Section 313 reporting under SARA Title III regulations.
MATERIAL AND METHODS
During the month of April 1989, Texaco Conroe Chemical Plant, Conroe,
Texas, collected approximately three hundred samples of benzene and
ethylene oxide in adsorption tubes at the upwind fenceline, the
downwind fenceline, the benzene processing unit, the ethylene oxide
processing unit, and the outside fenceline (Conservatex Company).
During collection of the samples, the following meteorological
measurements were recorded:
    (1) Wind speed
    (2) Wind direction
    (3) Relative humidity
    (4) Temperatures at 3 foot and 75 foot elevations above grade
Figures 1 and 2 show the ambient air monitoring sites for benzene and
ethylene oxide, respectively.  Figures 1 and 2 show that on days 1
and 2, points A, C, E, F, and FS-20 represented the monitoring sites
at the downwind fenceline, the upwind fenceline, the outside
fenceline, at pump number 8, and at the process unit, respectively.
Figures 1 and 2 also show that on days 3, 4, 5, 6, 7, and 8, points
B, D, E, F, and FS-20 represented the monitoring sites at the upwind
fenceline, the downwind fenceline, the outside fenceline, at pump
number 8, and at the process unit, respectively.
In this study, the Occupational.Safety and Health Administration
(OSHA) Method 50^ 'and Method 7^ ' were used to sample and
analyze for ethylene oxide and benzene, respectively.  All of the
on-site samples in this study were duplicated and analyzed at the
Southwest Research Institute Laboratories, San Antonio, Texas.  The
quality control protocol outlined in OSHA Method 50 and Method 7 were
used to determine the average desorption efficiency for the samples.
                                H-363
-------
                               -3-

RESULTS AND DISCUSSIONS
Table I shows the benzene concentrations measured during the
eight day ambient air monitoring program for five sites.  The
five sites are:  (1) downwind at fencellne, (2) upwind at
fenceline, (3) process unit, (4) outside fenceline (Conservatex
Company), and (5) pump number 8.  Pump number 8 is located
between the process unit and the downwind fenceline.
Table II shows the maximum and minimum concentrations of benzene
measured for each of the five sites during the eight day
monitoring program.  Table II also shows that the highest
measured value of benzene concentration was found at the process
unit.  This measured value of 246.6 micrograms per cubic meter
for benzene is below the OSHA 8-hour occupational Threshold Limit
Value (TLV) of 3,187 micrograms per cubic meter.  The OSHA 8-hour
occupational TLV is a time weighted average concentration limit,
based on eight hour workday and 40 hours per week, to which
workers can be repeatedly exposed without adverse effect.
Furthermore, Table II also shows that the maximum concentration
values of benzene at the downwind fenceline, the upwind
fenceline, the outside fenceline, and pump number 8 are 45.64,
60.75, 51.66, and 53.18 micrograms per cubic meter,
respectively.  The measured values of benzene concentrations for
these sites are below the OSHA 8-hour occupational TLV of 3,187
micrograms per cubic meter.  In this study, the Texas Air Control
Board (TACB) Effects Screening Level (ESCL) and the monitored
concentrations are not directly comparable due to differences
between the concentration averaging times used in the TACB ESCLs
(30-minute and annual) and the sample collection period (3.0-11.5
hours).
Table III shows the ethylene oxide concentrations measured during
the eight day ambient air monitoring program for the the same
five sites as described above.  Table IV shows the maximum
concentrations of ethylene oxide measured for each of the five
sites during the eight day monitoring program.  The highest
measured value of ethylene oxide concentration was found at the
process unit.  This measured value of 429.4 micrograms per cubic
meter for ethylene oxide is below the OSHA 8-hour occupational
TLV of 1,798 micrograms per cubic meter.  Furthermore, Table IV
also shows that the maximum concentration values of ethylene
oxide at the downwind fenceline, the upwind fencellne, the
                                 E-364
-------
                               -4-
outside fenceline, and pump number 8 are 45.67, <26.0, <26.00,
and 77.6 micrograms per cubic meter, respectively.  The measured
values of ethylene oxide concentrations for these sites are below
the OSHA 8-hour occupational TLV of 1,798 micrograms per cubic
meter.
In this study, all of the desorption efficiency samples were
analyzed at the Southwest Research Institute Laboratories, San
Antonio, Texas.  Table V shows the average percent desorption
efficiency values of benzene and ethylene oxide are 75% and 76%,
respectively.
CONCLUSIONS
The results of benzene and ethylene oxide concentrations at the
upwind fenceline, the downwind fenceline, the processing unit,
the outside fenceline, and pump number 8 are below the OSHA
8-hour occupational TLV of 3,187 and 1,798 micrograms per cubic
meter, respectively.  The advantages of conducting ambient air
monitoring in this study are listed as follows:  (1) measured
ambient air quality directly, (2) was more accurate than air
dispersion modeling, (3) used the results to address public
exposure limits, and (4) reduced community concern of toxic air
emissions.
REFERENCES
1.  C. B. Doty and C. C. Travis, "The Superfund Remedial
    Action Decision Prosess:  A Review of Fifty Records
    Decision," JAPCA 39:1535(1989).

2.  K. J. Cummins, "OSHA Method No. 50, Ethylene Oxide,"
    January, 1985, OSHA Analytical Laboratory, Salt Lake City,
    Utah 84115.

3.  M. Shulsky, "OSHA Method No. 07, Benzene," March, 1985, OSHA
    Analytical Laboratory, Salt Lake City, Utah 84115.
                                 H-365
-------
                                    -5-
t=J
OJ
                                FIGURE  1
              TEXACO  CONROE  CHEMICAL  PLANT
         AMBIENT  AIR  MONITORING  FOR  BENZENE
                                      B
                UNIT
             F-C-29 
           ETHYLENEDIAMINE/ALKY
                CFS-20)
          A REPRESENTS EMISSION SOURCE

          (•(INDICATES SAMPLING POINT
            DAYS 1
     UPWIND
    FENCELINE
                                                0 100 800 300
                GRAPHIC SCALE IN FEET
     ~—- FORMALDEHYDE STORAGE


    BODE STORAGE
          FS-86
           D
                                       5-23
                                  F
                                   PUMP # Q
                                         FS-2
                                                    WIND
                                                  DIRECTION'
                     DDWNVIND JJ
                     FENCELINE
 DUTSIDE
FfiNCELINE
                                              A
DDVNWIND
FENCELINE
                                DAYS 3,4,5,
                                6,7, & 8
-------
                                   -6-
                              FIGURE  2
           TEXACO  CONROE  CHEMICAL PLANT
AMBIENT  AIR  MONITORING  FOR  ETHYLENE  OXIDE
              UNIT
           AMINES PRODUCTION
          NWPHDUie PRODUCTION
              CFS—£09
            oxne STORAGE
           SPILL TC HANDLING
            
-------
                                                                 -7-
                                                               TABLE  I
                                                    TEXACO CONROE CHEMICAL PLANT
                                            RESULTS OF AMBIENT AIR MONITORING  FOR  BENZENE
DAY 1
DATE
4-10-89
TO
4-11-89
DAY 1
DAY 1
DAY 1
DAY 1
DAY 1
DAY 1
DAY 1
DAY 1
DAY 1
DAY 1
DAY 1
DAY 1
SAMPLE
NUMBER
4
5
6
8
10
12
14
16
20
21
24
26
LOCATION
PROCESS
UNIT
PROCESS
UNIT
PROCESS
UNIT
DOWN
WIND
DOWN
WIND
UP WIND
UP WIND
CONSER-
VATEX
PROCESS
UNIT
PROCESS
UNIT
DOWN
WIND
DOWN
WIND
POINT
FS-20
FS-20
FS-20
A
A
C
C
E
FS-20
FS-20
A
A
AVERAGE
WIND
SPEED
(MPH)
12.0
12.0
12.0
12.0
12.0
12.0
12.0
12.0
6.5
6.5
6.5
6.5
DIR.
6.3
6.3
6.3
6.3
6.3
6.3
6.3
6.3
332
332
332
332
AVERAGE
RELATIVE
HUMIDITY
U)
60.2
60.2
60.2
60.2
60.2
60.2
60.2
60.2
65.3
65.3
65.3
65.3
AVERAGE TEMP.
(F)
AT
3 FT.
N.D.
N.D.
N.D.
N.D.
N.D.
N.D.
N.D.
N.D.
N.D.
N.D.
N.D.
N.D.
AT
70 FT.
N.D.
N.D.
N.D.
N.D.
N.D.
N.D.
N.D.
N.D.
N.D.
N.D.
N.D.
N.D.
TIME SAMPLE
(HOUR)
12:OOAM-4 :20PM
(4.33)
9:14AM-4:28PM
(7.23)
9:14AM-4:28PM
(7.23)
9:24AM-5:02PM
(7.63)
9:24AM-5:05PM
(7.64)
9:48AM-5:05PM
(7.64)
9:14AM-4:50PM
(7.60)
9:45AM-5:28PM
(7.72)
4:36PM-11:38PM
(7.03)
4:40PM-11:38PM
(7.04)
5:13PM-11:51PM
(6.63)
5:13PM-11:51PM
(6.63)
SAMPLE
VOLUME
(M3)
0.0995
0.0381
0.0381
0.0391
0.1015
0.0353
0.0339
0.0382
0.1115
0.0528
0.1127
0.0434
SAMPLE
MASS
(UG)
2.050
0.054
2.228
0.082
0.067
0.128
0.058
0.062
2.36
3.61
0.16
0.073
SAMPLE CONG.
(UG/M3)
20.61
1.416
58.47
2.097
0.660
3.626
1.711
1.623
21.17
68.37
1.42
1.682
OSHA 8-HOUR OCCUPATIONAL THRESHOLD LIMIT VALUE OF BENZENE = 3,187 UG/M3

N.D. = NO DATA
-------
                                                                                          -8-
                                                                                  TABLE I  (CONTINUED)
                                                                              TEXACO CONROE CHEMICAL PLANT
                                                                     RESULTS OF AMBIENT AIR MONITORING FOR BENZENE
H
o\
DAY 1
DATE
4-10-89
TO
4-11-89
DAY 1
DAY 1
DAY 1
DAY 1
DAY 1
DAY 1
DAY 1
DAY 1
DAY 1
SAMPLE
NUMBER
28
29
31
35
36
38
39
40
46
LOCATION
DOWN
WIND
UP WIND
CONSER-
VATEX
PROCESS
UNIT
PROCESS
UNIT
PROCESS
UNIT
DOWN
WIND
DOWN
WIND
CONSER-
VATEX
POINT
A
C
E
FS-20
FS-20
FS-20
A
A
E
AVERAGE
WIND
SPEED
(MPH)
6.5
6.5
6.5
5.0
5.0
5.0
5.0
5.0
5.0
DIR.
332
332
332
152
152
152
152
152
152
AVERAGE
RELATIVE
HUMIDITY
U)
65.3
65.3
65.3
85.3
85.3
85.3
85.3
85.3
85.3
AVERAGE TEMP.
(F)
AT
3 FT.
N.D.
N.D.
N.D.
44.6
44.6
44.6
44.6
44.6
44.6
AT
70 FT.
N.D.
N.D.
N.D.
N.D.
N.D.
N.D.
N.D.
N.D.
N.D.
TIME SAMPLE
(HOUR)
5.13PM-11:55PM
(6.64)
5:OOPM-11:45PM
(6.75)
5:30PM-11.11PM
(5.80)
11:39PM-8:35AM
(8.93)
11:39PM-8:35AM
(8.93)
11:45PM-8:40AM
(8.92)
11:53PM-9:03PM
(9.17)
11:53PM-9:03AM
(9.17)
11:20PM-9:12AM
(9.87)
SAMPLE
VOLUME
(M3)
0.0462
0.0514
0.0490
0.0446
0.0446
0.0411
0.0446
0.0446
0.0407
SAMPLE
MASS
(UG)
0.150
0.084
0.084
10.50
11.00
0.087
0.096
0.096
0.098
SAMPLE CONC.
(UG/M3)
3.247
1.634
2.054
235.4
246.6
2.117
2.152
2.085
2.408
                          OSHA 8-HOUR OCCUPATIONAL THRESHOLD LIMIT VALUE OF BENZENE = 3,187 UG/M3
-------
                                                                 -9-
                                                         TABLE I (CONTINUED)
                                                     TEXACO  CONROE CHEMICAL PLANT
                                            RESULTS OF AMBIENT AIR MONITORING FOR BENZENE
DAY 2
DATE
4-11-89
TO
4-12-89
DAY 2
DAY 2
DAY 2
DAY 2
DAY 2
DAY 2
DAY 2
DAY 2
DAY 2
DAY 2
DAY 2
DAY 2
DAY 2
SAMPLE
NUMBER
49
50
52
54
57
58
60
63
64
66
69
70
72
LOCATION
PROCESS
UNIT
PROCESS
UNIT
UP WIND
DOWN
WIND
DOWN
WIND
DOWN
WIND
CONSER-
VATEX
PROCESS
UNIT
PROCESS
UNIT
PUMP #8
DOWN
WIND
DOWN
WIND
UP
WIND
POINT
FS-20
FS-20
C
A
A
A
E
FS-20
FS-20
F
A
A
C
AVERAGE
WIND
SPEED
(MPH)
5.9
5.9
5.9
5.9
5.9
5.9
5.9
5.7
5.7
5.7
5.7
5.7
5.7
DIR.
75.6
75.6
75.6
75.6
75.6
75.6
75.6
119
119
119
119
119
119
AVERAGE
RELATIVE
HUMIDITY
(X)
59.8
59.8
59.8
59.8
59.8
59.8
59.8
51.6
51.6
51.6
51.6
51.6
51.6
AVERAGE TEMP.
(F)
AT
3 FT.
61.3
61.3
61.3
61.3
61.3
61.3
61.3
65.7
65.7
65.7
65.7
65.7
65.7
AT
70 FT.
N.D.
N.D.
N.D.
N.D.
N.D.
N.D.
N.D.
62.1
62.1
62.1
62.1
62.1
62.1
TIME SAMPLE
(HOUR)
8:40AM-3:35PM
(6.92)
8:40AM-3:35PM
(6.92)
8:42AM-3:40PM
(6.97)
8:23AM-3:27PM
(7.07)
9:15AM-3:48PM
(6.55)
9:15AM-3:48PM
(6.55)
9:17AM-3:18PM
(6.00)
3:36AM-10:13PM
(6.60)
3:36PM-10:13PM
(6.60)
3:38PM-9: 43PM
(6.08)
3:40PM-10:04PM
(6.40)
3:40PM-10:04PM
(6.40)
3.50PM-9:55PM
(6.08)
SAMPLE
VOLUME
(M3)
0.0519
0.0519
0.0530
0.0463
0.0462
0.0462
0.0247
0.0465
0.0465
0.0398
0.0309
0.0309
0.0280
SAMPLE
MASS
(UG)
0.110
0.110
0.130
0.140
0.150
0.140
0.240
1.600
1.578
0.150
0.210
0.220
0.200
SAMPLE CONG.
(UG/M3)
2.119
2.119
2.453
3.024
3.247
3.030
9.717
34.40
33.93
3.769
6.800
7.120
7.143
OSHA 8-HOUR OCCUPATIONAL THRESHOLD LIMIT VALUE OF BENZENE = 3,187 UG/M3

N.D. = NO DATA
-------
                                                               -10-

                                                        TABLE I (CONTINUED)
                                                   TEXACO CONROE CHEMICAL PLANT
                                           RESULTS OF AMBIENT AIR MONITORING FOR BENZENE
DAY 2
DATE
4-11-89
TO
4-12-89
DAY 2
DAY 2
DAY 2
DAY 2
DAY 2
DAY 2
DAY 2
DAY 2
SAMPLE
NUMBER
74
77
78
80
83
84
86
88
LOCATION
CONSER-
VATEX
PROCESS
UNIT
PROCESS
UNIT
UP
WIND
DOWN
WIND
DOWN
WIND
PUMP #8
CONSER-
VATEX
POINT
E
FS-20
FS-20
C
A
A
F
E
AVERAGE
WIND
SPEED
(MPH)
5.7
4.4
4.4
4.4
4.4
4.4
4.4
4.4
DIE.
119
87.6
87.6
87.6
87.6
87.6
87.6
87.6
AVERAGE
RELATIVE
HUMIDITY
(X)
51.6
71.3
71.3
71.3
71.3
71.3
71.3
71.3
AVERAGE TEMP.
(F)
AT
3 FT.
65.7
56.5
56.5
56.5
56.5
56.5
56.5
56.5
AT
70 FT.
62.1
52.5
52.5
52.5
52.5
52.5
52.5
52.5
TIME SAMPLE
(HOUR)
3:20PM-9:48PM
(6.47)
10:14PM-8:38AM
(10.40)
10:14PM-8:38AM
(10.40)
9 : 58PM-8 : 24AM
(10.43)
10:05PM-8:28AM
(10.38)
10:05PM-8:28AM
(10.38)
9 : 45PM-8 : 45AM
(11.00)
9:50PM-8:19AM
(10.48)
SAMPLE
VOLUME
(M3)
0.0301
0.0780
0.0780
0.0794
0,0722
0.0722
0.0721
0.0432
SAMPLE
MASS
(UG)
0.240
2.67
2.84
0.250
0.290
0.280
0.210
0.370
SAMPLE CONG.
(UG/M3)
7.973
34.23
36.41
3.149
4.017
3.878
2.913
6.944
OSHA 8-HOUR OCCUPATIONAL THRESHOLD LIMIT VALUE OF BENZENE = 3,187 UG/M3
-------
                                                                                           -11-

                                                                                   TABLE I (CONTINUED)
                                                                               TEXACO CONROE  CHEMICAL PLANT
                                                                      RESULTS OF AMBIENT AIR MONITORING FOR BENZENE
B
OJ
DAY 3
DATE
4-12-89
TO
4-13-89
DAY 3
DAY 3
DAY 3
DAY 3
DAY 3
DAY 3
DAY 3
DAY 3
DAY 3
DAY 3
DAY 3
DAY 3
SAMPLE
NUMBER
91
92
94
97
98
100
102
105
108
111
112
114
LOCATION
PROCESS
UNIT
PROCESS
UNIT
PUMP #8
DOWN
WIND
DOWN
WIND
UP WIND
CONSER-
-VATEX
PROCESS
UNIT
PROCESS
UNIT
DOWN
WIND
DOWN
WIND
UP
WIND
POINT
FS-20
FS-20
F
D
D
B
E
FS-20
FS-20
D
D
B
AVERAGE
WIND
SPEED
(MPH)
7.0
7.0
7.0
7.0
7.0
7.0
7.0
7.0
7.0
7.0
7.0
7.0
DIR.
80.5
80.5
80.5
80.5
80.5
80.5
80.5
66.4
66.4
66.4
66.4
66.4
AVERAGE
RELATIVE
HUMIDITY
(*)
69.0
69.0
69.0
69.0
69.0
69.0
69.0
69.7
69.7
69.7
69.7
69.7
AVERAGE TEMP.
(F)
AT
3 FT.
59.1
59.1
59.1
59.1
59.1
59.1
59.1
62.5
62.5
62.5
62.5
62.5
AT
70 FT.
55.5
55.5
55.5
55.5
55.5
55.5
55.5
58.1
58.1
58.1
58.1
58.1
TIME SAMPLE
(HOUR)
8:40AM-3:18PM
(6.63)
8:40AM-3:18PM
(6.63)
8:50AM-2:OOPM
(5.17)
8:30AM-3:07PM
(5.62)
8:30AM-3:07PM
(5.62)
8 : 25AM-2 : 58PM
(6.55)
8:20AM-2:50PM
(6.54)
3:18PM-8:OOPM
(4.73)
2:57PM-7:58PM
(5.02)
3:10PM-8:OOPM
(4.83)
3:10PM-8:OOPM
(4.83)
2:56PM-8:OOPM
(5.07)
SAMPLE
VOLUME
(M3)
0.0467
0.0467
0.0339
0.0271
0.0271
0.0301
0.0323
0.0291
0.0289
0.0347
0.0347
0.0357
SAMPLE
MASS
(UG)
2.970
3.090
0.250
0.190
0.170
0.044
0.082
2.470
0.055
0.410
0.440
0.059
SAMPLE CONC.
(UG/M3)
63.60
66.17
7.375
7.011
6.273
1.462
2.539
84.88
1.903
11.81
12.68
1.653
                          OSHA 8-HOUR OCCUPATIONAL THRESHOLD LIMIT VALUE OF BENZENE =  3,187  UG/M3
-------
                                                               -12-

                                                        TABLE I  (CONTINUED)
                                                   TEXACO CONROE CHEMICAL PLANT
                                           RESULTS OF AMBIENT AIR MONITORING FOR BENZENE
DAY 4
DATE
4-12-89
TO
4-13-89
DAY 3
DAY 3
DAY 3
DAY 3
DAY 3
DAY 3
DAY 3
SAMPLE
NUMBER
116
119
120
122
125
126
128
LOCATION
PUMP #8
PROCESS
UNIT
PROCESS
UNIT
PUMP #8
DOWN
WIND
DOWN
WIND
CONSER-
VATEX
POINT
F
FS-20
FS-20
F
D
D
E
AVERAGE
WIND
SPEED
(MPH)
7.0
4.6
4.6
4.6
4.6
4.6
4.6
DIR.
66.4
101
101
101
101
101
101
AVERAGE
RELATIVE
HUMIDITY
(X)
69.7
88.8
88.8
88.8
88.8
88.8
1
88.8
AVERAGE TEMP.
(F)
AT
3 FT.
62.5
54.3
54.3
54.3
54.3
54.3
54.3
AT
70 FT.
58.1
52.1
52.1
52.1
52.1
52.1
52.1
TIME SAMPLE
(HOUR)
3:35PM-8:OOPM
(4.42)
9: 05PM-7:49AM
(10.73)
9:05PM-7:49AM
(10.73)
8:55PM-7:50AM
(10.93)
8:52PM-7:47AM
(10.93)
8:52PM-7:47AM
(10.93)
9:12PM-7:43AM
(10.52)
SAMPLE
VOLUME
(M3)
0.0213
0.0662
0.0662
0.0731
0.0671
0.0671
0.0725
SAMPLE
MASS
(UG)
0.440
6.190
5.930
0.890
0.460
0.210
0.210
SAMPLE CONG.
(UG/M3)
20.66
93.50
89.58
12.18
6.86
3.129
2.897
OSHA 8-HOUR OCCUPATIONAL THRESHOLD LIMIT VALUE OF BENZENE = 3,187 UG/M3

N.D. = NO DATA
-------
                                                                                           -13-
OJ
-J
                                                                                   TABLE I (CONTINUED)
                                                                               TEXACO  CONROE  CHEMICAL  PLANT
                                                                      RESULTS OF AMBIENT AIR MONITORING FOR BENZENE
DAY 4
DATE
4-13-89
TO
4-14-89
DAY 4
DAY 4
DAY 4
DAY 4
DAY 4
DAY 4
DAY 4
DAY 4
DAY 4
DAY 4
DAY 4
DAY 4
DAY 4
DAY 4
SAMPLE
NUMBER
131
132
134
137
138
138
140
142
145
146
148
151
152
154
LOCATION
PROCESS
UNIT
PROCESS
UNIT
PUMP #8
DOWN
WIND
DOWN
WIND
DOWN
WIND
UP WIND
CONSER-
VATEX
PROCESS
UNIT
PROCESS
UNIT
PUMP #8
DOWN
WIND
DOWN
WIND
CONSER-
VATEX
POINT
FS-20
FS-20
F
D
D
D
B
E
FS-20
FS-20
F
D
D
E
AVERAGE
WIND
SPEED
(MPH)
8.0
8.0
8.0
8.0
8.0
8.0
8.0
8.0
N.D.
N.D.
N.D.
N.D.
N.D.
N.D.
DIR.
47.3
47.3
47.3
47.3
47.3
47.3
47.3
47.3
N.D.
N.D.
N.D.
N.D.
N.D.
N.D.
AVERAGE
RELATIVE
HUMIDITY
(X)
88.3
88.3
88.3
88.3
88.3
88.3
88.3
88.3
N.D.
N.D.
N.D.
N.D.
N.D.
N.D.
AVERAGE TEMP.
(F)
AT
3 FT.
63.1
63.1
63.1
63.1
63.1
63.1
63.1
63.1
63.1
63.1
63.1
63.1
63.1
63.1
AT
70 FT.
60.2
60.2
60.2
60.2
60.2
60.2
60.2
60.2
60.2
60.2
60.2
60.2
60.2
60.2
TIME SAMPLE
(HOUR)
00:58AM-8:02PM
(7.13)
0:58AM-8:02PM
(7.23)
0: 50AM- 7: 55PM
(7.08)
0:55AM-7:59PM
(7.05)
0:55AM-7:59PM
(7.05)
0: 55AM- 7: 59PM
(7.05)
0:41AM-7:48PM
(7.12)
0:46AM-7:44PM
(7.12)
8:40PM-7:41AM
(11.0)
8:40PM-7:41AM
(11.0)
8:30PM-7:38AM
(11-13)
8: 3 3PM- 7 :35 AM
(11.05)
8:33PM-7:35AM
(11.05)
8:26PM-7:28AM
(11.02)
SAMPLE
VOLUME
(M3)
0.0344
0.0344
0.0498
0.0407
0.0407
0.0407
0.0463
0.0496
0.0675
0.0675
0.0694
0.0768
0.0768
0.0678
SAMPLE
MASS
(UG)
3.110
3.180
0.930
0.033
0.140
0.140
0.150
0.290
1.280
3.150
1.050
0.690
0.980
0.580
SAMPLE CONG.
(UG/M3)
90.41
92.44
18.67
0.811
3.440
3.440
3.240
5.847
18.96
46.66
15.13
8.98
12.76
8.55
OSHA 8-HOUR OCCUPATIONAL
N.D. = NO DATA
                                                    THRESHOLD LIMIT VALUE OF BENZENE = 3,187 UG/M3
-------
                                                                -14-

                                                        TABLE I (CONTINUED)
                                                    TEXACO CONROE  CHEMICAL  PLANT
                                           RESULTS OF AMBIENT AIR MONITORING FOR BENZENE
DAY 5
DATE
4-15-89
TO
4-16-89
DAY 5
DAY 5
DAY 5
DAY 5
SAMPLE
NUMBER
157
158
164
166
LOCATION
PROCESS
UNIT
PROCESS
UNIT
CONSER-
VATEX
UP
WIND
POINT
FS-20
FS-20
E
B
AVERAGE
WIND
SPEED
(MPH)
N.D.
N.D.
N.D.
N.D
DIR.
N.D.
N.D.
N.D.
N.D.
AVERAGE
RELATIVE
HUMIDITY
(X)
N.D.
N.D.
N.D.
N.D.
AVERAGE TEMP.
(F)
AT
3 FT.
54.3
54.3
54.3
54.3
AT
70 FT.
52.1
52.1
52.1
52.1
TIME SAMPLE
(HOUR)
10:30PM-8:41AM
(10.18)
10:30PM-8:41AM
(10.18)
10:45PM-8:35AM
(10.83)
11:02PM-8:33AM
(9.51)
SAMPLE
VOLUME
(M3)
0.0491
0.0491
0.0625
0.0674
SAMPLE
MASS
(UG)
1.680
1.740
0.440
0.447
SAMPLE CONG.
(UG/M3)
34.22
35.44
7.040
6.63
OSHA 8-HOUR OCCUPATIONAL THRESHOLD LIMIT VALUE OF BENZENE =  3,187 UG/M3
N.D. = NO DATA
-------
                                                                                           -15-
                                                                                    TABLE  I  (CONTINUED)

                                                                               TEXACO CONROE CHEMICAL PLANT

                                                                       RESULTS OF AMBIENT  AIR MONITORING  FOR  BENZENE
-J
O\
DAY 6
DATE
4-15-89
TO
4-16-89
DAY 6
DAY 6
DAY 6
DAY 6
DAY 6
DAY 6
DAY 6
DAY 6
DAY 6
DAY 6
DAY 6
DAY 6
DAY 6
DAY 6
DAY 6
DAY 6
DAY 6
DAY 6
SAMPLE
NUMBER
169
170
172
175
178
180
183
184
186
189
190
192
194
197
198
200
203
206
LOCATION
PROCESS
UNIT
PROCESS
UNIT
PUMP #8
DOWN
WIND
CONSER-
VATEX
UP
WIND
PROCESS
UNIT
PROCESS
UNIT
PUMP #8
DOWN
WIND
DOWN
WIND
CONSER-
VATEX
UP
WIND
PROCESS
UNIT
PROCESS
UNIT
PUMP #8
DOWN
WIND
CONSER-
VATEX
POINT
FS-20
FS-20
F
D
E
B
FS-20
FS-20
F
D
D
E
B
FS-20
FS-20
F
D
E
AVERAGE
WIND
SPEED
(MPH)
5.1
5.1
5.1
5.1
5.1
5.1
2.8
2.8
2.8
2.8
2.8
2.8
2.8
2.0
2.0
2.0
2.0
2.0
DIR.
293
293
293
293
293
293
238
238
238
238
238
238
238
206
206
206
206
206
AVERAGE
RELATIVE
HUMIDITY
(X)
73.3
73.3
73.3
73.3
73.3
73.3
66.7
66.7
66.7
66.7
66.7
66.7
66.7
92.4
92.4
92.4
92.4
92.4
AVERAGE TEMP.
(F)
AT
3 FT.
68.2
68.2
68.2
68.2
68.2
68.2
70.4
70.4
70.4
70.4
70.4
70.4
70.4
N.D.
N.D.
44.6
44.6
44.6
AT
70 FT.
67.4
67.4
67.4
67.4
67.4
67.4
70.0
70.0
70.0
70.0
70.0
70.0
70.0
N.D.
N.D.
N.D.
N.D.
N.D.
TIME SAMPLE
(HOUR)
8:40AM-3:45PM
(7.08)
8:40AM-3:45PM
(7.08)
9:21AM-3:40PM
(6.32)
9: 21AM- 3: 34PM
(6.34)
9:35AM-3:31PM
(5.93)
9:28AM-3:26PM
(5.97)
4:20AM-10:29PM
(6.15)
4:20AM-10:29PM
(6.15)
4:15PM-10:36PM
(6.35)
4:11PM-10:56PM
(6.76)
4:11PM-10:56PM
(6.76)
4:28PM-10:56PM
(6.50)
4:04PM-10:49PM
(6.75)
10:28PM-8:58AM
(10.38)
10:28PM-8:58AM
(10.38)
10:38PM-8:50AM
(10.20)
10:50PM-8:52AM
(10.03)
11:03PM-8:45AM
(9.70)
SAMPLE
VOLUME
(M3)
0.0456
0.0456
0.0414
0.0432
0.0306
0.0348
0.0432
0.0432
0.0413
0.0437
0.0437
0.0420
0.0418
0.0627
0.0627
0.0656
0.0658
0.0565
SAMPLE
MASS
(UG)
0.520
3.680
0.010
0.029
0.031
0.056
0.400
0.780
0.690
0.010
0.310
0.200
0.230
0.710
2.510
0.010
0.470
0.430
SAMPLE CONG.
(UG/M3)
11.40
80.70
0.245
0.671
1.013
1.609
9.259
18.06
16.71
0.229
7.094
4.762
5.502
11.32
40.03
0.152
7.143
7.611
                           OSHA 8-HOOR OCCUPATIONAL THRESHOLD LIMIT  VALUE OF BENZENE = 3,187 UG/M3
-------
                                                                                          -16-

                                                                                   TABLE  I  (CONTINUED)
                                                                              TEXACO CONROE CHEMICAL PLANT
                                                                      RESULTS  OF AMBIENT  AIR MONITORING  FOR BENZENE
s
-J
DAY 7
DATE
4-17-89
TO
4-18-89
DAY 7
DAY 7
DAY 7
DAY 7
DAY 7
DAY 7
DAY 7
DAY 7
DAY 7
DAY 7
DAY 7
DAY 7
DAY 7
SAMPLE
NUMBER
211
212
214
217
220
222
225
226
228
231
232
234
236
LOCATION
PROCESS
UNIT
PROCESS
UNIT
PUMP #8
DOWN
WIND
CONSER-
VATEX
UP WIND
PROCESS
UNIT
PROCESS
UNIT
PUMP #8
DOWN
WIND
DOWN
WIND
CONSER-
VATEX
UP
WIND
POINT
FS-20
FS-20
F
D
E
B
FS-20
FS-20
F
D
D
E
B
AVERAGE
WIND
SPEED
(MPH)
N.D.
N.D.
N.D.
N.D.
N.D.
N.D.
N.D.
N.D.
N.D.
N.D.
N.D.
N.D.
N.D.
DIR.
N.D.
N.D.
N.D.
N.D.
N.D.
N.D.
N.D.
N.D.
N.D.
N.D.
N.D.
N.D.
N.D.
AVERAGE
RELATIVE
HUMIDITY
(X)
N.D.
N.D.
N.D.
N.D.
N.D.
N.D.
N.D.
N.D.
N.D.
N.D.
N.D.
N.D.
N.D.
AVERAGE TEMP.
(F)
AT
3 FT.
69.0
69.0
69.0
69.0
69.0
69.0
85.7
85.7
85.7
85.7
85.7
85.7
85.7
AT
70 FT.
66.2
66.2
66.2
66.2
66.2
66.2
79.1
79.1
79.1
79.1
79.1
79.1
79.1
TIME SAMPLE
(HOUR)
9:48AM-1:32PM
(3.75)
9:48AM-1:32PM
(3.75)
9:40AM-1:25PM
(3.75)
9:39AM-0:37PM
(3.01)
9:58AM-0:40PM
(2.70)
10:02AM-1:35PM
(2.55)
1 : 25PM-5 : 00PM
(3.58)
1:25PM-5:OOPM
(3.58)
1:28PM-5:OOPM
(3.55)
1:15PM-5:OOPM
(3.75)
1:15PM-5:OOPM
(3.75)
1:30PM-5:OOPM
(3.50)
0:55PM-5:OOPM
(4.08)
SAMPLE
VOLUME
(M3)
0.0239
0.0239
0.0233
0.0195
0.0190
0.0158
0.0185
0.0185
0.0173
0.0241
0.0241
0.0211
0.0238
SAMPLE
MASS
(UG)
1.58
0.53
0.74
0.83
0.83
0.96
1.11
2.00
0.92
1.03
1.10
1.09
1.06
SAMPLE CONC.
(UG/M3)
66.11
22.18
31.76
42.54
43.68
60.75
60.00
108.1
53.18
42.74
45.64
51.66
44.53
                          OSHA 8-HOUR OCCUPATIONAL THRESHOLD LIMIT VALUE OF BENZENE = 3,187 UG/M3

                          N.D.  = NO DATA
-------
                                                                                           -17-

                                                                                   TABLE I (CONTINUED)
                                                                               TEXACO CONROE  CHEMICAL  PLANT
                                                                      RESULTS OF AMBIENT AIR MONITORING FOR BENZENE
H
-------
                        -18-

                     TABLE II
            TEXACO CONROE CHEMICAL PLANT
    MAXIMUM AND MINIMUM BENZENE CONCENTRATIONS
            FROM AMBIENT AIR MONITORING
MONITORING SITE
DOWNWIND
AT FENCELINE
UPWIND
AT FENCELINE
PROCESS UNIT
OUTSIDE FENCELINE
(CONSERVATEX CO)
PUMP NUMBER 8
CONCENTRATION (UG/M3)
MINIMUM
0.229
1.462
0.304
1.013
0.150
MAXIMUM
45.64
60.75
246.6
51.66
53.18
OSHA 8-HOUR OCCUPATIONAL THRESHOLD LIMIT VALUE OF
BENZENE = 3,187 UG/M3
                           H-379
-------
                                                                                          -19-
                                                                                        TABLE  III

                                                                              TEXACO CONROE CHEMICAL PLANT

                                                                  RESULTS OF AMBIENT AIR MONITORING FOR ETHYLENE OXIDE
DAY 1
DATE
4-10-89
TO
4-11-89
DAY 1
SAMPLE
NUMBER
17
LOCATION
PROCESS
UNIT
POINT
FS-20
AVERAGE
WIND
SPEED
(MPH)
6.5
DIR.
332
AVERAGE
RELATIVE
HUMIDITY
(*)
65.3
AVERAGE TEMP.
(F)
AT
3 FT.
N.D.
AT
70 FT.
N.D.
TIME SAMPLE
(HOUR)
4:24PM-11:32PM
(7.15)
SAMPLE
VOLUME
(M3)
0.1136
SAMPLE
MASS
(UG)
9.60
SAMPLE CONG.
(UG/M3)
84.51
oo
o
DAY 2
DATE
4-11-89
TO
4-12-89
DAY 2
DAY 2
SAMPLE
NUMBER
61
75
LOCATION
PROCESS
UNIT
PROCESS
UNIT
POINT
FS-20
FS-20
AVERAGE
WIND
SPEED
(MPH)
5.7
4.4
DIR.
119
87.6
AVERAGE
RELATIVE
HUMIDITY
U)
51.6
71.3
AVERAGE TEMP.
(F)
AT
3 FT.
65.7
56.5
AT
70 FT.
62.1
52.5
TIME SAMPLE
(HOUR)
3:33PM-10:09PM
(6.60)
10:10PM-8:34AM
(10.4)
SAMPLE
VOLUME
(M3)
0.0341
0.0708
SAMPLE
MASS
(UG)
6.51
2.82
SAMPLE CONG.
(UG/M3)
190.9
39.83
DAY 3
DATE
4-12-89
TO
4-13-89
DAY 3
DAY 3
DAY 3
DAY 3
SAMPLE
NUMBER
90
117
118
121
LOCATION
PROCESS
UNIT
PROCESS
UNIT
PUMP #8
PUMP #8
POINT
FS-20
FS-20
F
F
AVERAGE
WIND
SPEED
(MPH)
7.0
4.6
4.6
4.6
DIR.
80.5
101
101
101
AVERAGE
RELATIVE
HUMIDITY
(X)
69.0
88.8
88.8
88.8
AVERAGE TEMP.
(F)
AT
3 FT.
59.1
54.3
54.3
54.3
AT
70 FT.
55.5
52.1
52.1
52.1
TIME SAMPLE
(HOUR)
8:35AM-3:14PM
(6.65)
8:52PM-7:48AM
(10.93)
8:52PM-7:48AM
(10.93)
8:55PM-7:50AM
(10.93)
SAMPLE
VOLUME
(M3)
0.0344
0.0681
0.0681
0.0731
SAMPLE
MASS
(UG)
2.90
3.00
1.86
2.83
SAMPLE CONG.
(UG/M3)
84.30
44.05
27.31
38.71
                          OSHA 8-HOUR OCCUPATIONAL  THRESHOLD LIMIT VALUE OF ETHYLENE OXIDE = 1,798 UG/M3

                          N.D. = NO DATA
-------
                                                                                          -20-
t=J
oo
                                                                                  TABLE  III  (CONTINUED)
                                                                              TEXACO CONROE CHEMICAL PLANT
                                                                  RESULTS OF AMBIENT AIR MONITORING FOR ETHYLENE OXIDE
DAY 4
DATE
4-14-89
TO
4-15-89
DAY 4
DAY 4
DAY 4
DAY 4
DAY 4
SAMPLE
NUMBER
143
144
147
149
150
LOCATION
PROCESS
UNIT
PROCESS
UNIT
PUMP #8
DOWN
WIND
DOWN
WIND
POINT
FS-20
FS-20
F
D
D
AVERAGE
WIND
SPEED
(MPH)
N.D.
N.D.
N.D.
N.D.
N.D.
DIR.
N.D.
N.D.
N.D.
N.D.
N.D.
AVERAGE
RELATIVE
HUMIDITY
(%)
N.D.
N.D.
N.D.
N.D.
N.D.
AVERAGE TEMP.
(F)
AT
3 FT.
63.1
63.1
63.1
63.1
63.1
AT
70 FT.
60.2
60.2
60.2
60.2
60.2
TIME SAMPLE
(HOUR)
8:36PM-7:40AM
(11.07)
8:36PM-7:40AM
(11.07)
8:30PM-7:38AM
(11.13)
8:32PM-7:35AM
(11.05)
8: 32PM- 7: 35AM
(11.05)
SAMPLE
VOLUME
(M3)
0.0672
0.0672
0.0694
0.0738
0.0738
SAMPLE
MASS
(UG)
1.24
3.15
2.28
1.49
0.98
SAMPLE CONG.
(UG/M3)
18.45
46.88
32.85
20.19
13.28
                          OSHA  8-HOUR  OCCUPATIONAL THRESHOLD LIMIT VALUE OF ETHYLENE OXIDE = 1,798 UG/M3

                          N.D.  = NO DATA
DAY 5
DATE
4-15-89
TO
4-16-89
DAY 5
DAY 5
SAMPLE
NUMBER
159
162
LOCATION
PUMP #8
DOWN
WIND
POINT
F
D
AVERAGE
WIND
SPEED
(MPH)
N.D.
N.D.
DIR.
N.D.
N.D.
AVERAGE
RELATIVE
HUMIDITY
(X)
N.D.
N.D.
AVERAGE TEMP.
(F)
AT
3 FT.
54.3
54.3
AT
70 FT.
52.1
52.1
TIME SAMPLE
(HOUR)
10:25PM-8:38AM
(10.22)
10:32PM-8:40AM
(10.13)
SAMPLE
VOLUME
(M3)
0.0719
0.0624
SAMPLE
MASS
(UG)
2.06
2.85
SAMPLE CONG.
(UG/M3)
28.65
45.67
-------
                                                                                           -21-
oo
K)
                                                                                  TABLE III (CONTINUED)

                                                                               TEXACO  COKROE CHEMICAL PLANT

                                                                   RESULTS OF  AMBIENT  AIR MONITORING FOR ETHYLENE OXIDE
DAY 6
DATE
4-16-89
TO
4-17-89
DAY 6
SAMPLE
NUMBER
182
LOCATION
PROCESS
UNIT
POINT
FS-20
AVERAGE
WIND
SPEED
(MPH)
2.8
DIR.
238
AVERAGE
RELATIVE
HUMIDITY
(X)
66.7
AVERAGE TEMP.
(F)
AT
3 FT.
70.4
AT
70 FT.
70.0
TIME SAMPLE
(HOUR)
4:15PM-10:30PM
(6.25)
SAMPLE
VOLUME
(M3)
0.0463
SAMPLE
MASS
(UG)
1.220
SAMPLE CONC.
(UG/M3)
26.35
                          OSHA 8-HOUR OCCUPATIONAL THRESHOLD LIMIT VALUE OF ETHYLENE OXIDE = 1,798 UG/M3



                          N.D. = NO DATA
DAY 7
DATE
4-17-89
TO
4-18-89
DAY 7
SAMPLE
NUMBER
223
LOCATION
PROCESS
UNIT
POINT
FS-20
AVERAGE
WIND
SPEED
(MPH)
N.D.
DIR.
N.D.
AVERAGE
RELATIVE
HUMIDITY
(X)
N.D.
AVERAGE TEMP.
(F)
AT
3 FT.
69.0
AT
70 FT.
66.2
TIME SAMPLE
(HOUR)
1:25PM-5:OOPM
(3.58)
SAMPLE
VOLUME
(M3)
0.0235
SAMPLE
MASS
(UG)
1.55
SAMPLE CONC.
(UG/M3)
65.96
DAY 8
DATE
4-18-89
TO
4-19-89
DAY 8
DAY 8
DAY 8
SAMPLE
NUMBER
237
241
284
LOCATION
PROCESS
UNIT
PUMP #8
PROCESS
UNIT
POINT
FS-20
F
FS-20
AVERAGE
WIND
SPEED
(MPH)
6.0
6.0
N.D.
DIR.
218
218
N.D.
AVERAGE
RELATIVE
HUMIDITY
(2)
59.2
59.2
N.D.
AVERAGE TEMP.
(F)
AT
3 FT.
85.4
85.4
N.D.
AT
70 FT.
82.3
82.3
N.D.
TIME SAMPLE
(HOUR)
2:18PM-7:36PM
(5.30)
2:10PM-7:30PM
(5.33)
9: 14AM- 4: 28PM
(11.8)
SAMPLE
VOLUME
(M3)
0.0345
0.0375
0.0687
SAMPLE
MASS
(UG)
9.100
2.910
29.5
SAMPLE CONC.
(UG/M3)
263.8
77.6
429.4
-------
                          -22-

                        TABLE IV
              TEXACO CONROE CHEMICAL PLANT
          MAXIMUM ETHYLENE OXIDE CONCENTRATIONS
               FROM AMBIENT AIR MONITORING
MONITORING SITE
DOWNWIND
AT FENCELINE
UPWIND
AT FENCELINE
PROCESS UNIT
OUTSIDE FENCELINE
(CONSERVATEX CO)
PUMP NUMBER 8
MAXIMUM CONCENTRATION
(UG/M3)
45.67
<26.00
429.4
<26.00
77.6
OSHA 8-HOUR OCCUPATIONAL THRESHOLD LIMIT VALUE OF ETHYLENE
OXIDE= 1,798 UG/M3
                              H-383
-------
                   -23-

                 TABLE V
       TEXACO CONROE CHEMICAL PLANT
RESULTS OF DESORPTION EFFICIENCY STUDIES
      FOR  BENZENE AND ETHYLENE  OXIDE
AMOUNT SPIKE PER
(MICROGRAMS)
440
220
44
8.8
1.76
BENZENE
AMOUNT RECOVERED
(MICROGRAMS)
376
211
39.7
6.9
0.43
AVERAGE % RECOVERED FOR BENZENE
AMOUNT SPIKE PER
(MICROGRAMS)
18
35.9
RECOVERED
(%)
85
96
90
78
25
75
ETHYLENE OXIDE
AMOUNT RECOVERED
(MICROGRAMS)
14.6
25.6
AVERAGE % RECOVERED FOR
| ETHYLENE OXIDE
RECOVERED
(%)
81
71
76
                      H-384
-------
                        USEPA PROPOSED METHOD 25D
         FOR DETERMINING THE VOLATILE ORGANIC CONTENT OF WASTES:
                 EVALUATION ON REAL-WORLD WASTE SAMPLES
Thomas  L.  Dawson,  Ph.D.,  Group  Leader,  Union  Carbide  Chemicals  and
Plastics Company  Inc.,  P. 0.  Box 8361,  South  Charleston,  West Virginia
25303; Larry  I.  Bone,  Ph.D.,  Senior  Associate  Environmental Consultant,
Dow  Chemical  Company,  3867  Plaza  Tower Drive  Ste  140,   Baton  Rouge,
Louisiana  70816;  Brenda  A.   Cuccherini,   Ph.D.,   Associate  Director,
Chemical  Manufacturers  Association,   2501  M  Street,  NW,  Washington,
District  of   Columbia  20037;   Neil   E.   Prange,   Senior  Environmental
Consultant,  Monsanto  Company,   800   N.   Lindbergh  Blvd.,   St.   Louis,
Missouri 63167

ABSTRACT

USEPA  is  developing  regulations  to  control  secondary  emissions  from
hazardous  waste   TSDFs  and  industrial  wastewater systems.   Supporting
this  effort,  they have  developed proposed  Method  25D to  determine  the
aggregate  volatile  organic  compound   (VOC)  content of hazardous  wastes
and  wastewaters.   The  Chemical  Manufacturers  Association  (CMA)  has
conducted  a  study using  four  real-world wastes to compare  the proposed
method to routinely used methods and to evaluate its reproducibility.

Based  on  this  CMA  study,  proposed  Method  25D  appears  to  give  a
"ballpark",  but   somewhat  high,  estimate  of  the total  volatiles;  it
apparently  picks  up  a significant  amount  (10 to  30 percent)  of  the
semi-volatiles.   It  does,   therefore,  show potential  as  a  screening
tool.  Although  there  was some  heterogeneity of the  waste  samples,  the
magnitude  of  the  lab  to  lab  variation,  and  the  sometimes  large,
variable  RSDs, suggest that  the reproducibility of the current proposed
method  is  lacking.   Recommendations  are  made  to  study/improve  the
accuracy and precision and  to then set  the  conditions to  simulate, more
accurately,  actual   expected  secondary  emissions  of  volatile  organic
compounds from real waste facilities.

INTRODUCTION

The U.S. Environmental Protection Agency (EPA)  is  developing regulations
to control  secondary  emissions  from  hazardous waste  treatment,  storage
and  disposal  facilities  (TSDFs)  and   industrial  wastewater  systems.
Supporting this effort,  the  Agency's Office  of  Air Quality Planning and
Standards  (OAQPS)  at  Research  Triangle   Park,  NC,  has  developed  a
proposed method   to  determine  the  aggregate  volatile organic  compound
(VOC) content  of  hazardous  wastes and  wastewaters.   EPA  has conducted a
multi-lab  study   of  the  proposed  method  which   focused   on  synthetic
wastes,  but  they  also plan  to evaluate  it  further on real  wastes.   In
cooperation  with EPA,  CMA  has  conducted  a separate study using four
actual  chemical   industry  wastes  to  compare   proposed  Method  25D  to
                                  H-385
-------
routinely used methods,  i.e.,  SW-846 Methods  8240  and 8270.   The  major
objectives of this  study were to  determine  whether proposed Method  25D
gives a  reasonable estimate  for  the  VOC content  and  to evaluate  its
reproducibility.

EPA PROPOSED METHOD 25D

As proposed Method 25D, Determination of  the Volatile  Organic  Content of
Waste Samples, has  been published by  EPA,  the  details  of the proposed
method will not  be repeated here.

According  to  the   proposed   Method  25D   protocol,   the  samples   of
waste/wastewater  are  dissolved/dispersed in  polyethylene glycol  (PEG).
The  solutions/dispersions  are  heated to  75°C,  stirred in a chamber  and
purged with  nitrogen  for  thirty  minutes  at  a  rate  of  six liters  per
minute.   The effluent gas is split into  two  streams; one  is  analyzed  for
carbon content with a  flame ionization  detector (FID), and the other is
analyzed  for  chlorine content  (as  chloride)  with  a  Hall  electrolytic
conductivity detector (HECD).

The  detectors are calibrated  with gas(es),  in the  form  of propane  and
dichloroethane,  containing  known  amounts of  carbon (C)  as methane  and
chlorine  (Cl)  per liter.   Curves are  determined  for  each detector  to
establish  detector  response   and  linearity.    Sample  quantitation   is
accomplished  through  the  use  of  average  response  factors   from  the
multipoint calibration curves and,  if  required according to the  method,
the  end  of  day  calibration checks.  The VOC by  proposed Method 25D  is
the sum of the carbon (as methane) and  chlorine contents  of the sample.
COMPARISON METHODS

Two  of  the  samples  were  also  analyzed  by  GC/MS  for  volatiles  and
semi-volatiles  as a means  of  comparing the results  obtained  by  proposed
Method 25D to  existing,  routinely used  methodology.   Samples CMA-3  and
CMA-4  were  analyzed by  Radian (Austin,  TX)  by  Methods  8240DI  (Direct
Injection),  8240  (purgeable volatiles),  and  8270  (semi-volatiles).   The
GC/MS  analytical  methodology  followed   EPA  SW-846   (Third   Edition)
protocol.

THE REAL-WORLD  WASTE SAMPLES

The  four chemical industry wastes were  purposely  selected  to be  very
different in nature and  to challenge  the  proposed method across  a  wide
range of waste types and concentrations.   Since the  purpose  of  the study
was to test the  proposed Method 25D,  i.e., not  to determine the  actual
contents of the  wastes,  two of the wastes were  modified  to  broaden  the
scope  of the  test;  one  is  a mixture of  high-salt  and  high-organic
content  waste  streams  and  one was diluted to broaden  the concentration
range of the  organics  in the waste samples.  The  identities  of  the waste
samples tested  are listed in Table I,  as follows:
                                  n-386
-------
                                 TABLE I
                    IDENTITIES OF REAL WASTE SAMPLES
SAMPLE ID       SOURCE                               DESCRIPTION
CMA-1           Partially-stripped styrene tar       Highly-viscous tar

CMA-2           Mixture of high-salt and high-       Wastewater with
                organic content wastewater           slight sheen
                streams

CMA-3           Wastewater treatment plant           Primary sludge

CMA-4           Wastewater treatment plant           Diluted two-phase
                skimming tank                        supernatant from
                                                     top of clarifier
SAMPLING TECHNIQUES

Single  samples were  collected in  clean bottles  from each  source,  and
some were combined  and/or  diluted to give the  final  samples.   CMA-2 was
prepared  by mixing  the two  wastewater  stream samples  in  a  separate,
single  bottle;  the  final sample  mixture  had  a slight sheen.   CMA-4 was
prepared  by  mixing  samples   of  organic  and  aqueous layers  from  the
skimmer, and then diluting  the mixture  further with distilled,  deionized
water;  the  final  mixture had  two distinct  phases.  CMA-3 was  not  mixed
with any other material or diluted; it was the most homogeneous.

Radian  provided cleaned,  labeled  and  weighed  sample  bottles containing
polyethylene glycol (PEG)  with  approximately 10  to  15  ml  of  headspace
available for  sample  collection.   The four  final  CMA samples were shaken
or  stirred  in  their  single sample  bottles  to improve homogeneity  just
before  they were  added  to  the vials;  care  was taken to not  spill the
waste on  the  outside  of the  bottles.   Samples were  shipped  to  Radian
under ice  and  received at  approximately ice  temperatures  (4°C).   The
sample  bottles received  at Radian were  then reweighed  to obtain,  by
difference,  the weight of the sample collected.

Five replicate  proposed Method 25D  samples  were   collected  from  each of
the four  final sample  bottles.   Three  of  the samples  from each  waste
stream  were analyzed  to provide  an estimate  of   laboratory precision.
Two  samples were  kept  in  reserve in  case  of   breakage   or  equipment
failure.  Five  additional  samples were  collected  from each  of  the  same
four final  sample bottles  and sent  to  an EPA  contracted  laboratory for
analysis.
                                  H-387
-------
Samples of CMA-3 and  CMA-4 were also  collected for analysis  by Methods
8240,   8240  Direct  Injection,  and  8270.   Samples  for  Method  8240  were
collected in PEG in the  same  manner  as the proposed Method  25D samples.
PEG was  used  as the collection  media  because  it  does  not purge  and  it
inhibits  sample  volatility prior  to  analysis.   The  8240 samples  were
collected (diluted)  in PEG due  to the  expected high  concentrations  of
volatiles.  Samples  collected for  the  8240  Direct  Injection  and  8270
analyses were not collected in the PEG.   All of these  samples came  from
the same final sample bottles that were  used  to fill  the proposed Method
25D vials.

RESULTS AND DISCUSSION

Proposed Method 25D Analysis of CMA Waste Samples

The analytical results from the  proposed Method 25D analyses  of the CMA
waste  samples  are  presented  in  Table  II.   Each   sample,   with   the
exception of  CMA-4 at Radian,  was analyzed  in triplicate.   Triplicate
analyses  of  the  CMA-4 samples at  Radian were  not possible because the
flame  on the  FID was  extinguished  during  analysis   of  three of   the
replicate samples  and  a  data  system  failed on  one of  the two remaining
samples.  The reason for the  extinguished  flame on the FID has not  been
determined.

The  concentrations of  volatile  organic  compounds in the  four waste
samples, as measured with proposed Method  25D,  ranged from about 350 ppm
to  about  20,000 ppm.   Three  of  the  samples  contained  1-3 ppm chlorine
and the  other about  300  to  600 ppm chlorine.   Broad ranges  of chlorine,
carbon as methane, and volatile  organic  content, as measured  by proposed
Method 25D,  were, therefore, covered  by these four real-world wastes.

The averages, the  percent  relative standard deviations, and  the percent
differences between the averages from  the  two laboratories are presented
in  Table III.    In some  cases the percent relative standard  deviations
were fairly large;  this is possibly due  in part to some heterogeneity of
the  samples,  but  the variabilities  in the  percent  RSDs suggest  that
measures need to be  taken  to  improve the  reproducibility  of  the method.
The lab  to lab  variation is even more  severe, with differences  of 14 to
71  percent for  carbon  (as  methane) and  of  the  order  of 100 percent for
chlorine.  Statistical analysis  of the  data  showed significant differ-
ences  in  intra-laboratory  precision, and significant  differences  in the
results from the two  laboratories.

GC/MS Analysis of CMA Waste Samples

Compounds (volatiles)  detected in samples CMA-3 and CMA-4 by Methods
8240 and 8240DI are listed in Tables IV  and VI, respectively.   Compounds
(semi-volatiles) detected in  samples CMA-3 and CMA-4 by Method 8270 are
listed in Tables V and  VII,  respectively.   Several  different  types  of
volatile and semi-volatile  compounds were  present  in both of  these  real
wa s t e s.
                                  E-388
-------
Each  of  the  Tentatively  Identified  Compounds  (TICs)  and  the  unknown
compounds  have  been  quantitated  using  the  response  factor  from  the
nearest  internal  standard  as  they  are  not  target  analytes  for  the
particular method  and  response  factors  have not  been  determined.   Since
the actual response factor is not  known,  the concentrations  for the TICs
can be very close  to  the true concentration or  vary widely.   Therefore,
the values of the TICs are estimated concentrations.

The  carbon  and   chlorine   contents  which   were  calculated   from  the
empirical formulas  of  the compounds are also  shown in Tables  IV  - VII.
For the unknowns,  75 percent  carbon and  zero percent  chlorine  were used,
i.e., the same percentages as for methane.
Comparison  of  Results from  Proposed  Method  25D to Results  from Methods
8240, 8240DI and 8270

Table VIII  provides  comparisons of the results  from  proposed  Method 25D
to the results  from  Methods  8240 and 8240 DI  (volatiles)  for  both CMA-3
and CMA-4.  Note that the  comparison  here  is  for the  results as reported
for the two methods,  i.e.,  observed for the 8000 methods  and  the  sum of
carbon (as methane)  plus chlorine  for 25D.   There  are wide variations in
the  results from  the different  labs and  the two samples,  but  on the
average  it appears   that  proposed  Method  25D overpredicts  the  total
volatiles  content  of real-world wastes,  i.e., that  proposed  Method 25D
picks up  a  significant  amount (10 to 30 percent)  of  the semi-volatiles.
In  two   out   of   three  cases,  the  proposed  Method   25D  data  were
statistically  significantly  higher than  those from  Method  8240/8240DI.
Perhaps the proposed  Method  25D test  conditions are too  vigorous causing
excessive amounts of  semi-volatiles to also be driven off the wastes.

The numbers in  parentheses are  an attempt to  make  the comparisions  on a
more  uniform  basis, i.e.,   carbon  as  carbon  plus  chlorine   for  both
methods.  The differences are not  great,  but  it does  illustrate that the
method of presenting  the results does influence the reported contents as
well as the comparisons.

Table IX  provides comparisons of the  results  from proposed Method 25D to
the sum of  the results from  8240/8240DI  and  8270,  i.e.,  the  sum  of the
volatiles plus  semi-volatiles.   Again, the variations are large.

CONCLUSIONS

Based  on  triplicate  analyses  by  proposed  Method 25D  at  two separate
laboratories   of  four  real-world   chemical   industry   wastes,   and
comparisons of  the proposed  Method 25D results  with  SW-846 Methods  8240
and 8270  analyses of  two of  the  real wastes, it appears  that:
                                  H-389
-------
Proposed Method  25D  gives a  "ballpark",  but somewhat high,  estimate of
the total volatiles  contents  of the two  real-world wastes.   In  two out
of three cases, the  proposed  Method  25D  data were statistically signifi-
cantly higher  than  those from  Method  8240/8240DI.   It  also picks  up  a
significant  amount  (10  to 30 percent)  of the  semi-volatiles.   It does,
therefore,  show potential as  a  screening tool.   It  is  expected that the
extent  to  which  it  will overpredict  the volatiles content  will  be  a
function of  the actual composition of the waste.

In  some  cases,  the  percent  relative  standard  deviations  were  fairly
large.  This  is  possibly due in part  to  observed heterogeneity  of the
samples, but  there  were significant variabilities  in the  percent RSDs.
The   lab  to   lab  variations  are  even  more   severe,   with  observed
differences  of 14 and 71 percent carbon (as  methane)  and  of the order of
100  percent   for  chlorine.    Statistical  analysis  of  the  data  shoxred
significant  differences  in  the intra-laboratory  precision, and  signi-
ficant  differences  in  the  results  from  the  two  laboratories.  The
magnitude of the  lab  to  lab variation, and the  sometimes  large, variable
percent RSDs,  suggest that the  reproducibility  of  the proposed method is
lacking.

RECOMMENDATIONS

Accuracy

It is recommended that  the  comparison  of proposed Method  25D  to  Methods
8240,  8240DI  and  8270   be  pursued  further.   Artificial  waste  samples
should   be   prepared   containing   known  concentrations    of   specific
compounds,  such as  those identified in the  CMA wastes, and  analyzed by
proposed Method   25D and Methods   8240,  8240DI  and  8270 using  actual
standards  for the   compounds  of  interest.   Audit   sample  results  for
Methods 8240  and  8270 should  be reviewed  for accuracy and  precision to
ensure proper identifications and  quantitation.   Direct  comparison could
then  be  made  between  the  known  concentrations,   the  alternate  method
results, and proposed Method 25D results.

Should the results  of proposed  Method 25D not  agree with  the alternate
methods, it  is  also strongly  suggested  that  a  Quality   Control  Check
Sample be prepared for proposed  Method 25D using  known concentrations of
specific compounds to determine if the problem originates  with proposed
Method 25D.    This check sample  would  then be  used  to validate  the gas
based calibration curve.  A check  sample  would allow evaluation  of the
entire apparatus  in  a mode  equivalent to  that  used  in  sample analysis.
Verification of the  check sample concentration by Method  8240  or 8240DI
would also  provide a measure of  the accuracy  of  proposed Method 25D.

The calibration gas  is not introduced  through the entire  purge apparatus
when  calibrating  proposed Method   25D.   It   is,  therefore,  recommended
that  this  feature  be  altered   to  introduce  the  standard through  the
entire  apparatus.    It   is   currently  being  introduced   through  the
                                  H-390
-------
coalescing  filter  and  not  through  the   purge   chamber.   The  gaseous
standards should also  be  compared to liquid standards (calibration check
standards)  on  a  routine  basis  to  insure  the  validity  of  a  gaseous
calibration.

Once  the  accuracy  of  the  proposed method is  established,  the specified
purge  conditions  should  be  set  to  simulate  more  accurately  actual
expected  secondary  emissions of  volatile  organic compounds  from actual
waste facilities.

Precision

The  precision  of  the  proposed  method  needs  to  be  improved.    The
stability/reproducibility  of the detectors and  the  effects of  gas  flow
rates need  to be  investigated further.  Improvements  in  calibration and
quality  control  procedures  may  be  required.   Since  many  real-world
wastes  are  nonhomogeneous,  homogeneity  must  be  addressed  from  the
sampling  viewpoint.   Addition of  the  static  mixer,  as described  in the
proposed method, to streams  which contain  multiple phases should help to
obtain  more  homogeneous   samples   of  liquid  wastes.    This  sampling
apparatus should be evaluated using real-world wastes.

Detection Limit

A detection  limit  study using artificial  samples  of  known concentration
should be initiated to  determine the lower working range  of the  proposed
method.    Seven  replicate  samples at  a concentration  considered to  be
within a  factor of  10 of  the method detection limit would  be analyzed.
The  detection  limit  would  be  defined  as  3.14  times  the  standard
deviation  from the  seven  replicates.   The  detection   limit  will,  of
course,  vary depending  on  the compounds present,  even in  water,  and  also
vary further in different matrices.

ACKNOWLEDGEMENTS

This  work  was  supported   by  funding  from  CMA.   Radian  Corporation
(Austin, TX)  performed the proposed Method 25D and Methods 8240,  8240DI
and  8270   analyses   for  CMA,   provided  the   sample   containers,   and
participated  in helpful discussions.  Thanks also to USEPA for  allowing
Radian to use the proposed Method  25D apparatus  for  the  analyses and for
sharing  the  results  of their  analyses.   The  authors also  acknowledge
with  thanks  the  helpful  discussions  with  the  other   members  of  the
Environmental Monitoring  Task Group  of CMA,  Kenneth  DuPuis  (Ciba-Geigy
Corporation), Kathy  Hillig,  Ph.D.,   (BASF  Corporation),   Richard  Javick
(FMC  Corporation),  William  Krochta,  Ph.D.,  (PPG  Industries, Inc.)  and
George Stanko (Shell Development  Company).  Thanks also  to Kathy Hillig,
Ph.D.,  (BASF Corporation)   and  Deborah Miller,   Ph.D.,   (Union  Carbide
Chemicals and Plastics  Company  Inc.)  for  their  help with  the review of
the data.
                                  n-39i
-------
TABLE  II
RESULTS
PROPOSED METHOD 25D ANALYSIS OF CMA WASTE SAMPLES
BY CMA (RADIAN,
AUSTIN, TX) AND EPA CONTRACT LABORATORIES
CONCENTRATIONS IN PARTS PER MILLION BY WEIGHT
SAMPLE ID CARBON (AS CH4)
CMA- RADIAN
1 11300
18600
29800
2 13000
7910
4120
3 3050
3000
3500
4 163
	
	
EPA
14300
15100
14700
6930
6600
8150
3680
3800
3860
296
320
415
CHLORINE
RADIAN
ND
ND
ND
1.20
0.42
ND
301
232
334
ND
	
	
EPA
1.75
1.47
1. 34
1.71
1.84
1.97
601
606
645
4.24
2.35
1.64
TOTAL VO
RADIAN
11300
18600
29800
13000
7910
4120
3350
3230
3830
163
	
	
CONTENT
EPA
14300
15100
14700
6930
6600
8150
4280
4410
4500
300
322
417
 H-392
-------
                                TABLE III
                                PRECISION
            PROPOSED METHOD 25D ANALYSIS  OF CMA WASTE SAMPLES
        BY  CMA  (RADIAN,  AUSTIN, TX)  AND EPA CONTRACT LABORATORIES
      PERCENT RELATIVE STANDARD DEVIATION AND LAB TO LAB VARIATION
             CONCENTRATIONS IN PARTS  PER MILLION BY WEIGHT
            CARBON (AS CH4)
            RADIAN    EPA
                     PERCENT
                    DIFFERENCE
                            CHLORINE
                         RADIAN
                    EPA
                  PERCENT
                 DIFFERENCE
CMA-1
Average
RSD (%)
19900    14700
47       2.7
               30
          ND
          1.52
          14.0
CMA-2
Average
RSD (%)
8340
55
7230
11.2
                                    14
                                  0.54(1)   1.84
                                  113       7.2
                                              109
CMA-3

Average
RSD (%)
3180
8.7
3780
2.4
17
289
18
617
3.9
72
CMA-4

Average
RSD (%)
163(2)   344
         18.3
               71
          ND(2)
          2.74
          49
(1)  One ND averaged in as zero.
(2)  Value for analysis of a single sample only.
                                   H-393
-------
                                TABLE IV
                        VOLATILES IN SAMPLE CMA-3
              COMPOUNDS DETECTED BY METHODS 8240 AND 8240DI
COMPOUNDS
                                            CONCENTRATION (mg/L)
OBSERVED
CARBON
CHLORINE
Toluene
  1220
 1114
2-Propylfuran
             (1)
97(580)
       (2)
  443
Ethanediolc Acid,

   Dibutyl Ester(1)
    58
   35
Chiorotoluene isomer
                    (1)
Chlorotoluene isomer
                    (1)
  1120
   500
  747
  334
   313


   140
Unknown
       (3)
    (54)
        (2)
                                                     41
    Total Volatiles
  3532
                                                   2714
                                    453
(1)  Tentatively  Identified  Compound,   quantitated   using   an  assumed
     relative response factor of  one.

(2)  Identified  in  direct  injection   sample.    Values  used   in  the
     concentration totals/calculations.   All others by Method 8240.

(3)  Quantitated using an assumed  relative  response factor of  one.   The
     carbon  content  was assumed  to  be  75  percent   and   the  chlorine
     content  to be zero percent.
                                  H-394
-------
                                 TABLE V
                     SEMI-VOLATILES IN SAMPLE CMA-3
                    COMPOUNDS  DETECTED BY METHOD 8270
                                               CONCENTRATION (mg/L)
COMPOUND

Benzyl Alcohol
Butylbenzylphthalate
Di-n-oc tyIphthalat e
Phenol
Benzaldehyde
TributyIpho sphat e
       (2)
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
(2)
(2)
(2)
(2)
(2)
OBSERVED
190
2900
160
260
2.4
12
39
1.9
0.42
1.5
13
5.2
CARBON
148
2120
118
199
1.9
6.5
29
1.4
0.32
1.1
9.8
3.9
CHLORINE
...
	
	
	
	
	
	
	
	
	
	
	
    Total Semi-Volatiles
                                    3585
                                                       2639
(1)  Tentatively  Identified  Compounds,  quantitated  using  an  assumed
     relative response factor of one.
(2)  Quantitated using  an assumed relative response  factor  of  one.  The
     carbon  content  was assumed  to  be  75  percent   and  the  chlorine
     content to be zero percent.
                                   H-395
-------
                                TABLE VI
VOLATILES
COMPOUNDS DETECTED

COMPOUNDS
Benzene
S tyrene
Toluene
Xylenes
C -Alkene^
(1)
Diisopropyl Ether
C -Alkane(1)
7
C7-Alkane(-1-)
Propanoic Acid,
(1)
Methylpropyl Ester
( 1)
Chlorotoluene
(21
Ethyl Acetate
(3)
Unknown
(2")
2-Propyl Furan '
Propanoic Acid,
IN SAMPLE
BY METHODS

OBSERVED
1.24
1.84
2.92
2.99
4.68
7.80(14/
1.58

1.04
6.60(8.8)



0.78

7.9
42

70
36
CMA-4
8240 AND 8240DI
CONCENTRATION (rag/L)
CARBON
1.14
1.71
2.67
2.71
4.01
2) 9.88
1.33

0.87
(2) 5.78



0.52

4.31
31.5

53.5
20.7

CHLORINE
	
	
	
	
	
	
	

	
	



0.22

	
	

	

   Ethoxyethyl Ester
                    (2)
     Total Volatiles
196
141
0.22
(1)  Tentatively  Identified  Compound,   quantitated  using   an  assumed
     relative response factor of  one.
(2)  TIC  from   the  8240DI   analysis.    Used   in  the
     totals/calculations.   All others  by  Method 8240.
                            concentration
(3)   Identified in the  8240DI  analysis and  used  in   the   concentration
     totals/calculations.    Quantitated    using    an  assumed   relative
     response factor of one.   The   carbon content  was  assumed   to  be
     75 percent and the chlorine content  to be zero percent.
                                 H-396
-------
                                TABLE VII
                     SEMI-VOLATILES IN SAMPLE CMA-4
                    COMPOUNDS  DETECTED BY METHOD 8270
COMPOUND
                                             CONCENTRATION (mg/L)
OBSERVED
CARBON
CHLORINE
Acenaphthene
Acenaphthylene
Anthracene
Butylbenzylphthalate
Fluorene
2-Methylnaphthalene
Naphthalene
Phenanthrene
Pyrene
Unknown^-"-)
C6~Hydrocarbon(
1,2,3, 4-Tetrahydronaphthalene( 2)
1-Methylnaphthalene^ *•'
Unknown(l)
Cl4-Alkane(2)
C15-Alkane(2)
Cl7-Alkane(2)
Unknown(l)

    Total Semi-Volatiles
   0.67
   0.73
   0.20
   0.60
   1.20
  18
  27
   1. 40
   0.24
  30
  13
   7.9
  35
  13
   8.1
   9.1
   7.2
   7.1
  13

 193
  0.63
  0.69
  0.19
  0.44
  1.13
 16.7
 25.3
  1.32
  0.23
 22.5
 10.9
  6.61
 31.8
 12.1
  6.08
  7.72
  6.11
  6.03
  9.75

166
(1)  Quantitated  using  an  assumed  relative   response  factor  of  one.
     Carbon content was assumed  to  be 75 percent and chlorine content to
     be zero percent.

(2)  Tentatively  Identified  Compound,   quantitated  using  an  assumed
     relative response factor of one.
                                   H-397
-------
                                TABLE VIII
   COMPARISON OF  PROPOSED METHOD  25D  TO METHODS  8240,  8240DI (VOLATILES)
 PROPOSED  METHOD  25D  (METHANE  PLUS  CHLORINE)  VERSUS  8240,  8240DI (OBSERVED)
               CONCENTRATIONS  IN  PARTS PER MILLION BY  WEIGHT
                 PROPOSED METHOD 25D
                 (METHANE + CHLORINE)
CMA-3
  Radian                3470
  EPA Contract Lab      4400
  Average               3940
8240,  8240DI
 (OBSERVED)
  3530
 (3530)
 (3530)
(1)
(1)
         25D AS PERCENT
         OF 8240, 8240DI
         METHODS (C +C1)
 98
125
112
 (84)
(110)
 (98)
CMA-4
  Radian                 163
  EPA Contract  Lab       347
  Average                255
                            (2)
   196
  (196)
  (196)
(1)
(1)
 83      (87)
177     (185)
130     (136)
(1)   Measured  at  Radian  (Austin, TX)  only.
(2)   Single  analysis.   All  others  are  averages  of  three   analyses.
                                 n-398
-------
                                TABLE IX
                  COMPARISON OF PROPOSED METHOD 25D TO
     METHODS  8240,  8240DI  AND  8270  (VOLATILES  PLUS  SEMI-VOLATILES)
              PROPOSED METHOD  25D (METHANE  PLUS  CHLORINE)
                VERSUS 8240,  8240DI AND  8270 (OBSERVED)
              CONCENTRATIONS  IN PARTS PER MILLION BY WEIGHT
CMA-3
                PROPOSED METHOD 25D
                (METHANE + CHLORINE)
   Radian               3470
   EPA Contract Lab     4400
   Average              3940
   8240
8240DI, 8270
 (OBSERVED)
   7117
  (7117)
  (7117)
(1)
(1)
        25D AS PERCENT OF
        8240.  8240DI.  8270
        METHODS  (C + Cl)
49
62
55
(46)
(59)
(52)
CMA-4
   Radian                163
   EPA Contract Lab      347
   Average               255
                            (2)
    389
   (389)
   (389)
(1)
(1)
42
89
66
(40)
(85)
(63)
(1)  Measured at Radian (Austin, TX) only.
(2)  Single  analysis.   All  others  are  averages  of  three  analyses.
                                   H-399
-------
HI                               GROUND  WATER SAMPLING PROCEDURES NECESSARY
                                     TO OBTAIN DEFENSIBLE ANALYTICAL DATA

           Terr?  M.  McKee. Laboratory Director,  Craig McPherson,  Field Supervisor, G.  Shawn Meenihan,
           Service Coordinator,  Browning-Ferris Industries,  Houston Laboratory,  6630 Guhn Road, Houston,
           Texas  77040

           ABSTRACT
           Various types of sampling equipment and procedures necessary to obtain representative samples
           of ground water will  be discussed with emphasis placed on how to:
                 Prevent contamination of the well and well waters
                 Prevent cross  contamination of well and well waters
                 Prevent contamination of samples during collection, containerization and storage.
                 Obtain and record well  and field data
                 Initiate, maintain and  document groundwater sampling quality control (QC) procedures
                 Prepare sample labels and sample seals
                 Initiate, maintain and  document sample Chain of Custody
                 Provide adequate instructions to the analytical laboratory
           Utilizing the equipment and following the procedures discussed in  the  paper should result in
           pertinent, defensible  analytical  data from a  quality environmental laboratory.   Laboratory
           Chain of  Custody and  QA/QC will not be discussed.

           INTRODUCTION
           The purpose of a good groundwater sampling is to:
                  Obtain representative  samples of ground water.
                  Prevent contamination  of well and well waters.
                  Prevent croaa  contamination of wells and well waters.
                  Prevent contamination  of samples during collection,
                  containerization and storage.
                  Obtain and record proper well data.
                  Prepare proper sample  labels.
                  Maintain Chain-Of-Custody.
                  Provide adequate instructions to laboratory.
           Resulting in:
                  Obtaining meaningful and defensible groundwater data from a qualified laboratory.

           I. CONTAINER PREPARATION
           1)  The  laboratory performing the  groundwater analyses should supply  all  necessary coolers,
           pre-cleaned  containers,  trip  blanks,  field water  supply,  chemical preservatives,  packaged
           refrigerant, labels,  custody  seals, chain of  custody and  shipping  forms.   The field sampling
           log and  sample analysis request forms should  be provided  by  the  individuals or firm collect-
           ing the   samples.  It is highly recommended that the sample collection  team be employed by or
           employees  of  the  laboratory  responsible for  the  analytical  data.   One  firm,  preferably the
           laboratory, should be totally  responsible for all of the data concerning any sampling event.

           2)   The container needs to be constructed of a material compatible and non-reactive with the
           material it is  to contain.  Consult  Attachment  1,  Recommended  Containerization and Preserva-
           tion of   Samples. to determine  the number, type and volume of containers needed.  The contain-
           ers designated  may be  purchased  thru local container distributors with  the exception of the
           septum vial  and the  Teflon lined  caps  needed for  the  various analyses which are available
           thru  laboratory  supply companies.   Metal  lids should  not be  utilized.    Plastic  lids with
           polyethylene liners are acceptable  in many cases.   However,  Teflon lined lids are to be used
           for samples requiring organic  analyses.

           2.a)  Individual containers are not necessarily required  for  each  determination or test.  If
           two or more tests require the  same  container  and  preservation,  and a container of sufficient
           size is  available, the samples may be combined.

           2.b)  The cleanliness of the  containers, evacuating and sampling equipment  is most important.
           It is recommended that bottles and  lids  to  contain samples  for metals  or conventional  analy-
           ses be hand washed  with  a non-phosphate  detergent,  rinsed  in hot tap water,  rinsed with
           chemically pure or reagent grade nitric acid, rinsed  with distilled  or deionized water and
           let to air dry.  Glass  bottles used  to collect  samples for  the determination  of organic
           compounds by  GC,  GC/MS,  or HPLC  analysis should  be  washed with  a  non-phosphate detergent,
           rinsed with  hot tap  water,  rinsed  with pesticide grade hexane  and  methanol,  rinsed with
           copious   amounts  of  D.I.  water  (at least  six rinses),  and  kiln  baked  at 300°C.   Caps and
           Teflon liners should be prepared in the  same  manner,  except  without  the  kiln bake.  When the
           bottles   are  cool,  the  caps and  liners  completely dry, cap  the  bottles  and  store them in  a
           clean and dry environment.  Additionally, all equipment used to bail or sample a well must be
           cleaned  in the same manner prescribed for cleaning the caps and liners above,  and  stored in  a
                                                      H-400
-------
clean and  dry environment.  It  is  suggested  that clean bailers be  wrapped in foil or Kraft
paper for  storage.

3)  If  the sample locations and tests are known  prior  to  field collection, it may be desir-
able to label the  containers  and  add  the  chemical  preservatives  (where applicable) before
sampling.

II. BLANKS
1)  One complete set of trip,  field,  and equipment blanks  should be  prepared for each sam-
pling event,  or for  a  minimum  of  one  in ten  monitor  well and/or  field samples.   The  trip
blank samples  are to be prepared  in the laboratory by  filling the  appropriate clean sample
bottles with  laboratory  grade  carbon free deionized water with a specific conductance of 1.0
umhos/cm or  less that has  been passed through activated  carbon,  and  adding the appropriate
chemical preservatives  (if any)  as  indicated  in Attachment 1 for each type of sample.  These
bottles are then to  be  labeled "trip blank",  the analyses to be performed indicated on each,
and placed in the appropriate  sample  shuttle  cooler(s) to  be utilized  for sample transport to
the field  and back to the  Laboratory.  This procedure accounts  for any contamination that may
occur as  a result  of the  containers,  the sample  coolers,  the cleaning operations,  or  the
chemical preservatives.

l.a)  The  field  blank samples  are to be prepared  in  the  field  or  site at which the sampling
event is   taking  place.   Field  blanks  are to  be  prepared when dedicated  or non-dedicated
sampling equipment  is used in  the  sampling  project.  The field blank  samples are prepared in
the field  by filling the appropriate  sample bottles from the field supply of laboratory grade
carbon  free deionized water.   This  procedure  accounts for  any contamination that may occur as
a result of site  ambient air conditions,  and serves as an additional  check for contamination
in the  containers,  the  sample coolers,   the  cleaning operations,  or  the chemical preserva-
tives.

l.b)  Equipment  (rinsate)  blank  samples are to  be  prepared in the field immediately following
any  decontamination  cleaning  procedures  and  before non-dedicated  equipment  is  used  for
evacuation, sampling,  or  sample preparation.   Bailers,  funnels, filters,  and vacuum filter
pump vessels may be considered possible sources of cross contamination.   Following decontami-
nation,  laboratory  grade carbon free deionized water from the  field supply source is passed
through the non-dedicated  equipment in the same manner as the  sample  itself,  contacting  the
equipment  in question.   This  procedure  serves to  validate any  decontamination  procedures
carried out  on non-dedicated  equipment  utilized  in  the  field  or cleaned  in  the  laboratory
prior to field use.

III.  WELL  EVACUATION
1)  An  Organic Vapor Analyzer  (OVA) or  portable gas chromatograph (GC) should be  used to
determine  the presence  of total organic volatile  compound emissions prior  to  the  evacuation
of the  monitor  wells.   The measurements  are taken  immediately after  the well  cap  has  been
removed.   Historically  dry  wells  should also  be  tested  in the same manner.  All  necessary
calibrations and maintenance procedures found in the manufacture's operation manual are to be
followed.  Calibrations  and readings  are  to be  recorded on the applicable field sampling log
(see Attachment 2) for each well tested.

l.a)  If organic vapors are determined  to  be present at  the well  head,  a water-hydrocarbon
interface  probe  is to be employed  to determine  if an immiscible layer is present in the well
waters.   Depth to any light or dense immiscible layer(s),  and  thickness  of each layer should
be measured from a marked  measurement reference point (see  note  below) at the  top of the well
casing  to  the immiscible layer(s)  and through the immiscible layer(s).   If immiscible layers
are present, measurements  are  to be recorded  on the well's  field sampling log.

Note:   Each  well should have  a reference point  from which its water level  is taken.   The
reference point  should be  established by  a  licensed surveyor and is typically located at the
top of  the well  casing  with the locking cap  off.   The  reference point should be established
in relation to  mean sea level and  the  survey must also note  the  well location coordinates.
The device which is used to detect  the  water  level surface must be  sufficiently sensitive so
that a  measurement to + 0.01 foot can be obtained  reliably.

l.b)  Sampling of any immiscible layer(s) present  will precede  well  evacuation procedures.  A
bottom  filling  Teflon bailer  is lowered to  the  levels  at which  the light  or dense phase
immiscibles are  found and  a sample taken.  Care must be taken  to  gently lower the bailer to
avoid,   as  much  as possible,  disturbing  the  interface  between  the  hydrocarbon  and water
layers.

2)  Prepare equipment blanks by  passing  laboratory grade carbon free  deionized water  through
the purging  equipment,  if  possible,  then into  the appropriate  pre-labeled  container  for
transport to the Laboratory.   This  blank  is to  be prepared in the field, utilizing any clean
                                           11-401
-------
purging equipment,  prior to evacuating the wells.  These containers are to be labeled  "evacu-
ation equipment blank",  and placed in the coolers utilized to ship the samples to the  Labora-
tory.

3)  All equipment,  except that dedicated to the  well,  is  to  be  washed before evacuating each
well.  It should be detergent washed, rinsed with  tap  water,  deionized water, and chemically
pure hexane/methanol,  and allowed to air dry.   See Section I.2.b.

4)  Prior to sampling a monitor well, it should be flushed or evacuated no more than 24 hours
in  advance  of sampling.  A   maximum of one  day's time  is  suggested  to  allow  the  well to
properly recharge prior to sampling, and yet prevent the  accumulation of stagnant or  "spoil-
ed" water.  For rapidly  recovering  wells,  the  time between well  flushing  and sampling could
be considerably less.   At the time of sampling, there should be a net flow of formation water
into the well, increasing the  probability  that the samples consist  of  fresh formation water
only.  The requirements for flushing a well are as follows:
     -  For  a  low  yield or  slow recovering  well, evacuate  to  dryness.   If  time  permits,
additional evacuation is suggested.   Sample immediately upon recovery.
        Four  well volumes  should be removed  from a rapidly recovering well.    If  the  well
cannot be evacuated to  dryness,  it  is  recommended that pH and  specific  conductance  be moni-
tored  to  stabilization  during evacuation  to  assure that  a  representative  sample  of ground
water  is collected.  Sample as the well is recovering.

4.a)   In  order  to  prevent  well  contamination, either  of  the following  well flushing proce-
dures  are recommended for wells less than 26 feet deep:
        Bailers constructed of the same material as the well casing, PVC, Teflon  or stainless
steel  are recommended purging  equipment.   The  bailers  should be  attached  to a a down-rigger
reel with a polyethylene (PE) coated steel cable, single strand stainless steel wire,  or a PE
monofilament line.   The reel should be mounted on a tripod with the cable pulley  set directly
above  the well  opening.   The bailer  and  cable should  be  the  only sampling  equipment to
contact the  internal  well  casing and its  contents.  As  the  last bailer-full  is pulled,  the
cable  is  to be wiped with a  reagent  grade  methanol/D.I. water  saturated cloth.   If monofila-
ment  line  is used, it  may be  practical and a good  practice to  simply  provide clean, fresh
line at each  well.
        Evacuation  may also be  accomplished by  means of a  dedicated  non-collapsible  1/2 inch
discharge tube or pipe, preferably of Teflon or high density polyethylene,  placed inside the

well,  and a  portable  vacuum, peristaltic  or centrifugal  pump operating at  the well  collar.
In  an effort to collect all  the stagnant  water during draw-down,  the inlet of  the tube or
pipe must be  moved down as the well  draws  down,  maintaining  it  just  below  the surface of the
water.

4.b)    For wells deeper than 26  feet, one of the following flushing methods  is recommended:
        A  dedicated bailer constructed  of  the  same material as the well  casing, of  Teflon,
stainless steel  or PVC,  attached to  a down-rigger reel with  a polyethylene  coated steel
cable,  single strand stainless  steel  wire, or  a  monofilament  line  may be  used.   The reel
should be mounted as noted above in Section 3.a.
        A less  time  consuming  method  is   the  electrically powered Teflon/stainless steel
submersible pump,  or  a gas operated positive displacement (bladder  or piston) Teflon/stain-
less  steel  pump.   A dedicated submersible  pump, or  at least dedicated  discharge tubing, is
preferred.  The discharge tube and fittings should be of Teflon or high density polyethylene.
A dedicated  submersible electrical  pump should be  retained about a  foot above the bottom of
the  screened  area  to  prevent burn-out upon total evacuation,  and should not be operated dry.
If  a pump does not  evacuate  the  well to dryness. the inlet of  the pump should be placed just
below  the surface of the water column,  and lowered as the well draws down.


IV.  FIELD RECORDS   EVACUATION
1)  A  bound field book  containing numbered pages or a separate field log for each well should
be  maintained  by the  collector to record all pertinent information regarding the evacuation
and  sampling  of  monitor wells.  See  Attachment  2.  This recorded  information is necessary to
maintain  well sampling  data and should  become  part  of  the analytical report.  The sample
collector should sign and  date each page  of the field book or the  log.   The following data
should be determined as follows and recorded upon the evacuation of each well:
     a)  Collector's name,  date,  and time that evacuation was initiated and  completed.
     b)  Site and Location   Name of the facility,  city and state
     c)  Well Identification   i.e., monitor well number,  code or name.
     d)  Well Depth   Measure  from  the  reference point at the  top  of the  well casing to  the
     bottom of the well to the nearest  0.01 foot wit a weighted measuring tape  or a  calibrat-
     ed water  level  indicator  prior to  purging.   The  water level  indicator should be cali-
     brated and any correction factors  noted on the meter and  in the manufacture's  instruc-
     tion book which accompanies the meter.  See note  in Section  Ill.l.a.
                                           H-402
-------
     e)  Water Level  Depth   Measure from the  reference  point  at top of the  well  casing to
     the water surface  to the  nearest  0.01  foot with  a pre-cleaned calibrated  water  level
     indicator.  As the probe  and  line  are pulled from the wells, the lines  should be  wiped
     with a  fresh reagent grade methanol/D.I.  water saturated  cloth.   The probe  should be
     washed with a non-phosphate detergent and rinsed twice with  laboratory  grade  deionized
     water.
     f)  Depth to any  light  or dense  immiscible layer(s),  and thickness of each  layer
     Measure from the reference point at the top of  the well casing to the  immiscible lay-
     er(s) and through the immiscible layer(s) utilizing a water-hydrocarbon interface probe.
     If immiscible layers  are  present,  document the method of  collection  and identification
     of samples.   See Section III.l.
     g)  Well casing inside diameter to the nearest 0.01 inch.
     h)  The well volume   Calculate  the amount  of water  in gallons  occupying the well  prior
     to bailing.
               Volume (gallons) = (3.14) r2h/231.
               r = inside well casing radius in inches.
               h = height of water in well in inches (well depth minus the  water level
                   depth).
     i)  Total gallons evacuated   Well yield.
     j)  Water level following evacuation,  ft.
     k)  If organic vapors are  to  be measured in the well head, utilize a  OVA or a portable
     gas  chromatograph  and record  measurements.  Field  calibration must  be  conducted  and
     recorded.
     1)  Method of Evacuation   type of bailer,  pump,  etc. including description.
     m)  Comments    Information pertaining  to  the condition of the well, such as  no  cap,
     broken casing,  grout deterioration, etc.

V.  SAMPLING
1)  The wells  should have been properly flushed or  evacuated,  the containers and samplers
prepared and the initial log data entered upon evacuation.

l.a)  Determine the  water level depth  to  the nearest 0.01 of a foot    Record on  the  field
log.

l.b)  All non-dedicated equipment used  to  sample the well  (e.g., bailer, funnel, etc.)  must
be cleaned and stored as per the cleaning procedures outlined in Section I.

l.c)   Prepare equipment  blanks by passing  laboratory  grade  deionized water through  the
sampling equipment,  if possible, (and filtering  device,  if  applicable),  then  into the appro-
priate pre-labeled container  for  transport to the Laboratory.   This blank is to be prepared
in the  field,  utilizing the pre-cleaned sampling equipment,  but prior to  the collection of
samples.  These containers are  to  be labeled "sampling equipment blank"),   and placed in the
coolers utilized to ship the samples to the Laboratory.

l.d)  A dedicated positive displacement Teflon/stainless steel pump may be  utilized  to sample
the well.  When collecting samples  for the analyses of volatile constituents,  such as  vola-
tile organics, total organic carbon (TOC) or total organic halide (TOX),  pumping rates should
be minimized,  if possible.   No headspace  shall exist  in the  sample  containers  for  these
parameters.   A dedicated  discharge/peristaltic  pump  system may also  be utilized  to sample
groundwater monitoring wells,  but  is only recommended when sampling  for non-volatile compo-
nents .

l.e)  Samples may also  be  withdrawn from a well utilizing  a pre-cleaned or dedicated bailer
constructed  from  the same material  as the  well casing.  Teflon,  PVC,  or  stainless  steel
attached to a clean polyethylene coated steel cable,  single strand stainless steel wire, or a
monofilament line. The cable is raised  and lowered by means of  a down-rigger  reel mounted on
a tripod and set directly  above the  well opening.  The  first  bailer-full,  if  adequate sample
exists, should  be used  to rinse  the bailer  and discarded.    Subsequent  samples  should be
slowly  transferred from the bailer to  the container and preserved  immediately according to
the specific  test requirements (see Attachment  1).   Upon withdrawing the  last bailer-full,
wipe the cable with a clean cloth saturated with laboratory grade deionized water and reagent
grade methanol.  If  monofilament  line is  utilized, discarding  the used portion  of the line
may be more appropriate.

l.f)  Samples withdrawn from the well should  be  collected in  the following order if there is
sufficient volume:
          field pH,  Specific Conductance, Temperature
          Volatile organics
          Total organic halogen
          Total organic carbon
          Extractable organics
                                           n-403
-------
          Total recoverable metals
          Dissolved metals
          Wet chemistry parameters

2)  All samples collected for transport to  the Laboratory  should be chemically preserved  (if
applicable) and immediately placed  on wet ice.   See Attachment  1  for specific requirements.
Samples collected for dissolved metal  analyses should  be  immediately filtered through a 0.4B
micron glass fiber or membrane filter prior to transfer to the sample container and preserva-
tion.   All containers,  especially  those containing  samples to  be  analyzed  for  volatile
constituents,  should be  filled to  the top to maintain  anaerobic conditions  and to prevent
volatilization.

3)   The  following determinations should  be made in the  field  at  the time  of  sampling  and
recorded on the field logs:
        pH*     Specific Conductance*      Temperature     Calibration checks
     (* in quadruplicate)
Field monitoring  equipment  or test  probes  should  be  calibrated in the field prior  to each
sample  collection according  to  the  manufacturers specifications and in compliance  with  EPA
recommended methods.  Calibration checks  utilizing  standards  for pH and specific conductance
should be performed and documented at the time of sampling.

4)   Sample Shipment   Samples are to be shipped  in  sealed  insulated shipping containers,  ice
chests  or  coolers supplied by the  analytical  laboratory  conducting  the  analyses.   Shipment
and  receipt of  samples must be  coordinated with the laboratory  to  minimize time in transit.
All  samples for organic analysis  (and  many  other parameters)  should arrive at the laboratory
within  one day after  sampling  and  be  maintained  at  zero  to 4°C  with wet  ice  or  packaged
refrigerant ("blue  ice").  Wet  ice  should be replaced with frozen  packaged refrigerant just
prior to shipment.  To insure arrival at the laboratory in good condition,  the samples should
be  sent in sturdy  insulated ice  chests  (coolers)   equipped  with  bottle  dividers.    An  air
courier or equivalent over-night courier service  is  recommended.

VI.  FIELD RECORDS   SAMPLING
1).   It is most  important to maintain an accurate and  thorough field log in  case  one  is
required to recall particular information concerning the evacuation and sampling  of a monitor
well.   In  addition to the information to be logged covered in Section III and IV, the follow-
ing  information is also to be recorded at the time of sampling.  See Attachment 2.
     a)  Collector's name, date and  time of sampling.
     b)  Water Level Depth   Measure from the reference point at the top of casing to the
     water surface to the nearest 0.01 foot with a calibrated water level indicator.
     c)  Sample identification number.
     d)  Sample pH and specific conductance in quadruplicate,  and temperature.  See Section
     V.3.
     e)  Method of sample collection   type of bailer,  pump, etc.  Include a description of
     the pump   brand name, model, etc.
     f)  Sample characteristics   color, turbidity,  odor,  sediment,  surface oil,   etc.
     g)  Sample volume, containers, preservatives.
     h)  Test to  be performed on each sample (if  known).
     i)  Weather  conditions at the time of sampling.
     j)  Sample sequence number   Order in which  well was  sampled with respect to other wells
     on site.
     k)  Any additional field observations, comments or recommendations   e.g., split
     sampling (with whom), re-sampling, equipment failures, condition of the well, etc.
     1)  Sample Custody Statement   If the samples are  transferred to the receiving
     laboratory by the collector or an over-night common courier, a statement to  this effect
     should be noted on the field log.

2)   Prepare a sample label for each  sample  container employing a waterproof pen and adhesive
label.  The following is to be indicated on the  label
     a)  Collector's name, date and time of sampling.
     b)  Sample source.
     c)  Sample Identification number.
     d)  Sample preservatives.
     e)  Test(s) to be performed on the sample,  if known.

3)   The samples must  be  sealed to protect  their worth.   The  collector is  to date,  sign  and
identify each sample on a seal and attach  it to  each sample container and lid.  A waterproof
adhesive seal  and pen must be used.  The  sample  shuttle kit (cooler)  is sealed with a tamper
proof uniquely numbered  seal,  and this number  is recorded on the  chain  of  custody record.
See Attachment 3.
                                           E-404
-------
VII.  CHAIH-DF-CUSTODY
1).  Proper  Chain-of-Custody records are  necessary  to insure the  integrity  of the sampling
event and the analyses.

l.a)  A Chain-of-Custody  record (see Attachment  3)  shall be completed  for  all monitor well
samples collected by the sample collector.

l.b)  The sample  collector,  or the initiator, is to forward  the  original  with the sample to
the laboratory performing the analyses.

l.c)   Upon  receipt  of the  samples at  the  laboratory,  the  sample  coordinator  or  his/her
representative is to complete  the  record,  make  a  copy  for his files,  and return the original
with the analytical data to  the appropriate party.

VIII.  INSTRUCTIONS TO THE LABORATORY
1).  Written instructions to  the  analytical laboratory  must  accompany the  sample  to  avert
communication difficulties.  The instructions will be  recorded  on a form  such as  the  Sample
Analysis Request  Form, Attachment  4.  or  equivalent.   The instructions  should specifically
note the sample identification, date of  sampling,  and  analyses  to be performed.  The samples
must be delivered to the  laboratory as soon as  possible.   If the  samples cannot be delivered
immediately, they should be secured and  where necessary, maintained  at 4  C.   At no  time
should the  samples  be  delivered to the  laboratory for analysis  after  the permitted holding
times have expired.

2).  As soon as field  personnel are ready  to transport samples  from the field to the labora-
tory, they  shall  notify the laboratory  by telephone.   If the samples are  shipped by  common
carrier, the laboratory should  be  telephoned as soon as the shipping containers are coaigned
to the shipper, and provided with  shipping document numbers  .  In  addition to the estimated
time of arrival,  the field  personnel should provide  the following minimum information  to the
laboratory:
        Date of Shipment
        Time of Shipment
        Number of Containers Shipped
        Mode of Shipment
        Sample Type
     -  Source of Samples

3).  Some recommended specific  laboratory  requirements are as follows
        Maintain preservation of the samples  - refrigerate.
        Log-in samples -  record pertinent information, note the condition of samples,
     the sample shuttle and  sample  seals.
     -  Maintain Chain-of-Custody   External, and in-house or intra-lab Chain-of-Custody.
     -  Analyses  - perform analyses within prescribed holding time limits   record date and
     time of analysis.
          o  Identify methods of analyses.
          o  Use only methods acceptable to the involved regulatory agency.
        Employ good analytical  practices and  techniques,  such as:
          o  Clean glassware and analytical tools, e..g.,  pipettes, syringes, etc.  Use
          sulfuric-dichromate cleaning solution when applicable.
          o  Analytical reagent grade reagents  and certified standards.
          o  Distilled and/or deionized water with a conductivity of 1.0 umhos/cm or less,
          "organic free" where  necessary.
          o  Adequately trained, experienced  personnel, with special emphasis on laboratory
          safety.
          o  Adequate physical  facilities  and equipment.
          o  Frequent documented servicing and calibration of instruments.
        Maintain a quality assurance/quality  control program.  The quality assurance/quality
     control program must include the following minimum components:
          o  Calibration of laboratory instruments to within acceptable EPA and
          manufacturer's limits.
          o  Inspection, maintenance, and  servicing of all laboratory instruments and
          equipment.
          o  Use of reference standards and quality control samples (blanks,  spikes,
          duplicates, etc.).
          o  Use of thorough, documented QC procedures to monitor accuracy and  precision of
          data.
          o  Regular participation  in external laboratory evaluations (i.e. EPA Performance
          Audit Program).
          o  Continuous in-house training  program.  New analysts  must become  thoroughly
          familiarized with all laboratory safety procedures and  equipment.
          o  Maintenance of laboratory notebooks for each analytical method and copies of all
                                           n-405
-------
          analytical reports.   All raw data produced must be checked for validity before
          reported and permanently stored.

IX.  DATA REPORTING
1)  All analytical reports will  be  complete with analytical data,  sample  ID,  sample source,
date sampled,  date received,  parameters tested,  results,  percent recovery,  date extracted  (if
applicable) and analyzed,  analyst, referenced methodologies, QA/QC data, field logs, analysis
request forms,  and chain-of-custody forms.

2)  The report should be tamper proof bound,and include the following information:
        Title  Page   The title page should  include the site/project name, date of report,
     date(s) the samples were  received, the laboratory's  name and address,  and the laboratory
     supervisor's name and signature.
        Analytical Methodologies   A table  should be prepared listing all the analytical test
     method employed in the analyses of the samples with  a reference for each made to the
     method manual and test method.
        Field  Logs and Sample  Analysis Request Forms   These forms list pertinent field
     information and all requested analyses.
        Chain-of-Custody   Both field and laboratory chain-of-custody records should be
     included  in the final report package.   This should include a summary of sample movement
     through the laboratory with date(s)  of receipt,  date(s) of analysis, and date(s) of
     sample storage/disposal.
        Analytical Data   including sample  source,  date sampled,  date received,  parameters
     tested, results with the  appropriate units  of measurement,  data extracted (if
     applicable) and analyzed,  and the analyst's initials or name.
        All applicable QC Data not mentioned above   for  example,  laboratory blank data,
     internal  standard and surrogate recoveries,  chromatograms,  and tuning  data.
                                          n-406
-------
                    ATTACHMENT 1
RECOMMENDED CONTAINERIZATiaN AND PRESERVATION OF SAMPLES
Volume
Required
Measurement mL
Physical Properties
Color
Conductance
Hardness

Odor
pH (Per Replicate)
Residue
Filterable
Non-Filterable
Total
Volatile
Settleable Matter
Temperature
Turbidity
Metals ^except mercury)
Dissolved

Suspended
Total
Total Recoverable
Mercury-Dissolved

-Total
Chromium
(Hexavalent)

50
100
100

200
60

200
200
200
200
1000
1000
100

BOO

500
500
600
300

300

200
Container1

P. G
P. G
P, G

G only
P, G

P. G
P. G
P. G
P. G
P, G
P, G
P, G

P, G

P, G
P, G
P, G
P. G

P. G

P. G
Inorganics, Non-Metallics
Acidity
Alkalinity
Boron
Bromide
Chloride
Chlorine
Cyanides

Fluoride
Iodide
Nitrogen Ammonia

Kjeldahl. Total

200
200
100
200
200
200
BOO

BO
100
400

600

Nitrate plus Nitrite 200

Nitrate
Nitrite
Dissolved Oxygen
Probe
Winkler
Phosphorus
Ortho-phosphate ,
Dissolved
Hydrolyzable

Total

Total, Dissolved

Silica
Sulfate
Sulfide


Sulfite

100
60

300
300
100

100

100

100

60
100
250


100
P. G
P, G
P only
P, G
P. G
P. G
P. G

P
P. G
P. G

P. G

P. G

P. G
P. G

G only
G only
P, G

P, G

P. G

P, G

P only
P, G
P. G


P. G
Holding

Preservative Times Reference

Cool, 4°C 48 Hrs
Cool, 4°C 28 Days
Cool. 4°C 6 Mos
HN03 to pH <2
Cool, 4°C 24 Hrs
None Det. on Site

Cool, 4°C 7 Days
Cool, 4°C 7 Days
Cool, 4°C 7 Days
Cool, 4°C 7 Days
Cool, 4°C 48 Hrs
None Det. on Site
Cool, 4°C 48 Hrs

Filter on Site 6 Mos
HN03 to pH <2
Filter on Site 6 Mos
HN03 to pH <2 6 Mos
HN03 to pH <2 6 Mos
Filter on Site 28 Days
HN03 to pH <2
HN03 to pH <2 28 Days

Cool, 4°C 24 Hrs

Cool, 4°C 14 Days
Cool. 4°C 14 Days
Cool, 4°C 28 Days
None 28 Days
None 28 Days
None Det. on Site
Cool, 4°C 14 Days
NaOH to pH >12
None 28 Days
Cool, 4°C 24 Hrs
Cool, 4°C 28 Days
H2S04 to pH <2
Cool, 4°C 28 Days
H2S04 to pH <2
Cool, 4°C 28 Days
H2S04 to pH <2
Cool, 4°C 48 Hrs
Cool, 4°C 48 Hrs

None Det. on Site
Fix on Site 8 Hrs
Filter on Site 48 Hrs
Cool, 4°C
Cool. 4°C 28 Days
H2S04 to pH <2
Cool, 4°C 28 Days
H2S04 to pH <2
Filter on Site 24 Hrs
Cool, 4°C
Cool, 4°C 28 Days
Cool. 4°C 28 Days
Cool, 4°C 7 Days
2mL zinc acetate
plus NaOH to pH >9
None Det. on Site

1
1
1

1
1

1
1
1
1
1
1
1

1, 2

1, 2
1. 2
1. 2
1, 2

1. 2

1, 2

1. 2
1. 2
1
1. 2
1. 2
1. 2
1, 2

1, 2
1
1. 2

1, 2

1, 2

1, 2
1, 2

1, 2
1. 2
1. 2

1. 2

1. 2

1, 2

1. 2
1, 2
1. 2


1
                       H-407
-------
Measurement
                   Volume
                  Required                            Holding
                    mL      Container1  Preservative   Times   Reference
Coliform, Total
and Fecal
Gross Alpha, Gross
Beta, Radium
BOD
COD
Oil & Grease
(One Replicate)

Organic Carbon


Phenolics

MBAS (Surfactants)
TOX (2 Rep)
(4 Rep)

Drganica
Volatile Organics
by GC (2 vials

100

4000
1000
50
1000


100


1000

1000
600
1000


100
« 40mL)
Sterile
P. G

P, G
P. G
P, G
G only


G only
Teflon
lined cap
G only

P. G
G only
Teflon
lined cap

G. Teflon
septum cap

Cool,

HN03
Cool,
H2S04
Cool,
H2S04
pH <2
Cool,
H2S04
pH <2
Cool,
H2S04
Cool,
Cool,
H2S04


Cool,
Cool.

4°C

to pH <2
4°C
to pH <2
4°C
or HC1 to

4°C
or HC1 to

4°C
to pH <2
4°C
4°C
to pH <2


4°C
4°C.

6

6
48
28
28


28


28

48
7



7
14

Hours

Mos.
Hrs
Days
Days


Days


Days

Hrs
Days



Days
Days




1
1
1


1


1

1
1



2


3

3
,
.
,


,


,

,
,



,





2
2
2


2


2

2
2



3

HC1 to pH <2
Volatile Organics
by GC/MS (2 vials
100
0 40mL)
G. Teflon
septum cap
Cool,
Cool,
4°C
4°C,
7
14
HC1 to pH <2
Phenols by GC


Benzidines by GC




Phthalate Ester by GC

1000


1000




1000

G, Teflon
cap liner

Amber G,
Teflon cap
liner
zero head-
space
G, Teflon
cap liner
Cool,


Cool,


4°C


4°C


prepare
7
30
40
7


7
oxidant free
Cool,

4°C

zero he ad space

Nitrosamines
by GC



Organochlorine
Pesticides/PCBs
by GC
Nitroaromatics
and Isophorone
by GC
Polynuclear Aromatic
Hydrocarbons
by GC
Organopho sphorus
Pesticides by GC
Haloethers
by GC
Chlorinated
Hydrocarbons
by GC
Drganics
Chlorinated
Herbicides by GC
Semi-Volatiles
by GC/MS


1000




1000


1000


1000


1000

1000

1000



1000

2000



Amber G,
Teflon cap
liner
zero head-
space
G, Teflon
cap liner

G, Teflon
cap liners

Amber G,
teflon cap
liners
G, Teflon
cap liners
G. Teflon
cap liners
G, Teflon
cap liners


G, Teflon
cap liner
G, Teflon
cap liner


Cool,



4°C


prepare oxi-
dant
Cool,


Cool,


Cool,


Cool,

Cool,

Cool,



Cool,

Cool,
free
4°C


4°C


4°C


4°C

4°C

4°C



4°C

4°C
7

30
40
7


40

7
30
40
7
30
40
7
30
40
14
30
7
40
7
30
40

7
30
7
Days
Days

Daysb
Daysd
Daysc
Daysb


Daysc

Daysb

Daysd
Daysc
Daysb


Daysc

Daysb
Dayad
Daysc
Daysb
Daysd
Daysc
Daysb
Daysd
Daysc
Daysb
Dayad
Daysb
Daysc
Daysb
Daysd
Daysc

Daysb
Daysd
Davsb



2







2








2


2


2






3
2

,
2
3
3


3

t

2
3
3


3


2
3

2
3

2
3
2
2
3
3
2.






40 Daysc


14
Days"

40 Daysa
2
3

2
2
3
3
2
2



3







3








3


3


3






3









                                   n-408
-------
          NOTES:
a   Plastic (P) or Glass (G).  For metals, polyethylene with an all
     polypropylene cap is preferred.
b - Maximum holding time from sampling to extraction.
c - Maximum holding time from extraction to analysis.
d - Maximum holding time from sampling to analysis.

REFERENCES:
1 - Methods for Chemical Analysis of Water and Wastes. March 1983. USEPA.
     600/4-79-020 and additions thereto.
2 - Test Methods for Evaluating Solid Waste. Physical/Chemical Method,
    November, 1986, Third Edition, USEPA, SW-846 and additions thereto.
3   "Guidelines Establishing Test Procedures for the Analysis of Pollutants
     Under the Clean Water Act", Environmental Protection Agency, Code of_
     Federal Regulations (CFR), Title 40, Part 136.
 Rev:  2/89
                                       n-409
-------
                                                  ATTACHMENT 2
                                                    FIELD LOG
Facility
                                                              Location
Collector/Operator

Sample Point ID 	
                                                              BFI Lab No.
                                                                                         Shuttle  No.
EVACUATION:   Date/Time

 Water Level Depth,  Ft.

 Casing Diameter,  in.
                                 Method of evac.

                                 Well Depth,  Ft
                                 Well Volume, gal
 Well level after evac., ft.	

IMMISCIBLE LAYER:  Detected  ( ) Yes ( ) No ( ) N/A
                                 Total gallons evac.

                                 Completed-Date/Time
 Sample Collected
( )  Yes (  )  No (  )  N/A
                     Depth to top of layer(s). Ft.

                     Depth to bottom of layer(s), Ft.
ORGANIC VAPORS:  Detected  ( ) Yes  ( ) No  ( ) N/A   Method of Detection
  Concentration                                   Cone, of Calib.
  measured, ppm 	 as 	  Standard, ppm •	
                                                      Amt of Calib.
                                                      Stdn found, ppm
SAMPLING:  Collector
 Initiated
 Date/Time 	
 Completed
 Date/Time
                  Sample Sequence
Well
Stick-up, Ft.	
Method of
Sample Collection
                                          Water Level
                                          Depth, ft _
PARAMETERS:~l  ) annual~(  ) semi-annual (  ) quarterly ~(  ) monthly (  ) other

REPLICATE  SAMPLE DATA:  pH and Conductance are automatically temperature compensated at  time  of  measurement
 Temp. Deg C

 Temp. Deg C

 Temp. Deg C

 Temp. Deg C
    pH 0 25

    pH 0 26

    pH « 26

    pH 0 25
  Instrument Calibration Check Data.
 pH  4 stdn: 	
 100 umhos
 Cond. stdn C 26:
                          Cond. fl 26, umhos/cm:

                          Cond. C 26, umhos/cm
                          Cond. 0 25, umhos/cm_

                          Cond. 0 26, umhos/cm
         pH  7 stdn: 	
         1000 unhos
         Cond. stdn 0 26:
                                     pH 10 stdn: 	
                                     10000 umhos
                                     Cond. stdn 0 26:
GENERAL INFORMATION:
Weather Conditions at time of sampling:
 Sample Characteristics:
 Sample Containers, Volumes, Preservatives and Tests to be Performed:
Comments and Observations:
Recommendations:
                                                                  Certification:
rev 03/90
                                                                          Signed
                                                                        Date
                                                     H-410
-------
                                        ATTACHMENT 3

                                      CHAIN-DF-CUSTODY
Shuttle
Number:
Seal
No.
Prepared/
Sealed By:
                                                              (.print name;
Laboratory:
                                                              (.signature;
SHIP TO
Company:

Address:
                              Attn:
                              Phone:
SAMPLE IDENTIFICATION
Facility/Site:
SHUTTLE CONTENTS
LAB
I.D.











Bite Sample
Source I.D.











# of
Bottles Size




















p*
G*






Preser- j
vative ' Parameter(s)
1
1
1
1
|




1


.







Container"type: F* = Plastic.  G*  = Glass
(1)

(2)
(3)

Shuttle
(.print
opened by:
name;
Shuttle prepared
shipment by:
(.print
Shuttle
by:
Sprint
name;
received
name;
	
for

at Lab
	
Date
Time
Date Time
1
1
I
Date
Time
Seal
No.


New
No.


Seal
No.
Lab""
Intact


Seal Installed


Intact
s Name :

Yes No
(.signature;
Yes No
(.signature;
Yes No

(.signature;
DP/dp
Rev: 12/89
                                          n-4ii
-------
Assigned to
Laboratory:_
               ATTACHMENT 4

SAMPLE ANALYSIS REQUEST / CHAIN OF CUSTODY

                                BFI Shuttle
                                Kit ID:
                                BFI LAB
Location: Proiect ID:
Lab Laboratory
Phone: Contact:
BFI No(s):













Sample
Point













Matrix
code













Date
Sampled













volume/
container













preser-
vative













Analysis (.OodesJ
Requested













Matrix Codes: bW = groundwater;  br = surface water;  oU = soil; MH = multi-phase;
OR = organic/oil; SW = solid waste;  WW = wastewater;  L = Leachate. Other Codes:
MT = tot. metal; MD = dissol.  metal; MR = tot.  recoverable metal
P.O. No:
Remarks :
Safety
Precautions :
Sample Custody
Statement:
Special Handling/
Storage :
Quote Per:

Ilormal
Laboratory Hygiene,
$

Avoid skin/
eye contact,
All samples properly preserved,
iced and hand delivered to

Receiving Lab's
Comments : (Headspace
Date

Avoid breathing
vapors/dust,

, etc.)
Sample(s) submitted by:
Custody Seal Intact:    yes
     /  no
Sample(s) received by:
Custody Seal Intact:   yes  /  no
Name
Title
                                           Name
                                              Title
Company
Date/Time
                    Company
                          Date/Time
REPORT RESULTS TO:
PROJECT MANAGER.  ______
Browning-Ferris Industries
6630 Gufin Rd.
                         SEND INVOICE(S) TO:
                         ACCOUNTING MANAGER
                         Browning-Ferris Industries
                         5630 Gufin Rd.
                         Houston. Texas 77040
Houston, Texas 77040                            .__.__,,„_
Return a copy of this signed form after receiving samples and indicate lab project IDs

Project IDs:  	
Rev:  11/89   ~~	
                                       0-412
-------
112               TECHNIQUES & QUALITY CONTROL IN GROUNDWATER SAMPLING

            Frank  Perugini, Manager-Sample Management and
            Frank  H.  Jarke, Manager-Quality Programs,
            WMI-Environmental Monitoring Laboratories, Inc.,
            2100  Cleanwater Drive, Geneva, Illinois  60134

            INTRODUCTION
            The   basic  objective of a groundwater monitoring program at a
            land   disposal facility  is  to determine whether  or not the
            facility   has impacted the  groundwater.  Federal,  state  and
            local  regulatory agencies have established criteria that must
            be  met at  each  facility.  These criteria  involve standards
            that   the  groundwater  samples  must  meet  with  respect to
            chemical   concentrations.   In  order  to  comply with  these
            criteria,  an effective and comprehensive  monitoring program
            must  be established for the facility.

            The   quality of  the data  collected at  these land  disposal
            facilities    is  dependent  upon  the   following  six  major
            activities:

                 1.    Define the Geology and Hydrogeology.
                 2.    Groundwater Monitoring System.
                 3.    Define the Analyte Requirements.
                 4.    Select Dedicated Monitoring Equipment.
                 5.    Establish Proper Sampling Procedures.
                 6.    Establish   Chain-of-Custody   Record    and   Field
                      Information Documentation.

            Each  of these activities are of equal importance adherence to
            the   six   items  listed  above  will  lead  to  a  successful
            groundwater   monitoring program with quality assurance at the
            time  of sample collection.

            1.  DEFINE  GEOLOGY AND HYDROGEOLOGY

            Hydrogeological  investigation  is  an  attempt   to  identify
            geological    structures  and  characterize  the   movement  of
            groundwater   within a  specific  area.  The first  step is to
            describe    the  regional  and  site  topographic  conditions.
            Topographical maps are available from a number of sources and
            in  a  variety  of scales.  The U/S.G.S.  is widely  used and
            provides  maps covering a quadrangle range bounded by lines of
            latitude    and  longitude.   The  U.S.G.S.   quadrangle  maps
            indicate   ground  surface  contours,  surface  water  bodies,
            building   sites,  cemeteries,  and  private  well  locations.
            Other  information may  be  obtained from local   town offices
            which  will   identify   surrounding  land   uses  such   as
            residential,   commercial,   agricultural   or  recreational.
                                  n-4i3
-------
Defining  the   hydrogeology  at  the   facility will   provide a
clearer  understanding of the potential migration pathways to
the  groundwater.    Therefore,   development  of  a  conceptual
model  can  be  accomplished  in  two phases  -  an  initial
reconnaissance   and   a   field  investigation.     When  the
hydrogeology  of a project  area is  relatively  uncomplicated
and   well  documented   in  the  literature,  the    initial
reconnaissance may provide sufficient information to identify
flow  paths and the target monitoring zones.  However, where
little  background  data  is  available   or  the  geology  is
complicated,  a field investigation  is  necessary.   The field
investigation   routinely  involves   performing numerous  soil
borings,   soil  samples  and  installation   of  piezometers
followed by monitoring wells in  the  target monitoring zones.


2.  GROUNDWATER MONITORING SYSTEM

Before  samples can be collected,  an overall summary of  the
monitoring  wells should be identified.   Table 1 is  a typical
chart   identifying  the  wells    used  in   the   groundwater
monitoring program and summarizes construction and  monitoring
information.   Geological boring  logs  and  well construction
logs  for each monitoring well would  also  be  attached to Table
1.
 3.  DEFINE THE ANALYTE REQUIREMENTS
One  important component of any  groundwater  monitoring  program
is defining the analyte requirements.   The  primary  source  for
 this  information is the  facility  permit.   Facility
 usually  contain details concerning the  frequency  of
 required,  a list  of analytes   to  be
 requirements,   the  reporting   limits
 required  and in  some  cases the test
 used.   The following  list  should   be
 analyte requirements for a groundwater
                                                      permits
                                                     sampling
                                       tested,  the  reporting
                                        or  detection  limits
                                        methods that must  be
                                       used  in defining  the
                                       monitoring program.
               o Determine  the  list  of  analytes  required.

               o  Identify  holding   times,  preservative   and
                 sample volume  requirements.
                 Define any  additional
                 report due  date.
                                       QA/QC requirements and
 Determining  the List of Analytes - There  is a wide variation
             facilities permits  specify the  analytes to  be
               permits will   provide tables that  contain the
in  the
tested.
way
  Some
                       E-414
-------
                                                                   CROUNDWATF.R MONITORING SYSTEM SUMMARY
                                                                                                                                                                   52
                                                                                                                                                                   co
                                                                                                                                                                   t-
                                                                                                                                                                   m
SITE:  Example
( OOP )
DATE:
                                                     SAMPLE  POINT TYPE,  LOCATION,  AND SAMPLING EQUIPMENT

WHI
Well
ID No.
PW19
B07D
BIB
SOI









Former
Well ID (if
different
fro. WMI)
PW-19
B-7D
B-18
S-l









Type
Agency
Well ID (if
different
from WMI)
19
7D
18
SI









Active/
Inactive/
De.coamj.
(A/I/D)
SEE
WELL
ID
CHART1
Location Sampling Equipment
iource
Code
W
W
W
R









Program
Type
S
P, S
S
S









Singling
Frequency
Q
S. Q
NA
S









On-Site
Off-Site
Off
On
On
Off










SE
NW
SE
SE









N/S Coord.
F./W Coord.
808 110N
214600E
808000N
2U7000E
80821SN
2147500E
808100N
2146500E









Formation
at Screen
Outwaih
Out wash
Till
NA









Gradient
SEE
WELL
ID
CHART2
Purge and
Sample
Equipment
NA/GR
W/WPHOO
NA
NA/GR









Filtering
Equipment
NA
IN
NA
NA









Sample
Tubing
Material
NA
PP
NA
NA









-------
                                                                                                                                                                      w
                                                                                                                                                                      I-1
                                                                                                                                                                      m
SITE:  Example
( 000 )
                                                                GROUNDWATER MONITORING SYSTEM SUMMARY (continued)
DATE:
                                                                    WELL CONSTRUCTION INFORMATION
Veil
ID
to.
PWI9
<07D
US
501








Well
Depth
Jt«-L_
SEE
WELL
ID
CHART-1
Bottom
of Well
Elev.
(ft MSL)
UK
650.23
690.63
NA








Ground
Elev.
(ft MSL)
UK
705.23
712.63
NA








Internal
Casing
Material
S
PVC
PVC
NA








Internal
Casing
ID (in.)
8.0
2.0
2.0
NA








Top of
Internal
Casing
or Well
Wizard
Cap
(ft MSL)
SEE
WELL
ID
CHART4



Bottom
of
Internal
Casing
(ft MSI.)
UK
645.33
685.63
NA








Tnterna 1
Casing
Length
UK
53
20.0
NA








Internal
Casing
Stick up
(ft)
UK
3.0
3.0
NA








Top of
Screen
Kiev.
(ft MSL)
UK
645.53
, 685.53
NA








Bottom
of
Screen
Elev.
Jft MSL)
UK
650.33
690.63
NA








Screened
Length
(ft)
UK
5.0
5.0
NA








Screen
Material
UK
PVC
PVC
NA








External
Casing
Material
NA
A
S
NA








External
Casing
ID (in.)
NA
4.0
4.0
NA








Comments












-------
SITE:  Example_
( OOP )
                                                            GROUNDVATER MONITORING SYSTEM SUMMARY (continued)
                                                          WELL CONSTRUCTION INFORMATION (continued)
DATE:
                                                                                                                                                                  CD
                                                                                                                                                                  r1
WMI
Well
ID
No.
PW19
B07D
B18
SOI








Top of
External
Casing
(ft MSL)
UK
708.73
716.10
NA








Bottom
of
External
Casing
(ft MSL)
UK
704.73
UK
NA








External
Casing
Stick up
(ft)
UK
3.5
3.5
NA








Cons.
to WMI
Spec.
(Y/N)
N
Y
N
NA








Cons.
Date
UK
5-24-
85
6-23-
83
NA








Drilling
Firm
UK
OK
Drillers
NG
Drillers
NA








Drilling
Method
UK
HSA
WRM
NA








Devel .
Method
UK
AS&A
UK
NA








Packing
Material
UK
S
S
NA








Packing
Length
UK
10.0
5.0
NA








Length
of
Filter
Sand
Above
Packing
(ft)
UK
2.0
2.0
NA








Length of
Bentonite
Seal
(ft)
UK
3.0
NA
NA








Length
of
Filter
Sand
Above
Seal
(in.)
UK
6.0
NA
NA








Type
of
Grout
UK
Volclay
0
NA







i
Grout
Length
UK
41.5
17
NA








Comments












-------
                                                                                                                                                                  H

                                                                                                                                                                  00
                                                                                                                                                                  r-
                                                                                                                                                                  n
SITE:  Example
(  OOP )
GROUNDWATER MONITORING SYSTEM SUMMARY (continued)





            WELL [0 CHART


                        Normal Range
DATE:
WMI
Well
ID No.
PW19
B07D
R1R
SOI









1 Active/
Inactive/
Decomm.
(A/I/D)
A
A
D
A









2
Gradient
D
U
D
D









3 Depth
of
Well
(feet)
60
58
25
NA









4 Elevation
at Top
of Casing
(msl)
UK
708.24
715.63
NA









Purge
Volume
(gal Ions)
NA
8. 15
NA
NA









Depth to
Water
(feet)
UK
5.0
NA
NA









Recharge
Time
(hrs)
NA
3.0
NA
NA









Temp .
(°C)
15°
10°
NA
2-25°









pH
(Std.)
7.0-7.3
6.8-7.3
NA
7.3-7.5









Specific
Conduct .
(pmhos)
at 25°C
340-345
850-960
NA
1000-1500









Comments
Private well; sample taken at
outside faucet

Well grouted 8/84










-------
specific analytes,  reporting limits or detection limits to be
achieved and in some cases the analytical methods to be used.
Other  permits  simply  indicate the  analytes or  classes of
analytes  and allow the permittee discretion on methods to be
used and what reporting limits are acceptable.

If  specific methods  are  not specified, it  is advised that
40CFR Part 136 and the tables contained in this regulation be
used  in  determining  the  most  appropriate  method  to  be
followed.   Tables  are  included for  biological, inorganic,
organic and radiological analytes.  The methods specified are
widely used and understood by the analytical community.

Reporting limit requirements vary widely from state to state.
In  the absence of specific  requirements in the permit,  the
Contract  Required Detection Limits (CRDLs)  specified in the
USEPA  Contract Lab Program  (CLP) or Practical  Quantitation
Limits  (PQLs)  are  recommended.   One  should  consult  the
laboratory  to  be  used  for  guidance.   If  the laboratory
personnel  are not familiar  with these concepts  a different
laboratory should be considered.

Identify   Holding  Times,  Preservative  and  Sample  Volume
Requi rements   -   Many  permits   will  not  specify   these
requirements.    In  setting  up  a   groundwater  monitoring
program,   these  three  factors  should  be  determined  and
discussed  with the laboratory personnel to insure that there
are no misunderstandings after the event.

Table  II of 40CFR Part  136 specifies precise holding  times
and  preservation requirements for most common analytes.  The
most  important part of this table are  the notes that follow
it.  Particular attention must be paid to differences between
certain  analytes that  may  or may not  be analyzed together
depending  on the preservation  and holding times  used.  For
example,    purgeable    halocarbons,    purgeable   aromatic
hydrocarbons, acrolein and acrylonitrile can only be analyzed
together  from the same bottle, if no preservative other than
cooling to 4°C is used and the analysis is completed within 3
days  after sampling.  If acid preservation is used to extend
the  holding time, then these  compounds must be analyzed  in
three separate analytical tests and three separate containers
must be used.

Additional  Special Requirement -   Examples of the  types of
data  that may  be  included as QA/QC  data are results  from
specific  samples included in the batch of samples run during
one shift or on one analytical instrument.  These may  include
calibration  standards, lab blanks, matrix  spikes and matrix
spike  duplicates and  blank  spikes.   Each of  these  samples
provides  information that  is  useful  in determining   if the
analysis  met  the  requirements  of   accuracy  and  precision
                      n-4i9
-------
usually  published with  the  method.  Additional  information
that  may be required by a facility's permit  include  the date
of  analysis,   the  name of  the analyst,  chronicles showing
which  samples were rerun because of  matrix  interferences or
dilutions.

Some permits require additional QA/QC data be  submitted along
with   the   analytical  results   and/or  be   available   for
inspection   at the facility.  The laboratory  should  know what
additional  QA/QC is required prior to analysis and what is to
be  reported  to  the  client  with the  sample  results.   NO
ANALYTICAL   PROCEDURES SHOULD BE CONDUCTED IN  THE ABSENCE OF
QA/QC.

The  laboratory  should  be  aware  of the  regulatory agency
report  due dates.  This generally is expressed  in days after
the  sampling  event  or in  some cases  after completion  of
analysis.   In either case, the laboratory must  be  held to a
specific  due date to insure that ample  time  is available to
review  the  results  prior to  submission to  the regulatory
agency.   Penalties  for  failing  to  meet   the   agreed upon
"turnaround  times" should be agreed  to in advance  with  the
laboratory.

4.  SELECT DEDICATED MONITORING EQUIPMENT

Proper  selection  of  monitoring equipment   is  an  important
factor  in controlling sampling  quality and  sampling  costs.
Dedicated groundwater sampling systems are an  example of cost
effective  monitoring equipment  which  provides high quality
samples, while drastically reducing labor costs.

Groundwater   monitoring  focuses  on  very    low  levels  of
contaminant  concentration and slight changes  in physical and
chemical  properties.   As  contaminant measuring  techniques
advance  and  detection  levels  of  part per  billion become
common, consistent sampling procedures and reliable  equipment
are needed  to provide useful, accurate and high  quality data.
Sampling  procedures which  allow exposure  to air,   airborne
particulate  matter,  temperature  changes  and  contact with
contaminated  surfaces can  significantly  alter the physical
and chemical properties of a sample.

The    extended  length  of~ monitoring    required  at  modern
facilities  also  place  a  heavy  burden  on the monitoring
program,   from  the standpoint  of both sampling  and quality.
Over   many  years  of  monitoring,   changes  in  pe-rsonnel and
sampling    procedures  may  combine  to   introduce  erroneous
results  to the data base,  thereby jeopardizing  the  integrity
of  the  facility's  data   base.   The   sampling  methods and
equipment   should be selected with the   objective  of limiting
the  potential for variability over the monitoring  period.
                       n-420
-------
Based on the needs  of the groundwater monitoring program, the
equipment  selected for use  should provide for:  1. accurate
sampling,  2. ease  of use, 3. ease of set up and calibration,
4.  rugged to withstand continued  use and 5. INDEPENDENT  OF
OPERATOR TECHNIQUES.

Dedicated  bladder  pumps provide an  advantage by eliminating
air/water  contact, no loss of  volatile organics due to  air
stripping  and no  change in sample  temperature.  The bladder
pump  is  permanently  dedicated  in  the  well,  at a  fixed
position  within the  well  screen.  Water enters  the bladder
via  the bottom  check  valve assembly.  When  the bladder is
full,  compressed  air  is  delivered via an  air line to  the
space  between the  body and  the bladder, which squeezes  the
bladder.   This  action  closes  the  bottom check  valve and
forces  the water   contained  inside the bladder  through the
upper check valve  and into the attached water discharge tube.
After the bladder  is  fully evacuated, the pressure inside the
pump  is released  back through  the air line to  the surface.
This  action causes the upper check valve to seat (close) and
the  bottom  check   valve  to  open,  refilling  the bladder.
Continuous operation  of this pump relies upon alternating the
compression/release cycle.

5.  ESTABLISH PROPER SAMPLING PROCEDURES

The  objective  of   a  groundwater  sampling procedure  is to
obtain   a   representative   sample.   The   techniques  for
collecting   a  representative  sample  should   include  the
following:

               o Establish purge volume.
               o Consistent sampling procedures.
               o Filtration procedures.
               o Sample preservation and shipment.

Establish Purge Volume -  The purpose of purging is to obtain
fresh,  representative, formation  water.   In this instance,
"representative"    means   characteristic   of   geochemical
conditions  in undisturbed  parts  of the formation  (1)- The
equivalent  of  three  to  five  standing  water  volumes  is
generally  agreed upon in permeable formation as the standard
purging  volume.  This procedure  should insure that  samples
are  drawn from the aquifer, not from  stagnant water left in
the well between sampling events.

Consistent  Sampling  Procedures  - A  combination of  proper
sampling  procedures and sampling  equipment is necessary  to
maintain   consistency  during   the   sampling  event.   The
dedicated  equipment  should  be  used   for both  purging and
sampling  and information such  as purge volume,  purge  rate,
groundwater  elevation and material of the  equipment must be
documented  each time samples are collected. The flow rate of
                       n-42i
-------
the  pump must  be adjustable  in order   to  collect  volatile
organic   samples  and  other  sensitive   parameters  without
aeration.   Whenever possible, groundwater samples  should  be
collected  immediately after purging  is   completed ensuring a
representative sample from the formation  water  is collected.

Specific   conductance,    pH   and  groundwater   temperature
measurements  are taken after the well  is purged.  The  field
meters  should be calibrated daily and  checked every 4  hours
during the sampling event.  All calibration  techniques should
be  in  accordance  with  manufacturers   specifications.  All
results are recorded on the Field Information  Form.

Field  and trip blanks are used as  external QA/QC samples  to
detect  contamination that may be introduced   in the field  or
during  transportation to and from the site.   The blanks will
also  reflect any contamination that  may   occur during bottle
preparation  and storage within the   laboratory.  Upon return
to  the laboratory, field and  trip blanks are  logged-in  and
analyzed as if they were another sample.

Trip  blanks are samples of  organic  free water in  40 ml VOA
vials  which are prepared at the same time sample bottles are
prepared  for shipment.   They remain  with  the  sample bottles
under  Chain-Of-Custody while in transit  to  the site,  during
sampling and during the return trip to the laboratory.  At  no
time during these procedures are the  vials opened.

Field  blanks  are  samples  of  deonized water  used by  the
sampling  teams (at a specific  sample point)  to clean  field
equipment  (meters, depth  to water   probe and   non-dedicated
filtration  equipment).   The deonized water  is  exposed to the
air  and transferred into empty sample  bottles, and returned
to the laboratory.

Filtration  Procedures -  Filtering is necessary  in order  to
analyze   ions  and  compounds  that  are  dissolved  in  the
groundwater.   Detection monitoring wells are  not constructed
like  drinking  water  wells  and  often   contain   suspended
materials   like  silt  and  clay.    Any   suspended  sediment
contained  in the sample can react with the  sample and change
the  concentration  of  some of  the  dissolved   constituents,
yielding   a  sample  that  is  not   representative  of  true
groundwater   quality  (2).  The  sample   must    be  free   of
particulates  and must not be  exposed to air  (3).   Positive
pressure  filtering  should  be  performed  at  the  same time
samples  are collected-thereby,  minimizing change in  sample
temperature and preventing the de-gassing of CO-.  An in-line
0.45  (jut filter is recommended.   The filter is connected   to
the  discharge tubing of dedicated bladder pumps and filtered
samples  are collected  directly  in  the  sample  bottle.  The
in-line  filter is  disposed of  after each sample point   is
collected,   minimizing  cross-contamination  between  sample
points.
                      n-422
-------
 WIVII
         Environmental Monitoring Laboratories, Inc.
FIELD CHAIN-OF-CUSTODY RECORD


SITE/FACILITY #L_LJ_I_J SITE NAME: 	
Sample Point
         int:  II
AqyaPak™ PREP
AquaPalT
                                                                Date Seaied I   I   I  i   1  I
                                                                           YY  / MM  /  DD
                                                                By:
SAMPLE DATE: -
SAMPLE TIME:
                                MATRIX CODE:
Source Codes:
Dewatenng/Pressure Relief . . . . (D
Surface Water Impoundment . . . (
ENS#
SAMPLE
I.D.















l
(2400 HR )
) Leachate System
) Gas Condensate .
) Air
a
. . (C) Pretreatment Facility. . . (P) River/Slream/Brook
. . (A) Elfluent 	 (Tl Outfall 	
Dil 	
. . (R)
. . . (L)
. . . 101
	 (S) Other
Bottom Sediment . . . (B)
NnidP _ fNl
	 (X)
Generation Pt , . . . (G)



#OF
BOTTLES

















BOTTLE
TYPE
















AquaPak™ CONTENT
PRESERVATIVE
TYPE
















ANALYTES/LAB GROUPS
















FILTER
Y-N
^^N"
J^N"
ts^H
^^n
^^N
Z^N
2^N\
J>^N"
h>"""NJ
Z^N
^^N
J^^NJ
l^N
^^N
J>^N
Y Js^'
^ N
FIELD
COMMENTS
















E.M.L.
COMMENTS

















CHAIN OF CUSTODY CHRONICLE
AquaPak™ Opened By: (prir
Signature: 	
,f\ Oatp- / / Tirpp'
; 2400 HR.
Rpal «• Intart-
    I have received these materials in good condition from the above person.


 2.  Name:	-      Signature:—


    DatP-      /     /         T'—:.
                                    2400 HR.
    I have received these materials     od condition from the above person.
o. roame: 	 VJIMMCHUIW.- — 	
Datp- / / Tims" * Remarks:
2400 HR.
Am toDoLni OI-N/I/-!*-^ Dv/- /^rir»t\ DfltO"
4.
Sinnatiirp' Sftal #:

boftnmf Rv' ^J"8*"™* , , , , riflfp' , / '
MViaPnl- ff - TEMP. °c SFAI «


/ . Tirnp-
2400 HR.
Intact'

^V ' :
-v. .;•. 2400 HR.
INTACT
                                           H-423
-------
documented  on the C-O-C form.  The C-O-C  must  be  signed each
time  the samples  are  transferred to  the  responsibility of
another  person.   The C-O-C form must   be  signed  and enclosed
with  the samples when the  cooler is sealed  for   transported
back to the lab.   A tamper proof seal is placed on the cooler
and the seal number is recorded on the  C-O-C  form.

In  addition to documenting custody,  the  C-O-C form is  also
used  to identify field sample  point,  source code,  date  and
time  sampled. It also documents which  sample ID'S were field
filtered.   The C-O-C form lists  all sample  ID'S, number  of
sample  bottles and  type, analyte  groups and  preservatives
contained  in the cooler.    There's also   additional room for
comments specific to each sample ID.  Refer  to  Figure 1.

Field  Information Form  -  This form documents  the sampling
event.   The form  contains information    regarding  site  and
well  conditions, sampling and purging  procedures  used,  field
measurements  and field comments.  The  FI  form  must  be filled
out for each sample point. Refer to Figure 2.

Purging  Information  -  This  documents   the   date   and time
purging  began, the elapsed time of purging  (hrs), the volume
of  water  calculated  in  one  well  casing  and  the  volume
actually purged.

Purging  and Sampling Equipment - This  refers to  the types  of
dedicated  equipment, materials  and tubing  used  during  the
sampling event.

Field   Measurements  -   During  any   sampling   event   the
groundwater  elevation  (depth  to  water  adjusted  to  MSL),
specific  conductance  at  25°C, pH  and temperature  must  be
recorded.   Additional parameters may also be  required based
on  site specific conditions.

Field Comments - The section on field comments  should include
sample  appearance (if  applicable,  odor, color,  turbidity),
weather conditions (wind speed, direction  and precipitation)
and  other comments such as condition   of  the well,  dedicated
equipment,   field  meter   calibration    and   general  field
observations.

Sampling  Certification  -  The person  signing the   sampling
certification  must  be  present during  the  entire   sampling
event.
                      E-424
-------
 WIVII  Environmental Monitoring Laboratoriesrlnc.

FIELD  INFORMATION FORM
                                                                                       Site #
                                                                            Bottle Set:
                                                                     Sample Point
                                                        «:  Ml   I   I   I  I  II
                                           PURGING INFORMATION
                                    ill        LI  i  I         I
             PURGE DATE
             (YYMMDO)
                                                   ELAPSED MRS
                                                                          11
                                                                   VWER VOL W CASING
                                                                       (Gafcm)
Purging Equipment
          START PURGE
          (2400 Hr dock)
            PURGING AND SAMPLING EQUIPMENT
7 Dedicated I Y I   I N I            Sampling Equipment	Dedicated | Y
                                                                         1
                                                     ACTUAL VOLUME PURGED
                                                          (Gallons)
                                    (circle on*)
                                                                                               ,
                                                                                           (clrcrt o
-------
                   REFERENCES

"Sampling  Strategies,"  US  EPA  Ground-Water Monitoring
Seminar Series,  CERI-87-7 (Technical Papers) 1987, 4-16.

 Lindorff,   D.E.;   ;Feld, J. ;  Connelly, J.,  Groundwater
Sampling  Procedures  Guidelines, Wisconsin  Department of
Natural Resources,  PUBL WR-153-87, 1987, 44.

 Smith, J.S.;  Steele,   D.P. ;  Malley, M.J.;  Bryant, M.A.,
"Groundwater  Sampling," ACS  Principles  of Environmental
Sampling, 1988,  255-60.

 Manual for Groundwater Sampling, Waste Management, Inc.,
Oak Brook,  IL.,  1986  (unpublished).

  Site  Specific   Groundwater   Monitoring  Plan,  Waste
Management, Inc., Oak  Brook,  IL., 1987 (unpublished).

 Site Assessment Manual, Waste  Management North America,
Oak Brook,  IL.,  1989  (unpublished).
                  H-426
-------
113     ADAPTATION OF A SIMPLE COLORIMETRIC METHOD FOR FORMALDEHYDE FOR
       USE WITH GROUNDWATER MATRICES

       John DeWald. Turner Smith, S-CUBED, A Division of Maxwell Laboratories, Inc., 3398 Carmel
       Moutain Road, San Diego, California 92121-1095

       ABSTRACT. The NIOSH Method 3500 for formaldehyde in air has been adapted by the authors
       for use with groundwater samples. This simple colorimetric method provides sensitive analysis
       with detection limits of < 100 ppb. It also provides good precision and accuracy, with recoveries
       generally between 90 and 110%. The authors present data from detection limit studies, precision
       and accuracy studies, and ruggedness testing.  A discussion of the background of this method,
       and its application to  real-world samples with formaldehyde contamination is provided. A
       modification of this method may also be useful for soil and sludge analysis.  This method
       provides many advantages over traditional methods for soil and water analysis which have
       utilized HPLC analytical techniques.
                                            H-427
-------
114           A dual bio-monitoring system for the genotoxicity of air and water
                           at  the  site  of  hazardous waste  mixtures

                                          Te-Hsiu Ma
                     Department of  Biological  Sciences,  Western  Illinois  University
                                       Macomb,  IL  61455
         Abstract


         Chromosomes  of  the meiotic pollen mother cells of Tradescantia  (spiderwort)
         are the clastogenic targets in the Tradescantia-micronucleus (Trad-MCN) bioassay,
         while  the  genes (blue/pink alleles)  of the stamen hair cells of  Tradescantia
         are  the genetic end points of  the  Tradescantia-Stamen-Hair mutation (Trad-SHM)
         bioassay.   Tradescantia clone #4430 or 03 can be used concurrently for both of
         these  two  well  established bioassays as  a  dual bio-monitoring system to detect.
         the  clastogenicity and mutagenicity of  the  gaseous  or liquid mixtures  at  the
         hazardous  waste site or conduct laboratory tests on the water or  soil samples
         collected  at the sites.  Both Trad-MCN and Trad-SHM have the  extensive database
         accumulated  in  the past 10 -  15 years.   This  paper presents  a review of the on
         site monitoring and laboratory  testing results in  the  earlier publications  and
         current studies, and suggests  that this  dual  system  is suitable  for monitoring
         waste  sites  for their potential hazards, and  follow  up monitoring  of the sites
         after  clean-up  operation for  quality assurance.

         Plant  cuttings  (15  -  20  per  group) of Tradescantia   clone #4430  or 03  (a
         blue/pink  heterozygote) were  maintained in Hoagland solution for on site exposure
         or laboratory  tests on water sample or  the  solution washed off from the soil
         samples collected  at the waste sites.    For Trad-MCN  assay, the  inflorescence
         (series of flower  buds) were fixed in aceto-alcohol (1:3 ratio) after a 24 hr
         recovery time  for  preparation slides of  tetrads  (the  4-cell stage  of meiosis)
         of pollen  mother cells.   Micronuclei in  the tetrads derived from chromosome
         breakage were used as the  indicators of  clastogenicity of the pollutants.   For
         Trad-SHM assay,  an  11  - 14 day recovery time was  needed to  reveal the peak rate
         of pink mutation events in the  stamen hairs.  The  pink cells as  the results of
         somatic mutation in  the predominantly blue cells  in the stamen  hair were scored
         immediately  after  the recovery  time.

         Positive results of on site  air monitoring were  obtained from parking garages,
         truck  stops  and bus  depot,  industrial districts  of  various cities  in the  US,
         People's Republic  of China and Mexico,  and from college dormitories and other-
         indoor conditions.   On site  monitoring of  radiation  effects  from nuclear power
         plants, radon-contaminated houses  were obtained.  External and internal radiation
         effects of X-rays,  Gamma rays, beta and neutron particles were detected at very
         low  levels.  Positive test results were obtained from  the wastewater and drinking
         water  samples collected from Mexico,  and  People's Republic of China, and the  US.
         Recently,  on site  monitoring results were  obtained  from  the Lake  Superior,
         Canada, as well as  the  results of laboratory tests on 7 chemicals selected from
         the  US EPA Superfund Priority 1 list  and  the mixtures of some of these chemicals
         were also  included..
                                           n-428
-------
Introduction
Chromosome  is the  carrier  of the  DNA templates  which  determine the  genetic
traits, metabolic,  developmental  and all  the vital  processes of  life.    It  is
the  most fragile  structure  in  the  cell  during  its  interphase stage  while
replication  takes  place  through  the synthesis  of  the  new  halves  from the
dissociated  double helix without  the protection  of  its  nucleoproteins.   The
physical or chemical agents in the environment may excise the diester  bonds,  or
peptide  bonds and/or  interfere  with  the  proper  fusion  of  these bonds  which
maintain the  continuity of  the double helix,  thus  results in chromosome breaks
in the dividing cells.  The deletion,  duplication  and/or  alteration of  the DNA
bases caused by the foreign agents  in  the cell may lead to  the changes of  genetic
codes and result in gene mutation.   Chromosomes in the  young inflorescences  of
Tradescantia  have  the  easy  access  to  the gaseous or  liquid agents by  diffusion
through the porous tissues, or by absorption through the efficient transporting
vascular systems.   Based  upon the fragility of the chromosomes of the  meiotic
pollen mother cells and  the  easy  accessibility  to the  foreign  agents,   the
Tradescantia-Micronucleus (Trad-MCN) bioassay was developed [17, 22].   By taking
the advantage of the high  mutability of the  blue/pink gene locus in the dividing
cells  of the stamen  hair,   the  Tradescantia-Stamen-Hair-Mutation  (Trad-SHM)
bioassays was established  [34,  35,  49].     A dual  bio-monitoring  system was
developed by  applying  both Trad-MCN and Trad-SHM bioassays concurrently to the
same  group  of  Tradescantia  plant cuttings  (clone  4430, or 03)  which  were
subjected to  the  same treatment or  exposed at the same  site.   Both  of these
bioassays were able to  detect  gaseous,  liquid or radioactive pollutants [21,26],
and  already  had a  broad  database which  includes   in situ  and  in laboratory
studies and the test results  of 7  chemicals from the US EPA  Superfund Priority
1 List of the hazardous waste sites [33,39].   This paper presents a  review of
the  publications  of earlier  studies  and  some of  the  current investigations
[9,31,33,36] and suggests that this dual bio-monitoring system may be utilized
for monitoring  industrial  waste  sites for  the potenital hazards and  follow up
monitoring of the waste sites after clean-up operation for the  quality assurance.
Materials and Methods

For in situ monitoring of clastogenicity and mutagenicity of hazardous chemical
mixtures, this dual  system  is  one of the most efficient bioassays.  There are
two major approaches to  conduct  the  test.   One approach is to bring the plant
cuttings of Tradescantia clone #4430, or #03 (both are small in size and having
long blooming season) to the hazardous waste sites for a short exposure (24 hr
or less).  This is referred to as in  situ monitoring.   The  other approach is to
grow these special clones of  plants  on or near the site.  This is referred to
as the sentinel approach.   When  this system is used as an in situ monitor for
air pollutants, 15 plant cuttings bearing  young  inflorescences  are  held in a
screen cage  and  left  at the  selected sites  for an  appropriate  duration of
exposure, and brought back  to  allowed a 24 hr recovery time under the control
conditions before the inflorescences  are fixed in an  aceto-alcohol (1:3 ratio)
solution.  For  water pollutants,  the sample plant cuttings  are  carried on a
floating device called  "Aquatoon" for a continuous  exposure of  30 hr without
recovery time, and fixed  in aceto-alcohol.   The detailed  procedure for tetrad
selection, staining,  slide  preparation and micronucleus scoring are described
                                   H-429
-------
in the earlier publications  [17,22].   An  image  analysis system [50] specially
designed for scoring Tradescantia micronuclei  was developed  for Trad-MCN.  Thus
the scoring and data analysis can be carried out automatically to increase the
efficiency.  If the stamen hair pink mutation is used as the end point, a 9-14
day recovery time after treatment is needed to capture the peak mutation rate.
Scoring of pink mutation events from the predominantly blue  cells in the staman
hair should be done  immediately  after  the  recovery  time.   When this system is
used as a sentinel, young plantlets are transplanted to the appropriate locations
of the hazardous waste sites where the minimum survival requirements for these
plants are met. The pink mutation scoring procedure and the data analysis were
described in earlier publications [34,35,49]  The micronuclei  frequency in the
tetrads of the meiotic pollen  mother cells, or  the  pink mutation  rates of the
flower samples from the mature  plants grown on the site are  the indicator of the
clastogenicity  and  mutagenicity  of the  total  environmental  condition.    A
comparable lot, in the field  which is  relatively  free of pollutants of all kinds
should be used for growing the control  group for sentinel monitoring.   In both
of these two approaches  of in situ monitoring, climatic conditions  of the site
several days before, and during the  days of test should be recorded for better
interpretation of the data obtained.
Database and Discussion
The current literature survey covers the studies of  in  situ  monitoring of the
gaseous mixtures at  the  outdoor and indoor  sites,  the in situ  monitoring of
liquid mixtures.  The studies on the pure gases in the controlled chambers, and
the wastewater or drinking water and the solution extracted from the soil samples
collected from the  polluted sites are also reviewed.  Recent  bioassay on seven
chemicals selected  from the Super-fund  Priotority 1  list and earlier studies on
the effects of internal and external radiation, especially the  in situ monitoring
at the nuclear power  plants  are  included In order to corroborate the efficiency
and suitability  of  this dual system.   The location of the sites monitored, the
sources of the samples collected for laboratory testes  and the  references for
each of these studies are listed in  Table  1.
                                  H-430
-------
Table 1.  Literatures  pertaining to the studies  on  the in situ monitoring   of
pollutants and laboratory testing of samples collected at the site of pollution,
as well as the chemicals commonly  found at the hazardous waste  site.
Type of
pollutants
 Type of
 assays
 Site location
 or chemical classes
Reference cited
Gas mixtures
  Outdoor
 in situ
 monitor
Trad-MCN
Trad-SHM
   Indoor
  In Chamber
 in situ
 Trad-SHM
 Trad-MCN
 Lab test
 Trad-MCN
 Trad-SHM
Liquid mixture

   wastewater  in situ
               Trad-MCN
               Trad-SHM

   Wastewater  Lab test
    Tapwater    Lab test
Priority list
   chemicals
Radiation
  Lab test
  Trad-MCN
                Trad-SHM
Bus depot,  PRC, US
parking garage, US
Petro Company, US
Truck stops
Industrial district,
                                                  US
                   PRC
                Mexico

College Residential hall
Smoking rooms
Radon contaminated house
Trailers
Household cleaning agents

Ethylene dibromide, EMS
Pesticides  Malathion. DDV
Formaldehyde
Air fresheners
Sulfur dioxide, Ozone, N02
Diesel exhaust fumes
Lake superior, Canada
Sea water, PRC
Sludge, Chicago, US

Arena canal, Mexico
Fujian,      PRC

Spring Lake,US; Sichuan,PRC
Shallow well, Lewistown, US

Lead tetra acetate
Tetrachlorethylene,Aldrin
Dieldrin, Arsenic trioxide
Heptachlor,Benz(a)anthracene
other mutagens, Japan
[19,21,26]
[19,21,26]
[21,26]
[19,21,26]
[15,26,40,41,42,43]
[6,21,26]
[31]

[31]
[21,26,29,31]
[9]
[29]
[31]

[16,37, 45]
[6,17,23,26]
[36]
[9,29]
[21,45]
[20,21,24]
[8]
[3,5,7]
[11]

[38]
[51]

[14,25]
[30]

[33,39]
[33,39]
[33,39]
[33,39]
[48]

Internal
External
In situ
Trad-SHM
Trad-MCN
Trad-SHM
Trad-MCN
Trad-SHM

Nuclear power plant, Japan
P-32, H-3, 1-131
X-rays,
Gamma rays, Neutrons
Soil, Bikini, Island

[12]
[1,26,46,47]
[17,18,19,37
[2,37,44]
[13]

,44]
                                   n-43i
-------
Based upon the test results of the common environmental pollutants, both Trad
MCN and Trad-SHM bioassays  can detect  very  low concentration of chemicals at uM
level, and  low  dose of radiation  at  the pCi and  mR levels.   This dual bio-
monitoring system could be  claimed  to  be  the most sensitive and easy to operate
as well as cost effective (28)  one among the well known in situ bio-monitors [27,
32].    This system can be operated under the common climatic conditions  and there
are more than 20 species distributed  through  the  U.  S.  which could be used as
sentinels around the hazardous waste sites if the clones #4430 and #03 are not
suitable  in  certain  geographic  areas.    There  are  more than  10 clones  of
Tradescantia plants with heterozygous blue-pink locus  which are  suitable for
Trad-SHM tests.   The plant cuttings can survive in  the  in vitro conditions for
indefinite length  of  time  and function  as the  intact  live plants.   Thus the
plant cuttings used for testing are the  portable   in vitro  test  materials but
at the same time  they can provide  effectively  the  results comparable  to the data
obtained from the in vivo  tests.

The  results  of  this dual  system  can elucidate  the  relationship  between  the
chromosome damage and  gene  mutation.   The damage  on the meiotic chromosomes in
the  gametes  of  the Trad-MCN  system can   also serve as the indicator  of  the
genetic effects which may  be  passed on  to the future generations.   This dual
bio-monitoring system  is specially  suitable for monitoring hazardous industrial
waste sites where fumes, contaminated water  and soil can be monitored in the form
of mixtures.  This bio-monitoring  process accompanied  with  chemical  analysis
would  also be able  to  isolate the  prime hazardous agents from the  mixtures.
Application of this kind of bio-monitors to the industrial waste  site prior to
the  costly  chemical analysis  would  be   able  to  rank  the relative degree  of
hazardous conditions of many waste  sites,  and set the priority for the abatement
operations.  This would  make the waste  site removal operation more effective and
economical.
References

[1]   Anderson,  V-  A.  and T.  H.  Ma [1981]  Micronuclei induced by internal beta
      irradiation from incorporated phosphorus-32 in Tradescantia pollen mother
      cells.  Environ.  Mutagenesis,  3:398.

[2]   Anderson,  V. A.  and T, H.  Ma [1982]   Micronuclei induced  by low-dose
      cobalt-60  gamma-irradiation in Tradescantia pollen mother cells.
      Environ.  Mutagenesis,  4:348.

[3]   Chen,   D.   and  T.  Fang  [1981]    A  preliminary  study  on  the use  of
      Tradescantia-Micronucleus  technique  in  monitoring of  marine pollution.
      J.  Shandong Coll.  Oceanology 11:81-85.

[4]     Chen,  D.  and  D.  Xiang  [1983]  Preliminary  results of  Tradescantia-
      Micronucleus tests  on  the  wastewater  samples  from  several industrial
      factories  in Qingdao,    J.  Environ.  Sci.,  4:45-47.

[5]   Chen, D. ,  C. Li and B. Han  [1988] Application of Tradescantia-Micronucleus
      (Trad-MCN) technique to study the clastogenicity  of  heavy metals before
      and after  decontamination with marine yeast. Acta Scientiae Circumstantae
      8:79-83   (in Chinese  with English abstract).
                                  E-432
-------
[6]   Fang, T  [1981]   A report  on the  studies  of  effects  of environmental
      pollutants on chromosomes - A Sino-American  collaborated research project
      1980 II.   Tradescantia-Micronucleus bioassay on environmental mutagens in
      the air and water samples from some industrial  areas of Qingdao, PRC and
      on the pesticide-DDV.  J. Shandong Coll. Oceanology 11:0-11.

[7]   Fang, T.  [1981] A preliminary study on the use  of Tradescantia-
      Micronucleus technique  in  monitoring  mutagens in sea water-  J. Shandong
      Coll, Oceanology, 11:74-79  (in Chinese with English abstract).

[8]   Grant, W. F. ,  H.  G. Lee,  D.  M.  Logan  and M.  F. Salamone [1989]  The use
      of Tradescantia  and Vicia  faba  bioassays  for  the in  situ detection of
      mutagens in an aquatic environment,  Environ. Mol. Mutagenesis,  14 (Suppl.
      15):  75.

[9]   Harris,  M. M. and T. H. Ma  [1983]  Tradescantia-Micronucleus test on the
      mutagenicity of air fresheners,  Environ. Mutagenesis, 4:65.

[10]  Ho, J.,  R.  Zhou  and  T.  Fang [1983]  Tradescantia-Micronucleus tests  on
      fluoride contaminated  soil, Chinese  Soil Utilization  and Conservation,
      In: Proc. 5th   Ann. Meet. Chinese soil Sci. Soc. ,  2:325-326 (in Chinese
      with English   abstract).

[11]  Hopke, P. K. ,  M.  J. Plewa,  J. B.  Johnson, D.  Weaver,  S.  G.  Wood,   R.  a.
      Larson and T. Hindley [1982]  Multitechnique screening of Chicago
      municipal sludge mutagenicity.  Environ.  Sci. Technol., 16:140-147.

[12]  Ichikawa, S  [1981]  In  situ monitoring with Tradescantia around nuclear
      power plants, Environ. Health Perspect.,  37:145-164.

[13]  Ichikawa, S.and C. Nagashima [1979]  Changes  in  somatic mutation frequency
      in the stamen hairs of Tradescantia grown in the soil samples from Bikini
      Island,  Japanese. J. Genetics (in Japanese).

[14]  Lo, M.  [1985]   Tradescantia-Micronucleus tests on drinking water,  Sichuan
      Environment, 4:45-47 (in Chinese).

[15]  Lower, W. R., P-  S.  Rose,  and V.  K.  Drobney  [1978]  In situ monitoring of
      mutagenic and other effects  associated with lead smelting,  Mutation Res.,
      54:83-93.

[16]  Ma, T. H., A. H.  Sparrow, L. A. Schairer and A. F.  Nauman  [1978] Effects
      of  1,2-dibromoethane   (DBE)  on  meiotic  chromosomes  of  Tradescantia,
      Mutation Res., 58: 251-258.

[17]  Ma, T. H. [1979]  Micronuclei  induced  by  X-rays and chemical mutagens in
      meiotic pollen mother  cells of Tradescantia  - A promising mutagen test
      system,  Mutation Res., 64:307-313.

[18]  Ma, T. H. G. J. Kontos  and V. A. Anderson [1980] Stage sensitivity and dose
      response of meiotic chromosomes of pollen mother cells of Tradescantia to
      X-rays,   Environ. Expt. Bot., 20:169-174.

[19]  Ma, T.  H.   [1981]  Tradescantia-MCN-in-Tetrad  Mutagen test  for on site
                                   H-433
-------
      monitoring and further validation,  U.  S.  EPA Report EPA-600/SI-81-019.

[20]   Ma,  T.  H.,  V.  A. Anderson, and S. S. Sandhu  [1982] A preliminary study of
      the   clastogenic  effects  of diesel  exhaust  fumes using  Tradescantia-
      Micronucleus bioassay,  In: Short-term Bioassays in the Analysis of Complex
      Environmental Mixtures II,   Waters, Sandhu,  Huisingh,  claxton and Nesnow
      (Eds.)  Plenum Publ.  Corp.,  NY pp.  352-358.

[21]   Ma,  T.  H., V. A.  Anderson and I. Ahmed  [1982]   Environmental clastogens
      detected by meiotic  pollen  mother  cells of  Tradescantia,  In:  Genotoxic
      Effects of Airborne  Agents,  Tice,  Costa  and  Schaich (Eds.)   Plenum Publ.
      Corp.,  NY  pp.141-157.

[22]   Ma,   T.   H.   [1983]   Tradescantia-Micronucleus   (Trad-MCN)  test   for
      environmental  clastogens, In:  In vitro Toxicity Testing of Environmental
      Agents: current and  Future  Possibilities,  Pt.  A., Kolber,  Wong,  Grant,
      DeWoskin and Hughes  (Eds.)  Plenum  Publ.  Corp.,  NY pp.  191-214.

[23]   Ma,  T.  H., V.  A.  Anderson,  M.  M. Harris  and  J.  L.  Bare  [1983]
      Tradescantia-Micronucleus (Trad-MCN) test on genotoxicity  of Malathion,
      Environ. Mutagenesis, 5:  127-137.

[24]   Ma,  T.  H., W. R.  Lower, F. D. Harris, J. Poku, V. A. Anderson, M. M. Harris
      and  J.  L.  Bare [1983]  Evaluation  by the  Tradescantia-Micronucleus tests
      on the  mutagenicity  of  internatioanl combustion engine exhaust fumes from
      diesel  and   diesel-soybean oil mixed fuels,  In:  Short-term Bioassay in
      the   Analysis  of  Complex Environmental  Mixtures  III,   Waters,  Sandhu,
      Lewtas, claxton,  Chernoff and Nesnow  (Eds.) Plenum Publ.  cor.p. ,  NY pp.
      191-214.

[25]   Ma,  T.  H., V. A.  Anderson,  M. M.  Harris,  R. E.  Neas,  T. S.  Lee  [1984]
      Mutagenicity of drinking  water  detected  by  the  Tradescantia-Micronucleus
      test,  Can.  J.  Genet. Cytol.  27:143-150.

[26]   Ma,  T.  H., M. M. Harris, V. A. Anderson, I.  Ahmed,   K. Mohammad, J. L. Bare
      and  G.  Lin [1984]  Tradescantia-Micronucleus (Trad-MCN)  tests on 140
      health  related agents.  Mutation Res.,  138:157-167.

[27]   Ma,   T.  H. ,  M. M.   Harris  [1985]   In  situ  monitoring  of  environmental
      mutagens,  In: Hazard  Assessment of chemicals,  Current Development, Academic
      Press Inc.  pp.  4:77-106.

[28]   Ma,  T.  H.  and G.  L.  Cabrera  [1986]   Development  and application  of quick
      and   easy  bioassays  for environmental mutagens,    In:  II  Semana  de las
      Ecologia y Proteccions del  Ambiente,   Universidad autonoma  de Queretaro,
      QRO,  Mexico,  May  19-23, 1986,  pp.  130-132  (in Spanish).

L29]   Ma,  T.  H.  and M.  M,   Harris  [1987]   Tradescantia-Micronucleus (Trad-MCN)
      bioassay - A promising indoor pollution  monitoring system,  In:  Proc. 4th
      Internl.  Conf. Indoor  Air Quality  and Climate,  Berlin(W)  August  17-21,
      1987, pp.  243-247.

[30]   Ma,  T.  H., R. E.  Neas, M.  M.  Harris, Z. Xu, C. Cook and D. Swofford [1987]
      In vivo  tests (Tradescantia-  and Mouse-Micronucleus) and chemical analysis
                                   HM34
-------
      on drinking  water  or rural communities,  In:  Short-term Bioassay  in  the
      Analysis  of  complex Environmental  Mixtures V,  Sandhu,  DeMarini,  Mass,
      Moore and Mumford  (Eds.) Plenum Publ.  corp., NY  pp.  189-205.

[31]  Ma, T.  H.,  Y. Peng,  T.  D. Chen,   S.  S.  Sandhu and  E.  F.  Ruiz  [1989]
      Tradescantia-Micronucleus  (Trad-MCN)  and  Stamen  Hair Mutation  (Trad-SHM)
      assays on the clastogenicity of chemical  mixtures and  in situ  monitoring
      of air and water  pollution,  Environ. Mol. Mutagenesis,  14  (Suppl.  15):37.

[32]  Ma, T.  H.  [1989] In situ  monitoring  of environmental  clastogens using
      Tradescantia-Micronucleus bioassay,  In: Proc.  the First Symposium on  in
      situ Evaluation of biological hazards of environmental  pollutants, Chapel
      Hill, NC, Dec. 5-7,  1988 (in press).

[33]  Ma,  T.  H.,   S.  S.   Sandhu,  Y.   Peng,  T.  D.  Chen and  T.  W.  Kim  [1990]
      Tradescantia-Micronucleus  (Trad-MCN)  tests  on   the  mixtures of  four
      chemicals commonly found at the  hazardous waste sites,  Mutation Res.,
      (Manuscript in preparation)

[34]  Mericle, L.  W. and R. P.  Mericle [1967]  Mechanism of  somatic mutation for
      flowers of hybrid Tradescantia  (clone 02) Genetics, 56:576-577.

[35]  Mericle, L.  W. and  R. A. Mericle  [1971] Somatic  mutations  in clone 02 _
      Tradescantia: A search for genetic identity, J. Hered.  62:323-328.

[36]  Mohammad, K.  and T. H.  Ma  [1983]   Tradescantia-Micronucleus  (Trad-MCN)
      and  Tradescantia-Stamen   Hair  Mutation   (Trad-SHM)  tests  on  common
      pesticides,  Environ. Mutagenesis, 5:65.

[37]  Nauman, C. H., A. H. Sparrow, A. G. Underbrink and L. A.  Schairer [1977]
      Low dose mutation response relationships  in Tradescantia:   Principle and
      Comparison to mutagenesis  following  low  dose gaseous  chemical mutagen
      exposures, In: Radiobiological  Protection, First European  Symposium  on
      Rad-equiva.lence"  Commission of the European Community, Luxemberg, EUR 5725e
      13-23.

[38]  Ruiz, E.  F. ,  E.  R.  Valtierra and  T.  H.  Ma  [1987]  Presencia de agentes
      genetoxicos en aquas residuales empleadas para riego  utilizando  el sistema
      de micronucleos  en  cellulas  gameticas de  Tradescantia  clone #4430.  In:
      III Semana  dela  Ecologia  y  Proteccion  del  AmbjLenie_,  Jan  22-26,  1987,
      Universidad Autonoma de Queretaro,  QRO   Mexico,  p.  50.

[39]  Sandhu,  S.  S. T.  H. Ma,  Y.  Peng  and X.  Zhou   [1989]   Clastogenicity
      evaluation of seven chemicals commonly found at hazardous industrial waste
      sites,  Mutation  Res., 224:437-445.

[40]  Schairer, L.  A., J. Van't  Hof,  C.  C.  Hayes,  R,  M,  Burton  and F. J.  de
      Serres [1978] Exploratory  monitoring  of air pollutants for genotoxicity
      activity with Tradescantia stamen hair  system, Environ. Health Perspect.
      27:51-60.

[41]  Schairer, L.  A., J. Van't  Hof,  C.  C.  Hayes,  R.  m.  Burton  and F. J.  de
      Serres [1978] Measurement of biological activity of  ambient air   mixtures
      using a mobile laboratory for in situ   exposure,  Preliminary results from
                                   H-435
-------
      the Tradescantia  plant  test  system,    In:  Application  of  Short-term
      Bioassays in  the  Fractionation  and Analysis  of complex  Environmental
      Mixtures, Waters,  Nesnow,  Huisingh,  Sandhu  and  Claxtori (Eds.)   Plenum
      Press,   NY pp.  419-440.

[42]   Schairer, L.  A., R.  C.  Sautkulis  and N.  R.  Tempel [1982] Monitoring
      ambient air  for mutagenicity using the  higher plant  Tradescantia, In:
      Genotoxic Effect of Airborne Agents, Tice, Costa and Schaich (Eds.) Plenum
      Press,  NY  pp.  123-140.

[43]   Schairer, L.  A. and R.  C. Sautkulis  [1982]   Detection  of  ambient levels
      of mutagenic  atmospheric  pollutants with higher plant Tradescantia,   In:
      Environmental  Mutagenesis,  Carcinogenesis,  and  Plant Biology,  E.  J.
      Kleiowski,  Jr.  (Ed.), Vol.  2,  Praeger Publ.,  pp.  155-194.

[44]   Sparrow,  A. H., A. G. Underbrink and H.   H.  Rossi [1972] Mutation induced
      in Tradescantia by small doses of  X-rays,  and  neutrons:  Analysis of dose-
      response curve  of  X-rays  and neutrons,   Science,  176:916-918.

[45]   Sparrow,  A.  H., L.  A.  Schairer [1974]  The effects of  chemical mutagens
      (EMS,  DDE) and specific  air  pollutants  (03,  S02,  N02, N20)  on  somatic
      mutation  rate   in   Tradescantia   (in  Rus. )   in:  N.P.  Dubinin  (Ed.),
      Geneticheskie Poaledstviya Zagryazneniya Okruzhayuschchei  Sredy (Genetic
      Effects  of Pollution  in  the  Environment),  Institut Obshchei  Genetiki,
      Moscow,  pp.  50-61.

[46]   Tano,  S.  and  H  Yamaguchi  (1979) Effects  of  low  dose irradiation from 131I
      on the  induction  of somatic  mutations  in  Tradescantia,  Radiat.  Res.,
      80:549-555.

[47]   Tano,  S. , H. Yamaguchi and S. Ueda  [1984]  Effects  of  low  dose  tritium
      labeled Lhymidine  and uridine on  the induction  of somatic  mutations  in
      Tradescantia, Environ.  Expt.  Bot.  24:173-177.

[48]   Tano,  S.  [1989] In  situ detection of induced  mutations with chemicals by
      Tradescantia, Environ.  Mol.  Mutagenesis,  14 (Suppl.  15):197.

[49]   Underbrink,  A.  G.,  L.  A.  Schairer and A. H.  Sparrow  [1973]  Tradescantia
      stamen hairs:   A  radiobiological  test  system  applicable  to  chemical
      mutagenesis,  in:  Chemical Mutagenesis:  Principles and  Methods for their
      detection,  Vol. 3,  A. Hollaender  (Ed.),  Plenum  Press,  NY pp.171-207.

[50]   Xu, J.,  T.  H. Ma,  W. Xia, X.  Jong, W. Sun  [1989]  Image analysis system
      for rapid data processing in Tradescantia-Micronucleus bioassay, J.American
      Soc.  Test.  Materials (in  press).

[51]   Zheng,   D.  [1985]   Tradescantia-Micronucleus   tests  on  the  industrial
      wastewater from a  printing and dying factory in  Fuzhou city.   J. Fujian
      Normal  University,    21:5-7.
                                  IM36
-------
ENFORCEMENT
-------
115                       USE OF WASTE STREAM AUDITS TO DETERMINE
                      THE REGULATORY STATUS  OF  SURFACE  IMPOUNDMENTS

       William R.  Davis, Environmental  Scientist, Region  IV,  U.S.  Environmental
       Protection Agency, Athens, Georgia  30613;

       Ralph  R.  Stewart,  Environmental Engineer,  Region  IV,  U.S.   Environmental
       Protection Agency, Atlanta, Georgia   30365

       ABSTRACT

       The  Resource  Conservation  and  Recovery  Act (RCRA)  regulations  list many
       hazardous wastes  by  describing  a waste  stream generated  by  a particular
       industrial  process.    Frequently,  enforcement  personnel discover  a waste
       management  unit  or   spill   area  which  has  high  levels  of  hazardous
       constituents.   In the  absence of a thorough  understanding of the industrial
       process, it is difficult  for  the inspector to legally identify the unit as
       a "regulated hazardous waste management unit" and thus to require corrective
       action using regulations applicable to regulated units.

       This paper  will  discuss the corrective  action  required  for RCRA regulated
       units  as  well  as  non-regulated  units   and will  provide  procedures  for
       performing a waste stream audit to aid enforcement personnel in determining
       which units should  be  regulated.   Waste stream audits  involve  a detailed
       evaluation of  the chemical and/or manufacturing processes which are involved
       in the synthesis of chemical compounds and/or products.  The audit includes
       raw materials  information, chemical reaction and process details, and waste
       stream generation and  disposal procedures.  The key to the successful waste
       stream  audit  is  the  combination of  RCRA personnel  who are  familiar with
       hazardous waste listings  and  regulations and field investigative personnel
       who  are  experienced  in  the  chemical   process   industry  and  subsequent
       wastewater treatment and solid waste management.

       A case study is presented in  which  both  a field sampling investigation and
       a waste stream audit were performed  at  a chemical manufacturing facility.
       The  sampling  investigation  identified hazardous  constituents  in  surface
       impoundments which were used  as  a basis  for investigating processes during
       the waste stream audit.   The  audit  concentrated on identifying "F" and "U"
       coded wastes  which could  have  been  discharged  into the  impoundment with
       process wastewater.     The  procedures used in  performing  the  audit  are
       presented.   The results of the audit  are  summarized and a corrective action
       outline is developed based on the results of the entire study.

       INTRODUCTION

       In 1976, Congress passed  the  Resource Conservation and Recovery Act  (RCRA)
       which required EPA to  develop "cradle to grave" tracking and regulation of
                                          H-437
-------
hazardous wastes.   The RCRA regulations were written so that generators could
easily  identify  which  of their  wastes were  hazardous without  extensive
laboratory analyses.  The generator  could  then follow  the  requirements for
storage, treatment, or disposal of the waste.

To  facilitate  easy identification  of hazardous  wastes,  RCRA  lists waste
streams by industrial process.  Waste  chemicals  are identified by how they
are  used  in  a   manufacturing  process.      RCRA  also   identifies  four
characteristics that would make a waste legally defined as a hazardous waste.
These  are  Ignitability,  Corrosivity, Reactivity,  and  Extraction Procedure
(EP) Toxicity.  Each  of these characteristics has a definitive laboratory
test procedure which  will provide conclusive  evidence  of whether  a waste
falls  within  the   criteria  of a  characteristic  hazardous waste.    The  EP
toxicity  test  is  in  the   process   of being replaced  by  the  Toxicity
Characteristic Leachate Procedure (TCLP) test.   EPA estimates that the new
rule will  nearly  triple the  amount  of waste  that  is  considered hazardous
under RCRA.

The  intent of  the regulations to facilitate easy  identification  of listed
wastes by identifying the industrial process has made it difficult for RCRA
permitting and enforcement personnel to identify hazardous waste management
units.   When an  inspector  discovers  an  unknown  unit  or  spill  area,  the
inspector  must  rely on  facility  personnel  to identify the  process which
generated the waste.  When the inspector is  not familiar with a wide variety
of  chemical  manufacturing  processes,   the  facility's   identification  goes
unchecked.

By conducting an audit of the wastes generated at a facility,  the inspector
can become familiar with  the  processes  at  a particular  plant  and will gain
knowledge which will enable him to determine the regulatory  status of a waste
management unit in question.

WHY IDENTIFICATION OF REGULATED UNITS IS IMPORTANT

Any  land based unit which received hazardous waste after  July 26,  1982,  is
termed a "regulated unit" under RCRA.   The corrective action requirements for
regulated units are given in Title 40 of the Code of Federal Regulations (40
CFR), Sections 264.90  through 264.100.   These 11 sections of the regulations
specify such requirements as a ground-water protection standard, monitoring
programs,  concentration limits,   a  point  of compliance,  and  a compliance
period.  These programs  are implemented by  issuing either an operating permit
or a post-closure permit to the facility.

Hazardous waste management units which have  not received hazardous waste and
thus  cannot  be classified  as a  "regulated unit"  are  termed  solid waste
management units  (SWMUs).  Corrective action for SWMUs is addressed in 40 CFR
Section 264.101.   This single section in the regulations does not explicitly
state the clean-up  requirements for  SWMUs.   Corrective  action for SWMUs is
implemented by issuing  a permit to the facility  known  as  the  "HSWA permit"
(Hazardous and Solid  Waste  Amendments Permit), or by  issuing a corrective
                                  EW38
-------
action enforcement order known as a Section  3008(h)  Order.   The  HSWA permit
or the  3008(h)  order directs  the facility  to  perform an investigation  to
determine  the  extent of contamination as well  as to propose  a method  to
rectify and clean-up the contamination.

Although the corrective action process under a HSWA permit  should parallel
the  corrective  action  process  for a  regulated  unit,  there  are  several
significant differences.  Corrective action  under  a RCRA operating or post-
closure permit theoretically happens more  expeditiously.  This is  because the
investigation under HSWA can be delayed and stretched out over several years
before any corrective action begins.  Once a SWMU is cleaned-up, the facility
is under no obligation to continuously monitor groundwater unless the permit
writer or  the 3008(h) order can  adequately justify the necessity to do so.
For  regulated units,  the post-closure  period is  required to be 30 years.
This  ensures  long term monitoring of  groundwater and early  detection of
continuing releases.

More  significantly,  Section 264.94  requires contaminated groundwater under
a regulated unit  to  be  remedied  to  background levels present in upgradient
wells.  The HSWA process allows the facility to propose clean-up levels based
on the facility's investigation and current health  based levels.  A complete
understanding of  the migration of contamination is difficult to obtain.   If
the investigation is not extremely thorough, then  non-conservative clean-up
levels  could  be established   based  on  erroneous  conclusions  of  the
investigation.

Finally, under current EPA  policy, once contaminated groundwater  is detected
under a  newly discovered  regulated unit  (one  that  does not have  interim
status  or  an operation  permit), the  unit must be  closed  as  a landfill.
Closure  as a landfill  requires  a  cap which  precludes  the facility  from
continuing  to use the unit  to  dispose of waste,  either hazardous  or non-
hazardous.  This  enforcement hammer is especially  useful  to  the regulatory
agency  when a  facility has a  waste  water treatment  system  in  surface
impoundments  and  the system is  suspected  of receiving routine  spills  of
hazardous waste.

WASTE STREAM AUDIT PROCEDURES AND OBJECTIVES

The specific objectives of  the  waste stream audit are  to (1)  secure accurate
production  information,  i.e. identify  chemicals  and  products made  at  the
facility,  (2)   acquire  a  workable  understanding  of  the  basic chemical,
thermodynamic, and kinetic  principles of the individual unit processes at the
plant, (3) develop an inventory of raw  chemicals, intermediate products,  and
catalysts  used  in the reactions and/or manufacturing,  (4)  identify waste
streams and characterize their chemical contents through materials balance,
and (5)  determine the fate  of the chemicals identified in the waste streams.

The procedures used by EPA in  conducting a waste  stream audit are outlined
as follows:  (1) the appropriate  company officials  are notified and informed
of the intent to  perform  an audit.   This  allows  the company to assign the
                                   H-439
-------
proper personnel to  assist with the audit,  to prepare the needed information,
and to schedule those who may be  required to  provide  technical assistance,
(2) an initial  conference is conducted which includes the opening remarks and
clarifications of the audit objectives, (3) the audit is performed by guiding
the company  officials  through the objectives.   This phase  is technically
specific to  the unit processes  and includes a detailed  discussion of each
process and  each and every waste  stream associated with  the reactions, (4)
when  the  waste streams  have been  identified, and  the  investigators are
comfortable  with  the  facility  nomenclature  and  process  units,   a  visual
inspection of those  units  is  made,  and (5)  a final exit interview is made to
finalize the audit and to give the company an opportunity to make any  final
comments.

CASE STUDY

A Region IV chemical manufacturing facility was known to treat large volumes
of wastewater  in unlined surface  impoundments.  The  facility was  suspected
of having routine spills of  listed hazardous  wastes  in  process areas which
would drain  into the chemical sewer and into the surface impoundments.   EPA
was not familiar with all of the processes  at the plant, but it seemed likely
that listed  waste would be associated with some of the  many processes  that
occurred at  the facility.  An investigation was initiated to determine the
regulatory status of the surface impoundments.

Much of the  success of a waste  stream audit depends upon  the information
generated prior to the  actual conduct  of the audit.  EPA investigators  first
conducted a RCRA case development inspection/evaluation (CDIE) which provided
the auditors with an  opportunity to  familiarize themselves  with  the plant
layout.    During  the  CDIE,   a  thorough   investigation  was  conducted  of
identified SWMUs.  The wastewater treatment lagoon  system, storm water runoff
migration routes, and  an  old landfill with ancillary waste  piles  were the
primary areas  of  concern.  Also a series  of  ground water  monitoring wells
was sampled  to determine if contaminants  were reaching the shallow aquifer.
All of  the analytical  data were used to  develop a chemical  profile  of the
wastes  distributed  throughout  the  wastewater treatment  system water and
sludge, landfill  soil, waste pile soil,  runoff ditch sediments  and ground
water.  Both water  and  sludge  samples were  obtained  from  several  of the
impoundments  which  were  expected  to have  the highest  concentration  of
hazardous  constituents.   Figure 1 shows  the  approximate location  of  each
sample.  Table  1  shows the highest  concentrations  of  contaminants  found in
selected samples.  All samples were collected using  EPA  Region IV Standard
Operating Procedures and analyses were conducted according to  SW-846 methods.

The high  concentrations  of  Benzene  (5300 ug/1),  Xylene  (110,000  ug/1),
Toluene (5000  ug/1)  and Phenol  (24,000 ug/1),  gave evidence that  spills of
these chemicals could be entering the wastewater treatment system.  The waste
stream audit was scheduled to tie the detected chemicals to hazardous waste
listings.
                                  E-440
-------
PERFORMING THE AUDIT

The waste stream audit was begun by discussing the findings of the CDIE with
the company officials who were  given  an opportunity  to  explain the presence
of specific RCRA listed compounds  in their waste treatment system.   Armed
with chemical evidence of specific  chemicals  that routinely appeared in the
samples collected during the CDIE, the auditor's goal was to determine their
source.

This particular audit was technically compounded  because  the plant produces
more than 200 chemical products,  most by batch  reactions.   Before  the  audit
was initiated, company literature,  advertisements and the  1988  Directory of
United States Chemical Producers was used to develop  a potential list of the
major  chemical products  manufactured  in volume.   The company was  asked to
verify the list generated by the  auditors and to  add any other  products not
on the list.  Since the products  made at the  plant are  closely  related, and
since  it  would  have   been  impossible  to  analyze  every  process,  14
representative processes from the various production  areas  were  selected for
a detailed process review.

Company  and  corporate engineers  were  asked  to  explain  the  selected unit
processes.  Detailed process flow sheets outlined the chemical  reactions by
showing  the  sequence of raw materials added, the catalysts  used, and the
reactions, refluxes, decanting,  centrifugation,  filtration,  and  distillation
operations.  Solvent additions and recovery steps,  washing operations and any
source of  a waste  stream,  either liquid or  solid was  verified.  Materials
balances of raw materials,  intermediate products,  and  finished products were
roughly  calculated  for each process   audited.  The  amount  of  waste being
generated  for  disposal must be  inventoried  under RCRA regulations.   The
information prepared for manifests and regulatory annual  reports  should agree
with the  waste  streams  detailed in  the process  flow  sheets,   and can be
helpful in substantiating the previously cataloged chemicals found during the
field investigation.

After a meticulous indoctrination  in the various unit processes under review,
the  auditors  then  walked  through  the  various  units  to  verify   critical
information covered in  the  flow sheets.   Sources  of  waste  streams were
identified, and  the  containment or  treatment of these materials was noted.
Also the route for any spills or  intentional  dumps within the unit  area was
identified.  At this facility, most  of the units were  surrounded by   chemical
sewers which eventually drained into the waste water treatment system.   Some
process buildings had  stained  floors  which  gave  evidence  that spills were
migrating to the chemical sewer system.

After the visual inspection, the  auditors and company officials  conducted a
closing interview.   Any final questions were settled and a brief explanation
of  the   outcome  of  the  audit   was  made   to   the  company  officials.
Confidentiality was  maintained on  all  process  information not commonly listed
in chemical engineering textbooks and chemical  literature.
                                   H-441
-------
The final responsibility for the investigative scientist or engineer was to
consolidate the audit  findings  into a final report  which supplemented and
expanded the initial CDIE report.

CONCLUSIONS OF THE AUDIT

The nature of the operations at the facility made it difficult to associate
listed waste codes  with  the process waste streams.   Many "U" constituents
were  found in  the  impoundments,  but  these chemicals  were  used  as  raw
materials in the process and thus  did not meet  the listing requirements of
being a waste when they  were  introduced into the process.   The  F listings
were not met either because in  the  14 processes  reviewed,  benzene, xylene,
and toluene entered into the chemical reaction  and were  not used solely as
solvents as is required by  the  listings.  The evidence of spills migrating
to  the  chemical  sewer, coupled with the knowledge  of the use  of xylene,
toluene, and benzene  in the process gave the most  likely  source  of these
chemicals in the impoundments.  But  because  these  chemicals did enter into
every reaction investigated during  the audit, the  hazardous waste listings
were  not met  and  the surface  impoundments could  not  be  identified  as
regulated hazardous waste management units.

There were  two processes  where xylene was used  as a wash  solvent for the
reaction vessels  after  the reaction had occurred.  This xylene would meet the
F003 listing, but  the spent xylene was sent to an on-site distillation column
for reclamation.   There is no waste associated with the distillation column
that is discharged to the impoundments.

In  this  case  study,  EPA was not  able  to identify listed wastes  that were
discharged  into  the impoundments.   Therefore,   EPA did  not  issue  a post
closure permit to  require corrective action.   Although the  study did not show
that the  impoundments  were  regulated,  the information obtained  was indeed
very valuable in developing the HSWA permit.  By knowing the concentration
of  hazardous  constituents  in  the  impoundment  and  the  construction  and
operational details of the impoundments,  EPA was able to write specific and
detailed investigative requirements in the HSWA permit.  The facility is now
able  to delineate  the plume  of  groundwater  contamination  by  using  the
constituents found during the audit as a basis for  the investigation.

This investigation was  performed before the RCRA regulations were amended to
include the TCLP test.  EPA has added 25 organic constituents to the 14 EP
toxicity characteristic constituents resulting in a new characteristic test
which covers 39 constituents.  The  level for Benzene proposed in the March
29, 1990, Federal  Register is  0.5  mg/1.  Since the available data shows that
the level of Benzene in the impoundment is  in the range of 5 mg/1, then the
impoundment  would   be  considered  a   regulated   unit  having   the  new
characteristic of TC toxicity.   If the investigation shows that groundwater
is  contaminated,  the  facility  will need  to close  and  cap  the  unit,  or
retrofit  the  impoundment to  meet  minimum  technological requirements  and
remediate contaminated groundwater.
                                  n-442
-------
SUMMARY

The  regulatory  procedures  for  implementing  corrective  action  are  more
specific for RCRA regulated  units  than for SWMUs.   Waste stream audits are
an invaluable tool to use in determining the status of questionable or newly
discovered land based waste management units.  Although a waste stream audit
is time consuming, the information obtained undoubtedly proves to be valuable
in developing  corrective  action  scenarios.   Whether or not the audit shows
that the unit is regulated, a superior HSWA permit  can be written based upon
information obtained  from  the study.
                                   H-443
-------
                                  Table  1
Sample Number

W-l
W-2
S-l
Constituent

Phenol
Xylene (m&p)
Benzene
Toluene

Acetone
Tetrahydrofuran

Xylene (m&p)
Trichloroethene
Toluene
Benzene
Chlorobenzene
Concentration

 24,000 ug/1
110,000 ug/1
  5,300 ug/1
  5,000 ug/1

  7,100 ug/1
     80 ug/1

 25,000 mg/kg
 10,000 mg/kg
  1,600 mg/kg
    410 mg/kg
    650 mg/kg
Waste Code

 U002
 F003
 F005
 F005

 U188
 U213

 F003
 U228
 F003
 F005
 F003
-------
       INFLUENT
        rw-i)
        (W-2)
                   EQUILIZATION POND
                       POND 1
                    AERATION POND
                      POND 4
                                  (S-4)
                                                WASTEWATER
                                                         NEUTRALIZATION
                           ~i
                                                         POND 5
                                       OUTFALL
                                                        POND 6
                                (W-4
                                (S-5
        POND 2
    SUBSURFACE AERATION
        (W-2)
        (S-2)
 POND 3
(MUD POND)
                             (W-3}
                             (S-3)
                                              LAMELLA-
                                                              SLUDGE
                                FIGURE 1
                   WASTEWATER  TREATMENT  PLANT
&EPA
                                KEY:
                                W = WATER SAMPLE
                                S = SLUDGE SAMPLE
                                  H-445
-------
                                References

Kastner, K.  H.,  "Complying with the  New RCRA Toxicity  Characteristic and
         TCLP,"  Environmental Reporter,  March 23, 1990,  Bureau of National
         Affairs, Inc.

Lowrance, S. K. ,  "Guidance on Demonstrating Equivalence of  Part  265 Clean
         Closure with Part 264 Requirements," OSWER Policy Directive
         #9476.00-18, May 12,  1989.
                                  n-446
-------
116                            DESIGNING A LI MS
                      TO MEET ENFORCEMENT REQUIREMENTS
        Jeffrey C. WortKington
        Director of Quality Assurance
        TechLaw, Inc.
        ABSTRACT

        Many laboratory information management systems  (LIMS) do not
        include considerations for the following activities related
        to enforcement requirements!

             o Verification of the condition of the sample
               at receipt
             o Documentation of computer-resident data
             o Accountability of computer-generated bench sheets and
               analysis records
             o Quality assurance of software
             o Security of LIMS
             o Custody and tracking of physical samples
             o Verification of correct data entry
             o Documentation of direct electronic data transfer
             o Standard Operating Procedures for LIMS

        The author will review the highlights of two of the current
        commercial LIMS that are in use in many laboratories.  The
        author will also provide detailed requirements that each
        laboratory should consider in designing a LIMS or purchasing
        a commercial LIMS.  This paper will provide suggestions for
        systematic approaches to sample and document management with
        a LIMS and will include examples of user-friendly entry
        screens.  Examples of computer-generated bench sheets that
        are subject to easy modification to meet the needs of an
        individual laboratory will also be provided.

        A LIMS needs to be flexible to enable the laboratory to
        modify the software to meet unforeseen future sample
        tracking and technical considerations.  A well designed LIMS
        is a powerful management and quality assurance tool.  Adding
        functions to meet enforcement requirements will benefit
        laboratories and their clients.
                                   Tl-447
-------
117   The Use of Confirmed, "Tentatively Identified Compounds" to Build
     a Case Against the Primary Responsible Party at a Superfund Site

                            Thomas H. Pritchett
                    U.S.  EPA  Environmental  Response  Team

                         John Syslo  & Tony  Lo  Surdo
                         Roy F. Weston, Inc.,  REAC


     During most GC/MS semi-volatile analyses, an attempt is made to
     identify the unknown compounds detected by using searches against
     the standard mass spectral analyses.  Unfortunately, because of
     the large number of samples that must be processed by the typical
     CLP laboratory, the identification efforts rarely proceed further
     than this "tentative" identification step.  This past summer, the
     Region II Removal program requested the ERT's assistance in
     identifying the unknown compounds that were found throughout one
     of their site at levels approaching per cent concentrations.
     Because these compounds were listed as only "tentatively
     identified", there presence could not be used by the Agency for
     Toxic Substances and Disease Registry (ATSDR) in their health
     assessment nor by the Region in their enforcement case.  By
     acquiring standards for several of the tentatively identified
     compounds and other suspected compounds prior to starting the
     analyses, our laboratory was able to confirm the presence of
     several of these compounds using both GC retention times and
     spectral matches on the same instrument.   Based upon the
     recommendations of the GC/MS operator, additional standards were
     obtained and more unknowns were confirmed.

     When the newly identified compounds were compared against
     compound classes listed in responses to the EPA requests for
     information under 42 U.S.C. Section 9604 and 42 U.S.C Section
     6927, one company became a readily apparent suspect.  At the
     ERT's request the Regional Enforcement staff proceeded to request
     additional information on the commercial products listed in the
     response.  Based upon a comparison of the supplied information
     with the compounds and spectra  found in the site samples, samples
     were then requested of four of  the commercial products.  These
     products, two of which were mixtures of five or more distinct but
     related compounds, were then analyzed on the GC/MS.  When the
     acquired spectra and relative GC retention times were compared
     against the data from the site  samples, three of the products,
     one of which was a five component mixture, were confirmed to be
     present and in several cases were the primary wastes found.
     Based upon searches performed by the Analytical Operations Branch
     against the CARDs database, the compounds from two of the
     products were not commonly found at other sites by themselves and
     had been found together at only one other site in Superfund.
     Since these were all specialty  chemicals, their presence at the
     site strongly implicated the PRP.
-------
118                           EVIDENCE AUDITS t
                A CASB STUDY AND OVERVIEW OF AUDIT FINDINGS
       Jeffrey C. Worthington
       Director of Quality Assurance
       TechLaw, inc.

       Paula Smith
       Section Chief
       Evidence Audit and Quality Assurance Unit
       EPA-National Enforcement Investigations Center


       ABSTRACT

       Evidence auditors inspect the following activities to
       determine if complete documentation is provided by field and
       laboratory personnel to ensure that the chain of custody of
       the sample is maintainedi

            0  Sample collection
            o  Sample preparation in field staging areas
            o  On-aite field measurements
            o  Sample transfers
            o  Sample receipt at laboratory (or other location)
            o  Sample tracking
            o  Sample preparation and analysis
            o  Sample storage, archive, and disposal
            o  Site file development

       The authors will present a case study of a series of field
       and laboratory audits related to a specific site.  This case
       study will include a description of the audit planning
       process and a review of the findings of this specific series
       of audits.

       The authors will also present a summary of the problems that
       are typically found during both field and laboratory
       evidence audits.  Suggestions and a plan for the successful
       implementation of corrective action for evidence-related
       findings will be provided.  Thoughtful planning and follow-
       up review of these issues will enable the quality assurance
       officer for each site to ensure that enforcement-related
       activities are dealt with in a efficient manner.
                                    H-449
-------
119    EPA Oversight of Federal Facility Cleanup of Radiologically
   Contaminated Mixed Waste Sites under the Superfund and RCRA Programs


   Melanie S.  Barger,  Chemical Engineer, Office of Waste Programs
   Enforcement/Federal Facility Hazardous Waste Compliance Office,
   United States Environmental Protection Agency, 401 M Street, SW,
   Washington, D.C.  20460


       The Department  of Energy (DOE)  consists of 17 Defense Production
   Facilities.  Many of these facilities have sites that are
   contaminated with both radioactive  and hazardous contaminants.   EPA
   is using its Superfund and Resource Conservation and Recovery Act
   (RCRA) enforcement  authorities to oversee cleanup at many of these
   facilities.  DOE implements the Atomic Energy Act (AEA) through DOE
   orders for the radionuclides.   This paper will summarize the
   integration and use of all the RCRA and Superfund authorities for
   cleanup of the sites.  State involvement in the agreement and
   oversight process will be discussed.   A brief discussion of the
   EPA enforcement status of the  cleanup progress at DOE's weapons
   complex facilities  will be given.
                                   H-450
-------
120           HAZARDOUS WASTE INCINERATION ENFORCEMENT PROGRAM
    ABSTRACT

         The  Hazardous  and  Solid  Waste  Amendments   (HSWA)  to  the
    Resource  Conservation  and Recovery Act (RCRA)  required the EPA to
    either issue or deny permits for hazardous waste incinerator (HWI)
    facilities  before November 8, 1989.   Since  most of the operating
    incinerators are  permitted recently,  the EPA is strengthening its
    enforcement program in inspecting these facilities.

         This paper describes the current HWI universe, the status of
    applicable  regulations,  the  efforts put  forth by the  Office  of
    Waste  Program  Enforcement  to  support the EPA  regional  and  state
    HWI  enforcement  programs,  and the  past and  current  enforcement
    actions.
                                     n-45i
-------
121           Evaluation of the Draft High Concentration, Multi-media Protocol
                              Versus An  Historical  Database

      Joe Lowry, Ed Bour, U.S. EPA, National Enforcement Investigations Center,
      Box 25227, Denver Federal Center, Denver, CO 80225

      Timothy J. Meszaros*,  Lockheed Engineering & Sciences Company, 1050 E. Flamingo
      Rd. ,   Ste.  120,  Las  Vegas,  NV  89119,  under  contract  with  the  U.S.  EPA,
      Environmental Monitoring Systems  Laboratory, Las Vegas, NV
      The U.S. Environmental Protection Agency (U.S. EPA) has developed a draft

      protocol  for  the evaluation  of high concentration  wastes.  These  wastes are

      generated from hazardous waste  disposal  sites   with  the analyses conducted at

      Contract Laboratory Program (CLP) facilities in the majority of cases. Analytical

      services  under  this  protocol   include  pneumatic  nebulization  and  hydride

      generation inductively coupled plasma emission (ICP) spectrometric analyses. The

      pneumatic nebulization  technique  for  ICP involves preparation  of  the samples

      using a potassium hydroxide (KOH) fusion process. An historical database for 738

      high concentration samples and the quality control associated with measurement

      of  those   samples   was   developed  at   the  U.S.   EPA  National  Enforcement

      Investigations Center (NEIC) during the development of the KOH  fusion preparation

      technique. It will be demonstrated that two elements,  titanium and molybdenum,

      should be included in the list of target  analytes as both elements are found in

      a significantly large number  of  high concentration samples at levels above 1000

      mg/kg. Data will be presented to show that barium and calcium may not meet the
                                                      /
      Contract Required Quantitation  Limits specified7by  the protocol.  The possible

      need to adjust the accuracy control limits to allow KOH fusions to be employed

      on soil and oil matrices will  be  discussed. Silver may not meet acceptable matrix

      spike  recoveries  with  the   spike  levels   suggested  by  the  protocol.  The

      interference check sample called for in the  high concentration protocol will be

      evaluated.
                                           n-452
-------
Notice: Although the  research described in this article has been
supported by  the Environmental  Protection  Agency under contract
68-03-3249 with Lockheed Engineering & Sciences Company, it has not
been subjected to Agency review and therefore does not  necessarily
reflect the views of the Agency and no official  endorsement should
be inferred.
                              H-453
-------
H-454
-------
AUTHOR INDEX
-------
AUTHOR INDEX
Author
  Paper
Number
Author
                                                                      Paper
                                                                    Number
                                       Author
 Paper
Number
Abraham, B	50
Ainsworth, C	106
Alchowiak, J	12
Alforque, M.M	100
Allen, H	49
Aim, P.M	108
Anderson, B	54
Anderson, D	54
Asowata, C.A	68
Atwood, R.A	90
Baca, J.L	37
Baldin, E	77
Barger, M.S	119
Barich ffl, J.J	11
Bath, R J	8
Beckert, W.F. 	58,77,78
Behymer, T.D	52
Bellar, T.A	52
Benedicto, J	77
Berg,S	5
Berges, J.A	13,76
Bergman, F. 	18
Betowski, L.D	51,55,75,82
Bisson, D.L	105
Blackburn, W.B	3
Bone, L.1	110
Boomer, B	18
Bour, E	121
Boyer, D	31
Boyer, D.S	59
Brent, L	23
Brockhoff, C.A	93
Brown, M.A	53
Bruce, M.L	64
Buchanan, M.V.  	107
Budde, W.L	52
Buhle, H.G	69
Buirge, A.W. 	36
Calderoni, J	14
Campbell, R.M	61
Carlson, R.E	36
Carr, V. 	23
Caton, Jr., J.E	65
Chiang, T.C	75
Chamerlik, M	36
Chan, S.K.  	86
Chen, P.H	70
Chow, E.S	120
Coakley, W. 	25
Comeau, R	92
Cornet, H	71
Cramer, P.H	63
Creelman, L.W.  	15
Cross, K.  	59
Cuccherini, B.A	110
Davis, C.B	16,17
Davis, W.R	US
Dawson, T.L	110
DeMars, B	23
          DeWald, J	72,112
          DiGiulio, D.C	45
          Dias, F. 	94,97
          Dodd, J.A	4
          Dodhiwala, N.S	78
          Dovi, R	94,97
          Downs, J	101
          Draper, W.M	67
          Drinkwine, A	35
          DuBose, G. J	102,103,104
          Dugan, T.  	23
          Dux, T.P. 	18
          Dymerski, M	66
          Edwards, M.D	65
          Eichelberger, J.W. 	63
          Erickson, D.C	73
          Felmy, A	106
          Ferro, S	74
          Fitzgerald,!	2
          Fleming, G.S	65,79
          Fletcher, T.A	98
          Floyd, T.  	96
          Fribush, H	33,47
          Garcia, M.E	65
          George, G	14
          Gibbons, R.D	19
          Gilbert, C.W. 	44
          Gordon, N	73
          Grams, N.E	19,20
          Greenlaw, P.D	8
          Griest, W.H	65
          Grillo, A	96
          Guerin, M.R	107
          Gurka, D.F.  	56
          Guterriez, J	102,103,104
          Hable, M.A	68
          Hamilton, J	16
          Hansen, G	102,103,104
          Harmon, S.H	79
          Hartwell, S	102,103
          Hawthorne, S.B	60
          Hecht, C.R	62
          Heithmar, E.M	21,85
          Hernandez, T. 	23
          Hill, J.G	38
          Hillman, D	31
          Hinners, T.A	21
          Hinners, T.A	85
          Ho, J.S	52
          Hoffmann, K.D	42,83
          Hoffman, E.J	22
          Hewlett, L	15
          Hull, D.R	90
          Hull, K. 	23
          Humphrey, A	49
          Ilgner, R.H	107
          Jackson, D.R	105
          Janowski, J.W. 	42,83
          Jarke, F.H	19,116
                                       Jennings, K.F. 	42,83
                                       Johnson, G.L	10
                                       Jones, L	12
                                       Jones, R.R	11
                                       Jones, T.L	51,75
                                       Kaelin , L.P.  	48
                                       Kantor, E. J	13
                                       Kapustka, L.A	39
                                       Karu, A.E	36
                                       Kell, R.A	95,98

                                       Kendall, D	91
                                       Kim, I.S	53
                                       Kimbrough, D.E	87
                                       Kingsbury, G	46
                                       Kittikul, P. 	109
                                       Klesta, E.J	1
                                       Kobus, M.A	62
                                       Koglin, E.N	41
                                       Kohorst, K	31
                                       Kolopanis, J	6
                                       Kuehn, J.D	43,80
                                       Labor, L	108
                                       Langenfeld, J.J	60
                                       Lawson, T. 	66
                                       Leach, L.E	.,	45
                                       Lee, R.P. 	64
                                       Lee, Y.J	76
                                       Leibman, C.P. 	29,37
                                       Lenssen, G	16
                                       Levy, J.M	59
                                       Lim, A.K	36
                                       Linder, G	39
                                       Loconto, P.R	24
                                       Loehr, R.C	73
                                       Loeper, J	40
                                       Lopez-Avila, V. 	35,58,77,78
                                       Lowry, J	121
                                       Ma, T.H	114
                                       MacMinn, H.M	32
                                       Marcus, M.F.	6,62
                                       Marsden, P.J	55,57,82
                                       Maskarinec, M.P. 	79
                                       Mateo, J.M	25
                                       Mayahi, M	72
                                       McKee, T.M	Ill
                                       McNichols, R	16,17
                                       McPherson, C.A	Ill
                                       Meenihan, G.S	Ill
                                       Meszaros, T.J	121
                                       Milanos, J	77
                                       Miller, A.G	26
                                       Miller, D.J	60
                                       Miller, M	2
                                       Morotti, J.M	27
                                       Moul, R	66
                                       Murarka, I.P. 	106
                                       Neptune, D	...9
                                       Neulicht, R	18
                                                       H-455
-------
Author
  Paper
Number
Author
  Paper
Number
Author
                                Paper
                              Number
Tatro, M.E.  	**» 8V
                        	72
Newberry, W.R.	31
Northington, DJ	99
Notich, MJ)	7
O'Brien, K.0	5
O'Dell, J.W.	93
O'Quinn, C.M	42,43, 80,83
Osborn, R.J	95
Pace, C.M	.55,82
Pandit, N	25
Papp, M	.31
Peak, R	5
Perry, G	40
Perugini, F. 	112
Pfaff, J.D	„	93
Pierce, M.A	23
Polansky, G	104
Pospisil, P.A	62, 90
Poziomek, E.J	„	41
Prange, N.E	110
Pritchett, T.H	48,49,117
Rai, D	106
Rajani, A.H	38
Raymer, J.H	81
Rehm, R.P.	22
Richards, M	61
Riviello, J.M	85
Robbat Jr., A	50
Roberts, D.F. 	70
Robertson, G.L	69, 76
Roby, M.R	55, 82
Rodriguez-Padro, J	15
Roehl, R	100
Rollins, C.A	28
Roman, M	94, 97
          Rose, S	..71
          Rosenbacher, D	6
          Ross, R.R	45
          Rosseili, A.C	..59
          Roudebush, W.  	43,80
          Roudybush, L	1
          Rowan, J.T.	85
          Rzeszutko, C.P.  	29
          Sadowski, P. 	16,17
          Sasinos, F.I	53
          Sauter, A.D	101
          Schenley, R.L	65
          Schmidt, D.J	36
          Sheldon, L.S	34,81
          Shelton, M	99
          Sherman, L	16,17
          Shmookler, M	42,43,80,83
          Simes, G	3
          Slovaeck, M.J	15
          Smith, P. 	118
          Smith, T. 	113
          Solecki, M.F. 	48
          Spear, R.D	8
          Spurlin, S.R	35,108
          Stephens, M.W. 	64
          Stephens, R.D	53
          Stern, C.M	68
          Stewart, J	84
          Stewart, R	115
          Stockwell, P. 	92
          Stoub, K.P.  	19
          Surdo, T.L	117
          Swanson, T.A	36
          Syslo, J	117
                                        Tellez, D	
                                        Thomas, F.  	42,43,80,83
                                        Tomkins, B.A	65
                                        Trenholm, D	I8
                                        Tsiagkouris, L.A	29
                                        Tuschall, J	40
                                        Van Bueren, D.L	71
                                        Van Emon, J	35
                                        VanAusdale, W.A	70
                                        Vanderveer, E.P. 	~	.37
                                        Velez, G.R	81
                                        Vincent, H.A	31
                                        Wachter, LJ	65
                                        Wakakuwa, J	87
                                        Wallace, J	3
                                        Warden, B	94,97
                                        Weathington, B.C	32
                                        White, P.  	103,104
                                        Wiggenhauser, G.W. 	98
                                        Wijekoon, D	67
                                        Williams, K.R	1Q5
                                        Wilner, J	63
                                        Wise, M.B	107
                                        Wolf, M.A	37
                                        Worthington, J.C	33,116,118
                                        Wunsch, D.M	7
                                        Wynn, L.H	10
                                        Wynnyk, R.E	48
                                        Xyrafas, G	50
                                        Yanak, M.M	86
                                        Yoder, R	94
                                        Zully, R.S	95
                                                        n-456
-------