5383
                DEVELOPMENT AND APPLICATION OF TEST PROCEDURES FOR
                 SPECIFIC ORGANIC TOXIC SUBSTANCES IN WASTEWATERS
                               DRAFT FINAL REPORT
                                                           RECEIVED
                                                               FEB29  1980
                                                               LIBRARY, REGION V
                                                                            AGENCY
                                       by

                                John W. Rhoades
                               Richard E. Thomas
                               Donald E. Johnson
                         CATEGORY 1:  Phthalate Esters
                            EPA Contract 68-03-2606
                           SwRI Project 01-5068-101
                                 Prepared for

                   Mr. James E. Longbottom, Project Officer
                  Environmental Monitoring & Support Laboratory
                        Environmental Protection Agency
                            Cincinnati, Ohio  45268
                                 January 1980
               SOUTHWEST    RESEARCH
INSTITUTE

-------
It

-------
                            DISCLAIMER
     This report has been  reviewed  by  the  Environmental  Monitoring
and Support Laboratory, U.S. Environmental Protection  Agency,  and
approved for publication.  Approval  does not  signify  that the  contents
necessarily reflect the views and policies of the  U.S.  Environmental
Protection Agency, nor does mention  of trade  names or  commercial  products
constitute endorsement or  recommendation for  use.

-------
                                    CONTENTS
Abstract	iv
Figures	vi
Tables	:	vii

   1.  Introduction 	   1
   2.  Conclusions	3
   3.  Recommendations	5
   4.  Phase 1	7
          Objectives	7
          Program Direction 	   7
          Experimental	8
             Literature Search	8
             Gas Chromatography 	   9
             Extraction Study 	  13
             Preservation Study 	  25
   5.  Phase II	37
          Objectives	37
          Program Direction 	  37
          Liquid-Solid Column Chromatography	38
          Wastewater Application	42
          Accuracy and Precision	52
             Approach	52
             Results	53
             Summary	56

References	57
Appendices

   A.  Key References from Literature  Search	58
   B.  Statistical Analyses Phase I  Extraction Study	59
   C.  ANOVA Tables Phase I Extraction Study	66
   D.  Chromatograms of Wastewater Fractions - Phase II  	  73
   E.  Chromatograms of Wastewaters  1  through 5 Single
          Fraction Collection - Phase  II	76

-------
                              FIGURES

Number                                                      Page
  1.   Standard for DMP,  DEP  and DBP	10
  2.   Standard for BBP,  DEHP and OOP	11
  3.   Unspiked water extract (100% DCM)  	  17
                               vi

-------
                              TABLES

Number                                                       Page
  1.   Retention Times	12
  2.   Sensitivity Values	12
  3.   Quantities of Spike for Extraction Study	14
  4.   Extraction Study Results - 100% DCM % Recovery.  ...   19
  5.   Extraction Study Results - 15% DCM in Hexane
          % Recovery	19
  6.   Summary of Extraction Study Analyses	26
  7.   Results of Preservation Study - DMP, DEP,  DMP  .  .  .  .   29
  8.   Summary of Results of Preservation Study	35
  9.   Recovery {%) from Florisil PR Column  	   41
 10.   Recovery (%) from Alumina(N) Column 	   41
 11.   Analysis of Wastewater 1 for Phthalates 	   45
 12.   Analysis of Wastewater 2 for Phthalates	45
 13.   Analysis of Wastewater 3 for Phthalates	46
 14.   Analysis of Wastewater 4 for Phthalates 	   46
 15.   Analysis of Wastewater 5 for Phthalates 	   47
 16.   Analysis of Wastewater 1 Dosed with Phthalates.  ...   48
 17.   Analysis of Wastewater 3 Dosed with Phthalates.  ...   49
 18.   Analysis of Wastewater 4 Dosed with Phthalates.  ...   50
 19.   Analysis of Wastewater 5 Dosed with Phthalates.  ...   51
                              vii

-------
                            SECTION 1
                           INTRODUCTION
     This project was part of an overall program of the Environmental

Protection Agency to establish test procedures for the analysis of

114 toxic substances in wastewaters.  These toxic compounds have been

separated into 12 categories as follows:

     Category 1:   Phthalate Esters
     Category 2:   Haloethers
     Category 3:   Chlorinated Hydrocarbons
     Category 4:   Nitrobenzenes and Isophorone
     Category 5:   Nitrosamines
     Category 6:   Dioxin
     Category 7:   Benzidines
     Category 8:   Phenols
     Category 9:   Polynuclear Aromatics
     Category 10:  Pesticides and PCB's
     Category 11:  Purgeables
     Category 12:  Acrolein, Acrylonitrile, Dichlorodifluoromethane

     This is the report on Category 1:  Phthalate Esters.

     The specific compounds studied in this effort were:

     1.  Dimethyl Phthalate (DMP)  4.  Benzyl Butyl Phthalate (BBP)
     2.  Diethyl Phthalate (DEP)   5.  Diethylhexyl Phthalate (DEHP)
     3.  Dibutyl Phthalate (DBP) -  6.  Dioctyl Phthalate (OOP)

     It was desirable that common sample treatment for the various

categories be employed, where possible, to minimize cost of analysis

of unrelated compounds in any given water sample.  The efforts reported

under the following performance headings were designed to provide

information relative to this common purpose.

     The study was conducted in two phases.  In Phase I, work was

conducted with clean water and was intended to provide information

                                   1

-------
which would give direction  to Phase  II  work,  conducted on actual  wastewaters,
and serve as a comparison base  for  the  information developed.
     The ultimate objective was  to  develop  a  method,  having a maximum
of applicability, which  could be  published  in the Federal  Register
and carried out in most  analytical  laboratories.   Other objectives
were to develop accuracy and precision  information for the various
steps in the method,  to  develop  procedure variations  as required by
a particular wastewater, and to  maximize  the  use  of common sample
treatment steps for all  of  the  categories of  organic  toxic substances
listed above.

-------
                            SECTION  2
                           CONCLUSIONS
     The six phthalate  esters  can  be  gas  chromatographed very well.
Sensitivity quantities  (producing  a  response  10X noise level) for
the electron capture  detector  range  from  20 to  130  picograms.  Sensitivity
quantities for the flame  ionization  detector  are  roughly 200-700 times
greater than for the  electron  capture detector.
     Extraction of the  phthalate esters  from  clean  water at low parts
per billion (yg/L) concentrations  when the  pH was 2,  7,  and 10 was
generally 90 percent  or better when  using 15  percent  DCM in hexane
or 100 percent DCM as the  extracting  solvent.
     At the concentrations mentioned  above, some of the  conditions
of the preservation study  produced notable  effects  on some of the
phthalate esters.  DMP  and BBP extraction recoveries  were reduced
to 30 percent and 40  percent,  respectively, after  storage for 7 days
in the dark at pH 10  and 24C.  Storage  at 4C  under  acid conditions
proved satisfactory.
     Recoveries from Florisil  and  alumina cleanup columns were excellent
if the adsorbent was  slightly  deactivated by  the addition of 3 percent
water.
     Five wastewaters were analyzed  and  none  was found to be completely
free of phthalate esters.  The number of  phthalate  esters and amounts

-------
did vary considerably with  the  different  wastewaters.
     The recoveries of  the  phthalate  esters  dosed into the wastewaters
was, in general, good.   Recoveries  of selected phthalate esters was
somewhat reduced with some  wastewaters and marked reduction in recoveries
from some wastewaters was noted after storage of the dosed wastewaters
for 7 days at 4C.
     The principal problems encountered with the method are contamination
of glassware and nonlinearity of response of the DMP and DEP with
the electron capture detector.
     The method is basically acceptable for  use on wastewaters having
a low level of interferences.   It does not work well when background
levels are medium to high.

-------
                            SECTION  3
                         RECOMMENDATIONS
     The gas chromatographic  column  used  in  this  effort (1.5  percent
SP-2250 + 1.95 percent SP-2401 on 100/120 mesh  Supelcoport),  although
satisfactory for separation of all six  phthalate  esters,  is  only one
of many GC columns that could be used.  The  six phthalate esters are
separated primarily on the basis of  boiling  point and many other column
packings would be satisfactory.
     The electron capture detector,  being much  more sensitive to the
phthalate esters than the flame ionization detector,  is recommended
for most wastewater applications.  The  electron capture detector would
be less desirable with wastewaters contaminated with  chlorinated compounds.
     Either 15 percent DCM in hexane or pure DCM  is satisfactory for
extraction of phthalate esters from  wastewaters in the pH range of
2-10.  If wastewater is to be stored prior to  analysis for phthalate
esters, the water should be adjusted to acidic  conditions and the
samples should be stored in a cold (4"C)  room.  A preliminary evaluation
of the wastewater would still be necessary to  ensure  that there was
no degradation under these conditions.
     Florisil and alumina used in column  chromatography must  be deactivated
by addition of 3 percent water to insure  recovery of  all  phthalate
esters.

-------
     Great care should be exercised  in  preparing  the  reagents  and
glassware used in the analysis.  Phthalate  esters,  due  to  their occurrence
in many manufactured products, are a common source  of sample contamination
in the laboratory, and frequent reagent blanks  should be  run to establish
levels of contamination.
     Only chemists experienced with  the method  and  with trace  organic
analysis in general should conduct analysis for phthalate  esters.
     The method is recommended primarily for wastewaters with  only
low levels of contamination.  The collection of two fractions,  instead
of one, from column chromatography cleanup  may  be helpful  in some
cases and is recommended when higher levels of  early  eluting interferences
are suspected.

-------
                            SECTION 4
                             PHASE I
OBJECTIVES
     The objectives of Phase I were  limited  to:
1.   Review of scientific literature  for  key  references  relative  to
     determination of the phthalate  esters  in water  and  other  matrices.
2.   Determination of the best gas chromatographic columns  and conditions
     for detection of phthalate esters when  using the  detectors of
     interest.
3.   Determination of extraction efficiency  for  phthalate esters  from
     clean water samples.
PROGRAM DIRECTION
     Before experimental work began,  representatives of  SwRI met  in
Cincinnati with the project officer  and others of concern in the  EPA
Cincinnati laboratories to discuss and decide on certain specific
aspects of the activities in Phase I.  In addition to  the decisions
reached on GC columns, extraction solvents,  and  other  experimental
details which will be given in subsequent sections of  this  report,
guidelines were developed as follows:
       *  Unusual, exotic, or overly  expensive items were not  desirable.
          Simplicity was to be preferred wherever the  option existed.
          For example, mass spectrometry  and  WCOT or SCOT columns
          were ruled out.

-------
       *  Use of chemical  treatment  to  produce  a  derivative  which
          cannot be unequivocally  related  to  the  parent  substance
          was not acceptable.
       *  Quantisation  by  electronic  integration  of  peak  area  or by
          peak height measurement  was acceptable.
EXPERIMENTAL
Literature Search
     A comprehensive literature  search  was made using  a  terminal  linked
to Lockheed Information Systems' DIALOG*.   Chemical  Abstract Condensates
(CAC) were searched for the years  1972  through  the present.  Also,
NTIS (1964 to present), Enviroline (1971 to present),  and Pollution
Abstracts (1970 to present) were searched.  In  some  cases, both  on-
line and off-line prints were obtained,  the former in  order  to rapidly
obtain a small enriched set of references  to  work with and the latter
to provide comprehensive coverage  by  retrieval  of a  larger set of
references.
     The literature search as conducted yielded over 150  references.
Examination of titles and  in many  cases  abstracts drastically  reduced
the number to those listed in Appendix A as being of interest  to  this
program.  The papers of G. S. Giam and  his  coworkers at Texas  A&M
are of particular interest.  The paper  by  Giam  et al.  entitled "Sensitive
Method for Determination of Phthalate Ester Plasticizer  in Open-Ocean
Biota Samples" (Anal. Chem., Vol.  47, No.  13, November 1975) was probably
the most important product of the  literature  search.   The authors
report several possible laboratory sources  of phthalate esters which

-------
can/could result 1n contamination of  samples.  They  also  report  the
necessity of deactivation of Florisil  to  prevent  loss  of  DEHP when
column chromatography employing Florisil  is  used  in  a  sample  cleanup
procedure.
Gas Chromatography
     All six (6) phthalates have been  chromatographed  on  two  columns.
The primary column (Column 1)  is 1.8-m x  4-mm  ID  glass tubing packed
with 1.5-percent SP-2250 plus  1.95-percent SP-2401 on  100/120 mesh
Supelcoport.  The secondary column  (Column 2)  has 3-percent OV-1  as
its liquid phase but is otherwise the  same as  Column 1.
     Initial investigations with Column 1 indicated  that  all  six  phthalates
could be resolved at a column  temperature of 200C.  However, the
retention time for OOP was excessive  (34  minutes) and  the early  eluting
phthalates would be difficult  to quantitate  due to the proximity  of
the solvent peak.  This would  be even  more critical  on extracts  of
wastewater where the early eluters  would  be  more  likely to co-elute
with interferences.  Therefore, the six phthalates were divided  into
two groups of three for chromatography.   The "low" temperature (160C)
group includes DMP, DEP, DBP,  while the "high" temperature (225C)
group includes BBP, DEHP, and  DOP.  Table 1  gives the  retention  times
of the six phthalate esters on Column  1.  Example chromatograms  of
these compounds are shown as Figures 1 and 2.

-------
. 1 .
1 ; 
 i i


;-4+
^ 7-1
I  
r
f
^-r-
. 

f iJ.tj 	 H-; i. i.t,i;i;.i.-u-i--j 	 ri--~ r
-4i-^J:;"
-1 	



Co
 |  . ; i


j; OMP  *^ 	
i-H-'
I -; .
1 ^L |
\^\
l+~<
-! 	 iJ

2
0P
~. 
-!-
4-
i--' '-i

--'""-u
r ! ! i '
1 ' '  1
i '' i
OH>
OEP
DSP
turn 1
 ! : :



r ' '
1

.-J-K'i ._). 	
- 10 n
- 10 n
- 10 n
- 160-
.;.; : i
FS^
+H-,
i  i  <
^-~


ill'
h-H-
I ; | j
i ; i . i

9
9
9
C - 16
_

(
"OBP
F=4I|-
1 1 ;
-Tj i '

46 3 10 1
minutes
k
*
 T.(f~
  
i I
:
pr
rr-
rH-frJ
, ; 1 M
 	
.

 i
-i >-h
I'l'j"' ll
1 , '
^^d.
1
-  1
4
Figure 1.  Standard for DMP, DEP and DBP,
                     10

-------
               i lH.   r . io-"?i - .
   0    2-4    6     8    10   12
                          BBP - 0.6 ng
                          OEHP - 6.0 ng
                          OOP - 9,8 ng


                       O Colunn 1 - 225'C - 16X - EC
Figure 2.    Standard for BBP, DEHP and OOP
                       11

-------
                         TABLE 1.  RETENTION TIMES
                           (Column 1; 60 mL/min)
Phtha
DMP
DEP
OBP
BBP
DEHP
OOP
late Ester
> 160C
> 225C
Minutes
1.9
2.9
12.6
4.1
5.1
9.0
     Investigations with Column 2 gave  comparable  results,  leaving
little basis for recommending either column  over the  other.   Both
columns perform satisfactorily, although  somewhat  greater  sensitivities
were obtained with Column 1  (see below).
     At the beginning of this work, the electron capture detector
was considered to be the primary detector to  be used  in the  analysis
of phthalates, and the flame ionization detector was  designated  the
alternate detector.  Experimentation confirmed that  the electron capture
detector is preferred over the flame ionization detector on  the  bases
of the greater sensitivity and selectivity.   Table 2  gives  comparative
sensitivity values for the two detectors.
                     TABLE 2.  SENSITIVITY VALUES
  (Amount in nanograms required to  produce response  1QX noise level)
Detector
Phthalate Ester
DMP
DEP
DBP
BBP
DEHP
OOP
Electron
Column 1
0.11
0.13
0.02
0.02
0.04
0.11
Capture
Column 2
1.3
1.9
0.3
0.5
0.7
0.9
Flame Ionization
Column 1
18.8
31.3
13.9
15.3
20.3
31.3
                                   12

-------
     The gas chromatographs used in this work were  a  Hewlett-Packard
5710A with a Ni-63 electron capture detector and  a  Hewlett-Packard
5710A with a flame ionization detector.  Operating  conditions  of  these
instruments were:
                               	ECD	         FID
     Oven temperature          variable                 variable
     Detector temperature      300C                    2508C
     Carrier gas               5% methane in argon      helium
     Carrier gas flow rate     60 mL/min                60  mL/min
     Flame gases                                     H2  + air
Extraction Study
     The extraction study was initiated to  determine  the  recoveries
of the six phthalates of interest from clean water  at pH  2, 7,  and
10 using 15 percent dichloromethane  (DCM) in hexane and 100 percent
DCM as the extracting solvents.
     The water used in  the extraction study is  a  naturally buffered
well water obtained from the SwRI supply line prior to chlorination.
Samples of this water were collected and transported  to the laboratory
in empty solvent bottles which had previously contained Burdick and
Jackson solvents.  The  pH of the untreated  water  is close  to 8.  Adjustment
to the required pH's was accomplished by adding strong acid or base
as follows:
               pH 2  -  2.5 ml 18N H2S04
               pH 7  -  0.1 ml 18N H2S04
               pH 10-0.2 ml ION NaOH
The water was found to  be very low in electron  capture sensitive  materials
as determined through comparison of an extract  of the water with  a
glassware blank.  Chromatograms in these instances  were very similar,
strongly indicating that the water could not be the source of  any
                                    13

-------
large ir .erferences.
     The water was dosed with  acetone  solutions  of  the  six  phthalate
esters, each solution containing 100 times  the  sensitivity  amount
per micro!iter, as determined  earlier  for each  phthalate.   By  adding
100 vL of the appropriate acetone  solution  to a  liter of water,  dosed
water samples for extraction were  prepared.
     Poor reproducibility, contamination problems  and loss  of  sensitivity
after replacement of a  defective EC detector prompted an increase
in dosing levels after  two attempts failed  to get  usable data  from
the extraction study.   Only the data from the third effort  are reported,
as shown in Table 3.
             TABLE  3.   QUANTITIES  OF SPIKE FOR EXTRACTION STUDY
Phthalate
DMP
DEP
DBP
BBP
DEHP
OOP
Dosing Solution
ug/mL
100
100
100
6
60
97.5
Cone, in l Liter
ng/mL
10
10
10
0.6
6.0
9.8
     The one-liter samples of water were  dosed  while  in  one-liter
Erlenmeyer flasks, then poured  into two-liter  separatory funnels  for
extraction by either 15 percent DCM in  hexane  or  100  percent DCM  in
accordance with the procedure found in  the  Federal  Register, Volume
38, No. 125, Friday, June 29, 1973, page  17320.   Three  dosed samples
were extracted at pH 2 and 10 with each solvent.  At  pH  7,  four dosed
samples were extracted with each  solvent.   Thus,  20 samples  were  extracted
for each group.  The final volume of  the  extract  concentrate was  always
                                     14

-------
            10 ml, and 10 yL of the concentrate  was  always  Injected into the GC.
            At the time' dosing of the  samples  occurred,  100 uL of the dosing solution
*          was added to 10 ml of hexane which then  served  as the standard against
            which all recoveries were  measured.
                 The Federal Register  procedure  calls  for the separated extracts
          to be drained through three to  four  inches of anhydrous sodium sulfate
            into a Kuderna-Danish flask.  Frequent  plugging of the tube containing
            the sodium sulfate was encountered.   Usually unplugging could be achieved
          by applying a slight pressure to  the top of  the tube by means of a
            2-ounce rubber bulb.  It was discovered  that the column of sodium
            sulfate could be decreased in height to  3  to 4  centimeters without
*          any obvious effect on the  analytical  results and with a more rapid
            delivery of extract into the Kuderna-Danish  flask.  In the course
            of unplugging the sodium sulfate  tube,  a small  amount of water was
at
            sometimes forced into the  Kuderna-Danish flask.  This situation produces
            no harmful effects unless  the quantity  is  so great as to cause an
            intolerable volume error in determining  the  final 10-mL volume or
            unless the water phase should be  accidentally injected into the GC.
                 When the extracting solvent  is  100  percent DCM, it is necessary
            to remove essentially all  of the DCM prior to analysis.  This was

            done by taking the extract to a volume  of  10 to 15 ml, adding 75 to
            100 ml of hexane, and then reconcentrating to the final volume.  The
            DCM extracts usually produced a wider "solvent" peak than those produced
            by 15 percent DCM in hexane.  This peak  broadening was not reduced
            when the amount of hexane  added before  reconcentration was increased
                                                15

-------
in amounts up to 200 ml.  The peak  broadening  incurred  is  not a serious
problem, however.
     Occasionally, an emulsion  resulted  during the extraction process.
This occurred usually with the  15 percent OCM  in  hexane solvent and
at pH 2 and 7.  After draining  off  most  of  the water phase,  the emulsion
could be broken almost always by shaking or swirling the separatory
funnel.  When emu!sification occurred, the  solvent phase was usually
not as well separated from the  water  phase  as  when emu!sification
did not occur, resulting  in a higher  rate of plugging of the sodium
sulfate tube.
     The most serious problem encountered was  the inability  to obtain
chromatograms of blanks for water,  glassware and  sodium sulfate which
were devoid of interfering peaks.
     The most consistent  interfering  peaks  were observed at  the same
retention times as BBP and DEHP.  Other  interfering peaks  with retention
times matching the other  four phthalates were  present at a lesser
frequency.  Figure 3 is a chromatogram (at  225C)  of an unspiked water
extract using DCM.  This  chromatogram illustrates the magnitude of
the contamination problem.
     The interfering peak from  Figure 3  was concentrated and analyzed
by GC/MS to determine the nature of the  contaminant. This interference
was identified as an adipate, possibly the  ethylhexyl.
     Elimination of contamination was never completely  accomplished
and periodically caused serious problems with  the phthalate  determinations,
Whether glassware was cleaned and heated to 275C overnight  before
                                    16

-------
         I
          -M-
                      -C Column 1 - 225'C - 8X
              zn  SBP q: OEHP (?n
                                 EC
                               OOP
        0   2    4   6    8    10   12
Figure  3.  Unspiked water extract (100%  DCM)
                          17

-------
use or was heated for several  hours  at 450C  before use seemed to
have little effect.  Leaving the  glassware  in a chromic acid solution
(40 mL saturated I^C^Oy - H2S04  solution  in  1  liter of water) for
12 to 18 hours and rinsing several times with deionized water, acetone
and petroleum ether prior to use  seemed  to  give the best results.
Early in the study, the sodium sulfate was  heated overnight at 450-
500C before use.  Some improvement  resulted  when this practice was
supplanted by Soxhlet extraction  of  the  sodium sulfate with DCM for
two days.
     The sodium sulfate, the holding tube  and the glass wool  plug
were all rinsed with DCM and hexane  immediately before use.
     After increasing the phthalate  dosing  levels,  the problem with
contamination became manageable,  and the data acquired during the
third extraction effort allowed reliable conclusions to be made with
respect to the extraction efficiencies of  the solvents tested.
     Percent recovery data were obtained by both computer integrated
peak areas and by peak height  measurement.   Only data based on peak
height measurement are reported here.
     The data acquired in the  extraction study are  presented in Tables
4 and 5.  The data have not been  corrected  for blank extractions.
Data Analysis
     The results of the extraction study were submitted to statistical
analysis to determine which of the factors  affected the recovery efficiency
of the method.  The statistical model  for  these data was a two-way
analysis of variance (ANOVA) defined by:
                                    18

-------
     TABLE  4.    EXTRACTION STUDY  RESULTS - 100%  DCM % RECOVERY

pH
2



7




10



Extract
1
2
3
Mean
1
2 .
3
4
Mean
1
2
3
Mean
DMP
96
125
67
96
135
143
144
115
134
108
119
113
113
OEP
89
98
71
86
103
106
103
99
103
100
115
111
109
DBP
63
81
40
61
97
95
95
94
95
96
100
88
95
BBP
88
88
90
89
113
110
109
98
109
91
91
90
91
DEHP
102
92 
101
98
104
99
98
98
100
100
99
112
104
OOP
94
90
96
93
99
97
96
94
97
90
95
94
93

TABLE 5.   EXTRACTION STUDY RESULTS - 15% DCM  IN  HEXANE % RECOVERY

oH Extract
2 1
2
3
Mean
7 1
2
3
4
Mean
10 1
2
3
Mean
OMP
107
105
100
104
116
111
109
104
110
110
110
113
111
DEP
104
101
99
101
101
101
103
101
102
104
103
104
104
OBP
100
87
87
91
104
_
104
97
102
95
93
95
94
BBP
93
92
84
90
_
90
85
94
90
98
94
98
97
OEHP
99
96
127
107

101
104
98
101
105
102
112
110
OOP
95
95
91
94
91
92
91
93
92
97
96
99
97

     - data not available - contamination  (?)
                                   19

-------
          Yijk  u * Ti +  pj  +  Tpij  + eijk
where     Yjik - observed  recovery  for solvent system i and pH j on
            J   replicate  k,  1-1,2,  j=l,2,3,  k=l,2,3 or 1.....4
          u - overall mean
          TJ - effect due  to  solvent system i
          PJ - effect due  to  pH j
          Tpjj - interactive  effect between solvent i and pH j
and       Ejjjj - random error associated with Y^
Both solvent and pH are fixed effects, and  all  F-ratios were calculated
using the mean square for  error as  the denominator.  The 5-percent
significance level was used on  all  tests.
     The interaction between  solvent and pH,  if significant, implies
that the combination of a  particular pH and a particular solvent performed
better  (worse) than could  be  predicted by using the results from the
two factors taken separately.   For  example, if a difference of 7 percent
is observed between solvents  and 6  percent  between pH's, then using
the best combination would result  in a predicted difference of 13
percent, the sum of the two differences.  If the observed difference
were to be 20 percent, then the interaction term in the ANOVA would
be significant.  This merely  implies that the best estimate of the
expected recovery should be obtained from the results on that particular
combination, not from the  pH  or solvent means themselves.  Interaction
is used in terms of a data analytical  rather than physical  approach.
     A  factor or interaction  is considered  to be significant if an
F-ratio calculated as part of the ANOVA procedure exceeds the corresponding
tabled  value at the 5-percent significance  level.  This implies that
                                     20

-------
there was more variability  among  the  levels  of a factor, e.g., among
the three pH's, than there  was  among  replicate analyses under identical
conditions.  In some cases,  however,  the  precision of the individual
analyses allows differences  to  be termed  significant that are too
small to be of any practical  importance.   This should be considered
in evaluating the results contained  in  this  section.
     Separate ANOVA's were  conducted  for  each of the compounds in
the group, and the results  were then  synthesized to give the recommended
extraction conditions.  There are several  values in the data set which
appear unusual, either  high  or  low,  but none of these were eliminated
or replaced prior to conducting the ANOVA.   The rationale behind this
was that there was insufficient information  concerning the method
to warrant the use of an outlier  test,  which may eliminate values
likely to occur in actual use of  the  procedure.  It was noted where
a particular significant result occurred  due to one or more of these
values and a qualifying statement was added, where appropriate.
     For those factors  determined to  be significant, a means separation
procedure was used, Fisher's LSD  statistic,  to determine where the
differences lay.  The LSD is  calculated according to the formula:
          LSD.05 -
where     LSD Qg - test statistic at  5-percent significance level
          t.05 (v) - student's  value  with  v  degrees of freedom at
                        the  5-percent level

          n^, r\2 - number of  observations  used to obtain the mean values
and       MSE - mean square  for error from the ANOVA table.

                                     21

-------
F  tctor means are presented  in  the  statistical  summary tables in Appendix
B, along with the LSD values.  The  results  obtained for each of the
phthalate esters and the overall conclusions  are  summarized below.
     DMPThe dimethyl phthalate results  are  summarized in Table B.I
of Appendix B.  The only significant  term in  the  ANOVA was pH,  with
pH 2 giving 100 percent recovery with statistically equivalent  results
at pH 10, 112.1 percent recovery.   These  results  are considered somewhat
spurious, however, in light of complicating factors.  There was considerable
variability in the data, with  an overall  standard deviation of  13.1
percent recovery, due largely  to the  100  percent  DCM data.  The DCM-
pH 2 results ranged from 67 to 125  percent  recovery, and the average"
result of 96 percent recovery  seems fortuitous.   For the 15 percent
DCM in hexane data, the replicate  values  and  the  pH-to-pH values were
fairly consistent, with no  real tendency  observable.  In view of the
data, it is felt that little weight should  be given to this result.
     Neither the solvent system nor the  interaction term was significant,
and no preference for a solvent or  solvent-pH combination was indicated
for these data.  The conclusion then  for  the  DMP  is that no clear
tendency for better recovery was observed for any of the factors in
the experiment.
     PEPThe statistical analyses  for diethyl  phthalate are summarized
in Table B.2 of Appendix B.  For these data,  both the pH and the pH-
solvent interaction terms were significant.  The  pH means indicate
that pH's 7 and 10 were equivalent  with  sample means of 102.1 and
106.2 percent recovery, respectively, and the pH  2 results were significantly
                                     22

-------
lower, with an average of 93.7 percent  recovery.   The  pH  2  data  were
affected by one low value with 100 percent DCM  but would  have  indicated
that a different true mean could be  expectd  even  if this  value had
been excluded.  Similarly, the solvent-pH interaction  term  indicates
that the low average value for 100 percent DCM-pH 2 was the principal
difference among the means,  but this would remain true if the  average
of two results, excluding the one low one, was  used.  The interaction
term revealed that the low pH 2 results  are  a  function of the  low
100 percent DCM-pH 2 data and not a  true pH  effect.  The  indication
is then that if 100 percent  DCM is used  as the  extraction solvent,
the wastewater should first  be adjusted  to a neutral-to-basic  condition.
If 15 percent DCM in hexane  is used, no  pH adjustment  would be necessary.
     DBPThe summary of the statistical analyses on the  dibutyl  phthalate
data is given in Table B.3 of Appendix B.  Both solvent and pH terms
were significant, as well as the solvent-pH  interaction.  Higher recoveries
were obtained using the 15 percent DCM-hexane  system with 95.8 percent
recovery compared to 84.9 percent for the 100 percent  DCM.   The  mean
recoveries for the varying pH's indicate equivalent recoveries at
pH's 7 and 10, 98.0 and 94.5 percent, respectively,  with  lower overall
recovery at pH 2 of 76.3 percent.
     The interaction means indicate, however,  that in  both  of  these
cases low recovery with 100  percent  DCM-pH 2 was  the reason for  these
differences.  For the remaining solvent-pH combinations,  the average
recovery ranged from 91.3 to 101.7 percent,  all acceptable  results.
As in the previous data, the conclusion  is that if 100 percent DCM
                                     23

-------
is used, an adjustment  to  a  neutral-to-basic  condition would be required,
but no adjustment of pH need be made  if  15  percent DCM in  hexane is
used as the extraction  solvent.
     BBP--The summary of the ANOVA on the  benzyl-butyl phthalate data
is shown in Table B.4 of Appendix B.   The  significant terms were the
solvent, the pH, and the solvent-pH  interaction.   The preferred solvent
was 100 percent DCM with a recovery  of 97.3 percent compared to 93.7
percent for 15 percent DCM in  hexane. Better results were obtained
at pH 7 than at pH 2 or 10,  100.6 percent  recovery versus  89.2 and
93.7, respectively.  The separation  of the  interaction means,  however,
revealed that these differences were  influenced by a high  average
result for 100 percent DCM-pH  7 and  are  not truly  pH and solvent effects.
All of the solvent-pH combinations were  equivalent with the exception
of the 100 percent DCM-pH  7, which had an  average  recovery of  108.8
percent.  The remaining combinations  showed recoveries of  88.7 to
96.7 percent and were all  acceptable  recoveries.   The conclusion for
BBP then is that either solvent may  be used,  but  if 100 percent DCM
is used, the pH of the wastewater should be adjusted to a  nonneutral
condition.
     DEHPThe summary  of  the  statistical  analyses on the  diethylhexyl
phthalate data is contained  in Table  B.5 of Appendix B. For these
data there were no significant terms  in  the ANOVA, and the conclusion
is that neither solvent nor  pH affected  the recovery of this compound.
The overall mean recovery  was 103.1 percent.
     D|OP_The statistical  analyses on the  dioctyl  phthalate data are
summarized in Table B.6 of Appendix B.   These data gave one significant
                                   24

-------
term 1n the ANOVA, the  solvent-pH  interaction  term.   The best results
were obtained with 100  percent DCM-pH  7  and  15 percent DCM-hexane-
pH 10 with 96.5 and 97.3 percent  recovery,  respectively.  The other
four combinations were  equivalent  to each  other ranging from 91.8
to 93.3 percent recovery so  all combinations produced good recoveries.
The preference for those two sets  of conditions is slight and may
not be of practical significance.
Summary ~
     The results of the analyses  on  the  six  compounds are summarized
in Table 6.  From this  it  can be  seen  that no  clear tendency can be
detected for one solvent system to produce superior recoveries or
for the recoveries to be influenced  by the pH  of the water.  The principal
differences occurred when  100 percent  DCM  was  used and a particular
pH gave unacceptable results, especially in  the cases of DEP and DBP.
When 15 percent DCM in  hexane was  used as  the  extraction solvent,
no differences were detected among the pH's.  Another factor in the
evaluation is that the  15  pecent DCM-hexane  solvent system produced
more consistent results with fewer of  the  apparent contaminations
and none of the low recoveries.   The conclusion then is that 15 percent
DCM-hexane should be used  for extracting the wastewater and no adjustment
of the pH need be made.  The mean  recoveries obtained with 15 percent
DCM-hexane at the three pH's are  shown at  the  bottom of the summary
table to indicate the recovery obtained.
Preservation Study
     The preservation study  was conducted  to determine the effects
of a 7-day storage period  at various conditions on the recovery of
                                     25

-------
          TABLE  6.   SUMMARY OF  EXTRACTION STUDY ANALYSES
Compound
OMP
OEP
OBP
BBP
OEHP
OOP

Solvent
-.*

15% DCM-hexane
1002 OCM
~
_w
FACTOR
PH
U*
7,10
7,10
7

 <
-
Solvent x pH

not 1002 DCM-pH 2
not 1002 DCM-pH 2
not 1002 DCM-pH 7
...
1002 DCM-oH 7
                        Means (152 OCM in Hexane)
                                   COMPOUND
                                                          152 DCM-hexane-pH 10
EH
2
7
10
DMP
104.0
110.0
111.0
DEP
101.3
101.5
103.7
DBP
91.3
101.7
94.3
BBP
89.7
89.7
96.7
DEHP
107.31
101.3
109.?!
OOP
93.7
91.8
97.3

*  Line  indicates no  significant difference.
t  U indicates data do not lend themselves to ready interpretation.
!  Influenced by single high value.
                                       26

-------
the six phthalate  esters  of  interest from dosed water samples.   Each
sample consisted of  one quart  of  water dosed with six phthalates,
as in the extraction  study (see Table 3).   Two replicates for each
of twelve conditions  of pH,  temperature,  and residual chlorine  were
prepared for each  group as shown  in  the following model:

pH 2
pH 7
pH 10
4"
0 ppm Cl
2
2
2
C
2 ppm Cl
2
2
2

0 ppm Cl
2
2
2
rc
2 ppm Cl
2
2
2
The 2 ppm  residual  chlorine  level  was  obtained,  where required,  by
adding 160 microliters  of Mallinckrodt sodium hydrochlorite analytical
reagent  (5 percent  minimum available Cl).   This  quantity of hypochlorite
solution was  determined by preparing a standard  chlorine solution
which was  diluted to  five different concentrations  that were used
to prepare an absorbance versus  concentration curve.    Storage containers
were one-quart,  flint-glass,  round, narrow-mouth bottles with aluminum
foil-lined caps.  A sample was prepared by  filling  the bottle about
two-thirds full  with  pH-adjusted water, adding 160  microliters of
sodium hypochlorite (when required), swirling vigorously,  adding the
100 microliters  of  dosing solution into the vortex, and adding the
remainder  of  the pH-adjusted  water while that already in the bottle
was still  swirling.   The bottle  was capped  immediately and stored
in a closed cardboard box at  4C or 24C.   Care  was taken  not to slosh
the bottle contents onto the  aluminum  lining of  the cap after closure.
     After storage, samples were extracted  and analyzed as in the
extraction study.
                                     27

-------
     The third extraction  study had not been completed when the preservation
study was started.  This made  it necessary to select an extraction
solvent and extraction  pH  without the benefit of prior data.  The
solvent selected was 15 percent DCM-hexane,  Considerations in selecting
this solvent were  its proven extraction efficiency with other priority
pollutants (pesticides  and RGB's),  its relative general use, and the
ideal situation of conforming  as many priority pollutant methodologies
to a standard procedure as possible.  The preservation study samples
were extracted at  their given  pH without any prior adjustment.
     Data obtained in the  preservation study was given in Table 7.
The data presented are  uncorrected for values obtained for blanks
at each of the 12  test  conditions because of the variation of these
values.
Data Analysis
     The results of the preservation study were examined by submitting
the percent recoveries  to  an analysis of variance (ANOVA) procedure
to determine which of the  factors affected the recovery upon storage.
     The ANOVA for these data  is based on a 3x2x2 factorial experiment
with two replicates of  each combination.  The model  is presented by:
     Y-i-iH ~ ^ "*" P-i "*" "^-i "*" Y|  "*" PT-i-i + PY-Jlx "*" "nf-iU + PTY-HI/ "*"  -5H
      IJI\I         I     J     IV      IJ     IK     JN      IJK    IJKI
where
     Yijkl " Dserved recovery for pH i, temperature j, chlorine
             level  k and repetition 1
     y - overall mean
     P1 - effect of pH  i,  1=1,2,3
                                     28

-------

a.
O
o
*
OL
LU
a
a.
CO
CQ
a.-
ca o
a o
H- 1
"OS
a. LU
LU a.
o
^<
o o
i co
Q<
=> O
CO 1
0 *"*
t 1
> 5
C LU
LU >
CO O
LU O
Of LU
a. a:
Lu h-
o z
LU
CO O
I as
_1 LU
=D a.
co ~-*
LU
a:

*
r^.

LU
ca
H-

j



i


a.
eo
CO



~ a.
0



a.


9=
&


a
u
a

at

4)
c
i- S
0 9
* Q
e-
0

01
J^
3
2<_>
(DO
a.
S
l

=cl
i>- o Oktn > ** cMin fo CM o ok in r-i
O Ok Ok Ok Ok Ok Ok Ok Ok Ok Ok Ok 4O 41 Ok Ok


.
(3 Ot ^K Ok C3 ^4 CO C3 00 ^q CO ^^ Ok 41 CO C3


^S99**^ oo*-^inf2 co^^r^^* c*i oco
O OO Ok Ok CO Ok 00 O f^ \O ^* ^^ f^ 41 l^% IO
1-4 -



Ok Ok Ok OO v^Ok If) T t- CO CM OMA
 r~ coao aaok enoo coco otco Ok coo



^ m ^p ci ok o <*i oo i^k o o o oo o
41 ok oko> oot oo ok en oo 041 OkO
^H f( f-t I-H-I - 1-4


f) ^r r o ok vo m COCM r~*r CM oo
4IOOO i-^O*"*^^ OkOkOkOk -^4i OO



 CM CM CM CM CM CM CM CM







O CM O CM OCM O : CM






^   *
CM CM




CM r^
CO 00 CO Irt ^"4 ^^ CO ^^



SCM ^*in en-* m-
Ok 00 ** 00 Ok Ok Ok


i co COP* *r- coco




o^* in^h << co^*
ok Ok ok Ok r^ QO to P*



o i^* f1^ (*i m in o oo
ook cno coco cor^


*r o af- OCM 10 in
Ok Ok 00 00 CO f) CM CM



 CM CM CM CM







O CM; O CM






r *
CM




O











' "









^
C
0

trt
C
*
5
c
o
(J
1
a>
n
a
^~
*!o
1TJ

o

10
o
*
29

-------
     T- - effect of temperature  j, j*l,2
     Yk - effect of chlorine  level k,  k=l,2
     pTij p^fk Tyjk " tw-fac
-------
for pH, temperature, and/or  chlorine  taken  Individually.   The conclusion
is that the means of the combinations should be used to provide the
estimates of percent recovery,  and  no physical  interpretation is implied.
The precision of the analyses  allowed,  in  some  cases, relatively small
differences to be termed significant, while there may be no practical
value to the difference between average recoveries in these cases.
     DMPThe results  for  dimethyl  phthalate showed that all  main
effects and interactions had a significant  effect on the recoveries
obtained.  The most prominent  effects,  as  determined by the magnitude
of the mean square for each  factor, were pH, temperature, and the
pH-temperature interaction.  The LSD  for the pH factor showed that
all three sample means could be considered  to be estimating different
true (population) means.   At pH 7,  the average  recovery was 100.0
percent, while 108.0 percent was obtained  for pH 2 and 59.1 percent
for pH 10.  The best conditions for this phthalate then was storage
at neutral conditions  with basic conditions clearly unacceptable.
The better temperature for storage  was 4C  with an average recovery
of 95.8 percent versus 82.2  percent at room temperature.
     The pH-temperature combination means  gave  more information.
At pH's 2 and 7, averages  above 100 percent were obtained at room
temperature storage (112.5,  106), while the pH  2-4C mean was 103.5
and the pH 7-4C mean  was  94.0.  The  mean  recovery for pH 10-4C was
90.0 percent, an acceptable  result, but the recovery for pH 10-24C
was only 28.2 percent.  The  average recovery shown above for pH 10
thus was greatly affected  by the results from room temperature storage.
                                    31

-------
     Clearly then the data Indicate that cold  storage  is  recommended
for preserving samples for DMP, and some preference  is noted  for neutral
conditions.  None of the pH's is unacceptable  under  storage at 4C,
but the results at pH 2 were closest to 100 percent, although slightly
above.  There was no effect on the results at  pH  2 or  7 resulting
from the presence or absence of residual chlorine in the  water.
     PEPThere were three significant terms in the ANOVA of  the diethyl
phthalate data, pH, temperature, and the pH-temperature interaction.
The pH results indicated that better recovery  was obtained at pH's
of 2 and 7  (98.5 and 99.1 percent recovery, respectively)  than  at
pH 10 (90.6 percent recovery).  Storage at 4C produced a significantly
higher average result, but the difference was  slight,  with 97.9  percent
observed for 4C and 94.2 percent recovery for 24C.
     The pH-temperature interaction means revealed a similar  pattern
to the DMP results.  The principal difference  among  the means was
the low average recovery (82.0 percent) for the samples stored  at
pH 10 and at room temperature.  Neither the chlorine term nor any
of the interactions involving chlorine were significant,  and  the conclusion
is that residual chlorine in the water will not affect the results
obtained for DEP upon storage of the samples.
     DBFThe dibutyl phthalate results were analyzed  according  to
the ANOVA model above and gave two significant terms,  pH  and  the pH-
temperature interaction.  The best results were obtained  under  storage
at pH 7 with 90.8 percent recovery on average. The  mean  for  pH  2
(86.5 percent recovery) was indistinguishable  from the pH 7 mean using
                                   32

-------
the LSD statistic, but storage at  pH  10 produced  a significantly lower
result (83.4 percent  recovery).  As  In the  previous data,  however,
the Interaction term  revealed that the principal  reason for this difference
was the low result for pH  10-24C.   No effect  was noted due to the
presence or absence of residual  chlorine  in the water,  and this does
not appear to be a concern  in analyzing for this  compound.
     BBPThe recoveries for benzyl-butyl phthalate gave three significant
terms in the ANOVA, pH, temperature,  and  the pH-temperature interaction.
The three pH's produced distinct means for  this compound,  with storage
at pH 2 giving the best results, 93.8 percent  recovery, followed by
pH 7 and pH 10 (72.8  and 60.3 percent recovery, respectively).  Higher
results were obtained on average when storage  was at 4C as opposed
to room temperature,  with  mean recoveries of 84.3 percent and 67.0
percent, respectively.  From the pH-temperature interaction, however,
no significant difference  was observed between the 4C  and 24C results
at pH 2 (96.0 and 91.8 percent recovery,  respectively).  As a result,
for this compound an  adjustment  to pH 2 prior  to  storage would remove
the necessity for cold storage.  As  in the  previous analyses, the
significance of the pH-temperature interaction is largely due to the
loss of BBP at pH 10-24C,  and no  chlorine  effect, either alone or
in combination, was observed in  the  results.
     DEHPThe diethylhexyl phthalate data  gave one significant factor
in the ANOVA, the pH  of the water.   The best results were obtained
when the samples were stored at  pH 2, with  an  average of 100.8 percent
recovery.  The average for  pH 7, 91.8 percent  recovery, was indistinguishable
                                   33

-------
from the pH 2 mean using the LSD statistic,  but  only  marginally.
No other factors or  Interactions were  significant,  and  the  results
were not affected by either the presence  or  absence of  chlorine.
     POP-The dioctyl phthalate results gave one significant main
effect, pH, and three significant  interaction terms,  pH-temperature,
pH-chlorine, and temperature-chlorine.  The  pH results  were consistent
with the other compounds,  showing  best recovery  at  pH 2 (95.2 percent)
with pH's 7 and 10 lower and equivalent to each  other (86.2 and 84.2
percent, respectively).  The pH-temperature  interaction term revealed
that equivalent results were obtained  at  pH  2 for either 4 or 24C,
and at pH 7-4C.  The other combinations  were equivalent to each other
but lower.
     The two interaction terms involving  chlorine provided  no information
that would indicate  a problem for  these analyses.   For  the  pH 2 data,
the same average result (95.2 percent  recovery)  was obtained for both
0 and 2 ppm residual chlorine in the water,  which was the highest
average obtained.  For the temperature-chlorine  interaction term,
the data indicated that storage at 24C-0 ppm chlorine  gave a lower
average result than  the other combinations.   However, previous results
indicate a preference for cold storage so the presence  or absence
of chlorine, assuming storage at 4C,  is  not a problem  for  this compound.
Summary
     The results of  the analyses on the six  compounds in this category
are summarized in Table 8.  On the basis  of  the  trends  shown, the
recommended conditions for storage would  specify an adjustment of
                                    34

-------






Q
=5
CO

O
H-l

UJ
CO
UJ
a:
a.

0

i_

=3
CO
UJ
o;
^ t
o
^^
a:

2?
ZD
CO




CO

UJ
CO
^f
LM.
r^






























ce
o

u

*
u.






































u
X
Q
5
X
X
a


u
X
I


^m
u
X
Q



a
1
X
z




&





o
E
01





^2
a







a
s
3
o
a
e
8



311 III



O O
* *" 0 *
4J CM S r
ia i i it -* en
cr o
a* e
CM

a. a. CM
x 

3 1 1 1 1 r-
* Z

? ~ -. st?
CM   00
^^    CM CM^ O^
CM^ CM^  " <> CM -~ ee CM a. o
O O *O 1 * C ZC 03 <0
*. ** ^ r  -x
O O 
O





0.
Q. Q. Q. CL 3C O.
uj ca eo LU o
o o eo a a
















.
e
o
4J 
1 1
1 1
S u
c
>> S 3
ff e-o
a ai-^
01 u 
 01 01
s 
a o
VI
01 ** O
> c u
i ia c
oi u at
< ^3
j= "e--2.
+j o> a>
r- O
-a n e
e v
01 O w>
i CO)

*j ui a.
o a*
e  o
o u
0 i- 0
a e
> e i-

3 01 oi
Wl CO.
01 -r- 41
I. ---O

1 1 1

?-!-+-
35

-------
the water to acidic conditions,  and  storage at a temperature of 4C.
Under these conditions,  no  Interference  can be expected from residual
chlorine up to the 2 ppm level.
     Adjustment of wastewater  samples  to pH 2 may not be practical
under field conditions and  may  be  avoided with minimal  effect on the
determination of phthalate  esters, provided the samples are stored
at 4C.
                                    36

-------
                            SECTION  5
                             PHASE II
OBJECTIVES
     The objectives of Phase II were  as  follows:
1.   Investigation of two  liquid-solid chromatographic  cleanup procedures,
     including the determination  of the  separation  patterns  and recoveries
     for all of the substances of interest.
2.   Application of the method to five wastewater samples  from industrial
     discharges.  This aspect of  the  work  included  modifications of
     the method as required to analyze the wastewaters.  These modifications
     would be based on experience gained earlier  in the  program and
     on new developments reported in  scientific  literature.
3.   Generation of accuracy and precision  data  for  all  substances
     of interest by dosing them into  the five wastewaters.   Determinations
     would be made immediately after  dosing  and  after one  week of
     storage at the best conditions observed in Phase I.
PROGRAM DIRECTION
     Representatives of SwRI met  with the  project officer  in Cincinnati
before Phase II work commenced.   Discussion  with  respect to  Phase
II was primarily on the cleanup columns  and  procedures  to  be used.
Florisil and alumina were  selected as the  adsorptive solids.  Fractionation
of compounds was limited to that  which can be obtained  on  the above-
                                    37

-------
named solids.  Fractionation  on  secondary  columns  such  as silica gel
or charcoal was considered  to  be  beyond  the  scope  of this program.
     DCM was chosen as  the  extracting  solvent to  be used with the
wastewater samples.
LIQUID-SOLID COLUMN CHROMATOGRAPHY
     The first column chromatography method  to be  investigated for
cleanup of phthalate esters was  based  on the work  reported by C. S.
Giam et al.^   It was found by Giam et al. that incomplete recoveries
of phthalate esters were  obtained from Florisil  unless  the Florisil
was deactivated with 3  percent water.  Two procedures were developed:
one employing 35 grams  of Florisil for samples with a high lipid content
and the second procedure  using only 10 grams of Florisil  for samples
with a low lipid content.
     Initial studies using  the 10 grams  of deactivated  Florisil  (3
percent fr^O) and the reported  solvent  system indicated  that modification
of the procedure was required.   The dimethyl and  diethyl  phthalates
are more polar than the other  phthalates used by  Giam and were not
eluted from the column  when following  the  reported procedure.
                                                      /
     The modified procedure,  used in this  study,  is as  follows:
Prepare Florisil  (PR)  by  heating  to  400C  overnight.   Then add 3 percent
HpO (w/w) and mix  thoroughly.   Let  stand for at least 2 hours.  Use
10-mm ID column  chromatography  tube  with glass  frit and Teflon stopcock
(Kontes Catalog  K420540-0213).
1.   Add 10 g of Florisil  (3 percent ^0)  to the 1-cm ID chromatography
     tube and tap  to  settle.
2.   Add  =1 cm     $0   to  top of column.
                                    38

-------
 3.   Add hexane to tube and pre-elute 40 ml.  Discard.

 4.   Drain until solvent level is even with top of ^SC^.

 5.   Add sample in =2 ml hexane to top of column.

 6.   Drain sample onto column (to top of Na2S04).

 7.   Wash sides of tube with =2-mL hexane and drain  into  top of ^SO^j.

 8.   Add hexane to top of column and elute 40 ml of  hexane.  Discard
      this fraction.

 9.   Drain hexane to top of Na2S04 and add -2 ml 20  percent ether/hexane.

10.   Drain new solvent to top of Na2S04.  Add more 20 percent ether/hexane.

11.   Place a 500-mL Kuderna-Danish evaporator below  the column.

12.   Collect 100 ml of the 20 percent ether/hexane eluate in the K-D
      evaporator.

13.   Evaporate the above eluate to <10 ml on a steam bath.

14.   Transfer to a solvent-rinsed 4-dram vial with Teflon cap liner.

15.   Adjust final volume of concentrate to 10 ml.  Evaporate under
 	nitrogen if necessary.	

 The second column procedure developed for this program, using alumina

 as the adsorbent, is as follows:

 Prepare Alumina (Woelm N) by heating to 400C overnight.   Then add
 3 percent H20 (w/w) and mix thoroughly.  Let stand overnight in sealed
 jar.  Use 10-mm ID column chromatography tube with glass  frit and
 Teflon stopcock (Kontes Catalog K420540-0213).

 1.   Add 10 g of Alumina (3 percent H20) to the 1-cm ID chromatography
      tube and tap lightly to settle.

 2.   Add -1 cm Na2S04 to top of column.

 3.   Add hexane to tube and pre-elute 40 ml.  Discard.
                                     39

-------
 4.   Drain until solvent level is even with top of ^SO^
 5.   Add sample in -2 ml hexane to top of column.
 6.   Drain sample onto column (to top of ^SO^).
 7.   Wash sides of tube with = 2 ml hexane and  drain  into  top  of  ^SO^.
 8.   Add hexane to top of column and elute 35  ml of  hexane.   Discard
      this fraction.
 9.   Place a 500 ml Kuderna-Danish evaporator  below  the column.
10.   Drain hexane to top of ^$04 and add =2-ml 20  percent ether/hexane.
11.   Drain new solvent to top of ^SO^.  Add  more 20 percent ether/hexane.
12.   Collect 140 ml of the 20 percent ether/hexane eluate in  the K-D
      evaporator.
13.   Evaporate the above eluate to -10 ml on a steam bath.
14.   Transfer to a solvent-rinsed 4-dram vial  with Teflon cap liner.
      Rinse twice with hexane and add rinses to vial.
15.   Adjust final volume of concentrate to 10  ml.  Evaporate  under
 	nitrogen if necessary.	
      A gel permeation column using Sephadex was considered for use
 in this work, but when a computer search of the literature failed
 to give any references relative to the use of  this substance  other
 than for removal of lipids, no further use was contemplated.   Lipids
 are not a problem contaminant with most nondomestic  wastewaters.
      The cleanup procedures using Florisil and alumina were evaluated
 as to the recoveries that could be obtained when doses of the six
 phthalates were applied to columns of these adsorbents.   Table 9 shows
 the recoveries obtained from five replicate samples  when  using the
 Florisil cleanup column.  Table 10 shows the recoveries obtained from
                                    40

-------
     TABLE 9.  RECOVERY* (.%) FROM FLORISIL PR COLUMN

Sampl e
1
2
3
4
5
Avg
DMP
114
115
114
115
117
115
DEP
114
115
114
117
118
116
DBP
99
97
97
98
99
98
BBP
93
92
92
93
94
93
DEHP
98
96
97
96
97
97
DOP
97
94
96
94
95
95


TABLE 10.
RECOVERY*
(*)
FROM ALUMINA(N)
COLUMN

Sample
1
2
3
4
5
Avg
DMP
114
112
114
115
115
114
DFP
108
107
109
111
111
109
DBP
97
97
98
99
97
98
BBP
100
102
99
99
98
100
DEHP
87
85
97
99
84
91
DOP
95
95
98
99
93
96

* Amount added to column in 2 mL hexane.
  DMP-15yg, DEP-15yg, DBP-8yg, BBP-lyg, DEHP-8yg, DOP-20yg
                             41

-------
five replicate s mples using the Alumina N cleanup  column.   The recoveries
for all six test phthalates from both materials  appear  to  be very
good, averaging 90 percent or  better.  The high  recoveries  for DMP
and DEP were unexplained  since blank tests showed only  small  (2-3
percent) interferences or contamination  for  these two compounds.
A lack of a linear detector response could have  been  responsible.
WASTEWATER APPLICATION
     With the assistance  and approval of the project  officer,  five
wastewaters were procured and  analyzed.  All  samples  were  put  in  clean,
one-gallon bottles and shipped unrefrigerated via air to Southwest
Research Institute laboratories.  Upon receipt at SwRI, all  the wastewater
from a particular source  (except Wastewater  2) were pooled,  adjusted
to pH 5-7 if necessary, returned to  the  bottles  in  which they  were
shipped, and stored at 4C (in the dark) until used.  Due  to the  high
acid content of Wastewater 2 (approximately  48 g/L  NaOH required  to
neutralize), it was stored at 4C as received.
     In order to develop  method improvements and to provide  base  data
for dosing and recovery experiments  and  for  the  accuracy and precision
evaluations to follow, each wastewater was analyzed in  triplicate
for each substance of interest in this program.
     One liter of wastewaters was extracted  three times using  60-mL
DCM for each extraction.  The  combined extract was  dried with  ^$04
and placed in a Kuderna-Danish evaporator.   The  DCM extract  was concentrated
to 5-10 ml, 90 ml of hexane was added, and the extract  was  concentrated
to slightly less than 10  ml in the K-D.  The sample was then transferred
                                    42

-------
to small vial and concentrated to 2 ml.  This 2-ml concentrated  extract
was then subjected to the Florisil cleanup  procedure  as  previously
described.  One procedural change, used  only with Wastewaters 3, 4,
and 5, was attempted and with some success.  In  the Florisil cleanup
procedure, instead of collecting a single fraction of 100 ml of  eluting
solvent containing all the phthalate esters of interest, a  two-fraction
collection was made.  Fraction 1 (first  60 ml) contained nearly  all
of the OOP and DEHP, most of the DBP and BBP and, in  some instances,
much of the early GC-eluting material.   Fraction 2  (next 40 ml)  contained
all of the DMP, most of the DEP, some DEP and BBP and,  in some  instances,
reduced amounts of early GC-eluting material(s)  which otherwise  interfere
with detection of the DMP and DEP.  This method  of collecting two
fractions instead of one may be of some  help in  the analysis for DMP
and DEP.  For the analysis of the other  phthalate esters, it is  best
to combine fractions (or collect only the single fraction)  before
GC analysis and thus avoid the complications of  partial  separation
of some of the phthalate esters.
     Wastewaters 1 and 2 were analyzed by collection  and analysis
of only one fraction.  Wastewaters 3,4, and 5 were examined by  collection
of the two fractions.  Chromatograms of  Fraction 2, analyzed for DMP
and DEP only and of combined Fractions 1 and 2,  analyzed for the remaining
phthalate esters, are shown 1n Figures D.I and D.2 in Appendix D for
dosed Wastewater 3.
     Chromatograms of a standard mixture and Wastewater 1 through
5  (single fraction collection) are shown in Appendix  E.
                                    43

-------
     The gas chromatography  for  the wastewater  analysis  was  performed
using Column 1.  Samples were  analyzed  by  holding  the  column at 180C
for 16 minutes for  determination of DMP, DEP, and  DBP.   The  temperature
was then raised rapidly to 220C to decrease  elution time  of the BBP,
DEHP, and OOP.  All analytical data were obtained  using  peak height
measurements.  All  extracts  were diluted so compounds  of interest
would not go off scale at 8X,  thus assuring linearity  of the EC detector
for DBP, BBP, DEHP, and DOP.   No linear range for  DMP  and  DEP was
found with the EC detector,  and  analytical data for these  two compounds
were obtained from  calibration curves.
     Tables 11 through 15 show the results of the  analysis of raw
Wastewaters 1 through 5.
     Tables 16 through 19 show the results of the  analysis of Wastewaters
1, 3, 4, and 5 dosed with the  six phthalate esters under study.
     The data presented in the tables referred  to  below  were used
to calculate a "Method Recovery", i.e., the percent recovered of the
amount added at zero time, and a "Preservation  Recovery",  i.e., the
percent remaining after seven  days of the  total  amount in  the dosed
wastewater at zero  time.  This was performed  as follows:
     let k = background (amount  found in the  undosed wastewater)
     let y = quantity added  to wastewater
     let XQ = quantity found in  dosed wastewater at zero time for
                replicates 1,  2, or 3.

     let XQ av_ = average quantity found in dosed  wastewater at zero
                  time
                                   44

-------
     TABLE 11.   ANALYSIS OF WASTEWATER 1 FOR PHTHALATES,  ug/L

Replicate DMP DEP DBP
1 I* I 157
2 I I 151
3 I I 157
"k" ug/L
Average I I 155
BBP DEHP OOP
8.0 11.6 0
12.8 13.9 0
6.2 13.9 0
9.0 13.1 0

*I - Interfering Peak(s)
TABLE 12. ANALYSIS OF WASTEWATER

2 FOR PHTHALATES, ug/L

Replicate DMP DEP DBP
1 I* 1430 I
2 I 1500 I
3 I 1580 I
"k" ug/L
Average I 1500 I
BBP DEHP OOP
232 I 3260
304 I 3820
304 I 4020
280 I 3700

*I - Interfering Peak(s)
                                    45

-------
TABLE 13.  ANALYSIS OF WASTEWATER 3 FOR PHTHALATES,  ug/L

Replicate
1
2
3
11 k" ug/L
Average
DMP DEP DBP
0 0 2.4
0 0 1.8
0 '0 2.9

0 0 2.3
BBP
0.8
0.6
1.0

0.8
DEHP
6.6
6.5
6.6

6.6
DOP
0
0
0

0

TABLE 14.
ANALYSIS OF WASTEWATER 4
FOR
PHTHALATES, u<
3/L

Replicate
1
2
3
"k" ug/L
Average
DMP DEP DBP
000
000
000
000
BBP
0.7
0.4
0.9
0.7
DEHP
1.0
1.0
1.0 
1.0
DOP
0
0
0
0
                               46

-------
     TABLE 15.   ANALYSIS OF WASTEWATER 5 FOR PHTHALATES,  ug/L
Replicate      DMP      DEP      DBP      BBP      DEHP      OOP
1
2
3
"k" wg/L
Average
0
0
0
0
5.8
5.8
3.8
5.1
1.9
3.7
3.7
3.1
3.2
2.5
2.1
2.6
255
269
285
270
0
0
0
0
                                    47

-------
       TABLE 16.   ANALYSIS OF WASTEWATER 1 DOSED WITH PHTHALATES
Replicate
Zero time
                 1
  Method recovery.^
                 2
  Method recovery,%
                 3
  Method recovery,%
       XQ Avg.
       k
  Method recovery,% Avg.
7 days at 4C
                 1
  Preservation recovery,%
                 2
  Preservation recovery,%
                 3
  Preservation recovery,%
  Preservation Recovery,
  % Average                I*
                                           Dose,  ug/L
0 0 1500
MP DEP DBP*



* I
1380
82
1360
80
1360
80
1370
* 155 
81



1390
101
1370
100
1610
118
100
BBP*
86
77
86
77
86
77
86
9.0
77
85
99
80
93
77
90
150
DEHP*
122
69
116
69
117
73
118
13.1
70
111
94
109
92
116
98
150
OOP*
119
79
116
77
116
77
117
0
78
112
96
121
103
116
99
I*
106
94
95
99
*  Huge interferences made spiking unfeasible at ug/L levels.
t  Extract diluted 500x to see DBP in linear range of EC detector.
4=  Extract diluted 50x to see BBP, DEHP, & DOP in linear range of EC detector.
                                    48

-------
       TABLE  17.  ANALYSIS OF WASTEWATER 3 DOSED  WITH  PHTHALATES

Replicate

45
DMP*

45
DEP*
Dose,
24
DBPt
uq/L
3
BBPt

24
DEHPt

60
DOPt
Zero Time
                1          45     45    25.6      3.8    24.1    46.4
  Method Recovery,%       100    100    97       100     73      77
                2          45     45    23.2      3.6    20.2    34.4
  Method Recovery,%       100    100    87       93      57      57
                3          45     45    23.8      4.4    22.2    37.6
  Method Recovery,%       100    100    90       120     65      63
       Xn Avg.             45     45    24.2      4.0    22.2    39.5
       ku                   00     2.3      0.8     6.6     0
  Method Recovery,%Avg.   100    100    91       124     65      66

7 Days at 4C
                1         36.0   39.3   18.7     3.1     23.4    36.5
  Preservation Recovery, 80     87     77       79      105     92
                2         29.7   38.1   16.4     3.4     24.5    34.4
  Preservation Recovery,% 66     85     68       85      110     87
                3         32.1   39.3   14.5     3.0     24.6    32.2
  Preservation Recovery,% 71     87     60       76      111     82
  Preservation Recovery
    % Avg.                72     86     68       80      109     87
*DMP & DEP determined from Fraction 2 of Florisil eluate diluted 15x to
 come within EC detector calibration range.
fl)BP, BBP, DEHP, & OOP determined from total Florisil eluate (Fracs. 1&2
  recombined) diluted 3x to come within linear range of EC detector.
                                   49

-------
        TABLE 18.  ANALYSIS OF WASTEWATER 4 DOSED WITH PHTHALATES

Dose, pg/L

Replicate
Zero Time
1
Method Recovery, %
2
Method Recovery ,%
3
Method Recovery, %
X0 Avg.
k
Method Recovery, % Avg.
7 Days at 4C
1
Preservation Recovery, %
2
Preservation Recovery, %
3
Preservation Recovery, %
15
DMP*

12.9
86
13.2
88
13.3
89
13.1
0
88

13.7
105
14.3
109
14.1
108
15
PEP

12.2
81
12.3
82
12.6
84
12.4
0
82

12.9
104
13.2
106
13.1
106
20
DBP

17.6
.88
18.2
91
18.9
95
18.2
0
91

18.2
100
18.6
102
18.1
99
5
BBP

4.2
70
4.6
78
4.7
80
4.5
0.7
76

4.3
96
4.3
96
4.3
96
40
DEHP

19.4
47
19.4
47
23.0
56
20.6
1.0
50

27.9
134
29.5
143
29.3
142

40
POP

18.3
46
19.4
49
21.7
54
19.8
0
50

26.5
134
28.7
145
27.6
139
  Preservation Recovery
       % Avg.               107    106    100    96     140     139
* All  components determined from total  extract diluted 1:5 to come in
  linear or calibration range of EC detector.
                                   50

-------
     TABLE 19.  ANALYSIS OF WASTEWATER 5 DOSED WITH PHTHALATES

Dose, yg/L

Replicate
Zero Time
1
Method Recovery, %
2
Method Recovery, %
3
Method Recovery, %
X0 Avg.
u
k
Method Recovery t% Avg.
7 Days at 4C
1
Preservation Recovery, %
2
Preservation Recovery, %
3
Preservation Recovery, %
Preservation Recovery
% Avg.
50
DMP*

50
100
49
98
47
94
49
0
97

50
102
49
100
48
98

100
50
DEP*

52
94
51
92
50
90
51
5.1
92

49
96
49
96
48
94

95
30
DBF*

18
49
15
40
13
33
15.3
3.1
41

9
59
9
59
10
65

61
30
BBP*

23
68
22
65
19
55
21.3
2.6
63

13
61
15
70
14
65

65
1000
DEHP*

980
71
980
71
980
71
980
270
71

740
76
770
79
780
80

78
150
POP*

140
93
140
93
140
93
140
0
93

110
79
110
79
110
79

79
DMP & DEP determined from Fraction 2 of Florisil eluate diluted 50x
to come within EC detector calibration range.

DBP, BBP, DEHP, & OOP determined from total  Florisil eluate
(Fracs. 1&2 recombined) diluted 25x to come within linear range of
EC detector.
                                51

-------
     let Xy avg * average  quantity  found in dosed wastewater after
                    seven  days  for  replicates  1,  2,  and 3.

Then
    "Method Recovery" =  (xQ-k)100/y %  (calculated for each  replicate)
and
    "Preservation Recovery"  =  (x7)(100)/x0 avg %  (calculated for
               each replicate)
No dosed recovery or preservation data were obtained on Wastewater
2 due to obvious deterioration  of the  sample.
ACCURACY AND PRECISION
Approach
     The accuracy and precision assessment for the method was of a
limited nature due to the  number and  types of  analytical results obtained,
According to the design  of the  program three replicates of  each of
five wastewaters were to be  dosed for  the compounds of interest and
analyzed, both at a zero time  and after seven  days storage  under the
appropriate conditions.  This  provided 15 determinations of recovery
of the dose for statistical  analyses.   However, the five wastewaters
could not be considered  equivalent, either in  their nature  or in the
behavior of the method when  applied,  so combining the results to form
an overall evaluation of the method was inappropriate.  Instead, the
results of each wastewater were treated individually, and no general
statement about the method was  attempted.
     The accuracy of a method  refers  to its ability to obtain on average
the true value, within the limits of  the variability inherent in the
                                 "  52

-------
method.  For this study,  the  accuracy  standard  used was the percent
recovery of the amount of the particular  phthalate  ester dosed Into
the wastewater, with 100  percent  representing an  accurate method.
Precision relates to the  variability among  measurements arising from
various sources and is considered separately from the accuracy evaluation.
It 1s possible for a method to  be precise,  meaning  having little variability
among measurements, without being considered accurate.   For this study,
the precision measurement used  was  the  range or the difference between
the highest and lowest recoveries.  While the range is  sensitive to
single extreme values, it was deemed more appropriate for these data
than other measures, given the  number  of  observations available.
     The recoveries obtained  after  seven  days storage were examined
separately to evaluate the effects  of  storage on  the observed levels
of these compounds.  These recoveries  are based on  the  total  amount
present at time zero rather than  the dosing level,  since it can be
assumed that any degradation would  affect both  the  initial  amount
and the dosed amount in equal proportions.
Results
Wastewater 1
     Neither the DMP nor  the DEP  was dosed  into Wastewater 1, either
at time zero or after storage due to the  interferences.  Consistent
results were obtained for the other four  compounds  studied, however,
with recoveries from 70 percent for DEHP  to 81  percent  for DBP.  In
each of these cases, the  precision  of  the analyses  was  good,  with
ranges of 2, 0, 4, and 2  percent  recovery for DBP,  BBP, DEHP, and
DOP, respectively.
                                    53

-------
     The preservation  recoveries were  generally  good for the four
higher-boiling phthalates,  but more  variable  than  the zero-day analyses.
The recoveries were from a  low of 94 percent  to  a  high of 106, with
ranges of 6, 7, 9, and 18.  The 106  average for  DBP  was influenced
by a single value of 118, and the indication  is  that recovery for
these compounds was not affected by  the  storage  conditions in this
wastewater.
Wastewater 3
     The analyses on this wastewater produced results which ranged
from low to very good, depending upon  the  particular compound.  Results
for both DMP and DEP were very good, with  triplicate analyses indicating
100 percent recovery of the spike.   The  recovery of  DBP was 91 percent,
on average, but with a range of 10 percent recovery.   The results
for the remaining phthalates were not  good in this wastewater, however.
The average recovery of BBP was 104  percent,  but the individual  recoveries
had a range of 27 percent (93-120).  DEHP  and OOP  were recovered at
65 and 66 percent, respectively, of  the  dose  level with ranges of
16 and 20 percent recovery.
    The preservation data were also  inconsistent from one compound
to the next.  For the  six compounds  studied,  the average recovery
relative to the zero-day analyses went from 68 percent for DBP to
109 percent for DEHP.  These analyses  were more  variable than the
zero-day analyses in the cases of DMP  and  DBP, equivalent for DEP,
and considerably less  variable for the remainder.  The ranges of the
triplicate analyses at zero-time were  27,  16,  and  20  for DBP, DEHP,
                                     54

-------
and OOP, respectively, at the initial  nalyses compared  to 9, 6,  and
10 for these compounds after seven-days  storage.
Wastewater 4
     The analyses on Wastewater 4 for DMP and DEP produced good  recovery,
88 and 82 percent, respectively, with ranges of 3 and 2  percent.
Recovery of 91 percent on average was noted for DBP but  more variability
with a range of 10 percent recovery.  Lower recoveries were obtained
for the remaining three phthalates, with average percent recoveries
of 76, 50, and 50 for BBP, DEHP, and OOP, respectively,  with ranges
of 10, 9, and 8.
     The preservation data were good for the first four  compounds,
going from 96 to 107 percent with ranges from 0 to 4 percent recovery.
For the DEHP and OOP analyses, however,  the average recoveries were
140 and 139 percent with ranges of 9 and 11, respectively.  These
recoveries are comparable to 70 and 69 percent, respectively, of  the
original dosed amount and likely indicate a problem with the initial
analyses.
Wastewater 5
     The results for both DMP and DEP were good in this  wastewater,
with average recoveries of 97 and 92 percent, respectively, and  ranges
of 6 and 4 percent recovery.  The DBP and BBP results were low and
variable, with average recoveries of 41 and 63 and ranges of 16 and
13 percent recovery, respectively.  The DEHP and OOP values were  consistent
but only 71 percent of the dose was recovered on the DEHP, while  93
percent was recovered on the DOP.  For both of these compounds there
was a zero range, with all three analyses showing the same recovery.
                                    55

-------
     The preservation  recoveries were  fairly  consistent for all  of
the compounds, with  ranges  of  recovery of 0 to 9  percent.   However,
the level of recovery  could be broken  down into three groups:   DMP-
DEP, DBP-BBP, DEHP-OOP.  The recovery  after storage was 95 to  100
for the first pair,  61 to 65 for the  second and 78 to 79 for the third.
The 61 percent recovery  represents  only 25 percent of the  initial
dosed amount remaining after seven  days and indicates that storage
in this wastewater would not be recommended for these analyses.
Summary
     The accuracy and  precision evaluations on phthalate esters  in
wastewater lead  to the following conclusions.   There are no assurances
that acceptable  data will be obtained  when analyzing for phthalate
esters using this method, as in the case of Wastewater 2 where no
usable results were  obtainable,  and in Wastewater 1, where none  were
obtainable for DMP or DEP.   The recoveries that can be expected  for
the compounds studied  ranged from 40 to 100 percent, depending upon
which compound and which wastewater.   In general, the precision  of
the analyses was acceptable to good, with ranges  of less than  10 percent
recovery common.
     The overriding  conclusion is that the recovery and the ability
to store the water for later analysis  are a function of the wastewater.
Storage frequently resulted in significant losses of the study materials
and in less precise  determinations  and cannot be  recommended as  a
general rule.
                                     56

-------
                              REFERENCES
1.   Thompson, J. F., Reid, S. F.,  and Kantor, E.  J.y  Arch.  Environ.
     Contam. Toxlcol., 6:143-157,  1977.

2.   Standard Methods, 409H, pp.  342-343,  Free Available Chlorine
     Test, Syringaldazine, and 409A,  lodometric Method I.

3.   Giam, C. S., Chan, H. S., and Neff, G. S., "Sensitive Method
     for Determination of Phthalate Ester Plasticizers in  Open-Ocean
     Biota Samples," Analytical Chemistry. 47:2225,  November 1975.
                                    57

-------
                              APPENDIX A

                KEY REFERENCES FROM LITERATURE SEARCH
C. S, Giam, H. S. Chan, and G. S. Neff, "Sensitive Method for Determination
     of Phthalate Ester Plasticizers in Open-Ocean Biota Samples,"
     Analytical Chemistry, 47:2225, November 1975.

R. A. Kites and K. Biemann, "Water Pollution.  Organic Compounds in
     the Charles River, Boston," Science, Pub. 72, Series 178, Issue
     4057, pp. 158-160.

R.D.J. Webster and G. Nicklers, "Problems in the Environmental Analysis
     of Phthalate Esters," Proc. Anal. Div.  Chem. Soc., Pub.  76, Series
     13, Issue 11, pp. 333-335.

V. W. Saeger and E. S. Tucker, "Biodegradation of Phthalic Acid Esters
     in River Water and Activated Sludge," Applied and Environmental
     Microbiology, 31(1), pp. 29-34, January 1976.

C. S. Giam and H. S. Chan, "Control of Blanks in the Analysis of Phthalates
     in Air and Ocean Biota Samples," Nat'l. Bur. Stand. (U.S.), Spec.
     Pub 442, pp. 701-708, 1976.
                                    58

-------
       APPENDIX B

  STATISTICAL ANALYSES
PHASE I EXTRACTION STUDY
          59

-------
  TABLE  B-l.   STATISTICAL SUMMARY  FOR DMP  - EXTRACTION STUDY
Anova Table
Source
df
SS
MS
Solvents
pH's
Interaction
Error
Total
Mean
1
2
2
14
19
-112.5
320.00
1679.29
960.29
2391.42
5351.00

320.00
839.65
480.15
170.82


2.87
4.92*
2.81



Cell Means

,


1
116.5
Solvent



2
108.5



                                  pH  (LSD- 16-2)
             7
           1UU.U "
             7
           122.1
                       10
                      112.1
                             Interaction
(1.2)          (1.7)        (1.10)        (2.2)          (2.7)       (2.10)
 96.0         134.2         113.3        "I5OT        lld.O        111.0
  Significant at 0.05 level.
                                    60

-------
TABLE B-2.   STATISTICAL SUMMARY FOR DEP  - EXTRACTION STUDY
Anova Table
Source
Solvents
pH's
Interaction
Error
Total
Mean
df SS MS F
1 33.80 33.80 <1
2 492.16 246.08 6.38*
2 359.49 179.75 4.66*
14 539.75 38.55
19 1425.20
- 100.8
Cell Means


Sol vent
1 2
"99TT 102.1
                                 pH   (LSD  -4.7)
            93.7
                            Interaction  (LSD- 6.7)
(1.2)
86.0
(1.7) ,
102.8
(1.10)
108.7
(2,2)
101.3
(2,7)
101.5
(2,10)
103.7

 * Significant at 0.05 level.
                                 61

-------
  TABLE B-3.   STATISTICAL SUMMARY FOR DBP  - EXTRACTION STUDY
Anova Table
Source
Solvents
pH's
Interaction
Error
Total
Mean-
df
1
2
2
13
18
90.0
SS MS
560.49 560.49
1690.11 845.06
860.27 430.14
1072.08 82.47
4182.S5

F
6.80*
10.25*
5.22*


Cell Means



1
Solvent
2
1OT


                                  pH  (LSD -10.6)
            76.3
                              	7_
                               98.0
                                       10
                                     94.5
                             Interaction  (LSD- 14.9)
(1.2)
61.3
(1.7)
95.2
(1.10)
94.7
(2.2)
 91.3
(2.7)
101.7
(2.10)
 94.3
* Significant .at 0.05 level.
                                   62

-------
 TABLE B-4.   STATISTICAL SUIWARY  FOR BBP  - EXTRACTION  STUDY
Anova Table
Source
Solvents
pH's
Interaction
Error
Total
Mean

df
1
2
2
13
18
- 94.8

SS MS
133.06 133.06
431.28 215.64
546.74 273.37
306.08 23.54
1417.16


F
5.65*
9.16*
11.61



Cell Means
Solvent
                                 pH
                            Interaction
 (1.2)
'88.7""
(1.7)
108.8'
 (1.10)
~S077~"
(2.2)
(217)
89.7
 (2.10)
1BT7-
* Significant at 0.05  level.
                                 63

-------
   TABLE B-5.   STATISTICAL SUMMARY  FOR  DEHP  - EXTRACTION STUDY
 Anova Table
 Source
Total
       df
                              SS
Solvents
pH's
Interaction
Error
1
2
2
. r^ .
143.29
132.19
34.89
fl 25.42
143.29
66.10
17.45
78.88
1.82
<1
<1

                     18
              1335.79
           Mean - 103.1
                                    Means
                     _1	
                     100.5
                                Solvent
                              _J	
                               106.0
              2
            102.8
                                   PH
                    7
                  100.3
                                                   10
                                                   106.7
                              Interaction
(1.2)
 98.3
fl.7)
99.8
                          (1.101
                          103.7
(2.2)
107.3
(2.7)
101.3
(2.10)
109.7
* Significant at 0.05 level.
                                     64

-------
TABLE B-6.  STATISTICAL SUMMARY  FOR  DDP -  EXTRACTION STUDY
Anova  Table
Source
                   df
               SS
               MS
Solvents
pH's
Interaction
Error
Total
Mean -94.2
1
2
2
14^
19

1.25
8.54
72.21
63.75
145.75

1.25 <1
4.27 <1
36.11 7.94*
4.55


Cell Means




1
94.5
Solvent



2
94.0
                                  pH
            93.5
                  7
                94.1
                          10
                         95.2
 (1.2)
 93.3
(1.7)
96.5
                             Interaction (LSD a 3.5)
(1.10?
 93.0
                                       (2.2)
                                       93.7
(2.7)
91.3
(2.10)
97.3
* Significant at 0.05 level.
                                  65

-------
        APPENDIX C

       ANOVA TABLES
PHASE I PRESERVATION STUDY
           66

-------
   TABLE C-T.  STATISTICAL SUMMARY  FOR DMP - PRESERVATION  STUDY
AKOVA Table
Source
PH
Temperature
Chlorine
pH x T
pH x C
Tx C
pH x- T x C
Residual
TOTAL
Mean - 89.0
(2.4)
103.5
df SS MS F
2 10",996.08 5,498.04 2,972*
1 1,107.04 1,107.04 598*
1 22.04 22.04 11.91*
2 6,969.08 3,484.54 1,884*
2 88.08 44.04 23.82*
1 22.04 22.04 11.91*
2 104.08 52.04 28*
10 18.50 1.85
21 19,326.94
Cell Means
pH (LSEH.4)
_L. JL 10
108.0 100.0 59.1
Temperature, C
_4_ 24
95.3 82.2
pH-Temperature
(2.24) (7.4) (7.24) (10.4) (10.24)
112.5 94.0 106.0 90.0 28.2
* Significant at 0.05 level.
                                 67

-------
 TABLE C-2.  STATISTICAL  SUWARY .FOR DEP  -  PRESERVATION STUDY
ANOYA Table
Source
PH
Temperature
Chlorine
pH x T
pH x C
Tx C
pH X T x C
Residual
TOTAL
Mean  96.1
df
2
1
1
2
2
1
2
10..
21

SS
359.08
80.67
4.17
576.08
9.08
10.67
23,08
40.00
1,107.83

MS
179.54
80.67
4.17
288.04
4.54
10.67
14.04
4.00


F
45*
20.17*
1.04
72.01*
1.14
.2.67
3.51



                             Cell Means
                             pH (LSD2.1)
                   -1       -2       J2_
                   98.5        99.1         90.6
                           Temperature. "C
                        J_             24
                        97.9           94.2
                       pH-Temperature (LSD-3.0)
(2.4)
95.8
(2.24)
101.2
(7.4)
98.3
(7.24)
99.5
(10.4)
99.2
(10.24
82.0
* Significant at 0.05 level.
                                  68

-------
  TABLE  C-3.   STATISTICAL SUMMARY FOR DBP -  PRESERVATION STUDY

ANOVA Table
PH
Temperature
Chlorine
pH x T
pH x C
T x C
pH X T x C
Residual
TOTAL
Mean - 86.9


2
86.5
df
2
1
1
2
2
1
2
10
21





SS
219.25
35.04
45.38
907.58
91.75
57.04
30.08
108.50
1,566.62

Cell Means
oH (LSO-4.6)
7
90.3
MS
109.62
35.04
45.38
453.79
45.88
57.04
15.04
18.05




10
83.4
F
6.07*
1.94
2.51
25.14*
2.54
3.16-
<1







pH-Temperature (LSD-6.3)
(2.4) (2.24)
84.0 89.0
iL
87.0
L (7.24)
94.5
(10.4)
 93.3
(10.24)
73.5
* Significant at 0.05 level.
                                69

-------
   TABLE  C-4.   STATISTICAL SUMMARY FOR BBP - PRESERVATION STUDY

ANOVA Table
Source
PH
Temperature
Chlorine
pH x T
pH x C
Tx C
pH x T x C
Residual
TOTAL
Mean * 75.6


2
93.8




!L*L (2.24)
96.0 91.8
df
2
1
1
2
2
1
2
.11
22






4
84.

ILS
74.8
SS
4,616.08
1,802.67.
0.67
2,133.58
23.58
6.00
49.75
433.00
9,065.33

Cell Means
pH (LSO-6.9)
7
72.8
Temperature, C
24
3 67.0
pH-Temperature
1 (7.24)
71.0
MS F
2,308.04 58*
1 ,802.67 45*
0.67 <1
1,066.79 27*
11.79 <1
6.00 <1
24.88 <1
39.68




10
60.3




(10.4) (10.24)
82.3 38.3

* Significant at 0.05 level.
                                 70

-------
   TABLE C-5,  STATISTICAL SUMMARY FOR.DEHP - PRESERVATION  STUDY
 ANOVA Table
Source
PH
Temperature
Chlorine
pHxT
pH x C
Tx C
pH x T x C
Residual
TOTAL
Mean * 93.0
df
2
1
1
2
2
1
2
11
22

SS
709.08
287.04
108.38
29.08
57.25
2.04
188.53
681.50
2,062.90

MS
354.54
287.04
108.38
14.54
28.62
2.04
94.29
61.95


F
5.72
4.63
1.75

-------
TABLE C-6.   STATISTICAL SUMMARY FOR OOP - PRESERVATION STUDY
ANOVA Table
Source
PH
Temperature
Chlorine
pH x T
pH x C
Tx C
pH X T x C
Residual
TOTAL
Mean  88.5




df
2
1
1
2
2
1
2
11
22



2
95.2
SS
549.33
88.17
60.17
214.33
212.33
228.17
129.33
254.00
1,735.83

Cell Means
OH (LSD=5.3)
7
86.2
MS
274.67
88.17
60.17
107.17
106.17
228.17
64.67
23.09




10
84.2
F
11.90*
3.82
2.61
4.64*
4.60*
9.88*
2.80







pH-Temperature (LSD*7.5)
(2.4) (2.24)
97.8 92.8
(7.4)
91.5
1LSSI
81.0
(10.4)
82.2
(10.24)
86.2
oH-Chlorine (LSD-7.5)
(2.0) (2.2)
95.2 95.2
(7.0)
80.5
JL2L
92.0
Temperature-Chlorine (LSO=6
(4.0)
92.0
(4.2)
89.0
III
82
(10.0)
85.2
1)
.0)
.0
(10.2)
83.2

(24.2)
91.3
  Significant at 0.05 level.
                                72

-------
             APPENDIX D

CHROMATOGRAMS OF WASTEWATER FRACTIONS
              PHASE II
                 73

-------
                       PHTHALATE ESTERS
                       WASTEWATER 3-OOSED
                       FRACTION 2
Figure D.I.   Chromatogram Wastewater 3  -  Fraction 2
                             74

-------
                               PHTHAUATE ESTERS
                               WASTEUATER 13-DOSED
                               FRACTIONS 1 & 2 COMBINED
Figure D.2.   Chromatogram Wastewater 3  -  Fractions  1  & 2 Combined
                                    75

-------
               APPENDIX E

CHROMATOGRAMS OF WASTEWATERS 1 THROUGH 5
 SINGLE FRACTION COLLECTION - PHASE II
                 76

-------
      PHTHAUTE ESTERS
         STANDARD
            8X       Col.
Figure E.I.   Standard
             77

-------
      PHTHALATE ESTERS
      UASTEUATER II
Figure E.2.   Wastewater  1
               78

-------
      PHTHALATE ESTERS           +
      UASTEWATER 12             T
      FINAL EXTRACT VOL. 2000 mL   ]
      5 UL      ax      col. i   i

                                  DEHP f' i I ~l i
Figure E-3.   Wastewater 2
               79

-------
         PHTHALATE ESTERS
         WASTEWATER #3            _
         FINAL EXTRACTION VOL. 10W.>-H
         5uL    8X       Col. 1  !
Figure  E.4.   Wastewater  3
                80

-------
      PHTHALATE ESTERS
      WASTEWATER 14
      FINAL EXTRACT VOL. 20 ml
Figure  E.5.   Wastewater  4
             81

-------

-------
        PHTHALATE ESTERS
        WASTEWATER 15
        FINAL EXTRACT VOL. 500;mL
Figure  E.6.   Wastewater  5
                82

-------