PB82-23929"
Laboratory Evaluation of Level 1
Organic Analysis Procedures
Arthur D. Little, Inc.
Cambridge, MA
Prepared for
Industrial Environmental Research Lab.
Research Triangle Park, NC
Jun 82
nHMwufiHHwinaw
U.S. Department of Commerce
Rational Techmczl Information Service
-------
PU82-2J9 29 4
EPA 600/7-82-048
June 1982
Laboratory Evaluation of Level 1 Organic Analysis Procedures
by:
Judith C. Harris
Zoe A. Grosser
Philip L. Levins
Debra J. Sorlin
Clifford H. Summers
Arthur D. Little, Inc.
Acorn Park
Cambridge, Massachusetts 02140
Contract No. 68-02-2150, T.D. 10102
EPA Project Officer: Larry D. Johnson
Industrial Environmental Research Laboratory
Techr.j'.al Support Staff
Research Triangle Park, NC 27711
Prepared for
U.S. Environmental Protection Agency
Office of Research and Development
Washington, DC 20460
c: 79347-16
-------
TFPT PTD l^Ofi M TECHNICAL REPORT DATA
¦LCj ab-ui r-iouu (nease read jMtt/ucttons on the reverie before completing)
t REPORT NO 2
3 RECIPIENT'S ACCESSION NO
4 title anosubtitlE
Laboratory Evaluation of Level 1 Organic Analysis
Procedures
B REPORT OATE
6. PERFORMING ORGANIZATION CODE
7 AUTHOR1SI
J.C.Harris, Z.A. Grosser, P.L. Levins,
D.J.Sorlin, and C.H.Summers
8 PERFORMING ORGANIZATION REPORT NO
9 PERFORMING ORGANIZATION NAME AND AODRESS
Arthur D. Little, Inc.
Acorn Park
Cambridge, Massachusetts 02140
10 PROGRAM ELEMENT NO
11 CONTRACT/GRANT NO
68-02-2150, Task 10102
12 SPONSORING AGENCY NAME ANO ADDRESS
EPA, Office of Research and Development
Industrial Environmental Research Laboratory
Research Triangle Park, NC 27711
13 TYPE OF REPORT AND PERIOD COVERED
Task Final: 1/77-12/78
14 SPONSORING AGENCY COPE
B PA/600/13
is supplementary notes ieRL-RTP project officer is Larry
919/541-2557.
D. Johnson, Mail Drop 62,
i6 abstract rj.^e rep0rj- describes an evaluation of the Level 1 organic sampling and
analysis procedures proposed in 'IERL-RTP Procedures Manual: Level 1 Environ-
mental Assessment,' EPA-600/2-76-160a, June 1976. (This manual has been super-
seded by EPA-600/7-78-201, October 1978.) Priorities of the study included: devel-
opment of a resource of information concerning the behavior of compounds and
classes of compounds when subjected to Level 1 procedures, and identification of
problems and limitations of the proposed procedures that might require revision in
methodology. The report gives results of a series of experimental studies of the
organic analysis procedures as proposed and as eventually modified. Laboratory
studies are described in discrete independent chapters that deal with the individual
experimental investigations.
17 KEY WORDS AND DOCUMENT ANALYSIS
1 DESCRIPTORS
b IDENTIFIERS/OPEN ENDED TERMS
c cosati 1 irld/Gioup
Pollution
Analyzing
Organic Compounds
Sampling
Assessments
Pollution Control
Stationary Sources
Environmental Assess-
ment
13B
14 B
07C
19 DISTRIBUTION STATEMENT
Release to Public
19 SECURITY CLASS (This Report)
Unclassified
21 NO Of PAGES
205
20 SECURITY CLASS (This *agt)
Unclassified
22 PRICE
CPA Form 2220-1 |t>71)
-------
TABLE OF CONTENTS
Page
List of Tables iii
List of Figures viii
Abstract x
Acknowledgement xi
I. INTRODUCTION 1
II. FIELD CC ANALYSIS OF ORGANIC GASES 5
A. Sampling 5
B. Analysis 5
III. PREPARATION AND CHARACTERIZATION OF XAD-2 RESIN 14
A. Clean-up Procedures 14
B. Blank Determination 17
C. Wet vs. Dry Storage of Cleaned Resin 22
IV. EXTRACTION OF AQUEOUS SAMPLES 24
V. EXTRACTION OF SLUDGE/SLURRY SAMPLES 28
VI. ANALYSIS OF VOLATILE SPECIES IN ORGANIC EXTRACTS:
TCO AND SOLVENT EXCHANGE 30
A. Problem Definition 30
B. Quantitative Analysis of Volntiles: TCO 36
C. Sample Preparation for Ext'acts Prior to LC 40
VII. ELUTION PATTERNS IN LEVEL 1 LC 46
A. Model Co-npounds 46
B. Sodium Sulfate Drying of Extracts for LC 50
C. LC Column Blank Due to Silica Gel 53
VIII. RUGGLDNESS TESTING OF LEVEL 1 LC PROCEDURE 54
IX. REPORT FORMATS FOR LEVEL 1 ORGANIC ANALYSIS RESULTS 67
A. Introduction 67
B. Results of the LC Fractionation 69
C. Results of the IR Spectroscopy 71
D. Results of Low Resolution Mass Spectroscopy 75
E. Organic Analysis Summary Tables 80
F. Comparison with Decision Criteria 86
i ^
-------
TABLE OF CONTENTS (Continued)
Page
X. EXAMPLE OF APPLICATION OF LEVEL 1 PROCEDURES
TO FUEL SAMPLES 59
A. Introduction 89
B. Level 1 Sampling and Analysis of Fuels 90
C. Level 1 Organic Analysis Results for a
Fuel Oil Sample 92
D. Level 1 Organic Analysis Results for a
Coal Sample 105
XI. EXAMPLE OF APPLICATION OF LEVEL 1 PROCEDURES
TO A SASS TRAIN SAMPLE 120
A. Introduction 120
B. Sumnary of Level 1 Results 120
C. Detailed Level 1 Results 126
XII. REFERENCES 150
APPENDIX A: LEVEL 1 ORGANIC ANALYSIS TECHNIQUES A-l
APPENDIX B: PREPARATION OF XAD-2 SORBENT RESIN B-l
ii
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
LIST OF TABLES
Page
Rate of Pressure Change of Previously
Evacuated Glass Bulb Fitted with Critical
Flow Orifice 6
Recommended Procedures for Field GC of
Organic Gases 10
Elution Volumes of XAD-2 Prepared by
Two Methods 16
Summary of LRMS Res.ults for Aromatic
Fractions (LC Fraccion 2) of Sorbent
Trap Extracts 21
Blank Levels for Wet vs. Dry Storage
of Cleaned XAD-2 Resin 23
Illustrative Values of Solvent;
Water Partition Coefficients 26
Relative Recoveries of Various Compounds
After Extraction from Water 27
Rate of Evaporation of Model Compounds 31
Effect of Concentration of Pentane Extracts
on Loss of Model Compounds 33
Recovery of Phenol After Drying Solvent
Extracts Containing High Molecular Weight
Organics 35
Recoveries of Model Compounds After
Solvent Exchange with Hexane and
Cyclopentana 41
Breakthrough of Aroraatics into LC Fraction 1 43
Recoveries of Model Compounds After Solvent
Exchange 45
Distribution of Model Compounds among
LC Fractions 47
% Distribution in LC Fractions 48
iii
-------
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
LIST OF TABLES (Continued)
Page
Categories for Reporting of LRMS Data 49
Recovery of TCO and of Several Model
Compounds from Sodium Sulfate Columns:
Dry Solutions 51
Recovery of TCO and of Several Model
Compounds from Sodium Sulfate Columns:
Wet Solutions 52
Change in LC Solvent Composition with Time 57
Summary of Expected Data from Level 1
Organic Analysis 68
Level 1 LC Results Report Form 70
Distribution of Organic Compound Classes
among LC Fractions 72
Level 1 IR Results Report Form 73
Categories for Reporting of LRMS Data 76
Level 1 LRKS Results Report Form 79
Level 1 Organic Extract Summary Table 81
Summary of Results for Organic Extracts
for SASS Train Samples 87
Organic Extract Summary Table for
//4 Fuel Oil 93
LC Report for Fuel Oil 94
IR Reports: Fuel Oil Fractions LCI, LC2 95
IR Reports: Fuel Oil Fractions LC3, LC4 96
IR Reports: Fuel Oil Fractions LC5, LC6 97
IR Report: Fuel Oil Fraction LC7 98
LRMS Report: Fuel Oil Fraction LCI 99
iv
-------
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
LIST OF TABLES (Continued)
Page
LRMS Report: Fuel Oil Fraction LC2 100
LRMS Report: Fuel Oil Fraction LC3 101
LRMS Report: Fuel Oil Fraction LC4 102
LRMS Report: Fuel Oil Fraction LC6 103
LRMS Report: Fuel Oil Fraction LC7 104
Organic Extract Summary Table for
Coal Sample 106
LC Report for Coal Extract 107
IR Reports: Coal Extract LCI, LC2 108
IR Reports: Coal Extract IC3, LC4 109
IR Reports: Coal Extract LC5, LC6 110
IR Report: Coal Extract LC7 111
LRMS Report: Coal Extract LCI 112
LRMS Report: Coal Extract LC2 113
LRMS Report: Coal Extract LC3 114
LRMS Report: Coal Extract LC4 115
LRMS Report: Coal Extract LC5 116
LRMS Report: Coal Extract LC6 117
LRMS Report: Coal Extract LC7 118
Total Extractable Organics for SASS Train
Samples of Ferromanganese Process Emissions 121
Organic Extract Summary Table: Ferro-
manganese Process SASS Condensate 122
Organic Extract Summary Table 123
Organic Extract Summary Table 124
-------
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
LIST OF TABLES (Continued)
Page
Total Organics for SASS Train Samples
of Ferromanganese Process Effluent 125
LC Report: Ferromanganese Probe Wash Sample 127
IR Report: Ferromanganese Probe Wash
LCI, LC2 128
IR Report: Ferromanganese Probe Wash
LC3, LC4 129
IR Report: Ferromanganese Probe Wash
LC5, LC6 130
IR Report: Ferromanganese Probe Wash
LC7 131
LRMS Report: Ferromanganese Probe Wash
LCI 132
LRMS Report: Ferromanganese Probe Wash
LC2 133
LRMS Report: Ferromanganese Probe Wash
LC3 134
LRMS Report: Ferromanganese Probe Wash
LC4 135
LRMS Report: Ferromanganese Probe Wash
LC6 136
LRMS Report: Ferromanganese Probe Wash
LC7 137
LC Report: Ferromanganese XAD-2 Sorbent
Trap Sample 138
IR Report: Ferromanganese XAD-2 Sample
LCI, LC2 139
IR Reports: Ferromanganese XAD-2 Sample
LC3, ~C4
IR Report: Ferromanganese XAD-2 Sample
LC5, LC6 141
vi
-------
L1S1 OF TABLES (Continued)
Table No. Page
73 IR Report: Ferromanganese XAD-2 Sample
LC7 142
74 LRMS Report: Ferromanganese XAD-2 Sample
LCI 143
75 LRMS Report: Ferromanganese XAD-2 Sample
LC2 144
76 LRMS Report: Ferromanganese XAD-2 Sample
LC3 145
77 LRMS Report: Ferromanganese XAD-2 Sample
LC4 146
78 LRMS Report: Ferromanganese XAD-2 Sample
LC5 147
79 LRUS Report: Ferromanganese XAD-2 Sample
LC6 148
80 LRMS Report: Ferromanganese XAD-2 Sample
LC7 149
vii
-------
2
3
4
5
6
7
8
9
10
11
12
13
14
>ge
3
4
7
11
12
34
37
38
39
55
58
59
60
61
LIST OF FIGURES
Original Multimedia Overview of Samples
for Organic Analysis
Original Level 1 Organic Analysis Flow
Scheme
Filling Rate of Previously Evacuated
2L Glass Bulb Fitted with 0.2 LPM
Critical Flow Orifice
GC of -160°C to 30 °C B.P. Range
Organic Gases
GC of 30°C to 100°C B.P. Range
Organic Gases
Recoveries of N-Hydrocarbons vs. Boiling
Point after Concentration of Pentane
Solutions
TCO Chromatogram for Cg, C^2» and Cj^
Quantitative Calibration Mixture
TCO Chromatogram for Aqueous Sample
Extract
Example Calibration Curve for Level 1
TCO Analysis
Distribution of Model Compounds in the
LC Scheme with Moderate Change in Proceduie
Distribution of Model Compounds in the
LC Scheme with Moderate Change in Procedure
Distribution of Model Compounds in the
LC Scheme with Moderate Change in Procedure
Distribution of Model Compounds in the
LC Scheme with Moderate Change in Procedure
Distribution of Model Compounds in the
LC Scheme with Moderate Change in Procedure
viii
-------
LIST OF FIGURES (Continued)
Figure No. Page
15 Distribution of Model Compounds in the
LC Scheme with Moderate Change in Procedure 62
16 Distribution of Model Compounds in the
LC Scheme with Moderate Change in trocedure 63
17 Distribution of Model Compounds in the
LC Scheme with Moderate Change in Procedure 65
18 Distribution of Model Compounds in the
LC Scheme with Moderate Change in Procedure 66
19 Example Calculations of Concentration
Estimates from LRMS Data 83
20 Estimation of Fraction Composition from
IR and LC Data Only 85
ix
-------
ABSTRACT
Level 1 is the first stage in a three-tiered approach to performing an
environmental source assessment. Level 1 is designed to provide enough
information about the composition of effluent and process streams to
permit them to be ranked in order of priority for probable environmental
hazard. A set of sampling and analysis procedures designed to achieve
these objectives was developed and published in an EPA report in
June, 1976. That methodology was based on previously available laboratory
procedures, which had not, however, been specifically tested to determine
their suitability for this particular purpose. The overall objective of
the work described in this report was to evaluate the Level 1 organic
sampling and analysis procedures as proposed in the June, 1976 manual.
Priorities of this study included: the development of a resource of
information concerning the behavior of compounds and classes of compounds
when subjected to Level 1 procedures; and identification of problems
and limitations of the proposed procedures that might require revisions
in methodology. This report presents the results of a series of
experimental studies of the organic analysis procedures as proposed
and as eventually modified. The report also includes several examples
of Level 1 organic analysis data for samples analyzed according to the
revised procedures. The results for a coal sample, a fuel oil sample,
and a SASS train sample of an actual emission source are presented in
the Level 1 report format.
x
-------
ACKNOWLEDGMENT
This report has been submitted by Arthur D. Little, Inc., in partial
fulfillment of the requirements of EPA Contract No. 68-02-2150,
Technical Directive 10102. The work on which this report is based
was performed foi the Process Measurements Branch/Technical Support
Staff of the Industrial Environmental Research Laboratory at Research
Triangle Park, North Carolina. Throughout this method evaluation and
development effort, the EPA Project Officer, Dr. Larry Johnson, and
also Dr. Raymond Merrill and Mr. James Dorsey provided valuable
guidance and support. The authors are also grateful to the following
employees of Arthur D. Little, Inc. who contributed to the laboratory
work described here: Paulina Alexander, Rafael Cruz-Alvarez, Leslie
Root, Julie Rjdolph, Emmett Smith, James Stauffer, =ind Karen Weaver.
xi
-------
I. INTRODUCTION
The Process Measurement Branch of IERL/RTP has developed a three-
tiered or phased approach to performing an environmental source assess-
ment. The phased sampling and analytical strategy was developed to
focus available resources (both manpower and dollars) on emissions that
have a high potential for causing measurable health or ecological effects,
and to provide comprehensive chemical and biological information on all
sources of industrial emissions.
The phased approach requires three separate levels of sampling and
analytical effort. The first, Level 1, Is designed to provide enough
information about the composition of effluent and process streams to
permit them to be ranked in order of priority for probable environmental
hazard. The Level 1 assessment is intended to: 1) provide preliminary
environmental assessment data, 2) identify principal problem areas, and
3) provide the data needed for prioritization of energy and industrial
processes, streams within a process, components within a stream, and
classes of materials, for further consideration in the overall
assessment. The second phase of sampling and analysis effort, Level 2,
is designed to provide additional information that will confirm and
expand the information gathered in Level 1. The third phase, Level 3,
will permit quantitative monitoring of specific pollutants identified
in Level 2.
A set of sampling and analysis procedures designed to achieve the
objectives of Level 1 environmental assessment was developed and
published In an EPA report In June, 1976. (1) The methodology selected
was based on previously available laboratory procedures. However,
time constraints mitigated against experimental Investigation of the
entire, integrated protocol prior to its initial publication. The
Level 1 procedures are comprehensive, covering inorganic and organic
chemical analysis, and biological testing.
1
-------
The overall objective of the work described in this report was to
evaluate the Level 1 organic sampling and analysis procedures as
proposed in the June, 1976 manual. Priorities of this study included:
development of a resource of information concerning the behavior of
compounds and classes of compounds when subjected to Level 1 procedures;
and identification of problems and limitations of the proposed
procedures that might require revision in methodology. The question of
stability of Level 1 samples collected for organic analysis has been
addressed in two separate studies (2,3).
The Level 1 organic analysis methodology is designed to identify the
major classes of organic compounds present in a process or effluent
stream and to estimate their concentrations. The overall organic
procedures as originally proposed are summarized in Figure 1 (sampling)
and Figure 2 (analysis) .
This report presents the results of a series of experimental studies of
the organic analysis procedures as proposed and as eventually modified.
Laboratory studies are described here in discrete independent chapters
that deal with the individual experimental investigations. The topics
are presented in an order that seems to be logical in retrospect. The
experimental work was carried out at Arthur D. Little, Inc. over the
period January, 1977 to December, 1978 and documented in monthly
progress reports. An integrated overview of the evaluated and
revised procedures can be found in Chapter Q of the second edition of
the Level 1 Procedures Manual (4). For convenience, that chapter is
reproduced as Appendix A of this report.
This report also includes several examples of Level 1 organic analysis
data for samples analyzed according to the revised procedures. The
results for a coal sample and a fuel oil sample and a SASS train sample
of an actual emission source are presented in the Level 1 report format.
2
-------
Of«t> limrfm
I
I
S«u Tinn
i, lOltl tAMPtt
¦
) ICItAOtHMf!
l4"hl ¦ lIM,
I, UilAl
t, IC ikAi i h'liAitn
I lOUt SAMT1I
/ IllMl )«/iWll
J IC
Sou ice ni.li.rciM.ti I
FIGURE 1 ORIGINAL MULT . tEDIA OVERVIEW OF SAMPLES fOfl ORGANIC ANALYSIS
-------
ON FRACTIONS IXCLIDINC
IHRf SHOID CRIK RIA
XAD-?
SORdlNT
TRAP
SOLIDS
I (CHI FRAC1IONS
•M IGHI
SOLID
MATERIALS
ORGANIC
LIQUID^
LIOUJD
CHROMATOGRAPHIC
SEPARATION
CONCf NTRATl
ALIO' Of
(8 - 600 mgl
ON Sill G.C.
fORC,-C6
HYOROCARBONS
INFRARED
ANA! YSlS
SOLVl NT
[XTRACltON
LIQUIDS
IOW Rl SOLUTION
MASS SPtCTRA
ANALYSIS
CONCENTRAM
AllOtOT FOR
C RAVI METRIC
AND IR
ORGANIC
ANALYSIS
ac iauor
K>R HYDRO-
CARBONS
IOC.,
Source. Reference 1.
FIGURE 2 ORIGINAL LEVEL 1 ORGANIC ANALYSIS FLOW SCHEME
4
-------
II. FIELD GC ANALYSIS OF ORGANIC GASES
The original Level 1 Procedures (1) recommended that organic gases
(species with boiling points < 100°C) be collected in glass sampling
bulbs and analyzed on-site by gas chromatography (Field GC procedure).
A. Sampling
Although the possibility of substituting polymeric plastic sampling
bags was advocated, systematic studies (3) indicated that glass bulbs
were the sampling system of choice for Level 1 purposes.
The instantaneous nature of gas grab samples with evacuated glass
bulbs was of some concern. Experiments were performed to verify that
a degree of time-integrated sampling with a previouly evacuated glass
bulb can be achieved. If a glass fiber filter and an 0.2 Lpm critical
flow orifice are used upstream, a 2 L bulb f^lls gradually over a 12-
minute period after the stopcock is opened. The data are presented in
Table 1 and Figure 3.
B. Analysis
The GC procedures listed in the June, 1976 Procedures Manual (1) for
gas chromatographic analysis of sulfur gases and gaseous hydrocarbons
were reviewed and found not to be satisfactory for Level 1 purposes. The
revised procedures below were tested and recommended as replacements.
1. Sulfur Gases
Column: 36' x 1/8" TeflonR, 12% polyphenylether and .5%
H3PO4 on 40/60 mesh Chromosorb T (Note 1)
Detector: Flame photometric (Note 2)
5
-------
0
1
2
3
4
5
6
7
8
9
10
11
12
TABLE 1
RATE OF PRESSURE CHANGE OF PREVIOUSLY EVACUATFD
GLASS BULB FITTED WITH CRITICAL FLOW ORIFICE
Differential Pressure (mm)
Run 1 Run 2 Run 3
710 690 710
620 600
520
420 424 i»40
340 340
260 254
184 188 190
120 126
70 80 86
40 40
20 20
10 10
0 0 6
6
-------
a.
at
Q>
5
700
600
500
400
300
200
100
0
0
3
10
4
6
Time
-------
Conditions: Helium carrier at 30 mL/min.
Injector and detector temperatures: 110°C (Note 3)
Column oven temperature program:
isothermal at 60°C for 5 min.
increase at 10°min. to 100°C
hold at 100°C for 10 min.
(Note 4)
Notes: 1. Exit end of column should be fitted directly into the
base of the detector, eliminating any metal transfer
lines. On-column injection should be used.
2. The FPD response tends to saturate at concentrations
greater than 2 ppm for a 10 mL injection volume.
Injection should be repeated with smaller (1 or 0.1 mL)
sample size if apparent concentration exceeds 2 ppm.
3. Detector '.emperatures above 130°C are reported to
result in losses of sulfur species.
4. The conditions specified have been selected to give
the following approximate retention times:
H2S 3 min.
S02 4 min.
CH3SH 6 min.
C3H7SH 10 min.
Slight modifications in the temperature and duration of the
initial hold period may be necessary to accommodate variations
in individual column performance.
2. Gaseous Hydrocarbons
The column choices and temperature programs specified in the June,
1976 manual for field GC of organic gases need to be modified because:
a. The "C -j-C analysis was incorporated into an additional
(TCO) procedure. (See Chapter VI)
b. A single GC analysis procedure could not be found that would
cover the entire -160°C (methane) to 96°C (heptane) b.p. range.
8
-------
In the new procedure, described below, the -160°C to 30°C b.p. range
is analyzed using a temperature programmed run on Porapak Q. The
30°C to 100°C b.p. range is analyzed isothermally on 20% OV-101.
Two separate GC procedures are required in order to allow resolution
of the very volatile organics while eluting the higher boiling gases in
a reasonable period of time. The GC systems will primarily be separating
and analyzing mixtures of materials with given boiling point ranges,
although the separations will also be influenced by polarity in some
cases. A flame ionization detector is used.
The conditions recommended for these analyses are specified in Table 2.
Slight modifications in temperature and duration of the isothermal hold
periods and/or rate of temperature increases may be necessary to
accomodate variations in individual column performance.
Figures 4 and 5 present example chromatograms for the low boiling and
higher boiling calibration mixtures, respectively, under the recommended
conditions.
Mixtures of r rmal hydrocarbons are used to calibrate the field GC
procedures. A retention time vs boiling point calibration curve is
prepared for each procedure in the standard manner. The Level 1
boiling point ranges and the hydrocarbons falling in each range are:
Level 1
B.p
. Range
B.p.
Designation
Cc)
Hydrocarbon
(°C)
GC1
-160
to
-100
Methane, Cj
-161
GC2
-100
to
-50
Ethane, Cj
-88
GC3
-50
to
0
Propane, C3
-42
GC4
0
to
30
Butane, C4
0
GC5
30
to
60
Pentane, C5
36
GC6
60
to
90
Hexane, Cg
69
GC7
90
to
100
Heptane, C7
96
9
-------
TABLE 2
RECOMMENDFD PROCEDURES FOR FIELD GO OF ORGANIC GASES
>-*
o
B.p. Range
-160°C to 30°C
(GC1 to GC 4)
30°C to 100°C
(GC5 ro GC7)
Column
2 m x 2 nun
(6* x 1/6")
Stainless Steel,
Porapak Q
2 m x 2 mm
(6' x 1/8")
Stainless Steel,
20% OV-101 on
100/120 mesh
Chromosorb W-HP
Temperature <*itd Conditions
Helium carrier at 20 mL/min
Injector and Detector
temperatures: 200 C.
Column oven temperature program:
isothermal at gO for 4 min;
increase at 20 /min^to
110°; hold at 110°.
Bake-out at 170 C as necessary
between injections.
Helium carrier at 20 mL/min
Injector and Detector
temperatures: 200 C.
Column oven temperature:
isothermal at 30°C.
Calibration
Mixture
10 ppm (v/v)
each of:
methane
ethane
propane
n-butane
n-pentane
10 ppm (v/v)
each of:
n-butane
n-pentar.e
n-hexane
n-heptane
* Under these conditions the approximate retention times of the n-hydrocarbons are: C^, 1 min;
C2» 3 min; C^* 8 min; C^, 15 min; C^, 40 min.
** Under these conditions the approximate retention times of the n-hydrocarbons are: C,, 1 min;
Cy 2 min; Cg» 5 min; C^, 12 min.
-------
Condition*—60° C for 4 minute* than program to 110° "? 16°/mm
2S0°C Iniecrion Port
300°C Detector
C, c2
r-HY—H
Concentration—Approximately 10 ppm each component
C) ¦ Methart
Cj ¦ Ethane
C3 ¦ Propane
CA ¦ Isobutane
C5 • Pentane
Souic* Court«$y of Raymond MerritJ. PMB-SPA-RTP
FIGURE 4 GCOF-160°CT0 30°C8J» RANGE ORGANIC GASES
11
-------
c
E
ID
o
Conditions. As Specified in Table 2
Concentration: Approximately 10 ppm each component
C^ = n-Butane
Cg = n-Pentane
Ce = n-Hexane
d ®
E C-j = n-Heptane
\jv
C4 C5
Source: Arthur D Little, Inc
FIGURE 5 GC OF 30°C TO 100°C B P. RANGE ORGANIC GASES
12
-------
The GC systems should also be calibrated for quantitative analysis with
the normal hydrocarbon mixtures. Assumption of uniform FID response
for varying compound classes is acceptable in Level 1 analysis.
Since the chromatogran peaks for Level 1 samples will usually represent
mixtures of materials present in a certain boiling range, rather than
pure individual compounds, it is recommended that the chromatographic
data be reported as follows:
GC1
GC 2
GC3
GC 4
GC5
GC6
GC7
It is important to recognize that in many, if not most, cases, the
species present will not, in fact, be identical to those used for
calibration of the on-site procedure.
Level 1
Designation
No. of Peaks
Observed
13
-------
III. PREPARATION AND CHARACTERIZATION OF XAD-2 RESIN
A macroreticular crosslinked polystyrene resin, XAD-2, was selected
for use in the SASS train organic vapor sampling module (1) because of
its high collection efficiency for a variety of organic compounds. The
optimum procedures for preparation (clean-up) and analysis of the blank
for XAD-2 resin were not specified in the June, 1976 manual.
A. Clean-up Procedures
The Procedures Manual for cleaning XAD-2 called for sequential
extraction with water, methanol, ether and pentane. It was felt
desirable to change the extraction sequence to one of water, methanol and
methylene chloride for several reasons: there are some hazards associated
with reflux of diethyl ether; a three-step sequence would be time
saving; use of methylene chloride would result in one common solvent
for all of the organic extracts; and the solubility of many compounds is
better in methylene chloride than pentane.
Two different 3-step conditioning procedures (water + methanol + pentane
and water + methanol + methylene chloride) were compared to the original
4-step procedure (water + methanol + diethyl ether + pentane). Para-
meters investigated include the surface properties of resin (pore volume
and pore size distribution) and the quantity and nature of residual
solvent extractable materia]. Further experiments examined the
possibility of resin self-contamination by thermal processes (at 20°,
40° and 60°C) and the efficiency of recovery of spiked materials.
Based on these studies, the new recommended procedure is to clean the
XAD-2 resin with the water, methanol, methylene chloride overnight
extraction sequence and to extract the collected SASS sample with
methylene chloride.
14
-------
Several batches of XAD-2 were prepared by the solvent sequence described
above. All batches appeared to behave similarly based upon general
observations. In order to determine whetner there was any difference
between the old extraction sequence (pentane finish) and the new
sequence (methylene chloride finish), compound elution behavior, surface
areas, and residual blank levels were examined.
The elution volumes, Vg(mL/g), were compared for XAD-2 prepared by the
two routes by testing with phenol, octane and benzene. The results
are shown in Table 3. There was no difference in Vg between the batches
of resin prepared by the two techniques, within experimental error.
Surface areas of the XAD-2 prepared by the two methods were:
Surface Area Pore Volume
Final Solvent (m2/g) (cc/g)
Pentane 366 0.891
Methylene Chloride 376 0.886
To compare residual (blank) contaminant levels, 20 g samples of XAD-2
prepared by each method vere extracted.with both pentane and methylene
chloride. Half of the extract from each sample was taken to dryness
for gravimetric analysis. Thus the residue data obtained represented
that derived from 10 g of XAD-2 or about 7% of the amount of XAD-2
in a SASS module. The results obtained were as follows (mg):
Pentane Methylene Chloride
Extract Extract
Pentane Clean 0.0 0.6
Methylene Chloride Clean 0.0 0.1
The data showed that acceptable levels of residue were obtained by
both methods. If the differences were taken tc be significant, then
it would appear that methylene chloride did a better initial job in
cleaning the XAD-2 resin.
15
-------
TABLE 3
ELUTION VOLUMES OF XAD-2 PREPARED BY TWO METHODS
Challenge
Compound \ Solvent**
C5H12
(mL/g) at Temp.
96°C
135°C
ch2ci2
C5H12
CH2C12
Phenol
Octane
Benzene
4230
2390
425C
2400
382
215
47
386
211
41
Vg is the volume of gaseous sample that can be passed over a 1 g
sorbent bed in a frontal chromatography experiment before the exit
concentration of organic compound exceeds 50% of the challenge
concentration.
Final solvent in clean up.
16
-------
Details of a revised clean-up procedure, including a quality control
check to verify acceptable blank levels of extractable material, are
given in Appendix B.
B. Blank Determination
A series of experiments were conducted in an effort to quantify
and characterize the material that might result as a blank
on XAD-2 sorbent traps. These experiments were done at one-fifth scale
compared to the SASS Level 1 Procedure. A glass sorbent trap was
used; this trap has the same sorbent bed depth as the SASS sorbent
module but is smaller in diameter. The glass trap held 41.±1. g of
XAD-2 resin, previously cleaned by sequential extraction with water,
methanol, diethyl ether, and pentane.
A clean gas stream was generated by drawing nitrogen gas from a
cylinder of liquid nitrogen. The gas stream temperature was controlled
at 20, 40 or 60°C by passing the gas through a 182 cm (6') x 12.5 cm
(0.5") diameter heated glass-lined probe immediately upstream of the
XAD-2 trap. The trap itself was jacketed with a heating mantel. Tem-
peratures were monitored by thermocouples at the probe exit and at the
outer surface of the sorbent trap.
A Research Appliance Corporation Staksampler^ control module was used
to control the nitrogen flow rate and to measure the total volume of
nitrogen passed through the trap. The measured volume was 6.53 ± 0.03
standard cubic meters (230.7 ± 0.9 standard cubic feet).
After exposure of the traps to the nitrogen stream, the XAD-2 resin
was transferred to pre-extracted Soxhlet thimbles and extracted for
24 hours with methylene chloride. Each extract was transferred to a
250 mL volumetric flask and made to volume with additional methylene
chloride. The weight of material extracted was determined by evapora-
tion of 150 mL of the extract to dryness in a tared container at ambient
conditions.
17
-------
Found
mg/150 mL Calculated
of Extract mg/trap
20° Trap 2.985 A.98
40° Trap 1.731 2.88
60° Trap 1.628 2.72
The quantities of material found were only somewhat higher than the 1-2
mg amounts that have routinely been found on evaporation of 200 mL of
solvent. The quantities found for the traps exposed to nitrogen at
elevated temperature were not significantly higher than the 1.5-3.8 mg
amounts that had been found in previous work for unexposed (40 g) traps
extracted with pentane and methanol.
The residue from the 150 iaL of extract was mixed with a small quantity
of silica gel and transferred to the top of a silica gel column prepared
according to the Level 1 LC procedures. A four fraction separation was
performed using the sequence of solvents shown below:
Fraction No. Elution Solvent
1 20 mL pentane
2 10 mL methylene chloride
3 10 mL 50% methanol in
methylene chloride
A 10 mL of 70-30-5: methanol,
methylene chloride, cone HC1.
A control column, to which no sample was added, was run in parallel with
the samples.
In this 4-fraction LC procedure, fraction 1 contains the aliphatic
hydrocarbons, fraction 2 the aromatlcs, fraction 3 the polar (oxygenated)
material and fraction h any very polar compounds (carboxylic acids).
The aromatic fraction was investigated most thoroughly because it was
expected that any decomposition products of the styrene-divinyl benzene
resin would be aromatic compounds.
18
-------
Quantitative gas chromatographic analysis was done both on the unconcen-
trated initial total extracts and on the aromatic fractions from »-ne LC
separation. The resulting estimates of total quantities of material
with boiling points lower than about 320°C were:
mg/trap (40 g XAD-2)
Unconcentrated Aromatic
Extract Fraction
20° Trap <0.8 <0.000
40° Trap t0.4 £0.004
60° Trap <0.5 <0.002
Control LC Column (NA) £0.003
(no sample)
A 1 idL aliquot of the aromatic fraction from the 40° trap LC separation
was evaporated to dryness in a tared aluminum pan. After evaporation of
solvent, no change in pan weight was found; the quantity of non-volatile
material was therefore
-------
Examination of the aromatic fractions by LRMS, using the direct insertion
probe led to the results shown in Table 4. Because the total signal
intensity was found to be so low, a one mL sample (rather than the more
usualy 100 yL) was dried down for the probe. The results in T?.^le A
indicate that the major aromatic species produced by exposure of XAD-2
to heated nitrogen gas were alkyl benzenes, small polymers of styrene,
and fragments of polymers. The composition of the polymers and fragments
was confirmed by HRMS peak matching of a number of peaks. Both the
relative and the absolute quantities of stryrene-like polymers and frag-
ments increased as the trap exposure temperature increased. The total
amount of material that was volatile in the prote also increased. This
indicates that some thermal decomposition or thermal desorption of en-
trained material is occurring and, in fact, the species found include the
types of materials which would be expected to arise from decomposition
of the styrene-divinyl benzene resin. The polymers and fragments appear
in the LRMS at probe temperatures of about 150°C. The phthalates and
polynuclear aromatic hydrocarbons found in roughly equal amounts in all
four cases, including the aromatic fraction from the control column, are
most probably contaminants introduced by the solvent, the silica gel or
other sources.
In summary, the experiments reported here indicate that the only signifi-
cant contaminants that can be expected as a sorbent trap "blank" due to
thermal decomposition of the XAD-2, in the temperature range of 20-60°C,
are small polymers and fragme-its of polymers that include 1 to 4 styrene
units. These species would appear in the aromatic fraction(s) of the LC
scheme and would be detected by the LRMS analysis. The quantity of material
which these species represent is expected to be too small to detect by
gravimetric analysis (i.e., <0.1 mg), although the apparent total "blank"
—including solvent-introduced and other contaminants like silicones and
phthalates—may be higher.
20
-------
TABLE 4. SUMMARY OF LRMS RESULTS FOR AROMATIC FRACTIONS
(LC FRACTTOT1 2) OF SORBKNT TRAP EXTRACTS
Control Column 20" 40® 60°
(no sample) Trap Trap Trap
Total LRMS signal
intensity (arbitrary units)
for 1 ml sample 1000 800 1700 3600
Distribution of Species:
Polynuclear Aromatics
(napthalene, anthracene,
pyrene) 37, 13 7. 4% 3 7.
Phthalates 4% 8% 4% 27,
Alkyl benzenes, MW
91 and 105 3% 33% 15% 162
Styrene Polymers and Fragments
1 styrene unit 25% 38% 40%
2-4 styrene units 18% 22% 28%
Others mw 112, 20% nw 145, 3% mw 129, 4% mw 129
129, 25% 145, 67, 145
167, 2% 1 167
-------
C. Wet vs. Dry Storage of Cleaned Resin
An experiment was carried out to ascertain the effect on cleaned XAD-2
resin of dry storage vs. storage under methanol vs. storage under methy-
lene chloride until just prior to use.
A single lot of resin was cleaned according to Level 1 procedure using
Burdick 6 Jackson solvents. The cleaned resin was divided into three
portions: one was stored under methanol, the second was stored under
methylene chloride, and the third was dried (at a later date) according
to Level 1 procedure.
The three portions (dried, methanol storage, and methylene chloride
storage) were extracted after allowing the dry material to stand 60 days
and the two wet portions 90 days. The extracts were concentrated from
200 mL to 10 mL by K-D concentrators, and portions taken for TCO* and
GRAV. Results are in Table 5.
The total blank levels measured for dry storage and for methylene chloride
storage were about what one would expect for samples that have passed the
Level 1 QC check (Appendix B). Considerably higher blank levels were
found for XAD-2 stored under methanol.
These data indicate that wet storage under methanol is not the
method of choice for XAD-2 to be used in Level 1 air sampling.
*Total Chromatographable Organics. See Section VI B
22
-------
TABLE 5
BLANK LEVELS FOR WET VS DRY STORAGE OF CLEANED XAD-2 RESIN
TCP*
GRAV
TOTAL
Dried Resin
0.001 mg/g
0.044 mg/g
0.04 mg/g resin
MeOH Storage
0.024 mg/g
0.182 mg/g
0.21 mg/g resin
CH2CI2 Storage
0.005 mg/g
0.030 mg/g
0.04 mg/g resin
Total Chromatographic Organics. See Section VI.
23
-------
IV. EXTRACTION OF AQUEOUS SAMPLES
The June, 1976 Level 1 Procedures Manual (1) suggested
Chat aqueous samples be adjusted to neutral pH and then sequentially ex-
tracted with three 500-raL portions of methylene chloride for a 10 liter
water sample.
A literature study of the efficiency to be expected from methylene
chloride solvent extraction of aqueous samples, based on octanol-water
partition coefficients, was conducted. That study indicated that the
original Level 1 procedure (1), which called for three extractions at
neutral pH, was inadequate for Level 1 purposes, since moderately strong
organic acids and bases would be <0.1% extracted.
The proposed revised procedure for Level 1 extractions was as follows:
Extraction of aqueous solutions should be carried out with
methylene chloride using a standard separatory funnel fitted
with a Teflon stopcock. The pH should be adjusted to 2.0 ±
0.5 with hydrochloric acid, and then to 12.0 ± 0.5 with
sodium hydroxide, using multirange pH paper for indication.
Two extractions at each pH using 500 mL portions of methylene
chloride for a 10 liter sample should be sufficient.
(Extraction first at acid pH is recommended to reduce
emulsion formation by removing some surfactants in their
neutral form. If emulsions do occur, however, they can
generally be broken by centrifugation due to the high
density of chlorinated solvents.)
Neutral species, whose recovery is independent of pH, are essentially
extracted four times by this procedure, while acidic and basic species
are each extracted twice. The following shows the organic solvent/water
partition coefficients necessary to achieve a specified % recovery by the
24
-------
modified, 2 pH, procedure:
Necessary P , _ for
Type of Effective No. org/H2u
Species of Extractions 50% Recovery 75% Recovery
Neutral 4 3.8 8.3
Acidic, Basic 2 8.3 20.0
Table 6 presents estimated PQrg^p q values for some species that are
readily extractable and some that are essentially non-extractable, even
by the revised procedure.
Partition coefficient data suggest that etner may be superior to methyl-
lene chloride for-extracting certain acidic compounds, such as phenols
or carboxylic acids. This was tested experimentally to see if the ad-
vantage was significant. A stock solution of phenol, N-methyl aniline,
acenaphthene.n-hexadecane and 4,4,'dichlorobiphenyl in methanol was
spiked into samples of reagent grade water. Samples were made basic
(pH 11) and extracted twice with 50 mL of methylene chloride, then
made acidic to approximately pH 2. One was extracted twice with 50 mL
of methylene chloride and the extract combined with that obtained at pH
11, while the ether was extracted twice with 50 mL of ethe" and the ex-
tract combined with that obtained at pH 11.
The experimental data are presented in Table 7. Four of the five model
compounds were ejtracted efficiently with either methylene chloride or
ether. Phenol was extracted four times more efficiently by ether than
by methylene chloride yielding an 82% recovery versus only a 20% recovery.
In view of the inconvenience (lighter than water) and potential peroxide
explosion hazard associated with ether use, the improvement in efficiency
was considered to be or.ly marginally important.
Since, furthermore, higher homologs of phenol such as cresols were ade-
quately extractable with methylene chloride, the use of diethyl ether
was not recommended for Level 1 purposes.
25
-------
TABLE 6
ILLUSTRATIVE VALUES OF SOLVENTS: WATER PARTITION COEFFICIENTS*
Compound
P P P
Efficiently Extracted CHCl^:H^O heptane: 1^0 ether: 1^0
Propyl Ether 580. 34. 18.
Hexyl Amine 490. 30. 16.
Butyl Mercaptan 1,200. 98. 34.
Benzene 460. 24. 15.
Chlorobenzene 6,200. 1,100. 150.
Indene 78,000. 1,500. 180.
Acridine 20,000. 9,300. 575.
Thiophene 301. 13. 10.
Inefficiently Extracted
Acetic Acid
Phenol
Benzene Sulfonic Acid
Ethylene Glycol
0.03
2.
0.0001
0.002
1.0
0.05
0.000006
0.00001
0.4
31.
0.0C2
0.005
Estimated using data base and methods of Reference _>.
26
-------
TABLE 7
RELATIVE RECOVERIES OF VARIOUS COMPOUNDS AFTER EXTRACTION FROM WATER*
% Recovery
Model Compound Methylene Chloride
Phenol
20
% Recovery
Methylene Chloride
-Ether
82
Ratio
Methylene Chloride-
Ether to
Methylene Chloride
4.0
N-methylaniline
Accnaphthene
n-hexadecane
4,4' Dichloro-
biphenyl
95
95
94
95
108
99
100
100
1.1
1.04
1.06
1.05
Each number is the average of two analyses.
-------
V. EXTRACTION OF SLUDGE/SLURRY SAMPLES
The June, 1976 Procedures manual (1) Included no explicit Level 1 pro-
cedure for preparation of sludge/slurry samples prior to organic analysis.
This sample category can span a tremendous range, including slurries and
solids or semisolid sludges containing up to 95% water. Some of these
materials are very difficult to handle and no one procedure will work
for all of them. Nevertheless, it was desirable to define a sample
preparation protocol that could be applied consistently and that would
minimize variability. The following protocol was tested using a variety
of complex sludge/slurry samples that had been acquired in other programs.
The basic approach is to determine whether the sample is best treated as
a solid, a liquid, or by a combination of procedures. The protocol will
involve, in most cases, tests on small portions of the 1 Kg sample to
determine the best procedure prior to committing the entire sample. The
stepwise protocol is:
• If the physical character of the sample permits, treat it as a
solid. That is, transfer the whole sample to a Soxhlet thimble
and extract for 24 hrs with methylene chloride. Do not dry-sample
before extracting. Determine wet weight of sample taken by
weighing sample container before and after transferring sample to
thimble. If an aqueous phase is noticed in the organic extract,
this should be separated and removed prior to concentration.
• If the sample state does not appear compatible with direct Soxhlet
extraction (i.e., wet sludge/slurry of high liquid content) select
a treatment procedure as follows:
• Take a 10 mL portion of the sample (shaking vigorously first,
if necessary, to facilitate a fairly representati\e sampling)
and place in a 15 mL centrifuge tube. Add 2 mL methylene
chloride. Shake well and allow to settle for at least 30 min.
Then:
28
-------
If sample dissolves completely, treat the sample like a neat
organic liquid—no sample preparation is required.
If a clean two-phase (organic-aqueous) separation is achieved,
treat the sample like an aqueous sample.
If a clean two-phase separation is not achieved (i.e., if
an emulsion forms, or if the apparent solvent recovery is
low, or if a three-phase system with solids suspended in
organic layer or between organic and aqueous layers is present),
centrifuge the mixture. If a clean two-phase system then
results, treat sample as indicated above. If not, test sample
as suggested below.
Take a 10 mL portion of the sample (shaking vigorously first,
if necessary, to facilitate a fairly representative sampling)
and place in a 15 mL centrifuge tube. Centrifuge. If phases
separate, treat the solid phase by Soxhlet extraction. The
liquid phase is treated like an aqueous sample,or, if organic,
like a neat organic liquid. The several extracts generated
for this type of sample should be recombined—taking the same
fraction of each—prior to organic analysis.
This protocol was evaluated using two fairly intractable samples—one
sludge and one slurry—acquired in the course of other programs. The
protocol was found to be generally satisfactory for those samples. It
was also found that when difficult-to-break emulsions are encountered,
the most effective and convenient strategies for breaking them are (in
decreasing order of preference):
centrifugation
addition of more organic solvent
freezing.
29
-------
VI. ANALYSES OF VOLATILE SPECIES IN ORGANIC EXTRACTS: TCO AND SOLVENT
EXCHANGE
A. Problem Definition
The original Level 1 Organic Analysis Procedure, as specified in the
June 1976 Procedures Manual (1), was designed in a manner which provided
qualitative compound class identification and quantitative (gravimetric)
data for those components of a sample that are retained when an organic
sample extract is evaporated to dryness.* It was implicitly assumed
that all compounds with foiling points i ca. 220°C and higher normal
hydrocarbon range) would be retained and carried through the LC-IR-LRMS
procedure, while those boiling below 220°C (C12 anc* lower normal hydro-
carbons down to C7*) might be lost to varying extents during sample
evaporation to dryness.
Quite early In the experimental investigation of the LC-IR-LRMS organic
analysis procedure, it became apparent that the range of material lost
when the sample extract was evaporated to dryness for the gravimetric
analysis and preparation of the LC sample was considerably higher than
expected. Data were obtained (Table 8) which suggest that most compounds
in the <216°C boiling point rcnge will be completely lost if a sample
extract is evaporated to constant weight (several hours or more of drying).
Quantitative retention of the model compounds appeared to be achievable
only for species with boiling points of about 300°C and above. At about
the same time, it was recognized that many, if not most, of the organic
compounds that were considered to be of primary concern in Environmental
Assessment [for example, the compounds 01 the multimedia Environmental
Coals (MEG) list (6)], were in the <300°C boiling point range that would
*A gas chromatographic procedure had been incorporated in the procedures
(1) to provide some quantitative information (quantities of material in
defined boiling point ranges)on the "Cg to C '' range species. Very
volatile organics, with boiling points Delow about 100°C, including
the Cg and lower hydrocarbons, were to be analyzed by on-site chroma-
tography.
JO
-------
TABLE 8
RATE OF EVAPORATION OF MODEL COMPOUNDS*
mg
% Loss/
Compound
B.P.
Taken
Time Interval
Tetrachloroethane
146°
111
52%/15m; 88%/35m
Cumene
152°
98
79%/12m; 99%/30m
n-Decane
174°
—
89%/60m
Benzaldehyde
180°
100
6%/10m; 21%/30m
Phenol
182°
105
6%/45m; ll%/80m
2-Ethyl-l-hexanol
184°
113
10%/3Oji; 23%/70m
N-Methyl aniline
.196°
104
ll%/30m; 25%/70m
n-Dodecane
216°
70
12%/60m
O-Nitrotoluene
220°
115
7%/85m
Dihexyl ether
223°
104
15%/85m
Diethyl phthalate
296°
128
0%/85m
* Sample pipetted into tared aluminum pan and allowed to stand
uncovered at room temperature.
31
-------
correspond to loss prior to LC. Thus, no compound class information
would be available from Level 1 methods as originally published, to in-
dicate the possible presence of significant amounts of these materials.
It was considered desirable to modify the Level 1 procedures to improve
the qualitative information obtainable for organics boiling below about
300°C.
Preliminary studies indicated that the dramatic loss of moderately volatile
material occurs only after the sample has been evaporated to dryness.
Concentration of an extract by a factor of 10 to 100 (1 mL or 0.1 mL
final volume from 10 mL extract) produces less drastic losses (Table 9).
Figure 6 illustrates comparable results of a second series of concentra-
tion and recovery experiments. Quantitative recovery, after concentration
of pentane extracts to 100 yL, was achievable only for test species with
boiling points over 300°C. Other data showed that recoveries similar to
those in Figure 6 could be achieved when a 10 mL aliquot of pentane solu-
tion was evaporated to dryness after addition of 200 mg of silica gel.
At one point it was suggested that the presence of high molecular weight
materials, which are frequently found in extracts from real environmental
samples, might aid in retaining volatile materials when the extract is
evaporated to dryness. A brief experimental check of this possibility
was undertaken using phenol (b.p. 182°C) as a model volatile organic
compound. "Extracts" were prepared by dissolving 20 mg of phenol and
50 or 100 mg of previously dried high molecular weight organic waste in
10 mL of methylene chloride. These solutions were washed into tared
pans and dried to constant weight. Results (Table 10) implied that
phenol recoveries were no higher in the presence of complex organic
matrices, since the weight loss in each case corresponded approximately
to the quantity of phenol present in the extract.
It was decided to modify the procedures for preparation of sample extracts
prior to LC to avoid evaporating the sample to dryness and thus allow
acquisition of some qualitative chemical information about species in
32
-------
TABLE 9
EFFECT OF CONCENTRATION OF PENTANE EXTRACTS
ON LOSS OF MODEL COMPOUNDS
% Recovery After
Concentration to:
Compound
B.P., °C
1 mL
100 11L
C7
n-heptane
96
78
45
c8
n-octane
117
86
53
C9
n-nonane
151
88
52
C10
n-decane
174
89
56
Cll
n-undecane
196
88
52
C12
n-Dodecane
216
89
53
Cumene
152
89
31
Benzaldehyde
180
86
53
10 mL samples concentrated to 1.0 mL or 0.1 mL at room
temperature under a nitrogen stream. Original concentration
*^50 ppm.
Volume restored to 10 mL and samples analyzed by GC.
33
-------
100
%
o
o
n—octadecane (C^g)
c
o
o
60
<
2:
Q>
>
O
8
QC
40
20
c
n—dodecane (C^)
n—heptane (Oy)
_L
100 140 180 220 260
Boiling Point, °C
300
340
FIGURE 6 RECOVERIES OF N-HYDROCARBONS VS. BOILING POINT
AFTER CONCENT RATION OF PENTANE SOLUTIONS
34
-------
TABLE 10
RECOVERY OF PHENOL AFTER DRYING SOLVENT EXTRACTS
CONTAINING HIGH MOLECULAR WEIGHT ORGANICS
mg
20 mg Phenol
20 mg Phenol +
50 mg Waste I
20 mg Phenol +
100 mg Waste I
20 mg Phenol
20 mg Phenol +
100 mg Waste II
Recovered**
47.1
Taken*
20
70
120 93.9
20 20
121 102
Average Loss
Standard Deviation
Lost
20.0
22.9
26.1
20.0
19.0
22 mg
3 mg
* Total mg of organic material dissolved in pentane.
Residue remaining after evaporation to constant weight (+ 0.1 mg).
35
-------
the <300°C boiling point range. It was also lecessary to specify an
alternative procedure for quantitative analysis of materials in the 100-
300 °C range.
B. Quantitative Analysis of Volatiles: TCP
A procedure based on gas chromatography with a flame ionization detector
was investigated as a method of quantitative analysis of extracted or-
ganic materials in the 100°-300°C boiling point range. The approach
taken was to select a GC column and temperature program conditions that
would: (a) adequately resolve species with 100°C boiling point from the
solvent peak and (b) allow elution of species with 300°C boiling point
in a reasonable period of time. No importance was placed on resolution
of peaks eluting within the desired range (corresponding to 100-300°C
boiling point); the objective was to obtain a single estimate of the
quantity of volatile material, analogous to the "mg" value achieved by
gravimetric quantification of non-volatile species in the extract.
A satisfactory column for this Total Chromatographable Organics, or TOO,*
was found to be 1.8 m x 3 ram O.D. (6' x 1/8") 10% OV-101 on 100/120 mesh
Supelcoport. The GC was operated isothermally at about 30"C—or room
temperature—for five minutes after sample injection and then programmed
rapidly to 250°C and held as long as necessary. Figure 7 shows a typical
chromatogram for TCO analysis of standards composed of pure normal hydro-
carbons. Figure 8 shows a chromatogram for TCO analysis of an environ-
mental sample extract. Figure 9 presents a calibration curve for TCO,
based on automatic integration of the FID signals over a retention time
window defined by the C7 and Cj7 n-hydrocarbons.
*The acronym originally proposed was GCMA, for Gas Chromatographic Mass
Analysis. This was discarded to avoid confusion with GC/MS. An alterna-
tive to TCO may be required to avoid confusion with TOC.
36
-------
i :--i-
j i '
¦I
FILL
V
1
IL
J w
*}>/
J J
3
sy
4« v %#
1U/
1 J
5v 7
U„ j
7 s ^
bs> J
7y.
i>» 7
o o u
j j •*
3 i w
] ublJ -
bi 7
1 3 ?
o y 6
3^
;¦ j
:>
i \ *
] L/J ]
lua..
1 J 1 9 7 h
11
17; j
1 t 3c
UtD
jt'26„ j
F F
- L
1 1
T u
PL
f t
21 August
1979 (1500 ng/mL std I IxlO"10 * 16 (SfiL samp)
FIGURE 7 TCO CHROMATOGRAM TOR Cg. C12 AND C16 QUANTITATIVE CALIBRATION MIXTURE
37
-------
I
fi
alu
ID
1(J f«
30 ss
5 BL
60 7P
*00 T 1
J bit T 4
TIME
A fit A
418
Io5bo2 6
472
UdbO 6
boe
17462 6
561
51S39 1
613
Icblb 1
632
5043 1
666
20*700 1
700
3eGV4 1
712
51664 1
737
47065 1
7iJ
42742 1
791
9065b 1
619
313714 1
441
2 31263 1
361
4101i 1
675
555oL 1
Sua
15)266 1
90S
24C5ofe 1
925
139336 1
937
b 16 30 1
946
? £ 9 70 1
957
13107c 1
969
1 ^ 2 1 Jtt 1
933
10tl6i t
1 j02
131720 1
1 01 5
9«tb9 1
102 9
133662 1
104#
415223 1
<060
OS.1B7 1
I07u
btSVi 1
109 9
179765 1
Tolal
3657931
#2A Inlet 1x10'10x16 5jiL 13 February, 1979, Rep.
1650
FIGURE 8 TCO CHROMATOGRAM FOR AQUEOUS SAMPLE EXTRACT
38
-------
1 1 I I 1 I I I
10 30 100 300 1000 3OC.0 10,000 30,000 100,000
TCO, Total Quantity of Material Iniected, ng
FIGURE 9 EXAMPLE CALIBRATION CURVE FOR LEVEL 1 TCO ANALYSIS
J9
-------
From the calibration curve anc! the total integrated area of the GC
trace in Figure 8, the TCO of the example environmental sample extract
is estimated to be 350 mg of volatile organics.
C. Sample Preparation for Extracts Prior to LC
The data presented in Part A. above, showed that it would be necessary
to modify the Level 1 procedures, to avoid evaporating the extract to
dryness, in order to obtain some qualitative information about organic^
in the TCO range (100-300°C boiling point). On the other hand, it was
necessary to eliminate the methylene chloride from the sample prior to
the LC step, since it would alter the desired elution pattern. The
feasibility of a solvent exchange procedure was, therefore, explored.
Table 11 presents some data for recovery of a number of model compounds
that were subjected to a two-stage solvent exchange procedure. A 100 mL
portion of methylene chloride solution containing the indicated com-
pounds at known (ppm) concentrations was concentrated to 10 mL. This
concentrated solution was added to 200 mg of silica gel in a graduated
container, and the volume was reduced to 1 mL, One mL of the exchange
solvent (hexane or cyclopentane) was added, and the mixture concentrated
to 1 mL again. A second 1 r.L portion of exchange solvent was acded, and
the mixture reconcentrated to 1 mL. Recoveries of the model compounds
were determined by gas chromatographic analysis. Cyclopentane gave
measurably higher recoveries, than hexane and was, therefore, the
exchange solvent of choice.
After a two-step solvent exchange, an estimated 14-20% of methylene
chloride remained. One volatile and one relatively non-volatile
aromatic hydrocarbon were selected for use as indicators of the effect
that traces of methylene chloride have on the LC separation. rhe
compounds chosen were curaene and ncenaphthene.
40
-------
TAHIh 11
KLCOVLKIkS OK HODKL JUHI'OIINOS AKIhK MIIVKNl tXHIANUt Willi IIUCANK AND CYCIAII'UfrANE
N
I Kc< iiyt ry
Cftmpiiiiikl B.I*., 'r Ileum1 CyiItipcutaim
u-llyjroi arliuna
n-llcplanc 96 5,5« 13.11*
ii-(k ijho 126 25,32 47 44
ii-N»miuc 151 42, 54 71,61
ii-l>eiuuc 174 55,64 85,77
n-UiiJvi'jiio 195 65,72 OB,HI
n-Ti IJctaiio 243 82,89 9J,94
n-lVnljdccaue 2/0 91,91 95,97
n-lliiilaJecune 301 91,88 92,92
Otln.ru
Tell acliloruulhaue 146 48 95,100
152 55,4} 70,80
Ili-Iizal JUiydc 180 h0 70,80
Plienol 182 58, JO
2-tlliyl llcx.imil ld5 74 78,80
N-HcLliylaiil line 196 J5.55
u-Nllrululucnu 220 75 85,85
IMIiexyUtlier 223 88 100,100
Qiilmillnu 237 65 54,56
Ai un.i|.litlieim 278 88,95 95,90
|> .|>'-I>Klilurobi|.lienyl J|6 9U,IOi 101,100
* Iflurr iwii nunJitTB are fiven, tlu-y represent result)) of replicate experiments.
-------
CCuroene
'Acenaphthene
^|CH(CH,)2
One mL samples of these compounds in cyclopentane solution containing
know amounts of methylene chloride were put through the LC scheme.
The parameter of primary interest in these experiments was the percentage
of the aromatic compounds that was found in fraction _1 (which is supposed
to contain only aliphatic hydrocarbons) relative to the total amount re-
covered in fractions 1-4. The results in Table 12 indicate that the per-
cent breakthrough of aroraatics into fraction 1 is roughly equal to the
percentage of methylene chloride in the sample applied to the column.
It seems that 10% breakthrough represents the upper limit of acceptability.
It was decided to aim for a 5% or lower level of methylene chloride re-
maining in the sample after completion of the solvent exchange.
Subsequent experiments showed that a three step solvent exchange with
cyclopentane gave 4-6% residues of methylene cloride without markedly in-
creasing the losses of volatile materials. The three step exchange was
done in two ways:
A. The methylene chloride solution was evaporated to b mL, then
5 mL of cyclopentane added. The mixture was then concentrated
to 1 mL and 2 successive exchanges, each with 1 mL of cyclopentane,
were carried out. This procedure does not increase the number of
times the sample is reduced to 1 mL total volume.
B. The methylene chloride solution was evaporated to 1 mL. Three
successive exchanges, each with 1 mL of cyclopentane, were carried
out.
42
-------
TABLE 12
BREAKTHROUGH OF AROMATICS INTO LC FRACTION 1
, % Recovery in Fraction 1
% Methylene Chloride z
in Cyclopentane Curoene Acenapthene Average
20 17 11 19
15 11 18 15
10 3 14 9
5 3 8 6
hi
-------
The recoveries are shown in Table 13, along with data obtained for a two
step cyclopentane exchange done at the same time. The Table 13 data imply
that recoveries from a three step exchange are not significantly lower
than those from a two step exchange with the same solvent. (No plausible
explanation, other than operator variability, has been found for the fact
that the two step exchange experiment of Table 13 gave generally lower
recoveries than the analogous experiment of Table 11.)
The three step exchange was therefore recommended for Level 1 LC of all
sample extracts, except those known or shown to contain no TCO range
material. In the latter case, evaporation to dryness prior to LC was
recommended.
44
-------
TABLE 13
RECOVERIES OF MODEL COMPOUNDS AFTER SOLVENT EXCHANGE
% Recovery
Compound
2-Step
Exchange
3-Step ExchanRi
a* b*
Tetrachloroethane
58
52
45
Cumene
60
54
43
Benzaldehyde
52
48
48
2-Ethyl Hexanol
71
69
75
o-Nitrotoluene
56
56
49
Quinoline
68
58
68
Di-hexylether
83
83
70
*See text for full explanation. Procedure ji involves a total of 3
reductions to a final volume of 1 mL, while involves a total of
4 reductions.
45
-------
VII. ELUTION PATTERNS IN LEVEL 1 LC
A. Model Compounds
Several sets of mixtures of pure compounds were prepared to evaluate the
elutlon patterns of the Level 1 LC separation (1). The mixtures were
generally applied to the columns at levels corresponding to <25 mg of
each compound to avoid overloading the column and therefore distorting
the separation.
Table 14 shows the results of the first such experiment. The data il-
lustrate a phenomenon that has been borne out in subsequent work: it
Is uncommon to find any given material Isolated in one LC fraction.
Several years of cumulative experience with the Level 1 LC separation
have led to the inference that the band-broadening in this low resolution
chromatographic method leads to elution peaks that are about one LC
fraction wide.
Table 15 shows the results of LC elution pattern studies on a total of
17 model compounds. Based on these data and on observations of elution
patterns in real samples subjected to Level 1 separation, Table 16 was
prepared to indicate the most probably LC fractions for various compound
categories. It should be emphasized that, because of band spreading
and the low resolution separation, each of the indicated assignments
should be regarded as uncertain to within 1 LC fraction.
Even the most polar compounds (e.g., p-toluene sulfonic acid) tested in
the LC elution experiments were found in Fractions 6 and 7, and not in
Fraction 8. Furthermore, the early pilot tests of the Level 1 methodology
sponsored by Process Measurements Branch found no organic materials in
Fraction 8. For these reasons, it was determined by PMB that LC 8 should
be dropped and the LC separation be limited to 7 fractions.
46
-------
TABLE 14
DISTRIBUTION OF MODEL COMPOUNDS AMONG LC FRACTIONS
Column spiked with 5 mL of solution containing four
model compounds:
Cumene 4.7 mg/mL
Benzaldehyde 8.7 mg/mL
2-Ethyl-l-Hexanol 4.4 mg/mL
N-Methyl aniline 3.4 mg/mL
Fract.
No. Solvent
1 Pentane
2 20% CH2C1Z in Pentane
3 50% CH2Cl2 in Pentane
4 CH2C12
5 5% CH30H in CH2C12
6 20% CH3OH in CH2C12
7 50% CH3OH in CH2C12
Distribution in Column Effluent
N-Methyl 2-Ethyl
Cumene Benzaldehyde Aniline Hexanol
82%
17% 22%
75%
3% 3%
94% 99%
2% 0.7%
-------
TABLE 15
% DISTRIBUTION IN LC FRACTIONS
¦e-
oo
Compound 1 2 3 4 5 6 7 8 Ref.
Hexadecane 85 15 18
Cumene 82 17 18
Dlchlorobiphenyl 25 69 5 18
Acenaphthene 69 31 18
Tetrachloroethane 81 19 18
o-Nitrotoluene 30 70 18
Dibenzofuran MOO 13
Benzaldehyde 22 75 3 18
Dihexyl ether 18 77 4 18
N-raethyl aniline 3 94 2 18
Quinoline 100 18
Diethyl Phthalate 100 18
2-Ethyl hexanol 99 0.7 18
Phenol 100 18
2-Hexanone ^100 13
Acetophenone ^100 13
27 68 13
p-Toluene sulfonic
acid
-------
JAtil.K 16
CAYtUiKltS MIR KllOKflmJ UK LKI1S DATA
Category
_ mwui?ryj _ _
Aliplitit 1< Ityilroc.iiboittf
(Alkjju s)
(Al kuiLb)
(Alkyneti)
Must 1'rohiMti
IC Kr.ictloi*
1
1
1
1
ll.ilogiiialcd Al IpluI les 1,2
(S • Lilt tiled) 1 ,2
(Hiisjcni juJ) 1,2
Ar ohm lie Hydrocarbons 2,1
(ItuJlll'IU h) 2, J
ltdogcajIcd jrooutlc hydrocarbons 2,3
Nltru aruuvjilc hydrocarbons 4,5
Itihcil alternate, non-u Ite mate hydrocarbons 2,3
Htf < ^16 (iiil.tiiy] pyruuc) 2,J
MW > 216 2,3
blliera A
(lid logcnutcJ Ktlicra) 4
Ipoxideu 4
Aldthydeu and Ketones 4
Heterocyclic Oxygen Compounds 3,4
Ntlrl lea 4
(Al Iplutlc) 4
(Arotujl 1<) 4
Alcohol:* 6
(rrlimry, Secondary, Tertiary) 6
((ilycula) 6
M'osulble utisLgiiaicnts. Fraction 4-5, i-6, 6*7
gmcratly overlap to a conblderablc extcot.
Category
(Subcategory) _
Phenol*
(Alkyl, etc.)
(llu to gun«i ted Plieno I s )
(NI Lrophenolti)
hat era
(HlllldljJtCli)
Am'ueb
(I'llmary, Secondary, Tertiary)
(Hydrazines. nzo coupouud»)
(Nlirottoamlnes)
Heterocyclic Nitrogen Compounds
(1 ndo leu, Curbazoles)
(tytluolIocs, AcrJdfncs)
AlkyJ Sulfur Compounds
(Mcrc«iptani»)
(Sulfide*, disulfide*)
Heterocyclic Sulfur Compounds
(llciuothloplienc*)
Sulfonic Acids, Sulfoxide*
Aail dc»S
Cdiboxylic Ac Ida
bl 11 conet*
11to:>phdtes
Ho ft ProlMlili
I.C Krji_ 11 oij
6
6
6
6
6
6
4
7
0
6.7
2,1.4
5.6.7
-------
B. Sodium Sulfate Drying of Extracts for LC
There was concern that the presence of water in Level 1 organic extracts
was leading to irreproducible deactivation of the silica gel and, there-
fore, to irreproducible results of the LC separation. Experimental re-
sults indicated that sodium sulfate may be used as a drying agent for
methylene chloride extracts without causing unacceptable losses of
sample components.
A sample solution of model organic compounds was diluted to approximately
1000 ppm total concentration with methylene chloride. This was used
as the stock solution. Sodium sulfate (Fisher certified ACS grade) was
sequentially soxhlet extracted for 24 hours each with methanol,
methylene chloride an.; pentane. After drying at 110°C for several days
and cooling in a dessicator, the sodium sulfate was dry-packed into a
column 2.5 cm in diameter (with fritted glass disk) to a depth of
*v3.0 cm. This required ^12 gns of sodium sulfate. Aliquots of the
stock solution were passed through the colutiin. The eluent was collected
and analyzed. Samples were analyzed by gas chromatography using the
Level 1 TCO procedure.
Table 17 presents data for recovery of model compounds from dry methy-
lene chloride extracts. Table 18 presents comparable data for
recovery of model compounds from methylene chloride solutions that had
been saturated by shaking with water. In each experiment, the total
TCO value was un"hanged, within 10%, by passing the sample through the
sodium sulfate column.
The Level 1 LC column preparation procedure was, therefore, modified
by adding a short bed of sodium sulfate at the top of the silica gel
to dry the sample immediately prior to the separation.
50
-------
TABLE 17
RECOVERY OF TCO AND OF SEVERAL MODEL COMPOUNDS* FROM
SODIUM SULFATE COLUMNS: DRY SOLUTIONS
% Recovery
Experiment 1
leak 1 109
Peak 2 106
Peak 3 98
Peak 4 101
Total TCO 103
Experiment 2
Peak
1
89
Peak
2
97
Peak
3
101
Peak
4
102
Total TCO 97
*Model Compounds were identified only by retention times in these
Experiments. Both polar and non-polar species were included.
51
-------
TABLE 18
RECOVERY OF TCO AND OF SEVERAL MODEL COMPOUNDS FROM
SODIUM SULFATE COLUMNS: WET SOLUTIONS
% Recovery
Experiment 1
Peak 1 122
Peak 2 130
Peak 3 101
Peak 4 99
Peak 5 99
Peak 6 106
Peak 7 106
Peak 8 119
Peak 9 106
Total TCO 110
Experiment 2
Peak 1 92
Peak 2 91
Peak 3 84
Peak 4 95
Peak 5 94
Peak 6 95
Peak 7 95
Peak 8 97
Peak 9 108
Total TCO 96
52
-------
C. LC Column Blank Due to Silica Gel
It was suspected that the most polar solvents used In the LC scheme might
dissolve some of the silica gel from the column bed, giving rise to
spurious high values in the gravimetric analysis. The following data
were obtained for blank columns with no sample added:
Fraction
Run No. No. 7
1 1.0 mg
2 1.2
3 0.8
4 0.7
mean and standard
deviation
0.9 ± 0.2 mg
No weighable material was found in a»iy of fractions 1 through 6.
In an attempt to determine whether the silica gel could be separated from
any sample components of fraction 7, the residues in the dishes were
washed with hexane and the solvent was decanted. The washing was re-
peated with methylene chloride. The data below verify that only small
quantities of silica gel are suspended in those solvents in this procedure.
Fraction
No. 7
original wt, mg 0.9 ± 0.2
after hexane, mg 0.8 ± 0.2
after CH2CI2, mg 0.7 ± 0.2
It appeared to be necessary to correct the apparent weights of fraction
7 arithmetically, by subtracting the appropriate blank value, rather
than by actually separating the sample from the silica gel.
53
-------
VIII. RUCGLDNESS TESTING OF LEVEL 1 LC PROCEDURE
It was desirable to determine the change in results oL the LC separation
that might occur if a slight deviation in procedure vas made. This
margin of error within which an experiment can be varied and the results
remain unchanged is commonly referred to as the ruggeuness of the experi-
ment. If the procedure could be done slightly differently each time "nd
the same results obtained, chance are increased that intralaboratory or
even interlaboratory results will be consistent and comparable.
The Level 1 LC procedure was studied for "ruggedness" in three
different aspects.
• continuity of elution
• exact composition of eluants
• activity of silica gel
In all cases a solution of model compounds with a convenient elution
pattern was employed to test the procedures. The fractions were analyzed
by GC using the Level 1 TCO program and compared to controls.
To determine whether the continuity of elution affected the separation
pattern of the LC, a column was run with a 0.5 hr break after fraction A
and before fraction 5. The results show no significant change in the
elution pattern in comparison to the LC procedure with no break outside
the error associated with the LC procedure. (Figure 10.)
The stability of the LC separation relative to exact eluant composition
is of interest for storing the solvents. It is convenient to be able to
keep the solvent mixtures for several weeks and thus reduce preparation
time for the LC procedure. T'lis parameter was monitored both from the
view of solvent composition change with time and LC separation change
with time. The solvent mixtures were stored in amber bottles and tightly
capped.
54
-------
TEST LC LC INTERRUPTED FOR 05 HOURS BETWEEN FRACTIONS 4 AND 5
Model Compound
Fraction 1
Fraction 2
Fraction 3
Fraction 4
Fraction S
Fraction 6
Fraction 7
Hexadecane
Teuachloro-
ethane
Dtberuofuran
Benzylcyamde
Acetophenone
p-Tofuenesul*
fomc Acid
1
<5
nuiiih i
vmitummrth
¦ ¦ Control LC
W11U\ Test LC
FIGURE 10 DISTRIBUTION OF MODEL COMPOUNDS IN THE LC SCHEME
WITH MODERATE CHANGE IN PROCEDURE
-------
The composition of the LC solvent mixtures changed very little over the
testing period of four weeks. The largest variation observed was an in-
crease in methanol percentage in fractions 6 and 7 of approximately 8%,
as shown in Table 19.
The LC separation, as expected, also did not change significantly over
the four week period. The results are shown in Figure 11.
The state of activity of the silica gel may affect the elution sequence.
The effects of moderate variations in the duration and temperature of
thermal activation and in the time the silica gel is stored after activa-
tion were tested.
• The silica gel was activated for 24 hours at 110°C, instead
of the prescribed two hours, cooled in a dessicator and used
limnedlately. The I.C separation showed insigificant difference
when compared to a separation done on silica gel activated for
the usual two hours (Figure 12).
• The silica gel was activated for two hours at 100°C, but cooled
for 16 hours in a dessicator, rather than just until cool enough
to use. The pattern of LC elution was unchanged from the control
LC with silica gel cooled ior the mirimum amount of time (Figure
13).
• Silica gel activated for two hours at 100°C was cooled in the
ambient air rather than a dessicator and compared with silica
gel cooled in a dessicator. The LC separation differed by an
insignificant amount between control and air cooled silica gel
(Figure 14).
• The activation temperature of the silica gel was tested at 90°C
and 200°C. The separation at each temperature correlated very
closely with the control activated at 100°C (Figures 15 and 16).
56
-------
TABLE 19
CHA.VGE r» IC SOLVENT COXtOSITZOU VITH TIME
"SACTtOH N'O.
SOLVENT
rreah
Methanol 5.85 15.IT. 45.9*
Mechvletie Chloride 19.St 59.61 100S 94.2X 84.9S 54.IS
?entin« 100: SO.2: 40.4S
Two tfteks Old
Methanol 7.IS 20.9X 44.9;
Mec'.ivler.a Chloride 19.3? 54.3Z 1001 9:.9Z 79.IX 55.11
Pentane 100* 80.?* 45.2?
Four Weeks Old
Methanol 6.3: 22.7: 52.3:
Methylene Chloride 20.9? 52.35 100X 93.2X 77.2? 47.7:
Pentane 100: 79.11 47.7*.
57
-------
TEST LC:
LC SOLVENT 4 WEEKS OLD
Model Compound
Fraction 1
Fraction 2
Fraction 3
Fraction 4
Fraction 5
Fraction 6
Fraction 7
Hexadecane
Tetrachloro-
ethane
Dibenzofuran
Benzylcyamde
Acetophenone
p-Toluenesul-
fonic Acid
jrTTTTrrmr-rr
=3
nbr
q
I | Control LC
V))l)>)\ Test LC
FIGURE 11 DISTRIBUTION OF MODEL COMPOUNDS IN THE LC SCHEME
WITH MODERATE CHANGE IN PROCEDURE
-------
TEST LC SILICA GEL ACTIVATED FOR 24 HOURS
Med el Compound
Fraction 1
Fraction 2
Fraction 3
Fraction 4
Fraction 5
Fraction 6
Fraction 6
Hexadecane
Tetrochloro-
ethone
Dibenzofuran
Benzylcyanide
Acetophenone
p Toluenesul-
fonic Acid
ilrnnhm/ll/mrri
3
\»ml*
cd
vtmrntmm
I | Control LC
EZZZZZZ] Test LC
FIGURE 12 DISTRIBU'. ION OF MODEL COMPOUNDS IN THE LC SCHEME
WITH MODERATE CHANGE IN PROCEDURE
-------
TESTLC. SILICA GEL COOLED OVERNIGHT
Model Compound
Fraction 1
Fraction 2
Fraction 3
Fraction 4
Friction 5
Fraction 6
Fraction 7
Hexadecane
Tetrachloro-
ethane
Dtbenzofuran
Benzylcyamde
Acetophenune
p-Toluenesul-
fonic Acid
,,,,,,,,,
vhi/im
ii£i
uumm 1
J
viniamu
rrjrKm
£
1
I I Control LC
ezzzzza lc
FIGURE 13 DISTRIBUTION OF MODEL COMPOUNDS IN THE LC SCHEME
WITH MODERATE CHANGE IN THE PROCEDURE
-------
TEST LC SILICA GEL COOLEO IN OPEN AIR
Model Compound
Fraction 1
Fraction 2
Fraction 3
Fraction 4
Fraction 5
Fraction 6
Fraction 7
Hexadecane
Tetrachloro-
ethane
Dtbenzoluran
Benzylcyamdo
Acetophenone
p-Toluenesul-
torvic Acid
•frrrrtmrrrn rrrA
thinnunu
xrhm rrrrrrrr
C
I I Control LC
I hi u >i Tejl LC
FIGURE 14 DISTRIBUTION OF MODEL COMPOUNDS IN THE LC SCHEME
WITH MODERATE CHANGE IN PROCEDURE
-------
TEST LC: LC SILICA GEL ACTIVATED FOB 2 HOURS AT 90°C
Model Compound
Fraction 1
Fraction 2
Fraction 3
Fraction 4
Fraction 5
Fraction 6
Fraction 7
Mexadecane
Telrachloro-
ethane
Dtberuofuran
Beniylcyamde
Acctophenone
p-Toluenesul-
fonic Acid
C
yuatuah
uI.mhiu J
rirrrirrrrrrrht
T7h
*5
[fazmnwii
baasnSj
I I Control LC
umpn Test LC
FIGURE 15 DISTRIBUTION OF MODEL COMPOUNDS IN THE LC SCHEME WITH MODERATE CHANGE IN PROCEDURE
-------
TEST LC: SILICA GEL ACTIVATED FOR 2 HOURS AT 200°C
Model Compound
Fraction 1
Fraction 2
Fraction 3
Fraction 4
Fraction 5
Fraction 6
Fraction 7
Hexadecane
Tetrachloro-
ethane
Dibenzofuran
Benzylcyarude
Acetophenone
p Tofuenesul-
fomc Acid
mArrm
\mimnnhn
I I Control LC
V/JIHHX Test LC
FIGURE 16 DISTRIBUTION OF MODEL COMPOUNDS IN THE LC SCHEME
WITH MODERATE CHANGE IN PROCEDURE
-------
None of the experiments done to test different activation procedures of
the silica gel showed a significant difference in LC separation from
control samples run concurrently.
To evaluate the potential effects of variations in LC column temperature,
five LC columns were run using the same solution of model compounds and
with all conditions identical except for jacket temperature. Four
jacketed columns were held at 14.5, 18, 20 and 22°C. The fifth column
was run using acetone-cooling applied as described in the early versions
of the LC method.
No significant changes were observed to result from these temperature
variations. Figure 17 shows the elution pattern observed for the highest
and lowest water jacket temperatures and for the acetone cooling. It is
clear that over this temperature range, the elution behavior of each
model compound was not altered by as much as on
-------
Model Compound
Fraction 1
Fraction 2
Reaction 3
Fraction 4
Fraction 5
FrBctioi 6
Fraction 7
Hexadecane
Teuachloro-
ethane
Oibenzofuran
Benzylcyanide
Aceiophenone
p-Toluenesul-
fonic Acid
-
—
mam
WAWilril
I I Local Cooling Using Acetone
\ll»l»H>rm Column Then-iostatted at 22°C
mmn.u-"- iiiMt Column Thermostatted at 14 5°C
FIGURE 17
-------
TEST LC. SAMPLE NOT SOLVENT EXCHANGED FROM METHYLENE CHLORIDE INTO CYCLOPENTANE
Model Compound
Fraction 1
Fraction 2
Fraction 3
Fraction 4
Fraction 5
Fraction 6
Fraction 7
Hexadccane
Telrachloro-
ethane
Dibenzofuran
Benzylcyanide
Acetophenone
vnmnm //;
uir/irt
7773
umui
3
s
-------
IX. REPORT FORMATS FOR LEVEL 1 OrjSANIC ANALYSIS RESULTS
A. Introduction
A particularly important aspect of the evaluation and evolution of the
Level 1 organic analysis procedures was the development of systematic
reporting formats that summarize and integrate the data in a useful
way. The report formats developed in this study have been incorporated
into the Second Edition of the Level 1 Procedures Manual. (A) They
are described briefly in this chapter and examples are given in
Chapters X and XI.
Table 20 lists the major types of samples that are encountered in
Level 1 organic analysis and indicates the data that are to be generated
for each. In interpreting and integrating the Level 1 organic analysis
data and reporting the results, a list of organic compound categories,
based on the Multimedia Environmental Goals (MEG) (6) categorization
scheme (slightly modified), is used to provide a means of organizing
the Level 1 data. It should be emphasized that it is the list of MEG
categories-, and not the list of specific MEG compounds, that is referred
to in the context of Level 1 organic analysis. With the addition of
a very few non-MEG compound classes, this list of categories represents
almost all of the organic chemistry likely to be encountered in most
sources.
In developing this report format for Level 1 organic analysis results,
the working assumption was that the users of Level 1 EA d^ta will be
interested in comparing estimated mass loadings in various streams
with decision criteria expressed as concentrations. This is the type
of approach, for example, that is incorporated in the SAM models (7)
being developed for IERL, using DMEG (6) values as decision criteria.
The format, however, is flexible enough for use in other models and
with other decision criteria.
67
-------
.YBLE 20
SUMMARY OF EXPECTED DATA ROM LEVEL 1 ORGANIC ANALYSIS
On-Site
SAMPLE GC
TCO
GRAV
_IR
LC*
Gases - Grab Sample /
SASS
>10 vjm particulate
/
/
/
3-10 pm particulate
/
/
~
1-3 pm particulate
/
/
/
<1 vm particulate
~
/
/
Rinse of particulate
modules
/
/
/
XAD-2 Resin combined with
rinse of sorbent module
/
/
/
/
Sorbent Module Condensate
/
/
/
/
SOLIDS
Flyash; Clinker
/
/
/
Organic Feed Stock
/
/
. /
/
Coal
J
/
/
/
LIQUIDS
Effluent Water
~
/
/
~
Organic Feed Stock
/
~
/
/
Fuels
/
~
/
/
Includes GRAV + IR on all fractions; TCO
quantity is sufficient.
and
LRMS when
sample
68
-------
A requirement of the approach is that all of the data reduction steps
described in this document will be performed by personnel who have
access to the original Level 1 organic analysis data and/or direct
communication with the original analyst(s). The reason for this is
that the organic analysis data outputs (i.e., spectra) contain a
great deal of information (i.e., presence or absence of particular
peaks) that is not easily reduced to tables or other forms intelligible
to nonchemists, but is very valuable in making an assessment of the
Level 1 sampling and analysis results. It is recognized, however, that
guidance as to the appropriate models and decision criteria (concentrations)
for comparison with the Level 1 data will usually be provided by others.
For purposes of illustration (only), a SAM-1A type model has been used
along with DMEG values as decision criteria in this report.
B. Results of the LC Fractionation
The results of the LC fractionation procedure include quantitative
ft
estimates of TCO and GRAV range mateiials in each of seven fractions.
In most cases, the quantity of material actually taken for the LC
separation is only a portion of the total sample, and the amount taken
should be stated in the report. The actual, measured TCO and GRAV
values for the LC fractions should be multiplied by the appropriate
factor (total sample quantity * quantity taken for LC) to give the
corresponding total sample values. It is then useful to convert these
quantitative estimates into equivalent concentrations at the source,
in order to facilitate comparisons with various decision criteria.
Table 21 illustrates the format for reporting LC fractionation data.
GRAV analyses involve weighing to the nearest 0.1 mg. TCO vilues
and all Level 1 concentration data are reported to 1 or 2 significant
figures.
On sorbent module extracts and other extracts or organic liquids
found to contain more than 2 mg of TCO prior to LC fractionation.
TCO analyses are not done on SASS particulate, ash, or clinker samples.
69
-------
i AM I .'I
itvti. I Ijc hi-l.uits RM'imr mm
SAMPIk
TCO
m*
GHAV
rng
TCO < GRAV
Totaling
CoAccotralwa
mvjj tin*. L.or tt|)
Total ttfiH*1
Tali m lor LC2
fllCIMA
TCO U mo
GftAV m mg
TCO •
GIIAV
Total tug
Conccftliatwn
mg/
Found In
FlKtWfl
BUnk
Cur
i«cud
T»U*
found m
ft act tow
QUnk
Cot
ractad
lull)4
1
—
2
)
—
4
—
5
ft
7
Sum
—
—
I Quantity tntmiratantpla determuwd before LC
1 Poitn>ri ol whol« innpti uwl for tC, actual mg
3 Quantity iccoveicd fiom L'J culumii, actual n»y
4 Total mg coni|mt*d back to total Minolta
-------
The Level 1 LC separation schene is based on differences in polarity
among various organic compound classes. Table 22 lists the compound
classes that can be expected to be found in the LC fractions. The LC
separation is not a high resolution technique and is not sufficiently
precise to serve as a reliable qualitative identification tool.
Considerable blurring between fractions is inevitable for complex
samples. However, the LC separation scheme can and should be used
to gu^de the interpretation of thu LRMS and IR spectra. For example,
an apparent "aliphatic hydrocarbon" moiety in LC6 could be an aliphatic
acid, alcohol, amine, etc., but not a paraffin.
C. Results of the IR Spectroscopy
The total sample extracc, or neat organic liquid, and the seven LC
fractions are analyzed by IR spectroscopy. Spectra are interpreted in
terms of the functional group types present in the major components of
the sample or LC fraction. The many reference texts in this area are
of considerable help in interpreting the IR spectra. The interpretation
of the spectrum should also be guided by consideration of the LC
fractionation scheme (Tables 3 and 22) and the LRMS results (when available).
Components amounting to <—10% of the total sample will not contribute
significantly to the IR spectrum and cannot be detected by this
technique.
The report format for the results of the IR analysis is shown in
Table 23. Guidelines for reporting the data are:
1. Major peaks and assignments:
• The frequency reported should be the peak maximum, or a
range may be reported instead for broad peaks with no
well-defined maximum.
n
-------
TABLE 22
DISTRIBUTION OF ORGANIC COMPOUND CLASSES AMONG LC FRACTIONS
Organic Classes MEG List
LC Fraction Predominating Category
1 Aliphatic Hydrocarbons 1
Halogenated Aliphatics 2
2,3,4 Aromatic Hydrocarbons 15,21,22
Halogenated Aromatics 16
4.5 Heterocyclic N,0,S Compounds 23,24,25
Sulfides, Disulfides * 13B
Nitriles * 9
Ethers 3,4
5.6 Aldehydes. Ketones * 7
Nitroaromatics 17
Alcohols 5,6
Amines 10,11,12
6.7 Alcohrils 5,6
Phenols, Halo and Nitrophenols 18,19,20
Esters, Amides * 8C,3D
Amines 10,11,12
Mercaptans * 13A
Carboxylic Acids 8A,8B
Sulfoxides 14B
These compound classes have not yet been tested in the LC scheme,
so these assignments should be regarded as more tentative than
the others.
72
-------
TABLE 23
LZVEL I I* &£SLT.7S REPORT ? CK*
IR REPORT
SAMPLE __
Ww NumMf liwtmity j Anignmtnt Cdmmtms
(cm" M ' i
!
r
L
1
73
-------
• The intensity is reported relative to the strongest peak
in the spectrum on a transmittance basis,
s = strong - 50-100% of strongest peak
m = medium - 20-50% of strongest peak
w =» weak - 5-20% of strongest peak
When a peak is of borderline intensity, it should be
labeled "m." Finer intensity ratings such as m-s or w-m
are not appropriate.
o The assignment/comments column indicates the functional
group(s) to which the peak is attributed. It may also
contain single word descriptors of peak shape such as:
"broad," "doublet," "shoulder."
2. Unassigned weak bands:
This data category is merely a list of frequencies, in cm of weak
bands the analyst considers possibly significant but cannot assign to
a particular functional group. Any medium or strong peak, even if
unassigned, should be reported in Section 1 of the IR report.
3. Other remarks
This reporting point will not always be required. An example of the
type of other remarks to be included is: "The IR does not confirm
LRMS indication of quinones: no 1690-1660 cm ^ bands."
It is not necessary to report every peak in the IR spectrum. It is
expected that £10 "major peaks and assignments" entries, and <6
"unassigned weak bands," will be typical.
74
-------
D. Results of Low Resolution Mass Spectroscopy
Low resolution mass spectra (LRMS) are obtained on each LC fraction
which has sufficient quantity, when referenced back to the source. All
samples that meet the decision criteria for quantity (TCO plus GRAV)
will require analysis via the direct insertion probe. Samples with
significant quantities of TCO range material should also be analyzed
by insertion in the batch inlet. The mass spectroscopist is encouraged
to integrate the interpretation of the mass spectra obtained on a
particular sample to provide one report describing sample chemistry.
Interpretation of the spectra is guided by knowledge of the LC separation
scheme, the IR spectra, and other information about the source. In
reporting the results of the LRMS analysis, the basic philosophy is
to present increasingly more specific data, as the complexity (or
simplicity) of the spectra will allow. The first level of reporting
is to identify compound classes. If possible, or appropriate, one
should then attempt to identify the subcategory compound classes
present in the fraction. Finally, specific compounds should be
identified if possible to do so from the spectra. Whore possible,
the molecular weight range and composition of each of the categories
should be reported. The relative abundance of each category should
be estimated with a rating of 100 = major, 10 = minor, 1 = trace.
It should be possibe, using this methodology, to account for nearly all
observed species by selection from a relatively small list of compound
categories and subcategories. A list, of such categories has been
assembled in Table 2\. The primary reference for selecting these
caregories was the MEG list (6), which seems to do a good job of
representing all probable major compound classes. Some few categories
were not in the MEG list and have beep added here.
75
-------
TABLE ;i
CATEGORIES FOR REPORTING OF LK1S DATA
Category MEC List
(Subcategory' Category 'to.
Aliphatic Pvdroeirbont *
( A1'nines)
(Alkene s)
(AlVynei)
Halogeiaced Aliphatlcs 2
(Saturated)
( Unsaturated)
Arcnaclc Hvdrocirbons
Benzenes 15
Halogenatcd aroaatlc hydrocarbons 16
Nitro aronatic Hydrocarbons 17
Fused alternate, non-alternate
nydrocarsons
M'«" < 21b {aethvl pvreae) 21,22
MW » 216 21,22
Ethers
(Kalogenated Ethers) 3,6
Epoxides 6
Alde'nydes and Ketones 7
Heterocyclic Oxygen Compounds <5
Nicriles 9
( Aliphatic)
( Aronatic)
Alcohols 5,6
(Priaary, Seeondarv, Terlary) (j)
( Clvcols) (6)
Phenols 18
(Alitvl, etc.)
(Halogenated Phenols) l1*
(Si "onhaso Ls) 20
Esters gc
Phchalates
(Priraary, Secondar", Terri \r.) (10)
(hvdnzlnes, a:o cospounds) (11)
(Mtrasoanlnes ) (12)
Probable
LC Fraction
1
1
1
1
1.2
1.2
1.2
2.3
2.3
1,3
4.5
2.3
2.3
2.3
4
4
&
&
4
4. ?
4. »
4.?
6
6
6
6
6
6
6
6
6
6
6
&
centlnued....
76
-------
TABLE 24 (Continued)
Category »(EC List
(Subcase zo r *.) Category Nc. -C Frac t Ion
Heterocyclic Nitrogen Coapounda 23
(Indoles, Carbasoles) 4
(Qulnollnes, AcrlJines) 6
Alky! Sulfur Compounds 13 6
( Mercap tans) 6
(Sulfides, disulfides) 6
Heterocyclic Sulfur Compounds
( 3enzothlopnenes) 24 4
Sulfonic Acids, Sulfoxides 14 '
Aaides 3C 6
Carboxylic Acids 3 6,7
Silicones ?
Phosphates ?
Pesticides ?
Oyestuffs ,
Possible assignaents. Fraction 4-5, 5-6, 6-7, generally
overlap to a considerable extent.
77
-------
The list in Table 24 is also organized somewhat differently than the
MEG list to be more compatible with the nature of the mass spectrometry
data. It should be strongly emphasized that the list probably does
not include all compound categories that it will be possible for one
to identify. If interpretation of the spectra yields the identification
of a category not Included in this table, the category should be
reported. At the same time, EPA should be notified of the need to add
that category to the list. It will be possible in most cases to
identify the spectra In terms of the compound categories listed in
Table 24, but one should avoid "force fits" to the categories if another
one seems more appropriate.
Interpretation of the mass spectra data should take full advantage of
all other information known about the same source, LC fraction and IR
spectra. Since the LC separation does a reasonable job of dividing
compound classes, the categories listed in Table 24 have been listed
in order of their possible elution from the LC column. Where possible
some indication has been made as to the LC fraction in which the
category might elute. These fraction assignments are known to be
correct in some cases and are only estimates in others. Sometimes the
sample characteristics will have minor effects on the -fraction elution
behavior. Again the LC fraction indications should only be taken as a
guideline.
Table 25 shows the report form for the LRMS data.
It is once again emphasized that interpretation of the LRMS data Is
best done using all available information such as what one knows about
the chemistry of the source being sampled, what species generally
elute in the LC fraction being examined and functional group data
derived from the IR spectra.
78
-------
LEVEL 1 LJiitS RESULTS nE?OB.T ?GrU
CAMS REPORT
SAMPLE.
MilOf Catt^oriit
Inttnoty | CtMfory { MW R«r>9« I
i • ,
' J
1 I
t 1 i
SpkiOc Compound*
InttratY J Cctrvory | m/« | Composition
I
T
! I
t »
I
i
1
r
79
-------
E. Organic Analysis Summary Tables
At the end of the Level 1 organic analysis procedure, there will be an
LC report, 8 IR reports, and up to 7 LRMS reports for each organic
extract or neat organic sample. This is an unwieldy body of data from
which to make a decision. The first step in reducing these data to a
workable form is to prepare a single table that summarizes the organic
analysis results for each extract. Table 26 illustrates this table and
the following paragraphs describe the various entries.
Space is allotted in the table heading for a sample identification code.
It was assumed that each laboratory will have devised its own coding
system for uniquely identifying the various samples. It may be desirable
to include the date, the name of the analyst, or other similar information
as well.
The body of the table includes one column for each of the LC fractions and
one column for summing of the data. The first set of data entries is the
quantitative analysis results, transcribed from the LC report (Table 21).
3
The calculated total organic loading, in mg/m , corresponding to each
fraction is entered in the first row. This value is used in estimating
the abundances of the various organic compound classes. The next two rows
contain the estimated TOO arid GRAV values, to indicate the distribution of
total organics between volatile (b.p. 100-300°C) and nonvolatile
(b.p. >300°C) materials. This information can be useful later in the
selection of appropriate DMEG or other decision criteria values for
comparison with the Level 1 results.
The results of the LRMS analyses are summarized in the table as follows.
The major categories present (item 1 of Table 26) in LCI are listed at the
left-hand side of the table and the approximate intensity (100, 10, or 1)
for each category is entered in the LCI column. To convert the LRMS
intensity index to a concentration estimate for each organic compound
category, the individual intensity value is divided by the sum of all in-
tensities for that LC fraction and then multiplied by the total organic
80
-------
TAftlK >6
ItVM. I 4tHCW\NU tXIKA< r i>UHHl\KY 1AHI Y
ORGANIC EXIHACr SUMMARY TAHtE
ICI
LC2
LC3
IC4
LCb
LC6
icr
l-
lot*) CH«jjmuci, tt«g
rCO, my
CiKAV, n»j
---
- — -
Aixgnad Initmnly MW
-------
3
loading (mg/m ) estimated for the fraction. This procedure Js then
repeated for the other LC fractions. Examples are worked in Figure 19
for two LC fractions, LC2 and LO, of the same XAD-2 extract.
The results presented in Figure 19 illv.s.rat^ ue c ''jiderable overlap
in chemical class composition that can be e>.4 >_cted between some frac-
tions in the Level 1 LC scheme. As noted earlier, this is not a high
resolution separation technique and the various compound categories
cannot be uniquely assigned to particular LC fractions. The data do
show the expected trend, In that LC2 is relatively richer in light
aromatics (benzene and fused species with MW <216) than is LC4. The
fact that adjacent LC fractions can be expected to show gradual changes
in chemistry can serve as a useful guide for the analyst in detecting
contamination. Very abrupt changes in apparent composition, or the
appearance of a con pound class in an entirely unexpected fraction
(i.e., phthalates in LC2 or paraffins in LC6) should be regarded with
suspicion.
The overlap between fractions can also l>e used in estinauing the conpo-
sition of those fractions which did not contain sufficient material to
trigger an LKMS analysis. For any LC fraction that was not analyzed
by LRMS, it is suggested that the IR spectrum be compared with the IR's
of the adjacent fractions. If a close correlation is found between two
IR spectra then this, together with the known behavior of the LC scheme,
suggests that the two LC fractions have similar, though not identical,
qualitative composition. It is therefore proposed that the total
organics in the non-LRMS LC fraction be distributed over the same
classes and in the same proportion as was done for the adjacent LC
fraction whose IR spectrum was the best- match. The error introduced
by this procedure into the overall description ol sample chemistry will
be small, since only fractions with small amounts of material are
excluded from the LRUS.
Concentrations -jstimuced by this procedure should be identified with
an asterisk In the organic extract summary table aud an explanatory
footnote included.
82
-------
LC 2
Total organics = 0.57 mg/m3
LC 4
Total crganics = 6.6 mg/m3
Aromatic HC's-benzenes 10
Fused Arom <216 MW 100
Fused Arom >216 MW 10
Heteroyclic S Compounds 10
Calculation of Concentration
intensities = 130
00
W
Aromatic HC's - benzene: x 0.57 = 0.04 mg/m3
Fused Arom <216 MW: x 0.57 = 0." mg/m3
Fused Arom >216 MW: x 0.57 = 0.04 mg/m3
Heterocyclic S Cmpds: x 0-57 = C.04 mg/m3
Fused Arom <216 MW 100
Fused Arom >216 MW 100
Heterocyclic S Compounds 10
Heterocyclic 0 Compounds 10
Estimates by Category
X! intensities = 220
Fused Arom <216 MW:
oto
oh
X
6.6 = 3 mg/m3
Fused Arom >215 fW:
100
220
X
6.6 = 3 r-g/m3
Heterocyclic S Cmpds:
10
220
X
6.6 =0.3 mg/m3
Heterocyclic 0 Cmpds:
10
220
X
6.6 =0.3 mg/m3
FIGURE 19. EXAMPLE CALCULATIONS OF CONCENTRATION ESTIMATES FROM LRMS DATA
-------
If it is not possible to relate the composition of an LC fraction to
LRMS data, either directly or by analogy to another LC fraction, then
it will not be possible to describe the LC fraction chemistry with the
same degree of specificity. If only the IR and LC analyses are per-
formed, a compromise procedure might be to: 1) List all categories
that could be in that fraction chosen from Tables 3 and 4; 2) Assign
an intensity of 100 to each category for which functional groups were
identified in the IR; and 3) Assign an intensity of 10 to each category
for which functional group frequencies were not identified. Then, in
the absence of evidence to the contrary, assume that categories with
an intensity of 100 may constitute up to 50% of the total sample and
those with an intensity of 10 up to 10%. Clearly, this procedure
will "account for" more than 100% of the total sample. However, this
conservative method of estimation seems necessary because it is not
possible to determine from the IR spectrum of a mixture how the various
functional groups are assembled into molecules or classes.
Figure 20 illustrates this procedure for an LC6 sample, assuming only
IK and LC data were available. Concentrations estimated on the basis
of IR and LC data on!y should be marked with a double asterisk in the
organic extract summary table and an explanatory footnote should be
included.
Once concentrations have been estimated for all compound categories
identified in the seven LC fractions for a particular organic extract,
or neat organic liquid, these values are summed across each row of
the table. This procedure condenses the information obtained on the
various LC fractions to provide an integrated description of the
chemical composition of the organic extract. It is inappropriate to
state the concentrations of various categories to more than one signifi-
cant figure at this stage of the data reduction. Examples of completed
organic extract summary tables are given in Chapters X and XI.
84
-------
Stinifile IC S
Total Organic! » 0.28 ny/m1
IR Ro|>ort
-1
y. Lm
34iin
3050
21IS0-2940
22 JO
1700
1600
1'jV) fl
1420-14SO S
l:)'>o-noo 'I
700-850 S
A$s!unmefU/Coupon ti
OH or Nil/Broad
Aromatic (II
Aliphatic 01
C II or C C
Ketone. Carboxyllc acid
ConJ. C C or aromatic
C H. N-NOj. C-N'O
Carboxylate
C N, C-NO?
CHj, HIl,'
Multiple peaks
Sul>st. benzene rings
Categories Possible
in IC b
Heterocyclic N compounds
Heterocyclic 0 confounds
Heterocyclic S compounds
Sulfides, Disulfides
Nitriles
Ithers
Aldehydes, ketones
Nitroaromatics
Alcoliols
Amines
Assigned
Intensity
100
100
10
10
100
100
100
100
100
100
Haxiujo
Possible
Conientiat inn
0 14 nrj/ni1
0.14
0.01
0.01
0.14
0.14
0.14
0.14
0.14
0.14
MUIKK 20. INTIMATION OK KKACTlOtl ('(IMitlbn ll)H KKtlH IR AMI) IjC DATA tINI.Y
-------
If the organic, analysis results summarized in the last column of Table 26
were for a neat organic liquid sample or for the extract of an aqueous
or solid process or effluent stream, then the "E" data would be appro-
priate for comparison with MATE values or other decision criteria as
described in the next section. If, however, the results in Table 26
were for one of the several organic extracts obtained for a single SASS
train sample, the "I" data for each of the extracts should be pooled to
yield one estimate of stream chemistry for comparison with decision cri-
teria. Table 27 indicates a format for that summary table. If the cri-
teria concentrations are exceeded, of course, it is possible to refer to
the data in Table 27 to determine whether the material of concern is in
the particulate, vapor, or both.
F. Comparison with Decision Criteria
The previous section has described a set of procedures for interpreting
and integrating the Level 1 organic analysis data. The results of those
procedures is a list of organic compound classes and estimated streams
sampled. These concentrations can be compared with appropriate decision
criteria values in making a Level 1 environmental assessment decision.
If one or more of the estimated concentrations exceeds the levels of
concern, it may then be desirable to reexamine the organic analysis in
detail to determine more specifically the nature of the material respon-
sible for clie positive "trigger."
As noted in the introduction to this document, it can be assumed that
many, if not most, users of EA data will be interested in using decision
criteria such as DMEG values for effluent streams. The DME* values are
specific for individual chemical species, while the organic an.'ly ;is re-
sults give concentration data for chemical categories. Prudence dictates
that, for each chemical category, the DMEG value moit appropriate for
comparison is the value for the most hazardous compound (lowest DMEG
value) in that category that could possibly be present in the sample.
In many cases, especially when LRMS data were obtained, it will be
86
-------
TABLE 27
SUMMARY OF RESULTS FOR ORGANIC EXTRACTS FOR SASS TRAIN SAMPLES
E,* mg/m3
Particulate Module Sorbent Module
J"
Resin +
Rinses >3 nm <3 iim Rinse • Condensate
Total Organics
CATEGORIES
Sulfur
Aliphatic HC's
Aromatics - Benzenes
Fused Aromatics <216
Fused Aromatics >216
Heterocyclic S
Heterocyclic N
Heterocyclic 0
Carboxylic Acids
Phenols
Esters
*Sum of results for all LC fractions for each extract.
87
-------
possible to rule out the presence of the "worst case" compound in a
particular MEG category. The LRMS data may indicate, for instance,
that the fused polynuclear aromatic hydrocarbons in a sample are all of
MW <216 and that, therefore, one need not use the very low DMEG value
for benzo[a]pyrene, MW 252, in category 21.
The recommended procedure would be as follows:
1. List the categories of compounds found in the stream, with MEG
category number (Table A) and estimated concentration.
2. For each MEG category listed above, identify the compound with
the lowest DMEG value.
3. If the estimated concentration in the stream is lower than this
DMEG value for a category, go on to the next category.
A. Review the data to see if there is an;/ evidence that that compound
cannot be prerent in the sample. This evidence may be LRMS data
on MW ranges or compound types, such as the fact that amines
present are not nitrosamines. Other evidence may be that the
compound has a boiling point <100°C and therefore was not possibly
present in an SASS sample.
5. If the compound with the lowest DMEG can be ruled out, identify
the compound with the next lowest DMEG in that category and re-
iterate steps 3, 4, and 5 until DMEGs have been chosen for all
categories identified in the sample.
6. The highest DMEG value in a MEG category becomes the default value,
to use if all lower DMEG correspond to compounds not possibly
present in the sample.
7. List compounds and DMEG values chosen as appropriate for each
category.
8. Tabulate ratio of estimated concentration to DMEG value for all
categories identified.
88
-------
X. EXAMPLE OF APPLICATION OF LEVEL 1 PROCEDURES TO FUEL SAMPLES
A. Introduction
In designing the present phased apnroach to environmental assessment,
the authors recognized the diagnostic importance of examining feedstocks
and fuels — the input material to industrial manufacturing and energy-
producing processes — in order to identify potential environmental
pollutants at their source. Presumably, direct information of this
kind would be useful in selecting and applying pollution control tech-
nology to those industrial processes, or in specifying and controlling
the composition of fuels and feedstocks so as to minimize the amount of
effluent control required in certain cases.
A 1977 report by L.N. Davidson et^ al^. , "Technical Manual for the Analysis
of Fuels," (8) identified the principal established methods for sampling
and analysis of hydrocarbon fuels. Those methods were generally developed
for the use of engineers concerned primarily with the efficient utiliza-
tion of fuels as energy sources. While many of them are capable of
providing valuable information about the composition of fuels, and
particularly about components such as total sulfur, nitrogen and chlorine,
and residual ash, they leave much to be desired from the standpoint of
direct analysis for trace elements and for specific organic substances
other than those constituting the fuel itself.
One approach to the identification and measurement of potential pollutants
in fuels Is to subject those fuels to an appropriately modified Level 1
and/or Level 2 analytical scheme. As an example of the kind of information
that would be obtained from conducting the organic analysis portions of
the Level 1 protocol on fuels, two sets of results are included here,
pertaining to coal and to //A fuel oil. Certain obvious limitations can
be seen in the use of the Level 1 protocol on such samples, especially
in regard to analysis for organic species. These limitations will be
discussed, and some suggestions will be offered which should help, at
89
-------
least partially, to overcome them. These suggestions constitute the
best guidance that can be offered at present for sampling and analysis
of fuels beyond the use of conventional methods and of the Level 1
protocol.
B. Level 1 Sampling and Analysis of Tuels
The information obtainable from a conplete Level 1 assay of fuel samples
is likely to be useful in detecting potential pollutants in fuels or in
identifying the source of such pollutants when they are discovered in
effluent. To make best use of that information, however, one needs to
know something about the chemical and/or physical proc rses to which the
fuel will be subjected in the course of producing various kinds of
effluents. For example, a pile of loose coal, standing uncovered and
exposed to the weather, might produce a windblown dust which would be
chemically unchanged and which might need to be considered only as an
upreactive particulate. Leaching of the coal pile by rain could produce
an aqueous leachate that night be contaminated with soluble ions such
as sulfate and which might also be highly acidic and contain various
dissolved tTPce elements in ionic form. Conversion of the coal to coke
or to a liquid or gaseous form involves chemical as well as physical
changes, and results in the production and/or liberation of some chemical
species that would not be found by a Level 1 analysis of the original
material.
Combustion of a solid or liquid fuel may produce a great variety of
organic substances as well as a smaller variety or inorganic species,
depending not only on the initial composition of the fuel but also on
the conditions of combustion and of interaction between the effluents
and the environment into which they are discharged. The large influence
of these latter parameters (i.e., combustion and discharge) on the
identities and quantities of emitted organic substances places severe
limits on the value of any analysis of the composition of such a fuel
for purposes of predicting the nature of its potential for environmental
pollution.
90
-------
The protocol for a Level 1 analysis calls for treatment of gaseous fuels
as gases, to be collected as "grab" samples and analyzed accordingly by
on-site GC. If, in addition, it were decided to determine whether a
gaseous fuel might contain trace-level components in the higher-molecular
weight range, the Level 1 gas analysis might be supplemented by a vapor
3
analysis. A 30 in sample might be passed through the sorbent (XAD-2)
trap of a SASS train, Just as if it were the mixture of air and combustion
prodjcts in a stack effluent. Any material collected on the sorbent would
then be removed and analyzed by the recommended procedure, involving
liquid chromatography followed by infrared spectrophotometry, gas chromo-
tography (TCO), gravimetry and low resolution mass spectrometry (LRMS).
Liquid fuels include, in addition to liquid petroleum fuels, shale oil,
coal liquids and methyl fuels. The principal distinguishing characteristic
of the last-named group is its miscibility with water, since methyl fuels
consist essentially of methanol, and alwr.ys contain small quantities of
dissolved residual water. A Level 1 organic analytical procedure for
any fuel should be carried out directly on a small sample of the fuel
which can be treated as if it were a methylene chloride extract of a
sorbent module or of a solid sample. If the fuel is a liquid petroleum
product, the sample will consist primarily of aliphatic hydrocarbons.
The Level 1 liquid chromatography (LC) procedure will, therefore, show
predominantly high yields in fraction fll, with smaller but still substantial
quantities appearing in fractions fc'2 and /'3. Shale oils and coal liquids
can also be expected to show large yields in these three fractions.
Methanol fuels will consist primarily of compounds that appear in fraction /'6.
Infrared and LRMS analysis othe fractions will help to characterize both
the matrix materials and the minor components of all of these liquid fuels.
Analysis of coal fuels and other solid fuels by the Level 1 protocol may
be limited value, as the foregoing discussion has suggested. The bulk of
such fuels, although organic, is not readily soluble in ordinary liquid
organic solvents under near-ambient conditions of temperature and pressure,
91
-------
even vhen the original sample is ground to a fine powder before extraction.
The results of a Level 1 analysis of such fuels are, therefore, applicable
only to the description of a minor portion of the total combustible fuel.
C. Level 1 Organic Analysis Results for a Fuel Oil Sample
An example of an actual Level 1 analysis of a liquid petroleum, //A fuel
oil, is summarized in Table 28. A 100 mg portion of the oil was applied
directly to the Level 1 LC column, with no sample preparation. A separate
portion was dissolved in pentane and analyzed according to the Level 1
TCO procedures. More than 98% of the fuel oi] was found to be CRAV range
material. The data show that the major organic components of //4 fuel oil
are aliphatic hydrocarbons (68%), aromatic hydrocarbons (14%) and fused
aromatic hydrocarbons in the molecular weight range below m/e = 216 (11%).
Trace quantities of fused aromatic hydrocarbons of m/e > 216 were found,
as well as traces of carboxylic acids.
The Level 1 LC, IR and LRMS reports which were integrated to produce the
organic extract summary table are presented as Tables 29-39.
No TCO analyses were performed on the LC fractions for this sample because
the TCO value for the unfractionated sample was less than 2% of the GRAV
level. Therefore, the TCO portion of the LC report, Table 29, is left
blank.
Comparison of the IR reports (Tables 30-33) with the LRMS reports (Tables
34-39), for the same fractions reveals no inconsistencies between the two
types of qualitative analysis data. There is no LRMS report for LC 5
because the total quantity of material in this fraction (0.4 mg) was so
small. The chlorinated hydrocarbon material identified in LC 6 by LRMS,
confirmed by IR in LC 5, LC 6 and LC 7, was not reported ir. the organic
summary table because it Is thought to be a contaminant introduced in
the laboratory. As indicated in Tables 16 and 24, chlorinated hydrocarbons
are expected to elute in fractions LC 1, LC 2 and/or LC 3 of the Level 1
92
-------
I Alt) t 2H
hm.ank: txrktarr mkhaky iaiul h»u ia tuu. on
Sample ***** Pt* Analysed
LCI
LC2
ICJ
LC4
LCU
LCB
LCT
I
Tout Oxjjiuci, mj/m^
ICO 1119 »
6~i
~2l
1.4
*«0.4
<>r 216
100/1.3
10/0.1
Hi-rurixji 1 Ic NtCiof;en Compounds
1 4. r-i
1/0.6
1/0.6
1 to. 1
0.8
0 11 boxyl Ic Acids
"1/0.4
1/0.1
1.1
—.
* Not done, due Co tho low content of volatile constituents.
*A<«n(M.iittLr.itlon cstlmileJ from IK Jitu only.
-------
1 AH! t jy
II' KMMltr KOIt nit-1 OIL
SAMPIE. No. 4 Mh>J Oil (Until hit it)
ICO
GHAV
ICO * GHAV
loui (Bg
CurarsffMM
ToM Sunpb^
fmkva for LC2
fUce*«Ml'
< /
loo
IOU
lirti
a cm
Ifit>
t;M*
101)
\Q
£-
FlKtMXt
TCO
ificng
GHAV m tu§
ICO*
Cbw>W4m>
hound in
Fraction
QUtik
Cm
reeled
Toul4
Found m
FfKlMA
bUnk
Cm
90CtW
Tut**
GHAV
TwUifflf
i""*'
1
f.%.4
bS
2
6 1
-
0 4
6.4
»
?
-
JU
^4
4
1 <
-
1.4
1.4
MM
6
(1.4
_
f>.4
0 4
e
L9
.2—
_ ^9
1.
7
1 2
0.9
0.1
l».l
Sum
101.1
urn
KHJ
1 Qmirtity in mttii wtrpla, 4iltimm«il bctMe IC
2 PioiImi ol whoW um^lc used f«M tC„ tcluJ rnf
3 Quantity recovered liom LC cutuimi ecluel nttf
4 ToiJ mo bid to (u(«t ttnipU
A
H/A - Ni!UubU
-------
TABLS 30
IS REPORTS: FIEL on. rtACTIOSS IC 1, LC
IR REPORT
SAMPLE rMaI Oil. t.C rr»ccion I
¦ j
Wat* Numftw |
lem'M j
lAtmmy
| AMigrmM
' — T
Comnum* |
3C*30-:300 i
S
1 Uirfarie C1-C3i
i
::60. 13'0
y
1 ilimaclc CH*
i
720
*;
i Lena Chain
1
J „ I
I
T
I
IR REPORT
SAMPLE *'• "u«l 311. I.C Fraction 2
V»«* Numtar I
(cm") |
lAtirwty
| Aaatfptmtflt
1
Comment)
210Z-20C0
V
Aror^cic CH
>000-2?00 1
5
t Aliphatic CM
1600, 1500 !
*
Aronacic CH
li*9. 1]70 1
i Aliphatic CH
»:0. T50
V
f \ronaclc Susstitucicr
1
I
i
95
-------
TABLE 31
IS REPORTS: F^EL 01*. "ACTIONS 1C 3. "_C -
IR REPORT
SAMPLE 'i ruel Oil. -C Fraction 3
Wan Numb* j
(em"1 ( |
tnttnwtv
Auignmmt
Comrntnu |
j
2 no-"WOO 1
M
s.rcna-ic CH
I
3003-2900
s
CH
i
1500. !JC0
U
Ar-nat1c CH
;
Ii50. lite 1
<1
\llpnaclc CH
1
1020 1
V
Arccaric Sub«cicjcion
.
«0. 810. "501
M
Vronatic Substitution
!
:
i
1 1
1
1
I
;
;
IR REPORT
SAMPLE 'i "uel
Ml. LC -net
fen -
W«« Number {
(on"1) |
Intimity
Amfwxni
Commtmi j
I
t
3500-3iC0
V
or ''"K. 3road
,
3100-27C0
• *
OH 3rD3d
'
3050 1
*•
Aro~-acic CH
i
3000-:soo
s
Ulshatic CH
1
roo
V
\cU C - 0
1620. i:oo 1
•«
\«V>- . STU-
¦
1-50. iro
\l.rhatie CH
:;jo
Vrino. N'itro. J the"
i
30C. -:Of
w
Arciaclc Substitjt:o«
1
i
I
96
-------
TABLE 31
:a a£?e<*rs: fi=:l cil factions lc 5. lc 0
IR REPORT
1AMPLE. tj«I .HI. 1C rnctlon 3
Win Number
lem"1)
**00-3200
lAtifWiy
AUl^HMRt
Cemmwiu
32QC-£™0
OH. \cld
3000-:300
A., lyhac ic CH
i-*0. 13"0
\1 iohaci: CH
1150, 1020
Alcohol, Ester
7:0
ton., CP* Chain
1R jtEPOAT
SAMPLE " el . LC —6
Wm NumW
(cm"')
Imtnaty
Antgravant Cemmenti |
3»oo-::oo
V
GH i
300C-2800
s
Alioharic CH
1*00. 1600
,*
Carsorwl. id 1
1-60. 1370
V
U1 snarls CM 1
1020
M
Mcoht?!. 51-3
"50
M
C-Cl
i
i
1
1
1
!
¦
97
-------
TABLE 33
IR REPORT: FLIL OIL FRACTION LC 7
Ift REPORT
SAMPLt -- Fuel Oil. LC Fraction 7
Wan fJumtoir 1
Intwvtry
Amyiwuit
Comment!
(cm"11 j
3*00-:730 j
X
OH
)ooo-:soo >
S
Aliphatic CH
1700. :aoo I
V
¦ - "
C " O, Acid, Vetono
1460. 137P 1
M
AlJrharic f*H
io:o 1
7
Alcohol. Si-0
750 !
vw
C-Cl
1
t 1
j 1
1 1
; 1
i
1 1
1
1
IR RSPORT
SAAffLE
Wm Numbftf
(cm"J)
Attignmtfft
Commintt
98
-------
TABLE 34
LRMS REPORT. nTIL OiL FACTION LC 1
LAMS REPORT
SAMPLE ' * Fuel Oil. LC lrictior, I
MafOf Ctt*9MMl
InunutY
I Cttagory
M*| Rang* 1
lOO
[ Allchaclc Fvdrocirbon*
L'O-632 '
I
J
i
Svb-Ottgofwt. Sp«Cif*e Compounds
Intanuty
Category
m/« | Composition
100
Sjtjriced Miohatic HCs
170-^2 ^c Cusi. -
100
Lnaatt-racec \liohatic HCs
."JO-'frS 'l-H-.-C--F»a
!
1
!
1
1
1
1
i
I ! 1
:
1
i
1
1
1
1
!
1
i ii.
i
1 1
i
'
I
i i :
i 1 ! ;
: i
Ottar
99
-------
TACLc 35
LRXS REPORT: FUEL OTL FRACTION LC 2
LRMS RCPOBT
SAMPLE _ ^ ?i»ol Oil. LC F-jc:lon :
ftfcjot C«t«90fNi
IniMWty
I Gitvgorv
MW Ranfa
:o
1 Arc-vatic Hv«Jrocarbons
io<>-*2a
- o
CvrUc Allnh.iric Hvdrocarbons
to •>¦)!
1
1
1
i
SufrCtttgorttf S«wofiC Comooundi
Inunuty
Catteory
m/t
CompowtTOA
10
A1V I 24*izenes
lOft-*
!4 C.u,„ to C ,Hf.
10
Dent tcclcparaf firs
tO
} to C'iHc r
1
1
; i i
> ' i
i i l
! i
! I
i !
¦
i i
'
! !
I 1
i
!
I III
Otfccr
J
I
J
100
-------
TABLE 3b
LRSS IE"CRT- FUEL OIL FRACTION LC 3
LRMS REPORT
(AMPLE rutfl 1C Fraction 3
Mi (Of CattfDoM
Inutility
| Ctueorv
| WW Rang* 1
100
1 \rona:ic Hvdrjcirhona
I 0?-|Of> |
J 00
1 Ar?nei: H/Jroearbens. < 216
j n 1.19a !
10
1 cused rtroffaclc Hvdri-carbons, > 216
1:5:-ino !
1 1
i 1 1
i ' i
Sitb-Cragorm Sjaafic Compound!
Intftuitv
C«c»9o*y
1 m/»
Composition
100
A,1V"L Sersenes
®2-l*>0 C-rts- C< -i* -
iro
A1V* 1 \mhrh.i1t»n<»i
• "M- ?
r
r|-H-- r*-b-,
10
A!kvL ^nthrnceies/Pner intbrene*
or Cs-4»-
1 ! 1
1
i
1
i ;
: i !
! !
I
I
i
1 !
1 1 1
! II!
1
<
i
i
»
Ottwr
I
1
101
-------
TABLE J7
LRMS *£PORT: REL OIL FRACTION LC 4
LAMS REPORT
SAMPLE _ al* ruel Oil, 1C gracelon 4
Mi jo* Ctueofttf
Intmiity
| C«t«eory
MHV rang* ,
10
1 Heterocvclic Nitrogen Compounds
i
I
1 Eiter#
i
1
I
I
i
1
i 1 ;
Su^CnafOftn, SMcrfie Compound*
Intoftatv
Cittgory
m/9
Composition
10
C irbnsMes
1*7
C«-H,N
:o
BenzocarbasDles
:i;
10 1 Mk/laced Cirbazoles
C.-H.-N
1 i Phchalnes
1
i
i
;
i
i
i
l
i
i
i !
t
i I
1
1
J
1
Other
I
1
j
102
-------
T\BLE 38
LRMS REPORT: *TEL OIL FRACTION LC 6
CRMS RCPORT
SAMPLE ru«l Oil. 1.C rrictlon 6
Atop* C«t»90f*M
Inunuty
1 C*t»«ort
MM Ran*
1
1 Carboxvlie \clds
1
| Esters
I
1 Chlorinated Hvdrocarbors
! 1
Swb-Catagofiw. Sp«of*c Compound!
Int* rutty
Cltffory
m/.
CompoutioA
1
ralaltic Acid
:><>
r _h,-o-
^toarlr Acid
:si
C-=rl,.n.
i
Chlorinated H"drocarbois*
i
Chlorinated Hsdrccarbons*
230 I
t
Dloctvl Pnth.ilitcs
1
i
i i I
1 i
1
1
i
i
I
i
1
1
i i
OttMf
*probibly cortiolnantfl Introduced dtirin2 lanole handling
103
-------
TABLE J9
LR.YS REPORT: FL'EL OIL FRACTION LC 7
V.RMS RtPOflT
IAMPIE 114 Oil, tc rractlon 7
Mjjof Caugmm
1
Cit*9sry
MW Rtnga I
1
! Cjrboxvlle Acid
122-284 '
1
i Asters
1
t
i
1 i
h^CittlorMi, Spaotic Compound!
Intansiry
Cftttjonr
Composition |
1
Den/ole Acid
1
1
Scetric *cid
C' '
1 ' Dfoct I 'hr'iitace 1 1 !
II!
' !!>
i
\
1 1
* 1
i
I
1 |
1
i 1
1 ! 1
i !
¦
i i
| 1
•
! i
i
! !
OUm>
J
104
-------
scheme. A positive finding of these species in the later fractions only,
with no indication of their presence in the more probable fractions, is
a strong evidence that these are spurious results due to contaminants.
The fatty acids and phthalates reported in Tables 38 and 39 may also be
contaminants. However, since these species would be expected to be most
abundant In I.C 6 and LC 7 where they are, in fact, found, there is no
basis for reacting them as possible sample components. The carboxylic
acids, therefore, are reported in the summary table (Table 28).
D. l.evel 1 Organic Analysl-t Results for a Coal Sample
The results of the application of Level 1 organic analysis procedures
to a coal sample are presented In Table 40. These data were obtained by
analysis of a methylene chloride extract prepared by 24-hour Soxhlet
extraction of a 26.3 g sample of coal that had been ground and sieved
to pass a 60 mesh screen. Less than 1% of the mass of the coal, 3500 rg/Kg,
was extracted. Of the extracted material, approximately 22% was TCO range
and 78% was O.RAV range material; this is a higher level of volatile
material than might have been expected.
Analysis of the composition of the extract showed the nredominant species
to be the lower molecular weight (m/e > 216) fused aromatic hydrocarbons
(367. of the total sample after LC separation). Heterocyclic nitrogen
compounds accounted for 22% of the total, and phenols another 21%. High
molecular weight fused aromatic hydrocarbons (m/e > 216) comprised only
5% of the sample and although LR.MS analysis detected elemental sulfur in
LC fraction //1, no organosulfur species were detected.
Tables 41 through 52 show the results of the LC, IR and LR>IS analyses
that were integrated in constructing Table 40. The LC report. Table 41
shows that 80% of the GRAV range material, but only 39% of the TCO range
material, was recovered from the LC column. The overall sample recovery
was 70%. The "TOTAL" values in Table 41 have not been corrected by this
recovery factor; the only correction applied i<3 the factor of 2.7 reflecting
105
-------
TABl.t 40
ORCAHIC EXTRACT S'J^tAKY TAHIL HtR <<>At SAUl'Lt
Sjmpl* (IMN extrict of ?f> 3 k rual
LCI
LC2
72
0.J
1 .6
LCI
__990
6.V
19
LC4
270 _
2.5
4.7
LC5
lcs
LC7
X
Tot«1 0*9Jnia n»g /k^
200
_99
240
hOO
J25<)0
9.7
TCO. mg
5.1
.
GRAV mg
2.6
ft 3
16
55
Cateyoiy Aliiuswi lnt»m*lv MG/
All |>li*tC i«_ llydrucjrbons
10/100
10/100
100/71
l/l
I00/V00
10/»)0 __
*\
* t
_ JOO
_yoo
120
550
Sil 1 f IT
t-nscj Ariunitlcs < 216
fused Aruci.iLIon > 216
JO/31
10/33
HcUnuvrllc N Cnnpnundtf
10/245
10/270
HiLnnIs
10/240
10/270
510
Cnrhnxyl li AlMii
1/27
27
—
-------
TABI b 41
ir kHMKr M)N l'oai ixiKArr
SAMPLE f-Mimt fr<».3
5.3
2d0 tuj /K^
2
Q 1
-
0.1
O.J
0.6
-
0.6
l.h
1.9
72
3
2.5
-
2.5
6 9
7 2
7.2
14.
26.
990
4
0.2
.
0.2
2.5
1 .8
_
1.8
4.7
7.2
270
5
-
-
-
I 0
-
1.0
2.6
2.6
99
6
-
-
-
2 k
-
2
6.1
6.3
240
7
-
.
-
6.8
0 9
5.9
16.
16.
1.00
Sum
3 5
-
3.S
9.7
21.8
0 <*
.'0.9
¦>s.
6S.
25(K)
1 Quantity «n «nlu««*fnpto, det«rmm«d b«for« LC
2 Portwn of whola tampfc utcd for LC, actual mg
3 Quantity r«covcr«d from LC column adud mg
4 Total mg computvd back to lotal ianipta
-------
TAjJLE 42
IR REPORTS * COAL EXTRACT LC 1, LC 2
IR REPORT
SAMPLE Coil Extract, LC traction 1
Wav* Number
Imamtiy
Aniywwt Cornmtntt
(em'M
3 *00
Cf/verv oroad
3000-2800
S
Aliphatic CH
l«i60
S
\llohjtl'. CH
1380
M
Al'.nhitlc CH
730. 720
Doublet CH» .-hiln tenner thnn * units
1 i
1
1
1
1 I
; i
i i
' »
! 1
IR REPORT
SAMPLE Coal Extract, LC rrnction 2
W Number
(cm*1!
Inttiwtv
A*sipun»nt Comnwnn j
3400
V
OH or \*H/brjnd 1
3000-2300
s
Aliphatic TH. ur«;aturated CH 1
liuO
M
\I iohnt ic CH !
1330
V
Aiiohatic CH I
720
V
Alirhitic CH. rH-> rock '
! i
i 1 j
I : !
i i ;
i
t
i
i
i
!
i
i
i
108
-------
TXSLE O
la REPORTS: COAL EXTRACT IC 3, LC i
(R REPORT
tAMPU Coal Extract, 1C rraction 3
Watt Numfetf
Iranotv
Anlpuiwit Comwnu
(em*1)
o
0
o
1
o
©
X
\roracic C4
30oo-:soo
S
A1Iphaclc CH
'*00
Xromatlc rlisj
1450
s
Alfchatlc CH
iro
M
Aliphatic I'M
370. 905. 740
1
Arenacic substitution
•
1 1
i !
! !
!
i !
; t
i i
IR REPORT
SAMPLE Coal Extract, LC friction '*
f.lrt Humbut
tnurmtv
AMiqrmtm Comments
lon"'l
3«.20
W
OH _»r Nrl/broad '
3080-3000
u
Atom 11 c CH 1
3C *0-2500
s
j
A1 »phitic CH !
1600
1.'
Aronitic rins. isiine* nlt-oso 1
1430
Mlohitlc CH
1370
V
\liphatlc CM
970. S00. 740
Aromatic «ub«>tlti.tlcn
1 t i
i 1 !
i !
! i
]
1
1
i
' 1
109
-------
TVSLE i*
IR REPORTS: COAL EXTRACT LC 5, LC 6
IR REPORT
SAMPLE CojI Extract. LC S
Wan Numbar
(cm-'l
Intimity
AsMgnmem Comments
3i00
•J
or NH'bmd
3080-3000 !
Aromatic CH
3O0Q-2Sn0
S
Al1r»h itic CH
1700
u
Acid, lector?
1600
J
!i60
U
Aliphatic CH
1370
J
Allrhitic CS
750 !
Aron.itif imSr?'rjrion ':M-
; 1
1
i
|
!
IR REPORT
SAMPLE Co.il Ertrnt. LC ft
Wm Number
(cm"1)
im*mity
' 1
AtMpwtwnt Commtntt
3300
M
CH o. \*H/broad
30S0-3000
V
Aroaaclc CH
3000-2300
s
A1 irjh.irltr CV.
i
16°0.1650.1600
Acid, ketone. m!rp !
1450
A11 oh.ittc CH. n!tri<«.in(»f 1
1220
w
Ether, nheno! 1
1050
-•
llrrnnl
?so. 9:o. :¦¦¦)
s
i
i
i !
'
i
i
i
|
'
i
i :
110
-------
T\BLE 45
IR REPORT. COAL EXTRACT LC 7
IA REPORT
SAMPLE Coal Extract. LC craction 7
Win lumbar |
(em"1) |
ifttmuty
Autpnimnt
Commwu
3350 !
M
•¦>!! ir yf/hmJ
3coo-:son 1
s
AllDhiclc CH
1730 !
v
Ester
16=0. 1610 i
Ketone, iMd. anfne
1*60 1
M
\llnhatlc CH
1370 !
•>
Aliphatic CH
i:eo, 1220 . 1
V
,
Estor. other, alcohol
mo !
•4
£ther. alcohol. ester
1020 1
Alcohol, icld
750 '
->
\ro*n i 1
IFI REPORT
SAMPLE
Win Numfc«f
Inttfmtv
Awgn^
Comment*
(cm*4)
I
i I
i .
Ill
-------
TABLE 16
LR.XS REPORT: COAL EXTRACT LC 1
IRMJ REPORT
SAMPLE Caal Extract. LC Fraction 1
Mlior CtU«oiiM
IntMwty
C«t*9orv
MW Rtngc
10
Aliohatlc h' dTOcarbona
Sulfur
i
1 i
i i
i !
Sub-Citrjixm. SokiIic CamoouMt
Inunftty
Category
i mr.
Compouuon
10
'
Saturated aliphatic hydrocarbons
212-.A C« -Mil to
1 -22
10
Lnsaiuratttd aliphatic Vv>drocjrDons
i 39S C-sVU, to
1-.B2
C i c>ir ¦>
1C
Sulfur
236
5c
1 1
1 1
t 1
: 1
i
i !
•
i i
i i
t
; ! 1
; • !
i
1
i
1
1 1 f 1
1
i ;
Oihtr
(
I
1
112
-------
TABU 47
LR.MS REPORT * COAL EXTRACT LC 2
LRMS REPORT
SAMPLE Coal Extract, LC rractlon 2
M»)0f Ciugorm
Inttfiirty
| CtMforv
100
^ \llrhttic h'orrcirh.ins
' *0-^22
1
Sulfur
Sub-C4t*90vit*r Specific Compounds
Intimity
Catttory
ffl/t
Competition I
10
?it irat«?d allDhitic hydrocarbons
173 te
Ci -»H-to i
1/.-
10
I'nSic united a I irhaclc nvrfracarbon*
182 t J Ci ->c to 1
;:o
C i n H !
10
*.n?a£uraeed aliphatic hvJrccarbons
130 c
-------
TVBLF 48
L1XS REPORT: COAL EVTRACT LC 3
LRMS REPORT
SAMPLE Coal Extract. IC rractlon 3
Miior CjOtwiti
IflUdMTV
r
1
C*t*gory
MV» Rang*
100
t
• rused
aroraacic hvdrjc.trb^ns. xA? * 216
10
I ru?pj
arc-itic hydrocarbons. m'e s 21*
>w-snr>
1
! 1
! i
i
Cwb-Cawgoi in. Sp*ef>e Compounds
fntamicv
Ctr»9atv
my#
Composition |
10
\lit^.l a iphrhal*;m;s
i:3-:u: c.-h= - wh-- i
10
Ac-'nashthene 'M^nt-n 1
i i
is—:ao c. s?.- - u .'i ,
10
\-.chncene/->,i i
1
1
' 1
1
1 !
1
1
1
1 :
! 1
> (
! !
i ! :
1 ,
I
! i
t
i
OtM -
. ,
114
-------
TABLiI *9
LRMS REPORT: COAL EXTRACT LC 4
LRMS REPORT
SAMPLE __ C^al Extract* LC Fraction 4
Major Caufwin
Intiftuty
| Catagory
MW Rang*
10
| Heterocyclic iltroncn coooou'.ds
1 ft 7 —2 7 3
1
i
1 Esters
•
i
;
! 1
! 1
i !
SuthCatagorm, Specific Compounds
Inunntv
Category
m/a
Composition
10
Carba»ole
167
r ] -Hi
i j r
10
Benzocnrbasole
217
C.cH,,N
;o
Alkylated benzocarbazoles
:7i
C t r 41 o\
i
9hthalates
'
1
i
*
i
i i
I i
i
i
I
!
1
I ! !
1 :
I I
! 1
' 1 :
Othar
115
-------
TABLE 50
LR.MS REPORT- COAL EXTRACT LC 5
LRMS REPORT
SAMPLE Coal Extract, LC ^ractloa 5
Mftjor C«t«foriM
Intmitv
Ciugm>
MW Rang* !
10
r«ers
i
10
Heterocyclic ni-rceen "-opaounds
10
Fjcod .iromtic hxiroc-., bens
¦>40-400 !
J ;
: 1 i
1 1 !
$wb>C«ttgoriw, Sotefic Compound*
Inttmttv
C*tc«onr
m/«
Compovtiofl 1
10 iDietSwl nhtSal.ite
¦- -v.. .0, 1
10
A Ik/I ejrbizeies
167
r -h-,\ 1
.0
Uk» t cTbi-ol.cs
r, •<.->,
10
AIkvI naohth..Ivies
... ;
13--:is c. p.e - Cirth- 1
10
Fused arociutic hvd^ocarbons. a/e s 216
lifl-i'lo 1
1 t 1 ;
1 ll:
! ; :
i ;
;
1
— - {
i 1
t
i
1
1
.
1
|
_ . __ _ _ A
I
I ! i :
i it;
1 ! I !
Othtf
_I
,
116
-------
TABLE 51
LR^S REPORT: COAL E'CTRaCT LC 6
LRMS REPORT
SAMPLE _ Coal ExtTic;« LC Fraction 6
Miioc Cit*9orwt
Inttmifv
Crogory
MW Rftnp*
10
Phenol s
108-150
1
1 !
Si»frC«t»gQr«t Sptcmc Compounds
m/t
CompotJtion
10
Aik\late
-------
TABLE 52
LRMS REPORT: COAL EXTRACT LC 7
LRMS REPORT
SAMPLE Coal Extract, LC Fraction 7
btigwitl
lnt«ntitv
f
Gutfory
MW Rang*
10
i
1 "•heaols
94-179
10
•Heterocyclic nitrocen
compounds
145-251
1
;Esters
1 .CarboxvJtc Ids
i
i
1
1
I !
Sub-C*t«forrtt Specific Com pounds
Inttnutv
Catvgonr s
m/a
Composition |
iO
*1!^ 1 ated ohenoJs
o;-> 7
* _ r. .j. .,i 1
10
AIkvI Indoles
•i5-?ni c .fi .v - c . h.oM
10
AlLvl cjrbaioles
,
167-15*1 c, - _ c. .Hi.S i
1
Boiroie acid
122
C-HcO, |
1
Phchilate
1
1
1 :
1 1 ! 1
1 i
i !
1
1
i
1
i |
i 1
'
i !
1
!
I
i
V
|
1
1 1
Othtr
10 ruscd aro^.ittc hydrocarbons i/c rOQ-^OQ
118
-------
the fraction of the total sample (38%) taken for LC.
As in the case of the fuel oil, comparison of the IR reports (Tables 42-
45) and LRMS reports (Tables 46-52) reveal no inconsistencies Jn the
qualitative analysis results. The phthalnte ester species found by
LRMS probably indicates sample contamination, since the material appears
to be present at equal amounts in LC 4, LC 5 and LC 7 but absent from
LC 6 which is the most probable LC fraction for phthalate ester elution.
The phthalate esters are, therefore, not reported in Table 40.
119
-------
XI. EXAMPLE OF APPLICATION OF LEVEL 1'PROCEDURES TO A SASS TRAIN SAMPLE
A. Introduction
This chapter presents an example of application of Level 1 organic analy-
sis procedures to a sample of gaseous effluent from a sealed metallurgical
furnace at a ferroalloy production facility. The sample was collected
by personnel of Monsanto Research Corporation using the IERI. Source Assess-
ment Sampling System (SASS train) (1) and was sent to Arthur D. Little,
Inc., for analysis. A report of the complete study has been published (9).
B. Summary of Level 1 Results
Data on the total extractable organic material for the various SASS train
components for this ferromanganese process sample are summarized in Table
53. Substantial amounts of organic matter were found in the extracts of
all SASS train components, except the sorbent condensate extract, from
the ferromanganese process. About 92% of the material was found in the
XAD-2 extract in this case, of which about 82% was found to be high-
boiling (b.p. >300°C) material.
The extracts containing more than 0.5 mg/m3 of total organic material
were taken through LC separations, and the seven LC fractions collected
from each extract were analyzed for TCO and GRAV as well as by IR and
LRMS. Tables 54-56 are organic extract summary tables for the probe
wash, fine particulate fraction and sorbent module portions of this SASS
sample and Table 57 presents the overall organic summary for the ferro-
alloy process stream sampled;
Some interesting aspects of these data are:
Extremely hig.h quantities of fused aromatics (Table 56) over all
molecular weight ranges were found in the XAD-2 module extract in-
cluding many non-volatile species. Also present in this sampls were:
120
-------
TABLE 53
TOTAL EXTRACTABLE ORGANICS FOR SASS TRAIN SAMPLES
OF FLRROMANGA.^ESE PROCESS EMISSIONS, mg/m3
TCP GRAV
Particulate extracts
10 + 3pm — 6.6
1m + filter — 48.
probe and cyclone — 37.
rinse extract
XAD-2 extract 205 910
Sorbent condensate 0.41 ^0.1'
extract
Source: Reference 9
TOTAL
6.6
48.
37.
1100
0.41
121
-------
TAblh 54
HK(.ANIC LX1 KACT SUMfWY ITA&U. Y l-UKIHAM.ANtSL I'ROi KSS SALS Cl'NDLNbAI t
S#itvto 11 *w» P>obo WaaU, rcrrrmnganege
LCI
LC2
LC3
28
IC4
7.6
LC8
0.54
LC6
LC7
£
~4J
Total OiQima, mg/m^
1.95
<0.1
7.2
1.44
TCO n*
2.05
<0.1
-
-
-
-
-
GRAV mg
39
11
0.73
9.8
1.96
65
Caie^ory fnr/mfl/m^
AUnl.etJc Hydrocarbon*)
<4.(1.2
1.9
ArcKutlc llvdrocarbons
i#1.07
100/7.7.
1/0 07
<<0.2
Fubcd AromatKe <*U
ion/14
100/14
14.
Fubed Aroffuiilcfi >216
100/0.2*
22
Heterocyclic N cotcuotinds
10/0.6
1/0.1
0.8
ketones (polvcycllc aromatic)
10C/0.2*
10/0.0^'
iq/q.o'>«
10/0^02*
100/6.S
10/1.3
7.8
fcuter*
1/0.06
0.08
Alcclioln
0.02
t"l 11 r 11 e
0.0,;
Concentration estimated fror I.C,IR data with rcicruue to U.MS «W.tj of LC4 and 1X6
-------
rwil t 55
okoanic extract summary table
S.0
66
Cllefnry Inl/mg/rr'
Ali|ihuclc llydrucorbonB
<'0.2
<<0.2
SI _
<0.2
<'0.2
Aroiuirlc l<>drocarbons
Fuu«.d Aroia.itlcti <2]6
1/0.2
10U/.J3
10/10
100/6.4
0.2
J3
ruflcd Arouallcu >21C
vO.2
<•¦0.2
Kctom-s (polycyclfc aron-atlc)
6.4
Heterocyclic ^ coupounda
100/6.4
1/1.5
7.9
Esrera
1/C.C6
J/0.06
0.06
CarLoi.vlK Ailds
0.06
-------
TABLL S&
ORGANIC EXTRACT SUMMARY TABLE
Straplt lll(' UD-2 txtrare, FerfpMjniianeHC
LCI
3.9
LC2
LCI
._i?5
870
LC4
LIS
LC0
LC7
£
Tattl Oigjntct. mg/m^
J. 0
36.
6,7
es
11
11
940
ICO, mg
2.25
VJl
<0.1
1.21-
IB.
6.2
10
21J
CRAV nig
.1,0...
1.0
90
12
94Q
Cttigory Int/mg/m'
Allpliatlc Hydrocarbons
100/3.9
3.2
1.5
Aromatic Hydrocarbons
10/3_-S_
^Kj/3.5..
.JQQ/iL
Kuscd Aromatlcs «2]b
100/372
11)0/372
_ 1QL17 .
¦wn
(used AromutlcD >216
39J
Heterocyclic S cnmpnurds
12
53
Hi-tprocM 1 Ir N cutnimunJa
_100/12_
IUG/J1.2.
Klfi/3,2
lsmiki
100/42
10/8.1
10/B.l
Nitrl Us
10/O.1
0.3
(' irlioxylli* m l«M
10/4.2
k.l
tsti ri
1/0.8
0.8
-------
TABLF 57
TOTAL ORGANICS (mg/m3) FOR SASS TRAIN SAMPLES OF
FERROMANGANESE PROCESS EFFLUENT
Compound Categories
AUphatic Hydrocarbons
Aromatic Hydrocarbons
Fused Aromatics <216
Fused Aromatics >216
Heterocyclic S
Heterocyclic N
Ketones
Alcohols
Nitriles
Esters
Carboxylic Acids
Particulates
>3y* <3p Rinses
-vO.l 1.9
^0.1 M).l
0.2 14
4.5 33 22
1.1 7.9 0.8
0.8 6.4 7.8
0.02
0.02
0.06 0.08
0.06
Sorbent Module Total
Resin Condens.*
3.9 6.9
3.5 3.7
370 380
390 0.3 450
37 37
70 0.07 80
53 0.05 67
~0.1
0.3 0.3
0.8 0.9
4.2 4.3
Concentrations estimated from IR and total TGO and Crav data only.
SOURCE: Reference 9.
125
-------
heterocyclic nitrogen and sulfur compounds, polycyclic aromatic
ketones, and trace amounts of nitriles and esters. The LC separations
between aromatic and polar species were very good.
The most abundant organic species present in the particulate ex-
tracts (Tables 54 and 55) were similar to those found in the XAD-2
extract (II X), i.e., fused aromatics in LC 3 and heterocyclic
nitrogen compounds and ketones in LC 6.
The extracts that had insufficient organic material for LC separations
and subsequent analysis were examined by infrared only. By combining
the IR data with the TCO and GRAV results, the organic materials in each
extract were very roughly categorized and approximate concentrations
estimated. These data, along with the data in Tables 54-56 were
integrated to constuct summary tables describing the concentration
distribution of compound categories from each SASS train (Table 57).
C. Detailed Level 1 Results
The LC, IR and LRMS results for the ferromanganese process effluent SASS
sample are presented in Tables 58-68 for the Probe Wash/Rinse extract
and in Tables 69-80 for the XAD-2 sorbent trap extract.
126
-------
TAIJLF 5B \£ REPORT; HJtK'INANGANtSt I'Kdlih VASfl SAMPl t
SAMPLE 11 PW, PK01SK UAMI, Ki.RK<*lA]iCAMSE
TCO
GRAV
ICO ~ GRAV
CofWMiratKKi
mg
Toltl mg
mjj/ L.o* kg)
Total Sampta'
11
51
17
Takan for LC^
25
2b
Rlwiltad^
12
12
Fraction
TCO h* mg
GRAV n mg
TCO +
GRAV
TouJ mg
Cencantrat»a
m^l
Found ui
Fraction
BUnk
Cor
racttd
ToU*
Found tn
Fraction
BUnk
Cor
ractad
ToW4
1
_I
;
2.65
MJ6
J'J
2^65
KB
1.^5
Uli
2
1
J'J
28
4
LI
0.71
9.87
11
7.84
S
0.7J
0.54
6
9.81
7.21
7
1.96
1.96
1.44
Swn
1. Quantity in «ntift dcUtminod btf of» LC
2 Portwo of wholt wmpl« tncd loi LC, actual mg
9 Quintily Mcovtrid Irom LC column, ictuil ing
4 Total mgc Tiputad back to tola) um^
5. Total 14; disltlvii \»y tola) volume
6. Not Ucltcftible
-------
TABLE 59 IR REPORT: FERRCMANGANESE PROBE WASH LCI. LC2
SAMPLE i: Ptf-I. LCI. Probo Va3h, Fefrooanganeae
Wm Numb*
IfftWHtty
As**gn#m#it
Comments i
(cm 1)
I
3000-2300 1 $ I iH. allpnatlc
isoo-1300 ' sn J ch. .inphicie
I
I
IB REPORT
(AMPLE II ?W-2. LC2. Probe Vaah. Fetroaansanese
Win Numfctf
(cm'1) j
Intvmdv
Acuflnmam
Cotwiwnti I
1
3000-2900 !
u
CH, aliphatic I
]
I
128
-------
SAMPLE
TABLE 60 IK REPORT: FERROMANGANESE PROBE WASH LC3, LC4
II PV-3, LC3, Probe Wash, Ferroaangariese
¥¥•*• Number
(cm"J)
Innnuty
AMiyumnt Comment*
3100-3000
S
CH, jronacic olefinlc
L930
OOCH-* allene
3000-2300
w
CH, allohatlc
1600-1000
numerous sharp bands
I I aromatic rlnz
900-700
S | aultiple bands, aron.itic
* substitution, fused rings
I
j
IP REPORT
SAMPLE H Ptf-4, LC4, Probe Wash, Ferroaanganese
Win Number
Ion"')
Intimity
AMtflnmtftt Comwnt»
3500-3200
V
SH or OH \
3100-3000
H
CH, aromatic, oleflnlc I
3000^2800
U
CH* aliphatic |
1600
*
C*C, aromatic ring \
1500-1000
M-«rf
multiple bands, aromatic ring j
900-700
S
aroraatic substitution
i ! 1
i i i
! 1 !
1 i |
I ! 1
1
i I
129
-------
TABLE 61 IR REPORT: FERRCHANGANESE PROBE WASH, LCS, LC6
sample II ?V-5. LCS. Probe Vash. Ferrooanganese
VSav« Numb«f I
(em"1) j
Intimity j AwqnwwM
Comiwma |
3500-J 200 '
M
Oil. SH
3100-3000 1
M
Crt, aroeatlc, oleflnlc ;
3000-2800
S
CH, aliphatic
2210 1
W
C-^, iltrlle
1730 1
M
C-0, eater
1700 1
s
C*0, ketone, acid, carbamate
i 1 lalde, cvcllc Lilde '
1650 >
v
OC I
1600. 1530 1
arosatlc substitution
1450. 1350 !
aaines ;
1520. 1350
M
nitro aronatic '
1280. 1120
*
aronatic ester
820, 750 i
S
aronatic substitution
'
1 1
^ PV-6, LC6, Probe Wash, Ferroaoniganese
Wi*« Number
(cm-1)
Intensity
Aunwmnt
-
Comments
3500*3200
V
o
X
3100-3000
M
CH, aronatic, oleflnlc
3000-2300
S
CH, allphitic
1720
s
C*0, ester
1700
S
00, acid, ketone
1660.
1620
s
aside, nitrite
1580.
1300
s
S-N*0^, nltraelne
1270,
1120
5
aronatic ester
1230,
1210
ester 1
1130,
1C60
6
alconol •
750
S
aromatic substitution, C-Cl {
,
'
130
-------
TABLE 62 IS REPORT: FERSOttCIGANESE PROBE VASH, LCT
*AMP« LI n-7. Praiie Vash. rerrocanmic3e
Wave Numb«f
(cm'M
Immily
Atugmnwvl Cumwmu
JiOO-llOO
S
OH or . broad
3100-3000
VI
CH, arcoatlc
3000-2300
SM
CH. aliphatic
3oco;-:;4oo
m
OK, acid, broad
1720
s
1*0, ester, ketone
1670.mO.160C s
ketone, amide, amldlae, nlcrlce
i ' nitrate
1250
s
ester, phosphate, ?*S cvclic
i 1 CFi, C-Cl. Si-rH^
10h0, 1000
alcohol
750
s
arocatlc subst, t»F, CF*> ~ C-Cl
1 I
, 1
, !
1
SAMPLE
Wj*« Numbtr
(em'11
IcttmrtY
Amgnment Commwits
! 1
131
-------
SAMPLE _
TABLE 63 UCtS REPORT: FERROtlANGANESE PROBE VASH, LCI
Monsanto Perro Alloy. II PV.T-1
Miioi C*t*90iin
lounwty I CftUgory | MW Ringa |
1 ! AllpKaclca ' 1200—50 i
I
J
Sub-Cfttvgor**, $p*ctfrc Compounds
Iniiniity
Cftttgory
m/« | Co(nocfi«Mi !
! i
1 '
i i i .
1 iii
! 1
i
1 1
| I
i
!
i 1
; I
1
l i
t
¦
i
i
,
J
1
i
i
1
: 1
;
[
|
1
i
1
i
1
i
Ofhtr
aliphatlcs with 200 ^ a/e ^450
132
-------
sample
TABLE bi LRMS REPORT: FERJIOMA.NCA.SESE PROBE WASH. LC2
Monatnto Farro Alloy. II PJ-2
Miiot CitKMn
lAMmnv
| C*t*«ory
MW fUnQt J
10
! aromatic hydrocarbons
225 1
l
1
1
' I
i I
bb-Cmgorm Sevafte Compounds
lnt»n«tv | C*u?onr
Co#npo«it>on j
1
1
1
1
.
! > ! i
j i l i
' - i
j
i
!
¦
¦
i
:
i
,
>
i
!
i i
|
:
1
i
,
i
!
i
i
1 !
1
1 j
1
! '
1
1 1
Ortwr
10 benzene substituted with allphatlr chain and N and CI |
exterior to ring with a/e - 225 ;H/*'iCl) ~j
I
133
-------
TABLE »5 LR-MS REPORT: rERROJttXGANESE P70BE HASH, LCJ
SAMPLE Monsanto Ferro Alloy, XI Ptf-3
Mi)6f Cit«9onil
Inmwty
Ctttgorv
MW 1Urxp |
100
fused alternace/r.on-alterr.ite hydrocarbons
'216 |
100
fused alternate/non-alternace hydrocarbons
>216 |
1
i
I t
i 1 !
1 '
Sub-Cttagorm. Sp*oftc Compounds
Inttmity
Cattgory
m/«
Competition |
100
pvrene/fluoranthcne
:o2
C ;H- 3 !
100
benzoantnraccnc, ate.
223
,
100
benzoovrcnes, etc.
252
C--ri1? 1
10
anthracene/pnenantSrene
173
C:uH,3 |
10
bertzotlttocene. etc.
216
C.-W.J |
10
213
c.th-.u ;
10
Tethvl-benzanthracene
242
CioHir, !
10
dibenzchrvsene
276
c--h.- :
10
dlbenzanthracene
278
10
dlbenieoyrene
-302
j
10 I
324
1
Other
1 polycvcllcs vlch 326 < o/e s4 50
134
-------
SAMPLE
TABLE 66 LRMS REPORT: FERROMANCANESE PROBE MASH, LC4
•Monsanto Ferro Alloy, II P'J-4
Mtioc C*t«9U«s
Inttnuty
I Grugory
MW Rang* |
100
| fusej alternae^/non-aicernace hvdrocarbons
<:i*> |
X
| fused aLt«r*uce/non-alt**rnate hvdroc^rbons
-J16 1
1
1 heterocyclic nitrogen cotspounos
«; '
• 1
1 1
! 1
1 !
Sut>-C«t»9oriM Specific Compound!
Innitnty
Gittgory
mh
I Competition
100
chrvsene/bcnzanthracene
"»"3
; C.1HW
100
benzopyrene. ecc.
:52
C?-H -
10
dlbenzchrvscne
' :?6
iO
dibenzanthracene
273
10
dlbensDvrenc
302
Cr-H-w
io ;
326 |
i
benzocarbazolc
217
Ci-H..N
l
242
C >H-u
1 1 ! 258
pyrene/fluorantbene
202
C-.cH" 1
1
! 376
1
'
i !
1 1
! 1 !
! 1 1
i 1
1
i i :
OltMT
<1 polvcciics belov 200
1 polycvclics 350 £ a/e <460
135
-------
SAMPLE
TXSIX 67 ULMS REPORT: FERROStANGAKESE VROBE WASH, LC6
Monsanto F«»rro MLoy, 11 y,/-6
Ma for CittffOfiH
Inivmirr
Cittgwr
MWRtnf 1
100
ketones
3004- 1
10
heterocyclic nitrogen compounds
:oc-ioo+ !
I
eaters
1
1
! "
! 1 ;
Sufe-Catttorm Spaafte Compounds
tntinatv
Cattywv
m/0
Composition 1
100
benzainnrone, etc.
230
O -M -.0 !
LO '• phthaiite 1 1 '
10
L
?30
1
|
1° I
202
i
I
10 1
203
10 I
243 1 I
10
.244
'
i
10
5 r'ng-N
253
C,r>K, \ i
10
5 ring-0 (polvcvcllc aromatic ketone)
254
C» 3 m flO |
10
253
10
dibonzacrldlne. etc.
279
Cr,Hf-S |
10 )
280
1
10 j
302
'
10
6-ring N ! 303
;
10
6-ring 0 (ketone on 6-fused rings)
304
i
1
fluorenone, etc.
IdO
CljHqC !
!
i
Olh«r
1 polvcyclic< vt^h 305 g a/g *-460
136
-------
sample-
table 60 LRMS REPORT. FERRtttANGANESE PROSE WASH. LC7
Moosanco Ferro Allov, II PW-7
Mifoc Csttgo»wt
Inunuty | Cat*«orv
MW R«n*
10 | ketones
200-300+
1 | heterocyclic nitrogen compounds
179-300+
1
, 1
1
Sub-Cit»90fi«*, Specific Compounds
Intimity
Cragory
m/«
Composition
10
benzanthrone, ecc.
230
Cl7H::0
1
a^ridine
179
c.
1 1
203
C.crt:N
: i
1
204
1 1
229
Ci7rt.jS
1 ! ! 253
C1l"
1 I
254
1 I
279
1 I { 230
1
6-rlng \
303
|
I 1 | 304
!
1
I 1
I i
,
1 !
|
j
i
1
1
!
! ! !
1 1 i
Other
1 polvcycllcs up co n/e 350
137
-------
TABLE 69 1C RtrOtiT: Kl KKitlANCANPSt XAD-2 SORRENT TRAP SAMPLK
SAMPLE
TCO
GHAV
TCO ~ GRAV
Conotn (ration
mg
To«l mg
mqt Jm'.L.or kg)
ToUl Sampla'
279
12 JO
1510
1110
Takart for LC2
la.b
7^1.2
92.0
Rfoonfnt'^
17.0
6H. 2
»5.2
Fraction
TCO in mg
GHAV m
mg
TCO ~
ConoratralKMi
Found in
Fraction
Blank
Co#
ractad
Tom4
Found hi
FfKlwo
BUnk
Cor
itdeJ
Tam4
GRAV
Told mg
mtl
(ro1. L.kfll
1
2.25
3.0
5 25
3. 86
2
9.45
SL)6
9.i5
6.95
3
m
16?
1062
781
4
1.21
AH
49
3b
6
6.15
3.0
9.15
6.73
6
29. 7
90
120
B8
1
11
12
23
17
Sum
1 Quantity in tnttre umpl*. d«t*«mmed before LC
2 Pwttort aI wliok urnplt intd tor IC, kIui) mg
3 Qiuntily ricuftitd liom LC (nlumn, Mu^l mg
4 Total mgcompuUd back to totd tampW
3. Tot.il mg divided l»y t«»U I voltiuu
6. Not detectable.
-------
TABLE 70 IR REPORT: FERROMA.NGANESE XAD-2 SAMPLE, LCI, LC2
xi X-l, LCI, XAD Extract, Ferro Vllov
Wm Number
icm~1 )
ItttMVitV
Aatgnimnt Coauiwnu
300o-:uoo
5
CH, allohadc
1500-1350
•«
CH, aliphatic
1
1
1
!
1
i i
i !
1
1
i
| 1
1
SAMPLE II X-2. LC2, XAJ Extract, Ferroranganeae
Wm Numbtr
(cm*1)
IntanMty
Amgrurant Comm«nti
3000-2800
....
s
CH, aliphatic
1500-1350
M
CH, aliohatic
1 •
i 1
' i
1
,
i
1
1
t
r
< I
1
139
-------
TABLE 71 IR REPORTS: rSR&OMANGANESE XAD-2 SAMPLE, LC3, 1C4
SAMPLE II K-1% LC3, KaD Extract, ferroaangaaeae
Wa*« Numbar
inttftuty
Aiugrwnant
Comments '
i
i
3100-30GO
s
CH» aromatic, olefiilc |
1600-1050
y,y
numerous sharp bands, aromatic
900-700
s
numerous sharp band3, aro-natic
I ringst fuaed rings
I i
¦ i
SAMPLE II X-6« LC*. XAP Extract, ferromanganese
W«v» Number
(cm"11
lnt*m«tv
Anttniram
Conmnrti
I
3400
5
VH or OH 1
3100-3000
—I
V
CHV arooatlc t
1600. 1500
M
aromatic rlns
i
1320(d)
i H
C-N
i
4
I2i0
M
aroraadc H rocking
|
800. 7 50
aromatic subst.
:
720
3^
fu9«d rlncs
i
! j I
i i !
I
|
'
1
x positi.e
identification ofi caroazoLe
1
140
-------
TABLE 72 IR REPORT: FERROMANCANESE \AD-2 SAMPLE, LC5, LC6
SAMPLE II X-5, LC5. XAJ) Extract, Ferrocanqanese
War« Numbar
Icfn*4)
r
intimity
AMtflfwmnt Conmrts
3400
NH or OH
3050
w
CH, jronatlc
2220
V
C-N, S-C-0
1700
V
OO
1600, 1450
u
sharp bands, aronatle rl-13
7j0, 700
X
aroaaclc subac.
!
I
i
1 i
i
1
1
f
1
II X-6, LC6, KAD Extract, Ferrraanganese
Win NumMf
[cm"1)
Inttimrr
Awegrowit Commtnu
3500-3100
u
OH or
3100-3000
w
CH, arooatlc, oleflnlc !
3000-1800
y
CH, aliphatic 1
1710
s
c-o !
1670
w
c-c 1
1600. 1500
V
aromatic rin^ 1
1450-1100
aromatic rint;, sharp bands 1
820
w
aromatic subst \
750, 720
s
aronatic subsc. 1
1
1
(
j
i
i
1
1
ui
-------
sample
TABLE 73 1R REPORT: FEJJtCWAJ.GA:,ESE KAD-2 SWffLE, LC7
II \-7, LC7. t-O Extract, Fcr r-M;angan«se
r/m Numbar
J
Inunaty
Awgmnwt
Cwmnbii
So significant
IR bands.
1
I
i I
T 1
SAMPLE
Vim Number
Intimity |
Annnwwt
C'^nmnti |
(cm*1*
1
1
j
142
-------
TABLE 74 LRMS REPORT; FERROMANCAS*ESE XAD-2 SAMPLE, LCI
SAMPLE H X-l, HAD—2 Extract, Ferromariganeae
Miior CftltyxiM
*
ImtfiitTy
Cmgory
MW Rtng* |
100
aliphatic hydrocarbons
200—4S0 f
i
I
I
I
1
1
1
1 1 !
|
Sub-Cttrfor.tt, So*cifrc Compounds
Inuraty
Catvgory
mj*
CompouDGA {
I
I
i
j
i
j |
1 !
i
I
! I
: -- i- -
1
l
i
i
i
Othtf
143
-------
iAMPLE.
TABLE 75 LRMS REPORT: FERROKANGASESE XAD-2 SAMPLE, LC2
II \-2, XAD-2 Extract, Fferrotnanqanese
Ml id* Cit*9ortf»
Intmity
| CingofT
MW R«n«a
10
1 al'&ylaced ?ol»c\cltcB rr aroonttcs
200-330
10
J fused arooatlC3 <216
202
1
1
1 1
!
Sub-CttrgotiM. ScMofic Compounds
Inttraty
Gat*gory
m/t
CompotittOfi
10
pyrene
202
C« -H, -
1
!
1
J
1
i
1
i
i
i
i
i i
i
1
I
i
!
i .
i
!
i i
Otfwr
1 i*u
-------
TABLE ?6 UMS SZPORT: FERRCMABCAKESE XAD-2 SAHPli, LC3
UftVU
It X-3, KA3-2 Extract. ferronaQ^anftse
Mi 10/ CiUiafW
tnttmry
Catigocy
UW Rmqi
100
fused aromatic* <216
152-210
10&
fused jr©nation >216
216-350
10
hecccocvclic S coupda..
i
!
1
Swb-C*t»goriti. So*t»fie Cameoonda
Intcmtry
Cattforv
m/t
Compoutron
100
•inchr4cene/?hsainthren«
i? a
CuH^
IG0
svfc .«f/flooranthene
202
Cs1!-
100
cir*rseii, benzdrcnrdccndfl
22B
C] »H--
100
bentopvrcnet peryleno
252
IJ
acenaphchjlone
152
C-:P,
10
fluorene
166
Ci:H„
10
methyl acentpKthalesit
1M
10
dlbenith loohene
18i
C] ;Ha 5
10
oechvl arthracene
192
.
Cl eHj *
10
benr&tluor^nes
•216
Ci-H,-
10
benionaohthaicne
218
C-rfU,
1
1
i
i
J 1 !
S I !
i 1 1
! ! 1
1 ! 1
Ot*m
10
P\H at a/e
230-J02 !
1
P\H at n/e
190-550 J
1
145
-------
TABLE 77 LRMS REPORT: FERROMANGANESE XAD-2 SAMPLE, LC4
II X-4, XAD-2 Extract, ferrooanganese
Mtioi CtwgoriM
Intantitv
r^ttfory
MW Ringt
100
heterocyclic N coopds.
167-267
100
fused aroraatlca 216
C28-302
10
fused aroaatlc* 216
178-202
1
I
f
Svb-Ciitfonn, Specific Compounds
Inumity
Category
r»/#
Composition
100
carbazole
167
C<-»9S
100
benzocarbazole
217
O cH" 1 N
100
benzop /renes
252
C;iH':
10
~nthraccne/ohen3nthrene
173
C; -Hio
10
scchvl carbazole
131
C.-MliN
10
dinetir/1 carbazole
1S5
¦
C1 -Hi:N
10
0''tCM
202
Ci5iin
10
benzoanthraceries, chrvsene
223
C, "-H'
10
aethvl bensocarbazole
231
C'-H'
10
dlbenzocarbazole
267
i
10
benzoperviene
276
C2:Hij !
10
aiethvl choiaathrene
263
C21H1S ;
10
dlberizochrvsenes
302
C.'iHm
' 1
1 1
1
1 '
1 i
i i ;
Oth *r
10 PAH at a/a 191, 241, 243, 245. 257, 253, 326 |
1 P^l at a/e 200 co over 400 j
146
-------
SAMPLE
TABLE 78 L3MS REPORT: FElROi'A.SGAKESE XAD-2 SA.HPLE, LCS
II X-5, XAD-2 Extrace, Ferronanganeso
Miiov CfW^orm
Inimity
Ct(*«Drv
MWRanta
100
heterocyclic N cotnpds.
167-251
100
ketones
230-280
10
nlcriles
153
!
i 1
1 1
Sub^*tt9or«M, $p*cifte CoN i
10
borrociroozold
i w
ci.m-.K j
10
•i-r tn^ N
, ::7
10
5-rln* M
, 253
Ci^Hl N
10
dlbetuoiluorerione
I 230
CrHl;0 ;
! i
1 1
1
I i
! 1 1
I l|!
! 1 !
1 ! J ;
1 ill
1 1 ! ;
OtlMf
10
PMl at m/e I
I
heteroc/cllc N, s/e 179-239 1
1
©r^an crwjs. a/e 1*0-3?0 1
147
-------
SAMPLE-
TABLE 79 LRMS REPORT: FERROMAKGANESS XAD-2 SAMPLE, LC6
II X-6, XAD-2 Extract, Ferromanganese
MifOf Cit*9W«»
Intaniitv
Cattiory
MWAngi
100
'leccrocvcllc N coapdj.
179-303
100
ketones
180-304
10
carboxvltc jo Ids
122
1
1
Specific Compounds
Inttnatv
C«t»9ory
m/«
f
Competition |
130
acrldine
179
C;iH-}K |
100
iluorenone
130
Ci'H^o !
100
••-ring heterocyclic V
203
CISH^N 1
100
4-rinij heterocyclic 0
204
C.5H?0 |
100
anchraquinollne
229
ClTHtlN |
100
oenzanthrore
230
Ci-H<.:0 j
10
benzoic acid
122
C'HsO: 1
10
methvl acrldine
193
1
CltHi'.N I
10
aechvle fluorenonc
194
Cl<.H",0 j
10
dlaethvl acrldine
207
Cl'Hl'N 1
10
anthraqui tone
203
CnHlOi j
10
beazocarbazole
217
ci,Ht:s
10
~ethvl anehraouinoline
243
CieHi-jN
10
5-rlnn S
255
10
6-ring N
303
t
10
dibenz acrldine
279
C;iHi3N |
1
t
1
1
1
OttMT
10
P\H at
n/e
219, 244, 254, 258, 230, 304 I
1
PAH at
n/e
265-380 !
!
148
-------
TA3LE 80 UttS REPORT: FERROP.A.NCANESE XAD-2 SAMPLE, LC7
SAMPLEII X-7, XA3-2 Extract, Feroaanganese
MijOf Citi^orM
Inttnittv
Cdtftfort
MW Rtnga 1
10
heterocyclic S ccnpds.
129-303 1
10
ketones
230-280 |
'
1
esters
136 1
1
!
i 1 :
Sut^Citi^ofiti, Swafic Compounds
Intensity
Cjt®90*v
m/a
Composition
10
acrid lne
179
Cl -H-»N
10
4-rin? N
203
c. :;i
1
10
benzocarbazole
217
C* Hi 1 V
1C
bcnzophcnancnrldene
229
C-7H1.N
10
benzanthrone
230
Cl-H'30
10
5-rin? hetcrocvclic N
253
ci qhi : n
10
fr-rlnij hecerocvcltc N
279
' C:'H'3S
10
dibenzofluorenene
230
C:02 1
1
mechvl acrldinc
193
C:uHl!N
1 1 !
i
i i
i 1
i
1 !
1
1 III
¦ ! i I
I !
Other
1 PAH at a/e 200-329
I
149
-------
XII. REFERENCES
(1) Hamerema, J. W., S. L. Reynolds and R. F. Maddalone, "IERL-RTP
Procedures Manual: Level 1 Environmental Assessment," EPA-600/
2-76-160a. NTIS No. PB 257-850/AS, June 1976.
(2) Adams, J. W., T. E. Doerfler and C. H. Summers, "Effect of
Handling Procedures on Sample Quality," EPA-600/7-7L-017, NTIS
No. PB 279-910/AS, February 1978.
(3) Thrun, K. E., J. C. Harris and K. Beltis, "Gas Sample Storage,"
EPA-600/7-79-095, NTIS No. PB 298-350/AS, April 1979.
(4) Lentzen, D. E., D. E. Wagoner, E. D. Estes and W. F. Gutknecht,
"IERL-RTP Procedures Manual: Level 1 Environmental Assessment
(Second Edition)," EPA-600/7-78-201, NTIS No. PB 293-795.
October 1978.
(5) Leo, A., C. Hansch and D. Elkins, "Partition Coefficients and
Their Uses," Chem. Rev. 71, 525-621 (1971). See also: Hansch, C.
and A. J. Leo, Substituent Constants for Correlation Analysis in
Chemistry and Biology, John Wiley, New York, 1979.
(6) Cleland, J. G. and G. L.Kingsbury, "Multimedia Environmental
Goals for Environmental Assessmert." Volume I. EPA-600/7-77-
136a, NTIS No. PB 276-919/AS, November 1977.
(7) Herther, M. A. and L. R. Waterland, "SAM LA: A Rapid Screening
Method for Environmental Assessment of For.sil Lnergy Process
Effluents," Draft Report en EPA Contract No. 68-02-3176, Acurex
Technical Report No. TR-77-50C, December 1980.
(8) Davidscn, L. N., W. J. Lyman, D. Shooter and J. R. Valentine,
"Technical Manual for the Analysis of Fuels," EPA-600/7-77-143,
NTIS No. PB 279-196/AS, December 1977.
(9) Rudolph, J. L., J. C. Harris, Z. A. Grosser and P. L. Levins,
"Ferroalloy Process Emission* Measurement," EPA-600/2-79-045,
NTIS No. PB 293-171/AS , February 1979.
150
-------
APPENDIX A
LEVEL 1 ORGANIC ANALYSIS TECHNIQUES
This material is reproduced exactly as it
appears in the second edition of the Leve] 1
procedures manual (Reference 4, p. 150 of
this document). Figure, Table and Reference
numbers within this Appendix are those of
Lentz »n et. .al.* and do not correspond '.o those
in the body of this report.
A-l
-------
CHAPTER 9
LEVEL 1 ORGANIC ANALYSIS TECHNIQUES
9.1 INTRODUCTION
The objective of Level 1 organic analysis is to identify the major
classes of organic compounds present in a process or effluent stream and to
estimate their concentrations. An example of the kind of information the
methodology is designed to provide is givun in Table 14.
Samples obtained in accordance with the procedures outlined in Chapters
3 through 7 will be either gases, liquids, or solids. The multimedia analysis
flow scheme presented in Figure 29 shows how each of the sample types is
split for organic analysis.
In Level 1 organic analysis, quantitative information is provided by
gas chromatography (total chrci?jtographsble orrjdnics--TCO) and by gravimetry
(GRAV). Qualitative and scniquantitetive inf^r .'dtion is obtained from
liquid chromatography fractionation, from infrared spectra, and from lew
resolution mass spectra. In order to achieve a satisfactory characterization
of the sample, the analyst must integrate all of these data, as well as any
other available information about the source. Knowledge gained from any one
part of the analysis scheme (e.g., LC separation) should be used, to the
maximum extent possible, in interpreting the results from other parts (e.g.,
IR or LRMS spectrum). Table 15 summarizes the data expectea from individual
samples undergoing Level 1 organic analysis.
9.2 LEVEL 1 ORGANIC ANALYSIS METHODOLOGY (ref. 95)
An overview of the methodology to be used for the Level 1 organic
analysis is shown in Figure 30. This methodology de<"\ls with the preparation
of the samples to provide a form suitable for analysis, and with their
subsequent analysis.
A-2
-------
TABLE 14. SUMMARY OF RESULTS FOR ORGANIC EXTRACTS FOR SASS TRAIN SAMPLE
, mg/m3
Particulate nodule Sorbgnt module
Categories Rinses* >3 <3 pm Resin Rinse Condensate! Totalf
Aliphatic hydrocarbons
Aromatic hydrocarbons—
benzenes
Fused aromatics, MW <216
Fused aromatics, MW >216
Heterocyclic N
Heterocyclic S
Heterocyclic 0
Phenols
Esters
Carboxylic acids
Sulfur
Inorganics
Unclassi fied
Si 1icones
<0.06
0.25
0.25
0.31
<0.06
<0.06
0.06
0.18
<0.06
<0.04
0.15
0.15
0.19
<0.04
<0.04
0.04
0.11
<0.04
0.06
0.06
<0.04
C 04
0.3
0.6
6.3
4.2
0.6
0.4
0.2
0.1
0.1
0.3
0.1
0.2
0.8
22
21
19
2
2
0.1
0.3
0.2
1.1
0.6
29.
26.
20.
2.4
2.2
0.2
0.5
0.6
0.2
0.1
0.3
0.1
*Rinses corresponded to 0.03 mg/m3 of organics and were not subjected to LC-IR-LRMS analysis.
tNo condensate was collected for this sample.
TRounded results.
-------
f torju Mtfil
¦AAAiYSiJ INOS MIRE IF RCSIDUC < lOflRCENT OF TOlAl PAIITICU1 ATE CA1CH
FiijUio 29. UuIiiii.hIu onj line an.ilym overview.
-------
TABLE 15. SUGARY OF EX?ECTE3 DATA FROM LEVEL 1 ORGANIC ANALYSIS
Sample
OnsIte
GC Weigh Extract TCO GRAV IR LC"
Gases'-grab sample
SASS
>10 msi particulate
3-10 pis particulate
1-3 pm particulate
<1 pi* particulate
Rinse of particulate
Todules and probe
XAO-2 resin conbired
with rfrse of
sorfcent nodu'e
Sorbent module
condensate
S0LIS3
Flyash, clinker
Organic feed stock
Coal
LIQUIDS
Effluent water
Organic feed stock
Fuels
v
v
Vt
V
V
V
V
V
V
V
V
V
J
V
V
t Vt
•/
J
V
V
V
"Includes CRAV ~ IR and TCO on all fractions; perforo LRMS when criteria
are exceeded.
tOo analysis If gravlretrlc results are greater than 10 percent of the total
particulate catch
A-5
-------
' r
TCO"
Analym
Gravimetric
Analym
Repeat TCO*
Analym
i< Neceuary
Aliquot Combining
15 100 mg
Infrared Analym
Concentrate
Extract
Liquid
Chromatographic
Separation
Low Retolution
Mju Spectra
Analym
Seven Fractioni
Organic Extract
or
Neat Orq^mc Liquid
Infrared Analym
*Se« lection 9 4 1
Figure 30. Organic analysis methodology.
A-6
-------
As indicated in Figure 29, the extent of the sample preparation required
varies with sample type. The low molecular weight, volatile species (boiling
point <100° C) are determined by gas chromatography onsite and require no
preparation. Organic liquids, such as fuel oils, will not need pretreatment
and are placed directly into the analysis scheme. However, the majority of
the samples, including the SASS train components, aqueous solutions such as
scrubber waters, and bulk solids such as coal or slag, require extraction
with solvent prior to analysis. This extraction separates the organic
portion of the samples from the inorganic species. The analysis of organic
extracts or organic liquids then proceeds to initial quantitative analyses
of volatile (TCO) and nonvolatile (GRAV) organic material and a preliminary
infrared (IR) spectral analysis. The IR spectrum provides an indication of
the types of functional groups present in the sample and a control check-
point for subsequent analyses. All functional groups identified in this
total sample should be accounted for in the succeeding steps.
The sample extract or organic liquid is separated by silica gel liquid
chromatography (LC) using a 7-fraction solvent series of varying polarity.
TCO and gravimetric analyses of each fraction are done to determine the
distribution of the sar.ple by the various class types. An IR spectrum is
then obtained on each LC fraction for determination of the types of func-
tional groups present. Low resolution mass spectra (LRMS) are also obtained
on all fractions that exceed the concentration threshold in order to deter-
mine the principal compound types present in each fraction. For, the sample
streams identified in the Level 1 scheme, these threshold concentrations
are:
a. Gas streams sampled with the SASS system--0.5 mg/m3 computed at
the source;
b. Aqueous slurry or solid samples—dependent on extract concentra-
tion;
c. Organic liquids--l mg/LC fraction.
The decision is based on the sum of the TCO and GRAV analyses (Section 9.4)
for each fraction.
It should be emphasized that sample contamination and solvent impur-
ities are common problems in orqanic analysis. The best possible laboratory
A-7
-------
procedures must be used along with verified pure solvents. Blanks and
controls are to be run for each stage in the analysis scheme, as specified
in Chapter 2.
9.3 PREPARATION OF SAMPLE EXTRACTS
This section presents sample preparation procedures that are appropriate
for most samples. The specific solvent indicated for the extraction is
methylene chloride, which was selected because of its good solvent properties
and high volatility (to facilitate concentration). This solvent should be
used except in cases with unusual requirements, for which an alternative
procedure can be suggested and used after approval by the project officer
and by the Process Measurements Branch, IERL-RTP.
Procedures for concentration and analysis of extracts are presented in
Section 9.4.
9.3.1 Aqueous Solutions
Extraction of aqueous solutions should be carried out with methylene
chloride using a standard separatory funnel fitted with a Teflon stopcock.
The pH of the aqueous phase should be adjusted first to 2.0 AO.5 with hydro-
M'
chloric acid and subsequently to 12.0 ±0.5with sodium hydroxide, using
multirange pU paper for indication. Two extractions are to be done at each
pH, using a 250-mL volume of methylene chloride for each of the four extrac-
tions of a JO-L sample. The extractions may be perform d in several batches
on convenient-sized sample portions with corresponding amounts of solvent,
but the entire 10-L sample must be extracted.
For the SASS train sorbent module condensate, the volume of aqueous
solution should be measured and the quantity of methylene chloride adjusted
proportionately.
To avoid the necessity of shippinq large quantities of water, the
extractions should preferably he done onsite whenever facilities will permit
contamination-free conditions. If formation of euulsions is encountered,
the samples may be shipped to the laboratory for extraction. Centrifugation
at about 2,000 rpm 1ms been found to be an effective way to b»eak the emul-
sion in several studies.
-------
9.3.2 Solid*-., Particulate Matter, and Ash
All solid material including waste products, raw n.ateridls, cyclone,
probe and filter particulate, and ash are extracted for 24 h with methylene
chloride in a Soxhlet apparatus. When sample quantities are limited, such
as in the case of SASS cyclone catches, the designated solid sample aliquots,
remaining after a small portion has been set aside for inorganic and particle
morphology analysis, should be taken for the extraction. (See Table 11,
Chapter 4.) When kilogram quantities of the sample material are available,
optimal sample sizes should be used for these determinations. Bulk lumpy
solids, such as coal, should be crushed to a size that will pass a 60-mesh
screen before extraction, using a procedure such as that in ASTM D 2013,
"Preparing Coal Samples for Analysis" (ref. 74). The sample is held in the
thimble with a plug of glass wool and a stainless steel screen during the
extraction to avoid carryover of the sample.
9.3.3 Slurries and Sludges
The sludge/slurry sample category can span a tremendous range, in-
cluding slurries and solid or semisolid sludges containing up to 95 percent
water. Some of these materials are very difficult to handle and no one
procedure will work for all of them. The basic Level 1 approach is to
determine whether the sample is best treated as a solid, a liquid, or by a
combination of procedures. The protocol will involve, in most cases, tests,
on small portions of the 1-kg sample to determine the best procedure prior
to committing the entire sample. The stepwise protocol follows.
If the physical character of the sample permits, treat
it as a solid by transferring the whole sample to a Soxhlet
thimble and extracting for 2-1 h with methylene chloride. Do
not dry saciple before extracting. Determine wet weight of
sample taken by weighinq sample container before and after
transferring sample to thimble. If an aqueous phase is
noticed in the organic extract, this should be separated and
removed prior to concentration.
If the sample state does not appear compatible with
direct Soxhlet extraction (i.e., wet sludge/slurry of high
liquid content), select a treatment procedure as follows:
1. Take a 10-rnL portion of tin; sample (shaking vigorously
first, if necessary, to facilitate a fuirly ropresenta-
A-9
-------
tive sampling) and place in a 15-mL centrifuge tube.
Add 2 mL methylene chloride. Shake well and allow to
settle for at least 30 min. Then:
If sample dissolves completely, treat the sample
like a neat organic liquid—no sample preparation
is required.
If a clean two-phase (orqanic-aqueous) separation
is achieved, treat the sample like an aqueous
sample.
If a clean two-phase separation is not achieved
(I.e., if an emulsion forms, or if the apparent
solvent recovery is low, or if a three-phase system
with solids suspended in organic layer or between
organic and aqueous layers is present), centrifuge
the mixture. If a clean two-phase system then
results, treat sample as in l.b. above. If not,
test sample as suggested in 2., below.
2. Take a 10-mL portion of the sample (shaking vigorously
first, if necessary, to facilitate a fairly representa-
tive sampling) and place in a 15-mL centrifuge tube.
Centrifuge. If phases scoarate, treat the solid phase
by Soxhlet extraction. 1 tie liquid phase is treated like
an aqueous sample o:\ if organic, liko a neat organic
liquid. The several extracts generated for this typo of
sample should be recombincd'-taking the same fraction of
each—prior to organic analysis.
Occasionally a sludge/slurry sample may be encountered for which none
of the above methods will be satisfactory. In those instances, the EPA
project officer and appropriate. PMB personnel-should be consulted for guid-
ance.
9.3.4 SASS Train Rinses
For each SASS train run there are two samples of this type, one from
the rinse of the particulate portions (cyclones and filters), and a second
from the rinse of the sorbent module. The solvent mixture used for the
particulate rinses is 1:1 (v:v) methylene chloride:methanol. This rinse
should be dried and weighed. If the residue quantity is greater than 10
percent of total filter and cyclone catch, the full cc.nplcment of analytical
tests should be performed. (See Figure 29.) The rinse for the gas condi-
tioner and sorbent module is methylene chloride alone. It should be added
to the-Soxhlet solvent reservoir prior to XAD-2 extraction. This will
b.
A-10
-------
result in a combined rinse and extract characterization for organic com-
pounds. (See Fiqure 29.)
9.3.5 Sorbent Trap
The XAD-2 resin from the sorbent trap is removed from the SASS train
cartridge and homogenized, and a 5-g portion is removed for the inorganic
analysis. The balance of the resin is extracted with methylene chloride to
remove the organic material. A large Soxhlet extraction apparatus, avail-
able from several manufacturers*, must be used to extract the 400 mL of
resin. The resin is transferred to a previously cleaned glass extraction
thimblet- A glass wool plug and stainless steel screen are used to secure
the resin, which would otherwise float on the methylene chloride. Approxi-
mately 1 L of methylene chloride is added to the 2-L reflux flask. (The
dumping volume of an appropriate commercial extractor is 750 mL.) A larger
Soxhlet extractor may be used if available and an appropriate increase in
the reflux solvent volume made. The boiling solvent in the flask should be
examined periodically because additional methylene chloride may be needed to
replace that lost by wetting the resin and by volatilization. The resin is
extracted for 24 h. If water is extracted fron the XAD-2 resin during this
procedure, as evidenced by two phases in the liquid portion, segregate them
with a separatory funnel before concentrating the methylene chloride. The
aqueous fraction from this separation is added to the condensate catch
before it is extracted at the two pH levels.
9.4 ANALYSIS OF SAMPLES FOR ORGANICS
The analysis of each of the prepared or isolated samples for organic
compounds follows the scheme introduced in Figure 30. The overall scheme is
based upon an initially recoTimended scheme (ref. 95), which has been revised
with information from subsequent laboratory evaluations (ref. 96).
Qualitative analyses of organic compounds are accomplished by the
LC/IR/LRHS procedure, which will provide reliable data on the compound types
*For example, Ace Glass Incorporated (catalog Ho. 6810-10) or Lab Glass
(catalog No. LG-6910-100).
tThimbles of borusilicate glass with fritted cilass dices must be specially
fabricated for this size Soxhlet extractor. Do not use cellulose thimbles.
A-Jl
-------
present 1n the sample or organic extract. However, compounds with boiling
points below about 100° C will not be captured in the SASS train or retained
In the organic extracts. Consequently, a separate field gas chromatography
procedure has been Included for the analysis of this range of materials (see
Chapter 3), which gives a lfmited amount of qualitative Information (reten-
tion times) as well as quantitative information.
Quantitative analysis of moderately volatile materials (bp 100° C to
300° C) is achieved by a gas chromatographic procedure applied to various
organic solvent extracts, organic liquids, and SASS sorbent module rinses.
This Total Chromatographable Organics (TC0, Section 9.4.1) analysis is not
appropriate for extracts from samples that do not contain low boiling or-
ganics, such as SASS particulate material collected at 200° C. Quantitative
analysis of nonvolatile organic sample components (bp >ca. 300° C) in all
extracts is achieved by evaporating an aliquot of extract to dryness and
weighing the residue (GRAV procedure).
In summary, TC0 analyses of extracts and organic liquids are performed
prior to any concentration step. It is then necessary to obtain an IR on a
portion of this material, to do a gravimetric analysis on an aliquot, and
to concentrate the extract for the LC separation. The appropriate stage to
conduct each of these steps (gravimetric analysis, IR, concentrate) will
depend on the quantity and solubility of the sample as described in the
following sections. For many samples, quantitative analyses (TC0 and/or
GRAV) may be required both before and after concentration.
9.4.1 Total Chromatographable Organics (TCP) Analysis
As previously stated, the TC0 analysis is necessary for quantification
of materials with boiling points in the range of 100° to 300° C. This
analysis is applied to all samples that might contain compounds in this
volatility range. These include organic liquids, many solid sample extracts,
aqueous sample extracts, extracts from the SASS train sorbent module samples,
and LC fractions obtained for those samples. However, particulate samples
collected at the 'pecified 205° C SASS train oven terrperature or residues
from high temperature processes do not require TCO analysis. If, for some
.special circumstance, the front half of the sampling train is run at a
A-12
-------
temperature lower than 200° C, then both the TCO and the gravimetric proce-
dures should be applied to the extracts of cyclone particulate catch.
Because materials in the TCO volatility ran.je may be lost to varying
degrees during solvent evaporation, It 1s Important that this analysis be
performed on extracts and solutions prior to any concentration stop. It
will also frequently be desirable to repeat the TCO analysis later on the
concentrated extract. When the original analysis shows a low TCO value,
corresponding to a concentration of less than about 40 mg/L in the extract,
the TCO analysis should be repeated on a concentrated extract. This will
give a more reliable estimate than that obtained by multiplying the original
low concentration estimate by the large volume of unconcentrated extract.
For determination of TCO, a 1- to 5-|jL portion of the extract is ana-
lyzed by GC using a flame ionization detector. A 1.8 m x 3 mm O.D. (6 ft x
1/0 in.) column of 10 percent 0V-101 on 100/120 mesh Supelcoport has been
used successfully for this analysis. Other silicone phases (0V-1, etc.) may
work as well, but a 10 percent loading is recommended. The GC is operated
Isothermally at about 30° C--or room temperature—for 6 min after sample
Injection and then programed a', approximately 20° C/min to 250° C and held
at 250° C as long as necessary for complete elution of sample. Injector
temperature of 275° C and detector temperature of 300° C are appropriate.
Slight modifications in the temperature and duration of the Initial hold
period may be necessary tc accommodate variation? 1n individual GC systems.
Quantitative calibration of the TCO procedure is accomplished by use of
mixtures of known concentrations of the normal hydrocarbons C8, Ci2> and
C16. The quantitative calibration standards should bo prepared to cover the
concentration range to be studied. Retention time limits corresponding to
the TCO range of boiling points are defined by the peak maxima for n-heptane
(C7, bp 90° C) and n-heptadecane (Ci7, bp 303° C). Therefore, integration
of detector response should begin at the retention time of C7 and terminate
at the retention time of CJ7. By this procedure, the integrated area will
cover r.-aterial in the boiling range of 100° C to 300° C. (The C7 and C17
peaks should not be included in the quantitative calibration.)
In the TCO analyses, it is important that the observed values of total
integVated area for samples be corrected by subtracting an appropriate
A-13
-------
solvent blank prepared (i.e., concentrrted) in tho same manner as the sam-
ples. lhe blank should be checked on each of the seven LC solvent mixtures.
The results of each TCO analysis should be reported as one number, in
milligrams, corresponding to the total quantity of material in the 100° to
300° C boiling ran^e in the original sample collected (see Figure 31). The
TCO data are thus analogous to the results obtained by gravimetric analysis.
The chromatograms themselves contain some additional data beyond the TCO
values (i.e., retention times and areas of individual peaks) and should be
retained until the Level 1 sampling and analysis effort has been completed.
9.4.2 Gravimetric (GRAV) Analysis
The gravimetric analysis is used for quantification of organic sample
components with boiling points higher than 300° C. This analysis should be
done after the sample extract has been concentrated, since it is recommended
to weigh at least 10 mg of sample in a gravimetric analysis, when possible.
Weighing to a precision of ±0.1 mg is adequate for purposes of Level 1
analysis. Sample and tare weights should be obtained by drying to "constant
weight" (.tO.l mg) in a desiccator over silica gel or Drierite. In perform-
ing a gravimetric analysis on a large volume sample (i.e., >50 mL), no more
than 5 mL of extract should be evaporated to dryness. For extracts concen-
trated to 10 mL, a 1-mL aliquot is taken for GRAV analysis. The GRAV results
should be reported as one number for the entire sample (see Figure 31). The
infrared analyses calied for in Level 1 organic analysis can be performed on
the residues from the GRAV procedure, provided that the weighing dishes were
successively rinsed with distilled water, methanol, and methylene chloride
before use.
9.4.3 Concentration of Extracts and Solvent Fxchnnne Procedure
After the initial TCO* analysis, it will usually be necessary to concen-
trate the organic extracts to a volume of 10 mL for subseauent analysis. It
is recommended that concentration to slightly less than 10 mL volume (i.e.,
8 or 9 mL) be accomplished using a Kuderna-Danish apparatus with a 3-ball
Snyder column for volumes less than 1 L, and a rotary evaporator for volumes
*0n all snnples except SASS particulate fractions and residues from high
temperaturo processes.
A-14
-------
that initially exceed 1 L. It is essential that the extract not be reduced
to dryness at this point in the scheme to prevent loss of TCO range material.
The concentrated extract should then be transferred to a convenient graduated
container (e.g., Kuderna-Danish receiver or centrifuge tube) and the volume
restored to 10 mL. The concentration process should be stopped if material
begins to drop out of solution. In that case, the extract should be restored
to a convenient volume in which the material is redissolved. The original
TCO* analysis may provide guidance as to the degree of concentration that is
required for each pa-ticular sample. The objective is to achieve a final
concentration of not more than 100 mg/mL and a final volume of not less than
10 mL.
At this stage in the analysis sequence, a GRAV determination is done
and the TCO* determination is repeated.! A 1-mL aliquot is used for the
GRAV analysis and a 5-pL aliquot is used for the TCO. If the sum of TCO
plus GRAV is <15 mg for the total sairple, the LC separation is not performed
and the Level 1 analysis is concluded by obtaining IR and LRMS spectra on the
sample. If the sum of TCO plus GRAV is >15 mg, the LC separation is per-
formed. An IR spr.ctrum is also obtained on the residue from the GRAV analy-
sis or on a separate aliquot of extract. A portion of the concentrated
extract that contains about 100 rag of organic material, if possible, is
taken for the LC. Smaller quantities down to a lower limit of 15 mg may be
used if necessary.
The LC separation procedure requires that both methylene chloride
solvent and water be eliminated before the sample extract is applied to the
silica gel column. Otherwise, the required aliphatic/aromatic and subse-
quent compound class separations will not be achieved. Extracts that do not
contain low boiling organics, such as SASS train particulate materials or
other materials collected at temperatures exceeding 200° C (400° F), can be
evaporated to dryness with silica gel, as detailed below, before LC analysis.
*0n all samples except SASS particulate fractions and residues from high
temperature processes.
"tUnloss the initial TCO values for the unconcentrated extract exceeded the
guideline of 40 mg/L (TCO).
A-15
-------
Extracts containing appreciable quantities of TCO material (>2 mg) must be
transferred to the LC columns without being evaporated to dryness. For
these samples, a solvent exchange procedure is required to minimize losses
of volatile sample components. An aliquot of the methylene chloride solu-
tion containing 15 mg (minimum) to 100 mg (preferred) of sample is taken for
solvent exchange into cyclo|.».-ntano; the detailed procedure is given below.
Normal hydrocarbon solvents are not to be substituted for the cyclopentane;
pentane boils too low (36° C, below methylene chloride) and liexane boils too
high (68° C) for the solvent exchange. The solvent exchange and removal of
v/ater from the sample extracts is accomplished as described in Section
9.4.4.3.
9.4.4 Liquid Chromatographic (LC) Separation
All sample extracts, neat organic liquids, and SASS-traiu-dried probe/
cyclone rinse extracts are subjected to LC separation if sample quantity is
adequate. An aliquot of the concentrated extract containing 100 mg of
organic matter is preferred for tt.e LC, but smaller quantities down to a
lower limit of about 15 mg may be used. The sample components are separated
according to polarity on silica gel using a stop gradient elution technique.
The detailed procedure for the LC separation is given below:
Column: 200 mm x 10.5 nvn ID, glass with Teflon stopcock, water-
jacketed with inlet water temperature in the range of
18° to 22° C and sufficient flow to maintain this tempera-
ture through to the outlet.
Adsorbent: Oavison, Silica Gel, 60-200 mesh, Grade 950 (available
from Fisher Scientific Company) is to be used; no other
types or grades of silir^ o-'l can he subst ltutcd. This
material should be cleaned prior to use by sequential
Soxhlet extractions with methanol, methylene chloride,
and penlane. This adsorbent is then activated at 110° C
for at least 2 h just prior to use, and cooled in a
desiccator.
A-16
-------
Drying Agent: Sodium Sulfate (Anhydrous, Reagent Grade). Clean by
sequential Soxhlet extraction for 24 h each with methanol,
methylene chloride, and pentane. Dry for at least 2 h
at 110° C just prior to use and cool in a desiccator.
9.4.4.1 Procedure for Column Preparation--
The chromatographic column, plugged at one end with a small portion of
glass wool, should be slurry packed with G.O g of freshly activated silica
gel in n-pentane. A portion of properly activated silica gel weighing 6.0
+0.2 g occupies 9 mL in a 10-mL graduated cylinder. The total height of the
silica bed in this packed column is 10 cm. The solvent void volume of the
column is 2 to 4 mL. When the column is fully prepared, allow the pentane
level in the column to drop to the top of the silica bed so that the sample
can be loaded for subsequent chromatographic olution.
After packing the silica gel column, add 3 g ±0.2 g clean sodium sul-
fate tc the top of the column. Vibrate for 1 min to compact. The sodium
sulfate should occupy 2 mL in a 10-mL graduated cylinder. The sodium sul-
fate will remove small quantities of water from the organic extract; how-
ever, appreciable quantities of water will solidify the sodium sulfate,
inhibiting proper flow through the column. Therefore, it is advisable that
if enough water is present in the sample to form two layers, it should be
removed by another method--pipette or separatory funnel.
9.4.4.2 Evaporation of Sample Extracts with Low TC0 (<2 mg original
sample)--
For these samples, the aliquot of extract containing 15 mg (minimum) to
100 mg (prpferred) of material is added to a small amount of silica gel, the
solvent is allowed to evaporate, and the residue plus silica gel is trans-
ferred to the LC column with the aid of a microspatula. The container is
rinsed as described in Section 9.4.4.5.
9.4.4.3 Solvent Exchange of Sample Extract with High TC0 (>2 rcq original
sample)--
An aliquot of methylene chloride extract containing 15 mg (minimuni) to
100 mg (preferred) of material is added to 200 mg of silica gel in a grad-
uated rereiver. The volume of extract is carefully reduced to 1 ml. at
ambient temperature under a gentle stream of nitrogen (tapped from a liquid
A-17
-------
nitrogen cylinder, if possible, to minimize impurities). The solvent evap-
porates rapidly so it is important that this operation he done under con-
stant surveillance to insure that the volume is not reduced below 1 mL. It
is also necessary to warm the samples slightly, either by hand or water
bath, at <40° C, to prevent condensation of atmospheric moisture in the
sample due to evaporative cooling. One mi 1 li 1 i r.er of cyclopentane is added
and mixed by gentle agitation. The volume is reduced to a total of 1 mL as
before. A second milliliter of cyclopentane is added, mixed, and the volume
is again reduced to 1 ml. The exchange is repeated with a third milliliter
of cyclopentane. After ttie volume has been reduced to 1 mL for this last
time, the solvent mixture will be <5 percent methylene chloride. This is
sufficiently low to prevent breakthrough of aromatic sample components into
the aliphatic hydrocarbon fraction, LCI.
The cyclopentane and silica gel are transferred to the top of the
previously prepared LC column using a Pasteur pipette. The container is
rinsed as described in Section 9.4.4.5.
9.4.4.4 Neat Organic Liquids--
A lOO-r-g sample is weighed into a tared glass weighing funnel and mixed
with about 200 mg of silica gel using a microspatula. The sample is then
transferred to the top of the colunu:. The container is rinsed as described
in Section 9.4.4.5.
When neat organic liqvids are fractionated by the liquid chromatography
scheme, they have the same theoretical gravimetric detection limitations as
other samples separated by this means, 0.1 mg/100 mg or 0.1 percent of the
sample applied. Since these aliquots are neat samples and do not have
concentration factors as multipliers, the resultant detection limits for
minor components are 1 g/kg at best.
9.4.4.5 Chromatographic Separation into Seven Fractions—
Table 10 shows the sequence for the chromatographic elution. In ort-'er
to insure adequate resolution and reproducibility, the column elution rate
is maintained at 1 mL/miit.
A-18
-------
TABLE 16. LIQUID CHROMATOGRAPHY ELUTION SEQUENCE
Fraction
Solvent composition
Volume (inL)
1
2
3
4
5
6
7
20% Methylene chloride in pentane
50% Methylene chloride in pentane
Methylene chloride
5% Methanol in methylene chloride
20% Methanol in methylene chloride
50% Methanol in methylene chloride
Pentane
25
10
10
10
10
10
10
The volume of solvents shown in Table 16 represents the solvent volume
added to the column for that fraction. If the volume of solvent collected
is less than the volume actually added due to evaporation, restore the
fraction volume to the proper level with fresh solvent. In all cases, the
solvent level in the column should be at the top of the gel bed, i.e., the
sample-containing zone, at the end of the collection of any sample fraction.
The fractions are retained as solutions for TC0 analyses.
After the first fraction is collected, rinse the original sample con-
tainer or weighing funnel with a few milliliters of Fraction 2 solvent (20
percent methylene chloride/pentane) and carefully transfer this rinsing into
the column. ' Repeat with each successive solvent mixture in turn.
Add each new solvent to the column slowly to minimize disturbing the
gel bed and eliminate the trapped air bubbles, particularly in the zone of
the sample-containing silica gel.
After each sample is collected, an aliquot (1 to 5 pL) is taken for TC0
analysis of each fraction (unless the sairple taken for LC had a TC0 of <2
mg). Also, an aliquot (10 ml for Fraction 1 and 5 mL for Fractions 2-7) is
transferred to a tared aluminum micro weighing dish for evaporation and
gravimetric analysis. The GRAV data for Fraction 7 must be corrected for a
blank contributed by a snail quantity of silica gel that dissolves in the
highly polar eluent. The blank value is determined by running an LC column
to which no sample is addfd; it is on the order of 0.9 ±0.1 mg in LC7 (10 mL).
A-19
-------
After TCO and GRAV determinations, the fractions are analyzed by IR and,
when the quantity is sufficient, by LRMS (see Section 9.4.6).
The objective of the LC procedure is to separate the sample into frac-
tions of varying chemical class type to facilitate subsequent analyses. The
"LC separation procedure is not a high resolution technique and, consequently,
there is overlap in class type between many of the fractions. Figure 31
shows a sample LC report with a number of compound clashes represented in
the eluent.
The results of the LC fractionation procedure include quantitative
estimates of TCO and GRAV range materials in each of seven fractions. In
most cases, the quantity of material actually taken for the LC separation is
only a Dortion of the total sample, and the amount taken should be stated in
the report. The actual, measured TCO and GRAV values for the LC fractions
should be multiplied by the appropriate factor (total sample quantity r
quantity taken for LC) to give the corresponding total sample values. It is
then useful to convert these quantitative estimates into equivalent concen-
trations at the source in ordc-r to facilitate comparisons with various
decision criteria. Figure 31 illustrates the format for reporting LC frac-
tionation data, with an example from a SA5S train sorhent trap extract.
Note that GRAV analyses involve weighing to the nearest 0.1 mg. TCO values
are reported to the nearest 0.1 mg, also.
9.4.5 Infrared Analysis
The total sample extract, or neat liquid, and the 7 LC fractions are
analyzed by infrared (IR) spectrophotometry. A grating spectrophotometer
should be used and the following instrument conditions adhered to.
1. Resolution: For dispcrsively measured spectra, the spectral
slit width should not exceed 4 cm * through at least 80 percent
of the wave number range.
2. Wave number accuracy: ±4 cm * below 2,000 cm * and ±15 cm *
above 2,000 cm
3. Noise level: No more than 2 percent peak to peak.
4. Baseline flatness: The I or 100 percent line must be fl?t to
witliin 5 percent across the recorded spectrum.
A-20
-------
LC REPORT
SAMPLE: 11-3 SQREEuT TRAP EXTRACT
TCO
GFAV
TCO + GRAV
Concentration
m3
«"S
Total mj
m3/ {m^, L, or kg)
Teal Sample^
103
3C3
432
82
TaV.r .'or LC^
23
C4
107
18
Rccoii Cfeil^
17
74
SI
IS
Fraction
TCO innij
CRAV in ng
TCO ~
GnAV
Total rrj
Concentration
irj/
(m^.L, or kg}
Found n
Fract on
Clank
Cor-
rects
To si4
Fcj-d in
Fr-Jcn
EI;Y<
Cor-
rected
Total4
1
0.3
ao
0.3
1.4
2.5
ao
2.5
11 5
12.9
1.9
*
33
ai
32
15.C
10
ai
09
4.1
19.1
12
3
12.0
as
11 5
540
Ki
2.5
53 9
2s 7.9
331.9
50 0
4
ai
ao
01
cs
2.2
01
2.1
97
112
1.7
5
02
ai
ai
0.5
c 3
C.5
E 3
29 0
23 5
4.9
6
07
aG
07
12
IE
02
3.4
15.6
138
3.1
7
0.1
ao
01
as
as
ao
06
2.8
13
0.6
SliTI
16 7
0.7
1C3
751
73.1
3.4
63 7
320.6
355 8
65.9
1. Q"3nlity in eitiT dc* ai :.ii ltd tj!j:b LC
2. Pc::.dp of Wiii.': crrpls ussd lo: LC, amud r.j
1 Quantity r:co.cr;J Iron LC ccljrr.t, actual ri3
4. Tot:l ni; comput'-d hack to total sa-n/e
Figure 31. Serr.ple LC report.
-------
5. Energy: The instrument should be purged with dry gas or evacuated
so that atmosphere water bands do not exceed the allowable noise
level (2 percent) when the instrument is ued in a double beam
mode.
6. Spectral range: Spectra shoulo be recorded, without gaps, over
the spectral range 3,800-G00 cm
7. False radiation: Not to exceed 2 percent.
IR spectra are obtained in absorbance units on samples held between two NaCl
salt plates using methylene chloride to transfer the sample to the plates.
KBr pellets can also be used if films (from MeCl2) will not give satisfactory
spectra, i.e., material is a crystalline solid. Sample quantity and instru-
ment parameters are adjusted so that the maximum signal of the strongest
peak is less than 1.0 absorbance.
Spectra are interpreted in terms of functional group types present in
the sample or LC fraction. The many reference texts (refs. 96-101) in this
area are of considerable help in interpreting the IR spectra. The interpre-
tation of the spectra should also be guided by consideration of the LC
fractionation scheme and the LRMS results (when available).
The results of the IR analysis should be reported in the format shown
in Figure 32 according to the following guidelines:
a. The frequency reported should be the peak maxinum, or a range may
be reported instead for broad peaks with no well-defined maximum.
b. Absorbance values should be measured by baseline technique. The
intensity is reported relative to the strongest peak in the spec-
trum on a percentage absorbance basis.
S = strong, 70-100 percent of the absorbance value of the
strongest peak.
M = medium, 30-70 percent of the absorbance value of the
strongest peak.
W = weak, 0-30 percent of the absorbance value of the
strongest peak.
When a peak is of borderline intensity, it should be labeled "m."
Finer intensity ratings such as M-S or W-M are not appropriate.
A-22
-------
in neronT
SAMPLL: II-3-LC6
Wave Number
(cm"')
Intensity
Alignment
Comments
3400
M
OH, Nil
Brodd
3050
W
imsat'd CH
2850. 2920
M
sjt'd CH
1710
S
acid, kclonos
1680
M
amide, ketone*
1600
M
aromatic C = C
1450
M
CH?
1060
S
Si—O, ether
Broad
740
M
subst. pyridine, C—CI
Figure 32. Sample IK report.
c. The assign .ient/comnients column indicates the functional group(s)
to which the peak is attributed*1. This column may also contain
single-word descriptors of peak shape such as "broad," "doublet,"
"shoulder."
d. All weak, medium, and strong peaks must be reported.
e. A copy of the infrared spectrum should be retained at the labor-
atory for 3 years should further reference to it be needed.
As a matter of quality control, the IR b3nds observed with each LC
fraction should be compared to those observed with the total sample. Addi-
tional bands in any of the fractions would indicate the introduction of
impurities or decomposition on the column. The IR bands observed with the
total sample but not with any of the LC fractions would indicate loss of
material onto the column.
*A11 features of the IR spoclru:i, such as presc-nce or absence of bands at
other rol.itfd IR frequencies, should be considered in rnkiiig the functional
group assignments.
A-23
-------
9.4.6 Low Resolution Mass Spectrometry (rofs. 102-105)
A low resolution mass spectrum (LRMS) is obtained on each LC fraction
that has sufficient quantity (TCO plus GRAV), when referenced back to the
source, to exceed decision criteria concentrations. The Process Measure-
ments Branch guidelines recommended for general Level 1 purposes are:
Gas—SASS train samples 0.5 mg/m3
Ambient air--particulate 1 pg/m3
Solids 1 mg/kg
Aqueous solutions 0.1 mg/L
An LRMS analysis is to be done on any LC fraction that corresponds to a
source concentration higher than these levels.
In the event that other, more stringent criteria are determined to be
appropriate for a particular environmental assessment, then the cutoff point
for performing LRMS will change accordingly. For example, if MEG/MATE
(Multimedia environmental Goals/Minimum Acute Toxicity Effluent) concen-
tration values are to be used in making Level 1 decisions, then LRUS must be
done on all LC fractions for a SASS sample. This is because MATE values are
sufficiently lev/ that in each LC fraction there is at least one compound
category that might be present whose level of concern would be exceeded by
any detectable quantity of material.
In order to minimize t^e cost of the-Level 1 organic analysis, it is
desirable to keep the total number of LRMS analyses required as low as
possible. An LRMS analysis must always be run on any LC fraction that
exceeds the Level 1 concentration criteria given above. However, if the
more stringent (MEG/MATE) cutoff criteria are used, it is acceptable, for
LRMS purposes only, to combine fractions falling below the Level 1 concen-
tration criteria, according to the following scheme:
LCI = LRMS-1
LC2 plus LC3 = LRMS-2, 3
LC4 plus LC5 = LRMS-4, 5
LC6 plus LC7 = LRMS-6, 7
This will allow LRMS results to be obtained on all LC fractions in a minimum
number of separate LRMS analyses. Tin* IR data obtained with each LC fraction
should be considered before making the decision to combine fractions. If
the IR spectra of two fractions considered for combination are vastly dif-
A-24
-------
"ferent, those fractions should probably not be combined as the LRMS of the
mixture will he especially difficult to interpret.
The mass spectrometer used in this determination should have a resolu-
tion (M/AM) of 800 to 1,000, batch and direct probe inlet, variable ionizing
voltage source, and electron multiplier detection. Samples with significant
quantities of TC0 range material (>2 mg) should be analyzed by insertion in
the batch inlet. All samples that meet the decision criteria for quantity
(TC0 plus GRAV) will require analysis via the direct insertion probe. A
small quantity of sample is placed in the probe capillary and inserted into
a cool source. The temperature is then programmed up to vaporize the sample.
Spectra are recorded periodically through this period. Spectra are normally
obtained at 70 cV ionizing voltage, but low voltage (10 eV) spectra may
provide much simpler data and thus aid in interpretation in some cases.
The mass spectroscopist should integrate the interpretation of the
batch and probe mass spectra obtained on a particular sample to provide one
report describing sample chemistry. Details of quantitation of LRMS data
are too numerous to be addressed in this manual.
Interpretation of the mass spectra is guided by knowledge of the LC
separation scheme, the 1R spectra, and other information about the source.
In reporting the results of the LRfiS analysis, the basic philosophy is to
present increasingly more specific data as the complexity (or simplicity) of
the spectra will allow. The first level of reporting is to identify compound
classes. If possible, or appropriate, one should then attempt to identify
the subcategory compound classes present in the fraction. Finally, specific
compounds should be identified, if possible to do so from the spectra.
Where possible, the molecular weight range and composition of each cnteqory
shojld be estimated with a rating of 100 - major, 10 = minor, and 1 = trace.
It should be possible, using this methodoloqy, to account for nearly
all observed species by selection from a relatively small list of compound
categories and subcategories. A tentative list of such categories has been
assembled in Table 17. Tfie primary reference for selecting these categories
was the I*LG list, which seems to do an aciequote jcb of representing all
probable major compound classes Sor.ie few categories were not in the MEG
list and have been added here.
A-25
-------
TABLE 17. CATEGORIES FOR REPORTING LRMS DATA
Category Kost prcboble Category Host probable
(Subcategory) IC fraction* (Subcategory) IC fraction*
Aliphatic hydrocarbons
1
Phenols
6
(Aikanes)
1
(Alkyl, etc.)
6
(A'kenes)
1
(Halccjpnatcd phenols)
6
(A1 kynes)
1
(Nitrophenols)
6
Haloge.uted aliphatlcs
1.2
Esters
6
(Satjijted)
1.2
(Phthalatcs)
6
(Unsaturated)
1.2
Ketones
6
Anines
6
Aro.-atic hydrocarbons
2,3
(Primary, secondary, tertiary)
6
(Sen;- nes)
2.3
(Hydrazines, azo compounds)
6
Halogcnated aro-iatic hydrocarbons
2.3
(Ni troso-:raines)
6
Kitro aroratic hydrocarbons
4.5
Fused alternate, nonalternate hydrocarbons
Heterocyclic nitrogen compounds
2.3
(Indoles, carbaioles)
4
MV < 216 (.Tsthyl pyrene)
2,3
(Quinolines, aendmes)
0
KW > 216
2.3
Alkyl sulfur compounds
e
Ethers
4
(Morcaptans)
6
(Hulogenated ethers)
4
(Sulfides, disulfides)
6
Epoxides
4
Heterocyclic sulfur compounds
(Eonzothiophenes)
4
Aldehydes
4
Heterocyclic oxygen compounds
3,4
Sulfonic acids, sulfoxides
7
Nitriles
4
Amides
6
(#\1 i?h.itic)
4
(Aro.-.atic)
4
Carboxylic acids
6,;
Alcohols
6
(Primary, secondary, tertiary)
6
Si Mcones
2.3
(Glycols)
6
Phosphates
5.6
nxe
'Possible assignrtcnts Fractions 4-5, 5-G, 6-7 generally overlap to a considerable extent. Alio, addi-
tional components of a particular roolccjle iray cause it to elute in an LC fraction other than that ex-
pected. For example, a short-chain ester would probab'y elute in LC fraction 5 or 6 whereas a long-
chain ester would elute in Fractions 3 or 4.
-------
The list in Table 17 is also organized somewhat differently than the
MEG list to be more compatible with the nature of the mass spectrometry
data. It should bo strongly emphasized that the list probably does not
include all identifiable compound categories. If interpretation of the
spectra yields the identification of a category not included in this table,
the category should be reported. At the same time, EPA/PMB should be noti-
fied of the need to add that category to the list. It will be possible in
most cases to identify the spectra in terms of the compound categories
listed in Table 17, but one should avoid force fits if another category
seems more appropriate.
Interpretation of the mass spectral data should take full advantage of
all other information known about the sample source, i.e., LC fraction and
IR spectra. Since the LC separation does a reasonable job of dividing
compound classes, the categories listed in Table .17 have been listed in
order of their possible elution from the LC column. Where possible, some
indication has been made as to the LC fraction in which the category might
elute. These fraction assignments are known to be correct in some cases and
are only estimates in others. Sometimes the sample characteristics will
have minor effects on the fraction elution behavior. Again, the LC fraction
indications should only be taken as a guideline. Figure 33 gives a comploted
example of a reporting format for the LRMS data. Blank forms may be found
in Appendix A.
It is once again emphasized that interpretation of the LRMS data is
best done using all available information, such as what one knows about the
chemistry of the source being sampled, what species generally elute in the
LC fraction being examined, and functional group data derived from the IR
spectra.
9.5 ORGANIC ANALYSIS SUMMARY TABLES
At the end of the Level 1 organic analysis procedure, there will be an
LC report, seven IR reports, and up to seven LRMS reports for each organic
extract or neat organic sample. This is an unwieldy body of data from which
to make a decision. The first step in reducing these data to a workable
form is to prepare a single table that sur..narizes the organic analysis
A-27
-------
LRMS REPORT
SAMPLC; IJ-3-1.C6
Major Catryodei
Intensity
Category
MW Rang#
100
Kctones
130-280
100
Heterocyclic Nitrogen Conpoundg
107-253
10
Enters
10
Cnrboxylic Acids
10
Phenola
Sub-Cotc(ioiioi. Specific Compound!
Intensity
Category
m/0
Composition
100
Acridinc
179
cn"9M
100
l'lucroionc
180
C,,H„0
10
1'licnol
94
C,H,0
10
Crcsol
3 on
i?2
C7,,-°
10
l.jnnoic Acid
C ,11,0.
10
Carbn/oJc
1£Z_
S2V
10
!'c rhvl.icridinc
JJO_
104
10
Jk-'thyllluorononti
C]4^]0°
10
Anthip juinol ir.e
2?9
Ci:MnN
10
llenznnthrone
230
C17i:]0°
10
Dihonzofluorcnonc
280
C2l"l2°
10
Mlmrylphrhnl .ifp
278
C]6!,22°4
10
Mp rhy 1 hen *.in fh rr>nr
2*4
C18H12U
O'her
Fi'jmo n3. Sninplc LI IMS report.
A-28
-------
results for each extract. Figure 34 Illustrates the organization of this
table and the following paragraphs describe the various entries.
Space is allotted in the table heading for a sample identification
code. It is assumed that each laboratory will have devised its own coding
system (see Chapter 2) for uniquely identifying the various samples. It may
be desirable to include the date, the name of the analyst, or other similar
Information.
The body of the table includes one column for each of the LC fractions
and one column for summing the data. The first set of data entries is the
quantitative analysis, transcribed from the LC report. The calculated total
organic loading corresponding to each fraction is entered in the first row.
This value is used in estimating the abundance of the various organic com-
pound classes. The next two rows contain the estimated TCO and GRAV values
to indicate the distribution of total volatile (bp 100°-300° C) and nonvola-
tile (bp >300° C) organic materials. This information can be useful later
in the selection of appropriate decision criteria (Level 1, MATE, etc.)
values for comparison with the Level 1 results. The results of the compound
category analysis (primarily from LRMS data) are summarized in the bottom
columns.
For those LC fractions that contained sufficient quantity to have been
analyzed by LRMS, the results of the LRMS analyses are summarized in the
table as follows: The major categories present in LC fraction 1 are listed
at the left-hand side of the table and the approximate intensity (100, 10,
or 1) for each category is entered in the LCI column. To convert the LRMS
intensity index to a concentration estimate for each organic compound cate-
gory, the individual intensity value is divided by the sum of all intensities
for the LC fraction and then multiplied by the total organic loading (mg/m3)
estimated for the fraction. This procedure is then repeated for the other
LC fractions. Examples are worked in Figure 35 for two LC fractions, LC2
and LC4, of the same XAD-2 extract.
The results presented in Figure 34 illustrate the considerable overlap
in chemical class composition that can be expected between some fractions in
the Level 1 LC scheme. As noted earlier, this is not a high resolution
separation technique and the various compound categories cannot be uniquely
A-29
-------
ORGANIC EXTRACT SUMMARY TABLE
Sample So'bcnt Extract-11-3
LC1
LC2
LC3
LC4
LCS
LCS
LC7
z
Teal C'gan.cs, mg
13 2
22 3
253
29 7
11 0
46 3
15 1
390
TCC, rvg
52
19
73.
6 7
37
53
0 1
110
G~>^V. mg
13
33
1S0.
23
73
41.
*.5.
230
Category
Assigned intensiiy-mg/ (m3, L, or kg)*
Sl! fur
1O0--O.6
0.6
A'-phjtiC HC's
10-0.0 5
0 06
Arcnnt'Ci—Benzenes
10-0 06
0 06
Fi slJ Ato'h 216
100-0 6
100—4
100-0 5
5
Fused Arom 216
10-0C6
100—4
100—C 5
5
fu'C-GC^ Ci. C S
10-0 C6
10-0 4
10-0 05
0 5
H:tcrocyc! c N
10-0 05
-0.11
100-0.7
10-0 02
1
-0 1*
10-0 07
100-0 2
03
Cn-bcuy'ic Acids
-0.1*
100-0 7
10-0 02
1 0
Phi 1015
-0 01*
10-0 07
10-0 02
0.1
E-iic:s
-0.01*
10-0 07
0 03
•Concentration for ocs samples " mg/m^, for liqu
oCtL-al L, or kg V3lue.
d samples =
ng/l, for solic
samples = m
g/kg. Fill in
^Esiimnt:d avumwg same relative intensities as LC6, since IR spectra of LC5 and LC3 ire very similar,
I t 1 i
H4. Crynmc cx;ra^t sur.mjry tab'a.
-------
LC 2
Total organics =0.57 mg/m3
LC 4
Total organics = 6.6 mg/m3
Aromatic HC's-bsnzenes 10
Fused Arom <216 MW 100
Fused Arom >216 MW 10
Heteroyclic S Compounds 10
Fused Arom <216 MW 100
Fused Arom >216 MW 100
Heterocyclic S Compounds 10
Heterocyclic 0 Compounds 10
Calculation of Concentration Estimates by Category
Yj intensities = 130 53 intensities = 220
Fused Arom <216 MW:
Fused Arom >216 MW:
Heterocyclic S Cmpds:
X
0.57
= 0.04 mg/m3
100 „
T3"0 x
0.57
= 0.4 mg/m3
Fused Arom <215 MW:
100
d-
o|o
X
0.57
=0.04 mg/m3
Fused Arom >216 Md:
100
220"
J0X
130 x
0.57
= 0.04 mg/m3
Heterocyclic S Cmpds:
10
120
Heterocyclic 0 Cmpds: x 6.6 = 0.3 mg/m3
Figure 35. Sample calculations of concentration estimates from LRMS data.
-------
assigned to particular LC fractions. The data do show the expected trend,
in that LC2 is relatively richer in light aromatics (benzene and fused
species with M <216) than is LC4. The fact that adjacent LC fractions can
be expected to show gradual changes in chemistry can serve as a uselul guide
for the analyst in detecting contamination. Very abrupt changes in apparent
composition or the appearance of a compound class in an entirely unexptcted
fraction (i.e., phthalates in LC2 or paraffins in LC6) should be regarded
with suspicion.
The overlap between fractions can also be used in estimating the composi-
tion of those fractions that did not contain sufficient material to trigger
an LRMS analysis. For any LC fraction that was not analyzed by LRMS, it is
suggested that the IR spectrum be compared with the IR's of the adjacent
fractions. If a close correlation is found between tv/o IR spectra, then
this, together with the known behavior of the LC scherce, suggests that the
two LC fractions nave similar, though not identical, qualitative composi-
tion. It is therefore suggested that the total organics in the non-LRMS LC
fraction be distributed over the same classes and in the same proportion as
was done for the adjacent LC fraction whose IR srjectrum was the best match.
The error introduced by this procedure into tht: overall description of
sample chemistry will be small, since only fractions with small amounts of
material are excluded from the LRMS.
Concentrations estimated by this procedure should be identified with an
asterisk in the organic extract summary table and an explanatory footnote
should be included.
For those LC fractions for which only IR data are available, the proce-
dure for estimating concentrations by compound class is as follows:
1. List all categories from Table 17 that could be in that fraction.
2. Assign a weighting factor of 100 to each category that appears to
be present in the sample based on the IR spectrum.
3. Assign a weighting factor of 10 to each category for which func-
tional groups were not identified.
Then, in the absence of evidence to the contrary, assume that cate-
gories with an intensity of 100 may constitute up to 50 percent of the total
sample and those with an intensity of 10 up to 10 percent. Clearly, this
A-32
-------
procedure may "account for" more than 100 percent of the total sample.
However, this conservative method of estimation seems necessary because it
is not always possible to determine from the Ili spectrum of a mixture alone
how the various functional groups are assembled into molecules or classes.
Table 18 illustrates this procedure for an LC5 sample, assuming only IR
and LC data v/ere available. Concentrations estimated on the basis of IR and
LC data only should be footnoted with a dagger in the organic extract summary
table and an explanatory footnote should be included.
Once concentrations have been estimated for all compound categories
identified in the seven LC fractions' for a particular organic extract, or
neat organic liquid, these values are summed across each row of the table.
(See Figure 34.) This procedure condenses the information obtained on the
various LC fractions to provide an integrated description of the chemical
composition. In the case of liquid and solid samples, the summation column
represents the total concentration/compound category information for the
stream sampled. For gaseous streams, the summary report for each component
of the SASS train must be added to determine the stream composition, as was
shown in Table 15.
9.6 QUALITY CONTROL IN LEVEL 1 ORGANIC ANALYSIS
This discussion has been written for the analytical chemist, who can be
presumed to be familiar with generally accepted standards of good laboratory
practice. It is worthwhile, however, to retmphasize the importance of some
procedures, in addition to sanple preparation and analysis per se, which
have a very significant impact on the overall quality of the analytical
results. Some of those procedures, in particular those related to prevent-
ing and/or recognizing sample contamination, are especially critical in a
Level 1 environmental assessment.
To insure adequate data quality, it is essential that blanks (controls)
be analyzed along with the samples, as discussed in Chapter 2. There is no
reliable way to identify spurious results and/or sample contamination other
than by finding the same contaminant in a control sample.
A-33
-------
TABLE 18. ESTIMATION OF FRACTION COMPOSITION FROM IR AND LC DATA ONLY
Sample LC 5
IR Report
Wave nur.ber
cm 1 I Assignments/comments
Total Organics = 0.28 mq/m3 Estimated
Assigned possible
Categories most probable weighting concentration
in LC 5 factor (mg/m3)*
2,400
3,050
2,850-2,950
2,230
1,700
1,600
1,530
1,-720-1,450
S
M
S
M
S
S
M
S
1,050-1,300 M
700-850
OH or NH (broad)
Aromatic CH
Aliphatic CH
ChN or C=C, cyanates or
isocyanates
Ketone, carboxylic acid,
aldehydes
Conj. C=C or aromatic
C=N, N-M02, C-N=0,
Carfcoxylate
C=N, C-N02
CH.,, NH4+, CH2, Si-pnenyl
caroonates
Alcohol C-0, phenol C-0,
ether C-0, C-F, various
P-0 compounds, various Si-0
or Si-C compounds
Subst. bsnzene rings, some
olefimc C-H, C-Cl, CF-CF,
some peroxides
Heterocyclic N compounds
100
0.14
Heterocyclic S compounds
10
0.03
Sulfides, disulfides
10
0.03
Nitriles
100
0.14
Ethers
100
0.14
Aldehydes, ketones
100
0.14
Nitroaliphatics
100
0.14
Alcohols
100
0.14
Nitroaromatics
100
0.14
Amines
100
0.14
Phenols
100
0.14
Esters
10
0.03
*Total organics = 0.28 mg/nr.
-------
Control sample, of course, means more than a simple reagent blank for
the analysis itself. The control sample and its handling throughout the
laboratory sequence should be identical to the real sample and its handling,
except that the real sample has been exposed to a process stream in a sam-
pling procedure. For instance, an unexposed XAD-2 cartridge must be dumped,
homogenized, and a 5-g aliquot reserved for the Parr bomb ashing and trace
element assays. The remainder must be Soxhlet extracted and the extract
subjected to the entire organics analysis sequence.
The procedures for cleaning the XAD-2 resin prior to use are specified
in Appendix B. The quality control checks described in the appendix should
be applied to each batch of resin before it is used in a field study.
Each lot of organic solvent (i.e., each new batch number) must be
checked for contamination. A volume of solvent equivalent to that used in
sample extraction should be evaporated to dryness. The residue should be
weighed and examined by IR. If any significant quantity of organic contam-
ination is found, the solvent batch must be redistilled or rejected entirely.
Solvents used should be Burdick and Jackson "distilled in glass" or equiva-
lent quality. Note that use of chemicals of the- specified grade does not.
eliminate the necessity of performing checks on the quality of each new
batch of material used. However, use of high-quality reagents is important
to minimize the probability of acquiring unsatisfactory lots of material,
which would require repurification or replacement. Sodium sulfate and
silica gel used in the LC separation will frequently require cleanup by
extraction with organic solvent prior to use, as described in Section 9.4.4.
These controls and blanks should permit the analyst to identify the
source of any background contaminant and to make corrections to the results
of sample analyses. If contamination is excessive (more than 10 percent of
the sample level), the source should be traced and the contamination elimi-
nated, if possible. Note that silicones (from lubricants/sealants) and
phthalates (from plastics) are major potential interferences and must be
avoided entirely in collection, storage, and handling of samples for organic
analysis.
A-J5
-------
APPENDIX B
PREPARATION OF XAD-2 SORBENT RESIN
B-l
-------
APPENDIX B
PREPARATION OF XAD-2 SORBENT RESIN
B.l SCOPE AND APPLICATION
XAD-2 resin, as supplied by the manufacturer, is impregnated with a
bicarbonate solution to inhibit microbial growth during storage. Both the
salt solution and any residual extractable monomer and polymer species must
be removed before use. The resin is prepared by a series of water and
organic extractions followed by careful drying.
6.2 EXTRACTION
B.2.1 Method 1
The procedure may be carriod out in a giant Soxhlet extractor, which
will contain enough XAD-2 for a single SASS module. An all-glass thimble
(55-90 mm 00 x 250 mm deep [top to frit]} containing an extra-coarse frit is
used for extraction of XA0-2. The frit is recessed 10-15 mm above a crenu-
lated ring at the bottom of the thimble to facilitate drainage. The resin
must be carefully retained in the extractor cup with a glass wool plug and
stainless steel screen since it floats on the final solvent, methylene
chloride. This process involves sequential extraction in the following
order.
Solvent Procedure
Water Initial rinse with 1 L H20 for 1 cycle, then
Water
Methyl alcohol
Methylene chloride
Methylene chloride (fiesh)
discard H20
Extract with H20 for 8 hours
Extract for 22 hours
Extract for 22 hours
Extract for 22 hours
B-2
-------
B.2.2 Method 2
As an alternate to Soxhlet extraction, a continuous extractor has also
been fabricated for the extraction sequence and is described in Figure B-l.
This extractor has been found to be acceptable. The particular canister
used for the apparatus shown in figure B-l contains about 500 g of finished
XAD-2 or enough for more than three sorbent modules. Any size may be con-
structed; the choice is dependent on the needs of the sampling programs.
The XAD-2 is held under light spring tension between a pair of coarse and
fine screens. Spacers under the bottom screen allow for even distribution
of clean solvent. The three-necked flask should be of sufficient size (3 L
in this case) to hold solvent equal to twice the dead volume of the XAD-2
canister. Solvent is refluxed through the Snyder column and the distillate
continuously cycled up through the XAD-2 for extraction and returned to the
flask. The flow is maintained upwards through the XAD-2 to allow maximum
solvent contact and prevent channeling. A valve at the bottom of the can-
ister allows removal of solvent from the canister between changes.
Experience has shown that it is very difficult to cycle sufficient
water in this mode. Therefore, the aqueous rinse is accomplished by simply
flushing the canister with about 20 L of distilled water. A small pump may
be useful for pumping the water through the canister. The water extraction
should be carried out at the rate of about 20-40 mL/min.
After draining the water, subsequent methyl alcohol and methylene
chloride extractions are carried out using the refluxing apparatus. An
overnight or 10- to 20-hour period is normally sufficient for each extrac-
tion.
All materials of construction are glass, Teflon, or stainless steel.
Pumps, if used, should not contain extractable materials. Pumps are not used
with methanol and methylene chloride.
B.3 DRYING
After evaluation of several methods of removing residual solvent, a
fluidized-bed technique has proven to be the fastest and most reliable
drying method.
A simple column with suitable retainers as shown in figure B-2 will
serve as a satisfactory column. A 10.2-cm (4-in.) Pyrex pipe 0.6 m (2 ft)
B-3
-------
Distillat* Tak* off
0.65 cm x 0 32 cm Union
(1/4 *1/8 in)
>o«
0 32 cm (1/3 in) Ttflon
v Tubing
0 32 en Un'on
:I*low
0.32 cm Union
Figure B-1. XAD 2 clcmiup attraction apparatus.
B-4
-------
Loose Weave Nylon
Fabric Cover
10.2 cm
(4 Inch) Pyrox
Pipe
Liquid Take off
0 95 cm (3/8 in) Tubing
FmeSc-eii
.jJ ^Czitu Sjpp'jrt
Figure B 2. XAD 2 ffuidi/cd b«"d drying apparatus
B-5
-------
long will hold all of the XAD-2 from the extractor shown in Figure B_1 or
the Soxhlet extractor, with sufficient space for fluidizing the bed while
generating a minimum resin load at the exit of the column.
B.3.1 Method 1
The gas used to remove the solvent is the key to preserving the clean-
liness of the XAO-2. Liquid nitrogen from a regular commercial liquid
nitrogen cylinder has routinely proven to be a reliable source of large
volumes of gas free from organic contaminants. The liquid nitrogen cylinder
is connected to the column by a length of precleaned 0.95-cm (3/8-in.)
copper tubing, coiled to pass through a heat source. As nitrogen is bled
from the cylinder, it is vaporized in the heat source and passes through the
column. A convenient heat source is a v/ater bath heated from a steam Tine.
The final nitrogen temperature should only be warm to the touch and not over
40° C. Experience has shown that about 500 g of XAD-2 may be dried over-
night consuming a full 160-L cylinder of liquid nitrogen.
B.3.2 Method 2
As a second choice, high purity tank nitrogen may be used to dry the
XAD-2. The high purity nitrogen must first be passed through a bed of
activated charcoal approximately 150 ml in volume. With either type of
drying method, the rate of flow should gently agitate t"ie bed. Excessive
fluidization may cause the particles to break up.
B.4 QUALITY C0NTI10L PROCEDURES
For both Methods 1 and 2, the quality control results must be reported
for the batch. The batch must be reextracted with methylene chloride if the
residual extractable organics are greater than 20 pg/mL or the gravimetric
residue is greater than 0.5 mg/20 g XAD-2 extracted.
Three control procedures are used with the final XAD-2 to check for (1)
residual methylene chloride, (2) extractable organics (TC0), and (3) residue
(GRAV).
8.4.1 Procedure for Residual Methyl one Chloride
B.4.1.1 Description--
A 1 ±0.1 g sample of dried resin is weighed into a small vial, 3 mL of
B-6
-------
toluene are added, and the vial is capped and well shaken. Five microliters
of toluene (now containing extracted methylene chloride) are injected into a
gas chromatograph, and the resulting integrated area is compared to a refer-
ence standard.
The reference solution consists cf 2.5 pL of methylene chloride in
100 mL of toluene, simulating 100 pg residual methylene chloride on the
resin. The acceptable maximum content is 1,000 pg/g resin.
B.4.1.2 Experimental —
6 ft. x 1/8 in. SS column containing 10% OV-101 on 100/120 Supel-
coport
Helium carrier at 30 mL/min
FID operated on 4 x 10 11 A/mV
Injection port temp 250° C, detector temp 305° C
Programmed: 30° C (4 min) 40°/min 250° C (hold)
Program terminated at 1000 seconds.
B.4.2 Procedure for Residual Fxtractdble Orrpnics
B.4.2.1 Dtscription--
A 20 10.] g sample of cleaned, dried resin is weighed into a precleaned
alundum or cellulose thimble which is plugged with cleaned, glass wool.
(Note that 20 g of resin will fill a thimble, and the resin will float out
unless well plugged.) The thimble containing the resin is extracted for
24 b with 200 mL of pesticide-grade methylene chloride.*
The 200-mL extract is reduced in volume to 10 mL using a nitrogen
evaporation strea-n. Five microliters of that solution are analyzed by gas
chromatography using the TC0 analysis procedure. The concentrated solution
should not contain more than ?0 |ij/mL of TC0 extracted from the XAD-2. This
is equivalent to 1U pg/g of TC0 in the XAD-2 and would correspond to 1.3 nig
of TCO in the extract of the 130-g XAD-2 module. Care should be taken to
correct the TCO data for a solvent blank prepared (200 -* 10 mL concentra-
tion) in a similar manner.
B.4.2.2 Experimentdl--
Use the 1C0 analysis conditions described in the revised Level 1 manual.
*Burdick & Jackson pesticide grade or equivalent purity.
B-7
-------
B.4.3 Methodology for Residual Gravimetric Determination.
After the TCO value is obtained for the rosin batch by the above
procedures, dry the remainder of the extract in a tared vessel. There must
be less than 0.5 mg residue registered or the batch of resin will have to be
extracted with fresh methylene chloride again until it meets this criteria.
This level corresponds to 25 pg/g in the XAD-2 or about 3.25 mg in a resin
charge of 130 g.
B-8
-------
This Page Intentionally Blank
------- |