MIDWEST RESEARCH INSTITUTE
¦¦¦1
/ANALYTICAL METHOD: THE ANALYSIS OF CHLORINATED BIPHENYLS
IN LIQUIDS AND SOLIDS
WORK ASSIGNMENT NO. 6
DRAFT SPECIAL REPORT NO. 4
EPA Contract No. 68-02-3938
MRI Project Mo. 8201-A(6)
February 21, 1985
For
U.S. Environmental Protection Agency
Office of Toxic Substances
Field Studies Branch, TS-798
401 M Street, SW
Washington, DC 20460
Attn: Frederick W. Kutz, Project Officer
Daniel T. Heggen, Work Assignment Manager
¦¦¦¦
MIDWEST RESEARCH INSTITUTE 425 VOLKER BOULEVARD, KANSAS CITY, MISSOURI 64110 • 816 753-7600
-------
\
ANALYTICAL METHOO: THE ANALYSIS OF CHLORINATED BIPHENYLS
Mitchell D. Erickson
John S. Stanley
J. Kay Turman
Gil Radolovich
WORK ASSIGNMENT NO. 6
DRAFT SPECIAL REPORT NO. 4
EPA Contract No. 68-02-3938
MRI Project No. 8201-A(6)
February 21, 1985
U.S. Environmental Protection Agency
Office of Toxic Substances
Field Studies Branch, TS-798
401 M Street, SW
Washington, DC 20460
Attn: Frederick W. Kutz, Project Officer
Daniel T. Heggem, Work Assignment Manager
IN LIQUIDS AND SOLIDS
by
For
J
MIDWEST RESEARCH INSTITUTE 425 VOLKER BOULEVARD, KANSAS CITY, MISSOURI 64110 • 816 753-7600
-------
DISCLAIMER
This document is a preliminary draft. It has not been released
formally by the Office of Toxic Substances, Office of Pesticides and Toxic
Substances, U.S. Environmental Protection Agency, and should not at this
stage be construed to represent Agency policy. It is being circulated for
comments on its technical merit and policy implication.
-------
PREFACE
This report describes a method for the analysis of polychlorinated
biphenyls in solids and liquids. The work was done on Work Assignment No. 6
on US Environmental Protection Agency Contract No. 68-02-3938. This report
was prepared by Mitchell Erickson, John S. Stanley, J. Kay Turman, and Gil
Radolovich. The work on previous, similar methods for by-product PCBs was
conducted by Or. Erickson, Or. Stanley, Ms. Turman, Mr. Radolovich, Karin Bauer,
Jon Onstot, Oonna Rose, Margaret Wickham, and Ruth Blair.
The EPA Work Assignment Manager, Daniel T. Heggem, of Field Studies
Branch provided helpful guidance.
Approved:
James L. Spigarelli, Director
Chemical and Biological Sciences
Department
MIDWEST RESEARCH INSTITUTE
C/LL
Clarence L. Haile
Deputy Program Manager
(T&L
John E. Going
Program Manager
i i
-------
TABLE OF CONTENTS
Page
1.0 Scope and Application 1
2.0 Summary 4
3.0 Definitions 11
4.0 Interferences 11
5.0 Safety 12
6.0 Apparatus and Materials 13
7.0 Reagents 17
8.0 Calibration 21
9.0 Sample Collection, Handling, and Preservation 31
10.0 Sample Preparation 34
11.0 Gas Chromatographic/Electron Impact Mass Spectrometric
Determination 50
12.0 Qualitative Identification 52
13.0 Quantitative Data Reduction 53
14.0 Confirmation 55
15.0 Quality Assurance 55
16.0 Quality Control 56
17.0 Method Performance 59
18.0 Documentation and Records 59
References 61
i i i
-------
LIST OF TABLES
Number paqe
1 Comparison of Commercial PCB Mixtures 2
2 Numbering of PCB Congeners 3
3 Standard Procedures of Analysis for PCBs 5
4 Average Molecular Composition (wt.%) of Some Aroclors ... 9
5 DFTPP Key Ions and Ion Abundance Criteria 15
6 Concentrations of Congeners in PCB Calibration Standards
for Full Scan Analysis (ng/pL) 18
7 Concentrations of Congeners in PCB Calibration Standards
for Selected Ion Monitoring and Limited Mass Scan
Analysis (pg/pL) 19
8 Composition of Internal Standard Spiking Solution (SS100)
Containing 13-Labeled PCBs 20
9 Operating Parameters for Capillary Column Gas Chroma-
tographic System 22
10 Operating Parameters for Packed Column Gas Chromatography
System 23
11 Operating Parameters for Quadrupole Mass Spectrometer
System 24
12 Operating Parameters for Magnetic Sector Mass
Spectrometer System 25
13 Limited Mass Scanning (LMS) Ranges for PCBs 27
14 Characteristic SIM Ions for PCBs 28
15 Pairings of Analyte and Calibration Compounds 30
16 Relative Retention Time (RRT) Ranges of PCB Homologs
Versus ds-3,3',4,4'-Tetrachlorobiphenyl 32
17 Method Performance Parameters 60
i v
-------
LIST OF FIGURES
Reconstructed ion chromatogram of calibration solution
FSlOOng PCB obtained in the full scan mode. The con-
centration of the 10 PCB calibration congeners, the
4 13C-labeled PCB recovery surrogates, and the 3
internal standards are in Table 6. See Table 2 for
PCB numbering system, Table 9 for capillary GC param-
eters, and Table 11 for mass spectrometer operating
parameters
v
-------
THE ANALYSIS OF CHLORINATED BIPHENYLS IN LIQUIDS AND SOLIDS
1.0 Scope and Application
1.1 Analytes - This is a gas chromatographic/electron impact mass spec-
trometric (GC/EIMS) method applicable to the determination of chlo-
rinated biphenyls (PCBs) in liquid or solid samples of environmental,
human, or commercial origin. The PCBs may originate either as con-
taminants derived from commercial PCB products (e.g., Aroclors) or
as synthetic by-products. This is a general method designed for
detection of any PCB mixture in any solid or liquid matrix. It is
recognized that in most cases, PCBs will be derived from commercial
mixtures (see Table 1), although other mixtures or single congeners
can also be determined. The PCBs may include any of the 209 conge-
ners from monochlorobiphenyl through decachlorobiphenyl listed in
Table 2 and may be present as single isomers or complex mixtures.
1.2 Matrices - Because of the inherent quality control exercised with
each sample, the method is applicable to virtually any solid or
liquid matrix. Air or other gases require special, detailed sam-
pling protocols. A similar method for ambient air and stack gas is
available (Erickson 1984b).
1.3 Quality Control - The method is highly dependent on quality control
measures. The validity of the results is based on the measured re-
covery of four 13C-labeled PCBs which are added to every sample
prior to extraction and cleanup. The use of these recovery surro-
gates to validate the data for each sample permits the use of any
common extraction or cleanup technique. Quantitative recovery of
these 13C~labeled surrogate PCBs will indicate that the sample prep-
aration techniques have been appropriately applied.
Conversely, low surrogate recoveries indicate an inappropriate cleanup
and the sample must be reanalyzed using a different preparation.
The validity of the results depends on equivalent recovery of the
analyte and 13C PCBs. If the 13C PCBs are not thoroughly incorpo-
rated in the matrix, the method is not applicable.
1.4 Method Performance - The detection and quantitation limits are de-
pendent upon the complexity of the sample matrix and the ability of
the analyst to remove interferents and properly maintain the analyt-
ical system. For a complex mixture of low level PCBs (0.05 to 0.2
M9/9)» the method precision appears to be about ± 60% based on a
very limited study. The method performance at higher levels (e.g.,
2 |JQ/g) will be determined in future studies.
1.5 Analyst Qualifications - This method is restricted to use by or under
the supervision of analysts experienced in the use of gas chromatog-
raphy/mass spectrometry (GC/MS) and in the interpretation of gas
1
-------
Table 1. Comparison of Commercial PCB Mixtures
Trade Names
Aroclor Clophen Phenoclor Pyralenec Kanechlor Fenclor*
Av. No. CI/
Molecule
Approx.
Wt.% CI
Approx.
"Mol. Wt."
1221
1.15
21
193.7
1232
2,000
200
2
32-33
223.0
1,500
2.5
38
240.3
1242,
A30
DP3
3,000
300
42
3
40-42
257.5
1016
1248
A40
DP4
400
4
48
291.9
1254
A50
DP5
500
54f
5
52-54
326.4
1260
A60
DP6
600
64
6-6.3
60
366.0
1262
f
6.8
62
388.4
70
7.7
65
419.4
1268
8.7
68
453.8
1270
9.5
70
481.4
DK
10
71
498.6
^Monsanto Industrial Chemicals Company, USA.
"Bayer, GFR.
dCaffaro, Italy.
e«anegafuchi Chemical Company, Japan.
^Prodelec, France.
The two-digit number should indicate the wt.% CI; however, this does not fit in with the manufacturer's
specifications.
Source: Brinkman and DeKok, 1980. Reproduced with permission, copyright 1980, Elsevier Biomedical
Publishers BV.
-------
No
I
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
SU
51
Table 2. Numbering of PCB Congeners
Structure
No
Structure
No
Structure
No
Monochlorobiphcm Is
Tetrachlorobtphcnvli
Pcnuchlerobiphcn>ls
2
52
2.2 5 5
105
2 3 3-4 4-
161
3
- 53
2JT.5 6-
106
2.3 3 4 5
162
4
54
2.2',6 6'
107
2.3 3 .4 5
163
Dichlorobiphenvls
55
56
2.3 3' 4
2.3 3 ,4'
108
109
2.3.3- 4,5'
2.3.3- 4,6
164
165
2.2'
57
2 3.3'.5
no
2.3.3 4 6
166
2.3
58
2.3.3.5"
111
2 3.3',5 5'
167
2.3'
59
2 3 3" 6
112
2.3 3 .5 6
168
2.4
60
2.3,4 4'
113
2,3,3'.5'.6
169
2.4'
61
2.3.4 5
114
2 3 4 4 5
2.5
62
2 3 4 6
115
2.3.4 4-6
2 6
63
2 3 4-5
116
2.3 4 5 6
170
33
64
2 3 4 6
117
2.3 4 5 6
171
34
65
2.3 5 6
118
2.3' 4 4' 5
172
3.4'
66
2.3' 4 4
119
2.3',4 4,6
173
3.5
67
2,3 .4.5
120
2.3 ,4 5.5"
174
4 4'
68
2 3 4 5
121
2.3 .4 5' 6
175
Trichlorobiphtnsls
69
70
2.3 ,4.6
2,3',4 .5
122
123
2',3,3',4 5
2',3 4 4'5
176
177
2.2',3
71
2.3' 4".6
124
2'.3 4.5 5'
178
2.2".4
72
2.3'.5.5'
125
2',3.4 5 6'
179
2.2'j
73
2 3' 5',6
126
3 3 ,4 4* 5
180
2.2'6
74
2.4,4 5
127
3 3-4 5 5'
181
2.3.3'
2.3.4
75
76
2 4 4 6
2' 3.4 5
Hcxachlorobiphenyls
182
183
2 3 4'
77
3.3',4 4'
128
2.2 .3.3 4,4'
184
2.3.5
78
3 3'4 5
129
•J-2',3 3' 4 5
185
2,3.6
79
3.3'4 5-
130
2,2',3 3 4 5'
186
2.3'4
80
3 3-5.5'
131
2 2' 3 3 4 6
187
2.3' 5
81
3.4.4' 5
132
2.2',3 3 46
188
2.3',6
2.4,4'
PmLiChlorobiphem Is
133
134
2.2'.3.3' 5 5-
2 2'.3 3 .5,6
189
190
245
82
2 2' 3 3'.4
135
2.2'.3.3'.5.6'
191
2,4,6
83
2.2' 3 3'.5
136
2.2' 3 3' 6.6
192
2.4'5
84
2 2'.3 3- 6
137
2 2' 3 4 4'.5
193
2.4'6
85
2.2' 3 4 4'
138
2.2'3 44 5
2',3.4
86
2.2'.3.4 5
139
2 2 3 4 4-6
2'.3.5
87
2,2'.3 4.5'
140
2.2'3.4 4' 6'
194
3.3',4
88
2.2' 3 4 6
141
2.2' 3 4 5 5"
195
3.3'.5
89
2.2 .3 4 6
142
2.2' 3.4 5 6
196
3.4 4'
90
2.2' 3 4' 5
143
2 2".3.4 5.6'
197
345
91
2.2' 3.4'.6
144
2.2' 3 4 5' 6
198
3.4'.5
92
2.2' 3.5.5'
145
2.2',3 4 6 6'
199
Tcirachlorobiphenyls
93
94
2.2'.3.5.6
2.2 3 5 6
146
147
2.2' 3.4' 5.5-
2.2'.3 4' 5.6
200
201
2.2',3.3'
95
2.2' 3.5',6
148
2.2',3 4 5 6'
202
2.2',3 4
96
2 2' 3.6.6'
149
2.2 3 4'.5" 6
203
2.2'.3,4
97
2,2'.3'.4.5
150
2 2'3 4'66
204
2.2'.3.5
98
2.2'.3 .4,6
151
2.2' 3 5 5' 6
205
2.2',3,5'
99
2.2'4 4",5
152
2.2',3 5.6 6-
2,2'.3.6
too
2.2' 4 4-6
153
2.2' 4.4- 5 5'
22 3 6'
101
2.2- 4.5 5
154
2.2 .44 5 6
206
2 2 4 4
102
2.2' 4 5 6'
155
22 44 66
207
2.2' 4 5
103
2.2 4 5 .6
156
2.3 3 4 4 5
208
2.2' 4 5
104
2 2 .4 6 6
157
2 3 3' 4 4 5'
2 2 4 6
15X
2.3 1 4 4 6
2 2 4 6'
159
160
233 455
233 456
209
3
-------
chromatograms and mass spectra. Prior to sample analysis, each ana-
lyst must demonstrate the ability to generate acceptable results
with this method by following the procedures described in Section
16.2.
1.6 Background - This method is derived from methods for by-product PCBs
in commercial products and wastes, water, and air (Erickson 1984a,
1984b, 1984c). These methods were developed with the mandate to be
able to quantitate any PCB or PCB mixture in virtually any matrix.
The rationale for the development of these by-product methods has
been published (USEPA, 1982c; Erickson and Stanley 1982; Erickson
et al. 1985).
A variety of general and specific sample preparation options are
presented in this method. In some cases, these options are taken
from the standard procedures for PCBs listed in Table 3. In other
cases, no standard procedure exists and general guidance is given
with appropriate literature references. This method takes a dif-
ferent approach from those which rely on Aroclor mixtures for cali-
bration and quantitation (EPA 1979a, 1979b; Longbottom and Lichtenberg
1982; EPA 1982a). In this method PCBs are detected and quantitated
by homolog group. The results can be summed to give a total PCB
value comparable to results generated by other methods or they may
be presented as 10 individual homolog values. This homolog distri-
bution can provide additional quantitative information on the com-
position and source of the PCBs. For example, the relative homolog
distributions may indicate how closely the sample resembles a com-
mercial mixture (see Tables 1 and 4).
The use of GC/EIMS yields qualitative information which not only
readily differentiates among the PCB homologs but also permits
elimination of other compounds from quantitation. The other common
PCB analysis technique, GC/ECD, does not discriminate between PCBs
and many common interferents. Thus, GC/ECD results are less reli-
able for complex matrices.
1.7 Protocol - This method contains many options because of the diver-
sity of matrices and interferences which may be encountered. Once
the appropriate options for each sample type have been selected,
each laboratory should prepare a written step-by-step protocol for
use by the analysts. The protocol may contain verbatim sections
from this method, more detailed steps for certain techniques, or
totally different extraction or cleanup techniques.
2.0 Summary
2.1 The process or product must be sampled such that the specimen col-
lected for analysis is representative of the whole. Statistically
designed selection of the sampling position, time, or discrete items
should be employed where appropriate. The sample must be preserved
to prevent PCB loss prior to analysis. Room temperature storage
4
-------
Table 3. Standard Procedures of Analysis for PCBs
Matrix
Extraction
CIeanupc
Procedure designation
Reference
cn
Water
Water
Water
Water
Water
Water
Water
Water
Water
SIudge
CH2C12
CH2C12
Hexane/CH,C1
2° 12
Several
Hexane
Hexane/CH2C12
Hexane
Hexane
Hexane
Hexane/CH2C12/
acetone (83/15/2)
(Florisi1)
(S removal)
None
Florisil/Silica
Gel (CH3CN)
(S removal)
Several
A1 umi na
(Florisi1)
(Silica Gel)
(H2S04)
(Saponi fi cati on)
(A1umina)
(H2S04)
(Saponification)
A1umi na
Si 1ica Gel
GPC
S removal
608
625
304h
EPA (by-products)
Monsanto
D3534-80
D3304-74
ANSI
UK - DOE
EPA (Halocarbon)
EPA 1979a;
Longbottom and
Lichtenberg 1982
EPA 1979b;
Longbottom and
Lichtenberg 1982
EPA 1978
Erickson et al.
1982, 1983a;
Erickson 1984c
Moein 1976
ASTM 1981a
ASTM 1981b
ANSI 1974
UK DOE 1979;
Devenish and
Harling-Bowen 1980
Rodriguez et al.
1980
-------
Matrix
Extraction
SIudge
SIudge
SIudge
Sol id waste
Sol id waste
Sol id waste
Soil, sediment
CH2C12 (base/
neutral and acid
fractions)
CH2C12 (base/
neutral and acid
fractions)
CH2C12
(3 fractions)
CH2C12
CH2C12
CH2C12
Acetone/hexane
Soil, sediment CH3CN
Sediment CH3CN
Blood
Adipose
Adipose
Hexane
Pet. ether/CHjCN
Pet. ether/CHjCN
Table 3 (continued)
Cleanupc
GPC
Procedure designation
Reference
Priority
pol1utant
EPA 1979c
Florisil, silica
gel or GPC
625-S
Haile and Lopez-Avila
1984
GPC
Silica gel
(Florisi1)
None
None
Florisi1
Silica gel
(S removal)
(Saponification)
(H2S04)
(A1umina)
Saponification
H2S04
A1umi na
(F1ori si 1)
Florisi1
Saponification
Florisi1
B100
8080
8250
8270
EPA
D3304-77
Monsanto
EPA [5,A,(3)]
EPA [5,A,(1)]
EPA (9,D)
Ballinger 1978
EPA 1982a
EPA 1982a
EPA 1982a
EPA 1982b
ASTM 1981b; ANSI
1974
Moein 1976
Watts 1980
Watts 1980
Watts 1980
-------
Matrix
Extraction
Mi 1 k Acetone/hexane
Food CH,CN/Pet. ether
Food, Pet. ether/CH,CN
Food Pet. ether/CH3CN
Paper and Saponification
paperboard
Transformer fluids DI
or waste oils
Mineral oil Dilute with
hexane or iso-
octane
Products or wastes Several
Table 3 (continued)
CIeanupc
Procedure designation
Reference
CH3CN
F1ori si 1
Silica acid
EPA (9,B)
Watts 1980
Sherma 1981
Florisil MgO/
Celite
Saponi fication
Silicic acid
(Saponi fication)
(Oxidation)
(Flori si 1)
Silica gel
Saponi fication
(Florisi1)
Florisil MgO/
Celi te
Saponi fication
(H2S04)
(Florisi1)
(A1umina)
(Silica gel)
(GPC) (CH3CN)
Florisil slurry
(H2S04)
(Florisil column)
Several
AOAC (29)
PAM
Japan
AOAC (29)
EPA (oil)
D4059-83
EPA (by-products)
AOAC 1980a
FDA 1977
Tanabe 1976
AOAC 1980b
EPA 1981;
Bellar and
Lichtenberg 1981
ASTM 1983
Erickson et al.
1982, 1983;
Erickson 1984a
-------
Table 3 (concluded)
Matrix
Extraction
Cleanup'
Procedure designation
Reference
Chlori nated
benzenes
DI
None
DOW
Dow 1981
3 Pigment types
Unspeci fied
A. Hexane/H2S0^
B. CH2C12
Hexane/acetone
None
Flori si 1
(CH3CN)
(Florisi1)
(Silica gel)
(mercury)
DCMA
EPA (spills)
DCMA 1982
Beard and Schaum
1978
techniques in parentheses are described as optional in the procedure.
CO
-------
Table 4. Average Molecular Composition
(wt.%) of Some Aroclors
Homolog
Aroclor
(Chlorines) 1221 1232a 1016 1242 1248 1254 1260
0
10
1
50
26
2
1
2
35
29
19
13
1
3
4
24
57
45
2
1
4
1
15
22
31
49
15
5
10
27
53
12
6
2
26
42
7
4
38
8
7
9
1
Five percent unidentified (biphenyl?).
Source: Brinkman and DeKok, 1980. Reproduced with permis-
sion, copyright 1980, Elsevier Biomedical Publishers BV.
9
-------
may be adequate for some samples if biological degradation will not
occur. Otherwise, subambient storage is recommended; refrigeration
for aqueous samples and freezing for all others.
2.2 The sample is mechanically homogenized and subsampled if necessary.
The sample must then be spiked with four 13C PCB,surrogates (4-chloro-
13C6-biphenyl; 3,3',4,4'-tetrachloro-13Cx2-biphenyl; 2,2',3,3',5,5',
6,6'-octachloro-13C12-biphenyl; and decachloro-13C12~biphenyl) and
the surrogates incorporated by further mechanical agitation.
2.3 The surrogate-spiked sample is extracted and cleaned up at the dis-
cretion of the analyst. Simple dilution or direct injection is per-
missible. Possible extraction techniques include liquid-liquid par-
tition, liquid-solid partition, thermal desorption, and sorption
onto resin columns followed by solvent desorption. Cleanup tech-
niques may include liquid-liquid partition, sulfuric acid cleanup,
saponification, adsorption chromatography, high performance liquid
chromatography, gel permeation chromatography, or a combination of
cleanup techniques. The sample is diluted or concentrated to a final
known volume for instrumental determination.
2.4 The PCB content of the sample extract must be determined by high
resolution (preferred) or packed column gas chromatography/electron
impact mass spectrometry (HRGC/EIMS or PGC/EIMS) operated in the
full scan, selected ion monitoring (SIM), or limited mass scan (LMS)
mode.
2.5 PCBs are identified by comparison of their retention time and mass
spectral intensity ratios to those in calibration standards.
2.6 PCBs are quantitated by the internal standard technique, using re-
sponse factors for a mixture of 10 PCB congeners. The recoveries
of four 13C surrogates are used to monitor for losses in workup and
determi nation.
2.7 The PCBs identified by the SIM technique may be confirmed by full
scan HRGC/EIMS, retention on alternate GC columns, other mass spec-
trometric techniques, infrared spectrometry, or other techniques,
provided that the sensitivity and selectivity of the technique are
demonstrated to be comparable or superior to GC/EIMS.
2.8 The analysis time is dependent on the extent of workup employed.
The time required for instrumental analysis of a single sample, ex-
cluding instrumental calibration, data reduction and reporting, is
typically 30 to 45 min.
2.9 A quality assurance (QA) plan must be developed for each laboratory.
2.10 Quality control (QC) measures include laboratory certification and
performance check sample analysis, procedural QC (instrumental per-
formance, calculation checks), and sample QC (blanks, replicates,
and standard addition).
10
-------
3.0 Definitions
3.1 PCB Congener - One of the 209 discrete chemical compounds that are
listed in Table 2 (e.g., 2',3,4,5-tetrachlorobiphenyl; congener num-
ber 76).
3.2 Homolog - One of the 10 degrees of chlorination of the PCBs (e.g.,
tetrachlorobiphenyl)
3.3 Isomers - PCBs of a given homolog with different chlorine substi-
tution patterns (e.g., 2,3,4-trichlorobiphenyl and 3,3',5-trichloro-
biphenyl are two of the 12 trichlorobiphenyl isomers).
3.4 Internal standard - A compound not expected to be found in the sam-
ple which is added to the sample extract prior to instrumental anal-
ysis, used as a measure of the instrumental response. In this method,
(d6)-3,4,3',4'-tetrachlorobiphenyl, 1-iodonaphthalene, or d12_chrysene
are the compounds recommended for use as internal standards.
3.5 Surrogate compound - A compound not expected to be found in the sam-
ple, which is added to the original sample, and measured with the
same procedures used to measure sample components. The recovery of
these surrogate compounds is indicative of the recovery of the native
PCBs. Also called a recovery surrogate. In this method, four Re-
labeled PCBs are used as surrogate compounds.
4.0 Interferences
4.1 Method interferences may be caused by contaminants in solvents, re-
agents, glassware, and other sample processing hardware, leading to
discrete artifacts and/or elevated baselines in the total ion cur-
rent profiles. All of these materials must be routinely demonstrated
to be free from interferences by the analysis of laboratory reagent
blanks as described in Section 16.4.
4.1.1 Glassware must be scrupulously cleaned. All glassware
should be cleaned as soon as possible after use by rins-
ing with the last solvent used. This should be followed
by detergent washing with hot water and rinses with tap
water and reagent water. The glassware should then be
drained dry and heated in a muffle furnace at 400°C for
15 to 30 min. Some thermally stable materials, such as
PCBs, may not be eliminated by this treatment. Solvent
rinses with acetone and pesticide quality hexane may be
substituted for the muffle furnace heating. Volumetric
ware should not be heated in a muffle furnace. After it
is dry and cool, glassware should be sealed and stored in
a clean environment to prevent any accumulation of dust
or other contaminants. It is stored inverted or capped
with aluminum foil.
11
-------
4.1.2 The use of high purity reagents and solvents helps to mini-
mize interference problems. Purification of solvents by
distillation in all-glass systems may be required. All
solvent lots must be checked for purity prior to use.
4.2 Matrix interferences may be caused by contaminants that are coex-
tracted from the sample. The extent of matrix interferences will
vary considerably from source to source, depending upon the nature
and diversity of the sources of samples.
Many common matrix interferences are removed by the suggested ex-
traction and cleanup techniques. Where possible, standard procedures
are referenced which have proven reliable at removing common inter-
ferences such as fats and chlorinated insecticides.
5.0 Safety
5.1 The toxicity or carcinogenicity of each reagent used in this method
has not been precisely defined; however, each chemical compound
should be treated as a potential health hazard. From this viewpoint,
exposure to these chemicals must be reduced to the lowest possible
level by whatever means available. The laboratory is responsible
for maintaining a current awareness file of OSHA regulations regard-
ing the safe handling of the chemicals specified in this method. A
reference file of material data handling sheets should also be made
available to all personnel involved in the chemical analysis.
5.2 Polychlorinated biphenyls have been tentatively classified as known
or suspected human or mammalian carcinogens. Primary standards of
these toxic compounds should be prepared in a hood. Personnel must
wear protective equipment, including gloves and safety glasses.
Congeners highly substituted at the meta and para positions and un-
substituted at the ortho positions are reported to be the most toxic.
Extreme caution should be taken when handling these compounds neat
or in concentrated solutions. This class includes 3,3',4,4'-tetra-
chlorobiphenyl (both natural abundance and isotopically labeled).
5.3 Diethyl ether should be monitored regularly to determine the perox-
ide content. Under no circumstances should diethyl ether be used
with a peroxide content in excess of 50 ppm, as an explosion could
result. Peroxide test strips manufactured by EM Laboratories (avail-
able from Scientific Products Company, Cat. No. P1126-8 and other
suppliers) are recommended for this test. Procedures for removal
of peroxides from diethyl ether are included in the instructions
supplied with the peroxide test kit.
5.4 Waste disposal must be in accordance with TSCA (USEPA 1983), RCRA
and applicable state rules.
12
-------
Apparatus and Materials
6.1 Sampling containers -
6.1.1 Liquids - Glass bottles, 1-L or other appropriate volume,
fitted with screw caps lined with Teflon are appropriate
for liquid samples. Cleaned foil may be substituted for
Teflon if the sample is not corrosive. Samples should be
protected from light using amber bottles (preferable),
foil or a light-tight outer container. Bottles must be
washed, rinsed with acetone or methylene chloride, and
dried before use to minimize contamination.
6.1.2 Solids - The volume, materials, and configuration of solid
sample containers are dependent on the amount of sample
to be collected and its physical properties. Glass jars,
as described for liquids, may be appropriate for some sam-
ples. In most cases, however, wide-mouthed bottles will
be required.
Other containers such as wide-mouthed ointment jars or
heavy-walled polyethylene bags may be appropriate for
smaller samples such as vegetation, dry powders, or adi-
pose, provided they are demonstrated to be free of con-
tamination and preserve the sample integrity.
6.1.3 All sample containers and caps should be washed in deter-
gent solution, rinsed with tap water, and then rinsed with
distilled water. The bottles and caps are allowed to drain
dry in a contaminant-free area. Then the caps are rinsed
with pesticide grade hexane and allowed to air dry.
6.1.4 Sample bottles are heated to 400°C for 15 to 20 min or
rinsed with pesticide grade acetone or hexane and allowed
to air dry.
6.1.5 The clean bottles are stored inverted and sealed until
use.
6.2 Glassware - All specifications are suggestions only. Catalog numbers
are included for illustration only.
6.2.1 Volumetric flasks - Assorted sizes.
6.2.2 Pipets - Assorted sizes, Mohr delivery.
6.2.3 Micro syringes - 10.0 pL for packed column GC analysis,
1.0 pL for on-column GC analysis.
6.2.4 Chromatographic column - Chromaflex, 400 mm long x 19 mm
ID (Kontes K-420540-9011 or equivalent).
13
-------
6.2.5 Kuderna-Danish Evaporative Concentrator Apparatus
6.2.5.1 Concentrator tube - 10 mL, graduated (Kontes
K-570050-1025 or equivalent). Calibration must
be checked. Ground glass stopper size ($19/22
joint) is used to prevent evaporation of solvent.
6.2.5.2 Evaporative flask - 500 mL (Kontes K-57001-0500
or equivalent). Attached to concentrator tube
with springs (Kontes K-662750-0012 or equivalent).
6.2.5.3 Snyder column - Three ball macro (Kontes K-503000-
0121 or equivalent).
6.3 Balance - Analytical, capable of accurately weighing 0.0001 g.
6.4 Gas chromatography/mass spectrometer system.
6.4.1 Gas chromatograph - An analytical system complete with a
temperature programmable gas chromatograph and all required
accessories including syringes, analytical columns, and
gases. The injection port must be designed for on-column
injection when using packed columns. Any capillary injec-
tion techniques (split, splitless, on-column, "Grob," etc.)
may be used provided the performance specifications stated
in Section 8.1 are met.
6.4.2 High resolution (capillary) GC column - A 10-30 m long x
0.25 mm ID fused silica column with a 0.25 pm thick DB-5
bonded silicone liquid phase (J&W Scientific) is recom-
mended. Alternate liquid phases may include 0V-101, SP-2100,
Apiezon L, Dexsil 300, or other liquid phases or columns
which meet the performance specifications stated in Section
8.1.
6.4.3 Packed GC column - A 180 cm x 0.2 cm ID glass column packed
with 3% SP-2250 on 100/120 mesh Supelcoport or equivalent
is recommended. Other liquid phases or columns which meet
the performance specifications stated in Section 8.1 may
be substituted.
6.4.4 Mass spectrometer - Must be capable of scanning from at
least m/z 150 to m/z 550 every 1.5 s or less, collecting
at least five spectra per chromatographic peak, utilizing
a 70-eV (nominal) electron energy in the electron impact
ionization mode and producing a mass spectrum which meets
all the criteria in Table 5 when 50 ng of decaf1uorotri-
phenyl phosphine [DFTPP, bis(perf1uorophenyl)phenyl phos-
phine] is injected through the GC inlet. Any GC-to-MS
interface that gives acceptable calibration points at 10
ng per injection for each PC8 isomer in the calibration
standard and achieves all acceptable performance criteria
14
-------
Table 5. DFTPP Key Ions and Ion Abundance Criteria
n/z Ion abundance criteria
197
Less than 1% of mass 198
198
100% relative abundance
199
5-9% of mass 198
275
10-30% of mass 198
365
Greater than 1% of mass 198
441
Present, but less than mass 443
442
Greater than 40% of mass 198
443
17-23% of mass 442
15
-------
(Section 10) may be used. Direct coupling of the fused
silica column to the MS is recommended. Alternatively,
GC-to-MS interfaces constructed of all glass or glass-
lined materials are recommended. Glass can be deactivated
by silanizing with dichlorodimethylsilane.
6.4.5 A computer system that allows the continuous acquisition
and storage on machine-readable media of all mass spectra
obtained throughout the duration of the chromatographic
program must be interfaced to the mass spectrometer. The
data system must have the capability of integrating the
abundances of the selected ions between specified limits
and relating integrated abundances to concentrations using
the calibration procedures described in this method. The
computer must have software that allows searching any GC/
MS data file for ions of a specific mass and plotting such
ion abundances versus time or scan number to yield an ex-
tracted ion current profile (EICP). Software must also
be available that allows integrating the abundance in any
EICP between specified time or scan number limits.
6.5 Chromatographic systems (optional; see Sections 10.3.2 and 10.4).
6.5.1 Gel permeation chromatography.
6.5.1.1 GPC Autoprep 1002 (Analytical Biochemistry Labo-
ratories, Inc.) or equivalent.
6.5.1.2 A Bio-Beads SX-3 (Bio-Rad) column.
6.5.2 High performance liquid chromatography.
6.5.2.1 Waters Model 6000A pump or equivalent.
6.5.2.2 Waters Model 440 UV detector or equivalent.
6.5.2.3 Rheodyne 7125 injector or equivalent.
6.5.2.4 Amine column (Waters nBondapak, 3.9 x 300 mm)
or equivalent.
6.5.3 Gas chromatograph for GC/FID screening.
6.5.3.1 A temperature-programmable GC equipped with a
flame ionization detector. Varian 3740 or
equivalent.
6.5.3.2 A 2 m x 2 mm ID glass column packed with 3% SP-
2250 on 100/120 mesh Supelcoport or equivalent.
A high resolution GC column may also be used.
16
-------
7.0 Reagents
7.1 Solvents - All solvents must be pesticide residue analysis grade.
New lots should be checked for purity by concentrating an aliquot
by at least as much as is used in the procedure. HPLC solvents
should have UV cutoffs of 210 nm or less.
7.2 Calibration standard congeners - Standards of the PCB congeners
listed in Tables 6 and 7 are available from Ultra Scientific, Hope,
Rhode Island; or Analabs, North Haven, Connecticut.
7.3 Calibration standard stock solutions - Primary dilutions of each of
the individual PCBs listed in Table 3 are prepared by weighing ap-
proximately 1-10 mg of material within 1% precision. The PCB is
then dissolved and diluted to 1.0 mL with hexane. The concentra-
tion is calculated in mg/mL. The primary dilutions are stored at
4°C in screw-cap vials with Teflon cap liners. The meniscus is
marked on the vial wall to monitor solvent evaporation. Primary
dilutions are stable indefinitely if the seals are maintained. The
stock solutions and dilutions should be clearly labeled with pert-
inent information such as sample code, solvent, date prepared, ini-
tials of person preparing the solution, and notebook reference.
7.4 Working calibration standards - Working calibration standards are
prepared that are similar in PCB composition and concentration to
the samples by mixing and diluting the individual standard stock
solutions. Example calibration solutions are shown in Tables 6 and
7. The mixture is diluted to volume with pesticide residue analy-
sis quality hexane. Dilutions are stored at 4°C in narrow-mouth,
screw-cap vials with Teflon cap liners. The meniscus is marked on
the vial wall to monitor solvent evaporation. These secondary di-
lutions can be stored indefinitely if the seals are maintained.
These solutions are designated FSxxx ng PCB and SIxxx pg PCB where
the xxx is used to encode the nominal concentration of the lower
congeners in ng/pL and pg/pL, respectively. The FS prefix helps
aid the analyst in identifying solutions which are appropriate for
full scan analysis; the SI prefix is for solutions to calibrate in
the selected ion monitoring and limited mass scan acquisition modes.
7.5 Alternatively, certified stock solutions similar to those listed in
Tables 6 and 7 may be available from a supplier, in lieu of the pro-
cedure described in Section 7.4.
7.6 DFTPP standard - A 50-ng/^L solution of decafluorotriphenylphosphine
(DFTPP, PCR Research Chemicals, Gainesville, FL) is prepared in ace-
tone or another appropriate solvent.
7.7 Surrogate standard stock solution - The four 13C-labeled PCBs listed
in Table 8 are available as a certified solution and may be requested
from the Toxic and Hazardous Materials Repository, U.S. Environmental
Protection Agency, Environmental Monitoring and Support Laboratory,
17
-------
Table 6. Concentrations of Congeners in PCB Calibration Standards
for Full Scan Analysis (ng/|jL)a
Congener
FS100
FS050
FS010
FS005
FS001
Homolog
no.
ng PCB
ng PCB
ng PBC
ng PCB
ng PCB
1
1
100
50
10
5
1
2
7
100
50
10
5
1
3
30
150
75
15
7.5
1.5
4
50
200
100
20
10
2
5
97
200
100
20
10
2
6
143
200
100
20
10
2
7
183
300
150
30
15
3
8
202
300
150
30
15
3
9
207
450
225
45
22.5
4.5
10
209
200
100
20
10
2
4
210 (IS)
250
250
250
250
250
-
C10H7I (IS)b
250
250
250
250
250
-
Ci8Di2 (IS)C
250
250
250
250
250
13C-C1x
211 (RS)
100
50
10
5
1
13c-ci4
212 (RS)
250
125
25
12.5
2.5
13c-ci8
213 (RS)
400
200
40
20
4
13c-ci10
214 (RS)
500
250
50
25
5
^^Concentrations given as examples only.
cl-Iodonaphthalene.
d12-Chrysene.
18
-------
Table 7. Concentrations of Congeners in PCB Calibration Standards
for Selected Ion Monitoring and Limited Mass Scan
Analysis (pg/|jl)
Congener
SI1000
SI100
SI050
SI010
Homolog
no.
pg PCB
pg pcb
pg pbc
pg pcb
1
1
1,000
100
50
10
2
7
1,000
100
50
10
3
30
1,500
150
75
15
4
50
2,000
200
100
20
5
97
2,000
200
100
20
6
143
2,000
200
100
20
7
183
3,000
300
150
30
8
202
3,000
300
150
30
9
207
4,500
450
225
45
10
209
2,000
200
100
20
4
210 (IS)
250
250
250
250
-
C10H7I (IS)b
250
250
250
250
-
C18D12 (IS)c
250
250
250
250
13C-C1x
211 (RS)
1,000
100
50
10
13C-C14
212 (RS)
2,500
250
125
25
13C-C18
213 (RS)
4,000
400
200
40
13C-C110
214 (RS)
5,000
500
250
50
^Concentrations given as examples only.
cl-Iodonaphthalene.
d12-Chrysene.
19
-------
Table 8. Composition of Internal Standard Spiking Solution (SS100)
Containing 13C-Labeled PCBs
Congener Concentration
no. Compound Abbreviations (pg/mL)
211 4-Chloro-d',2',3',4',5',6'-13C6)-biphenyl 13C-C1X 100
212 3,3',4,4'-Tetrachloro-(13C12)-biphenyl 13C-C14 250
213 2,2',3,3',5,5',6,6'-Octachloro-(13C12)-biphenyl 13C-C18 400
214 Decachloro-(13C12)_biphenyl 13C-C110 500
-------
26 West St. Clair Street, Cincinnati, Ohio 45268, (513) 684-7327.
This solution may be used as received or diluted further. These
solutions are designated "SSxxx," where the xxx is used to encode
the nominal concentration in |jg/mL.
7.8 Internal standard solutions - Solutions of d6-3,31,4,4'-tetrachloro-
biphenyl (KOR Isotopes, Cambridge, MA), 1-iodonaphthalene (Aldrich
Chemical Company, Milwaukee, WI) or d12-chrysene (KOR Isotopes,
Cambridge, MA) are prepared at nominal concentrations of 1-10 mg/mL
in hexane. The solutions are further diluted to give working stan-
dards.
Note: Any internal standard may be used, provided it meets the fol-
lowing criteria: (a) it is not already present in the sample, (b)
it gives a strong, recognizable mass spectrum, (c) it does not give
mass spectral ions which interfere with native or 13C-labeled PCB
quantitation, (d) it is chemically stable, and (e) it elutes in the
PCB retention window. Ideally, several internal standards are used
which have retention times spanning the PCB retention windows to
improve the response factor precision. A second internal standard
also provides an alternative if the primary internal standard cannot
be used because of interference.
7.9 Solution stability - The calibration standard, surrogate, and DFTPP
solutions should be checked frequently for stability. These solu-
tions should be replaced after 6 months, or sooner if comparison
with quality control check samples indicates compound degradation
or concentration change.
8.0 Calibration
8.1 The gas chromatograph must meet the minimum operating parameters
shown in Tables 9 and 10, daily. If all criteria are not met, the
analyst must adjust conditions and repeat the test until all cri-
teria are met.
8.2 The mass spectrometer must meet the minimum operating parameters
shown in Tables 5, 11, and 12, daily. If all criteria are not met,
the analyst must retune the spectrometer and repeat the test until
all conditions are met. Instrumental drift can be monitored in all
acquisition modes by comparing response factors (see Section 8.3)
obtained at the beginning and end of the day. If they do not agree
within 20%, the data quality should be considered suspect.
8.2.1 Full scan data acquisition - Quadrupole mass spectrometers
must meet the tuning criteria in Table 5. The spectrometer
must scan between m/z 150-550, although wider scan ranges
are permissible.
21
-------
Table 9. Operating Parameters for Capillary Column Gas Chromatographic System
Parameter
Recommended
Tolerance
Gas chromatograph
Finnigan 9610
Other3
Column
15 -30 m x 0.255 mm ID
Fused silica
Other
Liquid phase
DB-5 (J&W)
Other nonpolar
or semipolar
Liquid phase thickness
0.25 pm
< 1 pm
Carrier gas
Heli um
Hydrogen
Carrier gas velocity
30-45 cm/sb
Optimum performance
Injector
"Grob" (split/splitless
mode)
Other0
Injector temperature
250-270°C
Optimum performance
Injection volume
1.0-2.0 pL
Other
Initial column temperature
60-80°C (2 min)d
Other'"'
Column temperature program
70°-300°C at 10°C/mine
Other
Separator
None^
Gldss jet or other
Transfer line temperature
280°C
Optimum^
Tailing factor'1
0.7-1.5
0.4-3
Peak width1
7-10 s
< 15 s
Substitutions permitted with any common apparatus or technique provided
^performance criteria are met.
cMeasured by injection of air or methane at 270°C oven temperature.
Manufacturer's instructions should be followed regarding injection tech-
nique.
v/ith on-column injection, initial temperature equals boiling point of the
gsolvent; in this instance, hexane.
Ci2Cl10 elutes at 270°C. Programming above this temperature ensures a
^clean column and lower background on subsequent runs.
Fused silica columns may be routed directly into the ion source to prevent
separator discrimination and losses.
^High enough to elute all PCBs, but not high enough to degrade the column
hif routed through the transfer line.
Tailing factor is width of front half of peak at 10% height divided by
width of back half of peak at 10% height for single PCB congeners in solu-
• tion FSxxx ng PCB or SIxxx pg PCB.
Peak width at 10% height for a single PCB congener is FSxxx ng PCB or
SIxxx pg PCB.
22
-------
Table 10. Operating Parameters for Packed Column Gas Chromatography System
Parameter
Recommended
Tolerance
Gas chromatograph
Finnigan 9610
Other3
Column
180 cm x 0.2 cm ID
glass
Other
Column packing
3% SP-2250 on 100/
120 mesh Supelcoport
Other nonpolar
or semi polar
Carrier gas
Helium
Hydrogen
Carrier gas flow rate
30 mL/min
Optimum performance
Injector
On-column
Other
Injector temperature
250°C
Optimum
Injection volume
1.0 jjL
^ 5 |jL
Initial column temperature
150°C, 4 min
Other
Column temperature program
150°-260°C at 8°/min
Other
Separator
Glass jet
Other
Transfer line temperature
280°C
„ ^. b
Optimum
Tailing factor0
0.7-1.5
0.4-3
Peak width*"*
10-20 sec
< 30 sec
.Substitutions permitted if performance criteria are met.
cHigh enough to elute all PCBs.
Tailing factor is width of front half of peak at 10% height divided by
width of back half of peak at 10% height for single PCB congeners in solu-
.tion FSxxx ng PCB or SIxxx pg PCB.
Peak width at 10% height for a single PCB congener is FSxxx ng PCB or
SIxxx pg PCB.
23
-------
Table 11. Operating Parameters for Quadrupole Mass Spectrometer System
Parameter
Recommended
Tolerance
Mass spectrometer
Finnigan 4023
Other3
Data system
Incos 2400
Other,
Scan range
95-550
Other
Scan time
1 sec
Other'3
Resolution
Unit
Optimum performance
Ion source temperature
280°C
200°-300°C
Electron energy0
70 eV
70 eV
^Substitutions permitted if performance criteria are met.
greater than five data points over a GC peak is a minimum.
Filaments should be shut off during solvent elution to improve instrument
stability and prolong filament life, especially if no solvent venting is
used.
24
-------
Table 12. Operating Parameters for Magnetic Sector Mass Spectrometer System
Parameter • Recommended Tolerance
Mass spectrometer
Finnigan MAT 311A
Other3
Data system
Incos 2400
Other
Scan range
98-550
Other
Scan mode
Exponential
Other
Cycle time
1.2 sec
Other'5
Resolution
1,000
> 500
Ion source temperature
280°C
250-300
Electron energy0
70 eV
70 eV
^Substitutions permitted if performance criteria are met.
greater than five data points over a GC peak is a minimum.
Filaments should be shut off during solvent elution to improve instrument
stability and prolong filament life, especially if no solvent venting is
used.
25
-------
8.2.2 Limited mass scan data acquisition - Table 13 presents a
suggested set of LMS ranges. The mass spectrometer should
be set to at least unit resolution. The computer acquisi-
tion parameters should utilize the minimum threshold filter-
ing necessary so as not to lose pertinent data. Optimum
acquisition parameters will vary depending on the condi-
tion of the mass spectrometer and should be checked daily.
The dwell times for the mass ranges given in Table 13 will
vary with instrument and should be optimized to allow at
least five data points across a chromatographic peak.
Maximum sensitivity will be achieved when utilizing max-
imum dwell time.
Instruments having the capability to switch mass ranges
during an analysis require particular attention to the
switching points to assure minimal data loss. Switching
points can be initially determined by analyzing a high
level congener or Aroclor mixture while in the full scan
mode.
8.2.3 Selected ion monitoring data acquisition - Table 14 pre-
sents a suggested set of characteristic ions for SIM.
The SIM program must include at least two ions for each
analyte, generally the primary and secondary ions in Table
10. The mass spectrometer should be set to at least unit
resolution. The computer acquisition parameters should
utilize the minimum threshold filtering necessary so as
not to lose pertinent data. Optimum acquisition param-
eters will vary depending on the condition of the mass
spectrometer and should be checked daily.
The dwell times for the masses given in Table 14 will vary
with instrument and should be optimized to allow at least
five data points across a chromatographic peak. Maximum
sensitivity will be achieved when utilizing maximum dwell
time.
Instruments having the capability to switch mass ranges
during an analysis require particular attention to the
switching points to assure minimal data loss. Switching
points can be initially determined by analyzing a concen-
trated congener or Aroclor mixture while in the full scan
mode.
8.3 The PCB response factors (RF ) must be determined in triplicate or
other replication, as discussed below, using Equation 8-1 for the
analyte homologs.
A x M.
RFp =
26
-------
Table 13. Limited Mass Scanning (LMS) Ranges For PCBs
Compound Mass range (m/z)
C12HgCli +
13c612c6h9ci
186-198
^12^8^12
220-226
Ci2H7C13
254-260
C12HgCl4 +
c12o6ci4a + 13c12h6ci4
288-310
^12^5^15
322-328
Ci2H4C16
356-362
C12H3CI7
390-396
C12H2C13
426-434
C12HClg
460-468
Ci2Cl10
496-502
Ci0H7Ia
254
^18^12^
240
13Ci2H2C18
440-446
13c12ci10
508-514
aMonitor appropriate ion range for internal standard(s) being used.
27
-------
Table 14. Characteristic SIM Ions for PCBs
Ion (relative intensity)
Homolog Primary Secondary Tertiary
Cj^HgCl
188
(100)
190 (33)
Ci2H8C12
222
(100)
224 (66)
226
(11)
C12H7CI3
256
(100)
258 (99)
260
(33)
Cj^HgCl 4
292
(100)
290 (76)
294
(49)
C12H5C15
326
(100)
328 (66)
324
(61)
C12H4CI6
360
(100)
362 (82)
364
(36)
C12^3C17
394
(100)
396 (98)
398
(54)
Ci2H2C18
430
(100)
432 (66)
428
(87)
C12HC19
464
(100)
466 (76)
462
(76)
Ci2C110
498
(100)
500 (87)
496
(68)
C10H7Ia
254
(100)
-
C12D6Cl4a
298
(100)
300 (49)
296
(76)
C18D12
240
(100)
-
13c612c6h9ci
194
(100)
196 (33)
13c12h6ci4
304
(100)
306 (49)
302
(76)
13c12h2ci8
442
(100)
444 (65)
440
(87)
13c12ci10
510
(100)
512 (87)
508
(68)
aMonitor appropriate ions for internal standard(s) being used.
28
-------
where RFp = response factor of a given PCB congener
A = area of the characteristic ion for the PCB congener
^ peak
Mp = mass of PCB congener injected (nanograms)
Ai>s = area of the characteristic ion for the internal
standard peak
M^s = mass of internal standard injected (nanograms)
Using the same conditions as for RF , the surrogate response fac-
tors (RFS) must be determined usingpEquation 8-2.
A X
RFs ¦ 8"2
where Ag = area of the characteristic ion for the surrogate peak
= mass of surrogate injected (nanograms)
Other terms are the same as defined in Equation 8-1.
If specific congeners are known to be present and if standards are
available, selected RF values may be employed. For general samples,
solutions containing FSxxx ng PCB or SIxxx pg PCB, along with SSxxx
and an internal standard at appropriate levels (see, for example,
Tables 6 and 7), may be used as the response factor solution. The
PCB-surrogate pairs to be used in the RF calculation are listed in
Table 15.
Generally, only the primary ions of both the analyte and surrogate
are used to determine the RF values. If alternate ions are to be
used in the quantitation, the RF must be determined using that char-
acteristic ion.
The RF value must be determined in a manner to assure ± 20% preci-
sion. If replicate RF values differ by greater than ± 10% RSD, the
system performance should be monitored closely. If the RSD is greater
than ± 20%, the data set must be considered invalid and the RF re-
determined before further analyses are done. The analyst is respon-
sible for maintaining records of the RF precision.
For instruments with good day-to-day precision, a running mean (RF)
based on seven values may be appropriate. A new value is added each
day and the oldest dropped from the mean. Other options include,
but are not limited to, triplicate determinations of a single con-
centration spaced throughout a day or determination of the RF at
three different levels to establish a working curve. If samples
are being analyzed on successive days, a single RF determination
which is within ± 20% of the initial day's triplicate determination
may be used.
29
-------
Table 15. Pairings of Analyte and Calibration Compounds
Analyte Calibration standard
Congener3
Congener
no.
Compound
no.
Compound
1-3
2-C12H9Cl
1
2
4-15
Cj^HgCl 2
7
2,4
15-39
Ci2H7Cl3
30
2,4,6
40-81
Cj^HgCl 4
50
2,2',4,6
82-127
Ci2H5Cl5
97
2,2',3',4,5
128-169
Ci2H4C1s
143
2,2',3,4,5,6'
170-193
C12H3CI7
183
2,2",3',4,4',5', 6
194-205
Ci2H2C1s
202
2,2',3,3',5,5',6,6'
206-208
c12hci9
207
2,2',3,3',4,4',5,6,6
209
Ci2Cl 10
209
Ci2Cl 10
aBallschmiter numbering system, see Table 2.
30
-------
8.4 If the GC/EIMS system has not been demonstrated to yield a linear
response or if the analyte concentrations are more than two orders
of magnitude different from those in the RF solution, a calibration
curve must be prepared. If the analyte and RF solution concentra-
tions differ by more than one order of magnitude, a calibration curve
should be prepared. A calibration curve should be established with
triplicate determinations at three or more concentrations bracketing
the analyte levels. The precision of the triplicate determinations
must be within ± 20%.
8.5 The relative retention time (RRT) windows for the 10 homologs and
surrogates must be determined unless the entire chromatogram is
scanned for each homolog. If all congeners are not available, a
mixture of available congeners or an Aroclor mixture (e.g., 1016/
1254/1260) may be used to estimate the windows. The windows must
be set wider than observed if all isomers are not determined. Typ-
ical RRT windows for one column are listed in Table 16. The windows
may differ substantially if other GC parameters are used.
9.0 Sample Collection. Handling, and Preservation
9.1 Sampling design - Sample collection must be designed to meet sam-
pling objectives. Sampling design guidance is available (see for
example, Moser and Huibregtse 1976; EPA 1976; Mason 1982; Kelso
et al. 1985).
9.1.1 The primary consideration in sample collection is that
the sample collected be representative of the whole.
Therefore, sampling plans or protocols for each situation
will have to be developed. The recommendations presented
here describe general situations. The number of repli-
cates and sampling frequency also must be planned prior
to sampling.
9.1.2 Discrete item - If one or more discrete items would be
used as the analytical sample (e.g., apples), a statis-
tically random sampling approach is recommended.
9.1.3 Solids - Larger bulk solids (e.g., soil) which must be
subsampled to get a reasonably sized analytical sample
must be treated on a case-by-case basis. A representa-
tive sample should be obtained by designing a scheme such
that all parts of the whole have a finite, known probabil-
ity of inclusion. Based on such a scheme, the PCB con-
tent of the sample can be used to extrapolate to the con-
tent of the whole.
9.1.4 Liquids or free-flowing solids - The source should be mixed
thoroughly before collecting the sample, if possible. If
mixing is impractical, the sample should be collected from
a representative area of the source. If the liquid is
flowing through an enclosed system, sampling through a
valve should be timed either randomly or at fixed intervals.
31
-------
Table 16. Relative Retention Time (RRT) Ranges of PCB Homologs
Versus ds-3,31,4,4'-Tetrachlorobiphenyl
No. of Projected
PCB isomers Observed range Congener Observed range of
homolog measured of RRTs no. RRT RRTs
Monochloro
3
0.40-0.50
1
0.43
0.35-0.55
Dichloro
10
0.52-0.69
7
0.58
0.35-0.80
Trichloro
9
0.62-0.79
30
0.65
0.35-0.10
Tetrachloro
16
0.72-1.01
50
0.75
0.55-1.05
Pentachloro
12
0.82-1.08
97
0.98
0.80-1.10
Hexachloro
13
0.93-1.20
143
1.05
0.90-1.25
Heptachloro
4
1.09-1.30
183
1.15
1.05-1.35
Octachloro
6
1.19-1.36
202
1.19
1.10-1.50
Nonachloro
3
1.31-1.42
207
1.33
1.25-1.50
Decachloro
1
1.44-1.45
209
1.44
1.35-1.50
The RRTs of the 77 congeners and a mixture of Aroclor 1016/1254/1260 were
measured versus 3,3' ,4,4'-tetrachlorobiphenyl-d6 (internal standard) using
a 15-m J&W DB-5 fused silica column with a temperature program of 110°C
for 2 min, then 10°C/min to 325°C, helium carrier at 45 cm/sec, and an on-
column injector. A Finnigan 4023 Incos quadrupole mass spectrometer oper-
ating with a scan range of 95-550 daltons was used to detect each PCB
^congener.
The projected relative retention windows account for overlap of eluting
homologs and take into consideration differences in operating systems and
lack of all possible 209 PCB congeners.
32
-------
9.2 Sample containers - Samples should be placed in a suitable container
which will preserve the integrity of the sample and prevent the sample
from becoming contaminated. Suitable containers are generally glass
bottles or jars, prepared according to Section 6.1. Other containers
may be used, provided they meet the sample integrity requirements.
9.3 Sample collection
9.3.1 All sample collection apparatus must be cleaned to prevent
contamination. Where possible, the cleaning and storage
specifications in Section 9.2 should be followed. Sam-
pling apparatus should be glass, steel, Teflon, or other
non-contaminating surface.
9.3.2 Water - Water samples may be collected by grab or inte-
gration techniques. Grab sampling has been much more com-
mon. The integration techniques include automatic com-
posite samplers, adsorption onto a solid (XAD resin, poly-
urethane foam, etc.), and liquid-liquid extraction. Since
water sampling for PCB analysis may be readily adapted
from general water sampling procedures for semi volatile
organics, the reader is referred to these procedures for
specific guidance.. These include EPA's Priority Pollutant
Methods 608 and 625 (EPA 1979a, 1979b; Longbottom and
Lichtenberg 1982); ASTM's (1981b) Method D 3370; the EPA
Handbook for Sampling and Sample Preparation of Water and
Wastewater (Moser and Huibregtse 1976); and an EPA docu-
ment on Procedures for Handling and Chemical Analysis of
Sediment and Water Samples (Plumb 1981).
9.3.3 Solids - Soil, sediment, biota, and tissue sampling all
follow general trace organic sampling techniques. Sam-
pling at PCB spill sites may include scraping soil or sedi-
ment surfaces, water sampling, and surface wipes. Specific
guidance has been given by Kelso (1985), USWAG (1984),
and Consent Decree (1985). Additional specific guidance
on sampling is available for soil (Mason 1982), hazardous
waste (deVera 1980); sediment (Plumb 1981).
9.3.4 Oil, dielectric fluids, etc. - A large number of PCB analy-
ses involve transformer oil, hydraulic fluid, askarels,
and other similar oils which may be contaminated with PCBs.
Since these oils are generally contained in a transformer,
drum, tank, etc., obtaining a sample may be as simple as
opening a drain valve. The only consideration in sampling
is representativeness. Especially where the oil is highly
contaminated, the PCB content may not be homogeneous.
For example, a drum of waste oil may contain several phases--
a sludge, a water layer, and an oil layer. The PCB content
of the three layers would differ markedly. Thus, the sam-
pler must attempt to obtain a representative sample. Small
33
-------
containers may be mixed. With larger containers, subsamples
from various depths may need to be taken. Subsamples may
be composited or treated separately, if compositing would
hamper subsequent analysis steps.
9.4 Sample preservation - Samples should be stored such that no PCBs
are lost, the sample is not contaminated and the matrix is not
greatly altered by biological or other action.
9.4.1 Aqueous samples should be stored at 4°C. If there is a
possibility of microbial degradation, addition of H2S04
during collection to a pH < 2 is recommended. A test strip
is used to monitor pH. Storage times in excess of 7 days
for water samples and 4 weeks for organic extracts are
not recommended.
If residual chlorine is present in water samples, it should
be quenched with sodium thiosulfate. EPA Methods 330.4
and 330.5 may be used to measure the residual chlorine
(EPA 1979c). Field test kits are available for this pur-
pose.
9.4.2 Other samples should be frozen if there is any chance of
biological degradation. This would include soil, adipose,
fish, and food samples. Samples which are inert may be
stored at room temperature. Storage times should be kept
to a minimum.
10.0 Sample Preparation
Since a wide variety of matrices may be subjected to analysis by this
method, the extraction/cleanup procedure cannot be specified. This sec-
tion describes general guidelines for subsampling, addition of 13C sur-
rogates, dilution, extraction, cleanup, extract concentration, and other
sample preparation procedures.
10.1 Sample homogenization and subsampling - The sample is homogenized
by shaking, blending, shredding, crushing, or other appropriate
mechanical technique. A representative subsample of known mass is
then taken. The sample size is dependent upon the anticipated PCB
levels and difficulty of the subsequent extraction and cleanup steps.
Note: The precision of the mass determination at this step will be
reflected in the overall method precision. Therefore, an analytical
balance must be used to assure that the weight is accurate to ± 1%
or better.
10.2 Surrogate addition - An appropriate volume of surrogate solution
SSxxx must be pipetted into the sample. The final concentration of
the surrogates must be in the working range of the calibration and
well above the matrix background. The surrogates are thoroughly
34
-------
incorporated by further mechanical agitation. For nonviscous liq-
uids, shaking for 30 s should be sufficient. For viscous liquids
or free-flowing porous solids, 10-min tumbling is recommended. In
cases where inadequate incorporation may be expected, such as soils,
overnight equilibration with agitation is recommended.
Note: The volume measurement of the spiking solution is critical
to the overall method precision. The analyst must exercise caution
that the volume is known to ± 1% or better. Where necessary, cali-
bration of the pi pet is recommended.
10.3 Sample preparation (extraction/cleanup) - After addition of the sur-
rogates, the sample is further treated at the discretion of the ana-
lyst, provided that the GC/EIMS responses of the four surrogates
are sufficient for reliable quantitation. The literature pertaining
to these techniques has been reviewed (Erickson et al. 1983, 1985;
Erickson 1985). Several possible techniques are presented below
for guidance only. The applicability of any of these techniques to
a specific sample matrix must be determined by the precision and
accuracy of the 13C PCB surrogate recoveries, as discussed in Sec-
tion 16.4.1.
10.3.1 Extraction
10.3.1.1 General techniques
10.3.1.1.1 Dilution - In some cases, where
the PCB concentration is high, a
simple volumetric dilution with an
appropriate solvent may be suffi-
cient sample preparation.
10.3.1.1.2 Direct injection - If sample vis-
cosity and composition permits,
direct injection with no dilution
is permissible.
10.3.1.1.3 Kuderna-Danish (K-D) concentration -
If the sample is a solvent with
low PCB concentration, reduction
of sample volume may be sufficient
preparation. In this case, the
sample is placed in the K-D appa-
ratus and concentrated over steam
to an appropriate volume. (Alter-
natives to K-D, such as rotary
evaporation techniques may also be
used.)
10.3.1.1.4 Evaporative concentration using
nitrogen - For smaller volume (5-50
mL) solvent samples concentration
35
-------
may be achieved by blowing sample
with a gentle stream of prepurified
nitrogen. Concentrate sample to
final volume, rinse sides of vial
with small aliquots of solvent,
and again concentrate to final vol-
ume. In some cases, where back-
ground is a problem, concentration
to dryness will eliminate inter-
ference by low-boiling compounds,
e.g., chloroform, without removing
the more volatile PCBs. After
blowing to dryness, reconstitute
with solvent and sonicate for 5
min. Care must be taken to avoid
evaporative loss of the more vola-
tile PCBs. Do not heat a dry sam-
ple for extended periods.
10.3.1.1.5 Liquid-liquid extraction - If the
matrix is aqueous (or another sol-
vent in which PCBs are only slightly
soluble), a liquid-liquid parti-
tion may be effective. The solvent,
number of extractions, solvent-to-
sample ratio, and other parameters
are chosen at the analyst's dis-
cretion.
10.3.1.1.6 Liquid-solid extraction - Solids
may be extracted with an appropri-
ate solvent by shaking, homogeniz-
ing, or any other mechanism which
assures contact of the two phases.
The primary criteria are intimate
contact and good partition ratios.
The solvent, number of extractions,
solvent-to-sample ratio, and other
parameters are chosen at the ana-
lyst's discretion.
10.3.1.1.7 Sorbent column extraction - PCBs
may be isolated from free-flowing
liquids onto sorbent columns. The
selection of sorbent (XAD, Porapak,
carbon-polyurethane foam, etc.)
will depend on the nature of the
matrix. The available methods have
been reviewed (Erickson 1985).
36
-------
10.3.1.1.8 Thermal desorption - If the matrix
is nonvolatile, thermal desorption
of the PCBs onto a sorbent column,
filter, or cold trap may be an ef-
fective extraction/cleanup method.
10.3.1.1.9 Matrix destruction - Some matrices
may be easily degraded to a volatile
or extractable compound. Once the
matrix is degraded, the PCBs can
be isolated, either by evaporation
of the matrix or partitions of the
PCBs into a nonpolar solvent and
the matrix into a polar solvent
(e.g., water). Examples include:
(1) esters which may be saponified
with base to the acid anion and
then extracted with water, and (2)
acid chlorides which may be hydro-
lyzed to the acid anion with water
or base and then extracted with
water.
10.3.1.2 Matrix-specific techniques - The following sec-
tions describe extraction techniques which are
widely accepted for specific matrices. These
are included in this procedure for the informa-
tion of analysts who need to extract the matrices
listed. Their inclusion does not constitute an
endorsement of the techniques, nor does it ex-
clude the use of other extraction techniques
for these matrices.
10.3.1.2.1 Water - In most cases, the water
sample extraction prescribed in
EPA's Method 608 and 625 is suffi-
cient [see Table 3), EPA 1979a,b;
Longbottom and Lichtenberg, 1982].
A 1-L sample is serially extracted
with three 60-mL aliquots of di-
chloromethane. The combined di-
chloromethane extracts are dried
with sodium sulfate, and concen-
trated to a final volume of 1.0 mL
by K-D evaporation.
10.3.1.2.2 Sewage, sludge, and hazardous wastes -
These ill-defined matrices range
from "dirty water" to high-organic
solids. Thus, many different ex-
traction techniques have been em-
ployed. For raw municipal sewage
37
-------
with relatively low organic content,
the sample can be extracted using
a water technique, such as solvent
extraction in a separatory funnel.
More concentrated samples can also
be treated this way by simply di-
luting the matrix to the desired
consistency—at a sacrifice of the
method detection limit.
EPA's sludge extraction methods
(Table 3, EPA 1982e) utilize di-
chloromethane extraction with me-
chanical homogenization, followed
by centrifugation to separate the
organic and aqueous phases (EPA
1979c, 1982a; Ballinger 1978; Haile
and Lopez-Avila 1984).
10.3.1.2.3 Sediment and soil - Sediment and
soil differ from sewage sludge in
their generally lower organic con-
tent. As with sewage sludge, the
critical component of an extraction
technique for either sediment or
soil is the contact between the
solvent and the matrix. This is
generally accomplished by physical
mixing (e.g., manual shaking) and
use of a "wetting" solvent mixture
(e.g., hexane and acetone). With
soil samples, especially, thorough
mixing and pulverizing of chunks
is an important preextraction step.
The ASTM (1981b, Table 3) procedure
utilizes water and acetonitrile.
Specifically, the soil or sediment
(250 g) is centrifuged to remove
excess water, air dried, wetted
with 10 mL water and extracted three
times with 150, 50 and 50 mL ace-
toni tri le.
10.3.1.2.4 Blood - The extraction of blood
samples generally entails a simple
liquid-liquid partition of whole
blood, plasma, or serum with sol-
vent. The techniques differ from
those for other aqueous liquids in
how the cells and proteins are han-
dled to assure complete extraction
38
-------
and avoid emulsions. In addition,
blood samples are generally small
(e.g., 10 mL) relative to water
and many other samples, so the me-
chanical techniques of agitation,
etc., are different.
Whole blood is essentially the same
matrix as obtained from the body,
except for the possible addition
of an anticoagulant, such as hepari
to retard clot formation. Plasma
is the liquid portion of the blood
obtained after physical removal of
the cells from whole blood by cen-
trifugation. Serum is the fluid
remaining after the cells and plate
lets are allowed to coagulate.
PCB analyses have been conducted
on all three "blood" matrices.
The results may not De comparable,
since there are unknown portion of
the PCBs in whole blood are removed
with the cells or clot.
Blood extraction procedures have
been issued by two organizations:
The U.S. Environmental Protection
Agency (Watts 1980) and the U.S.
Centers for DiseaSe Control
(Needham et al. 1981; Burse et al.
1983a). The EPA manual for pesti-
cide residue analysis (Section 5,
A,(3),(a), Watts, 1980) describes
a hexane extraction for serum:
"A 2-mL aliquot of serum is ex-
tracted with 6 mL of hexane in a
round-bottom tube. The extraction
is conducted for 2 hours on a slow-
speed rotating mixer. The forma-
tion of emulsion is unlikely, but
if it should occur, centrifugation
may be used to effect separation
of the layers. A 5-mL aliquot of
the hexane layer is quantitatively
transferred to an evaporative con-
centrator tube to which is affixed
a modified micro-Snyder column.
The extract is concentrated in a
water or steam bath, and the final
volume is adjusted to correspond
-------
to the expected concentration of
the pesticide residue. A suitable
aliquot is analyzed by electron
capture gas chromatography."
10.3.1.2.5 Adipose tissue - Adipose tissue,
from either humans or animals, con-
sists of both lipids ("fat") and
connective tissue. A classic ex-
traction procedure developed by
Mills et al. (1963) involves first
extraction of the lipids from the
connective tissue with petroleum
ether and then extraction of the
PCBs and pesticides from the petro-
leum ether into acetonitrile. Like
many other techniques, it was orig-
inally developed for organochlorine
pesticides and subsequently adapted
for PCBs. This Mills-Onley-Gaither
("MOG") procedure has formed the
basis of many techniques subsequently
reported. For example, the EPA
method (Watts 1980) for pesticides
in adipose involves:
"A 5 g. sample is dry macerated
with sand and sodium sulfate and
the fat is isolated by repetitive
extractions with petroleum ether.
Pesticide residues are extracted
from the fat with acetonitrile and
then partitioned back into petro-
leum ether by aqueous dilution of
the acetonitrile extract. Petro-
leum ether extract is concentrated
to 5 mL by K-D evaporation and trans-
ferred to a Florisil column for
successive elutions with 6% and
15% ethyl ether/petroleum ether.
The respective eluates are both
concentrated to suitable volumes
in K-D evaporators and the final
extracts are examined by electron
capture gas-liquid chromatography."
The PCBs and pesticides are sepa-
rated in the Florisil step. PCBs
are resolved from some co-eluting
pesticides by silicic acid chro-
matography. Many users of this
40
-------
method now substitute pesticide
residue grade n-hexane for the pe-
troleum ether stipulated in the
above procedure.
10.3.1.2.6 Fish - Fish samples can often be
extracted by techniques similar to
those used for adipose tissue, since
fish contain up to several percent
lipids. The extraction step usually
involves simultaneous grinding of
the tissue with the solvent to as-
sure good contact and thus efficient
extraction. This technique is em-
ployed in the AOAC (1980a) procedure
for extraction of fish (Method 29.012(e)).
In this procedure 25-50 g of fish
are ground with sodium sulfate in
a high speed blender to disinte-
grate the sample and bind any water.
The sample is then extracted three
times with petroleum ether. The
supernatant solvent is filtered,
combined, dried with sodium sulfate,
and concentrated in a K-D evaporator.
A portion of the sample is removed
for percent fat determination.
The AOAC method is widely used,
sometimes with minor modifications,
such as substitution of n-hexane
for petroleum ether.
10.3.1.2.7 Food and plant Tissue - A classic
procedure (Mills et al. 1963; FDA
1977-Section 212) for organochlorine
pesticides in nonfatty foods employs
an acetonitrile extraction of the
chopped sample, followed by a pe-
troleum ether back extraction of
the acetonitrile after dilution
with water. After blending or chop-
ping, the 100-g sample is blended
with 200-mL acetonitrile and 10 g
Celite in a blender. The acetoni-
trile is separated from the solids
by filtration. The filtrate is
diluted with 600 mL water, 10 mL
saturated sodium chloride, and ex-
tracted with 100 mL petroleum ether.
The organic layer is washed with
2 x 100 mL water and dried with
sodium sulfate. The extract is
41
-------
then cleaned with Florisil column
chromatography and analyzed by gas
chromatography. The procedure has
not been validated for PCBs, but
quantitative recoveries of several
organochlorine pesticides from green
vegetables, soups, fruits, and other
foods (Mills et al. 1963) indicate
that it would be suitable for PCBs.
10.3.1.2.8 Paper products - By virtue of their
use in carbonless copy paper and
other paper products and the exten-
sive recycling of paper products,
analysis for PCBs in paper prod-
ucts has been of considerable in-
terest. An alcoholic potassium
hydroxide reflux has been adopted
as an official method (AOAC 1980b).
The samples (10 g) are cut up, mixed,
and refluxed with 60 mL 2% potassium
hydroxide in ethanol or methanol
for 30 min. The sample is diluted
with water and repeatedly extracted
with petroleum ether. The combined
extracts are then washed with water,
dried with sodium sulfate, concen-
trated in a K-D evaporator, and
cleaned on a Florisil column.
10.3.1.2.9 Organic Oils - Hexane dilution is
recommended for analysis of trans-
former fluid and waste oils by the
U.S. EPA (1981; Bellar and Lichtenberg
1981). The procedure specifies
1:100 or 1:1000 dilution of the
oil with pesticide grade hexane.
Specifically, a 1-g sample of oil
is weighed to the nearest 0.001 g
and then diluted to 100 mL in a
volumetric flask with hexane. The
procedure recommends screening the
sample to determine the approximate
concentration by X-ray fluorescence,
microcoulometry, density measurement,
or GC screening of a very dilute
(1:10,000) sample. Samples are
then analyzed by PGC/ HECD, PGC/ECD,
or PGC/EIMS. Several optional cleanup
techniques are presented.
42
-------
10.3.2 Cleanup - Several tested cleanup techniques are described
below. All but the base cleanup (10.3.2.8) were previously
validated for PCBs in transformer fluids (USEPA 1981;
Bellar and Lichtenberg 1981). Depending upon the complex-
ity of the sample, one or more of the techniques may be
required to fractionate the PCBs from interferences. For
most cleanups a concentrated (1-5 mL) extract should be
used.
10.3.2.1 Acid cleanup
10.3.2.1.1 Place 5 mL of concentrated sulfuric
acid into a 40-mL narrow-mouth
screw-cap bottle. Add the sample
extract. Seal the bottle with a
Teflon-lined screw cap and shake
for 1 min.
10.3.2.1.2 Allow the phases to separate, trans-
fer the sample (upper phase) with
three rinses of 1-2 mL solvent to
a clean container.
10.3.2.1.3 Back-extract sample extract with
5-10 drops of distilled water.
Pass through a short column of an-
hydrous sodium sulfate and concen-
trate to an appropriate volume.
10.3.2.1.4 Analyze as described in Section
11.0.
10.3.2.1.5 If the sample is highly contaminated,
a second or third acid cleanup may
be employed.
10.3.2.2 Florisil column cleanup
10.3.2.2.1 Variations among batches of Florisil
(PR grade or equivalent) may affect
the elution volume of the various
PCBs. For this reason, the volume
of solvent required to completely
elute all PCBs must be verified by
the analyst. The weight of Florisil
can then be adjusted accordingly.
10.3.2.2.2 Place a 20-g charge of Florisil,
activated overnight at 130°C, into
a Chromaflex column. Settle the
Florisil by tapping the column.
Add about 1 cm of anhydrous sodium
43
-------
sulfate to the top of the Florisil.
Pre-elute the column with 70-80 mL
of hexane. Just before the. exposure
of the sodium sulfate layer to air,
stop the flow. Discard the eluate.
10.3.2.2.3 Add the sample extract to the column.
10.3.2.2.4 Carefully wash down the inner wall
of the column with 5 mL of hexane.
10.3.2.2.5 Add 200 mL of 6% ethyl ether/hexane
and set the flow to about 5 mL/min.
10.3.2.2.6 Collect 200 mL of eluate in a Kuderna-
Danish (K-D) flask. All the PCBs
should be in this fraction. Concen-
trate to an appropriate volume.
10.3.2.2.7 Analyze the sample as described in
Section 11.0.
10.3.2.3 Alumina column cleanup
10.3.2.3.1 Adjust the activity of the alumina
(Fisher A540 or equivalent) by heat-
ing to 200°C for at least 2 hr.
When cool, add 3% water (wt.wt)
and mix until uniform. Allow the
deactivated alumina to equilibrate
at least 1/2 h before use. Store
in a tightly sealed bottle.
10.3.2.3.2 Variations between batches of alum-
ina may affect the elution volume
of the various PCBs. For this rea-
son, the volume of solvent required
to completely elute all of the PCBs
must be verified by the analyst.
The weight of alumina can then be
adjusted accordingly.
10.3.2.3.3 Place a 50-g charge of alumina into
a Chromaflex column. Settle the
alumina by tapping. Add about 1
cm of anhydrous sodium sulfate.
Pre-elute the column with 70-80 mL
of hexane. Just before exposure
of the sodium sulfate layer to air,
stop the flow. Discard the eluate.
44
-------
10.3.2.3.4 Add the sample extract to the column.
10.3.2.3.5 Carefully wash down the inner wall
of the column with 5 mL of hexane.
10.3.2.3.6 Add 295 mL of hexane to the column.
10.3.2.3.7 Discard the first 50 mL.
10.3.2.3.8 Collect 250 mL of the hexane in a
Kuderna-Danish flask. All of the
PCBs should be in this fraction.
Concentrate to an appropriate volume.
10.3.2.3.9 Analyze the sample as described in
Section 11.0.
10.3.2.4 Silica gel column cleanup
10.3.2.4.1 Activate silica gel (Davison Grade
950 or equivalent) at 135°C over-
night.
10.3.2.4.2 Variations between batches of silica
gel may affect the elution volume
of the various PCBs. For this rea-
son, the volume of solvent required
to completely elute all of the PCBs
must be verified by the analyst.
The weight of silica gel can then
be adjusted accordingly.
10.3.2.4.3 Place a 25-g charge of activated
silica gel into a Chromaflex column.
Settle the silica gel by tapping
the column. Add about 1 cm of an-
hydrous sodium sulfate to the top
of the silica gel.
10.3.2.4.4 Pre-elute the column with 70-80 mL
of hexane. Discard the eluate.
Just before exposing the sodium
sulfate layer to air, stop the flow.
10.3.2.4.5 Add the sample extract to the column.
10.3.2.4.6 Wash down the inner wall of the
column with 5 mL of hexane.
10.3.2.4.7 Elute the PCBs with 195 mL of 10%
diethyl ether in hexane (v:v).
45
-------
10.3.2.4.8 Collect 200 mL of the eluate in a
K-D flask. All of the PCBs should
be in this fraction. Concentrate
to an appropriate volume.
10.3.2.4.9 Analyze the sample as described in
Section 11.0.
10.3.2.5 Gel permeation cleanup
10.3.2.5.1 Set up and calibrate the gel perme-
ation chromatograph with an SX-3
column according to the Autoprep
instruction manual. Use 15% meth-
ylene chloride in cyclohexane (v:v)
as the mobile phase.
10.3.2.5.2 Inject 5.0 mL of the sample extract
into the instrument. Collect the
fraction containing the PCBs (see
Autoprep operator's manual) in a
K-D flask equipped with a 10-mL
ampule.
10.3.2.5.3 Concentrate the PCB fraction to an
appropriate volume.
10.3.2.5.4 Analyze the sample as described in
Section' 11.0.
10.3.2.6 Acetonitrile partition
10.3.2.6.1 Place the sample extract into a
125-mL separatory funnel with enough
hexane to bring the final volume
to 15 mL. Extract the sample four
times by shaking vigorously for
1 min with 30-mL portions of hexane-
saturated acetonitrile. Retain
hexane layer for combination with
other hexane extracts in 10.3.2.6.3.
10.3.2.6.2 Combine and transfer the acetonitrile
phases to a 1-L separatory funnel
and add 650 mL of distilled water
and 40 mL of saturated sodium chloride
solution. Mix thoroughly for about
30 sec. Extract with two 100-mL
portions of hexane by vigorously
shaking about 15 sec.
46
-------
10.3.2.6.3 Combine the hexane extracts in a
1-L separatory funnel and wash with
two 100-mL portions of distilled
water. Discard the water layer
and pour the hexane layer through
an 8-to 10-cm anhydrous sodium sul-
fate column into a 500-mL K-D flask
equipped with a 10-mL ampul. Rinse
the separatory funnel and column
with three 10-mL portions of hexane.
10.3.2.6.4 Concentrate the extracts to an ap-
propriate volume.
10.3.2.6.5 Analyze as described in Section
11.0.
10.3.2.7 Florisil slurry cleanup
10.3.2.7.1 Place the sample extract into a
20-mL narrow-mouth screw-cap con-
tainer. Add 0.25 g of Florisil
(PR grade or equivalent). Seal
with a Teflon-lined screw cap and
shake for 1 min.
10.3.2.7.2 Allow the Florisil to settle; then
decant the treated solution into a
second container with rinsing.
Concentrate the sample to an appro-
priate volume. Analyze as described
in Section 11.0.
10.3.2.8 Base cleanup (ASTM 1981b)
10.3.2.8.1 Quantitatively transfer the concen-
trated extract to a 125-mL extrac-
tion flask with the aid of several
small portions of solvent.
10.3.2.8.2 Evaporate the extract just to dry-
ness with a gentle stream of dry
filtered nitrogen, and add 25 mL
of 2.5% alcoholic potassium hydroxide.
10.3.2.8.3 Add a boiling chip, put a water
condenser in place, and allow the
solution to reflux on a hot plate
for 45 min.
10.3.2.8.4 After cooling, transfer the solution
to a 250-mL separatory funnel with
25 mL of distilled water.
47
-------
10.3.2.8.5 Rinse the extraction flask with 25
mL of hexane and add it to the sepa-
ratory funnel.
10.3.2.8.6 Stopper the separatory funnel and
shake vigorously for at least 1
min. Allow the layers to separate,
and transfer the lower aqueous phase
to a second separatory funnel.
10.3.2.8.7 Extract the saponification solution
with a second 25-mL portion of hexane.
After the layers have separated,
add the first hexane extract to
the second separatory funnel and
transfer the aqueous alcohol layer
to the original separatory funnel.
10.3.2.8.8 Repeat the extraction with a third
25-mL portion of hexane. Discard
the saponification solution, and
combine the hexane extracts.
10.3.2.8.9 Concentrate the hexane layer to an
appropriate volume, and analyze as
described in Section 11.0.
10.3.2.9 High performance liquid chromatographic cleanup
10.3.2.9.1 Quantitatively transfer the concen-
trated extract into the sample loop
or the barrel of a syringe. Rinse
the vial with several small portions
of solvent. It may be necessary
to inject several fractions.
10.3.2.9.2 Inject the extract and washes onto
the amine column (Waters pBondapak
3.9 x 300 mm or equivalent) and
elute the PCBs with 1.0 mL/min hexane.
The UV at 254 nm or lower should
be monitored.
10.3.2.9.3 Collect the eluent from 3 min to
9.5 min as it exits UV cell. The
elution time should be verified
using PCB standards covering a range
from monochlorobiphenyls to deca-
chlorobiphenyls.
48
-------
10.3.2.9.4 After collection, wash the column
by eluting with methylene chloride
until the absorbance attains a sta-
ble minimum. Return the system to
hexane.
10.3.2.9.5 Concentrate the hexane eluate under
a gentle stream of purified nitrogen
to an appropriate volume and analyze
as described in Section 11.0.
10.4 Optional screening for interferences using GC/FID
Note: Since many sample matrices are one of a kind or infrequently
encountered by the analyst, the effectiveness of the extraction and
cleanup for a matrix may be unknown. A simple screen to assess whe-
ther the interferences have been reduced to a tolerable level can
both save GC/MS time and prevent contamination of the GC/MS instru-
ment with very dirty samples. This screen should not be used to
determine PCB levels under this analytical method.
10.4.1 Using a GC system as described in Section 5.5.3, analyze
for background interferences.
10.4.2 A 2 m x 2 mm glass column packed with 3% SP-2250 on 100/
120 Supelcoport or equivalent is suggested. A flow rate
of 40 mL/min 95% air/5% methane or nitrogen is recommended.
The air and hydrogen flow rates should be sufficient to
keep the flame lit and to burn efficiently, e.g., 300 mL/
min air and 30 mL/min H2.
10.4.3 The recommended temperature program is from 50°C to 250°C
at 20°C/min with an initial hold of 3 min and a final hold
of 10 min. The injector temperature should be 200°C and
the detector 300°C.
10.4.4 Set instrumental sensitivity comparable to the anticipated
mass spectral sensitivity. It is advisable to establish
criteria for rejection of samples at a given attenuation
such as (1) any off-scale peaks in PCB elution window,
(2) a baseline rise of over 40% full-scale, and (3) other
criteria which are indicative of "problem" samples.
10.4.5 If the FID screen suggests that the sample is not amen-
able to analysis by GC/EIMS, the analyst may either (1)
cycle the sample through the same cleanup again if it ap-
pears that the cleanup technique was overloaded by the
matrix the first time, (2) submit the extract to another
cleanup technique which may remove more interferences, or
(3) analyze a new aliquot of sample by another extraction
or cleanup technique.
49
-------
11-0 Gas Chromatographic/Electron Impact Mass Spectrometric Determination
11.1 Internal standard addition - An appropriate volume of the internal
standard solution is pipetted into the sample. The final concen-
tration of the internal standard must be in the working range of
the calibration and well above the matrix background. The internal
standard is thoroughly incorporated by mechanical agitation.
Note: The volumetric measurement of the internal standard solution
is critical to the overall method precision. The analyst must exer-
cise caution that the volume is known to be ± 1% or better. Where
necessary, calibration of the pipet is recommended.
11.2 Tables 5, and 9 through 14 summarize the recommended operating con-
ditions for analysis. Figure 1 presents an example of a chromatogram.
The analyst may choose to operate the mass spectrometer at any appro-
priate sensitivity, using either full scan, limited mass scanning
or selected ion monitoring acquisition. The sensitivity selected
will depend on anticipated PCB levels and the instrumental LOQ needed
to meet the required method LOQ. In general, the more concentrated
the PCBs, the greater the precision, accuracy, and qualitative data
confidence. Thus, if possible, the amount of sample and the concen-
tration factor should be scaled so that full scan acquisition may
be utilized.
11.3 While the highest available chromatographic resolution is not a
necessary objective of this method, good chromatographic performance
is recommended. With the high resolution of HRGC, the probability
that the chromatographic peaks consist of single compounds is higher
than with PGC. Thus, qualitative and quantitative data reduction
should be more reliable.
11.4 After performance of the system has been certified for the day and
all instrument conditions set according to Tables 5, and 9 through
14, inject an aliquot of the sample onto the GC column. If the re-
sponse for any ion, including surrogates and internal standards,
exceeds the working range of the system, dilute the sample and re-
analyze. If the responses of surrogates, internal standards, or
analytes are below the working range, recheck the system performance.
If necessary, concentrate the sample and reanalyze.
11.5 Record all data on a digital storage device (magnetic disk, tape,
etc.) for qualitative and quantitative data reduction as discussed
below.
11.6 The instrumental performance must be monitored from run-to-run.
The areas of internal standards should be consistent (e.g., ± 20%).
If a low area is encountered, the injection may be suspect, unless
matrix affects account for the lower response.
50
-------
100.0-1
88960.
202
C18D12
+
13
C12H2C18
RIC
13
C12°6C14
+
C12H6C14
,3c612c6h9ci
c)0h7i
50
—I—
800
13:20
30
97
1000
16i 40
—I—
1200
20100
183
143
1
1400
23:20
209
13 +
C12C110
207
~~1
1600
26:40
—f
1800
30:00
SCAN
TIME
Figure 1. Reconstructed ion chromatogram of calibration solution FS100 ng PCB obtained in the full scan mode. The
concentration of the 10 PCB calibration congeners, the 4 l^c-iabeled PCB recovery surrogates, and the 3 internal
standards are in Table 6. See Table 2 for PCB numbering system, Table 9 for capillary GC parameters, and Table 11
for mass spectrometer operating parameters.
-------
The resolution and peak shape of the internal standards, surrogates,
and other peaks should be monitored during or immediately after data
acquisition. Poor chromatography may indicate a bad injection, ma-
trix interferences, or column degradation.
11.7 If a "dirty" sample is encountered, the analyst must employ appro-
priate measures to demonstrate that there is no memory or carry-ons
to subsequent samples. To assess the system cleanliness, a standard,
blank sample, or solvent blank may be run.
If the system is contaminated, remedial efforts may include (1) chang-
ing or cleaning the syringe, (2) cleaning the injector, (3) baking
out the column at its maximum temperature, (4) changing to a new
column, or (5) cleaning the ion source.
12.0 Qualitative Identification
12.1 Ful1 scan data
12.1.1 The peak must elute within the retention time windows set
for that homolog (as described in Section 8.5).
12.1.2 The unknown spectrum should be compared to that of an au-
thentic PCB. The intensity of the three largest ions in
the molecular cluster (two largest for monochlorobiphenyls)
must match the ratio observed for a standard within ± 20%.
Fragment clusters with proper intensity ratios should also
be present. System noise at low concentration or inter-
ferences may skew the ion ratio beyond the ± 20% criteria.
If the analyst's best judgement is that a peak, which does
not meet the qualitative criteria, is a PCB, the peak may
be included in the calculation, with a footnote explaining
the data and the reason for relaxing the criteria.
12.1.3 Alternatively, a spectral search may be used to automati-
cally reduce the data. The criteria for acceptable iden-
tification include a high index of similarity.
12.2 Selected ion monitoring (SIM) or limited mass scan (LMS) data - The
identification of a compound as a given PCB homolog requires that
two criteria be met:
12.2.1 (1) The peak must elute within the retention time window
set for that homolog (Section 8.5); and (2) the ratio of
two ions obtained by LMS (Table 13) or by SIM (Table 14)
must match the ratio observed for a standard within ± 20%.
The analyst should search the higher mass windows, in par-
ticular M+70, to prevent mis identification of a PCB frag-
ment ion cluster as the parent. System noise at low con-
centration or interferences may skew the ion ratio beyond
the ± 20% criteria. If the analyst's best judgement is
that a peak, which does not meet the qualitative criteria,
52
-------
is a PCB, the peak may be included in the calculation,
with a footnote explaining the data and the reason for
relaxing the criteria.
12.2.2 If one or the other of these criteria is not met, inter-
ferences may have affected the results, and a reanalysis
using full scan EIMS conditions is recommended.
12.3 Disputes in interpretation - Where there is reasonable doubt as to
the identity of a peak as a PCB, the analyst must either identify
the peak as a PCB or proceed to a confirmational analysis (see Sec-
tion 14.0).
.0 Quantitative Data Reduction
13.1 After a chromatographic peak has been identified as a PCB, the com-
pound is quantitated based on the integrated abundance of either
the EICP or the SIM data for the primary characteristic ion in
Tables 13 and 14. If interferences are observed for the primary
ion, use the secondary and then tertiary ion for quantitation. If
interferences in the parent cluster prevent quantitation, an ion
from a fragment cluster (e.g., M-70) may be used. Whichever ion is
used, the RF must be determined using that ion. The same criteria
should be applied to the surrogate compounds (Table 14).
Note: With the higher homologs, the mass defect from unity is sig-
nificant. For instance, the mass of the most intense peak for de-
cachlorobiphenyl is 497.6830. Areas, EICPs, etc. must be based on
the true mass, not the nominal mass, or erroneous results may be
obtained. In addition, the tuning of some quadrupoles may be less
stable at high masses. The data quality must be monitored especial
carefully for the higher homologs.
13.2 Using the appropriate analyte-internal standard pair and response
factor (RF ) as determined in Section 8.3, calculate the concentra-
tion of each peak using Equation 13-1 or Equation 13-2.
13.2.1 Aqueous samples (water wastewater, blood, etc.) should be
reported as micrograms per liter (pg/L), using Equation
13-1.
A - M. V
Concentration (pg/L) = a ' RF~ ' ~\T^ ' ~\T Eq' 13-1
is p s i
where A = area of the characteristic ion for the analyte PCB
p peak
A^ = area of the characteristic ion for the internal
standard peak
RFp = response factor of a given PCB congener
53
-------
M. = mass of internal standard injected (micrograms)
Vg = volume of sample extracted (liters)
= volume injected (microliters)
Vg = volume of sample extract (microliters)
13.2 Solids and nonaqueous liquid samples (soil, fish, oil, still bottoms,
etc.) should be reported in micrograms per gram (pg/g) using Equation
13-2.
A 1 M. V
Concentration (pg/g) = a ' TT ' ' ~v Eq" 13-2
is p s i
where = mass of sample extracted (grams)
13.3 If a peak appears to contain non-PCB interferences, which cannot be
circumvented by a secondary or tertiary ion, either:
13.3.1 Reanalyze the sample on a different column which separates
the PCB and interferents;
13.3.2 Perform additional chemical cleanup (Section 10) and then
reanalyze the sample; or
13.3.3 Quantitate the entire peak as PCB.
13.4 Calculate the recovery of the four 13C surrogates using the appro-
priate surrogate-internal standard pair and response factor (RF. )
as determined in Section 8.4 using Equation 13-3. 1s
A .. M.
Recovery (%)=—.-—. -IS . ioo Eq. 13-3
is s s
where Ag = area of the characteristic ion for the surrogate peak
RFg = response factor for the surrogate compound with respect
to the internal standard (Equation 8-2)
Mg = mass of surrogate, assuming 100% recovery (nanograms)
Other terms are the same as defined in Equations 13.1 and 13-2.
13.5 Sum all of the peaks for each homolog, and then sum those to yield
the total PCB concentration in the sample. Report all numbers in
pg/L (aqueous samples) or pg/g (solids and nonaqueous liquids).
The concentrations added and percent recovery of the four surrogates
are to be reported.
13.7 Round off all numbers reported to two significant figures.
54
-------
14.0 Confirmation
If there is significant reason to question the qualitative identifica-
tion (Section 12), the analyst may choose a confirmatory technique to
provide more information. Any technique may be chosen provided that it
is validated as having equivalent or superior selectivity and sensitivity
to GC/EIMS. Some candidate techniques include alternate GC columns (with
EIMS detection), GC/CIMS, GC/NCIMS, high resolution EIMS, and MS/MS tech-
niques. Each laboratory must validate confirmation techniques to show
equivalent or superior selectivity between PCBs and interferences and
sensitivity (limit of quantitation, LOQ). If a peak is confirmed as being
a non-PCB, it may be deleted from the calculation (Section 13).
15.0 Quality Assurance
Each participating laboratory must develop a quality assurance plan (QAP)
according to EPA guidelines (USEPA 1980). Additional guidance is also
available (USEPA 1983). The qual.ity assurance plan must be submitted to
the Agency (regional QA officer) for approval prior to analysis of samples.
The elements of a QAP include:
• Title Page
• Table of Contents
• Project Description
• Project Organization and Responsibility
QA Objectives for Measurement Data in Terms of Precision, Accuracy,
Completeness, Representativeness, and Comparability
• Sampling Procedures
• Sample Custody
• Calibration Procedures and Frequency
• Analytical Procedures
• Data Reduction, Validation and Reporting
• Internal Quality Control Checks
• Performance and System Audits
• Preventive Maintenance
Specific Routine Procedures Used to Assess Data Precision, Accuracy
and Completeness
55
-------
• Corrective Action
• Quality Assurance Reports to Management
16.0 Quality Control
16.1 Each laboratory that uses this method must operate a formal quality
control (QC) program. The minimum requirements of this program con-
sist of an initial and continuing demonstration of acceptable lab-
oratory performance by the analysis of check samples. The labora-
tory must maintain performance records to define the quality of data
that are generated.
16.2 Certification and performance checks - Prior to the analysis of sam-
ples, the laboratory must define its routine performance. At a mini-
mum, this must include demonstration of acceptable response factor
precision with at least three replicate analyses or calibration curve
over at least three levels (Section 8.3); and analysis of a blind
QC check sample (e.g., the response factor calibration solution at
unknown concentration submitted by an independent QA officer). Ac-
ceptable criteria for the response factor precision and the accuracy
of the QC check sample analysis must be presented in the QA plan.
Ongoing performance checks should include periodic repetition of
the initial demonstration or more elaborate measures. More elaborate
measures may include control charts and analysis of QC check samples
containing unknown PCBs, possibly with matrix interferences.
16.3 Procedural QC - The various steps of the analytical procedure should
have quality control measures. These include but are not limited
to:
16.3.1 GC performance - See Section 8.1 for performance criteria.
16.3.2 MS performance - See Section 8.2 for performance criteria.
16.3.3 Qualitative identification - At least 10% of the PCB iden-
tifications, as well as any questionable results, should
be confirmed by a second mass spectrometrist.
16.3.4 Quantitation - At least 10% of all manual calculations,
including peak area calculations, must be checked. After
changes in computer quantitation routines, the results
should be manually checked.
16.4 Sample QC - Each sample and each sample set must have QC measures
applied to it to establish the data quality for each analysis re-
sult. The recoveries of the surrogates, general spectral data qual-
ity, and consistency of the internal standard area are all measures
of the data quality on individual samples. Within a sample set,
analysis of replicates and standard addition samples are measures
of the precision and accuracy, respectively.
56
-------
16.4.1 The surrogate recoveries must be within an acceptable range,
as defined in the QAP, for each sample. Typically, re-
coveries below 50% are unacceptable, antf indicate that
the sample(s) must be reanalyzed. Recoveries much greater
than 100% indicate interferences, improper tuning, a pro-
blem with response factors, or an error in the concentra-
tion of the surrogate compound. The trend of low recov-
eries can be indicative of the cause of the loss. If the
lower homologs are poorly recovered, but the higher homo-
logs are recovered quantitatively, volatility or chemical
degradation losses may be suspected. If higher homologs
are selectively lost, or if the losses are irregular, a
fractionation cut on a chromatographic cleanup would be a
likely suspect to account for the loss.
16.4.2 The general spectral data quality is indicative of the
overall reliability of the data for a sample. The levels
of the background, intensity ratios within chlorine clus-
ters, etc., must all be evaluated. If the data quality
is marginal, the analyst may footnote results with an ex-
planation regarding any doubts about the data quality.
If the data are unacceptable (see Section 12.0), either
the GC/MS determination or the entire analysis must be
repeated.
16.4.3 An easy and significant assessment of the data quality is
the consistency of the internal standard areas. If the
internal standard area is consistent, the injection volume
was correct and the system is operating within general
tolerances (i.e., the chromatography column is transmitting
compounds and the spectrometer is detecting them). If
the internal standard area does not meet the criteria spe-
cified in the QAP (e.g., ± 20% of other injections), the
data must be reviewed. If the injection or the GC/MS per-
formance is suspect, the sample should be reanalyzed, or
other corrective action taken.
16.4.4 QC for small sample sets - For small sample sets (1-10
samples), the minimum QC requirements can be a heavy bur-
den. Analysts are encouraged to be efficient and group
similar samples to increase the size of a set. A set is
defined as a group of samples analyzed together by the
same extraction/cleanup technique and determined on the
GC/MS system on the same day or successive days under the
same conditions.
At least one method blank must be run. The blank must be
exposed to the same sources of contamination--solvent,
glassware, etc.--as the samples. If conditions change,
additional blanks must be generated. An example would be
a new lot of solvent, or a change in dishwashing protocol.
57
-------
At least one sample must be run in replicate. Triplicates
are preferable, but duplicates may be acceptable. The
acceptable precision among replicates must be specified
in the QAP.
At least one sample must be analyzed by the standard addi-
tion technique. The analyst may select the most difficult
sample, based on prior knowledge of the sample set, or a
random sample. Two aliquots of the sample are analyzed,
one "as is" and one spiked (surrogate spiking and equili-
bration techniques are described in Section 10.2) with
Solution FSxxx ng PCB or SIxxx pg PCB. If the analyst
has no prior knowledge of the sample, the spiking level
should be in the middle of the calibrated range for the
mass spectrometer. If the concentrations of PCBs are known
to be high or low, the amount added should be adjusted so
that the spiking level is 1.5 to 4 times the measured PCB
level in the unspiked sample. The samples should be ana-
lyzed together and the quantitative results calculated.
The recovery of the spiked compounds (calculated by dif-
ference) must be 70-130%. If the sample is known to con-
tain specific PCB isomers, these isomers may be substituted
for solution FSxxx ng PCB or SIxxx pg PCB.
16.4.5 QC for intermediate sample sets - With intermediate (ap-
proximately 10-100 samples) sample sets, the number of
method blanks, replicates, and standard addition samples
must comprise at least 10% each.
16.4.6 QC for large sample sets - When a large sample analysis
program is being planned, the QA plan may propose specific
QC measures. If none are proposed, the guidelines for
intermediate sets may be followed. One QC measure which
may increase efficiency is the use of control charts.
If, for example, the control charts establish that there
is no blank problem over the long term, the percent of
blanks may be reduced. Any changes in the procedure (e.g.,
a new lot of solvent) will still, of course, require a
blank.
16.5 It is recommended that the participating laboratory adopt additional
QC practices for use with this method. The specific practices that
are most productive depend upon the needs of the laboratory and the
nature of the samples. Field duplicates or triplicates may be ana-
lyzed to monitor the precision of the sampling technique. Whenever
possible, the laboratory should perform analysis of standard refer-
ence materials and participate in relevant performance evaluation
studies.
58
-------
17.0 Method Performance
The method has been evaluated for by-product PCBs using both intra- and
interlaboratory studies (Erickson et al. 1982, 1983c). Both studies must
be considered preliminary; only a limited number of matrices and method
options have been investigated. The preliminary interlaboratory study
involved only four participants and was conducted before the method was
ruggedized, so data from that study may not be representative of the po-
tential method performance. Further intra- and interlaboratory validation
is anticipated. Preliminary values for limits of quantitation; intrala-
boratory recoveries, precision, and accuracy; and interlaboratory recov-
eries, precision, and accuracy are presented in Table 17. The values in
the "best" and "worst" columns represent typical extremes of the measure-
ment, ignoring exceptional cases. For example, some mass spectrometers,
when running under optimal conditions, can probably quantitate less than
0.5 ng/|jL in the full scan mode for some matrices and PCB congeners. On
the other hand, the method LOQ will be much greater than the stated value
if a high concentration coeluting interference completely obscures any
PCB signal, even at the percent level.
The performance values in Table 17 were derived using either the SIM or
LMS options of the method and at concentrations at the lower end of the
working range. Working in the middle-to-upper concentration ranges with
full scan data collection, the precision, recovery, and accuracy should
all improve considerably.
The values in Table 17 represent best estimates of the parameters and
will be refined as additional intrar and interlaboratory studies produce-
more data. Performance better than or worse than any of these parameters
cannot, of itself, be construed as grounds for acceptable or unacceptable
data quality. Performance criteria must be stipulated in the QA Project
Plan (see Section 15).
18.0 Documentation and Records
Each laboratory is responsible for maintaining full records of the anal-
ysis. A detailed documentation plan should be prepared as part of the
QAP. Laboratory notebooks should be used for handwritten records. GC/
MS data must be archived on magnetic tape, disk, or a similar device.
Hard copy printouts may be kept in addition if desired. QC records should
be maintained separately from sample analysis records.
The documentation must describe completely how the analysis was performed.
Any variances from the protocol must be noted and fully described. Where
the protocol lists options (e.g., sample cleanup), the option used and
specifics (solvent volumes, digestion times, etc.) must be stated.
The remaining samples and extracts should be archived for at least 2 months
or until the analysis report is approved, whichever is longer, and then
disposed unless other arrangements are made. The magnetic tapes of the
analysis and hardcopy spectra, quantitation reports, work sheets, etc.,
must be archived for at least 3 years.
59
-------
Table 17. Method Performance Parameters
Measurement Best case Worst case Typical value
Instrumental LOQ (ng/pL)
Full scan 0.5 10 1
Limited mass scan 0.1 0.5 0.05
Selected ion monitoring 0.005 0.5 0.025
Sample concentration factor 1,000+1 1+100 -a
Matrix interference level (ng/pL) <0.01 > 1,000 -a
Sample injection volume (pL) 10 0.1 1
Method LOQ (jjg/g) 10'5 500 1
Intralaboratory*3 ,
Recovery (%) 90 ± 15c < 50, > 200° 70-130
Precision (%) - ± 34%
Accuracy - -
b 6
Interlaboratory '
f
Recovery (%) - 22-690
Precision (%) - ± 60
Accuracy - ± 62^
bVaries widely from matrix-to-matrix, no "typical value."
Data for preliminary validation at low (< 1 pg/g) levels using selected ion
monitoring mass spectrometry.
"Methods of Analysis for By-Product PCBs--Preliminary Validation and Interim
Methods," M. D. Erickson, J. S. Stanley, G. Radolovich, K. Turman, K. Bauer,
J. Onstot, D. Rose, and M. Wickham, Interim Report No. 4, Washington D.C.:
Office of Toxic Substances, EPA-500/5-82-006, October 1982, 243 pp. NTIS No.
.PB83 127 696.
Values outside the 50-200% range are generally considered unacceptable and
gthe analysis must be repeated.
Preliminary interlaboratory study involving four participants: "Analytical
Methods for By-Product and Destruction Derived PCBs--Interlaboratory Valida-
tion A," M. 0. Erickson, K. M. Bauer, and F. J. Bergman, Draft Interim Report
No. 6, Washington, D.C.: Office of Toxic Substances, U.S. Environmental Pro-
tection Agency, Draft Interim Report No. 6, Task 51, EPA Contract No. 68-01-
^5915, August 1983.
Based on a mean of the 10 homolog values from the analysis of two samples by
four laboratories.
^Based on the deviation of the reported values from the prepared value for 10
homologs x 2 samples x 4 laboratories.
60
-------
REFERENCES
American National Standards Institute, Inc. 1974. American national stan-
dard guidelines for handling and disposal of capicator- and transformer-grade
askarels containing polychlorinated biphenyls ANSI C107.1-1974. New York,
NY.
ANSI/ASTM. 1983. Analysis of polychlorinated biphenyls in mineral insulating
oils by gas chromatography. In: Annual Book of ASTM Standards. Part 40.
Philadelphia, Pennsylvania.
ASTM. 1981a. Standard method for polychlorinated biphenyls (PCBs) in water
ANSI/ASTM D 3534-80. In: Annual Book of ASTM Standards. Part 31. Philadelphia,
PA.
ASTM. 1981b. Standard method for analysis of environmental materials for
polychlorinated biphenyls ASNI/ASTM D 3304-77. In: Annual Book of ASTM
Standadrds. Part 31. Philadelphia, PA.
ASTM. 1980. Standard practice for sampling water. In: ASTM Annual Book of
Standards. Part 31. Philadelphia, PA.
Association of Official Analytical Chemists. 1980a. General method for or-
ganochlorine and organophosphorus pesticides, method 29.001. In: Official
Methods of Analysis of the Association of Official Analytical Chemists.
Horwitz W. ed. Washington 0C.
Association of Official Analytical Chemists. 1980b. PCP in paper and paper-
board, method 29029. Official Methods of Analysis of the Association of Offi-
cial Analytical Chemists. Horwitz W. ed. Washington DC.
Ballinger DG. 1978. Test procedures for priority organics in municipal waste-
water and sludges. Cincinnati, OH: U.S. Environmental Protection Agency.
Beard JH, III, Schaum J. 1978. Sampling methods and analytical procedures
manual for PCB disposal: interim report, revision 0. Interim report.
Washington, DC: Office of Solid Waste, U.S. Environmental Protection Agency.
Bellar TA, Lichtenberg JJ. 1981. The determination of polychlorinated bi-
phenyls in transformer fluid and waste oils. Prepared for U.S. Environmental
Protection Agency, EPA-600/4-81-045.
Brinkman U. , De Kok A. 1980. Halogenated biphenyls, terphenyls, naphthal-
enes, dibenzodioxins and related products. Production, properties and usage.
Chapter 1. Kimbrough RD, ed. New York, NY: Elsevier/North-Holland Biomedi-
cal Press.
Burse VW, Needham LL, Korver MP, Lapeza CR Jr, Liddle JA, Bayse DD. 1983a.
Gas-liquid chromatographic determination of polychlorinated biphenyls and a
selected number of chlorinated hydrocarbons in serum. J. Assoc. Off. Anal.
Chem. 66(1):32-39.
61
-------
Consent Decree. Civil Action No. 34 C1597. 1984. U.S. District Court for
the Northern District of Illinois, Eastern Division.
Devenish IW, Harling-Bowen L. 1980. The examination and estimation of the
performance characteristics of a standard method for organo-chlorine insecti-
cides and PCB. In: Hydrocarbons and Halogenated Hydrocarbons in the Aquatic
Environment. Afghan BK, Mackay D, eds. New York, NY: Plenum Press.
deVera ER, Simmons BP, Stephens RD, Storm DL. 1980. Samplers and sampling
procedures for hazardous waste streams. Cincinnati, OH: Municipal Environ-
mental Research Laboratory, U.S. Environmental Protection Agency. EPA-600/
2-80-018.
Dow Chemical Company. 1981. Determination of chlorinated biphenyls in the
presence of chlorinated benzenes. Midland, MI.
Dry Color Manufacturers Association. 1982. An analytical procedure for the
determination of polychlorinated biphenyls in dry phthalocyanine blue, phthalo-
cyanine green, and diarylide yellow pigments. Arlington, VA.
Erickson MD, Stanley JS. 1982. Methods of analysis for incidentally generated
PCBs--literature review and preliminary recommendations. Washington, DC:
Office of Toxic Substances, U.S. Environmental Protection Agency. EPA-560/5-
82-005; NTIS No. PB83 126573.
Erickson MD, Stanley JS, Radolovich G,'Turman K, Bauer K, Onstot J, Rose D,
Wickham M. 1982. Analytical methods for by-product PCBs—preliminary vali-
dation and interim methods. Washington, DC: Office of Toxic Substances, U.S.
Environmental Protection Agency. EPA-500/5-82-006; NTIS No. PB83 127 696.
Erickson MD, Stanley JS, Radolovich G, Blair RB. 1983a. Analytical method:
the analysis of by-product chlorinated biphenyls in commercial products and
product wastes. Revision 1, Prepared by Midwest Research Institute for Office
of Toxic Substances, U.S. Environmental Protection Agency, Washington, DC,
under Subcontract No. A-3044(8149)-271, Work Assignment No. 17 to Battelle,
Washington, DC, August 15, 1983.
Erickson MD, Stanley JS, Turman JK, Going JE, Redford DP. 1983b. Methods
for analysis of complex mixtures in complex, variable matrices—chlorinated
biphenyls in commercial products and wastes." In: Polynuclear Aromatic Hy-
drocarbons: Formation, Metabolism, and Measurement. Cooke M, Dennis AJ, eds.
Columbus, OH: Battelle Press.
Erickson MD, Bauer KM, Bergman FJ. 1983c. Analytical methods for by-product
and destruction derived PCBs--interlaboratory validation A. Draft Interim
Report No. 6, Task 51, EPA Contract No. 68-01-5915, Office of Toxic Substances,
U.S. Environmental Protection Agency, Washington, DC.
62
-------
Erickson MD. 1984a. Analytical method: the analysis of by-product chlor-
inated biphenyls in commercial products and product wastes, revision 2.
Washington, DC: Office of Toxic Substances, U.S. Environmental Protection
Agency. EPA 560/5-85-010.
Erickson MD. 1984b. Analytical method: the analysis of by-product chlor-
inated biphenyls in air, revision 2. Washington, DC: Office of Toxic Sub-
stances, U.S. Environmental Protection Agency. EPA 560/5-85-011.
Erickson MD. 1984c. Analytical method: the analysis of by-product chlor-
inated biphenyls in water, revision 2. Washington, DC: Office of Toxic Sub-
stances, U.S. Environmental Protection Agency. EPA 560/5-85-012.
Erickson MD, Stanley JS, Turman JK, Going JE, Redford DP. 1985. Determina-
tion of by-product PCBs in commercial products and wastes by high resolution
GC/EIMS. Anal. Chem. (submitted).
Erickson MD. 1985. Analytical chemistry of PCBs. Boston, MA: Butterworths
(in press).
Food and Drug Administration. 1977. Pesticide analytical manual.
Haile CL, Lopez-Avila V. 1984. Development of analytical test procedures
for the measurement of organic priority pollutants in sludge. Project sum-
mary. Cincinnati, OH: Environmental Monitoring and Support Laboratory, U.S.
Environmental Protection Agency. EPA-600/S4-84-001; NTIS No. PB 84-129 048.
Kelso GL, Erickson MD, Boomer BA, Swanson SE, Cox DC, Schultz BD. 1985.
Cleanup of PCB spills from capacitors and transformers. Draft report no. 1,
revision no. 1. Washington DC: Office of Toxic Substances, U.S. Environ-
mental Protection Agency. Contract 68-02-3938.
Longbottom JE, Lichtenberg JJ. 1982. Methods for organic chemical analysis
of municipal and industrial wastewater. U.S. Environmental Protection Agency.
Report EPA-600/4-82-057.
Mason BJ. 1982. Preparation of soil sampling protocol: techniques and stra-
tegies." McLean, VA: ETHURA. Under Subcontract to Environmental Research
Center, University of Nevada, for U.S. Environmental Protection Agency.
Mills PA, Onley JH, Gaither RA. 1963. Rapid method for chlorinated pesti-
cide residues in nonfatty foods. J. Ass. Offic. Anal. Chem. 46(2):186-191.
Moein GJ. 1976. Study of the distribution and fate of polychlorinated bi-
phenyls and benzenes after spill of transformer fluid. U.S. Environmental
Protection Agency. Report EPA 904/9-76-014; NTIS PB288484.
Moser JH, Huibregtse KR. 1976. Handbook for sampling and sample preserva-
tion of water and wastewater. U.S. Environmental Protection Agency. EPA-600/
4-76-049; NTIS PB-259 946.
63
-------
Needham LL, Burse VW, Price HA. 1981. Temperature-programmed gas chromato-
graphic determination of polychlorinated and polybromianted biphenyls in serum.
J. Assoc. Off. Anal. Chem. 64(5):1131-1137.
Plumb RH Jr. 1981. Procedures for handling and chemical analysis of sedi-
ment and water samples. U.S. Environmental Protection Agency/Corps of Engi-
neers Technical Committee on Criteria for Dredged and Fill Material. Contract
No. EPA-4805572010; NTIS No. AD A103788.
Rodriguez CF, McMahon WA, Thoams RE. 1980. Method development for determina-
tion of polychlorinated hydrocarbons in municipal sludge. U.S. Environmental
Protection Agency. EPA-600/2-80-029; NTIS No. PB 82-234 071.
Sherma J. 1981. Manual of analytical quality control for pesticides and re-
lated compounds in human and environmental samples. U.S. Environmental Pro-
tection Agency. EPA-600/2-81-059; NTIS PB81-222721.
Tanabe H. 1976. PCB microanalysis. In: PCB Poisoning and Pollution.
Higuchi K. ed. Tokyo: Kodansha, Ltd; New York: Academic Press.
United Kingdom Department of the Environment. 1979. Organochlorine insecti-
cides and polychlorinated biphenyls in waters 1978; tentative method. Methods
for the examination of waters and associated materials. Organochlorine Insectic.
Polychlorinated Biphenyls Waters.
USEPA. 1976. Sample collection manual, ecological monitoring branch. Office
of Pesticides Programs, Benefits and Field Studies Division. Washington, DC.
USEPA. 1978 (September). Methods for benzidine, chlorinated organic compounds,
pentachlorophenol and pesticides in water and wastewater. Interim report.
Cincinnati, OH: Environmental Monitoring and Support Laboratory.
USEAP. 1979a (December 3). Organochlorine pesticides and PCBs—Method 608.
(44 FR 69501-69509).
USEPA. 1979b (December 3). Base/neutrals, acids, and pesticides—Method 625.
(44 FR 69540-69552).
USEPA. 1979c. Methods 330.4 (titrimetric, DPD-FAS) and 330.5 (spectrophoto-
metry, DPD) for chlorine, total residual. Methods for Chemical Analysis of
Water and Wastes, U.S. Environmental Protection Agency, Environmental Monitor-
ing and Support Laboratory, Cincinnati, OH, EPA 600-4/79-020.
USEPA. 1979d (September). Analytical protocol for screening publicly owned
treatment works (P0TW) sludges for organic priority pollutants. Cincinnati,
OH: Environmental Montoring and Support Laboratory.
USEPA. 1981 (February). The anaysis of polychlorinated biphenyls in trans-
former fluid and waste oils. Cincinnati, OH: Office of Research and Devel-
opment, Environmental Monitoring and Support Laboratory.
64
-------
USEPA. 1982a (July). Test methods for evaluating solid waste. Washington, DC:
Office of Solid Waste and Emergency Response.
USEPA. 1982b (November 4). Analysis of pesticides, phthalates, and polychlor-
inated biphenyls in soils and bottom sediments. Alexandria, VA: HWI Sample
Management Office. (Unpublished method.)
USEPA. 1982c. 40 CFR 761, Polychlorinated biphenyls (PCBs); manufacturing,
processing, distribution in commerce, and use prohibitions; use in closed and
controlled waste manufacturing processes. 47 FR 46980-46996.
USEPA. 1983 (July 1). 40 CFR Part 761. Polychlorinated biphenyls (PCBs)
manufacturing, processing, distribution in commerce, and use prohibitions.
Washington, DC: U.S. Government Printing Office.
The Utility Solid Waste Activities Group. 1984. Proposed spill cleanup policy
and supporting studies. Submitted to EPA.
Watts RR. 1980. Analysis of pesticide residues in human and environmental
samples, a compilation of methods selected for use in pesticide monitoring
programs. Research Triangle Park, NC: U.S. Environmental Protection Agency.
EPA-600/8-80-038.
65
------- |