United States
Environmental Protection
Agency
Environmental Monitoring
Systems Laboratory
Las Vegas NV 89193-3478
Research and Development
EPA/600/S4-87/022 Jan. 1988
v>EPA Project Summary
Single-Laboratory Evaluation of
Method 8080 — Organochlorine
Pesticides and PCBs
Viorica Lopez-Avila, Sarah Schoen, June Milanes, and Werner F. Beckert
Method 8080 was developed for the
determination of certain organochlorine
pesticides (OCPs) and polychlorinated
biphenyls (PCBs) in liquids and solids.
Liquid samples are extracted according
to Method 3510 (separatory funnel) or
Method 3520 (continuous liquid-liquid
extractor) and solid samples according
to Method 3540 (Soxhlet extraction) or
Method 3550 (sonication). The extracts
are concentrated, fractionated on Florisil
and the fractions analyzed by gas
chromatography on packed columns.
EPA Method 8080, as published in
the Second Edition of "Test Methods
for Evaluating Solid Waste," Office of
Solid Waste Manual SW-846, has been
evaluated in a single-laboratory study.
The Florisil cleanup procedure recom-
mended in Method 8O8O does not
separate the OCPs from the PCBs.
Consequently, the gas chromatographic
analysis of the OCPs on the packed
columns specified in the method may
result in false identifications or in no
identifications at all when PCBs are
present. Toxaphene and chlordane pose
special problems because of their
multi-peak responses. Silica gel was
therefore substituted for Florisil, and
capillary columns for the packed
columns. Furthermore, a sulfur cleanup
procedure was incorporated in the
method.
The Method 8080 protocol was re-
vised accordingly and was evaluated
with extracts of environmental samples
spiked with the substances of interest
at known concentrations. The precision
and accuracy results indicate that the
revised Method 8080 can be reliably
applied to the determination of OCPs
and PCBs in liquid and solid matrices.
The method detection limits for liquid
matrices range from 0.02 to 0.09 M9/L
for the OCPs and from 0.5 to 0.9 ng/L
for PCBs. The method detection limits
for solid matrices range from 1 to 6
Mg/Kg for the OCPs and from 60 to 70
jug/Kg for PCBs.
The revised protocol is included in
this report as an appendix. Also included
as an appendix is an extensive literature
review covering analytical methods for
the determination of OCPs and PCBs in
water, soil, sediment and sludge
samples.
This Project Summary was developed
by EPA's Environmental Monitoring
Systems Laboratory, Las Vegas, NV, to
announce key findings of the research
project that Is fully documented In a
separate report of the same title (sea
Project Report ordering Information at
back).
Introduction
The determination of organochlorine
pesticides (OCPs) and polychlorinated
biphenyls (PCBs) in environmental sam-
ples by gas chromatography (GC) with
electron capture detection and by mass
spectrometry has been recommended.
Electron capture detection is preferred
over mass spectrometry because the
former is two to three orders of magnitude
more sensitive than the latter. Since PCBs
are extracted along with the OCPs and
since they interfere with the determina-
tion of the OCPs whenever electron
capture detectors are used, their presence
in the extracts together with the OCPs
-------
needs to be minimized. Therefore, several
cleanup techniques based on Florisil,
alumina, and silica gel chromatography
have been developed.
EPA Method 8080, as published in the
document "Test Methods for Evaluating
Solid Waste," Office of Solid Waste
Manual SW-846 (1), provides sample
extract cleanup and GC conditions for the
determination of the OCRs and the PCBs
listed in Table 1 in a variety of environ-
mental matrices including ground water,
liquids, and solids. Following solvent
extraction of liquid samples in a separatory
funnel (Method 3510) or in a continuous
liquid-liquid extractor (Method 3520) and
of solid samples in a Soxhlet extractor
(Method 3540) or with a sonicator
(Method 3550), the extracts are cleaned
up by Florisil chromatography. Elution of
the compounds from the Florisil column
is performed with 6, 15, and 50 percent
ethyl ether in hexane. All compounds
listed in Table 1 except six OCRs elute in
Fraction I (6 percent ether in hexane). Of
those six pesticides, four (dieldrin,
endosulfan I, endrin, and endrin aldehyde)
elute in Fraction II (15 percent ether in
hexane), and two (endosulfan II and
endosulfan sulfate) elute in Fraction III
(50 percent ether in hexane). Endrin
aldehyde was also reported in Fraction III.
There is no mention in Method 8080 of
possible overlapping of compounds be-
tween fractions and of the reproducibility
of the elution pattern.
Acurex, under contract to the EMSL-
LV, conducted an evaluation and improve-
ment study of Method 8080. In the first
phase of this study. Method 8080, as
written, was evaluated to: (a) determine
the recoveries of the OCRs and PCBs
listed in Table 1 in the absence of matrix
interferences, (b) determine the extent of
overlapping of compounds between frac-
tions, and (c) determine the efficiency of
the Florisil cleanup scheme with real
samples. Also, the GC determination of
the OCPs and PCBs using packed and
capillary columns was evaluated, and a
literature review of the analytical meth-
odologies for the determination of the
compounds listed in Method 8080 was
conducted. Because of the complex nature
of the PCB formulations, only Aroclor
1016 and Aroclor 1260 were used
throughout this r.tudy.
In the second phase of this study, the
focus was on developing a fractionation
procedure to separate the PCBs from the
bulk of the OCPs, and, at the same time,
to remove interfering compounds coex-
tracted with the OCPs and PCBs. Several
extract cleanup procedures based on
silica, alumina, silica gel/Celite, and
Florisil/charcoal chromatography were
Table 1. Compounds Listed in EPA Method 8080
Parameter8
Storet No.
CAS No.
alpha-BHC
beta-BHC
gamma-BHC (Lindane)
delta-BHC
Heptachlor
Aldrin
Heptachlor epoxide
gamma-Chlordane
Endosulfan 1
4,4' -DDE
Dieldrin
Endrin
Endosulfan II
4.4' -ODD
Endrin aldehyde
Endosulfan sulfate
4.4' -DDT
4,4 '-Methoxychlor
Toxaphene
Aroclor 1016
Aroclor 1221
Aroclor 1232
Aroclor 1242
Aroclor 1248
Aroclor 1254
Aroclor 1260
39337
39338
39340
34259
39410
39330
39420
39350
34361
39320
39380
39390
34356
39310
34366
34351
39300
NA
394OO
34671
39488
39492
39496
39500
39504
39508
319-84-6
319-85-7
58-89-9
319-86-8
76-44-8
309-00-2
1024-57-3
57-74-9
959-98-8
72-55-9
60-57-1
72-20-8
33212-65-9
72-54-8
7421-93-4
1031-O7-8
50-29-3
72-43-5
8001-35-2
12674-11-2
1 104-28-2
11141-16-5
53469-21-9
12672-29-6
11097-69-1
1 1096-82-5
NA — Storet number not available.
8 Kepone is included in Method 8080 in the second edition of SW-846 but is not included in
Method 8080 in the third edition.
investigated. Furthermore, a capillary GC
method was developed, and a method for
sulfur removal was tested and incor-
porated. Upon completion of this phase, a
revised protocol was prepared.
The analytical scheme given in the
revised Method 8080 protocol employs
silica gel fractionation (silica gel de-
activated with 3.3 percent water). Three
fractions are collected: Fraction I eluted
with 80 ml hexane, Fraction II eluted
with 50 mL hexane, and Fraction III eluted
with 15 ml methylene chloride. The
determination of the OCPs and PCBs
which are recovered in these three frac-
tions is performed by GC on fused silica
capillary columns and with electron
capture detection.
The revised protocol was evaluated in
Phase III with extracts of environmental
samples spiked with the substances of
interest at known concentrations. The
evaluation studies were conducted at
three concentrations, each in triplicate.
The precision and accuracy results in-
dicated that the revised Method 8080
could be reliably applied to the determi-
nation of the OCPs and PCBs in liquid
and solid matrices. A method detection
limit determination was performed for
both the aqueous and the solid matrices.
Experimental
Materials and Reagents
The materials and reagents were those
specified in Method 8080, as applicable.
All solvents and reagents used were
pesticide grade or analytical grade. The
two capillary GC columns used in the
second and third phase of this study
were a 30 m x 0.25 mm ID DB-5 fused-
silica capillary column (J & W Scientific
Inc., Folsom, California) with a 0.25-/*m
film thickness, and a 30 m x 0.25 mm ID
SPB-608 fused-silica capillary column
(Supelco Inc., Bellefonte, Pennsylvania)
with a 0.25-jum film thickness. The DB-5
column was held for 2 min at 100°C,
heated at 15°C/min from 100°C to
160°C, and finally heated at 5°C/min
from 160°C to 270°C. The SPB-608
column was held for 2 min at 160°C,
heated at 5°C/min from 160°C to 290°C
and held 1 min at 290°C. The gas
chromatograph was equipped with a
constant current pulsed frequency elec-
tron capture detector and a data system.
A Varian 8000 autosampler was used;
the injection volume was 2 juL.
Samples and Sample Extract
Preparation
The aqueous samples used in this study
-------
included distilled water, an aqueous
waste (obtained from a pesticide waste
storage facility) with a high concentration
of organics (Liquid Waste 1), and an
aqueous waste (obtained from a pesticide
manufacturing plant) high in total dis-
solved solids (50 g/L) and various organic
solvents (>9.5 percent). The solid matrices
used were NBS SRM-1645 (River Sedi-
ment), a sandy loam soil from Soils Inc.,
Puyallup, Washington, with a total or-
ganics content of approximately 1300
mg/Kg, and a waste consisting of acti-
vated charcoal mixed with a polymeric
material. Except for the evaluation of
Method 3520 and 3540, the aqueous
samples were extracted in a separatory
funnel with methylene chloride, and the
soil and sediment samples were extracted
with hexane/acetone by sonication. The
extract solvents were exchanged for
hexane, and the extract concentrates
were spiked with the OCRs and RGBs.
Sample Extract Cleanup
The Florisil cleanup was performed as
specified in Method 8080. The silica gel
cleanup procedure was performed ac-
cording to Biddleman et al. (2), with slight
modifications. Cleanup on alumina was
similar to the procedure specified in the
EPA Superfund Contract Laboratory Pro-
gram (3), the Florisil/charcoal cleanup
procedure was performed as described
by Berg et al. (4), and the silica! gel/Celite
cleanup according to the procedure of
Armour and Burke (5). The sulfur removal
procedure evaluated and adopted was
that described by Jensen et al. (6) with
tetrabutylammonium sulf ite as the active
reagent.
Results and Discussion
Sample Extraction
A brief evaluation of the four extraction
procedures recommended in Method
8080 showed, with the samples tested,
the following results:
• The efficiencies of Methods 3510
(separatory funnel) and 3520 (con-
tinuous liquid/liquid extraction) for
the extraction of aqueous samples
were approximately equal. Method
3510 was then used for the extrac-
tion of all aqueous samples.
• Method 3550 (sonication) for solids
showed a better precision than
Method 3540 (Soxhlet extraction);
the accuracies were similar for both
methods. Method 3550 was than
used for the extraction of all solid
samples.
Florisil Fractlonatlon
Florisil fractionation was performed as
recommended in Method 8080. Separate
experiments were run in duplicate for
PCBs, toxaphene, technical chlordane,
OCP group A (gamma-BHC, heptachlor,
aldrin, heptachlor epoxide, endosulfan I,
dieldrin, endosulfan II, 4,4'-DDT, and
endrin aldehyde), and OCP group B
(alpha-BHC, beta-BHC, delta-BHC, 4-4'-
DDE, endrin, 4,4'-DDD, endosulfan sul-
fate, and 4,4'-methoxychlor). The experi-
mental results are presented in Table 2.
The overall recoveries are quantitative,
and the agreement between the duplicate
experiments is in most cases excellent.
However, discrepancies have been found
between our data and the recovery data
listed in Method 8080. But regardless of
the reproducibility of the fractionation, it
is apparent that the Florisil fractionation
method is not suitable for samples that
contain both OCPs and PCBs. PCBs appear
in the same fraction as the bulk of the
OCPs, but these two types of compounds
need to be separated from each other to a
larger extent to avoid cross-interference.
To exemplify this on a real sample, we
extracted a liquid waste and spiked the
extract with known amounts of OCPs,
Aroclor 1016 and Aroclor 1260. Because
of the complexity of this sample matrix,
we chose to analyze the fractions on a
DB-5 capillary column; however, even
then we were not able to find the spiking
compounds. In view of these results, we
eliminated the Florisil fractionation
scheme from further evaluation.
S///ca/ Gel Fractionation
The silica gel fractionations were per-
formed in triplicate at two concentration
levels. Technical chlordane and toxaphene
fractionations were performed separately.
The distribution and percent recoveries
of the OCPs, Aroclor 1016, Aroclor 1260,
technical chlordane, and toxaphene are
presented in Table 3. The distribution
patterns of the OCPs and PCBs in the
three silica gel fractions were quite
reproducible. Compounds found to elute
in Fraction I (80 mL hexane) include:
heptachlor, aldrin, 4,4'-DDE, chlordane
(partially), and the PCBs. Almost all the
other OCPs elute in Fraction III. Total
recoveries were greater than 70 percent,
except for technical chlordane at con-
centration 1, with most values ranging
from 80 to 110 percent.
The evaluation of the silica gel frac-
tionation scheme with real sample ex-
tracts (e.g., liquid waste, NBS SRM-1645,
sandy loam, etc.) gave comparable reults.
Only a few compounds (4,4'-DDD, alpha-
BHC, gamma-chlordane) were split be-
tween fractions when the real samples
were fractionated.
These results show that Fraction II and
III may be combined before concentration
and analysis when the sample matrices
are relatively simple. However, when
samples with complex matrices have to
Table 2. Results of the Florisil Fractionation Study
Recovery (%f
alpha-BHC
beta-BHC
gamma-BHC
delta-BHC
Heptachlor
Aldrin
Heptachlor epoxide
Endosulfan 1
4.4'-DDT
Dieldrin
Endrin
Endosulfan II
4,4' -ODD
Endrin aldehyde
Endosulfan sulf ate
4,4' -DDE
4,4 '-Methoxychlor
Aroclor 1016
Aroclor 1260
Technical chlordane
Toxaphene
Spike
level
tM)
0.5
0.5
0.5
0.5
0.5
0.5
0.5
0.5
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
5.0
10.0
10.0
5.0
10.0
Fraction 1
94;
86;
76;
70;
75;
78;
79;
59;
95;
20;
37;
93;
97;
87;
86;
91;
93;
105;
92
83
77
60
78
80
80
64
95
24
43
90
93
84
76
78
94
108
Fraction II Fraction III
9.7 8.7
JO
8.5
31
5.3
6.0
15
29
8.7
71
64
60
11
80
23
9.6
29
4.6
10
9.6
; 0
8.1 1.5;<1.0
27
5.8
6.6 2.8; 3.8
15
30
6.7
70
51
79 30 ; 15
10
90 11 ; 4.4
23 57 ; 51
8.6
26
4.4
9.0
7.6
Total
1O4;
96;
86;
101;
80;
87;
94;
88;
104;
91;
101;
90;
104;
91;
80;
107;
116;
91;
101;
103;
105;
101
83
85
87
84
90
95
94
102
94
94
94
100
94
74
102
110
80
87
102
108
" The number of determinations is two.
-------
Table 3. Distribution and Percent Recoveries of Organochlorine Pesticides and PCBs in Silica Gel Column Fractions
a,b.c,d.e
Fraction I
Fraction II
Fraction III
Total recovery
Compound
Cone. 1
Cone. 2
Cone. 1
Cone. 2
Cone. 1
Cone. 2
Cone. 1
Cone. 2
alpha-BHC
beta-BHC
gamma-BHC
delta-BHC
Heptachlor
Aldrin
Heptachlor expoxide
Endosulfan 1
4,4' -DDE
Dieldrin
Endrin
Endosulfan II
4.4' -ODD
Endrin aldehyde
Endosulfan sulfate
4.4'-DDT
4,4 '-Methoxychlor
Aroclor 1016
Aroclor 1260
Technical chlordane
Toxaphene
109(4.1) 118(8.7)
97 (5.6> 104 (1.6)
86 (5.4) 94 (2.8)
86 (4.0) 87 (6. 1)
91 (4.1) 95 (5.0)
14 (5.5) 22 (5.3)
86 (13.4)
19
15
(6.8)
(2.4)
73 (9.1)
39 (3.6)
17 (1.4)
82 (1-7)
107 (2.1)
91 (3.6)
92 (3.5)
95 (4.7)
95 (5.1)
96 (6.0)
85 (10.5)
97 (4.4)
102 (4.6 J
81 (1.9)
93 (4.9)
15 (18.7)
99 (9.9)
29
73
(5.0)
(9.4)
74 (8.0)
98 (12.5)
85 (10.7)
83 (10.6)
88 (10.2)
87 (10.2)
87 (10.6)
71 (12.3)
86 (10.4)
92 (10.2)
76 (9.5)
82 (9.2)
8.7 (15.0)
82 (10.7)
37
84
(5.1)
(10.7)
82
107
91
92
109
97
95
95
86
96
85
97
102
81
93
101
99
86
91
62
88
(1.7)
(2.1)
(3.6)
(3.5)
(4.1)
(5.6)
(4.7)
(5.1)
(5.4)
(6.0)
(10.5)
(4.4)
(4.6)
(1.9)
(4.9)
(5.3)
(9.9)
(4.0)
(4.1)
(3.3)
(12.0)
74 (8.0)
98 (12.5)
85 (10.7)
83 (10.6)
118 (8.7)
104 (1.6)
88 (10.2)
87 (10.2)
94 (2.8)
87 (10.6)
71 (12.3)
86 (10.4)
92 (10.2)
76 (9.5)
82 (9.2)
82 (23.7)
82 (10.7)
87 (6. 1)
95 (5.0)
98 (1.9)
101 (10.1)
* Eluant composition: Fraction I — 80 mL hexane; Fraction II — SO mL hexane; Fraction III — 75 mL methylene chloride.
* Concentration 1 is 0.5 ng per column for BHCs, heptachlor, aldrin, heptachlor epoxide, endosulfan I; 1.0 i*g per column for dieldrin. endosulfan li
4,4 '-DDT, endrin aldehyde, 4,4'-DDD, 4.4-DDE, endrin, and endosulfan sulfate; 5 tig per column for 4,4'-methoxychlor and technical chlordane; 11
ng per column for toxaphene, Aroclor 1016, and Aroclor 1260.
c For concentration 2 the amounts spiked are 10 times those of concentration 1.
d The values listed represent the average recoveries from three determinations; the numbers in parentheses are the standard deviations. Thi
recovery cut-off point is 5 percent.
e Data obtained with standards, as indicated in footnotes b and c, dissolved in 2 mL hexane.
be extracted, especially matrices contain-
ing organic solvents, more crossover
between fractions may occur. In such
cases it is more advantageous to analyze
the three fractions separately.
Fractlonatlon on Alumina,
Florisil/Charcoal, and Silica
Gel/Cellte
It was found in a series of experiments
that none of these procedures was
superior to the relatively simple silica gel
fractionation.
Sulfur Removal
Elemental sulfur, which may be present
in extracts from sediments and from some
industrial samples, gives GC peaks which
mask the region of aldrin, BHCs, hepta-
chlor and heptachlor epoxide when the
analysis is performed on the 1.5 percent
OV-17/1.95 percent OV-210 on Chromo-
sorb-WHP column or on the 30 m DB-5
fused-silica capillary column. The proce-
dure of Jensen et al. (6) was used on five
sample extracts fortified with the OCRs
and PCBs to determine if removal of
sulfur is affected by matrix interferences
and if the OCP and PCB recoveries are
acceptable (>80 percent) when this
method is used. In addition to the real
sample extracts, three pesticide standards
were reacted with the tetrabutylam-
monium sulfite reagent to determine
compound recovery in the absence of
matrix interferences. The results showed
that sulfur was removed quantitatively,
regardless of the matrix, and that the
recoveries were acceptable except for
the endrin aldehyde recovery which was
only about 10 percent. This procedure
was therefore incorporated into the re-
vised method protocol.
Capillary Gas Chromatography
The gas chromatographic retention
times of 18 OCPs on the DB-5 and the
SPD-608 fused-silica capillary columns
are presented in Table 4. Toxaphene is
not included because of its multipeak
response. Aroclor mixtures have been
analyzed individually on the DB-5 column;
the retention times of the individual
chlorinated biphenyls in these mixtures
are included in the full report.
Those OCPs that elute on the DB-5
column at the same retention times as
some of the components of the Aroclor
mixtures are identified in Table 4. Of the
six OCP peaks that overlap with PCB
component peaks, only heptachlor and
gamma-chlordane are of concern because
the other four compounds are separated
from the PCBs during the silica gel
chromatography step. The heptachlor
peak overlaps with a PCB peak eluting at
15.93 (present in six of the seven PCB
mixtures), and the gamma-chlordane peak
overlaps with a PCB peak eluting at 19.5
minutes (present in four of the seven
PCB mixtures).
Method Performance
Method performance, as used here
includes the method precision and ac
curacy and the method detection limit. Tc
determine method precision and accuracy
clean hexane and extracts of environ
mental samples (Liquid Waste 1, NB£
River Sediment SRM-1645, and sand^
loam soil) were spiked with the OCR:
listed in Table 1 (except toxaphene), with
Aroclor 1016 and with Aroclor 1260 a
three concentrations (0.25 ng//iL, O.E
ng/^L, and 2.5 ng/juL for the OCPs anc
2.5 ng/>L, 5.0 ng//uL and 25 ng/juL foi
PCBs) and were processed through the
method. The average recoveries of tripli
cate determinations (method accuracy
and the relative standard deviation;
(method precision) are presented in Table
5.
Of all recovery determinations in Table
-------
5, 77 percent fall within the range 85 to
132 percent. An additional 5.2 percent
are below 70 percent, and 18 percent are
between 71 and 84 percent. Because of
interferants, delta BHC, endosulfan II,
and 4,4'-DDD could not be determined in
the liquid waste when spiked at con-
centrations 1 and 2, endrin aldehyde
could not be determined in the liquid
waste at any of the three spike levels,
and 4,4'-methoxychlor could not be deter-
mined in the NBS SRM-1645 when spiked
at concentration 1.
Of all relative standard deviations in
Table 5, 58 percent fall below 10 percent,
36 percent are between 11 percent and
Table 4. Summary of Retention Times for the Organochlorine Pesticides'
Retention time (min)
No.
Compound Name
DB-5b
SPB-608*
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
alpha-BHC
beta-BHC
gamma-BHC
delta-BHC
Heptachlord
Aldrin
Heptachlor epoxide
gamma-Chlordaned
Endosulfan f1
4.4'-DDE
Dieldrind
Endrin
Endosulfan Ir
4.4' -ODD
Endrin aldehyded
Endosulfan sulfate
4.4' -DDT
4.4 '-Methoxychlor
12.29 ±0.010
13.1 3 ±0.009
13.37 ±0.011
14.1 4 ±0.011
15.91 ±0.008
17.1 6 ±0.009
18.60 ± 0.009
19.48 ±0.012
19.94 ±0.010
20.83 ± 0.008
20.91 ±0.008
21.71 ±0.007
22.05 ± 0.006
22.38 ± 0.008
22.75 ±0.007
23.64 ± 0.008
23.79 ±0.008
25.94 ± 0.007
9.46
11.33
10.97
12.73
12.46
13.76
15.98
16.70
17.40
18.36
18.60
19.96
20.69
20.53
21.90
22.54
21.72
24.90
Toxaphene and PCBs are not included because of their muttipeak response.
The values given for the DB-5 column are average retention time ± standard deviation of 10
replicate determinations.
Single determinations.
Coeluting with Aroclor components on the DB-5 column.
20 percent, and only 6 percent are above
21 percent.
There seem to be no patterns between
the recovery and the concentration of the
OCRs and PCBs or the matrix. When the
percent recovery of each compound was
plotted as a function of matrix for each of
the three concentrations, no trend could
be found.
The method detection limits (MDL) were
determined for both distilled water and
sandy loam soil from the standard devia-
tions (SD) of 7 replicate measurements
(MDL equals 3 times the SD); they repre-
sent the minimum concentrations that
can be measured and repotted with 99
percent confidence. The MDLs for water
samples ranged from 0.02 to 0.09 ng/L
for OCRs and from 0.5 to 0.9 /ug/L for
PCBs, and those for soil samples ranged
from 1 to 6 jug/Kg for OCPs and from 60
to 70 /ug/Kg for PCBs. It should be kept in
mind that these values are representative
of clean sample matrices. For complex
matrices, the MDLs may be higher.
Conclusions
A revision of Method 8080 for the
determination of the organochlorine
pesticides and PCBs was necessary be-
cause the Florisil procedure does not
allow separation of the organochlorine
pesticides from the multi-component PCB
mixture (except for endosulfans and
endrin aldehyde) and because the packed
Table 5. Recoveries fin Percent) of the Method 8080 Compounds from Spike Extracts of Environmental Samples.
Compound
alpha-BHC
beta-BHC
gamma-BHC
delta-BHC
Heptachlor
Aldrin
Heptachlor epoxide
gamma-Chlordane
Endosulfan 1
4.4' -DDE
Dieldnn
Endrin
Endosulfan II
4.4--DDO
Endrin aldehyde
Endosulfan sulfate
4.4' -DDT
4,4 '-Methoxychlor
Aroclor 1016
Aroclor 1260
Concentration (ng/^L extract)
Number of determinations
Cone. 1
83 ± 16 (19)
86 ± 15 (17)
85 ± 15 (18)
87 ± 12 (14)
79 ±11 (13)
88 ±11 (13)
94 ±67 (7. 1)
94 ± 13 (14)
89 ± 12 (14)
92 ± 13 (14)
89 ± 12 (14)
66 ±11 (17)
86 ± 7.9 (9.2)
89 ± 12 (14)
83 ± 8.3 (10)
91 ±52 (5.7)
74 ±19 (26)
98 ±2. 6 (2.7)
94 ± 14 (15)
92 ± 12 (13)
025
3
Hexanea
Cone. 2
106 ± 6.8 (6.4)
110 ±10 (9.5)
108 ±11 (10)
109 ± 12 (11)
94 ± 9.5 (10)
107 ± 9.5 (8.9)
109 ± 14 (13)
110 ±11 (10)
108 ± 13 (12)
107 ± 15 (14)
112 ±13 (12)
65 ± 10 (16)
111 ±14 (13)
110 ±9.8 (8.9)
102 ± 19 (19)
112 ±21 (19)
88 ± 13 (15)
104 ± 18 (17)
93 ± 6.5 (7.0)
87 ± 15 (17)
05
3
Liquid waste extract"
Cone. 3
91 ±4.6 (5.1)
98 ± 2.0 (2. 1)
99 ± 2.3 (2.3)
97 ±1.6 (1.6)
83 ± 6.6 (7.9)
89 ±4.2 (4.7)
100 ±23 (23)
91 ± 1.2 (1.3)
99 ± 2.3 (23)
89 ± 4.5 (5.0)
102 ± 1.5 (1.5)
64 ± 8.3 (13)
101 ±0.6 (0.6)
97 ±1.7 (1.8)
95 ± 3.0 (3.2)
104 ± 2.5 (2.4)
73 ± 5.8 (8.0)
104 ± 3.2 (3. 1)
93 ±2.0 (2.2)
78 ± 4.0 (5. 1)
2.5
3
Cone. 1
96 ± 7.0 (7.3)
92 ± 10 (1 1)
91 ± 10 (11)
b
89 ± 14 (15)
96 ± 8.9 (9.3)
111 ±17 (15)
100 ± 8.5 (8.5)
95 ± 6.5 (6.8)
119 ±11 (8.9)
88 ± 3.2 (3.6)
101 ±5.9 (5.8)
b
b
b
132 ±17 (13)
101 ±23 (23)
49 ±14 (29)
1 14 ± 6.0 (5.3)
99 ± 4.6 (4.6)
0.25
3
Cone. 2
97 ± 3.5 (3.6)
100 ±4.0 (4.0)
100 ±5.5 (5.5)
b
94 ±10 (11)
98 ±9. 4 (9.6)
109 ± 14 (13)
103 ± 2.5 (2.4)
100 ± 12 (12)
1 13 ± 2.5 (2.2)
86 ± 9.2 (1 1)
90 ± 10 (1 1)
b
b
b
127 ±22 (17)
83 ±11 (13)
58 ±9.3 (16)
122 ± 10 (8.3)
102 ±4.7 (4.6)
0.5
3
Cone. 3
79 ±10 (13)
90 ±3.1 (3.4)
90 ± 4.0 (4.4)
90 ±11 (8.8)
90 ±11 (12)
92 ± 9.2 (1O)
89 ±4.1 (4.6)
95 ± 8.0 (8.4)
88 ± 3.8 (4.3)
95 ±16 (17)
82 ± 4.3 (5.3)
65 ±3.1 (4.7)
79 ±7.1 (9.0)
76 ±16 (21)
b
83 ± 4.0 (4.8)
88 ±18 (21)
75 ±4.6 (6.1)
1 18 ± 9.8 (8.3)
100 ±18 (18)
2.5
3
" Percent recovery ± standard deviation (triplicate determinations).
* Unable to determine recovery because of interference.
-------
Tab/0 5. (Continued)
NBSSRM -1645 extract?
Compound
alpha-BHC
beta-BHC
gamma-BHC
delta-BHC
Heptachlor
Aldrin
Heptachlor epoxide
gamma-Chlordane
Endosulfan 1
4.4' -DDE
Dieldrin
Endrin
Endosulfan II
4.4'-DDD
Endrin aldehyde
Endosulfan sulfate
4,4'-DDT
4,4 '-Methoxychlor
Aroclor 1016
Aroclor 1260
Concentration (ng/nL extract)
Number of determinations
Cone. 1
73 ±2.1 (2.9)
88 ±4.1 (4.7)
83 ± 3.0 (3.6)
85 ± 4.6 (5.4)
53 ±10 (19)
69 ± 3.6 (5.2)
91 ± 4.9 (5.4)
77 ± 5.3 (6.9)
85 ± 5.5 (6.5)
75 ± 5.3 (7. 1)
92 ±8.6 (9.4)
100 ± 9.5 (9.5)
80 ± 7.4 (9.2)
W6 ± 6.4 (6.0)
70 ±5.7 (8.2)
75 ±8.7 (12)
54 ±13 (24)
b
1O4±9.0 (8.7)
92 ± 9.5 (10)
0.25
3
Cone. 2
75 ± 6.0 (8.0)
94 ± 3.0 (3.2)
89 ±4.1 (4.6)
92 ± 5.2 (5.6)
70 ±7.7 (11)
65 ±4.6 (7.1)
91 ±5.7 (6.3)
81 ±4.9 (6.1)
88 ±5.1 (5.8)
76 ±7.1 (9.3)
85 ± 9.4 (1 1)
87 ± 6.4 (7.3)
81 ± 4.5 (5.5)
85 ±3.1 (3.6)
71 ± 9.2 (13)
86 ± 5.0 (5.8)
61 ± 7.9 (13)
99 ±17 (17)
104 ± 2.5 (2.4)
95 ± 7.5 (7.9)
0.5
3
Cone. 3
76 ± 5.6 (7.3)
92 ±7.1 (7.7)
93 ±8.1 (8.7)
94 ±8.7 (9.3)
88 ±4.1 (4.7)
72 ± 1.0 (1.4)
93 ± 8.6 (9.2)
85 ±1.0 (1.2)
91 ±9.1 (10)
84 ±1.0 (1.2)
94 ± 10 (1 1)
76 ± 9.9 (13)
91 ±12 (13)
90 ± 7.2 (8.0)
88 ±12 (14)
72 ±11 (IS)
76 ± 2.5 (3.3)
92 ±17 (19)
102 ± 4.6 (4.5)
91 ±4.O (4.4)
2.5
3
Sandy loam soil extract"
Cone. 1
86 ±9.5 (11)
94 ± 8.4 (8.9)
92 ±11 (12)
94 ± 12 (13)
89 ± 9.6 (1 1)
99 ± 4.4 (4.4)
96 ±11 (11)
100 ± 8.3 (8.3)
95 ± 10 (1 1)
105 ± 14 (13)
113 ±12 (11)
74 ± 5.6 (7.5)
97 ± 14 (14)
103 ± 9.6 (9.3)
86 ±11 (13)
112 ±19 (17)
107 ±25 (23)
91 ± 14 (15)
90 ± 15 (17)
99 ± 6.8 (6.9)
0.25
3
Cone. 2
87 ±4.9 (5.7)
90 ±4.1 (4.5)
91 ±3.0 (3.3)
89 ± 3.6 (4. 1)
83 ±8.7 (11)
88 ± 2.0 (2.3)
90 ±4.6 (5.1)
93 ± 3.0 (3.2)
89 ± 5.3 (5.9)
93 ±5.6 (6.0)
99 ±4.6 (4.6)
60 ± 7.6 (13)
86 ±6.7 (7.8)
88 ± 7.8 (8.9)
82 ± 7.6 (9.3)
91 ± 13 (14)
83 ± 5.3 (6.5)
89 ± 9.6 (1 1)
92 ± 4.5 (4.9)
89 ± 8.9 (10)
0.5
3
Cone. 3
89 ± 2.5 (2.8)
93 ± 5.0 (5.4)
95 ±4. 6 (4.8)
93 ±6.1 (6.5)
79 ±17 (21)
82 ± 12 (15)
94 ± 6.6 (7.0)
87 ±11 (13)
93 ± 7.5 (8 1)
82 ±11 (13)
93 ± 8.2 (8.8)
47 ±11 (24)
89 ±11 (12)
90 ±7.8 (8.6)
73 ±20 (28)
88 ±11 (13)
64 ± 14 (21)
86 ± 14 (16)
85 ± 10 (12)
85 ± 13 (15)
2.5
3
' Percent recovery ± standard deviation (triplicate determinations).
* Unable to determine recovery because of interference.
columns do not have enough resolving
power to handle complex environmental
samples.
The revised Method 8080 presented in
Appendix B has been evaluated in a
single laboratory with some relevant liquid
and solid wastes. The utilization of silica
gel fractionation and capillary column
analysis was found to be appropriate.
When silica gel fractionation was used,
three fractions were collected. The silica
gel procedure is tedious and does account
for a major part of the analysis time.
However, we have demonstrated that the
method precision is better than ±20
percent for all compounds, and the ac-
curacy is greater than 60 percent when
standards are processed through the silica
gel procedure. Fraction III may be com-
bined with Fraction II when the matrix is
not very complex, and thus the number of
analyses per sample may be reduced.
The use of a second capillary column as a
confirmatory column is recommended.
Toxaphene, if present in the sample at
concentrations 10 times as high as the
OCPs, is likely to cause problems in the
determination of the OCPs and PCBs
since it does not elute from the silica gel
column in a narrow band. Other analytical
techniques (e.g., chemical ionization mass
spectrometry) should be considered for
the determination of toxaphene.
References
1. Test Methods for Evaluating Solid
Waste. Second Edition (July 1982),
SW-846, U.S. Environmental Protec-
tion Agency, Washington, D.C.
2. Biddleman, T. F., J. R. Matthews, C. E.
Olney, and C. R. Rice. J. Assoc. Off.
Anal. Chem. 61, 820-828 (1978).
3. Contract Laboratory Program Protocol
for the Analysis of Hazardous Sub-
stances List (HSL) Compounds. Revised
July 1985, U.S. Environmental Protec-
tion Agency, Washington, D.C.
4. Berg, O. W., P. L. Diosady, and G.A.V.
Rees. Bull. Environ. Contamin. Toxicol.
7,338-345(1972).
5. Armour, J. A., and J. A. Burke. J.
Assoc. Off. Anal. Chem. 53, 761-768
(1970).
6. Jensen, S., L. Renberg, and L. Reuter-
gardh. Anal. Chem. 49, 316-318
(1977).
-------
Viorica Lopez-Avila, Sarah Schoen, and June Milanes are with Acurex
Corporation. Mountain View, CA 94039; the EPA author Werner F. Beckett
(also the EPA Project Officer, see below) is with the Environmental Monitoring
Systems Laboratory Las Vegas, NV 89193-3478.
The complete report, entitled "Single-Laboratory Evaluation of Method 8080—
Organochlorine Pesticides and PCBs," (Order No. PB 87-232 591 /AS; Cost:
$24.95, subject to change) will be available only from:
National Technical Information Service
5285 Port Royal Road
Springfield. VA 22161
TeJephone: 703-487-4650
The EPA Project Officer can be contacted at'
Environmental Monitoring Systems Laboratory
U.S. Environmental Protection Agency
PO Box 93478
Las Vegas, NV 89193-3478
United States
Environmental Protection
Agency
Center for Environmental Research
Information
Cincinnati OH 45268
••,V'-«V,X :. -•-.••»-, -:• - t
' '-
Official Business
Penalty for Private Use S300
EPA/600/S4-87/022
------- |