United States
Environmental Protection
Agency
Environmental Monitoring
Systems Laboratory
Cincinnati OH 45268
Research and Development
EPA/600/SR-93/032 April 1993
&EPA Project Summary
Determination of Volatile Organic
Compounds in Soils Using
Equilibrium Headspace
Analysis and Capillary Column
Gas Chromatography/Mass
Spectrometry—Evaluation of the
Tekmar 7000 HA Analyzer
Pedro Flores and Thomas A. Bellar
Existing methods for determination
of volatile organic compounds (VOCs)
in soil matrices using the purge and
trap technique with gas chromatogra-
phy/mass spectrometry (GC/MS) have
several problems, which include pre-
serving sample integrity from collec-
tion to analysis and efficiently
extracting a broad spectrum of VOCs
from the soil matrix. This investigation
was undertaken using the Tekmar 7000
headspace autosampler to evaluate its
ability to resolve these problems. The
objective of this study was to optimize
analytical conditions and then to study
the efficiency of the headspace tech-
nique to extract VOCs from soils. Varia-
tions of sample preparation procedures
were studied, and method analytes were
identified and measured using internal
standard calibration GC/MS. Using
these data, relative standard deviations
and percent recoveries are reported for
59 analytes in four different types of
soil matrices: sand, clay, garden soil,
and hazardous waste landfill soil. The
most accurate and precise results are
obtained with sand. Method detection
limits (MDLs), ranging from 0.2 to 7.9
ng/kg, were calculated for all analytes,
using results of replicate analyses of
sand, the matrix that had the least ma-
trix effect. It is concluded that the 7000-
HA headspace analyzer can be used to
determine VOCs in soils.
This Project Summary was devel-
oped by EPA 's Environmental Monitor-
ing Systems Laboratory, Cincinnati, OH,
to announce key findings of the research
project that is fully documented in a
separate report of the same title (see
Project report ordering information at
back).
Introduction
An accurate and precise procedure is
needed to effectively remove volatile or-
ganic compounds (VOCs) from soils for
identification and measurement using gas
chromatography/mass spectrometry (GC/
MS). Ten years ago, the number of VOCs
that could be determined by GC/MS was
limited by packed column GC technology.
The least volatile compounds included tolu-
ene and ethyl benzene. Currently, with
capillary column capabilities, the scope of
VOCs in aqueous samples has been ex-
panded for a single column analysis to
include non-polar compounds with boiling
points ranging from -30°C to >220°C.
Heated purge and trap methodology has
also been applied to soil samples using
capillary column technology.1 The results
illustrated that many compounds curretly
determined in water matrices can be in-
cluded in the list of compounds deter-
mined in soil matrices. However, the
method was subjected to matrix effects,
particularly for those compounds with high
boiling points.
Printed on Recycled Paper
-------
In this work, we evaluated the capabil-
ity of the Tekmar 7000 Headspace
Autosampler (7000-HA) to effectively in-
troduce VOCs partitioned from soil ma-
trices into a fused silica capillary column
using the static headspace technique.
The integral features of the procedure
evaluated include sample fortification, dif-
ferent 7000-HA extraction parameters for
a wide variety of volatile compounds, and
the quantitative capabilities of the method
using different soil matrices and different
internal standards.
The 7000-HA was chosen for evalua-
tion because of its potential to include the
VOCs contained in the headspace of the
sample collection vials and extract VOCs
from a soil into the gas volume above the
sample, as well as its ability to permit
analysis with a minimal amount of sample
preparation. The analytes used for the
evaluation and their characteristic ions are
listed in Table 1. The results of this evalu-
ation are discussed in the following sec-
tions of this report.
Experimental Approach
Each sample was prepared by adding
5.0 g of a soil matrix to a 20-mL 7000-HA
crimp-seal glass headspace vial. In rapid
succession, each soil sample was fortified
with the target analytes in methanol, the
matrix modifier solution (MMS) was added,
and the vial was sealed. The purpose of
the matrix modifier solution was to in-
crease the efficiency of the headspace
analysis by providing a salting-out effect
and to minimize dehydrohalogenation re-
actions through pH adjustment.2 The vials
were placed in the autosampler carousel
and maintained at room temperature. Ap-
proximately 1 h prior to analysis, the indi-
vidual vials were moved to a heating zone
and allowed to equilibrate for 50 min at
85°C. The sample was then mixed by
mechanical vibration for 8 min while the
temperature was maintained at 85°C. The
autosampler then raised the vial causing
a stationary needle to puncture the sep-
tum and pressurize the vial with helium at
7.5 psi. The vial was allowed to pressure
equilibrate for a 0.10 min to ensure com-
plete mixing of the pressurization gas with
the vial headspace. The pressurized
headspace was then vented through a 2-
ml_ sample loop to the atmosphere for 15
sec. In this manner, a representative vol-
ume of headspace was isolated within the
loop. Finally the carrier gas, at a flow rate
of 9.5 mL/min, backflushed the sample
loop, sweeping the sample through the
heated transfer line into the GC/MS sys-
tem for separation, identification, and mea-
surement of the method analytes.
Parameters studied were
1. Precision of the Tekmar 7000
Headspace Autosampler
2. Selection of the Matrix-Modified So-
lution
3. Verification of the Fortifying Proce-
dure
4. Assessment of Analyte Recoveries
Using the Matrix Modifying Solution
5. Analyte Recoveries From Various
Soil Matrices
6. Analyte Recoveries From Various
Soil Matrices Using Internal Stan-
dard Calibration
7. Different Headspace Volume Re-
covery
Results and Discussion
Replicate analyses of fortified samples
showed the headspace analyzer to be re-
producible for all 59 analytes tested in
aqueous matrices. Only one relative stan-
dard deviation (RSD) was in excess of 13
percent. The RSDs were comparable to
those obtained using standard purge and
trap technology. Results from experiments
to evaluate the use of a matrix-modified
solution to increase the recovery of the
analytes from a solid matrix showed that a
saturated solution of sodium sulfate was
the most suitable. Replicate analytes of
fortified soils showed this solution pro-
duced the highest recoveries of most com-
pounds and the lowest relative standard
deviations.
Analyte recoveries from various types
of soils were studied using both the exter-
nal standard and the internal standard cali-
bration approaches. Four soils were
selected: sand, clay, garden soil, and a
subsurface soil sample collected near a
hazardous waste landfill. Table 1 summa-
ries the results showing the relative per-
cent recovery obtained from each matrix.
These relative percent recoveries were
obtained by dividing the peak areas from
each analyte by the respective peak ar-
eas in a control sample, and multiplying
by 100. High ratios are indicative of high
recoveries. These data indicate that ma-
trix effects are evident when analyzing
certain types of soils.
Figure 1 illustrates recoveries obtained
for representative compounds from the
analyte list when the headspace volume
in the sample vials was varied. Errors are
introduced into the analytical results if the
headspace volume is not held constant in
every vial. This is especially true for the
less volatile compounds.
Conclusions and
Recommendations
The following conclusions were made
from evaluation of study results.
1. The accuracy and precision of the
7000-HA were acceptable when
used to determine VOCs in water,
the matrix-modifying solution
(MMS), and sand. The 7000-HA
produced somewhat lower recover-
ies from other tested soil matrices.
However, these lower recoveries
were not due to inefficient
headspace analysis, but to stron-
ger adsorption capacity of soil. This
is the matrix effect. The results ob-
tained with the 7000-HA are equiva-
lent or better than current
methodology for volatiles in soil.
2. The matrix fortifying procedure was
found to be reproducible for all the
compounds evaluated.
3. Comparing recoveries obtained in
the different experiments for differ-
ent matrices, indicated a definite
matrix effect.
4. In an attempt to correct for the ma-
trix effect, seven internal standards
were evaluated. Results suggest
that the use of one internal stan-
dard improved data quality but did
not completely overcome the ma-
trix effect problem. Adding additional
internal standards with chemical and
physical properties similar to those
of the problem compounds helped
resolve this problem.
5. The less volatile compounds, such
as trichlorobenzenes, did not ap-
pear to be good candidates for ac-
curate measurement using the
headspace technique with a single
internal standard.
6. Headspace volume had a definite
effect on the sensitivity of the
method. When headspace volume
is decreased, sensitivity increases.
This effect is greater as the volatil-
ity of the compound decreases.
7. The amount of the matrix-modify-
ing solution added to the matrix
had little effect on analyte recov-
ery. The percent difference between
experiments was within the experi-
mental error.
8. This work pointed out the definite
need to develop a mechanism to
collect an exact predetermined
sample size and establish the her-
metic seal in the field. Until this is
done, this method cannot be used
to its fullest potential.
-------
Table 1. Relative Analyte Recoveries for Four Matrices
Analyte
MMS/Sand
MMS/Garden
MMS/Horizon-C
MMS/Clay
Dichlorodifluoromethane
Chloromethane
Vinyl chloride
Bromomethane
Chloroethane
Trichlorofluoromethane
1, 1-Dichloroethene
Methylene chloride
trans- 1 ,2-Dichloroethene
1, 1-Dichloroethane
2,2-Dichloropropane
cis- 1 ,2-Dichloroethene
Bromochloromethane
Chloroform
1, 1, 1-Trichloroethane
1, 1-Dichloropropene
Carbon tetrachloride
Benzene
1 ,2-Dichloroethane
Trichloroethylene
1 ,2-Dichloropropane
Dibromomethane
Bromodichloromethane
Toluene
1, 1 ,2-Trichloroethane
Tetrachloroethylene
1 , 3-Dichloropropane
Dibromochloromethane
1 ,2-Dibromoethane
Chlorobenzene
1,1, 1 ,2-Tetrachlorethane
Ethyl benzene
p-Xylene
o-Xylene
Styrene
Bromoform
Isopropylbenzene
p-Bromofluorobenzene
Bromobenzene
1, 1,2,2- Tetrachloroethane
1 ,2, 3- Trichloropropane
n-Propyl benzene
2-Chlorotoluene
4-Chlorotoluene
tert-Butylbenzene
1 ,3,5-Trimethylbenzene
sec-Butyl benzene
1 ,2-Dibromo-3-chloropropane
1,2,4- Trimethylbenzene
1 , 3-Dichlorobenzene
p-lsopropyl toluene
1 , 4-Dichlorobenzene
1 ,2-Dichlorobenzene-d4
1 ,2-Dichlorobenzene
n-Butyl benzene
1,2,4- Trichlorobenzene
Hexachlorobutadiene
Naphthalene
1,2,3- Trichlorobenzene
Avg
Area
164
775
1656
148
600
3761
6259
7569
6001
9273
4237
7051
5472
8387
6531
5909
5549
14743
9184
6296
4802
3533
7310
19477
3983
3833
6476
5574
4726
11075
5317
26235
17955
22680
8348
3362
9334
Nl
17581
8022
7490
27107
24763
20877
20987
26847
31773
4276
26210
8182
22003
9024
Nl
8181
27107
5253
3885
16109
4641
Rsd
(%)
44
10
9
42
6
6
7
1
6
5
12
1
3
3
3
3
3
3
3
4
8
1
3
4
4
11
7
4
3
4
1
2
5
3
7
8
6
-
6
3
16
6
7
1
7
9
5
6
6
5
4
-
16
6
5
6
7
3
Recov-
ery Avg
(%) Area
33
65
67
89
65
71
78
85
81
87
73
77
94
88
89
84
87
87
89
81
90
98
87
164
100
90
89
98
91
83
92
85
103
80
48
89
87
-
93
95
80
87
87
95
95
90
94
92
95
80
86
91
-
86
87
75
87
77
70
161
402
479
69
103
2195
2840
4330
1718
3058
2749
1193
1175
3235
3387
2052
2795
5308
1739
1885
1740
745
2272
6488
1107
1053
1098
1141
717
1208
1043
4042
2634
4004
602
516
2118
Nl
1515
1902
1377
3577
1894
1270
2887
3231
3434
554
4795
422
2163
564
Nl
400
2073
75
92
1399
88
Rsd
(%)
20
37
7
74
73
1
4
2
9
13
24
63
15
18
14
12
19
7
20
14
15
23
21
13
27
16
31
31
24
28
23
20
18
18
79
24
18
-
47
22
53
21
64
30
24
13
18
64
10
22
21
22
-
4
24
47
27
27
53
Recov-
ery Avg
(%) Area
33
34
19
42
11
42
35
49
23
37
48
13
20
34
46
29
44
31
17
24
33
21
27
16
28
25
15
20
14
9
18
13
15
14
3
14
20
-
8
22
15
9
7
6
13
11
10
12
17
4
8
6
-
4
7
1
2
7
1
94
454
1330
53
476
2969
5027
6107
5188
7318
4298
5978
4421
6960
5515
4805
4572
10710
6381
5046
4000
2741
5891
12893
3145
2604
4793
4236
3904
8006
3796
19240
11383
19003
8362
2737
6437
Nl
12394
6126
642
20731
16840
12792
12863
155410
16769
3231
16259
5165
11430
5430
Nl
5459
12660
2687
1617
8811
2653
Rsd
(%)
33
22
8
12
34
3
2
5
6
4
4
3
4
2
5
5
5
3
3
4
5
1
2
5
4
9
6
6
1
11
4
8
19
13
19
7
11
.
2
4
126
18
15
11
22
16
22
6
23
15
18
15
.
16
23
24
22
20
24
Recov-
ery Avg
(%) Area
19
38
54
32
52
56
62
67
70
69
74
66
76
73
75
68
71
63
62
65
75
76
70
119
79
91
66
74
75
60
66
62
65
67
48
72
60
-
65
72
59
53
59
5448
58
52
48
70
59
50
45
55
.
57
41
38
36
42
40
753
885
1693
81
410
2310
4377
5151
3703
5390
'1863
4208
3373
4462
2957
2853
2424
6658
5236
2664
2629
1920
3782
6737
2025
946
3527
2263
2058
3460
1665
5726
3102
4774
2006
1192
1801
Nl
4719
3583
4142
5033
4145
3508
2811
3494
3188
1684
4222
1266
2287
1325
Nl
1302
2667
473
176
2637
461
Recov-
Rsd ery
(%) (%)
18
4
3
35
8
4
9
9
9
13
18
13
14
18
21
21
20
16
16
22
24
17
23
31
24
37
23
30
24
38
34
39
30
38
33
31
36
.
37
32
29
44
44
44
43
42
43
40
38
44
41
47
_
47
42
48
59
44
50
39
35
44
43
48
42
47
65
48
54
40
54
67
55
47
46
43
59
82
47
57
68
59
48
64
32
69
61
69
43
51
40
33
39
24
63
27
.
42
57
67
25
29
25
21
20
17
62
25
23
18
23
_
25
16
13
9
24
13
Nl = not included
-------
720
700
80
60
40
20
0
Area (thousands)
Decreasing Headspace Volume ...
DCDF Methane
Chlorobenzene
Chloroform
135 Trimethylben
Benzene
Naphthalene
Figure 1. Effect of different headspace volumes on analyte recovery.
The linear dynamic range of this method
extends from the MDL of each analyte to
approximately 1000 x the MDL. Because
vial contents cannot be diluted or sub-
sampled after the vial is sealed without
losing headspace, for high concentration
samples multiple static headspace analy-
sis techniques2 should be investigated to
complement the single headspace evalu-
ations reported here. The 7000-HA has
the capacity to perform this type of analy-
sis, and further investigation is encour-
aged. Moreover, work must continue on
developing multiple internal standard meth-
ods to correct for matrix effects in soil.
References
1.G. Plemmons-Ruesink, USEPA, De-
velopment and Validation of a Sample
Preparation Procedure for the Analy-
sis of Organic Compounds in Soil and
Solid Matrices: Evaluation of Dynatech
PTA-30 W/S Autosampler, November
1990.
2. B.V. loffe and A.G. Vitenberg,
Headspace Analysis and Related
Methods in Gas Chromatography,
John Wiley and Sons, 1984.
3.J.W. Eichelberger and W.L. Budde,
U.S. EPA Method 524.2, Office of
Research and Development, Mea-
surement of Purgeable Organic Com-
pounds in Water by Capillary Column
Gas Chromatography/Mass Spectrom-
etry. Revision 3.0, 1989.
'U.S. Government Printing Office: 1993— 750-071/60231
-------
-------
P. Flores and T.A. BellarfMr. Bellaris retired) are with Environmental Monitoring
Systems Laboratory, Cincinnati, OH 45268.
James W. Eichelberger is the EPA Project Officer (see below).
The complete report, entitled "Determination of Volatile Organic Compounds in
Soils Using Equilibrium Headspace Analysis and Capillary Column Gas Chroma-
tography/Mass Spectrometry, "(Order No. PB93-155992; Cost: $19.50; subjectto
change) will be available only from
National Technical Information Service
5285 Port Royal Road
Springfield, VA 22161
Telephone: 703-487-4650
The EPA Project Officer can be contacted at
Environmental Monitoring Systems Laboratory,
U.S. Environmental Protection Agency
Cincinnati, OH 45268.
United States
Environmental Protection Agency
Center for Environmental Research Information
Cincinnati, OH 45268
Official Business
Penalty for Private Use
$300
BULK RATE
POSTAGE & FEES PAID
EPA
PERMIT No. G-35
EPA/600/SR-93/032
------- |