TN-4120-85-02
October 1985
AUTOMATED CRYOGENIC SAMPLING AND GAS
CHROMATOGRAPHIC ANALYSIS OF AMBJENT
jr.
VAPOR-PHAS£ ORGANIC. COMPOUNDS:
PROCEDURES AND COMPARISON TESTS
Submitted to
U.S Environmental Protection Agency
Environmental Monitoring Systems Laboratory
Research Triangle Park, NC 27711
Under Contract 68-02-4035
NORTHROP
-------
TN-4120-85-02
AUTOMATED CRYOGENIC SAMPLING AND GAS CHROMATOGRAPHIC ANALYSIS OF
AMBIENT VAPOR-PHASE ORGANIC COMPOUNDS:
PROCEDURES AND COMPARISON TESTS
by
Karen D. Oliver and Joachim D. Pleil
Northrop Services, Inc. - Environmental Sciences
Research Triangle Park, NC 27709
Submitted to:
Dr. William A. McClenny
Advanced Analysis Techniques Branch
Environmental Monitoring Systems Laboratory
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
Contract No. 68-02-4035
Reviewed and Approved by:
e,
N.E. Short
Program Manager
NORTHROP SERVICES, INC.
ENVIRONMENTAL SCIENCES
P.O. Box12313
Research Triangle Park, NC 27709
-------
NORTHROP TN41208502
Environmental Sciences
DISCLAIMER
This report has been reviewed by Northrop Services, Inc. - Environmental Sciences and
approved for publication. Mention of trade names or commercial products does not constitute
endorsement or recommendation for use.
-------
NORTHROP ™4208502
Environmental Sciences
FOREWORD
This report presents the results of work performed by Northrop Services, Inc. - Environmental
Sciences under Contract Number 68-02-4035 for the Advanced Analysis Techniques Branch,
Environmental Monitoring Systems Laboratory, U.S. Environmental Protection Agency, Research
Triangle Park, North Carolina. This work was conducted in response to Technical Directives 1.0-7 and
1.0-13 during the period May 1984 through March 1986.
-------
NORTHROP
Environmental Sciences
TN-4120-85-02
TABLE OF CONTENTS
SECTION PAGE
1 INTRODUCTION 1
2 EXPERIMENTAL 2
2.1 Equipment 2
2.2 Standards 2
2.3 Techniques 3
2.4 System Precision and Reproducibility 4
2.5 Nonlinearity of the ECD 4
2.6 Intercomparison Tests 4
3 RESULTS AND DISCUSSION 7
3.1 System Precision and Reproducibility 7
3.2 Nonlinearity of the ECD 7
3.3 Intercomparison Tests 8
4 CONCLUSIONS AND RECOMMENDATIONS 22
5 REFERENCES 23
IV
-------
NORTHROP
Environmental Sciences
TN-4120-85-02
LIST OF FIGURES
FIGURE PAGE
1 Schematic of Calibration System and Manifold 6
2 Nonlinear Response of Carbon Tetrachloride on the ECD 18
3 Nonlinear Response of Tetrachloroethylene on the ECD 19
4 Nonlinear Response of Hexachlorobutadiene on the ECD 20
5 Linear Response of Chloroform on the ECD 21
LIST OF TABLES
TABLE PAGE
1 Statistical Evaluation of Instrument Response Factors 10
2 Summary of Percent Differences in Area Counts for Daily Calibrations 11
3 Results of Intercomparison Tests 12
4 Summary of Intercomparison Tests 17
-------
NORTHROP
Environmental Sciences
TN-4120-85-02
SECTION 1
INTRODUCTION
A large number of volatile organic compounds are present at low concentrations in ambient
air, often at less than one part per billion by volume per compound. A current concern of the U.S.
Environmental Protection Agency is the quantitation of certain ambient volatile organic compounds
that may pose threats to human health.
In this laboratory, volatile organic compounds are collected by pulling ambient air through a
reduced temperature trap, so that various compounds are concentrated on the trap, but the major
air components, nitrogen and oxygen, pass through. These compounds are then thermally desorbed
onto a high-resolution capillary column, where they are separated by gas chromatography in
conjunction with oven-temperature programming and then detected simultaneously by electron
capture and flame ionization. The system has been automated so that consecutive samples can be
analyzed with minimal operator intervention. System design provides for analysis of samples from a
manifold at a stationary site or from canisters from the field.
The hardware configuration and automation of the cryogenic sampling and gas chromato-
graphic system has been addressed elsewhere (1). The purpose of this note is to present the
methodology of calibration and sample analysis, system precision, and results of interlaboratory
comparisons.
-------
NORTHROP
Environmental Sciences
TIM-4120-85-02
SECTION 2
EXPERIMENTAL
2.1 EQUIPMENT
The automated cryogenic sampling and gas chromatographic (GC) system consists of a
Hewlett-Packard 5880A Level 4 GC equipped with a 50-m by 0.31-mm-i.d. by 0.17-nm film thickness
OV-1 fused silica, high resolution capillary column (Hewlett-Packard, Avondale, PA), an electron
capture detector (ECD), a flame ionization detector (FID), and a modified Nutech 320-01 cryogenic
preconcentration unit (Nutech Corp., Durham, NC). Analysis of each sample requires 64 min - 15 min
for system initialization, 14 min for sample collection, 30 min for analysis, and 5 min for post time,
during which a report is printed. A detailed discussion of the automated sampling and analysis
system can be found elsewhere (1,2).
A Model MD-125-48(T) Perma-Pure dryer (Perma-Pure Products, Farmingdale, NJ) is used to
remove water vapor from the sample gas stream prior to analyte preconcentration. To prevent
excessive moisture buildup and any memory effects in the dryer, an automated cleanup procedure
has been devised in which the dryer is periodically heated and purged with zero air. Experiments
have been conducted to ensure that this analytical procedure does not alter sample integrity (2,3).
For the most recent comparison tests, a Hewlett-Packard HP-1 fused silica capillary column
(30-m x 0.530-mm-i.d. x 0.88-nm film thickness) was installed, and the dryer was not used.
2.2 STANDARDS
Two Scott Environmental Technology, Inc., pressurized cylinders containing a mixture of
volatile organic compounds (VOCs) at nominal concentrations of 10ppmv in nitrogen are used as
working standards. These cylinders contain the following compounds:
• Cylinder 1
benzene, toluene, o-xylene
• Cylinder 2
vinyl chloride methyl chloroform 1,2-dibromoethane
vinylidene chloride carbon tetrachloride tetrachloroethylene
1,1,2-trichloro-1,2,2- trichloroethylene chlorobenzene
trifluoroethane c;s-1,3-dichloropropene benzyl chloride
chloroform trans-1,3-dichloropropene hexachloro-1,3-butadiene
1,2-dichloroethane
-------
NORTHROP
Environmental Sciences
TN-4120-85-02
2.3 TECHNIQUES
2.3.1 Calibration Procedure
A 1.25-in. i.d. glass manifold is used for sample handling and preparation. As shown in
Figure 1, this manifold consists of three lengths of glass tubing, two of which are aligned
horizontally, one above the other. These attach at right angles to a third length of tubing which
extends through the roof of the mobile laboratory and arches into a "candy-cane" designed for
sampling ambient air. A blower attached to the end of the upper manifold pulls sample gas through
the tubing at a rate of 40 L/min. The upper manifold is used for ambient air samples and the lower
manifold for calibration gas. The calibration manifold is insulated and heated to 60°C.
For calibrations, the Scott gas mixtures are each passed through 0- to 50-cm3/min mass flow
controllers (MFCs; Tylan Corp., Carson, CA) into the calibration manifold at nominal flow rates of
2 cm3/min. Zero air is passed through a 0- to 5-L/min MFC at a rate of 2 L/min and is humidified with a
500-mL impinger flask containing deionized water. The two gas streams from the Scott cylinders and
the humid dilution air meet in 0.25-in.-o.d. Teflon tubing before reaching the manifold. To ensure
thorough equilibration, this dynamic mixture, now at nominal concentrations of 8 ppbv for each
compound in Cylinder 2 and 11 ppbv for each compound in Cylinder 1, is allowed to flow through
the manifold for a minimum of 6 h before sampling begins.
The calibration mixture is sampled from the manifold by pullfng 34 mL/min through
0.25-in.-o.d. Teflon tubing into a Teflon particle filter, and then through the Perma-Pure dryer and
into the analysis system using a 0- to 50-mL/min MFC. A total sample volume of 476 ml is collected
on the trap by sampling for 14 min. Normally, two consecutive analyses are performed, and the area
counts for each compound are averaged. The flow rates for the gas mixtures and zero air are
audited daily using a soap film bubble flow meter. Finally, precise dilution ratios, calibration
concentrations, and instrument response factors (ppbv/area) are calculated and entered into the
calibration tables for both detectors.
2.3.2 Ambient Air Samples
Ambient outdoor air is pulled into the upper manifold at a rate of 40 L/min through, the glass
"candy-cane" inlet tube (Figure 1). Ambient air is analyzed by attaching the sample tubing to the
upper manifold and pulling sample gas into the system, as described for calibration samples.
2.3.3 Canister Samples
Field samples are often collected in stainless steel canisters for analysis in this laboratory (4). If
the pressure in the canister is less than 12 psig when received, the sample is pressurized to 15 or
20 psig with nitrogen or zero air to ensure that enough gas is available to complete the run. The
-------
NORTHROP TN 4120 85 02
Environmental Sciences
pressure both before and after adding nitrogen or zero air is recorded, and the dilution factor is
calculated. The canister samples are then analyzed by venting the sample past the inlet to the system
at a known rate of about 70mL/min using a 0- to 500-mL/min MFC. This oversupplies the system
inlet, since the sample is collected at a rate of 34 mL/min.
2.4 SYSTEM PRECISION AND REPRODUCIBIL1TY
A statistical evaluation of calibration data collected during a six-week period was performed
to determine the combined inherent error in the system and the various calibration procedures. As a
test of reproducibility, the two daily calibration runs were compared for the data set discussed
above. One set of afternoon calibrations was not included in the calculations because of a
contaminant in the first run.
2.5 NONLINEARITYOFTHEECD
In the past, the concentrations of carbon tetrachloride, tetrachloroethylene, and hexachloro-
butadiene detected were frequently lower on the BCD than on the FID. FID concentrations were
generally in the expected range. When these three compounds were present at concentrations close
to calibration levels, both detectors agreed on the reported concentration. The problem appeared
to be nonlinearity of the ECD.
To confirm this theory, calibration gas samples which ranged in concentration from 0.1 to
8 ppbv were analyzed. Samples were prepared by altering the flow rates of zero air and the working
standards into the manifold or by pressurizing evacuated canisters with calibration gas and diluting
the gas mixture with zero air to known concentrations.
2.6 INTERCOMPARISON TESTS
A number of comparison tests have been conducted to determine how well analyses on this
system compare with analyses performed on systems in other laboratories. An explanation of each
test is presented below.
1. NBS (National Bureau of Standards). A certified gaseous mixture of five compounds
was analyzed.
2. RTI (Research Triangle Institute, Research Triangle Park, NC). Cylinder AAL-11933
was obtained from RTI and analyzed. The gaseous mixture was analyzed by RTI
before and after analysis in this laboratory (5).
3. GKPB (Gas Kinetics and Photochemistry Branch, EPA, Research Triangle Park, NC). A
sample of benzene, toluene, and o-xylene was placed in a stainless steel canister,
diluted, and analyzed. The sample was then analyzed in a GKPB laboratory by
cryogenic preconcentration-direct FID measurement of non-methane organic
carbon (NMOC)(6). Concentrations from that laboratory were converted from
ppbCto ppbv for comparison.
-------
NORTHROP TN-412WS02
Environmental Sciences
4. Scott (Scott Specialty Gases, Plumsteadville, PA). A cylinder containing six
compounds was analyzed twice within 10 months.
5. NSI-ES (Northrop Services, Inc. - Environmental Sciences, Research Triangle Park,
NC). A gaseous mixture of many of the calibration compounds used in this
laboratory was prepared in the NSI-ES Standards Laboratory (7). A sample was
injected into an evacuated canister and diluted with nitrogen to known
concentrations for analysis in this laboratory.
6. Matheson (Morrow, GA). A cylinder containing a gaseous mixture of four
compounds was prepared and certified by Matheson and analyzed in this
laboratory.
7. 22-City Study. Field samples from an EPA study were analyzed in this laboratory and
in the GKPB of EPA (6). Concentrations of benzene, toluene, and o-xylene were
compared for those samples analyzed by both laboratories.
8. RTI. Three additional audit cylinders from RTI (AAL-11948, AAL-12427, and AAL-
15182) were obtained and analyzed- These analyses were performed using the HP-1
capillary column without the Perma-Pure dryer.
-------
Calibration Gas
Cylinder 1
Calibration Gas
Cylinder 2
Air Blower
r
Ambient Air
Zero Air
Cylinder
Perma-Pure
Dryer
Out
To Analytical System
Teflon Filter
^•r Dry Purge
"^ Gas
In
Humidifier
Figure 1. Schematic of Calibration System and Manifold.
o
ro
o
oo
-------
NORTHROP
Environmental Sciences
TN-4120-85-02
SECTION 3
RESULTS AND DISCUSSION
3.1 SYSTEM PRECISION AND REPRODUCIBILITY
For calibration data collected over a six-week period, the mean, standard deviation, and
percent relative standard deviation (RSD) of instrument response factors for each compound on each
detector were calculated and are presented in Table 1. The experimental error in the system
averaged 3.5% RSD for compounds which respond on the BCD and 8.2% RSD for compounds which
respond on the FID. With the exception of benzyl chloride, o-xylene, and hexachlorobutadiene (FID),
the RSD for individual compounds on each of the detectors was less than 10.0%.
To test run-to-run reproducibility, the percent difference in area counts for each compound in
the afternoon calibrations was calculated for each day. The mean and standard deviation of percent
differences for each compound were then calculated and are presented in Table 2. The mean
difference for afternoon calibrations was between -0.20 and + 6.0% for compounds detected on the
FID and between -1.0 and +2.0% for compounds detected on the ECD. The standard deviation for
FID compounds was within ±7.0%. For ECD compounds, the standard deviation was within ±3.0%,
with the exception of 1,2-dibromoethane, which was ±6.1%.
3.2 NONLINEARITY OF THE ECD
Concentration versus area was plotted for the 0.1- to 8-ppbv calibration compound mixtures.
The plots of concentration versus area for calibration compounds detected by the ECD confirmed
that the curves are nonlinear for carbon tetrachloride, tetrachloroethylene, and
hexachlorobutadiene, as shown in Figures 2-4. The remaining six ECD compounds and the FID
compounds scale linearly for the range 0 to 8 ppbv, as shown for chloroform in Figure 5. For the
three compounds that are not linear over the calibration range, area counts generally begin to roll
off between 3 and 4 ppbv. To correct for the nonlinearity of these compounds, an additional
calibration step is necessary. An evacuated stainless steel canister is pressurized with calibration gas
at a nominal concentration of 8 ppbv. The sample is then diluted to approximately 3.5 ppbv with
zero air and analyzed. The instrument response factor (ppbv/area) on the ECD for each of the three
compounds is calculated for the 3.5-ppbv sample. Then, both the 3.5-ppbv and the 8-ppbv response
factors are entered into the ECD calibration table. Since software that runs the analysis system is
designed to accommodate multilevel calibration entries, the correct response factors will be
automatically selected for calculating concentrations.
-------
NORTHROP TN 4120 85 02
Environmental Sciences
3.3 INTERCOMPARISON TESTS
Table 3 contains the results of the interlaboratory comparison tests. Concentrations in ppbv
are reported, along with the percent difference in the two values. The percent difference (A%) was
calculated as follows:
Difference in concentration
A% = x 100
Average of concentrations
A negative A% indicates that the concentration reported by this laboratory was the lower of the two
concentrations. The mean A% for individual compounds in the intercomparison tests are presented
in Table 4.
The percent differences in concentrations in Table 3 show that most results from this
laboratory are in line with results reported by other laboratories. Many of the concentrations were
within ±20% of each other. Also, concentrations from this laboratory were sometimes greater and
sometimes less than concentrations reported by other laboratories, rather than being consistently
biased in one direction.
The comparisons with NBS, GKPB, NSI-ES, the second Matheson sample, and RTI 15182
revealed differences in concentrations within ±21%. The results from the first Matheson
comparisons showed differences between -6.0 and +24.0%. The RTI 11933 comparisons showed
differences between -3.0 and -21.0%, with the exception of tetrachloroethylene. For
tetrachloroethylene, the A% was lower for the second RTI 11933 comparison than for the first. The
first Scott comparison showed differences in concentrations of -4.6 to +38.6%. The second test,
conducted 10 months later, showed greater agreement with the exception of toluene,
chlorobenzene, carbon tetrachloride (ECD), and tetrachloroethylene (BCD). In the second
comparison, the effects of ECD roH-off can be seen in the carbon tetrachloride and
tetrachloroethylene concentrations, which were lower than the FID concentrations. (ECD roll-off did
not affect the first comparison, which was performed in 1983 using a sample volume of less than
476mL) The comparisons for the 10 22-city study samples were within ± 22%, with the exception of
samples 2409 and 1033, for which the concentrations were between -33.0 and +40.0%. The RTI
11948 comparison test resulted in differences between -10.4 and +3.8%, with the exception of
chloroform on the ECD, for which the values differed by +41.4%. This high ECD concentration was
unexplained and out of line with the FID analysis in which a + 3.8 A% was observed. FID results for
the RTI 12427 comparison showed differences between -12.7 and + 13.6%. However, ECD results
from this laboratory were much higher than the RTI analyses ( + 24.3 to 108 A%). High ECD
concentrations were presumably due to a change in the FID/ECD split ratio between calibration and
analysis. This has since been corrected.
-------
NORTHROP ™4208502
Environmental Sciences
Table 4 summarizes the intercomparison results. The average percent differences ranged from
-14.1 to +12.3%, with the exceptions of methyl chloroform ( + 20.7%) and 1,2-dibromoethane
( + 59.5%), both on the ECD. The high value for 1,2-dibromoethane was the average of two
comparisons, one of which was completely out of line with the other results.
-------
NORTHROP
Environmental Sciences
TN-4120-85-02
TABLE 1. STATISTICAL EVALUATION OF INSTRUMENT RESPONSE FACTORS (ppbv/area)
Compound
FID
vinyl chloride
vinylidene chloride
trichlorotrifluoroethane
chloroform
1,2-dichloroethane
methyl chloroform
benzene
carbon tetrachloride
trichloroethylene
c/s-1 ,3-dichloropropene
trans-1,3-dichloropropene
toluene
1 ,2-dibromoethane
tetrachloroethylene
chlorobenzene
o-xylene
benzyl chloride
hexachlorobutadiene
ECO
trichlorotrifluoroethane
chloroform
methyl chloroform
carbon tetrachloride
trichloroethylene
1 ,2-dibromoethane
tetrachloroethylene
hexachlorobutadiene
Mean, N a 33
0.56
0.43
0.48
1.28
0.41
0.38
0.14
1.71
0.38
0.29
0.29
0.11
0.44
0.33
0.13
0.09
0.16
0.28
2.39
6.97
1.98
0.80
5.59
2.96
1.07
0.88
Std. Dev.
0.04
0.03
0.03
0.08
0.02
0.02
0.01
0.08
0.02
0.02
0.02
0.01
0.04
0.02
0.01
0.01
0.03
0.05
0.09
0.21
0.06
0.01
0.13
0.18
0.04
0.04
RSD, %*
7.14
6.98
6.25
6.25
4.88
5.26
7.14
4.68
5.26
6.90
6.90
9.09
9.09
6.06
7.69
11.1
18.8
17.9
3.77
3.01
3.03
1.25
2.33
6.08
3.74
4.55
*RSD = relative standard deviation.
10
-------
NORTHROP
Environmental Sciences
TN-4120-85-02
TABLE 2. SUMMARY OF PERCENT DIFFERENCES IN AREA COUNTS FOR DAILY CALIBRATIONS*
Compound
FID
vinyl chloride
vinylidene chloride
trichlorotrifluoroethane
chloroform
1,2-dichloroethane
methyl chloroform
benzene
carbon tetrachloride
trichloroethylene
c/s-1,3-dichloropropene
trans-1,3-dichloropropene
toluene
1,2-dibromoethane
tetrachloroethylene
chlorobenzene
o-xylene
benzyl chloride
hexachlorobutadiene
ECD
trichlorotrifluoroethane
chloroform
methyl chloroform
carbon tetrachloride
trichloroethylene
1,2-dibromoethane
tetrachloroethylene
hexachlorobutadiene
Mean
(N = 32)
0.36
1.12
0.05
0.88
0.51
0.17
-0.15
-0.98
0.77
1.28
1.82
-0.19
1.46
1.18
2.04
0.33
4.39
5.54
0.44
0.46
0.29
0.07
0.75
0.34
-0.91
1.79
% Difference
Std. Dev.
2.86
4.32
4.08
6.47
3.05
2.46
2.54
5.75
3.12
3.43
3.71
3.15
4.02
3.45
4.03
3.84
6.04
5.85
1.24
1.71
1.17
0.38
1.33
6.10
2.90
1.68
'Percent Difference = (difference in area counts -=- average of area counts) x 100.
11
-------
§"<
§'0
i*
(» .*
3 ""^
ES
fa
TABLE 3. RESULTS OF INTERCOMPARISON TESTS (in ppbv)* I !g
1. NBS
EMSL
Compound Standard (N = 4) A*/>
2. RTI (AAL-1 1933) 2. RTI (AAL-11933) 3. GKPB 4. Scott (AAL-1 1745)
EMSL,
RTI. EMSL RTI. EMSL EMSL initial
Initial (N = 2) A% Second (N-2) A% GKPB (N = 1) A% Scott (N = 8) A%
FID
vinyl chloride
vinylidene chloride
(richlorotrifluoroethane
chloroform
1,2-dichloroethane
methyl chloroform
benzene
carbon tetrachloride
irichloroelhylene
c is-1,3-dichloropropene
trans-1,3-dichloropropene
loluene
1,2-dibromoethane
lelrachloroethylene
chlorobenzene
o-xylene
benzyl chloride
hexachlorobutadiene
ECO
I richlorotrifluoroethane
chloroform
melhyl chloroform
carbon tetrachloride
trichloroethylene
1,2-dibromoethane
letrachloroethylene
hexachlorobutadiene
38
8.7
77
64
382 +0.52
9.02 +3.6
783 +167
7 57 + 168
124 138 +10.7
38
77
4 59 + 18 8
886 +140
124 1248 +064
403 331 -196 408 331 -208
128 105 -19.7 126 105 -182
379 330 -138
323 286 -12.2
368 330 -109
296 286 -344
119 79 -40.4 103 79 -264
128 118 -813 126 118 -656
323 284 -129 296 284 -414
119 87 -31.1 103 87 -168
163 183 +116
189 183 -323
191 184 -373
18 21.1 +159
20 224 +113
16 207 +256
14 207 +386
16 203 +237
18 229 +240
20 19 1 -4 60
16 207 +256
16 213 +284
oo
en
•
o
to
A ilestnprion ol each test ii presented in Section 2 6
(continued)
-------
TABLE 3. (Continued)
4. Scott (AAL-1 1745)
EMSL,
second
Compound Scott (N = 1) A%
FID
vinyl chloride
vinylidene chloride
trichlorotrifluoroelhane
chloroform
1 ,2-dichloroethane
methyl chloroform
benzene 18 1989 +9.98
carbon tetrachloride 20 19.02 -5.02
trichloroethylene 16 18.61 +15.1
cis- 1 ,3-dichloropropene
(fans 1 ,3-dichloropropene
toluene 14 9.10 -42.4
1 ,2-dibromoe thane
tetrachloroethylene 16 18.33 +13.6
chlorobenzene 18 8.13 -7.56
o-xylene
benzyl chloride
hexachlorobutadiene
ECO
trichlorolrifluoroethane
chloroform
methyl chloroform
carbon telrachloride 20 9.60f -70.3
trichloroethylene 16 1926 +185
1 ,2-dibromoethane
tetrachloroethylene 16 10.24t -43.9
hexachlorobutadiene
t ECD roll oil resulted in this low measurement, this comparison is not used
5
Calculated
11.79
7.91
1365
11.96
9.44
1060
9.75
10.51
1033
7.29
9.21
927
7.80
791
1365
944
975
1051
7 29
921
. NSI-ES 6. Matheson Cylinder
EMSL
(N = 3)
12.07
801
14.01
11 14
952
1099
1028
1000
973
6.47
866
792
627
828
15 10
11 01
891
11 92
8 14
972
EMSL.
first
A* Calculated (N = 1) A%
+ 235
+ 1 26
+ 260
-7.10
+ 084 677 812 +18.1
+ 361 6.98 6.61 -545
+ 529 705 7.29 +3.35
-497 719 682 -528
-598
-11.9
-6.16
-157
-21 8
+ 4.57
+ 10 1
+ 154 677 854 +231
-900 705 788 +111
+ 126 719 686 -470
+ 110
+ 5 39
in subsequent calculations
6. Matheson Cylinder
EMSL,
second
Calculated (N = 1) A%
713 8.05 +121
736 658 -112
7.43 800 +7.39
758 6.90 -939
713 8.71 +19.9
743 794 +664
758 713 -612
(continued)
K)
O
do
Ul
O
K)
-------
TABLE 3. (Continued) J;
7. 22-City Study 7. 22-City Study 7. 22-City Study 7. 22-City Study 7. 22-City Study
(230H) (2409) (1053) (1033) (1027)
EMSL EMSL EMSL EMSL EMSL
Compound GKPB (N = i) A% GKPB (N = I) AS GKPB (N = I) A% GKPB (w = i) A% GKPB (N = I) A%
FID ~~~~~ ~~~~~
vinyl chloride
vinylidene chloride
trichlorotrifluoroethane
chloroform
1,2-dichloroethane
methyl chloroform
benzene 405 381 -611 4.10 296 -32.3 3.39 349 +291 1.27 1.64 +254 669 575 -15.1
carbon tetrachloride
trichloroethylene
cis-1,3-dichloropropene
trans-1,3-dichloropropene
toluene 741 680 -859 8.11 583-327 260 261 +038 421 539+246 15.35 1368-115
1,2-dibromoethane
tetrachloroethylene
chlorobenzene
o-xylene 1.59 162 +187 1.53 121 -23.4 071 1.06 +396 237 238 +042
benzyl chloride
hexachlorobutadiene
ECD
trichlorotrifluoroethane
chloroform
methyl chloroform
carbon tetrachloride
trichloroethylene
1,2-dibromoethane
tetrachloroethylene _)
hexachlorobutadiene f-
Sample identification number (continued) KJ
O
oo
Ul
I
o
KJ
-------
TABLE 3. (Continued)
7. 22-City Study 7. 22-City Study 7. 22-City Study 7. 22-City Study 7. 22-City Study
(2250) (1984) (1985) (2368) (2269)
EMSL EMSL EMSl EMSL EMSL
Compound GKPB (N=I) A% GKPB (N = I) A% GKPB (n-1) A% GKPB (N = l) A% GKPB (N = I) A%
__
vinyl chloride
vinylidene chloride
trichlorotri fluoroethane
chloroform
1,2-dichloroethane
methyl chloroform
benzene 156 158 +127 1.29 1.26 -2.35 2.14 2.02 -577 464 4.74 +213 1.72 1.99 +146
carbon tetrachloride
trichloroethylene
cis-1,3-dichloropropene
trans-1,3-dichloropropene
toluene 333 329 -1.21 419 393 -6.40 552 5.21 -578 1478 14.35 -295 3.58 388 +804
1,2-dibromoethane
telrachloroethylene
chlorobenzene
o xylene 060 073 +196 0.53 0.45-163 0.87 1.07 +206 219 267 +198 066 082 +216
benzyl chloride
hexachlorobutadiene
ECD
Irichlorotri fluoroethane
chloroform
methyl chloroform
carbon tetrachloride
trichloroethylene —(
1 2-dibromoethane '
&
tetrachloroethylene -•
KJ
hexachlorobutadiene O
— — oo
(continued) V1
-------
M"
fi
i:
TABLE 3. (Continued)
8. RTI(AAL-11948)
EMSL
Compound RTI (N = 2) A%
FID
vinyl chloride 20 20.0 0
vinylidene chloride
trichlorotrifluoroethane
chloroform 37 38.4 +3.8
1 ,2-dichloroethane
methyl chloroform
benzene 19 197 +3.7
carbon tetrachloride 9.6 9.1 -5.2
trichloroethylene
c/'s- 1 ,3-dichloropropene
trans- 1 ,3-dichloropropene
toluene
1,2-dibromoethane
tetrachloroethylene 9.1 9.2 +1.1
chlorobenzene
o-xylene
benzyl chloride
hexachlorobuladiene
BCD
trichlorotrifluoroethane
chloroform 37 52.3 +41.4
methyl chloroform
carbon tetrachloride 9.6 8.6 -10.4
trichloroethylene
1 ,2-dibromoethane
tetrachloroethylene 9.1 8.4 -7.7
hexachlorobutadiene
8. RTI (AAL-1 2427) 8. RTI (AAL-15182)
EMSL EMSL
RTI
-------
NORTHROP
Environmental Sciences
TN-4120-85-02
TABLE 4. SUMMARY OF INTERCOMPARISON TESTS
Compound
FID
vinyl chloride
vinylidene chloride
trichlorotrifluoroethane
chloroform
1,2-dichloroethane
methyl chloroform
benzene
carbon tetrachloride
trichloroethylene
c/s-1 ,3-dichloropropene
trans- 1 ,3-dichloropropene
toluene
1,2-dibromoethane
tetrachloroethylene
chlorobenzene
o-xylene
benzyl chloride
hexachlorobutadiene
ECD
trichlorotrifluoroethane
chloroform
methyl chloroform
carbon tetrachloride
trichloroethylene
1,2-dibromoethane
tetrachloroethylene
hexachlorobutadiene
N
3
2
2
5
2
4
20
9
6
-
-
16
2
7
4
11
-
-
2
5
4
8
6
2
6
-
Avq. A%
-13.5
+ 1.73
-t-4.53
-6.20
-9.90
+ 7.76
-0.41
+ 0.35
+ 1.86
-
-
-1.37
+ 0.85
-3.41
-14.1
+ 5.29
-
-
-0.67
+ 11.1
+ 20.7
-1.16
+ 12.3
+ 59.5
-3.53
-
17
-------
NORTHROP
Environmental Sciences
TN-4120-85-02
4567
Concentration (ppbv)
10
Figure 2. Nonlinear Response of Carbon Tetrachloride on the ECO. The solid lines approximate the
multi-level calibration used by system software for concentration calculations.
18
-------
NORTHROP
Environmental Sciences
TN-4120-85-02
100,000
0 >
456
Concentration (ppbv)
10
Figure 3. Nonlinear Response of Tetrachloroethylene on the ECD. The solid lines approximate the
multi-level calibration used by system software for concentration calculations.
19
-------
NORTHROP
Environmental Sciences
TN-4120-85-02
100,000 -
01 23456789
Concentration (ppbv)
Figure 4. Nonlinear Response of Hexachlorobutadiene on the ECO. The solid lines approximate the
multi-level calibration used by system software for concentration calculations.
20
-------
NORTHROP
Environmental Sciences
TN-4120-85-02
160,000 T
140,000 •
120,000 •
100,000 •
c
3
U
-------
NORTHROP
Environmental Sciences
TN-4120-85-02
SECTION 4
CONCLUSIONS AND RECOMMENDATIONS
The automated cryogenic sampling and GC system has been used successfully for trapping a
number of VOCs. The inherent error in the system and procedures has been demonstrated to
average 8.2% RSD for the FID and 3.5% RSD for the BCD. Also, reproducibility when using this
system is good; the average percent difference in area counts of compounds varied between -1.0
and +5.6%, as shown in daily calibrations. The intercomparison tests demonstrated that results
obtained with this system are generally within ±20% of results from other laboratories. The
intercomparison tests also showed that historically the results from this laboratory have improved
and have become more internally consistent as the technique has been further refined.
It is recommended that the list of VOCs for which the system is calibrated be expanded. Also,
experiments should be conducted with other types of capillary columns to develop procedures to
eliminate the Perma-Pure dryer and to add polar compounds to the calibration list. Since FID and
ECD results were occasionally inconsistent, the parallel FID/ECD split arrangement should be
eliminated. Instead, the ECD and FID should be arranged in series. This configuration has been used
in the-past with only minimal loss of resolution on the FID. Greater sensitivity also would be achieved
since the total sample volume would pass through both detectors rather than being split between
the two.
22
-------
NORTHROP TN-4120-85-02
Environmental Sciences
SECTION 5
REFERENCES
1. Pleil, J.D. 1982. Automated cryogenic sampling and gas chromatographic analysis of ambient
vapor-phase organic compounds: system design. EPA Contract No. 68-02-2566. Research
Triangle Park, NC; Northrop Services, Inc. - Environmental Sciences.
2. McClenny, W.A., J.D. Pleil, M.W. Holdren, and R.N. Smith. 1984. Automated cryogenic
preconcentration and gas chromatographic determination of volatile organic compounds. Anal.
Chem. 56:2947.
3. Pleil, J.D. and K.D. Oliver. 1985. Evaluation of various configurations of Nafion dryers: water
removal from air samples prior to gas chromatographic analysis. EPA Contract No. 68-02-4035.
Research Triangle Park, NC: Northrop Services, Inc. - Environmental Sciences.
4. Oliver, K.D., J.D. Pleil, and W.A. McClenny. In Press. Sample integrity of trace level volatile
organic compounds in ambient air stored in SUMMA® polished canisters. Atmos. Environ.
5. Jayanty, R.K.M., W.F. Gutkrtecht, and C.E. Decker. 1983. Status report #5: stability of organic
audit materials and results of source test analysis audits. EPA Contract No. 68-02-3431. Research
Triangle Park, NC: Research Triangle Institute.
6. McElroy, F.F., V.L. Thompson, D.M. Holland, W.A. Lonneman, and R.L. Seila. In Press. Cryogenic
preconcentration-direct FID method for measurement of ambient NMOC: refinement and
comparison with GCspeciation. JAPCA.
7. U.S. Environmental Protection Agency. 1984. Standard operating procedure: volatile organics
standards laboratory - standards preparation. EMSL/RTP-SOP-QAD-528. Research Triangle Park,
NC: U.S. Environmental Protection Agency.
23
-------
0
4567
Concentration (ppbv)
8
10
-------
1,000,000 •
900,000 •
c
u
ru
cu
<
400,000 '
300,000 •
0
456
Concentration (ppbv)
8
10
-------
200,000 • •
100,000 --
i
/
Jr
0
456
Concentration (ppbv)
8
10
-------
160,000 T
140,000 •
120,000 -
100,000 ••
o
------- |