ENVIRONMENTAL PROTECTION AGENCY
NATIONAL ENFORCEMENT INVESTIGATIONS CENTER
P.O. Box 25227 Denver Federal Center
Denver, Colorado 80225
March 1983
National Enforcement Investigations Center, Denver
CIS. Environmental Protection Agency
Office of Enforcement
-------
ENVIRONMENTAL PROTECTION AGENCY
NATIONAL ENFORCEMENT INVESTIGATIONS CENTER
P.O. Box 25227 Denver Federal Center
Denver, Colorado 80225
March 1983
GUIDELINES FOR
ANALYSIS OF VOLATILE ORGANIC COMPOUNDS IN AIR
A INTRODUCTION AND GUIDANCE FOR PLANNING FIELD STUDIES
B SAMPLE COLLECTION
C THERMAL DESORPTION AND GC/MS ANALYSIS
D PERMEATION TUBE PREPARATION AND CALIBRATION
E TENAX-GC® SAMPLE TRAP PREPARATION AND SCREENING
® Tenax-GC is a registered trademark of Enka N.V., The Netherlands.
-------
INTRODUCTION AND GUIDANCE FOR PLANNING FIELD STUDIES
-------
This introduction to the NEIC air sampling and analysis procedures provides
background information on air sampling by various means and the reasoning
behind the selection of procedures for use at NEIC. Any environmental an-
alytical data can be no better than the sampling plan and the protocols
used to collect samples. This is especially true for air sampling for or-
ganic compounds. The objectives of a field study (for example: great in-
terest in one or a few known compounds) can indicate, if not require, the
use of modifications to standard sampling protocols. Stated more explicit-
ly, air sampling and analytical methods need to be tailored to the object-
ives of the field study. Therefore, is is important that both field and
laboratory personnel understand the basic physical processes and principles
of organic air pollutant analysis.
Approaches to Air Sampling
Almost every organic analysis, including organics in air, depends on get-
ting the sample or sample extract onto some type of chromatographic column
for the actual quantisation step. (This discussion ignores in situ spectro-
scopic methods such as infrared because such methods are limited in the
number of compounds which can be determined and because the potential for
interferences limits their application to known atmospheres.) The most
straightforward, and therefore the best, way to analyze air would be to
place a measured amount of the air directly onto the chromatographic column.
However, using gas chromatography limits the sample size to a few cm3 at
most, so that this direct approach does not provide the necessary sensitiv-
ity in mos* cases.1 In order to increase sensitivity, three categories of
methods have been employed to trap organic compounds from a stream of air
to increase the effective sample size. The three types of methods are liq-
uid sorbents, cryogenic trapping, and solid sorbents.
Liquid Sorbents
Liquid sorbents, or bubbler trains, are inconvenient to use both from a
field and laboratory standpoint. Their use ought to be limited to special
situations where a chemical selectivity can be achieved in the trapping
step which outweighs the disadvantages of the technique.
-------
A-2
Cryogenic Trapping
Cryogenic trapping involves drawing the air sample over a cold inert sur-
face so that when the trap is warmed, the organics are contained in a small
volume. The method works best for very volatile compounds (gases at room
temperature) although it has been used for compounds as involatile as tri-
chlorobenzene.2 An obvious disadvantage of the method is the need for cry-
ogenic material, such as liquid argon or liquid oxygen or dry ice at the
very least, both to collect and store the samples before analysis. Other
problems encountered with cryogenic trapping are plugging of the trap with
frozen water vapor and reproducibly obtaining an inert trapping surface.
The most extensive use of cryogenic trapping has been by a group headed by
Hanwant Singh. They perform analyses in the field using a mobile labora-
tory to avoid problems associated with storing and shipping samples. The
analyses are done by GC using classical (not mass spec) detectors. Refer-
ence 3 contains a brief description of their methods and summarizes results
they have obtained in urban ambient air.
Solid Sorbents
Solid sorbents are the most applicable to general purpose air sampling be-
cause of their relative ease of use and storage and the wide range of com-
pounds which can be sampled. Analysis can be performed either by solvent
elution of the sorbent or by thermal desorption. With thermal desorption,
the entire sample is analyzed giving lower detection limits than solvent
elution for a given sample size. However, solvent elution nllows reanalysis
of the sample, which is not possible with thermal desorption. Thermal de-
sorption and solvent elution should be thought of as complimentary tech-
niques with a range of overlap in their applicability. Except for special-
ized methods, compounds compatible with solvent elution analysis are solids
at room temperature, while thermal desorption can be used for liquids and
more volatile solids. Compounds which are gases at room temperature can be
analyzed by thermal desorption with a judicious choice of solid sorbent.
-------
A-3
Choice of Sampling Technique
The method expected to be most applicable to NEIC needs (for example, air
sampling near a hazardous waste site) is trapping on Tenax solid sorbent
with analysis by thermal desorption. In situations where only relatively
non-volatile compounds such as pesticides, PCBs, or benzo(a)pyrene are of
interest, trapping on combination polyurethane foam (PUF)/XAD resin car-
tridges followed by solvent elution is recommended. The latter technique
has been successfully applied to a wide variety of compounds including or-
ganophosphorous pesticides by EPA personnel at HERL-RTP.4
The properties of Tenax are described more completely below. Briefly,
among the solid sorbents available, Tenax is applicable to the widest var-
eity of compounds while not collecting large amounts of water. Water
causes problems during the analysis because thermal desorption includes a
cryogenic trapping step, and ice plugs the trap if enough water was col-
lected with the air sample. Other materials which have been used as solid
sorbents have fewer of the desirable properties possessed by Tenax. Nearly
all adsorb more water. Some do not have the thermal stability of Tenax
and others, such as charcoal, can irreversibly adsorb compounds limiting
their range of applicability.
Properties of Tenax-GC
The structure of Tenax is shown below.
-0-/V-
-------
A-4
Tenax-GC is poly-p-2,6-diphenylphenyleneoxide, a porous linear polymer
first used as a packing material for gas chromatography. Several proper-
ties of Tenax make it suitable for collection of organic volatiles in air.
They are:
1- Excellent thermal stability. Thermal analysis methods (Differential
Scanning Calorimetry) have been used to confirm the stability of
Tenax. Breakdown does not occur below temperatures of 400° C.5 This
characteristic makes thermal desorption feasible and temperatures as
high as 400° C. will not increase background due to breakdown of the
polymer.
2. Quantitative desorption. Tenax does not exhibit losses due to irre-
versible adsorption as charcoal does.6 Decomposition of sorbates on
charcoal has been reported.7
3- Low background contamination. Virgin Tenax is Sohxlet extracted for
==18 hours with methanol. It is then thermally conditioned at an ele-
vated temperature (275-325° C.) for 20 to 30 minutes. Traps are
easily rebaked and returned to a low background state for reuse.
After solvent extraction and thermal conditioning, Tenax shows negli-
gible background contribution. However, the sampling process exposes
any solid sorbent to ozone and other oxidants which can be expected to
cause artifacts. Artifacts reported from the thermal decomposition
and/or reactions of Tenax include ethylene oxide7, alkylbenzenes8,
styrene8, benzene8, alkylphenols8, acetophenone9, and benzaldehyde9.
In our experience at NEIC, only acetophenone and benzaldehyde have
been observed as artifacts in upwind field samples.
4. High collection efficiency. Tenax collection efficiency of 100% has
been reported with a wide variety of compounds10 (when breakthrough
volumes are not exceeded). Testing has also been conducted under com-
monly encountered analytical conditions. Results indicate that col-
lection efficiency does not drop with repeated thermal desorption10'll
-------
A-5
and that breakthrough volumes are somewhat dependent on relative humi-
dity12 and C02 concentration12, even though no apparent decrease in
collection efficiency is observed with relative humidity10/13. (Col-
lection efficiency and breakthrough volume are not the same thing.)
5. Low affinity for water. Water vapor is poorly retained on Tenax14/ 1S,
a distinct advantage. In addition to the plugging problem mentioned
previously, adsorbed water presents a potential medium for hydrolysis
reactions and for collecting potentially reactive gases such as NO
and S0215.
Limitations of Tenax
Although Tenax has many desirable properties, it does have limitations.
Because of these limitations, it may be desirable to use other sorbents in
conjunction with Tenax. Materials such as charcoal8,16 and Ambersorb
®
XE-340 8 retain very volatile compounds much better than Tenax. Ambersorb
XE-340 is essentially carbonized XAD resin which has the desirable adsorp-
tive properties of charcoal, hopefully without the undesirable features
such as active sites due to metals. If it is necessary to sample very
volatile compounds, Ambersorb XE-340 can be used by itself or behind Tenax
traps.
Breakthrough Volumes
Two terms commonly used to describe collection efficiency in air are Elu-
tion Volume and Breakthrough Volume. Due to ambiguity by early researchers,
these terms were often used interchangeably?;10. To clarify the meaning,
they are defined below.
Elution Volume - The volume of air sampled which is required to move
the mass transfer zone to the end of the available packing bed.
Ambersorb XE-340 and XAD are registered trademarks of Rohm & Haas
Company.
-------
A-6
Breakthrough Volume - That volume of air which purged 50% of the
adsorbed vapor out of the cartridge.
The two definitions can be represented diagramatically.
FIGURE 1
ELUTION PROFILE
Concentration
at exit of trap
Volume Sampled
V = point where actual breakthrough of the sorbate begins (Elution
Volume)
VB = 50% breakthrough (Breakthrough Volume)
Another way of visualizing this concept is shown in Figure 2.
100%
% of Challenge
Concentration
at Trap Exit
FIGURE 2
FRONTAL PROFILE
50X Breakthrouah
Elution Volume
Volume Sampled
-------
.A-7
Elution analysis entails injecting a small quantity of adsorbate onto a
cartridge in a very small time. During frontal analysis, the sample injec-
tion is continuous.
In actuality, the elution volume and the 50% breakthrough volume may be
very similar for compounds with low breakthrough volumes. Compounds with
high breakthrough volumes will have a larger difference between the two
values.
Superimposing Figures 1 and 2 will show the relationship between frontal
breakthrough (Figure 2) and an elution peak (Figure I)11.
Frontal Profile
Concentration
Figure 3
-p V \J I UUIC
[_ 50% Breakth'rough_ (V Specie Restitution (Nation}
Elution ,
I Volume ~~~"
Factors Affecting Breakthrough Volume
Factors which affect or may affect breakthrough volume include temperature,
humidity, chemical class of the sampled compound, C02 concentration, the
total organic concentration in the sampled atmosphere or the concentration
of particular compounds, and changes in the sorbent surface caused by reac-
tions with ozone or other oxidizing species. All porous polymer sorbents
-------
A-8
work by allowing airborne compounds to diffuse into pores within the poly-
mer. This process requires a finite time so that collection efficiency
will eventually decrease as the sample flow rate in increased.17 This prob-
lem would probably never be encountered because the flow rates are higher
than are typically used or even possible with common pumping systems. How-
ever, one should be aware of the possibility. The dependence of break-
through volume on temperature is well known and has a theoretical basis16'
*? \ extrapolation of breakthrough volumes to other temperatures is an often
used and accepted procedure. Tabulated breakthrough volumes are useful in
estimating relative detection limits.
The dependence on relative humidity and C02 concentration has been deter-
mined empirically for selected compounds on Tenax.12 The breakthrough
volumes of four compounds decreased 22 to 43% upon going from 0 to 87% rel-
ative humidity in a laboratory situation. When C02 was added in addition
to water, breakthrough volumes decreased by approximately an additional 25%.
C. R. McMillin and co-workers sampled indoor and outdoor air using Tenax as
the first stage of a three-sorbent trapping system.8 They found great and
unexplained differences in the indoor and outdoor breakthrough volumes on
Tenax (other sorbents were not tested), implying that breakthrough volumes
determined in the laboratory cannot be used in quantitating field results.
These differences are the observable effects of the many factors affecting
breakthrough volume.
The only practical approach to obtaining quantitative field data is tc
avoid breakthrough by limiting sample size and to include tests for break-
through in the quality assurance plan so that you are aware if breakthrough
did occur. Tabulated breakthrc igh volumes or past experience can be used
to set sample size, but should not be used for quantisation.
Generic Quality Assurance Plan for Air Sampling
The procedures listed are recommended as a general approach to any air
sampling study. The reasons for each procedure are also listed.
-------
A-9
1. Keep the sampling time to the minimum which will give adequate sample
volume. This point depends on the objectives of the study. Presum-
ably, you are trying to detect contributions from a particular source.
The shorter the sampling time, the more stable weather (wind speed and
direction) conditions will be during the sampling, resulting in better
directional resolution in the data. An obvious corollary is to col-
lect all samples as nearly simultaneously as possible.
2. Do not use a sample volume larger than necessary to give the desired
detection limits. During sampling, both the solid sorbent and adsorbed
compounds are exposed to ozone and other reactive species. Reactions
between adsorbed compounds are also possible. These possibilities
cannot be avoided, but minimizing the sample size also minimized their
effects.
3. Take all samples using tandem (connected in series) sorbent tubes with
a sample size chosen to avoid breakthrough on the first tube. This
procedure is based on the assumption that quantitative data for more
volatile compounds are given higher priority than lower detection lim-
its for less volatile compounds. Avoiding breakthrough is necessary
to obtain quantitative data; analyzing tandem tubes provides a check
for breakthrough.
4. Always take duplicate samples downwind of the source as a minimum;
triplicate samples are recommended. The downwind samples are the most
important samples because they are most likely to show positive results;
the upwind sample is the next most important because it shows back-
ground levels. Taking duplicate samples downwind helps to ensure that
data will be available for the most important sampling point « ven if
one laboratory analysis fails. (Using thermal desorption, there is
only one chance at the analysis.) The duplicate or triplicate samples
also serve as a check of sampling and analysis precision.
5. Take duplicate breakthrough spikes upwind of the source in addition to
the upwind sample(s). A breakthrough spike consists of a tandem pair
-------
A-10
of tubes with the first tube spiked with known amounts of target com-
pounds. The tandem pair is then sampled over in the normal manner. A
breakthrough spike is the most valid matrix spike which is practical
in most air sampling projects. The results obtained from the break-
through spike are the best performance that can be expected from the
sampling and analysis procedure. If a spiked compound is not retained
on a tube during sampling, it cannot be expected to be sampled relia-
bly. The duplicate breakthrough spikes are a check of the precision
and accuracy of the methodology.
6. Carry triplicate field spikes to the field. The field spikes are sam-
ple tubes which are spiked with target compounds but not sampled over.
The field spikes serve to check for losses of target compounds during
shipment and storage and as a check of analytical precision. At the
time the spiking for the breakthrough and field spikes is performed,
an additional spike is performed which is stored in the laboratory
under the best possible conditions. This spike serves as a reference
for the field and duplicate spikes.
7. Include a field blank in each container of sample tubes to check for
contamination during storage or shipment.
8. One can expect artifacts from any sampling technique. Be aware of the
artifacts which can be expected with the methodology employed.
9. Know the principles involved in air sampling methodologies and avoid
situations which might cause artifacts or errors. For example, two
situations to avoid would be exposing sample traps to high concentra-
tions of ,rganic vapor such as gasoline fumes, and exposing the traps
to high temperatures during shipment or storage.
-------
A-11
FOOTNOTES AND REFERENCES
1. Tom Spittler and co-workers in U.S. Environmental Protection Agency
Region I, Boston, have very successfully employed this direct approach
to monitor solvent-type compounds.
2. H. B. Singh, L. J. Salas, A. J. Smith, and H. Shigeishi, Atmos.
Environ., 15, 601 (1981).
3. H. B. Singh, L. J. Salas, and R. E. Stiles, Environ. Sci. Technol.,
16, 872 (1982).
4. R. G. Lewis and K. E. MacLeod, Ana2. Chem., 54, 310 (1982).
5. "Selection and Evaluation of Sorbent Resins for the Collection of
Organic Compounds", A. D. Little, Inc., EPA/600/7-77/044, April 1977.
6. E. 0. Pellizzari, B. H. Carpenter, J. E. Bunch, and E. Sawicki,
Environmental Science and Technology, 9, 556 (1975).
7. "neveJopment of Method for Carcinogenic Vapor Analysis in Ambient
Atmospheres", Research Triangle Institute, EPA-650/2-74-121, July 1974.
8. "Potential Atmospheric Carcinogens, Phase 2/3, Analytical Technique
and Field Evaluation", Monsanto Research Corp., EPA-600/2-81-106,
June 1981.
9. "Artifact Problems in Atmospheric Analysis of Organic Compounds and
Strategies for Minimization," R. E. Sievers, presented at National
Symposium on Monitoring Hazardous Organic Pollutants in Air, Raleigh,
N.C., April 28 to May 1, 1981. Also by NEIC experience.
10. "Development of Analytical Techniques for Measuring Ambient Atmospheric
Carcinogenic Vapors", Research Triangle Institute, EPA-600/2-75-076,
November, 1975.
11. "Characterization of Sorbent Resins for Use in Environmental Sampling",
Research Triangle Institute, EPA-600/7-78-054, March 1978.
12. "Further Characterization of Sorbents for Environmental Sampling",
A. D. Little, Inc., EPA-600/7-79-216, September 1979.
13. E. D. Pellizzari, J. E. Bunch, R. E. Berkley, J. McRae, Analytical
Letters, 9, 45 (1976).
14. Dravnieks, et al., Environmental Science and Technology, 5, 1220
(1971). ** -'
15. "Analysis of Organic Air Pollutants by Gas Chromatography and Mass Spec-
troscopy, Final Report", Research Triangle Institute, EPA-600/2-79-057,
March 1979.
16. K. J. Krost, E. D. Pellizzari, S. G. Walburn, and S. A. Hubbard,
Ana2. Chem., 54, 810 (1982).
17. "Characterization of SorJbent Resins for Environmental Sampling," A. D.
Little, Inc., EPA-600/7-78-054, March 1978.
-------
B SAMPLE COLLECTION
-------
ENVIRONMENTAL PROTECTION AGENCY
NATIONAL ENFORCEMENT INVESTIGATION CENTER
Box 25227 Denver Federal Center
Denver, Colorado 80225
Volatile Organic Air Pollutant Analysis
Sample Collection
March 1983
1.0 Introduction
1.1 This procedure describes the collection of air samples on sampling
tubes containing a solid sorbent. Tenax-GC is the most commonly
used sorbent. Adsorbed organic compounds are detected by thermal
desorption of the sorbent onto a gas chromatograph column for
GC/MS analysis. The procedure presented here was designed to per-
form reasonably well on a wide range of compounds. In general,
those organic compounds that are liquids at room temperature are
well suited for analysis using this method and Tenax as an adsor-
bent. If only compounds of a narrow volatility range are of in-
terest, it is probable that the sample size and/or sorbent mater-
ial could be changed to yield superior performance for the com-
pounds of interest. For example, compounds as low in volatility
as benzo(a)pyrene have been analyzed by similar procedures. Al-
though this procedure was specifically designed for the use of
Tenax sorbent, other sorbents can be used.
2.0 Limitations
2.1 The sample traps are essentially short chromatographic columns.
Retention of chemicals is dependent upon adsorption characteris-
tics of the chemical/resin system. Factors influencing retention
include: temperature, flow rate, air volume, vapor pressure of
the chemical, and sample matrix. Volatile species like vinyl
chloride are only moderately retained while other chemicals like
chlorobenzene are retained very well. All chemicals will experi-
ence breakthrough under the correct conditions. Table I lists
breakthrough volumes for some relevant chemicals. The volumes
represent the amount of air sampled when 50% of the collected
chemical is lost through the trap. This data was compiled by
Pellizzari in Reference 9.7. Data for chemicals where the sample
volume exceeds the breakthrough volume represent minimum concen-
trations. The data in Table 1 can be used to estimate an appro-
priate sample size. When doing this, one must decrease the vol-
umes shown in Table 1 by the factor §f because the NEIC sample
traps contain 0.8g of Tenax rather than 2.2g. In general, the
sample size should be 25 liters or less.
Registered Trademark of Enka N.V., The Netherlands; appears hereafter
without the ®.
-------
B-2
Chemical Class
•~^^^—^^«.^
Alcohols
Aldehydes
Amines
Aromatics
Esters
Ethers
Halogenated
Ethers
Halogenated
hydrocarbon
Table 1
Tenax GC Breakthrough Volumes for Selected Compounds at Various Temperatures
Data Taken From Reference 9.7
Compound
Methanol
n-Propanol
Allyl alcohol
Acetaldehyde
Benzaldehyde
Dimethyl amine
Isobutylamine
t-Butylamine
Di-(n-buty1)amine
Pyridine
Aniline
Benzene
Toluene
Ethyl benzene
Cumene
Ethyl acetate
Methyl acrylate
Methyl methacrylate
Diethyl ether
Propylene oxide
2-Chloroethyl ethyl ether
Bis-(chloromethyl)ether
Methyl chloride
Methy.l bromide
Vinyl chloride
Methylene chloride
Chloroform
Carbon tetrachloride
1,2-Dichloroethane
1,1,1-Trichloroethane
Tetrachloroethylene
Trichloroethylene
l-Chloro-2-methylpropene
3-Chloro-2-methylpropene
1,2-Dichloropropane
1,3-Dichloropropane
Epichlorohydnn (1-chloro-
2,3-epoxypropane)
3-Chloro-l-butene
Allyl chloride
4-Chloro-l-butene
l-Chloro-2-butene
^ — _
b p (°
64.7
97.4
97
20
179
7.4
69
89
159
115
184
80 1
110.6
136.2
152.4
77
80
100
34.6
35
108
-24
3.5
13
41
61
77
83
75
121
87
68
72
95
121
116
64
45
75
84
C)
1
27
32
3
7,586
9
71
6
9,506
378
8,128
108
494
1,393
3,076
162
164
736
29
13
468
995
8
3
2
11
42
34
53
23
361
90
26
29
229
348
200
19
21
47
146
60
Temperature (°F)
70
80
Breakthrough Volumes for 2.7 q
1
20
23
2
5,152
6
47
5
7,096
267
5,559
77
348
984
2,163
108
111
484
21
9
336
674
6
2
1.5
9
31
27
41
18
267
67
20
22
162
253
144
15
16
36
106
0.8
14
16
2
3,507
4
34
4
4,775
189
3,793
54
245
693
1,525
72
75
318
15
7
241
456
5
2
1.25
7
24
21
31
15
196
50
16
17
115
184
104
12
12
27
77
0.6
10
11
1
2,382
3
23
3
3,105
134
2,588
38
173
487
1,067
48
50
209
11
5
234
309
4
1
1.0
5
18
16
23
12
144
38
12
13
81
134
74
9
9
20
56
90
100
Tenax (liters)
0.4
7
8
0.9
1,622
2
16
2
2,168
95
1,766
27
122
344
750
32
34
137
3
4
124
209
3
1
0.8
4
13
13
18
9
106
28
9
10
58
97
54
7
6
15
40
0 3
6
0.7
1,101
11
1
1,462
67
1,205
19
86
243
527
22
23
90
3
89
142
2.5
0.9
0.6
3
10
10
14
j
78
21
7
i
a
41
70
39
12
29
-------
B-3
Table 1 (cont.)
Tenax GC Breakthrough Volumes for Selected Compounds at Various Temperatures
Data Taken From Reference 9.7
Chemical Class
Halogenated
hydrocarbon
(cont. )
Hydrocarbons
Inorganic gases
Ke tones
Nitrogenous
hydrocarbons
Oxygenated
hydrocarbons
Sulfur
Compounds
Compound
Chlorobenzene
o-Dichlorobenzene
m-Dichlorobenzene
Benzyl chloride
Bromoform
Ethyl ene di bromide
Bromobenzene
n-Hexane
n-Heptane
1-Hexene
1-Heptene
2,2-Dimethylbutane
2,4-Diemthylpentane
4-Methyl-l-pentene
Cyclohexane
Nitric oxide
Nitrogen dioxide
Chlorine
Sulfur dioxide
Water
Acetone
Methyl ethyl ketone
Methyl vinyl ketone
Acetophenone
Ni tromethane
Aniline
Acrolein
Glycidaldehyde
Propylene oxide
Butadiene diepoxide
Cyclohexene oxide
Styrene oxide
Phenol
Acetopheonone
b-Propiolactone
Oiethyl sulfate
Ethyl methane sulfate
b.p. ('
132
181
173
179
149
131
155
68 7
98.4
63.5
93.6
49.7
80.5
53.8
80.7
_
-
.
100
56
80-2
81
202
101
184
53
34
132
194
183
202
57
208
86
50
60
Temperature (°F)
70
°C) Breakthrouah Volumes
899
1,531
2,393
2,792
507
348
2,144
32
143
28
286
0.5
435
14
49
0
0
0
0.06
0.06
25
82
84
. 5,346
45
3.864
19
364
35
1,426
2,339
5,370
2,071
3.191
721
40
5,093
653
1,153
1,758
2,061
386
255
1,521
23
104
20
196
0.4
252
10
36
o
o
o
0.05
0 05
17
57
58
3,855
34
2.831
14
247
24
1,009
1,644
3,926
1,490
2,382
514
29
3,681
473
867
1,291
1,520
294
188
1,079
]?
75
15
135
0.3
146
8
26
o
o
o
0.03
0 04
12
39
40
2,767
25
2,075
10
168
17
714
1,153
2,870
1,072
1,778
366
21
2.564
80
90
100
for 2 2 q Tenax (liters)
344
656
948
1,125
224
138
764
17
1 £
55
11
93
0.2
84
6
19
o
g
0 02
0 03
27
28
2,000
19
1,520
8
114
11
506
811
2,094
769
1,327
261
15
1,914
249
494
697
830
171
101
542
39
8
64
0.2
49
4
14
0.02
0.01
19
19
1,439
14
1,114
6
77
8
358
570
1,531
554
991
186
11
1,384
181
372
510
612
131
74
384
29
6
44
0.1
28
3
10
0
0
0 01
0
13
14
1,037
11
817
4
52
5
253
400
1,119
398
740
132
8
998
-------
B-4
2.2 The accuracy of the data produced from the analysis of samples ob-
tained using this procedure depends on the care with which samp-
ling is performed. Particular attention must be given to the cal-
ibration of pumps, checking to demonstrate that the sampling rate
was constant, and to the handling of sample tubes to avoid contam-
ination. The sampling tubes should be kept in sealed culture
tubes except for the time required for set up and sampling. The
tubes should never be handled without using nylon gloves or tis-
sues to prevent contamination by body oils.
2.3 The data in Table 1 show that the effects of temperature on sample
breakthrough volume are significant. For many of the compounds
listed in Table 1, the breakthrough volume at 90° F is only 20 to
25% of the breakthrough volume at 50° F.
2.4 In order to check for sample breakthrough, each sample is taken
using a tandem tube arrangement. If a particular compound is de-
tected on the first tube but none is seen on the second tube,
then that compound did not experience breakthrough.
3.0 Equipment
3.1 Sampler - DuPont model P4000 or equivalent personnel sampler.
Capable of adjusting and monitoring the flow over the range of
0.1 to 1 liter per minute (2pm) with a trap in place.
3.2 Mass flow meter - Portable unit equipped with a teflon fitting to
measure the flow through a sampling trap. It should have a range
of 0-2 £pm.
3.3 Sample traps - Glass sampling traps packed with the selected sor-
bent. See the procedure "Tenax Sample Trap Preparation and
Screening".
3.4 Sampling line - 2-5 feet of 1/4" o.d. tygon tubing with a teflon
fitting at one end to attach to the sampling traps.
3.5 Swagelok - 5/8" union.
4.0 Calibration Procedure
4.1 A mass-flow meter is used in line between the pump and the adsor-
bent traps to calibrate the pump before san >ling begins at each
station. The pump is rechecked with the mass-flow meter after
the sampling period is complete.
5.0 Sample Collection
5.1 Sample tubes are packaged inside screw cap culture tubes placed
in metal cans with a compression fit closure for transport to the
field. The NEIC employs virgin paint cans for packaging. Using
a clean tissue or wearing a nylon cloth glove, remove a numbered
sample trap from its culture tube, and reseal the culture tube.
-------
B-5
5.2 Inspect the trap for damage such as broken glass, loose glass wool
plugs, or spilled resin. If the trap is damaged, replace in the
culture tube and return to the laboratory unused.
5.3 Attach the tandem traps to the calibrated sampling pump. See
Figure 1.
5.4 Begin sampling, noting the start time and sample pump flow meter
reading. Select sample volumes so as to avoid breakthrough of
target pollutants from the first trap to the second. For most
purposes a 25-liter sample collected at 0.5-1.0 liters per minute
is desirable.
5.5 Record the weather conditions occurring during sampling including
temperature, wind speed and direction, humidity, and barometric
pressure.
5.6 Stop sampling, noting the end time and sample pump flow meter
reading. Replace the trap into a culture tube. Reseal with the
teflon-lined septum cap and tag. Note on the tag the trap number
and whether the trap was the front or back of the tandem pair.
The front trap is the one sampled air passes through first.
5.7 Return culture tubes containing sample traps to the paint can
and reseal the can. Be sure to tag the "field blank" sample in
each can and any field spikes (will be total of at least three).
6.0 Quality Control
6.1 Sample pumps are calibrated daily. During sampling any flow rate
changes are noted by monitoring the flow meter on the sampler.
Changes in flow up to 10% are acceptable. If the change is greater
than 5%, the beginning and ending flow rates are averaged to give
the flow rate.
6.2 Triplicate samples indicate the reproducibility of the overall
sampling and analysis. Triplicate samples will be collected at
least at one sampling station. The triplicate sample station
(or one of the duplicate sample locations) should be that station
most directly downwind of the source being sampled. The tripli-
cates should be collected at the same place, at the same flow
rate and at the same time. It is very important to have the sam-
pling as identical as physically possible.
6.3 Breakthrough spikes give an indication of which compounds would
have broken through under the field conditions of the sampling.
A breakthrough spike [Figure 2] consists of a tandem pair of
traps, the front trap having been spiked with a standard set of
compounds in the laboratory. A sample is then collected in the
normal manner using the tandem traps. Duplicate breakthrough
spikes should be sampled at the same time and rate as the regular
field sample at the sample station most directly upwind of the
source being sampled.
-------
FIGURE 1: Tandem Air Sample Traps
FLOW
To Pump
— 1
(\ • 1
•» 1
. — j j ;N x^~ '•• '•' ' ••' " ' • • '••••• i < )
Ijl I"""! fc'';';::.:, •''-:. '•- :^: N I
>~j •* — ^^ ' ••'-.' -•-. • h i
1/4" Teflon Union Air Sample Trap \L F
- h
i
i
'• X
/ :_.'•;,--.'. - ^
':'•' Z. '•''•'•• s ' ''''•.
Air Sample Trap
5/8" Brass Union
with Teflon ferrules
Figure 1. Tandem Air Sample Traps
CO
CTl
-------
FIGURE 2: Breakthrough Spike
FLOW
To pump
1
1/4" Teflon Union
*'
!
!?fe ; ,".'.':'•• ,' • :> I
J&¥ '• .,'' >';.'.•/./';. '•" ;?--; 1 i
^*~*1 * ' ',.. ,..,... j »
Unspiked Sample Trap (j
I
!
)
i
!
. I
„, „ , 1
\ 1
.
r*
i
)
i
t
' S
\
l_
••:' -. . ' •:-.- ;. . • '^
•'•" •- "- ', ; -', - "---s -. - ('- ,
-v • •*.•"•.. • -. -:-y
Soiked Samole Trao
5/8" Brass Union with
Teflon ferrules
Figure 2. Breakthrough Spike
CD
I
-------
B-8
6.4 Contamination in each sample transport container (paint can) is
monitored by a "field blank". A field blank is a sample trap
which is not sampled. It is transported and stored the same as
the samples. Field personnel designate the trap to be used as
the field blank.
6.5 Deterioration of the samples is monitored by a "field spike". A
field spike is a sample trap which is spiked in the laboratory
prior to going out to the field. The field spike is stored and
transported alongside the samples.
6.6 At the time laboratory personnel prepare the field and breakthrough
spikes, they will prepare a reference spike which remains in the
laboratory.
6.7 Samples can be stored in a dark, organic vapor-free area at -20° C
for up to four weeks before analysis according to Reference 9.10.
7.0 Options
7.1 In the event of unknown atmospheres suspected of containing high
levels of contaminants, two samples could be collected, one at
the normal sampling rate, and another at one tenth the normal
rate.
7.2 If specific compounds are of special interest, flow rates and
sampling times may be changed (e.g., a compound with a high break-
through volume, suspected in low concentrations might be sampled
at 1 liter per min for 100 minutes).
7.3 If particulate matter may provide an unwanted contribution to sam-
pled organics, filters -are available which will prevent particu-
late greater than 0.5 urn in size from reaching the sorbent traps
(and thus being thermally desorbed when the trap is analyzed).
Any filter must be used with the realization that organic com-
pounds may be stripped from particulates on a filter by the sam-
pled air flow, so that a total elimination of the contribution
of organics from particulate is not possible. However, compounds
stripped from particulates by the sampling process would have to
be considered readily available for volatization, and would prob-
ably be of interest.
7.3.1 The filters used are sold commercially as filters for liq-
uid chromatography solvents. The filters are Millex -SR,
Millipore Corp., Bedford, MA, Catalog #SLSR025NS, 0.5 urn
PTFE (polytetrafluoroethene). The PTFE filter itself is
encased inside a hard plastic holder equipped with leur
fittings. Other filter pore sizes are available.
7.3.2 The filters can be connected to sample traps by forcing one
end of the plastic case directly into the quarter-inch end
of the glass trap.
-------
B-9
8.0 Sample Analysis
8.1 Samples are analyzed by the procedure "Thermal Desorption and
GC/NS Analysis of Air Samples" (see Section C).
9.0 References
9.1 Bertsch, Wolfgang, Chang, Ray C. and Albert Zlatkis, 'The Determi-
nation of Organic Volatiles in Air Pollution Studies: Characteri-
zation of Profiles", Journal of Chromatographic Science. Vol. 12,
pp 175-182, April 1974.
9.2 Pellizzari, Edo D., "Development of Method for Carcinogenic Vapor
Analysis in Ambient Atmospheres", EPA-650/2-74-121, July 1974.
9.3 Pellizzari, Edo D., Bunch, John E. , and Ben H. Carpenter, "Collec-
tion and Analysis of Trace Organic Vapor Pollutants in Ambient
Atmospheres: Technique for Evaluating Concentration of Vapors by
Sorbent Media", Environmental Science and Technology. Vol 9,
pp 552-553, 1975.
9.4 Pellizzari, Edo D., "Development of Analytical Techniques for Mea-
suring Ambient Atmospheric Carcinogenic Vapors", EPA 600/2-75-076,
November 1975.
9.5 Pellizzari, Edo 0., 'The Measurement of Carcinogenic Vapors in
Ambient Atmospheres", EPA 600/7-77-055, June 1977.
9.6 Pellizzari, Edo D., "Analysis of Organic Air Pollutants by Gas
Chromatography and Mass Spectroscopy: Final Report", EPA
600/2-79-057, March 1979.
9.7 Pellizzari, Edo D., "Ambient Air Carcinogenic Vapors: Improved
Sampling and Analytical Techniques and Field Studies", EPA
60012-79-081, May 1979.
9.8 "Volatile Organic Air Pollutant Analysis - Permeation Tube Prepa-
ration and Calibration", NEIC, March 1983.
9.9 "Volatile Organic Air Pollutant Analysis - Tenax Trap Preparation
and GC Screening", NEIC, March 1983.
9.10 Pellizzari, Edo D., "Analytical Protocol: Personal Monitoring of
Vapor Phase Organic Compounds in Ambient Air (RTl)".
-------
THERMAL DESORPTION AND GC/MS ANALYSIS
-------
ENVIRONMENTAL PROTECTION AGENCY
NATIONAL ENFORCEMENT INVESTIGATION CENTER
Box 25227, Denver, Colorado 80225
Volatile Organic Air Pollutant Analysis
Using Tenax GC , Thermal Desorption and GC/MS
Analysis of Air Samples
March 1983
1.0 Introduction
This method describes the GC/MS analysis of the organic components
of air samples collected on Tenax traps. The analysis depends on
proper procedures for the preparation of Tenax traps, and for the
collection of samples. Those procedures are documented as the NEIC
methods "Volatile Organic Air Pollutant Analysis - Tenax Trap Pre-
paration and GC Screening" and "Volatile Organics Air Pollutant
Analysis - Sample Collection".
2.0 Summary of Method
Samples are collected by drawing a known volume of air through an
adsorbent resin which traps organic components. The resin traps
are analyzed by thermal desorption of the organics into a cryogenic
trap which is subsequently flash-heated to transfer the compounds
onto a GC column for GC/MS analysis.
3.0 Detection Limits
Detection limits for air samples depend on sample size, retention
characteristics on Tenax, and the individual sample matrix, among
other things, but can generally be expected to be in the range of 5
to 50 |jg/meter3 for a 25-liter sample size. Table 1 lists minimum
amounts of representative compounds detectable by this thermal de-
sorption/GC/MS procedure.
4.0 Limitations
4.1 Often, standard reference materials are not available and only
tentative identifications of unknowns can be achieved.
4.2 Because of the long time required to prepare accurate permea-
tion tube quantisation standards, a limited number of chemi-
cals can be quantitated.
4.3 Quantisation may not be possible if breakthrough occurs during
sampling. The sample traps are essentially short chromatograph-
ic columns. Retention of chemicals is dependent upon adsorption
® Registered Trademark of Enka N.V., The Netherlands; appears hereafter
without.
-------
C-2
Table 1
GC/MS RESPONSE FACTORS AND DETECTION LIMITS FOR SELECTED COMPOUNDS
Compound
Hexafluorobenzene (IS)
ds-bromoethane (IS)
Bromochlorome thane (SU)
Bromopentafluorobenzene (SU)
Benzene
Carbon Tetrachloride
Chlorobenzene
1,2,4-tn'chlorobenzene
1,2-dichloroethane
1,1,1-trichloroethane
1,1-dichloroethane
1,1,2-trichloroethane
1,1,2,2-tetrachloroethane
Bis(2-chloroethyl)ether
Chloroform
1,1-dichloroethane
1,2-transdichloroethylene
1,2-dichloropropane
Ethyl benzene
Bromoform
Tetrachloroethene
Toluene
Acetone
Hexane
Trichlorofluoromethane
n-Octane
2-Chloro toluene
Spiking
Level (ng)
2500
2500
380
1000
630
1000
400
100
900
270
620
150
160
75
1600
2000
4000
270
160
100
990
320
1000
510
1600
50
70
* ?£RSD — Percent Relative Standard nt*iria+'
*Q*»**** * WA h«^J J l^ CV^AatvAVC tj ^CtJlUQ J U W Y ^CK C*.
Average
Response
Factor
1.000
0.867
2.494
2.101
3.668
0.890
3.146
0.959
1.056
1.333
2.431
1.489
1.694
1.548
1.273
1.280
0.644
1.816
5.024
0.643
1.114
4.694
1.076
0.548
1.139
6.549
4.595
.. STD DEV _ .
Average
%RSD*
.
6.4
16.3
15.7
16.4
14.2
18.3
29.1
17.5
18.2
18.3
20.5
26.4
23.8
23.8
16.4
17.8
18.0
15.0
16.2
17.7
16.6
21.4
12.1
15.3
19.6
20.7
Jnn°s
I Vl/^)
Lower
Limit of
Detection
(ng)
_
200
300
200
400
100
NA**
300
100
200
50
200
20
800
600
1000
60
20
NA
300
100
700
90
500
NA
NA
** Not Available
-------
C-3
characteristics of the chemical and the resin system. Some fac-
tors influencing retention include: temperature, flow rate, vol-
ume of air sampled, vapor pressure of the chemical, chemical class,
presence of other chemicals, and batch-to-batch variation in the
sorbent resin. Due to the many variables involved, predicting
breakthrough is very difficult. Tandem tubes and breakthrough
spikes are used in the field to determine if breakthrough has
occurred. The sample size should be chosen to try to avoid break-
through.
The breakthrough volume is defined as the amount of air which
causes 50% of the collected chemical to be lost through the trap.
Table 1 of the procedure "Sample Collection" lists breakthrough
volumes compiled by Pellizzari in Ref. 12.8. The temperature was
the only factor varied.
5.0 Equipment and Reagents
5.1 Thermal Desorber. Nu-Tech 320 or equivalent desorber with the
following important features:
5.1.1 Capable of desorbing resin traps at 200-270° C.
5.1.2 Nickel cold trap able to be cooled to liquid nitrogen
temperature (-196°) and then rapidly heated to 150-250° C.
The upper temperature limit on the heated trap should
be reached in less than 2 minutes.
5.1.3 Heated transfer line between nickel trap and GC oven.
5.2 Gas Chromatograph. Varian 3700 or equivalent equipped with
linear temperature programmer, cryogenic cooling (liquid car-
bon dioxide or liquid nitrogen), and capillary column capa-
bility.
5.3 Capillary Column. 15M fused silica, DB-5 thick film (1 micron)
column. Other capillary columns yielding the desired chro-
matographic separations may be used.
5.4 Packed Column (optional). 6' x 2 mm I.D. glass column packed
with 60/80 mesh Carbo-pak C coated with 1% SP1000. Condition
overnight at 220° C with 20 m£/min flow rate. Other packed col-
umns may be used if chromatographic separe-ion is satisfactory
for the compounds of interest.
5.5 Mass Spectrometer capable of scanning from 35-350 a.m.u. in 1 sec-
ond or less and with open-split or direct interface for capillary.
5.6 Data System. Finnigan INCOS or equivalent capable of acquir-
ing and storing continuous repetitive mass spectra from the
mass spectrometer. The system must be able to match unknown
spectra to the EPA/NIH/MSDC mass spectral library and integrate
-------
C-4
ions for quantisation. Automated processing of the data is
desirable.
5.7 Culture Tubes. Pyrex glass screw cap tubes 25 mm x 150 mm. Pyrex
9825 or equivalent washed, dried, baked and fitted with Teflon-
backed butyl rubber septa as described in Reference 12.10.
5.8 Pyrex glass wool. Prepared as in Reference 12.10.
5.9 Calcium sulfate or sodium sulfate. Anhydrous, non-indicating
Baked at 220° C for at least 1 hour prior to use.
5.10 5/8" Teflon Rod with V drilled hole. Sized to hold Resin traps
securely against septa of culture tubes.
5.11 Resin traps as described in Reference 12.10.
6.0 Instrument Conditions
6.1 Desorber
6.1.1 Block Temperature 220° C - 270° C. 220° is the usual op-
erating temperature.
6.1.2 Desorber flow rate 15 m£/min helium.
6.1.3 Cold trap temperature -190° C (Reads approximately 160 on
Nu-tech Model 320 thermal desorber).
6.1.4 Cryogenic trap desorb temperature 180° C.
6.1.5 Transfer line temperature 180° C.
6.2 Gas Chromatograph (fused silica capillary column).
6.2.1 Carrier (helium) pressure 14 psig.
6.2.2 Initial temperature -20° C.
6.2.3 Initial hold time 2 minutes.
6.2.4 Program rate 5° C/min.
6.2.5 Final temperature 220° C.
6.2.6 Final hold time 15 min.
6.2.7 GC/MS separator oven 240° C.
6.2.8 Make-up gas flow (for open split) 30 m£/min.
-------
C-5
6.3 Gas Chromatograph (packed column)
6.3.1 Carrier (helium) flow rate 30 mA/min.
6.3.2 Initial temperature 60°.
6.3.3 Initial hold time 4 minutes.
6.3.4 Program rate 8° C/minute.
6.3.5 Final temperature 220° C.
6.3.6 Final hold time 15 minutes.
6.3.7 GC/MS separator oven 240° C.
6.4 Mass Spectrometer
6.4.1 Source temperature 220° C.
6.4.2 Mass Range 35 to 350. Other higher mass ranges may be
used.
6.4.3 Scan time 0.95 seconds up, 0.05 seconds hold at the
bottom of the scan.
6.4.4 Electron energy 70 eV.
6.4.5 Emission current 1.5 mA.
6.4.6 Line-of-sight inlet 230° C.
6.5 Sample Introduction Timing Sequence
6.5.1 Before sample introduction the desorber valve is in
the desorb mode, the cold trap is at liquid nitrogen
temperature and the sample desorbtion chamber is at
operating temperature. Time zero is the time at which
the GC temperature program is started.
6.5.2 t = -8 min 0 sec Insert Resin trap into thermal
desorber.
6.5.3 t = -0 min 15 sec Turn on Ionizer.
6.5.4 t = 0 min 0 sec Start GC oven program. Begin
mass spectral data acquisition.
Remove liquid nitrogen bath from
the cold trap.
6.5.5 t = 0 min 30 sec Begin heating nickel cold trap
and switch to inject mode on
desorber.
-------
C-6
6.5.6 t = 2 min 30 sec Turn off heat on nickel cold trap.
Return desorber valve to desorb
mode.
6.5.7 t = 65 min 0 sec Analysis complete.
7.0 Procedure
7.1 At least 16 hours before analysis of a sample, traps should be
dried. In an organic vapor free area, transfer Tenax resin
traps to cool, clean culture tubes containing approximately 10 g
anhydrous sodium sulfate or calcium sulfate. The desiccant should
be held in place with clean glass wool. Securely cap the culture
tube. This removes water adsorbed onto the Tenax during sampling.
The culture tubes should be stored in a desiccator with activated
charcoal adsorbent at room temperature. This step may be omitted
if the humidity was less than 20% during sampling. Other drying
techniques may be necessary with different resins which adsorb
water more strongly.
7.2 Set up instrument conditions as described in Section 6.
7.3 Spike the trap with surrogate standards. See Section 9.
7.4 Spike the trap with 20 u£ Internal Standard. See Section 9.
7.5 Begin analysis. Use the procedure described in 6.5.
7.6 After analysis is complete, output, and evaluate data.
8.0 Storage and Holding Times
8.1 Samples prior to the drying step are stored in a dark organic-
free area and held at -20° C or less. (Ref.12.12).
8.2 The samples should be analyzed within 4 weeks of collection.
(Ref.12.12).
9.0 Standards
9.1 Internal Standard (Static)
9.1.1 To a clean 300-m2 glass gas sampling bulb prepurged with
inert gas add 9.3 u£ hexafluorobenzene and 10.3 pfi ds-
bromoethane.
9.1.2 Maintain the bulb in a water bath at 30° +0.1° C. The
bulb should be in the water bath for at least 1 hour
before sampling.
9.1.3 Withdraw 20 u£ aliquots using a gas-tight syringe. Slowly
inject into the center of the sample or standard trap.
This injects 1000 ng of each compound. See calculations
9.1.3.1 and 9.1.3.2.
-------
C-7
9.1.3.1 Hexafluorobenzene has a density of 1.607 q/m£
or 1.607 mg/u£.
-3
x onn £o X 20 US, X ——;
_3
1.0 x 10 mg = 1000 ng
9.1.3.2 Bromoethane has a density of 1.4606 g/m£.
1-4606 mg 10.3 u£ v 20 \tl x 10"3 m£ _
u£ 300 m£ x ji£
-3
1.0 x 10 mg = 1000 ng
9.2 Surrogate Standards (Dynamic)
9.2.1 Prepare permeation tubes of bromochloromethane and bromo-
pentafluorobenzene using the procedure outlined in Ref-
erence 12.9.
9.2.2 Spike surrogates onto sample and standard tubes immediately
prior to analysis. This is done by connecting the Tenax
trap to the gas exit flow from the permeation chamber.
Time to the nearest second the length of time that the trap
has flow from the permeation chamber going thru it. This
time (min) multiplied by the permeation rate (ng/min) gives
the amount (ng) of each compound on the Tenax trap. The
usual time is 3 min.
9.3 Mass Intensity Standard (Static)
9.3.1 A static standard of octafluorotoluene is prepared the same
as the internal standard (Section 9.1). Use a 250 m£ gas-
sampling bulb and 7.5 u£ octafluorotoluene. Injecting 20 u£
of the gas from the bulb places 1000 ng of octafluorotolu-
ene on the trap. See calculation 9.3.2.
9.3.2 Octafluorotoluene has a density of 1.663 mg/(j£
.3
1.0 x 10 mg = 1000 ng
9.4 Quantisation Standards (Dynamic)
9.4.1 Prepare permeation tubes of compounds to be quantified
as in Sec. 9.2.1.
-------
C-8
9.4.2 Spike standards onto blank resin traps for analysis using
the procedure described in 9.2.2. Add surrogate stand-
ard and internal standard before beginning analysis.
9.4.3 The amount of material spiked onto the traps is controlled
by how long the outflow from the permeation tubes is al-
lowed to flow through the resin trap. This is a linear,
reproducible relationship to times at least down to I min.
Figure 1 shows the linearity typically achieved between
FID area response and sampling time.
10.0 Quantification
10.1 Chemicals identified from their mass spectra may be quantified
by comparison of the responses of the unknowns to the responses
of known amounts of pure standards. The preferred method is the
use of relative responses and internal standards.
10.2 Calibration is performed by analyzing a mixture of chemicals at
known concentrations containing an internal standard (hexafluoro-
benzene for example) added at a fixed concentration. The instru-
ment responses for selected ions are measured and compared for
each component. A response factor is calculated for each com-
ponent by:
Resp. fact. = Area x Ref. Amt/(Ref. area x amt.) (Eq. 1).
Where: (Resp. fact. = response factor)
Area = area of ion in component
Ref. Area = area of ion in internal standard
Ref. Amt. = amount of internal standard added
Amt. = amount of component
10.3 Quantification of identified chemicals is done by determining
the areas of the appropriate ions and calculating the amount
from equation 1 re-arranged:
Amt = Area x Ref. amt/(Ref. area x amt.) (Eq. 2).
11.0 Quality Control
11.1 A laboratory blank spiked with the surrogate and internal
standards is run daily before the analysis of samples. If
there is a response change between the surrogate and internal
standard compounds the cause is investigated and corrected
before analysis of samples. There should not be peaks in the
blank which might interfere in the analysis. Tenax traps are
easily contaminated by solvents and other volatile organics.
11.2 Octafluorotoluene is run daily before the analysis of samples
as a check of mass-intensity calibration. (May be combined with
laboratory blank.) Mass-intensity criteria for 1000 ng of octa-
fluorotoluene are given on page C-10.
-------
Absolute
Area
(GC/FID)
Response
i
EFfl-HEIC
L in v a r- 1 e ar t 5 q u a r e = fit
53253
_ IT
5140.59
coefficient of determination IP 2': .?
3 4 5 t ?
Time permeation tube outflow is sampled (min).
•? 10
nEHVEP, CULOPHDO
Figure 1. Typical linearity curve (area vs time)
o
10
-------
C-10
ID/I. % Relative Abundance
69 30-60
79 5-15
93 10-30
117 40-65
167 10-25
186 55-85
217 100
236 60-85
11.3 Calibration of response factors is done daily at a mid-range
concentration. Linearity is determined at least once during
a set of analyses.
11.4 The response of the surrogate standards is monitored relative
to the internal standard. Any significant deviation is inves-
tigated and proper corrective action is taken before other
samples are run. It is very important to monitor this and
correct problems immediately, since it is not possible to re-
run a sample.
11.5 Air sample traps are easily contaminated. One sample trap per
shipping container taken to the field will be tagged in the
field as a field blank and returned with the samples to the
laboratory for analysis.
11.6 Field spikes are air sample traps which have been spiked in
the laboratory, taken to the field, tagged and returned to
the laboratory. These spikes indicate sample deterioration
due to shipping, handling and storage. Three field spikes
or a number equal to 10% of the field sampling points, which-
ever is greater, will be analyzed.
11.7 Breakthrough spikes are air sample traps which have been spiked
in the laboratory and subsequently sampled over in the field
with a clean sample trap as the back-up in the tandem sampling
arrangement [see Figure 2]. This spike indicates whether break-
through of spiked compounds has occurred under field conditions.
The breakthrough spike is done in duplicate at the upwind sampling
point.
11.8 Table 1 lists typical response factors, percent relative stand-
ard deviations, and lower limits of detection.
11.9 Figure 3 is a typical chromatogram obtained using the 15 M DBS
capillary column.
-------
FLOW
' _ .'i ir
/
i
[ i
vj 1
' 1 ' V >*£"• '..••:'-,. . , •'•{! i
H h- irKi: ':.•'•••<.' -W f i
J_M ' m -, ;•,••:...:••• ; ;•:; ji j
""^ -1 — •-^Xv^ — ..., , ''•'": \ U_ i
i i i
\ '• r~
% Teflon Union Air Sample Trao 'v~ '•
!
!
•
1 ^
_ , »
I
j
i
: J
\
*'"*,-: ' * v\
• ' : '. * ' " - .-
'• • x " ' ',
CDT l/cn n •. . TT._
5/8" Brass Union with
Teflon ferrules
Sample
Figure 2. Breakthrough Spike
o
-------
c-
Table 2
IDENTITIES AND CONCENTRATION OF PEAKS IN FIGURE 3
Peak #
I
2
3
4
5
6
7
8
9
10
11
12
13
14
14
15
16
17
18
19
20
21
22
23
24
25
26
HC
Compound
Carbon dioxide
Tr ichl orofl uoromethane
Acetone
trans- 1,2-Dichloroethene
Dichloromethane
1,1-Dichloroethene
1 , 1-Di chl oroethane
Hexane
Hexaf 1 uorobenzene
Tetrahydrofuran
1,2-Dichloroethane
1,1,1-Tri chl oroethane
Octafluorotoluene
Benzene (coelutes)
Methylcyclopentene (coelutes)
Carbon tetrachloride
Cyclohexane
Methyl hexane
1,2-Dichloropropane
1,1, 2-Tri chl oroethane
Toluene
Tetrachloroethene
Octane
Chlorobenzene
Ethylbenzene
o-Chlorotoluene
1, 2, 4-Tri Chlorobenzene
Unidentified alkane/alkene
Amount (ng)
***
600
980
4000
2
***
2000
620
510
2500
2
***
870
270
2
***
610
2
***
1000
***
2
***
260
150
380
950
2
***
380
150
2
***
70
***
1 Carbon dixode is an artifact of the analysis. Identification
is based on spectra and that it is an unretained compound.
2 New permeation tube; not calibrated.
-------
DATA:
CALI:
1C0.0
R!C
C4/30/82 12:36:69
SAMPLE: 3 mil ARRAY 1. 5 HIM ARRAY 2, 5Q UL HFB
RAtiOE: G 1,1309 LABEL: M 0, 4.€ GUAN: A 0, 1.0 BASE: U 23,
fv 9 V
D*r
AB 121.266 {(1
CC4334 i)4
SCAiJS
1 TO 13^
256
89
i
i
i
t
i
i
!
i
3!
t
3
! i !
i • i
1
| 1
i
~J
\Xw,
5'
ti
B
s
II fl
II
1 ii A 1
I ii
In
5
ii ii
i ii
I i
l|
i ' i
)
i
i
jt
jj . i
i
.1 i;
! i
L ^
«
I
t
L
t
1
A
£j
/O
i
•1 '
i
ij
1
ii
1
1 -
1 ,
L« t
,,
"' 1'
1
V]
a
c_ -/ 1 0 t' a.
, 1244 '2t
1392
1 '^1
| | . l I | r 1 . w
58Q 1680 1500 $a?J
8:20 16:4!li 25:63 1M£
-------
C-14
12.0 References
12.1 Bertsch, Wolfgang, Chang, Ray C. and Albert Zlatkis, "The
Determination of Organic Volatiles in Air Pollution Studies:
Characterization of Profiles", Journal of Chromatographlc
Science. Vol. 12, pp 175-182, April 1974.
12.2 Pellizzari, Edo D., "Development of Method for Carcinogenic
Vapor Analysis in Ambient Atmospheres", EPA-650/2-74-121, July
1974.
12.3 Pellizzari, Edo D., Bunch, John E., and Ben H. Carpenter,
"Collection and Analysis of Trace Organic Vapor Pollutants in
Ambient Atmospheres: Technique for Evaluating Concentration
of Vapors by Sorbent Media", Environmental Science and Tech-
nology, Vol 9, pp 552-553, 1975.
12.4 Ibid, pp 556-560.
12.5 Pellizzari, Edo D., "Development of Analytical Techniques for
Measuring Ambient Atmospheric Carcinogenic Vapors", EPA 600/2-
75-076, November 1975.
12.6 Pellizzari, Edo D., "The Measurement of Carcinogenic Vapors in
Ambient Atmospheres", EPA 600/7-77-055, June 1977.
12.7 Pellizzari, Edo D., "Analysis of Organic Air Pollutants by Gas
Chromatography and Mass Spectroscopy: Final Report", EPA
600/2-79-057, March 1979.
12.8 Pellizzari, Edo D., "Ambient Air Carcinogenic Vapors: Improved
Sampling and Analytical. Techniques and Field Studies", EPA
60012-79-081, May 1979.
12.9 "Volatiie Organic Air Pollutant Analysis - Permeation Tube
Preparation and Calibration", NEIC, July 1982.
12.10 "Volatile Organic Air PoJ-Zutant Analysis - Tenax Sample Trap Pre-
paration and Screening", NEIC, March 1983.
12.11 "Vo.Zati.Ze Organic Air PoJJutant Analysis - Sample Collection",
NEIC, March 1983.
12.12 Pellizzari, Edo D. , "AnaJyticaJ Protocol: Personal Monitoring
of Vapor Phase Organic Compounds in Ambient Air (RTI)".
-------
PERMEATION TUBE PREPARATION AND CALIBRATION
-------
Volatile Organic Air Pollutant Analysis
Permeation Tube Preoaration and Calibration
NEIC July 1982
1.0 Introduction
1.1 Primary standards are necessary to quantitatively analyze or-
ganic air pollutants. Standards are prepared by loading sam-
ling traps with known amounts of chemicals from permeation
tubes. This is accomplished by passing the effluent gas stream
from a chamber containing calibrated permeation tubes onto
sampling trans identical to those used in the field.
1.2 Permeation tubes are generally Teflon tubes containing a pure
chemical, plugged to form gas tight seals at each end The
organic chemical then oermeates through the Teflon tubing at a
rate dependent upon the temoerature and length of the tube
The rates are also dependent upon the chemical and vary over
serveral orders of magnitude. The permeation rate is deter-
mined gravametrically.
2.0 Safety
2.1 Many of the compounds of interest in air analysis are toxic
and/or carcinogenic. They are also volatile which increases
the potential for exposure to the compounds. Persons prepar-
ing permeation tube standards must be aware of the hazards of
the individual compounds handled, and use appropriate safety
precautions. All permeation tubes should be prepared in a
hood, in extreme cases other precautions may be necessary.
2.2 The permeation tubes slowly emit the standard compounds. The
erriuent from the permeation tubes should be routed through a
charcoa trap and into a fume hood. When weighing the tubes,
the analyst should keep handling to a minimum and avoid breath-
- ing fumes from the tubes.
3.0 Tube Materials
and propyiene
03blwail Tetraf1uoroethylene Po1ymer tubing 1/4"
3.3 Teflon Rod. Del-F rod 3/16" o.d.
3-4
3.5 Crimp tool. Nicopress 31-CJ tool to crimp to 1/4" o.d.
-------
D-2
4.0 Permeation Chamoer
\ Recirculatin9 heating/cooling bath capable
a temperature of 30 ± 0.1 deg. C? ^°u\e
4.2 Water Jacket. Glass water jacketed tube with Teflon screw in
Plugs at ends. Typical dimensions 3 cm i.d. x 20 cm
4'3 dlTvPr? t6lA ,nta1?l6SS Steel CflPi11ary ^be capable of
delivering 40-60 cc/min of N2 from a 30 psig supply.
4.4 Switching valve. Teflon 2-way solenoid valve.
coa with
5.0 Tube Preparation
Of the .
carcin°9enic, toxic or Haz-
C" FE" P'U9 in the °"e" e"" •"« erl^. a band
5.5 Visually inspect the tube for signs of leaking.
5'6 COTs?aSet™Se ''J the,Derm«"<'n chamber and maintain at a
n .. 2
6.0 Tube Calibration
ml/min1!! ^L^f" at C°,nStant temP^ature with about 40-60
r -a
of the balance calibration and/or repair
-------
D-3
6.5 Monitor the weight changes of calibrated tubes every 4-6 weeks
for the life of the tube.
7.0 Calculations
7.1 Average Rate
7.1.1 Rate:
Rate = weight change (ng)
minutes between weighings
7.1.2 Typically weight change is less than 10 mg and time
between weighings is 20000-40000 minutes (2-4 weeks).
7.1.3 Average the last 5 stable rates.
7.2 Regression Rate
7.2.1 Tabulate the weight of the tube vs. time from the point
the rate stabilized, or for the last 20 stable weighings.
7.2.2 Using least squares techniques, calculate the slope of
the weight vs. time data. The slope is the permeation
rate. Also calculate the correlation coefficient as an
indication of the stability of the calibration data.
7.3 Percent Relative Standard Deviation (%RSD)
7.3.1 SRSD = Standard Deviation
Average x IUU/J
7.3.2 The percent relative standard deviation should be 10%
or less for stable tubes.
8.0 Other Permeation Devices
-u Pfrmeation ^vices may be used such
, permeation bags or drilled rod .
9.0 References
0n' C' Ana1.ytical Chemistry. 39,
' P" Wood> R-> Chemistry and Industry. Dec. 28, 1968,
3. Scaringelli Frank P., O'Keefe, Andrew E. , Rosenberg, Ethan
Be11' John p-» Analytical Chemistry. Vol. 42, 871 (1970)
4. Analytical Chemistry. 49, 1278 (1977).
vapors '
-------
D-4
TABLE I
Compound
Acetone
Acrylonitrile
Benzene
Bis(2-chloroethyl)ether
Bromochloromethane
Bromoethane-dr
Bromoform
Bromopentaf1uorobenzene
Carbon tetrachloride
Chlorobenzene
Chloroform
1,4-Dichlorobutane
1,1-Dichloroethane
1,2-Dichloroethane
1,1-Dichloroethene
1,2-trans-Dichloroethene
Dichloromethane
1,2-Dichloropropane
Ethyl benzene
Hexane
Tetrachloroethene
Toluene
1,2,4-TriChlorobenzene
1,1,1-Trichloroethane
1,1,2-Trichloroethane
Trichloroethene
Trichlorofluoromethane
Cas ^
67-64-1
107-13-1
71-43-2
111-44-4
74-97-5
Not Available
75-25-2
344-04-7
56-23-5
108-90-7
67-66-3
110-56-5
75-34-3
107-06-2
75-35-4
156-60-5
75-09-2
78-87-5
100-41-4
110-54-3
127-18-1
108-88-3
120-82-1
71-55-6
79-00-5
79-01-6
75-69-4
Tube Type
TFE
TFE
TFE
TFE
FEP
TFE
TFE
FEP
TFE
TFE
TFE
TFE
TFE
TFE
FEP
FEP
TFE
TFE
TFE
TFE
TFE
TFE
TFE
TFE
• - TFE
TFE
FEP
Rate/length*
ng/min/cm
64
88
40
2
25
170
4
71
20
25
110
4
40
58
64
150
320
8
10
37
62
26
4
9
9
200
100
*This should only be used as a guide. Each batch of tubing will be different.
-------
TENAX-G(T SAMPLE TRAP PREPARATION AND SCREENING
-------
ENVIRONMENTAL PROTECTION AGENCY
NATIONAL ENFORCEMENT INVESTIGATIONS CENTER
Box 25227 Denver Federal Center
Denver, Colorado 80225
e Organic Air Pollutant Analysis
Tenax Sample Trap Preparation and Screening
March 1983
1.0 Introduction
1.1 Sampling for organics in air is performed by drawing air
through a glass tube packed with the porous polymer resin
Tenax-GC. The traps and resin must be thoroughly cleaned
before use to minimize the trap background. Clean traps
ready for field use must also be carefully packed in clean
glass tubes to avoid contamination during handling.
2.0 Materials
2.1 Glass sampling traps. Pyrex glass traps constructed as shown
in Figure 1.
2.2 Resin. Tenax-GC, 35/60 mesh.
2.3 Glass wool.
2.4 Culture tubes. Pyrex glass screw cap tubes 25 mm x 150 mm.
Pyrex 9825-20X or equivalent.
®
2.5 Teflon -backed silicone septa. Pierce 12722 or equivalent.
2.6 Bakelite screw caps to fit culture tubes. Pierce 13219 or
equivalent.
2.7 Desiccator. Glass desiccator with activated charcoal
adsorbant.
2.8 Virgin gallon paint cans with pressure fit lids.
2.9 Polyurethane foam. Used for packaging of culture tubes to
avoid breakage.
2.10 Teflon spacers made from 5/8" Teflon rod with V drilled hole.
Sized to hold traps securely against septa of culture tube.
3.0 Resin Preparation
3.1 Extract new and used Tenax-GC with methanol followed by pen-
tane in a soxhlet extractor. Extract with each solvent 18
hours.
3.2 Dry the resin under vacuum for a least 6 hours.
® Registered trademark; appears hereafter without the ®.
-------
E-2
3.3 Sieve the dried resin to the 35/60 mesh particle size range.
3.4 Seal the cleaned and sieved resin in a glass jar capped with
a Teflon liner. Store in a desiccator containing activated
carbon.
4.0 Trap and Container Cleanup
4.1 Wash new and used glass sampling traps and culture tubes with
lab detergent and hot tap water. Rinse at least three times
with organics-free water. Rinse with methanol and let air
dry.
4.2 Bake the cleaned tubes in an oven at 220° C for at least 1 hour.
Remove from the oven and store in a desiccator containing acti-
vated carbon.
4.3 Place glass wool in boiling water for 15 minutes. Remove glass
wool from boiling water, drain, and place on absorbent paper
towels to remove excess water. Rinse by dipping in acetone and
placing in a soxhlet extractor. Soxhlet extract glass wool with
methanol at least 8 hours, air dry and bake in an oven at 200° C
for at least 1 hour. Spread the glass wool out in a baking dish
and place in a muffle furnace at 450° C for 1 hour. Remove from
the oven, cool and store in a desiccator containing activated
charcoal.
4.4 Bake Teflon-backed septa in an oven at 80° C for 30 minutes.
Remove from the oven and store in a desiccator containing
activated charcoal.
4.5 Bake paint cans in oven at 100° C for 1 hour.
4.6 Sohxlet extract foam in methanol for 18 hours. Drain foam and
dry under vacuum for 24 hours.
5.0 Trap Preparation
5.1 Pack about a 1-cm plug of glass wool into the trap and weigh
it. Add about 7 cm of cleaned Tenax-GC. Lightly tap the trap
on the bench to pack the resin. Weigh the trap and adjust the
weight of the Tenax to 0.8 ± 0.02 gram by adding or removing
small amounts of the polymer. Add another 1-cm glass wool plug
to hold the resin in place.
5.2 Condition each trap by placing it in the bakeout manifold.
The manifold is an apparatus which provides flow to six traps
simultaneously and can be placed in a GC oven. Connect a k"
to 1/16" reducing filter on the top of each trap and turn on
the helium flow. Measure the flow on each trap to assure it
is 20-30 m£/minute. Set the GC for bakeout conditions of
270° for 30 minutes. Set the oven initial temperature for
cooldown to 35°. After cooldown, turn off the helium flow
-------
E-3
5.3
and place the traps in culture tubes with Teflon spacers.
Cap the culture tube using a Teflon-lined septum [Figure 1].
Record the bakeout batch number on each tube label.
Tubes are stored in desiccators containing activated charcoal
until ready for shipment.
6.0 Screening
6.1 Screen at least one trap from each bakeout batch on a thermal
desorber system connected to an FID-GC. Thermal desorber con-
ditions are the same as those used for GC/MS analysis. Elec-
trometer setting is 10-12 amps and attenuation 32. Label each
chromatogram with a GC run number and store the chromatograms
by ascending number for future reference. Figure 2 shows a
typical acceptable screening chromatogram.
GC Conditions
Desorber Conditions
Column:
Phase: DB-5 or Equivalent
Length: 15 meter
Injection Temp: 110° C
Detection Temp: 260° C
Oven: Initial Temp. 35° C
Initial Time: 3 min.
Prog. Rate: 15 deg/min.
Final Tern: 230° C
Final Hold: 15 min.
Cap Pressure: 10 PSI
Temperatures:
Line: 150° C
Block: 220° C
Trap: 180° C
Time: 0:00 min. Desorb at
LN2 Temp, on the
trap.
8 min. - Start GC and
Integrator and Chart
and Lower LN2 bath.
8 min-30 sec. - Valve
to Inject trap switch
to heat.
10 min-30 sec. - Valve
to desorb trap switch
to cool.
6.2 Mark the label on the culture tube with the GC run number of
the screening chromatogram. If the trap is clean, the batch
is acceptable for use.
7.0 Packaging for Shipment
7.1
Place culture tubes containing Tenax traps in special foam inserts
in 1-gallon paint cans. Insert a charcoal-filled pouch around the
top of the foam insert and pour a few drops of liquid nitrogen
onto the foam. After a few seconds place the lid securely on the
paint can. If excessive bulging of the can occurs, immediately
remove the lid to release the nitrogen gas. Secure can lids
with clips. Place a label on the can lid indicating the number
and types of traps inside. Indicate the type of adsorbant and
whether spikes or blanks are present.
-------
Attachment to: Sample Trap Preparation
E-4
& ^
4
I 5 fr)m
_£_.
^ /A
JvrV_
Figure 1. Sampling trap and culture tube holder design.
-------
Figure 2. Typical FID screening of a Tenax trap after bakeout at 270°C. This result is
acceptable for sample traps to be used for GC/MS analysis. Because of the dif-
ference in sensitivity between FID and MS, only the C02 peak would appear in an
MS analyis. Chromatographic conditions are given in Sec. 6.1.
Ill
-ll-rn+H-r;
I ! ' ; i : ; '
._].„.
I I '
~rt"TTTJT
I
: !|"i"! ii'M ! J;'M '! i" il!"::"|
. : i .:•:.!. . . ;
. 1 , i ' , • •
1 i : • i
I f • , • 1 : • •
: : ; i i ! : i ! : • : , i ; - !
L. ...„! - ... : .
; i ' |
'. I ' i
1
r. • ;
i
1
; '• '
I i I :
— rv-i itpi'-r-'
^^.__3m^I I I ' *• (
I
I !_.. _j._. ._L.L.LLill
H-
-H
i i
i i '
1 • . ; . ' '.
• i :
1 - i '
1' : • i
1 i I i
MIL
'
i : .
> J,
! . ir
i : I '
I : : i ! : . : i
1
!-..,.,. . . ...
'
-L^LL
, i i
-'- f—
nii«-3-
LlP-iFAR
! H-l
I
! I
: : i •
_l_ _ I
"i r —
j : i ; | i
-1 •'
3
-------
E-6
8.0 References
8.1 "Selection and Evaluation of Sorbant Resins for the Collection
of Organic Compounds", EPA-600/7-77044, April 1977.
8.2 "Development of Method for Carcinogenic Vapor Analysis in Am-
bient Atmospheres", EPA-650/2-74-121, July 1974.
------- |