METHOD TO-3
REVISION 1.0
April, 1984
METHOD FOR THE DETERMINATION OF VOLATILE ORGANIC COMPOUNDS
IN AMBIENT AIR USING CRYOGENIC PRECONCENTRATION TECHNIQUES
AND GAS CHROMATOGRAPHY WITH FLAME IONIZATION AND
ELECTRON CAPTURE DETECTION
1. Scope
1.1 This document describes a method for the determination of
highly volatile compounds having boiling points in the
range of -10 to 200°C.
1.2 The methodology detailed in this document is currently
employed by numerous laboratories (1-4;8-11).
Modifications to this methodology should be accompanied
by appropriate documentation of the validity and
reliability of these changes.
2. Applicable Documents
2.1 ASTM Standards
D1356 Definition of Terms Related to Atmospheric Sampling
and Analysis
E 355 Recommended Practice for Gas Chromatography Terms
and Relationships
2.2 Other Documents
Ambient Air Studies (1-4).
U. S. EPA Technical Assistance Document (5).
3. Summary of Method
3.1 Ambient air analyses are performed as follows. A
collection trap, as illustrated in Figure 1, is submerged
in either liquid oxygen or argon. Liquid argon is highly
recommended for use because of the safety hazard
associated with liquid oxygen. With the sampling valve
in the fill position an air sample is then admitted into
the trap by a volume measuring apparatus. In the
meantime, the column oven is cooled to a sub-ambient
temperature (-50°C). Once sample collection is
completed, the valve is switched so that the carrier gas
sweeps the contents of the trap onto the head of the
cooled GC column. Simultaneously, the liquid cryogen is
-------
removed and the trap is heated to assist the sample
transfer process. The GC column is temperature
programmed and the component peaks eluting from the
columns are identified and quantified using flame
ionization and/or electron capture detection. Alternate
detectors (e.g., photoionization) can be used as
appropriate. An automated system incorporating these
various operations as well as the data processing
function has been described in the literature (8,9).
3.2 Due to the complexity of ambient air samples, high
resolution (capillary column) GC techniques are
recommended. However, when highly selective detectors
(such as the electron capture detector) are employed,
packed column technology without cryogenic temperature
programming can be effectively utilized in some cases.
Significance
4.1 Volatile organic compounds are emitted into the
atmosphere from a variety of sources including industrial
and commercial facilities, hazardous waste storage
facilities, etc. Many of these compounds are toxic,
hence knowledge of the levels of such materials in the
ambient atmosphere is required in order to determine
human health impacts.
4.2 Because these organic species are present at ppb levels
or below, some means of sample preconcentration is
necessary in order to acquire sufficient material for
identification and quantification. The two primary
preconcentration techniques are cryogenic collection and
the use of solid adsorbents. The method described herein
involves the former technique.
Definitions
Definitions used in this document and any user prepared SOPs
should be consistent with ASTM D1356(6). All abbreviations
and symbols are defined within this document at the point of
use.
Interferences/Limitations
6.1 Compounds having similar GC retention times will
interfere in the method. Replacing the flame ionization
detector with more selective detection systems will help
to minimize these interferences. Chlorinated species, in
particular, should be determined using the electron
capture detector to avoid interference from volatile
hydrocarbons.
-------
6.2 An important limitation of the technique is the
condensation of moisture in the collection trap. The
possibility of ice plugging the trap and stopping the
flow is of concern, and water subsequently transferred to
the capillary column may also result in flow stoppage and
cause deleterious effects to certain column materials.
Use of permaselective Nafion® tubing in-line before the
cryogenic trap avoids this problem; however, the material
must be used with caution because of possible losses of
certain compounds. Another potential problem is
contamination from the Nafion® tubing. The user should
consult the literature (7-12) for details on the use of
permeation-type driers.
7. Apparatus
7.1 Gas chromatograph/Flame Ionization/Electron Capture
Detection System - must be capable of subambient
temperature programming. A recent publication (8)
describes an automated GC system in which the cryogenic
sampling and analysis features are combined. This system
allows simultaneous flame ionization and electron capture
detection.
7.2 Six-port sampling valve - modified to accept a sample
collection trap (Figure 1).
7.3 Collection trap - 20 cm x 0.2 cm I.D. stainless steel
tubing packed with 60/80 mesh silanized glass beads and
sealed with glass wool. For the manual system (Section
9.2) the trap is externally wrapped with 28 gauge (duplex
and fiberglass insulated) type "K" thermocouple wire.
This wire, beaded at one end, is connected to a powerstat
during the heating cycle. A thermocouple is also
attached to the trap as shown in Figure 1.
7.4 Powerstat - for heating trap.
7.5 Temperature readout device - for measuring trap
temperature during heating cycle.
7.6 Glass dewar flask - for holding cryogen.
7.7 Sample volume measuring apparatus - capable of accurately
and precisely measuring a total sample volume up to 500
cc at sampling rates between 10 and 200 cc/minute. See
Section 9.
Stopwatch.
-------
7.9 Dilution container for standards preparation - glass
flasks or Teflon (Tedlar) bags, .002 inch film thickness
(see Figure 2) .
7.10 Liquid microliter syringes - 5-50 //I for injecting liquid
standards into dilution container.
7.11 Volumetric flasks - various sizes, 1-10 mL.
7.12 GC column - Hewlett Packard 50 meter methyl silicone
cross-linked fused silica column (.3 mm I.D., thick film)
or equivalent.
7.13 Mass flow controller - 10-200 mL/minute flow control
range.
7.14 Permeation drier - PermaPure® - Model MD-125F, or
equivalent. Alternate designs described in the
literature (7-12) may also be acceptable.
8. Reagents and Materials
8.1 Glass beads - 60/80 mesh, silanized.
8.2 Glasswool - silanized.
8.3 Helium - zero grade compressed gas, 99.9999%.
8.4 Hydrogen - zero grade compressed gas, 99.9999%.
8.5 Air - zero grade compressed gas.
8.6 Liquid argon (or liquid oxygen).
8.7 Liquid nitrogen.
8.8 SRM 1805 - benzene in nitrogen standard. Available from
the National Bureau of Standards. Additional such
standards will become available in the future.
8.9 Chemical standards - neat compounds of interest, highest
purity available.
9. Sampling and Analysis Apparatus
Two systems are described below which allow collection of an
accurately known volume of air (100-1000 mL) onto a
cryogenically cooled trap. The first system (Section 9.1) is
an automated device described in the literature (8,9) . The
second system (Section 9.2) is a manual device, also described
in the literature(2) .
-------
9.1 The automated sampling and analysis system is shown in
Figure 3. This system is composed of an automated GC
system (Hewlett Packard Model 5880A, Level 4, or
equivalent) and a sample collection system (Nutech Model
320-01, or equivalent) . The overall system is described
in the literature (8) .
9.1.1 The electronic console of the sampling unit
controls the mechanical operation of the six-
port valve and cryogenic trapping components
as well as the temperatures in each of the
three zones (sample trap, transfer line, and
valve).
9.1.2 The valve (six-port air activated, Seiscor
Model 8 or equivalent) and transfer line are
constantly maintained at 120°C. During sample
collection the trap temperature is maintained
at -160 ± 5°C by a flow of liquid nitrogen
controlled by a solenoid valve. A cylindrical
250 with heater, held in direct contact with
the trap, is used to heat the trap to 120°C in
60 seconds or less during the sample
desorption step. The construction of the
sample trap is described in Section 7.3.
9.1.3 The sample flow is controlled by a pump/mass
flow controller assembly, as shown in Figure
3. A sample flow of 10-100 mL/minute is
generally employed, depending on the desired
sampling period. A total volume of 100-1000
mL is commonly collected.
9.1.4 In many situations a permaselective drier
(e.g., Nafion®) may be required to remove
moisture from the sample. Such a device is
installed at the sample inlet. Two
configurations for such devices are available.
The first configuration is the tube and shell
type in which the sample flow tube is
surrounded by an outer shell through which a
countercurrent flow of clean, dry air is
maintained. The dry air stream must be free
from contaminants and its flow rate should be
3-4 times greater than the sample flow to
achieve effective drying. A second
configuration (7) involves placing a drying
agent, e.g., magnesium carbonate, on the
outside of the sample flow tube. This
approach eliminates the need for a source of
clean air in the field. However, contamination
from the drying agent can be a problem.
-------
9.2 The manual sampling consists of the sample volume
measuring apparatus shown in Figure 4 connected to the
cryogenic trap/GC assembly shown in Figure 1. The
operation of this assembly is described below.
9.2.1 Pump-Down Position
The purpose of the pump-down mode of operation
is to evacuate the ballast tank in preparation
for collecting a sample as illustrated in
Figure 4. (While in this position, helium can
also be utilized to backflush the sample line,
trap, etc. However, this cleaning procedure
is not normally needed during most sampling
operations). The pump used for evacuating the
system should be capable of attaining 200 torr
pressure.
9.2.2 Volume Measuring Position
Once the system has been sufficiently
evacuated, the 4-way ball valve is switched to
prepare for sample collection. The 3-position
valve is used to initiate sample flow while
the needle valve controls the rate of flow.
9.2.3 Sample Volume Calculation
The volume of air that has passed through the
collection trap corresponds to a known change
in pressure within the ballast tank (as
measure by the Wallace Tiernan gauge).
Knowing the volume, pressure change, and
temperature of the system, the ideal gas law
can be used to calculate the number of moles
of air sampled. On a volume basis, this
converts to the following equation:
A P 298
V = x
s 760 Ta+2 7 3
where
Vs = Volume sampled at 7 60 mm Hg pressure and
25°C.
AP = Change in pressure within the ballast
tank, mm of Hg.
V = Volume of ballast tank and gauge.
Ta = Temperature of ballast tank, °C.
The internal volume of the ballast tank and gauge can be
determined either by H20 displacement or by injecting
calibrated volumes of air into the system using large
volume syringes, etc.
-------
10. Sampling and Analysis Procedure - Manual Device
10.1 This procedure assumes the use of the manual sampling
system described in Section 9.2.
10.2 Prior to sample collection, the entire assembly should be
leak-checked. This task is accomplished by sealing the
sampling inlet line, pumping the unit down and placing
the unit in the flow measuring mode of operation. An
initial reading on the absolute pressure gauge is taken
and rechecked after 10 minutes. No apparent change
should be detected.
10.3 Preparation for sample collection is carried out by
switching the 6-port valve to the "fill" position and
connecting the heated sample line to the sample source.
Meanwhile the collection trap is heated to 150°C (or
other appropriate temperature). The volume measuring
apparatus is pumped-down and switched to the flow
measuring mode. The 3-position valve is opened and a
known volume of sample is then passed through the heated
sample line and trap to purge the system.
10.4 After the system purge is completed, the 3-position valve
is closed and the corresponding gauge pressure is
recorded. The collection trap is then immersed into a
dewar of liquid argon (or liquid oxygen) and the 3-
position valve is temporarily opened to draw in a known
volume of air, i.e. a change in pressure corresponds to
a specific volume of air (see Section 9) . Liquid
nitrogen cannot be used as the cryogen since it will also
condense oxygen from the air. Liquid oxygen represents
a potential fire hazard and should not be employed unless
absolutely necessary.
10.5 After sample collection is completed, the 6-port valve is
switched to the inject position, the dewar is removed and
the trap is heated to 150°C to transfer the sample
components to the head of the GC column which is
initially maintained at -50°C. Temperature programming
is initiated to elute the compounds of interest.
10.6 A GC integrator (or data system if available) is
activated during the injection cycle to provide component
identification and quantification.
-------
11. Sampling and Analysis Procedure - Automated Device
11.1 This procedure assumes the use of the automated system
shown in Figure 3. The components of this system are
discussed in Section 9.1.
11.2 Prior to initial sample collection the entire assembly
should be leak-checked. This task is completed by
sealing the sample inlet line and noting that the flow
indication or the mass flow controller drops to zero
(less than 1 mL/minute).
11.3 The sample trap, valve, and transfer line are heated to
120°C and ambient air is drawn through the apparatus
(~60mL/minute) for a period of time 5-10 minutes to flush
the system, with the sample valve in the inject position.
During this time the GC column is maintained at 150°C to
condition the column.
11.4 The sample trap is then cooled to -160 ± 5°C using a
controlled flow of liquid nitrogen. Once the trap
temperature has stabilized, sample flow through the trap
is initiated by placing the valve in the ini ect position
and the desired volume of air is collected.
11.5 During the sample collection period the GC column is
stabilized at -50°C to allow for immediate injection of
the sample after collection.
11.6 At the end of the collection period the valve is
immediately placed in the inject position, and the
cryogenic trap is rapidly heated to 120°C to desorb the
components onto GC column. The GC temperature program
and data acquisition are initiated at this time.
11.7 At the desired time the cryogenic trap is cooled to
-160°C, the valve is returned to the collect position and
the next sample collection is initiated (to coincide with
the completion of the GC analysis of the previous
sample).
12. Calibration Procedure
Prior to sample analysis, and approximately every 4-6 hours
thereafter, a calibration standard must be analyzed, using the
identical procedure employed for ambient air samples (either
Section 10 or 11) . This section describes three alternative
approaches for preparing suitable standards.
-------
12.1 Teflon® (or Tedlar®) Bags
12.1.1 The bag (nominal size; 20L) is filled with
zero air and leak checked. This can be easily
accomplished by placing a moderate weight
(text book) on the inflated bag and leaving
overnight. No visible change in bag volume
indicates a good seal. The bag should also be
equipped with a quick-connect fitting for
sample withdrawal and an insertion port for
liquid injections (Figure 2).
12.1.2 Before preparing a standard mixture, the bag
is sequentially filled and evacuated with zero
air (5 times) . After the 5th filling, a
sample blank is obtained using the sampling
procedure outlined in Section 10.
12.1.3 In order to prepare a standard mixture, the
bag is filled with a known volume of zero air.
This flow should be measured via a calibrated
mass flow controller or equivalent flow
measuring device. A measured aliquot of each
analyte of interest is injected into the bag
through the insertion port using a microliter
syringe. For those compounds with vapor
pressures lower than benzene or for strongly
adsorbed species, the bag should be heated
(60° C oven) during the entire calibration
period.
12.1.4 To withdraw a sample for analysis, the
sampling line is directly connected to the
bag. Quick connect fittings allow this hook-
up to be easily accomplished and also
minimizes bag contamination from laboratory
air. Sample collection is initiated as
described.
12.2 Glass Flasks
12.2.1 If a glass flask is employed (Figure 2) the
exact volume is determined by weighing the
flask before and after filling with deionized
water. The flask is dried by heating at
2 0 0 °C.
12 .2 .2
To prepare a standard, the dried flask is
flushed with zero air until cleaned (i.e., a
blank run is made) . An appropriate aliquot of
-------
each analyte is injected using the same
procedures as described for preparing bag
standards.
12.2.3 To withdraw a standard for analysis, the GC
sampling line is directly connected to the
flask and a sample obtained. However, because
the flask is a rigid container, it will not
remain at atmospheric pressure after sampling
has commenced. In order to prevent room air
leakage into the flask, it is recommended that
no more than 10% of the initial volume be
exhausted during the calibration period (i.e.,
200cc if a 2 liter flask is used).
12.3 Pressurized Gas Cylinders
12.3.1 Pressurized gas cylinders containing selected
analytes at ppb concentrations in air can be
prepared or purchased. A limited number of
analytes (e.g., benzene, propane) are
available from NBS.
12.3.2 Specialty gas suppliers will prepare custom
gas mixtures, and will cross reference the
analyte concentrations to an NBS standard for
an additional charge. In general, the user
should purchase such custom mixtures,
ratherthan attempting to prepare them because
of the special high pressure filling apparatus
required. However, the concentrations should
be checked, either by the supplier or the user
using NBS reference materials.
12.3.3 Generally, aluminum cylinders are suitable
since most analytes of potential interest in
this method have been shown to be stable for
at least several months in such cylinders.
Regulators constructed of stainless steel and
Teflon® (no silicon or neoprene rubber).
12.3.4 Before use the tank regulator should be
flushed by alternately pressuring with the
tank mixture, closing the tank valve, and
venting the regulator contents to the
atmosphere several times.
12.3.5 For calibration, a continuous flow of the gas
mixture should be maintained through a glass
or Teflon® manifold from which the calibration
-------
standard is drawn. To generate various
calibration concentrations, the pressurized
gas mixture can be diluted, as desired, with
zero grade air using a dynamic dilution system
(e.g., CSI Model 1700) .
13. Calibration Strategy
13.1 Vapor phase standards can be prepared with either neat
liquids or diluted liquid mixtures depending upon the
concentration levels desired. It is recommended that
benzene also be included in this preparation scheme so
that flame ionization detector response factors, relative
to benzene, can be determined for the other compounds.
The benzene concentration generated in this fashion
should be cross-checked with an NBS (e.g., SRM 1805) for
accuracy determinations.
13.2 Under normal conditions, weekly multipoint calibrations
should be conducted. Each multipoint calibration should
include a blank run and four concentration levels for the
target species. The generated concentrations should
bracket the expected concentration of ambient air
samples.
13.3 A plot of nanograms injected versus area using a linear
least squares fit of the calibration data will yield the
following equation:
Y=A+BX
where
Y = quantity of component, nanograms
A = intercept
B = slope (response factor)
If substantial nonlinearity is present in the calibration
curve a quadratic fit of the data can be used:
Y=A+BX+CX2
where
C = constant
Alternatively, a stepwise multilevel calibration scheme may be
used if more convenient for the data system in use.
-------
Performance Criteria and Quality Assurance
This section summarizes the quality assurance (QA) measures
and provides guidance concerning performance criteria which
should be achieved within each laboratory.
14.1 Standard Operating Procedures (SOPs)
14.1.1 Each user should generate SOPs describing the
following activities as accomplished in their
laboratories:
assembly, calibration and operation of
the sampling system.
preparation and handling of calibration
standards.
assembly, calibration and operation of
the GC/FID system and
all aspects of data recording and
processing.
14.1.2 SOPs should provide specific stepwise
instructions and should be readily available
to, and understood by, the laboratory
personnel conducting the work.
14.2 Method Sensitivity, Precision and Accuracy
14.2.1 System sensitivity (detection limit) for each
component is calculated from the data obtained
for calibration standards. The detection
limit is defined as
DL=A+3.3S
where
DL = calculated detection limit in nanograms
inj ected.
A = intercept calculated in Section 13.
S = standard deviation of replicate
determination of the lowest level
standard (at least three determinations
are required).
For many compounds detection limits of 1 to 5
nanograms are found using the flame ionization
detection. Lower detection limits can be
obtained for chlorinated hydrocarbons using
the electron capture detector.
-------
14.2.2 A precision of ± 5% (relative standard
deviation) can be readily achieved at
concentrations 10 times the detection limit.
Typical performance data are included in Table
1.
14.2.3 Method accuracy is estimated to be within ±
10%, based on National Bureau of Standard
calibrated mixtures.
-------
REFERENCES
1. Holdren, M., Spicer, C., Sticksel, P., Nepsund, K., Ward, G.,
and Smith, R., "Implementation and Analysis of Hydrocarbon
Grab Samples from Cleveland and Cincinnati 1981 Ozone
Monitoring Study", EPA-905/4-82-001. U.S. Environmental
Protection Agency, Research Triangle Park, North Carolina,
1982 .
2. Westberg, H., Rasmussen, R., and Holdren, M., "Gas
Chromatographic Analysis of Ambient Air for Light Hydrocarbons
Using a Chemically Bonded Stationary Phase", Anal. Chem. 46,
1852-1854, 1974.
3. Lonneman, W. A., "Ozone and Hydrocarbon Measurements in Recent
Oxidant Transport Studies", in Int. Conf. on Photochemical
Oxidant Pollutant and Its Control Proceedings, EPA-600/3-77-
001a, 1977.
4. Singh, H., "Guidance for the Collection and Use of Ambient
Hydrocarbon Species Data in Development of Ozone Control
Strategies", EPA-450/4-80-008. U.S. Environmental Protection
Agency, Research Triangle Park, North Carolina, 1980.
5. Riggin, R. M., "Technical Assistance Document for Sampling and
Analysis of Toxic Organic Compounds in Ambient Air", EPA-
600/4-83-027. U.S. Environmental Protection Agency, Research
Triangle Park, North Carolina, 1983.
6. Annual Book of ASTM Standards, Part 11.03, "Atmospheric
Analysis", American Society for Testing and Materials,
Philadelphia, Pennsylvania, 1983.
7. Foulger, B. E. and P. G. Sinamouds, "Drier for Field Use in
the Determination of Trace Atmospheric Gases", Anal. Chem.,
51, 1089-1090, 1979.
8. Pheil, J. D. and W. A. McClenney, "Reduced Temperature
Preconcentration Gas Chromatographic Analysis of Ambient
Vapor-Phase Organic Compounds: System Automation", Anal.
Chem., submitted, 1984.
9. Holdren, M. W., W. A. McClenney, and R. N. Smith "Reduced
Temperature Preconcentration and Gas Chromatographic Analysis
of Ambient Vapor-Phase Organic Compounds: System
Performance", Anal. Chem., submitted, 1984.
10. Holdren, M., S. Rust, R. Smith, and J. Koetz, "Evaluation of
Cryogenic Trapping as a Means for Collecting Organic Compounds
in Ambient Air", Draft Final Report on Contract No. 68-02-
3487, 1984.
-------
11. Cox, R. D. and R. E. Earp, "Determination of Trace Level
Organics in Ambient Air by High-Resolution Gas Chromatography
with Simultaneous Photoionization and Flame Ionization
Detection", Anal. Chem. .54, 2265-2270, 1982.
12. Burns, W. F., 0. T. Tingy, R. C. Evans and E. H. Bates,
"Problems with a Nation® Membrane Dryer for Chromatographic
Samples", J. Chrom. 2 69, 1-9, 1983.
-------
rOY 'Or? QYY c.O uxinD
Figure 1. Schematic of Six-Port Valve Used for Sample Collection.
-------
W'xYq PL)(pcpQY'
QYY;1X(pŁ X&> ''
Figure 2. Dilution Containers for Standard Mixtures
-------
Figure 3. Automated Sampling and Analysis System for Cryogenic Trapping
-------
V
)
I I
yuil x Y- 2 Qfo
-------
TABLE 1. VOLATILE ORGANIC COMPOUNDS FOR WHICH THE CRYOGENIC SAMPLING METHOD HAS BEEN EVALUATED(a)
4444444444444444444444444444444444444444444444444444444444444444444444444444444444444444
Test 1 Test 2
(4 runs, 2 0 0cc samples) (8 runs, 2 0 0-cc samples)
Retention Time, Mean Mean
Compound Minutes(b) (ppb) %RSD (ppb) %RSD
))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))
Vinylidene Chloride
9.26
144
4.4
6.1
3.9
Chloroform
12.16
84
3 . 8
3 . 5
5.8
1,2-Dichloroethane
12 .80
44
3.7
1 . 9
5.1
Methylchloroform
13 . 00
63
4.5
2.7
4 . 9
Benzene
13.41
93
4 . 0
3.9
5.1
Trichloroethylene
14 .48
84
3.7
3 . 5
4 . 1
Tetrachloroethylene
17 . 37
69
3.7
2 . 9
4.3
Chlorobenzene
18 . 09
46
3 . 3
1 . 9
3.2
4444444444444444444444444444444444444444444444444444444444444444444444444444444444444444
(a) Recovery efficiencies were 100 + 5% as determined by comparing direct sample loop (5cc)
injections with cryogenic collection techniques (using test 1 data). Data from
reference 10.
(b) GC conditions as follows:
Column - Hewlett Packard, crosslinked methyl silicone, 0.32 m ID x 50 m long,
thick film, fused silica.
Temperature Program - 50°C for 2 minutes, then increased at 8°C/minute to 150°C.
------- |