EVALUATION OF AN EPA HIGH-VOLUME AIR SAMPLER
FOR POLYCHLORINATED DIBENZO-P-DIOXINS
AND POLYCHLORINATED DIBENZOFURANS
by
F. L. DeRoos, J. E. Tabor, S. E. Miller,
S. C. Watson, and J. A. Hatchel
Battelle Columbus Division
Columbus, Ohio 43201-2593
Contract Number 68-02-4127
Project Officers
Robert G. Lewis and Nancy K. Wilson
Methods Development and Analysis Division
Environmental Monitoring Systems Laboratory
Research Triangle Park, North Carolina 27711
ENVIRONMENTAL MONITORING SYSTEMS LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U. S. ENVIRONMENTAL PROTECTION AGENCY
RESEARCH TRIANGLE PARK, NORTH CAROLINA 27711
-------
DISCLAIMER
The acquisition of the information in this document has been funded by
the United States Environmental Protection Agency under Contract 68-02-4127 to
Battelle Columbus Division. It has been subjected to the Agency's peer review
and it has been approved for publication as an EPA document. Mention of trade
names or commercial products does not constitute endorsement or recommendation
for use.
ii
-------
FOREWORD
Measurement and monitoring research efforts are designed to anticipate
environmental problems, to support regulatory actions by developing an in-
depth understanding of the nature and processes that affect health and the
ecology, to provide innovative means of monitoring compliance with
regulations, and to evaluate the effectiveness of health and environmental
protection efforts through the monitoring of long-term trends. The
Environmental Monitoring Systems Laboratory, Research Triangle Park, North
Carolina, has responsibility for assessment of environmental monitoring
technology and systems, implementation of agency-wide quality assurance
programs for air pollution measurement systems, and supplying technical
support to other groups in the Agency including the Office of Air and
Radiation, the Office of Toxic Substances, and the Office of Solid Waste.
The determination of human exposure to toxic organic compounds is an
area of increasing significance to EPA. The evaluation of air sampling
methodology for polychlorinated dibenzo-p-dioxins and polychlorinated
dibenzofurans provides important information that can be applied to the
measurement of the extent of potential human exposure to these compounds.
John C. Puzak
Acting Director
Environmental Monitoring Systems Laboratory
Research Triangle Park, North Carolina 27711
in
-------
ABSTRACT
An EPA High-Volume air sampler was evaluated for retention and migration
of polychlorinated dibenzo-p-dioxins (PCDD) and polychlorinated dibenzofurans
(PCDF) within the sampler. This sampler, which is available from General
Metal Works as the Model PS-1 Sampler, consists of a filter, polyurethane foam
adsorbent cartridge, air pump, and environmental housing. The use of an
alternative adsorbent, silica gel, was also studied. Because of the high
toxicity of selected PCDD/PCDF isomers and the limited availability of pure
isomers, the study was carried out using 1,2,3,4-tetrachlorodibenzo-p-dioxin,
1,2,4,8-tetrachlorodibenzofuran, 1,2,3,4,7,8-hexachlorodibenzo-p-dioxin,
1,2,3,6,7,8-hexachlorodibenzofuran, octachlorodibenzo-p-dioxin, and octa-
chlorodibenzofuran.
The sampler retained the isomers with approximately equal efficiencies
when either PDF or silica gel was used as the adsorbent. The median retention
efficiencies for the PCDD/PCDF isomers ranged from 67 to 124 percent when PUF
was used, and from 47 to 133 percent when silica gel was used. In general,
the lowest retention efficiencies were observed for the PCDF isomers and the
highest retention efficiencies for the PCDDs. The overall average retention
efficiency for all of the isomers at two concentration levels was 99 percent
for both the PUF and the silica gel adsorbents.
Silica gel produced lower levels of background interferences than did
PUF. The detection limits were therefore approximately four times lower for
the tetrachlorinated isomers and ten times lower for the hexachlorinated
isomers when silica gel was used as the adsorbent. The difference in
detection limit was approximately a factor of two for the octachlorinated
isomers, which are of higher molecular weight than are the tetrachloro isomer,
and consequently are less susceptible to interference.
iv
-------
The desorption efficiency of the PCDD/PCDF isomers from spiked filters
was evaluated to assess the extent of migration of these compounds from the
filter to the adsorbent. Migration was dependent upon the isomers1
chlorination level with the less chlorinated, more volatile isomers generally
desorbing more efficiently. Tetrachlorinated isomers desorbed almost
completely from the filter and were collected on the adsorbent, whereas the
octachlorinated isomers were retained on the filters. Hexachlorinated isomers
gave intermediate values of desorption from the filters.
This report was submitted in addition to work previously completed in
fulfillment of Contract 68-02-4127 by Battelle Columbus Division under the
sponsorship of the U.S. Environmental Protection Agency. This report
consolidates two preliminary reports previously submitted describing work
carried out during the period of June 1, 1985 to April 30, 1986.
-------
CONTENTS
Page
Foreword iii
Abstract iv
Figures vii
Tables vii
1. Introduction 1
2. Conclusions 2
3. Recommendations 4
4. Experimental Procedures 5
Flow rate study 5
Retention study 6
Materials 10
Preparation of standard solutions 10
PDF cartridge cleanup 10
Silica gel cartridge cleanup 11
Extract cleanup 11
HRGC/HRMS analyses 13
5. Results and Discussion 19
Flow rate study 19
Retention study 19
References 27
vi
-------
FIGURES
Number Page
1 Setup for desorption experiments , 7
2 Multilayered silica column 12
3 Chromatogram obtained with high resolution capillary column
coupled directly to mass spectrometer 14
TABLES
1 Summary of flow rate experiments 5
2a Summary of retention experiments (PDF) 8
2b Summary of retention experiments (silica gel) 9
3 HRGC/HRMS operating parameters 16
4 Exact masses and MID acquisition dwell times .17
5 Flow rates for silica gel adsorbent 19
6 Retention of test compounds spiked into PS-1 Samplers
using PDF as the adsorbent 20
7 Retention of test compounds spiked into PS-1 Samplers
using silica gel as the adsorbent. 22
8 Retention efficiency performance of the PS-1 Sampler for PCDFs
and PCDDs using PUF as the adsorbent . 25
9 Retention efficiency performance of the PS-1 Sampler for PCDFs
and PCDDs using silica gel as the adsorbent 26
-------
SECTION 1
INTRODUCTION
Polychlorinated dibenzo-p-dioxins (PCDD) and polychlorinated dibenzo-
furans (PCDF) are classes of tricyclic compounds that are extremely toxic and
are of major environmental concern. Certain isomers, including 2,3,7,8-tetra-
chlorodibenzo-p-dioxin (2,3,7,8-TCDD) and 2,3,7,8-tetrachlorodibenzofuran
(2,3,7,8-TCDF), have LDsg values in the parts-per-trillion range for some
animal species (1). Major sources of these compounds have been commercial
processes involving polychlorinated phenols and polychlorinated biphenyls
(PCB). Recently, however, combustion has been shown to be a source of PCDD
and PCDF (2). A particularly significant source of these compounds is burning
transformers and/or capacitors that contain PCBs and chlorobenzenes.
The objectives of this project were to determine the retention efficiency
of the EPA High-Volume air sampler (3) by measuring the retention and
migration of selected PCDD and PCDF isomers within the sampler and to evaluate
the utility of using silica gel as the adsorbent. Previous studies involving
the collection of pesticides, PCBs, semivolatile industrial organic compounds,
1,2,3,4-TCDD and octachlorodibenzofuran (OCDF) (4,5,6,7) had been successful.
However, it was anticipated that silica gel would provide a lower background
interference level and thus allow lower detection limits to be achieved.
The study consisted of spiking the filters or adsorbent cartridges of EPA
High-Volume air samplers with selected PCDD/PCDF isomers. The PCDD/PCDF
levels that remained on the filter or the adsorbent cartridge were then
measured after a volume of approximately 325 m^ of air had been pulled through
each sampler.
-------
SECTION 2
CONCLUSIONS
The retention efficiencies for the PCDD/PCDF isomers spiked into the
samplers were in general quantitative agreement within experimental accuracy.
The median retention efficiencies ranged from 67 to 124 percent when PDF was
used as the adsorbent and from 47 to 133 percent when silica gel was used.
The lowest retention efficiencies were observed for the PCDF isomers,
particularly the tetrachlorinated and the hexachlorinated isomers.
Silica gel was found to be suitable as a replacement adsorbent for PUF.
It can be packed into the same cartridge as the PUF and produces minimal
restriction to the air flow. It does not degrade like PUF, and therefore
produces lower levels of interfering compounds. Thus, the detection limits
for the PCDD/PCDF were between two and ten times lower when silica gel was
used as the adsorbent than when PUF was used.
When the PCDD/PCDF isomers were spiked onto the filter and approximately
325 m^ of clean air was drawn through the samples, the isomers desorbed and
were collected on the adsorbent. The degree of desorption was dependent upon
the volatility of the isomer and tended to follow the level of chlorination.
The tetrachlorinated isomers were almost completely desorbed from the filter,
while the octachlorinated isomers showed only minimal migration. The
hexachlorinated isomers desorbed at a degree intermediate to the
tetrachlorinated and octachlorinated isomers.
The EPA High-Volume air sampler should be a suitable sampler for
collection of PCDD/PCDF isomers from ambient air when either PUF or silica gel
is used as the adsorbent. When PUF was used in these studies, the analytical
detection limit was approximately 0.2 ng for TCDD and TCDF, which would
compare to a theoretical ambient air detection limit of about 0.6 pg/m^ for a
24-hour sample. The use of chromatographic-grade silica gel improved
-------
detection limits by a factor of nearly four, to about 0.15 pg/m3. The silica
gel, however, was less convenient to work with and required more steps for
cleanup.
Due to the desorption of PCDD/PCDF isomers from the filter, the sampler
will not provide samples that can be used to determine the particulate
matter/vapor concentration distribution of the PCDD/PCDF isomers. If the
lowest possible detection limits are needed, silica gel should be used as the
adsorbent since it is more stable than PDF and will, therefore, minimize
interferences and provide the lowest possible detection limits.
-------
SECTION 3
RECOMMENDATIONS
Because the retention efficiencies observed for '1,2,3,6,7,8-
hexachlorodibenzofuran (HxCDF) covered a wide range, additional work is
recommended to determine if the measured retentions are realistic or if they
resulted from unreliable analytical methodology. This additional work should
include the use of an isotopically labelled HxCDF isomer as an internal
standard to correct for sample workup losses. It should also include the
evaluation of the retention efficiencies of several HxCDF isomers so that
variations due to volatility, if significant, could be observed.
Additional work should also be carried out to evaluate further the
desorption of PCDD/PCDF isomers from particulate matter. The influence of
parameters such as the time between spiking and sampling, the spiking level,
and surface characteristics of the particulate matter should be investigated.
-------
SECTION 4
EXPERIMENTAL PROCEDURES
FLOW RATE STUDY
The air flow rates through various bed heights of silica gel were
measured to determine if sufficient flow could be achieved. The silica gel
was packed into glass cartridges and was held in place using a copper screen
and thin layers of glass wool. The cartridge containing the silica gel was
placed into a standard High-Volume sampler and flow was established. The
experimental setup consisted of a High-Volume Sampler air pump, a dry gas
meter, an EPA gas flow calibrator, a filter holder, and a test cartridge
containing the silica gel. The system was allowed to equilibrate for 30
minutes prior to recording the dry gas meter readings. The weights, bed
heights, and mesh ranges of silica gel that were evaluated are summarized in
Table 1.
TABLE 1. SUMMARY OF FLOW RATE EXPERIMENTS
Experiment
1
2
3
4
5
6
7
Mesh
35-70
35-70
6-12
6-12
6-12
6-12
6-12
Weight (g)
30
60
30
60
90
120
150
Bed Height (cm)
2.5
4.5
1.9
3.2
5.1
6.4
7.6
-------
RETENTION STUDY
This evaluation consisted of spiking known levels of PCDD/PCDF isomer
into the sampler and measuring the levels of the PCDD/PCDF isomers remaining
on the adsorbent cartridges and glass fiber filters following the sampling of
approximately 325 m3 of clean air. Two PCDD/PCDF levels, 150 ng and 5 ng of
each isomer, were evaluated in triplicate. The PCDD/PCDF isomers were spiked
into the sampler as n-decane solutions. For the experiment in which the spike
was placed on the adsorbent cartridge, only the cartridge was analyzed. When
the spike was placed on the filter, the filter and cartridge were analyzed
separately. Two additional spiked cartridges, one each at the low and high
levels, were also prepared for each adsorbent and held in the laboratory
during the sampling sessions. Air was not pulled through these cartridges.
These were used as reference samples to indicate if irreversible adsorption
occurred as a function of time.
The test setup, shown in Figure 1, consisted of two high-volume sampler
heads connected in series. The first sampler contained a microfiber glass
filter and activated carbon to purify the air going into the second sampler,
which contained the test filter and adsorbent cartridge. The retention
experiments are summarized in Table 2.
Following the sampling, each of the cartridges and filters was spiked
with 2,3,7,8-TCDD-13Ci2 and octachlorodibenzo-p-dioxin-13Ci2 (OCDD-13Ci2) and
Soxhlet-extracted with benzene for 18 hours. The high level samples, e.g.
those spiked with 150 ng of the native isomers, were spiked with 50 ng of the
labelled internal standards, while the low level samples were spiked with 5 ng
of each internal standard. The benzene extracts were concentrated using a 3-
stage Snyder column, diluted 1:1 with hexane, and cleaned up using acid/base-
treated silica and alumina column chromatography. The final solutions were
analyzed by high resolution gas chromatography/high resolution mass
spectrometry (HRGC/HRMS). The spiking solutions were used to prepare response
factor standards, thus eliminating the spiking solution concentration as a
variable in the retention efficiency calculations.
-------
ROOM AIR
INTAKE
O
CRITICAL
ORIFICE
EPA CALIB. NO. 750
CHARCOAL
CANISTER
FILTER
(4" PALLFLEX
OSAT 2500)
CARTRIDGE
(ADSORBENT)
DRY GAS METER
FIGURE 1. SETUP FOR DESORPTION EXPERIMENTS
-------
TABLE 2a. SUMMARY OF RETENTION EXPERIMENTS (PUF)
Experiment
Number
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
Test
Description
PUF High Level Spike
PUF High Level Spike
PUF High Level Spike
Lab Blank
Filter High Level Spike
Filter High Level Spike
Filter High Level Spike
Lab Blank
PUF High Level Spike (held)
PUF Low Level Spike
PUF Low Level Spike
PUF Low Level Spike
Lab Blank
Filter Low Level Spike
Filter Low Level Spike
Filter Low Level Spike
Lab Blank
PUF Low Level Spike (held)
PUF Blank
Spike
Quantity, ng
150
150
150
--
150
150
150
--
150
5
5
5
--
5
5
5
--
5
--
-------
TABLE 2b. SUMMARY OF RETENTION EXPERIMENTS (SILICA GEL)
Experiment
Number
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
Test
Description
Silica Gel High-Level Spike
Silica Gel High-Level Spike
Silica Gel High-Level Spike
Lab Blank
Filter High-Level Spike
Filter High-Level Spike
Filter High-Level Spike
Lab Blank
Silica Gel High Level Spike (held)
Silica Gel Low-Level Spike
Silica Gel Low-Level Spike
Silica Gel Low-Level Spike
Lab Blank
Filter Low-Level Spike
Filter Low-Level Spike
Filter Low-Level Spike
Lab Blank
Silica Gel Low-Level Spike (held)
Silica Gel Blank
Spike
Quantity, ng
150
150
150
--
150
150
150
__
150
5
5
5
—
5
5
5
—
5
--
-------
Materials
The solvents used for sample workup and for cleaning the adsorbents
were Distilled-in-Glass grade purchased from Burdick and Jackson Laboratories,
Muskegum, MI. The 35-70 and 6-12 mesh silica gel was purchased from Aldrich
Chemical Company (21,439-6) and the alumina from BioRad, Richmond, CA. The
native and isotopically labelled PCDD/PCDF isomers were obtained from
Cambridge Isotopes, Cambridge, MA. The n-decane used to prepare the native
and isotopically labelled standard solutions was obtained from Aldrich
Chemicals, Milwaukee, WI as Gold Label grade (D90-1).
Preparation of Standard Solutions
The native PCDD/PCDF isomers were obtained as neat materials, while
the isotopically labelled isomers were obtained as isooctane solutions. The
native PCDD/PCDF standard solutions were prepared gravimetrically and were
used as the primary standards. The isotopically labelled solutions used as
internal standards were prepared as dilutions of the stock solutions obtained
from Cambridge Isotopes. All spiking solutions were stored in a freezer at
approximately -15°C except when being used.
The native spiking solutions were prepared at concentrations of 50
pg/ul and 600 pg/yl for the low level and high level spikes, respectively.
These concentrations required spiking volumes of 100 pi and 250 ul to achieve
the 5 ng and 150 ng spiking levels. The isotopically labelled internal
standards were prepared at concentrations of 50 pg/yl and 250 pg/yl.
PUF Cartridge Cleanup
The PUF cartridges were extracted with solvent before use in the
experiments. A modification of the cleanup described by Thrane and Mikelsen
(6) was used. Cartridges were rinsed sequentially with toluene, acetone, and
diethyl ether/hexane (5:95, v/v) by placing them in 3-L beakers and
compressing them with the base of a 1-L graduated cylinder. After the last
rinse they were compressed to force out as much solvent as possible and then
placed into the glass cartridge holders (2.3 in. I.D., 5 in. length). The
cartridge assemblies were placed in Soxhlet extractors and extracted with
benzene. After approximately 24 hours, they were removed from the extractors,
10
-------
drained, and placed in a vacuum oven. The oven was evacuated to approximately
250 torr, flushed twice with dry nitrogen gas and held evacuated at 25°C
overnight. The PDF cartridge assemblies were stored in screw-cap wide mouth
jars, which were wrapped with aluminum foil to protect them from light.
Silica Gel Cartridge Cleanup
Silica gel (chromatographic-grade silicic acid, 35-70 mesh, No. 24, 217-
9, Aldrich Chemical, Milwaukee, WI) was activated by heating in a Pyrex glass
tube furnace under purified nitrogen gas purge at a temperature of 180°C for
one hour. The glass tube was then removed from the furnace and cooled to
ambient temperature while the nitrogen gas flow was maintained. Nitrogen flow
was discontinued, and the silica gel was washed consecutively with 350 ml
aliquots of methanol and methylene chloride. The methylene chloride-saturated
adsorbent was returned to the furnace and heated to 50°C with nitrogen gas
purge. After 20 minutes, the temperature was raised gradually to 180°C and
maintained for 90 minutes. The dry silica gel was cooled to
ambient temperature and transferred to a fritted glass extraction thimble. It
was Soxhlet-extracted for 12 hours with methylene chloride, dried under
nitrogen gas and heated to 180°C for 1 hour. The silica gel was then cooled
and transferred to sampler cartridges. Each cartridge was loaded .with
approximately 60 g of silica gel, which was held in the sampler by plugs of
glass wool. The silica gel bed height was approximately 4.5 cm.
This quantity of silica gel was the maximum that would allow sufficient
air flow for sampling. The cartridges were placed into a Soxhlet extractor
and extracted for 18 hours using benzene. They were then removed from the
Soxhlet apparatus, placed in an oven and maintained at 120°C until they were
used.
Extract Cleanup
Column chromatography was employed both to isolate the PCDD and PCDF
isomers and to minimize coextracted interferences in the extracts (8,9,10).
The process consisted of eluting the extract through two adsorption columns.
The first column, illustrated in Figure 2, contained alternate layers of
activated silica gel, 44 percent concentrated sulfuric acid on silica gel,
11
-------
Silica Gel (~1 g)
Sulfuric Acid on Silica Gel (~5 g)
Silica Gel (~1 g)
Sodium Hydroxide on Silica Gel (~4 g)
Glass Wool
FIGURE 2. MULTILAYERED SILICA COLUMN
12
-------
silica gel, and 33 percent 1 M sodium hydroxide on silica gel. This column
was eluted with 70 ml of hexane.
The eluate from the multilayered silica column was collected,
concentrated to near dryness using a gentle stream of nitrogen gas and
redissolved in 1 ml of hexane. The hexane solution was then added to the top
of a column containing 2 g of basic alumina, which had been activated at 300°C
for 90 minutes. This column was eluted with 15 ml of hexane, 5 mL of
dichloromethane/hexane (3:97, v/v) and 20 ml of hexane/dichloromethane (1:1,
v/v) in sequence.
The hexane/dichloromethane (1:1, v/v) eluate contained the PCDD and PCDF
isomers. It was collected in a 18 ml concentrator tube and concentrated to
near dryness at 30°C with a gentle stream of ultrapure nitrogen gas. During
the concentration step, the sides of the tube were rinsed with 1 ml of
dichloromethane. The dichloromethane was allowed to evaporate to dryness (on
standing) without the use of the nitrogen gas stream. The residue was
dissolved in 20 yL of n-decane containing 10 ng of l,2,3,4,-TCDD-13Ci2 which
was used to calculate the absolute recoveries of the internal standards. The
extract was stored at 0°C and protected from light until it was analyzed.
HRGC/HRMS Analyses
The extracts were analyzed and the PCDO and PCDF were detected and
quantified with combined capillary column gas chromatography/high resolution
mass spectrometry (HRGC/HRMS) (10). The HRGC/HRMS consisted of a Carlo Erba
Model 4160 gas chromatograph interfaced directly into the ion source of a VG
Model 7070H mass spectrometer. Although zero dead volume couplers and
efficient transfer lines are available, they still degrade chromatographic
resolution because of analyte adsorption. The use of a direct-coupled
capillary column, used in this study, can be utilized to minimize the loss of
resolution. An example of the chromatographic resolution obtained by direct
interface of the capillary column to the mass spectrometer ion source is shown
in Figure 3.
The mass spectrometer was operated in the electron impact (El) ionization
mode at a mass resolution of 10,000-12,000 (M/AM, 10 percent valley
definition). This mass resolution is sufficient to resolve the test compounds
from most potential interferences. The operating parameters of the HRGC/HRMS
13
-------
SIR REPORT. FILE - 348318 .SO RUN - 1 HRHIO 2,3,7,8-TCDD + C13
UNCALIBRATED.
MASS RETENTION TIME
320
B HRS IB MINS 58 5ECS
HEIGHT
1B6 914B
AREA
545.8865
CONC.(H)
a.eeaa
322 8 HRS 18 HINS 58 SECS 114.8828 662.7338
B.BBB8
L
332 0 HRS 18 MINS 56 SECS 73.5B78 378.8467
a.eeee
334
B HRS 18 HINS 56 SECS
84.7866
442.3453
a.eeee
FIGURE 3. HIGH RESOLUTION CAPILLARY COLUMN CHROMATOGRAM
14
-------
are summarized in Table 3. The data were acquired by multiple ion detection
(MID), with two molecular ion masses from each of the analytes and two from
each internal standard being monitored. The masses were selected and the data
were acquired using a VG Model 2035 Data System. The exact masses and
acquisition dwell times are listed in Table 4. A perfluorokerosene (PFK) mass
was monitored during the analyses to calibrate the MID masses and to provide
correction for mass drift. Although the short-term stability of modern mass
spectrometers is typically better than 10 ppm, long-term stability during an
HRGC/HRMS analysis (20-30 minutes) may be as poor as 50 ppm without the use of
a lock mass.
Quantification
The PCDF/PCDD isomers were quantified by comparing the sum of the two
ions monitored for each congener class to the sum of the two ions monitored
for the corresponding internal standard. The 2,3,7,8-TCDD-13Ci2 was used to
quantify the TCDD and hexachloro isomers and the OCDD-^c^ used for the
octachloro isomers. Experimental relative response factors were calculated
from daily analyses of a test mixture prepared from the spiking solutions.
These response factors were included in all calculations used to quantify the
data. The response factors were calculated using the sum of the two ions
monitored for each class of isomers compared to the sum of the two ions
monitored for the corresponding internal standard. The average experimental
response factors were:
Silica
Analyte Gel Study PUF Study
TCDF
TCDD
HxCDF
HxCDD
OCDF
OCDD
--
1.00
0.506
0.583
1.14
1.17
1.788
0.938
1.727
2.388
0.986
1.052
The formula used for quantifying the isomers was:
n . , , Areas of Quantification Masses X Quantity of Int. Std,
quantity/sample - ftrea Qf Internal standard Masses X Res. Factor
15
-------
TABLE 3. HRGC/HRMS OPERATING PARAMETERS
Mass Resolution
Electron Energy
Accelerating Voltage
Source Temperature
Preamplifier Gain
Electron Multiplier Gain
Transfer Line Temperature
Capillary Column
Injector Temperature
Column Temp - Initial
Column Temp - Final
Carrier Gas
Flow Velocity
Injection Mode
Injection Volume
10,000-12,000 (M/AM, 10% valley definition)
70 eV
4,000 volts
2000C
107 volts/amp
-106
280°C
30 m DB-5 Fused Silica
3000C
160°C
290°C
Helium
30 cm/sec
Splitless
2 uL
16
-------
TABLE 4. EXACT MASSES AND MID ACQUISITION DWELL TIMES
Mass
303.9016
305.8987
315.9418
317.9389
319.8965
321.8936
331.9368
333.9338
373.8207
375.8178
380.9761
389.8156
391.8127
441.7428
443.7398
457.7377
459.7347
469.7779
471.7749
Dwell Time, ms
45
45
45
45
45
45
45
45
45
45
45
45
45
45
45
45
45
45
45
Description
Native TCDF
Native TCDF
TCDF-c13i2
TCDF-Cl312
Native TCDD
Native TCDD
TCDD-Cl3i2
TCDD-c13i2
Native HxCDF
Native HxCDF
PFK Lock Mass
Native HxCDD
Native HxCDD
Native OCDF
Native OCDF
Native OCDD
Native OCDD
OCDD-Cl312
OCDD-Cl312
17
-------
A limit of detection for each of the isomers. was calculated from the
laboratory method blank analyses. The formula used to calculate the limit of
detection was:
Limit of Detection/sample =
Heights of Quantification Masses X Quantity of Internal Standard X 2.5
Heights of Internal Standard Masses X Res. Factor
Quality Assurance
The operation of the HRGC/HRMS was evaluated at least once every 8 hours
by analyzing standard mixtures of PCDD/PCDF isomer. These included mixtures
of native and isotopically labelled isomers to evaluate the accuracy of
quantification, mixtures of selected PCDD/PCDF isomers to evaluate the
stability of the chromatographic elution windows, and a mixture of TCDD
isomers to evaluate the chromatographic resolution. The mass accuracy of the
MID unit was also evaluated at least every 4 hours by focusing selected ion
masses from perfluorokerosene (PFK) and correcting the slope to account for
minor variations. Mass focus stability was assured by the use of a reference
PFK "lock mass" to correct for any mass focus drift. Mass resolution was
checked every 4 hours by peak matching selected PFK ion masses. Solvents,
silica gel cartridges, and chromatographic adsorbents were periodically
analyzed as method blanks to demonstrate freedom from contamination. No
indication of contamination was observed in any of the method blanks. Decane
was also injected into the HRGC/HRMS to show that the PCDD/PCDF were not
carried over to subsequent analyses by contamination of syringes, septa, or
the capillary columns.
18
-------
SECTION 5
RESULTS AND DISCUSSION
FLOW RATE STUDY
Adequate flow air rates for sampling (>150 std. L/min) were achieved for
both meshes of silica gel and for all bed heights evaluated. Although the 6-
12 mesh silica gel could be used with bed heights of up to 7.6 cm, the finer
35-70 mesh material was chosen to provide the greatest surface area for
collection of the test PCDD/PCDF isomers. The results of the flow rate study
are summarized in Table 5.
TABLE 5. FLOW RATES FOR SILICA GEL ADSORBENT
Experiment Weight (g) Mesh Bed Height (cm) Flow Rate (L/min)
1
2
3
4
5
6
7
30
60
30
60
90
120
150
35-70
35-70
6-12
6-12
6-12
6-12
6-12
2.5
4.5
1.9
3.2
5.1
6.4
7.6
190
160
230
220
220
215
215
RETENTION STUDY
The results of the retention experiments are summarized in Tables 6 and
7. Table 6 contains the results from the experiments in which PUF was used as
the adsorbent, while Table 7 contains the results from the experiments in
which silica gel was used.
19
-------
TABLE 6. RETENTION OF TEST COMPOUNDS SPIKED INTO PS-1
SAMPLERS WITH PUF AS THE ADSORBENT
ro
o
Percent Recovery
Spike Filter/ Spike
Location PUF Level (ng)
Filter Filter 150
PUF
Filter + PUF
Filter Filter 150
PUF
Filter + PUF
Filter Filter 150
PUF
Filter + PUF
Median Filter + PUF 150
Filter Filter 5
PUF
Filter + PUF
Filter Filter 5
PUF
Filter + PUF
Filter Filter 5
PUF
Filter + PUF
TCDF
0.3
73
73
0.3
95
95
0.4
78
78
78
2
90
92
0
76
76
1
72
73
TCDD
1.5
95
96
1.8
98
100
2
90
92
96
0
116
116
0
88
88
0
81
81
HxCDF
4.2
112
116
4.7
94
99
5.4
102
107
107
0
91
91
0
103
103
3
86
89
HxCDD
9.9
114
124
9.2
90
99
12
89
101
101
4
101
105
0
110
110
4
100
104
OCDF
56
21
77
55
36
91
59
17
76
77
21
74
95
29
69
98
27
70
97
OCDD
88
16
104
83
14
97
88
9.8
98
98
32
57
89
45
57
102
45
88
133
Median
Filter + PUF
76
88
91
105
97
102
-------
TABLE 7. RETENTION OF TEST SRpOUNDS SPIKED INTO PS-1
SAMPLERS WITH SILICA GEL AS THE ADSORBENT
Percent Recovery
Spike Filter/
Location Silica Gel
Filter Filter
Silica Gel
Filter + Silica Gel
Filter Filter
Silica Gel
Filter + Silica Gel
Filter Filter
Silica Gel
Filter + Silica Gel
Median Filter + Silica Gel
Filter Filter
Silica Gel
Filter + Silica Gel
Filter Filter
Silica Gel
Filter + Silica Gel
Filter Filter
Silica Gel
Filter + Silica Gel
Spike
Level (ng) TCDD
150 3.7
77
81
150 4.9
75
80
150 4.9
69
74
150 80
5 1.6 "
61
63
5 1.3
88
89
5 1.8
73
75
HxCDF
5.8
65
71
7.8
49
57
6.6
73
80
71
7.1
68
75
4.1
74
78
5.9
46
52
HxCDD
43
78
121
60
73
133
27
124
151
133
9.4
120
129
6.2
100
106
7.1
29
36
OCDF
68
18
86
61
27
88
54
17
71
86
83
7.6
91
76
7.8
84
91
8.8
100
OCDD
96
2.0
98
99
3.1
102
88
2.3
90
98
99
6.9
106
106
6.7
113
127
6.7
134
Median
Filter + Silica Gel
75
75
106
91
113
-------
TABLE 7 . (Continued)
ro
Percent Recovery
Spike
Location
Silica Gel
Silica Gel
Blank
Blank
Silica Gel
Silica Gel
Silica Gel
Median
Silica Gel
Silica Gel
Silica Gel
Median
Blank
Blank
Blank
Filter/
Silica Gel
Silica Gel Spike
Silica Gel Spike
..
--
Filter
Silica Gel
Filter + Silica Gel
Filter + Silica Gel
Filter
Silica Gel
Filter + Silica Gel
Filter + Silica Gel
--
~ ~
Spike
Level (ng)
150
5
..
--
150
150
150
150
5
5
5
5
--
™ ™
TCDD
108
82
ND
ND
92
89
89
89
92
81
92
92
ND
ND
ND
HxCDF
46
38
ND
ND
4
47
61
47
38
64
116
64
ND
ND
ND
HxCDD
129
112
ND
ND
101
105
138
105
127
132
163
132
ND
ND
ND
OCDF
115
85
ND
ND
90
97
99
97
110
99
108
108
ND
ND
ND
OCDD
101
95
ND
ND
104
101
101
101
118
119
128
119
ND
ND
ND
ND = Not Detected at a detection limit of approximately 0.05 ng for TCDD and for HxCDF/HxCDD
and 0.5 ng for OCDF/OCDD.
-------
The average retention efficiency for the PCDD/PCDF isomers spiked on
filters was dependent upon the volatility of the particular isomer. In
general, the tetrachlorinated isomers desorbed from the filters and were
collected on the adsorbent, while the octachlorinated isomers remained on the
filters. The hexachlorinated isomers exhibited intermediate behavior. When
the data from the PUF and silica gel experiments are averaged at the two
spiking levels, approximately 1.8 percent of the tetrachlorinated isomer
spikes was retained on the filters, while 83 percent of the octachlorinated
isomer spikes was retained after sampling. The hexachloro isomer spikes were
retained on the filters at approximately 10 percent.
The average performance of the sampler for retention of PCDO/PCDF isomers
is summarized in Tables 8 and 9. Table 8 summarizes the results for the
sampler when PUF is used as the adsorbent and Table 9 the results when silica
gel is used. All of the retention efficiencies are within ^25 percent of
quantitative, except for the two hexachloro isomers which vary by
approximately +3Q percent.
The adsorbent retention efficiencies for the HxCDD and HxCDF isomers
were the lowest and the most variable of those measured. Although it is
possible that these isomers were breaking through the adsorbents, it is also
possible that they were lost during the sample workup. Since an isotopically
labelled hexachlorinated dioxin or furan was not available for use as an
internal standard, losses due to extraction and cleanup were corrected based
on the recovery of the labelled tetrachlorinated standards. It was assumed
that the volatility and extraction efficiency of these standards were similar
to the hexachlorinated isomers, however, the validity of these assumptions has
not been tested for all matrices.
24
-------
TABLE 8. RETENTION EFFICIENCY PERFORMANCE OF THE PS-1 SAMPLER FOR
PCDFs AND PCDDs USING PUF AS THE ADSORBENT
Average Percent Recovery
Spike
Level1,
Medium ng/nn
Filter 0.5l
PUF
Total Accountability
Filter 0.02
PUF
Total Accountability
TCDF
0.32
82
82
2
79
81
TCDD
1.5
95
96
0
95
95
HxCDF
4.8
103
108
1
93
94
HxCDD
10
97
107
3
104
107
OCDF
57
25
82
26
71
97
OCDD
86
13
99
41
67
108
1 Spike levels based on total weight of PCDD/PCDF isomer spiked, ,
assuming 325 m^ of air were sampled.
2 Average for three experiments.
25
-------
TABLE 9. RETENTION EFFICIENCY PERFORMANCE OF THE PS-1 SAMPLER FOR
PCDFs AND PCDDs USING SILICA GEL AS THE ADSORBENT
Spike Level
Medium ng/m3
Filter 0.5
Silica Gel
Total
Accountability
Filter 0.02
Silica Gel
Total
TCDD
4.5
74
78
1.6
74
76
Average
HxCDF
6.7
62
69
5.7
63
68
Percent
HxCDD
43
92
140
7.6
83
91
Recovery
OCDF
61
21
82
83
8.1
91
OCDD
94
2.5
97
110
6.8
120
Accountability
26
-------
REFERENCES
1. Polar, A., Kuntson, J., Ann. Rev. Pharmacol. , 22, pp. 517-554 (1982).
2. Lustenhouwer, J.W.A., Olie, K., Hutzinger, 0., Chemosphere, 9, pp. 501-
522 (1980).
3. Lewis, R. G., and Jackson, M. D., Anal. Chem., 54, 592-594 (1982).
4. Lewis, R. G., Brown, A. R., and Jackson, M. D., Anal. Chem., 49, 1668-1672
(1977).
5. Lewis, R. G., and MacLeod, K. E., Anal. Chem., J54, 310-315 (1982).
6. Thrane, K. E., and Mikalsen, A., "High Volume Sampling of Airborne
Polycyclic Aromatic Hydrocarbons Using Glass Fiber Filters and
Polyurethane Foam". Atmos. Environ., JL5_(6), 909-918 (1981).
7. "Evaluation of the EPA High-Volume Sampler for Collection of
Polychlorinated Dibenzo-p-dioxins", DeRoos, F. L., Tabor, J. E., Miller,
S. E., Watson, S. C., and Hatchel, J. A., Progress Report, EPA Contract
68-02-3487, Battelle Columbus Laboratories, Columbus, Ohio, September 24,
1984.
8. Determination of Tetra-Hexa CDF, Tetra-Penta CDD, and Tetra-Penta
Biphenylenes in Air Samples from Floors 3, 5, 7, and 9 of the Binghampton
New York State Office Building. Smith, R. M., Hiler, D. O'Keefe, P., and
Aldons, K. Report from Laboratory of Organic Analytical Chemistry, New
York State Dept. of Health, Albany, NY, June 27, 1983.
9. Determination of 2,3,7,8-TCDD in Soil and Sediment, September 1983, U.S.
Environmental Protection Agency, Region VII Laboratory, Kansas City,
Kansas.
10. Harless, R. L., Oswald, E. 0., Wilkinson, M. K., Dupuy, A. E., Jr.,
McDaniel, D. D., and Tai, H., "Sample Preparation and Gas Chromatography-
Mass Spectrometry. Determination of 2,3,7,8-Tetrachlorodibenzo-p-dioxin",
Anal. Chem., 52_, 1239-1245 (1980).
27
------- |