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

-------