United States
 Environmental Protection
 Agency
Environmental Sciences Research
Laboratory
Research Triangle Park NC 27711
 Research and Development
EPA-600/S3-83-058 Sept. 1983
 Project  Summary
 Optimization  of  GC/MS  Based
 Tenax  Collection   Method  for
 Toxic Organics
 Barbu Demian and Edo D. Pellizzari
  The sampling of vapor-phase organ-
ics in the presence of reactive inor-
ganics using the Tenax GC sampling
cartridge was investigated. To differen-
tiate between pollutants in air and
those formed from in situ artifact reac-
tions, deuterated organics were em-
ployed in both laboratory and field
experiments, and any subsequent arti-
fact products were detected by capil-
lary gas chromatography/mass spec-
trometiy/computer [(GC)VMS/COMP]
techniques. Artifacts produced by Tenax
GC itself and  by reactions between
adsorbed species were  investigated
using statistical methods.
  This Project Summary was developed
by EPA's Environmental Sciences Re-
search Laboratory, Research Triangle
Park. NC. to announce key findings of
the research project that is fully docu-
mented in a separate report of the
same title (see Project Report ordering
information at back).

Introduction

  The characterization and measurement
of minute amounts of hazardous com-
pounds in ambient air has, for many years,
been seriously hampered  by the lack of
reliable sampling systems. Techniques
have been developed to concentrate or-
ganic vapors from large volumes of atmos-
phere and to transfer the collected vapors
to an analytical system.
  There is, however,  a paucity of experi-
mental data concerning artifact studies on
collection devices for vapor-phase organ-
ics. Investigators have had a propensity for
accepting the viability of the collection
step prematurely while emphasizing de-
velopment  of other  steps in  analytical
procedures. In fact the literature is replete
with analytical methods for analysis of a few
vapor-phase components in complex mix-
tures using various instrumental methods.
Nonetheless, it should be noted that none
of these procedures can be considered
superior since the collection method is a
common denominator among all the ana-
lytical methods, and appears to be the
weak link
  With  the elucidation  of nitrosamine
presence in ambient air,  serious consid-
eration was given to potential artifact for-
mation during collection. In fact, a plethora
of collection methods for nitrosamines
occurred. Artifacts associated with cryo-
genic traps were readily apparent,and
modifications  ensued which  eventually
lead to using a sorbent cartridge impreg-
nated with sulfamic acid to inhibit the in
situ nitrosation of secondary amines.
  An approach to detecting, identifying,
and quantifying halogenated hydrocar-
bons in ambient air has involved using a
Tenax GC cartridge combined with  high
resolution gas chromatography/mass spec-
trometry/computer   [(GC)2/MS/COMP]
analysis. The primary concern was to use a
Tenax GC sorbent that has been previously
employed for collecting vapor-phase or-
ganics in ambient air. Since the Tenax GC
cartridge may concentrate reactive com-
pounds from an atmosphere containing,
for example, ozone (03), nitrogen oxides
(NOX), molecular halogen such as chlorine
(CI2) and bromine (Br2), in situ formation
of artifact products may occur, even though
inorganic gases do not appreciably ac-
cumulate  on  the sorbent. Preliminary
studies  indicated that a  more in-depth
study was needed.
  The objectives of this research program
were: (a) to examine the potentiaIly trou ble-

-------
some chemical group types that exhibit
reactivity with the commonly occurring
inorganic  gases in  ambient air; (b)  to
conduct a series of controlled laboratory
experiments to determine the extent,  if
any, of artifact problems associated with
the direct interaction of these inorganic
gases with Tenax GC itself; (c) to evaluate
the artifact problems, if any, associated
with organics on Tenax GC in the presence
of atmospheric concentrations of inorganic
gases; (d) to develop a way to minimize
potential artifacts if they exist; and (e) to
evaluate  and apply potential  corrective
measures that minimize artifact production
under field sampling conditions.

Effects of Ozone and Nitrogen
Oxides
  A 23 factorial experiment was  planned
to assess the ozone/nitric oxides/humid-
ified air (03/NOX/H20) combination ef-
fect on Tenax. For quantitative analysis,
the gas chromatographic system was cali-
brated with external standards. The identity
of the major Tenax  GC  decomposition
products was  established by  GC/MS
methods.
  Clean,  "virgin"  Tenax  GC cartridges
were exposed to air containing 03/NOX/
H20, and pertinent information regarding
Tenax GC cartridge exposure was obtained.
  The most important Tenax GC decom-
position  products were  benzaldehyde,
acetophenone,  and  phenol, indicating
ozonolysis to  be the  polymer's major
chemical reaction. After breakdown of the
polymer backbone, further reactions of the
primary degradation  products lead to the
relatively stable compounds actually de-
tected. This result corroborates the many
reports citing that interaction with atmos-
pheric O3 is  the main degradative mech-
anism of polymeric materials.
  Humidity had a strong effect This is
important since more attention should be
given to artifact formation from Tenax GC
when sampling air with  a high relative
humidity.
  The ozone concentration and N02 con-
centration behaved additively. The three
major decomposition products were found
in easily detectable and.dosable quantities,
even in  "clean" cartridges.  It  appears,
however, that analysis for any of the three
major decomposition  products at levels
below 100 ng/L with Tenax GC sampling
is not possible without taking  special
precautions.

Effects of Sulfur Oxides
   Tenax GC cartridges were exposed to
30 L of air  containing 03, sulfur oxides
(SOX), and NOX at different humidity levels.
The major Tenax GC decomposition prod-
ucts were confirmed but no compounds
that could identify the presence of SOX
could  be detected.

Effects of Chlorine
  Tenax GC cartridges were exposed to
03, CI2, and  humidified air. The  same
major decomposition products were ob-
served.

Effects of Tenax GC Aging
  Substantially lower levels of decompo-
sition  products were found in sampled air
containing SOX and CI2. This effect was
not related to the nature of the reactive
inorganic gas (in all experiments 03 was
present, the main factor in polymer degra-
dation), but to the history of the Tenax GC
cartridge.
  Recycled Tenax GC cartridges were used
in the SOX and CI2 series. The hypothesis
that cartridge history was the main factor
in the total quantity of degradation prod-
ucts was tested using the same cartridge
exposed to 30 L of air in 30 min, with 100
ppb 03, 1 50 ppb NOX, and 60% humidi-
fied air. The  quantities  of degradation
products decreased by an order of magni-
tude after five analysis cycles.

Artifact Products from
Adsorbed Compound
Reactions
  Tenax GC  cartridges (sorbent bed di-
mensions, 60 mm  length x 13 diameter)
were loaded with d10-cyclohexene or d5-
ethanol. For quantitation, 2-fluorobiphenyl
was used as an internal standard. Stand-
ard solutions of d^-cyclohexene and 2-
fluorobiphenyl were prepared by weighing
and dissolving  in pure methanol.
  The loaded cartridges were exposed to
an air flow containing inorganic gases in
the concentrations listed below (inorganic
gases are potential generators of artifacts
by reaction with the adsorbed d^-cyclo-
hexene:

  CI2        maximum concentration,
              564 ppb
  03        maximum concentration,
               1,000 ppb
  N0/N02   maximum concentration,
               500 ppb
  S02/S03  maximum concentration,
              400 ppb
  humidified 50% at 25°C
    air
  For the low concentration range of reac-
tive inorganic components in the sampled
air,  no deuterated  cyclohexene product
was detected. At higher chlorine concen-
trations,  d10-dichlorocyclohexane  was
 found. A nondeuterated dichlorocyclc-
 hexane was also identified in the same
 samples, probably from CI2  reacting with
 cyclohexene (a contaminant).  Deuterated
 compounds derived from d5-ethanol and
 contaminants  typical for Tenax GC were
 identified as early eluting compounds (i.e.,
 contaminants). The  potential oxidation
 products - acetaldehyde and acetic acid -
 were not detected during GC/MS analysis.

 Artifact Inhibition with
 Impregnated  Filters

  A series of filters was impregnated with
 various reducing agents for03. The sodium
 thiosulfate impregnated glass  fiber filters
 appear to be the best in quenching ozone.
 Relatively good - but less  spectacular -
 results were obtained for N02  quenching.
  The inclusion of sodium lauryl sulfate in
 the impregnating solution was proposed
 with the view that  a surfactant would
 lower the solution's  surface tension and
 allow a better wetting of the filter, thus
 improving overall performance. This be-
 havior was observed for the hydrophobic
 Teflon filters,  which cannot be impreg-
 nated  uniformly with reducing agents  in
 the absence of surfactants.

 Studies on Artifact Problems
 Under Field Sampling
 Conditions

  The results from laboratory experiments
suggested two types of filters for  best
overall  performance: glass fiber filters im-
pregnated with sodium  thiosulfate and
Teflon  filters impregnated with sodium
thiosulfate/sodium   lauryl  sulfate.  The
major objective in designing the field ex-
periments was to test these two types of
filters for inhibition of Tenax GC decom-
position and for reduction of in situ reac-
tions of adsorbed compounds.
  The  deuterated compounds selected
and the possible artifactual reactions were
as follows:
  (1) d10-cyclohexene: chlorination and
     oxidation;
  (2) d6-benzene: aromatic substitution
     and oxidation; and
  (3) d8-dioxane: oxidation (peroxide for-
      mation).
In conducting field  experiments,  two
sorbent cartridges were run parallel. One
cartridge was  protected by an impreg-
nated filter and the other cartridge by an
unimpregnated filter.
  The objectives using (GC)2/MS/COMP
analysis were as follows:
  (1) assess the recovery of deuterated
     compounds from exposed cartridges;

-------
  (2)  identify  reaction  products of deu-
      terated compounds; and
  (3)  quantify adsorbed  volatile organic
      compounds, i.e., determine the trends
      in the quantities  found, relative to
      using impregnated  filters.
  The statistical analysts of aromatic com-
pound data was restricted to the higher
concentration range,  i.e.,  more than 300
ng benzene/cartridge,  in order to avoid
unaccountable contamination effects. Only
for toluene was a positive difference statis-
tically significant. The loss of the aromatic
compounds by further reaction of adsorbed
species was, therefore, negligible at  the
level of sensitivity of the analytical method.
  The following were excluded from  the
statistical analysis of  the halogenated
compound data:

  (1)1,2-dichloroethane, due to the high
      concentration level in air;
  (2)  1,2-dibromoethane, since it is prac-
      tically absent; and
  (3)  carbon tetrachloride and dichloro-
      benzene, which are present at very
      low concentrations.

The statistical analysis was restricted to
the higher  concentration range for  the
halogenated compounds as well (same
Tenax cartridge as for aromatics).
  The collection  of chloroform  and  tn-
chloroethane data was erratic and, there-
fore,  useless in these  experiments.  For
trichloroethylene,  a  positive difference
was statistically significant fourtimes, and
for tetrachloroethylene, only once. Again,
there  was some evidence  that the impreg-
nated filter was preventing sample loss by
further reaction of the adsorbed compound
if it was relatively reactive.
  It was highly significant that the evi-
dence for sample loss by  artifact reaction
was positive only for unsaturated com-
pounds (cyclohexene,  trichloroethylene,
tetrachloroethylene), which were more re-
active toward 03.
  The phenol  data were  also unreliable.
Due to their strong acidic nature, phenol
and acetic acid are subject to serious
limitations in trace analysis by chromato-
graphic methods, the  best deactivated
capillaries bind significant quantities ir-
reversibly.

Conclusions and
Recommendations
(2)  The  tr-h'-i f.v^oit, uetcrmining  the
    qnantities of artifacts generated were,
    in order of decreasing importance, as
    follows:
    (a) 03 concentration of sampled air;
    (b) Tenax cartridge age;
    (••) humidity of sampled air;
    (d) "Concentration of NOX; and
    (e) concentrations  of other  reactive
       comp unds (SOX, CI2).
(3)  Artifact i -••ot'ons of adsorbed species
    were fount, to be:
    (a) relativeK unimportant for aromatic-;
       and saturated  halogenated com-
       pounds; and
    (b) significant for more  reactive com-
       pounds, e.c,.,  unsaturated com-
       pounds such as i'ichloroethylene
       and tetrachloroethyltirit.
(4)  Sodium thiosulfate impregnated fil-
    ters were found to be:
    (a) effective in artifacts  generated by
       Tenax GC itself; and
    (b) effective in reducing artifact reac-
       tions of reactive adsorbed species.
  Use of sodium thiosulfate impregnated
glass fiber filters for sampling with Tenax
GC cartridges is strongly recommended.
  Based on the results of laboratory and
field experiments on artifact formation,
three important  recommendations  are
n-.ade:
  (1) Sodium thiosulfate  impregnated
      glass fiber filters  should  be used,
      with an  expected improvement in
      these characteristics:
      (a)  collection  efficiency  of com-
         pounds  reactive to 03  (e.g.,  un-
         saturated compounds); and
      (b)  Tenax GC sorbent stability, which
         will result in fewer artifact prod-
         ucts.
  (2) Selection of the  sampling volume
      must consider the air's  humidity
      (80 to  100% relative humidity),
      thereby avoiding  excessive artifact
      formation, and, at the same time,
      early breakthrough for the  more
      volatile components sampled.
  (3) If the quantities found for benzalde-
      hyde, acetophenone, and phenol are
      more  than  1,000 no/cartridge,
      quantitation for these compounds
      can be conducted.
   Barbu Demian and Edo D. Pellizzari are with the Research Triangle Institute,
     Research Triangle Park, NC 27709.
   Kenneth Krost is the EPA Project Officer (see below).
   The complete report, entitled "Optimization of GC/MS Based Tenax Collection
     Method for Toxic Organics," (Order No. PB 83-229 476; Cost: $ 13.00, subject to
     change) will be available only from:
           National Technical Information Service
           5285 Port Royal Road
           Springfield,  VA22161
           Telephone: 703-487-4650
   The EPA Project Officer can be contacted at:
           Environmental Sciences Research Laboratory
           U.S. Environmental Protection Agency
           Research Triangle Park, NC 27717
                                                 *US GOVERNNtNT PRINTING OFFICE 1983-659-017/7189
(1)  The  main artifacts produced from
    Tenax GC were benzaldehyde, aceto-
    phenone, phenol, a-hydroxyacetophe-
    none, acetic acid, and ethylene oxide.

-------
United States
Environmental Protection
Agency
Center for Environmental Research
Information
Cincinnati OH 45268
Postage and
F-ees Paid
Environmental
Protection
Agency
EPA 335
Official Business
Penalty for Private Use $300
                                PS   0000339
                                U S  ENVIR  HROTtCTION  AGENCY
                                REGION  5 LIBRAKY
                                230  S  DEARBORN  STRtET
                                CHICAGO  IL 60604

-------