United States
 Environmental Protection
 Agency
 Environmental Monitoring and
 Support Laboratory
 Cincinnati, OH 45268
*
 Research and Development
EPA/600/S4-85/009 Mar. 1985
 Project Summary
Thermally Modulated  Electron
Affinity   Detector for  Priority
Pollutant  Analysis

R. C. Hanisch, L. D. Ogle, A. E. Jones, and R. C. Hall
  In the area of environmental monitor-
ing, a need exists for a rapid, sensitive,
and  selective method  to analyze for
chlorinated  organic compounds such
as pesticides, PCB, PCDD, and PCDF at
trace levels in complex samples.
  In  response to this need, a program
was  conducted to determine the feasi-
bility of using a new detector concept
in the gas chromatographic analysis of
certain priority pollutants. The concept
is based on the thermal alteration of a
compound's electron affinity in a flow-
through reactor, which can be used to
modify the selectivity and sensitivity of
the ECD  to certain  compounds.  The
Thermally Modulated Electron Capture
Detector  (TM ECD)  consists of  two
ECDs connected by  a temperature-
controlled reactor.  Different classes of
organic  compounds respond to  the
reactor conditions in different ways:
some compounds exhibit an enhanced
ECD  response after passing through
the reactor; some a diminished signal;
and others no change in the magnitude
of the signal. The ratio of a compound's
response from the  postreactor ECD to
that  obtained from the prereactor  ECD
appears to be a property characteristic
of each compound. This peak area  ratio
can  be used  in conjunction with its
retention  time to  increase the  con-
fidence level of the identity of a given
compound while still taking advantage
of the excellent sensitivity character-
istics of the ECD.
  This  Project Summary was devel-
oped by EPA's Environmental Monitor-
ing and Support Laboratory, Cincin-
nati,  OH,  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 potential for enhancing the selectivi-
ty and sensitivity of the ECD to certain
compounds  was demonstrated  in  pre-
liminary studies conducted by R.C. Hall at
Purdue University in 1973. These studies
utilized a detector system comprised of two
ECDs and  a flow-through  reactor.  The
detectors in this system were arranged in
series, but were separated by a reaction
chamber. A gold reaction tube and nitrogen
carrier gas were used in this design.
  It was  found that at  moderate reactor
temperatures (up to 800°C), the response
of the  second detector was:  reduced
relative to the first for chlorinated hydrocar-
bon pesticides; approximately the same for
PCBs; and greater for phthalate esters. At
high reactor temperatures (950°C), PCBs
and phthalates continued to produce a sig-
nificant response in the second detector;
most chlorinated  hydrocarbon pesticides
exhibited  little or  no  response.  The
response  ratio of the detectors was also
found to be compound-specific and useful
for confirmation of compound identity.
  These phenomena were used to enhance
detector selectivity by directly eliminating
the response of certain components and by
differentially  summing the two detector
signals to eliminate the response of stable
components. These techniques were  used
to  enhance  detector  selectivity to
chlorinated hydrocarbon pesticides in the
presence of PCBs and phthalates.

-------
  Although these preliminary results were
encouraging,  this  approach  was   not
studied in detail and did not cover a wide
variety of compounds. The stability of the
system was not verified and actual samples
were not investigated. Consequently, fund-
ing was  solicited to fully investigate  the
utility of  the technique and determine the
feasibility of constructing  a commercial
detector  system based on this principle.
  This followup program was designed to
demonstrate the feasibility of the concept
using existing off-the-shelf components.
The specific objectives were:
  • construct a detection system;
  • determine response characteristics for
    selected compounds as a function of
    reactor temperature using at least two
    different materials as reactor tubes;
  • investigate response characteristics in
    different reaction gas compositions in-
    cluding nitrogen, argon/methane, and
    helium/hydrogen;
  • determine the most promising reaction
    tube/reaction gas combination;
  • system evaluate with a limited number
    of model compounds; and
  • evaluate detector specificity.


Hardware Development and
Preliminary Evaluation
  The chromatographic system used in this
study consisted  of a  Radian  110B  Gas
Chromatograph modified to accept an ex-
perimental  detector system  which con-
sisted of  two modified Tracor ECDs with a
flow-through  reactor  between  the  two
detectors.
  The reactor used for this study consisted
of a  two-hole  ceramic tube wrapped with
resistance heating  wire and enclosed in
ceramic insulation. The  reaction tube was
inserted through one hole of the ceramic
tube and a thermocouple in the other hole.
Temperatures  of the reactor were  con-
trolled + 2°C and were varied from 350°C
to 900°C.
  A 1/8-inch  o.d. glass-lined stainless steel
column (1.8 mm i.d. x  168 cm long) was
used for all separations. The exit of the col-
umn was interfaced to the first detector via
a short piece  of y16-inch o.d.  (0.035-inch
i.d.) gold tubing. Gold was used for this in-
terface to prevent catalytic decomposition
of the sample prior to detection.
  Seventeen model compounds were se-
lected to determine the TM ECD's response
characteristics. The basis for their selection
was that they represent  electron-capturing
analytes  from  a  variety  of  compound
classes  including  chlorinated  hydrocar-
bon pesticides,  PCBs,  phthalate esters,
organophosphate  pesticides,  chloroaro-
matics, nitroaromatics, and chlorophenols.
The  test   compounds  were  grouped
into mixtures  each  of  which  contained
similar  compounds  that  could  be  re-
solved  chromatographically   under the
analytical  conditions employed. The mix-
tures  were  introduced to the chromato-
graph in 5 /iL injections.
  The response  ratios  determined  ex-
perimentally at  various temperatures with
argon/methane carrier and a gold  reaction
tube are presented in Table  1. The  ratios
shown are the average of three determina-
tions.  The  standard  deviations  for the
determinations  are  also  shown. These
ratios are based on the assumption that the
responses of ECD#1 and ECD#2 are equal for
all  compounds with an ambient reactor
temperature. An ambient temperature for
the reactor, however, was impossible to at-
tain due to the temperature of the detectors
(340°C each). Therefore,  equal response
was assumed and a reactor temperature of
350°C was used as the minimum tempera-
ture.
  At 350°C, the response ratios were not
equal to 1.0 for all  compounds. This in-
dicates  that some  rearrangement takes
place either in the first detector or in the
reaction tube  at 350°C.  The compounds
altered the  most were toxaphene, chlor-
dane (both chlorinated hydrocarbons), and
the phthlates.
  As the temperature of the reactor was in-
creased,  the chlorinated pesticides  were
degraded to species  less responsive to the
electron capture detector. This  resulted in
response  ratios  less than  1.0.  The  most
thermally stable chlorinated  pesticides were
p,p'-DDE  and p,p'-DDT. Toxaphene  and
chlordane, both aliphatic chlorinated hy-
drocarbons,  were  found  to  be  very
unstable.
  PCBs,   nitrobenzene,   1,2,4-trichloro-
benzene and 2,4,-dichlorophenol exhibited
high  thermal stability. The thermal stability
of PCBs is well documented. Therefore, the
response ratios were expected to be close
to 1.0 and show little change as a function
of temperature.  The  stability  of nitro-
benzene, 1,2,4-trichlorobenzene and 2,4-di-
chlorophenol was  surprising. These com-
pounds, particularly nitrobenzene and 2,4-
dichlorophenol, are chemically reactive and
were therefore expected to  exhibit thermal
instability.
  Phthalates and   the  parathions were
found to  be thermally unstable. The or-
ganophosphate  pesticides,  methyl  and
ethyl parathion, were thermally degraded to
products having very little electron affinity.
As a result, the respone ratios  were quite
Table 1,    Response Ratios' At Various Temperatures With Ar/CH, Carrier And A Gold Reaction Tube

                                                   Reactor Temperature (°C>

1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
Compound
Lindane
Heptachlor
Heptachlor Epoxide
p,p'-DDE
p,p'-DDT
Aroclor 1016
Aroclor 1254
Toxaphene
Chlordane
Diethyl Phthalate
Dibutyl Phthalate
bis(2-Ethylhexyl)Phthalate
Methyl Parathion
Ethyl Parathion
Nitrobenzene
1,2,4- Trichlorobenzene
2, 4-Dichlorophenol
'Based on equal response of ECD
350
0.93±0.01
O.S2±0.01
0.82±0.01
0.81 ±0.01
0.78+0.01
0.98±0.02
0.94±0.06
0.61 ±0.06
0.72 + 0.01
1.54+0.30
1.61 + 0.24
1.84+0.39
0.90+0.01
0.92+0.01
1.00 + 0.02
0.82 ±0.01
1.07+0.03
it land ECD #2.
600
0.93+0.01
0.77+0.01
0.82 + 0.01
0.83+0.01
0.78 ±0.05
1.00+0.01
1.01 + 0.01
0.64+0.02
0.69 + 0.02
1.58+0.10
1.55+0.04
1.80+0.05
0.89+0.01
0.91 ±0.01
0.84 + 0.01
1.02 + 0.02
1.04 + 0.01

700
0.90 + 0.01
0.41 + 0.06
0.77+0.02
0.82+0.01
0.54+0.03
0.98+0.02
1.01 ±0.02
0.05+0.03
0.27+0.10
1.82±0.01
1.57 ±0.04
1.85+0.07
0.31 ±0.01
0.26+0.01
0.90+0.03
0.97+0.01
0.90+0.05

800
0.11 + 0.01
0.05
0.17+0.01
0.59 + 0.02
0.21 + 0.03
0.92+0.03
0.92+0.03
<0.01
0.03
3.48+0.04
1 86+0.05
2.04+0.05
0.09+0.01
0.09+0.01
0.84 + 0.02
0.92 + 0.01
0.85+0.03
850
0.05
0.02
0.08+0.01
0.48+0.01
0.25+0.03
0.92+0.01
0.89+0.03
<0.01
0.01
6.55+0.04
2.52+0.02
2.44+0.08
0.08+0.01
0.09+0.01
0.85+0.01
0.93 + 0.01
0.75+0.01
900
0.04
0.02
0.05
0.40+0.01
0.15 + 0.01
0.85+0.01
0.77+0.04
<0.01
<0.01
8.09+0.15
3. 16 ±0.02
3.03 + 0.12
0.07+0.01
0.08 + 0.01
0.90+0.02
0.91 + 0.01
0.77+0.03

-------
small.  The  phthalates were  also  very
unstable. However, the products formed
during the  thermal  degradation of  the
phthalates had a greater electron affinity
than the parent compounds.
  In general, the results with nitrogen as
the carrier were similar to those obtained
with argon/methane. The chlorinated pes-
ticides, the parathions, and the phthalates
were again found to be thermally unstable.
The PCBs, nitrobenzene, trichlorobenzene,
and dichlorophenol, were found to be more
thermally stable.
  The response ratios  for the helium/hy-
drogen gas composition were obtained by
using a carrier of 30 mL/min helium and a
makeup gas of 30  mL/min  hydrogen.  The
makeup gas was added to  the column ef-
fluent immediately prior to entering the first
detector.
  Response ratios obtained with a helium/
hydrogen  carrier gas  did  not  yield  re-
producible   results  for  the  chlorinated
pesticides or  for the  phthalates.  Repro-
ducible results were obtained for the PCBs,
2,4-dichlorophenol,  and  1,2,4-trichloro-
benzene.  Toxaphene,   chlordane,  nitro-
benzene and the parathions had very low
response ratios at all  temperatures  and
were not studied in great detail.
  Heating  the  reactor to  temperatures
above 800°C with helium/hydrogen carrier
caused an apparent activation of the reac-
tion tube. The reactor would require several
days to return to its original level of activa-
tion once the  temperature had been  re-
duced. This activation caused very poor re-
producibility, especially for the chlorinated
pesticides and  phthlates. The reactor ac-
tivation was postulated to be due to a tem-
perature dependent reaction  between the
hydrogen  and some  substance  coating
either the inside of the ECDs or the reaction
tube.
  Replacement of  the  gold reaction tube
with a nickel tube did not alleviate the prob-
lems observed with gold and helium/hydro-
gen.  All   compounds,  except ' diethyl
phthalate, were completely destroyed  at a
reactor temperature of 900°C. The activa-
tion of the reactor was again  observed at
high temperatures. After reducing the reac-
tor temperature, several days were required
for  restoration of the original activity level.
  Response  ratios  determined  with  a
nitrogen carrier and a nickel reaction tube
were found to be lower at 350°C for every
compound except the phthalates and nitro-
benzene. The  phthalates  were  observed
earlier to have larger response ratios due to
the formation  of  a  degradation  product
with a  greater  electron affinity. The  in-
creased degradation of all  compounds at
350°C with nitrogen/nickel  suggested  that
catalytic reactions were occurring (i.e., the
gold tube is more inert than the  nickel
tube). The nickel tube displayed another
difference  from   the  gold   tube  with
nitrogen. At temperatures >800°C all com-
pounds  were degraded including the nor-
mally thermally  stable   PCBs,  chloro-
aromatics  and  nitroaromatics.  In  ad-
dition, the response ratios of the phthalates
were  smaller,  suggesting that  the  in-
termediate species formed  by the thermal
degradation of the phthalates were degrad-
ed further to species which have small elec-
tron affinities. Due to these difficulties, this
system was not investigated further.
  Response  ratios  obtained  using  an
argon/methane carrier and a nickel reaction
tube at 350°C are very similar to those .ob-
served for a nitrogen carrier when a nickel
tube is employed as the reactor. The reac-
tivity of the nickel tube appears to increase
more rapidly with a corresponding increase
in temperature when  argon/methane  is
used in place of nitrogen  as the carrier.
  At a reaction temperature of 800°C the
response ratios of all the compounds ex-
cept the phthalates were extremely small
for the  argon/methane system, whereas
the response ratios of the chlorinated com-
pounds  and the phthalates resulting from
the  nitrogen  system  were  significantly
greater  at the  same  temperature. The
response  ratios for the organophosphate
pesticides, the chloroaromatics, and the
nitroaromatics  are very  similar  for  both
systems at this temperature.
  In the case of the chlorinated pesticides,
these data may be-indicative of the occur-
rence of free radical formation in the argon/
methane atmosphere within the nickel reac-
tor, followed by subsequent recombination
reactions in which species that have lower
electron  affinities are formed. The nickel
tube appears  to act as  a catalyst under
these conditions since the test compounds
exhibit much greater thermal stability when
argon/methane is used in conjunction with
a gold  reaction tube.  Because  of these
characteristics,  this system  was  not
selected for further study.

Method Evaluation Study
  The optimum experimental configuration
was determined to be a gold catalyst at
850°C with a total flow rate of 60 mL/min
of 5% methane/95% argon. This configur-
ation was selected for further study. The
target compounds selected for use in the
method  evaluation  study  included
phthalate esters   and  toxaphene.  The
evaluation study itself was based  on the
analytical procedures employed in USEPA
Method 608 for organochlorine pesticides
and PCBs. The study was conducted based
on the assumption that the phthalate esters
were to act as interferents in the analysis of
water samples for toxaphene.
  The following phthalate  esters  were
selected  for  use in the method  evalua-
tion:  di-/?-butyl   phthalate,  benzyl  butyl
phthalate, and bis(2-ethylhexyl)  phthalate.
  Twenty 1-L samples  of  reagent water
were fortified with a 1.0 rnL aliquot of spik-
ing solution containing appropriate concen-
trations  of  toxaphene  and  the  three
phthalate esters.  After spiking, the 20 water
samples  were extracted according to the
protocol  detailed in  USEPA  Method 608.
Briefly, this involved extracting the sample
with three 60-mL portions of  methylene
chloride,  drying  the combined extracts on
an anhydrous sodium sulfate column, and
concentrating the dried extracts followed
by solvent exchange into hexane. After ex-
traction,  the samples  were  analyzed  on
both a conventional ECD and TM ECD.
  After analysis, the extracts were cleaned
up  on Florisil®*  columns using the pro-
cedure recommended in USEPA Method
608. The 6%  diethyl ether in hexane eluate
fraction  was  analyzed  under  the same
chromatographic conditions. Prior to be-
ginning  this phase of  the  study,  the
Florisil® elution pattern for toxaphene was
established using standard solutions.
  A comparison  of  the two detectors in
terms of accuracy and precision for both
the pre- and post-cleanup methods demon-
strated that the  commercial ECD and the
TM ECD  work equally well for the analysis
of toxaphene in fortified  reagent water
samples.  This is true for both pre- and post-
cleanup methods.
  Industrial  wastewater  samples  (1L
volumes) known  to contain  toxaphene as a
contaminant were extracted and analyzed
according to USEPA  Method 608. The
analyses were conducted prior to and after
Florisil® column  cleanup using both detec-
tion systems.
Four  toxaphene-containing  wastewater
samples  were obtained  for use  in the
evaluation study. One of the samples was
divided into three 1-L aliquots. Two of the
aliquots were extracted and analyzed as
previously described. The third aliquot was
fortified with  a toxaphene spiking solution
at a level that would yield  an  equivalent
sample  concentration  of  86 /ig/L  as  a
method recovery check. The three other in-
dustrial wastewater  samples  were  each
treated as single determinations. One of
these samples contained phthalate levels so
high that measurements  of background
"Mention of trade names or commercial products does
 not constitute endorsement or recommendation for
 use.

-------
   toxaphene concentrations (before cleanup)
   had to be made using a single toxaphene
   peak.
    The results of the method recovery check
   are listed  in Table 2 and indicate quan-
   titative recovery of toxaphene from the for-
   tified  industrial waste sample. As was il-
   lustrated for the reagent water samples, the
   TM ECD appears to be equivalent in this
   particular application.
    An additional series of analyses using the
   TM ECD were  performed in an attempt to
   identify the mechanism responsible for the
   enhanced  phthalate  response. It was hy-
   pothesized that phthalate esters were being
   converted to phthalic anhydride within the
   heated  reactor and  that the  enhanced
   signal was the result of the relatively greater
   electron affinity of the anhydride.
    The phthalic  anhydride chromatographic
   peak  tailed so badly, however, that it was
   impossible to get an accurate comparison.
   Nevertheless,  it was found  that  phthalic
   anhydride was more  sensitive than the
   phthalate ester. The difference between the
   sensitivities of the anhydride and the ester
   was approximately the same as the elevated
   response observed for the ester after pass-
   ing through the reactor at 850°C. Thus, the
   formation of phthalic anhydride is a plausi-
   ble reaction mechanism.

   Recommendations
    The evaluation of the detector prototype
   has demonstrated the validity of the TM
   ECD concept. A final determination of the
   detector's potential for widespread applica-
   tion  in the area of environmental monitor-
   ing is not possible without additional work.
   The  following  areas merit continued  in-
   vestigation:
         • determine  the  reaction  products
           responsible  for  the  post-reactor
           signals;
         • determine molecular positional effects
           on  the  response  factors  of  various
           isomers;


       Table 2.    Method Recovery Check
                     Parameter
 utilize capillary columns in conjunction
 with the TM ECD;
 define the  limits of  matrix effects on
 TM ECD response ratios; and
 optimize TM ECD selectivity and sen-
 sitivity for PCD and  PCDF.
         Concentration f^g/Li
   Pre-Cleanup         Post-Cleanup
ECD      TMECD    ECD     TMECD
Average toxaphene concentration (duplicates)
Toxaphene spiking level (fig/L)
Predicted toxaphene concentration (^g/Li
Analyzed toxaphene concentration lji.g/L)
% toxaphene recovery
122
86
208
235
113
114
86
200
195
98
120
86
206
196
95
108
86
194
183
94
          R C. Hanisch, L D Ogle, A. E. Jones, and P. C. Hall are with Radian Corporation,
            Austin, TX 78766.
          Stephen Billets (formerly with EMSL-Cincinnati) is the EPA Project Officer (see
            below).
          The complete report, entitled "Thermally Modulated Electron Affinity Detector for
            Prioity Pollutant Analysis," (Order No. PB  85-158  145/AS;  Cost: $10.00,
            subject to change) will be available only from:
                  National Technical Information Service
                  5285 Port Royal Road
                  Springfield, VA 22161
                  Telephone: 703-487-4650
          The EPA Project Officer can be contacted at:
                  Environmental Monitoring Systems Laboratory
                  U.S. Environmental Protection Agency
                  P.O. Box 15027
                  Las Vegas, IW 89114
                                              U S GOVERNMENT PRINTING OFFICE' 1985-559-016/27002
United States
Environmental Protection
Agency
Center for Environmental Research
Information
Cincinnati OH 45268
Official Business
Penalty for Private Use $300
          OCOC329   PS
          U S  ENVIR  PROTECTION  AGENCY
          nS'JM.WSTCT.j.T
          CHICAGO                It-   60604

-------