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
------- |