EPA-600/4-76-061
December 1976
Environmental Monitoring Series
ON-LINE MEASUREMENT OF THE INFRARED
SPECTRA OF GAS CHROMATOGRAPHIC
ELUENTS
Environmental Research Laboratory
Office of Research and Development
U.S. Environmental Protection Agency
Athens, Georgia 30601
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development, U.S. Environmental
Protection Agency, have been grouped into five series. These five broad
categories were established to facilitate further development and application of
environmental technology. Elimination of traditional grouping was consciously
planned to foster technology transfer and a maximum interface in related fields.
The five series are:
1. Environmental Health Effects Research
2. Environmental Protection Technology
3. Ecological Research
4. Environmental Monitoring
5. Socioeconomic Environmental Studies
This report has been assigned to the ENVIRONMENTAL MONITORING series.
This series describes research conducted to develop new or improved methods
and instrumentation for the identification and quantification of environmental
pollutants at the lowest conceivably significant concentrations. It also includes
studies to determine the ambient concentrations of pollutants in the environment
and/or the variance of pollutants as a function of time or meteorological factors.
This document is available to the public through the National Technical Informa-
tion Service, Springfield, Virginia 22161.
-------
EPA-600/4-76-061
December 1976
ON-LINE MEASUREMENT OF THE INFRARED SPECTRA OF
GAS CHROMATOGRAPHIC ELUEWTS
Peter R. Griffiths
Department of Chemistry
Ohio University
Athens, Ohio ^5701
Grant Number R80351T-01-0
Project Officer
Leo V. Azarraga
Environmental Research Laboratory
Athens, Georgia 30601
ENVIRONMENTAL RESEARCH LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
ATHENS, GEORGIA 30601
-------
DISCLAIMER
This report has been reviewed by the Athens Environmental Research
Laboratory, U.S. Environmental Protection Agency, and approved for publica-
tion. Approval does not signify that the contents necessarily reflect the
views and policies of the U.S. Environmental Protection Agency, nor does
mention of trade names or commercial products constitute endorsement or
recommendation for use.
ii
-------
FOREWORD
Nearly every phase of environmental protection depends on a capability
to identify and measure chemical pollutants in the environment. As part of
this Laboratory's research on the occurrence, movement, transformation,
impact and control of specific environmental contaminants, the Analytical
Chemistry Branch develops techniques for identifying and measuring chemical
pollutants in water and soil.
This report described the development of two systems for improving
detection limits for gas chromatography - Fourier Transform infrared spec-
troscopy (GC-FTIR). GC-FTIR is being evaluated by EPA for applic ations in
identifying organic pollutants at trace concentrations in water. If the
technique proves feasible, it will improve the assessment of health and
ecological effects of organic compounds in water and improve the development
and implementation of control measures.
David W. Duttweiler
Director
Environmental Research Laboratory
Athens, Georgia
111
-------
ABSTRACT
Techniques for increasing the sensitivity of the interface between a
gas chromatograph and a rapid-scanning Fourier transform infrared spectro-
meter (GC-IR) have been developed. A single-beam system, in which a tri-
glycine sulfate (TGS) detector is used to measure the interferogram, has been
designed and constructed. Identifiable infrared spectra of submicrogram
quantities eluting from a gas chromatograph have been measured without trap-
ping the sample using this system. A double-beam configuration for GC-IR has
also been designed so that a cooled mercury cadmium telluride detector can be
used to further decrease the detection limits without limiting the sensitivi-
ty by digitization noise. Each of these systems necessitates the use of
light-pipes with relatively long (30 cm) absorbing paths but low reflection
losses.
This report was submitted in fulfillment of Grant Number R803517-01-0
by Ohio University under the (partial) sponsorship of the U.S. Environmental
Protection Agency. Work on the task of further increasing the sensitivity
of the GC-IR interface is continuing under E.P.A. Grant R80^333-01-0. This
report covers the period from 1/20/75 to 1/19/76 and work was completed as of
1/19/76.
iv
-------
CONTENTS
Foreword iii
Abstract iv
Figures vi
Abbreviations vii
1. Introduction 1
2. Conclusions 2
3. Recommendations 3
4. Experimental 4
5. Results 6
6. Applications 8
7. Discussion and Future Work 9
References 12
-------
FIGURES
Number Paqe
On-line GC-IR spectrum of 2 ug and 800 ng of
salicylaldehyde measured using a TGS detector 7
Dual-beam FT-IR optical configuration for high
sensitivity GC-IR 11
VI
-------
DB-FT-IR
FT-IR
GC
GC-IR
GC-JB
JCT detector
(Jg
ng
SCOT columns
TGS detector
D*
ABBREVIATIONS
— dual-beam Fourier transform infrared spectres copy
— Fourier transform infrared spectroscopy
— gas chromatography
— the interface between a gas chromatograph and a rapid-
scanning infrared spectrometer for the on-line measurement
of the infrared spectra of components of mixtures separat-
ed by gas chromatography
— the interface between a gas chromatograph and a mass
spectrometer for the on-line measurement of the mass
spectra of components of mixtures separated by gas
chromatography
— mercury cadmium telluride photoconductive infrared
detector
— microgram, 10 6 grams
— nanogram, 10"9 grams
— support-coated open tubular columns for gas chromatography
— triglycine sulfate pyroelectric infrared detector
— figure of merit for detector sensitivity
VI1
-------
SECTION I
INTRODUCTION
Before this project was initiated, the most sensitive interface between
a gas chromatograph and an infrared spectrometer (GC-IR) designed for use
with a commercial Fourier transform infrared (FT-IR) spectrometer was an
accessory to the Model FTS-l^t- spectrometer marketed by Digilab Inc.
(Cambridge, MA). The detection limits for "on-the-fly" measurements of the
infrared spectra of most materials elating from a gas chromatograph that
were determined using this accessory were not very low, typically 3 to 20
micrograms (|jg); for most analytical purposes, submicrogram sensitivity is
desirable. It has been shown (l) that the light-pipe used in the Digilab
GC-IR accessory is too short (5 cm in length) and too wide (6 mm in diameter)
to permit the optimum GC-IR sensitivity to be achieved. The same calcula-
tions showed that a longer, narrower light-pipe would give increased sensitiv-
ity by providing a substantially increased optical path without seriously
attenuating the radiation by reflection losses. A light-pipe of approximate-
ly the optimum dimensions for GC-IR (300 x h x k mm) is sold by Norcon In-
struments Inc. (S. Norwalk, CT)» and the first phase of this project involved
interfacing this light-pipe to an FTS-l^l- spectrometer.
A different source from the nichrome wire source used on the FTS-14
spectrometer was also proposed for use with this GC-IR system. Nichrome
wire sources may not be operated regularly at a temperature much above dull
red heat, since at higher temperatures their lifetime is much reduced. The
decision was made to use a Nernst glower,since sources of this type operate
at the highest temperatures of all common infrared sources.
-------
SECTION II
CONCLUSIONS
A sensitive interface between a gas chromatograph and a rapid scanning
Fourier transform infrared spectrometer was developed. The improved
sensitivity was derived by modifications on the standard Digilab FTS-l*)-
Fourier transform spectrometer, viz.
(i) replacement of the existing nichrome wire sources of
the FTS-11!- with a modified Nernst glower source to
permit operation at higher temperatures than is
possible with nichrome wire;
(ii) installation of a light-pipe gas-cell of dimensions
300 x k x k mm using two specially designed aspherical
mirrors to focus the beam into the light-pipe and
refocus the beam emerging from the light-pipe onto
a 2 mm square detector.
With these modifications, identifiable spectra from submicrogram
quantities of strong infrared absorbers could be obtained without trapping
the sample.
When the triglycine sulfate (TGS) detector being used for the measure-
ment was replaced by an MCT detector, some improvement in GC-IR sensitivity
was demonstrated, but the principal limitation to the sensitivity of the
system changed from detector noise to digitization noise.
Longer and narrower light-pipes are needed for single beam GC-IR
measurements using the sensitive MCT detectors to circumment digitization
noise limitations. Since work along this line had been under intensive
investigation at the U. S. Environmental Protection Agency's laboratory at
Athens, Georgia, it was not pursued further in this project although it
constituted the second part of the proposal.
Feasibility studies showed that the dual beam Fourier transform
infrared (DB-FT-IR) technique is uniquely suited for the elimination of
digitization noise problems that are found in FT-ffi when a sensitive
detector is used for the measurement of absorption spectra with an optically
efficient sampling configuration. It is fundamentally a different approach
and holds promise for further improvement of the sensitivity of the GC-IR
interface. A complete optical system for GC-IR using this technique was
designed and is currently being built.
-------
SECTION III
RECOMMENDATIONS
The experimental part of this project was largely limited to the
relatively low sensitivity single-beam method for GC-IR. Substantially
increased sensitivity should be achieved through the use of dual-beam
Fourier transform infrared spectroscopy. With this method using light-pipes
of similar dimensions to the ones used in the single-beam experiment, it is
possible to utilize a liquid nitrogen cooled mercury cadmium telluride
detector without limiting the sensitivity of GC-IR by digitization noise.
It is recommended that such a system should be constructed and tested.
One potential method of improving the automation and, to a lesser
extent, the sensitivity of GC-IR is the development of an efficient trigger
to initiate and terminate data collection.
It should also be noted that if the concentration of the analyte in
any GC peak could be increased, for example in the same fashion as sample
concentrators for the interface between a gas chromatograph and a mass
spectrometer (GC-MS), a further increase in GC-IR sensitivity, by as much
as a factor of four, could be obtained.
-------
SECTION IV
EXPERIMENTAL
An optical bench was constructed with the following components.
Source. A modified Nernst glower (137 type source from Perkin-Elmer Corp-
oration, Norwalk, CT) was used. The resistance of this source is very low
at ambient temperature so that the source could not be heated directly using
any d.c. power supply available for this project, since the current required
to heat the source caused an overload for the power supply. To get around
this problem the source was initially heated by a.c. (using a Variac to
control the voltage). After the source had reached red heat, a 6 v.d.c.
power supply (Lambda Electronics, Melville, N.Y.) was switched on in place of
the a.c. supply. A 3" focal length 45° off-axis paraboloidal mirror (Special
Optics, Little Falls, N.J.) was used to collect the radiation emitted by this
source and produce an approximately collimated beam of radiation 2" in
diameter as input into the interferometer. Both the source and the
collimating mirror were mounted on a baseplate which butted directly up
against the interferometer.
Interferometer. A standard Digilab Model 296 Michelson interferometer was
used. The optical bench was kept close to the optical head of the FTS-14
spectrometer (from which the interferometer had been removed), and the
position of the electronic controller was not changed. The cables from the
controller to the interferometer were long enough that no extension cabling
was necessary for the system to be used in this configuration*
Focusing Paraboloid. The exit beam from the interferometer was brought to a
3-mm diameter focus using an identical off-axis paraboloid to the one mounted
in the source unit. The entrance to the light-pipe was at this focus.
Light-pipe and oven. A Norcon Instrument light-pipe and oven assembly was
used. The light-pipes were 30 cm long and square (4-mm side) in cross-
section. These light-pipes were constructed using four identical rectangular
strips of glaas with one side gold-coated. They were mounted in an oven, the
maximum temperature of which was 250°C. Only one of the two light-pipes in
this assembly was used in the first phase of this project. The effluent gas
from the gas chromatograph was passed down a flexible heated tube (Wilks
Scientific Corporation, S. Norwalk, CT) to the light-pipe. All exposed
surfaces of the delivery system were heated and wrapped in asbestos tape.
The temperature was monitored at several points to ensure no cold-spots were
present.
Focusing ellipsoid. The beam emerging from the light-pipe was focused onto
the detector using a 45° off-axis section of an ellipsoidal mirror whose two
focal lengths were 1.6" and 5.6". The diameter of the refocused beam was
2—mm.
-------
Detector. A 2-mm square triglycine sulfate (TGS) detector was generally
used for the measurements. This detector could be replaced by an MCT
detector, the element of which was also 2-mm square.
With these optics the transmittance between the interferometer and the
detector was 20$ to 25$ enabling low noise spectra to be measured using a
TGS detector with just a few seconds measurement time.
The typical experimental procedure was as follows. The effluent from
the gas chromatograph was monitored using a thermal conductivity detector
and passed through the light-pipe continuously (via the heated transfer
line). Interferograms were signal-averaged during the time that each peak
was present in the light-pipe, and interferograms from successive GC peaks
were stored in sequential arrays in the data system of the spectrometer.
At the end of the chromatogram, each interferogram was recalled and trans-
formed (generally using double-precision software), and the resulting
single-beam spectrum was ratioed against a stored low-noise background
spectrum of the empty cell. The resulting ratio-recorded spectrum could
be output in a linear transmittance or linear absorbance format.
-------
SECTION V
RESULTS
Early calculations (l) showed that the mini mum amount of a strongly
absorbing material (e.g. salicylaldehyde) from which a recognizable
spectrum (i.e. at least five bands with a signal-to-noise ratio greater
than 2:l) could be obtained with this system was approximately 800 nanograms
(ng). Figure 1 shows that this estimate was quite accurate.
It was found that the use of a Nernst glower source allowed spectra
to be measured at about twice the sensitivity attainable with a nichrome
wire source below 2000 cm"1. However above 2000 cm"1 the emissivity of the
Nernst glower falls off quite rapidly; and in spite of the fact that the
Nernst glower can be operated at a higher temperature than the nichrome wire
source, the spectral energy density of this source is actually slightly
lower than that of the nichrome wire about 3000 cm"1. Thus the noise-level
of transmittance spectra was substantially lower (by about a factor of five)
in the fingerprint region of the spectrum, below 2000 cm"1, than in the C-H
and 0-H stretching region between 3800 and 2800 cm"1.
The feasibility of using an MCT detector with this optical arrangement
was also studied. In order to avoid "clipping" the interferogram (2), it
was found that the source had to be operated below red heat. Even under
these far from optimum conditions, a recognizable spectrum of approximately
200 ng of salicylaldehyde could be obtained below 2000 cm x. Since the
response of the MCT detector falls off above 2000 cm"1 and because the
temperature of the source was so low, the noise around 3000 cm"1 in ratio-
recorded spectra was very high.
-------
LU
O
QQ
CC
O
GO
QQ
2.0
INSTRUMENTAL
ARTIFACT
I8<
ITU v I
800 ng '
i i . i . i . i .
in i
FREQUENCY (CM ')
•i
1
800
Figure 1. On-line GC-IR spectrum of 2 (jg and 800 ng of
salicylaldehyde measured using a TCS detector.
-------
SECTION VI
APPLICATIONS
As part of the feasibility studies to determine detection limits of
this system, a series of chlorinated pesticides was chromatographed and
their GC-IR spectra were measured. During this work it was found that one
of these pesticides, Endrin, decomposed on the column. We have been able
to study the decomposition products and go some way towards determining
the reaction kinetics by applying scaled absorbance subtraction routines
(3) on the GC-IR spectra. This work is not yet complete and will be
continued after the funding period for this grant is over.
Another study which was made using the system constructed in the
first phase of this project concerned the identification and quantification
of the products of the pyrolysis or combustion of certain polymers and
polymer composites. In this work (k) the gaseous species produced after a
small (l-2 mg) sample of polymer was pyrolyzed or burned were flushed
through the light-pipe with a stream of helium or air. Interferograms were
signal-aver aged over short periods of time and stored in the data system.
The process was repeated for as long as the pyrolysis was taking place, and
at the end of the experiment each interferogram was recalled and transformed
in precisely the same manner as a typical GC-IR experiment. This rapid
infrared method for studying the products and kinetics of pyrolysis and
combustion reactions has several advantages over the gas chromatographic
methods used previously, especially when reactive inorganic species (such as
HCl) are formed. Whereas using GC methods it is only possible to follow the
production of two or three species, by this infrared technique as many as
ten species could be followed simultaneously.
8
-------
SECTION VII
DISCUSSION AND FUTURE WORK
The second part of the proposal for this project was for a study of
the effect of using longer, narrower light-pipes than the Norcon light-pipes
used in the first phase of the project. In this case, the absorbing path
would be increased but reflection losses would decrease the overall trans-
mittance of the system. When the transmittance of the system is reduced
below about lOfo, an MCT detector can be used without running into the problems
of digitization noise (5)« However before this phase of the project was
started, Azarraga (6) had already made a study of the feasibility of using
optics of this type. His results showed detection limits close to those
predicted in the proposal for this project. It was therefore believed that
there was no point in repeating this approach, and work was started on another
method for further increasing the sensitivity of GC-IR measurements. This
system involved the use of dual-beam Fourier transform infrared (DB-FT-IR)
techniques. To recognize the reasons for adopting this approach, the limita-
tions of current GC-IR systems will first be discussed.
The system developed in the first phase of this project probably came
within a factor of two from meeting the lowest detection limits possible
with a state-of-the-art FT-IR spectrometer and a TGS detector. The light-
pipe provided a fairly long absorbing path without seriously attenuating the
beam due to reflection losses. Therefore a fairly high proportion of the
energy leaving the interferometer reaches the detector. If an MCT detector
is interfaced to this system with the source at its maximum temperature, the
interferogram would become digitization noise limited, so that the increase
in sensitivity gained through the use of a more sensitive detector is less
than the ratio of the D*'s of the two detectors. By using longer, narrower
light-pipes the energy reaching the detector may be attenuated to the point
that an MCT detector can be used without the problem of digitization noise.
If one could use an MCT detector with light-pipes of high transmittance
but without encountering the problems of digitization noise, it is apparent
that a further increase in GC-IR sensitivity could be achieved. The
application of DB-FT-IR techniques should allow this goal to be realized.
In DB-FT-IR (?)» use is made of the fact that the two beams emerging from an
interferometer are exactly 180° out-of-phase. For an ideal interferometer,
if these two complementary beams traverse optically identical paths and are
passed onto the same detector, no net a.c. interferogram would result. If a
sample of low absorbance is present in one beam and not the other, the
interferogram is due" solely to the sample absorption bands, and the smaller
is the quantity of sample present in the beam, the smaller is the
interferogram. (This is, of course, the reverse of the more conventional
-------
single-beam FT-IR methods, where the smaller the amount of sample in the
"beam, the greater is the intensity of the interferogram). Thus for DB-FT-IR,
a sensitive detector can be used to measure the spectra of samples at low
concentration using an MCT detector without encountering the problems of
digitization noise.
The final phase of this project was spent designing a dual-beam optical
system for GC-IR using the high transmittance Norcon light-pipes used in the
first phase of this study. In our final design, see Figure 2, the input
beam to the interferometer is slightly skewed, so that both output beams can
be collected and passed through the two parallel light-pipes in the Norcon
oven assembly. The beams emerging from the light-pipes are then focused on
a single MCT detector. After this system was designed, all components and
mirrors were ordered well before the completion date for this project.
However delays in the delivery of the two off-axis ellipsoidal mirrors
required to focus the two beams onto the detector have meant that we have not
yet been able to construct and test this system.
10
-------
J
DIGILAB
MODEL 296
INTERFEROMETER
IR SOURCE
I
OFF-AXIS
PARABOLOIDAL
MIRROR
i
DETECTOR
OFF-AXIS
ELLIPSOIDAL MIRROR
Figure 2. Dual-beam FT-IR optical configuration for
high sensitivity GC-IR.
11
-------
SECTION VIII
REFERENCES
1. Griffiths, P.R. Optimization of Parameters for On-line GC-IR Using
an FTS-l*)- Spectrometer. Ohio University, Athens, OH. (Presented at the
Pittsburgh Conference on Analytical Chemistry and Applied Spectroscopy.
March 5,
2. Mark, H. and M. J. D. Low. Background Elimination in Spectra
Generated by Fourier Transform Spectrometers. Appl. Spectrosc.
25:605-608. November 1971.
3. Kbenig, J. L. Application of Fourier Transform Infrared Spectroscopy
to Chemical Systems. Appl. Spectrosc. 22:293-308. July 1975.
4. Liebman, S. A., D. H. Ahlstrom and P. R. Griffiths. On-line Fourier
Transform Infrared Analysis of Pyrolysis and Combustion Products.
Appl. Spectrosc. 30:355-357. May 1976.
5. Griffiths, P. R. Sampling the Interferogram. In: Chemical Infrared
Fourier Transform Spectroscopy. New York, Wiley- Inter science, 1975?
P.6V75.
6. Azarraga, L. V. Improved Sensitivity of On- the- fly GCIR Spectroscopy.
U.S. Environmental Protection Agency, Athens, GA. (Presented at
the Pittsburgh Conference on Analytical Chemistry and Applied
Spectroscopy. March k, 1976.)
7. Griffiths, P. R. Dual-beam Fourier Transform Spectroscopy.
In: Chemical Infrared Fourier Transform Spectroscopy. New York,
Wiley- Interscience, 1975- p.l71-l80.
12
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing}
1. REPORT NO.
EPA-600/4-76-061
2.
3. RECIPIENT'S ACCESSION-NO.
4. TITLE AND SUBTITLE
On-Line Measurement of the Infrared Spectra of Gas
Chromatographic Eluents
5. REPORT DATE
December 1976 issuing date
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
Dr. Peter R. Griffiths
8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORG \NIZATION NAME AND ADDRESS
Ohio University
Department of Chemistry
Athens, Ohio 45701
10. PROGRAM ELEMENT NO.
ISA 027
11. CONTRACT/GRANT NO.
R 803517-01-0
12. SPONSORING AGENCY NAME AND ADDRESS
Environmental Research Laboratory - Athens, GA
Office of Research and Development
U.S. Environmental Protection Agency
Athens, Georgia 30601
13. TYPE OF REPORT AND PERIOD COVERED
Final, 1/20/75-1/19/76
14. SPONSORING AGENCY CODE
EPA/600/01
15. SUPPLEMENTARY NOTES
16. ABSTRACT
Techniques for increasing the sensitivity of the interface between a gas
chromatograph and a rapid-scanning Fourier transform infrared spectrometer (GC-IR)
have been developed. A single-beam system, in which a triglycine sulfate (TGS)
detector is used to measure the interferogram, has been designed and constructed.
Identifiable infrared spectra of submicrogram quantities eluting from a gas
chromatograph have been measured without trapping the sample using this system. A
double-beam configuration for GC-IR has also been designed so that a cooled mercury
cadmium telluride detector can be used to further decrease the detection limits
without limiting the sensitivity by digitization noise. Each of these systems
necessitates the use of light-pipes with relatively long (30 cm) absorbing paths but
low reflection losses.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
. b.lDENTIFIERS/OPEN ENDED TERMS
c. COS AT I Field/Group
Gas chromatography - Infrared Spectrometry
Fourier Transform Infrared Spectroscopy,
Infrared vapor spectra.
GC-IR
GC-FT-IR
07C
18. DISTRIBUTION STATEMENT
Release unlimited
19. SECURITY CLASS (ThisReport)
Unclassified (same)
21. NO. OF PAGES
21
20. SECURITY CLASS (This page)
Unclassified (same)
22. PRICE
EPA Form 2220-1 (9-73)
13
ft U. S. GOVERNMENT PRINTING OFFICE: 1977-757-056/553't Region No. 5-11
------- |