Vinyl Chloride Detection Using Carbon Monoxide and Carbon Dioxide Infrared Lasers Samuel M. Freund Optical Physics Division, National Bureau of Standards, Washington,DC 20234
Daniel M. Sweger Analytical Chemistry Division, National Bureau of Standards, Washington, DC 20234
Vinyl chloride, the 22nd ranking chemical in total tonnage produced in the United States (1, 2), and starting material for many industrial processes has recently been recognized as a potential carcinogen, (3-6), and perhaps a mutagen (7-9). Although safe levels for continued or long term exposure to this compound have not yet been determined, it appears that sensitivities a t or below 1 ppm C2H3C1 in air will be required (10-12). It is the purpose of this paper t o describe a sensitive, selective detection procedure based upon Stark modulated absorption of infrared laser radiation. Consideration is also given t o the feasibility of incorporating the technique presented into a prototype C2H3Cl quantitative analyzer. Vinyl chloride has received extensive attention among laser researchers due to its use in passive Q-switching (13) and Stark modulation (14, 15) of several C 0 2 laser lines, as well as for generation of optically pumped laser emission in the far-infrared (16). In both applications, use is made of several near coincidences between COz laser lines and molecular transitions of C2HsCl. It is these near coincidences and the polar nature of the vinyl chloride that suggest t h e possibility of applying a form of electric field perturbation spectroscopy (17) as the basis for the development of a vinyl chloride detector. By superimposing a static electric field of the appropriate magnitude and sign, a n electric field sweep, and a modulating field, it is possible both t o Stark shift several vinyl chloride infrared transitions into exact coincidence with the laser and t o use phase-sensitive detection to observe the absorption signal. Vinyl chloride is a slightly asymmetric rotor (18-22) with a 1.4 D electric dipole moment (23). Although this gives rise t o a small second-order Stark effect, due to several near coincidences with laser lines it is only necessary to vary the electric field over approximately 5000 V/cm in order to tune through a CzH3Cl Doppler broadened transition and t o observe the entire absorption peak. Thus, the usual necessity of very low operating pressures (millitorr range) due t o requisite high electric fields (typically as high as 1OO,OOO V/cm) is no longer present. Several near coincidences with CO and C02 laser emissions other than those used for the C2H3C1 signal vs. concentration measurements are also reported. A catalog of the laser lines used and the electric fields required t o bring about exact coincidence will “fingerprint” the vinyl chloride molecule with two orders of magnitude greater resolution than is usually achieved using commercial infrared grating spectrophotometers. T h e high resolving power of t h e method and discrimination against molecules with different Stark effects should eliminate almost all difficulties with interfering species ( 2 4 ) .Moreover, sensitivities attainable with laser spectrometers (25) are not fully realized a t the 1 ppm CzH3C1 level in the experiments described, and possible improvements will be discussed further. 930
ANALYTICAL CHEMISTRY, VOL. 47, NO. 6, MAY 1975
EXPERIMENTAL The extracavity absorption cell contains two 40-cm long stainless steel Stark electrodes spaced 1 mm apart. Optics are arranged so that the electric field on the plates is perpendicular to the electric field of the laser. The experimental procedure employed is similar to that described in other recent Stark spectroscopy experiments (26, 27) with important differences occurring in the amplitude of the modulation, the total gas pressure, and the signal processing. Since vinyl chloride has a small second-order Stark effect, required modulation voltages were as high as 150 volts peak-to-peak at a frequency of 1-2 kHz. A 250-volt sawtooth voltage was also applied t o sweep through particular CZH3C1 absorption lines for display purposes. Superimposed dc voltages necessary to bring the vinyl chloride resonances into coincidence with the laser lines were of the order of several hundred volts. Therefore, the total voltage appearing on the plates never exceeded 750 volts and total pressures of up to 7 Torr were maintained in the cell for long periods of time without any evidence of electrical breakdown. At the total pressures and CzH3C1 concentrations investigated, the line width is primarily due to Doppler broadening (28). In a cell designed to operate at higher gas pressures, it should be possible to increase the sample pressure to 25 Torr without important pressure broadening effects. Data were taken simply by setting the CO or COn laser to a line with a near coincidence with vinyl chloride, setting the bias voltage on the Stark electrodes to tune the CzH3Cl energy levels into exact coincidence with the laser, and measuring the peak-to-peak height of the derivative signal, as displayed on an X-Y recorder, corresponding to the resonant absorption of the laser power from a single traversal of the absorption cell. The derivative response occurs as a result of the sine wave modulation applied to the Stark electrodes for the purpose of phase-sensitive detection of the resulting signal. The output from the lock-in amplifier was fed into a signal averager which permits averaging time constants of from 5 to 500 sec. For each vinyl chloride sample, data were accumulated for at least three time constants, and the storage of the averager was sent t o an X-Y recorder in order t o obtain a hard copy. Twelve high pressure samples of C2H3C1, ranging from 1 to 1000 ppm in air, were prepared by the Air Pollution Analysis Section of the Analytical Chemistry Division at the National Bureau of Standards. Concentrations determined by gas chromatographic techniques agreed well in most cases with those calculated from dilution techniques used in the preparation of the samples. Eight series of absorption measurements, four with each laser, were made at pressures ranging from 1.5 to 4 Torr. A Bourdon tube gauge (020 Torr) was used for measuring the pressure in the Stark cell and was sensitive enough to permit measurement of pressures to within a few percent. While each series of measurements did not include all the samples, each did include the 27.8-ppm sample, which was therefore used to normalize the data in order to compare runs. The CO laser power was approximately 0.2 watt and that for the COz laser about 0.5 watt. Detectors for the laser radiations were Au:Ge and PbSnTe, respectively. The linearity of signal processing was checked at all stages to ensure that the final signal read out onto the X-Y recorder bore a proportional relationship to the concentration of vinyl chloride.
RESULTS Table I gives a partial survey of the near coincidences of vinyl chloride with COn laser emission lines. These coinci-
Table I. Partial Survey of the Near Coincidences of Vinyl Chloride with COz Laser Emission Lines Signal
CO2 laser transition
(001-100 band)
Bias (Vlcmf
Power ("atts)'
Comments
(peak-to-peak volts)c
two weak lines nothing observed
Daniel M. Sweger Analytical Chemistry Division, National Bureau of Standards, Washington, DC 20234
Vinyl chloride, the 22nd ranking chemical in total tonnage produced in the United States (1, 2), and starting material for many industrial processes has recently been recognized as a potential carcinogen, (3-6), and perhaps a mutagen (7-9). Although safe levels for continued or long term exposure to this compound have not yet been determined, it appears that sensitivities a t or below 1 ppm C2H3C1 in air will be required (10-12). It is the purpose of this paper t o describe a sensitive, selective detection procedure based upon Stark modulated absorption of infrared laser radiation. Consideration is also given t o the feasibility of incorporating the technique presented into a prototype C2H3Cl quantitative analyzer. Vinyl chloride has received extensive attention among laser researchers due to its use in passive Q-switching (13) and Stark modulation (14, 15) of several C 0 2 laser lines, as well as for generation of optically pumped laser emission in the far-infrared (16). In both applications, use is made of several near coincidences between COz laser lines and molecular transitions of C2HsCl. It is these near coincidences and the polar nature of the vinyl chloride that suggest t h e possibility of applying a form of electric field perturbation spectroscopy (17) as the basis for the development of a vinyl chloride detector. By superimposing a static electric field of the appropriate magnitude and sign, a n electric field sweep, and a modulating field, it is possible both t o Stark shift several vinyl chloride infrared transitions into exact coincidence with the laser and t o use phase-sensitive detection to observe the absorption signal. Vinyl chloride is a slightly asymmetric rotor (18-22) with a 1.4 D electric dipole moment (23). Although this gives rise t o a small second-order Stark effect, due to several near coincidences with laser lines it is only necessary to vary the electric field over approximately 5000 V/cm in order to tune through a CzH3Cl Doppler broadened transition and t o observe the entire absorption peak. Thus, the usual necessity of very low operating pressures (millitorr range) due t o requisite high electric fields (typically as high as 1OO,OOO V/cm) is no longer present. Several near coincidences with CO and C02 laser emissions other than those used for the C2H3C1 signal vs. concentration measurements are also reported. A catalog of the laser lines used and the electric fields required t o bring about exact coincidence will “fingerprint” the vinyl chloride molecule with two orders of magnitude greater resolution than is usually achieved using commercial infrared grating spectrophotometers. T h e high resolving power of t h e method and discrimination against molecules with different Stark effects should eliminate almost all difficulties with interfering species ( 2 4 ) .Moreover, sensitivities attainable with laser spectrometers (25) are not fully realized a t the 1 ppm CzH3C1 level in the experiments described, and possible improvements will be discussed further. 930
ANALYTICAL CHEMISTRY, VOL. 47, NO. 6, MAY 1975
EXPERIMENTAL The extracavity absorption cell contains two 40-cm long stainless steel Stark electrodes spaced 1 mm apart. Optics are arranged so that the electric field on the plates is perpendicular to the electric field of the laser. The experimental procedure employed is similar to that described in other recent Stark spectroscopy experiments (26, 27) with important differences occurring in the amplitude of the modulation, the total gas pressure, and the signal processing. Since vinyl chloride has a small second-order Stark effect, required modulation voltages were as high as 150 volts peak-to-peak at a frequency of 1-2 kHz. A 250-volt sawtooth voltage was also applied t o sweep through particular CZH3C1 absorption lines for display purposes. Superimposed dc voltages necessary to bring the vinyl chloride resonances into coincidence with the laser lines were of the order of several hundred volts. Therefore, the total voltage appearing on the plates never exceeded 750 volts and total pressures of up to 7 Torr were maintained in the cell for long periods of time without any evidence of electrical breakdown. At the total pressures and CzH3C1 concentrations investigated, the line width is primarily due to Doppler broadening (28). In a cell designed to operate at higher gas pressures, it should be possible to increase the sample pressure to 25 Torr without important pressure broadening effects. Data were taken simply by setting the CO or COn laser to a line with a near coincidence with vinyl chloride, setting the bias voltage on the Stark electrodes to tune the CzH3Cl energy levels into exact coincidence with the laser, and measuring the peak-to-peak height of the derivative signal, as displayed on an X-Y recorder, corresponding to the resonant absorption of the laser power from a single traversal of the absorption cell. The derivative response occurs as a result of the sine wave modulation applied to the Stark electrodes for the purpose of phase-sensitive detection of the resulting signal. The output from the lock-in amplifier was fed into a signal averager which permits averaging time constants of from 5 to 500 sec. For each vinyl chloride sample, data were accumulated for at least three time constants, and the storage of the averager was sent t o an X-Y recorder in order t o obtain a hard copy. Twelve high pressure samples of C2H3C1, ranging from 1 to 1000 ppm in air, were prepared by the Air Pollution Analysis Section of the Analytical Chemistry Division at the National Bureau of Standards. Concentrations determined by gas chromatographic techniques agreed well in most cases with those calculated from dilution techniques used in the preparation of the samples. Eight series of absorption measurements, four with each laser, were made at pressures ranging from 1.5 to 4 Torr. A Bourdon tube gauge (020 Torr) was used for measuring the pressure in the Stark cell and was sensitive enough to permit measurement of pressures to within a few percent. While each series of measurements did not include all the samples, each did include the 27.8-ppm sample, which was therefore used to normalize the data in order to compare runs. The CO laser power was approximately 0.2 watt and that for the COz laser about 0.5 watt. Detectors for the laser radiations were Au:Ge and PbSnTe, respectively. The linearity of signal processing was checked at all stages to ensure that the final signal read out onto the X-Y recorder bore a proportional relationship to the concentration of vinyl chloride.
RESULTS Table I gives a partial survey of the near coincidences of vinyl chloride with COn laser emission lines. These coinci-
Table I. Partial Survey of the Near Coincidences of Vinyl Chloride with COz Laser Emission Lines Signal
CO2 laser transition
(001-100 band)
Bias (Vlcmf
Power ("atts)'
Comments
(peak-to-peak volts)c
two weak lines nothing observed
