*To whom correspondence should be sent.

Resonance enhanced multiphoton ionization (REMPI) has been used to quantitatively detect benzene in an Ar DC plasma. The plasma reactor was fabricated from a 5 cm diameter pyrex tube that was about 30 cm long. The reactor also possessed two plasma electrodes that were separated by 3 cm. The plasma was generated at 160 mTorr pressure under the continuous flow of Ar that was seeded with 100-1000 ppm of benzene. Ions generated by REMPI were detected using a separate electrode placed at the edge of the plasma zone. We were able to obtain well-resolved and unambiguous benzene REMPI spectra in the 250-265 nm laser wavelength under a broad range of plasma conditions. The spectra obtained was well resolved, exhibited sharp spectral features and was in complete agreement with the results reported in literature under non-plasma conditions. Measured REMPI signal intensities also were linearly proportional to the benzene concentration, suggesting that REMPI can be a useful quantitative plasma diagnostic.

Optogalvanic effect is the transient change in the electrical impedance of the plasma caused by the absorption of pulsed radiation by a plasma moiety. This change can be either due to direct perturbation of the equilibrium ion/electron density by photoionization,^[8] or due to the transient redistribution of the charged species as a result of a pressure wave created by the photoacoustic effect.^[9] Optogalvanic spectroscopy has been applied for the detection of elements which have large densities of excited states, such as actinides, lanthanides and some transition elements^[8]. Previously rotational states of H₂ were also determined in a hydrogen DC plasma using an ArF excimer laser ^[10]. Laser induced optogalvanic signals, that were attributed to 2+1 resonance enhanced multiphoton ionization of the abundant H₂ gas, were measured as perturbations in the potential across the discharge. As a consequence of this measurement technique a large DC noise was encountered, thereby limiting the applicability of this approach to trace plasma species. In related studies, attempts were made to measure species in inductively coupled plasmas (ICP) using laser enhanced ionization (LEI) ^[11,12]. However, the high densities of ions and electrons associated with ICP overwhelmed the signals generated by LEI, necessitating the turning off of the plasma to achieve acceptable signal to noise ratios.

Here we report a novel application of resonance enhanced multiphoton ionization (REMPI) for the detection of a broad range of neutral moieties, i.e. molecules and radical species, in glow discharge plasmas. The technique exploits pulsed UV lasers to selectively photoionize neutral species followed by the measurement of the resulting photoions or photoelectrons by a dedicated detection electrode placed at the edge, i.e. the perimeter, of the plasma. We are not aware of any prior work in which the REMPI technique was used for the determination of trace species concentrations in glow discharge plasmas. As we shall demonstrate below, excellent REMPI signals can be obtained in glow discharge plasmas under a wide range of operating conditions, thereby opening up the possibility that the technique can be used in a broad range of plasma processing applications.

The REMPI technique is based on the fact that when the laser frequency is tuned to a real intermediate electronic state of a molecule or a radical, the cross section for ionization is significantly enhanced. On the other hand, when the laser wavelength is not tuned to a real electronic state, the probability for photoionization is much smaller.^[13] Therein lies the strength of REMPI, in which a specific molecule or radical can be selectively ionized with very high efficiency using a suitable laser frequency, while avoiding the simultaneous photoionization of other species.

where P^* is a real electronic excited state, P⁺ is the photoion and e^- is the photoelectron. By varying the photon energy, which can be accomplished using a tunable laser, the ionization spectrum of the target species P can be mapped to determine a suitable laser frequency that can be used to exclusively ionize it without simultaneously ionizing other species in the mixture. In plasmas, it is also possible that photo-excitation and electron impact processes may be coupled to induce ionization. There are several advantages of the REMPI approach. First, REMPI is inherently a high spectral resolution technique in which absorption features of any species can be determined with high precision. Second, species are ionized from a selected vibrational level of an electronically excited state, thereby allowing for the specific photoionization of target species. The latter can be used to distinguish between isomers because of their different electronic structures. Third, the REMPI process can also be repeated to sequentially detect different products (using different laser frequencies) and thus can allow the determination of a multitude of species in the plasma. Fourth, REMPI is a high sensitivity technique with real-time detection of species at low parts per billion ^[14] and high parts per trillion ^[15] already demonstrated in our laboratories and elsewhere.^[16] Finally, REMPI can be applied to a wider spectrum of polyatomic species than LIF, both molecules and radicals, thereby holding the promise to be a versatile plasma diagnostics tool.

In Figure 1 a sketch of the experimental system is shown. The plasma reactor was fabricated from a 5 cm diameter pyrex tube that was about 30 cm long. The reactor was equipped with fittings for gas inlet and outlet, and for the placement of electrodes. Plasma electrodes were

Figure 1. Sketch of the Experimental System

made from 1.8 cm diameter stainless steel plates, and were spaced about 3 cm apart during the experiments. One of the electrodes was connected to a DC power supply (Stanford Research Systems, PS 350) and the other was grounded. The plasma reactor was also fitted with optical access windows and a port for the signal detection. Optical windows were made from UV transparent silica (Acton Research, MA) and were connected to the reactor walls using o-ring seals. The detection electrode was made from a 0.2 mm diameter stainless steel wire that was insulated by pyrex glass. The pulsed, tunable UV beam was obtained from a solid state laser (Opotek, Carlsbad, CA, 5 ns pulse width). The laser power was uniform at about 1.4 mJ/pulse over the entire 248-300 nm range. A pulse rate of 10 Hz was used in all the experiments. The laser beam was focused to a waist of about 2 mm using a series of long focal length lenses. The laser beam passed through the center of the plasma reactor as indicated in Figure 1. Signal detection was accomplished by a digital oscilloscope.(Hewlett-Packard 54540C).

Plasma was generated under the continuous flow of Ar (99.99%, Matheson, CA) seeded with about 1000 ppm benzene at a pressure of 160 mTorr. We selected benzene as the test species to demonstrate the utility of REMPI in a plasma environment because its REMPI features are well known.^[16] Argon plasma was observed to strike when -300 V was applied to the electrode. It was visually shaped as a rounded cylinder sandwiched between the electrodes. Increasing the DC bias voltage increased the glow and appeared to stabilize the plasma until about -400 V. Beyond -400 V, the plasma exhibited transient behavior and became anchored at multiple locations. Between -300 V and -350 V, the plasma possessed a steady glow and remained securely anchored between the electrodes over a period of hours. Thus the undertaking of extensive measurements were possible. During the experiments, the total current passing through the plasma was about 0.05 mA.

Experiments were conducted first to determine if the benzene REMPI spectra can be obtained in the presence of plasma. For this the DC bias was set first at -320 V, where a stable plasma was present. This plasma was also characterized using the same electrostatic probe used in REMPI measurements. The plasma parameters were determined using a commercial system (Hiden Analytical, Warrington, UK) based on the orbital motion limited analysis. The mean energy and the number density of the electrons were measured to be 3 eV and 8x10⁸/cc, respectively. These parameters are consistent with a low pressure glow discharge plasmas^[3] and also suggest an approximate plasma sheath thickness of 0.25 mm. This value is about a factor of 15 larger than the detection probe diameter, thereby supporting the validity of the orbital motion limited analysis. The laser wavelength was then scanned from 250 nm to 265 nm in 0.1 nm increments. In this wavelength range, benzene is known to have significant 1+1 REMPI features^[16], while argon has none. In this and subsequent experiments, the detection electrode tip was kept at about 1.5 cm from the center of the reactor, very close to the edge of the visible part of the plasma (see Figure 1). The spectra obtained under these conditions are presented in Figure 2. As evident from this figure, we were able to obtain well-resolved and unambiguous benzene REMPI spectra over the 250-265 nm at the -320 V bias voltage. Benzene exhibited sharp REMPI maxima both at 253.8 and 259.7 nm, respectively, in complete agreement with the results reported in literature in the absence of plasma, which is also shown in Figure 2 ^[16]. These results suggest that REMPI, together with optogalvanic detection probe should have a place in plasma technology as a diagnostics tool.

We also measured the benzene REMPI signals when the tip of the detection electrode was placed at different positions in the plasma. No changes in signal intensity were observed

Figure 2. Benzene REMPI Spectra in Ar Plasma and in Air

when the detection probe was moved over a 1cm distance relative to the laser beam. Evidently the plasma efficiently directs the photoelectrons and/or photoions to the detection electrode, regardless of the position of the probe relative to the laser beam. This is in contrast to the extreme sensitivity of the measured REMPI signals to the distance between the probe tip and the laser beam in non-plasma gases.^[17] Since the position of the detection electrode is unimportant, as long as it is in contact with the plasma, it can be placed at the edge of the discharge thereby minimally perturbing the plasma interior where laser beam passes. This renders REMPI electrode detection essentially a non-intrusive technique for the determination of species concentration profiles in plasmas.

In order to assess the utility of the present technique as a quantitative tool, REMPI signals must be correlated with the concentration of benzene in the plasma as well. This was accomplished by preparing different Ar/benzene calibration mixtures and then measuring the resulting REMPI signals as a function of benzene concentration. In Figure 3 the results of such an undertaking are presented at a laser wavelength of 253.8 nm, where benzene has been shown to give a prominent REMPI signal (see Figure 2). As evident from this figure, the measured REMPI signals are well correlated with the benzene levels in the calibration mixtures over an order of magnitude change in concentration. However, at high concentrations,

i.e. above 600 ppm, some signal saturation was observed. Possible causes of this signal saturation include the limitations of laser power, plasma conductivity and detection electronics.

Figure 3. Benzene REMPI Signal Strength as function of concentration

Figure 4. Benzene REMPI Signal at 253.8 nm as function of DC Bias Voltage

figure, the benzene REMPI signals were remarkably uniform over a broad range of plasma voltages, i.e. -200 to -400 V. The insensitivity of benzene signals to plasma density suggests that the ionization process must be strictly due to photon absorption and that electron impact events are not involved in the ionization process. This result also suggests that REMPI can be a useful diagnostic for the determination of neutral species in plasmas under a broad range of conditions. Voltages above -400V, the glow changed into an arc. Therefore, the plasma was not stable and led to unreliable measurements. The critical lower voltage -200 V approximately corresponds to the minimum in the Paschen curve for argon where the glow discharge potential is equal to the breakdown potential. Further reductions in applied potential leads to the loss of luminosity and to a Townsend discharge^[2], concomitant with a decrease in gas conductivity and REMPI signal intensity.

In the current experimental configuration, the benzene ions were generated along the entire cylindrical laser beam path. Consequently, this arrangement allows the determination of spatially resolved concentration profiles only in two dimensions, i.e. on the plane perpendicular to the laser beam. Although this may be adequate for many applications where the conditions are axisymmetric, the ability to determine three dimensionally resolved concentration profiles clearly would be desirable. This can readily be accomplished by crossing two laser beams inside the plasma, where the laser beams can be derived either from the same laser using a beam splitter, or from two or more different lasers. These arrangements would then allow for the implementation of 1- or 2-color REMPI schemes, respectively.^[14]

It should also be noted that although the application of REMPI has been demonstrated with benzene, a stable molecule, the technique is equally applicable for the detection of radical species as well. In fact, due to their lower ionization potentials, radical species are particularly well suited for REMPI.^[18]

We would like to thank Professor. F. Chen, Dr. J.D. Evans, Dr. P. Gillis, and Dr. D. Blackwell for useful discussions on plasmas and plasma reactors and Dr. V. Zengin for his help with the use of the tunable solid state laser system.

1. Murarka, S.P., Peckerar, M.C., "Electronic Materials Science and Technology", Academic Press, New York 1989.

2. Grill, A., "Cold Plasma in Materials Fabrication", IEEE Press, New York, 1994.

3. Manos, D.M., Flamm, D.L., Eds. "Plasma Etching", Academic Press, San Diego, CA1989.

4. Auciello, O., Flamm, D.L., Eds. "Plasma Diagnostics", Academic Press, Washington, DC 1989.

5. Vasile,M.J., Dylla, H.F., "Mass Spectrometry of Plasmas", in "Plasma Diagnostics", Auciello,M.J., Flamm, D.L., Eds., Academic Press, Washington, DC 1989.

6. Hershkowitz, N., "How Langmuir Probes Work", in "Plasma Diagnostics", Auciello,M.J., Flamm, D.L., Eds., Academic Press, Washington, DC 1989; Coburn,J.W., Chen, M.J., J. App. Physics, 51, 3134, 1980.

7. Donnely, V.M., "Optical Diagnostic Technique for Low Pressure Plasmas and Plasma Processing", in "Plasma Diagnostics", Auciello,M.J., Flamm, D.L., Eds., Academic Press, Washington, DC 1989; Manos, D.M., Dylla, H.F., "Diagnostics of Plasmas for Material Processing, in "Plasma Etching", Manos, D.M., Flamm, D.L., Eds., Academic Press, San Diego, CA 1989.

8. Levesque,S., Gagne, J.M., Babin, F., "Optogalvanic Photoionization Spectroscopy", J. Phys. B: At. Mol. Opt. Phys., 30, 1331-1346 (1997).

9. Kumar,D., Zinn, R.R., Armstrong, T.D., McGlynn, S.P., "Optogalvanic Effect as a Probe of Plasma Processes", J. Phys. Chem., 99, 7530-7536 (1995).

10. Hitchcock, L.M., Reck, G.P., Rothe, E.W., Tung, C.C., "Diagnostics in H₂ discharges using a tunable 193 nm excimer laser", paper presented at the Microwave and Particle Beam Sources and Directed Energy Concepts meeting 16-20 January 1989, Los Angeles, CA.

11. Turk, G.C., Yu, L., Watters,R.L., Travis, J.C., "Laser-Induced Ionization of Atoms in a Power-Modulated Inductively Coupled Plasma"" Appl. Spectroscopy, 46, 1217-1222 (1992).

12. Ng, K.C., Angebranndt, M.J., Winefordner, J.D., "Laser-Enhanced Ionization Spectroscopy in an Entended Inductively Coupled Plasma", Anal. Chem., 62, 2506-2509 (1990).

13. Lubman,D., Li, L.,"Resonant Two-Photon Ionization Spectroscopy of Biological Molecules in Supersoni Jets Volatilized by Pulsed Laser Desportion", in Lasers and Mass Spectrometry,D.M. Lubman, Ed., Oxford Univ. Press, New York 1990.

14. Gittins,C.M., Castaldi,M.J., Senkan,S.M., Rohlfing,E.A,"Real-Time Quantitative Analysis of Combustion Generated Polycyclic Aromatic Hydrocarbons by Resonance Enhanced Multiphoton Ionization Time of Flight Mass Spectrometry", Anal. Chem., 69, 286 (1997).

15. Senkan, S.M., Castaldi,M.J., "Real-time, Ultrasensitive Monitoring of Air Toxics by Laser Photoionization Time of Flight Mass Spectrometry", J. Air and Waste Mgmgt. Assoc., 48, 77-81 (1998).

16. Marshall, A., Ledingham, K.W.D., Singhai, R.P., "Trace Detection of Benzene Vapor in a Simple Ion Chamber Employing Laser Induced Resonance Enhanced Multiphoton Ionization in the 246-265 nm Wavelength Region", Analyst, 120, 2069-2073 (1995).

17. Cool,T.A., "Quantitative measurement of NO density by resonance three-photon ionization", App. Optics, 23, 1559 (1984).

18. Lias,S.G., Bartmess, J.E., Liebman, J.F., Holmes, J.L., Levin, R.D., Mallard, W.G., "Gas-Phase Ion and Neutral Thermochemistry", J. Phys. Chem. Ref. Data, 17, Supplement No.1, 1988.