Plasma Generation In An Organic Molecular Gas By An .

FIG. 2. Plasma column produced in TMAEFIG. 3. Electron density profile along the direction of laser beam,vapor.cm) when the Th1AE pressure is 500 mTorr, with theintensity decreasing from the window towards the center.The light column exists all the way to the other side of thechamber, a total length of 50 cm, if the TMAE vaporpressure is reduced to 100 mTorr or lower.A. Photon absorptioncross-sectionmeasurementEquations ( 1) and (2) provide general descriptions ofthe photon flux in a gas medium where the ionization potential of the molecules is less than the photon energy. Ifthe photon flux is a short pulse of width 7, during that timethe electron density will not, be affected by diffusion, recombination, or attachment processes,and the percentageof ionization will remain small, i.e., n&t,,,; then Eqs. ( 1)and (2) can be simplified towhere n,o;-k,is the absorption coeffiicient andite nf,,pir.(6)The solution of (5) isn,(z) n,ONexp( -n,,q23,(7)where R,(O) - r(O) z,,a;r. The absorption cross section ofTM.4E vapor can then be determined after measurementof the electron density spatial profile along the direction oflaser propagation.To measure the electron density spatial distribution,we used a 2.4 mm-diam by O.l-mm-thick tantalum diskLangmuir probe, which can be moved axially or transverseto the direction of the laser beam. The electron collectingsurfaces of the disk are parallel to the direction of the laserbeam to minimize t.he cross-sectional area in the beampath. The threshold of ablat.ion of metals by a UV laser isthe order of lOsW/cm’, which is at least two orders higherthan the laser pulse power in our experimental chamber.l7Since tantalum has a work function of 3.76 eV and thephoton energy of the 193 nm laser is 6.4 eV, photoelectricemission from the laser beam striking the edge of the tantalum disk has to be considered. We checked this effectexperimentally as follows. The probe was positioned in the1 4781J. Appl. Phys., Vol. 73, No. IO, 15 May 1993laser beam close to the window and biased negatively( -20 V) through a resistor ( 1.2 kS1). The voltage on theresistor was monitored by connecting it directly to the highimpedance input of a digital storage scope which has amaximum sensitivity of 2 mV/division. When the chamberis filled with air from 20 mTorr to atmospheric pressure,and TMAE is not introduced, the recorded signal is about5 mV which comes from the radio frequency interferencegenerated by the laser discharge circuitry. No noticeablechangeis observedwhether we block the laser beam or not.Therefore, the maximum current that could be producedby photoelectric emission is 4.2 PA which is 2-3 orderssmaller than the ion saturation current measured.The probe is then biased negatively to measure thepeak ion saturation current in the plasma as a function ofdistance from the window. The current is proportional toelectron density generatedat the position which is in turnproportional to the photon flux according to Eq. (6). Themeasurementis conducted at a vapor pressure of 0.5 Torrat a room temperature of 2 1.5 “C corresponding to theTMAE molecular density of n,k 1.65 1016/cm3. Themaximum electron density measured is 5 10’3/cm3, sone(nm. Figure 3 shows the peak electron density along theaxial direction. The exponential fit of the experimental datagives the absorption coefficient ki O.l8/cm, a photonmean free path of I 5.4 cm, and a photon absorption crosssection of CT, 1.1ho.2 x lo-l7 cm’. This value is about30% of that measured by Holroyd et al. l4 There are manyfactors that could lead to the difference in results. Thelinewidth of the laser output is 1 nm for our laser according to the manufacturer’s supplied information. In Holroyd et al.‘s experiment, a synchrotron light beam is used,which needs a monochromator to filter out the unwantedspectrum, and no information about the linewidth is given.In our experiment, the photon intensity is measured alongthe beam path at intervals of 1 cm for a total length of 30cm, and an exponential function is then fitted to the data toobtain the absorption coefficient. This should lead toimproved accuracy compared to fitting an exponentialfunction to two data points as Holroyd et al. have done.Also, the condensation of TMAE on their windows couldhave contributed to the photon loss in their experiment.The condensation does not affect our measurement,Y. S. Zhang and J. E. Scharer4781Downloaded 03 Oct 2008 to 128.104.1.219. Redistribution subject to AIP license or copyright; see http://jap.aip.org/jap/copyright.jsp

3.53-2.5m‘-01.025a, 0.80-a,I0.55FIG. 4. Radial electron density profile.FIG. 5. Electron temperature vs time.since the photon intensity is effectively measured insidethe chamber.To determine the ionization efficiency, we need toknow the total number of electrons generatedand the laserpulse energy. The total number of electrons generatedin ahalf spaceis* sn,(O)exp( --n,a )dz,18)J0where the s is the laser beam cross-sectional area. Theionization efficiency is thereforeiv, (9)where the NP is the total number of photons in the pulsewhich can be determind from pulse energy measurement.In this experiment, 7 is determined to be 0.52.In the above calculation, we have assumed that theelectron density is uniform inside the beam and zero outside. The actual electron density radial profile has beenexamined. A cylindrical laser beam is obtained by passingthe rectangular beam through a ‘I-mm-diam iris. TheLangmuir probe is moved radially from the beam center;the peak electron density along the radial direction isshown in Fig. 4. It shows that the electrons are concentrated in the region illuminated by the laser beam, and theradial diffusion is insignificant. This is consistent with thevisual observation that the light emission column coincideswith the laser beam.B. ElectrontemperaturemeasurementWhen a photon having an energy E, ionizes a TMAEmolecule, the free electron is expected to have a kineticenergy of E,,- EP, where EP is the ionization potential.Hence, right after the laser pulse, the electron velocity distribution is highly non Maxwellian. The Langmuir probetemperature measurementwill deviate from that predictedby probe theory for a non-Maxwellian plasma. But theTMAE molecule (which has a formula weight of 200) hasa large electron collision cross section. If we estimate theelectron neutral collision frequency Y to be - 10gp, wherepis the pressure in Torr, a reasonableestimation of the re4782J. Apple Phys., Vol. 73, No. 10, 15 May 1993laxation time is on the order of 10/v, i.e., 10 ns. Therefore,we can consider that the electron density distribution function will relax to a Maxwellian almost immediately afterthe UV pulse. Figure 5 is an electron temperature plotversus time at z l cm and z 5 cm, where t O corresponds to the leading edge of the laser pulse. The electrontemperatures measuredat t 0.2 ps are 1.0 eV for z 1 cmand 0.92 eV for z 5 cm.C. Electron-ionmeasurementrecombinationcoefficientThe plasma decay processcan be described by Eq. (2)by setting I’ 0an,a%,z ad-p-apt-aa,tb.(10)TMAE has a molecular weight of 200. The ambipolar diffusion coefficient of TMAE is estimated to be 1.65X 10’cm”/s and the plasma decay time constant is estimated tobe 0.59 lo-” s at room temperature. Therefore, the diffusion can be neglectedon a time scale shorter than a ms.In TMAE vapor, the impurity TMU has about the sameorder of magnitude of electron capture cross section asoxygen.16The impurity TM0 electron capture cross section is several times larger.16Rewick et al’s analysis16ofcommercial TMAE shows that the concentration of TM0is 0.04% and that of ThlU is 0.31%. We can, therefore,estimate the effective attachment rate using the availabledata on electron attachment to oxygen.” The electron attachment frequency to neutral oxygen molecules is % 2x 10-13ptimfor an electron attachment cross section of3 x lo-z1 cm’, where tZimis the impurity density of molecules which have a large electron capture cross section.The oxygen in our vacuum system is calculated from theleak rate of our vacuum system to be less than 0.2%. Assuming the impurities total 1% in the system, the corresponding decay time constant is calculated to be on theorder of 10 ms. So the initial decay of electron density isdominated by electron-ion recombination, and it obeys thewell-known simple equationY. S. Zhang and J. E. Scharer4782Downloaded 03 Oct 2008 to 128.104.1.219. Redistribution subject to AIP license or copyright; see http://jap.aip.org/jap/copyright.jsp

crowavereceiverN2-TMAE3.5FIG. 8. Schematic of wave scattering in TMAEt(W)plasma tilled waveguide.The reflected wave is measured by a diode detectorthrough a directional coupler, which is shown in Fig. 7.The peak reflected power is 0.33 mW. For comparison, the8.7 GHz microwave signal reflected from an X-band waven(t) !I0(11)ha# 1’guide (25 mm X 13 mm) placed at the same position as theplasma column is measured to be 0.45 mW. The resultwhere pztois the initial electron density.demonstrates that reflection due to the TMAE plasma canThe initial electron decay is measuredwith a Langmuirbe very effective. In the future we plan to produce a sheetprobe. Figure 6 shows the inverse of electron density versusuhravioletbeam to create a high density gradient TMAEtime. The good fit of the experimental data to a linearplasmasheetwith a width several times the wavelength offunction shows that the decay is, indeed, recombination.themicrowaves.The plasma sheet should be able to effecThe recombinat.ioncoefficient.is determined to be 5.4 f O.5tivelyreflectmicrowaves.If the UV laser beam is opticallyX lo-” cm’/s.rotated, a rotating mirror reflector is produced whichwould be attractive for radar scanning applications at miD. Microwave scattering measurementcrowave or millimeter wavelengths.The microwave reflection experimental setup is shownFigure 8 shows another setup for the microwavein Fig. 1. The microwave beam is incident on the plasmacolumn in the direction transverse to the laser beam. TheIOcenter of the antenna horn is 18 cm from the end plate. Theantenna is a 7.5 cm x 9.5 cm rectangular horn placed just0.0outside of the glass chamber. The dist.ante from the antenna to the plasma column is about 7 cm. The microwavepower ( 10 mW) is incident on the narrow side (0.7 cm) ofthe plasma column. The microwave frequency is adjustedto allow the reflected microwave power from the chamberwalls to be in phase with the peak reflection from theplasma column, and is experimentally determined to be 8.7GHz. The peak electron density in the region in front ofthe antenna is measured by a Langmuir probe to be theorder of lOI3 cnle3 at a pressure of 250 mTorr.FIG. 6. Inverse of electron density vs time.FIG. 7. Microwave power reflected from TMAE4783plasma column.J. Appl. Phys., Vol. 73, No. IO, 15 May 1993FIG. 9. Microwave transmission through a waveguide filled with a transient TMAE plasma generated by an UV pulse: (a) high initial density,(b) lower initial density.Y. S. Zhang and J. E. Scharer4783Downloaded 03 Oct 2008 to 128.104.1.219. Redistribution subject to AIP license or copyright; see http://jap.aip.org/jap/copyright.jsp

electron density decay at high initial densities can be utilized to rotate the reflecting plasma sheet very rapidly.IV. SUMMARYU!. )FIG. 10. Electron density decay in TMAE plasma by recombination fordifferent initial densities.plasma interaction experiment to demonstratewave shielding effects. In this arrangement, the laser beam is injectedthrough a magic-teeinto an X-band waveguide.The magictee also couples microwaves into the same waveguidewhere the plasma generated by the laser fills the wholewaveguide.Figure 9 (a) shows that the receivedsignal leveldrops by 38 dB indicating a total cutoff due to the highelectron density produced by the 4 mJ laser pulse; thesignal quickly recovers to a -8 dB level in less than 2 psand then decays at a much slower rate. Figure 9(b) illustrates an 8 dB signal level reduction when the laser pulseenergy is reduced to an estimated value of 0.02 mJ. Sincethe initial density generatedis much lower in this case,thesignal level recovers slowly indicating a slow electron density decay. These features are in good agreement with thenature of recombination loss in the TMAE plasma.To get a quantitative idea of electron recombinationloss in the TMAE plasma, we plot the electron densitydecay given by Eq. ( 11) with the measured electron-ionrecombination coefficient in Fig. 10, for five different initialelectron densities. The high initial electron density of 10”drops to l/e of its initial value in about 0.05 ps; and thelower initial density of 10”’decaysto l/e of its initial valuein about 10 ps. If one uses repetitive pulses to sustain theelectron density above certain value, e.g., l/e of the initialdensity, it can be shown that the power required is proportional to the square of the initial electron density. Therefore, it is more economical to use the afterglow TMAEplasma for the purpose of wave absorption and shielding inthe lower frequency range, e.g., 3 GHz and below, since itrequires much less power. On the other hand, the TMAEplasma is more suitable for rotating mirror applications athigh frequencies, e.g., 10 GHz and above, since the fast4784J. Appl. Phys., Vol. 73, No. IO, 15 May 1393We have investigated TMAE plasma generated by aUV laser beam. The photon absorption cross section for193 nm UV radiation in TMAE has been determined fromelectron density profile measurements. The experimentshows that the plasma decay in TMAE is primarily viaelectron-ion recombination. The electron-ion recombination coefficient has been measured. Effective microwavereflection and shielding effects are demonstrated by scattering measurements.A set of two partial differential equations are derived to model the one photon ionization process in a low ionization potential gas. The experimentalresults show that this model is a good description of theTMAE plasma generationby the UV ( 193 nm) laser. Withthis model and given parameters, the numerical solutioncan help in the design and control of TMAE or other lowionization energy gas plasmaswhich have attractive properties for applications of microwave reflection, shielding,and absorption.ACKNOWLEDGMENTSWe wish to thank Professor James C. Weisshaar forthe use of his laser equipment, and Robert J. Noll, OwenEldridge, and Binshen Meng for valuable technical help.This work was supported by AFOSR Grant 89-0353B.‘K. R. Stalder, R. J. Vidmar, and D, J. Eckstrom, J. Appl. Phys. 72,5089 ( 19923.‘W. W. Destler, J. E. DeGrange, H. H. Fleischmann, J. Rodgers, and 2.Segalov, J. Appl. Phys. 69, 6313 (1991).3S. P. Kuo and Y. S. Zhang, Phys. Fluids B 2, 667 (1990).‘J. E. Scharer, 0. C Eldridge, S. F. Chang, Y. S. Zhang, M. Bettenhausen, and N. T. Lam, IEEE Trans. PIasma Sci. (to be published).‘P. A. Kossey, R. A. Shanny, and E. C. Field, AGARD ConferenceProceedings, Bergen, Norway, May 28-31, 1990, No. 485, paper 17A.%. P. Kuo, Y. S, Zhang, R. J. Barker, and P. Kossey, AGARD Conference Proceedings, Bergen, Norway, May 28-31, 1990, NO. 485, paper 18B.‘W. M. Manheimer, IEEE Trans. Plasma Sci. 19, 1228 (1991).s A. E. Robson, R. L. Morgan, and R A. Meger (unpublished).‘R. J. Vidmar, IEEE Trans. Plasma Sci. 18, 733 (1990).“J. R. Woodworth, T. A. Green, and Frost, .I. Appl. Phys. 57 (1985).“Y. Nakato, M. Or&i, A. Egawa, and H. Tsubomum, Chem. Phys. Lett.15 June, 615 (1971).“D. F. Anderson, IEEE Trans. Nucl. Sci. NS-28, 842 [ 198 I ).t3G. Melchart, G. Charpak, and F. Sauli, IEEE Trans. Nucl. Sci. ( 1990).iJR. A. Holroyd, J. M. Preses, C. L. Woody, and R. A. Johnson, Nucl.Instrum. Methods in Phys. Research A 261,440 (1987).“K. R. Stakler and D. J. Eckstrom, J. Appl. Phys. 72, (1992).“R. T. Rewick and M. L. Schumacher, Analyt. Chem. 60,2095 (1988).“D. V. Gaidarenko, A. G. Leonov, and I. V. Novobrantsev, Sov. Tech.Phys. Lett. 15, 112 (1989).‘*S. C. Brown, Basic Data of Plasma Physics (M.I.T., Cambridge Press,1966), p. 199.Y. S. Zhang and J. E. Scharer4784Downloaded 03 Oct 2008 to 128.104.1.219. Redistribution subject to AIP license or copyright; see http://jap.aip.org/jap/copyright.jsp

tetrakis (dimethylamino ) ethylene (TMAE) seeded in a buffer gas to produce a collisional plasma.” TMAE has one of the lowest ionization potentials known today (5.36 eV).‘ ” Its vapor can be ionized efficiently with photons i