Session QP1 - Poster Session.
POSTER session, Thursday morning, November 18
Grand III, The Westin Seattle

[QP1.01] The Effect of Flow Rate on Competitive Growth Mechanisms in Nitrogen-Methane-Hydrogen Diamond Thin Film Plasma Deposition

Flow rates in plasma CVD reactors determine the radical resonance time and may greatly influence the characteristics of the deposited films, especially if the film growth is catalyzed by any particular radical species. In these investigations, polycrystalline diamond thin films are grown for 8 hours in a microwave plasma CVD reactor on single crystal (100) p-type silicon substrates, using 1methane/hydrogen ratios, and nitrogen levels between 0 and 1500 ppm. The addition of ppm of nitrogen has been shown in previous investigations to catalytically affect both morphology and texture, at a flow rate of 200-217 sccm in the Michigan State University wide-area reactor geometry. Flow rates between 50 and 400 sccm are investigated in the present research. Plasma properties during deposition are analyzed by Optical Emission Spectroscopy, optical pyrometry, and by Langmuir probes. The resulting diamond film properties are analyzed by Tapping Modeä atomic force microscopy, X-ray diffraction texture analysis and micro-Raman.

[QP1.02] Environmental Behavior of Laser Ablation Induced Plasma on Diamond Films

Pulsed laser ablation has been recognized to be a promising method for micromachining and patterning of diamond films. Due to the characteristics of the short or ultrashort pulsed laser irradiation, plasma formation is inevitable upon the ablation processing. Formation and expansion of the plasma result in thermal damages, which limit the resolution of the process. Behavior of observed plasma was significantly affected by altering the ionization potential of the processing environment. Thermal damages were dependent on the behavior of the plasma in the various gas processing environments. In our experiments, an excimer (KrF) laser with an optical imaging system is used for the pulsed laser ablation. Diagnostics for the environmental behavior of the plasma are performed to characterize formation and expansion of the ablation induced plasma. Images of the plasma will be optically filtered and captured by charge coupled device (CCD) camera. Spectral information of the plasma will be obtained by an optical multichannel spectroscope.

[QP1.03] Athermal Annealing of Silicon

Current semiconductor annealing methods are based on thermal processes which are accompanied by diffusion that degrades the definition of device features or causes other problems. This will be a serious obstacle for the production of next-generation ultra-high density, low power semiconductor devices. Experiments underway at NRL utilize a new annealing method which is much faster than thermal annealing and does not depend upon thermal energy (J. Grun, et al)., Phys. Rev. Letters \bf78, 1584 (1997).. A 10 J, 30 nsec, 1.053 nm wavelength laser pulse is focussed to approximately 1 mm diameter on a silicon sample. Acoustic and shock waves propagate from the impact region, which deposit mechanical energy into the material and anneal the silicon. Experimental results will be presented on annealing neutron-transmutation-doped (NTD) and ion implanted silicon samples with impurity concentrations from 1 \times 10^15-3 \times 10^20/cm^3.

[QP1.04] Material Testing at Pulsed High Voltages

High power microwave tubes typically all share the same characteristic in that large electric field stresses, both pulsed and rf, can be present in the tube. These large field stresses can cause a variety of problems such as spurious beam emission, arcing, and other similar problems. We review experiments to explore the behavior of various materials (stainless steel, molybdenum coated stainless steel, molybdenum, and titanium) exposed to a high voltage pulse of 1 microsecond duration and field stresses up to 260 kV/cm. Further, the evolved gas content of the vacuum is reviewed. The voltage source used for the testing is the Cathode Test Bed, a pulse forming network based pulser with a 100 Ohm impedance. We review the results of tests on these different materials and show the field stresses at which breakdown of the material occurred.

[QP1.05] Materials Processing with Intense Pulsed Ion Beams*

Materials applications are being investigated on the 700 kV RHEPP-1 facility at Sandia National Laboratories. Surface modification for property improvement is possible in the fluence range 1-5 J/cm^2, with ablation and thin-film synthesis at 5-20 J/cm^2 fluences. Differences from previous efforts include selectability of accelerating ions (H, He, C, N, Ne, Ar, and Xe), and repetitive pulsing of the MAP (Magnetically Confined Anode Plasma) gas-breakdown ion source. Surface modification using melt-resolidification cycles has led to improvement in hardness and corrosion resistance of various metals. Mixing of pre-applied thin-films into the bulk has led to even greater performance improvements. Characterization is ongoing to determine the microstructural basis for these improvements. We have characterized liquid-phase diffusion of implanted elements in Ti and Si during the power pulse. Experiments with Si device processing and polymer modification are also ongoing. Thin-films are being formed from graphite, YBCO, and ZnO targets for various applications, including hard-coatings and optical coatings. Surface topography, stoichiometry, and optical and infrared absorption measurements have been made.

*Sandia is a multiprogram laboratory operated by Sandia Corporation, a Lockheed Martin Co., under US DOE Contract DE-AC04-94AL85000.

[QP1.06] Spectroscopy Diagnostics for Helicon Plasmas

Measuring plasma parameters via non-intrusive diagnostics is a matter of necessity in steady-state helicon sources. Contrary to probes, spectroscopy measurements are only weakly affected by RF fields and do not contaminate the plasma. We discuss spectroscopic diagnostics designed to evaluate the different plasma parameters of the HELIX helicon source. The wide range in density that can generated by HELIX (for He, 10^10 to 10^13 cm^-3) makes it an ideal source to develop and validate diagnostics that can be used in RF plasmas. For He discharges, a spectroscopy technique based on the relative intensities of He I lines is used to measure T_e in the plasma. This diagnostic is based on the fact that the dependence on the electron energy of the excitation rate differs between singlet and triplet lines of the He atom. In addition, using an absolute calibration, the study of specific neutral and ion transitions can predict the population of the different excited levels and the ion density. For Ar plasmas, the possibility of using a line ratio technique to evaluate the electron temperature is investigated. An alternative technique using the ratio of bound-bound radiation versus continuous radiation is also presented. Finally, absolute intensity measurements of specific Ar I and Ar II transitions are used to predict the excited levels populations and the ion density in the plasma, respectively.

[QP1.07] Ion Temperature Anisotropies in Helicon Plasmas

Laser induced fluorescence measurements of the ion temperature in an argon helicon plasma indicate a substantial ion temperature anisotropy (perpendicular over parallel). The perpendicular ion temperature scales linearly with the applied magnetic field strength and also depends on the RF driving frequency and amplitude. We will present measurements of the ion temperature as a function of magnetic field, RF amplitude, and RF frequency. We will also present measurements of the electromagnetic fluctuation spectrum in the source. Preliminary measurements suggest that the RF driving wave may parametrically decay into waves that can couple to the ions, thus providing a path for RF energy directly into the ions.

[QP1.08] Helicon Studies at Auburn University

Auburn University currently operates a single helicon source, the Auburn Linear Experiment for Space Plasma Investigations (ALESPI). It is also in the process of constructing research facility for basic plasma studies incorporating a multi-helicon array for creating large-area high-density plasmas. ALESPI uses a 10 cm diameter antenna to create a 2 m long discharge in both argon and helium. Peak densities are of order 10^20m^-3. The Auburn Large-area Plasma Helicon Array (\alphalpha) consists of a 4m vacuum chamber with 13 magnetic field coils to create a maximum field of 0.2T on axis. The helicon array uses seven helical twist type helicon antenna surrounding quartz tubes of 10cm diameter, for a total diameter of about 30cm. We anticipate peak densities of order 10^20m^-3. Results from both ALESPI as well as initial results from \alphalpha will be presented. This work supported by the U.S. Department of Energy and the Auburn University College of Science and Math.

[QP1.09] Observation of resonant cavity modes and copropagating m = +1 and m = -1 helicons in a helicon discharge

Five distinct discharge modes are observed in a large-diameter, weakly-magnetized, short helicon source (13.5-cm diameter, 55 G, 62-cm length) excited by a double-saddle antenna. Each mode has been characterized by its plasma impedance, wave field structure, density distribution and floating potential structure. The lowest two discharge modes are capacitive (E) and inductive (H), while higher modes are helicon wave (W) modes. Successive wave modes correspond to helicon cavity resonances of the plasma-filled vacuum vessel, each with a well-defined wavevector, density and impedance. The first helicon discharge mode has m = ±1 azimuthal symmetry, as expected for a double-saddle antenna, while the second and third wave modes have m = 0 symmetry in the diffusion chamber. The relative amplitudes of the m = +1 and m = -1 modes has been determined by decomposing on-axis polarization data into left-handed and right-handed components. The amplitude of the m = +1 helicon is generally larger than that of m = -1, and the parallel wavenumber of m = +1 is greater than that of m = -1, in agreement with theory.

[QP1.10] Low Frequency Electrostatic Instability in a Helicon Discharge

The increase in plasma density with static magnetic field strength in helicon plasma discharge sources is well known. Recently, with light ions, the density has been observed to reach a maximum and saturate, or decrease, as the magnetic field strength is increased further. We find that the density maximum correlates with the onset of low frequency (1- 10 kHz) density fluctuations and an eventual turbulent fluctuation spectrum as the magnetic field is increased beyond the maximum. Quiescent, coherent, and turbulent regions of low frequency oscillations are observed in the ion saturation region of a RF compensated Langmuir probe at various magnetic field strengths. Azimuthally separated probes show rotation in density fluctuations in the electron diamagnetic drift direction slightly below the magnetic field corresponding to the plasma density maximum. Measured radial profiles reveal large radial potential and density gradients and off-axis fluctuation amplitude maximums. These results show the onset and growth of a low frequency drift or Kelvin-Helmholtz instability in the helicon discharge plasma.

[QP1.11] Detection of Trivelpiece-Gould Modes in Helicon Discharges

It has been suggested that the high RF power absorption efficiency in helicon discharges is due to the transfer of energy to electron cyclotron waves called Trivelpiece-Gould (TG) modes(K.P. Shamrai, V.P. Pavlenko, and V.B. Taranov, Plasma Phys. Control. Fusion 39), 505 (1997). To test this theory, measurements have been performed using magnetic and current density probes to detect TG waves.

The experiments are performed in a 10-cm diam chamber with B_0=20-55 Gauss, p=3 mTorr of argon, and p_rf=0.2-1kW @ 11 MHz. A two turn m=0 antenna is used to excite the helicon wave plasma.

Measurements are compared with numerical calculations which take into account radial and axial density gradients, antenna coupling, and ion mass. The results show clear evidence of the TG wave contribution to the radial magnetic field and current profiles. Details of computational and diagnostic methods will be discussed.

[QP1.12] Investigation of Magnetized ICP and low-B Helicon Discharges for Plasma Processing

The transition between ICP, magnetized ICP and helicon mode is investigated in a device consisting of a dome-shaped Pyrex top attached to a magnetic bucket. The antenna is a multi-turn loop of wire wrapped around the top. Bo is produced by a coil located near the antenna. Rtop=15cm \sim height of top, Rbucket \sim 18cm, bucket height\sim30cm, Np\sim 10^11cm^-3, Te\sim3eV, Prf\leq1.4kW, frf = 2-27.12MHz, Po=1-20mT, Bo\leq200G. A Bz-dot probe is used to measure radial Bz-field profiles. Bz decays exponentially with decreasing R at low Bo. Characteristic decay lengths of Bz (Lsd) are measured as Po is varied. Good agreement between experiment and theory is observed when the self-B-field produced by the antenna current is included. Variations of Lsd with Bo have been measured, and good agreement between experiment and theory is observed. Regions of enhanced Bz (``wings") are observed near the edge for Bo>0, accompanied by a bright``halo" near the antenna. As Bo increases, Lsd increases, and Bz in the interior increases. Coupling to a low-B helicon mode is expected when Lsd\simRtop. Under these conditions, the Bz-field behavior changes from evanescence to propagation.

[QP1.13] A Steady State Helicon Plasma Source

We are using a 27 MHz helicon plasma source to create steady state argon plasma for development of several novel diagnostics such as motional Stark effect diagnostic for low magnetic fields, 2D laser induced fluorescence (LIF) imaging of ion turbulence and for the magnetic nozzle experiment at PPPL. This plasma source can be operated under a wide range of conditions: RF power 0.3-2 kW, pressure 0.3-20 mTorr, magnetic field 1-3.5 kG. Langmuir probe measurements were used to get radial plasma density and electron temperature profile. The electron density at the center of the vacuum chamber ranged from 10^18-10^20 m^-3 and the electron temperature was T_e \approx 5 eV. Using LIF diagnostic the average perpendicular ion temperature was found to be T_i \approx 0.8 eV. Future plans include detailed radial profile measurements of the ion and electron temperature and electron density.

[QP1.14] Plasma Production by m = 0 Standing Helicon Waves.

The m=0 standing helicon waves (SHWs) are used to produce high-density Ar plasmas of the order of 10^13 cm^-3 in a bounded cylindrical plasma column [1]. Axial mode number N=1, 3, and 5 (N is the number of half-wavelengths in the plasma column) SHWs are strongly excited at plasma length L=10-30 cm with the rf power \leq 3.5 kW and external static magnetic field <1 kG. Off-axis and parabolic density profiles, which depend on the axial mode of SHW, are observed. Standing wave ratio decreases with L, and travelling helicon waves dominate the plasma production at L>30 cm. The measured results are compared with a two-dimensional wave calculation. [1] M. Nisoa, Y. Sakawa, and T. Shoji, Jpn. J. Appl. Phys. 38, L777 (1999).

[QP1.15] Ion Heating in the HELIX helicon plasma source

At the 1997 DPP meeting, first results of ion heating in a helicon plasma were presented. The most efficient heating perpendicular to the magnetic field occurred at a semi-integer value of the ion cyclotron frequency at a single neutral pressure. At the 1998 DPP meeting, data showing that the ion heating was anisotropic with respect to the magnetic field and more efficient at lower neutral pressures were presented. The perpendicular ion heating efficiency was 800value of the ion cyclotron frequency. These results suggested that the heating may have been the result of ion Bernstein waves. We present new results that indicate the ion heating is due to ion cyclotron waves. The occurrence of the maximum heating efficiency at a semi-integer value of the ion cyclotron frequency is most likely due to the plasma conditions at a particular magnetic filed strength and not resonant wave-particle effects. Using a technique developed by Skiff et al. [1987] for identifying electrostatic waves through laser induced fluorescent (LIF) measurements of the first order perturbation to the ion velocity distribution function, the data confirms the presence of an ion cyclotron wave generated by the ion heating antenna. Further measurements of the ion density fluctuations using LIF show that the wave amplitude decreases with increasing ion heating efficiency.

[QP1.16] Plasma Production by Helicon Waves in VHF Band

High-density plasmas of the order of n_p \simeq 10^13 \rmcm^-3 are produced in helicon discharge using VHF band. Two types of antenna are used to investigate the role of RF current and voltage; one is a single loop antenna\footnote[1]Y. Sakawa, N. Koshikawa, and T. Shoji, Appl. Phys. Lett. \bf69, 12 (1996), in which azimuthal RF current excites the m=0 helicon waves, the other is a capacitive antenna, in which only RF voltage is applied. Operating conditions were frequency = 144 MHz, power P_ rf \leq 1 kW, B_0 = 1 kG and 5 mTorr of Ar. With the single-loop antenna, the denstiy-jump\footnotemark[1] occures at P_rf = 0.3 kW. After the desity-jump, helicon waves contribute to the plasma production, and n_p reaches 8 \times 10^12 \rmcm^-3 at P_rf = 0.7 kW. Whereas with the capcitive antenna, the density-jump does not occure, but the radial position of the maximum n_p varies from r \simeq 0 to r \simeq 2 cm at P_\rmrf \simeq 0.4 kW. At P_\rmrf \leq 0.3 kW, the dependence of n_p on P_rf is the same as that by the single loop antenna. Therefore, RF voltage plays an important role in the plasma production before the density-jump.

[QP1.17] Helicon wave density limit

Using a simple model for a cold plasma in a cylindrical waveguide where the plasma is surrounded by an extensive vacuum layer, It has been found that there is a range of parameters where two propagating modes coalesce as the density increases, beyond which no modes exist for real k_z. Comparing the observed density limit for a helicon plasma, the theoretical density limit lies remarkably close to that observed. Following this coalescense as a function of ømega/ømega_ci\gg1, the model shows that a 5% increase in this ratio leads to a density decrease of about 15% for the dimensions of the experiment, a trend which is in agreement with theory. The model assumes a uniform density out to radius a, vacuum to radius b where there is a conducting boundary. The wave is an m=3D+1 fast wave, and the bundary condition comes from Section 5.6 and problem 5.18 (with a typo corrected on page 318)(D.~G.~Swanson, Plasma Waves), (Academic Press, Boston, 1989).

abstract.

[QP1.18] LIF Diagnosis of Plasma Processing Sources

A variety of plasma sources are used in plasma processing. Laser-induced fluorescence (LIF) can be used to characterize the plasma ion population at different locations in some sources and processing chambers. LIF measurements of ion distributions in plasma processing sources will be presented. For example, in a hollow cathode dc argon ion source, the ion temperature is about a tenth of an eV. Convective speeds and plasma density profiles are under study. RF generated plasma sources are being studied for comparison. The ion distributions may be taken with inexpensive diode lasers described in: G.D. Severn, D.A. Edrich, and R. McWilliams, Rev. Sci. Instrum. 69 (1), 10 (1998).

[QP1.19] Characterization of a Cylindrical Corona Discharge Reactor

Temporal variation of current was measured in a pulsed corona discharge reactor for dissociation of NOx and SOx from the flue gas. Industrial scale reactor with height of 3 m and diameter of 0.5 m was made for dissociation of 500 liters of flue gas per hour with temperature of 80 - 120 degrees of Celcius. Reactor is made as a three-stage cyclinder-cylinder type with a rugged inner cylindrical electrode. Electric probes have been used to measure the current density variation with potential. Effects of gas temperature, flow rate, power input, and additives on the dissociation efficiency will be addressed.

[QP1.20] Reactor Geometries for Efficient Nitrogen Dissociation

Effective Nitrogen dissociation by plasma electrons requires operation in non-Maxwellian electron energy distribution function (EEDF) regimes, with reaction and ionization rates determined by local E/N values rather than the electron ``temperature". For such low pressures (0.05 - 0.5 Torr) and ionization fractions (10^-5 - 10^-3), self-consistent solutions of RF field and electron density profiles in N_2 are dominated by dissociative recombination in the interior, leading to hollow n_e profiles as the input power is increased. We show how annular antenna geometries for an inductive reactor can make the best use of the absorbed power for atomic nitrogen production by maintaining a higher electron density over more of the reactor volume. An Ar/N seeding scheme for actinometry is suggested to verify the dissociation fractions and excited state populations expected from the discharge model.

[QP1.21] Recent Resonant and Surface Wave Plasma Research at UC-Berkeley

Resonant and surface wave sustained plasma sources are a promising technology for next generation processing plasmas. Electrostatic theories show resonances in device impedance and admittance and are extended to fully electromagnetic models in order to accurately model large area sources. A resonant discharge experiment performed in [1] has been recreated in 1D3V PIC-MCC simulation and extensively analyzed. The simulation reveals new phenomenology including bunching of electrons in phase space and hystersis of the plasma and external circuit. We are extending the resonant discharge analysis in [2] to cylindrical and spherical geometries and comparing the results to 1D3V simulations. We have also applied an electrostatic surface wave model to explain analomous transport in magnetrons. Other recent work has focused on the development of high performance simulations in order to model the high density large area sources and to study mechanisms to couple circuits with plasma surface waves.

This work is supported by DOE DE-FG03-97ER5446, ONR N00173-98-1-G001 and the Hertz Foundation Fellowship Program.

[1] V.A. Godyak. Paper 3C9. ICOPS 1994. [2] V.A. Godyak. ``Steady-State Low-Pressure RF Discharges." Soviet Journal of Plasma Physics. Vol. 2. No. 1. Jan - Feb 1976.

[QP1.22] Investigation on Improvement Density Uniformity of Planar Type Surface Wave Plasma Device

The planar type Surface Wave Plasma (SWP) devices have enabled us to generate uniform, high density and large area plasmas without external magnetic fields.

Their applications have been considered to expand into various kinds of areas of the plasma processing.

The understanding of the discharge mechanism, specifically, the mechanism of uniform plasma production has still been one of the major research subjects.

On SWP device, it has been found that the uniformity of electron density distribution is improved by the placement of conductive plates (aluminum plates) on the dielectric window.

We have observed this uniformity improvement on this device, and have compared between the result of experiment with that of numerical simulation.

In the experiment, we have used electrostatic probes and loop antenna to measure the spatial profiles of the electron density and electric fields in the surface area.

In the simulation, we have used the Finite Difference Time Domain (FDTD) method for 2 dimensional configurations where the plasma is assumed to be cold electron plasma with a density gradient at the plasma surface.

The result of experiment shows that the uniformity of electron density distribution depends on the interval of aluminum plates.

And the result of simulation reasonably interprets that of experiment.

[QP1.23] Particle-in-cell simulation of two-dimensional axially symmmetric plasma processing tools

We have developed a two-dimensional axially symmetirc pariticle-in-cell (PIC) simulasion code for glow discharges. In this presentation, we focus on simulation of capacitively-coupled argon discharges. The PIC simulation consists of (1)charge assignment from particles to grid points, (2) determination of the electric field, (3) interpolation of the field at the position of each particle, and (4) integration of the equation of motion. Electlon-neutral and ion-neutral collisions are taken into account through the Monte-Carlo Collision (MCC) method. At the boundary, paticles are assumed to be reflected or absorbed with given probabilities and may cause secondary electron emission. In addition to detailed simulation methods that we have employed, preliminary simulation results for radio-frequency (RF) discharges with an external circuit model will be presented.

[QP1.24] Electrode-Size Effects on Plasma Sheath Expansion

If a negative high-voltage pulse is applied to a metal piece (target) immersed in plasma, a large volume of ion sheath is developed near the target. Plasma ions in this ion sheath are accelerated by the electrical potential and are implanted into the target surface, changing surface properties. This plasma process is called the plasma source ion implantation (PSII), which is a simple, cost-effective alternative to conventional ion implantation processes. It is essential for successful PSII process to correctly determine the sheath front in terms of the negative-voltage pulse profile. One of the fundamental issues in most plasma applications is also the plasma sheath appeared in the vicinity of plasma boundaries. This plasma sheath is either useful or troublesome, depending on the plasma applications. Therefore, it is again very important to correctly understand properties of the plasma sheath for various plasma applications. Therefore, influence of the electrode-size effects on plasma sheath expansion is investigated for a negative voltage at the cathode with maximum amplitude of V0, assuming that the cathode is a thin, flat, conducting, circular disk of radius R. It is shown that propagation of the sheath front is proportional to the one-third power of the combination, 2V0R/B. The propagation of sheath front is also proportional to the five-ninth power of time during the pulse rise and is proportional to the two-ninth power of time after the pulse rise. Experimental measurements have been carried out to verify theoretical predictions. The experimental data agree remarkably well with theoretical results in every aspects.

[QP1.25] Time-dependent equilibrium in the negative hydrogen ion rich plasma.

In the sheet plasma experiment, we showed that negative hydrogen ions existed in the following equilibrium space-dependently;^1) \begincenter e + e + H \leftrightarrow e + H^- \endcenter Where e is an electron and H is a hydrogen atom. Accordingly H^- means its negative ion. If the equilibrium (not thermal equilibrium) really existed, the equilibrium should be presented not only space-dependently but also time dependently. In the cold region of the sheet plasma, the mean free paths of both electrons and negative hydrogen ions were larger than 10cm. There electron density was less than one-tenth of negative hydrogen ion density and electrons were not considered to be a fluid anymore. We are going to present a data which may prove the equilibrium time-dependently. Therefore we conclude that a kind of equilibrium exists among negative hydrogen ions, hydrogen atoms and electrons in the sheet plasma. 1. K.Jimbo:J.Phys.Soc.Jpn.65, 2455 (1996).

[QP1.26] Cross-field Ion Transport in LAPPS (Large-Area Plasma Processing System)*

In LAPPS a substrate is placed next to a large flat plasma sheet (typical dimensions 1 m X 1 m X 1 cm) created by a magnetically-guided sheet electron beam. The flux of neutral radicals to the substrate depends only on the product of beam current density J_b and gas pressure P, while the flux of ions across the magnetic field to the substrate depends separately on the beam/substrate separation distance d and the magnetic field B, as well as on J_b and P; the ion bombardment energy can of course be controlled by the bias. It is possible to vary the ion bombardment from very strong to essentially nil, while separately choosing the neutral radical flux. We present initial theoretical calculations, based on a fluid model, showing the scaling of ion flux with each of the parameters for the case of insulating or conducting collectors, and for the cases of no bias, dc bias or rf bias. The theory also indicates that the ion flux should be quite uniform over the surface of the collector, with end effects playing a significant role only in the outer few cm. The theory will be compared with initial experimental results for electric current flow to negatively-biased collector plates. *Work sponsored by the Office of Naval Research

[QP1.27] Plasma and Plasma-Surface Diagnostics in LAPPS

NRL's Large Area Plasma Processing System (LAPPS) uses a magnetically confined sheet of high-energy electrons to ionize a neutral background gas, producing a high-density planar plasma for use in materials processing. The plasma is scalable to large areas (meters^2) and decoupled from the reactor geometry, allowing independent positioning and biasing of the processing stage. Currently, a pulsed electron beam is produced by a hollow cathode, confined by a 100-300 Gauss magnetic field, and injected into 10-500 mTorr of background gas producing plasmas with densities of 10^9-10^12 cm^-3. The results from temporally resolved Langmuir probe, microwave transmission/absorption, and optical emission experiments are presented for oxygen and noble gas mixtures. In situ mass and energy analysis of ion and neutral fluxes to the processing stage are also presented to understand the plasma-surface interactions. The relative species fluxes and ion energy distributions are correlated to stage positioning and biasing. Additional details concerning LAPPS are presented at this conference.(See papers by R.A. Meger and R.F. Fernsler, this conference.)

[QP1.28] Properties of Striations in Coplanar PDP cells

Although AC-Plasma Display Panels(AC-PDPs) are now on the market, many discharge-related phenomena have not been fully understood yet. Striations in AC-PDPs are one of them. Generation mechanism and characteristics of the striations have not been systematically investigated. Striations on a positive column have been observed at the beginning of several glow-discharge experiments. The nature of positive column striations was identified as the ion-plasma acoustic waves due to the presence of propagating perturbations. The striations observed on the electrode surface in AC-PDP without the positive column might be very different from the positive column striations. In this paper, we report the experimental observations of striations in various electrode geometries, including the actual coplanar AC-PDPs. In order to find out the dependence of striations on the electrode shapes and on the DC and AC discharge types, we performed the discharge experiments in parallel diode plates, in wedge-shaped electrodes having a wedge angle and in coplanar electrodes. The observed data indicate that the striations are related to ion charge waves of the wall generated by the self-sustained perturbations that originate from the force balance between the ion and the electron wall-charge clouds accumulated on the dielectric layers over the metal electrodes.

[QP1.29] Characterization of Plasma by Electric Probes in an AC High Pressure Plasma

Plasma density, temperature and potentials were measured by electric probes in an AC high pressure plasma. Co-planar electrodes are installed in order to simulate the plasma display panel(PDP) cell with applied voltage of 3 kV with frequency of 20 kHz, and gas pressure of 1 - 10 torr. Effects of pressure (collisionality), types of gases (mass), and thermalization on the probe analysis are studied. Applicability of this study to PDP cell will be addressed from the temporal and spatial variations pf plasma parameters.

[QP1.30] Influence of Driving Frequency on Capacitances in Surface Discharge AC PDP

In ac plasma display panel (AC-PDP), wall voltages have important role in lowering the sustaining voltage due to wall charges accumulated on the dielectric surfaces. It is of great importance to experimentally investigate influence of driving frequency on the wall charges and various capacitances in AC-PDP to determine the wall voltages, since they are dependent on the driving frequency. Thus, in this paper, we present quantitative experimental results of wall charges, voltages, as well as the capacitances in terms of driving frequency. The square pulse with risetime of 150 ns and duty ratio of 40 percent has been used in the experiment. It has been found that capacitances of surface discharge at the gas-filled region are nearly invariant to be about 0.3 pF/cell, while those for the dielectric region decrease significantly from 7.5 pF/cell to 0.8 pF/cell, as the driving frequency increases from 10 kHz to 200 kHz. It is also found that both the wall charge and wall voltage decrease from 34.5 pC/cell to 15.6 pC/cell and from 104.4 V to 76.5 V, respectively, as the driving frequency changes from 10 kHz to 200 kHz in the experiment.

[QP1.31] Study of MgO secondary electron emission coefficient and micro-plasmas for application to plasma display panels (PDP)*

Reduction of plasma cell size by ~40an essential step in bringing current flat panel television technology to meet true HDTV standards in the future. In order to reduce the cell size effectively while maintaining and/or improving the brightness, modeling techniques more sophisticated than simple fluid or time-consuming kinetic codes are vital. More importantly, any model has to be supported by experimentally measured plasma parameters such as density and temperature in order to accurately predict time dependent behavior. Note that the diagnosis of micro-plasmas is a challenging problem, as most conventional techniques are inapplicable. On behalf of this effort, we have constructed a single cell in which plasmas produced across ITO electrodes coated with MgO are studied. The measured secondary electron emission coefficient of MgO will be used to benchmark a new, time-efficient kinetic based code. This paper will also address a technique for a direct measurement of plasma density in a plasma cell comparable to the real PDP cell size.

[QP1.32] Electrical and VUV/Vis. Emission Characteristics of a Xe/He microdischarge

Commercially available flat panel displays operate with various gas compositions over a broad range of pressure. Few experimental studies have addressed the question of optimum gas mixture and total pressure for maximum luminance as well as discharge control for the minimization of sputtering damage. In this study, the vacuum emission and breakdown characteristics of a high surface area to volume ratio direct-current discharge have been measured for various xenon and xenon/helium gas mixtures. The breakdown characteristics in pure helium are consistent with previous studies. A database of the breakdown characteristics for a range of gas mixtures is collected and compared to the pure gas reference cases. Steady-state operation of the discharge at high pressures (>200 Torr) gives rise to operating discharge voltages that are relatively insensitive to discharge current densities. Vacuum emission spectra are obtained over the spectral range from 110 nm to 200 nm, pressure range from 100 Torr to 400 Torr, voltage range from 200 V to 500 V, and electrode gap from 50 microns to 500 microns. These experimental measurements will provide the basis for collisional-radiative model development and insight on the discharge physics that controls the population of excited ultraviolet-emitting states.

[QP1.33] Properties of Discharge Plasma at High Pressure

The estimate of the electron energy and density right at an electrical breakdown is one of the most important issues in applications of various electrical discharge plasmas in high-pressure gas. Plasma properties during and right after electrical breakdown in high-pressure gas are investigated. The mean electron energy in plasmas generated by Townsend discharge in a gas is proportional to the electric field E/p. Introducing the normalized net ionization-rate >, the electron temperature at breakdown is uniquely described in terms of the ionization properties of the gas, the second ionization coefficient ( at cathode, and the gas pressure pd. The electron attachment process plays a decisive role in breakdown phenomenon for a high-pressure gas, whereas it is not important in low-pressure discharge. An analytical expression of high-pressure plasma density is obtained by making use of the electron rate equation. A simple analytical expression of plasma generation in a high-pressure gas provides important scaling laws in the dc electrical discharge system. It is found that the logarithm of the electron density at the breakdown is proportional to the discharge time Jb, and is inversely proportional to the pulse risetime and gap distance d between the two electrodes. The plasma density at breakdown is also an increasing function of the gas ionization energy ,i.

[QP1.34] Properties of Electrical Breakdown in Flames

Properties of electrical discharge in flames and influence of plasma electrons on gas neutrals are investigated by making use of the ionization cross section of air. An analytical expression of air ionization rate is obtained from tabulated data of the ionization cross section of oxygen and nitrogen, and is compared with air ionization rate measured with the applied electric field. The influence of gas temperature on electrical discharge properties is investigated by making use of electron energy-gain in the electric field. Electrical breakdown occurs whenever ionization of neutrals dominates the electron attachment of oxygen molecules. It is found that the breakdown electric field in flames is inversely proportional to the flame temperature Tg, thereby easily generating plasmas in flames. A swarm of low-energy electrons in flames would allow a significant population of electronically excited states of flame molecules to be formed. The analysis shows that the electronic excitation of flame molecules may also considerably reduce the breakdown field. Plasma electrons generate atomic oxygen by the electron attachment of oxygen molecules in high-pressure flames. An example calculation shows that more than 63 percent of oxygen molecules are converted into atoms within 760 microseconds dwelling time for the plasma with density of np = 1013 cm-3 and temperature of Te . 2.5 eV. Oxygen atoms are the most reactive radicals in flames for material oxidation.

[QP1.35] Spectroscopic Analysis of a LiAg Plasma Plume

In a series of experiments performed at Sandia National Laboratories to characterize a Li ion source, LiAg plasma plumes were produced by laser ablation of solid targets. Production and detailed characterization of these plasma plumes are also important in laser ablation applications for synthesis of materials. Targets were irradiated with 10 ns FWHM Gaussian pulses from a Nd YAG laser, at fluences of 0.7 J/cm^2. Time- and spatially-resolved optical spectra were recorded with a framing spectrograph using a technique similar to that of microdot spectroscopy. The most prominent lines in the observed spectra include L-shell line emission from Li atoms as well as 5d-5p lines in Ag atoms. The analysis of the spectra is done using a detailed collisional-radiative atomic kinetics model that self-consistently calculates the line emission of Li and Ag atoms. Opacity effects are taken into account using an escape probability approximation, and line shapes are determined by detailed Stark-broadening calculations. This model indicates that line intensity ratios are useful for determining the plasma temperature, while the intensity of forbidden components and line broadening can be used to estimate plasma density. In addition, temperature and density time-histories from hydrodynamic simulations are used in conjunction with time-dependent atomic kinetics to investigate the importance of time-dependent effects in the spectra formation. Analysis of the data permits plasma characterization as a function of time and distance from the surface of the target.

[QP1.36] Optical Diagnostics of Laser Ablation Plumes

Experiments have been performed to investigate the temporal and spatial properties of plasma plumes produced by laser ablation of metal targets. A KrF laser is focused to generate incident fluences of 17-19 J/cm^2 with a FWHM pulse duration of 30 ns. Spatially-resolved and temporally-gated optical emission spectroscopy have been performed utilizing a spectrograph and an intensified CCD detector. Atomic Boltzmann plots yield electronic temperatures exceeding 5 eV near the target at early times. The plasma electron temperature then rapidly decays to the 1 eV range in approximately 500 ns. Dye laser resonance absorption photography experiments are underway. Inferred plasma streaming velocities are in the range of cm/microsecond.

[QP1.37] Development of a New-type of Plasma Jet by using a Modified Pseudospark Discharge

A new type of plasma jet by using a modified pseudospark discharge(PSD) is proposed. The PSD is a low-pressure gas discharge with a pair of parallel disk electrode. In this discharge, the single plane cathode is replaced by a cathode with a circular hole on the axis. The cathode electrode has a cylindrical cavity behind the circular hole. PSD is characterized as follows: (i) Discharge is formed through the electrode holes. (ii) Breakdown voltage is a function of plasma pressure. (Paschen'law-like relation) (iii) In the hollow cathode discharge, the current is lager than that for the parallel plane gap (hollow cathode effect). (iv) High current glow discharge is formed. For our plasma jet experiment, geometry of electrode is modified. The force of plasma acceleration is Lorentz force (j�~B), similar to MPD arc jet. The difference between MPD arc jet and our plasma jet is the mode of discharge. The discharge of our plasma jet is not an arc mode but a high current glow mode, as a result, the impurity production into the plasma is low and lifetime of electrodes is long. As the high current discharge and its pinch effect, high temperature and low impurity plasma is formed and spouted from the anode hole.

[QP1.38] Atmospheric Pressure Plasma Jet for Chem/Bio Warfare Decontamination

Atmospheric Pressure Plasma Jet (APPJ) technology may provide a much needed method of CBW decontamination which, unlike traditional decon methods, is dry and nondestructive to sensitive equipment and materials. The APPJ discharge uses a high-flow feedgas consisting primarily of an inert carrier gas, such as He, and a small amount of a reactive additive, such as O2, which flows between capacitively-coupled electrodes powered at 13.56 MHz. The plasma generates highly reactive metastable and atomic species of oxygen which are then directed onto a contaminated surface. The reactive effluent of the APPJ has been shown to effectively neutralize VX nerve agent as well as simulants for anthrax and mustard blister agent. Research efforts are now being directed towards reducing He consumption and increasing the allowable stand-off distance. Recent results demonstrate that by replacing the O2 reactive additive with CO2, ozone formation is greatly reduced. This has the result of extending the lifetime of atomic oxygen by an order of magnitude or more. A recirculating APP Decon Chamber which combines heat, vacuum, forced convection and reactivity is currently being developed for enhanced decontamination of sensitive equipment. Several techniques are also being evaluated for use in an APP Decon Jet for decontamination of items which cannot be placed inside a chamber.

[QP1.39] Plasma Effects on Hypersonic Flows

We numerically investigate the effect of ionization on hypersonic flows by using an approximate Riemann fully three-dimensional MHD solver, WARP3, which includes multiple temperature. The code calculates the ionization fraction as a function of temperature. We treat the flow as a single fluid with three (3) constituents, or namely ions, electons, and neutrals.

It is believed that the inclusion of multi-temperatures effects may explain the experimentally measured increase of shock stand-off distance encountered when an ionized hypersonic flow stagnates bluff body.

[QP1.40] The magnetic nozzle experiment

A facility to study the flow of magnetized plasmas expanding through a constriction formed by increased magnetic field intensity has been constructed and operated. The experiments are to study plasma recombination promoted by expansion cooling. This has applications to the fields of fusion physics, space propulsion, materials processing, and lasing systems. Through the helicon-wave method, 2 cm diameter Ar plasmas with densities up to 10^14 cm^-3 and temperatures near 5 eV have been produced in a region of nearly uniform magnetic field (B < 0.5 T), formed by a 25 cm ID Helmholtz coil pair. Inside the Helmholtz coil remote from the helicon antenna, the field is compressed to more than 1 T by a coaxial coil with 2 cm ID. Spectroscopic and electric probe techniques have been used to study plasmas in both the uniform and expanding field regions. We present numerical results from a model of 2-species plasma expansion, including 3-body effects and electron-ion coupling. More than 50% helium recombination is predicted for initial plasma densities above 10^14 cm^-3. Cool ions aid recombination.

[QP1.41] First results from the Magnetic Nozzle Experiment.

A weak magnetic nozzle producing a maximum magnetic field of more than 1 Tesla and 2 cm ID has been installed on a helicon source. The initial studies have concentrated on what effects the nozzle has on the overall plasma parameters and spectra in an Argon plasma. Spectra from a spectrometer with 1200 l/mm grating were digitized with a CCD camera with 1024x256 pixels. The probe diagnostics was carried out with a Langmuir probe which could be positioned axially and radially. We present the first results of spectrometer measurements together with Langmuir probe electron temperatures and densities.

[QP1.42] A comparison of spectroscopic measurements of an inductive plasma source with the INDUCT mode.

Noninvasive spectroscopic measurements of an inductively driven hydrogen plasma source with density and temperature characteristic of plasma processing tools have been made. Full radial and axial profiles of electron density and temperature are reduced from absolutely calibrated multichannel spectroscopic measurements of upper state number densities and a collisional radiative model. Profiles were obtained over a range of powers from 50-200 W and pressures from 5-50 mtorr in a small cylindrical source. The hydrogen working gas and simple cylindrical geometry was chosen to simplify detailed comparisions with a 2D computational model (INDUCT95) which uses a fluid approximation for the plasma and neutral gas. The code calculates the inductive coupling of the 13.56 MHz RF source, the collisional radiative, and wall losses as well as a complete chemistry model. We found agreement between the model and experimental data over part of the operational range. Ranges of agreement and divergences will be discussed.

[QP1.43] Impedance Calculation of a Solenoidal Indcutively Coupled Plasma

The plasma impedance is calculated for a solenoidal inductively coupled plasma (ICP) discharge, which is one of the important sources for plasma processing. To calculate this impedance, the electromagnetic field quantities are obtained by solving the two-dimensional Maxwell equations in a realistic geometry. Also considered in the calculation is the anomalous skin effect which is regarded as a collisionless heating mechanism of ICP discharge. This effect is presented by solving the Boltzmann equation in one-dimensional [1] and two-dimensional velocity spaces. The results show that the plasma impedance is a function of various discharge parameters, such as plasma density, electron temperature, antenna position, collision frequency, excitation frequency, and chamber geometry.

[1] K.-I. You and N. S. Yoon, Phys. Rev. E, 59, 7074 (1999).

[QP1.44] Ionization Profiling of a Star Mode IEC Device

The Inertial Electrostatic Confinement (IEC) concept has been noted as having an attractive, staged, commerical development plan for fusion technology. DaimlerChrysler Aerospace has recently produced a portable, low-cost, neutron generator for material assay based on the UI Star Mode IEC discharge. However, further investigation is required to increase the efficiency of these systems and to acquire the knowledge base for future scale-up and deployment.

This presentation will focus on recent studies in examining the ionization efficiency of the Star Mode discharge, present beamlet diagnostic data, and benchmark experiments with complimentary computational studies. Considerations for next-generation IEC scale-up will be discussed.

[QP1.45] Computational Advances in Charge Exchange Modeling of a Spherical IEC Device

The spherical Inertial Electrostatic Confinement (IEC) device produces neutrons from the fusion of an electrostatically confined deuterium plasma. Charge-exchange (CX) collisions are an important energy sink in this fusion plasma. During these collisions, fast ions become high-energy neutrals, leaving behind thermal energy ions. The CX process represents a net transfer of energy from ions to neutrals, decreasing the fusion rate and the efficiency of the system. Maximizing the fusion neutron production rate and the IEC efficiency are essential for the further development of commercial fusion neutron sources.

The Analytic Charge Exchange code was developed to show the effects of voltage, pressure, and cathode radius on the CX process in a spherical IEC. The code output includes values for the fusion rate, ion and neutral particle energy distributions, and average ion time-of-flight in the IEC. Results show a fusion scaling function that generally matches experimental results. Further insights about optimal operating pressure and the use of additional ion sources (e.g. electron emitters) can be drawn from these results. Such information will improve the design of future IEC fusion neutron sources.

[QP1.46] ECRH Produced Plasma for Materials Processing

A device is being constructed to produce high density quiet plasmas for materials processing. The device consists of a solenoid magnetic field of strength up to 10 kG, and a 1.2 m long vacuum system. The plasma is produced by ECRH at either 10 GHz, or 18 GHz, with the resonance zone located on one end of the device. Plasma can be made from any gas simply by flowing the gas though the resonant zone, however, an alternative plasma production method is to sputter source material from a plate. This can be done by placing a biased plate near the ECRH resonant zone and accelerating ions from the resonant zone that will in tern impact the sputter plate and release more neutrals that will become ionized. Using this technique it is possible to produce purely metallic plasmas from several different materials. These plasmas can be used for materials processing, and have applications in the semi-conductor industry.

Diagnostics include electrostatic probes, energy analyzers, LIF, and deposition measurements. The device is also equipped with an ICRH drive that can be used to change the pitch-angle of the plasma ions.

[QP1.47] Ion Transport in Chlorine/Argon ECR Plasma*

We have used the quasineutral particle simulation code QUASI-rz (which includes Monte Carlo and Langevin representations of the relevant charged-neutral and charged-charged collision processes) to study transport and distributions of various species in an axisymmetric ECR reactor, operating with pressures on the order of a few mTorr and plasma density up to the order of 10^12 cm-3. In this high-density electronegative plasma, ion-ion Coulomb collisions play an important role, since positive and negative ions are driven in opposite directions by the potential gradients, but are strongly coupled by collisions. The velocity distributions of both negative and positive ions are significantly isotropized as a result of ion-ion collisions, and negative ions also contribute to the heating of all heavy species. The kinetics of the plasma is spatially varying and is controlled by the interplay of species transport and a number of processes, including dissociative recombination of Cl_2^+ ions, recombination and collisional detachment of Cl^– ions, and wall recombination of atomic Cl. Densities and distribution functions of various species will be shown, at different locations in the reactor and for operating times up to 500 msec. *Work sponsored by the Office of Naval Research

[QP1.48] Control of the Electron Temperature in an ECR Plasma for Thin Film Deposition

Amorphous Silicon (a-Si:H) films with high deposition rate and less damage required from industry. The ion bombardment energy is considered to have a great influence upon the film properties, however, only few studies have been made from the point of view of the ion bombardment. In this report, we employed an electron cyclotron resonance (ECR) plasma with high electron density and attempted to control the electron temperature related to the ion bombardment energy by pulse modulation of incident microwave power or changing the magnetic field configuration. As a result, the mean electron temperature decreased from 4 eV to 2 eV by the pulsed ECR plasma. On the other hand, the electron temperature perpendicular to the substrate was about 1 eV and the electron density was 1.7 \times 10^11 cm^-3 under the optimal magnetic field configuration. Furthermore, high quality a-Si:H thin films (\sigma_p / \sigma_d \sim 10^6) with high deposition rate (14 A/sec) were obtained without heating the substrate. In addition, we have tried to perform the deposition of films by an ECR plasma of 915 MHz.

[QP1.49] Role of Peripheral Vacuum Regions in the Control of the Electron Cyclotron Resonance Plasma Uniformity

Spatial measurements of the ion saturation current density indicate stable vacuum regions have a possibility to contribute to plasma uniformity by behaving as a waveguide for the incident electromagnetic waves. Mode conversion of electromagnetic waves with long wavelength to the right circular polarized wave was observed experimentally at a certain radial position. Furthermore, microwave propagation in a partially filled plasma chamber was examined numerically. The simulation indicated that the electromagnetic waves with long wavelengths propagated in a periphery of the plasma were converted into the extraordinary wave or electrostatic waves outside the ECR region and that the power absorption took place at the local regions. Physical considerations toward these results imply the reason why the plasma uniformity in influenced by magnetic field gradient.

[QP1.50] Frequency Measurement by Time of Flight Method from DC to AC Radiation Converter with Perpendicular DC Magnetic Field

We have demonstrated the short microwave pulse generation by a DC to AC radiation converter (DARC) principle with/without a DC magnetic field along the electrostatic field. A periodic electrostatic field is excited by a capacitor array in an X-band waveguide with a cut-off frequency of 6.6 GHz. An ionization front is created by 4 ømega_0 light (266 nm, 100 mJ, 6 ns (FWHM)) of the Nd:YAG laser and a front velocity is approximately equal to a speed of light. An emitted frequency with a perpendicular DC magnetic field is given by ømega \approx k_0 v_f / 2 + (ømega_p^2 / 2 k_0 v_f) [1+ømega_c^2/(ømega^2 - ømega_h^2)], where ømega_ h and ømega_c are the upper hybrid frequency and the cyclotron frequency, respectively. Thus, further frequency upshift is expected by adding the DC magnetic field. In our experiments, the electrostatic field consists of the wavelength of 2 cm with a gap of 0.6 cm. A working gas is TMAE (U_i = 5.36 eV). The emitted frequency from DARC is measured by the time of flight method of the delay waveguide line. The velocity of the wave in the waveguide depends on the frequency. The observed frequency is 9.5 GHz without the magnetic field and is 12 GHz with the magnetic field of 2 kGauss in the plasma density of \sim 6 \times 10^11 cm^-3. The experimental value is good agreement with the theoretical value.

[QP1.51] Slicing Off and Adding Up of Short Microwave Pulse by Laser Produced Overdense Plasma

An ultrashort electromagnetic wave pulse is expected to be used for the excitation of large amplitude wavefield and the diagnostic of temporal evolution in the plasma wave phenomena. We have demonstrated the slicing off short microwave pulse from CW or long pulse sources by using the laser produced overdense plasma. The sliced short microwave pulse is created by shut off the long pulse at several points in the waveguide by employing the overdense plasma. The pulse width is equal to \tau = L / v_g, where L and v_g = \mboxconst. are the waveguide length between two spatial points and the group velocity of the microwave pulse in the waveguide. In our experiments, a low power Gunn diode (\sim 5 mW) and high power magnetron (\sim 100 kW, 1 \mus) at the frequency of 9 GHz are used as a microwave source. An ionizing laser is the 4 ømega_0 light of the Nd:YAG laser (266 nm, 100 mJ, 6 ns, h \nu = 4.7 eV) and a working gas is TMAE (U_ i = 5.6 eV). The group velocity in the waveguide is equal to v_ g = 0.68 c at 9 GHz. The observed pulse width is 5 ns for the waveguide length of 2 m. The observed pulse width is the good agreement with the theoretical value. In low power experiments, we have also demonstrated the adding up two separately created pulses into twice of amplitude using the microwave components. We will present the experimental results of adding up in higher power microwave source at the conference.

[QP1.52] Experimental Studies on Frequency Upshift of Microwave by Flash Ionizatio

Frequency Upshift of microwave radiation has been observed in the interaction experiments between the microwave and laser plasma. Output radiation frequency ømega depends on both a plasma frequency ømega_p and the microwave frequency ømega_0. The expected emitting frequency is ømega=\sqrtømega_p^2+ømega_0^2. Fourth harmonics of Nd:YAG laser light at a wavelength of 265 nm with maximum energy of 100 mJ is used for ionizing the gas to produce plasma. Line focus is created in the waveguide by a cylindrical lens. We used TMAE as a working gas. The frequency of the used microwave is 9 GHz with the maximum power of 5 mW generated by Gunn diode. The frequency of the microwave is detected by the crystal detector and several specific waveguidest. The observed maximum frequency of the electromagnetic wave is higher than 31.4 GHz with the pulse duration of 6 ns. We also detected the current in the plasma by using two tiny dB/dt probes which were located both side of plasma inside the waveguide. This current is due to the electron accelerated by the electric field of microwave. The detected current is good agreement with the theory.

[QP1.53] Cherenkov Radiation from Magnetized Plasmas

Large amplitude electro-static (es) plasma waves are generated in plasma acceleration experiments: >100 GeV/m in Laser Wakefield Acceleration (LWFA), >30 GeV/m in Plasma Beatwave Acceleration (PBWA), and >1 GeV/m in Plasma Wakefield Acceleration (PWFA). In these experiments, the energy stored in these es waves is ultimately dissipated in the plasma. When applying a static magnetic field (B), the perturbation (laser pulse, the laser beatwave, or the electron bunch) couples to the L branch of the XO mode of the magnetized plasma through Cerenkov radiation. The Cerenkov radiation is emitted essentially in the forward direction, at the plasma frequency. The LWFA and the PWFA correspond to Cerenkov radiation from photons. When reaching the plasma/vacuum boundary, the electromagnetic (em) component of the XO mode couples to the vacuum em mode. The transmission coefficient is equal to 1 in the case of a sharp boundary. With the parameters of the UCLA PBWA experiment, MW to GW of THz radiation are expected with magnetic fields in the 6 to 180 kG range. Numerical simulations as well as plans for experiments for laser and beam drivers cases will be presented.

[QP1.54] Thomson backscattered x-rays at 400 eV in the NRL Laser Synchrotron Source Experiment

A Laser Synchrotron Source (LSS)(P. Sprangle, et al., J. Appl. Phys. 72), 5032 (1993) experiment is in progrss at the Naval Research Laboratory to demonstrate the generation of monochromatic, tunable x-rays by Thomson backscattering of laser photons from a relativistic electron beam. Laser photons from a Nd:Glass laser at 1.053 \mum wavelength are backscattered from a 4 MeV electron beam generated by a one and a half cell S-band RF gun. X-ray photons have been observed at 400 eV. Characterization of the energy spectrum, flux, and angular distribution of the x-rays are being performed and the experimental progress will be reported.

[QP1.55] Stochastic Electron Gas Theory of Coherence in Compton Scattering

The transition from coherent to incoherent Compton scattering is studied within the framework of a stochastic electron gas model, where the statistical distribution of the electron phase is taken into account. The fundamental difference between this approach and a relativistic fluid model resides in the fact that, for any number of incoherently phased point electrons, the four-current contains Fourier components at arbitrarily short wavelengths, whereas the fluid model introduces an unphysical cutoff scale. The coherence factor is found to scale from N^2 to N as the ratio of the bunch to the wave length varies between zero and infinity, and the discrete nature of electric charge is shown to play a fundamental role in the physics of Compton scattering.

[QP1.56] Generation of Single Cycle Laser Pulses by Photon Acceleration/Deceleration

When a short-pulse laser propagates through an underdense plasma it generates a plasma wave wake. The laser pulse’s energy decreases as the wake is generated while the number of photons, or equivalently, the classical action of the electromagnetic wave is conserved. If the pulse is shorter than a plasma period then the energy of each photon, \hbarømega, gradually decreases; this is referred to as photon deceleration. Photon deceleration will continue until ømega approaches ømega_p at which time only a single wavelength of radiation is left. In addition, since the number of photons, E^2/ømega, is conserved and ømega is decreasing, then the normalized laser momentum, p/(mc)=(eE)/(mcømega), will increase despite the fact that E is decreasing. Therefore, this mechanism is also an p_0/(mc) amplifier. We have observed this behavior in both 1D and 2D PIC (particle-in-cell) simulations for laser and plasma parameters, e.g., n_0=4\times 10^19cm^-3, \lambda_0=1\mu m, I=10^19W/cm^2, and \tau_L=15fs, which are now experimentally feasible. We will describe the physics of photon deceleration and present 3D PIC simulation results to determine whether photon deceleration can be used to generate attosecond pulses.

Work supported by DoE, NSF and LLNL.

[QP1.57] Observation of ionization instability of intense laser pulses

The scattering of intense laser pulses in ionizing gas jets is observed. The scattering is characterized by ‘feather’ shaped ionization tracks, which are left after the laser pulse passes through the gas. The scattering is accompanied by blue shifting of the laser spectrum and depends on the plasma density and the atomic number of the target gas. Numerical simulations in which the dominant process is an ionization scattering instability qualitatively reproduce the experimentally observed plasma density profile.

[QP1.58] Acceleration of Plasma Electrons by a Short Intense Laser Pulse

We have studied the generation of moderately energetic electrons by a single, short laser pulse propagating through plasma. The plasma is formed by complete ionization of the helium gas due to the high power of the laser pulse while at the same time accelerating the ionized electrons. The main mechanism responsible for electron generation in this case is wave breaking. We have studied the conditions for producing energetic electrons (>200keV) by varying three different parameters; the laser pulse length, the peak power and the helium gas pressure. First, we fixed the laser pulse length and varied the pressure for different laser energies. Energetic electrons were generated only when the gas pressure was higher than the self-focusing critical pressure. Next, the laser peak power was fixed and the pressure was varied for several laser pulse lengths. Once again, no energetic electrons were generated below the critical pressure. The dependence of the electrons spectrum on pulse parameters will be discussed.

[QP1.59] Measurement of Spatial and Temporal Profiles of Electron Plasma Oscillation Excited by Ultrashort Laser Pulse

Large amplitude electron plasma waves (EPW), which are produced by ultrashort laser pulses, are of great interest for particle acceleration or photon acceleration. In this study, we present the temporally and spatially resolved measurements of the electron density perturbation produced by the laser wakefield (LWF) process.

0.6 TW Ti:sapphire laser pulse ionized the helium gas of \sim 1 Torr near the focus and excited the electron density perturbation. We observed this electron density perturbation by the frequency-domain interferometry technique. The probe pulse was the second harmonic of the partially separated pulse from the main pump pulse. The probe pulse was sent into the Michelson interferometer and make two colinear pulses. These two probe pulses go through the EPW, and are affected by EPW of which phase velocity is almost equal to the light velocity. Each pulse obtains a phase shift depending on the phase of EPW. These two pulses interfer each other in the spectometer. Spatialy resolved relative phase shift can be obtained from the interferogram.

With varying the relative delay between the two probe pulses, 2 THz periodic change of the relative phase shift was observed. It was caused by 2THz electron density oscillation in LWF.

[QP1.60] Particle Acceleration via Laser-Cluster Interaction: Multi-Cluster Distribution

Highly energetic electrons and ions are known to be produced during the interaction of a laser pulse with atomic and molecular clusters. Here, we report on simulation studies of cluster explosions under irradiation with high-intensity, ultra-short laser pulses and apply it to the case when a multi-cluster array is present. By using a 2D fully relativistic Particle-in-Cell code we investigate the influence of the cluster distribution on the dynamics of the explosion and compare it with the single-cluster case, providing information about the time-resolved position, momentum and energy of electrons and ions, for different laser intensities, and cluster separation.

[QP1.61] Design of a High Gradient Vacuum Laser Accelerator

The inverse free-electron laser (IFEL) interaction is studied theoretically and computationally in the case where the drive laser intensity approaches the relativistic regime, and the pulse duration is only a few optical cycles long. It is shown that by using an ultrashort, ultrahigh-intensity drive laser pulse, the IFEL interaction bandwidth and accelerating gradient are increased considerably, thus yielding high energy gains. Using a chirped pulse and negative dispersion focusing optics allows one to take further advantage of the laser optical bandwidth and produce a chromatic line focus maximizing the gradient. The combination of these novel ideas results in a compact vacuum laser accelerator capable of accelerating picosecond electron bunches with a high gradient (GeV/m) and low energy spread.

[QP1.62] A Covariant Approach to Intense Laser-Plasma Interactions

We present a relativistically covariant fluid model of intense laser-plasma interactions for the purpose of studying laser wakefield accelerators. This approach has the advantage that in any (Lorentz) frame the dynamical equations are naturally in ``conservative'' form, greatly simplifying their numerical solution. To study ponderomotive effects we perform a two scale analysis covariantly\/, thus preserving the conservative form of the equations. We present numerical results for a variety of plasma configurations.

[QP1.63] Petawatt Laser Pulse Interaction with Plasmas

An ultrashort laser pulse in the petawatt range with field amplitude larger than E_h=ømega c^2\sqrtm_em_i/e interacting with an underdense plasma exhibits effects that are typical of overdense-plasmas at more moderate pulse amplitudes. Ion acceleration by petawatt laser radiation is studied with 2D3V-PIC (Particle in Cell) numerical simulations. These simulations show that the laser pulse drills a channel through the plasma slab, electrons and ions expand in vacuum. Fast electrons escape first from the electron-ion cloud. Later ions gain a high energy, due to the Coulomb explosion of the cloud and inductive electric field which appears due to fast change of the magnetic field generated by the laser pulse.

[QP1.64] Optical guiding of terawatt laser pulses in the plasma waveguide

We report coupling and guiding of pulses of peak power ~0.5 TW in 1.5 cm long preformed plasma waveguides generated in a high repetition rate argon gas jet. Greater than 50 percent coupling was measured in the injection of 50 mJ, 100 fs pulses, giving guided intensities up to ~10^17 W/cm^2. For short delays between waveguide generation and pulse injection, refraction-induced pulse shortening occurred, with this effect reduced either by increasing the delay between waveguide generation and injection or by injecting a prepulse into the waveguide. We will also describe recent experiments which attempt to reduce the avalanche ionization threshold for the gases in which the waveguide is generated.

This work is supported by the US Department of Energy (DEF G0297 ER 41039) and the National Science Foundation (PHY-9515509).

[QP1.65] Exponential and Explosive Growth of Relativistic Plasma Wave From Beat Wave Excitation at \Deltaømega=2ømega_p and 3ømega_p

The study of the excitation of relativistic plasma waves by beating two lasers is important to the plasma-based accelerator research. It is well-known that in their original paper, Rosenbluth and Liu [Phys. Rev. Lett. 29, 701 (1972)] showed that if \Deltaømega=ømega_p, the plasma wave grows secularly in time and that it saturates due to a relativistic nonlinear frequency shift. It is less well-known that, in addition, they also predicted that if \Deltaømega=2ømega_p, the plasma wave will grow exponentially. We will show that if \Deltaømega=3ømega_p, a new explosive instability is theoretically predicted. The growth rates of these two instabilities are independent of \epsilon\equivømega_p/ømega_0 while the growth rate of the Raman forward scattering is proportional to \epsilon. This provides a way to separate the two instabilities from the Raman scattering. We will also use particle-in-cell (PIC) codes to study these two instabilities.

[QP1.66] Photon Kinetic Theory and Simulations of Laser-Plasma Interactions

The propagation of electromagnetic waves in underdense plasmas can be described by a Vlasov-like transport equation for the Wigner distribution function of the e.m.field. We determine the temporal growth rates for Raman Scattering for all angles from the photon kinetic theory, and compare them with the standard growth rates, thus establishing the limits of validity of the theory. The kinetic structure of the e.m.field propagation equation leads naturally to a ``photon in cell" numerical model. We compare 1D simulations (neglecting plasma response) and 2D simulations (using QWAKE, the photon kinetic version of WAKE (P.Mora and T.M.Antonsen, PoP 4), 217 (1997)) of ultraintense laser pulse (a_0 \simeq 1) propagation in underdense plasmas, with analytic results for the pulse spread/compression, pulse steepening and the velocity of energy transport, and we find good agreement between the simulations and the theory. Applications of the photon kinetic theory to intense broadband radiation-plasma interaction will also be discussed.

[QP1.67] Wakefield Generation in Plasma Channels

Laser wakefield generation in plasma channels is experimentally studied. Plasma channels, produced using the ignitor-heater method [1] in hydrogen and nitrogen, have been used to guide intense (> 5 x 10^17 W/cm^2), short (<70 fs) infrared (800 nm) laser pulses. Laser pulses injected into these channels produce a plasma wake with a phase velocity close to the speed of light. The transverse density profile of the channel determines the properties of the laser mode as well as of the plasma wave mode. The longitudinally integrated properties of the channel are measured with a Mach-Zehnder interferometer using 400 nm radiation. The probe and reference beam are combined directly on a CCD camera to provide two-dimensional interferograms and also through a spectrometer to allow Fourier domain interferometry. Progress on measuring the transverse channel profile and wakefield amplitudes will be presented. [1] P. Volfbeyn, E. Esarey and W.P. Leemans, Phys. Plasmas 6, 2269 (1999).

[QP1.68] Progress towards a Colliding Pulse Laser Wakefield Accelerator

The possibility of generating ultra-short (1-10 fs), low energy spread (< 5 %), low emittance (< \pi mm-mrad) electron bunches with energies of 10's of MeV's using the colliding pulse optical injection scheme [1] is being studied experimentally at LBNL. In the colliding pulse scheme three synchronized laser pulses are used. The first pulse is used to excite a fast phase-velocity plasma wave and the other two collinear but counterpropagating pulses are used to produce a slow phase velocity beatwave allowing background electrons to be injected into the plasma wave. Progress on implementing this all-optical acceleration method using a high power (10 TW), short pulse (< 70 fs) Ti:Al_2O_3 laser system will be presented.

[1] E.Esarey et al., Phys. Rev. Lett. 79, 2682 (1997).

[2] C.B. Schroeder et al., Phys. Rev. E 59, 6037 (1999).

[QP1.69] Raman Forward Scattering of Intense Laser Pulses in Plasma Channels

In this work we study Raman Forward Scattering of an intense laser pulse in a plasma channel which only supports a single laser mode. Due to the radial variation of plasma frequency, several distinguished features of RFS in inhomogeneous plasma are discussed, including: the temporal growth rate of RFS in the center of a plasma channel is smaller than the well known results for homegenous plasma, \gamma_0 = |a_0| ømega_p/\sqrt8 ømega_0 because of the dephasing of plasma excitation between different radial plasma elements. We also study under which condition the damped quasi-modes (G.~Shvets and X.~Li, ``Theory of Laser Wakes in Plasma Channels", Phys.~Plasmas 6), 591 (1999) which are unique to inhomogeneous plasma can be coupled to the pump wave resonantly and emerge as global plasma waves with a well-defined frequency throughout the channel.

[QP1.70] Recent Results from Variational Principle Approaches to Studying Short-Pulse Laser-Plasma Interactions

Over the past few years we have developed a variational principle approach for describing the evolution of short pulse lasers. The starting point is an action integral whose Euler-Lagrange equations recover the model equations. In its simplest form the action can be thought of as S=\int d\tau d\Psi d\vecx_\perpL(a, a^\ast•,\phi), where a is the vector potential of the laser and \phi is the scalor potential of the potential. Substituting trial functions for a and \phi and performing the \int d\vecx_\perp (or \int d\Psi) integration provides a reduced action whose Euler-Lagrange equations give envelope equations for the parameters of the trial functions. We will present recent results using the variational principle on the fully nonlinear set of fluid equations, on propagation in channels, on the coupling of direct Forward Raman scattering to whole beam instabilities, and on the group velocity of short-pulse lasers.

Work supported by DoE, NSF and LLNL.

Brian J. Duda and W.B. Mori, Submitted to Phys. Rev. E

[QP1.71] Coherent Control of Stimulated Raman Scattering Using Chirped Laser Pulses

A novel method for the control of stimulated Raman scattering in short-pulse laser-plasma interactions is proposed. Linear frequency chirp in a non-bandwidth limited pulse is used to selectively increase or decrease the instability growth rate, depending on the sign of the chirp. Theoretical calculations show that a chirped pulse with 12% bandwidth is needed to eliminate forward Raman scattering. The predicted control over the growth rate is confirmed in two-dimensional particle-in-cell simulations of 20% bandwidth pulses. Relevance to areas of current research are also discussed.

[QP1.72] Characterization of a laser ionization electron injector for laser based accelerators

The laser ionization and ponderomotive acceleration (LIPA)(C.I. Moore et al., PRL, 82, 1688 (1999)) of electrons released from noble gases produces high brightness electron beams. Results from previous experiments and numerical simulations have shown that these electron beams are very promising as an electron source for laser based particle accelerators. Knowledge of the parameters of these electron beams is essential for understanding the full potential of this injector mechanism. Experiments have been performed at NRL to characterize the electron beams and verify simulation predictions. Results from these experiments will be presented.

[QP1.73] Electron Injection in the NRL Laser Wakefield Accelerator Experiment

Experiments on the external injection of electrons into the Laser Wakefield Accelerator (LWFA) are being performed at the Naval Research Laboratory. The laser ionization and ponderomotive acceleration (LIPA) (C.I. Moore, et al., Phys. Rev. Lett. 82), 1688 (1999) technique is being used to generate short pulses of energetic electrons for injection into the wakefield. These optically generated electron bunches are perfectly synchronized to the wakefield and have well defined energy and spatial characteristics. Both the wakefields and the injection electron bunches are generated by the same short pulse T^3 laser which is operating at 2.5 TW and 400 fsec pulse length. The T^3 laser is being upgrade to produce over 20 TW with a vacuum compressor. Progress on the experiment of electron injection and subsequent acceleration by the wakefields will be presented.

[QP1.74] NRL Capillary Optical Guiding Experiments for Laser Plasma Accelerators

Extended propagation of TW laser pulses has been shown in plasma channels formed with a capillary discharge.( D. Kaganovich, et al., Phys. Rev. E 59) R4769 (1999). Capillary optical guiding experiments at NRL include plasma channels formed using both laser preionization and electrical discharge. Present experiments aim to characterize plasma channel conditions within

the capillary. Time-resolved backward-scattered Raman spectra, measured with a picosecond resolution streak camera, yield plasma densities within the capillary as low as 10^18 cm^-3. Other capillary guiding diagnostics include forward Raman spectra, 90^\circ Thomson scattering imaging, and output beam mode imaging. Progress on the experiment and plans for capillary guided laser acceleration experiments will be presented.

[QP1.75] Characterization of Short Pulse Laser Propagation in the NRL Laser Wakefield Accelerator Experiment

As a short intense laser pulse propagates in a plasma medium and generates wakefields, it will continuously experience both temporal and phase distortions. It is important to understand the effects of these distortions on the evolution of both the laser pulse and the wakefields as the pulse propagates through the plasma in a Laser Wakefield Accelerator (LWFA). These distortions can be studied with the use of frequency-resolved optical gating (FROG) technique. Progress and results on the characterization of the laser pulse from the NRL LWFA will be reported.

[QP1.76] Intense Short Laser Pulses Propagating Through Channels

Optical guiding of intense short laser pulses in plasmas is important in applications ranging from x-ray lasers to laser driven electron accelerators. The dynamics of short intense laser pulses propagating in channels is analyzed. The model describing the pulse envelope dynamics includes nonlinear, finite pulse length and group velocity dispersion effects. The intense short laser pulse propagating in the channel undergoes self phase modulation resulting in a large frequency spread. The nonlinear evolution of the pulse's frequency spread is analyzed. In addition to ultra broadband radiation generation, laser driven electron acceleration in plasma channels will be discussed.

[QP1.77] Analysis of Laser Ionization and Ponderomotive Acceleration.

Laser Ionization and Ponderomotive Acceleration (LIPA) is a new approach for the generation of electron beams from a gas. In this process an intense laser beam strips electrons from atoms, leading to a short pulse of relatively monoenergetic electrons with low emittance.(C.I. Moore et al.), Phys. Rev. Lett. 82, 1688 (1999). The energy of the electrons depends on the ponderomotive potential of the laser beam. Additionally the electrons emerge from the interaction region in directions that are correlated with the final energy. An analysis of LIPA based on tunneling ionization and subsequent classical motion is presented. Electron distribution functions are obtained for linear and circular polarization laser beams and compared with Monte Carlo simulations.

[QP1.78] Self-Modulation in a Channel-Guided Laser Wakefield Accelerator.

Most laser wakefield accelerator (LWFA) experiments to date have operated in the high plasma density self-modulated (SM) regime in which the laser pulse length c \tau_L is much longer than the plasma wavelength \lambda_p = 2 \pi c / ømega_p. These experiments rely on a self-guiding effect to overcome diffraction of the laser pulse and extend the interaction distance. In a channel-guided SM-LWFA, a preformed plasma channel provides the guiding. Simulations indicate that for a wide range of laser and channel parameters, the laser pulse will become modulated at \lambda_p, eventually producing accelerating gradients which may exceed 10 GV/m. The resulting wakefields are generally more regular than those produced in the self-guided regime, and the accelerating gradients are usually much higher than in the resonant (c \tau_L \sim \lambda_p / 2) LWFA regime. Simulations of a recent channel guiding experiment at NRL(D. Kaganovich, et al., Phys. Rev. E \bf59), R4769 (1999) show that self-modulation and high accelerating gradients may have been achieved in that experiment.

[QP1.79] Langmuir Probe Analysis for "High Density" Processing Plasmas

Semiconductor etchers have moved from RIE plasmas in the range 10^10-11 cm^-3 to ICPs in the range 10^11-12 cm^-3. In the lower range, saturation ion currents were found to vary as V^1/2, yielding linear I^2-V curves. This has been erroneously attributed to orbital-motion-limited (OML) theory. In a paper^1 specifically written on this point, we had previously shown that the parabolic behavior was accidental and was caused by sheath expansion. At densities > 10^13 cm^-3, sheaths are so thin that probes are essentially planar, and linear extrapolation of the I-V curves can be used. In the intermediate regime 10^11-12 cm^-3, the exact solution of Bernstein and Rabinowitz*, as extended to Maxwellians by Laframboise*, must be used. Because of nonuniform convergence, T_i = 0 cannot be used for cylindrical probes. As has been done by more recent workers*, we have parametrized the computed curves so that automatic fitting of I-V curves can be done in real time. However, the ratio \lambda _D/r_p cannot exceed 3 for the collisionless theory to hold; a larger probe diameter must then be used. ^1F.F. Chen, J.Appl. Phys. 36, 675 (1965). *References excised by word limit.

[QP1.80] Measurements of the negative ion density in reactive gas plasmas

The reactive gas plasmas, such as C_4F_8, SiH_4 and SF_6 gas plasmas, have been widely used in plasma etching or CVD. The radicals and ions species in these plasmas have been reported in a lot of study. However, the negative ion density has not been measured quantitatively, since the conventional Langmuir probe cannot be used due to film depositions on its surface. In this study, the negative ion density in the reactive gas plasmas was measured with a heated Langmuir probe and an 8-mm microwave interferometer as a function of gas flow rate and radial position. Furthermore, the following equation was suggested to estimate the negative ion density only from the probe measurements: \fracI_+(X )I_+(Ar) = \left[ \fracI_-(X)I_-(Ar) + \fracN_-(X )N_+(Ar)\sqrt\fracT_e(X)T_e(Ar) \right] \sqrt\fracM_+(Ar)M_+(X), where N_- denotes negative ion density and the other characteristics represent the conventional ones. The positive ion mass M_+ should be assumed properly. It was confirmed that this equation provides the negative ion density both in the magnetized plasmas, such as ECR plasmas, and non-magnetized plasmas.

[QP1.81] Detection of RF Perturbations in a Helicon Plasma Using a Heavy Ion Beam Probe

We are conducting experiments with the ultimate goal of developing a non-perturbing diagnostic technique for measuring RF field and density fluctuations in fusion plasmas. In order to develop this technique, a helicon plasma with a heavy ion beam probe (HIBP) diagnostic is being constructed. The helicon plasma has a magnetic field of up to 1.5 kG produced by a set of circular coils. A 1 kW, 13.56 MHz RF generator will be used to drive helicon waves in the plasma. Diagnostic apparatus utilizes a 60 keV HIBP diagnostic beam and detector that were previously used on the Tokamak de Varenne. The detection electronics is being modified to operate at higher frequencies. Present status of the experiment as well as key issues in extending HIBP measurements to higher frequencies will be discussed. These key issues include noise levels and path effects on the signal.

[QP1.82] Single-Gas Bubble Neutron Detectors for Alpha Knock-On Tail Measurements

Measurement of the neutron energy spectrum above \sim16~MeV will yield information on the spatial and energy distributions of confined fast alphas in DT tokamaks. Standard two-gas bubble neutron detectors, designed to only detect neutrons with energies above a selectable threshold determined by the gas mixture, were used in preliminary attempts to measure the knock-on neutrons from DT plasmas in TFTR and JET. Subsequent measurements at accelerator neutron sources showed an unexpected below-threshold detector response that prevented observations of the alpha-induced neutron tails. Spontaneous bubble nucleation measurements show that this below-threshold response is due to slight variations in the gas mixture, and is not present in single-gas detectors. Single-gas detectors will be tested at Ohio University and at UC Berkeley to determine the neutron energy threshold as a function of detector operating temperature, and to confirm the lack of a below-threshold response. An array of single-gas detectors operating at different temperatures should allow measurements of the alpha knock-on neutron tail during the planned DTE2 experiments on JET.

[QP1.83] Reconstruction of Beam Shape Using Fluctuation Spectroscopy

Information about a particle beam's shape and structure can be determined from noise in or produced by the beam. Fluctuations in the incoherent radiation emitted by a particle beam during interaction with external fields or in a medium are not ``totally random'' white noise, but encode certain information about the phase space structure of the beam, which can in principle be extracted by a careful statistical analysis of the spectral properties of the noise. The frequency scale at which correlations in the spectral noise decay is inversely related to the temporal duration of the beam, and more generally, structure in the spectral covariance matrix of the radiation encodes information about both the longitudinal and transverse shape of the beam, so that certain features of the beam can be reconstructed from these data. This technique of beam reconstruction by fluctuation spectroscopy was first suggested by M. Zolotorov and G. Stupakov(M.S. Zolotorov and G.V. Stupakov, SLAC-PUB-7132, 1996; in Proceedings of the 1997 Particle Accelerator Conference) (IEEE, Piscataway, NJ, 1998)., and a proof-of-principle experiment and data analysis have recently been performed by P. Catravas, et al.(P. Catravas, W.P. Leemans, J.S. Wurtele, M.S. Zolotorov, M. Babzien, I. Ben-Szi, Z. Segalov, X.-J. Wang, V. Yakimenko, Phys. Rev. Lett.), 82, 5261 (1999)., where the bunch length and emittance were estimated using shot-noise driven fluctuations in the spontaneous emission of a relativistic electron microbunch in a wiggler. Here we aim to improve the accuracy of the reconstructed beam morphology by developing a more elaborate statistical analysis. This technique shows promise as a diagnostic for particle beams which is non-destructive, can work either with single shots or with averages over many shots, can maintain accuracy even at short bunch lengths, and uses avaialble technology. Only the case of spontaneous emission of an electron beam in a wiggler has been studied in depth, but the technique can be applied more generally, for example to the cases of Cerenkov or transistion radiation of a beam in a medium, or perhaps even to distinct areas of physics, where noise-driven fluctuations reveal real underlying structure.

[QP1.84] Array Technology Development for Millimeter-Wave Plasma Imaging Diagnostics

One of the problems encountered in the application of reflectometric imaging to fusion plasmas is that the transmit antenna must be aligned such that the reflected waveform is successfully collected by the receiver mixer array. One would ideally like to have antennas which may be steered in real-time to compensate for changes in plasma shape and/or position. Another problem lies in providing low cost, wide bandwidth sources capable of supplying the required 100-500 mW power levels required of the transmit beam. A program has been initiated at UC Davis to address both of these problems, and develop (i) microwave beam steerers suitable for electronically-scanned reflectometers, and (ii) wide bandwidth frequency mutliplier arrays which can efficiently frequency multiply the output of low frequency sources. Laboratory test results will be presented, along with a description of ongoing research activities.

[QP1.85] Ultrashort Pulse Reflectometry (USPR) Density Profile Measurements on GAMMA-10

Ultrashort pulse reflectometry (USPR) involves time-of-flight measurements of extremely broadband, high speed chirped signals (~ns sweep times). A multichannel USPR system has been installed on the central cell of the GAMMA-10 mirror machine located at the University of Tsukuba, Japan. Here, the output from a 65 ps FWHM impulse generator is stretched and amplified to form a ~10 ns duration, 11-18 GHz chirp signal. A five channel X-mode USPR receiver, with frequency channels at 12, 13, 15, 16 and 17 GHz, measures the double-pass time delay of each reflected subpacket simultaneously with 25 ps time resolution. Density profile and fluctuation data collected on GAMMA-10 will be presented.

[QP1.86] 2-D Numerical Simulations of a Reflectometric Imaging System

We present 2-D numerical simulations of a reflectometric imaging system for measurement of density fluctuations in support of a 3-D experimental apparatus to be constructed in FY1999 and installed on a major tokamak or stellarator in FY2000. The imaging system includes lenses to ensure locally normal incidence of the reflectometer beam onto curved plasma isodensity surfaces and a spatially distributed array of receiving antennae to collect the reflected radiation. The effects of beam spot size, non-normal incidence of the illuminating beam, and locations of receiving antennae on the system’s measurement capabilities will be presented.

[QP1.87] Investigation of Correlation Reflectometry for Magnetic Field and Turbulence Measurements in a Fusion-Relevant Plasma

Investigations of fundamental aspects of correlation reflectometry in both single mode (O-O, X-X) and crossed mode (O-X) configurations are presented. Experiments have been conducted in the LArge Plasma Device (LAPD) at UCLA, a low temperature linear device where detailed measurements of turbulence with probe arrays are possible. In the LAPD, density profiles are well known, and experiments have been conducted over a large number of repeatable discharges, resulting in low levels of measurement and statistical error. Comparisons between reflectometer and probe measurements show good agreement between homodyne and probe correlation lengths and power spectra over a range of L_n's, while the reflectometer phase shows agreement only in cases of low fluctuation levels. In addition, proof of principle experiments of a local magnetic field measurement using O-X cross-correlations are presented. It is shown that the magnetic field strength can be determined from O-X reflectometer measurements via a one-dimensional numerical model, given estimates of L_n and the k-spectral width of the turbulence.

[QP1.88] Development of a Two-Dimensional, Time-Dependent, Parallelized Wave Propagation Code as an Aide to Interpreting Reflectometry Data.

The wave propagation equation is solved in a two-dimensional geometry, consisting of three regions: (1) vacuum, including a realistic antennae geometry, (2) the inhomogeneous plasma, which is currently described in the fluid approximation, and (3) perfectly matched absorbing layers (``An Anisotropic Perfectly Matched Layer-Absorbing Medium for the Truncation of FDTD Lattices)'', Gedney, S., IEEE Transactions on Antennas and Propagation, 44, (1996). at the boundaries. For studying systems of experimentally relevant size, the processing capacity of single CPU systems is greatly exceeded, and therefore the use of massively parallel computers was necessary. An explicit finite-difference time domain method (``Waves and Fields in Inhomogeneous Media)'', W.~C.~Chew (Van Nostrand Reinhold, New York, 1990). enables efficient use of parallel architecture computers. Results will be benchmarked against existing serial codes. Ultimately such simulations should provide a useful tool for more detailed interpretation of experimental reflectometry measurements.

[QP1.89] 3-D Reflectometric and ECE Imaging Diagnostic for TEXTOR

A multi-dimensional, high temporal and spatial resolution imaging diagnostic, capable of simultaneous reflectometric and electron cyclotron emission (ECE) imaging, is under development for 2-D mapping of electron density and temperature turbulence spectra on the TEXTOR tokamak. The goal is to make localized, multi-dimensional correlation measurements between Te and ne turbulence over a broad region of the TEXTOR plasma. Such data are crucial to understanding the underlying physics between microturbulence and anomolous transport. A detailed description of the system design and implementation plans will be presented.

[QP1.90] Fast ion Collective Thomson Scattering diagnostic for TEXTOR

A Collective Thomson Scattering diagnostic is being built for diagnosing confined fast ions in the TEXTOR tokamak. ICRH and NBI produce highly non--thermal fast ion distributions in TEXTOR, where deuterons with energies up to 400 keV can be confined. The probing radiation will be provided by a gyrotron delivering 350 kW for up to 200 ms at 110 GHz. At B_\phi = 2.6 Tesla the ECE and other plasma noise is below 10 eV, while the detector noise is \approx 5 eV. This should permit 20 measurements per shot with a temporal resolution of 10 ms and adequate velocity space resolution. Initially a near back--scattering geometry will be implemented, providing a radial resolution of \approx10 cm. Radial location of the measuring volume and direction of the resolved velocity component can be varied from shot to shot. Later an additional receiver will provide near 90 degrees scattering angle, facilitating simultaneous resolution of the 1--D velocity distributions near parallel and perpendicular to the magnetic field with a radial resolution in both cases of \approx5 cm. The diagnostic capability is aimed at addressing both generic issues of fast ion dynamics such as phase space diffusion and redistribution at sawtooth events, and specific issues for ICRH physics such as efficiency of coupling to high energy tails.

[QP1.91] TEXTOR Collective Thomson Scattering Receiver

A receiver system is being constructed for a collective Thomson scattering experiment to measure the density and velocity distribution function of confined energetic ions which are heated by ion cyclotron resonance heating (ICRH) and neutral beam injection in the TEXTOR tokamak. The receiver system is matched to a scattering source which is a 350 kW, 110 GHz gyrotron with a pulse length of up to 200 ms. The receiver system will initially have 32 receiver channels of 80 MHz bandwidth, and 8 channels of 750 MHz bandwidth which will sample an overall 10 GHz bandwidth. The receiver noise temperature is estimated at approximately 5 eV. When the tokamak is operated at 2.6 Tesla, the electron cyclotron emission ranges up to approximately 10 eV. A notch filter will be used to reject the gyrotron radiation which is not Doppler-shifted by the plasma. An IF multiplexer just after the RF mixer will be used to prevent IF amplifier saturation by stray gyrotron radiaton by channeling the Doppler broadened spectrum into separate IF amplifier chains. This system should allow for 10 ms temporal resolution with adequate signal to noise performance.

[QP1.92] Compact Cassegrain Antenna Design for Collective Thomson Scattering Exeriments on TEXTOR

Collective Thomson scattering (CTS) of 110GHz radiation is being implemented on TEXTOR to yield localized information on confined high energy ions. Optimizing these measurements requires careful shaping of the receiver antenna beam to maximize signal and ensure beam overlap. The experiment will be implemented in two stages. The first stage antenna will be installed in the poloidal plane of the probe, using near back scattering geometry, near perpendicular to the magnetic field. The second stage will operate at large angles to the field and a second antenna, implementing a 90° scattering geometry, will be installed. All antennas will be steerable and of compact Cassegrain design. In the first system an on-axis smooth rectangular horn is used as feed for a 10cm diameter parabolic mirror with focal length of 5cm. The second system will use antennas with corrugated feeds and mirrors that produce gaussian beams focussed on the mid-plane. Work supported by the US DOE.

[QP1.93] Improvements on the Belief Time and Magnetic Field Resolution of the Transient Internal Probe (TIP) Plasma Diagnostic

TIP is a unique diagnostic used for direct measurement of core plasma B fields at both high spatial (on the order of 1 cm) and temporal (1 \mus) resolution. A glass verdet probe illuminated by a 514.5 nm, polarized laser traverses a plasma at a speed of approximately 2 km/s, causing rotation of the polarization angle by an amount proportional to the local B field component along the transit, due to the Faraday effect. For the probe to survive transit of a dense plasma with temperatures above about 100 eV, a refractory cladding is necessary. This paper presents results of tests on probes clad with sapphire and accelerated with the TIP gas gun. Data documenting their structural integrity at an average speed of 1.6 km/s will be shown, along with computations of their belief times inside the SSPX, DIII--D, and NSTX plasmas. Probe response to applied fields will also be measured, for demonstration of an absolute B field resolution on the order of 20 G. Finally, ray-tracing analyses and lab calibrations of a ``cat's eye'' optical system will be presented, as an alternative sapphire clad probe design for improved retro-reflection of the laser to the TIP ellipsometer.

[QP1.94] Internal Magnetic Field Measurements on the FIX Experiment

FIX (FRC Injection Experiment) is a Field Reversed Configuration (FRC) used for translation, confinement, and NBI experiments. A Faraday effect probe is being developed for internal magnetic field measurements on FIX. The Faraday effect: in the presence of a magnetic field, a Faraday rotator material will rotate the polarization of light by an angle \theta = VBL, where V is the Verdet constant of the material, B is the field component along the direction of propagation, and L is the length of material that the beam traverses. The FIX probe uses Yttrium Iron Garnet (YIG), which has a Verdet constant of 0.1 ^o/gauss-cm. It is dropped into the FIX confinement chamber and illuminated with a 1.064 \mum YAG laser. The polarization of the returned signal is measured and inverted to give the poloidal magnetic field. Profiles may be measured over several shots by changing the time delay between the dropping of the probe and the plasma discharge.

[QP1.95] Spectroscopic Imaging of Plasma Ion Temperature and Flow in the H-1 Heliac

It has been recently shown that fixed delay modulated Fourier transform spectrometers have important fundamental and instrumental advantages over frequency domain instruments such as grating spectrometers ^1. In this paper we describe new 2-D imaging time-domain spectrometers used for tomography of the flow vorticity and ion distribution function in H-1 plasmas. The spectrometers utilize one or more electrooptic crystals to visualize the coherence envelope and phase of an isolated spectral emission line. Recent results obtained using these instruments will be presented. \footnotesize 1. J. Howard, J. Opt. Soc. Am. A (Submitted 1999)

[QP1.96] Multi-channel H_\alpha Diagnostics for Position Determination of a Tokamak Plasma

A multi-channel H_\alpha spectroscopic diagnostic system was developed on KAIST-Tokamak to be utilized for diagnosing the plasma position in the early phase of ohmic discharges. Since the measured intensity is line-integrated along the line of sight, an Abel inversion computer program was developed. Using the inversion program, the vertical (similar to minor-radial) H_\alpha intensity profile was obtained at several different time steps in the start-up phase of KAIST-Tokamak ohmic discharges. The center position of the intensity is an indirect indication of the plasma center. Comparison with the magnetics data shows reasonable agreement. This suggests that the multi-channel H_\alpha diagnostic may be a good candidate to determine the plasma position which can be useful especially for plasma position control in the tokamak start-up phase.

[QP1.97] Measurement of Z_eff Profile in MST

The effective ion charge (Z_eff) profile is essential to make accurate statements regarding resistivity, power deposition and thus confinement time. The Z_eff profile on MST has been determined via a measurement of near-infrared bremsstrahlung emission, which relates to Z_eff by: \epsilon (\lambda ) = \frac 1.89e^\rm-27 g Z_\rmeff n_\rme^2 \sqrtT_\rme \lambda ^2 \frac W cm^3 nm . Spectroscopic measurements in MST, obtained through the use of both a CCD spectrometer and a NIR spectrometer, reveal the absence of spectral lines in a narrow window centered at 1040 nm where bremsstrahlung dominates. Detectors on a 17-chord array on MST have been developed with optical filters and signal amplifiers optimized to measure the line integrated emission in this spectral range. A 2-dimensional inversion in MSTFit, along with accurate Thomson Scattering electron temperature (T_e) and FIR electron density (n_e) profiles yield the Z_eff profile. Results are shown for both standard and enhanced confinement (PPCD) discharges, illustrating an appreciable decrease in Z_eff during the enhanced confinement periods.

[QP1.98] Two-dimensional measurements of turbulent structures using laser-induced fluorescence

An LIF-based diagnostic system for measuring ion density fluctuations in two spatial dimensions is under development to be used on the Magnetic Reconnection Experiment~(M.~Yamada et al.), Phys.~Plasmas, 4, 1936, (1997). 2D images of turbulent structures will be useful in clarifying our understanding of reconnection physics. A sheet beam of a pulsed tunable laser is used to excite the ion emission from Ar II, Kr II, Xe II. The fluorescence emitted from the plane of the laser beam is detected with a filter and an intensified CCD camera, providing an image of relative ion density fluctuations. It is expected that the entire plasma volume in the Magnetic Reconnection Experiment, CDX-U, as well as edge plasma on fusion experiments such as NSTX will be accessible to this technique~(C.~H.~Skinner et al.) Rev.~Sci.~Instr. 70, 917, (1999). Initial LIF measurements on a helicon source of argon plasma are planned for the near future, followed by experiments on the MRX. Preliminary results will be presented.

[QP1.99] A Check on the Relative Importance of Higher Order Effects in the Unified Theory of Electron Broadening

Some features seen in the spectra of hot, dense plasmas can reach over such a large spectral range that an all-order electron broadening model may be needed. To check this, comparisons will be made between the results of a semiclassical all-order model, a semiclassical second-order model using similar approximations, and a quantum mechanical second-order model. In addition to this, calculations of a spectra series (Ly-alpha, Ly-beta, Ly-gamma) will be made treating it as a single object instead of as individual lines and therefore including the mixing effects. These calculations along with calculations of overlapping lines will also be used to determine the relative importance of the higher order terms in the all-order model.

[QP1.100] Multichannel Optical Pyrometry Temperature Measurements of Imploding Metal Liners

A five channel optical pyrometer has been used to measure the temperature of metal samples on the inner diameter of a z-pinch, imploding liner. Temperature increase in the material under test of ~50 to 1000 degrees Kelvin are observed due to the plastic deformation caused by the radial convergence (~3:1). (1, 2) Sub-microsecond time resolution is required to track the time dependence of the temperature and permit unfolding of material strength from the temperature data. Interpretation of temperature from the observed radiances, including effects of emissivity variations, and the shielding from stray radiation will be discussed.

1. R.R. Bartsch, et al, ``Imploding Liner Material Strength Measurements at High-Strain and High Strain Rate", LA-UR-98-2669, Paper 6-07, Proceedings of the VIIIth International Conference On Megagauss Magnetic Field Generation, Tallahassee, FLA., Oct. 18-23, 1998 2. H. Lee, et al, Paper MO-29, ibid

This work supported by the US DOE under contract W-7405-ENG-36

[QP1.101] Filtered Silicon Photodiodes for Plasma Diagnostics

Silicon photodiodes possess excellent properties for measuring low temperature plasmas. These include a nominally flat response, insensitivity to surface contamination, low voltage biasing requirements, sensitivity to low energy photons, excellent detector to detector response reproducibility, and ability to operate in poor vacuum or gas backfilled experiments. We have characterized silicon photodiodes from 1 eV to 10 keV photon energy for time response and signal saturation levels. Our 'standard' detector assembly, the XUV-7, provides seven 0.2x0.2 mm photodiodes arrayed on a 5.3 mm circle in a vacuum leak-tight, electrically isolated, low noise, high bandwidth, x-ray filtered assembly in a compact package 94 mm long by 37 mm diameter. Other custom-built assemblies provide readouts for transmission grating spectrometers, and act as tiny x-ray detectors used in confined space applications. We have used our silicon photodiodes to diagnose low-density plasmas produced on explosive pulsed power experiments with Bremsstrahlung temperatures of about 100 eV and also high-density z-pinch plasmas with Planckian temperatures up to 200 eV.

[QP1.102] Spectral Signatures of Anisotropies in Si Plasmas Driven by High-Intensity, Femtosecond Pulsed Lasers

Plasmas driven by high-intensity femtosecond-duration pulsed lasers are investigated because of their potential as sources of directional energetic electrons and bright ultrafast X-rays. We study the polarization properties of the He-like Si satellite line emission of the Ly-alpha line. Hydrodynamic simulations and transient ionization balance calculations are used to estimate the overall plasma and satellite line emission behavior. Furthermore, 2-D PIC simulations are performed to model the electron kinetics and magnetic field generation. These results are used to feed a detailed, transient, magnetic sublevel atomic kinetics model for the calculation of level alignment and polarization-dependent satellite line spectra. We discuss the time evolution of the polarized spectra and identify polarization markers that are sensitive to the anisotropy of the plasma.

[QP1.103] A Two-Dimensional, Time-Gated X-ray/EUV Imaging Spectrometer

A novel two-dimensional x-ray/EUV imaging spectrometer (2DXIS) has been developed to investigate complex, rapidly-evolving plasmas. This novel device uses glass capillaries to multiplex a two-dimensional image of the plasma into an output array of spatially-separated pixels. This array is then spectrally dispersed by a crystal or a multilayer mirror, recorded by a temporally-gated imager, and reconstituted as an image by a computer. The result is a time-gated spectrum for each image point. Polycapillaries are employed, rather than monocapillaries, for a uniform output angular distribution of captured radition. Modular design allows observation of a wide selection of spectral lines. Efficient computer calculations of resonance and satellite lines analyze the high volume of K-shell/L-shell spectroscopic data. Spatial resolution, field of view, wavelength range, and spectroscopic resolution are chosen to yield time-resolved two-dimensional maps of plasma temperature, density, and ionization state. A 100-channel 2DXIS is being installed on SNL-Z to observe Ti K-shell radiation from imploding-Ti-wire-array z-pinches. Two-dimensional x-ray spectroscopy of other wire materials (Al, Ni, Cu, Mo, W, and Au) will also be developed.

[QP1.104] X-ray Spectropolarimetry of Dense Z-pinch Plasma

X-ray spectropolarimetry is a powerful new tool for investigating the anisotropy of dense z-pinch plasmas. It is sensitive to the magnetic field and energetic electron distribution function, plasma characteristics important to z-pinch behavior, that, in general, have not been measured adequately. The influence of the magnetic field on the polarization of L-shell line radiation is large enough to be measured for a broad range of experimentally-realizable z-pinch conditions. Calculations of x-ray line polarization have been performed, and prospective diagnostic configurations developed, for magnetic field measurements on SNL-Z and NTF-Zebra. In particular, the Ti L-shell spectrum provides good candidate lines for measuring magnetic field through line polarization. The diagnostic also makes use of the L-shell spectra to monitor plasma ionization state, density and temperature, and novel analysis techniques to accomplish this have been developed for Al, Ti, Ni, and Cu ions.

[QP1.105] Demonstration of soft x-ray laser plasma interferometry with a tabletop laser and an amplitude division interferometer based on diffraction ratings.

We have performed soft x-ray interferometry of a laser-created plasma using a tabletop capillary discharge laser operating at 46.9nm in combination with a novel amplitude division interferometer. The interferometer utilizes diffraction gratings as beam splitters in a Mach-Zehnder configuration to generate high visibility fringes over a large field of view. This table-top system was used to probe a large scale (=3mm long) plasma created by a Nd:YAG laser. The short wavelength of the probe laser has allowed mapping of the electron density in plasma regions where steep density gradients exceed those that could be probed with 265nm radiation (fourth harmonic of YAG).

[QP1.106] Edge diagnostics for KSTAR using hyperthermal neutrals

Hyperthermal neutrals with energy level between 5 eV and 100 eV can be utilized to understand how sputtered and recycled neutrals move and to measure the electron temperature and density of the edge plasma in magnetic confinement devices. DEGAS 2, Monte-Carlo neutral transport code, follows trajectory of each neutral in the pancake-style TCP source in which the reflective neutrals are generated. The code optimizes the height of the TCP source to maximize the neutral flux out of the TCP source. Using DEGAS 2 code we can calculate the flux and energy distribution of the hyperthermal beam at interesting positions. It is essential to analyze neutral's behavior in the 3D cylindrical space, which is the geometry of the TCP source.

[QP1.107] KSTAR Magnetic Diagnostics

Magnetic diagnostics are essential tools for operation and control of the KSTAR tokamak and for understanding the plasma behavior. For this purpose, a magnetic diagnostic test chamber (MDTC) was fabricated to test sample magnetic diagnostics under similar conditions with the KSTAR vacuum vessel. Design specifications, fabrication concepts for the KSTAR magnetic diagnostics, and results from the MDTC will be presented.

[QP1.108] Designing an AVS/Express interface to MDSPLUS

We are in the process of designing a visualization system using AVS/express. Our focus is in designing a data analysis system for MDSPLUS data, which will allow users to see 1 to n dimensional plots of different signals with individual signals placed in different windows. Users will be able to rotate, to animate, and to merge the plots interactively and will be able to perform mathematical operations on the data sets. Another key feature of this system is that it will be collaborative. Researchers will be able to visualize and to perform mathematical operations on the data with other researchers at other institutions. Our focus will not be on large scale simulations, but rather small scale data sets. Since we are using MDSPLUS data, we will not have to send data sets from one client to another. Scripting commands will be sent from individual clients, which results in small packets being sent from individual users. We will also allow researchers to share custom modules.

Part Q of program listing