Papers

AVS1997 Session PS-ThA: Plasma Surface Interactions II

Thursday, October 23, 1997 2:00 PM in Room A5/6

Thursday Afternoon

Start	Invited?	Item
2:00 PM		PS-ThA-1 Beam Scattering Studies of Plasma-Surface Interactions in Chlorine Etching of Photoresist, Polysilicon and Silicon Dioxide J.P. Chang, H.H. Sawin (Massachusetts Institute of Technology) Ion-enhanced chlorine etching of photoresist, silicon dioxide, and polysilicon were studied with a direct beam scattering technique to determine the reaction rate coefficients required to predict the etching of patterned polysilicon. To measure the surface kinetics, the etching rates were measured with the beam fluxes (30-100eV Cl⁺, Cl, Cl₂, SiCl₂) independently varied at levels comparable to those encountered in a high density plasma reactor. The reduction of etching rates caused by deposition/redeposition of the etching products/by-products was explored by utilizing a SiCl₂ beam - SiCl₂ is typically produced by etching or by electron impact dissociation of SiCl₄ in the plasma¹. This approach permits a thorough understanding of the fundamental reaction mechanisms and allows formulation of a kinetic model useable in a profile simulator to model the profile evolution during plasma etching processes. The chlorine ion-enhanced etching of photoresist, silicon dioxide, and polysilicon are characterized as a function of Cl⁺ ion energy, ion flux, Cl/Cl⁺ and SiCl₂/Cl⁺ flux ratios, and angle of ion impingement. The sticking coefficient of SiCl₂ to form a stable SiCl_x film was measured and calculated to be approximately 0.3; use of this sticking coefficient would suggest deposition rather than etching of polysilicon. As a result, etching of the polysilicon by Cl⁺ or Cl⁺ and Cl was reduced by a factor of two in the presence of SiCl₂ (~10% of the Cl flux ); however, the apparent deposition probability of SiCl₂ was reduced by an order of magnitude when simultaneously exposed to Cl⁺ or Cl⁺ and Cl. This indicates that the chlorination and/or other modification of the surface caused by the Cl⁺ and Cl fluxes greatly reduces the apparent sticking probability of SiCl₂. Possible reaction pathways are proposed and reduced into a phenomenological model which is useable in a profile simulator. ¹C. Lee, D. B. Graves and M. A. Lieberman, Plasma Chem. and Plasma Proc. 16, 99 (1996)
2:40 PM		PS-ThA-3 Ion Bombardment Energies in Continuous and Pulsed High Frequency Plasma for Materials Processing. O. Zabeida, J.E. Klemberg-Sapieha, L. Martinu (Ecole Polytechnique, Canada) Dual-mode microwave/radiofrequency (MW/RF) plasma is now being used in our laboratory for materials processing such as polymer surface modification for enhanced adhesion and deposition of optical, protective or gas permeation barrier coatings. Generally, the energy of impinging ions is an important factor which determines the film and surface properties ( density, hardness, stress, roughness etc. ). In the present work we use a large area MW/RF plasma reactor in which the sample can be exposed to a MW (2.45 GHz) or an RF (13.56 MHz) discharge individually or simultaneously. We use a four-grid differentially pumped ion energy analyzer to measure ion energy distribution functions (IEDF) on the grounded and on the RF powered electrodes. We present the IEDF for different process gases, such as Ar, He, N₂, NH₃ and N₂O, and for different plasma excitation modes. The IEDF characteristics, such as their shape, the maximum ion energy and the average ion energy, E_a, are correlated with plasma density and electron temperature (measured by electrostatic probes) for a wide range of the gas pressure and excitation power. We show that the E_a values are between 5 and 10 eV for the MW plasma and 10-20 eV in the RF plasma on the grounded electrode, while they reach several hundreds of eV on the RF powered electrode. The effect of pulse frequency and of the duty cycle in MW discharge is also studied: we observed that these parameters affect the IEDF and the ion flux. We clearly demonstrate that relative number of ions in the low energy part of the IEDF increases with decreasing the duty time. We correlate the ion bombardment phenomena under different modes of operation ( MW, RF, MW/RF, continuous or pulsed) with the characteristics of plasma treated polymers and plasma-deposited silicon nitride films.
3:00 PM		PS-ThA-4 Positive Ion Species in Inductively Coupled rf Discharges Containing Mixtures of Cl₂, BCl₃, Ar, and N₂ J.R. Woodworth, C.A. Nichols (Sandia National Laboratories); T.W. Hamilton (Applied Physics International, Inc.) We are using a quadrupole mass spectrometer (QMS) to determine the positive ion species at the wafer position in inductively coupled plasma (ICP) discharges in a Gaseous Electronics Conference Reference Cell. This work examined gas mixtures containing Cl₂, BCl₃, Ar, and N₂. These mixtures have been used in the etching of metal films for semiconductor interconnects. The QMS is placed in a differentially pumped region below a ~5 micron diameter pinhole in the center of the grounded lower electrode. Transmission of the QMS as a function of mass is calibrated, allowing the relative flux of each ion to be known. In most cases, the dominant ions we detect are Cl+, Cl₂+ and BCl₂+. In 20 mTorr, 300 W discharges containing a 3/1 Cl₂/BCl₃ mixture and a stainless steel lower electrode, the Cl+/Cl₂+/BCl₂+ ratios were 1.0/2.2/2.0. The Cl+/BCl₂+ ratio increases by a factor of seven in going from 200 w to 400 w discharge power, and decreases by a factor of five in going from 15 mTorr to 50 mTorr. The Cl+ and Cl₂+ fluxes decrease as the discharge becomes BCl₃ rich, suggesting that BCl₃ is not a significant source of Cl₂ in the discharge. The nature of the "wafer" and the wall materials surrounding the discharge have a large impact on the discharge parameters. When a Si wafer was placed on the electrode, the discharge became more highly dissociated, with the Cl +/Cl₂+ ratio increasing by about a factor of 3. Placing an aluminum wafer on the electrode resulted in a discharge with little free chlorine and many AlCl_x compounds. Addition of N₂ to the discharge lead to a rapid buildup of material on the walls and an increase in the degree of dissociation in the discharge. This work was supported by the United States Department of Energy under Contract DE-AC04-94AL85000 and by SEMATECH. Sandia is a multiprogram laboratory operated by the Sandia Corporation, a Lockheed Martin Company, for the United States Government.
3:20 PM		PS-ThA-5 Surface Reactivity, and Velocity Distributions of NH₂ Radicals in an NH₃ Plasma Molecular Beam and Scattered from a Surface. P.R. McCurdy, V.A. Venturo, E.R. Fisher (Colorado State University) Using spatially and temporally resolved LIF we have measured the velocity distributions of NH₂ molecules in an effusive NH₃ plasma molecular beam as a function of applied rf plasma power. We have also measured the velocity distribution of NH₂ radicals scattered from a Si substrate at 300 K. Monte Carlo simulation methods were used to model spatial profiles of radicals in the molecular beam and those scattered from the Si surface. The model assumes an initial Gaussian distribution for radicals both in the molecular beam and scattered from the surface and calculates time dependent changes in the profiles using Maxwell-Boltzmann distributions. For the scattered molecules, a cosine distribution of velocities from the surface normal is included. The translational temperature of the NH₂ radicals in the molecular beam increases with rf power, from 512 ± 8 K at 25 W to 664 ± 35 K at 150 W. Velocity profiles for NH₂ radicals scattered from a Si substrate were determined to be ~330 K. The surface reactivity of NH₂ in this system was determined to be near zero. The implications of these data for plasma surface modification processes will also be discussed.
3:40 PM		PS-ThA-6 Measurements and Modeling of Low-Energy Sputtering as a Function of Target Temperature and Grain Size D.N. Ruzic, J.P. Allain (University of Illinois) The angular distribution of sputtered material and the absolute sputtering yield of metal targets by Argon ions at energies less than 1000 eV has been measured in previous work for a number of materials. This paper focuses on the temperature dependence and grain size dependence of the distributions and yields. The ion beam is produced by a Colutron ion gun and is decelerated near the target. At the target the beam diameter can be as small as 100 microns or can be rastered over larger areas to include grain boundaries. The target assembly is fixed and monitored by a thermocouple. A "cold finger" which can deliver liquid N₂ or boiling water is attached to the target. The diagnostics are rotated near the target to intercept the sputtered flux. They include a quartz crystal oscillator to measure the total yield and a pyrolytic graphite collector plate. The graphite plate is vacuum transferred to a nearby Auger spectrometer to obtain the areal densities and thus the angular distribution. The modeling of the systems are conducted using the latest version of VFTRIM-3D, a code which includes fractal geometry and a non-binary collision model.
4:00 PM		PS-ThA-7 Molecular Dynamics Simulations of Low Energy (25-200 eV) Ion-Surface Interactions N.A. Kubota, D.J. Economou (University of Houston); S.J. Plimpton (Sandia National Laboratories) The etch yield and sub-surface radiation damage are important issues in low energy (<200 eV) ion interactions with surfaces. In particular, the atomic layer etching (ALET) process requires etching of electronic materials with monolayer precision and minimal inter-layer atomic mixing. In this study, the molecular dynamics technique is used to simulate the impact of argon ions on chlorine-free and chlorine-passivated silicon surfaces. The silicon sputter yield is reported for Ar ion energies ranging from 25-200 eV. Thousands of individual impact simulations are performed on an nCUBE-2 massively parallel supercomputer. Sputter yields from both defect-free and amorphized surfaces are also reported and compared with different sources of available experimental results. Sputtering of initially defect-free surfaces reveal the tendency of volatile product formation reactions to follow distinct local trajectories. The formation of volatile product species due to 120 eV Ar ions impacting into Si(100)(2 x 1) at normal incidence has been found to occur mainly by a mechanism in which the Ar ion impacts directly in between a surface silicon dimer pairing. The energetic recoiled silicon atoms undercut nearby silicon atoms resulting in volatile product formation. Several other product formation pathways will also be reported.
4:20 PM		PS-ThA-8 Surface Chemistry of NF₃ Plasma and Ion Beam Interactions with Silicon T.W. Little, F.S. Ohuchi (University of Washington) Nitrogen trifluoride (NF₃) is rapidly gaining popularity as an etchant gas for a wide variety of silicon-based materials, but mostly for silicon (Si) itself. In addition to fast etching rates and high selectivity, NF₃ is also preferable to more established etchant gases such as carbon tetrafluoride (CF₄) and other perfluorocarbons in terms of environmental compatibility. Although NF₃ is not known to form any non-volatile reaction products at a macroscopic level during silicon etching, it is not clear if this is truly the case at the nanometer-order length scales typically measured by surface characterization techniques since there have been few studies of this type. In fact, ion beam studies of NF₃ interaction with Si have indicated both etching and nitridation processes. The results of such studies will be presented in this work along with the extension to plasma processing using mixtures of NF₃ and nitrogen (N₂) feed gases. Optical emission spectroscopy (OES) has been used to monitor atomic fluorine and molecular nitrogen species in the plasmas which are then correlated to the results of surface characterization by techniques such as x-ray photoelectron spectroscopy (XPS) following in vacuo sample transfer. The presentation will focus on the competition between etching and nitridation reactions as a function of gas mixtures and processing parameters. Results indicate that both nitridation and etching/fluorination do occur and seem to be quite sensitive to the presence of oxygen. Possible reaction mechanisms based on proposed bonding configurations will be presented.
4:40 PM		PS-ThA-9 Interactions of Molecular Fragments from Silane/Hydrogen Plasma with Si Surfaces: An Atomic Scale Computational Study S. Ramalingam, D. Maroudas, E.S. Aydil (University of California, Santa Barbara) Plasma enhanced chemical vapor deposition of hydrogenated amorphous silicon and nanocrystalline silicon is an important process in manufacturing of solar cells and thin film transistors for flat panel displays. Despite large number of studies, the surface reactions and processes that determine the film structure and composition remain unknown. In this study, we employed atomic-scale computer simulations to study the interactions of molecular fragments originating from SiH₄/H₂ discharge with surfaces of hydrogenated amorphous and crystalline silicon films. The objective is to provide insight into the chemical and physical processes that occur on surfaces during plasma deposition of Si films. The atomic scale studies are based on molecular dynamics simulations and aided by structural relaxation and Monte Carlo methods. The interatomic interactions in the Si:H system are described by recently developed interatomic potentials that extend Tersoff's potential for Si to incorporate Si-H, H-H, and the corresponding three body interactions. Specifically, we have studied the reactions of SiH, SiH₂ and SiH₃ with (i) a- Si:H film surfaces with varying degrees of hydrogen coverage, (ii) with Si(001)-(2x1) surfaces and hydrogen terminated Si(001)-(2x1) surfaces. Bond-breaking and bond-forming processes as well as the mobility of the radicals on the surfaces are monitored in detail during the MD simulations of radical impingement and reaction with the surfaces. Results of the atomic scale study reveal that, in addition to the identity of the gas phase radical, the reactivity also depends on the H coverage, the location of impingement on the surface, and the molecular orientation of the radical with respect to the surface. Surface reactivity is as important as the reactivity of the radical in determining the reaction probabilities or the so-called 'sticking coefficients'.
5:00 PM		PS-ThA-10 Molecular Dynamics Simulations of Physisorbed Halogen Atoms on Halogenated Silicon Surfaces B.A. Helmer, D.B. Graves (University of California, Berkeley) The mechanistic details of the surface interactions of halogen atoms and other radical species created in the plasma are still poorly understood. We present the results of molecular dynamics (MD) simulations of the trapping, desorption, surface diffusion, and surface recombination reactions of weakly-bound (physisorbed) halogen atoms (F, Cl, and Br) on halogenated silicon surfaces. The simulations have been conducted to provide insight into the possible mechanisms of the surface recombination reactions of the halogen atoms. The simulations were motivated by the recent experiments of Kota, Coburn, and Graves¹, who measured surface recombination reaction coefficients for F, Cl, and Br, on a variety of surfaces for a range of temperatures. Their data suggest that physisorbed halogen atoms participate in the recombination reactions. We have simulated the dynamics of these weakly-bound atoms, for comparison with experiment when possible. We examined the effects of surface temperature (100 to 600 K), surface structure (e.g., surface roughness), halogen species, and the assumed parameters in the interatomic potentials (Stillinger-Weber and Lennard-Jones forms). Because the halogen atoms are weakly bound to the surface (binding energies on the order of 0.1 eV), the time-scales for desorption and diffusion can approach the time-scales of the MD trajectories, depending on the surface temperature. The MD results and transition state theory are used to estimate trapping probabilities, and the rates of desorption, diffusion, and recombination. The rates of these processes are discussed as a function of surface temperature and the other parameters listed above. ¹Kota et al., unpublished