This research program is focusing on the demonstration and development of rare-earth-doped fiber and planar waveguide lasers operating in the visible and ultraviolet. Recently, the first ultraviolet fiber laser was demonstrated. Current work focuses on developing new blue, violet, and UV fiber lasers that are pumped by near-IR diode lasers. Recently observed transitions in Nd ³⁺ in Tm ^;s3+ are pursued.

Microdischarge devices in silicon have recently been demonstrated. Having apertures ;lt400 ;gmm, these devices are intense sources of visible and UV radiation and are being developed into arrays. New sources of coherent UV and VUV radiation based on parametric four-wave mixing and SHG in atoms and small molecules are also being developed.

Experiments are in progress to develop a visible-emitting fiber laser pumped by red or near-IR semiconductor diodes. Emphasis is placed on the praseodymium- or holmium-doped fluorozirconate glass (ZBLAN) fiber lasers that operate in the red (;gl ;sl 635 nm) or green (;gl ;sl 549 nm), respectively. By optimization of the pump wavelength and fiber length, the pump power threshold of the Ho:ZBLAN has been reduced below 15 mW, which brings the pumping requirements for the device within the realm of commercially available red diodes. Efforts to frequency double the Pr:ZBLAN laser into the UV are underway.

The focus of these experiments is the development of fiber and planar waveguides in rare earth-doped LiNbO;i3 or fluoride glass that are capable of generating or amplifying coherent visible or UV radiation. The optical quality of LiNbO;i3 or glass films produced by sol-gel processes are being examined by laser techniques. Recently, a gas phase process for converting oxide films to fluoride glass was demonstrated in collaboration with colleagues in the Department of Materials Science and Engineering. The ultimate goal of this work is the development of waveguides suitable for being driven by semiconductor laser diodes.

Plasmas are becoming increasingly important in high-technology manufacturing. Plasmas are used to deposit films, harden materials, modify surfaces, and synthesize bulk materials. We are supporting the University of Wisconsin National Science Foundation Engineering Research Center for Plasma-Aided Manufacturing by developing computer models for systems of interest to the manufacture of microelectronics and index gradable materials. We have developed nonequilibrium electron and ion kinetics models of reactive ion etching (RIE) and inductively coupled plasma (ICP) reactors; and cleanup of plasma sources for ion implantation (PSII) using Fourier transform mass spectroscopy (FTMS). We are also addressing the consequences of particulate contamination in polymerizing plasmas.

One of the greatest causes of reduced yields during manufacturing of microelectronics devices is contamination of the wafer by particles (sizes 10s nm to many ;gmm). The contamination of wafers during plasma etching results from particles being driven to the wafer by ``ion wind,'' electrostatic, fluid drag, and thermophoretic forces. We are developing comprehensive computer models of plasma etching tools that include the forces on and transport of dust particles. We have quantified the manner in which electrode-topography may lead to electrostatic traps for particles in the plasma. We are collaborating with semiconductor manufacturers to minimize particle contamination.

Nonplanar structures in contact with a plasma can cause perturbations in the electric potentials and plasma densities. Plasma etching reactors for fabrication of microelectronic devices are carefully designed to minimize these perturbing effects. It is less well known that nonuniform structures below the wafer can also perturb the plasma. Observations of dust particle traps and nonuniform etching have been correlated with dielectric structures beneath the wafer. In this research program, we are developing plasma equipment models to investigate the effects of electrode topography (above and below the wafer) on the uniformity of plasma generation, ion fluxes to the wafer, and dust particle traps.

Increasing environmental awareness and regulatory controls have motivated research into new methods to cleanse toxic gases from air streams. We are computationally investigating the use of plasmas for cleansing of NO;yx (oxides of nitrogen) from diesel exhaust. Reducing (or oxidizing) radicals, which react with NO;yx to produce less harmful substances, are readily and efficiently produced by plasmas. Production of these radicals can be easily and rapidly controlled by plasma sources such as dielectric barrier discharges or coronas, thereby providing a practical method to adapt to changing road conditions. In this work, we are developing multidimensional computer models of the plasma chemistry and gas dynamics during the remediation process.

More stringent regulations on the allowable levels of toxic gases in the exhausts from internal combustion engines and industrial processes have motivated research into more efficient methods to treat those gas streams. We are developing multidimensional plasma chemistry models to investigate plasma remediation as a method to remove toxins from atmospheric pressure gases. In plasma remediation, electron impact reactions produce oxidizing or reducing radicals which either convert the toxin to a harmless gas or to a gas for which conventional remediation methods can be used. We are investigating reaction mechanisms and the plasma hydrodynamics in dielectric barrier and corona discharges for remediation of NO;zx;za and VOCs (volatile organic compounds). Scaling laws for energy efficiency are being developed.

Of the hundreds of processing steps in the manufacture of silicon microelectronics devices such as memory chips and microprocessors, approximately one-third use plasmas for etching, deposition, cleaning, or passivation. As wafer sizes continue to increase, the need for highly uniform, particle-free plasma-processing equipment also increases. In this project we are developing plasma equipment models (PEMs) to study important plasma generation and transport processes and to investigate methods to scale plasma reactors to process larger wafers. The PEMs are geometrically flexible and are able to address a variety of chemistries. Particular attention is being paid to processes which generate particle contamination of the wafers, and to coupling plasma transport codes to feature profile simulators.

As wafer sizes continue to increase, improvements in plasma processing for microelectronics fabrication will soon require that real-time control be applied to stabilize and improve manufacturing steps. We are investigating real-time control methodologies for obtaining spatially uniform etch rates and to signal the end point of the etch. This work is being performed in collaboration with the University of Michigan, where experiments on real-time control of plasma etching reactors are being performed. We are developing computer models of those processes to determine the physical processes, sensor outputs, and control points that are best monitored to achieve these goals.

As the feature size of microelectronics components shrinks to below 100nm, the interfaces between structures become increasingly more important to the performance of the device. Remote plasma processing provides a means whereby surfaces may be cleaned and passivated and dielectric layers grown while carefully controlling the flux of reactants to the substrate. In a collaborative research project with North Carolina State University, we are developing computer simulations of remote plasma reactors and surface growth kinetics. These models are being used to optimize reactor conditions for producing thin films for microelectronics devices.

The development of plasma models for materials processing has rapidly progressed during the past five years. In many cases, model development has outpaced validation. These are few instances of comprehensive and iterative validation of plasma models with detailed experimental data. In this cooperative research program, 2-D optical diagnostics of excited state species in the GEC plasma reference cell are being performed at NIST/Gaithersburg to provide benchmarking data for plasma equipment models being developed at the University of Illinois. Excited state densities in A;zr, A;zr/0;i2, Ar/CF;i4, and Ar/Cl;i2 plasmas have been computed and validated against measurements. These comparisons have provided insight to the underlying physics and have provided a mechanism for validation of complex computer simulations.

Particle contamination of wafers during plasma processing of microelectronics is a major source of reduced yield. We have previously developed algorithms and computer models (the dust transport simulation, DTS) to predict particle contamination and growth in plasma tools. In this research project, these algorithms are being transferred to CFD Research Corp. for use in comprehensive computational fluid dynamics computer-aided design tools.