Summary, Findings, Conclusions, And Recommendations
This study focuses on the plasma processing of materials, a technology that impacts and is of vital importance to several of the largest manufacturing industries in the world. Foremost among these industries is the electronics industry, in which plasma-based processes are indispensable for the manufacture of very large-scale integrated (VLSI) microelectronic circuits (or chips). Plasma processing of materials is also a critical technology in the aerospace, automotive, steel, biomedical, and toxic waste management industries. Because plasma processing is an integral part of the infrastructure of so many American industries, it is important for both the economy and the national security that America maintain a strong leadership role in this technology.
A plasma is a partially or fully ionized gas containing electrons, ions, and neutral atoms or molecules. In Chapter 2, the panel categorizes different kinds of plasmas and focuses on properties of man-made low-energy, highly collisional plasmas that are particularly useful in materials processing applications. The outstanding properties of most plasmas applied to processing of materials are associated with nonequilibrium conditions. These properties present a challenge to the plasma scientist and an opportunity to the technologist. The opportunities for materials processing stem from the ability of a plasma to provide a highly excited medium that has no chemical or physical counterpart in a natural, equilibrium environment. Plasmas alter the normal pathways through which chemical systems evolve from one stable state to another, thus providing the potential to produce materials with properties that are not attainable by any other means.
Applications of plasma-based systems used to process materials are diverse because of the broad range of plasma conditions, geometries, and excitation methods that may be used. The scientific underpinnings of plasma applications are multidisciplinary and include elements of electrodynamics, atomic science, surface science, computer science, and industrial process control. Because of the diversity of applications and the multidisciplinary nature of the science, scientific understanding lags technology. This report highlights this critical issue.
A summary of the many industrial applications of plasma-based systems for processing materials is included in Chapter 2. Electronics and aerospace are the two major industries that are served by plasma processing technologies, although the automotive industry is likely to become a significant user of plasma-processed materials like those now in widespread use in the aerospace industry. The critical role of plasma processing technology in industry is illustrated in Chapter 2.
For the electronics industry more than for any other considered by the panel, the impact of—and the critical and urgent need for—plasma-based materials processing is overwhelming. Thus Chapter 3 further elucidates plasma processing of electronic materials and, in particular, the use of plasmas in fabricating microelectronic components. The plasma equipment industry is an integral part of the electronics industry and has experienced dramatic growth in recent years because of the increasing use of plasma processes to meet the demands of fabricating devices with continually shrinking dimensions. In this country, the plasma equipment industry
is composed of many small companies loosely connected to integrated circuit manufacturers. In Japan, on the other hand, equipment vendors and device manufacturers are tightly linked and are often parts of the same company.
Plasma processes used today in fabricating microelectronic devices have been developed largely by time-consuming, costly, empirical exploration. The chemical and physical complexity of plasma-surface interactions has so far eluded the accurate numerical simulation that would enable process design. Similarly, plasma reactors have also been developed by trial and error. This is due, in part, to the fact that reactor design is intimately intertwined with the materials process for which it will be used. Nonetheless, fundamental studies of surface processes and plasma phenomena—both experimental and numerical—have contributed to process development by providing key insights that enable limitation of the broad process-variable operating space. The state of the science that underpins plasma processing technology in the United States is outlined in Chapter 4. Although an impressive arsenal of both experimental and numerical tools has been developed, significant gaps in understanding and lack of instrumentation limit progress.
The broad interdisciplinary nature of plasma processing is highlighted in the discussion of education issues outlined in Chapter 5, which addresses the challenges and opportunities associated with providing a science education in the area of plasma processing. For example, graduate programs specifically focused on plasma processing are rare because of insufficient funding of university research programs in this field. By contrast, both Japan and France have national initiatives that support education and research in plasma processing.
FINDINGS, CONCLUSIONS, AND RECOMMENDATIONS
Finding and Conclusion: In recent years, the number of applications requiring plasmas in the processing of materials has increased dramatically. Plasma processing is now indispensable to the fabrication of electronic components and is widely used in the aerospace industry and other industries. However, the United States is seeing a serious decline in plasma reactor development that is critical to plasma processing steps in the manufacture of VLSI microelectronic circuits. In the interest of the U.S. economy and national defense, renewed support for low-energy plasma science is imperative.
Finding and Conclusion: The demand for technology development is outstripping scientific understanding of many low-energy plasma processes. The central scientific problem underlying plasma processing concerns the interaction of low-energy collisional plasmas with solid surfaces. Understanding this problem requires knowledge and expertise drawn from plasma physics, atomic physics, condensed matter physics, chemistry, chemical engineering, electrical engineering, materials science, computer science, and computer engineering. In the absence of a coordinated approach, the diversity of the applications and of the science tends to diffuse the focus of both.
Finding: Technically, U.S. laboratories have made many excellent contributions to plasma processing research—making fundamental discoveries, developing numerical algorithms, and inventing new diagnostic techniques. However, poor coordination and inefficient transfer of insights gained from this research have inhibited its use in the design of new plasma reactors and processes.
Finding: The Panel on Plasma Processing of Materials finds that plasma processing of materials is a critical technology that is necessary to implement key recommendations contained in the National Research Council report Materials Science and Engineering for the 1990s (National Academy Press, Washington, D.C., 1989) and to enhance the health of technologies as identified in Report of the National Critical Technologies Panel (U.S. Government Printing Office, Washington, D.C., 1991). Specifically, plasma processing is an essential element in the synthesis and processing arsenal for manufacturing electronic, photonic, ceramic, composite, high-performance metal, and alloy materials.
Accordingly, the panel recommends:
Plasma processing should be identified as a component program of the Federal Initiative on advanced materials synthesis and processing that currently is being developed by the Office of Science and Technology Policy.
Through such a Plasma Processing Program, federal funds should be allocated specifically to stimulate focused research in plasma processing, both basic and applied, consistent with the long-term economic and defense goals of the nation.
The Plasma Processing Program should not only provide focus on common goals and promote coordination of the research performed by the national laboratories, universities, and industrial laboratories, but also integrate plasma equipment suppliers into the program.
Finding and Conclusion: Currently, computer-based modeling and plasma simulation are inadequate for developing plasma reactors. As a result, the detailed descriptions required to guide the transfer of processes from one reactor to another or to scale processes from a small to a large reactor are not available. Until we understand how geometry, electromagnetic design, and plasma-surface interactions affect material properties, the choice of plasma reactor for a given process will not be obvious, and costly trial-and-error methods will continue to be used. Yet there is no fundamental obstacle to improved modeling and simulation nor to the eventual creation of computer-aided design (CAD) tools for designing plasma reactors. The key missing ingredients are the following:
A reliable and extensive plasma data base against which the accuracy of simulations of plasmas can be compared. Plasma measurement technologies are sophisticated, but at present experiments are performed on a large variety of different reactors under widely varying conditions. A coordinated effort to diagnose simple, reference reactors is necessary to generate the necessary data base for evaluation of simulation results and to test new and old experimental methodology.
A reliable and extensive input data base for calculating plasma generation, transport, and surface interaction. The dearth of basic data needed for simulation of plasma generation, transport, and surface reaction processes results directly from insufficient generation of data, insufficient data compilation, insufficient distribution of data, and insufficient funding of these activities. The critical basic data needed for simulations and experiments have not been prioritized. For plasma-surface interactions, in particular, lack of data has precluded the formation of mechanistic models on which simulation tools are based. Further experimental studies are needed to elucidate these mechanisms.
Efficient numerical algorithms and supercomputers for simulating magnetized plasmas in three dimensions. The advent of unprecedented supercomputer capability in the next 5 to 10 years will have a major impact in this area, provided that current simulation methods are expanded to account for multidimensional effects in magnetized plasmas.
Accordingly, the panel recommends:
The Plasma Processing Program should include a thrust toward development of computer-aided design tools for developing and designing new plasma reactors.
The Plasma Processing Program should emphasize a coordinated approach toward generating the diagnostic and basic data needed for improved plasma and plasma-surface simulation capability.
A program to extend current algorithms for plasma reactor simulation should be included among the activities funded under the umbrella of the federal High
Performance Computing and Communications program1developed by the Office of Science and Technology Policy and started in FY92.
Finding and Conclusion: In the coming decade, custom-designed and custom-manufactured chips, i.e., application-specific integrated circuits (ASICs), will gain an increasing fraction of the world market in microelectronic components. This market, in turn, will belong to the flexible manufacturer who uses a common set of processes and equipment to fabricate many different circuit designs. Such flexibility in processing will result only from real understanding of processes and reactors. On the other hand, plasma processes in use today have been developed using a combination of intuition, empiricism, and statistical optimization. Although it is unlikely that detailed, quantitative, first-principles-based simulation tools will be available for process design in the near future, design aids such as expert systems, which can be used to guide engineers in selecting initial conditions from which the final process is derived, could be developed if gaps in our fundamental understanding of plasma chemistry were filled.
Finding and Conclusion: Three areas are recognized by the panel as needing concerted, coordinated experimental and theoretical research: surface processes, plasma generation and transport, and plasma-surface interactions. For surface processes, studies using well-controlled reactive beams impinging on well-characterized surfaces are essential for enhancing our understanding and developing mechanistic models. For plasma generation and transport, chemical kinetic data and diagnostic data are needed to augment the basic plasma reactor CAD tool. For studying plasma-surface interactions, there is an urgent need for in situ analytical tools that provide information on surface composition, electronic structure, and material properties.
Finding and Conclusion: Breakthroughs in understanding the science will be paced by development of tools for the characterization of the systems. To meet the coming demands for flexible device manufacturing, plasma processes will have to be actively and precisely controlled. But today no diagnostic techniques exist that can be used unambiguously to determine material properties related to device yield. Moreover, the parametric models needed to relate diagnostic data to process variables are also lacking.
According, the panel recommends:
The Plasma Processing Program should be dedicated in part to the development of plasma process expert systems.
A coordinated program should be supported to generate basic data and simulation of surface processes, plasma generation and transport, and plasma-surface interactions.
A program should be supported that focuses on development of new instrumentation for real-time, in situ monitoring for control and analysis.
Finding: Research resources in low-energy plasma science in the United States are eroding at an alarming rate. U.S. scientists trained in this area in the 1950s and early 1960s are retiring or are moving to other areas of science for which support is more forthcoming. When compared to those in Japan and France, the U.S. educational infrastructure in plasma processing lacks focus, coordination, and funding. As a result, the United States will not be prepared to maintain its leading market position in plasma processing, let alone capture more market share as the plasma process industry grows into the 21st century.
Finding: Graduate programs are not offering adequate educational opportunities in the science of weakly ionized, highly collisional plasmas. An informal survey by the panel indicated that only a few U.S. universities offer formal course work in this science and that there are
insufficient texts on collisional plasmas and plasma processing. These deficiencies are a direct result of low-level funding for graduate research in plasma processing and low-energy plasmas.
Finding and Conclusion: The most serious need in undergraduate education is adequate, modern teaching laboratories. Due to the largely empirical nature of many aspects of plasma processing, proper training in the traditional scientific method, as provided in laboratory classes, is a necessary component of undergraduate education. The Instrumentation and Laboratory Improvement Program sponsored by the National Science Foundation has been partly successful in fulfilling these needs, but it is not sufficient.
Finding and Conclusion: Research experiences for undergraduates made available through industrial cooperative programs or internships are essential for high-quality technical education. But teachers and professors themselves must first be educated in low-energy plasma science and plasma processing before they can be expected to educate students. Industrial-university links can also help to impart a much needed, longer-term view to industrial research efforts.
Accordingly, the panel recommends:
As part of the Plasma Processing Program, government and industry together should support cooperative programs specific to plasma processing with universities and national laboratories.
A program should be established to provide industrial internships for teachers and professors in the area of plasma processing.