II. Materials R&D in a Changing World

Report of the Committee on Condensed-Matter and Materials Physics

Venkatesh Narayanamurti

Harvard University

Chair, Committee on Condensed-Matter and Materials Physics

Condensed-matter and materials physics (CMMP) plays a central role in many of the scientific and technological advances that have changed our lives so dramatically in the last 50 years. CMMP gave birth to the transistor, the integrated circuit, the laser, and low-loss optical fibers so important to the modern computer and communication industries. The years ahead promise equally dramatic advances, making this an era of great scientific excitement for research in the field. Communicating this excitement and ensuring further progress are the main goals of the CMMP report.

Over the decade since the last major assessment of the field, important results and discoveries have come rapidly and often in unexpected ways. These advances range from development of new experimental tools for atomic-scale manipulation and visualization, to creation of new synthetic materials (such as bucky balls and high-temperature superconductors), to discovery of new physical phenomena such as giant magnetoresistance and the fractional quantum Hall effect.

An enormous increase in computing power has yielded qualitative changes in visualization and simulation of complex phenomena in large-scale many-atom systems. Progress in synthesis, visualization, manipulation, and computation will continue to have an impact on many areas of research spanning different length scales from atomic to macroscopic. Strong impact may also be expected in “soft” condensed-matter physics, particularly at the interfaces with biology and chemistry.

The priorities of society are shifting from military security to economic well-being and health. Changing societal priorities, in turn, create shifting demands on CMMP. Among these growing demands are improving public understanding of science, allowing better education of scientists and engineers for today's employment marketplace, and making new contributions to the nation's industrial competitiveness.

The key challenges facing condensed-matter and materials physics are the following:

  • Nurturing the intellectual vitality of the field—particularly the facilitation of the research of individual investigators and small teams in areas that cross disciplinary boundaries;

  • Providing the facilities infrastructure for research—for example, creation of laboratory-scale microcharacterization facilities at universities and large-scale facilities at national laboratories;

  • Enhancing efforts in research universities to improve integration of CMMP education and research, particularly at the boundaries of disciplines, and to prepare flexible and adaptable physicists for the future; and

  • Developing new modes of cooperation among universities, colleges, government laboratories, and industry to ensure the connectivity of the field with the needs of society and to preserve the fertile, innovative climate of major industrial laboratories, which have played a dominant role in CMMP research.

The different modes of research—benchtop experiments, larger collaborations, and so on—are evolving steadily. The work that is carried on in these varied venues is complex and diverse, and the committee has paid special attention to describing the forefronts of research in terms of a small number of research themes. These themes, listed in Box 1, are discussed in some detail in the Overview of the CMMP report and reappear in each of the chapters of the report.

One of the themes that has captured the imagination of theorists and experimenters alike is the structure and properties of materials at reduced dimensionality—for example, in planar structures. Large-scale integrated circuits depend on understanding the behavior of semiconductors in such configurations, so the potential for impact is apparent.

NOTE: This article was prepared from written material provided to the Solid State Sciences Committee by the speaker.



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 25
Materials in a New Era: Proceedings of the 1999 Solid State Sciences Committee Forum II. Materials R&D in a Changing World Report of the Committee on Condensed-Matter and Materials Physics Venkatesh Narayanamurti Harvard University Chair, Committee on Condensed-Matter and Materials Physics Condensed-matter and materials physics (CMMP) plays a central role in many of the scientific and technological advances that have changed our lives so dramatically in the last 50 years. CMMP gave birth to the transistor, the integrated circuit, the laser, and low-loss optical fibers so important to the modern computer and communication industries. The years ahead promise equally dramatic advances, making this an era of great scientific excitement for research in the field. Communicating this excitement and ensuring further progress are the main goals of the CMMP report. Over the decade since the last major assessment of the field, important results and discoveries have come rapidly and often in unexpected ways. These advances range from development of new experimental tools for atomic-scale manipulation and visualization, to creation of new synthetic materials (such as bucky balls and high-temperature superconductors), to discovery of new physical phenomena such as giant magnetoresistance and the fractional quantum Hall effect. An enormous increase in computing power has yielded qualitative changes in visualization and simulation of complex phenomena in large-scale many-atom systems. Progress in synthesis, visualization, manipulation, and computation will continue to have an impact on many areas of research spanning different length scales from atomic to macroscopic. Strong impact may also be expected in “soft” condensed-matter physics, particularly at the interfaces with biology and chemistry. The priorities of society are shifting from military security to economic well-being and health. Changing societal priorities, in turn, create shifting demands on CMMP. Among these growing demands are improving public understanding of science, allowing better education of scientists and engineers for today's employment marketplace, and making new contributions to the nation's industrial competitiveness. The key challenges facing condensed-matter and materials physics are the following: Nurturing the intellectual vitality of the field—particularly the facilitation of the research of individual investigators and small teams in areas that cross disciplinary boundaries; Providing the facilities infrastructure for research—for example, creation of laboratory-scale microcharacterization facilities at universities and large-scale facilities at national laboratories; Enhancing efforts in research universities to improve integration of CMMP education and research, particularly at the boundaries of disciplines, and to prepare flexible and adaptable physicists for the future; and Developing new modes of cooperation among universities, colleges, government laboratories, and industry to ensure the connectivity of the field with the needs of society and to preserve the fertile, innovative climate of major industrial laboratories, which have played a dominant role in CMMP research. The different modes of research—benchtop experiments, larger collaborations, and so on—are evolving steadily. The work that is carried on in these varied venues is complex and diverse, and the committee has paid special attention to describing the forefronts of research in terms of a small number of research themes. These themes, listed in Box 1, are discussed in some detail in the Overview of the CMMP report and reappear in each of the chapters of the report. One of the themes that has captured the imagination of theorists and experimenters alike is the structure and properties of materials at reduced dimensionality—for example, in planar structures. Large-scale integrated circuits depend on understanding the behavior of semiconductors in such configurations, so the potential for impact is apparent. NOTE: This article was prepared from written material provided to the Solid State Sciences Committee by the speaker.

OCR for page 25
Materials in a New Era: Proceedings of the 1999 Solid State Sciences Committee Forum BOX 1: Research Themes in CMMP The quantum mechanics of large, interacting systems The structure and properties of materials at reduced dimensionality Materials with increasing levels of compositional, structural, and functional complexity Nonequilibrium processes and the relationship between molecular and mesoscopic properties Soft condensed matter and the physics of large molecules, including biological structures Controlling electrons and photons in solids on the atomic scale Understanding magnetism and superconductivity A number of actions are required to maintain and enhance the productivity of the field of condensed-matter and materials physics. These actions involve each level of the hierarchy of research modalities and the interactions among the various levels and the various performers. The principal recommendations of the cmmittee are summarized as follows: The National Science Foundation (NSF), the Department of Energy (DOE), and other agencies that support research should continue to nurture the core research that is at the heart of condensed-matter and materials physics. The research themes listed in Box 1 provide a guide to the forefronts of this work. The agencies that support and direct research in CMMP should plan for increased investment in modernization of the CMMP research infrastructure at universities and government laboratories. The NSF should increase its investment in state-of-the-art instrumentation and fabrication capabilities, including centers for instrumentation R&D, nanofabrication, and materials synthesis and processing at universities. The DOE should strengthen its support for such programs at national laboratories and universities. The gap in neutron sources in the United States should be addressed in the short term by upgrading existing neutron-scattering facilities and in the longer term by moving forward with the construction of the Spallation Neutron Source. Support for operations and upgrades at synchrotron facilities, including research and development on fourth-generation light sources, should be strengthened. The broad utilization of synchrotron and neutron facilities across scientific disciplines and sectors should be considered when agency budgets are established. Federal agencies should provide incentives for formarion of partnerships among universities and government and industry laboratories that carry out research in condensed-matter and materials physics. Universities should endeavor to enhance their students' understanding of the role of knowledge integration and transfer as well as knowledge creation. In this area, experience is the best teacher. Action on these issues will allow us to capture the opportunities for intellectual progress and technological impact that continue to emerge in condensed-matter and materials physics. Materials R&D in Industry Cherry A. Murray Bell Laboratories, Lucent Technologies The Changing Role for Physical Science Research in Industry in the “Information Age” Today industry funds about two-thirds of total U.S. R&D, amounting to nearly $130 billion in 1997. The U.S. government supplies the remaining one-third of the funds spent on R&D in the United States (nearly $65 billion). Of federal funding in 1997, about 35 percent (roughly $23 billion) went directly to industry, 28 percent to national laboratories, 22 percent to universities, and the remainder to federally funded R&D centers and nonprofit organizations. Today, total federal funding of industrial R&D has declined to about three-fifths of its high point of a decade ago. “Blue sky” corporate research has declined sharply in this decade in all economic sectors, to about one-tenth to one-third of its 1988 extent, depending on the company. Also NOTE: This article was prepared from written material provided to the Solid State Sciences Committee by the speaker.

OCR for page 25
Materials in a New Era: Proceedings of the 1999 Solid State Sciences Committee Forum companies are attempting to measure and monitor more closely the output of both research and development, linking both to new product development. I define “blue sky” research as research that is not linked in any way to a possible product or is at least 15 years away from becoming a product. An emerging trend in the last several years that offsets this bleak picture for the future of industrial research is the recently renewed support for long-term research in the information technology (IT) sector. I define long-term research as that which could be at least 5 to l5 years out but can be linked to possible future potential products of interest to the company. R&D spending in the United States varies dramatically by economic sector: About 20 percent of corporate revenues are spent for R&D in the pharmaceutical sector, from 15 percent to 18 percent in the IT software sector, from 10 percent to 15 percent in the IT hardware sector, about 5 percent in the chemical/materials sector, at most 5 percent in the automotive/transportation sector, and only 1 to 2 percent in the energy/power sector of the economy. In this article, I will focus on the fast-growing IT sector, whose revenues constitute about 10 percent of the current gross domestic product. This includes U.S.-based companies such as AT&T, IBM, Motorola, Lucent, General Electric, Intel, Lotus Development, Microsoft, Silicon Graphics, Bay Networks, Adobe Systems, Tandem Computers, and so on. In this sector in 1967, hardware products accounted for 89 percent of the revenues; software, 2.6 percent; and services, 8 percent. Over the years, those percentages have dramatically changed, with the current balance more strongly favoring faster-growing software and services over hardware: In 1996, 50 percent of the revenues were related to hardware, 20 percent to software, and 30 percent to services. IT corporations have reacted by rebalancing their hardware (physical sciences) versus software and services (mathematics and computer science) research mix. For example, at Bell Laboratories, the central corporate research has evolved from a traditional hardware:software 70:30 split in the 1970s to closer to 50:50 today. Some of the trends in corporate research in the IT sector can be summarized as follows: The years of the 1970s and 1980s were the era of the Cold War, the last of Vannevar Bush's “New Frontier,” and the age of the hardware near-monopolies (the Bell System, IBM, GE, and so on). The justification of corporate research could be characterized as “just in case.” Blue sky research in the physical sciences flourished in industry. During the mid-1980s to the 1990s, the Cold War ended; monopolies such as AT&T, IBM, and GE broke apart; software grew in importance; many corporations retrenched; and major U.S. consortia such as Sematech, SRC, MONET, and NSIC were formed in an attempt to pool resources for precompetitive research and involve universities, as well as to train a technical work force necessary for development and manufacturing. The justification for corporate research can be characterized in this era as “just in time.” Many corporate central research laboratories were broken up and distributed to the various business units and focused on shorter-term product development. In many accounts, this created fears of the existence of a “valley of death” for research on products that are 5 to 10 years in the future: Companies were focusing most of their efforts on the period 0 to 5 years out, and universities and government laboratories were focusing on the period beyond 10 years out. In the silicon integrated circuit industry, this has been the justification for the formation of industry-university-federal laboratory consortia such as MARCO (the Microelectronics Advanced Research Consortium) to fill this gap. In the late 1990s into 2000, because of the technological advances of this sector of the economy, we have entered the “Information Age ”: Corporations are global, there is exponential growth in both technology and profits for the IT sector, there is strong competition in hardware while new monopolies emerge in software and speed-to-market is essential, and the gap is closing between research and products while hardware information technologies are approaching their fundamental limits. Now, because of the exponential trends in all information technology, devices are becoming smaller and faster and acquiring more functionality, all at lower cost, and are rapidly approaching real fundamental limits. The justification for corporate physical sciences research for much of the hardware IT sector is that it is “just indispensable.” If the corporation wants to be a technology leader, being first to market is viewed as critical. This speed-to-market requires a much tighter coupling of research to products combined with a longer-term in-house research effort —allowing the fastest innovation to occur while avoiding being blindsided by competitors. We will be approaching some fundamental limits in the next 20 years (around the year 2010). For example, silicon device scaling will produce metal-oxide semiconductor field-effect transistors with gate oxide thickness of less than 5 atoms, magnetic data storage spot sizes will approach the paramagnetic limit, and transmission of optical pulses through optical fi-

OCR for page 25
Materials in a New Era: Proceedings of the 1999 Solid State Sciences Committee Forum bers will be approaching the Shannon information theory limit. Many companies in the hardware IT sector are actually increasing their support of internal long-term research as a result. Why do companies spend money on internal R&D? If they are in the leading-edge technology market for the long haul and have evolved past the startup phase, they must maintain critical competencies in house, maintain an infusion of new technology, stimulate innovation, fuel growth and business development, extend their product horizons, and recruit top people. It is difficult to accomplish these objectives by funding external research at a university or within a consortium or by buying small companies. Companies spend money on external or cooperative R&D to leverage their internal R &D efforts, develop new applications for existing technology, make use of facilities and equipment that are too costly to develop internally, and acquire access to a skilled work force in the technologies relevant to their products. The major problems encountered by corporations in carrying out external or cooperative R&D at universities or government laboratories are the complications in intellectual property ownership and licensing, the relatively slow cycle times, and the focus on process rather than product that is natural for universities and government laboratories. A 1999 Battelle R&D magazine survey of all industries found that a large majority of respondents, 37 percent, used joint development agreements with other companies for external R&D and 23 percent purchased services from commercial laboratories. Only about 26 percent of respondents used academia for external R &D and about 3.5 percent used federal laboratories. Why do companies spend money on internal long-term research? There are many reasons, depending on the competitive environment and growth of the economic sector. First, internal research stimulates invention leading to innovation. It provides insurance—it allows a company to maintain a breadth of technological expertise to make use of when it is suddenly needed. When integrated into the R&D community, long-term research can broaden horizons and provides a future beyond several product cycles. Often an internal research organization can be useful in recruiting top people into the business, enthralling customers, and challenging competitors. Internal research also allows companies to keep trade secrets and create a strong intellectual property portfolio, which is essential to become and stay a technology leader. Two examples where physical science and materials research paid off, causing a factor of two increase in logarithmic slope in the technology “Moore” plots for magnetic data storage and optical networking, respectively, are the invention and development of giant magnetoresistance materials in magnetic read-heads by IBM research and the development of the Er-doped optical amplifier by Bell Laboratories research. There are also many other examples where long-term industrial research has resulted in a paradigm shift in technology and business opportunity for the parent company because it was the first to get products out on the market. In the 1980s, the chain from physical sciences research to products in industry was either nonexistent (blue sky research) or linear —from applied research in the corporate research organization handing off to a development organization handing off to a manufacturing organization who would eventually create a product and be in contact with the customers. At every handoff along the way, there were roadblocks and bottlenecks—the entire process could take as long as 5 to 10 years, with a relatively low success rate of getting through the whole process and little communication along the way. That approach no longer is a viable way of innovation. In a typical corporate research group, there are at least four types of research: Long-term research in areas related to future technology needs, often in consultations with customers and marketing; Cooperative long-term research with federal laboratories and/or universities not directly related to near-term future products; Cooperative short- and long-term R&D with other companies in joint development agreements; and A “massively parallel” approach to marketing, research, development, and manufacturing that brings a closely knit team of people from all organizations together to produce a product from a research concept in as little time as months to a few years, maintaining close customer contact and competitor awareness at all times. In summary, industrial R&D is what has created the “Information Age.” IT corporate research has evolved over the decades, but physical sciences research is as essential as ever for leading-edge high-technology companies.

OCR for page 25
Materials in a New Era: Proceedings of the 1999 Solid State Sciences Committee Forum Changing Roles for Research Universities J. David Litster Massachusetts Institute of Technology MIT: A University in Transition David Litster discussed the changing environment for research at the Massachusetts Institute of Technology (MIT). He pointed out that MIT is in some ways unique among the top 20 research universities in the United States. Of this group, it receives the largest amount of industrial research support and was among the lowest in self-support of research. For these reasons, he warned that one must be careful not to generalize the observations that he made. The overall federal funding picture, excluding the National Institutes of Health (NIH), is that of flat or decreasing budgets. The Department of Defense (DOD), which supported some 65 percent of materials engineering research between 1993 and 1995, has experienced an overall decrease in funding throughout the 1990s. Its funding for research has not been spared and has fallen from about $16 billion in 1989 to about $9 billion in 1999. DOD-sponsored research at MIT has fared somewhat better, remaining relatively level at about $35 million (in 1993 dollars). The flat or declining budgets of the federal agencies has put enormous pressure on overhead recovery rates at MIT. Between 1980 and 1990, the amount of indirect costs at MIT covered by the federal government hovered around 50 to 56 percent, whereas the amount covered by internal MIT sources was between 32 and 38 percent (the remainder was picked up by state sources). Since then, the proportion of the indirect costs borne by MIT has climbed steadily while the federal share has plummeted. The change was so rapid that by 1996 MIT was covering about 52 percent of the costs and the funding agencies were covering only about 36 percent—a complete reversal of roles. Furthermore, during the 1980s, the federal government cut its financial aid for students. This trend is true for all schools that continue a need-blind admission policy. In 1980, MIT supplied about 50 percent of the financial aid to students. By 1990, MIT's share had grown to about 80 percent and has remained there since. One thing MIT has done to counter these trends is to engage in partnerships with industry. Industrial partners who support research at MIT include Amgen, Ford Motor Company, Merck, and Merrill Lynch. MIT Research Support Industrial Partnerships are long term (5 to 10 years). The support provided by these partners amounts to about $3 million per year. A joint committee of MIT and industry representatives allocates $2.5 million of this. The remaining $0.5 million is a “discretionary” fund. The normal MIT policies on research support are followed in these partnerships. This means that the research should be of intellectual interest to the principal investigator on the project. The principal investigator is responsible for directing the project, which should provide some mix of thesis opportunities for students, the advancement of knowledge, or advancement of the state of the art. Any visiting scholars on the project are chosen by the faculty and are expected to make significant contributions to the research project. Ideally, these industrially sponsored research projects should balance MIT 's educational purposes and the search for knowledge to meet the needs of industry. The results of these partnership projects must be freely published —this is a requirement for MIT to maintain its tax-exempt status. Thus the results are available to anyone, regardless of the source of the research support. There is no delay in giving students academic credit for the work; however, to protect patent rights, publication of the results may be delayed by as much as 30 days (60 days in extreme circumstances). The industrial sponsor approves any thesis proposal and agrees in advance that anything falling within the proposal can be freely published. The sponsor has 30 days to review the thesis and publications to ensure that they contain no proprietary information. MIT retains title to all intellectual property developed by employees of MIT using significant funds or facilities administered by the university. All research sponsorship agreements at MIT are negotiated by the university and, regardless of the sponsor, transfer the rights to intellectual property to MIT. The university, in turn, licenses the intellectual property to encourage technology transfer for development by industry in the public interest. Intellectual property that is developed by visiting scholars also belongs to MIT. If products produced under license from MIT are to be sold in the United States, then MIT requires a substantial amount of the manufacturing of that product to NOTE: This article was prepared from notes taken by a staff member of the Board on Physics and Astronomy.

OCR for page 25
Materials in a New Era: Proceedings of the 1999 Solid State Sciences Committee Forum be carried out in the United States. The royalty split is one-third to the inventor before expenses; the remaining after-expense income is divided equally between the inventor's department or laboratory and the central administration of MIT. The sponsor may have a nonexclusive royalty-free license to these inventions for internal use. The sponsor may also obtain exclusive commercial rights and pay royalties to MIT, or the sponsor may obtain a nonexclusive commercial license in exchange for paying the patent maintenance costs, or the sponsor may waive all rights and receive 25 percent of the after-expense income from the patent. If MIT chooses not to file a patent, then the sponsor has the right to do so in MIT's name. The sponsor must then choose from the four options listed above. Changing Roles for Government Laboratories John P. McTague Vice President (retired), Ford Motor Company John McTague, in introducing his subject, declared that he would focus primarily on trends within the Department of Energy (DOE) laboratories, a focus sharpened by the perspectives that he has gained as co-chair of the Secretary of Energy's Laboratory Operations Board. He noted that the DOE laboratory complex has received considerable criticism over recent years, primarily in two areas: (1) for alleged, or perceived, duplication of effort within the laboratories; and (2) for the laboratories ' tendency (in view of ever-tightening budgets and decreasing programmatic support in some of their traditional areas) toward “mission creep. ” Laboratories invent new missions—in particular, the formerly popular and politically correct mission of “industrial competitiveness,” which many of the DOE laboratories embraced as the Cold War came to an end and as the defense-driven support for R&D began to wane. These criticisms and the associated perception of DOE's inefficient management of its laboratories were highlighted in the Galvin Commission Report of 1994. McTague noted that the aforementioned critics are at least somewhat off the mark in the sense that the laboratories do not have missions from Congress—it is the Department of Energy that has the missions. The laboratories are premiere among the DOE's resources to execute its missions. Therefore, it is the Department of Energy's job to bring the laboratories, universities, and other R&D providers to bear, both collectively and individually, to accomplish its missions. A challenge that has been gaining increasing acceptance and discussion in recent years is how to get the laboratories to operate as a true system and, more generally, how laboratories and other partners can work more effectively together to attack problems of national importance. In attempting to address this question, McTague described four examples of more-or-less successful orchestration of the suites of instruments represented by these laboratories and their partners. The first example is the Center of Excellence for the Synthesis and Processing of Advanced Materials. It is a virtual center, directed by George Samara of Sandia National Laboratories, and involves 12 DOE laboratories, as well as several industries and some industrial partners. The idea is, with a modest incremental investment of only about $2.5 million a year, to provide value-added, enhanced coupling among projects of related natures that are already taking place within these many laboratories. Selective projects from within their suite of capabilities are coordinated and joined together, so that the whole exceeds the sum of the parts. This has been a very successful enterprise, and McTague identified some key elements of that success through the example of aluminum alloy formability. In the auto industry, which worldwide produces something like two vehicles per second, a small advance can have a huge integrated impact. The center has achieved such an advance in this highly leveraged application. Similar opportunities arise in the areas of joining and welding, and McTague described some examples of using a transparent welding analog where the effect of weld freeze rates on joint shapes can be directly observed. One important attribute of this center's approach is that it is multidimensional in the sense that the number of component projects is large enough that statistically a reasonable number of them will succeed, and the success of the whole project does not depend on every single element being successful on its own. This fairly loose coupling among the projects minimizes, therefore, the chances of overall failure and gives high leverage to the added value of the investment in affect-

OCR for page 25
Materials in a New Era: Proceedings of the 1999 Solid State Sciences Committee Forum ing the coupling. McTague's second example is “The Partnership for a New Generation of Vehicles,” an extremely large project costing approximately $600 million a year. It involves USCAR (a consortium of the Big Three automakers in the United States) and a Department-of-Commerce-led government consortium of the Departments of Commerce, Defense, Energy, Transportation, and Interior, as well as the Environmental Protection Agency, the National Aeronautics and Space Administration, and the National Science Foundation (NSF). Industry has a very large influence on how the government participants allocate their resources in this case. The good news of this project is the high parallelism of interest among all the partners in reaching the stated goal of achieving a 300 percent increase in fuel efficiency in the next generation of vehicles. The specific goals are stated as outcomes, but the paths to those outcomes are not specified. So, for example, the flywheel approach has been tried and abandoned. Fuel cells, on the other hand, look considerably more promising. The bad news is that there is an extremely high overhead as a result of working together because there are so many players. Indeed, it took 18 months to get the detailed agreements and working relationships in place for this collaboration. Nevertheless, it represents a collaboration at an interagency level, involving many of the Department of Energy laboratories and funding from several of the agencies. It is a good example of the kind of orchestration that is required for large advances in technology. The third example is the Spallation Neutron Source project. This $1.3 billion facility, to be built at Oak Ridge National Laboratories, involves a collaboration among five DOE laboratories. The major risk and disadvantage of this project, aside from its considerable expense and complexity, is that it is “one-dimensional.” Every single element, from the source through the accelerator, to the accumulator ring to the target station to the laboratory instruments is in series. The overall project can succeed if, and only if, every one of the elements, namely every one of the laboratories, does its job 100 percent correctly. It is questionable whether proper orchestration of such an effort is really possible given the current structure of the laboratories and the Department of Energy. This project lacks what McTague would call the “statistical safety” of his first two examples. It is very much an open and important question as to whether the five DOE laboratories in this case can operate as a system. The answer is that they simply have to. The question remains, can they? The fourth example is a major new initiative in information technology. “Information Technology for the 21st Century,” or IT2, adds approximately $150 million to the research portfolio —primarily in the NSF, but with additional funding for several of the other agencies as well. The idea is to push forward the development of distributed and highly interconnected computing capabilities. It is very important because it might lead to a revolution in engineering and product development, including safety, through the use of more realistic simulations. McTague mentioned, as an aside, that at the Ford Motor Company, simulation has become an integral part of design and also product worthiness. Indeed the company is now at point where the simulation of an automobile crash can be “more accurate than the experiment” in the sense that in the simulation you can keep track, reproducibly and quantitatively, of every element of the vehicle and the event, whereas an experiment must typically be repeated many times without assurance that the exact initial conditions are reproduced or that measurements are done in sufficient detail. The IT2 initiative resembles the DOE's major simulation advances in the Accelerated Strategic Computing Initiative Program and the Strategic Science Initiative. Even within the DOE, those two projects are really not being coordinated. The involvement of the DOE with the NSF and other agencies in this major IT2 initiative is still incompletely formed. Indeed the ownership and leadership of the whole IT2 enterprise are still unsettled. Even with these caveats, though, McTague felt that the prospects for success are reasonable because the project is multidimensional. In other words, the success of the whole enterprise does not depend on the success of every single element working to perfection. The elements in this case are not catastrophically interdependent. McTague concluded by noting that he felt cautiously optimistic that the agencies and the laboratories would develop more coherent and fully orchestrated ways of working together and that they might in fact evolve from a collection to a system in the foreseeable future. During the questions following his talk, David Litster of MIT noted that the DOE's problems include having a number of “associate conductors” in their orchestra in the form of program managers within the headquarters and also at the field offices. McTague's comment was that the DOE often lacks the necessary in-house technical expertise among its managers, and therefore it is not managing the laboratories as a sys-

OCR for page 25
Materials in a New Era: Proceedings of the 1999 Solid State Sciences Committee Forum tem. The Laboratory Operations Board has been examining the question of laboratory governance for a couple of years now, and a report on this is being prepared with a set of recommendations. At present, the preliminary grade is no better than a C+ in improving the systems approach to laboratory management by the DOE, and the rate of progress has been slow. Murray Gibson of the University of Illinois asked about the notion of “corporatization” of the laboratories, as suggested by the Galvin Report. McTague 's response was that the Galvin Report really indicated that it was the management dynamics of the laboratories that is poor, which stems from the government's tendency to focus on inputs as metrics rather than on outcomes. The Galvin Report suggested strongly that this emphasis is backward. Industry measures outputs and outcomes in order to determine its success. The most easily measured is the bottom line in the profit and loss statement. But McTague also noted that the idea of industrial or good business practices in management was the original basis for the concept of the government-owned, contractor-operated approach to the national laboratories. The question with respect to the DOE laboratories then becomes, “Is the path from the present input and compliance-driven focus to output and performance-driven focus adiabatically reversible?” The Galvin Report suggests that the answer is “no,” and the Laboratory Operations Board is trying to figure out ways to approach the answer in the affirmative. Jack Rush raised concerns about the DOE rules that impute increasing liability to the maintenance and operation contractors for the laboratories, implying that these rules inhibit the performance of existing contractors or the participation of new contractors who would do a good job at running the laboratories. McTague's answer was that the DOE “mega rule” allows the nonprofit maintenance and operation contractors to get relief from these increased liability concerns, but because for-profit contractors generally get large fees to run the laboratories, they are expected to address the liabilities. It remains unclear whether the shift of liability burden to the contractor is actually improving performance and lowering costs. He suggested that we need to have more data to examine the cost/benefit analysis of this “mega-rule” approach. Although it is clear that the costs to manage the laboratories have gone up, particularly for environmental safety and health concerns, it is not clear whether the health and safety of the environment in the laboratories have increased correspondingly. Panel Discussion of the Future of Materials R&D Moderator: Tom Russell, Chair, Solid State Sciences Committee Panel: Cherry A. Murray, Committee on Condensed-Matter and Materials Physics; Venkatesh Narayanamurti, Chair, Committee on Condensed-Matter and Materials Physics; Skip Stiles, House Science Committee Minority Staff; William Oosterhuis, Department of Energy; Harlan Watson, House Science Committee Majority Staff; Thomas Weber, National Science Foundation A panel discussion was held at the end of the first day. The panelists were asked to comment on the future of condensed-matter and materials physics research. There was a general discussion about changes in research funding during the previous decade and how the community can cope with them, the need for improved education at all levels, and the future directions of the field. Skip Stiles pointed out that Congress will be living under spending caps for the next 2 fiscal years and that there will be no additional money until then unless Congress lifts the caps. Harlan Watson concurred with Stiles and said that he sees little likelihood that the caps will be lifted—at least in the near term. According to Watson, the Administration's outyear projections for Department of Energy (DOE) research funding will remain flat for the next few years. To meet the flat budgets, the major DOE research facilities may need to cut costs by reducing their operating hours. Thomas Weber pointed out that funding in his division has also been flat. He postulated that if the NSF were a mission agency it would: (1) strengthen NOTE: This article was prepared from notes taken by a staff member of the Board on Physics and Astronomy.

OCR for page 25
Materials in a New Era: Proceedings of the 1999 Solid State Sciences Committee Forum the physical infrastructure, (2) integrate research and education, and (3) promote partnerships. Weber said that, despite some resistance among program managers at first, he sees funding of education as a win-win situation for his division. Venkatesh Narayanamurti pointed out that the NSF is one of the major stakeholders in the field and that although we have come a long way in preparing students to work in interdisciplinary teams, there is still much room for improvement. Cherry Murray pointed out industry's need for highly skilled people and that university training is a key to the success of domestic and global industries. She said that industry often presents a “problem-rich” environment, with exciting science, and pointed out that condensed-matter and materials physics (CMMP) has applications that can be important for the Department of Defense (DOD), the National Science Foundation (NSF), DOE, or industry. Stiles said that the CMMP community can make a compelling argument for increased funding but that the money will go to other needs unless the university and industrial communities present a unified case to Congress on a continuing basis. He pointed to The Physics of Materials as a document with the right tone and shape for Congress and said that much more along that line needs to be done. Narayanamurti expressed a note of caution regarding changes at the DOE. He was concerned that the balance in the number of laboratories is very delicate. He said that if the DOE tries to fund too many laboratories, then rivalry between laboratories working in the same area will drag down the system and that, on the other hand, if the DOE funds too few laboratories, then there will not be enough interaction between them to stimulate progress. In his view, a key component in maintaining this balance is strong technical leadership. Weber was concerned about the need he sees to promote international collaboration without giving away our advantages in research. He sees the main challenges of the future as making links to biologically inspired materials, inventing and improving microscopes and other instruments, making sense of complex phenomena, and producing students interested in science. William Oosterhuis reminded the audience that scientists are involved in CMMP because of the stimulating questions that can be attacked using modern instruments and techniques. He would like to see a strengthening of the neutron infrastructure and new experimental techniques using them. He is concerned with how we can best study the growing array of self-organizing materials. He claimed that one technique is to combine the Scientific Simulation Initiative with combinatorial chemistry to fine-tune the computations. He believes that this combination will allow DOE-funded researchers to do things never before possible in modeling complex materials. Following the panel discussion, the panelists received several questions from the audience. Question: We are in danger of losing a cadre of excellent young researchers. How can this be addressed in an era of flat or declining budgets? Narayanamurti replied that more Young Investigator awards need to be funded. NSF and the Office of Naval Research are funding some, but more needs to be done. Oosterhuis said that the DOE is interested in funding young people but there needs to be a formal program set up at the agency for funding them. Weber said that the NSF has established the Faculty Early Career Development (CAREER) program and that although his division has no formal program, he has set aside a small pool of money to fund a competitive program geared toward a diverse pool of applicants doing risky research. Question: We all heard a lot about DOD's downsizing. Agencies used to look around and say, “If someone else is funding research in topic X, then we don't need to.” Now there are areas that are not being investigated at all. Should agencies get money to pick up the areas that are not being adequately funded? Stiles responded that under the current conditions, science has virtually no voice in Congress. He said that the science community needs to step up and help Congress set its priorities for funding and that it must be an ongoing process. Watson countered that Congress is not the right body to make these decisions. He said that the President 's budget submission is where these priorities are set and the Office of Science and Technology Policy (OSTP) is probably the place to start. Question: The scientific community is fairly effective at pointing out the benefits of doing research in particular fields. Is it less effective at pointing out the consequences of not funding certain research? Stiles replied that if the community were prepared, it could help get more money for research, but the

OCR for page 25
Materials in a New Era: Proceedings of the 1999 Solid State Sciences Committee Forum groundwork must be done. Furthermore, he said that the funding does not need to be tied to a particular crisis. Question: Does the scientific community need a lobby to put pressure on Congress and OSTP? According to Stiles, members of Congress do not need to have a deep understanding of the field. He said that what matters to them is how it touches the lives of people in their districts. Weber reiterated that the community should not focus all of its attention on Congress but should talk to the executive branch because it makes the initial budget proposal. Question: We have heard many times that CMMP contributes to prosperity. How can we get the message across that if our government invests in CMMP, our society will get the largest return on the investment? Stiles asked, “How is Congress getting this message?” He said that someone has to tell people in Congress on an individual level and inform public policy structure—and that information needs to be consistent. Judy Franz, Executive Officer of the American Physical Society, pointed out that the American Physical Society has worked hard to get the message out and that grassroots lobbying is the best way to do this. She added, however, that to be effective it has to be done continuously and making this happen is difficult because the community does not see this activity as an essential part of the life of a scientist. Weber warned that the community needs to be careful in making economic arguments. He said that it is not possible to say how today's funding of research will impact the economy of the future. Stiles advised the community to stay close to Congress because Congress can do things by accident that can harm the community, for example, the recent provision that makes data available to the public. He observed that research is becoming a commodity—we are going from a system of grants to one of contracts. Watson added that there is no free lunch in research and that most researchers get their money from the federal government. Question: What are the statistics on collaborations? Is this a way to get federal funding? Oosterhuis replied that research on significant problems needs to done by teams. He believes that the problems we currently face are more difficult and require input from diverse sources. He is trying to encourage collaboration at DOE laboratories in areas where it makes sense to do so. He pointed out that the DOE has the PAIR program to encourage these interactions. Weber called attention to the GOALI program at the NSF as an example of a program that encourages collaborations. Question: How can we convert S. 1305 from an authorization bill to an appropriations bill? Stiles replied that S. 1305 is a good organizing tool but will not produce more money for science. He said that in the short term, only having Congress declare an emergency or lift the budget caps will accomplish that. He urged the community to ensure that the subcommittees that fund the community's work get all the funding they need. Watson responded that the requests need to get into the Administration' s budget request; otherwise, it will be difficult to get the requests into the final budget. He said that Congress can only “tweak” the numbers around the edges. Furthermore, he said that the community should bear in mind that there are huge numbers of claimants for any pot of money. Because one Congress cannot bind future Congresses, he thinks a large effort at passing “feel-good” legislation is a waste of time and effort.