Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
9 Industrial Laboratories and Research in Condensed- Matter and Materials Physics History of Industrial Research Laboratories The 20th century was an era of large, well-funded corporate research labo- ratories focusing on research in the physical sciences and related engineering disciplines. General Electric (GE) Laboratories founded in 1900, Bell Laboratories founded in 1925, IBM T.J. Watson Laboratories founded in 1945, and Xerox Palo Alto Research Center (PARC) founded in 1970 are examples of major corporate- funded laboratories that encouraged large groups to work on long-range research. Breakthroughs such as x-ray tubes, transistors, lasers, cellular telephones, graphical user interfaces, and other technologies that define the beginning of the 21st century were the result, as summarized in histories of the industrial laboratories. Some Nobel Prizeâwinning contributions from industrial laboratories are summarized in Table 9.1. These powerful corporate laboratories, harboring some of the greatest scien- tists and engineers of the time, were enabled by the market dominance of their parent corporations. GE, Xerox, and IBM dominated lighting, copiers, and com- puters, respectively. Kodak, Polaroid, and Westinghouse had similar histories. As a government-regulated monopoly, AT&T ran the telephone network. These compa- nies generated large amounts of cash that enabled them to invest in the future with fundamental, long-term research. The results have been stunning. Fundamental âRichard S. Rosenbloom and William J. Spencer, eds., Engines of Innovation: U.S. Industrial Research at the End of an Era, Boston, Mass.: Harvard Business School Press, 1996. 165
166 C o n d e n s e d - M at t e r and M at e r i a l s P h ys i c s TABLE 9.1â Some Nobel PrizeâWinning Contributions from Industrial Laboratories Name of Researcher(s) and Activity Corporate Sponsor Date of Prize Surface chemistry GE Laboratories Langmuir, 1932 Electron diffraction Bell Laboratories Davisson and Thomson, 1937 Transistor Bell Laboratories Bardeen, Brattain, and Shockley, 1956 Maser-laser Bell Laboratories/Columbia Townes, Basov, and Prokhorov, 1964 University Quantum tunnel junctions IBM T.J. Watson Laboratories/ Esaki and Giaever, 1973 GE Laboratories Theory of disordered materials Bell Laboratories Anderson, Mott, and van Vleck, 1977 Cosmic microwave background Bell Laboratories Penzias and Wilson, 1978 radiation Scanning tunneling microscopy IBM Zurich Research Laboratory Binnig and Rohrer, 1986 High-temperature IBM Zurich Research Laboratory Bednorz and Mueller, 1987 superconductivity Quantum Hall effect Bell Laboratories Laughlin, Stormer, and Tsui, 1998 Integrated circuit Texas Instruments Kilby, 2000 SOURCE: See http://nobelprize.org. inventions, such as the transistor from AT&T, the semiconductor diode (or com- munications) laser from IBM and GE, and many others enabled the revolution in consumer electronics, information technology, and digital communications that is still sweeping the world today. Ironically, the global changes sparked by these and many other inventions have, over decades, weakened U.S. industrial research in the physical sciences. In particular, information technology has been a key enabler of globalization and its resulting intensified economic competition. Information technology has upset natural monopolies in communications and computers and is sweeping en- tire job categories, products, and industries into the dustbin of history, even as it creates new ones. Newer industrial research laboratories established by Microsoft, Google, IBM, and others focus principally on software, systems, and services. This has also been the growth direction for some of the longer-established industrial laboratories. And some of these new software-focused laboratories have been set up in other countries to attract local talent and to improve understanding of and participation in rapidly growing local markets. In sum, these changes have led to the downsizing or elimination of some once-great industrial laboratories and have greatly reduced the focus on physical sciences research in others. In addition to generating countless inventions that have driven the U.S. econ- omy, this core of industrial laboratories has also provided large numbers of sci- entific and technological leaders to industry, academia, national laboratories, and the government. This training ground for future leaders in science, education, and policy is also diminished by the changes in the industrial laboratories. After
I n d u s t r i a l L a b o r ato r i e s and Research in CMM P 167 a century of scientific and technological leadership, the consequences to the U.S. economy and national pride of going from best-in-class to technology followers would be devastating. So, what research organizations and institutions will drive innovation in the physical sciences for the next hundred years? In the decades following World War II, the federal government was the major provider of U.S. funds for research and development (R&D). Since 1980, the indus- trial investment has exceeded the government investment and today represents over two-thirds of the national effort. However, the bulk of the industrial investment in R&D is in incremental improvements to existing products, while longer-range research has declined; today it almost certainly represents less than 10 percent of the industrial investment (see, for instance, the discussion in Chapter 10 of Figure 10.15). The federal government is now the larger investor in fundamental long- range research, and this is especially true in condensed-matter and materials phys- ics (CMMP). Other investment in U.S. R&D comes from foundations, states, and private individuals. This funding is increasing, and today some of the longer-range work in CMMP comes from state sources. The states of New York, California, and Texas have been leaders in this area. The sharp rise of venture capital in the 1990s is also changing the face of R&D investments. Some of todayâs venture-funded start-up companies are pursuing research models that they hope will be more efficient than the older model of the centralized industrial research laboratory. Through the licensing of intellectual property from universities at low rates and the hiring of the graduate students who generated that intellectual property, these companies strive to translate academic research rapidly into product innovations that they can license to large companies. Note, however, that these companies do not themselves, as a rule, pursue basic or long-term research. The BayhâDole Act (or University and Small Business Patent Procedures Act) of 1980 (Public Law No. 96-517) has also contributed to changes in R&D invest- ment practices. It reversed the presumption of title so that universities, small businesses, or nonprofit institutions pursuing government-funded research can elect to pursue ownership of a resulting invention before the government. The act thus encouraged universities and small companies to pursue the licensing and development of such inventions. However, some observers maintain that the act has encouraged some universities to be very protective of all potential intellectual property development and that this practice has hindered university-industry research partnerships. Filling the Gap: New Approaches to Long-Term Research With reduced participation by industrial laboratories, breakthroughs resulting from long-term research must increasingly come from universities, national labo- ratories, and/or industry-led consortia. For example, the National Nanotechnology
168 C o n d e n s e d - M at t e r and M at e r i a l s P h ys i c s Initiative (NNI) led to the creation of the Department of Energy (DOE) Nanoscale Science Research Centers (NSRCs), which represent an organizational innova- tionâan effort to include university students, faculty, and industrial researchers in government-funded, interdisciplinary centers focused on nanoscience. The NSRCs thus differ in some significant ways from other DOE beamline-based facilities. The five NSRCs collectively are similar in cost to a single large DOE synchrotron, in terms of cost for construction (about $380 million total) and operation (about $100 million per year). As for the large DOE facilities, users from any institution (in the United States or elsewhere) may submit a proposal that is peer-reviewed for science to be done at the facility. For the DOE NSRCs to re-energize long-term basic research in CMMP, they need, in addition to âuserâ facilities, a large in-house base of world-class scientific talent. This staff should be challenged to foster great new ideas. It is not yet clear that the new research institutions are focusing sufficiently on nurturing a creative intellectual environment. While these centers represent a significant investment in an important emerging scientific research area, to be successful they must also attract and serve the needs of many industrial users, including start-ups and small companies. Can the ownership of intellectual property from the research provide protection to an organization developing commercial products? Is there a path- way for product development from the research, and does the government policy on indemnification restrict participation in the centers? There will undoubtedly be other questions that arise as work is performed at the NSRCs. If successful in producing research results and structured for other organizations to bring those research results quickly to market, these centers can serve as a model for future government-university-industry cooperation. However, due to the newness of these centers, the research community awaits further experience with their operation before drawing conclusions from this experience. A second possibility for replacing some of the CMMP physics research done by the U.S. industrial laboratories is to enhance the research sponsored by industry in research universities. Today, U.S. industry funds less than 10 percent of univer- sity research, although the funding varies greatly by institution, with some of the b Â etter-known universities having greater than 30 percent of their research funded by industry. The focus of much of this research is in engineering and business schools and little is in physics or materials research. An example of industry-university cooperative research is the Semiconductor âDetails on the construction and operation budgets of the DOE NSRCs can be found at http:// www.sc.doe.gov/bes/archives/budget/BES_FY2008budget.pdf; last accessed September 17, 2007. âNational Science Foundation, Division of Science Resources Statistics, Where Has the Money Gone? Declining Industrial Support of Academic R&D, NSF 06-328, Arlington, Va., 2006. Available at http://www.nsf.gov/statistics/infbrief/nsf06328; last accessed September 17, 2007.
I n d u s t r i a l L a b o r ato r i e s and Research in CMM P 169 Research Corporation (SRC) founded by the U.S. Semiconductor Industry Associa- tion (SIA) in 1982. The SRC cultivated a pivotal cultural change for the semicon- ductor business, sharply raising the funding of relevant university research and the transfer of university research results into the semiconductor industry. Talented faculty members have received incentives to tackle outstanding semiconductor industry research problems and to train students who then seek employment with semiconductor companies. For about 25 years, much of the SRC funding has gone to engineering departments and has addressed incremental improvements to established devices and manufacturing process technologies. However, given increasing industry concern that transistor technology is approaching fundamental physical limits, some SRC programs are beginning to focus on long-term research in CMMP. One of these is the Focus Center Research Program (FCRP); another is the Nanoelectronics Research Initiative (NRI). The SRC-FCRP consists of five centers, each composed of research groups from several geographically dispersed universities with funding from a subset of the SRC member companies and the Defense Advanced Research Projects Agency (DARPA). The FCRP research portfolio is longer term and higher risk than older SRC pro- grams. A portion of this portfolio is aimed at exploratory materials and devices. The SRC-NRI consists of three major research centers with lead schools located in New York, Texas, and California. Funding comes from six of the leading U.S. semiconductor manufacturers and various state governments. NRI has also teamed with the National Science Foundation (NSF) to provide additional joint funding of existing NSF-funded research centers. NRI research focuses on new materials and device concepts that might someday replace the field-effect transistor as the foun- dation of information technology. The new materials might not be semiconductors. The new devices may operate by as-yet-undiscovered physical principles. This is a remarkable development, motivating top condensed-matter physicists to explore exotic physical systems with the goal of revolutionizing an entire great industry. For more on the motivation for this research, see Chapter 7. In the SRC programs, there is broad and intimate contact between the engi- neers and scientists in the SRC member companies and the university students and faculty. The research directions are set by the university faculty with input from the SRC members. This process has been nurtured and modified over nearly 25 years and is considered a major success by universities and by the global semiconductor industry. It could serve as a model for enhancing condensed-matter and materials research in physics, chemistry, and materials science departments. Finally, it might be possible to form research consortia that would involve interÂ national companies that would profit by funding university research and having increased interaction with students and faculty. There are consortia today in many economic regions around the world that have involved cooperative R&D. In the United States, SEMATECH was initially an effort by the semiconductor industry to
170 C o n d e n s e d - M at t e r and M at e r i a l s P h ys i c s solve short-term research needs. As part of this program, there was a $10 million program to fund SEMATECH Centers of Excellence at several U.S. universities. In 1995, SEMATECH became an international organization and withdrew from U.S. federal funding, permitting funding of research internationally. In Europe, the Interuniversity Microelectronics Consortium has both short- and long-term research in several universities. In Japan, there has been a history of cooperative efforts in semiconductors and computers. The pharmaceutical industry has a his- tory of funding university research. Again, while much of this research is focused on short-term results, some of the consortia are investing in longer-range work that will be of interest to CMMP scientists and offer examples of possible future funding possibilities. Conclusions The United States has lost a major source of innovation in CMMP with the changes that occurred in the great industrial laboratories of the 20th century. Replacement of this source of invention and leadership will be a major challenge. As described above, new research models are being tried. These include coopera- tive university-industry agreements, new interdisciplinary centers at the national laboratories that welcome university and industry researchers, and a variety of university-government-industry research consortia. There are undoubtedly other possibilities as well. The current approaches described here are focused on electron- ics and information technology, but other CMMP research areas that serve society can also benefit and advance from these approaches, such as energy research. These new approaches to long-term research are expected to evolve over the next decade, so the next decade will be one of exploration (see also related discussions at the end of Chapter 7). It would be highly desirable for the physics community and the federal govern- ment to establish mechanisms to measure and compare the effectiveness of these evolving models for conducting scientific researchâparticularly those that are funded primarily by the government. Can we as a nation make the organizational and funding mechanisms work well for this new purposeâfor creating scientific and technological breakthroughs and providing future scientific leadership? The national laboratories were not originally formulated to capture commercial eco- nomic dominance, and universities historically have existed to advance knowl- edge, not to develop new products. The semiconductor industryâs effort in SRC and particularly in the new NRI program is one model to look to for lessons for introducing change. The DOE Nanoscale Science Research Centers are another example of this new mode of research. An evaluation of efforts such as these could identify the positive lessons to apply in the future about how to provide funding to stimulate vigorous innovation in CMMP research from which the United States
I n d u s t r i a l L a b o r ato r i e s and Research in CMM P 171 reaps major economic rewards. Further, the DOE should evaluate the new NSRCs by metrics that include success in attracting a diverse set of industrial users and by other metrics as highlighted in the previous section. The National Nanotechnology Coordination Office, in its arrangement of the triennial review of the NNI, should evaluate all NNI-funded centers and networks of centers by similar metrics. For further discussion, see Chapter 11. Recommendation Recommendation:â The Office of Science and Technology Policy (OSTP) should convene a study with participation from the Department of Energy, the Department of Defense, the National Science Foundation, and the National Institute of Standards and Technology, the physics community, and U.S. cor- porations to evaluate the performance of research and development (R&D) activities that might replace the basic science previously done by the large in- dustrial laboratories and the contributions that those laboratories made to the training of future scientific leaders and educators. This next decade will involve a series of new approaches to long-term R&D designed to recapture the ability to work on large difficult projects based on fundamental CMMP research. Such an evaluation should be an ongoing activity of OSTP, since it may be several years before the performance of these activities can be adequately evaluated.