Perspectives on Facilities and Instrumentation for Materials Research

Introduction

DEAN E.EASTMAN

In the past few decades, materials research in the United States has emerged as a large national effort vital to our technological and economic welfare. Materials research is interdisciplinary and is carried out through important programs in the university, government, and industrial sectors. Facilities and instrumentation, an essential element of these research programs, are becoming more sophisticated and costly. This chapter presents several perspectives on that element of materials research programs.

Large-Scale Facilities for Materials Research

MARTIN BLUME

Many of the large facilities and the large-scale aspects of materials research originated at Department of Energy (DOE) national laboratories many years ago. The quintessential large facilities are, of course, the high-energy physics facilities. In materials research and in other areas with a strong tradition of small science, these large-scale laboratories evolved gradually; in fact, the first were not built as materials research facilities. They were supported with funds designated for neutron scattering research, for example, from reactor programs.



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 335
Advancing Materials Research Perspectives on Facilities and Instrumentation for Materials Research Introduction DEAN E.EASTMAN In the past few decades, materials research in the United States has emerged as a large national effort vital to our technological and economic welfare. Materials research is interdisciplinary and is carried out through important programs in the university, government, and industrial sectors. Facilities and instrumentation, an essential element of these research programs, are becoming more sophisticated and costly. This chapter presents several perspectives on that element of materials research programs. Large-Scale Facilities for Materials Research MARTIN BLUME Many of the large facilities and the large-scale aspects of materials research originated at Department of Energy (DOE) national laboratories many years ago. The quintessential large facilities are, of course, the high-energy physics facilities. In materials research and in other areas with a strong tradition of small science, these large-scale laboratories evolved gradually; in fact, the first were not built as materials research facilities. They were supported with funds designated for neutron scattering research, for example, from reactor programs.

OCR for page 335
Advancing Materials Research As a result, there was no problem with funding arrangements until a decade ago, when such facilities started to turn up in materials research budgets. The national laboratories, of course, had their own problems and research programs connected with atomic energy in the days of the Atomic Energy Commission. Thus, DOE not only had these internal programs, but became an agency that also provided large research facilities to universities and, more recently, to the industrial community as well. The synchrotron light source at the Brookhaven National Laboratory is an example of the large facilities available for materials research. The research carried out at these facilities, as opposed to the high-energy physics facilities, remains basically in the small-science mode and in effect provides research opportunities similar to those in the small laboratories. For neutron scattering, a fair number of research facilities are available: the intense pulsed neutron source at Argonne, the pulsed source at Los Alamos, and the reactors at Brookhaven and Oak Ridge. In synchrotron radiation the DOE-supported facilities are at Stanford and Brookhaven, with National Science Foundation (NSF)-supported facilities at the University of Wisconsin, Cornell University, and elsewhere. In addition, an electron microscope facility is available at the Lawrence Berkeley Laboratory, a high-magnetic-field facility is available at the Massachusetts Institute of Technology, and there are others. All of these large research facilities are open to users, and pressures for their use have grown in the last decade. These pressures have had to be responded to by the agencies that fund research in materials science, as opposed to other areas. In the past, materials scientists were accustomed to working parasitically on either a high-energy physics facility or a reactor facility. The pressures for increased use of synchrotron radiation sources arise from the relatively simple fact that for many generations, x-ray tubes provided more or less the same intensity. With the advent of synchrotron radiation sources, however, came an exponential increase in the intensity of electromagnetic radiation available for research. Brookhaven has two synchrotron radiation storage rings—an ultraviolet ring that runs at 750 million electron volts (MeV) and provides radiation up to the soft x-ray part of the spectrum, and a high-brightness x-ray ring that runs at 2.5 billion electron volts (GeV) and provides the harder part of the radiation. There are 16 ports for radiation on the ultraviolet ring, each of which is capable of providing up to four experimental beam lines. Similarly, there are 28 ports with perhaps three experimental beam lines possible on each of those ports. Thus, it is possible to carry out many experiments simultaneously. This provides important advantages, social as well as scientific, but at the same time produces tremendous problems. The operation of a facility like this differs considerably from that of a high-energy physics facility (where there is only one primary user of the beam) in

OCR for page 335
Advancing Materials Research that two or three experiments may be going on at one time. How is such a facility organized? How are all of those beam lines built? One way is for the laboratory itself to provide all of the experimental beam lines and then take proposals from each of the users. A difficulty with this approach is that it engenders a large bureaucracy and is counter to the way in which materials science researchers as well as biologists, chemists, and others who use the facility are accustomed to working. The bureaucracy also tends to eliminate spontaneity in the conduct of research. (This is one of the major advantages of having an x-ray source in your basement laboratory. You can go down there without having to ask a committee to use it at a particular time; you can make mistakes and try new things.) The management of concurrent research at Brookhaven is of interest because it involves a different organizational method—having users build and operate the beam lines. The compromise adopted at Brookhaven is to ask for the organization of participating research teams. These are groups that propose to place instruments at the facility. If a team’s proposal is accepted, the instruments are installed and the Department of Energy provides the photons for research. In return for those photons, the research team makes this instrumentation available one quarter of the time to small users who just want to come in and do a single experiment. This mode of operation has worked very well. The participating research teams are left to themselves to organize and to carry out their own experiments. A further advantage is that industry is investing in this instrumentation—something that is strongly encouraged. Thus, a system that amounts to time-sharing has succeeded in attracting a fair amount of money and instrumentation expertise. Many institutions, including governmental laboratories, corporations, and universities, have taken part in this system through the participating research teams. All of them are involved in beam lines at various places. Many of these are beam lines that have been installed by Materials Research Laboratories (MRLs) and are used as parts of the MRLs. Many of the MRLs located near to one another, including those at the University of Pennsylvania, Cornell University, Massachusetts Institute of Technology, and Harvard University, have been actively involved in this way. Some corporations participating in the Brookhaven system are not known for basic research. Indeed, assistance had to be provided to researchers at some of these corporate research laboratories to enable them to make even a relatively small investment in this equipment outside their own institutions. Thus, some of the corporate research centers have been opened up to basic research. There has also been a good deal of “marriage brokering” to bring together joint university and corporation programs. Despite the large number of participants in research at Brookhaven, the facility still functions like a small-science facility. It is as if all of the experiments that

OCR for page 335
Advancing Materials Research required electric power had to be done right at the power plant. Thus, from a research activity viewpoint, facilities like Brookhaven should be viewed not as extremely large single units, but as impressive concatenations of many different facilities and many different types of science. At Brookhaven, for instance, chemists and biologists sit together as members of participating research teams at that early stage. It is important, however, not to overlook the large core cost associated with such a large facility. Despite the size and complexity of the facility, the operating cost for individual experiments is relatively low. The cost of a shift on one of the beam lines is $80 an hour just for the photons. Although overall operating costs of $14 million per year are not particularly low, the number of beam lines in use is relatively high. In addition to the participating research teams, many small groups use the facility. For instance, it is not uncommon to see a single professor and a graduate student using one of the beam lines. These small groups can come in at a relatively low initial cost and do this kind of research. Brookhaven has the possibility of providing for travel grants, although this presents one important difficulty—such grants are very useful for small groups, but they can distort the research agenda. They create the possibility that a small group with a good idea but unable to get a research grant will push its efforts in directions dictated by the availability of these facilities. This important question needs careful attention. This is one reason why it is important to avoid what might be described as “giving away lollipops” with each of the experiments that is funded. It is important not to make research at large facilities (such as Brookhaven) so desirable that people will distort their research in this direction. Balance must be maintained overall in the research program. It is unfortunate that the funds that are necessary to operate these large facilities often are not fully realized. Consequently, there often is strong pressure to cut back on internal small-science programs at the host laboratory and to use that money for the operation of the large facility. As a consequence of this, at Brookhaven virtually all of the internal research is now based on large facilities. National Commitment to Facilities and Instrumentation for Materials Research C.PETER FLYNN Most university and national laboratory materials research is supported by the National Science Foundation (NSF) through its Division of Materials Research (DMR) and by the Department of Energy (DOE) through the Di-

OCR for page 335
Advancing Materials Research vision of Materials Sciences (DMS) in Basic Energy Sciences. These two agencies have “grown up” as the field of materials science has come into being over the past two decades. Together they are responsible for about $300 million of yearly materials science funding. This approaches half the annual total in materials research funding for the nation, including that provided from the Department of Defense, industry, the National Bureau of Standards, and so on. The relevant point for present purposes is that both DMR and DMS commit roughly 25 to 30 percent of their yearly resources to the support of various types of facilities. The details differ in the two cases. Most DOE facility support passes into major Centers for Collaborative Research in such areas as neutron scattering, synchrotron radiation, and electron microscopy, which are established at institutions (both university and governmental) in the DOE Laboratories Program. NSF also supports major centers for synchrotron radiation, microscopy, and so on. Through its Materials Research Laboratories (MRLs), Materials Research Groups (MRGs), and Instrumentation programs, it also funds smaller-scale facilities on a number of university campuses. While the details differ, a massive commitment to the support of facilities is evident in both agencies. Still further facilities for materials research are operated by other organizations, including the National Bureau of Standards and the weapons laboratories at Livermore, Los Alamos, and Sandia. It is a contemporary phenomenon that such a large portion of research funds is directed to facilities. At the time the MRLs were founded in the early 1960s, there were far fewer facilities, of which neutron sources operated by the Atomic Energy Commission constituted the major part. Without question, the current prominence of facilities funding is in direct recognition of the important role that research facilities play in modern materials science and of the unique research avenues that they open to the enterprising researcher. Such growth in difficult times has naturally caused tension in funding decisions at both NSF and DOE. A further growth of facilities expenditures by a factor of two to 50 to 60 percent of the total appears unlikely, at least without major new resources, because facilities only contribute to a portion of the entire materials field. To help judge whether the present balance is appropriate, one must be familiar with the level of marginal declinations of research proposals in non-facility-related areas and with the level of marginal research supported by facilities-related programs. The decisions are complex and involve many considerations. These include the fact that facilities are justified in part by the finest work to which they give rise, the long time scales required to establish facilities, the cumulative distortion of the research field and the funding patterns they produce downstream, and many others. These are complex issues on which opinions differ.

OCR for page 335
Advancing Materials Research Despite the current large investment in materials research, the United States lacks desirable research facilities in a number of areas. At the same time, the marginal rejections of research proposals at both NSF and DOE are alarmingly high in the materials sciences, and the ability to fund new proposals from the brightest young scientists entering the field is dangerously low. The competition between these factors presents a critical dilemma in the disposition of available resources. In the following brief commentary on the roles that research facilities play, the different types of facilities are referred to as infrastructure, research facilities, and collaborative research centers. The term infrastructure refers to durable, shareable equipment established in a given research environment for use by several or many researchers to whose work the equipment is, to some degree, beneficial. Examples of such environments might be a campus or department. Equipment typically costs between $100,000 and $300,000. It might consist of a VAX computer, mechanical testing equipment, fairly simple x-ray systems, or a robust scanning electron microscope. Such equipment can be kept up and used to mutual benefit by a number of scientists whose main research directions differ, provided that means for maintenance and occasional expert consultation are available. To be well used, infrastructure equipment must nevertheless exist inside an organizational framework. If there is an MRL or similar organization on campus, these matters are easily handled. The MRGs—surely a much-needed funding initiative—can bring a leadership structure to many other campuses. Organization is required for maintenance and replacement of infrastructure equipment. A maintenance contract on a computer costs perhaps 10 percent of its purchase price per year, and on an electron microscope perhaps 3 percent. These and other operating costs must generally be defrayed by a system of usage charges. In general, few research universities lack instrumentation of this type, although what exists may not be optimal. The term research facilities refers to instrumentation that is more specialized, more fragile, and much more expensive than infrastructure equipment. Often these are commercial systems that perform the primary research itself. Examples are high-resolution transmission electron microscopes, surface science systems, machines for advanced materials synthesis, as in molecular beam epitaxy or microfabrication, and complexes of laser equipment. One machine may cost a million dollars. The facility may consist of a single instrument or several. It may be operated by an organization, such as an MRL, or it may be separately funded. Examples of larger complexes are the electron microscopy facilities operated by DOE at Argonne, the University of California at Berkeley, the University of Illinois, and Oak Ridge, and by NSF at Arizona State University, and the NSF surface science facility at Montana State University. These are generally identified with user programs that draw investigators from an extended geographical region.

OCR for page 335
Advancing Materials Research Research facilities face a number of organizational difficulties. Local expertise at an advanced research level is generally needed to justify the expense. Costs for maintenance, operating, and technical assistance may be considerable. Again, the need to have experts maintain fragile equipment for nonexpert users raises obvious problems. Yet, these questions must be faced. In electron microscopy, for example, the United States still is not self-sustaining in the training of research talent, despite the major role these instruments have played in revealing the structure of solids on the scales of 1 micron to a few angstroms. Social factors enter into the operation of a research facility and can influence its effectiveness. Because an expert’s involvement is essential, the instruments tend to become captive rather than appropriately accessible. To maintain such equipment at the state of the art can become a funding burden that inhibits other new initiatives. The peer review system has not easily adapted to decisions about organizations with the complexity of MRLs or surface science facilities. The task of handling research facility funding in the best interest of the nation is both delicate and vital. The third category of facilities is collaborative research centers. These facilities include neutron sources for spectroscopy and synchrotron radiation sources (one or two electron microscope centers with uniquely engineered instruments could possibly be included). Research centers involve large-scale, complex engineering and have price tags of at least $50 million for synchrotron radiation and an order of magnitude more for neutrons. When instrumented, the facilities accommodate 10 to 100 independent projects simultaneously, often operating around the clock. Collaborative research centers provide the nation with research opportunities that would otherwise be inaccessible. Neutron scattering, for example, has revealed much that is known about phonons in crystals and about magnetic structure. Synchrotron radiation is heir to both x-ray and ultraviolet spectroscopies and has played a key role in the contemporary development of surface science. Existing U.S. neutron reactors at Brookhaven, the National Bureau of Standards, and Oak Ridge are powerful and well used but aging; new facilities are needed. Institut Laue-Langevin in Europe has become a center of activity. The past decade has seen new synchrotron radiation centers built at Brookhaven, Stanford, and Madison to join existing sources. None is yet fully developed. At least two more are planned for special production of hard x rays and high-intensity ultraviolet. Although recent U.S. investment in these areas is more evident than in neutron reactors, these developments only keep us abreast of comparable advances abroad. Materials science has emerged as a field only over the past two and a half decades—the same period over which the MRLs have existed. A significant part of this self-identification in the United States has occurred in concert with the Division of Materials Research at NSF and the Division of Materials Sciences at DOE. These agencies and the field now face a critical problem:

OCR for page 335
Advancing Materials Research How can we channel more funds into new research facilities when the funding criteria in other areas of the field are already unrealistically high? Either choice will damage existing programs and cause major research opportunities to be lost. There are two points to be made. First, the field of materials science is not yet organized so that decisions of this type can be made in the context of the overall national program. Second, the field is not well organized to present its needs appropriately in the national arena. One major deficiency of the field is the lack of a forum for national consensus. This is not a surprising problem for a field that has drawn itself together from the diverse disciplines of metallurgy, ceramics and polymers, solid-state physics, and chemistry. The National Academy of Engineering and National Academy of Sciences have sponsored symposia on materials science topics and organized bodies such as the Solid State Sciences Committee of the National Research Council. These efforts contribute to the broad exchange of information at a level at least comparable with that of the professional societies in the several areas of materials science. It seems clear, however, that a further ingredient is needed to ensure that representational factors in this diverse field are correctly balanced in the consensus. The funding agencies have charted these difficult waters for a decade or more and have operated representative committees. Their experience is now needed in pulling together an appropriate forum in which national issues in materials science can be discussed and collective decisions can be made in the best interests of the field as a whole. A representative body of this type would not, of course, eliminate the difficulties mentioned above. The debate over major facility developments would still have charismatic leaders urging decisions that are to their own benefit, and laboratories would still seek to have their own machine concepts funded. Small science would still feel threatened by the encroachment of large machines onto the funding base. The advantage lies in having the debate focused in an arena of continuing, rational discussion. Recommendations could be fitted into a logical pattern in which commitments and priorities evolve hand in hand. It would be possible to consider the way infrastructure, research facility, and collaborative research center funds balance with each other and with science issues unrelated to facilities; whether facilities are in fact paid for substantially with “extra” funds that would not otherwise be available to the field; and whether DOD, industry, and others should contribute more to facility costs to ease the burden on the NSF and DOE materials sciences programs. The best interests of the field are not served by having different bodies recommending solutions to each problem separately. Materials science could reap a final major benefit from organizing a representational body. By doing so it would identify its own voice in the public debate over funding priorities. Authoritative statements could be made about

OCR for page 335
Advancing Materials Research the needs of materials science and about the consequences of their neglect. After all, materials science plays as critical a role in national defense and in improving the quality of life as it does in the nation’s industrial well-being and its intellectual progress. The problems of the field are not so much in the division of funds between science and facilities as in the fact that $600 million annually is much too small a national investment in this ubiquitous and still youthful branch of science. Materials scientists need to organize so that this viewpoint becomes recognized and accepted in the national debate. Instrumentation for Materials Research J.DAVID LITSTER Questions of instrumentation for materials research are addressed in the recent report Financing and Managing University Research Equipment, a study carried out under the supervision of the Association of American Universities, the National Association of State Universities and Land-Grant Colleges, and the Council on Governmental Relations. With support from six government funding agencies and the Research Corporation, a three-member field research team, of which I was a member, visited 23 universities, government laboratories, and industrial research laboratories and spoke with approximately 500 people. Recognizing that the existence of a problem in research instrumentation in universities had been well documented by previous studies, we asked the following questions: What changes in federal and state regulations and policies would help solve the problem? What changes should universities make? What changes in tax and other laws might help? What can be accomplished by alternative or creative methods of financing? Changes can be made in all of these areas to improve the efficiency of university acquisition and management of research equipment. The problem is so large, however, that its solution requires substantial and sustained investment from all available sources. Let me begin by reviewing the nature of the problem, drawing heavily upon work carried out by the National Science Foundation in its survey of academic research instrumentation in 1982 and 1983. More than 70 percent of the departments surveyed reported that lack of equipment prevented crucial experiments. About 20 percent of the equipment in their inventory was obsolete. Of the equipment in use, about 22 percent was more than 10 years old; only about 50 percent of the equipment in use is in excellent condition. The report stresses that maintenance and operation of equipment is as serious a problem as getting the money for its initial

OCR for page 335
Advancing Materials Research Federal expenditures for R&D in universities and colleges from 1965 to 1983. The solid line shows expenditures in current dollars, and the dashed line is corrected to current dollars using the Consumer Price Index relative to 1972. The dotted line shows expenditures in current dollars corrected for inflation in the cost of scientific equipment. purchase. With respect to infrastructure, about 50 percent of the departments reported inadequate or nonexistent support facilities. A further, important ingredient in the problem is the high start-up cost for new projects and for new faculty members. There has been a 78 percent decline in bricks-and-mortar expenditures in real dollars since 1968. This decline also affects instrumentation, since new facilities generally come equipped with instrumentation. Finally, there is the increased sophistication and cost of research equipment in all fields, not just in materials science. Data in the figure (see above) from the report give a quantitative picture of the research equipment problem. The figure shows the total federal R&D spending in colleges and universities from 1965 to 1983 in current dollars and corrected to constant dollars using the Consumer Price Index. The best data I could find on the proper rate of inflation for costs of research equipment come from an unpublished study by Robert Melcher, a scientist and manager at IBM’s Thomas J.Watson Research Center, Yorktown Heights, New York. Melcher examined the costs of the type of research equipment purchased by

OCR for page 335
Advancing Materials Research IBM between 1976 and 1981 and found a rate of inflation 1.7 times that of the Consumer Price Index. I have applied Melcher’s correction for inflation and show the results as the dotted line in the figure. This gives an overly pessimistic view of the overall support of research, since most research costs probably increase at a rate closer to the Consumer Price Index. However, it underestimates the seriousness of the problem for research instrumentation, because over most of the period represented in the figure the federal agencies and the universities were reducing the fraction of research dollars that were spent on equipment. What can be done about this problem? It is important to keep in mind where the resources come from—well over 50 percent of funding for research equipment in materials science is provided by federal agencies; industrial support, which has never been large, accounts for 3 to 5 percent; the universities themselves have been the second major funder of research equipment and have paid for approximately 30 percent of the cost of equipment in use. What can these various parties do to ease the problem? Federal agencies, for instance, could interpret their regulations, rules, and policies in a consistent way. The present situation tends to make universities unnecessarily conservative in their management practices. It is sometimes difficult to spread the costs of major equipment across Fiscal Year boundaries and certainly across grant boundaries, but frequently this would help. Numerous administrative barriers increase the viscosity of the systems: for example, excessive inventory requirements and the Defense Industry Plant Equipment Center screening for DOD contracts. In many cases, realistic depreciation allowances for equipment would help, providing that the funds so generated were put toward the purchase of new equipment. This is not a cure-all, of course, because universities can depreciate only the share of equipment that they paid for themselves. The policies of state agencies raise similar problems because state regulations are frequently more troublesome than those of the federal government. State agencies can help by improving or removing burdensome regulations. In addition, they can help with tax-exempt financing, although it is not clear whether this will be possible if the current proposed federal legislation goes through. In fact, many universities are now seeking to float tax-exempt bonds just to put money in the bank so they will have it in a year or two. Finally, the states could set up agencies to promote science and industry, as North Carolina has already done. What can the universities do? First, they should recognize that university research differs from that in industry or government laboratories. University research tends to be much more decentralized than it is in industry or government, and significant funding originates from individual principal investigators within the university. However, it is important that, if universities use creative forms of debt financing to acquire equipment, they must not go

OCR for page 335
Advancing Materials Research into debt in a decentralized fashion. Therefore, it seems likely that resource allocation and planning will become more centralized in universities. Of course, this has its undesirable side effects, and universities will have to make some hard decisions. Universities will have to cut back on some programs to provide the increased support necessary to maintain the health of others. Each university must investigate its individual potential for university-supported maintenance and repair facilities and perhaps limited inventories of research equipment that could be shared. Iowa State University, for example, has an excellent equipment-sharing program called REAP, elements of which could perhaps be adopted by other universities. In our survey, the field research team investigated carefully the issue of sharing research equipment: Is there enough sharing going on? Should there be more? Are instruments sitting unused? A considerable amount of sharing is already going on in universities, much of which is made possible by the Materials Research Laboratories. We did find, however, that not in all cases did the universities properly prepare for the realistic costs of operation and maintenance when they were buying research equipment. The universities should try harder to recover realistic depreciation costs. These will, of course, either increase the indirect cost base or increase the direct costs of doing research. Nevertheless, these are real costs that must be met in some way. We found a further need to work with funding agencies to find an incentive for investigators to transfer equipment to other investigators who might make good use of it, perhaps in other universities. There is little incentive to do that now. Our overall conclusion was that in the last 10 or 15 years, universities have supported research by supporting people, not instrumentation. Funding by the National Institutes of Health for permanent equipment declined from about 12 percent in 1966 to about 3 percent in 1985, which is clearly too low. Similarly, NSF support for equipment went through a minimum in the period between 1969 and 1976 and has since come back up as the agency recognized the problem. In summary, an effective and balanced national research program requires that a larger percentage—probably greater than 20 percent—of our resources be devoted to instrumentation, and this must be done on a sustained basis. It will probably be necessary also to increase the size of grants in order to provide this support and to meet the increased costs of operating and maintaining this more sophisticated equipment. If there is no increase in total funding, it may be necessary to reduce the number of grants and the number of investigators supported.