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 61
Midsize Facilities: The Infrastructure for Materials Research 3 Challenges for Midsize Facilities The capabilities of new instruments are advancing at a remarkable rate. Such instruments enable and are essential to significant advances in materials research and development. But two competing trends are apparent when considering facilities that support the materials research enterprise: (1) Capital costs as well as the costs of maintenance and support are escalating to such an extent that individual institutions are experiencing serious difficulty in providing equipment to their user base and also maintaining it at a state-of-the-art level. (2) Individual sophisticated machines are becoming easier to operate, opening them up to a wider range of users and providing useful data more quickly and efficiently; in other words, automation is more common (for example, in x-ray diffractometers or scanning electron microscopes), but interpretation of the data acquired can be more difficult. As facilities become increasingly sophisticated in structure and content, they require a complex network of support to maximize their effectiveness. As best put by the National Science Board in its recent report Science and Engineering Infrastructure for the 21st Century, “The current 22 percent of the NSF [National Science Foundation] budget devoted to infrastructure is too low to provide adequate small-and medium-scale infrastructure.”1 1 National Science Board, Science and Engineering Infrastructure for the 21st Century: The Role of the National Science Foundation, NSB 02-190, Arlington, Va.: National Science Foundation, 2003, p. 2.
OCR for page 62
Midsize Facilities: The Infrastructure for Materials Research LONG-TERM VIABILITY Midsize facilities provide a stable and user-friendly interface between much of today’s most advanced instrumentation and the research community. A core challenge for midsize facilities is that of maintaining their infrastructure for the long term in order to fully exploit the initial capital investments (often in the tens of millions of dollars in aggregate). Midsize facilities operate best with long-term commitments. Because these facilities necessarily outlast (and transcend) the single investigator’s research project, their sustenance must come from sources that are more stable and long term than is an amalgamation of individual users with an overlapping need. This challenge is especially acute for smaller schools that often do not have the administration and management overhead or experience to develop a sustainable plan for the long-term operation of a midsize facility. Diverse and Stable Funding Funding sources for existing successful facilities are highly diverse and often depend on the local environment and the resourcefulness of the persons involved. Funding may come from a combination of federal and state government support, institutional funds (whether state or private), user fees, and donations. Funding is one of the most significant challenges facing midsize facilities for the following reasons: The escalating costs of instrumentation make it more and more difficult to identify sources of funding for the initial capital investment. The changing landscape of cost-sharing requirements for federal grants has placed many institutions at a disadvantage for securing the initial capital funds. In addition to the initial capital investment, a plan for stable long-term funding is necessary to cover the plethora of recurring expenses involved in sustaining facility operations. The escalating cost of obtaining, operating, and maintaining advanced instrumentation is one of the greatest challenges facing materials research. Instrumentation typically evolves in several directions: it simplifies certain experimental procedures, it automates and standardizes certain techniques, or it provides entirely new functionality and capability through innovation. These forces tend to escalate the cost of instrumentation because of the substantial enhancements in capabilities provided. A common observation is that the cost of flagship tools has increased over time well beyond standard inflationary rates. See Box 3.1, “Escalating Costs of
OCR for page 63
Midsize Facilities: The Infrastructure for Materials Research Instrumentation,” for several examples. Critical equipment is routinely so expensive that a sophisticated organization for promoting the utility of and access to the equipment is required in order to make it worth the capital investment. As capital costs rise, so do the long-term maintenance and operations costs. In terms of the initial accumulation of funds, a major issue identified is that of cost sharing for the acquisition of expensive equipment. Cost sharing is usually a standard requirement for funding by federal agencies, but the proportion is highly variable and has recently been removed entirely! For example, the NSF cost-sharing requirement increased from 0 to 30 percent over the past few years and has just recently been reduced again to 0.2 For a $1.5 million transmission electron microscope (TEM), this cost-sharing requirement amounted to as much as $450,000. Meeting such an obligation would be a major obstacle for smaller institutions. Although the committee recognizes that an important objective of cost sharing was to encourage the host institution to be more responsible and committed to the facility, it observes that these financial requirements had become a serious barrier for some institutions wishing to participate. In its recent announcement, the National Science Board approved the removal of all cost-sharing requirements for major awards and grants.3 On the one hand, this change will help level the playing field by allowing institutions with fewer resources to compete more equitably with larger schools, but on the other hand, this change effectively decreases the buying power of NSF’s grants by up to 30 percent. Investigators no longer have the leverage to entice host institutions to put in their own money, which would often have entrained other forms of institutional (and moral) support. As a result, fewer resources will likely be directed to facilities’ infrastructure. Alternatively, institutions with access to additional resources may voluntarily choose to continue to use these resources in order to maintain the institution’s preeminence and leadership position. Additionally, suitable incentives to encourage institutions to team up as partners may go a long way in overcoming limited funds and the consequent loss or denial of opportunities to participate in groundbreaking materials research. A conservative estimate of (recurring) support costs for a $2.0 million instrument such as a focused-ion beam or a secondary ion mass spectrometer (SIMS) would have three components: 2 National Science Board, Memorandum to Members and Consultants of the National Science Board, NSB 04-157, Arlington, Va.: National Science Foundation, 2004, p. 2. 3 National Science Board, Memorandum to Members and Consultants of the National Science Board, NSB 04-157, Arlington, Va.: National Science Foundation, 2004.
OCR for page 64
Midsize Facilities: The Infrastructure for Materials Research BOX 3.1 Escalating Costs of Instrumentation Nuclear Magnetic Resonance Spectrometers Nuclear magnetic resonance (NMR) is a technique used by many researchers, often for determining the structure and the relationship between chemical components of a substance (see Figure 3.1.1). Through the first half of the 1980s, the cost of state-of-the-art high-field NMR instruments with a standard bore size was approximately $1,000 per megahertz (MHz). Thus, a 300 MHz instrument was priced at about $300,000, and a 500 MHz spectrometer (the highest-field NMR instrument commercially available at that time) sold for about $500,000. Wide-bore magnets often used in solid-state NMR were more expensive, but the cost increases experienced for spectrometers with these magnets have been similar to those outlined below. When the new-generation 600 MHz instruments incorporating a higher-field magnet were introduced in the mid- to late 1980s, their price was over $1 million. This change represented a significant increase in the cost per megahertz for instruments operating at this higher frequency. The cost for the lower-field spectrometers also increased as a result of technology advances, but the increases were modest, on the order of 10 percent. The next step up in frequency occurred in the mid- and late 1990s, when 750 MHz and then 800 MHz magnets were developed. Spectrometers incorporating these magnets were priced in the $2.0 million to $2.4 million range, another major jump in the cost per megahertz ratio. Later, in 2002, the first commercial 900 MHz systems were sold, with a price tag of approximately $5 million, representing another very significant jump in price. These trends in instrument pricing are illustrated in Figures 3.1.2 and 3.1.3. Throughout the period of the 1990s to the present, the cost of the lower-field instruments, which have become workhorse spectrometers, remained roughly constant, although their capabilities increased. As each high-field magnet was developed, however, it provided increased sensitivity and resolution, which in turn enabled new-frontier research. Thus, despite the escalating costs, these instruments were attractive additions to shared instrumentation resources, and in fact they were essential for facilities concerned about providing state-of-the-art capabilities for users. In the NMR area, another essential accessory, NMR probes (see Figure 3.1.1), also experienced significant price escalation as they incorporated incremental technical advances. The probe fits in the magnet and measures the NMR signal, which is the molecular signature of the sample being studied. In the 1980s, probes for solution NMR cost $12,000 to $15,000 each. During the mid-1990s, they were priced at around $20,000, and in 2005 they were $30,000 to $35,000. Probes with totally new designs that have been developed during the past 5 years to help enable innovative research are priced even higher. Microcoil probes range from $50,000 to $75,000, and cryogenically cooled probes, which were introduced at $180,000 in 2000, now are priced at $275,000 to $350,000. Well-equipped research instrumentation resources need these capabilities to support competitive programmatic research. Secondary Ion Mass Spectrometers Secondary ion mass spectrometry (SIMS) is a standard tool in the characterization suite of a materials research laboratory. A leading manufacturer of SIMS instruments is CAMECA, a company based in France. As Figure 3.1.4 shows, the average purchase price for popular configurations of SIMS instruments has steadily increased.a Note, however, that even after correcting for
OCR for page 65
Midsize Facilities: The Infrastructure for Materials Research FIGURE 3.1.1 (Top) Cryogenic probes with ultrahigh-field NMR magnets enable efficient structural characterization of biomaterials (biological polymers or biological macromolecules). Pictured here is a Varian, Inc., cryogenic probe next to an 800 MHz NMR magnet that also has a cryogenic probe installed. Also shown are ribbon diagrams of the backbone structures (bottom left) for a functional mutant protein of thioredoxin (L78K), which to date has resisted crystallization, and (bottom right) for human carbonic anhydrase II, a 29 kilodalton enzyme. These structures were determined in solution by NMR to 0.22 nm and 0.29 nm resolution, respectively, using very high sensitivity probes such as the one pictured. Courtesy of Duke University Nuclear Magnetic Resonance Center.
OCR for page 66
Midsize Facilities: The Infrastructure for Materials Research FIGURE 3.1.2 One measure of the capability of an NMR spectrometer is the frequency response of hydrogen in megahertz (MHz). Higher magnetic fields allow higher-precision measurements at higher frequencies. In this figure, the cost per megahertz of the highest-magnetic-field instrument available is shown over the past 20 years by the top line. The cost year-by-year is shown and compared to the Consumer Price Index-inflated cost of the 1980 500 MHz machine (bottom line). inflation (as estimated from the Consumer Price Index) to measure the cost in constant dollars, the instrumentation’s average cost grew significantly in the 1980s. This growth coincided with the introduction of significantly enhanced functionality. SIMS characterization techniques were first introduced in the 1950s but were not commercialized until the 1970s. The initial configurations were fairly simple. The first instruments incorporated a large-beam oxygen or cesium primary ion source (not mounted simultaneously) and a simple electron-beam neutralization gun. Over the ensuing years, key advances were made to the system, including the following: Primary beam mass filter, which improved the primary ion beam purity and allowed the simultaneous mounting of two ion sources; Ion sources and ion optics that allowed microfocusing of the primary beam; An improved charge neutralization system for bulk insulators; Dramatic improvement in the quality of the vacuum, permitting better detection limits for the atmospherics; and Sample rotation for improved depth resolution. The foregoing discussion concerns the evolution of what might be considered the main-
OCR for page 67
Midsize Facilities: The Infrastructure for Materials Research FIGURE 3.1.3 The purchase price of the highest-field NMR instruments—from 500 MHz to 900 MHz systems—when they became commercial products over the past two decades (top line). For comparison, the Consumer Price Index-inflated price of the best machine in 1980 is shown (bottom line). As higher-field magnets were introduced, the cost of these instruments has increased in major increments. stream SIMS depth profiling instrumentation. Over the past 10 to 15 years, other configurations of SIMS instruments have become available to the materials research community. These include the following: Time-of-flight SIMS for sensitive, near-surface analysis of elemental and molecular species; Quadrupole SIMS designed for high-resolution depth studies and extremely low bombardment energies; Very large geometry instruments for sensitive, in situ isotopic ratio measurements; Specialty optics for submicron imaging, <50 nm resolution, with parallel detection for precise isotopic ratio measurements; and The coupling of MeV accelerators, for molecular ion fragmentation, to conventional SIMS instrumentation to measure very low-lying, non-naturally occurring (generally radioactive) isotopes. Again, a caveat: the last three embodiments are very specialized and more sophisticated
OCR for page 68
Midsize Facilities: The Infrastructure for Materials Research FIGURE 3.1.4 History of the average purchase price of top-selling secondary ion mass spectrometry (SIMS) instruments over the past approximately two decades. The cost year-by-year is shown (top line) and compared to the Consumer Price Index-inflated cost of the instrument package, circa 1980 (bottom line). The growing separation in the two trajectories indicates that the increase in the purchase price is driven by much more than just inflation. Note, however, that in some instances, the purchase price of SIMS instruments actually dropped, indicating that without the introduction of enhancements and new capabilities, instrumentation becomes more inexpensive as time passes. than are the conventional analytical SIMS instruments. As such, they are much more expensive, with accelerator-based configuration generally built as a one-off, custom instrument. Electron-Beam Lithography Systems In 1981 a group of researchers in microfabrication at Bell Laboratories began requesting proposals from vendors to build an electron-beam lithography tool that combined a high-voltage, high-resolution electron column with the precise scanning and the field stitching capability made possible by a laser interferometer controlled stage. The result was that in 1985 a Japanese company, JEOL, shipped its first JBX 5DII, a 50 kiloelectronvolt (keV) system with a 3 inch x-y stage, to Holmdel, New Jersey. The price tag was about $1.3 million. The PDP 11 computer that controlled the system cost about $200,000. Since that time computers have become less expensive, but electron-beam systems have not. The cost of the electron-beam
OCR for page 69
Midsize Facilities: The Infrastructure for Materials Research FIGURE 3.1.5 History of typical costs of electron-beam lithography machines over the past 20 years. system has steadily grown as the high-precision stages have been required to accommodate first 4 inch, then 6 inch, then 8 inch, and soon 12 inch wafer sizes. In addition, vendors have included higher-brightness electron sources (e.g., thermal field emitters), which require ultrahigh vacuum column components. Deflection amplifiers and scanning rates have improved from 2 MHz then to 50 MHz today. Nanolithography researchers have pushed for higher accelerating voltages to improve image resolution and contrast in order to explore ever-finer fabrication frontiers, and optoelectronics applications have pushed vendors to provide finer addressability. These improvements have driven the 1980s price (thermionic source, 3 inch, 50 keV) to $3 million to $4 million in the 1990s (thermal field emission [TFE] source, 6 inch, 50 keV) to $5 million to $7 million today (TFE, 8 to 12 inch, 100 keV) (see Figure 3.1.5). The escalating cost of these systems stems partly from the small number of them built and delivered worldwide each year. The amount of engineering that goes into each generation of tool is enormous. The small volume leaves vendors reluctant to consider lower cost options,
OCR for page 70
Midsize Facilities: The Infrastructure for Materials Research such as placing a newer column on an older or smaller stage. This route would create orphaned machines, which are difficult to service and upgrade. Also responsible for the trend, however, is the varied nature of the user community. Systems placed in shared user facilities need to have the near-Angstrom-scale spatial coherence to produce gratings for optoelectronics one day and the ultrahigh resolution to attach contact electrodes to nanotubes or molecular electronics the next day. They need to have automated loaders to allow around-the-clock sample exchange and multisize substrate compatibility. The improved scanning speed allows access to greater design complexity or simply exposes faster and therefore performs for more users each year. The seemingly long list of specifications and features is driven by the need to support the broadest possible research base. a The reader should note that there is an alternative SIMS that employs a quadrupole mass filter rather than the double-focusing magnetic spectrometer of the CAMECA. These instruments have gone through similar price increases (albeit from a lower base) and enhancement of capabilities in the time frame from the 1980s to the present. A maintenance contract (approximately 5 percent of the capital cost per annum); A full-time professional and/or technical support staff person (about $100,000 per annum, including benefits); and Consumables, supplies, and replacement parts (about $50,000 per annum). These costs amount to a total of about $250,000 per year. If that amount was to be recovered from user fees alone, a cost-recovery rate of $125 per hour would be required for a 2,000 hour year. That hourly fee significantly exceeds the stated level of tolerance (among academic users) of $100 per hour identified by the committee’s survey, as discussed in the section “User Comments” in Chapter 2. (See Figure 3.1 for another example.) Any instrument downtime then becomes a further serious problem. This cost analysis excludes the tool cost or depreciation that would be a major factor for establishing a commercial analytical service rate; that is, the commercial rate for the same services would be significantly higher. However, this additional constraint further exacerbates the cost-recovery gap. For its commercial users, a grant-subsidized facility must take care to avoid providing services that are otherwise available in the marketplace. This important precept is discussed further in the section below, “Cooperation and Noncompetition with Commercial Interests.”
OCR for page 71
Midsize Facilities: The Infrastructure for Materials Research FIGURE 3.1 Example of rising costs of maintenance at midsize facilities: comparison of the capital expenditures and the cumulative operations and maintenance cost for a typical and modest field-emission transmission electron microscope. Because the annual maintenance costs are typically covered under a service contract, the price is the same in the first year as it is in the tenth. To bridge the gap between cost recovery through user fees and the recurring expenses of maintaining and operating the facility, midsize facilities must identify alternative sources of ongoing support. The committee found that it is more difficult to cover ongoing maintenance costs than to obtain one-time capital funding! This lack of continuity is a critical issue—funding policy needs to be addressed in a serious fashion. The committee believes strongly that when a shared user facility is established and its operation has proven successful, continuity or stability of funding is required in order to realize the original objective. This funding is extremely important to enable a long-term plan of maintenance, equipment replacement, and staffing. To maintain operations with a relatively high degree of confidence in their not being interrupted, many successful facilities visited by the committee have diverse sources of funding rather than relying on a single source (see Figure 2.2 in Chapter 2). A diverse portfolio of continuing support enables such facilities to minimize the risks associated with the reduction of any particular funding source,
OCR for page 78
Midsize Facilities: The Infrastructure for Materials Research such as clean rooms and expensive instruments, or at the very least to facilitating an informed prioritization. Regional cooperation in these efforts can have a similar effect, and it should be encouraged as a best practice. See Box 3.3, “Consortium Approach to Extraterrestrial Sample-Return Missions,” for an example of a regionally networked approach to a fascinating scientific opportunity. The committee carefully considered the notion of quantitative performance metrics. Ultimately, of course, midsize facilities should be optimized for “usage”—some intangible measure of the ratio of the level of invested resources to the delivered output. With the data from its questionnaires and its site visits, the committee found it tempting to construct estimators of usage per unit of investment capital or of operating budget. However, in part because midsize facilities are so diverse and in part because of the limited extrapolating power of the data sample, there were no clear answers. A more complete and statistically robust data set with consistent and uniform definitions would enable more quantitative measures. For instance, accounting schemes are not standardized across facilities, across programs, or across agencies, so even just the denominator is hard to determine. Likewise, not all midsize facilities have the same purpose (some are directed at advancing research, some focus on education, and so on). This understanding of the diversity involved is one of the reasons that the committee advocates the combined use of user-feedback mechanisms and facility reviews by committees of scientific and management experts. Maintaining a Balanced Suite of Equipment It is important that facilities have a balanced suite of equipment that enables users to exploit the full capabilities of the instrumentation. For instance, a midsize TEM facility requires instrumentation for x-ray energy dispersive spectroscopy and electron energy loss spectroscopy as well as sample-preparation equipment. Synthesis and/or preparation of materials samples, for example, can present a serious obstacle, especially for users obliged to travel significant distances—a problem so critical, in fact, that the nation’s need for crystal-growing facilities was the subject of an October 2003 DOE workshop.10 In the committee’s discussions with facility managers and users, it was clear that an important attractor for users is one-stop shopping for the fabrication, synthesis, and characterization of materials and the measurement of their properties. 10 Department of Energy, Design, Discovery and Growth of Novel Materials for Basic Research: An Urgent U.S. Need, Washington, D.C.: Department of Energy, October 2003. Available online at http://www.science.doe.gov/bes/dms/Publications/DMSE_Sponsored/Xtal-Growth.pdf; last accessed June 1, 2005.
OCR for page 79
Midsize Facilities: The Infrastructure for Materials Research BOX 3.3 Consortium Approach to Extraterrestrial Sample-Return Missions The San Francisco Bay Area is the home of two national laboratories and several universities. Each of these institutions offers extensive analytical and technical capabilities. In 2003, the Bay Area Particle Analysis Consortium (BayPAC) was formed with the Lawrence Livermore National Laboratory (LLNL), the Advanced Light Source (ALS), the Space Sciences Laboratory (SSL) at the University of California at Berkeley, and the Stanford Synchrotron Radiation Laboratory (SSRL) (see Figure 3.3.1). The approach of the consortium is to leverage technical and analytical expertise at each of the institutions so as to maximize the science return from the Stardust mission and future sample-return missions of the National Aeronautics and Space Administration (NASA). FIGURE 3.3.1 A diagram showing the interaction of institutions in the Bay Area Particle Analysis Consortium (BayPAC). Note the hub-and-spoke arrangement. See the text later in this box for definitions of the acronyms.
OCR for page 80
Midsize Facilities: The Infrastructure for Materials Research Stardust was launched in February 1999. It is the first “solid matter” sample-return mission since Apollo 17 in 1972. Unlike the Apollo missions that returned rocks to the laboratory, Stardust will return just micrograms of dust in the form of several thousand individual particles, most of them less than 10 micrometers in diameter. BayPAC is preparing to confront the technical and analytical challenges that will be presented by the Stardust samples when the sample-return capsule is opened in 2006. The SSL is developing methods to extract captured particles from the aerogel collectors (silica aerogel is the capture medium used on Stardust); LLNL is developing NanoSIMS (-secondary ion mass spectrometry), ion microprobe, nuclear microprobe, dual-beam focused-ion beam (FIB), and Super-STEM (-scanning transmission electron microscopy) methods for the analyses of individual particles; SSRL is refining total reflection x-ray fluorescence (TXRF), x-ray absorption fine-structure spectroscopy (XAFS), and x-ray absorption near-edge structure (XANES) techniques; and ALS is refining x-ray fluorescence (XRF) techniques and providing a beam line for Fourier transform infrared (FTIR) spectroscopy. Since the consortium was formed, NASA has provided several new grants to BayPAC members. The membership of BayPAC provides a wide range of expertise from the fields of chemistry, materials science, geology, physics, and astrophysics that enables a unique level of interpretation and analysis of acquired data. The investment in the analytical and technical capabilities developed for Stardust by BayPAC can be further utilized on future sample-return opportunities such as Hayabusa (a Japanese mission to an asteroid) and NASA’s proposed Gulliver mission to Mars satellite Deimos. Multiple particle-capturing studies are in progress because small sample return appears to be the way of the future, and NASA appears to be committed to a long-term effort in this arena. Specialty technologies are being developed to capture samples from asteroid surfaces, and hypervelocity capture is being studied as well. Facilities at the Johnson Space Center, McDonnell Center for Space Sciences at Washington University, and in England are also involved in these efforts. Users were more satisfied with midsize facilities that provide a variety of techniques and tools to accomplish their tasks; the complementary nature of synthesis/fabrication tools and characterization/measurement tools was noted by many users. Similarly, the committee found that, in general, facilities that offer equipment for training, education, and practice alongside topflight cutting-edge tools are regarded more highly and engage a broader group of users. Suites of instruments at successful midsize facilities strongly reflected the usage patterns of the local community. If certain instruments were not sustainable (through lack of use, high use by relatively few researchers, or costliness of operation), they were passed off to individual investigators. Heavily used instruments became top priorities of facilities; in some cases, a facility’s director sought to obtain equipment
OCR for page 81
Midsize Facilities: The Infrastructure for Materials Research that was popular elsewhere in order to attract additional users to the facility. The committee found space and resources to be highly optimized at successful facilities. A balanced instrument suite not only includes existing equipment but some flexibility to develop new tools in response to the individual and sometimes unique needs of researchers. A potential consequence of the development of new instrumentation or techniques is the creation of new intellectual property (IP). Facilities housed at national laboratories and academic institutions are generally not well equipped to turn IP into a value statement, particularly when the development is incremental or very application-specific. In part because facilities do not have secure resources, the instrumentation efforts must be less exploratory and more focused. Although universities are increasingly encouraging the licensing and transfer of new technologies developed on campus, the committee doubts that development of IP could become a serious component of a facility’s revenue stream: the lucrativeness of technology transfer is simply too unpredictable, and the institutional overhead is not designed to allow the facility itself to profit directly. NETWORKING WITH OTHER FACILITIES TO PROVIDE BALANCED RESOURCES In some instances it is not financially possible to support a complete suite of instrumentation at a single facility or at one institution. A useful solution has been to develop a network with other facilities that can effectively provide users with needed access in a timely fashion. Support for such networking should be a recognized priority of federal agencies. In other cases, particularly where different but complementary equipment might be needed, it is often not feasible for a single institution to support all of the technologies required. Recognition of this limitation is becoming increasingly important for projects that are interdisciplinary—many activities in the materials science area are increasingly so, especially those dealing with biological materials. For example, investigations of macromolecular structures might very well need good access to synchrotron sources for x-ray crystallography, a neutron source such as the Spallation Neutron Source for neutron scattering, ultrahigh-field NMR instrumentation, and cryogenic electron microscopy. Facilities for each of these instrumentation types range from midsize to very large and are rarely, if ever, present in one location. In such situations, effective models for the support and networking of these capabilities would have a very positive impact on the timely advancement of materials research projects and the continued development of frontier science. The National Nanotechnology Infrastructure Network (NNIN) (see Appendix F for details) and the Network for Computational Nanotechnology approaches recently developed by NSF are a major step forward in this regard.
OCR for page 82
Midsize Facilities: The Infrastructure for Materials Research One of the primary motivations for the formation of the NNIN was the benefit of pooling resources (tools, staff, and funds), and its success can prove the virtue of this goal. NNIN has a formal governance structure for coordinating activities across the different sites; it is highly organized because of its central mission. The management structure for NNIN is probably larger than would be appropriate for a national association of midsize facilities loosely affiliated for the purposes of communication, referrals, and shared planning. As noted earlier, midsize facility users predominantly come from the local area—in academic settings, this might be just the on-campus community. In fact, several major universities (e.g., Rice University and Harvard University) have completed a campus-wide inventory of their instrumentation, motivated by the opportunity to optimize their investments. These efforts also provided information on instrument use. Certain materials research tasks, such as fabrication, require quick and easy access to tools, driving the need for local facilities. Other tasks, such as analysis and synthesis, are already handled predominantly through remote service providers. These observations suggest that, at second glance, regional consolidation of midsize facilities would not be easy. In fact, it might hinder certain phases of the research enterprise, at least until the scientific user community adapted. The fact remains, however, that in a revenue-neutral environment, solutions that make existing resources available to a broader community will be necessary. It is equally important that, within a given university, efforts are made to avoid duplication of facilities and to coordinate and consolidate capabilities, particularly in large institutions with multiple user facilities (such as major research universities). Sometimes the greatest barriers to networking or teaming with other facilities exist within different facilities at the same site. That is, the committee found that midsize facilities are dominated by local users, and in such circumstances, networking with another facility on campus can be perceived to have a negative impact—the new user community could easily dominate the facility. Similarly, connecting across departments in a university situation may incur additional difficulty because of administrative involvement. The committee recognizes the challenges described above, but successful proofs of principle have been demonstrated by a number of universities, and most national laboratories have been required to address this issue for quite some time. Networking with a neighboring facility is a partnership for coordination and communication, not a way to transfer users. For instance, regular discussions taking place at Purdue University between two microscopy facilities (one for physical sciences and one for life sciences) represent a form of networking. While on-site networking represents a prime example of small regional networking, it may not evolve at the same rate that longer-distance networking relationships develop.
OCR for page 83
Midsize Facilities: The Infrastructure for Materials Research Networking has a concrete meaning in terms of cyberinfrastructure as well; modern computer interfaces and high-speed networks make operation of many instruments “at the console” a virtual experience. Remote access has often been touted as a route to engage nonlocal users; through video or computer interfaces, remote researchers can either directly manipulate the controls or watch as other experts perform the tasks. The committee’s judgment is that, while remote access can enhance the radius of effective impact for casual education and training of a midsize facility, much of the vitality and value of facilities is obtained in person and on site. That is, routine analysis tasks can be dispatched to a remote facility, but researchers pushing the limits of scientific and technological know-how invariably prefer to visit the facility in order to interact with the instrumentation and professional staff personally. BALANCING COMPETING PURPOSES Smaller and midsize facilities often face problems posed by competing purposes. It is important that these facilities recognize these potential challenges and take steps to minimize conflicts, both when planning the creation of a facility and also during its operation. A major issue in this regard is that of achieving the appropriate balance between the need to train students (with equipment breakage being an inevitable part of the learning process) and the need to maintain state-of-the-art performance. Several midsize facilities, such as the Microfabrication Laboratory at the University of California at Berkeley and the NanoScale Science and Technology Facility at Cornell University, achieve this goal with a strictly organized training and sign-up system for each tool: users are simply not permitted to use a tool until they have received training and certification. However, this system appears to work best for larger facilities with multiple tools, where a large pool of technicians or advanced students is available to conduct training. In other facilities, such as NCEM, users are generally expected to be competent in the use of basic instrumentation when they arrive, and then they receive technicians’ specialized help in the use of advanced equipment. This approach has proved a challenge to NCEM staff, given some users’ expectation that the staff will train students in basic microscopy! For facilities that are in great demand, this problem is usually minimized by giving preferential treatment to users who already have the required basic knowledge, thus reducing the facility’s teaching load. However, researchers from smaller universities that cannot provide such basic training can thus be put at a disadvantage. Given adequate resources, each facility would ideally operate basic as well as advanced equipment, but this approach can be a challenge,
OCR for page 84
Midsize Facilities: The Infrastructure for Materials Research given the reality of limited resources and the preference (of researchers and agencies) for funding advanced equipment. A related and equally important issue is the need to balance the development of instrumentation, staff-initiated research, and the routine provision of services. Taking an advanced tool offline so that it can be used for instrument development, for example, can disrupt the operation of a facility. Facility managers who schedule this sort of activity view instrument development as an essential part of their mission and as one that benefits their facility in the long term. In many cases noted by the committee, facility members’ research activities have in fact been based on development of the facility’s instruments. For example, 10 years ago the Center for X-Ray Optics at LBNL undertook a major effort to adapt an electron-beam lithography tool so that it could operate with polar pattern generators rather than conventional rectangular scanners. This system became a research focal point because of its unique fabrication capability, and it has been used since to make world-class diffractive optics for the ALS at LBNL and elsewhere. Facility members given an opportunity to enhance and extend the equipment, even by making small modifications or developing accessories, have a much greater stake in ensuring that their tools operate at as high a level as possible. An ongoing instrument development activity, however small, should be considered in planning for a facility to ensure balance in the basic activities of a midsize materials research facility.11 Also important is achieving a balance between outreach and research. Time spent on educational outreach, for example, to local elementary and high school students, although an important and rewarding experience, can detract from research time. This issue, of course, is common to every NSF program, but there are alternative forms of outreach that may in fact help to enhance the unique opportunities offered by midsize facilities. For instance, an interesting option would be to direct outreach toward the training or support of technicians, alleviating to some extent the staffing problems typical of many midsize facilities. A related issue, the importance of which is increasing given the availability of high-speed computer links, is that of balancing remote operation and hands-on experience. From one perspective, as a teaching tool, hands-on experience is irreplaceable; in addition, visits by students to a facility allow them to interact with 11 One caution, however, is that one-of-a-kind instruments can become orphaned because commercial repairs may be unavailable, a consequence that should be weighed, anticipated, and mitigated where possible.
OCR for page 85
Midsize Facilities: The Infrastructure for Materials Research other researchers and thus broaden their experience. From another perspective, remote operation can be more cost-effective and can engage a broader community. One final issue involving balance relates to the openness of access to facilities for research in materials. The process that potential users must follow to gain access needs to reflect a balance that minimizes the time before experiments can be scheduled while maintaining safety and security. Once potential users have been approved and are present at a facility, their having the greatest possible access to the facility’s tools—especially outside normal working hours—is extremely important. Such access will ensure the best use of limited time and resources and will promote the rapid developments associated with progress in areas such as nanotechnology. Restrictions that limit such access are detrimental to the research enterprise. Indeed, the Berkeley Microlab’s experience was that tools that can be self-scheduled by students for use outside standard business hours were used more efficiently than were tools run by staff during business hours. In general, universities offer a different environment for balancing efficient access with security and safety than do the national laboratories; each environment achieves its own best compromise. COOPERATION AND NONCOMPETITION WITH COMMERCIAL INTERESTS It can be tempting for midsize facilities to recoup operating costs by providing analytical services to commercial users. The issue is that government-funded facilities can, in principle and sometimes in fact, perform analytical research at costs substantially below those of commercial analytical service laboratories or nanofabrication laboratories, since the commercial laboratories must recover substantially greater expenses. For instance, because the recovery of the cost of depreciation for government-funded capital investments at universities is not allowed, many overhead costs are not recognized, and no profit is generally sought; such facilities arrive at low figures when they prepare their “cost-recovery” user fees. These facilities might thus charge only an amount necessary to cover their direct, out-of-pocket costs. This accounting method is perceived by commercial laboratories as allowing unfair competition, since they must realize all of their costs in their pricing structure. It is viewed as tax dollars in effect supporting a competitor. This issue has been a problem since the 1970s, when major instrument purchases were funded by the federal government and commercial laboratories began operation in the same time frame. Recognizing the problem, the NSF issued “Important Notice to Presidents of Universities and Colleges and Heads of Other National Science Foundation Grantee Organizations, Notice No. 91” (Important
OCR for page 86
Midsize Facilities: The Infrastructure for Materials Research Notice 91)12 during the 1980s. The issue was again addressed in 1998 by NSF in “Important Notice to Presidents of Universities and Colleges and Heads of Other National Science Foundation Grantee Organizations, Notice No. 122” (Important Notice 122)13 and by the Office of Management and Budget (OMB) in Circular A-21.14 Current NSF policy recognizes and encourages cooperation between universities and the industrial and manufacturing sectors. In advanced study and research, such cooperation not only promotes a more rapid development and dissemination of knowledge, but it also contributes to economic development. Use of NSF-sponsored facilities for direct collaboration with the industrial and commercial sector is allowed. If the collaboration is open and nonproprietary and the results are published in a timely manner, it may be appropriate to charge the lower fees generally charged to academic users. Otherwise, full commercial rates should be charged, as explained below. Note that the preceding statements distinguish between “collaboration” and “competition” with a commercial partner—often the area of difficult judgment. Additional local oversight of this issue could be helpful. For instance, when preparing the annual reports for their facilities, managers should describe all partnerships with commercial entities; each activity should be categorized as collaborative, full-cost-recovery services. This effort would clearly delineate all nonproprietary work and put it in the public domain, even though it might never be formally published in a scientific journal. It is the responsibility of each university’s administration to ensure that a policy is in place and is properly followed for determining which fee structure to use. This committee recommends that all facilities funded in part or in whole by the federal government adhere to the federally mandated guidelines. Some facilities have in fact developed policies that convey the letter and spirit of the NSF and OMB notices. Usage of NSF-sponsored facilities is specifically governed by NSF Important Notice 122. Potential commercial users of such a facility are encouraged to become familiar with this announcement. OMB Circular A-21 is meant to govern all facilities that are funded in whole or part by the federal government. Because of the keen sensitivity to the issue of unfair competition between commercial analytical service and fabrication laboratories and federally funded 12 NSF Important Notice 91 was replaced by Important Notice 122, but the earlier notice can still be viewed online at http://prism.mit.edu/nsf.in91/in91txt.htm; last accessed June 1, 2005. 13 NSF Important Notice 122 is available online at http://www.nsf.gov/pubs/1998/iin122/iin122.txt; last accessed June 1, 2005. 14 OMB Circular A-21 is available online at http://www.whitehouse.gov/omb/circulars/a021/a021.html; last accessed June 1, 2005.
OCR for page 87
Midsize Facilities: The Infrastructure for Materials Research facilities, the committee endorses several guidelines here for the provision of services by midsize facilities to the industrial and commercial sector. These guidelines are a modified version of policies developed by the Cornell Center for Materials Research:15 Commercial use of the facility must not interfere with the research mission of the facility. Appropriate fees must be charged to recover full costs. Fees for services to commercial businesses must not be less than fees charged for equivalent services from viable commercial vendors or facilities. Excess capacity must be available to provide the services to the commercial sector. It is the responsibility of each facility’s manager to establish (using reasonable judgment) whether equivalent services are available from the private sector. In ambiguous cases, a member of the organization, educational institution, or national laboratory management who is not directly involved in the management of the facility should assume responsibility for establishing whether services are equivalent. In determining whether equivalent services are available, a facility’s manager should do as follows: Identify the specific activities required by the potential commercial user. Take into account the capability of the compared instruments, the fragility of the specimens involved, the specialized expertise of the technicians involved, the need for special adjustments or accessories, and the number of samples being tested or the frequency with which the user must repeat processes. Document the results of the assessment of equivalent services on a case-by-case basis as requests from potential commercial users are received. Have on hand current pricing information from representative commercial laboratories if the manager intends to provide services to the commercial sector. Avoid allowing such documentation to significantly interfere with the research-related activities of the facility’s manager. 15 Cornell Center for Materials Research, Policies and Procedures for Shared Experimental Facilities, Ithaca, N.Y.: Cornell University, November 2000. Available online at http://www.ccmr.cornell.edu/images/pdf/CCMR_SEF_Policy.pdf; last accessed June 1, 2005.
OCR for page 88
Midsize Facilities: The Infrastructure for Materials Research SUMMARY Long-term infrastructure affects every aspect of a facility’s long-term viability, ranging from the ability to plan for instrument acquisition and replacement to the retention of skilled staff and the ability to define and carry out a sound management plan. A related challenge for midsize facilities is that of publicizing their capabilities to attract the best research. As discussed in Chapter 4, current funding models (within agency programs) do not address these needs well. Midsize facilities in materials research have been traditionally thought of as independent units that function individually. Because of growing opportunities and increased demands, there is an increasing need for efficient networking and interaction between facilities in order to avoid unnecessary duplication—and to ensure that instrumentation is present in regional or local facilities in proportion to the needs of the regional or local communities. There is currently no framework to facilitate these types of interactions for planned or existing facilities. Another common challenge involves the difficulty of balancing competing purposes, such as training versus research. A key strength of midsize facilities is their flexibility, but this characteristic can also be a weakness—efforts to maximize usage, train students, and facilitate world-class research (for instance) can interfere with one another. Finally, the close parallels between midsize facilities and commercial analytical service laboratories in materials research can provide challenges. Compliance with federal guidelines and regulations in these areas is critically important to maintaining a healthy symbiosis between midsize facilities and commercial ventures.