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.
10 Structure and Level of the Current Research Effort Condensed-matter and materials physics (CMMP) is the largest and, in its scope, most diverse subfield within all of physics. By its nature, CMMP is also fun- damentally a crosscutting research enterprise; it connects a multitude of subfields within physics, from semiconductors to biological physics, and it links physics with a multitude of other disciplines, from engineering to computer science to applied mathematics. This chapter looks at the available data for assessing the current level of federal research support, the demographics of the field, and the research output as measured by publications. Where data were available, the Committee on CMMP 2010 also compares the United States with other countries. Federal Funding for CMMP Research Federal support for CMMP research activities occurs through several fund- ing venues, supporting work at universities and national laboratories. Agencies were asked to provide data for the past 10 years on funding levels, grant sizes, and success rates, as well as on the demographics of the CMMP-supported programs. Additional questions concerned the number and size of centers and of facilities supported. The breadth and range of research activities falling under CMMP make it difficult to arrive at a clean breakdown, especially since the federal agencies delineate their programs differently and much of the research is interdisciplinary. Besides condensed-matter physics and condensed-matter and materials theory, CMMP-related programs include aspects of ceramics, metals, polymers, electronic materials, and solid-state chemistry. More recently, nanoscience and nanotechnol- 172
Structure and Level of the Current Research Effort 173 ogy have emerged as areas with a large and inherent overlap with CMMP activities. For the funding trends discussed here, the committee has made an effort to focus on CMMP-related physics efforts. The primary funding sources in terms of the total amount of funds allocated for CMMP research over the past 10 years have been the National Science Foun- dationâs Division of Materials Research (NSF DMR); the Department of Energy Basic Energy Sciences (DOE BES); the Department of Defense Army Research Office, Air Force Office of Scientific Research, and Naval Research Laboratory (DOD ARO, AFOSR, and NRL, respectively); and the National Aeronautics and Space Administrationâs Materials Physics and Condensed-Matter Physics programs (NASA MP and CMP). The total funding for CMMP basic research from the sources listed above has varied little over the past 10 years. When corrected for inflation, using Office of Management and Budget (OMB) deflators, the net increase over this period is roughly 10 percent. Figure 10.1 gives the overall funding level for CMMP basic research at universities in inflation-adjusted fiscal year (FY) 2006 dollars. The total amount of support shown in this figure, roughly $600 million per year, includes the direct funding of research through grants to individual investigators and small 700 600 500 FY 2006 $ (millions) 400 300 200 100 0 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 NSF DOE DOD NASA FIGURE 10.1â Federal investment in condensed-matter and materials physics basic research in Âterms of inflation-adjusted FY 2006 dollars. SOURCE: Data supplied to the Committee on CMMP 2010 by the respective agencies.
174 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 groups of investigators (such as Focused Research Groups at NSF) both at univer- sities and national laboratories, and NSF DMR support for research centers such as the Materials Research Science and Engineering Centers (MRSECs), Science and Technology Centers (STCs), and Nanoscale Science and Engineering Centers (NSECs). In addition to the research support shown in Figure 10.1, CMMP ben- efits from investments by the federal government through the support of national laboratories, including the new DOE nanocenters, and investments into the con- struction and maintenance of large facilities, such as facilities for x-ray and neutron scattering, electron microscopy, high magnetic fields, and large-scale computation. These facilities are discussed in more detail in Chapter 11. It is important to note that within each agency, individual program areas have evolved during the past 10 years, with some growing at modest amounts and oth- ers declining somewhat. However, there are situations in which downward fluc- tuations in specific programs of several agencies occur at the same time, thereby threatening whole areas of CMMP. This is particularly critical for smaller research efforts and emerging areas. During the past couple of years this has happened, for example, for research on colloids and granular materials. In other instances, whole programs have been eliminated, such as NASA CMP (at the end of 2006). Specific research areas strongly dependent on these funding sources are now in jeopardy in the United States. For research at universities, a major cost factor is the support of graduate students. A survey by the Committee on CMMP 2010 of nine state and private universities indicates an average increase over the past 10 years of about 5 percent per year in the cost per graduate student, as charged to federal grants related to CMMP research. This increase includes the salary, health fees and tuition, and overhead. As pointed out in the National Research Council (NRC) study on the Materials Research Science and Engineering Centers, the use of this deflator more realistically captures the actual costs associated with carrying out research at universities. In the present report, the committee identifies this deflator with the funding âbuying power.â Figure 10.2 indicates that, using this measure, the overall support for CMMP research has steadily declined over the previous decade. The NRC study Midsize Facilities also found that instrumentation costs over the past 10 years have risen far more quickly than inflation. Since the actual cost of do- ing research is closely tied to the cost of supporting personnel, such as graduate âNational Research Council, The National Science Foundationâs Materials Research Science and Engi- neering Center Program: Looking Back, Moving Forward, Washington, D.C.: The National Academies Press, 2007. âNational Research Council, Midsize Facilities: The Infrastructure for Materials Research, ÂWashington, D.C.: The National Academies Press, 2006.
Structure and Level of the Current Research Effort 175 1.15 1.10 1.05 (FY 2006 $, normalized to 1996) CMMP Research Support 1.00 0.95 0.90 0.85 0.80 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 FY 2006 inflation-adjusted dollars FY 2006 5% per year inflation FIGURE 10.2â U.S. funding for basic research in condensed-matter and materials physics (CMMP), in FY 2006 dollars, normalized to 1996. The top curve takes inflation into account by using the Office of Management and Budget deflator (the same data as for Figure 10.1 are used). The bottom curve uses the average yearly increase in costs to support a graduate student (5 percent) as the deflator. This clearly indicates that CMMP research support experienced a decrease in ânet buying powerâ over the past decade. 10-2 students, and instrumentation, CMMP research in the United States can only stay competitive if the level of funding increases in a commensurate way. Within individual programs, funding can be delineated by support for indi- vidual investigators (or small teams), support for facilities, and support for larger groups such as research centers. Figure 10.3 gives this breakdown for NSF DMR for the past 10 years. It is clear from these figures that, in NSF DMR, the support for individual principal investigators (PIs) has closely tracked overall DMR growth. Both have grown so modestly that, in taking the graduate student cost inflation into account, the âbuying powerâ of an NSF grant has decreased (see Figure 10.2). For centers, the funding trend has been even worse, as they have been essentially flat-funded in inflation-adjusted dollars. At DOE BES, grant support for CMMP research has also increased, from $181 million in 1996 to $220 million in 2006. This corresponds to an approximate 20 percent increase in as-spent dollars over the time period. Hence, when the
176 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 a 300 NSF DMR Budget (FY 2006 $, millions) 250 200 150 100 50 0 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 Single PIs Centers Facilities DMR Total b 1.6 1.5 (FY 2006 $, normalized to 1996) 1.4 NSF DMR Budget 1.3 1.2 1.1 1.0 0.9 0.8 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 Single PIs Centers Facilities DMR Total FIGURE 10.3â The National Science Foundation Division of Materials Research (NSF DMR) funding profile for the past 10 years. The data are in inflation-adjusted FY 2006 dollars. (Top) Funding levels for the identified programs (data provided by NSF). (Bottom) These same funding levels normalized 10-3a, b to 1996 levels. Note that the data shown for facilities include the Major Research Instrumentation and Instrumentation for Materials Research programs, which support small teams of principal investiga- tors (PIs) in addition to supporting large-scale NSF facilities. a,b see note on fig 10-4
Structure and Level of the Current Research Effort 177 i Â ncreasing cost for supporting a graduate student is factored in (5 percent per year on average, based on the past 10 years), the net âbuying powerâ has declined. Funding Success Rates The essentially flat federal funding for CMMP basic research over the past 10 years, coupled with an overall increase in the number of grant applications, has led to an increase in pressure on grant allocations. A modest increase in average grant size for DOE and NSF (discussed below) thus implies a significant drop in success rate, defined as the ratio of the number of grants awarded by a program to the number of applications (Figure 10.4). For NSF DMR, the average success rate dropped precipitously over the past 5 years, from 38 percent in 2000 to 22 percent by 2005. Success rates for applica- tions by new investigators, that is, investigators not supported by NSF within the 5 years prior to the application, have always been lower, but until 2001 they were in the 20Â percent range. For NSF DMR, the two rates track each other very closely (see FigureÂ 10.4, bottom); success rates for ânew investigatorsâ have fallen from 28Â percent in 2000 to 12 percent in 2005. At DOE BES, new grant applications to the Division of Materials Sciences and Engineering (DMS&E) have had fairly steady average success rates of around 25 to 30 percent over the past decade (FigureÂ 10.5). Success rates for renewal applications, however, during 2000 to 2005 dropped from percentages in the high 90s to 87 percent for condensed-matter physics and materials chemistry (CMP and MC), and to about 62 percent for materials and engineering physics (MEP). Because the DOE, unlike NSF, actively encourages, or discourages, full proposals on the basis of previously submitted âwhite paperâ pre- proposals, success rates reported by the DOE tend to be significantly higher than at NSF. This makes these drops in funding success rates especially significant. As success rates drop, the effort spent by PIs in preparing more grant appli- cations and the effort spent by the funding agencies in managing and reviewing increased numbers of applications become a burden on the federal funding system. Certainly when success rates drop as low as 10 percent, the overall amount of en- ergy spent in the application process becomes counterproductive. As the data show, the CMMP research community is now reaching this point for new applications at NSF DMR and is heading in the same direction for established PIs. These are extremely worrisome trends. As the above data for the increases in submitted proposals and concomitant drops in success rates indicate, there is tremendous pressure on the CMMP fund- ing system. This pressure is further increased by the loss of the premier industrial laboratories. However, the importance, vigor, and increasing breadth of CMMP, as outlined in this report, demonstrate that the field is healthy and indeed poised
178 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 a 1.5 1.4 1.3 1.2 (normalized to 1996) NSF DMR Statistics 1.1 1.0 0.9 0.8 0.7 0.6 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 Proposals Submitted Success Rate Median Grant Size, 2006$ b 40 35 30 Success Rate (%) 25 20 15 10 5 0 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 DMR Average New PI, DMR Average FIGURE 10.4â Grant proposals, success rates, and grant sizes for the National Science Foundation 10-4 a, b Division of Materials Research (NSF DMR). (Top) DMR statistics normalized to 1996. Shown are the these two figures were marked for Left and Right number of proposals submitted, the median grant size (in inflation-adjusted FY 2006 dollars), and the success rate for established principal investigators (PIs) (same as the upper curve in the bottom panel). (Bottom) Success rates for PIs (average over all DMR programs). The upper curve represents success rates for PIs who received funding from the NSF within the previous 5Â years, while the lower curve represents PIs new to NSF. a b 1.5 40 1.4 35 1.3 30 1.2 SF DMR Statistics ormalized to 1996) Success Rate (%) 25 1.1 20 1.0 15 0.9
Structure and Level of the Current Research Effort 179 FIGURE 10.5â Success rates for new and renewal proposals at the Department of Energy (DOE) Basic Energy Sciencesâ Division of Materials Sciences and Engineering as analyzed using the Information Management for the Office of Science database. NOTE: CMP, condensed-matter physics; MC, materials chemistry; MEP, materials and engineering physics. SOURCE: DOE Office of Basic Energy Sciences, Division of Materials Sciences and Engineering. to grow. Recent reports, strongly recommend an increase in the funding of the physical sciences, such as CMMP, to the level of a doubling in the funding over a 10-year period to maintain U.S. economic innovation relative to the rest of the world. When considering the increasing costs of doing CMMP research in terms of graduate student support, a doubling of funding levels over 10 years would repre- sent a growth of approximately 20 percent over a decade. The committee believes this growth level to be necessary for the United States to sustain a competitive position in CMMP worldwide, also taking into account the need to nurture the new subfields that are developing at the interdisciplinary frontiers between CMMP and other physics subfields and other disciplines. âNational Academy of Sciences, National Academy of Engineering, and Institute of Medicine, Ris- ing Above the Gathering Storm: Energizing and Employing America for a Brighter Economic Future, Washington, D.C.: The National Academies Press, 2007. âDomestic Policy Council, Office of Science and Technology Policy, American Competitiveness Ini- tiative: Leading the World in Innovation, Washington, D.C., 2006. Available at http://www.whitehouse. gov/stateoftheunion/2006/aci/aci06-booklet.pdf; last accessed September 17, 2007.
180 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 Grant Sizes Over the past 10 years, the median grant size has increased moderately. For NSF DMR, it rose from $84,000 to $112,000 (see Figure 10.4), and for DOE BES it rose from approximately $120,000 to $150,000 per year. However, costs associated with research at universities have outpaced inflation. As discussed above, the committeeâs survey of state and private universities indicates an average increase of 5 percent per year in the cost per graduate student. Median-size grants today often can only support one graduate student in full, plus the costs of materials and supplies and of small pieces of equipment. International Data The committee endeavored to obtain data to compare the overall funding trends for CMMP in the United States with those in other countries. The breadth and diversity of the CMMP enterprise and the different venues for funding CMMP- related activities in different countries made this difficult. However, funding for various nanoscience and nanotechnology initiatives abroad can provide a basis for an estimate of the increasing rate of CMMP funding in foreign countries. Fig- ure 10.6 provides such data for four Asian countries, indicating a wide variation in funding trends over the 5-year period, with China showing the largest percentage increase (more than a factor of two) and Japan showing the smallest increase (about 10 percent). The data are corrected for economic inflation within each country and normalized, to provide an estimate of recent increases in the âbuying powerâ related to total expenditures. A comparison of research output in terms of publications is described later in the chapter. Demographics of CMMP According to data compiled by the American Institute of Physics (AIP), after the rapid and historic increase in the number of U.S. physicists after World War II, the number of Ph.D. degrees peaked around 1970 (see Figure 10.7). The peaks and valleys in recent decades are influenced by many factors, including changing perceptions of job opportunities in physics, changes in the number of foreign students enrolling in U.S. graduate programs, and changes in the number of those receiving physics bachelorâs degrees in the United States and in the proportion of those who choose to continue with graduate study in physics. Economic and political changes also have an effect on degree production, along with available funding for physics research. The substantial 25 percent decline in Ph.D. produc- tion over the past decade has ended with a sharp increase in the number of degrees conferred in the class of 2005. Following the large gains seen in first-year Ph.D.
Structure and Level of the Current Research Effort 181 1.6 (as-spent domestic currency, corrected by inflation 1.5 1.4 1.3 Medical Research Funding rate, normalized to 2002) 1.2 1.1 1.0 0.9 0.8 0.7 0.6 0.5 0.4 1999 2000 2001 2002 2003 2004 2005 Japan China Taiwan Korea U.S. FIGURE 10.6â International trends in funding of materials science, normalized to 2002 with correction for inflation rates of each country (data provided by colleagues in each country). Since each country has a different way to categorize research, the research topics included in the figure vary, depending on the country. The committee was unable to locate data for Europe. 10-6 FIGURE 10.7â The number of physics doctorates conferred in the United States, 1900-2005. SOURCE: American Institute of Physics, Enrollments and Degrees Report, College Park, Md., 2006. Available at http://www.aip.org/statistics/trends/reports/ed.pdf.
182 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 student enrollments from 1998 to 2003, increases in physics Ph.D. production are expected in future years. Over the preceding decades, CMMP has been the largest subfield of physics in the United States and worldwide. The number of Ph.D. degrees issued in CMMP over the past 20 years is compared to other subfields in Figure 10.8. The decline in CMMP Ph.D. degrees from the mid-1990s to 2003 mirrors the decline in total phys- ics degrees over the same period. These data were obtained from the NSF through the Survey of Earned Doctorates, in which new doctoral recipients are asked to identify their primary field of dissertation research from a list of categories. This survey has about a 95 percent response rate from new doctoral recipients. In the 2004 survey, âApplied Physicsâ and âBiophysicsâ were introduced as new categories (the committee chose to include applied physics as part of CMMP, but to graph biophysics separately). Doctoral recipients in these areas prior to 2004 likely chose âPhysics, Generalâ or âPhysics, Otherâ as their field of dissertation research, thus de- 500 450 400 350 Number of Doctorates 300 250 200 150 100 50 0 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 Condensed-Matter and Materials Physics Atomic, Molecular, Optical Physics Elementary Particle Physics Nuclear Physics Plasma Physics Physics, General Physics, Other Biophysics FIGURE 10.8â Physics doctorates awarded by subfield, 1985-2005. Condensed-Matter and Materials Physics includes Solid-State/Low-Temperature Physics, Polymer Physics, Fluids Physics (category discontinued in 2004), and Applied Physics (new category in 2004). Biophysics is a new category as of 2004 and is shown separately here. SOURCES: National Science Foundation, Selected Data on Sci- ence and Engineering Doctorate Awards: 1994, NSF 95-337, Arlington, Va., 1995; and National Science Foundation, Science and Engineering Doctorate Awards: 2005, NSF 07-305, Arlington, Va., 2006. 10-8
Structure and Level of the Current Research Effort 183 creasing the CMMP count. Because of these variations in the categorization of dis- sertation research, the data in Figure 10.8 can only give an approximate indication and are likely an underestimate of the number of Ph.D. recipients in CMMP. Women and Underrepresented Minorities in CMMP Data from the American Institute of Physics Statistical Research Center show the number of degrees granted to women from 1978 to 2005 (see Figure 10.9). While the total number of women entering physics each year has reflected, to some extent, the ups and downs in total degrees granted, the dominant trend is a steady rise in the fraction of degrees granted to women, from about 6 percent of Ph.D. degrees in 1978 to about 14 percent of Ph.D. degrees in 2005. FIGURE 10.9â Percentage of bachelorâs degrees, masterâs degrees, and doctorates in physics earned by women, 1978-2005. NOTE: A form change occurred in 1994 resulting in a more accurate repre- sentation of women among holders of a bachelorâs degree in physics. Some of the increase in 1994 only may be a result of that change. SOURCE: American Institute of Physics, Enrollments and Degrees Report, College Park, Md., 2006. Available at http://www.aip.org/statistics/trends/reports/ed.pdf.
184 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 Data from the AIP Statistical Research Center (see Table 10.1) also show a sig- nificant increase between 2000 and 2004 in the number of African-ÂAmerican and Hispanic university faculty, particularly those working in Ph.D.-granting departÂ ments. Much of this increase is due to one institution, Florida Agricultural and Mechanical (A&M) University, a historically black university, which switched from granting the masterâs degree as its highest degree to granting Ph.D.âs. However, African-Americans, Hispanics, and women remain underrepresented compared with their representation in other disciplines. Figure 10.10 shows just how imbalanced the situation still is for women faculty. In 2006, approximately 40 percent of all physics departments in the United States count fewer than two women among their faculty. As the largest Ph.D.-producing physics subfield, CMMP has both an obliga- tion and an opportunity to improve this situation. Data compiled by NSF show a rise in the number of degrees granted to women in CMMP (see Figure 10.11; NSF physics subfield categories are discussed above in the section âDemographics of CMMPâ). The number of degrees granted to women roughly doubled over the period from 1986 to 2005, with a peak around 1996, which mirrors a peak in total CMMP doctoral degrees around the same time. Over the same period, the percent- age of CMMP Ph.D. degrees earned by women increased steadily from around 8 percent in 1986 to around 14 percent in 2005. This is a very encouraging trend, but the CMMP community must continue to encourage women to enter the field and continue to knock down remaining obstacles to their advancement. Several efforts within CMMP are actively increasing the participation by women and underrepresented minorities. In particular, the NSF Materials ÂResearch Science TABLE 10.1â Percentage of Physics Faculty Who Are African-American and Hispanic and Number of African-American and Hispanic Physics Faculty by Degree-Granting Department, 2004 and 2000 Highest Degree Granted by Percentage Department of Physics Faculty Ph.D. Masterâs Bachelorâs Number of departments Total number of physics departments in 2004 185 72 503 2004 Faculty African-American 2.0 64 29 78 Hispanic 2.7 107 56 60 2000 Faculty African-American 1.8 38 41 62 Hispanic 2.0 81 32 42 SOURCE: American Institute of Physics, 2004 Physics and Astronomy Academic Workforce Survey, College Park, Md., 2005. Available at http://www.aip.org/statistics/trends/reports/awf.pdf.
Structure and Level of the Current Research Effort 185 FIGURE 10.10â Percentage of Ph.D. physics departments, by number of women faculty in professo- rial ranks, 2006. SOURCE: American Institute of Physics, 2006 Physics and Astronomy Academic Workforce Survey, College Park, Md., 2007. and Engineering Centers are playing an important role in CMMP in attractÂing and retaining women and minorities. Averaged over all MRSECs in 2005, about 13Â per- cent of MRSEC faculty were women and 3 percent were from underÂrepresented minorities. At the level of graduate students participating in MRSECs, the numbers are 27 percent and 5 percent, respectively (for a detailed discussion, see the NRC MRSEC report). MRSECs are also helping to address the pipeline issue through their K-12 outreach programs and, at the college level, their Research Experience for Undergraduates programs. In addition, NSF DMR recently initiated the Partner- ships for Research and Education in Materials program that seeks to enhance diver- sity by stimulating long-term partnerships between Âminority-Âserving institutions and DMR-supported centers and facilities. Together, these efforts provide CMMP âNational Research Council, The National Science Foundationâs Materials Research Science and Engi- neering Center Program: Looking Back, Moving Forward, Washington, D.C.: The National Academies Press, 2007.
186 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 80 70 Number of Female Doctorates 60 50 40 30 20 10 0 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 Condensed-Matter and Materials Physics Atomic, Molecular, Optical Physics Elementary Particle Physics Nuclear Physics Plasma Physics Physics, General Physics, Other Biophysics FIGURE 10.11â Physics doctorates awarded to women, by subfield, 1985-2005. Condensed-ÂMatter and Materials Physics includes Solid-State/Low-Temperature Physics, Polymer Physics, Fluids Â Physics (category discontinued in 2004), and Applied Physics (new category in 2004). Biophysics is a new category as of 2004 and is shown separately here. SOURCES: National Science Foundation, S Â elected Data on Science and Engineering Doctorate Awards: 1994, NSF 95-337, Arlington, Va., 1995; 10-11 and Â National Science Foundation, Science and Engineering Doctorate Awards: 2005, NSF 07-305, A Â rlington, Va.,Â 2006. with a leading role in increasing the number of women and underÂrepresented minorities in physics. Doctoral Degrees in Physics by Citizenship In contrast to the ups and downs in total Ph.D. production discussed earlier, data compiled by the AIP Statistical Research Center show a steady decline in the proportion of physics Ph.D. degrees granted by U.S. universities to U.S. citizens over the period from 1966 to 2005 (see Figure 10.12). This decline has been com- pensated for by a steady rise in the proportion of degrees granted by U.S. universi- ties to citizens of other nations. In past decades, many of these advanced degree holders have remained in the United States to pursue employment opportunities. Now, with rapidly improving opportunities outside the United States, many are
Structure and Level of the Current Research Effort 187 FIGURE 10.12â Citizenship of individuals granted physics doctoral degrees by U.S. universities, 1966- 2005. SOURCE: American Institute of Physics, Enrollments and Degrees Report, College Park, Md., 2006. Available at http://www.aip.org/statistics/trends/reports/ed.pdf. Foreign citizens include individu- als with permanent residence and those with temporary visas. taking their knowledge and creative energies elsewhere. This is, no doubt, a posi- tive development for the world, but some have predicted undesirable consequences for the U.S. economy as high-paying research and development (R&D) jobs move to other countries and the United States loses clear leadership in various fields of science. However, this downward trend in Ph.D. degrees granted to U.S. citizens is expected to reverse soon. According to the AIP (see Figure 10.13), the number of U.S. citizens enrolling as first-year graduate students was up nearly 50 percent from the recent low in 1998. It should therefore follow that the number of Ph.D. degrees awarded to U.S. citizens in coming years will grow sharply. Degree production in the subfield of CMMP is expected to follow the same course. Publication Trends The Committee on CMMP 2010 considered the number of articles published in scholarly journals as a good proxy for scientific productivity. The committee obtained data for the numbers of articles published by and submitted to several journals of the CMMP-relevant American Physical Society and the American Institute of Physics. Comparisons between the United States and other countries and between academia, industry, and national laboratories can be made with these data. The committeeâs most important observations and conclusions are described below.
188 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 FIGURE 10.13â First-year U.S. and foreign graduate physics students, 1964-2005. NOTE: A change in wording on the 2000 questionnaire resulted in more accurate data on first-year graduate students. This change was responsible for 3 percent of the reported 8 percent increase in total first-year students between 1999 and 2000. SOURCE: American Institute of Physics, Enrollments and Degrees Report, College Park, Md., 2006. Available at http://www.aip.org/statistics/trends/reports/ed.pdf. Figure 10.14 compares the combined numbers of articles originating in the United States with those coming from the rest of the world. The committee bases this comparison on two journals that capture the core disciplines within CMMP: Physical Review B, whose subject areas include hard condensed-matter and Âmaterials physics, and Physical Review E, which covers statistical, nonlinear, and soft-Âmatter physics. While the number of publications contributed from the United States has remained essentially flat for the past 13 years, publications from the rest of the world nearly doubled. The United States has lost ground to both Western Europe and Asia during this period, as shown in Figure 10.14. Articles originating in the United States constituted the largest component at the beginning of the period, but the United States was overtaken by Western Europe in about the middle of the past decade. The committee emphasizes that the publication trends are recent (since 1993) and occur long after the reestablishment of Western Europe after World War II. If the current trends continue, the United States will also be overtaken within the next decade by Asia, consisting of primarily Japan and China, and to a lesser extent, South Korea. In fact, articles submitted to Physical Review B and Physical
Structure and Level of the Current Research Effort 189 4000 3500 Articles in Two Journals (total number) 3000 2500 2000 1500 1000 500 0 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 Western Europe United States Asia (including India) Rest of World FIGURE 10.14â Comparison among the United States and other regions of the world in the total number of articles published in Physical Review B plus Physical Review E for the past 13 years (1993- 2005). SOURCE: Publication data supplied to the Committee on CMMP 2010. 10-14 Review E by scientists from Asia already exceed those from the United States, but the article rejection rate for Asia is currently larger than that for the United States. The committee acknowledges that the growth of CMMP research in Asia in the next decade is indeed expected because of the growths of Asian nationsâ economies. A general trend across all the journals surveyed, however, is that the rejection-rate gap between the United States and the rest of the world has narrowed, and continues to narrow, presumably reflecting a relative improvement in scientific quality of the rest of the world relative to the United States, although the quality of the U.S. publications remains high as judged by publication citations, as discussed below. Publication data can also expose shifts in activity at industrial and national laboratories. Figure 10.15 is a plot of the numbers of publications by several no- table U.S. industrial laboratories in Physical Review B. The data reveal a precipitous drop in publications between 1996 and 2001 by the two major industrial research laboratories, Bell Laboratories and the IBM Research Centers, indicating a dramatic drop in CMMP basic research being performed at these institutions. Bell Labo- ratories, in particular, experienced an eight-fold decrease in publications in just 5 years, followed by a slight increase in output for the past 5-year period. Data for the national laboratories do not exhibit any particularly significant trends.
190 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 90 80 70 Articles Published (no.) 60 50 40 30 20 10 0 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 Total No. of Articles/100 Bell Labs IBM Hewlett-Packard DuPont FIGURE 10.15â Numbers of articles published in Physical Review B by major industrial laboratories in the past 10 years (1996-2005), compared to the total number of articles (divided by 100 to fit on the same graph) published in the same journal. 10-15 In addition to looking at the geographical distribution of CMMP publications in leading U.S. journals, the committee also tried to assess the quality of the U.S. output. Here the picture is rather more encouraging. For example, although the U.S. share of the overall output in journals such as Physical Review B and Physical Review E has fallen in absolute terms during the past decade, the fraction of U.S. papers in the top 100 cited publications has remained approximately constant. The committee concludes from this observation that the quality of work in the nationâs leading CMMP research groups remains world-class, even if the total output is flat. There is no room for complacency, however, as it is also clear that the impact of publications from emerging nations such as China is growing rapidly. CMMP publications are one source of evidence that individual investigators and small numbers of collaborators dominate advances in the field. Furthermore, CMMP research at large facilities also tends to be done by individual groups or by small groups of collaborators. CMMP researchers view large facilities as a means to do individual or small-group research, and the publications from using large facilities still have a small number of authors.
Structure and Level of the Current Research Effort 191 Recommendations The Committee on CMMP 2010 bases the following recommendations on its assessment regarding the most efficient use of resources and projected growth in funding for the field of 7 percent per year over the next 10 years. This rate of growth reflects levels recommended in the presidentâs American Competitiveness Initiative, which seeks a doubling of the physical science research budget of NSF, DOE, and the National Institute of Standards and Technology in 10 years. In some cases, improved research quality and efficiency can be obtained through changes in the structure of funding. In other cases, additional funding is necessary in order to retain current expertise and to nurture emerging fields of science. Some new facilities are also required to advance science and to keep the U.S. research effort at the forefront. What is critical for success in the coming decade is strong support for research at the single- and few-investigator level. Centers and national facilities provide the infrastructure, but they will be useless without a strong community of scientists pushing the science and technology forward. The field of CMMP is large, heterogeneous, and complex. It progresses through the integration of a large number of small breakthroughs rather than by a few high- profile achievements. Unlike some other fields of science, these breakthroughs are predominately made by small teams following their particular visions rather than by large groups pursuing a limited number of specific, clearly defined goals. This means that strong support at the individual-investigator level is essential. The pro- posed doubling of the physical sciences budget, with particular focus on supporting individual investigators (by increasing both average grant size and the number of awards), would be an excellent first step toward accomplishing this goal. As described above, the funding situation in CMMP has made it difficult for researchers to focus on their research programs. The reduced success rates for proposals, the reduction in buying power of grants, and the increasing number of submitted proposals have changed dramatically over the past 10 years. The pros- pects are particularly dim for young scientists and for those entering for the first time a field in which funding success rates are approximately 10 percent. It is also difficult for researchers involved in interdisciplinary research to obtain support. Finding the correct funding agency and program for work that lies between phys- ics and engineering or physics and chemistry or biology has been difficult. The recommendations below to the National Science Foundation, the Department of âDomestic Policy Council, Office of Science and Technology Policy, American Competitiveness Ini- tiative: Leading the World in Innovation, Washington, D.C., 2006. Available at http://www.whitehouse. gov/stateoftheunion/2006/aci/aci06-booklet.pdf; last accessed September 17, 2007.
192 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 Energy, and the CMMP community are aimed at maintaining U.S. innovation in CMMP research. Recommendation: Strong support should be maintained for individual and small groups of investigators, which are historically the primary source of in- novation in CMMP. The ratio of support for individual and small groups of investigators relative to support for centers and facilities should not decline in the next decade. Recommendation:â The average success rates for the funding of proposals should be increased to more than 30 percent over the next 5 years in order to give junior scientists the opportunity to obtain research results before the ten- ure decision and to enable currently funded researchers to maintain continuity in their research programs. Recommendation:â The size of grants to individual and small groups of in- vestigators should be increased to maintain the buying power of the average grant and to retain scientific talent in the United States. Recommendation:â The CMMP community should work to improve the rep- resentation of women and underrepresented minorities in CMMP through mentoring; providing flexible working conditions, day-care opportunities, and viable career paths; and developing outreach programs targeted to students and the public and aimed at increasing the numbers of prospective researchers. Recommendation:â The Faculty Early Career Development awards and other research grant awards should be made on the basis of research criteria, not education and outreach. Providing financial incentive for education and out- reach activities would more effectively engage the community and encour- age the best programs. The committee recommends a separate program for funding outreach and education (see Chapter 8 for further discussion of this recommendation).