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 193
11 Tools, Instrumentation, and Facilities for Condensed-Matter and Materials Physics Research The quest to observe, predict, and control the arrangements and motions of the particles that constitute condensed-matter systems is central to the condensed- matter and materials physics (CMMP) enterprise. The constituent particles span an enormous range of sizes—from electrons and atoms to macromolecules—and their motions span a correspondingly immense range of timescales. As a result, the experimental, computational, and theoretical tools required to study them are extremely diverse. Many of these tools are developed by individual research groups; other tools, such as synchrotron x-ray and neutron scattering, are developed at large-scale national laboratory facilities. Technical innovations that extend the limits of measurement and prediction lie at the forefront of CMMP research. For example, scanning probe microscopes were developed to image surfaces at scales too small to be resolved by ordinary optical microscopy, and they immediately transformed the fundamental understanding of surface science. Moreover, the benefits of new techniques often stretch far beyond condensed-matter physics; scanning probe microscopes have now evolved into universal tools at the nanoscale for the physical and life sciences. Experimental condensed-matter tools underlie many noninvasive medical diagnostics, while theoretical and computational tools from CMMP, such as local electron density approximations and numerical simu- lation methods, are now used by pharmaceutical companies. The past decade has seen the advent of promising techniques, such as coherent and pulsed x-rays, novel optics based on exotic materials, multiscale modeling, and topological approaches to the study of magnetic and superconducting materials. As CMMP researchers seek to answer fundamental questions about materials, they will continue to design 

OCR for page 193
c o n d e n s e d - m at t e r m at e r i a l s P h ys i c s  and tools, or adapt tools for new applications, that will benefit CMMP, other scientific disciplines, and society at large. TOOLS AND INSTRUMENTATION FOR CMMP RESEARCH Measurement techniques designed to probe the properties of matter at smaller length, time, or energy scales or with greater quantitative resolution and sensitivity advance the forefront of condensed-matter and materials physics research. Likewise, techniques designed to synthesize high-quality materials with precisely controlled structures underpin many great CMMP discoveries. By pushing the boundaries of materials fabrication and measurement forward, experimental CMMP researchers have uncovered new phenomena that were often unanticipated. These discoveries have not only transformed CMMP, but they themselves have led in turn to new ways to manipulate and image matter, crucial to many new technological advances with a broad range of applications. New computational and theoretical techniques that push forward the boundar- ies of prediction also play a prominent role in advancing CMMP. To some extent, theory and computation are interlinked—theory nearly always forms the basis for new approximations or algorithms that substantially increase the efficiency of computations. Conversely, numerical computation is often indispensable in theory. Theoretical innovations, such as the application of field theories to condensed- matter systems and linear response theory have not only allowed researchers to tackle previously intractable problems, but, like many of the greatest experimental and computational techniques, have also changed the landscape of CMMP by revealing unexpected phenomena or deep, previously hidden connections among phenomena. As discussed later, computation can dramatically amplify the power of analytical tools. Indeed, the first electronic digital computer itself was built in order to carry out theoretical CMMP calculations. The research community is at the brink of an era in which powerful computer simulations will be integrated into measurement tools, enabling the extraction of information in unprecedented detail from measured quantities. Simulations will extend the reach of analytical theoretical techniques, connecting conceptual devel- opments to experimental measurements. The results will guide researchers through the realms of materials possibilities so vastly expanded by the ability to control the structure of the material at the nanoscale. Closing the loop, new detectors and de- vices will be made possible by new, purposefully designed, functional materials to further increase the power of measurements. Some of these breakthroughs will be made in single-investigator laboratories and, following the example of the scanning tunneling microscope, will turn into commodity instruments. Other advances will rely on the unique powers of staggeringly expensive large-scale instruments and teams of experts supported by large national facilities; these tools will need to be

OCR for page 193
t o o l s , i n s t ru m e n tat i o n , facilities cmmP research  and for broadly available and “user friendly” for a wide cross section of researchers. Thus, all agencies supporting CMMP research should provide strong instrumentation programs to enable the research work in the CMMP field to be carried out effi- ciently and well. There is a need for keeping the infrastructure supporting CMMP research at universities up to date and for providing modern instruments for the training of the next generation of researchers. Instrumentation in CMMP Research CMMP researchers have a track record of developing new measurement tools that have enabled advances not only within CMMP, but also in other areas that encompass the physical, chemical, biological, and medical sciences. Indeed, the continued development of techniques with sufficient spatial resolution and sen- sitivity to measure structure, composition, and properties of condensed-matter over various length scales (from nano to macro), dimensionalities, and timescales is essential. During the past decade there have been significant advances in the use of tools in imaging, scattering, and spectroscopy. In this section, the Committee on CMMP 2010 briefly highlights advances in some of these areas. Imaging Techniques Imaging techniques provide structural images, direct and indirect, and prop- erty maps. Microscopy alone and microscopy combined with tomographic tech- niques are the most commonly used techniques to create images in two and three dimensions, respectively. Recent developments in x-ray microscopy, based largely on improvements of the fabrication of the optics, have enabled the observation of molecular length-scale height variations on a surface. Image measurements are now accomplished over an area of many microns, with a resolution of 200 nanometers (nm) and a step height of 0.6 nm with this technique. X-ray imaging has also enabled imaging at greater depths within a sample than is possible with electrons. With third-generation synchrotron sources, it is possible to study opaque objects using hard x-rays, while soft x-rays are used for soft materials and to probe near the surface of materials. X-ray beams can be focused to dimensions on the order of 100 nm to enable better resolution; the limiting resolution with x-rays is yet to be reached. A new development involves imaging by neutrons. Imaging is based on the notion that neutrons are characterized by a de Broglie wave packet with a spatial distribution that is sufficiently large to permit interference, very much in the same way as light. With the development of appropriate “optics,” including transmission gratings based on differences in neutron-capture cross sections and incoherent scattering cross sections, two-dimensional images of various materials can be

OCR for page 193
c o n d e n s e d - m at t e r m at e r i a l s P h ys i c s  and created. In fact, a three-dimensional image, based on the scattering length density distribution, can be reconstructed. While scanning probe techniques have become ubiquitous in CMMP and have had an enormous impact on the understanding of materials, particularly at the nanoscale, many challenges remain. The inability to scan large areas of samples rapidly and issues related to thermal drift remain to be solved. Other issues await- ing resolution relate to the interpreting of data that are influenced by interactions between the cantilever tip and the sample. In the future, further progress in in- strumentation and data analysis, smaller cantilevers together with better deflection sensors, and improved sample-preparation techniques will lead to greater sensitiv- ity and resolution. Multifunctional cantilevers, wherein a local “field” is applied while simultaneously probing the local response of the system, are part of a future strategy to ensure the increased impact of these techniques in the understanding of nanoscale properties. More sophisticated detection systems to enhance sensitivity further and improved computer algorithms for data interpretation and analysis will increase the utility and wide accessibility of these techniques that are so essential to modern CMMP materials characterization. Scattering Techniques Scattering is also used to provide information about the structure and dy- namics of condensed matter. For example, diffraction techniques, the best known of the scattering processes, provide information about the long-range order of a sample, with tenths-of-a-nanometer resolution, and the use of x-rays, combined with information gleaned from neutron measurements, has led to a better under- standing of the crystallography of complex macromolecules. Neutron scattering has grown in recent years as a regular tool for the characterization of samples. Neutron scattering techniques provide information about dynamics from 10–12 seconds to seconds and structure at length scales from 0.1 nm to 103 nm. Neutrons convey information about interatomic forces on the basis of measurement of the energy of the scattered neutrons. The incident intensity (flux) of neutrons and the efficiency with which the scattered neutrons are detected are key factors that determine the performance of a neutron source. The third-generation neutron sources (for example, the Spallation Neutron Source [SNS] at Oak Ridge National Laboratory) provide large increases in sensitivity that will result in better speed in data acquisition (seconds or minutes versus hours or a day, depending on the system and the information collected) and will enable measurements of lower concentrations of a given species. The latter is important because it will enable the analysis of multicomponent systems. Other types of scattering techniques provide direct spatial depth profiling in- formation of materials, including Rutherford backscattering spectrometry, forward

OCR for page 193
t o o l s , i n s t ru m e n tat i o n , facilities cmmP research  and for recoil spectrometry, nuclear reaction analysis, and secondary ion mass spectrom- etry. These techniques involve the use of incident energetic ion beams; analysis of species emanating from the sample provides information about the concentration of a given species as a function of depth. The effective use of combinations of neutron and x-ray scattering with ion-beam techniques can provide more detailed information about the structure and dynamics of nanocomposites, heterostruc- tures, and complex liquids at smaller length scales. Spectroscopy Techniques The use of spectroscopy techniques for imaging has grown rapidly in re- cent years. These techniques, which have been very successful in the imaging of soft materials, including biological materials and polymers, have been invaluable. New developments that involve the use of scanning force probes, scanning force magnetic resonance (a sample is placed on a cantilever in the presence of a small ferroelectric tip which creates an inhomogeneous field that has the effect of polar- izing the spins in the sample) have enabled the three-dimensional imaging of an individual atom as well as single spins. Infrared and Raman techniques have been used to image samples based on a vibrational signature associated with a molecule. Researchers have been suc- cessful in using the Raman effect, inelastically scattered light that is shifted in wavelength relative to the incident wavelength, to improve the sensitivity of the identity of certain molecules within a sample. Surface-enhanced Raman scatter- ing has laid the foundation for the development of surface-enhanced spectros- copies that include surface-enhanced fluorescence and surface-enhanced infrared spectroscopy. The latest developments include single-molecule surface-enhanced Raman spectroscopy and tip-enhanced scanning near-field optical microprobe Raman spectroscopy. Simultaneous Measurement Capabilities New strategies that involve in situ characterization of materials using x-rays or neutrons are becoming routine. Specifically, some research groups use x-rays or neutrons to measure the properties of materials (dynamics structure, phase transi- tions, and so on) that are simultaneously subjected to external perturbations (stress, temperature, and various kinds of fields). With the use of tomographic techniques, information about the interior of samples can now be learned without the need to section them destructively for analysis with transmission electron microscopy (TEM) or scanning probe techniques. The availability of these combined techniques enables increased spatial and temporal resolution and rapid data acquisition. In some cases the duration of measurements could be reduced from tens of hours to minutes.

OCR for page 193
c o n d e n s e d - m at t e r m at e r i a l s P h ys i c s  and Another significant advance is the use of scanning force techniques in con- junction with other techniques for learning about properties, such as electrical conductivity and magnetism, in unprecedented detail at the nanoscale. Instruments are approaching the stage at which the resolution of joint probes is comparable to atomic force microscope measurements of topography. In situ TEM capabilities are also being developed to enable the direct observation of changes in atomic ar- rangement (structure) of a material while it simultaneously experiences external perturbations owing to changes in temperature, mechanical stresses, or electric fields. This is a powerful technique that is currently exploited by a number of elec- tron microscopists. The development of more sophisticated theory and multiscale algorithms that enable better use of experimental data to characterize samples will be a continuing challenge. Computation in CMMP Research As the materials and phenomena of interest have become increasingly com- plex, computation has emerged as an essential tool in the process of interpreting experimental data and analyzing theoretical models. Over the past decade or two, computational CMMP has developed fully into a branch of study in its own right, on a par with experimental and theoretical CMMP. From the beginning, computational CMMP has not only allowed researchers to confront previously insoluble problems but has also provided a means to discover new phenomena. There are two paradigms in computational CMMP. One extends the power of theoretical modeling by numerical solution of “simple” models, both classical and quantum, which capture the essential physics of the system of interest. These results are used either directly in interpreting and predicting experimentally observed phenomena or as an aid to “pencil and paper” analysis by developing and validating approximations to study models that cannot be solved exactly. The sec- ond paradigm is the direct solution of the quantum mechanical equations to make quantitative predictions about the behavior of particular materials at the atomic scale. In both, breakthroughs in the development of theory and algorithms, aided by enormous increases in computer speed and memory, have enabled dramatic progress in the past decade. Techniques developed for the numerical investigation of simple models have had widespread applicability beyond CMMP. Monte Carlo methods are now stan- dard tools in all fields of science and engineering and are even used in industrial contexts. Some recent approaches that have promise for significant impact in CMMP are phase retrieval methods and new forms of Monte Carlo algorithms, including ones that can evolve dynamically. New field theoretical algorithms are having increasing impact. With the density matrix renormalization group (DMRG) method, significant progress has been made toward eliminating the “sign problem”

OCR for page 193
t o o l s , i n s t ru m e n tat i o n , facilities cmmP research  and for bottleneck, ubiquitous in numerical studies of systems of interacting electrons; this method has also had a significant impact in quantum chemistry, quantum information theory, and nuclear and high-energy physics. With these methods, it is now possible quantitatively to study models that cap- ture aspects of the essential physics of materials with strongly correlated electrons, such as complex oxides, including the high-temperature superconducting cuprates and magnetoresistive manganites, and two-dimensional electron gases. Figure 11.1 shows an example of the rich variety of ground-state orderings that have been observed in models of the cuprates with the DMRG method. Methods for the direct solution of the underlying quantum mechanical equa- tions allow quantitative, material-specific, first-principles prediction of structure and properties. The ongoing development of efficient algorithms allows the study of ever-more-complex structures, including crystals with very large unit cells, and nanostructured systems. New classes of algorithms and the incorporation of many-body physics allow the extension of these methods to a broader range of materials—notably, correlated electron systems with magnetic, orbital, and FIGURE 11.1 Plots showing the stripe ordering of the charge and spin on a two-dimensional CuO 2 plane in the high-temperature superconductor (La,Nd,Sr)CuO4. (Left) As suggested by neutron- scattering experiments. (Right) As calculated for the t-J model using the density matrix renormaliza- tion group method. SOURCE: Reprinted with permission from S.R. White and D.J. Scalapino, “Density Matrix Renormalization Group Study of the Striped Phase in the 2D t-J Model,” Phys. Rev. Lett. 80, 1272 (1998). Copyright 1998 by the American Physical Society.

OCR for page 193
c o n d e n s e d - m at t e r m at e r i a l s P h ys i c s 00 and charge ordering, and systems under ultrahigh pressure relevant to geophysics. New capabilities are being developed to study systems in applied electric and magnetic fields and to extend computations for excited states. Such computational capability is necessary for predicting optical and transport properties. An example is shown in Figure 11.2. The direct solution of the underlying quantum mechanical equations also plays a key enabling role in the design of new materials. In this work, the target is particular structures and properties, requiring the solution of the “inverse problem” to find a corresponding material. In an experimental framework, combinatorial solid-state methods survey the structure and properties for entire compositional ranges for complex solids containing three or more different elements. Similarly, computational methods for the prediction of structure and properties of solids now are accurate and fast enough to allow first-principles materials design, in which the FIGURE 11.2 The current induced by varying voltages across carbon chains of varying lengths sandwiched between gold and aluminum leads can be computed using first-principles methods. The carbon nanowires differ from conventional wires in that the current is not proportional to the voltage. SOURCE: J.B. Neaton, K.H. Khoo, C.D. Spataru, and S.G. Louie, “Electron Transport and Optical Prop- erties of Carbon Nanostructures from First Principles,” Comput. Phys. Commun. 169, 1-8 (2005).

OCR for page 193
t o o l s , i n s t ru m e n tat i o n , facilities cmmP research 0 and for FIGURE 11.3 The structure of lithium nickel manganese oxide, a promising new battery material designed using computational methods, consists of layers of transition metal (nickel and manganese, blue layer) separated from lithium layers (green) by oxygen (red). SOURCE: K. Kang, Y.S. Meng, J. Bréger, C.P. Grey, and G. Ceder, “Electrodes with High Power and High Capacity for Rechargeable Lithium Batteries,” Science 311, 977 (2006). Reprinted with permission from the American Associa- tion for the Advancement of Science. structural parameters and selected properties for large sets of real and hypothetical structures can be surveyed to identify interesting materials for new physics and applications, including room-temperature ferromagnetic semiconductors for spin- tronics1 and new battery materials2 (see Figure 11.3). By using the first-principles calculations as input into parameterizations of the composition space, searches can be extended to even larger spaces of materials. Similar principles can be applied to the design of heterogeneous materials and devices. Much of the important physics in materials systems takes place at length scales well beyond which fully first-principles methods are practical. This range is extended by molecular dynamics simulations with parameterized interatomic potentials. The associated loss in accuracy at the atomic level is compensated by the ability to use tremendously larger numbers of atoms (100 million or more) and the ability to study the time evolution of phenomena. One area in which such computations have proved particularly valuable is in the study of mechanical prop- erties, such as strength of materials, plastic deformation, fracture, and friction, in 1 S.C. Erwin and I. Zutic, “Tailoring Ferromagnetic Chalcopyrites,” Nat. Mater. 3, 410 (2004). 2 C.C. Fischer, K.J. Tibbetts, D. Morgan, and G. Ceder, “Predicting Crystal Structure by Merging Data Mining with Quantum Mechanics,” Nat. Mater. 5, 641 (2006).

OCR for page 193
c o n d e n s e d - m at t e r m at e r i a l s P h ys i c s 0 and which the behavior is determined by line and planar defects that are created and propagated through the system. An example of such calculational results is given in Figure 11.4. As impressive as such simulations are, they are still far short of macroscopic scales: a solid cube 1 micron on a side contains over a thousand times more atoms. Within the past decade, priority has been given to developing truly multiscale methods for modeling materials properties, with seamless integration of atomic scale, intermediate length scale, and continuum methods. Similar multiscale ap- proaches are needed to treat materials physics involving time evolution (dynam- ics) on a wide range of timescales. While great progress has been made, additional breakthroughs are needed. As new measurement tools are developed, computational approaches will be essential to interpreting larger amounts of data and extracting subtle signals and correlations. Simulations can be invaluable in separating artifacts from intrinsic behavior. Both in the numerical study of simple models and in first-principles FIGURE 11.4 Snapshot from a molecular-dynamics simulation showing the behavior of nanocrystal- line aluminum during deformation. The crystal grain at the center is 70 nm in diameter and is defined by clear grain boundaries (blue atoms). Deformation is seen to drive the formation of planar defects (red atoms) that start at the grain boundary and grow into the grain’s interior. SOURCE: V. Yamakov, D. Wolf, S.R. Phillpot, A.K. Mukherjee, H. Gleiter, “Dislocation Processes in the Deformation of Nano- crystalline Aluminium by Molecular-Dynamics Simulation,” Nat. Mater. 1, 45 (2002).

OCR for page 193
t o o l s , i n s t ru m e n tat i o n , facilities cmmP research 0 and for simulations, novel unanticipated physical behavior can arise, giving new insights and guiding experimental investigation. Fundamental issues need to be addressed in the next decade to build on this progress. More efficient, accurate, and broadly applicable methods must be de- veloped to study dynamics, effects of thermal fluctuations, and excited states of materials. There are many promising avenues for further progress in techniques for studying systems with strong correlations. Particular attention should be paid to improving the conceptual and algorithmic framework for studying energy trans- formation in solids, such as in electromagnetic radiation, energetic particles, and heat generation. Efforts should continue to be made to increase the efficiency of algorithms by drawing on forefront research in numerical methods and computer science. New approaches to multiscale methods for spatial and temporal variations should be pursued. A concerted effort should be made to integrate simulations into experimental data analysis and help with the proper interpretation of the experimental measurements to increase the power of the developing experimental probes described in this chapter. Lastly, the push to integrate simulations into new materials design should intensify, with work continuing in parallel both on realiza- tions for particular systems and on the development of broadly applicable tools based on knowledge gained from these collaborations. CENTERS AND FACILITIES IN CMMP RESEARCH Both the complexity of scientific challenges and the resources required to con- duct a successful CMMP research program have increased in recent years. A major scientific challenge to the field is how to synthesize or fabricate materials in which the electronic, atomic, and molecular organization varies spatially, and in some systems, temporally. A related challenge is how to understand principles, or rules, that govern the behavior of materials over different length scales and timescales. To address these challenges, sophisticated tools (experimental, computational, and theoretical) are needed to probe the structure and properties of materials over a wide range of length scales and timescales. For synthesis, fabrication tools such as focused ion beams, molecular beam epitaxy, and lithography have become prohibi- tively expensive for operation by a single principal investigator (PI). Measurement tools to probe structure and properties are also very expensive, with centers and facilities addressing many of these needs. The associated requirements to educate students on how to perform experiments using new techniques and facilities are a pressing and constantly evolving need. In this section, the committee describes the current status of the research infrastructure and its ability to address, for example, the six CMMP challenges introduced in Chapter 1 and explicitly discussed in this report. The status and role of centers and mid- and large-scale facilities in rela- tion to single and small-group principal investigators are discussed. This chapter

OCR for page 193
c o n d e n s e d - m at t e r m at e r i a l s P h ys i c s  and Recommendation: Hold crosscutting workshops. There are major opportunities to reach out and connect with communities that use other, yet related techniques to image nanoscale phenomena, such as atom-probe and ion microscopes for three-dimensional imaging at the atomic scale, and x-ray nanoprobes. All of these communities, from electron microscopy to x-ray nanoprobe, are now gearing up to study similar materials problems and will face many similar scientific as well as technique-related challenges; yet the communities operate in parallel. Interdisciplinary, broadly based, and forward-looking workshops to address such common issues should be encouraged. High-Magnetic-Field Facilities Magnetic fields interact with moving charges. Because the typical length scales associated with this interaction scale decrease with increasing magnetic-field strength, high magnetic fields can probe small spatial features and the associated fast processes. In order to achieve magnetic lengths comparable to the size of a quantum dot of 6-nanometer diameter, fields of about 20 tesla (T) are required; 80 T are necessary to shrink this length by another factor of two. As a consequence, the study of magnetic phenomena on the scale of a few nanometers, and from there on down to atomic dimensions, necessitates pushing the limits of what is possible with current magnet technology. Traditional areas of success for high-field research have been the study of fun- damental mechanisms in correlated quantum systems such as low-dimensional magnetism, the quantum Hall effect, and superconductivity, as well as the investiga- tion of the properties of interacting magnetic flux bundles (“vortex matter”) inside superconductors. Separately, high-field research has enabled magnetic resonance studies in organic materials, providing important insights into membrane protein structures, hemoglobin, and the underpinnings of photosynthesis. Furthermore, CMMP research provides advanced materials, including superconductors with better performance, special conductors, and high-strength alloys. These materials form the critical components for magnets used in applications ranging from atomic particle accelerators to medical magnetic resonance imaging (MRI). Two recent studies have looked into the current status and the potential for future developments of high-field magnet research. For more detailed informa- tion and discussions of the various technical issues, the committee refers to these reports.8,9 8 National Research Council, Opportunities in High Magnetic Field Science, Washington, D.C.: The National Academies Press, 2005. 9 Report of the International Union of Pure and Applied Physics working group on Facilities for Condensed Matter Physics: High Magnetic Fields, 2004. Available at http://www.iupap.org/wg/fcmp/ hmff/highmagneticreport.pdf; last accessed September 17, 2007.

OCR for page 193
t o o l s , i n s t ru m e n tat i o n , facilities cmmP research  and for Current Status of High-Magnetic-Field Facilities Magnet facilities fall into two categories, delineated by magnet technology and, thus, by the maximum achievable magnet field strength. Smaller high-field mag- nets (<20 T) are currently based on technology using superconducting niobium compounds and are available commercially. These magnets are found in single PI laboratories as well as in local multiuser facilities. Costs rise steeply with increas- ing field. Niobium-titanium magnets deliver up to 11 T, while Nb3Sn goes up to 20 T in driven magnets (at a cost of $1 million to $2 million) and up to ~22 T in persistent-mode NMR magnets (at a cost of $5 million to $15 million per system). For these smaller magnet systems there have been no major technological advances in recent years. Large magnets (>30 T) comprise both continuous-field (direct current [dc]) and pulsed systems, are typically unique in design, and, because of their complexity and costs, are mostly located at dedicated high-field facilities, such as the National High Magnetic Field Laboratory (NHMFL) in the United States. At U.S. national facilities, large magnets are currently available that can reach up to 45 T in con- tinuous mode (hybrid superconducting/resistive magnets), and up to 60 T for 100 microseconds in pulsed mode. As pointed out in the reports mentioned above, the value of the maximum achievable field is not the only important parameter for high-field research. Depending on the application, the quality and usefulness of a facility are determined also by factors such as the homogeneity of the field, the diameter of the magnet bore, or the availability of an environment amenable to low-noise measurements. Furthermore, for much of CMMP research, another important factor is the simultaneous access to low sample temperatures, that is, a large ratio of magnetic-field strength to temperature. In this area, the NHMFL has been a leader with its High B/T Facility. With high-field magnet user facilities come challenges of energy costs in the face of increasing magnet hours driven by user demand. This challenge motivates higher-efficiency magnets, but they involve larger capital investment. As each magnet technology becomes more broadly used (for example, resistive magnets for nuclear and electron resonance), the issues shift toward addressing and integrating different magnet specifications (for example, peak field, time at fixed field, and field homogeneity) desired by the CMMP, chemistry, and biology user communities. Medium-Term Developments High-magnetic-field research in CMMP is driven by the prospect of using the field as an exquisitely sensitive tuning parameter to explore emergent quantum phases of matter and by being able to perform precision spectroscopy using tech- niques such as NMR. In the area of complex fluids, the same spectroscopic methods can be used to track trace elements, while quadrupolar NMR opens up almost the

OCR for page 193
c o n d e n s e d - m at t e r m at e r i a l s P h ys i c s 0 and entire periodic table as candidate nuclei. This capability will allow for a new level of structure-function correlation in glasses, ceramics, catalysts, and porous materials (for example, zeolites and batteries). Technological challenges for the coming decade center on the development of new magnet technology beyond niobium. The recent report Opportunities in High Magnetic Field Science10 identified a 30-T high-resolution NMR magnet, a 60-T dc hybrid magnet, and a 100-T long-pulse magnet as grand challenges. All of these require conductor materials in forms that are not yet commercially available, which in itself poses a materials research and development challenge. The NHMFL has been taking the lead in meeting these challenges and, furthermore, has embarked on developing additional magnet systems for low power consumption, complex fluids research, and ultrahigh fields (200 T/1 microsecond pulsed magnet). New superconducting materials, such as MgB2 or high-Tc materials such as yttrium barium copper oxide (YBCO) or bismuth strontium calcium copper oxide (BSCCO), offer several advantages in terms of larger upper critical field strength and higher operating temperatures (eliminating the need to cool with cryogens such as liquid helium). Mulifilament MgB2-based technology currently can go as high as 10 T, but 30 T or more appear eventually achievable. Commercial magnets based on this new technology are around the corner, with MRI applications as a major driver. Bi-2212 magnet wires promise greater than 50-T fields, among other advantages, while YBCO offers the highest fields. However, there are still many research and development challenges in terms of fabricating sufficiently long wires or tapes out of these materials and in improving their tensile strength, as required to withstand the forces generated in high-field magnets. Successful development of these materials could lead to relatively low cost and easy-to-operate magnets and would broaden the accessibility of high fields to small groups. Special pulsed and hybrid magnets also will benefit from the integration of high-Tc components. Resistive plus high-Tc technology should get well beyond 50 T. For pulsed magnets, multishot 100-T fields are within reach. An important direction besides magnet development will be the integration of high fields with beam lines. This would allow the investigation of the neutron and x-ray scattering properties of materials in high magnetic fields. Currently, there are interesting design proposals to add hybrid magnets of fields of about 30 T to beam lines at the SNS at ORNL, and at the APS at ANL. Another plan, involving a collaboration of NHMFL, Jefferson Laboratory, and the University of California at Santa Barbara, envisions combining advanced magnet technology with an infrared free-electron laser. This would allow access to the terahertz regime that is resonant with magnetic-field energy scales. 10 National Research Council, Opportunities in High Magnetic Field Science, Washington, D.C.: The National Academies Press, 2005.

OCR for page 193
t o o l s , i n s t ru m e n tat i o n , facilities cmmP research  and for Recommendations for High-Magnetic-Field Facilities in CMMP Research Recommendation: NSF should continue the support of the National High Magnetic Field Laboratory and high-magnetic-field instrumentation develop- ment following the priorities recommended by the recent National Research Council report Opportunities in High Magnetic Field Science.11 Recommendation: The research community, with support from the federal agencies, should exploit the opportunities for superconducting magnets pro- vided by the recent and imminent high-Tc conductor forms. This will benefit small-scale users and high-field NMR users, and will allow for more powerful hybrid magnets. Nanocenters and Materials Synthesis The past decade has given rise to significant investment in the establishment of a diverse portfolio of nanoscience research centers. This development was made possible by the stewardship of the multiagency National Nanotechnology Initiative (NNI). The centers complement traditional major neutron and photon sources for CMMP research and include strong user support in their mission statements. The centers differ in character from one another according to the directives of their sponsoring agencies. But, more significantly, they are in many ways distinct in character from large-scale facilities such as neutron and photon sources. The primary focus of the nanocenters is on the creation of new materials as well as on the advanced characterization of materials, while the other major facilities deal pri- marily with advanced characterizations. This focus represents a turning point, an acknowledgment of the central importance of the need for new materials in order to invigorate CMMP. This is a theme that needs to be extended and broadened in the next decade in order for the United States to recapture its leadership in the area of the discovery of new materials. In this subsection, nanocenters are discussed and the model is considered for the design and discovery of new materials of interest to CMMP researchers, such as bulk crystals, novel thin films, and superlattices. Current Status of Nanocenters and Materials Synthesis Researchers at many institutions face challenges associated with the avail- ability of materials. They may lack the expertise or the appropriate equipment 11 National Research Council, Opportunities in High Magnetic Field Science, Washington, D.C.: The National Academies Press, 2005.

OCR for page 193
c o n d e n s e d - m at t e r m at e r i a l s P h ys i c s  and for the synthesis or growth of new or high-quality materials. The NSF National Nanotechnology Infrastructure Network (NNIN) program is intended to address these issues. The NNIN program is largely directed at providing capabilities for the synthesis and fabrication of materials and for providing computational and theoretical tools and expertise. A network of 13 universities around the country (see Figure 11.14) participates in this program to provide and share facilities for nanoscience and engineering research. In addition to the NSF NNIN program, DOE has established Nanoscale Science Research Centers at five national laboratories: the Center for Nanophase Materials Sciences at ORNL, the Molecular Foundry at LBNL, the Center for Integrated Nan- otechnologies jointly operated by Sandia National Laboratories and Los Alamos National Laboratory, the Center for Nanoscale Materials at ANL, and the Center for Functional Nanomaterials at Brookhaven National Laboratory. These centers are largely dedicated to materials synthesis, fabrication, and characterization. They provide access to electron-beam nanowriters, lithography and stamping for nano- fabrication; x-ray nanoprobes and facilities for complex materials formation and soft hybrid materials; and infrastructure for theory simulations. The nanocenters are distributed facilities that embrace interdisciplinary ap- proaches to solving nanoscience and nanotechnology problems using a full suite of modern instrumentation. At many of the nanocenters, theory and simulation FIGURE 11.14 Institutions participating in the National Nanotechnology Infrastructure Network pro- gram. NOTE: UCSB, University of California at Santa Barbara; PSU, Pennsylvania State University; TNLC (NCSU), Triangle National Lithography Center (North Carolina State University). SOURCE: See http://www.nnin.org.

OCR for page 193
t o o l s , i n s t ru m e n tat i o n , facilities cmmP research  and for are on a similar footing with experimental science. Also, the pursuit of world-class, in-house scientific research programs is on a similar footing with user support services. The nanocenters provide a pathway from fundamental science to applica- tions with the possibility of commercialization and the creation of new start-up companies. The balance between science and user-service programs at the nanocenters is approached differently, depending on the sponsoring agency. For example, the DOE NSRCs encourage a model whereby each Ph.D. staff member pursues basic science research and user-support services. The NIST Center for Nanoscale Science and Technology has separate divisions that emphasize scientific programs and user sup- port. The NIST scientific programs focus on solving major measurement-related obstacles in the path from discovery to production. The Department of Defense (DOD) supports an in-house, mission-oriented Institute for Nanoscience at the Naval Research Laboratory. Medium-Term Developments The challenge ahead is to learn how to sustain the progress of the nanoscience era and to optimize accessibility to a diverse range of instruments and facilities. In cases where nanocenters are co-located with other major facilities, the planning of one-stop shopping needs to be perfected so that newly created nanosystems can be interrogated with electrons, neutrons, and x-rays in a single visit. Metrics need to be refined for monitoring the success of the new nanocenters. The funding for operations needs to support the diverse suite of equipment at the nanocenters. Models need to be evaluated for a balance of in-house science and user sup- port. Barriers will need to be lowered to facilitate the transition from science to commercialization. The nanocenters have addressed a gap in research culture by acknowledging the importance of synthesis, processing, and fabrication of new materials and systems. Recognizing this, it is imperative to accelerate the momentum and to energize other areas of new-materials exploration and discovery of vital interest to CMMP. The design and synthesis of novel systems are the foundation to address all of the CMMP grand challenges. The energy challenge needs new materials for storing hydrogen, thermoelectrics, organic light-emitting diode (LED) crystals, and high-performance superconductors and ferromagnetics. Information technology needs new materials for spintronic, organic, and molecular electronics that ex- hibit quantum coherence properties suitable for quantum computation prototyp- ing. Multiferroics, magnetic semiconductors, and half-metallic ferromagnets are specific systems also of great interest to spintronics. As the art of crystal growth matures into a science, the resulting insights might apply to the understanding of the physics of soft-matter crystallization. Protein crystallography data collection at

OCR for page 193
c o n d e n s e d - m at t e r m at e r i a l s P h ys i c s  and photon sources would benefit immensely from such a development. This example highlights the multidisciplinary nature of the quest that brings together chemists, biologists, engineers, and physicists. The new crystal discovery centers of the future can be distributed, as are the nanocenters. The models for creating and operating them might also benefit from examining the NSF NNIN and DOE NSRC models. Presumably viable hybrid organizational structures will evolve that best serve the particular materials missions of these future efforts. A target budgetary level per year might be similar to that for NSF of its NNIN program, and for DOE a level of the equivalent of one or two of its five NSRC facilities (for further discussion, see Chapter 9). Why is it imperative to move forward on a new-materials discovery agenda now? The United States is not in a lead role in the creation of new materials. It needs to recapture its lost status, because the consequences of delay and neglect are long- term erosion of the U.S. competitive edge and a loss of intellectual property. New materials invigorate all of the CMMP grand challenges. While new-materials dis- covery is cross-disciplinary, at present there is no obvious academic home for new- materials initiatives. This problem needs to be remedied. New-materials discovery embraces theory and simulation in the sense of virtual fabrication. New materials created via computer models, including electronic band structure codes, provide insights and guidelines to direct the design of new materials in the laboratory. The new-materials discovery centers of the future will also provide fertile training grounds for future generations of graduate students. The nanocenters started the culture change by emphasizing the creation of new materials. The transformation needs to be extended to embrace the larger landscape of new-materials discovery beyond the nano-realm. The time is ripe to focus on this strategic scientific goal, to plan multidisciplinary team approaches, and to identify visionary management, scientific advisers, and stakeholders, as stated above. Balance must be sought be- tween support of the individual investigators and small groups of investigators relative to centers, instrumentation, and major facilities investments. Recommendations for Materials Synthesis and Nanocenters in CMMP Research Nanoscience is a core discipline whose advances will affect all of the other challenges, from emergent phenomena (Chapter 2) to information technology (Chapter 7). The past decade has already seen significant federal investment in nanotechnology infrastructure. Notable are the NSF-funded Nanoscale Science and Engineering Centers and the National Nanotechnology Infrastructure Net- work, as well as the new DOE-funded Nanoscale Science Research Centers at the national laboratories. These facilities serve a critical need and deserve continued support. Nanoscience by its very nature spans an enormously wide variety of disciplines, from condensed-matter physics to engineering to chemistry and biol-

OCR for page 193
t o o l s , i n s t ru m e n tat i o n , facilities cmmP research  and for ogy. This makes it all the more critical to develop an intellectual resource network that allows scientists from one discipline to have access to the knowledge of all of the others. There is need for training opportunities for students, postdoctoral researchers, and faculty that allow them to reach beyond the standard disciplinary boxes. There is also need to develop knowledge repositories like those that exist in biology, where genomes and so on are stored and made widely available. The NSF and DOE-funded nanoscience centers should take the lead in meeting these needs, teaching short courses on particular techniques and subfields, as well as providing repositories of information. Recommendation: DOE and NSF should develop distributed national facili- ties in support of the design, discovery, and growth of new materials for both fundamental and applied CMMP research.12 Recommendation: DOE should evaluate the new NSRCs by metrics described in Chapter 9. The National Nanotechnology Coordination Office (NNCO), in its arrangement of the triennial review of the NNI, should evaluate all NNI- funded centers and networks of centers by similar metrics. Large-Scale High-Performance Computing Facilities High-performance computing is well recognized as a prerequisite for scientific and technological preeminence. High-priority, significant resources at the federal level are therefore directed toward the ongoing development and maintenance of state-of-the-art computational facilities for general scientific research, includ- ing CMMP. In understanding how such resources address the needs of CMMP researchers, it is important to note that large-scale computation is an important component of many scientific fields that share these resources. Below, the com- mittee describes the major U.S. high-performance computing facilities and shows data as to how the available resources are shared among disciplines. Current Status of Computing Facilities The largest and most powerful systems define the limits of the types of compu- tational studies that can be carried out at present. For the U.S. CMMP community, these computational facilities are supported by NSF, DOE, and DOD. Building on the system of NSF supercomputing centers of the 1990s, the devel- 12 The National Research Council study Assessment of and Outlook for New Materials Synthesis and Crystal Growth will make detailed recommendations on how best to support this need. The report is expected to be released in the summer of 2008.

OCR for page 193
c o n d e n s e d - m at t e r m at e r i a l s P h ys i c s  and opment of the TeraGrid began in 2000 as the world’s largest, most comprehensive distributed cyberinfrastructure for open scientific research. Partners in this dis- tributed framework include the National Center for Supercomputing Applications (NCSA) at the University of Illinois at Urbana-Champaign, the San Diego Super- computer Center (SDSC) at the University of California at San Diego, Argonne National Laboratory, the Center for Advanced Computing Research (CACR) at the California Institute of Technology, Pittsburgh Supercomputing Center, Indiana and Purdue Universities, Oak Ridge National Laboratory, and the Texas Advanced Computing Center at the University of Texas at Austin. As of 2005, the TeraGrid had about 1600 users. A relatively small fraction is used for materials and (all of) physics research (see Figure 11.15). DOE supports scientific computing primarily through the National Energy Re- search Scientific Computing Center (NERSC) and through Leadership Computing Facilities (LCF) at the national laboratories. NERSC is described by DOE as “one of the largest facilities in the world devoted to providing computational resources and expertise for basic scientific research,” with 2677 users in 2005. Reflecting the broad DOE mission, a relatively small fraction of resources (about 9 percent) is devoted to computation for materials research (see Figure 11.15). At national laboratories, such as ORNL where materials are a larger compo- nent of research, the fraction of resources at the LCF is correspondingly higher. a b FIGURE 11.15 (Left) National Science Foundation TeraGrid usage, by discipline, in FY 2005. NOTE: “Computer & Info,” computer science and information technology. (Right) Department of Energy Na- tional Energy Research Scientific Computing Center usage, by discipline, in 2005. NOTE: “Lattice QCD,” lattice quantum chromodynamics. SOURCES: (Left) National Science Foundation TeraGrid. (Right) National Energy Research Scientific Computing Center, Lawrence Berkeley National Laboratory. 11.15 a,b

OCR for page 193
t o o l s , i n s t ru m e n tat i o n , facilities cmmP research  and for At ORNL’s National Center for Computational Sciences, about 25 percent of the center’s resources are used for materials computations. A number of the LCFs are also partners in the NSF TeraGrid. DOD has a large network of supercomputer centers (the High Performance Computing Modernization Program) to support the computing needs of DOD researchers, with 4550 users in 2005. Materials research falls in the category “CCM” (Computational Chemistry, Biology, and Materials Science). The share for this category can be seen in Figure 11.16. While the focus of this discussion has been on high-performance computing, there is much interesting and innovative work done in computational materials that does not demand computational resources at the highest available level, but where accessibility and throughput are key considerations. Much valuable work is done at computing facilities at the state level, at individual universities, in depart- ments, and by research groups. Support for computational facilities from sources such as the NSF Major Research Instrumentation program should be encouraged, and budgeting for computer equipment in theoretical and computational CMMP FIGURE 11.16 FY 2006 Department of Defense high-performance computing requirements, alloca- tions, and utilization breakdown for individual “computational technology areas.” Computational Chemistry, Biology, and Materials Science (CCM) is third from the left; other areas are Computational Structural Mechanics (CSM), Computational Fluid Dynamics (CFD), Computational Electromagnetics and Acoustics (CEA), Climate/Weather/Ocean Modeling and Simulation (CWO), Signal/Image Process- ing (SIP), Forces Modeling and Simulation (FMS), Environmental Quality Modeling and Simulation (EQM), Electronics, Networking and Systems (ENS), and Integrated Modeling and Test Environments (IMT). SOURCE: C.J. Henry, Department of Defense High Performance Computing Modernization Program.

OCR for page 193
c o n d e n s e d - m at t e r m at e r i a l s P h ys i c s  and individual and small-group proposals should be considered the norm. However, this hierarchical structure, while it evolved largely to meet the needs of researchers, does come with problems of its own. As computational power increases, issues of professional systems administration and user support personnel for computing clusters become increasingly important. The diversity of facilities can make the system hard to navigate for researchers seeking resources. This latter challenge is particularly common for computational junior faculty members starting careers in computational CMMP. CONCLUSIONS The need for sophisticated tools (experimental, computational, and theoreti- cal) to probe the structure and properties of materials over a wide range of length scales is essential for continued progress in CMMP research. The new-generation facilities (light and neutron sources, magnetic-field facilities, and electron micro- scopes), which offer higher fluxes and energies, provide significant advantages with regard to resolution, sensitivity, and data acquisition. Two additional challenges will continue to be important in the future: the simultaneous measurement of structure and dynamics over various time and length scales and dimensions, and the simul- taneous measurement of structure and dynamics while the system is perturbed independently by an external field (magnetic, stress, electric, and so on). The synthesis, structure, and properties of materials are all intimately con- nected, so researchers will increasingly need to be intimately familiar with this entire spectrum of activities. Lessons learned from one class of materials will increasingly be used to understand the behavior of seemingly different classes of materials. For the first time in history, the complexity of CMMP is such that new advances in the field will depend on strong support for large-scale facilities, mid- scale facilities, interdisciplinary research centers, and individual investigators who actually carry out the research. Students will have to understand computational methods, together with the full spectrum of experimental endeavors (synthesis, fabrication, and measurement) to become successful researchers.