This case was written by Juan D. Rogers. Comments or questions should be directed to Juan D. Rogers at (404) 894-6697 or, by email, juan.rogers@pubpolicy.gatech.edu. The research was sponsored by the Department of Energy, Office of Basic Energy Sciences, Contract ER45562. The views presented here are the case author’s and do not necessarily represent those of the Department of Energy, the University of California, or Georgia Institute of Technology.

The Nuclear Magnetic Resonance Spectroscopy case examines the research on NMR techniques for the study of solids and quasi-solids developed mainly under the leadership of professor Alexander Pines at UC Berkeley. This work was able to overcome inherent limitations of NMR for chemical analysis of materials other than liquids increasing it sensitivity and resolution by several orders of magnitude. It now has numerous applications in materials sciences, electronics, biology, the oil industry, and general analytical chemistry instrumentation. More than 50 graduate students worked on PhDs under professor Pines and have been hired in top academic institutions and private industry.

The funding for this stream of research was originally provided by NSF (at a 75% level) with supplemental support from the university. In 1978, the program received full funding from DOE/BES and has continued that way to the present. It is a good example of multifaceted research with a very important fundamental science component with implications for many disciplines and significant impact in private industry. The diversity in its scientific and industrial implications contrasts with the central role of a single person over the span of 25 years. He led several teams of researchers and students at his own lab and conducted collaborations with many scientists at other institutions in the country and abroad. From the sources available, it seems he is the only person linking them all together.

Sources: This case is based on an interview with professor Pines (August 14^th, 1997), conducted by Barry Bozeman and Juan D. Rogers. The historical analysis draws as well on materials provided by professor Pines: proposals, field reports, articles, patent citations, news releases, etc. covering the period of 1973-1997.

The Nuclear Magnetic Resonance (NMR) Spectroscopy case did not have a single scientific output that can be identified as the objective of this stream of research. Rather, a team of researchers led by Alex Pines worked with fundamental principles of nuclear magnetic resonance to develop numerous techniques, instruments and applications that on several occasions even spanned several disciplines including chemistry, engineering, materials science, and biology. The main contributions were in the use of NMR techniques for the study of solids.

The use of NMR to study materials can be traced back to the work of Felix Bloch, Edward Purcell, Robert Pound, and Martin Packard during the 1940s. Bloch and Purcell, team leaders at Stanford and Harvard respectively, were awarded the Nobel prize in 1952 for its application to solids. "Nuclear Magnetic Resonance involves the absorption and emission of radio-frequency energy by nuclear spins as they oscillate and reorient themselves in the presence of internally and externally applied magnetic fields" (Chmelka and Pines 1989, 71). In other words, NMR makes use of the fact that spinning nuclei are charged particles in movement and generate magnetic fields. When irradiated with external electromagnetic energy, the interactions will follow patterns that depend on the local environment in the sample. Materials are subjected to radio-frequency radiation of known characteristics. After interacting with the material, the energy absorbed and emitted is analyzed to detect its frequency patterns, which contain information about the characteristics of the material.

Depending on the experimental setup, an image of the sample or the spectrum of its frequency components can be constructed. The former is NMR imaging (MRI) and the second NMR spectroscopy. In the case of imaging, rf radiation of constant frequency is applied in the presence of variable magnetic field. The detected energy will have patterns that correspond to the total number of spins in the sample yielding an image of the atom density of the sample at every point.

In the case of NMR spectroscopy, the frequency of rf radiation is varied and all the frequencies emitted and absorbed by a given material are detected resulting in the resonance spectrum of the material. This spectrum contains a wealth of information about the composition, structure, and dynamics of the sample at the molecular scale.

NMR has three main features that make it a valuable tool for laboratory analysis. First, NMR is able to focus on individual atomic sites through the nuclei. This contrasts with optical spectroscopy that is able to visualize only whole molecules. Second, NMR can show the relationships between atoms in different sites. These are structural relationships concerning the relative position of atoms in a sample. Dynamic relationships concern the relative motion of atoms. NMR can reveal these relationships because they produce characteristic frequencies in the spectrum. Finally, NMR is non-invasive and non-destructive. The interactions of spinning nuclei with radio frequency energy occur without altering the fundamental characteristics of the material.

The advantages of NMR come at a price set by several inherent drawbacks. These are: low sensitivity, poor resolution, and weak selectivity. The first comes about because of the low energy of systems in the radio frequency portion of the spectrum. Therefore, large numbers of atoms and high magnetic fields are required in order to see the effects. Second, selectivity is low because of the long wave-length of radio frequency. This makes it difficult to perturb only a selected part of the sample. Finally, NMR has two sorts of resolution problem that result in the broadening of the lines of the spectrum. One occurs when a spectrum is inherently complicated and there are so many sharp lines that the spectrum cannot be interpreted. The other occurs when simple spectra overlap and cannot be separated from each other.

The main knowledge contributions of this stream of research under the leadership of Alex Pines involved overcoming some of these drawbacks of NMR while maintaining its advantages and, therefore, extending its use to a broader range of materials. The development of instruments must be counted as significant scientific contributions as well because of their enabling features in other fields of research.

Alex Pines received his PhD at MIT in 1972 under the direction of one of the pioneers of NMR, John Waugh. Pines’ decision to enter the field, according to his recollection, was the result of his inclination to combine fundamental science with technical development. The fact that Waugh both knew the physics of nuclear magnetic resonance and could build electronic devices attracted him.

His work with Waugh led to the development of Proton-Enhanced Nuclear Induction Spectroscopy (PENIS), a technique that allowed high-resolution spectroscopy of dilute spins that occur, for example, in the study of impurities in solids. At that time, NMR was an established tool for the study of liquids and liquid-like systems. It is much easier to obtain high-resolution spectra for these materials because the natural random motion present in liquids averages out the influences that broaden spectral lines. This occurs because spins flip easily in and out of alignment when an external magnetic field is applied. The easy realignment does not occur in solids and the lack of mobility results in the rf energy being dispersed in a larger frequency interval. The new technique developed in Pines’ PhD work allowed higher resolution spectra for certain materials that had not been susceptible to characterization by these techniques before. One such application was the analysis of the bonds of fuel to the source material in oil shales. The technique permitted the extraction of NMR spectra of ¹³C in non-isotropic media. The PENIS technique also reached commercial applications as Bruker developed and marketed an accessory package for NMR spectroscopes based on Pines’ work.

In 1973, Pines joined the faculty of the department of chemistry at the University of California, Berkeley. He already realized that his work would have most of its implications for the study of materials. Therefore, he contacted the director of the Inorganic Materials Research Division, Leo Brewer, and requested becoming a part the unit. IMRD provided some of the funding required to develop the devices needed to implement the techniques. A grant from NSF for work in the period 1973-1977 covered the rest. The documentation reveals that the budget required constant negotiation because the development of devices to implement the techniques required fairly large, indivisible, pieces of equipment such as the high power magnets. Therefore, funding of proposals at a fraction of the requested budget did not allow even for partial implementation of the research program.

The main scientific results of the period between 1973-1980 led to the development of multipolar spectroscopy. For this, a high field double resonance spectrometer was developed that had high sensitivity in the detection of low frequency nuclei. The development of this instrument and its associated techniques required the study of some fundamental aspects of particle interaction, especially, multiple nuclear resonance in solid state samples. The elucidation of these processes led to the development of detection schemes which allowed characterization of a variety of systems involving the observation of excited molecular states in organic and biologically oriented solids and quasi solids.

As the applications of NMR to a broader range of materials became clearer, Pines was able to secure funding for his research program from DOE’s Office of Basic Energy Sciences. His initial strategy was to diversify the funding sources given the earlier problems with partial funding and the impossibility of obtaining large expensive components for building his novel instruments. However, the program managers at DOE showed considerable interest in Pines’ techniques and offered to fully fund the program. The composition of support for Pines’ work changed in 1978 to a single source, BES. A small amount of about 10% of the total budget was provided by industry partners interested in either the applications of NMR techniques or the commercial development of NMR instrumentation.

The period of the early to mid 1980s gave rise to two highly successful techniques developed by Pines and his group: multiple-quantum NMR and zero-field NMR. Multiple-quantum NMR is a technique designed to avoid the problem of "homogeneous broadening," which reduces the resolution of NMR spectra. "In this technique, with each pulse of the radio wave the molecule absorbs a group of photons –up to several hundred at a time—rather than a single quantum of energy" (Goldhaber 1986). When a large array of photons are irradiated on the sample, there is only a limited number of ways in which it can be absorbed. Multiple-quantum excitation forces the group or nuclei to respond together eliminating the coupling between nuclei that creates multiple spectral lines clustering in the same region of the spectrum. These improvements are especially useful in the case of characterizing the distribution of atoms in amorphous materials. Pines’ group collaborated with Jeffrey Reimer of LBL’s Center for Advanced Materials in the development of multiple-quantum NMR spectroscopy because of its potential in the analysis of amorphous silicon hydride.

Zero-field NMR was developed to counter the problem of "inhomogeneous broadening." In some samples, there are many molecules that have simple spectra but may overlap producing a "smear." Because NMR spectroscopy is a sort of diffraction, the formation of spectra is the result of the interaction of radiation with a regular pattern of particles in the sample. When the material is amorphous, like glass, the spectrum is a scatter with little information. The problem occurs because amorphous materials are disordered only with respect to the magnetic field that is introduced to observe the NMR effect. When the field is not present, the symmetry of space is restored. By switching the magnetic field off in an adiabatic demagnetization process, that is, without allowing the sample to exchange heat with its surroundings, the NMR effect occurs, but the orientation information is not lost by the imposition of the preferred direction of the magnetic field. The spectrum then contains sharp lines at the frequencies defined by the local environments of the nuclei.

At this stage, the variety of possible applications of NMR spectroscopy led to a fairly large number of collaborations. These included researchers from UC San Francisco Medial School, several of LBL’s divisions, such as the Biology and Medicine Division, Cornell University’s Department of Mathematics, the Solid State Physics Institute of the University of Stuttgart, Exxon, US Naval Research Lab, Mobil Oil, and Monsanto. Pines has emphasized repeatedly that the Lawrence Berkeley Laboratory has been a major source of collaborative efforts contributing to this stream of research.

Two more techniques were developed during this period and resulted from mathematical work that suggested ways in which a solid could be made to behave more like liquids. These are Double Rotation (DOR) and Dynamic-Angle Spinning (DAS). By spinning the sample around two axes the symmetry with respect to the external magnetic field appears to be closer to that of a sphere. The orientations of the individual atoms approximate the random orientations of a liquid and increase the degree of mutual cancellation yielding a spectrum with much higher resolution (around 2 orders of magnitude). This technique received a US patent and was licensed to several private companies. It was also awarded the "R&D 100" award by the publication Research and Technology for outstanding achievement in technology.

The techniques developed during this stage appear to have had the greatest impact of all developments in this research stream. The productivity in scholarly journals was high at all stages (an average of 10 co-authored papers published in refereed journals per year), however, at this time the results were cited in a variety of contexts and types of publications. It led to two publications in Science, others in Nature as well as several featured articles in industry magazines and mentions in sections on "hot" research news in non-specialized journals.

During the 1990s a technique that improved the poor sensitivity of NMR was developed. An inert gas, Xenon, was used as an intermediary to enhance the signals from inherently weak sources such as the atoms on a surface. Xenon is polarized with a laser that aligns the spins of the gas in one direction. Then the polarized gas is allowed to enter into contact with the sample binding to surfaces without altering them. The polarization of the Xe atoms is transferred to the surface atoms and these can be detected with sensitivity increasing in four orders of magnitude. These results were again of high visibility and the technique was featured in journals of general scientific interest as well as news and R&D industry publications.

Beginning in 1978, almost all the support for this research program was provided by BES. Therefore, this section will report on outputs corresponding to that period. The data for the tables that follow were compiled from a set of LBL annual reports that extends to 1990. The period of 1991-96 is not covered by the available materials.

Year	Papers	Technical Reports	Dissertations
1979	8	10	-
1980	9	8	1
1981	6	7	3
1982	7	8	4
1983	ND	ND	ND
1984	13	15	-
1985	11	20	2
1986	19	31	4
1987	25	24	4
1988	12	14	1
1989	8	22	3
1990	11	20

In order to interpret these figures correctly it must be noted that almost all papers published in refereed journals, listed in the "Papers" column, go previously through a report stage. Therefore, most reports in one year, listed in the last column, become papers in the following year. Theses, dissertations, and drafts of chapters in books are also counted as reports until they are published in another medium. Then, the figures in the last column are also included in the total of the "Reports" column.

Throughout the period for which detailed reports are available, Pines and/or his team members attended numerous invited talks. The number ranged from 11 in 1982 to 48 in 1988, 38 of which were given by Pines himself. Most of the presentations during the earlier years were given at universities or academic conferences. Around 1987, the number of presentations at private industry sites increased significantly, including IBM, AT&T, DuPont, Shell, Mobil. However, the predominance of presentations at academic institutions and meetings continued with a larger number of presentations abroad.

During the interview, Pines mentioned that conferences and lecture presentations are one of the most significant occasions for setting up collaborative efforts.

There is no exhaustive list available of students and their placement after graduation. However, from interview data we can determine that many students went into fields of the applications of NMR spectroscopy rather than continuing to focus in spectroscopy itself. Pines mentioned the fields of biochemistry, electronics as well as chemistry and physical chemistry.

The available data does not include an exact, year by year, report on funds requested and granted. The amounts varied considerably in time depending on the requirements for major equipment purchases, which in 1973, for example, were $131,000 plus $64,000 for other operating costs taking the total to $195,000. The following year, total operating costs were $80,000. This example comes from years in which the research program was supported by NSF with supplemental funds from the university. In 1987, the major equipment budget was $1.78M and the other operating costs $500,000 taking the total budget to around $2.3M. During the latter period of existence of this research program, the staff and operating costs have remained at the same level of magnitude but there are no data on the details of equipment purchases.

There is a direct relation between Pines’ scientific agenda and the patterns of collaborations and career paths. The people trained under him go, more often than not, into fields of expertise related applications of Pines’ central concerns. The impact of Pines’ work, therefore, has two dimensions related to the scientific agenda: the specific knowledge contributions and experts in applications fields that have a different professional profile from his own.

The relation with industry is another important impact area. It is reflected in his patent and licensing agreements. It results from the same peculiarities of his scientific style.

The fact that NMR spectroscopy works on any kind of materials makes its implications interdisciplinary. This has had an impact on the publication strategies of the PIs who have made a deliberate effort to publish their research results in journals that reach a broad scientific audience, such as Science or Nature. The main competitive publications were published for the most part in the Journal of Chemical Physics, a journal that grew in importance in the field of Physical Chemistry in part through the success of NMR, among other things. This publication strategy went hand in hand with patterns of collaborations because the details of the materials NMR were applied to demanded expertise that cannot be expected in a single research team. It also led to special patterns in training students that have participated in this research stream. The focus is not the same as most research programs that single out a system and apply multiple techniques to uncover the characteristics of that particular system. Rather, a single set of NMR techniques was used on many systems in biochemistry, electronics, and so on. Therefore, students were encouraged to work on systems they found interesting and develop expertise in them while contributing to the team effort of extending NMR across disciplinary boundaries. As a result, graduates have been hired in various academic departments and by industries related to the particular systems they worked on.

In spite of the great diversity of activities and systems to which NMR is applied in this research, the lab is organized around its leader. He is co-author of all the papers produced by the group and maintains the coherence of the research done by people with quite different interests and career paths. His role seems to have changed gradually over the years, if we are to infer from his role in publications and the sorts of activities that appear over the years. The first ten years have Pines as the main author of almost all publications and the reports indicate his direct involvement in every aspect of fundamental research activities. The second period, during the mid to late 1980s, sees his role as a mentor and research leader greatly increased. He is the last co-author in almost every paper but the number and diversity in collaborative efforts as well as the sub-groups in his own lab increased dramatically. This is the period when his reputation as one of the central figures in the field is definitively established and the major impacts of the work occur in transferring techniques and personnel to other fields and industry. The last period, which covers the 1990s, he still seems to be quite involved in mentoring and leading his lab but much less of a protagonist in the actual day to day fundamental research. The excitement of the action in NMR seems now to be in the work of the next generation that looks up to him and organizes symposia to honor his career contributions.

The description of the research activity of Pines’ lab has become more formalized in later years (since around 1990). It is described as comprising four areas of activity. First, the work in fundamental science is always central because the application of NMR to any sample raises basic questions about the nature of the phenomena. Second, within the framework provided by possible answers to those basic questions, new techniques to study various samples with NMR are proposed and developed. Third, these techniques are tried and extended to systems in various fields. Fourth, particular technological implementations of instruments based on the techniques are developed.

Chmelka, B. F. and A. Pines. (1989). "Some Developments in Nuclear Magnetic Resonance of Solids." Science, 6 October 1989, Vol. 246, pp. 71-77.

Goldhaber, J. (1986). "Nuclear Magnetic Resonance: Two New Techniques Expand NMR’s Potential for Laboratory Analysis of Materials." LBL Research Review, Spring 1986, pp. 20-25.

-----, (1991). "Sharpening the Spectra: New NMR Techniques are Clarifying the Structure of Solids." LBL Research Review, Fall/Winter 1991, pp. 12-19.

Maciel, Gary. (1984). "High-Resolution Nuclear Magnetic Resonance of Solids." Science, 19 October 1984, Vol. 226, pp. 282-288.

Munowitz, M. and A. Pines. (1986). "Multiple-Quantum Nuclear Magnetic Resonance Spectroscopy." Science, 1 August 1986, Vol. 233, pp. 5525-531.