Reports of ICAM Workshops held in FY 1999

Adaptive Atoms in Physics, Chemistry, Biology and the Environment


Daniel Cox (Dept. of Physics, UC Davis)

Zachary Fisk (National High Magnetic Field Laboratory, Florida State University)

John Kaszuba (Environmental Chemistry Group, Los Alamos National Laboratory)

Andrew Shreve (Bioscience and Biotechnology Group, Los Alamos National Laboratory).

Dates: 6/28/99-7/2/99.

Location: Center for Materials Science, Los Alamos National Laboratory

Synopsis prepared by workshop leaders

A small and lively workshop with participants from the solid-state physics, chemistry, environmental science, and bioscience communities was held. The primary goal was to assess the potential for developing multidisciplinary approaches to enhance understanding of mixed- and intermediate-valence phenomena as they occur in physical, chemical, biological and environmental contexts. By the end of the workshop, an enthusiastic consensus was reached that such approaches can be developed and will be important. There was general agreement that the study of (adaptive) mixed-valence atoms could provide a strong organizing principle for a research focal area of the Institute for Complex Adaptive Matter. In this synopsis, we present a brief overview of a unifying conceptual cross-disciplinary framework that emerged from the workshop and provide some examples of how application of this framework can help facilitate collaborative approaches. In the future, a longer and more detailed technical document will be developed by the organizers with input from workshop participants.

Conceptual Framework

Metal ions existing in mixed, intermediate or fluctuating valence states are ubiquitous in solid-state physics, biology, chemistry and the environment. Historically, the study of such systems within the context of each discipline has yielded great insight and understanding of the issues that arise within each discipline. However, increasingly, new scientifically and technically important materials and problems are emerging that will require treatments that merge and extend single disciplinary approaches. Examples include:

- High-temperature superconducting materials

- Colossal magnetoresistance materials

- Nanoscale electronic materials that are intermediate between molecular and extended systems

- f-element ions in both intermetallic materials and environmentally important heterogeneous solids and solutions

- Synthetically tuned molecular clusters and molecular crystals with current and potential applications in molecular electronic devices and molecularly designed magnetic materials

- Mimicking biological energy transduction, including the coupling of electron transfer processes to specific structural responses.

Based on the workshop, the organizers are proposing an organizing framework to facilitate cross-disciplinary discussions of each area in terms of a unified language. The framework and language are based on the concept of competition between electronic and structural responses, which in turn, determines behavior through the separability, or lack thereof, of electronic and structural time scales. To some degree, similar frameworks exist separately within each discipline (albeit with each having quite different nomenclature and language). Thus, a clear and unified description in terms of electronic and structural dynamics, which is often not the explicit focus of the treatments that have been developed, can build from this base to provide a language that will aid in establishing connections between strong efforts already existing within each discipline and in developing new multidisciplinary experimental and theoretical approaches.

To illustrate this framework, some specific examples are discussed (as noted above, a more technical discussion of these examples will be presented in a future document). In solid-state physics, a common manifestation of intermediate valency phenomena occurs in the context of the Anderson impurity model. A generic manifestation considers an ion embedded within a metal, with electronic coupling present between the ion's valence electrons and a conduction band of the metal. This electronic coupling, including strong correlation effects (occurring when the electronic repulsion energy scale exceeds the electronic kinetic energy at a given temperature), leads to a solution consisting of a (coherent) amplitude level superposition of states involving different valencies of the ion. Thus intermediate (or fluctuating) valencies are obtained. In this discussion, the time scale that characterizes the electronic coupling is much faster than the time scale associated with the structural response of the material, and indeed, successful treatments of this problem can often completely ignore phonons.

When comparing to molecular systems, casting the problem in terms of time scales leads to clear connections between this case and the so-called delocalized limit, in which the time scale associated with the inverse strength of the electronic coupling between ions (or between an ion and its ligands) in a molecule or cluster is much faster than the time scale of symmetry breaking vibrations (scaled by the coupling between electronic and structural degrees of freedom). In this limit, however, the treatment of correlation effects often can become quite important, and an open and interesting question that arose repeatedly during the workshop is the degree to which treatments derived in the solid-state physics community could be usefully applied to molecular systems to address such issues. A particular example discussed by several speakers was the understanding of the electronic structure of cerocene or other metallocenes in which a metal ion can be taken to be electronically coupled to a (finite) space of mobile electrons, and in which the valency of the metal ion is found to be effectively non-integral. Are there useful insights into the structure of such complexes based on impurity models? What are the formal connections between such treatments and quantum chemical approaches, and what are the advantages and disadvantages of each for these problems? These are questions that were discussed during the week, and which may serve as focal points for specific collaborative and cross-disciplinary projects.

To continue, for many important materials the electronic time scale can also become comparable to or even much slower than structural time scales. This can occur as either the electronic coupling energy becomes smaller, the size of the electronic "bath" involved in the coupling decreases (ranging through mesoscale nanomaterials all the way to molecular systems), or the coupling of structural motions to electronic states becomes stronger. In all such cases, structural degrees of freedom must now be included in order to obtain a useful understanding of mixed-valency. If the relevant electronic coupling is quite weak, and the corresponding time scale of electronic delocalization is very slow relative to structural responses, the near complete separability of these time scales allows for good understanding and control of mixed-valence phenomena, and more generally, electron transfer reactions. Such a scenario applies to much of biological electron transfer reactivity (although we note that the ability to use electron transfer reactions to drive specific, functionally useful, structural responses in complex environments remains a generally enigmatic aspect of biological energy transduction), and also to many synthetic molecular systems in which multiple metal ions are weakly electronically coupled.

However, much interesting and important chemistry and physics remains to be understood in the case where the relevant electronic and structural time scales are truly comparable. In such cases, one approach discussed extensively in the workshop is the use of numerically exact methods for solving model Hamiltonians. The application of such methods for examining molecular systems whose properties can tuned synthetically and that are amenable to study using advanced optical techniques that allow independent probing of electronic and structural responses on ultrafast time scales (including the role of coherent state preparation and evolution), is not yet well developed. Collaborations that address such issues are likely to be useful for both understanding the responses of the molecular systems and for benchmarking the Hamiltonians, and were discussed during the workshop.


The framework that is proposed here, in which mixed valency is presented in terms of competing structural and electronic time scales, has arisen independently (at least implicitly) in each discipline. Precisely for this reason, the organizers believe that it will provide a useful language in which to discuss and connect various treatments of mixed valency. Within this framework, collaborations and useful experimental and theoretical approaches have already been identified. For example, the role of materials synthesis in tuning properties, either through spatial restriction such as in nanoscale materials, through control of heterogeneity (which often can be usefully treated as providing coupled structural degrees of freedom whose characteristic time scales are extremely slow), or through control of electronic and structural responses is obviously important. Also, the application of novel experimental probes that could directly address ultrafast dynamics in strongly-correlated solid state physics materials is an interesting, provocative (and difficult) area. Exploring connections and developing synergy between strong correlation approaches and quantum chemical treatments will be useful in many contexts.


Vicki Colvin, Rice University

Daniel Cox, University of California, Davis

Deborah Evans, University of New Mexico

Zachary Fisk, NHMFL-Florida State University

Lev Gorkov, NHMFL-Florida State University

David Logan, University of Oxford

George McLendon, Princeton University

Joel Miller, University of Utah

Alexandra Navrotsky, University of California, Davis

Peter Nordlander, Rice University

Martin Plihal, Rutgers University

Gustavo Scuseria, Rice University

Rajiv Singh, University of California, Davis

David Clark, LANL

Robert Heffner, LANL

John Kaszuba, LANL

Richard Martin, LANL

David Morris, LANL

Mary Neu, LANL

Don Parkin, LANL

David Pines, LANL

Andew Shreve, LANL

William Woodruff, LANL 


Mesoscopic Organization in Soft, Hard, and Biological Matter

Leaders: Alexander Balatsky (Theoretical Division, Los Alamos )

Robert Laughlin (Physics Department, Stanford University)

David Pines (ICAM and LANSCE Division, Los Alamos)

Peter Wolynes (Chemistry and Physics Departments, UIUC)

Dates: August 24-28, 1999

Location: Los Alamos National Laboratory

Workshop Summary Prepared by Robert Laughlin and David Pines

One of the great triumphs of science in this century has been the discovery that the properties of matter at human length scales involve protectorates, i.e. rules of organization that are generic and universal. 1 The knowability of these properties is very different from those of atoms and small molecules, where first-principles computations starting from the underlying laws of nonrelativistic quantum mechanics are both routine and reliable. It is instead phenomenology coordinated with model-building and facilitated by the ready availability of accurate probes, such as optical spectroscopy in its various forms, neutron scattering, and magnetic resonance suitable for the length scale. That there should be such principles is fortunate, for the notorious cases in which they fail, such as the Kondo insulators and cuprate superconductors, demonstrate that without them our understanding is nearly nonexistent.

In between the regime of atoms and the quantum protection of stable states of matter is a scale at which neither first-principles computation nor measurement using existing technology sheds much light. This middle regime is that of mesoscopic organization, length scales between 10-4 and 10-7 - the domain of the glass phenomenon in solids and all of the machinery of life. Glass behavior is the physical scientist's name for a broad spectrum of strange phenomena seen by the spectroscopic tools at his disposal, including noncrystallinity, non-ergodicity, hysteresis, long-term memory, a bewildering variety of relaxation times, and sample-dependence.3 Were it not for the example of biology one might be tempted to dismiss this as an embarrassing exception to the beautiful rules by which matter organizes itself, a reason to discard samples before wasting one's time studying them. But the miracles of protein functionality revealed by traditional wet chemistry show that this need not be the case. The common ignoble view of the glassy state might simply be a false perception caused by having brought the wrong tools to bear and having asked the wrong questions. Whether or not this is the case is impossible to resolve with our present concepts and technological capabilities, for we understand only what we can see, and the universal mesoscope has yet to be invented. Understanding mesoscopic organization of matter, including biological matter, is one of the great unsolved problems of our time.

On August 23 - 27, 1999, the newly-organized Institute for Complex Adaptive Matter at Los Alamos held an interdisciplinary workshop on the subject of Mesoscopic Organization in Hard, Soft, and Biological Condensed Matter. The purpose of this workshop was to bring together experts from a variety of scientific disciplines, including optical and neutron spectroscopy, protein crystallography, gel chemistry, bioinformatics, and theoretical chemistry and physics, to discuss the status, future, and organizational needs of this emerging discipline. The topics discussed included:

The potential role of physical constraints in understanding gene expression. The inadequacy of current experimental techniques for addressing mesoscopic organization and ideas for their improvement. The intrinsic difficulty of understanding these systems through ab-initio computer modeling alone. The fundamental difficulties in proceeding from sequence to structure to function in biology and whether predictive power is possible even in principle. Whether and when it is appropriate to apply concepts borrowed from the large-scale organization of matter to the mesoscopic domain.

The difficulty of asking cleanly falsifiable questions. An important issue discussed at the meeting was the idea that protein folding is much more like crystallization than like glass formation. Peter Wolynes of the University of Illinois, Jose Onuchic of UCSD, and Angel Garcia of Los Alamos talked about protein-folding theory and the discovery that efficient folding proceeds through "funneling", the evolution of the shape through a sequence of configuration classes, each smaller than the previous one, with the special property that no configuration in a class is metastable.4 An energy landscape that funnels efficiently differs from one that does not by providing many routes to the folded state and few opportunities for arrest of the folding process. The difference between amino acid sequences that fold and those that do not is thus like the difference between liquids that solidify into crystals and those that do not, i.e. between crystals and glasses. Since crystallization is the simplest known example of the emergent property known as continuous symmetry breaking, the work raises the interesting possibility that there exists a "folding protectorate", in that some properties of the folded state might be protected and universal as a matter of principle.

Self-organization on the mesoscopic scale is also found in non-living things. David Whitten of Los Alamos described an astonishing new class of materials called organogels that fail to crystallize easily out of solvent but instead form organized arrays of fibers with complex substructures not unlike conventional organic gelatine, despite being composed of nothing but simple molecular monomers.6 David Oxtoby of the University of Chicago7 talked about the spontaneous formation of micelles, vesicles, and membranes by small amphiphilic molecules in water, mesoscopic-scale structures ubiquitous in biology but also easily created by artificial means, while Branko Stojkovic of Los Alamos8 talked about their counterparts in transition metal oxides and Langmuir films. In either case complex, non-ergodic behavior evidently resulted from elementary microscopic rules. This was also true in other systems that fail to order, in particular traditional solid glasses, polymers, and spin glasses. The low-energy excitation spectra of these were discussed by Feri Mezei of Los Alamos, who used neutron time-of-flight and spin-echo measurements to show that the famous Kohlrausch phenomenon, the nearly universal occurrence in amorphous solids of long-time dynamics characterized by stretched-exponential relaxation had significant length-scale dependence between 5 and 50 \AA. Elihu Abrahams of Rutgers reviewed some theoretical ideas about the Kohlrausch effect based on the concept of hierarchically-constrained dynamics, an idea related to self-organized criticality.\cite{abrahams} Glasses and spin glasses also have long-term memory effects known collectively as "remanence", a new and startling instance of which with demonstrable quantum content was reported by Gabriel Aeppli of NEC and also discussed by Dirk Morr of Los Alamos.11 Taken together these experiments hint at the possibility of mesoscopic organization in traditional glassy materials.

One of the most extensively discussed topics at the workshop was the presence of glassy behavior in systems with a high degree of crystalline order, suggesting that mesoscopic organization need not occur only in the context of chemistry but can also be purely electronic in origin. This understanding has come about as a result of the enormous effort over the past decade to improve the quality of correlated-electron materials as a result of the discovery of cuprate superconductivity. The observations of complex phase diagrams in these materials, for example the manganites discussed by Andy Millis of Rutgers and Yeong Soh of NEC, is slowly ruling out the possibility that these strange effects in correlated-electron materials are extrinsic. Chris Hammel of Los Alamos reported magnetic resonance measurements at a wide variety of doping levels in La2-xSrxCuO4 showing the universal presence of slow, inhomogeneous spin fluctuations not associated with either pure antiferromagnetism or superconductivity. There was a heated discussion about whether this might be defect magnetism associated with textures in the pattern of "stripes" or some other kind of order parameter that does not quite acquire long-range order. Sudip Chakravarty of UCLA noted that sum rules on the inelastic neutron scattering spectrum pointed to the persistence of pairing across the metal-superconductor boundary, as though the superconductivity were destroyed by massive amounts of atomic disorder, notwithstanding the cleanliness indicated by other experiments.12 There was also discussion of behavior, suspected by many theorists of indicating quantum criticality, found in both underdoped cuprate superconductors and heavy-electron systems. New experimental results on the latter were reported at the workshop by Joe Thompson of Los Alamos. Theoretical presentations suggesting a connection between mesoscopic organization and this behavior, as well as the phase diagram of the cuprates generally, were made by Alexander Balatsky of Los Alamos, Doug Scalapino of UCSB, Joerg Schmalian of Iowa State, David Pines of Los Alamos, and Branko Stojkovic. Further contributions to the problem of strongly-correlated electrons were made by Amit Chattopadhay of the University of Maryland, and Robert Schrieffer of FSU. There was general agreement that the limitations of measurement technology and sample quality had badly obscured the potential role of mesoscopic organization in correlated-electron materials.

There was great concern expressed by the participants about the task of developing new measurement strategies capable of shedding light in a meaningful way on mesoscopic organization. It was generally acknowledged that the limitations to measurement were fundamental, and that technological progress would require deep physical insight, qualitatively new approaches, and luck. A specific example of a relevant new measurement technology under development, the Scanning Probe NMR Force Microscope, was described by Chris Hammel. Several new ideas for reducing the effective wavelength of visible light by means of gratings and waveguides, so as to extend conventional optics, the most powerful probe at our disposal, to the mesoscopic domain were discussed by George Gruner of UCLA. Another strategy, to increase the power and resolution of existing techniques until they can overcome the limitations of extremely small sample size, was discussed by Feri Mezei and Jill Trewhella of Los Alamos. They compared the relative merits of X-ray and neutron technology in this context and explained the capabilities of hoped-for improvements in LANSCE. The relative merits of other emerging techniques, such as novel electron microscopy, conventional force microscopy, and optical tweezers were also discussed.

Perhaps the most important idea about the relationship between the physical sciences and the life sciences articulated again and again by the participants, was the hope that the culture of ideas engendered by mastery of measurement at long length scales might be combined with the culture of respect for the ingenuity of evolution to create a discipline greater than the sum of its parts. In her presentation to the workshop Jill Trewhella placed great emphasis on the scientific challenge of determining how macromolecular dynamics control biological function and described some current preliminary work on this topic. There was a lively discussion during her presentation about the value of structure in identifying and understanding protein function, and whether qualitatively new kinds of question based on mesoscopic thinking and measurement strategies might be an important missing ingredient. A specific example of such a question was raised by Robert Laughlin of Stanford in his discussion of (-independent transcription termination in procaryotes. He reviewed the experimental evidence that termination has a statistical aspect, i.e. that RNA polymerase sometimes reads through the termination command and sometimes not, and then argued that the observed balance in the branching ratio was unnatural and therefore suggestive of deterministic termination regulated by an on-board memory on the polymerase. He proposed some transcription experiments that could falsify this idea and also discussed the possibility that polymerase might be programmed by variations in the terminator. There was general agreement by the participants that this kind of idea-based interaction between theory and experiment was an important sense in which physical science could contribute to biology, and that the approach could aptly be termed physical bioinformatics.

In summary there was unanimous agreement among the participants that the mesoscopic world is a key frontier in science, and that the grand challenge in the field is establishing the existence or nonexistence of "mesoscopic protectorates", rules of organization of matter on this scale that transcend details. There was also agreement that sharp disciplinary boundaries are counterproductive to this end, and that great untapped opportunities lie at the interface between the cultures of physical and biological science.

In summary, this workshop brought leaders from the condensed matter, biological, and chemistry communities together to focus on the topic of mesoscopic organization in matter. Mesoscopic organization is the middle world between the regime of atoms and the quantum protection of stable states of matter. It is a scale at which neither first-principles computation nor measurement using existing technology sheds much light. There was unanimous agreement among the participants that the mesoscopic world is a key frontier in science, and that the grand challenge in the field is establishing the existence or nonexistence of "mesoscopic protectorates", rules of organization of matter on this scale that transcend details. There was also agreement that sharp disciplinary boundaries are counterproductive to this end, and that great-untapped opportunities lie at the interface between the cultures of physical and biological science

P. W. Anderson, Science, 177, 393 (1972).

R. B. Laughlin, Rev. Mod. Phys., 71, 863 (1999).

C. A. Angell, Science, 267, 1924 (1995).

P. G. Wolynes, J. N. Onuchic, and D. Thirumalai, Science, 267, 1619 (1995).

J. N. Onuchic et al, Proc. Natl. Acad. Sci., 92, 3626 (1995).

C. Geiger, et. al., Langmuir, 15, 2441 (1999).

V. Talanquer, D.W. Oxtoby, J. Chem. Phys. 106, 3673 (1997).

B.P. Stojkovic, et. al., Phys. Rev. Lett. 82, 4679 (1999).

F. Mezei, W. Knaak, B. Farago, Physica B, 156, 182 (1989).

R. Palmer et al, Phys. Rev. Lett., 53, 958 (1984).

J. Brooke, D. Bitko, T. F., Rosenbaum, G. Aeppli, Science, 284, 779 (1999).

S. Chakravarty and Hae-Young Kee, cond-mat 9908205.


Elihu Abrahams, Rutgers University

Gabriel Aeppli, NEC Research Institute

Kevin Bedell, Boston College

Sudip Chakravarty, University of California

Amitav Chattopadhyay, University of Maryland

George Gruner, University of California

Kevin Harrington, Stanford University

Robert Laughlin, Stanford University

Andrew Millis, Rutgers University

Jose Onuchic, University of California

David Oxtoby, University of Chicago

John Portman, University of Illinois

Douglas Scalapino, University of California

Joerg Schmalian, Rutherford Appleton Lab.

Robert Schrieffer, NHMFL/FSU

Yeong-ah Soh, NEC Research Institute

Devarajan Thirumalai, IPST

Benjamin Turek, Stanford University

John Wilkins, Ohio State University

Peter Wolynes, University of Illinois

David Alexander, LANL

Alexander Balatsky, LANL

Alan Bishop, LANL

Steven Conradson, LANL

Angel Garcia, LANL

P. Chris Hammel, LANL

Niels Gronbech Jensen, LANL

Stuart Maloy, LANL

Ferenc Mezei, LANL

Dirk Morr, LANL

David Morris, LANL

Don Parkin, LANL

David Pines, LANL

John L. Sarrao, LANL

Branko Stojkovic, LANL

Joe Thompson, LANL

Jill Trewhella, LANL

Timothy Weeks, LANL

David G. Whitten, LANL