December 1998 ICAM White Paper
Professor Zachary Fisk
Florida State University - NHMFL
1800 East Paul Dirac Drive, B223
Tallahassee, FL 32306-3016
A broad category of rare earth and transition metal materials owe their interesting properties to the physics and chemistry of mixed (or intermediate) valence, in which the d- or f-shell occupation of a constituent atom is non-integral. These properties involve, for example, metal/insulator transitions, magnetic/non-magnetic behavior, high Tc superconductivity and catalytic behavior. Mixed valence is an important problem of long standing, and we can express our ignorance concerning it by stating that we have no capability whatsoever to predict how it will play out in any given situation. The essential non-linearity here from the ICAM perspective is the non-linearity of properties with respect to small variations in the chemistry of a material.
We note first that not all-crystal structures are equally interesting: certain structures are host to a broad variety of interesting properties, other structures remain in general bland. While very little is known (surprisingly) concerning structure/property relationships, a working hypothesis is that certain structures have a generic band structure favorable for the playing out of enhanced correlated electron behavior. Our idea is that there is a fairly sizeable group of materials which has an inherent closed-shell electronic structure that both gives the material its stability and provides an inert background against which the strongly correlated behavior of atoms containing active and partially filled inner electronic shells can dominate the electric and magnetic properties.
Further, we suggest that the fixed-point case of the so-called Kondo insulators provides the origin from which to think about the entire problem of mixed valence. For these, there is good reason to believe that exactly one occupied magnetic orbital interacts with a single half-filled electron band: nothing else is relevant. The ground state for such materials is a non-magnetic insulator, our closed shell structure, arrived at through a highly non-linear process. We make the important point that these materials are very much mixed valent, and our contention is that the essential physics and chemistry of mixed valent materials will always involve, albeit often hidden, Kondo insulator behavior.
A rational program to discover new classes of unusual materials with properties driven by mixed valence can involve bringing the more intuitive bonding ideas from chemistry to bear on the many body ideas of physicists. We believe there is a particular type of electronic structure in certain solids which favors the mixed valent state, and that this electronic structure can be evident in the detailed chemical understanding of bonding requirements in intermetallic compounds. The point to emphasize is that intermetallic structures are not in general approached from the standpoint of chemical bonds, and the metallic bond is in fact quite peripheral to most chemistsí interests. We believe there is a great deal to be gained by bringing this thinking to bear on the mixed valence problem and the search for new materials.