Neutron Scattering and Magnetism
Laboratory for Solid State Physics · ETH Zurich

Research

Some (arguably, material scientists) study the physics of new materials. We, as physicists, are more interested in new physics of materials. We hardly ever start with a material. We start with an interesting theoretical concept (e.g. confinement), model (e.g., the quantum XXZ model on a triangular lattice) or phenomenon (e.g., dielectric relaxation by Bose-Einstein condensing magnons).

We then look for materials where these can be studied experimentally. For the examples above it would be, correspondingly, the triangular lattice antiferromagnet Cs2CoBr4, where fractional excitations become confined into a "Zeeman ladder" of bound states [1]; the cobaltate K2Co(SeO3)2, which is an excellent realization of that model and in applied fields even becomes a magnetic analog of a supersolid [2], or the quantum multiferroic Cs2Cu2Mo3O12, where dipolar degrees of freedom relax through the soft mode of Bose-Einstein condensing magnons [3]. For us, materials are no more than a means to the end of studying the underlying fundamental physics. But they are also a validation of that this physics is real and not just a form of mathematics.

We are primarily interested in complex collective quantum behavior (quantum phases and excitations) that emerges in simple systems. This is the main reason why our focus is on magnetic insulators. These materials may be chemically complex, but the relevant degrees of freedom are perfectly well defined: the spins of the magnetic ions, often just one or two per crystallographic unit cell. The interactions between them are quite transparent as well: Heisenberg or Ising couplings, typically between nearest-neighbor spins only. The resulting models contain just a few parameters, and these can usually be determined experimentally, leaving no unknowns in the problem. And yet, these embarrassingly simple models and the quantum magnets that realize them demonstrate the most amazing collective behavior: quantum liquids, excitations with fractional quantum numbers, Majorana fermions, a zoo of the most bizarre symmetry-broken phases, states with non-trivial topology, quantum phase transitions of every stripe, etc.

There is another reason why quantum magnets for our purposes are: observables. We have experimental techniques that couple to the relevant degrees of freedom directly. This giving us experimental access to correlation and response functions that are key to understanding emergent collective behavior. The most important of these techniques is neutron scattering , which can probe spin correlations not only in time (frequency) but also in space (momentum), and do so on a level allowing a direct quantitative comparison with theory.