David Goldhaber-Gordon

Stanford University

Designer Hamiltonians:
Single-Electron Transistors as a Laboratory for Strongly-Correlated Electron Physics

Abstract: Physicists and material scientists have long designed materials with remarkable and varied electronic behaviors, from charge-density waves to high-temperature superconductivity. However, this approach has two important limitations: We often lack a microscopic understanding of the Hamiltonian of such systems, and we usually cannot tune (or even accurately measure) important parameters of the underlying Hamiltonian, such as energy levels and couplings. I propose a complementary strategy for obtaining and understanding new electronic phenomena, by building nanostructures out of electronically-simple semiconductors. In this talk, I will explore an initial example of this strategy.

The two-channel Kondo Hamiltonian is a prototype for non-Fermi-liquid behavior in correlated electron systems. A spin-1/2 local moment is coupled antiferromagnetically with equal strength to two independent reservoirs of conduction electrons. The two reservoirs both try to screen the local moment, with the result that neither fully succeeds: the moment retains entropy all the way down to zero temperature. Proposed experimental realizations of two-channel Kondo physics in ballistic metallic point contacts and heavy fermion alloys are intriguing but inconclusive, largely because we do not know and cannot tune the underlying Hamiltonian.

I will discuss a realization of the two-channel Kondo Hamiltonian in a semiconductor nanostructure. Using electrostatic gates, we should be able to tune continuously between two distinct Fermi-liquid regimes, which are characterized by different values of conductance through the nanostructure. Between these two regimes should reside the two-channel Kondo state. I will explore recent experimental progress and open questions.