Spatially separated, gate voltage controlled electron-hole plasma in semiconductor bilayers. Recent progress and challenges

Complete text of publication follows. Recent developments in solid-state semiconductor devices have made it possible to prepare a layer of electrons and a layer of holes in close proximity (∼ 10-20 nm), comparable to the excitonic Bohr radius of the host semiconductor (Gallium Arsenide). The physics of this system is driven by the attractive interlayer (electron-hole) Coulomb interaction and the in-plane repulsive interaction between the electrons (holes) themselves. But the binding energy scales (few meV) and the mass ratio of the positive and negative charges are very different from those in ionized Hydrogen plasmas. The electron and hole densities of this 2-component plasma can be individually controlled by gate voltages, set by the experimenter. Electrical current can be passed through each layer by independent ohmic contacts. The primary transport based tool for probing the interlayer scattering rate in these devices is the 'Coulomb drag' method, in which a current is passed through one layer and Coulomb interaction mediated momentum transfer to the other layer is measured. The interesting physics in these bilayers can be observed at T ∼ 1K or lower - readily accessible in liquid helium cooled cryostats. A rich phase diagram of the ground state of the electron-hole bilayer, consisting of excitonic phases[2], Charge density waves, Wigner crystals, excitonic condensates has been anticipated for many years. Recent data from two experimental groups have clearly shown that at low temperatures the interlayer scattering rate can no longer be correctly described by Fermi liquid theory. I will discuss the design concepts behind these devices, results of the ongoing experiments and several interesting open questions that have come up. Experimentally we have reached a strongly interacting regime, where the electron and hole densities can be tuned down to ∼ 5 x 1010cm-2 (rs ∼ 10 for holes, rs ∼ 2 for electrons) at an interlayer separation of 10-20 nm. This is a regime where the finite thickness of the wavefunctions are comparable to their separation and the intra-layer separation between the similarly charge particles are larger than the interlayer separation. The interlayer scattering and possible binding/density modulations must both be driven by the screened Coulomb interaction under these conditions. The behaviour of the dynamic bilayer dielectric screening ε(q; ω) in presence of disorder would be crucial in understanding the results.