We show that band-structure effects are likely to prevent superfluidity in semiconductor electron-hole double-layer systems. We suggest the possibility that superfluidity could be realized by the application of uniaxial pressure perpendicular to the electron and hole layers. ͓S0163-1829͑98͒51712-2͔The possibility of realizing a superconducting condensate of electron-hole pairs in a system consisting of two spatially separated layers of electrons and holes was suggested some time ago.1 Only recently, however, has it become feasible [2][3][4][5] to produce systems where the electrons and holes are close enough to interact strongly, and, at the same time, sufficiently isolated to inhibit optical recombination in nonequilibrium systems and tunneling between electron and hole bands. Since the overlap of the electron and hole wave functions in these systems can be made negligibly small, the joint motion of condensed electron-hole pairs turns out to be superfluid; antiparallel currents can flow in the two layers without dissipation. 1,6 Although the electron-hole condensation temperature has been predicted to be in an accessible range, and signatures of its occurrence have been discussed in the literature, 7,8 compelling evidence of a superfluid state is yet to appear. In this paper we propose a strategy for the realization of electron-hole superconductivity in double well systems. We point out that at high sufficiently densities, the anisotropy of the hole band in realistic wells is a major obstacle to the occurrence of superconductivity. We propose that the application of a moderate uniaxial stress (ϳ10 kbar) could reduce the anisotropy enough to permit the formation of a condensate.Microscopic theories of superfluidity in electron-hole liquids have usually been developed in the framework of a simple mean field theory 9 similar to the BCS theory of superconductivity. Recently, detailed numerical solutions of the BCS gap equation have been obtained for models of epitaxially grown double-layer structures. [10][11][12] We are interested in the high carrier density regime for which the underlying fermionic degrees of freedom of electrons and holes play an essential role in the pairing physics, and mean-field theory estimates of transition temperatures can be reliable.
13Indeed, recent variational 14 and diffusion 15 Monte Carlo calculations of the ground-state energy of an electron-hole double layer appear to qualitatively confirm BCS theory predictions for the dependence of the zero temperature gap on interlayer separation, provided that the attractive electronhole interaction is appropriately screened in estimating the BCS theory coupling constant. Although transition temperatures calculated with unscreened interactions ͑as high as 10 K with typical parameters͒ are expected to be overestimates, the naive expectation from these calculations is that the superfluid state should be within reach.An aspect of the problem which is potentially important at high densities, and to which little attention has been paid thus far, i...