Spatially Indirect Interfacial Excitons in c-Oriented n-ZnO/p-GaN Heterostructures

Zinc oxide (ZnO) with wide direct band gap (3.37 eV) and large excitonic binding energy (~60 meV) has significant prospects in UV optoelectronics. However, stable p-type doping is a big challenge. Recently, there are efforts to study ZnO/GaN np-heterojunctions as GaN, which is a semiconductor with a direct band gap of ~3.5 eV, has a very close lattice matching with ZnO and can be controllably doped p-type. A type-II band alignment at the interface is expected between the two materials. Moreover, the difference in spontaneous polarizations along the c-direction for the two semiconductors can result in a net positive polarization charge at the interface. We have shown theoretically that the combination of the two effects can lead to the formation of a unique type of spatially indirect excitons (SIDXs) at the n-ZnO/p-GaN interface. While the electron part of such excitons is quantum confined along c-direction in the ZnO side, the hole part stays at the GaN side of the junction[1]. Like interlayer excitons reported in homo-/hetero-bilayers of transition metal dichalcogenides (TMDs), these excitons have planer nature with finite dipole moment along c-direction. Unlike conventional direct excitons, SIDXs can interact repulsively with each other through dipole-dipole coupling, which should favour Bose-Einstein condensation (BEC). Though, BEC of the excitons has long been anticipated theoretically, its experimental demonstration remains to be elusive until now. Moreover, due to the spatial separation between the electrons and holes, SIDXs are expected to have lifetime much longer than that of the conventional direct excitons. Long lifetime means that the excitons can travel long distances before annihilation. The added advantage of the present system is that the lifetime of the SIDXs is expected to be controlled by applying bias across the junction. Note that the longer is the survival of the excitons, better would be the scope for controlling their spins. Long lifetime of the excitons as well as a control over their lifespan are essential for the development of exciton-based logic circuits, which are supposed to be better than the electronic devices in terms of the energy efficiency and compatibility with the optical communication. Experimentally, we investigate the behaviour of electroluminescence (EL) as a function of applied bias at different temperatures in pn-heterojunctions consisting of c-oriented n-ZnO layer grown epitaxially on p-type c-GaN/sapphire templates [1]. In certain samples, where the acceptor concentration in GaN is more than a critical value, an EL peak is found at low temperatures, which shows a redshift with increasing forward bias. Our study assigns this peak to the SIDXs. Binding energy of these excitons can be significantly increased by changing the applied bias and can even be made larger than that of ZnO bulk. We also study the time dynamics of electroluminescence (EL) in these heterostructures. EL is found to have an extremely slow rise (decay) time (hundreds of milliseconds). Our study finds a strong connection between the existence of SIDXs and the extraordinarily slow EL rise(decay) property observed in these structures. The effect can thus be attributed to the SIDXs, meaning that the lifetime of SIDXs in this system must be hundreds of milliseconds and the lifespan can be controlled by external bias. All these features make ZnO/GaN pn-heterojunction system a unique platform not only to study the physics of excitonic Bose-Einstein condensation but also to investigate the transport of the excitonic spin and its manipulation via electrical means towards the realization of exciton-based electronics.<br/>[1] S. Arora and S. Dhar, Appl. Phys. Lett. 122, 202102 (2023)