Electroluminescence from indirect band gap semiconductor ReS$_2$

TL;DR

This study uses ionic liquid gated FETs to determine that bulk ReS$_2$ is an indirect band gap semiconductor with a gap of 1.41 eV, and observes electroluminescence, suggesting potential for opto-electronic applications.

Contribution

First direct measurement confirming bulk ReS$_2$ as an indirect band gap semiconductor using FETs with high mobility.

Findings

Band gap of ReS$_2$ is 1.41 eV, smaller than the direct optical transition.

ReS$_2$ FETs exhibit electroluminescence in the ambipolar regime.

ReS$_2$ is confirmed as an indirect band gap semiconductor.

Abstract

It has been recently claimed that bulk crystals of transition metal dichalcogenide (TMD) ReS$_2$ are direct band gap semiconductors, which would make this material an ideal candidate, among all TMDs, for the realization of efficient opto-electronic devices. The situation is however unclear, because even more recently an indirect transition in the photoluminescence spectra of this material has been detected, whose energy is smaller than the supposed direct gap. To address this issue we exploit the properties of ionic liquid gated field-effect transistors (FETs) to investigate the gap structure of bulk ReS$_2$. Using these devices, whose high quality is demonstrated by a record high electron FET mobility of 1,100 cm$^2$/Vs at 4K, we can induce hole transport at the surface of the material and determine quantitatively the smallest band gap present in the material, irrespective of its direct or indirect nature. The value of the band gap is found to be 1.41 eV, smaller than the 1.5 eV direct optical transition but in good agreement with the energy of the indirect optical transition, providing an independent confirmation that bulk ReS$_2$ is an indirect band gap semiconductor. Nevertheless, contrary to the case of more commonly studied semiconducting TMDs (e.g., MoS$_2$, WS$_2$, etc.) in their bulk form, we also find that ReS$_2$ FETs fabricated on bulk crystals do exhibit electroluminescence when driven in the ambipolar injection regime, likely because the difference between the direct and indirect gap is only 100 meV. We conclude that ReS$_2$ does deserve more in-depth investigations in relation to possible opto-electronic applications.