Indirect-to-direct band-gap crossover in few-layer MoTe$_2$

TL;DR

This study investigates the evolution of the electronic band structure in few-layer MoTe₂, revealing a crossover from direct to indirect band gaps as the number of layers increases, which differs from other TMDs.

Contribution

It provides the first detailed analysis showing that mono and bilayer MoTe₂ are direct band-gap semiconductors, with a transition to indirect gaps in thicker layers, challenging previous assumptions about TMDs.

Findings

Mono and bilayer MoTe₂ are direct band-gap semiconductors.

Trilayer MoTe₂ has nearly equal direct and indirect gaps.

Tetralayer MoTe₂ exhibits an indirect band gap.

Abstract

We study the evolution of the band-gap structure in few-layer MoTe$_2$ crystals, by means of low-temperature micro-reflectance (MR) and temperature-dependent photoluminescence (PL) measurements. The analysis of the measurements indicate that, in complete analogy with other semiconducting transition metal dichalchogenides (TMDs), the dominant PL emission peaks originate from direct transitions associated to recombination of excitons and trions. When we follow the evolution of the PL intensity as a function of layer thickness, however, we observe that MoTe$_2$ behaves differently from other semiconducting TMDs investigated earlier. Specifically, the exciton PL yield (integrated PL intensity) is identical for mono and bilayer and it starts decreasing for trilayers. A quantitative analysis of this behavior and of all our experimental observations is fully consistent with mono and bilayer MoTe$_2$ being direct band-gap semiconductors, with tetralayer MoTe$_2$ being an indirect gap semiconductor, and with trilayers having nearly identical direct and indirect gaps.This conclusion is different from the one reached for other recently investigated semiconducting transition metal dichalcogenides, for which only monolayers are found to be direct band-gap semiconductors, with thicker layers having indirect band gaps that are significantly smaller, by hundreds of meV, than the direct gap. We discuss the relevance of our findings for experiments of fundamental interest and possible future device applications.