Multidimensional analysis of excitonic spectra of monolayers of tungsten disulphide: toward computer-aided identification of structural and environmental perturbations of 2D materials

Pavel V Kolesnichenko,Changxi Zheng,Michael S Fuhrer,Jeffrey A Davis,Qianhui Zhang

doi:10.1088/2632-2153/abd87c

Abstract

Despite 2D materials holding great promise for a broad range of applications, the proliferation of devices and their fulfillment of real-life demands are still far from being realized. Experimentally obtainable samples commonly experience a wide range of perturbations (ripples and wrinkles, point and line defects, grain boundaries, strain field, doping, water intercalation, oxidation, edge reconstructions) significantly deviating the properties from idealistic models. These perturbations, in general, can be entangled or occur in groups with each group forming a complex perturbation making the interpretations of observable physical properties and the disentanglement of simultaneously acting effects a highly non-trivial task even for an experienced researcher. Here we generalise statistical correlation analysis of excitonic spectra of monolayer WS2, acquired by hyperspectral absorption and photoluminescence imaging, to a multidimensional case, and examine multidimensional correlations via unsupervised machine learning algorithms. Using principal component analysis we are able to identify four dominant components that are correlated with tensile strain, disorder induced by adsorption or intercalation of environmental molecules, multi-layer regions and charge doping, respectively. This approach has the potential to determine the local environment of WS2 monolayers or other 2D materials from simple optical measurements, and paves the way toward advanced, machine-aided, characterization of monolayer matter.

Highlights

Since the realization of exfoliation of a single layer of graphite and confirmation of its extraordinary physical properties [1], a wave of efforts aiming at synthesizing other two-dimensional (2D) materials has naturally emerged
Absorption spectra are approximated here by differential reflectance [50, 51, 36] and feature two distinct peaks corresponding to spin-orbit split A- and B-exciton transitions occurring at K symmetry points in the first Brillouin zone [52]
Similar trends are observed in the spatial maps of PL emission (Figure 1d,e): the absorption and emission are blue-shifted in the regions spanning from the center of the flake towards its apexes

Summary

Introduction

Since the realization of exfoliation of a single layer of graphite (graphene) and confirmation of its extraordinary physical properties [1], a wave of efforts aiming at synthesizing other two-dimensional (2D) materials has naturally emerged. The spectra are fully parameterised, leading to a multi-dimensional parametric phase-space (hypercube) where a single data point represents the set of values corresponding to all parameters at a given spatial location on the monolayer sample This allows us to apply principal component analysis [44, 45, 46] (PCA) to identify the parameters that vary together and ideally combine to quantify specific perturbations and how they vary across the monolayer flake. By using unsupervised K-means clustering [47, 48, 49] of the data-points in this PCA-plane, regions of the sample with similar properties can be identified and provide further insight into how the perturbations combine and vary across the monolayer sample

Results and discussion

Conclusions

Sample preparation

Experimental realization

Electric-field assisted SNOMb PLa