Use of the Indirect Photoluminescence Peak as an Optical Probe of Interface Defectivity in MoS₂

Abstract

technique which can give insight in the trap density inside the MoS 2 bandgap, while being non-invasive and fast. One of the factors that can strongly influence the performance of 2D materials is the substrate interaction. For both graphene and transition metal dichalcogenides (TMDCs), it has been shown [3-8] that amorphous dielectric substrates such as SiO 2 cause increased scattering, while enhanced performance is achieved when atomically flat 2D materials, such as hexagonal boron nitride (hBN), are employed as substrates. Performance, however, is often measured from devices, where mobility, hysteresis, or interface trap density (D IT) are assessed. This requires full device processing before the metrics can be acquired, where such processing can introduce non-idealities in the material which are difficult to deconvolve using electrical measurements. [9-12] This strongly supports the need for characterization techniques and metrics which can correlate material and interface quality before any processing takes place or/and between fabrication steps. For the case of graphene and carbon nanotubes, Raman spectroscopy is a powerful in-line quality assessment tool, with information on defectivity, flatness, thickness, etc. [13-15] The same fast and non-invasive quality metrics are not mature for the case of 2D TMDCs. One technique which can potentially fill this role is photoluminescence (PL) spectroscopy. Several works have used photoluminescence to study the properties of the interface between 2D TMDC and different substrates, with particular focus on the direct photoluminescence peak from monolayers, arising from direct radiative transitions from the conduction band to the valence band. [16-19] The direct photoluminescence peak of MoS 2 , around 1.8 eV, contains information on thickness, doping, and strain. [20-22] Additionally, direct exciton lifetime measurements can give insights on interface scattering and defect dynamics. [23,24] However, as thickness increases above the monolayer limit, the direct bandgap of MoS 2 transitions to an indirect bandgap. Notably, not much attention has been given to the peak originating from indirect bandgap transitions, present in thicker layers (⩾2 monolayers). [20] From a technological perspective, thicker layers have a series of advantages compared to monolayers: their lower volume-to-surface ratio makes them more tolerant towards ambient or process-induced degradation; [25] their smaller bandgap results in higher current density and lower contact resistance. [26-28] Here, we investigate the behavior Defect characterization of 2D materials is a critical aspect for their successful integration in future electronic devices. Here, a simple characterization technique is proposed that opens a path for fast, non-invasive, quality assessment of transition metal dichalcogenide layers, such as MoS 2 , and their interfaces. It relates to the correlation between substrate-induced traps and the indirect-to-direct photoluminescence peak ratio. It is shown that the indirect peak is quenche...

Trap Density Assessment on Multilayer WS2 using Power-Dependent Indirect Photoluminescence

“…Whereas WS 2 thicker than 2ML is a semiconductor with an indirect gap, the direct transition can still be excited by the 2.33 eV laser used in this study, thus resulting in the spectra observed in Fig.1 d-f). The relatively high indirect PL intensity of the selected flakes indicate their high material/interface quality[19].Upon confirming that the selected flakes present a distinct indirect photoluminescence peak, PL spectra are collected at different excitation intensities. Fig.2shows the typical normalized power-dependent PL characteristics of a WS 2 flake.…”

“…1 d-f). The relatively high indirect PL intensity of the selected flakes indicate their high material/interface quality [19].…”

“…The indirect photoluminescence, on the other hand, is strongly sensitive to defectivity, due to the phononassisted momentum conservation requirements which makes the radiative recombination less efficient. Indeed, for the indirect PL process, the optical indirect transitions strongly compete with non-radiative trap-assisted recombination, making the indirect PL peak a less ambiguous probe of material/interface traps [19]. Furthermore, when analyzing both direct and indirect PL peaks, the indirect-to-direct intensity ratio (I ind /I dir ) provides a simple quality metric which largely eliminates the effect of factors difficult to account for (e.g.…”

“…Increasing the photoexcited carrier concentration enhances defect-assisted Auger recombination, the main non-radiative recombination mechanism in 2D TMDCs [20,23]. This, in turn, reduces the I ind /I dir ratio proportionally to the defect density, since the indirect photoluminescence peak is more sensitive to non-radiative recombination [19]. This is experimentally achieved by modulating the incident laser power density and collecting the photoluminescence spectra, from which the I ind /I dir ratio is calculated.…”

A microprocessor-aided platform enabling surface differential reflectivity and reflectance anisotropy spectroscopy

“…Nowadays, light probes are successfully applied in semiconductor technology [1][2][3], surface physics [4][5][6], chemical detection [7][8][9], biological analysis [10][11][12], etc. More specifically, the surface and interface physics raised great interest in the optical characterization of layered systems [13][14][15] and molecular films [16][17][18].…”