Substrate-induced strain and charge doping in CVD-grown monolayer MoS2

Abstract

Due to its electronic-grade quality and potential for scalability, two-dimensional (2D) MoS2 synthesized by chemical vapor deposition (CVD) has been widely explored for electronic/optoelectronic applications. As 2D MoS2 can be considered a 100% surface, its unique intrinsic properties are inevitably altered by the substrate upon which it is grown. However, systematic studies of substrate-layer interactions in CVD-grown MoS2 are lacking. In this study, we have analyzed built-in strain and charge doping using Raman and photoluminescence spectroscopy in 2D MoS2 grown by CVD on four unique substrates: SiO2/Si, sapphire, Muscovite mica, and hexagonal boron nitride. We observed decreasing strain and charge doping in grown MoS2 as the substrates become less rough and more chemically inert. The possible origin of strain was investigated through atomic force microscopy roughness measurements of the as-grown layer and substrate. Our results provide direction for device optimization through careful selection of the growth substrate and pave the way for further investigations to unravel the complex nature of the 2D monolayer-substrate interface.

Photoluminescence Study of B‐Trions in MoS2 Monolayers with High Density of Defects

“…In addition, we noticed a faint continuous background emission and it was subtracted during further analysis. All PL bands are red‐shifted with respect to peak positions of high quality exfoliated monolayers and this is typical for CVD grown strained MoS 2 on SiO 2 /Si substrates . At the same time, we did not observe any features corresponding to indirect transition usually visible in multilayers at about 1.5 eV .…”

“…As it was shown in refs. , the reduced electron concentration leads to a blue shift of the peak compared with the peak position in exfoliated high quality monolayer. It was also shown that the nitrogen doping could be an effective way to produce hole doping in MoS 2 monolayers by creating N S acceptor defects.…”

“…The separation between these peaks is Δ = 23.3 cm −1 and this value is higher than usually observed separation in high quality monolayers (Δ ≈19 cm −1 ). Vertical‐dashed lines in Figure mark the peak positions of high quality exfoliated monolayer and, with respect to these lines, our CVD‐grown MoS 2 has peaks that are either softened ( E ′) or stiffened (), which can be attributed to the presence of substrate‐induced tensile strain and charge doping . The origin of tensile strain in CVD‐grown MoS 2 layers has been attributed to the mismatch of the thermal expansion coefficient between the layer and a SiO 2 /Si substrate or/and to the surface roughness.…”

“…The actual peak position of A and B bands depends on type of the substrate and on quality of monolayers. It has been shown that monolayers grown by CVD on SiO 2 /Si substrates usually show PL bands at lower energy than in exfoliated monolayers and this is caused by intrinsic tensile strain in MoS 2 monolayer . The strain and the presence of lattice defects affect not only PL bands, but also the position of Raman peaks of monolayer .…”

“…It has been shown that monolayers grown by CVD on SiO 2 /Si substrates usually show PL bands at lower energy than in exfoliated monolayers [17,18] and this is caused by intrinsic tensile strain in MoS 2 monolayer. [19,20] The strain and the presence of lattice defects affect not only PL bands, but also the position of Raman peaks of monolayer. [20][21][22][23] It has been shown that the A band at room temperature is often a sum of neutral (A 0 ) and charged (A À ) exciton emissions.…”