PhD Defence by Denys Igorevich Miakota

Title: Pulsed laser deposition of two-dimensional and quasi-one-dimensional transition metal dichalcogenides

Supervisors:
Principal supervisor: Senior Researcher Stela Canulescu, Department of Electrical and Photonics Engineering, Technical University of Denmark (DTU), Denmark
Co-supervisor: Prof. Arkady Yartsev, Lund University, Sweden

Evaluation Board
Associate Professor, Haiyan Ou, Department of Electrical and Photonics Engineering, DTU, Denmark
Professor Husam N. Alshareef, King Abdullah University of Science & Technology (KAUST), Saudi Arabia

Associate Professor David McCloskeyTrinity College Dublin, Ireland

Master of the Ceremony
Emeritus Jørgen Schou, DTU, Denmark

Abstract

2D semiconductors are among the most promising materials for next-generation electronic and photonic devices owing to their strong light-matter coupling and low thickness. However, the growth of 2D monolayer semiconductors at the wafer scale possesses many unsolved challenges. This thesis’s most scientifically relevant results are 2D transition metal dichalcogenides (TMDs) synthesis using pulsed laser deposition (PLD). The aim is advanced studies of 2D materials growth conditions in PLD, both in direct and two-step processes. The key challenges faced in 2D TMD crystal synthesis via direct deposition of chalcogenides in PLD are discussed, and a detailed study of the precursor influence in a two-step 2D TMDs synthesis is presented.

First, we synthesize 2D MoS2 monolayers at 700oC using direct PLD deposition and a MoS2 target. Then, we systematically study intrinsic defects in 2D TMDs grown in a single-step PLD synthesis using atomic resolution imaging and first-principles calculations. We use the MoS2 monolayer as a prototype material and show that sulfur vacancies, antisite defects, and mirror twin GBs with 4|8-membered ring motifs are predominant defects in PLD specimens. The intrinsic point defects are thermo-dynamically favourable under Mo-rich/S-poor conditions specific to the PLD-grown films.

Second, the synthesis of 2D MoSe2 monolayers using PLD is explored. The question is if it is possible to overcome the challenges faced by a single-step PLD using chalcogenide-rich targets. The formation of the intrinsic point defects in 2D MoSe2 monolayers can be slightly suppressed under Mo-poor/Se-rich conditions. Addition-ally, we find that the photoluminescence (PL) response of the PLD-grown films is dominated by A0 neutral excitons, unlike the PL for chemical vapour deposition (CVD) grown 2D MoSe2, which is dominated by negatively charged A− trions.

Third, the synthesis of 2D WS2 monolayers using a PLD-assisted CVD process was studied. Here, we explore how intrinsic oxygen vacancies present in a non-stoichiometric WO3−x solid oxide precursor lead to a more facile conversion from WO3−x to WS2 monolayers. A two-stage growth process was developed, employing tunability of the oxygen vacancies in uniform WO3−x precursors to control the nucleation, lateral growth independently, and ultimately, the WS2 domain size. This study suggests that native oxygen vacancies in the oxide films can serve as active sites through which sulfur atoms enter the lattice and facilitate the growth of WS2 crystals with high PL emission and large domain size. 

Fourth, we apply our findings from WS2 monolayer growth on the influence of the precursor composition and crystallinity to synthesize MoS2 monolayers in a similar PLD-CVD process. Oxygen vacancies in a reduced MoO3−x solid oxide precursor (0 < x < 1) lead to an easier conversion from MoO3−x to 2D MoS2. It was found that the resulting MoS2 monolayers demonstrate good optical response and strong photo-luminescence, and the PL signal for 2L MoS2 is dominated by negatively charged X− trions.

Fifth, based on the previous outcomes on individual 2D TMDs synthesis both in a single-step and a two-step processes, a comprehensive investigation on a few possible ways to grow 2D MoS2-WS2 heterostructures directly was performed. These results follow the advanced studies of the growth parameters in PLD and present a few possible routes to synthesize 2D MoS2-WS2 heterostructures. The heterostructures can be grown using bilayer MoOx/WOx oxide conversion and sequential synthesis. Here, different stages of the heterostructure synthesis are investigated, and an optical characterization of their various stacking types is presented. For all the cases, the PL emission from the heterostructures appears to be quenched as it emerges from the overlapping areas. The resulting films were classified as n-type/n-type heterostructures because the constituent semiconducting materials exhibit different bandgaps.

Sixth, the influence of the efficiency of various growth promoters, which aims to minimize 2D TMD growth temperature and maximize area coverage by inducing an intermediate liquid phase, was studied. It was found that upon certain growth conditions, vapour-liquid-solid (VLS) phase reaction leads to the formation of long single crystalline MoS2 quasi-1D nanoribbons. Notably, the alkali metal halide layer (NaF) acts as a promoter for the directional growth of quasi-1D MoS2 nanoribbons. Here, a possible growth mechanism of such MoS2 nanoribbons was suggested, and an intrinsic biaxial strain present in the nanoribbons was discussed. The demonstration of multilayer quasi-1D MoS2 nanoribbons opens new opportunities to potentially achieve nanoscale devices with novel functionalities.

In conclusion, a 2D materials synthesis was established, and a vast amount of exciting phenomena related to their synthesis were explored.