UC Berkeley

Excited carriers relax in semiconductors through various recombination pathways. Some of these processes can be radiative, where photons are created by carrier recombination. It is imperative to understand and increase the efficiency of these radiative processes for the betterment of many optoelectronic devices, such as light emitting diodes, lasers, photodetectors/photovoltaics, and solar cells. Two-dimensional (2D) semiconductors possess van der Waals bonding in the out-of-plane direction and have emerged as a promising material system for high-performance optoelectronic and electronic applications. Due to their reduced Coulomb interaction recombination dynamics of two dimensional semiconductors are significantly different from conventional bulk semiconductors. To achieve their full potential, it is crucial to understand and enhance radiative recombination in 2D semiconductors. Photoluminescence (PL) quantum yield (QY) is the ratio of the number of photons emitted as a fraction of the number of photons absorbed. PL QY governs the ultimate performance of optoelectronic devices and a key indicator of the efficiency of the radiative processes in a semiconductor. The room-temperature PL QY for most as-exfoliated monolayer 2D semiconductors are extremely poor; a prototypical 2D material such as monolayer MoS$_2$ has a QY of $~0.1$\%. Traditionally, this low QY has been attributed to the large defect density, as defects in conventional semiconductors drastically reduce their PL QY. In the first section of this thesis, I will show how the low QY in monolayer transition metal dichalcogenides (TMDCs) such as MoS$_2$ is not from defects but from background doping. I will show that the PL QYof as-processed MoS$_2$ and WS$_2$ monolayers reaches near-unity when they are made intrinsic through electrostatic or chemical counterdoping, without any chemical passivation. Surprisingly, neutral exciton recombination is entirely radiative even in the presence of a high native defect density. Most optoelectronic devices operate at high photocarrier densities, where all semiconductors suffer from enhanced multiparticle nonradiative recombination and 2D semiconductors are no exception. Although TMDC monolayers exhibit near-unity PL QY at low exciton densities, nonradiative exciton-exciton annihilation (EEA) rapidly degrades their PL QY at high exciton densities and limits their utility in practical applications. In the next section, we will discuss how by applying small mechanical strain (less than 1\%), we can markedly suppressed EEA in monolayer TMDCs, resulting in near-unity PL QY at all exciton densities despite the presence of a high native defect density. Next, I will discuss how these knobs of controlling the photophysics of 2D materials, such as strain and electrostatic doping effect other aspects such as exciton diffusion, electroluminescence and indirect-to-direct transition. As excitonic systems show robust radiative recombination even in the presence of defects, it is desirable to tune the exciton binding energy in the same material system without changing the defect density. However most 2D semiconductors becomes indirect when thickness is increased which obfuscates excitonic radiative recombination. In the final section I will explore exciton to free-carrier transition in black phosphorus (BP), as exceptional system that remains direct at all thicknesses. I will show in the excitonic regime the PL QY decreases with thickness and shows the highest PL QY of $\sim 20$\% when it is completely excitonic at the monolayer limit. When recombination is dominated by free carriers PL QY increases with thickness, and surface recombination velocity in BP is found to be two orders of magnitude lower than in passivated silicon: the most electrically inactive surface known to the modern semiconductor industry. The rich excitonic photophysics of monolayer 2D semiconductors have already garnered enormous amount of research interest. Even in thick 2D semiconductors where excitonic effects are absent, the recombination can be strikingly different from covalently bonded bulk semiconductors. My findings here will highlight the drastic difference in radiative recombination mechanisms in 2D semiconductors and can enable light-emitting devices that retain high efficiency at all brightness levels despite large defect density.