Colloidal Semiconductor Nanocrystals: Tools For and Insights From First Principles Investigations

Abstract

Colloidal semiconductor nanocrystals, also known as quantum dots (QDs), are at once systems of great promise for diverse applications, systems with complex quantum physics that remain poorly understood at a fundamental level, and systems that are difficult to describe with conventional first principles computational chemistry methods due to their large size. Whereas QDs made from toxic materials like Cd and Pb are able to emit and absorb light efficiently and precisely, non-toxic alternative III-V QDs suffer from difficult to control defects that interfere with their radiative processes. Motivated primarily by the problem of understanding defects in III-V quantum dots, in this thesis we develop and apply new frameworks for the computational study of quantum dots. These frameworks include orbital localization techniques for de-convoluting ground-state band structures, a ∆SCF procedure for modeling the x-ray photoelectron spectra (XPS), and a low-cost machine learning framework for predicting Hamiltonians that can be used to extend molecular dynamics simulations. Through these frameworks, we reveal a more comprehensive picture of the defects that interfere with the optical performance of III-V quantum dots. We find that both three-coordinate indium and phosphorus can cause trap states in InP, explore how geometry and charge modulate trap depth, discover and explain a new source of four-coordinate trap states in InP and GaP, and assign shifts in P 2p XPS to specific III-V surface defects.