Quantum dots, also known as artificial atoms, are created by tightly confining electrons, and thereby quantizing their energies. They are important components in the emerging fields of nanotechnology where their potential uses vary from dyes to quantum computing qubits. Interesting properties to investigate are e.g. the existence of atom-like shell structures and lifetimes of prepared states.

Stability and controllability are important properties in finding applications to quantum dots. The ability to prepare a state and change it in a controlled manner without it loosing coherence is very useful, and in some semiconductor quantum dots, lifetimes of up to several milliseconds have been realized. Here we focus on dots in semiconductor materials and investigate how the confined electrons are effected by their experienced potential.

The shape of the dot will effect its properties, and is important when considering a suitable model. Structures elongated in one dimension, often called nanowires, or shaped as rings have more one-dimensional characteristics than completely round or square dots. The two-dimensional dots investigated here are usually modeled as harmonic oscillators, however we will also consider circular well models.

The effective potential confining the electrons is investigated both in regard to how elliptical it is, as well as how results differ when using a harmonic oscillator or a circular well potential. By mixing spin states through spin-orbit interaction transitioning between singlet and triplet states becomes possible with spin independent processes such as phonon relaxation. We solve the spin-mixing two-electron problem numerically for some confinement, and calculate the phonon transition rate between the lowest energy singlet and triplet states using Fermi's golden rule.

The strength of the spin-orbit interaction is varied both by changing the coupling constants, and by applying an external, tilted, magnetic field. The relation between magnetic field parameters and dot parameters are used to maximize state lifetimes, and to model experimental results.