Silicon in the Quantum Limit: Quantum Computing and Decoherence in Silicon Architectures

TL;DR

This paper explores the potential of silicon-based quantum architectures, analyzing spin coherence and relaxation, and proposing a new spin readout scheme to advance silicon quantum computing.

Contribution

It provides a detailed analysis of spin relaxation in silicon, including effects of phonons and spin-orbit coupling, and introduces a novel spin readout method with automatic initialization.

Findings

Silicon and strained silicon exhibit excellent spin coherence properties.

The complex band structure of silicon influences spin relaxation mechanisms.

A new spin readout scheme with automatic initialization is proposed.

Abstract

Semiconductor architectures hold promise for quantum information processing (QIP) applications due to their large industrial base and perceived scalability potential. Electron spins in silicon in particular may be an excellent architecture for QIP and also for spin electronics (spintronics) applications. While the charge of an electron is easily manipulated by charged gates, the spin degree of freedom is well isolated from charge fluctuations. Inherently small spin-orbit coupling and the existence of a spin-zero Si isotope facilitate long single spin qubit coherence times. Here we consider the relaxation properties of localized electronic states in silicon due to donors, quantum wells, and quantum dots, including effects due to phonons and Rashba spin-orbit coupling. Our analysis is impeded by the complicated, many-valley band structure of silicon and previously unaddressed physics in silicon quantum wells. We find that electron spins in silicon and especially strained silicon have excellent decoherence properties. Where possible we compare with experiment to test our theories. We go beyond issues of coherence in a quantum computer to problems of control and measurement. Precisely what makes spin relaxation so long in semiconductor architectures makes spin measurement so difficult. To address this, we propose a new scheme for spin readout which has the added benefit of automatic spin initialization, a vital component of quantum computing and quantum error correction. Our results represent important practical milestones on the way to the design and construction of a silicon-based quantum computer.