Acceptor-based silicon quantum computing

A solid-state quantum computer with dipolar coupling between qubits is proposed. The qubits are formed by the low-lying states of an isolated acceptor in silicon. The system has the scalability inherent to spin-based solid state systems, but the spatial separation between the qubits is an order of magnitude larger. Despite strong dipolar inter-qubit coupling, the decoherence rate, as measured by electric dipolar echoes at an energy splitting of 1.5 GHz, is less than 1 kHz at low temperatures. For inter-acceptor distances of 100 nm and for modest microwave field amplitudes (50 V/cm) the clock frequency of the quantum computer is 0.1 GHz, which yields a quality factor of 10. This paper describes ideas for detection and operation of the quantum computer, and examines limitations imposed by noise sources. The concept of a solid-state quantum computer (QC) based on silicon has many attractive features. In comparison to a QC based on atoms, ions or liquid-state NMR, the anticipated integration with existing silicon device fabrication and nanoscale technology should provide overwhelming advantages. The major challenge is to identify a scalable system that has a sufficiently long decoherence time, can be controlled with high precision, and whose quantum state can be measured. Examples of systems based on silicon proposed to date include localized electron spins in semiconductor quantum dots(1-4); donor electron spins in Si/Ge heterostructures(5); a zero nuclear spin Si matrix with nuclear spin qubits on P donors interacting via overlap of localized electron wave functions(6, 7); chains of Si nuclei(8); and deep impurity states coupled with an optical field (9). Spin-based systems are natural candidates for quantum computers, as they take advantage of extremely long spin coherence times. However, their implementation is complicated by a number of issues yet to be resolved. For nuclear spins of donors, for example, interdonor separations should be less than 20 nm(6, 7). Execution of gate operations is relatively slow. Detection using a single-electron transistor appears feasible but the close donor separation makes fabrication highly demanding. These arguments suggest that there is a need to consider new alternatives for quantum computing in the condensed phase. In this paper we describe a scalable solid-state quantum computer with “spinless” qubits. The states of the qubit are the low-lying states of a substitutional group III acceptor in silicon. The proposed system has many features in common with the NMR systems in liquids and semiconductors. It has the scalability inherent to spin-based solid state systems. However, the required spatial separation between the qubits is an order of magnitude larger than in the impurity-spin system(6), which greatly simplifies fabrication. This is made possible by strong dipolar coupling between the qubits. Measurements described here indicate a decoherence rate of order 1 kHz for this system for qubits with level separation 1-2 GHz. The frequency of gate operations is controlled by the interaction between the qubits and by the Rabi frequency in a resonant microwave field. For inter-acceptor distances of 100 nm and for moderate microwave power, the clock frequency of the quantum computer is 0.1 GHz. This implies a quality factor of 10,