Can single-electron integrated circuits and quantum computers be fabricated in silicon?

Single-electronics and quantum computers have tremendous potential, but obstacles to fabricating them are enormous. Here we consider the physical difficulties in terms of both circuit architecture and plausible future advancements in silicon technology. Our discussion will focus on a set of recent proposals which involve tunnelling between 2D arrays of Coulomb islands or ‘quantum dots’. Planar architectures of this type can potentially be realized through in situ e-beam patterning of self-ordered dopant precursor molecules on hydrogen-terminated silicon surfaces. Low-temperature molecular beam epitaxy would then be used to overgrow them into Si/SiGe heterolayers. Epitaxial structures of this kind could potentially eliminate the random polarization offsets which plague all single-electron devices today, hopefully to levels permitting large-scale (fault-tolerant) integration. Individual donor atoms might be incorporated via STM lithography to fabricate the critical elements of a quantum computer. This approach offers many potential advantages, but the degree of material perfection it requires will be very challenging.