Nuclei in Semiconductors

Semiconductors are substances who electronic structure is such that they are neither good electrical conductors nor insulators, but whose conduction properties may be altered by various means, chemical and physical to produce materials which can do all sorts of amazing things. Things like transistors, and all the technology that comes from them, things like solar cells, things like cheap LED lighting, like lasers for BlueRay players - like a lot of the neat things that get developed by my colleagues in the Advanced Technology Institute here at the University of Surrey.

Usually, people interested in semiconductor materials are not terribly concerned with the nuclei that hold the electrons in place, except that the nuclei have to be the right element, say Silicon, in order to have the right number of electrons and so the right electron structure. It's not always the case, though. One cutting edge of semiconductor research involves using quantum "spins" to make quantum computers. Spin is a kind of quantum angular momentum - to do with things rotating - though in the quantum world things don't have to rotate to have angular momentum. In spin-based semiconductor research (or "spintronics"), one tries to manipulate the orientation of a spinning electron to store information, rather than by presence or absence of a charge. This is one of the promising ways of creating a quantum computer.

Once one starts to deal with electron spins, however, the nuclei can start becoming interested. Silicon is element number 14, with 14 protons in each nucleus, and 14 electrons around a neutral silicon atom. Silicon comes in three naturally occurring isotopes, though: Si-28 with 24 protons and 24 neutrons but also Si-29 and Si-30 with one and two extra neutrons respectively. Si-28 and Si-30 have no nuclear spin, so they don't interfere with spin-based quantum computers, but Si-29 (like all odd-numbered isotopes) has a non-zero nuclear spin, and their presence in naturally occurring silicon causes the quantum computer states to decay, or "decohere". The solution? Make isotopically-enriched silicon, without the natural Si-29. It turns out not to be that easy to either make a sample sufficiently isotopically pure, or even get rid of other contaminants, as a paper published last week, cited below, shows. Surprising sometimes where nuclear physics issues pop up.


Witzel, W., Carroll, M., Morello, A., Cywiński, L., & Das Sarma, S. (2010). Electron Spin Decoherence in Isotope-Enriched Silicon Physical Review Letters, 105 (18) DOI: 10.1103/PhysRevLett.105.187602