News & Views | Published:

Solid-state physics

Super silicon

Naturevolume 444pages427428 (2006) | Download Citation


Silicon is the archetypal semiconductor, and base material of the microelectronic age. But it turns out that, treated the right way, silicon the semiconductor can become silicon the superconductor.

Window on the world: silicon is arguably the material most central to modern life. Credit: D. HALLINAN/ALAMY

If someone were to stop me in the street and ask me to name the most important materials on Earth, I would say concrete, steel, glass and silicon. To witness the importance of the first three, just look up from your page or screen. For the last, close your eyes and imagine yourself back in the BS (before-silicon) world of, say, Myrna Loy in The Thin Man or Humphrey Bogart in Casablanca. One might reasonably argue a preference for the softer focus of those earlier times; but the differences in lifestyle between then and now make it hard to argue against the assertion that silicon has become the technologically most important material of the past 50 years.

It is for this reason that Bustarret and colleagues' report (page 465 of this issue)1 is such a breakthrough: they have succeeded in turning silicon, the consummate semiconductor, into a superconductor at ambient pressure. Admittedly, the treatment they meted out to silicon to force its conversion ('doping' with high levels of boron) can only be termed abusive, and the temperature at which they measured it (0.3 degrees above absolute zero) frigid. That raises questions as to how useful this turncoat-silicon might be; but its existence is impressive in its own right.

Ironically, the chemical characteristic of silicon that makes it so useful as a semiconductor turns out to be the same characteristic that had stopped it from being made into a superconductor: it generally allows only very small amounts of other elements to be incorporated into its solid form. Silicon can therefore be made extremely pure relatively straightforwardly. Pure silicon is the ultimate semiconductor: it has only a tiny number of free electrons, or positively charged 'holes' (equivalent to electrons missing from the crystal lattice), available to carry electrical current, and so it is effectively electrically insulating. This characteristic can be changed by introducing into the pure silicon small concentrations of atoms that have either one electron more than silicon, such as phosphorus, or one electron fewer, such as boron. This procedure — known as doping — is used to control the number of electrons or holes in carefully patterned regions. Manipulation of these extra charge carriers is the basis of the myriad products based on microelectronics that we know and love.

Over the past decades, doped silicon has proved the perfect platform for testing ideas about the difference, at a fundamental level, between electrically insulating and electrically metallic states of matter (see refs 2,3 and references therein). It is right at the crossover between these two states that superconductivity — when electrons flow without encountering resistance — often occurs.

Here is the gist of what has been found so far. When pure silicon is doped, at low temperatures, with small concentrations of phosphorus — that is, phosphorus atoms replace silicon atoms in the crystal lattice — the extra electron that does not form a crystal bond is localized around the phosphorus nucleus, which has an extra positive charge. Electron and nucleus form a bound state, the one 'orbiting' the other. As the number of phosphorus dopants increases, the nuclei eventually become so close that their electron orbitals overlap, forming one extended orbital. Once that happens, metallic conductivity occurs. This state persists even down to very low temperatures, despite the disorder caused by the random substitutions of dopant for silicon atoms4. The same effect applies to a boron dopant5; here, however, the nucleus is negatively charged and the carriers are positively charged.

In the early experiments, silicon was doped with sufficient numbers of electrons and holes to make it a metallic conductor, and the properties of the resulting materials were carefully characterized down to temperatures below 1 kelvin. But superconductivity was never observed.

Bustarret et al.1 adopt a new approach. They dope silicon with boron, to a level well beyond that possible with normal synthesis techniques, by using a pulsed laser to heat a thin silicon film on which a layer of boron-containing gas has been adsorbed. These high-intensity laser blasts cause the boron atoms to invade the silicon, where they freeze in place before the silicon can push them out again. Strange though the resulting structure might be, superconductivity is clearly observed in the treated silicon, turning on at a temperature of 0.3 K.

Performing experiments using such high-powered lasers, and testing materials for superconductivity at such low temperatures, is no small matter. So why bother? The authors are motivated by the possibility that, if silicon could be made superconducting — even under conditions too extreme to be useful in practical devices — the integration of superconducting silicon into the sophisticated world of microelectronics processing might uncover new electronic functions. It will be interesting, for example, to see whether an electron-rich superconductor can be made out of silicon through extreme doping with electron-rich phosphorus or arsenic, rather than hole-rich boron. That would allow the gamut of microelectronics concepts and processing to be applied to superconductors, but is far from an obvious extension of the present work.

So, are Bustarret and colleagues' results1 just an amusing diversion in the search for new superconductors, or a herald of more and better devices and materials? It's too early to tell. The main thrusts in the search for new superconducting materials nowadays are towards exotic systems in which magnetism can be transformed into superconductivity by changing a carefully controlled experimental parameter, and towards materials based on metallic elements that have high superconducting transition temperatures and are easy to process. Such a material could change the way electrical current is carried. I remain convinced that we could wake up any morning to the announcement that someone, somewhere, has found it. If that superconductor is made by doping concrete, I'll know it's time for me to retire.


  1. 1

    Bustarret, E. et al. Nature 444, 465–468 (2006).

  2. 2

    Lee, P. A. & Ramakrishnan, T. V. Rev. Mod. Phys. 57, 287–337 (1985).

  3. 3

    Thomas, G. A. Phil. Mag. B52, 479–498 (1985).

  4. 4

    Rosenbaum, T. F. et al. Phys. Rev. Lett. 45, 1723–1726 (1980).

  5. 5

    Dai, P., Zhang, Y. & Sarachick, M. P. Phys. Rev. Lett. 66, 1914–1916 (1991).

Download references

Author information


  1. the Department of Chemistry, Princeton University, Princeton, 08544, New Jersey, USA

    • Robert J. Cava


  1. Search for Robert J. Cava in:

About this article

Publication history


Issue Date


Further reading


By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.

Newsletter Get the most important science stories of the day, free in your inbox. Sign up for Nature Briefing