Extraordinarily conventional

Article metrics

Attitudes to high-temperature superconductivity have swung from disbelief to a conviction that it occurs only 'unconventionally'. But conventional superconductivity is now reported at record high temperatures. See Letter p.73

In 1911, the physicist Heike Kamelingh Onnes was puzzled to observe1 that mercury became an ideal conductor below 4.2 kelvin. How could all the electrons in a metal cooperate so as to carry electric current without resistance? Common wisdom dictates that there is nothing ideal in this world. Nobody's perfect! No crystals without defects can be created, no wheel can roll without friction, no glass can be 100% transparent. Yet subsequent experiments confirmed that the resistivity of many metals suddenly drops to exactly zero at a sufficiently low temperature. Chaotic motion introduced by heat destroys electronic cooperation, so for many years it was believed that this phenomenon, now known as superconductivity, is limited to ultra-low temperatures. But on page 73 of this issue, Drozdov and colleagues2 report a superconductor that works at about 200 K — a temperature that actually exists on Earth's surface.

For decades, physicists were in the dark about the origin of superconductivity. The discovery in 1938 of another cooperative phenomenon3 — superfluidity in helium at 2 K — offered the first clue4. This complete lack of viscosity turned out to be a direct consequence of quantum mechanics. All quantum particles are characterized by a 'spin' number; if this is an integer (as for helium-4 atoms), the particles can combine into a single object so large that it cannot be disturbed by such nuisances as friction or viscosity. This effect is called Bose–Einstein condensation.

But electrons, which conduct electricity, have a spin of ±½, and so are not subject to Bose–Einstein condensation. In 1957, John Bardeen, Leon Cooper and Bob Schrieffer therefore proposed that the interaction of electrons with metal ions creates an attractive interaction that forces the electrons to combine in pairs5. These 'Cooper pairs' have a net spin of zero, and can form a Bose–Einstein condensate. This theory also allows the transition temperature, Tc, below which superconductivity occurs for a given metal, to be estimated.

Most elemental superconductors had been discovered by 1957, and all had Tc values of less than 10 K. For the next two decades, scientists worked with various compounds, but failed to increase Tc by even a factor of three. Not surprisingly, most physicists began to believe that nature imposes a fundamental, but as-yet unexplained, Tc limit of 25–30 K. The problem was succinctly formulated by the materials scientist Bernd Matthias in 1964: “Why has it been relatively easy, within the last 10 years, to reach transition temperatures of 17 to 18 K in many intermetallic systems and impossible to raise this value even by as little as half a degree?”6 Eight years later, Marvin Cohen and Phillip Anderson pointed out that if electrons interact too strongly with the ions in a metal, they can break the lattice apart7. On this basis, they estimated that the highest Tc for conventional superconductors (those driven by the electron–ion interaction) is approximately 30 K.

Although the argument seemed convincing, some physicists remained hesitant. In the early 1970s, Vitaly Ginzburg — one of the top theorists of the time — organized a group in Moscow to explore routes to high-temperature superconductivity. One of his team's principal results was that a key assumption by Cohen and Anderson was flawed, and that Tc could, in principle, be arbitrarily high even in a conventional superconductor8.

Another prominent physicist who did not subscribe to the idea of a universal limit was Neil Ashcroft. In the late 1960s, he9 and Ginzburg10 proposed that, if hydrogen could become metallic, the energy of its ionic vibrations would be so high that even a moderately strong electron–ion coupling could result in a rather high Tc. Unfortunately, metallization of hydrogen has proved to be extremely difficult. It was then pointed out that hydrogen-rich compounds might be better targets11,12, but it is only now that this idea has been realized, as reported by Drozdov and colleagues.

In the meantime, three major breakthroughs occurred in superconductivity. First, cuprate superconductors were discovered13 in 1986; within seven years, the Tc for these compounds reached 133 K (ref. 14). These have been recognized as 'unconventional' superconductors, driven by interactions among electrons, rather than by electron–ion interactions.

The second was the discovery15, in 2001, of magnesium diboride — a conventional superconductor whose Tc is 40 K. This relatively high number is due to the low mass of boron, and to the fact that strong electron–ion coupling is ensured because the conducting electrons come from the boron, and the boron ions form a rigid sublattice. The physics of superconductivity in magnesium diboride turned out to be considerably more complex than for other conventional superconductors known at the time, but was understood within a year of the original discovery. At last, theorists could accurately calculate the critical temperature of a rather complicated material. This encouraged scientists to seek quantitative predictions for new superconducting materials.

The third breakthrough was the discovery16 of iron-based superconductors in 2008. These materials seem to be unconventional and, although of great interest, have never surpassed the Tc of the cuprates.

Drozdov and co-workers report a fourth breakthrough: superconductivity at approximately 200 K in a hydrogen-rich compound, sulfur hydride, at about 90 gigapascals — a pressure hardly achievable just a few years ago. Not only is this a 50% increase over the previous record for Tc, but the authors convincingly argue that the observed superconductivity is conventional, vindicating the ideas of Ashcroft and Ginzburg.

“A previously unknown material predicted to be a high-temperature superconductor has been experimentally confirmed to be one.”

Moreover, this is the first time that a previously unknown material predicted to be a high-temperature superconductor has been experimentally confirmed to be one. A computational study17 of hydrogen-rich materials under pressure had reported that the sulfur hydride H3S would be a superconductor with Tc in the range 190–200 K at 200 gigapascals — very close to the now-reported experimental value. Drozdov et al. studied H2S, but it seems that at high pressure this decomposes into elemental sulfur and hydrogen-rich H3S. It is therefore highly likely that the superconducting material is H3S. More-accurate calculations18,19,20 yielded a Tc value approximately 20% higher than in the earlier computational study17. There is some disagreement about which small effects, not accounted for in standard computations, are responsible for this overestimate, but it is amazing that theorists quibble about a 20% inaccuracy in first-principles calculations when even an order-of-magnitude estimation was considered practically impossible only 40 years ago.

In 1796, the philosopher Wilhelm Hegel introduced the concept of spiral progress: an intellectual proposition is superseded by its negation, but later the negation itself is negated; the original thesis is then reinstated, but at a higher level of development. The generality of this concept can be philosophically disputed, but Hegel's idea seems to be confirmed by the fact that the holy grail of superconductors has been discovered in the same group of materials as the first known superconductors, after a tiresome quest along exotic routes.

Footnote 1


  1. 1.

    See all news & views


  1. 1

    Kamerlingh Onnes, H. Commun. Phys. Lab. Univ. Leiden No. 120 (1911).

  2. 2

    Drozdov, A. P., Eremets, M. I., Troyan, I. A., Ksenofontov, V. & Shylin, S. I. Nature 525, 73–76 (2015).

  3. 3

    Kapitza, P. Nature 141, 74 (1938).

  4. 4

    London, F. Nature 141, 643–644 (1938).

  5. 5

    Bardeen, J., Cooper, L. N. & Schrieffer, J. R. Phys. Rev. 108, 1175 (1957).

  6. 6

    Anderson, P. W. & Matthias, B. T. Science 144, 373–381 (1964).

  7. 7

    Cohen, M. L. & Anderson, P. W. in Superconductivity in d- and f-Band Metals (ed. Douglass, D. H.) 17–27 (American Inst. Physics, 1972).

  8. 8

    Dolgov, O. V., Kirzhnits, D. A. & Maksimov, E. G. Rev. Mod. Phys. 53, 81 (1981).

  9. 9

    Ashcroft, N. W. Phys. Rev. Lett. 21, 1748 (1968).

  10. 10

    Ginzburg, V. L. J. Stat. Phys. 1, 3–24 (1969).

  11. 11

    Ginzburg, V. L. & Kirzhnits, D. A. (eds) High-Temperature Superconductivity Chapter 1 (Consultants Bureau, 1982) [transl. from Russian].

  12. 12

    Ashcroft, N. W. Phys. Rev. Lett. 92, 187002 (2004).

  13. 13

    Bednorz, J. G. & Müller, K. A. Z. Physik B 64, 189–193 (1986).

  14. 14

    Schilling, A., Cantoni, M., Guo, J. D. & Ott, H. R. Nature 363, 56–58 (1993).

  15. 15

    Nagamatsu, J., Nakagawa, N., Muranaka, T., Zenitani, Y. & Akimitsu, J. Nature 410, 63–64 (2001).

  16. 16

    Kamihara, K. Y., Watanabe, T., Hirano, M. & Hosono, H. J. Am. Chem. Soc. 130, 3296–3297 (2008).

  17. 17

    Duan, D. et al. Sci. Rep. 4, 6968 (2014).

  18. 18

    Errea, I. et al. Phys. Rev. Lett. 114, 157004 (2015).

  19. 19

    Flores-Livas, J. A., Sanna, A. & Gross, E. K. U. Preprint at (2015).

  20. 20

    Akashi, R., Kawamura, M., Tsuneyuki, S., Nomura, Y. & Arita, R. Phys. Rev. B 91, 224513 (2015).

Download references

Author information

Correspondence to Igor I. Mazin.

Related links

Related links

Related links in Nature Research

Superconductivity record bolstered by magnetic data

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

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.