News & Views | Published:

High-temperature superconductivity

Electron mirages in an iron salt

Nature volume 515, pages 205206 (13 November 2014) | Download Citation

The detection of unusual 'mirage' energy bands in photoemission spectra of single-atom layers of iron selenide reveals the probable cause of high-temperature superconductivity in these artificial structures. See Letter p.245

Engineering artificial structures on the nanometre scale with the aim of controlling the quantum properties of their electrons is a mainstay of modern materials science. But it is a grand challenge to try to apply this technology to manipulate the rich, collective quantum phenomena of strongly interacting electron systems, because these systems invariably involve complicated chemistry that is hard to tame on the nanoscale. An iconic example of such phenomena is superconductivity (electrical conduction without resistance) in copper- and iron-based metallic salts at unusually high temperatures. Discovered1 in the 1980s, this type of superconductivity is still poorly understood2,3, and only in the past couple of years has it become possible to manufacture high-quality single-atom layers of an iron selenide superconductor4. The surprise is that these nanolayers superconduct at higher temperatures4,5 — perhaps6 up to a whopping 109 kelvin — than the rather poorly superconducting bulk form of this metallic salt. On page 245 of this issue, Lee et al.7 report data that seem to uncover the culprit behind this intriguing observation.

The researchers used photoemission spectroscopy to directly probe the electron worlds of iron selenide (FeSe) films, finding in their spectra highly unusual 'mirage' energy bands. It turns out that these bands can be explained in terms of physics that is unique to these nanostructures: the special nature of the interaction between the FeSe electrons and phonons (quantized crystal-lattice vibrations) in the electrically insulating strontium titanate (SrTiO3) substrate on which the films are grown. The interaction seems to provide an ideal way of boosting the superconductivity of the films. This result is ironic, given the contentious role of electron–phonon interactions in the long history of high-temperature superconductivity8.

Conventional superconductivity, which occurs at temperatures close to absolute zero in simple metals such as aluminium, was explained by a theory developed by John Bardeen, Leon Cooper and John Schrieffer in the mid-twentieth century. This theory describes the superconducting phase as a collective quantum state — a Bose–Einstein condensate — of 'Cooper pairs' of electrons. Cooper pairs form as a result of a quantum-mechanical exchange of phonons that produces an attractive interaction between the electrons; on their own, electrons just repel each other.

However, when the news of the discovery1 of high-temperature superconductivity broke in 1986, it was clear that phonons could not be responsible for the phenomenon8; these are already ineffective at rather low temperatures, and the repulsive interactions between the electrons are exceedingly strong in the metallic salts in which the phenomenon was observed. After nearly 30 years of struggle with intricate quantum many-body theory, eventually a consensus was reached that high-temperature superconductivity can nevertheless arise in these materials. This came about because quantum processes had been identified that could transform the strong electrostatic repulsion between the electrons into effective interactions stabilizing Cooper pairs at high temperature2. This electronic pairing mechanism is very different from that involving phonons2,8, to the extent that the two mechanisms have generically opposing effects in bulk metals. But according to Lee and colleagues' experiments on FeSe films, phonons in the underlying substrate can work closely together with this electronic pairing to promote superconductivity in the films.

Photoemission spectroscopy measures the probability that an electron of given energy and momentum will be removed from a sample subjected to photon excitation. In simple molecules such as hydrogen (H2), the nature of the electron–phonon interactions is easily discerned from such spectroscopic information9. After an electron has been removed, the molecule becomes an ion (H2+) and so acquires an altered bond length. This is indicated in the photoemission spectrum as a progression of peaks: the lowest-energy peak corresponds to the pure electronic excitation, and is followed by peaks at higher energies that are associated with the emission of one, two, and so on, phonons 'repairing' the mismatch in bond length.

Lee et al. observed this spectral pattern for the specific case of the electrons in FeSe monolayers on SrTiO3 substrates. Instead of the lowest-energy peak, the authors detected an energy band that reflects the way in which the electrons quantum-mechanically delocalize in the FeSe, forming the material's metallic and superconducting states. But the authors also observed intense mirages — replicas of this 'normal' band showing up at higher excitation energies. Remarkably, this extra energy is coincident with the energy required to excite dominant phonons of the SrTiO3 substrate.

Such mirage bands have not been observed in any other solid. The reason is that in bulk metals the electron–phonon interaction becomes short-range owing to 'metallic screening' effects. As a result, the electrons excite phonons of all wavelengths, and these phonon 'shake-offs' result in featureless backgrounds in the photoemission spectrum.But in the present case, the nanometre-scale structure of the films results in the electrons exciting only phonons of long wavelength in the insulating substrate, because of unscreened interactions in this medium. This wavelength selectivity causes the appearance of mirage bands, and the strength of the mirages indicates that the electron–phonon interaction is strong. What is more, such long-wavelength phonons actually cooperate with the electronic-pairing mechanism7, explaining why the films superconduct at higher temperatures than their bulk analogues.

“The results suggest new routes by which high-temperature superconductors can be engineered on the nanoscale.”

These are exciting results for superconductivity researchers, because they suggest new routes by which high-temperature superconductors can be engineered on the nanoscale. But perhaps more importantly, they are an example of the kind of surprise that might emerge from studies that attempt to harness the power of nanotechnology to explore the rich but largely uncharted territory of strongly interacting quantum systems.


  1. 1.

    & Z. Phys. B 64, 189–193 (1986).

  2. 2.

    , , , & Preprint at (2014).

  3. 3.

    Rev. Mod. Phys. 83, 1589–1652 (2011).

  4. 4.

    et al. Chin. Phys. Lett. 29, 037402 (2012).

  5. 5.

    et al. Nature Commun. 3, 931; (2012).

  6. 6.

    et al. Preprint at (2014).

  7. 7.

    et al. Nature 515, 245–248 (2014).

  8. 8.

    in 100 Years of Superconductivity (eds Rogalla, H. & Kes, P. H.) Ch. 2.4, 92–117 (Chapman & Hall, 2011); .

  9. 9.

    Molecular Photoelectron Spectroscopy (Wiley, 1970).

Download references

Author information


  1. Jan Zaanen is at the Instituut-Lorentz for Theoretical Physics, Leiden University, 2300 RA Leiden, the Netherlands.

    • Jan Zaanen


  1. Search for Jan Zaanen in:

Corresponding author

Correspondence to Jan Zaanen.

About this article

Publication history




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