Superconductivity - the absence of electrical resistance in a material - is observed when the material is cooled below the superconducting transition temperature. In 1957 John Bardeen, Leon Cooper and Robert Schrieffer explained that interactions between electrons and phonons allowed the electrons to overcome their mutual electrostatic repulsion and bind together to form Cooper pairs, which lead to superconductivity.

By the mid-1980s the highest known transition temperature was 23 Kelvin. In 1986, however, materials that became superconducting at much higher temperatures were discovered, and the record transition temperature is now around 130 Kelvin. Most of these materials contain copper oxide layers and are known as cuprates. However, the BCS theory could not explain their behaviour, and ever since both theorists and experimentalists have been looking for an alternative mechanism to explain the formation of Cooper pairs in the cuprates.

Zhi-xun Shen of Stanford University in California, and co-workers at Stanford, the Lawrence Berkeley National Laboratory, also in California, and the University of Tokyo have used angle-resolved photoemission spectroscopy (ARPES) to study the electron dynamics in three different families of cuprate superconductors. Shen and co-workers used the ARPES technique - in which synchrotron radiation causes electrons to be ejected from a sample - to measure the electron velocity and scattering rate as a function of energy.

In all three families of material they observed kinks in the electron distributions which, they claim, cannot be explained by any known process other than interactions between electrons and phonons. "This suggests," they write, "that electron-phonon coupling strongly influences the electron dynamics in the high-temperature superconductors, and must therefore be included in any microscopic theory of superconductivity."

In an accompanying article Philip Allen of the State University of New York at Stony Brook writes that this "interpretation will be controversial."