The temperature below which a material loses its resistance to electricity is known as its superconducting transition temperature. In an earlier experiment, Batlogg and colleagues eliminated electrical resistance from carbon-60 at 54 kelvin by adding positive 'holes' to it. But when they added tribromomethane, the resistance disappeared at the much higher temperature of 117 kelvin.

Carbon-60 molecules form a crystal with a face-centred cubic structure. Its lattice constant - the separation of the centres of two adjacent molecules - is 1.417 nanometres. When trichloromethane was added to the molecule, this stretched to 1.428 nanometres and the transition temperature reached about 70 kelvin. But when tribromomethane was added, the lattice constant grew to 1.443 nanometres, and superconductivity persisted up to 117 kelvin.

As these results show, the transition temperature of carbon-60 increases linearly with lattice constant, and Batlogg's team believes that boosting this constant is the key to achieving superconductivity at higher temperatures. But this will be a challenge: the weak electrostatic attractions that bind the carbon-60 crystal lattice - known as van der Waals forces - will rapidly weaken even further as the carbon-60 molecules accept larger neutral molecules.

This work is the latest in a string of experiments conducted by Batlogg and colleagues into the intriguing properties of carbon-60. The electron-phonon interactions thought to give rise to its superconductivity are dominant inside individual molecules of carbon-60. But the electronic 'density of states' of the material - which helps to determine how well it conducts electricity - is associated with the bulk material.