By studying an isolated neutron star, astronomers may have found the first observational indication of a strange quantum effect called vacuum birefringence, which was predicted in the 1930s by Werner Heisenberg and Hans Heinrich Euler.

Neutron stars are the very dense remnant cores of massive stars – at least 10 times more massive than our Sun – that have exploded as supernovae at the ends of their lives. In the 1990s, the Germany-led ROSAT space mission for soft X-ray astronomy discovered a new class of seven neutron stars that are known as the Magnificent Seven. The faint isolated objects emit pulses of X-rays every three to 11 seconds or so, but unlike most pulsars they have no detectable radio emission. The ultra-dense stars have an extremely high dipolar magnetic field (of the order 109–1010 T) and display an almost perfect black-body emission, making them unique laboratories to study neutron-star cooling processes.

A team led by Roberto Mignani from INAF Milan in Italy and the University of Zielona Gora, Poland, used ESO’s Very Large Telescope (VLT) at the Paranal Observatory in Chile to observe the neutron star RX J1856.5-3754. Despite being the brightest of the Magnificent Seven and located only around 400 light-years from Earth, its extreme dimness is at the limit of the VLT’s current capabilities to measure polarisation. The aim of the measurement was to detect a quantum effect predicted 80 years ago: since the vacuum is full of virtual particles that appear and vanish, a very strong magnetic field could polarise empty space and hence also light passing through it. Vacuum birefringence is too weak to be observed in laboratory experiments, but the phenomenon should be visible in the very strong magnetic fields around neutron stars.

After careful analysis of the VLT data, Mignani and collaborators detected a significant degree (16%) of linear polarisation, which they say is likely due to vacuum birefringence occurring in the empty space surrounding RX J1856.5-3754. They claim that such a level of polarisation is not easily explained by other sources. For example, the contribution from dust grains in the interstellar medium were estimated to be less than 1%, which was corroborated by the detection of almost zero polarisation in the light from 42 nearby stars. The genuine thermal radiation of the neutron star is also expected to be polarised by its surface magnetic field, but this effect should cancel out if the emission comes from the entire surface of the neutron star over which the magnetic-field direction changes substantially.

The polarisation measurement in this neutron star constitutes the very first observational support for the predictions of QED vacuum polarisation effects. ESO’s future European Extremely Large Telescope will allow astronomers to study this effect around many more neutron stars, while the advent of X-ray polarimetric space missions offers another perspective to this new field of research.