Have Particle Masses Changed since the Early Universe?

Indications of a change in the proton-to-electron mass ratio have shown up in comparisons of the spectra of hydrogen gas as recorded in a lab with spectra of light coming from hydrogen clouds at the distance of quasars. This is another of those tests of so-called physical constants that might not be absolutely constant. For example, the steadiness of the fine structure constant (denoted by the letter alpha), defined as the square of the electron's charge divided by the speed of light times Planck's constant, has been in dispute (see PNU 410). Some tests say alpha is changing, others say it isn't.

This is an important issue since alpha sets the overall strength of the electromagnetic force, the force that holds atoms together. Similarly, the proton-to-electron mass ratio (denoted by the letter mu) figures in setting the scale of the strong nuclear force.

There is at present no explanation why the proton's mass should be 1,836 times that of the electron's. The new search for a varying mu was carried out by Wim Ubachs of the Vrije Universiteit Amsterdam. He and his colleagues approach their task by studying hydrogen gas in the lab, performing ultra-high-resolution spectroscopy in the difficult-to-access extreme-ultraviolet range. This data is compared to accurate observations of absorption spectra of distant hydrogen (which absorbs light from even more distant quasars) as recorded with the European Southern Observatory (ESO) in Chile.

The astronomical hydrogen is essentially hydrogen as it was 12 billion years ago, so one can seek hints of a changing value for mu. Why the comparison? Because the position of a particular spectral line depends on the value of mu; locate the spectral line accurately (that is, its wavelength) and you can infer a value for mu. In this way, the researchers report that they see evidence that mu has decreased by 0.002 percent over those 12 billion years. According to Ubachs (wimu@nat.vu.nl), the statistical confidence of his spectroscopic comparison is at the level of 3.5 standard deviations.

Reinhold et al., Physical Review Letters, 21 April 2006
Laser Centre Vrije Universiteit (lab Web site)
Contact Wim Ubachs, wimu@nat.vu.nl