Among its marvelous consequences, general relativity asserts that a stationary clock at Earth’s surface will run slower than one high in a tower where the gravitational potential is weaker; the phenomenon is called gravitational redshift (see the article by Neil Ashby, PHYSICS TODAY, May 2002, page 41). Now Holger Müller (University of California, Berkeley and Lawrence Berkeley National Laboratory) and colleagues report that the redshift idea, first experimentally confirmed 45 years ago, has passed its strictest test yet. In its analysis, the group reanalyzed a 10-year-old experiment that used atom interferometry to determine the gravitational acceleration. In that earlier work, an upward-directed atom interacted with a pair of laser pulses that put it in a superposition of states with differing momentum. As the figure shows, the phase of the atomic wavefunction evolves along each of the two paths, but with a lower frequency along the bottom path. A second pair of pulses tweaked the atom so that the diverging paths would reconverge; an experimental measurement of the probability that the atom is observed at the convergence point yields the phase difference between the two paths. As Müller and company discuss, the earlier-measured phase difference receives contributions due to the relative motion of the atom in its different states and from the laser interactions, but the two effects cancel. The total phase difference is attributable to the redshift. And to better than one part in 108, it is precisely what is predicted by general relativity. (H. Müller, A. Peters, S. Chu, Nature 463, 926, 2010.) —Steven K. Blau
PACS: 03.75.Dg, 04.80.Cc
