The basic idea is to fire coherent pulses of cold atoms upwards and then let them fall.

Discussions linking quantum theory and gravity can be traced at least back to Einstein, who famously argued that, if energy is conserved, a photon must be red-shifted on climbing through a gravitational field. This argument alone shows that special relativity cannot adequately describe a universe with gravity and, in the 1960s, experiments using the Mossbauer effect, another quantum phenomenon, confirmed it. Many other manifestations of general relativity have now been observed, the most demanding typically involving planets or stars, or light moving over astronomical distances.

However, our understanding of the relationship between gravity and quantum physics remains speculative. So it is surprising that quantum effects may be put to use again in improving tests of general relativity — they may even be able to detect general relativistic influences in atomic motion within human-scale laboratories.

For atoms falling near the Earth's surface, the magnitude of the relativistic component of the gravitational field is only about 10−15 g, which would seem to make detection almost hopeless. But the accuracy of atom interferometry may be up to the challenge. A recent proposal suggests that this technique should be able to probe general relativity at the 10−15 g level and beyond (S.Dimopoulos et al. http://arxiv.org/abs/gr-qc/0610047). Devices tracking the coherent dynamics of atomic beams may even surpass large-scale astrophysical tests in accuracy.

The basic idea is to fire coherent pulses of cold atoms upwards and then let them fall. Laser pulses interacting with the atoms can recreate the logic of a Mach–Zender interferometer. A series of pulses can act first to split the beam, sending its components along distinct space-time paths, and then to recombine the components for the detection of interference fringes. A rough estimate of the numbers shows that with such a device — roughly 10 m in height, with atoms falling for a little more than 1 s and the results from repeated pulses integrated over the course of a day — general relativistic effects could be just within reach.

This proposed experiment seems also to offer a natural means for distinguishing fundamentally different effects. Departures from newtonian gravity arise in part from the nonlinearity of general relativity, as the field is itself a source for further gravity. Moving atoms also contribute to the field in proportion to their kinetic energy, which also gravitates. As Dimopoulos and colleagues argue, however, these distinct contributions should scale differently with the physical parameters of an experiment — the inital atomic velocity, the time of fall, and so on — suggesting that systematic experiments could probe them separately.

All of which is encouraging, given recent findings concerning dark matter and the apparently accelerating expansion of the Universe, which hint that we may still have a lot to learn about general relativity. We may for a long time lack any way to unify quantum theory with gravity, but understanding the one may give us unprecedented means for learning about the other.