Abstract
We study the theory behind a Talbot–Lau interferometer. This consists of three gratings in each other's Fresnel near field and accepts spatially incoherent illumination. Our formalism gives a clear physical picture and permits efficient numerical simulations. We concentrate on the case of matter waves and provide an adequate description of recent fullerene experiments taking into account the Casimir–Polder interaction between a molecular beam and mechanical gratings. For more massive molecules, the influence of this interaction is more drastic and leads to a forbiddingly narrow velocity distribution requirement for future experiments with very massive molecules. This problem can be avoided by configurations where at least one mechanical grating is replaced with an optical grating, i.e. a standing light wave. Such interferometers show improved scaling behaviour. Magnifying or demagnifying interferometer variants are also discussed.