Gravitational and Aharonov-Bohm phases in neutron interferometry

This thesis describes two experiments with the interferometer for very-cold-neutrons (VCN) at the Institut Laue-Langevin in Grenoble, France. This interferometer operates with neutrons that are 40 m/s slow and uses micro-fabricated phase gratings as beam-splitters. The first experiment is a demonstration of the scalar neutron Aharonov-Bohm (AB) effect, which is the neutron analogue of the electrostatic AB effect for electrons. Aharonov and Bohm (1959) proposed this latter effect together with the magnetic AB effect to clarify the significance of electromagnetic potentials in quantum mechanics. The experiment described here focuses at for the first time simultaneously demonstrating the two essential operational signatures of all AB-type effects: their nondispersivity, i.e. wavelength independence, and the fact that they are not locally observable in a simply connected space-time. In this thesis it will be shown, as was already pointed out by Zeilinger (1984, 1986), that the neutron analogue also has both signatures and thus can be used to demonstrate them. In the experiment the paths of the neutron interferometer are enclosed by two anti-parallel solenoids. When, determined by a chopper, the neutron is completely inside the homogeneous field region of the coils the field is turned on and off, resulting in a phase shift due to the difference in magnetic energy. High-visibility fringes were observed even for phase shifts exceeding the limit given by the coherence length, which is a clear demonstration of the wavelength independence of this effect. Because the interferometer has spatially separated beams and uses unpolarized neutrons the effect of the magnetic field on the neutron is not locally observable in either of the interferometer arms. The second experiment is what is commonly known as a COW-experiment. In these experiments a neutron interferometer is tilted around its optical axis, resulting in a phase shift due to the difference in gravitational potential acting along the different beam paths. It is a long-standing issue in the field of neutron interferometry, that a whole series of experiments using silicon single-crystal interferometers, starting with the original experiment by Colella, Overhauser, and Werner (1975), consistently show a discrepancy between experiment and theory, after correction for dynamical diffraction effects and other known systematic errors. This situation prompted the measurement of the COW phase shift with the VCN interferometer, which, because of its different characteristics, is not expected to suffer from the same systematic problems. Weber (1998) obtained results that were in agreement with theory within the accuracy of 1 %, excluding some, but not all of the earlier results. By tailoring the incoming neutron spectrum and by more thoroughly investigating and eliminating the systematic effects, the new measurements described in this thesis have a significantly higher accuracy of 0.4 %, confirming theory while excluding at the same time all earlier anamolous results on a 1σ-level. (author)