Spin dynamics in polarized neutron interferometry

Since its first implementation in 1974, perfect crystal neutron interferometry has become an extremely successful method applicable to a variety of research fields. Moreover, it proved as an illustrative and didactically valuable experiment for the demonstration of the fundamental principles of quantum mechanics, the neutron being an almost ideal probe for the detection of various effects, as it interacts by all four forces of nature. For instance, the first experimental verification of the 4-pi-periodicity of spinor wave functions was performed with perfect crystal neutron interferometry, and it remains the only method known which demonstrates the quantum mechanical wave-particle-duality of massive particles at a macroscopic separation of the coherent matter waves of several centimeters. A particular position is taken herein by polarized neutron interferometry, which as a collective term comprises all techniques and experiments which not only aim at the coherent splitting and macroscopic separation of neutron beams in the interferometer with the purpose of their separate treatment, but which aim to do so with explicit employment of the spin-magnetic properties of the neutron as a fermion. Remarkable aspects may arise, for example, if nuclear and magnetic potentials are concurrently applied to a partial beam of the interferometer: among other results, it is found that - in perfect agreement to the theoretical predictions - the neutron beam leaving the interferometer features non-zero polarization, even if the incident neutron beam, and hence either of the partial beams, is unpolarized. The main emphasis of the present work lies on the development of an appropriate formalism that describes the effect of simultaneous occurrence of nuclear and magnetic interaction on the emerging intensity and polarization for an arbitrary number of sequential magnetic regions, so-called domains. The confrontation with subtle theoretical problems was inevitable during the experimental work in polarized neutron interferometry, wherefore in the present work some of them are specified and treated in detail or solved for the first time. Among these was, in particular, the spin-dependent refraction of thermal neutrons at magnetic prisms, and magnetic field calculations for special coil systems used in the manipulation of polarized neutron beams. Ultimately, this leads to the concept for a new type of polarizer based on magnetic bi-refringence and to considerations on the realization of a completely new method for the spin-tomographic visualization of ferromagnetic domain structures. (author)