Magnetic interaction and phase smearing

Abstract

In evaluating the consequences for the de Haas-van Alphen effect of supposing thatB rather thanH is the field acting on the electrons(“magnetic interaction”), it is usually assumed that the amplitude of the oscillations should be reduced by the Dingle factor before magnetic interaction is introduced. However, if, as is often true in pure samples, the Dingle factor is mainly due to sample inhomogeneity and consequent phase smearing, rather than to electron scattering, the usual procedure is clearly wrong and does indeed make inaccurate predictions. This paper explores the consequences of a new treatment in which magnetic interaction is applied to a perfect sample and the Dingle reduction factor is inserted only afterward. This is, in a sense, an opposite limiting approximation to the usual one and a crude justification for it is presented. This new treatment is applied to calculate the line shape of the de Haas-van Alphen oscillations and good agreement is obtained with experimental results which could not be adequately explained by the conventional treatment. The problem of frequency and amplitude modulation of a high-frequency oscillation by a low frequency is also examined and it proves possible to interpret previously puzzling experimental results by the new treatment if some degree of correlation between the phase departures of the high and the low frequencies in different parts of the sample is assumed.