| Macroscopic tunneling of dislocations through a pinning point potential is considered in terms of deviations from the classical behavior expected by Teutonico, Granato, and Lucke, and the Isaac-Granato Brownian motion rate theory incorporating viscous damping and inertial overshoot. Internal friction measurements are obtained on three single crystal dilute aluminum alloys down to 0.6 K which do not show any deviation from the classically expected results. This gives an upper bound on the crossover temperature of 1 K. However, in measurements extended to 0.15 K, a flattening of the characteristic T2/3 temperature dependence of the unpinning stress is observed in the superconducting state data of a zone-refined aluminum single crystal, indicating that an athermal process controls the dislocation motion below about 0.45 K in the superconducting state. Although the temperature dependence of the normal state data is weaker than expected classically, it suggests that temperature independent tunneling is not the dominant mechanism of dislocation breakaway for temperatures above 0.15 K in the normal state.;Both the estimates of the crossover temperature and the effective tunneling mass based on a simple geometrical approach introduced by Mott and our preliminary results using a discrete mass approximation are consistent with the data. However, the predictions by Natsik based on a quantum harmonic oscillator approximation do not agree with our observations. A more satisfactory treatment of the multi-dimensionality of the system is given by Link, Epstein, and Baym using a truncated harmonic oscillator approximation. The theory has the same form as the one-dimensional theory with a renormalization of the frequency factor. Using the effective frequency for dislocations in the classical limit determined from the Granato, Teutonico, Lucke piecewise quadratic potential approximation, which is also a multi-dimensional treatment, a crossover temperature in fair agreement with the data is obtained.;An attempt is made to explain the tunneling suppression in the normal state data in terms of the Calderia-Leggett theory. However, using our best estimates of the relevant parameters, the calculated influence of the viscosity seems to be smaller than observed. The origin of this discrepancy is still an open question. |