Mathematical theory of quantum tunneling decay at positive temperature (Q698186)

Perturbative and, predominantly, power-series approximations are one of the basic tools in quantum theory but a true challenge only emerges when one has to deal with the much (i.e., typically, exponentially) smaller corrections. This type of task often emerges in connection with tunneling, and the paper deals with it on a high level of mathematical rigor. Its main goal is formulated as an initiation of the mathematical (i.e., more than just hand-waving) study tunneling at positive (physicists would say `finite', meaning `non-zero', with a certain logarithmic-scale under-tone) temperatures. The authors offer the first rigorous treatment of several aspects of the problem of the error estimate in the common exponential formula for non-escape probability. Their ``ab initio'' treatment of the estimates based on the time-evolution of the density matrix is pioneering and impressively consequent, proceeding, roughly speaking, through deformations (i.e., complexifications) and Feynman-Kac formula extended to the case of resonances. The task was difficult, requiring a number of new ideas. Thus, the selection of the reasonable assumptions combines the technical feasibility (by keeping, say, the class of Hamiltonians ``simple enough'') with the insightful reduction of the model (treating, e.g., the interaction with the external reservoir with appropriate care). The effort is rendered successful via the new notion of a ``free resonance energy'' which is a function of temperature and whose imaginary part defines the line-width. The latter quantity is, by the way, assigned also a new semi-classical estimate as a byproduct.