Speaker: Jun Li

Beijing Computational Science Research Center    

Abstract: We present an efficient approach, based on a number-conditioned master equation, for large-deviation analysis in mesoscopic transports. Beyond the conventional full-counting-statistics study, the large-deviation approach encodes complete information of both the typical trajectorie and the rare ones, in terms of revealing a continuous change of the dynamical phase in trajectory space. The approach is illustrated with two examples: (i) transport through a single quantum dot, where we reveal the inhomogeneous distribution of trajectories in a general case and find a particular scale invariance point in trajectory statistics; and (ii) transport through double dots, where we find a dynamical phase transition between two distinct phases induced by the Coulomb correlation and quantum interference.

Date&Time: October 12, 2012 (Friday), 14:30–15:30
Location: 606 Conference Room