Meter calibration and the geometric pumping process in open quantum systems

TL;DR

This paper develops a geometric framework for charge pumping in open quantum systems, linking meter calibration to geometric phases and reconciling different theoretical approaches to transport statistics.

Contribution

It introduces a unified geometric approach to charge pumping in open quantum systems, accounting for meter calibration and reconciling FCS and AR methods.

Findings

Meter calibration emerges as a gauge freedom in geometric pumping.

Physical constraints prevent independent application of geometric and physical considerations.

The approach explains the equivalence of FCS and AR results for average charge transport.

Abstract

We consider the process of pumping charge through an open quantum system, motivated by the example of a quantum dot with strong repulsive or attractive electron-electron interaction. Using the geometric formulation of adiabatic nonunitary evolution put forward by Sarandy and Lidar, we derive an encompassing approach to ideal charge measurements of time-dependently driven transport, that stays near the familiar approach to closed systems. Following Schaller, Kie{\ss}lich and Brandes we explicitly account for a meter that registers the transported charge outside the system. The gauge freedom underlying geometric pumping effects in all moments of the transported charge emerges naturally as the calibration of this meter. Remarkably, we find that geometric and physical considerations cannot be applied independently as done in closed systems: physical recalibrations do not form a group due to constraints of positivity (Bochner's theorem). This complication goes unnoticed when considering only the average charge but it is relevant for understanding the origin of geometric effects in the higher moments of the charge-transport statistics. As an application we derive two prominent existing approaches to pumping, based on full-counting-statistics (FCS) and adiabatic-response (AR), respectively, from our approach in a transparent way. This allows us to reconcile all their apparent incompatibilities, in particular the puzzle how for the average charge the nonadiabatic stationary-state AR result can exactly agree with the adiabatic nonstationary FCS result. We relate this and other difficulties to a single characteristic of geometric approaches to open systems: the system-environment boundary can always be chosen to either include or exclude the ideal charge meter. This leads to a physically motivated relation between the mixed-state Berry phase and the entirely different geometric phase of Landsberg.