NMR Protocol for Black-Box Ergotropy Estimation via Feedback Algorithm

Considering the emerging applications of quantum technologies, studying energy storage and usage at the quantum level is of great interest. In this context, there is a significant contemporary interest in studying ergotropy, the maximum amount of work that can be extracted unitarily from an energy-storing quantum device. Here, we propose and experimentally demonstrate a feedback-based algorithm (FQErgo) for estimating ergotropy. This method also transforms an arbitrary initial state to its passive state, which allows no further unitary work extraction. FQErgo applies drive fields whose strengths are iteratively adjusted via certain expectation values, conveniently read using a single probe qubit. Thus, FQErgo provides a practical way for unitary energy extraction and for preparing passive states. By numerically analyzing FQErgo on random initial states, we confirm the successful preparation of passive states and estimation of ergotropy, even in the presence of drive errors. Finally, we implement FQErgo on two- and three-qubit NMR registers, prepare their passive states, and accurately estimate their ergotropy.