thermoDynamics in Cavity Quantum ElectrodynamIcs – Qu-DICE

Since the XIXth century, classical thermodynamics has developed the conceptual and practical tools to harness thermal noise, turning it into a resource. Heat engines efficiently convert heat produced by thermal baths into useful and possibly storable work, by engineering the transformations of working substances. Reciprocally, irreversibility quantified by entropy production measures our lack of control over noise, putting fundamental bounds to the engines' yield. Eventually, work extraction can be powered by information in so-called Maxwell's demon engines. In these devices, information appears as a physical resource that can be consumed to produce work and conversely, whose encoding in physical memories costs work.

Recently, stochastic thermodynamics has extended the concepts of heat, work and entropy production to small, out-of-equilibrium working substances. It provides an especially convenient framework to study the energetic and entropic footprints of information at the scale of single bits, and to assess the performances of nano-heat engines. Fluctuation theorems have been measured in various experimental situations, allowing to derive tight bounds for work extraction and information-to-work conversion.

Quantum thermodynamics represents the latest state of development of the field. It aims at studying the laws of thermodynamics if the working substances, batteries and baths are quantum entities. This emerging field addresses problems of both fundamental and practical nature, like e.g. what are the causes of irreversibility in the quantum regime, and how can we quantify it? Is quantum coherence some energetic resource, able to enhance the performances of quantum engines? What is the energetic cost of harnessing quantum noise, for instance to perform quantum feedback?

Qu-DICE will bring elements of answers to these questions, taking advantage of advanced platforms of cavity quantum electrodynamics, that have shown over years a high level of quantum control. We will address experimentally and theoretically the three following objectives:

(i) Characterize elementary transfers of energy, entropy and information between simple quantum entities, e.g. single atoms and photons. To do so, we will develop a new generation of Maxwell's demons based on Rydberg atoms and superconducting cavities. We target to measure generalized fluctuation theorems in this new scenery, bringing new quantitative evidences of the energetic and entropic footprints associated to classical and quantum correlations.
(ii) Evidence work extraction from quantum coherence by designing and realizing a new type of solid-state quantum engine. The engine will consist of a single semi-conducting quantum dot coupled to an optical cavity, extracting work from an engineered, out-of-equilibrium bath. The physics at play behind coherence powered quantum engines is still an open problem in quantum thermodynamics. An experimental demonstration involving the observation of clear quantum signatures would be a premiere
(iii) Derive fundamental bounds relating work extraction and information in the genuinely quantum scenarios investigated in Qu-DICE, allowing to elaborate the relevant figures of merit for our engines. Ultimately, this study will provide new tools to assess the energetic cost of an elementary feedback loop.

By evidencing and measuring the energetic and entropic footprints of quantum coherence and entanglement in textbook experiments, Qu-DICE will bring a striking demonstration of the importance of quantum thermodynamics for quantum technologies. It will contribute to build a framework to assess the energetic cost of elementary tasks of quantum control, paving the way towards an energetics of quantum information processing. Within the current search to scale quantum processors, developing such tools is bound to become strategic to benchmark the possible future computing architectures.