The main result of the project is to introduce the 'Periodically Refreshed Baths' (PReB) scheme. This was initially envisioned as a numerical technique which would allow driven dissipative quantum many-body systems far beyond their present limitations. Later, we realized this as a situation possible to devise experimentally, and we analyzed the thermodynamics of the PReB scheme, discovering, what we call 'Periodically refreshed quantum thermal machines'. This required combining various state-of-the-art concepts from the theories of open quantum systems, condensed matter, statistical physics, quantum information and thermodynamics. These were then further put together with state-of-the-art tensor network based computational techniques.
The complexity of simulating dynamics of driven dissipative quantum many-body systems comes from (a) the macroscopic nature of the environments (baths), (b) the many degrees of freedom in the system, (c) the evolution of the system not depending only of the instantaneous state of the system, but on at least a finite history of the states of the system at previous times. The point (c) is called 'finite memory time'. The PReB scheme hinges on the idea that finite memory time implies that the system effectively only sees a finite part of the macroscopic baths. So by repeatedly obtaining finite-time dynamics in presence of finite-size baths one can reconstruct the full dynamics of the system (we prove this). But, simulating the finite-time dynamics with finite-size baths can still be quite difficult, the difficulty scaling exponentially with the number of degrees of freedom. This exponential growth is essentially due to exponential memory requirement. This is handled by using state-of-the-art tensor network techniques which borrow from the standard idea of data compression, like used daily in computers.
The results with this numerical technique were published in Phys. Rev B, 104, 045417 (2021). They were also presented in several conferences and invited talks across the world (including Republic of Korea, India, UK, Poland, USA).
Subsequently, we realized that the PReB scheme is not just a numerical technique. Instead, a particular driving protocol can be devised with realizes the PReB process experimentally. Intrigued by this, we explored the thermodynamics of such a process. Note that, this is very far from the regime of applicability of standard thermodynamics. So, the notion of thermodynamics had to be extended. We used a quantum information-inspired approach to thermodynamics, that has already been formulated in various previous works by other authors. This led to obtaining 'Periodically refreshed quantum thermal machines', which are quantum thermal machines, based on the PReB scheme. This connects various concepts from quantum information, thermodynamics, condensed matter, statistical physics.
These results are in arxiv:2202.05264 which is accepted for publication in Quantum.
