Thermalization of finite complexity and its application to heat bath algorithmic cooling
Xueyuan Hu, Valerio Scarani · 2025
Read Thermalization of finite complexity and its application to heat bath algorithmic cooling by Xueyuan Hu, Valerio Scarani (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2505-10054.txt.
Introduction
Xueyuan Hu and Valerio Scarani study how small systems cool when they meet bath units one by one. The work dates to 2025. Each bath unit is no larger than the system, which keeps the whole setup simple. The bath touch keeps total energy fixed.
The pair finds a clear rule for cooling below the bath heat in one touch. The system must start with no set heat level at all, not even a cold one below zero. A step-by-step plan can then reach sub-bath cold with no extra machine. Adding a tiny two-state machine lifts both the cold limit and the yield of the plan.
Catalog record
| CREATOR | Xueyuan Hu, Valerio Scarani |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 73 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2505-10054
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Thermalization of finite complexity and its application to heat bath algorithmic cooling Xueyuan Hu, Valerio Scarani Abstract We introduce a class of thermal operations based on the collision model, where the system sequentially interacts with uncorrelated bath molecules via energy-preserving unitaries. To ensure finite complexity, each molecule is constrained to be no larger than the system. We identify a necessary condition for cooling below the bath temperature via a single collision: the system must initially lack a well-defined effective temperature, even a negative one. By constructing a iterative protocol, we demonstrate that sub-bath cooling is achievable without a machine under these restricted thermal operations. Moreover, introducing a qubit machine further enhances both the cooling limit and energy efficiency. These findings contribute to the broader study of cooling with finite resources. I Introduction Thermodynamics establishes the fundamental laws governing state transformations of matter. However, when the system size approaches the microscopic scale, stricter conditions than those dictated by traditional thermodynamics must be satisfied [ 1 ] . Within the framework of the resource-theoretic approach [ 2 ] , significant progresses have been made over the past two decades [ 3 ] . Generalized second laws have been formulated to describe both population dynamics [ 4 ] and coherence dynamics [ 5 ] in quantum state evolution under thermal operations (TO)—a set of operations modeling interactions between the system and a thermal reservoir at a fixed temperature. Furthermore, necessary conditions for state transitions assisted by ancillary systems, such as catalysts [ 6 ] and batteries [ 7 ] , have also been established. The third law of thermodynamics has similarly been revisited and refined: not only have no-go theorems been rigorously proved through careful identification of relevant resources [ 8 , 9 , 10 ] , but the fundamental cooling limits achievable with finite resources have also been derived [ 11 , 12 , 13 , 14 ] . Heat-bath algorithmic cooling (HBAC) [ 15 ] leverages quantum control techniques to transfer entropy from the target system to an auxiliary machine system, which subsequently dissipates the entropy into a thermal reservoir. This process enables cooling of the target system to temperatures below that of the surrounding heat bath. Beyond its foundational role in quantum thermodynamics, HBAC-based techniques have also been applied to improve sampling efficiency in quantum machine learning [ 16 ] . Traditionally, the reset subsystem of the machine is assumed to undergo full thermalization with the heat bath [ 17 , 18 , 19 ] . However, recent progresses demonstrate that improved cooling performance can be achieved by optimizing over partial thermalization protocols. These include thermalizing only specific subspaces, known as state reset (SR) [ 20 ] , and optimization over the full set of thermal operations, as in extended HBAC (xHBAC) [ 21 ] . Remarkably, xHBAC allows for asymptotic cooling of any finite-dimensional system to absolute zero, even in the absence of a machine. It is also observed that, because the reservoir is infinitely large, the complexity, measured by the dimension of the interaction unitary between the system and reservoir [ 9 ] , diverges. In this paper, we address the following two questions: • Can cooling below the bath temperature be achieved without the use of a machine, under thermal operations of restricted complexity? • In the presence of a machine, can the achievable energy efficiency or cooling limit be improved? For this purpose, we propose a set of thermal operations based on the collision model [ 22 , 23 , 24 ] , where the heat bath consists of non-interacting identical molecules, and the system interacts sequentially with these molecules via energy-preserving unitaries. Crucially, we impose a constraint that each molecule is no larger than the system, ensuring that the overall operation remains of finite complexity. Within this framework, we analyze the condition on the system state before collision, construct a sub-bath cooling protocol based on the collision thermalization, and study the improvement of the energy efficiency and cooling limit by employing a qubit machine. Our main results are the following: 1. For single collisions, we discuss when it is possible to cool the system below the bath’s temperature (Sec. III ). We prove that this is impossible if there is an effective temperature, above the bath temperature or negative, defined for its initial state. This result follows from limiting the complexity of the thermal operations: when this complexity is unlimited, sub-bath cooling can be realized by the optimal \beta -permutation introduced in Ref. [ 21 ] , which is an essential step in the xHBAC protocol. However, if the system state before collision lacks an effective temperature, sub-bath cooling may become possible, and we show an explicit case when it is. 2. For the case where the system and the molecule have equally spaced and nondegenerate Hamiltonian, we construct an iterative protocol to cool the system below the bath temperature without the use of a machine (Sec. IV.2 ). 3. By introducing a single-qubit machine (Sec. IV.3 ), we show that the energy efficiency (quantified via the cumulative coefficient of performance, CoP), can be improved from zero to a strictly positive value. Moreover, if energy efficiency is not constrained, the inclusion of the machine also allows for an enhancement of the ultimate cooling limit. Figure 1: The configuration of one round in HBAC under coherent control. The target system, the machine and the heat bath are labeled as S , M , and R , respectively. The energy-non-preserving unitary V is the recharging routine, and \Gamma , realized by interacting with a heat bath via a energy-preserving unitary U , is the thermalizing routine. Thermalization Machine cooling limit cumulative CoP \Gamma_{1}(\cdot)=\mathrm{tr}_{M}(\cdot)\otimes\tau(H_{M},\beta) [ 11 ] m qubits \beta^{*}=m\beta positive \Gamma_{2}(\cdot)=\mathrm{tr}_{m}(\cdot)\otimes\tau(h,\beta) (PPA, [ 19 ] ) m qubits \beta^{*}=2^{m-1}\beta positive Full thermalization within a subspace of the composed system SM (SR, [ 20 ] ) m qubits \beta^{*}=(2^{m+1}-1)\beta zero Full set of TO (xHBAC, [ 21 ] ) w/o \beta^{*}\rightarrow\infty zero \mathcal{C}_{\beta}(H^{(d_{S})},H^{(d_{r})},1) with d_{r}\leq d_{S} (this work, Sec. IV.2 ) w/o \beta^{*}\approx(d_{r}-1)\beta zero \mathcal{C}_{\beta}(H^{(3)}+H^{(2)},H^{(3)},1) (this work, Sec. IV.3 ) 1 qubit \beta^{*}=2\beta positive \beta^{*}=4\beta zero Table 1: Performance of protocols with different thermalizing process, in the limit N\rightarrow\infty of infinite number of steps. The cumulative coefficient of performance [CoP, Eq. ( 12 )] is a measure of energy efficiency. It is zero, respectively positive, if the processes requires an unbounded, resp. bounded, amount of energy. II thermal operations based on the collision model II.1 definition In the collision model, the reservoir R consists of a collection of identical non-interacting d_{r} -dimensional molecules, each with Hamiltonian H_{r} and initially in the Gibbs state. The target system S under consideration is a d_{S} -dimensional system with Hamiltonian H_{S} . The target system successively collides with the molecules, which is described by a energy-preserving unitary U acting on the system and one molecule. Here and in the following, we use \tau(H,\beta):=e^{-\beta H}/\mathrm{tr}(e^{-\beta H}) to denote the Gibbs state of a system with Hamiltonian H and at inverse temperature \beta . Also, for simplicity, we label the Gibbs state of the molecule and the system as \tau^{r}\equiv\tau(H_{r},\beta) and \tau^{S}\equiv\tau(H_{S},\beta) , respectively, and the corresponding energy distributions as \{\tau_{j}^{r}\}_{j=0}^{d_{r}-1} and \{\tau^{S}_{k}\}_{k=0}^{d_{S}-1} .