Quantum Heat Engine Exploits Indefinite Causal Order to Provide Work and Refrigeration at Once
Physicists built a photonic quantum heat engine that uses indefinite causal order to generate work and refrigeration simultaneously, reversing the everyday direction of heat flow.
Editor's Note ·
- Clarification:
- This article attributes several direct quotes (from Zhong-Xiao Man, Giulio Chiribella, and Rosario Lo Franco) and specific details — the researchers' institutional affiliations and the claim of publication in Physical Review Letters on July 26, 2026 — solely to a Phys.org article. That source could not be archived by our pipeline (HTTP 403, bot-blocked) and could not be independently re-verified word-for-word by editorial review at time of publication. The article's core scientific claim is independently confirmed via the underlying arXiv preprint (arXiv:2511.04028), but readers should treat the exact wording of the quoted passages and the journal/publication-date detail as sourced to Phys.org alone, pending independent re-verification.
Overview
Researchers at Qufu Normal University, the University of Hong Kong, and the University of Palermo have built a quantum heat engine that both refrigerates and generates mechanical work at the same time, by exploiting a purely quantum effect called indefinite causal order, according to Phys.org. The result, published in Physical Review Letters on July 26, 2026, targets an assumption baked into everyday thermodynamics: that heat only moves from hot to cold.
What We Know
Classical thermodynamics holds that heat naturally flows from hotter systems or regions to colder systems or regions until a state of thermal equilibrium is reached, the baseline assumption the team set out to test. The group’s own paper, posted to arXiv, states that heat “can sometime flow from a colder entity to a hotter one” when “the order of interactions between two identical thermalization channels is indefinite.”
That reversal is driven by indefinite causal order (ICO). As co-senior author Zhong-Xiao Man told Phys.org, “Indefinite causal order (ICO) allows two events to occur in a superposition of orders.” In the team’s setup, the two interactions are modeled “as thermalization channels acting on a system via a control qubit, creating an indefinite order,” Phys.org reports.
Using that anomalous heat flow, the researchers then built what their arXiv preprint calls a “quantum Otto cycle with indefinite causal order, which not only achieves refrigeration but also generates work,” folding two separate thermodynamic jobs — cooling and power generation — into a single cycle.
The result was not purely theoretical. “We experimentally realized both the anomalous flow and the engine on a photonic platform,” according to Phys.org, a claim the arXiv paper frames as “a proof-of-principle demonstration of the theory” carried out in “a photonic quantum setup.”
Co-author Giulio Chiribella described the significance in terms of coherence rather than raw output: “First, we identified a new form of anomalous heat flow, showing that quantum coherence can fundamentally alter the way heat is exchanged between systems,” he told Phys.org.
The paper, titled “Anomalous heat flow and quantum Otto cycle with indefinite causal order” on arXiv, was first posted in November 2025 and is authored by Qing-Feng Xue, Qi Zhang, Xu-Cai Zhuang, Yun-Jie Xia, Enrico Russo, Giulio Chiribella, Rosario Lo Franco, and Zhong-Xiao Man.
What We Don’t Know
Neither Phys.org nor the arXiv abstract reports an efficiency figure, coefficient of performance, or power output for the demonstration — there is no quantitative measure of how much cooling or how much work the photonic setup actually produced.
The demonstration is explicitly small-scale: a proof-of-principle simulation using photons, not a working refrigerator or generator, and the sources describe no path yet to a device that could cool or power anything beyond the experimental optics table. Co-senior author Rosario Lo Franco pointed to that gap directly, telling Phys.org, “One area for future research is to move beyond idealized thermodynamic cycles and investigate more realistic implementations where all operations occur in a finite time” — meaning the current result assumes idealized, instantaneous operations that a real-world device could not have.
Phys.org notes the work “could potentially open new avenues for cooling quantum processors and managing heat in quantum sensors, imaging technologies or nanoscale devices,” but that framing is speculative and is not tied to any specific engineering roadmap or timeline in either source.