Escaping the quantum work trilemma: exotic measurement schemes between energy conservation and the Jarzynski equality, and the negative-spectrum classical limit of the work operator

The Hovhannisyan–Imparato no-go theorem (Quantum 8, 1336 (2024)) proves that no quantum work measurement scheme whose POVM elements and outcomes are differentiable functions of the initial state, and which satisfies two mild structural constraints, can simultaneously satisfy first-law energy bookkeeping (measured mean work equals the unmeasured energy difference for every initial state; untouched systems exchange exactly zero work) and the Jarzynski equality. The theorem leaves a precisely delineated set of escape routes, which this problem records: (i) either a physically reasonable 'exotic' scheme — violating one of the structural constraints or the differentiability hypotheses — jointly satisfies all three conditions, or a strengthened, assumption-free incompatibility theorem holds; and (ii) whether the Heisenberg operator of work recovers the Jarzynski equality in the classical limit for driven harmonic oscillators when the work operator is negative definite, the one case left open by the paper's supplementary analysis. The 2024–2026 follow-up literature has converged on quantum-corrected fluctuation relations that presuppose rather than resolve the incompatibility, so both cores remain open.

Background

In classical stochastic thermodynamics, the work $W$ performed on a driven system prepared in a thermal state is a random variable whose fluctuations obey the Jarzynski equality: the thermal average of $e^{-\beta W}$ equals $e^{-\beta\Delta F}$, where $\beta$ is the inverse temperature and $\Delta F$ is the equilibrium free-energy difference between the initial and final Hamiltonians. In quantum mechanics there is no unique way to assign such a random variable to work: work characterizes a process rather than an instantaneous observable, and measurement back-action on coherent (superposition) states makes competing definitions inequivalent. The two paradigmatic definitions are the two-point measurement (TPM) scheme of Kurchan and Tasaki, which projects onto energy eigenstates before and after the driving and quotes the energy difference — it satisfies the Jarzynski equality, but the first energy measurement destroys coherences, so the measured average work disagrees with the unmeasured energy difference — and the Heisenberg operator of work (HOW) $\Omega = U^{\dagger}H'U - H$ of Allahverdyan and Nieuwenhuizen, where $H$ and $H'$ are the initial and final Hamiltonians and $U$ the unitary evolution: its spectral statistics satisfy energy bookkeeping by construction but violate the Jarzynski equality whenever $H$ and $U^{\dagger}H'U$ fail to commute, as follows from the Golden–Thompson inequality. Talkner, Lutz, and Hänggi argued that work is not an observable at all, and Perarnau-Llobet et al. (2017) proved an earlier no-go: no scheme reproduces both the mean and the variance of work for coherent initial states.

Hovhannisyan and Imparato posed the question in maximal generality. Any work statistics can be modeled as a measurement scheme: a positive operator-valued measure (POVM — a collection of positive operators $M_W \geq 0$ summing to the identity, generalizing a projective measurement) with outcomes $W$, where both the elements and the outcomes may depend on the initial state $\rho$ and on the process $(H, H', U)$. They single out three conditions: $\mathfrak{A}_1$ (first law) — for every initial state, the measured average work $\sum_W W\,\mathrm{Tr}(\rho M_W)$ equals $\mathrm{Tr}(\rho\Omega)$, the unmeasured energy difference; $\mathfrak{A}_2$ (untouched systems) — if the Hamiltonian is not driven, the scheme outputs exactly zero work with probability 1 for every state; and $\mathfrak{B}$ (fluctuation theorem) — for every thermal initial state $\tau_\beta = e^{-\beta H}/Z_\beta[H]$ and every $\beta > 0$,

$\sum_{W} e^{-\beta W}\,\mathrm{Tr}(\tau_\beta M_W) = e^{-\beta(F_\beta[H'] - F_\beta[H])}.$

Their Theorem 1 shows that any state-independent scheme satisfying $\mathfrak{B}$ is incompatible with $\mathfrak{A}_1$; Theorem 2 shows that the schemes satisfying $\mathfrak{B}$ which obey $\mathfrak{A}_1$ on thermal states are equivalent to TPM; and the main Theorem 3 shows that no state-dependent scheme — one whose POVM elements and outcomes are at least once differentiable in $\rho$ in some open ball $B_{\epsilon_0}$ around the maximally mixed state $\tau_0 = \mathbb{1}/d$ and at least twice differentiable in $U$ at $U = \mathbb{1}$, and satisfying two structural constraints — can satisfy $\mathfrak{A}_1$, $\mathfrak{A}_2$, and $\mathfrak{B}$ simultaneously. Constraint 1 stipulates that, in the eigenbasis of $H$, the complexness of the first-order (in the driving strength) POVM elements can originate only from the noncommutativity of $U$ and $\rho$ with $H$; Constraint 2 requires the thermal expectation of the anticommutator of the positive and negative parts of the first-order outcome operator to be nonnegative when $U$ is real in that eigenbasis. All schemes known to the authors satisfy both. Crucially, a fully general but highly discontinuous scheme — apply TPM when $[\rho, H] = 0$ and HOW otherwise — does satisfy all three conditions formally, so what Theorem 3 leaves open is precisely whether a physically reasonable such scheme exists, or whether the incompatibility holds without the technical assumptions.

The paper also analyzes the classical limit $\beta\hbar \to 0$, where all three conditions become compatible for TPM and, in finite-dimensional systems, for HOW. For continuous variables this limit is not continuous — operators such as $\beta^2[p^2, q^2]$ have infinite norm for any nonzero $\beta\hbar$ — yet in Supplementary Note S the authors explicitly compute, for driven harmonic oscillators with quadratic $H$, $H'$ and Gaussian $U$, that $\langle e^{-\beta\Omega}\rangle_{\tau_\beta} = e^{-\beta\Delta F} + O(\beta\hbar)$ whenever the work operator $\Omega$ is positive definite ($\Omega > 0$). The complementary case $\Omega < 0$, where the spectrum of $\Omega$ is unbounded below — so that $e^{-\beta\Omega}$ is unbounded above and the relevant traces may diverge — was left explicitly open.

Problem Statement

Determine whether the mutual incompatibility of energy bookkeeping and the fluctuation theorem proven by Hovhannisyan and Imparato extends beyond the technical hypotheses of their Theorem 3, and complete the classical-limit analysis of the Heisenberg operator of work. The problem comprises two questions whose common objective is the exact delineation of when the first-law conditions and the Jarzynski equality can be observed simultaneously in a quantum work measurement.

(1) Escape or strengthen the no-go. Either exhibit a physically reasonable quantum work measurement scheme — a POVM with outcomes, depending on the state $\rho$ and on $(H, H', U)$ — that lies outside Theorem 3's hypotheses (violating structural Constraint 1 or Constraint 2, or failing differentiability in $\rho$ in a ball around $\tau_0 = \mathbb{1}/d$ or in $U$ at $U = \mathbb{1}$) and simultaneously satisfies $\mathfrak{A}_1$ (measured mean work equals $\mathrm{Tr}(\rho(U^{\dagger}H'U - H))$ for every initial state), $\mathfrak{A}_2$ (an undriven system outputs exactly zero work with probability 1 for every state), and $\mathfrak{B}$ (the exact Jarzynski equality for every thermal initial state and every $\beta > 0$); or prove a strengthened no-go theorem establishing the mutual incompatibility of $\mathfrak{A}_1$, $\mathfrak{A}_2$, and $\mathfrak{B}$ without the differentiability assumptions and Constraints 1–2, replacing them, if needed, by an explicitly stated and physically justified weaker hypothesis. Since the trivial discontinuous scheme (TPM when $[\rho, H] = 0$, HOW otherwise) already satisfies all three conditions, any construction must be accompanied by a precise and defensible operational criterion of physical reasonableness — for example continuity in $\rho$, or implementability without full state tomography — and any strengthened no-go must cover all schemes meeting that criterion.

(2) Classical limit of the HOW scheme for negative-definite work operator. For a driven harmonic oscillator with quadratic $H$ and $H'$ and Gaussian $U$ such that $\Omega = U^{\dagger}H'U - H$ is negative definite ($\Omega < 0$), determine whether $\langle e^{-\beta\Omega}\rangle_{\tau_\beta} = \mathrm{Tr}(e^{-\beta H}e^{-\beta\Omega})/Z_\beta[H]$ converges to $e^{-\beta(F_\beta[H'] - F_\beta[H])}$ as $\beta\hbar \to 0$, diverges, or converges to something else — completing Supplementary Note S of the source paper, which established the positive-definite case. The answer decides whether operator-based definitions of work recover the Jarzynski equality in the classical limit for the full family of Gaussian processes, or whether the sign of the work operator marks a genuine boundary of classical-limit consistency, given that the $\beta\hbar \to 0$ limit is not continuous and the relevant traces may diverge.

The authors' fourth explicitly stated open issue — adapting state-dependent schemes to fluctuation statistics of heat, currents, entropy production, and the energetic cost of measurements — is an open-ended research direction with no recognizable completion boundary; it is part of the surrounding context but not a required deliverable of this problem.

References

1. Karen V. Hovhannisyan and Alberto Imparato, 'Energy conservation and fluctuation theorem are incompatible for quantum work', Quantum 8, 1336 (2024). DOI: 10.22331/q-2024-05-06-1336, arXiv:2104.09364, https://arxiv.org/abs/2104.09364
2. C. Jarzynski, 'Nonequilibrium Equality for Free Energy Differences', Physical Review Letters 78, 2690 (1997). DOI: 10.1103/PhysRevLett.78.2690, https://doi.org/10.1103/PhysRevLett.78.2690
3. Jorge Kurchan, 'A Quantum Fluctuation Theorem', arXiv:cond-mat/0007360 (2000), https://arxiv.org/abs/cond-mat/0007360
4. Hal Tasaki, 'Jarzynski Relations for Quantum Systems and Some Applications', arXiv:cond-mat/0009244 (2000), https://arxiv.org/abs/cond-mat/0009244
5. A. E. Allahverdyan and Th. M. Nieuwenhuizen, 'Fluctuations of work from quantum subensembles: The case against quantum work-fluctuation theorems', Physical Review E 71, 066102 (2005). DOI: 10.1103/PhysRevE.71.066102, https://doi.org/10.1103/PhysRevE.71.066102
6. Peter Talkner, Eric Lutz, and Peter Hänggi, 'Fluctuation theorems: Work is not an observable', Physical Review E 75, 050102(R) (2007). DOI: 10.1103/PhysRevE.75.050102, https://doi.org/10.1103/PhysRevE.75.050102
7. Martí Perarnau-Llobet, Elisa Bäumer, Karen V. Hovhannisyan, Marcus Huber, and Antonio Acín, 'No-Go Theorem for the Characterization of Work Fluctuations in Coherent Quantum Systems', Physical Review Letters 118, 070601 (2017). DOI: 10.1103/PhysRevLett.118.070601, https://doi.org/10.1103/PhysRevLett.118.070601
8. S. Gherardini, A. Belenchia, M. Paternostro, and A. Trombettoni, 'End-point measurement approach to assess quantum coherence in energy fluctuations', Physical Review A 104, L050203 (2021). DOI: 10.1103/PhysRevA.104.L050203, https://doi.org/10.1103/PhysRevA.104.L050203
9. Kaonan Micadei, Gabriel T. Landi, and Eric Lutz, 'Extracting Bayesian networks from multiple copies of a quantum system', EPL 144, 60002 (2023). DOI: 10.1209/0295-5075/ad177d, arXiv:2103.14570, https://arxiv.org/abs/2103.14570
10. Konstantin Beyer, Roope Uola, Kimmo Luoma, and Walter T. Strunz, 'Joint measurability in nonequilibrium quantum thermodynamics', Physical Review E 106, L022101 (2022). DOI: 10.1103/PhysRevE.106.L022101, https://doi.org/10.1103/PhysRevE.106.L022101
11. Konstantin Beyer, Kimmo Luoma, and Walter T. Strunz, 'Work as an external quantum observable and an operational quantum work fluctuation theorem', Physical Review Research 2, 033508 (2020). DOI: 10.1103/PhysRevResearch.2.033508, https://doi.org/10.1103/PhysRevResearch.2.033508
12. Giulia Rubino, Karen V. Hovhannisyan, and Paul Skrzypczyk, 'Revising the quantum work fluctuation framework to encompass energy conservation', npj Quantum Information 11, 102 (2025). DOI: 10.1038/s41534-025-01053-6, arXiv:2406.18632, https://arxiv.org/abs/2406.18632
13. Konstantin Beyer and Walter T. Strunz, 'Operational Work Fluctuation Theorem for Open Quantum Systems', Physical Review Letters 134, 140403 (2025). DOI: 10.1103/PhysRevLett.134.140403, arXiv:2408.13417, https://arxiv.org/abs/2408.13417
14. Thales Augusto Barbosa Pinto Silva and David Gelbwaser-Klimovsky, 'On the consistency of measurement protocols for quantum processes fluctuations', arXiv:2502.12905 (2025), https://arxiv.org/abs/2502.12905
15. Ge Bai, Francesco Buscemi, and Valerio Scarani, 'A quantum entropy production operator', Journal of Physics A: Mathematical and Theoretical 59, 295301 (2026). DOI: 10.1088/1751-8121/ae86cb, arXiv:2412.12489, https://arxiv.org/abs/2412.12489
16. Milton Aguilar, Franklin L. S. Rodrigues, and Eric Lutz, 'Minimal-backaction work statistics of coherent engines', arXiv:2603.01962 (2026), https://arxiv.org/abs/2603.01962
17. Cyril Elouard, Karen Hovhannisyan, and Giulia Rubino, 'Minimally invasive measurement of work in coherent quantum systems', arXiv:2607.02652 (2026), https://arxiv.org/abs/2607.02652
18. Eugene Y. S. Chua, 'Looking for Work in Quantum Thermodynamics', British Journal for the Philosophy of Science (2026). DOI: 10.1086/740608, arXiv:2601.06312, https://arxiv.org/abs/2601.06312
19. Ji-Hui Pei, Tingzhang Shi, Jin-Fu Chen, and H. T. Quan, 'Extending Covariant Fluctuation Theorems into Quantum Regime through Quasiprobability Approach', arXiv:2606.14519 (2026), https://arxiv.org/abs/2606.14519