Source
Vinjanampathy & Anders (2015; Contemporary Physics 57, 2016), LIT-010, read in full in NOTE-011.
What was actually shown
The review adds no new results. Its one synthesising observation is that the strands of the field define “work” differently (§5.2, §7):
- Ensemble work. ⟨W⟩ = ∫₀^τ tr[ρḢ]dt, with heat ⟨Q⟩ = ∫ tr[ρ̇H]dt and ⟨Q⟩ + ⟨W⟩ = ΔU (Eqs. 1–3). Average heat and work are process-dependent and are not observables (§2.1, citing Talkner–Lutz–Hänggi).
- Two-point-measurement work. The difference of projective energy measurements at the start and end of a closed unitary process, W = E_m^(τ) − E_n^(0) (Eq. 24), a random variable with distribution Eq. 25. The measurements destroy energy coherences (§3.6).
- ε-deterministic single-shot work content. A^ε ≈ −kT ln(Z_Λ*/Z) (Eq. 49, after Åberg), with Λ* the smallest energy subspace carrying probability above 1 − ε.
- Resource-theory work. w^max_ε = F^min_ε(ρ_S) − F(τ_S) (Eq. 50, after Horodecki–Oppenheim and Gemmer–Anders). The work of formation (Eq. 51) is generally different.
The authors compare the plurality to the many entanglement measures, each suited to a task (§7, p. 39). Only Eq. 49 and Eq. 50 coincide. The review shows this on p. 26 by observing that Z_Λ*/Z equals the sum in Eq. 50. NOTE-011 corrects §7’s pointer, which names Eq. 48, the random single-shot yield; the coinciding quantity is Eq. 49.
The single-shot and resource-theory notions are stated for states diagonal in the energy eigenbasis (Eqs. 48–50, 53). For states with coherence §5.4 gives only necessary conditions (Eq. 57), and says sufficiency is unknown.
What could have come out otherwise. A derivation could have shown the ensemble work to be the i.i.d. limit of the single-shot work, or the two-point work to reproduce the ensemble average in general. The review lists the operational link between ensemble and single-shot work as open (§7).
What this does not say
- That the notions cannot be unified. “Not unified” is the review’s informed judgement as of early 2016 (NOTE-011, C12, moderate). The field was called rapidly evolving, and the record has not checked what came later.
- That the single-shot and resource-theory works are proved equal. The agreement is an identification of two published formulas, one of them stated with ”≈”, not a derivation (NOTE-011, C6, weak to moderate).
- Anything for states with energy coherence. The single-shot second laws (Eq. 53) and the coincidence hold for energy-diagonal states only.
- Which notion is right. Each answers a different question, so a claim about quantum work has to name its notion before it can be assessed. That is the review’s point, and it does not rank them.
- That thermodynamic entropy is von Neumann entropy out of equilibrium. The review assumes it for the rest of the article and calls it debated (§2.1, p. 5). Results downstream inherit the assumption.
Connections
- The fluctuation relations for the two-point work are filed as THEORY-482, from the same reading.
- The classical fluctuation theorems are THEORY-431 (Crooks, Jarzynski, Seifert). Landauer’s principle, which the review states for mixed states “following from the second law”, is THEORY-030.
- The smooth entropies use the unsquared fidelity whose operational meaning is THEORY-584 (Fuchs, LIT-018).