{"id":"c6ad0795-9db7-4409-abcb-09b9008bf07c","arxiv_id":"1908.05139","paper_version":2,"verdict":"CONDITIONAL","confidence":"HIGH","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":4,"one_line_summary":"A tunable mobility edge in the generalized Aubry-André-Harper model acts as an energy filter that yields thermoelectric figures of merit up to ZT ≈ 60 and efficiencies near 40% of Carnot at maximum power.","lead":"Researchers propose a new type of nanoscale heat engine built from a one-dimensional chain of atoms with a special quasiperiodic pattern that separates conducting from insulating states. This mobility edge acts as a tunable energy filter, and numerical calculations suggest it can convert heat to electricity far more efficiently than typical thermoelectric devices.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Phase-averaged Onsager coefficients at a single N=987 may not represent any fixed-phase device; the ZT≈60/40 peaks are nonlinear functions of the averaged L_ij, so phase-resolved distributions are needed.","rationale":"The most load-bearing step is the passage from numerically computed transmission functions to quantitative performance claims. The Landauer-Büttiker formalism and the GAAH model are standard, and the boxcar argument in Sec. IV E supports the qualitative conclusion that an asymmetric transmission with a mobility edge and a ballistic band yields good thermoelectric response. What is not secured is the quantitative level of the headline: ZT≈60/40 and η≈0.4η_C. These values come from a single parameter set and from phase-averaged Onsager coefficients. Because ZT is a nonlinear function of L_ij, the average of the coefficients does not equal the coefficient of an average device, and the fractal quasiperiodic spectrum makes the band-edge transmission especially sensitive to the phase. This is precisely where the claimed peaks sit. The manuscript's own caveat ('we focus on a single, representative example rather than performing an exhaustive study' and 'averaged over the phase φ') flags the missing support. A phase-resolved and N-resolved numerical check would settle whether the numbers are robust. Unless such a check is supplied, the correct verdict is conditional: the qualitative mechanism is credible, but the quantitative predictions need qualification.","tokens_in":18454,"tokens_out":10551,"duration_ms":124800,"concrete_test":"Recompute the transport coefficients without phase averaging for the same chain parameters (α=0.792, λ=−0.8t, γ=t, T=0.1t/k_B) and for N=987, and optionally N=1597/2584. For each of at least 100 phases φ uniformly in [0,2π), evaluate τ_φ(E) via Eq. (20) and L_ij(φ) via Eq. (12), then compute the phase-resolved ZT(μ) and η(P_max)(μ). At the reported peak positions (just above E_c and at the first and second band edges), report the median and the 10–90% spread. If the peaks are rare, shift in μ by more than k_B T, or have a spread comparable to their magnitude, the phase-average numbers are not representative of a fixed-phase device; if the distribution is narrow, the claim is supported.","verdict_should_be":"UNCHANGED","load_bearing_attack":"Sec. IV A states that all transport quantities are averaged over 40 phases φ, and the manuscript uses one representative parameter set at N=987. The central quantitative predictions—ZT≈60 and ≈40 at band edges (Sec. IV B) and η≈0.4η_C (Sec. IV C)—are nonlinear functions of the Onsager coefficients. Computing ZT from phase-averaged L_ij is not equivalent to the performance of a single realization, and in a finite incommensurate system the fractal, phase-dependent transmission is expected to fluctuate most strongly exactly at the band edges where the peaks occur. The paper provides no variance, no phase-resolved data, and no finite-size scaling. If the high-ZT peaks are produced by averaging and are rare in individual phases, the headline numbers do not describe a single device; the qualitative statement that a mobility edge acts as an energy filter could survive, but the quantitative claims would need to be restated as ensemble averages or supplemented by robustness data.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript proposes a steady-state thermoelectric heat engine based on the generalized Aubry-André-Harper (GAAH) model, whose single-particle mobility edge can be tuned by potential parameters. Using non-equilibrium Green's function and Landauer-Büttiker calculations for a 987-site chain with γ = t, the authors compute the Onsager coefficients for one parameter set (α = 0.792, λ = -0.8t), average over 40 quasiperiodic phases, and report low-temperature figures of merit ZT ≈ 10 near the mobility edge and ZT ≈ 60 and ≈ 40 at the edges of the first two ballistic bands, with efficiency at maximum power up to about 0.4η_C. They further compare with a clean wire and with a boxcar approximation to argue that the qualitative behavior is generic for 1D quasiperiodic systems with a mobility edge.","tokens_in":18694,"tokens_out":10744,"duration_ms":115401,"significance":"If the quantitative results are robust, this would be a notable contribution: it identifies a concrete, experimentally relevant 1D system in which a mobility edge acts as a tunable energy filter, and it opens a new application of quasiperiodic systems in quantum thermodynamics. The work is built on standard and internally consistent NEGF transport theory; the comparison against a clean wire and a boxcar transmission function is a useful sanity check; and the use of the exact analytic mobility-edge expression of Ref. [34] makes the energy-filter interpretation transparent. The main limitations are that the headline quantitative figures are obtained from phase-averaged transport coefficients without statistical or finite-size validation, and the 'several orders of magnitude' claim in the abstract is not substantiated by any explicit comparison to prior predictions.","major_comments":[{"comment":"The statement 'All quantities shown in this section are obtained numerically and averaged over the phase φ' is not sufficient support for the headline values ZT ≈ 60 and ZT ≈ 40. ZT is a nonlinear function of the Onsager coefficients (ZT = L12^2/det L), so computing it from phase-averaged L_ij yields an ensemble quantity that need not describe any single fixed-phase device. In a finite incommensurate chain the spectrum and transmission are strongly φ-dependent, especially at band edges where the ZT peaks occur. As written, the paper provides no variance, no phase-resolved histogram, and no finite-size scaling (only N = 987 is used). Please provide phase-resolved distributions of ZT and η at the peak positions, or alternatively present a convincing self-averaging argument, and report results for at least two other system sizes. Without this, the quantitative predictions in Figs. 6 and 7 cannot be distinguished from phase-averaging artifacts.","section":"Sec. IV A and Fig. 6"},{"comment":"The abstract claims that the effects 'exceeding existing predictions by several orders of magnitude' is not supported by the manuscript. No specific prior prediction is cited and compared; the only quantitative baseline in the paper is the clean-wire calculation of Sec. IV D, which is not an 'existing prediction' from the literature. Please either cite and compare against specific earlier thermoelectric predictions or reformulate the abstract to state the actual comparison made, for instance by comparing directly with the clean-wire and AAH-model results obtained in the paper. In its current form the claim is too vague to be checked.","section":"Abstract and Sec. I"}],"minor_comments":[{"comment":"The name 'Aubrey-André-Harper' should be 'Aubry-André-Harper'.","section":"Abstract"},{"comment":"The text says the particles are 'spinless electrons', but the Landauer integrals include a factor 2 'due to spin degeneracy'. This is internally inconsistent; if the model is spinless the factor should be 1, while if the factor 2 is retained the model is spin-degenerate. The inconsistency does not affect S, ZT, or η because the factor cancels in ratios, but it should be fixed.","section":"Sec. II C, Eq. (10)"},{"comment":"The sentence 'the last site of the system is coupled to the right lead (denoted by the subscript L)' should say 'denoted by the subscript R'.","section":"Sec. II A, after Eq. (16)"},{"comment":"The caption says '(b)-(d) The transmission functions associated respectively to the first and second configuration', but panels (c) is a spectrum and only (b) and (d) are transmission functions. Please rephrase to ' (b) and (d) Transmission functions ...'.","section":"Fig. 4 caption"},{"comment":"The decision to focus on a single representative parameter set is reasonable for a first study, but the abstract's word 'versatile' would be better supported by at least one additional parameter set with thermoelectric results; Fig. 4(c)-(d) provides a second spectrum and transmission but no corresponding transport coefficients or ZT.","section":"Sec. IV A"},{"comment":"The claim that 'the forms of L11, L12 and L22 ... remain the same regardless of γ, up to an overall factor' is supported in Fig. 11 only for L11 and for the ratio ZT. Showing the analogous curves for L12 and L22 would make the γ-independence argument more complete.","section":"Appendix A"}],"recommendation":"major_revision","confidential_remarks":"The paper is likely to be of interest to the readership if the robustness concerns are addressed. I would not accept it without the phase-resolved and finite-size data described in major comment 1, and the abstract's 'several orders of magnitude' claim should be either substantiated or removed before publication."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Dear colleague,\n\nThe thing to know about this paper is that it is the first to put a generalized Aubry-André-Harper chain—a 1D quasiperiodic model with a tunable mobility edge—into the thermoelectric engine framework, and it does so with standard, internally consistent Landauer-Büttiker calculations. The central idea is sound: the mobility edge separates ballistic from localized states, breaking electron-hole symmetry, and the boxcar comparison shows that the fine fractal structure is not what matters. The comparison to a clean wire also supports the qualitative claim that the mobility edge boosts efficiency at low/intermediate temperature. For that, the paper is worth reading.\n\nThe soft spots are in the quantitative claims. The abstract says the effects 'exceed existing predictions by several orders of magnitude,' but the text never makes that comparison. That phrasing should go or be substantiated. The bigger issue is the phase averaging. The paper says all quantities are 'averaged over the phase φ' (40 values), but there is no variance, no phase-resolved data, and no finite-size scaling. Because ZT and efficiency are nonlinear functions of the Onsager coefficients, using phase-averaged L_ij may not describe any single realization, and the fluctuations are likely largest exactly at the band edges where the ZT peaks occur. The peaks may be real, but as presented the reader cannot tell if they belong to a typical device or are an artifact of the averaging. This needs a phase-resolved analysis or at least error bars.\n\nThere are also a couple of minor things: the 'spinless electrons' label conflicts with the factor 2 in the Landauer formulas (a harmless overall constant, but a typo that should be fixed), and the weak-coupling 'analytical' results in Appendix B rely on fitted proportionality constants—fine for ratios, but not a parameter-free derivation.\n\nWho should read it: people working on thermoelectrics in coherent nanostructures and on quasiperiodic systems. It is a reasonable proof-of-principle, not a definitive demonstration. I would accept it for peer review—the topic is timely and the qualitative message is credible—but I would ask the authors to tone down the abstract, add phase-resolved data, and provide a proper comparison to existing high-ZT proposals.","headline":"A solid proof-of-principle that a 1D quasiperiodic mobility edge can act as a thermoelectric energy filter, but the headline ZT numbers and the 'orders of magnitude' claim outrun the evidence as presented.","tokens_in":19224,"tokens_out":3985,"would_cite":true,"duration_ms":39322,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"The mobility edge of a quasiperiodic 1D wire acts as a tunable energy filter that yields thermoelectric figures of merit up to ZT ≈ 60.","keywords":["quantum heat engine","thermoelectricity","mobility edge","Aubry-André-Harper model","quasiperiodic system","energy filtering","Landauer-Büttiker transport","efficiency at maximum power"],"falsifier":"Take a GAAH chain of 987 sites at $T = 0.1\\, t/k_B$, fix one phase $\\varphi$ instead of averaging over 40 phases, and compute $ZT(\\mu)$ around the first two ballistic bands; if the peaks near $ZT \\approx 60$ and $ZT \\approx 40$ shift or vanish under phase sampling, the single-device prediction fails.","tokens_in":18239,"feed_emoji":"🔥","tokens_out":8350,"duration_ms":77384,"temperature":0.7,"pith_summary":"This paper proposes that a one-dimensional quasiperiodic wire whose spectrum contains a mobility edge can serve as a high-performance, tunable quantum heat engine. The mobility edge — the energy separating localized, insulating states from extended, conducting ones — produces an asymmetric transmission function that acts as an energy filter for thermoelectric conversion. Using the generalized Aubry-André-Harper model, the authors compute linear-response transport coefficients and report thermoelectric figures of merit ($ZT$, a standard dimensionless efficiency measure) of $ZT \\approx 10$ just above the mobility edge, $ZT \\approx 60$ and $ZT \\approx 40$ at the edges of the first two ballistic bands, and an efficiency at maximum power of roughly 40% of the Carnot limit. These values, if correct, exceed existing predictions by orders of magnitude and would make a single quasiperiodic chain a highly attractive heat-to-work converter.","feed_headline":"Mobility edge turns a 1D wire into a ZT≈60 heat engine","feed_subtitle":"Quasiperiodic disorder separates conducting and insulating states, creating an energy filter that beats typical thermoelectrics.","key_machinery":"The load-bearing object is the mobility edge $E_c = (1/[\\alpha\\,\\mathrm{sign}(\\lambda)])(|t|-|\\lambda|)$ of the GAAH model. It splits the single-particle spectrum into localized states below $E_c$ and extended, ballistic states above, making the transmission function $\\tau(E) = \\gamma^2/|\\det[M(E)]|^2$ strongly energy-asymmetric. That asymmetry breaks the particle-hole symmetry of the Fermi-Dirac transport window, so the Seebeck coefficient $S = (1/eT)(\\int dE\\,(E-\\mu)\\,\\tau(E)[-f'(E)] / \\int dE\\,\\tau(E)[-f'(E)])$ is nonzero and large near the edge. The paper shows that only the coarse-grained profile of the conducting bands, not their fractal fine structure, controls the ratio quantities $S$, $ZT$, and $\\eta$, which is why a boxcar approximation to $\\tau(E)$ reproduces the thermodynamic performance.","core_discovery":"The central claim is that the mobility edge of the generalized Aubry-André-Harper (GAAH) model is an exceptionally effective energy filter for steady-state thermoelectric conversion. The model is a tight-binding chain with on-site potential $V_i = 2\\lambda \\cos(2\\pi b i + \\varphi)/[1-\\alpha \\cos(2\\pi b i + \\varphi)]$ and an analytically known mobility edge $E_c = (1/[\\alpha\\,\\mathrm{sign}(\\lambda)])(|t|-|\\lambda|)$. Eigenstates below $E_c$ are localized, so they do not conduct; states above it are extended and ballistic. The resulting transmission function is strongly asymmetric, breaking the particle-hole symmetry that would otherwise cancel the thermoelectric current. Computing the Onsager coefficients through the Landauer-Büttiker formula, the paper finds $ZT \\approx 10$ just above the mobility edge, $ZT \\approx 60$ and $ZT \\approx 40$ at the edges of the first two ballistic bands, and an efficiency at maximum power of $\\eta \\approx 0.4\\,\\eta_C$ at $T = 0.1\\, t/k_B$. The authors argue the effect is generic: a boxcar-shaped transmission with the same coarse features reproduces the ratio quantities, so any quasiperiodic system with a mobility edge separating ballistic and localized states should behave similarly.","pith_inferences":["The paper does not report the spread of $ZT$ across the 40 phase values it averages over; measuring that spread for a fixed wire is the natural next step and would show whether a single realization can actually deliver $ZT \\approx 60$.","If the boxcar argument generalizes, the design principle is broader than quasiperiodic systems: any transmission profile with sharp conducting windows separated by gaps could be engineered into a high-$ZT$ thermoelectric, for instance in superlattices or nanostructured wires.","Because the ratio quantities are independent of the system-bath coupling $\\gamma$ while power grows up to an optimal $\\gamma$, experimental implementations could maximize power by tuning the coupling without sacrificing efficiency.","The linear-response $ZT$ values may not persist under large temperature biases; extending the calculations beyond linear response would reveal whether the mobility-edge filter remains advantageous in the strongly driven regime."],"forward_implications":["Linear-response thermoelectric figures of merit of order 10–60 become available in a single non-interacting one-dimensional chain, far above bulk values around $ZT \\approx 1$.","Tuning $\\alpha$ and $\\lambda$ moves the mobility edge according to $E_c = (1/[\\alpha\\,\\mathrm{sign}(\\lambda)])(|t|-|\\lambda|)$, so one wire can be reconfigured between high-power and high-efficiency operating points.","Because the fractal fine structure of the transmission function is irrelevant to $ZT$ and $\\eta$, the prediction should carry over to other quasiperiodic systems with a mobility edge and to experimentally realized versions with beyond-nearest-neighbour hopping.","The proposed platform is testable with ultracold neutral atoms in bichromatic optical lattices, where two-terminal transport measurements are available."],"supporting_citations":[{"why":"Introduced the generalized Aubry-André-Harper model and derived the analytical mobility-edge expression used throughout.","marker":"[34]"},{"why":"Characterized transport in the GAAH model and established ballistic transport above the mobility edge, grounding the engine's power output.","marker":"[53]"},{"why":"Defined the original Aubry-André-Harper model whose delocalization-localization transition the GAAH model generalizes.","marker":"[26]"},{"why":"Provides the linear-response Onsager and ZT framework used to quantify thermoelectric performance.","marker":"[11]"},{"why":"Supplies the weak-coupling eigenstate expression for currents, used in Appendix B to show the underlying physics analytically.","marker":"[52]"},{"why":"Argues for energy filtering as a route to high efficiency at finite power, the design principle the paper exploits.","marker":"[19]"},{"why":"Demonstrates two-terminal transport measurements with ultracold atoms, the proposed experimental platform.","marker":"[61]"},{"why":"Shows two-terminal ultracold-atom transport techniques used to motivate experimental testability.","marker":"[57]"}],"fun_headline_variants":["Mobility edge filter drives heat engine to ZT≈60","Quasiperiodic chain yields ZT≈60 heat engine","Tunable energy filter gives ZT≈60 thermoelectric","Disorder-based heat engine hits ZT≈60","Mobility edge enables ZT≈60 from quasiperiodic wire"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The quantitative results treat the phase-averaged transport coefficients as the behavior of a single physical wire, although the fractal spectrum of the quasiperiodic potential can make individual realizations differ strongly.","fun_headline_variants_meta":{"raw":{"variants":["Mobility edge filter drives heat engine to ZT≈60","Quasiperiodic chain yields ZT≈60 heat engine","Tunable energy filter gives ZT≈60 thermoelectric","Disorder-based heat engine hits ZT≈60","Mobility edge enables ZT≈60 from quasiperiodic wire"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000456,"raw_usage":{"total_tokens":2340,"prompt_tokens":1043,"completion_tokens":1297,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":659,"completion_tokens_details":{"reasoning_tokens":1212}},"tokens_in":659,"tokens_out":1297,"duration_ms":10966,"temperature":1.0,"reasoning_tokens":1212,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T13:22:36.700575+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Take a GAAH chain of 987 sites at $T = 0.1\\, t/k_B$, fix one phase $\\varphi$ instead of averaging over 40 phases, and compute $ZT(\\mu)$ around the first two ballistic bands; if the peaks near $ZT \\approx 60$ and $ZT \\approx 40$ shift or vanish under phase sampling, the single-device prediction fails.","supporting_citations":[{"cited_title":"Aubry and G","cited_arxiv_id":null,"evidence_quote":"Defined the original Aubry-André-Harper model whose delocalization-localization transition the GAAH model generalizes."},{"cited_title":"Benenti, G","cited_arxiv_id":null,"evidence_quote":"Provides the linear-response Onsager and ZT framework used to quantify thermoelectric performance."},{"cited_title":"Whitney, Phys","cited_arxiv_id":null,"evidence_quote":"Argues for energy filtering as a route to high efficiency at finite power, the design principle the paper exploits."},{"cited_title":"Brantut, C","cited_arxiv_id":null,"evidence_quote":"Demonstrates two-terminal transport measurements with ultracold atoms, the proposed experimental platform."},{"cited_title":"Lebrat, P","cited_arxiv_id":null,"evidence_quote":"Shows two-terminal ultracold-atom transport techniques used to motivate experimental testability."}],"review_version":1}