{"id":"7599cf50-c77b-4bcc-951e-7b8b315d709d","arxiv_id":"1908.04782","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":4,"one_line_summary":"DMC pairing gaps in 2D Fermi gases, extracted from odd-even energy staggering at up to 58 particles, are suppressed relative to mean-field BCS and consistent with Gorkov-Melik-Barkhudarov theory.","lead":"This paper computes pairing gaps in two-dimensional Fermi gases using Diffusion Monte Carlo and a finite-range BCS equation. It finds pairing is weaker than the standard mean-field prediction across the BEC-BCS crossover, consistent with an earlier analytic correction.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The thermodynamic-limit gap in Fig. 9 is an average of two finite-N DMC energy differences with no extrapolation or uncertainty estimate; since fixed-node and finite-size errors are uncontrolled in the odd-even difference, the claimed BCS-side suppression is not quantitatively secured.","rationale":"I reviewed the finite-range BCS derivation in Section III; the partial-wave reduction to Eq. (14) and the iterative solution of Eqs. (14)-(15) are internally consistent, and Fig. 2 supports the claim that kF*re ~ 0.006 is effectively zero-range. The two-body check in Section II likewise supports the range choice. The DMC methodology is standard, and the paper is transparent about limitations: Section IV notes that the variational property does not apply to energy differences, and Section V explicitly labels the thermodynamic-limit estimate as an average of two finite-N values. Those two passages are exactly the weak point. The main physical conclusion — BCS-side suppression — depends on the finite-N gaps being representative of the thermodynamic limit. Since no finite-size extrapolation, no error bars, and no fixed-node sensitivity analysis are provided, the benchmark status of Fig. 9 is conditional. This is the same load-bearing concern identified by the reader, so I agree. I do not see a separate internal inconsistency that would require rejection; the requested controls can be supplied, so the appropriate verdict remains CONDITIONAL, and the reader's verdict is unchanged.","tokens_in":12001,"tokens_out":7018,"duration_ms":73556,"concrete_test":"At each eta shown in Fig. 9, compute DMC gaps for the next available closed shells (e.g., N=74 and N=90) with the same trial-wave-function protocol, and plot the gap from Eq. (20) versus 1/N together with N=42, 50, 58, 74, and 90. Extrapolate to N approaching infinity, or use twist-averaged boundary conditions. If the extrapolated value differs from the N=50/58 average by more than the statistical uncertainty, or if the N=50 and N=58 points do not bracket the extrapolated value, the quoted thermodynamic-limit gap and the Gorkov-like suppression are finite-size artifacts rather than benchmarks.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim — suppression of the pairing gap relative to mean-field BCS on the BCS side, similar to Gorkov-Melik-Barkhudarov — is carried by the DMC gaps in Figs. 9 and 10. These are obtained from Eq. (20) using DMC total energies, and Section V explicitly approximates the thermodynamic limit as 'the average of our best two sets of points', i.e., the odd-even staggering at N=50 and N=58. This is not a controlled extrapolation. Because each DMC energy is a variational upper bound, the energy difference in Eq. (20) is not variational; the fixed-node error can differ between closed-shell even systems and open-shell odd systems, and no sensitivity test is reported. The two finite-N estimates are also adjacent closed shells with the same shell structure, so averaging them does not remove a common finite-size bias. The non-interacting finite-size behavior is shown in Fig. 6 and discussed qualitatively, but it is not subtracted from the interacting gaps, and no uncertainty is assigned to the quoted benchmark values. If the uncontrolled bias is comparable to the suppression being claimed, the qualitative conclusion could change.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies pairing in two-dimensional Fermi gases across the BEC-BCS crossover. The authors first derive and solve the mean-field BCS gap equation for a finite-range coordinate-space potential (modified Pöschl-Teller), showing that for the small effective ranges used in their Quantum Monte Carlo simulations (k_F r_e ≈ 0.006) the finite-range corrections are negligible compared with the zero-range analytical result. They then perform variational and diffusion Monte Carlo calculations for particle numbers N = 10, 18, 26, 42, 50, 58 and their N+1, N+2 neighbors, at interaction strengths η = ln(k_F a) from 0.5 to 3.0. The pairing gap is extracted from the odd-even energy staggering of Eq. (20), and the thermodynamic-limit gap is approximated by averaging the N = 50 and N = 58 values. The central physical claim is that on the BCS side of the crossover the DMC pairing gap is suppressed relative to the mean-field BCS prediction, in qualitative agreement with Gorkov-Melik-Barkhudarov theory, and that these results provide thermodynamic-limit benchmarks.","tokens_in":12269,"tokens_out":2495,"duration_ms":27743,"significance":"If the extracted gaps are reliable, the paper provides valuable microscopic benchmarks for pairing in 2D Fermi gases, a system of direct experimental relevance. The finite-range BCS derivation in Sec. III is a clean, useful contribution, and the systematic DMC energy calculations for many particle numbers at several interaction strengths go beyond earlier work that used only N = 26. The authors are also transparent that the variational property does not apply to the energy differences used in Eq. (20), and their two-body binding-energy check in Fig. 1 gives a sensible justification for neglecting finite-range effects. The comparison with Gorkov-Melik-Barkhudarov theory and with the dynamical-cluster results of Vitali et al. frames the main physics claim appropriately. However, the quantitative benchmark status of the DMC gaps is not yet secured because the thermodynamic-limit extraction and the associated uncertainties are not controlled.","major_comments":[{"comment":"The central claim of BCS-side suppression is carried by the DMC gaps shown in Figs. 9 and 10, but these gaps are presented without any statistical or systematic uncertainty estimate. The text states in Sec. IV that the statistical errors in the DMC energies of Figs. 7 and 8 are smaller than the symbols, yet no error bars or confidence intervals are propagated through Eq. (20) into the final gap values. Without quantified uncertainties, the reader cannot judge whether the suppression relative to mean-field BCS, which is the paper's main conclusion, is statistically significant.","section":"Section V, Eq. (20), Figs. 9-10"},{"comment":"The thermodynamic-limit gap is approximated as 'the average of our best two sets of points,' namely the odd-even staggering at N = 50 and N = 58. This is not a controlled finite-size extrapolation: the two particle numbers are adjacent closed shells with the same shell structure, so averaging them does not remove a common finite-size bias. The non-interacting finite-size behavior shown in Fig. 6 is discussed qualitatively but is not subtracted from the interacting gaps, and no sensitivity test (e.g., using N = 42 or including a free-gas-based correction) is reported. The resulting systematic bias could be comparable in magnitude to the claimed suppression, so the benchmark values in Fig. 9 require a more robust extrapolation or an explicit error band.","section":"Section V, paragraph beginning 'To re-cap'"},{"comment":"The paper correctly notes that a variational property applies to the total DMC energy but not to the energy difference of Eq. (20). This is a real limitation for the gap extraction: the fixed-node error may differ between the closed-shell even systems and the open-shell odd systems, and no test of this nodal-error imbalance is provided. A practical check would be to compare VMC and DMC estimates of the staggering, or to vary the pairing-orbital parameters and observe the stability of the gap. As it stands, neither the fixed-node error nor the finite-size error in the gap is controlled, and both could affect the qualitative conclusion of BCS-side suppression.","section":"Section IV, discussion following Eq. (20)"}],"minor_comments":[{"comment":"The manuscript contains numerous typographical errors that should be corrected, including 'inticate' (Sec. III), 'eaqulity' (Eq. (13) derivation), 'p aticle' and 'N mber' (Fig. 5 axis), 'uni s' (Fig. 10 axis), and '/uni0394fg' (Fig. 6 axis).","section":"General"},{"comment":"The notation ∆(N) in Eq. (20) is used for the odd-even staggering, while the physical pairing gap is denoted ∆_gap and ∆(k) appears in the BCS section. Please clarify the relation between these quantities to avoid confusion when reading Figs. 9 and 10.","section":"Eq. (20) and surrounding text"},{"comment":"Reference [52] is incomplete as printed, and reference [27] appears to duplicate [22]. Please update these entries.","section":"References"},{"comment":"The caption of Fig. 6 says the pairing gap applied to the non-interacting gas is zero in the large-system limit, but the finite values for small systems are finite-size artifacts. Since this figure is used to motivate the choice of N = 50 and N = 58, it would be helpful to show the numerical values of ∆_fg(N) in a table or in the text.","section":"Fig. 6"}],"recommendation":"major_revision","confidential_remarks":"The finite-range mean-field derivation and the DMC energy calculations are solid and worth publishing, but the main quantitative claim - the BCS-side suppression of the pairing gap - relies on an uncontrolled thermodynamic-limit approximation and unquantified error bars. This is fixable within the scope of the manuscript by adding uncertainty estimates, a more systematic finite-size analysis, and sensitivity tests for the fixed-node error in the odd-even difference. I therefore recommend major revision rather than rejection."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Quick take: this is a solid but incomplete paper. The finite-range BCS gap equation is new, a routine but useful extension, and the DMC calculations at N up to 58 are a genuine step beyond the earlier N=26 results. What you should know before citing Fig. 9: the central claim—suppression of the pairing gap relative to mean-field BCS on the BCS side—is not quantitatively secured. The thermodynamic-limit gap is an average of two finite-N odd-even staggering values, with no error bars and no controlled extrapolation.\n\nWhat's actually good. Section III derives the 2D BCS gap equation for a coordinate-space potential and checks that the effective ranges used later are short enough for range corrections to be negligible. That is careful and useful. The DMC energy curves show the expected shell structure, and the comparison with the non-interacting gas is a sensible way to think about finite-size effects. The authors are also honest that the variational property does not carry over to the energy differences in Eq. (20). The citation pattern looks fine; they cite their own earlier DMC work because they are extending it.\n\nWhere it softens. The key result in Fig. 9 comes from literally 'the average of our best two sets of points,' N=50 and N=58. Those are adjacent closed shells with similar shell structure, so averaging them does not remove a common finite-size bias. The paper shows the non-interacting odd-even staggering in Fig. 6 but does not subtract it from the interacting gaps or test how sensitive the conclusion is to that choice. The fixed-node error could differ between closed-shell even and open-shell odd systems, and no sensitivity test is reported. Since the claimed suppression is only of order 20-30% relative to BCS, an uncontrolled bias of that size could change the conclusion. There are also no numeric tables of the gaps, which makes it hard for others to use these as benchmarks.\n\nThe comparison with Gorkov-Melik-Barkhudarov is qualitative and appropriately cautious. I think the direction of the effect is plausible, and the earlier N=26 results in Refs. [35,43] certainly needed revisiting. But the paper is not at benchmark status yet.\n\nBottom line: worth a serious referee. The methodology is sound, the data are new, and the issue is real. The revision needs a table of energies and gaps, a heuristic or systematic finite-size correction (the non-interacting subtraction would be a start), and an uncertainty estimate. If that is too much, the claim should be demoted to a qualitative trend.","headline":"Useful finite-range BCS check and larger-N DMC data, but the headline gap suppression rests on an uncontrolled average of two finite-N points and needs a real extrapolation before the benchmarks can be trusted.","tokens_in":12780,"tokens_out":3965,"would_cite":false,"duration_ms":34087,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":["03.75.Ss","03.75.Hh","67.85.Lm","05.30.Fk"],"model":"deepseek-v4-flash","headline":"A two-dimensional Fermi gas's pairing gap is suppressed relative to mean-field BCS theory across the BEC-BCS crossover.","keywords":["two-dimensional Fermi gas","BEC-BCS crossover","pairing gap","diffusion Monte Carlo","odd-even staggering","finite-range potential","Pöschl-Teller potential","superfluidity"],"falsifier":"Run the same trial wave function and DMC procedure at larger closed shells such as N=74 or N=90 and extrapolate $\\Delta(N)$ in $1/N$; if the extrapolated value moves outside the statistical errors of the N=50/N=58 average, the thermodynamic-limit claim is not settled.","tokens_in":11791,"feed_emoji":"⚛️","tokens_out":6703,"duration_ms":62385,"temperature":0.7,"pith_summary":"This paper asks how large the superfluid pairing gap is in a two-dimensional Fermi gas across the crossover from tightly bound dimers (BEC) to weakly bound Cooper pairs (BCS). It derives and solves a mean-field BCS gap equation for a finite-range coordinate-space potential, showing that the effective ranges used in the quantum Monte Carlo calculations are short enough to match the zero-range analytic result. It then extracts the pairing gap from diffusion Monte Carlo ground-state energies of 10 to 58 particles using odd-even energy staggering. The central result is that, toward the BCS regime, the Monte Carlo gaps are suppressed relative to the mean-field BCS prediction, a suppression qualitatively similar to a standard many-body correction. These gaps are offered as thermodynamic-limit benchmarks for other theoretical and experimental studies of two-dimensional Fermi superfluidity.","feed_headline":"2D Fermi gas pairing gaps fall below mean-field BCS","feed_subtitle":"Quantum Monte Carlo energies at N=50 and N=58 yield thermodynamic-limit benchmarks across the crossover.","key_machinery":"The load-bearing analytic object is the two-dimensional BCS gap equation for a coordinate-space potential, derived by expanding the gap function and the potential in angular momentum eigenstates and Bessel functions, Eq. (14): $\\Delta_l(k) = -\\int_0^\\infty dk'\\, k' V_l(k,k') \\Delta_l(k') / (2 E(k'))$, where $V_l(k,k')$ is the Bessel-transformed partial-wave potential and $E(k)$ is the quasiparticle energy. This equation is solved self-consistently with the density equation to show that $k_F r_e = 0.006$ reproduces the zero-range analytic gap. The many-body extraction machinery is a Jastrow-BCS trial wave function with 10 optimized plane-wave pairing orbitals, propagated in imaginary time by fixed-node diffusion Monte Carlo, combined with the odd-even staggering formula Eq. (20) applied to closed-shell particle numbers N = 10, 18, 26, 42, 50, 58 and their neighbors.","core_discovery":"The paper's main claim is that the zero-temperature pairing gap in a two-dimensional Fermi gas, obtained from fixed-node diffusion Monte Carlo energies via $\\Delta(N)=E(N+1)-\\frac{1}{2}[E(N+2)+E(N)]$, is smaller than the mean-field BCS prediction on the BCS side of the crossover, with the suppression growing as $\\ln(k_F a)$ increases. The extracted gaps at $\\eta = 0.5, 1.0, 1.5, 2.0, 3.0$ are compared with the analytic zero-range BCS result, and the DMC points fall below that curve, consistent in shape with the suppression predicted by a standard many-body perturbation theory. The paper also argues that the finite effective range of the interatomic potential, $k_F r_e \\approx 0.006$, is irrelevant at the two-body and mean-field levels, so the DMC results can be interpreted as zero-range physics. The new ingredient that makes this statement non-trivial is that previous small-system DMC calculations found gaps larger than mean-field BCS on the BCS side; the larger particle numbers used here reverse that trend.","pith_inferences":["A direct corollary the paper leaves implicit: the same coordinate-space BCS equation could be applied to neutron-star crust pairing, where the effective range of the nuclear interaction is not negligible, to check how much finite-range effects change the gap.","The paper's finite-size analysis suggests that non-interacting-gas shell structure can be used to design a systematic $1/N$ extrapolation of the pairing gap; implementing such an extrapolation at say N=74 or N=90 would test the thermodynamic-limit claim.","If cold-atom experiments in quasi-2D geometries can measure the pairing gap directly, the predicted suppression could be tested against experiment; agreement would confirm that beyond-mean-field corrections are essential in two dimensions."],"forward_implications":["If the DMC gaps are correct, they provide thermodynamic-limit benchmarks across the two-dimensional BEC-BCS crossover that other many-body methods can be tested against.","On the BCS side, the suppression relative to mean-field BCS indicates that mean-field theory overestimates pairing in two dimensions, and the magnitude of the effect agrees qualitatively with the standard many-body correction.","The newly derived finite-range gap equation shows that coordinate-space potentials with $k_F r_e \\lesssim 0.1$ can be used in quantum Monte Carlo without corrupting the zero-range physics for $\\eta \\ge 0$.","The odd-even staggering at N=50 and N=58 gives a smaller gap than earlier N=26 calculations, so finite-size effects in two-dimensional pairing gaps are significant and must be handled by going to larger systems."],"supporting_citations":[{"why":"Provides the zero-range two-dimensional BCS gap and chemical potential formulas, Eqs. (6)-(7), that the numerical solutions must reproduce.","marker":"[54, 55]"},{"why":"Gives the many-body correction benchmark (suppression by a factor of e relative to mean-field BCS) against which the DMC gaps are compared.","marker":"[56, 57]"},{"why":"Earlier DMC studies with N=26 and the Jastrow-BCS trial wave function; this work extends them to larger N and corrects the earlier finite-size trend.","marker":"[43, 45]"},{"why":"Imaginary-time Green's function results used for comparison in the intermediate crossover region.","marker":"[46]"},{"why":"Earlier two-dimensional QMC pairing-gap calculations that found values above mean-field BCS on the BCS side, the result this paper overturns with larger systems.","marker":"[35]"}],"fun_headline_variants":["DMC gaps drop below BCS in 2D Fermi gas","2D Fermi gas: gaps fall below mean-field BCS","Quantum Monte Carlo sees 2D pairing gap shrink","2D pairing gap: QMC beats mean-field BCS","Larger-N Monte Carlo reverses 2D gap trend"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that the odd-even staggering of diffusion Monte Carlo energies at N=50 and N=58, averaged together, already equals the infinite-system pairing gap; if finite-size or fixed-node errors shift these energy differences, the reported suppression could change.","fun_headline_variants_meta":{"raw":{"variants":["DMC gaps drop below BCS in 2D Fermi gas","2D Fermi gas: gaps fall below mean-field BCS","Quantum Monte Carlo sees 2D pairing gap shrink","2D pairing gap: QMC beats mean-field BCS","Larger-N Monte Carlo reverses 2D gap trend"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000225,"raw_usage":{"total_tokens":1443,"prompt_tokens":899,"completion_tokens":544,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":515,"completion_tokens_details":{"reasoning_tokens":460}},"tokens_in":515,"tokens_out":544,"duration_ms":4833,"temperature":1.0,"reasoning_tokens":460,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T13:33:24.734666+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Run the same trial wave function and DMC procedure at larger closed shells such as N=74 or N=90 and extrapolate $\\Delta(N)$ in $1/N$; if the extrapolated value moves outside the statistical errors of the N=50/N=58 average, the thermodynamic-limit claim is not settled.","supporting_citations":[{"cited_title":"Vitali, H","cited_arxiv_id":null,"evidence_quote":"Imaginary-time Green's function results used for comparison in the intermediate crossover region."},{"cited_title":"Bertaina and S","cited_arxiv_id":null,"evidence_quote":"Earlier two-dimensional QMC pairing-gap calculations that found values above mean-field BCS on the BCS side, the result this paper overturns with larger systems."}],"review_version":1}