{"id":"e6e23de2-d9fe-4d7e-9a72-a22b2b60f637","arxiv_id":"1908.05364","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"low","formal_verification":"none","parameter_count":0,"one_line_summary":"A Tonks-Girardeau gas expanding in one dimension is observed to evolve from a bosonic to a fermionic momentum distribution, providing the first measurement of a rapidity distribution in an interacting many-body system.","lead":"Ultracold rubidium atoms in hundreds of parallel one-dimensional tubes were released along the tubes, and their momentum distribution morphed from a bosonic to a fermionic shape. This is the first direct measurement of rapidities, the conserved quantities that govern integrable many-body quantum systems.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Measurement of 'rapidities' requires the finite-tev TOF profile to equal the asymptotic momentum distribution; Fig. S1B shows the profile is still broadening at the latest usable time, so the claim rests on the no-free-parameter simulation rather than on a directly observed asymptotic limit.","rationale":"The reader's weakest assumption concerned the mapping from the measured 1D TOF distribution to the momentum distribution at tev, focusing on the 32 µs transverse shutoff and the 70 ms TOF. The concern identified here is adjacent but distinct: even if that mapping is exact, the central claim requires the momentum distribution at the accessible tev to equal the asymptotic rapidity distribution. The paper's own Fig. S1B shows the FWHM still increasing at the latest times, and the axial potential is only flat over a limited range, so the asymptotic limit is not directly observed. The no-free-parameter hard-core-boson simulation is strong supporting evidence and likely suffices for the main observation of dynamical fermionization, but the stronger 'first measurement of rapidities' claim would be more secure with a quantitative comparison to the true asymptotic distribution. This reinforces the reader's CONDITIONAL verdict rather than overturning it, so no change to the verdict is recommended.","tokens_in":12991,"tokens_out":13123,"duration_ms":145315,"concrete_test":"Deconvolve the experimental TOF profiles at tev=12 ms and tev=15 ms using the independently characterized axial density at tev (e.g., the simulated or measured spatial kernel) and compare the resulting experimental momentum distributions directly to the theoretical rapidity distribution — the noninteracting-fermion momentum distribution of the initial trap — rather than to the finite-time hard-core-boson simulation. If the deconvolved profiles agree with the asymptotic rapidity distribution within the shot-to-shot noise, the finite-tev extrapolation is validated. If they agree only with the finite-time simulation, the measurement of rapidities remains model-dependent and the paper should state that it observes the approach to, rather than the direct observation of, the rapidity distribution.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim that the asymptotic TOF profiles constitute a direct measurement of the rapidity distribution requires the momentum distribution at the measurement time tev to have converged to the asymptotic rapidity distribution. The experimental window is bounded: the axial potential is flat only over about 40 µm, and for tev>15 ms the residual potential begins to distort the expansion. Within this window, Fig. S1B shows the FWHM of the TOF profiles still increasing at 15 ms, including in the flat-potential simulation, so the distribution has not fully reached its asymptotic form. The no-free-parameter hard-core-boson simulation bridges this gap, but the comparison is presented only as overlaid curves without error bars on the main profiles (Fig. 2D), and the simulation itself assumes the T-G limit (γ=∞) and zero temperature. The measured observable is therefore a finite-time TOF distribution whose agreement with a finite-time simulation does not by itself demonstrate that the asymptotic rapidity distribution was directly observed; the 'measurement of rapidities' is an inference mediated by the model. The authors' own discussion of temperature effects at early times and the finite-γ deviations in the quench experiments show that these model assumptions are not idle.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript reports time-of-flight (TOF) measurements of a 1D Tonks-Girardeau (T-G) gas of strongly interacting ultracold bosons after the axial confining potential is suddenly removed, observing the predicted evolution of the momentum distribution from a peaked bosonic shape to a rounded fermionic shape (dynamical fermionization). The authors also study the momentum dynamics after a quench to a new finite axial trap depth, observing bosonic-fermionic oscillations. They compare their measured TOF profiles with no-free-parameter exact numerical simulations of hard-core bosons in the T-G limit, reporting good agreement, and claim that the asymptotic TOF distributions constitute the first direct measurement of the distribution of rapidities in an interacting many-body quantum system.","tokens_in":13211,"tokens_out":10799,"duration_ms":104596,"significance":"If the central claim holds, this is a significant result: it provides a direct experimental window into the conserved quantities (rapidities) of an integrable many-body system, a long-sought observable in quantum integrable physics. The paper's strengths are the no-free-parameter comparison between experiment and exact hard-core-boson simulations, the use of predictions made in prior independent theoretical work (Refs. [7,8]), and the clean 1D expansion protocol that isolates the momentum dynamics. The experimental observation of dynamical fermionization itself appears robust and well supported, as is the observation of the bosonic-fermionic oscillations. However, the stronger claim of 'directly measuring rapidities' is not fully supported by the data as presented, because the measured TOF profiles are taken at finite evolution times and the asymptotic limit is inferred through the simulation rather than demonstrated experimentally.","major_comments":[{"comment":"The claim that the asymptotic rapidity distribution has been directly measured is not supported by the data shown in Fig. S1B, where the FWHM of the TOF profiles is still increasing at tev = 15 ms, the latest usable time before the axial potential becomes non-flat. Because the total detection time tdet is fixed and the remaining TOF is tdet - tev, a fully converged momentum distribution would produce a decreasing FWHM as tev increases; the observed increase indicates that the momentum distribution at tev is still evolving. The connection between the measured finite-time TOF profiles and the asymptotic rapidity distribution is therefore made by the no-free-parameter hard-core-boson simulation, not by a directly observed asymptotic limit. The authors should either provide an experimental demonstration that the normalized profiles have converged (for example, by showing that the profile shape is stationary after rescaling by tTOF) or explicitly state that the rapidity distribution is inferred from agreement with the T-G simulation, thereby tempering the word 'directly'.","section":"Page 6, paragraph after Fig. 2D ('We have thus measured the distribution of rapidities')"},{"comment":"The 'almost perfect agreement' and 'essentially indistinguishable' claims are based on overlaid curves without error bars on the experimental profiles and without any quantitative goodness-of-fit metric. Since the central claim rests on this agreement, the authors should provide error bars or confidence intervals on the measured TOF profiles and a quantitative measure of the deviation between experiment and theory at each tev, especially at the longest evolution times. This would allow the reader to assess whether the agreement is indeed statistically consistent.","section":"Fig. 2D and the associated discussion of agreement"},{"comment":"There is an internal inconsistency between the main text and the Supplemental Material regarding the breathing period of the T-G gas after the quench. The main text states 'The theoretical period is ∼9% shorter than in the experiment' and explains that finite γ makes the experimental period longer, while SI Section E states 'We observe that the experimental breathing period is 9±0.6% smaller than the T-G theory.' These two statements have opposite signs. Because this period comparison is used to support the finite-γ interpretation of the quench data, the sign error must be corrected and the claim reconciled with the data shown in Fig. 3A.","section":"SI Section E (Quench from low to high ωz)"}],"minor_comments":[{"comment":"There is a typographical error: 'experimentaly' should be 'experimentally.'","section":"Page 5, text near Fig. 3"},{"comment":"The phrase 'just starts to effect the TOF distribution' uses 'effect' as a verb; it should be 'affect.'","section":"Page 4, text near Fig. 2"},{"comment":"The caption lists 'the first six times shown in A' but does not identify which color corresponds to which tev value; please add a legend or describe the color order.","section":"Fig. 2D caption"},{"comment":"The statement 'Rapidities have not previously been measured in any interacting many-body quantum system' is a strong negative claim; it should be qualified with 'to the best of our knowledge' unless an exhaustive literature search is provided.","section":"Abstract and introduction"},{"comment":"There is a typographical error: 'shutoff' appears to be a misspelling of 'shutoff.'","section":"SI Section C"},{"comment":"Because the initial dimensionless coupling is only γ ≈ 8.5 (with values as low as 4.2), the statement 'The agreement at long times suggests that the T-G gas model is sufficient for our finite γ system' would be strengthened by a brief quantitative argument or a finite-γ calculation showing that corrections are expected to be small in the asymptotic regime.","section":"Page 4, discussion of finite-γ agreement"}],"recommendation":"major_revision","confidential_remarks":"The experiment appears well executed and the observation of dynamical fermionization is likely correct, but the paper's headline claim of a 'direct measurement of rapidities' is stronger than the finite-time TOF data support; the manuscript needs to either demonstrate convergence to the asymptotic limit or soften the claim. The sign inconsistency between the main text and SI about the breathing period is the kind of error that can undermine reader confidence and must be fixed. Note that M. Rigol is both a coauthor and an author of prior theoretical predictions (Refs. [7,8]); this is not a circularity issue because the predictions predate the experiment, but the independent origin of the predictions should be made clear."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"First, the headline: this paper reports the first experimental observation of dynamical fermionization, and the first measurement of a rapidity distribution in an interacting many-body quantum system. The central new content is the experiment itself—a clean 1D expansion of a Tonks-Girardeau gas with a clever lattice shutoff that lets them probe the momentum distribution at variable times. The asymptotic TOF profiles match no-free-parameter hard-core-boson simulations almost perfectly, which is strong evidence the measurement is correct. The paper is also honest about its regime: finite-γ deviations in the quench experiments, temperature effects at early times, and anharmonic-trap artifacts are all discussed openly rather than hidden.\n\nThe soft spots are real but not disqualifying. First, the phrase 'directly measured the distribution of rapidities' is slightly stronger than what the experiment actually does. The measured TOF distributions are finite-time objects; Fig. S1B shows the FWHM still increasing at 15 ms, the last usable time, so the profile has not fully converged to its asymptotic form. The bridge to the rapidity distribution is the exact simulation, which incorporates all the experimental details and shows the measured profiles are close to the true momentum distribution. That is a legitimate inference, but it is model-mediated, not purely direct. The stress-test note has this right. It would be more careful to say 'we infer the rapidity distribution from TOF measurements and no-free-parameter simulations.' That said, the agreement is good enough that the central physics claim holds up.\n\nSecond, the main fermionization profiles in Fig. 2D have no error bars. The FWHM comparisons in the oscillation data do have error bars, but the key claim rests on curves that are overlaid without uncertainties. That makes it harder to assess how significant the remaining discrepancies are. It is a minor presentation issue, but worth fixing.\n\nThird, the Bose-Fermi oscillation results are a nice extra but are not the core of the paper. The qualitative agreement is there, and the period shift is plausibly attributed to finite γ, but the authors explicitly say no quantitative finite-γ theory was attempted. That's fine; it doesn't undermine the main result.\n\nOn citation practice: having Rigol as a coauthor and comparing to his earlier predictions is not circular, because the predictions are external and the experiment genuinely tests them. The work is not self-citation in any problematic sense.\n\nBottom line: this is a real experimental advance, honestly presented. It deserves serious peer review and, after minor revisions, publication. The lack of error bars on the key comparison and the slight overstatement of 'direct' are the two things I'd ask the referees to push on.","headline":"First clean experimental observation of dynamical fermionization and the first measurement of a rapidity distribution, with strong no-free-parameter simulation agreement; the 'direct' claim is slightly model-mediated and the main curves lack error bars.","tokens_in":13771,"tokens_out":2927,"would_cite":true,"duration_ms":28107,"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":"Expanding a one-dimensional gas of strongly interacting bosons flips its momentum distribution from bosonic to fermionic, directly exposing the system's rapidities.","keywords":["dynamical fermionization","Tonks-Girardeau gas","rapidities","1D Bose gas","momentum distribution","time-of-flight imaging","integrable quantum systems","Bose-Fermi oscillations"],"falsifier":"Take the same initial T-G gas and repeat the fermionization measurement with several lattice shutoff durations (0, 16, 32, 108, 270 microseconds as in the paper's control). If the asymptotic 1D profiles differ by more than the small broadening already reported, the claim that the late-time profile equals the momentum distribution—and hence the rapidity distribution—is falsified. A second decisive check is to compare the late-time profile directly with the computed momentum distribution of the noninteracting Fermi gas that shares the initial density; any systematic mismatch beyond experimental noise would falsify the identification.","tokens_in":12783,"feed_emoji":"⚛️","tokens_out":9082,"duration_ms":86948,"temperature":0.7,"pith_summary":"The paper reports the first direct measurement of rapidities—the conserved quantities that govern integrable many-body quantum systems—by watching the momentum distribution of a one-dimensional Tonks-Girardeau gas of strongly interacting bosons transform from a bosonic peak into a fermionic, rounded shape after the axial trap is switched off. In the Tonks-Girardeau limit, the asymptotic momentum distribution after free 1D expansion is exactly the distribution of rapidities, so a time-of-flight image taken late in the expansion amounts to a readout of those conserved quantities. The measured profiles agree almost perfectly with hard-core-boson simulations that contain no free parameters, and the same apparatus also shows the predicted oscillation between bosonic and fermionic momentum shapes after a sudden change of trap depth. The paper's contribution is to bring integrability, previously a mostly theoretical structure, into direct experimental view.","feed_headline":"Expanding 1D gas flips its momentum shape","feed_subtitle":"Strongly interacting 1D atoms reveal hidden conserved quantities for the first time.","key_machinery":"The load-bearing object is the Tonks-Girardeau (T-G) gas: a 1D Bose gas whose contact interaction is so strong that no two atoms can occupy the same point. Its many-body wavefunction is the absolute value of a noninteracting spinless-fermion wavefunction, so local observables match those of fermions while the momentum distribution remains bosonic; the distribution of rapidities, however, is identical to the noninteracting Fermi momentum distribution. The experiment uses a 2D optical lattice to create many independent 1D tubes, suddenly flattens the axial potential to start 1D expansion, then shuts off the transverse lattice in 32 microseconds at a chosen time to freeze interactions and let the cloud expand for 70 milliseconds before absorption imaging. The numerical twin is a continuum-limit lattice of hard-core bosons, evolved exactly through a Jordan-Wigner mapping onto free fermions; it includes the initial tube populations, the Gaussian axial trap with anti-trap, the time-of-flight propagation, instrumental resolution, and the sum over tubes, with no adjustable parameters.","core_discovery":"On its own terms, the paper establishes that a strongly interacting 1D Bose gas, when released from its axial confinement while remaining transversely confined, dynamically fermionizes: the momentum distribution, initially peaked like that of a bosonic condensate, deforms within about 12 ms into a rounded distribution that matches the momentum distribution of a noninteracting Fermi gas. Because the Tonks-Girardeau wavefunction equals the absolute value of the noninteracting Fermi wavefunction, the asymptotic momentum distribution of the expanding gas is precisely the distribution of rapidities, the conserved quantities of the integrable Lieb-Liniger model. The experiment therefore claims to have measured rapidities in an interacting many-body system for the first time, with no-free-parameter simulations reproducing the data at long expansion times. It further claims that after a sudden change in axial trap depth, the momentum distribution oscillates between fermionic and bosonic shapes with the predicted period, with small deviations from T-G theory traced to the finite interaction parameter $\\gamma$.","pith_inferences":["If the asymptotic profile is truly the rapidity distribution, the same protocol can map how rapidity distributions evolve after arbitrary quenches, offering a direct test of generalized-hydrodynamics predictions for non-thermalizing dynamics.","An unstated corollary is that varying the transverse-lattice shutoff speed and extrapolating to zero shutoff time would quantify the systematic error of the sudden-freeze assumption; this is a simple extension of the control already shown in the supplement.","Because rapidities encode the full many-body state in integrable systems, this readout could be combined with local density or noise-correlation measurements to distinguish integrable from thermalizing dynamics in one and the same experimental run."],"forward_implications":["Rapidities become experimentally accessible observables, turning integrable models from theoretical constructs into testable systems.","The no-free-parameter agreement validates the hard-core-boson description of 1D expansion dynamics in strongly interacting gases.","The observed bosonic-fermionic oscillations after trap quenches confirm the predicted coherent breathing of momentum space and locate where finite-$\\gamma$ corrections become visible.","The same time-of-flight protocol can be applied after more general quenches and to other integrable models, such as the 1D Fermi-Hubbard model, where rapidity distributions feed into generalized hydrodynamics.","With rapidity distributions known, integrable-system theory can predict correlation functions and subsequent dynamics, not just the momentum profile."],"supporting_citations":[{"why":"Shows that the asymptotic momentum distribution after 1D expansion is the rapidity distribution.","marker":"[6]"},{"why":"Predicts that the momentum distribution of a T-G gas dynamically fermionizes after release.","marker":"[7]"},{"why":"Predicts the bosonic-fermionic momentum oscillations after a sudden change of trap strength.","marker":"[8]"},{"why":"Defines the integrable Lieb-Liniger model of 1D bosons with contact interactions.","marker":"[12]"},{"why":"Provides the first experimental realization of Tonks-Girardeau gases in 1D tubes.","marker":"[13]"},{"why":"Derives the effective 1D interaction parameter $\\gamma$ that sets the T-G regime.","marker":"[17]"},{"why":"Supplies the continuum-limit hard-core boson lattice method used for the no-free-parameter simulations.","marker":"[18]"},{"why":"Gives the finite-$\\gamma$ dependence of breathing frequency used to interpret the observed period shift.","marker":"[21]"}],"fun_headline_variants":["Momentum flip: bosons become fermions","First measurement of rapidities in quantum gas","Expansion fermionizes 1D Bose gas","Hidden quantum invariants emerge in expansion","Bosonic gas takes on fermionic identity"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The measurement assumes that the 70 ms time-of-flight spatial distribution is the momentum distribution at the moment the transverse lattice is switched off, which requires the 32 microsecond lattice shutoff to remove all interaction energy before the axial wavefunction changes while keeping the flight long enough for the asymptotic distribution to form.","fun_headline_variants_meta":{"raw":{"variants":["Momentum flip: bosons become fermions","First measurement of rapidities in quantum gas","Expansion fermionizes 1D Bose gas","Hidden quantum invariants emerge in expansion","Bosonic gas takes on fermionic identity"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000283,"raw_usage":{"total_tokens":1624,"prompt_tokens":850,"completion_tokens":774,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":466,"completion_tokens_details":{"reasoning_tokens":705}},"tokens_in":466,"tokens_out":774,"duration_ms":8117,"temperature":1.0,"reasoning_tokens":705,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T13:16:23.052350+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Take the same initial T-G gas and repeat the fermionization measurement with several lattice shutoff durations (0, 16, 32, 108, 270 microseconds as in the paper's control). If the asymptotic 1D profiles differ by more than the small broadening already reported, the claim that the late-time profile equals the momentum distribution—and hence the rapidity distribution—is falsified. A second decisive check is to compare the late-time profile directly with the computed momentum distribution of the noninteracting Fermi gas that shares the initial density; any systematic mismatch beyond experimental noise would falsify the identification.","supporting_citations":[{"cited_title":"Sutherland, Phys","cited_arxiv_id":null,"evidence_quote":"Shows that the asymptotic momentum distribution after 1D expansion is the rapidity distribution."},{"cited_title":"Rigol and A","cited_arxiv_id":null,"evidence_quote":"Predicts that the momentum distribution of a T-G gas dynamically fermionizes after release."},{"cited_title":"Minguzzi and D","cited_arxiv_id":null,"evidence_quote":"Predicts the bosonic-fermionic momentum oscillations after a sudden change of trap strength."}],"review_version":1}