{"id":"4e4de153-5576-48f4-8c9b-b8fb1e77c741","arxiv_id":"2501.13499","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":2,"one_line_summary":"Ultracold 87Rb atoms in a programmable 2D momentum lattice realize a 2D SSH model and show the dynamics of higher-order topological bound states in the continuum.","lead":"Researchers built a 2D lattice in the momentum space of ultracold rubidium atoms and used it to realize a model with corner states that stay localized even though they overlap the bulk energy bands. The work brings a new platform to quantum simulation: programmable 2D synthetic lattices that can host higher-order topological states.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Adiabatic preparation claim relies on C4 symmetry broken by the single-corner initial state; the paper's own odd-L simulations show fidelity below unity, so observed localization may not be the target BIC.","rationale":"The most load-bearing concern is the adiabatic preparation of the BIC. The central claim hinges on the prepared state being a symmetry-protected higher-order BIC, but the experimental protocol starts from a single corner, which manifestly breaks the C4 symmetry invoked for protection. The paper itself states that the actual process resembles the odd-L case, where the fidelity converges to a value below unity unless lambda is close to 1. This is an internal admission that the protection is imperfect for the actual initial condition. The experimental observable (population localization) is not a direct measure of BIC identity; a slow ramp could produce a localized wave packet that is not the topological BIC, and the lack of a gap makes adiabaticity nontrivial. Other aspects of the paper, such as the 2D momentum-lattice realization and the bulk mean-chiral-displacement measurement, are more robust and support the phase-transition claim. The reader's verdict of CONDITIONAL already reflects the need for additional evidence; my concern reinforces that condition rather than changing the verdict. The proposed simulation is a concrete, low-cost check that would directly test whether the experimentally prepared state has high overlap with the target BIC; if it does, the concern is resolved. If not, the central claim of observing a higher-order topological BIC via adiabatic preparation would need to be substantially weakened. This is an addressable, not fatal, issue, consistent with a CONDITIONAL verdict.","tokens_in":8956,"tokens_out":5587,"duration_ms":50062,"concrete_test":"Run exact numerical simulation of the adiabatic ramp on the 16x16 lattice with the experimental single-corner initial state and the actual ramp (tau = 0.75 ms, t2 = 1.25 kHz, tf = 0.20, 0.55, 1.30 kHz), and compute the fidelity F = |<psi_final|psi_BIC>|^2 to the target zero-energy corner eigenstate(s). If F < 0.9 for tf = 0.55 kHz, the claim of adiabatic preparation of the BIC is not supported; also check the overlap of psi_final with the bulk zero-energy subspace. Alternatively, repeat the ramp starting from the symmetric superposition of all four corners and compare the final fidelity; a marked difference would confirm the C4-breaking effect.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim requires the prepared corner-localized state to be the symmetry-protected higher-order topological BIC. The adiabatic strategy (Eq. 2) and the BIC protection argument invoke C4 symmetry: the zero-energy corner state is decoupled from the bulk continuum because of its C4 irreducible representation, despite the absence of an energy gap. But the experiment initializes the BEC in a single corner of the 16x16 lattice, which is not a C4 eigenstate. The paper acknowledges this: 'the actual adiabatic preparation process closely resembles the odd-L case' (Fig. 3a inset), and for odd L 'the finally fidelity to converge to a fixed value, which approaches unity as lambda approaches 1.' Thus for the experimentally relevant symmetry-broken initial condition, fidelity is not perfect and can be significantly below unity for the lambda values used (e.g., tf = 0.55 kHz corresponds to lambda ~ 0.39). The experimental evidence for successful preparation is a localized population distribution (Figs. 3c-d), not a measured overlap with the exact target BIC eigenstate. Localized population could also arise from the slow ramp freezing the initial corner occupation or from finite-size decoupling, without the state being the topological BIC. Since the defining property of a BIC is its coexistence with the bulk continuum and its protection by symmetry, the broken-C4 preparation protocol does not directly establish BIC character. The load-bearing assumption is therefore that residual C4-breaking effects (finite lattice, single-corner state) are negligible; the paper's own simulations contradict this for moderate lambda.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript reports a cold-atom experiment that realizes a two-dimensional (2D) momentum lattice for ultracold 87Rb atoms and uses it to implement the 2D Su-Schrieffer-Heeger (SSH) model. The authors observe corner- and edge-localized dynamics, identify the zero-energy corner states as higher-order topological bound states in the continuum (BICs), demonstrate an adiabatic preparation protocol for these states, and probe the higher-order topological phase transition by measuring a 2D mean chiral displacement as well as boundary dynamics. The central claim is that the observed corner-localized states are symmetry-protected higher-order topological BICs, as stated in the section 'Implementing two-dimensional momentum lattice'.","tokens_in":9238,"tokens_out":3223,"duration_ms":32818,"significance":"If the claims are substantiated, this would be a significant experimental advance: a 2D momentum lattice platform for ultracold atoms, and the first cold-atom realization of higher-order topological bound states in the continuum. The paper has several strengths. The topological classification is not fitted to the data: the 2D Zak phase and phase diagram come from external literature (Liu and Wakabayashi, Ref. [59]), and the hopping amplitudes are experimentally calibrated rather than optimized to reproduce corner states. The authors provide numerical simulations of the quench dynamics and of the adiabatic preparation fidelity, and they compare the observed phase transition with theoretical predictions. The 2D momentum lattice itself is a new capability for synthetic quantum matter. However, as detailed below, the evidence for the specific BIC character of the prepared state and for the adiabatic preparation fidelity is not yet conclusive, and the experimental data lack statistical uncertainties throughout.","major_comments":[{"comment":"The adiabatic preparation claim relies on C4 symmetry protection, but the experiment starts from a single-corner initial state that itself breaks C4 symmetry. The authors acknowledge this in the text: 'the actual adiabatic preparation process closely resembles the odd-L case' (Fig. 3a inset), and for odd L the fidelity converges to a value below unity unless λ is close to 1. For the experimental parameters, e.g., tf = 0.55 kHz corresponds to λ ≈ 0.39, so the expected fidelity is not near unity. The experimental evidence in Figs. 3c–3e is a localized population distribution, not a measured overlap with the target BIC eigenstate. A localized final distribution could also arise from the slow ramp freezing the initial corner occupation or from finite-size decoupling, without the state being the symmetry-protected topological BIC. To support the claim, the authors should provide a quantitative comparison between the measured final state (e.g., the full site-resolved population) and the simulated odd-L adiabatic preparation, including the expected fidelity for the ramps used, and ideally a direct or indirect measurement of the overlap with the target zero-energy corner state.","section":"Symmetry-protected adiabatic preparation of BICs, Eq. (2) and Fig. 3"},{"comment":"The identification of the corner states as BICs embedded in the zero-energy bulk continuum is based on the computed spectrum and the D2 participation parameter. The experimental corner-injection dynamics in Fig. 2(e) show local population retention, but such a result is also consistent with a conventional corner-localized bound state in a finite lattice or with slow spreading due to finite evolution time. The authors do not provide an experimental measure of the bulk population or of the overlap of the time-evolved state with the candidate BIC eigenstate. A more direct test would be to measure the time-resolved population spread and compare it quantitatively with the simulated spreading for the BIC, and to contrast this with the trivial-phase corner dynamics at the same hopping parameters.","section":"Observing bound-states dynamics, Fig. 2(d)–2(f)"},{"comment":"The experimental determination of the 2D winding number from the mean chiral displacement in Fig. 4(b) is presented without error bars, without a clear statement of statistical uncertainty, and without a quantitative measure of agreement with the numerical simulation (solid lines). The claim that v2d 'oscillates around 0' in the trivial phase and 'around 1' in the nontrivial phase needs a quantitative analysis, especially because finite evolution time (τ = 0.6 ms) and finite lattice size can shift the time-averaged value away from the ideal integer. The authors should provide the standard deviation or confidence intervals for each data point and a test of whether the observed values are statistically distinguishable from 0 and 1.","section":"Measuring higher-order topological phase transition, Eq. (3) and Fig. 4(b)"},{"comment":"The paper asserts that the zero-energy corner states are BICs because they lie within the zero-energy bulk continuum while remaining localized. However, no experimental observable directly establishes coexistence with the continuum or the symmetry-protected decoupling. The argument that nearby bulk states have 'vanishingly small overlap' with the corner state is relegated to the Supplementary Material, but the main text does not show how this protection survives the experimental single-corner initial condition that breaks C4 symmetry. The authors should either provide the relevant supplementary proof in the main text or clearly delineate the symmetry assumptions under which the experimental preparation protocol still produces the target BIC, and explain how the data distinguish a BIC from a merely localized corner state.","section":"Discussion of BIC nature, Fig. 2(d) and Supplementary Material [58]"}],"minor_comments":[{"comment":"There are several typographical errors, including 'Figiure 3(a)' in the section on adiabatic preparation, 'programable' in the Discussion, and inconsistent notation such as λx/y versus λ. These should be corrected.","section":"Throughout"},{"comment":"The D2 values used to classify eigenstates as extended or localized are not quantified in the figure or the text; the reader cannot tell from Eq. (1) or the figure what threshold separates D2 ~ 1 from D2 ~ 0. Please provide the actual D2 values for the corner states and for representative bulk states.","section":"Fig. 2(d)"},{"comment":"The main text gives only limited experimental details and refers to the Supplementary Material for calibration procedures. I recommend stating the typical atom number, temperature, and any loss or heating rates in the main text, because these affect the interpretation of the residual condensate fraction in Fig. 3(b).","section":"Experimental methods"}],"recommendation":"major_revision","confidential_remarks":"The manuscript reports an impressive experimental platform and a timely physics claim. However, the central claim that the prepared corner-localized state is a symmetry-protected higher-order topological BIC is not yet fully supported, because the authors' own odd-L simulations admit fidelities below unity and the experimental evidence is limited to population distributions without overlap measurements or error bars. I recommend major revision with a focus on providing quantitative evidence for the BIC character of the prepared state and statistical uncertainties for the phase-transition measurements."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The paper does something new that actually matters: it pushes the momentum-lattice technique into two dimensions and reports the first cold-atom observation of a higher-order topological bound state in the continuum. That is a real experimental advance, not a parameter scan. The 2D SSH model is a known model, but the platform is the story, and the authors use it to show corner-localized dynamics, a bulk topological phase transition via the 2D Zak phase, and a purported adiabatic preparation of the corner BIC. The writing is clear and the main text is honest enough to flag the odd-L fidelity issue in the adiabatic protocol.\n\nThe measurement of the 2D winding number through mean chiral displacement is a nice bulk probe, and the corner vs. edge dynamics are consistent with the model. The paper does not overclaim wildly; it explicitly states that the single-corner initial state makes the preparation resemble the odd-L case with fidelity below unity for moderate λ. That self-criticism is a point in its favor. The central claim—that the corner state is a higher-order topological BIC—is supported by the combination of localized dynamics, the measured topological invariant, and the agreement with simulations. I buy it as a headline result.\n\nThe soft spots are real but not fatal. Error bars are absent from the main figures, which makes it hard to judge how much of the localization is signal vs. noise. More importantly, the stress-test note is on target: the adiabatic argument relies on C4 symmetry, but a single-corner initial state breaks that symmetry. The paper's own odd-L simulations show fidelity converging below unity for the λ values actually used (tf = 0.55 kHz corresponds to λ ~ 0.39), so the observed localized population could partly be a slow-ramp artifact rather than the symmetry-protected BIC. The direct corner-state dynamics in Fig. 2e partially mitigate this, as does the Zak phase measurement, but the authors do not demonstrate that the prepared state actually overlaps a single, well-identified BIC eigenstate. The supplementary file is not included, so the main symmetry-protection argument is deferred. No raw data or code are provided, which limits reproducibility.\n\nFor whom is this? The cold-atom quantum simulation community, and anyone working on higher-order topology in synthetic matter. It deserves a serious referee, not a desk rejection. The issues are addressable with error bars, a state-overlap measurement, and better disclosure of the adiabatic fidelity in the symmetry-broken setup. I would send it to peer review, and I'd expect the referee report to push for those clarifications.","headline":"Genuine first: a 2D momentum lattice in cold atoms with a plausible higher-order topological BIC, but the adiabatic preparation evidence is weaker than the abstract suggests.","tokens_in":9784,"tokens_out":3160,"would_cite":true,"duration_ms":31152,"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":"Ultracold rubidium atoms in a 2D momentum lattice realize higher-order topological bound states in the continuum.","keywords":["higher-order topological insulators","bound states in the continuum","Su-Schrieffer-Heeger model","momentum lattice","ultracold atoms","topological phase transition","adiabatic preparation","Zak phase"],"falsifier":"Measure the corner-site population for evolution times well beyond the 0.9 ms window at $\\lambda=0.5$: a true BIC should keep a localized population plateau, whereas any residual coupling to the bulk continuum would spread the population across the lattice on a characteristic diffusion timescale.","tokens_in":8745,"feed_emoji":"⚛️","tokens_out":11179,"duration_ms":83522,"temperature":0.7,"pith_summary":"This paper reports the observation of higher-order topological bound states in the continuum (BICs) in an ultracold atomic gas. The authors build a two-dimensional momentum lattice for 87Rb atoms using two perpendicular pairs of Bragg lasers, and program it to realize the two-dimensional Su-Schrieffer-Heeger (SSH) model. They show that the model's corner-localized zero-energy states remain localized even though they sit at the same energy as the extended bulk states, identifying them as higher-order topological BICs protected by the bulk topological invariant. They also demonstrate adiabatic preparation of these corner states through a symmetry-preserving ramp of the hopping amplitudes, and map the higher-order topological phase transition via the 2D Zak phase and corner-state dynamics.","feed_headline":"Cold atoms host corner states that are bound in the continuum","feed_subtitle":"A 2D rubidium momentum lattice realizes the 2D SSH model; corner states stay localized inside the bulk continuum.","key_machinery":"The central object is the two-dimensional Su-Schrieffer-Heeger model on a synthetic momentum lattice: a tight-binding model with alternating intra- and inter-cell hopping amplitudes $t_1 = t(1-\\lambda)$ and $t_2 = t(1+\\lambda)$ per direction, whose zero-energy corner states are protected by $C_{4v}$ and chiral (sublattice) symmetries. The load-bearing mechanism is the symmetry protection that keeps a corner state decoupled from the bulk continuum even at zero energy, which the paper probes through a time-averaged mean chiral displacement that extracts the 2D winding number, and through an adiabatic ramp from a single populated corner that tests whether the BIC can be prepared deterministically.","core_discovery":"The central claim is that the zero-dimensional corner states of the two-dimensional Su-Schrieffer-Heeger model, realized in a synthetic momentum lattice of ultracold 87Rb atoms, are higher-order topological bound states in the continuum. Unlike ordinary corner states that sit inside a band gap, these corner states are degenerate with the bulk energy bands, yet they remain localized because the $C_{4v}$ and chiral symmetries forbid hybridization with the bulk continuum. The paper supports this by showing corner- and edge-localized dynamics, by adiabatically preparing the corner state through a ramp that turns on the intracell hopping $t_1$ from zero, and by measuring the 2D Zak phase, which takes value $(\\pi,\\pi)$ in the nontrivial phase and matches the appearance of the corner BIC.","pith_inferences":["Because the experiment initializes a single corner and thereby breaks the $C_4$ symmetry, the measured preparation fidelity should be compared with the paper's odd-lattice simulation; a systematic scan of final corner population versus ramp time and $\\lambda$ could separate genuine symmetry protection from finite-size confinement.","The same time-averaged chiral displacement observable could be used to detect higher-order topology in other symmetry classes, such as models with time-reversal or particle-hole symmetry, without requiring edge spectroscopy.","Introducing interactions on the momentum lattice could test whether the corner BIC remains decoupled from the continuum or acquires a finite lifetime, connecting to the strongly correlated regime the paper identifies as a future direction."],"forward_implications":["The 2D momentum lattice can be extended to more complicated geometries and internal-state configurations, opening higher synthetic dimensions to cold-atom topological simulation.","The adiabatic preparation scheme can prepare a higher-order topological BIC on demand, providing a controllable initial state for studies of dynamics and of the interplay between topology and interactions.","The bulk measurement of the 2D Zak phase via mean chiral displacement offers a boundary-free probe of the higher-order topological phase transition.","The tunable long-range interactions of the momentum lattice allow the fate of these BICs to be studied in the strongly correlated regime."],"supporting_citations":[{"why":"Supplies the 2D SSH model and the 2D Zak phase used to define the topological phase and the transition.","marker":"[59]"},{"why":"Provides the theoretical concept of higher-order topological bound states in the continuum that the paper claims to observe.","marker":"[38]"},{"why":"Earlier experimental demonstration of higher-order topological BICs in photonic waveguides, the benchmark this cold-atom result extends.","marker":"[39]"},{"why":"Introduced higher-order topological insulators whose boundary states are protected by crystalline symmetries.","marker":"[7]"},{"why":"Establishes the one-dimensional momentum-lattice technique in ultracold atoms on which the new 2D implementation is based.","marker":"[54]"}],"fun_headline_variants":["Corner states bound in the continuum seen in cold atoms","Ultracold atoms realize higher-order topological BIC","Symmetry-protected corner BIC in a 2D SSH model","Bound states in the continuum from 87Rb momentum lattice","Edge and corner states survive inside bulk energy bands"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The claim that the prepared corner state is a true topological bound state in the continuum assumes that the $C_{4v}$ and chiral symmetries keep the zero-energy corner state decoupled from the degenerate bulk continuum even though the experimental lattice is finite and the initial single-corner state breaks the $C_4$ symmetry.","fun_headline_variants_meta":{"raw":{"variants":["Corner states bound in the continuum seen in cold atoms","Ultracold atoms realize higher-order topological BIC","Symmetry-protected corner BIC in a 2D SSH model","Bound states in the continuum from 87Rb momentum lattice","Edge and corner states survive inside bulk energy bands"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000216,"raw_usage":{"total_tokens":1393,"prompt_tokens":865,"completion_tokens":528,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":481,"completion_tokens_details":{"reasoning_tokens":447}},"tokens_in":481,"tokens_out":528,"duration_ms":5412,"temperature":1.0,"reasoning_tokens":447,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-10T15:53:31.270749+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Measure the corner-site population for evolution times well beyond the 0.9 ms window at $\\lambda=0.5$: a true BIC should keep a localized population plateau, whereas any residual coupling to the bulk continuum would spread the population across the lattice on a characteristic diffusion timescale.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Establishes the one-dimensional momentum-lattice technique in ultracold atoms on which the new 2D implementation is based."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides the theoretical concept of higher-order topological bound states in the continuum that the paper claims to observe."},{"cited_title":"Cerjan, M","cited_arxiv_id":null,"evidence_quote":"Earlier experimental demonstration of higher-order topological BICs in photonic waveguides, the benchmark this cold-atom result extends."}],"review_version":1}