{"id":"af7c68f9-cbed-4ea1-afcc-79199f1bfdfa","arxiv_id":"1908.04482","paper_version":3,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":1,"one_line_summary":"For alkaline-earth atoms in a mixed-dimensional optical lattice, two tight-binding Kondo models are derived from exact two-body scattering, with parameters controllable by confinement; the common projection approximation is shown to fail at surprisingly small scattering lengths.","lead":"Two ultracold atoms in a cigar-shaped trap, one moving in an optical lattice and one fixed at a site, can swap nuclear spins and mimic a Kondo magnet. This paper derives the lattice model parameters from exact scattering, and shows that a simple approximation and the experiment's assumed momentum both need correction.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The fitted quasi-momentum k=1.8k0 lies in the second band, but the paper only asserts the second-band scattering generalization in footnote [27]; without an explicit derivation or numerical check, the experimental finite-momentum conclusion is unsupported.","rationale":"The paper's main theoretical contribution, the microscopic derivation of the tight-binding interaction parameters, is internally consistent: Eqs. (39)-(41) follow from matching the 1D scattering lengths of the tight-binding Hamiltonians in Appendix B, and the limit a_o -> 0 correctly recovers model (I). The projection-approximation failure is a concrete, parameter-free quantitative claim. The weakest link is therefore the experimental application in Sec. V, which is also part of the central claim. The reader correctly identified that the finite-momentum and second-band use of amplitudes matched only at zero momentum is the load-bearing assumption. My reading confirms this: the exact scattering amplitudes used in Rse are derived for a lowest-band incident state, and the second-band extension is asserted in footnote [27] without derivation or numerical check. Fig. 4 does not cover the second band or the large scattering lengths relevant to 173Yb. Because the paper itself flags additional uncontrolled effects (footnote [28], final paragraph of Sec. V), the experimental conclusion should remain conditional. I find no other concern that would change the verdict from CONDITIONAL; the model derivation and projection-approximation results are not seriously threatened. Verdict: UNCHANGED.","tokens_in":26411,"tokens_out":6117,"duration_ms":61980,"concrete_test":"Recompute the two-body scattering amplitude for a second-band incident Bloch state by solving Eq. (A22) with the second-band Green's function, i.e., the Mathieu solution of Eq. (A6) with the appropriate band index and characteristic exponent k > k0, and then evaluate Rse(k) at k=1.8k0 for the 173Yb parameters of Fig. 8. Compare the resulting peak positions CA and CB with the experimental peaks pA and pB in Fig. 8(a); if the shift exceeds the experimental peak widths, the k=1.8k0 fit is not controlled. A simpler cross-check is to compare the exact second-band on-shell amplitude with the lowest-band amplitude at the same physical energy; if they differ substantially, the implicit assumption that band extension of Sec. IV A is benign fails.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central experimental claim of Sec. V is that the resonance positions of Ref. [4] are reproduced when the incident g-atom has quasi-momentum k=1.8k0, which is in the second band of the axial lattice. The spin-exchange rate Rse(k) is built from the exact scattering amplitudes f_xi^(e/o)(k) via Eqs. (50)-(52), and those amplitudes are defined in Sec. IV A for an incident lowest-band Bloch state: Eq. (29) explicitly takes psi(zg) as the lowest-band Bloch function, and the asymptotic form Eq. (34) uses F(|zg|) from the corresponding Mathieu solution. Footnote [27] states that the calculation is 'generalized' to the second band, but gives no equations, boundary conditions, or validation. The low-energy expansions (36)-(37) and the zero-momentum matching of Sec. IV B are threshold results and do not control on-shell amplitudes at k=1.8k0. Fig. 4 checks finite momentum only in the lowest band (0 <= k <= k0) and for a single small scattering length a_s^(xi)/a_perp = 0.032, far from the 173Yb parameters (a_s^(+) = 1878 a0) used in the fit. If the unshown second-band generalization shifts the resonance peaks by more than their observed widths, the fitted k=1.8k0, and the claim that the finite-momentum effect is 'very significant', is not established. The paper's own caveats (footnote [28] and the last paragraph of Sec. V) that many-body and anharmonicity effects could also shift k underscore that this part of the argument is not settled.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies two ultracold alkaline-earth-like atoms in a mixed-dimensional geometry: a quasi-1D tube, with the g-atom moving in a 1D optical lattice and the e-atom localized by an axial harmonic trap. It proposes two tight-binding Kondo-type models, depending on whether odd-wave scattering is negligible, and derives the interaction parameters u_0^(I), u_0^(II), u_1^(II) by matching the exact low-energy even- and odd-wave scattering lengths of the quasi-(1+0)D scattering problem. The scattering calculation is performed with a Mathieu-function Green's function and a Lippmann-Schwinger integral equation for the regularized wave-function derivative eta(z). The paper also tests the standard projection approximation for the interaction parameters, finding that it fails when the 3D scattering length is already of order 10% of the confinement lengths. Finally, it computes the spin-exchange collision rate R_se(k) for the 173Yb experiment of Ref. [4] and claims that the finite incident quasi-momentum effect is very significant, inferring that the g-atom quasi-momentum in that experiment may be as large as k=1.8k0, i.e., in the second band of the axial lattice.","tokens_in":26787,"tokens_out":6894,"duration_ms":74547,"significance":"The central derivation is a substantial technical contribution: it gives closed-form expressions for the tight-binding interaction parameters, Eqs. (39)-(41), anchors them to microscopic 3D scattering lengths, and provides a quantitative and convincing demonstration that the naive projection approximation is uncontrolled in realistic alkaline-earth lattice systems. The explicit Green's-function construction with virtual transitions to excited transverse, axial, and band states is detailed and internally consistent. If the finite-momentum part were as well supported as the zero-momentum derivation, the paper would be a strong candidate for publication. The experimental claim, however, rests on an unproven second-band generalization and on a fitted value of k, so the significance of the paper currently lies mainly in the tight-binding parameter derivation and the projection-approximation analysis, not in the quantitative comparison with Ref. [4].","major_comments":[{"comment":"The spin-exchange rate R_se(k) used for the experimental comparison is evaluated at k=1.8k0, which the authors identify with the second band of the axial lattice. However, the scattering amplitudes f_xi^(e/o)(k) entering Eqs. (50)-(52) are defined in Sec. IV A for an incident lowest-band Bloch state: Eq. (29) and the asymptotic boundary condition Eq. (34) are written for the lowest band, and the low-energy expansions (36)-(37) are threshold results. Footnote [27] states that the calculation is \"generalized\" to the second band, but no equations, boundary conditions, or validation are provided. The only finite-momentum test, Fig. 4, covers k in [0,k0] and a single small scattering length a_s^(xi)/a_perp=0.032, far from the 173Yb parameters used in the fit. The authors should either provide the explicit second-band generalization or soften the quantitative claim of Sec. V; as written, the k=1.8k0 result is not established.","section":"Sec. V and footnote [27]"},{"comment":"The value k=1.8k0 is chosen by matching the computed peak positions to the experimental resonance peaks; it is not predicted independently from the model or from a measurement of k. Since the authors themselves list alternative mechanisms that could shift the effective k (many-body and multi-collision effects in footnote [28], and anharmonicity of the axial potential in the last paragraph of Sec. V), the conclusion that the finite-momentum effect is \"very significant\" and that the g-atoms may be in the second band is one consistent interpretation rather than a falsifiable prediction. The paper should quantify how robust the inferred k is against these uncertainties, for example by showing how much the peak locations move under reasonable variations of the model, or by comparing the thermal average in Fig. 8(d) with the experimental peak widths before drawing the strong conclusion.","section":"Sec. V, Fig. 8, and last paragraph"},{"comment":"The u-parameters in Eqs. (39)-(41) are obtained by matching only the zero-momentum scattering lengths a_e and a_o. The finite-momentum validity of the tight-binding models is checked for one parameter set and for k up to about 0.2k0, where the amplitudes agree; for k>0.2k0 the paper itself reports differences of order 0.05 in the even-wave amplitude. Because the models are used in Sec. V at k=1.8k0 and for 173Yb scattering lengths (a_s^(+) = 1878 a0) much larger than the tested value, the matching at threshold does not control the on-shell amplitudes used in the experimental fit. A statement about the expected size of the higher-order terms in Eqs. (36)-(37), or an additional numerical test at the experimental parameters, is needed to justify the application of the single-band tight-binding amplitudes at large k.","section":"Sec. IV B and Fig. 4"}],"minor_comments":[{"comment":"The abstract contains typos: \"we further exam\" should be \"we further examine\" and \"which dependents on\" should be \"which depends on\".","section":"Abstract"},{"comment":"The notation is sometimes unclear about whether k is the quasi-momentum or the characteristic exponent of the Mathieu equation; a short sentence clarifying the relation between k, the band index, and the first Brillouin zone would help readers, especially because the paper later uses k=1.8k0.","section":"Sec. IV A, Eq. (34) and Appendix A"},{"comment":"The numerical solution of the integral equation for eta(z) is described only by \"we numerically solve this equation\"; a brief description of the discretization, the number of grid points, and a convergence check would increase confidence in the quantitative results, particularly the projection-approximation error percentages quoted in Sec. IV C.","section":"Appendix A, Eq. (A22)"},{"comment":"The statement that s_e is approximately 3.3s_g for both 173Yb and 171Yb is used to relate experimental parameters, but the accuracy of this relation is not discussed; a sentence on its expected uncertainty would be useful for the comparison with Ref. [4].","section":"Sec. IV D"},{"comment":"The caption contains a typo: \"the the exact Hamiltonian\" should read \"the exact Hamiltonian\".","section":"Fig. 4 caption"},{"comment":"Footnote [27] contains substantive technical content about the second-band generalization and should be promoted to the main text or an appendix, with the promised equations.","section":"Footnote [27]"}],"recommendation":"major_revision","confidential_remarks":"The core tight-binding derivation and the projection-approximation analysis are sound and valuable; my concerns are concentrated in Sec. V. The stress-test concern about the second-band generalization lands: the manuscript relies on a footnote for a load-bearing step, and the finite-momentum fit is presented as a finding rather than as one possible interpretation. If the authors can supply the second-band scattering derivation or a numerical validation at the experimental parameters, and reframe the Sec. V conclusion accordingly, I would support publication."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Dear colleague,\n\nThe thing to know about this paper: it gives a believable microscopic derivation of tight-binding Kondo models for alkaline-earth atoms in a mixed-dimensional lattice, and the warning about the projection approximation is worth taking seriously. The second part, claiming a large finite-momentum effect in the Riegger experiment, is more fragile: the key number k=1.8k0 is fitted, not predicted, and the second-band extension lives entirely in a footnote.\n\nWhat's actually new: the u-parameters in Eqs. (39)-(41), obtained by matching the low-energy scattering lengths of the exact two-body problem to the lattice model. The derivation in Appendix A is detailed and internally consistent. They also test the projection approximation quantitatively and show it fails when the 3D scattering length is already ~10% of the confinement lengths. That's a concrete, useful result. The experimental section includes honest caveats about anharmonicity and many-body effects, which I appreciate.\n\nThe soft spots are in proportion. First, the experimental fit: k=1.8k0 is a free parameter chosen to match the observed peaks. The paper says the finite-momentum effect is \"very significant,\" but since the model is matched at zero momentum and validated for k up to ~0.2k0 in the lowest band, extrapolating to the second band at 1.8k0 requires the unstated generalization in footnote [27]. Maybe it's fine, but the reader can't check. Second, the tight-binding model's finite-momentum accuracy is only tested for one parameter set and small scattering length, far from the 173Yb parameters used in the fit. The authors' own caveats at the end of Sec. V lower the stakes: they explicitly say the real quasi-momentum might not be so high. So the experimental conclusion is suggestive, not settled. None of this undermines the central model derivation.\n\nWho should read it: anyone building lattice Kondo models with alkaline-earth atoms, and anyone using the projection approximation in quasi-low-dimensional systems. It deserves a serious referee; the experimental section needs either a real second-band derivation or a softer claim.\n\nI'd engage with it.","headline":"Solid microscopic derivation of lattice Kondo parameters; the experimental finite-momentum claim is a fit, not a prediction.","tokens_in":27286,"tokens_out":2948,"would_cite":true,"duration_ms":28750,"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":"This paper derives the interaction parameters of tight-binding Kondo models for alkaline-earth atoms in a mixed-dimensional lattice by matching the exact two-body scattering lengths.","keywords":["tight-binding Kondo model","alkaline-earth atoms","mixed-dimensional optical lattice","confinement-induced resonance","spin-exchange collision rate","scattering length matching","Mathieu Green's function","finite quasi-momentum effect"],"falsifier":"A decisive check is to compute the exact two-body scattering amplitude in the second Bloch band at $k=1.8k_0$ and compare it with model (II) using the matched parameters; a discrepancy beyond the few-percent level found at $k\\lesssim0.2k_0$ would show the zero-momentum matching does not control the regime used. Experimentally, a 173Yb gas cooled to the lowest band at known small quasi-momentum should show spin-flip resonance peaks at the small-$k$ lattice depths predicted by the theory, not at the $k=1.8k_0$ positions.","tokens_in":26168,"feed_emoji":"⚛️","tokens_out":10246,"duration_ms":95340,"temperature":0.7,"pith_summary":"The paper addresses how to describe ultracold alkaline-earth atoms in a mixed-dimensional trap—a mobile g-atom moving along an optical lattice and an e-atom localized in a harmonic well—by a tight-binding Kondo model. Its central claim is that the interaction parameters of such models can be derived microscopically by solving the exact two-body scattering problem and matching the low-energy even- and odd-wave scattering lengths. The resulting formulas make the spin-exchange coupling a calculable function of the lattice and trap parameters, with confinement-induced resonances providing control knobs. The same calculation shows that the commonly used projection approximation fails already when the three-dimensional scattering length is about 10 percent of the confinement lengths. Applied to a recent 173Yb experiment, the theory reproduces the observed spin-flip resonances only if the incident quasi-momentum is high, around $1.8k_0$, so finite-momentum effects are significant.","feed_headline":"Lattice Kondo parameters fixed by scattering lengths","feed_subtitle":"Matching scattering lengths makes spin exchange controllable and explains observed peaks.","key_machinery":"The central object is a pair of one-dimensional tight-binding Hamiltonians for a mobile g-atom on an optical lattice and a fixed e-atom impurity at site 0: model (I) with an on-site interaction $u_0$ and model (II) with an additional nearest-neighbor interaction $u_1$ at sites $\\pm1$. The argument is carried by a matching procedure. For each spin channel, the exact low-energy even- and odd-wave scattering lengths $a_{e,\\xi}$ and $a_{o,\\xi}$ are computed from the full Hamiltonian using a Mathieu-function Green's function for the lattice and a Lippmann-Schwinger equation regularized by the Huang-Yang pseudopotential; the same scattering lengths are computed for the tight-binding models by solving the lattice Schrödinger equation. Equating the two determines the $u$-parameters while explicitly including virtual transitions to transverse excited states, axial trap states, and higher lattice bands.","core_discovery":"The authors establish that the quasi-(1+0)D alkaline-earth system with an axial lattice is faithfully represented by two tight-binding Hamiltonians, provided the interaction parameters are fixed by exact low-energy scattering lengths rather than by naive projection. Model (I) keeps only a local interaction at the e-atom site and is valid when odd-wave scattering is negligible; model (II) adds nearest-neighbor interactions and reproduces both partial-wave channels. The central results are the parameter formulas $u_{0,\\xi}^{(I)}/t = -2l_0/a_{e,\\xi}$, $u_{0,\\xi}^{(II)}/t = -2(l_0^2-2l_0a_{o,\\xi}+a_{e,\\xi}a_{o,\\xi})/(l_0(a_{e,\\xi}-a_{o,\\xi}))$, and $u_{1,\\xi}^{(II)}/t = a_{o,\\xi}/(l_0-a_{o,\\xi})$, which are obtained by equating the scattering lengths of the exact Hamiltonian with those of the lattice models. These parameters can be tuned through the transverse and axial confinements and the lattice depth, and even- or odd-wave confinement-induced resonances appear where the spin-exchange terms are resonantly enhanced. The paper further shows that the spin-exchange collision rate in the 173Yb experiment depends strongly on the incident quasi-momentum, and that the observed resonance locations are reproduced only for $k\\simeq1.8k_0$, already inside the second Bloch band.","pith_inferences":["I infer that the same scattering-length matching could be applied to arrays with several e-impurities or to higher band occupations, provided three-body losses remain negligible, turning the computed parameter curves into a design map for Kondo-lattice quantum simulators.","I infer that a direct exact calculation of the two-body scattering amplitude in the second Bloch band would settle whether the zero-momentum matching controls the $k=1.8k_0$ regime; the paper asserts this extension only in a footnote.","I infer that the predicted failure of the projection approximation can be probed experimentally by tuning $a_s/a_\\perp$ through 0.1 and watching the interaction parameter diverge from the projected value, a sharp signature of virtual higher-band processes.","I infer that the nearest-neighbor term $u_1$ of model (II) provides a route to realize odd-wave Hubbard-type interactions in optical lattices by working near an odd-wave confinement-induced resonance."],"forward_implications":["The matched $u$-parameters turn each confinement-induced resonance into a quantitative input for the Kondo model, so the spin-exchange strength can be resonantly enhanced while the spin-independent term stays finite.","Model (II) reduces exactly to model (I) when the odd-wave scattering length vanishes, so the two models cover the even-wave-only and full cases without an adjustable switch.","Under the projection approximation the interaction parameter is off by about 70 percent already at $a_s/a_\\perp=0.1$, so quantitative simulations in this geometry need the matched parameters or direct calibration.","The resonance positions of the spin-exchange collision rate shift significantly with incident quasi-momentum; matching the 173Yb data requires $k\\simeq1.8k_0$, implying many g-atoms occupy the second lattice band and further cooling is needed for a single-band Kondo simulator.","Thermal averaging over the g-atom quasi-momentum distribution broadens the predicted collision-rate peaks toward the observed widths, giving an experimentally testable line shape."],"supporting_citations":[{"why":"It supplies the previous quasi-(1+0)D scattering treatment and the regularization of the singular Green's function that the present lattice calculation generalizes.","marker":"[3]"},{"why":"It provides the 173Yb experimental spin-flip data and the trap-lattice parameters used for the finite-quasi-momentum comparison.","marker":"[4]"},{"why":"It gives the 173Yb singlet and triplet three-dimensional scattering lengths used to compute the model parameters and collision rates.","marker":"[5]"},{"why":"It gives the 171Yb scattering lengths used for the model-parameter illustrations.","marker":"[6]"},{"why":"It introduces confinement-induced resonance, the mechanism by which the derived parameters become tunable.","marker":"[12]"},{"why":"It provides the harmonic-trap confinement-induced resonance analysis underlying the even- and odd-wave CIR interpretation.","marker":"[13]"},{"why":"It supplies updated 173Yb scattering lengths for the experimental parameter regime.","marker":"[21]"},{"why":"It supplies the Mathieu-function framework used to construct the lattice Green's function and Bloch states.","marker":"[22]"}],"fun_headline_variants":["Scattering lengths dictate lattice Kondo spin-exchange","Spin-exchange rates hinge on exact scattering lengths","Finite momentum explains experimental spin-exchange peaks","Tight-binding models calibrated by scattering lengths","Odd-wave resonances enhance spin-exchange in lattice"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The argument stands or falls on whether matching the zero-momentum scattering lengths also makes the lattice model accurate at the high quasi-momenta (up to $1.8k_0$, in the second Bloch band) where the model is then used.","fun_headline_variants_meta":{"raw":{"variants":["Scattering lengths dictate lattice Kondo spin-exchange","Spin-exchange rates hinge on exact scattering lengths","Finite momentum explains experimental spin-exchange peaks","Tight-binding models calibrated by scattering lengths","Odd-wave resonances enhance spin-exchange in lattice"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000263,"raw_usage":{"total_tokens":1747,"prompt_tokens":1242,"completion_tokens":505,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":858,"completion_tokens_details":{"reasoning_tokens":437}},"tokens_in":858,"tokens_out":505,"duration_ms":5937,"temperature":1.0,"reasoning_tokens":437,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T13:41:32.632795+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A decisive check is to compute the exact two-body scattering amplitude in the second Bloch band at $k=1.8k_0$ and compare it with model (II) using the matched parameters; a discrepancy beyond the few-percent level found at $k\\lesssim0.2k_0$ would show the zero-momentum matching does not control the regime used. Experimentally, a 173Yb gas cooled to the lowest band at known small quasi-momentum should show spin-flip resonance peaks at the small-$k$ lattice depths predicted by the theory, not at the $k=1.8k_0$ positions.","supporting_citations":[{"cited_title":"projection approximation","cited_arxiv_id":null,"evidence_quote":"It supplies the previous quasi-(1+0)D scattering treatment and the regularization of the singular Green's function that the present lattice calculation generalizes."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"It provides the 173Yb experimental spin-flip data and the trap-lattice parameters used for the finite-quasi-momentum comparison."},{"cited_title":"Zhang, D","cited_arxiv_id":null,"evidence_quote":"It gives the 173Yb singlet and triplet three-dimensional scattering lengths used to compute the model parameters and collision rates."},{"cited_title":"Cheng, R","cited_arxiv_id":null,"evidence_quote":"It gives the 171Yb scattering lengths used for the model-parameter illustrations."},{"cited_title":"Scazza, C","cited_arxiv_id":null,"evidence_quote":"It introduces confinement-induced resonance, the mechanism by which the derived parameters become tunable."},{"cited_title":"Zhang, M","cited_arxiv_id":null,"evidence_quote":"It provides the harmonic-trap confinement-induced resonance analysis underlying the even- and odd-wave CIR interpretation."},{"cited_title":"Bettelheim, Journal of Physics A: Mathematical and Theoretical 48, 165003 (2015)","cited_arxiv_id":null,"evidence_quote":"It supplies updated 173Yb scattering lengths for the experimental parameter regime."},{"cited_title":"Altland, B","cited_arxiv_id":null,"evidence_quote":"It supplies the Mathieu-function framework used to construct the lattice Green's function and Bloch states."}],"review_version":1}