{"id":"47c02463-0233-4b79-86f6-e471508e5bfb","arxiv_id":"2506.01831","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"Including the first excited electronic state in full-dimensional quantum scattering raises the computed K + KRb ultracold rate coefficient to about 1.32 x 10^-10 cm3/s, closer to the measured 1.7 x 10^-10 cm3/s.","lead":"This paper reports the first full quantum scattering calculation of the K + KRb reaction that includes the first excited electronic state, raising the computed ultracold rate coefficient from 0.9 to about 1.3 x 10^-10 cm3/s and bringing it into the error bar of the 2010 experiment. It matters because it indicates that non-adiabatic coupling and quantum interference at short range, not just the ground-state surface, can set ultracold chemical rates.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The claimed rate improvement (1.32 vs 0.9/1.1 ×10^-10 cm^3/s) is tested against a single experimental value using a scattering model with no hyperfine structure, an omission the authors admit 'cannot be neglected' for related systems; until a quantitative hyperfine estimate is supplied, the…","rationale":"The reader's weakest assumption correctly identifies the omitted hyperfine structure; I agree that is the load-bearing point. The paper is otherwise a plausible first non-adiabatic treatment: the 2×2 formalism, convergence checks in ρmax and ρmatch, and reproduction of the NGP value all support the dynamical methodology. But the central claim is quantitative agreement with a single experimental rate, and the comparison is not apples-to-apples in the entrance channel. The authors' own acknowledgment of hyperfine-dependent outcomes in K + NaK and Rb + KRb converts this from a routine approximation into a testable alternative explanation. A secondary concern is that the improvement is partly entangled with a new ground-state PES (their NGP rate is 0.9 vs Croft's 1.1 at 250 nK), so the figure should show their own NGP baseline; this reinforces the need for care but does not change the verdict. CONDITIONAL remains the right disposition: accept only if the hyperfine sensitivity is quantified or the claim is softened.","tokens_in":11669,"tokens_out":8776,"duration_ms":99942,"concrete_test":"Carry out a hyperfine-coupled close-coupling (or effective-boundary-condition) calculation for the experimental entrance state K |F=9/2,mF=-9/2> + KRb |mK=-4,mRb=3/2> at the relevant magnetic field, using the same ground-state PES and a hyperfine interaction for the entrance arrangement, and compare the zero-temperature loss rate with the 2×2 result of 1.32×10^-10 cm^3/s. If the hyperfine-resolved rate differs by more than about 0.4×10^-10 cm^3/s (i.e., outside the 2×2 vs NGP difference), the non-adiabatic explanation of the experimental gap is not established.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central comparison is to the Ospelkaus et al. rate measured for K in |F=9/2,mF=-9/2> and KRb in |mK=-4,mRb=3/2>. The scattering calculation (Eq. 1) contains only electronic-state channels and omits all hyperfine terms, as the authors state in the penultimate paragraph: 'it does not yet account for the hyperfine structure within the entrance channel', and they cite experiments showing that hyperfine-dependent outcomes 'cannot be neglected'. At ultracold temperatures the total loss rate is controlled by the short-range quantum boundary condition, and hyperfine interactions modify that condition by coupling entrance-channel spin states and by shifting thresholds; the cited Rb + KRb and K + NaK experiments show order-unity variations with hyperfine state. The claimed non-adiabatic effect is an increase of roughly 0.4×10^-10 cm^3/s (from the present NGP value 0.9 to the 2×2 zero-temperature limit 1.32), which is only slightly larger than the experimental error bar (±0.3). If a hyperfine-resolved calculation for the experimental initial state shifts the rate by a comparable amount, the agreement with experiment would be coincidental rather than evidence for non-adiabatic coupling. No estimate, model, or bound on the hyperfine shift for this specific system is provided.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript reports the first non-adiabatic quantum reactive scattering study of the ultracold K + KRb(X 1Σ+, v=0, j=0) -> Rb + K2(X 1Σg+, v', j') reaction. The authors construct new diabatic potential energy surfaces for the lowest two electronic states of K2Rb from MRCI/ECP calculations, propagate a 2x2 diabatic coupled-channel system in hyperspherical coordinates, and compare the resulting total rate coefficients with the NGP (no-geometric-phase) calculation of Croft et al. and with the experimental value of Ospelkaus et al. They obtain a zero-temperature 2x2 total rate of about 1.32 x 10^-10 cm3/s, closer to the measured 1.7 ± 0.3 x 10^-10 cm3/s than the previous NGP value (0.9–1.1 x 10^-10 cm3/s), and attribute the improvement to non-adiabatic coupling and quantum interference near a conical intersection. The authors explicitly acknowledge that the calculation does not include hyperfine structure in the entrance channel.","tokens_in":11973,"tokens_out":4432,"duration_ms":44131,"significance":"If the calculation is correct, it would be the first non-adiabatic treatment of this benchmark ultracold atom–diatom reaction and would indicate that short-range non-adiabatic effects can substantially close a long-standing theory–experiment gap. The work has notable strengths: no parameter is fitted to the experimental rate, the rate emerges from ab initio PESs and coupled-channel scattering, and the authors reproduce the earlier NGP calculation to within about 20%. However, the central quantitative claim is not yet fully established: the calculation omits hyperfine structure, the authors themselves cite experiments showing hyperfine-dependent effects that 'cannot be neglected', and the claimed non-adiabatic correction is comparable in size to the experimental uncertainty. In addition, the specific claim that non-adiabatic interference changes state-to-state product distributions is not directly demonstrated by any NGP-versus-2x2 state-to-state comparison.","major_comments":[{"comment":"The central comparison with the Ospelkaus et al. measurement is made against a scattering calculation that omits all hyperfine structure, as the authors state: 'it does not yet account for the hyperfine structure within the entrance channel.' At ultracold temperatures the total loss rate is controlled by the short-range quantum boundary condition, and the cited K + NaK and Rb + KRb experiments show hyperfine-dependent collisional outcomes of order unity. The non-adiabatic correction claimed here is an increase from 0.9 x 10^-10 to 1.32 x 10^-10 cm3/s, i.e. about 0.4 x 10^-10 cm3/s, which is only slightly larger than the experimental error bar of ±0.3 x 10^-10 cm3/s. Without a quantitative estimate or at least a bound on the hyperfine shift for the specific K + KRb initial state used in the experiment, the improved agreement cannot be uniquely attributed to non-adiabatic coupling; the paper's main conclusion is therefore not yet established.","section":"Penultimate paragraph (hyperfine caveat)"},{"comment":"The manuscript claims that 'short-range dynamics mediated by coupling with the excited electronic state introduces quantum interference effects that influence both the state-to-state rate coefficients and the overall reaction rates.' However, no NGP-versus-2x2 comparison of state-to-state rate coefficients is presented. Figure 3 shows only the 2x2 rotationally resolved rates at 211 nK, while Figure 4 compares only total rates. To support the state-to-state interference claim, the authors should provide state-to-state NGP and 2x2 rate coefficients (or an equivalent decomposition of the scattering amplitude into direct, loop, and interference terms) and show specifically where the interference changes sign or redistributes population.","section":"Abstract and Fig. 4"},{"comment":"The NGP baseline computed in this work is 0.9 x 10^-10 cm3/s at 250 nK, whereas the original Croft et al. value is 1.1 x 10^-10 cm3/s; the authors attribute this 18% discrepancy to differences in the adiabatic PES V1. This baseline uncertainty is roughly half the size of the claimed non-adiabatic correction (about 0.42 x 10^-10 cm3/s). The quantitative statement of 'better agreement with experiment' would be considerably strengthened by an error estimate for the present PESs or by repeating the 2x2 calculation with the original Croft et al. V1 to separate PES sensitivity from genuine non-adiabatic effects.","section":"'As a first step' paragraph and Fig. 4"}],"minor_comments":[{"comment":"The text reads 'the NATC matrix was evaluated'; this should be 'NACT matrix' (non-adiabatic coupling term).","section":"Page 6, electronic-structure paragraph"},{"comment":"The notation 'KRb−' appears to be a typographical artifact; the reaction is K + KRb, not K + KRb−. The minus sign is also inconsistent with the body of the paper.","section":"Title and abstract"},{"comment":"'spd fdiffuse functions' should read 'spdf diffuse functions'.","section":"Page 6, basis-set description"},{"comment":"The author name 'Fraschio' appears to be misspelled; please verify against the original publication.","section":"Reference 45"},{"comment":"The sentence 'presents every hundred V1 effective coupled potential curves ... compared to the first and the hundredth corresponding V2 curves' is unclear; please specify exactly which curves are shown and whether every 100th curve of V1 is plotted.","section":"Page 7, discussion of Fig. 2"}],"recommendation":"major_revision","confidential_remarks":"The paper is within the scope of the journal and presents a technically demanding calculation on an important benchmark reaction. The main barrier is the unquantified hyperfine omission: the authors themselves concede that hyperfine effects 'cannot be neglected' for related systems, and the claimed non-adiabatic correction is comparable to the experimental error bar. I would encourage the editor to request either a quantitative hyperfine estimate (or a carefully bounded claim) and a direct state-to-state NGP/2x2 comparison before reconsidering for publication."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Dear colleague,\n\nI read the K+KRb non-adiabatic paper. Here's the short version.\n\nThe paper is the first non-adiabatic quantum scattering calculation for the ultracold K+KRb reaction, built on new MRCI/ECP diabatic surfaces for the two lowest electronic states of K2Rb. The method is not new — the same 2x2 formalism was applied to Li+LiNa — but the application to this benchmark system is, and the physics is different enough to matter. The zero-temperature total rate lands at about 1.32 x 10^-10 cm3/s, clearly closer to the measured 1.7 ± 0.3 x 10^-10 than the earlier ground-state-only result (0.9–1.1 x 10^-10). They also reproduce the NGP result at higher collision energies, and the convergence checks on rho_max are careful. That is real progress.\n\nThe main weakness is the one the authors themselves flag at the end: the calculation omits all hyperfine structure in the entrance channel, even though the experiment uses a specific hyperfine state and recent work on similar systems shows hyperfine effects that \"cannot be neglected.\" The non-adiabatic shift they attribute to coupling is about 0.4 x 10^-10, only slightly larger than the experimental error bar. If a hyperfine-resolved calculation for the experimental initial state shifts the rate by a comparable amount, the improved agreement could be coincidental. I don't think that kills the paper—it is honest about the limitation—but it does make the abstract's \"better agreement\" claim more conditional than it sounds.\n\nA second, quieter issue: the paper says non-adiabatic coupling introduces quantum interference effects that influence state-to-state rates, but there is no direct NGP vs 2x2 state-to-state comparison. In fact, they say product vibrational populations are \"somewhat similar\" to the earlier NGP work, which undercuts the interference story. A side-by-side plot would help a lot.\n\nThere is also no uncertainty quantification: no error bars on the PESs, the basis, or the hyperfine shift. For a single value compared to a single experimental number, that is a real gap.\n\nWho is this for? Anyone working on ultracold atom-molecule collisions or non-adiabatic effects in chemical dynamics. It deserves a serious referee. I would send it to peer review, and ask the authors to either provide an estimate of hyperfine effects or soften the claim that the non-adiabatic treatment explains the experimental discrepancy.","headline":"A solid first non-adiabatic calculation for K+KRb that improves agreement with experiment, but the hyperfine-free comparison makes the headline claim less clean than it looks.","tokens_in":12532,"tokens_out":4302,"would_cite":true,"duration_ms":39890,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":["34.50.-s","34.50.Lf"],"model":"deepseek-v4-flash","headline":"Non-adiabatic coupling to an excited electronic state brings the ultracold K + KRb reaction rate into line with experiment.","keywords":["ultracold chemistry","non-adiabatic quantum dynamics","conical intersection","geometric phase","K + KRb reaction","rate coefficients","quantum interference","coupled-channel scattering"],"falsifier":"A hyperfine-resolved scattering calculation or measurement using the same initial hyperfine states that changes the zero-temperature rate by roughly 40% or more would show that the agreement is not caused by the non-adiabatic coupling. Alternatively, a product-state-resolved measurement of the K2 rotational distribution below 1 μK, if it matched the single-surface prediction rather than the two-state interference pattern, would rule out the claimed mechanism.","tokens_in":11495,"feed_emoji":"🧪","tokens_out":7430,"duration_ms":72032,"temperature":0.7,"pith_summary":"The paper reports the first non-adiabatic quantum scattering study of the ultracold K + KRb → Rb + K2 reaction, the first atom-diatom reaction measured below 1 μK. Previous single-surface calculations that ignored the first excited electronic state produced a rate coefficient about 35% below the measured value. By coupling the ground and first excited electronic states through a 2×2 diabatic treatment, the authors obtain a zero-temperature total rate of roughly 1.32×$10^{-10}$ cm3/s, close to the lower edge of the measured 1.7±0.3×$10^{-10}$ cm3/s. They argue that short-range dynamics near a conical intersection introduces quantum interference that changes both the overall rate and product state-to-state distributions, and that such non-adiabatic effects are confined to collision energies below about a millikelvin.","feed_headline":"Excited-state coupling closes ultracold K+KRb gap","feed_subtitle":"Adding the first excited electronic state lifts the zero-temperature rate to 1.32e-10 cm3/s, inside the experimental error bar.","key_machinery":"The central object is a 2×2 diabatic representation of the time-independent Schrödinger equation, written in hyperspherical coordinates, in which the ground and first excited electronic states become the diagonal potentials and are connected by an off-diagonal coupling. The diabatic potentials are built by transforming ab initio adiabatic surfaces using an adiabatic-to-diabatic transformation angle obtained by integrating the non-adiabatic coupling term. In the region where the excited-state channels are closed but strongly coupled, this 2×2 system is propagated, then transformed back to a single-surface form at a large hyperradius. Physically, the scattering amplitude splits into a direct path and a path encircling the conical intersection; the non-adiabatic treatment changes the sign of their interference cross term, converting constructive interference into destructive interference and vice versa.","core_discovery":"The central claim is that the long-standing theory-experiment gap for the ultracold K + KRb reaction is mostly accounted for by non-adiabatic coupling to the first excited electronic state, which is energetically accessible even in the zero-temperature limit. In a conventional Born-Oppenheimer no-geometric-phase calculation the total rate at 250 nK is about 0.9×$10^{-10}$ cm3/s; including the excited state in a two-state coupled-channel calculation raises it to about 1.32×$10^{-10}$ cm3/s, within the experimental error bars of the 2010 measurement. The authors further claim that the coupling flips the sign of the quantum interference term between the direct reactive path and the path that loops around the conical intersection, so the effect is genuinely quantum mechanical and grows as the temperature approaches zero.","pith_inferences":["The authors' own caveat implies the natural next test is a hyperfine-resolved version of this calculation; until that exists, the agreement with experiment should be read as conditional.","If the mechanism is general, other ultracold alkali atom + alkali dimer reactions with accessible excited states may show similar non-adiabatic corrections, and the present treatment offers a template for computing them.","A product-state-resolved ultracold experiment on K + KRb could distinguish the two-state interference pattern from the single-surface one, giving a direct observable signature of the geometric-phase sign flip."],"forward_implications":["At zero temperature the 2×2 total rate converges to about 1.32×10^-10 cm3/s, versus about 0.9×10^-10 cm3/s in the no-geometric-phase calculation, placing theory at the lower edge of the experimental 1.7±0.3×10^-10 cm3/s.","Above roughly 1 mK the two-state and single-surface calculations agree, so non-adiabatic interference is an ultracold-limit phenomenon tied to the Wigner threshold regime.","The products remain dominated by highly rotationally excited K2 in the v'=0 and 1 manifolds, meaning the main non-adiabatic effect is on the magnitude and interference pattern of the rate rather than on the gross energy disposal.","Short-range coupling to the excited electronic state, not only long-range forces, must be included to reproduce ultracold reactive rates, with consequences for other near-universal-loss bialkali systems."],"supporting_citations":[{"why":"Supplies the experimental rate coefficient 1.7±0.3×10^-10 cm3/s and the hyperfine-state preparation that the theory is compared against.","marker":"[18]"},{"why":"Provides the previous ground-electronic-state no-geometric-phase quantum scattering result and the ~35% shortfall that motivates the new calculations.","marker":"[35]"},{"why":"Original full-dimensional J=0 quantum scattering calculations on the ground electronic state, establishing the baseline rate coefficient.","marker":"[20]"},{"why":"Demonstrates the non-adiabatic two-state treatment on the analogous Li + LiNa reaction, supplying the methodological precedent.","marker":"[8]"},{"why":"Presents the 2×2 non-adiabatic coupled-channel scattering formalism in hyperspherical coordinates used throughout the paper.","marker":"[26]"},{"why":"Describes how non-adiabatic coupling terms are evaluated to construct the adiabatic-to-diabatic transformation angle.","marker":"[43]"},{"why":"Reports hyperfine-dependent atom-molecule loss outcomes that motivate the caveat that hyperfine structure cannot be neglected in the entrance channel.","marker":"[52]"},{"why":"Documents hyperfine-to-rotational energy transfer in related atom-molecule collisions, further supporting the hyperfine caveat.","marker":"[53]"}],"fun_headline_variants":["Non-adiabatic path fixes ultracold reaction rates","Quantum interference lifts K+KRb rate to experiment","Excited state bridges ultracold chemistry gap","Interference flips sign, matches ultracold experiment","Two-state model hits experimental K+KRb rate"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The calculation omits hyperfine structure in the entrance channel, so it assumes the measured rate, taken with K and KRb in specific hyperfine states, can be compared to a scattering calculation without any hyperfine resolution; if hyperfine effects shift the rate by tens of percent, the improved agreement could be coincidental.","fun_headline_variants_meta":{"raw":{"variants":["Non-adiabatic path fixes ultracold reaction rates","Quantum interference lifts K+KRb rate to experiment","Excited state bridges ultracold chemistry gap","Interference flips sign, matches ultracold experiment","Two-state model hits experimental K+KRb rate"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000196,"raw_usage":{"total_tokens":1323,"prompt_tokens":870,"completion_tokens":453,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":486,"completion_tokens_details":{"reasoning_tokens":375}},"tokens_in":486,"tokens_out":453,"duration_ms":4535,"temperature":1.0,"reasoning_tokens":375,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-07T11:33:16.910278+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A hyperfine-resolved scattering calculation or measurement using the same initial hyperfine states that changes the zero-temperature rate by roughly 40% or more would show that the agreement is not caused by the non-adiabatic coupling. Alternatively, a product-state-resolved measurement of the K2 rotational distribution below 1 μK, if it matched the single-surface prediction rather than the two-state interference pattern, would rule out the claimed mechanism.","supporting_citations":[{"cited_title":"K.; Balakrishnan, N","cited_arxiv_id":null,"evidence_quote":"Demonstrates the non-adiabatic two-state treatment on the analogous Li + LiNa reaction, supplying the methodological precedent."}],"review_version":1}