{"id":"3355cc79-a45c-4720-b70f-486d41a48afe","arxiv_id":"1908.00809","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":3,"one_line_summary":"A 12D variational computation on CH4·F- reports converged vibrational band origins and tunneling splittings, showing small splittings at low energies, increasing splittings upon excitation, and sizeable heavy-atom tunneling near the barrier.","lead":"This paper computes the full 12-dimensional quantum vibrational states of the complex formed by a methane molecule and a fluoride ion, using a variational method with a sparse grid. It finds that tunneling splittings between the four equivalent binding sites are very small at low energies but grow with excitation, and that a full-dimensional treatment is needed to get the energies right.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Excited-state tunneling splitting claim relies on an undemonstrated error-cancellation between the small b=2 methane basis and the large angular basis; the abstract's <0.02 cm^-1 statement also conflicts with the body's own >0.02 cm^-1 splittings above 680 cm^-1.","rationale":"The reader's weakest assumption is the accuracy of the CBB08 PES. That is a real limitation for statements about the real CH4·F- complex, but it is largely external to the paper's central claim, which is explicitly about vibrational states on this PES; the authors also acknowledge the PES's limited coverage. The more load-bearing concern is internal to the numerical argument: the excited-state tunneling splittings are computed with the b=2 methane basis, and the 0.02 cm^-1 convergence is justified only by an error-cancellation argument that is not directly demonstrated. This concern targets the central claims about the increase of splittings with excitation and the onset of heavy-atom tunneling. In addition, the abstract's '<0.02 cm^-1' statement conflicts with Section V's reported splittings above 0.02 cm^-1, creating an internal inconsistency that should be fixed before publication. Because the concern is substantial but not fatal—it can be tested by a larger-b computation or by clearer convergence evidence—the reader's CONDITIONAL verdict remains appropriate. The PES-fidelity concern raised by the reader is valid as an external limitation, so I partially agree rather than fully disagree.","tokens_in":13300,"tokens_out":5973,"duration_ms":59006,"concrete_test":"Recompute the BT-style 12D tunneling splittings for the states between 680 and 856 cm^-1 using b=3 (and, if feasible, b=4) while keeping the 31-point sin-cot DVR for cosθ and the 45-function Fourier basis for φ; compare each splitting with the b=2 values in Figure 2 and the Supplementary Material. If any splitting shifts by more than 0.02 cm^-1, the claimed 0.02 cm^-1 convergence of excited-state splittings is not established. Also re-run the b=2 calculation with a different R basis size to test whether the radial basis participates in the alleged error cancellation.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central numerical claim—converged 12D tunneling splittings up to 856 cm^-1, including the 'sizeable' splittings near the barrier—is carried by the BT calculation in Section V. BT uses b=2 pruning of the 9D methane basis, and the authors state that band origins then carry 3–4 cm^-1 uncertainty, while the splittings are converged to 0.02 cm^-1 through error cancellation. No comparison is shown between b=2 and b=3/4 splittings at a fixed large angular basis; the only b-convergence data in Table I use a small angular DVR (23/25 sin-cot functions) and show shifts up to about 2 cm^-1 in band origins and artificial splittings up to 0.2–0.3 cm^-1. Thus the assertion that methane-basis error cancels in near-degenerate tunneling manifolds is plausible but unverified. If it fails, the headline conclusions that 'tunneling splittings increase upon vibrational excitation' and that 'heavy-atom tunneling' appears above ~680 cm^-1 are not established. A second internal inconsistency: the abstract states that tunneling splittings are smaller than 0.02 cm^-1, while Section V reports several splittings larger than 0.02 cm^-1 above ca. 680 cm^-1. This is not merely cosmetic: the abstract's unqualified claim is false under the paper's own numbers, and the body's Figure 2 labels sub-0.02 splittings as possibly numerical artifacts, so the convergence threshold and the physical conclusions attached to it need to be reconciled.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"This paper reports full-dimensional (12D) variational vibrational computations for the CH4·F− complex on the CBB08 potential energy surface, using the GENIUSH program with Smolyak quadrature. The authors compute vibrational band origins up to about 730 cm−1 above the zero-point energy and tunneling splittings up to about 856 cm−1, with stated convergence of 1 cm−1 for band origins and 0.05 cm−1 (or better) for splittings. They compare their results with earlier MCTDH, MULTIMODE, and reduced-dimensionality GENIUSH calculations, and conclude that tunneling splittings are very small for low-lying states, that a multi-well full-dimensional treatment is necessary despite the small splittings, and that the splittings increase with vibrational excitation, leading to non-negligible 'heavy-atom' tunneling near the barrier.","tokens_in":13585,"tokens_out":9931,"duration_ms":95749,"significance":"If the reported calculations are correct, this work provides a valuable benchmark for a challenging 12D, four-well vibrational problem with large-amplitude motion. Its strengths are the systematic basis-pruning and grid-convergence studies (b=2,3,4; Nc=23 and 25; and a larger BT angular basis), the honest reporting of artificial splittings, the explicit labeling of the fitted 3D(Gfit) model rather than disguising its adjustable parameters, and the detailed comparison with previous methods. The results should be useful as numerical benchmarks on the CBB08 PES, with the important caveat that this surface itself has an rmsd of 42 cm−1 to the underlying CCSD(T) energies, so the numbers are benchmarks for the surface rather than direct experimental predictions.","major_comments":[{"comment":"The central new conclusion that tunneling splittings become significant above about 680 cm−1 rests on the BT 12D calculation, which uses b=2 pruning of the methane basis. The paper states that although the band origins have 3–4 cm−1 uncertainty, the splittings are converged to better than 0.02 cm−1 because the methane-basis error cancels in the near-degenerate tunneling manifold. This error-cancellation assumption is plausible but is not demonstrated. Table I provides b=2, b=3, and b=4 comparisons only with a small angular basis (23 or 25 sin-cot functions), and that table shows artificial splittings of 0.2–0.3 cm−1 for some higher states at b=2 and b=3; the larger angular basis in BT tests angular convergence but not the assumed cancellation of the methane-basis truncation. Because the claims that 'tunneling splittings increase upon vibrational excitation' and that 'heavy-atom tunneling' appears are carried by the BT splittings, the paper should provide a direct comparison of b=2 and b=3 (or b=4) splittings at a fixed large angular basis, at least for representative states above 680 cm−1.","section":"Section V (Excited-state tunneling manifolds)"},{"comment":"The abstract states as a confirmed result that 'the tunneling splittings are smaller than 0.02 cm−1.' This unqualified statement is contradicted by the body of the paper: Section V reports that several splittings larger than 0.02 cm−1 appear above about 680 cm−1, Figure 1 highlights these states in red, and Section VI summarizes 'sizeable tunneling splittings, >0.1 cm−1' near the barrier. The abstract should be qualified (for example, by limiting the statement to states below about 680 cm−1 or to the low-energy manifold), and the convergence threshold should be reconciled consistently between the abstract, Section V, Figure 1, and the summary.","section":"Abstract and Section V/Figure 1"}],"minor_comments":[{"comment":"In Table I, the ZPVE row for the two 3D models (461.0 and 378.8 cm−1) is not on the same absolute energy scale as the 12D ZPVE values, because the 3D models exclude methane internal zero-point energy; a footnote explaining this difference would prevent confusion.","section":"Table I and Section IV"},{"comment":"The sentence stating that the barrier is '1104 cm−1 from the PES minimum and ca. 700 cm−1 measured from the intermolecular ZPVE' is potentially confusing because Figure 1 plots energies measured from the full 12D zero-point vibration; please clarify which model's ZPVE the 700 cm−1 value refers to.","section":"Section V"},{"comment":"The main text does not contain a numerical table of the BT excited-state tunneling splittings, even though those splittings drive the main conclusions; including the BT energies and splittings for the states above 680 cm−1 in the main text (or at least a clear pointer to the specific supplementary table) would strengthen the presentation.","section":"Section V and Supplementary Material"}],"recommendation":"major_revision","confidential_remarks":"The manuscript is within scope for the journal. The main technical concern is the unverified error-cancellation in the BT basis; the authors should be asked to either provide the b=3 convergence check or soften the claims. The abstract inconsistency is fixable locally but must be corrected. No concerns about citation practice or novelty disclosure."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"What you should know: this is a serious piece of computational quantum dynamics. The authors report the first full-dimensional (12D) variational vibrational computation for CH4·F- using GENIUSH with Smolyak quadrature, and they get converged band origins and tunneling splittings on the published CBB08 surface. The core qualitative result—tunneling splittings are very small below roughly 700 cm^-1 and become sizeable near the barrier—looks like it holds. But the paper's own abstract overstates the smallness claim, and the convergence evidence for the excited-state splittings is thinner than the text admits.\n\nWhat is genuinely good: the methodology is established, the basis-set and grid convergence is checked systematically (b = 2, 3, 4; Nc = 23 and 25), the angular basis is taken to near saturation for the splitting calculations, and the uncertainty estimates are honest. The comparison with the earlier MCTDH and MULTIMODE results is careful, and the discrepancies are explained sensibly. The only fitted quantity is the explicitly labelled 3D(Gfit) model, which is used as a diagnostic; the 12D results themselves are direct variational solutions on an externally published PES. That is reproducible, checkable work.\n\nSoft spots, in order of importance. First, the abstract states that tunneling splittings are smaller than 0.02 cm^-1, while Section V and Figure 2 report several splittings larger than 0.02 cm^-1 above about 680 cm^-1, and Figure 2 even labels sub-0.02 splittings as possibly numerical artifacts. That is an internal contradiction in the paper's central claim. It is probably a careless wording—the body text is more careful—but it needs to be fixed. Second, the excited-state splittings are computed with the BT basis, which uses a small methane basis (b = 2) and a large angular basis. The authors argue the methane-basis error cancels in near-degenerate tunneling manifolds, so the splittings are converged to 0.02 cm^-1 even though the band origins have 3–4 cm^-1 uncertainty. That is plausible, but it is not demonstrated. No b = 2 versus b = 3/4 comparison is shown at a fixed large angular basis. If that cancellation fails, the \"splittings increase with excitation\" conclusion is on shakier ground. The paper would be much stronger with one direct check. Third, the ZPVE is 5.1 cm^-1 above the MCTDH variational upper bound. The authors call the computation \"near-variational\" and claim 1 cm^-1 convergence for band origins, but this gap means absolute energies are not benchmark-quality; only relative energies are. The benchmark claim should be qualified.\n\nNone of these are fatal. The central variational computation is sound, and the qualitative physics is probably right. The paper deserves a serious referee. I would send it out and ask for a revision that fixes the abstract, adds a splitting-convergence check against the methane basis, and states the ZPVE discrepancy plainly.","headline":"Serious 12D variational study with a plausible core result, but the abstract overstates the tunneling-splitting smallness and the excited-state splitting convergence rests on an unverified error-cancellation argument.","tokens_in":14204,"tokens_out":2654,"would_cite":true,"duration_ms":25338,"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":"A full-dimensional 12D variational computation determines the vibrational states of CH4·F−, showing that tunneling splittings stay tiny below 730 cm⁻¹ and grow near the 1104 cm⁻¹ barrier.","keywords":["CH4·F− complex","full-dimensional vibrational computation","variational quantum dynamics","tunneling splittings","multiple wells","GENIUSH-Smolyak","anharmonicity","molecular symmetry"],"falsifier":"An infrared spectrum of CH4·F− resolving the region up to 730 cm$^{-1}$ above the zero-point level: any tunneling splitting larger than 0.05 cm$^{-1}$ in that range would contradict the central numerical claim. Alternatively, a 12D variational calculation on an independently fitted potential surface that differs from these band origins by more than about 2 cm$^{-1}$ would show the results are artifacts of the CBB08 surface.","tokens_in":12946,"feed_emoji":"⚛️","tokens_out":15751,"duration_ms":128830,"temperature":0.7,"pith_summary":"In this work, the authors use a full-dimensional (12D) variational vibrational calculation, based on the GENIUSH–Smolyak algorithm and the CBB08 potential energy surface, to determine the vibrational band origins and tunneling splittings of the CH4·F− ion-molecule complex. They aim to settle earlier discrepancies among full- and reduced-dimensionality studies. The computation achieves converged band origins within about 1 cm$^{-1}$ and tunneling splittings within 0.05 cm$^{-1}$ (0.02 cm$^{-1}$ for excited-state splittings). The results confirm that tunneling splittings are smaller than 0.02 cm$^{-1}$ below roughly 730 cm$^{-1}$ above the zero-point level, and that a single-well description fails because of significant multi-well anharmonicity. They also show that splittings increase with vibrational excitation and become sizeable, around 0.1–0.3 cm$^{-1}$, near the 1104 cm$^{-1}$ barrier, indicating 'heavy-atom' tunneling.","feed_headline":"12D quantum computation settles CH4·F− vibrational levels","feed_subtitle":"Band origins are converged to ~1 cm⁻¹; tunneling splittings stay under 0.05 cm⁻¹ up to 730 cm⁻¹, then grow.","key_machinery":"The load-bearing machinery is the GENIUSH–Smolyak algorithm, a combination of the numerical kinetic energy operator (KEO) approach with a Smolyak (non-product) quadrature grid, which makes the 12D variational computation tractable. Methane's nine internal modes are described with normal coordinates and a harmonic-oscillator basis pruned by the condition $\\sum_{i=1}^{9} n_{q_i} \\le b$, while the three intermolecular coordinates ($R$, $\\cos\\theta$, $\\phi$) keep a full direct-product basis; the $\\cos\\theta$ degree is treated with a sin-cot discrete variable representation (DVR) to avoid singular KEO terms. Convergence is established by increasing the pruning parameter $b$ (2, 3, 4) and the angular grid size, giving the claimed 1 cm$^{-1}$/0.05 cm$^{-1}$ accuracy. The symmetry analysis uses the $T_d(M)$ molecular symmetry group with the local $C_{3v}$ irreps lifted as $\\Gamma(A_1^{C_{3v}}) = A_1 \\oplus F_2$, $\\Gamma(A_2^{C_{3v}}) = A_2 \\oplus F_1$, and $\\Gamma(E^{C_{3v}}) = E \\oplus F_1 \\oplus F_2$, which converts computed eigenvalues into assigned tunneling manifolds.","core_discovery":"The central claim is that a genuinely full-dimensional, multi-well variational treatment is both necessary and sufficient to describe the low-energy vibrational states of CH4·F− on the CBB08 surface. The 12D GENIUSH–Smolyak computation yields benchmark-quality band origins, converged to within 1 cm$^{-1}$, and tunneling splittings converged to within 0.05 cm$^{-1}$ (excited-state splittings within 0.02 cm$^{-1}$), for all states up to about 730 cm$^{-1}$ above the zero-point vibrational energy. Up to that energy the computed splittings are smaller than 0.02 cm$^{-1}$, so the four equivalent wells are not resolved as spectral splittings; nevertheless, a single-well normal-coordinate computation disagrees with the multi-well band origins by about 20 cm$^{-1}$ (rmsd), showing that the wavefunctions feel the multi-well anharmonicity even when the tunneling splitting is tiny. In the higher-energy range, the splittings increase with vibrational excitation and reach values larger than 0.1 cm$^{-1}$ near the lowest electronic barrier (1104 cm$^{-1}$), which the authors interpret as non-negligible tunneling of the heavy (methane–fluoride) relative motion. These results reconcile the earlier 3D and 12D studies: the 3D model that reproduces the band origins fails to reproduce the splittings by orders of magnitude, while the rigid-monomer 3D model gives reasonable splittings but poor band origins, so only the full-dimensional treatment captures both.","pith_inferences":["If the CBB08 surface is accurate, the predicted onset of sizeable splittings around 680–730 cm$^{-1}$ above the zero-point level is a sharp, observable prediction for action-spectroscopy experiments on CH4·F−.","The same pruned-basis Smolyak strategy could be applied to other four-well complexes such as CH4·Cl−, where the pattern of 3D-versus-12D disagreement would likely be similar.","The 1104 cm$^{-1}$ barrier height is the key control parameter; recomputing the splittings on a surface with a deliberately modified barrier (e.g., ±50 cm$^{-1}$) would quantify how sensitively the near-threshold splittings respond, which the paper does not report.","One could also compute the tunneling path non-perturbatively from the 12D wavefunctions (for example, the nodal surfaces of the relative-orientation coordinates) to verify the 'heavy-atom tunneling' interpretation; the paper assigns splittings but does not map the path."],"forward_implications":["The computed band origins up to 730 cm$^{-1}$ provide a benchmark that future approximate quantum-dynamics treatments of this complex can be tested against.","Infrared spectra below 730 cm$^{-1}$ should show unsplit (or nearly unsplit) vibrational bands, while spectra excited near or above the barrier should develop resolvable tunneling multiplets.","Single-well normal-coordinate calculations, despite giving reasonable zero-point energies, are not reliable for vibrational excitation energies of this complex.","No fixed-monomer 3D reduction can simultaneously reproduce both band origins and tunneling splittings; accurate results require explicit coupling of methane's vibrations to the hindered rotation of the ion.","Near the 1104 cm$^{-1}$ barrier, tunneling splittings of order 0.1 cm$^{-1}$ appear, which is the signature of heavy-atom tunneling in the relative methane–fluoride motion."],"supporting_citations":[{"why":"Supplies the full-dimensional CBB08 potential energy surface used as the sole input for all computed energies.","marker":"[8]"},{"why":"Previous full-dimensional computation that provides the comparison set and the claimed variational upper bound.","marker":"[9]"},{"why":"Previous reduced-dimensionality computation that first reported small tunneling splittings and is used for comparison.","marker":"[10]"},{"why":"Describes the GENIUSH–Smolyak algorithm and kinetic energy operator that the present full-dimensional computation applies.","marker":"[18]"},{"why":"Introduces the Smolyak non-product quadrature grid that makes the high-dimensional integrals tractable.","marker":"[19]"},{"why":"Provides the GENIUSH program with the numerical kinetic energy operator used for the reduced-dimensionality computations.","marker":"[14]"},{"why":"Defines the sin-cot discrete variable representation used for the cosθ coordinate to avoid singular kinetic energy terms.","marker":"[31]"},{"why":"Supplies the molecular symmetry group analysis adapted for assigning the tunneling manifolds in Td(M).","marker":"[34]"}],"fun_headline_variants":["12D variational benchmark for CH4·F− vibrational levels","Full 12D treatment reveals CH4·F− tunneling splittings","Heavy-atom tunneling in CH4·F− from 12D quantum levels","12D computation benchmarks CH4·F− multi-well vibrational states"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The single fitted potential energy surface (CBB08), with a reported 42 cm$^{-1}$ root-mean-square fitting error, accurately represents the true CH4·F− interaction over the coordinate regions sampled by the vibrational states.","fun_headline_variants_meta":{"raw":{"variants":["12D variational benchmark for CH4·F− vibrational levels","Full 12D treatment reveals CH4·F− tunneling splittings","Heavy-atom tunneling in CH4·F− from 12D quantum levels","12D computation benchmarks CH4·F− multi-well vibrational states"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000966,"raw_usage":{"total_tokens":4178,"prompt_tokens":1077,"completion_tokens":3101,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":693,"completion_tokens_details":{"reasoning_tokens":3022}},"tokens_in":693,"tokens_out":3101,"duration_ms":22513,"temperature":1.0,"reasoning_tokens":3022,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T15:33:06.002923+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"An infrared spectrum of CH4·F− resolving the region up to 730 cm$^{-1}$ above the zero-point level: any tunneling splitting larger than 0.05 cm$^{-1}$ in that range would contradict the central numerical claim. Alternatively, a 12D variational calculation on an independently fitted potential surface that differs from these band origins by more than about 2 cm$^{-1}$ would show the results are artifacts of the CBB08 surface.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the full-dimensional CBB08 potential energy surface used as the sole input for all computed energies."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Previous reduced-dimensionality computation that first reported small tunneling splittings and is used for comparison."},{"cited_title":"Meyer, F","cited_arxiv_id":null,"evidence_quote":"Describes the GENIUSH–Smolyak algorithm and kinetic energy operator that the present full-dimensional computation applies."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Introduces the Smolyak non-product quadrature grid that makes the high-dimensional integrals tractable."},{"cited_title":"Czak´ o, B","cited_arxiv_id":null,"evidence_quote":"Provides the GENIUSH program with the numerical kinetic energy operator used for the reduced-dimensionality computations."},{"cited_title":"F´ abri, J","cited_arxiv_id":null,"evidence_quote":"Defines the sin-cot discrete variable representation used for the cosθ coordinate to avoid singular kinetic energy terms."},{"cited_title":"Sarka, A","cited_arxiv_id":null,"evidence_quote":"Supplies the molecular symmetry group analysis adapted for assigning the tunneling manifolds in Td(M)."}],"review_version":1}