{"id":"25be6526-42d1-4372-b9ca-13b5367d8412","arxiv_id":"2412.21060","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":2,"one_line_summary":"Tuned hybrid-functional calculations reproduce the optical absorption and emission of most LiF vacancy color centers to about 100 nm, with per-defect Koopmans parameter sets and a triplet-state assignment for F3- and F3+ emission.","lead":"This computational study uses tuned hybrid density functional theory to compute the optical and spin properties of vacancy color centers in lithium fluoride, finding that most defect transitions match experiment to within about 100 nanometers. The work tests whether a lightweight hybrid-functional approach can reliably predict color center properties in polar materials, relevant for quantum sensing and single-photon applications.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The half-electron ΔSCF 'singlet' is a spin-mixed ensemble, not a pure singlet; its uncontrolled bias is comparable in magnitude to the acknowledged F2+/F3- failures.","rationale":"The reader's weakest_assumption correctly identifies the constrained-occupation half-electron ΔSCF method as the load-bearing approximation. My stress-test sharpens this: the half-electron occupation is not merely 'crude' but is a spin-ensemble state with a known, unquantified contamination by ground-state and triplet determinants. The paper provides singlet–triplet splittings that can be used to estimate the bias magnitude, and the two exceptions (F2+, F3-) are consistent with such a bias. No independent excited-state method is benchmarked, so the ~100 nm accuracy claim for the remaining centers is not yet secured. This does not change the reader's CONDITIONAL verdict; it reinforces the need for the requested error quantification and a reference excited-state calculation. The concrete test proposed (TDDFT/BSE or singlet–triplet extraction) would directly settle whether the half-electron representation introduces a >100 nm error.","tokens_in":14622,"tokens_out":7762,"duration_ms":79331,"concrete_test":"Compute the vertical absorption of the F2 center at the singlet ground-state geometry with a method beyond ΔSCF, e.g., TDDFT or BSE using the same HSE functional and 216-atom supercell, and compare with the reported 2.475 eV (501 nm). Alternatively, perform a spin-polarized integer ΔSCF calculation for the triplet vertical excitation at the same geometry and extract the pure singlet energy as E_singlet = 2×E_half − E_triplet. If the reference singlet absorption differs from 501 nm by more than ~100 nm, the half-electron ensemble bias is the dominant error and the central accuracy claim must be re-evaluated; if it agrees within the claimed tolerance, the approximation is validated.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central accuracy claim relies on the constrained-occupation excited states described in 'Details of optical properties' and Table I. To avoid charge-transfer instabilities, the authors model spin-conserving singlet excitations by placing half an electron in the LUMO of each spin channel (Fig. 5). This half-electron state is not a pure singlet: it is an ensemble of Slater determinants whose energy includes contributions from the ground-state, triplet, and double-excitation configurations. The deviation from the true singlet excitation energy is therefore uncontrolled and can be as large as half the singlet–triplet splitting. The paper itself reports sizable singlet–triplet splittings for the relevant centers (290 meV for F2, 266 meV for F3-), implying a possible bias of ~0.1–0.15 eV, i.e., tens of nm in the visible, which is a substantial fraction of the claimed ~100 nm error bar. The two acknowledged failures (F2+ and F3-) are attributed to weak lattice coupling rather than to this ensemble contamination, but no correlated excited-state reference (TDDFT, BSE, or quantum chemistry) is provided to rule out a systematic methodological bias. Because the printing claim generalizes to polar materials, the half-electron ΔSCF representation is the load-bearing approximation and its error has not been quantified.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper reports hybrid-functional (HSE) calculations of vacancy-cluster color centers in LiF, covering the F center and the F2 and F3 vacancy aggregates in several charge and spin states. The HSE parameters αHSE and μHSE are tuned along the 14 eV band-gap isoline and then further constrained per defect by the generalized Koopmans' theorem. Excited states are computed with a constrained-occupation ΔSCF scheme, using half-electron excitations in both spin channels to mimic spin-conserving singlet excitations and avoid charge-transfer instabilities. The authors report absorption/emission wavelengths, transition dipole moments, radiative lifetimes, and zero-field splittings, and they compare to experiment. They claim that most of the computed optical transitions agree with experiment within about 100 nm, with the F2+ and F3- centers as acknowledged exceptions, and they propose that S=1 states of F3+ and F3- account for the observed emission from those centers.","tokens_in":14878,"tokens_out":4646,"duration_ms":49745,"significance":"If the accuracy claim holds, the work provides a computationally light hybrid-functional route for color centers in wide-gap polar materials, with a parameter-selection protocol that does not fit the optical transitions themselves. The systematic comparison across charge states of the same vacancy clusters, the treatment of symmetry-lowering distortions, and the spin-state assignments for triplet emission are valuable contributions to the color-center literature. The paper is also honest about the per-defect nature of the gKT tuning and about the acknowledged failures. However, the central claim of ~100 nm accuracy rests on an excited-state method whose systematic error is not quantified, and no convergence tests are reported, so the generality of the claim is not yet established.","major_comments":[{"comment":"The half-electron ΔSCF representation of singlet excitations is a spin-mixed ensemble, not a pure singlet. The paper reports ground-state singlet–triplet splittings of 290 meV for F2 and 266 meV for F3-; for a two-electron/two-orbital model the half-electron energy is the average of the singlet and triplet energies, so the error in a nominally singlet transition can be as large as half the relevant excited-state singlet–triplet splitting. That is tens of nm in the visible, a substantial fraction of the claimed ~100 nm accuracy. No reference calculation (e.g., BSE, TDDFT, or a spin-pure constrained-DFT calculation) is provided to bound this error for any of the reported transitions. Since F2+ and F3- already deviate by 100–340 nm, the possibility of a systematic methodological bias is not ruled out. Please provide a quantitative estimate of this bias for at least one center, or re-scope the accuracy claim accordingly.","section":"Details of optical properties (Fig. 5) and Table I"},{"comment":"The statement that 'with the exception of F2+ and F3-, our computed absorption/emission wavelengths compare well with experimental values to within ~100 nm' is supported by only four favorable comparisons: F0 (204 vs 250 nm), F2- (912/1048 vs 956/1113 nm), F2 singlet (501/649 vs 443/678 nm), and F3+ singlet/triplet (440/471 vs 443/542 nm and 497/563 nm). The paper attributes the F2+ and F3- failures to weak lattice coupling, but no supporting calculation or alternative excited-state treatment is offered. To substantiate the generality claim, the authors should analyze whether the two failures are consistent with the half-electron ΔSCF bias or with a missing physical mechanism, rather than treating them as isolated exceptions.","section":"Table I and Discussion"},{"comment":"No convergence tests are reported for supercell size (a single 216-atom Gamma-only cell), plane-wave cutoff, or finite-size corrections for charged defects beyond the brief charge-correction term in the generalized Koopmans' theorem section. For a polar insulator with strongly localized defect states, supercell-size effects on vertical transition energies can be a few tenths of an eV. At minimum, a comparison of the F0 and F2 optical transitions in a 3×3×3 supercell with a larger cell (or a 2×2×2 k-point sampling) should be provided to establish numerical error bars for the reported wavelengths and for the ~100 nm accuracy claim.","section":"Methods and Computational Details"}],"minor_comments":[{"comment":"The text states that the green and red emission bands at 528 nm and 670 nm are associated with the M band (F2 center), while Figure 1's caption says the ~525 nm and ~650 nm peaks correspond to the F2 center and F3+ center, respectively. Please reconcile these attributions.","section":"Introduction / Figure 1"},{"comment":"The sentence 'the vertical ionization energy along the 14 eV isoline, chosen to reproduce the LiF bandgap, expectedly remains constant' would benefit from a one-sentence justification, since the vertical ionization energy is generally a function of αHSE and μHSE even at fixed band gap.","section":"Generalized Koopmans' theorem"},{"comment":"The spin states in Table I are listed as S=0, 1/2, 1 in a column, but the column header is not explicit; consider adding a clear 'S' heading and placing all transitions for each defect in a single contiguous block.","section":"Table I"},{"comment":"References [37] and [56] are the same article (Okuda, J. Phys. Soc. Japan 16, 1746 (1961)) and should be merged or cited separately with distinct note.","section":"References"},{"comment":"The heading uses 'Jahn-Teller' for an instability that is avoided by half-electron occupations, but no Jahn-Teller energy or distortion coordinate is actually computed; a more neutral heading such as 'Excited-state occupation instabilities' would describe the content more accurately.","section":"Section 'Spin-dependent Jahn-Teller instability'"},{"comment":"Several minor typographical issues appear, including 'an M -center formed from two associated F -centers' (missing 'from'), 'V ASP' spacing in the VASP program name, and inconsistent hyphenation of 'generalized Koopmans’ theorem.'","section":"Throughout"}],"recommendation":"major_revision","confidential_remarks":"The paper's per-defect tuning of HSE parameters to satisfy gKT, while not circular for the optical transitions, means the method is not a single predictive functional; this should be stated more prominently in the introduction or discussion if the authors wish to avoid overclaiming transferability to other polar materials. The main technical risk is the half-electron ΔSCF singlet, which deserves at least one benchmark against a higher-level method before the ~100 nm accuracy claim is accepted."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Colleague,\n\nThis is a solid computational study of LiF vacancy color centers with tuned HSE functionals. The genuinely new parts: first gKT-tuned calculations for the F2 and F3 charge states, a parameter-sensitivity analysis along the 14 eV band-gap isoline, and symmetry-breaking distortions that prior work missed. The paper also reports triplet states, zero-field splittings, and lifetimes for several defects, which is useful for the quantum defect community. Credit where due: they are honest about the two exceptions (F2+ and F3-) and the limitations of their excited-state method.\n\nWhat I like: the parameter-tuning protocol along the band-gap isoline is sensible, and checking gKT per defect is a good discipline. The discussion of electron-phonon coupling via absorption-emission separation is insightful. Their computed triplet state for F3- being lower in energy than the singlet is a concrete, falsifiable prediction.\n\nThe soft spots, in proportion: the half-electron delta-SCF treatment of singlet excitations is the load-bearing approximation. As the authors admit, it is crude. The stress-test point is correct: that state is a spin-mixed ensemble, not a pure singlet, and the reported singlet-triplet splittings (290 meV for F2, 266 meV for F3-) imply a possible bias of tens of nm in the visible. They offer no correlated reference (TDDFT, BSE, quantum chemistry) to quantify it. The two acknowledged failures may be symptoms of this approximation rather than lattice-coupling effects, as the authors suggest. So the ~100 nm accuracy claim is plausible but not fully supported. Second, no convergence tests or error bars on supercell size or k-points appear in the paper. That is a minor but real omission. Third, the generalization to polar materials is speculative from one material.\n\nAll that said, this deserves a serious referee. The authors are careful, the results are new, and the honest reporting of exceptions is a sign of good science. I would send it to peer review, but I would not cite it in my own work until the delta-SCF excited-state method is benchmarked against a correlated method, at least on one of the failing defects.\n\nRecommendation: engage with it. Ask for convergence data, error bars, and ideally a BSE/TDDFT test case.","headline":"A careful HSE study of LiF color centers with real new results, but the central accuracy claim rests on a crude delta-SCF singlet approximation whose error is not quantified.","tokens_in":15464,"tokens_out":2733,"would_cite":false,"duration_ms":30554,"reading_group":"maybe","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 shows that hybrid functional calculations with tuned parameters reproduce the optical wavelengths of most lithium fluoride color centers to within about 100 nanometers, and identifies triplet states as the emitters for two…","keywords":["lithium fluoride","color centers","F-center","hybrid functional","generalized Koopman's theorem","optical absorption and emission","triplet states","delta-SCF"],"falsifier":"Compare the half-electron $\\Delta$-SCF wavelengths for $F_2^+$ and $F_3^-$ against GW+BSE or TDDFT results; if the rigorous method matches experiment where $\\Delta$-SCF does not, the half-electron approximation is the weak link.","tokens_in":14430,"feed_emoji":"🔬","tokens_out":10370,"duration_ms":88498,"temperature":0.7,"pith_summary":"The paper tries to establish that a computationally affordable hybrid-functional approach can predict the optical and spin properties of vacancy-cluster color centers in lithium fluoride well enough for quantum-applications screening. By tuning the Heyd-Scuseria-Ernzerhof parameters $\\alpha_{\\mathrm{HSE}}$ and $\\mu_{\\mathrm{HSE}}$ so that the computed band gap matches the experimental 14.0 eV and the generalized Koopman's theorem is satisfied, the authors compute absorption and emission wavelengths for the F, F$_2$, and F$_3$ centers. With the exception of $F_2^+$ and $F_3^-$, the computed transitions fall within about 100 nm of measured values. The paper further argues that the spin-triplet ($S=1$) states of $F_3^-$ and $F_3^+$ are the likely sources of the observed emission from these defects. If correct, the work offers a lightweight protocol for identifying and characterizing quantum-relevant color centers in polar insulators.","feed_headline":"Most LiF color-center wavelengths computed within ~100 nm","feed_subtitle":"Hybrid functional tuning reproduces vacancy-defect optics and reveals triplet emitters.","key_machinery":"The load-bearing object is the Heyd-Scuseria-Ernzerhof (HSE) hybrid functional, whose fraction of exact exchange $\\alpha_{\\mathrm{HSE}}$ and range-separation parameter $\\mu_{\\mathrm{HSE}}$ are tuned together so that the band gap of LiF stays at the experimental value of about 14.0 eV. The generalized Koopman's theorem is used as a self-consistent constraint: the HOMO energy of the $N$-electron defect, the LUMO energy of the ionized $N-1$ state, and the vertical ionization energy are required to align, which cancels self-interaction error and fixes the defect level positions. Excited states are computed with the constrained-occupation $\\Delta$-SCF method, using half-electron occupations in both spin channels to represent spin-conserving singlet excitations and to prevent the charge-transfer instability that would otherwise break convergence in aggregate defects such as $F_2$ and $F_3$.","core_discovery":"The central discovery is that a tuned hybrid functional plus a careful constrained-occupation treatment reproduces most of the known optical transitions of LiF color centers to an accuracy of about 100 nm, and that the correct functional parameters differ from defect to defect. For example, the computed absorption of the neutral $F$ center is 204 nm versus 250 nm experimentally; for the $F_2^0$ singlet, computed absorption is 501 nm (443 nm experiment) and emission is 649 nm (678 nm experiment); for $F_3^+$, computed absorption is 440 nm (443 nm experiment) and emission 471 nm (542 nm experiment). The parameters that satisfy the generalized Koopman's theorem vary strongly, from $\\alpha_{\\mathrm{HSE}}=0.45$, $\\mu_{\\mathrm{HSE}}=0.125$ for $F^0$ to $\\alpha_{\\mathrm{HSE}}=0.368$, $\\mu_{\\mathrm{HSE}}=0.010$ for $F_3^+$, even though all lie on the same 14.0 eV band-gap isoline. The paper also reports a previously unnoticed symmetry-breaking distortion that splits the $p$-like excited state of the $F$-center, and a spin-dependent Jahn-Teller-like instability in aggregate centers that is controlled by using half-electron excitations in each spin channel. These results establish the hybrid-functional approach as an accurate and computationally lightweight tool for color-center properties in polar materials.","pith_inferences":["The success of the same HSE framework across multiple charge states suggests that gKT tuning, rather than a universal functional, is the transferable part of the method; this may be the main methodological takeaway for other materials.","The $F_2^+$ and $F_3^-$ failures could indicate that the half-electron approximation is most accurate when the excited state is strongly lattice-coupled; testing that correlation across a wider defect set would refine the method's domain.","The identification of triplet states as emitters implies that control of the singlet-triplet intersystem crossing rate could tune LiF color centers as single-photon sources, a route not explored in the paper.","Because the gKT-optimal parameters vary per charge state, a single fixed HSE parameter set will not be predictive; future database-driven defect screening may need to store per-defect functional parameters."],"forward_implications":["The hybrid-functional protocol with gKT constraints can be applied to other charge states and vacancy aggregates in LiF without re-fitting to experiment, enabling predictions for defects that have not yet been measured.","The reported spin-triplet assignments for $F_3^-$ and $F_3^+$ give specific emission wavelengths and lifetimes (e.g., about 1.6 $\\mu$s for $F_3^+$), which can be checked by time-resolved photoluminescence and magnetic-field experiments.","If the weak dependence of optical properties on HSE parameters holds for other polar insulators, high-throughput computational screening of color centers in such materials becomes feasible.","The computed large zero-field splittings of the triplet ground states identify LiF color centers as candidate spin-photon interfaces for quantum information.","The symmetry-breaking distortion of the F-center excited state suggests that at low temperature, emission peaks may split; this is a concrete, testable signature."],"supporting_citations":[{"why":"Supplies the experimental absorption and emission wavelengths and electron-phonon coupling assignments against which the computed values are compared in Table I.","marker":"[38]"},{"why":"Provides the G0W0 reference for the F-center and the prior gKT-optimized HSE parameter ($\\alpha_{\\mathrm{HSE}}=0.47$) that the current tuning reproduces.","marker":"[47]"},{"why":"Provides GW+BSE results for the F-center that the HSE defect-level positions are compared against.","marker":"[45]"},{"why":"Introduces the HSE band-gap tuning procedure that this work extends along the 14.0 eV isoline.","marker":"[64]"},{"why":"Defines the generalized Koopman's theorem conditions used to set the HSE parameters for each defect.","marker":"[65]"},{"why":"Establishes the constrained-occupation $\\Delta$-SCF method for computing color-center optical transitions.","marker":"[69]"},{"why":"Justifies the half-electron occupation trick for avoiding Jahn-Teller-induced convergence failures in degenerate excited states.","marker":"[72]"},{"why":"Supplies experimental evidence for the triplet state and microsecond lifetime of the $F_3^+$ center that the triplet assignment is compared against.","marker":"[4]"}],"fun_headline_variants":["Hybrid functional nails LiF color-center optics to ~100 nm","LiF color-center wavelengths computed within ~100 nm","Tuned hybrid functional reproduces LiF defect optics","First-principles LiF color centers: accurate and light","Defect-specific parameters key to LiF optical accuracy"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The entire accuracy claim rests on the assumption that representing a singlet excitation by half an electron in each spin channel is a faithful enough model of the correlated excited state; if that approximation shifts vertical transition energies by more than about 100 nm, the central claim fails.","fun_headline_variants_meta":{"raw":{"variants":["Hybrid functional nails LiF color-center optics to ~100 nm","LiF color-center wavelengths computed within ~100 nm","Tuned hybrid functional reproduces LiF defect optics","First-principles LiF color centers: accurate and light","Defect-specific parameters key to LiF optical accuracy"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000587,"raw_usage":{"total_tokens":2790,"prompt_tokens":1010,"completion_tokens":1780,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":626,"completion_tokens_details":{"reasoning_tokens":1700}},"tokens_in":626,"tokens_out":1780,"duration_ms":13945,"temperature":1.0,"reasoning_tokens":1700,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-10T23:03:50.978742+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Compare the half-electron $\\Delta$-SCF wavelengths for $F_2^+$ and $F_3^-$ against GW+BSE or TDDFT results; if the rigorous method matches experiment where $\\Delta$-SCF does not, the half-electron approximation is the weak link.","supporting_citations":[{"cited_title":"Baldacchini, Journal of Luminescence 100, 333 (2002), ISSN 0022-2313, URL https: //www.sciencedirect.com/science/article/pii/ S002223130200460X","cited_arxiv_id":null,"evidence_quote":"Supplies the experimental absorption and emission wavelengths and electron-phonon coupling assignments against which the computed values are compared in Table I."},{"cited_title":"Chen and A","cited_arxiv_id":null,"evidence_quote":"Provides the G0W0 reference for the F-center and the prior gKT-optimized HSE parameter ($\\alpha_{\\mathrm{HSE}}=0.47$) that the current tuning reproduces."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Establishes the constrained-occupation $\\Delta$-SCF method for computing color-center optical transitions."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Justifies the half-electron occupation trick for avoiding Jahn-Teller-induced convergence failures in degenerate excited states."}],"review_version":1}