{"id":"f077255c-d3a2-4918-95a7-49f7b8fe443d","arxiv_id":"2411.15641","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":3,"one_line_summary":"The paper derives the internal conversion rate for the thorium-229 nuclear isomer in crystals and shows that it quenches the isomer in about a millisecond whenever the resonance is energetically allowed, compared to the roughly 1000 second radiative lifetime.","lead":"This paper derives a formula for the rate at which a thorium-229 nuclear isomer in a crystal dumps its 8.4 eV energy into an electronic defect, creating a hole and emitting phonons. It estimates that whenever this process is energetically possible, it kills the isomer in milliseconds, far faster than the natural radiative decay, which matters for choosing crystals for proposed solid-state nuclear clocks.","discovery_kind":"first_principles","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Defect-state energy uncertainty is binary: if CASPT2's ~11 eV value is right, the IC channel is closed in Th:CaF2, so the millisecond-quenching warning does not apply to current hosts.","rationale":"The reader's weakest assumption identifies exactly the load-bearing point. The concern is more than a parameter uncertainty because the resonance condition is a hard inequality; a Kohn-Sham eigenvalue error of order 1 eV can move εd from below to above ω_nuc and change the rate discontinuously from 10^3 s^-1 to zero. The paper's own text flags this, and the CASPT2 benchmark is the only externally computed defect energy and it sits outside the window. The authors' counterarguments (particle-hole shifts, relaxation corrections) are invoked as order-of-magnitude statements and are not turned into a corrected εd. The SI's nonperturbative particle-hole discussion is also a genuine soft spot: |Δ|≫1 means the second-order expression is not controlled, and the 'qualitative' rescue is not a derivation. But that concern modifies the magnitude of an already-fast rate, whereas the defect energy decides whether the rate is finite at all; hence the defect energy is the most load-bearing. I do not change the reader's verdict: CONDITIONAL remains the right call, with the condition being a benchmarked defect energy for at least one current host. The paper deserves credit for the transparent derivation, the explicit conditional framing, and the order-of-magnitude agreement with the Ref. [25] photo-quenching rate.","tokens_in":12678,"tokens_out":7713,"duration_ms":76908,"concrete_test":"Compute the vertical 5f defect excitation energy for Th:CaF2 using the same 3×3×3 supercell and 90° interstitial-F geometry with an embedded-cluster CASPT2 calculation (as in Ref. [23]) and with the paper's PBE/MBJ DFT setup, both at fixed ground-state lattice geometry. Then apply the paper's relaxation shift εd → εd − E_R (E_R ≈ 0.5 eV) and test conditions (5)-(6). Repeat for Th:LiSrAlF6. If the corrected εd exceeds ω_nuc ≈ 8.4 eV in both hosts, the millisecond bound-IC quenching scenario is closed for the current experimental hosts; if it falls below, the central warning is supported.","verdict_should_be":"UNCHANGED","load_bearing_attack":"Eq. (19) predicts Γ_IC ≈ 1.2×10^4 ρ5f(ε_hr) s^-1, but this rate is identically zero unless the resonance window (5)-(6) is satisfied: εd ≤ ω_nuc and εd − εmin ≥ ω_nuc. The paper explicitly concedes in the main text that 'predicting if the IC channel is open critically depends on the reliability of computing the defect state energies εd' and that DFT typically underestimates excited-state energies. The only higher-level benchmark it cites, CASPT2 (Ref. [23]), gives defect state energies ~11 eV for Th:CaF2, above ω_nuc ≈ 8.4 eV. The DFT defect levels used in Figs. 2-3 and in the LiSrAlF6 rate estimate are PBE/MBJ Kohn-Sham eigenvalues; no error bar or correction is supplied, and the CASPT2 discrepancy is not propagated into the central estimate. This is a binary sensitivity: a 2-3 eV error in εd flips the conclusion from 10^3 s^-1 quenching to exactly zero in that host. The later agreement with the photo-quenching rate in Ref. [25] is real supporting evidence, but that measurement is in LiSrAlF6 and accesses a conduction-band PDOS channel, so it does not calibrate the ground-state defect energy εd that gates the bound-IC resonance. Thus the paper's practical claim about which hosts are unusable rests on an unbenchmarked defect energy.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper develops a quantitative theory of internal conversion (IC) of the 229Th nuclear isomer in solid-state hosts. The isomer is treated as a discrete state embedded in a continuum of particle-hole excitations, coupled through the hyperfine interaction, and the decay rate is obtained from Fermi's golden rule. The rate is reduced to a practical formula, Eq. (19), expressed in terms of the measured 5f5/2 hyperfine constant A5f5/2, a nuclear M1 matrix element extracted from measured lifetimes, and the Th-5f projected density of states at the resonant hole energy. DFT calculations for Th:LiSrAlF6 and Th:CaF2 identify 5f-like defect states and yield an estimated IC rate of order 10^3 s^-1 whenever the resonance conditions (5)-(6) are met. The paper also discusses lattice-relaxation and particle-hole corrections, compares with a recent photo-quenching experiment, and draws conclusions for choosing solid-state hosts for nuclear clock applications.","tokens_in":12975,"tokens_out":6060,"duration_ms":58253,"significance":"If the central claim holds, the paper resolves a genuine controversy in the 229Th solid-state clock literature and provides a concrete, testable design criterion: hosts whose defect-state energies satisfy the resonance window will quench the isomer on millisecond timescales, making them unusable for clocks. The main strengths are the transparent derivation from Fermi's golden rule through the relativistic hyperfine matrix elements, the use of independently measured atomic and nuclear inputs (A5f5/2 and the M1 matrix element), the basis-independence proof in the SI, and the order-of-magnitude agreement with the photo-quenching rate extracted in Ref. [25]. The principal weakness is that the practical applicability of the central result is controlled by DFT-level defect-state energies with no demonstrated error bar, and the paper itself cites a CASPT2 benchmark that would close the IC channel in Th:CaF2. This does not invalidate the rate formula, but it does mean the paper's strongest practical conclusion, 'if energetically allowed, it generally quenches,' is not yet quantitatively supported for currently used hosts.","major_comments":[{"comment":"The central rate formula Eq. (19) is nonzero only when the resonance window (5)-(6) is satisfied, making the practical conclusion a binary function of the defect-state energy ε_d. The paper uses PBE/MBJ Kohn-Sham eigenvalues to estimate ε_d for Th:LiSrAlF6 and Th:CaF2 but does not report the numerical values of ε_d, the relaxation shift E_R, or their uncertainties. The cited CASPT2 study (Ref. [23]) gives defect-state energies around 11 eV for Th:CaF2, above ω_nuc ≈ 8.4 eV, which would close the IC channel in that host; a 2-3 eV error in ε_d therefore flips the predicted rate from ~10^3 s^-1 to exactly zero. The SI's own perturbative particle-hole correction is found to have |Δ| ≫ 1, meaning the estimate breaks down, and the subsequent qualitative argument that ρ_5f ~ 0.1 state/eV 'regardless' of particle-hole effects does not repair this. The authors explicitly concede that 'predicting if the IC channel is open critically depends on the reliability of computing the defect state energies ε_d.' I recommend that the paper either provide a benchmarked calculation or a defensible uncertainty range for ε_d in each host, or reformulate the headline claim as a conditional design rule that requires experimental or higher-level verification of ε_d before a host is declared unusable.","section":"Main text, Eqs. (5)-(6) and Eq. (19); SI 'Effects of particle-hole interaction'"},{"comment":"The value ⟨g‖µ‖e⟩ = (0.84 ± 0.11) μ_N is obtained by a weighted average of the two measured isomer lifetimes with the weight of the CaF2 measurement rescaled by an arbitrarily chosen factor k = 2 ('we pick k = 2'). This ad hoc rescaling is not derived from a statistical model or from a quantified refractive-index uncertainty, and it propagates directly through ξ in Eq. (18) into the prefactor of Eq. (19). The resulting uncertainty in the absolute rate is therefore understated. Please replace the ad hoc prescription with a principled combination of the two measurements (for example, a hierarchical model or a sensitivity scan over k) and show how the final rate changes over a plausible range of k.","section":"SI 'Lifetimes and the off-diagonal M1 nuclear matrix element', Eqs. (20)-(22)"},{"comment":"The numerical prefactor in Eq. (19) uses the isotropic-environment assumption ⟨|c5f5/2|^2⟩/⟨|c5f|^2⟩ = 1/2 in deriving ξ ≈ 11. The hosts LiSrAlF6 and CaF2 are not isotropic, and the DFT-calculated defect wavefunctions shown in Fig. 2 reflect the local crystal environment. The paper does not quantify how much the ratio deviates from 1/2 in the calculated structures. Since this ratio enters Eq. (19) linearly, the quoted prefactor 1.2 × 10^4 carries an unquantified systematic uncertainty that could be larger than the stated statistical uncertainties from A5f5/2 and ⟨g‖µ‖e⟩. Please provide the computed ratio from the DFT wavefunctions, or bound its effect on the final rate estimate.","section":"Main text, Eqs. (14)-(15) and Eq. (18)"}],"minor_comments":[{"comment":"The phrase 'on a millisecond timescales' should be 'on a millisecond timescale'.","section":"Abstract"},{"comment":"The units 'states/eV s^-1' are ambiguous; they should be written as '(states/eV) s^-1' to make clear that ρ5f carries the states/eV factor.","section":"Eq. (19)"},{"comment":"Reference [10] is the same article as Reference [7]; one of the two entries should be removed or the citation should be deduplicated.","section":"Reference list"},{"comment":"In the sentence about 6f, 7f, ... orbitals, 'orbits' should read 'orbitals'.","section":"SI, 'Hyperfine interaction'"},{"comment":"The phrase 'The recent experiments [2, 3, 5] relied on observing a nuclear decay on a much longer time-scale' would be clearer with an explicit statement of what 'longer' is relative to (the IC timescale) and with 'timescale' spelled as one word.","section":"Main text, final summary paragraph"},{"comment":"The caption introduces ε_min but does not define it in the figure itself; please add a short definition or a pointer to the text where ε_min is defined.","section":"Fig. 1 caption"}],"recommendation":"major_revision","confidential_remarks":"To the editor: This is a technically serious and timely manuscript that fits the journal's scope. The derivation of the rate formula is sound and the conditional claim is well posed. The main concern is that the practical conclusion about which hosts are unusable depends on defect-state energies that the authors themselves acknowledge are not reliably determined at the DFT level; the cited CASPT2 benchmark points in the opposite direction for Th:CaF2. A revision that adds an explicit uncertainty budget for ε_d and E_R, or that softens the practical claim to a verifiable criterion, would address the load-bearing issue. The ad hoc k=2 weighting of the lifetime measurements should also be justified or replaced."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Colleague,\n\nThe paper gives the first quantitative internal conversion rate for 229Th in solid-state hosts, and the central result is conditional but real: if the resonance condition on the defect energy is met, IC quenches the isomer in milliseconds. The rate formula, Eq. (19), is transparent, depends on the Th-5f projected DOS, and uses independent inputs: a measured hyperfine constant, a nuclear M1 element from lifetime measurements, and DFT band structure. That matters.\n\nThe best part is the honesty of the conditional claim. The authors state plainly that predicting whether IC is open depends on the reliability of defect-state energies from DFT, and they cite a CASPT2 calculation that puts the defect at ~11 eV in CaF2, above the ~8.4 eV nuclear transition, which would close the channel. So the practical warning about unusable hosts hangs on defect energies that are not yet benchmarked. This is the soft spot, and it is a genuine one. The stress-test note is right that it is binary: a 2-3 eV shift flips the rate from 10^3 s^-1 to zero. That doesn't sink the paper, because the conclusion is explicitly conditional, but it does limit the reach of the \"general quenching\" claim until those energies are pinned down.\n\nThe other flaws are minor. The k=2 lifetime-adjustment factor is ad hoc, but the resulting M1 element is within uncertainty. The particle-hole nonperturbative justification is qualitative, and the authors admit the perturbative estimate fails, but the argument that it only strengthens the rate is plausible. The photo-quenching comparison is a good check of the rate scale, though it probes conduction-band PDOS, not the gating ground-state defect energy.\n\nWho is this for? Anyone choosing or designing 229Th-doped hosts for nuclear clocks, and theorists working on non-radiative nuclear decay in crystals. It deserves a serious referee: the derivation is coherent, the formula is measurable, and the conditional claim is the right shape. I would send it out and ask the referee to focus on the defect-energy benchmark and the lifetime averaging.\n\nRecommendation: accept conditional on the authors either benchmarking defect energies against a higher-level method or clearly listing which hosts satisfy the resonance window. If they can't, the paper still stands as a framework, but the headline \"millisecond quenching\" should be presented as conditional on defect placement, which it already is in the abstract.","headline":"A solid, conditional theory of IC quenching in 229Th-doped crystals; the rate formula is honest and useful, but the practical warning hangs on unbenchmarked DFT defect energies.","tokens_in":13598,"tokens_out":2731,"would_cite":true,"duration_ms":24721,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":["23.20.Nx","71.55.-i"],"model":"deepseek-v4-flash","headline":"Internal conversion quenches the 229Th isomer in milliseconds when energetically allowed.","keywords":["thorium-229 isomer","internal conversion","solid-state nuclear clock","defect states","hyperfine interaction","projected density of states","nuclear quenching","density functional theory"],"falsifier":"Compute the defect-state energy of Th:CaF2 with a many-body method that includes particle-hole interactions; if it remains above 8.4 eV, the paper's own resonance condition excludes IC for this host. Alternatively, measure the isomer lifetime as a function of applied strain or doping density: a sharp drop from ~$10^{3}$ s to ~$10^{-3}$ s as the defect level crosses the nuclear transition would confirm the mechanism, while its absence would refute it.","tokens_in":12422,"feed_emoji":"⚛️","tokens_out":9503,"duration_ms":76027,"temperature":0.7,"pith_summary":"This paper develops a quantitative theory of internal conversion for the thorium-229 nuclear isomer embedded in solid-state hosts, and argues that whenever the process is energetically allowed, it quenches the isomer on a millisecond timescale — vastly faster than the isomer's measured radiative lifetime of about $10^{3}$ seconds. The mechanism is a resonant transfer of the nuclear excitation energy to a valence-band electron, which is promoted into a thorium 5f defect state through the hyperfine interaction. If the paper is right, the practical consequence for nuclear clock development is immediate: only hosts whose defect states lie safely above or below the resonance window can preserve the long-lived isomeric state needed for clock operation.","feed_headline":"Internal conversion quenches 229Th isomer in ~1 ms","feed_subtitle":"A new rate formula explains why only hosts with defect states above the nuclear transition are safe for 229Th clocks.","key_machinery":"The engine of the argument is the mapping of the nuclear decay to a textbook discrete-state-in-continuum problem: the isomeric nuclear state is coupled by the magnetic-dipole hyperfine interaction to a continuum of particle-hole excitations in which a valence-band electron enters a thorium 5f defect state. Fermi's golden rule then gives the rate as the product of the hyperfine matrix element squared and the thorium 5f-projected density of states at the resonant hole energy. The many-body electronic part is collapsed into a single host-dependent number, $\\rho_{5f}(\\varepsilon_{\\mathrm{hr}})$, while the nuclear and relativistic atomic matrix elements are evaluated ab initio and expressed through the measured $5f_{5/2}$ hyperfine constant, yielding the compact practical formula Eq. (19). Lattice relaxation shifts the defect energy by about 0.5 eV and is absorbed as a $\\varepsilon_d \\to \\varepsilon_d - E_R$ shift, with multi-phonon emission changing the rate only by an order-unity factor.","core_discovery":"The central claim is that internal conversion is generically fast in 229Th-doped crystals when two energy conditions are met: the defect state energy $\\varepsilon_d$ must not exceed the nuclear transition energy $\\omega_{\\mathrm{nuc}} \\approx 8.4$ eV, and the resulting resonant hole must fall within the valence band. Under these conditions the decay rate is $\\Gamma_{\\mathrm{IC}} \\approx 1.2\\times 10^4\\, \\rho_{5f}(\\varepsilon_{\\mathrm{hr}})\\,\\mathrm{s}^{-1}$, where $\\rho_{5f}$ is the thorium 5f-projected density of states at the resonant hole energy. Using density functional theory for Th:LiSrAlF$_6$ and Th:CaF$_2$, the paper estimates $\\rho_{5f}\\sim 0.1$ states/eV, yielding $\\Gamma_{\\mathrm{IC}}\\sim 10^3\\,\\mathrm{s}^{-1}$, i.e., a millisecond-scale lifetime. The paper concludes that the ~$10^{3}$ s radiative lifetimes seen in recent experiments therefore imply that the IC channel is closed at the emitting defect sites, and it identifies the 'safe' energy windows for host selection.","pith_inferences":["If the defect-state energies computed by cluster-model quantum chemistry (≈11 eV for Th:CaF$_2$) are more accurate than the density-functional values used here, then the practical landscape is less bleak: the default host may already sit in a safe window, and the design question becomes one of keeping defect states above $\\omega_{\\mathrm{nuc}}$ rather than avoiding them.","The rate formula suggests a tunable test: applying strain or an electric field to shift $\\varepsilon_d$ across the resonance boundary should produce an abrupt, orders-of-magnitude change in isomer lifetime, providing a clean experimental signature of the mechanism.","One could also look for hosts with very small thorium 5f hybridization, since a reduced $\\rho_{5f}$ suppresses IC even within the resonance window, adding a second selection axis beyond energy placement."],"forward_implications":["Any host that satisfies the resonance window $\\varepsilon_d \\le \\omega_{\\mathrm{nuc}}$ and $\\varepsilon_d - \\varepsilon_{\\mathrm{min}} \\ge \\omega_{\\mathrm{nuc}}$ will quench the isomer in roughly a millisecond, disqualifying it as a nuclear clock medium.","The long radiative lifetimes measured in Th:CaF$_2$ and Th:LiSrAlF$_6$ imply that, at the fluorescent doping sites, the IC channel is energetically closed; the paper locates the allowed windows explicitly.","The same rate formula accounts for the recently observed photo-quenching of the isomer in LiSrAlF$_6$, with the predicted $10^3\\,\\mathrm{s}^{-1}$ rate matching the experimental extraction of $2\\pi\\times10^2\\,\\mathrm{s}^{-1}$.","Because defect-defect interactions lower defect-state energies, increasing the thorium doping density can open the IC channel, setting a critical doping limit for clock operation."],"supporting_citations":[{"why":"Reports the measured radiative lifetime of the isomer in CaF2, the experimental baseline the paper's fast-quenching claim is compared against.","marker":"[2]"},{"why":"Reports the lifetime and defect-state observations in LiSrAlF6 that motivate the paper's DFT treatment of the thorium defect manifold.","marker":"[3]"},{"why":"Provides the measured 5f5/2 hyperfine constant used to convert the hyperfine matrix elements into the compact rate formula.","marker":"[22]"},{"why":"Cluster-model quantum-chemistry calculation finding defect-state energies near 11 eV, the key counterpoint that defines whether the IC channel is open.","marker":"[23]"},{"why":"Photo-quenching experiment whose extracted rate (2π×10^2 s^-1) is compared against the paper's predicted 10^3 s^-1 rate.","marker":"[25]"},{"why":"Supplies the nuclear magnetic moment value used in the geometric factor of the rate estimate.","marker":"[21]"},{"why":"Provides the relativistic many-body atomic-structure method used to evaluate the hyperfine matrix elements.","marker":"[18]"},{"why":"Proposes the solid-state nuclear clock concept, defining why the quenching rate matters for clock performance.","marker":"[1]"}],"fun_headline_variants":["Internal conversion kills 229Th isomer in milliseconds","Millisecond quench: IC limits 229Th solid-state clocks","229Th isomer quenched in ~1 ms by internal conversion","Fast IC quench: 229Th clocks need high defect states"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The prediction depends entirely on whether the vacancy-like electron level in the crystal actually lies below the nuclear transition energy; if that level is a few electronvolts higher than the calculation says, the rapid quenching never happens.","fun_headline_variants_meta":{"raw":{"variants":["Internal conversion kills 229Th isomer in milliseconds","Millisecond quench: IC limits 229Th solid-state clocks","229Th isomer quenched in ~1 ms by internal conversion","Fast IC quench: 229Th clocks need high defect states"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000517,"raw_usage":{"total_tokens":2485,"prompt_tokens":898,"completion_tokens":1587,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":514,"completion_tokens_details":{"reasoning_tokens":1517}},"tokens_in":514,"tokens_out":1587,"duration_ms":11205,"temperature":1.0,"reasoning_tokens":1517,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-12T14:04:40.116439+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Compute the defect-state energy of Th:CaF2 with a many-body method that includes particle-hole interactions; if it remains above 8.4 eV, the paper's own resonance condition excludes IC for this host. Alternatively, measure the isomer lifetime as a function of applied strain or doping density: a sharp drop from ~$10^{3}$ s to ~$10^{-3}$ s as the defect level crosses the nuclear transition would confirm the mechanism, while its absence would refute it.","supporting_citations":[{"cited_title":"To get an estimate for ⟨g||µ||e⟩, we seek to combine these two values for τ","cited_arxiv_id":null,"evidence_quote":"Reports the lifetime and defect-state observations in LiSrAlF6 that motivate the paper's DFT treatment of the thorium defect manifold."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides the measured 5f5/2 hyperfine constant used to convert the hyperfine matrix elements into the compact rate formula."},{"cited_title":"Campbell, A","cited_arxiv_id":null,"evidence_quote":"Cluster-model quantum-chemistry calculation finding defect-state energies near 11 eV, the key counterpoint that defines whether the IC channel is open."},{"cited_title":"Dresselhaus, G","cited_arxiv_id":null,"evidence_quote":"Photo-quenching experiment whose extracted rate (2π×10^2 s^-1) is compared against the paper's predicted 10^3 s^-1 rate."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the nuclear magnetic moment value used in the geometric factor of the rate estimate."},{"cited_title":"Dessovic, P","cited_arxiv_id":null,"evidence_quote":"Provides the relativistic many-body atomic-structure method used to evaluate the hyperfine matrix elements."}],"review_version":1}