{"id":"6e8f854a-c38e-42a8-9bf6-c4fc27dce2a8","arxiv_id":"2505.24747","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":7,"one_line_summary":"The Stark shift of the T center in silicon is computed from first principles, yielding a small linear dipole change of -0.79 D along X and about 0.09 D along Y, with a 25.6 meV exciton binding energy.","lead":"This paper computes from first principles how the zero-phonon emission line of the T center in silicon shifts under an electric field, using density functional theory with large supercells. It reports a small dipole moment change of about -0.8 debye along one axis and near zero along the other, and proposes that charged impurities explain why ensemble measurements see a larger shift.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The Y-component Stark dipole is not established: PBE yields 0.09 D as a near-cancellation of two ~3.5 D contributions, while the paper's own HSE data give 0.9–2.1 D; the 'insensitive T center' claim needs a large-cell HSE check.","rationale":"The paper's most valuable result is the carefully converged binding energy (25.59±0.69 meV), and the X-component Stark dipole is plausible and consistent with the experimental X value. The Y component is where the central 'modest/insensitive' claim is won or lost. Sec II C shows the Y value is a small remainder of two large, opposite contributions, so it is extremely sensitive to functional choice, relaxation, and fitting. The HSE data reported in Supp S2/S3 do not corroborate the PBE remainder: the extrapolated HSE Y values (2.09 D all cells, 0.89 D excluding two small cells) bracket a much larger response, and the claim that PBE and HSE 'usually' agree is supported only by molecular and NV benchmarks that do not involve a 35 Å-bound hole. This is not a dispute with consensus; it is an internal tension within the paper's own data. The discrepancy between the abstract's ΔμY=0.03 D and main-text 0.09 D is a symptom of this fragility rather than an independent flaw. A large-cell HSE (or equivalent higher-level) calculation of ΔμY would settle the point. Until then, the quantitative Y value and the local-field-based explanation of the experiment–theory mismatch should be treated as conditional. The reader's CONDITIONAL verdict is appropriate; no adjustment is needed.","tokens_in":13054,"tokens_out":8798,"duration_ms":104463,"concrete_test":"Converge ΔμY with HSE at the largest cells used for PBE (or at least N=7/N=8, 2744–4096 atoms) using the same ΔSCF + modern-polarization workflow. If the resulting extrapolated ΔμY,HSE remains within, say, ±0.3 D of the PBE 0.09 D, the PBE result is corroborated. If it falls in the 0.9–2.1 D range suggested by the smaller HSE cells, the published Y value and the 'insensitive along Y' conclusion are unsupported, and the local-field explanation for the experiment–theory Y discrepancy is not needed. A cheaper auxiliary check: decompose μ_a″ and μ_hole at PBE and HSE for the same N=6 cell and show that the ±0.15–0.2 D uncertainty in each cannot flip the remainder.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central conclusion that the T center has a small, anisotropic Stark response rests on ΔμX=-0.79 D and especially ΔμY=+0.09 D (abstract: 0.03 D). The Y value is the most fragile part of the argument. Sec II C decomposes the exciton dipole into the a″ occupation (+3.50 D along Y) and the hole (-3.47 D along Y); the published result is the 0.03–0.09 D remainder of two ~3.5 D terms. At the 5% level of uncertainty typical for DFT dipole errors, either term could shift the remainder by ~0.2 D and change its sign. The paper's HSE data, although unconverged, do not support the PBE remainder: Supp S2 reports ΔμY,HSE=2.09 D from all cells and 0.89 D when the two smallest cells are excluded, and states it is 'difficult to obtain a definite value.' The appeals to molecular benchmarks (Refs. 30,31) and NV center (Ref. 16) concern compact dipoles, not a 35 Å bound-exciton hole; the same PBE/HSE functional difference changes the exciton binding energy of this system by roughly a factor of two (14.6 vs 28.5 meV). Thus the implicit assumption that PBE and HSE give similar dipole changes for the T center is not merely unverified; the only direct T-center HSE evidence is in tension with it. Because the experimental comparison and the local-field explanation in Sec II D treat ΔμY≈0 as the signature of a field-insensitive defect, this is the load-bearing uncertainty. The X component is better supported by the convergence trend and experiment.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"This paper reports first-principles calculations of the zero-phonon-line (ZPL) Stark shift of the T center in silicon, treating the excited state as a defect-bound exciton. Using supercell-size convergence studies with PBE and HSE functionals on cells up to 5834 atoms, the authors extract an exciton binding energy of 25.59 ± 0.69 meV after applying the Swift et al. slope-correction scheme, in agreement with the experimental range of 22.5–35 meV. The dipole moment change between the ground and excited states is computed with the modern theory of polarization, yielding Δµ_X = -0.79 D and Δµ_Y = +0.09 D (the abstract states 0.03 D), which the authors describe as a modest and anisotropic linear Stark response. A band-by-band decomposition shows that the small Y component arises from the near-cancellation of a +3.50 D contribution from the occupied a″ state and a -3.47 D contribution from the delocalized valence-band hole. The discrepancy with the ensemble Stark measurement of Clear et al. (Δµ_Y = +1.49 D) is attributed to local field effects from charged impurities, supported by an order-of-magnitude estimate of impurity-induced dipoles, and the computed exciton polarizability change (0.0127 Hz·m²/V²) is offered as evidence of environmental sensitivity.","tokens_in":13451,"tokens_out":14291,"duration_ms":148617,"significance":"The binding-energy part of this paper is a significant and carefully executed contribution: the supercell convergence is documented, confidence intervals are quoted, the PBE/HSE slope-correction method is explained in the supplementary material, and the corrected value of 25.59 meV sits within the experimental range. The application of the modern theory of polarization to a delocalized bound exciton, including the band-by-band decomposition that isolates the hole and a″ contributions, goes beyond the authors' earlier NV-center work and provides a methodological template for other shallow and bound-exciton defects. The predicted weak X-component response and the large polarizability volume are falsifiable statements that single-defect Stark experiments could test. The central caveat is the quantitative security of Δµ_Y: the paper's own supplementary material concedes that the PBE and HSE dipole results cannot currently be reconciled, and the (unconverged) HSE data actually point toward a larger Y-component that would bring theory closer to the ensemble experiment.","major_comments":[{"comment":"The central claim of a small, anisotropic Stark response rests on Δµ_Y = +0.09 D (abstract: 0.03 D), yet this value is the near-cancellation of two roughly 3.5 D contributions (hole -3.47 D and a″ +3.50 D along Y, §II C). At the few-percent level of accuracy typical of DFT dipole errors, either term could shift the remainder by approximately 0.2 D and change its sign, and no uncertainty is quoted for Δµ_Y. The paper's own HSE data are in direct tension with the PBE result: S2 reports Δµ_Y,HSE = 2.09 D when all cells are included and 0.89 D when the two smallest cells are excluded, and states that it is 'difficult to obtain a definite value', while S3 concedes that 'it is difficult to conciliate the PBE and HSE dipole moment changes'. Because the comparison with experiment and the local-field interpretation in §II D treat Δµ_Y ≈ 0 as the signature of a field-insensitive defect, this is a load-bearing uncertainty; the revision should either provide a converged large-cell HSE value for Δµ_Y or report a quantified uncertainty spanning the HSE range and correspondingly temper the conclusions.","section":"§II B, Fig. 3, §II C, S2, S3"},{"comment":"The assumption that PBE and HSE give comparable dipole moment changes for the T center is argued by analogy to a compact defect (the NV center: Δµ_PBE = 2.68 D vs Δµ_HSE = 2.23 D) and to molecular benchmarks (Refs. 30 and 31), but none of these benchmarks involves a charge distribution as delocalized as the approximately 35 Å bound-exciton hole. The same functional difference changes the exciton binding energy of the T center by nearly a factor of two (PBE 14.62 meV vs HSE 28.51 meV, §II A), demonstrating that PBE and HSE are not interchangeable for this state. Since the authors note in S2 that larger supercells would be required to obtain a converged HSE dipole moment, the equivalence of PBE and HSE for this specific system remains an unverified inference rather than a demonstrated fact, and the published Δµ values (especially Δµ_Y) should be reported with an uncertainty that reflects this.","section":"§II B, §IV, S3"},{"comment":"The dipole convergence analysis does not meet the reporting standard set by the binding-energy analysis in the same paper. The binding-energy fits quote 95% confidence intervals throughout (§II A), whereas Δµ_X = -0.79 D and Δµ_Y = +0.09 D are reported without any uncertainty. In addition, smaller supercells are excluded from the dipole fit 'due to the large deviations with respect to the calculated slopes', but no quantitative criterion for the cutoff is given, and the number of retained points, the fit residuals, and the sensitivity of the extrapolated intercept to the cutoff are not reported. Since the extrapolated intercepts are the published results, the fit should be documented with the same rigor as the binding-energy extrapolation, including the excluded data points in the figure.","section":"§II B, Fig. 3"},{"comment":"The value of Δµ_Y is quoted inconsistently across the manuscript: the abstract and the decomposition analysis in §II C state 0.03 D, whereas §II B, Table I, and the conclusion report 0.09 D. This is a headline quantity that readers will propagate from the abstract, and the discrepancy must be resolved so that a single value is used throughout.","section":"Abstract; §II B; Table I; Conclusion"}],"minor_comments":[{"comment":"The abstract's Δµ_Y = 0.03 D coincides with the scheme-1 decomposition value in §II C rather than with the converged supercell result of 0.09 D reported in §II B; the abstract appears to conflate the decomposition with the final converged value.","section":"Abstract; §II C"},{"comment":"The critical-field estimate uses a Bohr radius of 35.64 Å while §II B reports 34.99 Å from the largest supercell; the origin of this difference should be stated.","section":"§II D"},{"comment":"The notation 'N at >1000' for the cell-size threshold is unclear; 'N at' should be defined as the number of atoms and the threshold stated precisely.","section":"§II B"},{"comment":"The impurity-induced dipole estimate in S6 uses α = 0.123 Hz·m²/V², which is the experimental polarizability change from Ref. [13] and an order of magnitude larger than the computed Δα = 0.0127 Hz·m²/V² of §II B. Using the experimental value is reasonable if the estimate is meant to describe the experimental environment, but the text should state this explicitly, because with the computed polarizability the induced dipole at the quoted concentrations would be roughly ten times smaller.","section":"S6, Table III"},{"comment":"There are several typos and inconsistent notations: 'estimate the extend' (§II B), 'the larger the fields applied' (§II D), 'a a density functional theory' (§IV), '22.5eV to 32meV' (§II A), inconsistent spellings of Heyd-Scuseria-Ernzerhof ('Ernzherhof', 'Erzherhof'), and the Table III units 'Hz.m/V' should be Hz·m²/V².","section":"Throughout"}],"recommendation":"major_revision","confidential_remarks":"The stress-test concern about Δµ_Y lands on reading the manuscript: the authors' own supplementary text (S2, S3) concedes the central tension between the PBE and HSE dipole results, so the major-comments request is to make the uncertainty quantitative rather than to reassert the PBE result. The paper is within the journal's scope, the prior work is cited appropriately, and I see no concerns requiring confidentiality beyond what is stated in the report."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Read this for the binding-energy convergence study, not for the Y-component Stark shift. The paper does something genuinely new—first supercell evaluation of the T-center Stark shift via modern theory of polarization—and the binding-energy part is the strongest section. The extrapolated 25.59 ± 0.69 meV, using the Swift et al. slope-correction scheme, matches experimental values, and the careful treatment of the delocalized hole is convincing. The finding that PBE underestimates the binding energy by roughly a factor of two relative to HSE is clearly presented.\n\nThe dipole part is where the paper is soft. The X component, -0.79 D, looks plausible from the convergence trend and is close to experiment. The Y component, however, is the remainder of two ~3.5 D contributions (hole and a''), and the PBE value is 0.09 D in the main text but 0.03 D in the abstract. The paper's own HSE data don't settle it: excluding the two smallest cells gives 0.89 D, including all cells gives 2.09 D, and the supplementary text says it is 'difficult to obtain a definite value.' The argument that PBE and HSE give similar dipole changes relies on compact molecular benchmarks and the NV center, not on a 35 Å exciton. Given that PBE versus HSE changes the binding energy of this very system by a factor of two, the equivalence assumption is not safe. The smaller supercells are excluded from the fit, and no error bars are quoted for the dipole values. The local-field explanation for the experimental discrepancy is labeled as a hypothesis, which is fine, but it rests partly on the near-zero Y value.\n\nSo the central claim that the T center is relatively insensitive to uniform electric fields is not yet established. The X component alone seems fine, but the Y component could have either sign and a magnitude of several tenths of a debye. The authors are honest about the HSE difficulty, and the paper doesn't tune any parameter to experiment, so I don't see circularity. The binding energy justifies publication; the dipole claims need a large-cell HSE calculation or at minimum a clear statement that the Y value is a tentative PBE-level estimate.\n\nFor a referee: yes, this deserves serious review. The methods are solid, the problem is relevant, and the binding-energy result alone is worth publishing. But the referee should push on the Y dipole and the PBE/HSE equivalence before the quantitative Stark-shift claim goes through. I'd also want the abstract fixed to match the main text.","headline":"Read this for the binding-energy convergence study, not for the Y-component Stark shift.","tokens_in":14013,"tokens_out":2272,"would_cite":true,"duration_ms":25482,"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":"First-principles calculations predict a small, anisotropic linear Stark shift for the T center's zero-phonon line, with a dipole moment change of $-0.79$ D along X and near zero along Y.","keywords":["T center in silicon","Stark shift","zero-phonon line","defect-bound exciton","density functional theory","modern theory of polarization","supercell convergence","quantum defects"],"falsifier":"Measure the Stark shift of one isolated T center, or a dilute ensemble with controlled impurity density, at low electric field: if the Y-axis dipole change remains near +1.5 D instead of the predicted near-zero value, the claim that ensemble values are inflated by local impurity fields is falsified.","tokens_in":12838,"feed_emoji":"⚛️","tokens_out":8972,"duration_ms":95272,"temperature":0.7,"pith_summary":"Using density functional theory with careful supercell-size convergence, this paper computes the electric-field response of the T center in silicon, a defect whose telecom-band emission makes it a candidate quantum light source. It finds that the zero-phonon line shifts little under a uniform electric field: the dipole moment change of the optical transition is $-0.79$ D along one in-plane axis and about $+0.09$ D along the other (the abstract quotes $+0.03$ D). The small response matters because a large Stark shift is usually linked to spectral diffusion and poor optical coherence; a line that barely moves is preferable. The paper also argues that the modest intrinsic response is consistent with ensemble measurements only if those measurements are inflated by local fields from nearby charged impurities, a consequence of the exciton's large spatial extent.","feed_headline":"Silicon T center resists uniform electric fields, theory finds","feed_subtitle":"The zero-phonon line shifts little under uniform fields, pointing to local-field effects in ensemble measurements.","key_machinery":"The central object is the defect-bound exciton of the T center, treated in a supercell with the excited state constrained by $\\Delta$-SCF occupation of the $a''$ level. The argument is carried by two tools: supercell-size convergence of the exciton binding energy via Kohn-Sham eigenvalues, with a correction that replaces the PBE exchange slope by the HSE exchange slope, and calculation of the dipole moment change via the modern theory of polarization, a formalism that obtains dipole moments from bulk polarization rather than from a slab with vacuum. The load-bearing decomposition is band-by-band: the total $\\Delta\\mu$ is split into a hole contribution and an $a''$-level contribution, and the small net value is shown to be a cancellation of two large numbers. The paper also extracts the exciton Bohr radius from convoluted charge-density envelopes, giving an elongated wavefunction with a longest decay length of about 35 Å and a corresponding large polarizability change.","core_discovery":"The central claim is that the T center's zero-phonon line has a modest, anisotropic linear Stark shift, with $\\Delta\\mu_X = -0.79$ D and $\\Delta\\mu_Y \\approx +0.09$ D, computed from first principles using the modern theory of polarization applied to defect supercells. The excited state is a defect-bound exciton: a delocalized hole bound to a negatively charged T center, analogous to a group-III acceptor. The paper shows that this delocalization demands supercells far larger than the standard 512-atom cell; only for cells above about 1000 atoms does a linear convergence trend emerge, giving an extrapolated exciton binding energy of $25.59 \\pm 0.69$ meV, in line with the measured 22 to 35 meV range. A band-by-band decomposition reveals that the small total dipole change is the near-cancellation of large hole contributions ($-2.03$ D and $-3.47$ D) and large $a''$-state contributions ($+1.12$ D and $+3.50$ D). The paper attributes the disagreement with ensemble Stark measurements, which give $\\Delta\\mu_Y \\approx +1.49$ D, to local field effects from charged impurities, and notes the T center's large polarizability makes it unusually susceptible to such effects.","pith_inferences":["Beyond the paper: the same supercell-convergence and polarization approach should apply to other bound-exciton defects, such as the G center or shallow acceptors in silicon, where slab-based field calculations are unreliable because the exciton is delocalized.","Beyond the paper: a single-T-center Stark measurement is the cleanest test; an isolated emitter showing Y-axis response near +1.5 D would cast doubt on the local-field explanation, while near-zero response would confirm it.","Beyond the paper: because the net dipole is an accidental cancellation of large hole and $a''$-state contributions, modest perturbations such as strain or nearby charges could change the Stark coefficient by a large relative amount even though the absolute shift stays small.","Beyond the paper: the finding that a 512-atom cell gives a binding energy about three times too large suggests that previous defect calculations on delocalized excitons may need re-examination with the same convergence protocol."],"forward_implications":["A uniform electric field detunes the T center's zero-phonon line only weakly, particularly along the Y axis, so uniform-field spectral diffusion should be small.","Ensemble Stark measurements may not reflect the intrinsic T center response: local fields from charged impurities can induce dipoles on the order of 1 D at impurity concentrations around $10^{16}$ cm$^{-3}$, comparable to the calculated intrinsic dipole change.","The extrapolated binding energy of about 25.6 meV implies a critical field for exciton dissociation near 70 kV/cm, well above the fields used in recent experiments."],"supporting_citations":[{"why":"Provides the experimental optical-transition parameters (dipole and polarizability changes) that the computed Stark shift is compared against.","marker":"[13]"},{"why":"Establishes the modern theory of polarization approach for defect zero-phonon-line dipole moments and supplies the NV-center PBE/HSE comparison.","marker":"[16]"},{"why":"Supplies the Kohn-Sham eigenvalue method and the slope-correction scheme used to extrapolate binding energies in shallow impurities.","marker":"[22]"},{"why":"Previous first-principles T center study that provides the defect model, eigenvalues, and the overestimated 85 meV binding energy motivating the convergence study.","marker":"[11]"},{"why":"Experimental characterization of the T center's telecom emission and an estimated binding energy of 35 meV used as a reference.","marker":"[9]"},{"why":"Experimental PLE measurement giving binding energies in the 22.5 to 32 meV range used as a reference.","marker":"[17]"},{"why":"Benchmark of DFT molecular dipole moments used to argue that PBE-level dipole moments are accurate.","marker":"[30]"},{"why":"Benchmark of DFT dipole moments across 200 molecules, supporting the expectation that PBE and HSE give similar dipole changes.","marker":"[31]"}],"fun_headline_variants":["First-principles T center Stark shift is small and anisotropic","T center's tiny Stark shift points to local-field effects","Silicon T center's Stark shift: modest, anisotropic, first-principles","T center exciton Stark shift computed: small, directional"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The final dipole result assumes that PBE and HSE give essentially the same dipole-moment change for the T center's extended bound exciton, since the reported values come from PBE while HSE could not be converged for the large supercells that contain the full wavefunction.","fun_headline_variants_meta":{"raw":{"variants":["First-principles T center Stark shift is small and anisotropic","T center's tiny Stark shift points to local-field effects","Silicon T center's Stark shift: modest, anisotropic, first-principles","T center exciton Stark shift computed: small, directional"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000232,"raw_usage":{"total_tokens":1534,"prompt_tokens":1037,"completion_tokens":497,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":653,"completion_tokens_details":{"reasoning_tokens":426}},"tokens_in":653,"tokens_out":497,"duration_ms":6414,"temperature":1.0,"reasoning_tokens":426,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-07T12:14:39.500446+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Measure the Stark shift of one isolated T center, or a dilute ensemble with controlled impurity density, at low electric field: if the Y-axis dipole change remains near +1.5 D instead of the predicted near-zero value, the claim that ensemble values are inflated by local impurity fields is falsified.","supporting_citations":[{"cited_title":"The dipole moments are given in D and the polarizability in Hz.m 2/V2 C","cited_arxiv_id":null,"evidence_quote":"Provides the experimental optical-transition parameters (dipole and polarizability changes) that the computed Stark shift is compared against."},{"cited_title":"Udvarhelyi, R","cited_arxiv_id":null,"evidence_quote":"Establishes the modern theory of polarization approach for defect zero-phonon-line dipole moments and supplies the NV-center PBE/HSE comparison."},{"cited_title":"Wang, Density functional calculations of shallow acceptor levels in si, Journal of Applied Physics105 (2009)","cited_arxiv_id":null,"evidence_quote":"Supplies the Kohn-Sham eigenvalue method and the slope-correction scheme used to extrapolate binding energies in shallow impurities."},{"cited_title":"Bergeron, C","cited_arxiv_id":null,"evidence_quote":"Previous first-principles T center study that provides the defect model, eigenvalues, and the overestimated 85 meV binding energy motivating the convergence study."}],"review_version":1}