{"id":"234124ca-125f-4a0a-b942-2948ef63a39e","arxiv_id":"1908.06516","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":6,"one_line_summary":"Carrier density and mobility in disordered silicon and silicon carbide obey anti-Meyer-Neldel statistics with proportional activation energies, yielding a power-law mobility-density relation.","lead":"Temperature-dependent Hall measurements on microcrystalline silicon and silicon carbide show that conductivity, carrier density, and mobility follow anti-Meyer-Neldel behavior. The paper reports the first sample switching between normal and anti-Meyer-Neldel rule as defects are annealed, and derives a power-law relation between mobility and carrier density.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The claimed first sample switching between anti-MNR and normal MNR rests on a by-eye split of six points in Fig. 5; no statistical test shows that the three 'MNR' points are not consistent with the anti-MNR line.","rationale":"The reader's verdict is CONDITIONAL, and their weakest assumption targets the linear E_n-E_mu relation used in the theoretical derivation. That concern is valid but it does not threaten the paper's stated central claim—the first sample switching between anti-MNR and normal MNR. The stress-test should target the argument that actually bears the headline. That argument is made in Section IV A and Fig. 5, where six data points from sample Ma-15 are partitioned into two regimes. The partition is performed by eye: no fit is reported for the normal-MNR subgroup, the dashed MNR line is explicitly a guide, and the paper admits the right-hand points have large error bars. Since the Arrhenius curves are convex and the MNR coordinates are local room-temperature slopes, the E_a and prefactor values are not model-independent; small changes in the fit window could move points across the claimed boundary. The authors also do not report a statistical comparison of the one-line and two-line hypotheses. If the right-hand points are within 2 sigma of the anti-MNR line, the switch is not demonstrated. The concrete test I propose—a covariance-based model comparison with varying fit windows—would settle this. I do not recommend changing the verdict label: the paper should still be considered for publication conditional on the authors providing that test, because the underlying Hall data are valuable and the theoretical derivation may remain of interest. But the condition should be focused on the robustness of the central claim, not just on the E_n-E_mu correlation.","tokens_in":8563,"tokens_out":10242,"duration_ms":105880,"concrete_test":"Refit the six Ma-15 data points of Fig. 5 using a full covariance-based analysis: for each annealing state, recompute E_sigma and sigma_0 from the Arrhenius fit over a fixed window (e.g., 300 ± 20 K) and propagate uncertainties. Then perform a formal model comparison: fit all six points to a single line (the anti-MNR relation) versus a two-line model with a breakpoint at the claimed demarcation (~120 meV), and report the F-test or chi-square difference. If the single-line model is not rejected at the 5% level—or if the grouping into anti-MNR and MNR is not stable under a ±10 K shift of the fit window—the 'first sample switching' claim is not established.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central novelty of the paper—the first observation of a single sample switching between anti-MNR and normal MNR—rests on six data points of sample Ma-15 in Fig. 5, which the authors split into three left-hand points following the anti-MNR line and three right-hand points following a normal-MNR trend. This split is asserted visually: the normal-MNR trend is drawn only as a dashed 'guide to the eye,' no fit or goodness-of-fit test is reported for the right-hand group, and the text acknowledges that these points have 'larger error estimates' (Section IV A). Because the Arrhenius curves are explicitly convex and the MNR coordinates are local slopes at room temperature (Section III), the E_a and sigma_0 values depend on the fitting procedure. If the three 'normal-MNR' points are statistically compatible with the anti-MNR line fitted from the other annealed states (given the large error bars), or if the apparent grouping changes when the Arrhenius fit window is varied, then the claimed transition evaporates and the paper's headline novelty is unsupported. This is more load-bearing than the E_n-E_mu proportionality concern raised by the reader, because the proportionality issue affects only the theoretical derivation, whereas the sample-switching claim is the stated central result.","agreement_with_reader":"disagree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper reports temperature-dependent Hall measurements of conductivity, carrier concentration, and mobility for lightly phosphorus-doped microcrystalline silicon (μc-Si:H) and hydrogenated silicon carbide (μc-SiC:H), including a series of electron-irradiated μc-Si:H samples measured after step-wise annealing. The authors analyze the data in Meyer-Neldel rule (MNR) plots and report anti-MNR behavior for most samples, the first observation of anti-MNR in μc-SiC:H, and a claim that a single μc-Si:H sample (Ma-15) switches from anti-MNR to normal MNR as its defect density is changed by annealing. The theoretical part of the paper shows that MNR in the carrier concentration does not by itself imply MNR in conductivity unless the mobility satisfies one of three conditions, identifies the proportionality E_n^a = a E_μ^a + b as the operative condition, and derives from it a power-law relation between room-temperature mobility and carrier concentration, μ_r = μ'_00 (n_r/n_00)^κ.","tokens_in":8875,"tokens_out":2048,"duration_ms":22089,"significance":"If the empirical claims hold, the paper provides a valuable experimental dataset that extends the MNR phenomenology: the first reported anti-MNR in μc-SiC:H, a systematic comparison of MNR in σ, n, and μ for the same samples, and a claimed single-sample transition between anti-MNR and MNR that would support the Fermi-level position as the controlling parameter. The theoretical derivation connecting the linear E_n^a-E_μ^a relation to a μ(n) power law is a useful formal step, even though its practical value depends on the independence of the fitted parameters. The paper also gives credit by explicitly flagging the speculative origin of the Arrhenius curvature and the large uncertainties in the mobility data, which is appropriate for a study of this type.","major_comments":[{"comment":"The central claim that sample Ma-15 switches from anti-MNR to normal MNR is based on a visual dichotomy of six data points, three assigned to the anti-MNR line and three to a dashed 'guide to the eye' for normal MNR. No fit is reported for the right-hand group, no goodness-of-fit test is compared with the anti-MNR line, and the text states that these points have 'larger error estimates.' The authors should quantify whether the three right-hand points are statistically incompatible with the anti-MNR line fitted from the annealed samples, including the effects of the large error bars and of the choice of Arrhenius fit window, before the switching claim can be accepted.","section":"IV A, Fig. 5"},{"comment":"The power-law exponent κ in Eq. (7) is computed from E^n_MN, E^μ_MN, a, and b, all of which are fitted from the same Hall data used to produce Fig. 9. The subsequent fit of κ in Fig. 9 (values 0.48 and 0.61) is therefore not an independent test of the derived relation; it only checks that the direct fit lies in the broad range allowed by the large uncertainties. The authors should either provide a proper uncertainty propagation for κ and state whether the directly fitted exponents are consistent with the derived values under that propagation, or clearly label the power law as a consistency check rather than a prediction.","section":"V B, Eq. (7) and Fig. 9"},{"comment":"The MNR coordinates (activation energy and prefactor) are extracted from Arrhenius curves that are explicitly convex rather than straight, and the text concedes that the cause of the curvature is speculative. Because each MNR point depends on the local slope and intercept at room temperature, the classification of samples into MNR and anti-MNR regimes, including the claimed Ma-15 switching, is sensitive to the fitting procedure. The manuscript should specify the exact fitting window and method, and ideally show how the MNR plot changes when the window is varied within the measurement range, so that the reader can assess whether the reported separation is robust.","section":"III and Figs. 5-8"}],"minor_comments":[{"comment":"The text states that the demarcation between MNR and anti-MNR lies at 'approx. 120 eV'; this should presumably be 120 meV.","section":"IV A"},{"comment":"There are typographical errors: 'degration' in Section II B, 'poylnomial' in the caption of Fig. 9, and 'suprising' in Section IV B. The sample identifiers '05B-054' and '06B-276' in Section V B are not defined in the paper.","section":"II B and Fig. 9"},{"comment":"The appendix title says 'at root temperature'; this should be 'at room temperature.'","section":"Appendix A"},{"comment":"The notation 'terms without Ea’s' is imprecise; the authors should state explicitly which constants (such as n_00 and μ_00) are absorbed into the y-intercept, to make the derivation easier to follow.","section":"V A, Eq. (4)"}],"recommendation":"major_revision","confidential_remarks":"The paper addresses a topic that fits the journal well, and the experimental dataset is potentially valuable. However, the headline claim of the first sample switching between anti-MNR and normal MNR is currently supported mainly by a visual grouping of six points in Fig. 5, and the theoretical power-law relation is tested against the same data that were used to determine its ingredients. Both issues are fixable in revision, but they need to be addressed quantitatively rather than by assertion. I would not recommend rejection, as the underlying measurements and the formal derivation are likely sound; I would want to see a revised version that either provides the missing statistical analysis and error propagation or substantially softens the claims."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The paper is more interesting than its low-key framing suggests, but the one claim the authors push hardest -- that a single sample switches between anti-MNR and normal MNR when annealed -- is also the weakest supported.\n\nWhat's actually new: first report of anti-MNR in μc-SiC:H, with σ00 and E_MN matching the silicon values, which strengthens the universality observation. And the proportionality E_n = 1.25 E_mu − 18 meV across very different samples is a genuinely useful empirical result, especially if it holds up under more careful fitting. The theoretical section is a good piece of clarity: it spells out that MNR in n does not guarantee MNR in σ, and the three conditions for σ MNR are worth having. The derivation of the mu(n) power law from the proportionality is algebraically fine, though the authors overstate it as a prediction when it's really a rearrangement of fitted parameters.\n\nThe soft spot is the switching claim. In Fig. 5, six points of the 15 ppm sample are split into three left and three right by eye, with no fit and no goodness-of-fit test for the right group. The text concedes those points have larger error bars. If those three points are statistically consistent with the anti-MNR line, the transition evaporates. Given the Arrhenius curves are convex, the activation energies are local slopes at room temperature, so the grouping could shift with the fitting window. The authors need a simple chi-square or t-test on the split, or a combined fit with a breakpoint, to make the claim stick.\n\nMinor issues: Fig. 8 fit has no reported uncertainties; the text says the numbers 'must be taken with care' but does not propagate them into kappa. And the power-law exponent is fitted again in Fig. 9 after being derived, which is circular -- report it as a consistency check.\n\nCitation pattern is normal; prior work is cited appropriately.\n\nBottom line: this deserves peer review, not desk rejection. The experimental data on n and mu in these materials is rare and presumably hard to get. The switching claim needs real statistical support, and the theoretical claims need to be scaled back to consistency checks. A serious referee would ask for those changes, and the paper would be stronger for it.","headline":"A useful experimental paper with a clever theoretical framing, but the headline 'first sample switching' rests on a by-eye split of six points and needs statistical backup before it can be believed.","tokens_in":9417,"tokens_out":2369,"would_cite":true,"duration_ms":25172,"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 single microcrystalline silicon sample switches between Meyer–Neldel and anti-Meyer–Neldel behavior when its defect density changes, tying the rule to Fermi-level position.","keywords":["Meyer-Neldel rule","anti-Meyer-Neldel","microcrystalline silicon","silicon carbide","Hall effect","statistical shift","carrier concentration","mobility"],"falsifier":"Measure Hall transport on a sample whose defect density is stepped finely across the anti-MNR-to-MNR boundary: if the exponent $\\kappa$ in $\\mu_r(n_r)$ changes with the transition, or if $E_n^a$ versus $E_\\mu^a$ stops lying on the $1.25$ slope line, then the derived power law and the claim that this proportionality governs the anti-MNR domain are wrong.","tokens_in":8324,"feed_emoji":"⚡","tokens_out":10260,"duration_ms":86200,"temperature":0.7,"pith_summary":"The paper examines the Meyer–Neldel rule (MNR) and its inverse, anti-MNR, in hydrogenated microcrystalline silicon and silicon carbide, using temperature-dependent Hall measurements to separate conductivity into carrier concentration and mobility. Its central claim is that the same 15 ppm microcrystalline silicon sample switches from anti-MNR to normal MNR when its defect density is changed by electron bombardment and stepwise annealing, the first such transition achieved by changing material properties rather than gate voltage. The authors argue this supports the statistical-shift picture, in which the Fermi-level position, not sample-to-sample variation, determines MNR behavior. They also report anti-MNR in the carrier concentration and conductivity of silicon carbide with parameters close to silicon, and they derive a power-law relation between room-temperature mobility and carrier concentration from an observed linear proportionality between their activation energies.","feed_headline":"Same silicon sample flips from anti-Meyer–Neldel to Meyer–Neldel","feed_subtitle":"Defect-density changes flip a silicon sample's transport law and yield a mobility–carrier power law.","key_machinery":"The central object is the Meyer–Neldel rule itself, the observation that Arrhenius prefactors grow exponentially with activation energy so that extrapolated Arrhenius lines meet at one point, together with its negative-slope counterpart anti-MNR. The theoretical machinery is the statistical-shift model, which predicts MNR for carrier concentration because the Fermi level moves with temperature. The paper shows this model alone does not give MNR in conductivity; the mobility must meet one of three conditions. The load-bearing condition in this work is the empirical proportionality $E_n^a = a\\,E_\\mu^a + b$, observed with $a \\approx 1.25$ and $b \\approx -18$ meV in the anti-MNR domain of both $\\mu$c-Si:H and $\\mu$c-SiC:H. Substituting this relation into the Arrhenius forms for $n$ and $\\mu$ yields the room-temperature power law $\\mu_r(n_r) = \\mu'_{00} (n_r/n_{00})^\\kappa$ with $\\kappa = E^n_{MN}(E^\\mu_{MN} - kT_r)/(a\\,E^\\mu_{MN}(E^n_{MN} - kT_r))$.","core_discovery":"Using Hall measurements on seven microcrystalline silicon layers, one amorphous layer, two irradiated-and-annealed silicon samples, and five silicon carbide layers, the paper establishes three results. First, all three transport quantities—conductivity, carrier concentration, and mobility—show anti-Meyer–Neldel behavior in lightly doped microcrystalline silicon, and silicon carbide shows anti-MNR in conductivity and carrier concentration with Meyer–Neldel prefactors and energies close to silicon's. Second, the 15 ppm sample Ma-15 moves from the anti-MNR branch to the normal MNR branch as annealing removes bombardment-induced defects; the authors state this is the first report of such a transition by changing material properties, and take it as evidence that Fermi-level position controls MNR. Third, in the anti-MNR domain, the activation energies for carrier concentration and mobility obey a common linear relation $E_n^a = 1.25\\,E_\\mu^a - 18$ meV for both materials; combined with anti-MNR this yields a power law $\\mu_r(n_r) \\propto n_r^\\kappa$, with $\\kappa$ between 0.5 and 1.5, which the annealed-sample data reproduce with exponents around 0.48 and 0.61.","pith_inferences":["If the proportionality $E_n^a = 1.25\\,E_\\mu^a - 18$ meV holds beyond these two material families, then the ratio between carrier-concentration and conductivity Meyer–Neldel energies (about 1.7 here) should be derivable from that line alone; a future analysis could check that consistency.","The same-sample switch suggests a device-relevant handle: irradiation and low-temperature annealing can toggle a film's transport regime without changing its chemistry.","A sharper test than the paper's would fix the exact defect densities and measure $\\kappa$ continuously across the anti-MNR-to-MNR boundary; if $\\kappa$ is constant within the anti-MNR branch, the power-law picture is robust, and if it drifts, the boundary is not a clean phase-like division.","Because the paper lists barrier-height distributions and differential mobility as alternative causes of the Arrhenius curvature, a natural next step is to check whether either mechanism also predicts the measured $E_n^a$–$E_\\mu^a$ slope; if one does, it would unify the curvature and the power law."],"forward_implications":["A fixed sample can be moved across the Meyer–Neldel/anti-Meyer–Neldel boundary solely by changing its defect density, so Fermi-level position, not sample identity, is what selects the regime.","Observing MNR in conductivity is not automatic once carrier concentration follows the statistical shift; the mobility activation energy must satisfy one of three stated conditions, and in these data it does so through the linear $E_n^a$–$E_\\mu^a$ relation.","Silicon carbide joins silicon with nearly the same anti-MNR parameters, strengthening the case for a common Meyer–Neldel behavior across disordered semiconductors.","The derived power law $\\mu_r \\propto n_r^\\kappa$ with $\\kappa$ between 0.5 and 1.5 gives a compact way to predict room-temperature mobility from carrier concentration in the anti-MNR regime.","The unresolved item the paper leaves open is a microscopic explanation for why $E_n^a$ and $E_\\mu^a$ are proportional in the first place."],"supporting_citations":[{"why":"Together they supply the statistical-shift model that predicts MNR for carrier concentration, the theoretical base the paper extends.","marker":"[7, 8]"},{"why":"Reports anti-MNR in undoped microcrystalline silicon, the earlier observation the new samples are compared with.","marker":"[4]"},{"why":"Shows an MNR-to-anti-MNR transition in TFTs via gate voltage, the prior case the paper's same-sample material-property switch extends.","marker":"[6]"},{"why":"Reports a similar transition in TFT structures, establishing that the new sample-alone transition has not been shown before.","marker":"[22]"},{"why":"Provides the earlier Hall measurements and the room-temperature mobility-carrier correlation for the same silicon layers.","marker":"[13]"},{"why":"Supplies the electron-bombardment method used to create and then anneal defects in the silicon samples.","marker":"[11]"},{"why":"Describes the deposition of the microcrystalline silicon carbide samples whose MNR data are new.","marker":"[12]"}],"fun_headline_variants":["First sample to switch between Meyer–Neldel and anti-Meyer–Neldel","Annealing flips one silicon sample between two Meyer–Neldel regimes","Defect tuning makes a silicon sample obey both Meyer–Neldel rules","Hall measurements show a sample flipping Meyer–Neldel behavior","Silicon samples reveal anti-Meyer–Neldel and a rare flip to MNR"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"All the quantitative conclusions rest on treating the measured transport as a single channel with well-defined activation energies, even though the temperature plots bend and the paper says the cause of the bending is speculative.","fun_headline_variants_meta":{"raw":{"variants":["First sample to switch between Meyer–Neldel and anti-Meyer–Neldel","Annealing flips one silicon sample between two Meyer–Neldel regimes","Defect tuning makes a silicon sample obey both Meyer–Neldel rules","Hall measurements show a sample flipping Meyer–Neldel behavior","Silicon samples reveal anti-Meyer–Neldel and a rare flip to MNR"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.001137,"raw_usage":{"total_tokens":4716,"prompt_tokens":933,"completion_tokens":3783,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":549,"completion_tokens_details":{"reasoning_tokens":3679}},"tokens_in":549,"tokens_out":3783,"duration_ms":24677,"temperature":1.0,"reasoning_tokens":3679,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T12:44:26.272185+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Measure Hall transport on a sample whose defect density is stepped finely across the anti-MNR-to-MNR boundary: if the exponent $\\kappa$ in $\\mu_r(n_r)$ changes with the transition, or if $E_n^a$ versus $E_\\mu^a$ stops lying on the $1.25$ slope line, then the derived power law and the claim that this proportionality governs the anti-MNR domain are wrong.","supporting_citations":[{"cited_title":"the conductivity in a-Si:H exhibits a MNR because the statistical shift causes the carrier density to follow a MNR as well as the mobility","cited_arxiv_id":null,"evidence_quote":"Reports anti-MNR in undoped microcrystalline silicon, the earlier observation the new samples are compared with."},{"cited_title":"Kondo, Y","cited_arxiv_id":null,"evidence_quote":"Shows an MNR-to-anti-MNR transition in TFTs via gate voltage, the prior case the paper's same-sample material-property switch extends."},{"cited_title":"Carius, F","cited_arxiv_id":null,"evidence_quote":"Reports a similar transition in TFT structures, establishing that the new sample-alone transition has not been shown before."},{"cited_title":"Beyer and H","cited_arxiv_id":null,"evidence_quote":"Provides the earlier Hall measurements and the room-temperature mobility-carrier correlation for the same silicon layers."},{"cited_title":"Overhof and W","cited_arxiv_id":null,"evidence_quote":"Supplies the electron-bombardment method used to create and then anneal defects in the silicon samples."},{"cited_title":"Overhof and W","cited_arxiv_id":null,"evidence_quote":"Describes the deposition of the microcrystalline silicon carbide samples whose MNR data are new."}],"review_version":1}