{"id":"0da2a7ae-e6b1-4085-81c7-4f74e547313e","arxiv_id":"2411.17460","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":2,"one_line_summary":"Quantum Monte Carlo data on a rubrene-like lattice model show the relaxation time approximation is inconsistent with the computed correlator, while a one-parameter fit yields a low-frequency mobility estimate below RTA and experiment.","lead":"This paper compares high-precision Quantum Monte Carlo data for charge transport in a rubrene-like lattice model with simpler approximations, and reports that the standard relaxation time approximation is inconsistent with the data. It then fits a one-parameter mobility function to the QMC correlator and argues this gives an improved low-frequency mobility estimate, a claim that depends on the fitted ansatz.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Eq. (9) uses ε_k+ε_l in the Lorentzian denominator; the static limit then selects ω=ε_k+ε_l instead of ω=ε_l−ε_k from the Kubo formula (4), so the RTA curves in Fig. 1 may not represent the standard relaxation-time approximation.","rationale":"The reader's conditional verdict is reasonable and I keep it, but the condition I would attach is different. The paper's main empirical claim is the RTA inconsistency; the comparison is well-posed in principle and the QMC data are high-precision, with public codes available. However, Eq. (9) as printed has a sign error in the Lorentzian transition frequency. The exact static limit from Eq. (4) requires ω=ε_l-ε_k; Eq. (9) requires ω=ε_k+ε_l. This is not a matter of convention: a global shift of all eigenvalues changes the RTA spectral function, so the printed formula is not gauge invariant and cannot reduce to the static approximation. The reader's concern about rare negative QMC weights is legitimate and should be checked, but the Eq. (9) issue directly threatens the definition of the approximation being tested and is therefore more load-bearing. I also note the text claims 'any finite τ_loc increases deviations' while showing only two finite τ_loc values; after correcting the denominator, a scan over τ_loc is needed. The μ0 estimate from Eq. (11) remains ansatz-dependent and should carry an uncertainty, as the reader said. Verdict remains conditional: require correction of Eq. (9), re-run of the RTA comparison, and a sign-weight/code check.","tokens_in":9634,"tokens_out":15682,"duration_ms":145151,"concrete_test":"One check: with frozen phonon fields, take one disorder configuration and compute μ_RTA(ω) from Eq. (9) for both denominators, τ_loc^{-1}→0 and small; compare to the exact Kubo mobility from Eq. (4) for the same configuration. Equivalently, read the denominator used in the public TLPP/SSHExDiag code. If the code uses (ω-(ε_l-ε_k))^2, correct Eq. (9) and re-plot Fig. 1; if deviations still grow for all finite τ_loc, the conclusion survives. If the code uses (ω-ε_k-ε_l)^2, the published RTA curves are nonstandard and the conclusion is unsupported until recomputed.","verdict_should_be":"UNCHANGED","load_bearing_attack":"Section 3 bases the RTA-inconsistency claim on G_RTA(τ) computed from μ_RTA(ω) in Eq. (9). As printed, Eq. (9) is internally inconsistent with the exact frozen-phonon limit. Starting from Eq. (4) with static eigenstates |ψ_k⟩, ε_k, one uses (e^{-β ε_k}-e^{-β ε_l}) = (e^{-β ε_k}+e^{-β ε_l}) tanh(β(ε_l-ε_k)/2) and replaces δ(ω-(ε_l-ε_k)) by a Lorentzian; this yields a denominator (ω-(ε_l-ε_k))^2+τ_loc^{-2}. The printed denominator (ω-ε_k-ε_l)^2 makes the τ_loc^{-1}→0 limit select ω=ε_k+ε_l instead of ε_l-ε_k. A uniform shift of all ε_k then shifts the RTA spectrum, violating gauge invariance; therefore Eq. (9) cannot be a valid static/transient-localization approximation. If the code used for Fig. 1 follows the printed formula, the comparison is not a test of the standard RTA. If the code uses ε_l-ε_k, the paper's central equation is misprinted and the comparison must be rerun and documented. This is the most load-bearing unresolved issue; the reader's sign-weight concern is plausible but secondary to the definition of the approximation being tested.","agreement_with_reader":"disagree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript reports high-precision Hybrid Monte-Carlo (QMC) results for the imaginary-time velocity correlator G(τ) of a single charge carrier in the two-dimensional triangular-lattice electron-phonon model of Ref. [11], with parameters appropriate to rubrene. The central claims are (i) that the QMC data, with sub-permille statistical errors, can discriminate between phenomenological transport models; (ii) that the static approximation agrees with G(τ) to about 0.1%, while any finite relaxation time in the simple relaxation-time approximation (RTA) Eq. (9) increases the deviation from QMC, so the RTA is not uniformly valid; and (iii) that a single-parameter phenomenological mobility ansatz Eq. (11) fitted to the QMC G(τ) yields a zero-frequency mobility μ0 that is smaller than the RTA prediction and below the experimental rubrene value. The paper also presents exact-diagonalization (ED) benchmarks on a 5-site 1D chain and discusses connections to quark-gluon plasma transport.","tokens_in":9950,"tokens_out":3432,"duration_ms":35724,"significance":"If the claims are correct, the paper makes a useful methodological contribution: it shows that sufficiently precise imaginary-time data can constrain real-frequency transport models in a regime where direct analytic continuation has poor frequency resolution, and it provides a concrete falsifiable statement about the inadequacy of a simple RTA. The companion open-source codes on GitHub are a strength that should facilitate reproduction. The claim that the RTA is inconsistent with QMC is potentially important for the transient-localization community. However, the validity of this central claim hinges on the correctness of the RTA expression Eq. (9), on the unbiasedness of the QMC samples, and on a statistically meaningful estimate of μ0; each of these points needs attention before the conclusions can be accepted.","major_comments":[{"comment":"The denominator in Eq. (9) is printed as (ω − ε_k − ε_l)^2 + τ_loc^{-2}. Starting from the Kubo formula (4) and taking the static limit, the δ-function is δ(ω − (ε_l − ε_k)), so a Lorentzian broadening should be (ω − (ε_l − ε_k))^2 + τ_loc^{-2}. The printed form is not gauge invariant: shifting all single-particle energies by a constant changes ε_k + ε_l and therefore shifts the RTA spectrum, whereas the mobility from Eq. (4) depends only on energy differences. If the code used for Fig. 1 follows the printed Eq. (9), the comparison tests a non-standard 'RTA' and the conclusion that the RTA is inconsistent with QMC is not established. If the code instead uses ε_l − ε_k, then Eq. (9) is a misprint and the comparison must be rerun and documented with the correct formula. In either case, the central claim of Section 3 requires a corrected and re-validated Eq. (9).","section":"Section 3, Eq. (9)"},{"comment":"The zero-frequency mobility μ0 is introduced as the single free parameter of a phenomenological ansatz and is determined by minimizing χ² against the QMC G(τ). The manuscript reports that 'the optimal value of μ0 ... turns out to be smaller than the RTA prediction, and also lower than the experimental value', but it does not report any uncertainty interval, confidence level, or cross-validation for μ0. Given that the ansatz itself is not derived from the Hamiltonian, the fitted value alone is not strong evidence for the stated comparison with the experimental rubrene mobility. The authors should provide an error bar on μ0 (e.g., from the covariance matrix, bootstrap, or a scan over allowed α and ω1) and ideally show the sensitivity of the conclusion to the functional form of the ansatz.","section":"Section 3, Eqs. (11) and (12)"},{"comment":"The path-integral weight Tr U(0,β) exp(−S_B) is stated to be non-positive in general, and the text asserts that 'negative values only occur for phonon field configurations with very strong τ dependence' and that 'we never encounter negative weights in our simulations'. This is an empirical claim, not a proven bound. Since all QMC results and the sub-permille G(τ) comparisons rely on the absence of sign-related bias, the paper should provide quantitative evidence: for example, a monitored histogram of the weights, an explicit bound on the range of configurations sampled, or a statement of how many samples were checked. Without such evidence, a reader cannot rule out a small but systematic shift in G(τ) from rare negative-weight configurations, which would affect the chi²-based conclusions in Section 3.","section":"Section 2, paragraph on sign weight"}],"minor_comments":[{"comment":"The claim that 'any finite value of τ_loc increases deviations between G_RTA(τ) and G(τ)' is stronger than what is shown in Fig. 1, which presents only two values of 1/τ_loc (6 meV and 2.7 meV) for the 2D lattice and one value for the 1D case. If this statement is based on a systematic scan over τ_loc, the scan range and step should be reported.","section":"Section 3, Fig. 1 and text"},{"comment":"Reference [11] contains the typo 'A. Trosi'; it should be 'A. Troisi'.","section":"References"},{"comment":"The manuscript twice calls the QMC method 'first-principle' even though the phenomenological ansatz in Eq. (11) is used in conjunction with the QMC data. The wording is not internally inconsistent, but a brief clarification of which parts of the analysis are first-principle and which are phenomenological would avoid confusion.","section":"Introduction/Conclusions"},{"comment":"The convergence of ED with respect to the phonon occupation cutoff N_M is slow (relative difference up to 10%), and the paper notes that the low-frequency enhancement at ω ≲ 0.5 meV is not clearly converged. This limitation is stated, but it would be helpful to state explicitly that the ED data are used only for the 1D 5-site lattice and are not used in the RTA-inconsistency claim for the 2D system.","section":"Section 3, exact diagonalization"}],"recommendation":"major_revision","confidential_remarks":"The most serious issue is Eq. (9). It is possible that the correct denominator ε_l − ε_k was used in the code and the paper contains only a typographical error, but the manuscript must be corrected and the comparison rerun or verified. Because the RTA-inconsistency claim is the paper's main scientific message, this is a load-bearing point that prevents acceptance in the present form. The fit-parameter issue (no uncertainty on μ0) is also substantive but easier to address. I recommend major revision rather than rejection, since the underlying QMC data and the overall methodology appear sound and the issues are fixable within the manuscript's scope."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The paper has a good idea: use sub-permille QMC data for G(τ) to test the static and relaxation-time approximations for μ(ω) in a rubrene-like model. The covariance-aware χ² comparison is well posed, the QMC data look genuinely high precision, and the one-parameter phenomenological fit that beats both approximations is a legitimate empirical result. The code is public. But the central RTA claim is compromised by Eq. (9) as printed: the Lorentzian denominator is (ω−ε_k−ε_l)^2, so the τ_loc^{-1}→0 limit selects ω=ε_k+ε_l instead of ω=ε_l−ε_k from the Kubo formula (4). A uniform shift of all ε_k then shifts the RTA spectrum, which violates gauge invariance. If the code used the printed formula, the Fig. 1 RTA curves are not the standard RTA; if the code used ε_l−ε_k, the equation is misprinted and the comparison must be rerun and documented. Either way, the paper needs to address this before the central inconsistency claim can be evaluated.\n\nThe mu0 estimate also rests on a single-parameter ansatz with no reported uncertainty, so the claim that mu0 is lower than experiment is not yet quantitative. The tau_loc scan uses only two values (6 meV and 2.7 meV), making the strong statement that 'any finite tau_loc worsens the fit' a bit thinly supported. The positive-weight assumption for HMC is empirical rather than proven, but at T≈25 meV that is a minor concern.\n\nWhat the paper does well: it demonstrates that transient localization shows up as a tiny correction to G(τ) yet dominates low-frequency μ(ω), and it gives a concrete route to constrain phenomenological models. That is worth a serious referee.\n\nMy recommendation: send it to peer review, but only after the sign issue is resolved. The comparison must be rerun or the equation corrected, and the mu0 uncertainty should be reported.","headline":"A promising QMC-vs-phenomenology comparison that is currently undermined by a sign error in the central RTA formula, Eq. (9).","tokens_in":10476,"tokens_out":2006,"would_cite":false,"duration_ms":21425,"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":"QMC rules out relaxation-time model for organic mobility.","keywords":["organic semiconductors","charge transport","transient localization","Quantum Monte-Carlo","relaxation time approximation","optical conductivity","tight-binding model","low-frequency mobility"],"falsifier":"If a sign-reweighted QMC calculation at $T=25$ meV on the same $21\\times 21$ lattice finds any configurations with negative weight, and restoring the exact sign shifts $G(\\tau)$ by more than the quoted statistical error, then the RTA-inconsistency comparison is not settled.","tokens_in":9406,"feed_emoji":"⚡","tokens_out":10602,"duration_ms":92801,"temperature":0.7,"pith_summary":"The paper argues that high-precision Quantum Monte-Carlo (QMC) simulations of a rubrene-like tight-binding model can discriminate between competing pictures of low-frequency charge transport in molecular organic semiconductors, even though the transport-relevant effects are only about a 0.1% correction to the simulated correlators. It establishes that the simple relaxation-time approximation (RTA), in which a single phenomenological parameter broadens the static-disorder spectrum, is not consistent with the exact QMC data: moving from static disorder to any finite relaxation time increases the deviation of the predicted imaginary-time correlator from the QMC result. Combining the QMC correlators with the static-disorder approximation yields a fitted zero-frequency mobility that is smaller than the RTA prediction and below the experimental rubrene mobility of $(9.25 \\pm 0.75)$ cm$^2$/(V s). The paper concludes that RTA must be modified, for instance by allowing an energy-dependent relaxation time, and that a similar QMC-plus-effective-theory strategy could improve spectral reconstruction in other systems.","feed_headline":"QMC rules out relaxation-time model for organic mobility","feed_subtitle":"Sub-permille precision shows any finite relaxation time worsens the match, pointing to energy-dependent scattering.","key_machinery":"The central object is the imaginary-time velocity correlator $G(\\tau)$, computed by Hybrid Monte-Carlo from the path-integral representation in which the weight includes $\\operatorname{Tr} U(0,\\beta)$ times $e^{-S_B}$, and velocity correlators are normalized by $\\operatorname{Tr} U(0,\\beta)$ to suppress heavy-tailed log-normal fluctuations. Since QMC outputs $G(\\tau)$ rather than the mobility $\\mu(\\omega)$, every approximation (static disorder, RTA, and the phenomenological fit) is translated into $G(\\tau)$ through the Green-Kubo relation and compared directly with the QMC data. The static and RTA approximations are built from eigenstates of the single-particle Hamiltonian in random static phonon fields, with RTA adding a $\\tau_{loc}$ broadening; the phenomenological fit splices a quadratic low-frequency form onto the static mobility and fixes its normalization from $G(0)$.","core_discovery":"The central discovery is that the relaxation time approximation (RTA) for charge mobility in organic semiconductors cannot be uniformly valid. When the RTA mobility is converted through the standard imaginary-time relation into a correlator $G_{RTA}(\\tau)$ and compared with the exact QMC correlator $G(\\tau)$, the match is worse for any finite relaxation time $\\tau_{loc}$ than for the static-disorder limit $\\tau_{loc}^{-1}\\to 0$; the paper states this holds for both the $5\\times 1$ and $21\\times 21$ lattices. Transient localization therefore enters QMC data as a tiny, sub-permille deviation, yet it dominates the real-time dynamics at low frequencies. A phenomenological fit with a single zero-frequency mobility parameter $\\mu_0$ describes the QMC correlator to better than 0.05%, and the optimal $\\mu_0$ is below both the RTA prediction and the experimental rubrene value.","pith_inferences":["The paper does not pursue it, but the same protocol—translate any candidate $\\mu(\\omega)$ into $G(\\tau)$ and compare against sub-permille QMC data—could be turned into a general validity test for transport models, assigning a quantitative error to a spectrum rather than a qualitative pass/fail.","If the RTA failure is cured by an energy-dependent $\\tau_{loc}$, the QMC data could be used to fit that energy dependence directly; a successful two-parameter form would then make testable predictions for other transport observables such as the Seebeck coefficient or Hall angle.","The tentative low-frequency enhancement below 0.5 meV seen in exact diagonalization, if it survives the $N_M\\to\\infty$ limit, would indicate a transport channel absent from both the static and RTA spectra; current statistical errors are too large to confirm or exclude it.","The analogy to quark-gluon plasma, which the paper notes at the end, suggests that lattice QCD transport coefficients could benefit from the same hybrid strategy of using effective-theory spectral shapes as priors on high-precision Euclidean correlators."],"forward_implications":["The simple RTA with a single $\\tau_{loc}$ cannot be uniformly valid; phenomenological transport models should allow an energy-dependent relaxation time.","High-precision QMC data, even when the target effect is only a ~0.1% correction, can meaningfully discriminate between competing mobility spectra.","For rubrene-like parameters, the fitted zero-frequency mobility is below the RTA prediction and below the measured $9.25$ cm$^2$/(V s), so the RTA overestimates low-frequency mobility.","Finite-volume effects are large for small lattices, but the qualitative behaviour of $\\mu(\\omega)$ is similar for the $5\\times 1$ and $21\\times 21$ systems, so small-lattice exact diagonalization remains useful for identifying low-frequency features."],"supporting_citations":[{"why":"Supplies the triangular-lattice tight-binding Hamiltonian, the rubrene parameter set, the transient-localization framework, and the experimental mobility reference.","marker":"[1]"},{"why":"Provides the charge-transport regime analysis of thermal off-diagonal disorder that motivates the dynamical-disorder picture.","marker":"[5]"},{"why":"Provides the displaced-Drude-peak spectral feature of transient localization that the low-frequency comparisons are designed to resolve.","marker":"[8]"},{"why":"Presents the Hybrid Monte-Carlo algorithm and the high-precision QMC data for the single carrier Hamiltonian that this paper extends to low-frequency mobility.","marker":"[11]"},{"why":"Introduces the exact Fourier acceleration that keeps HMC autocorrelations low, underpinning the sub-permille statistical errors in $G(\\tau)$.","marker":"[12]"},{"why":"Formulates the transient localization scenario and the relaxation time approximation used as the baseline comparison.","marker":"[21]"},{"why":"Fixes the variances of the normally distributed static phonon fields used in the static and RTA mobility calculations.","marker":"[22]"},{"why":"Provides the experimental hole mobility of rubrene, $(9.25 \\pm 0.75)$ cm$^2$/(V s), against which the fitted $\\mu_0$ is compared.","marker":"[23]"}],"fun_headline_variants":["High-precision QMC nixes relaxation-time mobility for organics","Transient localization defeats relaxation-time mobility","Sub-permille deviation invalidates relaxation-time model"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"Two load-bearing premises are that the QMC path-integral weight stays positive in the simulated regime, so the sub-permille correlator is unbiased, and that the phenomenological fitting form of Eq. (11), though not derived from the Hamiltonian, adequately represents the low-frequency mobility spectrum.","fun_headline_variants_meta":{"raw":{"variants":["High-precision QMC nixes relaxation-time mobility for organics","Transient localization defeats relaxation-time mobility","Sub-permille deviation invalidates relaxation-time model"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.001244,"raw_usage":{"total_tokens":5073,"prompt_tokens":881,"completion_tokens":4192,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":497,"completion_tokens_details":{"reasoning_tokens":4142}},"tokens_in":497,"tokens_out":4192,"duration_ms":31555,"temperature":1.0,"reasoning_tokens":4142,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-12T12:06:48.911888+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"If a sign-reweighted QMC calculation at $T=25$ meV on the same $21\\times 21$ lattice finds any configurations with negative weight, and restoring the exact sign shifts $G(\\tau)$ by more than the quoted statistical error, then the RTA-inconsistency comparison is not settled.","supporting_citations":[{"cited_title":"8 Lattice model of organic semiconductors Pavel Buividovich","cited_arxiv_id":null,"evidence_quote":"Supplies the triangular-lattice tight-binding Hamiltonian, the rubrene parameter set, the transient-localization framework, and the experimental mobility reference."},{"cited_title":"Troisi, G","cited_arxiv_id":null,"evidence_quote":"Provides the charge-transport regime analysis of thermal off-diagonal disorder that motivates the dynamical-disorder picture."},{"cited_title":"Transient localization from the interaction with quantum bosons","cited_arxiv_id":"2312.03840","evidence_quote":"Provides the displaced-Drude-peak spectral feature of transient localization that the low-frequency comparisons are designed to resolve."},{"cited_title":"The transient localization scenario for charge transport in crystalline organic materials","cited_arxiv_id":"1505.02686","evidence_quote":"Formulates the transient localization scenario and the relaxation time approximation used as the baseline comparison."},{"cited_title":"Nematiaram, A","cited_arxiv_id":null,"evidence_quote":"Fixes the variances of the normally distributed static phonon fields used in the static and RTA mobility calculations."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides the experimental hole mobility of rubrene, $(9.25 \\pm 0.75)$ cm$^2$/(V s), against which the fitted $\\mu_0$ is compared."}],"review_version":1}