{"id":"90f5af45-dbe1-4857-95d4-fe3f76bd62a0","arxiv_id":"2412.12995","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":5,"one_line_summary":"DMRG simulations of RIXS in the 1D Hubbard-Holstein model show phonon peak intensities gain momentum dependence from electron itinerancy, beyond the single-site Lang-Firsov prediction.","lead":"This paper uses a numerical method called DMRG to compute what happens when X-rays scatter off a one-dimensional correlated material with electron-lattice coupling, going beyond the standard single-site approximation. The results show that electron motion, not just local lattice distortion, shapes the phonon signal, which may change how experimental RIXS data are analyzed in cuprates and other quantum materials.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The central q-dependence of I2/I1 in Fig. 4 is extracted from DMRG RIXS spectra with no convergence tests in m, M, or L; a truncation or center-site artifact could mimic the itinerancy effect.","rationale":"To support the central claim--that RIXS phonon intensities acquire momentum dependence from electron itinerancy even for momentum-independent coupling--the computed spectra must faithfully represent the spatial extent of the phonon-generating intermediate state. The observable used, I2(q)/I1(q), is a ratio of peak intensities at the two momenta where magnetic background is claimed to vanish; it is a delicate quantity. The authors provide XAS convergence checks (App. D), but XAS is a local response and is insensitive to the same long-range propagation errors that affect RIXS. In DMRG, keeping m=500 states is usually adequate for ground-state energies, but dynamical response functions--especially those involving two-point correlations with oscillatory phases--converge more slowly. The center-site approximation likewise introduces an asymmetry that could produce an artificial momentum dependence at the zone boundary. Thus the single most load-bearing concern is the unverified numerical reliability of the exact observable on which the headline conclusion rests. The proposed test--rerunning at larger m, M, and L, plus an exact-diagonalization cross-check--would directly determine whether the q-dependence is physical. This does not require rejecting the paper; it requires supplying the missing validation. The reader's CONDITIONAL verdict is appropriate, and my analysis agrees with the reader's weakest_assumption.","tokens_in":21166,"tokens_out":10547,"duration_ms":104868,"concrete_test":"Recompute the Fig. 4(a) spectra with m=1000, M=20, and L=48 and extract I2(q)/I1(q) at q=0 and q=pi/a; additionally, cross-check the DMRG pipeline against exact diagonalization on L=4-6 chains (with phonon truncation M=12) for the same U, Omega, lambda, Vc, and Gamma. If the difference [I2(pi/a)/I1(pi/a) - I2(0)/I1(0)] changes sign or shrinks by more than 20%, the mobility-induced q-dependence claim is not established.","verdict_should_be":"UNCHANGED","load_bearing_attack":"Section III B and Fig. 4 present the central evidence: for Vc=-4t, I2(q)/I1(q) is larger at q=pi/a than at q=0, while for Vc=-12t the ratio is q-independent; this is interpreted as electron-mobility-induced delocalization pathways. These ratios are extracted from DMRG RIXS spectra computed with m=500, M=10-17, on L=24 open chains, using the center-site approximation (Eq. B2). Appendix D demonstrates convergence only for XAS, a one-point response, comparing m=500 vs 600 and M=16 vs 17; no RIXS convergence data are given. RIXS requires propagating the core-hole intermediate state to sites j across the chain; truncation errors grow with distance and are not captured by a local XAS test. The q=0 and q=pi/a Fourier components are the most sensitive to long-range tails and boundary effects of the open chain, so the specific observable used to support the central claim is exactly the one most likely to be biased. The paper's 'numerically exact' phrasing is also undermined by the unvalidated center-site approximation for the RIXS response. If the q-dependent intensity ratio disappears under stricter truncation or larger L, the central distinction from the SS-LF prediction collapses.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"This manuscript develops a DMRG-based theoretical framework for computing XAS and RIXS spectra of the one-dimensional half-filled Hubbard-Holstein model, explicitly including the core-hole potential and core-hole-lattice coupling. The central claim is that, even for momentum-independent bare electron-phonon coupling, the RIXS phonon intensities acquire momentum dependence due to electron itinerancy, and that this effect persists deep into the Mott-insulating regime. The authors compare their numerical spectra with the single-site Lang-Firsov (SS-LF) model, emphasizing that the common SS-LF analysis of experiments can underestimate or mischaracterize the electron-phonon coupling. They also discuss how the core-hole potential, core-hole lifetime, Hubbard U, and core-hole-lattice coupling modify the spectra.","tokens_in":21424,"tokens_out":2704,"duration_ms":29439,"significance":"If the central claim is correct, the paper makes a valuable contribution: it provides a numerically exact (in the DMRG sense) treatment of RIXS in an extended correlated electron-phonon system and directly challenges the widely used SS-LF interpretation of phonon RIXS intensities. The parameter-free nature of the calculations, in the sense that no experimental data are fitted, is a strength, as is the explicit treatment of both spin and phonon excitations and the comparison against a well-defined limiting model. The paper also gives concrete falsifiable predictions, such as the momentum dependence of I2/I1 and its dependence on the core-hole potential, which could be tested experimentally at different absorption edges. However, the credibility of the central claim rests on numerical convergence evidence that is currently incomplete.","major_comments":[{"comment":"The convergence tests reported in Appendix D and referenced in Sec. II C validate only the XAS spectra, not the RIXS response or the intensity ratios I2(q)/I1(q) that carry the paper's central claim. The central evidence in Sec. III B and Fig. 4 is obtained from RIXS spectra computed with m=500, M=10-17, on L=24 open chains, but no comparison is shown for the RIXS spectra at larger m, larger M, or larger L. Since RIXS requires propagating the intermediate state to sites j across the chain, truncation errors can be larger and less localized than in XAS. The authors should provide convergence checks for the RIXS spectra and, in particular, for the q=0 and q=pi/a values of I2/I1, as a function of m, M, and L.","section":"Sec. II C and Appendix D"},{"comment":"The center-site approximation in Eqs. (B2) and (B6) is used for the RIXS calculations but is never validated for the RIXS response itself. Equation (B2) involves a sum over all sites j with phase factors exp[iq*(Rj - Rc)], so the Fourier components at q=0 and q=pi/a are precisely the ones most sensitive to long-range contributions and to boundary effects of the open chain. The statement in Sec. II C that the approach is 'numerically exact' is also too strong given that this approximation is unvalidated for the computed observable. The authors should test the center-site approximation against a full summation on smaller systems, or at least compare results obtained with different center-site choices and different chain lengths.","section":"Sec. II C and Eq. (B2)"},{"comment":"The conclusion that I2(q)/I1(q) is larger at q=pi/a than at q=0 for Vc=-4t, and q-independent for Vc=-12t, is drawn from normalized spectra shown in Fig. 4 without reporting the numerical values of these ratios or any estimate of their uncertainty. The authors state that the phonon peaks are isolated at these momenta because the spin-conserving magnetic weight vanishes at q=0 and pi/a, but the spectra shown still contain overlapping features at finite energy width. To make the central claim quantitative and verifiable, the authors should extract I1 and I2 by fitting the peaks (e.g., with Lorentzians or a defined integration window) and report the resulting ratios with error estimates, for each parameter set and for at least a few convergence parameters.","section":"Sec. III B and Fig. 4"}],"minor_comments":[{"comment":"There are several typographical errors in Sec. III D, including 'We nowxturn toases', 'wawhich favorslectron delocalization', and 'allow cs mpeting CDW correlations'; these should be corrected.","section":"Sec. III D"},{"comment":"The notation for the core-hole potential is inconsistent: Eq. (4) uses VCH while the text and figures use Vc. Please unify the notation.","section":"Eq. (4) and throughout"},{"comment":"The caption statement 'The SS-LF predicts no momentum dependence for all cases, while our calculations only observe this in (c)' is ambiguous; it should be rephrased to clarify that only the Vc=-12t case shows q-independent intensity in the calculations.","section":"Fig. 4 caption"},{"comment":"The convergence tests are shown only for U=8t and lambda=0.5; the paper would benefit from at least brief convergence statements for the smaller-U and nonzero-lambda_CH parameter regimes where the conclusions about itinerancy and core-hole-lattice effects are drawn.","section":"Appendix D"}],"recommendation":"major_revision","confidential_remarks":"The manuscript addresses an important problem and the DMRG approach is state of the art. The main concern is that the central observable, the momentum-dependent RIXS phonon intensity ratio, lacks direct convergence validation. This is fixable by adding RIXS-specific convergence tests and quantitative peak-intensity ratios; it does not appear to require new physics or a change of scope. If the authors provide those, the paper could be suitable for publication."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Honest take: this is the first DMRG-based RIXS calculation for a correlated electron-phonon model on an extended lattice, and it delivers a clean, well-motivated result. The main finding—that electron mobility produces momentum-dependent phonon intensities even for momentum-independent bare coupling, with a deep core-hole potential suppressing the effect—extends the earlier dilute-limit work to finite density and the Mott regime. The treatment of core-hole-lattice coupling is also new in this setting. The paper is clearly written, the parameter choices are transparent, and no experimental data were fitted. That is real value.\n\nThe soft spot is the convergence evidence. Appendix D only checks XAS, not the RIXS response, and the central observable in Fig. 4—the I2/I1 ratio at q=0 and q=pi/a—is exactly the kind of quantity that could be biased by truncation and open-boundary artifacts. The center-site approximation adds further uncertainty. The authors call the results 'numerically exact,' which fails to acknowledge these approximations. The difference in Fig. 4a is small; without convergence tests on the RIXS spectra themselves, a truncation artifact could plausibly mimic the itinerancy effect. This is a legitimate concern, and the authors should be asked to address it.\n\nThat said, I don't think the concern is fatal. The qualitative physics is consistent with earlier dilute-limit results, and the Vc-dependence provides an internal check. The experimental implications for doped cuprates are speculative, but clearly framed as implications rather than direct predictions. The paper deserves a serious referee, not a desk rejection. With RIXS convergence tests added, or the 'numerically exact' language softened, I'd be comfortable with it. I'd bring it to a reading group interested in either RIXS or tensor-network methods.","headline":"A useful, clearly-written DMRG-RIXS study of the Hubbard-Holstein model whose central claim is plausible but whose lack of RIXS convergence tests keeps it from being 'numerically exact'.","tokens_in":21997,"tokens_out":3496,"would_cite":true,"duration_ms":33782,"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":"Phonon peak ratios in RIXS reveal electron motion, not just coupling strength.","keywords":["resonant inelastic x-ray scattering","electron-phonon coupling","Hubbard-Holstein model","density matrix renormalization group","Mott insulator","core-hole potential","phonon excitations","one-dimensional correlated systems"],"falsifier":"Recompute the RIXS spectra for the same parameters with larger bond dimension and phonon truncation, or with a time-dependent DMRG approach, and check whether the momentum dependence of $I_2(q)/I_1(q)$ at $V_c = -4t$ survives; if it disappears, the central claim collapses. Alternatively, measure the first-phonon-overtone intensity ratio across the Brillouin zone at both the Cu L-edge and the O K-edge of a 1D cuprate such as Sr$_2$CuO$_3$: the O K-edge, with a deeper core-hole potential, should show a flatter ratio if the core-hole-potential mechanism is right.","tokens_in":20934,"feed_emoji":"🔬","tokens_out":5775,"duration_ms":48689,"temperature":0.7,"pith_summary":"This paper asks whether resonant inelastic x-ray scattering (RIXS) spectra of a strongly correlated electron-phonon system can be read with the simple single-site Lang-Firsov model commonly used to extract coupling strengths from experiments. Using DMRG simulations of the one-dimensional half-filled Hubbard-Holstein model, the authors compute the full RIXS response while treating electron hopping, Hubbard repulsion, core-hole potential, and core-hole-lattice coupling on equal footing. They find that phonon excitations acquire momentum-dependent intensities even when the bare electron-phonon coupling is momentum-independent, because the intermediate-state doublon can hop before the core hole decays. The size of this effect is controlled by the core-hole potential: deep potentials localize the doublon and restore the single-site prediction, while shallow potentials do not. If correct, the work implies that previously published electron-phonon coupling strengths extracted from RIXS data with single-site models could be systematically underestimated.","feed_headline":"Phonon peak ratios in RIXS reveal electron motion","feed_subtitle":"Standard single-site fits miss this effect, so published electron-phonon coupling strengths may be too low.","key_machinery":"The load-bearing object is the intermediate-state doublon and its delocalization pathways. In the Kramers-Heisenberg scattering amplitude, the photon creates a core hole and an excited electron; the resulting doublon can either stay at the core-hole site and shake the local oscillator (the single-site Lang-Firsov channel) or hop away and back, creating phonons on neighboring sites. The relative weight of these pathways is set by the core-hole potential $V_c$ and the core-hole lifetime $\\Gamma$. The paper computes the spectra with the DMRG correction-vector method on 24-site chains, which treats all these processes without operator or diagrammatic expansions; the single-site Lang-Firsov amplitude is derived in Appendix A as the reference limit.","core_discovery":"The central result is that in the half-filled Hubbard-Holstein model, the RIXS phonon overtone intensities become momentum-dependent even when the bare electron-phonon coupling $g$ is constant. In the single-site Lang-Firsov limit, the ratio $I_2(q)/I_1(q)$ of two-phonon to one-phonon intensity is constant in momentum; the DMRG spectra instead show $I_2(\\pi/a)/I_1(\\pi/a)$ larger than the value at $q=0$ for a shallow core-hole potential $V_c = -4t$. This momentum dependence arises from electron mobility: the doublon created in the intermediate state can delocalize, excite phonons on neighboring sites, and return, processes that interfere with the purely local Franck-Condon pathway. Deepening the core-hole potential to $V_c = -12t$ suppresses these delocalization pathways and the spectra approach the single-site prediction, identifying the core-hole potential as the switch that controls how much itinerancy appears in the spectra. A second result is that the phonon feature observed in RIXS tracks the renormalized phonon dispersion obtained from the dynamical structure factor, including softening and Kohn-anomaly features when charge fluctuations hybridize with phonons at smaller $U$.","pith_inferences":["A direct experimental test is to measure the momentum-dependent phonon overtone intensity ratio at the Cu L-edge and the O K-edge of the same quasi-1D cuprate: the deeper core-hole potential at the K-edge should suppress the momentum dependence if the mechanism is right.","The center-site approximation used in the DMRG correction-vector method is the main uncontrolled approximation for the RIXS response; since the paper validates convergence only for the absorption spectra, a time-dependent DMRG or larger-bond-dimension check of the intensity ratios would settle whether the central claim is an artifact.","If the effect survives in two dimensions, the momentum dependence of phonon overtones could serve as a probe of polaron delocalization and core-hole screening in layered cuprates, not just chain compounds.","The results suggest a practical guideline for future analyses: single-site fits are safer for deep core-hole edges and strongly correlated regimes, while extended-lattice modeling is required for shallow edges in itinerant or doped systems."],"forward_implications":["Electron-phonon coupling strengths extracted from RIXS data with the single-site Lang-Firsov model are likely underestimated when the core-hole potential is shallow, for example at transition-metal L-edges.","Doping a Mott insulator will reduce phonon excitation intensity independent of screening, so coupling strengths extracted from doped cuprates may need revision.","RIXS phonon dispersion reflects the renormalized phonon dispersion including hybridization with charge excitations, supporting interpretations of dispersive CDW phonon anomalies in cuprates.","A core-hole-lattice coupling acts mostly as a local effective coupling $g_{\\mathrm{eff}} = g + g_{\\mathrm{CH}}$, but with nontrivial momentum-dependent residuals at shallow core-hole potentials, such as zone-boundary phonons when the local effective coupling vanishes.","Comparing edges with different core-hole potentials (for example ligand K-edges versus transition-metal L-edges) could disentangle the intrinsic electron-phonon coupling from core-hole-lattice coupling and delocalization effects."],"supporting_citations":[{"why":"Supplies the single-site Lang-Firsov model and the formula relating phonon overtone intensities to the coupling strength, the baseline that the DMRG results are compared against.","marker":"[2]"},{"why":"Shows that relaxing the single-site assumption to multiple modes in the atomic limit changes line shapes and relative intensities, motivating the extended-lattice treatment.","marker":"[24]"},{"why":"Demonstrates in the dilute limit that electron mobility introduces additional momentum dependence, which this paper extends to finite density and the Mott-insulating regime.","marker":"[25]"},{"why":"Provides the concept of core-hole-lattice (exciton-phonon) coupling, parameterized here as $g_{\\mathrm{CH}}$, and its expected sign from electrostatic interactions.","marker":"[27]"},{"why":"Introduces the DMRG method for computing RIXS spectra that is used for all calculations in this work.","marker":"[28]"},{"why":"Identifies spin-conserving and non-spin-conserving channels in quasi-1D RIXS, which the paper exploits to isolate phonon contributions at $q=0$ and $q=\\pi/a$.","marker":"[35]"},{"why":"Underpins the Krylov-space correction vector algorithm that the DMRG spectral calculations are built on.","marker":"[38]"},{"why":"Assigns the well-screened and poor-screened XAS resonances used to tune the incident photon energies in the RIXS calculations.","marker":"[45]"}],"fun_headline_variants":["RIXS phonon overtone ratio turns momentum-dependent for shallow core holes","Shallow core holes make RIXS phonon ratios reveal electron motion","Electron delocalization changes RIXS phonon overtone momentum signature","Core-hole depth controls momentum dependence of RIXS phonon peaks","Mobile electrons distort RIXS phonon overtone intensity pattern"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The whole argument relies on the DMRG spectra being numerically converged with respect to bond dimension, phonon truncation, and the center-site approximation, but the paper only shows convergence tests for the absorption spectra, not for the RIXS response or the intensity ratios that carry the central claim.","fun_headline_variants_meta":{"raw":{"variants":["RIXS phonon overtone ratio turns momentum-dependent for shallow core holes","Shallow core holes make RIXS phonon ratios reveal electron motion","Electron delocalization changes RIXS phonon overtone momentum signature","Core-hole depth controls momentum dependence of RIXS phonon peaks","Mobile electrons distort RIXS phonon overtone intensity pattern"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000538,"raw_usage":{"total_tokens":2607,"prompt_tokens":995,"completion_tokens":1612,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":611,"completion_tokens_details":{"reasoning_tokens":1518}},"tokens_in":611,"tokens_out":1612,"duration_ms":12952,"temperature":1.0,"reasoning_tokens":1518,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-11T13:30:46.046916+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Recompute the RIXS spectra for the same parameters with larger bond dimension and phonon truncation, or with a time-dependent DMRG approach, and check whether the momentum dependence of $I_2(q)/I_1(q)$ at $V_c = -4t$ survives; if it disappears, the central claim collapses. Alternatively, measure the first-phonon-overtone intensity ratio across the Brillouin zone at both the Cu L-edge and the O K-edge of a 1D cuprate such as Sr$_2$CuO$_3$: the O K-edge, with a deeper core-hole potential, should show a flatter ratio if the core-hole-potential mechanism is right.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the single-site Lang-Firsov model and the formula relating phonon overtone intensities to the coupling strength, the baseline that the DMRG results are compared against."},{"cited_title":"Geondzhian and K","cited_arxiv_id":null,"evidence_quote":"Shows that relaxing the single-site assumption to multiple modes in the atomic limit changes line shapes and relative intensities, motivating the extended-lattice treatment."},{"cited_title":"Bieniasz, S","cited_arxiv_id":null,"evidence_quote":"Demonstrates in the dilute limit that electron mobility introduces additional momentum dependence, which this paper extends to finite density and the Mott-insulating regime."},{"cited_title":"Geondzhian and K","cited_arxiv_id":null,"evidence_quote":"Provides the concept of core-hole-lattice (exciton-phonon) coupling, parameterized here as $g_{\\mathrm{CH}}$, and its expected sign from electrostatic interactions."},{"cited_title":"Nocera and G","cited_arxiv_id":null,"evidence_quote":"Underpins the Krylov-space correction vector algorithm that the DMRG spectral calculations are built on."},{"cited_title":"Kourtis, J","cited_arxiv_id":null,"evidence_quote":"Assigns the well-screened and poor-screened XAS resonances used to tune the incident photon energies in the RIXS calculations."}],"review_version":1}