{"id":"490000e7-50ec-46f0-ac6a-339edbcbd596","arxiv_id":"1908.09259","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":3,"one_line_summary":"STM/STS of MoSe2 twin domain boundaries reveals quantum well states and power-law density-of-states suppression attributed to a Tomonaga-Luttinger liquid, with no charge density wave down to 5 K.","lead":"This paper reports scanning tunneling spectroscopy of one-dimensional twin boundaries in monolayer MoSe2. The results show length-dependent energy gaps and power-law suppression of the electronic density of states, which the authors interpret as quantum confinement in a Tomonaga-Luttinger liquid and as evidence against a charge density wave.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"TLL claim rests on a power-law exponent the authors themselves call unreliable; the TLL signature may be contaminated by quantum confinement in the same long DB.","rationale":"The reader's weakest_assumption—that ruling out CDW assumes a length-independent Peierls gap—is a genuine concern, but I do not regard it as the most load-bearing element of the central claim. The no-CDW conclusion does not rest solely on the 1/L scaling: the paper reports that long >30 nm DBs show essentially no hard gap, only a power-law suppression, which argues against a conventional CDW gap. The DFT relaxation test, while limited, provides additional support. By contrast, the positive TLL claim hinges decisively on the exponent α derived from the power-law fit. The authors explicitly flag that the spatially-resolved spectra on the same long DB show quantum-confinement-like features, making the Kc estimate less reliable. No alternative explanations are fitted or excluded, and no statistics are provided. This is an internally stated limitation, not merely a disagreement with external consensus. Because the TLL identification is the headline claim, this is the weakest link. The paper remains a competent experimental study with a plausible but not quantitatively established TLL interpretation, so the CONDITIONAL verdict should stand, with the condition being a clean and statistically supported extraction of α or another decisive TLL observable.","tokens_in":4925,"tokens_out":10140,"duration_ms":108617,"concrete_test":"Fit the long-DB dI/dV spectra (multiple DBs, L>30 nm, and multiple positions along each DB) with a combined model N(E) = A|E|^α + B·N_conf(E;L) and also with a finite-gap model N(E) = A·Re((E−iΓ)/√((E−iΓ)^2−Δ^2)); report ΔAIC or χ² per degree of freedom. If the extracted α varies with L or with tip position, or if the finite-gap model fits equally well, the reported α=0.47 is not a clean TLL exponent and the TLL identification is unsubstantiated.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The most load-bearing weakness is the quantitative basis for the TLL claim. The exponent α=0.47±0.05 is extracted from a power-law fit to dI/dV on one >30 nm DB (Fig. 3(a,b)), but the spatially-resolved data on that same DB (Fig. 3(c)) show oscillations 'reminiscent of the quantum confinement effect,' and the authors concede this 'would make the estimate of the parameter Kc shown above less reliable.' If finite-size quantization contributes to the spectrum, the apparent power-law may not be the intrinsic TLL density of states; the fit range, background subtraction, and statistics across multiple DBs/positions are not reported, and alternative models (small hard gap, disordered wire, zero-bias anomaly) are not tested. Since α is the only quantitative TLL observable, the central claim of TLL signatures is not yet established. The CDW length-dependence assumption flagged by the reader is a real but secondary concern, because the absence of a hard gap in long DBs provides independent (though not airtight) evidence against CDW.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript reports low-temperature scanning tunneling microscopy/spectroscopy (LT-STM/S) measurements on 4|4P-type twin domain boundaries (DBs) in monolayer MoSe2. The authors observe (i) an energy gap at the Fermi level whose size scales approximately as 1/L for DB segments of length L from ~3 nm to ~40 nm, which they attribute to quantum confinement of a Tomonaga-Luttinger liquid (TLL); (ii) power-law suppression of the density of states near EF in a DB longer than 30 nm, with exponent alpha = 0.47 +/- 0.05 and derived Luttinger parameter Kc ~ 0.28; and (iii) a spatial shift of the DOS peak energy near a DB end consistent with r*E = constant. From the length dependence of the gap and from DFT relaxation tests of artificially distorted structures, the authors conclude that a charge density wave (CDW) is not supported down to ~5 K.","tokens_in":5149,"tokens_out":2087,"duration_ms":22453,"significance":"If the conclusions hold, this work would help resolve conflicting reports on the electronic nature of twin domain boundaries in MoSe2 (quantum well states vs. CDW vs. TLL) and would provide a clean, isolated 1D system for studying correlated-electron physics. The paper's strengths include the systematic length-dependent STS data, the direct visualization of quantum well states in finite DBs, and the explicit attempt to rule out CDW via a DFT relaxation test. However, the quantitative TLL claim rests on a single power-law fit whose reliability the authors themselves qualify, and the no-CDW conclusion depends on an unstated assumption about the length dependence of CDW gaps; these issues are load-bearing for the central claims.","major_comments":[{"comment":"The power-law exponent alpha = 0.47 +/- 0.05, the only quantitative TLL observable in this work, is extracted from a fit to one DB longer than 30 nm (Fig. 3(a,b)). The authors immediately concede in the same paragraph that the spatially resolved data on that same DB (Fig. 3(c)) are 'reminiscent of the quantum confinement effect' and that this 'would make the estimate of the parameter Kc shown above less reliable.' Because the fit range, the background subtraction, the number of independent DBs/positions, and the statistical uncertainty are not reported, the apparent power law cannot be distinguished from a finite-size quantization artifact or from alternative models (e.g., a small hard gap, a disordered wire, or a zero-bias anomaly). This is a central, load-bearing weakness for the TLL assignment; please provide additional spectra from multiple long DBs, quantify the fit quality and systematic dependence on the fit window, and test competing functional forms.","section":"Fig. 2(b) and the paragraph beginning 'In Fig. 2(a), we show...'"},{"comment":"The conclusion that the observed ~1/L gap scaling rules out CDW assumes that a Peierls/CDW gap would be independent of segment length. This assumption is not justified for finite segments, where boundary conditions, pinning, or confinement can modify the CDW gap. The authors state 'a constant, length-independent, gap would be otherwise expected [4],' but finite-size effects in 1D CDW systems can produce length-dependent gaps. The DFT relaxation test described in the same paragraph only tests an artificial 3a periodic distortion in the infinite periodic calculation, not a finite-length segment; it therefore does not directly address the length-dependence assumption. Please either provide a supporting argument or literature for length-independent CDW gaps in finite DBs, or soften the no-CDW claim accordingly.","section":"Fig. 2(b) and surrounding text"},{"comment":"The definition of the measured gap Eg and the quoted power law are not fully specified. For Fig. 2(a), the STS spectra are shown on a logarithmic scale but the criterion for extracting the gap size (e.g., crossing of linear fits, onset of nonzero dI/dV, or a fixed threshold) is not given. For Fig. 3(b), the power-law fit is shown as a red line but the fit range, the treatment of the finite temperature broadening, and the number of independent measurements are not stated. Without this information, the reader cannot assess whether the reported scaling and exponent are robust. Please add the numerical fitting procedure, error bars on Eg for each DB, and the number of spectra averaged per DB.","section":"Figs. 2 and 3"}],"minor_comments":[{"comment":"There are several typographical errors: 'lest square fitting' should be 'least squares fitting' (Fig. 2(b) caption); 'Tomonago-Luttinger' should be 'Tomonaga-Luttinger' in the summary; 'ddIV' in Fig. 4(b) text should be formatted as dI/dV.","section":"Abstract and figure captions"},{"comment":"The DFT Fermi-level upshift is a key input for comparing the calculated band structure to the experimental data, but the text only states that the Fermi level was upshifted to match experiment. Please specify the magnitude of the upshift and whether the resulting EF position (near one-third of the BZ edge) is consistent with the known doping level of the MBE-grown MoSe2 monolayer.","section":"Fig. 1(c) and the paragraph beginning 'Fig. 1(a) shows...'"},{"comment":"The r*E = constant analysis in Fig. 4 appears in the text but would benefit from a quantitative statement: how many spectra were used, over what range of r, and what is the uncertainty in the fitted constant 0.49 nm*eV? The current description is qualitative and the statistical significance is unclear.","section":"Paragraph beginning 'A more direct evidence of TLL...'"}],"recommendation":"major_revision","confidential_remarks":"This is a compact experimental report that addresses an active controversy, and the qualitative observations (length-dependent gaps, power-law-like DOS suppression) are interesting. However, the quantitative TLL claim and the no-CDW conclusion both require stronger support. In particular, the power-law exponent is presented as the central evidence but is self-described as unreliable due to quantum confinement contamination; additional data and analysis are needed. The no-CDW conclusion relies on an assumption about CDW gap length dependence that I would like to see justified or relaxed. I believe the manuscript is within the journal's scope and that the issues are fixable, but the revision needs to be substantive rather than cosmetic."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"You should know this paper is a clean LT-STM/STS study of 4|4P twin boundaries in monolayer MoSe2. The new thing is systematic length-dependent spectroscopy: short segments show gaps that scale roughly as 1/L, long segments show power-law DOS suppression near EF with exponent α≈0.47, and the authors argue against CDW using both the length dependence and DFT relaxation tests. That combination hasn't been reported before for this specific boundary type.\n\nThe paper does several things well. The length-dependent gap is the strongest observational claim, and the 1/L trend is easy to see in the data. The DFT test for Peierls distortion is a nice check: they show that artificially compressed lattices relax back to the undistorted structure, which independently weakens the CDW case. The authors are also honest about their own limitations, explicitly saying the Kc estimate is 'less reliable' because the long DB still shows quantum-confinement oscillations in spatially resolved spectra. Credit where due: that is the right way to flag a load-bearing uncertainty.\n\nThe soft spots are real but not fatal. The TLL assignment rests on a single power-law fit from one long DB, with no reported fit range, background subtraction, or statistics across multiple positions or DBs. Because the same DB shows standing-wave features, the power-law could be contaminated by finite-size quantization, and the extracted α may not be the intrinsic TLL exponent. The authors don't test alternative explanations like a small hard gap or a disordered wire. That is a genuine weakness, but the claimed α is presented as a signature, not a precision measurement, and they explicitly qualify it.\n\nThe second concern is the no-CDW conclusion, which assumes a CDW gap would be length-independent. For finite segments, pinning can alter the gap, so that assumption deserves justification. But this is a secondary issue, because the absence of a hard gap in long DBs and the DFT relaxation result provide independent, though not airtight, evidence against CDW.\n\nWho benefits? Researchers working on 1D physics in TMD defects, especially the STM/STS community. The paper is worth a serious referee, not a desk reject. A referee should ask for more extensive data: several long DBs, fits at multiple positions, and explicit error bars. As is, the central claim of TLL is plausible but not fully established; the confinement and CDW-negative parts are closer to solid.\n\nRecommendation: send it to peer review with a request for revision and more statistics.","headline":"A solid experimental report of quantum confinement and possible TLL signatures in MoSe2 twin boundaries; the CDW-negative part is stronger than the TLL-positive part.","tokens_in":5708,"tokens_out":1106,"would_cite":true,"duration_ms":14145,"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":"Monolayer MoSe2 twin domain boundaries are quantum-confined Tomonaga-Luttinger liquids, not charge density waves.","keywords":["twin domain boundary","monolayer MoSe2","Tomonaga-Luttinger liquid","quantum confinement","quantum well states","charge density wave","scanning tunneling spectroscopy","power-law density of states"],"falsifier":"A decisive experiment would measure a set of 3 nm to 40 nm boundaries at about 5 K with atomic resolution and test whether any long segment shows a constant, length-independent gap while shorter segments deviate from $1/L$ scaling; if such a length-independent gap appears, the argument against CDW would lose its main observational support.","tokens_in":4760,"feed_emoji":"🔬","tokens_out":9493,"duration_ms":81047,"temperature":0.7,"pith_summary":"This paper reports low-temperature scanning tunneling spectroscopy of the metallic 4|4P-type twin domain boundaries that form in monolayer MoSe2, line defects where two mirror-oriented semiconducting domains meet. It argues that short boundary segments behave as quantum wells, with the energy gap growing as $E_g \\sim 1/L$ as the segment length $L$ shrinks, while long segments show a power-law suppression of the density of states near the Fermi level with exponent $\\alpha \\approx 0.47$. Together these are taken as the signatures of a Tomonaga-Luttinger liquid, an interacting one-dimensional electron state. The data do not support the competing charge-density-wave picture at temperatures down to about 5 K.","feed_headline":"MoSe2 twin boundaries host a one-dimensional electron liquid","feed_subtitle":"Scanning tunneling data show gaps that shrink with segment length and power-law density suppression, not a charge-density wave.","key_machinery":"The central object is the 4|4P twin domain boundary, a one-dimensional metallic line defect embedded in semiconducting MoSe2 and confined by van der Waals substrates. The argument is carried by two spectroscopic signatures measured with a scanning tunneling microscope: the inverse-length scaling of the zero-mode gap, $E_g \\sim 1/L$, which marks quantum confinement of a Tomonaga-Luttinger liquid, and the power-law density-of-states suppression near the Fermi level, $N(E) \\sim E^\\alpha$, with $\\alpha = (K_c + K_c^{-1} - 2)/4$ and $K_c \\approx 0.28$. A third observable, the shift of the density-of-states peak energy with distance $r$ from the boundary end according to $E \\sim 1/r$, ties the data to the TLL edge theory. These signatures together separate the TLL picture from a charge density wave, which would instead produce a constant, length-independent gap and a periodic lattice distortion.","core_discovery":"The central claim is that 4|4P twin domain boundaries in monolayer MoSe2 are one-dimensional metals whose low-energy physics is set by quantum confinement and strong electron-electron interactions rather than by a Peierls charge-density-wave distortion. Short boundaries, with lengths from about 3 nm to 40 nm, show STS energy gaps that fit $E_g \\propto 1/L$, as expected for the zero-mode gap of a confined Tomonaga-Luttinger liquid. Long boundaries, over 30 nm, show no hard gap but a zero-bias density of states that vanishes as $N(E) \\propto E^\\alpha$ with $\\alpha = 0.47 \\pm 0.05$, from which the paper derives a Luttinger parameter $K_c \\approx 0.28$, indicating strong repulsive interactions. The paper also reports that the density-of-states peak moves to higher energy closer to a boundary end, consistent with the predicted $E \\propto 1/r$ edge behavior. It reads the absence of a length-independent gap and the relaxation of artificially Peierls-distorted DFT structures as evidence against CDW formation.","pith_inferences":["The paper does not directly observe spin-charge separation; a spin- and energy-resolved STS experiment on a single long boundary would test whether the TLL interpretation is complete.","The $1/L$ gap could also be read as ordinary Coulomb blockade in a metallic island; gate-dependent transport on individual segments would separate charging energy from confinement.","If $K_c \\approx 0.28$ is set by the defect geometry, similar exponents should appear in other twin-boundary types; a head-to-head measurement across different domain boundaries would show whether the defect structure controls the interaction strength.","The density-functional theory relaxation test only probes uniform 3a distortions, so local disorder-pinned distortions are not excluded; the no-CDW conclusion is strongest for uniform Peierls order."],"forward_implications":["Boundary length becomes a tuning knob: cutting the same defect into shorter segments raises the zero-mode gap continuously.","Long boundaries are not gapped insulators at 5 K; they conduct as interacting one-dimensional metals with strongly suppressed density of states at the Fermi level.","The strong repulsive interactions implied by $K_c \\approx 0.28$ mean transport through these wires should be power-law in temperature and bias rather than ohmic.","The edge peak shift $E \\propto 1/r$ provides a spatially local fingerprint that can identify Tomonaga-Luttinger behavior in other one-dimensional defects in monolayer materials."],"supporting_citations":[{"why":"supplies the finite-length zero-mode gap $E_g \\sim 1/L$ and the Tomonaga-Luttinger signatures used as the comparison for the 4|4P boundaries.","marker":"[4]"},{"why":"the earlier claim of a Peierls-type charge density wave in these boundaries, which the length-dependent gaps must rule out.","marker":"[2]"},{"why":"the prior photoemission report of TLL behavior in MoSe2 twin boundaries that motivates the low-temperature STS test.","marker":"[3]"},{"why":"the original observation of quantum well states and moiré effects in finite-length domain boundaries.","marker":"[1]"},{"why":"the Tomonaga-Luttinger liquid review providing the power-law density of states and the relation between $\\alpha$ and $K_c$.","marker":"[20]"},{"why":"the theory predicting that the density-of-states peak energy scales as $E \\sim 1/r$ from the one-dimensional edge.","marker":"[37]"},{"why":"the experimental STS demonstration of edge TLL power-law behavior used as a template for the $E \\sim 1/r$ analysis.","marker":"[38]"}],"fun_headline_variants":["MoSe2 twin boundaries reveal a confined Luttinger liquid","Twin boundaries in MoSe2 host a Luttinger liquid","No CDW: MoSe2 twin boundaries show Luttinger liquid","Quantum confinement drives TLL in MoSe2 twin boundaries","One-dimensional electron liquid without CDW in MoSe2 twin boundaries"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The no-CDW conclusion rests on assuming that a Peierls gap would be independent of segment length, so the measured $1/L$ dependence rules it out; finite segments can have confinement-dependent CDW gaps, and the paper does not justify why that should not apply here.","fun_headline_variants_meta":{"raw":{"variants":["MoSe2 twin boundaries reveal a confined Luttinger liquid","Twin boundaries in MoSe2 host a Luttinger liquid","No CDW: MoSe2 twin boundaries show Luttinger liquid","Quantum confinement drives TLL in MoSe2 twin boundaries","One-dimensional electron liquid without CDW in MoSe2 twin boundaries"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000721,"raw_usage":{"total_tokens":3183,"prompt_tokens":839,"completion_tokens":2344,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":455,"completion_tokens_details":{"reasoning_tokens":2255}},"tokens_in":455,"tokens_out":2344,"duration_ms":14949,"temperature":1.0,"reasoning_tokens":2255,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T11:16:36.752379+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A decisive experiment would measure a set of 3 nm to 40 nm boundaries at about 5 K with atomic resolution and test whether any long segment shows a constant, length-independent gap while shorter segments deviate from $1/L$ scaling; if such a length-independent gap appears, the argument against CDW would lose its main observational support.","supporting_citations":[{"cited_title":"Jolie et al., Physical Review X 9, 011055 (2019)","cited_arxiv_id":null,"evidence_quote":"supplies the finite-length zero-mode gap $E_g \\sim 1/L$ and the Tomonaga-Luttinger signatures used as the comparison for the 4|4P boundaries."},{"cited_title":"Barja et al., Nature Physics 12, 751 (2016)","cited_arxiv_id":null,"evidence_quote":"the earlier claim of a Peierls-type charge density wave in these boundaries, which the length-dependent gaps must rule out."},{"cited_title":"Ma et al., Nature communications 8, 14231 (2017)","cited_arxiv_id":null,"evidence_quote":"the prior photoemission report of TLL behavior in MoSe2 twin boundaries that motivates the low-temperature STS test."},{"cited_title":"Liu et al., Physical Review Letters 113, 066105 (2014)","cited_arxiv_id":null,"evidence_quote":"the original observation of quantum well states and moiré effects in finite-length domain boundaries."},{"cited_title":"Eggert, Phys Rev Lett 84, 4413 (2000)","cited_arxiv_id":null,"evidence_quote":"the theory predicting that the density-of-states peak energy scales as $E \\sim 1/r$ from the one-dimensional edge."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"the experimental STS demonstration of edge TLL power-law behavior used as a template for the $E \\sim 1/r$ analysis."}],"review_version":1}