{"id":"2b18f84d-a651-454f-ba70-594181785ebb","arxiv_id":"2505.05865","paper_version":1,"verdict":"CONDITIONAL","confidence":"HIGH","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":3,"one_line_summary":"The direction of light reflection and emission from molecular monolayers is controlled by the thickness of the hBN substrate through Fabry-Perot interference.","lead":"Researchers measured how a molecular monolayer on a thin hBN crystal reflects and emits light at different angles. They show that simply changing the hBN thickness steers the direction of the light, because of optical interference inside the flake.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Per-sample oscillator-strength fitting weakens the thickness-only explanation of the exact emission direction; a fixed-parameter re-simulation would settle it.","rationale":"Read in good faith: the paper reports a clear angle-resolved optical study. The experimental angular maps for different hBN thicknesses visibly differ, and the interference mechanism is standard and credible. The simulations reproduce qualitative trends with fixed oscillator strengths for the main-text thicknesses, which is real support. My concern is not that the phenomenon is absent, but about the precise causal claim. The central assertion that the direction is dictated by thickness requires that thickness be the only relevant variable. Each experimental thickness is a different flake or batch, and the appendix fits oscillator strength per sample; this is a self-acknowledged confound. A fixed-parameter re-simulation across all samples is a tractable check. The reader's weakest assumption pointed to oscillator strength; I agree, and add that the unstated monolayer thickness and the discrete-sample design make the quantitative angle prediction underdetermined. The verdict remains conditional; I would not reject because the qualitative reversal and the fixed-parameter main-text simulations provide credible support.","tokens_in":12103,"tokens_out":5245,"duration_ms":57530,"concrete_test":"Re-run the PyGTM transfer-matrix simulations for all measured hBN thicknesses (10, 45, 78, 95, 98, 115 nm, plus the Fig. 6 series) using only the single oscillator-strength set fitted on the 6 nm sample and a fixed monolayer thickness, with no per-sample optimization. Compare the predicted k∥/k0 position of the maximum reflectivity dip (and the corresponding PL angular maximum, via the calculated emission pattern) against the experimental values. If the predicted angles match within the experimental uncertainty (e.g., ±0.05 k∥/k0) for all thicknesses, the per-sample fitting is not load-bearing; if any thickness requires a different oscillator strength to reproduce the observed angle, the thickness-only conclusion is weakened.","verdict_should_be":"UNCHANGED","load_bearing_attack":"Section II.C states that the two Lorentz oscillator strengths L1 and L2 were optimized for the 6 nm hBN sample and then kept constant 'so as to isolate effects from substrate from oscillator strength.' However, the main experimental comparison uses only discrete flakes (PL: 95 nm vs 45 nm; reflectivity: 6 nm vs 115 nm) with no continuous thickness control, and Appendix Fig. 8 explicitly says that for the additional series 'the oscillator strength in the simulations was optimized to match best the optical response of each sample.' The text also concedes that quantitative differences could arise from 'oscillator strength variations between different molecular monolayers.' Because the monolayer thickness is not stated in Methods and the transfer-matrix model treats the monolayer as a homogeneous uniaxial dielectric layer, the simulation has adjustable parameters beyond hBN thickness. If those parameters are free per sample, the transfer-matrix agreement does not uniquely prove that thickness controls the exact angle of maximum emission; it only shows that the model class is flexible enough to fit each sample. The qualitative reversal (maximum at k∥/k0=0 for 95 nm versus off-center for 45 nm) is well supported by the data, but the stronger claim that the direction is dictated by hBN thickness, including the quantitative angle shifts, is less secure.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies angle-resolved reflectivity and photoluminescence of MePTCDI molecular monolayers on hBN flakes of different thicknesses using back focal plane imaging at 3.5 K. The authors report that the angular distribution of reflectivity and photoluminescence strongly depends on hBN thickness, with the maximum signal occurring at normal incidence for 95 nm hBN and at off-normal angles for 45 nm hBN, and analogous trends for other thicknesses. Transfer matrix simulations reproduce the main qualitative trends and attribute the effect to angle-dependent Fabry-Pérot interference in the hBN substrate. The authors conclude that the substrate thickness controls the direction of light reflection and emission and discuss applications in directional lighting and optoelectronic devices.","tokens_in":12291,"tokens_out":4002,"duration_ms":40791,"significance":"If the quantitative claim holds, the work provides a simple and general mechanism for controlling the emission direction of atomically thin emitters on layered substrates, which is relevant for LEDs, lasers, and angle-resolved spectroscopy. The qualitative observation that the angular distribution flips between on-axis and off-axis maxima with hBN thickness is well supported by the experimental data and by the transfer-matrix interpretation. The paper's strengths are the systematic use of back focal plane imaging, the explicit attempt to separate substrate interference from oscillator-strength effects in the main text, and the open-source transfer-matrix code. The main weakness is that the quantitative thickness-dictated claim is weakened by the per-sample oscillator-strength fitting used in the Appendix, by the unspecified monolayer thickness in the simulations, and by the acknowledged quantitative discrepancies without error analysis.","major_comments":[{"comment":"The simulation methodology is internally inconsistent regarding oscillator strength fitting. The main text states that the two Lorentz oscillator strengths L1 and L2 were optimized for the 6 nm hBN sample and then kept constant 'so as to isolate effects from substrate from oscillator strength.' However, the Appendix states that for the additional thickness series (10, 78, 98 nm), 'the oscillator strength in the simulations was optimized to match best the optical response of each sample.' This means the simulations for the extended thickness series are not independent predictions with fixed material parameters, and the agreement shown in Fig. 8 could stem from per-sample flexibility. To support the central quantitative claim that the direction of maximum reflection and emission is dictated by hBN thickness alone, the authors should either re-run all simulations with the single fixed set of oscillator strengths and show the angular maps, or provide a sensitivity analysis showing how much the angle of maximum reflectivity/PL shifts for plausible variations of L1 and L2. Without this, the quantitative angle predictions are underdetermined.","section":"Section II.C and Appendix Fig. 8"},{"comment":"The monolayer thickness is not stated anywhere in the Methods. Since the transfer-matrix model treats the MePTCDI monolayer as a homogeneous uniaxial dielectric layer, its thickness is a necessary input parameter. If the thickness was taken from a previous work or fitted, this should be stated explicitly, including whether the same value was used for all samples or whether it was varied. As written, the simulated angular maps are not reproducible by an independent group, and the claim that thickness is the only controlling parameter is incomplete without specifying all other geometric inputs.","section":"Section II.C"},{"comment":"The quantitative comparison between experiment and simulation is acknowledged to be imperfect, with the authors listing normalization issues and oscillator-strength variations as possible causes. However, the abstract and conclusion claim that 'the direction of light reflection and emission is dictated by the hBN flake thickness.' The experimental support for this quantitative directionality rests on two photoluminescence thicknesses (95 and 45 nm) and two reflectivity thicknesses (6 and 115 nm) in the main text, and the additional thicknesses in the Appendix are not compared to fixed-parameter simulations. To test the thickness-dictated prediction explicitly, the authors should plot the experimental angle of maximum reflectivity dip and photoluminescence intensity as a function of hBN thickness, overlaid on the simulated curve from Fig. 6(a), including error bars for the experimental angle determination. This would convert the qualitative visual comparison into a quantitative test of the central claim.","section":"Sections III.B, III.C, and Fig. 6(a)"}],"minor_comments":[{"comment":"The caption says '(d) 20◦ and (d) 110◦'; the first should be '(c) 20◦'.","section":"Fig. 3 caption"},{"comment":"The sentence 'we note that here that in contrast to the main text' contains a duplicated 'that'; it should read 'we note that here, in contrast to the main text'.","section":"Appendix Fig. 8 caption"},{"comment":"The sentence 'which changes with temperature can be neglected' is grammatically incomplete; it should read 'whose changes with temperature can be neglected.'","section":"Section IV"},{"comment":"The statement says data 'will be posted in an online repository' but provides no repository name or identifier; please provide a persistent link or state that data are available on request.","section":"Data Availability Statement"},{"comment":"The quantitative comparison of reflectivity dip depths (about 5% vs 15%) is given without any uncertainty estimate; adding error bars from repeated measurements or from sample-to-sample variation would strengthen the discussion of quantitative agreement.","section":"Fig. 5 and related text"}],"recommendation":"major_revision","confidential_remarks":"The paper is within scope for cond-mat.mes-hall and addresses a topic of current interest to the 2D materials and molecular aggregate communities. The qualitative observation is convincing and the transfer-matrix interpretation is plausible. The main technical concern is the internal inconsistency between the fixed-parameter approach in the main text and the per-sample fitting in the Appendix; this is fixable with a re-simulation and a sensitivity analysis, so I recommend major revision rather than rejection. The novelty relative to prior interference studies in 2D materials is incremental but sufficient for a specialized journal."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Short version: this is a solid, honest experimental paper. The qualitative claim—that hBN thickness controls the angular distribution of reflectivity and PL from an ordered molecular monolayer—is well supported by the data and by transfer-matrix simulations. The physics is textbook Fabry-Pérot interference, but the systematic thickness-dependent angle-resolved mapping on this system is new and useful.\n\nWhat it does well: two independently grown batches, AFM thicknesses, careful normalization to the bare hBN after desorbing the molecules, and both reflectivity and PL at multiple thicknesses. The reversal between the 95 nm and 45 nm PL patterns is clear, and the 6 nm vs 115 nm reflectivity comparison is convincing. The simulations reproduce the main trend without needing exotic physics.\n\nThe soft spots are proportionate. The transfer-matrix model fixes two Lorentz oscillator strengths from one 6 nm sample for the main comparison, which is fine, but the appendix then optimizes oscillator strength per sample for the additional thickness series. The text acknowledges oscillator strength variations between monolayers. That means the quantitative angle of maximum emission is not an independent prediction; the model class is flexible enough to fit each sample. A fixed-parameter re-simulation across all thicknesses would settle it. The monolayer thickness is not stated in Methods, and there are no error bars on the angular maps. The generalization to all 2D materials is plausible but not demonstrated.\n\nThe paper is honest about its limitations and cites the relevant prior work on interference at normal incidence. It doesn't oversell the mechanism. The central argument holds up; the issues are about precision of the predictive claim, not about the existence of the effect.\n\nFor whom: anyone doing angle-resolved spectroscopy on monolayer/substrate systems, and people designing experiments with back focal plane imaging. It deserves a serious referee. I'd send it out and ask for the fixed-parameter re-simulation, the monolayer thickness, and error bars. That's a reasonable revision, not a fundamental flaw.","headline":"A careful angle-resolved study showing hBN thickness flips the emission direction of MePTCDI monolayers; the qualitative effect holds, but per-sample oscillator fitting in the appendix keeps the quantitative claim from being fully predictive.","tokens_in":12863,"tokens_out":2036,"would_cite":true,"duration_ms":19989,"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":"The hBN flake thickness determines the angle at which a molecular monolayer reflects and emits light, through Fabry-Pérot interference.","keywords":["angle-resolved spectroscopy","back focal plane imaging","hexagonal boron nitride","MePTCDI monolayer","optical interference","Fabry-Pérot cavity","transfer matrix simulation","directionality of light emission"],"falsifier":"Grow the same molecular monolayer on hBN flakes whose thicknesses span at least one full interference period at the exciton energy and record angle-resolved photoluminescence. If the direction of maximum emission does not sweep through the predicted sequence of angles, or if a monolayer transferred off the hBN onto a non-interfering substrate still shows the same directed pattern, the thickness-interference explanation is wrong.","tokens_in":11892,"feed_emoji":"💡","tokens_out":4580,"duration_ms":44866,"temperature":0.7,"pith_summary":"This paper reports that a monolayer of MePTCDI molecules on a hexagonal boron nitride (hBN) flake reflects and emits light in a strongly directional pattern, and that the direction of that pattern is set by the thickness of the hBN flake. Across flakes from 6 to 150 nm, angle-resolved reflectivity and photoluminescence maps change character: some flakes emit brightest straight up, others emit a ring at large angles. Transfer matrix simulations reproduce the angular maps and trace the effect to angle-dependent Fabry-Pérot interference inside the hBN flake and substrate. The authors argue the result is general: any monolayer on a substrate that supports such interference should show the same substrate-controlled directionality.","feed_headline":"hBN thickness sets the angle of monolayer light emission","feed_subtitle":"Angle-resolved maps show optical interference in the flake steers both reflection and photoluminescence.","key_machinery":"The load-bearing object is the hBN flake as an angle-dependent Fabry-Pérot cavity. For light arriving at an oblique angle, the optical path through the flake changes, shifting the interference condition; this makes the substrate's reflectivity, and therefore the standing-wave field at the monolayer, strongly angle-dependent. The simulations model the flake and the monolayer as uniaxial dielectric layers with in-plane and out-of-plane permittivities, with the monolayer's resonance described by two Lorentz oscillators; this generic model is what lets the authors extend the conclusion beyond MePTCDI.","core_discovery":"At the exciton resonance of the molecular monolayer, both the reflectivity dip and the photoluminescence peak are concentrated at specific emission angles that vary with hBN thickness. For 95 nm hBN the photoluminescence maximum sits at k∥/k0 = 0, while for 45 nm hBN it is a minimum there; reflectivity shows the complementary trend for 6 nm and 115 nm flakes. The paper's central explanation is that the hBN flake acts as a Fabry-Pérot etalon whose resonance condition depends on angle; when the cavity is on resonance, substrate reflection drops and absorption in the monolayer is enhanced, so the monolayer emits preferentially at the matching angle. Transfer matrix simulations with a generic Lorentz oscillator model reproduce the measured angular maps, and the authors generalize the conclusion to any 2D material on an interference-supporting substrate.","pith_inferences":["Editorial inference: the same interference mechanism should imprint angle-dependent patterns on other monolayer signals such as Raman scattering and second-harmonic generation, so thickness scans could be used to isolate those signals from background.","Editorial inference: a dynamically tunable emission direction might be achieved by changing the refractive index of the cavity through temperature, gating, or an added tunable layer, a direction the paper leaves unexplored.","Editorial inference: because the effect depends only on the dielectric stack, the design principle transfers to encapsulated emitters and multi-layer stacks, making thickness choice a general layout rule for directional outcoupling."],"forward_implications":["Substrate thickness can be used as a design knob: choosing a flake thickness pre-selects the angle at which a monolayer emits, which is relevant for lighting and laser geometries.","Experiments that collect light with a limited numerical aperture may miss most or all of the signal at unfavourable hBN thicknesses, so substrate choice must be matched to the measurement geometry.","Angle-averaged spectra are not a safe guide to monolayer properties; the same monolayer can appear bright or almost dark depending only on the flake under it.","Other 2D materials on hBN or similar substrates should exhibit the same thickness-controlled directionality, since the model only needs their dielectric function."],"supporting_citations":[{"why":"Establishes the ordered MePTCDI/hBN system whose collective excitonic resonance is the object of the angle-resolved study.","marker":"[26]"},{"why":"Supplies the generalized transfer-matrix formalism for anisotropic layered heterostructures used in all simulations.","marker":"[41, 42]"},{"why":"Documents the normal-incidence interference effects on optical signals of monolayer materials that the paper extends to oblique angles.","marker":"[4]-[9]"},{"why":"Reports the strong optical response and light emission of MePTCDI monolayers that motivates the choice of the emitter system.","marker":"[25]"},{"why":"Provides angle-resolved studies of dipole emission in layered systems showing the angle-dependent interference effects this work systematizes as a function of thickness.","marker":"[10]-[12]"},{"why":"Accounts for the nonzero epsilon infinity of the organic layer that produces structure in the normalized simulated spectra.","marker":"[43]"}],"fun_headline_variants":["hBN thickness steers monolayer light direction","Optical interference directs monolayer emission","Cavity thickness tunes emission angle in 2D","Thickness of hBN controls light's exit angle","Tuning 2D emission angle via substrate thickness"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The simulations fix the monolayer's optical strength once, using data from the thinnest flake, and apply it to every other sample; if that strength actually varies from flake to flake, the exact angle of brightest emission could shift.","fun_headline_variants_meta":{"raw":{"variants":["hBN thickness steers monolayer light direction","Optical interference directs monolayer emission","Cavity thickness tunes emission angle in 2D","Thickness of hBN controls light's exit angle","Tuning 2D emission angle via substrate thickness"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000162,"raw_usage":{"total_tokens":1236,"prompt_tokens":940,"completion_tokens":296,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":556,"completion_tokens_details":{"reasoning_tokens":225}},"tokens_in":556,"tokens_out":296,"duration_ms":3132,"temperature":1.0,"reasoning_tokens":225,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-15T22:53:30.817737+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Grow the same molecular monolayer on hBN flakes whose thicknesses span at least one full interference period at the exciton energy and record angle-resolved photoluminescence. If the direction of maximum emission does not sweep through the predicted sequence of angles, or if a monolayer transferred off the hBN onto a non-interfering substrate still shows the same directed pattern, the thickness-interference explanation is wrong.","supporting_citations":[{"cited_title":"Juergensen , author M","cited_arxiv_id":null,"evidence_quote":"Establishes the ordered MePTCDI/hBN system whose collective excitonic resonance is the object of the angle-resolved study."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Accounts for the nonzero epsilon infinity of the organic layer that produces structure in the normalized simulated spectra."}],"review_version":1}