{"id":"afacc0cb-ea5a-4dc3-94a5-f5312cd78cb0","arxiv_id":"1908.05206","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":5,"one_line_summary":"In telecom-band InGaAs/GaAs quantum dots, the s-p splitting decreases as emission energy increases, and 8-band k·p simulations attribute this inverse trend to indium content rather than dot size.","lead":"Researchers measured the energy gap between the ground and first excited states in single indium-gallium-arsenide quantum dots that emit light at telecommunication wavelengths. They found that dots emitting at shorter wavelengths have smaller gaps, opposite to the usual size-driven expectation, and their simulations trace the effect to indium content rather than dot size.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Attribution to indium content is undermined because the model's height-variation curve reproduces the same inverse trend, and its dismissal relies on unverified growth assumptions.","rationale":"The reader's weakest assumption identifies the model-dependence of the structural attribution. My concern sharpens this: the model itself admits a specific alternative channel (height/aspect-ratio variation) that yields the same inverse trend, and the authors' rejection of that channel rests on general growth expectations rather than sample-specific structural measurements. This is a real soft spot, but it does not overturn the paper's value: the experiment is careful, the model is state-of-the-art, and the composition interpretation is plausible. It does, however, strengthen the need for the conditional elements already requested in the reader's verdict: error bars, structural characterization, and code/data availability. I therefore leave the verdict unchanged.","tokens_in":9121,"tokens_out":6221,"duration_ms":61681,"concrete_test":"Perform cross-sectional scanning tunneling microscopy or atom probe tomography on a set of the same MOCVD-grown QDs spanning the measured emission range, extracting per-dot height, base diameter, and indium composition. Feed the measured joint distribution into the same 8-band k·p model and compare the predicted s-p splitting versus emission energy relation with the PLE data. If the model reproduces the experimental trend only when In content is varied while height remains nearly constant, the attribution holds; if the measured height variation alone reproduces the trend, the central claim that indium content dominates is not supported.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim is that composition variation, not size, causes the inverse s-p splitting versus emission energy trend. The authors' own 8-band k·p calculation shows that varying QD height at constant base diameter also produces an inverse trend (grey dashed curve in Fig. 3(b)). They dismiss this because (i) aspect ratio 'does not usually change significantly' (Refs. 13, 30, 31) and (ii) explaining the experimental range would require heights exceeding 10 nm, near the plastic relaxation limit. Neither argument is supported by structural data for this specific MOCVD sample: the reported height range is 6-8 nm, but no per-dot height distribution or correlation with emission energy is provided. Since the height-only curve already matches the observed sign of the trend, the conclusion that the trend is 'predominantly' due to indium content depends on an untested assumption about the ensemble's height/base-diameter covariation. The paper itself concedes that 'some changes in the QDs' height cannot be ruled out', which is precisely the ambiguity that prevents the causal attribution from being secure without direct structural input.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript reports photoluminescence excitation (PLE) spectroscopy of single MOCVD-grown InGaAs/GaAs quantum dots with an InGaAs strain-reducing layer emitting in the telecommunication O band. The authors observe a PLE resonance above each emission line and attribute it to p-shell absorption, supported by an ensemble PL energy difference of about 70 meV and a forward 8-band k·p calculation with realistic parameters. Extracting the s-p splitting for many single dots, they find an inverse dependence on emission energy, from about 80 meV down to 60 meV. Comparing with 8-band k·p simulations for varying QD size, average composition, SRL composition, and height, they conclude that the inverse trend is predominantly caused by variation of indium content within individual dots, and they frame this as a route for engineering and selecting telecom-wavelength QDs.","tokens_in":9320,"tokens_out":5702,"duration_ms":53514,"significance":"If the causal attribution is correct, the result provides a practically useful design rule: for InGaAs/GaAs QDs with strain-reducing layers, the ground-state emission energy and the s-p splitting can be mutually tuned by controlling indium content, which is relevant for single-photon sources pumped via p-shell quasi-resonant excitation. The measured PLE dataset is a valuable systematic single-QD study in a wavelength range where such data are scarce. The work combines careful PLE mapping with a moderately sophisticated 8-band k·p model that includes strain, piezoelectric effects, and an indium composition gradient; the forward calculation reproduces the absolute ground-state energy and s-p splitting for one representative dot, which supports the p-shell assignment. However, as detailed in the major comments, the central causal attribution is not uniquely supported by the evidence presented, and the experimental trend lacks quantified uncertainties.","major_comments":[{"comment":"The height-variation curve (grey dashed line in Fig. 3(b)) reproduces the same inverse s-p splitting versus emission energy trend as the experimental data, as the authors acknowledge. The dismissal of this scenario rests on two arguments: (i) the general expectation that the aspect ratio of self-assembled QDs does not change significantly (Refs. 13, 30, 31), and (ii) the claim that covering the experimental emission-energy range would require QD heights exceeding 10 nm, near the plastic relaxation limit. Neither argument is supported by structural data for this specific MOCVD sample: the reported height range of 6-8 nm is an ensemble average, and no per-dot height distribution or height-emission correlation is provided. Since the height-only curve already matches the sign of the observed trend, the conclusion that the trend is 'predominantly' caused by indium content depends on an untested assumption about the height/base-diameter covariation within this ensemble. The manuscript itself concedes that 'some changes in the QDs' height cannot be ruled out from our considerations, and these perhaps contribute to the obtained inverse s-p splitting vs emission energy dependence' (last paragraph of the modelling section). To make the central claim load-bearing, the authors should either provide structural evidence (e.g., cross-sectional TEM/STEM or atom-probe tomography on identical samples) that the height/base-diameter ratio is narrowly distributed, or quantitatively demonstrate that the combined size-composition model, rather than the height model, is required to explain the observed magnitude and slope of the s-p splitting versus emission energy.","section":"Confined states calculations (pp. 6-7), Fig. 3(b)"},{"comment":"The experimental s-p splitting values are presented without any error bars or peak-position uncertainties. The PLE resonances have a linewidth of about 2 meV (p. 5), while the claimed trend spans from 80 to 60 meV; without per-dot uncertainties, the statistical significance of the inverse dependence cannot be assessed. The authors should report the uncertainty of each PLE peak position (e.g., from Lorentzian fits) and, ideally, show a linear regression of the data with confidence intervals to substantiate the 'clear dependence' stated in the text.","section":"Fig. 3(a)"},{"comment":"The comparison between experiment and model is made visually via slopes in Fig. 3(b), and the authors state that 'the absolute energy values from the simulations do not correspond to the experimental s-p splitting values precisely, but support (or do not) the observed trends.' Because the central inference—composition over size or height—rests on which model curve best reproduces the experimental trend, a quantitative comparison is needed. For example, the authors could fit the experimental s-p splitting versus emission energy with a line and compare its slope with the slopes of the composition, height, and size curves, including propagated uncertainties. Without such a comparison, 'closer match' remains a qualitative judgement, and the competing hypothesis of height variation cannot be excluded at any stated confidence level.","section":"Confined states calculations (p. 7)"}],"minor_comments":[{"comment":"There are numerous typographical errors, including 'with in' in the abstract, 'T he' at the start of a sentence, and inconsistent spacing in 's -shell' / 'p -shell'. A thorough proofread is needed.","section":"Abstract and throughout"},{"comment":"The energy differences between QD states in the ensemble PL spectrum are described as 'estimated'; please specify the fitting procedure (e.g., Gaussian fits) and the associated uncertainties.","section":"p. 4, Fig. 1(a)"},{"comment":"The statement that the accuracy of experimentally determined In content in Ref. 27 was ±0.13 suggests substantial uncertainty in composition; please indicate how this uncertainty propagates into the simulated s-p splitting and whether it affects the comparison with experiment.","section":"p. 7, indium content discussion"},{"comment":"The sentence 'The PLE maxima are rather broad (~2 meV), which may be related to the relatively large excitation laser linewidth (~0.7 meV)' is unclear because a 0.7 meV laser linewidth would not by itself produce 2 meV peaks; please clarify the broadening mechanism or revise the phrasing.","section":"p. 5, PLE linewidth"},{"comment":"The paper relies heavily on the authors' previous work (Refs. 20, 25, 27, 29) for key structural and material parameters; a clear statement of which parameters are taken from which source, and any independent validation, would improve transparency.","section":"References"}],"recommendation":"major_revision","confidential_remarks":"The manuscript is within the journal's scope, and the experimental dataset in Fig. 3(a) is interesting. However, the central causal claim—that the inverse s-p splitting trend is dominated by indium content—is not uniquely supported because the model's height-variation curve produces the same trend, and the manuscript itself acknowledges this ambiguity. The absence of error bars on the experimental s-p splitting values further weakens the strength of the claim. I recommend major revision rather than rejection because the issue is addressable in principle with additional analysis or structural data, but as it stands the central attribution is not secure."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Plainly: this paper gives the QD community something it did not have—a single-dot PLE map of s-p splitting versus emission energy for InGaAs/GaAs dots with a strain-reducing layer in the O band, with a clear inverse trend. That dataset is worth having, and the k·p calculations are competent and transparent about being a parameter study rather than a fit. Credit is due for the ensemble corroboration and for flagging the Coulomb correction as small.\n\nThe soft spot is the structural attribution. The model shows that simultaneous scaling of the dot (size multiplier) pushes the splitting the wrong way, and that varying average In content produces the right trend. But their own grey dashed curve—height only at constant base—also produces an inverse trend. They dismiss it with two arguments: aspect ratio is \"not usually\" varying, and the height range needed (>10 nm) would be near plastic relaxation. Neither is a measurement on this sample. The paper itself concedes \"some changes in height cannot be ruled out.\" That concession lands on the headline claim: saying the dependence is \"predominantly\" due to indium content is stronger than the evidence supports. The trend itself is real; the cause is not uniquely identified. The stress-test note is right, and I don't think the authors' rebuttal closes it.\n\nAlso minor but real: the s-p splitting values in Fig. 3(a) have no error bars, despite the PLE peak widths (~2 meV) making determination error estimable. No code or data are deposited, so the model curves are hard to reproduce. Self-citation is not a problem here; the cited prior work is theirs but the new result is distinct.\n\nWho is this for? People engineering telecom single-photon sources who need p-shell pumping; they will get a useful empirical relation and a plausible design knob (In content). A cautious reader should treat the causal attribution as provisional.\n\nRecommendation: send to a serious referee. The experiment is on solid ground, the modeling is sophisticated, and the paper is honest about limitations. I would accept it with major revisions requiring either error bars and statistical treatment, or a softened claim—ideally both.","headline":"New single-dot PLE data show a real inverse s-p splitting trend in O-band InGaAs dots, but the claim that indium content is the dominant cause is not fully secured against the height-variation alternative.","tokens_in":9957,"tokens_out":2237,"would_cite":true,"duration_ms":22974,"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":"The s-p splitting in O-band quantum dots is set by indium content, not dot size.","keywords":["quantum dots","InGaAs/GaAs","s-p splitting","photoluminescence excitation spectroscopy","8-band k·p model","strain-reducing layer","O-band emission","indium composition"],"falsifier":"Measure individual dots' indium content and size directly (e.g., by atom-probe tomography or scanning transmission electron microscopy) on the same sample and correlate them with each dot's emission energy and s-p splitting: if higher-energy dots do not have lower In content, or if size differences alone reproduce the splitting trend, the attribution fails. Alternatively, if two dots emit at the same energy but have different measured sizes and show the same s-p splitting, composition would be confirmed as the cause.","tokens_in":8935,"feed_emoji":"🔬","tokens_out":6172,"duration_ms":56042,"temperature":0.7,"pith_summary":"This paper uses single-dot photoluminescence excitation spectroscopy to measure the energy gap between the s-shell and p-shell states in MOCVD-grown InGaAs/GaAs quantum dots that emit in the telecom O band around 1.3 μm. It finds that this s-p splitting falls from about 80 meV to about 60 meV as the emission energy increases. Because smaller dots would normally raise both the emission energy and the splitting, the observed inverse trend points away from size as the controlling factor. Comparing the data with 8-band k·p calculations, the paper attributes the trend mainly to variations in indium content between dots, with strain-reducing-layer composition and dot height playing secondary roles. This matters because p-shell quasi-resonant excitation is a route to clean single-photon sources, so knowing what sets the s-p splitting lets growers engineer and select dots suited to telecom applications.","feed_headline":"Quantum dot s-p splitting follows indium content, not size","feed_subtitle":"Single-dot PLE and k·p modelling explain the 80-to-60 meV trend, guiding telecom photon-source design.","key_machinery":"The load-bearing machinery is the pair of curves in Fig. 3(b) produced by 8-band k·p calculations: a red curve for simultaneous rescaling of all dot dimensions, a green curve for varying average indium content in the dot, a curve for varying strain-reducing-layer composition, and a dashed curve for varying dot height, each plotted as s-p splitting versus ground-state energy. The model uses continuous-elasticity strain, second-order deformation potentials, an indium gradient concentrated at the dot centre, and configuration-interaction excitonic states; its role is to show which structural parameter yields the same slope as the measured single-dot data. Since the composition curve matches the inverse trend and the size curve has the opposite sign, the machinery converts the measured trend into a causal attribution.","core_discovery":"On the paper's own terms, the central discovery is that in MOCVD-grown InGaAs/GaAs quantum dots capped with an InGaAs strain-reducing layer and emitting above 1.3 μm, the energy separation between the s-shell and p-shell states, measured on many single dots by photoluminescence excitation spectroscopy, decreases from about 80 meV to about 60 meV as the ground-state emission energy increases. This inverse dependence is opposite to what would be expected if the dots differed only in size, because smaller dots should show both higher emission energy and stronger confinement, hence larger s-p splitting. 8-band k·p calculations that vary the dot size, the average indium content in the dot, the indium content in the strain-reducing layer, and the dot height show that only the composition variation reproduces the experimentally observed slope; the strain-reducing layer alone gives too weak an effect, and height variations would require implausibly broad changes. The paper therefore attributes the trend predominantly to differences in indium content between individual dots, with Coulomb interactions contributing only a few meV, and concludes that emission wavelength and s-p splitting can be co-engineered by controlling composition.","pith_inferences":["A testable extension would be to measure individual-dot composition directly (e.g., by atom-probe tomography) on the same samples to confirm that higher-energy emitters are indeed indium-poorer; the paper's attribution rests on model slope matching.","If the trend is compositional, the same inverse relation should appear in other O-band dots grown by different methods, and its absence would point to size effects; this could be checked with existing MBE-grown samples with and without SRLs.","The anticorrelation between emission energy and s-p splitting also implies that wavelength-selective filtering of an ensemble can preselect dots with a desired excited-state ladder, which may simplify integration into photonic cavities.","One could use magnetic-field or pressure tuning of single dots to vary confinement and composition independently, providing a cleaner separation of the two effects than the ensemble trend."],"forward_implications":["If composition is the dominant lever, the p-shell resonance energy for quasi-resonant pumping can be read off from the emission wavelength of a chosen dot.","Engineering the average indium content (e.g., by overgrowth or thermal interdiffusion) should shift emission wavelength and s-p splitting together, giving ensemble-level control.","Dots emitting at the same wavelength can still differ in s-p splitting when size and composition are traded off, so selection from a broad ensemble can optimize both.","The large ~80 meV splitting supports thermal stability at cryocooler temperatures without additional engineering.","Since strain-reducing-layer composition alone moves the splitting only weakly, the SRL is a secondary tuning knob."],"supporting_citations":[{"why":"Provides the ensemble-level reference: QD structural parameters (30 nm base, 6–8 nm height) and the ~70 meV s-p splitting expected for these dots.","marker":"[20]"},{"why":"Supplies the photoluminescence excitation spectroscopy setup adapted to the >1 μm range, the experimental method that yields single-dot s-p splittings.","marker":"[21]"},{"why":"Gives the continuous-elasticity strain calculation used in the model to evaluate the strain field around the dots.","marker":"[24]"},{"why":"Provides the improved second-order deformation potentials used in the strain and band-structure calculations.","marker":"[25]"},{"why":"Supplies the indium-gradient concentration profile inside the QD and the structural data on which the modelled composition distribution is based.","marker":"[27]"},{"why":"Defines the 8-band k·p model used to compute single-particle electron and hole states.","marker":"[28]"},{"why":"Details the confined-state calculations and material parameters, including the excitonic configuration-interaction treatment from which s-p splitting is extracted.","marker":"[29]"},{"why":"Supports the assumption that III-V self-assembled QD size changes scale all dimensions together, which underpins the size-scaling curve.","marker":"[13]"}],"fun_headline_variants":["Indium content sets s-p splitting in 1.3-μm QDs","Quantum dot s-p gap shrinks as emission energy rises","Composition, not size, drives quantum dot s-p splitting","S-p splitting in InGaAs QDs reveals indium control"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The whole attribution rests on the assumption that the 8-band k·p model with the assumed indium-gradient profile and deformation potentials predicts the correct relative slopes of s-p splitting versus emission energy for composition and size variations, since no direct per-dot composition or size measurements are made.","fun_headline_variants_meta":{"raw":{"variants":["Indium content sets s-p splitting in 1.3-μm QDs","Quantum dot s-p gap shrinks as emission energy rises","Composition, not size, drives quantum dot s-p splitting","S-p splitting in InGaAs QDs reveals indium control"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000315,"raw_usage":{"total_tokens":1783,"prompt_tokens":938,"completion_tokens":845,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":554,"completion_tokens_details":{"reasoning_tokens":770}},"tokens_in":554,"tokens_out":845,"duration_ms":7897,"temperature":1.0,"reasoning_tokens":770,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T13:20:29.065137+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Measure individual dots' indium content and size directly (e.g., by atom-probe tomography or scanning transmission electron microscopy) on the same sample and correlate them with each dot's emission energy and s-p splitting: if higher-energy dots do not have lower In content, or if size differences alone reproduce the splitting trend, the attribution fails. Alternatively, if two dots emit at the same energy but have different measured sizes and show the same s-p splitting, composition would be confirmed as the cause.","supporting_citations":[],"review_version":1}