{"id":"d3cf5a63-82eb-45f6-bab9-da3343c50e37","arxiv_id":"1908.08863","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"Combined single-nanowire electroluminescence and I-V measurements reveal around 12% of nanowires emit in the working ensemble LED, with per-wire current densities of 20 to 1000 A/cm2.","lead":"Researchers measured light emission and current-voltage behavior of single nanowire LEDs while they were still standing in the as-grown array, using a probe tip inside a scanning electron microscope. Comparing single-wire efficiency peaks with the whole-device peak let them estimate that only about 12% of nanowires actually emit in the processed ensemble.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The 12% wire-density estimate assumes every active nanowire carries exactly 80 nA at the ensemble EQE peak; the measured spread in single-wire I-V characteristics and the possibility of non-emitting conductive paths leave that assumption unverified, so the count could be biased.","rationale":"The paper's central quantitative claim is the 12% active-wire density and the resulting per-wire current densities, obtained by equating the device current density at the ensemble EQE maximum with n_active times the single-wire current at its EQE maximum. The method is clever and the measurements are internally consistent, and the paper does flag the equal-current assumption before applying it. However, the load-bearing step is not proven: the processed device connects all contacted wires in parallel, so at the ensemble peak each wire's current is set by its own I-V curve, not by an externally imposed current command. The reported spread in Rtot and Vth among the three single wires, which are already a selected subset, implies a distribution that will broaden and shift the ensemble EQE curve relative to the single-wire curve; a single representative I_NW cannot be used without knowing the full distribution. Additionally, the terms 'emitting' and 'conducting' are not the same: any contacted but non-emitting wire contributes to the denominator of the ensemble EQE and to J_device but not to the EL numerator, so the ratio J_device/I_NW overestimates the emitting density to the extent such leakage exists. The proposed simulation using the paper's own Shockley parameters and EQE data is a direct test: if the ensemble EQE curve can be matched with f = 12% and negligible non-emitting conduction, the concern is resolved; if not, the headline number needs a correction or an explicit uncertainty. Since the reader's verdict is already CONDITIONAL on exactly this class of unverified assumption, the conditional verdict should stand unchanged rather than being upgraded.","tokens_in":12102,"tokens_out":10513,"duration_ms":109574,"concrete_test":"Digitize the single-NW EQE-versus-current curves of wires C, E, and G from Fig. 6(b) and combine them with the modified Shockley parameters reported for the single wires (I0 = 80 pA, n = 50, RP = 1.5 GOhm, RS = 18 MOhm). Simulate the processed LED as a parallel array of N_total = 5 x 10^9 cm^-2 nanowires sharing one voltage node, with a fraction f of wires emitting (using the measured EQE(I) curves) and the remainder conducting according to the same diode model but contributing zero EL. Compute the ensemble EQE versus total current density for several f and overlay the measured ensemble curve (inset of Fig. 6b), focusing on both the 47 A/cm^2 saturation point and the overall curvature.","verdict_should_be":"UNCHANGED","load_bearing_attack":"Section 3 derives don = J_EQEmax_device / I_EQEmax_NW = 47 A/cm^2 divided by 80 nA, giving about 6 x 10^8 cm^-2, i.e., roughly 12% of the as-grown density. This inversion is valid only if, at the ensemble EQE peak, every nanowire that conducts current also emits, and every emitting wire carries the same current, equal to the 80 nA at which the probe-contacted single wires peak. The paper acknowledges the equal-sharing condition (\"this relation is only true if the NWs have similar contact resistances...\") but does not establish it. Its own Table 1 shows series resistances of 34-58 MOhm and threshold voltages of 7.0-7.8 V for the three single wires, so under a common ITO top contact each wire will carry a different current at a given device voltage. The ensemble EQE is then a current-weighted average over shifted single-wire EQE(I) curves, not a simple replication of one wire at 80 nA. In addition, J_device includes current through nanowires that are contacted but do not emit (leaky or poorly injected wires). With the fitted parallel resistance RP of about 1.5 GOhm per wire, even a few nA of leakage through a large fraction of the 5 x 10^9 cm^-2 wires would add tens of A/cm^2 to J_device, making the inferred density an overestimate. The caveat about homogeneous current spreading flags the issue, but a caveat does not supply evidence that the processed ITO contact actually enforces it.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The authors use a tungsten probe tip inside an SEM to contact individual (In,Ga)N/GaN nanowire LEDs in an unprocessed as-grown ensemble and simultaneously measure their EL spectra and I-V characteristics. They observe several emission lines per single-wire spectrum, a linear increase of the spectral FWHM with decreasing emission energy, and diode-like I-V curves that are fitted with a modified Shockley equation, including very high ideality factors attributed to the tungsten/p-GaN contact. The main quantitative result is obtained by comparing the current at which the relative EQE peaks for single nanowires (about 80 nA) with the current density at which the processed ensemble device EQE peaks (47 A/cm2), yielding an active emitting nanowire density of 6e8 cm-2, about 12% of the as-grown density. From this value the authors estimate per-wire current densities of 20-1000 A/cm2 in the working device.","tokens_in":12463,"tokens_out":7082,"duration_ms":72582,"significance":"If the central estimate is correct, the finding that only about one in eight nanowires emits in a processed ensemble LED is important and practically relevant, and the ratio method J_device/I_NW is an elegant, parameter-free way to count active wires. The paper also introduces a useful measurement capability: simultaneous EL and I-V on single nanowires within their native ensemble, avoiding the artifacts of dispersion-based approaches. The authors are transparent about the main assumptions, and the independent cross-check via the ensemble series resistance adds credibility. The significance is tempered, however, by the fact that the headline 12% number depends on two unquantified assumptions: equal current sharing among active wires and transferability of the single-wire EQE-peak current from a tungsten-probe contact to an ITO contact.","major_comments":[{"comment":"The central estimate d_on = 6e8 cm-2 assumes that, at the ensemble EQE peak, every current-carrying nanowire emits and carries exactly I_EQEmax_NW = 80 nA. The paper acknowledges the equal-current condition, but the data in Table 1 show a spread in single-wire series resistances (34-58 MOhm) and threshold voltages (7.0-7.8 V); under a common top contact such variations produce unequal current sharing, so the ensemble EQE is a current-weighted average over shifted single-wire EQE(I) curves rather than N copies of one wire at 80 nA. In addition, the fitted parallel resistance RP ~ 1.5 GOhm per wire implies that even a few nA of leakage through a large fraction of the 5e9 cm-2 wires adds tens of A/cm2 to J_device without contributing to EL, biasing d_on upward. A quantitative sensitivity analysis, such as propagating the measured Rtot and Vth distributions or bounding the leakage contribution, is needed to support the abstract's claim of a 'quite accurate' determination.","section":"Sec. 3, Fig. 6 and the relation J_EQEmax_device = d_on I_EQEmax_NW"},{"comment":"The single-wire value I_EQEmax_NW is measured through a tungsten probe that produces ideality factors around 50 and series resistances of tens of MOhm, whereas wires in the processed device are contacted by ITO. The text assumes that a high contact resistance has negligible influence on the integrated EL and relative EQE, but this is not demonstrated and is not obviously guaranteed: a contact-dominated junction can alter the injection geometry, local current density, and heating in the nanowire, and the ITO contact injects over the full nanowire tip. Because d_on is inversely proportional to I_EQEmax_NW, a contact-induced shift of the single-wire EQE peak by even a factor of two changes the headline result. Please provide a direct test, for example measuring the same nanowire under different probe pressures or contact areas, or cross-checking against confocal EL maps of the processed device as in Ref. [7].","section":"Sec. 3, Fig. 6(b) and the assumption stated after the definition of relative EQE"},{"comment":"The assignment of spectra C, E, and G to single nanowires is indirect: the SEM cannot resolve individual wires at a density of 5e9 cm-2, and the single-wire assignment relies on spectral shape and on one supporting example in the supplementary information. If any of these spectra actually contains two or more nanowires, the inferred I_EQEmax_NW is too large and d_on is correspondingly underestimated; the contrast with point F, which is explicitly a bundle and peaks at 300 nA, demonstrates the sensitivity. The paper should state how many of the more than 20 positions were classified as single-wire and provide an estimate of the misclassification rate, or corroborate the assignment with a spatially resolving measurement.","section":"Sec. 3, Fig. 2 and the paragraph beginning 'Within a set of more than 20 measurement positions'"}],"minor_comments":[{"comment":"The phrase 'we sweeped the voltage' should be 'we swept the voltage'.","section":"Sec. 2"},{"comment":"Reference [22] is cited as '[22, ?]' in the text, and the reference list contains a stray '?' and malformed URL prefixes; please clean up the bibliography.","section":"References"},{"comment":"The table caption contains a typo ('T able 1. . Series resistance'); the column headers could also include the units (Ohm, V) to avoid repetition.","section":"Table 1"},{"comment":"The FWHM trend is based on ten selected spectra without error bars or a statement of the fitting procedure; please add uncertainty information or soften the linear-trend wording.","section":"Sec. 3, Fig. 4(b)"},{"comment":"The text says the ensemble relative EQE 'saturates' at 47 A/cm2, while the single-wire EQE is described as having a maximum; please use consistent terminology because the ensemble value is used as the EQE-peak current density.","section":"Sec. 3, inset of Fig. 6(b)"}],"recommendation":"major_revision","confidential_remarks":"The manuscript is within the journal's applied-physics scope and presents a useful experimental method. The main quantitative claim is interesting but the uncertainty budget for the 12% active-wire estimate is not provided; the equal-current-sharing and contact-transferability assumptions are acknowledged but not substantiated. I recommend major revision to address these points, after which the paper could be acceptable."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"This paper's real contribution is the simultaneous EL and I-V measurement of single as-grown nanowire LEDs in an unprocessed ensemble, plus the idea of comparing single-wire and ensemble EQE-peak currents to back out the active wire density. That combination is new as far as the cited literature goes, and the execution looks careful: the probe-contact procedure is described in detail, the modified Shockley fit with ideality factor around 50 is physically attributed to the tungsten/p-GaN contact rather than to the wire itself, and the FWHM-versus-energy trend is consistent with earlier work. The authors also deserve credit for being explicit about the assumptions in the density estimate, even if those assumptions are not verified.\n\nThe central number, 12% active wires, is the main event. The derivation is simple: ensemble EQE-peak current density divided by single-wire EQE-peak current. But that ratio only yields a true count if every conducting wire emits and every emitting wire carries the same current at the ensemble peak. The paper acknowledges the equal-sharing condition and cites ITO contact homogeneity, but their own Table 1 shows series resistances from 34 to 58 MOhm across three single wires. Under a common ITO contact, those wires will not draw equal currents at a given voltage, so the ensemble EQE is a current-weighted average over shifted single-wire curves, not a replication of one wire at 80 nA. The stress-test note is fair: leaky non-emitting wires would inflate J_device and bias the density upward. Their parallel-resistance fit of 1.5 GOhm per wire means even a few nA of leakage across a large fraction of the 5e9 cm^-2 wires could add tens of A/cm^2. The caveat is stated, but a caveat does not validate the inversion.\n\nOther soft spots are minor. The single-wire identification for spectrum C is indirect—based on spectral similarity and probe-movement behavior—though the argument is reasonable. The FWHM trend uses ten spectra without error bars, which is fine for a qualitative observation. The phrase \"quite accurate\" overstates what the method can deliver; this is an order-of-magnitude estimate with probably a factor-of-two uncertainty.\n\nThe citation pattern is clean: they cite the relevant single-wire I-V, confocal EL, and their own prior ensemble work where appropriate. No circularity problem; the two quantities in the ratio are measured independently.\n\nWho is this for? Anyone working on nanowire LED processing or device simulation who needs a per-wire current density estimate. It is a useful tool paper, not a field-reshaping result.\n\nRecommendation: send it to peer review. The measurement method is sound and the manuscript is honest about its limitations; a good referee should push for a quantitative sensitivity analysis of the active-wire fraction against the contact-resistance spread and leakage current.","headline":"A careful single-wire EL/I-V study that delivers a plausible but assumption-laden estimate of active nanowire density; worth refereeing with revisions, not a home run.","tokens_in":12996,"tokens_out":1332,"would_cite":true,"duration_ms":15601,"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":"Single-wire electroluminescence probing reveals how many nanowires actually emit in an ensemble LED.","keywords":["nanowire LED","electroluminescence","current-voltage characteristics","external quantum efficiency","(In,Ga)N/GaN","scanning electron microscopy","single nanowire","ensemble device"],"falsifier":"Measure the EQE-versus-current curve of the same individual nanowires before and after depositing the ITO top contact, or count emitting spots in the working device with confocal microscopy: if the single-wire EQE-peak current shifts or the counted active-wire density differs from $6\\times10^8$ cm$^{-2}$, the central relation fails.","tokens_in":11954,"feed_emoji":"💡","tokens_out":8690,"duration_ms":76005,"temperature":0.7,"pith_summary":"This paper tries to establish a way to count how many nanowires actually emit inside a working (In,Ga)N/GaN nanowire LED by comparing the efficiency-versus-current behavior of single nanowires with that of the whole device. A probe tip in a scanning electron microscope contacts individual as-grown nanowires and records their electroluminescence and current-voltage curves simultaneously. The authors find that the relative external quantum efficiency of single wires peaks near 80 nA, while the processed ensemble peaks at a device current density of 47 A/cm$^2$; their ratio yields an emitting-wire density of $6\\times10^8$ cm$^{-2}$, about 12% of the as-grown density of $5\\times10^9$ cm$^{-2}$. If correct, this gives the actual current density per active wire, 20 to 1000 A/cm$^2$, a quantity needed for meaningful device comparison and simulation.","feed_headline":"Single-wire probe shows only 12% of nanowires emit in LED","feed_subtitle":"Efficiency-peak comparison between one wire and the whole device gives real current per emitting wire.","key_machinery":"The load-bearing identity is $J_{\\mathrm{EQE}_{\\max}}^{\\mathrm{device}} = d_{\\mathrm{on}} I_{\\mathrm{EQE}_{\\max}}^{\\mathrm{NW}}$, which equates the ensemble device current density at its relative-EQE maximum with the per-wire current at the single-wire relative-EQE maximum times the number density of emitting nanowires. The relative EQE itself is defined as integrated electroluminescence divided by driving current. Fitting single-wire I-V curves with the modified Shockley equation and using the EQE-maximum current as a common fingerprint lets the authors transfer a single-wire measurement to the ensemble device.","core_discovery":"The central claim is that the relative external quantum efficiency of a single nanowire LED peaks at a well-defined current, about 80 nA for wires in this ensemble, and that the same peak appears in the processed ensemble device at a device current density of 47 A/cm$^2$. Dividing the latter by the former gives the number density of emitting nanowires, $d_{\\mathrm{on}} = 6\\times10^8$ cm$^{-2}$, only about 12% of the as-grown density $5\\times10^9$ cm$^{-2}$. Active wires in the working device therefore carry current densities of 20 to 1000 A/cm$^2$ for device current densities of 0.9 to 47 A/cm$^2$. The paper also reports that single-wire EL spectra consist of several quantum-well emission lines whose widths grow as peak energy decreases, and that single-wire I-V curves follow the modified Shockley equation with high ideality factors caused by the probe contact.","pith_inferences":["Beyond the paper: the same EQE-peak ratio could be used as a contact-free process monitor, counting active emitters in any working LED from one representative single-wire calibration.","Beyond the paper: if the tungsten-probe contact changes the single-wire EQE-peak current, the 12% figure would shift; measuring the same wires through the ITO contact after processing would settle that calibration.","Beyond the paper: the inferred spread of 20 to 1000 A/cm$^2$ across active wires implies strong current crowding, which could be tested by comparing degradation patterns with the estimated per-wire current density.","Beyond the paper: the approach could be applied to other self-assembled emitter ensembles, and the per-wire current densities could feed efficiency-droop models that normally rely on device-averaged current density."],"forward_implications":["The ensemble EQE maximum shifts in proportion to the density of active wires, so a lower active fraction moves the device EQE peak to lower current densities.","In this sample only about one in eight nanowires emits, meaning the processed LED is far from using its full as-grown wire density and processing optimization has a large headroom.","Active nanowires in the working device experience current densities of 20 to 1000 A/cm$^2$, one to two orders of magnitude above the naive average over all as-grown wires.","The ensemble EL spectrum is a superposition of strongly varying single-wire spectra, and single-wire linewidths increase linearly as peak emission energy decreases.","Reverse leakage observed in the ensemble is also present in single as-grown wires, so it is intrinsic to the nanowire structure rather than a processing artifact."],"supporting_citations":[{"why":"supplies the processed ensemble LED's EQE-versus-current-density curve, series resistance, ideality factor, and the earlier rough active-wire estimate used for comparison.","marker":"[12]"},{"why":"provides top-view EL mapping and modeling that attributes the single-wire emission lines to the four quantum wells and supports identifying single contacted wires.","marker":"[7]"},{"why":"demonstrates probe-based I-V characterization of single GaN nanowire LEDs in an ensemble, the electrical method extended here with simultaneous EL detection.","marker":"[6]"},{"why":"established diffraction-limited EL spot measurements in working ensemble devices without per-wire currents, motivating the need to correlate EL with current.","marker":"[4]"},{"why":"reports processing-related variations in contacting nanowires and supports the conclusion that the low active fraction stems from processing.","marker":"[5]"}],"fun_headline_variants":["Single-wire LEDs reveal true emitting density: only 12%","Nanowire LED probe: 88% of wires stay dark","One-wire EQE peak unlocks real current per nanowire","How many nanowires really light up? Single-wire answer: 12%","Per-wire efficiency exposes active fraction of nanowire LED"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The count assumes that the current at which a single nanowire's relative EQE peaks is the same under a tungsten probe contact as for nanowires in the processed device with an ITO top contact, and that current spreads evenly across all active nanowires.","fun_headline_variants_meta":{"raw":{"variants":["Single-wire LEDs reveal true emitting density: only 12%","Nanowire LED probe: 88% of wires stay dark","One-wire EQE peak unlocks real current per nanowire","How many nanowires really light up? Single-wire answer: 12%","Per-wire efficiency exposes active fraction of nanowire LED"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000652,"raw_usage":{"total_tokens":3014,"prompt_tokens":992,"completion_tokens":2022,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":608,"completion_tokens_details":{"reasoning_tokens":1931}},"tokens_in":608,"tokens_out":2022,"duration_ms":13250,"temperature":1.0,"reasoning_tokens":1931,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T11:27:13.985357+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Measure the EQE-versus-current curve of the same individual nanowires before and after depositing the ITO top contact, or count emitting spots in the working device with confocal microscopy: if the single-wire EQE-peak current shifts or the counted active-wire density differs from $6\\times10^8$ cm$^{-2}$, the central relation fails.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"supplies the processed ensemble LED's EQE-versus-current-density curve, series resistance, ideality factor, and the earlier rough active-wire estimate used for comparison."},{"cited_title":"The effect of the three-dimensional strain variation on the emission properties of light-emitting diodes based on (In,Ga)N/GaN nanowires","cited_arxiv_id":"1704.01569","evidence_quote":"provides top-view EL mapping and modeling that attributes the single-wire emission lines to the four quantum wells and supports identifying single contacted wires."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"demonstrates probe-based I-V characterization of single GaN nanowire LEDs in an ensemble, the electrical method extended here with simultaneous EL detection."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"established diffraction-limited EL spot measurements in working ensemble devices without per-wire currents, motivating the need to correlate EL with current."},{"cited_title":"Current path in light emitting diodes based on nanowire ensembles","cited_arxiv_id":"1210.7144","evidence_quote":"reports processing-related variations in contacting nanowires and supports the conclusion that the low active fraction stems from processing."}],"review_version":1}