{"id":"e318315f-cee8-46ae-b0e2-1a80302ac010","arxiv_id":"2501.00455","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":1,"one_line_summary":"A single porous GaN layer in green micro-LEDs boosts measured EL intensity up to 22x and creates narrow resonant peaks in square and hexagonal mesas.","lead":"Researchers made tiny green LEDs with a porous layer inside and measured up to 22 times brighter electroluminescence than identical-looking LEDs without it. The gain depends on the LED's shape, with square and hexagonal devices showing sharp resonant peaks instead of a broad glow.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Central mechanism rests on unverified porous-layer formation; no cross-section or PL-before/after data shown.","rationale":"The reader's weakest assumption identified the lack of cross-section microscopy and before/after PL as a key uncertainty, specifically questioning whether the 100 nm u-GaN spacer protects the MQWs and whether the porous layer is uniform. My analysis reaches the same structural concern: the central mechanism hinges on the presence and integrity of the porous layer, and the paper supplies no direct structural or spectroscopic verification of it. This is more load-bearing than the statistical weakness of the 22× comparison, because even a perfectly reproducible intensity enhancement would not validate the proposed photon-confinement mechanism if the porous layer is not actually formed as intended. The reader's verdict of CONDITIONAL remains appropriate: the paper should be accepted only if the authors provide cross-sectional structure data and PL before/after etching. Since my concern coincides with the reader's, no change to the verdict is needed.","tokens_in":8462,"tokens_out":5048,"duration_ms":49921,"concrete_test":"Perform cross-sectional SEM/TEM on the hexagonal porous Micro-LED and on a normal device from the same sample to confirm the presence, continuity, thickness, and porosity of the porous layer beneath the active region. Additionally, measure micro-PL spectra from the MQW region on the same sample before and after the electrochemical etching step. If the porous layer is absent or non-uniform, or if the MQW PL intensity/linewidth is degraded, the claimed refractive-index confinement mechanism is unsupported.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The paper attributes the 22× intensity enhancement and the resonant emission to a single porous GaN layer located under the active region. However, the only structural evidence presented is the optical microscope images in Figure 2; no SEM/TEM cross-section, porosity measurement, or thickness profile of the porous layer is provided. The method section asserts that a 100 nm u-GaN spacer 'prevents the active region from being damaged during the etching process', but this protective role is not verified by any before/after PL or TEM data. If the electrochemical etch only partially converts the highly doped n-GaN layer, or produces a non-uniform porous region, the effective refractive index and current paths would differ from the assumed model. If the etch damages the MQWs, the observed spectral narrowing and intensity increase could arise from defect-assisted recombination or altered injection rather than from photon confinement. The quantitative comparison (single current point, no error bars, arbitrary units) is also weak, but the structural precondition is more fundamental: without confirmation that the porous layer exists as described, the central mechanism is not securely established.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper reports GaN-based micro-LEDs on Si substrates in which a highly doped n-GaN layer is electrochemically converted into a single porous layer beneath the InGaN/GaN MQW active region. Circular, square, and hexagonal mesas are fabricated, and the porous devices are compared with unetched 'normal' devices from the same sample. The main claims are that the porous layer enhances luminous intensity by up to 22 times (hexagonal device), that the porous devices show narrower EL linewidths and resonant peaks (especially square and hexagonal mesas), and that the porous layer confines photons without significantly impairing current conduction. The paper includes a simple resistance model based on assumed pore diameters and presents J-V curves, EL spectra, and peak-wavelength versus current-density data.","tokens_in":8643,"tokens_out":3780,"duration_ms":40903,"significance":"If the central claims are correct, a single electrochemically formed porous GaN layer would be a simple and potentially practical route to improved micro-LED brightness and narrow-linewidth emission, compared with more elaborate DBR-based resonant-cavity approaches. The paper has useful strengths: it gives a detailed process flow, compares several mesa geometries on the same sample, and explicitly quantifies a plausible resistance trade-off. However, the central mechanism rests on the unverified existence and geometry of the porous layer, and the headline 22x number comes from uncalibrated single-sample EL spectra. The significance of the result is therefore conditional on additional structural and radiometric evidence.","major_comments":[{"comment":"The manuscript asserts that the highly doped n-GaN layer is converted into a porous layer by electrochemical etching and that a 100 nm u-GaN spacer 'prevents the active region from being damaged during the etching process,' but it provides no cross-sectional SEM/TEM image, no porosity measurement, and no comparison of MQW quality before and after etching. The optical microscope images in Figure 2 only show top-view mesa outlines. Because every subsequent claim (effective refractive index change, photon confinement, resonant emission, and the 22x enhancement) is attributed to this porous layer, the central mechanism is not yet supported. Add cross-sectional imaging, measured pore size/density/thickness, and PL or EQE data before and after etching for both the active region and the full device.","section":"Material and Methods and Figure 1"},{"comment":"The central quantitative claim of 'approximately 22 times' is based on EL spectra in arbitrary units at a single chosen current (10 mA) with no error bars, no reported number of devices, and no absolute power calibration. Without calibrated radiometry or integrating-sphere measurements, the 22x ratio could reflect collection efficiency, focusing, or processing variability rather than device performance. Report absolute output power or EQE with uncertainty across at least several devices and over a range of currents.","section":"Results and Discussion, Figure 5"},{"comment":"The assignment of spectral features to resonant modes is not quantitatively tested. The text invokes prior work (refs 35 and 36) for mode distributions, but it does not compare the observed peak positions or spacings with calculated cavity modes for these specific diameters/side lengths and for the unknown porous-layer refractive index. Provide a mode analysis (for example, peak spacing versus cavity size, or simulation) or otherwise the claim that square and hexagonal mesas produce 'resonant emission' remains speculative.","section":"Results and Discussion, Figures 4 and 5"},{"comment":"The resistance model assumes a pore diameter Dhole and spacing d that are not measured; the conclusion that the added resistance is 'still acceptable' is therefore illustrative only. More importantly, the model treats the pores as a static area reduction and ignores current spreading and lateral injection, so it cannot rule out altered current paths. If direct electrical measurements (for example, differential resistance or ideality factor) show that the current path is unchanged, state that explicitly; otherwise present the resistance analysis as a rough bound rather than as evidence for unchanged injection.","section":"Results and Discussion, Equations (1)-(3)"}],"minor_comments":[{"comment":"The Author Contributions lists Jing Zhou, Bin Liu, and Hong Zhao, none of whom appear in the author list; please align the contributions with the actual author list.","section":"Author Contributions"},{"comment":"There are typographical errors and inconsistencies: 'FWMH' appears in the Introduction where FWHM is meant, 'MWQs' should be 'MQWs', and 'electrochemistry' is capitalized inconsistently.","section":"Throughout"},{"comment":"The caption and text do not specify the pore spacing d or mesa radius r used in Equations (1)-(2); please define all model parameters in the caption or in the text so the calculated resistance ratios are reproducible.","section":"Figure 3"},{"comment":"The comparison current of 10 mA is described as being near the inflection points, but the inflection points are given as current densities (354.6, 320, and 213 A/cm2); please state the current density used for each device and justify the selection.","section":"Results and Discussion, Figure 5"},{"comment":"The sentence 'The peak wavelength of the normal Micro-LEDs also exhibits a blue shift, followed by continuous small amplitude changes' is vague; please specify the numerical wavelength range of the shifts.","section":"Results and Discussion, Figure 5(f)"}],"recommendation":"major_revision","confidential_remarks":"The manuscript relies heavily on the authors' prior papers (refs 34-36) for the optical mechanism, so the novel contribution is essentially the device-level comparison with a single porous layer. That is potentially useful, but the missing structural evidence makes the central mechanism unverified as it stands. The author contribution list contains names not in the author list, which the editor should check. The paper fits the scope of physics.optics as an experimental device study."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"You should know two things about this paper before reading it. First, the qualitative effect is probably real: putting a single electrochemically etched porous GaN layer under the MQWs does change the EL spectrum, and the shape-dependence (square/hexagon showing sharper peaks than circle) is visible in the figures. Second, the quantitative headline—22x intensity enhancement—is not supported by the evidence as presented, and neither is the mechanism, because the porous layer itself is never directly characterized.\n\nWhat is actually new: prior work used porous GaN in DBRs or resonant-cavity LEDs; this paper applies a single porous layer to micro-LEDs and compares mesa shapes. That combination is not in the cited literature. The fabrication is described in enough detail to attempt replication, and the J-V curves show the porous layer does not wreck current conduction—useful practical information. The resistance model, though it assumes a pore diameter that is never measured, is simple and clearly stated; it correctly shows the resistance increase is modest.\n\nThe soft spots are structural, not cosmetic. There is no SEM/TEM cross-section, no porosity measurement, no thickness profile of the porous layer. The pore size Dhole is a free parameter, not a measured quantity. The 22x claim comes from uncalibrated EL spectra at a single hand-picked current per device, in arbitrary units, with no error bars and no absolute power or EQE. The resonance explanation leans heavily on the authors' prior papers (refs 34-36) without showing mode calculations for these specific geometries. The 100 nm u-GaN spacer is asserted to protect the MQWs during etching, but no before/after PL or TEM is shown, so the possibility that the etch damages the active region and that the spectral changes come from altered injection or defect recombination is not ruled out.\n\nThese are significant gaps for a quantitative claim, but the qualitative observations are not trivial. The paper is a fabrication study, not a theoretical advance, and its value lies in the recipe and the shape-dependent spectral behavior, not in the unverified 22x number. I would not cite it for the enhancement factor, but I might cite it as an example of a single porous layer in micro-LEDs if I were working on porous GaN.\n\nIt deserves a serious referee, not a desk reject: the topic is timely for micro-display work, the fabrication is replicable, and the missing structural data can be requested in revision. I would send it out, but I would expect the reviewers to demand cross-section imaging and calibrated intensity measurements before publication.","headline":"Plausible fabrication variant, unverified structure, and the 22x headline outruns the evidence.","tokens_in":9212,"tokens_out":1688,"would_cite":false,"duration_ms":18952,"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":"A single porous layer boosts micro-LED brightness 22-fold","keywords":["porous GaN","Micro-LED","electrochemical etching","resonant emission","light confinement","efficiency-on-size effect","InGaN/GaN quantum wells","mesa geometry"],"falsifier":"Take a cross-sectional TEM or STEM image of the porous Micro-LED and compare the quantum-well region with an unetched device; visible etching damage, thickness fluctuations, or a non-continuous porous layer under the mesa would show that the intensity gain and resonance peaks come from uncontrolled damage rather than the intended porous confinement layer.","tokens_in":8282,"feed_emoji":"💡","tokens_out":6480,"duration_ms":62278,"temperature":0.7,"pith_summary":"Micro-LEDs lose efficiency as their mesas shrink below tens of micrometers, the \"efficiency-on-size effect.\" This paper tries to show that a single porous GaN layer, made by electrochemically etching a heavily doped n-GaN layer under the quantum wells, confines photons and recovers that loss. On the same wafer, porous devices reached roughly 22 times the luminous intensity of normal devices (in hexagonal mesas) and narrowed the emission to about 5.9 nm FWHM, with square and hexagonal mesas showing resonant peaks. If correct, this offers a simpler route than multi-layer DBR mirrors to bright, narrow-line Micro-LEDs for displays and lasers.","feed_headline":"Single porous layer makes micro-LEDs 22x brighter","feed_subtitle":"Electrochemical etching under the quantum wells confines light, narrows emission to about 5.9 nm, and shifts the peak to 510 nm.","key_machinery":"The central object is a single porous GaN layer formed by electrochemical etching of the heavily doped n-GaN layer, located 100 nm below the InGaN/GaN quantum wells. The pores are modeled as air cylinders of diameter $D_{\\mathrm{hole}}$ with spacing $d$, which lower the layer's effective refractive index and, the paper argues, raise the index contrast around the active region so photons are better confined and cavity resonance is enhanced. The same geometry raises series resistance only mildly (up to about a factor of 2 at the largest pore size), so vertical current conduction is preserved. This layer plus the mesa shape carries the entire argument: square and hexagonal mesas spread resonant modes across the device, whereas circular mesas keep whispering-gallery modes near the damaged etched edge.","core_discovery":"The paper reports that converting a highly doped n-GaN layer into a single porous layer below the InGaN/GaN multiple quantum wells, while leaving a 100 nm u-GaN spacer, enhances emitted light and changes the spectral character of green Micro-LEDs. Electroluminescence at 10 mA shows the porous circular, square, and hexagonal devices are 2.64, 4.86, and approximately 22 times brighter than the normal counterparts on the same sample. The main peak shifts from 535 nm toward 510 nm and stays near 510 nm as current rises, while square and hexagonal porous devices develop multiple resonance peaks; circular porous devices remain mostly spontaneous-emission dominated. The authors attribute this to the air-filled pores lowering the effective refractive index, strengthening photon confinement and microcavity resonance, with mode distribution depending on mesa shape.","pith_inferences":["If the confinement mechanism is correct, the same single-porous-layer design should transfer to other emission colors by adjusting layer thickness and doping; the paper does not test this.","The resonant peaks and roughly 5.9 nm linewidth hint at a path to electrically injected microcavity lasers, but true lasing (threshold kink, coherence, beam pattern) is not demonstrated here.","A decisive check would be angle-resolved or spatially resolved electroluminescence: genuine cavity modes should show distinct angular and spatial structure, while random scattering from pores would not.","The 22x figure is a peak-intensity comparison at one operating point; integrated-power and efficiency measurements would show whether the enhancement is total light output or a redistribution into the measured direction."],"forward_implications":["A single porous layer could replace multi-layer DBR stacks for some resonant-cavity LED applications, reducing epitaxial and fabrication complexity.","Mesa shape becomes a design knob: square and hexagonal geometries produce narrower, resonant emission, while circular geometry keeps a spontaneous-emission-like spectrum.","The wavelength locking near 510 nm suggests the porous cavity suppresses the usual current-induced blue shift, stabilizing color over a range of drive currents.","Because the resistance penalty stays acceptable, the approach could be added to existing contacted Micro-LED process flows without a separate mirror deposition step."],"supporting_citations":[{"why":"Supplies the effective-refractive-index model of the single porous GaN layer that the electrical-injection results build on.","marker":"[34]"},{"why":"Provides the regular resonance-mode analysis for polygonal GaN microdisks on Si used to explain the square and hexagonal spectra.","marker":"[35]"},{"why":"Shows single-mode-dominated resonant emission in square GaN microdisks, motivating the polygonal mesa designs.","marker":"[36]"},{"why":"Gives the nanoporous GaN DBR resonant-cavity LED with 3.4 nm FWHM that the single-porous-layer approach aims to simplify.","marker":"[32]"},{"why":"Demonstrates porous GaN in an InGaN resonant microcavity narrowing the EL linewidth, supporting the photon-confinement mechanism.","marker":"[33]"}],"fun_headline_variants":["Porous layer boosts micro-LED brightness 22-fold","Etched pores make green micro-LEDs 22x brighter","Hexagonal porous micro-LEDs shine 22x brighter","Single porous layer lifts micro-LED output 22x","Porous GaN layer turns micro-LEDs up to 22x brighter"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The paper assumes the 100 nm undoped spacer keeps the electrochemical etch from damaging the quantum wells, but it shows no cross-section images or before-and-after luminescence data to confirm that the active region is intact.","fun_headline_variants_meta":{"raw":{"variants":["Porous layer boosts micro-LED brightness 22-fold","Etched pores make green micro-LEDs 22x brighter","Hexagonal porous micro-LEDs shine 22x brighter","Single porous layer lifts micro-LED output 22x","Porous GaN layer turns micro-LEDs up to 22x brighter"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.0002,"raw_usage":{"total_tokens":1376,"prompt_tokens":945,"completion_tokens":431,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":561,"completion_tokens_details":{"reasoning_tokens":342}},"tokens_in":561,"tokens_out":431,"duration_ms":3873,"temperature":1.0,"reasoning_tokens":342,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-10T22:50:05.333201+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Take a cross-sectional TEM or STEM image of the porous Micro-LED and compare the quantum-well region with an unetched device; visible etching damage, thickness fluctuations, or a non-continuous porous layer under the mesa would show that the intensity gain and resonance peaks come from uncontrolled damage rather than the intended porous confinement layer.","supporting_citations":[],"review_version":1}