{"id":"f3e0503a-ca45-48fa-a888-73f9a3427202","arxiv_id":"2501.10142","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":4,"one_line_summary":"Room-temperature electroluminescence at 866 nm is demonstrated in 166-nm-wide n-i-n GaAs nanopillar arrays, with an estimated internal quantum efficiency around 2%.","lead":"Researchers made tiny light-emitting pillars from n-type GaAs that glow without any p-type doping. The devices emit at 866 nanometers at room temperature, a first for sub-micrometer unipolar LEDs, and could simplify future nanoscale light sources.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The IQE>2% claim is not independently measured: the same lifetime sets both S and A, and the stated A=3.34e9 s^-1 is numerically inconsistent with S=2.08e4 cm/s and da=490 nm; recalculating with consistent values can drop the low-injection IQE below 2%.","rationale":"I read the central claim as two-part: (a) room-temperature EL at ~866 nm from n-i-n GaAs nanopillar arrays is demonstrated; (b) lifetimes >300 ps imply IQE>2%. Part (a) is supported by the spectral data in Fig. 3 and the size-dependent comparison in Supplement 1; it is credible. Part (b) is where the paper is least secure. The Supplement's IQE formula is a standard ABC model, but the A coefficient is not independently measured: it is set by inverting the measured lifetime with τ^-1=4S/da. This makes the IQE a rearrangement of the lifetime, not a measurement. More seriously, the inserted numerical values are internally inconsistent: for da=490 nm and S=2.08e4 cm/s, A would be 1.70e9 s^-1, whereas the paper states 3.34e9 s^-1; the latter requires da≈250 nm. This is not a stylistic discrepancy; it changes the inferred IQE by roughly a factor of two and can push the low-injection value below the advertised 2%. The reader's weakest assumption is closely related, but I locate the failure one step earlier: even the S value is an inference from a single lifetime at one device size, and the paper supplies neither raw decays nor an independent surface-recombination measurement. The demonstration of EL itself is not affected, so I do not recommend rejection; the efficiency claim should be recalculated and the raw data released before the quantitative result is accepted.","tokens_in":9805,"tokens_out":9048,"duration_ms":89459,"concrete_test":"Obtain the raw TREL decay traces and refit them to extract τ with confidence intervals; then recompute IQE enforcing A=4S/da with the stated da=490 nm and S either independently measured or consistently derived from τ. Specifically, use A=1.70×10^9 s^-1 (or the value from the refit) in Supplement Eq. (E.2) with B=1.8×10^-10 cm^3/s, C=3.5×10^-30 cm^6/s, ηi=0.11, and Nd=2×10^17-2×10^18 cm^-3. If the resulting IQE is <2%, the abstract/conclusion efficiency claim must be downgraded. As a stronger independent check, measure τ at three pillar diameters under identical low-injection bias and verify τ^-1 vs 4/da is linear with zero intercept; non-linearity invalidates the S-extraction and the IQE model.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The quantitative headline ('IQE >2% at sub-mA') is a model-based quantity, and the model is partly circular. In the IQE section the authors estimate S from the low-injection differential lifetime via τ^-1≈4S/da (Fig. 4), then set A=4S/da in Supplement Eq. (E.2) and call the result IQE. Because A is fixed by the same measured lifetime, the calculation cannot validate the assumption that radiative and Auger recombination are negligible at the operating point; indeed the claimed ηr≈2-10% requires B Nd not to be negligible compared with A, so assigning τ^-1 entirely to S is internally inconsistent. The arithmetic is also inconsistent: with the quoted S=2.08×10^4 cm/s and da=490 nm, A=4S/da=1.70×10^9 s^-1, not 3.34×10^9 s^-1 as stated; matching A=3.34×10^9 would require da≈250 nm. Recomputing Eq. (E.2) with A=1.70×10^9, B=1.8×10^-10 cm^3/s, C=3.5×10^-30 cm^6/s, and ηi=0.11 gives IQE≈0.2-1.9% over Nd=2×10^17-2×10^18 cm^-3, below the claimed 2% low-injection value. No raw TREL traces, error bars, absolute photon-count calibration, or independent S measurement are supplied. The qualitative room-temperature EL demonstration is not endangered, but the efficiency number is not supported as stated.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript reports room-temperature electroluminescence (EL) at approximately 866 nm from sub-micrometer n-i-n unipolar GaAs nanopillar LED arrays with a top pillar diameter of 166 nm. The devices show diode-like I-V characteristics, a nonlinear L-I curve, and time-resolved EL decay lifetimes larger than 300 ps. The authors attribute hole generation to impact ionization and Zener tunneling in an AlAs/GaAs/AlAs double-barrier quantum well, and they estimate an internal quantum efficiency (IQE) higher than 2% at sub-mA injection using an ABC recombination model with parameters taken from prior work and an assumed injection efficiency.","tokens_in":10187,"tokens_out":5765,"duration_ms":52011,"significance":"If the quantitative efficiency estimate were reliable, the demonstration of room-temperature EL in an n-type-only GaAs nanoLED would represent a genuinely new class of nanoscale emitters without p-type doping, potentially mitigating contact resistance and optical absorption losses. The qualitative observation is well supported: the EL peak at 866 nm is consistent with the GaAs band-edge transition, the spectral comparison across pillar sizes (Supplementary Figs. S1 and S4) is informative, and the surface recombination velocity estimate is usefully compared with an independent PL-based benchmark (S = 1.1 x 10^4 cm/s from [18]). However, the headline IQE claim is not independently established; it is a model-dependent recasting of the lifetime measurement with assumed parameters, and it contains a numerical inconsistency. The qualitative room-temperature EL demonstration is not endangered, but the efficiency number needs substantial revision or re-framing.","major_comments":[{"comment":"The stated value A = 3.34 x 10^9 s^-1 is inconsistent with the given S = 2.08 x 10^4 cm/s and da = 490 nm, since 4S/da = 4*(2.08 x 10^4 cm/s)/(4.90 x 10^-5 cm) = 1.70 x 10^9 s^-1. Recomputing Eq. (E.2) with A = 1.70 x 10^9 s^-1, B = 1.8 x 10^-10 cm^3/s, C = 3.5 x 10^-30 cm^6/s, and eta_i = 0.11 gives IQE values from approximately 0.2% at Nd = 2 x 10^17 cm^-3 to 2.3% at Nd = 2 x 10^18 cm^-3; the claim of 'IQE >2% at sub-mA' is therefore not supported across the stated carrier-density range. The authors should correct the arithmetic and re-evaluate the quantitative conclusion.","section":"Surface recombination velocity and IQE (main text) and Supplement Eq. (E.2)"},{"comment":"The IQE estimate is partly circular: the surface recombination velocity S is extracted from the measured differential lifetime via tau^-1 ~ 4S/da, and the same 4S/da is then used to set the A coefficient in the ABC model. As a result, the IQE is a recasting of the lifetime measurement with literature values for B and C, not an independent determination. Moreover, the assumption that surface recombination dominates at low injection, used to extract S, is in tension with the derived radiative efficiencies: for the IQE to reach 2% with eta_i = 0.11 requires eta_r ~ 18%, which from Eq. (E.2) implies B*Nd is not negligible compared with A at the upper end of the assumed Nd range. Please state this limitation explicitly or support the IQE with an independent S measurement.","section":"Surface recombination velocity and IQE (main text) and Supplement Eqs. (E.1) and (E.2)"},{"comment":"The IQE calculation relies on assumed values of eta_i = 0.11 (taken from micro-sized devices [15]) and an assumed donor density range Nd = 2 x 10^17 - 2 x 10^18 cm^-3, with no direct measurement of the carrier density under electrical operation. Since the IQE varies by roughly an order of magnitude across this assumed Nd range, the claimed 'IQE >2%' is not robust. The authors should provide an experimental constraint on Nd (for example, from geometry and injection current density or from capacitance measurements) or present the IQE as a sensitive range with the assumed parameters explicitly identified.","section":"Surface recombination velocity and IQE (main text)"}],"minor_comments":[{"comment":"The sentence 'TREL measurements were used to estimate the differential carrier lifetimes of the EL of the unipolar microLEDs' should refer to the nanoLEDs, since the measurements presented in Fig. 4 are on the nanopillar arrays.","section":"Time-resolved electroluminescence"},{"comment":"The manuscript reports only fitted lifetime values without showing the raw TREL histograms or the mono-exponential fits; including representative decay traces with residuals would allow the reader to assess the quality of the fit.","section":"Fig. 4 and Supplement Eq. (E.1)"},{"comment":"The text contains typographical errors: 'Zenner tunneling' should read 'Zener tunneling', and 'simplifing' should be 'simplifying'.","section":"Introduction and Fig. 1 caption"},{"comment":"The data availability statement indicates that data may be obtained from the authors upon reasonable request; depositing the raw EL spectra and time-resolved decay histograms in a public repository would improve reproducibility.","section":"Data availability"}],"recommendation":"major_revision","confidential_remarks":"The qualitative demonstration of room-temperature EL from n-type unipolar GaAs nanoLEDs is novel and likely of interest to the readership. However, the quantitative IQE>2% claim is the main headline and is not supported as stated: besides the numerical inconsistency in A, the estimate is a model-dependent recasting of the lifetime measurement with several assumed parameters. The authors should be asked to correct the arithmetic, clearly acknowledge the circularity, and either provide independent support for the efficiency or soften the claim. The paper is publishable after such revision."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The paper reports room-temperature electroluminescence at ~866 nm from n-i-n GaAs nanopillar arrays with 166 nm top diameter, which is genuinely new. Prior unipolar III-V EL was cryogenic or microscale, so the sub-micrometer demonstration is a real step. The fabrication is careful, the passivation comparison with their earlier work is sensible, and the size-dependent disappearance of the 806 nm DBQW peak is a nice observation with a plausible surface-recombination explanation. Credit is due for the L-I-V and TREL data, and for acknowledging that the EQE is limited by metal-coating extraction losses.\n\nThe soft spots are concentrated in the efficiency estimate. The stress-test arithmetic holds up: with S=2.08e4 cm/s and da=490 nm, A=4S/da gives 1.70e9 s^-1, not 3.34e9. The higher value requires da~250 nm, which contradicts the stated cone-correction. Recomputing the ABC radiative efficiency with consistent parameters puts the low-injection IQE closer to 1-2% rather than clearly above 2%, so the headline number is not supported as written. There is also a methodological circularity: S is extracted from the measured lifetime, then used to set A, so the IQE is a recasting of the lifetime with model constants rather than an independent measurement. That can be acceptable as a rough estimate, but it should be labeled as such, and the absence of error bars on the lifetime fits and of raw TREL traces makes the quantitative claim hard to evaluate.\n\nThe core observation itself is credible: the spectrum is a clean GaAs band-edge peak, the I-V shows diode-like turn-on, and the EL clearly tracks current. The paper is worth a serious referee because the device class is new and the qualitative result is likely reproducible. My recommendation is to send it to peer review with a request for major revision: correct the A inconsistency, present the IQE as a model-dependent estimate with a sensitivity range, and include the raw TREL fits with uncertainties. If the efficiency claim is softened and the data are provided, this becomes a solid letter.","headline":"A credible first demonstration of room-temperature EL from sub-micrometer n-i-n GaAs nanopillars, but the IQE>2% headline rests on an arithmetic error and a model-dependent estimate; the qualitative result deserves review, the efficiency number needs fixing.","tokens_in":10741,"tokens_out":2013,"would_cite":false,"duration_ms":21665,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":["78.60.Fi","85.60.Jb"],"model":"deepseek-v4-flash","headline":"First room-temperature light from n-type-only GaAs nanoLEDs","keywords":["n-i-n unipolar nanoLED","GaAs electroluminescence","double-barrier quantum well","impact ionization","Zener tunneling","time-resolved electroluminescence","internal quantum efficiency","nanopillar array"],"falsifier":"Measure the differential carrier lifetime for a series of pillar diameters under identical bias and check whether $\\tau^{-1}$ scales linearly with $1/d_a$ with a constant surface velocity; alternatively, collect the emitted light with an integrating sphere and count photons to obtain an absolute external quantum efficiency. If the lifetime does not follow the $4S/d_a$ scaling, or if the absolute photon count implies an efficiency far below the estimated IQE after accounting for extraction, the central efficiency claim would be unsupported.","tokens_in":9610,"feed_emoji":"💡","tokens_out":5313,"duration_ms":44637,"temperature":0.7,"pith_summary":"This paper reports the first room-temperature electroluminescence from sub-micrometer n-i-n unipolar GaAs light-emitting diodes, devices that contain no p-type doping at all. The emitters are arrays of nanopillars with top diameters as small as 166 nm, and they shine at the GaAs band-edge wavelength of about 866 nm. Time-resolved electroluminescence shows decay lifetimes above 300 ps, and the authors estimate an internal quantum efficiency above 2% at sub-mA injection, rising toward 6% at higher injection. If correct, the result shows that the p-type layer long thought necessary for III-V light emitters can be replaced by hole generation through impact ionization and Zener tunneling inside a double-barrier quantum well, opening a simpler path to nanoscale light sources.","feed_headline":"First room-temperature light from n-type-only GaAs nanoLEDs","feed_subtitle":"166 nm n-type-only nanopillars emit 866 nm light, with internal efficiency above 2 percent at sub-mA injection.","key_machinery":"The load-bearing element is the AlAs/GaAs/AlAs double-barrier quantum well embedded in the n-i-n diode, which under bias generates minority holes through impact ionization and Zener (interband) tunneling. Around it, the argument uses the relation $\\tau^{-1}\\approx 4S/d_a$ to convert measured differential carrier lifetimes into a surface recombination velocity $S\\approx 2\\times10^4$ cm/s, and then inserts this $S$ as the linear coefficient $A=4S/d_a$ in the ABC recombination model $\\eta_r = BN_d/(A+BN_d+CN_d^2)$ to estimate internal quantum efficiency. The claimed performance therefore rests on combining a passivation mechanism that keeps $S$ low in nanoscale pillars with a hole-generation mechanism that does not require p-type doping.","core_discovery":"The central claim is that a purely n-type (electron-transporting) n-i-n GaAs diode can emit light at room temperature when shaped into a nanopillar array. Holes are not supplied by doping; they are created inside the device by impact ionization and Zener tunneling in the high electric field of an AlAs/GaAs/AlAs double-barrier quantum well, then recombine radiatively with electrons in the n-doped GaAs emitter near the band edge, producing the observed 866 nm peak. The paper reports differential carrier lifetimes above 300 ps for 166 nm top-diameter pillars, only 2.8-fold shorter than for 6 µm unipolar microLEDs from the same epilayer design, and attributes the relatively long lifetime to sulfur chemical treatment plus a 10 nm Si$_3$N$_4$ coating that suppresses surface recombination. From these lifetimes the authors estimate an internal quantum efficiency of about 2% at low injection and up to 6% before Auger saturation, with a very small external quantum efficiency below $10^{-5}$ dominated by light blocked by the metal contact coating.","pith_inferences":["The IQE figure is essentially an interpretation of the measured lifetime under the assumption that surface recombination dominates at low injection; a direct photon-counting measurement of the external quantum efficiency would test whether the internal efficiency is really as high as 2-6%.","The disappearance of the 806 nm quantum-well emission in pillars below 1 µm suggests that shrinking the device diverts current away from coherent resonant-tunneling pathways; this could be exploited or countered by engineering the barrier to favor resonant injection.","If the 2.8-fold lifetime penalty between micro- and nanopillars is mostly due to surface effects, even better passivation or core-shell designs could bring nanoscale unipolar devices close to the efficiency of their micron-scale counterparts.","The large gap between internal (2-6%) and external ($<10^{-5}$) efficiency is dominated by the metal contact blocking emission, so switching to transparent contacts or optimizing the angled deposition could yield orders-of-magnitude improvements in usable light output without changing the emitter physics."],"forward_implications":["If the central claim holds, nanoscale III-V emitters no longer require p-type doping, removing a source of series resistance, optical absorption, and fabrication complexity.","The n-i-n architecture with a double-barrier quantum well could be transferred to other III-V materials such as InP, extending unipolar emission across visible and infrared wavelengths.","Sub-300 ps lifetimes imply the devices can be modulated at gigahertz rates, making them candidates for fast nanoscale light sources.","The measured lifetimes indicate that sulfur-plus-Si$_3$N$_4$ passivation keeps surface recombination low even for pillars with high surface-to-volume ratio, a result that would apply to other etched nanostructures.","With improved carrier injection efficiency and transparent contacts, the paper projects internal quantum efficiency could exceed 15% and external efficiency could rise well above the current $10^{-5}$."],"supporting_citations":[{"why":"Provides the identical epilayer design and the microLED baselines (806 nm and 866 nm emission, ~821 ps lifetime, S~2e5 cm/s) against which the nanoLED results are compared.","marker":"[15]"},{"why":"Supplies the sulfur treatment and low-frequency Si3N4 passivation procedure and the optically pumped GaAs nanopillar surface recombination velocity S=1.1e4 cm/s used as a benchmark.","marker":"[18]"},{"why":"Shows electroluminescence in unipolar GaN/AlN double-barrier structures via interband tunneling, establishing the mechanism class the paper extends to NIR GaAs.","marker":"[10]"},{"why":"Reports unipolar InGaAs/AlAs resonant-tunneling diode electroluminescence and suggests higher injection efficiency values used in the IQE projections.","marker":"[16]"},{"why":"Documents impact-ionization-driven electroluminescence in GaAs/AlGaAs heterostructures, supporting the hole-generation mechanism invoked for the low-field regime.","marker":"[17]"},{"why":"Motivates unipolar n-type quantum dot diode structures for quantum light sources, the application direction the paper positions its nanoLEDs toward.","marker":"[8]"}],"fun_headline_variants":["N-type-only GaAs nanopillars emit light via tunneling","Unipolar GaAs nanoLEDs glow at 866 nm","Electron-only GaAs nanoLEDs shine at room temperature","Holes from Zener tunneling light up n-type nanoLEDs"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The efficiency estimate assumes that the measured carrier lifetime at low injection is set entirely by surface recombination, through $\\tau^{-1}\\approx 4S/d_a$, with no separate radiative or Auger contribution, and the same surface velocity is then used as the linear coefficient in the ABC model, so the reported IQE is a model-dependent recasting of that lifetime rather than an independent measurement.","fun_headline_variants_meta":{"raw":{"variants":["N-type-only GaAs nanopillars emit light via tunneling","Unipolar GaAs nanoLEDs glow at 866 nm","Electron-only GaAs nanoLEDs shine at room temperature","Holes from Zener tunneling light up n-type nanoLEDs"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.00124,"raw_usage":{"total_tokens":5074,"prompt_tokens":911,"completion_tokens":4163,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":527,"completion_tokens_details":{"reasoning_tokens":4091}},"tokens_in":527,"tokens_out":4163,"duration_ms":28367,"temperature":1.0,"reasoning_tokens":4091,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-10T19:23:08.171278+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Measure the differential carrier lifetime for a series of pillar diameters under identical bias and check whether $\\tau^{-1}$ scales linearly with $1/d_a$ with a constant surface velocity; alternatively, collect the emitted light with an integrating sphere and count photons to obtain an absolute external quantum efficiency. If the lifetime does not follow the $4S/d_a$ scaling, or if the absolute photon count implies an efficiency far below the estimated IQE after accounting for extraction, the central efficiency claim would be unsupported.","supporting_citations":[{"cited_title":"Room-temperature electroluminescence and light detection from III-V unipolar microLEDs without p-type doping,","cited_arxiv_id":null,"evidence_quote":"Provides the identical epilayer design and the microLED baselines (806 nm and 866 nm emission, ~821 ps lifetime, S~2e5 cm/s) against which the nanoLED results are compared."},{"cited_title":"Surface Passivation of III –V GaAs Nanopillars by Low-Frequency Plasma Deposition of Silicon Nitride for Active Nanophotonic Devices,","cited_arxiv_id":null,"evidence_quote":"Supplies the sulfur treatment and low-frequency Si3N4 passivation procedure and the optically pumped GaAs nanopillar surface recombination velocity S=1.1e4 cm/s used as a benchmark."},{"cited_title":"Near -UV electroluminescence in unipolar - doped, bipolar -tunneling GaN/AlN heterostructures,","cited_arxiv_id":null,"evidence_quote":"Shows electroluminescence in unipolar GaN/AlN double-barrier structures via interband tunneling, establishing the mechanism class the paper extends to NIR GaAs."},{"cited_title":"Electroluminescence in Unipolar -Doped In0.53Ga0.47As/AlAs Resonant -Tunneling Diodes: A Competition between Interband Tunneling and Impact Ionization,","cited_arxiv_id":null,"evidence_quote":"Reports unipolar InGaAs/AlAs resonant-tunneling diode electroluminescence and suggests higher injection efficiency values used in the IQE projections."},{"cited_title":"Electroluminescence from GaAs/AlGaAs Heterostructures in Strong in-Plane Electric Fields: Evidence for k- and Real-Space Charge Transfer,","cited_arxiv_id":null,"evidence_quote":"Documents impact-ionization-driven electroluminescence in GaAs/AlGaAs heterostructures, supporting the hole-generation mechanism invoked for the low-field regime."},{"cited_title":"A Unipolar Quantum Dot Diode Structure for Advanced Quantum Light Sources,","cited_arxiv_id":null,"evidence_quote":"Motivates unipolar n-type quantum dot diode structures for quantum light sources, the application direction the paper positions its nanoLEDs toward."}],"review_version":1}