arxiv: 2604.03485 · v1 · submitted 2026-04-03 · ⚛️ physics.app-ph

Microwave Performance of all MOCVD-grown AlScN/GaN MIS-HEMTs on Semi-Insulating GaN Substrates

Can Cao , Vijay Gopal Thirupakuzi Vangipuram , Abdul Mukit , Motahareh Helli , Jinwoo Hwang , Hongping Zhao , Wu Lu This is my paper

Pith reviewed 2026-05-13 17:41 UTC · model grok-4.3

classification ⚛️ physics.app-ph

keywords AlScNGaNMIS-HEMTMOCVDsubthreshold swingpower densitynoise figureinterface traps

0 comments p. Extension

The pith

All-MOCVD AlScN/GaN MIS-HEMTs on semi-insulating GaN substrates reach 1 A/mm drain current and 25.8 GHz fT in 1-micron gate devices.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper shows that long-gate metal-insulator-semiconductor high electron mobility transistors can be fabricated entirely by metal-organic chemical vapor deposition using an aluminum scandium nitride barrier layer on a gallium nitride channel, grown on semi-insulating gallium nitride substrates. These 1-micrometer gate devices deliver a maximum drain current density of 1 A/mm, an on/off ratio of 2 times 10 to the 5, a subthreshold swing of 63 mV/decade, and current dispersion as low as 7.8 percent, linked to an interface trap density of 2.11 times 10 to the 11 per square centimeter per electron volt. Radio-frequency testing yields an fT of 25.8 GHz and fmax of 51.1 GHz, while load-pull measurements at 10 GHz give 4.04 W/mm output power with 22.7 percent power-added efficiency and noise figures below 2.5 dB across a wide current range. A reader would care because the results move the material system from prior short-gate demonstrations into practical high-voltage microwave operation, showing that the growth approach and substrate choice sustain performance without added complexity.

Core claim

The authors establish that all-MOCVD-grown AlScN/GaN MIS-HEMTs on semi-insulating GaN substrates with 1 μm gate length and 0.9 μm gate-drain spacing exhibit a maximum drain current density of 1 A/mm, on/off ratio of 2×10^5, three-terminal breakdown of 63 V, subthreshold swing of 63 mV/dec, current dispersion of 7.8% at 10 V, fT/fmax of 25.8/51.1 GHz, output power density of 4.04 W/mm with 22.7% PAE at 10 GHz, and minimum noise figure below 2.5 dB below 6 GHz. This performance stems from the low interface trap density of 2.11×10^11 cm^{-2}eV^{-1} together with the low threading dislocation density of the semi-insulating substrate, confirming the material system works in the long-gate high-bia

What carries the argument

The all-MOCVD growth of the AlScN barrier on GaN channel atop a semi-insulating GaN substrate that produces low threading dislocation density and enables the measured low interface trap density Dit of 2.11×10^11 cm^{-2}eV^{-1} in the metal-insulator-semiconductor gate stack.

If this is right

The same growth sequence supports both high-frequency small-signal metrics and high-power large-signal operation at 10 GHz.
Low Dit directly enables the near-ideal subthreshold characteristics and reduced trapping effects.
Microwave noise performance stays below 2.5 dB across drain currents from 100 to 700 mA/mm.
The structure maintains a 63 V breakdown voltage at short gate-drain spacing suitable for power applications.
Current dispersion remains low at 7.8% under 10 V bias, indicating stable operation.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

The single-growth-process approach could reduce steps compared with hybrid growth methods that switch between techniques.
Low noise across a broad current range suggests the devices suit both power amplifiers and low-noise receiver chains on the same platform.
Extending gate lengths further or testing at higher frequencies would test whether the performance scaling holds without new trap mechanisms.
The semi-insulating substrate may allow monolithic integration of passive elements with better isolation than on conductive templates.

Load-bearing premise

The low current dispersion and near-ideal subthreshold swing arise mainly from the low interface trap density achieved by the specific MOCVD growth parameters on this semi-insulating GaN substrate.

What would settle it

Reproducing the identical 1 μm gate devices on a conventional GaN substrate with higher dislocation density and measuring current dispersion well above 7.8% or subthreshold swing well above 63 mV/dec would falsify the claimed role of the substrate and interface quality.

Figures

Figures reproduced from arXiv: 2604.03485 by Abdul Mukit, Can Cao, Hongping Zhao, Jinwoo Hwang, Motahareh Helli, Vijay Gopal Thirupakuzi Vangipuram, Wu Lu.

**Figure 1.** Figure 1: (a) Schematic of epitaxial and processed device structure of the AlScN/GaN MIS-HEMTs. (b) Microscope image of an AlScN/GaN MIS-HEMT. (c) High resolution STEM cross-sectional imaging of the AlScN/AlN/GaN MIS-HEMT structure. (d) STEM-EDX elemental color map of the MIS-HEMT layers, and (e) 1D simulated energy band diagram and electron density profile of the AlScN/GaN MIS-HEMT with SiN cap layer. The simulated… view at source ↗

read the original abstract

We report on the design, fabrication, and characterization of all MOCVD-grown long-gate AlScN/GaN metal-insulator-semiconductor high electron mobility transistors (MIS-HEMTs) on semi-insulating GaN substrates. Devices with a gate length of $1~\mu$m and gate-drain spacing of $0.9~\mu$ m exhibit a maximum drain current density of 1 A/mm, an on/off current ratio of $2\times 10^5$, and a three-terminal breakdown voltage of 63 V. The device has near-ideal subthreshold characteristics with a subthreshold swing of 63 mV/dec and a current dispersion as low as 7.8$\%$ at 10 V due to the excellent interfacial quality with a trap density ($D_\mathrm{it}$) of $2.11\times{10}^{11}~\mathrm{cm}^{-2}eV^{-1}$ and the semi-insulating GaN substrate with a low threading dislocation density. Small-signal RF measurements reveal an $f_\mathrm{T}/f_\mathrm{max}$ of 25.8/51.1 GHz, while large-signal load-pull characterization at 10 GHz demonstrates an output power density of 4.04 W/mm with a power-added efficiency of 22.7$\%$. In addition, a minimum noise figure below 2.5 dB was measured over a wide drain current range from 100 mA/mm to 700 mA/mm below 6 GHz. These results extend previous demonstrations of short-gate MOCVD-grown AlScN/GaN HEMTs to the long-gate, high-voltage regime, confirming the robustness of this material system for both high-frequency and high-power device applications with favorable microwave noise performance.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Desk Editor's Note private letter to a colleague

The paper delivers usable measured RF metrics for these long-gate devices but the explanation for the low dispersion and ideal swing relies on unproven links to the substrate and interface without control experiments.

read the letter

The main thing to know is that this reports solid measured performance for 1 μm gate AlScN/GaN MIS-HEMTs on semi-insulating GaN, with max current of 1 A/mm, fT/fmax of 25.8/51.1 GHz, 4 W/mm power at 10 GHz, and low noise. It extends earlier short-gate MOCVD work into the high-voltage long-gate area. They do a good job showing the DC characteristics, small-signal RF, load-pull, and noise data all in one place. The numbers line up with the device sizes they give, and the low dispersion of 7.8% plus 63 mV/dec swing are competitive. Reporting both power and noise performance is practical for the target applications. The soft spot is how they explain it. They tie the good results to low Dit of 2.11×10^11 cm^{-2}eV^{-1} and the substrate's low dislocation density, but the abstract has no XRD data, no TEM, no growth specifics, and no comparison devices on other substrates. If the dispersion reduction comes from passivation or dielectric instead, the story about the substrate enabling this doesn't hold as strongly. The stress-test note flags this correctly based on what's visible. Full paper might include more details, but the attribution looks under-supported without those controls. This paper is for researchers in GaN RF power devices looking for data on AlScN barriers in practical voltage ranges. It deserves a serious referee because it has real experimental results that could be useful even if the interpretation needs tightening with additional characterization. I would send it to peer review.

Referee Report

2 major / 2 minor

Summary. The manuscript reports fabrication and characterization of all-MOCVD-grown AlScN/GaN MIS-HEMTs on semi-insulating GaN substrates. For devices with 1 μm gate length and 0.9 μm gate-drain spacing, it presents DC metrics (Id,max = 1 A/mm, Ion/Ioff = 2×10^5, SS = 63 mV/dec, current dispersion = 7.8% at 10 V), RF metrics (fT/fmax = 25.8/51.1 GHz), large-signal performance (4.04 W/mm output power density and 22.7% PAE at 10 GHz), and noise performance (NFmin < 2.5 dB below 6 GHz over a wide bias range), attributing the low dispersion and near-ideal subthreshold behavior to an extracted Dit of 2.11×10^11 cm^{-2}eV^{-1} and the low threading-dislocation-density substrate.

Significance. If the central attribution holds, the work would be significant for demonstrating that MOCVD-grown AlScN/GaN can deliver competitive high-voltage, high-frequency, and low-noise performance in the long-gate regime, extending prior short-gate results and supporting the material system for high-power RF applications. The reported metrics are internally consistent with the stated dimensions and represent concrete, falsifiable device data.

major comments (2)

Abstract and §3 (Device Characterization): The claim that the 7.8% current dispersion and 63 mV/dec subthreshold swing result primarily from Dit = 2.11×10^11 cm^{-2}eV^{-1} and the semi-insulating GaN substrate's low threading dislocation density is load-bearing for the headline explanation yet unsupported by quantitative TDD data (e.g., XRD FWHM values or TEM counts) or by control devices on higher-TDD substrates such as sapphire or SiC. Without these, the attribution cannot be isolated from possible contributions of unstated surface passivation or gate-stack processing.
§2 (Growth and Fabrication): No MOCVD growth parameters (temperature, pressure, Sc flux, or V/III ratios) or substrate qualification metrics are provided, preventing independent assessment of how the reported Dit and low dispersion were achieved or reproduced.

minor comments (2)

Abstract and §4 (RF Measurements): The manuscript should report error bars, number of devices measured, and full S-parameter or load-pull data sets rather than single-point values to allow verification of the quoted fT/fmax and power-density figures.
Figure captions and methods: Clarify whether the semi-insulating GaN substrate itself was MOCVD-grown or commercially supplied, and specify the exact procedure used to extract Dit (e.g., conductance or Terman method).

Simulated Author's Rebuttal

2 responses · 1 unresolved

We thank the referee for the constructive comments and the opportunity to clarify key aspects of our work. We address each major comment below and indicate where revisions will be made.

read point-by-point responses

Referee: [—] Abstract and §3 (Device Characterization): The claim that the 7.8% current dispersion and 63 mV/dec subthreshold swing result primarily from Dit = 2.11×10^11 cm^{-2}eV^{-1} and the semi-insulating GaN substrate's low threading dislocation density is load-bearing for the headline explanation yet unsupported by quantitative TDD data (e.g., XRD FWHM values or TEM counts) or by control devices on higher-TDD substrates such as sapphire or SiC. Without these, the attribution cannot be isolated from possible contributions of unstated surface passivation or gate-stack processing.

Authors: We agree that the attribution would be strengthened by quantitative TDD data and control devices. The reported Dit value was extracted from frequency-dependent C-V measurements on fabricated MIS capacitors using the Terman method. The substrate is a commercial semi-insulating GaN wafer specified by the supplier as having low threading dislocation density, which is consistent with the observed low dispersion and near-ideal subthreshold swing. We will revise the manuscript to explicitly describe the Dit extraction procedure, add a statement acknowledging that contributions from gate-stack processing cannot be fully isolated without additional controls, and note the absence of direct TDD quantification in this study. We maintain that the combination of the extracted Dit and the substrate properties provides a plausible explanation for the device metrics, but we accept that the causal isolation is not definitive. revision: partial
Referee: [—] §2 (Growth and Fabrication): No MOCVD growth parameters (temperature, pressure, Sc flux, or V/III ratios) or substrate qualification metrics are provided, preventing independent assessment of how the reported Dit and low dispersion were achieved or reproduced.

Authors: We will add the requested MOCVD growth parameters (temperature, pressure, Sc flux, and V/III ratios) and any available substrate qualification metrics to the revised §2 to improve reproducibility. revision: yes

standing simulated objections not resolved

Absence of quantitative TDD data (XRD FWHM or TEM counts) and control devices on higher-TDD substrates to definitively isolate the substrate contribution from gate-stack effects.

Circularity Check

0 steps flagged

No circularity; results are direct experimental measurements with no derivations or self-referential reductions

full rationale

The manuscript presents only measured device metrics (Id,max = 1 A/mm, SS = 63 mV/dec, dispersion = 7.8%, fT/fmax = 25.8/51.1 GHz, Pout = 4.04 W/mm) obtained from fabricated long-gate AlScN/GaN MIS-HEMTs. No equations, fitted parameters, or predictions appear that reduce to their own inputs by construction. The stated attribution of performance to measured Dit = 2.11e11 cm^{-2}eV^{-1} and low-TDD substrate is an interpretive claim resting on experimental data, not a derivation that loops back on itself. No self-citations, ansatzes, or uniqueness theorems are invoked in the provided text. The work is therefore self-contained as a characterization study.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 0 invented entities

Central claims rest entirely on experimental device characterization rather than theoretical derivations; no free parameters, new entities, or non-standard axioms are introduced beyond routine semiconductor physics assumptions about interface traps and carrier transport.

axioms (1)

domain assumption Standard assumptions in HEMT physics that subthreshold swing and Dit extraction accurately reflect interface quality.
Invoked implicitly when linking measured 63 mV/dec swing and Dit value to low dispersion.

pith-pipeline@v0.9.0 · 5664 in / 1508 out tokens · 44938 ms · 2026-05-13T17:41:03.467297+00:00 · methodology

discussion (0)

Reference graph

Works this paper leans on

19 extracted references · 19 canonical work pages

[1]

Polarization induced interface and electron sheet charges of pseudomorphic ScAlN/GaN, GaAlN/GaN, InAlN/GaN, and InAlN/InN heterostructures,

O. Ambacher, B. Christian, M. Yassine, M. Baeumler, S. Leone, and R. Quay, “Polarization induced interface and electron sheet charges of pseudomorphic ScAlN/GaN, GaAlN/GaN, InAlN/GaN, and InAlN/InN heterostructures,” J. Appl. Phys. 129(20), 204501 (2021)

work page 2021
[2]

Wurtzite ScAlN, InAlN, and GaAlN crystals, a comparison of structural, elastic, dielectric, and piezoelectric properties,

O. Ambacher, B. Christian, N. Feil, D. F. Urban, C. Elsässer, M. Prescher, and L. Kirste, “ Wurtzite ScAlN, InAlN, and GaAlN crystals, a comparison of structural, elastic, dielectric, and piezoelectric properties,” J. Appl. Phys. 130(4), 045102 (2021)

work page 2021
[3]

Transport properties of polarization-induced 2D electron gases in epitaxial AlScN/GaN heterojunctions,

J. Casamento, T. S. Nguyen, Y. Cho, et al. , “Transport properties of polarization-induced 2D electron gases in epitaxial AlScN/GaN heterojunctions,” Appl. Phys. Lett. 121(19), 192101 (2022)

work page 2022
[4]

Experimental determination of giant polarization in wurtzite III - nitride semiconductors,

H. Ye, P. Wang, R. Wang, J. Wang , et al., “Experimental determination of giant polarization in wurtzite III - nitride semiconductors,” Nat. Commun. 16(1), 3863 (2025)

work page 2025
[5]

Dawn of nitride ferroelectric semiconductors: From materials to devices,

P. Wang, D. Wang, S. Mondal, M. Hu, J. Liu, and Z. Mi, “Dawn of nitride ferroelectric semiconductors: From materials to devices,” Semicond. Sci. Technol. 38(4), 043002 (2023)

work page 2023
[6]

Recent advancement in ScAlN/GaN high electron mobility transistors: Materials, properties, and device performance,

P. Murugapandiyan, S. Maheswari, A. S. A. Fletcher, G. Saranya, and P. Anandan, “Recent advancement in ScAlN/GaN high electron mobility transistors: Materials, properties, and device performance,” Mater. Sci. Semicond. Process. 193(7), 109509 (2025)

work page 2025
[7]

ScAlN-GaN Transistor Technology for Millimeter-wave Ultra-high Power and Efficient MMICs,

E. M. Chumbes et al., "ScAlN-GaN Transistor Technology for Millimeter-wave Ultra-high Power and Efficient MMICs," in 2022 IEEE/MTT-S International Microwave Symposium - IMS 2022, 295 ( 2022)

work page 2022
[8]

RF Power Performance of Sc(Al,Ga)N/GaN HEMTs at Ka-Band,

A. J. Green et al., "RF Power Performance of Sc(Al,Ga)N/GaN HEMTs at Ka-Band," in IEEE Electron Device Lett., 41(8), 1181 (2020)

work page 2020
[9]

FerroHEMTs: High -Current and High -Speed All -Epitaxial AlScN/GaN Ferroelectric Transistors,

J. Casamento et al., "FerroHEMTs: High -Current and High -Speed All -Epitaxial AlScN/GaN Ferroelectric Transistors," in 2022 International Electron Devices Meeting (IEDM), 111.1.1 (2022)

work page 2022
[10]

Strain-Balanced AlScN/GaN HEMTs with fT/fMAX of 173/321 GHz,

T. S. Nguyen, K. Nomoto, W. Zhao, C. Savant, H. G. Xing and D. Jena, "Strain-Balanced AlScN/GaN HEMTs with fT/fMAX of 173/321 GHz," in 2024 IEEE International Electron Devices Meeting (IEDM), 1 (2024),

work page 2024
[11]

AlScN/GaN HEMTs Grown by Metal-Organic Chemical Vapor Deposition With 8.4 W/mm Output Power and 48 % Power-Added Efficiency at 30 GHz,

S. Krause, I. Streicher, P. Waltereit, L. Kirste, P. Brückner and S. Leone, "AlScN/GaN HEMTs Grown by Metal-Organic Chemical Vapor Deposition With 8.4 W/mm Output Power and 48 % Power-Added Efficiency at 30 GHz," IEEE Electron Device Lett., 44(1), 17 (2023)

work page 2023
[12]

Voltage-margin limiting mechanisms of AlScN -based HEMTs

P. Döring, S. Krause, P. Waltereit, et al., “ Voltage-margin limiting mechanisms of AlScN -based HEMTs” Appl. Phys. Lett., 123(3), 032101 (2023)

work page 2023
[13]

ScAlN/GaN high-electron-mobility transistors with 2.4-A/mm current density and 0.67-S/mm transconductance,

A. J. Green, J. K. Gillespie, R. C. Fitch, et al. , “ScAlN/GaN high-electron-mobility transistors with 2.4-A/mm current density and 0.67-S/mm transconductance,” in IEEE Electron Device Lett., 40(7), 1056 (2019)

work page 2019
[14]

Progress in Near-Equilibrium Ammonothermal (NEAT) growth of GaN substrates for GaN-on-GaN semiconductor devices,

T. Hashimoto, E. R. Letts and D. Key, "Progress in Near-Equilibrium Ammonothermal (NEAT) growth of GaN substrates for GaN-on-GaN semiconductor devices," Crystals, 12, 8 (2022)

work page 2022
[15]

Metalorganic Chemical Vapor Deposition of AlScN Thin Films and AlScN/AlN/GaN Heterostructures

V. G. T. Vangipuram, A. Mukit, K. Zhang, S. Salmani -Rezaie, and H. Zhao, “Metalorganic Chemical Vapor Deposition of AlScN Thin Films and AlScN/AlN/GaN Heterostructures”, ACS Crys. Growth Des. 26 (2), 8 45 (2026)

work page 2026
[16]

High -mobility window for two -dimensional electron gases at ultrathin AlN∕GaN heterojunctions,

Y. Cao and D. Jena, “High -mobility window for two -dimensional electron gases at ultrathin AlN∕GaN heterojunctions,” Appl. Phys. Lett. 90(18), 182112 (2007)

work page 2007
[17]

The impact of post gate annealing on microwave noise performance of AlGaN/GaN HEMTs

D. Liu, J. Lee, and W. Lu, “The impact of post gate annealing on microwave noise performance of AlGaN/GaN HEMTs”, in Solid-State Electronics, 51(1), 90 (2007)

work page 2007
[18]

Investigation of High Frequency Noise and Power in AlGaN/GaN HEMTs

P. Sakalas, M. Schroter, H. Xing, et al., “ Investigation of High Frequency Noise and Power in AlGaN/GaN HEMTs” in AIP Conf. Proc. 922(1), 171 (2007)

work page 2007
[19]

High power density gallium nitride radio frequency transistors via enhanced nucleation in heteroepitaxy,

H. Zhou, C. Zhang, K. Zhang, et al., “High power density gallium nitride radio frequency transistors via enhanced nucleation in heteroepitaxy,” Nat Commun, vol. 17, no. 1, p. 556, Dec. 2025. [S1] E. H. Nicollian and J. R. Brews, MOS (metal oxide semiconductor) physics and technology (Wiley, New York, 1982), p. 176

work page 2025