Pith. sign in

REVIEW 3 major objections 4 minor 2 references

Two-step growth of (In,Ga)N pseudo-substrates on GaN templates by plasma-assisted molecular beam epitaxy

T0 review · 3 major / 4 minor · reviewed 2026-08-01 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read A two-step PAMBE growth protocol produces relaxed (In,Ga)N pseudo-substrates with in-plane lattice constant ≈3.26 Å, smooth pit-free surfaces, and narrow photoluminescence linewidths, making them attractive for red-emitting (In,Ga)N LEDs.

desk verdict Two-step PAMBE gives smooth, relaxed (In,Ga)N pseudo-substrates with the right lattice constant, but the paper's key comparative claim is undercut by a cross-system control. read the letter →

arxiv 2607.15748 v1 pith:D2WMJ5C7 submitted 2026-07-17 cond-mat.mtrl-sci physics.app-ph

Huaide Zhang , Jingxuan Kang , Aidan F. Campbell , Jonas Lähnemann , Oliver Brandt , Lutz Geelhaar This is my paper · ORCID
classification cond-mat.mtrl-sciphysics.app-ph PACS 81.15.Hi78.55.Cr61.05.cp
keywords molecularbeamepitaxy(InGa)Npseudo-substratestrainrelaxationlatticemismatchredLEDsX-raydiffractionphotoluminescence
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

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

The reading

This paper tries to establish that a simple two-step molecular beam epitaxy recipe, entirely carried out in vacuum, can produce relaxed (In,Ga)N pseudo-substrates on GaN templates. The first step grows a rough, pitted layer under nitrogen-rich conditions to relieve strain, and the second step overgrows it under metal-rich conditions to smooth the surface. The authors report an in-plane lattice constant of about 3.26 Å, roughly 70–75% strain relaxation, and photoluminescence linewidths of 117–132 meV, which indicate good compositional uniformity. If correct, this offers a scalable route to substrates that could improve the efficiency of red (In,Ga)N LEDs by reducing the strain and polarization fields in quantum wells.

What carries the argument

The central mechanism is the two-step growth protocol: first grow (In,Ga)N under N-stable (metal-lean) conditions at about 550 °C to create a rough, pit-covered layer that relaxes strain plastically; then switch to metal-stable (Ga+In-rich) conditions to grow a smooth over-layer. The rough intermediate surface is expected to bend threading dislocations and promote their annihilation, following earlier demonstrations for GaN, and the relaxation state of the first layer appears to be imprinted on the second, giving a coherent in-plane lattice constant across the stack.

What would settle it

Grow the two-layer sample (#A-L1+2) and the direct single-layer sample (#A-L2) in the same MBE chamber under identical fluxes and compare their in-plane lattice constants from reciprocal space maps; if the difference disappears, the two-step protocol is not the cause. Alternatively, cross-sectional TEM at the L1/L2 interface would show whether dislocations actually bend and annihilate as assumed.

Watch

Extended reading notes

Core claim

The paper claims that a two-step plasma-assisted molecular beam epitaxy protocol—first N-stable, then metal-stable—produces (In,Ga)N pseudo-substrates with the lattice constant needed for red LEDs. The rough first layer relaxes elastically and plastically, and the smooth second layer inherits a larger in-plane lattice constant than a layer grown directly on GaN. XRD gives an in-plane lattice constant of ≈3.26 Å for the two-layer stack, and room-temperature PL shows a narrow linewidth (117 meV at 543 nm, 132 meV at 587 nm), which the authors interpret as excellent compositional homogeneity. They conclude that these layers are attractive as pseudo-substrates for red-emitting (In,Ga)N LEDs and

Load-bearing premise

The claim that the two-step protocol gives the decisive advantage relies on comparing samples grown in two different MBE systems whose nitrogen fluxes differ by about 13%, and on the assumption that dislocation bending and annihilation observed for GaN also occur in (In,Ga)N; if either premise fails, the relaxation gain attributed to the rough intermediate layer may not hold.

Editorial extensions

If this is right

  • Relaxed (In,Ga)N pseudo-substrates with in-plane lattice constants of 3.26–3.27 Å can be obtained on standard GaN templates without porous layers, wafer bonding, or piezoelectric buffer layers.
  • Red LEDs grown on such pseudo-substrates should experience lower compressive strain and smaller polarization fields in the quantum wells, improving indium incorporation and internal quantum efficiency.
  • The protocol is robust to the indium content of the first layer: using 0.3 or 0.4 indium fraction in layer L1 both gave in-plane lattice constants in the desired range, suggesting tunability.
  • The achieved surface roughness (≈2.5 nm RMS), high relaxation degree (75%), and narrow PL linewidth compare favorably with other reported pseudo-substrate approaches.
  • Because the entire fabrication is done by molecular beam epitaxy, it is scalable and eliminates time-consuming ex-situ processing steps.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • If the dislocation-bending mechanism demonstrated for GaN transfers to (In,Ga)N, the two-step protocol should reduce threading dislocation density in the upper layer; direct cross-sectional TEM or etch-pit counting would confirm this and could further boost LED performance.
  • The apparent link between higher relaxation degree and narrower PL linewidth suggests a testable extension: systematically varying the thickness or composition of layer L2 could quantify how relaxation improves compositional homogeneity.
  • Repeating the #A-L1+2 and #A-L2 growths in the same MBE chamber under identical fluxes would cleanly separate the protocol's effect from the 13% difference in nitrogen flux between the two systems used here.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

3 major / 4 minor

Summary. The manuscript reports a two-step PAMBE protocol for growing relaxed (In,Ga)N pseudo-substrates on GaN templates. The first step is N-stable growth of rough (In,Ga)N-L1, followed by metal-stable growth of smooth (In,Ga)N-L2. For the two-step sample #A-L1+2 the authors report an in-plane lattice constant of about 3.260 Å, In content 0.28–0.30, relaxation degree 70–75%, RMS roughness about 2.5 nm, and a narrow RT PL line at 543 nm with FWHM 28 nm (117 meV). A second two-step sample with higher In in L1 (#B-L1+2) gives a≈3.27 Å and PL at 587 nm (FWHM 132 meV). The authors conclude that the two-step approach yields smooth, compositionally homogeneous, relaxed (In,Ga)N layers suitable as pseudo-substrates for red μ-LEDs, without external processing. A key comparative claim is that L2 relaxes more when grown on L1 than directly on GaN (the latter, #A-L2, has R≈50% and a≈3.235 Å).

Significance. If the two-step protocol is as effective as claimed, it provides a simple, scalable, all-MBE route to (In,Ga)N pseudo-substrates with lattice constants in the 3.24–3.27 Å range desired for red-LED active regions. The reported absolute properties of the two-step samples—particularly the narrow PL linewidths, smooth morphologies, and high relaxation degrees—are valuable and are documented with standard characterization (RHEED, SEM/AFM, XRD RSM, PL). The work also shows robustness of the approach when the L1 composition is changed. The main weakness is that the 'decisive advantage' of the two-step protocol over direct L2 growth is not established by a controlled experiment: the comparison crosses two MBE systems with different N fluxes, and the dislocation-reduction mechanism is only inferred by analogy. These issues are fixable with additional control growth or more cautious claims.

major comments (3)
  1. [Sections II and III (comparative claim)] The central comparative claim—that L2 relaxes more on L1 than directly on GaN—rests on comparing #A-L1+2 (MBE system I, N flux 6.0×10^14 s^-1cm^-2) with #A-L2 (MBE system II, N flux 6.8×10^14 s^-1cm^-2). Under metal-stable growth with (Ga+In)/N=1.2, the growth rate is N-limited, so the L2 in #A-L2 grows roughly 13% faster (and thicker) than the L2 in #A-L1+2. Since strain relaxation in (In,Ga)N depends on thickness and kinetic conditions, the observed difference in relaxation degree (50±10% vs 70±5%/75±5%) could be caused by the system/flux change rather than by the presence of L1. Moreover, the L1 reference #A-L1 was also grown in system II, so the L1 comparison is similarly confounded. A same-system direct-L2 control (and ideally a two-step sample in system II) is needed to substantiate the 'decisive advantage' claim. Without it, the conclusion should be limited to the absolute propert
  2. [Section III / Fig. 3(b) (RSM deconvolution)] The RSM of #A-L1+2 contains overlapping contributions from L1 and L2; the authors state that these contributions 'cannot easily be deconvoluted.' Nevertheless, the conclusion that L2 has a higher relaxation degree than #A-L2 is drawn from an 'additional feature labelled L2' being closer to the fully relaxed line. Since L1 and L2 in sample #A have nominally the same In content (0.30) and the same measured in-plane lattice constant (same Qx), the L1 and L2 contributions have the same relaxation degree by construction; the feature cannot be unambiguously assigned to L2 alone. A quantitative decomposition, or a deliberately grown L2-only sample on an L1 template in the same MBE system, is required to support the claim that layer L2 itself relaxes more than direct L2 growth.
  3. [Section III (dislocation mechanism)] The mechanism invoked for the benefit of the rough L1 layer—bending and annihilation of dislocations, as reported for GaN (refs 17, 18)—is not directly demonstrated in this work. The text says 'We expect similar phenomena for our samples' and uses schematic black lines in Fig. 1, but no dislocation-density measurements (TEM, etch pit density, or XRD peak-width analysis) are presented. If the two-step protocol is claimed to reduce threading dislocation density in L2, this needs direct evidence; otherwise the discussion should be clearly marked as a hypothesis that motivates the growth protocol.
minor comments (4)
  1. [Section III, caption of Fig. 3] The text says 'The PL spectra acquired at RT are shown in Fig. 3(d)' but Fig. 3(d) is a 2θ-ω scan; the PL spectra are in Fig. 3(f). Please correct the cross-reference.
  2. [Section III, paragraph on PL] Typo: 'sample A#-L1+2' should read 'sample #A-L1+2'.
  3. [Section III, thickness statement] The thickness values (L1 ≈200 nm, L2 ≈500 nm) are stated to come from 'cross-sectional micrographs... (not shown)'. Since thickness enters the relaxation discussion, please include the micrographs or a supplementary figure.
  4. [General] The numerical values for fluxes are written as '6 × 1014' in the text; use proper superscript formatting (6 × 10^14) consistently.

Circularity Check

0 steps flagged · score 0.0 of 10

No circular reasoning found: the paper is an experimental demonstration with externally calibrated measurements and no fitted-parameter predictions.

full rationale

The paper's central claims—that the two-step N-stable/metal-stable PAMBE protocol yields smooth, compositionally homogeneous (In,Ga)N pseudo-substrates with in-plane lattice constants of about 3.26–3.27 Å—are supported by direct measurements. XRD lattice constants, In contents, and relaxation degrees are extracted from 2θ-ω scans and reciprocal-space maps using standard elastic constants and reference peak positions of GaN; no parameter is fitted to the target result and then renamed a prediction. The PL linewidth is used as a qualitative indicator of compositional homogeneity, not to back-calculate composition. The strain-relaxation comparison between #A-L1+2 and #A-L2 may be experimentally confounded by the fact that these two samples were grown in different MBE systems with a 13% difference in N flux, but this is a question of experimental control and validity, not circularity. The only self-citations (e.g., ref. 23, ref. 22) are used for corroborating surface reconstructions, roughness trends, and growth morphology; they are not the load-bearing derivation of the main result, which rests on the reported XRD and microscopy data. The dislocation-bending mechanism is borrowed from external GaN studies (refs. 17, 18) and is explicitly labeled an expectation rather than an established result for (In,Ga)N. No equation in the paper defines an output in terms of the quantity it is supposed to predict, and no fit is presented as an independent prediction. The derivation chain is therefore self-contained with respect to the paper's own inputs.

Assumptions & free parameters 0 free parameters · 5 assumptions · 0 invented entities

The central claims rest on standard III-nitride characterization assumptions and on the unverified transferability of dislocation-bending from GaN to (In,Ga)N; no free parameters are fit in this experimental study and no new physical entities are introduced.

assumptions (5)
  • domain assumption XRD analysis of In content and relaxation degree relies on Vegard's law and standard III-nitride elastic constants.
    Used in §III to convert measured lattice constants to In content x and relaxation R; systematic errors in these constants propagate to the reported x and R values.
  • domain assumption A rough intermediate surface reduces threading dislocation density in the overgrown layer via bending and annihilation, as reported for GaN growth.
    Invoked in §III ('We expect similar phenomena for our samples') and in the schematic; not directly measured in (In,Ga)N here.
  • domain assumption PL linewidth narrowing is attributed to improved compositional homogeneity, assuming no other dominant broadening mechanisms.
    Used in §III to infer 'excellent compositional homogeneity' from a 117 meV FWHM; carrier localization and other broadening are not excluded.
  • domain assumption PL emission from the two-layer sample #A-L1+2 originates entirely from the upper layer L2 because ~500 nm L2 absorbs all 325 nm excitation.
    Stated in §III ('all exciting laser light should be absorbed in this layer L2'); absorption coefficient not measured.
  • ad hoc to paper Growth runs in MBE systems I and II are comparable apart from the stated N-flux difference, so the relaxation difference between #A-L1+2 and #A-L2 is attributed to the two-step protocol.
    The key comparison crosses two MBE systems with N fluxes 6.0 vs 6.8 ×10^14 s^-1cm^-2; the paper does not correct for chamber-to-chamber variation.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Two-step growth of (In,Ga)N pseudo-substrates on GaN templates by plasma-assisted molecular beam epitaxy." pith.science (2026). https://pith.science/paper/D2WMJ5C7

@misc{pith2026260715748,
  author       = {Pith},
  title        = {Pith review of: Two-step growth of (In,Ga)N pseudo-substrates on GaN templates by plasma-assisted molecular beam epitaxy},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/D2WMJ5C7}},
  note         = {Machine review of arXiv:2607.15748}
}
abstract

(In,Ga)N layers are grown by plasma-assisted molecular beam epitaxy on GaN templates. We introduce a two-step protocol that involves switching the growth conditions from initially N-stable to metal-stable. Reflection high-energy electron diffraction as well as scanning electron and atomic force microscopy reveal that the first step results in a rough intermediate surface with open pits, whereas the final surface is smooth. The narrow linewidth of the photoluminescence band indicates an excellent compositional homogeneity of the upper layer. Its in-plane lattice constant is determined to be $\approx$3.26 \AA from X-ray diffraction measurements. This combination of favorable properties makes these layers attractive as pseudo-substrates for the growth of red-emitting (In,Ga)N light-emitting diodes. In particular, the approach presented here does not require any complex external processing and is, thus, scalable and economical.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

2 extracted references

  1. [1]

    Future trends of display technology: micro- LEDs toward transparent, free -form, and near- eye displays,

    1 T.S. Kim, J.-E. Ryu, J. Park, R.-J. Liu, J. Choi, J. Kim, Y.J. Hong, D. -H. Kim, and J. Shin, “Future trends of display technology: micro- LEDs toward transparent, free -form, and near- eye displays,” Light Sci. Appl. 14(1), 335 (2025). 2 M. Sheen, Y. Ko, D. Kim, J. Kim, J. Byun, Y. Choi, J. Ha, K.Y. Yeon, D. Kim, J. Jung, J. Choi, R. Kim, J. Yoo, I. Ki...

    2025

  2. [2018]

    Investigation of Statistical Broadening in InGaN Alloys,

    36 D.S. Arteev, A.V. Sakharov, E.E. Zavarin, W.V. Lundin, A.N. Smirnov, V.Y. Davydov, M.A. Yagovkina, S.O. Usov, and A.F. Tsatsulnikov, “Investigation of Statistical Broadening in InGaN Alloys,” J. Phys. Conf. Ser. 1135, 012050 (2018)

    2018

Pith tools

Reviewed August 1, 2026 · model on record in the stance chip above.