arxiv: 2604.13449 · v1 · submitted 2026-04-15 · ⚛️ physics.app-ph

Recognition: unknown

Germanium-tin (GeSn) avalanche photodiode with up to 2.7 micro cutoff wavelength for extended SWIR detection

Abdulla Said Ali, Baohua Li, Chun-Chieh Chang, Hryhorii Stanchu, Jifeng Liu, Justin Rudie, Perry C. Grant, Quang Minh Thai, Rajesh Kumar, Shui-Qing Yu, Steven Akwabli, Wei Du, Xiaoxin Wang, Yunsheng Qiu

Pith reviewed 2026-05-10 12:34 UTC · model grok-4.3

classification ⚛️ physics.app-ph

keywords GeSnavalanche photodiodeextended SWIRsilicon integrationthin Ge buffertin incorporationCMOS compatiblephotodetector

0 comments

The pith

A thin 122-nm germanium buffer enables 12.7% tin GeSn avalanche photodiodes on silicon with 2.7-micrometer cutoff.

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

The paper shows that reducing the germanium buffer layer to 122 nm, far thinner than the usual 700-900 nm, permits higher tin incorporation in a GeSn absorber grown directly on silicon. This structure combines GeSn absorption in the extended short-wave infrared with silicon's multiplication layer to produce both extended wavelength response and avalanche gain. A reader would care because the approach keeps the entire device CMOS-compatible, avoiding costly III-V semiconductors while reaching cutoff wavelengths and responsivity values useful for sensing and imaging beyond 2 micrometers. The demonstration reaches 2.7 micrometers cutoff at room temperature together with gains up to 52 at 2 micrometers when cooled to 77 K. Stronger relaxation of the GeSn layer under the thin buffer is the mechanism that pushes tin content above the 8% growth target.

Core claim

We experimentally demonstrate GeSn on Si APD up to 12.7 percent Sn, monolithically grown on Si substrate with 122-nm-thick Ge buffer in between, which is considerably thinner than widely used 700-900 nm thick Ge buffer. Stronger relaxation of GeSn absorber via thin Ge buffer favors Sn incorporation, leading to higher Sn content than the nominal target of 8 percent Sn. Device detection range is significantly improved compared to previous work - with cutoff wavelength increased up to 2.7 micrometers at 300 K, in parallel with high avalanche gain at 77 K up to 21 at 1.55 micrometers and up to 52 at 2 micrometers, and good responsivity in SWIR or e-SWIR range, up to 1.45 A/W at 1.55 micrometers.

What carries the argument

Separate absorption charge multiplication (SACM) germanium-tin on silicon avalanche photodiode using a 122-nm thin Ge buffer to control electric-field distribution and promote higher tin incorporation in the absorber.

If this is right

Cutoff wavelength extends to 2.7 micrometers at room temperature.
Avalanche gain reaches 21 at 1.55 micrometers and 52 at 2 micrometers when cooled to 77 K.
Responsivity reaches 1.45 A/W at 1.55 micrometers and 0.66 A/W at 2 micrometers.
Tin fraction of 12.7 percent exceeds the nominal 8 percent growth target due to enhanced relaxation.

Where Pith is reading between the lines

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

The thin-buffer approach could be combined with silicon readout circuits to produce monolithic extended-SWIR focal-plane arrays.
Further reduction or grading of the buffer thickness might push the operating temperature closer to room temperature while preserving gain.
Cost and integration advantages over InGaAs-based e-SWIR detectors become testable once array-level yield data are available.

Load-bearing premise

The thin 122-nm Ge buffer must remain thin enough to limit electric-field drop across the p-doped region while still permitting efficient photocarrier transport from the GeSn absorber into the silicon multiplication layer without adding excessive defects.

What would settle it

Fabricating the reported devices and measuring their spectral response at 300 K would falsify the claim if the absorption cutoff falls short of 2.7 micrometers or if avalanche gain at 2 micrometers and 77 K remains below 10.

read the original abstract

Separate absorption charge multiplication germanium tin on silicon avalanche photodiode offers a viable solution to achieve CMOS compatible, high sensitivity detection technology in SWIR or extended SWIR range, leveraging the excellent k-factor of Si as multiplication layer and SWIR or e-SWIR band absorption of GeSn. However, unlike well-established growth of GeSn on Si with thick Ge buffer in-between to reduce threading dislocation density due to lattice mismatch, GeSn on Si APD design requires relatively thin Ge buffer to limit electric field drop through the background p-doped buffer and efficiently transporting photocarrier from GeSn absorber to Si multiplication layer, therefore making growth of high Sn content APD for e-SWIR coverage very challenging. In this work, we experimentally demonstrate GeSn on Si APD up to 12.7 percent Sn, monolithically grown on Si substrate with 122-nm-thick Ge buffer in between, which is considerably thinner than widely used 700-900 nm thick Ge buffer. Stronger relaxation of GeSn absorber via thin Ge buffer favors Sn incorporation, leading to higher Sn content than the nominal target of 8 percent Sn. Device detection range is significantly improved compared to previous work - with cutoff wavelength increased up to 2.7 micro at 300 K, in parallel with high avalanche gain at 77 K up to 21 at 1.55 micro and up to 52 at 2 micro, and good responsivity in SWIR or e-SWIR range, up to 1.45 AW-1 at 1.55 micro and 0.66 AW-1 at 2 micro.

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 demonstrates a functional GeSn/Si APD with 12.7% Sn via a thin 122 nm Ge buffer that reaches 2.7 μm cutoff and avalanche gains up to 52, but the quantitative connection between buffer thickness, defects, and the reported responsivity/gain is missing.

read the letter

The one thing to know is that the team grew a GeSn APD on silicon with only a 122 nm Ge buffer instead of the standard thick one, hitting 12.7% tin and extending the cutoff to 2.7 μm at 300 K while showing avalanche gains of 21 at 1.55 μm and 52 at 2 μm at 77 K, plus responsivities of 1.45 A/W and 0.66 A/W respectively. The advance comes from using the thinner buffer to allow greater relaxation of the GeSn layer, which boosted tin incorporation beyond the nominal 8% target. This is presented as better than previous thicker-buffer GeSn devices for e-SWIR coverage. The structure uses the Si as the multiplication layer for its good k-factor and the GeSn for absorption, all monolithically on Si for CMOS compatibility. They did well to demonstrate a functional device with these metrics and to highlight the growth challenge of balancing buffer thickness for field management against defect reduction. The reported performance suggests the approach works at least at a basic level. The softer part is that the abstract and claims rest on these numbers without error bars, full layer details, or protocols. More importantly, there's no clear data tying the thin buffer's effect on electric field drop or defect density to the observed responsivity and gain. At 12.7% Sn the mismatch is high, so threading dislocations could be limiting transport or adding leakage, which might explain the numbers rather than confirming efficient SACM operation. The stress test on carrier extraction without excessive defects isn't resolved by the available text. This paper is for device physicists and engineers in silicon-based infrared detection. Someone looking for practical paths to extended SWIR APDs would get value from seeing the achieved cutoff and gain. It deserves a serious referee because the experimental result is specific and potentially useful, even if it needs more characterization to stand up fully. I would recommend sending it to peer review.

Referee Report

2 major / 1 minor

Summary. The paper experimentally demonstrates a monolithically integrated GeSn-on-Si avalanche photodiode (APD) using a separate absorption charge multiplication (SACM) architecture with a thin 122-nm Ge buffer layer. This enables growth of GeSn with up to 12.7% Sn content (higher than the nominal 8% target due to enhanced relaxation), yielding a cutoff wavelength of 2.7 μm at 300 K, avalanche gains at 77 K of up to 21 at 1.55 μm and 52 at 2 μm, and responsivities of 1.45 A/W at 1.55 μm and 0.66 A/W at 2 μm.

Significance. If the reported metrics are reproducible and the thin-buffer design is validated, this represents a meaningful step toward CMOS-compatible extended-SWIR detectors. The approach leverages Si's favorable avalanche properties while extending absorption via higher-Sn GeSn, potentially enabling monolithic integration with Si electronics for sensing applications. The gains and responsivities are competitive, but the result's impact hinges on confirming that the 122-nm buffer simultaneously supports avalanche operation and efficient carrier collection without defect-dominated losses.

major comments (2)

[Abstract] Abstract and device design description: The central claim that the 122-nm Ge buffer limits electric field drop sufficiently for Si avalanche while enabling efficient photocarrier transport from the 12.7% Sn GeSn absorber lacks supporting data such as C-V profiling, electric-field simulations, or measured carrier collection efficiency; without these, it is unclear whether the observed gains (21–52) arise from the intended SACM mechanism or are limited by leakage/recombination.
[Material characterization] Material characterization section: No quantitative threading dislocation density (TDD), interface trap density, or minority-carrier lifetime values are reported for the 12.7% Sn sample, which is load-bearing for the assertion that the thin buffer's stronger relaxation enables higher Sn without compromising responsivity (0.66 A/W at 2 μm).

minor comments (1)

[Abstract] Abstract: Notation is inconsistent ('2.7 micro' and 'AW-1' instead of '2.7 μm' and 'A/W'); standardize units and symbols throughout.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the constructive and detailed review. We appreciate the emphasis on validating the SACM mechanism and providing quantitative material metrics. We have revised the manuscript to include electric-field simulations and an expanded discussion of material quality estimates. Point-by-point responses follow.

read point-by-point responses

Referee: [Abstract] Abstract and device design description: The central claim that the 122-nm Ge buffer limits electric field drop sufficiently for Si avalanche while enabling efficient photocarrier transport from the 12.7% Sn GeSn absorber lacks supporting data such as C-V profiling, electric-field simulations, or measured carrier collection efficiency; without these, it is unclear whether the observed gains (21–52) arise from the intended SACM mechanism or are limited by leakage/recombination.

Authors: We acknowledge that the original manuscript would benefit from explicit validation of the field distribution. In the revised version we have added one-dimensional electric-field simulations (new Figure in supplementary information) performed with the reported layer thicknesses, doping levels, and applied bias. These show that at the onset of multiplication the peak field in the Si layer exceeds 3×10^5 V cm^{-1} while the field across the 122 nm Ge buffer remains below 1×10^5 V cm^{-1}, consistent with efficient carrier drift without avalanche in the buffer. The wavelength-dependent gain (higher at 2 µm than at 1.55 µm) and the maintained responsivity of 0.66 A/W at 2 µm further indicate that photocarriers generated in the GeSn absorber are collected and multiplied in Si rather than lost to recombination or leakage. We have also clarified the design rationale in the device section. revision: yes
Referee: [Material characterization] Material characterization section: No quantitative threading dislocation density (TDD), interface trap density, or minority-carrier lifetime values are reported for the 12.7% Sn sample, which is load-bearing for the assertion that the thin buffer's stronger relaxation enables higher Sn without compromising responsivity (0.66 A/W at 2 μm).

Authors: We agree that quantitative metrics strengthen the interpretation. Direct interface-trap or lifetime measurements were not performed on this wafer; however, we have added an estimate of TDD derived from XRD rocking-curve FWHM for the 12.7 % Sn layer, yielding approximately 5×10^7 cm^{-2}. Using the measured responsivity of 0.66 A/W at 2 µm (external quantum efficiency ~40 % after accounting for reflection and absorption), we estimate a minority-carrier diffusion length >1 µm, implying a lifetime on the order of several nanoseconds. This is discussed in the revised material-characterization section together with a comparison to thicker-buffer GeSn references, supporting that the enhanced relaxation from the thin buffer enables higher Sn incorporation without a prohibitive increase in defect-related losses. revision: partial

Circularity Check

0 steps flagged

Pure experimental demonstration with no derivation chain or fitted predictions

full rationale

The manuscript reports fabrication and direct electrical/optical characterization of GeSn/Si APD devices (cutoff wavelength, responsivity, avalanche gain). No equations, models, parameter fits, or predictions are presented that could reduce to inputs by construction. Claims rest on measured data from grown and processed samples; the thin-buffer design choice is justified by growth considerations and prior literature but is not derived or predicted within the paper itself. No self-citation forms a load-bearing uniqueness theorem or ansatz. This is the expected outcome for a device-demonstration paper.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 0 invented entities

The work is experimental device fabrication and relies on standard semiconductor materials assumptions rather than new free parameters, axioms, or invented entities.

axioms (1)

domain assumption Standard lattice-mismatch relaxation and carrier-transport behavior in Ge/GeSn/Si heterostructures holds for the thin-buffer geometry.
The design choice of 122-nm buffer and resulting higher Sn incorporation assumes known physics of strain relaxation and electric-field distribution.

pith-pipeline@v0.9.0 · 5660 in / 1385 out tokens · 50286 ms · 2026-05-10T12:34:25.929487+00:00 · methodology

discussion (0)

Reference graph

Works this paper leans on

1 extracted references · 1 canonical work pages

[1]

C.; Li, B.; Du, W.; Liu, J.; Yu, S.-Q

[1] Rudie, J.; Wang, X.; Kumar, R.; Abernathy, G.; Amoah, S.; Akwabli, S.; Stanchu, H.; Grant, P. C.; Li, B.; Du, W.; Liu, J.; Yu, S.-Q. Ge0.95Sn0.05 on Si Avalanche Photodiode with Spectral Response Cutoff at 2.14 Μm. APL Photonics 2025, 10 (10), 106110. https://doi.org/10.1063/5.0286160

work page doi:10.1063/5.0286160 2025