pith. sign in

arxiv: 2604.23394 · v1 · submitted 2026-04-25 · ⚛️ physics.optics

Moth's eye-inspired perfectly vertical subwavelength grating coupler for silicon photonics

Ivan A. Kazakov , Ilona Popova , Arkady Shipulin This is my paper

Pith reviewed 2026-05-08 07:43 UTC · model grok-4.3

classification ⚛️ physics.optics
keywords grating couplersilicon photonicssubwavelength gratingvertical couplinganisotropyunidirectionalitymoth eyeVCSEL integration
0
0 comments X

The pith

Anisotropy in grating stripes directs vertically incident light unidirectionally into silicon chips.

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

This paper proposes a bio-inspired design for perfectly vertical grating couplers on silicon-on-insulator that introduces anisotropy to the grating stripes. The anisotropy breaks symmetry so light from a vertical fiber couples preferentially in one direction along the chip. The approach requires only one etching step and targets 1550 nm operation for easy integration of vertical-cavity lasers. Simulations predict 41 percent coupling efficiency, more than 10 dB unidirectionality, and 50 nm bandwidth; fabricated devices confirm 12.8 dB unidirectionality.

Core claim

By adding anisotropy to the grating stripes, the coupler directs light to one side at vertical incidence, yielding 41 percent in-coupling efficiency from an SMF-28 fiber, over 10 dB unidirectionality, and a 50 nm 1 dB bandwidth in simulation on the SOI platform; experiments measured 12.80 dB unidirectionality and 8.35 dB insertion loss near 1528 nm.

What carries the argument

Anisotropic grating stripes that break left-right symmetry to direct light unidirectionally while permitting single-step etching.

If this is right

  • Fabrication reduces to a single etch step on standard SOI wafers.
  • The design supports direct vertical placement of VCSELs for on-chip light sources.
  • Unidirectionality above 10 dB reduces unwanted back-reflection into the fiber.
  • The 50 nm bandwidth at the 1 dB level covers typical C-band operation.
  • Experimental results already confirm strong unidirectionality in real devices.

Where Pith is reading between the lines

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

  • The same anisotropy principle could be tested on other photonic platforms such as silicon nitride for different wavelengths.
  • Higher simulated efficiencies might be reached by combining this stripe shape with existing apodization techniques.
  • Direct VCSEL integration would allow compact transmitters for LIDAR or quantum communication without fiber-angle adjustments.

Load-bearing premise

The anisotropy in the grating stripes will reliably steer light in one direction in actual fabricated devices without large back-reflections or efficiency drops that simulations miss.

What would settle it

Fabricating the coupler and measuring coupling efficiency below 25 percent or unidirectionality under 5 dB at exact vertical incidence would show the anisotropy does not deliver the claimed performance.

Figures

Figures reproduced from arXiv: 2604.23394 by Arkady Shipulin, Ilona Popova, Ivan A. Kazakov.

Figure 1
Figure 1. Figure 1: (a) Side view of the proposed grating, which includes a sub-wavelength grating (SWG) view at source ↗
Figure 2
Figure 2. Figure 2: Approximated refractive index of the sub-wavelength grating for different fill-factors, view at source ↗
Figure 3
Figure 3. Figure 3: Cross section of the FDTD simulation setup view at source ↗
Figure 4
Figure 4. Figure 4: Coupling efficiency (blue, left axis) and unidirectionality (orange, right axis) as a function view at source ↗
Figure 5
Figure 5. Figure 5: Distribution of Ey for the optimized grating. Black lines indicate silicon and blue lines indicate subwavelength grating regions (a) (b) view at source ↗
Figure 6
Figure 6. Figure 6: Experimental setup for ME-vGC characterization on the probe station: (a) top view and view at source ↗
Figure 7
Figure 7. Figure 7: Measured transmission of a single ME-vGC (blue, left axis) and unidirectionality (orange, view at source ↗
read the original abstract

We propose a novel bio-inspired design principle for the perfectly vertical grating coupler. The main idea of our design is to introduce anisotropy to the grating stripe to direct the light to one side of the grating. This grating design is easy to manufacture, only requiring a single etching step, and it is designed to efficiently couple vertically incident light. This makes it a good candidate for heterogeneous integration of light sources, especially VCSELs, on chip for applications in classical and quantum communications, LIDARs, sensing systems, and others. The grating coupler was designed for the SOI material platform with a central wavelength of 1550 nm. We obtained the efficiency of in-coupling from the SMF-28 fiber of 41% at vertical incidence and unidirectionality of over 10 dB, with a bandwidth of 50 nm at a 1 dB level in simulation. Experimental measurements confirmed unidirectionality, with observed unidirectionality of 12.80+-0.02 dB and a single-coupler insertion loss of 8.35+-0.02 dB around 1528 nm.

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.

Referee Report

2 major / 2 minor

Summary. The paper proposes a moth's eye-inspired anisotropic subwavelength grating coupler for perfectly vertical incidence in silicon-on-insulator photonics. By introducing anisotropy to the grating stripes, the design achieves simulated in-coupling efficiency of 41% from an SMF-28 fiber at 1550 nm, unidirectionality exceeding 10 dB, and a 50 nm 1 dB bandwidth using only a single etch step. Experimental results confirm unidirectionality of 12.80 dB but report a single-coupler insertion loss of 8.35 dB near 1528 nm.

Significance. If the simulated efficiency can be realized in fabricated devices, the approach would provide a simple, single-etch solution for vertical grating couplers, facilitating heterogeneous integration of VCSELs and other sources for classical/quantum communications, LIDAR, and sensing. The bio-inspired anisotropy principle for enforcing unidirectionality without additional etches or reflectors represents a potentially useful design heuristic for subwavelength gratings.

major comments (2)
  1. [Abstract / Experimental Results] Abstract and experimental measurements: The simulated in-coupling efficiency of 41% at vertical incidence is not reproduced experimentally, where the reported single-coupler insertion loss of 8.35 dB implies only ~14.6% efficiency—a gap exceeding 4 dB. The manuscript provides no loss budget, fabrication tolerance analysis, or calibrated absolute power measurements to identify whether the discrepancy arises from etch-depth variation, sidewall angle, subwavelength feature fidelity, or unmodeled scattering in the anisotropic stripes; this directly affects the central claim that the anisotropy enables efficient vertical coupling.
  2. [Design Principle] Design section: The anisotropy ratio and specific stripe geometry are listed as free parameters, yet no quantitative sensitivity analysis or Monte-Carlo tolerance study is presented to show that the >10 dB unidirectionality and 41% efficiency remain robust under realistic fabrication variations (e.g., ±10 nm critical-dimension error). Without this, the translation from simulation to experiment remains an unverified assumption.
minor comments (2)
  1. [Abstract] The abstract states a central wavelength of 1550 nm but reports experimental data at 1528 nm; a brief note on the wavelength shift or measurement conditions would improve clarity.
  2. [Methods / Experimental] Full fabrication process details, SEM images of the fabricated anisotropic features, and error bars on all efficiency/unidirectionality values are referenced only in passing; expanding these in the methods section would aid reproducibility.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the constructive feedback and for recognizing the potential of the bio-inspired anisotropy approach. We address each major comment point by point below, agreeing where revisions are warranted and providing the strongest honest defense of the manuscript's claims.

read point-by-point responses

  1. Referee: [Abstract / Experimental Results] Abstract and experimental measurements: The simulated in-coupling efficiency of 41% at vertical incidence is not reproduced experimentally, where the reported single-coupler insertion loss of 8.35 dB implies only ~14.6% efficiency—a gap exceeding 4 dB. The manuscript provides no loss budget, fabrication tolerance analysis, or calibrated absolute power measurements to identify whether the discrepancy arises from etch-depth variation, sidewall angle, subwavelength feature fidelity, or unmodeled scattering in the anisotropic stripes; this directly affects the central claim that the anisotropy enables efficient vertical coupling.

    Authors: We acknowledge the discrepancy between the simulated 41% in-coupling efficiency and the experimental value of approximately 14.6% inferred from the 8.35 dB single-coupler insertion loss. This gap exceeds 4 dB and represents a genuine limitation in the current presentation. In the revised manuscript we will add a dedicated loss-budget subsection that decomposes expected contributions from etch-depth variation, sidewall angle deviations, subwavelength feature fidelity, and scattering within the anisotropic stripes, supported by additional FDTD simulations that incorporate realistic fabrication imperfections. We will also clarify the absolute power calibration procedure used in the measurements. The experimental confirmation of 12.8 dB unidirectionality still substantiates the core anisotropy principle for enforcing directionality without extra etches or reflectors, even if overall efficiency requires further optimization. revision: yes

  2. Referee: [Design Principle] Design section: The anisotropy ratio and specific stripe geometry are listed as free parameters, yet no quantitative sensitivity analysis or Monte-Carlo tolerance study is presented to show that the >10 dB unidirectionality and 41% efficiency remain robust under realistic fabrication variations (e.g., ±10 nm critical-dimension error). Without this, the translation from simulation to experiment remains an unverified assumption.

    Authors: We agree that a quantitative sensitivity analysis strengthens the claim of robustness. In the revised manuscript we will insert a new subsection under Design that presents Monte-Carlo tolerance results. Key parameters including anisotropy ratio, stripe width, period, and etch depth will be varied within ±10 nm ranges drawn from typical fabrication statistics; the resulting distributions will show that unidirectionality remains above 9 dB and efficiency above 25% in the majority of realizations. These data will directly address the translation from ideal simulation to experimental conditions and demonstrate that the anisotropy heuristic retains its utility under realistic process variations. revision: yes

Circularity Check

0 steps flagged

No circularity: design rests on independent EM simulation and experiment

full rationale

The paper proposes an anisotropic grating geometry inspired by moth-eye structures, simulates its performance using standard finite-difference time-domain methods for the SOI platform at 1550 nm, and validates unidirectionality experimentally. No load-bearing step reduces a claimed prediction or first-principles result to a fitted parameter, self-citation, or definitional tautology; the 41% simulated efficiency and measured 12.8 dB unidirectionality are outputs of independent numerical modeling and fabrication/measurement, not rearrangements of the inputs. The sim-to-exp efficiency gap is a separate correctness issue outside the circularity analysis.

Axiom & Free-Parameter Ledger

1 free parameters · 1 axioms · 0 invented entities

Limited to abstract; the design assumes standard SOI platform behavior and electromagnetic simulation validity. No new physical entities are introduced.

free parameters (1)
  • Grating geometry and anisotropy ratio
    Specific dimensions and anisotropy strength optimized in simulation for 1550 nm; exact values not reported in abstract.
axioms (1)
  • domain assumption Anisotropy in grating stripes directs incident light unidirectionally with minimal back-reflection
    This is the central design principle invoked to achieve vertical coupling and high unidirectionality.

pith-pipeline@v0.9.0 · 5503 in / 1539 out tokens · 65806 ms · 2026-05-08T07:43:09.104432+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

11 extracted references · 11 canonical work pages

  1. [1]

    Coupling strategies for silicon photonics integrated chips.Photonics Research, 7(2):201–239, 2019

    Riccardo Marchetti, Cosimo Lacava, Lee Carroll, Kamil Gradkowski, and Paolo Minzioni. Coupling strategies for silicon photonics integrated chips.Photonics Research, 7(2):201–239, 2019

    work page 2019

  2. [2]

    Grating couplers on silicon photonics: Design principles, emerging trends and practical issues.Micromachines, 11(7):666, 2020

    Lirong Cheng, Simei Mao, Zhi Li, Yaqi Han, and HY Fu. Grating couplers on silicon photonics: Design principles, emerging trends and practical issues.Micromachines, 11(7):666, 2020

    work page 2020

  3. [3]

    Optical interfaces for silicon photonic circuits

    Diedrik Vermeulen and Christopher V Poulton. Optical interfaces for silicon photonic circuits. Proceedings of the IEEE, 106(12):2270–2280, 2018

    work page 2018

  4. [4]

    Flip-chip integration of tilted vcsels onto a silicon photonic integrated circuit.Optics express, 24(15):16258–16266, 2016

    Huihui Lu, Jun Su Lee, Yan Zhao, Carmelo Scarcella, Paolo Cardile, Aidan Daly, Markus Ortsiefer, Lee Carroll, and Peter O’Brien. Flip-chip integration of tilted vcsels onto a silicon photonic integrated circuit.Optics express, 24(15):16258–16266, 2016

    work page 2016

  5. [5]

    Waveguide sub-wavelength structures: a review of principles and applications

    Robert Halir, Przemek J Bock, Pavel Cheben, Alejandro Ortega-Mo˜ nux, Carlos Alonso-Ramos, Jens H Schmid, Jean Lapointe, Dan-Xia Xu, J Gonzalo Wang¨ uemert-P´ erez, ´I˜ nigo Molina- Fern´ andez, et al. Waveguide sub-wavelength structures: a review of principles and applications. Laser & Photonics Reviews, 9(1):25–49, 2015

    work page 2015

  6. [6]

    Subwave- length integrated photonics.Nature, 560(7720):565–572, 2018

    Pavel Cheben, Robert Halir, Jens H Schmid, Harry A Atwater, and David R Smith. Subwave- length integrated photonics.Nature, 560(7720):565–572, 2018

    work page 2018

  7. [7]

    Electromagnetic properties of a finely stratified medium.Soviet Physics JEPT, 2:466–475, 1956

    S Rytov. Electromagnetic properties of a finely stratified medium.Soviet Physics JEPT, 2:466–475, 1956

    work page 1956

  8. [8]

    On the effective medium theory of sub- wavelength periodic structures.Journal of Modern Optics, 43(10):2063–2085, 1996

    Philippe Lalanne and Dominique Lemercier-Lalanne. On the effective medium theory of sub- wavelength periodic structures.Journal of Modern Optics, 43(10):2063–2085, 1996

    work page 2063

  9. [9]

    High-order effective-medium theory of subwave- length gratings in classical mounting: application to volume holograms.JOSA A, 15(7):1843– 1851, 1998

    Philippe Lalanne and Jean-Paul Hugonin. High-order effective-medium theory of subwave- length gratings in classical mounting: application to volume holograms.JOSA A, 15(7):1843– 1851, 1998

    work page 1998

  10. [10]

    Antireflective surface inspired from biology: A review.Biosurface and Biotribology, 2(4):137–150, 2016

    ZW Han, Ze Wang, XM Feng, Bo Li, ZZ Mu, JQ Zhang, SC Niu, and LQ Ren. Antireflective surface inspired from biology: A review.Biosurface and Biotribology, 2(4):137–150, 2016

    work page 2016

  11. [11]

    Meep: A flexible free-software package for electromagnetic simulations by the fdtd method.Computer Physics Communications, 181(3):687–702, 2010

    Ardavan F Oskooi, David Roundy, Mihai Ibanescu, Peter Bermel, John D Joannopoulos, and Steven G Johnson. Meep: A flexible free-software package for electromagnetic simulations by the fdtd method.Computer Physics Communications, 181(3):687–702, 2010. 8

    work page 2010