Inverse Design of an All-Dielectric Nonlinear Polaritonic Metasurface.
Open accessThe DOI matches its Crossref record; 62 of the 62 references it cites were found in a registry or an index; 9 of the 10 sources its claims cite were checked for support.
Abstract
Nonlinear metasurfaces offer the ability to realize optical nonlinear devices with unparalleled properties compared to nonlinear crystals, due to the interplay between photonic resonances and materials properties. The complicated interdependency between efficiency and emission directionality of the nonlinear optical signal on the existence, localization, and lifetimes of photonic resonances, as well as on the nonlinear susceptibility, makes it extremely difficult to design optimal metasurfaces using conventional materials and geometries. Inverse design using topology optimization is a powerful design tool for photonic structures, but traditional approaches developed for linear photonics are not suitable for such high dimensional nonlinear problems. Here, we use a topology optimization approach to inverse-design a fabrication-robust nonlinear metasurface that includes quantum-engineered resonant nonlinearities in semiconductor heterostructures for efficient and directional second harmonic generation. Furthermore, we also demonstrate that under practical constraints, among all the parameters, the nonlinear modal overlap emerges as the dominant parameter that enhances conversion efficiency, a finding that contrasts with intuition-driven studies that often emphasize Purcell enhancement. Our results demonstrate an efficient approach for optimizing nonlinear processes in nanophotonic structures for classical and quantum light sources, quantum information applications, and communication.
Overview
CoChat Summary
TL;DR
The paper inverse-designs an all-dielectric nonlinear polaritonic metasurface using topology optimization with quantum-engineered intersubband transitions, simultaneously targeting SHG efficiency and normal-direction emission. The fabricated constrained design achieves record-high SHG efficiency for all-dielectric metasurfaces and shows that nonlinear modal overlap, not Purcell/high-Q enhancement, dominates under practical fabrication constraints.
Key findings
- The inverse-designed metasurface achieved a record-high SHG efficiency of 0.38 mW/W² versus 0.15 mW/W² for the cylindrical benchmark under NA = 0.85 collection; zero-order emission was 0.047 mW/W², about 10% of total SHG efficiency.
- Simulations predicted the inverse-designed structure to be ~11× more efficient than cylinders at λ_IST = 7.9 µm, and ~5× higher than the cylinder peak at λ_C; it emitted ~10% of total SH power along the normal direction, whereas cylindrical resonators had zero normal emission due to symmetry.
- Under fabrication constraints, the nonlinear modal overlap β was the dominant optimized parameter, ~40× larger than for cylinders; without constraints, topology optimization instead favored a ~9× higher Q at the fundamental, with β enhanced only ~2.6×.
- The unconstrained design was simulated to be ~17× more efficient than cylinders at 7.9 µm and ~7.5× higher than the cylinder peak, but after fabrication its measured SHG efficiency was lower than cylinders because of mode red-shifting.
Methods
- Density-based topology optimization was used with a 50 nm square grid, adjoint-method gradients, method-of-moving-asymptotes optimization, and on-the-fly filtering/binarization; optimization was sequential, first maximizing total SH power and then far-field z-component SH Poynting flux, with fabrication constraints.
- Full-wave simulations used frequency-domain finite-element COMSOL with an anisotropic Lorentzian multi-QW permittivity, Floquet boundary conditions, plane-wave ports, undepleted pump approximation, 4.2 µm period, 1.6 µm thickness, pump at 7.9 µm, and SH at 3.95 µm.
- Fabrication used epi-side-down BCB bonding to sapphire, substrate removal, electron-beam lithography with HSQ, and Cl/Ar reactive ion etching; SHG measurements used a pulsed tunable QCL (3–13 µm, 400 ns, 250 kHz) and an MCT detector with NA = 0.18 and NA = 0.85 collection.
Limitations
- The paper states simulations predicted approximately 5× larger SHG efficiency for the inverse-designed structure than experiments, attributed to fabrication imperfections.
- The unconstrained inverse-designed structure required aspect ratios of approximately 1:16 and isolated empty regions that fabrication could not fully resolve; this caused red-shifted modes, with M2 instead of M1 overlapping the IST resonance, and measured SHG efficiency lower than cylinders.
- The paper notes SHG conversion efficiency saturates at higher pump powers due to IST intensity saturation, and NA = 0.85 measurements showed sub-linear scaling because of smaller pump spot size and larger intensities.
- Evident from the text: pre-compensating for fabrication imperfections such as etching profile and blurring would require a more elaborate computational framework and is left for future studies.
Model-written summary. Check it against the paper before relying on it.
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
2 Valid · 4 Not valid · 0 Not checked · 74 Own results and interpretations
The cited sources that could be checked couldn’t confirm or dispute this claim.
In the paper
“ISTs offer one of the largest χ (2) available in condensed matter systems [37]”
What the sources say
- Could not verifycites [37] sourceExperimental — accuracy not yet measured
The Results section discusses χ(2) in metasurface design and identifies InGaP as the nonlinear material used in the examples. Searches for ISTs, intersubband transitions, and a comparison of χ(2) values across condensed-matter systems found no passage addressing the claim. (LLM, conf=0.74)
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Could not verify
The cited sources that could be checked couldn’t confirm or dispute this claim.
In the paper
“the cylindrical resonators support a magnetic dipole-type Mie mode at the pump frequency [22]”
What the sources say
- Could not verifycites [22] sourceExperimental — accuracy not yet measured
The available indexed passages are from an archive listing (section labeled “Abstract > 1.1 Articolo in rivista,” no page number), which links to a PDF but does not provide the paper’s substantive text. Searches for the magnetic dipole-type Mie mode and cylindrical resonators at the pump frequency found no passage that addresses the claim. (LLM, conf=0.76)
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Could not verify
The cited sources that could be checked couldn’t confirm or dispute this claim.
In the paper
“the adjoint method allows us to compute the objective gradient at a similar cost to computing the objective function itself, through the solving of an adjoint partial differential equation (PDE) [47].”
What the sources say
- Could not verifycites [47] sourceExperimental — accuracy not yet measured
The paper’s modeling passage (section labeled Abstract; page unavailable) describes temporal coupled-mode theory and analytical expressions for absorption and reflectivity. Searches for the adjoint method, objective gradients, comparable computation costs, and adjoint PDEs found no passage addressing the cited claim. (LLM, conf=0.76)
plainstates a cause- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Could not verify
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“Compared to the maximum SHG efficiency of the cylindrical metasurface at 𝜆 , the inverse-designed structure has ~ 5x higher SHG efficiency.”
plainstates a cause- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“removing some of these isolated regions of the optimized structure significantly reduces the SHG efficiency to values below the SHG efficiency of the structure shown in Figure”
plainstates a cause- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
50 more claims
Create a free account to see all 80 claims and every source verdict.
- It may state two separate claims as one. 20
- Probably not a substantial claim. 4
- The quote only partly states this claim. 2
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
Each contradiction between two passages listed here was confirmed by an independent check.
Rated minor: it would not touch a result — a value no result rests on, a p-value that stays on the same side of the significance level, or a conflict that is clerical, resolved elsewhere in the paper, or outside its own findings.
In the paper
§ FIGURES · p. 21
“The metasurface with no constraints has a peak of SHG efficiency that is approximately 17 times larger compared to the peak of the SHG efficiency of the metasurface with cylindrical resonators at pump wavelength λ IST ~ 7.9 µm.”
§ Importance of Constraints on Inverse Design · p. 9
“Compared to the peak of SHG at 𝜆 for the cylindrical metasurface, the same inverse-designed structure has approximately 7.5 times higher SHG efficiency compared to cylinders.”
Citations
Cites
Inverse design in nanophotonics
Molesky, S., Lin, Z., Piggott, A. Y., Jin, W., Vucković, J., & Rodriguez, A. W. (2018) · Nature Photonics · doi:10.1038/s41566-018-0246-9 · Cited by 1,818 · cited 2×
“The computational approaches for discovering optical structures based on desired functional characteristics have recently redefined the limits of nanophotonic devices in the linear regime [30, 31].”
BackgroundMethodSubwavelength dielectric resonators for nonlinear nanophotonics
Koshelev, K., Kruk, S., Melik-Gaykazyan, E., Choi, J.-H., Bogdanov, A., Park, H.-G., & Kivshar, Y. (2020) · Science · doi:10.1126/science.aaz3985 · Cited by 997
High- Q Quasibound States in the Continuum for Nonlinear Metasurfaces
Liu, Z., Xu, Y., Lin, Y., Xiang, J., Feng, T., Cao, Q., Li, J., Lan, S., & Liu, J. (2019) · Physical Review Letters · doi:10.1103/physrevlett.123.253901 · Cited by 863 · cited 1×
“For example, recently bound-state-in-the-continuum (BIC) photonic modes have been utilized to achieve high quality factors at the fundamental wavelength [14-17].”
BackgroundTopology optimization for nano‐photonics
Jensen, J. S., & Sigmund, O. (2011) · Laser & Photonics Reviews · doi:10.1002/lpor.201000014 · Cited by 767 · cited 2×
“The computational approaches for discovering optical structures based on desired functional characteristics have recently redefined the limits of nanophotonic devices in the linear regime [30, 31].”
BackgroundMethodObservation of intense second harmonic generation from MoS2atomic crystals
Malard, L. M., Alencar, T. V., Barboza, A. P. M., Mak, K. F., & de Paula, A. M. (2013) · Physical Review B · doi:10.1103/physrevb.87.201401 · Cited by 760
Adjoint shape optimization applied to electromagnetic design
Lalau-Keraly, C. M., Bhargava, S., Miller, O. D., & Yablonovitch, E. (2013) · Optics Express · doi:10.1364/oe.21.021693 · Cited by 753
Giant nonlinear response from plasmonic metasurfaces coupled to intersubband transitions
Lee, J., Tymchenko, M., Argyropoulos, C., Chen, P.-Y., Lu, F., Demmerle, F., Boehm, G., Amann, M.-C., Alù, A., & Belkin, M. A. (2014) · Nature · doi:10.1038/nature13455 · Cited by 669
Self-citationNonlinear metasurfaces: a paradigm shift in nonlinear optics
Krasnok, A., Tymchenko, M., & Alù, A. (2018) · Materials Today · doi:10.1016/j.mattod.2017.06.007 · Cited by 613 · cited 2×
“…support different types of localized and de-localized photonic resonances capable of dramatically enhancing light-matter interactions [5-7], enabling significant nonlinear responses over much smaller optical volumes compared to bulk crystals while relaxing cumbersome phase matching constraints [8].”
BackgroundRoadmap for Optical Metasurfaces
Kuznetsov, A. I., Brongersma, M. L., Yao, J., Chen, M. K., Levy, U., Tsai, D. P., Zheludev, N. I., Faraon, A., Arbabi, A., Yu, N., et al. (2024) · ACS Photonics · doi:10.1021/acsphotonics.3c00457 · Cited by 420
Resonant metasurfaces for generating complex quantum states
Santiago-Cruz, T., Gennaro, S. D., Mitrofanov, O., Addamane, S., Reno, J., Brener, I., & Chekhova, M. V. (2022) · Science · doi:10.1126/science.abq8684 · Cited by 326
Self-citation
Research works printed but not linked: 1
- Brener I., 2019, Dielectric Metamaterials: Fundamentals, Designs and Applications
Cited by
Inverse design for scalable photonic systems
Schul, L., Mason, S., Eun, S., Ahn, G. H., & Vučković, J. (2026) · Nature Reviews Materials · doi:10.1038/s41578-026-00915-5 · Cited by 28
“110).”
Abdelraouf, O. A. M. (2025) · Optics Communications · doi:10.1016/j.optcom.2025.132526 · Cited by 12
Gradient-based optimization of scatterer arrangements based on the T-matrix method
Asadova, N. H., Fischbach, J. D., Vallée, R. A. L., Kuster, O., Augenstein, Y., Vovchuk, D., Kharchevskii, A. A., Ginzburg, P., & Rockstuhl, C. (2026) · APL Photonics · doi:10.1063/5.0316808 · Cited by 3
Aoueille, A. T., Hurley, N., Bustamante, E., Martinez, N. P., Reinke, C. M., El-Kady, I. F., & Lin, Y. (2025) · Optics Letters · doi:10.1364/ol.583903 · Cited by 1
Artificial intelligence empowered metasurface design
Wang, Y., Tang, X., Liu, C., & Ma, W. (2026) · Intelligent Opto-Electronics · doi:10.67704/ioe.2026.260007 · Cited by 1
Free space few-photon nonlinearity in critically coupled polaritonic metasurfaces
Fang, J., Kala, A., Johnson, R. A., Sharp, D., Chen, R., Chang, C., Munley, C., Fröch, J. E., Naresh, V., Tang, A., et al. (2025) · Nature Communications · doi:10.1038/s41467-025-65088-1
“While at longer wavelength like mid-infrared strong polaritonic nonlinearity has been frequently studied 22,23 , the demonstration at visible wave-lengths is usually dif fi cult.”
Plasmonic metasurfaces for spin-selective optical vortex generation and transverse bifocusing
Zhao, X., Luo, S., Wang, X., et al. (2026) · Physical Review Applied · doi:10.1103/q8h2-6xjy
Inverse design of multiwavelength multifocal metalens based on the tandem neural networks
Song, Y., Wu, B., & Sun, M. (2026) · EPJ Applied Metamaterials · doi:10.1051/epjam/2025017
Ionochromic optical sensor of amino-functionalized alginate brush grating
Chen, C., Iqbal, M., Cheng, C., et al. (2026) · International Journal of Biological Macromolecules · doi:10.1016/j.ijbiomac.2026.153129
Nonlinear meta-devices for high-dimensional optical applications
Wang, Z., Yao, J. L., & Tsai, D. P. (2026) · APL Photonics · doi:10.1063/5.0339283
Metadata
- Authors
- Stich, S., Mohajan, J., de Ceglia, D., Carletti, L., Jung, H., Karl, N., Brener, I., Rodriguez, A. W., Belkin, M. A., & Sarma, R.
- Affiliations
- Technical University of Munich (DE)
- Princeton University (US)
- National Institute of Optics─National Research Council (INO−CNR) and Department of Information Engineering
- National Academies of Sciences, Engineering, and Medicine (US)
- Sandia National Laboratories California (US)
- Department of Electrical and Computer Engineering
- Corresponding author
- Belkin, Mikhail A.
- Sarma, Raktim
- ORCID
- Venue
- ACS Nano
- Publisher
- American Chemical Society (ACS)
- Published
- 2025-04-09
- Volume
- 19
- Issue
- 18
- Pages
- 17374-17384
- ISSN
- 1936-0851 (ISSN-L)
- 1936-086X
- OpenAlex
- W4409315190
- Cited by
- 13
- References
- 55
- Citations by year
- 2026: 8 · 2025: 5
- FWCI
- 2.08
- Type
- journal article
- Language
- en
- Open access
- green
- arxiv.org
- Licence
- Locations
- ACS Nano (publishedVersion)
- PubMed (publishedVersion)
- arXiv (Cornell University) (submittedVersion)
- Institutional Research Information System (Università degli Studi di Brescia) (submittedVersion)
- arXiv (Cornell University)
- Funders
- Division of Materials Sciences and Engineering
- Topics
- Physical Sciences › Materials Science › Electronic, Optical and Magnetic Materials › Metamaterials and Metasurfaces Applications
- Physical Sciences › Physics and Astronomy › Atomic and Molecular Physics, and Optics › Plasmonic and Surface Plasmon Research
- Physical Sciences › Physics and Astronomy › Atomic and Molecular Physics, and Optics › Photonic Crystals and Applications
- Keywords
- nonlinear metasurfaces, inverse design, topology optimization, dielectric metasurfaces, second harmonic generation, semiconductor heterostructures, directional emission, photonic resonances, nonlinear optical susceptibility, Purcell effect
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
The claim verdicts this report counts are experimental: their accuracy is not yet measured.
A language model drew the claims from the paper on 2026-10-29, under version 3 of the claim definition (model deepseek-v4.1-flash, prompt extract-v3).
A second model labelled each claim on 2026-10-29 (model jev-1.13-20260917, definition revision gold-guideline-v0.2).
Cited sources are checked with a language model’s help (gpt-6-sol).
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.