An ultrathin gradient metasurface can, in principle, transform an incident electromagnetic wave into an arbitrarily prescribed wave front. Estakhri and Alù (Phys. Rev. X 6, 041008, 2016) gave an exact pointwise boundary-condition synthesis for two-dimensional, scalar-polarization reflection, and proved within that local model that unitary (100%) efficiency requires locally balanced gain and loss, that passive-lossless phase-only designs scatter power into undesired diffraction orders, that infinite periodic couplers to guided modes are inherently ill posed, and that perfect focusing assumes an ideal drain (time-reversed source) at the focal point. The open problem is to close the gap between this ideal theory and physically realizable, fully general surfaces: derive the exact synthesis and sharp efficiency/passivity bounds for three-dimensional vectorial, polarization-converting, and transmitting configurations; determine which passive nonlocal (or active balanced loss-gain) implementations can realize the required power redistribution; provide finite-size coupler and drain-less-lens synthesis procedures with quantified performance gaps; and develop systematic joint optimization under passivity, discretization, bandwidth, and fabrication constraints. Later literature has substantially resolved the two-dimensional passive-reflection and plane-to-surface-wave conversion facets, while the three-dimensional/vectorial generalization, the general physical limits of anomalous scattering, and the realization questions remain open.

Background

A metasurface is an ultrathin, planar array of subwavelength scatterers engineered to control reflected and transmitted electromagnetic waves. The modern era began with phase-discontinuity metasurfaces, in which a transversely varying phase shift imprinted on the wave implements generalized (gradient-modified) laws of reflection and refraction (Yu et al., Science 2011), and with Huygens' surfaces, electrically and magnetically polarizable sheets that tailor transmitted wave fronts with near-zero reflection (Pfeiffer and Grbić, Phys. Rev. Lett. 2013). A general tensor ('bianisotropic', i.e. coupling electric and magnetic responses) synthesis framework based on surface susceptibility tensors, covering transmission and polarization conversion, was formulated by Achouri, Salem and Caloz (IEEE Trans. Antennas Propag. 2015). In this context, Estakhri and Alù (Phys. Rev. X 6, 041008, 2016) asked what passive gradient metasurfaces can fundamentally do for arbitrary wave-front shaping. They model a reflecting sheet by position-dependent electric surface admittance Ye(x)Y_e(x) and magnetic surface impedance Zm(x)Z_m(x) (a 'magnetoelectric' or Huygens-type sheet; for an isotropic reflecting sheet YeZm=4Y_e Z_m = 4) and impose pointwise boundary conditions linking the tangential fields to the desired scattered field, which determines the local complex reflection coefficient R(x)=r(x)ejϕr(x)R(x) = r(x)e^{j\phi_r(x)} everywhere. Because the surface must be periodic with the superlattice, i.e. the smallest common period compatible with the incident and desired scattered wave vectors, the scattered power partitions into discrete diffraction harmonics of that periodicity (Floquet modes), and efficiency means the fraction of power in the intended order. Their central findings are: (i) local phase compensation, the ray-optics recipe behind most gradient-metasurface and flat-lens designs (linear phase gradient for beam steering, hyperbolic phase profile for focusing), is essentially inaccurate — the optimal phase and amplitude profiles deviate strongly at large deflection angles and high numerical aperture (NA, the sine of the half-angle of the focused cone); (ii) within the local model, unitary-efficiency transformation requires regions of locally balanced gain and loss, since a locally passive-lossy surface that couples exclusively to the desired order is limited to relative reflected amplitude ∣Ar∣≤min⁡[1,cos⁡θi/cos⁡θr]|A_r| \le \min[1, \cos\theta_i / \cos\theta_r] for TE (transverse electric) and TM (transverse magnetic) incidence, i.e. the two linear polarizations whose electric or, respectively, magnetic field is perpendicular to the plane of incidence; (iii) the natural extreme of beam steering — coupling a normally incident plane wave into a guided mode over an infinite periodic surface — is 'inherently ill posed', since the lateral guided power is constant while the incident normal power is finite, and practical finite couplers instead require a surface wave that grows gradually and adiabatically matches the fed waveguide mode; (iv) the optimal focusing solution is exact only in the presence of an 'ideal drain' at the focal point, i.e. a time-reversed source that absorbs the focused power, and its absence in a passive lens shifts and weakens the focus; (v) quantizing (discretizing) the ideal impedance profile into finitely many levels scatters residual power into undesired diffraction orders. The authors state that they consider two-dimensional reflection scenarios with scalar polarization, and that 'similar restrictions may be derived for three dimensions, polarization coupling surfaces, and transmitting metasurfaces'; they also flag that active elements are difficult to realize and introduce stability limitations, and that a concrete surface-wave-coupler design goes beyond the scope of the paper. Terms used below: 'anomalous reflection/refraction' means steering a wave to a non-specular (or non-Snell) direction; 'nonlocal' or 'spatially dispersive' metasurfaces have responses coupling spatially separated points, so power can flow along the sheet before being reradiated; 'metagratings' are sparse-scatterer gratings engineered beyond the graded-metasurface paradigm; 'reconfigurable intelligent surfaces' (RIS) are programmable metasurfaces considered for wireless coverage, for which efficiency limits are a design-critical quantity.

Problem Statement

The source paper derived, for two-dimensional reflection of a scalar-polarized wave, the exact magnetoelectric surface-impedance profiles that transform an incident wave into an arbitrary prescribed reflected wave front, and showed that within the local-response model unitary efficiency demands locally balanced gain and loss, that passive phase-only designs leak power into undesired diffraction orders, that infinite periodic couplers to guided modes are ill posed, and that unitary focal efficiency presumes an ideal drain at the focus. The open problem, stated by the paper as its own scope limitations — bridging the ideal local two-dimensional theory and physically realizable, fully general surfaces — is: (1) extend the exact pointwise boundary-condition synthesis and its passivity/efficiency restrictions to fully three-dimensional geometries, to polarization-coupling (tensorial/bianisotropic) surfaces, and to transmitting metasurfaces with simultaneous control of reflected and transmitted fields, and establish sharp physical bounds on the attainable efficiency of anomalous reflection, refraction, and polarization conversion in each class (later literature having shown that passive nonlocal surfaces beat the local loss/gain restriction, the bounds must cover nonlocal and multiport surface responses); (2) determine feasible physical realizations of the required power redistribution — passive nonlocal or spatially dispersive implementations, or stable, broadband, actively balanced loss-gain surfaces — including their stability, bandwidth, and noise margins; (3) derive concrete synthesis procedures for finite-size metasurface couplers that convert a realistic finite feed beam into a guided mode whose amplitude grows adiabatically along the surface while the surface remains locally passive; (4) quantify how the absence of an ideal drain (time-reversed source) degrades the efficiency and shifts the focus of high-numerical-aperture passive flat lenses, and engineer realistic absorber/drain configurations that mitigate it; and (5) develop systematic design methods that jointly trade local passivity, discretization granularity, bandwidth, and fabrication constraints to maximize usable throughput in practical devices. Facet (1) — the exact synthesis and sharp efficiency/passivity bounds for the general classes above — is the primary scientific target; facets (2)–(5) are further independent open questions. For the facets that involve a device demonstration, the scientific target is a design whose measured efficiency approaches the corresponding derived bound without the idealizations (infinite periodicity, ideal drain, lossless local sheets) of the 2016 two-dimensional theory.

References

  1. Nasim Mohammadi Estakhri and Andrea Alù, "Wave-front transformation with gradient metasurfaces," Phys. Rev. X 6, 041008 (2016). DOI: 10.1103/PhysRevX.6.041008, https://doi.org/10.1103/PhysRevX.6.041008
  2. Nanfang Yu, Patrice Genevet, Mikhail A. Kats, Francesco Aieta, Jean-Philippe Tetienne, Federico Capasso, and Zeno Gaburro, "Light propagation with phase discontinuities: generalized laws of reflection and refraction," Science 334, 333-337 (2011). DOI: 10.1126/science.1210713, https://doi.org/10.1126/science.1210713
  3. Carl Pfeiffer and Anthony Grbić, "Metamaterial Huygens' surfaces: tailoring wave fronts with reflectionless sheets," Phys. Rev. Lett. 110, 197401 (2013). DOI: 10.1103/PhysRevLett.110.197401, https://doi.org/10.1103/PhysRevLett.110.197401
  4. Karim Achouri, Mohamed A. Salem, and Christophe Caloz, "General metasurface synthesis based on susceptibility tensors," IEEE Transactions on Antennas and Propagation 63, 2977-2991 (2015). DOI: 10.1109/TAP.2015.2423700, https://doi.org/10.1109/TAP.2015.2423700
  5. V. S. Asadchy, M. Albooyeh, S. N. Tcvetkova, A. Díaz-Rubio, Y. Ra'di, and S. A. Tretyakov, "Perfect control of reflection and refraction using spatially dispersive metasurfaces," Phys. Rev. B 94, 075142 (2016). DOI: 10.1103/PhysRevB.94.075142, https://doi.org/10.1103/PhysRevB.94.075142
  6. Younes Ra’di, Dimitrios L. Sounas, and Andrea Alù, "Metagratings: beyond the limits of graded metasurfaces for wave front control," Phys. Rev. Lett. 119, 067404 (2017). DOI: 10.1103/PhysRevLett.119.067404, https://doi.org/10.1103/PhysRevLett.119.067404
  7. Do-Hoon Kwon and Sergei A. Tretyakov, "Perfect reflection control for impenetrable surfaces using surface waves of orthogonal polarization," Phys. Rev. B 96, 085438 (2017). DOI: 10.1103/PhysRevB.96.085438, https://doi.org/10.1103/PhysRevB.96.085438
  8. S. N. Tcvetkova, D.-H. Kwon, A. Díaz-Rubio, and S. A. Tretyakov, "Near-perfect conversion of a propagating plane wave into a surface wave using metasurfaces," Phys. Rev. B 97, 115447 (2018). DOI: 10.1103/PhysRevB.97.115447, https://doi.org/10.1103/PhysRevB.97.115447
  9. Hamidreza Taghvaee, Fu Liu, Ana Díaz-Rubio, and Sergei Tretyakov, "Perfect-lens theory enables metasurface reflectors for subwavelength focusing," Phys. Rev. Applied 19, 014004 (2023). DOI: 10.1103/PhysRevApplied.19.014004, https://doi.org/10.1103/PhysRevApplied.19.014004
  10. Shiyu Li and Chia Wei Hsu, "Transmission efficiency limit for nonlocal metalenses," Laser & Photonics Reviews 17, 2300201 (2023). DOI: 10.1002/lpor.202300201, https://doi.org/10.1002/lpor.202300201
  11. Shiyu Li, Ho-Chun Lin, and Chia Wei Hsu, "High-efficiency high-numerical-aperture metalens designed by maximizing the efficiency limit," Optica 11, 454 (2024). DOI: 10.1364/OPTICA.514907, https://doi.org/10.1364/OPTICA.514907
  12. Mats Gustafsson, "Physical limits and optimal synthesis of beyond diagonal anomalous scatterers," arXiv:2505.00691 (2025). DOI: 10.48550/arXiv.2505.00691, https://arxiv.org/abs/2505.00691
  13. Jonathon George Smith, Ian Richard Hooper, Nathan Clow, Alastair Paul Hibbins, and Simon Arthur Robin Horsley, "Simultaneous anomalous reflection and polarization conversion from a tensorial impedance metasurface," npj Metamaterials 2 (2026). DOI: 10.1038/s44455-026-00033-w, https://doi.org/10.1038/s44455-026-00033-w
  14. Serena Assefa Asfaw, Federico Giusti, Matteo Albani, and Enrica Martini, "Canonical problem of out-of-plane anomalous reflection from modulated opaque impedance surfaces," in 2026 20th European Conference on Antennas and Propagation (EuCAP), pp. 1-4 (2026). DOI: 10.23919/EuCAP68105.2026.11612120, https://doi.org/10.23919/EuCAP68105.2026.11612120