Ultra-thin, entirely flat, Umklapp lenses

Gregory J. Chaplain; Richard V. Craster

REVIEW 5 major objections 6 minor 35 references

Ultra-thin, entirely flat, Umklapp lenses

T0 review · 5 major / 6 minor · reviewed 2026-08-14 · deepseek-v4-flash

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

Pith's one-line read An entirely flat, one-unit-cell-thick dielectric array can focus light like a lens with negative refraction by using crystal-momentum transfer at an abrupt change in periodicity.

desk verdict A genuinely new flat-lensing mechanism backed by qualitative simulation, but missing the quantitative check that would make the Umklapp claim stick. read the letter →

arxiv 1908.07797 v1 pith:34GFFXPU submitted 2019-08-21 physics.optics

Gregory J. Chaplain , Richard V. Craster This is my paper · ORCID

classification physics.optics

keywords UmklappscatteringcrystalmomentumtransferflatlensnegativerefractiondielectricarraysurfacewavesBrillouinzonePendry-Veselago

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 designs ultra-thin, entirely flat dielectric lenses that achieve focusing without any negative-index material. At an abrupt junction between two periodic regions of a silicon-nitride array, guided surface waves scatter into backward-propagating beams via an Umklapp process, transferring crystal momentum through a reciprocal lattice vector. The authors show that a line source placed on one side of the array produces real focal spots on both sides, emulating a Pendry-Veselago lens. The effect is demonstrated numerically at visible frequencies between 420 and 500 THz, and the design principle is argued to be independent of the specific material or size.

What carries the argument

The central mechanism is the Umklapp process, the transfer of crystal momentum via a reciprocal lattice vector $G=2\pi/a$, described by $\kappa_1+\kappa_2-\kappa_3=G$ for a U-process. The design exploits an overlap between the first Brillouin zone of a region with period $a_1$ and the second Brillouin zone of a region with larger period $a_2>a_1$, so that a guided wavevector lying in the first zone of region 1 falls into the second zone of region 2. By matching mode shapes and isofrequency contours across the abrupt interface, the incident guided mode couples to a reversed propagating beam whose angle is set by the phase-matching condition $\kappa_{\mathrm{wg}} = \kappa n_0\sin\theta + 2\pi/a_2$. This construction replaces explicit negative refractive index with a purely geometric periodicity change.

What would settle it

Fabricate the two-periodicity Si3N4 array with the specified dimensions and measure the far-field pattern of a line-source-excited surface wave: if no backward-propagating beam appears at the angle predicted by $\kappa_{\mathrm{wg}} = \kappa n_0\sin\theta + 2\pi/a_2$, or if a single-periodicity array with no transition region produces the same pattern, the Umklapp mechanism is not doing the claimed work.

Watch

Extended reading notes

Core claim

A structured dielectric slab partitioned into two regions of different periodicities can act as a flat lens. An array-guided mode in the first region, with wavevector within the first Brillouin zone, crosses into the second region where the same wavevector lies in the second Brillouin zone. Because the second region has a larger unit cell, its reciprocal lattice vector $G=2\pi/a_2$ is smaller, so the incident wavevector can be folded back into the first zone via an Umklapp flip-over process. This promotes reversed conversion: the guided surface wave radiates into a backward-directed beam in the exterior medium, at an angle predicted by phase matching with the first negative diffractive order of standard mode-coupling theory. By placing the periodicity transitions symmetrically about a central point, the reversed beams refocus a line source into two images on the opposite side of the slab, realizing flat lensing with a device that is one unit cell in width and has completely flat edges.

Load-bearing premise

The design assumes that at the abrupt boundary between the two periodic regions, the incident array-guided mode couples efficiently into a second-Brillouin-zone mode of the other region and that this desired reversed beam dominates over parasitic scattering at the interface, with no quantitative coupling efficiency given.

Editorial extensions

If this is right

The proposed device works as a flat lens at visible/near-visible frequencies (420-500 THz) using a positive-index dielectric, with thickness of just one unit cell and no protruding features.
Focal spot positions are tunable by changing the location of the periodicity-transition regions, the relative periodicities $a_1$ and $a_2$, and the symmetry of the array.
The mechanism is broadband: reversed conversion and focusing are shown at 484 THz and 418 THz, with the conversion angle shifting as frequency changes.
Loss in the dielectric weakens but does not eliminate the effect; placing the transition regions near the excitation point compensates for decay of the guided wave.
Because the analysis relies only on dispersion-curve overlap and mode symmetry, the design principle transfers to other materials and wavelength regimes by geometric scaling.

Reading between the lines

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

Extending beyond the paper: the same Umklapp conversion mechanism should operate for acoustic or elastic surface waves, since the argument depends only on periodicity overlap and symmetry of guided modes, not on electromagnetism.
A quantitative prediction the paper does not state: conversion efficiency should be governed by the overlap integral between the two regions' modal field profiles; maximizing that overlap at design frequencies should increase focal intensity at the cost of bandwidth.
The design implies a testable scaling rule: scaling both periodicities and inclusion sizes by a common factor should shift the operating frequency band proportionally in the lossless case, because the dispersion curves scale with the unit-cell size.
Using piezoelectric or otherwise tunable materials could make the periodicity actively controllable, turning the passive lens into an electrically adjustable beam-steering or zoom element.

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.

Desk Editor's Note

read the letter

Two things you should know. First, the paper introduces a genuinely new lensing mechanism: an abrupt change in the periodicity of a dielectric line array folds a guided surface wave from the first Brillouin zone of one region into the second zone of another, and the resulting Umklapp transfer launches a backward beam into the surrounding medium. Two such transitions act like a Pendry-Veselago lens without negative index. Second, the evidence is good but incomplete: full-wave simulations at two frequencies with a control array show the effect, yet the authors never compare the beam angle predicted by their mode-coupling equation (3) to the angle in the simulated field, and they report no coupling efficiency. That gap is the difference between demonstrating an effect and demonstrating the mechanism.

What's actually new: the prior line-array work by the same group used adiabatic grading inside the first Brillouin zone. Here it's an abrupt interface and transfer of crystal momentum, which is a different trick. The design method is clearly explained with dispersion curves and isofrequency contours, and the simulations at 484 and 418 THz both give focal spots, with a control showing no focusing when the transition is absent. That's a fair amount of support for a numerical design paper.

Where the soft spots are: the stress-test note is on target. The predicted angle from Eq. (3) is never computed and checked against Fig. 1(a) or S1(a). Without that, the field plots don't rule out ordinary scattering at the discontinuity or direct source radiation. The control shows the transition is necessary, but not that U-process phase matching is what's happening. The paper also says 'undesirable scattering is anticipated, but by carefully engineering the design we can recapture the scattered field'—yet no number is given for how much of the incident power ends up in the reversed beam. Loss analysis is included, which is nice, but there's no efficiency or spot-size metric. Text sloppiness: permittivity is called permeability, and the operating range is described as both visible (420–500 THz) and terahertz. None of this is fatal, but it fits a paper that hasn't been tightened.

Bottom line: the central idea is plausible and the simulations back it qualitatively, but the quantitative case needs to be made before this would convince a skeptical reader. I'd send it to peer review, with a required revision that adds a predicted-versus-simulated angle comparison, a coupling or scattering-loss estimate, and cleanup of the text. It's worth bringing to a reading group and worth citing as a design concept once the numbers are in.

Referee Report

5 major / 6 minor

Summary. Chaplain and Craster propose an ultra-thin, entirely flat dielectric lens based on Umklapp scattering at an abrupt change in the periodicity of a singly periodic array. The device is a Si3N4 slab in air containing two regions with different lattice constants and different elliptical-inclusion sizes. Because the first Brillouin zone of region 1 overlaps the second Brillouin zone of region 2, an array-guided mode arriving at the transition can be transferred by a reciprocal lattice vector G = 2π/a2 into a backward-propagating exterior beam, as expressed in Eqs. (2) and (3). Full-wave simulations at 484 and 418 THz show reversed conversion and two-sided focusing for a line source, and a control array without the periodicity change shows no focusing. The authors claim the device emulates a Pendry-Veselago lens one unit cell in width without requiring a negative refractive index, and they examine the effect of material loss.

Significance. If quantitatively confirmed, the paper would introduce a genuinely new flat-lensing mechanism: using Umklapp processes in singly periodic dielectric structures, with no resonant elements, no negative index, and flat exterior faces. The mode-coupling picture is concrete, and the design is not fitted to a target lensing curve: the geometry is chosen from dispersion overlap, and the central effect is demonstrated by full-wave simulation with a control case. These are real strengths. The significance is conditional, however, because the evidence is currently qualitative: there is no predicted-versus-simulated angle comparison, no coupling efficiency or insertion-loss accounting, and no quantitative focusing metric.

major comments (5)

[Design Methods, Eq. (3), Fig. 1(a)] The central mechanism is not quantitatively established. The text states that the reversed conversion angle is 'explicitly predicted from mode coupling analysis' by rearranging Eq. (3) with Λ = 2π/a2, but no predicted angle is ever computed and compared with the outgoing beam in Fig. 1(a) or Fig. S1(a). The isofrequency-contour construction in Fig. 4(d) is schematic: for a one-dimensional guided mode the dispersion is a curve in (ω, κx), not an isotropic circular contour, so the phase-matching argument should be presented directly from Eq. (3) and the computed band structure. Please provide the predicted beam angle (and the modal wavevector used in Eq. (3)) and compare it with the simulated field for both operating frequencies; without this comparison, the field plots do not discriminate the proposed U-process from ordinary scattering at the discontinuity or from direct source radiation.
[Introduction and Design Methods] The claim that 'undesirable scattering of the field at such an interface is anticipated, but by carefully engineering the design we can recapture the scattered field' is not supported by any quantitative measure. No coupling efficiency, insertion loss, or fraction of incident array-guided power converted into the reversed beam is reported, nor is any modal overlap integral between the region-1 and region-2 modes computed. Without such an accounting, the assertion that U-processes dominate over parasitic scattering remains unsupported. Please add a power-flow or S-parameter analysis at the transition so that the converted fraction can be assessed.
[Supplemental Fig. S1(b)] The control array without the transition shows that the abrupt periodicity change is necessary for the observed reversal, but it does not show that Umklapp phase matching is the operative mechanism. A more discriminating control would be a transition in which the periodicity changes but the Brillouin-zone mode overlap of Fig. 3 is suppressed, or a systematic comparison of simulated beam angles with the prediction of Eq. (3). Please add such a control or otherwise rule out competing explanations such as geometric scattering at the interface.
[Umklapp Lensing, Fig. 1(b)] The claim that the device operates as a Pendry-Veselago lens is supported only by visual inspection of field plots. There is no quantitative focusing metric: no spot size relative to the diffraction limit, no focal contrast against background, no throughput efficiency, and no comparison of the image position with the source position expected from symmetry and from the predicted conversion angle. Please characterize the focal spots quantitatively so that 'focusing' is separated from near-field artifact.
[Introduction, Eq. (1), Table I] The material identification is internally inconsistent. The text introduces Si3N4 as a low-index material with n ≈ 2, but the simulations use εr = 9.7, which corresponds to n ≈ 3.1. Since the abstract and conclusions present the device as a Si3N4 lens at visible frequencies, the simulation parameters must be reconciled with the actual material (εr ≈ 4.0 for Si3N4 at visible wavelengths) or the material should be relabeled. This is not a conceptual flaw in the Umklapp mechanism, but it is a factual error in the demonstrated device.

minor comments (6)

[Eq. (1)] The quantity κ0 is not defined; please state κ0 = ω/c and specify the time-harmonic convention (e^{-iωt} or e^{+iωt}) used in the Comsol simulations.
[Introduction, after Eq. (1)] The phrase 'high relative permeability (εr = 9.7)' should read 'high relative permittivity'; the symbol εr denotes a permittivity throughout the paper.
[Abstract and Introduction] The abstract states operation at 'visible wavelengths between 420–500 THz', while the text later says the devices operate 'in the terahertz frequency range'; these statements are inconsistent and should be reconciled.
[Fig. 2 and Table I] The figure and table do not state whether the elliptical inclusions are air voids or inclusions of another dielectric; please specify the inclusion material and the exact unit-cell geometry used in the full-wave simulations.
[Fig. 3 and Concluding remarks] The claim of broadband performance over 420–500 THz is based on overlapping dispersion bands, but only two frequencies (484 and 418 THz) are simulated; please state the predicted usable bandwidth or add a frequency sweep of the focal-spot quality.
[References and Fig. 5(b)] Reference [10] is incomplete (journal volume and pages missing), and the quantity plotted as the 'normalised electric field norm' in Fig. 5(b) should be defined (e.g., |E|/|E|max) with labeled axes and units.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: the Umklapp-lensing claim is self-contained; prior self-citations are contextual and not load-bearing.

full rationale

Walking the derivation chain: the design starts from standard Floquet-Bloch dispersion curves computed with COMSOL, identifies a frequency range where region 1's first Brillouin zone overlaps region 2's second Brillouin zone (Fig. 3 and Fig. S2), and then uses the conventional phase-matching relation Eq. (3) with Lambda = 2 pi / a2 to predict the output beam angle. No parameter of Eq. (3) is fitted to the full-wave simulations: the device geometry (Table I) is fixed before the field plots, and the claimed focal spots are checked against a control array with no periodicity transition (Fig. S1(b)). The only author-overlap citations ([14], [18]) are used as background for adiabatic array rainbow trapping and for homogenization estimates of decay length; they are not invoked as a uniqueness theorem or as the justification that Umklapp transfer produces lensing. The admitted absence of a quantitative comparison of predicted versus simulated beam angles, and of a coupling efficiency for the U-process, is a completeness/correctness concern, not a circular reduction: Eq. (2) is not defined in terms of the simulated focal spots, and no fitted quantity is renamed as a prediction. The typos (permeability for permittivity; terahertz vs visible) do not affect the derivation chain. Therefore no circularity is found.

Assumptions & free parameters 5 free parameters · 3 assumptions · 0 invented entities

The central mechanism rests on the hand-picked geometry (periods a1=300nm and a2=330nm, width 500nm, inclusion axes listed in Table I) and on the standard assumptions of scalar TE wave propagation, Bloch band theory, and grating phase matching. No new physical entity is introduced. The free parameters are not fitted to the target result, but they are chosen specifically to create the band overlap that makes Umklapp coupling possible.

free parameters (5)

a1, region 1 periodicity = 300 nm
Chosen manually to produce the band structure and mode overlap needed for Umklapp coupling; no first-principles constraint fixes it.
a2, region 2 periodicity = 330 nm
Larger period chosen so the second Brillouin zone of region 2 overlaps the first of region 1; hand-picked for the effect.
slab width w = 500 nm
Set as the waveguide thickness in Table I; chosen to support the guided surface modes used in the design.
inclusion semi-axes in region 1 (ra, rb) = 170 nm, 110 nm
Ellipse geometry in region 1 selected to give mode shapes that overlap with region 2 modes.
inclusion semi-axes in region 2 (ra, rb) = 80 nm, 130 nm
Ellipse geometry in region 2 selected to create the desired dispersion and modal overlap.

assumptions (3)

domain assumption The scalar TE wave equation (Eq. 1) with out-of-plane electric field adequately describes the electromagnetic response of the dielectric array.
Invoked at the start of the modeling; assumes time-harmonic, z-invariant fields and ignores vectorial coupling at the hole edges.
domain assumption Floquet-Bloch band structures of the two infinite periodic regions predict the behavior of the finite, abruptly joined device.
Dispersion curves in Fig. 3 and Fig. S2 are used to select operating frequencies and to assert mode overlap, without a full multimode matching at the interface.
standard math Mode-coupling phase matching (Eq. 3), with reciprocal lattice vector Lambda = 2*pi/a2, governs the emission angle of the reversed beam.
This is standard grating/waveguide coupling theory; the paper uses it to interpret the Umklapp process but does not verify it quantitatively against the simulations.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Ultra-thin, entirely flat, Umklapp lenses." pith.science (2026). https://pith.science/paper/34GFFXPU

@misc{pith2026190807797,
  author       = {Pith},
  title        = {Pith review of: Ultra-thin, entirely flat, Umklapp lenses},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/34GFFXPU}},
  note         = {Machine review of arXiv:1908.07797}
}

abstract

We design ultra-thin, entirely flat, dielectric lenses using crystal momentum transfer, so-called Umklapp processes, achieving the required wave control for a new mechanism of flat lensing; physically, these lenses take advantage of abrupt changes in the periodicity of a structured line array so there is an overlap between the first Brillouin zone of one medium with the second Brillouin zone of the other. At the interface between regions of different periodicity, surface, array guided, waves hybridise into reversed propagating beams directed into the material exterior to the array. This control, and redirection, of waves then enables the device to operate as a Pendry-Veselago lens that is one unit cell in width, with no need for an explicit negative refractive index. Simulations using an array embedded in a slab of silicon nitride ($\text{Si}_3\text{N}_4$) in air, operating at visible wavelengths between $420 - 500\text{THz}$ demonstrate the effect.

Figures

Figures reproduced from arXiv: 1908.07797 by the authors.

**Figure 2.** Typical electric fields, of similar mode symmetry, [PITH_FULL_IMAGE:figures/full_fig_p002_2.png] view at source ↗

**Figure 4.** Mode coupling and conversion: (a,c) Conventional [PITH_FULL_IMAGE:figures/full_fig_p003_4.png] view at source ↗

Figures from the paper (1 more)

**Figure 5.** Figure 5: (a) Normalised electric field, |E|, for lossy case with r = 9.7 + 0.2i, excited by line source at the centre of the array at frequency 484 THz. The effect is less pronounced due to weaker propagation along the array. (b) Comparison of normalised electric field norm be…

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

35 extracted references · 32 canonical work pages

[1]

J. B. Pendry, Physical Review Letters 85, 3966 (2000)

work page 2000
[2]

Yu and F

N. Yu and F. Capasso, Nature Materials 13, 139 (2014)

2014
[3]

Lalanne and P

P. Lalanne and P. Chavel, Laser & Photonics Reviews 11, 1600295 (2017)

2017
[4]

Verhagen, R

E. Verhagen, R. de Waele, L. Kuipers, and A. Polman, Physical Review Letters 105, 223901 (2010)

work page 2010
[5]

T. Xu, A. Agrawal, M. Abashin, K. J. Chau, and H. J. Lezec, Nature 497, 470 (2013)

work page 2013
[6]

M. K. T. Al-Nuaimi, W. Hong, and W. X. Zhang, 2014 Inter- national Workshop on Antenna Technology: Small Antennas, Novel EM Structures and Materials, and Applications (iW AT) , 277 (2014)

work page 2014
[7]

Khorasaninejad, W

M. Khorasaninejad, W. T. Chen, R. C. Devlin, J. Oh, A. Y . Zhu, and F. Capasso, Science (New York, N.Y .)352, 1190 (2016)

work page 2016
[8]

Genevet, F

P. Genevet, F. Capasso, F. Aieta, M. Khorasaninejad, and R. Devlin, Review Article 4, 139 (2017)

work page 2017

Show all 35 references

[9]

Y . Jin, R. Kumar, O. Poncelet, O. Mondain-Monval, and T. Brunet, Nature Communications 10, 143 (2019)

2019
[10]

C. Luo, S. G. Johnson, J. D. Joannopoulos, and J. B. Pendry, RAPID COMMUNICATIONS PHYSICAL REVIEW B 65 (2011), 10.1103/PhysRevB.65.201104

2011 doi
[11]

Banerji, M

S. Banerji, M. Meem, A. Majumder, F. G. Vasquez, B. Sensale- Rodriguez, and R. Menon, Optica 6, 805 (2019)

2019
[12]

Peierls, Annalen der Physik 395, 1055 (1929)

R. Peierls, Annalen der Physik 395, 1055 (1929)

1929
[13]

K. L. Tsakmakidis, A. D. Boardman, and O. Hess, Nature 450, 397 (2007)

2007
[14]

Chaplain, M

G. Chaplain, M. Makwana, and R. Craster, Wave Motion 86, 162 (2019)

2019
[15]

Colombi, D

A. Colombi, D. Colquitt, P. Roux, S. Guenneau, and R. V . Cras- ter, Sci. Rep. 6, 27717 (2016)

2016
[16]

Colombi, V

A. Colombi, V . Ageeva, R. J. Smith, A. Clare, R. Patel, M. Clark, D. Colquitt, P. Roux, S. Guenneau, and R. V . Craster, Scientiﬁc Reports 7, 6750 (2017)

2017
[17]

J. M. D. Ponti, A. Colombi, R. Ardito, F. Braghin, A. Corigliano, and R. V . Craster, (2019), arXiv:1907.09297

2019 arXiv
[18]

G. J. Chaplain and R. V . Craster, Phys. Rev. B 99, 220102 (2019)

2019
[19]

Naqvi, IEEE Transactions on Antennas and Propagation 38, 584 (1990)

S. Naqvi, IEEE Transactions on Antennas and Propagation 38, 584 (1990)

1990
[20]

H. O. Pierson, Handbook of chemical vapor deposition: princi- ples, technology and applications (William Andrew, 1999)

1999
[21]

A. Zhan, S. Colburn, R. Trivedi, T. K. Fryett, C. M. Dodson, and A. Majumdar, ACS Photonics 3, 209 (2016)

2016
[22]

R. H. Ritchie, Physical Review 106, 874 (1957)

1957
[23]

J. B. Pendry, L. Mart´ın-Moreno, and F. J. Garcia-Vidal, Science (New York, N.Y .)305, 847 (2004)

2004
[24]

Mu ˜noz, G

P. Mu ˜noz, G. Mic ´o, L. Bru, D. Pastor, D. P ´erez, J. Dom ´enech, J. Fern´andez, R. Ba˜nos, B. Gargallo, R. Alemany,et al., Sensors 17, 2088 (2017)

2017
[25]

A. H. Ghadimi, S. A. Fedorov, N. J. Engelsen, M. J. Bereyhi, R. Schilling, D. J. Wilson, and T. J. Kippenberg, Science 360, 764 (2018)

2018
[26]

Y . T. Lo,Antenna handbook: antenna fundamentals and mathe- matical techniques(Springer Science & Business Media, 1993)

1993
[27]

Wang and W

Z. Wang and W. Dou, Journal of Electromagnetic waves and Applications 20, 1643 (2006)

2006
[28]

Brillouin, Wave propagation in periodic structures: electric ﬁlters and crystal lattices (Courier Corporation, 2003)

L. Brillouin, Wave propagation in periodic structures: electric ﬁlters and crystal lattices (Courier Corporation, 2003)

2003
[29]

R. E. Peierls, Quantum theory of solids (Clarendon Press, 1996)

1996
[30]

P. L. Taylor, P. L. Taylor, and O. Heinonen, A quantum ap- proach to condensed matter physics (Cambridge University Press, 2002)

2002
[31]

A. A. Maznev and O. B. Wright, American Journal of Physics 82, 1062 (2014)

2014
[32]

J. R. Wallbank, R. Krishna Kumar, M. Holwill, Z. Wang, G. H. Auton, J. Birkbeck, A. Mishchenko, L. A. Ponomarenko, K. Watanabe, T. Taniguchi, K. S. Novoselov, I. L. Aleiner, A. K. Geim, and V . I. Fal’ko, Nature Physics15, 32 (2019)

2019
[33]

Bauer, The Journal of Chemical Physics 26, 1440 (1957)

E. Bauer, The Journal of Chemical Physics 26, 1440 (1957)

1957
[34]

S. L. Chuang, Physics of photonic devices, V ol. 80 (John Wiley & Sons, 2012)

2012
[35]

A. A. Maradudin, Structured surfaces as optical metamaterials (Cambridge University Press, 2011) Chap. 5, pp. 158–181. 6 ULTRA-THIN, ENTIRELY FLAT, UMKLAPP LENSES SUPPLEMENTAL MATERIAL To demonstrate the ﬂat lensing behaviour over a broad range of frequencies, we present anoth...

2011

Pith tools

Reviewed August 14, 2026 · model on record in the stance chip above.

[1] [1]

J. B. Pendry, Physical Review Letters 85, 3966 (2000)

work page 2000

[2] [2]

Yu and F

N. Yu and F. Capasso, Nature Materials 13, 139 (2014)

2014

[3] [3]

Lalanne and P

P. Lalanne and P. Chavel, Laser & Photonics Reviews 11, 1600295 (2017)

2017

[4] [4]

Verhagen, R

E. Verhagen, R. de Waele, L. Kuipers, and A. Polman, Physical Review Letters 105, 223901 (2010)

work page 2010

[5] [5]

T. Xu, A. Agrawal, M. Abashin, K. J. Chau, and H. J. Lezec, Nature 497, 470 (2013)

work page 2013

[6] [6]

M. K. T. Al-Nuaimi, W. Hong, and W. X. Zhang, 2014 Inter- national Workshop on Antenna Technology: Small Antennas, Novel EM Structures and Materials, and Applications (iW AT) , 277 (2014)

work page 2014

[7] [7]

Khorasaninejad, W

M. Khorasaninejad, W. T. Chen, R. C. Devlin, J. Oh, A. Y . Zhu, and F. Capasso, Science (New York, N.Y .)352, 1190 (2016)

work page 2016

[8] [8]

Genevet, F

P. Genevet, F. Capasso, F. Aieta, M. Khorasaninejad, and R. Devlin, Review Article 4, 139 (2017)

work page 2017