Optical computation of the divergence of a vector field using a metal-dielectric multilayer

Artem I. Kashapov; Dmitry A. Bykov; Evgeni A. Bezus; Leonid L. Doskolovich

Optical computation of the divergence of a vector field using a metal-dielectric multilayer

T0 review · 0 major / 4 minor · reviewed 2026-08-07 · deepseek-v4-flash

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

Pith's one-line read A metal-dielectric multilayer satisfying a TM reflection zero and a derivative-matching condition computes the divergence of the incident transverse field in a single reflected component, and the same stack also yields the gradient…

desk verdict A clean, honest design for all-optical divergence computation with a simple multilayer; the main limitation is the stated narrow angular bandwidth. read the letter →

arxiv 2505.21023 v1 pith:UBP5RIIP submitted 2025-05-27 physics.optics

Leonid L. Doskolovich , Artem I. Kashapov , Evgeni A. Bezus , Dmitry A. Bykov This is my paper

classification physics.optics

keywords analogopticalcomputingdivergenceoperatorgradientmetal-dielectricmultilayerspinHalleffectoflightobliqueincidenceisotropicdifferentiationvectorvortexbeams

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

The paper claims that a single passive multilayer can compute the divergence of a two-dimensional vector field encoded in the transverse electric field of an obliquely incident beam. In the linear (paraxial) approximation, the reflected beam's transverse $x$-component becomes proportional to $\operatorname{div}\mathbf{E}_{\mathrm{inc},\perp}$ when the structure satisfies $r_{\mathrm{TM}}(\theta_0)=0$ and $r'_{\mathrm{TM}}(\theta_0)=-r_{\mathrm{TE}}(\theta_0)\cot\theta_0$. For a linearly polarized input, the same conditions make the structure output the gradient of the scalar field and an intensity pattern proportional to $|\nabla f|^2$, and a second reflection computes the Laplacian. As a concrete realization, the authors design a four-layer Au/TiO$_2$ stack on a gold substrate and report numerical simulations reproducing divergence, gradient, directional derivatives, and double-reflection Laplacians with normalized RMS errors of a few percent. If correct, this turns divergence, gradient, and Laplacian into compact analog optical operations for edge detection and polarization-singularity analysis.

What carries the argument

The carrying object is the vectorial transfer function of Eq. (15), which maps the incident transverse field to the reflected field through TM and TE plane-wave coefficients. Expanding this transfer function to first order in the transverse wave vector gives the approximate forms of Eqs. (16)–(17); the two conditions in Eq. (22) collapse the $x$-component of the reflected field to the divergence operator. The design problem reduces to satisfying a complex-valued condition on the reflection coefficients, which is why the structure needs the two extra thicknesses of a second metal-dielectric pair: the lower layers act as an effective substrate in the reflection-zero equations (27)–(28), providing the free parameters that make all conditions simultaneously satisfiable.

What would settle it

Fabricate the designed stack and illuminate it at $633$ nm and $25^\circ$ incidence with a vector vortex beam of order $m=3$: the reflected $x$-component should vanish along the four angular directions $\varphi=\pi/4+n\pi/2$, and any measurable displacement or loss of these nodal lines would show the transfer function is not the divergence operator. Alternatively, compute the rigorous transfer functions at $|k_x|/k_0=0.2$ and compare them with the linear approximation, which must show growing discrepancy with spatial frequency and thereby delineate the paraxial validity limit.

Watch

Extended reading notes

Core claim

The central claim is that computing the divergence of a vector light field reduces to engineering a layered structure whose TM reflection coefficient vanishes at the design angle and whose TM derivative at that angle is locked to the TE coefficient: $r_{\mathrm{TM}}(\theta_0)=0$ and $r'_{\mathrm{TM}}(\theta_0)=-r_{\mathrm{TE}}(\theta_0)\cot\theta_0$. Under these conditions, the linearized vectorial transfer function of the structure makes the reflected field's $x$-component equal to $-i\alpha_x\,\operatorname{div}\mathbf{E}_{\mathrm{inc},\perp}$ up to the proportionality factor, with the cross-polarization derivative term attributed to the optical analogue of the spin Hall effect and the reflection zero to a resonant (critical-coupling) effect. The same conditions yield the gradient of a scalar field for a linearly polarized input beam, and two successive reflections compute the Laplacian. A four-layer Au/TiO$_2$ multilayer with thicknesses $h_{m,1}=5.2$ nm, $h_{d,1}=39.6$ nm, $h_{m,2}=52.5$ nm, $h_{d,2}=70.7$ nm is found to satisfy the conditions at $\lambda=633$ nm and $\theta_0=25^\circ$, and rigorous simulations confirm the operators for directional derivatives, gradient, isotropic differentiation, a double-reflection Laplacian, and the divergence of a Gauss–Bessel vector vortex beam of order $m=3$.

Load-bearing premise

The whole construction presupposes that the incident beam's angular spectrum is concentrated near the design direction $k_\perp=0$, so the first-order Taylor expansion of the transfer function is accurate; the paper's accuracy figures are stated for $|k_x|/k_0<0.05$ and $|k_y|/k_0<0.05$, and beams with wider angular content will not be transformed as the exact divergence operator.

Editorial extensions

If this is right

A single passive multilayer implements divergence, gradient, and isotropic differentiation of optical fields in reflection, making these vector operations available at the speed of light.
Because $\operatorname{div}\operatorname{grad} f=\Delta f$, the same structure computes the Laplacian of an input scalar field with two consecutive reflections, giving a compact all-optical way to obtain a second-order operator.
For vector vortex beams, the divergence output vanishes at $2m-2$ angular positions determined by the singularity order, so the reflected pattern directly reveals the polarization-singularity structure.
Rotating a linear polarizer changes the incident polarization direction and thereby selects the direction along which the derivative is taken, giving orientation-tunable edge enhancement.

Reading between the lines

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

The same transfer-function engineering could be applied to other linear differential operators: any operator whose Fourier symbol is a homogeneous linear function of $k_\perp$ could in principle be realized by an analogous pair of conditions, making this a template for a broader class of vector derivatives and mixed operators.
Because the two derivative coefficients $\alpha_x$ and $\alpha_y$ are complex and only approximately equal in the designed stack, modest depolarization or anisotropic phase response may appear for inputs with sizable angular content; a quantitative study of the complex-valued mismatch would reveal whether the quoted RMS errors persist across polarization states.
The double-reflection Laplacian idea suggests a cascade architecture in which multiple reflections from identical stacks implement higher-order differential operators, at the cost of optical loss per pass; the paper does not analyze that trade-off.
A natural experimental follow-up is to fabricate the stack and image the reflected field of a known vector field; comparison of measured and predicted nodal lines would simultaneously test the paraxial and material assumptions.

Share X Bluesky LinkedIn Reddit HN

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

Short version: this is a clean theoretical design paper that does what it claims, within a clearly stated paraxial limit. The new thing is the divergence conditions (Eq. 22) and a concrete four-layer Au/TiO2 stack that satisfies them; the same structure also gives gradient and, with two reflections, Laplacian. The math is straightforward and internally consistent: Taylor expand the vectorial transfer functions, require a zero TM reflection and matching slopes, and the reflected x-component is proportional to the divergence. There are no fitted parameters; the layer thicknesses are chosen to satisfy the derived conditions and the simulations use rigorous transfer-matrix calculations checked against analytic derivatives. The RMS errors (0.4% to 1.6%) are respectable for this kind of design.

The main caveat, correctly flagged in the stress test, is that Eq. (23) is a first-order approximation. The divergence identity holds only for incident spectra with |k_perp|/k0 < 0.05. The paper states this restriction and all the numerical examples use smooth beams well inside that band. So it is an honest limitation of scope, not a hidden flaw. Someone processing sharp edges or high-NA fields will not be served by this particular structure, but the authors never claim it would be.

The only other soft spot I'd note is that the required equality alpha_x = alpha_y is satisfied only approximately (about 1% amplitude mismatch and a small phase offset). That is minor, and the paper says 'approximately satisfied with good accuracy.' For most applications the residual error is probably acceptable; for quantitative work it would matter, but it is not a load-bearing flaw.

Citation pattern looks reasonable. The prior divergence implementations [27,28] are cited and contrasted; the authors' own vectorial differentiation work [13,26] is the natural foundation, and the extension to divergence is clearly delimited. No circular fitting, no invented entities.

Bottom line: this is a solid, honest contribution to analog optical computing. It deserves a serious referee and should be published, likely with a minor revision that more explicitly frames the angular bandwidth as an application constraint and perhaps quantifies the effect of the alpha_x/alpha_y mismatch on the divergence output. I'd cite it if I were working on multilayer optical differentiators. Send it to peer review.

Referee Report

0 major / 4 minor

Summary. The paper derives a vectorial transfer function for reflection of an obliquely incident three-dimensional beam from a layered structure, and shows that when the TM reflection coefficient and its derivative satisfy the two conditions of Eq. (22), the transverse x-component of the reflected field is proportional, in the paraxial (first-order) approximation, to the divergence of the two-dimensional vector field formed by the incident transverse electric field components. The authors design a four-layer Au/TiO2 multilayer on a gold substrate that approximately satisfies these conditions, and verify by rigorous plane-wave simulations that the structure computes directional derivatives, gradients, the Laplacian (via two successive reflections), and the divergence of vector vortex beams, with normalized RMS deviations between 0.4% and 1.6% against analytic references.

Significance. If the result holds, this is a substantial practical advance: it replaces the previously proposed plasmonic-grating-plus-phase-plate scheme and the bulky tetrahedron-based structure with a compact, planar multilayer that computes the divergence operator in reflection. The derivation is algebraically clean and self-consistent, with no fitted parameters; the layer thicknesses are design variables chosen from resonance conditions, and the transfer-function approximations are validated against rigorous calculations. The numerical demonstrations are carefully performed using the full transfer functions and checked against independently computed analytic derivatives, divergence, and Laplacian. The paper also contributes a useful closed-form expression for the divergence zeros of general vector vortex beams, Eq. (34). These strengths make the work a credible contribution to analog optical computing.

minor comments (4)

[Section 5, Eq. (29)] The designed multilayer satisfies the condition alpha_x = alpha_y only approximately: the reported values 0.260*exp(-2.995i) and 0.263*exp(-2.998i) differ by about 1% in amplitude and by 0.003 rad in phase. Consequently, Eq. (23) is not exact for the designed structure; an additional term proportional to (alpha_y - alpha_x) * dE_y/dy appears. The numerical examples show small RMS errors, but the paper does not quantify the maximum systematic error introduced by this mismatch for arbitrary in-band inputs. A short paragraph giving an error bound for |kx|/k0, |ky|/k0 < 0.05 would strengthen the claim.
[Section 3, Eqs. (16)-(23)] The divergence identity in Eq. (23) is derived in the linear (paraxial) approximation and is therefore conditional on the incident plane-wave spectrum being concentrated near k_perp = 0. The paper states this and shows in Fig. 3 that the rigorous transfer functions deviate from the linear approximations by at most 2.3% for |kx|/k0, |ky|/k0 < 0.05. It would be helpful to add an explicit sentence in the conclusion reminding readers of this bandwidth limitation and its practical implication for the minimum feature size of the processed field.
[Section 6.2, Figs. 5(b)-5(e)] In the double-reflection Laplacian demonstration, the field produced after the first reflection is only approximately equal to the gradient (due to the coefficient mismatch and higher-order transfer-function terms). The paper reports RMS deviations of 1.6% for the gradient and 1.08% for the Laplacian, but it does not comment on whether these deviations are dominated by the alpha-coefficient mismatch or by the finite spectral width of the input. A brief analysis would clarify the origin of the residual errors.
[Throughout] There are a few typographical and formatting issues: the abstract contains a spurious space in "di vergence"; the vectors ATM and ATE in Eq. (15) are evaluated with theta0 replaced by -theta0 and kz = +sqrt(...), which is stated in the text but could be repeated explicitly in the equation display for clarity; and the sentence after Eq. (22) says "in the/u1D465reﬂ electric field component" with a stray slash. These do not affect the technical content.

Circularity Check

0 steps flagged · score 1.0 of 10

No significant circularity: the divergence identity is a first-order Taylor construction whose design conditions are independently verified against rigorous simulation and analytic operators.

full rationale

The paper derives Eq. (23) from the plane-wave expansion of the reflected field and the first-order Taylor approximations of the transfer functions in Eqs. (16) and (17). The conditions in Eq. (22) are chosen so that, in this linear approximation, the reflected x-component becomes proportional to div E_inc,perp. This is a constructive mathematical derivation, not a fit renamed as a prediction: the layer thicknesses in Eq. (29) are design variables found by satisfying the derived reflection-zero and slope-matching conditions, and no parameter is fitted to the divergence, gradient, or Laplacian outputs. The numerical demonstrations compare the rigorously computed reflected fields (Eqs. (14) and (15)) against independently calculated analytic divergence, gradient, and Laplacian, with reported RMS deviations (0.4%, 1.6%, 1.08%, 1.3%). The self-citations (e.g., refs. [10,11,13,26]) supply background design recipes for reflection zeros and prior differentiation frameworks; they are not used as a uniqueness theorem or as the sole justification for the central identity. The main caveat is the explicitly stated narrow-spectrum assumption (|k_x|/k0, |k_y|/k0 < 0.05 in Fig. 3), but this is a stated approximation of scope, not a circular step. Overall, the derivation is self-contained and the circularity burden is very low.

Assumptions & free parameters 2 free parameters · 4 assumptions · 0 invented entities

The central claim rests on standard plane-wave scattering theory and a stated narrow-spectrum assumption. The only hand-chosen numbers are the four layer thicknesses, which are design variables rather than fitted constants. No new physical entities are introduced. The design search does not create circularity, because the verification compares full transfer-matrix simulations against independently computed analytic operators.

free parameters (2)

lower layer thicknesses h_m2, h_d2 = 52.5 nm, 70.7 nm
Chosen by numerical search to approximately satisfy the divergence conditions Eq. (22); they are engineering design parameters, not data-fitted constants, but they set the accuracy of the final device.
upper layer thicknesses h_m1, h_d1 = 5.2 nm, 39.6 nm
Determined from the reflection-zero conditions Eqs. (27)-(28) once the lower layers are fixed; design parameters for the specific device.

assumptions (4)

domain assumption The incident beam is represented as a superposition of TM and TE plane waves, each multiplied by a scalar reflection coefficient r_TM(theta) or r_TE(theta) upon reflection from the multilayer.
Standard electromagnetic model for a laterally homogeneous layered structure; introduced in Section 2.2 and used to derive Eqs. (14)-(15).
domain assumption The angular spectrum of the incident beam is concentrated near k_perp=0, allowing the vectorial transfer functions to be replaced by their first-order Taylor expansions.
Stated at the start of Section 3; this is the paraxial premise that limits the range of validity of the divergence computation.
domain assumption The dielectric permittivities of gold and TiO2 from the reference database [33] accurately describe the materials.
Used in the design search and in the numerical simulations; no sensitivity analysis or experimental data are provided.
domain assumption A reflection zero can be obtained for the chosen materials by solving Eqs. (27)-(28), following the critical-coupling design from prior work [10,11].
The design procedure assumes the existence and computability of such a reflection zero for each lower-layer pair.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Optical computation of the divergence of a vector field using a metal-dielectric multilayer." pith.science (2026). https://pith.science/paper/UBP5RIIP

@misc{pith2026250521023,
  author       = {Pith},
  title        = {Pith review of: Optical computation of the divergence of a vector field using a metal-dielectric multilayer},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/UBP5RIIP}},
  note         = {Machine review of arXiv:2505.21023}
}

read the original abstract

We theoretically describe the optical computation of the divergence of a two-dimensional vector field, which is composed by the transverse electric field components of an incident light beam. The divergence is computed in reflection at oblique incidence of light on a layered structure. We show that in the particular case of a linearly polarized incident beam, the layered structure implementing the divergence operator also allows one to compute the gradient and perform the isotropic differentiation. As an example of a layered structure computing the divergence, we propose a metal-dielectric multilayer consisting of two pairs of metal and dielectric layers on a metal substrate. The presented numerical simulation results of the designed multilayer confirm that the divergence operator is computed with high accuracy. We also demonstrate the possibility of using the designed structure for optical directional differentiation and computation of the gradient and Laplace operators.

Figures

Figures reproduced from arXiv: 2505.21023 by the authors.

**Figure 2.** Geometry of the two-layer metal-dielectric-metal s [PITH_FULL_IMAGE:figures/full_fig_p008_2.png] view at source ↗

**Figure 3.** Absolute values of the calculated TFs Θ, (k⊥) (a) and Θ, (k⊥) (b) describing the formation of the refl electric field component of the beam reflected from the multilayer of Eq. (29). (c), (d) Cross-sections of (a) and (b) along the corresponding solid lines. 6.1. Computing a directional derivative First, let us consider the simplest case of an incident beam linearly polarized along a certain direction P() = (cos , s… view at source ↗

Figures from the paper (3 more)

**Figure 4.** Figure 4: Incident Gaussian beam [PITH_FULL_IMAGE:figures/full_fig_p010_4.png]

**Figure 5.** Figure 5: (a) Incident field corresponding to the image of the na [PITH_FULL_IMAGE:figures/full_fig_p011_5.png]

**Figure 6.** Figure 6: (a), (b) - and - electric field components of the incident beam defined by Eq. (36), (c) the calculated refl electric field component of the reflected beam refl, formed upon reflection of the incident field shown in (a) and (b) from the designed metal-dielectric multil…

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

34 extracted references · 34 canonical work pages

[1]

Performing mathematical operations with metamaterials ,

A. Silva, F. Monticone, G. Castaldi, et al., “Performing mathematical operations with metamaterials ,” Science 343, 160–163 (2014)

work page 2014
[2]

Inver se-designed metastructures that solve equations,

N. Mohammadi Estakhri, B. Edwards, and N. Engheta, “Inver se-designed metastructures that solve equations,” Science 363, 1333–1338 (2019)

work page 2019
[3]

Fla t optics for image diﬀerentiation,

Y . Zhou, H. Zheng, I. I. Kravchenko, and J. Valentine, “Fla t optics for image diﬀerentiation,” Nat. Photonics 14, 316–323 (2020)

work page 2020
[4]

Temporal diﬀerentiation of optical signals using resonant gratings,

D. A. Bykov, L. L. Doskolovich, and V . A. Soifer, “Temporal diﬀerentiation of optical signals using resonant gratings,” Opt. Lett. 36, 3509–3511 (2011)

work page 2011
[5]

Optical spatial diﬀeren tiator based on subwavelength high-contrast gratings,

Z. Dong, J. Si, X. Yu, and X. Deng, “Optical spatial diﬀeren tiator based on subwavelength high-contrast gratings,” Appl. Phys. Lett. 112, 181102 (2018)

work page 2018
[6]

First-order optical spatial diﬀerentiator based on a guided-mode resonant grating,

D. A. Bykov, L. L. Doskolovich, A. A. Morozov, et al. , “First-order optical spatial diﬀerentiator based on a guided-mode resonant grating,” Opt. Express 26, 10997–11006 (2018)

work page 2018
[7]

Plasmonic transmitt ed optical diﬀerentiator based on the subwavelength gold gratings,

W. Y ang, X. Yu, J. Zhang, and X. Deng, “Plasmonic transmitt ed optical diﬀerentiator based on the subwavelength gold gratings,” Opt. Lett. 45, 2295–2298 (2020)

work page 2020
[8]

Spatiotemporal di ﬀerentiators generating optical vortices with transverse orbital angular momentum and detecting sharp change of puls e envelope,

J. Huang, J. Zhang, T. Zhu, and Z. Ruan, “Spatiotemporal di ﬀerentiators generating optical vortices with transverse orbital angular momentum and detecting sharp change of puls e envelope,” Laser & Photonics Rev. 16, 2100357 (2022)

work page 2022

Show all 34 references

[9]

Spatial diﬀerentiation of optical beams using phase-shifted Bragg grating,

L. L. Doskolovich, D. A. Bykov, E. A. Bezus, and V . A. Soifer , “Spatial diﬀerentiation of optical beams using phase-shifted Bragg grating,” Opt. Lett. 39, 1278–1281 (2014)

2014
[10]

Spatial diﬀerentiation of optical beams using a resonant metal-dielectric-metal structure,

A. I. Kashapov, L. L. Doskolovich, E. A. Bezus, et al., “Spatial diﬀerentiation of optical beams using a resonant metal-dielectric-metal structure,” J. Opt. 23, 023501 (2021)

2021
[11]

Optical properties of cascaded metal-dielectric- metal structures and their application to the diﬀerentiation of optical signals,

L. L. Doskolovich, A. I. Kashapov, E. A. Bezus, and D. A. By kov, “Optical properties of cascaded metal-dielectric- metal structures and their application to the diﬀerentiation of optical signals,” Photonics Nanostructures — Fundam. Appl. 52, 101069 (2022)

2022
[12]

Spatiotemporal optical diﬀerentiation and vortex generation with metal-dielectric-metal multilayers,

L. L. Doskolovich, A. I. Kashapov, E. A. Bezus, and D. A. Bykov, “Spatiotemporal optical diﬀerentiation and vortex generation with metal-dielectric-metal multilayers,” Phys. Rev. A 106, 033523 (2022)

2022
[13]

Vectorial spatial diﬀerentiation of optical beams with metal–dielectric multilayers enabled by spin Hall eﬀe ct of light and resonant reﬂection zero,

L. L. Doskolovich, A. I. Kashapov, E. A. Bezus, and D. A. By kov, “Vectorial spatial diﬀerentiation of optical beams with metal–dielectric multilayers enabled by spin Hall eﬀe ct of light and resonant reﬂection zero,” Opt. & Laser Technol. 181, 111884 (2025)

2025
[14]

Plasmonic computing of spatial diﬀerentiation,

T. Zhu, Y . Zhou, Y . Lou,et al., “Plasmonic computing of spatial diﬀerentiation,” Nat. Co mmun. 8, 15391 (2017)

2017
[15]

Analogue optical spatiotemporal diﬀerentiator,

Y . Zhou, J. Zhan, R. Chen, et al., “ Analogue optical spatiotemporal diﬀerentiator,” Adv. O pt. Mater. 9, 2002088 (2021)

2021
[16]

Temporal diﬀerentiation of optical signals using a phase -shifted ﬁber Bragg grating,

N. K. Berger, B. Levit, B. Fischer, et al. , “Temporal diﬀerentiation of optical signals using a phase -shifted ﬁber Bragg grating,” Opt. Express 15, 371–381 (2007)

2007
[17]

Design of high-order all-optical temporal diﬀerentiators based on multiple-phase-shifted ﬁber Bragg gratings,

M. Kulishov and J. Azaña, “Design of high-order all-optical temporal diﬀerentiators based on multiple-phase-shifted ﬁber Bragg gratings,” Opt. Express 15, 6152–6166 (2007)

2007
[18]

High-order photonic diﬀerentiator employing on-chip ca scaded microring resonators,

J. Dong, A. Zheng, D. Gao, et al. , “High-order photonic diﬀerentiator employing on-chip ca scaded microring resonators,” Opt. Lett. 38, 628–630 (2013)

2013
[19]

Use of photonic crystal cavities for temporal diﬀerentiation of optical signals,

N. L. Kazanskiy, P . G. Seraﬁmovich, and S. N. Khonina, “Use of photonic crystal cavities for temporal diﬀerentiation of optical signals,” Opt. Lett. 38, 1149–1151 (2013)

2013
[20]

Subpicosecond ﬂat- top pulse shaping using a hybrid plasmonic microring- based temporal diﬀerentiator,

A. Karimi, A. Zarifkar, and M. Miri, “Subpicosecond ﬂat- top pulse shaping using a hybrid plasmonic microring- based temporal diﬀerentiator,” J. Opt. Soc. Am. B 36, 1738–1747 (2019)

2019
[21]

Analog computing by Brewster eﬀect,

A. Y ousseﬁ, F. Zangeneh-Nejad, S. Abdollahramezani, an d A. Khavasi, “ Analog computing by Brewster eﬀect,” Opt. Lett. 41, 3467–3470 (2016)

2016
[22]

Optical diﬀerentiation based on the Brewster eﬀect,

D. Nesterenko, M. Kolesnikova, and A. Lyubarskaya, “Optical diﬀerentiation based on the Brewster eﬀect,” Comput. Opt. 42, 758–763 (2018)

2018
[23]

Optical computation of the Laplace operator using phase-shifted Bragg grating,

D. A. Bykov, L. L. Doskolovich, E. A. Bezus, and V . A. Soifer, “Optical computation of the Laplace operator using phase-shifted Bragg grating,” Opt. Express 22, 25084–25092 (2014)

2014
[24]

Photonic crystal slab Laplace operator for image diﬀeren tiation,

C. Guo, M. Xiao, M. Minkov, et al., “Photonic crystal slab Laplace operator for image diﬀeren tiation,” Optica 5, 251–256 (2018)

2018
[25]

Laplace metasurfaces for optical analog computing based on quasi-bound states in the continuum,

D. Pan, L. Wan, M. Ouyang, et al., “Laplace metasurfaces for optical analog computing based on quasi-bound states in the continuum,” Photonics Res. 9, 1758–1766 (2021)

2021
[26]

Optical computation of the Laplace operator at oblique incidence using a multilayer metal-dielectric structure,

L. L. Doskolovich, A. I. Kashapov, E. A. Bezus, et al., “Optical computation of the Laplace operator at oblique incidence using a multilayer metal-dielectric structure, ” Opt. Express 31, 17050–17064 (2023)

2023
[27]

Optical computation of divergence operation for vector ﬁelds,

Y . Lou, Y . Fang, and Z. Ruan, “Optical computation of divergence operation for vector ﬁelds,” Phys. Rev. Appl. 14, 034013 (2020)

2020
[28]

All optical div ergence and gradient operators using surface plasmon polaritons,

H. Mohammadi, M. Akbari, and A. Khavasi, “ All optical div ergence and gradient operators using surface plasmon polaritons,” Opt. Express 30, 17806–17823 (2022)

2022
[29]

Generalized spatial diﬀerentiation from the spin Hall eﬀect of light and its application in image processing of edge detection,

T. Zhu, Y . Lou, Y . Zhou,et al., “Generalized spatial diﬀerentiation from the spin Hall eﬀect of light and its application in image processing of edge detection,” Phys. Rev. Appl. 11, 034043 (2019)

2019
[30]

Optical phase mining by adjustable spatial diﬀerentiator,

T. Zhu, J. Huang, and Z. Ruan, “Optical phase mining by adjustable spatial diﬀerentiator,” Adv. Photonics 2, 016001 (2020)

2020
[31]

Born and E

M. Born and E. Wolf, Principles of Optics (Cambridge University Press, 1999), 7th ed

1999
[32]

Stable implementation of the rigorous coupled-wave analysis for surface-relief gratings: enhanced transmitt ance matrix approach,

M. G. Moharam, T. K. Gaylord, D. A. Pommet, and E. B. Grann,“Stable implementation of the rigorous coupled-wave analysis for surface-relief gratings: enhanced transmitt ance matrix approach,” J. Opt. Soc. Am. A 12, 1077–1086 (1995)

1995
[33]

Refractiveindex.info database of op tical constants,

M. N. Polyanskiy, “Refractiveindex.info database of op tical constants,” Sci. Data 11, 94 (2024)

2024
[34]

Gouy phase induced polarization transition of focused ve ctor vortex beams,

Y . Zhang, X. Guo, L. Han, et al., “Gouy phase induced polarization transition of focused ve ctor vortex beams,” Opt. Express 25, 25725–25733 (2017)

2017

Pith tools

Reviewed August 7, 2026 · model on record in the stance chip above.

[1] [1]

Performing mathematical operations with metamaterials ,

A. Silva, F. Monticone, G. Castaldi, et al., “Performing mathematical operations with metamaterials ,” Science 343, 160–163 (2014)

work page 2014

[2] [2]

Inver se-designed metastructures that solve equations,

N. Mohammadi Estakhri, B. Edwards, and N. Engheta, “Inver se-designed metastructures that solve equations,” Science 363, 1333–1338 (2019)

work page 2019

[3] [3]

Fla t optics for image diﬀerentiation,

Y . Zhou, H. Zheng, I. I. Kravchenko, and J. Valentine, “Fla t optics for image diﬀerentiation,” Nat. Photonics 14, 316–323 (2020)

work page 2020

[4] [4]

Temporal diﬀerentiation of optical signals using resonant gratings,

D. A. Bykov, L. L. Doskolovich, and V . A. Soifer, “Temporal diﬀerentiation of optical signals using resonant gratings,” Opt. Lett. 36, 3509–3511 (2011)

work page 2011

[5] [5]

Optical spatial diﬀeren tiator based on subwavelength high-contrast gratings,

Z. Dong, J. Si, X. Yu, and X. Deng, “Optical spatial diﬀeren tiator based on subwavelength high-contrast gratings,” Appl. Phys. Lett. 112, 181102 (2018)

work page 2018

[6] [6]

First-order optical spatial diﬀerentiator based on a guided-mode resonant grating,

D. A. Bykov, L. L. Doskolovich, A. A. Morozov, et al. , “First-order optical spatial diﬀerentiator based on a guided-mode resonant grating,” Opt. Express 26, 10997–11006 (2018)

work page 2018

[7] [7]

Plasmonic transmitt ed optical diﬀerentiator based on the subwavelength gold gratings,

W. Y ang, X. Yu, J. Zhang, and X. Deng, “Plasmonic transmitt ed optical diﬀerentiator based on the subwavelength gold gratings,” Opt. Lett. 45, 2295–2298 (2020)

work page 2020

[8] [8]

Spatiotemporal di ﬀerentiators generating optical vortices with transverse orbital angular momentum and detecting sharp change of puls e envelope,

J. Huang, J. Zhang, T. Zhu, and Z. Ruan, “Spatiotemporal di ﬀerentiators generating optical vortices with transverse orbital angular momentum and detecting sharp change of puls e envelope,” Laser & Photonics Rev. 16, 2100357 (2022)

work page 2022