Expand this Topic clickable element to expand a topic
Skip to content
Optica Publishing Group

Spatial integration and differentiation of optical beams in a slab waveguide by a dielectric ridge supporting high-Q resonances

Open Access Open Access

Abstract

We show that a very simple structure consisting of a single subwavelength dielectric ridge on the surface of a slab waveguide enables spatial integration and differentiation of the profile of optical beams propagating in the waveguide. The integration and differentiation operations are performed in reflection and in transmission, respectively, at oblique incidence of the beam impinging on the ridge. The implementation of these operations is associated with the resonant excitation of a cross-polarized eigenmode of the ridge. We demonstrate that the quality factor of the resonances strongly varies along the dispersion curves and allows one to achieve the required tradeoff between the integration (or differentiation) quality and the amplitude of the resulting beam. The presented rigorous numerical simulation results confirm high-quality integration and differentiation. The proposed integrated structure may find application in ultrafast all-optical analog computing and signal processing systems.

© 2018 Optical Society of America under the terms of the OSA Open Access Publishing Agreement

1. Introduction

Over the past few years, analog optical computing has attracted increasing attention, since it offers high-performance solution of several important computational tasks. Among the basic operations of analog processing of optical signals are spatial differentiation and integration of the profile of an optical beam. Traditionally, for the implementation of these operations, bulky optical systems consisting of lenses and filters are utilized [1]. Recently, spatial differentiators and integrators based on nanophotonic structures having a thickness comparable to the wavelength of the processed optical signal were proposed. In particular, in [2–5], compact analogues of the optical Fourier correlator are considered, in which the spatial filter comprises a metasurface encoding the complex transmission coefficient of a differentiating or an integrating filter. In [2, 6–16], various resonant structures containing systems of homogeneous layers [2,6–12] and diffraction gratings [13–16] are used for spatial differentiation and integration of optical beams. The possibility to utilize such structures for optical differentiation or integration is based on the fact that the Fano profile describing the reflection or the transmission coefficient of the structure in the vicinity of the resonance can approximate the transfer function of a differentiating or an integrating filter.

Development of planar (on-chip) differentiators and integrators, in which the processed optical signal propagates in some guiding structure, is of great interest. In particular, in a wide class of planar (integrated) optoelectronic systems, spectral or spatial filtering of optical signals is performed in a slab waveguide (“insulator-on-insulator” platform) [17, 18]. In this case, the processed signal corresponds to a superposition of slab waveguide modes with different propagation directions (in the case of spatial filtering) or with different frequencies (in the case of spectral filtering). In a recent work, the present authors proposed a simple planar differentiator based on the so-called W-type structure and consisting of two grooves on the surface of a slab waveguide and operating in reflection [19]. The differentiator operation is associated with the excitation of an eigenmode localized between the grooves. A similar graphene-based W-type spatial integrator operating in transmission was considered in [10].

In this work, we for the first time show that an extremely simple structure consisting of a single subwavelength ridge on the surface of a slab waveguide can be used as a spatial integrator and differentiator. The implementation of these operations is associated with the resonant excitation of a cross-polarized eigenmode of the ridge. We demonstrate that the Q-factor of the resonances strongly varies along the dispersion curves, enabling the structure to be used both as an efficient optical integrator and differentiator. In particular, the Q-factor can be arbitrarily high due to the existence of the bound states in the continuum in the investigated structure [20,21]. That makes this structure especially suitable for analog optical integration. In contrast, high-quality differentiation is achieved at relatively low-Q resonances, which are also supported by the proposed structure. The presented results of rigorous numerical simulations confirm high accuracy of spatial integration and differentiation.

2. Diffraction of slab waveguide modes on a ridge

In order to explain the resonant effects that make it possible to implement spatial integration and differentiation, let us first discuss the diffraction of slab waveguide modes on the ridge located on the surface of a slab waveguide. The geometry of the structure is shown in Fig. 1. For the analysis, the following parameters were chosen: refractive index of the waveguide core layer and of the ridge nc = 3.3212 (GaP at the wavelength λ = 630 nm), refractive indices of the substrate and superstrate nsub = 1.45 (fused silica) and nsup = 1, respectively, waveguide thickness hc = 80 nm, waveguide thickness at the ridge hr = 110 nm. In this case, at λ = 630 nm the waveguide is single-mode for both TE- and TM-polarizations, and the effective refractive indices of the modes amount to nwg,TE = 2.5913 and nwg,TM = 1.6327, respectively. Effective refractive indices of the modes at hr = 110 nm equal nr,TE = 2.8192 and nr,TM = 2.1867.

 figure: Fig. 1

Fig. 1 Geometry of the problem of diffraction of a waveguide mode on a dielectric step.

Download Full Size | PDF

Figure 2 shows the reflectance |RTE (l, θ)|2 and the transmittance |TTE (l, θ)|2 of the ridge vs. the angle of incidence θ and the ridge length l in the case of an incident TE-polarized mode, where RTE (l, θ) and TTE (l, θ) are the complex reflection and transmission coefficients, respectively. The plots were calculated using an in-house implementation of the aperiodic rigorous coupled-wave analysis (aRCWA) technique [22,23]. The RCWA, also called the Fourier modal method, is an established numerical technique for solving Maxwell’s equations.

It is evident that several resonances (sharp reflectance maxima and transmittance minima) are present in the spectra of Fig. 2. The resonant reflectance peaks (transmittance dips) are located in the angle of incidence range 39.05° < θ < 57.55°, which is marked with dashed lines in Fig. 2. It is important to note that the transmittance strictly vanishes at the resonances, and therefore the reflectance reaches unity.

 figure: Fig. 2

Fig. 2 Reflectance (a) and transmittance (b) of the ridge vs. the ridge length l (horizontal axis) and the angle of incidence θ (vertical axis). Dashed lines show the boundaries of the resonance region and correspond to the cutoff angles of the TM-modes outside the ridge (at hc = 80 nm, upper line) and inside the ridge (hr = 110 nm, lower line). Dotted curves show the dispersion of the cross-polarized modes of the ridge. The points marked with white and black asterisks correspond to the spatial integration and differentiation examples considered below, respectively.

Download Full Size | PDF

Let us explain these effects. In a general case, at oblique incidence of a TE-polarized guided mode on a ridge located on the waveguide surface, reflected and transmitted TE- and TM-polarized modes are generated, as well as a continuum of plane waves arising from “parasitic” scattering out of the guiding layer to the superstrate and substrate. However, if the angle of incidence exceeds a certain value θcr,1, the reflected and transmitted fields contain only the TE-polarized mode, and no out-of-plane scattering and polarization conversion occur [19,24]. Indeed, let us denote by kx,inc = k0nwg,TE sin θ the x-component of the wavevector of the incident TE-mode, which is parallel to the ridge boundaries. Here, k0 = 2π/λ is the wavenumber. According to the boundary conditions for the Maxwell’s equations, the kx,inc component has to be conserved, i.e. it is the same for all the waves constituting the reflected and the transmitted fields, including the radiation scattered out of the waveguide. Therefore, at angles of incidence θ > θcr,1 = arcsin (nwg,TM/nwg,TE) = 39.05°, reflected and transmitted TM-polarized modes become evanescent. At θ > θcr,1, kx,inc also exceeds the wave vector magnitudes of the propagating plane waves over and under the waveguide (in the regions with the refractive indices nsup and nsub, respectively) and therefore the incident TE-mode is not scattered out of the waveguide layer. Thus, at θ > θcr,1 the reflection and transmission coefficients corresponding to the TE-polarized mode satisfy the equality |RTE|2 + |TTE|2 = 1 according to the energy conservation law, since the materials of the structure are assumed to be lossless. The angle θcr,1 is shown with a dashed line in Fig. 2, which corresponds to the upper boundary of the “resonance region”. The lower boundary of this region corresponds to the cutoff angle of the TM-polarized mode in the ridge region, which amounts to θcr,2 = arcsin (nr,TM/nr,TE) = 57.55°.

The presented analysis suggests that the resonances in Fig. 2 are associated with the excitation of quasi-TM eigenmodes of the ridge, which in this case acts as a leaky rib waveguide. Within the framework of the effective index method (EIM) [25], the quasi-TM modes of a rib waveguide can be approximately described by the dispersion relation of TE-polarized modes of a dielectric slab waveguide with the thickness l, in which the values nr,TM and nwg,TM are used as the refractive indices of the core layer and claddings, respectively. The dispersion relation of a symmetric slab waveguide can be written in the following form [26]:

l=πm+argr(θ)k0nr,TMcosθ,
where the integer m is the mode order and r(θ) is the complex reflection coefficient of a TE-polarized plane wave from the interface between the media with refractive indices nr,TM and nwg,TM. The EIM-based dispersion relation (1) can be made more accurate by using the reflection coefficient of the TM-polarized mode rmode, TM (θ) from the interface between two waveguides with the thicknesses hr = 110 nm and hc = 80 nm instead of the plane-wave reflection coefficient r(θ). The dispersion curves l = l(θ) corresponding to this refinement are shown in Fig. 2 with dotted curves and are in an excellent agreement with the resonance locations. This confirms the hypothesis that the resonant features in the spectra of Fig. 2 are indeed due to the excitation of the cross-polarized modes of the ridge.

Since in the resonant regime (at θcr,1 < θ < θcr,2) there is neither out-of-plane scattering nor polarization conversion in the reflected and transmitted radiation, the transmission coefficient strictly vanishes at the resonances [27,28]. Moreover, it is evident from Fig. 2 that at different ridge lengths the resonant peaks (dips) have different angular widths (different quality factor), which vary from units to thousandths of a degree and less. This change in the quality factor is caused by the interaction between the resonances of two types: TE Fabry–Pérot resonance and TM guided-mode resonance. When two types of resonances (two modes) interact, the so-called matrix Fabry–Pérot resonances occur, which can have a very high Q-factor [29]. In fact, the Q-factor in the considered structure can even reach infinity, i.e. the structure supports the so-called bound states in the continuum (BICs) [20,21]. A detailed theoretical investigation of the Q-factor of the resonances and of the BIC formation mechanism in the studied structure will be published elsewhere [21].

3. Transformation of the profile of a beam propagating in the waveguide upon diffraction on a ridge

Let us now consider the transformation of the spatial profile of a TE-polarized beam propagating in the waveguide, which occurs upon reflection from the ridge and propagation through the ridge. In the coordinate system associated with the incident beam (uinc, vinc) (Fig. 1), the obliquely incident beam with the angle of incidence θ0 can be represented as a superposition of slab waveguide modes with different spatial frequencies ku,inc = k0nwg,TE sin θinc, where θinc is the angle between the propagation direction of the mode and the vinc axis. We assume that the spatial spectrum of the incident beam G (ku,inc), |ku,inc| ≤ g is narrow enough so that gk0nwg,TE. In this case, the field of the beam can be represented as [19]

Ψinc,wg(uinc,vinc)=exp(ik0nwg,TEvinc)Pinc,wg(uinc)=exp(ik0nwg,TEvinc)G(ku,inc)exp(iku,incuinc)dku,inc,
where Pinc,wg (uinc) is the profile of the beam in a certain plane inside the waveguide, e.g. at y = hc/2. The transformation of the profile of the beam upon diffraction on a ridge can be described in terms of the linear system theory. The transfer functions (TFs) describing the transformation of the beam profile in reflection and transmission have the form [6,19]
HR(ku,inc)=RTE(kx(ku,inc)),HT(ku,inc)=TTE(kx(ku,inc)),
where RTE (kx) and TTE (kx) are the complex reflection and transmission coefficients of the ridge, respectively, and kx (ku,inc) are the spatial frequencies of the TE-polarized modes constituting the incident beam written in the coordinate system associated with the structure:
kx(ku,inc)=k0nwg,TEsin(θinc+θ0)ku,inccosθ0+kx,0,
where kx,0 = k0nwg,TE sin θ0. According to Eqs. (3) and (4), the profiles of the reflected and transmitted beams in the respective coordinate systems (Fig. 1) have the form
Prefl(urefl)=G(ku,inc)R(kx(ku,inc))exp(iku,incurefl)dku,inc,Ptr(utr)=G(ku,inc)T(kx(ku,inc))exp(iku,incutr)dku,inc.

4. Spatial integration of a beam propagating in the waveguide

The presence of high-Q resonances in the spectra in Fig. 2 makes it possible to use the considered structure as an optical spatial integrator operating in reflection [7,10]. Indeed, in the vicinity of the resonance, the reflection coefficient can be approximated by the Fano lineshape [7]

RTE(kr)r+bkxkx,p,
where r is the non-resonant reflection coefficient, b is the coefficient describing the coupling of the incident light to an eigenmode supported by the structure (by the rib waveguide), and kx,p = k′x,p + ik″x,p is the complex propagation constant of the eigenmode corresponding to the pole of the function RTE (kx). Under the assumptions that kx,0 = k′x,p, RTE (kx,0) = 1, and that the non-resonant reflection coefficient r in Eq. (6) can be neglected, we can write the TF HR (ku,inc) of (3) in the following form:
HR(ku,inc)=γku,inciγ,
where γ = k″x,p/cos θ0. The TF in Eq. (7) is an approximation of the TF of a perfect integrator with respect to the spatial variable Hint (ku,inc) = 1/(iku,inc). As the Q-factor of the resonance increases (i.e. as the imaginary part of the mode propagation constant k″x,p decreases), the accuracy of the approximation also increases.

As an example, solid curves in Figs. 3(a), 3(b) and 3(d), 3(e) show the amplitudes and phases of the TFs HR (ku,inc) calculated at two resonant points, at which the reflectance reaches unity: (lunity,1, θunity,1) = (0.355 μm, 53.28°) (point 1) and (lunity,2, θunity,2) = (0.36 μm, 53.37°) (point 2). These points are marked with white asterisks in the inset of Fig. 2(a). Let us note that since the structure works at relatively high angles of incidence, the incident and reflected fields become spatially separated at a reasonable distance from the ridge. Therefore, in the experimental studies, the reflected and transmitted signals can be registered after outcoupling by appropriately designed diffraction gratings located on the waveguide surface. In practical applications, more sophisticated fully integrated optical setups similar to the ones designed for digital planar holograms [17] can be used.

The resonant approximations calculated using Eq. (7) are shown with dashed curves in Figs. 3(a), 3(b) and 3(d), 3(e) and are in a good agreement with the rigorously calculated TFs. The parameters of the approximations γ = γ1 = 0.0011 μm−1 (point 1) and γ = γ2 = 0.0018 μm−1 (point 2) were found by fitting Eq. (7) to the computed angular spectra.

Let us now consider the integration of a beam propagating in the waveguide and having the profile Pinc(uinc)=(2uinc/σ2)exp(uinc2/σ2), which corresponds to the derivative of a Gaussian function g(uinc)=exp(uinc2/σ2). The normalized spatial spectrum of the beam iG(ku,inc)=σ[ku,inc/(2π)]exp(ku,inc2σ2/4) at σ = 15 μm is shown with a dashed curve in Figs. 3(a) and 3(d). The spatial bandwidth of the input beam (the width of its spectrum amplitude) at the 0.1 level with respect to the maximum value amounts to W = 0.052k0. In order to assess the integration quality, we calculated the profile of the reflected beam Prefl (urefl) using Eq. (5). Figures 3(c) and 3(f) show the incident beam (dotted curves), analytically calculated integral of the incident beam profile (dashed curves), and absolute values of the profiles of the reflected beams (solid curves) corresponding to the TFs of Figs. 3(a) and 3(d), respectively. The reflected beams are normalized by the maximum amplitude of the incident beam. The analytically calculated integral [the function g (urefl)] is shown scaled so that its value coincides with the amplitude of the reflected signal at urefl = 0.

The resonance at the point (lunity,1, θunity,1) [Fig. 3(a)] has a higher Q-factor than the resonance at the point (lunity,2, θunity,2) [Fig. 3(d)]. Therefore, one can expect the quality of integration at the point (lunity,1, θunity,1) [Fig. 3(c)] to be higher than that at the point (lunity,2, θunity,2) [Fig. 3(f)]. As a measure of the integration quality, we use the normalized root-mean-square deviation (NRMSD) of the absolute value of the reflected beam from the analytically calculated integral corresponding to the Gaussian function. For the considered examples, the NRMSD values amount to 1.2% [Fig. 3(c)] and 4.2% [Fig. 3(f)]. Let us note that for the first integrator corresponding to the point (lunity,1, θunity,1) the NRMSD does not exceed 5% for the incident signals with spatial bandwidth in the range 0.013k0W ≤ 0.1k0. The maximum bandwidth is comparable to the one reported in [10].

 figure: Fig. 3

Fig. 3 Amplitudes (absolute values) and phases (arguments) of the reflection transfer functions HR (ku,inc) calculated at the points (lunity,1, θunity,1) = (0.355 μm, 53.28°) (a), (b) and (lunity,2, θunity,2) = (0.36 μm, 53.37°) (d), (e). Dotted curves in (a) and (d) show the spectrum of the incident beam, dashed curves in (a), (b) and (d), (e) show the amplitudes and phases of the TF approximations calculated using Eq. (7). (c), (f) Absolute values of the calculated profiles of the reflected beams corresponding to the TFs shown in (a), (b) and (d), (e), respectively. Dotted curves in (c) and (f) show the incident beam, dashed curves show the absolute values of the analytically calculated integral.

Download Full Size | PDF

It is evident from Fig. 3 that the higher the integration quality (the higher the Q-factor), the lower the amplitude of the reflected signal. Maximum amplitudes of the reflected signal normalized by the maximum amplitude of the incident beam in the considered cases equal 0.02 [Fig. 3(c)] and 0.038 [Fig. 3(f)]. Thus, it is possible to achieve the required tradeoff between the quality of integration and the energy (amplitude) of the reflected beam by choosing the Q-factor of the resonance. It is interesting to note that in contrast to the previously proposed optical integrators based on phase-shifted Bragg gratings [7] and W-type structures [10], an increase in the quality factor (which is necessary for an increase in the integration quality) is not necessarily accompanied by an increase in the footprint of the structure in the propagation direction.

The authors believe that higher-order integration can be implemented using several ridges arranged in a configuration similar to the one used for achieving higher-order differentiation in reflection with acoustic metamaterials (see Fig. 4 in [30]).

 figure: Fig. 4

Fig. 4 Amplitudes (absolute values) and phases (arguments) of the transmission transfer functions HT (ku,inc) calculated at the points (lzero,1, θzero,1) = (0.2 μm, 48.62°) (a) and (lzero,2, θzero,2) = (0.24 μm, 50.14°) (c). Dotted curves in (a) and (c) show the normalized spectrum of the incident Gaussian beam. (b), (d) Absolute values of the calculated profiles of the transmitted beams corresponding to the TFs shown in (a) and (c), respectively. Dotted curves show the incident Gaussian beam, dashed curves show the absolute value of the analytically calculated derivative of the Gaussian function.

Download Full Size | PDF

5. Spatial differentiation of a beam propagating in the waveguide

Let us now discuss the application of the considered structure for the computation of the spatial derivative of the incident beam profile. As shown in Section 2, the transmission coefficient of the ridge strictly vanishes at the resonances, which enables using this structure as an optical differentiator operating in transmission [6, 19]. Indeed, let us assume that the transmission coefficient TTE (kx) vanishes at certain angle of incidence θ0 and ridge length l. Expanding the TF HT (ku,inc) of Eq. (3) into Taylor series at ku,inc = 0 up to the linear term, we obtain:

HT(ku,inc)αTku,inc.
Thus, in the first approximation this TF is proportional to the TF of an exact differentiator Hdiff (ku,inc) = iku,inc.

As an example, Fig. 4(a) shows the amplitude (absolute value) and phase (argument) of the TF HT (ku,inc) calculated at the point (lzero,1, θzero,1) = 0.2 μm, 48.62°) marked with a black asterisk in Fig. 2(b), where lzero,1 and θzero,1 are the ridge length and the angle of incidence, at which the transmission coefficient vanishes. Figure 4(a) demonstrates that in the vicinity of the point ku,inc = 0 (i.e. at θ0 = θzero,1) the TF of the ridge is in good agreement with the TF of an exact differentiator. Let us note that the linear phase of the TF leads only to the shift of the transmitted beam similarly to the Goos–Hänchen effect and does not affect the differentiation quality.

Consider the differentiation of a beam propagating in the waveguide and having a Gaussian profile Pinc(uinc)=exp(uinc2/σ2). The normalized spectrum of the beam (2π/σ)G(ku,inc)=exp(ku,inc2σ2/4) at σ = 25 μm is shown with a dashed curve in Fig. 4(a). The widths of the beam Pinc (uinc) and of its spectrum G (ku,inc) (the spatial bandwidth W) at the level 0.1 with respect to the maximum value amount to 75.9 μm and 0.024k0, respectively.

In order to assess the differentiation quality, we calculated the profile of the transmitted beam Ptr (utr) using Eq. (5). Figure 4(b) shows the incident beam (dotted curve), the absolute value of the calculated profile of the transmitted beam (solid curve), and the absolute value of the analytically calculated derivative of the Gaussian function (dashed curve). The analytically calculated derivative is normalized so that its maximum value coincides with that of the transmitted signal. Similarly to the previous section, we use the NRMSD of the transmitted beam from the analytically calculated derivative as the measure of the quality of differentiation. It is evident from Fig. 4(b) that the differentiation quality is high, the NRMSD value in this case is only 1.2%. Note that in the NRMSD calculation, the shift of the transmitted beam was not taken into account (i.e. the minima of the transmitted beam and of the exact derivative were shifted to the same point). The maximum of the amplitude of the transmitted beam for this example equals 0.123. Let us note that for this structure, the differentiation NRMSD does not exceed 5% at spatial bandwidth values of the incident beam W ≤ 0.06k0.

Similarly to the integration case considered above, by choosing the quality factor of the resonance, one can achieve the required tradeoff between the differentiation quality and the energy (amplitude) of the transmitted beam. For example, let us consider another resonant point (lzero,2, θzero,2) = (0.24 μm, 50.14°) also marked with a black asterisk in Fig. 2(b), at which the transmission coefficient also vanishes. At this point, the resonant dip is narrower (i.e. the resonance has a higher quality factor). Therefore, the corresponding TF shown in Fig. 4(c) has a smaller linearity interval, but a larger amplitude. Thus, in this case, we should expect a decrease in the differentiation quality accompanied by an increase in the amplitude of the transmitted beam. The absolute value of the calculated profile of the transmitted beam corresponding to the point (lzero,2, θzero,2) is shown in Fig. 4(d). In this case, the NRMSD value is indeed higher than in the previous example and amounts to 3.5%. At the same time, maximum amplitude of the transmitted signal is also higher and equals 0.21.

Let us note that in the considered ridge-based differentiators (Fig. 4), the ridge length is of about 3 times smaller than the free-space wavelength of the incident beam λ = 630 nm, while the ridge height hrhc is deep-subwavelength. It follows from Fig. 2 that we can control the ridge length of the designed differentiator or integrator by choosing different (l, θ) points located on the dispersion curves of the resonances. In particular, in the vicinity of the cutoff angle θcr,1, the ridge length can be 10–100 nm depending on the required Q-factor of the resonance.

It is worth mentioning that higher-order differentiation can be performed simply by using several consecutive ridges on the waveguide surface (the number of the ridges equals to the order of the computed derivative), similar to how it was done in paper [31] for the case of grating-based optical differentiators also operating in transmission.

6. Conclusion

In the present work, we demonstrated that in the case of oblique incidence of a TE-polarized mode of a slab waveguide on a subwavelength dielectric ridge located on the waveguide surface, resonant changes in the reflectance and the transmittance occur. Using the effective index method, we explained these resonant effects by the excitation of a cross-polarized mode of the ridge.

The discovered resonances enable optical implementation of the operations of spatial integration and differentiation of the profile of an optical beam propagating in the waveguide. The computation of the integral is performed in reflection, whereas the computation of the derivative is performed in transmission. Rigorous numerical simulation results confirm the possibility of high-quality spatial integration and differentiation. The proposed ridge structure has a simpler geometry comparing to the previously proposed integrators and differentiators based on phase-shifted Bragg gratings [7,32] and the so-called W-type structures [10,19].

The obtained results may find application in the design of on-chip systems for all-optical analog computing. The authors also believe that the presented planar structure can be used as an integrated optical spatial (angular) or spectral filter. Moreover, the existence of the bound states in the continuum in the structure also makes it promising for lasing and sensing, among other applications.

Funding

Russian Foundation for Basic Research (16-29-11683); Ministry of Science and Higher Education of the Russian Federation (State assignment to the FSRC “Crystallography and Photonics” RAS); Russian Science Foundation (14-19-00796).

Acknowledgments

This work was funded by Russian Foundation for Basic Research (numerical investigation of the optical properties of the ridge, Section 2), Ministry of Science and Higher Education of the Russian Federation (theoretical analysis of the resonances of the ridge using the effective index method, Section 2), and by Russian Science Foundation (application of the structure to spatial integration and differentiation of optical beams, Sections 3–5).

References

1. J. Goodman, Introduction to Fourier Optics (McGraw-Hill, 1996).

2. A. Silva, F. Monticone, G. Castaldi, V. Galdi, A. Alù, and N. Engheta, “Performing mathematical operations with metamaterials,” Science 343(6167), 160–163 (2014). [CrossRef]   [PubMed]  

3. A. Pors, M. G. Nielsen, and S. I. Bozhevolnyi, “Analog computing using reflective plasmonic metasurfaces,” Nano Lett. 15(1), 791–797 (2015). [CrossRef]  

4. A. Chizari, S. Abdollahramezani, M. V. Jamali, and J. A. Salehi, “Analog optical computing based on a dielectric meta-reflect array,” Opt. Lett. 41(15), 3451–3454 (2016). [CrossRef]   [PubMed]  

5. H. Babashah, Z. Kavehvash, S. Koohi, and A. Khavasi, “Integration in analog optical computing using metasurfaces revisited: toward ideal optical integration,” J. Opt. Soc. Am. B 34(6), 1270–1279 (2017). [CrossRef]  

6. 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(5), 1278–1281 (2014). [CrossRef]   [PubMed]  

7. N. V. Golovastikov, D. A. Bykov, L. L. Doskolovich, and E. A. Bezus, “Spatial optical integrator based on phase-shifted Bragg gratings,” Opt. Commun. 338, 457–460 (2015). [CrossRef]  

8. T. Zhu, Y. Zhou, Y. Lou, H. Ye, M. Qiu, Z. Ruan, and S. Fan, “Plasmonic computing of spatial differentiation,” Nat. Commun. 8, 15391 (2017). [CrossRef]   [PubMed]  

9. Z. Ruan, “Spatial mode control of surface plasmon polariton excitation with gain medium: from spatial differentiator to integrator,” Opt. Lett. 40(4), 601–604 (2015). [CrossRef]   [PubMed]  

10. F. Zangeneh-Nejad and A. Khavasi, “Spatial integration by a dielectric slab and its planar graphene-based counterpart,” Opt. Lett. 42(10), 1954–1957 (2017). [CrossRef]   [PubMed]  

11. F. Zangeneh-Nejad, A. Khavasi, and B. Rejaei, “Analog optical computing by half-wavelength slabs,” Opt. Commun. 407, 338–343 (2018). [CrossRef]  

12. A. Youssefi, F. Zangeneh-Nejad, S. Abdollahramezani, and A. Khavasi, “Analog computing by Brewster effect,” Opt. Lett. 41(15), 3467–3470 (2016). [CrossRef]   [PubMed]  

13. N. V. Golovastikov, D. A. Bykov, and L. L. Doskolovich, “Spatiotemporal pulse shaping using resonant diffraction gratings,” Opt. Lett. 40(15), 3492–3495 (2015). [CrossRef]   [PubMed]  

14. N. V. Golovastikov, D. A. Bykov, and L. L. Doskolovich, “Resonant diffraction gratings for spatial differentiation of optical beams,” Quantum Electron. 44(10), 984–988 (2014). [CrossRef]  

15. D. A. Bykov, L. L. Doskolovich, A. A. Morozov, V. V. Podlipnov, E. A. Bezus, P. Verma, and V. A. Soifer, “First-order optical spatial differentiator based on a guided-mode resonant grating,” Opt. Express 26(8), 10997–11006 (2018). [CrossRef]   [PubMed]  

16. Z. Dong, J. Si, X. Yu, and X. Deng, “Optical spatial differentiator based on subwavelength high-contrast gratings,” Appl. Phys. Lett. 112(18), 181102 (2018). [CrossRef]  

17. G. Calafiore, A. Koshelev, S. Dhuey, A. Goltsov, P. Sasorov, S. Babin, V. Yankov, S. Cabrini, and C. Peroz, “Holographic planar lightwave circuit for on-chip spectroscopy,” Light. Sci. Appl. 3, e203 (2014). [CrossRef]  

18. T. Mossberg, “Planar holographic optical processing devices,” Opt. Lett. 26(7), 414–416 (2001). [CrossRef]  

19. L. L. Doskolovich, E. A. Bezus, N. V. Golovastikov, D. A. Bykov, and V. A. Soifer, “Planar two-groove optical differentiator in a slab waveguide,” Opt. Express 25(19), 22328–22340 (2017). [CrossRef]   [PubMed]  

20. C. L. Zou, J. M. Cui, F. W. Sun, X. Xiong, X. B. Zou, Z. F. Han, and G. C. Guo, “Guiding light through optical bound states in the continuum for ultrahigh-Q microresonators,” Laser Photon. Rev. 9(1), 114–119 (2015). [CrossRef]  

21. E. A. Bezus, D. A. Bykov, and L. L. Doskolovich, “Bound states in the continuum and high-Q resonances supported by a dielectric ridge on a slab waveguide,” https://arxiv.org/abs/1807.01888.

22. E. Silberstein, P. Lalanne, J.-P. Hugonin, and Q. Cao, “Use of grating theories in integrated optics,” J. Opt. Soc. Am. A 18(11), 2865–2875 (2001). [CrossRef]  

23. J. P. Hugonin and P. Lalanne, “Perfectly matched layers as nonlinear coordinate transforms: a generalized formalization,” J. Opt. Soc. Am. A 22(9), 1844–1849 (2005). [CrossRef]  

24. M. Hammer, A. Hildebrandt, and J. Förstner, “How planar optical waves can be made to climb dielectric steps,” Opt. Lett. 40(16), 3711–3714 (2015). [CrossRef]   [PubMed]  

25. R. D. Kekatpure, A. C. Hryciw, E. S. Barnard, and M. L. Brongersma, “Solving dielectric and plasmonic waveguide dispersion relations on a pocket calculator,” Opt. Express 17(26), 24112–24129 (2009). [CrossRef]  

26. G. Lifante, Integrated Photonics: Fundamentals (Wiley, 2003). [CrossRef]  

27. D. A. Bykov, L. L. Doskolovich, N. V. Golovastikov, and V. A. Soifer, “Time-domain differentiation of optical pulses in reflection and in transmission using the same resonant grating,” J. Opt. 15(10), 105703 (2013). [CrossRef]  

28. N. A. Gippius, S. G. Tikhodeev, and T. Ishihara, “Optical properties of photonic crystal slabs with an asymmetrical unit cell,” Phys. Rev. B 72(4), 045138 (2005). [CrossRef]  

29. V. Karagodsky, C. Chase, and C. J. Chang-Hasnain, “Matrix Fabry–Perot resonance mechanism in high-contrast gratings,” Opt. Lett. 36(9), 1704–1706 (2011). [CrossRef]   [PubMed]  

30. F. Zangeneh-Nejad and R. Fleury, “Performing mathematical operations using high-index acoustic metamaterials,” New J. Phys. 20, 073001 (2018). [CrossRef]  

31. D. A. Bykov, L. L. Doskolovich, and V. A. Soifer, “On the ability of resonant diffraction gratings to differentiate a pulsed optical signal,” J. Exp. Theor. Phys. 114(5), 724–730 (2012). [CrossRef]  

32. L. L. Doskolovich, E. A. Bezus, D. A. Bykov, and V. A. Soifer, “Spatial differentiation of Bloch surface wave beams using an on-chip phase-shifted Bragg grating,” J. Opt. 18(11), 115006 (2016). [CrossRef]  

Cited By

Optica participates in Crossref's Cited-By Linking service. Citing articles from Optica Publishing Group journals and other participating publishers are listed here.

Alert me when this article is cited.


Figures (4)

Fig. 1
Fig. 1 Geometry of the problem of diffraction of a waveguide mode on a dielectric step.
Fig. 2
Fig. 2 Reflectance (a) and transmittance (b) of the ridge vs. the ridge length l (horizontal axis) and the angle of incidence θ (vertical axis). Dashed lines show the boundaries of the resonance region and correspond to the cutoff angles of the TM-modes outside the ridge (at hc = 80 nm, upper line) and inside the ridge (hr = 110 nm, lower line). Dotted curves show the dispersion of the cross-polarized modes of the ridge. The points marked with white and black asterisks correspond to the spatial integration and differentiation examples considered below, respectively.
Fig. 3
Fig. 3 Amplitudes (absolute values) and phases (arguments) of the reflection transfer functions HR (ku,inc) calculated at the points (lunity,1, θunity,1) = (0.355 μm, 53.28°) (a), (b) and (lunity,2, θunity,2) = (0.36 μm, 53.37°) (d), (e). Dotted curves in (a) and (d) show the spectrum of the incident beam, dashed curves in (a), (b) and (d), (e) show the amplitudes and phases of the TF approximations calculated using Eq. (7). (c), (f) Absolute values of the calculated profiles of the reflected beams corresponding to the TFs shown in (a), (b) and (d), (e), respectively. Dotted curves in (c) and (f) show the incident beam, dashed curves show the absolute values of the analytically calculated integral.
Fig. 4
Fig. 4 Amplitudes (absolute values) and phases (arguments) of the transmission transfer functions HT (ku,inc) calculated at the points (lzero,1, θzero,1) = (0.2 μm, 48.62°) (a) and (lzero,2, θzero,2) = (0.24 μm, 50.14°) (c). Dotted curves in (a) and (c) show the normalized spectrum of the incident Gaussian beam. (b), (d) Absolute values of the calculated profiles of the transmitted beams corresponding to the TFs shown in (a) and (c), respectively. Dotted curves show the incident Gaussian beam, dashed curves show the absolute value of the analytically calculated derivative of the Gaussian function.

Equations (8)

Equations on this page are rendered with MathJax. Learn more.

l = π m + arg r ( θ ) k 0 n r , TM cos θ ,
Ψ inc , wg ( u inc , v inc ) = exp ( i k 0 n wg , TE v inc ) P inc , wg ( u inc ) = exp ( i k 0 n wg , TE v inc ) G ( k u , inc ) exp ( i k u , inc u inc ) d k u , inc ,
H R ( k u , inc ) = R TE ( k x ( k u , inc ) ) , H T ( k u , inc ) = T TE ( k x ( k u , inc ) ) ,
k x ( k u , inc ) = k 0 n wg , TE sin ( θ inc + θ 0 ) k u , inc cos θ 0 + k x , 0 ,
P refl ( u refl ) = G ( k u , inc ) R ( k x ( k u , inc ) ) exp ( i k u , inc u refl ) d k u , inc , P tr ( u tr ) = G ( k u , inc ) T ( k x ( k u , inc ) ) exp ( i k u , inc u tr ) d k u , inc .
R TE ( k r ) r + b k x k x , p ,
H R ( k u , inc ) = γ k u , inc i γ ,
H T ( k u , inc ) α T k u , inc .
Select as filters


Select Topics Cancel
© Copyright 2024 | Optica Publishing Group. All rights reserved, including rights for text and data mining and training of artificial technologies or similar technologies.