betaFeedback

WorksPhysics and Astronomy

Anisotropic Virtual Gain and Large Tuning of Particles' Scattering by Complex-Frequency Excitations

Abstract

Active tuning of the scattering of particles and metasurfaces is a highly sought-after property for a host of electromagnetic and photonic applications, but it normally requires challenging-to-control tunable (reconfigurable) or active (gain) media. Here, we introduce the concepts of anisotropic virtual gain and oblique Kerker effect, where a completely lossy anisotropic medium behaves exactly as its anisotropic gain counterpart upon excitation by a synthetic complex-frequency wave. The strategy allows one to largely tune the magnitude and angle of a particle's scattering simply by changing the shape (envelope) of the incoming radiation, rather than by an involved medium-tuning mechanism. The so-attained anisotropic virtual gain enables directional super-scattering at an oblique direction with fine-management of the scattering angle. Our study, opening a unique light-management method, is based on analytical techniques that allow multipolar decomposition of the scattered field, and is found, throughout, to be in excellent agreement with full-wave simulations.

Directional scattering of light is a fundamental concept in modern technology with various applications ranging from improving the efficiency of optical communication and information processing [1, 2], enhancing the performance of solar energy harvesting [3], to refining imaging techniques and treatment methods in biomedicine [4, 5], the directional control of scattering [6], and all-optical switching [7]. Furthermore, it is essential for the correct operation of Lidar systems [8], precise particle manipulation in optical tweezers [9], and the creation of nanoscale devices, particularly metasurfaces [10]. Among the established strategies for achieving directionality, the Kerker effect is the most prominently utilized [11]. However, the effect still remains confined either to forward/backward directions, or when both blocked, to the transverse directions [12].

On the other hand, gain (amplification) plays a vital role in the fields of electromagnetics and optics, allowing not only for the compensation of losses or the attainment of lasing, but also for providing additional degrees of freedom for the active tuning of electromagnetic and optical devices. However, the attainment of gain requires gain media (or other parametric and coherent mechanisms) [13], which are either challenging to integrate with electromagnetic devices or do not exist at all in certain spectral regimes. Moreover, the use of gain media is always accompanied by unwanted side-effects, such as (amplified) spontaneous emission and noise, and is in most cases isotropic [14] (the same in all directions), thereby usually preventing the exploitation of anisotropy for the attainment of active tuning.

It is the objective of the present work to introduce a new - anisotropic - virtual-gain mechanism that, simply by adjusting the incident radiation (rather than each time the medium itself), allows for effectively converting a completely passive (lossy) anisotropic medium to a corresponding amplifying anisotropic medium, giving rise to unique, hitherto unattainable, degrees of freedom in the active steering of scattering by particles. Our methodology makes use of, so called, complex-frequency excitations, which in the recent past have been investigated in the context of slow and stopped light [15, 16, 17], the attainment of effective PT-symmetry [18], subwavelength focusing [19, 20], virtual critical coupling [21], virtual optical pulling force [22], and transient non-Hermitian skin effects [23]. To those ends, we also introduce for the first time to our knowledge, the concept of the oblique Kerker effect, where destructive interference of light can be achieved at an arbitrary angle, resulting in a controllable steering of scattered radiation at an oblique direction with respect to that of the incident field.

The physical mechanism and the underlying designing rules are demonstrated in a system that consists of a dielectric cylinder coated by an arbitrary, hypothetical, gyroelectric anisotropic magneto-optical (AMO) medium (where for simplicity we used InSb [24, 25] and we may occasionally reverse the sign of the imaginary part of its permittivities to ‘turn it’ amplifying). The physics is elucidated by our analytical method that allows multipolar decomposition of the scattered field, while our results are further validated by full-wave finite-element simulations.

The configuration is depicted in Figure 1 and consists of a 15 μm\mu\mathrm{m} bb-radius core from a high-permittivity dielectric material ϵc=ϵcrϵ0\epsilon_{\mathrm{c}} = \epsilon_{\mathrm{cr}}\epsilon_0, where ϵcr=25\epsilon_{\mathrm{cr}} = 25, with $\epsilon_0 the free space permittivity. The core is coated by a 20 μm20\ \mu\mathrm{m} aa-radius AMO shell. The magnetic permeability of both media is that of the free space, i.e., μ0\mu_0. The whole setup is located in free space and is exposed in a tunable external magnetic flux density B0=B0ez\mathbf{B}_0=B_0\mathbf{e}_z. In general, AMO’s permittivity tensor ϵˉˉ\bar{\bar{\epsilon}} is of gyroelectric type (i.e., features three relative permittivity elements ϵ1r\epsilon_{1r}, ϵ2r\epsilon_{2r} and ϵ3r\epsilon_{3r}), and B0B_0-dependent. When the bias B0≠0B_0\ne0 (on-state), the AMO becomes gyroelectric which, in Cartesian coordinates,

Figure 1. (a) Schematic representation of a hypothetical anisotropic magneto-optical (AMO) 2-D structure under study. (b) Scattering cross-section spectrum assuming gain in the AMO material (v=−0.001ωpv=-0.001\omega_p) and zero magnetic flux density, at the setup of (a). A transparency window appears close to the frequency pointed out with the dotted orange arrowhead. Blue curve (stars): total normalized QscQ_{\mathrm{sc}} under analytical solution (simulation using COMSOL); red curve: MD term; green curve: ED term; black curve: EQ term. (c) Same as (b) but with B0=0.1B_0=0.1 T. The magenta circle indicates the ED-EQ intersection (higher-order Kerker point) and the magenta dotted arrowhead indicates the respective frequency f=2.016f=2.016 THz. (d) Radiation pattern of normalized σ\sigma under f=2.016f=2.016 THz excitation. Magenta: on-state/bended when B0=0.1B_0=0.1 T/curve: analytical solution/stars: COMSOL; orange: off-state/transparency when B0=0B_0=0 T.

ϵˉˉ(B0)=ϵ0[ϵ1r(B0)(exexT+eyeyT)+iϵ2r(B0)(exeyT−eyexT)+ϵ3rezezT],\bar{\bar{\epsilon}}(B_0)=\epsilon_0\left[\epsilon_{1r}(B_0)\left(\mathbf{e}_x\mathbf{e}_x^{T}+\mathbf{e}_y\mathbf{e}_y^{T}\right)+i\epsilon_{2r}(B_0)\left(\mathbf{e}_x\mathbf{e}_y^{T}-\mathbf{e}_y\mathbf{e}_x^{T}\right)+\epsilon_{3r}\mathbf{e}_z\mathbf{e}_z^{T}\right],

where TT denotes transposition. It should be emphasized that only ϵ1r(B0)\epsilon_{1r}(B_0), ϵ2r(B0)\epsilon_{2r}(B_0) depend on B0B_0, while ϵ3r\epsilon_{3r} does not. When the bias B0=0B_0=0 (off-state), AMO turns isotropic with ϵ1r(0)=ϵ3r\epsilon_{1r}(0)=\epsilon_{3r} and ϵ2r(0)=0\epsilon_{2r}(0)=0. In particular, the relative elements coincide with that of InSb [24, 25] and are given by

ϵ1r(B0)=ϵ∞{1−(ω−iv)ωp2ω[(ω−iv)2−ωc2]},\epsilon_{1r}(B_0)=\epsilon_\infty\left\{1-\frac{(\omega-iv)\omega_p^2}{\omega\left[(\omega-iv)^2-\omega_c^2\right]}\right\},
ϵ2r(B0)=ϵ∞{ωcωp2ω[(ω−iv)2−ωc2]}\epsilon_{2r}(B_0)=\epsilon_\infty\left\{\frac{\omega_c\omega_p^2}{\omega\left[(\omega-iv)^2-\omega_c^2\right]}\right\}

and

ϵ3r=ϵ∞{1−ωp2ω(ω−iv)}.\epsilon_{3r}=\epsilon_\infty\left\{1-\frac{\omega_p^2}{\omega(\omega-iv)}\right\}.

Here, ϵ∞=15.6\epsilon_\infty=15.6 accounts for interband transitions, ωp=[Nee2/(ϵ0ϵ∞m∗)]1/2=4π×1012 rad s−1\omega_p=\left[N_e e^2/(\epsilon_0\epsilon_\infty m^*)\right]^{1/2}=4\pi\times10^{12}\ \mathrm{rad}\ \mathrm{s}^{-1} is the plasma angular frequency (with NeN_e the electron density, ee the elementary charge and m∗=0.0142mem^*=0.0142m_e electron’s effective mass, where mem_e is electron’s rest mass), ωc=eB0/m∗\omega_c=eB_0/m^* is cyclotron angular frequency, and v=αωpv=\alpha\omega_p the damping angular frequency (with α\alpha a dimensionless parameter). In what follows, the employed analytical solution adopts the exp⁡(iωt)\exp(i\omega t) time dependence. This means that, if α>0\alpha>0, the AMO is lossy (AMO becomes InSb), while, if α<0\alpha<0, AMO is active (gain).

To examine the electromagnetic (EM) response of the system, we employ the exact analytical solution of EM plane-wave scattering by gyrotropic cylinders [26], and extend it to dielectric-gyroelectric core-shell setups. We also employ COMSOL, whenever possible, for the sake of comparison. As it is explained in [27], the gyroelectric elements iϵ0ϵ2r(B0)(exeyT−eyexT)i\epsilon_0\epsilon_{2r}(B_0)(\mathbf{e}_x\mathbf{e}_y^{T}-\mathbf{e}_y\mathbf{e}_x^{T}) affect only the transverse electric (TE) illumination (the incident electric field Einc\mathbf{E}^{\mathrm{inc}} is normal to cylinder’s zz-axis), while the transverse magnetic (TM) illumination (Einc\mathbf{E}^{\mathrm{inc}} is parallel to cylinder’s zz-axis) is unaffected by such an anisotropy. Therefore, we only focus on TE plane-wave scattering, as is illustrated in Figure 1(a). Since the plane-wave impinges from the negative towards the positive values of xx, Einc\mathbf{E}^{\mathrm{inc}} is yy-polarized and the incident wavevector kinc\mathbf{k}^{\mathrm{inc}} is normal to B0\mathbf{B}_0. To explore the spectrum of the system we employ the total scattering cross-section QscQ_{\mathrm{sc}}, in meter (m) units for two-dimensional (2-D) structures, given by

Qsc=∫02π∣fφ(φ)∣2 dφ=4k0∑m=−∞∞∣am∣2,(1)Q_{\mathrm{sc}}=\int_{0}^{2\pi}\lvert f_\varphi(\varphi)\rvert^2\,d\varphi=\frac{4}{k_0}\sum_{m=-\infty}^{\infty}\lvert a_m\rvert^2, \tag*{(1)}

where fφ(φ)=1/k01/2f^φ(φ)f_\varphi(\varphi)=1/k_0^{1/2}\hat{f}_\varphi(\varphi), f^φ(φ)=(2i/π)1/2∑m=−∞∞im+1e−imφam\hat{f}_\varphi(\varphi)=(2i/\pi)^{1/2}\sum_{m=-\infty}^{\infty}i^{m+1}e^{-im\varphi}a_m, k0k_0 is the free space wavenumber, and ama_m are the expansion coefficients of the scattered electric field, with m=0,±1,±2,…m=0,\pm1,\pm2,\ldots. It should be mentioned that the expression (1) is different from the one used in three-dimensional (3-D) problems with units m2\mathrm{m}^2, as (1) is a per-unit-length quantity. Generally, the excitation angular frequency ω=2πν\omega=2\pi\nu is introduced in k0=ω(ϵ0μ0)1/2k_0=\omega(\epsilon_0\mu_0)^{1/2}. For complex-frequency excitation, ν=f+iγ\nu=f+i\gamma, where f=Re⁡νf=\operatorname{Re}\nu and γ=Im⁡ν\gamma=\operatorname{Im}\nu represent the real and imaginary parts of the complex ν\nu; in this case, k0k_0 becomes complex. For real-frequency excitation, ν=f\nu=f, γ=0\gamma=0 and k0k_0 is real. A multipolar decomposition treatment is inherent in this analytical method. In fact, the magnetic dipolar (MD) contribution to the QscQ_{\mathrm{sc}} (dominant term in the subwavelength regime) is obtained by keeping the m=0m=0 term in (1), while the electric dipolar (ED) and the electric quadrupolar (EQ) contributions to the QscQ_{\mathrm{sc}} are obtained by keeping the m=±1m=\pm1 and the m=±2m=\pm2 terms, respectively, in (1) [28]. In contrast to the multipolar decomposition, the total QscQ_{\mathrm{sc}} discussed below is obtained exactly by truncating (1) with an upper bound of 20 terms, i.e., m=−20,…,−1,0,1,…,20m = -20, \ldots, -1, 0, 1, \ldots, 20, a selection that yields machine precision results.

At first, we consider an AMO shell with gain, by setting v=−0.001ωpv = -0.001\omega_p (α=−0.001\alpha= -0.001) in ϵˉ\bar{\epsilon}, and illuminating the configuration by a real-frequency plane-wave (ν≡f\nu\equiv f, γ=0\gamma= 0). In Figure 1(b) we depict the Qsc/aQ_{\mathrm{sc}}/a spectrum at off-state (B0=0B_0 = 0), in a frequency window that includes AMO’s plasma frequency fp=2f_p = 2 THz, where the total scattering is negligibly small. In particular, the scattering is fully suppressed with almost null contribution from all MD/ED/EQ terms, resulting in a transparency window around f=2.016f = 2.016 THz, with the latter pointed out by the dotted orange arrowhead. The validity of the analytical solution (blue curve) is verified by considering the same setup in COMSOL (blue stars). In Figure 1(c) we again plot the Qsc/aQ_{\mathrm{sc}}/a in the same frequency window as in Figure 1(b), by turning on the bias at B0=0.1B_0 = 0.1 T (on-state). As evident, a magnetoplasmonic mode arises due to the interaction between the EM wave and plasma oscillations in the presence of the external B0B_0. From the multipolar decomposition, the MD term is negligible and the contribution to the total Qsc/aQ_{\mathrm{sc}}/a stems, mainly, from the ED/EQ terms, which indicates the plasmonic origin of this mode. The magenta circle in Figure 1(c) indicates the ED-EQ intersection, which corresponds to a generalized higher-order Kerker point, characterized by the same nature (electric) but different order (dipolar and quadrupolar) multipoles, as in the case of isotropic particles [29]. The magenta dotted arrowhead shows the location of this Kerker point at f=2.016f = 2.016 THz. At f=2.016f = 2.016 THz, the total Qsc/a=11.77Q_{\mathrm{sc}}/a = 11.77, a value that overpasses the 2-D super-scattering limit Qsc/a=2c0/(πaf)=4.73Q_{\mathrm{sc}}/a = 2c_0/(\pi af) = 4.73 [30], where c0c_0 is the speed of light in free space (by super-scattering it is meant that an arbitrarily large QscQ_{\mathrm{sc}} can be achieved). To visualize the far-field behavior, we calculate the scattering width (or radiation pattern) σ(φ)\sigma(\varphi),

σ(φ)=lim⁡r→∞2πr∣Esc(r,φ)∣2∣Einc(r,φ)∣2=2πk0∣f^φ(φ)∣2,(2)\sigma(\varphi) = \lim_{r\to\infty} 2\pi r \frac{\left|\mathbf{E}^{\mathrm{sc}}(r,\varphi)\right|^2}{\left|\mathbf{E}^{\mathrm{inc}}(r,\varphi)\right|^2} = \frac{2\pi}{k_0}\left|\hat{f}_\varphi(\varphi)\right|^2, \tag*{(2)}

where Esc(r,φ)\mathbf{E}^{\mathrm{sc}}(r,\varphi) is the scattered electric far-field. Figure 1(d) depicts the σ/a\sigma/a for the off- and on-state, when the system is excited at the same frequency f=2.016f = 2.016 THz. Evidently, the negligible off-state radiation pattern (tiny orange polar curve) gives its position to a high-scattering pattern with a maximum at φ=305∘\varphi= 305^\circ (or −55∘-55^\circ). The analytical solution (magenta curve) is verified by COMSOL (magenta stars). Notably, the magnetically-assisted active-gain system, excited by real frequency, allows a switching from a transparency window to a directional super-scattering pattern [25, 32].

To elucidate the mechanism behind the angle steering shown in Figure 1(d), in Figure 2(a) we plot the magnitudes ∣am∣|a_m|, m=0,±1,±2m = 0, \pm1, \pm2, of the involved expansion coefficients that contribute to the MD/ED/EQ scattering response, in the same frequency window as in Figure 1(c). We notice that the dominant contribution to the peak stems from the m=−1m = -1 and m=−2m = -2 coefficients, while their m=+1m = +1 and m=+2m = +2 counterparts are found at different frequency windows. This is due to the degeneracy-lifting of the mm index in the presence of the magnetic field [25, 33, 34]. Regarding the a0a_0 coefficient, we have ∣a0∣<1|a_0| < 1, meaning that its squared magnitude in (1) yields an insignificant MD response in the Qsc/aQ_{\mathrm{sc}}/a spectrum. At frequency f=2.016f = 2.016 THz, indicated by the magenta dotted arrowhead, there is a same-magnitude intersection of a−1a_{-1} and a−2a_{-2}. This point indicates a possible Kerker condition, although requirements for their respective phases must also be fulfilled. Therefore, in Figure 2(b) we show the coefficients a−1a_{-1} and a−2a_{-2} on the complex plane, and their phase difference Δϕ\Delta\phi. The spectrum in Figure 2(a) corresponds to the gain case of v=−0.001ωpv = -0.001\omega_p (α=−0.001\alpha= -0.001), while in Figure 2(b) we also consider the increased-gain cases v=−0.002ωpv = -0.002\omega_p (α=−0.002\alpha= -0.002) and v=−0.003ωpv = -0.003\omega_p (α=−0.003\alpha= -0.003) upon the same excitation f=2.016f = 2.016 THz. In parallel, Figure 2(c) depicts the radiation pattern σ/a\sigma/a for the respective gain cases of Figure 2(b), to investigate the dependence of the steering angle from the phase difference Δϕ\Delta\phi. As evident, the increase in gain yields an increase in Δϕ\Delta\phi, from 126∘126^\circ (ν=−0.001ωp\nu= -0.001\omega_p) to 161∘161^\circ (ν=−0.003ωp\nu= -0.003\omega_p). This change in Δϕ\Delta\phi impacts the σ/a\sigma/a behavior via (2), precisely through the f^(φ)\hat{f}(\varphi) quantity. Specifically, the increase in Δϕ\Delta\phi further steers to greater negative angles and enhances the radiation pattern, as shown in Figure 2(c). For the depicted instances, if we define the rotation angle (RA) as the angle at which the radiation pattern achieves its maximum value, we get RA =−55∘= -55^\circ, −70∘-70^\circ and −82∘-82^\circ for the respective gain cases of ν=−0.001ωp\nu= -0.001\omega_p, −0.002ωp-0.002\omega_p and −0.003ωp-0.003\omega_p. A systematic study on the Δϕ\Delta\phi and RA vs α\alpha can be found in Section I of the Supplemental Material (SM). Furthermore, in Section II of the SM we show the near field profile for the ν=−0.001ωp\nu= -0.001\omega_p case. The proposed strategy for oblique directional scattering also applies for generic, non-circular 2-D cylinders, as well as for 3-D particles (rods), as demonstrated in Sections III and IV of the SM. We note here that, micron-sized AMO rods can be fabricated in the lab [35], where also measuring THz radiation patterns is within current experimental reach [36].

Figure 2. (a) Magnitudes of ama_m for m=0,±1,±2m = 0, \pm1, \pm2. Red curve: ∣a0∣|a_0|; green solid curve: ∣a−1∣|a_{-1}|; green dashed curve: ∣a1∣|a_1|; black solid curve: ∣a−2∣|a_{-2}|; black dashed curve: ∣a2∣|a_2|. The magenta circle indicates the intersection of ∣a−1∣|a_{-1}| and ∣a−2∣|a_{-2}| and the magenta dotted arrowhead indicates the frequency of the higher-order ED-EQ Kerker point (f=2.016f = 2.016 THz). (b) Representation of the complex numbers a−1a_{-1} and a−2a_{-2} on the complex-plane, together with their phase difference, for different values of gain vv. Magenta: v=−0.001ωpv = -0.001\omega_p; brown: v=−0.002ωpv = -0.002\omega_p; dark green: v=−0.003ωpv = -0.003\omega_p. (c) Radiation pattern of normalized σ\sigma at f=2.016f = 2.016 THz, for the values of vv shown in (b). Magenta: v=−0.001ωpv = -0.001\omega_p; brown: v=−0.002ωpv = -0.002\omega_p; dark green: v=−0.003ωpv = -0.003\omega_p.

Having established the broad tunability of angle-steering via an anisotropic actual-gain system, in the following we show that a completely passive (lossy) anisotropic medium, excited by a complex-frequency wave (see, e.g., [22]), behaves exactly as its anisotropic gain counterpart excited by a real-frequency wave. In particular, we show that it is possible to attain angle steering of particles’ scattering, simply by changing the shape (envelope) of the incoming radiation rather than by changing the gain of the medium. This equivalence is based on the concept of virtual gain (which in our case,

is present in the non-diagonal terms of the permittivity tensor as well) since the outgoing wave carries away more energy than the incoming one, due to the transient decay of the latter. For this purpose, we examine exactly the same configuration discussed above, except that now the InSb shell is passive and lossy. This is achieved by setting positive values to the damping angular frequency ν\nu in InSb’s permittivity tensor. In Figure 3(a) we depict the Qsc/aQ_{\mathrm{sc}}/a spectrum at on-state (B0=0.1B_0 = 0.1 T), in the same frequency window as in Figure 1(c), using the lossy InSb with ν=+0.001ωp\nu= +0.001\omega_p (α=+0.001\alpha= +0.001), under real-frequency excitation (ν≡f\nu\equiv f, γ=0\gamma= 0). We observe that the lossy system under real-frequency excitation does not yield an ED-EQ intersection, meaning that a higher-order Kerker point does not exist for such material choices. However, the higher-order Kerker point emerges if the system is excited by a complex-frequency wave. To show this, in Figure 3(b) we plot the Qsc/aQ_{\mathrm{sc}}/a spectrum when the lossy system is illuminated by a complex-frequency plane-wave having ν=f+iγ\nu= f+i\gamma with γ=2\gamma= 2 GHz. This can be interpreted as a decaying wave in time-domain (as illustrated in the inset of Figure 3(b)) of the form cos⁡(2πft)exp⁡(−Γt)\cos(2\pi f t)\exp(-\Gamma t), where Γ=2πγ\Gamma= 2\pi\gamma is the decay factor that defines the envelope of the incoming radiation [37]. The higher-order Kerker point is then emerged at the same position (f=2.016f = 2.016 THz) as the respective one of the active system (see Figure 1(c)). It is worth noting that, the spectra in Figure 1(c) and Figure 3(b) have striking similarities. This is explained by the fact that the material response in those two cases is identical, as detailed in Section V of the SM.

Figure 3. (a) Scattering cross-section spectrum assuming the typical lossy InSb material (ν=0.001ωp\nu= 0.001\omega_p), for the same frequency window as in Figure 1(c). The rest values of parameters and legends are the same as in Figure 1(c). The gray dotted arrowhead indicates that there is no intersection between ED-EQ. (b) Same as (a) but assuming a complex-frequency incident wave with γ=2\gamma= 2 GHz. Inset: the system is now illuminated by a decaying plane-wave with decay factor Γ=2πγ\Gamma= 2\pi\gamma. The magenta circle indicates the appearance of the ED-EQ intersection and the magenta dotted arrowhead indicates the f=2.016f = 2.016 THz of the higher-order ED-EQ Kerker point.

Figure 4(a) shows the radiation patterns for three different values of γ\gamma, namely, γ=2\gamma= 2 GHz, 33 GHz and 44 GHz. The 22 GHz case is equivalent to the ν=−0.001ωp\nu= -0.001\omega_p active AMO, as explained in Figure 3(c)–(Figure 3d). The 33 GHz and 44 GHz cases correspond to ν=−0.002ωp\nu= -0.002\omega_p and ν=−0.003ωp\nu= -0.003\omega_p, respectively, therefore AMO increases its virtual gain. The σ/a\sigma/a for ν=−0.002ωp\nu= -0.002\omega_p and ν=−0.003ωp\nu= -0.003\omega_p values, however, is also plotted in Figure 2(c). The radiation patterns depicted in Figure 4(a) and Figure 2(c) are identical. In Section VI of the SM we show that our main conclusions are not limited to the specific material and geometrical parameters employed in the above analysis. In case where the multipolar decomposition is not available, Kerker points can be located by obtaining σ/a\sigma/a patterns similar to those of Figure 4(a) and Figure 2(c), as we show for the non-circular structures in Section III of the SM, or experimentally as shown in Ref. [38]. In addition, the lossy system also supports a switching from transparency to a directional super-scattering pattern, similar to the one of the active system depicted in Figure 1(d). This is further discussed in Section VII of the SM. These observations justify the equivalence between the active system/real-frequency excitation (actual gain) and the lossy system/complex-frequency excitation (virtual gain). The dependence of Δϕ\Delta\phi and RA vs γ\gamma is further discussed in Section I of the SM.

Figure 4. (a) Radiation pattern of normalized σ/a\sigma/a under complex-frequency ν=f+iγ\nu= f+i\gamma excitation; the real-frequency f=2.016f = 2.016 THz is kept fixed. Magenta: γ=2\gamma= 2 GHz; brown: γ=3\gamma= 3 GHz; dark green: γ=4\gamma= 4 GHz. (b) Poles of ama_m on complex plane ν=f+iγ\nu= f+i\gamma. Crosses: poles (ν1=1.911+i0.004057\nu_1 = 1.911+i0.004057 THz for m=+2m=+2, ν2=2.018+i0.00445\nu_2 = 2.018+i0.00445 THz for m=−2m=-2); squares: location of the complex-frequency excitations used in (b).

Finally, in Figure 4(b) we perform a stability analysis by seeking the poles of the ama_m expansion coefficients on the complex plane ν=f+iγ\nu=f+i\gamma. As shown, poles exist only for m=±2m=\pm2 and only in the upper-half complex plane (time dependence convention exp⁡(iωt)\exp(i\omega t)), i.e., our lossy system is stable under real- and complex-frequency excitations. These poles are attributed solely to the shell, as is deduced by the analysis offered in Section VIII of the SM.

In conclusion, we presented the concepts of anisotropic virtual gain and oblique Kerker effect, where a passive anisotropic medium mimicks the behavior of its active (gain) counterpart upon judicious complex-frequency excitation. This allows for large tuning of, both, the angle and magnitude of a particle’s scattering, simply by modulating wisely in time the incident radiation. Our work introduces a new, practical method for tuning the scattering properties of particles, which are of paramount importance for modern photonic applications. In future studies we similarly aim at studying/constructing further classes of lossy/active media [39, 40], as well as magneto-optical [41, 42] or time-varying ones [43].

Acknowledgement

G.P.Z., I.L., E.A. and K.L.T. acknowledge support for this research by the General Secretariat for Research and Technology (GSRT) and the Hellenic Foundation for Research and Innovation (HFRI) under Grant No. 4509.

References

References

  1. [1]H. Zhang, L. Zhang, and X. Yu, China Communications 18, 153 (2021).
  2. [2]S. Schwarz and S. Pratschner, IEEE Communications Magazine 61, 64 (2023).
  3. [3]J.-X. Wang, M. Zhong, Z. Wu, M. Guo, X. Liang, and B. Qi, Applied Energy 322, 119517 (2022).
  4. [4]G. Valušis, A. Lisauskas, H. Yuan, W. Knap, and H. G. Roskos, Sensors 21, 4092 (2021).
  5. [5]Y. Chen, W. Du, Q. Zhang, O. Ávalos-Ovando, J. Wu, Q.-H. Xu, N. Liu, H. Okamoto, A. O. Govorov, Q. Xiong, et al., Nature Reviews Physics 4, 113 (2022).
  6. [6]R. R. Naraghi, S. Sukhov, and A. Dogariu, Opt. Lett. 40, 585 (2015).
  7. [7]C. Argyropoulos, P.-Y. Chen, F. Monticone, G. D’Agauanno, and A. Alù, Phys. Rev. Lett. 108, 263905 (2012).
  8. [8]J. Xue, X. Zeng, L. Guo, R. Guo, Z. Zhang, C. Chu, Z. Cheng, N. Zhang, L. Lin, and W. Liu, Optics & Laser Technology (2023).
  9. [9]X. Xu, M. Nieto-Vesperinas, C.-W. Qiu, X. Liu, D. Gao, Y. Zhang, and B. Li, Laser & Photonics Reviews 14, 1900265 (2020).
  10. [10]Y. Yang, A. E. Miroshnichenko, S. V. Kostinski, M. Odit, P. Kapitanova, M. Qiu, and Y. S. Kivshar, Physical Review B 95, 165426 (2017).
  11. [11]M. Kerker, The scattering of light and other electromagnetic radiation (Elsevier, 2016).
  12. [12]H. K. Shamkhi, K. V. Baryshnikova, A. Sayan-skiy, P. Kapitanova, P. D. Terekhov, P. Belov, A. Karabchevsky, A. B. Evlyukhin, Y. Kivshar, and A. S. Shalin, Physical review letters 122, 193905 (2019).
  13. [13]A. Krasnok and A. Alù, Proceedings of the IEEE 108, 628 (2020).
  14. [14]K. L. Tsakmakidis, Metamaterials and Nanophotonics: Principles, Techniques and Applications (World Scientific, 2022).DOI
  15. [15]K. L. Tsakmakidis, T. W. Pickering, J. M. Hamm, A. F. Page, and O. Hess, Phys. Rev. Lett. 112, 167401 (2014).
  16. [16]E. I. Kirby, J. M. Hamm, T. W. Pickering, K. L. Tsakmakidis, and O. Hess, Phys. Rev. B 84, 041103 (2011).
  17. [17]K. L. Tsakmakidis, L. Shen, S. A. Schulz, X. Zheng, J. Upham, X. Deng, H. Altug, A. F. Vakakis, and R. Boyd, Science 356, 1260 (2017).
  18. [18]H. Li, A. Mekawy, A. Krasnok, and A. Alù, Phys. Rev. Lett. 124, 193901 (2020).
  19. [19]S. Kim, Y.-G. Peng, S. Yves, and A. Alù, Loss compensation and super-resolution with excitations at complex frequencies (2023), 2303.17699.arxiv.org/abs/2303.17699
  20. [20]F. Guan, X. Guo, K. Zeng, S. Zhang, Z. Nie, S. Ma, Q. Dai, J. Pendry, X. Zhang, and S. Zhang, Science 381, 766 (2023).
  21. [21]Y. Ra’di, A. Krasnok, and A. Alù, ACS Photonics 7, 1468 (2020).
  22. [22]S. Leshpov and A. Krasnok, Optica 7, 1024 (2020).
  23. [23]Z. Gu, H. Gao, H. Xue, J. Li, Z. Su, and J. Zhu, Nature Communications 13, 7668 (2022).
  24. [24]K. L. Tsakmakidis, L. Shen, S. A. Schulz, X. Zheng, J. Upham, X. Deng, H. Altug, A. F. Vakakis, and R. W. Boyd, Science 356, 1260 (2017).
  25. [25]G. P. Zouros, G. D. Kolezas, E. Almpanis, and K. L. Tsakmakidis, ACS Photonics 8, 1407 (2021).
  26. [26]J. C. Palais, IEEE Trans. Antennas Propag. 11, 505 (1963).
  27. [27]K. Katsinos, G. P. Zouros, and J. A. Roumeliotis, IEEE Trans. Antennas Propag. 69, 2256 (2021).
  28. [28]M. Q. Liu, C. Y. Zhao, and B. X. Wang, Nanoscale 10, 18282 (2018).
  29. [29]W. Liu and Y. S. Kivshar, Opt. Express 26, 13085 (2018).
  30. [30]C. Qian, X. Lin, Y. Yang, X. Xiong, H. Wang, E. Li, I. Kaminer, B. Zhang, and H. Chen, Phys. Rev. Lett. 122, 063901 (2019).
  31. [32]S. Lepeshov, A. Krasnok, and A. Alù, ACS Photonics 6, 2126 (2019).
  32. [33]E. Almpanis, Physical Review B 97, 184406 (2018).
  33. [34]K. Katsinos, G. P. Zouros, G. D. Kolezas, G. K. Pagiatakis, and J. A. Roumeliotis, IET Microwaves, Antennas & Propagation 15, 1206 (2021).
  34. [35]F. Mignerot, B. Kedjar, H. Bahsoun, and L. Thilly, Scientific Reports 11, 19441 (2021).
  35. [36]Y. Yang, B. Cui, Z. Geng, and S. Feng, Applied Physics Letters 106 (2015).
  36. [37]I. Loulas, G. P. Zouros, E. Almpanis, and K. L. Tsakmakidis, arXiv preprint arXiv:2305.04509 (2023).arxiv.org/abs/2305.04509
  37. [38]J.-M. Geffrin, B. García-Cámara, R. Gómez-Medina, P. Albella, L. Froufe-Pérez, C. Eyraud, A. Litman, R. Vaillon, F. González, M. Nieto-Vesperinas, et al., Nat. Commun. 3, 1171 (2012).
  38. [39]K. L. Tsakmakidis, A. D. Boardman, and O. Hess, Nature 455, E11 (2008).
  39. [40]E. I. Kirby, J. M. Hamm, T. W. Pickering, K. L. Tsakmakidis, and O. Hess, Phys. Rev. B Condens. Matter Mater. Phys. 84 (2011).
  40. [41]E. Almpanis, G. P. Zouros, P. A. Pantazopoulos, K. L. Tsakmakidis, N. Papanikolaou, and N. Stefanou, Phys. Rev. B. 101 (2020).
  41. [42]P. A. Pantazopoulos, K. L. Tsakmakidis, E. Almpanis, G. P. Zouros, and N. Stefanou, New J. Phys. 21, 095001 (2019).
  42. [43]F. Y. Gardes, K. L. Tsakmakidis, D. Thomson, G. T. Reed, G. Z. Mashanovich, O. Hess, and D. Avitabile, Opt. Express 15, 5879 (2007).

Paper details

vlnc:W_t63mfdjv
Version
v1 · latest
Published
Fresh — added to the Hub within the last 30 days
Added
License
CC BY 4.0
Usage

views · 0

downloads · 0

A view counts once per browser per day, a download each time the PDF is requested.

  • Physics and Astronomy

Not recorded for this paper

View statement
Looking for related papers…
Contents