• Chinese Optics Letters
  • Vol. 22, Issue 3, 031901 (2024)
Zhaofeng Liu1, Siwei Tang1, Zengrun Wen1、*, Yuanmei Gao1, Yangjian Cai1、2、**, and Liren Zheng1、***
Author Affiliations
  • 1Shandong Provincial Engineering and Technical Center of Light Manipulations & Shandong Provincial Key Laboratory of Optics and Photonic Device, School of Physics and Electronics, Shandong Normal University, Jinan 250358, China
  • 2School of Physical Science and Technology, Soochow University, Suzhou 215006, China
  • show less
    DOI: 10.3788/COL202422.031901 Cite this Article Set citation alerts
    Zhaofeng Liu, Siwei Tang, Zengrun Wen, Yuanmei Gao, Yangjian Cai, Liren Zheng, "Talbot effect in anti-PT symmetric synthetic photonic lattices," Chin. Opt. Lett. 22, 031901 (2024) Copy Citation Text show less

    Abstract

    We investigated the Talbot effect in an anti-parity-time (PT) symmetric synthetic photonic lattice composed of two coupled fiber loops. We calculated the band structures and found that with an increase in the gain-loss parameter, the band transitions from a real spectrum to a complex spectrum. We study the influence of phase in the Hermitian operator on the Talbot effect, and the Talbot effect disappears when the period of the input field is N > 8. Further study shows that the variation of Talbot distance can also be modulated by non-Hermitian coefficients of gain and loss. This work may find significant applications in pulse repetition-rate multiplication, temporal invisibility, and tunable intensity amplifiers.

    1. Introduction

    The Talbot effect, initially discovered by H. F. Talbot in 1836[1], refers to the self-replication phenomenon of periodic structures[2] under quasi-monochromatic coherent light illumination, which directly arises from near-field Fresnel diffraction[3]. The Talbot distance ZT is defined as the position at which waves periodically reproduce themselves along the propagation direction[4]. In addition to optics, this effect has also been widely studied and applied in condensed matter[5,6], quantum revivals[7,8], X-ray[9,10], plasmonic[11,12], electron[13,14], prime-number decomposition[15], and so on. Currently, research on the optical Talbot effect has expanded from the spatial domain[1618] to the temporal domain[1921], from linear to nonlinear[22], and from continuous media to discrete systems[1618,23]. Unlike in continuous media, the occurrence of the Talbot effect in discrete media requires the pattern period to meet specific conditions[4,18], whether in one-dimension[24] or two-dimensions[25].

    The non-Hermitian Hamiltonians have completely real spectra if they obey parity-time (PT) symmetry, which was proved by Bender and their colleagues[26,27]. The anti-PT symmetry demands that the refractive index n(x)=nR(x)+inI(x) should satisfy n(x)=n*(x)[28,29]. This means that the real part nR(x), the phase distribution, should be an odd function, while the imaginary part nI(x), the gain/loss, should be an even function. PT-symmetry in optics was initially observed in the context of two coupled waveguides[30,31]. However, experimental achievement on an anti-PT symmetric large-scale lattice is highly limited. To overcome limitations, recently, there has been a generalization of the lattice concept from real space to synthetic dimensions, including temporal and spectral lattices[32,33]. By representing a temporal grid lattice, the evolution of pulses in two coupled fiber rings can be analogous to the dynamics of light beams in a spatial grid lattice[19]. Through temporal control of gain and loss within the lattice structure[34,35], researchers have, to the best of their knowledge, successfully achieved experimental realization of PT-symmetry in large-scale lattices for the first time[36].

    In this study, we employ an anti-PT symmetric synthetic photonic lattice composed of two coupled fiber loops. Electro-optic and acousto-optic modulators are utilized to adjust the phase distribution as well as gain and loss coefficients of the system. First, the band structure is calculated, followed by a theoretical analysis of the conditions for generating the Talbot effect and the calculation method of the Talbot distance. Additionally, simulations are conducted under two scenarios, Hermitian phase modulation and non-Hermitian modulation, to investigate the Talbot self-imaging phenomenon. The goal is to explore the influence of these different modulation methods on the Talbot effect and the characteristics of self-imaging.

    2. Model and Theory

    The anti-PT symmetric synthetic photonic lattice is formed by two coupled fiber loops, which are shown in Fig. 1(a). A 50:50 directional coupler connects two fiber rings with a length difference ΔL[19,29]. Once the coupler receives the initial light pulse injected into the long loop, it generates two new pulses that flow into each loop. After completing one round trip, the pulses in the short loop return to the coupler first, and subsequently propagate in the long loop. As a result, the two pulse trains interfere with each other through the coupler[20]. Like the dynamics of beams in spatial grids, the pulse train undergoes incremental evolution as the number of round trips increases. The equivalent lattice in the time domain is shown in Fig. 1(b), where m is the number of pulses entering the coupler, and n is the position of the pulse in the lattice.

    Schematic of anti-PT symmetric synthetic photonic lattice. (a) Two coupling fiber loops with a length difference and (b) corresponding temporal mesh lattice with anti-PT symmetry. (c) Distribution of the phase, gain, and loss.

    Figure 1.Schematic of anti-PT symmetric synthetic photonic lattice. (a) Two coupling fiber loops with a length difference and (b) corresponding temporal mesh lattice with anti-PT symmetry. (c) Distribution of the phase, gain, and loss.

    As mentioned in the introduction section, to guarantee the odd symmetry of the real part, the phase modulator needs to be adjusted so that φn satisfies φn={φ,n=2k+φ,n=2k+1,where φ is the amplitude of phase modulation, kZ. The distributions of φn, gain, and loss are shown in Fig. 1(c).

    Similarly, to ensure that the imaginary part is an even function, the acousto-optic modulator needs to be adjusted, and the gain (G>0) and loss (γ<0) are set as {γp=γ,γq=G,formod(n+3;4)=0γp=γ,γq=γ,formod(n+3;4)=1γp=G,γq=γ,formod(n+3;4)=2γp=G,γq=G.otherwise.Assuming that the amplitudes of the pulses in the short and long loops are Unm and Dnm, respectively, the pulse iteration equation governing the dynamics in the anti-PT symmetric synthetic photonic lattice can be expressed as {Unm+1=12eγp(Un+1m+iDn+1m)eiφnDnm+1=12eγq(Dn1m+iUn1m).

    To better study the optical pulse transmission characteristics in the synthesized photonic lattice, we use eikneiβm to solve the dispersion relation of the system. The Bloch mode is (Unm,Dnm)T=(U,D)Teikneiβm, where (U,D)T is the periodic Bloch function, k and β represent the transverse Bloch momentum and longitudinal propagation constant, respectively[36,37]. By inserting Eqs. (1) and (2) and the Bloch mode (Unm,Dnm)T=(U,D)Teikneiβm into Eq. (3), we can obtain the dispersion relation cos(4k)=2+4cos(4β)+cos(2φ)+8cos(2β)cos  φcosh(2G)+2cosh(4G).

    The band structure can be manipulated by adjusting the wave vector k, the phase φ, and the gain G (loss γ). For G=0, Eq. (4) becomes cos(4k)=4+4cos(4β)+cos(2φ)+8cos(2β)cos(φ). The band structures with different values of φ have been investigated[29].

    Next, as φ=π/2, the dispersion relationship is simplified as cos(4k)=1+4cos(4β)+2cosh(4G). Figure 2 illustrates the band structures with various gain G. In Fig. 2(a), the band structures become entirely real and exhibit anomalies. In Figs. 2(b)2(e), the bands exhibit both real and imaginary parts. However, in Fig. 2(f), the bands become entirely imaginary.

    Band structures for G = 0, G = 0.1, G = 0.2, G = 0.3, G = 0.327, and G = 0.33, respectively. φ = π/2. The blue line represents the real part of the band, and the red line represents the imaginary part of the band.

    Figure 2.Band structures for G = 0, G = 0.1, G = 0.2, G = 0.3, G = 0.327, and G = 0.33, respectively. φ = π/2. The blue line represents the real part of the band, and the red line represents the imaginary part of the band.

    Considering the essential prerequisites for the occurrence of the Talbot effect in the synthetic photonic lattice, we assume that the input field has the period N in the time domain grid. Due to transverse discreteness, N is an integer[19]. Therefore, the period of the incident pulse train in the coupling loop is NΔT. To satisfy the periodic boundary condition, the Bloch momentum is kl=2πl/M, where M=N/4 and l=0,1,2,,M1. The propagation constant corresponding to the Bloch momentum kl is expressed as βj,l, where j is the band index and is equal to 1, 2, 3, 4, 5, 6, 7, and 8. The propagation constant at the fourth band is β4,l=14arccos(14(1+cos(4kl)2cosh(4G))).

    The Bloch superposition is used to describe the evolution of the field, (UnmDnm)=j=18l=0M1cj,l(Uj,lDj,l)exp(ikln)exp(iβj,lm).

    Recovery can occur only when βj,l satisfies λj,lβj,l/2=2μj,lπ, where λj,l is the recovery distance, and μj,l is a nonzero integer. Since the time-domain grid is discrete in the propagation direction and λj,l is an integer, βj,l should be a rational multiple of π[19]. Therefore, the Talbot distance is ZT=LCM(λ1,0,λ1,1,,λ1,M,λ2,0,,λ8,M), where LCM is the least common multiple.

    3. Results and Discussion

    3.1. Talbot effect of Hermitian phase modulation

    First, we study the influence of the phase of the Hermitian operator on the Talbot effect. For N=4, the gain factor takes a value of G=0. Considering the periodic boundary conditions with k0=0, the propagation constant is β4,0=12arccos(12(1cosφ)), and the Talbot effect occurs when the mode of β4,0 can be replicated during propagation. Assuming efficient recovery of the β4,0 mode, we can express β4,0 as pπ/q, where p and q are relatively prime integers.

    For odd p and q, the modal recovery distance for β4,0 is λj,l=4q. For G=0, φ=0, and β4,0=π/4, according to the symmetry of the bands, the propagation constants are β1,0=3π/4, β2,0=π/2, and β3,0=π/2. The recovery distances are λ1,0=16, λ2,0=8, λ3,0=8, and λ4,0=16. Thus, the Talbot distance is ZT=16. In Fig. 3(a1), the evolution of the pulse intensity is the Talbot effect plot, where the pulse train achieves complete recovery every 16 steps. At m=8, the intensity distribution is the same as the initial distribution at m=0, which only accumulates a π phase delay. For φ=π/2, according to the symmetry of the bands, the corresponding propagation constants of the bands are β1,0=5π/6, β2,0=2π/3, β3,0=π/3, and β4,0=π/6. The recovery distances are λ1,0=24, λ2,0=12, λ3,0=12, and λ4,0=24. Thus, the Talbot distance is ZT=24. The corresponding Talbot diagram is presented in Fig. 3(b1). For N=8 and φ=π/2, the Bloch momentums are k0=0 and k1=π, and the propagation constants of the fourth band are β4,0=14arccos(12cosh(4G)),β4,1=14arccos(12cosh(4G)).

    (a1), (b1) The pulse intensity evolution in the long loop is shown for φ = 0 and φ = 0.5π, respectively. The input period is N = 4. (c1), (d1) The pulse intensity evolution in the long loop is shown for N = 8 and N = 12 with φ = 0. The green dotted line is the location of the Talbot images.

    Figure 3.(a1), (b1) The pulse intensity evolution in the long loop is shown for φ = 0 and φ = 0.5π, respectively. The input period is N = 4. (c1), (d1) The pulse intensity evolution in the long loop is shown for N = 8 and N = 12 with φ = 0. The green dotted line is the location of the Talbot images.

    To support recovery, β4,0 and β4,1 are both in the form of pπ/q. The Talbot effect occurs only when G equals zero, and the Talbot distance is ZT=24, as shown in Fig. 3(c1). Due to the periodic incidence conditions and transverse discreteness, the set of N should be restricted to {4,8,12,,2ML}, where ML is the longest transmission distance. For N>8, the Talbot effect does not occur. We use the case of N=12 as an example for illustration. For N=12, the corresponding Bloch momentum is k1=2π/3, and the propagation constant is β4,1=14arccos(78) and is no longer rational multiples of π. Consequently, the condition required for the Talbot effect, as illustrated in Fig. 3(d1), is not satisfied, the propagation image along the propagation direction does not exhibit periodic recovery, and the energy shows irregular oscillation.

    On the other hand, the Talbot distance can be manipulated by applying linear phase modulation to the incident pulse train. We denote the interval between adjacent pulse phase shifts as φ0. In Fig. 4, the input field of the temporal grid also has an initial momentum φ0, which causes the Bloch momentum to become kl=2πl/M+φ0. For N=4, the fourth band with the corresponding transmission constant is β4,0=14arccos(14(1+cos(4φ0)2cosh(4G))). The Talbot effect exists only when β4,0 is a rational multiple of π, and we express β4,0 as aπ/b, where a and b are positive integers of relative prime numbers. Since π<φ0<π and 0G<0.33, the range of values for β4,0 is (π/4,π/3]. b should be at least 3, and the corresponding Talbot distance is ZT=12. The set of Talbot distances is denoted by {12,16,20,,4b,,ML}. For G=0, when φ0 takes the values of 0.36π and 0.46π, it corresponds to ZT=28 and ZT=16, respectively. For G=0.21, when φ0 takes the values of 0.15π and 0.5π, it corresponds to ZT=28 and ZT=36, respectively. The corresponding Talbot images are depicted in Figs. 4(a1)4(d1).

    Input pulse period N = 4. (a1), (b1) Talbot carpets for φ0 = 0.36π and φ0 = 0.46π. G = 0. (c1), (d1) Talbot carpets for φ0 = 0.15π and φ0 = 0.5π. G = 0.21.

    Figure 4.Input pulse period N = 4. (a1), (b1) Talbot carpets for φ0 = 0.36π and φ0 = 0.46π. G = 0. (c1), (d1) Talbot carpets for φ0 = 0.15π and φ0 = 0.5π. G = 0.21.

    3.2. Non-Hermitian modulated Talbot effect

    The Talbot effect can also be observed by modifying the non-Hermitian coefficients of gain and loss. For N=4 and φ=0, the fourth band propagation constant is β4,0=arccos12cosh(G)24sinh(G)24+1414cosh(G)2sinh(G)2. When the gain coefficients are G=0.29 and G=0.35, respectively, the propagation constants of the fourth band are β4,0=3π/8 and β4,0=3π/10. The corresponding Talbot distance is ZT=32 and ZT=40, and the Talbot images are shown in Figs. 5(a1) and 5(b1).

    (a1), (b1) Talbot carpets for G = 0.29 and G = 0.35. φ = 0. (c1), (d1) Talbot carpets for G = 0.25 and G = 0.31. φ = π/2. (e1), (f1) Talbot carpets for G = 0.29 and G = 0.35. φ = π.

    Figure 5.(a1), (b1) Talbot carpets for G = 0.29 and G = 0.35. φ = 0. (c1), (d1) Talbot carpets for G = 0.25 and G = 0.31. φ = π/2. (e1), (f1) Talbot carpets for G = 0.29 and G = 0.35. φ = π.

    Then, we investigated the case where N equals 4 and φ equals π. We found that when the gain-loss coefficient and the input pulse period are identical, the Talbot effect and the amplitude of energy remain consistent for the two initial phases of φ=0 and φ=π.

    For N=4 and φ=π/2, according to Fig. 2, the first and second bands are symmetrical with respect to β=3π/4, the third band is symmetrical with the fourth band β=π/4, and the first (second) band is symmetrical with the fourth (third) band about β=π/2. The fourth band propagation coefficient is β4,0=14arccos(12cosh(4G)). The Talbot effect occurs when β4,0 is a rational multiple of π, expressed as pπ/q. In this case, the Talbot distance can be calculated as 4q. Considering that the gain-loss factor falls within the range of 0G<0.33, the propagation constant β4,0 is in the range of π/8 to π/4, with q being at least 5. Thus, the Talbot distance falls within the range {20,24,28,,4q,,ML}. Accompanying the energy oscillations during Talbot imaging, for G=0.25, the propagation constant of the fourth band is β4,0=π/5. Consequently, the corresponding Talbot distance is ZT=20, which is illustrated in Fig. 5(c1). For G=0.31, the propagation constant of the fourth band is β4,0=2π/9, corresponding to the Talbot distance ZT=36, as shown in Fig. 5(d1). From Figs. 5(c1) and 5(d1), the amplitude of energy oscillation increases as the gain-loss factor augments. For G0.33, the band structure becomes entirely imaginary; as illustrated in Fig. 2(f), the input field distribution cannot be recovered due to varying degrees of amplification or attenuation experienced by these modes. Therefore, when the included modes are excited in the purely imaginary region of the band structure, the Talbot effect does not occur. We attempted to utilize the enumeration method to detect the presence of the Talbot effect for N>4. However, the Talbot effect does not occur.

    The influence of parameters φ and γ on the energy bands is depicted in Fig. 6. When above the curve, all eigenvalues of the system are completely imaginary. When on the curve, symmetry breaking occurs. Below the line, the eigenvalues of the system are complex, representing a complex spectrum with both real and imaginary components. It is evident that the curve exhibits symmetry with respect to the φ=π/2 axis.

    The curve representing the anti-PT symmetry breaking threshold in a two-dimensional parameter space of φ and γ.

    Figure 6.The curve representing the anti-PT symmetry breaking threshold in a two-dimensional parameter space of φ and γ.

    4. Conclusion

    In summary, we constructed an anti-PT symmetric synthetic photonic lattice using two coupled fiber loops. We presented the corresponding equivalent lattice diagram and derived the pulse amplitude expression for light pulses in the double loops. Then, we obtained the dispersion relation of the system using discrete plane wave solutions and plotted the band structure by varying phase values and gain-loss coefficients. Furthermore, we analyzed and studied the Talbot effect in the anti-PT symmetric synthetic photonic lattice under both Hermitian and non-Hermitian phase modulation. The research revealed that when the input period is N=2,4,6,8, the Talbot distance can be flexibly controlled by changing the gain and loss factors or linearly modulating the input pulse sequence. Additionally, at φ=π/2, when the gain-loss factor is greater than or equal to 0.33, the Talbot effect disappears. If the pulse input period, gain, and loss coefficients are the same, the band structures corresponding to the pulse phases φ=0 and φ=π exhibit the same patterns, resulting in the same Talbot effect and associated energy oscillations. This work may find promising applications in pulse repetition-rate multiplication, temporal cloaking, and tunable intensity amplifiers.

    References

    [1] H. F. Talbot. LXXVI. Facts relating to optical science. No. IV. Philos. Mag., 9, 401(2009).

    [2] L. A. Hall, S. Ponomarenko, A. F. Abouraddy. Temporal Talbot effect in free space. Opt. Lett., 46, 3107(2021).

    [3] J. T. Winthrop, C. R. Worthington. Theory of Fresnel images. I. Plane periodic objects in monochromatic light. J. Opt. Soc. Am. A., 55, 373(1965).

    [4] K. Zhan, X. Kang, Q. Zhang et al. Discrete Talbot effect in reciprocal and nonreciprocal dimer lattices. J. Opt. Soc. Am. B., 39, 3283(2022).

    [5] V. Makhalov, A. Turlapov. Order in the interference of a long chain of Bose condensates with unrestricted phases. Phys. Rev. Lett., 122, 090403(2019).

    [6] J. Feng, Z. Zhang, M. Xiao. Talbot effects induced by gain-loss modulated optical lattices in a coherent atomic medium. Phys. Rev. A., 103, 063516(2021).

    [7] F. Chamizo, O. P. Santillán. About the quantum Talbot effect on the sphere. J. Phys. A, 56, 255302(2023).

    [8] P. Mantashyan, G. Mantashian, D. Hayrapetyan. Talbot effect in InAs/GaAs coupled cylindrical quantum dots ensemble. Phys. E, 148, 115662(2023).

    [9] M. Balyan, L. Levonyan, K. Trouni. X-ray dynamical diffraction Talbot effect behind a crystal in free space. Acta. Cryst. A, 77, 149(2021).

    [10] M. K. Balyan. X-ray dynamical diffraction analogues of the integer and fractional Talbot effects. J. Synchrotron Rad., 26, 1650(2019).

    [11] H. Ma, S. Su, H. Zhou et al. Tunable plasmonic talbot effect based on graphene monolayer. Appl. Sci., 10, 4782(2020).

    [12] S. Farhadi, A. Farmani, A. Hamidi. Figure of merit enhancement of surface plasmon resonance biosensor based on Talbot effect. Opt. Quantum Electron., 53, 518(2021).

    [13] M. V. Tsarev, A. Ryabov, P. Baum. Free-electron qubits and maximum-contrast attosecond pulses via temporal Talbot revivals. Phys. Rev. Res., 3, 043033(2021).

    [14] G. Calogero, N. R. Papior, B. Kretz et al. Electron transport in nanoporous graphene: probing the Talbot effect. Nano. Lett., 19, 576(2019).

    [15] K. Pelka, J. Graf, T. Mehringer et al. Prime number decomposition using the Talbot effect. Opt. Express, 26, 15009(2018).

    [16] K. Li, F. Xia, M. Wang et al. Discrete Talbot effect in dielectric graphene plasmonic waveguide arrays. Carbon, 118, 192(2017).

    [17] K. Zhan, L. Dou, R. Jiao et al. Talbot effect in arrays of helical waveguides. Opt. Lett., 46, 322(2021).

    [18] X. Kang, H. Wang, L. Dou et al. Discrete Talbot effect in modulated lattices with PT symmetric perturbations. J. Opt. Soc. Am. B., 40, 856(2023).

    [19] S. Wang, B. Wang, P. Lu. PT-symmetric Talbot effect in a temporal mesh lattice. Phys. Rev. A., 98, 043832(2018).

    [20] S. Wang, C. Qin, B. Wang et al. Discrete temporal Talbot effect in synthetic mesh lattices. Opt. Express, 26, 19235(2018).

    [21] Z. Wen, B. Lu, K. Wang et al. Discrete optical propagation in one-dimensional synthetic mesh lattice. J. Opt. Soc. Am. B, 37, 3152(2020).

    [22] L. Li, H. Liu, X. Chen. Dynamic manipulation of nonlinear Talbot effect with structured light. Opt. Lett., 46, 1281(2021).

    [23] K. Zhan, L. Dou, X. Kang et al. Controllable discrete Talbot self-imaging effect in Hermitian and non-Hermitian Floquet superlattices. Opt. Express, 30, 35256(2022).

    [24] R. Iwanow, D. A. May-Arrioja, D. N. Christodoulides et al. Discrete Talbot effect in waveguide arrays. Phys. Rev. Lett., 95, 053902(2005).

    [25] Z. Chen, Y. Zhang, M. Xiao. Discrete Talbot effect in two-dimensional waveguide arrays. Opt. Express, 23, 14724(2015).

    [26] C. M. Bender, S. Boettcher. Real spectra in non-Hermitian Hamiltonians having PT symmetry. Phys. Rev. Lett., 80, 5243(1998).

    [27] D. C. Brody, C. M. Bender, H. F. Jones. Complex extension of quantum mechanics. Phys. Rev. Lett., 89, 270401(2002).

    [28] L. Ge, H. E. Türeci. Antisymmetric PT-photonic structures with balanced positive- and negative-index materials. Phys. Rev. A., 88, 053810(2013).

    [29] Y. Yue, J. Li, Z. Zhang et al. Interaction between double solitons in anti-PT symmetric synthetic photonic lattices. Opt. Commun., 529, 129082(2023).

    [30] A. Guo, G. J. Salamo, D. Duchesne et al. Observation of PT-symmetry breaking in complex optical potentials. Phys. Rev. Lett., 103, 093902(2009).

    [31] C. E. Rüter, K. G. Makris, R. El-Ganainy et al. Observation of parity–time symmetry in optics. Nat. Phys., 6, 192(2010).

    [32] S.-L. Wang, B. Wang, C.-Z. Qin et al. Rabi oscillations of optical modes in a waveguide with dynamic modulation. Opt. Quantum Electron., 49, 389(2017).

    [33] I. D. Vatnik, A. Tikan, G. Onishchukov et al. Anderson localization in synthetic photonic lattices. Sci. Rep., 7, 4301(2017).

    [34] R. El-Ganainy, K. G. Makris, D. N. Christodoulides et al. Theory of coupled optical PT-symmetric structures. Opt. Lett., 32, 2632(2007).

    [35] Z. H. Musslimani, K. G. Makris, R. El-Ganainy et al. Optical solitons in PT periodic potentials. Phys. Rev. Lett., 100, 030402(2008).

    [36] A. Regensburger, C. Bersch, M. A. Miri et al. Parity-time synthetic photonic lattices. Nature, 488, 167(2012).

    [37] M. Mohammad-Ali, A. Regensburger, U. Peschel et al. Optical mesh lattices with PT symmetry. Phys. Rev. A, 86, 023807(2012).

    Zhaofeng Liu, Siwei Tang, Zengrun Wen, Yuanmei Gao, Yangjian Cai, Liren Zheng, "Talbot effect in anti-PT symmetric synthetic photonic lattices," Chin. Opt. Lett. 22, 031901 (2024)
    Download Citation