## Abstract

Utilizing the robust transport properties of the topological photonic crystal interface, we experimentally realize two-dimensional topological photonic crystal cavities, where discrete whispering gallery modes can propagate unidirectionally along the cavity circumference. Different from traditional cavities, these topological whispering galley modes are insensitive to cavity shapes. Our microwave demonstration has a good agreement with numerical simulations. Using pure dielectrics, by scaling down to the optical wavelength, an optical directional coupler based on the same topological photonic crystal scheme is also proposed. We here show that topological photonics can provide more novel designs for optical devices.

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

## 1. Introduction

Recently, topological insulators have received considerable attentions because of intriguing properties that topologically protected surface states are immune to backscattering, even at the presence of large impurities [1,2]. In analogy to electronic systems [3], optical quantum Hall effect was proposed theoretically using photonic crystals (PCs) [4] and unidirectional electromagnetic propagations were observed in microwave experiments [5]. These efforts opened the research area of topological photonics. PCs, consisting of periodic arrangement of low loss dielectrics, possess the ability to control the propagation of electromagnetic waves for a wide range of frequencies [6–9] which are a perfect platform to realize topological photonic properties. Optical and acoustic analogy to quantum spin Hall effect [10,11] were soon proposed and realized [12–18], where spin-polarized one-way edge states exist at the interfaces of photonic topological insulators. Using dielectric materials only is crucial that these unidirectional edge states can be extended to optical frequencies contributing to future optical communications. By engineering *C _{6v}* lattice symmetry in dielectric PCs [14] at microwave frequencies, we succeeded in demonstrating unidirectional photonic channels [17].

Whispering gallery mode (WGM) exists due to the total internal reflection on concave surfaces. As light wave of certain frequencies can be confined and allowed given a certain geometry, even a single molecular biosensing is possible using WGMs [19–22]. Various structures have been explored to sustain WGMs including microspheres, micro-rings and optical fibers [23–26]. By combining optical topological insulators and microcavity geometries, novel design principles of lasers were proposed [27,28] and other microcavities were also been designed for different applications [29]. Finite topological PC particles with discrete interface states were also theoretically discussed [30].

In this paper, we experimentally investigate the WGMs existing in two-dimensional (2D) PC cavities consisting of dielectric materials. Different from previously studied infinite system, we observe discrete WGMs propagating unidirectionally along the topological PC cavity circumference. In contrast to previous WGMs that have a strong dependence to cavity geometry, topological WGMs remain unchanged if we keep the same circumferences but allow different shapes in topological PC cavities. As pure dielectric PCs with reasonable permittivities are used in our microwave experiments, the structure can be scaled down to optical frequencies and a topological optical coupler is suggested.

## 2. Topological photonic crystal cavity

The schematic to construct WGM topological cavity using 2D PCs is depicted in Fig. 1(a). Six alumina cylinders in air constitute one unit cell of the triangular lattice of PC_{1} and PC_{2}, whose lattice constant is *a*, relative permittivity of alumina *ε* = 7.5 and their diameters *d* = 0.24*a*. As enlarged in the inset, for such a composite artificial molecule, two geometrical parameters determine its photonic band properties: the intra (inter) couple distance *h _{1}* (

*h*) defined as the distance between neighboring (nearest) cylinders, where 2

_{2}*h*+

_{1}*h*=

_{2}*a*. The intra (inter) couple distance is

*h*= 0.36

_{1}*a*(

*h*= 0.28

_{2}*a*) for PC

_{1}, while

*h*= 0.30

_{1}*a*(

*h*= 0.4

_{2}*a*) for PC

_{2}. As discussed in [17], the change between

*h*and

_{1}*h*induces a topological phase transition of the photonic band properties of the corresponding PCs and a topological (with

_{2}*h*>

_{1}*h*) PC

_{2}_{1}and a trivial (with

*h*<

_{1}*h*) PC

_{2}_{2}can be constructed, as shown in red and blue cylinders respectively. In [17], at the interface between semi-infinite PC

_{1}and PC

_{2}, topological interface states with in-plane orbital angular momenta were observed and their unidirectional propagations were visualized.

Now we extend to a finite situation, where finite cavities made of topological PC_{1} are considered. We start with a hexagon-shape cavity first. The side length of the topological PC cavity is *R* = 3*a* surrounding by trivial PC_{2}. The surrounding trivial PC_{2} layers are set larger than the topological PC cavity side length to avoid unwanted coupling between topological cavities when periodic boundary conditions are applied. We consider Transverse Magnetic (TM, with ${E}_{z}$ polarization) polarization only in this work. Though this is a finite system, we still can obtain its photonic modes by doing eigenmode analyses of a periodic supercell as in Fig. 1(a) using finite element numerical solver COMSOL Multiphysics when setting the wave vector ** k** = 0. The calculated modes are shown in the left panel of Fig. 1(b). The modes obtained within the common bulk photonic band gap frequencies between PC

_{1}and PC

_{2}correlate to the interface between topological and trivial PCs, as indicated by red dots where modes corresponding to either bulk PC

_{1}or PC

_{2}bands are shown in black dots. Five sets of discrete topological interface states are found numerically for the

*R*= 3

*a*cavity, as labeled next to the dots. The reason that WGMs exist in a microcavity is that constructive interference occurs when the interface states wind around the microcavity:

*L*is the circumference of the cavity, and

*m*is the index. In Fig. 1(c), we plot the eigenfield distributions of all WGMs, where plus ( + ) and minus (-) signs are for eye-guiding purpose to indicate the phase difference in eigenfields. From Eq. (1),

*m*shall reflect the repeat times of +/− phase variations and we use it to index the WGMs. The number of sets of modes can thus be understood as a purely geometrical effect. For a cavity with certain circumference

*L*and index

*m*, only one ${k}_{//}(\omega )$ lies in the original dispersion of interface state shall satisfy Eq. (1). As we only have a limited choice of ${k}_{//}(\omega )$ determined by the photonic band properties of PC

_{1}and PC

_{2}, the number of sets of modes is expected to positively correlate to the circumference

*L*. We maintain the topological cavity to be hexagon and consider different sizes. When

*R*= 2

*a*as shown in the middle panel of Fig. 1(b), only four sets of topological interface modes exist. While the side length of a hexagonal cavity is increased to

*R*= 6

*a*as shown in the right panel of Fig. 1(b), the number of sets increases to nine. Similar to traditional WGMs, the number of modes increases if the cavity size increases. Owing to the finite-sized effect of PC cavity, single WGM may split in spectrum, as indicated by a prime.

For traditional WGMs, Eq. (1) is satisfied only in the situation that an adiabatic change on the circumference occurs. When there is a sharp bending, the propagating wave number ${k}_{//}(\omega )$will change dramatically and thus WGM frequencies are very sensitive to the shapes of cavities. However, the interface state between a topological PC and a trivial one can even propagate at the existence of sharp corners [17]. In other words, these sharp bending and corners will not alter its propagation properties and WGM frequencies shall be insensitive to the shape of topological cavities. When we modify the shape of topological cavity from a hexagon to a rhombus, the number of topological WGMs and their eigenfield distributions are unchanged. Figure 2(a) depicts schematics of four rhombic cavities (left panels) and their eigenfield distributions of the zeroth-order modes (right panels). Even though the shapes of rhombic cavities are different, we maintain their circumferences to be the same as 18*a*, which is also the circumference of the hexagonal cavity in Fig. 1. The bottom edges of four rhombic topological cavities are 5*a*, 4*a*, 3*a* and 2*a* in turn. In all scenarios, five sets of WGMs are observed, which is a direct consequence of Eq. (1) with the same circumference *L*, and these modes frequencies and corresponding eigenfield distributions can be found in Figs. 2(b) and 2(a) respectively. This unique mode property of topological PC cavity owes to the unidirectional propagation of interface states allowed on the cavity circumference. Even if there are defects or fabrication imperfections on the PC cavities, as no backscattering is allowed, the WGMs are not perturbed. We acquire similar results for cavities with rectangle or triangle shapes in numerical simulations.

## 3. Experimental results

We carry out microwave experiments to verify and detect topological WGMs in PC cavity. The photo of experimental setup can be found in Fig. 3(a). Alumina cylinders with height *h* = 8 mm, relative permittivity *ε* = 7.5 and diameter *d* = 6 mm are used to assemble into PC_{1} and PC_{2} structures as shown in Fig. 1(a). The intra (inter) couple distance is *h _{1}* = 9 mm (

*h*= 7 mm) for topological PC

_{2}_{1}highlighted by orange hexagonal region, while

*h*= 7.5 mm (

_{1}*h*= 10 mm) for trivial PC

_{2}_{2}. The lattice constant is

*a*= 25 mm and the hexagon cavity size is

*R*= 3

*a*= 75 mm. A chiral excited source indicated by a dashed circle locates at the center of one unit cell of topological PC

_{1}. The pseudospin up topological interfacial state (with positive in-plane angular momentum, as denoted by the black circular arrow) is excited and propagates anti-clockwise along the cavity. The chiral source is constructed by four antennas with an anticlockwise π/2 phase decrease between neighbors [31,32]. Absorbing materials (in blue) are surrounding PCs to avoid unwanted scattering. As we are interested in TM polarization, the sample and source are assembled inside a parallel-plate microwave field mapper system where the time-harmonic ${E}_{z}$ field distributions are measured by a probing antenna connected to a vector network analyzer (VNA, Agilent E5071C) through the top plate (not shown here). The distance between the two parallel plates is 8.5 mm, where an extra 0.5 mm air layer between alumina cylinders and top plate is allowed to guarantee a smooth movement of the top plate in order to automatically measure field distributions.

In experiments, we inject electromagnetic waves at different frequencies through the antenna-array and the corresponding *E _{z}* field distributions are obtained. We then integrate amplitudes of electric fields of the whole sample region including both topological PC

_{1}and trivial PC

_{2}. The frequencies are inside the common photonic band gap between PC

_{1}and PC

_{2}. If a WGM existing along the cavity circumference is excited, we shall expect a largely enhanced integrated field. Figure 3(c) depicts our experimental results that the fields are enhanced at discrete frequencies. For comparison, we performed numerical simulations by COMSOL Multiphysics of electric field distributions of WGMs in topological cavity, whose parameters used are the same as those in experiments. We calculate electric field intensity of the simulated topological cavity at different frequencies shown in Fig. 3(b). Only discrete modes with very high fidelity are obtained and the frequencies are consistent with the discrete WGMs obtained from eigenmode analyses (Fig. 1, by setting

*a*= 25 mm) and we acquire similar mode distributions. The corresponding mode indices are labeled. Due to the existence of air gap in our microwave field mapper system, the corresponding frequencies slightly shift from numerical simulations and the inevitable leakage of electromagnetic wave through the air gap lowers the mode fidelity of WGMs as in experiments. However, good consistence between the number of modes and corresponding mode indices can be found between experiments and simulations. We observe discrete WGMs existing on a topological PC cavity. The excited WGM distributions at the microwave frequencies 7.531 and 7.709 GHz are shown in the upper and lower panels of Fig. 3(d) respectively, corresponding to the upward and downward triangles in Fig. 3(c). We observe good confinement of the WGMs on the interface between PC

_{1}and PC

_{2}and the field distribution reflects the corresponding index

*m*of WGMs. Moreover, by replacing the chiral source with a monopole antenna with no preferred in-plane angular momentum, identical mode spectrum as in Fig. 3(b) is obtained. However, the WGM will wind around the topological PC cavity clockwise/anti-clockwise if pseudospin excitation is used. This leads to the directional coupler design as follows.

## 4. Coupler

Our topological PC cavities are constructed with low loss dielectrics and the permittivity required is possible at optical and infrared frequencies. Thus their physical properties demonstrated in our microwave experiments remain unchanged if the sample is scaled to micron size. In numerical simulations, we propose a topological optical coupler device using silicon dielectric cylinders with lattice constant *a* = 0.25 μm, relative permittivity *ε* = 12.5 and diameter *d* = 0.06 μm. The intra (inter) couple distance is *h _{1}* = 0.09 μm (

*h*= 0.07 μm) for topological PC

_{2}_{1}, while

*h*= 0.075 μm (

_{1}*h*= 0.1 μm) for trivial PC

_{2}_{2}. The configuration of coupler is shown in the left panel of Fig. 4(a). The topological PC

_{1}(orange regions) and trivial PC

_{2}(gray region) constitute a topological cavity and a waveguide where blue regions represent absorbing layers around PC structures. A topological interface state excited by a chiral source with pseudospin up state (indicated by a black circular arrow) at 561 THz propagates unidirectionally along the topological waveguide and couples evanescently to the WGM in topological cavity. Light can thus be trapped inside the topological PC cavity.

Moreover, WGMs in topological PC cavity can also directionally couple to external waveguide. The pseudospin up (down) mode will wind clockwise (anticlockwise) and propagate to left (right) of the external waveguide. Thus by selecting the proper excitation of different pseudospin states, a directional coupler is achieved. As shown in Fig. 4(b), at 539 THz, as pseudospin up source is used (indicated by the black circular arrow), WGM winding clockwise around the cavity can only couple to the rightward propagating mode in the external mode and its emission is to the right output of the waveguide. If we change the excited source to be pseudospin down with reversed in-plane angular momentum, anticlockwise winding along the cavity is expected and the emission is to the left output.

## 5. Conclusion

In conclusion, topological WGMs in 2D finite dielectric PC cavity are demonstrated in microwave experiments. Fruitful physics occurs when considering a finite geometry. The original continuous topological interface state dispersion on a semi-infinite structure becomes discrete and its physical origin is discussed. The number of modes is solely dependent to the circumference of the topological cavity. Different to traditional cavities, as the topological WGMs are backscattering immune, the detailed geometry of a topological PC cavity does not affect the number of WGMs. Using low loss dielectrics only, our topological PC cavity scheme shall be compatible to modern semiconductor fabrication process and a directional coupler working at optical frequencies is also proposed.

## Funding

National Natural Science Foundation of China (No. 11574226); Natural Science Foundation of Jiangsu Province (No. BK20170058); a Project Funded by the Priority Academic Program Development of Jiangsu Higher Education Institutions (PAPD).

## Acknowledgments

The authors would like to acknowledge Prof. Vincenzo Giannini of Imperial College London for his kind suggestions. We also thank Prof. Hua Jiang of Soochow University for fruitful discussions.

## References and links

**1. **L. Lu, J. D. Joannopoulos, and M. Soljačić, “Topological photonics,” Nat. Photonics **8**(11), 821–829 (2014). [CrossRef]

**2. **A. B. Khanikaev and G. Shvets, “Two-dimensional topological photonics,” Nat. Photonics **11**(12), 763–773 (2017). [CrossRef]

**3. **K. V. Klitzing, G. Dorda, and M. Pepper, “New method for high-accuracy determination of the fine-structure constant based on quantized Hall resistance,” Phys. Rev. Lett. **45**(6), 494–497 (1980). [CrossRef]

**4. **F. D. Haldane and S. Raghu, “Possible realization of directional optical waveguides in photonic crystals with broken time-reversal symmetry,” Phys. Rev. Lett. **100**(1), 013904 (2008). [CrossRef] [PubMed]

**5. **Z. Wang, Y. Chong, J. D. Joannopoulos, and M. Soljacić, “Observation of unidirectional backscattering-immune topological electromagnetic states,” Nature **461**(7265), 772–775 (2009). [CrossRef] [PubMed]

**6. **E. Yablonovitch, “Inhibited spontaneous emission in solid-state physics and electronics,” Phys. Rev. Lett. **58**(20), 2059–2062 (1987). [CrossRef] [PubMed]

**7. **S. John, “Strong localization of photons in certain disordered dielectric superlattices,” Phys. Rev. Lett. **58**(23), 2486–2489 (1987). [CrossRef] [PubMed]

**8. **K. Sakoda, *Optical properties of photonic crystals* (Springer Science & Business Media, 2004).

**9. **J. D. Joannopoulos, S. G. Johnson, J. N. Winn, and R. D. Meade, *Photonic crystals: molding the flow of light* (Princeton university press, 2011).

**10. **C. L. Kane and E. J. Mele, “Quantum spin Hall effect in graphene,” Phys. Rev. Lett. **95**(22), 226801 (2005). [CrossRef] [PubMed]

**11. **B. A. Bernevig, T. L. Hughes, and S.-C. Zhang, “Quantum spin Hall effect and topological phase transition in HgTe quantum wells,” Science **314**(5806), 1757–1761 (2006). [CrossRef] [PubMed]

**12. **A. B. Khanikaev, S. H. Mousavi, W.-K. Tse, M. Kargarian, A. H. MacDonald, and G. Shvets, “Photonic topological insulators,” Nat. Mater. **12**(3), 233–239 (2013). [CrossRef] [PubMed]

**13. **W.-J. Chen, S.-J. Jiang, X.-D. Chen, B. Zhu, L. Zhou, J.-W. Dong, and C. T. Chan, “Experimental realization of photonic topological insulator in a uniaxial metacrystal waveguide,” Nat. Commun. **5**(1), 5782 (2014). [CrossRef] [PubMed]

**14. **L.-H. Wu and X. Hu, “Scheme for achieving a topological photonic crystal by using dielectric material,” Phys. Rev. Lett. **114**(22), 223901 (2015). [CrossRef] [PubMed]

**15. **T. Ma and G. Shvets, “All-Si valley-Hall photonic topological insulator,” New J. Phys. **18**(2), 025012 (2016). [CrossRef]

**16. **C. He, X. Ni, H. Ge, X.-C. Sun, Y.-B. Chen, M.-H. Lu, X.-P. Liu, and Y.-F. Chen, “Acoustic topological insulator and robust one-way sound transport,” Nat. Phys. **12**(12), 1124–1129 (2016). [CrossRef]

**17. **Y. Yang, Y. F. Xu, T. Xu, H.-X. Wang, J.-H. Jiang, X. Hu, and Z. H. Hang, “Visualization of a Unidirectional Electromagnetic Waveguide Using Topological Photonic Crystals Made of Dielectric Materials,” Phys. Rev. Lett. **120**(21), 217401 (2018). [CrossRef] [PubMed]

**18. **S. Yves, R. Fleury, T. Berthelot, M. Fink, F. Lemoult, and G. Lerosey, “Crystalline metamaterials for topological properties at subwavelength scales,” Nat. Commun. **8**, 16023 (2017). [CrossRef] [PubMed]

**19. **S. Yang, Y. Wang, and H. Sun, “Advances and prospects for whispering gallery mode microcavities,” Adv. Opt. Materials **3**(9), 1136–1162 (2015). [CrossRef]

**20. **M. R. Foreman, J. D. Swaim, and F. Vollmer, “Whispering gallery mode sensors,” Adv. Opt. Photonics **7**(2), 168–240 (2015). [CrossRef] [PubMed]

**21. **R. Chen, V. D. Ta, and H. Sun, “Bending-induced bidirectional tuning of whispering gallery mode lasing from flexible polymer fibers,” ACS Photonics **1**(1), 11–16 (2014). [CrossRef]

**22. **F. Vollmer and S. Arnold, “Whispering-gallery-mode biosensing: label-free detection down to single molecules,” Nat. Methods **5**(7), 591–596 (2008). [CrossRef] [PubMed]

**23. **S. Blair and Y. Chen, “Resonant-enhanced evanescent-wave fluorescence biosensing with cylindrical optical cavities,” Appl. Opt. **40**(4), 570–582 (2001). [CrossRef] [PubMed]

**24. **C.-Y. Chao and L. J. Guo, “Biochemical sensors based on polymer microrings with sharp asymmetrical resonance,” Appl. Phys. Lett. **83**(8), 1527–1529 (2003). [CrossRef]

**25. **I. M. White, H. Oveys, and X. Fan, “Liquid-core optical ring-resonator sensors,” Opt. Lett. **31**(9), 1319–1321 (2006). [CrossRef] [PubMed]

**26. **J. D. Suter, W. Lee, D. J. Howard, E. Hoppmann, I. M. White, and X. Fan, “Demonstration of the coupling of optofluidic ring resonator lasers with liquid waveguides,” Opt. Lett. **35**(17), 2997–2999 (2010). [CrossRef] [PubMed]

**27. **B. Bahari, A. Ndao, F. Vallini, A. El Amili, Y. Fainman, and B. Kanté, “Nonreciprocal lasing in topological cavities of arbitrary geometries,” Science **358**(6363), 636–640 (2017). [CrossRef] [PubMed]

**28. **M. A. Bandres, S. Wittek, G. Harari, M. Parto, J. Ren, M. Segev, D. N. Christodoulides, and M. Khajavikhan, “Topological insulator laser: Experiments,” Science **359**(6381), 4005 (2018).

**29. **Q. Wang, M. Xiao, H. Liu, S. Zhu, and C. T. Chan, “Optical Interface States Protected by Synthetic Weyl Points,” Phys. Rev. X **7**(3), 031032 (2017). [CrossRef]

**30. **G. Siroki, P. A. Huidobro, and V. Giannini, “Topological photonics: From crystals to particles,” Phys. Rev. B **96**(4), 041408 (2017). [CrossRef]

**31. **Y. Yang, H. Jiang, and Z. H. Hang, “Topological Valley Transport in Two-dimensional Honeycomb Photonic Crystals,” Sci. Rep. **8**(1), 1588 (2018). [CrossRef] [PubMed]

**32. **L. Ye, Y. Yang, Z. Hong Hang, C. Qiu, and Z. Liu, “Observation of valley-selective microwave transport in photonic crystals,” Appl. Phys. Lett. **111**(25), 251107 (2017). [CrossRef]