Semiconductor lasers have the advantages of small size, high efficiency, long service life, wide wavelength range, easy integration and modulation, and are widely used in communication, processing, medical and military fields. Among them, single-mode devices have become the first choice for many applications because of their ideal linewidth and beam quality, and the key to single-mode operation is mode selection, which depends on photonic crystal structure (figure 1). For example, the light source of the whole optical fiber interconnection network is a distributed feedback laser (DFB, figure 1, upper left). Early DFB lasers used one-dimensional periodic grating structure to select modes, but the single-mode output was not stable because of the competition between two sideband modes. The textbook solution is to introduce a defect (quarter-wavelength phase shift, figure 1 upper right) and then generate a defect mode in the middle of the photonic band gap, thus ensuring stable single-mode operation. In addition, the resonator of vertical cavity surface emitting lasers (VCSELs), which are widely used in short-distance communication, optical mouse, laser printer and face recognition, also uses the interband defect state to select the mode. However, because the above two mainstream products use one-dimensional photonic crystals to select modes, in the case of no periodic structure, the dimensions in the other two directions cannot exceed the wavelength order, because there is no mode selection mechanism, otherwise it is multimode lasing. If the device size does not go up, single-mode power will encounter bottlenecks. A natural scheme to improve single-mode power is to adopt two-dimensional photonic crystal structure. The product of two-dimensional photonic crystal surface-emitting laser (PCSEL, figure 1, bottom left) was successfully launched by Hamamatsu Company of Japan in 20 17, which has many advantages such as large-area single-mode output, high power and narrow divergence angle, but PCSEL also has at least two high angles. Therefore, if we can design a robust two-dimensional interband defect mode like one-dimensional mainstream products DFB and VCSEL, it may become the mainstream direction of high-power single-mode lasers in the future.
The research team of the Institute of Physics designed an optical resonator with two-dimensional band gap defect mode by using topological principle. The team first realized that the one-dimensional defect state in DFB and VCSEL is actually topological, which is equivalent to many well-known one-dimensional topological models, including Shockely, Jackiw-Rebbi and SSH modes. Especially the one-dimensional Jackiw-Rebbi mode in high-energy physics has a direct two-dimensional correspondence, namely Jackiw-Rossi mode, which is the mass vortex solution of Dirac equation and can be realized in principle by generalized Kekule modulation in the honeycomb lattice of condensed matter system (HCM model). The team designed this topological optical cavity by vortex modulating Dirac photonic crystal, and realized this kind of Dirac vortex cavity in the silicon chip (SOI) and optical communication band (1550nm) (figure 1, bottom right). The cavity can realize the excellent characteristics of single mode between bands, arbitrary multi-degenerate modes, maximum free spectral range, small far-field divergence angle, vector light field output, adjustable mode area from micron to millimeter, and compatibility with various substrates.
The best large-area single mode is the most unique advantage of Dirac vortex cavity which is different from other known optical cavities. Large area single mode is beneficial to improve the power and stability of single mode laser. The market demand for electricity is always increasing, and the existing products have reached the bottleneck in single-mode energy output, which requires new ideas. Moreover, high power and single mode itself are a pair of contradictions, because high power requires a large area of optical cavity, and the number of modes will inevitably increase with the size of optical cavity, making it more difficult to maintain single mode operation stably. Now the appearance of Dirac vortex cavity is a potential new technical route. The singlet state of optical cavity can be characterized by the free spectral range: FSR. As we all know, the mode spacing (FSR) of all optical cavities is inversely proportional to the mode volume (V-1), so the way to increase the FSR is to reduce the cavity volume. However, the FSR of Dirac optical cavity is inversely proportional to the root number of the mode system (V-1/2, figure 1 lower right), so the FSR is much higher than that of ordinary optical cavity (one or two orders of magnitude larger) under the same mode volume. The reason for this difference is that the photon state density in ordinary optical cavity is non-zero constant and the modes are arranged at equal intervals; However, the density of photon states at Dirac point frequency is equal to zero, and the mode spacing (FSR) on both sides is maximized (Figure 2, left).
The degeneracy of arbitrary modes is another unique feature of Dirac vortex cavity. Because the topological invariant of the system is the winding number (w) of the vortex, the modulus in the topological center cavity is equal to w, which can be any positive integer or negative integer, and all w topological modes are close to frequency degeneracy. The experimental spectra of w=+ 1, +2, +3 are shown on the right of fig. 2. Highly degenerate optical cavity can reduce the spatial coherence of multimode laser and can be used in laser illumination technology.
The correspondent of the thesis is Ling Lv, a researcher at the Institute of Physics. The first author is Gao Xiaomei, a doctoral student jointly trained by Nankai University and the Institute of Physics, and Yang, a doctoral student at the Institute of Physics. Other authors include Lin Hao, Ph.D. student of the Institute of Physics, Zhang Lang, an undergraduate student of Nankai University (now a Ph.D. student of Yale University), Wang Zhong, a researcher at the Institute of Advanced Studies of Tsinghua University, Li Jiafang, an associate researcher at the Institute of Physics of Beijing Institute of Technology, and Fang Bo, a professor at the School of Physical Sciences of Nankai University. The sample preparation of topological microcavity was completed in the micromachining laboratory of Institute of Physics, Chinese Academy of Sciences, and Li Guangrui, a postdoctoral fellow of Institute of Physics, participated in the later discussion of the work. This work has won the national key research and development plan (20 17yfa0303800, 20 16yfa0302400), the national natural science foundation (1 172 1404) and the pilot project of Chinese Academy of Sciences (XDB).