Figure 1 - available from: Scientific Reports
This content is subject to copyright. Terms and conditions apply.
Design and simulation results of the wavelength demultiplexer. (a) Initial silicon slab before design optimization. (b) Final optimized structure. (c,d) are the simulated optical field distributions. (e) is the transmittance spectrum of each output port.
Source publication
In this paper, we use the inverse design method to design an optical interconnection system composed of wavelength demultiplexer and the same direction waveguide crossing on silicon-on-insulator (SOI) platform. A 2.4 μm × 3.6 μm wavelength demultiplexer with an input wavelength of 1.3–1.6 μm is designed. When the target wavelength of the device is...
Contexts in source publication
Context 1
... interconnection component. Wavelength demultiplexer. Based on the "method", We first use DBS algorithm to design a wavelength demultiplexer on the standard commercial SOI substrate. The thickness of the silicon core layer is 220 nm, and the under-cladding silicon dioxide is 2 μm. As shown in Fig. 1a, the photonic-like crystal material structure 34,35 is used for design, and its footprint is 3.6 μm × 2.4 μm. The advantage of the photonic-like crystal material structure is that regardless of the arrangement of the circular holes optimized by the final algorithm, the final circular holes are independent of each other and have the ...
Context 2
... is complete, which is an iteration. For each iteration, the algorithm stops until FOM converges. We utilize FDTD solutions to solve the problem. In the wavelength range of 1300-1600 n, the input and output modes are both TE 0 mode. For the wavelength demultiplexer, the final optimization result is obtained after 4 iterations, as shown in Fig. 1b. The target wavelength of output port 1 is 1.4 μm, and the target wavelength of output port 2 is 1.6 μm. The performance test of the final structure is performed, and the simulated optical field profile of the device shown in Fig. 1c,d is obtained. It can be seen that the target wavelength is well output from the target output port. ...
Context 3
... mode. For the wavelength demultiplexer, the final optimization result is obtained after 4 iterations, as shown in Fig. 1b. The target wavelength of output port 1 is 1.4 μm, and the target wavelength of output port 2 is 1.6 μm. The performance test of the final structure is performed, and the simulated optical field profile of the device shown in Fig. 1c,d is obtained. It can be seen that the target wavelength is well output from the target output port. Here, the insertion loss (IL) is defined ...
Context 4
... t denotes the transmission efficiency of the target wavelength of the non-target output port, and T is the transmittance of the input port. Figure 1e shows the transmittance spectrum of the last two output ports. Finally, the insertion loss of output port 1 is − 0.93 dB, and the crosstalk is − 18.4 dB. ...
Similar publications
Silicon photonic micro-ring resonators (MRR) developed on the silicon-on-insulator (SOI) platform, owing to their high sensitivity and small footprint, show great potential for many chemical and biological sensing applications such as label-free detection in environmental monitoring, biomedical engineering, and food analysis. In this tutorial, we p...
Citations
... specified footprints or structures) without manually tedious parameter adjustment works, saving massive resources. There are many reports on developments or improvements of nano-photonic devices using inverse design with various optimization algorithms [47][48][49][50][51][52][53][54][55][56][57][58][59][60], but only few studies about the applications of inverse design on nonreciprocal devices [61][62][63][64], which are normally based on magneto-optical effects. In this study, we introduce several magnetic-free nonreciprocal devices that avoid use of magneto-optical materials and magnets, which are difficult to be compatible with current semiconductor integration technology. ...
Inserting nonreciprocal devices at the doorways of Alice and Bob is a widely recognized countermeasure against quantum hacking attacks in quantum key distribution (QKD) systems. However, traditional integrated nonreciprocal devices, which are typically based on magneto-optical effects, face challenges in compatibility with current semiconductor integration technology. As a result, earlier chip-based QKD systems were unable to integrate nonreciprocal components and were vulnerable to injecting-type attacks. Based on the actual parameters of SOI integration, we employed the inverse design with the direct binary search algorithm to construct several magnetic-free nonreciprocal devices, facilitating their integration into chip-based QKD systems while meeting various chip configuration design requirements. The designed devices have sizes of only a few square micrometers, yet the quasi-isolator can achieve an isolation level exceeding 27 dB. To demonstrate their practical utility in QKD, we employed the designed devices to safeguard the QKD system against Trojan-horse attacks. The simulation results demonstrate that our proposed devices effectively secure the BB84 and measure-device-independent QKD systems against Trojan-horse attacks.
... Numerous Si on-chip optical components have been designed to steer, split, and demultiplex light waves [6][7][8][9][10][11][12][13][14][15][16][17][18][19][20][21][22]. The standard, forward, method to design the optical components utilizes Maxwell's equations, then hand-tunes some parameters, then simulates the structures to determine their characteristics; this design requires many time-consuming simulations and usually results in big devices. ...
... The inverse design of the direct binary search method can optimize binary structures but requires high computational time and resources [9,47]. Bayesian and rotatable direct-binary-search algorithms reduce the optimization time [41,48]. ...
... Another was a 220-nm silicon-on-insulator platform with a footprint of 5.76 μm 2 ; at 1.4 ≤ ≤ 1.6 μm, the simulation results showed < 0.5 dB and < −19 dB. The inverse design process used a direct-binary-search algorithm and took about 36 h on an eightcore PC [9]. In 2020, another used a 340-nm-thick silicon-on-insulator waveguide crossing with a footprint of 16 μm 2 ; at 1.5 ≤ ≤ 1.6 μm, it had < 0.07 dB and < −25 dB [62]. ...
This paper describes the inverse design of the particle swarm optimization algorithm combined with the three-dimensional finite-difference time-domain simulation to design a waveguide crossing that resembles a binary code. The device consists of 15 × 15 air holes in a 220-nm silicon slab on a 3-μm-thick SiO2 substrate with one input port and three output waveguide ports. The designed device has a small footprint of 4μm² and a short simulation time of 1.7 h. In the wavelength range of 1.5–1.6μm, the device has insertion loss IL<0.85 dB and crosstalk XT<−14.5 dB. This device tolerates air hole-position disordering of ηp=23%, and hole-radius disordering of ηr=35%, so it is appropriate to use in the complementary metal–oxide–semiconductor fabrication process. The paper also proposes integrated interconnections that contain 2 × 2 waveguide crossings and have IL<2.9 dB and XT<−13 dB, and 3 × 3 waveguide crossings that have IL<5 and XT<−12.5 dB.
We proposed and demonstrated integrated photonic networks composed of inverse-designed components for spectral reconstruction. We reconstructed spectral lines separated by 100 pm and continuous spectra from a device with an ultracompact footprint of 88×300 μm ² .
A 90-degree bending silicon-on-insulator waveguide is constructed using a self-developed inverse design optimization algorithm. The measured insertion loss is about 1.7 dB at 1550 nm with a footprint of only 1.5µm×1.5µm.
Inverse design has been widely studied as an efficient method to reduce footprint and improve performance for integrated silicon photonic (SiP) devices. In this study, we have used inverse design to develop a series of ultra-compact dual-band wavelength demultiplexing power splitters (WDPSs) that can simultaneously perform both wavelength demultiplexing and 1:1 optical power splitting. These WDPSs could facilitate the potential coexistence of dual-band passive optical networks (PONs). The design is performed on a standard silicon-on-insulator (SOI) platform using, what we believe to be, a novel two-step direct binary search (TS-DBS) method and the impact of different hyperparameters related to the physical structure and the optimization algorithm is analyzed in detail. Our inverse-designed WDPS with a minimum feature size of 130 nm achieves a 12.77-times reduction in footprint and a slight increase in performance compared with the forward-designed WDPS. We utilize the optimal combination of hyperparameters to design another WDPS with a minimum feature size reduced to 65 nm, which achieves ultra-low insertion losses of 0.36 dB and 0.37 dB and crosstalk values of -19.91 dB and -17.02 dB at wavelength channels of 1310 nm and 1550 nm, respectively. To the best of our knowledge, the hyperparameters of optimization-based inverse design are systematically discussed for the first time. Our work demonstrates that appropriate setting of hyperparameters greatly improves device performance, throwing light on the manipulation of hyperparameters for future inverse design.
In this paper, numerical calculations of the Berry curvature and Chern number of two types of two-dimensional photonic crystals consisting isotropic dielectric and anisotropic magneto-optical, gyromagnetic, rods in air in a square lattice are studied. The Chern number, an integer number, is a key parameter to distinguish between trivial and non-trivial photonic crystals. Trivial and non-trivial photonic crystals reveal zero and non-zero Chern numbers. A non-zero Chern number is achieved through the breaking of time-reversal and inversion symmetries. The results for two-dimensional photonic crystals containing isotropic dielectric and gyromagnetic materials under TM mode illustrate zero and 0, 1, -2, and -1 Chern numbers for the first four bands, respectively. The creation of non-zero Chern numbers brings a new way of designing one-way, robust to arbitrary disorder, and zero back-reflection photonic components
