Open Access
Add to BibSonomy
Abstract
An ultra-compact on-chip spectrometer was demonstrated based on an array of add-drop micro-donut resonators (MDRs). The filter array was thermally tuned by a single TiN microheater, enabling simultaneous spectral scanning across all physical channels. The MDR was designed to achieve large free spectral ranges with multimode waveguide bends and asymmetric coupling waveguides, covering a spectral range of 40 nm at the telecom waveband with five physical channels (which could be further expanded). Benefiting from the ultra-small device footprint of 150 µm2, the spectrometer achieved a low power consumption of 16 mW. Additionally, it is CMOS-compatible and enables mass fabrication, which may have potential applications in personal terminals and the consumer industry.
© 2023 Optica Publishing Group under the terms of the Optica Open Access Publishing Agreement
1. Introduction
Photonic integration facilitates the development of miniaturized spectrometers that find applications in portable sensing, lab-on-a-chip technology, aerospace imaging, and underwater scientific research [1–7]. In comparison to conventional benchtop spectrometers, the footprint of these devices is drastically reduced, albeit resulting in compromises concerning resolution, spectral range, or signal-to-noise ratio to some extent [1]. As a result, on-chip spectrometers are typically engineered to meet specific performance requirements. For rapid indicative measurements in situ, key considerations include ultra-low power consumption, ultra-compact size, and a wide spectral range.
Integrated spectrometers could be implemented using dispersive optics (or narrowband filters) to split (or select) spectral channels [8–11], Fourier transform interferometers to exploit Fellgett’s advantage [12–14], or computational algorithms to decode the incident signal via spectral-to-spatial mapping [15–19]. Among these methods, Fourier transform spectrometers offer a high signal-to-noise ratio but suffer from drawbacks such as high power consumption and a large physical footprint. Reconstructive spectrometers possess wide bandwidth and high-resolution characteristics but may pose certain challenges when analyzing complex spectra. Narrowband add-drop filter-based spectrometers, commonly employing ring resonators, provide an effective solution for general-purpose spectral analysis with high resolution and CMOS compatibility. However, the spectral bandwidth of the narrowband filter-based spectrometer is limited by the free spectral range (FSR) of the filters.
To extend the spectral range of ring resonator-based spectrometers, approaches such as employing arrayed waveguide gratings (AWGs) [18,20,21] or diffraction gratings [11] in tandem with micro-rings were employed. However, they sacrificed the overall size of the spectrometer. In an effort to address this issue, a spectrometer based on FSR-free filters with a double-side-coupled grating-assisted Fabry-Perot cavity was designed [22–24]. Unfortunately, this design exhibits high power consumption, exceeding 600 mW, and a relatively large device length, which limits its application in scenarios where power supply and equipment size are strictly constrained, such as mobile terminals and outdoor spectrometry. Thankfully, ring resonators could achieve an ultra-large FSR by designing a multimode waveguide bend with an extremely small radius [25,26]. Using these micro-donut structures, a passive spectrometer can cover an operating bandwidth of approximately 50 nm with 84 physical channels [27]. Nevertheless, it is important to note that actively tuning [28] the filter array is a promising approach to significantly reducing the number of physical channels and required detectors, thus mitigating large-scale fabrication imperfections. Note that the reduced size of micro-donut resonators (MDRs) is expected to significantly decrease the power consumption involved in dynamic tuning.
In this study, we present the design and fabrication of a highly compact spectrometer with low power consumption, achieved through implementing a large-FSR MDR array. The spectrometer was fabricated on the silicon-on-insulator (SOI) platform using the CMOS foundry, ensuring good scalability and compatibility. All the resonators were simultaneously heated by a single microheater to cover a wide spectral range of 40 nm (at a central wavelength of 1542 nm) with only 5 physical channels, enabling a significantly simplified fabrication and operation process.
2. Device design
Figure 1(a) shows a schematic diagram of the spectrometer based on large-FSR add-drop MDRs. The device was designed on the SOI platform with a 220 nm-thick device layer and a 2 µm-thick box layer. The light was coupled into and out of the device through a grating coupler, as shown in the inset within the blue dashed box, which was connected to each port. The functional section of the spectrometer consisted of a series of MDRs with varying radii, as indicated in the inset within the green dashed box. The narrower bent coupling waveguides were connected to 500 nm-wide single-mode strip waveguides via tapered waveguides. A TiN microheater, attached to two AlCu electrodes, was positioned above the resonator array. To prevent optical absorption loss introduced by the metal, a layer of SiO2 was inserted between the Si waveguides and the microheater.
Fig. 1. Schematic diagram of the (a) device structure and (b) operating principle of the spectrometer based on the tunable add-drop MDR array.
The operational principle of the spectrometer is depicted in Fig. 1(b). When broadband input light is introduced, optical signals with specific wavelengths that match the resonance wavelengths of the five MDRs are sequentially coupled into each resonator and subsequently output through the drop port (corresponding to channel Ch1∼Ch5). At the same time, light at other wavelengths continues to propagate through the device and exits via the through port. By applying different voltages to the electrode pads, the resonator array is heated by the TiN microheater, causing the resonance wavelengths within each channel to simultaneously redshift and sweep across a certain wavelength range. Hence, the input spectra within the spectral response range of the spectrometer are sampled by the tunable MDR array and subsequently detected. The original spectra can then be reconstructed from the detected signals, taking into account the predetermined transmission characteristics of the filter array.
In order to accommodate more channels in an ultra-compact spectrometer, large-FSR and small-footprint MDRs with small radii and asymmetrical bent directional couplers were utilized, as shown in Fig. 2(a). To achieve a larger FSR, the group index ng of the ring waveguide and circumference L of the resonator needed to be reduced, since FSR was determined by λ2/(ngL). This was confirmed through simulation, as demonstrated in Fig. 2(c) and 2(d), where a smaller radius Rr lead to smaller L and larger ng. Meanwhile, decreasing the waveguide width of a resonator (wr) also resulted in a larger FSR. The influence of wr became more pronounced as Rr decreased. It could be observed that for Rr < 1.8 µm, an FSR larger than 50 nm could be achieved for various wr values. However, despite the use of a wide multimode waveguide to mitigate bending loss, excessively small Rr and wr could still result in a substantial optical loss for the fundamental TE0 mode in the MDR (see the shifted mode shape in the inset of Fig. 2 (b)). Bending losses became prohibitive for resonators with a radius smaller than 0.8 µm, as shown in Fig. 2(b). Therefore, a minimum radius of Rr_min = 0.8 µm and a minimum waveguide width of wr_min = 0.8 µm were selected for the donut array to ensure that the angular loss remained below 0.05 dB/rad. Considering the larger FSR and reduced bending loss, wr was set to be 0.8 µm. The radii of donuts ranging from 0.8 µm to 1.6 µm were chosen for the filter array to achieve the same resonance wavelength spacing. Figure 3(d) shows the simulated resonance peak position of the donut array. Specifically, Rr = 0.8 µm, 1 µm, 1.2 µm, 1.4 µm, and 1.6 µm were assigned to channels 5 to 1 (Ch5 to Ch1), respectively. Well linearity was observed as the resonance peaks shifted in correlation with the resonator’s radius. Specifically, each 0.2 µm increase in Rr resulted in an approximate resonant wavelength spacing of ∼8 nm.
Fig. 2. (a) Schematic diagram of a miniaturized MDR with bent asymmetrical directional couplers. The simulated (b) angular loss, (c) group index, and (d) FSR of the resonator with different waveguide widths vary with the radius Rr.
Fig. 3. Design of the donut array. The MDR for channel 1: (a) The change of coupling ratio of TE1 mode with the coupling angle and the coupling waveguide width. The (b) calculated transmission spectrum and (c) mode field of the designed resonator (λ = 1563 nm). (d) Design of the resonance peak position of the donut array.
After the parameters of micro-donut waveguides were determined, the design of the coupling waveguide was then conducted. In directional couplers, effective refractive index (neff) multiplied by the bending radius is equal or approximately equal for both the coupling waveguide and the micro-donut waveguide. Specifically, here it can be expressed as neff(coupling waveguide)·Rc = neff(donut)·Rr, where Rc took the value of Rr + wr /2 + wgap + wc /2 for the coupling waveguide. Considering the patterning resolution of photolithography and ease of SiO2 filling during fabrication, a fixed value of 200 nm was chosen for wgap. Various coupling waveguide widths around 0.3 µm can achieve an approximate phase match. However, when selecting a value for wc, additional considerations need to be taken into account. Since the MDR waveguide is wider than a single-mode waveguide, it also supports higher-order modes, primarily the TE1 mode, which can result in increased loss and crosstalk. Thus, a properly designed bent coupling waveguide is crucial to minimize the coupling ratio of higher-order modes, and the coupling angle (θ) plays a critical role. Here, the design of the MDR for channel 1 (Ch1) was presented, for instance. As shown in Fig. 3(b), the coupling ratio of the TE1 mode varies with the coupling angle (θ) and the width of the coupling waveguide. It could be observed that the minimum coupling ratio of the TE1 mode at a specific θ decreased for a wider coupling waveguide. Therefore, considering the balance between phase matching and higher-mode suppression, a coupling angle of θ = 30° and a coupling waveguide width of wc = 350 nm were chosen. Figure 3(c) displays the simulated transmittance of the MDR for channel 1, and Fig. 3(d) showcases the simulated optical field distribution of the fundamental TE mode in the MDR on resonance. FSR for channel 1 was simulated to be 55 nm, which represents the smallest FSR among channels in the array. The contrast between the resonances of the fundamental mode and the higher-order mode reaches 20 dB, indicating good suppression of higher-order modes.
3. Results and discussions
Figure 4(a) shows an optical microscope image of the fabricated spectrometer. The device was fabricated in a multi-project-wafer (MPW) run offered by the Institute of Microelectronics of the Chinese Academy of Science (IMECAS). The grating couplers and strip waveguides were patterned using deep-ultraviolet (DUV) photolithography, followed by inductively coupled plasma (ICP) etching of the silicon device layer of the SOI substrate. The etching depths of the waveguides and the gratings are 220 nm and 70 nm, respectively. A SiO2 spacer was then grown through plasma-enhanced chemical vapor deposition (PECVD) and planarized to 1 µm. Next, the 50 nm-thick TiN microheater and 700 nm-thick AlCu electrodes were deposited and subsequently capped by another layer of PECVD SiO2. Finally, windows were opened to expose the electrode pads.
Fig. 4. (a) An optical microscope image of the fabricated tunable filter array. The transmission spectra of (b) Ch1 and (c) Ch2 showing the FSR of the two MDRs. (d) The static transmission spectra and (e) the static resonance wavelengths of the fabricated tunable filter array.
To validate the design of the large-FSR MDR, the transmission spectra of the resonators were measured. Figures 4(b) and 4(c) show the transmission spectra of Ch1 and Ch2, respectively. The spectra were normalized by subtracting the transmission spectra obtained from two reference grating couplers connected by a 500 nm-wide straight strip waveguide. For Ch1, the measured FSR was found to be 54 nm, and the peak intensity of the fundamental TE mode was observed to be 18 dB higher than the higher-order mode (i.e., inter-modal crosstalk <-18 dB), which agrees with the design. Channel 2 exhibited an even larger FSR of 60 nm with a resonator that is 0.2-µm smaller in radius, and an inter-channel crosstalk of -15 dB caused by the coupling of the light with longer wavelengths into the neighboring channel. Although measurements of the FSRs for the other channels were unavailable due to the limited wavelength range of the tunable laser used in this study, FDTD simulations indicated that the largest FSR among all the five channels (Ch5) was expected to exceed 80 nm.
Figure 4(d) exhibits the normalized static transmission spectra from the five drop channels and the through port when no voltage is applied. The 3-dB bandwidths (BWs) of the five resonance peaks from the drop channels (Ch1 to Ch5) were measured as 0.32 nm, 0.47 nm, 0.67 nm, 0.735 nm, and 1.13 nm, respectively. For a spectrometer based on narrowband filters with low crosstalk, the resolution of the spectrometer is expected to depend on the 3-dB bandwidths of MDRs only [21]. The insertion losses (ILs) ranged from 0.74 dB to 3.86 dB. The inter-channel crosstalk was lower than -10 dB and could be further lowered by placing the MDR with a longer resonant wavelength closer to the input port. Figure 4(e) shows the measured resonance peak wavelength of each channel. It presents good linearity as the resonance peaks shift with the radius of the MDR, and the fitted slope agrees well with the simulated results shown in Fig. 3(d). A minor blueshift of approximately 9 nm was found in the starting wavelength of the spectral range compared to simulations due to fabrication errors in the filter array.
When electrical power was applied through electrodes, resonance peaks of the five channels redshifted simultaneously due to the thermo-optic effect of silicon. Figure 5 shows the transmission spectra of the spectrometer for various applied powers from 0 to 16 mW. The regular fluctuation of the peak intensity in each channel during thermo-optic tuning was attributed to reflection-induced interference in the device. However, this fluctuation does not affect spectral retrieval, as it can be eliminated through calibration prior to measurements. Remarkably, even with low applied powers, the resonance peaks were able to scan across a wavelength range of ∼40 nm.
Figure 6(a) shows the measured resonance wavelengths and their linear fitting of all five channels when various heating powers were applied. A low total power consumption of 16 mW was measured for the spectrometer. A well-linear relationship was exhibited between the resonant wavelength shift and heating power. The slopes of the fitting lines in Fig. 6(a) represent the energy efficiency or thermo-optic efficiency of the resonator array, as shown in Fig. 6(b). The energy efficiencies were determined to be 0.51, 0.68, 0.69, 0.64, and 0.49 nm/mW for Ch1 to Ch5, respectively. The energy efficiency of Ch1 and Ch5 was relatively lower due to the heat exchange between the microheater and the connected metallic electrodes. This could be further optimized by designing curved heater boundaries. The filter array possessed a low power consumption and a high energy efficiency, benefiting from the small heating area of the device due to the small footprint of MDRs.
Fig. 6. (a) Linear fitting of the resonance wavelengths versus the electrical power applied on the spectrometer. The (b) energy efficiency, (c) insertion loss, and 3-dB bandwidth of each channel of the spectrometer.
Figure 6(c) shows the insertion loss and 3-dB bandwidth of the spectrometer when heating power is applied. The error bars represent the variation and fluctuation in each channel as the applied power changes. The IL variation within the same channel was found to be below 2.9 dB, and the minimum and maximum insertion losses measured across all peaks in the five channels were 0.74 dB and 5.52 dB, respectively. The difference in coupling efficiencies between the grating couplers in the spectrometer and those in the reference structure used for transmittance normalization, as well as the reflectance resulting from the geometric change of waveguide may contribute to IL variation. Despite this, the fluctuation of insertion loss has no impact on spectral reconstruction due to calibration before measurements.
Among previously reported on-chip spectrometers for general-purpose spectra reconstruction based on tunable narrow-bandwidth filters, our spectrometer exhibited the smallest device footprint and highest energy efficiency, to the best of our knowledge, as shown in Table 1. While a passive spectrometer demonstrated in Ref. [27] operates without power consumption, it necessitates the use of delicate electron beam lithography to create variation steps of 1 nm among radii of tens of ring resonators. This greatly increases the complexity of large-scale fabrication. In contrast, our spectrometer claims an extremely low total power consumption of 16 mW for 5 physical channels and 70 spectral channels. The total number of spectral channels in our spectrometer is determined by summing the spectral channels within each physical channel (i.e., dividing the channel spectral range by the 3-dB bandwidth of the MDR in each channel). The power consumption per spectral channel is significantly low among reported spectrometers with a large dynamic range.
Table 1. Comparison of integrated spectrometers based on wavelength-selective filters.
Reconstructive spectrometers based on ring resonators [31] (with a central wavelength of 809 nm) were also demonstrated recently, and a similar scheme could also be realized based on nanobeam cavities [32]. Reconstructive spectrometers offer an effective way to improve the resolution, but the reconstruction error of algorithms can vary among spectra with different complexity [33]; hence, the deficient robustness makes them more suitable for particular spectra reconstruction. Our design needs no subsequent computational reconstruction, and only a one-time calibration is performed on each spectrometer chip before use. The resolution of this spectrometer could also be further improved by reconstruction algorithms.
4. Conclusion
We proposed and demonstrated an integrated spectrometer that utilizes a thermo-optically tuned MDR array. This spectrometer covers a large spectral range of 40 nm at a central wavelength of 1542 nm with 5 physical channels. Benefiting from the small radii of the resonators, the spectrometer possesses an ultra-compact footprint of 150 µm2 and a low total power consumption of 16 mW, making it one of the most energy-efficient spectrometers with dynamic tuning capabilities. The minimum FSR of resonators in the array was measured to be as large as 54 nm, manifesting that both physical and spectral channels could be further extended to broader wavelength regions. The spectrometer is compatible with mass production using CMOS technology on a silicon photonic platform and is potentially applicable for low-cost and portable spectral recognition.
Funding
National Natural Science Foundation of China (62105287, 62175202, 62205274,); National Key Research and Development Program of China (2021YFB2801300).
Acknowledgments
The authors thank ZJU Micro-Nano Fabrication Center at Zhejiang University for the facility support.
Disclosures
The authors declare no conflicts of interest.
Data availability
Data underlying the results presented in this paper are not publicly available at this time but may be obtained from the authors upon reasonable request.
References
1. Z. Yang, T. Albrow-Owen, W. Cai, and T. Hasan, “Miniaturization of optical spectrometers,” Science 371(6528), eabe0722 (2021). [CrossRef]
2. K. Gajos, M. Angelopoulou, P. Petrou, K. Awsiuk, S. Kakabakos, W. Haasnoot, A. Bernasik, J. Rysz, M. M. Marzec, K. Misiakos, I. Raptis, and A. Budkowski, “Imaging and chemical surface analysis of biomolecular functionalization of monolithically integrated on silicon Mach-Zehnder interferometric immunosensors,” Appl. Surf. Sci. 385, 529–542 (2016). [CrossRef]
3. N. Blind, E. Le Coarer, P. Kern, and S. Gousset, “Spectrographs for astrophotonics,” Opt. Express 25(22), 27341–27369 (2017). [CrossRef]
4. A. Nascetti, M. Mirasoli, E. Marchegiani, M. Zangheri, F. Costantini, A. Porchetta, L. Iannascoli, N. Lovecchio, D. Caputo, G. de Cesare, S. Pirrotta, and A. Roda, “Integrated chemiluminescence-based lab-on-chip for detection of life markers in extraterrestrial environments,” Biosens. Bioelectron. 123, 195–203 (2019). [CrossRef]
5. J. Zhou, Z. Zhang, B. Dong, Z. Ren, W. Liu, and C. Lee, “Midinfrared Spectroscopic Analysis of Aqueous Mixtures Using Artificial-Intelligence-Enhanced Metamaterial Waveguide Sensing Platform,” ACS Nano 17(1), 711–724 (2023). [CrossRef]
6. Y. Ma, B. Dong, and C. Lee, “Progress of infrared guided-wave nanophotonic sensors and devices,” Nano Convergence 7(1), 12 (2020). [CrossRef]
7. X. Liu, W. Liu, Z. Ren, Y. Ma, B. Dong, G. Zhou, and C. Lee, “Progress of optomechanical micro/nano sensors: a review,” International Journal of Optomechatronics 15(1), 120–159 (2021). [CrossRef]
8. G. Calafiore, A. Koshelev, S. Dhuey, A. Goltsov, P. Sasorov, S. Babin, V. Yankov, S. Cabrini, and C. Peroz, “Holographic planar lightwave circuit for on-chip spectroscopy,” Light: Sci. Appl. 3(9), e203 (2014). [CrossRef]
9. Z. Zhang, Y. Liu, Z. Wang, Y. Zhang, X. Guo, S. Xiao, K. Xu, and Q. Song, “Folded Digital Meta-Lenses for on-Chip Spectrometer,” Nano Lett. 23(8), 3459–3466 (2023). [CrossRef]
10. M. Faraji-Dana, E. Arbabi, A. Arbabi, S. M. Kamali, H. Kwon, and A. Faraon, “Compact folded metasurface spectrometer,” Nat. Commun. 9(1), 4196 (2018). [CrossRef]
11. B. B. C. Kyotoku, L. Chen, and M. Lipson, “Sub-nm resolution cavity enhanced micro-spectrometer,” Opt. Express 18(1), 102–107 (2010). [CrossRef]
12. D. M. Kita, B. Miranda, D. Favela, D. Bono, J. Michon, H. Lin, T. Gu, and J. Hu, “High-performance and scalable on-chip digital Fourier transform spectroscopy,” Nat. Commun. 9(1), 4405 (2018). [CrossRef]
13. M. Souza, A. Grieco, N. C. Frateschi, and Y. Fainman, “Fourier transform spectrometer on silicon with thermo-optic non-linearity and dispersion correction,” Nat. Commun. 9(1), 665 (2018). [CrossRef]
14. S. N. Zheng, J. Zou, H. Cai, J. F. Song, L. K. Chin, P. Y. Liu, Z. P. Lin, D. L. Kwong, and A. Q. Liu, “Microring resonator-assisted Fourier transform spectrometer with enhanced resolution and large bandwidth in single chip solution,” Nat. Commun. 10(1), 2349 (2019). [CrossRef]
15. Z. Wang, S. Yi, A. Chen, M. Zhou, T. S. Luk, A. James, J. Nogan, W. Ross, G. Joe, A. Shahsafi, K. X. Wang, M. A. Kats, and Z. Yu, “Single-shot on-chip spectral sensors based on photonic crystal slabs,” Nat. Commun. 10(1), 1020 (2019). [CrossRef]
16. Q. Sun, P. Falak, T. Vettenburg, T. Lee, D. B. Phillips, G. Brambilla, and M. Beresna, “Compact nano-void spectrometer based on a stable engineered scattering system,” Photonics Res. 10(10), 2328 (2022). [CrossRef]
17. A. Li and Y. Fainman, “On-chip spectrometers using stratified waveguide filters,” Nat. Commun. 12(1), 2704 (2021). [CrossRef]
18. Z. Zhang, Y. Wang, J. Wang, D. Yi, D. W. U. Chan, W. Yuan, and H. K. Tsang, “Integrated scanning spectrometer with a tunable micro-ring resonator and an arrayed waveguide grating,” Photonics Res. 10(5), A74 (2022). [CrossRef]
19. Q. Qiao, X. Liu, Z. Ren, B. Dong, J. Xia, H. Sun, C. Lee, and G. Zhou, “MEMS-Enabled On-Chip Computational Mid-Infrared Spectrometer Using Silicon Photonics,” ACS Photonics 9(7), 2367–2377 (2022). [CrossRef]
20. Z. Zhang, Y. Wang, and H. K. Tsang, “Tandem Configuration of Microrings and Arrayed Waveguide Gratings for a High-Resolution and Broadband Stationary Optical Spectrometer at 860 nm,” ACS Photonics 8(5), 1251–1257 (2021). [CrossRef]
21. S. Zheng, H. Cai, J. Song, J. Zou, P. Y. Liu, Z. Lin, D.-L. Kwong, and A.-Q. Liu, “A Single-Chip Integrated Spectrometer via Tunable Microring Resonator Array,” IEEE Photonics J. 11(6), 1–11 (2019). [CrossRef]
22. C. Sun, Z. Chen, Y. Yin, Y. Ye, Y. Luo, H. Ma, J. Jian, Y. Shi, C. Zhong, D. Zhang, H. Lin, and L. Li, “Broadband and High-Resolution Integrated Spectrometer Based on a Tunable FSR-Free Optical Filter Array,” ACS Photonics 9(9), 2973–2980 (2022). [CrossRef]
23. C. Sun, Y. Yin, Z. Chen, Y. Ye, Y. Luo, H. Ma, L. Wang, M. Wei, J. Jian, R. Tang, H. Dai, J. Wu, J. Li, D. Zhang, H. Lin, and L. Li, “Tunable narrow-band single-channel add-drop integrated optical filter with ultrawide FSR,” PhotoniX 3(1), 12 (2022). [CrossRef]
24. C. Sun, C. Zhong, M. Wei, H. Ma, Y. Luo, Z. Chen, R. Tang, J. Jian, H. Lin, and L. Li, “Free-spectral-range-free filters with ultrawide tunability across the S + C + L band,” Photonics Res. 9(6), 1013 (2021). [CrossRef]
25. D. Liu, C. Zhang, D. Liang, and D. Dai, “Submicron-resonator-based add-drop optical filter with an ultra-large free spectral range,” Opt. Express 27(2), 416–422 (2019). [CrossRef]
26. A. M. Prabhu, A. Tsay, H. Zhanghua, and V. Vien, “Extreme Miniaturization of Silicon Add–Drop Microring Filters for VLSI Photonics Applications,” IEEE Photonics J. 2(3), 436–444 (2010). [CrossRef]
27. A. A. E. Zhixuan Xia, Mohammad Soltani, Babak Momeni, Qing Li, Maysamreza Chamanzar, Siva Yegnanarayanan, and Ali Adibi, “High resolution on-chip spectroscopy based on miniaturized microdonut resonators,” Opt. Express 19(13), 12356–12364 (2011). [CrossRef]
28. S. Liu, J. Feng, Y. Tian, H. Zhao, L. Jin, B. Ouyang, J. Zhu, and J. Guo, “Thermo-optic phase shifters based on silicon-on-insulator platform: state-of-the-art and a review,” Front. Optoelectron. 15(1), 9 (2022). [CrossRef]
29. L. Zhang, M. Zhang, T. Chen, D. Liu, S. Hong, and D. Dai, “Ultrahigh-resolution on-chip spectrometer with silicon photonic resonators,” Opto-Electron. Adv. 5(7), 210100 (2022). [CrossRef]
30. C. Sun, Z. Chen, Y. Ye, K. Lei, H. Ma, M. Wei, R. Tang, J. Wu, H. Lin, and L. Li, “Scalable On-Chip Microdisk Resonator Spectrometer,” Laser Photonics Rev. 17(5), 2200792 (2023). [CrossRef]
31. X. Chen, X. Gan, Y. Zhu, and J. Zhang, “On-chip micro-ring resonator array spectrum detection system based on convex optimization algorithm,” Nanophotonics 12(4), 715–724 (2023). [CrossRef]
32. J. Zhang, Z. Cheng, J. Dong, and X. Zhang, “Cascaded nanobeam spectrometer with high resolution and scalability,” Optica 9(5), 517 (2022). [CrossRef]
33. B. Redding, S. Fatt Liew, Y. Bromberg, R. Sarma, and H. Cao, “Evanescently coupled multimode spiral spectrometer,” Optica 3(9), 956 (2016). [CrossRef]
