Silicon Photonics Technology Promotes Advanced High-Performance Computing
Research Team Led by Professor Pei-Wen Li, Institute of Electronics,
National Yang Ming Chiao Tung University
-
1. Silicon Photonics Technology Advances Classic High Performance, Energy-Efficient Computing |
Data centers around the world and popular artificial intelligence products (such as ChatGPT, DeepSeek, etc.) all have an urgent need for highly effective, energy efficient hardware (including processors and circuit systems) in order to manage and process massive amounts of data. However, the performance (i.e. the number of computations per Joule) of commercial core processors such as CPUs and GPUs reached saturation in the mid 2010s and have since been unable to progress further. This is because Moore’s Law, which has long guided the development of computing technology, has reached its limit. Computing efficiency and energy consumption can no longer be addressed simultaneously by simply shrinking the critical length and technology node of transistors. The short channel effect causes a transistor’s leakage current (and energy consumption) to increase dramatically. In addition, the extreme reduction of metal wire diameters and the use of electrical interconnections with metal wiring over long distances for the sake of increasing the packing density of electronic components also increases energy consumption. These issues form the main bottleneck when it comes to speed. The performance of the computing system overall can be improved through the use of multi-core processors, but the processor system’s board-to-board, chip-to-chip circuitry is still connected via cables or metal wires. The more multi-core processors there are, the more complex the electrical connections in the interface. This is an enormous disadvantage when it comes to improving the overall performance of the computing system.
In recent years, silicon photonics technology has become the best booster for promoting effective, energy-efficient computing systems. If we can make use of optical signals with their ideal properties, such as their great speed and zero-energy consumption, and supplement or replace “electrical interconnections” with “optical interconnections, it should be possible to improve or even overcome the delays and energy consumption issues of charge signal transmissions. Currently, data centers are using the first generation of silicon photonics technology to replace the cable connections between circuit boards. Using pluggable optical transceiver modules, optical signals can be transmitted via optical fibers then converted into electrical signals and transmitted through metal wires into the switches. Though this type of transmission, which uses both optical fibers and metal wires, can effectively reduce the quantity and length of cables used, it still has problems with the “heat” generated by the metal wires and the attenuation of signal strength. Therefore, processor manufacturers, such as NVIDIA, AMD, and Qualcomm, etc., have been proposing support for second or 2.5 generation co-packaged optics (CPO) modules to chip manufacturers, such as intel, TSMC, Samsung, and GlobalFoundries, etc.. The idea is that the optoelectric modules (such as optical receivers, optical waveguides, optical modulators, current and voltage amplifiers, driver ICs, switches and other components) originally scattered all over the printed circuit board (PCB) would be co-packaged onto a single silicon chip. This could significantly reduce the quantity of wire connections used, thereby increasing computing speed and bandwidth while reducing power consumption.
At the 2024 IEDM international conference held in San Francisco, TSMC announced the launch of a new advanced silicon photonics and packaging technology for data center applications—the “Compact Universal Photonic Engine (COUPE)”. [1–4] There are process design kits (PDK) for individual optical components (germanium photodetectors, micro-ring light modulators, silicon nitride/silicon optical waveguides, Grating Coupler and Edge Coupler for connecting to external optical fibers, and temperature sensors, etc.) already available to chip designers and manufacturers. TSMC has published several papers on silicon photonics engines, demonstrating the integration of electronic integrated circuit chips and silicon photonics circuit chips using CoWoS (Chip-on-Wafer-on-Substrate) packaging technology. TSMC expects to officially launch a pluggable optical transceiver module co-packaged on a PCB in 2025, followed by the launch of planar integration of photonic and electronic integrated circuits on substrates and the vertically-stacked integration of photonic and electronic integrated circuits on interposers in 2026.
Even though many advanced electronic processors and integrated circuit systems have been successfully manufactured using advanced and robust CMOS process technology, it remains difficult to realize vertically stacked integrated silicon photonics-electronics integrated circuits on silicon platforms. First, there is the issue of the geometric size (for example, the characteristic length or film thickness tends to be around sub-µm or µm, depending on the wavelength of the transmitted light) of the optical components in the photonic integrated circuits (such as semiconductor lasers, photodetectors, optical modulators, and especially optical waveguides, etc.). They are much larger than the electronic components (such as transistors, capacitors, etc.) in electronic integrated circuits. Furthermore, the connection between optical components focuses on optical coupling. In order to reduce optical coupling loss, the coupling area or coupling length usually needs to be greater than µm2 or µm. This is hundreds or even thousands of times larger than the electrical contacts of electrical components. However, in order to reduce light scattering and transmission loss, the sidewall roughness of optical components must be at the nm-scale. In contrast, electrical components rarely require nanometer-level sidewall roughness. In order to realize silicon photonics technology, chip manufacturers such as IBM, the predecessor to GlobalFoundries, began developing silicon photonics technology in the early 2000s (or even earlier) and formally proposed the concept and prototype for silicon photonic optical connections in 2008. Long-term, strategic collaboration between Intel and the University of California, Santa Barbara (UCSB), Stanford University, and MIT has also established deep roots in silicon photonics technology. In addition to the single-crystal thin film epitaxial growth required for individual optical components, they have enabled the development of optimized processes for structural design and optical performance and the development of photonic design automation (PDA) software, as well as furthered development of the coupling design and process integration of active optical components and optical waveguides. Even optical integrated circuits and systems have been demonstrated (such as the high-density wavelength division multiplexing fiber backbone transmission system). At present, Intel’s cutting-edge silicon photonics technology holds a leading position in the chip manufacturing industry. Currently, silicon photonics chips have been integrated into data centers, 5G communications and other technical fields, opening up new and extensive business opportunities. Companies such as Google, Apple, Meta, Amazon, and Microsoft are also actively developing high efficiency silicon photonics interconnection chip technology, building optical links within short distance data centers (850nm laser and multimode fiber) and between long distance data centers (1310nm laser and single-mode fiber). Other European and American companies and research institutions, such as Cisco, STM and IMEC, AIM Photonics, IME, etc., have already laid out the technical groundwork for integrating silicon photonics integrated optical circuits with CMOS integrated circuits, hoping to further improve the speed and bandwidth of on-chip signal processing.
Although wafer manufacturers have now launched foundry services for silicon photonics technology for circuit board to circuit board optical connection applications, in order to fully realize the benefits of high speed, energy-efficient computing promised by silicon photonics technology, it is necessary to start employing on-chip optical connections. However, chip foundries are currently unable to provide lasers capable of being directly integrated. Professor John Bower of the University of California, Santa Barbara (UCSB) gave a special presentation at the 2024 IEDM international conference that provided detailed and insightful explanations of the key technologies required for fabricating directly integrated quantum dot laser sources on silicon platforms and the technological evolution for achieving high-capacity silicon photonics integrated circuits. [5] Typically, semiconductor thin film laser light sources need to be triggered using extremely large driving currents. However, during long-term operation, as the operating temperature gradually increases, the wavelength of the light can undergo a red shift, and the light intensity will decrease. As a result, they often suffer from issues with temperature stability, reliability and service life. What’s more, silicon is an indirect bandgap semiconductor with a rather low light-to-electricity conversion efficiency. Fortunately, in 2008, MIT was able to demonstrate the functionality of germanium semiconductor lasers. [6] Using heavy n-type doping and stretch deformation techniques, the indirect bandgap of germanium films can be converted into a pseudo-direct bandgap. However, the Free Carrier Absorption effect greatly increases the laser threshold (threshold current value: about 280 mA/cm2). This results in the germanium laser having a short operating life and is detrimental to both the temperature reliability and stability of operation. Therefore, at present, the only way to integrate III-V laser light sources and active/passive silicon photonics components is through packaging or wafer bonding. However, the costs of such packaging and wafer bonding are high and thus not conducive to market competitiveness.
On the other hand, thanks to the three-dimensional quantum confinement effect, quantum dots can effectively keep excitons locked inside so that the driving current threshold required to start laser luminescence is relatively small. They can also operate stably in higher temperature environments. In addition, the dispersed energy levels within the quantum dots help purify the monochromatic wavelength of the quantum dot lasers (spectral line width of tens of nm or narrower, which is only 1/10 the line width of a typical laser source), enabling stable pulse laser modes and longer laser life (10 thousand times longer). [5] Moreover, quantum dots have quantum physical properties of tunable electronic energy structures. The wavelength of the light emissions can be adjusted by changing the diameter of the quantum dots during formulation. If combined with engineering designs such as surface plasmon resonance (SPR), the performance of quantum structure lasers can be even further optimized. Professor John Bower specifically introduced the process used by his laboratory where silicon dioxide film on a silicon platform is etched to reveal the silicon before the selective epitaxial growth of InAs quantum dots. This process demonstrated concrete results for creating high quality quantum dot laser light sources.
In addition to quantum dot lasers, germanium photodetectors that can be integrated with silicon nitride/silicon optical waveguides are also an important cornerstone for the realization of the silicon photonic platform. Due to germanium semiconductors’ relatively small bandgap, about 0.66 eV, a high dark current is a common concern. It affects the signal to noise ratio of photoelectric conversion and detectivity in addition to causing serious energy loss. Many studies have shown that incorporating quantum dots into the light absorption structure of photodetectors can reduce dark current, improve the lifetime, thermal stability and photoresponse of photocarriers, adjust the wavelength of detection light and even reduce the thickness/area of devices. [7] Therefore, research into quantum dot light sources and light detectors has also been booming.
In addition to the manufacturing challenges facing components such as quantum dot laser light sources and germanium photodetectors, another practical technical challenge for the commercialization of silicon photonics technology is how to accurately and efficiently simulate and design silicon photonics’ main components and passive components (such as semiconductor lasers, photodetectors, optical modulators, optical waveguides, spectrometers, and light collectors, etc.) as well as how to evaluate the performance of the entire photonic integrated circuit system. Professor Jelena Vuckovic of Stanford University gave a special presentation at the 2024 IEDM international conference on the high-speed, high-performance electromagnetic (EM) simulation analysis software and hardware (Stanford Photonics Inverse Design Software, SPINS) developed by her laboratory in collaboration with other leading research institutions, laboratories and wafer manufacturers. Through the mature processes of industry-university cooperation, miniaturized and integrated photonic integrated circuit systems are being produced for verification. [8] Using inverse design, the laboratory and its partners successfully demonstrated Scalable Quantum and Classical Photonics integrated optical circuit systems. For example, they can provide broadband multi-channel optical transceiver modules for data center applications or use a self-made miniaturized Ti:Sapphire laser on a silicon carbide platform to control the electron spin quantum bit of silicon holes.
2. Silicon Photonics Technology Plays a Key Role in Quantum Computing |
As Professor Jelena Vuckovic said, silicon photonics technology can not only improve the computing speed of traditional classical computing and enhance the transmission efficiency of data centers but also promote the practical development of quantum bit technology for quantum computing. In fact, whether it’s silicon hole or ion trap qubits, silicon photonic integrated circuits will be needed to control and read quantum states in order to fully realize the operation of quantum bits.
At present, most ion trap qubits, which claim to be capable of operating at room temperature, are placed in vacuum chambers and controlled remotely using light or microwave signals on an optical table [9]. When reading the quantum state information of the ion trap qubit, the high aperture lens on the optical table focuses to a single-photon detector. Although the functionality of ion trap qubits has been verified in a room temperature, vacuum environment, in terms of expanding the number of qubits and improving the fidelity of initializing/manipulating/detecting quantum states, ion trap quantum technology still faces many technical challenges. [9][9]
This is due to the complex alignment of optical components and optical fibers as well as environmental disturbances such as mechanical vibrations/thermal noise, which generates numerous additional noise sources, limiting the reading fidelity of ion trap qubits. Ideally, a laser light source would be used to directly control ion trap qubits through an optical waveguide in a vacuum environment while the single-photon detector connected with an optical waveguide is used to directly read the quantum state information then output it to a CMOS integrated circuit for subsequent signal processing. This would not only eliminate the jitter/drift problem of optical components in free space and the ubiquitous electrical noise but also eliminates the problems of excessively long optical fibers and the complex alignment between optical components. It would go a long way to improving the reading fidelity of ion trap qubits.
2-1. Technical Challenges in the Application of Silicon Photonics in Quantum Bit Technology |
Although the research and development of silicon photonics technology has been going on for over thirty years, most of that research and development has been aimed at applications such as communications and classical computing. The silicon photonics connection technology that has been developed (components such as optical waveguides, optical modulators, optical detectors and even light sources) have therefore focused on processing high-speed, high bandwidth, high response or high wattage optical/electrical signals for communications applications. If silicon photonics devices are to be used to control or read ion trap qubits, they must be able to handle low noise, low dark current or extremely low wattage (with few photons) near ultraviolet and visible light signals. They must even be able to operate stably in low temperature environments. The following is a brief description of the key technical requirements and challenges for the application of silicon photonics devices to ion trap qubit technology.
Integrated Silicon Nitride Optical Waveguides and Gratings |
he wavelength range of laser light sources commonly used to manipulate ion trap qubits is approximately 300-2000 nm. This covers the near ultraviolet, visible and near infrared spectrums. Unfortunately, the silicon waveguides commonly used in current silicon photonics technology have high absorption rates in the ultraviolet and visible light bands. The high degradation of optical properties is unsuited for ion trap qubit technology. In contrast, silicon nitride (Si3N4) optical waveguides are transparent in the ultraviolet-visible wavelength range and transparent and non-light-absorbent with low optical loss in the ultraviolet-visible wavelength range [10,11]. Silicon nitride is commonly used in insulating layers, spacer layers, protective layers and more in CMOS process technology. It can be deposited using conventional chemical vapor deposition (CVD) process methods. The chemical vapor deposition process recipe can also be fine-tuned according to actual application requirements, adjusting the chemical composition of the silicon nitride (for instance, SixNy or even SiOxNy film) and its refractive index. This helps modulate the mode number, optical limitations and transmission loss of silicon nitride optical waveguides. Because silicon nitride optical waveguides can cover the visible light (400-1000nm) band that silicon optical waveguides cannot cover, they have become the best platform for various experimental chips in recent years, especially for quantum communications/computing [12].
In 2020, the Swiss Federal Institute of Technology in Zurich [13] reported an ion trap quantum logic gate that integrated an ion trap chip with a silicon nitride optical waveguide. This ion trap quantum logic gate demonstrated the use of a single-mode optical fiber to inject 729 nm visible light into a silicon nitride optical waveguide then transmit to the ion trap chip in a low-temperature, vacuum environment. This method can eliminate the problems of optical alignment on the optical table and mechanical vibration and beam point drift, thus improving the fidelity of quantum logic gates. However, the ion trap chip demonstrated by the Swiss Federal Institute of Technology in Zurich has not yet been integrated with active silicon photonics components such as light modulators and single-photon detectors. This is because the nucleation incubation time for germanium and silicon germanium on silicon nitride film is very short, making it difficult to grow high quality, single-crystal germanium or silicon germanium films on silicon nitride films via selective epitaxial growth. This makes it impossible to proceed to making silicon photonic active components. One option is to use wafer bonding. After bonding SOI to a silicon nitride platform, epitaxial growth can be used to grow the optical active layer—germanium or silicon germanium film [14]. Another option is to imitate STM, IHP and the University of Toronto and make silicon germanium modulators and germanium photodetectors on SOI platforms then depositing PECVD silicon nitride film, performing CMP polishing and finally making the top layer of the silicon nitride optical waveguide. [14-16] However, the process for the top silicon nitride optical waveguide in the second proposal makes it difficult to proceed with the high temperature annealing process for the dehydrogenation or densification of silicon nitride, which means the number of defects inside the silicon nitride optical waveguide cannot be reduced. This will cause the lattice relaxation of the optical active area—the silicon germanium and germanium epitaxial film—at the bottom, resulting in performance degradation of the optical active component. So far, the literature on monolithic integrated high-speed optical modulators [14], high-speed photodetectors [14,15] and laser sources [17] using Ge/SiGe on SiN optical platforms is very limited. As such, the fabrication of optical active components and the integration of optical active/passive components on silicon nitride platforms are important research topics.
Integrated On-Chip Single-Photon Detector |
The quantum state signal of the quantum bit is very weak, so it is easy for noise in the surrounding environment to interfere with it. As such, it is necessary to have a single-photon detector that can be directly “built-in” to quickly and accurately detect and read changes in the small number of photons in the ion trap qubit. Ideally, a single-photon detector and the ion trap quantum chip would be connected directly with a silicon nitride optical waveguide to minimize the crosstalk between the “collection” and “detection” of photons. This could also further expand and measure the feasibility of large ion trap bit arrays. Generally speaking, the wavelength of photons emitted by ion trap qubits is mostly 300–500 nm. However, the “most mature” silicon avalanche photodetector can only detect 850 nm light. It is unable to directly detect the state of ion trap qubits. NIST researchers in America used a self-made, built-in “superconducting” single-photon detector with no need for imaging lenses or cameras that could read the quantum state of beryllium ions with near-perfect accuracy (reading accuracy exceeded 99.9%). [18] However, “superconducting” single-photon detectors can only operate normally in an environment close to absolute zero. According to the NIST’s report, in order to effectively improve the detection efficiency and reduce the dark count rate, there is an urgent need for near ultraviolet silicon-based single-photon detectors compatible with CMOS technology. It is also necessary to be able to monolithically integrate silicon nitride optical waveguides/gratings with silicon-based single-photon detectors,to further reduce coupling loss and noise and expand the number of ion trap qubits.
Integrated On-Chip Light Sources |
In addition to single-photon detectors that can be integrated with ion trap quantum chips, visible light sources coupled to silicon nitride optical waveguides are also key components for the manipulation of ion trap quantum chips. As Professor John Bower said in his IEDM presentation, realizing the integration of a light source on a silicon substrate has always been the biggest obstacle facing silicon photonics technology—let alone the realization of a visible light source that can be integrated with silicon nitride optical waveguides.
The literature reports that germanium nanostructures such as quantum wells, quantum wires and even quantum dots can effectively reduce the problem of defects when growing single-crystal germanium films on silicon wafers. Especially in small germanium quantum dots, the quantum confinement effect causes strong overlapping coupling of electron-hole wave functions, greatly enhancing the optical transition oscillation intensity of the germanium quantum dots and breaking the curse that germanium bulk materials must adhere strictly to energy-momentum (E-k) conservation. What’s more, by adjusting the diameter of a single-material germanium quantum dot, the luminescence energy gap can be adjusted to emit light of different wavelengths, overcoming the limitation of needing to select different materials to make light sources with different wavelengths. However, the light emitted by a single quantum dot is small. It needs to be placed in a resonance chamber. When the laser light irradiates the quantum dot/resonance chamber, the “Purcell effect” rapidly increases the number of photon excitations within the quantum dot, thus improving the overall luminosity. Commonly used quantum dot resonance chamber structures include photonic crystals, micro-disks and micro-rings. Given the complexity of the photonic crystal array structure design (the template thickness, hole diameter, period, and defect mode, etc.) and the extremely high requirements for process precision (advanced lithography systems must be used to expose sub-micron-level holes (diameter or period)), the design and fabrication of micro-disk and micro-ring resonance chambers at the micron level are comparatively easy. Moreover, the emitted in-plane light can be coupled with the adjacent bus waveguide, which is beneficial for on-chip integration.
The micro-disk resonator is comprised mainly of a light field confined in a disk-shaped optically dense medium. At the edge of the micro-ring resonator, resonance is achieved along the radial direction of the disk, generating the fiber whispering gallery mode (WGM). The micro-disk resonator structure offers considerable flexibility and many cost advantages for the design and manufacture of electrodes and waveguides. In recent years, European and American research institutions have demonstrated photo-induced micro lasers by embedding various quantum dots (including silicon, germanium and CdSe) in suspended silicon, germanium, silicon dioxide or silicon nitride micro-disk resonators. CNRS-Univ. in France published a series of studies [19] in which a 300-nm thick n+- germanium layer is first grown via epitaxy on a GaAs substrate. After the suspended germanium micro-disk was made via lithography, silicon nitride was deposited to coat the germanium micro-disks, forming a tensile deformation n+-Ge active light-emitting layer and demonstrating a photo-excited germanium micro laser. However, the germanium on gallium arsenide (Ge-on-GaAs) method is difficult to transfer to a silicon platform. Tokyo City University proposed a P-I-N germanium quantum dot micro-disk diode, [20], which can be coupled with adjacent waveguides to generate electroluminescence. However, most quantum dot micro-disks are made on SOI platforms and therefore not suitable for visible light sources. There is an urgent need to develop quantum dot silicon nitride micro-disk visible light sources to enable smooth integration with ion trap quantum chips.
At the 2022 flagship IEDM international conference [21], our research team reported a monolithically integrated silicon nitride waveguide (involving a grating coupler and waveguide cone) with a germanium quantum dot micro-disk light source and photon detector components that can be used for near ultraviolet-visible light ion trap sensing applications, as shown in Figure 1.
![]() Fig. 1 Ge quantum-dot photodiodes and light emitter embedded in Silicon-Nitride |
We used CMOS process technology to manufacture germanium quantum dot arrays with adjustable diameters and spatial positions. Using selective oxidation, it is possible to transform polycrystalline silicon germanium pillars defined via lithography on a silicon nitride film into spherical germanium quantum dots embedded in silicon nitride in a single step. The most important feature of our germanium quantum dots is that they are prepared via thermal oxidation at 900 °C. This means that, they have the advantage of high-temperature thermal stability, as shown in Figure 2. This inherent thermal stability opens up the possibility of using Ge quantum dot photodetectors and emitters with top or bottom SiN waveguides through evanescent wave coupling. From the perspective of device manufacturing and integration, the top waveguide coupling structure offers flexibility in both device (light detectors and emitters) design and three-dimensional integration material selection. The top waveguide coupling structure can eliminate the need for the “waveguide” and “substrate” to be made from the same material. Our SiN embedded germanium quantum dot array structure provides the flexibility to integrate SiN micro-disk light emitters and PIN photodetectors with top or bottom SiN waveguides and the feasibility of implementing three-dimensional PIC integration. [22] The germanium quantum dot production method we developed directly utilizes CMOS process technology. With excellent process control and engineering advantages for component design, it can directly produce quantum bits, single-electron transistors and photoelectric transistors, etc.. As such, it has great potential for practical application and industrialization. It can also assist in the development of technologies such as quantum computing and optical connection. |
![]() Fig. 2 Formation of self-organized heterostructures of capping SiO2/Ge QDs within host of Si3N4 on top of SOI as evidenced by TEM, HAADF STEM and EDS maps of elemental Ge (green), N (red), and O (white) micrographs. After P. W. Li et al., IEDM Tech. Dig. pp. 451-454 (2022). |
Reference:
[1] Y. J. Mii, “Semiconductor industry outlook and new technology frontiers,” IEDM Tech. Digest, 1.1, Dec. 2024
[2] S. K. Yeh et al., “Silicon photonics platform for next generation data communication technologies,” IEDM Tech. Digest, 23.3, Dec. 2024
[3] C. H. Fann et al., “Novel parallel digital optical computing system (DOC) for generative AI,” IEDM Tech. Digest, 31.7, Dec. 2024
[4] H. Hsia et al., “EPIC-BOE: An electronic-photonic chiplet integration technology with IC processes for broadband optical engine applications,” IEDM Tech. Digest, 31.8, Dec. 2024
[5] J. Bower, “Integrated quantum dot lasers and high capacity silicon photonic integrated circuits,” IEDM Tech. Digest, 23.1, Dec. 2024
[6] RE Camacho-Aguilera et al., “An electrically pumped germanium laser,” Optics Express, vol. 20, no. 10, pp. 11316-11320, 2012
[7] Hongmei Liu, Fangfang Zhang, Jianqi Zhang, Guojing He, “Performance analysis of quantum dots infrared photodetector,” Proceedings of SPIE, vol. 8193, 81930J (2011)
[8] Jelena Vuckovic, “Scalable quantum and classical photonics,” IEDM Digest, 26.3, Dec. 2024
[9] K. Brown et al., “Materials challenges for trapped-ion quantum computers,” Nat. Rev. Mater. vol. 6, pp. 892–905 (2021).
[10] D. J. Blumenthal et al., “Silicon nitride in silicon photonics,” Proc. IEEE, vol.106, 12 (2018).
[11] P. Munoz et al., “Silicon nitride photonic integration platforms for visible, near-infrared and mid-infrared applications,” Sensors, vol.17, 2088(2017).
[12] A. Orieux et al., “Recent advances on integrated quantum communications,” J. Opt., vol. 18, 083002 (2016).
[13] K. Mehta et al., “Integrated optical multi-ion quantum logic,” Nature vol. 586, pp. 533–537 (2020).
[14] F. Boeuf et al., “A silicon photonics technology for 400 Gbit/s applications,” IEDM Tech. Digest, pp. 775 (2019)
[15] S. Lischke et al., “Silicon nitride waveguide coupled 67GHZ Ge photodiode for non-SOI PIC and EPIC platforms,” IEDM Tech. Digest, pp. 779 (2019)
[16] W. Sacner et al., “Monolithically integrated multilayer silicon nitride-on-Si waveguide platforms,” Proc. IEEE, vol. 16, 2232 (2018)
[17] S. Bao et al., “Low-threshold optically pumped lasing in highly strained germanium nanowires,” Nat Comm. vol. 8, 1845 (2017)
[18] S. L. Todaro et al., “State readout of a trapped ion qubit using a trap-integrated superconducting photon detector” Phys. Rev. Lett. 126, 010501 (2021)
[19] A. Ghrib et al., “Tensile-strained germanium microdisks,” Appl. Phys. vol. 102, 221112 (2013)
[20] X. Xu, T. Maruizumi, and Yasuhiro Shiraki, “Waveguide-integrated microdisk light-emitting diode and photodetector based on Ge quantum dots,” Optics Exp., vol. 22, 3905 (2014)
[21] C. H. Lin, P. Y. Hong, B. J. Lee, H. C. Lin, T. George, and P. W. Li, “Monolithic integration of top Si3N4-waveguided germanium quantum-Dots microdisk light emitters and PIN photodetectors for on-chip ultrafine sensing,” IEDM Tech. Dig. pp. 451454, Dec. 2022.
[22] C. H. Lin, P. Y. Hong, B. J. Lee, H. C. Lin, T. George, and P. W. Li, “Self-organized germanium quantum-dots/Si3N4 enabling monolithic integration of top Si3N4 waveguided microdisk light emitters and PIN photodetectors for on-chip sensing,” IEEE Trans. Electron Dev. vol. 70, no. 4, pp. 2113-2120 (2023)