Review: Prospects and applications of lasers on a chip

At the beginning of the new year, eLight has prepared 3 inaugural reviews for all readers.

Today, we first understand the first article, the “review of lasers on chip” by the team of Professor John Bowers of the University of California, Santa Barbara and the team of Assistant Professor Wan Yating of King Abdullah University of Science and Technology.

The other two opening reviews, namely “Metal Halide Perovskite Review” by the team of Academician Huang Wei of Northwestern Polytechnical University, and “Review of Light Field Control Based on Weyl Semimetals” by Professor Shanhui Fan’s team of Stanford University in the United States, have been launched today, and Chinese interpretation will be launched this week.

Humans produce about 2.5 trillion bytes of data every day, and this number will rise rapidly with the spread of 5G and big data. According to a 2018 Forbes survey, 90% of the world’s data was generated in the past two years. With the explosive growth of data traffic, traditional electronic information interconnection architectures can no longer meet the increasing bandwidth and energy consumption requirements.

Therefore, silicon photonics (Silicon photonics) came into being. Silicon-based optoelectronic chips can not only take advantage of the advantages of microelectronics technology in low-cost, large-scale CMOS integration, but also have the advantages of small attenuation of optical signals during transmission, high transmission bandwidth, fast transmission rate, strong anti-interference performance, and low power consumption. As one of the most promising solutions for realizing on-chip optical interconnection in the post-Moore’s Law era, silicon photonics technology has great potential in the fields of lidar, biochemical sensing, quantum information processing and high-performance computing.

However, silicon is an indirect bandgap semiconductor and cannot achieve high efficiency in luminescence. With the gradual maturity of photonic integrated devices such as silicon-based modulators, detectors, and couplers, efficient light sources on silicon substrates have become a bottleneck restricting the development of silicon photonics technology. In order to break through this bottleneck, on-chip lasers, as the “heart” of silicon photonics chips, have become one of the most active fields of photonics research in the past decade.

Recently, the team of Professor John Bowers of the University of California, Santa Barbara (UCSB) and the team of Assistant Professor Wan Yating of King Abdullah University of Science and Technology (KAUST) were invited to write a review article entitled “Prospects and applications of on-chip lasers”, which was published in the new journal eLight of the Excellence Program. This paper systematically introduces the different schemes and latest research progress of light sources on silicon substrates, and looks forward to the application prospects of this technology in communications, lidar, sensing, quantum information processing and optical computing.

Figure 1: Schematic diagram of a silicon-based integrated photonic chip

Implementation of the on-chip laser

An ideal integrated on-chip laser should meet the following requirements:

First, high luminous efficiency;

Second, electric pump excitation can be realized;

Third, compatibility with existing CMOS processes;

Fourth, the wavelength of the outgoing light matches the existing optical communication wavelength (1310 nm/1550 nm).

In addition, as the core of the Photonic integrated circuit (PIC), the integrated on-chip laser must also ensure stable operation (30 to 150°C) and a sufficiently long operating life in a wide temperature range.

At present, integrated on-chip lasers are mainly divided into germanium-silicon IV light sources and silicon-based III-V light sources in terms of materials. Germanium silicon IV light source can transform germanium from indirect bandgap material to direct bandgap material by n-type doping, stress stretching and germanium-tin (GeSn) alloy, which greatly improves the original luminous efficiency. Germanium and silicon are both group four elements, and the process of growing germanium on silicon is relatively mature. However, the performance indicators such as threshold current and luminous efficiency of germanium-silicon IV lasers currently realized are far behind those of III-V lasers and cannot meet the requirements of actual use.

III-V group materials are direct bandgap semiconductors with inherent advantages in their luminescence properties. At present, there are three main integration schemes for integrating III-V materials into silicon substrates: hybrid integration, Heterogeneous integration based on wafer bonding, and Monolithic integration based on direct epitaxial growth.

Hybrid integration uses a coupler to introduce an external light source into the silicon waveguide, allowing each device to be tested and characterized prior to packaging, and selecting a chip that performs well for packaging, resulting in high flexibility. However, limited by complex packaging technology and the use of III-V group substrates, hybrid integration is expensive and large-scale, making it difficult to achieve large-scale integration of on-chip lasers. Heterogeneous integration uses techniques such as low-temperature plasma bonding to bond III-V substrates to silicon wafers, and then perform subsequent device fabrication. It avoids the problems of insufficient coupling efficiency and long alignment adjustment time in the first-generation flip-chip solution, and has directly spawned three related start-ups. Intel, the industry leader in this field, has developed more than ten years on this basis, and the photoelectric transceivers based on the silicon photonics technology have reached about 2 million shipments per year, and the products have been rapidly iterated from 100G to 200/400G or even higher speeds. However, heterobonding schemes are limited by the use of III-V substrates, which limits further cost reduction.

The method of direct epitaxial III-V materials on silicon substrates is suitable for large-scale growth and mass production, and is an ideal solution to solve the lack of core light source for silicon-based optoelectronic integration. However, limited by the polarity, lattice mismatch and thermal expansion coefficient of III-V and silicon materials, direct growth of III-V materials on silicon will cause material defects such as reversed-phase domains, penetration dislocations and microcracks, which will have a serious impact on device life and working performance. In order to solve this problem, the researchers used asymmetric slow filter layer, capture layer and other structural designs to reduce the dislocation density of the material, and also adopted a quantum dot structure that is insensitive to dislocation defects to further reduce the impact of dislocation defects on laser performance. Laser lifetime test results for this structure showed that the device showed only a 6.8% change in threshold current over 4000 hours of 80 °C high temperature test environment. This shows that the working life of the laser is up to one million hours, which can fully meet the actual needs of data centers and supercomputing centers. Further, the researchers introduced wafer bonding on the basis of epitaxial growth quantum dots, and transferred the quantum dot laser mode field generated by electric pumping to the silicon waveguide through the tapered waveguide. This technology of integrating quantum dot lasers into silicon photonics chips not only has significant low-cost advantages, but also has the performance advantages brought by the synergistic effect of the two, providing a solution with large-scale industrial application prospects for the further development of silicon photonics technology.

Figure 2: Schematic diagram of the on-chip quantum dot laser and device picture (the entire process is completed on a 4-inch silicon-based wafer)

Applications of lasers on a chip

Silicon-based III-V group on-chip lasers can combine the high luminous efficiency of III-V materials, mature process, high integration and low cost of silicon materials, and are expected to lead the rapid development of silicon-based optoelectronic integration in different application scenarios in the future:

Communication: The biggest driver for the development of silicon-based optoelectronic chips is still data communication. Companies such as Intel, Broadcom, Cisco and Hewlett Packard Enterprise continue to make efforts in silicon photonics chips and continue to propose innovative solutions, pushing the transmission rate of silicon photonics chips from Gbps to Tbps orders.

Lidar: Frequency-modulated continuous wave (FMCW) lidar based on optical phased array (OPA) has the potential to achieve long detection ranges, direct speed measurement, and powerful interference-resistant lidar systems. As one of the most promising platforms for chip-scale lidar, silicon-based optoelectronic platforms have achieved many key technological breakthroughs in recent years.

Biochemical sensing: Since Covid-19, there has been a huge market demand for wearable devices with biological health monitoring capabilities. Integrated silicon photoelectric sensing technology, mainly has two schemes: spectral absorption type and refractive index change type. Silicon photosensing technology is one of the most promising solutions to achieve high-sensitivity, portable sensors, and has driven the implementation of several products.

Quantum information processing: Optical quantum technology uses the quantum properties of photons for information processing, and has been proved to have important application prospects in confidential communication and molecular simulation in several frontier works reported in recent years. Traditional quantum optical paths are composed of discrete optical lenses, and the optical paths are complex and susceptible to interference. After the integration of quantum light source and linear network on silicon photonics chip, the volume of optical quantum information processing loop can be greatly reduced, and it has anti-interference, programming control and other properties, which is expected to become one of the core technologies in the field of quantum information.

Light calculation: Compared with microelectronic chips, silicon photonics chips have the characteristics of high throughput, high energy efficiency ratio and ultra-low latency, which have significant advantages in the computing field. At present, the research of silicon photonics computing chips mainly focuses on all-optical logic, photoelectric fusion neural network implementation, etc., and these innovative computing architectures will provide new solutions to the current problems of Moore’s Law constraints and von Neumann bottlenecks.

Figure 3: Application of silicon-based photonics integrated chips with integrated on-chip lasers

Challenges and prospects for on-chip lasers and their silicon-based integrated optoelectronic chips

Figure 4: Progress in silicon-based photonic integration at different stages of development since 1992

The two core problems in silicon-based optoelectronics are silicon-based light sources and silicon-based integration technology. While the integration of light sources on silicon substrates has made remarkable progress in recent years, it still faces the following serious challenges

First, silicon-based quantum dot lasers are expected to meet the requirements of luminous efficiency, outgoing power and high temperature working environment in practical applications, and their many advantages such as insensitivity to light reflection and anti-irradiation are ideal solutions to solve the problem of lack of core light sources in silicon-based optoelectronic integration. However, at present, most of the research on silicon-based quantum dot lasers is still limited to a single device, and the silicon substrate only plays the role of a substrate. How to realize the integration and co-optimization of multiple materials and new mechanisms on silicon-based platforms, and how to introduce light on quantum dot devices into silicon waveguides to form silicon photonics chips with complete functions of light generation, transmission, modulation, processing and detection are still unconquered problems.

Second, the design optimization focus of on-chip lasers shifts from the individual device to the system level. How to shift from design optimization of single discrete devices to system-level design optimization for large-scale integrated optical paths such as on-chip optical interconnect and optical computing, how to achieve the integration of optical devices and electrical devices, and how to balance the integration density of optoelectronic devices and the difficulty of process preparation are all issues that we need to consider when designing optical chips.

Third, silicon-based optoelectronic technology is still in its infancy, and many advanced silicon-based optoelectronic technologies still stop at the door of the laboratory and have not yet entered the research and development stage of the industry. Moreover, there are still great differences between the design ideas and process routes of major manufacturers in the industry, and the integrated on-chip laser solution with the best cost performance has not yet been determined, and the collaborative efforts of academia and industry are still needed to find the best solution.

From the first realization of on-chip lasers on silicon-based platforms in 2006 to now, in just 16 years, the research of on-chip lasers and silicon-based optoelectronic chips has achieved many remarkable results, which have become the key technology of future data communication. This is a science that “stands tall”, from basic science to practical applications, from the Internet connected by long-distance optical fiber to on-chip optical interconnection, lidar, sensing, and then to the future all-optical computing, with the continuous improvement of on-chip light source integration technology, silicon photonics is becoming the core driving force to promote the development of a new generation of information technology. (Source: China Optics WeChat public account)

Related paper information:

Special statement: This article is reproduced only for the need to disseminate information, and does not mean to represent the views of this website or confirm the authenticity of its content; If other media, websites or individuals reprint and use from this website, they must retain the “source” indicated on this website and bear their own legal responsibilities such as copyright; If the author does not wish to be reprinted or contact the reprint fee, please contact us.

Source link

Related Articles

Leave a Reply

Your email address will not be published. Required fields are marked *

Back to top button