Recently, we are jointly developing a LiDAR system based on multi-junction vertical cavity surface emitting lasers (VCSELs), which will be used in future in-vehicle autonomous driving systems. Professor Wang Jun from Sichuan University expressed, "I believe that in the near future, such LiDARs will become mainstream."
Recently, he and his research team successfully created a 15-junction vertical cavity surface emitting laser with an electro-optical conversion efficiency of 74%.
This is the highest efficiency reported in the field of vertical cavity surface emitting lasers to date, ending a 20-year stagnation in the electro-optical conversion efficiency of such devices.
In the field of semiconductor lasers, efficiency has always been one of the most prominent advantages of semiconductor lasers compared to other types of lasers.
However, in the past decade, there has been no core breakthrough in the technical route for improving the efficiency of semiconductor lasers, and this achievement provides a new idea for enhancing the efficiency of semiconductor lasers.The vertical cavity surface emitting laser (VCSEL) based on multi-junction cascading not only shows a significant advantage in the multiplication of power.
Advertisement
At the same time, the team also focused on analyzing its advantages in efficiency improvement, which plays an important role in the development of green energy photonics.
It is reported that after the power and efficiency of multi-junction VCSELs are greatly enhanced, the sensing accuracy and energy consumption of sensors in mobile terminals can be greatly improved.
Meanwhile, multi-junction VCSELs also show a very low cost advantage, which can greatly reduce the cost of LiDAR and have a great impact on LiDAR solutions based on other light sources.
In addition, the rapid development of AI has put forward higher demands for computing power and data centers. Recently, with the use of the PAM4+ modulation scheme, the power of VCSELs in the field of optical communication has become extremely important.Therefore, high-power, high-efficiency vertical-cavity surface-emitting lasers (VCSELs) also have great potential in long-distance, high-bit-rate communication.
Additionally, this achievement can also promote the development of AI computing power data centers.
Computing power is a very core element. The growth in demand for computing power also drives a significant increase in the data exchange capacity and speed between data centers and terminal devices.
This allows multi-junction vertical-cavity surface-emitting lasers to play a certain role in data centers and longer-distance communication.Vertical-Cavity Surface-Emitting Lasers, once used in iPhone facial recognition modules, are now hindered by energy issues.
It is understood that Vertical-Cavity Surface-Emitting Lasers (VCSELs) are a type of microcavity laser with advantages such as low cost, high reliability, narrow spectral linewidth, and a symmetrical circular light spot.
Due to the advantages of VCSELs, they quickly found commercial applications after their introduction. In the early days, the application of VCSELs was mainly focused on consumer electronics that require small size and low power light sources, such as mice, printers, and short-distance optical communication in data centers.
With the development of intelligent technology, VCSELs have become the core light source of intelligent sensing systems and have been widely and maturely applied in facial recognition and short-range sensing systems.
In particular, in 2017, VCSELs were successfully used in iPhone facial recognition modules, which attracted large-scale attention, and their application scenarios and demand began to explode rapidly.In recent years, with the development of higher-order AI technology, vertical-cavity surface-emitting lasers (VCSELs) face tremendous prospects and challenges in the fields of sensing, communication, atomic clocks, optical or quantum computing, topological lasers, and medical diagnostics.
With the rapid advancement of long-range sensing technology in autonomous driving, such as AI large models like ChatGPT that demand high data capacity and speed, as well as the rapid growth of intelligent and quantum technology applications based on VCSELs, such as deep learning, they inevitably face a common challenge: the issue of energy consumption.
Whether it is the battery consumption of mobile terminals or the energy consumption of data centers, VCSELs as light sources constitute an important part of energy loss.
Especially with the rapid development of AI computing, the energy consumption requirements of data centers will further increase, and it is expected to grow an order of magnitude by 2030.
Secondly, since the advent of semiconductor lasers in 1962, they have undergone rapid development.Compared to gas lasers, solid-state lasers, and fiber lasers, semiconductor lasers have many advantages, the most notable of which is the ability to achieve extremely high electro-optical conversion efficiency.
Pursuing the ultra-high efficiency of semiconductor lasers has always been an important goal in the field of photonics and laser physics.
Since the advent of edge-emitting semiconductor lasers, the record for power conversion efficiency has been continuously broken.
In 2006, edge-emitting semiconductor lasers achieved a conversion efficiency of 85% at -50°C. In 2007, they achieved a maximum power conversion efficiency of 76% at room temperature.
However, no new records have emerged in the past years. Therefore, these power conversion efficiency records have always represented the top records for all semiconductor lasers.Compared to edge-emitting semiconductor lasers, the power conversion efficiency of vertical cavity surface-emitting lasers (VCSELs) increases very slowly, and there are significant differences.
Since the maximum power conversion efficiency of single-junction VCSELs reached 62% in 2009, there has been no breakthrough progress for more than a decade.
It is generally believed that for microcavity lasers, VCSELs, to achieve high power conversion efficiency records has always been one of the long-term unattainable goals in photonics.
The main reason for the persistently low efficiency of VCSELs is that, as a microcavity laser, its cavity volume is extremely small, which leads to a significant reduction in its round-trip gain. The lower gain leads to a higher threshold current.
To achieve the goal of a lower threshold, the structural design of VCSELs usually adopts high-reflectivity mirrors formed by distributed Bragg reflectors grown from the top and the bottom.However, this design leads to a significant increase in resistance, which in turn results in Joule heating. In addition, doping within the distributed Bragg reflector causes carrier absorption loss. These two factors together limit the efficiency of vertical cavity surface emitting lasers (VCSELs).
It is against this backdrop that Wang Jun's team conducted this research.
After the impact of the pandemic and "academic disputes," they finally obtained three patent grants.
In fact, the team's research on multi-junction vertical cavity surface emitting lasers had already started in 2020.
At that time, the concept of early multi-junction vertical cavity surface emitting lasers had been proposed, and many well-known universities and companies around the world had applied for patent protection for multi-junction vertical cavity surface emitting lasers.At the beginning, the research group focused on reviewing literature and reproducing experimental results. Later on, they grew 3 junction vertical cavity surface-emitting lasers and achieved an efficiency of 62%.
"This efficiency is based on a commercially viable structural design, which is an extremely high level. Therefore, we also reported it at the 2021 International Conference on Photonics West in the United States," said Wang Jun.
In the spring of 2021, the COVID-19 pandemic forced the experimental work to be suspended. Subsequently, he began to organize researchers for online discussions.
During one of the discussions, they proposed a new type of multi-junction cascade vertical cavity surface-emitting laser structure. The research group quickly applied for an invention patent and obtained authorization from China.
Later, they also applied for an international invention patent, which has also been authorized by the United States Patent and Trademark Office and the European Patent Office.In fact, the technology of multi-junction vertical cavity surface-emitting lasers has been patented for protection by foreign teams long ago.
"This has always been the risk point I think in the process of domestication, so I have studied the layout of international patents on multi-junction vertical cavity surface-emitting lasers," said Wang Jun.
He was determined to "break the situation of this chip" and develop a structure with independent intellectual property rights.
With such a goal in mind, Wang Jun analyzed the patent protection item by item, and later decided to use a new "abnormal" structural design.
After simulating this structure, he found that the technical advantages were considerable, and subsequently obtained patent authorization in China, the United States, and Europe.By April 2021, production was finally able to resume. At this time, they began to verify the validity of the patent theory and started research on extremely high junction counts.
"The mainstream research institutions and companies at that time were still focusing on the research of 6-8 junctions. The higher the cascade junction count, the worse the epitaxial growth quality is prone to become. Therefore, we targeted the optimization of material growth for chip optimization, promoting the cascade junction count to 15," said Wang Jun.
Promoting the development of multi-junction vertical cavity surface emitting lasers towards a higher number of junctions is also in line with the research goals in the field.
Because multi-junction vertical cavity surface emitting lasers can multiply the power, which can compensate for the disadvantage of early vertical cavity surface emitting lasers that can only work in low-power scenarios.
However, what Wang Jun has been thinking about is that the power increase of multi-junction vertical cavity surface emitting lasers will not bring very obvious advantages, and there is a lack of basic physical research on this.Therefore, he and his students began to construct theoretical models again, starting from the perspective of fundamental physics, and through simulation, they found that multi-junction vertical cavity surface-emitting lasers (VCSELs) showed great advantages after efficiency improvement.
From the simulation results, the efficiency of a 20-junction VCSEL can reach 88%. "We were somewhat shocked when we saw this result, because this level can basically surpass the efficiency level of all current lasers," said Wang Jun.
Later, they carried out structural optimization design based on the early experimental foundation of 15-junction, and achieved an efficiency of 74% in the experiment.
Promoting the development of multi-junction VCSELs towards a higher number of junctions is not an easy task. As the number of junctions increases, the difficulty of epitaxial growth will also increase exponentially.
In nearly a year, the epitaxial growth verification method of the research group has been changed 10 times, and the micro-adjustment of the device structure has replaced nearly a hundred design schemes.In the end, everyone was at a loss. During this period, they didn't lack academic disputes.
The epitaxial growth personnel complained before the device design was finalized, believing that the unreasonable structural design led to the epitaxial growth issues.
The design personnel also began to feel a headache: the fluctuations in epitaxial growth caused significant performance fluctuations in the device, making it difficult to conduct experiments.
Just when the thinking was exhausted, Wang Jun pointed out the core of the problem: the way the device design personnel and epitaxial growth personnel worked independently was wrong.
Subsequently, he led the team members to reorganize, allowing the personnel from both R&D sections to work towards the same direction.Soon after, they successfully proposed a new design structure and epitaxial growth verification scheme.
In July 2023, at an international seminar, Wang Jun presented the aforementioned simulation and experimental results.
"At the meeting, several pioneering professors in the field of vertical cavity surface-emitting lasers from Germany and Japan were shocked by this result and had many discussions," he said.
He continued, "This shows that many times, people's research is more about following the trend, that is, focusing on the problems that are generally considered by everyone, and lack in-depth research on basic physics."
However, the team's research started from the perspective of basic physics and saw a different scenery, surpassing people's general cognition. "This is a kind of innovation," Wang Jun said.Recently, a related paper titled "Multi-junction cascaded vertical-cavity surface-emitting laser with a high power conversion efficiency of 74%" was published in Light: Science & Applications (IF 19.4).
Yao Xiao is the first author, and Wang Jun serves as the corresponding author [1].
One of the reviewers frankly stated: "This is indeed a new breakthrough in a stagnant field, and the paper is worth publishing."
Benefiting AI while being nurtured by AI.While the achievements of this study benefit the development of AI, the research team will also be nourished by AI in return.
In this study, they have not yet adopted AI technology, but Wang Jun believes that combining AI for the design and search of chip structures will be greatly beneficial, which can greatly improve R&D efficiency and reduce experimental costs.
At present, the research group has a large amount of full-process data of chips. Based on these database resources and AI concepts, large models can be trained to build a model capable of detecting the impact of the full process, thereby assisting in chip research and development and production.
In addition to using AI to assist in research and development, they will also carry out the following research. First, they will continue to tackle the difficult problem of higher junction number epitaxial growth, focusing on solving the problem of stress accumulation and quality degradation brought about after the accumulation of multi-junction active area layers.At the same time, the device structure design will also be optimized, based on the development of 20-layer vertical cavity surface emitting lasers (VCSELs), further breakthroughs in the efficiency of VCSELs will be achieved.
Secondly, leveraging the advantages of high-efficiency VCSELs, the application of high-efficiency, high-power multi-junction VCSELs in the communication field will be expanded.
Currently, they have achieved further advancements in power and efficiency levels on single-mode multi-junction VCSELs.
In the future, they will collaborate with external partners to jointly promote the structural design of multi-junction communication VCSELs, achieving multidimensional improvements in power, efficiency, and modulation rate.
Thirdly, the expansion to a broader range of wavelengths, especially for special wavelengths targeting quantum sensing such as atomic clocks, will greatly enhance the performance of this device, providing a larger light source space for the development of related functions.Fourthly, based on the results of this project, more application scenarios will be expanded, such as using it for intelligent devices that combine virtual and reality, as well as providing more distance and three-dimensional sensing space for low-altitude flying car terminals.
Fifthly, at the fundamental physics research level, the team will conduct research on photonic systems such as topological energy bands and bosonic systems, based on the integration of multi-junction vertical cavity surface-emitting lasers with micro-nano optical structures.
