Zhongkemi Micro's technical director: Never give up, never abandon, and tackle the "bottleneck" of MEMS chips

If it is difficult to start a chip business, then it is even more difficult to start a MEMS chip business. The difficulty mainly lies in the difficulty in transforming research and development results. Since the manufacturing process is completely different from that of integrated circuit chips, and the manufacturing processes of different types of MEMS chips are also different, chip research and development companies often find it difficult to find wafer manufacturers for mass production after making experimental chips, and even if they achieve mass production, it is difficult to gain a foothold in the extremely fragmented market. Therefore, there have been very few domestic companies engaged in MEMS chip research and development and production for a long time.

In order to break this passive situation in the international market and even achieve import substitution, Zhongkemi Microelectronics Technology Co., Ltd. came into being in 2018. In recent years, Zhongkemi has been persistent and determined to tackle tough problems, successfully solving a series of core key technical problems such as chip design, manufacturing, packaging and testing, and thus opening up the complete industry chain of optical communication MEMS chips. Its commercial products can completely replace imports of similar products, and it also has certain advantages over its foreign counterparts in terms of "cost" and "service".

After becoming the world's third and the only domestic supplier of key chips in optical communication core modules, Zhongke Miwei has further branched out and developed a series of MEMS scanning mirrors that can be used for lidar and head-up displays, as well as MEMS sensor chips. Based on this, Zhongke Miwei has currently developed a number of MEMS chip products, some of which have been mass-produced and supplied to nearly 20 major domestic optical device companies, with revenue of more than 32 million yuan in 2021.

Although the road to innovation is full of difficulties and challenges, Zhongke Miwei has never given up, and with its unyielding will, it has solved the relevant "bottleneck" technologies in MEMS chip R&D and mass production, breaking the foreign monopoly and contributing to the localization of MEMS chips. So, how did Zhongke Miwei conquer one fortress after another? What key technological breakthroughs have been achieved? What are the future development plans? To this end, Jiwei.com interviewed Xu Jing, technical director of Zhongke Miwei, and Li Wei, deputy technical director, to answer them one by one.

Jiwei.com: As the only domestic and third largest supplier of key chips for core modules of optical communications, how did Zhongke Miwei do it?

Zhongke Miwei: I think the key to Zhongke Miwei's success today is persistence. We actually made samples a long time ago, but it took a long time to reach mass production because of the immature conditions for mass production, lack of full trust from customers, and imperfect supply chain. During this period, even though we faced many difficulties, we did not give up our goals, but worked hard in the planned direction with the support of all walks of life, which led to the current situation.

Secondly, the early technical accumulation is also very important. The technical team of Zhongke Miwei comes from the Shanghai Institute of Microsystem and Information Technology of the Chinese Academy of Sciences. Since 2004, it has been working on related topics of the combination of optical communication and MEMS, including cooperative projects with leading domestic optical device companies. Although the products were not very mature at that time, we firmly believed that MEMS technology had its absolute advantages and place in the intelligent control of optical communications. Later, the Shanghai Institute of Micro-Technology also gave us strong support in the tape-out verification and test analysis platform. In short, it is only after a long period of technical accumulation that Zhongke Miwei has achieved its current success.

In addition, although the product was released relatively late, the company was able to provide targeted solutions to customer pain points and upgraded and transformed the shortcomings of products of international peers, thus gaining certain latecomer advantages.

Jiwei.com: In the process of tackling the MEMS chip of the core module of optical communications, what key technologies or important breakthroughs has Zhongke Miwei achieved?

Zhongke Miwei: In terms of key processes, we have made breakthroughs in multi-step etching, silicon-silicon bonding, step alignment, and optical mirror degumming. However, to truly achieve mass production, we need to fully open up the production process from tape-out, testing to dicing, and post-processing. Completely different from IC chip Fabless companies that have specialized chip design, tape-out, foundry, and packaging and testing, there are no ready-made specifications for each link in the optical MEMS chip production process, such as how to test, which indicators to measure, and what equipment to use for testing. Therefore, we must grope and solve the problems that arise in each link one by one.

In addition, for optical communication MEMS micromirror chips, different customers may use different packaging methods, so they cannot be protected by standard packaging like IC chips, and are usually delivered with bare dies. However, bare die delivery is very risky and requires overcoming key issues such as dust and particle contamination. In response to this, Zhongke Miwei has gradually established a complete set of quality inspection processes to ensure that the quality of micromirror chips meets customer requirements.

Jiwei.com: Domestic manufacturers usually only focus on a certain link in the semiconductor industry chain, but Zhongke Miwei has opened up the entire optical communication MEMS chip industry chain. Why did you have such a plan in the beginning? How long did it take to open up the industry chain?

Zhongke Miwei: For optical communication MEMS chips to truly achieve commercialization, the industrial chain must be opened up. Any problem in any link will result in the failure of commercialization in the end. Based on the previous accumulation and "the right time, the right place, and the right people", Zhongke Miwei has reached the current relatively good state. Our first product that opened up the entire process should have started in 2013. At that time, it was first taped out domestically for verification, then taped out abroad for mass production, and finally returned to domestic mass production. But the real independent mass production should be in 2019, and the overall process took about 6 years.

Jiwei.com: Compared with its international competitors, what are the outstanding advantages of Zhongke Miwei's optical communication MEMS chips such as VOA (tunable optical attenuator), OSW (optical switch), and TOF (tunable optical filter)?

Zhongke Miwei: Overall, I think our levels are comparable, and we may have an advantage in some individual indicators. First, the design principles of these chips are basically similar, and second, we will optimize them according to customer application needs. In addition, on the basis of import substitution products, we can also develop new product models according to customer needs. For example, some customers want the product to have a larger angle slope and a larger mirror. We will comprehensively analyze market demand and develop new products. In comparison, we have closer ties with domestic customers and can provide more convenient and higher-quality customized services.

Jiwei.com: Zhongke Miwei's products are both domestic substitutes and foreign counterparts. So, from the customer's perspective, is their motivation for replacement more reflected in the flexibility of the company's products or the cost?

Zhongke Miwei: Objectively speaking, it is both. On the one hand, it is the international environment. Now domestic customers are willing to try domestic chips, otherwise they can only be "choked" by foreign suppliers. On the other hand, because Zhongke Miwei's product industry chain is all in China, it must have certain advantages in cost. In general, unlike other industries, optical communications are relatively more concerned about performance. Sometimes for safety reasons, customers may not really care about price, but put performance first. However, if you want to occupy a larger market share, you must show advantages in cost.

Zhongkemiwei MEMS micromirror products (Photo provided by the interviewee)

Jiwei.com: In the field of scanning mirror products, what are the advantages of Zhongke Miwei's hybrid solid-state MEMS lidar scanning mirror compared to traditional mechanical radars?

Zhongke Miwei: First, compared with traditional mechanical radars that require motors, bearings and other scanning components, hybrid solid-state MEMS lidar scanning components are smaller and lighter. Secondly, hybrid solid-state radars can be mass-produced to reduce costs and have a longer lifespan. Because of this, it can gradually move from a very niche surveying and mapping market to a relatively large consumer market.

Traditional mechanical radars are too large to be installed on cars. Hybrid solid-state MEMS lidars or pure solid-state radars may not be as good as traditional mechanical radars at present, but they are being developed and upgraded. If the technology is more optimized and stable, and the size is smaller and easier to use, it will become the mainstream trend in the future. However, in some areas that are not restricted by size, lifespan, etc., traditional mechanical radars still have the advantage of high precision, so they will coexist with hybrid solid-state and pure solid-state radars in the future.

Jiwei.com: What is the development progress of Zhongke Miwei's MEMS scanning mirror chip? In what fields will it be used in the future?

Zhongkemiwei: The larger the scanning mirror, the more difficult the design and manufacturing process is, so the development of Zhongkemiwei's scanning mirror chip is also gradually promoted according to the difficulty. At present, the first product we developed, the two-millimeter diameter MEMS scanning mirror, is relatively mature, and the horizontal and vertical scanning angles of the light beam can reach 60 degrees and 30 degrees respectively. In addition, the first version of the four-millimeter diameter scanning mirror product has completed the first tape-out, which can basically meet certain indicators, but there are still areas that need improvement. After the optimization is completed, it is expected to be finally applied. In short, Zhongkemiwei's MEMS scanning mirror products use electrostatic drive, mature technology, and are easy to mass produce. At present, they are mainly positioned in the application fields of laser radars for automotive assisted driving, service robots, intelligent transportation, and other needs.