The Unknown Victory in The Semiconductor Field: Another Link To Achieve Domestic Substitution

On December 29, 2020, a lawsuit that was dubbed by the media as "US semiconductor companies stealing Chinese technology" started in Hong Kong, China.

Yes, you read that right, the United States "piracy" Chinese technology, and it is semiconductor technology.

The initiator of the lawsuit is Keystone, the leader of domestic ion implanters, and the defendant is AIBT in the United States.

The latter is not some little-known pheasant company, but the current industry third in the world.

The cause of the lawsuit, according to public reports, is that AIBT purchased advanced ion implanter technology from Keystone, and agreed to purchase equipment components from Keystone within 3 years.

However, AIBT directly surpassed Keystone and signed orders with upstream parts suppliers. Keystone formally filed a lawsuit against AIBT after multiple active communications with AIBT regarding this breach of contract were unsuccessful.

From the official content, the media reports are slightly wrong.

Because AIBT has been acquired by the Chinese Taiwanese company Hanmin Group, and its headquarters has also moved to Taiwan, it is not an "American company".

Also, by nature, the term "plagiarism" is inappropriate. Of course, infringement of intellectual copyright is a certainty.

The new generation of ion implantation equipment just released by AIBT is also facing risks such as production stoppage.

Some people may ask, is the domestic ion implanter already so strong?

But more people may want to ask, what is an ion implanter?

1

Chip manufacturing can be divided into 10 key steps, namely oxidation, gluing, photolithography, etching, ion implantation, physical vapor deposition, chemical vapor deposition, polishing, wafer inspection, and cleaning.

Ion implantation is an indispensable link, which can be simply understood as doping some impurities into a pure single crystal silicon wafer.

Some people may be very strange. Why do you want to dope the silicon wafers after you have worked so hard to make 99.999999999% pure silicon wafers from sand?

To answer this question, we must first clarify the concept of semiconductor.

Although in the current public context, semiconductors are basically equivalent to integrated circuits, and the "semiconductor industry" is equivalent to the chip industry, but in the original sense, semiconductors actually refer to a large class of substances, that is, the electrical conductivity at room temperature. material between conductors and insulators.

As we all know, those substances around us that are usually non-conductive, such as rubber, glass, ceramics, etc., are called insulators; while metal substances such as gold, silver, copper, and iron are conductive and are called conductors, which can be simply The material between a conductor and an insulator is called a semiconductor.

Semiconductors are useful because of this intermediate property of conducting and non-conducting.

But the problem is that at room temperature, pure silicon has a very small conductivity, close to that of an insulator. At the same time, a whole piece of silicon wafer cannot control its current direction, so impurities must be doped in specific places to change its conductivity and conductivity.

For a simple analogy, the lithography machine "draws" a circuit diagram on the wafer, the etching machine "carves" the drawn circuit diagram, and the doping of impurities makes these engraved patterns conductive. the "circuit".

In the past, the diffusion process was usually used to incorporate impurities into silicon wafers, that is, the required impurities and silicon wafers were placed in a high temperature environment, and the principle of particles moving from high concentration to low concentration was used to make impurities naturally diffuse to in silicon.

However, as the chip manufacturing process continues to rise, the integration level is getting higher and higher, and the size is getting smaller and smaller, the concentration, depth and distribution range of impurities diffused in the silicon wafer need to be controlled more precisely, so small-sized chips are now using Ion implantation process.

Compared with diffusion, ion implantation has many advantages, including precise control of doping concentration and doping depth, arbitrary impurity distribution, good uniformity and repeatability of impurity concentration, low doping temperature, less contamination, No solid solubility limit.

It can be said that the technical level of ion implantation determines the performance of the device on the chip after it is formed.

In 2020, the global integrated circuit ion implanter market size is only about 1.8 billion US dollars, accounting for only about 3% of the wafer manufacturing process equipment market, but it is also known as the four major manufacturers of lithography machines, etching machines and coating machines. Core equipment, precisely because of its importance.

As mentioned above, ion implantation has many advantages, but its disadvantages are also very significant. The first is to bombard the silicon wafer with high-energy impurity ions. If the control is not good, the crystal lattice of the silicon wafer may be damaged or even lose its crystal characteristics. Another disadvantage is more significant for our country, that is, crystal implantation requires very complicated and expensive machines, and the technical threshold is very high.

In general, the ion implanter consists of 5 parts: ion source, ion introduction and mass analyzer, acceleration tube, scanning system and process chamber.

Although it looks like there are only 5 "district" parts, the ion source alone includes so many things below, and the complexity of the entire system can be imagined.

The complexity of the system is second, and the internal operating procedures are also very cumbersome, which contains a lot of difficulties. For example, the wafer is easily broken when transferred between different machines, the beam stability is insufficient, the life of the ion source is insufficient, the cooling cannot reach the target, and the wafer may be damaged. Breakdown, etc., have extremely high requirements on the technical indicators of the machine.

In the past, this field was under the absolute monopoly of the United States.

Among them, one Applied Materials company accounted for 70% of the market share, followed by Axcelis, which accounted for nearly 20% of the market share.

According to different uses, ion implanters are divided into low-energy large-beam ion implanters, high-energy ion implanters, and medium-low beam ion implanters according to different energy and beam current indicators.

As the chip manufacturing process is getting smaller and smaller, the depth requirements of ion implantation are correspondingly reduced, and the low-energy large-beam ion machine used for shallow doping has become more and more mainstream.

According to Gartner, low-energy high-beam ion implanters account for 61% of the total ion implanter market, while low- and medium-beam ion implanters and high-energy ion implanters account for 20% and 18%, respectively.

In this segment, the degree of monopoly is higher. At present, the largest low-energy large-beam ion implanter is mainly controlled by three leading companies, including Applied Materials with a market share of 40% and Axcelis with a market share of 32%.

There is also the AIBT we mentioned at the beginning, with a market share of 25%.

Why is such a well-known giant in the industry caught in the turmoil of "piracy" Chinese technology?

2

Interestingly, the two companies in this round of lawsuits actually have a lot of origins.

The five founders of Keystone, headed by Chen Jiong, are actually the founding team of AIBT, and they are all world-class ion implantation equipment experts. Chen Jiong is also the chief technology officer of AIBT, and his technical strength is quite solid.

After the establishment of the company in 1999, it was under the leadership of Chen Jiong that AIBT successfully developed a second-generation large-beam ion implanter, which entered the 28nm key process of integrated circuit manufacturers.

In 2009, China's "02 Special" research team found Chen Jiong in Silicon Valley.

Regarding the 02 special, our article has given a detailed introduction before.

To put it simply, the state put forward a plan in 2006 to support the development of 16 major special projects in the high-tech field based on the next 15 years. .

The "very large-scale integrated circuit manufacturing technology and complete set of processes" project on chip manufacturing ranks second, so it is called "02 special" in the industry, including lithography machines, etching machines, ion implantation machines, etc. in.

The country's far-sighted and strong support, coupled with the domestic economic construction in full swing, provided unlimited imagination for the prospects of the semiconductor industry. Chen Jiong's team immediately decided to return to China for development. "Everyone starts from scratch and builds a high-end equipment brand that belongs to China."

Obviously, technically, it's not that hard for them. From the results, this is indeed the case, as evidenced by the ability to sell technology in reverse.

But in the beginning, they got off to a rough start.

Under the circumstance that the international situation is not yet severe, domestic and foreign semiconductor manufacturers prefer foreign equipment. In such a technology-intensive industry, the first-mover advantage is extremely difficult to shake.

Therefore, after the establishment of Keystone, it can only start with the parts of the ion implanter, and abandoned the integrated circuit field, which the founding team was good at, and turned to the photovoltaic ion implanter market, which has the same technology, but the application is still blank. 1 in the world.

After the great success of photovoltaic Dongfeng, it coincided with the rise of the "domestic substitution" in the domestic integrated circuit field, and Keystone turned back to its "old business" in a timely manner.

According to the data of the Prospective Industry Research Institute in 2020, the domestic ion implantation equipment market size is 4.45 billion yuan, of which the scale of the integrated circuit field reaches 4.32 billion yuan, and the vast majority of the shares belong to the three giants of Applied Materials, Axcelis and AIBT.

However, after that, domestic equipment manufacturers quickly broke through.

The first is the "national team". As a subsidiary of the central enterprise CEC, which is responsible for tackling key problems in ion implanters, China Science and Technology announced in 2021 that it has successfully achieved the localization of a full spectrum of ion implanters, and can provide global chip manufacturers with one-stop solutions for ion implanters. Scheme, the process segment covers to 28nm.

In fact, earlier in 2018, its 28nm products had already entered the SMIC production line. On June 15, 2020, its 12-inch medium beam ion implanter was successfully moved into the large-scale integrated circuit production line.

In comparison, Keystone's technology and industrialization level, as a representative of private enterprises, seems to be better.

In terms of business, Keystone is the only company in the world that covers all fields of ion implantation equipment, including ion implanters for photovoltaics, chips, AMOLED, IGBT, etc. They all have business.

In 2020, Keystone won several orders for different types of 12-inch low-energy large-beam ion implanters and high-energy ion implanters, and the integrated circuit ion implanter products have entered the stage of customer verification.

In 2021, the first low-energy large-beam ion implanter and high-energy ion implanter independently developed by Keystone will pass the verification and acceptance of mainstream 12-inch wafer fabs in China.

Judging from its published technical indicators, the characteristic line width, implantation energy, implantation dose, and maximum productivity covered by its low-energy large-beam ion implanter are basically the same as those of similar foreign mainstream products, and the implanted beam can reach the mainstream of foreign countries. product nearly 2 times.

In addition, in July this year, its parent company, Wanye Enterprise, revealed in response to investor questions that Keystone's 7nm ion implantation platform has passed customer verification and received acceptance.

There is also news that its integrated circuit ion implanter has broken through the 3nm process, and its main parameters are better than mainstream similar foreign products.

For the future, according to the plan of "Made in China 2025", the core basic components and key basic materials of semiconductors should achieve 70% independent guarantee in 2025, which is a huge space.

Although the current domestic technology level is still relatively backward, the huge cake is bound to become the driving force for local enterprises to break through.