Analysis of EpiWafer's Advantages and Existing Problems

It has been reported recently that NexWafe GmbH, a subsidiary of Fraunhofer ISE in Germany, is developing monocrystalline silicon wafers through the EpiWafer process, without melting, crystallization, ingot casting or cutting, with a thickness as low as 120um, and has achieved 23-24% cell efficiency. EpiWafer not only significantly reduces production costs, but also closely matches n-type single crystal technology, and will be the best material for future TOPCon and heterojunction technologies.

What is EpiWafer, which has won the favor of many capitals and is about to enter mass production?

One of the reasons that EpiWafer has won the favor of many capitals is that NexWafe told an unimaginable story: instead of producing silicon wafers by the traditional routine of melting, crystal growth and slicing, they ingeniously produced silicon wafers directly from chlorosilane, the raw material of polysilicon.

The traditional silicon wafer production process is to first purify industrial silicon into high-purity polysilicon, then melt and pull crystals to produce silicon ingots, and then cut the silicon ingots into square wafers.

From industrial silicon to high-purity polysilicon, it is not the physical process that you imagine by melting and removing impurities and then purifying, but converting industrial silicon into chlorosilane, and then reducing it to polysilicon. The chlorosilanes here are not only explosive, but also highly toxic gases, which is why the polysilicon factory you see is like a chemical factory.

Pictured: Polysilicon production plant

In addition, the subsequent processes of melting, crystal growth, ingot formation, and slicing need to consume a large amount of electric energy, and a large amount of scraps will be generated when the cylindrical silicon rod is cut into square silicon wafers. together become low-value silica mud). On April 5, Yunnan announced the cancellation of electricity price discounts in accordance with the requirements of the National Development and Reform Commission, which directly led to a 5.51% drop in LONGi’s stock price on April 6.

Previously, 1366 Technologies of the United States told a story of not playing cards according to the routine. The idea of 1366 is to crystallize the molten polysilicon directly in the die of the silicon wafer and form it into a silicon wafer in one step, eliminating redundant steps such as pulling rods for crystal growth and cutting, and greatly reduces the loss of raw materials. However, this process has not been promoted due to the high cost and the fact that single crystal has formed an absolute substitute for polycrystalline.

In Fraunhofer ISE's view, the process of 1366 is too complicated, and it can be further simplified from industrial silicon to silicon wafers, just like reading Du Mu's "There are rains during the Qingming Festival, and pedestrians on the road want to break their souls. Can I ask where is the restaurant? The shepherd boy pointed out. Like "Xinghua Village", it is too long-winded and can be simplified as "Qingming rain, people die. Where is the wine? Xinghua Village"

The idea of Fraunhofer ISE is to directly produce standard thickness individual wafers from chlorosilanes produced from industrial silicon.

Since there is no need to produce polysilicon material in the manufacturing value chain, and no need to melt polysilicon material for crystal pulling, and no need for subsequent slicing and trimming, it not only saves energy, reduces waste, but also simplifies the process. This value chain occurs With the fundamental change, the production cost will definitely be significantly lower compared to the traditional wafer fabrication process.

This is EpiWafer, a silicon wafer made with an ingenious process.

Pictured: Epitaxial wafer (right) separated from reusable seed wafer (left)

Ideas can be wild, but they must be tested in practice. Can Fraunhofer ISE's idea produce solar wafers?

Don't worry, Epiwafer technology is not the invention of Fraunhofer ISE. In the electronics industry, the Epiwafer process is already a mature method of manufacturing wafers.

Epiwafer translated into Chinese is "silicon epitaxial wafer". Since the late 1950s, silicon epitaxial wafers have been successfully used in the manufacture of high-frequency and high-power transistors. In order to meet the needs of various semiconductor devices, various silicon epitaxy technologies have been produced accordingly. From the perspective of device manufacturing, it can be divided into positive epitaxy and reverse epitaxy, and from the chemical composition, it can be divided into homoepitaxy and heteroepitaxy.

Methods for preparing silicon epitaxial wafers include vapor phase epitaxy, liquid phase epitaxy, molecular beam epitaxy, and the like. Among them, chemical vapor deposition (CVD)-based vapor phase epitaxy is the mainstream method for producing silicon epitaxial wafers. The commonly used gas sources are SiCl4, SiHCl3, SiH2Cl2 and SiH44, and SiCl4 source is currently the most widely used.

In EpiWafer technology, silicon wafers can be fabricated by epitaxial deposition of chlorosilane gases as thick crystalline silicon layers, which are then separated after growth to produce individual wafers of standard thickness, either n-type or p-type doped silicon monoliths. Crystal wafers can be produced.

How come such a good technology has not been used in solar silicon? It is very simple, efficient and cost. Since EpiWafer has been used in the semiconductor electronics industry, think about how many times the cost of the semiconductor electronics industry is that of solar energy.

In 2012, Japan's SUMCO, the world's second largest semiconductor wafer factory, shut down the production of its 12-inch mainstream silicon epitaxial wafers because of high costs. SUMCO also has solar silicon wafers, but also withdrew because the cost was too high. Although SUMCO masters the production technology of silicon epitaxial wafers, it is not used for the production of solar wafers.

Technically speaking, the silicon wafers used in the solar industry still need to solve a problem of power generation, which is still different from the electronics industry, and the number of silicon wafers used in the solar industry is several orders of magnitude multiples of the semiconductor industry. one question.

That's what Fraunhofer ISE is trying to solve.