Third Generation Semiconductors—Fabrication Process and Analysis of Silicon Carbide Materials

SiC power electronics is the driving force that will accelerate the arrival of the electric vehicle era. By replacing the current Si IGBT with the SiC MOSFET, it will be possible not only to reduce switching loss by more than 80% but also to dramatically reduce energy losses during power transfers. At the same time, chip modules can be reduced to 1/10 of their original size. This will help improve the performance of electric vehicles and realize the goals of extending cruising ranges and shortening charging times. The Yole Developpement market research agency estimates that, by 2025, the market size of SiC in the field of electric vehicles and charging poles will reach 1.778 billion USD, which will be about 70% of the total SiC applications market.

Driven by the green energy and carbon reduction policies and subsidies of various countries, the global electric vehicle market is booming. In the past two years, major auto manufacturers have successively launched 800V high voltage models, sparking an explosion in demand for SiC substrate materials. According to research done by TrendForce, by 2025, the demand for 6-inch SiC wafers in the global electric vehicle market will reach 1.69 million. However, the total output of SiC wafers annually is only about 4 hundred thousand to 6 hundred thousand pieces around the world currently, with a mainstream wafer size of 6 inches. This quantity falls far short of the demand for downstream substrates in the industry chain. In truth, crystal growth is the most difficult part of producing SiC substrates. The existing process is not only complex but also slow in terms of crystal growth. As such, it is extremely difficult to manufacture these wafers in large quantities. There are only a handful of companies in the market with SiC wafer mass production capabilities. Among them, Wolfspeed (formerly Cree) is the most prominent supplier. This one company’s shipments account for about 50% of the total market capacity. It is followed by ROHM and II-VI, who together account for about 35% of the total market.

Due to the difficulty of SiC wafer manufacturing and the serious shortage in market supply, SiC components are very expensive. Among them, the substrate part accounts for about 50% of the total cost of a chip. Recently, in lieu of the huge business opportunities in the global automotive SiC power components market, many well-known SiC wafer manufacturers have begun to expand their production capacities through acquisitions or strategic cooperation with wafer suppliers, etc. to obtain a stable supply of wafers. Whoever can grasp the key technologies for mass production of upstream SiC substrates in a timely manner can also seize a leading position in the future era of third generation semiconductors.

In this issue of the “New Technology Channel | Collaboration Column”, MA-tek has specially invited Professor Quanpu Liu, an expert in the field of semiconductor materials, to comprehensively introduce the current SiC waver manufacturing methods and challenges as well as provide an overview of the technological developments in the industry surrounding this key raw material to our readers.

Director of R&D Center & Marketing Division, Chris Chen, 2022/05/20

Third Generation Semiconductors—Fabrication Process and Analysis of Silicon Carbide Materials

Professor Chuan-Pu Liu

Ph.D. students： Yi-CHANG Li

Department of Materials Science and Engineering, National Cheng Kung University

-

According to the analysis of the characteristics of new generation semiconductors (Figure 1), silicon carbide is a wide bandgap material with an excellent breakdown field and an intrinsic carrier concentration much lower than that of silicon. It remains stable in high voltage operating environments and does not easily generate leakage currents. These qualities make it very suitable for use in power components such as power converters. Furthermore, silicon carbide has two to three times the thermal conductivity of silicon. This means that, compared to silicon, the heat generated by a component using SiC can be more effectively carried away of the system. As such, it is quite suitable for bearing gallium nitride (GaN). GaN on SiC can be used for high frequency or RF components.

Figure 1.　Summary of Key Characteristics of Semiconductors for Power and RF Components [1]

With Third Generation Semiconductor-Silicon Carbide emerging as ideal for new applications, people all around the world are actively getting involved in related industry chains. For example, the upstream and downstream industry chain can be simply divided into SiC wafers, epitaxy and power and RF components. Relevant benchmark manufacturers are listed in Figure 2. The number of manufacturers expected to enter the market in the future is also gradually increasing.

Figure 2.　Diagram of the Correlation Between the SiC Front-End Applications Industry Chain and Key Benchmark Manufacturers [2]

At present, the cost of the growth of silicon carbide accounts for about 38% of the total cost of a component [3]. The three mainstream production methods are High Temperature Chemical Vapor Deposition (HTCVD), Solution Growth and Physical Vapor Transport (PVT). The relevant characteristics of each method and the manufacturers that use it are listed in Figure 3.

Figure 3.　Current Mainstream Silicon Carbide Growth Methods [3]

Chemical vapor deposition, as shown in Figure 4, passes silicon-based gas (SiH4) and carbon-based gas (C2H4) into the HTCVD reactor. The reaction of the high temperature gases in the reactor produces crystal clusters. The crystal clusters will then sublimate due to the high temperature. It is then deposited onto the seed crystal in the low temperature region to form a single crystal wafer. Although this method does not require high purity of silicon carbide powder as a raw material, instead, using gas directly as a source of silicon carbide, this method involves many kinetically unstable factors. As such, it is difficult to control the stability and consistency of the overall process. There are still many problems to solve before it can enter the mass production stage [4].

Figure 4.　Schematic Diagram of the High Temperature Chemical Vapor Deposition Method [5]

The liquid-based method is the traditional single crystal growth technique, which has advantages in terms of crystal size and production speed. This method is to immerse a silicon carbide seed crystal in a silicon melt containing a carbon-based solvent. When the cooler seed crystal is pulled from the high-temperature silicon solution, silicon carbide will grow on the seed crystal, forming a crystal ingot, as shown in Figure 5. However, problems such as solid-liquid interface control, the removal of the solvent that for increasing carbon solid content, and the uniformity of large-scale components are still to be solved. At present, they are still only in the research stage. There are still many challenges to overcome for commercialization [6].

Figure 5.　Schematic Diagram of the Solution Growth Method[6]

The physical vapor transport method is currently the most commonly used in the silicon carbide semiconductor industry. It has the fastest output among the methods currently available. It also has the advantages of scalable wafer sizes and a relatively low equipment cost. As shown in Figure 6, the manufacturing process uses high purity of polycrystalline silicon carbide powder as the source, heating it until it sublimates to produce gas (Si, C, Si2C and SiC2). The gas is deposited on the Monocrystalline SiC Seed Crystal in the low temperature region to obtain high-purity single crystals. This method is currently the fastest process of silicon carbide production in the industry and produces SiC boule with the largest area and greatest length. The biggest difficulty in this method lies in the need to balance the Si, C, Si2C and SiC2 gases in a closed system to ensure the consistency and uniformity of the wafer’s composition ratio. The temperature and growth rate are also important factors. If it grows too fast, it is easy to get defects. It is also necessary to ensure that the wafers maintain a low dislocation density. If the wafer has too many defects, those defects can become nucleation sites, and often the multiphase crystal phase of silicon carbide will form and grow at these sites. This phenomenon causes the wafer to be unable to maintain a single crystal phase. When the defects are too numerous, they can even cause the wafer to crack [7].

Figure 6.　Schematic Diagram of the Physical Vapor Transport Method [8]

Impurities in the SiC powder can become the origins of defects. Therefore, extremely high purity is required to ensure that wafers can maintain low dislocation densities. In addition to controlling wafer growth parameters, the selection of upstream raw material of silicon carbide powder is crucial. As for the synthesis of silicon carbide powder, in addition to the appropriate synthesis reactions, the purity of the synthesis should be maintained as much as possible to reduce the influence of impurities. Below is an introduction to various silicon carbide powder synthesis methods.

The Acheson Process was the first method that enabled the high-volume manufacturing of silicon carbide powder. Its principles are shown in Figure 7. Place the silicon oxide powder and organic, carbon-based reaction mixture in the furnace and pass a current through the graphite electrode running through the furnace body, causing a carbothermic reaction. The reaction formula is [9]　SiO2+3C→SiC+2CO. In turn, silicon carbide is produced on the graphite electrode.

Figure 7.　A Simple Diagram of the Acheson Process [11]

However, because there are often many metal impurities in the furnace wall and in the atmosphere, there is often a problem with oxygen pollution in the reaction [10]. The silicon carbide produced this way can have a purity of up to 99.8%, which falls considerably short of what the semiconductor industry demands for crystal growth. Therefore, the material produced via this process is mostly used in the cutting and grinding industry [10].

Chemical vapor deposition is often the method of choice for making high purity materials. In addition to controlling the many gas dynamics parameters involved, this process also requires the careful selection of precursors according to reaction methods such as thermolysis, hydrolysis, oxidation, reduction and carboration, etc.

At present, SiH4 and C2H2 are the main precursors of the CVD process. The process structure is shown in Figure 8. With the aid of a vacuum pump, the internal environment can maintain a high vacuum to avoid the influence of external impurities. The gas flow controller is used to make SiH4 and C2H2 flow into the reactor at an appropriate rate. N2 is used as the carrier gas. It is premixed with SiH4 and helps the reaction gas flow smoothly into the reaction furnace to produce the reaction. The reaction formula is 2SiH4 (g) + C2H2 (g)→2SiC(s) + 5H2(g). After the reaction, the silicon carbide powder produced falls into the powder collection barrel to form the finished product.

Figure 8. Simple Schematic Diagram of the Vapor Deposition Method (1. Gas Flow Controller; 2. Reaction Furnace; 3. Thermocouple; 4. Pressure Detector; 5. Powder Collection Barrel; 6. Vacuum Pump) [12]

Under the right operating parameters, high purity, nano-scale silicon carbide powder will be obtained, as shown in Figure 9(a). The primary size of the powder particles is about 100-200nm, and the particle size after agglomeration is only 500nm. This means that the particle size of the powder produced is fairly uniform and consistent. However, when the three parameters of gas, temperature and pressure are not optimally adjusted (See Figure 9(b) and (c)), it is easy for incompletely reacted intermediates to be produced. It can also easily cause the finished product to have an excessive particle size distribution (100nm~500nm), potentially even resulting in a post-agglomeration monomer particle size of more than 1000nm. In addition, if not properly regulated, the intermediates will contain O-H, C-H, Si-H, C-H and other bonds to a large extent, resulting in a powder that is not pure silicon carbide. It will contain residual un-reacted silicon-based and carbon-based organic compounds.

Although the surface of the material produced by CVD easily meets the quality standard required by the semiconductor industry, the process requires very precise control of all the parameters of the reaction system. This means that it is both time-consuming and has limited production capacity. Before CVD can be used to mass produce silicon carbide, there needs to be further research and exploration of ways to balance the complex reactions and gas conditions [12].

The sol-gel method can also be used for the production of silicon carbide powder. The advantage of this method is that silicon carbide materials can be produced at relatively low temperatures. In other words, unlike in other processes, it does not require the temperature of the reactant to be raised above 2000°C. Therefore, this method has a much lower energy consumption. Also, because the precursor reacts uniformly in the solution, the uniformity of the final product is also quite high.

The sol-gel process consists mainly of two stages—hydrolysis and polycondensation. Begin by preparing the silicon alkoxide, and put it into a solution containing water and a suitable carbon-based solvent to facilitate hydrolysis to form functional silanol groups (SiOH). These functional groups will continue to condense and form siloxane bonds (SiOSi). That’s when you get silica gel. Finally, you remove the solution to obtain xerogel, which you heat to above 1600°C to obtain silicon carbide [10].

There is literature on the use of TEOS (tetraethyl orthosilicate) and adding different proportions of phenolic resin to obtain silicon carbide powder [13] then conducting crystallographic analysis, as shown in Figure 10. It was found that, when the proportion of carbon is too low, the heat-treated precursor will fail to form silicon carbide. However, if the proportion of carbon is increased, it will also increase the production of non-silicon carbide materials. Furthermore, it can be found via transmission electron microscope (Figure 11) that there are many incompletely reacted intermediates (carbon residue and silicon oxide) left in the silicon carbide powder. However, if the residual carbon and silicon oxide are removed using hydrofluoric acid and atmospheric sintering, pure silicon carbide powder can be obtained.

The proportion of carbon also affects the process’ yield. According to the experiments conducted, when the carbon ratio is higher, the yield will be higher, but it will also be easy for there to be the problem of incompletely reacted intermediates. Therefore, although this method is capable of growing silicon carbide with high uniformity, it is easy to get residual intermediates that then affect the purity. So, in consideration of these reactions, the design of the process still needs to be refined. In addition, the silicon-based alkoxide liquid used in the reaction often contains metal impurities, which can seriously affect the quality of the silicon carbide produced. In order to meet the raw material crystal growth standards of the semiconductor industry, we still need better solutions for purity control.

Mechanical alloying is also a potential production method. The process involves putting the appropriate proportion of carbon powder and silicon powder into a high energy ball mill (HEBM) and adding zirconium beads to induce high energy rotation and vibration. The collision force of the zirconium beads is used to destabilize the surface of the carbon and silicon powders, resulting in localized combustion synthesis (CS).

As shown in Figure 12, the HEBM produces combustion synthesis after a certain length of time, causing the temperature to rise sharply to 1750K. This begins the production of silicon carbide powder. As shown in Figure 13, silicon carbide powder can be produced quickly using the HEBM. Additionally, when the action time is longer, the residual silicon powder and carbon powder are more likely to react completely. This process also does not require the raising of the ambient temperature to the reaction temperature. However, this method uses the collision force generated by zirconium beads to ignite a reaction, so impurities are easily incorporated during the production process. In addition, this process is suitable only for small-volume manufacturing. It is not suitable for mass production. With the basic requirement for high purity in mind, the feasibility of this method in subsequent high-volume production still needs to be improved [14].

Figure 12. HEBM Time vs. Temperature Diagram [14]

Figure 13. XRD of Silicon Carbide Powder Produced by HEBM at Different Times (1 Silicon Crystal; 2 Silicon Carbide Crystal) [14]

Finally, there is the solid combustion synthesis method. The costs for this method are more controllable, and its yield can meet industrial demands. It involves uniformly mixing silicon powder and carbon powder in the appropriate proportions and placing them in a vacuum furnace. Pump to high vacuum to prevent the influence of external impurities and raise the ambient temperature to above 2100°C to obtain SiC powder [15, 16].

Although this method is much simpler than those mentioned above, it faces other challenges. Because the melting points of carbon and silicon carbide exceed 2500°C, the silicon melts first into a liquid state. This means that the size of the carbon particles limits the depth to which the liquid silicon reacts. When silicon carbide is formed on the surface of liquid silicon and solid carbon, the liquid silicon cannot easily enter the core of the carbon particles for a complete reaction. This phenomenon easily causes there to be residual silicon and carbon, resulting in a decrease in purity.

In addition, due to the fluidity of liquid silicon, during high temperature heat treatment, gravity easily makes the liquid flow to the lowest points, resulting in a decrease in overall uniformity. You can see in Figure 14 that there is no residual silicon on the upper part of the SiC bulk. However, there are signals from un-reacted silicon crystals in the middle and bottom. As a result, its yield rate is very low. This problem also seriously affects the stability of its silicon carbide quality [16]. In this method, the overall uniformity, time and other key conditions of the reaction must be closely controlled.

Figure 14. (a) Top (b) Middle and (c) Bottom XRD Image of Bulk Silicon Carbide Generated via SCS [16]

Silicon carbide is a polytype material with more than 100 known crystal structures, and different crystal structures have different properties. 3C-SiC, 4H-SiC and 6H-SiC are currently the most well-known and widely used in the semiconductor industry. 3C-SiC or β-SiC is the only cubic crystal structure of silicon carbide. Crystal phases with non-cubic crystal structures are referred to collectively as α-SiC. 4H-SiC and 6H-SiC are among the crystals known to have hexagonal crystal structures [17]. Determining the crystal phases contained in a finished product allows for the timely control of the basic properties of the material. Generally, XRD is used to analyze the powder and wafer crystal phases of three silicon carbide materials (3C-SiC, 4H-SiC and 6H-SiC) to determine wafer crystal orientations and defects, etc.

In addition to the crystal phase, the silicon carbide wafer and silicon carbide powder production processes need to take purity into consideration. Some finished products have to be as much as 99.9999% pure (6N). Otherwise, it is easy for there to be problems such as defects, the presence of polymorphic phases in finished products, and even wafer breakages. Therefore, qualitative—and even quantitative—material analysis is a necessity. Based on foreign silicon carbide powder specifications, the purity needs to be above 5N5 to 6N to ensure that the wafers grown can maintain a low defect density.

At present, silicon carbide wafers can be divided into semi-insulating and N-type according to doping type and resistivity. Among them, the production of semi-insulating single crystal silicon carbide wafers in particular needs the impurity concentration that may be generated in the manufacturing system to be minimized. These impurities may come from graphite materials, silicon or silica raw materials, process gases and other sources. The main elements of impurities include nitrogen (N), boron (B) and other elements with low excitation energies. The N comes mainly from the N₂ in the air, absorption of silicon or silicon dioxide feedstock via graphite (including graphite crucibles and graphite liners in heating chambers), and the air introduced in the process. The B comes mainly from impurities in graphite materials and silicon or silicon dioxide raw materials themselves. Therefore, in order to produce high purity, stable-quality silicon carbide powders and wafers, the development and reliability of purity detection and analysis technology is extremely critical. ICP-MS, SIMS, GDMS and other methods are all currently being used.

The synthesis and crystallization processes cause slight changes in an area, and heterostructures can form in the inner region. These powders and wafer materials can be microscopically analyzed using transmission electron microscopy (TEM) and the focused ion beam (FIB). Combining TEM microscopic technology with the FIB’s micro-positioning function to cut at the nano level and conducting energy spectrum analysis, such as EELS and EDS, enables the analysis of the crystal structure, related interfaces (such as SiOx/SiC), defect distribution, and chemical composition at specific positions on the silicon carbide so as to clarify the influence of process parameters.

The demand for the semiconductor material silicon carbide is on the rise as key components with new specifications are developed for 5G, electric vehicles, high power applications (e.g., fast-charging components), radars and more. Improvements in device characteristics and yields come from the fundamental properties of the high quality materials available upstream. On wafer growth, in order to obtain high quality wafers, it is not enough to merely have high purity seeds as growth guides. It is necessary to maintain precise control of the subsequent growth conditions, including all the parameters in the crystal growth furnace. By combining material analysis and its ability to monitor the wafer’s dislocation density, internal impurities and more with crystal phase control, etc., it becomes possible to make the appropriate corrections to crystal growth in a timely manner. As for the synthesis of silicon carbide, in order to obtain high purity silicon carbide powder, it is necessary to pay attention to the selection of process parameters and environment, etc. in addition to the selection of carbon and silicon sources. It is also important to follow up with material analysis to understand the powder’s properties, the composition of its internal impurities, and the un-reacted residue, etc. to monitor powder quality at all times.

Future development is moving gradually towards high voltage and high frequency applications. In order to meet the huge market demand, silicon carbide material manufacturing technology is being improved. Equipment is being effectively connected in series and integrated into the overall industry chain. This will affect the long-term development and key output of subsequent silicon carbide components.

Reference:

[1] 高頻用化合物半導體材料需求趨勢, 張致吉, 2021

[2] 碳化矽(SiC)之新興市場應用與發展, 鄭華琦, 2021

[3] 碳化矽(SiC)長晶與晶圓薄化設備的技術發展趨勢, 張雯琪, 2021

[4] M. A. Fraga, M. Bosi, and M. Negri, "Silicon Carbide in Microsystem Technology — Thin Film Versus Bulk Material," in Advanced Silicon Carbide Devices and Processing, 2015

[5] A. Ellison et al., "SiC Crystal Growth by HTCVD," Materials Science Forum, vol. 457-460, pp. 9-14, 2004.

[6] 蕭達慶, "功率元件用新半導體基版材料發展現況," 工研院材化所-工業材料雜誌, no. 380, 2018/08 2018.

[7] 蕭達慶, "從2012ECSCRM看碳化矽昇華法晶體技術發展近況," 工研院材化所, 2012

[8] Wikipedia. Lely method.

https://en.wikipedia.org/w/index.php?title=Lely_method&oldid=896204204

[9] SiC Power Materials (Springer Series in Materials Science). 2004

[10] H. Abderrazak and E. S. B. H. Hmida, "Silicon Carbide: Synthesis and Properties," in Properties and Applications of Silicon Carbide, 2011

[11] Wikipedia. Acheson process. https://en.wikipedia.org/w/index.php?title=Acheson_process&oldid=950228730

[12] S. Kavecky, B. Janekova, J. Madejova, and P. Sajgalik, "Silicon carbide powder synthesis by chemical vapour deposition from silane/acetylene reaction system," (in English), J Eur Ceram Soc, vol. 20, no. 12, pp. 1939-1946, Nov 2000.

[13] M. R. Youm, S.-I. Yun, S. C. Choi, and S. W. Park, "Synthesis of β-SiC powders by the carbothermal reduction of porous SiO2–C hybrid precursors with controlled surface area," Ceram Int, vol. 46, no. 4, pp. 4870-4877, 2020.

[14] A. S. Mukasyan, Y.-C. Lin, A. S. Rogachev, D. O. Moskovskikh, and R. Cutler, "Direct Combustion Synthesis of Silicon Carbide Nanopowder from the Elements," J Am Ceram Soc, vol. 96, no. 1, pp. 111-117, 2013

[15] A. Mukasyan, "Combustion Synthesis of Silicon Carbide," 2011.

[16] L. Wang, X. B. Hu, X. G. Xu, S. Z. Jiang, L. Ning, and M. H. Jiang, "Synthesis of high purity SiC powder for high-resistivity SiC single crystals growth," (in English), J Mater Sci Technol, vol. 23, no. 1, pp. 118-122, Jan 2007.

[17] L. Pizzagalli, "Stability and mobility of screw dislocations in 4H, 2H and 3C silicon carbide," Acta Mater., vol. 78, pp. 236-244, 2014.

At present, most of the global supply of SiC power components are provided by American and Japanese manufacturers. This is mainly because the U.S. and Japan have mastered the manufacturing and mass production of SiC substrate materials. Among them, the crystal growth technology for this material is the most critical and the most difficult. As for the manufacturing process used in the industry, the three major technical barriers to SiC crystal growth are the “crystal growth conditions”, “growth speed”, and “defect density”. At present, the temperature for SiC vapor growth must be above 2,300°C, and the pressure needs to be about 350MPa. Silicon, on the other hand, only needs a temperature of about 1,600°C. This high temperature condition places extremely high demands on equipment and process control. Any inadvertent loss of control over the crystal growth temperature can lead to the scrapping of all the ingots grown over several days.

In addition, traditional silicon materials usually take only about 3 days to grow a crystal ingot of about 300cm. The current SiC crystal growth process, however, is unable to grow a 10cm ingot even after 2 weeks, resulting in very limited production. Furthermore, SiC is comprised of as many as 200 types of crystals. Among them, only the handful of single crystal SiC with hexagonal structures (such as 4H-SiC) are desired semiconductor materials. Therefore, precise control over the silicon/carbon ratio, process temperature, growth rate, and gas flow and pressure, etc. is required during the growth of SiC crystals. Otherwise, it is very easy to produce polymorphic inclusions or failed crystals with Screw Dislocation defects.

Companies with SiC wafer mass production capabilities in the market today include Wolfspeed, ROHM, and II-VI, etc. The key equipment they use, such as the crystal growth furnace the processing technology, have all been developed over several years. Those manufacturers now seeking to become competitive in this field quickly have to overcome an extremely high technical threshold. Moreover, more and more countries are beginning to consider SiC materials to be strategic resources and are adopting export control measures, thus forming an oligopoly where SiC upstream materials are supplied by only a few manufacturers and leading to the current situation where there is a severe imbalance between supply and demand in the industry.

Recently, however, in the face of huge future market opportunities, many international manufacturers have begun actively laying the groundwork through acquisitions and strategic cooperation with other manufacturers who have mastered crystal growth technology, etc. to quickly build up their production capacities in order to guarantee themselves a stable supply of substrates. For example, in 2021, Onsemi acquired the SiC manufacturer GT Advanced Technologies (GTAT) for 415 million USD to support its market growth plans for products related to electric vehicles, charging stations, and other energy infrastructures by obtaining sufficient SiC wafers. Similarly, Infineon signed a two-year contract with Japan’s Showa Denko in order to ensure a sufficient supply of various SiC materials. According to statements made by Infineon, the annual growth rate of the SiC market in the next five years is estimated to reach 30-40%, and this cooperation with Showa Denko will rapidly expand its market share of power chips for emerging applications such as electric vehicles and 5G.

The development of SiC third generation semiconductors has entered a stage of fierce competition. Although the American and Japanese manufacturers that have mastered the technology for making the substrate materials have created a three-kingdoms situation, governments around the world are setting their sights on the huge SiC power electronics market opportunity. They are actively expanding the strategic layout of their industry chains and combining the resources and strengths of government, academic and industry researchers to accelerate the development of core SiC equipment and process technology, striving to obtain a leading or niche position in the market in the future. Taiwan has always supported the entire ICT industry chain with its mature silicon semiconductor technology, playing a key role in the global technology industry. However, its development of third generation semiconductor technology and materials has been relatively slow.

In view of this, the government launched the four-year “Compound Semiconductor Project” in 2022, using policies to connect the upstream and downstream nodes of the semiconductor industry chain to accelerate the development of domestic 8-inch key process equipment and promote the independent production of SiC powder and 8-inch SiC wafers. In terms of SiC growth and epitaxial technology, it is known that several well-known manufactures, including Universal Crystal, Shengxin Materials and Wensheng, have been investing in their own development. We anticipate that Taiwan will maintain its industrial advantages in the field of silicon-based semiconductors through the early establishment of leading technologies for third generation semiconductors and continue to contribute to the global technology industry and economic development.

The author of this article, Professor Quanpu Liu, and his laboratory team are nationally recognized for their work in third generation semiconductor materials research. Professor Liu earned his PhD in materials from the University of Cambridge, UK in 1999, and is currently a distinguished professor of the Department of Materials at National Cheng Kung University. He is also an internationally renowned scholar. His team’s nanomaterials research has been fruitful, and they have published more than 225 articles in various journals and twice won the Outstanding Research award of the Ministry of Science and Technology as well as several international academic awards. In recent years, Professor Liu has been committed to assisting the industry with accelerating the application of R&D results. MA-tek is very honored to join hands this year with Professor Liu to carry out industry-university cooperation by providing the full range of analytical services his team needs in the research of ultra-high purity SiC materials. MA-tek has a comprehensive set of testing equipment and the professional, technical experience to fully meet the various testing and analysis needs of compound semiconductors, process packaging and failure analysis.