Highlights from the Future Days Semiconductor Q&A: Discover key insights into silicon carbide!

by Dirk Kok, Thursday, 26th September 2024

This year, our Future Days event for semiconductors focused on silicon carbide (SiC), a key material for the semiconductors of the future. It was a great opportunity for viewers to catch up with the most recent developments in SiC growth, including the analytical instruments, techniques, and even simulation software solutions that can help optimize the SiC crystal growth process.

We’ve already published a recap of the presentations given during the event, but the Q&A sessions provided some of the most insightful moments of the day. Read on for our highlights!

What’s so special about silicon carbide?

The focus of this Future Days event was SiC growth, so naturally many questions were about this material.

Why is SiC more expensive and difficult to produce than silicon (Si), and is SiC was more difficult to measure compared to other materials.

Our expert explained that conventional crystal growth techniques like the Czochralski method can’t be used for SiC due to its elevated melting point. As a result, it can take three times as long to produce a 6-to-8-inch-diameter SiC crystal as it takes to produce a 12-inch-diameter silicon crystal.

There is no difference in terms of measuring technology between SiC and similar materials like Si and gallium nitride (GaN), although new accessories facilitate the accuracy levels customers are looking for. SiC is a unique material, however, with several hundred polytypes, and while it is very robust and thermally conductive, it is also very difficult to grind, cut, and polish. While the metrology works the same way as with other materials, SiC therefore does present unique challenges.

What is the current state of the competition between SiC and GaN?

Dennis Seibert answered that SiC and GaN have very different applications, with SiC intended for higher-power applications, and GaN for higher-frequency applications. So, there is no competition in that sense. “The market for both is expanding and the expectation is that they won’t grow at the expense of each other, but they will grow together,” Dennis explained.

SiC in research and industry

The market for SiC includes applications in research and industry. Several questions asked whether there were any differences in trends or analytical solutions between the production environment and the R&D lab.

Which Malvern Panalytical instruments are aimed at R&D metrology as opposed to production metrology?

Our expert answered that differences in equipment are very small, and production devices can typically handle higher throughputs, but any system can be used for either R&D or production.

What are the focuses for investment in SiC R&D and SiC production, and how does sustainability fit into the picture?

The R&D agenda is primarily focused on solving the issues encountered when transitioning from the manufacture of 6-inch crystals to 8-inch crystals. These include productivity, defects, and dislocation density.

Sustainability concerns find their way into discussions about SiC in two ways. Firstly, in exploring how to make the most efficient use of the material, as it is needed in key technologies such as electric car power converters and battery shortage grids. And secondly, in improving the efficiency of the production process itself, which currently requires very high energy consumption.

What are the capabilities of the presented solutions?

Our Future Days event involved the presentation of hardware and software solutions for the SiC context, so many questions focused on the capabilities of these solutions. Slobodan Mitic and Dennis Seibert also presented the equipment and techniques included in PVA Tepla’s powder-to-wafer Technology Hub Service (THUB) portfolio.

Which parts of the SiC growth process does THUB impact, and what are the detection limits of scanning acoustic microscopy and infrared depolarization?

Slobodan explained that THUB helps ensure that materials with the right properties are used, for example. Combined with the use of growth simulations, the production setup can be fine-tuned to achieve the best outcomes. For example, THUB can help differentiate between good and bad boules for processing, cutting down production costs and increasing yields by reducing waste. The detection limits of the two techniques the viewer asked about are below 10 micrometers.

Can STR’s virtual reactor take into account the effect of powder granule structure on evaporation rate?

Andrey Smirnov presented STR’s virtual reactor software, and how it can be used to optimize designs – and process efficiency as a result. To this question, he answered that powder size, grain size, and chemical composition could indeed be specified. The virtual reactor can also model the deposition on the reactor walls, serving as both a starting point and an optimization tool for process design.

Can Malvern Panalytical instrument accessories handle larger boules, and what is the largest diameter sample the instruments can map?

Having given a detailed presentation of Malvern Panalytical’s product range for the semiconductor context, Dirk Kok explained that the mapping table can handle 285 mm samples in the default configuration, but if the metal reference bars are removed, either a 300 mm sample can be fully mapped, or just the central part of a 450 mm sample. The accessories can indeed measure even large chunks of silicon. “The limiting issue,” he added, “is only how much the operator can lift!”

Thanks again to all the experts who delivered talks during our event, and to everyone who joined in online!

Have any questions of your own? Contact our experts today.

Further reading: