US12698573: Resonac's SiC Patent That Quietly Decides EV Range

💡 Resonac Corporation, the world's largest independent supplier of silicon carbide epitaxial wafers, has just been granted US Patent 12,698,573 for a new SiC crystal growth method - classified under C30B 29/36 by the USPTO and issued on August 4, 2026. The patent advances the manufacturing process that determines defect density and yield in the wafers at the heart of every EV power inverter. Better crystals mean better chips, and better chips mean more kilometers per charge.

What silicon carbide is - and why the growth step decides everything

Most people know silicon as the material inside every smartphone and laptop. Silicon carbide (SiC) is the harder, hotter, faster cousin: it withstands three times the voltage of silicon, operates at temperatures silicon cannot survive, and converts electrical power with far less waste heat. That last property is why SiC is rapidly replacing silicon inside the inverters of every serious electric vehicle on the road today.

But SiC comes with a fundamental catch: growing its crystals is enormously difficult. Silicon crystals grow relatively quickly from a molten bath. SiC does not melt - it sublimes directly from solid to gas at around 2,700°C. Crystal growth must happen in a carefully controlled vapor phase, grain by grain, over days or weeks. A single defect introduced during that process - a basal-plane dislocation, a micropipe, or a surface irregularity - can render the finished wafer useless for high-voltage applications. SiC crystal quality is not just a materials science problem; it is the central bottleneck of the entire EV power semiconductor supply chain. That bottleneck is exactly what US 12,698,573 is designed to address.

What US 12,698,573 actually covers

Granted by the USPTO on August 4, 2026, US Patent 12,698,573 is assigned to RESONAC CORPORATION and was developed by inventors Tanaka, Kindaichi, and Ohuchi. The patent is classified under C30B 29/36 - the international patent code for single-crystal growth of silicon carbide from vapors, the epitaxial process at the foundation of all SiC wafer manufacturing.

The title "Crystal growth method" is deliberately broad but precise in scope: it covers a specific improvement to the vapor-phase process by which SiC epitaxial layers are deposited onto a crystalline substrate. Classification C30B 29/36 places it firmly in the territory of methods that control the nucleation and layer-by-layer growth of SiC single crystals - the step where defect density is set, where doping uniformity is established, and where the economic yield of the entire downstream supply chain is won or lost.

Resonac's closely related patent US 12,356,687, granted in July 2025, focused on achieving high doping concentration uniformity in SiC layers by controlling the carbon-to-silicon (C/Si) gas ratio during epitaxial growth - using a C/Si ratio between 1.1 and 1.7 to compensate for nitrogen's inhibitory effect on crystalline growth. US 12,698,573 continues this systematic push to master every controllable variable in SiC crystal growth at manufacturing scale. Together they build an IP stack that protects different aspects of the same critical process.

The real-world problem it solves

DENSO, one of the world's largest automotive component suppliers, adopted Resonac's SiC epitaxial wafers for the inverter in the Lexus RZ - the first battery electric vehicle in the Lexus lineup. DENSO's selection criteria were revealing: they cited Resonac's "industry-leading low levels of surface-defect density and basal-plane dislocation." These two defect types are precisely the ones that crystal growth methods must control.

Without low defect density, the SiC power transistors inside a vehicle's inverter will fail under the high voltages and temperatures of real-world EV operation. With it, the inverter converts battery DC power to motor AC power with lower energy loss, directly extending the vehicle's driving range. This is not an abstract materials science story: it is connected directly to why your next EV will drive farther on a single charge than the one before it. The quality of a crystal grown weeks earlier in a Japanese factory quietly determines the range number on the window sticker.

Resonac's position - and the scale of what is at stake

Resonac Corporation, formerly known as Showa Denko K.K., is the world's largest independent supplier and technology leader in SiC epitaxial wafers. This position is unusual: Resonac chooses not to manufacture SiC semiconductor devices itself, remaining purely a wafer supplier. That independence allows the company to supply multiple device manufacturers simultaneously, pooling defect-density feedback across customers to accelerate improvements faster than any vertically integrated rival could.

The company's patent portfolio reflects this focus. According to an assessment by France-based KnowMade, Resonac's patent value in SiC epitaxial wafers more than doubled that of industry peers in 2024. The company has begun test shipments of 8-inch (200mm) SiC epitaxial wafers, which offer significantly more chip area per growth run than the current 6-inch standard, and has constructed a new production facility at its Yamagata Plant to support scaled output. US 12,698,573 is one brick in the IP fortress that protects this leadership position - and every competitor attempting to manufacture higher-quality, larger-diameter SiC wafers must navigate around Resonac's documented prior-art landscape.

Who this threatens - and why SiC supply is a geopolitical story

The SiC power semiconductor supply chain involves at least three distinct competitive layers. The crystal growth layer - Resonac's home territory - is where US 12,698,573 operates. Competitors here include Wolfspeed (formerly Cree, United States), SiCrystal (Germany, a subsidiary of Rohm of Japan), and ON Semiconductor's SiC substrate business. The device manufacturing layer is where companies like Wolfspeed, STMicroelectronics, Infineon, Rohm, and BYD Semiconductor convert wafers into SiC power transistors. The system integration layer is where automotive suppliers like DENSO, Vitesco, and Bosch package those transistors into inverters.

A patent on the crystal growth method at the top of this stack has compounding leverage: every improvement in growth quality and yield reduces cost at every layer below it. If Resonac's crystal growth methods produce fewer defects, rivals must either find their own path or fall behind on yield. The supply chain is also a geopolitical story: SiC crystal growth expertise is concentrated in Japan, the United States, and Europe. A Resonac patent granted in the US protects their position in the world's largest automotive market and gives them grounds to enforce in disputes. That is why these seemingly narrow materials science patents matter beyond the laboratory.

The Soitec connection - and what 70% CO2 reduction means

In September 2024, Resonac and French semiconductor materials company Soitec signed an agreement to jointly develop 200mm (8-inch) SiC bonded substrates. The technology works by bonding the epitaxial surface of a high-quality SiC single crystal to a polycrystalline SiC support wafer using Soitec's proprietary SmartSiC technology, then splitting the original substrate into thin films. Each SiC boule, which previously yielded one substrate, can now yield multiple substrates - multiplying wafer output without multiplying the slow, expensive crystal growth step.

This bonded-substrate approach reduces manufacturing CO2 emissions by up to 70% compared to conventional SiC substrate production, because fewer growth cycles are needed per usable wafer. It is the manufacturing equivalent of building a single foundation and extruding multiple floors from it. SiC crystal growth method patents like US 12,698,573 directly underpin this stack: the bonded-substrate approach is only as good as the initial crystal quality that enters the bonding process. A better crystal growth method means higher-quality starting material, which means better bonded substrates, which means cheaper high-quality wafers at scale.

Key patent facts at a glance

Field	Detail
Patent number	US 12,698,573
Title	Crystal growth method (SiC epitaxial layer)
Assignee	RESONAC CORPORATION
Inventors	Tanaka, Kindaichi, Ohuchi
Issue date	August 4, 2026
Jurisdiction	United States (USPTO)
Classification	C30B 29/36 (SiC single-crystal vapor-phase growth)
Related patent	US 12,356,687 (Resonac, July 2025)

So what does it mean for us?

Three takeaways from US 12,698,573 and the field it sits in.

First: the competition for EV range is fought inside crystal growth chambers, not on the showroom floor. Every patent granted on SiC manufacturing is a claim on the efficiency of future electric vehicles. The companies that master crystal growth cheaply and at scale will set the floor price of the power semiconductors that determine how far an EV travels. Resonac's filing record shows they intend to own a dominant position in that floor.

Second: this supply chain is fragile in exactly the way semiconductor supply chains always are. SiC crystal growth takes weeks. A single facility running below specification can ripple through automotive production lines on multiple continents. The DENSO-Resonac relationship for the Lexus RZ is instructive: DENSO did not choose the cheapest wafer supplier; they chose the one with the lowest defect density and the longest proven track record. Quality protected by patents is the durable competitive advantage here.

Third, a cautious note: translating a materials science improvement into automotive-scale production takes years. US 12,698,573 is part of a portfolio that will only realize its full value over a five-to-ten-year production ramp. The Yamagata Plant was planned for September 2025 completion; its actual volume ramp will shape whether this IP translates into sustained market share or remains primarily a defensive IP position.

FAQ

Sources: USPTO Official Gazette Week 31, August 4, 2026 - Resonac patents listing | Resonac SiC Business Strategy (2024) | Resonac SiC Epitaxial Wafer Products | DENSO adopts Resonac SiC epi-wafer for Lexus RZ inverter (March 2023) | Resonac and Soitec develop bonded SiC substrates (September 2024) | Google Patents - US 12,356,687 (Resonac SiC doping uniformity, July 2025)

About the author

Dao Huy (Lucas) is a professional translator (English, Chinese, French into Vietnamese) with over 7 years of experience in technical, patent, and IP documentation. His work spans patent applications, engineering specifications, and technology product localization for clients in the semiconductor materials, power electronics, and automotive technology sectors. As SiC wafer patents cross borders - from a USPTO grant to filings in Japan, Europe, and Vietnam's growing EV component supply chain - the precision of patent translation and technical translation is not a formality: a mistranslated materials specification in a patent claim can shift the boundary of protection across an entire jurisdiction.

If your organization needs patent translation, IP translation, or English to Vietnamese technical translation for semiconductor materials, SiC technology, or EV engineering documents, contact Dao Huy at daohuy.com for a consultation and quote.