What are ultra-wide bandgap semiconductors and why do they matter now?

Ultra-wide bandgap (UWBG) semiconductors are materials with bandgap energies generally exceeding 4 eV—a threshold that places them beyond conventional wide bandgap materials such as silicon carbide (SiC) and gallium nitride (GaN). The principal UWBG material classes are gallium oxide (Ga₂O₃), aluminum nitride (AlN), boron nitride (BN), and diamond. Their combination of superior breakdown fields, high-temperature stability, and radiation hardness promises step-change advances that SiC and GaN cannot deliver at the system level.

The urgency behind UWBG investment is structural. Power electronics for new-energy vehicles, smart-grid HVDC transmission, and 5G radio front ends are all pushing against the physical limits of silicon—and increasingly of first-generation SiC and GaN devices. According to filings in this dataset, Infineon Technologies has targeted blocking voltages above 1,000 V and up to 6,500 V for grid-scale power conversion, a regime where UWBG breakdown fields become decisive. Meanwhile, the UVC sterilization market—water and air disinfection at 200–280 nm—requires light sources that III-nitride AlGaN alone cannot efficiently produce below 215 nm, creating an opening for metal-oxide UWBG emitters.

A defining characteristic across retrieved filings is the recognition that bandgaps above approximately 4 eV impose severe materials challenges: low p-type dopability (especially in Ga₂O₃), inadequate thermal conductivity, limited native substrates, and extreme growth temperatures that exceed standard MOCVD equipment limits. The patents in this dataset directly address these constraints through UV-assisted epitaxy, selective lateral growth, and heterogeneous integration strategies—a pattern that signals the field has moved from material curiosity to engineering problem-solving.

From SiC foundations to Ga₂O₃ frontiers: the UWBG innovation timeline

UWBG patent activity in this dataset divides into three distinct developmental eras, each building on the device architectures and process knowledge of the previous generation. The trajectory runs from volatile compound semiconductor synthesis in the 1970s to explicit Ga₂O₃/diamond complementary integration in 2024.

The foundational period (pre-2015) established the conceptual basis. The earliest relevant record—a Sumitomo Electric Industries filing from 1974—addresses volatile compound semiconductor synthesis under pressure. Fuji Electric’s wide bandgap semiconductor device (JP, 2014) and SemiSouth Laboratories’ normally-off integrated JFET power switch (CN, 2008) reflect consolidation of SiC and GaN as the first commercial WBG generation, establishing device architectures that were later adapted for UWBG materials.

The intermediate maturation period (2015–2022) saw power device architecture innovations intensify. Xidian University filed a series on wide bandgap/silicon heterojunction IGBTs between 2018 and 2021, directly addressing the insertion of UWBG-class breakdown layers into hybrid structures. Texas Instruments began filing silicon-recess hybrid WBG components in 2022. Infineon Technologies filed on WBG wafer processing methods with non-crystalline support layers (CN, 2019), tackling the substrate cost barrier that remains a primary commercialization obstacle.

The emerging UWBG-specific period (2023–2026) is characterized by explicit targeting of Ga₂O₃ and diamond. Southeast University’s high power density UWBG device (CN, 2024) directly names Ga₂O₃ and diamond in a heterogeneous complementary power circuit. Hitachi Energy’s lateral selective epitaxy patents span EP (2023), US (2025), and JP (2026)—the most recent manufacturing-process filings in this dataset.

“Filings from 2023–2026 account for approximately 10 of the directly UWBG-relevant records in this dataset, indicating accelerating activity in materials explicitly beyond the SiC/GaN generation.”

Four technology clusters defining UWBG patent activity

Patent activity in this dataset organizes into four distinct technology clusters, each targeting a different rate-limiting constraint in UWBG device development. Together they map the engineering agenda that must be solved before UWBG materials reach volume manufacturing.

Cluster 1: UV-Assisted and High-Temperature Epitaxial Growth

Conventional MOCVD requires temperatures approaching or exceeding 1,400 °C to achieve adequate adatom migration for UWBG materials such as AlN and BN. The Ningbo Institute of Materials Technology and Engineering (Chinese Academy of Sciences) addresses this in a 2022 CN filing by applying pulsed UV irradiation at intensities above 0.2 W/cm²—with photon energy exceeding the UWBG material bandgap—during MOCVD growth. The UV field substitutes for a portion of the thermal energy, reducing reactor temperature demands while improving crystal quality, stress profiles, and optical and electrical breakdown characteristics.

Hitachi Energy’s selective lateral epitaxial growth patents (EP, 2023; US, 2025) describe forming WBG semiconductor material within insulating-layer cavities on a crystalline seed substrate. The lateral geometry decouples crystalline quality from substrate defect density—a critical advantage when native Ga₂O₃ and AlN substrates carry high dislocation densities. The US 2025 filing adds claims covering dual-template structures enabling simultaneous fabrication of both n+/p-/n-/n+ and n+/n-/p-/n+ device polarity variants within a single lateral growth process.

Cluster 2: Heterogeneous Integration — UWBG on Silicon/SiC Substrates

Because Ga₂O₃ substrates are expensive and diamond substrates remain near-laboratory scale, several assignees pursue inserting UWBG-characteristic breakdown layers into conventional silicon or SiC platforms. Southeast University’s 2024 CN filing cascades a Ga₂O₃ depletion-mode power FET with a diamond enhancement-mode FET on a diamond semi-insulating substrate, exploiting diamond’s approximately 10 W/(cm·K) thermal conductivity to manage heat dissipation that Ga₂O₃ alone—with a thermal conductivity of approximately 0.27 W/(cm·K)—cannot sustain.

Texas Instruments’ approach (US, 2022) forms a WBG structure on silicon within a silicon recess, with separate current terminals on each material region, enabling a hybrid switching element that combines silicon process maturity with WBG breakdown performance. Xidian University’s IGBT series (CN, 2018–2021) bonds a WBG substrate—SiC, GaN, or diamond—to an n-type silicon active region via crystal bonding, exploiting the high critical breakdown field of the WBG layer while retaining the well-characterized silicon IGBT channel process.

Cluster 3: Deep-UV Light Emission Using UWBG Active Layers

Deep-UV LEDs operating below 280 nm require active layers in AlGaN with greater than 60% aluminum content, or metal-oxide systems with bandgaps extending into the UVC window. Silanna UV Technologies’ 2023 CN filing on metal-oxide semiconductor LEDs targets 150–425 nm emission, directly addressing the crystalline quality deficiencies of III-nitride UVC LEDs caused by large lattice mismatch with sapphire. The filing notes that AlN limits UVC operation to approximately 215 nm and that metal oxides offer a pathway below that limit—a claim that, if validated in manufacturing, would represent a significant competitive displacement of III-nitride AlGaN.

Wavelo Inc.’s thin-film UV LED patent (CN, 2025) develops a laser-scribing and thermal self-separation process to remove SiC growth substrates from AlN/AlGaN-based deep-UV LED epi-stacks operating at or below 280 nm, converting them to thin-film chip structures on high-thermal-conductivity support substrates for operation beyond 350 mA. The Guangdong Institute of Semiconductor Technology (CN, 2024) exploits differential thermal expansion between support substrate and AlGaN thin-film LED to introduce controlled strain that switches the dominant emission polarization from TM to TE mode, substantially improving light extraction through the device surface.

Cluster 4: Gate Engineering and Ohmic Contact Formation

Reliable gate dielectrics and low-resistance ohmic contacts are rate-limiting issues for Ga₂O₃ and diamond FETs, because standard thermal budgets used for SiC are incompatible with these materials’ thermal properties. General Electric (JP, 2022) discloses a vertical n-channel FET gate stack in which a semiconductor layer is doped in-situ during deposition directly above the gate dielectric, followed by a metal-containing layer. This modulates the effective gate work function, improving threshold voltage control without post-implant annealing.

Hitachi Energy’s ohmic contact patent (JP, 2025) achieves contact formation using implantation energies below 15 keV and annealing temperatures below 1,100 °C—well below the approximately 1,650 °C conventionally required for SiC—making the process compatible with thin-film UWBG device stacks that cannot tolerate high thermal budgets. This is one of the most practically significant process-level innovations in the dataset, as it removes a key barrier to wafer-level UWBG integration.

Application domains: power electronics, deep-UV, RF, and extreme environments

UWBG semiconductor filings in this dataset cluster around four distinct application domains, each with a different maturity level and a different set of incumbent technologies that UWBG materials must displace or complement.

Power Electronics and Electric Vehicles

The largest cluster of UWBG-adjacent filings targets power conversion. Tianjin Saiwei Industrial Technology (CN, 2025) explicitly extends SiC/GaN composite systems—proposed at 1:1 to 3:1 ratios via heteroepitaxy or bonding—to new-energy-vehicle motor drive systems and smart-grid HVDC transmission. The University of Science and Technology of China (CN, 2024) addresses the p/n channel mobility mismatch in GaN-only CMOS by substituting diamond p-FETs for improved complementary logic. Infineon’s wafer processing filing (CN, 2019) targets devices with blocking voltages above 1,000 V and up to 6,500 V for grid-scale power conversion—a regime inaccessible to standard silicon.

Deep-UV Sterilization, Disinfection, and Sensing

Silanna UV Technologies’ two patents (CN, 2023 and 2024), Wavelo’s thin-film UV LED patent (CN, 2025), and the Guangdong thin-film LED work (CN, 2024) all converge on deep-UV sources for water and air sterilization in the UVC band (200–280 nm). According to WHO guidance on water quality, UVC disinfection is a critical public health technology, and solid-state UVC sources at high efficiency remain an unmet need. King Abdullah University of Science and Technology (CN, 2024) addresses UV-to-visible wavelength conversion for high-speed optical wireless communication receivers, extending UWBG photonics into data communications.

RF and Telecommunications

Huawei Technologies (CN, 2024) integrates WBG compound semiconductor layers on silicon substrates with an ESD protection architecture, directly targeting power amplifier and low-noise amplifier chips in 5G radio front ends where GaN-on-Si is the incumbent technology. According to ITU spectrum analysis, 5G mmWave deployments demand RF front-end components with higher power density and linearity than existing silicon CMOS can provide—a gap that WBG and UWBG materials are positioned to address.

Extreme-Environment and Radiation-Hardened Electronics

SemiSouth Laboratories’ normally-off integrated JFET power switch (CN, 2008) identifies radiation tolerance and high-temperature operation as primary UWBG use cases for aerospace and defense. The Southeast University Ga₂O₃/diamond device (CN, 2024) highlights radiation hardness as a motivating property for the heterogeneous complementary architecture. These use cases align with NASA‘s long-standing requirements for electronics capable of operating in high-radiation space environments where silicon devices degrade rapidly.

Geographic and assignee landscape: who holds the IP

Innovation in this dataset is geographically concentrated, with China dominating by filing volume and non-Chinese multinationals holding the most commercially significant manufacturing-process patents. This divergence between concept origination and process-level IP has direct implications for freedom-to-operate analysis.

China accounts for the majority of retrieved records with UWBG-specific content. Key assignees include Ningbo Institute of Materials Technology and Engineering (Chinese Academy of Sciences), Southeast University, University of Science and Technology of China, Xidian University (multiple IGBT heterojunction patents, 2018–2021), Tianjin Saiwei Industrial Technology, Guangdong Institute of Semiconductor Technology, and Huawei Technologies. This concentration reflects China’s substantial national investment in third- and fourth-generation semiconductor independence, consistent with policy priorities tracked by WIPO in its annual technology trend reports.

Hitachi Energy Ltd (Switzerland/UK) is the most active non-Asian assignee in UWBG power device manufacturing, with filings in EP (2023), US (2025), CN (2025), and JP (2026)—a coordinated multi-jurisdiction strategy that covers the major commercial markets simultaneously. Texas Instruments files two active/pending hybrid WBG-silicon patents (both 2022 US), signaling a strategic move to embed WBG layers within CMOS-compatible architectures. Silanna UV Technologies (Australia/US) occupies a specialized niche in epitaxial oxide UV devices with no close competitors identified in this dataset.

Strategic implications for IP teams and R&D leaders

Five directional signals from 2023–2026 filings in this dataset carry specific strategic consequences for IP strategists, R&D leaders, and product developers working in power electronics and photonics.

1. Ga₂O₃/Diamond Complementary Integration

The Southeast University (CN, 2024) and University of Science and Technology of China (CN, 2024) filings both propose heterogeneous complementary circuits combining Ga₂O₃ n-channel and diamond p-channel devices. This directly addresses Ga₂O₃’s inability to support p-type doping and diamond’s lack of high-quality n-type channel—each material contributes its strength in a complementary architecture. If manufacturable, this approach would unlock CMOS-like logic at UWBG performance levels. R&D teams should monitor these assignees for follow-on process-level filings that could indicate scale-up activity.

2. Hitachi Energy’s Multi-Jurisdiction Manufacturing Position

Hitachi Energy has established a broad, multi-jurisdiction manufacturing patent position in lateral selective epitaxy and low-thermal-budget ohmic contacts for UWBG power devices. IP strategists seeking freedom-to-operate in UWBG power device manufacturing should conduct detailed claim mapping against the EP/US/JP Hitachi Energy portfolio filed between 2023 and 2026. The JP 2026 filing represents the most recent manufacturing-method patent in this dataset. PatSnap’s patent analytics platform provides claim-level mapping tools suited to this type of FTO analysis.

3. Texas Instruments’ Silicon-Recess Integration Pathway

Texas Instruments’ silicon-recess hybrid WBG integration approach (US, 2022) offers a commercially pragmatic pathway—inserting WBG performance into existing CMOS fabs—and may reach volume production faster than pure UWBG material platforms. Competitors developing discrete UWBG devices should model market displacement risk from this integration-first approach.

4. Silanna’s Metal Oxide Epitaxy Position in Deep-UV

Metal oxide epitaxy (Ga₂O₃/Al₂O₃) may bypass III-nitride limitations in deep-UV, but Silanna UV Technologies appears to be building a dominant patent position in this specific niche. Silanna’s 2024 CN filing covers Ga₂O₃, Al₂O₃, MgO, NiO, and their ternary combinations across multiple crystal symmetries (monoclinic, rhombohedral, triclinic) for applications spanning UV LEDs, laser diodes, photodetectors, high-power diodes, and HEMTs. Entrants developing sub-215 nm UV sources should assess whether the Silanna oxide heterostructure claims cover their proposed material stacks. PatSnap’s landscape analysis tools can accelerate this assessment.

5. Thin-Film Substrate Removal for Sub-280 nm LEDs

The Wavelo thin-film UV LED patent (CN, 2025) using laser-induced thermal self-splitting of SiC growth substrates from AlGaN/AlN stacks, and the Guangdong strain-engineering patent (CN, 2024), represent a maturing thin-film chip architecture analogous to the substrate-removal processes that enabled high-brightness InGaN LEDs. This approach is now being adapted for sub-280 nm AlGaN devices operating beyond 350 mA—a power level relevant to commercial UVC sterilization systems.

“Chinese academic institutions collectively constitute the most prolific source of UWBG device architecture concepts in this dataset. Manufacturing and process-level patents—which typically govern commercial access—are predominantly held by non-Chinese multinationals.”