Why is β-Ga₂O₃ considered a breakthrough for RF power?

The semiconductor industry's RF power hierarchy has been Si → GaAs → GaN → SiC, each step increasing bandgap and critical field. β-Ga₂O₃ represents the next step with the highest Baliga figure of merit of any commercially viable semiconductor. The BFOM = ε·μ·E_c³ captures the fundamental trade-off between on-resistance and breakdown voltage in a power device. For β-Ga₂O₃: ε_r = 10, μ = 200 cm²/V·s, E_c = 8 MV/cm. BFOM = 10 × 200 × 8³ = 1,024,000 (relative units). Normalized to silicon (BFOM_Si = 1): β-Ga₂O₃ ≈ 3,444× Si. Compare: GaN ≈ 870× Si, 4H-SiC ≈ 340× Si. This means a β-Ga₂O₃ device can achieve the same breakdown voltage with 3,444/870 ≈ 4× lower on-resistance than GaN, or the same on-resistance at much higher voltage. For RF applications, the Johnson figure of merit (JFOM = E_c·v_sat/2π) is also relevant: β-Ga₂O₃ v_sat ≈ 1.5-2×10⁷ cm/s (comparable to GaN). JFOM ≈ 8×10⁶ × 2×10⁷ / 6.28 = 2.5×10¹³ V/s, roughly 2× GaN. However, the real disruption is cost: melt-grown substrates at $100-500 per 2-inch wafer vs. $2,000-5,000 for SiC and $5,000-20,000 for freestanding GaN.

What are the limitations of β-Ga₂O₃ for RF?

Despite the exceptional BFOM, β-Ga₂O₃ faces three significant challenges: 1) No p-type doping: all known acceptor impurities (Mg, Zn, Fe, N) create deep levels (>1 eV above VB) that do not ionize at room temperature. This eliminates bipolar devices (BJTs, thyristors), pn junction diodes, and complementary logic. RF devices are limited to unipolar structures: MOSFETs, MESFETs, and Schottky diodes. Enhancement-mode operation requires careful gate dielectric engineering. 2) Low thermal conductivity: β-Ga₂O₃ k = 10-27 W/m·K (anisotropic, lowest along [100]). Compare: GaN k = 130 W/m·K (13× higher), SiC k = 370 W/m·K (37× higher). For RF power amplifiers dissipating 5-10 W/mm, junction temperature rise through the substrate is severe. Mitigation strategies: thin β-Ga₂O₃ epilayers on diamond substrates (k = 2,000 W/m·K), flip-chip bonding to high-k heat spreaders, or pulsed operation (radar). 3) Epitaxial maturity: MBE and HVPE produce β-Ga₂O₃ epilayers with dislocation densities of 10⁴-10⁶ cm⁻² (vs. 10²-10⁴ for GaN-on-SiC). Electron mobility in epitaxial layers: 100-180 cm²/V·s (vs. theoretical 250-300). Interface trap density at β-Ga₂O₃/dielectric: 10¹¹-10¹² cm⁻² eV⁻¹ (needs 10× reduction for competitive RF FETs).

What is the current state of β-Ga₂O₃ RF device development?

As of 2025-2026, β-Ga₂O₃ RF devices are in the research-to-prototype transition: MOSFET performance: best reported f_T = 27 GHz, f_max = 37 GHz (Cornell, 2023). Gate length 100 nm, T-gate on 2 μm thick β-Ga₂O₃ epilayer. Compare GaN HEMT: f_T > 300 GHz demonstrated. β-Ga₂O₃ is 5-10 years behind GaN in RF maturity. Power density: 3.8 W/mm at 1 GHz (AFRL, 2024) in pulsed mode. GaN: 40+ W/mm demonstrated. The gap narrows at higher voltages where β-Ga₂O₃'s higher breakdown is advantageous. Breakdown voltage: > 8 kV demonstrated in lateral MOSFETs (Tamura/NICT, 2023). This exceeds GaN lateral devices (< 2 kV typical) and approaches SiC vertical devices. Substrate availability: 2-inch EFG wafers commercially available (Novel Crystal Technology, Japan). 4-inch demonstration reported. Cost: approximately $200-500/wafer in research quantities, projected $50-100 at volume. Key developers: AFRL (US DoD), Cornell, Ohio State, Tamura/NCT (Japan), University of Utah. DARPA DREaM program funded β-Ga₂O₃ RF development 2020-2025. Commercial timeline: Schottky diodes: 2027-2028 projected. RF MOSFETs competitive with GaN: 2030+ (requires thermal solution and mobility improvement).

Semiconductor Materials

Beta Gallium Oxide

/BAY-tuh GAL-ee-um OX-ide/ — β-Ga&sub2;O&sub3;

Ultra-wide bandgap semiconductor (E_g = 4.8 eV). Critical field: 8 MV/cm. Baliga FOM: 3,444× Si (vs. GaN 870×, SiC 340×). Melt-grown substrates via EFG/Czochralski at 10 to 100× lower cost than SiC/GaN. Electron mobility: 100 to 300 cm²/V·s. Challenges: no p-type doping, low thermal conductivity (10 to 27 W/m·K). Applications: high-voltage RF MOSFETs, Schottky diodes, UV-C detectors.

E_g: 4.8 eV

E_c: 8 MV/cm

BFOM: 3,444× Si

Understanding β-Ga&sub2;O&sub3;

The semiconductor industry's relentless push toward higher power density and higher voltage has driven a progression from silicon to GaAs to GaN to SiC. Beta gallium oxide represents the next frontier: an ultra-wide bandgap material whose critical electric field is more than double that of GaN. This translates to a Baliga figure of merit nearly 4× higher than GaN and 10× higher than SiC, promising dramatically lower on-resistance at a given breakdown voltage.

What makes β-Ga&sub2;O&sub3; uniquely disruptive is its substrate economics. Unlike GaN (which requires expensive HVPE or ammonothermal growth) and SiC (which requires the extreme temperatures of physical vapor transport), gallium oxide melts at 1,795°C and can be pulled from the melt using conventional crystal growth methods. This opens a path to large-area, low-cost native substrates that could fundamentally change the power semiconductor cost structure.

Figures of Merit

Baliga FOM (power switching):
BFOM = ε_r·μ·E_c³
β-Ga&sub2;O&sub3;: 10×200×8³ = 1,024,000
GaN: 9.5×1500×3.3³ ≈ 512,000
4H-SiC: 9.7×900×2.5³ ≈ 136,000

Johnson FOM (RF power-frequency):
JFOM = E_c·v_sat / (2π)
β-Ga&sub2;O&sub3;: 8×10⁶×2×10⁷/6.28 ≈ 2.5×10¹³
GaN: 3.3×10⁶×2.5×10⁷/6.28 ≈ 1.3×10¹³

Specific On-Resistance:
R_on,sp = 4V_BR² / (ε·μ·E_c³)

Wide Bandgap Semiconductor Comparison

Property	Si	4H-SiC	GaN	β-Ga&sub2;O&sub3;	Diamond
E_g (eV)	1.1	3.3	3.4	4.8	5.5
E_c (MV/cm)	0.3	2.5	3.3	8.0	10.0
μ (cm²/V·s)	1,400	900	1,500	200	2,000
k (W/m·K)	150	370	130	10–27	2,000
BFOM (×Si)	1	340	870	3,444	24,000
Substrate cost	$	$$$$	$$$$$	$$	$$$$$$

Common Questions

Frequently Asked Questions

Why is it a breakthrough?

BFOM 3,444× Si (4× GaN, 10× SiC). Melt-grown substrates at $200 to 500/wafer vs. $2K to 5K (SiC), $5K to 20K (GaN). E_c = 8 MV/cm enables higher voltage at lower R_on. Projected substrate cost at volume: $50 to 100.

Limitations?

No p-type doping (deep acceptors >1 eV). Low k = 10 to 27 W/m·K (SiC: 370). Mobility 100 to 200 cm²/V·s (GaN: 1,500). Solutions: diamond substrates, flip-chip, pulsed operation.

Device status?

Best RF MOSFET: f_T = 27 GHz, f_max = 37 GHz (100 nm gate). Power: 3.8 W/mm pulsed at 1 GHz. BV > 8 kV lateral. Commercial Schottky diodes projected 2027 to 2028. RF competitive with GaN: 2030+.

Related Terms

← Beta Factor Bethe-Hole Coupler →

Power Semiconductors

Precision RF Components

RF Essentials provides precision terminations and custom waveguide assemblies for high-power device characterization, load-pull measurement, and thermal management testing.

Request a Quote