GaN’s RF promise depends on measuring defects

For decades, gallium arsenide (GaAs) has been the default material for high-performance radio-frequency semiconductors used in satellite communications, radar, and mobile networks. Its advantages are practical as much as technical: GaAs is mature, well understood, and can be produced on large, high-quality wafers at an economical cost.

That balance is now being challenged by gallium nitride (GaN). GaN can operate at higher voltages, temperatures, and power densities, making it attractive for next-generation RF systems, including 5G and 6G networks, defense equipment, and space technologies. But the shift is not a simple swap of one material for another. The harder problem is proving that GaN can deliver its theoretical advantages consistently, at scale, and at a commercially viable cost.

Why GaN is harder to manufacture

Large, defect-free GaN substrates remain difficult and expensive to produce. Instead, most GaN devices rely on heteroepitaxy: growing a thin GaN layer on a different substrate, typically silicon or silicon carbide, so manufacturers can use more established wafer technologies.

The trade-off is that the materials do not perfectly match. Differences in lattice structure and thermal expansion introduce defects into the GaN layer. Those defects can influence electrical performance and long-term reliability, making results less predictable across wafers and individual devices.

Recommended reading

Intel shares Atom x86 designs with new startup

GaN therefore offers superior theoretical performance to GaAs, but realizing that performance requires tighter control over material quality and manufacturing. A GaN module also requires a broader system redesign rather than a straightforward replacement of a GaAs component.

That makes measurement a central engineering problem. Manufacturers need to understand not only whether a process produces GaN, but whether small changes in temperature, deposition rates, or substrate preparation improve the material or make it worse.

Metrology has three jobs

Measurement, or metrology, plays three distinct roles in the development and commercialization of GaN technology:

• Process development: Measurements provide feedback on heteroepitaxial growth, helping engineers refine manufacturing conditions and reduce defect densities.
• Material-quality claims: Manufacturers need accurate, comparable methods to demonstrate that their GaN is better than a competitor’s. Customers must be able to trust that the same quality claim means the same thing at different organizations.
• Device validation: Material measurements must ultimately connect to application-level results such as power output, efficiency, frequency response, and thermal stability.

The third link is particularly important. If a manufacturer improves its growth process, it must show that the improvement produces a measurable gain in the finished RF device. Connecting material quality to device behavior turns a laboratory result into evidence that customers can use.

This is also why the measurement challenge extends beyond a single wafer or transistor. Our recent coverage of a Keysight and WIN Semiconductors GaN development platform showed the same industry push toward evaluating RF devices as complete systems rather than isolated components.

From materials to complete systems

As semiconductor devices become more specialized and operate in harsher conditions, performance increasingly depends on interactions among materials, structures, and manufacturing processes. RF systems are especially sensitive: small imperfections can have outsized effects on signal integrity and efficiency.

The issue is not unique to GaN. Photonics is facing similar measurement problems as heterogeneous integration brings different materials and device types together in a single system. In both fields, engineers must control complexity across multiple scales, from material properties to device structures and final system behavior.

That creates a need for measurement methods that are accurate inside one organization and comparable across the industry. Without trusted benchmarks, customers may struggle to distinguish a genuine material improvement from a difference in testing conditions or reporting methods.

Measurement will shape GaN adoption

The transition from GaAs to GaN is therefore more than a material substitution. It shifts competitive advantage away from established manufacturing alone and toward the ability to engineer, characterize, and validate newer materials.

Countries and companies with strong capabilities in materials science, process development, and measurement will be better positioned to influence how GaN evolves. The speed of adoption across global RF markets will depend on whether manufacturers can measure material quality, correlate it with device performance, and establish credible standards.

GaN’s intrinsic properties may be stronger, but those properties do not automatically become reliable products. Consistent material growth, dependable device fabrication, and credible performance validation are prerequisites. In high-frequency semiconductors, metrology is no longer merely a downstream quality check; it is part of the technology’s development, manufacturing, and route to market.