The race for the next great semiconductor material.
by Caitlin Ryan, Scientific Writer at Ossila
For decades, Moore’s Law—the doubling of transistors on a chip every two years—has defined progress in computing. However, with transistors now operating at atomic scales and quantum effects causing power and thermal issues, the path forward depends on new materials, not just smaller silicon.
Gallium Nitride
It’s impossible to discuss the next generation of semiconductors without bringing up Gallium Nitride (GaN). As a wide-bandgap material, GaN thrives in high-frequency, high-power applications where silicon struggles. It has high electron mobility, enabling rapid switching capabilities, and superior thermal performance that addresses heat dissipation issues. Additionally, GaN is widely recognized for its superior power performance, enabling the creation of more efficient transistors and ultimately smaller systems.
Silicon Carbide
Silicon Carbide (SiC) is another wide-bandgap powerhouse, and it’s the closest that we currently have to mainstream adoption. Again, SiC handles far greater voltages than silicon, but also has the advantage of being durable, making it ideal for use in harsh environments, such as industrial systems and aerospace applications.
SiC’s most significant opportunity lies in electric vehicles (EVs), where efficiency and heat resistance are particularly important. As such, the SiC market is projected to hit $4.5 billion by 2027, making it one of the fastest-growing segments in the semiconductor industry.
2D materials
At the cutting edge of semiconductor research are 2D materials which are atomically thin layers with extraordinary electrical, optical, and mechanical properties. The most well-known of these is graphene, a single layer of carbon atoms in a hexagonal lattice that’s stronger than steel, nearly transparent, and boasts electron mobility up to 100 times that of silicon. These traits make it an ideal candidate for futuristic applications, such as flexible electronics and high-sensitivity sensors.
We’re not quite there yet
Silicon dominates the industry for a reason. After decades of refinement, manufacturing is fast, cheap, and highly optimised. Switching to a new material isn’t just a simple upgrade; it’s a complete reset that requires new processes and tools. That’s an expensive gamble for an industry that is built on scale and predictability. Silicon is also abundant and easy to source, unlike many next-generation materials that depend on rare elements or volatile by-products, such as GaN’s reliance on aluminum and zinc mining.
Compatibility issues may be one of the biggest hurdles to overcome. Even if tomorrow’s semiconductors outperform silicon, they still need to fit into a world that’s built for silicon. For now, the industry is leaning on hybrid solutions, where new materials are layered onto existing silicon chips. It’s a clever workaround, but it adds design complexity, introduces new points of failure, and forces engineers to rethink decades of assumptions about transistor architecture.
The road ahead
Looking forward, the semiconductor landscape is unlikely to shift overnight. What will decide the winner? It’ll come down to more than just performance, and manufacturing economics will determine whether a material is viable on the massive scale that the semiconductor industry demands.
Silicon isn’t going anywhere soon, but as demand for higher efficiency grows, we’ll start to see these new materials increasingly take centre stage in specific applications.