In the years after graphene was isolated, a common prediction was that it would replace silicon in computer chips. Its electrons move with extraordinary mobility, it is atomically thin, and it handles high current densities. A graphene processor seemed only a matter of engineering.
It has not happened, and for a reason that is easy to understand once it is explained.
A Transistor Must Turn Off
A digital chip is built from billions of transistors acting as switches. Each switch must conduct strongly when on and almost completely stop conducting when off. The ratio between the two — the on/off ratio — needs to be very large, typically many orders of magnitude, so that switched-off transistors do not leak current, waste power, and generate heat.
Silicon can switch off because it has a bandgap: an energy range in which electrons are not allowed. With the right gate voltage, the transistor channel is pushed into that gap and conduction nearly stops.
Graphene has no bandgap. Its conduction and valence bands meet at a point. A gate voltage can reduce graphene’s conductivity, but it cannot shut it off. Room-temperature on/off ratios for plain graphene transistors are small — far too low for logic. A chip built from such switches would leak current everywhere and overheat.
That single property is why graphene has not replaced silicon.
The Workarounds — and Their Costs
Researchers have tried several ways to open a bandgap. Each works to a degree, and each sacrifices something.
Nanoribbons. Cutting graphene into very narrow strips, a few nanometres wide, confines the electrons and opens a gap. But the gap depends critically on width and edge structure, and rough edges scatter electrons, destroying the high mobility that made graphene attractive. Making billions of identical ribbons with atomically precise edges is not yet possible at manufacturing scale.
Bilayer graphene under an electric field. Two stacked layers can develop a tunable gap when a strong perpendicular field is applied. The gap is small, and the device structure is complex.
Chemical modification. Hydrogenating or oxidising graphene opens a gap but converts carbon atoms from sp² to sp³, which degrades the electronic properties that made graphene special.
Substrate engineering. In 2024, a Georgia Tech-led team reported a semiconducting form of epitaxial graphene grown on silicon carbide, with a usable bandgap and good mobility. It is a significant research result. Turning it into a manufacturable technology competitive with silicon remains a long road.
The general pattern holds: making graphene behave like a semiconductor tends to remove what made it better than a semiconductor.
Silicon Kept Moving
Even if a perfect graphene switch existed, it would compete against an industry that has spent decades and enormous capital refining silicon processes. New transistor architectures and advanced packaging have continued to improve chip performance. A replacement material must not only work but justify rebuilding a manufacturing ecosystem.
This is why the semiconductor industry’s interest in two-dimensional channel materials has shifted toward transition metal dichalcogenides such as molybdenum disulphide, which have a natural bandgap. Our article on materials beyond graphene covers them.
Where Graphene Is Entering Chips
Graphene’s future in electronics is not as the switch, but around it.
Interconnects. The copper wiring that connects transistors suffers rising resistance as dimensions shrink. Graphene and related carbon layers are being studied as capping layers, barrier layers, and interconnect materials to reduce resistance and improve reliability.
Analogue and high-frequency devices. Amplifiers and radio-frequency components do not need to switch fully off. Graphene’s high mobility suits some of these roles.
Sensors integrated with chips. Graphene Hall-effect sensors, gas sensors, and photodetectors can be built on top of conventional circuitry, adding sensing functions to silicon chips. Several companies are commercialising this approach.
Photonics. Graphene modulators and photodetectors integrated with silicon photonics are under active development for data communications.
Pilot lines. Europe’s 2D-Experimental Pilot Line, linked to the Graphene Flagship initiative, is working on processes for integrating 2D materials into semiconductor manufacturing — a recognition that integration, not discovery, is now the bottleneck.
Why This Matters for Non-Electronics Readers
The bandgap story is a useful lesson for anyone assessing graphene claims. A headline property — in this case electron mobility — can be extraordinary and still irrelevant if another property, here the ability to switch off, is missing. Many graphene applications that stalled did so for the same reason: one outstanding number, one disqualifying gap.
The question to ask of any proposed application is not “what is graphene best at?” but “what does this application require, and does graphene have all of it?”
This article is part of our Fundamentals series. For market context, see The Top 10 Graphene Companies to Watch in 2026.
