
Graphene Instead of Silicon? Simulations From Kiel Show Light-Controlled Electrons in the Femtosecond Range
Key Takeaways
- •Epitaxial graphene on SiC shows 0.6 eV bandgap, 5,000 cm²/Vs mobility.
- •Simulations reveal femtosecond, spatially selective electron excitation in nanoribbons.
- •Potential petahertz switching could outpace current gigahertz silicon circuits.
- •Manufacturing, defect control, and integration remain major hurdles.
- •Graphene may complement, not replace, silicon in future nanoelectronics.
Pulse Analysis
The semiconductor roadmap is hitting the wall of diminishing returns on classic Moore’s Law scaling. As transistor dimensions shrink below 5 nm, lithography costs soar and power density spikes, prompting manufacturers to explore alternative materials. Graphene, a single‑atom‑thick carbon lattice, offers carrier mobilities an order of magnitude higher than silicon, but its lack of an intrinsic bandgap has long blocked digital logic. The January 2024 Nature paper from Georgia Tech and Tianjin University shows epitaxial graphene on silicon carbide achieving a 0.6 eV bandgap while keeping mobility above 5,000 cm² V⁻¹ s⁻¹, bringing graphene nearer to practical transistors.
Separately, the September 2025 simulation study from the University of Kiel demonstrates that ultrashort laser pulses can launch localized electron and exciton excitations within graphene nanoribbons on a femtosecond timescale. By exploiting the material’s topological states, a uniform light field produces spatially selective charge separation, effectively acting as an optical switch that could operate in the petahertz regime—roughly ten thousand times faster than today’s gigahertz silicon transistors. While the results are currently limited to computational models, they reveal a pathway to combine photonics and nanoelectronics, potentially reducing the energy overhead of voltage‑driven switching.
Turning these laboratory insights into commercial chips will require solving three hard problems: reproducible large‑scale graphene synthesis with low defect density, integration of optical control mechanisms into existing fab lines, and a cost structure that can compete with silicon’s multi‑billion‑dollar ecosystem. Even if those hurdles are cleared, graphene‑based logic is likely to appear first in niche roles—high‑frequency interconnects, photonic modulators, or specialized AI accelerators—where its speed and thinness offset integration complexity. For the broader data‑center market, silicon will dominate for the foreseeable decade, but sustained research on bandgap‑engineered graphene and femtosecond switching keeps the material on the industry’s radar as a potential long‑term enabler of ultra‑low‑power, petahertz nanoelectronics.
Graphene instead of silicon? Simulations from Kiel show light-controlled electrons in the femtosecond range
Comments
Want to join the conversation?