Caltech chip steers light beam in 74 femtoseconds

September 18, 2026 at 05:03 PM2 min

This digest was compiled by AI from multiple sources — links to the originals are below.

Caltech researchers have developed a device that uses one light beam to redirect another in 74 femtoseconds. The team reports the result in Nature Nanotechnology, with lead author Claudio Hail now at UC Berkeley. The approach bypasses the electronic relaxation bottleneck that limits conventional light steering to nanosecond or picosecond timescales.

Key Facts

The device redirects a probe beam using a pump beam in 74 femtoseconds, roughly the time light takes to cross a human hair.
The work is published in Nature Nanotechnology with lead author Claudio Hail, formerly a postdoctoral scholar at Caltech and now an assistant professor at UC Berkeley.
Conventional light steering based on electronic excitation is limited to nanosecond or picosecond timescales because of electron relaxation.
The Caltech device uses the optical Kerr effect, where an intense pump beam briefly changes the refractive index of a material without exciting electrons into long-lived states.

Device Operation

The Caltech team eliminated the need for an electrical signal by using a powerful pump beam with a carefully designed pattern to temporarily alter the optical behavior of a material. A second, weaker probe beam then passed through that material and changed direction according to the pattern created by the pump beam. The system relies on the optical Kerr effect, in which an intense beam of light briefly produces a very small change in the material's refractive index. The change comes from electron motion within orbitals, not from pushing electrons into separate, longer-lasting excited states.

Speed Limit Comparison

Existing technologies for steering or modulating light depend on changing the electronic properties of a material, such as liquid-crystal panels in projectors and optical chips in telecommunications systems. In those devices, electrons are pushed into higher energy states and then relax, creating a bottleneck that limits light modulation to nanosecond or picosecond timescales. The Caltech approach avoids that relaxation bottleneck, allowing the change to appear and vanish almost as quickly as the light pulse itself.