| May 20, 2026 |
A silver silicon nanodisk antenna performs all-optical modulation in 37 femtoseconds, bypassing the electron-phonon bottleneck that limits plasmonic devices.
(Nanowerk News) Researchers at Xiamen University and Hangzhou Dianzi University have measured all-optical modulation times of 37 femtoseconds in a plasmonic device, breaking a speed limit set by the electron–phonon relaxation that has held plasmonic modulators to picosecond response. The silver–single-crystal silicon nanodisk antenna also delivered an on-off switching contrast above 100.
|
|
The work appears in Nano-Micro Letters (“Sub-100 Femtosecond All-Optical Modulation Beyond Electron–Phonon Limits”).
|
Key Findings
- The nanodisk antenna achieved modulation time constants of 37 ± 9 femtoseconds, outpacing the electron–phonon thermalization that normally limits plasmonic devices.
- Inserting a 30 nanometer aluminum oxide layer to block electron transfer reverted the device to picosecond-scale dynamics, confirming an interface-driven carrier extraction mechanism.
- The device produced an on-off switching ratio above 100, among the highest reported for plasmonic ultrafast systems.
|
The electron-phonon bottleneck
|
|
Plasmonic modulators are attractive for ultrafast optics because they manipulate light at deep-subwavelength scales and respond on extremely short timescales. Their speed has been capped by how quickly excited electrons can dump energy into the surrounding lattice, a process called electron–phonon relaxation that typically holds modulation in the picosecond range.
|
 |
| Engineered Ag–Si interfacial plasmonic modes enable deconvolution-verified sub-100 fs all-optical modulation, overcoming the temporal bottleneck imposed by electron–phonon relaxation in conventional plasmonic structures. A unified electromagnetic–thermal physical model reveals an interface-governed nonthermal carrier extraction pathway faster than electron–electron and electron–phonon scattering, quantitatively explaining the observed sub-100 fs dynamics. (Image: Reproduced from DOI:10.1007/s40820-026-02166-z, CC BY)
|
|
The relaxation also produces a persistent tail in the device response, dragging out recovery and lowering switching contrast even when the initial switching is fast. Eliminating that tail is the central problem for pushing plasmonic modulators into the femtosecond regime.
|
How the nanodisk antenna works
|
|
Corresponding authors Ming-De Li and Zhilin Yang and colleagues built a silver–single-crystal silicon nanodisk antenna they call an SSDMA. The structure places the metal–semiconductor boundary inside the same nanoscale volume where plasmonic energy is deposited, so excited carriers do not need to travel far before they can be extracted across the interface.
|
|
Confining the plasmonic mode this way shortens the transport path for hot carriers and activates an interfacial extraction route that operates before electron–phonon thermalization sets in. Photogenerated energy is pulled out of the metal as electronic charge transfer into the silicon, instead of being deposited as heat in the silver lattice. The transient optical response then tracks intrinsic electronic timescales rather than slower thermal ones.
|
Femtosecond pump-probe verification
|
|
The team verified the mechanism with femtosecond pump-probe spectroscopy. Deconvolution of the measured signals resolved modulation time constants of 37 ± 9 femtoseconds, well below the lattice-mediated relaxation seen in standard plasmonic structures. A unified electromagnetic–thermal model developed alongside the experiments reproduced the observed dynamics and identified the interface-governed nonthermal carrier extraction pathway as faster than both electron–electron and electron–phonon scattering.
|
|
To test whether the interface really was responsible, the researchers inserted a 30 nanometer aluminum oxide layer between the silver and the silicon. The insulating barrier blocked electron transfer across the junction, and the device response immediately slowed to the picosecond range characteristic of thermally limited plasmonic systems.
|
|
The contrast between the two cases established the interface as the source of the femtosecond speed. The unblocked device also produced an on-off switching ratio above 100, among the highest reported for plasmonic ultrafast platforms.
|
Applications for all-optical modulation at electronic limits
|
|
Femtosecond free-space photonic computing is one direct application, where the modulator can act as a high-speed optical logic element. The architecture also fits temporal optical gating, supplying the kind of fast shutter needed to capture transient phenomena that occur on similar timescales. The team additionally reports sub-100 femtosecond modulation across multiple discrete wavelengths, useful for multi-frequency processing.
|
|
The broader contribution is a physical foundation for ultrafast photonic systems whose performance is currently constrained by carrier or cavity lifetimes. Engineered interfacial plasmonic modes replace the usual reliance on bulk material response, giving designers a template for next-generation modulators governed by interfacial dynamics.
|