How quickly can a laser be made to jump between competing operating states? A team at the School of Physics at Peking University answered this for ultrahigh-quality microcavity lasers and found the answer obeys a simple rule. Switching time scales approximately as the inverse square root of the speed at which the control parameter is swept, an exponent close to 0.5, which sets a practical speed limit on reconfiguring such devices.
The experimental system is a silica microsphere only tens of micrometres across, with a quality factor high enough that photons circulate many millions of times before leaking out. Clockwise and counter-clockwise travelling waves couple to form two competing standing-wave states of opposite symmetry. A feedback loop reinjects a small portion of the laser output, and controlling its phase tunes the loss balance between the two states, sweeping the system across a critical point.
Resolving the transient required care. The team used a radio-frequency beat-note method, mixing the laser output with a stable reference, achieving sub-ten-nanosecond time resolution, which was necessary because the switching itself can unfold on nanosecond timescales. Repeating the measurement at different sweep speeds gave a well-defined transition time for each event: faster sweeps switch faster, but with diminishing returns that follow the power law.
The same scaling behaviour appears in studies of coupled-cavity laser networks, which suggests a broader principle governing non-equilibrium switching in driven, dissipative photonic systems rather than a quirk of this geometry. Published in Physical Review Letters, the result bears both on reconfigurable microlasers for on-chip photonics and on coupled laser networks proposed for optimisation and analogue computing.