Time-Resolved dynamics of semiconductor nanolaser via four-wave mixing gating

We experimentally demonstrate the direct time-domain characterization of photonic-crystal nanolasers at telecom wavelengths using a nonlinear optical gating technique based on four-wave mixing. This approach enables the temporal characterization of the ultrafast emission dynamics under short-pulse excitation with picosecond time resolution. When a weak continuous-wave component is added to the pulsed pump, the emission becomes deterministic and the build-up time is considerably reduced. The difference between purely pulsed and hybrid excitation regimes points to the influence of pulse-to-pulse timing fluctuations. To elucidate this effect, we perform Langevin-based simulations that reproduce the experimentally observed broadening and confirm that time jitter, originating from spontaneous-emission noise near threshold, dominates the temporal dispersion. These results establish four-wave-mixing gating as a powerful method to probe nanolaser dynamics with picosecond precision.

💡 Research Summary

In this work the authors present a direct time‑domain investigation of photonic‑crystal nanolasers operating at the telecom wavelength of 1550 nm, using a nonlinear optical gating technique based on degenerate four‑wave mixing (FWM). Conventional infrared detectors suffer from limited sensitivity and bandwidth, preventing accurate measurement of the tens‑of‑picosecond build‑up dynamics that characterize high‑β nanolasers. By overlapping a short‑duration “gate” pulse with the nanolaser emission inside a dispersion‑shifted fiber, an idler photon at ωi = 2 ωg − ωs is generated. Because the idler intensity scales with the temporal overlap of gate and signal, recording the idler as a function of relative delay yields a reconstruction of the laser’s temporal profile with an effective resolution of ≈1.9 ps (the square‑root of the 3.5 ps gate pulse).

The device under test is a 1‑D InP‑based nanobeam cavity (650 nm × 285 nm) incorporating four InGaAsP quantum wells and a graded‑lattice photonic crystal that confines a mode near 1550 nm with a quality factor of 10⁵–10⁷. The cavity is heterogeneously integrated on a silicon‑on‑insulator waveguide, allowing out‑coupling into a standard optical fiber. Under continuous‑wave (CW) pumping at 1180 nm the laser exhibits a threshold of ≈250 µW, while under 250 fs pulsed excitation the threshold drops to ≈4.5 µW.

Time‑resolved measurements with pure pulsed pumping reveal a build‑up time that shortens from ~80 ps near threshold to ~34 ps at high pump powers, and a decay time that reduces from ~60 ps to ~30 ps. However, near the threshold the emission is strongly influenced by spontaneous‑emission noise, leading to large pulse‑to‑pulse timing jitter (≈85 ps). Averaging many pulses therefore artificially broadens the measured temporal profile.

To suppress this jitter the authors add a weak CW component (≈250 µW, i.e., at the CW threshold) together with the pump pulse. In this hybrid excitation regime the laser turns on deterministically, the build‑up delay collapses to ~15 ps, and the temporal profiles become much sharper. The peak intensity rises immediately with pulse power, the build‑up time drops from ~50 ps to ~20 ps, and the decay time saturates around 35 ps.

For quantitative interpretation the authors employ a set of semiconductor laser rate equations that include a Gaussian pump term, carrier‑saturation, non‑radiative and radiative recombination, and the spontaneous‑emission factor β ≈ 0.13. By fitting all measured time traces simultaneously they extract physical parameters such as photon lifetime (τph ≈ 25 ps), carrier transparency density, and saturation carrier density. The deterministic model reproduces the CW‑pumped steady‑state curve and the pulsed dynamics when noise is neglected.

To capture the observed jitter they augment the model with a Langevin noise term, treating spontaneous emission as a stochastic diffusion process with coefficient D = ½ n(t)² β τnr τph / τrad. They run 10 000 stochastic simulations for each pump level. The simulations show that above the pulsed threshold the distribution of build‑up times narrows dramatically, and the ensemble‑averaged trace converges to the deterministic solution. The extracted jitter decreases abruptly from ~85 ps at low power to ~5 ps just above threshold, confirming that spontaneous‑emission fluctuations dominate the temporal dispersion near threshold.

Overall, the study demonstrates that four‑wave‑mixing gating provides picosecond‑scale temporal resolution for infrared nanolasers without requiring ultrafast detectors. It also clarifies how spontaneous‑emission noise governs the build‑up dynamics and timing jitter of high‑β nanolasers, and shows that a modest CW seed can render the emission deterministic and dramatically improve modulation speed. These insights are directly relevant to the design of ultra‑low‑energy on‑chip light sources for high‑speed optical interconnects, quantum photonic circuits, and other applications where precise control of nanolaser dynamics is essential.