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

TL;DR

Time-domain characterization of nanolasers at telecom wavelengths is impeded by detector bandwidth and sensitivity. The paper introduces all-optical four-wave-mixing (FWM) gating with two synchronized OPAs to sample nanolaser emission via the idler at $\omega_i=2\omega_g-\omega_s$, achieving a temporal resolution of about $1.9~\mathrm{ps}$ and gating the emission with gate at $\lambda_g=1564~\mathrm{nm}$. Under pulsed pumping, the nanolaser shows spectral broadening with a blue tail and a threshold around $P_{\mathrm{th}}\approx 4.5$, while the build-up time decreases from $\sim 80~\mathrm{ps}$ to $\sim 34~\mathrm{ps}$ and decay time from $\sim 60~\mathrm{ps}$ to $\sim 30~\mathrm{ps}$ as power increases; adding a weak CW component reduces build-up time and imposes deterministic pulses. The authors fit a normalized rate-equation model with parameters $\beta$, $\tau_{\mathrm{ph}}$, $n_{\mathrm{tr}}$, $n_{\mathrm{sat}}$, and $R_0$, and perform Langevin-based simulations with diffusion $D=\frac{1}{2}n(t)^2\beta\tau_{\mathrm{nr}}\tau_{\mathrm{ph}}/\tau_{\mathrm{rad}}$ to quantify time jitter, finding it drops from about $85~\mathrm{ps}$ below threshold to $\sim 5~\mathrm{ps}$ above threshold. Overall, the work establishes FWM gating as a powerful, broadly applicable technique for probing ultrafast nanolaser dynamics with picosecond precision.

Abstract

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.

Figures (7)

• Figure 1: a. Schematics of the integrated nanolaser. b. SEM images of the photonic-crystal nanolaser. Top-view of the full PhC nanobeam (top), zoom-in top-view (left) and angled-view (right) of the nanobeam ending. c. Characterization of the laser curve under CW pumping. Experimental data (blue squares) and simulation obtained from the calibrated model (red), where $\beta=0.13$.
• Figure 2: Experimental configuration.a. Schematic of the FWM time-gating experimental setup. b. Measurement of the SOI waveguide output power as a function of the optical power of the pulsed pump, $P_\mathrm{in}$. c. Optical spectra of the output emission for $P_\mathrm{in}=$3.5 (green), 5 (blue), and 16 (red).
• Figure 3: Temporal response measurements for increasing pulsed pump powers.a. The nanolaser is pumped by 250 fs optical pulses. Data are fitted (gray lines) with the numerical integration of \ref{['eq:master']}. b. Extracted pulse amplitudes, c. pulses build up time, and d. pulses decay times. The same parameters obtained from the fits are shown in red.
• Figure 4: Effect of a CW pump power $P_\mathrm{cw}=\qty{250}{\micro\watt}$.a. The nanolaser is pumped by 250 fs optical pulses. Data are fitted (gray lines) with the numerical integration of \ref{['eq:master']}. b. Extracted pulse amplitudes, c. pulses build up time, and d. pulses decay times. The same parameters obtained from the fits are shown in red.
• Figure 5: Stochastic simulationsa. For $P_\mathrm{in}$ set to 3 (top), 4 (middle), and 5 (bottom), the nanolaser emission is simulated using the Langevin model. A hundred independent realizations (light grey curves) are shown with the average response (blue) and the deterministic simulation (black dashed). b. time jitter as a function of the excitation pump power. It is computed as the standard deviation of the build-up time obtained over 10,000 realizations for each power value.