Gain recovery dynamics of a terahertz quantum cascade laser

The gain recovery time of a bound-to-continuum terahertz quantum cascade laser is measured using a stimulated emission photocurrent technique. At low temperatures a value $\sim 50\text{ps}$ is found. This tends to decrease with rising temperature, and suggests that class B laser dynamics should be expected from these devices. Both the value and the trend in the gain recovery time are found to be consistent with hopping transport through the miniband. Power-dependent measurements of the photocurrent clearly show the effects of inhomogeneous broadening of the gain transition.

Figure 1

Schematic showing the two-beamsplitter method used to obtain the variable delay between the two collinear, copolarized pulses used in the experiment. Components solely used to equalize the length of the two paths have been omitted for clarity.

Figure 2

(a) Photocurrent magnitude measured as a function of the applied bias current, using a FELIX peak power of $\sim 3\text{kW}$ at $\lambda \sim 94\mu \text{m}$. The device temperature was $\sim 20\text{K}$. (b) Continuous-wave $I\text{−}V$ characteristic measured from the device under the same operating conditions.

Figure 3

Calculated band diagrams for the structure under (a) zero bias and (b) expected operating bias, showing ${\left|\Psi \right|}^2$ for selected wave functions. Bold lines indicate the energy levels involved in radiative transitions.

Figure 4

A typical trace of photocurrent against delay time between the two beams, measured at 300 mA bias current. The thick gray line shows a single exponential fit to the negative delay data. Letters indicate different regions of the curve as detailed in the text.

Figure 5

(a) Gain recovery times measured as a function of the bias current. The measurements were carried out using $94\mu \text{m}$ laser excitation at a power of $\sim 1\text{kW}$. (b) Gain recovery times measured as a function of temperature. A 300 mA bias current was used, with a FELIX peak power of $\sim 3\text{kW}$.

Figure 6

(a) Symbols represent the measured photocurrent as a function of FELIX peak power, at $\lambda =94\mu \text{m}$, 300 mA bias current and 20 K device temperature. The line is a fit to the data using the expression for inhomogeneous broadening and a saturation power of 240 mW. (b) Arrhenius plot of the photocurrent measured as a function of temperature. The same conditions were used, but the FELIX peak power was fixed at $\sim 1\text{kW}$. The dashed line represents a fit to the high-temperature data, using an activation energy of 5.5 meV.