Investigation of Thermal Effects in Quantum-Cascade Lasers

Self Cite

Terahertz-frequency quantum cascade lasers (THz QCLs) based on bound-to-continuum active regions are difficult to model owing to their large number of quantum states. We present a computationally efficient reduced rate equation (RE) model that reproduces the experimentally observed variation of THz power with respect to drive current and heat-sink temperature. We also present dynamic (time-domain) simulations under a range of drive currents and predict an increase in modulation bandwidth as the current approaches the peak of the light–current curve, as observed experimentally in mid-infrared QCLs. We account for temperature and bias dependence of the carrier lifetimes, gain, and injection efficiency, calculated from a full rate equation model. The temperature dependence of the simulated threshold current, emitted power, and cut-off current are thus all reproduced accurately with only one fitting parameter, the interface roughness, in the full REs. We propose that the model could therefore be used for rapid dynamical simulation of QCL designs.

mentioning

confidence: 99%

Efficient prediction of terahertz quantum cascade laser dynamics from steady-state simulations

Agnew

Grier

Taimre

et al. 2015

Self Cite

“…The wavelength tuning dynamics is also of prime interest for the understanding of the underlying mechanism of frequencynoise generation and linewidth broadening in free-running QCLs. 15,16 Moreover, while the thermal resistance and the heat extraction of pulsed mid-IR [17][18][19] and THz QCLs 20 have been deeply investigated, our measurements allow us to discuss the dynamic of thermal effects in continuous-wave (CW) DFB-QCLs under direct current modulation. A simple thermal model allows us to fit the experimental results and extract the different thermal time constants involved in the system.…”

mentioning

confidence: 99%

“…Moreover, these experimental thermal time constants are on the same order of magnitude as the experimental values reported for a THz QCL 20 and simulation results obtained for pulsed QCLs. 19 The highest cut-off frequency, on the order of f 1 ¼ $200 kHz (corresponding to a time constant s 1 ¼ 1/2pf 1 shorter than 1 ls), reflects the heat dissipation in the small volume of the active region itself, along the planes of the heterostructure. The second characteristic frequency f 2 ¼ $20 kHz likely corresponds to the heating due to heat extraction perpendicular to the planes of the heterostructure and through the waveguide layers.…”

mentioning

confidence: 99%

Wavelength tuning and thermal dynamics of continuous-wave mid-infrared distributed feedback quantum cascade lasers

Tombez

Cappelli

Schilt

et al. 2013

We report on the wavelength tuning dynamics in continuous-wave distributed feedback quantum cascade lasers (QCLs). The wavelength tuning response for direct current modulation of two mid-IR QCLs from different suppliers was measured from 10 Hz up to several MHz using ro-vibrational molecular resonances as frequency-to-intensity converters. Unlike the output intensity, which can be modulated up to several gigahertz, the frequency-modulation bandwidth was found to be on the order of 200 kHz, limited by the laser thermal dynamics. A non-negligible roll-off and a significant phase shift are observed above a few hundred hertz already and explained by a thermal model.Since the first demonstration of quantum cascade lasers (QCLs) in 1994, 1 the number of promising application in the field of chemical sensing for biomedical and environmental sciences has been constantly rising during the last years. Indeed, thanks to the ability to tailor their emission wavelength and to target precisely selected ro-vibrational molecular transitions in the fingerprint region, QCLs were demonstrated to be very sensitive probes for a wide variety of molecules.2 Single-frequency QCLs are generally required for trace gas sensing and a common approach is to use a distributed feedback (DFB) grating etched at the surface of the QCL active region in order to force laser operation at a precise wavelength.3 Moreover, promising developments in the field of high-resolution spectroscopy can lead to the possibility of performing measurements with unprecedented precision, especially by linking spectrally narrow QCLs to optical frequency combs. 4,5 In order to push the limits of those highresolution experiments, narrow-linewidth sources of coherent light which can be achieved by active stabilization of DFB-QCLs to optical references with high-bandwidth servoloops are required. 6,7 For the most demanding applications in the field of high-resolution spectroscopy, frequency-noise analysis revealed that feedback loop bandwidths of several hundred of kHz are necessary for linewidth narrowing of DFB-QCLs. 7,8 While the picosecond carrier lifetime in QCLs allows a very fast modulation of the intensity above 10 GHz, 9,10 the modulation of the optical frequency-or wavelength-is limited by the thermal dynamics of the device. Indeed, unlike interband semiconductor laser diodes whose wavelength can be modulated at high speed through carrier density modulation, 11,12 the latter has no effect in QCLs because of the symmetric gain curve and associated independence of the refractive index at the gain peak (zero alpha parameter). 13 The wavelength tuning in DFB-QCLs is therefore mainly governed by the temperature dependence of the refractive index with a tuning rate of 1/k dk/dT % 7 Â 10 À5

“…Although PL measurements show a much steeper temperature gradient across the AR than the substrate during continuous-wave operation, 16 theoretical work predicts a more complex situation in pulsed mode. 7 In this case, the temperature difference between the top and bottom of the AR fluctuates rapidly within each pulse, while the temperature at the bottom of the AR increases gradually over many pulses owing to the large heat capacity of the substrate. Since we use the THz emission as the thermometric property, the absolute value of T AR,n would represent both a temporal average (over the duration of the n th pulse) and a spatial average of the temperature across the active-region (weighted by the photon emission at each point).…”

mentioning

confidence: 97%

“…However, the thermal relaxation path involves several regions with significant variation in heat capacity and thermal conductance, each being characterized by a different thermal time-constant. This results in a complicated time-variation in device temperature when driven by a a) Electronic mail: a.valavanis@leeds.ac.uk pulsed electrical source, 7,8 and a range of thermal characterization techniques have been used to investigate such thermal effects experimentally. The steady-state thermal resistance has been inferred by measuring the timeaveraged THz power using a pyroelectric detector.…”

mentioning

confidence: 99%

Time-resolved measurement of pulse-to-pulse heating effects in a terahertz quantum cascade laser using an NbN superconducting detector

Valavanis

Dean

Scheuring

et al. 2013

EH (2013) Time-resolved measurement of pulse-to-pulse heating effects in a terahertz quantum cascade laser using an NbN superconducting detector. Applied Physics Letters, 103 (6). 061120-1 -061120-4 (4).http://dx.doi.org/10.1063/1.4818584Time-resolved measurement of pulse-to-pulse heating effects in a terahertz quantum cascade laser using an NbN superconducting detector Joule heating causes significant degradation in the power emitted from terahertz-frequency quantum-cascade lasers (THz QCLs). However, to date, it has not been possible to characterize the thermal equilibration time of these devices, since THz power degradation over sub-millisecond time-scales cannot be resolved using conventional bolometric or pyroelectric detectors. In this letter, we use a superconducting antenna-coupled niobium nitride detector to measure the emission from a THz QCL with a nanosecond-scale time-resolution. The emitted THz power is shown to decay more rapidly at higher heat-sink temperatures, and in steadystate the power reduces as the repetition rate of the driving pulses increases. The pulse-to-pulse variation in active-region temperature is inferred by comparing the THz signals with those obtained from low duty-cycle measurements. A thermal resistance of 8.2 ± 0.6 K/W is determined, which is in good agreement with earlier measurements, and we calculate a 370 ± 90-µs bulk heat-storage time, which corresponds to the simulated heat capacity of the device substrate.