We have demonstrated a room temperature, low power consumption, high-speed, and single-mode tunable quantum cascade laser at λ ∼ 4.6 µm. Combining a 300 µm buried hetero-structure ultra-short cavity with optimized facet coatings, a single-mode tuning range covering ∼40 nm is obtained at room temperature under low electrical power consumption below 1.1 W. The single-mode emission is observed with more than 25 dB side mode suppression ratio in the whole tuning range. The high frequency modulation is achieved with a microwave package, resulting in the −3 dB cutoff frequency of 5.6 GHz.

Quantum cascade lasers (QCLs) are unipolar light sources based on intersubband transition.1 Their subband structure can be controlled by tuning the composition and thickness of the active superlattice materials. Therefore, compared with the traditional interband semiconductor lasers, QCLs show two remarkable and special features. One is the large spectral tuning range covering from mid-infrared to terahertz.2–5 Another important characteristic of QCLs is their capability for fast modulation because of ultra-short carrier lifetime limited by phonon assisted intersubband scattering.6,7 A number of particular applications were triggered by these intrinsic advantages, such as the laser spectroscopic trace-gas sensing and high bit-rate free-space optical communication (FSOC).8–12 Chirped laser dispersion spectroscopy (CLaDS) utilizes the optical dispersion effect in the vicinity of a resonant transition of the target gas molecule to extract useful signals,9 which make this detection scheme realizable for inherent linear response to the sample concentration, immune to source intensity-fluctuations, and capable of baseline-free detection. QCL based CLaDS is very attractive and efficient for remote gas sensing applications. On the other hand, due to the robust transmission characteristics in mid- and far-infrared atmospheric windows, QCLs are the most promising light sources for high-speed FSOC. Therefore, room temperature, high-frequency directly modulated single-mode QCLs are of particular interest for CLaDS and FSOC.

Clearly, QCL materials are characterized by intrinsic high-speed response, so the final direct modulation speed is usually determined by the chip packaging parasitics. The optimization of device structure and processing is much desired in order to fully exploit the high frequency properties of QCLs. High frequency modulation with different device packages for mid-infrared QCLs has been previously reported. A chalcogenide lateral waveguide QCL maintained a modulation response of up to roughly 7 GHz at a temperature of 20 K.13 A flat frequency response of up to ∼15 GHz for mid-infrared QCLs embedded into the microstrip line operating at 77 K was obtained.7 A single-mode distributed feedback (DFB) QCL integrated with a three-terminal microwave coplanar waveguide structure displayed a modulation bandwidth of up to 23.5 GHz (optical) and 25.5 GHz (electrical).14 A room temperature single-mode buried grating quantum cascade laser with a thick SiO2 insulating layer was demonstrated to have a flat frequency response of up to roughly 3 GHz.15 However, emergence of more promising technologies toward yielding room temperature high-speed QCLs with high single-mode tunability is still expected for advancing the development of CLaDS and FSOC.

In this paper, we present a high-speed single-mode tunable QCL with an ultra-short resonant cavity. By tuning the heat-sink temperature and injection current cooperatively, a single-mode tuning range covering ∼40 nm is obtained in the vicinity of room temperature with low electrical power consumption below 1.1 W. Measured by microwave rectification technology, the −3 dB cutoff frequency is 5.6 GHz with a parasitic resistance of 30 Ω and capacitance of 1.5 pF.

The QCL wafer was grown on an n-doped (Si, 2 × 1017 cm−3) InP substrate by solid source molecular beam epitaxy (MBE). The laser core was designed with a two phonon resonance scheme similar to that presented in Ref. 16. The layer sequence and average doping levels were as follows: a 1.2 µm InP lower cladding (Si, 2.2 × 1016 cm−3), a 0.3 µm In0.53Ga0.47As layer (Si, 4 × 1016 cm−3), 30 stages of an active/injector laser core (Si, 2.4 × 1016 cm−3), a 0.3 µm In0.53Ga0.47As layer (Si, 4 × 1016 cm−3), a 2.4 µm InP upper cladding (Si, 2.2 × 1016 cm−3), and a 0.6 µm InP cap cladding (Si, 1 × 1019 cm−3). After growth, buried hetero-structure processing with an average ridge width of 8 µm was performed for reducing parasitic capacitance and improving the heat transport. Selective-area regrowth of Fe-doped semi-insulating (SI) InP to planarize the channels was carried out in a metal–organic chemical vapor deposition (MOCVD) system.17 Following the regrowth, a 470-nm-thick SiO2 layer was deposited by chemical vapor deposition (CVD) for electrical insulation. A 5-μm-wide window was opened on top of the ridge for current injection, and electrical contact was provided by a Ti/Au layer. An additional 5-μm-thick gold layer was subsequently electroplated to further improve heat dissipation. The resulting buried ridge structure is illustrated in Fig. 1(a).

FIG. 1.

(a) Scanning electron microscope image of a buried hetero-structure ultra-short cavity QCL from the front facet, where AR represents the whole active region of the laser. (b) Photograph of our microwave package of an ultra-short cavity QCL.

FIG. 1.

(a) Scanning electron microscope image of a buried hetero-structure ultra-short cavity QCL from the front facet, where AR represents the whole active region of the laser. (b) Photograph of our microwave package of an ultra-short cavity QCL.

Close modal

The wafer was cleaved into 300-μm-long bars to pursue high frequency modulation and single longitudinal mode operation. Due to the small gain area and large mirror loss, it is very important to optimize the facet coating parameters for ultra-short cavity QCLs. The high-reflectivity (HR) coating consists of Al2O3/Ti/Au/Al2O3 (200/10/100/120 nm, with a reflectivity of ∼100%) and Al2O3/Ge (650/350 nm, with a reflectivity of ∼82%) deposited on the back and front facets, respectively, to tune the mirror losses and ensure a high mode gain. The laser is mounted epilayer side down on a coplanar waveguide (CPW) SiC transition heat-sink with indium solder. To reduce the influence of the external circuit, a 50 Ω CPW transmission line is designed to connect the laser and a 50 Ω SMA adapter to deliver the power of the RF signal to the laser, just as shown in Fig. 1(b). In this design, by shortening the cavity length to increase the longitudinal mode spacing (>5 cm−1), the mode gain is tuned and single-mode operation is guaranteed. The precisely designed HR coatings will reduce the mode loss and ensure the device to work at room temperature. On the other hand, the buried hetero-structure waveguide and ultra-short cavity help reduce the parasitics of the chip for high-frequency response.

Device characterizations were performed on an open-air copper stage outfitted with thermoelectric temperature stabilization. The output power was measured using a polished metallic pipe guiding the light from the laser front facet directly onto the surface of a calibrated thermopile detector. Figure 2 shows the continuous wave (cw) power-current–voltage (P–I–V) curves for this device at various heat-sink temperatures. A cw output power of 14 mW and 1.5 mW with wall-plug efficiencies (WPE) of 13% and 1.4% was obtained at 283 K and 298 K, respectively. At a heat-sink temperature of 293 K, a typical optical power more than 5 mW and WPE higher than 6% was demonstrated under the electrical power consumption less than 1 W. The threshold current density Jth was 2.38 kA/cm2 at 298 K corresponding to a slope efficiency of 300 mW/A. At 283 K, the Jth was as low as 1.72 kA/cm2 and the slope efficiency was 535 mW/A. The effective width of the active region is considered in the calculation of Jth. The temperature dependence of the Jth leads to a characteristic temperature of T0 = 45.4 K in the vicinity of room temperature. To avoid damage, lasers were not tested to their maximum operating current in this configuration. According to the I–V curves, the differential series resistances near the threshold of this ultra-short cavity laser are ∼40 Ω–50 Ω at temperatures between 283 K and 298 K. Although the larger differential resistance increases the parasitic parameters of the chip to a certain extent, it can achieve almost perfect impedance matching with most application systems with a characteristic impedance of 50 Ω.

FIG. 2.

Typical cw P–I–V characteristics of the ultra-short cavity QCL tested at various heat-sink temperatures.

FIG. 2.

Typical cw P–I–V characteristics of the ultra-short cavity QCL tested at various heat-sink temperatures.

Close modal

The frequency tunability of the ultra-short cavity QCL was studied by the emission spectra at different heat-sink temperatures and injection currents. The spectral characterization was performed at a resolution of 0.5 cm−1 with a Fourier transform infrared spectrometer equipped with a deuterated triglycine sulfate (DTGS) detector. As shown in Fig. 3(a), a single-mode emission with a side mode suppression ratio (SMSR) of ∼25 dB is observed in the whole current and heat-sink temperature range. The frequency shift of the same laser as a function of input electrical power in the temperature range from 283 K to 298 K is shown in Fig. 3(b). The lasing frequency varies with the input electrical power due to heating effects. The correlation between lasing frequency and input electrical power is ν = ν0 + βνTsink + βνPelecRth, where β=1/νΔν/ΔT is the thermal tuning coefficient, Rth is the thermal resistance, Pelec is the electrical power, and ΔT is the temperature difference between the active region and the heat-sink.18 By fitting the sets of experimental data using the above function and extracting from the two of them at 283 K and 287.5 K, the values of Rth = 5.65 K/W and β = −7 × 104 K−1 are estimated. The corresponding specific thermal conductance is Gth = 1/(RthA) = 7374.6 W/K cm2, where A is the laser surface area. The high thermal conductance value is mainly due to the effective heat dissipation process including the high thermal conductivity SI InP regrowth, thick electroplated gold layer, and optimized device geometry. The value of the tuning coefficient is much larger than the typical value of the order of −6 × 105 K−1, which comes only from the temperature-induced refractive index change.19 We believe that the thermal induced deformation plays a more important role in mode selection in the ultra-short FP cavity QCLs. However, in the DFB mode, this effect is usually negligible relative to the change in refractive index. The larger temperature tuning coefficient also leads to mode hopping during the heat-sink temperature tuning when increasing from 287.5 K to 293 K, as shown in Fig. 3(b). A broad range of single-mode tuning can be achieved by the combination of coarse tuning of heat-sink temperature and fine tuning of injection current. As shown in Fig. 3, a single-mode tuning range covering ∼40 nm is obtained in the vicinity of room temperature with low electrical power consumption below 1.1 W.

FIG. 3.

(a) Single-mode cw spectra at different heat-sink temperatures and injection currents. (b) Lasing frequency changes as a function of input electrical power (symbols) for the temperature range of 283 K–298 K. The solid line is the fitting result using the function of ν = ν0 + βνTsink + βνPelecRth.

FIG. 3.

(a) Single-mode cw spectra at different heat-sink temperatures and injection currents. (b) Lasing frequency changes as a function of input electrical power (symbols) for the temperature range of 283 K–298 K. The solid line is the fitting result using the function of ν = ν0 + βνTsink + βνPelecRth.

Close modal

To examine the high frequency response of the QCLs, we applied a microwave rectification technology.20 In this method, the rectified voltage relies on the inherent nonlinear V–I characteristics of the devices. The experimental setup is shown in the inset of Fig. 4, and the QCL device is modeled by a simplified RC-circuit. In this measurement, a microwave signal and a direct current (DC) bias are injected into the laser through a bias-T. The laser is DC biased at different currents (298 K), while the RF power is kept at 7 dBm. 50 kHz amplitude modulation is added to the RF generator by an additional arbitrary function generator while sweeping from 100 MHz to 6 GHz in steps of 100 MHz. The variations of the DC voltage are recorded with the help of a lock-in amplifier. Figure 4 shows the normalized electrical rectification curves as a function of the modulation frequency at different DC biases ranging from below to above lasing threshold Ith = 68 mA. A strong dependence of modulation bandwidth on the DC bias is observed. It is clear that the curves remain relatively flat up to roughly 5 GHz at bias-currents above the threshold and the widest band is obtained for the highest DC bias. It is in good agreement with other reports,7,14 where similar experiments were carried out on different device packages. When fitting the measurement with the theoretical curves from the rectification-model,14 as shown in the dashed line in Fig. 4, the −3 dB cutoff frequency is about 5.6 GHz at DC = 81 mA, which is 19% larger than the threshold current. The parameters for the fit were listed as RF power PRF = 5 mW, transmission line resistance RL = 50 Ω, the second derivative of the voltage–current curve V″ = −200 V2/A2, the differential resistance of the QCL R = 30 Ω, and the capacitance C = 1.5 pF. Although the differential resistance increases with the decrease in the device size, the ultra-short cavity quantum cascade laser in a buried hetero-structure achieves high bandwidth operation mainly due to the very small parasitic capacitance level.

FIG. 4.

Normalized electrical rectification data of an ultra-short cavity QCL at different bias-currents between IDC = 38 mA (below lasing threshold Ith = 68 mA) and 81 mA (above the threshold). The dashed line is the fitting result using the rectification-model in Ref. 14. The inset figure is the experimental setup for a microwave rectification measurement.

FIG. 4.

Normalized electrical rectification data of an ultra-short cavity QCL at different bias-currents between IDC = 38 mA (below lasing threshold Ith = 68 mA) and 81 mA (above the threshold). The dashed line is the fitting result using the rectification-model in Ref. 14. The inset figure is the experimental setup for a microwave rectification measurement.

Close modal

In conclusion, we have presented high-speed operation of a single-mode tunable quantum cascade laser at λ ∼ 4.6 µm based on a buried hetero-structure ultra-short resonant cavity. A broad single-mode tuning range covers ∼40 nm at room temperature under low electrical power consumption below 1.1 W. High frequency modulation is achieved with the −3 dB cutoff frequency of 5.6 GHz. All these make it a very suitable source for applications in the field of high bit-rate FSOC and high efficiency CLaDS.

This work was supported by the National Natural Science Foundation of China under Grant Nos. 61835011, 61991430, and 61674144 and the Key Program of the Chinese Academy of Sciences under Grant Nos. XDB43000000 and QYZDJ-SSW-JSC027. The authors would like to thank Ping Liang, Ying Hu, and Ke Yang for their help in device processing and measurement.

The data that support the findings of this study are available from the corresponding author upon reasonable request.

1.
J.
Faist
,
F.
Capasso
,
D. L.
Sivco
,
C.
Sirtori
,
A. L.
Hutchinson
, and
A. Y.
Cho
,
Science
264
,
553
(
1994
).
2.
O.
Cathabard
,
R.
Teissier
,
J.
Devenson
,
J. C.
Moreno
, and
A. N.
Baranov
,
Appl. Phys. Lett.
96
,
141110
(
2010
).
3.
M.
Razeghi
,
Proc. SPIE
11296
,
11296C
(
2020
).
4.
R.
Colombelli
,
F.
Capasso
,
C.
Gmachl
,
A. L.
Hutchinson
,
D. L.
Sivco
,
A.
Tredicucci
,
M. C.
Wanke
,
A. M.
Sergent
, and
A. Y.
Cho
,
Appl. Phys. Lett.
78
(
18
),
2620
(
2001
).
5.
R.
Köhler
,
A.
Tredicucci
,
F.
BEltram
,
H. E.
Beere
,
E. H.
Linfield
,
A. G.
Davies
,
D. A.
Ritchie
,
R. C.
Iotti
, and
F.
Rossi
, “
Terahertz semiconductor-heterostructure laser
,”
Nature
417
,
156
(
2002
).
6.
R.
Paiella
,
R.
Martini
,
F.
Capasso
,
C.
Gmachl
,
H. Y.
Hwang
,
D. L.
Sivco
,
J. N.
Ballargeon
,
A. Y.
Cho
,
E. A.
Whyttaker
, and
H. C.
Liu
,
Appl. Phys. Lett.
79
,
2526
(
2001
).
7.
A.
Calvar
,
M. I.
Amanti
,
M.
Renaudat St-Jean
,
S.
Barbieri
,
A.
Bismuto
,
E.
Gini
,
M.
Beck
,
J.
Faist
, and
C.
Sirtori
,
Appl. Phys. Lett.
102
,
181114
(
2013
).
8.
J. H.
van Helden
,
N.
Lang
,
U.
Macherius
,
H.
Zimmermann
, and
J.
Röpcke
,
Appl. Phys. Lett.
103
,
131114
(
2013
).
9.
A.
Hangauer
,
G.
Spinner
,
M.
Nikodem
, and
G.
Wysocki
,
Appl. Phys. Lett.
103
,
191107
(
2013
).
10.
R.
Martini
,
C.
Gmachl
,
J.
Falciglia
,
F. G.
Curti
,
C. G.
Bethea
,
F.
Capasso
,
E. A.
Whittaker
,
R.
Paiella
,
A.
Tredicucci
,
A. L.
Hutchinson
,
D. L.
Sivco
, and
A. Y.
Cho
,
Electron. Lett.
37
(
3
),
191
(
2001
).
11.
P.
Pang
,
O.
Ozolins
,
J.
Falciglia
,
R.
Schatz
,
J.
Storck
,
A.
Udalcovs
,
J. R.
Navarro
,
A.
Kakkar
,
G.
Maisons
,
M.
Carras
,
G.
Jacobsen
,
S.
Popov
, and
S.
Lourdudoss
,
Opt. Lett.
42
(
18
),
3646
(
2017
).
12.
A.
Delga
and
L.
Leviandier
,
Proc. SPIE
10926
,
1092617
(
2019
).
13.
R.
Martini
,
R.
Paiella
,
C.
Gmachl
,
F.
Capasso
,
E. A.
Whittaker
,
H. C.
Liu
,
H. Y.
Hwang
,
D. L.
Sivco
,
J. N.
Baillargeon
, and
A. Y.
Cho
,
Electron. Lett.
37
(
21
),
1290
(
2001
)
14.
B.
Hinkov
,
A.
Hugi
,
M.
Beck
, and
J.
Faist
,
Opt. Express
24
(
4
),
3294
(
2016
).
15.
Y.
Zhou
,
S.
Zhai
,
J.
Liu
,
F.
Liu
,
J.
Zhang
,
N.
Zhuo
,
L.
Wang
, and
Z.
Wang
,
Electron. Lett.
52
(
7
),
548
(
2016
).
16.
F. Q.
Liu
,
L.
Li
,
L. J.
Wang
,
J. Q.
Liu
,
W.
Zhang
,
Q. D.
Zhang
,
W. F.
Liu
,
Q. Y.
Lu
, and
Z. G.
Wang
,
Appl. Phys. A
97
,
527
(
2009
).
17.
A.
Evans
,
S. R.
Darvish
,
S.
Slivken
,
J.
Nguyen
,
Y.
Bai
, and
M.
Razeghi
,
Appl. Phys. Lett.
91
,
071101
(
2007
).
18.
S.
Blaser
,
D. A.
Yarekha
,
L.
Hvozdara
,
Y.
Bonetti
,
A.
Muller
,
M.
Giovannini
, and
J.
Faist
,
Appl. Phys. Lett.
86
,
041109
(
2005
).
19.
J. C.
Zhang
,
F. Q.
Liu
,
S.
Tan
,
D. Y.
Yao
,
L. J.
Wang
,
L.
Li
,
J. Q.
Liu
, and
Z. G.
Wang
,
Appl. Phys. Lett.
100
,
112105
(
2012
).
20.
H. C.
Liu
,
J.
Li
,
M.
Buchanan
, and
Z. R.
Wasilewski
,
IEEE J. Quantum Electron.
32
(
6
),
1024
(
1996
).