Document [original]

This version is available at https://doi.org/10.14279/depositonce-10918

or electronic copy may be made for personal use only. Systematic reproduction and

distribution, duplication of any material in this publication for a fee or for commercial

purposes, and modification of the contents of the publication are prohibited.

Grillot, F., Huang, H., Lin, L.-C., Chen, C.-Y., Arsenijevic, D., Bimberg, D. & Lin, F.-Y. (2018). Recent

advances in InAs/GaAs quantum dot lasers with short optical feedback. Proc. SPIE 10682, Semiconductor

Lasers and Laser Dynamics VIII, 106820Y (9 May 2018). https://doi.org/10.1117/12.2314708.

Frédéric Grillot, Heming Huang, Lyu-Chih Lin, Chih-Ying Chen, Dejan

Arsenijevic, Dieter Bimberg, Fan-Yi Lin

Recent advances in InAs/GaAs quantum

dot lasers with short optical feedback

Accepted manuscript (Postprint)Conference paper |

Recent advances in InAs/GaAs quantum dot lasers with

short optical feedback

F. Grillota,b, H. Huanga, L. C. Linc, C. Y. Chenc, D. Arsenijevi`cd, D. Bimbergd,e, and F. Y.

Linc

aLTCI, T´el´ecom ParisTech, Universit´e Paris-Saclay, 46 rue Barrault, Paris, France

bCenter for High Technology Materials, University of New-Mexico, Albuquerque, USA

cInstitute of Photonics Technologies, Department of Electrical Engineering, National Tsing

Hua University, Hsinchu 300, Taiwan

dInstitut f¨ur Festk¨orperphysik, Technische Universit¨at Berlin, Berlin 10623, Germany

eChinese-German Green Photonics Center at the Chinese Academy of Sciences, CIOMP,

Changchun, China

ABSTRACT

The optical feedback dynamics of two multimode InAs/GaAs quantum dot lasers emitting exclusively on sole

ground or excited lasing states is investigated under the short delay configuration. Although the two lasers are

made from the same active medium, their responses to the external perturbation are found not much alike. By

varying the feedback parameters, various periodic and chaotic oscillatory states are unveiled. The ground state

laser is found to be much more resistant to optical feedback, benefiting from its strong relaxation oscillation

damping. In contrast, the excited state laser can easily be driven into very complex dynamics. While the ground

state laser is of importance for the development of isolator-free transmitters, the excited one is essential for

applications taking advantages of chaos such as chaos lidar, chaos radar, and random number generation.

Keywords: Quantum dot lasers, optical feedback, dynamical states

1. INTRODUCTION

The nonlinear dynamics of semiconductor lasers operating under external optical feedback has been widely stud-

ied in the literature.1Indeed, by controlling the feedback strength and the external cavity length, multiple

complex dynamical states and their routes to chaos can be observed.1Compared to quantum well (QW) lasers,

significant breakthroughs have been achieved by using quantum dots (QD) as gain media2.3Their inherent prop-

erties allow producing energy- and cost-efficient devices with outstanding temperature stability, low threshold

current, ultrafast gain dynamics, and low amplified spontaneous emission.3A low threshold current density and

high internal quantum efficiency results in a reduced amount of dissipated heat. Even at high operation temper-

atures, temperature insensitive threshold currents have been observed for p-doped devices.4Most advantages

of QD lasers have been demonstrated for InAs/GaAs QD lasers operating at 1310 nm, hence mostly targeting

short communication links such as metro and access networks. QD lasers usually show three possible regimes of

lasing operation, depending on the bias conditions: (i) ground state (GS) lasing; (ii) dual state emission showing

an interplay dynamic between the GS and the first excited state (ES), and (iii) ES lasing5.6While feedback

dynamics of QD lasers has been widely studied in particular in the context of the dual-state dynamics,7none of

this work focused on the feedback dynamics of QD devices emitting exclusively on single lasing states. Here, we

report for the first time on the feedback dynamics of two InAs/GaAs QD Fabry-Perot (FP) lasers sharing the

same active regions but emitting exclusively on either the GS or the ES. In other words, the QD lasers under

study do not exhibit the GS-ES interplay dynamics where ES and GS lasing take place simultaneously. By

varying the feedback strength under the short delay configuration, multiple dynamical states such as periodic

(P), regular pulse package (RPP), quasi-chaos pulse package (QCPP), and chaotic (C) states are unveiled. The

Further author information: (Send correspondence to F.G.)

F.G.: E-mail: grillot@telecom-paristech.fr

GS laser is found to be rather insensitive to optical feedback due to the large damping factor. In contrast, the

ES laser can be driven more easily into complex dynamical regimes including chaotic states. We believe that

these results are of primary importance for utilizing the nonlinear dynamics for ultrafast devices in particular

for the development of isolator-free transmitters, integrated self-pulsating devices and other applications using

diode laser chaos.8

2. LASER STRUCTURE AND EXPERIMENTAL SETUP

The dynamical characteristics of the two FP multimode QD lasers, one emitting exclusively on the GS and one

on the ES, are investigated and compared under different operation and feedback conditions. For both lasers,

their active regions comprise 10 InAs dot sheets grown in InGaAs QWs by molecular beam epitaxy (MBE) with

a dot-in-well structure. The dot densities are around 3 ∼5×1010cm−2per layer and their lateral extensions

approach 30 nm. Both laser types have internal cavities of 1 mm long and 2 µm waveguides etched through the

active area. Lasers are left as-cleaved and cavity lengths are both 1 mm long while the ridge waveguide (RWG)

etched through the active region is 2 µm wide. For the GS laser, the threshold current Ith is of 16.5 mA and the

external efficiency is 21%. For the ES one, the threshold current Ith is of about 88 mA and the external efficiency

is 11%. Figures 1(a) and 1(b) display the two optical spectra taken at 1.5 times the threshold. The first FP laser

(in red) emits at ∼1300 nm on the sole GS transition whereas the second (in blue) is at ∼1230 nm on the sole ES

transition. The insets highlight the center of the emission of both lasers respectively, where the ES one exhibits

a modulated optical spectral envelope in contrast to the GS laser.9In this work, the ES selection was obtained

by exploiting the natural wavelength dispersion of the photoluminescence peak across the entire wafer.9Figure

-80

-60

-40

-20

1200 1220 1240 1260 1280 1300 1320 1340

-80

-60

-40

-20

(b)

Spectral Power (dBm)

(a)

Wavelength (nm)

1298 1300 1302 1304 1306

-60

-40

-20

1226 1228 1230 1232 1234

-60

-40

-20

Figure 1. Optical spectra of the (a) GS and (b) ES lasers measured at 1.5 times the threshold under free-running condition

(no feedback applied).

2 shows the schematic setup of the QD laser subject to optical feedback. The output of the laser towards the

left is fed back to the laser cavity through a partially reflecting mirror to form an external cavity with length

Lext with a minimum cavity length of 2 cm. The ratio between the frequency of the external cavity fext and the

relaxation oscillation frequency fRO is such as fext/fRO >1 meaning that the short cavity regime is investigated

in this paper. A variable optical attenuator is used to adjust the feedback strength ξfdefined as the ratio of the

feedback field to the laser output field. The optical signals are analyzed by an optical spectrum analyzer and

the optical power is measured by a power meter. The electrical signals are detected by two identical high-speed

photodetectors and analyzed by an electrical spectrum analyzer and a real-time oscilloscope.

ISO

PD ESA

OSC

OSA

96%

20%

80%

50%

VOA

Translation Lens

fiber

Mirror

Figure 2. Experimental setup of a QD laser subject to optical feedback. LD: QD laser; ISO: isolator; PD: high-speed

photodetector; ESA: electrical spectrum analyzer; OSA: optical spectrum analyzer; OSC: oscilloscope; PM: power meter.

3. RESULTS AND DISCUSSION

Figure 3 shows the time series and the corresponding power spectra of the dynamical states measured at 1.7

times the threshold for the GS laser. At Lext = 30 mm, a RPP state is demonstrated in Figures 3(a) and 3(b)

for ξf= 0.826. The RPP regime is a typical complex dynamics arising under optical feedback, which manifests

as regular pulsations in the output power.10 It occurs in the very short cavity regime meaning that a very small

Intensity (dBm)

Amplitude (mV)

(a)

5 10 15 20 25 30

-10

-5

0 5 10 15

(d)

(c)

Time (ns) Frequency (GHz)

0 5 10 15

−90

−80

−70

−60

−50

−40

−30 (b)

RPP

−90

−80

−70

−60

−50

2 4 6 8 10

−4

−2

ext

Figure 3. Time series and corresponding power spectra of the dynamical states measured at 1.7 times the threshold for

the GS laser with (ξf,Lext) = (a)(b) (0.826, 30 mm), (c)(d) (0.731, 30 mm). Red-dashed lines depict the external cavity

frequency fext and black curves are the spectra without feedback for reference.

number of modes is involved. Unlike the low-frequency fluctuation regime taking place when fext << fRO ,

the path in the phase space for the RPP regime is one direction only, which means that the crisis is no longer

present, and the trajectory always visits the same external cavity modes leading to the birth of fast and regular

oscillations at fext. Different from the RPP state described in10 for QW lasers showing a fast oscillation frequency

fFO that coincides with fext, the RPP state observed in this work has fFO around fRO = 3 GHz instead. This

difference might be attributed to the fact the QD laser does not operate in the ultra short cavity regime for which

fext >> fRO, but rather in a short delay configuration for which the relaxation oscillation frequency remains

still pretty close to the external cavity frequency. In addition, it has to be stressed that the slow oscillation has a

frequency fSO = 72 MHz that is much lower than both fext and fRO. Last but not the least, as compared to the

RPP states reported in10 with less than 10 oscillations in each package, more than 40 oscillations are packed in

the RPP states found in this GS laser. Figure 4 shows the evolution of fSO as a function of the external cavity

frequency for two values of the bias current namely 1.5 times and 1.7 times the threshold respectively. Results

show that there is a slight dependence of the slow oscillation with the bias current which is in agreement with

prior works on the RPP regime in QW lasers11.12 For instance, for an external cavity frequency, experiments

show that fSO increases by about 20 MHz with the bias current. To the best of our knowledge, this RPP regime

has never been observed so far in QD lasers.

f ext (GHz)

0 1 2 3 4 5 6

SO (MHz)

100

1.75

1.5

Figure 4. Evolution of the SO frequency fSO measured in the RPP regime for the GS laser at 1.5 times and 1.7 times the

threshold respectively.

When ξfdecreases to 0.731, the laser becomes more stable and a P state oscillating at fRO is found as shown

in Figures 4(c) and 4(d) is found. Further increase of the external cavity length to the long cavity regime leads

to frequency-locking regimes as reported elsewhere.13 Experimental results reveal that the GS laser is perfectly

stable such that no chaotic state was found whatever the feedback strength level used in this setup. Besides, the

GS laser remains under the steady-state condition providing the feedback strength does not exceed 30% which

is already much larger compared to any typical reflection levels taking place in a transmission system13.14

Figure 5 shows the time series and corresponding power spectra of the dynamical states from the ES laser

still measured at 1.7 times the threshold. With Lext = 20 mm and with a strong feedback of ξf= 0.703, a C state

with irregular intensity modulation and a spectrum broadly elevated from the noise floor is shown in Figures

5(a) and 5(b). When ξfdecreases to 0.686, a QCPP state is found and shown in Figures 5(c) and 5(d). Different

from a RPP state that has a fast oscillation at fext,10 the fast oscillation of this QCPP oscillates at fRO. The fSO

does not coincide with either fRO or fext but with a frequency of 340 MHz instead. Moreover, unlike the spectra

of typical RPP states that have frequency components of only fast or slow oscillations and their corresponding

harmonics and beats, the spectrum of the QCPP state preserves the broadband characteristics as present in the

C state. When ξffurther decreases to 0.48, as shown in Figures 5(e) and 5(f), a P state stably oscillating at

fFO = 4 GHz is obtained. Finally, let us stress that in the long cavity regime, more complex dynamical states are

observable in the ES laser hence leading to a complete route to chaos13 which which is fundamentally different

than the dynamics displayed in the GS laser.

Amplitude (mV)

Frequency (GHz)Time (ns)

Intensity (dBm)

(c)

-10

-5

06102 4 8 0 5 10 15

−90

−80

−70

−60

−50

−40

−30 (d)

QCPP

(e)

-10

-5

06102 4 8

-10

-5

06102 4 8

(a)

−90

−80

−70

−60

−50

−40

−30

0 5 10 15

(f)

0 5 10 15

−90

−80

−70

−60

−50

−40

−30 (b)

ext

Figure 5. Time series and corresponding power spectra of the dynamical states measured at 1.7 times the threshold for

the ES laser with (ξf,Lext) = (a)(b) (0.703, 20 mm), (c)(d) (0.686, 20 mm), (e)(f)(0.48, 20 mm). Red-dashed lines depict

the external cavity frequency fext and black curves are the spectra without feedback for reference.

Figures 6 show fSO extracted from the RPP states of the GS laser and the QCPP states of the ES laser under

different fext. As can be seen, while fSO does not coincide exactly with either fRO or fext,fSO in both states

increases linearly as fext increases.

Compared to the GS laser which is in general stable and insensitive to feedback, the ES laser is more easily

moving to complex dynamics. While a GS laser is of large importance for the development of isolator-free

transmitters in short-reach networks, an ES laser on the other hand can be essential for applications taking

advantages of chaos such as chaos lidars, chaos radars, and high-speed random number generations.15–18 From

these results it is obvious that, although the GS and ES lasers have the same active medium, their response to

the feedback are very different. Unlike the ES laser, the carrier dynamics of the GS laser involves transport,

capture, and relaxation, leading to a larger damping rate that stabilizes the laser and prevents the development

Figure 6. Slow oscillation frequencies fSO of the RPP and QCPP states under different external cavity frequencies fext of

both GS and ES lasers respectively. Red lines are the linear fits.

of complex dynamics. As reported in,9the damping is as large as 18 GHz for the GS laser whereas it is below 1

GHz for the ES one. Moreover, the ES laser has a stronger modal competition9which also makes it easier to be

driven into instabilities.

4. CONCLUSIONS

This work investigates for the first time the dynamical states and their spectral characteristics of optical feedback

InAs/GaAs QD lasers emitting exclusively on single lasing states in the short cavity regime. Although the GS

and ES QD lasers are made from the same active medium, their feedback dynamics are found not much alike.

The GS laser is shown to be almost insensitive to feedback, especially at higher bias levels. No chaotic states

was found which is of vital importance for the development of high-speed transmission links operating without

isolator, in agreement with some earliest results. In contrast, the ES laser exhibits a plethora of complex dynamics

including chaotic states, especially at higher bias levels, thus being useful for integrated self-pulsating devices,

chaotic lidars and radars as well as high-speed random bit generations.

5. ACKNOWLEDGMENTS

This work was funded by Campus France (PhC Orchid No. 33721SC), the Institut Mines T´el´ecom (IMT) through

the Futurs & Ruptures program and the Ministry of Science and Technology, Taiwan, under contracts MOST

105-2911-I-007-501 and 103-2112-M-007-019-MY3.

REFERENCES

[1] Kane, D. M. and Shore, K. A., [Unlocking Dynamical Diversity: Optical Feedback Effects on Semiconductor

Lasers], Wiley (2005).

[2] Crowley, M. T., Naderi, N. A., Su, H., Grillot, F., and Lester, L. F., [GaAs-Based Quantum Dot Lasers, in

Advances in Semiconductor Lasers], Academic Press (2012).

[3] Eisenstein, G. and Bimberg, D., [Green Photonics and Electronics], Springer (2017).

[4] Shchekin, O. and Deppe, D., “1.3 µm InAs quantum dot laser with T0=161◦K from 0 to 80◦C,” Appl. Phys.

Lett. 80, 3277 (2002).

[5] Luedge, K., [Nonlinear Laser Dynamics: From Quantum Dots to Cryptography], Wiley (2012).

[6] Arsenijevi´c, D. and Bimberg, D., “Quantum-dot lasers for 35 Gbit/s pulse-amplitude modulation and 160

Gbit/s differential quadrature phase-shift keying,” Proc. of SPIE 9892, 98920 (2016).

[7] Virte, M., Breuer, S., Sciamanna, M., and Panajotov, K., “Switching between ground and excited states by

optical feedback in a quantum dot laser diode,” Appl. Phys. Lett. 105, 121109 (2014).

[8] Sciamanna, M. and Shore, K. A., “Physics and applications of laser diode chaos,” Nat. Photonics 9, 151

(2015).

[9] Huang, H., Arsenijevi´c, D., Schires, K., Sadeev, T., Bimberg, D., and Grillot, F., “Multimode optical feed-

back dynamics of InAs/GaAs quantum-dot lasers emitting on different lasing states,” AIP Advances 6(12),

125114 (2016).

[10] Heil, T., Fischer, I., Els¨aßer, W., and Gavrielides, A., “Dynamics of semiconductor lasers subject to delayed

optical feedback: The short cavity regime,” Phys. Rev. Lett. 87(24), 243901 (2001).

[11] Sciamanna, M., Tabaka, A., Thienpont, H., and Panajotov, K., “Intensity behavior underlying pulse pack-

ages in semiconductor lasers that are subject to optical feedback,” J. Opt. Soc. Am. B 22(4), 777 (2005).

[12] Tabaka, A., Panajotov, K., Veretennicoff, I., and Sciamanna, M., “Bifurcation study of regular pulse pack-

ages in laser diodes subject to optical feedback,” Phys. Rev. E 70(3), 036211 (2004).

[13] Lin, L. C., Chen, C. Y., Huang, H., Arsenijevic, D., Bimberg, D., Grillot, F., and Lin, F. Y., “Comparison

of optical feedback dynamics of InAs/GaAs quantum-dot lasers emitting solely on ground or excited states,”

Optics Lett. 43, 210 (2018).

[14] Huang, H., Lin, L. C., Chen, C. Y., Asenijevic, D., Bimberg, D., Lin, F. Y., and Grillot, F., “Multimode

optical feedback dynamics in InAs/GaAs quantum dot lasers emitting exclusively on ground or excited

states: transition from short- to long-delay regimes,” Optics Express 26, 1743 (2018).

[15] Lin, F. Y. and Liu, J. M., “Chaotic lidar,” IEEE J. Sel. Topics Quantum Electron. 10, 991 (2004).

[16] Cheng, C. H., Chen, C. Y., Chen, J. D., Pan, D. K., Ting, K. T., and Lin, F. Y., “3D pulsed chaos lidar

system,” Optics Express 26, 12230 (2018).

[17] Lin, F. Y. and Liu, J. M., “Chaotic radar using nonlinear laser dynamics,” J. Quantum Electron. 40, 815

(2004).

[18] Takahashi, R., Akizawa, Y., Uchida, A., Harayama, T., Tsuzuki, K., Sunada, S., Arai, K., Yoshimura, K.,

and Davis, P., “Fast physical random bit generation with chaotic semiconductor lasers,” Opt. Express 22(10),

11727 (2014).