In optical communication, a controllable variable optical buffer is one of the most critical components. In such buffer, optical data would be kept in optical format throughout the storage time without being converted into electronic format. The turn-on and turn-off time should be variable with an external control. The most important application of optical buffers is perhaps all-optical routers in packet-switched networks. A router is used in networks to interconnect end-user systems to each other where packets are the basic units of information that are transported. A router often connects many networks and performs decisions on how to send packets from its source to its destination in the network. Packet switching is a method of communication whereby information is broken up into blocks of limited length called packets. They are then switched in a network by routers. The key building blocks of an electronic router include a switch fabric, processors, and buffers. The key missing component for an all-optical router is an all-optical buffer. With optical buffers, one packet can be stored in the buffer temporarily, allowing the other packet to go first. Until the traffic is cleared at the output port, the packet stored in the buffer is released. An all-optical router can potentially alleviate the traffic congestion in future very-high-bandwidth networks. However, variable optical delay lines (also called “slow light” devices) and buffers will be crucial components in
optical communication, phased array antenna, and signal processing systems. There are several technical approaches to implement an optical buffer. No matter what methods, the medium which the optical signal travels in should be varied by either increasing the path length or reducing the signal group velocity. The former can be achieved by using of a fiber delay line or resonant cavity and the latter has several possibilities, such as EIT, CPO. Recently, tunable slow light using an InGaAsP quantum well Fabry-Perot laser has been reported [15]. Tunable optical group delays 14 ps for 2 GHz sinusoidal signal have been demonstrated, and the delay-bandwidth product was 0.028. Moreover, tunable slow light using a quantum well VCSEL fabricated on InP-based materials also has been proposed [16]. The modulation frequency of probe signal between 1-3 GHz was presented, and the delay-bandwidth product was 0.28 (100 ps × 2.8 GHz).
1.4 Organization of the thesis
This thesis consists of three related parts. In Chapter 2, we demonstrate monolithically single-mode QD VCSELs with high side-mode suppression ratio (> 30 dB) and report the temperature performance and dynamic properties including bandwidth and eye diagram.
In Chapter 3, we report the experimental characterization of 1.3µm QD VCSEL with and without external light injection. Significant frequency response enhancement has been observed. Furthermore, we demonstrate that this frequency response enhancement allows us to improve the performance of subcarrier multiplexed (SCM) system. We also report the third-order intermodulation distortion (IMD3) of QD VCSEL with and without external light injection.
In Chapter 4, we report the slow light in the monolithically single-mode QD VCSEL.
Tunable optical group delay can be achieved by adjusting the bias current. A 10 GHz modulation signal with tunable optical group delays 42 ps has been demonstrated. We
Reference
[1] N. Tansu and L. J. Mawst, "Temperature sensitivity of 1300-nm InGaAsN quantum-well lasers", IEEE Photonics Technology Letters, vol. 14, no. 8, pp.
1052-1054, 2002.
[2] C. W. Tu and Pkl Yu, “Material properties of III-V semiconductors for lasers and detectors,” MRS Bulletin, vol. 28, pp. 345-349, 2003.
[3] M. Kawaguchi, T. Miyamoto, E. Gouardes, D. Schlenker, T. Kondo, F. Koyama and K. Iga “Lasing Characteristics of Low-Threshold GaInNAs Lasers Grown by Metalorganic Chemical Vapor Deposition,”Japanese Journal of Applied Physics, vol.
40, pp. L744–L746, 2001.
[4] S. L. Chuang, N. Holonyak, Jr. “Quantum-well assisted tunneling injection quantum-dot lasers,” Conference on Lasers and Electro-Optics, Technical Digest, vol.
1, pp. 297, 2002.
[5] T. Yang, J. Tatebayashi, S. Tsukamoto and Y. Arakawa,” Highly uniform self-assembled InAs/GaAs quantum dots emitting at 1.3µm by metalorganic chemical vapor deposition,” Physica E, vol. 26, pp. 77-80, 2005.
[6] N. N. Ledentsov, "Long-wavelength quantum-dot lasers on GaAs substrates: from media to device concepts," IEEE Journal of Selected Topics in Quantum Electronics, vol. 8, pp. 1015 - 1024, 2002.
[7] N. N. Ledentsov, M. Grundmann, F. Heinrichsdor, D. Bimberg, V.M. Ustinov, A. E.
Zhukov, M. V. Maximov, Zh. I. Alferov, J. A. Lott, “Quantum-Dot Heterostructure Lasers,” IEEE Journal of Selected Topics in Quantum Electronics, vol. 6, no. 3, pp.
439-451, 2000.
[8] J. A. Lott, N. N. Ledentsov, V. M. Ustinov, A. Yu. Egorov, A. E. Zhukov, P. S. Kop’ev, Zh. I. Alferov, and D. Bimberg, “Vertical cavity lasers based on vertically coupled quantum dots,” Electronics Letters, vol. 33, pp. 1150-1151, 1997.
[9] Y. Yamamoto and T. Kimura, “Coherent optical fiber transmission systems,”IEEE J.
Quantum Electron., vol. QE-17, pp. 919–935, June 1981.
[10] P. Gallion, H. Nakajima, G. Debarge, and C. Chabran, “Contribution ofspontaneous emission to the linewidth of an injection-locked semiconductorlaser,” Electron. Lett., vol. 22, pp. 626–628, 1985.
[11] K. Iwashita and K. Nakagawa, “Suppression of mode partition noise by laser diode light injection,” IEEE J. Quantum Electron., vol. QE-18, pp.1669–1674, Oct. 1982.
[12] N. A. Olsson, H. Temkin, R. A. Logan, L. F. Johnson, G. J. Dolan, J.P. Van der Ziel, and J. C. Campbell, “Chirp-free transmission over 82.5km of single mode fibers at 2 Gbit/s with injection locked DFB semiconductor lasers,” IEEE J. Lightwave Technol., vol. LT-3, pp. 63–67, Feb.1985.
[13] X. Meng, T. Chau, and M. C.Wu, “Experimental demonstration of modulation bandwidth enhancement in distributed feedback lasers with external light injection,”
Electron. Lett., vol. 34, no. 21, pp. 2031–2032, 1998.
[14] R. Olshansky,””, J. Lightwave Technology 7(1989) 1329
[15] S. Minin, M. R. Fisher, S. L. Chuang, "Current-controlled group delay using a semiconductor Fabry-Perot amplifier," Applied Physics Letters, vol. 84, pp.
3238-3240, 2004.
[16] X. Zhao, P. Palinginis, B. Pesala, C. J. Chang-Hasnain, P. Hemmer, "Tunable ultraslow light in vertical-cavity surface-emitting laser amplifier," Optics Express, vol.
13, pp. 7899-7904, 2005.
Chapter 2 High speed modulation of Quantum Dot VCSEL
In this chapter, we present monolithic quantum-dot vertical-cavity surface-emitting laser (QD VCSELs) operating in the 1.3 µm optical communication wavelength. The QD VCSELs have adapted fully doped structure on GaAs substrate. The output power is ~ 330 µW with slope efficiency of 0.18 W/A at room temperature. Single mode operation was obtained with side-mode suppression ratio of > 30 dB. Modulation bandwidth and eye diagram in 2.5 Gb/s was also presented.
2.1 Sample structure and Fabrication process
All structures were grown on GaAs (100) substrates by molecular beam epitaxy (MBE). The epitaxial structure was as follows (from bottom to top) - n+-GaAs buffer, 33.5-pair n+-Al0.9Ga0.1As/n+-GaAs (Si-doped) distributed Bragg reflector (DBR), undoped active region, p-Al0.98Ga0.02As oxidation layer, 22-pair p+-Al0.9Ga0.1As/p+-GaAs DBR (carbon-doped) and p+-GaAs (carbon-doped) contact layer. The graded-index separate confinement heterostructure (GRINSCH) active region consisted mainly of three stack of QDs active region, with PL emission at 1.266 µm, embedded between two linear-graded AlxGa1-xAs (x = 0 to 0.9 and x = 0.9 to 0) confinement layers. The thickness of the cavity active region was 1λ. Carbon was used as the p-type dopant in the DBR to increase the carrier concentration (2-3×1018 cm-3). The interfaces of both the p-type and n-type Al0.9Ga0.1As/GaAs DBR layers are linearly graded to reduce the series resistance. The optical characteristics of QDs were optimized through PL measurement and structural analysis. The details of the process were fully described in our previous works [1]. The mesa diameter of the fabricated device is 26µm with a 5 µm oxide aperture, and the device surface is quasi-planar so that the annular p-contact metal and the bond pad are on the same level. The device structure is shown in Fig.2-1. The p-contact was formed by directly depositing Ti/Pt/Au on the upper heavily doped p+ GaAs contact layer, and Au/Ge/Ni/Au was deposited on the bottom side of the substrate after it had been thinned down, and the shaded region beneath the bond pad represents the implanted region. It is worth noting that neither the intra-cavity /
co-planar metal contact nor polyimide resin planarization been used in the process.
After metal annealing, the sample was immediately probe tested on the wafer level to extract the static operation characteristics, and was subsequently divided into two pieces.
One piece was diced for packaging later, and the other was sent to be subjected to H+ implantation with a dose of 1015cm-2 for the purpose of further reducing the parasitic capacitance. Preliminarily we adopt four different proton energies in the range from 300 to 420keV according to the stopping and range of ions in matter (SRIM) simulation results, but the ideal combination of proton implantation energies in relative experiments currently we finished has not been optimized yet. The proton energy we used in this experiment was stronger than that used for conventional proton-implanted VCSELs. The reason for using such high-energy protons to bombard the sample is simply that the surface of the implanted region has already been passivated by the presence of 1500-Å-thick SiO2 and 4000-Å-thick metal bond pad. When a charged particle such as H+ penetrates a dielectric material such as SiO2, it is subjected to scattering because of the built-in electric field result from the dipole and metal-insulator-semiconductor interfaces.
However the incident proton has seldom been influenced by the thicker metal bond pad because there is no electric field inside the metal. Hence, we need a higher energy to make sure the implanted proton can penetrate the SiO2 and eventually locate at the right depth, that is, as close as possible to the active region. The implantation region was kept apart the mesa to prevent the damage which was caused by ion bombardment from destroying the active region.
2.2 DC Characteristics of QD VCSEL
2.2.1 Experimental Setup
In order to precisely measure characteristics of QD VCSEL, such as LIV curve, optical spectrum, or near field pattern, we need to setup a system which can test our sample on wafer level or packaged level. Probe station was a basic instrument to meet our needs.
Scheme of probe station system, illustrated in Figure 2-2, include probe station, current source, and power-meter module. Keithley 238 can provide precisely continuous current with laser diode and measure relative voltage synchronously. Newport power meter module (model 1835C) with photodiode and power meter can measure the light output power of the laser diode. An integration sphere was used to pick up whole emitting power from VCSEL to improve the accuracy of power measurement.
The VCSEL device was placed on a platform of the probe station and was injected bias current with microprobe. Threshold condition, slope efficiency, turn-on voltage and differential resistance can be obtained from L-I-V information by sweeping bias current.
Near-field pattern was obtained by specific CCD. Emission spectrum of the device was measured by optical spectrum analyzer (OSA, Advantest 8381). A multi-mode fiber probe was placed close to the emission aperture to take optical spectra. The OSA had spectrum resolution of 0.1nm which was adequate to measure VCSEL lasing spectra.
2.2.2 Results and Discussion
Fig. 2-3 plots curves of light output and voltage versus current (LIV). The threshold current is ~ 1.8 mA and the threshold current density is 7.6 kA/cm2. The output power rollover occurs as the current increases above 4mA with maximum optical output of 0.33mW at 20°C. In addition, the VCSEL is single polarization in the full operating range with ratio of ~20dB. Inset of Fig 2-3 is the near-filed pattern of the QD VCSEL biased at 3mA which indicated the fundamental mode lasing. Fig 2-4 shows the typical emission spectra of the quantum-dot VCSELs, which indicate single transverse mode operation in the whole operation range with a lasing wavelength of ~1.278 µm and side mode suppression ratio (SMSR) > 30dB.
To investigate the temperature dependence of the QD VCSEL, LI curves were measured from room temperature to 55oC with current step of 0.01 mA, as shown in Fig.
2-5. The threshold current varies only 0.15 mA (< 10% of Ith ) with temperatures from 10°C to 45°C and the slope efficiency drops from 0.18 to 0.1 W/A. The small temperature dependence of threshold current corresponds to a characteristic temperature (T0) of 450K, a high value comparing with the InGaAs(N) VCSEL in 1.3 µm. The high characteristic temperature was attributed to the wide gain spectra of the quantum dots gain media. When temperature increases, the wide gain spectra make the alignment between gain spectra and cavity resonance not sensitive, and therefore improve the T0. However, increase of threshold current with temperature and the quench of output power were also observed after 50oC which implies the gain of the quantum dots decreases severely in higher temperature.
2.3 High speed modulation of QD VCSEL
2.3.1 Theory：Small signal modulation
Under small signal modulation, the carrier and photon density rate equation are used to calculate relaxation resonance frequency and its relationship to laser modulation bandwidth.
Consider the application of an above-threshold DC current, Io, carried with a small AC current, Im, to a diode laser. The small modulation signal with some possible harmonics of the drive frequency, ω. Small signal approximation, assumes Im << Io bias and spontaneous emission term, β, is neglected, is expressed as
jwt
Before applying these equations, the rate equation is rewritten for the gain. Assumption under DC current is sufficiently above threshold that the spontaneous emission can be neglected. Without loss of generality, we suppose full overlap between the active region and photon field, Γ=1; furthermore, internal quantum efficiency, ηi , is neglected. That is,
substitute Eq (2-1) into Eq(2-2) and Eq(2-3), it is similarly expressed modulation terms as
p
The small signal terms in frequency domain of carrier and photon are given by
jwt
substitute into Eq(2-4) and (2-5), the equations become
(2-1)
(2-2)
(2-3)
(2-4)
(2-5)
τ
Carrier modulation term in frequency domain is simplified as
)
Photon modulation term in frequency domain is simplified as
) response of two arranged equations as below
))
With the Eq(2-8) and (2-9), we observe the coupling between the small signal photon, npm, and carrier, n . Small signal carrier injection induces photon achieved oscillation. m This phenomenon produces a natural resonance in the laser cavity which shows up the output power of the laser in response to sudden changes in the input current. The natural frequency of oscillation associated with this mutual dependence between n and m npm. Modulation response is expanded the small signal modulation relationship to steady-state.
From Eq.(2-8) and (2-9), the modulation response is denoted as
(2-6)
2 2
Modulation bandwidth is determined as cutoff frequency, f , which is the position with c half response written as
2
Transfer function, H(w), is the identical term in Eq(2-12). and respectively obtained with Cramer’s rule. It is similar to modulation response, M(w), describing the response of the laser intensity to small variations in the drive current through the active region. That is,
πγ to Eq.(2-11) , and C is a constant. Accounting for additional extrinsic limitations due to carrier transport and parasitic elements related to the laser structure results in an extra pole in the small signal modulation transfer function
)
where f is the cutoff frequency of the low pass filter characterizing the extrinsic p limitations. It is crucial for microwave applications that the modulation bandwidth of the VCSEL is sufficiently large so that efficient modulation is achieved as the modulation frequency.
(2-12)
(2-13)
(2-14)
(2-15)
2.3.2 Experimental Setup：Microwave test system
The microwave test system was mainly consisted of network analyzer, Bias-Tee and high speed photodetector, as illustrated with Figure 2-6. Agilent 8720ES network analyzer was a crucial instrument of this microwave measurement. Transmitter of network analyzer produced -10dBm RF signal. Laser diode drivers (New port, model 525) provided direct bias current with the laser diode. Bias-Tee combined AC and DC signal transmission through the coaxial cable. The laser diode was hermetically sealed by a standard TO-Can laser package (TO-46) with a built-in lens. The laser diode TO-Can package and the single-mode fiber are assembled by laser welding technique, as shown in Figure 2-7. Then, we welded our device on a high speed SMA connector and connected with the coaxial cable. 25-GHz near-IR photodetector (New Focus, model 1414) was received the modulation light signal from the laser diode and was conversed into electrical signal and fed to network analyzer. Comparing two channels microwave signal by network analyzer, information of transmission and reflection characteristics could be expressed as vector(magnitude and phase), scalar(magnitude only), phase-only quantities, that was, S-parameter.
2.3.3 Results and Discussion
The small signal response of VCSELs as a function of bias current was measured at 25oC using a calibrated vector network analyzer (Agilent 8720ES). Fig. 2-8 indicates the modulation frequency increases with the bias current low current range. At a bias current of larger than 2.5mA, the bandwidth saturate and the maximum 3dB modulation frequency response is measured as ~2 GHz. In Fig. 2-9, the 3dB bandwidth (f3dB) is plotted as a function of the bias current. At low bias currents, the bandwidth increase in proportion to the square root of the current as expected from the rate equation analysis. The saturation of bandwidth was clearly observed as bias current increases above 2.5mA which might be attributed to heating effect. Carrier de-population in QDs subjected to the heated active region may suppress the material gain and put the intrinsic limit of high speed modulation. The modulation current efficiency factor (MCEF) is ~ 2.5 GHz/(mA)1/2. Improvement can be done by increasing the quantum dots stacks and reducing the current density of each dot simultaneously. Finally, we illustrate the eye diagram at 1.25 Gb/s and 2.5 Gb/s, as shown in Fig 2-10. The QD VCSEL shows clear and symmetrical eye diagram at 1.25 Gb/s. At 2.5 Gb/s, the eye was degraded due to the overshoot and insufficient bandwidth. Future work will focus on enhancement of high speed performance by reducing the device parasitic and thermal impedance.
2.4 Conclusion
We present monolithic quantum-dot vertical-cavity surface-emitting laser (QD VCSELs) operating in the 1.3 µm optical communication wavelength. The QD VCSELs have adapted fully doped structure on GaAs substrate. The output power is ~ 330 µW with slope efficiency of 0.18 W/A at room temperature. Single mode operation was obtained with side-mode suppression ratio of > 30 dB. The high speed characteristics were also investigated. The free running bandwidth of QD VCSEL is
~2GHz. The modulation current efficiency factor (MCEF) is ~ 2.5 GHz/(mA)1/2. Finally, we illustrate the eye diagram at 1.25 Gb/s and 2.5 Gb/s.
Reference
[1]H. C. Yu, S. J. Chang, Y. K. Su, C. P. Sung, Y. W. Lin, H. P. Yang, C. Y. Huang and J. M. Wang, “A simple method for fabrication of high speed vertical cavity surface emitting lasers,” Materials Science Engineering B, vol. 106, pp. 101-104, 2004.
Fig. 2-1 QD VCSELs device structure. Inset is the top view image of the QD VCSEL.
CCD
Objective Keithley 238 driver
Power supply
Fig. 2-2 Probe station measurement instrument setup
1272 1274 1276 1278 1280 1282
1272 1274 1276 1278 1280 1282 -80
Fig. 2-4 Emission spectra of QD VCSEL at room temperature.
Current driver Network analyzer
Laser diode Photodetector
0 5 10 15 20
-65 -60 -55 -50 -45 -40 -35 -30 -25 -20 -15 -10 -5
Response (dB)
Frequency (GHz)
S11 : Reflection coefficient S21: Frequency response
HP 8720ES VNA
New Port 525
New Focus 1414 Bias Tee
QD VCSEL
Fiber
Fig. 2-6 Microwave test system setup
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 -30
-20 -10 0
Ma g n it ude (d B)
Frequency(GHz)
2mA 2.2mA 2.4mA 2.6mA 2.8mA 3mA 3.2mA
0.0 0.2 0.4 0.6 0.8 1.0 1.2
0.0 0.5 1.0 1.5 2.0
MCEF=2.5 GHz/(mA)1/2
3- dB F re q uency ( GH z )
(I - I
th
)
1/2(mA)
1/2Fig. 2-8 Small signal modulation response of QD VCSEL
Fig. 2-9 3dB frequency as a function of square root of current
Fig.2-10 Eye diagram of the QD VCSELs at (a) 1.25 Gb/s and (b) 2.5 Gb/s ( The time scale was 200 ps/div and 100 ps/div)
(b)
(a)
Chapter 3 Injection Locking of Quantum Dot VCSEL
This investigation experimentally demonstrates the dynamic characteristics of quantum dot vertical-cavity surface-emitting lasers (QD VCSEL) without and with light injection. The QD VCSEL is fully doped structure on GaAs substrate and operates in the 1.3 µm optical communication wavelength. The eye diagram, frequency response, and intermodulation distortion are presented. We also demonstrate that the frequency response enhancement by light injection technique allows us to improve the performance of subcarrier multiplexed system.
3.1 Modulation Response Enhancement by injection locking technique
3.1.1 Theory：Injection Locking Theorem
A rate equations based model is usually used to describe the interaction between photons and carriers inside a laser cavity. When an additional light source is injected into the cavity, the system preserves the general form of the original equations, but with extra terms describing the effects of the injection [1]. These extra terms play an important role in this nonlinear dynamic system and the equations are shown below.
n
In these equations, S,φ, and N denote the photon number, the phase, and the carrier number inside the follower laser cavity, respectively. G0 denotes the gain coefficient,
In these equations, S,φ, and N denote the photon number, the phase, and the carrier number inside the follower laser cavity, respectively. G0 denotes the gain coefficient,