Quantum Dot Lasers. Huizi Diwu and Betul Arda

Transcription

1 Huizi Diwu and Betul Arda Abstract In this paper, we reviewed the recent literature on quantum dot lasers. First of all, we start with physics of quantum dots. These nanostructures provide limitless opportunities to create new technologies. To understand the applications of quantum dots, we talked about quantum confinement effect versus dimensionality and different fabrication techniques of quantum dots. Secondly, we examined the physical properties of quantum dot lasers along with history and development of quantum dot laser technology and different kinds of quantum dot lasers comparing with other types of lasers. Thirdly, since engineering is a practical science, we made a market search on the practical usage of quantum dot lasers. And lastly, we predicted a future for quantum dot lasers. I. INTRODUCTION 1.1. Quantum Dots Quantum dots (QD) are semiconductor nanostructures with vast applications across many industries. Their small size (~2-10 nanometers or ~10-50 atoms in diameter) gives quantum dots unique tunability. Like that of traditional semiconductors, the importance of QDs is originated from the fact that their electrical conductivity can be altered by an external stimulus such as voltage or photon flux. One of the main differences between quantum dots and traditional semiconductors is that the peak emission frequencies of quantum dots are very sensitive to both the dot's size and composition. [1], [2] Figure 1 size comparison of a QD [2] QDs utilize the motion of conduction band electrons, valence band holes or excitons. Excitons are pairs of conduction band electrons and valence band holes and defined to describe the motion of 1

2 electrons and holes less complicatedly. The confinement of the motion of holes and electrons can be created by electrostatic potentials o e.g. doping, strain, impurities, external electrodes the presence of an interface between different semiconductor materials o e.g. in the case of self-assembled QDs the presence of the semiconductor surface o e.g. in the case of a semiconductor nanocrystal or by a combination of these [2] Quantum Confinement Effect To understand the QD concept, first of all, we should consider the quantum confinement effects on electrons. Quantum confinement occurs when one or more of the dimensions of a nanocrystal approach the Exciton Bohr radius. The concepts of energy levels, bandgap, conduction band and valence band still apply. However, the electron energy levels can no longer be treated as continuous - they must be treated as discrete. [3], [4] Figure 2 Comparisons of quantum wells, wires, rods and dots a. Geometries of the different structures. b. Plots of E g (the increase in the bandgap over the bulk value) against d (the thickness or diameter) for rectangular quantum wells, cylindrical quantum wires and spherical QDs obtained from particle-in-a-box approximations. The grey area between the dot and wire curves is the intermediate zone corresponding to quantum rods. The vertical dotted line and points qualitatively represent the expected variation in the bandgap for InAs quantum rods of varying length/diameter ratio, as studied by Kan et al. 1. c. A plot of E g against length/diameter ratio for the InAs quantum rods synthesized by Kan et al., showing the dependence of the bandgap on the shape of the quantum rods. The dotted line represents the variation expected from a particle-in-a-box approximation [5]. 2

3 Quantum well, or quantum wire confinements give the electron at least one degree of freedom. Although this kind of confinement leads to quantization of the electron spectrum which changes the density of states, and results in one or two-dimensional energy subbands, it still gives the electron at least one direction to propagate. On the other hand, today s technology allows us to create QD structures, in which all existing degrees of freedom of electron propagation are quantized. We can think this confinement as a box of volume d 1 d 2 d 3. The energy is therefore quantized to E = E q1 + E q2 + E q3 where E qn = h 2 (q 1 π/d n ) 2 / 2m c q 1, q 2 and q 3 are the quantum numbers associated with an energy subband [6]. Since the allowed energy levels are discrete and separated, we can represent the density of states as delta functions. (Figure 3) The energy levels of a QD can be adjusted with a proper design according to the needs of the application. For instance, the addition or subtraction of just a few atoms to the QD has the effect of altering the boundaries of the bandgap. Changing the geometry of the surface of the QD also changes the bandgap energy, owing again to the small size of the dot, and the effects of quantum confinement. Figure 3 Comparison of the quantization of density of states: (a) bulk, (b) quantum well, (c) quantum wire, (d) quantum dot. The conduction and valence bands split into overlapping subbands that get successively narrower as the electron motion is restricted in more dimensions. [6] 3

4 1.3. Fabrication Techniques of Quantum Dots Core-Shell Quantum Structures: QDs are small regions of one material buried in another with a larger band gap. Self Assembled QDs and Stranski-Krastanov growth: Self-assembled QDs nucleate spontaneously under certain conditions during molecular beam epitaxy (MBE) and metalorganics vapor phase epitaxy (MOVPE), when a material is grown on a substrate to which it is not lattice matched. The resulting strain produces coherently strained islands on top of a two-dimensional "wetting-layer". This growth mode is known as Stranski- Krastanov growth. The islands can be subsequently buried to form the QD. The main limitations of this method are the cost of fabrication and the lack of control over positioning of individual dots. o MBE: A technique that grows atomic-sized layers on a chip rather than creating layers by diffusion. o MOVPE: is a chemical vapor deposition method of epitaxial growth of materials, especially compound semiconductors from the surface reaction of organic compounds or metalorganics and metal hydrides containing the required chemical elements. In contrast to molecular beam epitaxy (MBE) the growth of crystals is by chemical reaction and not physical deposition. This takes place not in a vacuum, but from the gas phase at moderate pressures (2 to 100 kpa). Monolayer fluctuations: QDs can occur spontaneously in QW structures due to monolayer fluctuations in the well's thickness. Individual QDs can be created from two-dimensional electron or hole gases present in remotely doped quantum wells or semiconductor heterostructures. The sample surface is coated with a thin layer of resist. A lateral pattern is then defined in the resist by electron beam lithography. This pattern can then be transferred to the electron or hole gas by etching, or by depositing metal electrodes (lift-off process) that allow the application of external voltages between the electron gas and the electrodes. Such QDs are mainly of interest for experiments and applications involving electron or hole transport, i.e., an electrical current.[1], [4], [7] 4

5 Figure 4 Schematic representation of different approaches to fabrication of nanostructures: (a) microcrystallites in glass, (b) artificial patterning of thin film structures, (c) self-organized growth of nanostructures. [7] 2.1. Development of II. QUANTUM DOT LASERS The laser operation is based on producing radiative emission by coupling electrons and holes at nonequilibrium conditions to an optical field. The advantages of quantum well lasers on traditional lasers first predicted in 1970s (Dingle and Henry 1976), and first quantum well lasers which were very inefficient were demonstrated at those dates (van der Ziel et al. 1975). The advantages recognized were: The confinement and nature of the electronic density of states result in more efficient devices operating at lower threshold currents than lasers with bulk active layers. The laser threshold current density can be reduced by decreasing the thickness of the active layer. Discrete energy levels provide a means of "tuning" the resulting wavelength of the material. Since the thickness of the quantum well-depends on the desired spacing between energy levels, tuning can be done by changing the quantum well dimensions or 5

6 thickness. For energy levels of greater than a few tens of mev s, the critical dimension is approximately a few hundred angstroms. The in efficiency of quantum well lasers were eliminated in 1980s by the use of new materials growth capabilities (molecular beam epitaxy), and optimization of the heterostructure laser design (Tsang 1982). [8], [9] Figure 5 The historical evolution of QD lasers from the beginning. [8] Since the quantum confinement in a QD is in all three dimensions, tunability of a quantum dot laser (QDL) is higher than a quantum well laser (QWL). The concept of semiconductor QDs was proposed for semiconductor laser applications by Arakawa and Sakaki in 1982, predicting suppression of temperature dependence of the threshold current. Henceforth, reduction in threshold current density, reduction in total threshold current, enhanced differential gain and high spectral purity/no-chirping were theoretically discussed in 1980 s (Asada et al. 1986). [8], [9]. At this point, we should examine the basics of laser operation. (Figures 5 and 6) A laser utilizes stimulated that is triggered by an incident photon of the same energy. This occurs when a medium has more population of electrons in the excited quantum level than in the ground level. This artificially situation, called population inversion, is produced by either electrical stimulation (electroluminescence) or optical stimulation and is different from the spontaneous emission, whereby the electron returns to the ground state in the natural course (within the lifetime of the excited states), even in the absence of any photon to stimulate it. These two processes are represented in Figure 7. [10], [9] 6

7 In Figure 5, stimulated recombination of electron-hole pairs takes place in the GaAs quantum well region, where the confinement of carriers and the confinement of the optical mode enhance the interaction between carriers and radiation. In Figure 6, we can observe the changes in density of states for different dimensionalities. The population inversion necessary for lasing occurs more efficiently as the active layer material is scaled down from bulk (3D) to QDs (0D). However, the advantages in operation depends both on the absolute size of the nanostructures in the active region, and on the uniformity of size. A broad distribution of sizes smears the density of states, producing behavior similar to that of bulk material. [10] Figure 5 Schematic of a semiconductor laser 7

8 Figure 6 Density of electronic states as a function of structure size [9] QD lasers acquired more importance after significant progress in nanostructure growth in the 1990 s such as the self-assembling growth technique for InAs QDs. The first demonstration of a quantum dot laser with high threshold density was reported by Ledentsov and colleagues in Bimberg et al. (1996) achieved improved operation by increasing the density of the QD structures, stacking successive, strain-aligned rows of QDs and therefore achieving vertical as well as lateral coupling of the QDs. In addition to utilizing their quantum size effects in edgeemitting lasers, self-assembled QDs have also been incorporated within vertical cavity surfaceemitting lasers. [9] QD lasers are not as temperature dependent as traditional semiconductor lasers. This theory was utilized by applications and in 2004; temperature-independent QD lasers were invented in Fujitsu Laboratories. (which is discussed later in the text) We can summarize the predicted advantages of QD-lasers as [12]: 1. Emits light at wavelengths determined by the energy levels of the dots, rather than the band gap energy. Thus, they offer the possibility of improved device performance and increased flexibility to adjust the wavelength [13]. 8

9 2. Has the maximum material gain and differential gain, at least 2-3 orders higher than QW lasers [14]. 3. advantages of small volume: a. low power high frequency operation, b. large modulation bandwidth, c. small dynamic chirp, d. small linewidth enhancement factor, e. and low threshold current. 4. Shows superior temperature stability of the threshold current. The threshold current is given by the relation: I threshold (T) = I threshold (T ref).exp ((T-(T ref))/ (T 0)), where T is the active region temperature, (T ref) is the reference temperature, and (T 0) is an empirically-determined "characteristic temperature", which is itself a function of temperature and device length. In QDLs T 0 can be high, because one can effectively decouple electronphonon interaction by increasing the intersubband separation. This leads to undiminished room-temperature performance without external thermal stabilization. 5. QD lasers suppress the diffusion of non-equilibrium carriers, resulting in reduced leakage from the active region Basic Characteristics of QD Lasers In a laser, the stimulated emission is amplified by passing the emitted photons to stimulate emission at other locations. (Figure 7) The basic components of a laser are [15]: An active medium (gain medium which is the QD in our case) where population inversion is created by a proper pumping mechanism. The spontaneously emitted photons at some site in the medium stimulate emission at other sites as it travels through it. An energy pump source (electric power supply for QDLs) Two reflectors (rear mirror and output coupler) to reflect the light in phase (determined by the length of the cavity) so that the light will be further amplified by the active medium in each round-trip (multipass amplification). The output is partially transmitted through a partially transmissive output coupler where the output exits as a laser beam (R = 80% in the figure) Figure 7 shows schematic view of the band structure of a typical quantum dot laser. An ideal QD laser consists of a 3D-array of dots with equal size and shape (middle of the figure), surrounded by a higher band-gap material which confines the injected carriers. The whole structure is embedded in an optical waveguide consisting of lower and upper cladding layers (n-doped and p- doped shields). [9] 9

10 Figure 7 Schematic band structure of a quantum dot laser with self-organized dots under forward bias. A 3D array of dots vertically aligned along the growth direction, which is formed during the growth, of multiple QD layers is illustrated schematically. Typically the dot area density in the (100)plane is cm 2 and the dot size distribution is around 10%. The distance between the dot layers is 5 nm and the real dot density in the recombination volume with a thickness of 200 nm is cm 3 for three QD layers. [9] Figure 8 Schematics of a laser cavity [15] 10

11 Figure 9 Scheme of double heterostructure semiconductor laser [15] QD lasers are established by spontaneous formation of QDs at growth temperatures between 460 and 550 C. Fabrication method for such QDs is Stranski-Krastanov growth. The low growth temperature and the low dot density can cause several problems concerning threshold and gain. The cladding layer and the GaAs QD barrier typically grown at these lower temperatures are a possible source for current leakage and non-radiative recombination. On the other hand, the QDs exhibit some intermixing with the surrounding barrier material if temperatures of about 700 C are used to grow high quality cladding layers. [16] Figure 10 Self-organized QDs [16] The self-organization of nanoscale three-dimensional coherent strained islands following Stranski-Krastanov growth mechanism is considered as the most promising way of in-situ QDs fabrication. The ordered arrays so formed may result in distributed feedback and in stabilization of single-mode lasing. In addition, intrinsically buried QDs spatially localize carriers and prevent 11

12 them from recombining non-radiatively at resonator facets. Overheating of facets at high power operation may thus be avoided. A real challenge lies in the optimization of growth parameters to achieve a dense and uniform array of QDs, identical in size and shape. [9], [15], [16] Figure 11 Schematic representation of different approaches to fabrication of nanostructures: Selforganized growth of nanostructures [7] Figure 12 Estimated shape of a self-assembled QD made of InGaAs.[17] 2.3. Required Characteristics for Quantum Dot Laser Applications Quantum dot lasers utilize an oscillator strength that is condensed into a narrow energy width. Because of that reason, the absolute energy level of the QDs should be the same. In other words, the size, shape and alloy composition of QDs should be close to identical. Therefore, the inhomogeneous broadening of QD luminescence is eliminated, and real concentration of the electron energy states can be obtained. If a macroscopic physical parameter is desired, such as light output in laser devices, the density (the number of interacting QDs) should be as high as possible. [17] The reduction of nonradiative centers in QDs is important for QDL applications. Nanostructures made by high-energy beam patterning cannot be used damage is incurred from the beam around the nanostructures. Since the surface-to-volume ratio of QDs is drastically increased compared to 12

13 QWs, this type of damage around the surface of self-assembled QDs is critical for the development of the QDL applications. QDs are put into layered structures to create lasers. At this point, electrical control is very important because an electric field applied to the structure can change certain physical properties of QDs in a desirable way and carriers can be injected into the structure to create light emission. [15], [17] Figure 13 comparison of efficiency between a QWL and a QDL [15] In order for QD lasers compete with QW lasers, two major issues have to be addressed: A large array of QDs has to be used because their active volume is very small. An array of QDs with a narrow size distribution has to be produced to reduce in homogeneous broadening. Furthermore, that array has to be without defects that degrade the optical emission by providing alternate nonradiative defect channels. The phonon bottleneck created by confinement limits the number of states that are efficiently coupled by phonons due to energy conversation. Therefore, it also limits the relaxation of excited carriers into lasing states. This bottleneck causes degradation of stimulated emission (Benisty et al., 1991). However, other mechanisms can be used to suppress that bottleneck effect. (e.g. Auger interactions) [14], [15], [16] 13

14 2.4. Different Types of High speed quantum dot lasers There are several epitaxials were proposed to get the predicted advantages of QD lasers, among them are: overgrowth of QDs with quantum well layers, stacking of quantum dots, close stacking of quantum dots leading to the vertical coupling of quantum dot layers, p-doping of the GaAs barrier layers, etc[18]. Directly modulated quantum dot lasers: Being the key point of the fiber-based datacom application, directly modulated quantum dot lasers could convert electrical signals into digital optical signals at the rate of around 10Gb/s. The modulation speed needs to be further improved, the power consumption should be reduced and the temperature performance needs to be better. Figure 14 (online colour at: BER measurement of QD laser module at 8 Gb/s and 10 Gb/s (a) and at 10 Gb/s for different temperatures (b), inset shows the corresponding eye patterns.[18] Mode-Locked quantum dot lasers With the applications of Mode-Locked quantum dot lasers, several advantages could be received: short optical pulses, narrow spectral width with a small footprint device. Besides, Mode-Locked quantum dot lasers are able to provide a much broader gain spectrum (>50nm), longer cavities (approximately 1cm) [19, 20], sub-ps width and a very low α factor [21] which leads to low chirp. 14

15 Figure 15 (online colour at: Autocorrelation trace of a passively mode-locked quantum dot laser at 1.3 µm and 80 GHz repetition rate. The side peaks correspond to the cross-correlation of two successive pulses, while the middle peak presents the autocorrelation of a pulse (a). Field scan of autocorrelation traces with colorcoded FWHM pulse widths of a 80 GHz passively mode-locked QD laser. Three regimes of operation can be distinguished (b). [18] InP based quantum dot lasers Compared with QW lasers, the emission wavelength of the InP based quantum dot lasers is much lower (0.2nm/K compared to 0.55nm/K). This property could allow this kind of quantum dot lasers operate within a much wider temperature range. Although there still exist some limitations in speed due to the inhomogeneous linewidth broadening, the data transmission could still be possibly over 10Gb/s for InP based quantum dot lasers [22]. Figure 16 Small signal modulation response for a QDash DFB laser at four different pulsed (10 µs pulses at a duty cycle of 2%) bias levels.[23] High power quantum dot lasers With several promising properties of the quantum dot materials, it is widely realized that quantum dot lasers are able to get a good power performance. The advantages of quantum dot materials to be suitable applied to high power application fields are: zero linewidth enhancement factor, the 15

16 free geometric parameters of the quantum dots, e. g. quantum dots size, dots density and size distribution could allow to get the gain without considering the material composition [22]. Extra expensive cooling by Peltier elements is then not needed. QD lasers for coolerless pump sources. The devices with these properties are recently developed based on the GaInAs/Ga(Al)As QD layers emitting at 920nm. In this research, the size of the quantum dot is reduced by modifying the growth parameter and In composition with a constant emission wavelength of the transition. A power splitting of 65ev could be received at room temperature wavelength of 920nm [24] with the size reduced quantum dot structures. Figure 17 (online color at: PL spectra of dot layers with different dot sizes resulting in different transition energy splittings between fundamental and first excited state transitions (a = 47 mev, b = 56 mev, c = 65 mev) [22]. Figure 18 (online color at: Total output power and wall-plug efficiency of a quantum dot high power laser with 1 mm cavity length and 100 µm broad contact stripes. The facets are cleaved without coatings. Maximum cw output power of 3.02 W and a maximum wall-plug efficiency of 55% at 1.5 W are obtained [22]. 16

17 Single mode tapered lasers New device geometry is used to get a similar performance as multi-mode emitting devices. Such kind of lasers enables the amplification of the single mode during the propagation. Compared with quantum well lasers, the wavelength shift is smaller due to the better temperature performance (temperature sensitivity). Figure 19 Emission spectra for 3 different temperatures of a single mode emitting QD tapered laser are shown [24] III. MARKET DEMAND AND NEW TECHNOLOGY 3.1. Market demand Because of the approved advantages of Quantum Dots Lasers, such as low threshold current, enhanced differential gain, lower chirp/high spectral purity, independent of the threshold current on temperature and a decreased a factor, QDs Lasers were intensively researched all through the previous decade. They are suitable to be used in optical applications, microwave or millimeter wave transmission with optical fibers and other telecom and datacom networks. However, QD lasers were commonly regarded as only a theoretical topic which is almost impossible to be brought to the market. The early models were based on the assumptions: Only one confined electron level and hole level Infinite barriers Equilibrium carrier distribution Lattice matched heterostructures The emerge of self-assembling growth technology which forms today the very basis of optoelectronic devices such as edge emitting lasers, which has great potential for the future applications, pushes quantum dot lasers to the boundary between theoretical field and commercial 17

18 applications. Those updated QD based lasers employ fundamentally different models compared to the original models: Lots of electron levels and hole levels Finite barriers Non-equilibrium carrier distribution Strained heterostructures [18] The predictions of decreased α factor and wavelength chirp have already been proved on real devices. In the lightwave applications, lasing in the 1.3um spectral range, using GaAs substrate, both surface and edge emitters have been commercially produced at 6 inch diameter [25]. Nevertheless, as can be expected, due to the challenges listed below, the way of fulfill the QD based lasers into commercial markets is not smooth. First, the lack of uniformity. Second, Quantum Dots density is insufficient. Third, the lack of good coupling between QD and QD. Recently, a Tunnel Coupling Layer for Efficient technology has been published as a Commercial Opportunity Announcement. In order to enhance transportation of electron-hole pairs among quantum dots, get more efficient quantum dot lasers and break the limitation of the older QD technologies, a solution of coupling the sheet of uniform and dense layer of quantum dots, via a thin barrier, to a quantum well (QW) layer. This technology has been proved in the visible red wavelength. InAlGaP was used as the coupling barrier layers and InGaP was used in quantum well layers [26]. [18] has stated that GaAs-based QD lasers will be a good choice for light wave communication networks in terms of performance and expense. Although difficulties were met on the way of realizing QD lasers, with those attractive properties, Quantum Dot laser is still predicted to maintain a hot research field in a few years. Some researchers are seeking some other ways to push their research toward Technology Trends Although Professor Yasuhiko Arakawa of the University of Tokyo predicted that quantum dot lasers do not rely on temperature, the theory is only valid for very low temperatures as room temperature for a long time. Since quantum dot lasers are used in high speed applications such as optical transmitters in metro excess optical systems and optical-lans, a high speed quantum dot lasers which can also operate under high temperatures without cooler are needed to be realized [27]. 18

19 Figure 20 Structure of the new quantum dot laser [27] Figure 21 Temperature dependence of light-current characteristics[10] Figure 22 Modulation waveform at 10 Gbps at 20 C and 70 C with no current adjustments [27] 19

20 Figure 23 Average optical output fluctuations with no current adjustments [27] Using the 3-dimensional quantum dots in the emitting area, the new developed technology successfully received high speed performance at the temperature exceeds room temperature (10Gb/s at as high as 85[27]). The exciting property of temperature independence is achieved by implementing multi-layering quantum dots into 10 layers, p-doping, increasing the layer density (quantum dots density) and using a new structure could get optimized modulation at high speed with low parasitic capacitance. With the results got from the new developed quantum dot lasers, the applications of optical transmitters will be further simplified. The realization of such quantum dot lasers in the future optical metro-access systems and high-speed optical LANs is exciting. Until recent years, most of the QD laser studies were based on InAs/ GaAs system which operates in the 1.3um window, Since the most commonly interesting telecom window is between 1.4um to 1.6um[28] which is the region InAs/ InP based on QD lasers operat in, the direction of QD lasers research is beginning to change. However, the production of InAs/InP based QD laser is even more difficult compared to that of InAs/ GaAs based laser in terms of the growth of isotropic dots. Some progress has been demonstrated, but there still exist several challenges to be solved before the new QD laser could be realized into the market. The first problem is we need to get high gain without degrading the threshold current density. It is required that the density of quantum dots is large enough and we must also be careful of the stacking of QD layers. Similar as mentioned before, good coupling between the QD electronic states and three-dimension electronic states is needed to achieve efficient electron transport. The second challenge rises from the requirement of high T0 to get better temperature performance (temperature insensitivity) especially in uncooled operations. Methods to deal with this problem have been developed: epitaxial layers optimization, P-doping and tunnel injection. The third problem needs to be solved is that in optical-fiber transmission, the lasers must have the ability to operate at high bit rate without or with very low chirp [28]. Since they have really narrow mode-beating linewidth, the QD based lasers are quite suitable to be used in clock-recovery applications. Because of this, the development of low-jitter pulses in mode-locked lasers is a work should to be done. Fulfillment of this development will lead to wider applications of the QD based lasers to high-repetition rate sources, microwave applications and all-optical clock recovery. 20

21 Two months ago, a group of researchers who were with Alcatel Thales III V Laboratory, Route D epartementale published their recent work on InAs/InP quantum dots in a barrier and dots in a well (DWELL) heterostructures which operates at 1.5um. They demonstrated that, on InP (0 0 1), for both dots in a barrier and dots in a well structures, short cavity length was achieve under CW room-temperature operation. In contrary with the 1.3-um InAs/GaAs, the 1.5-um InAs/ InP based lasers can operated on ground state for much shorter cavity length. In the paper, the group also reported a high T0 achieved for both P-doping and optimized DWELL structures. For optimized DWELL structures, T0 is higher than 100K for Broad Area (BA) lasers and 80K for singletransverse mode lasers within the temperature region without increasing the threshold current. For the first time, this research group presented in realizing the buried ridge stripe (BRS)- type single-mode distributed feedback (DFB) lasers, a 45 db high side-mode suppression ration (SMSR) is achieved. The first buried DFB DWELL which operates at 10Gb/s in the 1.55um range strongly demonstrated the potential of the use of InAs/InP system in future applications such as optical sources for telecommunication. Besides, the work they ve done also includes the demonstration of an surprisingly narrow linewidth. Compared to the QW lasers or bulk counterparts, this merit will bring a very good phase noise and time-jitter characteristics under the condition that the lasers are actively mode-locked [27]. Figure 24 Schematic representation of QD-based active structures: Dashes inserted directly in the barrier, or within an intermediate quantum well [27]. As can be seen from above, there are several drawbacks block off the way of InAs/GaAs based QD lasers to the commercial market. Although the implementation of new substrate material (i. e. InP ) instead of traditional GaAs has proved to be promising, there still exist some directions need to be researched. Since the 3-D quantum structure needs to be enhanced by further refining the manipulation on the nanostructure shape. Second, due to the disfigurement of the current QD lasers heterostructures, refinement should be implemented on the structures to get efficient carrier transportation leads to much better dynamic performance. Third, the noise is also an issue which needs to be solved. Quantum Dot based mode-locked lasers can still be further researched to broaden their applications in both optical fiber communications and millimeter-wave generations by their good performance on phase noise and time-jitter characteristics. 21

22 IV. FUTURE The advantages of quantum dot based lasers compared to other conventional technologies have been realized for several years. Especially the free geometric parameters of quantum dot layers give probabilities to tailor the spectral gain profile applied to different types of QD lasers applications. [23] Nevertheless, due to the intrinsic limitation of technologies, to realize quantum dot lasers with predicted properties met several difficulties. The requirement of further widening the parameters range in order to reducing the inhomogeneous linewidth broadening (we need homogeneous linewidth) is one of the aspects of developing quantum dot lasers. Using surface preparation technologies, lots of groups are working on the issue of further controlling the position and dot size for the self-organized technology. Once the developed methods can be implemented in the high density systems, the new technology will become the breakthrough in the history of quantum dot lasers development. Since the speed of carrier capture extremely increase the transport time and affects the modulation bandwidth, it is required to decouple the carrier capture from the escape procedure. Employing tunnel injections to quantum dots is a choice. Allowing the injection of cooled carriers, this method is able to achieve good performance without loosing the extra carriers which often happens before due to the thermal relaxation. With the experiment done by comparing the QW lasers and QD lasers in term of raised gain at the fundamental transition energy with the constant broad band characteristics of quantum dot lasers, it is concluded the combination use of quantum dot and quantum well would tailor the material properties in a much wider range than using quantum dots or quantum wells alone [24], [34]. With the employment of further control of parameters and better coupling technology and the breakthroughs which are already done, realizing quantum dot lasers as well as other quantum dot optoelectronic devices in commercial market is not so far away. V. CONCLUSION During the previous decade, there was an intensive interest on the development of quantum dot lasers. The unique properties of quantum dots allow QD lasers obtain several excellent properties and performances compared to traditional lasers and even QW lasers. Although bottlenecks block the way of realizing quantum dot lasers to commercial markets, breakthroughs in the aspects of material and other properties will still keep the research area active in a few years. According to the market demand and higher requirements of applications, future research directions are figured out and needed to be realized soon. 22

23 REFERENCES [1] Wikipedia, the Free Encyclopedia, Quantum Dots. [Online]. Available: [2] Evident Technologies, How Quantum Dots Work. [Online]. Available: [3] V. Mitin, V. Kochelap, M. Stroscio, Quantum Heterostructures: Microelectronics and Optoelectronics, ch. 6. [4] M. Heninia, M. Bugajskib, Advances in self-assembled semiconductor quantum dot lasers. [5] W. Buhro, V. Colvin, Semiconductor Nanocrystals: Shape matters [6] B. Saleh, M. Teich, Fundamentals of Photonics, ch.15. [7] D. Bimberg, M. Grundmann, N. Ledenstov, Quantum Dot Heterostructures [8] Y. Arakawa, T. Someya, and K. Tachibana, Progress in Growth and Physics of Nitride- Based Quantum Dots [9] H. Goronkin, P. Allmen, R. Tsui, T. Zhu, Functional Nanoscale Devices, ch.5. [10] D. Bimberga, M. Grundmanna, F. Heinrichsdorffa, N.N. Ledentsovb, V.M. Ustinovb, [11] A. Zhukovb, A. Kovshb, M. Maximovb, Y. Shernyakovb, B. Volovikb, A. Tsatsul'nikovb, P.S. Kop'evb, Zh. Alferovb, Quantum dot lasers: breakthrough in optoelectronics [12] M. Datta, Z. Dilli, L. Wasiczko, [13] K. Eberl, Quantum-dot Lasers [14] N. N. Ledentsov, Quantum Dot Heterostructures: Fabrication, Properties, Lasers [15] P. Prasad, Nanophotonics [16] D. Bhattacharyya, A. Bryce, J. Marsh, C. Sotomayor-Torres, Self-organised InGaAs / GaAs [Online]. Available: [17] Y. Masumoto, T. Takagahara, Semiconductor quantum dots : physics, spectroscopy, and applications [18] D. Bimberg, G. Fiol, M. kuntz et al. High speed nanophotonic devices based on quantum dots. phys, stat, sol.(a)203, [19] M. Kuntz et al, New J. Phys , [20] A. Gubanko et al. Electron. Latt. 41 (20) , 2005, [21] A. Martinez et al. Appl. Phys. Lett. 86(21), 21115, 2005 [22] J. P. Reithmaier, A.Somers, W. Kaiser, et al. Semiconductor quantum dots devices recent advances and application prospects, phys, stat, sol(b), 243, [23] D. Hadass, R. Alizon, H. Dery, V. Mikhelashvili, et al. Phys. Lett , [24] S. Deubert, R. Debusmann, J. P. Reithmaier and A. Forchel. Electron. Lett. 41, [25] D. Bimberg, M. Kuntz, M. Laemmlin, Auantum Dot Photonic Devices for Lightwave Communication. Applied Physics. A. 80, 2005, pp [26] Website: [27] Fujitsu, University of Tokyo Develop World s First 10Gbit/s Quantum Dot Laser Featuring Breakthrough Temperature-Independent Out put, [28] Jean Landreau, Olivier Drisse, Estelle Derouin, et al. Recent Advances on InAs/InP Quantum Dash Based Semiconductor Lasers and Optical Amplifiers Operating at 1.55um, IEEE Journal of Selected Topics in Quantum Electronics, vol 13, [29] Ephraim Suhir. Microelectronics and Photonics-the Future. Proc. 22 nd International Conference on Microelectronics, [30] L. Joulaud, C. Paranthoen, A. Le Corre, et al. InAs Self-assembled quantum dot and quantum dash lasers on Inp for 1.55um optical telecommunications IEEE,

24 [31] Y. Arakawa, T. Someya, and K. Tachibana. Editor s Choice Progress in Growth and Physics of Nitride-Based Quantum Dots, Phys, Stot. Sol [32] R. Schwertberger, D. Gold, J. P. Reithmaier, and A. Forchel. Long-Wavelength InP-Based Quantum-Dash Lasers. IEEE Photonics Technology Letters, Bol. 14, 2002 [33] Nikolai N. Ledentsov, Vitaly A. Shchukin. Novel Concepts for injection lasers. Society of Photo-Optical Instrumentation Engineers, 2002 [34] John E Midwinter. The Future Development of Optical Communication Systems. IEEE,