21-01-2010, 11:43 AM
[attachment=1258]
ABSTRACT
Quantum Dots(QDs) are small conductive regions in a semiconductor, containing a number of charge carriers(from one to thousand) that occupy well defined discrete quantum states. They have typical dimensions between nanometers to a few microns .When the space at any side, around a material shrinks to 100Ã…, quantisation of the energy levels at the reduced side will occur. In quantum dots electrons are confined in all directions to a volume in space with dimensions on the order of their de Broglie wave length, ie, they have no kinetic energy and as a result they occupy spectrally sharp energy levels like those found in atoms. This is the reason why they are often referred as artificial atoms. Quantum dots can used to emit lasers, that is they are semiconductor type lasers. Because of fundamentally different physics that stem from zero dimensional electronic states, quantum dot lasers surpass other type of laser technologies in several respects, these include their minimum threshold current density, the threshold dependence on temperature, slightly larger wall plug efficiency etc¦
Quantum dots can naturally form self buried heterostructures and for the production of QD lasers it consist of a gain medium provided by the semiconductor material and optical feed back by resonator. Pumping over P-N junction is achieved by current injection and carriers are free into an active region, where lasing transitions take place. Semiconductor lasing structures require optical confinement. For quantum dot laser production this is achieved by sandwiching the quantum dots in the structure within a separate confinement heterostructure(SCH)
1. INTRODUCTION
The past decade has seen a tremendous amount of research in the fabrication of semiconductor structures, which was stimulated by the drive towards increasing miniaturization and performance of solid-state devices. One major step in these developments has been the development of low dimensional devices.
Engineering of less than three-dimensional semiconductors began during early 1970s, when groups at Bell Labs and IBM made the first two-dimensional quantum wells. These structures have provided new kinds of optical and electronic devices like lasers in CD player; low noise amplifier in direct broadcast satellite receivers and laser sources for fiber optic communication. Modification of these structures to confine the electrons to one-dimension (quantum wires), and zero-dimension (quantum dots) lead to new electrical and optical properties.
1.1. QUANTUM DOTS
Quantum dots (QDs) are small conductive regions in a semiconductor containing a variable number of charge carriers( from one to thousand) that occupy wellâ€defined discrete quantum states. They have typical dimensions between nanometers to a few microns. When the space, at any side around a material shrinks to 100A, quantization of the energy levels at the reduced side will occur. In quantum dots electrons are confined in all directions to a volume in space with dimensions on the order of their de Broglie wavelength. Therefore they have no kinetic energy and as a result, they occupy spectrally sharp energy levels like those found in atoms. This is the reason why they are often referred to as artificial atoms. However unlike atoms, a solidâ€state environment always surrounds quantum dots and they are three orders of magnitude larger than that of the real atom, Also, they can be easily connected to electrodes and are therefore excellent tools to study atomic-like properties. The size and shape of these structures and therefore the number of electrons they contain can be precisely controlled; a quantum dot can have anything from a single electron to a collection of several thousands and an experimentalist can scan through the entire periodic table by simply changing a voltage.
While it is possible to make quantum dots with different types of materials, using Semiconductor quantum dots has an important obvious advantage: the possibility of integrating the quantum dots with other structures and devices, taking advantage of sophisticated technology that has already been developed. Achieving quantum dots in semiconductors requires having a very small volume of a low band gap material imbedded in larger bandgap material(s), thus creating the potential barrier for quantum confinement of the carriers.
Because of the rich, novel physical properties, exhibited by these systems such as Coulomb Blockade, quantum confinement, exchange enhancement and shape dependent spectroscopy and also because of their promise for applications such as all-optical Switches, logic gates, lasers, LEDs, transistors etc, there has recently been considerable interest in the electronic, optical, transport and structural properties of semiconductor quantum dots. In this report, we are going to focus only on a specific quantum dot device, viz, the semiconductor quantum-dot lasers.
2. QD LASER
Since the 1994 demonstration of a QD semiconductor laser, the research progress in developing lasers based on QDs has been impressive. Because of their fundamentally different physics that stem from zero-dimensional electronic states, QD lasers now surpass the established planar quantum well laser technology in several respects. These include their minimum threshold current density, the threshold dependence on temperature, slightly larger wall-plug efficiency, and range of wavelength obtainable in given strained layer material systems.
Since the QDs grow to naturally form self-buried heterostructures, they represent a key potential component of future microcavity light emitters based on oxide-confined vertical-cavity surface-emitting lasers (VCSELs), photonic band-gap defect lasers, and micro-curved resonators. Advanced crystal growth, such as molecular beam epitaxy (MBE) or metal-organic chemical vapour deposition (MOCVD), allows artificial 111-V materials to be made using elaborate heterostructures need for these quantum devices.
The hindrance came from the fabrication of the Quantum dot array structure needed. There have been many attempts to reduce the size of Quantum dots to increase the quantum confinement. Another factor is the difficulty in producing the highly regular quantum dots in the array. Highly regular quantum dots are required because the more regular the quantum dots, the sharper the gain profile and hence better spectral quality in the laser output.
2.1. PRINCIPLE OF OPERATION: QUANTIZATION OF
DENSITY OF STATES
In conventional double heterostructure lasers, a thin (0. 1 -0.m) active region of lower bandgap material (e.g. GaAs) bounded on either side by a larger bandgap material (e.g. AIGaAs), acts as a trap for electrons and holes, thereby reducing the required threshold current density. If the thickness of the active layer is reduced to 50-10OA, the dimension become comparable with the de Broglie wavelength of the thermalized electrons, and the confined electrons and holes display quantum effects. Since the movement of the charge carriers is restricted in all three dimensions in a QD, the degeneracy of energy levels is largely lifted, and the density of states becomes extremely quantized,as found by solving the time-independent three-dimensional Schrodinger equation. The smaller the dimensions of the QD, the larger the separation between adjacent energy levels. In fig 2.1, we compare the density states in different confinement configurations. Ideally, a QD is a point with zero dimensions. Hence, the density of states should be a sequence of delta functions at the allowed energies. However, since there is finite dimensions associated in all three directions, the physical structure is like a quantum box of volume dld2d3, rather than a dot, and the DOS spectrum shows a finite linewidth, though infinitesimally small. Energy is quantized as,
E=Ec+EQ1+EQ2+EQ3
Where EQn = h2(qn/dn)2/(82mc),n = 1,2,3
Here q 1,2,3 are the triplet of quantum numbers associated with an energy subband. Each of q 1,2,3 takes integral values, and d1,2,3 are the dimensions of the dot.
fig 2.1 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 subband, which get successively narrower as the electron motion is restricted in more dimensions.
For efficient lasing, it is desirable to have a large density of carriers in both electron and hole bands at energies close to the band-edge, so that population inversion (i.e. there should be more electrons in the excited state than in the ground state) becomes easier. Short population time of ground states in QD lasers ensures a huge number of optical transitions per unit volume, most of them being radiative recombination, rather than non- radiative, resulting in high internal efficiency. Energy pumped into the system raises the charge carriers from one energy level to the next; none of it goes to random motion, because there are no other degrees of freedom. Thus, one expects any lasing from the dots to occur with high efficiency and at lower threshold current than in either quantum wells or quantum wire.
2.2 EXPECTED ADVANTAGES
1. QD lasers should be able to emit 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.
2. QD lasers have the maximum material gain and differential gain, at least 2-3 orders higher than quantum-well lasers.
3. The small active volume translates to multiple benefits, such as low power high frequency operation, large modulation bandwidth, small dynamic chirp, small Iinewidth enhancement factor, and low threshold current.
4. QD lasers also show superior temperature stability of the threshold current. The threshold current is given by the relation,
Ithreshold(T) = Ithreshold(Tref). exp ((Ti- Tref)/To),
where T is the active region temperature, Tref is the reference temperature, and T0 is an empirically-determined Ëœcharacteristic temperature, which is itself a function of temperature and device length. In QD lasers T0 can be high, because one can effectively decouple electron-phonon interaction by increasing the intersubband separation. This leads to undiminished room-temperature performance without external thermal stabilization.
5. In addition, QD lasers suppress the diffusion of non-equilibrium carriers, resulting in reduced leakage from the active region.
6. More novel structures such as distributed feedback lasers and single-dot VCSELs promise ultra-stable single mode operation.
2.3 QUANTUM DOT REQUIREMENTS
To be useful for devices, a QD should satisfy several important requirements. The lower limit (Dmin)of the size of a quantum dot is defined as the size at which only one electronic level exists in the quantum dot. This critical dimension (Dmin) strongly depends on the conduction band discontinuity ( Ec) in the corresponding heterojunction used to fabricate the quantum dot according to the relation,
EC =(hbar)22/(2.Me.Dmin)2
AssumingEc~0.3 eV for e.g. direct bandgap GaAs †A10.4Ga0.6As quantum wells, this would mean that the diameter of the quantum dot be smaller than 40 Ã…. This is an absolute lower limit of the QD size, since for QDs of this or even slightly larger size the separation between the electron level and the barrier continuum energy is very small and at finite temperatures evaporation of carriers from QDs will result in their depletion. For the InAs †A1GaAs system the conduction band offset is larger, while the electron effective mass is smaller, and the critical size is similar.
If the separation between energy levels becomes comparable to the thermal energy (kT), population of higher-lying energy levels cannot be neglected and both static and dynamic properties deteriorate to some extent. This equation establishes an upper limit (Dmin) for GaAs †A1GaAs QDs of~120 Ã… , and of~200Ã… for InAs †GaAs QDs, due to the much lower electron effective mass in the latter case. Statistical carrier capture into thermally decoupled QDs might lead, however, to excited state population still at low excitation levels. Thus the upper limit of size is decided by the requirement,
Thermal energy, kT>distance between energy levels,
Also, for applications to optoelectronic devices, the QDs should not contain dislocations or point defects. Dense arrays of QD (~1011 cm2) are required to realize high modal gainof laser. Ordering of QDs in the substrate plane and the possibility of creating periodic lattices of QDs in all three dimensions are also desirable in a number of cases.
2.4. BASIC STRUCTURE OF QD LASER
Fig. 2.4.1 A semiconductor Laser
The essential components required for a laser is a gain medium provided by the semiconductor material and optical feedback by the resonator. The amplification is by radiative recombination through QW or QD transitions. Pumping over p-n junction is achieved by current injection while cleaved crystal facets works as mirrors ( Fabry-Perot resonator) -
Stimulated Emission in Semiconductor is shown in figure 2.4.2,
In the figure Nph = photon density
B21 = Einstein coefficient
v, c = density of states
fv, fc = Fermi functions
The gain g is a function of the density of states.
Fig. 2.4.2 stimulated emission in semiconductor
In all semiconductor diode lasers, electrons and holes are injected from 3D contact regions, where carriers are free, into an active region, where lasing transitions take place and where carriers may be dimensionally confined. In lasers with 3D (bulk) or 2D (Q\V) active regions, there is always a population of carriers distributed according to Fermi Dirac statistics within some energy range (determined by the temperature and the injection level) around the lasing transition energy (Fig. 2.4.3). These carriers reside in the active region itself and their recombination contributes to a T-dependent threshold.
Semiconductor lasing structures require optical confinement. For quantum dot laser production, this is obtained b sandwiching the quantum dots in the structure within a separate confinement heterostructure (SCH).
Consider the QD laser heterostructure (Fig.2.44). Carriers are injected from cladding layers into the optical confinement layer (OCL) where an approximate equilibrium with the QDs is established at room temperature. High occupation of QD electron and hole levels embedded in the OCL is therefore accompanied by an appreciable 3D population of both types of carriers in the OCL itself.
Fig.2.4.3 Career Population in bulk Fig 2.4.4 schematic
QW, and Single QD. The dashed energy band diagram of a
Arrow shows the excited state trasnsition in QD QD laser
2.5 CAPTURE TIMES AND THE PHONON BOTTLENECK
It was recognized very early that the relaxation of carriers in QDs would be much longer than in QWs. Theoretical studies identified what has since become known as the phonon bottleneck in QDs: since the excited and ground states are not typically separated by phonon energies of ~36meV, single-phonon-assisted relaxation events between these levels are forbidden. Multiple phonon events, while permitted, are typically much slower (>lns). If the only available mechanism for carrier relaxation were carrier phonon scattering, the bandwidth of the QD lasers would be forever limited to a few GHz.
However there are several other mechanisms for carrier relaxation that has been suggested that provide a much faster relaxation path. In particular, methods relying on carrier-carrier interaction rather that carrier-phonon interaction can take place much more rapidly. Both Auger-like mechanisms, in which a relaxing electron transfers energy to another electron that is promoted into the continuum, and electron-hole scattering, in which a relaxing electron transfers energy to a ground state hole, have been suggested.
The Carrier statistics in quantum dots requires a fundamentally different treatment than in any higher dimensional semiconductor structure. Detailed description using microstates can help explain the phenomena of level population, laser threshold and recombination dynamics.
2.6 BANDWIDTH LIMITS
Before discussing in detail how the dynamics of QDs affect the performance of QD devices, in particular directly modulated lasers, it is important to mention briefly what generally limits the bandwidth of semiconductor lasers and the typical methodology for analyzing semiconductor laser performance. Typically high-speed lasers are analyzed using a three-rate-equation model, in which the number of photons, carriers in the active region, and carriers in the core are modeled in three distinct equations.
This model gives rise to two fundamental limits on modulation bandwidths in semiconductor lasers. Since the modulation is output light/input current, one limit f-3dB-capture, derives from how quickly carriers can be transported to the active region. This capture time limit includes transport to the active region, relaxation down the lasing wavelength. The second limit ( the K-factor limit, F3dB K )comes from the fundamental dynamics, such as differential gain and gain compression, of the active region. Although the differential gain is increased in QD lasers, the gain compression increases along with it, and current experimental evidence (along with the theory for a particular physical cause for gain compression) suggests that this K-factor limit also increases with capture time.
3. QD LASER DESIGNS
3.1. In0.67Ga0.33Al/GaInAsP/InP tensile strained QD Laser
In 1994 In0.67Ga0.33Al/GaInAsP/InP tensile strained QD Laser structure under current injection was reported. In this report it was argued that using a tensile strain (TS) instead of a compressive strain (CS) provide larger size quantum dots for longer wavelength applications.
It was grown by two-step MOVE, EPX writing, and wet chemical etching. The quantum dots are situated in the middle of an optical confinement layer (OCL), which has a total thickness of 400nm. Figure 4 shows the lasing spectrum and the output verses current characteristics of the laser operating at 77K.
Fig 3.1 figure and schematic of a tensile strained (TS,) QD laser
3.2 Self-Organized In0.5Ga0.5As QD Laser
In 1995 Shorji et al. published an article on successfully obtaining Laser oscillation from In0.5Ga0.5As self-organized quantum dots. The laser was operated by current injection at 80K. They argued that the size of the quantum dots fabricated by MBE or MOVPE are too large to provide significant quantum confinement effect, and the quantum dots have very poor uniformity. Therefore in their design they used Atomic Layer Epitaxy (ALE) instead.
Figure 3.2 shows the schematic of the quantum dot laser. Each of the quantum dots has a diameter of 2Onm and height of l0nm. In0.1Ga0.9As barriers separate the In0.5Ga0.5As quantum dots and the whole array is sandwiched between two GaAs separate confinement heterostructure layers. Most of the light energy was confined between the two cladding layers i.e. the p-Al0.45Ga0.55As and the n-In0.47Ga0.53P layer. Hence the laser output is emitted in a direction parallel to the dot array.
Fig 3.2 Schematic of the Self Orgwiized QD Laser
3.3. High Power Continuous-Wave lnGaAs/AIGaAs QD laser
In 1998 M. Maximov et al. reported a successful demonstration of high power (1W) InGaAs/A1GaAs QD laser. After many years of QD laser development they realized that at high temperature (180K) electrons and holes are evaporated from the confinement of quantum dots and thus decreases the gain and increases Jth. In this design they implemented a structure called Vertically Coupled Quantum dots (VCQDs) to correct this problem.
In the VCQD design layers of quantum dots are grown, and each quantum dot is aligned with a quantum dot in the layer below and quantum dot in the layer above hence forming columns of quantum dots called VCQDs. In each VCQD column the distance between the adjacent dots are small that their waveforms overlap. As a result VCQD can be considered as a quantum confined structure.
3.4. InGaN QD Laser
In 1999 Tachibana et al. reported successful lasing action from an InGaN QD laser under optical pumping, GaN is one of the newer semiconductor materials in the field of photonics and optoelectronics; many of its properties are still under investigative study
Fig 3:4 Schematic of a inGaN QD laser
The design of this laser is very similar to the VCQD laser. There are 10 layers of quantum dots stacked on top of each other. Each layer has a thickness of 5nm and QD density of 6x109cm-2. The quantum dots are made from In0.2Gas0.8N and the barrier region is made of In0.02Ga0.98N. The layers are grown by MOCVD, and the laser cavity was fabricated by low electron cyclotron reactive ion etching (ME) with Al/Cl2 gases.
3.5. Vertical Cavity QD Laser design
This design reported by Nishioka et al. is a more exotic vertical emitting laser, i.e. the laser output is emitted perpendicular to the deposited Layer and the QD array. The schematic of the laser is shown in the figure 3.5
Fig 3.5 Schematic of the vertical cavity QD laser
The structure was fabricated in a low-pressure horizontal MOCVD quartz reactor. The quantum dots were created using SK growth mode. This design made use of Distributed Bragg Reflectors (DBR) whose function is to control the emission spectrum by creating and enhancing the resonator mode of desired emission wavelength. The Vertical Cavity Surface Emitting Quantum Dot Laser will be discussed in detail in section 8.
4. FABRICATION OF DOTS
The unique advantages of QD structures can be realized only if the dots are as uniform as possible in shape and size. Conventional semiconductor-processing techniques that are based on lithography and etching face inherent problems such as limited resolution, and the introduction of surface defects during production. As a result, several research groups have started working on the direct synthesis of quantum nanostructures either by combining epitaxial growth techniques (MBE or MOCVD) with photolithography,or by introducing just the right amount of crystal strain so that planar epitaxial growth becomes thermodynamically unfavorable. The second approach is more popular because of the self-organizing characteristics.
4.1 COHERENTLY STRAINED, SELF-ASSEMBLED THREE- DIMENSIONAL ISLANDS
When a semiconductor material is deposited on a substrate that is made from a material with significantly smaller lattice constant (< 8% difference in lattice constants), the atoms of the overlayer. arrange themselves in a two-dimensional layer, called the wetting layer, and as the growth proceeds, clusters themselves into pyramid-shaped 3- dimensional islands, connected by the wetting layer called the Stranski-Krastanow growth mode. (it is discussed in detail in section 4.1.1 ). The competition between the surface and interface tensions causes the atoms to bunch up, so that the lattice can elastically relax the compressive strain, and thus reduce the strain energy within the islands.
Fig 4.1 island formation, when the lattice mismatch is relatively large
(b) Pseudomorphic growth when the mismatch is small
In an alternative growth mode known as Pseudomorphic growth, the epilayer is laterally compressed to match the substrate lattice, The lateral strain is automatically introduced as the growth proceeds. Once dots are formed, they are covered with lattice- matched material. This is called passivation. In other words, the smaller bandgap material must be completely embedded in the barrier material without any crystal defects.
4.1.1 The Stranski-Krastanow growth mode
The formation of Stranski-Krastanow islands is, as mentioned above, closely related to an epitaxial misfit and the accumulation of elastic strain energy in the epilayer. Strain relaxation takes place through the rearrangement of the deposited material when 3D islands are formed. The formation of 3D islands changes the strain situation completely. Deposition starts with complete wetting of the substrate. The total energy of the system decreases until the substrate is covered by one monolayer of deposit. This wetting is due to the energy contribution from the substrate/epilayer interface. Further deposition will form a uniformly strained film on a rigid substrate, and the elastic strain energy, Estrain will increase linearly with the layer thickness, t. We can write this as:
Estrain = 2 At
Where is the elastic modulus, =a/a is the misfit (a is the lattice parameter) and A is the surface area.
The formation process can be illustrated as in Figure 4.1 .1. The process can be divided into three main steps:
¢ The 2D growth period (A)
¢ The 2D-3D transition period (B), and
¢ Ripening of islands period ©.
The process starts with the growth of stable 2D layers and a perfect wetting layer is formed. When we reach the critical thickness, the growth is still 2D but now metastable, which means we have a supercritical thick wetting layer. Metastable growth continues until the excess strain energy, Ee is high enough to overcome the activation energy, EA, when the 3D growth starts. Material from the strained 2D layers is relocated in this 2D/3D transition and we end up with a wetting layer of a thickness close to t
Fig 4.1.1 Schematic of total energy versus time for the 2D-3D-morphology Transition A) the 2D growth period, B) the 2D-3D transition period, and C) the ripening of the island period
4.2 ARRAYS OF VERTICALLY COUPLED QUANTUM DOTS
One can add a new dimension to the study of quantum-dot lasers by stacking the individual layers. Because of the interacting strain fields, islands in one layer tend to form above those in the layer below, if the separation is not too great (fig 4.2.1). Stacking layers of dots in the laser structure increase the amount of active materials in devices.
Significant reduction in photoluminescence linewidth seen from the stacked layers is attributed to the more uniform dot sizes caused by the correlated nucleation of the Dots. Better uniformity of dots is achieved at the top surface of a stack than an iso1ated monolayer(fig 4.2.2). Vertical correlation and electronic coupling/decoupling between the dots open up the possibility of some novel micro- and optoelectronic devices.
Fig 6.2.3 Atomic force micrograph shows greater homogeneity of dots on the top surfaceâ„¢ of a stack of 20 layers of dots (left,) than a single quantum-dot lover. Dots were made by depositing SiGe on a Si substrate.
5. QUANTUM DOT VCSELS
Much of the present focus on quantum dots is driven by the promise of inexpensive lasers and detectors for the 1 .m telecommunications wavelength, utilizing the zero- dispersion window of an optical fiber. There has been an additional incentive to develop lasers grown on GaAs substrates, for easy integration of optical devices with the relatively mature GaAs electronic device technology, moving towards the development of high- speed optical communication systems.
In a very generic grouping, semiconductor lasers can be divided into two types: edge-emitters and vertical cavity lasers. For review, the pictures listed below show a broad area laser (edge-emitter) and a Vertical Cavity Surface Emitting Laser (VCSEL). The basic difference between these two types of lasers is that in the VCSEL, the optical beam travels perpendicular to the active region, while in the more conventional edge-emitter; the optical beam oscillates in the plane of the active region.
Fig 5.1 QD VCSEL with a 10 layer In GaAs active region.
Fig 5.2 (left) shows the lasing spectrum of a QD VCSEL at 960.4 nm.
Fig. 5. 3(right) the emission spectra of the 10 period dot region at varying temperatures. The dashed line indicates the resonant wavelength of the vertical cavity.
The interesting effect seen in Fig. 5.3 is that the peak of the emission spectra shifts to shorter wavelengths as the current is increased. The QD gain will saturate and in turn shift the peak to a shorter wavelength. It is likely that this ground state saturation is due to hole burning or carrier escape (Hole burning is the phenomena where a large flux of single frequency photons is applied to an inhomogeneously broadened.medium, causing the gain to saturate for that frequency where the lineshape function overlaps the photon frequency).
In VCSELs, the number of available cavity modes is limited by the aperture size. Higher modes tend to have much more loss because of the small size of the aperture. Therefore, if the cavity is designed correctly, the researcher can effectively force the lasing from mode the QDs, even though there may be a larger gain at higher energy states. In an edge-emitting laser, it can be clearly be observed that there will always be one direction where the cavity will be very long relative to the other two directions. Therefore, there is a greater distribution of optical modes available for the laser to lase at. If a higher energy transition lases first (implying shorter wavelength lasing), the Fermi level of the cavity will be pinned, and the ground state and first excited state may never get the opportunity to lase.
Fig. 5.4 Schematic showing the cavity mode and electronic density of states of the gain medium compared for a quantum well and quantum dots
Quantum dot VCSELs are perhaps one of the most promising photonic applications which intends to utilize the QD to its fullest, The cavity design of the VCSEL can force the QD ensemble to lase in lower order states, thereby maximizing the temperature independence and minimizing the threshold current of the laser. Figure 5.4 shows clearly that the greatest advantage of using a vertical cavity design. Clearly, aligning the high optical gain of the QD to the discrete density of states results in a highly efficient laser.
6. TEMPERATURE INSENSITIVE QD LASER
Despite significant recent progress in the fabrication of QD lasers, their temperature stability has fallen far short of expectations. Even though the best results for the all important parameter, describing empirically the temperature dependence of the threshold current density and defined as, T0 = l/(Injth /T) are quite respectable for QD lasers, matching, and even exceeding the best results reported for QW lasers at room temperature, so far they have been nowhere near the predicted infinite values that would allow one to regard the laser as temperature insensitive.
The dominant source of the temperature dependence of jth is parasitic recombination outside the QDs, primarily in the OCL. In the conventional design, the OCL is a conductive medium where the QDs are embedded in such a way that carriers in the OCL and in QDs are in thermal equilibrium at room temperature. Consequently. the component of jth associated with recombination in the OCL depends exponentially on T and the total threshold current becomes temperature dependent.
Another mechanism of the -dependence in QD lasers is the inhomogeneous line broadening due to the QD size dispersion. Experimental progress in controlling the QD parameters during the structure growth has been impressive; nevertheless, even in the best devices the measured gain and spontaneous emission spectra still indicate a significant QD size dispersion. Physically, the effect of QD size dispersion on T is similar to that due to recombination in the OCL in the sense that the inhomogeneous line broadening is associated with undesired pumping of nonlasing QDs. So long as the electron and hole populations in the nonlasing QDs are in equilibrium with those in the active QDs, the fraction of Jth arising from the recombination in nonlasing QDs depends on T and the characteristic temperature T0 is no longer infinite. Quantitatively, the effect of inhomogeneous broadening on is further discussed below.
Still another mechanism of the T-dependence of jth is associated with the violation of charge neutrality in QDs. This leads to a temperature dependence of the recombination current in the lasing QDs themselves arising from the fact that carrier populations there are no longer fixed by the generation condition. Violation of charge neutrality is the dominant mechanism of temperature sensitivity at low temperatures but is unimportant at 300 K.
Elimination of the OCL recombination alone results in a dramatic improvement of the temperature stability. T0 accomplish this we propose a novel QD laser design, based on tunneling injection of carriers into the QDs where in they recombine radiatively. Our design allows to both suppress the parasitic components of threshold current and diminish the effect of inhomogeneous line broadening.
Carrier injection by tunneling has been successfully tested in the context of QW lasers. Bhattacharya and coworkers have realized tunneling-injection QW lasers and: demonstrated improved modulation characteristics, lower wavelength chirp, and superior high-temperature performance as compared to conventional QW lasers.
One approach is based on direct injection of carriers into the QDs, resulting in a strong depletion of minority carriers in the regions outside the QDs. Recombination in these regions, which is the dominant source of the temperature dependence, is thereby suppressed, raising the characteristic temperature above 1500 K. Still further enhancement of T0 results from the resonant nature of tunneling injection.
6.1 TUNNELING INJECTION QD LASER STRUCTURE
A.schematic view of the structure and its energy band diagram are shown in Fig. 6. 1. Basically, we have a separate confinement double-heterostructure laser. Electrons and holes are injected from - and -cladding layers, respectively. The QD layer, located in the central part of the OCL, is clad on both sides by QWs separated from the QDs by thin barrier layers. Injection of carriers into QDs occurs by tunneling from the QWs.
The key idea of the device is that the QWs themselves are not connected by a current path that bypasses the QDs. Electrons (coming from. the left in Fig. 6.1) can approach the right QW only through the confined states in the QDs. Similarly; holes cannot directly approach the left QW.
Fig 6. 1. (a) Schematic view and (b) the energy band diagram of a tunneling-injection QD laser. The tunnel barrier on the electron-injecting side is made thicker to suppress the leakage of the holes from the QD.
To realize this idea, the following conditions must be met.
1. The QW material and the thickness should be chosen so that the lowest subband edge in the injecting QW matches the quantized energy level for the corresponding type of carrier in the average-sized QD (the QWs may or may not be of the same material as the QDs). -
2. The barriers should be reasonably high to suppress the thermal emission of carriers from the QWs.
3. The material separating QDs. from each other in the QD layer should have a sufficiently wide bandgap to suppress all tunneling other than via the QD levels. This material may be the same as that of the barrier layers;
4. The barrier layers should be thin enough to ensure effective tunneling between the QW and .QD states. At the same time, the separation between the adjacent QDs in the QD layer should be large enough to prevent any significant tunnel splitting of the energy levels in neighbouring QDs (otherwise, such a splitting would effectively play the same role as the inhomogeneous line broadening).
7. DYNAMIC PROPERTIES
The gain compression coefficient is equal to the inverse of the saturation photon density coefficient Ssat in the laser resonator. It defines the deviation of the frequency response of the device from the. equation valid at lower currents. The gain compression coefficient is more than one order of magnitude higher than typical values (~10-17cm3) in InGaAs/InGaAlAs QW lasers and may be connected to the slower relaxation rate into the QDs at low temperatures. The fitted differential gain of 2xl0-12cm2, is close to the value derived from the QD parameters.
Fig 7. Temperature dependence of relaxation oscillation frequency. F3dB~ 10.2 GHz (210K)
The PL rise time of the QD luminescence is reduced from 40 to 13 ps by increasing the coupling strength between QDs by growing dots on top of each other with very thin barriers. This gives the possibility to suppress the limitation due to the carrier capture time into the QDs. From direct observations of relaxation oscillations, cut-off frequencies larger than 10 GHz have been determined (see Fig. 7).
Potential applications of a QD laser as directly modulated light source for data transmission via glass fibers requires minimum chirp. The chirp is determined from the shift of the wavelength a laser exhibits during current modulation, and is described by a parameter called factor. The physical origin of this shift is related to the fact that any absorption or gain peak causes modulation of the refractive index near the resonance energy in agreement with the Kramers-Kronig (KK) relation. The linewidth enhancement factor can be calculated from the gain spectrum via the KK relation. In the case of a QD laser with a dot ensemble showing a perfect Gaussian energy distribution and well- separated QD energy level for electrons and for holes, the gain spectrum is symmetric and the derivative of the gain spectrum, the modulation of the refractive index, and, thus, the a- factor should be close to zero around the peak gain energy. As opposite, highly asymmetric absorption and gain profiles in QWs cause the -factor to be about two.
8. NEW DIRECTIONS FOR QD LASERS
RESONANT WAVEGUIDING AND LASING. Excitons in QDs cannot be screened or heated, as opposite to excitons in QWs. Thus, effect can be used intentionally to induce resonant refractive index enhancement (reduction) on the low energy side of the QD absorption (gain). This effect can lead to entirely QD exciton-induced wave guiding and lasing. This approach is particularly attractive in materials where a high density of QDs can be realized and no suitable lattice-matched heterostructure with significantly different refractive indices for its components exists, such as in some II-VI and III-V systems, diamond, silicon, etc.
LIGHT-EMISSION IN SILICON. If it is possible to insert narrow-gap QDs (e.g. made from InAs) which have high probability of radiative recombination in such a way that electrons and holes we will be trapped in these QDs, then silicon-based devices with efficient radiative recombination may become possible.
INFRARED LASERS BASED ON QDS. In the QD case, the population time of the ground state and the depopulation time of excited, state coincide and are about 10â€40 ps. Thus, while the total relaxation time, important for high frequency operation of lasers, is comparable for QWs and QDs, the excited state depopulation time is much longer in QDs, This slow relaxation increases the relative importance of the competing carrier relaxation mechanism via emission of middle-infrared (MIR) photons.
EXTENSION OF THE SPECTRAL RANGE ON GAAS SUBSTRATES. Success in GaAs-based QD lasers for the 1 .3 m range, including high-power cw operation and VCSELs stimulates further attempts to shift the emission wavelength in GaAs-based structures towards the 1 .55m range.
9. CONCLUSION
Though quantum dot lasers show immense potential for superior device performances, there are still some significant problems associated with the control of emission wavelengths reproducibility of the dots, high-temperature reliability and long- term stability of the dots. The current challenge is to match and surpass the performance of quantum well lasers. There is still need for the development of a quantum dot structure lasing around 1.55 micrometer, which is a principal wavelength in fiber optic communications. This would give QD lasers a chance to move into applications such as ultrafast optical data transfer. A key aspect of quantum-dot production challenge will be to improve our control over the dot distribution produced in the self-assembly process. Reliable continuous wave room temperature operation of QD lasers has already been reported;structure improvements are required to get the operation characteristics more desirable, especially the elimination of the several mechanisms that have a detrimental effect at room temperature.
From a birdâ„¢s eye view, the research on QD lasers is still newly emerging from its beginning stages. Several prominent groups of researchers around the world are all going down their own avenues, grappling with a portion of the overall problem, identifying and overcoming obstacles one-by-one individually. This is not surprising, considering the research on QD lasers, as opposed to the some what more well established research on basic QDs themselves, began to hit the stage truly only around 1995-1996. Still, considering the flux of effort and the emergence of well-defined directions, there seems to be hope that the field will settle down and become established. If the collective effort succeeds in bettering the performance of quantum well lasers, which it might, then QD lasers can finally be up there along with the MOSFET, quantum well lasers, and monolithic integration technology.
REFERENCES
1. Novel Infrared Quantum Dot Lasers: Theory and Reality. phys. stat. sol. (b) 224, No. 3, 787â€796 (2001)
2. Quantum Dot Lasers: ENEE 697 Term Project byMadhumita Datta, Zeynep Dilli and Linda Wasiczko
3. Properties and Designs of Quantum Dot Laser: Maurice Cheung
4. Bringing quantum dots up to speed: IEEE Circuits and Devices January 2000.
5. A quantum leap in laser emission: IEEE Circuits and Devices May 1999.
URLs
1. http://howstuffworks.com
2. http://e1ectronicsforu.com
3. http ://spectrum . ieee.org
4. http://google.com
5. http :// us-epanorama.net
TABLE OF CONTENTS
1. INTRODUCTION 1
1.1 QUANTUM DOTS 1
2 QDLASER 3
2.1 PRINCIPLE OF OPERATJON: QUANTIZATJON OF
DENSITY OF STATES 4
2.2 EXPECTED ADVANTAGES 6
2.3 QUANTUM DOT REQUIREMENTS 7
2.4 BASIC STRUCTURE OF QD LASER 8
2.5 CAPTURE TIMES AND THE PHONON BOTTLENECK 10
2.6 BANDWIDTH LIMITS 11
3 QD LASER DESIGNS 12
3.1 1n0.67Ga0.33 AI/GaInAsP/InP TENSILE STRAINED QD LASER 12
3.2 SELF ORGANIZED In0.5 Ga0.5 As QD LASER 13
3.3 HIGH POWER CONTINUOUS - WAVE InGaAs/AlGaAs
QD LASER 14
3.4 InGaN QD LASER 14
3.5 VERTICAL CAVITY QD LASER DESIGN 15
4 FABRICATION OF DOTS 16
4.1 COHERENTLY STRAINED. SELF-ASSEMBLED
THREE DIMENSIONAL ISLANDS 16
4.2 ARRAYS OF VERTICALLY COUPLED QUANTUM DOTS 19
5 QUANTUM DOT VCSELS 20
6 TEMPERATURE INSENSITIVE QD LASER 23
6.1 TUNNELING INJECTION QD LASER STRUCTURE 25
7 DYNAMIC PROPERTIES 27
8 NEW DIRECTIONS QD LASER 29
9 CONCLUSION 30
10. REFERENCES 31
ACKNOWLEDGEMENT
I extend my sincere thanks to Prof. P.V.Abdul Hameed, Head of the Department, Electronics and Communication Engineering, for providing me his invaluable guidance for the Seminar.
I express my sincere gratitude to my Seminar Coordinator and Staff in Charge Mr. Manoj K, for his cooperation and guidance in the preparation and presentation of this seminars.
I also extend my sincere thanks to all the faculty members of Electronics and Communication Department for their support and encouragement.
SALEEL O.M.
