• 沒有找到結果。

Wavelength Switching in Quantum Dot Lasers

Chapter 1 Introduction

1.2 Wavelength Switching in Quantum Dot Lasers

Solid-state light emitting devices with the capability of electrically controlled wavelength switching may become important for read and write operations, chip-to-chip interconnects, mode-locked laser and wavelength division multiplexing. For InAs QD lasers, distinct peaks in the optical spectra reveal that both the ground state and the excited state transitions can participate in lasing action [24,25]. These distinctly separated levels, which are not encountered in quantum well lasers, are only observed in QD lasers. In chapter 4, dual state lasing (ground state~1.3 μm and excited state~1.2 μm) was observed in conventional single cavity ridge waveguide lasers. In addition, we also have demonstrated of electrically controlled wavelength switching in a couple-cavity QD laser.

1.3 QD growth on metamorphic structures

To this date, only InP/InGaAsP material system has been used for 1.3–1.55 μm telecommunication semiconductor lasers. The characteristics of such lasers are limited by several fundamental problems. One of the problems is the insufficient confinement of carriers in the active region [26,27] leading to a poor temperature stability and a high value of threshold current density. InAs/GaAs QDs laser at 1.3 μm with high temperature stability and low threshold current densities have been demonstrated. It is desirable to extent the wavelength of InAs/GaAs QDs lasers to 1.55 μm. Recently, reported papers have demonstrated InAs QDs laser can emit wavelengths at both 1.55 μm(ground state) [28,29] and 1.3 μm(first excited state)[30] using the structure of metamorphic buffers with GaAs substrates. It provides an opportunity for long wavelength, high performance lasers with GaAs substrates.

The metamorphic buffer layer allows the large lattice mismatch between GaAs substrates

and the device layers. The buffers serve as the transition region for the lattice parameter to change from that of GaAs to a larger value by growing relaxed, compositionally graded epilayers. In other words, the metamorphic buffers behave as a virtual substrate that can be designed to accommodate the lattice parameter of the topmost structure. Device performance is influenced by the structure, composition and strain relaxation rate of metamorphic buffer.

Most studies in InAs QDs grown on the metamorphic buffer focus on the relation of the emission wavelength and the metamorphic buffer layers. The growth behavior of InAs QDs on the metamorphic buffer is still under studies including the size, the shape and the position distribution of QDs. In this work, we present the growth behavior of InAs QDs including an interesting and unique finding discovered in this study.

1.4 Organization of Dissertation

In chapter 2, the experimental techniques in my research are presented. In Section 2.1 we introduce our Molecular Beam Expitaxy system and growth condition of sample. Section 2.2 describes the basic principle of the ion channeling technique. Section 2.3 describes the photoluminescence system.

In chapter 3, we study the strain in self-assembled InAs QDs buried in GaAs matrix using MeV C++ ion-channeling technique. Section 3.2 describes why InAs QDs was probed using heavy ions. Section 3.4 shows the angular scan curve. A little strain behavior was observed in this section. Section 3.5 presents how to determine the strain state along the growth direction. Section 3.6 shows the relationship between the strain state of InAs QDs and the optical property.

In chapter 4, we present wavelength switching transition of QD lasers. Section 4.2, section 4.3 and section 4.4 introduce the standard procedure of laser process and measurement system in this study. Section 4.5 presents the wavelength switching characteristic of InAs QD

lasers with single cavity. Various methods can achieve the ground to excited state lasing switching. Section 4.6 presented wavelength switching transition in couple cavity laser.

In chapter 5, we present InA QD growth on metamorphic hetero-structure. Section 5.2 explains that the indium composition profile affect the residual strain in the hetero-structure.

Section 5.4 shows the growth behavior of InAs QDs on metamorphic buffer by adjusting the indium composition profile in the buffer layer. The results are presented from images of the atomic force microscopy and transmission electron microscopy. Section 5.3 shows the growth behavior of QDs was affected by growth temperature of the buffer layer.

In chapter 6, we concluded our researches. The further work also was mentioned in this chapter.

Chapter 2

Experimental Techniques

In this chapter, the experimental techniques in my research were presented. The contents consist of three parts: (1) the molecular beam epitaxy (MBE) system in our laboratory (2) Ion channeling theory, application for material analysis and system setup. (3) Photoluminescence measurement system.

2.1 Molecular Beam Epitaxy System

2.1.1 Introduction of System

Semiconductor Hetero-structures are important materials system for both fundamental physics researches and device applications. The growth of a perfect hetero-structure need own an abrupt interface, uniform and exact compositions of the grown materials with high quality, and a precise doping distribution. In the past decades, many epitaxy techniques have been developed for obtaining high-quality epi-layer and ideal hetero-structure growth. They are liquid-phase epitaxy (LPE), vapor-phase epitaxy (VPE), metal-organic vapor-phase epitaxy (MOVPE), and molecular beam epitaxy (MBE). Among all these techniques, MBE system has been testified to be the most powerful and reliable system due to its ultra-high vacuum (UHV) environment and relatively simple growth mechanism. The high controllability during growth has made it more and more valuable for various applications, especially for advanced researches.

FIG.2.1 A schematic of our Gen II MBE system.

MBE system in our lab is Vecco Gen II molecular beam epitaxy system. All samples in this dissertation were grown in this system. Fig. 2.1 shows the scheme of the system. The MBE system consists of three chambers, the entry/exit (E/E) chamber, the buffer chamber, and the growth chamber. Gate valves are used to connect and isolate the buffer chamber and the other two chambers. Oil-free pumping systems were used in MBE system in order to not containment the growth environment. In the entry/exit chamber, a turbo pump station and a sorption pump are for roughly pumping when we complete the processes of loading and unloading. A cryo-pump was used to maintain the chamber in 10-9 order. The buffer chamber mainly function is to pre-bake samples and isolated the growth chamber from the water in air.

Therefore, the chamber generally main the 10-10 vacuum by a high efficiently ion-pump and an assistance TSP. The pumps in the growth chamber include an ion-pump, an cryo-pump and a TSP pump. Because the volume of growth chamber is bigger than other chambers, pumping rate of the pumps in the chamber are higher than in other chambers. The liquid nitrogen was injected into the cryo panel to assist the growth chamber with better vacuum when we grow the samples. Eight effusion cells are installed in growth chamber. Ground III elements were provided from two gallium dells, one aluminum cell and one indium cell. As2 and As4 were supplied from arsenic cracker cell and arsenic kundsen cell, respectively. silicon and berullium effusion cell offer as n and p type dopant sources in this system. In our MBE system, material systems that could be grown include AlGaAs, InGaAs on GaAs substrates or AlGaAs, InGaAs, and AlInGaAs on InP substrates.

Besides, in the growth chamber, two analysis instruments are installed. One is the residue gas analyzer (RGA), and another is the reflection high-energy electron diffraction (RHEED) monitor. The RGA is used to analyze the residual gas in the chamber and, thereby, to help us to understand the cleanness in the growth chamber. By setting its detection range of the atomic mass unit (AMU) to that one of helium (He, 4), it can also serve as a very sensitive leak detector.

2.1.2 Reflection High-Energy Electron Diffraction (RHEED)

The RHEED allow us to in-situ monitor the surface construction of wafers and epi-layers.

In RHEED analysis, a collimated monoenergetic electron beam is directed towards the surface at a grazing angle of about 2 degree and orthogonal to the molecular beam paths, the primary electron energy lies in the range 5~40KeV. Since the energy component perpendicular to the surface is of the order of ~100eV. The penetration depth of the incident electron beam is limited to only the first few atomic layers, as a result, the crystal surface acts as a two dimensional grating and diffracts the incident electron beam. A fluorescent screen placed diametrically opposite the electron gun recorded the diffraction pattern. Because the interaction (diffraction) of the electron beam is essentially with a two-dimensional atomic net for the crystal surface structure, we area concerned with its reciprocal lattice in the reciprocal space which is composed of rods that have a direction normal to the real surface. Fig. 2.2 shows the Ewald sphere and reciprocal lattice rods for a simples square net and the formation of diffraction lines on a plane screen, the surface condition of epitaxial layer can therefore be clearly indicated from the fine details of RHEED patterns.

FIG. 2.2. The schematic representation showing the interaction of the Ewald sphere with reciprocal lattice rods in RHEED analysis of a two-dimensional surface net.

By using RHEED, the construction of a few monolayers on the surface can be analyzed with the diffraction patterns. It makes desorption of native oxide on the surface easy and exact.

It can determine whether lattice mismatched hetero-epitaxy retained the layer by layer growth.

In general, the 2x4 or 2x8 patterns means the smooth surface during 2D layer growth and native oxide is removed. It is also useful for the growth of quantum dots (QDs), because the 2-D to 3-D transition in QDs formation on the surface could be observed easily from the RHEED pattern transition. Somehow, if the growth mode transformed from the two-dimensional layer by layer growth to the three dimensional island growths, which is the case for the self-assembled quantum dots, the RHEED patterns would change from streaky into spotty patterns. For hetero-structure growth, RHEED is useful tools.

2.1.3 Pre-growth Procedure and Growth Conditions

To maintain the cleanness of the growth chamber and to ensure the quality of grown samples, a standard treatment procedure for wafers before growth is important and necessary.

The epi-ready GaAs wafers are mounted on Mo substrate holders with indium or indium-free springs and then put into E/E chamber for bake-out at 200oC, which can remove the water molecules absorbed during the loading procedure. After about two hours bake-out and four hours pumping down, the pressure of the E/E chamber can be lower than 10-8 torr. After that, we transfer the wafers into the buffer chamber, in which the pressure is about 1-5x10-10 torr typically. In this chamber, each wafer is heated to 400oC for over one hour and the chamber pressure is lower than 3x10-9 torr, in order to remove residual water and organic species.

Finally, the wafer could be transferred into the growth chamber. Just before growth, the thin native oxide formed on the wafer surface is decomposed and desorpted by heating under enough arsenic flux. The desorption temperature are about 610oC for GaAs. It should be noted that, before the cells warmed up for sample growth, the growth chamber is cooled down with

the continuously flowed liquid nitrogen and then, typically, the pressure of the chamber lowers from about 3x10-9 torr to less than 2x10-10 torr.

The growth rate is typically around one µm/hr. The III-V BEP ratio (group V versus group III) and growth temperature must be kept within suitable range for different grown materials for high-quality epi-layers. In general, the growth temperature for AlAs, GaAs and AlxGa1-xAs materials on GaAs substrates are near 575oC for the all the samples studied in this dissertation. The III-V BEP ratio keeps around 10 and 20 for As2 and As4, respectively, for the growth on GaAs substrate. However, 3-D quantum dots were grown under various growth conditions for the specific propose. They were described in respective chapters for clarity. For InAlAs metamorphic buffer layer, the growth temperature is near 400 oC in order to obtain better strain relaxation layer. The growth conditions depended on the structure recipes. The III-V BEP ratio was more than 20 for As2 in growing metamorphic structure.

2.2 MeV Ion Channeling as Instrument for Materials analysis 2.2.1 Rutherford Backscattering Spectrum

The arrangement of atoms determines the properties of materials. Ion-channeling analysis is a powerful technique to probe the arrangement of atoms in the solid crystal.

Rutherford backscattering spectrum (RBS) can resolve the arrangement of lattices and provide angular scan curves, which reveal the deformation information of the epi-layer. In this subsection, we first describe RBS in detail.

In ion channeling technique, the sample is bombarded by using an ion beam operated in the MeV energy range. When a sample is bombarded with MeV ions, the energy of incoming ions transfers to atoms of the sample. The energy transfer depends on how close incoming ions approaches to the atomic nucleus. When the distance between the incoming ion and the atom is at the order of 1 A o, small amount of the energy transfer from incoming ions to the

valence electrons. The energy of incoming ions start losing gradually when the ions penetrate deeper into samples. At smaller distance than 1A°, incoming ions interact with inner electrons and have larger energy transfers. When the distance between the ion and the atom reaches minimum, i.e.the nuclear size, large energy transfers occur and result in the large angle scattering of the incident ion. The process with the large angle scattering of the incoming ion is called the close impact.

The energy of the incoming ion after large angle scattering event can be revealed by RBS.

This energy is determined by the mass and the depth of atoms when the scattering event takes place. An example shown in Fig. 2.3 describes the RBS of an epi-layer structure. A MeV ion beam bombards a sample consisted of a buried epi-layer and a substrate. The atomic mass of the epi-layer is higher than that of the substrate. The bottom portion of Fig. 2.3 shows the RBS of backscattered ions detected by an energy-scan detector. The x-axis represents the energy of backscattered ions. The y-axis represents the scattering yield, which is the number of the backscattered ions. The energy of backscattered ions from the epi-layer with larger atomic mass is higher with finite energy width (i.e., the shadow area in the figure). The broad and continuous spectrum at lower energy represents the energy of backscattered ions from the substrate. The energy distribution of backscattered ions from the epi-layer is separable from the energy of backscattered ions from the substrate when atomic mass is high. Therefore, RBS is suitable for the analysis of heavy atoms in a light substrate.

FIG. 2.3. Schematic of the energy spectrum of particles scattering from a solid composed of a substrate (M2,Z2) and the epilayer (M3,Z3). The left of this figure show the angular scan curves of the substrate and the epilayer.

In Fig. 2.3, the ion beam enters at an incident angle (ψ) respect to the crystal axis within 1°. Less than 2% of incoming ions are backscattered when the incident ion beam aligns with the crystal axis. Thus, the scattering yield in RBS is low as the spectrum with mark,

“ALIGNEN”, in the figure. The scattering yield increases when the incident angle of the ion beam increases. It reach maximum when the incident angle is 1° respect to the crystal axis (i.e.

the dash line with mark <RANDOM> in the RBS of Fig.2.3)

2.2.2 Angular Scan Curve

A so-called angular scan is the integral of the scattering yield within an interest energy interval as a function of the angle between the incident ion beam and the crystal axis. The right portion of Fig. 2.3 shows the relationship between the angular scan curves and RBS. The angular scan curves of the epi-layer and the substrate are obtained from the energy interval ΔE1 and ΔE2, respectively, as shown with RBS plot. The integral yield from ΔE1 and ΔE2 is minimum when the incident beam aligns with the crystal axis (angle=0°), and it increases with increasing incident angle.

2.2.3Tetragonal Deformation of Epi-layer in MeV Measurement

The tetragonal deformation of the coherent strained epi-layer can be measured from the angular scan curve of the ion-channeling analysis, specifically, by comparing the two minimum yields of the angular scan curves from the strained epi-layer and the substrate. Fig.

2.4 shows deformation of the epi-layer causes the different off-normal crystal axes between the epi-layer and the substrate. It results in the incident angle of the ion beam different relative to the crystal axis of the epi-layer and the substrate. Because the minimum of the angular scan curve is generated when the ion beam is aligned with the crystal axis, so the minimum yield position of the strained epi-layer is different from that of the substrate.

FIG. 2.4. The deformation of the epi-layer causes the different off-normal crystal axes between the epi-layer and the substrate.

Fig. 2.5 shows angular scans for Ge in the strained epi-layer (Si0.8Ge0.2) and Si in the substrate.

In the figure, it indicates that the crystal axis is different between the strained epi-layer and the substrate such that the minima of the strained layer and the substrate are different. The shift between these two curves is to measure the tetragonal distortion in the Si0.8Ge0.2

epi-layer. Thus ion-channeling is an excellent tool for analyzing lattice deformations in crystalline semiconductor structures.

FIG. 2.5. The measured Si (substrate) and SiGe (strained epi-layer) angular scans of the [111] axis. (after L. J. M. Selen. et al., 2003)

2.2.4 Ion Channeling Set-up

An ion channeling experiment requires three basic components: a particle accelerator, vacuum chamber, detector of scattered particles, and an accurate crystal manipulator, a goniometer. In channeling experiment, ion beams are extracted in the energy range of 2-6MeV from the particle accelerator. The ions are transported to the scattering chamber through a beam line. In order to preserve a low angular spread in the ion beam, the pressure in the scattering chamber and the beam line is below 3x10-8 and 3x10-5 mbar, respectively.

A high precision goniometer is needed to obtain proper alignment of the target crystal axes or planes with the ion beam. Such a goniometer was designed and constructed at the National Tsing Hua University, with three axes of rotation and three directions of translation. The angular resolution of the goniometer is better than 0.005°. The detector is mounted on a stepper-motor controlled detector disk to obtain a high flexibility in the detector position and

angle, for RBS and transmission channeling experiments. A schematic drawing of the set-up is shown in Fig. 2.6.

2.3 Optical Measurement

Photoluminescence (PL) is a non-destructive characterization technique for optical property of semiconductor. The emission wavelength, peak intensity, and full width at half maximum are the most important for characterizing the quality of epitaxial layers. Fig. 2.7 depicts the schematic of our PL measurement. The exciting light is an Argon laser. The intense spectral emissions around 488 nm and 515 nm are well above the bandgap of investigated materials, such as GaAs, InGaAs, InAs and AlGaAs. The samples are kept in the Helium cryostat which provides for the various temperature measurements. The excitation laser is focused by one lens and the output luminescence is collected by another lens into the input slit of the spectrometer. High-pass filter is put in the input of the spectrometer to block the laser and light with wavelength below 600nm.

The spectrometer is SPEX-1404, 0.85 m double grating monochrometer with ultimate

The spectrometer is SPEX-1404, 0.85 m double grating monochrometer with ultimate

相關文件