Characteristic of Semiconductor Nanostructure

Nanostructure is defined as the materials with size within 1-100 nm, and it exhibits unique properties because of the smaller radius than the bulk exciton Bohr radius. With the small particle size, the quantum confinement effect, surface effect and size effect make the nanostructure different from the bulk. As the size reducing to nano-scale, the area of 80% ~ 90% of the material atoms is exposed to the surface. So the chemical activity or quantum confinement effect is increasing apparently. The quantum confinement effect plays an important role in determining the optical and electronics properties of the semiconductor materials. At such a small size, excited electrons and holes are confined in all three dimensions (quantum dots), two dimensions (quantum wires), or one dimension (quantum well), as displayed in Figure 1.2. Quantum dot (QD) is the prototype of zero-dimensional (0-D) system and possesses many unique properties. Regarding the QD, there have been lots of researches devoted to realizing the predicted potential of 0-D quantum-confined structures, for instance, their unique atomic-like discrete states with a δ-function density of state as shown in Figure 1.3. And the widening of band-gap also appears. Investigations of semiconductor QD have been very extensive particularly in the last decade. These are fuelled by unique physical phenomena and potential device applications.

Recently, Group- nitrides lowⅢ -dimensional structures such as nanorod, quantum dot have attracted much attention [2]. QD structure could still improve their performance enormously. LDs with QDs structures in the active layer have been theoretically predicted to have superior characteristics, including lower threshold currents, enhanced modulation frequency, and narrowing spectral line width [3]. Moreover, because of the obvious advantage of enhancing radiative recombination through carrier localization, QDs structures have been expected to increase the efficiency of the luminescence of LEDs

1.4 Indium Aggregation and Quantum dot-like Structure in InGaN MQW Structure The large difference in interatomic spacing between GaN and InN was found to give rise to a solid phase miscibility gap as shown in Figure 1.4 [4]. And the low miscibility leads to indium aggregation and phase separation [5-7].

It was proposed that nanoscale indium composition fluctuations or In-rich clusters act as

(EDX) spectroscopy [10]. The Indium composition was probed at various positions as displayed in Figure 1.5. It is shown that indium composition in QD-like regions (area B) is higher than those in the adjacent well area. This is a direct evidence of In-rich clusters in InGaN MQWs structure.

As a result of carrier localization, many phenomena are different from those in conventional Ⅲ-Ⅴ semiconductor. For example, the performance of an InGaN-based device is less sensitive to the dislocation than those of other Ⅲ-Ⅴ semiconductors. These In-rich clusters act as the role of QDs to deeply localize the carriers and to obstruct their migration toward non-radiative defects such as dislocations, stacking faults…etc. Also, the temperature-dependent photoluminescence (PL) peaks energy exhibits an S-shape behavior.

Meanwhile, a large Stokes’ Shift of PL peak with respect to photoluminescence excitation (PLE) absorption peak was often observed in an InGaN/GaN QW structure. The behavior was usually attributed to the carrier localization effect [9, 11]. It was also pronounced the contribution of the piezoelectric field effect [12]. Both models of carrier localization and strain-induced piezoelectric field can provided certain explanations of the observed optical phenomena in such a QW structure. In my thesis, the results strongly suggest that the emission mechanism of InGaN/GaN MQWs originates from radiative recombination within the localized states of In-rich clusters.

1.5 GaN Quantum Dots structure

The growth of GaN QDs had not been investigated before 1996 owing to the fact that even growth of InGaN quantum wells of high quality had been difficult to be grown. GaN QDs were first formed on an AlGaN layer with low-pressure MOCVD [13] and gas-source MBE [14] by anti-surfactant method as shown in Figure 1.6. In anti-surfactant method, the surface was covered with silicon, by with the growth mode may be changed from 2-D to 3-D island grown by this anti-surfactant. After these reports, the S-K growth mode has been also investigated by MBE for GaN or InGaN QDs on AlN layer [15-16] as shown in Figure 1.7.

The formation of the QDs with the S-K mode is mainly by the 2.5 % lattice mismatch of the GaN-AlN system.

In a highly disordered system such as GaN exhibits a large number of dislocations (> 10⁸ cm^-2). The strong carrier localization in quantum dots (QDs) can reduce the dislocation-induced non-radiative channels. The carrier localization effect in QD plays the role of potential fluctuations and it has been proposed to explain the high quantum

efficiencies on InGaN/GaN LED [9]. Recently, GaN QDs have attracted a lot of attention for development of single electron transistors [17], ultraviolet sources [18] and detector [19].

1-6 Research Motivation

In recent years, there have been heated arguments concerning the emission mechanism in InGaN-based semiconductors [20, 21]. Understanding the emission mechanism of these materials in these materials is very crucial, not only from the viewpoint of physical interest, but also in designing practical device. Rectangular QW structure is generally used with constant Indium composition in InGaN for the well region. In this case, spatial indirect recombination is expected due to internal piezoelectric field, and thereby poor light emission is expected. However, the emission efficiency is in fact extraordinarily high. This high emission is generally believed to be due to In-content fluctuations or QD-like formation in the QWs. Indeed, we have observed well-defined In-rich QD-like in the InGaN/GaN MQW structure [10].

Many authors consider the In-rich QD-like regions as the origin of the high brightness emission in the InGaN-active region [3, 8-9]. A great deal of studies have been considered using different conditions, such as the doping [22], thickness [23], and growth temperature [24] of barrier layers, well thickness [23], and growth interruption between the growth of well and barrier layers [25] as well as the insertion of a graded InGaN barriers [26].

But the effect of δ-TMIn flow rate in the well layers on the properties of the InGaN/GaN MQWs grown by MOPVE is still not reported. Chapter3 will be the key issues of InGaN QD-like structure, including a) their optical characteristics and photon emission mechanisms and b) the microstructure formed by different growth condition.

To obtain GaN QDs, the most straightforward techniques are the S-K growth mode and anti-surfactant method. Several methods have been used to define the nucleation sites in the growth plane including electron-beam lithography [27], scanning tunneling microscope-assisted nanolithography [28], and optical lithography [2]. The growth of patterned In(Ga)As QDs has been accomplished using MOPVE [29]. Recently, GaN QDs on self-assembled nanoholes have been reported [30]. However, there are no detail optical studies of GaN QDs grown by this method. Chapter 4 will be the optical properties of GaN QDs grown by in-situ etching of AlN nanoholes and regrowth of pyramid shape GaN QDs.

Figure1.1 Lattice constant and energy band-gap of GaN-based materials and related materials.

Figure 1.2 Electrons and holes are in (a) bulk, (b) quantum well, (c) quantum wire, and (d) quantum dot.

Figure 1.3 Density of states in one band of semiconductor as a function of dimension.

Figure 1.4 Solubility of GaN in InN (InN:Ga) and InN in GaN (GaN:In) [4].

Figure 1.5 (a) Cross-section HRTEM image of MQW. White arrows indicate the indium rich regions. EDX spectra obtained form (b) area A (solid line) and (c) area B (dashed line) shown in Fig 1.4 (a) [10].

Figure 1.6 (a) The nanoholes were formed as a result of anti-surfactant. (b) The TEM image of GaN QDs grown with anti-surfactant mode [30].

Figure 1.7 (a) The Stranski-Krastanow (S-K) growth mode. (b) The TEM image of GaN QDs grown with S-K mode.

Chapter 2 Experimental Instrument Setup

2.1 Photoluminescence

Photoluminescence (PL) is the spontaneous emission of light from a material under optical excitation. When light with sufficient energy is incident on a specimen, photons are absorbed and electronic excitations are created. Eventually, these excitations relax and the electrons return to the ground state. If radiative relaxation occurs, the emitted light is called PL. PL analysis is a simple, versatile and nondestructive technique to exam the optical characteristic of optical semiconductor. Feature of the emission spectrum can be used to determine electronic energy level, such as identify surface, interface, and impurity levels and to gauge alloy disorder and interface roughness. The intensity of the PL signal provides information on the quality of surfaces and interfaces. Variation of the PL intensity with external parameters like temperature and applied voltage can be used to characterize further the underlying electronic states and bands.

2.1.1 Principle of Photoluminescence

Photoluminescence properties of semiconductors can be classed as following:

(a) Radiative transition: Luminescence may involve radiation electronic transition emitting a photon when an electron drops from upper to a lower level of either intrinsic band states or impurities levels. Several different types of raidative transitions are described and shown in Figure 2.1.

(i) Free-to-bound transition:

As shown in Figure 2.1(c)(d), transitions between intrinsic band and impurity state, the so called free-to-bound transition, which may occur between deep impurity and one of the bands (i.e. C-band to acceptor or donor to V-band) with momentum conservation even in indirect band-gap materials. For nonzero temperature, the impurities are partially occupied so that some impurity centers are neutral while others are ionized. If the impurity is a donor, then these two transitions would be (a) electron to ionized donor transition (e^-D⁺), and (b) hole to neutral donor transition (h-D).

Transitions of type (a) occur in the far infrared spectral range. Because of the small energy involved, phonon emission offers every effective competition and radiative efficiency is quire low. Transitions of type (b) occur close to the fundamental band-gap energy and have been

(ii) Band-to-band transitions:

As shown in Figure 2.1 (a), band-to-band transitions involving free electrons and holes.

Such transitions usually occur in the direct band-gap materials, such as Ⅲ-Ⅴ compounds, between C- and V-bands with conservation of momentum. The e-h pairs will recombine radiatively with a high probability.

(iii) Donor-acceptor pairs (DAP) recombination:

Transition between donor (activators) and acceptor (co-activators) levels is shown in Figure 2.1 (e). By optical excitation, electrons and holes can be treated at the D⁺ and A^- sites to produce D⁰ and A⁰ centers. In returning to equilibrium, some of the electrons on the neutral donors will recombine radiatively with holes on the neutral acceptors. It can be represented by the reaction,

0 0

D +A →hν +D⁺+A⁻ (2-1) The energy EDA of a photon emitted from such a transition would be

2

where ED and EA are the binding energies of donor and acceptor respectively. Q is the charge, ε the dielectric constant of the material, RDA the donor-acceptor separation. Larger values of RDA broaden the emission spectrum with less probability of radiative tunneling.

(iv) Free exciton Transition

As shown in Figure 2.1 (b), free exciton (FE) represents the lowest energy intrinsic excitation of electrons and holes in pure materials at low excitation density. In most semiconductors, the FE state is adequately described by Wannier-Mott approximation where the carriers are treated as nearly independence, oppositely charged particles, interacting through the Coulomb fields. The energy of FE is given by

2 * 4

where m^* is the reduced mass, n the quantum number, ε the dielectric constant. The FE results in a lowering of the total energy of the e-h pair as

g n

hν =E −E (2-4) (b) Non-radiative transitions

Several possible mechanism leading to non-radiative transitions, competing with the radiative ones, and adversely affecting the luminescence efficiency, can be described as:

(i) Generation of phonons due to thermal vibrations.

(ii) Recombination at surface states, dislocations, grain boundaries, pores etc., by losing the excess energy through step-wise transitions, so called cascade-process.

(iii) All the defect sites may not act as recombination centers to allow the carriers to recombine radiatively.

(iv) Auger process, in which the energy lost by the captured carrier excites another nearby carrier in the crystal and may give rise to non-radiative loss of energy. The other carrier can return to a lower energy state by multiple phonon emission.

2.1.2 Setup of Photoluminescence Measurement System

Photoluminescence measurement system is shown in Figure 2.2. The pumping source was a non polarized and multi-mode Melles Griot Helium-Cadmium laser operated at 325 nm with 25 mW. After reflecting by three mirrors, the laser was focused by a lens, which focal length was 5 cm with 300 m in diameter and the luminescence light was collected by the same 　 focused lens. The luminescence light was dispersed by 0.32 m monochromator (Jobin-Yvon Triax-320) equipped with 300, 1200 and 1800 grooves/cm grating. And the maximum width of the entrance slit was 1 mm.

300 grooves/cm grating and the slit of 0.1 mm was performed in the experiment. Under this condition, the wavelength resolution was approximately 1 nm. In order to prevent the spray laser light from the sample surface entered the detector, a long pass filter with a cut-off wavelength at 360 nm in front of the entrance slit was used. Finally, the collected luminescence light was detected by the charge coupled device (CCD).

2.1.3 Setup of Micro-Photoluminescence Measurement System

The setup of micro-photoluminescence (　-PL) was combined with Scanning Near Field Optical Microscopy (Alpha SNOM) manufactured by WITec. The pumping source was a Helium-Cadmium laser operated at 325 nm with 21 mW. The setup of 　-PL is displayed in Figure 2.3. The laser beam was focused by a 15X objective lens (N. A. = 0.32) and the focal size of the laser beam was 2 m under these conditions. Such a confocal optical system 　 enables the highly spatial resolution beyond the diffraction limit of a light wave. Therefore, the lateral spatial resolution of this system was as small as 2 m. Th　 e PL light was collected by the same objective lens through a multimode fiber (core = 50 m) and detected by charge 　 coupled device (CCD) detector through a 0.32 m monochromator (Jobin-Yvon Triax-320)

In order to perform the low temperature measurement, the sample was mounted in the micro-miniature refrigerator of liquid N2 flow cryogenic system that can cool down to 80K.

2.2 Photoluminescence Excitation

2.2.1 Principle of Photoluminescence Excitation

In PL measurement, which is performed at fixed excitation energy, the luminescence properties are generally investigated. While photoluminescence excitation spectroscopy (PLE), which is carried out at fixed detection energy, provides mainly information about the absorption properties. Apart from absorption and PL experiments, photoluminescence excitation (PLE) measurement is a widely used spectroscopic tool for the characterization of optical transitions in semiconductors.

It is also very important to note that the PLE also depends strongly on the different carrier relaxation processes that connect the absorbing state to the luminescent state. For example, it is possible to recognize the absorption in a quantum well (QW) from that of the substrate if they have different emission energies. In which case, it can be assumed that carrier transfer between substrate and QW is negligible. Nevertheless, in many cases it is difficult to separate the influence of relaxation from that of absorption. The PLE spectrum is strongly influenced by the relaxation depending on different samples.

2.2.2 Setup of Photoluminescence Excitation Measurement System

Except for the excited source, the light collection system and spectrometer (Jobin-Yvon Triax-320) are the same as the PL system. As shown in Figure 2.4, the sample was mounted in the low temperature vacuum chamber and cooled down to 10 K. The pumping source was Xe lamp with 450 W dispersed by double-grating monochromator (Jobin-Yvon Gemini 180) and then focused to the sample by two lenses. And then the PLE signals were collected by PL system. The detector was a high sensitive Hamamatsu photomultiplier tube (PMT) with GaAs photocathode.

2.3 X-ray Diffraction Apparatus

High resolution X-ray diffraction (HRXRD) is a standard method for the non-destructive structural analysis of semiconductor thin films and hetero-structures. The general principle of XRD is that X-ray intensity measured as a function of angle. Using Bragg’s law, the angle a peak appears is related to the separation of atoms (also known as the lattice constant). From this lattice constant, and by knowing the lattice constant in difference directions in the crystal

and in which types of scan they appear, other parameters about the material can be calculated (such as layer composition, relaxation, texture, strain, tilt angles etc).

Bede D1 triple-axis diffractometer with a parabolic graded multiplayer Gutman mirror collimator, followed by a fourbounce channel-cut Si (220) monochromator, delivering a CuKα

line of wavelength λ=0.154054 nm. The asymmetry two-bounce Si (220) analyzer crystal was place in front of detector. The equipment equipped with one four circle diffractometer. We can control the θ, 2θ, Chi and Phi angles of diffractometer to move the diffractive plane that we want to measure. The apparatus of XRD is shown in Figure 2.5.

In my thesis, the x-ray diffraction measurement was performed to determine the crystal orientation of the samples and to analyze the sample structure in the InGaN/GaN hetero-structures. The regular arrays of atoms in a crystalline semiconductor form a three-dimensional diffraction grating for waves with a wavelength around that of the distance between atoms as illustrated in Figure 2.6. When waves enter a crystal, they are scattered in all direction by atoms. In certain directions, these scattered waves can interfere destructively in some direction producing zero or weak intensity in those directions. In other directions, constructive interfere can occur, producing a strong maximum in the scattered wave intensity.

This scattering and interference is generally known as crystal diffraction. The diffraction pattern results in a map of the reciprocal lattice of the crystal and can be used to determine the structure of the crystal. This model of the diffraction process then leads to the Bragg diffraction law

2d_hklsinθ_B =nλ (2-5) where dhkl is the reciprocal lattice spacing and can be expressed by

2 2 2

hkl

d a

h k l

= + + (2-6)

where a is the crystal lattice constant; (h, k, l) is known as the Miller indices of the plane.

Besides, θB is the incident angle, n is an integer representing the diffraction order and λ is the wavelength of the incident radiation. The scheme diagram is shown in Figure 2.6 (a).

To simulate the x-ray diffraction pattern, assume that the crystal structure is a perfect structure without defect. The x-ray scattering from each mono-layer is shown in Figure 2.6 (b).

The wavefunction of incident beam can be expressed as Ee^ikxj. The wavefunction of total reflected beam can be written by the form

total scattered x-ray intensity is

where * represents the complex conjugate. Utilizing equation we can obtain the theoretical diffraction curve by choosing the lattice constant of each monolayer and thickness of quantum well, according to the growth condition. By fitting the theoretical diffraction curve to the experimental one, the structure of the sample can be determined.

2.4 Cross-Section Transmission Electron Microscopy Sample Preparation

Due to the strong interaction between electrons and matter, the specimens have to be rather thin (<<1000 nm) for Transmission electron microscopy (TEM) investigation. Thus, bulk materials have to be thinned enough to make the electron transparent.

Cross-section TEM specimens were prepared using a Tripod polisher with diamond abrasive films. This method is very powerful for the characterization of interface roughness, composition segregation…etc. First of all, the samples were constructed with Si and were grinded mechanically with both two faces. The diamond abrasive films with several particle sizes (30, 15, 6, 3 and 1 m) were utilized on the wet plate. After grinding both faces, the 　 thickness of film was smaller than 1 m. Then, a copper grid with a 2 mm X 2 mm hole was 　 epoxied onto the mirror surface of the mechanically polished specimen. The use of the copper grid prevents from generating cracks in sapphire substrates and it plays the role of supporting layer compared to specimen without copper grid.

For ion milling process, a GATAN 691 Precision Ion Polishing System (PIPS) is used as displayed in Figure 2.7. In this machine, two focused Ar ion beams mill the dimple-ground sample in such a way that a hole results at the desired position. In general, the parameters for

For ion milling process, a GATAN 691 Precision Ion Polishing System (PIPS) is used as displayed in Figure 2.7. In this machine, two focused Ar ion beams mill the dimple-ground sample in such a way that a hole results at the desired position. In general, the parameters for