Semiconductor quantum dots are promising candidates as bright, indistinguishable, single-photon sources---making them desirable for applications in quantum computing and quantum cryptography protocols. By embedding the quantum dots in III-V nanowires, the collection efficiency from the quantum dot is greatly increased. Our goal is to develop a platform that allows for the stable and efficient generation of single-photons on chip. This on-chip design offers an enhanced degree of stability and miniaturization, important in many applications involving the processing of quantum information. In this thesis, we demonstrate the efficient coupling of quantum light generated in a III-V photonic nanowire to a silicon-based photonic integrated circuit. We use high quality SiN waveguide devices fabricated by a foundry (LIGENTEC) to minimize coupling and propagation losses through the waveguide. A hybrid integration of these single-photon sources with a photonic integrated circuit is developed by employing a "pick & place" method which uses a nanomanipulator in a scanning electron microscope setup. By tailoring the nanowire geometry, we are able to maximize the efficient coupling between the optical mode of the photonic nanowire and an accompanying SiN waveguide through evanescent coupling. To determine the effectiveness of our integration method, we compare our hybrid devices with free-standing nanowires on their growth substrate. For each set, we measured the optical properties (brightness, spectral purity, lifetime, and single-photon purity) and efficiencies of the devices. We have shown that using tapered nanowires with embedded quantum dots coupled to on-chip photonic structures is a viable route for the fabrication of stable, high-efficiency, single-photon sources. Although the measured collection efficiencies from device to device were substantially different 9.6%~93%, we have found that the optical properties of the hybrid devices were hardly impacted from the transfer process. In fact, from the same nanowire that achieved 93% coupling efficiency, we were able to measure a single photon purity of 97%. By comparing the amount of emitted light collected from both ends of the nanowire (taper and base), we confirmed that the coupling efficiency of the devices have a strong dependence on the geometry of the nanowire as collection from the taper yielded count rates at least 10x greater than from the base. From our promising results, we can envision integrating the nanowire devices with different types of photonic structures such as ring resonators.

This thesis presents advancements on the modeling of quantum dots using Fourier-space k.p theory and on the use of InGaN quantum dots for ratchet band solar cells. Fourier-space based methods have generally assumed sharp material interfaces for electronic structure, strain and piezoelectric potential calculations in quantum dot systems. Additionally, standard Fourier-space methods have often assumed uniform elastic and dielectric constants for the strain and piezoelectric potential calculations. We present generalized methods to include smoothly varying alloy profiles for the quantum dots, including spatially varying elastic and dielectric constants for the strain and piezoelectric potential calculations. For the case of InGaN/GaN quantum dots, we show that the elastic and dielectric constants corrections are important for accurate strain, piezoelectric potentials, and electronic structure. The smooth alloy profiles are constructed by convolving sharp alloy profiles with a Gaussian, and we show that the electronic structure strongly depends on the smoothing kernel, indicating the need for precise alloy profiles for accurate electronic structures. We also present a new method that facilitates the coupling of strain into the k.p Hamiltonian when considering isolated dots, greatly reducing the computational costs of calculating the Hamiltonian matrix elements. Using the methods, we investigate the use of InGaN/GaN quantum dot superlattices as ratchet band solar cells, where we propose to use the piezoelectric potential to generate a ratchet. The piezoelectric potential can spatially separate confined electron and hole states, creating a spatial ratchet in order to reduce recombination. From our quantum dot k.p model, we calculate optical light absorption cross sections and present an improved method to calculate bound-to-continuum absorption, where electrons are excited out of the dots. In this method, we approximate the continuum states as bulk k.p states for GaN. By coupling our k.p model absorptions into a detailed balance model, we predict power conversion efficiencies. We find that a large number of QD layers is necessary to achieve sufficiently strong absorption as to reach high efficiencies, highlighting one of the key issues of QD-based solar cells. We consider systems which consists of up to 131000 layers of quantum dots and an ideal Lambertian back reflector with two different QD geometries. The first geometry possesses a strong spatial ratchet and can reach a maximum efficiency of 36% under a 1-sun 6000 K black-body spectrum. We verify the existence of the spatial ratchet through the optical properties of the system, showing that it truly has the potential to block recombination. We also present an efficiency optimized system that reaches a 42% detailed balance efficiency but does not have a spatial ratchet.

In this dissertation, a theory of modulation response of a semiconductor quantum dot (QD) laser is developed. The effect of the following factors on the modulation bandwidth of a QD laser is studied and the following results are obtained:<br /><br />1) Carrier capture delay from the optical confinement layer into QDs<br /><br />Closed-form analytical expressions are obtained for the modulation bandwidth omega_{-3 dB} of a QD laser in the limiting cases of fast and slow capture into QDs. omega_{-3 dB} is highest in the case of instantaneous capture into QDs, when the cross-section of carrier capture into a QD sigma_n = infinity. With reducing sigma_n, omega_{-3 dB} decreases and becomes zero at a certain non-vanishing sigma_n^{min}. This sigma_n^{min} presents the minimum tolerable capture cross-section for the lasing to occur at a given dc component j_0 of the injection current density. The higher is j_0, the smaller is sigma_n^{min} and hence the direct modulation of the output power is possible at a slower capture. The use of multiple layers with QDs is shown to considerably improve the modulation response of the laser -- the same omega_{-3 dB} is obtained in a multi-layer structure at a much lower j_0 than in a single-layer structure.<br /><br />2) Internal optical loss in the optical confinement layer<br /><br />The internal optical loss, which increases with free-carrier density in the waveguide region, considerably reduces the modulation bandwidth omega_{-3 dB} of a QD laser. With internal loss cross-section sigma_int increasing and approaching its maximum tolerable value, the modulation bandwidth decreases and becomes zero. There exists the optimum cavity length, at which omega_{-3 dB} is highest; the larger is sigma_int, the longer is the optimum cavity.<br /> <br />3) Excited states in QDs<br /><br />Direct and indirect (excited-state-mediated) mechanisms of capture of carriers from the waveguide region into the lasing ground state in QDs are considered, and the modulation response of a laser is calculated. It is shown that, when only indirect capture is involved, the excited-to-ground-state relaxation delay strongly limits the ground-state modulation bandwidth of the laser -- at the longest tolerable relaxation time, the bandwidth becomes zero. When direct capture is also involved, the effect of excited-to-ground-state relaxation is less significant and the modulation bandwidth is considerably higher.<br /> / Ph. D.

This work develops a comprehensive theoretical model for a semiconductor laser, which exploits tunneling-injection of electrons and holes into quantum dots (QDs) from two separate quantum wells (QWs). The potential of such a tunneling-injection QD laser for temperature-stable and high-power operation is studied under the realistic conditions of out-tunneling leakage of carriers from QDs (and hence parasitic recombination outside QDs) and the presence of the wetting layer (WL). The following topics are included in the dissertation: 1) Characteristic temperature of a tunneling-injection QD laser The threshold current density jth and the characteristic temperature T0 are mainly controlled by the recombination in the QWs. Even in the presence of out-tunneling from QDs and recombination outside QDs, the tunneling-injection laser shows the potential for significant improvement of temperature stability of jth — the characteristic temperature T0 remains very high (above 300 K at room temperature) and not significantly affected by the QD size fluctuations. 2) Output power of a tunneling-injection QD laser Closed-form expressions for the light-current characteristic (LCC) and carrier population across the layered structure are derived. Even in the presence of out-tunneling leakage from QDs, the intensity of parasitic recombination outside QDs is shown to remain restricted with increasing injection current. As a consequence, the LCC of a tunneling-injection QD laser exhibits a remarkable feature — it becomes increasingly linear, and the slope efficiency grows closer to unity at high injection currents. The linearity is due to the fact that the current paths connecting the opposite sides of the structure lie entirely within QDs — in view of the three-dimensional confinement in QDs, the out-tunneling fluxes of carriers from dots are limited. 3) Effect of the WL on the output power of a tunneling-injection QD laser In the Stranski-Krastanow self-assembling growth mode, a two-dimensional WL is initially grown followed by the formation of QDs. Due to thermal escape of carriers from QDs, there will be bipolar population and hence electron-hole recombination in the WL, even in a tunneling-injection structure. Since the opposite sides of a tunneling-injection structure are only connected by the current paths through QDs, and the WL is located in the n-side of the structure, the only source of holes for the WL is provided by QDs. It is shown that, due to the zero-dimensional nature of QDs, the rate of the hole supply to the WL remains limited with increasing injection current. For this reason, as in the other parts of the structure outside QDs (QWs and optical confinement layer), the parasitic electron-hole recombination remains restricted in the WL. As a result, even in the presence of the WL, the LCC of a tunneling-injection QD laser becomes increasingly linear at high injection currents, which is a further demonstration of the potential of such a laser for high-power operation. / Ph. D.

The effect of different factors on the operating characteristics of a semiconductor quantum dot (QD) laser is studied. Specifically, the following topics are included in the dissertation: 1) Effect of carrier-density-dependent internal loss in the optical confinement layer (OCL) on the characteristic temperature. Internal optical loss in a QD laser couples the confined-carrier level occupancy in QDs to the free-carrier density in the OCL. Due to this coupling, which is controlled by the threshold condition, the free-carrier density is increased and more temperature-sensitive, and also the confined-carrier level occupancy becomes temperature-dependent. As a result, the characteristic temperature of a laser is considerably reduced. Carrier-density-dependent internal loss also sets an upper limit for operating temperatures of a QD laser and constrains the shallowest potential well depth and the smallest tolerable size of a QD at which the lasing can be attained. The dependences of the characteristic temperature, maximum operating temperature, and shallowest potential well depth on the parameters of the structure are obtained. At the maximum operating temperature or when any parameter of the structure is equal to its critical tolerable value, the characteristic temperature reduces to zero. 2) Effect of excited-states in QDs on the light-current characteristic (LCC). The carrier capture from the three-dimensional reservoir (optical confinement layer – OCL) into the QD ground-state and escape from the ground-state to the OCL are assumed to occur via the QD excited-state. Such a two-step capture places a fundamental limitation on ground-state lasing—the output power saturates at high injection currents. The saturation power is controlled by the transition time between the excited- and ground-state in a QD. The longest, cut-off transition time exists, beyond which no ground-state lasing is possible. The following characteristics are analyzed versus the injection current density and the transition time: occupancies of the ground- and excited-state, free carrier density in the OCL, threshold current density, number of stimulated photons emitted, output power, internal and external differential quantum efficiencies. 3) Effect of longitudinal spatial hole burning (SHB) and multimode lasing on the LCC. The number of modes is shown to remain limited with increasing injection current. The maximum number of modes that can oscillate in a QD laser is analytically estimated. While this number increases with increasing surface density of QDs or cavity length, it remains limited (first increases and then decreases) with increasing scatter in the QD-size. The critical tolerable values of the structure parameters are derived beyond which higher-order longitudinal modes can not oscillate. It is notable that, in addition to the maximum tolerable scatter, there also exists the minimum scatter in the QD-size for each higher-order mode to start lasing. The threshold currents and output powers of modes are computed numerically. The power of the main mode is reduced due to lasing of higher-order modes and spatially nonuniform carrier distribution. As a new mode turns on, kinks appear in the LCCs of existing modes. SHB reduces the total optical power of a laser and contributes to nonlinearity of the overall LCC. The effect is more significant when any of the structure parameters is close to its critical tolerable value. The LCC becomes more linear with improving QD-size uniformity or increasing surface density of QDs or cavity length. / Ph. D.

In this thesis, our primary objective was to theoretically analyze the real world modulation bandwidth of a DTI QD laser and this was done by analyzing the effect of out-tunneling leakage of carriers from QDs, and by analyzing the effect of electron-hole asymmetry on the device characteristics. We are confronted with the following results: 1) Effect of Out-Tunneling Leakage on Modulation Bandwidth in Double Tunneling Injection Quantum Dot Lasers To purely focus on this effect, the conditions of instantaneous carrier exchange between the OCL and QW (on each side of the structure) and tunneling injection into QDs are assumed and closed-form analytical expressions for modulation bandwidth are obtained. The relative decrease in modulation bandwidth, due to this effect, in a DTI QD laser (from plots of modulation bandwidth vs j on increasing wout) is then shown to be small, and at ranges of injection currents of operational interest, nearly negligible. Consequently, it is shown that the DTI laser is a robust device in terms of sensitivity to out-tunneling leakage i.e. much effort need not be paid in suppressing this phenomenon. 2) Effect of Electron-Hole Asymmetry on Modulation Bandwidth of Double Tunneling Injection Quantum Dot Lasers On analyzing the effect of electron-hole asymmetry on the device characteristics of a DTI QD laser, it can be noted (from plots of modulation bandwidth vs injection current) that there is no reduction in the maximum modulation bandwidth i.e. electron-hole asymmetry does not indicate a reduction in the effectiveness of such a DTI design. This is shown to occur as the maximum modulation bandwidth depends on both, the effective differential non-stimulated recombination time as well the photon lifetime in the optical cavity. The photon lifetime being much smaller than the former acts as the dominating factor, and hence we see no appreciable change in the maximum modulation bandwidth. In the course of this analysis, we also see that the actual condition i.e. that of electron hole asymmetry is closer, among the cases of symmetry, to symmetry assuming hole parameters rather than electron parameters. As such, in cases where electron-hole symmetry must be used (in order to facilitate numerical simplifications), a recommendation of this study is to use hole parameters instead. / Master of Science / In this age of internet and optical communications, semiconductor lasers have a profound impact on the way we interact with our world. They act as intermediaries converting digital signals into optical pulses (in order to be transmitted) and then back into digital code. Understandably, the maximum speed at which these lasers can encode and decode information limits the speed of this entire communication network. This speed can be defined as the modulation bandwidth. A new design, the double tunneling-injection (DTI) quantum dot (QD) laser shows considerable promise, however its modulation bandwidth under real world operating conditions was yet to be analyzed. The aim of this thesis was to then theoretically analyze the real world modulation bandwidth of this new semiconductor laser design. This was done by analyzing the effect of unwanted leakage of carriers (out-tunneling) from the active region (Quantum Dots), and by analyzing the effect of electron-hole asymmetry on the device characteristics. The relative decrease in modulation bandwidth, due to leakage of carriers, in a DTI QD laser is then shown to be nearly negligible. Consequently, it is shown that the DTI QD laser is a robust device in terms of sensitivity to out-tunneling leakage, i.e., much effort need not be paid in suppressing this phenomenon. On analyzing the effect of electron-hole asymmetry on the device characteristics of a DTI QD laser, it is shown that there is no reduction in the maximum modulation bandwidth, i.e., electron-hole asymmetry does not indicate a reduction in the effectiveness of such a design. Thus, this analysis reiterates the fact that DTI QD lasers indeed show incredible potential to drastically improve modulation bandwidth and must be investigated further.

Lab-on-chip systems have been developed for various applications like point of care diagnostics and compact imaging systems. Compact, on-chip imaging systems face a challenge in the integration of multicolor light sources on-chip. This is because of the unavailability of compact, individually addressable, multicolor light sources on a single planar substrate. Colloidal Quantum Dot based Light Emitting Diodes (QDLEDs), which have found wide appeal, due to their unique properties like their tunable and narrow emission bandwidth and easy fabrication, are ideal for lab-on-chip integration. Among different types of QDLED structures implemented, inorganic QDLEDs have shown great promise. We have demonstrated designs and fabrication strategies for creating QDLEDs with enhanced performance. In particular: (I) We introduce a sandwich structure with a spin coated inorganic hole transporting layer of nickel oxide underlying the QD layer and with a spin coated zinc oxide electron transporting layer, with patterning of anode and cathode on the substrate. Compared to the use of sputtered thin films, solution processed charge transporting layers (CTLs) improve robustness of the device, as crystalline ZnO shows low CB and VB edge energy levels, efficiently suppressing hole leakage current resulting in LEDs with longer lifetimes. We also use Atomic Layer Deposition to deposit an additional hole injecting layer to protect the QDs from direct contact with the anode. With this device design, we demonstrate a working lifetime of more than 12 hours and a shelf-life of more than 240 days for the devices. Our solution based process is applicable to micro-contact printed and also spin-coated QD films. QDLEDs with spin-coated CTLs show a lifetime increase of more than three orders of magnitude compared to devices made using sputtered CTLs. (II) We implement strategies of the enhancement of light extraction from the fabricated QDLEDs. We discuss the integration of a two dimensional grating structure based on a metal-dielectric-metal plasmonic waveguide with the metal electrode of a QDLED, with the aim of enhancing the light intensity by resonant suppression of transmitted light. The grating structure reflects the light coupled with the metal electrode in the QDLED and we found an increase of 34.72% in the electroluminescence intensity from the area of the pattern and an increase of 32.63% from photoluminescence of QDs deposited on a metal surface. (III) We demonstrate the capability of our fabricated devices as a light source by measuring intensity across stained cells with QDLEDs of two different wavelengths and show the correlation as expected with the absorption profile of the fluorescent dye. We measure the absorption from the biological samples using QDLEDs fabricated with various design modifications, as a quantification of the improvements in device performance, directly affecting to our target application. / text

Progress in the burgeoning field of organic electronics is enabling the development of novel technologies such as low-cost, printable solar cells and flexible, high-resolution displays. One exciting avenue of research in this field is nanostructured hybrid organics such as quantum dot (QD)-polymer devices. The incorporation of QDs can greatly improve a device’s efficiency and gives one the ability to tune its electrical and optical characteristics. In order for such technologies to be commercially viable, it is important to classify their mechanical integrity and reliability. Surprisingly little is known about the mechanical properties of QD-polymer thin films (<100 nm). This is in part due to challenges of: (1) isolating the mechanical response of a thin film from the underlying substrate, (2) obtaining a homogeneous dispersion of QDs in the film, and (3) the sensitivity of mechanical properties to the inherent rate dependence of polymer deformation (i.e., viscoelasticity). All of these challenges can introduce significant errors in the measurement of mechanical properties. Furthermore, the deformation mechanisms in nanocomposites are not well understood, so it is difficult to predict the effect of adding QDs on the mechanical behavior of films. In this thesis, these challenges are addressed for characterizing the mechanical properties of thin films of CdSe QD-poly[2-methoxy-5-2(2΄-ethylhexyloxy-p-phenylenevinylene)] (MEH-PPV) nanocomposites using quasi-static nanoindentation testing. Elastic modulus, hardness, and creep are measured as a function of QD concentration and loading and unloading rates. The QDs' ligands are removed by pyridine treatment prior to mixing with MEH-PPV to improve dispersion. The films are prepared via spin-coating onto glass substrates and subsequent annealing in air. Efforts are taken in the mechanical testing to minimize errors due to viscoelastic creep and interference from the substrate. Transmission electron microscopy reveals that the QDs are relatively well-dispersed in the polymer matrix. It is observed that adding QDs increases the elastic modulus (E) and hardness (H) of the films, while reducing the viscoelastic creep. Both E and H increase linearly with the volume percent of QDs. E ranges from 14.5 GPa to 52.7 GPa for pure MEH-PPV (0% QDs) and 100% QD films, respectively, while H ranges from 220 MPa to 1430 MPa for the same films, respectively. The films behave viscoelastically at lower QD loading, but assume a more granular character as the loading approaches 100%.

This thesis is composed of three projects including the AC magnetic susceptibility study of high-temperature superconductor YBa$_2$Cu$_3$O$_{7-\delta}$, the ionic-liquid gating study of the Mott insulator Sr$_2$IrO$_4$, and the single-electron study of quantum dot device with self-assembled nanocrystal PbS. Chapter 1 covers a general introduction to all three projects. The basic background and the motivation for each project are presented. Project I is covered in Chapter 2, Chapter 3, and Chapter 4. The first part of Chapter 2 is a theoretical introduction to the Bardeen-Cooper-Schrieffer theory of superconductivity with its main conclusions presented. This chapter builds a basis for the use of high pressure technique to YBa$_2$Cu$_3$O$_{7-\delta}$ in the later chapters. The rest of Chapter 2 reviews the work in the study of high-temperature superconductors, especially on YBa$_2$Cu$_3$O$_{7-\delta}$, on both experiments and theories and the possible applications of high-temperature superconductors. Chapter 3 introduces the YBa$_2$Cu$_3$O$_{7-\delta}$ sample preparation process and the characterisation. A dry cryomagnetic equipment was employed for the measurement. The results and the discussion are presented in Chapter 4. Project II is described in Chapter 5, Chapter 6, and Chapter 7. Chapter 5 firstly introduces the background knowledge of the gated material SrTiO$_3$ and the technical details of the ionic-liquid gating technique. Then the sample growth and the characterisation are presented. The fabrication process of Sr$_2$IrO$_4$ and SrTiO$_3$ (material for a control experiment) are described in Chapter 6. Chapter 7 covers the measurement and the result of the fabricated devices and related discussion. Project III ranges from Chapter 8, and Chapter 9. A literature review of quantum-dot devices and self-assembled nanocrystals is presented in Chapter 8. The experimental design of this nanocrystal quantum dot device is also included. Following it, the fabrication process of quantum-dot devices and the techniques used for fabrication are introduced in the start of Chapter 9. Chapter 9 also gives a description of the probe-station for measurements. The results and discussion of the measurements are covered in the last section of Chapter 9. Chapter 10 summarises and concludes the three projects stated above and gives some suggestions about the directions for future work.

Colloidal quantum dots (QDs) are promising candidates for the next generation of solar cells due to their tunable band gaps, solution processability and the potential for multiple exciton generation. However their stability and the reduction of surface defects are big challenges and effective surface passivation is needed. Passivations via organic ligands have been shown to be imperfect and hinder the charge transfer in devices. Three different QD systems, chosen as exemplars of different approaches to surface passivation, have been investigated with synchrotron-radiation (SR) depth- profiling X-ray photoelectron spectroscopy (XPS). With this technique the chemical composition of the top few nanometres of a sample can be studied with depth. The study of CdTe QDs with and without a chloride treatment revealed the presence of stoichiometric particles prior to, and the likely coexistence of Cl atoms and organic ligands on the surfaces of the QDs after the treatment. The chloride treatment led to a better surface passivation of the QDs resulting in photoluminescence quantum yields of up to 97.2%. Shell thickness estimations using a core/shell/shell model were performed of the chloride treated sample and XPS highlighted the complexity of the structure of the sample. CdTe QDs passivated by a thick CdSe shell were investigated. Indications for an improvement of the stability of the QDs against oxidation were found. The Se:Te ratio was equivalent to a CdSe shell of 0.3-0.4 nm which was significantly smaller than intended, indicating that the butylamine ligand exchange and/or the washing of the sample reduced the thickness of the CdSe shell drastically. The third system studied was PbS QDs that were passivated with a thin CdS shell. XPS of the thoroughly washed QDs confirmed the presence of Cd in an amount equivalent to a 0.13-0.18 nm thick shell. This is thicker than the 0.05 nm shell expected from absorption spectroscopy. A study of ageing of the PbS/CdS QDs revealed that oxidation took place on the surface of the QDs. It was found that sulfur oxidised in stages leading to highly oxidised SO4^2- components. Upon long-term ageing Pb oxidised more rapidly than S, and either some Pb and/or Cd migration or some decomposition of the QDs occurred. The PbS/CdS nanoparticles were more stable than a comparable PbS colloidal quantum dot sample from the literature. The study of the PbS/CdS QDs prior to and after the second wash- ing cycle after a mercaptopropionic acid (MPA) ligand exchange revealed, amongst other things, the removal of MPA and a reduction of the Cd:Pb ratio indicating that (parts of) the QDs decomposed through the ligand exchange or the washing. In addition to the results of the nanoparticles studied some limitations of the study of colloidal QDs with SR depth-profiling XPS are discussed.