Dissertations / Theses on the topic 'InAs QDs on silicon'

2012/2013
Riassunto (Abstract) La fabbricazione di punti quantici (quantum dots, QD) auto-assemblati è una tematica di particolare rilevanza a causa delle loro possibilità di applicazione in dispositivi optoelettronici. Nel presente lavoro, ci siamo prefissi di ottenere array di QD di semiconduttore altamente uniformi, con lo scopo di raggiungere un controllo completo sulla loro distribuzione spaziale, ed un’uniformità spettrale superiore, rispetto a QD auto-assemblati convenzionali. Il metodo consiste in un approccio combinato top-down ebottom-up: QD auto-assemblati vengono cresciuti tramite Epitassia a Fasci Molecolari (MolecularBeam Epitaxy, MBE) su una superficie pre-patternata con un array regolare di buchi. Nella prima parte di questo lavoro di tesi, abbiamo ottimizzato la crescita di punti di InAs su substrati patternati di GaAs. Nella maggior parte dei lavori precedenti, i substrati vengono patternati tramite litografia elettronica (Electron BeamLitography, EBL), che non è la tecnica ottimale per l’applicazione a dispositivi broad-area, a causa della sua natura seriale e degli alti costi. Il metodo più indicato per superare questa limitazione è la scelta di un approccio litografico alternativo come la litografia a nanostampa (Nanoimprint Lithography,NIL), che ha come vantaggi un alto volume di produzione e dei costi più ridotti. Comunque, esistono soltanto pochi studi sull’uso della NIL per il patterning di superfici di GaAs. Nella maggior parte dei casi, viene usata la variante dell’UV-NIL, che richiede l’uso addizionale di una maschera di SiO2, con conseguente complicazione del processo. In questo lavoro, abbiamo utilizzato la forma di NIL più semplice, che non richiede alcun processo addizionale (quali il trattamento in UV o uno strato di SiO2). L’attacco chimico del GaAs è stato effettuato tramite wetetcìhing, per ottenere superfici prive di difetti, ed è stato ottimizzato per ottenere nanopori con le dimensioni laterali e la profondità desiderate. I substrati patternati di GaAs, dopo un processo di pulizia, vengono usati per la crescita controllata di QD nell’MBE. Abbiamo ottimizzato i protocolli di crescita per migliorare le proprietà strutturali ed ottiche dei QD, con lo scopo di migliorare le prestazioni di dispositivi optoelettronici. Abbiamo esaminato l’effetto del desorbimento dell’ossido superficiale a bassa temperatura tramite fasci di gallio (tecnica già applicata con successo in precedenza)e l’effetto della crescita sul substratodi uno strato di GaAs sulla forma e dimensione dei buchi, per migliorare l’occupazione di QD singoli nei buchi stessi. Abbiamo adottato un approccio alternativo per rimuovere l’ossido superficiale usando fasci di indio. Questo approccio è stato adottato siccome l’eventuale indio in eccesso può essere desorbito facilmente dalla superficie di GaAs scaldando a temperature che non alterino il profilo dei buchi. Inoltre, abbiamo osservato che nei nostri array patternati il desorbimento tramite fasci di indio ha anche l’effetto benefico di preservare meglio la forma dei buchi dopo la deposizione dello strato di GaAs, il che è di grande aiuto per migliorare l’occupazione di QD singoli e l’uniformità dei QD. Abbiamo usato la spettroscopia di fotoluminescenza (photoluminescence, PL) a bassa temperatura per esaminare le proprietà ottiche dei punti di InAs/GaAs. I risultati sono confrontabili o migliori, rispetto ai pochi studi simili effettuati su punti cresciuti su substrati definiti con la NIL, ed hanno il potenziale di raggiungere l’allargamento spettrale non-omogeneo ottenuto su QD controllati ottimizzati realizzati con la tecnica EBL. Nella seconda parte della tesi, abbiamo esteso questa tecnica alla crescita selettiva di punti InAs/GaAs su silicio patternato. Il silicio è il materiale principale per i dispositivi a semiconduttore (90%). Comunque, il gap di energia indiretto del silicio limita la realizzazione di dispositivi optoelettronici efficienti. D’altra parte, a causa del loro gap in molti casi diretto, i semiconduttori III-V hanno proprietà ottiche ed optoelettroniche eccellenti. L’integrazione di semiconduttori composti III-V su Si ha una grande prospettiva per la realizzazione di circuiti fotonici integrati. Tuttavia, l’integrazione GaAs/Si è limitata fortemente dalla loro differenza del 4% nel parametro reticolare, che induce dislocazioni e difetti nel materiale cresciuto sul substrato di silicio e previene l’emissione di luce. Svariati approcci sono stati esplorati per l’integrazione III-V/Si, quali la crescita di nanofili, la fusione dei substrati, l’utilizzo di strati spessi di rilassamento e l’applicazione di composizioni di materiali per accomodare i parametri reticolari. La crescita di strati spessi aumenta la complessità dei processi ed il costo dei materiali. Un’alternativa sarebbe quindi la crescita diretta di nanostrutture III-V su silicio. La crescita di QD III-V su substrati Si patternati dovrebbe aiutare a ridurre i difetti a causa di effetti di taglia e rilassamento della tensione laterale. Tuttavia, ad oggi esistono solo pochi studi su substrati di Si patternati, e la maggior parte includono maschere di SiO2con grosse periodicità dei buchi (1 µm) e piccole aree patternate (1 mm2) definite tramite EBL. In questo lavoro abbiamo cresciuto nanostrutture III-V direttamente su substrati patternati di Si con array di buchi densi (periodo 300 nm) e grandi aree patternate attraverso la tecnica NIL. Abbiamo ottimizzato l’attacco chimico del SI e la passivazione della sua superficie tramite terminazione in idrogeno, allo scopo di facilitare il desorbimento dell’ossido nella camera MBE a temperature relativamente basse. Il desorbimento dell’ossido ed i protocolli di crescita sono stati ottimizzati allo scopo di ottenere array di punti uniformi. Abbiamo usato la spettroscopia EDS (Energy Dispersive X-rayspectroscopy) per verificare la presenza di InAs e GaAs solo in corrispondenza dei buchi. Verranno anche mostrati risultati preliminari di spettroscopia PL per controllare le proprietà ottiche dei punti InAs/GaAs.
Abstract The fabrication of self-assembled quantum dots (QDs) is a topic of high current interest due to their vast applications in optical devices. In this research work, our aim is to obtain highly uniform arrays of semiconductor QDs to reach a complete control on their spatial distribution and a superior spectral uniformity, with respect to conventional self-assembled dots. The method consists of a combined top-down and bottom-up approach: self-assembled QDs are grown by molecular beam epitaxy (MBE) on a pre-patterned surface with a regular array of holes. In the first part of this thesis work, we have optimized the growth of InAs dots on patterned GaAs substrates. In most of the research efforts, GaAs substrates are patterned through electron beam lithography (EBL), which is not the optimal technique for application in broad-area devices, due to its serial nature and high cost. The finest way to overcome this limitation can be through choosing an alternative lithographic approach like nanoimprint lithography (NIL) for patterning of the GaAs surfaces, which has the advantage of high throughput and low cost. However, there are only few studies available that have used NIL for the patterning of GaAs surfaces. In most of the cases, instead of NIL, UV-NIL is being used for patterning, that requires an additional layer of SiO2 for masking, which also complicates the process. In this work, we have tried to use the simplest form of NIL for patterning, which requires no additional processing (like UV treatment or SiO2 layer). Wet etching process is chosen for GaAs etching to get defect-free surfaces, and is optimized to get the nanopores with required lateral dimensions and depth. These patterned GaAs substrates after optimizing all cleaning procedures are used for further growth of site-controlled QDs in MBE. We have optimized the growth protocols to improve the structural and optical properties of the dots, with the aim of improving the performance of optoelectronic devices. We have examined the effect of the low-temperature oxide desorption by means of Ga beams (which was already applied successfully to patterned GaAs surfaces) and the effect of the GaAs buffer layer growth on the hole shape and size, to improve the single-dot occupancy of the patterned holes. We have adopted an alternative approach to remove the oxide layer using In beams. This approach have been adopted because excess In can be easily desorbed from GaAs surface just by heating it to temperatures that do not alter the hole profiles. Furthermore, we have observed that for our patterned arrays In-assisted desorption has also the beneficial effect to better preserve the hole shape after the growth of the GaAs buffer layer, which ultimately helps in improving the single-dot occupancy, as well as the structural uniformity of the dots. We have used low temperature photoluminescence (PL) spectroscopy to assess the optical properties of InAs/GaAs dots. The results compare favourably with the few similar dot arrays previously grown on NIL-defined patterns, and have the potential to match the inhomogeneous broadening reported for optimized site-controlled dots on EBL-defined patterns. In the second part of the thesis, we have extended this technique to the selective growth of InAs/GaAs QDs on patterned silicon. Silicon is the main material for semiconductor devices (90%). However, the indirect bandgap of silicon prevents the realization of efficient light emitting devices. On the other hand, due to their direct bandgap in many cases, III-V semiconductors have excellent optical properties and optoelectronic capabilities. Integration of III-V compound semiconductor with Si has a broad prospective for the realization of photonic integrated circuits. However, GaAs/Si integration is largely limited by their 4% lattice mismatch, which induces dislocations and defects in the grown material on the Si substrate and ultimately prevents light emission. Several approaches are under exploration for III-V/Si integration like nanowire growth, wafer fusion techniques, using thick relaxation layers and applying lattice matched material compositions. Growth of these buffer layers increases the process complexity and material cost. A perfect alternative would thus be the direct epitaxial growth of III-V nanostructures on silicon. Growth of III-V quantum dots on pre-patterned Si substrates should help to reduce defects because of size effect and effective lateral stress relaxation due to the presence of facet edges and side walls. However, there are limited research efforts available on patterned Si substrates including mostly on SiO2 as mask with large periods (1µm) and small patterned areas (1mm2) defined by EBL. In our work, we have grown III-V nanostructures directly on patterned Si substrates with dense hole arrays (period 300nm) and larger patterned areas through nanoimprint lithography. We optimized the dry etching of Si and its surface passivation with H-termination, in order to facilitate oxide desorption in the MBE at relatively low temperatures. Oxide desorption and growth protocols were optimized in order to obtain uniform dot arrays. We used energy-dispersive X-ray spectroscopy (EDS) analysis for the characterization of InAs/GaAs QDs to verify the presence of GaAs and InAs only at the hole location. We will also show preliminary results using photoluminescence spectroscopy to assess the optical properties of InAs/GaAs dots.
XXVI Ciclo
1976

The main focus of this master thesis work has been to image InAs emph{quantum dots} (QDs) using emph{atomic force microscopy} (AFM), to identify and evaluate various image processing methods used to estimate the volume of the InAs QDs. The InAs QDs studied in this thesis work, had been deposited on GaAs substrates, using solid-source emph{molecular beam epitaxy} (MBE) before the thesis work started. The total QD volume was determined for all samples, using eight different estimation methods. The purpose of estimating the total QD volume, was to compare the total volume to the deposited volume.Previous studies on similar samples, have indicated that the total volume can be larger than the deposited volume during MBE growth. This discrepancy is explained by incorporation of Ga from the substrate during growth. This was not observed in this thesis work. One possible explanation is that the samples have oxidized; resulting in a lower measured height.In addition, the relationship between atomic steps, defects and the appearance of large QDs were studied. parTwo series of samples were studied: one in which the QD growth temperature was varied and one in which the amount of deposited InAs was varied. The total QD volumes were found to increase with the QD growth temperature and the deposited InAs thickness. Square-shaped defects and contours of 2D islands were observed in nearly all samples. Higher/multiple terraces seem to be related to regions of higher QD density.

This diploma project has been focused on optimization of the D2B IR detector fabrication process using different mesa sidewall treatments and passivation methods. X-ray photoelectron spectroscopy (XPS), scanning electron microscopy (SEM) and atomic force microscope (AFM) measurements have been carried out on samples treated by different wet etching methods, to analyze their surface chemical composition and roughness. The surface roughness has been improved by critic etching, annealing and NaClO sequential treatment steps. Then these results have been utilized to improve the process of the D2B IR detectors. The dark current of the fabricated detectors passivated with various techniques have been characterized by I-V measurements at low (77 K) and room temperatures. The dark current mechanisms owing to surface shunt or bulk leakage are investigated by dark current temperature dependence analysis. By photoresist passivation devices with least leakage current are achieved.

Si-based light emitting sources are highly demanded for applications in optoelectronic integration circuits. Unfortunately, Si has an indirect bandgap and thus a low efficiency in photon emission. On the other hand, III–V semiconductors have superior optical properties and are considered as strong candidates to achieve efficient light emitting sources on Si platforms via wafer bonding or monolithically epitaxy growth. III–V materials monolithically grown on Si substrate could introduce various types of defects including antiphase domain, threading dislocation, misfit dislocation. These defects must be dealt with satisfactorily in order to fulfill the potential of III–V/Si integration. In this thesis, buffer layers for InAs/GaAs quantum dots (QDs) monolithically grown Si substrate have been investigated. The buffer layer study is mainly focused on the different types of defect filter layers (DFLs). The measurements of atomic force microscopy, photoluminescence and transmission electron microscopy are carried out to investigate the effectiveness of each type of DFLs. The results of lasers and superluminescent diodes (SLDs) have been presented based on the studies of DFLs. In order to improve the performance of InAs/GaAs QDs grown on Si substrates, a GaAs buffer layer and DFLs have been used to reduce the defect density from ~1010 to 106 cm-2 after three sets of DFLs, which consists of strained layer superlattices (SLSs). In the thesis, the optimisation of DFLs has been carried out. Different types of DFLs are investigated in the Chapter 3, including InAs/GaAs QDs, InGaAs submonolayer QDs, InGaAs/GaAs SLSs and InAlAs/GaAs SLSs. DFLs made of InAlAs/GaAs SLSs show the strongest performance, based on the measurements of atomic force microscopy, photoluminescence and transmission electron microscopy. The high performance InAs/GaAs QDs lasers with low threshold current density (194 A/cm2 ) and high operating temperature (85 ̊C) has been obtained for the samples with optimised DFLs. In addition to III–V/Si lasers, III–V SLDs monolithically grown on silicon substrates would further enrich the silicon photonics toolbox, enabling low-cost, highly scalable, high-functional, and streamlined on-chip light sources. In this thesis, the first InAs/GaAs QD SLDs monolithically grown on a Si substrate have been demonstrated based on the similar growth structure of laser devices. The fabricated two-section InAs/GaAs QD SLD produces a close- 4 to-Gaussian emission spectrum of 114 nm centred at ∼1255 nm wavelength, with a maximum output power of 2.6 mW at room temperature. The optimisation of InGaAs/GaAs SLSs DFLs has been carried out in the Chapter 5. The optimisation includes introducing different growth methods into GaAs spacer layer between each set of DFL, indium composition and GaAs thickness in InGaAs/GaAs SLSs. The optimisation is examined by atomic force microscopy, photoluminescence and transmission electron microscopy. The laser device with optimised InGaAs/GaAs SLSs DFLs has a lower threshold current density, higher operating temperature and characteristic temperature. In conclusion, InAs/GaAs QDs lasers with low threshold current density and the first QDs SLDs monolithically grown on Si substrates have been demonstrated. InAlAs/GaAs SLSs DFLs have been proved that as considerable solution to reduce the threading dislocation density significantly. The optimisations of InGaAs/GaAs SLSs DFLs successfully improve the QDs laser performance which could also be used in III–V/Si monolithically integration. The III–V QDs lasers and SLDs monolithically grown on Si substrate are essential steps for Si photonics integration, which will fill the “holy grail” of opto-electronic integration circuits.

To imitate the way electrical components evolved from discrete devices to devices integrated on Si platform, the next stage for integrated circuits is to integrate photonic components with electrical components on one chip, with active devices known as optoelectronic integrated circuits (OEIC). An ideal solution for this would be to have an all-Si laser. However due to the indirect bandgap of Si this is difficult to achieve. Therefore attention has been focused on trying to integrate the existing and mature III-V laser technology with Si. The difference in lattice constant between GaAs and Si makes direct, monolithic growth of GaAs on Si difficult due to the generation of high defect densities. But the advances in quantum dot (QD) technology and in III-V buffer layer techniques have led to the improvements of direct growth integration. In this thesis an AlAs nucleation layer (NL) in the place of a GaAs nucleation layer was found to increase the photoluminescence intensity and reduce defect density in active layers. Lasers were fabricated with lower threshold current densities than similar devices with GaAs NL. Lasing operation at 1.28 μm was achieved up to 63 °C with a threshold current density of 675 A/cm2 at room temperature. In addition, Ge-on-Si substrates have been used to demonstrate the lasers on Si substrates with a very low pulsed threshold current density of 64 A/cm2, which is significantly lower than any other laser integrated with Si substrates. Also this was the first demonstration of a CW laser on Si with a threshold current density of 163 A/cm2. Lasers were operated up to 30 °C for CW devices and 84 °C for pulsed devices. The difference in threshold currents and temperature performance between CW and pulsed operation is due to high device resistances caused by a combination of poor contact resistance and the introduction of defects from the Si/Ge interface. In conclusion, lasers on Si substrates have been fabricated with low threshold current densities. A very low threshold current density of 64 A/cm2 has been achieved with a Ge-on-Si device and is the lowest result for any type of Si laser at the time of writing, which shows good potential for future integration with Si electronics.

Grâce à leur propriétés uniques, les nanofils d'InAs et de Bi1-xSbx sont important pour les domaines de la nanoélectronique et de l'informatique quantique. Alors que la mobilité électronique de l'InAs est intéressante pour les nanoélectroniques; l'aspect isolant topologique du Bi1-xSbx peut être utilisé pour la réalisation de Qubits basés sur les fermions de Majorana. Dans les deux cas, l'amélioration de la qualité du matériau est obligatoire et ceci est l'objectif principal cette thèse ou` nous étudions l'intégration des nanofils InAs sur silicium (compatibles CMOS) et où nous développons un nouvel isolant topologique nanométrique: le Bi1-xSbx. Pour une compatibilité CMOS complète, la croissance d'InAs sur Silicium nécessite d'être auto- catalysée, entièrement verticale et uniforme sans dépasser la limite thermique de 450 ° C. Ces normes CMOS, combineés à la différence de paramètre de maille entre l'InAs et le silicium, ont empêché l'intégration de nanofils InAs pour les dispositifs nanoélectroniques. Dans cette thèse, deux nouvelles préparations de surface du Si ont été étudiées impliquant des traitements Hydrogène in situ et conduisant à la croissance verticale et auto-catalysée de nanofils InAs compatible avec les limitations CMOS. Les différents mécanismes de croissance résultant de ces préparations de surface sont discutés en détail et un passage du mécanisme Vapor-Solid (VS) au mécanisme Vapor- Liquid-Solid (VLS) est rapporté. Les rapports d'aspect très élevé des nanofils d'InAs sont obtenus en condition VLS: jusqu'à 50 nm de diamètre et 3 microns de longueur. D'autre part, le Bi1-xSbx est le premier isolant topologique 3D confirmé expérimentalement. Dans ces nouveaux matériaux, la présence d'états surfacique conducteurs, entourant le coeur isolant, peut héberger les fermions de Majorana utilisés comme Qubits. Cependant, la composition du Bi1-xSbx doit être comprise entre 0,08 et 0,24 pour que le matériau se comporte comme un isolant topologique. Nous rapportons pour la première fois la croissance de nanofils Bi1-xSbx sans défaut et à composition contrôlée sur Si. Différentes morphologies sont obtenues, y compris des nanofils, des nanorubans et des nanoflakes. Leur diamètre peut être de 20 nm pour plus de 10 microns de long, ce qui en fait des candidats idéaux pour des dispositifs quantiques. Le rôle clé du flux Bi, du flux de Sb et de la température de croissance sur la densité, la composition et la géométrie des structures à l'échelle nanométrique est étudié et discuté en détail
InAs and Bi1-xSbx nanowires with their distinct material properites hold promises for nanoelec- tronics and quantum computing. While the high electron mobility of InAs is interesting for na- noelectronics applications, the 3D topological insulator behaviour of Bi1-xSbx can be used for the realization of Majorana Fermions based qubit devices. In both the cases improving the quality of the nanoscale material is mandatory and is the primary goal of the thesis, where we study CMOS compatible InAs nanowire integration on Silicon and where we develop a new nanoscale topological insulator. For a full CMOS compatiblity, the growth of InAs on Silicon requires to be self-catalyzed, fully vertical and uniform without crossing the thermal budge of 450 °C. These CMOS standards, combined with the high lattice mismatch of InAs with Silicon, prevented the integration of InAs nanowires for nanoelectronics devices. In this thesis, two new surface preparations of the Silicon were studied involving in-situ Hydrogen gas and in-situ Hydrogen plasma treatments and leading to the growth of fully vertical and self-catalyzed InAs nanowires compatible with the CMOS limitations. The different growth mechanisms resulting from these surface preparations are discussed in detail and a switch from Vapor-Solid (VS) to Vapor- Liquid-Solid (VLS) mechanism is reported. Very high aspect ratio InAs nanowires are obtained in VLS condition: upto 50 nm in diameter and 3 microns in length. On the other hand, Bi1-xSbx is the first experimentally confirmed 3D topololgical insulator. In this new material, the presence of robust 2D conducting states, surrounding the 3D insulating bulk can be engineered to host Majorana fermions used as Qubits. However, the compostion of Bi1-xSbx should be in the range of 0.08 to 0.24 for the material to behave as a topological insula- tor. We report growth of defect free and composition controlled Bi1-xSbx nanowires on Si for the first time. Different nanoscale morphologies are obtained including nanowires, nanoribbons and nanoflakes. Their diameter can be 20 nm thick for more than 10 microns in length, making them ideal candidates for quantum devices. The key role of the Bi flux, the Sb flux and the growth tem- perature on the density, the composition and the geometry of nanoscale structures is investigated and discussed in detail

We reported the structure peculiarities of nanocrystals formed in Si by means of high-fluence implantation at 25 and 500 °С followed by rapid thermal annealing (RTA). The structure of implanted samples has been investigated by means of transmission electron microscopy (TEM). The crystalline nature of the precipitates is proved by the Moiré fringe patterns presence in the TEM images. The Moiré fringe distance (Moiré period) is equal of 1.8 nm for small precipitates. This experimental value coincides with the calculated one for crystalline InAs. It is noted a Moiré period increasing in the case of large precipitates. We suppose that this feature is a result of surplus As or In atoms embedded in precipitates. One can see an interesting effect – “glowng” of nanocrystal/Si interfaces at the dark-field images of implanted and annealed samples. We ascribe this effect to a presence of misfit dislocation networks at the InAs/Si interfaces generated as a result of strain relaxation in highly mismatched InAs/Si system. When you are citing the document, use the following link http://essuir.sumdu.edu.ua/handle/123456789/35436

Les boîtes quantiques sous formes de nanocristaux semi-conducteurs permettent de réaliser des matériaux à énergie de gap variable, propriété très intéressante pour les composants optoélectroniques. Ce travail est dédié à la création de nanocristaux de silicium dopés enfouis dans SiO2 et de nanocistaux binaires (InAs et GaAs) et ternaires d’InxGa1-xAs enfouis dans Si et à leurs caractérisations structurales, électriques et optiques. La synthèse par faisceaux d’ions permet d’avoir un contrôle de la quantité et de la taille des nanocristaux synthétisés. Des caractérisations structurales ont pu démontrer le dopage des nanocristaux de silicium avec le phosphore et l’arsenic à une concentration atomique moyenne de 8 %. Nous avons également montré la possibilité de moduler la taille et la composition chimique des nanocristaux d’InxGa1-xAs sur une large gamme à l’aide de la dose d’implantation et de la température de recuit
Semiconductor nanocrystals can be used as quantum dots to produce band gap engineering by varying the nanocrystals size, which is a very interesting property for optoelectronic components. This work is dedicated to the creation of doped silicon nanocrystals embedded in SiO2 and binary (InAs and GaAs) and ternary nanocrystals of InxGa1-xAs embedded in Si and also to investigate their structural, electrical and optical properties. Ion beam synthesis allows a control of the nanocrystals amount and size. Structural characterizations were able to demonstrate the doping of silicon nanocrystals with phosphorus and arsenic at an average atomic concentration of 8 %. We have also shown the ability to modulate the size and the chemical composition of InxGa1-xAs nanocrystals in a large range by varying the implantation dose and the annealing temperature

Near-field optical microscopy has attracted remarkable attention, as it is the only technique that allows the investigation of local optical properties with a resolution far below the diffraction limit. Especially, the scattering-type near-field optical microscopy allows the nondestructive examination of surfaces without restrictions to the applicable wavelengths. However, its usability is limited by the availability of appropriate light sources. In the context of this work, this limit was overcome by the development of a scattering-type near-field microscope that uses a widely tunable free-electron laser as primary light source. In the theoretical part, it is shown that an optical near-field contrast can be expected when materials with different dielectric functions are combined. It is derived that these differences yield different scattering cross-sections for the coupled system of the probe and the sample. Those cross-sections define the strength of the near-field signal that can be measured for different materials. Hence, an optical contrast can be expected, when different scattering cross-sections are probed. This principle also applies to vertically stacked or even buried materials, as shown in this thesis experimentally for two sample systems. In the first example, the different dielectric functions were obtained by locally changing the carrier concentration in silicon by the implantation of boron. It is shown that the concentration of free charge-carriers can be deduced from the near-field contrast between implanted and pure silicon. For this purpose, two different experimental approaches were used, a non-interferometric one by using variable wavelengths and an interferometric one with a fixed wavelength. As those techniques yield complementary information, they can be used to quantitatively determine the effective carrier concentration. Both approaches yield consistent results for the carrier concentration, which excellently agrees with predictions from literature. While the structures of the first system were in the micrometer regime, the capability to probe buried nanostructures is demonstrated at a sample of indium arsenide quantum dots. Those dots are covered by a thick layer of gallium arsenide. For the first time ever, it is shown experimentally that transitions between electron states in single quantum dots can be investigated by near-field microscopy. By monitoring the near-field response of these quantum dots while scanning the wavelength of the incident light beam, it was possible to obtain characteristic near-field signatures of single dots. Near-field contrasts up to 30 % could be measured for resonant excitation of electrons in the conduction band of the indium arsenide dots
Die optische Nahfeldmikroskopie hat viel Beachtung auf sich gezogen, da sie die einzige Technologie ist, welche die Untersuchung lokaler optischer Eigenschaften mit Auflösungen unterhalb der Beugungsgrenze ermöglicht. Speziell die streuende Nahfeldmikroskopie erlaubt die zerstörungsfreie Untersuchung von Oberflächen ohne Einschränkung der verwendbaren Wellenlängen. Die Nutzung ist jedoch durch das Vorhandensein entsprechender Lichtquellen beschränkt. Im Rahmen dieser Arbeit wurde diese Beschränkung durch Entwicklung eines streuenden Nahfeldmikroskops überwunden, das einen weit stimmbaren Freie-Elektronen-Laser als primäre Lichtquelle benutzt. Im theoretischen Teil wird gezeigt, dass ein optischer Kontrast erwartet werden kann, wenn Materialien mit unterschiedlichen Dielektrizitätskonstanten kombiniert werden. Es wird hergeleitet, dass diese Unterschiede in unterschiedlichen Streuquerschnitten für das gekoppelte System aus Messkopf und Probe resultieren. Diese Streuquerschnitte definieren die Stärke des Nahfeldsignals, welches auf unterschiedlichen Materialien gemessen werden kann. Ein optischer Kontrast kann also erwartet werden, wenn unterschiedliche Streuquerschnitte untersucht werden. Dass dieses Prinzip auch auf übereinander geschichtete oder sogar verborgene Strukturen angewendet werden kann, wird in dieser Doktorarbeit an zwei Probensystemen experimentell gezeigt. Im ersten Beispiel wurden die unterschiedlichen Dielektrizitätskonstanten durch örtliches Ändern der Ladungsträgerdichte in Silizium durch Bor-Implantation erreicht. Es wird gezeigt, dass die Dichte der freien Ladungsträger an Hand des optischen Kontrastes zwischen implantiertem und reinem Silizium ermittelt werden kann. Zu diesem Zweck wurden zwei unterschiedliche Ansätze verwendet, ein nicht-interferometrischer mittels variabler Wellenlängen und ein interferometrischer mit einer konstanten Wellenlänge. Weil diese Techniken gegensätzliche Informationen liefern, können sie genutzt werden, um die effektive Ladungsträgerdichte quantitativ zu bestimmen. Beide Ansätze lieferten konsistente Resultate für die Trägerdichte, welche sehr gut mit den Vorhersagen der Literatur übereinstimmt. Während die Strukturen im ersten Beispiel im Mikrometer-Bereich lagen, wird die Möglichkeit, verborgene Nanostrukturen zu untersuchen, an Hand einer Probe mit Indiumarsenid Quantenpunkten demonstriert. Diese sind von einer dicken Schicht Galliumarsenid bedeckt. Zum ersten Mal wird experimentell gezeigt, dass Übergänge zwischen Elektronenzuständen in einzelnen Quantenpunkten mit Nahfeldmikroskopie untersucht werden können. Durch die Messung der Nahfeld-Antwort der Quantenpunkte unter Änderung der Wellenlänge des eingestrahlten Lichtes war es möglich, charakteristische Nahfeld-Signaturen der einzelnen Quantenpunkte zu erhalten. Nahfeld-Kontraste bis zu 30 Prozent konnten für die resonante Anregung der Elektronen im Leitungsband der Indiumarsenid Punkte beobachtet werden

The aim of this thesis is to understand the dynamics of the nucleation and growth of III-V semiconductor nanowires and associated heterostructures grown by chemical beam epitaxy. These nanowires represent well-controlled and high quality materials suitable for both fundamental physics and applications in optical and electronic devices. The first part of the thesis investigates growth recipes to obtain Au-catalyzed InAs NWs with controlled morphology. Good control of NW length and diameter distributions has been achieved by a systematic study of two different Au deposition techniques: Au thin film deposition and colloidal dispersion. Triggered by the issues of Au contamination and CMOS compatibility, the second part of the thesis is dedicated to the investigation of the nucleation and growth mechanisms of Au-free InAs NWs on Si (111) substrates. A thorough analysis of the silicon substrate preparation is conducted and an optimized silicon surface for the nucleation of Au-free nanowires is identified. We show that the silicon surface can be modified by in situ and ex situ parameters allowing us to control the density of NWs. Growth conditions were established for growing InAs NWs either by catalyst-free or self-catalyzed mechanisms on Si (111). The catalyst-free growth proceeds in the vapor-solid growth mechanism without the use of any catalyst particle while the self-catalyzed growth proceeds in the vapor-liquid-solid mechanism involving a liquid In droplet. Growth models are proposed in order to interpret the experimental findings. The third part of the thesis concerns the growth of axial and radial (core-shell) heterostructured NWs. Nanowire heterostructures combining either highly lattice mismatched materials (GaAs and InAs) or almost lattice matched materials (InAs and GaSb) are investigated. GaAs/InAs and InAs/GaAs axial heterostructures are grown by Au-catalyzed method. Here, it is demonstrated that the catalyst composition, rather than other growth parameters, as postulated so far, controls the growth mode and the resulting NW morphology. We have also explored the growth of core-shell InAs/GaSb heterostructures by catalyst-free mechanism. The morphology and structural properties of InAs/GaSb core-shell heterostructures are optimized to fabricate Esaki tunnel diodes exploiting their broken-gap band alignment.

Une source de photons uniques (SPU) dans la bande télécom, épitaxiées sur un substrat de silicium (Si), est le Saint Graal pour réaliser des dispositifs CMOS compatibles pour les technologies de l'information optiques. Pour atteindre cet objectif, nous proposons la croissance monolithique de Boîte Quantiques-Nanofils (BQ-NFs) InAs/InP sur des substrats de silicium par épitaxie par jet moléculaire (EJM) en utilisant la méthode vapeur-liquide-solide (VLS). Au début, nous avons concentré nos efforts sur l'optimisation des conditions de croissance afin d'obtenir une densité de NF ultra-faible sans effort avant ou après la croissance, ce qui nous permet d'exciter optiquement un seul BQ-NF sur l'échantillon tel qu'il a été épitaxiées et de préserver la croissance monolithique sur le silicium. Par la suite, nous avons porté notre attention sur l'amélioration de l'extraction de la lumière de la BQ InAs du guide d'onde InP NF vers l'espace libre pour obtenir une source lumineuse avec un profil d'émission en Champ Lointain (CL) gaussien afin de coupler efficacement les photons individuels à une fibre optique monomode. Cela a été réalisé en contrôlant la géométrie de NF pour obtenir des NFs en forme d'aiguille avec un très petit angle de conicité et un diamètre de NF adapté pour supporter un guide d'onde monomode. Une telle géométrie a été produite avec succès en utilisant un équilibre induit par la température sur les croissances axiale et radiale pendant la croissance des NFs catalysée par l'or. Des mesures optiques ont confirmé la nature mono-photonique des photons émis avec g2(0) = 0,05 et un profil d'émission gaussien en CL avec un angle d'émission θ = 30°. Pour obtenir des performances optimales, nous avons ensuite abordé une question cruciale dans cette géométrie de NF représentée par l'état de polarisation inconnu des photons émis. Pour résoudre ce problème, une solution consiste à intégrer un seul BQ dans un NF avec une section asymétrique optimisée pour inhiber un état de polarisation et améliorer l'efficacité d'émission de l'autre. Une stratégie de croissance originale a été proposée, permettant d'obtenir des photons à haut degré de polarisation linéaire parallèle à l'axe allongé des NFs asymétriques. Enfin, l'encapsulation des BQ-NFs dans des guides d'ondes en silicium amorphe (a-Si) a ouvert la voie à la production des dispositifs des SPU entièrement intégrés sur Si dans un avenir proche
A telecom band single photon source (SPS) monolithically grown on silicon (Si) substrate is the Holy Grail to realize CMOS compatible devices for optical-based information technologies. To reach this goal, we propose the monolithic growth of InAs/InP quantum dot-nanowires (QD-NWs) on silicon substrates by molecular beam epitaxy (MBE) using the vapour-liquid-solid (VLS) method. In the beginning, we have focused our efforts on optimizing the growth conditions aiming at achieving ultra-low NWs density without any pre-growth or post-growth efforts allowing us to optically excite a single QD-NW on the as-grown sample and to preserve the monolithic growth on silicon. Subsequently, we have turned our attention on enhancing the InAs QD light extraction from the InP NW waveguide towards the free space to achieve a bright source with a Gaussian far-field (FF) emission profile to efficiently couple the single photons to a single-mode optical fiber. This was done by controlling the NW geometry to obtain needlelike-tapered NWs with a very small taper angle and a NW diameter tailored to support a single mode waveguide. Such a geometry was successfully produced using a temperature-induced balance over axial and radial growths during the gold-catalyzed growth of the NWs. Optical measurements have confirmed the single photon nature of the emitted photons with g2(0) = 0.05 and a Gaussian FF emission profile with an emission angle θ = 30°. For optimal device performances, we have then tackled a crucial issue in such NW geometry represented by the unknown polarization state of the emitted photons. To solve this issue, one solution is to embed a single QD in a NW with an asymmetrical cross-section optimized to inhibit one polarization state and to improve the emission efficiency of the other one. An original growth strategy was proposed permitting us to obtain highly linearly polarized photons along the elongated axis of the asymmetrical NWs. Finally, the encapsulation of the QD-NWs within amorphous silicon (a-Si) waveguides have opened the path to produce fully integrated SPSs devices on Si in the near future

碩士
國立交通大學
分子科學研究所
95
I have investigated the carrier capture and relaxation processes in InAs/GaAs self-assembled quantum dots at room temperature by time-resolved photo- luminescence techniques previously. So a polarization-resolved photo-luminescence up-conversion spectroscopy was set up from the original time-resolved system. Then we can use it to study carrier spin capture and relaxation in quantum dots, wetting layers, and barrier. The electron and hole spin lifetimes are found (electrons spin: ~100ps; hole spin: ~20ps) at 77K. The rapid relaxation of hole spin is about 100fs in bulk GaAs at room temperature, but it is preserved in quantum dots longer time. Because up-conversion system is performed with high time resolution, we can do experiments to study carrier spin dynamics in InAs/GaAs quantum dots. The time-resolved measurements reveal that spin relaxation time can not be fit with monoexponential but can be fit with biexponential. The fast decay time is primarily attributed to the carrier-carrier scattering by coulomb interaction. And the slow decay time corresponds to the spin flip process with phonons.

碩士
國立交通大學
電子工程系所
97
In this paper, we are talking about the optical properties of single layer self-assembled InAs/GaAs QDs. At the low temperature（20K）, we have found that the peak position of the photoluminescence spectrum will have a blue-shift as the power of the photoluminescence increasing. This phenomenon comes from the tunneling time of the carriers between QDs is compatible to the carrier recombination time in QDs. The carrier will re-distribute in QDs during the radiative process, so each QD couldn’t be treated as independent one. The amount of the blue-shift is strongly affected by the density and the ground state energy of QDs. The higher density and the higher ground state energy, the bigger shift of the PL peak position. Based on the experiment above, we also set up a model to check out our explanation of “blue-shift” phenomenon. In the model, beside the “tunneling effect” we also introduce the “bi-exciton effect”. In high density QDs’ sample, simulated PL peak moves like experiment’s doing. So the blue-shift phenomenon comes from the carrier’s “re-distribution” in QDs in high density QDs’ sample. But in low density QDs’ sample, the simulated PL peak doesn’t change whether having “tunneling effect” or not. Until we introduce “bi-exciton effect” to our model, it moves like experiment’s doing. So the blue-shift in low density QDs’ sample comes from the “bi-exciton effect”.

碩士
國立中山大學
物理學系研究所
92
Abstract We have extended the spectral range of the current PL-upconversion apparatus to be operated in infrared. Using the IRPL-upconversion，we study the behavior of carrier cooling of InN film and the relationship between the spacer and lifetime in InAs/GaAs stacked QDs . We excited InN film of the band gap of 0.74eV with ultrafast Ti:sapphire laser of the wavelength 404nm. We found the phonon emission time by hot carriers of InN is 14fs. The hot carriers release their excess energy to the lattice of 35K with a timescale of 100ps. We observed in InAs/GaAs QDs that the shorter life time for samples with thin spacer is due to tunneling effect.