Incorporating label dependency into the binary relevance framework for multi-label classification

In multi-label classification, examples can be associated with multiple labels simultaneously. The task of learning from multi-label data can be addressed by methods that transform the multi-label classification problem into several single-label classification problems. The binary relevance approach is one of these methods, where the multi-label learning task is decomposed into several independent binary classification problems, one for each label in the set of labels, and the final labels for each example are determined by aggregating the predictions from all binary classifiers. However, this approach fails to consider any dependency among the labels. Aiming to accurately predict label combinations, in this paper we propose a simple approach that enables the binary classifiers to discover existing label dependency by themselves. An experimental study using decision trees, a kernel method as well as Naive Bayes as base-learning techniques shows the potential of the proposed approach to improve the multi-label classification performance.

Interaction of miltefosine with intercellular membranes of stratum corneum and biomimetic lipid vesicles

Miltefosine (MT) is an alkylphospholipid approved for breast cancer metastasis and visceral leishmaniasis treatments, although the respective action mechanisms at the molecular level remain poorly understood. In this work, the interaction of miltefosine with the lipid component of stratum corneum (SC), the uppermost skin layer, was studied by electron paramagnetic resonance (EPR) spectroscopy of several fatty acid spin-labels. In addition, the effect of miltefosine on (i) spherical lipid vesicles of 1,2-dipalmitoyl-sn-glycero-3-phosphatidylcholine (DPPC) and (ii) lipids extracted from SC was also investigated, by EPR and time-resolved polarized fluorescence methods. In SC of neonatal Wistar rats, 4% (w/w) miltefosine give rise to a large increase of the fluidity of the intercellular membranes, in the temperature range from 6 to about 50 degrees C. This effect becomes negligible at temperatures higher that ca. 60 degrees C. In large unilamelar vesicles of DPPC no significant changes could be observed with a miltefosine concentration 25% molar, in close analogy with the behavior of biomimetic vesicles prepared with bovine brain ceramide, behenic acid and cholesterol. In these last samples, a 25 mol% molar concentration of miltefosine produced only a modest decrease in the bilayer fluidity. Although miltefosine is not a feasible skin permeation enhancer due to its toxicity, the information provided in this work could be of utility in the development of a MT topical treatment of cutaneous leishmaniasis. Published by Elsevier B.V.

Uma análise dos principais elementos influenciadores da tomada de decisão de compra de produtos de marca própria de supermercados

A adoção de marcas próprias no mix de produtos de supermercados é uma ação estratégica que alguns varejistas estão utilizando para melhorar sua competitividade no setor. No Brasil, a participação de mercado das marcas próprias ainda é pouco representativa, o que reflete a atitude dos consumidores diante desses produtos. A proposta deste trabalho foi avaliar os fatores que influenciam no comportamento do consumidor em relação aos produtos de marcas próprias de supermercados. Foi desenvolvida uma pesquisa de campo de caráter exploratório com abordagem quantitativa, por meio de questionário auto-administrado, o que possibilitou a coleta de opiniões de uma amostra de 983 clientes de supermercados. Os resultados obtidos demonstram que os respondentes não possuem uma imagem positiva dos produtos de marcas próprias, devido: à grande variação na qualidade dos produtos dentro das categorias e entre as categorias ofertadas, ao risco percebido, à ausência de comunicação efetiva sobre os produtos e à imagem de inferioridade transmitida pelos atributos das marcas próprias como embalagem, forma de exposição e política de preços. Dos fatores obtidos, pode-se concluir que a imagem da loja, a comunicação e preço, a qualidade e preço destacaram-se como os mais importantes para os respondentes e que exerceram maior influência em seu comportamento.

Plasma kinetics of an LDL-like nanoemulsion and lipid transfer to HDL in subjects with glucose intolerance

OBJECTIVE: Glucose intolerance is frequently associated with an altered plasma lipid profile and increased cardiovascular disease risk. Nonetheless, lipid metabolism is scarcely studied in normolipidemic glucose-intolerant patients. The aim of this study was to investigate whether important lipid metabolic parameters, such as the kinetics of LDL free and esterified cholesterol and the transfer of lipids to HDL, are altered in glucose-intolerant patients with normal plasma lipids. METHODS: Fourteen glucose-intolerant patients and 15 control patients were studied; none of the patients had cardiovascular disease manifestations, and they were paired for age, sex, race and co-morbidities. A nanoemulsion resembling a LDL lipid composition (LDE) labeled with 14C-cholesteryl ester and ³H-free cholesterol was intravenously injected, and blood samples were collected over a 24-h period to determine the fractional clearance rate of the labels by compartmental analysis. The transfer of free and esterified cholesterol, triglycerides and phospholipids from the LDE to HDL was measured by the incubation of the LDE with plasma and radioactivity counting of the supernatant after chemical precipitation of non-HDL fractions. RESULTS: The levels of LDL, non-HDL and HDL cholesterol, triglycerides, apo A1 and apo B were equal in both groups. The 14C-esterified cholesterol fractional clearance rate was not different between glucose-intolerant and control patients, but the ³H-free-cholesterol fractional clearance rate was greater in glucose-intolerant patients than in control patients. The lipid transfer to HDL was equal in both groups. CONCLUSION: In these glucose-intolerant patients with normal plasma lipids, a faster removal of LDE free cholesterol was the only lipid metabolic alteration detected in our study. This finding suggests that the dissociation of free cholesterol from lipoprotein particles occurs in normolipidemic glucose intolerance and may participate in atherogenic signaling.

A semi-supervised classification technique based on interacting forces

Semi-supervised learning is a classification paradigm in which just a few labeled instances are available for the training process. To overcome this small amount of initial label information, the information provided by the unlabeled instances is also considered. In this paper, we propose a nature-inspired semi-supervised learning technique based on attraction forces. Instances are represented as points in a k-dimensional space, and the movement of data points is modeled as a dynamical system. As the system runs, data items with the same label cooperate with each other, and data items with different labels compete among them to attract unlabeled points by applying a specific force function. In this way, all unlabeled data items can be classified when the system reaches its stable state. Stability analysis for the proposed dynamical system is performed and some heuristics are proposed for parameter setting. Simulation results show that the proposed technique achieves good classification results on artificial data sets and is comparable to well-known semi-supervised techniques using benchmark data sets.

QuMinS: fast and scalable querying, mining and summarizing multi-modal databases

Given a large image set, in which very few images have labels, how to guess labels for the remaining majority? How to spot images that need brand new labels different from the predefined ones? How to summarize these data to route the user’s attention to what really matters? Here we answer all these questions. Specifically, we propose QuMinS, a fast, scalable solution to two problems: (i) Low-labor labeling (LLL) – given an image set, very few images have labels, find the most appropriate labels for the rest; and (ii) Mining and attention routing – in the same setting, find clusters, the top-'N IND.O' outlier images, and the 'N IND.R' images that best represent the data. Experiments on satellite images spanning up to 2.25 GB show that, contrasting to the state-of-the-art labeling techniques, QuMinS scales linearly on the data size, being up to 40 times faster than top competitors (GCap), still achieving better or equal accuracy, it spots images that potentially require unpredicted labels, and it works even with tiny initial label sets, i.e., nearly five examples. We also report a case study of our method’s practical usage to show that QuMinS is a viable tool for automatic coffee crop detection from remote sensing images.

Sex Differences in Memory and Other Cognitive Abilities

The aim of the present thesis was to study sex differences in memory and other cognitive bilities in healthy adults. In Study I, participants performed a number of episodic memory tasks that were more or less verbal in nature. Results showed that women performed on a higher level than did men in the episodic memory tasks where it was possible to use verbal labels, whereas men performed on a higher level than did women in a visuospatial episodic memory task. In Study II, women’s advantage in face recognition was investigated.Results showed that women performed at a higher level than did men only in the recognition of other women’s faces. In Study III, sex differences in cognitive tasks as well as brain measures were investigated in healthy older adults. Results showed that only the sex differences in a motor task could, to some extent, be explained by sex differences in one of the brain measures. The findings, as well as possible explanations for these patterns of results, are discussed in a theoretical context.

Luminescent Systems for Genomics and Proteomics

The present PhD project was focused on the development of new tools and methods for luminescence-based techniques. In particular, the ultimate goal was to present substantial improvements to the currently available technologies for both research and diagnostic in the fields of biology, proteomics and genomics. Different aspects and problems were investigated, requiring different strategies and approaches. The whole work was thus divided into separate chapters, each based on the study of one specific aspect of luminescence: Chemiluminescence, Fluorescence and Electrochemiluminescence. CHAPTER 1, Chemiluminescence The work on luminol-enhancer solution lead to a new luminol solution formulation with 1 order of magnitude lower detection limit for HRP. This technology was patented with Cyanagen brand and is now sold worldwide for Western Blot and ELISA applications. CHAPTER 2, Fluorescescence The work on dyed-doped silica nanoparticles is marking a new milestone in the development of nanotechnologies for biological applications. While the project is still in progress, preliminary studies on model structures are leading to very promising results. The improved brightness of these nano-sized objects, their simple synthesis and handling, their low toxicity will soon turn them, we strongly believe, into a new generation of fluorescent labels for many applications. CHAPTER 3, Electrochemiluminescence The work on electrochemiluminescence produced interesting results that can potentially turn into great improvements from an analytical point of view. Ru(bpy)3 derivatives were employed both for on-chip microarray (Chapter 3.1) and for microscopic imaging applications (Chapter 3.2). The development of these new techniques is still under investigation, but the obtained results confirm the possibility to achieve the final goal. Furthermore the development of new ECL-active species (Chapter 3.3, 3.4, 3.5) and their use in these applications can significantly improve overall performances, thus helping to spread ECL as powerful analytical tool for routinary techniques. To conclude, the results obtained are of strong value to largely increase the sensitivity of luminescence techniques, thus fulfilling the expectation we had at the beginning of this research work.

Kernel Methods for Tree Structured Data

Machine learning comprises a series of techniques for automatic extraction of meaningful information from large collections of noisy data. In many real world applications, data is naturally represented in structured form. Since traditional methods in machine learning deal with vectorial information, they require an a priori form of preprocessing. Among all the learning techniques for dealing with structured data, kernel methods are recognized to have a strong theoretical background and to be effective approaches. They do not require an explicit vectorial representation of the data in terms of features, but rely on a measure of similarity between any pair of objects of a domain, the kernel function. Designing fast and good kernel functions is a challenging problem. In the case of tree structured data two issues become relevant: kernel for trees should not be sparse and should be fast to compute. The sparsity problem arises when, given a dataset and a kernel function, most structures of the dataset are completely dissimilar to one another. In those cases the classifier has too few information for making correct predictions on unseen data. In fact, it tends to produce a discriminating function behaving as the nearest neighbour rule. Sparsity is likely to arise for some standard tree kernel functions, such as the subtree and subset tree kernel, when they are applied to datasets with node labels belonging to a large domain. A second drawback of using tree kernels is the time complexity required both in learning and classification phases. Such a complexity can sometimes prevents the kernel application in scenarios involving large amount of data. This thesis proposes three contributions for resolving the above issues of kernel for trees. A first contribution aims at creating kernel functions which adapt to the statistical properties of the dataset, thus reducing its sparsity with respect to traditional tree kernel functions. Specifically, we propose to encode the input trees by an algorithm able to project the data onto a lower dimensional space with the property that similar structures are mapped similarly. By building kernel functions on the lower dimensional representation, we are able to perform inexact matchings between different inputs in the original space. A second contribution is the proposal of a novel kernel function based on the convolution kernel framework. Convolution kernel measures the similarity of two objects in terms of the similarities of their subparts. Most convolution kernels are based on counting the number of shared substructures, partially discarding information about their position in the original structure. The kernel function we propose is, instead, especially focused on this aspect. A third contribution is devoted at reducing the computational burden related to the calculation of a kernel function between a tree and a forest of trees, which is a typical operation in the classification phase and, for some algorithms, also in the learning phase. We propose a general methodology applicable to convolution kernels. Moreover, we show an instantiation of our technique when kernels such as the subtree and subset tree kernels are employed. In those cases, Direct Acyclic Graphs can be used to compactly represent shared substructures in different trees, thus reducing the computational burden and storage requirements.

Anwendung der konfokalen Laser-Scanning-Mikroskopie zur Charakterisierung der Struktur von Zähnen und von oberflächenmodifiziertem Titan für Knochenimplantate

Zusammenfassung Mittels Fluoreszenzfarbstoffen können Strukturen sichtbar gemacht werden, die auf kon-ventionellem Weg nicht, oder nur schwer darzustellen sind. Besonders in Kombination mit der Konfokalen Laser Scanning Mikroskopie eröffnen sich neue Wege zum spezifischen Nachweis unterschiedlichster Komponenten biologischer Proben und gegebenenfalls deren dreidimensionale Widergabe.Die Visualisierung des Proteinanteils des Zahnhartgewebes kann mit Hilfe chemisch kopplungsfähiger Fluorochrome durchgeführt werden. Um zu zeigen, daß es sich bei dieser Markierung nicht um unspezifische Adsorption des Farbstoffes handelt, wurde zur Kontrolle die Proteinkomponente der Zahnproben durch enzymatischen Verdau beseitigt. Derartig behandelte Präparate wiesen eine sehr geringe Anfärbbarkeit auf.Weiterführend diente diese enzymatische Methode als Negativkontrolle zum Nachweis der Odontoblastenfortsätze im Dentin bzw. im Bereich der Schmelz-Dentin-Grenze. Hiermit konnte differenziert werden zwischen reinen Reflexionsbildern der Dentinkanäle und den Zellausläufern deren Membranen gezielt durch lipophile Fluoreszenzfarbstoffe markiert wurden.In einem weiteren Ansatz konnte gezeigt werden, daß reduzierte und daher nichtfluoreszente Fluoresceinabkömmlinge geeignet sind, die Penetration von Oxidationsmitteln (hier H2O2) in den Zahn nachzuweisen. Durch Oxidation dieser Verbindungen werden fluoreszierende Produkte generiert, die den Nachweis lieferten, daß die als Zahnbleichmittel eingesetzten Mittel rasch durch Schmelz und Dentin bis in die Pulpahöhle gelangen können.Die Abhängigkeit der Fluoreszenz bestimmter Fluorochrome von deren chemischer Um-gebung, im vorliegenden Fall dem pH-Wert, sollte eingesetzt werden, um den Säuregrad im Zahninneren fluoreszenzmikroskopisch darzustellen. Hierbei wurde versucht, ein ratio-metrisches Verfahren zu entwickeln, mit dem die pH-Bestimmung unter Verwendung eines pH-abhängigen und eines pH-unabhängigen Fluorochroms erfolgt. Diese Methode konnte nicht für diese spezielle Anwendung verifiziert werden, da Neutralisationseffekte der mineralischen Zahnsubstanz (Hydroxylapatit) die pH-Verteilung innerhalb der Probe beeinflußen. Fluoreszenztechniken wurden ebenfalls ergänzend eingesetzt zur Charakterisierung von kovalent modifizierten Implantatoberflächen. Die, durch Silanisierung von Titantestkörpern mit Triethoxyaminopropylsilan eingeführten freien Aminogruppen konnten qualitativ durch den Einsatz eines aminspezifischen Farbstoffes identifiziert werden. Diese Art der Funktionalisierung dient dem Zweck, Implantatoberflächen durch chemische Kopplung adhäsionsvermittelnder Proteine bzw. Peptide dem Einheilungsprozeß von Implantaten in den Knochen zugänglicher zu machen, indem knochenbildende Zellen zu verbessertem Anwachsverhalten stimuliert werden. Die Zellzahlbestimmung im Adhäsionstest wurde ebenfalls mittels Fluoreszenzfarbstoffen durchgeführt und lieferte Ergebnisse, die belegen, daß die durchgeführte Modifizierung einen günstigen Einfluß auf die Zelladhäsion besitzt.

Konformationsanalyse und Lipidbindung am rekombinanten Lichtsammlerprotein LHCIIb höherer Pflanzen

Die Struktur des Haupt-Lichtsammlerproteins II (LHCIIb) höherer Pflanzen ist aufgrund kristallographischer Strukturanalysen zu 94% aufgeklärt. Dennoch ist es bislang nicht gelungen, die aminoterminale Region des Komplexes vollständig zu lokalisieren. In einem ersten Abschnitt dieser Dissertation sollte anhand einer vergleichenden Bindungsstudie mit Hilfe von in vitro - Rekonstitutionen des LHCIIb geklärt werden, ob es sich bei dem so genannten N - terminalen Trimerisierungsmotiv des Lichtsammlerproteins um eine Interaktionsstelle mit dem Phospholipid Phosphatidylglyzerin handelt. Dazu wurden mehrere vergleichende Lipidbindungsstudien an rekombinantem Wildtyp - Protein und verschiedenen LHCIIb - Trimerisierungsmotiv - Mutanten durchgeführt, die allerdings nicht zu reproduzierbaren Ergebnissen führten.Im zweiten Teil dieser Arbeit wurden intra- und intermolekulare Distanzmessungen an rekombinantem LHCIIb mit Hilfe der Elektronenspin - Resonanz - Spektroskopie durchgeführt. Dazu wurden zweifach mit einem Spin - Label markierte LHCIIb - Monomere und Trimere mit je einer Markierungsposition pro Monomer benutzt. Im Anschluss an die Messungen wurden die erhaltenen Distanzinformationen zusammen mit den bereits zugänglichen Kristallstrukturdaten des Komplexes für eine Modellierung der aminoterminalen Region des LHCIIb verwendet. Die resultierenden Modelle lassen den Schluss zu, dass es im LHCIIb - Trimer zu konformativen Restriktionen des Aminoterminus kommt. Dem entgegen findet man eine größere konformative Diversität in den vermessenen monomeren Komplexen.

Photoreactive polyorganosiloxane nanoparticles and the fabrication of photocleavable microcapsules

In this thesis, we have presented the preparation of highly crosslinked spherical photoreactive colloidal particles of radius about 10 nm based on the monomer trimethoxysilane. These particles are labeled chemically with two different dye systems (coumarin, cinnamate) which are known to show reversible photodimerization. By analyzing the change in particle size upon UV irradiation with dynamic light scattering, we could demonstrate that the partially reversible photoreaction in principle can be utilized to control increase and decrease of colloidal clusters. Here, selection of the appropriate wavelengths during the irradiation employing suitable optical filters proved to be very important. Next, we showed how photocrosslinking of our nanoparticles within the micrometer-sized thin oil shell of water-oil-water emulsion droplets leads to a new species of optically addressable microcontainers. The inner water droplet of these emulsions may contain drugs, dyes or other water-soluble components, leading to filled containers. Thickness, mechanical stability and light resistance of the container walls can be controlled in a simple way by the amount and adjustable photoreactivity (= No. of labels/particle) of the nanoparticles. Importantly, the chemical bonds between the nanoparticles constituting the microcapsule shell can be cleaved photochemically by irradiation with uv light. An additional major advantage is that filling our microcapsules with water-soluble substrate molecules is extremely simple using a solution of the guest molecules as inner water phase of the W/O/W-emulsion. This optically controlled destruction of our microcontainers thus opens up a pathway to controlled release of the enclosed components as illustrated by the example of enclosed cyclodextrin molecules.

Supramolecular approaches to organized luminescent nanostructures for sensing, labeling and imaging applications

The common thread of this thesis is the will of investigating properties and behavior of assemblies. Groups of objects display peculiar properties, which can be very far from the simple sum of respective components’ properties. This is truer, the smaller is inter-objects distance, i.e. the higher is their density, and the smaller is the container size. “Confinement” is in fact a key concept in many topics explored and here reported. It can be conceived as a spatial limitation, that yet gives origin to unexpected processes and phenomena based on inter-objects communication. Such phenomena eventually result in “non-linear properties”, responsible for the low predictability of large assemblies. Chapter 1 provides two insights on surface chemistry, namely (i) on a supramolecular assembly based on orthogonal forces, and (ii) on selective and sensitive fluorescent sensing in thin polymeric film. In chapters 2 to 4 confinement of molecules plays a major role. Most of the work focuses on FRET within core-shell nanoparticles, investigated both through a simulation model and through experiments. Exciting results of great applicative interest are drawn, such as a method of tuning emission wavelength at constant excitation, and a way of overcoming self-quenching processes by setting up a competitive deactivation channel. We envisage applications of these materials as labels for multiplexing analysis, and in all fields of fluorescence imaging, where brightness coupled with biocompatibility and water solubility is required. Adducts of nanoparticles and molecular photoswitches are investigated in the context of superresolution techniques for fluorescence microscopy. In chapter 5 a method is proposed to prepare a library of functionalized Pluronic F127, which gives access to a twofold “smart” nanomaterial, namely both (i)luminescent and (ii)surface-functionalized SCSSNPs. Focus shifts in chapter 6 to confinement effects in an upper size scale. Moving from nanometers to micrometers, we investigate the interplay between microparticles flowing in microchannels where a constriction affects at very long ranges structure and dynamics of the colloidal paste.

Synthetic strategies of new molecular paramagnetic mechanically-interlocked complexes

Supramolecular chemistry is a multidisciplinary field which impinges on other disciplines, focusing on the systems made up of a discrete number of assembled molecular subunits. The forces responsible for the spatial organization are intermolecular reversible interactions. The supramolecular architectures I was interested in are Rotaxanes, mechanically-interlocked architectures consisting of a "dumbbell shaped molecule", threaded through a "macrocycle" where the stoppers at the end of the dumbbell prevent disassociation of components and catenanes, two or more interlocked macrocycles which cannot be separated without breaking the covalent bonds. The aim is to introduce one or more paramagnetic units to use the ESR spectroscopy to investigate complexation properties of these systems cause this technique works in the same time scale of supramolecular assemblies. Chapter 1 underlines the main concepts upon which supramolecular chemistry is based, clarifying the nature of supramolecular interactions and the principles of host-guest chemistry. In chapter 2 it is pointed out the use of ESR spectroscopy to investigate the properties of organic non-covalent assemblies in liquid solution by spin labels and spin probes. The chapter 3 deals with the synthesis of a new class of p-electron-deficient tetracationic cyclophane ring, carrying one or two paramagnetic side-arms based on 2,2,6,6-tetramethylpiperidine-N-oxyl (TEMPO) moiety. In the chapter 4, the Huisgen 1,3-dipolar cycloaddition is exploited to synthesize rotaxanes having paramagnetic cyclodextrins as wheels. In the chapter 5, the catalysis of Huisgen’s cycloaddition by CB[6] is exploited to synthesize paramagnetic CB[6]-based [3]-rotaxanes. In the chapter 6 I reported the first preliminary studies of Actinoid series as a new class of templates in catenanes’ synthesis. Being f-block elements, so having the property of expanding the valence state, they constitute promising candidates as chemical templates offering the possibility to create a complex with coordination number beyond 6.

Synthesis and surface modification of semiconductor nanocrystals

The last decade has witnessed an exponential growth of activities in the field of nanoscience and nanotechnology worldwide, driven both by the excitement of understanding new science and by the potential hope for applications and economic impacts. The largest activity in this field up to date has been in the synthesis and characterization of new materials consisting of particles with dimensions in the order of a few nanometers, so-called nanocrystalline materials. [1-8] Semiconductor nanomaterials such as III/V or II/VI compound semiconductors exhibit strong quantum confinement behavior in the size range from 1 to 10 nm. Therefore, preparation of high quality semiconductor nanocrystals has been a challenge for synthetic chemists, leading to the recent rapid progress in delivering a wide variety of semiconducting nanomaterials. Semiconductor nanocrystals, also called quantum dots, possess physical properties distinctly different from those of the bulk material. Typically, in the size range from 1 to 10 nm, when the particle size is changed, the band gap between the valence and the conduction band will change, too. In a simple approximation a particle in a box model has been used to describe the phenomenon[9]: at nanoscale dimensions the degenerate energy states of a semiconductor separate into discrete states and the system behaves like one big molecule. The size-dependent transformation of the energy levels of the particles is called “quantum size-effect”. Quantum confinement of both the electron and hole in all three dimensions leads to an increase in the effective bandgap of the material with decreasing crystallite size. Consequently, both the optical absorption and emission of semiconductor nanaocrystals shift to the blue (higher energies) as the size of the particles gets smaller. This color tuning is well documented for CdSe nanocrystals whose absorption and emission covers almost the whole visible spectral range. As particle sizes become smaller the ratio of surface atoms to those in the interior increases, which has a strong impact on particle properties, too. Prominent examples are the low melting point [8] and size/shape dependent pressure resistance [10] of semiconductor nanocrystals. Given the size dependence of particle properties, chemists and material scientists now have the unique opportunity to change the electronic and chemical properties of a material by simply controlling the particle size. In particular, CdSe nanocrystals have been widely investigated. Mainly due to their size-dependent optoelectronic properties [11, 12] and flexible chemical processibility [13], they have played a distinguished role for a number of seminal studies [11, 12, 14, 15]. Potential technical applications have been discussed, too. [8, 16-27] Improvement of the optoelectronic properties of semiconductor nanocrystals is still a prominent research topic. One of the most important approaches is fabricating composite type-I core-shell structures which exhibit improved properties, making them attractive from both a fundamental and a practical point of view. Overcoating of nanocrystallites with higher band gap inorganic materials has been shown to increase the photoluminescence quantum yields by eliminating surface nonradiative recombination sites. [28] Particles passivated with inorganic shells are more robust than nanocrystals covered by organic ligands only and have greater tolerance to processing conditions necessary for incorporation into solid state structures or for other applications. Some examples of core-shell nanocrystals reported earlier include CdS on CdSe [29], CdSe on CdS, [30], ZnS on CdS, [31] ZnS on CdSe[28, 32], ZnSe on CdSe [33] and CdS/HgS/CdS [34]. The characterization and preparation of a new core-shell structure, CdSe nanocrystals overcoated by different shells (CdS, ZnS), is presented in chapter 4. Type-I core-shell structures as mentioned above greatly improve the photoluminescence quantum yield and chemical and photochemical stability of nanocrystals. The emission wavelengths of type-I core/shell nanocrystals typically only shows a small red-shift when compared to the plain core nanocrystals. [30, 31, 35] In contrast to type-I core-shell nanocrystals, only few studies have been conducted on colloidal type-II core/shell structures [36-38] which are characterized by a staggered alignment of conduction and valence bands giving rise to a broad tunability of absorption and emission wavelengths, as was shown for CdTe/CdSe core-shell nanocrystals. [36] The emission of type-II core/shell nanocrystals mainly originates from the radiative recombination of electron-hole pairs across the core-shell interface leading to a long photoluminescence lifetime. Type-II core/shell nanocrystals are promising with respect to photoconduction or photovoltaic applications as has been discussed in the literature.[39] Novel type-II core-shell structures with ZnTe cores are reported in chapter 5. The recent progress in the shape control of semiconductor nanocrystals opens new fields of applications. For instance, rod shaped CdSe nanocrystals can enhance the photo-electro conversion efficiency of photovoltaic cells, [40, 41] and also allow for polarized emission in light emitting diodes. [42, 43] Shape control of anisotropic nanocrystals can be achieved by the use of surfactants, [44, 45] regular or inverse micelles as regulating agents, [46, 47] electrochemical processes, [48] template-assisted [49, 50] and solution-liquid-solution (SLS) growth mechnism. [51-53] Recently, formation of various CdSe nanocrystal shapes has been reported by the groups of Alivisatos [54] and Peng, [55] respectively. Furthermore, it has been reported by the group of Prasad [56] that noble metal nanoparticles can induce anisotropic growth of CdSe nanocrystals at lower temperatures than typically used in other methods for preparing anisotropic CdSe structures. Although several approaches for anisotropic crystal growth have been reported by now, developing new synthetic methods for the shape control of colloidal semiconductor nanocrystals remains an important goal. Accordingly, we have attempted to utilize a crystal phase control approach for the controllable synthesis of colloidal ZnE/CdSe (E = S, Se, Te) heterostructures in a variety of morphologies. The complex heterostructures obtained are presented in chapter 6. The unique optical properties of nanocrystals make them appealing as in vivo and in vitro fluorophores in a variety of biological and chemical investigations, in which traditional fluorescence labels based on organic molecules fall short of providing long-term stability and simultaneous detection of multiple emission colours [References]. The ability to prepare water soluble nanocrystals with high stability and quantum yield has led to promising applications in cellular labeling, [57, 58] deep-tissue imaging, [59, 60] and assay labeling [61, 62]. Furthermore, appropriately solubilized nanocrystals have been used as donors in fluorescence resonance energy transfer (FRET) couples. [63-65] Despite recent progress, much work still needs to be done to achieve reproducible and robust surface functionalization and develop flexible (bio-) conjugation techniques. Based on multi-shell CdSe nanocrystals, several new solubilization and ligand exchange protocols have been developed which are presented in chapter 7. The organization of this thesis is as follows: A short overview describing synthesis and properties of CdSe nanocrystals is given in chapter 2. Chapter 3 is the experimental part providing some background information about the optical and analytical methods used in this thesis. The following chapters report the results of this work: synthesis and characterization of type-I multi-shell and type-II core/shell nanocrystals are described in chapter 4 and chapter 5, respectively. In chapter 6, a high–yield synthesis of various CdSe architectures by crystal phase control is reported. Experiments about surface modification of nanocrystals are described in chapter 7. At last, a short summary of the results is given in chapter 8.