Grains on the surfaces of airless bodies are continually being modified due to their exposure to interplanetary space, a phenomenon known as space weathering. This dissertation uses a multi-faceted approach to understanding space weathering of the lunar and asteroidal surfaces. Chapters 1 and 2 provide an introduction to space weathering and a discussion of the methods employed in this work, respectively. Chapter 3 focuses on the analysis of returned samples from near-Earth asteroid Itokawa using the transmission electron microscope (TEM) and contributes to the first-ever comparison of microstructural and chemical features of space weathering in returned samples from two different airless bodies. This research uses high-resolution imaging and quantitative energy-dispersive x-ray spectroscopy (EDS) measurements to analyze space weathering characteristics in an Itokawa soil grain. These analyses confirm that space weathering is operating on the surface of Itokawa, and that many of the resulting features have similarities to those observed in lunar soils. Results show that while there is evidence that both major constituent space weathering processes are operating on the surface of Itokawa, solar wind irradiation, not micrometeorite impacts, appears to be the dominant contributor to changes in the microstructure and chemistry of surface material. Chapter 4 presents a detailed study of nanophase Fe (npFe) particles in lunar soil samples. For the first-time, the oxidation state of individual npFe particles was directly measured using electron energy-loss spectroscopy (EELS) in the TEM. The results show that npFe particles are oxidizing over their time on the lunar surface, and that the amount of oxidized Fe in the nanoparticles is correlated with soil maturity. The EELS data are also coupled to atomic-resolution imaging, which is used to determine the structure of the nanoparticles, confirming their mineral phase. This work challenges the long-standing paradigm that all npFe particles are composed of metallic Fe and that the chemical composition of these features remains static after their formation. A theoretical modeling investigation of the influence that npFe particles of different oxidation states have on the spectral properties of the material is also presented. The model results show that varied Fe-oxidation states of the nanoparticles can produce subtle changes in the optical properties of the soils, including the degree of reddening and the attenuation of characteristic absorption bands. These findings should be accounted for in future modeling of reflectance spectra. Chapter 5 presents a novel technique for simulating space weathering processes inside the TEM. Using an in situ heating holder, lunar soils were subjected to both slow- (~minutes) and rapid-heating (<seconds) events to simulate micrometeorite impacts. The slow-heating experiments show that npFe forms at ~575 ºC, providing a temperature constraint on initial npFe formation. Lunar soil grains that were subjected to a single, rapid, thermal pulse show the development of npFe particles and vesiculated textures near the grain rim. The vesicles were imaged and the npFe particles were imaged and then mapped with EDS. The oxidation state of the npFe particles was confirmed to be Fe^0 using EELS. Several lunar soil grains were subjected to multiple thermal shocks to simulate longer exposure times on the lunar surface. With each heating cycle, the number and size distribution of the npFe particles changed. The average size of npFe particles increased, and the size distribution became more gaussian after multiple heating events, versus the asymmetric distribution present after only one heating event. These results provide insight into the particle growth dynamics for space weathered soils and could offer a new way to place relative age constraints on grains in lunar soil.Chapter 6 provides a summary of the work presented here, discusses its implications for understanding space weathering processes across the solar system, and presents a perspective on the future of space weathering studies.

Atualmente, a demanda e a necessidade de metodologias analíticas e bioanalíticas que promovam análise rápida e seletiva tem impulsionado o constante desenvolvimento de sorventes para preparo de amostra e fases estacionárias para cromatografia líquida de alta eficiência (HPLC). Desta maneira, esta tese tem por objetivo desenvolver materiais utilizando precursores de sílica com aplicação em preparo de amostra e suportes cromatográficos para cromatografia líquida de ultra alta eficiência (UHPLC). Foram sintetizadas e caracterizadas duas fases extratoras as quais tiveram sua aplicabilidade demonstrada na extração de fármacos em fluidos biológicos empregando a técnica de extração em fase sólida (SPE). A primeira metodologia desenvolvida foi a síntese de partículas de sílica empregando precursores de baixo custo (silicato de sódio) pelo método sil-gel. Após a funcionalização com grupamento C18, e caracterização da fase extratora, o material sintetizado foi aplicado em SPE off-line para a determinação de fluoxetina e seu metabólito, norfluoxetina, em plasma humano por HPLC-UV. O método desenvolvido foi validado e aplicado em amostra de plasma de pacientes sob tratamento de fluoxetina. Além disso, o sorvente desenvolvido apresentou resultados similares quando comparado com fases extratoras comerciais. O segundo método desenvolvido utilizou a técnica de impressão molecular (MIP) a qual tem se tornado uma fase extratora atrativa devido a sua maior seletividade em relação as fases típicas empregadas em SPE. Ao mesmo tempo, o uso de SPE online com MIP (MISPE) é uma alternativa atraente no preparo de amostra, pois é simples, diminui o tempo total da análise, necessita de pequena quantidade de amostra, e pode ser automatizado. Assim, MIP amino-funcionalizado pelo processo sol-gel utilizando ibuprofeno como template foi sintetizado. Para comparação, o polímero não impresso (NIP) foi preparado nas mesmas condições, sem a presença do template. As análises por cromatografia líquida foram realizadas utilizando uma coluna de extração MIP e uma coluna analítica em fase reversa configuradas para operar no modo SPE online por column switching. O método desenvolvido, MISPE-HPLC-UV, foi validado e aplicado na extração seletiva de cinco anti-inflamatórios não esteroidais (NSAIDs) em urina humana com um tempo total de análise de 22 minutos. Além dos materiais voltados para preparo de amostra, foi avaliado a síntese e caracterização de suporte cromatográfico baseado em sílica com diâmetro de partícula abaixo de 2 &micro;m com aplicabilidade em análises rápidas em cromatografia (UHPLC). Nesta vertente, partículas de sílica esférica não porosa (0,9 &micro;m) foram sintetizadas pelo método sol-gel com DPR abaixo de 10 %. As partículas foram funcionalizadas com o grupamento C18 (ODS) e submetidas posteriormente ao processo de endcapping (TMS) para operar em modo reverso. Métodos físico-químicos de caracterização foram utilizados para determinar a morfologia, área superficial e quantidade de carbono na superfície do material. Adicionalmente, as técnicas de infravermelho e ressonância magnética nuclear forneceram informações sobre a estrutura da sílica e das ligações após a funcionalização. A fase estacionária foi empacotada em uma coluna (40 mm x 0,25 mm) e a sua aplicabilidade foi avaliada em cromatografia líquida capilar (cLC). Apesar de não alcançar a eficiência ótima da separação cromatográfica devido a limitação instrumental, a fase estacionária sub 1 &micro;m apresenta potencial para separações rápidas sem perda de eficiência. / Currently, the demand and need for bioanalytical and analytical methodologies that promotes rapid and selective analysis has stimulated the development of sorbents for sample preparation and stationary phases for high performance liquid chromatography (HPLC). Therefore, this thesis aims to develop materials using silica precursors with application in sample preparation and chromatographic supports for ultra high performance liquid chromatography (UHPLC). Thus, in this study two sorbents were synthesized and characterized and their applicability was demonstrated in the extraction of drugs in biological fluids employing the solid phase extraction technique (SPE). The first method developed was the synthesis of silica particles employing low cost precursors (sodium silicate) using the sil-gel method. After functionalization with C18 group and characterization of the extraction phase, the synthesized material was applied to SPE off-line for the determination of fluoxetine and its metabolite, norfluoxetine, in human plasma by HPLC-UV. The method was validated and applied to plasma samples from patients treated with fluoxetine. Furthermore, the labmade sorbent presented similar results when compared with commercial ones. The second method developed used the molecular imprinting technique (MIP) that has become an attractive sorbent due to its greater selectivity over the typical phases employed in SPE. At the same time, the use of online SPE with MIP (MISPE) is an attractive alternative in sample preparation because it is simple, reduces the total time of analysis, needs little amount of sample and can be automated. Thus, MIP amino-functionalized by the sol-gel method using ibuprofen as template was synthesized. For comparison, the non-imprinted polymer (NIP) was prepared under the same conditions without the presence of template. LC analysis was performed using a MIP extraction column and a reverse phase analytical column in order to perform column switching on-line sample separation. The developed method, MISPE-HPLC-UV, was validated and applied to the selective extraction of five anti-inflammatory drugs (NSAIDs) in human urine with a total analysis time of 22 minutes. Moreover it was evaluated the synthesis and characterization of chromatographic support based on silica with particle diameter under 2 &micro;m with applicability in fast LC analysis (UHPLC). In this instance, spherical particles of non-porous silica (0.9 &micro;m) were synthesized by sol-gel method with RSD below 10 %. The particles were functionalized with the C18 group and submitted to the endcapping process with TMS to operate in reverse phase mode. Physico-chemical methods were used to determine the morphology, surface area and amount of carbon in the silica surface. Additionally, the infrared and nuclear magnetic resonance techniques provided information about the structure of silica and bonds after the functionalization step. The stationary phase was packed into a column (40 mm x 0.25 mm), and its applicability was evaluated in capillary liquid chromatography (cLC). Although optimum chromatographic efficiency was not achieved separation due to instrumental limitations, the sub 1 &micro;m stationary phase has potential for rapid separations without loss of efficiency.

The aim of this thesis is to systematically study the statistical guarantee for two representative non-convex optimization problems arsing in the statistics community. The first one is the high-dimensional Gaussian mixture model, which is motivated by the estimation of multiple graphical models arising from heterogeneous observations. The second one is the low-rank tensor estimation model, which is motivated by high-dimensional interaction model. Both optimal statistical rates and numerical comparisons are studied in depth. In the first part of my thesis, we consider joint estimation of multiple graphical models arising from heterogeneous and high-dimensional observations. Unlike most previous approaches which assume that the cluster structure is given in advance, an appealing feature of our method is to learn cluster structure while estimating heterogeneous graphical models. This is achieved via a high dimensional version of Expectation Conditional Maximization (ECM) algorithm. A joint graphical lasso penalty is imposed on the conditional maximization step to extract both homogeneity and heterogeneity components across all clusters. Our algorithm is computationally efficient due to fast sparse learning routines and can be implemented without unsupervised learning knowledge. The superior performance of our method is demonstrated by extensive experiments and its application to a Glioblastoma cancer dataset reveals some new insights in understanding the Glioblastoma cancer. In theory, a non-asymptotic error bound is established for the output directly from our high dimensional ECM algorithm, and it consists of two quantities: statistical error (statistical accuracy) and optimization error (computational complexity). Such a result gives a theoretical guideline in terminating our ECM iterations. In the second part of my thesis, we propose a general framework for sparse and low-rank tensor estimation from cubic sketchings. A two-stage non-convex implementation is developed based on sparse tensor decomposition and thresholded gradient descent, which ensures exact recovery in the noiseless case and stable recovery in the noisy case with high probability. The non-asymptotic analysis sheds light on an interplay between optimization error and statistical error. The proposed procedure is shown to be rate-optimal under certain conditions. As a technical by-product, novel high-order concentration inequalities are derived for studying high-moment sub-Gaussian tensors. An interesting tensor formulation illustrates the potential application to high-order interaction pursuit in high-dimensional linear regression

X-­ray nanotomography is an imaging technique used to study three­-dimensional structures at submicrometer length scale. The samples to be studied should be only 10s of micrometers in diameter, and ideally cylindrical. Focused ion beam milling is the most common technique used to prepare samples for nanotomography experiments, and the current technique used at the NanoMAX beamline of the MAX IV synchrotron facility. It is a time­-consuming process as preparing one sample can take hours. With the aim of offering a faster, alternative sample preparation method, a CNC milling setup was developed, and is presented here. The CNC setup is based on two spindles placed on precision linear stages - one for the sample and one for the milling tool. The sample is rotated while being trimmed gently by the milling tool, resulting in a small sample cylinder. A Python script generating G­-code commands controls the procedure. The setup was used to trim copper samples down to 18.5 micrometers. Further work is needed to optimize milling parameters in order to reach similar diameters for other sample materials. The developed setup offers a time-­efficient, repeatable and low­-cost sample preparation method for X-­ray nanotomography. / Nanotomografi med röntgenstrålning är en metod som används för att urskilja och studera tre­-dimensionella strukturer i material av storleksordningar mindre än 100 nanometer. Proverna behöver vara små; endast tiotals mikrometer i diameter. Det ideala provet för tomografi är cylindriskt. Fokuserad jonstrålefräsning är en vanlig metod för att preparera prover för nanotomografi, och är den metod som idag används på NanoMAX vid synkrotronljusanläggningen MAX IV. Det är en tidskrävande process; att förbereda ett enda prov tar flera timmar. Med avsikt att erbjuda en snabbare, alternativ metod för provpreparering så utvecklades ett CNC-styrt frässystem, som presenteras här. Systemet är baserat på två spolar placerade på precisionsplattformar - en för själva provet och en för fräsverktyget. Ett Python-program utvecklades som skickar kommandon i G-­kod och därigenom kontrollerar precisionsplattformarnas rörelser. Provet trimmas till en viss diameter, och slutresultatet är ett nästintill cylindriskt prov som kan användas för nanotomografi. Den minsta diameter som nåddes var 18.5 mikrometer, för ett kopparprov. Ytterligare arbete krävs för att optimera parametrar för fräsproceduren för andra typer av material, för att lika små diametrar ska kunna nås. Systemet som utvecklats erbjuder en snabb, billig och repeterbar metod för att förbereda prover för nanotomografi med röntgenstrålning.

The mean absolute deviation is a simple and informative measure of variability. Approximate confidence intervals for mean absolute deviations in one-group and two-group designs are derived and are shown to have excellent small-sample properties under moderate nonnormality. Sample size planning formulas are derived.

Several large-sample confidence intervals for the ratio of independent binomial proportions are compared in terms of exact coverage probability and width. A non-iterative approximate Bayesian interval is derived and its frequency properties are superior to all of the non-iterative confidence intervals considered. The approximate Bayesian interval, which is very easy to compute, has performance characteristics that are very similar to the computationally intensive score method. Two sample size determination formulas are presented, one for desired absolute precision and the other for desired relative precision.

The small-sample bias and root mean squared error of several distribution-free estimators of the variance of the sample median are examined. A new estimator is proposed that is easy to compute and tends to have the smallest bias and root mean squared error.

Solid Phase Microextraction (SPME) associated with Microwave Assisted Extraction (MAE)^(1-3) , on-line headspace derivatization ^(4-6) and the selectivity of different SPME coatings ⁽⁷⁾ were studied. Trace amounts of Veltol ®, Veltol Plus® and short chain fatty acids in food samples were analyzed by GC/MS. Since SPME is not directly applicable to solid samples, SPME associated with MAE techniques was studied for solids, primarily food samples. The efficiency of classical solvent extraction and MAE was compared. The parameters which affect SPME, were optimized for the determination of Veltol® and Veltol Plus® in food products such as potato chips and coffee. The technique gave a detection limit of 2 ppb for Veltol Plus ® which is 200 times more sensitive than conventional GC technique. Headspace injection is characterized by simple and easy handling of complicated solid and solution matrices. Headspace injection, however, is not suitable for high molecular weight substances or non-volatile compounds. An on-line derivatization headspace technique was studied for short chain fatty acids. These samples are difficult to do by classical GC. The developed technique simplified the conventional derivatization procedures and combined the sample preparation and GC/MS analysis into one step. The thermostatting temperature, time, solvent and matrix effects were investigated. Low calorie fat and some agricultural samples were analyzed. The detection limit for acetic acid is 8 ppb. SPME is a novel sample introduction technique. The behavior of di(methylsiloxane), polyacrylate and Carbowax coatings on SPME fibers for compounds having different functional groups were investigated. The selectivities of the coating, sample pH and the sample temperature were investigated. / Ph. D.

The purpose of this study was to develop a test which can be used to compare the reliability performances of two types of equipment components to determine whether or not the new component satisfies a given feasibility criterion. Two types of tests were presented and compared: the fixed sample size test and the truncated sequential probability ratio test. Both of these tests involve use of a statistic which is approximately distributed as F. This study showed that the truncated sequential probability ratio test has good potential as a means of comparing two component types to see whether or not the reliability of the new component is at least a certain number of times greater than the reliability of the old component.

Memory-trace theories of remembering suggest that performance in delayed matching-to-sample (DMTS) procedures depends on a memory trace that degrades with time. By contrast, the theory of direct remembering suggests that increasing the delay between sample and comparison stimuli in DMTS procedures is functionally equivalent to decreasing the disparity between sample stimuli. The present dissertation tested this assumption by assessing the degree to which changes in the frequency of reinforcement for correct choices biased the distribution of choice responses in a conditional-discrimination procedure. Seven pigeons responded under a temporal-discrimination procedure in which temporal sample-stimuli were categorized as being of either short or long duration by a response to a corresponding comparison key. In the sample-stimulus disparity condition, the disparity between the sample stimuli (difference between the short and long samples) was manipulated. In the retention-interval condition, the delay between sample offset and presentation of the comparison stimuli was manipulated. Importantly, the same general procedure was used across conditions, facilitating conclusions regarding functional equivalence of the two manipulations. The theory of direct remembering suggests that the relation between sensitivity of behavior to changes in reinforcer frequency and discriminability (accuracy) should be similar in the sample-stimulus disparity and retention-interval condition. The results showed that discriminability decreased with both the sample-stimulus disparity and retention-interval manipulations. Overall estimates of sensitivity were similar to those obtained previously. There was, however, no difference in the estimates of sensitivity as a function of discriminability during either the sample-stimulus disparity or retention-interval conditions; sensitivity was independent of discriminability. These results are in contrast to most previous reports, and are interpreted in terms of the use of temporal-sample stimuli in the current experiment. Further analyses of the choice-response data showed that the effects of variation in reinforcer ratios differed across conditions as a function of trial type and trial difficulty. These results suggest the need for careful consideration of behavioral outcomes at several levels of analysis when assessing functional equivalence of experimental manipulations. The potential benefits and hindrances of characterization of behavioral outcomes in terms of functional equivalence are discussed.