Orientador: João Francisco Escobedo / Resumo: O trabalho apresenta uma análise estatística e climática das irradiações global (HG) e global infravermelha (HGNIR) e das frações global total (Kt = HG/H0) e global infravermelha (KtNIR = HGNIR/H0NIR), transmitidas na atmosfera no período de 2001 a 2006 em Botucatu/SP/Brasil. Os valores das frações médias Kt e KtNIR (horária e diária) calculadas estatisticamente foram: Kth = 50,0%, KtNIRh = 42,0%, e Ktd = 56,0%, KtNIRd = 48,0%, respectivamente. O estudo mostra ainda a obtenção e a validação da equação de estimativa para KtNIR em função de Kt obtidas por regressão linear (y = ax), onde KtNIRh = 0,850Kth e KtNIRd = 0,855Ktd apresentaram elevados coeficientes de determinação: R2 = 0,981 e R2 = 0,987, respectivamente. Os valores horários obtidos na validação (ano típico e atípico) foram (rMBE)h = 1,0% e 0,2%, (rRMSE)h = 6,3% e 5,3%; dh = 0,994 e 0,996 respectivamente, enquanto que os valores diários foram (rMBE)d = 0,9% e 0,05%; (rRMSE)d = 3,4% e 2,6%; dd = 0,995 e 0,998. Os resultados mostram que as equações podem ser utilizadas com precisão e exatidão nas estimativas de KtNIR. A análise sazonal mostrou que a equação KtNIRh = 0,850Kth anual superestima KtNIRh no verão e primavera em 2,8% e 1,0%, e subestima-o no outono e inverno em 1,0% e 2,6%, respectivamente. Similarmente, KtNIRd = 0,855Ktd anual superestima KtNIRd no verão e primavera em 2,8% e 1,3% e subestima-o no outono e inverno em 1,0% e 2,1%, respectivamente. A análise por cobertura de céu mostrou que: o valor KtNIRh an... (Resumo completo, clicar acesso eletrônico abaixo) / Abstract: This paper describes the statistical and climatic analysis of global (HG) and infrared global (HGNIR) irradiations, and total global (Kt = HG/H0) and infrared global (KtNIR = HGNIR/H0NIR) fractions, transmitted in the atmosphere from 2001 to 2006 in Botucatu/SP/Brazil: Values of mean fractions Kt and KtNIR (hourly and daily) calculated statistically were as follows: Kth = 50,0%, Kth NIR = 42,0%, and Ktd = 56,0%, KtNIR d = 48,0%, respectively. The study also presents obtaining and validation of the estimate equation for KtNIR as a function of Kt. The equations obtained by linear regression (y = ax) KtNIR h = 0,850Kth and KtNIR d = 0,855Ktd had high determination coefficient: R2 = 0,981 and 0,987 respectively. Hourly values based on validation (typical and atypical years) were as follows: (rMBE) h = 1,0% and 0,2%, (rMSE) h = 6,3% and 5,3%; d h = 0,994 and 0,996, respectively. Daily values were as follows: (rMBE) d = 0,9% and 0,05%; (rMSE) d = 3,4% and 2,6%; d d = 0,995 and 0,998. The results show that the equations can be used precisely and accurately for KtNIR estimates. Seasonal analysis showed that the annual KtNIR h = 0,850Kth equation overestimates KtNIR h in the summer and spring by 2,8% and 1,0% and underestimates it in the autumn and winter by 1,0% and 2,6%, respectively. Similarly, KtNIR d = 0,855Ktd equation anual overestimates KtNIR d in the summer and spring by 2,8% e 1,3% and underestimates it in the autumn and winter by 1,0% e 2,1%, respectively. Considering sky c... (Complete abstract click electronic access below) / Mestre

The fluid bed granulation is a well-established method how to improve such properties of powders as flowability and increase content uniformity of the tablets. In this thesis, there was evaluated a granulation process on a lab scale fluid Glatt bed granulator and optimized for highest possible yield. Product yield in the size range of 80-90 % of granules and process reproducibility were stated as most effective. The product was analysed for its particle size distribution, the API distribution within the different particle size fractions and the flowability of the final granules. For process optimization, the most critical parameters such as spraying rate, particle size of raw materials and fluid bed pressure were identified and evaluated. As the highest-yielding dosage for the powder binder was found the spraying rate of 9 g/min. Changes in bed fluid pressure and nozzle pressure showed no significant improvement. Different grades of caffeine were compared for their impact on the granulation properties. Sieved caffeine enhanced yield of the product and reproducibility compared to bulk or disagglomerated caffeine.

The selection of a representative sample is a delicate problem when drug seizures comprised of large number of units arrive at the Swedish National Forensic Centre (NFC). If deviating objects in the selected sample size are found, additional analyzes are required to investigate how representative the results are for the entire population. This generates further pressure on operational analysis flow. With the goal to provide a tool which forensic scientists at NFC can base their assessment of the representative nature of the selected sampling of large drug seizures on, this project investigated the possibilities of evaluating the level of homogeneity in drug seizures using near-infrared (NIR) hyperspectral imaging along with principal component analysis (PCA). A total of 27 sample groups (homogeneous, heterogeneous and seized sample groups) were analyzed and different predictive models were developed. The models were either based on quantifying the variation in NIR spectra or in PCA scores plots. It was shown that in the spectral range of 1300-2000 nm, using a pre-processing combination of area normalization, quadratic (second polynomial) detrending and mean centering, promising predictive abilities of the models in their evaluation of the level of homogeneity in drug seizures were achieved. A model where the approximated signal-dependent variation was related to the quotient of significant and noise explained variance given by PCA indicated most promising predictive abilities when quantifying the variation in NIR spectra. Similarly, a model where a rectangular area, defined by the maximum distances along PC1 and PC2, was related to the cumulative explained variance of the two PCs showed most promising predictive abilities when quantifying the variation in PCA scores plots. Different zones for which within sample groups are expected to appear based upon their degree of homogeneity could be established for both models. The two models differed in sensitivity. However, more comprehensive studies are required to evaluate the models applicability from an operational point-of-view.

As a result of the standardizability of SNe Ia light curves over a wide range of photometric bands, they are used as standard candles to accurately measure distances in the cosmos up to z ≈ 1 [22]. As dust extinction is smaller in the NIR than in the optical [21] there is less dispersion seen in the peak brightnesses of SNe Ia, making them truly standard candles. We use SNPY to fit light curves for 192 SNe Ia. The mean of all Hubble residuals of our sample is ≈ 0.101 mag with a standard deviation of ≈ 0.234 mag. After applying an original set of cuts, the mean of 173 Hubble residuals reduces to ≈ 0.080 mag with a standard deviation of 0.203 mag. We next estimate host galaxy stellar masses of 175 SNe. From our sample we detect a 0.039 ± 0.026 mag (1−2σ) mass-step. For reasons outlined in section 4.1.1 and 4.1.2 respectively, we increase our sBV cut to sBV &gt; 0.8 and decrease our extinction cut to E(B −V ) ≤ 0.2 mag to see the mass step disappear entirely (0.004 ± 0.034 mag). Fast-declining SNe occur with preference in high-mass galaxies, possibly pointing to an intrinsic contribution to this mass step [22]. As NIR data is seen to significantly reduce the 3−4σ [14] mass-step detected with optical data, it is concluded that extinction likely plays a large role in the mass-step, as proposed in Brout &amp; Scolnic 2020 [2]. / <p>Presentation given over zoom due to the COVID-19 crisis.</p>

Carbon nanotubes (CNTs) have special characteristics that made them good agents especially for biomedical applications. In this study, Fe3O4 nanoparticles were incorporated onto the surface of CNTs followed by polyethylene glycol (PEG) grafting forming CNTs-Fe3O4-PEG hybrids. The PEGylation improves their biocompatibility, water solubility, and increases blood circulation. CNTs-Fe3O4-PEG was used as T2-contrat agent for magnetic resonance imaging (MRI). In addition, doxorubicin (DOX) was loaded onto CNTs-Fe3O4-PEG. The release of DOX from DOX-loaded CNTs-Fe3O4-PEG was tested under different pH conditions (7.4, 6.3 and 5.2). The release increased at acidic pH compared to neutral pH, which shows the sensitivity of the system to pH change. Triggering the release by Near Infra-Red (NIR) irradiation was furthermore investigated. The release increased after irradiation with NIR compared to control sample. These result prove that the designed system can be triggered by an internal stimuli (pH) and external stimuli (NIR irradiation) making it a promising candidate to be used for biomedical applications.

京都大学 / 0048 / 新制・課程博士 / 博士(農学) / 甲第19019号 / 農博第2097号 / 新制||農||1029(附属図書館) / 学位論文||H27||N4901(農学部図書室) / 31970 / 京都大学大学院農学研究科地域環境科学専攻 / (主査)教授 近藤 直, 教授 清水 浩, 准教授 小川 雄一 / 学位規則第4条第1項該当 / Doctor of Agricultural Science / Kyoto University / DGAM

No description available.

Multiple object tracking (MOT) can be an efficient tool for finding patterns in video monitoring data. In this thesis, we investigate which type of video data works best for MOT in an indoor combat training scenario. The three types of camera data evaluated are color data, near-infrared (NIR) data, and depth data. In order to evaluate which of these lend themselves best for MOT, we develop object tracking models based on YOLOv5 and DeepSORT, and train the models on the respective types of data. In addition to the individual models, ensembles of the three models are also developed, to see if any increase in performance can be gained. The models are evaluated using the well-established MOT evaluation metrics, as well as studying the frame rate performance of each model. The results are rigorously analyzed using statistical significance tests, to ensure only well-supported conclusions are drawn. These evaluations and analyses show mixed results. Regarding the MOT metrics, the performance of most models were not shown to be significantly different from most other models, so while a difference in performance was observed, it cannot be assumed to hold over larger sample sizes. Regarding frame rate, we find that the ensemble models are significantly slower than the individual models on their own.

Spectroscopic methods were established decades ago in a wide variety of fields. This also applies to the pharmaceutical field, although they initially were mostly used for identity testing or structure elucidation only. Technical developments, such as miniaturization (NMR benchtop devices), Fourier transformations (for NMR, MIR spectroscopy) or the combination with chemometric evaluation (e.g., in Process Analytical Technology, PAT), have further increased their importance and opened up new applications. The aim of this work was to investigate further new approaches and to find new applications for already established methods and to show their benefits. By means of MIR, NIR and NMR data and their chemometric evaluation (principal component analysis, PCA; hierarchical cluster analysis, HCA; linear discriminant analysis, LDA), possibilities were presented to successfully determine the manufacturer or the pharmaceutical company of various paracetamol preparations. In the course of this, various similarities and correlations between the preparations of individual companies could also be identified. For this purpose, a suitable sample preparation was developed for each spectroscopic method, and suitable measurement parameters in order to obtain reproducible spectra for the chemometric evaluation were determined. Furthermore, the results of the two unsupervised methods (HCA, PCA) were compared with each other. The HCA was able to confirm those of the PCA for the very most part. Additionally, through these methods it was possible to characterize many of the preparations based on clusters formed by comparable tablet compositions. In order to be able to measure unmortared, whole tablets using the NIR spectrometer, an attachment was developed and manufactured using 3D printing. Its functionality was demonstrated by measuring and analyzing the tablets of two different batches of nine paracetamol preparations. The batches were clearly distinguished on the basis of a PCA and a significant difference was also demonstrated by means of statistical tests. For NMR spectroscopy, a method was developed to obtain optimized "fingerprint" spectra of drug formulations. For this purpose, a 1D DOSY measurement was elaborated, in which the signals of the active ingredient could be filtered out by the appropriate choice of measurement parameters. The chemometric evaluation can thus focus on the remaining signals of the excipients, on the basis of which the preparations of the same API can be distinguished. Especially in the case of formulations that consist largely of active ingredient, data pre processing of the spectra can thus be simplified and greater importance can be assigned to the originally very small excipient signals. A quantitative 1H NMR method was developed for the comparison of a high field spectrometer (400 MHz) with a benchtop spectrometer (80 MHz) for two finished drugs. It was shown that it is possible to obtain comparable results with both instruments, but that the influence of the excipients on the signals and the lower resolution of the benchtop instrument must be taken into account. Therefore, it was not possible to obtain comparable results without further optimization of the method for one of the active ingredients. In the investigation of various reactions between APIs and excipients using DOSY, its usefulness as a screening method in stability testing was demonstrated. For this purpose, three different APIs and excipients were stressed together and the reaction mixtures were subsequently measured using DOSY. Based on the translational diffusion coefficient, the reaction products could be identified and distinguished from the active ingredients and the excipients used. The importance of thoughtful processing could also be demonstrated. If all peak heights are selected when evaluating signals split by direct spin spin coupling, this allows the detection of hidden signals as long as not all signals have the same diffusion coefficient. The selective selection of individual peak heights in the case of split signals also enables the evaluation of signals that overlap slightly. However, the limitations of this method were also shown when two signals overlap too much and differ too little in their diffusion coefficients. Hence, it has been successfully demonstrated in the various projects that the new chemometric approaches, as well as the new applications of already established methods, enable in depth findings and thus have a clear added value. / Spektroskopische Methoden haben sich schon vor Jahrzehnten in den verschiedensten Bereichen etabliert. Dies betrifft auch den pharmazeutischen Bereich, auch wenn sie hier zunächst meist nur zur Identitätsprüfung oder Strukturaufklärung verwendet wurden. Durch technische Weiterentwicklungen, wie Miniaturisierungen (NMR benchtop Geräte), Fourier Transformationen (NMR, MIR) oder die Kombination mit einer chemometrischen Auswertung (z. B. bei Process Analytical Technology, PAT), haben sie weiter an Bedeutung gewonnen, und es wurden neue Einsatzbereiche erschlossen. Das Ziel der vorliegenden Arbeit war es, weitere neue Ansätze zu untersuchen und neue Anwendungen für bereits etablierte Methoden zu finden und deren Mehrwert aufzuzeigen. Es wurden Möglichkeiten aufgezeigt mittels MIR , NIR und NMR Daten und deren chemometrischen Auswertungen (Hauptkomponentenanalyse, PCA; hierarchische Clusteranalyse, HCA; lineare Diskriminanzanalyse, LDA) erfolgreich den Hersteller bzw. das pharmazeutische Unternehmen verschiedener Paracetamol Präparate zu bestimmen. In diesem Zuge konnten Ähnlichkeiten zwischen Präparaten unterschiedlicher Firmen identifiziert werden. Um dies zu erreichen, wurde für jede spektroskopische Methode eine geeignete Probenvorbereitung entwickelt sowie geeignete Messparameter festgelegt, um reproduzierbare Spektren für die chemometrische Auswertung zu erhalten. Weiterhin wurden die Ergebnisse der zwei unüberwachten Methoden (HCA, PCA) miteinander verglichen, wobei die HCA die der PCA zum allergrößten Teil bestätigen konnte. Zudem war es möglich durch diese Methoden viele der Präparate anhand von Clustern zu charakterisieren, die durch vergleichbare Tablettenzusammensetzungen gebildet wurden. Um mit Hilfe des NIR Spektrometers intakte Tabletten vermessen zu können, wurde ein Aufsatz entwickelt und mittels 3D Druck hergestellt. Dessen Funktionalität wurde überprüft, indem Tabletten aus je zwei unterschiedlichen Chargen von neun Paracetamol Präparaten vermessen und analysiert wurden. Dabei konnten die Batches anhand einer PCA eindeutig unterschieden und zudem mittels statistischer Tests ein signifikanter Unterschied nachgewiesen werden. Für die NMR Spektroskopie wurde eine Methode entwickelt, um optimierte „Fingerprint“ Spektren von Arzneimittelformulierungen zu erhalten. Dazu wurde eine 1D DOSY Messmethode erarbeitet, bei der durch die passende Wahl der Messparameter die Signale des Wirkstoffes herausgefiltert werden konnten. Die chemometrische Auswertung konnte sich somit auf die Signale der Hilfsstoffe beschränken, anhand derer die Präparate unterschieden werden können. Vor allem bei Formulierungen, die zum größten Teil aus Wirkstoff bestehen, kann so eine Datenvorverarbeitung der Spektren vereinfacht und den ursprünglich sehr kleinen Hilfsstoffsignalen eine größere Bedeutung beigemessen werden. Für den Vergleich eines Hochfeld Spektrometers (400 MHz) mit einem „benchtop“ Spektrometer (80 MHz) wurde für zwei Fertigarzneimittel eine quantitative 1H NMR Methode entwickelt. Es konnte gezeigt werden, dass es möglich ist, mit beiden Geräten vergleichbare Ergebnisse zu erzielen. Dabei ist jedoch der Einfluss der Hilfsstoffe auf die Signale sowie die geringere Auflösung des „benchtop“ Gerätes zu berücksichtigen. Aus diesen Gründen war es ohne eine weitere Optimierung der Methode für einen der Wirkstoffe nicht möglich vergleichbare Ergebnisse mit beiden Geräten zu erzielen. Bei der Untersuchung verschiedener Reaktionen zwischen Wirk- und Hilfsstoffen mittels DOSY konnte dessen Nutzen als Screening Methode bei Stabilitätstests gezeigt werden. Für diesen Zweck wurden drei verschiedene Wirk- und Hilfsstoffe gemeinsam gestresst und die Reaktionsgemische anschließend mittels DOSY vermessen. Anhand des translationalen Diffusionskoeffizienten konnten die Reaktionsprodukte identifiziert und von den eingesetzten Wirk- und Hilfsstoffen unterschieden werden. Ebenso konnte die Bedeutung einer sorgfältigen Prozessierung demonstriert werden. Werden bei der Auswertung von Signalen, die durch direkte Spin Spin Kopplung aufgespalten wurden, alle Peakhöhen ausgewählt, erlaubt dies die Detektion von versteckten Signalen, falls nicht alle Signale den gleichen Diffusionskoeffizienten besitzen. Die selektive Auswahl einzelner Peakhöhen bei aufgespaltenen Signalen ermöglicht zudem die Auswertung von leicht überlappenden Signalen. Es wurden jedoch auch die Grenzen dieser Methode aufgezeigt: wenn zwei Signale zu stark überlappen und sich dabei in ihrem Diffusionskoeffizienten zu wenig unterscheiden. Somit konnte in den verschiedenen Projekten erfolgreich gezeigt werden, dass die neuen chemometrischen Ansätze, sowie die neuen Anwendungen bereits etablierter Methoden vertiefte Erkenntnisse ermöglichen und somit einen deutlichen Mehrwert besitzen.

Femtosecond Laser Direct Writing (FLDW) is a viable technique for producing photonic devices in bulk materials. This novel manufacturing technique is versatile due to its full 3D fabrication capability. Typically, the only requirement for this process is that the base material must be transparent to the laser wavelength. The modification process itself is based on non-linear energy absorption of laser light within the focal volume of the incident beam. This thesis addresses the feasibility of this technique for introducing photonic structures into novel dielectric materials. Additionally, this work provides a deeper understanding of the lightmatter interaction mechanism occurring at high pulse repetition rates. A novel structure on the sample surface in the form of nano-fibers was observed when the bulk material was irradiated with high repetition rate pulse trains. To utilize the advantages of the FLDW technique even further, a transfer of the technology from dielectric to semiconductor materials is investigated. However, this demands detailed insight of the absorption and modification processes themselves. Experiments and the results suggested that non-linear absorption, specifically avalanche ionization, is the limiting factor inhibiting the application of FLDW to bulk semiconductors with today’s laser sources.