Rozprawy doktorskie na temat „Optimized spectroscopy”

Electric field mediated gene delivery modalities have preferable safety profiles with the ability to rapidly transfect cells in vitro and in vivo with high efficiency. However, the current state of the art has relied on trial and error studies that target the average cell within a population present in treated tissue to derive electric pulse parameters. This results in fixed gene electrotransfer (GET) parameters that are not universally optimum. Slow progress towards the validation of a mechanism that explains this phenomena has also hindered its advancement in the clinic. To date, GET methods utilizing feedback control as a means to optimize doses of electric field stimulation have not been investigated. However, with modern electric components the electric characteristics of tissue exposed to electric pulses can be measured in very short time scales allowing for a near instantaneous assessment of the effect these pulses have on cells and tissue. This information is ideal for use in optimizing GET parameters to ensure the conditions necessary for gene delivery can be created regardless of anisotropic tissue architecture and electrode geometry. Bioimpedance theory draws parallels between cell structures and circuit components in an attempt to use circuit theory to describe changes occurring at a cellular and tissue level. In short, a reduction in tissue impedance indicates a reduction to the opposition of current flow in a volume conductor indicating new pathways for current. It has been purported these new pathways exist in the cell membrane and indicate a degree of membrane permeability/destabilization that either indicates or facilitates the uptake of exogenous molecules, such as nucleic acids or plasmid DNA. This study evaluated the use of relative impedance changes from 10 Hz – 10 kHz that occur in tissue before and after GET to indicate relative increase in tissue and membrane permeability. An optimum reduction in impedance was then identified as an indicator of the degree of membrane permeability required to significantly enhance exogenous DNA uptake into cells. This study showed the use of impedance-based feedback control to optimize GET pulse number in real time to target 80% or 95% reduction in tissue impedance resulted in an 12 and 14 fold increase in transgene expression over controls and a 6 and 7 fold increase in transgene expression over fixed pulse open loop protocols.

β-glucan is a major component of the fungal cell wall consisting of (1→3)-β linked glucose polymers with (1→6)-β linked side chains. The published classical isolation procedure of β-glucan from Saccharomyces cerevisiae is expensive and time-consuming. Thus, the aim of this research was to develop an effective procedure for the extraction of glucans. We have developed a new method for glucan extraction that will be cost effective and will maintain the native structure of the glucan. The method that we developed is 80% faster and utilizes 1/3 of the reagents compared to the published classical method. Further, the method developed increases the yield from 2.9 % to 10.3 %. Our new process has a branching frequency of 18.4 down from 197 and a side chain of 5.1 up from 2.5. The data indicate a more preserved native structure of isolated glucans.

There is huge academic interest and clinical need associated with the development of biomarker immunoassays where general aims are the generation of highly specific, convenient and sensitive sensing formats. In this project, a powerful electrochemical technique, electrochemical impedance spectroscopy (EIS), is applied in the establishment of powerful biomarker detecting protocols. Firstly, ultrasensitive, label-free and reusable insulin sensors, based on an antibody-PEGylated thiol self-assembly monolayer (PEG thiol SAM) interface, were produced and characterised via Faradaic EIS, presenting a detection limit (LOD) of 1.2 pM, a linear range across four orders of magnitude, and high sensitivity in even 50 % serum. By applying similar surface chemistry, a label-free biosensor, specific for the detection of α-synuclein antibodies, was fabricated. The α-synuclein interfaces used enabled the reliable detecting of this biomarker in patient sample serum. The concentration levels in the control and a patient group were determined to be significantly different, and, significantly, this difference was consistently across two different cohorts. Strikingly, this could potentially underpin an entirely new means of early Parkinson’s disease (PD) diagnosis. Non-Faradaic EIS methods were additionally applied to label-free insulin assays at both PEG thiol SAM and zwitterionic polymer film interfaces. The latter presented not only an exceptionally non-fouling interface, but also one seemingly both highly biocompatible and facilitating enhanced receptor: target binding. Finally, impedance assays, though potent, generally, operate by sampling only one of a limited number of available experimental variables, typically, Rct for Faradaic EIS, or C or Z for non-Faradaic EIS. Work carried out herein also explores the generation and utility of a portfolio of mathematically derived immittance functions all obtained from the same raw data sets. A particular focus was the examination of whether these were capable of increasing assay sensitivity and efficiency above normal impedance treatments.

La résonance magnétique nucléaire (RMN) fournit des informations structurelles et dynamiques précieuses à l'échelle atomique, cependant, la faible sensibilité et résolution des signaux empêchent l’étude d'objets moléculaires plus importants. Nous présentons 3 stratégies de marquage isotopique pour différentes expériences RMN des protéines en solution et démontrons leur potentiel pour l'étude structurale des biomolécules. Parmi les stratégies envisagées, 2 utilisent l'expression in vitro pour obtenir des protéines marquées sélectivement sur un groupe chimique et/ou acide aminé dans un environnement perdeutéré. Avec l’utilisation de séquences d'impulsions TROSY, ces échantillons ont permis des gains spectraux importants lorsque ils étaient spécifiquement marqués sur des groupes amide ou sur le méthylène des glycines tout en maintenant un taux de deutération élevé sur les autres fonctions chimiques des protéines. La troisième stratégie de marquage protéique utilise des protocoles in vivo pour des applications RMN innovantes: l'hyperpolarisation de noyaux en solution qui augmente leur sensibilité de plusieurs ordres de grandeur. La durée de vie de cette hyperpolarisation est régie par le temps de relaxation longitudinale des noyaux, qui est réduit pour les protéines à température ambiante. En isolant les noyaux d'intérêt dans un environnement perdeutéré, les interactions dipolaires créées par les protons voisins sont éliminées et les noyaux hyperpolarisés relaxent beaucoup plus lentement. L'hyperpolarisation d'un petit domaine protéique a été entreprise avec succès mais les conditions de dissolution doivent encore être améliorées pour conserver une phase aqueuse homogène
Nuclear Magnetic Resonance (NMR) provides valuable structural and dynamic information at the atomic scale, however, the low sensitivity and resolution of signals rapidly preclude investigations of larger molecular objects. We present three isotopic labeling strategies for different protein-solution NMR experiments and demonstrate their potential for the structural study of biomolecules in solution. Among the strategies considered, two are based on the use of in vitro protein expression to obtain selectively labeled proteins of a certain chemical group and/or amino acid in a perdeuterated environment. Perdeuteration is essential for the optimal use of Transverse Relaxation Optimized Spectroscopy pulse sequences. They allowed significant spectral gains when samples were specifically labeled on amide groups or on the methylene of glycines while maintaining a very high rate of deuteration on the other chemical functions of the proteins. The third protein labeling strategy employed is based on in vivo protocols but used in innovative NMR applications: a technique of hyperpolarization of nuclei in solution which increases their sensitivity by several orders of magnitude. The lifetime of this hyperpolarization is governed by the longitudinal relaxation time of nuclei, which are reduced for proteins at room temperature. By isolating the nuclei of interest in a perdeuterated environment, dipolar interactions created by neighboring protons were eliminated and hyperpolarized nuclei relaxed much more slowly. Hyperpolarization of a small protein domain was successfully undertaken at 1K but the dissolution conditions need to be improved in order to preserve a homogeneous aqueous phase

NMR spectroscopy is a primary analytical approach of metabolomics. Although 1D 1H NMR spectroscopy is versatile, highly reproducible and widely used, analysis of complex biological samples yields congested spectra with many overlapping signals. This makes metabolite identification and quantification challenging. 1H J-resolved (JRES) experiments spreads this high signal density into a second dimension, simplifying the spectral analysis. This thesis analyses the approaches and suitability of JRES spectroscopy to analyse metabolomics data. Firstly, the robustness of the JRES experiment is investigated. Using spectral relative standard deviation, benchmarks of spectral robustness can be compared between disparate processing techniques, sample types and analytical platforms. JRES spectra were found to be suitable for metabolomics experiments. Secondly, the application of standard metabolomic analysis methods to JRES spectra was examined. Using principal component analysis, the classification accuracy of 1D 1H and JRES spectra were investigated using several data sets. Alongside, three scaling methods were also evaluated. It was found that 2D JRES spectra and the glog transformation could produce 100% classification accuracy. Finally, spectral deconvolution of 2D JRES spectra from line-shape fitting was investigated Here, the mathematical functions describing the JRES line-shape, under several different processing conditions, are derived and used to create a semi-automated metabolite identification and quantification algorithm. Furthermore, possible quantitation errors arising from using JRES spectra are investigated, evaluating effects such as the overlapping of dispersive tails of nearby signals. In conclusion, the JRES experiment is a suitable for use in the field of metabolomics.

Une variété d’applications techniques dans la conversion d’énergie sont basées sur la combustion turbulente. En dépit du fonctionnement avancé de la technologie, la science de la combustion (turbulente) en est à un stade relativement jeune. Des études détaillées des flammes de référence sont essentielles pour mieux comprendre la physique sous-jacente aux processus mentionnés, ainsi que pour fournir une base de données fiable permettant de valider les modèles numériques développés pour simuler des problèmes de combustion turbulents. La spectroscopie Raman / Rayleigh est une technique extrêmement utile qui permet d’accéder simultanément à des informations sur la température et la concentration en espèces chimiques principales dans les structures de flammes. La data reduction strategy appliquée avec cette technique est cruciale pour extraire des informations fiables des investigations expérimentales. Dans cette thèse, une version modifiée de la stratégie, basée sur la Hybrid Method de Fuest et al., a été développée et testée à l’aide de techniques de résolution de NLP problem, dont des méthodes d’optimisation globale et un algorithme génétique. La stratégie proposée permet une réduction significative du temps de traitement des données, nécessite moins d’expertise de l’utilisateur et réduit l’erreur de mesure. La procédure modifiée a été appliquée à un data set fourni par une étude expérimentale de deux jet flame méthane/air prémélangés turbulentes appartenant au régime flamelet de la combustion prémélangée turbulente. Le data set est composé de plusieurs scalaires, y compris les principales espèces et la température, mesurés simultanément avec la single-shot 1D Raman/Rayleigh spectroscopy. Les résultats des mesures sont analysés et discutés
A variety of technical applications in energy conversion are based on turbulent combustion. Despite the advanced contest of operation, (turbulent) combustion science is at a relatively young stage. Detailed investigations of benchmark flames are essential to achieve a better understanding of the physics behind the mentioned processes, as well as to provide reliable database for validating numerical models developed to simulate turbulent combustion problems. Raman/Rayleigh spectroscopy is a highly valuable technique which allows to access simultaneous information on temperature and main chemical species concentration within the flame structures. The data reduction strategy applied with this technique is crucial, in order to extract reliable information from the experimental investigations. In this this thesis, a modified version of the strategy, based on the Hybrid Method by Fuest et al., has been developed and tested using NLP problem solving techniques, including global optimization methods and a genetic algorithm. The proposed strategy allows for a significant reduction of the data processing time, requires less user’s expertise and returns better results in reduced measurements error. The modified routine has been applied to data set provided by an experimental investigation of two turbulent premixed methane/air jet flames belonging to the flamelet regime of turbulent premixed combustion. The data set is composed by multiple scalars, including major species and temperature, simultaneously measured with single-shot 1D Raman/Rayleigh spectroscopy. The results of the measurements are analyzed and discussed

Nuclear Magnetic Resonance (NMR) is a phenomenon in which certain nuclei in the presence of a magnetic field and radiofrequency (RF) radiation emit a certain amount of signal at a frequency equal to that of the RF radiation. Proton Magnetic Resonance Spectroscopy (1H-MRS) is an NMR technique capable of measuring the chemical composition, often referred to as metabolites, of the human body non-invasively and in vivo. It is commonly used as a research tool in the investigation of neurological disorders such as multiple sclerosis, brain tumors, stroke, clinical depression, and schizophrenia. Accurate quantification of the metabolites of interest requires a reference standard of known and fixed concentration. Brain tissue water has been previously reported to have a fairly constant and known concentration, and so has been suggested to be a suitable reference concentration in absolute quantitative 1H-MRS of the human brain. In practice, however, it is challenging to measure the actual tissue water concentration; hence, some studies choose to use estimates of tissue water concentration from the literature. These literature values are usually averages from a healthy study group. There are however indications that brain tissue water content could vary widely in certain disease conditions such as in brain tumors and inflammation. In such situations, absolute metabolite quantification using the literature estimates of tissue water content will be inaccurate while the measurement of cerebral water content using the available techniques will be impractical for the patients due to scanning time considerations. It is therefore necessary to develop a technique that can be used to quantify both the reference water and metabolite concentrations, simultaneously without subject tolerance issues. The main objective of this thesis was to investigate the response of water NMR signal from human brain tissue under various measurement conditions using the single-voxel 1H-MRS technique. As part of the investigation, the thesis also focused on the development of methods for the absolute quantification of cerebral water and metabolite concentrations. A standard 1H-MRS water-suppressed acquisition on the General Electric (GE) MR scanner acquires some unsuppressed-water spectra at the beginning of the PRESS pulse sequence. Using the Spectroscopy Analysis by GE (SAGE) software package (version 7), this thesis developed methods to optimise the unsuppressed-water and suppressed-water signals from which, respectively, cerebral water and metabolite concentrations were estimated. The unsuppressed-water signal response characteristics were investigated in experiments at 3 T that involved: 1) variation of the MRS voxel position over a three-dimensional RF field within an eight-channel head coil; 2) measurement of the relaxation times of brain tissue water using standard saturation recovery and multi spin-echo MRS techniques; 3) measurement of brain tissue water content in peripheral inflammation; and 4) estimation of the BOLD effect on the water spectral peak. The stability of the MR scanner used for all the investigations was assessed. Over the project period, the worst precision measurements of the scanner (for both water and metabolite signals) were observed to be about 12 % and 26 % in serial phantom and human studies, respectively. The MRI/MRS scanner was therefore found to measure water and metabolite signals with good precision, both in vivo and in vitro. By recording the water NMR signal responses at various locations within the phased-array head coil, RF sensitivity profile (voxel position-dependent) equations of the head coil were obtained. The coordinates of any in vivo voxel could be substituted into an appropriate profile equation to estimate an unsuppressed-water signal area that could be used as a reference signal to quantify brain tissue water content. This novel technique of quantifying cerebral water content is superior to the previous techniques of performing multi-echo unsuppressed-water signal acquisitions. The method does not require extra unsuppressed-water acquisitions, or corrections for variations in the sensitivity of the eight-channel head coil as both the in vivo and reference signals are acquired from the same voxel position. Brain tissue water content was subsequently quantified accurately using the newly developed method of referencing. In frontal brain voxels, the average water content, WC of grey matter, GM was found to be higher than that of white matter, WM (GM/WM WC ± SE = 46.37 ± 2.58/42.86 ± 2.46 mol/kg; p = 0.02); parietal voxels also showed a similar comparison (GM/WM WC ± SE = 37.23 ± 1.70/34.14 ± 2.02 mol/kg; p = 0.03). These findings were consistent with previous reports of cerebral water content. For regions of mixed proportions of grey and white matter tissues, the average water contents of each tissue type considered separately (by voxel segmentation) and together were found to compare with literature estimates. Using data from five voxel positions, average brain tissue water content was observed to be uniformly distributed across the human brain by one-way ANOVA (p = 0.60), and did not vary significantly with gender (p > 0.05) and age (p > 0.05). For the first time, cerebral water content was observed in this thesis to remain fairly constant in psoriatic arthritis, a peripheral inflammatory condition (one-way ANOVA, p = 0.63). Among five brain metabolites quantified in the psoriasis patients, only the mean concentration of creatine, Cr was found to be significantly lower in the frontal grey matter voxels of the patients, PsA compared to healthy controls, HC at baseline (PsA/HC ± SE = 6.34 ± 0.38/7.78 ± 0.38 mM/kg; p = 0.01) and post-TNF-alpha blockade medication (PsA/HC ± SE = 6.69 ± 0.25/7.78 ± 0.38 mM/kg; p = 0.03). None of the metabolite concentrations, including Cr (p = 0.27), changed significantly with medication. The condition of PsA was not observed to affect the mood of the patients, as indicated by their BDI scores. The significant finding of Cr concentration alteration in psoriatic arthritis thus suggests that Cr may not be a reliable denominator in studies of psoriasis that express the metabolite levels as ratios. The T1 and T2 relaxation times of water and the metabolites were measured in the prefrontal grey matter (T1/T2 ± SE = 1574 ± 61/147 ± 6 ms) and bilateral Hippocampi (T1/T2 ± SE; left = 1475 ± 68/178 ± 83 ms, right = 1389 ± 58/273 ± 98 ms). The relaxation time estimates for the metabolites were in agreement with literature values; relaxation times for water however were measured for the first time in those regions and at 3 T. The measured relaxation times were used to correct the water and metabolite signals for relaxation effects during their absolute quantification, and could as well serve the same purpose in future studies. There is increasing interest in the BOLD response of cerebral metabolites and water during tasks. This thesis thus also assessed changes in brain tissue metabolite and water contents while a subject experienced a visual stimulus. In the presence of the visual stimulus, the BOLD effects on the metabolite and water spectral peaks were found to be comparable, as has been observed in previous studies. For the first time, this thesis further investigated the impact of temporal resolution (determined by NEX) on the amount of the BOLD signal acquired from cerebral water and metabolites. In a single visual activation paradigm, the BOLD effect resulted in increased water peak area which differed significantly between NEX values of 2 and 8 (p < 0.01); this observation also was true for NAA and Glu. The findings thus suggest that temporal resolution of the MRS data could result in significant differences in the results of functional MRS studies.

Il cancro della prostata (PCa) è il tumore maligno non-cutaneo più diffuso tra gli uomini ed è il secondo tumore che miete più vittime nei paesi occidentali. La necessità di nuove tecniche non invasive per la diagnosi precoce del PCa è aumentata negli anni. 1H-MRS (proton magnetic resonance spectroscopy) e 1H-MRSI (proton magnetic resonance spectroscopy imaging) sono tecniche avanzate di spettroscopia in risonanza magnetica che permettono di individuare presenza di metaboliti come citrato, colina, creatina e in alcuni casi poliammine in uno o più voxel nel tessuto prostatico. L’abbondanza o l’assenza di uno di questi metaboliti rende possibile discriminare un tessuto sano da uno patologico. Le tecniche di spettroscopia RM sono correntemente utilizzate nella pratica clinica per cervello e fegato, con l’utilizzo di software dedicati per l’analisi degli spettri. La quantificazione di metaboliti nella prostata invece può risultare difficile a causa del basso rapporto segnale/rumore (SNR) degli spettri e del forte accoppiamento-j del citrato. Lo scopo principale di questo lavoro è di proporre un software prototipo per la quantificazione automatica di citrato, colina e creatina nella prostata. Lo sviluppo del programma e dei suoi algoritmi è stato portato avanti all’interno dell’IRST (Istituto Romagnolo per lo Studio e la cura dei Tumori) con l’aiuto dell’unità di fisica sanitaria. Il cuore del programma è un algoritmo iterativo per il fit degli spettri che fa uso di simulazioni MRS sviluppate con il pacchetto di librerie GAMMA in C++. L’accuratezza delle quantificazioni è stata testata con dei fantocci realizzati all’interno dei laboratori dell’istituto. Tutte le misure spettroscopiche sono state eseguite con il nuovo scanner Philips Ingenia 3T, una delle machine di risonanza magnetica più avanzate per applicazioni cliniche. Infine, dopo aver eseguito i test in vitro sui fantocci, sono stati acquisiti gli spettri delle prostate di alcuni volontari sani, per testare se il programma fosse in grado di lavorare in condizioni di basso SNR.

Chemische Reaktionen in der Natur sowie Prozesse in synthetischen Materialien werden oft erst durch die Wechselwirkung von Licht mit Materie ausgelöst. Üblicherweise werden diese komplexen Prozesse mit Hilfe von Näherungen beschrieben. Im ersten Teil der Arbeit wird die Gültigkeit der Born-Oppenheimer Näherung in einem vibronischen Modellsystem (Trans-Polyacetylene) unter Photoelektronenspektroskopie im Gleichgewicht sowie zeitaufgelöster Photoelektronenspektroskopie im Nichtgleichgewicht überprüft. Die vibronische Spektralfunktion zeigt aufgrund des faktorisierten Anfangs- und Endzustandes in der Born-Oppenheimer Näherung zusätzliche Peaks, die in der exakten Spektralfunktion nicht auftreten. Im Nichtgleichgewicht zeigen wir für eine Franck-Condon Anregung und eine Anregung mit Pump-Probe Puls, wie die Bewegung des vibronischen Wellenpaktes im zeitabhängigen Photoelektronenspektrum verfolgt werden kann. Im zweiten Teil der Arbeit werden sowohl die Materie als auch das Licht quantisiert behandelt. Für eine volle quantenmechanische Beschreibung des Elektron-Licht Systems, verwenden wir die kürzlich entwickelte quantenelektrodynamische Dichtefunktionaltheorie (QEDFT) für gekoppelte Elektron-Photon Systeme. Wir zeigen erste numerische QEDFT-Berechnungen voll quantisierter Atome und Moleküle in optischen Kavitäten, die an das quantisierte elektromagnetische Feld gekoppelt sind. Mit Hilfe von Fixpunktiterationen berechnen wir das exakte Kohn-Sham Potential im diskreten Ortsraum, wobei unser Hauptaugenmerk auf dem Austausch-Korrelations-Potential liegt. Wir zeigen die erste Näherung des Austausch-Korrelations-Potentials mit Hilfe eines optimierten effektiven Potential Ansatzes angewandt auf einen Jaynes-Cummings-Dimer. Die dieser Arbeit zugrunde liegenden Erkenntnisse und Näherungen ermöglichen es neuartige Phänomene an der Schnittstelle zwischen den Materialwissenschaften und der Quantenoptik zu beschreiben.
Many natural and synthetic processes are triggered by the interaction of light and matter. All these complex processes are routinely explained by employing various approximations. In the first part of this work, we assess the validity of the Born-Oppenheimer approximation in the case of equilibrium and time-resolved nonequilibrium photoelectron spectra for a vibronic model system of Trans-Polyacetylene. We show that spurious peaks appear for the vibronic spectral function in the Born-Oppenheimer approximation, which are not present in the exact spectral function of the system. This effect can be traced back to the factorized nature of the Born-Oppenheimer initial and final photoemission states. In the nonequilibrium case, we illustrate for an initial Franck-Condon excitation and an explicit pump-pulse excitation how the vibronic wave packet motion can be traced in the time-resolved photoelectron spectra as function of the pump-probe delay. In the second part of this work, we aim at treating both, matter and light, on an equal quantized footing. We apply the recently developed quantum electrodynamical density-functional theory, (QEDFT), which allows to describe electron-photon systems fully quantum mechanically. We present the first numerical calculations in the framework of QEDFT. We focus on the electron-photon exchange-correlation contribution by calculating exact Kohn-Sham potentials in real space using fixed-point inversions and present the performance of the first approximate exchange-correlation potential based on an optimized effective potential approach for a Jaynes-Cummings-Hubbard dimer. This work opens new research lines at the interface between materials science and quantum optics.

La formulation des principes actifs (PA) pharmaceutiques à l'état amorphe est une préoccupation majeure actuelle du génie pharmaceutique car elle permet d'améliorer la solubilité des (PA) peu solubles. Son optimisation requiert une parfaite connaissance de la courbe de solubilité du (PA) dans le polymère, de manière à définir le taux maximum de (PA) pouvant être chargé dans le polymère sans risque de recristallisation ultérieure. La détermination de cette courbe de solubilité est cependant extrêmement difficile en raison de la forte viscosité des polymères qui rend les états saturés très difficiles à atteindre. Dans cette thèse, nous présentons une méthode efficace de détermination des courbes de solubilité qui permet de s'affranchir des limitations cinétiques liées à la viscosité. L'idée générale est d'utiliser des dispersions moléculaires et cristallines (DMC) dans lesquelles le (PA) est dispersé dans le polymère, en partie au niveau moléculaire et en partie sous la forme de minuscules cristallites. Cette microstructure particulière augmente considérablement la vitesse de dissolution des cristallites puisque : (1) La dispersion moléculaire a un effet plastifiant qui augmente considérablement la mobilité moléculaire dans la matrice amorphe. (2) La fine dispersion des cristallites réduit fortement les distances sur lesquelles les molécules de PA doivent diffuser pour envahir le polymère. La méthode a été mise au point et testée sur deux (PA) (le sulindac et le paracétamol) et un polymère (le polyvinylpyrrolidon). L'étude des courbes de solubilité nous a par ailleurs permis de préciser le polymorphisme du sulindac et du paracétamol, et en particulier les relations de stabilité des polymorphes étudiés
The formulation of pharmaceutical active principles (PA) in the amorphous state is a major current concern of pharmaceutical engineering because it improves the solubility of poorly soluble (PA). Its optimization requires a perfect knowledge of the solubility curve of (PA) in the polymer, so as to define the maximum level of (PA) that can be loaded into the polymer without the risk of subsequent recrystallization. The determination of this solubility curve is, however, extremely difficult because of the high viscosity of the polymers which makes the saturated states very difficult to achieve. In this thesis, we present an efficient method for determining solubility curves that allows us to overcome the kinetic limitations related to viscosity. The general idea is to use molecular and crystalline dispersions (MCDs) in which (PA) is dispersed in the polymer, partly at the molecular level and partly in the form of tiny crystallites. This particular microstructure greatly increases the dissolution rate of the crystallites since: (1) Molecular dispersion has a plasticizing effect which greatly increases the molecular mobility in the amorphous matrix. (2) The fine crystallite dispersion strongly reduces the distances over which the API molecules must diffuse to invade the polymer. The method was developed and tested on two (PA) (sulindac and paracetamol) and a polymer (polyvinylpyrrolidon). The study of solubility curves also allowed us to specify the polymorphism of sulindac and paracetamol, and in particular the stability relationships of the polymorphs studied

碩士
國立交通大學
光電科技學程
105
This thesis focuses on investigating the optimized ultraviolet (UV) curing time in forming PSBPs as determined by means of dielectric spectroscopy. In experiments, we firstly choose a well-known nematic liquid crystal (NLC) E7 as the host and two types of monomer (i.e., RM257 and TMPTA) and fabricate three NLC/monomer mixtures. They are E7/RM257, E7/TMPTA and E7/RM257/ TMPTA. In this stage, the UV exposure process was carried out at the temperature of T = 30 C. Based on dielectric results, we analyze the dielectric spectra within the frequency reigme of 1 kHz10 kHz to find out time dependence of dielectric properties of the NLC/monomer caused by the UV exposure and thus the induction of photopolymerization. We found that if we extend the UV exposure time the spectrum of the imaginary part of the dielectric will trend down gradually. After the minimum value of the trend, the dielectric properties are the same as the original LC. Extending this approach to polymer stabilized blue phases (PSBPs), we further investigate and confirm the optimized UV exposure time for BPLC/monomer mixture by means of the technology of dielectric phase transition temperature and optical texture to verify dielectric spectra results. Based on experimental results, the phase transition temperature of each sample will be identical if the UV irradiation time is longer than the critical time tc. and sample still keep the blue-phase state to low temperature (T = 10 ºC). However, the phase transition temperature of each sample would be different and the blue-phase state can not be maintained at low temperature (T = 10 ºC). For longer UV time (t > tc), the internal monomer has been completely cured to polymer network that stabilized the blue phase liquid crystal lattice, whereas the blue phase liquid crystal has been successfully formed PSBP. According to the above results, we can use the trend of dielectric spectrum to find best PSBP UV curing time (tc).

Background: As técnicas de ressonância magnética cardíaca por imagem (MRI) e espetroscopia (MRS) são ferramentas usadas para caraterizar, de forma não invasiva, modelos de rato com doenças cardíacas humanas. As experiências são tipicamente conduzidas em sistemas de Ressonância Magnética (MR) equipados com magnetos de elevada intensidade (≥ 7 Tesla). Um requisito fundamental da MR é a homogeneidade do campo magnético estático, B0 (Grutter, 1993), e as flutuações (inomogeneidades) do campo magnético principal na região de imagem devem ser menores a três partes por milhão (3 ppm). Inserindo uma amostra aumenta-se a inomogeneidade do campo (devido a diferentes graus de magnetização ao longo da amostra como resposta a B0 ("suscetibilidade magnética")), a qual necessita de ser compensada (Crijns et al, 2011; Koch et al, 2006). Homogeneizar (shimming) o campo magnético estático é uma tarefa crucial em qualquer experiência de MR para maximizar a resolução e a razão entre sinal e ruído. Isto é particularmente importante em campos magnéticos de elevada intensidade devido à dependência linear da suscetibilidade magnética com B0. O ajuste manual das bobinas de shim é laborioso e subjetivo. Para além disso, este processo é particularmente desafiante onde vários tecidos (por exemplo, osso, fluxo de sangue, entre outros) estão numa vizinhança próxima dentro do tórax, tendo cada um diferentes suscetibilidades magnéticas e movimentos relativos. Métodos automáticos de shimming, como o FASTMAP ou FASTERMAP (Shen et al, 1997), estão experimental e clinicamente bem estabelecidos no tecido cerebral mas falham no coração devido à fase de sinal mal definida de MR, particularmente no interior dos ventrículos. Com base numa técnica previamente implementada para o cérebro humano, foi investigada a implementação de uma nova abordagem para corações de ratos, in vivo, capaz de homogeneizar B0 na região de interesse, com uma forma aleatória. Objetivo: O objetivo deste projeto é investigar os parâmetros ótimos de digitalização e pós-processamento, por forma a otimizar e alcançar um procedimento automático de shimming, potenciando, assim, as técnicas de MRI e MRS cardíacas. Métodos: Diversos ratos (n=5) foram submetidos à técnica de MR, realizada num magneto horizontal de 9.4 Tesla (T). A aquisição de imagem foi conduzida através de sequências rápidas echo variando os seguintes parâmetros: resolução, compensação de fluxo (on / off), orientação (short-axis / axial) e dimensão (multi-cortes 2D vs 3D). Três diferentes configurações de bobinas de shim foram investigadas e a sequência ótima de MR foi avaliada. Resultados: O nível de 17% de threshold demonstrou ser aceitável para a remoção das discontinuidades de fase. A análise quantitativa do desempenho das diferentes abordagens de phase unwrapping mostrou que a abordagem 3D é a mais eficaz na resolução das discontinuidades de fase presentes nos mapas de campo. A aplicação de orientação axial, os dados de maior resolução, a ausência de compensação de fluxo e a introdução de bobinas de shim de maiores ordens demonstraram um peso significativo na redução das inomogeneidades de B0, quando aplicados. Conclusões: Este projeto permitiu estabelecer parâmetros ótimos de aquisição e opções de pós-processamento que melhoram a homogeneidade de B0, importantes na validação de futuros estudos complementares.
Background: Cardiac magnetic resonance imaging and spectroscopy are tools to non-invasively characterize rodent models of human heart disease. The experiments are typically carried out on dedicated MR systems equipped with ultra-high field magnets (≥ 7 Tesla). One fundamental requirement of MR is the homogeneity of the static magnetic field B0 (Grutter, 1993), and fluctuations of the main magnetic field (B0 inhomogeneities) within the scan region should be less than three parts per million (3 ppm). Inserting a sample inherently increases the field inhomogeneity (due to different degree of magnetization across the sample in response to the B0 field (“magnetic susceptibility”)), which needs to be compensated for (Crijns et al, 2011; Koch et al, 2006). Homogenizing (i.e. shimming) the static magnetic field is crucial for any MR experiment in order to maximize resolution and signal-to-noise. This is particularly important at ultra-high magnetic fields due to linear dependence of magnetic susceptibility. Adjusting the three linear and typically up to 14 higher order shims manually is laborious and subjective. Moreover, this process is particularly challenging where various tissues (i.e. heart and skeletal muscle, bone, lungs and flowing blood) are in close vicinity within the chest, each having different magnetic susceptibilities and relative motions. Auto-shim methods such as FASTMAP or FASTERMAP (Shen et al, 1997), are clinically and experimentally well established in brain tissue, but inevitably fail in the heart due to the ill-defined phase of the MR-signal, particularly inside the ventricles. Based on a technique, previously applied to human brain – implemented a novel approach for the application to mouse hearts in vivo, that is able to homogenize the B0-field in an arbitrarily shaped, but connected region of interest. Aim: The aim of this project is to investigate optimal scan parameters and post-processing approach to optimize and advance an automated shimming procedure for improved experimental cardiac magnetic resonance imaging and spectroscopy at ultra-high magnetic fields. Methods: Mice (n = 5) underwent MR experiments carried out in a 9.4 Tesla (T) horizontal magnet. The image acquisition was performed using fast gradient echo sequences varying the following parameters: resolution, flow compensation on / off, orientation (short-axis / axial), and dimension (2D multislice vs 3D). Three different shim coils’ configurations (shim coils up to the third order) were investigated and optimal MR sequence was assessed. Results: The threshold level of 17% proved to be acceptable for removal of phase discontinuities and hence it was used in subsequent studies. Quantitative analysis of the performance of different phase unwrapping approaches showed that the 3D approach is the most effective in resolving phase discontinuities present in field maps. The application of axial orientation, highest resolution data, absence of compensation flow and the introduction of higher order shim coils showed a significant reduction of B0 inhomogeneities when applied. Conclusions: This project established optimal acquisition parameters and post-processing options to improve the homogeneity of B0, and will aid the validation process in further follow-up studies.

The ability to interrogate structure-function photophysical properties on lanthanide-doped nanoscale materials will define their utility in next-generation applications and devices that capitalize on their size, light-conversion efficiencies, emissive wavelengths, syntheses, and environmental stabilities. The two main topics of this dissertation are (i) the interrogation of laser power-dependent quantum yield and total radiant flux metrics for a homogeneous, solution phase upconversion nanocrystal composition under both continuous wave and femtosecond-pulsed excitation utilizing a custom engineered absolute measurement system, and (ii) the synthesis, characterization, and power-dependent x-ray excited scintillation properties of [Y₂O₃; Eu] nanocrystals, and their integration into a fiber-optic radiation sensing device capable of in vivo dosimetry.

Presented herein is the laser power-dependent total radiant flux and absolute quantum yield measurements of homogeneous, solution-phase [NaYF₄; Yb (15%), Er (2%)] upconversion nanocrystals, and further compares the quantitative total radiant flux and absolute quantum yield measurements under both 970 nm continuous-wave and 976 nm pulsed Ti-Sapphire laser excitation (140 fs pulse-width, 80 MHz). This study demonstrates that at comparable excitation densities under continuous-wave and fs-pulsed excitation from 42 - 284 W/cm2, the absolute quantum yield, and the total radiant flux per unit volume, are within a factor of two when spectra are integrated over the 500 - 700 nm wavelength regime. This study further establishes the radiant flux as the true unit of merit for quantifying emissive output intensity of upconverting nanocrystals for application purposes, especially given the high uncertainty in solution phase upconversion nanocrystal quantum yield measurements due to their low absorption cross-section. Additionally, a commercially available bulk [NaYF₄; Yb (20%), Er (3%)] upconversion sample was measured in the solid-state to provide a total radiant flux and absolute quantum yield standard. The measurements were accomplished utilizing a custom-engineered, multi-detector integrating sphere measurement system that can measure spectral sample emission in Watts on a flux-calibrated (W/nm) CCD-spectrometer, enabling the direct measurement of the total radiant flux without need for an absorbance or quantum yield value.

Also presented is the development and characterization of a scintillating nanocrystalline composition, [Y_2-xO₃; Eu_x, Li_y], in which Eu and Li dopant ion concentrations were systematically varied in order to define the most emissive compositions under specific x-ray excitation conditions. It is shown that these optimized [Y_2-xO₃; Eu_x, Li_y] compositions display scintillation responses that: (i) correlate linearly with incident radiation exposure at x-ray energies spanning from 40 - 220 kVp, and (ii) manifest no evidence of scintillation intensity saturation at the highest evaluated radiation exposures [up to 4 Roentgen per second]. X-ray excitation energies of 40, 120, and 220 kVp were chosen to probe the dependence of the integrated emission intensity upon x-ray exposure-rate in energy regimes where either the photoelectric or the Compton effect governs the scintillation mechanism on the most emissive [Y_2-xO₃; Eu_x, Li_y] composition, [Y_1.9O₃; Eu_0.1, Li_0.16]. These experiments demonstrate for nanoscale [Y_2-xO₃; Eu_x], that for comparable radiation exposures, when scintillation is governed by the photoelectric effect (120 kVp excitation), greater integrated emission intensities are recorded relative to excitation energies where the Compton effect regulates scintillation (220 kVp excitation).

The nanoscale [Y_1.9O₃; Eu_0.1, Li_0.16] was further exploited as a detector material in a prototype fiber-optic radiation sensor. The scintillation intensity from a [Y_1.9O₃; Eu_0.1, Li_0.16]-modified optical fiber tip, recorded using a CCD-photodetector or a Si-photodiode, was correlated with radiation exposure using a Precision XRAD 225Cx small-animal image guided radiation therapy (IGRT) system, an orthovoltage cabinet-irradiator, and a clinical X-ray Computed Tomography (CT) machine. For all x-ray energies tested from 80 - 225 kVp, this near-radiotransparent device recorded scintillation intensities that tracked linearly with total radiation exposure, highlighting its capability to provide alternately accurate dosimetry measurements for both diagnostic imaging and radiation therapy treatment. Because Si-based CCD and photodiode detectors manifest maximal sensitivities over the emission range of nanoscale [Y_1.9O₃; Eu_0.1, Li_0.16], the timing speeds, sizes, and low power-consumption of these devices, coupled with the detection element's linear dependence of scintillation intensity with radiation dose, demonstrates the opportunity for next-generation radiation exposure measuring devices for in/ex vivo applications that are ultra-small, inexpensive, and accurate.

Dissertation

A variety of technical applications in energy conversion are based on turbulent combustion. Despite the advanced contest of operation, (turbulent) combustion science is at a relatively young stage. Detailed investigations of benchmark flames are essential to achieve a better understanding of the physics behind the mentioned processes, as well as to provide reliable database for validating numerical models developed to simulate turbulent combustion problems. Raman/Rayleigh spectroscopy is a highly valuable technique which allows to access simultaneous information on temperature and main chemical species concentration within the flame structures. The data reduction strategy applied with this technique is crucial, in order to extract reliable information from the experimental investigations. In this this thesis, a modified version of the strategy, based on the Hybrid Method by Fuest et al., has been developed and tested using NLP problem solving techniques, including global optimization methods and a genetic algorithm. The proposed strategy allows for a significant reduction of the data processing time, requires less user’s expertise and returns better results in reduced measurements error. The modified routine has been applied to data set provided by an experimental investigation of two turbulent premixed methane/air jet flames belonging to the flamelet regime of turbulent premixed combustion. The data set is composed by multiple scalars, including major species and temperature, simultaneously measured with single-shot 1D Raman/Rayleigh spectroscopy. The results of the measurements are analyzed and discussed.

Silicon has long been the most important material in semiconductor technology, having a huge impact in the field. This is predominantly due to its highly useful electrical and optical properties. It has also been shown to phase transform under pressure, from the diamond cubic Si-I that is ubiquitous in our modern society, to several different phases. Of particular interest is the rhombohedral Si-XII phase which, while showing some similar optical and electrical properties to Si-I, has potentially even more useful properties. A metastable region of Si-XII, in conjunction with Si-III, can be generated through point loading pressure via nanoindentation. This study addresses the challenges of creating large uniform volumes of Si-XII and compares the electrical properties of such volumes to a similar region created through a line of overlapping indents. This is motivated by the desire to broaden the range of uses for Si-XII and to allow for greater flexibility in the creation of regions of this material. Nanoindentation of Si-I is conducted using two separate spherical indenters; a larger, ~60 um diameter, indenter with a non-uniform surface and a smaller, ~20 um diameter, indenter with a smooth surface. Indents are made across a large range of maximum load forces. The maximum loads are separated into deformation regimes based on the presence of events in the load/unload curves. The characteristics of each of the regimes are determined using Raman microspectroscopy, cross-sectional transmission electron microscopy (XTEM), and scanning electron microscopy (SEM). For the ~20 um diameter tip, the regime in which only phase transformation occurs extends across the maximum loads 350 - 500 mN. Indentation made to a maximum load above 500 mN still display phase transformation but also display other competing deformation mechanisms such as dislocations, slip and fracturing. This causes a reduction in the amount of phase transformation at these higher maximum loads. For the ~60 um diameter tip, the phase transformation only regime ranges from 550 - 750 mN. Indentation above 750 mN displays competing deformation mechanisms. A maximum load of 2500 mN (>>750 mN) creates a larger transformed region than 750 mN but causes a large amount of dislocations in the underlying substrate. In the phase transformation only regime, the smaller diameter indenter with a smoother tip creates larger regions of Si-III/XII with a more consistent shape than the larger diameter indenter with pitting on the tip. The thus obtained understanding of the optimal indentation conditions are utilised to extract the electrical properties of Si-XII. Therefore, amorphous silicon (a-Si) resistive bar structures that isolate two contact pads are bridged with a region of Si-III/XII generated by a single indent and the electrical properties of this region is measured. These measurements from a single indentation are comparable to those of a line of overlapping indents. This shows that the process of overlapping indentation areas do not negatively affect the electrical properties of the region. This allows for greater flexibility in the "writing" of Si-III/XII zones through overlapping indentation areas, thus promises greater potential for future technological application.

Molecular imaging using fluorescence spectroscopy-based techniques is generally inefficient due to the low quantum yield of most naturally occurring biomolecules. Current fluorescence imaging techniques tag these biomolecules chemically or through genetic manipulation, increasing the complexity of the system. A technique capable of imaging these biomolecules without modifying the chromophore and/or its environment could provide vital biometric parameters and unique insights into various biological processes at a molecular level. Pump probe spectroscopy has been used extensively to study the molecular properties of poorly fluorescing biomolecules, because it utilizes the known absorption spectrum of these chromophores. Optical Coherence Tomography (OCT) is an optical imaging modality that harnesses the power of low coherence interferometry to measure the 3-D spatially resolved reflectivity of a tissue sample. We plan to develop a new molecular imaging modality that combines these techniques to provide 3-D, highresolution molecular images of various important biomolecules. The system uses a Fourier domain OCT setup with a modified sample arm that combines the "pump" and "probe" beams. The pump beam drives the molecules from the ground state to excited state and the probe interrogates the population change due to the pump and is detected interferometrically. The pump and the probe beam wavelengths are optimized to maximize absorption at the pump wavelength and maximize the penetration depth at the probe wavelength. The pump-probe delay can be varied to measure the rate at which the excited state repopulates the ground state, i.e., the ground state recovery time. The ground state recovery time varies for different chromophores and can potentially be used to identify different biomolecules. The system was designed and optimized to increase the SNR of the PPOCT signals. It was tested by imaging hemoglobin and melanin samples and yielded encouraging results. Potential applications of imaging hemoglobin using this technique include the mapping of tissue microvasculature and measuring blood-oxygen saturation levels. These applications could be used to identify hypoxic areas in tissue. Melanin imaging can provide means of demarcation of melanoma in various organs such as skin, eye and intestines.

Les matériaux moléculaires amorphes, aussi appelés verres moléculaires, sont constitués de molécules organiques de petite taille capables de s’organiser de façon désordonnée. En plus de présenter certaines des propriétés analogues à celles des polymères, ils offrent des avantages supplémentaires, puisqu’ils sont des espèces isomoléculaires dont la synthèse, la purification et la mise en œuvre sont facilitées par leur viscosité relativement faible. Toutefois, la préparation souvent exigeante de ces matériaux et leur durée de vie utile limitée par leur tendance à relaxer vers l’état cristallin demeurent des obstacles à leur utilisation pour certaines applications, e.g. opto-électronique, nanolithographie, pharmaceutique. Le développement de stratégies visant à faciliter la préparation de la phase vitreuse et éviter sa cristallisation est donc essentiel à la conception de matériaux moléculaires amorphes fonctionnels. L’objectif principal de cette thèse est d’établir des relations entre la structure moléculaire des verres moléculaires et leurs propriétés. Pour y arriver, différentes librairies de composés modèles, des dérivés analogues de triazine ayant démontré une excellente capacité à former une phase vitreuse, sont utilisées pour i) déterminer l’influence de la nature et de la position des groupements sur la triazine; ii) explorer l’influence des liaisons hydrogène sur les propriétés des verres lorsque leur structure comporte des groupements fonctionnels reconnus pour faciliter la cristallisation et lorsque leurs conditions de préparation se rapprochent de celles employées en industrie et iii) exploiter la phase amorphe afin d’étudier la photosensibilité des azobenzènes (azo) en vue d’optimiser leur utilisation dans des applications. Tout d’abord, l’influence des différents groupes substituants sur la triazine (groupements de tête, auxiliaires et liants) sur la capacité des composés à former une phase vitreuse (GFA), sur sa stabilité cinétique (GS) et sur sa température de transition vitreuse (Tg) est étudiée. Un système de classification des composés développé à partir de mesures de calorimétrie différentielle à balayage (DSC) et des mesures de spectroscopie infrarouge (IR) à température variable combinées à des analyses chimiométriques facilitent la rationalisation des rôles joués par chaque groupe. L’impact des liaisons hydrogène (H), de la barrière énergétique de rotation et de l’encombrement stérique des groupements est ainsi déterminé, permettant de conclure que le groupe de tête est le plus influent et que la présence de liaisons H n’est pas essentielle au GFA mais qu’elle est importante pour obtenir une Tg élevée. Ensuite, l’influence des liaisons H sur les propriétés des verres se rapprochant de ceux exploités dans l’industrie est explorée. Des mesures de spectroscopie IR à température variable, de DSC et de résolution de structures cristallines ont permis de conclure que les liaisons H réussissent à nuire à la cristallisation des composés et ce, même s’ils sont simultanément fonctionnalisés avec des motifs qui favorisent la cristallisation (empilements π-π entre dérivés stilbènes fluorés et non fluorés). De plus, trois composés analogues fonctionnalisés avec un groupement de tête possédant une capacité décroissante à établir des liaisons H (donneur, accepteur, aucune) ont été déposés en phase vapeur (PVD), une technique employée entre autres dans l’industrie opto-électronique pour évaluer leur capacité à former des verres ultrastables. Les films ainsi préparés présentent tous des propriétés similaires à celles des verres ultrastables précédemment étudiés, telles qu’une plus grande densité et anisotropie, et sont tous plus stables que ceux préparés par refroidissement à partir de l’état liquide. Toutefois, le verre formé du composé avec un groupement de tête donneur de liaisons H est moins stable que les autres d’au moins un ordre de grandeur, suggérant que les liaisons H limitent le niveau de stabilité atteignable par PVD. Finalement, un verre à base de triazine fonctionnalisé avec un groupement azo est employé pour étudier d’un point de vue moléculaire les perturbations provoquées par la photoisomérisation de l’azo. Grâce à une nouvelle méthode de spectroscopie IR, il est possible d’observer un gradient d’environnement moléculaire le long de la molécule lors de la photoisomérisation, permettant de soutenir certaines hypothèses relatives au déplacement macroscopique de la matière qui en résulte. Les mélanges de verres à base de triazine servent aussi de plateforme idéale pour découpler l’influence de la Tg et du contenu en azo sur la photo-orientation de l’azo, mais aussi sur la cinétique d’écriture et l’efficacité des réseaux de diffraction (SRG). Ce travail permet ainsi de déterminer une zone optimale de Tg pour l’inscription de SRG. Ces nouvelles connaissances mèneront à la conception plus rationnelle de nouveaux verres moléculaires, pouvant s’étendre à d’autres matériaux amorphes.
Amorphous molecular materials, also known as molecular glasses, are small organic molecules capable of being organized in a disordered manner. In addition to sharing some of the useful properties of polymers, they offer additional advantages because they are isomolecular species for which synthesis, purification and processing are facilitated by a relatively low viscosity. However, the usually demanding preparation conditions of these materials and their limited functional lifetime due to their tendency to relax to the crystalline state remain obstacles to their use for certain applications, e.g. opto-electronics, nanolithography, pharmaceuticals. The development of strategies to facilitate the preparation of the vitreous phase and avoid its crystallization is therefore essential for the design of functional amorphous molecular materials. The main objective of this thesis is to establish relationships between the molecular structure of molecular glasses and their properties. To achieve it, various libraries of model compounds, analogues of triazine derivatives that have demonstrated excellent glass-forming ability, are used to i) determine the influence of the nature and the position of the groups on the triazine; ii) explore the influence of hydrogen (H) bonds on the properties of glasses when their structure includes functional groups known to facilitate crystallization and when their preparation conditions are similar to those used in industry; and iii) exploit the amorphous phase in order to study the photoresponsiveness of azobenzenes (azo) in order to optimize their use in different applications. The influence of the various substituent groups on the triazine (headgroup, ancillary and linkers) on the glass-forming ability (GFA), the kinetic glass stability (GS) and the glass transition temperature (Tg) of the compounds is first studied. A classification system based on differential scanning calorimetry (DSC) and variable temperature infrared spectroscopy (IR) measurements combined to chemometrics analyses facilitate the rationalization of the roles played by each group. The impact of the H-bonds, the energy of the rotation barrier, and the steric hindrance of the groups is determined, leading to the conclusion that the headgroup is the most influential group and that the presence of H-bonds is not essential to the GFA, but important to obtain a high Tg. The influence of the H-bonds on the properties of glasses approaching those exploited in industry is then explored. Variable temperature IR spectroscopy measurements, DSC studies, and single crystal structure resolution have led to the conclusion that H-bonds impede the crystallization of the compounds even though they are simultaneously functionalized with moieties that promote crystallization (π-π stacking between fluorinated and non-fluorinated stilbene groups). In addition, three similar compounds functionalized with a headgroup presenting a decreasing capability to establish H-bonds (donor, acceptor, none) were vapor-deposited (PVD), a technique used, among others, in the opto-electronic industry, to evaluate their capability to form ultrastable glasses. These PVD glasses all show properties that are similar to those previously reported for ultrastable glasses, including higher density and anisotropy, and are all more kinetically stable than glasses prepared by cooling from the viscous state. However, the PVD glasses prepared with a H-bond donor headgroup are less stable than the others by at least an order of magnitude, suggesting that H-bonds limit the level of kinetic stability achievable by PVD. Finally, a triazine molecular glass functionalized with an azo group is used to study, from a molecular point of view, the perturbations caused by the photoisomerization of the azo. A new IR spectroscopy method was developed to observe a molecular environment gradient along the molecule during photoisomerization, making it possible to support certain hypotheses concerning the resulting macroscopic transport of the material. Triazine-based molecular glass blends are also used as an ideal platform for decoupling the influence of Tg and azo content on the azo photo-orientation, but also on the inscription kinetics and the diffraction efficiency of surface relief gratings (SRGs). This work enables the determination of an optimal Tg range for the inscription of SRGs. Altogether, these new insights will lead to a more rational design of new molecular glasses, which can extend to other amorphous molecular materials.