La cristallisation est un processus d’auto-assemblement de molécules à partir d’une phase désordonnée (liquide, amorphe ou gazeuse). De façon générale, la miscibilité à l’état solide entre un constituant d’intérêt et son impureté dépendent d’une part, de la structure moléculaire de ces deux espèces, et d’autre part, de la structure cristalline du réseau hôte, c’est-à-dire celui du composé d’intérêt. En cas d’absence de miscibilité, l’optimisation des conditions de cristallisation permet un retrait total de l’impureté du composé visé. Cette thèse vise à démontrer que, lorsque deux molécules sont semblables et que leur structure cristalline est peu dense, des solutions solides de substitutions peuvent se former et empêcher le retrait de l’impureté de l’espèce cible. Inversement, lorsque les impuretés sont très diférentes de l’espèce à purifer, l’absence de solution solide est observée et permet leur élimination totale grâce à une bonne discrimination à l’état solide. Le système du phénanthrène, un composé servant entre autres de base pour la synthèse de composésmorphiniques, a été choisi pour illustrer cette thèse. Diverses méthodes de cristallisation (fusion de zone, co-cristallisation, recristallisation en solution, sublimation-condensation) ont été testées et ont permis, après optimisation et combinaison, l’obtention à l’échelle préparative de phénanthrène pur à plus de 99,999 % en moles, ce qui le place sur l’échelle de l’ultra-pureté (pureté > 99,9 % en moles). / Crystallization is a process during which molecules self assembly from a disordered (liquid, amorphous or gaseous) phase. The miscibility in the solid state between a component of interest and its impurity depends on: (i) their molecular structures, (ii) the crystal structure of the host lattice (namely, that of the component of interest). When no solid solution exists, the impurity can be completely eliminated from the target product after optimization of the crystallization conditions.The present thesis intends to demonstrate that, when two compounds present similar molecular structures with low-density crystal structures, solid solution formation can occur which prevents from their complete separation. Conversely, in case of sufcient dissimilarity, no solid solution is stable and their separation is possible thanks to a large discrimination in the solid state. The phenanthrene system – a model compound used as base for the synthesis of morphine derivatives – was chosen to illustrate this thesis. Several crystallization methods (zone melting, co-crystallization, solvent assisted crystallization and sublimation-condensation) were investigated and permitted, after optimization and combination, to obtain 99.999(9) mole % purity phenanthrene (which is above the ultrapure grade of 99.9 mole %).

Narrow molecular weight fractions of Poly(ε-caprolactone) were successfully obtained using the successive precipitation fractionation technique with toluene/n-heptane as a solvent/nonsolvent pair. Calorimetric studies of the melting behavior of fractions that were crystallized either isothermally or under constant cooling rate conditions suggested that the isothermal crystallization of the samples should be used for a proper evaluation of the molecular weight dependence of the observed melting temperature and degree of crystallinity in PCL. The molecular weight and temperature dependence of the spherulitic growth rate of fractions was studied in the context of the Lauritzen-Hoffman two-phase model and the Strobl three-phase model of polymer crystallization. The zero-growth rate temperatures, determined from spherulitic growth rates using four different methods, are consistent with each other and increase with chain length. The concomitant increase in the apparent secondary nucleation constant was attributed to two factors. First, for longer chains there is an increase in the probability that crystalline stems belong to loose chain-folds, hence, an increase in fold surface free energy. It is speculated that the increase in loose folding and resulting decrease in crystallinity with increasing chain length are associated with the ester group registration requirement in PCL crystals. The second contribution to the apparent nucleation constant arises from chain friction associated with segmental transport across the melt/crystal interface. These factors were responsible for the much stronger chain length dependence of spherulitic growth rates at fixed undercooling observed here with PCL than previously reported for PE and PEO. In the case of PCL, the scaling exponent associated with the chain length dependence of spherulitic growth rates exceeds the upper theoretical bound of 2 predicted from the Brochard-DeGennes chain pullout model. Observation that zero-growth and equilibrium melting temperature values are identical with each other within the uncertainty of their determinations casts serious doubt on the validity of Strobl three-phase model. A novel method is proposed to determine the Porod constant necessary to extrapolate the small angle X-ray scattering intensity data to large scattering vectors. The one-dimensional correlation function determined using this Porod constant yielded the values of lamellar crystal thickness, which were similar to these estimated using the Hosemann-Bagchi Paracrystalline Lattice model. The temperature dependence of the lamellar crystal thickness was consistent with both LH and the Strobl model of polymer crystallization. However, in contrast to the predictions of Strobl’s model, the value of the mesomorph-to-crystal equilibrium transition temperature was very close to the zero-growth temperature. Moreover, the lateral block sizes (obtained using wide angle X-ray diffraction) and the lamellar thicknesses were not found to be controlled by the mesomorph-to-crystal equilibrium transition temperature. Hence, we concluded that the crystallization of PCL is not mediated by a mesophase. Metallocene-catalyzed linear low-density (m-LLDPE with 3.4 mol% 1-octene) and conventional low-density (LDPE) polyethylene blends of different compositions were investigated for their melt-state miscibility and concurrent crystallization tendency. Differential scanning calorimetric studies and morphological studies using atomic force microscopy confirm that these blends are miscible in the melt-state for all compositions. LDPE chains are found to crystallize concurrently with m-LLDPE chains during cooling in the m-LLDPE crystallization temperature range. While the extent of concurrent crystallization was found to be optimal in blends with highest m-LLDPE content studied, strong evidence was uncovered for the existence of a saturation effect in the concurrent crystallization behavior. This observation leads us to suggest that co-crystallization, rather than mere concurrent crystallization, of LDPE with m-LLDPE can indeed take place. Matching of the respective sequence length distributions in LDPE and m-LLDPE is suggested to control the extent of co-crystallization. / Ph. D.

This dissertation has focused on the solid-state characterization of different co-crystal system as well as the effect of co-crystallization of these systems on pharmaceutical dosage form performance. Urea/ 2-MB, caffeine/ malonic acid, caffeine/ oxalic acid and theophylline/ malonic acid co-crystals were prepared using co-grinding- and co-precipitation techniques. In addition, the synthesis of co-crystals through two novel methods has been demonstrated. This includes compaction and convection mixing. The solid-state characterization of the co-crystals has been carried out using XRPD, Raman spectroscopy, DSC, TGA, hot-stage microscopy and SEM. After preparation of co-crystals, tablets have been produced from co-ground-, co-precipitated-, and physical mixtures using Compaction Studies Press (Kaleva), and the data were recorded to compare between the different mixtures, regarding compactibilty, compressibility and deformational properties. The DSC results showed that the physical mixtures of all systems, formed co-crystals during heating process. For systems of urea/ 2-MB, caffeine/ malonic acid and theophylline/ malonic acid, the co-ground mixture produced tablets with higher tensile strength compared with either co-precipitated or physical mixture. However, for caffeine/ oxalic acid system, the tensile strengths of compacts produced from the physical mixture were greater than those obtained from either co-ground or co-precipitated mixtures. The Heckel data suggested that urea/ 2-MB, caffeine/ malonic acid and theophylline/ malonic acid systems are Type 1 materials, as an extensive linearity during compression was indicative of a plastic deformation mechanism, while the caffeine/ oxalic acid system was Type 2 materials. However, the co-precipitated mixture of urea/ 2-MB system was the least compressible, as it possessed the greatest value of yield pressure (85 MPa) and the highest elastic recovery (7.42%). The co-precipitated mixture of both of caffeine/ malonic acid and theophylline/ malonic acid systems was the most compressible with small yield pressure values of (44 & 80 MPa) and elastic recovery of (7.2% & 6.56%), respectively. The co-ground mixture of caffeine/ oxalic acid possessed the highest value of yield pressure (166 MPa) and thus the lowest compressibility among other mixtures. Furthermore, the addition of microcrystalline cellulose and α-lactose monohydrate has affected the crystallinity as well as the tableting properties of the co-crystals. After the addition of excipients, the tensile strength of compacts was about 2 times higher than any other mixture. Finally, urea/ 2-MB and caffeine/ malonic acid co-crystals were successfully synthesized through convection mixing and compaction. / Islamic University of Omdurman and the Ministry of Higher Education in Sudan

The thesis entitled “Structure-function control in organic co-crystals/salts via studies on polymorphism, phase transitions and stoichiometric variants” consists of five chapters. The main emphasis of the thesis is on two aspects, one to characterize co-crystal polymorphism in terms of propensity of intermolecular interactions to form co-crystals/salts or eutectics. The other aspect is to explore the feasibility of using such co-crystals/salts to exhibit properties like proton conduction, dielectric and ferroelectric behaviour. Gallic acid and its analogues possess functionalities to provide extensive hydrogen bonding capabilities and are chosen as the main component while the coformers are carefully selected such that they either accept or reject the hydrogen bonding offered. Such co-crystallization experiments therefore provide an opportunity to unravel the intricate details of the formation of crystalline polymorphs and/or eutectics at the molecular level. Further these co-crystal systems have been exploited to evaluate proton conductivity, dielectric and ferroelectric features since the focus is also on the design aspect of functional materials. In the context of identifying and utilizing Crystal Engineering tools, the discussions in the following chapters address not only the structural details but identify the required patterns and motifs to enable the design of multi-component co-crystals/salts and eutectics. In particular, the presence/absence of lattice water in gallic acid has been evaluated in terms of importing the required physical property to the system. Chapter 1 discusses the structural features of tetramorphic anhydrous co-crystals (1:1; which are synthon polymorphs) generated from a methanolic solution of gallic acid monohydrate and acetamide, all of which convert to a stable form on complete drying. The pathway to the stable form (1:3 co-crystal) is explained based on the variability in the hydrogen bonding patterns followed by lattice energy calculations. Chapter 2A studies the presence/absence and geometric disposition of hydroxyl functionality on hydroxybenzoic acids to drive the formation of co-crystal/eutectic in imide-carboxylic acid combinations. In Chapter 2B the crystal form diversity of gallic acid-succinimide co-crystals are evaluated with major implications towards the design and control of targeted multi-component crystal forms. The co-crystal obtained in this study shows a rare phenomenon of concomitant solvation besides concomitant polymorphism and thus making it difficult to obtain a phase-pure crystal form in bulk quantity. This issue has been resolved and formation of desired target solid form is demonstrated. Thus, this study addresses the nemesis issues of co-crystallization with implications in comprehending the kinetics and thermodynamics of the phenomenon in the goal of making desired materials. Chapter 3 focuses on the systematic co-crystallization of hydroxybenzoic acids with hexamine using liquid assisted grinding (LAG) which show facile solid state interconversion among different stoichiometric variants. The reversible interconversion brought about by varying both the acid and base components in tandem is shown to be a consequence of hydrogen bonded synthon modularity present in the crystal structures analyzed in this context. In Chapter 4A, the rationale for the proton conduction in hydrated/anhydrous salt/co-crystal of gallic acid - isoniazid is provided in terms of the structural characteristics and the conduction pathway is identified to follow Grotthuss like mechanism which is supplemented by theoretical calculations. Chapter 4B describes an extensive examination of the hydrated salt of gallic acid-isoniazid which unravels the irreversible nature of the dielectric property upon dehydration and suggests that the “ferroelectric like” behaviour is indeed not authenticated. This chapter brings out the significance role of lattice water in controlling the resulting physical property (dielectric/ferroelectric in this case). Chapter 5 describes the structural features of two hydrated quaternary salts of hydroxybenzoic acids-isoniazid-sulfuric acid and the phase transitions at both low and high temperatures are shown to be reversible. Single Crystal to Single Crystal (SCSC) in situ measurement corroborated by thermal and in situ Powder X-ray Diffraction studies proves the claim. Further, the properties exhibited by these materials are also governed by lattice water content.

The DNA binding proteins under starvation (Dps) from M. smegmatis are cage-like structures which internalize iron and bind DNA. They provide resistance to the cells from free radical damage, and physically protect the DNA from the harmful effects of reactive oxygen species by DNA compaction. The work compiled in this thesis has been an effort to study oligomerization and dynamics of iron metabolism by these nano-protein compartments. Chapter 1 gives a general introduction on stress, especially oxidative stress, and the ways bacteria fight back the host resistance systems. This has been elaborated from the point of view of the Dps proteins which is the focus of our work. Also, the competition for iron among the host and pathogens, and the modes of iron trafficking of the pathogens from host organisms has been summarized. Finally, the structural aspects of ferritin family proteins to which Dps belongs, has been discussed. Chapter 2 elaborates on the oligomerization pathways of the first M. smegmatis Dps MsDps1, which exists in vitro as two oligomeric forms. The GFP-tagging has been used to locate the Dps1 proteins by live cell imaging and the over-expression of these proteins during nutrient limiting conditions has been studied. The crystal structure of a point mutant F47E in the background of MsDps1, which shows no dodecamerization in vitro, has been solved. The possible ways of dodecamerization of MsDps1 has been concluded by analyzing the intermediates via glutaraldehyde cross-linking and native electrospray mass spectrometry. Chapter 3 documents the gating machinery of iron in MsDps2 protein, the second M. smegmatis Dps protein. Through graph theoretical approaches, a tight histidine-aspartate cluster was identified at the ferritin-like trimeric pore which harbors the channel for the entry and exit of iron. Sitespecific variants of MsDps2 were generated to disrupt this ionic knot, and the mutants were further assayed for ferroxidation, iron uptake and iron release properties. Our studies in MsDps2 show the importance of counter-acting positive and negatively charged residues for efficient assimilation and dispersion of iron. Chapter 4 describes crystallization studies of MsDps2 pore variants, done in an attempt to connect the changes in functional properties described in chapter 3, with structural alterations of the point mutants. We show here that the gating mechanism happens by alterations in side chain configuration at the pore and does not alter the over-all stability of the proteins. Chapter 5 is the final section where we have employed site specific mutations and cocrystallization studies to elucidate the behaviour of MsDps2 proteins upon the addition of iron. By studying the effect of substitutions at conserved sites near ferroxidation center, we attempt to arrive at a pathway which iron atoms take to reach the ferroxidation site. Also, by crystallization of proteins loaded with varying amounts of iron we tried to map the changes in the protein structure in the presence of its ligand. Chapter 6 concludes briefly the work that has been documented in this thesis. Appendix I relates the role of N-terminal tail for DNA binding in MsDp2. Appendix II gives the technical details of a modified protein preparation and oligomerization process for his-tagged MsDps1 protein. Appendix III gives the maps of the plasmids used in this study.