ON THE INVESTIGATION OF NUCLEATION, GROWTH, STRUCTURE AND MORPHOLOGY OF FUNCTIONAL MATERIALS THROUGH ADVANCED CRYSTALLOGRAPHIC TECHNIQUES

Title of the thesis On the investigation of nucleation, growth, structure and morphology of functional materials through advanced crystallographic techniques Description of the research work done I-Use of gel technique for the crystallization of coordination polymers A gel is a system where a liquid phase is dispersed within a solid phase. A network of micrometric channels and pores filled by solution characterizes the obtained material. Any mass transport mechanism but slow diffusion is avoided. This creates an interesting media to perform crystal nucleation and growth under purely diffusive conditions. Recently gel technique for crystallization has been extensively adopted, leading to intriguing results about crystals quality, dimension, mechanical and chemical stability. For certain organic molecule were reported the gel ability to induce the formation of new phases and polymorphs. Inside gel media, slightly different crystallization conditions can be simultaneously created. For a class of compounds known as coordination polymers or MOF, little difference in the crystallization environment can lead the synthesis towards the formation of network completely different from the point of view of dimensionality, topology, coordination geometry of metal centres and ligand behaviours. The aim of the following projects is to map the relevance of gel technique if applied for the synthesis of coordination polymers, considering parameters as solvent composition, gel nature, reagents concentrations ([L]+[M]) and ratio([L]/[M]). I-A system Cd(II) bpp We considered water/EtOH solutions of CdCl2 (M) and bis-Pyridyl-propane (bpp) and agarose gel at different concentration. Three different crystalline phases can be obtained: 1D [Cd(bpp)3Cl2]∙H2O (1D-Cd) (L/M=3) 3D [Cd(bpp)2Cl2]·H2O (3D-dia) (L/M=2) 3D [Cd2(bpp)3Cl2]·H2O (3D-sra) (L/M=1,5) The phases show different compositions in terms of L/M ratio, and also different dimensionality and topology. Their morphologies also are peculiar enough to let they be straightforwardly identified by optical microscopy. We made a strict comparison between system behaviour inside solution and jellified solution, to point out the most important gel effects on crystallization process. For solution studies, we applied batch technique. For gel studies, we used counterdiffusion methods. We used two different experimental set up: (a) Layering of the solutions of a reagent over the gelified solution of the other reagents inside a test tube; (b) Layering in the opposite branches of a U-shaped tube the solutions of the reagents over a gelified solvent. I-A.1 solution behaviour We determined the precipitated phases considering different reagent concentration and ratio at fixed EtOH amount (50%) and different amounts of EtOH and reagents ratio at a fixed concentration (20mM). To confirm the results obtained, we also study powders precipitated from solutions at higher concentrations with the same experimental conditions. 3D-sra phase forms at low L/M value (<2) for any solvent compositions. Increasing [L]/[M] value bring to two different behaviours. For EtOH quantity below than 30% 1D-Cd precipitates as powder, above 30% we observe precipitation of 3D-dia. From solutions, we never observe co-precipitation of all the three phases. But 3D-sra and 3D-dia coexist inside alcoholic solutions. I-A.2 Gel technique modifies the phase diagram When we laid ligand alcoholic solutions over water-gelified solutions of the metal salt, the crystals form more in the solutions than in gel. The sra phase appears for a narrower range of L/M values, while dia phase becomes predominant. Instead, when we layer metal salt solution over the ligand solution, the two phases coexist for all the considered value of L/M. They are segregated: sra phase inside the solution, the 3D-dia inside the gel. We suggest that the gel slows selectively the diffusion of bpp, segregating the phases. The gel strongly affects crystals morphologies. Dia phase forms largely elongated and distorted structures inside the gel, while 3D-sra inside gel media forms little spherulites. I-A.3 The phase diagram can be mapped inside a U-shaped tube We perform counterdiffusion experiment inside a U-shaped tube. In the first place we consider tubes whit the same gelled solutions (1% agarose, 40% ethanol), varying only reagents concentrations and ratio. We observed the formation, starting from the ligand charged branch, of phase dia and then phase sra. Inside tube containing the excess of metal, dia phase forms only inside ligand solution, while using ligand excess we observed a transition from a tube containing both the phases to the disappearance of the sra phase. This behaviour confirms that the gel phase probably slows the diffusion of the ligand solution. After that, we studied the effect of the solvent composition. Inside gel, we consider %EtOH from 15% to 40%. We observe coprecipitation of all the three phases for %EtOH<20%. They form starting from the ligand branch and their order is 1D-Cd, 3D-dia, 3D-sra. I-A.4- The use of different gel influences the phase diagram and crystals morphologies Agarose dispersed in water o ethanol-water solvents is the gel media that we choose. Agarose dispersion forms gel at different concentrations, expressed as %w/v, ranging from 0.1% to 3%. Increasing agarose % creates a stiffer material. We didn’t observe relevant effects on the phase diagram, while crystal morphology shows increasing habit distortion increasing gel strength. Then we considered different gel media to fill the U-shaped tube. As gelator, we considered silica, Polyacrylamide (PAA), polyvinylpyrrolidone (PVP). All gels considered kept the same reciprocal disposition of the phases. But 1D-Cd isn't formed inside PVP, PAA gels strongly inhibit 3D-dia nucleation, also shifting the crystal position nearer to the metal solution branch, while silica gels move phases position in opposite directions. Crystals morphology is strongly affected by gel strength; in fact, we observe no modification only inside the very light PVP gel. Instead, inside silica gels, 3D-dia crystal modifies its habit. In fact, We observed the prismatic form {100} becoming predominant. I-B System Cu(II)/isonicotinic acid Isonicotinic acid (HIso) is a widely used ligand. It is a linear, rigid connector. Its coordinative sites are also donor or acceptor of hydrogen bonds useful for the assembly of supramolecular frameworks. It’s an anionic ligand, so its MOF structures are neutral, avoiding any templating effect of the counter anion. In CSD databased is reported a large number of structures with different topologies and dimensionality. In particular, Copper(II) atom containing structure show peculiar ligand behaviour and coordination geometry. In this project, we mapped the species obtained from the mixing of Copper chloride salts and isonicotinic acid at room temperature, the parameters considered are the same as the previous project, total concentration ([Cu(II)]+[HIso]), the molar ratio([HIso]/ [Cu(II)]). As the solvents, we considered water, THF, and DMSO. We considered the previously described crystallization method. Five phases were obtained: 1 - 0D complex [Cu(Iso)2(H2O)4] formed by single distorted copper octahedral complexes. 2 - 2D MOF [Cu2(Iso)3Cl(H2O)2] ∙ THF formed by a 2D network staked along [001] directions 3 – 2D MOF [Cu(Iso)2(H2O)]∙THF formed by a 2D network staked along [001] directions 4 – 2D MOF [Cu4(Iso)6Cl2(H2O)2(DMSO)] formed by a 2D network staked along [110] directions 5 - 2D MOF [Cu2(Iso)2(H20)] ∙ DMSO formed by a 2D network staked along [001] Working inside water/THF solutions, phase 1 and 2 have been isolated. Phase 2 is strongly favoured both by low values of L/M ratio and high %THF in the precipitating solvent. Phase 1 precipitates from water solutions or solutions with high L/M value. To observed gel effect we used a limited amount of experimental conditions. We consider an equal amount of the reagents (L/M=1) or metal excess (L/M=0,5), and concentration 50 or 75 mM. About solvent compositions, working inside test tubes, we consider all the possible combination of water and water/THF solutions. The use of only water or water gel bring to the formation exclusively of phase 1, instead, water/THF gels and THF rich environment are characterized by the formation of phases 2 (L/M=0,5) and 3 (L/M=1). This last phase is highly unstable out of the gel and has never been detected in the solution. Counterdiffusion inside U-shaped tube, using water/THF solutions of equal concentration, lets to precipitate phase 1 (near ligand charged branch) and phase 3 (right behind it). This tells us an higher mobility of the metal salt respect to the ligand molecule inside the gel media, and the existence of a range of L/M value was phase 3 is favoured respect to phase 1, where from the solution we can only precipitate phase 1. Has to be determined if phase 3 is stabilized due to peculiar interaction interplayed with the gel media or it’s a metastable phase that can be observed due to the slowing of the degradation kinetic inside the gel media. Using as solvent DMSO/THF mixtures we obtain interesting results. Also in this case from solution can precipitate two different phases, both of them are 2D MOF. For this mixture, the presence of THF doesn’t change the nature of the obtained phase. For L/M=0,7 we obtained phase 4 while phase 5 was obtained for a higher concentration of ligand (L/M=2,7). The two phases can coexist for L/M=1,3, but phase 5 is unstable outside solution. Considering crystallization conditions, stability outside solution, topology, dimensionality, coordination geometry, and ligand behaviour phase 2 and 4 show strong similarity, as phases 3 and 5. We proposed that a simple parameter as L/M can have deep effects on the solution behaviours of the reagents and then determine the topology and dimensionality of the precipitated phase. II-Morphology prediction Inside gel media, the only viable mass transport mechanism is the diffusion troughs the small pores of the media. If reagent diffusion is enough slowed, it is the rate-determining step of crystal growth. That means that the observed morphology is determined only by the stability of the faces, that is proportional to the surface tensions. This happens because the energy request to bind a molecule on a certain face (attachment energy) that is the most important kinetic parameter, is less determining. The former situation leads to the formation of the equilibrium morphology, while the latter to growth morphologies. Being gel media pure diffusive, in principles crystals grew inside gel media should show a morphology nearer to the equilibrium one, not considering peculiar interactions between faces and gel media itself. The simulation of crystal habit and morphology is not trivial for coordination polymer due to two issues: the selection of a proper building unit that is actually added to the growing crystal and the choice of an appropriate potential to model the interactions. Our purpose is to study from an experimental point of view the effect on crystal habit of gel media, and then, to verify the feasibility of PIXEL package for simulation of the morphology. The 0D Cu-HISO (phase 1, project B) system shows important morphological modification if growth inside gel or inside solutions with different concentrations. We identify four different crystal habits. All of them can be described by some the following six forms: {101 ̅}, {010}, {011 ̅ }, {11 ̅0},{111 ̅},{001}. Inside the solutions, the crystals are platelets and the most important form is the {101 ̅ }. We observed an evolution of the habit increasing the reagents concentration, from rhombic platelet (forms {101 ̅}, {010}, {011 ̅}) to hexagonal platelets with increasing importance of form {001}. Inside gel, we observe only prisms, where forms {101 ̅} and {010} are predominant and they have the same importance. We embarked on the simulation of both equilibrium and growth morphology, in the framework of Hartman-Perdok (HP) theory, using PIXEL method for energetic calculations, to verify if habit simulation can shed some light on these behaviours. HP approach considers a crystal a fabric of chain of molecules bounded by strong interactions (polymeric bond chain, PBC). Every chain represents a crystallographic direction of fast growth. If a face is parallel to at least two of these chains, the energy request to attach a molecule over it is presumably higher than the energy requested to attach a molecule to its border. As a result, this faces will grow in extension and is classified as F (flat). The equilibrium morphology contains only flat faces. Attachment energy calculations, considering as the starting point a reasonable bunch of faces (the flat ones), let us know the growth morphology. Then, towards the broken bonds model, from this, we can calculate the surface tension of the single face and then the equilibrium morphology. All energetic evaluations were performed using PIXEL formalism. Originally developed for closed-shell, neutral organic compounds, it has been subsequently extended to deal with ionic compounds, salts, and zwitterions. The PIXEL formalism and parameterization have been recently extended systematically to compounds including first-row transition metals in combination with organic moieties. This extension has been carried out using an extensive set of thermochemical data, successfully reproducing the sublimation enthalpies of a wide range of crystalline materials. PIXEL treat pecking energies as a summation of single dimer interactions, bringing straightforwardly to the identification of strong bonds and then of PBC. The identified PBC let us classify as flat all the observed crystal forms apart from the plane (001), that is classified K (kinked) the unstable kind. Then, we calculate the attachment energies and surface tensions for all the identified flat faces. The calculated equilibrium morphology is very similar whit the bars morphology. This morphology is typical of low concentration environment and gel media. This is the experimental equilibrium habit. This confirms the goodness of PIXEL simulation. This also confirms that crystals grew inside gel media shows their equilibrium morphology, suggesting that inside gel mass transport is only diffusive and it’s the rate determining step of the growth. The balance between kinetic and thermodynamic factors for the determination of the morphology is well described by PIXEL calculated values. Form {111 ̅} from both calculation and experiments reveal to be relevant in the equilibrium morphology but almost invisible during the growth process, while form {011 ̅} shows exactly the opposite behavior. The importance of form {101 ̅} is not explained by energetic values. However, some considerations about its structure suggest that this is the only face without a net charge over itself, creating an easily poisoned surface. Form {001} is completely different, and here we can suggest an opposite behaviour, where the large negative charge on its surface can probably create a kinetic barrier that became more effective at high concentration fast growth process. However, its K nature is significant considering that its appearance is strongly affected by experimental conditions. III-Cholesterol crystallization from model lipid bilayers Atherosclerotic diseases are one of the major death cause in the western world. A crucial step in the development of the disease is the rupture of the atherosclerotic plaques end the resulting inflammatory response. It's well known that plaques are formed by different lipids, in particular, cholesterol, and the presence of cholesterol crystals is indicative of the pathologic ones. The presence of cholesterol crystals can be related also with the triggering of the inflammatory response itself, but a clear correlation was not yet established. There is not much information about the process of cholesterol crystal formation and deposition inside plaque, a lipid-rich environment. Different studies where performed inside bile, because cholesterol crystal triggers the formation of the gallbladders. From these studies was observed the ability of cholesterol to form different crystalline materials, from the typical platelet of triclinic monohydrate cholesterol to more unusual material as curved fibres and helices. In recent years some observations were performed on the behaviours of lipid mono and bilayer. A result from this study was the identification of 2D lipid structures called ‘’rafts’’, usually formed by saturated lipids as dipalmitoylphosphatidylcholine (DPPC) or sphingomyelin (SM), that move inside a liquid 2D phase of unsaturated lipids as palmitoyl oleoyl phosphatidylcholine (POPC). Cholesterol molecules also form ordered lipid domains, sometimes by mixing with other saturated lipids. For all the cited lipids was identified a concentration threshold above which Cholesterol phase separates and forms such rafts. Cholesterol rafts are characterized by a rectangular 2D symmetry, not compatible with the already known triclinic structure. Recently was discovered a correlation between the presence of such ordered cholesterol domains and the appearance of cholesterol crystal over lipid bilayer. The purpose of the project was to study the crystallization process. In the first place identifying the cholesterol phases that can be obtained, then, adopting super-resolved fluorescence microscopy (STORM method), obtaining more information about nucleation and growth processes. Lipid bilayers were formed by vesicles fusion and were cholesterol-enriched using a well know procedure adopting a cholesterol: cyclodextrin complex. We considered the mixture of cholesterol with unsaturated (POPC) and saturated (DPPC) lipids. For fluorescence microscopy, we used an IgM antibody developed in Lia Addadi group, able to recognize ordered cholesterol patterns as crystals or domains. TEM microscopy and electron diffraction were performed to identify the crystalline phases. Over POPC/Chol bilayer we observed exclusively the formation of square platelet crystals, while over POPC/DPPC/Chol bilayer we identified also a large number of curved fibres, needle, and hexagonal platelets. TEM diffraction demonstrated that the square platelets consist of triclinic cholesterol monohydrate phase. While all the other morphology belongs to crystals of a monoclinic cholesterol monohydrate phase, that was never be identified before. The morphology of such phase is intriguing because atherosclerotic plaques often contain needle crystal, that cannot be formed by the triclinic phase, suggesting a relevant role for the monoclinic phase for pathological environments. Crystals grown from murine cells fed by HDL particles show the same characteristics. Further analysis performed with STORM technique revealed very different behaviours of cholesterol domains inside the two considered bilayers (POPC/Chol, POPC/DPPC/Chol). Also, the crystallization environment of the two phases shows interesting differences. POPC bilayer is characterized by the presence of small Cholesterol domains, accordingly with the low solubility of Cholesterol in unsaturated lipids. They can aggregate forming larger domains that already show shapes similar to the triclinic platelet, suggesting a sudden aggregation and formation of triclinic phase crystals. DPPC bilayers show the presence of slightly bigger domains, homogenously diffused over the bilayer. The triclinic phase crystal is often surrounded by large not labelled areas, that can be related with large cholesterol ordered domains, suggesting a slower process of aggregation and crystal formation over the bilayer. Instead, monoclinic crystals are often found grown inside the bilayer itself. They reveal to be very flexible, bending and assuming curved morphologies from their very first appearance.