Since decades, nucleation, or the ability of certain organic or inorganic substances to trigger the crystal growth, has been empirically used in the plastics industry. Talc, for instance, is a well-known nucleating agent of poly(ethylene terephthalate) (PET) and other polymers, that allows one to enhance the crystallisation rate of the polymer material and to control its spherulites size. The exact mechanism involved in this nucleation had however remained unknown at the onset of this thesis. Through electron diffraction, performed on thin PET films nucleated by macroscopic talc particles as model samples, this work demonstrates an epitaxial relationship between polymer and substrate and thus confirms the seemingly ubiquitous role of epitaxy in the nucleation of polymers. However, in order to compare the talc-nucleated morphology of PET with the virgin one, new methods of sample preparation for transmission electron microscopy (TEM) have also been developed. Coupled with theoretically justified image analysis techniques, they allow the direct observation of PET crystalline lamellae, both in the bulk and in thin films. Analyses of the semicrystalline structure in the reciprocal and direct spaces were performed from small-angle X-ray scattering (SAXS) data and from observations by TEM on melt-crystallised samples. These independent results were shown to be in good agreement and bring strong evidence in favour of a semicrystalline space-filling model, where the average crystalline thickness is slightly smaller than the average width of the amorphous regions. Discrepancies between characteristic distances derived by several methods from the same experimental results were attributed to the broad distribution of thicknesses, in contrast with the ideal linear stack model commonly used to analyse the data.

Microstructural data on the orientation and distribution of minerals can be utilized to better understand the processes controlling mineral crystallization during metamorphism and the extent to which equilibrium versus kinetic factors control the evolution of metamorphic rocks. Four studies in this dissertation address this, finding that: i) garnet crystals crystallize via epitaxial nucleation in which garnet crystallizes by templating on the crystal structure of muscovite; ii) the distribution of grain boundary void space at quartz-quartz and garnet-quartz grain boundaries is a function of the orientation of quartz crystals on either side of the grain boundary. There are more voids, and in some cases larger voids, at grain boundaries in which the a-axis of a neighboring quartz grain is perpendicular to the grain boundary than any other orientation; iii) the chemical potentials of garnet-forming components evolve differently in samples in which garnet growth either significantly or minimally overstepped equilibrium garnet-forming reactions; iv) the southwestern Meguma Terrane, Nova Scotia, experienced peak metamorphic conditions of ~630ºC and 4.0 kbar, likely resulting from regional metamorphism during the Neoacadian orogeny. A case study on the mechanisms controlling garnet crystallization in one Nova Scotian sample reveals that the rate limiting step of garnet crystallization was probably the diffusional transport of Al through the intergranular matrix. Taken together, this work has implications for understanding: i) the properties of grain boundaries in metamorphic rocks and ii) the extent to which equilibrium versus kinetic factors impact metamorphic petrogenesis. / Doctor of Philosophy / A fundamental question in the development of metamorphic rocks, or rocks that form due to changes in pressure and temperature conditions deep in the Earth's mountain belts, is: what controls the crystallization of new minerals? While pressure, temperature, and bulk composition likely play a major role in this, it is also possible that the distribution of reactant minerals and the transport of elements through the rock may also play a role in mineral crystallization. This dissertation explores several projects related to this broad topic. In one example, garnet, an important metamorphic mineral, was found to crystallize by utilizing the atomic structure of another mineral in the rock. This creates a favorable pathway for the crystallization of garnet, which preferentially grows on this 'parent' mineral. Further, the distribution of porosity, or void space, at the interfaces between mineral grains in metamorphic rocks is found to be controlled by the orientation of those minerals. This porosity likely formed when the rocks were exhumed from deep in the Earth towards its surface. Metamorphic rocks can also tell the story of continental plates colliding millions of years ago. In an example from the formation of the Appalachian Mountains ~400 million years ago, a combination of collisional tectonic forces and the heat from magmas in the shallow crust resulted in metamorphic rock, which make up much of southern Nova Scotia today. This work has important implications for understanding: i) porosity in metamorphic rocks and ii) how minerals crystallize during metamorphism.