Metal–organic frameworks are emerging as a highly functional class of hybrid materials. Once believed to be rigid and well-ordered structures, a growing number of studies are proving this is not always the case. This thesis focuses on the atomic structure, physical properties and formation mechanism of the amorphous Fe-BTC framework whose structure is very poorly understood, compared to its crystalline counterpart MIL-100, despite its catalytic and industrial significance.

Chapters 1 and 2 introduce metal–organic frameworks, the concept of disorder, and the techniques necessary for characterising amorphous MOFs.

Chapter 3 explores the structure of Fe-BTC. Using a combination of X-ray, electron and neutron-based techniques, coupled with computational modelling, the first model for Fe-BTC is produced. Electron microscopy reveals the presence of a nanocomposite structure that is dominated by an amorphous phase, while X-ray pair distribution function analysis reveals a mixed hierarchical local structure. The Fe-BTC model enables us to establish key structure–property relationships.

Chapter 4 investigates how the disordered structure of Fe-BTC impacts its gas sorption ability compared to MIL-100. While the porous network within MIL-100 facilitates high uptakes of gas, the disordered structure of Fe-BTC leads to the emergence of highly sought-after propane/propene separation capabilities. A range of gases are measured and coupled with surface area, virial, ideal adsorption solution theory and non-local density functional theory analyses. The results suggest disorder can be used to tune the adsorption capacity and selectivity in framework materials.

Chapter 5 studies how disorder can be progressively introduced into the MIL-100 framework as motivated by the previous chapter. Disorder is introduced using mechanical milling to break the metal–linker bonds and ultimately cause the collapse of the framework. Analysis of the pair distribution functions of the series of disordered MIL-100 materials reveals a stepwise collapse of the hierarchical structure with retention of the local structural building unit. Progressive milling of MIL-100 is found to be a facile route to tune the degree of long-range order and porosity.

Chapter 6 addresses one of the key challenges faced in this thesis, namely the sheer complexity of MIL-100’s crystallographic unit cell. Using multivariate analysis, mechanisms of structural disordering are studied using pair distribution function data in a manner inaccessible to computational modelling techniques. Rich kinetic and mechanistic information is found to be deeply encoded within the pair distribution functions. The analysis methodology is then applied to a very different disordered system to explore the approach’s generality.

Chapter 7 investigates the formation of Fe-BTC using in situ X-ray absorption spectroscopy and UV-vis spectroscopy to probe the redox behaviour that occurs during Fe-BTC formation. Kinetic and mechanistic analysis reveals a non-linear mechanism that occurs via the formation of a transient intermediate. These results are used to propose the first mechanism for Fe-BTC formation.

Chapter 8 summarises the results and discusses the avenues for future research.