Although rare, some inclusions in so called superdeep diamonds appear to originate below the lithospheric mantle, raising questions as to how they form under these conditions and, therefore, how the deep carbon cycle operates. Ferropericlase, (Mg,Fe)O, makes up ~20 vol% of the lower mantle and is found in ~50% of superdeep diamonds. Although it is often argued that such diamonds come from the lower mantle, ferropericlase can in principle form at shallower depths from SiO₂-poor materials. In addition to Fe2+, synthetic and natural ferropericlase also contains Fe3+. At room pressure the Fe3+/Fetot ratio of ferropericlase varies strongly with oxygen fugacity (fO₂), reaching a minimum where ferropericlase is in equilibrium with iron metal and a maximum where it coexists with spinel ferrite (Fe₃O₄-MgFe₂O₄) solid solution. Ferropericlase inclusions in diamonds also show evidence for the exsolution of a spinel-ferrite phase. To interpret the fO₂ at which these inclusions formed and to understand why spinel-ferrite exsolution occurs, experiments were performed to measure ferropericlase Fe3+/Fetot over a range of fO₂, up to its coexistence with spinel-ferrite, and at higher pressures up to 30 GPa, where it coexists with recently discovered mixed-valance Fe and Mg-oxides. Once the fO₂ rises above a certain threshold, all samples contain nanocrystalline topotaxial exsolutions of spinel-ferrite, that exsolve during quenching. Characteristics of the Mössbauer spectra make it possible to determine the original ferropericlase Fe3+/Fetot ratio from the contribution of the exsolution and to reconstruct the high temperature composition. Using these results and measurements made using electron energy loss spectroscopy, a thermodynamic model has been developed that describes the entire range of ferropericlase compositions in the system Mg-Fe-O between 1atm and up to 30 GPa. A poorly constrained parameter in this thermodynamic treatment is the compressibility of the spinel-ferrite phase. To resolve this, X-ray diffraction experiments were performed in a diamond anvil cell on single crystals of pure MgFe₂O₄, as well as on an intermediate composition. The compressibility changes non-linearly with Mg-content across the solid solution, may be due to an interruption of Fe2+-Fe3+ electron hopping by introducing Mg cations into the spinel structure. The resulting thermodynamic model indicates that the ferropericlase Fe3+/Fetot ratio in equilibrium with spinel ferrite increases slightly with pressure. Above 10 GPa, however, after the phase transformation of spinel ferrite to the high-pressure oxide, [Fe2+, Mg]₂Fe3+₂O₅, the Fe3+/Fetot ratio declines rapidly reaching values at 30 GPa, that would exclude some ferropericlase inclusions in diamonds from being formed in the lower mantle, since they have Fe3+/Fetot ratios up to 0.12. Using this model, the ferropericlase Fe3+/Fetot ratio in equilibrium with diamond and (Mg,Fe)CO₃ magnesite can be calculated, defining a redox buffer, with the acronym FDM, that imposes an effective limit on the possible fO₂ at which ferropericlase inclusions in diamond can form. The ferropericlase Fe3+/Fetot ratio at FDM also decreases with pressure, indicating that only a few of the ferropericlase inclusions found in diamonds so far could have formed at the top of the lower mantle and if so, they must have formed at relatively oxidising conditions from carbonate minerals or melts. Many of the other inclusions have maximum formation depths in the transition zone. A few of the most oxidised inclusions could have only formed in the upper mantle and only from more oxidised precursors, implying that ferropericlase inclusions are formed through reduction of carbonates at depths mainly shallower than the lower mantle. Interestingly the FDM fO2 curves for ferropericlase inclusions with Fe/(Fe+Mg) ratios > 0.5 intersect the stability of [Fe2+, Mg]₂Fe3+₂O₅, implying that this, or an [Fe2+, Mg]₃ Fe3+₂O₆ high-pressure oxide, could be trapped in diamond at near lower mantle conditions. In a separate study performed during a three-month research visit to Tohoku University the nitrogen contents and nitrogen stable isotope ratios of several hydrous minerals and diamond were measured, after high pressure and temperature equilibration with a fluid produced by the decomposition of glycine (C₄H₅NO₂). Phlogopite, antigorite, lizardite and montmorillonite were equilibrated at 3 GPa and 500°C. The recovered samples incorporated N contents between 0.35 and 2.14 wt% which were enriched in ¹⁵N compared to the glycine starting material. This suggests that the passage of N-bearing fluids through hydrous mineral-bearing assemblages in subduction zones could cause N isotope fractionation in the mantle. Diamonds were also synthesised from graphite in the presence of an Fe-Ni-S catalyst and glycine at 13 GPa and 1700°C. N was found to partition evenly between diamond and the coexisting metallic catalyst but the diamond became depleted in ¹⁵N compared to the glycine starting material. This implies that, during diamond growth processes, coexisting fluids should become progressively ¹⁵N enriched.