In order to be able to have a solid understanding of the species, phases and chemical interactions present in the materials of a molten salt reactor, the gap between the neutronic, isotope-based fuel evolution codes and the chemical-thermodynamic codes has to be bridged. To achieve this, the EQL0D code, which uses Serpent and MATLAB to simulate the evolution of materials in an MSR, was combined with the Gibbs energy minimization software GEMS in the scope of this thesis. As a first step, basic cases were created and run for both GEMS as well as EQL0D, in order to validate the underlying chemical-thermodynamic database HERACLES and for familiarization purposes with EQL0D. For GEMS, these validation cases take the form of binary phase diagrams, wherein both the substance-specific, thermodynamic parameters as well as binary liquid phase interaction parameters were adjusted to match literature diagrams. Where necessary, new species were added to the HERACLES databse. In EQL0D, many old cases were rerun, adjusted and analysed in order to facilitate the creation of four simplified, homogenized input structures which are to serve as the baseline for speciation work performed in GEMS. Once the composition of the equilibrium salt mixture is obtained, it is put into GEMS, which will output the according speciation. In this work, the LiF-UF4 system was selected. For the sake of simplicity as well as reducing the amount of required database work of the simulations, only the ten most prevalent elements are included. Then, some analysis in terms of the influence of redox conditions as well as temperature on the outcome of the speciation is performed. In addition to the above chain of simulations, some tasks not falling within the chain of analyses outline above were completed. Using the updated thermodynamic database, cGEMS, a tool combining the release analysis of MELCOR with the powerful thermodynamics of GEMS was run in order to simulate the release from a given salt mixture. The results are compared with literature results obtained from the same input prior to the modifications made to the database. EQL0D is used extensively for calculations as part of the SAMOSAFER benchmarking process. Multiple codes from different institutes are compared against each other using incremental benchmarking, with the goal being to start simple and ramp up the complexity of the system to the realistic case. Results from various different cases are presented in this work. On top of the as of yet quite basic benchmarking runs, a full-scale MSFR case is run, analysed and presented.