Simulating heterogeneous catalytic processes occurring at solid-liquid interfaces requires the use of quantum mechanical methods to allow for a description of bond breaking and bond forming. These methods must also allow a sufficiently sized statistical sampling of the liquid to be gathered. Current approaches based on density functional theory (DFT) are too expensive to satisfy these requirements so it is therefore desirable to develop alternative, faster methods. Self consistent tight binding (TB) methods, which are derived from a second-order expansion of the Kohn-Sham energy functional and use a parametrised Hamiltonian matrix, can be significantly faster than DFT making them ideal for the above task. In this thesis we describe the development of a TB model for organic molecules containing carbon, hydrogen and oxygen. This model has been fitted using a combination of intuition, analytical fitting and genetic optimisation. We have also developed a framework for the description of inter-molecular Van der Waals interactions using long-ranged empirical potentials. We have shown that this model is capable of describing the gas-phase structural and dynamic properties of a range of organic molecules, as well as the structure and dynamics of liquid-phase alcohols and solvated organic molecules. We have also shown how the the model can be used to describe chemical reactions in solution. The model forms part of a larger collection which also includes models for water and titania. All models within this collection share a single transferable oxygen species. When combined with a parametrisation of a metal such as platinum this collection could be used for the study of catalytic processes.