In this thesis，we develop a new approach to calculate the solvent mediated interactionpotential of mean force (PMF) in binary mixtures of hard-core fluids.This approach is based on classical density functional theory in combination with solvation free energy concept.Comparing to integral equation method，simulation or other theoretical and experimental PMF determination techniques，this approach has advantages in calculation speed and computational expenses.To demonstrate，three types of systems are considered.　　(1) Two big spheres interact in solvent;　　(2) A big sphere interacts with a planar wall in solvent;　　(3) A big sphere interacts with a planar wall with a cavity in solvent.　　For all systems，the theoretical calculations are calibrated with the comparison to available simulation results.Afterwards，the solute-solvent size ratio effect is examined in addition to the solvent packing fraction effect.Particularly，both symmetric and asymmetric spheres are considered in the first system，and well-matched and ill-matched cases are investigated in the third system.We show that when the packing fraction is higher，the oscillation feature of PMF curve in term of separation is more significant，bringing larger attraction in the circumstance of close contact and higher energy barrier for two solutes approaching each other.When the solute-solvent size ratio is larger，the PMF is more profound; the contact energy in the case of perfect match between the sphere and the cavity is overall larger than that in the other case.This work provides theoretical guidance to the experimental processes associated with adsorption，surface modification and separation of biomolecules.