In order to build a model, one needs to know what your goal is, what you will do with that model. Only you know those things, and you have not communicated them to the world. Perhaps the solution is a differential equation system. Perhaps it is a simple formula. (I'm not a chemist, so I can't go too far here.)
Often your goals for the model drive the solution. Is this something you need to model as a function of time? If so, then often differential equation models are good choices. But you will then need to write down the equations. Transform the graphical form you wrote into mathematics. Then the MATLAB code will naturally be easier.
