An improperly structured silica nanoparticle is no better than a structureless sphere with defined interactions, the latter requiring much less effort. My opinion of course, point is, if you aren’t going to consider the surface chemistry, there’s no point in using silica in your model.