I think water by itself can randomly break a few loosely bound atoms off the surface of anything. But the rate of that dissolution is dwarfed by the wear effects of the sediment being transported by the water, which is the primary material remover.