On the possible role of the interfacial inversion layer in the improvement of the performance of hydrogenated amorphous silicon/crystalline silicon heterojunction solar cells (HIT)