On the doping problem of CdTe films: The bismuth case