Why can't we use a divalent or ahexavalent impurity while doping a semiconductor so that
Its conductivity increases??

Divalent and hexavalent impurities are used to dope group III or group V semiconductors, for example, Selenium (group VI) doped Gallium Indium Phosphide, or Beryllium (group II) doped Aluminium Arsenide or Gallium Arsenide.

I assume you're talking about using group II or VI dopants in group IV semiconductors. I would guess that using these would cause some kind of instability or mismatch in the crystal lattice - as one of the bonds wouldn't form. This probably gives rise to unwanted effects. However I can't find much discussion of this in the literature.