Dear Vignesh,

I'll supose that you are trying to produce an electric junction between Mg and ZnO (both bulk materials). In addition I'll supose that ZnO is intrinsic or slighly n-type. With these assumptions in mind, the workfunction from Mg is 3.7 eV while the electron affinity from ZnO is 4.6. Thus, the conduction band of Mg is above. This also means that when put together, free electrons from Mg will flow into the semiconductor conduction band (as you said) until reaching equilibrium. Take into account that in this configuration you are probably producing an ohmic junction, but things may somehow change if you are working at the nanoscale!

Dear Dario Gallach 

Thank you very much for your valuable explanation sir

From our Experiments with MOMBE and ALD ZnO layers on various substrates we normally get an ionization energy of 7.3eV, which will lead to affinities of ca. 3.8eV. making a junction to Mg will probably alter the interface by forming O vacancies in ZnO and an oxidised Mg intrfacelayer. In my oppinion the application of the Schottky rule or elektron affinity rule will lead to wrong results. I think a strong upward Bendig of the ZnO bands in a Shottky junction will be inhibited by an interface reaction, but have to be investigativ experimentally.