Dear Marcos Thanks a lot. But there is another problem that for the surface, I compared the pure surface'PDOS with surface-O2, it seems the shifting the Fermi level is not enough. Because the surface is semiconductor, siesta define the fermi level as surface's middle of HOMO and LUMO, but after adsorption, the fermi level become the HOMO of the surface. So in this case, we should not shift the PDOS by fermi level, right?
Regards, Gao Min