I agree with point 2 made by Mrs. Chamoire above. Any way of heating the material should help you. Higher temperatures or ways of removing the evaporated Sb (heating in vacuum, providing condensation sites) will enhance the effect. If you are talking about removal of Sb from the bulk material and not only the surface (a few 10 nm) temperature will be key because you need to enhance the diffusion within your material. I have 'seen' Sb evaporate from an amorphous material at 600 °C, while there was no effect in the same material when it was polycrystalline (this was done in vacuum). If you have the means, I would suggest to start by heating your material to temperatures between 600 °C and 800 °C in vacuum and see what happens.
If the temperature is too high, you might actually crack the bond between Co and Sb and end up with a material that is light on Sb though.