İf you apply uni- or bi-axial tension during the coating you may have the chance to obtain smooth surface. But in the case of compressive stresses just the opposite could occur. Otherwise one may not eliminate the surface roughness caused by the spontaneous relief of the stored elastic strain energy density created by the thermal contraction or expansions taking place during the coating process..
10 layers were deposited on the substrate...for 10 bilayer Sa=0.719 nm and for control it was 11.101 nm and if this is possible can you please explain the reason for this?