In the PVD of TiN thin films (~100A), I've found the resistivity of the TiN films dropped from ~100 to 60um*cm with N2 flow increasing up to the metallic mode ending point; then the resistivity jumped to ~200um*cm with N2 flowing in the poisoned mode (100sccm). How to explain this resistivity behavior vs N2 flow? What's the mechanism behind it?