Enhanced adhesion through local epitaxy of transition-metal nitride coatings on ferritic steel promoted by metal ion etching in a combined cathodic arc/unbalanced magnetron deposition system

In situ substrate cleaning by ion etching prior to deposition in physical v apor deposition processes is a key step in achieving good film adhesion, wh ich is essential for all coating applications. Irradiation with metal or ga s ions alters substrate surface chemistry, topography, and microstructure t hus affecting subsequent film growth. This study compares Ti1-xAlxN/ferriti c steel (x=0.54) interfaces formed after Cr ion bombardment at negative sub strate biases, U-s, ranging from 600 to 1200 V during a Cr cathodic are dis charge, stabilized with a 0.06 Pa Ar background pressure. Samples biased wi th -1200 V in an Ar glow discharge at a pressure of 0.6 Pa were also invest igated. Microstructure and microchemistry of the interfaces was studied by scanning transmission electron microscopy with energy dispersive x-ray anal ysis using cross-sectional samples. Cr ion etching with U-s= 1200 V resulte d in a net removal of over 100 nm of substrate material with the formation, through implantation, of a Cr-enriched near-surface region extending to a depth of similar to 10 nm. As U-s was reduced to 600 V, Cr accumulated at t he surface as a greater than or equal to 5 nm thick layer. Ar was incorpora ted at the surface to levels of 4-6 at. % during both Cr are and Ar glow di scharge etching. The microstructure of Ti1-xAlxN overlayers was dramaticall y affected by pretreatment procedures. Ar sputter cleaned steel surfaces (U -s=1200V) promote nucleation of randomly oriented grains leading to a compe titive column growth with small column size and open boundaries. In contras t, Cr irradiation at the same bias voltage results in local epitaxial growt h of Ti1-xAlxN on steel and lead to a superior performance in scratch testi ng compared to coatings deposited after Cr treatment with U-s = 600 V or Ar ion bombardment at U-s= 1200 V. Critical loads were 63, 47, and 27 N, resp ectively. (C) 2000 American Vacuum Society. [S0734-2101(00)01804-2].