High-Throughput Optimization of Adhesion in Multilayers by Superlayer Gradient

Abstract

We used thickness gradients for high throughput optimization of adhesion in film stacks. The idea is based on the so-called superlayer test where a top layer under high compression exerts a load onto the lower interfaces and may cause delamination and buckling. Thus, on one hand, the thickness gradient of the superlayer results in the gradient of the load. On the other hand, the adhesion gradient can be realized by changing the thickness of an adhesion enhancer (or an adhesion reducer). When applied in two perpendicular directions (cross-gradient), the gradient of the superlayer in one direction and of the adhesion enhancer in the other, the plane of the sample represents a map where the line of delamination relates the interfacial toughness to the thickness of the enhancer. In our tests we used Mo superlayers under compressive stress of the order of ˜1.5 GPa on a Si wafer with a native oxide. The adhesion reduction was observed with this methodology when Ag layer up to 10 nm thick was deposited onto the substrate prior to Mo deposition. The delamination occurred at Ag thicknesses starting from ˜6 nm. This thickness of Ag corresponds to the islands coalescence and formation of a continuous film which immediately results in adhesion reduction. The other test was performed with a step gradient of Ti enhancer placed under a 10 nm thick Ag layer in otherwise the same arrangement. A single test showed that 2.8 Å of Ti was sufficient to improve the adhesion between Ag and SiO_x by several times.