Preparation of a Well-Defined Amino-Terminated Self-Assembled Monolayer and Copper Microlines on a Polyimide Substrate Covered with an Oxide Nanoskin

Abstract

A well-ordered, uniform amino (NH(2))-terminated organosilane self-assembled monolayer (SAM) was prepared on a polyimide (PI) substrate, the surface of which had silica-like reactivity. First, through chemical vapor deposition of 1,3,5,7-tetramethylcyclotetrasiloxane and subsequent photooxidation using 172 nm vacuum ultraviolet light, an extremely thin silicon dioxide (SiO(2)) layer about 1 nm thick, which we call an "oxide nanoskin" (ONS), was prepared on a PI substrate. Due to the presence of this ONS layer, the PI surface's properties became almost identical with those of Si covered with native oxide (SiO(2)/Si) without any marked change in surface morphology, as evidenced by zeta-potential measurements, X-ray photoelectron spectroscopy (XPS), and atomic force microscopy (AFM). Next, this ONS-covered PI (ONS/PI) surface was exposed to vapor of a 12.5 vol % solution of N-(6-aminohexyl)(3-aminopropyl)trimethoxysilane (AHAPS) molecules diluted with absolute toluene. On the basis of contact angle analysis, the surface energy of this AHAPS/ONS/PI sample was mostly consistent with that of a SiO(2)/Si substrate covered with an AHAPS-SAM (AHAPS/SiO(2)/Si). On the other hand, the surface energy of an AHAPS-treated PI (AHAPS/PI) substrate was much smaller than that of the AHAPS/ONS/PI substrate due to insufficient surface coverage by the AHAPS molecules. This was also confirmed by lateral force microscopy using photolithographically micropatterned samples. Fabricated micropatterns composed of AHAPS- and SiO(2)-covered regions were clearly imaged on the AHAPS/ONS/PI substrate through their difference in friction, while the friction contrast of the micropatterned AHAPS/PI substrate was unclear. This marked difference in packing density of the AHAPS molecules had a direct influence on the adsorption behavior of palladium colloids and subsequent electroless plating of copper (Cu). As confirmed by AFM and XPS, metallization proceeded only on the AHAPS-covered regions, while the SiO(2)-covered regions remained free of deposits, resulting in the formation of 10-mum-wide Cu microlines on both samples. However, the plating rate achieved on the AHAPS/ONS/PI substrate was about 4.5 times faster than that on the AHAPS/PI substrate and the pattern resolution was considerably fine.