G-and i-line diazonaphthoquinone/novolak photoresist films are surface imaged with g-line, i-line and deep-UV steppers. Following optical exposure, the resist film is treated with aqueous solutions which deposit a catalyst for electroless metal deposition. Wet development of the exposed and catalyzed photoresist results in selective removal of catalyst along with the exposed portion of the underlying photoresist. Upon immersion in an aqueous electroless plating solution, metal is selectively deposited on the unexposed photoresist which is still bearing catalyst to yield a positive-tone plasma etch mask. Oxygen magnetron-enhanced reactive ion etching (02 MERlE) provides high polymer etch rates (4 im/min) with excellent selectivity (>300:1) to 70-1 70 A Ni films. In addition, large ion fluxes produce highly anisotropic etch profiles for faithful pattern transfer. The process has achieved 0.30 pm resolution with a 6:1 aspect ratio at 248 nm (0.35 NA). Printing of 0.40 im lines and spaces has been achieved at i-line (0.45 NA) over Al steps.

No abstract

The universality of a single percolation field scaling law to the tetramethylammonium hydroxide dissolution of derivatized novolac and poly(4-vinylphenol) polymers was investigated. According to this hypothesis, the dissolution ofphenolic polymers occurs through a percolative mechanism resulting from the diffusion of base along nascent channels formed by the spatial proximity of phenolic hydroxyl groups (presumably diffusion of the cation is rate limiting). Dissolution inhibition results from the the removal of sites from the percolation field. In this study, the polymers were derivatized with increasing amounts of either 2,1,5-diazonaphthoquinone groups or methylsuiphonyl ester groups and the dissolution rates of the films were measured. While our experimental data supported adherence to a percolation law (p =the scaled amount of free hydroxyl sites remaining on the polymer), Rate = Rate0 (ppc)t , we did not find that a single exponent of t = 2 universally described the dissolution behavior. Rather, our data indicated that t varied with different systems, with values of t greater than 5 being observed. These results are explained in terms of multiple simultaneously operant mechanisms of dissolution creating an environment where multiple percolation can occur. The relative shielding effects of the blocking groups are also compared.A computer simulation of Reiser's percolation model was constructed to gain a better understanding of the dissolution kinetics. We found that the function which governs the probability of transfer from a phenolate site to a hydroxyl site needed to be relaxed in order to allow a greater likelihood of transfer, particularly when the site separation distance exceeded 5 A. The simulated values of xc decreased as either the difficulty of intersite transfer was increased, or as the shielding effect per blocking group was increased.