Technology diversity.(EDITORIAL)

Scientific competition among companies has produced numerous solutions to the common problems of the semiconductor industry. That diversity of technical approaches has been built into the very DNA of the industry, resulting in the flexibility needed to confront the unexpected. Other industries--think transportation--suffer from narrow technological bases and experience widespread hardship when unexpected difficulties (e.g., high oil prices) impact all companies and segments simultaneously. However, having a wide range of solutions to each important problem has prevented anyone from claiming monopoly profits. That inability to reap outsized rewards has "de-motivated" managements and tightening R&D budgets. The range of options is narrowing as the industry matures--and that could be bad.

Just imagine what might have happened a few years ago when 193nm dry lithography was running out of steam: Both EUV and 153nm had been touted as follow-on paradigms, but neither was ready when needed. Had it "stayed the course," the industry would be in trouble now! Instead, two companies competed to quickly develop a maverick idea to use water immersion with 193nm exposure. The tools that emerged were quite different in detail, each with virtues and flaws. One had twin wafer stages, with metrology done on dry wafers and water flow being turned on and off . The immersion hood incorporated an air curtain to confine the liquid. The other used a single wafer stage and continuous flow and confined the liquid using surface tension alone. The projection lenses and illumination systems were also quite different.

What happened? It all worked, but not perfectly. Engineers had to deal with the unexpected (bubbles, particles, stages that bent under the force of the water, etc.). The problems were sorted out, rather than accepted, because of the dynamics of the competition. The resist situation was even more dramatic: Chemists devised photo-resist formulations that resisted leaching by water and topcoats that could protect dry-type resists, but then had to be removed (with solvent or developer). Additives appeared that self-segregated, forming an in-situ topcoat. Surface energies were all over the map, and some switched, depending on pH or exposure. Some things performed better than others; some cost ...