AVS 50th International Symposium
    Plasma Science and Technology Monday Sessions
       Session PS-MoM

Invited Paper PS-MoM1
Not Quite 50 Years of Plasma Etching

Monday, November 3, 2003, 8:20 am, Room 315

Session: Critical Dimension Etching
Presenter: R.A. Gottscho, Lam Research Corp.
Correspondent: Click to Email

Since the advent of plasma etching in the manufacturing of semiconductor devices in the 1970s, both industry and technology have changed dramatically. In the ‘70s, integrated device manufacturers (IDMs) dominated the industry. IDMs built systems by starting with "sand" and using equipment of their own design and fabrication. For example, hexode etchers were created by Bell Labs, the Reinberg reactor came from Texas Instruments, and electrostatic chucks came from IBM. Today, IDMs buy plasma processing equipment with processes developed, at least partly, by the equipment companies. Fabless companies have sprouted and grown ubiquitous as foundries have become a dominant source of chip supply. In the early ‘80s, debates raged over the relative merits of single wafer processing for 4-6" wafers. Today, batch etchers cannot be found in the fab despite their inherent throughput advantages. Single wafer processing of 12" wafers with unprecedented control is the norm. Now, we see the advent of integrated metrology and even more advanced process control. Over this time period, gate lengths have shrunk from >1 um to <0.13 um. Concomitant with this inexorable reduction in gate length, is a comparably relentless reduction in defect density and within wafer, lot-to-lot, and machine-to-machine CD non-uniformity. Where will etching technology go in the future and what further changes can we expect in the industry? I will attempt to answer these questions using the past as a guide.