Billions of dollars have been spent controlling or removing dissolved oxygen (DO2) from ultrapure water (UPW) to concentrations on the order of single- or double-digit part-per-billion. Entire industries — ranging from the development of unit process operations to metrology - have developed in the pursuit of removing as much DO2 as possible. ITRS, ASTM, and SEMI have each published roadmaps that target the dissolved oxygen concentrations in UPW at the single-digit per billion level over the next ten years. It is conceivable that semiconductor manufacturers could spend millions of dollars in capital investments and operational expenses to control to those setpoints. But since almost all industry-standard manufacturing tools perform critical wet processing in an oxygenated, ambient atmosphere, is there any added value to increase the complexity and cost associated with such precise DO2 control? Or is the semiconductor industry just following the conventional wisdom that UPW should be as pure as possible? To answer this question, one needs to understand: (1) the impact of DO2 on the wafer, (2) where the DO2 concentration is being measured prior to the water touching the wafer; (3) how the manufacturing customer is using the UPW; and (4) Fick's Law of Diffusion. By understanding these four items, the UPW engineer may quickly find that controlling DO2 or driving it as low as possible may be more conventional wisdom than a true process need.
展开▼
