Etching of silicon with molecular fluorine for the micromachining of micro electro mechanical systems (MEMS) has been evaluated. The etching process is carried out in a continuous flow etching system that uses a 25 vol% mixture of F-2 in N-2 and operates at room temperature and atmospheric pressure. Fluorine etches silicon isotropically at a rate of 0.2 mu m min(-1) and is a viable etchant for bulk silicon micromachining. The F-2 etch results in the formation of pits about 10-50 mu m in size within the silicon features and roughness in the micrometer and sub-micrometer length scales. SiO2, Pt, Ni, Al and Ta do not etch or roughen after several hours of F-2 exposure. In addition, F-2 does not etch low-stress silicon nitride; a solid layer forms on the silicon nitride surface upon exposure to F-2 that protects the underlying silicon nitride from further attack. In contrast, a commercial XeF2 etching system etches low-stress silicon nitride with a silicon nitride: silicon selectivity of approximately 1: 200. The F-2 etching system is used to release tubes of a MEMS fuel processor that have 2 mu m thick low-stress silicon nitride walls.

This study was designed to determine and compare the dose-response characteristics, speed of onset, and relative potency of single-dose epidural fentanyl (F) and sufentanil (S) for postoperative pain relief. Eighty women undergoing cesarean section (C/S) with epidural 2% lidocaine with epinephrine (1:200,000) were randomly assigned to receive double-blind epidural administration of F (25, 50, 100, or 200 micro g) or S (5, 10, 20, or 30 micro g) (n = 10 per group) upon complaint of pain postoperatively. Visual analog scales (VAS, 0-100 mm) were used to assess pain and sedation at baseline; at 3, 6, 9, 12, 15, 20, 25, 30, 45, and 60 min; and every 30 min until further analgesia was requested. The study was terminated at 30 min if satisfactory analgesia was not achieved. Side effects were recorded. A dose-response was demonstrated for both opioids. F 25 micro g and S 5 micro g were ineffective, with significantly fewer women achieving VAS scores <10 mm (P < 0.05 compared with F 100 or 200 micro g and S 20 or 30 micro g). F 100 and 200 micro g and S 20 and 30 micro g all achieved VAS scores <10 mm in all women with no differences in time to 50% reduction in VAS (mean 11-16 min) and no differences in duration of analgesia (mean 117-138 min). The 50% and 95% effective dose values for each opioid to achieve a VAS score <10 mm were F 33 micro g and 92 micro g and S 6.7 micro g and 17.5 micro g. There were no differences among groups in sedation scores or side effects. Our data suggest that the relative analgesic potency of epidural S:F is approximately 5 and that there are no differences between the opioids in the onset, duration, and effectiveness of analgesia when equianalgesic doses are administered postoperatively after lidocaine anesthesia for C/S. (Anesth Analg 1997;85:365-71)