invited paper for International Conference on Fracture, to be held Dec. 3-7, 2001, Honolulu, Hawaii, Hilton Hotel questions: (1) To what degree is MEMS adhesion controlled by areas of real contact versus by van der Waals forces across non-contacting portions of the surfaces? (2) What is the minimum achievable value of adhesion for contacting bodies with rough surfaces? (3) What is the optimum value of roughness in MEMS? SAMPLE PREPARATION Cantilevers were fabricated according to a three mask level process, schematically represented in Fig. 1. Nanotexturing of the lower layer of polysilicon (poly 0 in Fig. 1(a)) was accomplished by thermal oxidation in dry O2 at 900 °C for increasing times. Table I indicates the times and the rms roughness as measured by AFM. Fig. 2 shows SEM images of the surface textures achieved by this method after the release etch. Two features are noted. First, the main texturing effect is due to grains that protrude upwards from the surface. This occurs because the polysilicon grains are randomly oriented, and dry oxidation in the linear regime proceeds at different rates on different orientations of silicon [11]. Second, the grain boundaries are decorated at increasing oxidation times, giving rise to grooves. These do not contribute significantly to the desired texturing because they do not take up a large percentage of the surface area, and reach below the surface. Standard deposition, lithography and etch techniques were used to fabricate the cantilevers (Fig. 1(b)). They are supported on the left in Fig. 1 by a step-up support post, formed by filling a hole etched into the sacrificial oxide layer. A critical step is the release and coating of the cantilevers (Fig. 1(c)). We used a solvent-based coating procedure that applies a self-assembled monolayer coating of perfluorodecyltrichlorosilane (FDTS, (C8F17C2H4SiCl3), similar to ref. [8]. Critical cantilever dimensions, as indicated in Fig. 1, include gap height g, thickness t, width w, length L, and actuation pad length a. After actuation (Fig. 1(d)), the length of the unattached region is used to denote the crack length s. Using profilometry, g=1.90 mm and thickness t=2.62 mm were determined from freestanding cantilevers. Mask dimensions were w=20 mm, a=81.5 mm and L ranged from 100 to 1635 mm. Fig. 2 Polysilicon ground plane surface textures versus oxidations times (SEM, 70° tilt). (b) 100 Å oxidation (a) No oxidation (c) 300 Å oxidation (d) 600 Å oxidation 1 mm Table I Polysilicon roughness versus oxidation time Oxdn. Time (min) Target tox (nm) rms roughness (nm) 0 -- 2.8 20 10 4.5 136 30 7.8 400 60 12.1 SiN on Si substrate poly 0 thermal oxide (a) (b) (c) (d) SiN on Si substrate poly 0 sacrificial oxide cantilever SiN on Si substrate after release and PFTS coating s d Testing (electrostatic actuation, inteferometry) Fig. 1 (a) ground plane polysilicon (poly 0) oxidation (b) cantilever fabrication, (c) release and coating (d) testing t g a L
RkJQdWJsaXNoZXIy MjM0NDE=