• PDF / 2,552,184 Bytes
• 6 Pages / 417.6 x 639 pts Page_size
• 6 Downloads / 277 Views

DOWNLOAD

REPORT

---

A direct testing technique was developed to

characterize the strength and distribution in properties of polysilicon, it was applied to measure the strength and modulus of samples of amorphous diamond having dimensions typical of critical features of MEMS devices. This technique will be described as well as the results for the amorphous diamond MEMS. The results will be compared to indentation results on material from the same die. EXPERIMENT Material Description Recent work [1, 2] has shown that thermally stable, stress-free, smooth, hard, and stiff amorphous-diamond films can be grown using pulsed laser deposition at room temperature followed by a brief anneal at 600*C. Layers 1 ýtm thick were built up by depositing 4 layers of amorphous diamond (-0.25 gtm thick) by repeated deposition and annealing in situ. The films were then post-annealed in a vacuum furnace (550 °C for -60 hrs) until the residual stress (as measured by wafer curvature) was reduced to near zero ( < 10 MPa). The tensile properties of these films are expected to approach those of natural diamond, but they have not been directly measured. Amorphous diamond MEMS structures were patterned using contact lithography on silicon wafers that were prepared with a 2 gtm blanket layer of Si0 2, 60 nm of polysilicon for adhesion and 1 gim of amorphous diamond. 465

Mat. Res. Soc. Symp. Proc. Vol. 593 © 2000 Materials Research Society

Direct Tensile Testing Direct tensile measurements on structural thin films are challenging. A technique to perform automated testing has been developed that utilizes the lateral force capabilities of a nanoindenter to measure the small forces and displacements. Typical samples produced from polysilicon incorporate a freely rotating hub on the fixed end of the sample to avoid bending stresses on the sample [3, 4]. Two designs were attempted in amorphous diamond, a single layer sample utilizing a timed release of the underlying oxide, and a two layer structure with a patterned oxide between the two amorphous diamond layers. Difficulty in fabrication limited the samples to the first, simpler, style. Double ended "dog bone" shaped samples were used, where one end is attached to the silicon die via a layer of Si0 2 . The mask was designed so that the fixed ends would not release during a timed etch of sufficient duration to release the gage section and the ring used to engage the nanoindenter. Figure 1 shows two views of tested samples. The partially etched Si0 2 is visible through the amorphous diamond layer under the fixed ends of the samples.

Figure 1. Top and perspective views of tested amorphous diamond samples. The sample widths and lengths were measured in an SEM and compared to a calibration standard. The thickness was measured using a profilometer. For this group of samples. w = 2.25 ýtm and t = 1.02 .tm. The raw load-displacement data is processed in two steps to evaluate the strength and modulus of the material. First, a force balance is solved to correct for the losses due to the frictional sliding of the tip along the