Deformation and fracture of single-crystal silicon theta-like specimens

Abstract

Single-crystal silicon test specimens, fabricated by lithography and deep reactive ion etching (DRIE), were used to measure microscale deformation and fracture properties. The mechanical properties of two specimen geometries, both in the form of a Greek letter H (theta), were measured using an instrumented indentation system. The DRIE process generated two different surface structures leading to two strength distributions that were specimen geometry independent: One distribution, centered about 2.1 GPa, was controlled by 35 nm surface roughness of scallops; the second distribution, centered about 1.4 GPa, was controlled by larger, 150 nm, pitting defects. Finite element analyses (FEA) converted measured loads into strengths; tensile elastic measurements validated the FEA. Fractographic observations verified failure locations. The theta specimen and testing protocols are shown to be extremely effective at testing statistically relevant (hundreds) numbers of samples to establish processing-structure-property relationships at ultrasmall scales and for determining design parameters for components of microelectromechanical systems.

Fracture strength of micro- and nano-scale silicon components

“…A simple proof test at 1 GPa would have been enough to eliminate such an irregularity. Similar remarks can be made regarding the "unintended" etch that occurred during the fabrication of the theta specimens by Gaither et al 153 Boyce et al, however, note several precautions that should be considered if proof testing is to be applied to MEMS components: 261 (i) proof testing loading conditions should match in-service loading conditions, (ii) proof testing environments should be at least as aggressive as the worst case in-service environments, and (iii) proof testing may not be feasible for all component designs. Nevertheless, although adding another manufacturing step, proof testing provides MEMS manufacturers with a method to bypass processing-structure dependencies and directly engineer strength properties for optimized device performance.…”

“…In addition, the complex inner geometry of the original Durelli design was replaced with a simple arch, thereby reducing the size and extent of secondary stresses in the specimens on loading. Gaither et al used the original Durelli and new arch theta test specimens to examine both unintended 153 and intended 86 etching process effects on the surface structure, fracture strength, and reliability of SCS. In these studies, the load-displacement response at the loadpoint was translated into stress-strain behavior across the web region using FEA and r f for each specimen was calculated from the peak load at sample failure.…”

“…On the other hand, compressive test methods with an instrumented indenter or AFM are sometimes easier to conduct, in that they avoid many of the gripping, mounting, loading, and post-test characterization issues associated with tensile test methods. 153 Consequently, a number of researchers have recently developed test methods based on compressive loading schemes, especially for investigating the mechanical properties of nanoscale SCS components.…”

“…For example, Gaither et al used the original Durelli and new arch theta test specimens to examine both unintended 153 and intended 86 DRIE process effects on the surface structure, fracture strength, and reliability of SCS. Both specimen designs used tangential circular sections to incorporate the web and had the same diameter of 250 lm and web width of 8 lm, and as a result, the same stressed area of %6.25 Â 10 3 lm 2 .…”

Accurate spring constant calibration for very stiff atomic force microscopy cantilevers

“…The rapid development of MEMS and nanoelectromechanical systems (NEMS) necessitates the ability to predict and control the strength of micro-and nano-scale structures. In a recent survey of the strength of single crystal Si 26 (a material widely used in MEMS and NEMS), the strong dependence of Si strength on both component fabrication method and size was highlighted. The survey made clear that surface flaws introduced by the fabrication process were the dominant factor in controlling the strength of Si components and that flaws were larger (and hence strengths smaller) in larger components.…”