Francesco Rizzini

Sidewall Adhesion Evolution in Epitaxial Polysilicon as a Function of Impact Kinetic Energy and Stopper Area

This paper investigates the evolution of adhesion forces during dynamic contacts, which simulate realistic shocks. The developed method accounts for the differences in offset, resonance, and quality factor among the devices under test to precisely predict (and monitor) the kinetic impact energy during shocks. The results clearly show that adhesion force grows and stabilizes at an increasing amplitude with the applied impact energy. This dynamic adhesion is also dependent on the contact area, while the native adhesion on the same specimens was independent of it. Finally, a relaxation of the adhesion force with a very long time constant is observed after the end of the tests.

A study of adhesion forces in thick epitaxial polysilicon under dynamic impact loading

The work presents a structure and a method for the in-line characterization of impacts and adhesion phenomena between MEMS moving and fixed parts: the focus is on the monitoring of an inertial proof mass motion when colliding with a mechanical stopper. Through such measurements, one can evaluate the energy balance during impact events. The work analyzes the adhesion force evolution after a number of impact cycles comparable or larger than shocks in a 5-year operation. Results obtained on two different specimens show growing and then stabilizing adhesion forces of on average 170 nN, under impact cycles with about 500 fJ energy loss. No marked dependence on the specimen area is obtained. The possibility to change and track the impact kinetic energy is also demonstrated.

Experimental study of out-of-plane adhesion force evolution (and regression) for MEMS accelerometers

The work investigates în-operation adhesion evolution in out-of-plane (OOP) test devices fabricated in an industrial surface micromachining process. The aim is to mimic the behavior of OOP MEMS accelerometers in terms of package environment (pressure of ~ 100 mbar), characteristic frequency and quality factor, so that the contact dynamics during impact events is representative of such devices. The experiments include: (i) statistical measurements of pristine adhesion forces; (ii) application of dynamic impacts at known contact velocity (~ 1 cm/s) and adhesion evolution monitoring as a function of the elapsed collisions; (iii) adhesion force monitoring within 40 days after the former test ends. With a sub-nN measurement resolution, the results show an adhesion force increase during cycles up to a few hundred nN, followed by a regression (likely due to a slow relaxation mechanism) when the device is left still.