No abstract

Techniques to predict the reliability of microdevices are necessary to facilitate the transfer of MEMS designs from the laboratory to the marketplace. An important reliability concern for microfabricated structures is in-use stiction, the operational failure of devices due to surface adhesion. The current study determines the temperature dependence of in-use stiction for polycrystalline silicon microcantilevers subjected to three different release conditions: supercritical CO2 drying; laser-irradiation repair; and self-assembled monolayer post processing. The microcantilever beam arrays were electrostatically actuated at temperatures between 22°C and 300°C. The supercritical CO2 dried devices showed an overall decrease in sticking probability as the actuation temperature was raised to 300°C. After a distinct improvement in the failure rate between the first and second actuation temperatures, arrays released using laser-irradiation did not exhibit a consistent trend. Samples coated with an OTS monolayer had large increases in their sticking probability as the temperature was raised. However, at temperatures above 200°C, a decrease in in-use stiction was observed which continued through most of the cooling cycle.

A theoretical approach for estimating solutions to Maxwell’s equations for structures with spatially-varying electromagnetic properties is presented for conductive media containing surfaces modified with functionally graded, heterogeneous electrical conductivity. The basis of the approach is an equivalent depth technique that replaces a graded conductivity region consisting of a phase mixture with a series of thin layers with uniform, multi-phase properties locally matching the effective mixture properties of the graded region. Radio frequency field propagation within each layer is determined as if it had existed within a constant conductivity medium but its depth is electromagnetically equivalent to the replaced graded region existing prior to the layer. The equivalent depth approach was applied to planar, thin foil, and cylindrical media to enable comparison with experimental results. Model predictions were compared with total transmission results for Pt-doped titanium thin foils and steady-state temperature rise in closed wire loops made from Sn-modified copper wire. For the thin foil case, the model-predicted total transmissivity shows good agreement with trends in the experimental results due to property changes in the modified surface layers. In the cylindrical wire case, similar agreement between the predicted effective conductivity values for the modified layers and experimental results was observed. Thus, the equivalent depth approach is an effective method for estimating solutions to Maxwell’s equations in complex media and a useful tool for predicting the performance of tailored surface conductivity modifications.

A measurement system is developed utilizing electromagnetic compatibility test equipment for the study of induced current in conductive materials subjected to radio frequency (RF) magnetic field strengths similar to the 1.5 T magnetic resonance imaging (MRI) B1 magnetic field at ∼65 MHz. The intent of developing such a system was to produce μT range RF magnetic fields in the laboratory to facilitate characterization of induction in conductive materials with modified surface electromagnetic properties to address unintended eddy current issues like Joule heating caused by implanted devices during MRI. A Helmholtz coil (HHC) is used as the RF magnetic field source, and the radiated field is monitored using a receiving loop antenna positioned coaxially outside the HHC. The measurement system operates in continuous wave and pulsed wave modes. Analytical models of the system were derived, which calculate the spatial distribution of RF magnetic flux and the induced current within a coaxially located sample in the transmission path between the HHC and receiving (R/C) loop from output voltage measurements at a single coaxial position. Induced currents were evaluated at multiple flux densities and at different frequencies, showing direct proportionality over the flux densities tested. Induced current results recorded in samples of different sizes and electrical conductivities (ranging from 0.1 to 5.8 × 107 (Ω m)−1 produced changes, matching trends predicted by conductive, closed-loop antenna theory. Induced currents were also used with simultaneous temperature rise measurements to characterize the effective surface conductivity for wire with non-uniform properties at 65 MHz.

No abstract