A chemisorbed fluorocarbon self-assembled monolayer (FSAM) layer on MEMS surfaces can greatly
improve MEMS device reliability by reducing the in-use stiction force. In this work, a MEMS tribogauge that
measures stiction force between two interacting surfaces has been used for FSAM coating and process parameter
characterization. A commercially available nanocoating tool was used for depositing FSAM coatings. Five different
coating recipes with different injection times for the FSAM precursor were evaluated with the tribogauge. The
interacting surfaces of the tribogauge were exposed to many load cycles. These experiments allowed a preferred
process to be determined.
A MEMS tribogauge was used for on-chip and in-situ characterization of nano-tribological phenomena
(stiction, friction, and wear of coated polysilicon surfaces). The device was fabricated using the SUMMiT-V
process. Measurements were made on sidewall surfaces on the polysilicon-3 layer. The device consists of two
orthogonally positioned comb-drive assemblies that are used for both actuation and sensing. One assembly is used to
apply a normal load (F<sub>n</sub>) to contacting surface, while the other induces a tangential load (F<sub>T</sub>). Precise position
control is tracked by employing a LabVIEW controlled AD7747 capacitance sense mechanism. The resolution of
the characterization apparatus is ±10nm.
Three MEMS tribogauge devices are tested; two of them have a chemisorbed layer of self-assembled
monolayer (SAM) coatings and one with no SAM coating. The two types of SAM coatings are FOTS and 'Sandia
vapor-SAM' (SVSAM). The tribogauge with no FSAM coating is either UV-Ozone or 'air plasma' treated to
remove organic contaminants leaving behind -OH bonds on top of the MEMS surface (native oxide, SiO<sub>2</sub>).
Characterization using the tribogauge for each coating type includes: measurement of baseline stiction force [see manuscript], static and dynamic coefficient of friction [see manuscript], induced stiction force
calculated after specific load cycles [see manuscript]. Experiments showed that the induced stiction force
increases in proportion to the increase in the number of load cycles, indicating degradation of the FSAM coating and
topographical changes to the interacting surfaces. The UV-Ozone /air plasma treated pristine tribogauge was used to
measure the stiction force of the device with no SAM coating [see manuscript].
This paper presents the results of an experimental study of electrothermal poly-Si MEMS structures wherein
changes to the surface topography and material properties are observed due to use. The ex-situ AFM
characterization reveals changes in the surface topography after cyclic actuation. The extent of topical SiO2
appears to increase with the number of actuation cycles and increasing stress levels on the polysilicon surfaces.
The differences in the surface topography and oxide thickness are characterized as a function of fatigue cycling
and in-situ annealing of the electrothermal actuators. FEA analyses were performed to evaluate the magnitude
and distribution of stresses in the actuators to compare stress effects from oxide development on
electrothermomechanical structures. With the observation of topographical changes, the intrinsic material
property like resistivity was also affected. A change of 1.4% was seen for a 20% duty cycle, 3.1% for 50% duty
cycle and 4.1% for 80% duty cycle. Similar experiments were performed for sealed devices in order to observe
the changes in resistivity under inert conditions. A comparison of change in resistivity for sealed devices and nonsealed
devices was done. Finally, force-distance curves were plotted to ascertain the adhesion forces for the
actuator surfaces before and after actuation. The adhesion forces increases from ~7nN (un-actuated chevron) to
~40nN (10,000 cycles).