Impact of in situ oxygen plasma cleaning on the resistance of Ru and Au-Ru based rf microelectromechanical system contacts in vacuum

Abstract

Contact resistance measurements are reported for radio frequency microelectromechanical system switches operating in an ultrahigh vacuum system equipped with in situ oxygen plasma cleaning capabilities. Ru-based contacts were prepared by means of standard sputtering techniques, sputtering followed by postdeposition oxidation, ͑surface RuO 2 ͒ or reactive sputtering in the presence of oxygen ͑bulk RuO 2 ͒. In situ oxygen plasma cleaning lowered the resistance of Ru contacts by two or more orders of magnitude but not lower than Au contacts, irrespective of whether the Au contacts were cleaned. The time dependence of the resistance was fit to power law extrapolations to infer contact creep properties and resistance values at t = ϱ. Time-dependent creep properties of mixed Au-Ru contacts were observed to be similar to those of Au-Au contacts, while the absolute value of the resistance of such contacts was more comparable to Ru-Ru contacts. Prior to, and for short oxygen plasma exposure times, bulk RuO 2 resistance values exhibited much larger variations than values measured for surface RuO 2 . For O 2 plasma exposure times exceeding about 5 min, the bulk and surface RuO 2 resistance values converged, at both t = 0 and t = ϱ, with the t = ϱ values falling within experimental error of theoretical values predicted for ideal surfaces. The data strongly support prior reports in the surface science literature of oxygen plasma induced thickening of oxide layers present on Ru surfaces. In addition, they demonstrate that vacuum alone is insufficient to remove contaminants from the contact surfaces and/or prevent such contaminants from reforming after oxygen plasma exposure.

Electrical Contact Resistance and Device Lifetime Measurements of Au-RuO2-Based RF MEMS Exposed to Hydrocarbons in Vacuum and Nitrogen Environments

“…The system is equipped with in situ oxygen plasma cleaning capabilities and has the ability to immediately expose the samples to partial pressures of dodecane gas, or alternatively examine samples lifetimes in either vacuum or nitrogen environments. In this stringently controlled chamber, we have previously reported the baseline contact resistance values after cleaning to be close to those theoretically predicted for a contact with no adsorbed contaminant film [8]. Au-RuO 2 contacts were selected for our studies, as they are a promising materials combination for RFMEMS devices [8,37], and in particular remain operational in a wide range of conditions so as to allow controlled studies to be performed on them [8].…”

“…Walker et al [8], for example, reported contaminant films to reform after several days in UHV, and almost immediately in highvacuum conditions. Such contaminant layers are generally invoked whenever experimentally measured resistances are found to be higher than theoretical values for a clean contact [4,8]. The situation is complicated, however, by the fact that actual contact occurs only at a limited number of small ''contact spots'' [24][25][26] that can exhibit asperity creep and grow larger in time [27][28][29][30].…”

“…The layers are extremely difficult to remove [8,23], and even after in situ plasma cleaning in Ultra High Vacuum (UHV), they have been observed to readily reform. Walker et al [8], for example, reported contaminant films to reform after several days in UHV, and almost immediately in highvacuum conditions. Such contaminant layers are generally invoked whenever experimentally measured resistances are found to be higher than theoretical values for a clean contact [4,8].…”

“…It is believed that residual hydrocarbon films at the surface lead to decomposition and then deposition of highly resistive carbon deposits [3,4,[20][21][22]. The layers are extremely difficult to remove [8,23], and even after in situ plasma cleaning in Ultra High Vacuum (UHV), they have been observed to readily reform. Walker et al [8], for example, reported contaminant films to reform after several days in UHV, and almost immediately in highvacuum conditions.…”

A micro-model for elasto-plastic adhesive–contact in micro-switches: Application to cyclic loading

“…Ru films are deposited on the movable electrode, and also on a part of the substrate, and are characterized by a thickness t s and a Young modulus E s . Typical value for the Ru film thickness is reported in Table 2 [52], and the Young modulus is the one reported in Table 1.…”