Materials Science and Engineering B 163 (2009) 199–203 Contents lists available at ScienceDirect Materials Science and Engineering B journal homepage: www.elsevier.com/locate/mseb Short communication Electrical and mechanical properties of layered gold–chromium thin films for ohmic contacts in RF-MEMS switches V. Mulloni ∗ , R. Bartali, S. Colpo, F. Giacomozzi, N. Laidani, B. Margesin Fondazione B. Kessler- Centro Materiali e Microsistemi, Via Sommarive 18, 38050 Trento, Italy article info Article history: Received 3 February 2009 Received in revised form 1 June 2009 Accepted 6 June 2009 Keywords: Gold Chromium Annealing Diffusion Oxidation abstract The electrical contacts are the most critical part for the reliability of RF-MEMS switches. For this reason the development of new contact materials with higher hardness and low resistivity would be of great help in increasing MEMS switch reliability. In this paper a detailed study of mechanical, electrical and morphological properties of gold–chromium multilayers is presented and discussed. It will be shown how hardness of gold can be increased introducing thin layers of chromium inside the gold contact layer. However, some care must be taken, since this improvement can vanish because of thermal and oxygen plasma treatments normally involved in RF-switches fabrication, mainly due to chromium oxidation. © 2009 Elsevier B.V. All rights reserved. 1. Introduction Microelectromechanical (MEMS) switches for radiofrequency (RF) applications are in the recent years widely studied for their promising features with respect to conventional switches as low insertion loss, high isolation, small size and low weight, wide band of exploitable frequencies and very low power consumption. How- ever, in spite of their evident advantages, some questions regarding RF-MEMS switches performances are still open, and among them the most relevant are lifetime, power handling, number of switch- ing cycles and other important issues concerning reliability [1,2]. In MEMS switches, most of these critical issues are closely related to the metal-to-metal contacts in the switching area. From the reliability point of view, the metal-to-metal contact is the most critical part of ohmic RF-switches, because it directly influences not only the contact resistance and insertion loss, but also the most rel- evant failure mechanisms for this type of switch [3] that are wear of ohmic contact, adhesion and stiction phenomena [3–6]. The ideal contact should be hard, flat and very conductive. From this point of view gold is a promising material, because of its high conductivity and chemical inertness, and it is in fact widely used as contact metal in MEMS applications. However, pure gold is a soft metal, and strategies to improve its hardness may include incorpo- ration of hard particles [7,8], some sort of gold alloying with metals like nickel or copper, or with other noble metals like silver, palla- dium platinum [9], rhodium and ruthenium [10,6,11]. Some of this ∗ Corresponding author. Tel.: +39 0461 314441. E-mail address: mulloni@fbk.eu (V. Mulloni). strategies are effective in increasing hardness and maintain at the same time most the desirable properties of gold, like alloying gold with some other noble metals in low percentage. For instance, in the case of gold–platinum alloy, the best result are obtained for of Au–(6.3%)Pt, where hardness can be improved going from 1.77 of pure gold to 2.19 of Au–Pt alloy and the maximum number of switching cycles can be increased by of factor of three [9]. On the other hand, in the case of Au–Ru and Au–Rd alloy very high val- ues of hardness can be reached [10,11], but the highest values are obtained only alloying gold with a high percentage of rhodium or ruthenium and in this case most of the useful properties of gold are lost, leading to wearing of contacts due to chemical contamination or brittle fracture [9]. Inclusion of nanoparticles in the gold contact seems to be another promising approach [7,8], that gives a consid- erable increase in hardness with respect to pure gold, but that still need to be fully tested from the contact performances perspective. From the technology point of view, most of the above pre- sented approaches have some technical difficulties in being simply inserted in a fabrication cycle, that would require, when possible, a significant modification of a consolidated process flow. For instance, including the gold–platinum (or other noble metal) alloy in the fab- rication process is not always simple. Many of these metals are is not so easy to pattern, and usually to obtain a good alloy a dedicated equipment or high temperature processes are required, which are in addition not always compatible with the fabrication flow [12]. The same holds for the case of incorporation of zirconium nanoparticles [7], which looks very promising from hardness point of view, but requires a dedicate technical equipment they may not be present in all foundries. Finally, the processes typically included in a fabri- cation cycle may impact on the properties of these new materials, 0921-5107/$ – see front matter © 2009 Elsevier B.V. All rights reserved. doi:10.1016/j.mseb.2009.06.004