650 JOURNAL OF MICROELECTROMECHANICAL SYSTEMS, VOL. 16, NO. 3, JUNE 2007 A MEMS-Based Evaluation of the Mechanical Properties of Metallic Thin Films Arun Reddy, H. Kahn, and Arthur H. Heuer Abstract—Characterizing the mechanical properties of metal thin films is critical for the design and fabrication of metal mi- croelectromechanical systems and integrated circuit devices. This paper focuses on wafer-level determination of the mechanical be- havior of sputtered aluminum and nickel thin films, using a variety of measurement techniques. Elastic moduli have been determined in devices fabricated with standard micromachining techniques using bulge testing of square diaphragms and lateral resonator structures. We find a Young’s modulus of 70 GPa for Al and 200 GPa for Ni, in agreement with data for the bulk metals. Using pressurize/depressurize cycles, the load–deflection curves of the membranes have also been determined, and in conjunction with finite element simulations, were used to determine the yield strength and fracture strength of these films. Residual stresses in the films have also been investigated using wafer curvature, bulge testing, and X-ray diffraction. The merits of each measurement technique are discussed. [2006-0087] Index Terms—Metal thin films, micromachining, residual stresses, strength, Young’s modulus. I. I NTRODUCTION M ETAL thin films are one of the most versatile classes of materials used in the microelectromechanical systems (MEMS) and integrated circuit (IC) industries. Aluminum thin films are used in applications ranging from sensors [1] and thin- film transistors [2] to field emission displays [3]. To fully ex- ploit MEMS and microelectronics technologies, the reliability of metallic thin films used in making such miniature structures has to be established unambiguously. Characterizing the mechanical properties of thin films, in- cluding their elastic moduli, strength, and the nature and mag- nitude of residual stresses, allows estimation of their reliability in service. Although bulk metals are well understood, the same is not true for thin films. This arises due to intricacies and com- plexities involved in fabrication of specimens and subsequent handling during testing. Evidence exists [4], [5] that suggests that the properties of thin films cannot be extrapolated from their bulk counterparts because length-scale effects become important at small dimensions. Thus, specialized techniques for Manuscript received May 11, 2006; revised July 24, 2006. Subject Editor S. Spearing. A. Reddy was with the Department of Materials Science and Engineering, Case Western Reserve University, Cleaveland, OH 44106 USA. He is now with Rohm and Haas Electronic Materials, Phoenix, AZ 85034 USA (e-mail: areddy@rohmhaas.com). H. Kahn and A. H. Heuer are with the Department of Materials Science and Engineering, Case Western Reserve University, Cleaveland, OH 44106 USA (e-mail: harold.kahn@case.edu; heuer@case.edu). Color versions of one or more of the figures in this paper are available online at http://ieeexplore.ieee.org. Digital Object Identifier 10.1109/JMEMS.2007.892912 testing thin films have been developed [6], both for freestanding films and for films on substrates. However, reported mechanical property data are very variable, either due to the experimental errors attendant on a particular testing method or to differing process parameters used for thin film fabrication. The elastic anisotropy coefficients of Al and Ni (2(s 11 - s 12 )/(s 44 )) are 1.2 and 2.5, respectively [7]. These coefficients refer to the differences in elastic modulus as a function of crystallographic direction in the crystal. As a result, highly textured Ni films with varying orientation can display different elastic moduli, whereas those of Al will show less variation. Therefore, the processing path by which these films are deposited, and the resulting texture (if any) should also be taken into account before using the values for designing MEMS structures. Ex- perimental difficulties are thus implicated in the fact that the Young’s modulus of Al has been reported to range from 25 to 90 GPa [8], [9]. In this paper, elastic moduli of Al thin films were determined using both lateral resonators and bulge testing of membranes, with specimen dimensions comparable to typical MEMS de- vices. They were fabricated using standard MEMS bulk and surface micromachining techniques. Residual stresses in the Al films were obtained by bulge testing and wafer curvature techniques. Finally, burst testing, which is an extension of the bulge test, was used to determine the yield strength and fracture strength of the Al thin films. Similar but more limited data on sputtered Ni thin films are reported in Appendix A. II. THIN FILM/DEVICE FABRICATION The Al films used in this paper were deposited on Si wafers at room temperature in the Case Microfabrication Laboratory (MFL) class 100 clean room using dc magnetron sputtering. Cathode power was maintained at 750 W, and high-purity argon was used at 5 mTorr (0.7 Pa) pressure as the sputtering gas. The deposition rate of Al was 1.4 nm/s. The temperature of the substrate, which is initially at 18 C, increased to 30 C during the deposition. For membrane fabrication, a 0.2-μm-thick SiN x film was deposited by low-pressure chemical vapor deposition (LPCVD) onto both sides of a 100-mm-diameter double-side- polished (100) Si wafer. The backsides of the wafers were then patterned using standard photolithography to define square windows in the SiN x , which were etched using an SF 6 plasma. The exposed square windows of Si were etched anisotropically in aqueous 3.5 M KOH solution at 55 C, the Si (111) planes defining the SiN x membranes. The Al films (1.0 and 1.4 μm thick) were then deposited onto the front side of the wafers to form a metal–SiN x composite membrane. Finally, the SiN x 1057-7157/$25.00 © 2007 IEEE