Journal of Alloys and Compounds 480 (2009) L13–L16 Contents lists available at ScienceDirect Journal of Alloys and Compounds journal homepage: www.elsevier.com/locate/jallcom Letter Crystallization of amorphous NiTiCu thin films Ye Xu a , Xu Huang a , A.G. Ramirez a,b,∗ a Department of Mechanical Engineering, Yale University, New Haven, CT 06520, United States b Department of Materials Science and Engineering, Massachusetts Institute of Technology, Cambridge, MA 02139, United States article info Article history: Received 15 December 2008 Received in revised form 13 January 2009 Accepted 22 January 2009 Available online 6 February 2009 Keywords: Crystallization NiTiCu Interfacial energy Thin films in situ TEM abstract The effect of copper additions (1.3 at%) on the crystallization of amorphous nickel-titanium thin films was explored. Using differential scanning calorimetry (DSC), the overall activation energy and crystallization temperatures were found to be similar to pure NiTi. However, in situ transmission electron microscopy (TEM) showed the nucleation and growth behaviors were markedly different. NiTiCu exhibited a lower nucleation rate, suggesting a higher activation energy in comparison to NiTi. The large grains in NiTiCu microstructures are consistent with a lower growth activation energy. These data suggest copper additions create higher interfacial energies. This paper presents experimental measurements that corroborate this mechanism. © 2009 Elsevier B.V. All rights reserved. 1. Introduction Several efforts have been made to incorporate NiTi thin films into microelectromechanical systems (MEMS), since they have large deformation and recovery forces [1–3]. Despite their superior actu- ation properties, the integration of shape memory alloys (SMA) in MEMS is limited, because their behavior shows an acute compo- sitional sensitivity, i.e. a small compositional shift causes a large transformation-temperature change [4]. One method of mitigating this compositional sensitivity is through alloying [5]. Recent stud- ies have found that substituting copper (<25 at%) for nickel in NiTi significantly reduces the compositional sensitivity [6]. Like NiTi, the as-deposited state of sputtered NiTiCu thin films is amorphous; annealing is required for crystallization. However, knowledge of the microstructural evolution of NiTiCu thin films during crystallization is limited. Preliminary studies on the crystallization behavior of NiTiCu have been done using differential scanning calorimetry (DSC) and X-ray diffraction [7–14]. However, most studies describe only the overall crystallization process rather than determining the individ- ual contributions of nucleation and growth. Thus, the mechanism by which the addition of copper affects the components of crystal- lization remains unknown. Aiming to elucidate the effect of copper on both nucleation and growth, we observed the crystallization of NiTiCu thin films using in situ TEM. ∗ Corresponding author. Tel.: +1 203.432.2156; fax +1 203.432.6775. E-mail address: ramirez@stanfordalumni.org (A.G. Ramirez). 2. Experimental methods Using two dc magnetron sputtering targets of NiTi (at 302 W) and Ti (at 150 W), and one rf magnetron sputtering target of Cu at 30W at a deposition rate of 0.67 mm/s, we deposited 0.2 m amorphous NiTiCu thin films onto silicon-nitride TEM membrane samples [15,16] and silicon wafers coated with 0.2 m silicon- nitride. The resulting composition of Ni48.3Ti50.4Cu1.3 was determined by electron microprobe analysis. Real-time in situ TEM heating studies were performed in bright-field mode using a 200 kV field-emission gun. NiTiCu TEM samples were heated within the micro- scope on a single-tilt holder with heating controllable up to 1200 ◦ C (and measured with a thermocouple). Using a low magnification to achieve a wide field of view and resolution of 50 nm, digital images were continuously recorded and time-stamped, giving a precise time-temperature record. Samples were heated at a rate of 50 ◦ C/min to target temperatures of 480, 500, and 520 ◦ C and then held constant until crystal- lization was complete. For comparison, we also analyzed binary NiTi samples with the same amount of Ti (50.4 at%), since copper substitutes for nickel [6]; with a thick- ness of 1 m deposited on silicon wafers both NiTiCu and NiTi films were tested to 550 ◦ C at rates of 50, 20 and 10 ◦ C/min, using differential scanning calorime- try (DSC). The contact angles of both distilled water and formamide droplets on both amorphous and crystallized NiTi and NiTiCu thin films were measured using a contact-angle measurement system. 3. Results and discussion The crystallization temperatures and overall activation energies of NiTi and NiTiCu films were determined by DSC. Fig. 1 shows our DSC curves of NiTi and NiTiCu thin films under different heating rates with the peak crystallization temperature, T p , indicated. Pre- vious research has shown a decrease in crystallization temperatures and activation energies with a significant amount of copper addi- tion (≥10 at%) [7–14]. The peak crystallization temperatures appear to remain nearly the same after adding a small amount of copper (of 0925-8388/$ – see front matter © 2009 Elsevier B.V. All rights reserved. doi:10.1016/j.jallcom.2009.01.113