Properties of AlN film grown on Si (111) Yiquan Dai a , Shuiming Li a,b,c , Qian Sun b,c , Qing Peng d,e , Chengqun Gui d , Yu Zhou b,c , Sheng Liu d,n a Mechanical Science & Engineering, Huazhong University of Science & Technology, Wuhan 430074, China b Key Laboratory of Nanodevices and Applications, Chinese Academy of Sciences (CAS), Suzhou 215123, China c Suzhou Institute of Nano-Tech and Nano-Bionics (SINANO), Chinese Academy of Sciences (CAS), Suzhou 215123, China d School of Power and Mechanical Engineering, Wuhan University, Wuhan, Hubei 430072, China e Department of Mechanical, Aerospace and Nuclear Engineering, Rensselaer Polytechnic Institute, Troy, NY 12180, USA article info Article history: Received 9 June 2015 Received in revised form 13 November 2015 Accepted 18 November 2015 Communicated by C. Caneau Available online 26 November 2015 Keywords: A1. Characterization A1. X-ray microdiffraction A3. Metalorganic chemical vapor deposition B1. Nitride B2. Semiconductor III–V materials B2. Piezoelectric materials abstract Stress and strain in an AlN film grown on Si (111) substrate have been evaluated by measuring Raman frequency shifts. Mechanical properties and phonon deformation potentials of AlN are evaluated by first principles calculations. The calculation model is verified by comparing the calculated Raman frequencies and frequencies detected from a bulk single crystal. Results show that the two sets of frequencies agree very well with each other. Thus, with the same verified model and parameters, elastic constants and phonon deformation potentials are calculated. Additionally, we successfully develop a numerical model to verify the calculation above and the model itself is also useful to predict properties of crystal films. Finally, the stress, strain, and piezoelectric properties are analyzed and compared for films on different substrates. & 2015 Elsevier B.V. All rights reserved. 1. Introduction Aluminum nitride (AlN) is an attractive candidate material for semiconductor applications [1], such as acoustic wave devices, UV solid-state light sources, MIS (metal–insulator– semiconductor) devices, and as a buffer-layer for gallium nitride (GaN). Though these exciting applications have been reported for years, there are still many problems due to lattice and thermal mismatch [2,3]. Large misfits between films and sub- strates lead to a high dislocation density, even cracking [4]. In order to deal with this problem, different film growth technol- ogies, mechanical characterizations, and growth parameter optimizations have been developed [5]. Furthermore, the stress state of a film is strongly correlated to the reliability of the metal line interconnection in devices and the percentage of disloca- tions in post thermal processing [6]. Directly measuring the strain state of these nitride films or microscale semiconductors is one of the potential techniques used for reliability characterization. In terms of stress/strain measurement, Raman scattering is a very convenient non-destructive stress detection method, espe- cially for materials or electronic devices bearing unknown loading corresponding to intrinsic phonon frequency shifts. Moreover, Raman active phonons are straightforward signatures of bonds within a sample's mixed phase and bonds at interfaces, by which the quality of film crystallization and its in-situ evolution can also be characterized [7]. Currently, its utility in III-V nitride com- pounds mainly depends on the elastic constants' determination and phonon deformation potential (PDP) calibration [8–11]. (PDP is defined as the (constant) coefficient between the Raman fre- quency shift and strain or stress quantization [8]). First principles calculations are one of the effective ways to rebuild Raman frequencies [12,13], based on which elastic con- stants [14–16] and phonons [17–19] can be analyzed theoretically. From recent publications [20–22], and early articles [14,15], one can find comparisons on AlN, which show some fluctuation in results. Most of the calculations are compared or verified just by other calculations, lacking independent theoretical or experi- mental support. For example, the authors in references [9–10,23] confuse whose elastic constants are the best and even calculate several sets of PDP based on different elastic constants. At the same time, the inconsistency of PDP appearing in literature also Contents lists available at ScienceDirect journal homepage: www.elsevier.com/locate/jcrysgro Journal of Crystal Growth http://dx.doi.org/10.1016/j.jcrysgro.2015.11.016 0022-0248/& 2015 Elsevier B.V. All rights reserved. n Corresponding author. E-mail address: victor_liu63@126.com (S. Liu). Journal of Crystal Growth 435 (2016) 76–83