Structural and mechanical properties of a-C:H thin films grown by RF-PECVD

Laboratoire des Matériaux Inorganiques, UMR CNRS 6002, Université Blaise Pascal (Clermont-Ferrand 2), 24 Av des Landais, 63177 Aubière cedex, France
Diamond and Related Materials (Impact Factor: 1.57). 09/2004; 13(9):1618-1624. DOI: 10.1016/j.diamond.2004.01.017

ABSTRACT a-C:H thin films have been deposited by plasma enhanced chemical vapor deposition at 13.56 MHz in a large bias voltage range (200–600 V). Their mechanical properties have been studied through different techniques such as nanoindentation, profilometry and finally X-ray reflectometry in order to determine their hardness, elastic modulus, stress level, interfacial fracture energy and density. We show that the hardness, the Young modulus, the density and the stress level decrease with increasing bias voltage contrary to the interfacial fracture energy, hence the adhesion of the deposits. To understand this behavior, films microstructure has been characterized through the determination of hydrogen and sp3 hybridized carbon contents by ERDA, FTIR, Raman diffusion and ESR. We have demonstrated that a graphitization and a loss of hydrogen occur when the bias increases. Such evolutions are explained in terms of the subplantation model proposed by Lifshitz. Finally, hard and dense coatings can be grown at low bias as the sp3 carbon content is high, but their corresponding adhesion is low.


Available from: Laurent Thomas, May 07, 2015
1 Follower
  • [Show abstract] [Hide abstract]
    ABSTRACT: Amorphous hydrogenated carbon (a-C:H) and amorphous deuterated carbon (a-C:D) films were prepared using plasma-enhanced chemical vapor deposition (PECVD) from CD4, H2, CD4 and D2 source gases. Fourteen different samples were prepared by changing the source gas ratios of CH4/H2, CH4/D2, CD4/H2, and CD4/D2. The concentrations of hydrogen (H) and deuterium (D) relative to carbon (C) in the films were determined by elastic recoil detection analysis (ERDA) and Rutherford backscattering spectroscopy (RBS). The sp2/(sp2 + sp3) ratios of the films were analyzed by near-edge X-ray absorption fine structure (NEXAFS) measurements. Hardness and mass density of the films were measured using a nanoindenter and X-ray reflectivity (XRR), respectively. For all combinations of source gas, the H and D concentrations varied by only 4.0 at.%. For the CH4/D2 source gas, the D concentration in the film increased from 0 at.% with the D2/(CH4 + D2) source gas combination to 11.2 at.% for the 80% D2/(CH4 + D2) source gas combination. The increase in D concentration exceeded the increase in total H and D concentration (3.4 at.%). For CH4/D2 source gas, the H concentration decreased as the D concentration increased. For the CD4/H2 source gas, we observed the opposite tendency. Additionally, an isotope effect between the a-C:H films and the a-C:D films was observed, with preferential incorporation of H over D. From the NEXAFS measurements, the sp2/(sp2 + sp3) ratios in all of the samples were between 38.8% and 40.8%. A correlation between the sp2/(sp2 + sp3) ratio and the H2 or D2 gas source ratio was not observed. The hardness and density of the films decreased when the H2 or D2 source gas ratio increased. Even though the H concentration in the a-C:H films was higher than the D concentration in the a-C:D films, the a-C:D films had lower hardness and mass density values. These findings suggest that information concerning the voids, nanostructures, sp2/sp3 ratios and H concentrations of hydrogenated amorphous carbon films is crucial for evaluation of their mechanical properties.
    Applied Surface Science 01/2013; 265:750-757. DOI:10.1016/j.apsusc.2012.11.100 · 2.54 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: The thermophysical properties of hydrogenated amorphous carbon (a-C:H) thin films, namely the thermal diffusivity α, thermal effusivity ɛ, thermal conductivity κ and volumetric heat capacity Cf have been measured simultaneously using ultrafast optical pump–probe technique. The pump–probe method utilizes the film surface temperature decay following a heating pulse from a pump laser. From the temperature decay, the thermal diffusivity α and effusivity ɛ were obtained by curve-fittings of two different physical models. The thermal conductivity κ and volumetric heat capacity Cf were then deduced from the measured α and ɛ using the relations κ = α1/2ɛ and Cf = ɛ/α1/2. The thermal conductivity and volumetric heat capacity obtained compare favorably with those reported in the literatures. The method is independent of substrate and Al capping layers. The curve-fittings are simpler because independent of parameters of the films and the Kapitza boundary thermal resistance at Al/a-C:H interface.
    International Journal of Heat and Mass Transfer 08/2013; 63:199–203. DOI:10.1016/j.ijheatmasstransfer.2013.03.062 · 2.52 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: In this work, we focused on investigations of mechanical properties of SiNx and diamond-like carbon thin films deposited by plasma-enhanced chemical vapour deposition method for application in optical devices or solar cells. Mechanical properties of thin films deposited on clean and oxidized silicon substrates were determined by nanoindentation. The main difficulty with the characterization of thin films using nanoindentation method is related to the influence of the substrate on the measured properties of thin layers. We proposed a method to determine the mechanical properties (hardness and Young's modulus) of thin films in single-layer/substrate or double-layer/substrate system through combining the finite element method, nanoindentation experiments and numerical simulations. In this study, a three-dimensional numerical model of nanoindentation tests performed with Vickers diamond indenter was examined to determine the stress distributions during measurement with various maximum loads. The hardness and Young's modulus of the examined layers were determined using two types of procedures, depending on the von Mises equivalent stress distribution obtained at the maximum load. If the size of an elastically deformed region was sufficiently small compared with the thickness of the measured layers, we applied a standard method of measuring at the depth equal to 10% of the layer thickness; otherwise, an approximation method was used to reduce the substrate influence. Copyright © 2014 John Wiley & Sons, Ltd.
    Surface and Interface Analysis 10/2014; 46(10-11). DOI:10.1002/sia.5473 · 1.39 Impact Factor