Quantification of low levels of fluorine content in thin films F.J. Ferrer a,⇑ , J. Gil-Rostra b , A. Terriza b , G. Rey c , C. Jiménez c , J. García-López a , F. Yubero b a Centro Nacional de Aceleradores (CSIC – Univ. Sevilla), Av. Thomas A. Edison 7, E-41092 Sevilla, Spain b Instituto de Ciencia de Materiales (CSIC – Univ. Sevilla), Américo Vespucio 49, E-41092 Sevilla, Spain c Laboratoire des Matériaux et du Génie Physique-UMR 5628-INPGrenoble-Minatec 3, parvis Louis Néel, BP 257, 38016 Grenoble Cedex 1, France article info Article history: Received 24 October 2011 Received in revised form 24 November 2011 Available online 6 December 2011 Keywords: Fluorine content Rutherford Backscattering Spectrometry Particle induced gamma-ray emission Fluorine-doped tin oxides Fluorinated silica films Fluorinated carbon films abstract Fluorine quantification in thin film samples containing different amounts of fluorine atoms was accom- plished by combining proton-Rutherford Backscattering Spectrometry (p-RBS) and proton induced gamma-ray emission (PIGE) using proton beams of 1550 and 2330 keV for p-RBS and PIGE measure- ments, respectively. The capabilities of the proposed quantification method are illustrated with examples of the analysis of a series of samples of fluorine-doped tin oxides, fluorinated silica, and fluorinated dia- mond-like carbon films. It is shown that this procedure allows the quantification of F contents as low as 1 at.% in thin films with thicknesses in the 100–400 nm range. Ó 2011 Elsevier B.V. All rights reserved. 1. Introduction Quantification of the fluorine content in thin films is an impor- tant issue in many technological fields. For example, transport performance in transparent conductive oxide thin films based on fluorine-doped tin oxides (F-SnO 2 ) is strongly correlated to their fluorine content [1]. This type of films is also extensively used as gas sensors [2] or low-emissivity (low-e) coatings for the glass industry [3]. Precise knowledge of the fluorine content is also an important issue in low refractive index films and low dielectric constant (j) interlayer dielectrics [4,5] based on fluori- nated silica films (SiOF) [6,7]. Another example is the coatings with antiseptic functionality based on fluorinated carbon films (CF x ), where a correlation between hydrophobicity and bacterial anti-adherence performance with their F content has been reported [8]. Despite the importance of the control of the fluorine content for the applications mentioned above, the quantification of this ele- ment in thin films is not an easy task. Fluorine is a light element (Z = 9), difficult to detect with techniques based on X-ray emission, such as energy dispersive analysis of X-rays (EDAX). Other tech- niques, like X-ray photoelectron spectroscopy (XPS), can easily quantify the presence of fluorine content within a surface but often these surface measurements are not representative of the fluorine thin film composition. Infrared spectroscopy is also often used to this aim. However, although this technique is strongly sensitive to F–X bond type formation [9,10], its quantitative information is more doubtful. Standard Rutherford Backscattering Spectrometry (RBS), using alpha particles as probe projectiles, is also commonly used to determine the fluorine content in thin films. In this case, the fluorinated thin films are quite often deposited on artificial light substrates such as graphite [11] to simplify the elemental quanti- fication. This is due to the low scattering cross section and kinematic factor for this element that, in most cases, strongly handicaps its quantification. The choice of graphite as a substrate generally avoids the overlapping of signals from the substrate and the deposited fluorinated film, so that fluorine quantification becomes straightforward. However, this trick is not always possi- ble and quantification of low fluorine content in thin films depos- ited in other standard substrates, such as Si wafers or glass, is generally not an easy task. Nuclear Reaction Analysis (NRA) by means of several reactions (c.f., 19 F(p,a) 16 O [12], 19 F(p,ac) 16 O [13], 19 F(p,p 0 c) 19 F [14], 19 F(a,p) 22 Ne [15], 19 F(a,a 0 c) 22 Ne[16], 19 F(d,p) 20 F, 19 F(d,a) 17 O [17]) is also extensively used for fluorine quantification purposes. In this case, quantification relies on comparative studies with well-known ref- erence samples (NdF 3 , AlF 3 or samples prepared with known doses of fluorine implantation [18]). In this work we propose the combination of proton RBS (p-RBS) and proton-induced gamma-ray emission (PIGE) ( 19 F(p,p 0 c) 19 F reaction) to determine the fluorine content in thin films deposited on different substrates (Si wafers, glass, fused silica). The method 0168-583X/$ - see front matter Ó 2011 Elsevier B.V. All rights reserved. doi:10.1016/j.nimb.2011.11.042 ⇑ Corresponding author. Address: Centro Nacional de Aceleradores (CSIC – Univ. Sevilla), Isla de la Cartuja, E-41092 Sevilla, Spain. Tel.: +34 95 4460553; fax: +34 95 4460145. E-mail address: fjferrer@us.es (F.J. Ferrer). Nuclear Instruments and Methods in Physics Research B 274 (2012) 65–69 Contents lists available at SciVerse ScienceDirect Nuclear Instruments and Methods in Physics Research B journal homepage: www.elsevier.com/locate/nimb