Diamond and Related Materials 12 (2003) 974–978 0925-9635/03/$ - see front matter  2003 Elsevier Science B.V. All rights reserved. PII: S0925-9635 Ž 02 . 00374-6 XPS and laser Raman analysis of hydrogenated amorphous carbon films J. Filik , P.W. May *, S.R.J. Pearce , R.K. Wild , K.R. Hallam a a, a b b School of Chemistry, University of Bristol, Cantock’s Close, Bristol BS8 1TS, UK a Interface Analysis Centre, University of Bristol, Oldbury House, 121 St Michael’s Hill, Bristol BS2 8BS, UK b Abstract Hydrogenated amorphous carbon films were deposited in an RF parallel plate plasma reactor using various values of process pressure (10–50 mTorr) and DC self-bias (0–300 V). The films were then analysed by laser Raman spectroscopy (LRS) at 514.5 nm and X-ray photoelectron spectroscopy (XPS). Values for the ratio of sp :sp bonded carbon in the various films were obtained 2 3 by suitable fitting of the XPS carbon 1s energy peaks, using a three-curve fitting procedure, which recognises a portion of the peak attributable to CO surface bonding. The sp content was found to depend upon the DC self bias (and hence the ion impact 3 energy) during deposition, peaking at a value of 81% at approximately 150 V. The softer films grown at lower DC bias values still had an sp content of approximately 70%. Microcombustion analysis showed that films deposited with low DC bias contained 3 7 at.% H compared to less than 2 at.% for films deposited at biases greater than 100 V. This high sp content can be explained 3 by H-termination of dangling bonds, suggesting that sp content alone is not a reliable indication of film properties. Curve-fittings 3 of LRS spectra of the films showed that the Breit–Wigner–Fano lineshape is inappropriate for use with hydrogen containing films. Fitting using a Gaussian profile gave precise values for the FWHM, intensity, and Stokes’ shift of the G and D-peaks. A linear relationship between the intensity ratio of the D to G peaks and the width of the G peak was found for films deposited at high DC bias (with low H content), but not for films deposited at low DC bias. This is consistent with the increased H content of the films causing a change in the elastic constants andyor affecting the stress levels within the films.  2003 Elsevier Science B.V. All rights reserved. Keywords: Diamond-like carbon; Laser Raman analysis; XPS; Film composition 1. Introduction Diamond-like carbon (DLC) thin films are becoming increasingly important in many forms of industrial appli- cations, including wear-resistant coatings for hard-disk drives and optical components, as well as in semi- conductor devices w1–3x. However, the term DLC covers a range of materials with properties that can vary from those similar to graphite to those approaching those of natural diamond. Hydrogen-free DLC films (also known as amorphous carbon films, a-C w4x, are believed to consist of a mainly graphitic sp carbon matrix contain- 2 ing nm-sized clusters of sp diamond-like carbon. 3 Hydrogenated DLC films (also known as hydrogenated amorphous carbon films, a-C:H) add yet more degrees of complexity to the possible structure of the films w5x. One of the most crucial factors determining the film properties is the ratio of sp :sp bonded carbon. In a-C 2 3 *Corresponding author. Tel.: q44-0117-9289927; fax: q44-0117- 9251295. E-mail address: paul.may@bris.ac.uk (P.W. May). films, laser Raman spectroscopy (LRS) can be used to determine this ratio w6x, but in a-C:H films more sophis- ticated techniques such as electron energy loss spectro- scopy (EELS) or X-ray photoelectron spectroscopy (XPS) often need to be used. In this paper we use XPS in conjunction with LRS to study the sp :sp ratio in a- 2 3 C:H films grown with a variety of ion impact energies and deposition conditions, and to gain an insight into the structure and bonding within these films. 2. Experimental a-C:H films were deposited onto single crystal Si (100) substrates using a capacitively-coupled radio fre- quency (RF) parallel-plate plasma reactor. The process gas was methane (CH ) at a flow rate of 30 sccm and 4 a pressure ranging from 10–50 mTorr. RF power was used to strike a plasma and the ions were then acceler- ated onto the lower electrode by the DC self-bias. The substrate sat on the lower electrode, and the value of the DC bias in this system thus gives an approximate