Conference Paper

Composition and optical properties of silicon oxynitride films deposited by electron cyclotron resonance

  • A. Del Prado /
  • E. San Andrés /
  • F. L. Martínez /
  • I. Mártil /
  • G. González-Díaz /
  • W. Bohne /
  • J. Röhrich /
  • B. Selle /
  • M. Fernández
Journal cp
  • Volumen: 67
  • Número: 3-4
  • Fecha: 26 September 2002
  • Páginas: 507-512
  • ISSN: 0042207X
  • Source Type: Journal
  • DOI: 10.1016/S0042-207X(02)00220-8
  • Document Type: Conference Paper
Silicon oxynitride films covering the whole composition range from silicon nitride to silicon oxide have been deposited by electron cyclotron resonance chemical vapor deposition from SiH4, O2 and N2 gas mixtures. The composition of the films has been determined by heavy-ion elastic recoil detection analysis (HI-ERDA), providing absolute concentrations of all elements, including H, and by Auger electron spectroscopy. Additionally, Fourier transform infrared (FTIR) spectroscopy and ellipsometry measurements have been performed on the same samples for optical characterization. The concentration of the different species (Si, O, N and H) and the density of the films have been calculated and compared to the theoretical values for stoichiometric films. The presence of N-H bonds and non-bonded H results in a significant decrease of the Si concentration with respect to the theoretical value, especially for samples close to silicon nitride composition. The decrease of the Si concentration results in a decrease of both the N and O concentrations. The overall result is a decrease of the density and therefore a decrease of the refractive index with respect to stoichiometric films. The total H content determined by ERDA has been compared with the area of the FTIR N-H stretching band, which is frequently used to obtain the H content. It has been found that the calibration factor for this band depends on composition, increasing with increasing the O content. © 2002 Elsevier Science Ltd. All rights reserved.

Author keywords

    Indexed keywords

      Funding details