Abstract
When using electron-energy-loss spectroscopy (EELS) to quantify the bonding fraction in thin film carbon, the spectrum taken from the film must be compared to that of a suitable known standard. In contrast to the work of Berger [S. D. Berger and D. R. McKenzie, Philos. Mag. Lett. 57, 285 (1988)], fullerite is used in this work as the standard since highly oriented pyrolytic graphite (HOPG) is highly anisotropic, and can therefore lead to preferential orientation effects in EELS. It was found that had a shoulder on the high-energy side of the peak, which was centered at 287 eV. In theory, many Gaussian peaks can be fitted to the near edge structure (NES) of the energy-loss spectrum. In practice, we show that only three, centered at energies of 285, 287, and 293 eV, are necessary for semiquantitative analysis. The 285 eV peak is indicative of the bonding fraction; the second peak at 287 eV is attributed to molecular structure within the sample; the third at 293 eV is determined by bonding contributions in the sample. We show that by fitting these three peaks to fullerite and evaporated amorphous carbon (a-C), that the a-C has a bonding fraction compared with the fullerite standard. The importance of considering the 287 eV peak in highly bonded amorphous carbon is further illustrated by analysis of a-C:H:N thin films.
- Received 9 April 1999
DOI:https://doi.org/10.1103/PhysRevB.62.12628
©2000 American Physical Society