2012 Volume 6 Issue 2 Pages 129-139
The composition of the BN film was determined using Rutherford backscattering spectrometry (RBS) and nuclear reaction analysis (NRA). RBS can provide all the constituent concentrations in the films and the absolute determination of the number of atoms/cm2. However RBS is not suited to detection of light atoms deposited on a substrate material of higher atomic mass. On the other hand, the NRA has the advantage that it allows to measure the areal concentrations of nitrogen and boron in BNx films on Si substrates, although calibration is required using standard specimens. These experiments were carried out on the 2 MeV Van de Graaff accelerator connected to an ultra high vacuum (UHV) chamber. For RBS measurement, a 42He+ beam at 2.0 MeV, a total scattering angle of 168° and a beam incident angle to the substrate normal or 60 deg. were used. Zr and Pt films, 1150 Å to 3300 Å in thickness, deposited on vitreous carbon plates were used as a substrate. NRA was performed using a deuteron beam of 1.7 MeV and a beam incident angle of 20 deg. A peak from 10B(d,α)8Be in an NRA spectrum of a standard sample appeared clearly without significant background, however a broad signal from 11B(d, α)9Be appeared overlapping with a peak from 14N(d, α)12C. Therefore the 10B(d, α)8Be and 14N(d, α)12C yields were estimated, since the ratio of 11B : 10B measured by RBS was 0.83 : 17, which is well consistent with the natural isotopic ratio, 11B : 10B =0.802 :0.192. In the case of calculating the 14N(d, α)12C yields, the signal from 11B(d, α)9Be was deconvoluted by taking into account the shape of 11B(d, α)9Be signal. The areal ratio 14N/10B was 7.73 and the error was -3.5 to +3.2%. These values will be used for determining composition of BN films. The conversion factor allows obtaining the composition of BN thin films on Si substrate.
