Boron Doping of Microcrystalline and Nanocrystalline Diamond Films: Where is the Boron Going?
- PDF / 1,320,411 Bytes
- 9 Pages / 612 x 792 pts (letter) Page_size
- 111 Downloads / 216 Views
1039-P17-03
Boron Doping of Microcrystalline and Nanocrystalline Diamond Films: Where is the Boron Going? Paul William May1, William J Ludlow1, Matthew Hannaway1, James A Smith1, Keith N Rosser1, and Peter J Heard2 1 School of Chemistry, University of Bristol, Cantock's Close, Bristol, BS8 1TS, United Kingdom 2 Interface Analysis Centre, University of Bristol, 121 St. Michael's Hill, Bristol, BS2 8BS, United Kingdom
ABSTRACT We present data showing how the electrical conductivity and Raman spectra of boron doped ‘cauliflower’-type nanocrystalline (c-NCD) CVD diamond films vary as a function of B content. The conductivity is roughly linear as a function of B content between an onset threshold of ~5×1020 cm-3 up to ~6 ×1021 cm-3, with the higher concentrations giving near metallic conductivity values. The onset threshold may be due to compensating donors due to the large number of impurities and defects in these films. The position of the Lorentzian contribution to the 500 cm-1 Raman feature was used to estimate the B content and compared to the value measured using SIMS. We found that the Raman method overestimated the concentration of B by a factor of ~5 for these c-NCD films. The shortfall may be explained if only a small fraction of the B found in the small-grained films is being incorporated into substitutional sites. We conclude that in diamond films with a high concentration of grain boundaries, the majority of the B (80% in some cases) must be present at or in the grain boundaries. INTRODUCTION Diamond films produced by chemical vapour deposition (CVD) can be doped with boron to produce a p-type semiconducting material with electrical conductivity that ranges from insulating to metallic, depending upon the doping level [1]. The boron dopant atoms act as electron acceptors, and form a band located ~0.35 eV above the valence band edge. At low temperatures or at boron concentrations 5×1018 cm-3 it disappeared altogether giving a flat baseline. In order to see the small Raman features more clearly, and to reproduce the spectra from the lightly doped films on the same scale as the heavily doped ones, it was useful to subtract this PL background where necessary. This was done by fitting a polynomial curve to the background, and then subtracting this to obtain a processed spectrum. With increasing B content, the films became visibly darker and often looked bluish, and the Raman signals became correspondingly less intense. Longer accumulation times were required to obtain reasonable signal:noise levels in these samples. In those samples where the 500 cm-1 band was discernible (i.e. the more heavily doped samples), the band was fitted to a Gaussian curve centred ~500-550 cm-1 and a Lorentzian curve centred ~460-505 cm-1 following the procedure mentioned in the Introduction and outlined in ref.[20]. Using the relationship between the B content and the wavenumber of the Lorentzian component of the 500 cm-1 band (Eqn.1), it was possible to estimate the doping levels in these films, with which to compare the values from
Data Loading...
