Boron Doping of Microcrystalline and Nanocrystalline Diamond Films: Where is the Boron Going?

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 ~5×10 20 cm -3 up to ~6 ×10 21 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.

Scanning transmission electron microscopy (STEM) study on surface modified CVD diamond/Si(111) film post implanted Fe-B and NiFe-B related to GMR properties

“…At low dose of 1.0x10 15 ions/cm 2 (blue line), the sample shows strong Raman peak characteristic for diamond at wave number of 1328 cm -1 , while broaden profile with large full wave half maximum (FWHM) might be related to oxygen concentration at the interface between CVD diamond film and Silicon substrate which is in agreement with result of the STEM-EDX measurement. On the other hand, in sample A-E3D1 implanted with the same ion dose of 1.0x10 15 ions/cm 2 but higher ion energy (red line), the diamond characteristic peak is still stronger while the graphite peak invisible. This proves that the implanted layer is free from graphitic bonds which is in agreement with the study as reported in [13].…”

“…The first and second order of Silicon substrate Raman peaks are found at the wave number 500 and 1000 cm -1 as corfirmed by May et al [14]. Asymmetry peak profile at wave number 1120 cm -1 may be related to the Fano lineshape due to quantum mechanical interference induced by boron dopant as reported in [15] and this Fano interference effects were much reduced for the smaller grain sized. As indicated by the peak at wave number lower than 500 cm -1 the intensity of paired boron atoms increased and have tendency of build up at crystalline boundary by increasing of ion energy [16].…”

“…Two kinds of samples were used in this experiment, firstly sample is post NiFe-B at E=70keV and Dosis=10 15 ions/cm 2 (denoted as A-E3D1). Secondly, is post FeB at E=20 keV and dosis=10 15 ions/cm 2 (denoted as B-E1D1) followed by Argon ion sputtering at 70 keV and dose 10 16 ions/cm -2 .…”

“…Two kinds of samples were used in this experiment, firstly sample is post NiFe-B at E=70keV and Dosis=10 15 ions/cm 2 (denoted as A-E3D1). Secondly, is post FeB at E=20 keV and dosis=10 15 ions/cm 2 (denoted as B-E1D1) followed by Argon ion sputtering at 70 keV and dose 10 16 ions/cm -2 . Phase identification of the film was performed using Raman Spectroscopy while the magnetic properties were evaluated by the vibrating sample magnetometer and the magnetoresistance were studied by four point probe method.…”

“…All samples for TEM investigation were prepared in the Focus Ion Beam workstation (FIB) FEI Strata 205. Figure 1a shows a typical STEM darkfield image of CVD diamond film after NiFe+B ion implantation at energy 70 keV and dose 1.0x10 15 ion/cm 2 giving the over all film thickness around 1000 nm. The film from the top surface to the bottom as shown in Fig.…”