The interferences caused by reactions with fast neutrons may be corrected with the help of substances containing the interfering element, which are irradiated together with samples and standards. For the correction of the apparent sodium content caused by the28Si(n,p)-reaction we need silicon with an extremely low sodium content. Irradiation without such a standard needs calculation for interference with the help of flux monitors. Because of the highly effective threshold energy of the reaction an additional correction factor taking into account various neutron spectra in different irradiation positions is necessary.
The energy-dependent range of charged particles in activation analysis according to the reaction12C(d,n)13N permits the method to be applied to carbon determination in model epitaxial layers of sufficient thickness. We investigated
100 μm epitaxial layers of the n-type and undoped 50 μ layers as to p− Czochralski substrates. Deuterons were slowed down with Cu and Ta foils having a limiting energy of 13.5 MeV, to 4.2 MeV
and 2.9 MeV, respectively. In the resulting activation depths of 52 and 102 μm, the sensitivity of the method, which is 3·1014 at ·cm−3C at Ed=10 MeV in silicon, is reduced to 25% and 10%, respectively. An optimal flux of 0.9 μA·cm−2 was maintained. After irradiation, 20 or 10 μm were etched off. The sample was inductively fused at 1500 K in a Pb3O4/B2O3 mixture.13N was passed with He as carrier gas into an absorption vessel kept at 77 K, and its activity was measured in γ, γ-coincidence.
Mg, Ti, Ni and Zn have been determined in semiconductor silicon by deuteron activation analysis using radiochemical separation and coincidence spectrometric methods. Detection limits under the experimental conditions are 15, 3, 0.7, and 14 ppb, respectively.
Authors:M. Schulepnikov, H. Rausch, J. Dubnack, D. Birnstein, and S. Niese
In an interlaboratory comparison four samples of monocrystalline silicon were analyzed sequentially by INAA in four laboratories. The results obtained demonstrate the possiblity of repeated use of silicion samples for interlaboratory comparisons. *** DIRECT SUPPORT *** A0653052 00003
Authors:M. Böttger, D. Birnstein, W. Helbig, and S. Niese
A sensitive method of carbon determination in silicon by activation with deuterons and3He particles followed by chemical separation of the measured nuclides13N and11C has been developed. Aside from the etch removal after irradiation, it is essential for a low detection limit to avoid too
much thermal exposure of the target surface during the irradiation process, or else microflaw formation and diffusion on the
surface have to be anticipated. The ion current must have an optimal rather than a maximum value.
Authors:M. Böttger, S. Niese, D. Birnstein, and W. Helbig
NAA and CPAA are the most sensitive analytical tools for the determination of residual impurities in semiconductor silicon. Extremely small detection limits are found by measuring the activity in an underground laboratory. The accuracy and reliability of the methods were checked by several comparisons: interlaboratory comparison on NAA, another CPAA with IR-spectrometry as well as comparison of NAA-results, obtained instrumentally and by chemical separation.