Detection limits in Auger electron spectroscopy

The sources of background noise in Auger electron microanalysis are analyzed in order to evaluate the minimum detectable concentration x_m and the minimum number of detectable atoms y_m that can be reached. The best choices of operating conditions (the energy E₀, intensity I₀, and spot size d₀ of the incident electron beam, and the duration of the experiment t_e) are deduced for bulk and thin film analysis. The main results are: (i) The choice of E₀ is not very stringent, at least when E₀ ? 5E_i(A), where E_i(A) is the ionization energy, (ii) For a given electron dose received by the sample, x_m is improved by the use of the largest incident spot size while y_m is improved by the use of the finest spot size. The results also hold for other microanalytical techniques such as electron energy loss spectroscopy or electron probe microanalysis. (iii) Chemical identification of a single atom will be possible on samples able to tolerate very large electron doses by using incident electron beams 10 nm or less in diameter. The expected performance of a coincidence technique first suggested by Wittry is also discussed.