Comparative study of different modes of calibration and quantification in radiofrequency glow discharge optical emission spectroscopy (rf-GD-OES)

Abstract

Three modes of calibration (normal, internal standard and sputtering rate), available in commercial quantification software, were evaluated for the determination of aluminium, carbon, iron, phosphorus, and silicon in various NIST SRM reference materials using radiofrequency glow discharge optical emission spectroscopy (rf-GD-OES). Incorporation of the dc-bias potential (V_dc) correction in the normal and sputtering rate modes of calibration were also investigated. Statistical figures of merit, including the square of the least-squares correlation coefficient (R²), minimum least squares coefficient (chi square, χ²), and standard error of estimate (SEE) were used to evaluate the calibration of the five test elements. The internal standard mode, using the intensity of the Ar (I) 404.44 nm transition as the internal standard, performed better than the normal and sputtering rate modes. Use of the V_dc correction on the calibration functions of the five elements by the latter two modes provides enhanced performance. The accuracy and precision of the measured concentrations of the five elements in the test samples further validated the statistical results of the various types of calibration; with and without the V_dc correction. The best accuracy results for elements with no apparent self-absorption (C, P and Si), were found to be ≤7% error, using either internal standard or sputtering rate quantification with V_dc correction modes. For elements with noticeable self-absorption (Al and Fe), the accuracy was found to be best when using sputtering rate calibration with V_dc correction (error ≤ 5%). However, the internal standard mode of calibration produced the best precision in the concentrations for all five elements with a 1.3–7.7% RSD range.