Determination of relative sensitivity factors of elements in high purity copper by doping-melting and doping-pressed methods using glow discharge mass spectrometry

Glow discharge mass spectrometry (GDMS) with its high resolution, low limit of determination and wide linear dynamic range is a powerful technique for direct analysis of high purity solid metals. Similar to the drawback of other direct solid sampling methods, a drawback of this technique is the lack of appropriate pure certified reference materials to calibrate the instrument, so that it is often used for semi-quantity analysis. A new calibration procedure using synthetic calibration samples by the doping-melting method was developed to obtain more accurate and traceable analytical results. The calibration samples were prepared from high purity copper powder doped with standard solutions of graduated and defined concentration. And then a part of them was melted into bulks and another was pressed to tablets. The homogeneity of both types of samples was measured by GDMS and for the doped tablets, relative deviations of the IBRs of both sides for 45 elements were mostly less than 10%. For the doped bulks, some problems such as the segregation effect and analyte losses happened for most elements; so satisfactory homogeneity was obtained for only 9 elements, for which nearly all relative deviations of the IBRs of both sides were less than 5%. Relative sensitivity factors of 9 and 45 elements in the copper matrix were obtained by calibration of GDMS with doped bulks and doped tablets, respectively. The RSF values obtained by calibration with doped tablets were compared with values derived from standard RSFs and the large differences demonstrated that matrix-adapted calibration was necessary when the impurities were measured accurately by using the instrument, especially for the rare earth elements. Then the 9 RSFs determined using tablets and bulks were compared and the results showed that relative deviations of them in most cases were less than 10%. The correlation coefficients of the calibration curves obtained by using the doping-melting method were better than those obtained using the doping-pressed method and the uncertainties of RSFs were smaller. Several copper CRMs were measured to validate the calibration procedures. The results revealed that the measured values using the two types of calibration samples were in agreement with the certified values. The results revealed that the calibration results of Fe and Ni coming from calibration samples prepared by using the two types of methods were in agreement with the certified values.