Controlled dissolution of surface layers for elemental analysis by inductively coupled plasma-mass spectrometry

Abstract

Determining the composition of thin layers is increasingly important for a variety of industrial materials such as adhesives, coatings and microelectronics. Secondary ion mass spectrometry (SIMS), Auger electron spectroscopy (AES), X-ray photoelectron spectroscopy (XPS), glow discharge optical emission spectroscopy (GDOES), glow discharge mass spectrometry (GDMS), and laser ablation-inductively coupled plasma-mass spectrometry (LA-ICP-MS) are some of the techniques that are currently employed for the direct analysis of the sample surface. Although these techniques do not suffer from the contamination problems that often plague sample dissolution studies, they do require matrix matched standards for quantification. Often, these standards are not readily available. Despite the costs of clean hoods, Teflon pipette tips and bottles, and pure acids, partial sample dissolution is the primary method used in the semiconductor industry to quantify surface impurities. Specifically, vapor phase decomposition (VPD) coupled to ICP-MS or total reflection x-ray fluorescence (TXRF) provides elemental information from the top most surface layers at detection sensitivities in the 10{sup 7}-10{sup 10}atoms/cm{sup 2} range. The ability to quantify with standard solutions is a main advantage of these techniques. Li and Houk applied a VPD-like technique to steel. The signal ratio of trace element to matrix element was used for quantification. Although controlled dissolution concentrations determined for some of the dissolved elements agreed with the certified values, concentrations determined for refractory elements (Ti, Nb and Ta) were too low. LA-ICP-MS and scanning electron microscopy (SEM) measurements indicated that carbide grains distributed throughout the matrix were high in these refractory elements. These elements dissolved at a slower rate than the matrix element, Fe. If the analyte element is not removed at a rate similar to the matrix element a true representation of the sample layer cannot be realized. Specifically, the ratio of analyte signal to matrix element signal does not equal the actual ratio in the bulk sample. The objective of this work was to investigate the controlled dissolution of other materials, simpler than steel. Matrices of copper, high copper alloy and NIST C1100 brass were investigated but the matrix that showed the best agreement between measured and certified values was NIST 612 glass. Further studies were conducted to limit the amount of surface layers removed for the NIST 612 matrix.