Analysis of palladium by high resolution ICP-MS

Abstract

As a matter of fact the impact of increasing palladium emissions is a heavily discussed issue in environmental science and medicine. The major sources for environmental Pd are PGE-based car catalyst technology (59%), dental applications (10%), and electronics (25%). The introduction of Euro Stage III legislation from January 2000 may trigger a further distribution of Pd, since the Pd-rich catalysts can best meet the imposed strict emission specifications. Although assessment and evaluation of Pd toxicology is not within the area of this chapter, it is of vital importance to stress the need for continuous monitoring of PGE exposure in terms of environmental and human bio-monitoring. Through emission of PGE containing particulate matter (PM), palladium contaminates all important environmental compartments. Recent studies showed that Pd is the most mobile and hence bio-available of all platinum group elements (PGE) Pd>Pt>Rh (Ek et al, 2004; Jarvis et al, 2001). Environmental Pd contamination finally results in bioaccumulation of this element in the living organisms through diverse pathways. So far diverse studies could associate PGEs with several health problems in humans, e.g. asthma, nausea, increased hair loss, increased spontaneous abortion, dermatitis and others (Ravindra et al, 2004) Generally, concentration levels of palladium in biological samples are two orders of magnitude lower than in most geological samples, i.e. below microgram per gram levels. Even in contaminated environmental samples, the concentration levels do not exceed the ng g-1 level. The complexity of the matrix and the low concentrations make a direct measurement of the metal extremely challenging. Basically, palladium can be detected with any atomic spectroscopic method; however some of these methods can only deliver their full detection power when analysing samples with easy matrices (Schuster et al, 1999). Among atomic spectroscopic methods inductively coupled plasma mass spectrometry (ICP-MS) is unrivalled in terms of detection power (Bencs et al, 2003). The key features are excellent sensitivity, allowing detection limits at a sub-ng L-1 level in most environmental matrices, low sample consumption and high precision. However, determination of palladium by ICP-MS is not straightforward as spectral interferences impede accurate measurement. Indeed, trace analysis of palladium is still a topical research theme in the ICP-MS community. In these specific applications, i.e. palladium analysis in environmental and biological studies, it is beneficial if not necessary to employ high resolution ICP-MS (HR-ICP-MS) rather than quadrupole ICP-MS (Q-ICP-MS), because of the enhanced sensitivity, resulting in improved detection limits (factor 50-100). With no argument, HR-ICP-MS excels Q-ICP-MS concerning elimination of spectral interferences. However, direct determination of Pd is still problematic as even the application of the high resolution setting (m/δm > 9000) does not allow to separate all occurring interferences. Table 2.1.1 gives an overview of spectral interferences on Pd isotopes and the for separation necessary mass resolution. Several studies accentuate the extent of spectral interferences and recommend routes either for their elimination and/or mathematical correction. Rauch et al have critically investigated the spectral interferences in road dust and river sediment samples on PGEs, using HR-ICP-MS in combination with ultrasonic nebulization (Rauch et al, 2000). Excellent limits of detection (5.3 ng L-1) could be obtained exploiting the low-resolution setting of the instrument. Especially 105Pd and 106Pd were reported to be strongly interfered and prominent interferences could not be removed by increasing the resolution. The authors suggested mathematical correction as possible analytical strategy. A methodological study on direct ICP-MS measurement by our group showed that spectral interferences could be significantly reduced by application of membrane desolvation in combination with HR-ICP-MS (Köllensperger et al, 2000). Although oxide formation was reduced and mass resolution m/δm exceeded 8000, the determination of Pd in road dust was not possible without mathematical correction for Sr. Also other groups report the necessity of mathematical correction procedures, which have to be specifically designed for each sample matrix, even if the sample is less complex than road dust (i.e. urine, remote snow and ice samples) (Barbante et al, 2001; Krachler et al, 1998; Caroli et al, 2001).