Quantitative analysis of trace elements in environmental powders with laser ablation inductively coupled mass spectrometry using non-sample-corresponding reference materials for signal evaluation

• Cross-matrix-calibration: quantification of samples with reference materials not matching in matrix composition.
• Fast and easy in-house pelletization procedure for powdered environmental samples.
• Collision/reaction cell technology to overcome matrix-introduced interferences

Laser ablation inductively coupled plasma-mass spectrometry (LA-ICP-MS) is an attractive alternative to traditional procedures for the analysis of environmental samples (i.e., conventional liquid measurement after sample digestion). However, for accurate quantification, certified reference materials (CRM) are necessary which match the composition of the sample and include all elements of interest at the required concentration levels. The limited availability of appropriate CRMs hampers therefore substantial application. In this work, an LA-ICP-MS procedure allowing for accurate determination of trace element contents in powdered environmental samples is presented. For LA-ICP-MS analysis, the samples are mixed with an internal standard (silver oxide) and a binder (sodium tetra borate) and subsequently pressed to pellets. Quantification is accomplished using a calibration function determined using CRMs with varying matrix composition and analyte content, pre-treated and measured in the same way as the samples. With this approach, matrix-induced ablation differences resulting from varying physical/chemical properties of the individual CRMs could be compensated. Furthermore, ICP-related matrix-effects could be minimized using collision/reaction cell technology. Applicability of the procedure has been demonstrated by assessment of Cd, Cu, Ni, and Zn in four different environmental CRMs (NIST SRM1648a (urban particulate matter), NIST SRM2709 (San Joaquin Soil), BCR144 (sewage sludge), and BCR723 (road dust)). Signal evaluation was performed by alternative use of three CRMs for calculation of the calibration function whereas the remaining fourth CRM acted as unknown sample, resulting in a good agreement between measured and certified values for all elements and reference materials.