Multianalytical determination of trace elements in atmospheric biomonitors by k0-INAA, ICP-MS and AAS

Elemental contents of atmospheric biomonitors—epiphytic lichens and tree bark, exposed in continuous and discontinuous modes—have been assessed through k0-standardised instrumental neutron activation analysis (k0-INAA) (two different institutions), inductively coupled plasma mass spectrometry (ICP-MS) and atomic absorption spectrometry (AAS). Certified reference materials—ISE-921 (river clay), NIST-1547 (peach leaves), ICHTJ-INCT-TL-1 (tea leaves; TL-1 hereinafter) and IAEA-336 (lichen material), and nonparametric statistics—rank-order correlations (Spearman RS) and enhanced-sign tests (Wilcoxon T)—were used for analytical control and data comparison, respectively. In general, quality of procedures was deemed good, except for k0-INAA in determining Br, Cu and Na, all likely affected by high counting statistics, and/or contamination issues (the latter). Results for Cu, Ni, Pb and Sr (by both ICP-MS and AAS) revealed that, despite an outstanding correlation (asymptotic p=0.000p=0.000), they could be viewed as statistically equal for Cu only: AAS tended to yield higher values for Pb and Ni, and lower ones for Sr. The comparison between ICP-MS and k0-INAA data from TUDelft, for Al, Ca, Cu, Mg, Mn, Na, Ti and V, showed an excellent correlation (as above) and random (relative) magnitude for Cu, Mg, Mn and Ti only: ICP-MS tended to yield higher values for Al, Na and V, and lower ones for Ca, whereas between k0-INAA data from TUDelft and ITN, for Br, Ca and Na, resulted in systematically higher [Br] and [Ca] variates from TUDelft, even if all corresponding data sets were found to correlate at stringent significance levels. In a few cases, though—Ca, Sr in lichens; Pb in bark—matrix effects did appear to interfere in the outcome of matched-pairs, signed-rank tests, since random hierarchy of variates could be asserted just when lichen and bark data sets were processed separately.