Determining aerosol metals routinely requires an assessment of differences between test sample aerosol results and collection blanks. When using pad collection principles, the background levels associated with metals from the quartz pad can make the interpretation of test sample results difficult. The objectives of this study were to (1) quantify the within-package, among-package and among-lot variability for levels of 20 different metals from untreated quartz pads and (2) utilize that variability data to assess potential limitations in the determination of differences between metals measured in the particulate trapped on quartz pads from aerosols and collection (air) blanks.

Quartz pads were tested by removing the first five pads (closest to the packing material) from each package, prior to use. Each pad was extracted with 35 mL of 10 % (v/v) ultrahigh purity nitric acid solution, and analyzed directly for 20 metals using ICP-MS. Statistical analysis was performed on the analytical results to assess the within-package, among-package and among-lot variability for each metal.

Of the 20 metals tested, the largest proportion of total variance was attributed to the lot (65-70 %) for two metals (Zn, Pb), and the package (55-80 %) for nine metals (including Al, Cr, As, Mo). For the remaining nine metals (of which only Fe was quantifiable), the majority of the variance could not be explained by lot or package of the quartz pad. The pooled residual variance obtained (in units of ng/collection), after accounting for among-lot and among-package differences, was highest for Al and Mo (110, 58) and lowest for Sr and Pb (2.0, 1.1) for metals that were quantifiable on the quartz pads.

Both lot and package within lot are significant factors in the levels of quantifiable metals found on shipments of quartz pads. Thus, the pooled residual variances from the statistical analysis of the untreated quartz pads can be used to estimate the levels of “background” metals that exist on pads within any given package. This, in combination with instrument sensitivity (limit of quantification), help define the limitations of the overall test method.