The purposes of this study are: (1) to determine whether proficiency analytical test (PAT) materials from the American Industrial Hygiene Association can be used to provide quality data for portable X-ray fluorescence analysis (XRF) of lead in dust wipe surface samples; (2) to provide data to determine whether the on-site analysis of field dust wipe samples by XRF and the laboratory method of inductively coupled plasma emission analysis (ICP) are comparable; and (3) to determine if differences exist between different wipe materials. Several wipes meet the ASTM E1792 performance requirements of lead background level less than 5 microg per wipe, be only one layer thick, yield recovery rates of 80- 120% from spiked samples, remain damp throughout the sampling procedure, and do not contain aloe. The wipes used in this study were Pace Wipes, which are used for the PAT materials, and, for the field samples, Palintest Wipes, which were supplied by the instrument manufacturer, and Ghost Wipes, which are popular because they digest in hot, concentrated acid, so that chemical analysis is simplified. Twenty PAT wipe samples were obtained from four different proficiency test rounds. Surface wipe samples were taken at three different locations representing different industry types. All samples were analyzed using a portable XRF spectrometer and by ICP. Strong linear relationships were found for the analysis of wipe samples by ICP and by portable XRF. For the PAT samples, the results from the ICP and XRF analysis were not statistically equivalent, which indicates a bias in the ICP analysis. The bias was not excessive, since all ICP analyses fell within the acceptable range for the proficiency samples. The good correlation between the proficiency sample reference values and the XRF determinations is not surprising considering similar proficiency samples were used to calibrate the instrument response. Users of this portable XRF analyzer could enroll in the proficiency test program as part of their quality assurance program. For field samples, the relationship was strongest for Palintest wipes, and the values found for all three industries could be combined. However, the results from the ICP and XRF analysis were not statistically equivalent using the correction factor in the calculation algorithm as supplied with the instrument, and a new coefficient was derived. The mean relative error for the XRF analysis versus the ICP analysis was greater than 25%, such that the method falls within the realm of screening procedures. For Ghost Wipe samples, the precision was different for different industries, and the results could not be pooled. Differences between the two wipe materials may be related to the number of folds required for analysis.
Department of Environmental Health Sciences, School of Public Health, University of Alabama at Birmingham, 1530 3rd Ave. S., Birmingham, AL 35294-0022