precision
To ensure good accuracy, it is important to be aware of three sources of error: the supplier of the coating material, background variations and interference. It has been recognized, though not explained, that silicones from different suppliers give different sensitivities. When the different silicones were analyzed on the same basis weight paper, lines with different slopes were obtained . Therefore, to avoid this source of error, each vendor's silicon builds individual calibrations. Second, corrections for background variations help reduce errors.
Figure 2 shows two sets of standards for the same silica gel on Supercalendar kraft paper prepared on different dates. Clearly the samples give the same slope, they cancel out. The analyzer will correct for this background variation by the operator inserting an uncoated sample from the paper to be coated. Therefore, all subsequent analyzes will be corrected for apparent background changes.
A final source of error is the presence of another substance at high concentrations. Examples include silicone coatings on polyvinyl chloride (PVC) or titanium dioxide filled films. As the energy levels in these films change, so does the effect on the background of the silicon region in the spectrum. Consequently, the apparent weight of the silicone coating of the film sample will either increase or decrease. This type of interference is easily corrected by software built into the analyzer. Interference is identified with spectral scanning and automatically compensated for during calibration.
Repeatability and Reproducibility
Repeatability in terms of measurement accuracy is a statistical function that determines the variability in cumulative intensity and x-ray counts from repeated measurements. Since the iron-55 source excites titanium x-rays more efficiently than silicon x-rays, the sensitivity to titanium coatings is higher and therefore the measurements are more accurate. Accuracy may always be improved by increasing coating time, but this does not always result in a substantial increase in accuracy. Typically, for silicone coatings, the reported standard deviation corresponds to #0.01 g of silicone per square meter.
Figure 1 Differences in sensitivity of products from different silicone suppliers
Figure 2 Differences in paper backing
