Page:NIOSH Manual of Analytical Methods - Chapter L.pdf/11

 During PLM analysis, fibers are immersed in a fluid selected to have a known refractive index. When a fiber has a larger refractive index than the surrounding fluid medium, the bright halo (Becke line) around that fiber appears to move into it as the microscope focus is raised; when the fiber has a smaller refractive index, the Becke line move out of it. Placing the fibrous material into several different refractive index fluids allows the fiber refractive index to be bracketed. Dispersion, or refractive index change with wavelength, of a fiber can be used for identification. When particles are placed in a liquid whose dispersion is different from that of the particle, the particle may exhibit a color caused by the refraction of light. This technique requires the use of special "dispersion staining" optics. By using several refractive index liquids in series, the refractive index and the dispersion of the fiber can be established and compared with those of standard materials or published data [12]. Once the sample has been uniformly dispersed on a slide in the appropriate refractive index liquid, specific fiber types, e.g., asbestos, can be identified and the percent fibers estimated. Two approaches are typically used: visual comparison with prepared reference slides or pictures and point counting. When attempting to estimate whether a material is ACM (i.e., > 1% asbestos), the visual comparison technique is adequate when more than about 10% of the particles observed are asbestos. Point counting is used for lower concentration samples to provide higher accuracy [60]. It involves observing 400 or more randomly selected "points" (identified with a reticle crosshair) in the sample. The number of points containing asbestos is divided by the total number of points observed to give the percent asbestos. A combination of these approaches balances the analysis time and accuracy of the results [61]. PLM also can be used for qualitative analysis of air sample filters by collapsing the filter and using low temperature plasma etching of the surface to expose the fibers. Various refractive index liquids can then be placed on the etched surface to surround the fibers, allowing techniques noted above to be used [53]. The smallest fibers that can be identified by this method are about 1-:m diameter. c. Accuracy PLM analysis is primarily used for qualitative identification of fiber type. Accurate identification of asbestos and other fibers requires proper training in the crystallographic properties of particles as well as training and familiarization with the PLM. As with fiber counting, a laboratory quality assurance program is necessary to ensure consistently accurate results. The National Voluntary Laboratory Accreditation Program (NVLAP) operated by the National Institute for Standards and Technology (NIST) inspects laboratories for proper practice as well as providing unknown samples four times a year to check their performance in fiber identification. Under a predecessor to this program, approximately 350 laboratories correctly classified 98.5% of the samples as asbestos and correctly identified the specific asbestos types in approximately 97% of the samples. A blind test of 51 laboratories resulted in 97.5% correct classifications and 79.1% correct identifications [62]. The American Industrial Hygiene Association Proficiency Analytical Testing Program provides similar PLM audit samples to laboratories. Some common interferences for bulk analysis by PLM include sepiolite, vermiculite, and cleavage fragments of non-asbestos amphiboles.

3/15/03

153

NIOSH Manual of Analytical Methods