Neither a total particle or respirable particle inlet sampler is appropriate for sampling the inhalable fraction of continuous filament glass.
The discussion of scanning electron microscopy (SEM) and transmission electron microscopy (TEM) in the section on gravimetric analysis appears to be misplaced, in that electron microscopy is more applicable to fiber counting than to mass measurement. A better example of an alternative means of determining sample mass would be x-ray fluorescence analysis based on the metal content of the MVF.
The introductory paragraphs on fiber counting are appropriate, but the purpose and intent of the remainder of the section are less clear. The cautionary notes about the limited resolution of phase-contrast optical microscopy (PCOM) and SEM for fine fibers (diameter less than 0.3 µm) and superfine fibers (diameter less than 0.1 µm) are appropriate. However, although guidance is given on how to describe the concentration of fibers resolvable with SEM and TEM, there is no guidance provided for determining fiber concentrations with PCOM. Should NIOSH B rules apply to SEM and TEM measurements? If not, why not? If so, should the analysis list separately the fibers seen with SEM or TEM that have diameters above and below about 0.3 µm? If that were done, the concentration of fibers with diameters greater than 0.3 µm might be comparable with the concentration obtained with PCOM, whereas fibers smaller than 0.3 µm might indicate the limitations of PCOM analysis for the particular exposure and risk assessment.
It might also be appropriate to note that using the resolving power of an electron microscope to count the number concentration of fibers less than 5 µm in length is inappropriate because the hazard is largely confined to fibers longer than 5 µm. While the 5-µm length limit has long been used and is incorporated in the NIOSH and ACGIH Threshold Limit Value (TLV) protocols, there is evidence that a better length limit may be 10 or 20 µm (HEI 1991; Lippmann 1994). In any case, an important need when scanning electron microscopic images of sampled fibers is to measure the length distribution of at least several hundred fibers longer than 5 µm. Furthermore, microscopic fields to be counted should be separated by 10 fields to ensure that the fiber distribution is valid (Miller 1999).