6. SAMPLE DEPOSITION UNIFORMITY Some analytical methods require that sampled particles be deposited uniformly on the filter surface. For instance, asbestos fiber analysis by microscopy requires uniform deposition of fibers on the filter for accurate results. Direct silica analysis of collected filter samples also is improved with uniform particle deposition. Classifiers using inertial or gravitational forces tend to stratify the aerosol stream. A small, high velocity inlet in a sampler, such as the 4-mm opening in the 37-mm closed-face cassette, can also result in the larger particles being deposited in a small central area on the filter. Even sampling at high flow rates through more open inlets can cause a non-uniform deposit [73]. This results in particle deposits that vary in uniformity as a function of particle size. Such deposition patterns are visible when sampling colored particles [74]. Open-pore foam classifiers may improve the uniformity of particle deposits on the filter, but have not been thoroughly evaluated [75, 76]. Careful design of classifiers to ensure mixing of the aerosol prior to deposit on the filter may result in adequate uniformity [47]. Even inhalable samplers or samplers that have no classifier may be prone to non-uniform deposits under certain conditions of sampler orientation relative to gravitational settling, orientation relative to external winds, or when sampling charged particles [61, 62, 64, 70]. Flaring the inlet of such a sampler, as in the commercial “bell-mouth cowl,” (Figure 2f, Envirometrics, Charleston SC), is one approach to improving sample uniformity under anisokinetic conditions [73]. In another study, a sampler having an inlet screen (button sampler, Figure 2d, SKC, Inc. Eighty Four PA) exhibited improved filter deposit uniformity when compared to a closed-face cassette [77]. The filters in some samplers require support to prevent tearing or distortion of the filter. The support device may cause occlusion of parts of the filter surface, resulting in non-uniform particle deposits [78]. On occasion, it was observed that poorly-sized tubing connectors protruded into the 37-mm cassette and touched the filter surface. This caused all the airflow to pass through the filter adjacent to the small area of the connector opening. When undetected, this caused low sampling efficiency and pump failure because of the high pressure drop. 7. SAMPLER WALL LOSSES Particle deposits on internal surfaces (i.e., wall losses) of the 37-mm closed-face cassette for several hundred field measurements were found to be large and highly variable (2 - 100% of dust collected in the cassette) [79]. Another study found only 22% of the dust on the filter, 65% on the upstream portion of the cassette, and 22% downstream of the filter [58]. In a study of an in-line cassette of similar shape, it was found that the internal wall deposition of particles could be largely eliminated by: (a) making the cassette conductive, (b) creating an aerodynamically smooth surface having no corners for eddies to form, and (c) decreasing the diameter of the filtration area so that dust does not deposit on the filter adjacent to the upstream walls of the cassette [72]. By incorporating these three corrective measures, the wall losses in the latter cassette were reduced from 25-30% to 5%. These losses appear to be caused by a combination of electrostatic, inertial, gravitational and diffusion mechanisms. Another solution to the problem of not capturing 100% of the sampled particles on the filter is to use an internal cartridge sealed to the filter. All the particles collected in the combined
3/15/03
192
NIOSH Manual of Analytical Methods
