A standardized procedure for collecting dust samples from surfaces using a micro-vacuum sampling technique was evaluated. Experiments were carried out to investigate the collection efficiency of the vacuum sampling method described in ASTM Standard D7144, "Standard Practice for Collection of Surface Dust by Micro-Vacuum Sampling for Subsequent Metals Determination." Weighed masses (approximately equal to 5, approximately equal to 10 and approximately equal to 25 mg) of three NIST Standard Reference Materials (SRMs) were spiked onto surfaces of various substrates. The SRMs used were: (1) Powdered Lead-Based Paint; (2) Urban Particulate Matter; and (3) Trace Elements in Indoor Dust. Twelve different substrate materials were chosen to be representative of surfaces commonly encountered in occupational and/or indoor settings: (1) wood, (2) tile, (3) linoleum, (4) vinyl, (5) industrial carpet, (6) plush carpet, (7,8) concrete block (painted and unpainted), (9) car seat material, (10) denim, (11) steel, and (12) glass. Samples of SRMs originally spiked onto these surfaces were collected using the standardized micro-vacuum sampling procedure. Gravimetric analysis of material collected within preweighed Accucap inserts (housed within the samplers) was used to measure SRM recoveries. Recoveries ranged from 21.6% (+/- 10.4%, 95% confidence limit [CL]) for SRM 1579 from industrial carpet to 59.2% (+/- 11.0%, 95% CL) for SRM 1579 from glass. For most SRM/substrate combinations, recoveries ranged from approximately equal to 25% to approximately equal to 50%; variabilities differed appreciably. In general, SRM recoveries were higher from smooth and hard surfaces and lower from rough and porous surfaces. Material captured within collection nozzles attached to the sampler inlets was also weighed. A significant fraction of SRM originally spiked onto substrate surfaces was captured within collection nozzles. Percentages of SRMs captured within collection nozzles ranged from approximately equal to 13% (+/- 4 - +/- 5%, 95% CLs) for SRMs 1579 and 2583 from industrial carpet to approximately equal to 45% (+/- 7 - +/- 26%, 95% CLs) for SRM 1648 from glass, tile and steel. For some substrates, loose material from the substrate itself (i.e., substrate particles and fibers) was sometimes collected along with the SRM, both within Accucaps as well as collection nozzles. Co-collection of substrate material can bias results and contribute to sampling variability. The results of this work have provided performance data on the standardized micro-vacuum sampling procedure.
National Institute for Occupational Safety and Health, Division of Applied Research and Technology, Mail Stop R-7, 4676 Columbia Parkway, Cincinnati, OH 45226-1998