G. A. Klouda and H. J. Parish (SRI International)

Objective: To produce an air particulate matter (PM) filter Standard Reference Material (SRM 2784) by resuspending a Baltimore Dust (< 2.5 mm aerodynamic diameter) and collecting on quartz-fiber filters for calibration of organic carbon, elemental carbon, and individual compounds if relevance to PM2.5 at air monitoring sites throughout the U.S.

Problem: A "prototype" PM on quartz-fiber filters using SRM 1649 Urban Dust was produced in 1995 in collaboration with SRI. SRIs system was designed to resuspend particles in air and collect the aerosol on up to 320 filters simultaneously. For the 300 prototype-PM filters produced, the particle distribution was non-uniform near the filter edge and the PM loading varied up to 20% across filters. These problems were attributed to an inability to reproduce the seal provided by the filter pack thereby causing air to leak across the filter. Also, the inorganic residue remaining from the combustion of the carbonaceous fraction of the "prototype" visually indicated that SRIs system was not breaking up all the larger agglomerates present in the source material. Therefore, to produce a PM2.5 filter SRM, it was necessary to develop a custom filter pack to assure that only PM2.5 are collected and that the particle distribution is uniform to the filter edge. Also, a contemporary PM2.5 sample on the order of several grams would be needed for the production of SRM 2784.

Approach and Results: Through additional funding from SRMP, new filter packs were designed and constructed by URG Corp. to size-segregate particles by impaction and, thereby, allow only < 2.5 mm-size particles to reach the filters. Each filter pack included a gasket set and stainless-steel screen to provide an excellent seal and support to obtain a uniform distribution of particles throughout the filter. A test run of SRIs system with the new URG filter packs was completed using a fine coal-mine dust collected on quartz, TeflonÒ and NucleaporeÒ (polycarbonate) filters. Visually, the quartz-fiber filters, having a torturous-path, appeared homogeneous throughout. This was verified at the < 3% level through carbon measurements on filter aliquots. However, a visual inspection of the Teflon and Nucleapore membrane-type filters clearly revealed a non-uniform deposit. Tests conducted on a filter pack by URG using a fine-chalk dust showed that the stainless-steel screen had to be repositioned directly behind the filter within the gasket set. A second test run by SRI using SRM 1649a Urban Dust was completed and visual indications were that the quartz and Teflon filters appeared to have a uniform distribution of particles; the Nucleapore filter remains problematic. Carbon analysis of quartz filters will provide the assurance of within-filter homogeneity at an acceptable level and XRF measurements will show how uniform the elemental (inorganic) constituents are across Teflon filters. The variability in the PM loading across-filters remains unacceptably high, as much as 17%, therefore, requiring the weighing of each filter before and after resuspension and filtration. For the production of SRM 2784, ~10 g of PM2.5 was collected in Baltimore during the fall '98 and spring '99 in collaboration with John Ondov at the University of Maryland.

Future Plans: It is anticipated that the production of 2200 PM filters loaded with Baltimore Dust (SRM 2784) will be completed by the first half of FY 2000. Through an informal release of a modest number of "prototype" PM filters, an immediate need for this material has been recognized and we expect to release it in the fall '99 as a Reference Material. Laboratories that have assisted in the analysis of the "protoptype" have also agreed to assist in the certification of SRM 2784. U.S. EPA matching funds are anticipated to help defray some of the cost for production of SRM 2784 as well as support future development of NIST-traceable PM-filter SRMs.

Web Contact micro@nist.gov