Anal. Chem. 1980, 52, 1779-1780
1779
Filter Material for Sampling of Ambient Aerosols D. F. Leahy," M. F. Phillips, R. W. Garber, and R. L. Tanner Brookhaven National Laboratory, Associated Universities Inc., Upton, New York 11978
Analyses of aerosols collected on filters have become more numerous and sophisticated. In addition to ammonium and nitrate, investigations in this laboratory have been directed toward sulfate, acidity, and specific sulfuric acid measurements ( I ) . An ideal filter should be strong, inert, efficient, and capable of high linear flow rates as well as having low, consistent blank values. Initial sampling operations for aerosol sulfate were conducted using Whatman GF81 glass fiber filters (2). Efforts to modify this material for use in determining certain trace aerosol constituents were unsuccessful due principally to alkaline content of t h e glass fiber material. Commercially available, Teflon-coated glass fiber material, although improved with respect t o artifact sulfate formation, proved unsuitable for strong acid determinations. Teflon membrane-type filters were inert but not adaptable to high volume air sampling because of filter-limited low flow rates. Promising results were observed with a pretreated tissue quartz filter material produced by Pallflex Products Corporation, Putnam, Conn. T h e quartz material as commercially available is not suitable for use in typical sampling of ambient aerosols. Blank values were unacceptably high and variable (Table I). T h e material was delicate and easily damaged in the routine physical manipulations required. Investigation led to t h e conclusion t h a t a quartz material, i.e., Pallflex QAO 2500, would be most practical for ambient aerosol sampling, provided i t could be suitably modified. T h e first method used ( 3 ) involved heating the quartz material in concentrated phosphoric acid and firing the rinsed fiber to 750 "C in a furnace. Satisfactory results were recorded but the method was tedious and time consuming. The final product was somewhat variable in composition, often brittle and prone to cracking, the latter probably caused by insufficient rinsing of the filters after the phosphoric acid treatment step. T h e substitution of dilute phosphoric or various concentrations of acetic acid in the procedure proved ineffective in removing alkaline contaminants. An improved method has been adopted using a hydrochloric acid soaking step. This procedure is simpler, results in uniform filters (Table I), and requires approximately 3 min per filter in preparation time. No change in overall collection efficiency or homogeneity can be measured when treated filters are compared to the quartz material as received from the manufacturer. EXPERIMENTAL Ten oversized (10%) quartz circles (QAO 2600, Pall Corporation, Putnam, Conn.) are placed in a plastic Buchner funnel (Nalgene 4280) lined with two Whatman 41 circles. The material is washed with slight suction using deionized H20. The top section of the funnel is detached and immersed in a deep Pyrex dish containing hot (75-90 "C) 1.2 h-hydrochloric acid for 15 min. The Buchner section is lifted from the bath, drained, and reconnected to its bottom portion and the circles are thoroughly rinsed with deionized H 2 0 under slight suction. The quartz circles are then ignited to 750 "C for 10 min in a muffle furnace. The cooled material is cut to the final desired size, placed in an ultrasonic bath containing pH 4 phosphoric acid, and sonicated for 30 min. The circles are removed from the bath, placed individually on screens, and air dried in a laminar flow hood with final drying conducted in a vacuum oven at 100 "C. Treated filters are sealed 0003-2700/80/0352-1779$0 1 OO/O
Table I. Blank Value Ranges quartz sample*
SO,*-,
ng/cmz
",
nequivi cmZa
NH,+, ngicm'
NO;,
ng/ cmz
as received 300 to 3600 58 to 62 650 to 800 < 2 g d H,PO,260 to 420 -6 to - 7 c
