2.1. Aerosol Generator
An aerosol generator (Model SAG 410/U, TOPAS GmbH, Dresden, Germany) was purchased in 2010 as an option for dispersing aerosols from a bulk powder in the dry state. The generator incorporates both a dispersing unit and a dust feeder. The disperser is a venturi aspirator that operates at flow rates between 25 L min
−1 and 67 L min
−1. The dust feeder consists of multiple parts that combine to deliver powder to the suction nozzle of the venturi (
Figure 1). Powder is placed in a conical, rotating hopper. A rotating auger, consisting of wire wrapped loosely around a rod, pulls powder up through a tube. To enhance the auger’s ability to pick up powder, a wire extends straight down the side of the hopper wall to agitate powder as it moves past the wire. Powder that reaches the top of the tube forms a cloud of falling particles within a column. As shown in
Figure 1, the falling particles land on a rotating ring protruding into the column space. The particles settling on the upper edge of the ring are then scraped to an even height prior to exiting the column and carried to the suction nozzle of the venturi ejector to be aerosolized.
Figure 1.
Dust generator schematic diagram (not to scale).
Figure 1.
Dust generator schematic diagram (not to scale).
As supplied by the manufacturer, the powder hopper had a total volume of approximately 350 cm
3 but powder could only be added to half the height to prevent spill-over, and with the auger tube, resulted in an effective volume of powder of 135 cm
3. As shown in
Figure 1, an aluminum insert was fabricated to reduce the effective powder volume to 30 cm
3. This alteration allowed as little as 0.5 g of powder to be added to the hopper and still surround the auger to a depth sufficient to cause it to be consistently pulled up by the auger. No other adjustments were made to the purchased unit. A glass box was supplied by the manufacturer to surround the feeder and ejector. Since the ejector pulls in air along with particles on the ring edge, HEPA filtered air is supplied to the space within the glass box. We also built a polypropylene box large enough to house the entire generator and with a connection to laboratory vacuum air to maintain a negative pressure relative to room air to minimize the possibility of aerosol emission into the laboratory (
Figure 2).
Figure 2.
Aerosol delivery system consisting of the (a) aerosol generator in an acrylic plastic enclosure; (b) charge neutralizer; (c) whole body exposure chamber; (d) aerosol photometer; and (e) scanning mobility particle sizer.
Figure 2.
Aerosol delivery system consisting of the (a) aerosol generator in an acrylic plastic enclosure; (b) charge neutralizer; (c) whole body exposure chamber; (d) aerosol photometer; and (e) scanning mobility particle sizer.
2.2. Generator Characterization Trials
To characterize the performance of the generator for aerosolizing a CNT aerosol, double walled carbon nanotubes (DWCNTs) were purchased (Stock No. 1290NMG, Nanostructured & Amorphous Materials, Houston, TX, USA). The DWCNTs consisted of >90% carbon nanotubes of all types (single-walled and multiwalled) and >50% double-walled carbon nanotubes. As reported by the manufacturer, the outer diameter of these DWCNTs was < 5 nm, the inner diameter ranged from 1.3 to 2.0 nm, and with a length 5–15 μm and surface area ~400 m2/g. Additional characterization included the use of transmission electron microscopy (TEM) to size outer tube diameters, which ranged between 1 nm and 6 nm with an average of 3.3 nm. BET surface area analysis was also performed which resulted in a higher value of 575 m2/g than that reported by the manufacturer. The bulk powder was also analyzed for metals that can often appear as contaminants in DWCNTs. The University of Iowa State Hygienic Laboratory performed a qualitative screen for 21 metals following a modified EPA method 3050B (200.8), using inductively coupled plasma-mass spectroscopy (ICP-MS). No detectable level (>2 µg/g) of any metal was obtained.
The primary variables that affect the generation rate include the width of the top edge of the rotating ring; the rotation speed of the ring, and the height of powder on the ring. Two rings are provided: one with a top edge of 0.3 mm and the other with a top edge of 1 mm. The bulk characteristics of the powder may also influence the generation rate, but this study was restricted to the use of the generator for aerosolizing one type of carbon nanotubes. Factors that may not affect the generation rate but could influence the consistent dispersion of powder include the rotation speed of the auger wire, the depth of the powder in the hopper, and the air pressure to the venturi ejector.
The DWCNT bulk powder was placed in a drying oven at 100 °C overnight prior to its use in the aerosol generator. To test the effect of changes in the operational variables of the generator, the generated aerosol flowed into a small, 65 L, aluminum chamber (
Figure 2) used for mouse whole-body exposures [
14]. Aerosol exiting the generator first passed through a charge neutralizer (Model 3012, TSI Inc., Shoreview, MN, USA) before entering the chamber to follow our standard procedure for a mouse exposure. A vacuum pump on the exhaust side of the chamber was used to pull air at a slightly higher rate than provided by the generator to achieve a total flow of 32 L min
−1 and a negative static pressure of −10 mm H
2O.
The size distribution of the DWCNT aerosol was measured primarily with a scanning mobility particle sizer (SMPS, TSI Inc., Shoreview, MN, USA) set to size and count particles in 105 size channels between 7 nm and 289 nm. To verify that the entire particle size distribution was measured, a portable optical particle counter (Model 1.108, Grimm Technologies, Inc., Douglasville, GA, USA) was also used to provide counts in 15 size channels between 300 nm and 20,000 nm. Inlet hoses to these instruments were attached to a Y-connector that terminated at a port placed immediately above the exhaust-air port of the chamber. DWCNT mass concentration was determined from gravimetric analysis of filter samples taken by placing a 37-mm filter in a filter holder placed in-line with the chamber exhaust air. Filters were pre- and post-weighed in a dedicated, climate-controlled room housing a six-place microbalance (Model XP26, Mettler-Toledo, Inc., Columbus, OH, USA).
Generator characterization trials consisted of measuring changes in chamber aerosol concentration after step-changes in ring speed. Those measurements were made with an aerosol photometer (Model pDR-1200 RAM, Thermo-Electron Corp., Waltham, MA, USA) set to take readings every 1 s. The relationship between photometer readings and filter-based concentrations was also analyzed to determine whether that instrument could be employed as the sensor in an aerosol concentration feedback control system utilizing signals to the generator to vary ring speed. Measurements to establish the consistency of aerosol concentration over hour-long periods were also performed as part of a study to characterize the sub-acute toxicity of DWCNT using a murine model.