The research presented in this doctoral dissertation aimed to improve knowledge on methods to evaluate exposures to carbon-containing nanomaterials and to develop optimized respiratory filters properties to protect workers from these exposures while minimizing discomfort due to breathing resistance. In the initial study, a novel laboratory-based system generated aerosols of four carbon-containing powders (carbon black, a small-diameter (< 8 nm) multi-walled carbon nanotube (MWCNT), a large-diameter (50-80 nm) MWCNT, and a nickel-coated MWCNT) to evaluate the effectiveness of NIOSH Method 5040 for measuring masses as low as 1 μg. A targeted mass of a powder ranging from 1 to 30 μg was deposited on filters for gravimetric and elemental carbon (EC) analysis. The gravimetric mass was compared to the EC mass, and a regression model developed for each powder type. Additionally, the limit of detection (LOD) of the NIOSH Method 5040 for each powder type was determined. The regression models had significant slopes relative to zero for all powder types with all but carbon black demonstrating a statistical difference between the two methods. The LOD of NIOSH Method 5040 ranged from 4.5 for small-diameter MWCNTs to 31.8 μg for nickel-coated MWCNTs. Assuming a sample flow rate of 4.2 L/min and an 8-hour sample duration, the concentration-based LOD for NIOSH Method 5040 ranged from 2.2 μg/m3 for small-diameter MWCNTs to 15.8 μg/m3 for nickel-coated MWCNTs. These results indicate the analysis of EC is affected by the structure and elemental content of the CNTs. Additionally, based on the LOD determined for each powder type, the method may not be sufficient to assess exposures at and below the recommended exposure limit accurately without sampling durations longer than 8 hours. A second study used a laboratory-based system to evaluate an aethalometer response to carbon-containing nanomaterials including carbon black and MWCNTs. Concentrations ranging from 1 to 20 μg/m3 were generated to evaluate the device at concentrations expected in occupational settings. The concentration of the aerosol was measured by an aethalometer alongside a sample collected for EC analysis using NIOSH Method 5040. Additionally, NIOSH Method 7300 was used to determine the concentration of nickel during trials with a nickel-coated MWCNT to determine if the method along with the aethalometer can be used to assess metal-coated MWCNTs. A regression model was developed for each powder type, and the slopes for each were significant relative to zero. The LOD of the aethalometer ranged from 0.56 μg/m3 for nickel-coated MWCNTs to 7.2 μg/m3 for small-diameter MWCNTs. These results indicate the response of the aethalometer may be affected by particle structure and elemental content. NIOSH Method 5040 performed better than the aethalometer for all powder types except the nickel-coated MWCNT. Additionally, based on the LOD determined for each powder, an aethalometer may not be able to assess low-level exposures. In the third study, a mathematical model was used to predict the particle penetration and pressure drop of respirator filters with varying filter thickness, fiber diameter, solidity, and electrostatic charge. Particle penetration was determined experimentally for two different commercially available respirator filters against a sodium chloride aerosol using a scanning mobility particle sizer (SMPS). Optimized filter designs were developed using the model to minimize the pressure drop by adjusting the filter depth, fiber diameter, and solidity of the filter. The model and experimental data were used to maintain a 5% maximum penetration against nanoparticle exposures while minimizing breathing resistance. Model results indicated electrostatic charging played a significant role in improving collection efficiency of respirator filters while not increasing the breathing resistance of the filter. Filter thickness and solidity also played a key role in minimizing breathing resistance. Pressure drop decreased with decreasing solidity, however, the filter depth increased to maintain the collection efficiency of the filter. This increase in filter depth introduced a decision point of determining the practical implications of increased filter thickness on the end user. Filter depth increases dramatically as the solidity decreases below 0.20. The breathing resistance that corresponds to this design is heavily dependent on the face velocity and electrostatic charge of the filter. The electrostatic charge should be maximized during filter production as this was the dominant collection method for nanoparticle aerosols.

Objectives: 1 To correlate carbon nanotube (CNT) concentrations measured by Method 5040 relative to particle count concentrations; 2 to correlate CNT concentrations measured by Method 5040 relative to black carbon concentrations measured with an aethalometer; 3 to compare elemental carbon (EC) concentrations measured by Method 5040 among various CNT types and purities. Methods: CNT samples were collected using 25 mm quartz fiber filters and analyzed for EC by Method 5040. An aethalometer was simultaneously used to measure black carbon concentrations. Samples sent for EC analysis included various CNT types (multi-walled, single-walled) and purities (high, low). Levels of EC concentration were subjected to a two-way analysis of variance having two levels of CNT type and two levels of purity. Results: No correlation was established between CNT count and EC concentration, but a correlation was found between CNT volumetric and total carbon (TC) concentration. A significant correlation between black carbon and TC concentration was found. Method 5040 was found to have a positive bias for TC, and the aethalometer was found to have a positive bias for black carbon. Lastly, this study found that CNT type had no effect on EC concentration, but purity did have a significant effect on EC concentration. Conclusions: Samples analyzed by Method 5040 were found to have 6 - 19% EC content, and thus surprisingly high amounts of organic carbon. It is reasoned that significant amounts of impurities were introduced to CNT samples while travelling through the experimental apparatus. When TC concentrations were plotted against black carbon concentrations a significant relationship was found and the bias of Method 5040 and the aethalometer cancelled out. Future research is needed to investigate the aethalometer as a surrogate for Method 5040. Until then, those conducting CNT exposure assessments should use a 25 mm cassette and increase the volume sampled to achieve a reporting limit lower than the NIOSH recommended CNT REL of 7 µg/m3.