Evaluation of classification methods for identifying multiwalled carbon nanotubes collected on mixed cellulose ester filter media.

Enhanced darkfield microscopy (EDFM) and hyperspectral imaging (HSI) are being evaluated as a potential rapid screening modality to reduce the time-to-knowledge for direct visualisation and analysis of filter media used to sample nanoparticulate from work environments, as compared to the current analytical gold standard of transmission electron microscopy (TEM). Here, we compare accuracy, specificity, and sensitivity of several hyperspectral classification models and data preprocessing techniques to determine how to most effectively identify multiwalled carbon nanotubes (MWCNTs) in hyperspectral images. Several classification schemes were identified that are capable of classifying pixels as MWCNT(+) or MWCNT(-) in hyperspectral images with specificity and sensitivity over 99% on the test dataset. Functional principal component analysis (FPCA) was identified as an appropriate data preprocessing technique, testing optimally when coupled with a quadratic discriminant analysis (QDA) model with forward stepwise variable selection and with a support vector machines (SVM) model. The success of these methods suggests that EDFM-HSI may be reliably employed to assess filter media exposed to MWCNTs. Future work will evaluate the ability of EDFM-HSI to quantify MWCNTs collected on filter media using this classification algorithm framework using the best-performing model identified here - quadratic discriminant analysis with forward stepwise selection on functional principal component data - on an expanded sample set.

Occupational exposure to airborne nanomaterials: An assessment of worker exposure to aerosolized metal oxide nanoparticles in a semiconductor fab and subfab.

This occupational exposure assessment study characterized potential inhalation exposures of workers to engineered nanomaterials associated with chemical mechanical planarization wafer polishing processes in a semiconductor research and development facility. Air sampling methodology was designed to capture airborne metal oxide nanoparticles for characterization. The research team obtained air samples in the fab and subfab areas using a combination of filter-based capture methods to determine particle morphology and elemental composition and real-time direct-reading instruments to determine airborne particle counts. Filter-based samples were analyzed by electron microscopy and energy-dispersive x-ray spectroscopy while real-time particle counting data underwent statistical analysis. Sampling was conducted during worker tasks associated with preventive maintenance and quality control that were identified as having medium to high potential for inhalation exposure based on qualitative assessments. For each sampling event, data was collected for comparison between the background, task area, and personal breathing zone. Sampling conducted over nine months included five discrete sampling series events in coordination with on-site employees under real working conditions. The number of filter-based samples captured was: eight from worker personal breathing zones; seven from task areas; and five from backgrounds. A complementary suite of direct-reading instruments collected data for seven sample collection periods in the task area and six in the background. Engineered nanomaterials of interest (Si, Al, Ce) were identified in filter-based samples from all areas of collection, existing as agglomerates (>500 nm) and nanoparticles (100-500 nm). Particle counts showed an increase in number concentration above background during a subset of the job tasks, but particle counts in the task areas were otherwise not significantly higher than background. Additional data is needed to support further statistical analysis and determine trends; however, this initial investigation suggests that nanoparticles used or generated by the wafer polishing process become aerosolized and may be accessible for inhalation exposures by workers performing tasks in the subfab and fab. Additional research is needed to further quantify the degree of exposure and link these findings to related hazard research.

Occupational Exposure to Airborne Nanomaterials: An Assessment of Worker Exposure to Aerosolized Metal Oxide Nanoparticles in Semiconductor Wastewater Treatment.

This study characterized potential inhalation exposures of workers to nanometal oxides associated with industrial wastewater treatment processes in a semiconductor research and development facility. Exposure assessment methodology was designed to capture aerosolized engineered nanomaterials associated with the chemical mechanical planarization wafer polishing process that were accessible for worker contact via inhalation in the on-site wastewater treatment facility. The research team conducted air sampling using a combination of filter-based capture methods for particle identification and characterization and real-time direct-reading instruments for semi-quantitation of particle number concentration. Filter-based samples were analyzed using electron microscopy and energy-dispersive x-ray spectroscopy while real-time particle counting data underwent statistical analysis. Sampling conducted over 14 months included 5 discrete sampling series events for 7 job tasks in coordination with on-site employees. The number of filter-based samples captured for analysis by electron microscopy was: 5 from personal breathing zone, 4 from task areas, and 3 from the background. Direct-reading instruments collected data for 5 sample collection periods in the task area and the background, and 2 extended background collection periods. Engineered nanomaterials of interest (Si, Al, Ce) were identified by electron microscopy in filter-based samples from all areas of collection, existing as agglomerates (>500 nm) and nanoparticles (100 nm-500 nm). Particle counts showed an increase in number concentration during and after selected tasks above background. While additional data is needed to support further statistical analysis and determine trends, this initial investigation suggests that nanoparticles used or generated by chemical mechanical planarization become aerosolized and may be accessible for inhalation exposures by workers in wastewater treatment facilities. Additional research is needed to further quantify the level of exposure and determine the potential human health impacts.

Evaluating the Effectiveness of Ozone Management Efforts in the Presence of Meteorological Variability.

It is difficult to assess the effectiveness of regulatory programs in improving ozone air quality in the presence of meteorological fluctuations. In this paper, techniques are presented that improve upon previous methods for moderating the effects of meteorology on ozone concentrations. This approach entails the use of the relations between ozone and meteorological variables to construct meteorologically adjusted ozone time series. To this end, the effectiveness and usefulness of various methods for separating time series of ozone and meteorological data into long-term (climate- and policy-related), seasonal (solar-induced), and short-term (weather-related) components are examined. Correlations between baseline components (sum of long-term and seasonal variations) of ozone and meteorological variables are then investigated independently of correlations between short-term components (weather effects) of ozone and meteorological variables. This allows us to account for the effects of the dominant meteorological variables on each time scale embedded in time series of ozone data. Ozone time series that are devoid of seasonal and climatic variations as well as weather-related fluctuations can then be constructed to detect and track changes in ozone due to the emission control policies implemented. The results of this study reveal that the combination of solar radiation and specific humidity performs best in filtering the seasonal and climatic variations from the baseline component of the ozone data. The combination of temperature and dew point depression performs best in moderating the weather-related effects on the short-term component of ozone data. This method is able to explain about 65% of the variance in ozone data through meteorological variables at several locations examined here.

Moderating the Influence of Meteorological Conditions on Ambient Ozone Concentrations.

Because ambient ozone concentrations are so strongly influenced by stochastic and seasonal variations, it is difficult to assess the effectiveness of regulatory controls in improving ambient ozone air quality. The purpose of this paper is to present a method for moderating the influence of meteorological fluctuations on ambient ozone levels. Techniques presented here account for temperature and other meteorological variables that affect ambient ozone concentrations. To this end, we have examined the correlation between several meteorological variables and ozone concentrations. In addition, we have evaluated trends in ozone time series after removing the effects of these variables on ozone concentrations. The results indicate that inclusion of two meteorological variables strengthens the relationship between ozone and meteorological effects. Moreover, the meteorologically-independent ozone time series at one of the locations studied had a significant trend that was not detected in temperature-independent ozone concentrations.