Field portable X-Ray fluorescence for rapid analysis Of titanium dioxide on air filters

• Main Author: Hill, Barry Keith
• Other Authors: Peters, Thomas M.
• Format: Others
• Language: English
• Published: University of Iowa 2012
• Subjects: Occupational Health and Industrial Hygiene
• Online Access: https://ir.uiowa.edu/etd/3312; https://ir.uiowa.edu/cgi/viewcontent.cgi?article=3370&context=etd

Field portable x-ray fluorescence (FPXRF) was evaluated as a method for screening titanium dioxide (TiO2) on air filters. Mixed cellulose ester (MCE) and polycarbonate (PC) filter types were compared to gravimetric filters to determine if there was a benefit of using one over the other during FPXRF analysis. No significant difference (p-value = 0.92) was found between MCE and PC filters for FPXRF. MCE filters had a higher coefficient of determination (R2 = 0.97) with the FPXRF analyzer than the PC filters (R2 = 0.70) when compared to gravimetric filter results. The limit of detection (LOD) and limit of quantitation (LOQ) of the FPXRF analyzer were determined through the analysis of blank filters and filters with low levels of TiO2 analyte. The LOD for TiO2 using filter blanks was 7.3 μg/filter, and 25 μg/filter with low levels of TiO2 analyte. The LOQ was 12.8 μg/filter determined with filter blanks, and from the low level analyte samples was 0.82 μg/filter. Filter samples were collected at varying TiO2 concentrations and submitted to FPXRF and inductively couple plasma - mass spectrometry (ICP-MS) analysis. Linear regression was used to determine the relationship between the two methods for TiO2 assessment. A high coefficient of determination (R2 = 0.90) was found between FPXRF and ICP-MS at lower TiO2 concentrations while a low coefficient of determination (R2 = 0.24) was shown for the high TiO2 concentrations. Statistical analysis was used to determine the overall accuracy of the FPXRF method. The FPXRF method did not meet the NIOSH accuracy requirements to be considered an acceptable method; however there were unexplained anomalies within the ICP-MS data.