| 英文摘要 |
The purpose of this project is to evaluate and develop analysis methods for carbon components in fine particulate matter (PM2.5). Currently, Taiwan has not established standardized methods for analyzing carbon components in PM2.5, leading to potential discrepancies in analytical approaches across different projects, which makes data comparison challenging. Therefore, this project compares the most commonly used international methods for PM2.5 carbon component analysis and evaluates methods suitable for Taiwan's environmental context to establish a standard analytical method for carbon components in PM2.5 in the future.
The project tasks include literature review, establishing at least two analytical methods for PM2.5 carbon components, sampling and analyzing elemental carbon (EC) and organic carbon (OC) concentrations in industrial and metropolitan areas, and proposing recommendations for PM2.5 carbon component analysis methods. To date, the project has compiled 25 relevant studies, established three analytical methods for PM2.5 carbon components (IMPROVE_A, NIOSH 5040, and EUSAAR_2), and completed sampling and analysis in industrial areas (Fenglin Junior High School, Kaohsiung City) and metropolitan areas (Fuxing Elementary School, Kaohsiung City, and Linsen Bus Station, Tainan City) in June and September. Additionally, the uniformity of filter paper distribution was analyzed to confirm the variation ratio for repeated sampling and analysis of the same sample. The detection limits for the established methods IMPROVE_A, NIOSH 5040, and EUSAAR_2 were 0.23, 0.11, and 0.13 μgC/cm² (filter area), respectively. The ratio of standard deviation to mean for seven repeated analyses of the same sample was 0.06 for IMPROVE_A, 0.04 for NIOSH 5040, and 0.03 for EUSAAR_2, indicating acceptable uniformity. Field sampling results for total carbon (TC) using IMPROVE_A and NIOSH 5040 showed high correlation coefficients at Fenglin Station and Fuxing Station, with R² values of 0.9924 and 0.9933, respectively. However, TC concentrations obtained via IMPROVE_A were approximately 12% higher than those from NIOSH 5040, a result consistent with findings from the literature review. The project also conducted technical dissemination on field sampling, carbon analyzer operation, and standard/sample analysis procedures on June 25, June 27, and September 26.
To assess differences among instruments, optical corrections, and temperature protocols, the project used three standard samples of varying concentrations provided by Sunset Laboratory Inc., along with field samples collected in June, September, and various locations in Tainan City. The results were analyzed and discussed as follows: first, comparing analytical integrity across instruments; second, exploring differences among optical correction methods; and finally, evaluating the advantages and disadvantages of different temperature protocols.
The Sunset analyzer required four analyses to obtain values below the detection limit, whereas the DRI analyzer achieved this in just one analysis. Regarding standard sample analysis, the DRI analyzer demonstrated results closer to the reference values compared to the Sunset analyzer, leading to its selection for subsequent analyses. For the three concentration levels of standard samples, TOT results were closer to reference values than TOR. In atmospheric sample analysis, there was little difference in correlation between TOT and TOR optical correction methods. Lastly, among the three temperature protocols (IMPROVE_A, NIOSH 5040, and EUSAAR_2), results from IMPROVE_A and EUSAAR_2 were more consistent, while NIOSH 5040 showed greater variability. Based on DRI analysis spectra, the optical correction method under IMPROVE_A was the most accurate.
|