Evaluation of nitric acid by thermodynamic model-a case study at supersites in southern Taiwan

• Main Authors: Yi-Chan Chiang, 江宜展
• Other Authors: Yee-Lin Wu
• Format: Others
• Language: zh-TW
• Published: 2007
• Online Access: http://ndltd.ncl.edu.tw/handle/86499558313218739816

碩士 === 國立成功大學 === 環境工程學系碩博士班 === 95 === The purposes of this study were to establish a simplified thermodynamic model of NH4NO3 and evaluate whether the model was feasible to predict the real-time concentration of HNO3 in the southern supersites in Taiwan. The calculations were based on the studies by Stelson and Seinfeld (1982), used to compute the equilibrium constant of NH4NO3 (Kp). And HNO3 can be calculated by Kp and NH3 which was measured in the atmosphere. The parameters of this model include temperature (T), relative humidity (RH) and the impact of (NH4)2SO4 in the mixed sulfate/nitrate solution. Moreover, Annular Denuder Sampler (ADS) and Honeycomb Denuder system (HDS) were adopted to verify this model. The 4 times of sampling were in fall (September 2005), winter (December 2005), spring (March 2006) and summer (July 2006), respectively. The 4 sampling sites were Fooyin, Chiautou, Chenjen and Choujau all in the southern supersites. The sample numbers of ADS were 167 and which of HDS were 183. The result used by ADS showed that 39.25% samples could be predicted by this model with 2 parameters, T and RH. The results in winter and spring had the better fits. 50% and 60% samples in winter and spring could be predicted by this model, respectively. The result used by HDS showed that 36.06% samples could be predicted by this model . The results in winter and spring had the better fits which were the same as ADS. 47.2% and 57.4% samples could be predicted in winter and spring, respectively. In addition, because higher T and lower RH in summer and fall compared with winter and spring resulted in fewer formation of NH4NO3, the model couldn’t be used to compute Kp (NH4NO3). In winter and spring, there might be some factors we didn’t consider, so some samples couldn’t be predicted by this model. Furthermore, after (NH4)2SO4 was considered in the model with T and RH , the correlation coefficient between the predicted value and measured values used by ADS and HDS were almost the same as the result of model with just T and RH. The reason might be that (NH4)HSO4 and (NH4)2SO4 may exist in the atmospheric environment at the same time, so the mole ratio of NH4+ to SO42- won’t be equal to 2. As a result, the impact of (NH4)HSO4 couldn’t be ignored. The better fits with the simplified model applied in the 4 southern supersites were in winter and spring. The value-scale of HNO3 was smaller than 10ppb, and it was higher in daytime and lower in nighttime. The finding was identical with previous studies.