運用開徑式傅立葉轉換紅外光光譜儀測量懸浮微粒

Abstract

目的: 研究利用開徑式傅立葉轉換紅外光光譜儀(OP-FTIR)的測量結果回推氣膠粒徑分佈之可行性。 方法:首先由文獻中獲得有關水、硝酸銨和硫酸銨的複折射率(complex refractive index)資料，並依據Mie理論針對這些物質不同的粒徑分佈(幾何平均數=2-10微米，幾何標準差=1.1-2.5)模擬出消光光譜。然後發展最佳化演算法用以在已知複折射率的情況下回推幾何平均數和幾何標準差。也在產生模擬光譜時增加1%、4%、7%和10%的雜訊作一敏感度分析。更進一步收集在一控制環境下，經由開徑式傅立葉轉換紅外光光譜儀(OP-FTIR)量測所得的水微粒消光光譜，用其來驗證電腦模擬研究的結果。 結果: 對所有物質來說，在幾何標準差相同的情況下，較大的微粒具有較好的回推結果。而敏感度分析的結果顯示當雜訊程度小於4%且幾何平均數大(>3.5微米)的時候，模擬與回推光譜的符合情況良好，並且其光譜間R2可高於0.9。同時對於輸入和回推的幾何平均數與幾何標準差而言，彼此間的差異小於10%。在雜訊程度等於10%時，集中粒徑分佈的回推結果會優於離散粒徑分佈。在實驗過程中，收集了兩組水微粒的消光光譜。其光譜間的R2值介於0.8至0.9，回推的幾何平均數(幾何標準差)介於1(1.7)至4(2) 微米。實驗數據所得的回推結果與電腦模擬的結果相同，較大的微粒具有較好的回推結果。 結論: 對於大於3.5微米的微粒，最佳化演算法可良好執行。在雜訊程度小於10%的情況下，最佳化演算法對雜訊是不太具有敏感性的。我們的研究結果證明了可利用開徑式傅立葉轉換紅外光光譜儀(OP-FTIR)的測量結果回推氣膠粒徑分佈。 關鍵字: 懸浮微粒、開徑式傅立葉轉換紅外光光譜儀、消光光譜

Objective: To investigate the feasibility of retrieving aerosol size distribution information from OP-FTIR measurements. Method: We first obtained the complex refractive index of water, ammonium nitrate and ammonium sulfate from published literatures. The extinction spectra were then simulated with various size distribution (geometric mean ranged from 2 to 10μm; geometric standard deviation ranged from 1.1 to 2.5) based on the Mie theory. An optimization algorithm was developed to retrieve the geometric mean and standard deviation of the aerosols size distribution from the spectra, assuming the complex refractive index is known. We also added 1%, 4%, 7% and 10% noise levels to the simulated spectra for sensitivity analysis. We further collected the extinction spectra with a FTIR system for water aerosols in a controlled environment to verify the simulation study. Results: For all three compounds, when the geometric standard deviation was the same, the reconstruction results for larger particles were better than for the smaller particles. The sensitivity analysis results showed that when the noise level was less than 4% and geometric mean was large (>3.5μm), the fit was good with the R2 value between the input and reconstructed spectra ( ) greater than 0.9. The differences between the input and retrieved geometric mean and geometric standard deviation were less than 10% under the same condition. When the noise level was equal to 10%, the reconstruction results for a narrow aerosol size distribution were better than for a wide distribution. In the FTIR-aerosol experiments, two unique extinction spectra of water aerosols were collected. The ranged from 0.8 to 0.9 and the geometric mean (geometric standard deviation) ranged from 1 (1.7) to 4 (2) μm. Similar to the findings in the simulation study, the reconstructed spectra fit better to the input spectra for larger particles than for the small particles. Conclusion: The reconstruction algorithm performs reasonable well for particles larger than 3.5μm. It is also less sensitive to the spectral noise for the noise levels up to 10%. Our study results demonstrate that it is feasible to retrieve the aerosol size distribution from the OP-FTIR measurements. Key words: particulate matter, OP-FTIR, extinction spectrum