氣流溫度對PM2.5採樣準確性的影響

Abstract

本研究使用多濾紙PM10-PM2.5採樣器(Multi-filter PM10-PM2.5 Sampler, MFPPS)於交大環工館頂樓進行24小時PM2.5¬採樣，其中MFPPS的四個PM2.5頻道分別裝有一個可控溫式致冷濾紙匣(Chilled Teflon filter)、兩個多孔金屬片固氣分離採樣器(Porous-Metal Denuder Sampler, PDS)及一個單一鐵氟龍濾紙匣(Single Teflon filter)。本研究比較MFPPS各頻道的採樣結果，探討降低採樣氣流對半揮發性無機物質揮發損失的影響，並評估PDS以兩片濾紙(PDS2，以尼龍濾紙採集微粒)取代三片濾紙(PDS1，以鐵氟龍濾紙採集微粒)進行採樣的可行性。本研究採樣時間為2015年12月08日至2016年12月14日，共採30天，日平均PM2.5, act 濃度範圍在8.77至52.76 g/m3 (平均: 24.80 ± 12.20 g/m3)。 研究結果顯示，未經降溫的單一鐵氟龍濾紙測得之總離子濃度(Cion, ST)低於PDS1測得之實際總離子濃度(Cion, act)，第一階段平均低17.24 ± 13.75 %，第二階段平均低13.62 ± 6.55 %，總揮發量占PM2.5, act平均濃度為11.41 ± 9.58 %；經4℃致冷濾紙匣後測得之總離子濃度(Cion, CT和Cion, CT 24)與Cion, act兩種濃度之間的誤差並不顯著，第一階段平均相差0.41 ± 9.41 %，第二階段平均相差3.99 ± 5.79 %，總揮發量占PM2.5. act平均濃度為3.53 ± 5.94 %。故使用致冷濾紙匣可有效減少半揮發性物質的揮發損失，增加PM2.5 act測值的準確性。分析PDS1採集到的微粒揮發成分發現，在採樣過程中揮發掉的NH4+、NO3-及Cl-濃度占各自實際離子濃度(Cion, act)的比例分別為36.61 ± 17.32、60.21 ± 22.82及71.55 ± 23.06 %，而占PM2.5, act的比例約為5.46 ± 3.48、4.58 ± 3.57及1.14 ± 0.96 %。 評估致冷濾紙能否正確量測PM2.5濃度的部分，致冷濾紙測得之PM2.5質量濃度值(PM2.5, CT)與PDS1測得之真實PM2.5濃度(PM2.5, act)之間平均相差為15.56 ± 14.10 %，推測可能的原因為採樣溫度皆低於當日的露點溫度，因此會有大量的水冷凝並包覆在微粒上，使濾紙取出時表面會有一層水分，在24小時調理的過程中不足以將所有水分去除，導致PM2.5, CT較PM2.5, act高。 PDS1與PDS2總離子濃度相當接近，平均誤差為0.80 ± 4.88 %，顯示PDS使用兩片濾紙取代原本的三片濾紙採樣是可行的。

Field observations were done at National Chiao Tung University (NCTU), Taiwan from December 2015 to December 2016 to understand the evaporation loss of SVIM (Semi Volatile Inorganic Material) of fine particulate matter (PM2.5) collected by MFPPS (Multi-Filter PM10- PM2.5 Sampler) including four MFPPS PM2.5 channels: two Porous-Metal Denuder Samplers (PDSs) with different filter assemblies, a Teflon filter cassette, and one chilled filter holder. The measured PM2.5 mass concentration (PM2.5, act) ranged from 8.77 to 52.76 g/m3 (ave. ± std.: 24.80 ± 12.20 g/m3). Results show that the actual ion concentrations of the single Teflon filter (Cion, ST) are lower than those of the PDS1 (Cion, act) (period #1: -17.24 ± 13.75% and period #2: -13.62 ± 6.55 % in average). In contrast, no significant differences in ion species concentrations are observed between the chilled Teflon filter (Cion, CT and Cion, CT 24) and the PDS1 (period #1: 0.41 ± 9.41 % and period #2: 3.99 ± 5.79 % in average). During 24-h sampling, the evaporated concentrations of NH4+、NO3- and Cl- of the single Teflon filter are accounted for 36.61 ± 17.32、60.21 ± 22.82 and 71.55 ± 23.06 % of Cion, act in each species, respectively, or 5.46 ± 3.48、4.58 ± 3.57 and 1.14 ± 0.96 % of PM2.5, act, respectively. Meanwhile, the total evaporation loss of the single Teflon filter and the chilled Teflon filter are 11.41 ± 9.58 and 3.53 ± 5.94 % in PM2.5 concentrations, respectively. Therefore, chilled filter holder can effectively decrease the evaporation loss of PM2.5 mass concentration. Although no significant difference of total ion concentration measurement observed between the chilled Teflon filter and PDS1, it can be seen that the measured PM2.5 concentration of chilled Teflon filter (PM2.5, CT) is 15.56 ± 14.10% higher than PM2.5 act. The temperature in chilled filter holder (4℃) is lower than the dew point of incoming air during sampling period while RH is not low enough after drying (24.84 ± 5.20 %), leading to water condensation on the collected particles and chilled Teflon filter surface as observed during field test. The condensed water is not evaporated during the 24-h conditioning process effectively, resulting in PM2.5, CT is higher than PM2.5, act. No significant differences in ion species concentrations are measured by the PDS1 and PDS2 which have different filter assemblies are used, this means that using the PDS with the assembly of two filters pack (PDS2) to replace that of three filters pack (PDS1) for PM2.5 measurement is feasible.