| Abstract: | 本研究以恆壓過濾(dead-end)系統進行探討，使用OMEGA公司所生產之商業化高分子薄膜(1,000及5,000 Da MWCO)，就溶液性質、薄膜特性及操作參數等進行過濾實驗，探討在不同操作參數(透膜壓差、濃度、pH值、溶液組成、攪拌速度)下對濾速、阻力與阻隔率的影響，並由理論模式定量本實驗系統的操作條件對阻力值之影響及分析理論濾速與實驗濾速之差異。
在BSA與β-cyclodextrin的混合溶液系統中，薄膜對β-cyclodextrin之阻隔率隨溶液pH值不同而有所改變，且加入攪拌可使阻隔率提高；以MWCO 5k薄膜為例，在攪拌系統下，pH=4.9時，β-cyclodextrin之阻隔率約為50%，而在pH=10時，則只有10%左右之阻隔率，這是因為BSA粒子形成之極化濾餅層緊密度不同所致。當未加入攪拌時，不論1k或5k薄膜，其β-cyclodextrin之阻隔率均近乎於零，但加入攪拌後1k薄膜之阻隔率皆高於90%。顯示改變BSA性質具有對β-cyclodextrin阻隔的效果，但也造成分離選擇性降低。
以滲透壓模式計算理論濾速，可發現理論與實驗濾速之趨勢相符，在高透膜壓差下，理論計算值有明顯緩升之趨勢，但實驗值隨壓力改變之濾速下降程度相當不明顯。這可能是在攪拌系統中所產生之剪應力對BSA產生的濃度極化之移除具有相當良好的效果，使濾速無明顯下降，且滲透壓模式只考慮膜面濃度提高對滲透壓所產生之效應而忽略薄膜結垢對濾速下降之影響，而使理論濾速與實驗濾速無法吻合。
In this study, commercialized polymer membranes(1,000 and 5,000 Da MWCO, OMEGA) were employed in a dead-end filtration system to investigate the effect of operation conditions, solutions and membrane properties on filtration. The solution fluxes, resistance and solute rejection were discussed under various operating parameters such as transmembrane pressure, concentration, pH value, solution composition and stirred velocity. In addition, the calculated filtration resistance by using resistance-in-series model and the predicting flux of BSA solution using the osmotic pressure model had been discussed.
For BSA and β-cyclodextrin binary solution, the rejection of β-cyclodextrin varies with the pH value, and the rejection with stirring increases. In the case of MWCO 5k membrane, the rejection of β-cyclodextrin is approximately 50% at pH=4.9 and near 10% at pH=10. This is due to the fact that the porosity of the polarization layer of BSA on the membrane surface varies with pH value. No matter 1k or 5k membrane but without stirring, the rejection of β-cyclodextrin is near zero；However, in the stirring system, the rejection of β-cyclodextrin by the 1k membrane is higher than 90%. The result indicates that, the cake property of BSA layer has significant influence on the rejection of β-cyclodextrin, and lead to reducing separation selectivity.
In osmotic pressure model, the trends of theoretical flux agree with experiment flux. At high transmembrane pressures, the theoretical flux apparently reached a plain, while the flux decline tendency was not obvious in the experiment. This is due to the shear stress caused by the stirring system can eliminate the concentration polarization layer of BSA effectively. The osmotic pressure model considers the osmotic pressure caused the concentration polarization layer but neglects the effect of membrane fouling on flux. Therefore, the theoretical flux is higher than the experimental data. |
