纳滤去除水中中性药物的预测模型

Modeling the rejection of neutral pharmaceuticals by nanofiltration membranes

纳滤是去除饮用水中微量药物污染物的一种有效方法。本文选择中性药物安替比林（ANT）和异丙基安替比林（AMT）为研究对象,考察利用溶解-扩散模型和薄膜理论相结合的方法预测一种常用的纳滤膜NF-X对ANT和AMT的去除效果.模型建立过程中,首先在不同初始浓度和压力条件下进行实验,测得通量和去除率.试验结果表明,在温度一定的条件下,渗透通量与操作压力线性相关,而和初始浓度无关.根据已有的试验数据得到模型中的未知参数k和DK/δ,通过最优拟合的方法确定传质系数k与ANT和AMT初始浓度C_o关系为k=αC_o~β,而DK/δ与C_o无关,由此得到在不同操作压力和初始浓度条件下ANT和AMT去除率的预测模型（25℃,pH=7）.最后,通过选取其它初始浓度和压力对模型进行验证,实验值和预测值相对误差小于10%,预测效果较好.

Nanofiltration is an effective method to remove trace drug pollutants in water. In this paper, netural drugs antipyrine （ANT） and isopropylantipyrine （AMT） were chosen to evaluate the prediction results with membrane theory and solution-diffusion mode by NF-X. Firstly, the permeate flux and rejection were obtained at different pressure and feed concentrations. The results indicated that the permeate flux and pressure were linearly related at a certain temperature and it was independent on the feed concentration. Then the unknown parameters k and DK/8 in the model were determined by experimental data and the mass transfer coefficient k could be calculated by the feed concentration （k=aCo^β） through the best-fit method, while DK/8 was nearly constant. Thus the prediction model was established under different pressure and feed concentration conditions（25 ℃, pH=7）. Finally, the rejections of ANT and AMT at other concentrations and pressures were predicted with relative deviation less than 10%. This indicated that the prediction results were satisfactory.