饮用水中频繁检出的药物污染物对人类健康带来潜在威胁。由于这些污染物种类繁多、理化性质复杂，传统饮用水处理技术难以将其有效去除。本论文针对已有深度水处理技术的不足，构建了基于碳基纳米材料的新型电催化臭氧膜过滤耦合技术（EPMF）。该技术通过膜过滤传质与电催化臭氧化作用的有机结合，突破了以往吸附过滤的原位再生难题，降低了电催化臭氧过程的传质限制，实现了水中低浓度药物的高效、稳定降解去除，为饮用水安全保障提供了新途径。

论文首先制备筛选出2种不同结构特性的碳材料复合膜过滤电极，并分别构建了基于2种膜电极的电催化臭氧膜过滤系统（EPMF）。在此基础上，以布洛芬为模型污染物，依次探究了EPMF中电化学产过氧化氢、自由基生成、污染物去除的化学反应机制，以及过滤强化对流传质的作用机制。通过对以上机制解析，深入揭示了这些机制在碳布和碳纳米管复合膜过滤中的耦合效应，以及两种电极EPMF的布洛芬去除机理。最后，进一步验证了EPMF对多种药物污染物和天然水中低浓度药物的处理效果。

主要研究内容和结论如下：

（1）     碳布复合膜过滤电极EPMF去除布洛芬的催化反应过程。采用压制法制备而成碳布复合膜过滤电极，该电极具有疏松的三维网状膜孔结构，平均膜孔径为4.55 μm，电极活性层（碳布层）平均厚度为355.5 ± 12.74 μm。最佳条件时，该电极的过氧化氢产率为972.48 mg H₂O₂ m^-2 h^-1，相应的电流利用率为72.69 ± 1.53%。EPMF过程中对布洛芬降解起主要作用的氧化剂是·OH，由电化学原位生成的过氧化氢与臭氧反应（electro-peroxone反应）生成，浓度可达18.34 ± 2.15 μM；EPMF能显著提升布洛芬降解效率，膜通量为140.35 L m-2 h-1时单次过滤布洛芬的去除率达64.48 ± 1.56%，分别是电耦合膜过滤、臭氧耦合膜过滤和peroxone耦合膜过滤去除率的3.25，3.14和1.59倍。

（2）     碳纳米管复合膜过滤电极EPMF去除布洛芬的催化反应过程。与碳布相比，碳纳米管具有更好的导电性、催化活性，以及较高的比表面积。采用抽滤法制备而成碳纳米管复合膜电极，该电极具有紧密排布的微孔结构和均匀的微观导电特性，这些有利于增大复合膜电极的活性面积，提供更多活性位点。碳纳米管层通过催化/活化臭氧和电化学产过氧化氢的分解，以及臭氧和过氧化氢反应，协同促进·OH生成。系统内·OH浓度达到57.42 ± 4.86 μM，远高于碳布膜电极。基于碳纳米管复合膜的EPMF过程具有吸附、电化学反应、臭氧氧化、催化氧化、peroxone氧化等多效协同作用，对布洛芬的去除率达到95.75 ± 1.12%，显著高于碳布EPMF过程的布洛芬去除率。

（3）     跨膜对流传质对EPMF去除药物污染物的影响。过滤及分析实验结果表明，跨膜对流传质对膜电极内部的电子传递、自由基生成、污染物降解反应动力学具有重要影响。随着过滤流速增大，复合膜过滤电极电流逐渐增大，电化学产过氧化氢的浓度先升高，后随副反应的增强而下降。当电化学产过氧化氢足量时，增大过滤流速会促进·OH生成，从而强化污染物的降解去除；而当流速过大时，过氧化氢产量不足以满足peroxone反应需要，因而自由基产量下降，污染物去除率降低。

（4）EPMF去除药物污染物的应用效果。针对实际水处理进水pH存在波动，药物污染物种类繁多，以及存在其他共存物质的问题，分别开展过滤实验研究。结果显示，由于具有多效协同作用机制，在饮用水pH值波动范围内，EPMF对布洛芬具有良好且稳定的去除效果；此外，EPMF对具有不同臭氧反应速率的药物都具有长期（600 min）稳定去除效果，去除率最低为97.31 ± 0.67%（布洛芬）；同时，EPMF耦合系统能够有效去除天然水中低浓度药物污染物，与模拟配水实验相比，布洛芬的去除率仅下降8.85个百分点。因此，EPMF技术是一种具有良好潜力的饮用水深度处理技术。

The frequent detection of pharmaceutically active compounds (PhACs) in drinking water has posed a serious threat to human health. Due to their diverse physical and chemical properties, PhACs are difficult to be removed by conventional drinking water treatment technologies. In order to overcome the limitations of existing advanced water treatment processes, an electro-peroxone membrane filtration (EPMF) process was developed in this study for enhanced PhAC removal. Due to the convective mass transfer induced by the membrane permeate flow and the advanced oxidation processes incurred by electro-peroxone, the EPMF process was competent in eliminating the difficulty of membrane regeneration encountered in adsorption membrane filtration, as well as the restriction of mass transfer in batch electro-peroxone processes. These merits were beneficial for sustainable and efficient removals of PhACs in drinking water.

In this study, a carbon-cloth (CC) based composite membrane and a carbon nanotube (CNT) based composite membrane, each having its distinct structure, were prepared and used as the cathode for the EPMF system. Then, with ibuprofen as the model PhAC, filtration experiments and complemental analyses were conducted to sequentially investigate the mechanisms of H2O2 electrogeneration, hydroxyl radical production, ibuprofen degradation, and the effects of filtration-enhanced mass transfer. Based on these studies, the mechanisms of PhAC removal by the EPMF process was identified for the CC membrane and the CNT membrane, respectively. Finally, the applicability of the EPMF process was further investigated under conditions relevant to realistic drinking water treatment.

The main tasks and findings of this study are as follows:

(1)  Chemical processes involved in PhAC removal by the EPMF system with the CC-membrane cathode. The CC-membrane cathode was prepared by firmly pressing a selected CC on a substrate membrane made of polytetrafluoroethylene (PTFE). The prepared CC cathode possessed loose, 3-D porous structure, with nano-scale homogeneity in electro-conductivity. The main pore size of the CC/PTFE membrane was 4.55 μm, and the average thickness of the CC layer was 355.5 ± 12.74 μm. The in-situ electrogenerated concentration of H2O2 reached 6.92 ± 0.29 mg L^-1 in the filter permeate, equivalent to a unit generation rate of 972.48 mg H₂O₂ m^-2 h^-1 and a corresponding current efficiency of 72.69 ± 1.53%. Moreover, ibuprofen removal was mainly ascribed to advanced oxidation by ?OH that was formed in the EPMF system by the electroperoxone reaction. The permeate concentration of ·OH was 18.34 ± 2.15 μM. Overall, a single-pass EPMF at a permeate flux of 140.35 L m-2 h-1 achieved a stable IBU removal of 64.8%, which was 3.41 fold, 2.94 fold and 1.59 fold of those obtained in electrochemical filtration, ozonation-coupled filtration and peroxone-coupled filtration, respectively.

(2)  Chemical processes involved in PhAC removal by the EPMF system with CNT-membrane cathode. The CNT-membrane cathode was prepared by filtering dispersed CNT onto the PTFE membrane. A tightly entangled and conductive CNT layer was established on the PTFE membrane in this study. This nano-porous and conductive CNT structure provided a large surface area and abundant active sites for the reactions. The main pore size of the membrane was 0.09 - 0.1 μm, and the average thickness of the electrode active layer (CNT layer) is 78.8 ± 8.0 μm. Unlike the CC membrane, H2O2 and CNT-promoted O3 decomposition, along with the CNT-enhanced H2O2 conversion led to a high ·OH concentration of 57.42 ± 4.86 μM in the permeate of the CNT membrane. Meanwhile, the co-occurrence of various adsorption, chemical reaction, electrochemical reaction, ozonation, catalytic oxidation and peroxone reactions further enhanced the degradation and mineralization of ibuprofen. Overall, the EPMF system demonstrated stable IBU removal efficiency and mineralization efficiency of 95.75% and 87.83% in a single-pass filtration time, respectively, which were significantly higher when compared with CC membrane cathode.

(3)  The mechanism of enhanced IBU degradation due to permeation-induced convective mass transfer. The permeation-enhanced mass transfer significantly promoted the inner-cathode electron transfer, radical production, and contaminant degradation kinetics. The electron transfer effect was evidenced by the increased electrical current through the EPMF system with increasing feed flowrate. Moreover, the electrogeneration of H2O2 first increased and then decreased with increasing feed flowrate, due to the occurrences of side reactions. When the electrogenerated H2O2 was sufficient, the production of ?OH increased with increasing feed flowrate, thus enhancing the degradation and removal of ibuprofen. However, when the electrogenerated H2O2 was insufficient for the peroxone reaction, the production of ·OH decreased as the feed flowrate further increased, which led to lower ibuprofen removal efficiencies.

(4)  Applicability study of the EPMF for drinking water treatment. Due to the co-occurrence of synergistic mechanisms, the EPMF process performed equally well in removing ibuprofen in the typical pH range for drinking water treatment. In addition, extended filtration experiments demonstrated that the EPMF process was capable of sustainably and effectively removing various PhACs that had diverse reactivity with O3, as the lowest removal efficiency was observed for ibuprofen at 97.31 ± 0.67%. Furthermore, the ibuprofen removal efficiency remained at 86.90 ± 1.14% during the EPMF treatment of a realistic water, which was just 8.85% lower than that obtained with the synthetic model water. These findings showed a promising potential for EPMF to be applied in advanced drinking water treatment.