类蛋白类溶解性有机物（pDOM）是天然水体中具有与蛋白质标准物质类似性征的溶解性有机物的统称，已知包括蛋白质、氨基酸、微生物副产物及其各种混合物。作为含氮消毒副产物的重要前驱物，pDOM的组成和去除在饮用水处理领域受到广泛关注。本研究以实际水体中的pDOM为研究对象，通过与标准物质对照，建立了高效尺寸排阻色谱联合光电二极管阵列检测、荧光检测和有机碳检测（HPSEC-PDA/FLD/OCD）的分析方法，用于不同地表水中pDOM组成及理化特性表征，以及实际饮用水处理厂不同工段进出水pDOM组分的变化趋势研究。并在此基础上，进一步选取5种常见饮用水处理单元过程，通过在不同控制条件下对实际水样进行小试处理，系统探究了不同水处理工艺对pDOM组分的去除效果及机理。本论文的主要发现和结论如下：

（1） HPSEC-PDA/FLD/OCD分析能够快速、准确地对实际水体中pDOM各组分进行分离和表征。对不同来源地表水样的分析结果表明，pDOM普遍存在于地表水体中，而且不同于标准有机物质（蛋白、氨基酸等），其分子量分布范围广、组成复杂。其中，中、高分子pDOM主要组成为芳香性蛋白质（类色氨酸型和类酪氨酸型）和微生物副产物（类色氨酸型），小分子pDOM主要为氨基酸、多肽类物质等（类色氨酸和类酪氨酸型）。中分子组分在本研究选取的天然水中含量通常最高，浓度范围为0.91-6.11 mg L^-1，且芳香性较强。此外，在污染程度较轻、腐质化程度较高的水体中，溶解性有机物（DOM）通常以陆地衍生型为主，pDOM含量较低。而在受人为污染影响较大、微生物活动较强的水体中，DOM更倾向于微生物衍生型，并且这些水中的pDOM组分，尤其是分子量较低和疏水性较强的类色氨酸类组分含量较高。

（2）在实际饮用水厂的处理工艺流程中，不同分子量和结构的pDOM组分表现出不同的去除特性。预氧化主要作用于一些分子量较大的类色氨酸和类酪氨酸型pDOM，并且随氧化剂氧化作用的增强，更多的高分子组分被氧化分解为小分子，同时小分子pDOM的氧化效率也有所提升，但对于分子量较大的多糖和脂肪族类蛋白等组分，预氧化处理的去除效果较差。混凝对分子量较大的pDOM组分有良好去除效果，尤其是对疏水性较强的类酪氨酸型pDOM。活性炭过滤对中、低分子量pDOM组分去除效果较好，但在水厂运行过程中，可能由于微生物在炭床滋生，导致炭滤出水中pDOM组分，尤其是类色氨酸类组分含量升高。此外，紫外消毒在一定程度上可降解小分子pDOM组分。相比于类腐殖质类物质，pDOM的处理受水厂设施微生物活动及水质变化影响更为明显。由于所调研水厂运行过程中水质条件的复杂性和微生物活动的影响，不同水处理单元没有发挥协同作用，pDOM整体去除效果较差。

（3）通过在小试条件下对比不同水处理单元过程，发现纳滤和离子交换对天然水体中pDOM的去除效果优于其它过程。对比两者，阴离子交换树脂对高分子pDOM组分去除效果低于纳滤，且在处理过程中会释放一些具有芳香性结构的小分子疏水性物质，造成出水小分子有机物增多。而纳滤出水仅残留微量的有机组分，主要为名义分子量小于2569 Da，且疏水性较强的类酪氨酸类和脂肪族有机物。

（4）对照实际水厂和实验室小试结果，几种常规水处理单元对pDOM各分子量组分的去除规律相似，但由于各组分的化学结构等不同，去除效果存在一定差异。相比于类酪氨酸型pDOM，不同处理过程出水中类色氨酸型pDOM的含量及组分结构等对微生物活动和水质变化的影响更为敏感，去除效能易受水处理条件影响而变化。而类酪氨酸型pDOM组分则相对稳定，因此在水厂各工艺连续处理过程中，类酪氨酸型pDOM在组合工艺中显示更好的去除效果，而实际水厂对于类色氨酸型pDOM的去除效果与单一水处理单元相比变化较大，去除效率较低。

（5）相比于地表水样，污水处理厂二沉池出水中pDOM各组分浓度均较高，在不同物化水处理单元中难以有效去除，因此如果将其消毒后排放进入地表水，可能会提高水源水的消毒副产物污染风险。同时，处理后污水中残留pDOM将影响天然水体中DOM组成，增大饮用水处理过程中有机污染物的去除难度。

综上所述，本研究建立了HPSEC-PDA/FLD/OCD分析方法，并用于实际水体pDOM常见组分的理化特性及含量分布研究。所获结果初步揭示了pDOM不同分子量组分在实际饮用水处理厂的去除特性及迁移转化规律，探明了几种常用水处理技术对pDOM各组分的去除机理。这些成果为饮用水厂水源地保护、日常工艺水质检测、处理过程优化，以及出水消毒副产物控制提供了方法支持和理论参考。

Protein-like dissolved organic matter (pDOM) refers to dissolved organic matters in natural water bodies that possess the characteristics of standard protein compounds. These substances encompass a heterogeneous mixture of proteins, amino acids, microbial by-products and some other organics. The control and elimination of pDOM have been a growing concern for drinking water treatment plants due to its high formation potential for nitrogenous disinfection by-products (N-DBPs). In this study, a high performance size exclusion chromatography system coupled with photo-diode array, fluorescence, and online organic carbon detectors (HPSEC-PDA/FLD/OCD) was used to determine the occurrence and physicochemical properties of different-sized pDOM fractions in selected surface waters. Besides, the system is also applied to investigate the removal behaviors of different-sized pDOM fractions in full-scale drinking water treatment plants (DWTPs). Based on these results, five types of unit processes were further assessed at bench scale in order to determine their individual removal efficiencies and mechanisms for pDOM fractions in realistic waters. Major findings and conclusions of this study are presented as follows:

（1）HPSEC-PDA/FLD/OCD analysis was proven to be a convenient and accurate technique for the fractionation and characterization of pDOM in realistic waters. By using this technique, pDOM was found to be ubiquitous in the studied surface waters. However, different from the reference compounds, pDOM in surface waters possessed broader molecular weight (MW) distributions and more complex components. The medium MW fractions contained more aromatic structures at the highest concentration range of 0.91-6.11 mg L^-1. Additionally, for surface waters with less pollution, DOM is mainly terrestrial-derived type with high aromaticity. While for those affected by exogenous pollutions and strong microbial activities, DOM is mainly microbial-derived type and possessed more tryptophan-like components with low MW and high hydrophobicity.

（2）Different-sized pDOM components exhibited different removal behaviors in full-scale DWTPs. Pre-oxidation mainly targeted some tyrosine- and tryptophan-like components with high MW. As the oxidization effect was enhanced, greater portions of the high MW fractions were decomposed into low MW ones and the oxidation efficiency of the low MW components was also improved. However, some high MW fractions (e.g., polysaccharides and some aliphatic proteins) were not subject to pre-oxidation removal. Meanwhile, the subsequent coagulation was efficient in removing high MW fractions, especially some tyrosine-like pDOM with strong hydrophobicity. Then after, the granular activated carbon (GAC) filtration effectively removed residual medium and low MW pDOM fractions, but also increased the concentrations of some pDOM components, especially the tryptophan-like type, due to microbial growth on the filter media. Furthermore, the downstream UV disinfection process caused degradation of some low MW fractions. Unlike humic-like DOM, pDOM were susceptible to microbial activities and water quality changes that took place during the water treatment processes in DWTPs. As a result, the aforementioned water treatment processes did not exhibit synergistic effect on pDOM removal, which led to limited pDOM removals in the surveyed DWTPs.

（3）By comparing the results obtained in bench-scale experiments, it was found that nanofiltration and ion exchange were more efficient unit processes for pDOM removal. Among the two, ion exchange exhibited relatively low removal rate for high MW pDOM fractions. Besides, it released some low MW fractions with aromatic structures and high hydrophobicity, thereby increasing the contents of low MW organics. In comparison, only a small portion of organics remained in the permeate of nanofiltration, which were mainly hydrophobic tyrosine-like and aliphatic pDOM with MW under 2569 Da.

（4）By comparing the bench- and full-scale results, it was found that pDOM exhibited similar removal trends at both treatment scales. However, due to their diverse chemical structures, there were distinguishable differences in the removal efficiencies of different pDOM species. Overall, tryptophan-like pDOM was more sensitive to the effects exerted by microbial activities and water quality changes. Therefore, its removal showed greater differences at full- and bench-scale water treatment processes than other types of pDOM. Comparatively, tyrosine-like pDOM was more stable under different water quality conditions, and therefore, showed higher removal efficiencies in the full-scale DWTPs.

（5）The secondary wastewater effluent (SWE) contained relatively high concentrations of different-sized pDOM species, which made it more difficult to remove during water treatment process. Consequently, there would be a greater potential for DBPs pollution if SWE is discharged into natural water bodies after disinfection. Meanwhile, SWE discharge was also expected to affect DOM composition of natural waters, thus increasing the difficulty of DWTPs in DOM removal.

Overall, this study established an analytical method based on HPSEC-PDA/FLD/OCD and further applied it to the fractionation and characterization of pDOM in realistic water samples. The results provided a preliminarily insight into the removal behaviors of different-sized pDOM fractions in full-scale DWTPs, as well as their removal mechanisms during various unit treatment processes. Application of these results will assist DWTPs in improving source water protection, process monitoring and optimization, and implementation of DBPs formation control measures.