Since the events of aristolochic acid poisoning were reported widely around the world, the safety of Traditional Chinese Medicine (TCM) was greatly concerned. The thesis optimized the supercritical fluid extraction (SFE) conditions of aristolochic acids I & II (AAI & II) from Aristolochia plants---Fructus Aristolochiae and Caulis Aristolochiae manshuriensis. The influences of related factors on the extraction efficiency and possible reasons were discussed. Selectivity of the optimized SFE method was investigated for the potential application of removing aristolochic acids in the future. Extraction pressure (p), temperature (T) and mole fraction (x) of the entrainer---absolute ethanol were mainly optimized, which had great interactions. Based on the CO₂-ethanol P-T-x phase diagram, a special univariate design was introduced that was able to assist with fast analysis of results. Extraction time and mode were also investigated. In the experiments, the sample powder was extracted by supercritical fluid and the residue was then analyzed on the extraction rates of AAI & II by pressurized liquid extraction (PLE) and high performance liquid extraction (HPLC), which were all optimized. After optimization, the final SFE conditions were the extraction pressure of 194 bar, extraction temperature of 50°C, mole faction of the entrainer---absolute ethanol of 0.20, extraction time of 4 hours including 3-hour dynamic extraction. Under this method, the extraction rates of AAI & II for Fructus Aristolochiae and Caulis Aristolochiae manshuriensis were 65.23%, 59.07% and 81.25%, 81.19%, respectively. Mole fraction of the entrainer was dominant factors, while the extraction pressure and temperature had little influences. Besides, extraction mode had great effect. Dynamic mode was much better than the static one. Selectivity, depicted by the change of peak area of individual peak or peak combination, was investigated and the optimized SFE method did show its selectivity. In a word, the optimized SFE method was able to extract AAs from Fructus Aristolochiae and Caulis Aristolochiae manshuriensis. Mole fraction of the entrainer and extraction mode were the determinants. Selectivity of this method offered opportunity for the future application of SFE method in the removal of AAs from Aristolochia plants. The optimized HPLC condition was able to separate and detect the analytes satisfactorily. PLE method mainly influenced by the extraction solvent, parted site of the sample powder, extraction temperature and pressure, exhibited high efficiency and good applicability in the pretreatment of Aristolochia samples and was much better than conventional methods. Five chapters are included in this thesis. Chapter 1 is the introduction to the background, objective and the experimental design of this research and relate information on Aristolochia plants and aristolochic acids. Chapter 2 focuses on the optimization of the HPLC method and discusses the generations of errors. Chapter 3 gives the overview of the optimization of PLE method by the univariate way and recovery, applicable and comparison experiments were also presented. Chapter 4 is the key content of this thesis. SFE optimization is discussed including several factors such as p-T-x. as well as other factors based the phase diagram. Selectivity of SFE method is also investigated. Chapter 5 is the conclusion of this study. Key words: Supercritical fluid extraction; Fructus Aristolochiae; Caulis Aristolochiae manshuriensis; Aristolochic acid; HPLC; PLE; Univariate Design

中药的安全性一直是一个备受关注的问题。近年来马兜铃酸中毒的事件大量被报道，很多含有马兜铃酸的中药被禁用。本课题主要优化了超临界流体从马兜铃属植物（马兜铃和关木通）中提取马兜铃酸 (AAI & AAID） 的条件，并讨论了各因素的影响及可能的原因。最后，本文对超临界提取(SFE）的选择性进行了考察，以探讨该法将来从马兜铃属植物中去除马兜铃酸的可能性。 实验在维持药材前处理和收集方法不变的情况下，重点优化了提取过程中的压力、温度、携带剂含量等参数。这三个参数有着很强的交互作用，是影响提取过程的主要因素。本文依据乙醇-二氧化碳二元体系相图，对三者的优化进行了实验设计。该设计有助于我们较快地得到结果并分析可能的原因。实验亦对提取时间和提取方式等因素进行了考察。实验中，粉碎后的药材粉末首先经过超临界二氧化碳萃取，萃余物再经过加压溶剂提取(PLE) 及高效液相色谱 (HPIC) 分析，计算两个指标性成分即马兜铃酸I和II的提取率。PLE和HPLC 的条件也进行了优化。 最后优化的 SFE 条件是：压力194 bar，温度 50°C，携带剂乙醇的摩尔分数 0.20，提取时间 4 小时，其中包括 3小时动态提取。据此条件，马兜铃和关木通中马兜铃酸I和II的提取率分别为 65.23%，59.07% 和 81.25%，81.19%。本研究表明，携带剂的摩尔分数对提取效率有显著性影响，是主导性因素，而温度和压力对提取的影响却不大。另外，提取方式对于提取有非常大的影响，动态提取的效果要远远好于静态提取，可以达到较好的提取率。本实验亦对提取前后各峰的峰面积变化进行了比较以考察提取的选择性。该工作表明优化过的 SFE 方法具有一定的选择性。总之，本文建立的 SFE 方法可以较好地从马兜铃和关木通中提取马兜铃酸，其中携带剂的摩尔分数和提取方式是主要影响因素。该方法显示了一定的选择性，这为将来用 SFE 的方法去除马兜铃酸提供了可能。 本文共分五章。第一章简要介绍了本研究的背景、目的、目标，以及马兜铃属植物和马兜铃酸的有关情况。实验设计的思路也进行了介绍。第二章集中在色谱条件的优化上，并对误差产生原因进行了探讨。第三章介绍了加压溶剂提取优化的有关实验，包括各因素的影响，方法的回收率、适用性和与其它方法的比较等。第四章是本文的核心内容。该章优化了很多包括摩尔分数、温度、压力在内的和超临界提取有关的关键因素井对可能的机制做了探究。为了提高实验效率，方便结果分析，本文特别根据流体的相行为设计了实验。最后对 SFE提取的选择性进行了考察。第五章是本文的总结。 关键词:超临界流体萃取，马兜铃，关木通，马兜铃酸，高效液相色谱，加压溶剂提取，单因素考察