高效能液相層析(HPLC)及毛細管電泳(CE)是目前最常用來測定中藥成份含量的分析方法。綜合兩者之優點、合併運用，將可拓展中藥分析之範疇。本研究以此兩種儀器，開發桅子藥材、相關製劑及芍藥湯之分析方法，並比較CE與HPLC在實際分析中藥時的優劣。
中藥材桅子中的五個指標成分，可利用膠束電動力學毛細管層析(MEKC)於30分鐘內順利分離。當緩衝溶液為40mM硼酸鹽、60mM SDS及5M尿素(pH=9.48)的水溶液，電壓20kV，UV偵測波長於240nm可測定genipin-1-β-d-gentiobiosic acid(GN)、geniposide(GP)、gardenside(GD)與geniposidic acid(GPA)，於440nm則可偵測 crocin(CR)等之含量。定量濃度的線性範圍可達兩個數量級；各成分的偵測極限範圍在2.67-5.75μg/ml之間。遷移時間及面積的相對標準偏差皆小於2%(n=6)。本分析方法亦可運用於分析含桅子藥材之加味逍遙散、清肺湯、黃連解毒湯、辛夷清肺湯、龍膽瀉肝湯等五種製劑。
芍藥湯由芍藥、黃苓、大黃、黃連、肉桂、木香、甘草、當歸、檳榔等9種藥材組成，成分眾多，本研究選擇其中27種主要化合物，利用CE與HPLC兩種儀器分析。HPLC方面，沖提系統之流動相(A)為30mM KH2O4以0.01%H3PO4調至pH=3.25，流動相(B)為H2O/CH3CN3=20/80(V/V)，利用Cosmosile 5Cl8-MS之逆相分離管柱，以230nm及275nm為偵測波長，在70 分鐘內可順利完成27個成分的定量。定量濃度的線性範圍可達兩個數量級；各成分的偵測極限範圍在0.38-2.77μg/ml之間。滯留時間及積分面積的相對標準偏差皆小於2.3%(n=6)。
CE方面，開發了類梯度膠束電動力學毛細管層析(g-MEKC)。分析條件為：第一緩衝溶液含20mM硼酸鹽、40mMSC(pH=9.75)，第二緩衝溶液為AcCN；電壓20kV，毛細管長度ll0cm，偵測波長230nm及275nm，可分析芍藥湯中l7個指標成分。定量濃度的線性範圍可達兩個數量級；各成分的偵測極限範圍在3.23-l7.28μg/ml之間。遷移時間及積分面積的相對標準偏差皆小於3%(n=6)。g-MEKC之基本理論與分析作用機制於本研究中亦有所探討。
檢討芍藥湯的CE與HPLC的分析結果，發現在分析時間、基線穩定性、再現性、回收率及各波峰的理論板數等方面各有其優劣，兩者應可相輔相成，搭配使用。

High-performance liquid chromatography (HPLC) and capillary electrophoresis (CE) are analytical methods most commonly used presently to quantify components of Chinese herbal medicine. Working together, HPLC and CE can advantageously complement each other and broaden the field of analysis of Chinese herbal medicine. In this work, analyses of Gardeniae Fructus, applied to five relevant preparations and Peony Combination using HPLC and CE have been studied. The practical aspects and experimental factors of analyses by HPLC and CE have been compared and discussed.
Five standard components of Gardeniae Fructus can be separated by micellar electrokinetic chromatography (MEKC). The analytical conditions included a buffer solution containing 40mM borate, 60mM sodium dodecyl sulfate (SDS) and 5M urea (pH 9.48) and the analytical voltage set at 20kV. The wavelength of the UV-VIS spectrometer was fixed at 240nm for the detection of genipingentiobioside (GN), geniposide (GP), gardenside (GD) and geniposidic acid (GPA), and for the detection of crocin (CR) the wavelength was fixed at 440nm. Within 30 minutes, the five compounds could be quantified. The linearity range of the quantified concentration of all five compounds could reach 2 orders, and the detection limits fell between 2.67 and 5.75 μg/ml. The relative standard deviations of the migration times and the areas-were all smaller than 2.0% (n=6). The working analytical procedure can also be applied to the analyses of five relevant preparations.
Peony Combination contains nine component herbs, and 27 chief constituents were selected from all constituents of the formula. In this work,the analytical methods for Peony Combination using HPLC and CErespectively have been developed. In the respect of HPLC, the reverse-phase HPLC was used, and the elution system included dynaulic phase (A) (30mM KH2P04; pH3.25 adjusted by adding 0.01% H3P04) and dynamic phase (B) (H2O/CH3CN=20/80, V/V). With 5C18-MS separation column, the 27 chiefcomponents of Peony Combination were all quantified in 70 minutes by UV- VIS detection at 230nm and 275nm, respectively. The linearity range of the quantified concentration could reach 2 orders, and the detection limits fell between 0.38 and 2.77 μg/ml. The relative standard deviations of the migration times and the areas were all smaller than 2.3% (n=6).
In the respect of CE, gradient-MEKC (g-MEKC) has been developed to analyze Peony Combination. The analytical conditions used in this study included： the primary buffer solution contained 20mM borate and 40mM sodium cholate (SC) (pH9.75), and the secondary buffer solution was acetonitrile (AcCN). The analytical voltage was set at 20kV, and the length of capillary tube was 11Ocm. The wavelength of the UV-VIS spectrometer wasfixed at 230nm and 275nm respectively to detect 17 chief constituents. The linearity range of the quantified concentration could reach 2 orders, and the detection limits fell between 3.23 and 17.28 μg/ml. The relative standard deviations of the migration times and the areas were all smaller than 3.0% (n= 6). The basic theory and the proposed mechanism of separation and analysis of g-MEKC are also discussed in this study.
From the experimental results, it is concluded that for the analysis of Chinese herbal medicine, HPLC and CE possess some advantages and some disadvantages in terms of the analytical time, the stability of baseline, the order of movements of compounds and the numbers of theoretical plates of each compound, respectively. HPLC and CE complement each other in analysis of Chinese herbal medicine.