【摘要】：中藥產(chǎn)地及生長環(huán)境的改變會對其次生代謝產(chǎn)物、質(zhì)量產(chǎn)生整體性影響。近年來,三七產(chǎn)地從道地產(chǎn)地文山擴展到周邊縣市。為保證三七質(zhì)量,中國藥典以三種皂苷含量為指標對三七進行質(zhì)量控制,指標較為單一,難以對三七質(zhì)量進行整體性評價。通過紫外-可見分光光度法測定三七總黃酮含量,傅里葉變換紅外光譜結合化學計量學建立三七總黃酮含量快速預測模型;為三七快速及整體性質(zhì)量控制提供研究基礎。采集云南省12個產(chǎn)地96個三七樣品的紫外與紅外光譜。記錄樣品紫外光譜268nm處吸光度,結合蘆丁標準品線性方程計算樣品中總黃酮含量。預處理紅外光譜數(shù)據(jù)采用一階(1D)、二階導數(shù)(2D)結合SavitskyGolay平滑(7點、9點和11點)處理,Kennard-stone算法將96個個體分為2/3訓練集與1/3預測集。訓練集數(shù)據(jù)用于正交信號校正偏最小二乘回歸(OSC-PLSR)模型的建立,1/7交叉驗證用于篩選最佳主成分數(shù),預測集數(shù)據(jù)對OSC-PLSR模型預測能力進行驗證。結果顯示:(1)標準品蘆丁在268nm處吸光度與濃度相關系數(shù)r=0.999 7,線性濃度范圍為5.6~72.0μg·mL~(-1),線性關系良好;(2)道地產(chǎn)區(qū)文山州3個產(chǎn)地以及曲靖市羅平、昆明市石林等產(chǎn)地三七總黃酮含量較高,平均含量高于7mg·g-1;(3)相同點數(shù)SavitskyGolay平滑之后,二階導數(shù)模型預測能力優(yōu)于一階導數(shù),不同處理模型預測能力具有較大差別;(4)預測模型中,2D+SG 7+OSC-PLSR(R2pre=0.976 1,RMSEP=0.325 2)與2D+SG 11+OSC-PLSR(R2pre=0.946 9,RMSEP=0.382 0)模型預測效果較好,RMSEP均小于0.4,預測值與檢測值接近。傅里葉變換紅外光譜結合OSC-PLSR能夠對12個產(chǎn)地三七中總黃酮含量進行快速準確的預測,為三七整體性質(zhì)量控制提供一種快速、簡便、有效的檢測方法。
[Abstract]:The change of origin and growth environment of traditional Chinese medicine will have a holistic effect on the secondary metabolites and quality of traditional Chinese medicine. In recent years, the producing area of Panax notoginseng has expanded from Wenshan, the place of origin, to the surrounding counties and cities. In order to ensure the quality of Panax notoginseng, the Chinese Pharmacopoeia took the content of three kinds of saponins as the index to control the quality of Panax notoginseng, which was relatively single, so it was difficult to evaluate the quality of Panax notoginseng as a whole. The content of total flavonoids in Panax notoginseng was determined by UV-vis spectrophotometry, and the rapid prediction model of total flavonoids content of Panax notoginseng was established by Fourier transform infrared spectroscopy combined with chemometrics, which provided the research basis for the rapid and integral quality control of Panax notoginseng. The UV and IR spectra of 96 Panax notoginseng samples from 12 producing areas in Yunnan Province were collected. The absorbance at 268nm of UV spectrum of the sample was recorded, and the content of total flavonoids in the sample was calculated according to the linear equation of rutin standard product. The preprocessing infrared spectral data were processed by first order (1D), second derivative (2D) combined with SavitskyGolay smoothing (7 points, 9 points and 11:00). 96 individuals were divided into 2 鈮，

本文編號：2525317