基于天然低共熔溶剂提取川芎有效成分和抗氧化活性的工艺优化研究?

Optimization of extraction process of active components and antioxidant activity of Ligusticum Chuanxiong with natural low eutectic solvent

关键词:  天然低共熔溶剂  响应面试验  遗传神经网络  抗氧化  工艺优化

Key words:  nature deep eutectic solvents  response surface test  genetic neural network  antioxidant  process optimization

摘要:

摘要：目的 优选川芎中酚酸类成分及其抗氧化活性的提取工艺条件。方法 利用高效液相色谱法从8种天然低共熔溶剂、纯水和甲醇中筛选川芎酚酸类成分提取溶剂，再采用超声辅助开展相关提取实验。以Bet 13浓度（A）、固液比（B）、超声时间（C）、提取温度（D）为工艺参数，开展四因素三水平的响应面实验设计。通过熵权法计算川芎中阿魏酸、原儿茶酸、绿原酸、咖啡酸、香草酸五种成分提取率综合评分(Y1)，以及DPPH自由基清除率和抑制羟自由基能力的抗氧化活性综合评分（Y2），将Y1和Y2分别作为响应值。用响应面法和遗传神经网络分别建立提取工艺与响应值之间的关系，然后采用遗传算法优化得到最优条件，并进行实验验证。结果 比较发现，采用甜菜碱-甲基脲（Bet 13）作为提取溶剂具有相对优势。对于Y1，响应面和遗传神经网络模型的R2分别为0.79和0.94，相对误差分别为15.34%和6.46，遗传神经网络模型得到最优工艺条件为Bet 13浓度55%、固液比 96：1 mg/mL、超声时间36 min、提取温度36℃。对于Y2，响应面和遗传神经网络模型的R2分别为0.65和0.95，相对误差分别为19.32%和7.41，得到最优工艺条件为Bet 13浓度55%、固液比 71：1 mg/mL、超声时间39 min、提取温度38℃。结论 遗传神经网络与实验数据的拟合度较高，优化得到的提取工艺稳定可靠，为川芎工业化提取和实验研究提供参考。

Abstract:

ABSTRACT: OBJECTIVE To optimize the extraction process conditions for the phenolic acid components and their antioxidant activities in Ligusticum chuanxiong Hort. METHODS High-performance liquid chromatography (HPLC) was utilized to screen the extraction solvents for the phenolic acid components of Ligusticum chuanxiong from eight natural low eutectic solvents, pure water, and methanol, and then ultrasound-assisted to carry out the relevant extraction experiments. A four-factor, three-level response surface experimental design was carried out using Bet 13 concentration (A), solid-liquid ratio (B), ultrasound time (C), and extraction temperature (D) as process parameters. The composite scores (Y1) of the extraction rate of five components, ferulic acid, protocatechuic acid, chlorogenic acid, caffeic acid, and vanillic acid in Ligusticum chuanxiong Hort, as well as the composite scores (Y2) of the antioxidant activity of the DPPH radical scavenging rate and the ability to inhibit hydroxyl radicals were calculated by entropy weighting, and Y1 and Y2 were taken as the response values, respectively. The relationship between the extraction process and the response values was established by response surface methodology and genetic neural network, respectively, and then the optimal conditions were obtained by optimization using a genetic algorithm and experimentally verified. RESULTS Comparison revealed that using betaine-methylurea (Bet 13) as the extraction solvent has a relative advantage. For Y1, the R2 of the response surface and genetic neural network models were 0.79 and 0.94, respectively, and the relative errors were 15.34% and 6.46, respectively. Genetic neural network modeling yielded optimal process conditions of Bet 13 concentration of 55%, solid-liquid ratio of 96:1 mg/mL, ultrasonic time of 36 min, and extraction temperature of 36 °C. For Y2, the R2 of the response surface and genetic neural network models were 0.65 and 0.95, respectively, and the relative errors were 19.32% and 7.41, respectively, which gave the optimal process conditions of Bet 13 concentration of 55%, solid-liquid ratio of 71:1 mg/mL, ultrasonic time of 39 min, and extraction temperature of 38 °C. CONCLUSION The fit between the genetic neural network and the experimental data was high, and the extraction process obtained by optimization was stable and reliable, which provides a reference for the industrialized extraction and experimental research of Ligusticum chuanxiong Hort.