Prediction of Chromatographic Retention of Pyrazine and Alkylpyrazines in RP-LC

Retention prediction models for a group of pyrazines chromatographed under reversed-phase mode were developed using multiple linear regression (MLR) and artificial neural networks (ANNs). Using MLR, the retention of the analytes were satisfactorily described by a two-predictor model based on the logarithm of the partition coefficient of the analytes (log P) and the percentage of the organic modifier in the mobile phase (ACN or MeOH). ANN prediction models were also derived using the predictors derived from MLR as inputs and log k as outputs. The best network architecture was found to be 2-2-1 for both ACN and MeOH data sets. The optimized ANNs showed better predictive properties than the MLR models especially for the ACN data set. In the case of the MeOH data set, the MLR and ANN models have comparable predictive performance.     

Retention behavior of ginsenosides on a poly(vinyl alcohol)-bonded stationary phase in hydrophilic interaction chromatography

The influences of the organic component of the mobile phase and the column temperature on the retention of ginsenosides on a poly(vinyl alcohol) (PVA) bonded stationary phase operated under hydrophilic interaction chromatographic mode were investigated. The retention of the ginsenosides was found to increase with increasing amount of acetonitrile (MeCN) in the mobile phase, which is typical of hydrophilic interaction chromatographic behavior. It was also found that the retention of the analytes was highly affected by the type of the organic modifier used. Aqueous MeCN (75–90%) gave the most satisfactory retention and separation of ginsenosides Rf, Rg1, Rd, Re, Rc, Rb2 and Rb1 compared with aqueous methanol, isopropyl alcohol or tetrahydrofuran at the same composition levels. The effects of the different types of organic modifiers on the retention of the analytes were attributed to their solvent strength and hydrogen-bond accepting/donating properties. The effect of temperature on the retention of ginsenoside on the PVA-bonded phase was assessed by constructing van’t Hoff plots for two temperature ranges: subambient (273–293 K) and ambient-elevated (298–333 K) temperatures. van’t Hoff plots for all analytes were linear at the two temperature intervals; however, the slopes of the lines corresponding to ginsenosides Rg1 and Re were completely different from those for the rest of the analytes especially in the subambient temperature range. Enthalpy-entropy compensation (EEC) studies were conducted to verify the difference in thermodynamics observed for ginsenosides Rg1 and Re compared with the other analytes. EEC plots showed that Rf, Rd, Rc, Rb2 and Rb1 were possibly retained by the same retention mechanism, which was completely different from that of Rg1 and Re at subambient temperatures. Retention prediction models were derived using multiple linear regression to identify solute attributes that affected the retention of the analytes on the PVA-bonded phase. The mathematical models derived revealed that the number of hydrogen-bond donors and the ovality of the molecules are important molecular properties that govern the retention of the compounds on the chromatographic system.     

人参皂甙的反相高效液相色谱多台阶梯度优化方法

建立了一种反相高效液相色谱多台阶梯度分离人参皂甙的方法.该方法以乙腈-水溶液为流动相,通过一系列等度实验,获得了8种人参皂甙Rg1,Re,Rf,Rg2,Rb1,Rc,Rb2和Rd的色谱保留参数,发现两参数保留方程不适合用于人参皂甙这种天然产物的分离条件的优化,而三参数保留方程的高精度才可满足预测的要求.在三参数保留方程的基础上,通过计算确定了8种人参皂甙(包括3台阶梯度)的液相色谱分离条件.通过实验对此优化条件进行了验证,实验结果显示了较好的预测精度和分离度.将本方法用于分离人参皂甙,分析时间短且分离度高,显示了等度台阶梯度优化方法对确定色谱分离条件的优越性.     

High Performance Liquid Chromatographic Separation and Quantitative Analysis of Ginsenosides Using a Polyvinyl Alcohol-Bonded Stationary Phase

A simple high performance liquid chromatographic assay for the simultaneous quantitative analysis of seven ginsenosides, Rb1, Rb2, Rc, Rd, Re, Rf and Rg1 in commercial ginseng products is described. Chromatographic separation of the analytes was achieved in less than 20 min using a polyvinyl alcohol-bonded column with UV detection at 203 nm. Optimization of chromatographic conditions was determined by a three-factor central composite design, the variables being the percentage of acetonitrile in the mobile phase, column temperature and flow rate. A full quadratic model was found to be adequate in describing the separation of ginsenosides on the polyvinyl alcohol-bonded stationary phase. Complete separation of seven ginsenosides was achieved using acetonitrile–water (82.5/17.5) as the mobile phase run isocratically at a flow rate of 298 μL min^?1 and with the column temperature at 9 °C. The developed method was validated over the range of 10–120 μg mL^?1 using a 5 μL sample injection volume. Intra- and inter-day variation for three ginsenoside standards (Rf, Rd and Rb1) at three concentration levels ranged from 0.07 to 0.83% expressed as the relative standard deviation. The accuracy based on the nominal concentration values at three concentration levels was in the range 98.7–100.8%. The limit of detection was between 0.43 and 1.03 μg mL^?1 while the limit of quantification was from 1.42 to 3.13 μg mL^?1. The method is found to be applicable for the determination of ginsenosides in commercial ginseng products.     

Modeling of synergistic solvent extraction of Zr(IV) / HCl / TOA+Cyanex 921 system using statistical analysis and neural network

Multi-linear regression analysis (MLR), radial basis function (RBF) and multilayer perceptron (MLP) of artificial neural network (ANN) with five inputs (temperature, concentrations of HCl, TOA, Cyanex 921, Zr (IV) and percentage of extraction (%E)) as only output were employed for the construction of models. It was observed that ANN (RBF and MLP) performed better as compared to the MLR model. Based on the models proposed, the extraction of Zr(IV) could be predicted under variable experimental conditions of concentrations of HCl, TOA (Tri-n-octylamine), Cyanex 921 (Tri-n-octyl phosphineoxide), Zr(IV) and temperature. The nonlinear and complex relation between the percentage of extraction and operating variables have been determined using two and three layered feed forward neural network with back-propagation of error learning algorithm. Uncertainties in data have been determined in terms of statistical parameters such as root mean-squared error and R-squared values to check the efficiency of the model for prediction.