多环芳烃活性参数与色谱保留指数相关性分析

为了简化多环芳烃(PAHs)的Koc、Kow和BCF等活性参数的测定和预测工作,通过对PAHs的活性参数与气相色谱保留指数(I)、苯环个数(N)的相关关系分析,分别建立了I与lgKoc、lgKow和lgBCF,N与lgKoc、lgKow和lgBCF间的2类一元线性回归预测方程,以及I、N与lgKoc、lgKow和lgBCF间的二元线性回归预测方程。结果表明,PAHs类化合物的I值或N值分别与lgKoc、lgKow和lgBCF存在明显的线性相关性,均能对上述3种活性参数进行准确预测,其中,用I值建立的一元回归方程的预测精度更好;而利用I值和N值共同建立的二元预测方程经t检验发现,方程中变量N的偏回归系数的统计检验没有显著意义。本研究利用I值建立的一元线性预测模型能较好地预测PAHs类化合物的Koc、Kow和BCF等活性参数,从而为PAHs的活性参数预测提供了一种简便易行的方法。

Correlation Analysis on Active Parameters and Retention Index of Polycyclic Aromatic Hydrocarbons

To simplify the determination and prediction of Koc, Kow and BCF of polycyclic aromatic hydrocarbons (PAHs), the correlations of activity parameters with retention index (I) and benzene ring number (N) were analyzed in this paper. Unary linear regression equations including I (or N) with lgKoc, lgKow, lgBCF, and binary linear regression equations of I and N with lgKoc, lgKow, lgBCF were established, respectively. Results showed that a clear linear relationship was observed between I (or N) and lgKoc, lgKow, lgBCF. The established equations could accurately predict the three active parameters, and the linear regression equation of I showed a higher predicting ability. The correlation coefficients of binary linear regression equation of I and N with lgKoc, lgKow, lgBCF were analyzed with t-test, which revealed that the partial regression coefficient of variable N had no statistic significance. It was demonstrated that lgKoc, lgKow and lgBCF of PAHs could be well predicted based on the linear prediction model of I, providing a simple method for predicting the activity parameters of PAHs compounds.