摘要
采用AM1半经验算法,计算了多氯联苯(PCBs)所有209种可能分子结构的6种量子化学参数——分子平均极化率(α)、偶极距(μ)、分子最高占据轨道能(E_(homo))、分子最低未占据轨道能(E_(lumo))、分子中氢原子的最正形式电荷(qH^+)和分子中原子的最负形式电荷(q^-);在此基础上依据理论线性溶解能相关模型(TLSER),通过正向逐步多元线性回归方法建立了PCBs气相色谱保留值参数(RRT)和两种理化性质——水溶解度(-logSw)、正辛醇/水分配系数(logK_(ow))的定量预测方程。检验结果表明;方程相关性显著,精度较高,物理意义明确。
Six quantum chemical parameters were calculated using AM1 semiempirical quantum method for all 209 possible patterns of poly-chlorinated biphenyls (PCBs): molecular average polarizability (α), dipole moment (μ), energy of the highest occupied molecular orbit (Ehomo), energy of the lowest unoccupied molecular orbit (Elumo), the most positive charge of hydrogen atom (qH+) , and the most negative charge (q - ) in the molecule. With these data, three quantitative structure-property relationship (QSPR) equations of PCBs were established based on the theory linear solving energy relationship (TLSER) modei by fonvard stepwise multiple regression methodology, which involves the relative retention time and two physical-chemical properties: solubility( - logSw) , n-octanol/water partition coefficient(logKow) individually. These equations have high signiflcance and low Standard error Ievels, also apparent physical meanings.
出处
《计算机与应用化学》
CAS
CSCD
北大核心
2002年第6期755-758,共4页
Computers and Applied Chemistry
基金
国家重点基础研究专项经费资助(G1999045711)