拓扑量子方法估算液相链烷烃导热率

用烷烃分子C—C键和C—H键键邻接矩阵特征根衡量分子中价电子被束缚的强弱程度, 以该参数和烷烃的支化参数以及温度为变量, 对含有5～24个碳原子的液相烷烃在较大温度范围内的导热率建立一个4参数拟合方程: λ=0.7568?0.2728(ΣX1CC)/NC—C+1.5171(ΣX1CH)/NC—H+7.4×10?5ΣSij+2.0966T ?0.4 该模型用于导热率估算, 其标准偏差仅为0.0033 W?m?1?K?1, 相对平均误差仅为2.11%. 用上述方程对拟合集以外的9个烷烃分子的导热率进行预测, 其预测的相对平均误差仅为1.64%, 该预测精度在实验误差范围内. 因此, 此方程可用于化工设计中估算还未经实验测定的液相烷烃的导热率.

A Topological-Quantum Method for the Estimation of the Thermal Conductivity of Liquid Alkanes

The strength that the nucleus attracts the valence electrons in alkane molecules can be scaled by the eigenvalues of the bond orbital-connection matrices of C—C and C—H bonds. Using the above two kinds of eigenvalues, the branching parameter of alkane molecules and temperature parameter, a four-parameter equation was developed to fit the thermal conductivity λ of a training set with 45 alkanes bearing 5 to 24 carbon atoms: λ=0.7568?0.2728(ΣX1CC)/NC—C+1.5171(ΣX1CH)/NC—H+7.4×10?5ΣSij+2.0966T ?0.4 The standard error between the estimation values by above equation and the experimental values is only 0.0033 W?m?1?K?1 and the relative average error is only 2.11%. Further, the thermal conductivities of 9 another alkanes excluded in the training set were predicted with the obtained model, and the predicted relative average error is only 1.64%, which is within the experimental uncertainties. Therefore it is expected that this model can be used to estimate and predict the thermal conductivity of liquid alkanes whose λ values have not been measured by experiments.