借助密度泛函理论采用B3LYP/6-31G(d)方法对209种多溴联苯醚（PBDEs）同系物进行了结构优化和频率计算, 利用已知的PBDEs结构和频率参数值对计算结果进行验证, 平均模拟系数为0.989。 在此基础上, 选取苯环上只有H原子或Br取代基且红外光谱各振动形式间干扰较少的二苯醚和BDE-209为代表, 分别对PBDEs的苯环振动和C—Br振动进行归属。研究发现苯环三角呼吸振动具有普遍性、 强振动性和可区分性, 而红外光谱中C—Br振动常与其他振动形式相耦合不易区分, 故选取苯环三角呼吸振动作为PBDEs同系物的特征振动, 以分析其振动频率与溴取代基之间的规律。 研究结果表明: 邻对位取代基可显著提高苯环三角呼吸振动频率, 间位取代基作用相反; 低溴代物中邻位取代基越多特征振动频率越高, 而高溴代物中同时存在邻对位取代基的同系物频率较高; 取代类型相同时, 位于同一苯环取代基数越多的同系物振动频率较高。 利用SVM回归模型对溴取代基与苯环三角呼吸振动频率间变化规律进行模拟, 所建模型的模拟效率系数为0.956, 说明所建分析方法可为红外光谱检测和辨别PBDEs同系物提供一定的理论基础。

Structure optimization and frequency calculations of 209 kinds of PBDEs homologues were carried out at the B3LYP/6-31G(d) level based on density functional theory, the known experimental parameters of 4 PBDEs congeners were compared with the calculated values, and the average analog coefficient was 0.989. The main vibration of DE and BDE-209 which contained only H atoms or Br substituent in the benzene ring was assigned due to their less interference for various infrared spectrum vibrations. C-Br bond vibration which was often coupled with other forms of vibration was not easy to distinguish between PBDEs congeners. With the characteristics of universality, strong vibration and identifiable, triangular breathing vibration of the benzene rings in the infrared spectrum was selected as a characteristic vibration among different PBDEs homologues, the analysis results showed that the ortho-substituted and para-substituted Br could significantly improve the vibration frequency of triangular breathing vibration, while the meta-substituted Br played the opposite role; For the low bromide, the more the substituents in ortho position, the higher the frequency, while for the high bromide homologues, simultaneous existence in ortho-position and para-position would have a higher frequency; With the same substituted location, the more the substituent in the same benzene ring, the higher the vibration frequency. SVM regression model was used to verify the relative relationship between Br substituent and triangle breathing vibration frequency of the benzene rings, and the simulation efficiency coefficient of model was 0.956. The study provided a theoretical basis and guidance for actually testing and identifying PBDEs homologues in the environment.