冰毒作为一种新型毒品, 近年来泛滥速度不断加快, 社会危害日益突出, 对相关监管部门带来了严峻的挑战, 如何能够提供一种无损、 快速、 实时与准确的冰毒检测方法具有重要的实际意义和应用价值。 拉曼光谱是一种较为符合上述要求的新颖方法, 但由于冰毒分子构象上的不同会导致其拉曼光谱存在差异, 进而对冰毒现场实时的检测造成影响, 甚至产生误判, 并对毒品拉曼数据库的建立带来非常大的困难。 因此, 根据密度泛函理论, 采用Becke-3-Lee-Yang-Parr (B3LYP)杂化泛函方法在6-31G基组下分别以冰毒分子ϕ1, ϕ2和ϕ3三个二面角为变量, 在0°~360°范围内, 以10°为步长做了势能面扫描, 取能量的极小值点从而获得12个不同的冰毒分子稳定构象, 在此基础上, 采用B3LYP方法6-31G++(d, p)基组选取其中4种能量较低的构象做进一步优化以及振动频率的计算, 并将最终得到的计算拉曼光谱与实验光谱进行比较, 结果表明: 构象上的差异使计算得到的冰毒拉曼光谱中746, 837和1 356 cm^-1处的特征峰位置产生了不同程度的偏移, 而632, 1 003, 1 180和1 312 cm^-1处的特征峰则基本不受构象影响, 因此, 在对可疑样品进行识别时上述不受构象影响的特征峰可以作为冰毒主要的特征峰来进行快速匹配, 而这种通过关键特征峰进行匹配的方法相比于传统的相关系数匹配算法显然更为快捷; 同时在选取的4种能量较低的构象中, 由构象Ⅸ计算得到的结果与实验值最为吻合, 于是在文中采用该结果并结合各振动频率的势能分布以及相关文献对冰毒实验拉曼特征峰进行了详细的归属。 其中, 1 003 cm^-1为冰毒拉曼最强特征峰, 归属为苯环的环呼吸振动; 837 cm^-1处特征峰归属为NH摇摆振动; 此外, 1 180 cm^-1处的特征峰归属为C—N伸缩振动; 1 312 cm^-1处则归属为CH₂摇摆振动。 上述这些研究结果将为今后的毒品检测、 毒品拉曼数据库的建立以及药物分子拉曼光谱理论计算研究提供有益的参考。

As a new-type drug, methamphetamine has been spreading rapidly in recent years and its social harm has become increasingly serious, which has brought severe challenges to the relevant regulatory authorities. How to provide a non-destructive, fast and accurate drug detection method has important practical and application value. Based on the above conditions, Raman spectroscopy is a novel method. However, due to the differences of methamphetamine molecules, the Raman spectra detected are different, which affect the on-site methamphetamine detection, and even cause misjudgment, and bring great difficulties to the establishment of the database of Raman spectra of methamphetamine. Therefore, according to the density functional theory, the Becke-3-Lee-Yang-Parr (B3LYP) hybrid functional method was used to determine the three dihedral angles of the methamphetamine molecules ϕ1, ϕ2 and ϕ3 on the 6-31G basis. In the range of 0°~360°, the potential energy surface scanning was performed in steps of 10° respectively, and 12 different stable conformations of methamphetamine molecule were obtained based on minimal energy points. Besides those, four lower energy conformations were further optimized and their vibration frequencies were calculated on the 6-31G++(d, p) basis set. Finally the theoretical Raman spectra obtained were compared with the experimental spectra. The results showed that the differences of methamphetamine conformations produced various shifts on peak positions at 746, 837 and 1 356 cm^-1 in Raman spectra, however, the Raman peaks positions at 632, 1 003, 1 180 and 1 312 cm^-1 were basically unaffected. Therefore, when we identify suspicious samples, those unaffected peaks could be used as the identifier spectra of methamphetamine, and the matching method by identifier Raman spectra is obviously faster than most traditional correlation coefficient matching algorithms. Meanwhile, in the four lower energy conformations selected, the calculated results from the conformation IX were the closest to the experimental values. Therefore, in order to assign the experimental Raman identifier spectra of methamphetamine, the IX’ result was combined with the potential energy distribution of each vibration frequency and related literature. In those results, 1 003 cm^-1 was the strongest identifier peak of methamphetamine, which was assigned to the aromatic respiration vibration. The Raman peak at 837 cm^-1 was assigned to NH rocking vibration. In addition, the Raman peak at 1 180 cm^-1 was attributed to CN stretching vibration and 1 312 cm^-1 belonged to CH₂ wagging vibration. These researches have the potential to provide useful references for drug detection, database of drug Raman spectra establishment and theoretical calculation of Raman spectroscopy of drug molecular in future.