Comparative density functional theory study of the binding of ligands to Cu+ and Cu2+:: Influence of the coordination and oxidation state
Comparative density functional theory study of the binding of ligands to Cu+ and Cu2+:: Influence of the coordination and oxidation state
复制标题
DOI：
10.1021/jp047971b
复制
发表时间：
2005-01-20
影响因子：
2.9
通讯作者：
Berthomieu, D
中科院分区：
文献类型：
--
作者：
Ducéré, JM;Goursot, A;Berthomieu, D
比较了BP 86、B3 LYP和MP2方法对铜配合物键离解能的准确计算能力。研究了L为小配体的[Cu-L]（+）和[Cu-L]（2+）配合物以及高配位的[Cu（NH 3）（4）]（+）和[Cu（NH 3）（4）]（2+）配合物。对于单配位络合物，三种方法计算的溴化二苯醚相差2至60千卡/摩尔，[Cu-L]（2+）络合物的差异较大。使用B3 LYP泛函计算的BDE值一般接近实验值，而使用BP 86泛函计算的BDE值过高，使用MP2计算的BDE值过低。如果我们根据它们增加的键强度对整个配体进行排序，则对于[Cu-L]（+）络合物，用三种方法得到的结果顺序是不同的，B3 LYP给出的顺序与实验顺序相同。这一结果表明，使用B3 LYP比使用BP 86和MP2方法更好地模拟[Cu-L]（+）络合物的BDE。对于[Cu-L]（2+），B3 LYP也给出了最可靠的结果，而BP 86给出的BDE太大，MP2给出的BDE太小。然而，使用DFT和后Hartree-Fock方法，基态的对称性可以不同。对于[Cu-N2 O]（2+），B1 LYP的使用提供了比B3 LYP更好的络合物的对称性，正如最近在文献中对[CuH 2 O]（2+）所示的那样。MP2导致[Cu-N-2]（2+）的不正确的弯曲结构，而用其他方法（包括CCSD（T））获得的是线性结构。然而，由于缺乏[Cu-L]（2+）配合物的实验数据和对比的方法的结果，这是不可能肯定地得出结论。对于高配位的配合物[Cu（NH3）（4）]（+）和[Cu（NH3）（4）]（2+），除了BP 86，B3 LYP和MP2外，还使用了PBE计算方法。当反应过程中氧化态没有变化时，BDE值非常接近。基于这些计算，方法的选择对于高配位络合物[Cu（NH3）（4）]（+）和[Cu（NH3）（4）]（2+）不太重要，只要氧化态在反应期间保持相同即可。相反，当[Cu（NH3）（4）]（2+）在[Cu（NH3）（3）]（+）和NH3中被还原时，使用四种方法计算的BDE显著不同。
BP86, B3LYP and MP2 methods, generally used to study large systems containing transition metals, were compared for their ability to accuratly evaluate bond dissociation energies of copper complexes. Various [Cu-L](+) and [Cu-L](2+) complexes in which L are small ligands and the higher coordinated complexes, [Cu(NH3)(4)](+) and [Cu(NH3)(4)](2+) were studied. For monoligated complexes, the BDEs calculated by the three methods differed by 2 to 60 kcal/mol, the larger differences being obtained for [Cu-L](2+) complexes. The BDEs calculated using the B3LYP functional were in general close to the experimental values whereas the BDEs calculated using the BP86 functional were too high and the BDEs calculated using the MP2 were too low. If we rank the whole ligands according to their increased bond strength, the resulting orders obtained with the three methods are different for the [Cu-L](+) complexes, the B3LYP giving the same order as the experimental one. This result indicates that the BDEs of [Cu-L](+) complexes are better modeled using the B3LYP than using the BP86 and MP2 methods. For [Cu-L](2+), B3LYP also gave the most reliable results whereas BP86 gave too large BDEs and MP2 gave too small BDEs. However, symmetries of ground states can be different using DFT and post-Hartree-Fock methods. For [Cu-N2O](2+) the use of the B1LYP provides a better symmetry of the complex than the B3LYP, as has been recently shown in the literature for [CuH2O](2+). MP2 led to an incorrect bent structure for [Cu-N-2](2+) in contrast to a linear structure obtained with the other methods, including CCSD(T). However, due to the lack of experimental data for [Cu-L](2+) complexes and to contrasted results for the methods, it is not possible to conclude definitely. For the high coordinated complexes [Cu(NH3)(4)](+) and [Cu(NH3)(4)](2+), the PBE calculation method was used in addition to the BP86, B3LYP and MP2. The BDE values were very close to each other when there is no change of the oxidation state during the reaction. On the basis of these calculations, the choice of the method was less crucial for high coordinated complexes [Cu(NH3)(4)](+) and [Cu(NH3)(4)](2+) so long as the oxidation state remained the same during the reaction. In contrast, when [Cu(NH3)(4)](2+) is reduced in [Cu(NH3)(3)](+) and NH3, the BDE calculated using the four methods were markedly different.