化学键是化学科学中的重要概念。多重键作为化学键的重要形式之一，在有机、无机等化学领域均有广泛的应用。其中数量众多的含主族和过渡金属元素组成的多重键化合物已为化学界所熟知，但f-区元素（镧系和锕系）组成的多重键化合物尚甚少报道。因核能和放射化学等领域的发展，目前已经积累了大量关于f-区元素化学性质的实验数据，亟需从理论上进行分析和解释。但由于重元素的强电子相关作用和相对论效应，镧系和锕系化合物的理论研究进展缓慢。本论文运用相对论量子化学方法, 研究了f-区锕系(An)、镧系(Ln)元素与第IVA族C、Si等元素形成多重键化合物的可能性，揭示了重元素多重键化合物的成键规律和物理化学性质。运用含相对论校正的密度泛函理论(DFT)和从头算多组态Post-HF方法(如CCSD(T), CASSCF和CASPT2等)，本文首先研究了An或Ln与卤代甲烷和卤代硅甲烷反应所得的产物的最稳定结构和电子态。发现An + H2CXY (An = U, Th；X, Y= F, Cl)反应的最稳定产物包含C=An双键的具有C1对称性H2C=AnXY型化合物。它们因锥角化和抓氢键的形成而呈奇特的手性结构。在三或四卤代甲烷与U原子的反应产物中，通过理论计算结果与实验数据的对比首次发现了包含碳–铀三重键的锕系金属有机分子HC≡UX3 (X, Y = F, Cl, Br, I)，并揭示了该三重键的形成和稳定化的必要条件。对四种其它HCAnF3（An = Np, Pu, Am 和 Cm）化合物的研究发现，C-An键级随着原子序的增加逐渐下降而最后变为单键。对C-U键长最短的CUO分子全面的理论分析表明，在该分子中存在介于于三重和四重之间的碳-铀多重键。这是由未杂化的三个C2p原子轨道组成（C2s轨道部分参与）的一种新型成键方式。最后，将对f-区元素多重键化合物的考察扩展到了Si-An化合物和镧系C-Ln（Ln = La- Lu）化合物，发现了一类全新的包含•C-Ln•单键的多自由基化合物，并提出了双自由基卡宾的新概念。这类H2CLnF2分子中，电子定域在C和相邻的Ln原子上而未形成化学键，打破了传统的Lewis电子对理论和八电子规则。理论计算同样证实在XSi-AnF3 (An = U, Th) 体系中，我们发现Si-An之间没有形成期待的多重键，而是形成了X桥连的高自旋多重度的Si(u-X)–AnF3（X = H, F）分子，它们含罕见的三重态硅宾基团。
One of the most essential concepts in chemistry is chemical bonding. Multiple bonding is an important variant with versatile applications in inorganic and organic chemistry. While multiply bonded main-group and transition-metal compounds are abundant, multiply bonded lanthanoid (Ln) and actinoid (An) compounds are rather rare. With the development of electronic devices and nuclear energy systems, a large number of data were collected from experimental lanthanoid and actinoid chemistry, which require a deeper theoretical understanding. Therefore, we have performed theoretical investigations of multiple bonding of f-elements such as selected lanthanoids and actinoids, and of main- group 14 elements such as carbon and silicon. Complicated electron correlation effects in the bonding shells and relativistic effects significant in the valence shells of heavy atoms are accounted for.Both approximate density functional theory (DFT) and ab initio post-Hartree-Fock methods such as CCSD(T), CASSCF, and CASPT2 are exploited. We have searched for final products for unusual reactions of Ln or An atoms with molecules of type CXYZR or SiX3R, where X, Y, Z, R means F, Cl, Br, I or H. Starting from U + H2CXY, we predict the double bonded H2C=UXY molecular product with unexpected chirality, agosticity and pyramidality. In addition, we find unprecedented triply bonded RC≡UX3 uranium methylidyne molecules. Another discovery is the multiply bonded CUO molecule with an extremely short C-U distance, corresponding to an intermediate between triply bonded C≡U≡O and novel quadruply bonded C≣U=O, due to the interaction of three unhybridized C2p atomic orbitals (and part of C2s) with the various U atomic orbitals of 6p, 5f, 6d and 7s type.We have extended our investigations to a whole series of actinoid compounds, HC-AnF3, An = U, Np, Pu, Am and Cm, where the C-An bond becomes weaker and weaker. The extension to lanthanoid compounds such as H2C-CeF2 led to the discovery of another new interesting type of bond, namely the biradical •C-Ce• single bond, violating the Lewis electron-pair theory and the Octet rule for carbon atoms. Furthermore, the discovery of this type of carbene complex is an addition to the Fischer or Schrock types. Other H2C-LnF2 species exhibit similar multi-radical bonding. When replacing the C by a Si atom, we obtain new X-bridged structures with triplet or quintet spin-multiplicity in Si-(u-X)-AnF3 (X = H, F; An = Th, U) molecules. Multi-radical Si–An single bonds are another example for the breakdown of Lewis electron-pair theory. Interestingly, there are two unpaired electrons localized on the silylene group, yielding the very rare triplet state of silylene.We have shown in this dissertation, how quantum chemical approaches can pave the way into new fields of chemistry with challenging outlooks for the synthetic chemists.
