MOLCAS is an ab initio computational chemistry program package that is specialized in multiconfigurational wavefunction theory [101, 107]. The package has the CASSCF and restricted MCSCF wavefunctions restricted active space self‐consistent‐field (RASSCF) modules. The program can perform the geometries optimization for equilibrium and transition states and vibrational frequency calculations. Second‐order perturbation theory codes CASPT2 and RASPT2 were also implemented. A unique technique named Cholesky decomposition is implemented robustly and efficiently. This code is commercially available. Website: https://www.molcas.org/. It also has an open source version, OpenMolcas. Website: https://gitlab.com/Molcas/OpenMolcas.
PySCF is a quantum chemistry program and a lightweight and efficient platform for calculations and code development [108]. It uses the Python language to equilibrate the convenience in development and computational efficiency. The SCF module includes implementations of HF and DFT for restricted, unrestricted, closed‐shell, and open‐shell Slater determinant references. Post‐HF methods are also available, including Møller–Plesset second‐order perturbation theory, configuration interaction, and CC theory. Multiconfigurational calculations can be performed by an interface to external solvers. Relativistic correction can be added by relativistic effective basis set and other complexed methods. This computational chemistry code is fully open source available. Website: https://github.com/pyscf/pyscf.
2.8 The Limitation of Current Computational Methods
Although the application of DFT methods has greatly flourished computational organic and organometallic chemistry, there are still a series of shortcomings and challenges in this field, including the accuracy of DFT methods, exact solvation effect, evaluation of entropy effect, the computation of excited state and high spin state, speculation on the reaction mechanism, mechanism study in biological system, etc.
2.8.1 The Accuracy of DFT Methods
Over the past several decades, hundreds of DFT methods have been developed. Generally speaking, the accuracy has been improved according to the developed date of density functionals; however, their application scope and field are often restricted by their own profiles. In the field of computational organic and organometallic chemistry, the choice of density functional and basis set is often uncertain, which is more often based on past experience of experimenters. This results in computational data depending on the selection of density functional and basis, which may lead to inaccurate and incomparable computational results. The challenge now is to find a density functional that can be adapted to as many systems as possible and with sufficient precision.
2.8.2 Exact Solvation Effect
In computational organometallic chemistry, most of the computing systems are homogeneous catalysis in solution. However, the existing computational capability can only support the calculation method based on the implicit solvent model level. This model has a good effect in dealing with weak polar solvents, but still has a large error for polar solvent. Especially for protonic solvents, it is difficult to get satisfactory results from the implicit solvent model.
2.8.3 Evaluation of Entropy Effect
Entropy effect is very important for multimolecular reactions in organometallic chemistry. In the current computational studies, the entropy correction data often come from the calculation results in gas phase or in the implicit solvent model, which is not enough to reflect the influence of entropy in homogeneous solvent system. The influence of solvents on the entropy majorly restricted translational and rotational entropy and would make the actual free energy of a molecule between the calculated free energy and enthalpy. Sometimes, the inevitable calculation error of the entropy even affects the judgment of the reaction mechanism.
2.8.4 The Computation of Excited State and High Spin State
The d orbitals of transition metals are often occupied by unpaired electrons. Due to the limitation of principle, DFT method often exhibited a critical error in dealing with high spin states. On the other hand, photocatalytic processes involving transition metals currently became the focus of chemists, whose core steps would include the formation, transformation, and quenching of excited states. For this system, DFT calculation also cannot obtain reliable spectral results. In fact, some completely active space‐based methods can obtain high‐precision results for those systems, but due to the huge amount of calculation requirement, it is still difficult for the exploration of reaction mechanism in photocatalysis with transition metals and large ligands.
2.8.5 Speculation on the Reaction Mechanism
As mentioned in Section 2.6, the initial step of studying the reaction mechanism by computational chemistry is the hypotheses of possible pathways. The critical disadvantage of this approach is that the correctness of mechanism depends on the experience of researchers. If the correct pathway of one reaction is ignored, the true one cannot be obtained by theoretical calculation. Although some computational methods based on the full potential energy surface analysis are being applied to the study of reaction mechanism, the computational efficiency limits their application. Maybe artificial intelligence would be the way out for mechanism exploration in the near future. The possible mechanism of an unknown reaction can be obtained directly by analyzing and screening a large number of previous mechanism research information.
References
1 1 Cramer, C.J. (2002). Essential of Computational Chemistry: Theories and Models, 2nd Edition. New York: John Wiley & Sons.
2 2 Schrödinger, E. (1926). Quantisierung als Eigenwertproblem. Annals of Physics 384: 361–376.
3 3 Szabo, A. and Ostlund, N.S. (1996). Modern Quantum Chemistry, Introduction to Advanced Electronic Structure Theory. New York: Dover Publications.
4 4 Born, M. and Oppenheimer, R. (1927). Zur Quantentheorie der Molekeln. Annals of Physics 389: 457–484.
5 5 Hartree, D.R. (1928). The wave mechanics of an atom with a non‐Coulomb central field. Part I. Theory and methods. Mathematical Proceedings of the Cambridge Philosophical Society 24: 89.
6 6 Fock, V. (1930). Näherungsmethode zur Lösung des quantenmechanischen Mehrkörperproblems. Zeitschrift für Physik 61: 126.
7 7 Clark, T. and Koch, R. (1999). Linear combination of atomic orbitals. In: The Chemist's Electronic Book of Orbitals. Berlin, Heidelberg: Springer.
8 8 Roothaan, C.C.J. (1951). New developments in molecular orbital theory. Reviews of Modern Physics 23: 69.
9 9 Foresman, J.B., Head‐Gordon, M., Pople, J.A. et al. (1992). Toward a systematic molecular orbital theory for excited states. Journal of Chemical Physics 96 (1): 135–149.
10 10 Bartlett, R.J. and Purvis, G.D. (1978). Many‐body perturbation‐theory, coupled‐pair many‐electron theory, and importance of quadruple excitations for correlation problem. International Journal of Quantum Chemistry 14: 561–581.
11 11 Møller, C. and Plesset, M.S. (1934). Note on an approximation treatment for many‐electron systems. Physical Review 46: 618–622.
12 12 Pople, J.A., Binkley, J.S., Seeger, R. et al. (1976). Theoretical models incorporating electron correlation. International Journal of Quantum ChemistrySupply 10: 1–19.
13 13 Pople, J.A., Seeger, R., Krishnan, R. et al. (1977). Variational configuration interaction methods and comparison with perturbation theory. International Journal of Quantum Chemistry 12 (Suppl 11): 149–163.
14 14 Head‐Gordon, M., Pople, J.A., Frisch, M.J. et al. (1988). MP2 energy evaluation by direct methods. Chemical Physics Letters 153: 503–506.
15 15 Raghavachari, K. and Pople, J.A. (1978). Approximate