Predicting the origin of selectivity in NHC-catalyzed ring opening of formylcyclopropane: a theoretical investigation

Abstract

Using density functional theory, we investigated the origin of selectivity in the N-heterocyclic carbene (NHC)-catalyzed transformation of formylcyclopropane with an alkylidene oxindole. Our proposed mechanism includes the formation of a Breslow intermediate, ring opening of cyclopropane, formation of an enolate intermediate, formal [4 + 2] cycloaddition and disassociation of the catalyst. The transformation from an enolate intermediate into an alkylidene oxindole determines the chemoselectivity and stereoselectivity. Local reactivity analysis indicates that the chemoselectivity in the controlled transition state of the [4 + 2] cycloaddition is intrinsically favored by the higher reactivity of the C=C moiety of the alkylidene oxindole. Electron localization function analysis indicates that the NHC alleviates the electronic density between Cα and Cγ, and thus facilitates opening of the cyclopropane ring. This knowledge of NHC organocatalysis helps predict the chemoselectivity of divergent reactions involving cyclopropane rings.

---

Introduction

Donor–acceptor cyclopropanes (DACs), substituted with electron-withdrawing and electron-donating groups, are a unique class of reactive molecules.¹ DACs readily undergo ring opening because of this functional group synergy and thus generate an active intermediate that readily undergoes a wide range of reactions.²

Researchers commonly use Lewis acids to promote the ring opening of DACs, and apply DACs to enantioselective reactions through incorporation of chiral ligands.³ In general, the DAC acts as a three-carbon synthon, which undergoes formal [3 + n] (n = 2, 3, and 4) annulation reactions with numerous reactive partners under Lewis acid catalysis.⁴ Related density functional theory (DFT) investigations on the Lewis acid-catalyzed ring opening of DACs indicated that GaCl₃ promotes ring opening by changing the charge population (Scheme 1a).⁵ A Brønsted acid is used to promote the ring-opening of DACs by increasing the polarization of the C–C bond (Scheme 1b).⁶

Scheme 1 Representative strategies for the ring-opening of DACs.

Recently, N-heterocyclic carbenes (NHCs), which have been frequently employed as an organocatalyst in many transformations,⁷ have also been successfully applied to promote the ring opening of formylcyclopropane (Scheme 1c).⁸ However, the specific role of NHCs is unknown. Therefore, elucidating the role of NHCs is foundational to optimizing such ring-opening/transformation reactions.

Vicario et al. reported an NHC-catalyzed chemoselective and stereoselective ring opening and transformation reaction of formylcyclopropane.^8a This protocol provides a potential route for constructing diverse functional molecules (Scheme 2). We hypothesize that absorption of the NHC considerably affects the electronic density of the C–C bond in cyclopropane. Vicario et al. produced an ε-lactone product with excellent stereoselectivity (99% ee, Scheme 2); they did not observe the γ-lactone product obtained as the main species in other reports.⁹ Thus, we further hypothesize that the chemoselectivity depends on the reactivity of the interacting partners. To test our hypothesis, we used DFT, which has been commonly used to elucidate organocatalytic¹⁰ and organometallic¹¹ reaction mechanisms, to simulate the ring-opening and transformation reaction of formylcyclopropane R1 and alkylidene oxindole R2 reported by Vicario and co-workers. Our goal was to predict the origin of the chemoselectivity and stereoselectivity, and elucidate the role of NHCs, in the ring-opening and transformation reactions of formylcyclopropane.

Scheme 2 Possible activation and transformation pathways of cyclopropane under NHC organocatalysis.

Results and discussion

Reaction mechanism

Fig. 1 shows the energy profile of the entire NHC-catalyzed ring-opening/transformation reaction of formylcyclopropane with an alkylidene oxindole. As shown in Fig. 1, nucleophilic addition of the NHC to the carbonyl atom of the formyl group inititates the reaction. This process costs an energy of 10.6 kcal mol⁻¹ (TS1) to deliver intermediate M1. Then, a [1,2]-proton shift generates the Breslow-type intermediate M2. The energy barrier involved in the [1,2]-proton shift process is 20.7 kcal mol⁻¹, associated with transition state TS2, which was considerably lowered by implicating the in situ generated Brønsted acid.¹² Subsequently, the Breslow-type intermediate M2 quickly transforms into intermediate M3 via transition state TS3 through C–C bond cleavage. Then, γ-protonation gives enolate intermediate M4 via transition state TS4 through a [1,5]-proton shift, which costs an energy of 9.2 kcal mol⁻¹. Subsequently, ε-lactonization ([4 + 2] cycloaddition) between the formed enolate intermediate M4 and alkylidene oxindole R2 takes place via transition state TS5. We considered and calculated four possible diastereoselective pathways, which led to four diastereoisomers (the stereochemistry is provided in Scheme S1 in the ESI†). The SR-configured products are the most energetically favored, with an energy barrier of 6.0 kcal mol⁻¹ (the energy profiles of the other three pathways are provided in Fig. S1 in the ESI†). The last step of the catalytic cycle is desorption for catalyst turnover.

Fig. 1 Energy profile of the NHC-catalyzed DAC ring-opening/[4 + 2] cycloaddition reaction.

Alternatively, γ-lactonization, namely, the formal [2 + 2] cycloaddition, can occur between intermediate M4 and the C=O moiety of R2 to form the γ-lactone product. The energy barrier of this process is 22.3 kcal mol⁻¹ (TS5-CO, Fig. 1), which is much higher than that of ε-lactonization. Moreover, the Gibbs free energy of the [2 + 2] cycloproduct M5-CO is −1.6 kcal mol⁻¹ and this is 13.7 kcal mol⁻¹ higher than that of the [4 + 2] cycloproduct M5SR. These findings indicate that the [4 + 2] cycloaddition reaction is energetically favorable in terms of both kinetics and thermodynamics.

Origin of selectivities

We explored the mechanism of the NHC-catalyzed ring-opening/[4 + 2] cycloaddition reaction of formylcyclopropane with an alkylidene oxindole. Based on the computational results, we propose that the reaction between the enolate and alkylidene oxindole is the chemoselectivity- and stereoselectivity-determining step. The ε-lactonization reaction occurs prior to γ-lactonization and the SR-configured ε-lactone product formed predominately in ε-lactonization. As for the two reacting partners, the lower energy barrier of the corresponding transition state was mainly due to the higher nucleophilicity or electrophilicity.¹³ Therefore, we considered that the chemoselectivity could be modulated by the reactivity of the C=C–C=O and C=O moieties; a higher electrophilicity or nucleophilicity of the reacting center would lead to a lower energy barrier. Thus, we predicted the chemoselectivity by comparing the Parr functions of the nucleophilicity (P_k⁻) and electrophilicity (P_k⁺) of the two interactive partners (M4 and R2).

As shown in Table 1, the nucleophilic and electrophilic values of Cα are 0.64 and 0.13, respectively; the corresponding values of Cγ are 0.02 and 0.01, respectively. Thus, Cα is nucleophilic and this atom has a higher nucleophilicity than Cγ. The computational results also indicate that C1 and C3 in R2 are electrophilic. The reactivity of C1 (P_k⁺ = 0.30) is higher than that of the C3 atom (P_k⁺ = 0.09), revealing that the [4 + 2] cycloaddition occurs prior to the [2 + 2] cycloaddition. Therefore, we conclude that the greater electrophilicity and nucleophilicity lead to a lower energy barrier through local electrophilicity and nucleophilicity analysis, which agrees with our previous studies.¹⁴

Table 1 The local reactivity of enolate intermediate M4 and alkylidene oxindole R2

	R2				M4
	C1	C2	C3	O1	Cα	Cγ
P k ^−	0.14	−0.07	−0.04	0.09	0.64	0.02
P k ^+	0.30	0.15	0.09	0.12	0.13	0.01

---

In addition, some theoretical studies report that the electrostatic solvent–solute interaction is also responsible for the chemoselectivity.¹⁵ We tested whether the chemoselectivity is related to the intrinsic polarity of the corresponding transition state. Fig. 2 shows the electrostatic potential surfaces (EPSs) of the competing [4 + 2] and [2 + 2] cycloaddition transition states TS5SR and TS5-CO. In Fig. 2, blue represents the positively charged area, whereas red represents the negatively charged area. We found considerable charge separation in TS5-CO. Therefore, TS5-CO exhibits zwitterionic characteristics, which correspond to a larger dipole moment (μ = 13.99 D) and result in instability of the [2 + 2] cycloaddition transition state. In contrast, the [4 + 2] cycloaddition transition state TS5SR has a fairly neutral surface, which corresponds to a smaller dipole moment (μ = 2.38 D) and makes the [4 + 2] cycloaddition transition state stable. In order to clarify the effect of solvent polarity, we have additionally performed DFT calculations to re-compute the chemoselective [4 + 2] and [2 + 2] cyclization transition states in different solvents, including DMSO, cyclohexene, MeCN, toluene, and THF (the energy barriers in different solvents are summarized in Table S1 in the ESI†). The calculated results indicate that the main product always is the [4 + 2] cyclization product, so we can conclude that the solvent and its polarity should not be the main factor for controlling the chemoselectivity of the cyclization process.

Fig. 2 Electrostatic potential surfaces of the competing [4 + 2] and [2 + 2] cycloaddition transition states in dichloromethane with the NHC catalyst.

Next, we turn to disclosing the origin of the stereoselectivity. Non-covalent interaction (NCI) analysis, which has been widely used to qualitatively analyse the types of weak interactions between two interactive fragments, was performed to address this point. As can be seen from Fig. 3 (the other two NCI pictures of transition states TS5RR and TS5SS are shown in Fig. S2 in the ESI†), C–H⋯π weak interactions existed in both transition structures, with that in TS5SR found to be stronger. The C–H⋯O interaction made an almost equal contribution in both transition states. Therefore, the C–H⋯π interaction was key to determining the stereoselectivity. Overall, NCI analysis implied that the weak interactions existing in TS5SR were more predominant, with the C–H⋯π interaction contributing significantly to stabilizing the structure.

Fig. 3 NCI analysis of stereoselective transition states TS5RS and TS5SR (distance in Å).

Role of the catalyst

The literature indicates that NHC catalysts play a crucial role in opening three-membered rings. To understand the role of the NHC, we compared the electron density (∇²ρ), bond length (BL) and bond order (BD) between Cα and Cγ using electron localization function (ELF) analysis. The values of the electron density and bond order in formylcyclopropane R1 are 0.244 and 0.88, respectively; in M1, these are 0.228 and 0.89, respectively (Fig. 4). This is indicative of a strong σ-bond between Cα and Cγ. In Breslow intermediate M2, the values of the electron density and bond order are 0.193 and 0.77, respectively. This indicates that the Cα–Cγ σ-bond is weakened along with the Breslow intermediate formation and thus facilitates cleavage of the Cα–Cγ σ-bond. The electronic density in TS3 decreases to 0.009, indicating the C–C bond cleavage. Therefore, the NHC promotes the opening of the three membered-ring by alleviating the electron density between Cα and Cγ.

Fig. 4 ELF analysis graphs derived from the Laplacian electron density function (∇²ρ) in a selected plane.

Conclusions

We used DFT calculations to investigate the chemoselective and stereoselective mechanism of the NHC-catalyzed ring-opening/cycloaddition reaction of formylcyclopropane with an alkylidene oxindole. The reaction between the enolate intermediate and alkylidene oxindole is the chemoselectivity and stereoselectivity-determining step, and [4 + 2] cycloaddition occurs prior to [2 + 2] cycloaddition. The [2 + 2] cycloaddition pathway is unfavorable mainly because of the low reactivity of the C=O moiety, as also indicated by the electrostatic solvent–solute interaction. Our NCI analysis indicates that the stereoselectivity is mainly controlled by C–H⋯π and C–H⋯O interactions. ELF analysis indicates that the NHC catalyst promotes opening of the three membered-ring by alleviating the electronic density between Cα and Cγ, which would provide a new method to predict the ring-opening position of cyclopropane rings.

Experimental

Computational details

The Gaussian 09 program¹⁶ was employed to perform the theoretical study. The geometry optimization and frequency calculations were carried out by using the M06-2X¹⁷ functional with the 6-31G(d,p) basis set¹⁸ and accounting for the CH₂Cl₂ (DCM, ε = 8.93) solvent effect by employing the IEF-PCM¹⁹ solvation model (M06-2X/6-31G(d,p)//IEF-PCM_DCM, L1). Furthermore, single-point energy calculations were performed to refine the energies at the M06-2X-D3/6-311++G(2df,2pd)//IEF-PCMDCM level (L2). The free energies discussed in the text are obtained by the addition of Gibbs free energy correction calculated at L1 to the single-point energy calculated at L2. Moreover, the AIM and NCI analyses were plotted using Multiwfn.²⁰

Conflicts of interest

There are no conflicts to declare.

Acknowledgements

The authors gratefully thank the financial support from the National Natural Science Foundation of China (No. 21773214, 21772020, 21822303, and 21903072), the Outstanding Young Talent Research Fund of Zhengzhou University (No. 1521316001), the Startup Fund of Zhengzhou University of Light Industry (No. 2017BSJJ036), and the Henan Province Supercomputing Center. We also thank Michael Scott Long, from Liwen Bianji, Edanz Editing China (http://www.liwenbianji.cn/ac), for editing the English text of a draft of this manuscript.

Notes and references

1. For selected reviews: (a) C. A. Carson and M. A. Kerr, Chem. Soc. Rev., 2009, 38, 3051–3060; (b) H. K. Grover, M. R. Emmett and M. A. Kerr, Org. Biomol. Chem., 2015, 13, 655–671; (c) M. A. Cavitt, L. H. Phun and S. France, Chem. Soc. Rev., 2014, 43, 804–818; (d) P. Singh, R. K. Varshnaya, R. Dey and P. Banerjee, Adv. Synth. Catal., 2020, 362, 1447–1484; (e) O. A. Ivanova and I. V. Trushkov, Chem. Rec., 2019, 19, 2189–2208.
2. (a) H.-U. Reissig and R. Zimmer, Chem. Rev., 2003, 103, 1151–1196; (b) T. F. Schneider, J. Kaschel and D. B. Werz, Angew. Chem., Int. Ed., 2014, 53, 5504–5523; (c) E. M. Budynina, K. L. Ivanov, I. D. Sorokin and M. Y. Melnikov, Synthesis, 2017, 49, 3035–3068; (d) A. Kreft, P. G. Jones and D. B. Werz, Org. Lett., 2018, 20, 2059–2062; (e) O. A. Ivanova, A. O. Chagarovskiy, A. N. Shumsky, V. D. Krasnobrov, I. I. Levina and I. V. Trushkov, J. Org. Chem., 2018, 83, 543–560; (f) V. Pirenne, B. Muriel and J. Waser, Chem. Rev., 2020 DOI:10.1021/acs.chemrev.0c00109.
3. For selected examples: (a) Q.-K. Kang, L. Wang, Q.-J. Liu, J.-F. Li and Y. Tang, J. Am. Chem. Soc., 2015, 137, 14594–14597; (b) Y. Xia, L. L. Lin, F. Z. Chang, X. Fu, X. H. Liu and X. M. Feng, Angew. Chem., Int. Ed., 2015, 54, 13748–13752; (c) S. M. Wales, M. M. Morgan and J. S. Johnson, Org. Lett., 2013, 15, 2558–2561; (d) P.-J. Yang, L. Qi, Z. Liu, G. S. Yang and Z. Chai, J. Am. Chem. Soc., 2018, 140, 17211–17217; (e) F. Z. Chang, L. L. Lin, Y. Xia, H. Zhang, S. X. Dong, X. H. Liu and X. M. Feng, Angew. Chem., Int. Ed., 2018, 57, 2608–2612.
4. For selected examples: (a) A. U. Augustin, J. L. Merz, P. G. Jones, G. Mloston and D. B. Werz, Org. Lett., 2020, 22, 9405–9409; (b) L. K. B. Garve, A. Kreft, P. G. Jones and D. B. Werz, J. Org. Chem., 2017, 82, 9235–9242; (c) B. Q. Li, Z.-W. Qiu, A.-J. Ma, J.-B. Peng, N. Feng, J.-Y. Du, H.-P. Pan, X.-Z. Zhang and X.-T. Xu, Org. Lett., 2020, 22, 1903–1907; (d) M. A. Zotova, R. A. Novikov, E. V. Shulishov and Y. V. Tomilov, J. Org. Chem., 2018, 83, 8193–8207; (e) A. O. Chagarovskiy, V. S. Vasin, V. V. Kuznetsov, O. A. Ivanova, V. B. Rybakov, A. N. Shumsky, N. N. Makhova and I. V. Trushkov, Angew. Chem., Int. Ed., 2018, 57, 10338–10342; (f) R. Dey and P. Banerjee, Org. Lett., 2017, 19, 304–307; (g) A. U. Augustin, M. Sensse, P. G. Jones and D. B. Werz, Angew. Chem., Int. Ed., 2017, 56, 14293–14296.
5. P.-C. Tu, L. Zhou, A. M. Kirillov, R. Fang and L. Z. Yang, Org. Chem. Front., 2018, 5, 1702–1712.
6. (a) B. Cui, J. Ren and Z. W. Wang, J. Org. Chem., 2014, 79, 790–796; (b) E. Richmond, V. D. Vukovic and J. Moran, Org. Lett., 2018, 20, 574–577; (c) L. C. Irwin, C. R. Renwick and M. A. Kerr, J. Org. Chem., 2018, 83, 6235–6242; (d) Y. Z. Zhou, R.-C. Xue, Y. Q. Feng and L. Zhang, Asian J. Org. Chem., 2020, 9, 311–316.
7. For selected examples: (a) D. M. Flanigan, F. Romanov-Michailidis, N. A. White and T. Rovis, Chem. Rev., 2015, 115, 9307–9387; (b) X.-Y. Chen, S. Li, F. Vetica, M. Kumar and D. Enders, iScience, 2018, 2, 1–26; (c) Y. Wang, D. H. Wei and W. J. Zhang, ChemCatChem, 2018, 10, 338–360; (d) X.-Y. Chen, Q. Liu, P. Chauhan and D. Enders, Angew. Chem., Int. Ed., 2018, 57, 3862–3873; (e) S. J. Ryan, L. Candish and D. W. Lupton, Chem. Soc. Rev., 2013, 42, 4906–4917.
8. (a) L. Prieto, E. Sanchez-Diez, U. Uria, E. Reyes, L. Carrillo and J. L. Vicario, Adv. Synth. Catal., 2018, 359, 1678–1683; (b) S. S. Sohn and J. W. Bode, Angew. Chem., Int. Ed., 2006, 45, 6021–6024; (c) L. Candish and D. W. Lupton, Chem. Sci., 2012, 3, 380–383; (d) D. Du and Z. W. Wang, Eur. J. Org. Chem., 2008, 2008, 4949–4954; (e) L. X. Li, D. Du, J. Ren and Z. W. Wang, Eur. J. Org. Chem., 2011, 2011, 614–618; (f) J. W. Bode and S. S. Sohn, J. Am. Chem. Soc., 2007, 129, 13798–13799; (g) D. Du, L. X. Li and Z. W. Wang, J. Org. Chem., 2009, 74, 4379–4382; (h) G.-Q. Li, L.-X. Dai and S.-L. You, Org. Lett., 2009, 11, 1623–1625; (i) H. Lv, J. M. Mo, X. Q. Fang and Y. R. Chi, Org. Lett., 2011, 13, 5366–5369.
9. (a) A. Ametovski and D. W. Lupton, Org. Lett., 2019, 21, 1212–1215; (b) S. Mondal, S. Mukherjee, T. K. Das, R. Gonnade and A. T. Biju, ACS Catal., 2017, 7, 3995–3999.
10. (a) P.-C. Tu, L. Zhou, A. M. Kirillov, R. Fang and L. Z. Yang, Org. Chem. Front., 2018, 5, 1356–1365; (b) T. Liu, S. M. Han, Y. P. Li and S. W. Bi, J. Org. Chem., 2016, 81, 9775–9784; (c) P. Verma, P. A. Patni and R. B. Sunoj, J. Org. Chem., 2011, 76, 5606–5613; (d) X. X. Wei, R. Fang and L. Z. Yang, Catal. Sci. Technol., 2015, 5, 3352–3362; (e) Y. Wang and D. H. Wei, Mol. Catal., 2020, 489, 110944.
11. (a) R. Fang, L. Yang, L. Zhou, A. M. Kirillov and L. Z. Yang, Org. Lett., 2020, 22, 4043–4048; (b) Y. Liu, S. D. Feng and X. G. Bao, ChemCatChem, 2018, 10, 2817–2825; (c) J. Zhang, C. H. Shan, K. Lv, L. Zhu, Y. Y. Li, T. Liu and Y. Lan, ChemCatChem, 2020, 11, 1228–1237; (d) T. Zhang, X. J. Wu and J. Li, ChemCatChem, 2020, 12, 1385–1393; (e) B. Lian, L. Zhang and D.-C. Fang, Org. Chem. Front., 2019, 6, 2600–2606; (f) X. Y. Lv, X. T. Zhang, R. J. Sa, F. Huang and G. Lu, Org. Chem. Front., 2019, 6, 3629–3635.
12. (a) X. H. Wang, Y. Wang, J. S. Song and D. H. Wei, Org. Chem. Front., 2020, 7, 1113–1121; (b) Q. L. Liu, Y. Y. Wang, X. Li, L.-B. Qu, Y. Lan and D. H. Wei, Mol. Catal., 2020, 484, 110723; (c) Q. L. Liu, L. Sun, S.-J. Li, X. Li, L.-B. Qu, Y. Lan and D. H. Wei, Chem. – Asian J., 2019, 14, 2000–2007; (d) Y. Wang, L.-B. Qu, D. H. Wei and Y. Lan, J. Org. Chem., 2020, 85, 8437–8446; (e) X. Li, S.-J. Li, L.-B. Qu, Z. J. Li, Y. R. Chi, D. H. Wei and Y. Lan, Chem. Sci., 2020, 11, 7214–7225; (f) Y. Wang, L.-B. Qu and D. H. Wei, Chem. – Asian J., 2019, 14, 293–300; (g) J. X. Gao and Y. Wang, Org. Biomol. Chem., 2019, 17, 7442–7447; (h) Y. Wang and Y. Lan, Chin. Chem. Lett., 2020, 31, 736–738; (i) Y. Wang, L.-B. Qu, Y. Lan and D. H. Wei, ChemCatChem, 2019, 11, 2919–2925.
13. (a) Y. Wang, Q.-Y. Wu, T.-H. Lai, K.-J. Zheng, L.-B. Qu and D. H. Wei, Catal. Sci. Technol., 2019, 9, 465–476; (b) Y. Wang, L.-B. Qu, Y. Lan and D. H. Wei, ChemCatChem, 2020, 12, 1068–1074; (c) X. Li, R. H. Duan, Y. Y. Wang, L.-B. Qu, Z. J. Li and D. H. Wei, J. Org. Chem., 2019, 84, 6117–6125.
14. (a) L. J. Zheng, Y. Wang, D. H. Wei and Y. Qiao, Chem. – Asian J., 2016, 11, 3046–3054; (b) X. Li, M. S. Tang, Y. Y. Wang, Y. Wang, Z. J. Li, L.-B. Qu and D. H. Wei, Chem. – Asian J., 2018, 13, 1710–1718; (c) X. Li, Y. Y. Wang, Y. Wang, M. S. Tang, L.-B. Qu, Z. J. Li and D. H. Wei, J. Org. Chem., 2018, 83, 8543–8555.
15. (a) P.-P. Chen, H. Y. Zhang, B. Cheng, X. Chen, F. C. Cheng, S.-Q. Zhang, Z. Lu, F. K. Meng and X. Hong, ACS Catal., 2019, 9, 9589–9598; (b) Y. Wang, C. Du, Y. Y. Wang, X.-K. Guo, L. Fang, M.-P. Song, J.-L. Niu and D. H. Wei, Adv. Synth. Catal., 2018, 360, 2668–2677.
16. M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria, M. A. Robb, J. R. Cheeseman, G. Scalmani, V. Barone, B. Mennucci, G. A. Petersson, H. Nakatsuji, M. Caricato, X. Li, H. P. Hratchian, A. F. Izmaylov, J. Bloino, G. Zheng, J. L. Sonnenberg, M. Hada, M. Ehara, K. Toyota, R. Fukuda, J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda, O. Kitao, H. Nakai, T. Vreven, J. A. Montgomery Jr. , J. E. Peralta, F. Ogliaro, M. Bearpark, J. J. Heyd, E. Brothers, K. N. Kudin, V. N. Staroverov, R. Kobayashi, J. Normand, K. Raghavachari, A. Rendell, J. C. Burant, S. S. Iyengar, J. Tomasi, M. Cossi, N. Rega, J. M. Millam, M. Klene, J. E. Knox, J. B. Cross, V. Bakken, C. Adamo, J. Jaramillo, R. Gomperts, R. E. Stratmann, O. Yazyev, A. J. Austin, R. Cammi, C. Pomelli, J. W. Ochterski, R. L. Martin, K. Morokuma, V. G. Zakrzewski, G. A. Voth, P. Salvador, J. J. Dannenberg, S. Dapprich, A. D. Daniels, O. Farkas, J. B. Foresman, J. V. Ortiz, J. Cioslowski and D. J. Fox, Gaussian 09, Revision C.01, Gaussian, Inc., Wallingford, CT, 2010.
17. (a) Y. Zhao and D. G. Truhlar, Theor. Chem. Acc., 2008, 120, 215–241; (b) Y. Zhao and D. G. Truhlar, Acc. Chem. Res., 2008, 41, 157–167.
18. (a) W. J. Hehre, R. Ditchfield and J. A. Pople, J. Chem. Phys., 1972, 56, 2257–2261; (b) W. Sang-Aroon and V. Ruangpornvisuti, Int. J. Quantum Chem., 2007, 108, 1181–1188.
19. (a) B. Mennucci and J. Tomasi, J. Chem. Phys., 1997, 106, 5151–5158; (b) V. Barone and M. Cossi, J. Phys. Chem. A, 1998, 102, 1995–2001.
20. T. Lu and F. W. Chen, J. Comput. Chem., 2012, 33, 580–592.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/d0cy01768j

---