Theoretical investigation of boron-doped lithium clusters, BLi_n (n = 3–6), activating CO₂: an example of the carboxylation of C–H bonds

Abstract

This work reports the first example of boron-doped lithium clusters, BLi_n, activating CO₂. The investigation shows that a kind of novel BLi_n–CO₂ (n = 2–6) complex (I_n) with bidentate double oxygen M(η₂-O₂C) coordination is obtained through the bimetal 2Li of BLi_n clusters binding to the two O atoms of CO₂, and the structural integrities of BLi_n clusters are not destroyed. We find that the partial negative charge transfer from BLi_n to the π* orbital of CO₂ leads to weakening of the C=O bonds of CO₂ and an active CO₂ moiety, except for when n = 2. Further, we perform reactions between I_n (n = 3–6) and benzene to elucidate a novel alternative approach to direct carboxylation by inserting CO₂ into C–H bonds. We find that the carboxylation of the C–H bond of benzene can be achieved through the transition states (TS) of C–C bond formation and H-atom-transfer from C to O via two H₂O molecules acting as a H-transfer tunnel. Comparing the transition states of H-direct transfer and one H₂O molecule assisting H transfer, two H₂O molecules assisting H-transfer as a tunnel is shown to lower the barrier, due to this long H-bond bridge effectively easing the rotation of the dihedral angle between the C₆H₆ and CO₂ moiety planes. Considering the whole free energy profile, BLi₅ and BLi₆ clusters are more feasible for the carboxylation of C–H bonds.

---

1. Introduction

Seeking efficient methods of recycling greenhouse gas CO₂ from industrial emissions and of removing some of it, is a major challenge in catalysis research and has been the subject of many investigations. In principle, CO₂ is an ideal and attractive chemical feedstock and may provide access to high-value products, since it is a nontoxic, renewable, and low-cost resource.^1–3 However, only a small amount of CO₂ on Earth is currently being used for producing useful chemicals, because of its thermodynamic stability and the highest oxidative state of carbon. Therefore, the activation of CO₂ is at the heart of improving the feasibility of using CO₂ as carbon feedstock in chemical synthesis. Recently, assistance from transition metals is a very popular approach for CO₂ activation, which achieves C–C bond formation transformation. In nearly all cases, late transition metals (d⁸–d¹⁰) were used, since they are capable of binding a weak ligand such as CO₂ through backbonding, due to their basic behavior. The research groups of Inoue, Hoberg and Walther performed the pioneering studies, which proved that the [Ni(CO₂)] complex is very active toward olefins and alkynes.^4–7 Recently, Martin's group performed the Ni-catalyzed carboxylations of unactivated primary alkyl bromides and benzyl halides with CO₂ at atmospheric pressure.⁸ Sakaki et al. theoretically investigated that the formation of a Ni(I) species is crucial for Ni(0)-catalyzed carboxylation reaction of aryl chloride with CO₂.⁹ The motif of Pd catalysis is also widely employed to investigate the carboxylation of substrates with CO₂.^10–14 For example, Pd-catalyzed direct carboxylation of alkenyl C–H bond with CO₂ had been realized in 2013,¹⁵ which is a significant example to direct carboxylation of aryl bromides with CO₂ by Pd catalysis in the homogenous condition. For other transition metals, for example, Au, Ag, Cu, etc.^16–22 Recent investigations show that they play essential roles to facilitate high efficient transformations of CO₂. Although tremendous achievements have been made for the transformation of CO₂ via transition-metal-catalyzed reactions, no-transition-metal-catalyzed direct carboxylation reaction has not been reported.

Studies on clusters are of considerable interest in research due to their potential application in catalytic processes, chemisorptions, materials science. In particular, recently, cluster-catalyzed oxidation of CO to CO₂ has attracted more and more attentions. Bernhardt investigated the mechanism and energetic of the low-temperature oxidation of CO to CO₂ catalyzed by small free palladium clusters Pd^x+ (x = 2–7).²³ The CO oxidation on subnanometer gold clusters (Au₁₆–Au₁₈, Au₂₀, and Au₂₇–Au₃₅) has been systematically studied by Zeng' group.^24–26 Ramasubramaniam theoretically verified the catalytic activity of ultrasmall Pt nanoclusters through CO oxidation reaction.²⁷ These studies show that clusters indeed make CO be active through electron back-donation to the CO antibonding orbital. However, the activation of CO₂ mediated by clusters is regrettably scarce.²⁸

In this work, in order to explore cluster assisting the activation of CO₂, we selected boron-doped lithium clusters BLi_n (n = 2–6) to activate CO₂ and further investigate the transformation of it. The lithium–boron system has been extensively studied,^29–31 because of the practical importance of Li–B alloys as anode materials for the production of lithium batteries. Moreover, the advantages of complex I_n were concerned. First, in the complex I_n, BLi_n exhibits super-alkali atom characteristics, which provides the large charge transfer into the π* orbital of CO₂, and the structural integrities of BLi_n cluster. Second, BLi_n cluster is a representative mononuclear cluster, in which the nuclear B atom is trapped into the Li alliance. The BLi_n–CO₂ complex with a bidentate double oxygen M(η₂-O₂C) coordination provides a novel coordination modes to activate CO₂. Third, BLi_n cluster only contains the main group metal Li atoms and the nonmetal B atom, and does not contain any transition metals.

A key piece of information in metal-catalyzed CO₂ reactions is the large reactivity of metal–CO₂ complexes. Generally, the coordination of CO₂ to the metals which makes CO₂ be active has several basic modes. A single metal atom monodentate coordination M(η₁-CO₂) or bidentate coordination M(η₂-CO₂) are most common, corresponding to structures (A) and (B) in Scheme 1, respectively.^32–34 A bidentate double oxygen M(η₂-O₂C) coordination is also seen in the anionic monometal Mg complexes [XMgCO₂] (X = Cl, Br, OH)^35. Here, we present novel BLi_n–CO₂ complexes with a bidentate double oxygen M(η₂-O₂C) coordination (C), where bimetal 2Li of BLi_n clusters bind to O of CO₂, to activate CO₂ (the mode D in Scheme 1). Further, we theoretically calculated the reactions between the BLi_n–CO₂ complexes and benzene molecule, to elucidate the role of BLi_n clusters in the CO₂ transformation. Although this work is a theoretical design, I_n provides a idea theoretical model to study the novel mechanism of activation CO₂. We anticipate that our results guide the catalyst community to discover additional cluster catalyst in the transformation of CO₂.

Scheme 1 Several metal–CO₂ coordination modes.

2. Computational details

All the calculations were performed with the GAUSSIAN 09 programs.³⁶ The geometrical optimizations of the reactants, intermediates, transition states (TS) and products were performed at the density functional exchange correlation (B3LYP) method^37,38 with the augmented correlation consistent basis set 6-311+G(d, p), and frequency calculations at the same level were performed to confirm stationary points as minima (zero imaginary frequencies) or transition states (one imaginary frequency). Further, the single point energies were obtained at the MP2/6-311+G(d, P) level based on the optimized geometries. For open-shell systems of n = 4 and 6, the restricted open-shell MP2 (ROMP2) calculations were carried out. The zero-point energies are evaluated within the harmonic potential approximation at optimized structures. In several cases where TS are not easily confirmed by animation of their vibrations, intrinsic reaction coordinate (IRC)^39,40 calculations were performed to confirm the connection of each TS to its corresponding reactant and product. Natural bond orbital (NBO) analyses^41–43 at the MP2/6-311+G(d) were carried out to understand the charge transfer of complexes.

3. Results and discussions

3.1 Activation of CO₂ by assisting of clusters BLi_n (n = 3–6)

The activation of CO₂ is a key point for the successful transformation of CO₂. At present, the activation of CO₂ by metal complexes is receiving much interest. The neutral metal complexes containing a single metal atom monodentate coordination M(η₁-CO₂) or bidentate coordination M(η₂-CO₂) have been extensively studied. In this work, a new kinds of BLi_n–O₂C (I_n) (n = 2–6) complexes with a bidentate double oxygen M(η₂-O₂C) coordination, where bimetal 2Li of BLi_n clusters bind to O of CO₂, are present. Those complexes are depicted in Fig. 1. Here, we also considered linear end-on coordination mode, where one Li atom binds to one O of CO₂. But BLi_n–O₂C complexes prefer to the bidentate double oxygen M(η₂-O₂C) coordination, except for BLi₄–O₂C and BLi₅–O₂C, whose energies for the bidentate double oxygen M(η₂-O₂C) coordination is only slightly higher than for the linear end-on coordination. Comparisons of the structures and their energies between two coordination modes are provided in Fig. S1 of the ESI.†

Fig. 1 The bidentate double oxygen M(η₂-O₂C) coordination complexes BLi_n–O₂C (I_n).

BLi_n cluster is a representative mononuclear cluster, in which the nuclear B atom is trapped into the Li alliance. Therefore, the interaction between CO₂ and BLi_n cluster in BLi_n–O₂C complex adopts the interaction between large electronegativity O and large ionizing potential Li. However, the BLi₂–O₂C complex not only adopts Li–O interaction, but also contains B–C interaction, because the nuclear B atom is exposed due to no enough Li atom. So the BLi₂ acting on CO₂ affords a stable BLi₂–O₂C complex, where the C atom of CO₂ is not active. In contrast to BLi₂–O₂C, BLi_n–O₂C (n = 3–6) do not contains B–C interaction, which leads to possess a potential active site (C position).

From Fig. 1, we found that the bidentate double oxygen coordination leads to a bent CO₂ moiety as a result of the partial negative charge transfer into the π* orbital of CO₂,^44,45 and that the structural integrities of BLi_n clusters are not destroyed. When compared to neutral CO₂, the C–O bonds are obviously elongated by 0.034–0.090 Å in the BLi_n–O₂C complexes, and the O–C–O angels become to be 120–140° in the bent CO₂ moiety from the 180° in linear CO₂. Therefore, from the structural view point, CO₂ has been activated by BLi_n cluster. Further, the unscaled harmonic vibrational frequencies and intensities are calculated to demonstrate the activation of CO₂. The antisymmetric CO₂ stretching (v_AS), the CO₂ symmetric stretch (v_S) and the CO₂ bend (v_B) are listed in Fig. 1. In bare CO₂, v_AS, v_S and v_B is found at 2222, 1216 and 713 cm⁻¹, respectively. Obviously, the antisymmetric CO₂ stretching in the BLi_n–O₂C complex becomes weaker than that in the bare CO₂, in which a redshift of v_AS is 240–700 cm⁻¹. The symmetric stretch and the CO₂ bend become stronger than those in the bare CO₂, and their blueshift of v_AS and v_B is 110–610 cm⁻¹ and 23–154 cm⁻¹. This analysis then suggests that the action of BLi_n cluster on CO₂ leads to a weakening of the C=O bonds.

Here, the advantages of complex BLi_n were concerned. Comparing with BH_n and AlLi_n clusters, BLi_n not only provides the large charge transfer into the π* orbital of CO₂, but also holds the structural integrities of BLi_n clusters. First, we calculated the vertical ionization potential to describe the capabilities of the charge transfer of BLi_n (3.633–4.534 eV), BH_n (6.859–13.128 eV) and AlLi_n (3.367–5.317 eV), and found that the capabilities of charge transfer of BH_n is weakest. Second, although AlLi_n has the large capabilities of the charge transfer; it is difficult to maintain the structural integrities, due to weak Al–Li bond.

Comparing with the antisymmetric CO₂ stretching in other structure with carbene character, it is significant to understand the weakening of the C=O bonds. While isolated CO₂⁻ by itself is not stable in the gas phase with respect to electron autodetachment (EA = −0.6 eV), its antisymmetric CO₂ stretching is measured at 1658 cm⁻¹ in a Ne matrix.⁴⁶ In CO₂-solvated M(CO₂)⁻ complexes (M = Au, Ag),⁴⁷ corresponding to partial charge transfers ranging from 0.5 to 0.9 |e|, intermediate values for v_AS (1865–1680 cm⁻¹) are found. Further, a considerably more redshited v_AS (1128 cm⁻¹) is observed in [ClMgCO₂]⁻.³⁵ Therefore, the calculated large redshited v_AS (1506–1964 cm⁻¹) suggests that the bent CO₂ moiety in the BLi_n–O₂C complexes may also exhibit carbene character. A superficial scheme explains the carbene character of the bent CO₂ moiety in the BLi_n–O₂C complexes (see Scheme 2).

Scheme 2 The carbene character of the bent CO₂ moiety in the BLi_n–O₂C complexes through the partial negative charge transfer into the π* orbital of CO₂.

Natural bond orbital analyses were carried out to understand the charge transfer of BLi_n–O₂C complexes. The partial negative charge transfer into the π* orbital of CO₂, decreases the energies of the lowest empty orbitals (LUMOs) or the highest single occupation orbitals (SOMOs) in the BLi_n–O₂C complexes. As listed in Fig. 1, the SOMOs of BLi₂–O₂C, BLi₄–O₂C and BLi₆–O₂C is −0.28142, −0.25438 and −0.19137 a.u., respectively; and the LUMOs of BLi₃–O₂C and BLi₅–O₂C BLi₇–O₂C is −0.15327, and −0.01241 a.u. Comparing with the high energy LUMO of CO₂ (0.05419 a.u.), which indicates that the electrophilicities of the BLi_n–O₂C complexes are increased. Moreover, the low energies of the lowest empty orbitals in the BLi_n–O₂C complexes is close to that of unstable radical anion CO₂⁻˙ with high reactive activity. Therefore, it is expected that the BLi_n–O₂C complexes exhibited high reactive activity, and would react with some molecules.

3.2 Carboxyl functionalization of the C–H bond of benzene molecule via BLi_n (n = 3–6) assisting

Our analysis above shows that the bent CO₂ moiety in the BLi_n–O₂C complex is active by assisting of clusters BLi_n and possesses a potential active site at the C atom of CO₂. Then, a significant question is raised as to whether we can achieve a transformation of CO₂ via BLi_n assisting. Consequently, we investigated the reaction between the BLi_n–CO₂ complex and benzene molecule, to elucidate a novel alternative approach to direct C–H carboxylation by inserting CO₂ into C–H bond. Due to stable BLi₂–O₂C complex without active site C, here, we only consider other BLi_n (n = 3–6) clusters. The routes inserting CO₂ into C–H bond are shown in Scheme 3.

Scheme 3 The routes inserting CO₂ into C–H bond via BLi_n assisting.

At the beginning of the reaction, one CO₂ molecule approaches BLi_n, and forms slightly stable BLi_n–O₂C complexes I_n with a potential active site at the C atom of CO₂. Then, I_n would attack at benzene molecule to produce intermediates II_n via the C–C bond formation between the active site C atom and the carbon atom of C₆H₆. Further, Li-bond complexes are obtained through H transfer. The free energy profile is shown in Fig. 2, and the transition states and important intermediates are depicted in Fig. 3 and 4.

Fig. 2 Free energy profile of the catalytic cycle under two H₂O assisting. Blue, red, black and green lines represents n = 3, 4, 5 and 6, respectively.

Fig. 3 Optimized structures of C–C formation TS1 and H-transfer TS2 through two H₂O molecules acting as an H-transfer tunnel. Bond lengths are in Å, and frequencies are in cm⁻¹.

Fig. 4 Optimized structures of intermediates II and products. Bond lengths are in Å, and angles are in degree.

For the slightly stable complexes I_n (n = 3–6), they have lower energies of 12.69(n = 3), 28.46(n = 4), 4.29(n = 5) and 24.05(n = 6) kcal mol⁻¹ (see Fig. 2), respectively, than the total energies of BLi_n and CO₂. The decreased energies for n = 3 and 5 is smaller than for n = 4 and 6.

As discussion above, the partial negative charge transfer into the π* orbital of CO₂, leads to the high reactive activity of the bent CO₂ moiety in the complexes I_n with a potential active site at the C atom of CO₂. Consequently, the intermediates II_n are obtained via the C–C bond formation between the active site C atom and the carbon atom of C₆H₆. The barrier for the C–C formation transition state TS1 is 5.92, 24.00, 9.01 and 9.42 kcal mol⁻¹ for n = 3, 4, 5 and 6 (see Fig. 1, the molecular coordinates of TS1 are listed in the ESI†), respectively. In TS1, the C–C distance is 2.104, 2.153, 2.103 and 2.152 Å (see Fig. 3), respectively. For the obtained II_n, they have lower energies (9.42(n = 3), 4.44(n = 5) and 4.55 kcal mol⁻¹ (n = 6)) than the total energies of I_n and C₆H₆, except for II₄. The energy of II₄ is higher by 12.11 kcal mol⁻¹ than the total energies of I₄ and C₆H₆. Therefore, these indicate that intermediates (II₃, II₅ and II₆) are the resting states in the reaction process.

In intermediates II_n (see Fig. 4), the lengths of the new formed C–C bond are 1.549–1.562 Å, which are close to ordinary C–C bond length. Thus these reactions provide an alternative approach of C–C bond formation, which offers a new idea to employ CO₂ as an attractive carbon source in organic transformations. We also found that the bond length of C–H is longer (∼0.02 Å) than 1.086 Å of C₆H₆. This indicates that the C–H bond is slightly active due to the C–C formation. On the other hand, the CCC angles (135.2–139.7°) can approximately be looked upon as the dihedral angles between the C₆H₆ ring and the plane of CO₂, which are much larger than normal 109.5° of C-sp³.

To achieve direct C–H carboxylation through inserting CO₂ into C–H bond, the H-atom-transfer from C to O is required. As shown in Scheme 3, if II are the start point, the more straightforward pathway is the direct transfer of the H atom from C to the O of CO₂. However, the H-direct transfer transition states are only found for II₃ and II₅ (see Fig. 5), yet for II₄ and II₆, we have not obtained. This reason can be understood by the structure characters of II. Comparing the structures of II₃ and II₅ with II₄ and II₆, we found that the lengths of the new formed C–C bonds in II₃ (1.555 Å) and II₅ (1.562 Å), are slightly longer than those in II₄ (1.549 Å) and II₆ (1.549 Å), and the dihedral angles in II₃ (139.7°) and II₅ (139.5°) are also slightly larger than those in II₄ (135.2°) and II₆ (135.3°). Therefore, the new formed C–C bonds in II₃ and II₅ are able to easily rotate in comparison with those in II₄ and II₆. However, the predicted barriers for the H-direct transfer transition states are very large (up to 76.38 and 43.57 kcal mol⁻¹ for II₃ and II₅, respectively), which can be attributed to the large rotation of the dihedral angle between the C₆H₆ and CO₂ moiety planes. The large barriers are too high to accomplish under mild conditions. Therefore, an alternative H transfer route should be found. By understanding the reason of the large rotation barriers, we present that employing extra H₂O molecule can act as an H-transfer tunnel and decreases the rotation barriers. Comparing with the H-direct transfer route, we find that the extra H₂O molecule acting as an H-transfer tunnel indeed decreases the barriers for H transfer. As shown in Table 1, when the H transfer is mediated by one H₂O molecule acting as an H-transfer tunnel, the barrier decreases to 64.13 and 36.54 kcal mol⁻¹ for II₃ and II₅, respectively. The barrier is only 14.15 and 21.90 kcal mol⁻¹ for II₄ and II₆, respectively. Their structures of transition states are depicted in Fig. 5. Further, when the H transfer is mediated by two H₂O molecules acting as an H-transfer tunnel, the barrier decreases to 49.34, 5.81, 22.16 and 6.38 kcal mol⁻¹ for II₃, II₄, II₅ and II₆, respectively. The molecular coordinates of different H-transfer transition states are listed in the ESI.† Here, we mainly depict the H transfer transition states TS2 through two H₂O molecules acting as an H-transfer tunnel (see Fig. 2). In TS2 through two H₂O molecules acting as an H-transfer tunnel (see Fig. 3), the H-bond bridge is long enough to effectively ease the rotation of the dihedral angle between the C₆H₆ and CO₂ moiety planes. Comparing the H-direct transfer transition states and one H₂O molecule assisting H-transfer transition states, we found that the barrier would decrease 7–15 kcal mol⁻¹ per additional one H₂O (see Table 1), which is consistent with the 10–16 kcal mol⁻¹ barrier to internal rotation of the dihedral previously reported.^48–51 This is because the rotation of the dihedral angle between the C₆H₆ and CO₂ moiety planes decreases 20–40° (see Fig. 6), which leads to an added impetus to form an extended conjugated π system through the couple of π_C6H6 and π_CO2. The added impetus can promote to the H-atom-transfer from C to O.

Fig. 5 Optimized structures of the H-direct transfer transition states and one H₂O molecule assisting H-transfer transition states. Bond lengths are in Å, and frequencies are in cm⁻¹.

Table 1 Comparison of the barriers of different H-transfer tunnels

n	Direct H transfer	H transfer by one H2O assisting	H transfer by two H2O assisting
3	76.38	64.13	49.34
4	No found TS	14.15	5.81
5	43.57	36.54	22.16
6	No found TS	21.90	6.38

---

Fig. 6 The rotation of the dihedral angle between the C₆H₆ and CO₂ moiety planes.

Through H-atom-transfer from C to O, the product of the Li-bond complex P is obtained. In P, the C–C bond lengths are ∼1.49 Å, and the dihedral angles between the C₆H₆ and CO₂ moiety planes are close to 180°. Thus, the carboxylation of C–H bond is achieved through C–C bond formation and the H-atom-transfer from C to O via two H₂O molecules acting as an H-transfer tunnel.

The mechanism of the reaction between complex BLi_n–CO₂ and C₆H₆ can be described as follow. First, BLi_n–O₂C complex with the bidentate double oxygen M(η₂-O₂C) coordination possesses a potential active site at the C atom of CO₂ moiety. This is because the partial negative charge transfer from BLi_n moiety into the π* orbital of CO₂ moiety, leads to the carbene character of the bent CO₂ moiety in the BLi_n–O₂C complex. Then, the C–C bond is formed between the active site C atom and the carbon atom of C₆H₆ through radical attack. Finally, the H transfer is performed with the help of two H₂O molecules acting as an H-transfer tunnel. Here, we found that two H₂O molecules assisting long H-bond bridge lowers the barrier of H-transfer transition state through effectively easing the rotation of the dihedral angle between the C₆H₆ and CO₂ moiety planes. After comparing all relative free energies shown in Fig. 1, we can find that the C–C bond formation is more favorable for n = 3 and 5 than n = 4 and 6. But for the step of H-atom-transfer from C to O, the barriers for n = 4 and 6 are lower than those for n = 3 and 5. Therefore, we can conclude that BLi₅ (the barrier of 22.16 kcal mol⁻¹) and BLi₆ (9.52 kcal mol⁻¹) clusters are more feasible to the carboxylation of C–H bond of benzene through the two steps of the C–C bond formation and two H₂O molecules assisting H transfer.

4. Conclusions

In this paper, we have proposed the boron-doped lithium clusters BLi_n to activate CO₂, and found that the bimetal 2Li of BLin clusters bind to the two O atoms of CO₂ to form a kind of novel BLi_n–CO₂ complex (I_n) (n = 3–6) with a bidentate double oxygen M(η₂-O₂C) coordination. In I_n, because of the partial negative charge transfer from BLi_n into the π* orbital of CO₂, the antisymmetric CO₂ stretching in the BLi_n–O₂C (I_n) complex becomes weaker than that in the bare CO₂, and the symmetric stretch and the CO₂ bend become stronger than those in the bare CO₂. Meanwhile, the energies of LUMO and SOMO are decreased. These lead to a weakening of the C=O bonds of CO₂, and an active CO₂ moiety, yet the structural integrities of BLi_n clusters are not destroyed. Further, the reaction between the BLi_n–CO₂ complex and benzene molecule is investigated to elucidate a novel alternative approach to direct C–H carboxylation by inserting CO₂ into C–H bond. We found that the carboxylation of C–H bond can be achieved through C–C bond formation and the H-atom-transfer from C to O through two H₂O molecules acting as an H-transfer tunnel. Comparing the transition states of H-direct transfer and one H₂O molecule assisting H transfer, we found that two H₂O molecules acting as an H-transfer tunnel has a lower barrier, due to that long H-bond bridge is to effectively ease the rotation of the dihedral angle between the C₆H₆ and CO₂ moiety planes. Comparing all relative free energies in the whole free energy profile, we found that BLi₅ (the barrier of 22.16 kcal mol⁻¹) and BLi₆ (9.52 kcal mol⁻¹) clusters are more feasible to the carboxylation of C–H bond of benzene through the two steps of the C–C bond formation and two H₂O molecules acting as an H-transfer tunnel. Therefore, we hope that these results encourage scientists to investigate more clusters to active CO₂ and further catalyze the transformation of CO₂.

Acknowledgements

This work was supported by the National Natural Science Foundation of China (No. 21303065 and 21364009).

References

1. M. Aresta, Carbon Dioxide as Chemical Feedstock, Wiley-VCH, Weinheim, 2010.
2. A. Behr, Angew. Chem., 1988, 100, 681–698.
3. D. H. Gibson, Coord. Chem. Rev., 1999, 185–186, 335–355.
4. H. Hoberg and S. Schaefer, J. Organomet. Chem., 1982, 236, C28–C30.
5. H. Hoberg and S. Schaefer, J. Organomet. Chem., 1983, 251, C51–C53.
6. D. Walther, E. Dinjus, J. Sieler, L. Andersen and O. Lindqvist, J. Organomet. Chem., 1984, 276, 99–107.
7. H. Hoberg, A. Ballestros, A. Sigan, C. Jegat and A. Milchereit, Synthesis, 1991, 395–398.
8. Y. Liu, J. Cornella and R. Martin, J. Am. Chem. Soc., 2014, 136, 11212–11215.
9. F. B. Sayyed, Y. Tsuji and S. Sakaki, Chem. Commun., 2013, 49, 10715–10717.
10. A. Correa and R. Martin, J. Am. Chem. Soc., 2009, 131, 15974.
11. P. Knochel, M. I. Calaza and E. Hupe, Metal-Catalyzed Cross-Coupling Reactions, Wiley, Weinheim, 2004, p. 619.
12. G. E. Greco, B. L. Gleason, T. A. Lowery, M. J. Kier, L. B. Hollander, S. A. Gibbs and A. D. Worthy, Org. Lett., 2007, 9, 3817.
13. E. M. Beccalli, G. Broggini, M. Martinelli and S. Sottocornola, Chem. Rev., 2007, 107, 531.
14. K. Huang, C.-L. Sun and Z.-J. Shi, Chem. Soc. Rev., 2011, 40, 2435–2452.
15. K. Sasano, J. Takaya and N. Iwasawa, J. Am. Chem. Soc., 2013, 135, 10954–11095.
16. Z. Li, C. Brouwer and C. He, Chem. Rev., 2008, 108, 3239.
17. A. S. K. Hashmi, Chem. Rev., 2007, 107, 3180.
18. E. Jiménez-Núñez and A. M. Echavarren, Chem. Rev., 2008, 108, 3326.
19. L. Zhang, J. Cheng, T. Ohishi and Z. Hou, Angew. Chem., Int. Ed., 2010, 49, 8670.
20. I. I. F. Boogaerts, G. C. Fortman, M. R. L. Furst, C. S. J. Cazin and S. P. Nolan, Angew. Chem., Int. Ed., 2010, 122, 8674.
21. Q. W. Song, Z. H. Zhou, H. Yin and L. N. He, ChemSusChem, 2015, 23, 3967–3972.
22. S. Rasul, D. H. Anjum, A. Jedidi, Y. Minenkov, L. Cavallo and K. Takanabe, Angew. Chem., Int. Ed., 2015, 127, 2174–2178.
23. K. Judai, S. Abbet, A. S. Wörz, U. Heiz and C. R. Henry, J. Am. Chem. Soc., 2004, 126, 2732–2737.
24. Y. Gao, N. Shao, Y. Pei, Z. Chen and X. C. Zeng, ACS Nano, 2011, 5, 7818–7829.
25. L. Li, Y. Gao, H. Li, Y. Zhao, Y. Pei, Z. Chen and X. C. Zeng, J. Am. Chem. Soc., 2013, 135, 19336–19346.
26. L. Li and X. C. Zeng, J. Am. Chem. Soc., 2014, 136, 15857–15860.
27. I. Fampiou and A. Ramasubramaniam, J. Phys. Chem. C, 2015, 119, 8703–8710.
28. J. Li, P. González-Navarrete, M. Schlangen and H. Schwarz, Chem.–Eur. J., 2015, 21, 1–11.
29. A. I. Boldyrev, N. Gonzales and J. Simons, J. Phys. Chem., 1994, 98, 9931.
30. A. I. Boldyrev, J. Simons and P. V. R. Schleyer, J. Chem. Phys., 1993, 99, 8793.
31. A. V. Nemukhin, J. Almlof and A. Heiberg, Chem. Phys. Lett., 1980, 76, 601.
32. M. Aresta and A. Dibenedetto, Dalton Trans., 2007, 2975–2992.
33. M. Cokoja, C. Bruckmeier, B. Rieger, W. A. Herrmann and F. E. Kühn, Angew. Chem., Int. Ed., 2011, 50, 8510–8537.
34. M. Cokoja, C. Bruckmeier, B. Rieger, W. A. Herrmann and F. E. Kühn, Angew. Chem., 2011, 123, 8662–8690.
35. G. B. S. Miller, T. K. Esser, H. Knorke, S. Gewinner, W. Schölkopf, N. Heine, K. R. Asmis and E. Uggerud, Angew. Chem., Int. Ed., 2014, 53, 14407–14410.
36. 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. 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 A.2), Gaussian, Inc., Wallingford, CT, 2009.
37. A. D. Becke, J. Chem. Phys., 1993, 98, 5648–5652.
38. C. Lee, W. Yang and R. G. Parr, Phys. Rev. B: Condens. Matter Mater. Phys., 1988, 37, 785–789.
39. V. Barone, M. Cossi and J. Tomasi, J. Comput. Chem., 1998, 19, 404.
40. Y. Takano and K. N. Houk, J. Chem. Theory Comput., 2005, 1, 70.
41. A. E. Reed, R. B. Weinstock and F. Weinhold, J. Chem. Phys., 1985, 83, 735.
42. J. E. Carpenter and F. Weinhold, J. Mol. Struct.: THEOCHEM, 1988, 169, 41.
43. E. D. Glendenig, J. K. Badenhoop, A. E. Reed, J. E. Carpenter, J. A. Bohmann, C. M. Morales and F. Weinhold, NBO 5.0, Theoretical Chemistry Institute, University of Wisconsin, Madison, 2001.
44. B. J. Knurr and J. M. Weber, J. Phys. Chem. A, 2013, 117, 10764–10771.
45. B. J. Knurr and J. M. Weber, J. Phys. Chem. A, 2014, 118, 4056–4062.
46. W. E. Thompson and M. E. Jacox, J. Chem. Phys., 1999, 111, 4487.
47. B. J. Knurr and J. M. Weber, J. Am. Chem. Soc., 2012, 134, 18804–18808.
48. S. Y. Han, I. Chu, J. H. Kim, J. K. Song and S. K. Kim, J. Chem. Phys., 2000, 113, 596.
49. C. Lim, A. M. Holder and C. B. Musgrave, J. Am. Chem. Soc., 2013, 135, 142–154.
50. S. Qu, Y. Dang, C. Song, M. Wen, K.-W. Huang and Z.-X. Wang, J. Am. Chem. Soc., 2014, 136, 4974–4991.
51. T. Chen, H. Li, S. Qu, B. Zheng, L. He, Z. Lai, Z.-X. Wang and K.-W. Huang, Organometallics, 2014, 33, 4152–4155.

---

Footnote

† Electronic supplementary information (ESI) available: The structures and their energies between two coordination modes and molecular coordinates of key transition states. See DOI: 10.1039/c6ra15152c

---