Theoretical insights into copper(I)–NHC-catalyzed C–H carboxylation of terminal alkynes with CO₂: the reaction mechanisms and the roles of NHC

Abstract

The reaction mechanisms of copper(I)–NHC-catalyzed C–H carboxylation of terminal alkynes with CO₂ were investigated by DFT calculations (NHC=N-heterocyclic carbene). Three types of reaction mechanisms were designed, explored and compared. The optimal reaction channels of corresponding pathways were selected. It was investigated that the formation of new C–C bond in the insertion process of activated CO₂ by NHC was induced by the formation of Cu–O bond. Also, the functions of NHC were determined. Our calculations investigated that (1) the special difunctional roles of NHC can indeed facilitate the reaction process after the formation of CO₂–NHC–Cu cocatalyst, whereas the unexpected low energy of this cocatalyst results in its ultrastability and then hinders the dropping of energy barrier in the whole reaction and (2) the additional interaction of NHC with the same metal atom will promote the insertion process of CO₂ through increasing the electrophilicity of the metal center.

---

1. Introduction

Chemical capture, conversion and fixation of carbon dioxide have attracted much attention due to global warming.^1–3 As an environmentally friendly feedstock, utilizing CO₂ for organic synthesis is a highly attractive approach in making commodity chemicals for addressing the energy penalty problem.^1c,4 However, developing a more efficient catalyst to incorporate CO₂ into organic molecules under mild conditions remains a great challenge in real world applications.⁵

As carboxylic acids have been among the most important types of compounds in medicinal chemistry, numerous protocols have been well established.^6,7 One of the most attractive and straightforward reactions is the direct catalytic carboxylation of organic compounds with CO₂.^8–16 The essence of this method is the direct carboxylation of carbon nucleophiles using CO₂ as the electrophile.⁶ Typically, organometallic reagents such as organolithium are universally taken as precursors. However, this method is always limited by low catalytic performances, harsh reaction conditions, and restricted substrate scope (Scheme 1). Although some transition metal-catalyzed carboxylations of less reactive carbon such as organotin,¹⁶ organoboron,¹⁰ organozinc¹⁵ improve performance in functional group tolerance, stoichiometric consumption is still a disadvantage.

Scheme 1 Reaction equation of organometallic reagents and CO₂ for carboxylation.

Recently, other economical methods have been reported by direct C–H bond activation and carboxylation with CO₂.^8,17–21 For instance, Nolan et al. reported the carboxylation of C–H bonds of highly activated arenes and heterocycles using NHC–gold(I) and NHC–copper(I) complexes to catalyze CO₂.^17,18 Gooßen's group reported the C–H carboxylation of terminal alkynes catalyzed by low loadings of silver(I)/DMSO at ambient CO₂ pressure.¹⁹ Iwasawa et al. reported the direct carboxylation of arenes with CO₂ using rhodium(I)-catalyzed via chelation-assisted C–H bond activation.²⁰ Zhang Yugen et al. disclosed a copper–NHC-catalyzed transformation of CO₂ to carboxylic acid through C–H bond activation of terminal alkynes, which was carried out under ambient conditions and tolerant to a wide range of functional groups²¹ (Scheme 2). In theoretical aspects, some relevant researches have been carried out, such as the carboxylation of arene C–H bond,²² arylboronate esters,²³ and C–H bond of heteroarenes²⁴ with CO₂ catalyzed by metal complexes, respectively.

Scheme 2 Equation of generations of R1 and III-M1.

The reaction mechanisms of the carboxylation of terminal alkynes catalyzed by copper–NHC complexes with CO₂ are still unclear. Furthermore, the roles NHC plays in these reactions is also an interesting and key issue.²⁵ In order to solve these two issues, the mechanisms of the title reaction were studied by DFT calculations in this study. We designed, explored and compared three types of reaction mechanisms (Scheme 3) and investigated how additional NHC affects the reaction process. Optimal reaction channels for the title reaction were selected. Transformations of NHC types, still a hotspot as recently reported,²⁶ are discussed to help to interpret differences among the reaction mechanisms. We expect that understanding the reaction mechanisms and functions of NHC will help us devise more efficient synthetic strategies and catalysts for carboxylation reactions utilizing CO₂.

Scheme 3 Perspective of the three reaction pathways calculated.

2. Computational details

Molecular geometries of model complexes were optimized without constraints via DFT calculations using the B3LYP²⁷ functional within the Gaussian 09 programs.²⁸ Frequency calculations at the same level of theory have also been performed to identify all the stationary points as minima or transition states. For the Cu atom, the all-electron split valence basis set 6-311G* Wachters–Hay²⁹ was used. The 6-311G(d,p) basis set was used for all other atoms. Intrinsic reaction coordinates (IRC)³⁰ were calculated for the transition states to confirm that such structures indeed connect two relevant minima. To include the effect of the solvent on the optimal reaction channels, the conductor-like polarizable continuum model (CPCM)^31,32 using UAHF radii on it was employed with dimethylformamide (DMF) (ε = 37.219) in consideration of the solvent used in the experiment. Partial atomic charges were calculated on the basis of natural bond orbital (NBO) analyses,³³ and Mulliken charges were used in this study. Several papers³⁴ have demonstrated the suitability and correctness of this DFT method. Charge decomposition analysis was obtained by Aomix³⁵ and Multiwfn.³⁶

We calculated the energies of initial reactant R1 with different conformations and picked the lowest one, the two branches of whose structure is a trans- site was used in our study.

3. Results and discussion

Pyridine-1-(ethylbenzene-NHC)-5-(ethylbenzene-NHC-p-nitrobenzene copper acetylide) R1 and its isomeride III-M1 are generated as the equation shown in Scheme 2. Substituent NO₂ is also a strong electron-withdrawing group, but not a general H substituent, which is helpful for the reaction to continue. Scheme 3 gives the possible reaction mechanism between these and CO₂, respectively. Both pathway I (PW-I) and pathway II (PW-II) are catalyzed by R1 initially. Pathway I is the direct insertion process of CO₂ with two channels, including the concerted and stepwise channels. Pathway II is the insertion process of activated CO₂ by NHC, with four channels. After the formation of II-M3, channels A and B are distinguished from the conformations between copper and the two oxygen atoms, (B-M1 and A-M2), while channels C and D can proceed through two- and one-step reactions, respectively. Pathway III is the insertion process of CO₂ catalyzed by III-M1 initially, proceeding through one transition state III-TS. In these pathways, R1 and III-M1 are all recyclable.

3.1 Pathway I (PW-I): direct insertion of process of CO₂

The calculated potential energy profiles for the reaction process of the direct carboxylation of 4-nitro-1-ethynylbenzene with CO₂ catalyzed by the copper–NHC complex are shown in Fig. 1. As shown in Fig. 1, the formation of R2 lets the sum of Gibbs energy of R1 and CO₂ drop to −4.5 kcal mol⁻¹ after the common optimization. The electronic interaction energy between fragment R1 and fragment CO₂ in their adduct R2 is −7.17 kcal mol⁻¹ (Table 1). This is in line with the lowest electronic interaction energy during the process, which reveals that the interaction between these two fragments is strengthening gradually. From complex R2 (Fig. S1†) to P1 (Fig. S1†), two different possible reaction mechanisms for CO₂ insertion into the Cu–C bond are considered when CO₂ is directly catalyzed, including the concerted and stepwise channels.

Fig. 1 Potential energy profiles of PW-I. All energies listed are Gibbs free energies (kcal mol⁻¹).

Table 1 Charge decomposition analysis of different fragments in relevant complexes and transition states in PW-I^a

	d × 10^−3	b × 10^−2	(d − b) × 10^−2	r × 10^−2	E(int)/kcal mol^−1
a d, number of electrons donated from fragment (C6H5NO2)C≡C(CO2) to fragment L-Cu; b, number of electrons back donated from (C6H5NO2)C≡C(CO2) to fragment L-Cu; r, number of electrons involved in repulsive polarization; E(int), electronic interaction energy between fragment R1 and fragment CO2.
R2					−7.17
I-TS1					−19.26
HOMO + 12	0.044	1.189	−1.182	−1.557
HOMO + 22	1.665	2.981	−2.814	−3.694
I-M1					−69.81
HOMO + 28	0.985	1.986	−1.887	−2.797
I-TS2					−82.26
HOMO + 20	9.377	0.388	0.550	0.027

---

In the concerted channel, CO₂ synchronously interacts with both the Cu metal center and the associated C atom, generating complex P1 via the transition state I-TS3 (Fig. 2). The energy barrier is 30.5 kcal mol⁻¹.

Fig. 2 Structures of I-TS1, I-M1, I-TS3 in PW-I. All hydrogen atoms are omitted. Bond distances and bond angles are measured by Å and degree, respectively.

Nevertheless, the stepwise channel has the lower maximum energy barrier of 26.5 kcal mol⁻¹ from complex R2 to I-TS1 (Fig. 2), providing an intermediate I-M1 (Fig. 2). The forming of C1–C2 bond occurs first, and then the Cu–O1 bond forms, accompanied by the breaking of the Cu–C2 bond. Since the coordination ability of carboxylate oxygen vs. C2–C3 bond π system, with the forming of C1–C2 bond, the bond distance of C2–C3 elongates to 1.25 Å in I-TS1 first and then has a slight shortening in I-M1 (1.24 Å). Meanwhile, the Cu–C2 bond distance elongates from 2.00 Å in I-TS1 to 2.01 Å in I-M1. This also results in the Cu–C2–C3 angles in both I-TS1 and I-M1 being acute angles, which is different from the corresponding obtuse angles in I-TS3 of the concerted channel.

From stage I-M1 to P1, the energy barrier is small, only 1.6 kcal mol⁻¹. The structure of I-TS2 is more profitable for the formation of Cu–O1 bond and the weakening of the interaction between the Cu atom and C≡C bond with the strengthening of the coordination ability of carboxylate O vs. C2–C3 π system (Table 1 and Fig. S3†). Meanwhile, the dihedral angle Cu–C2–C1–O1 in I-TS3 (−40.7°) is larger than that in I-TS1 (−31.5°), which is also favorable for the concerted interaction between the Cu–C2 single bond and the O1–C1 double bond.

In conclusion, the asynchronous channel is superior to the concerted one. P1 and P2 are isomerides. CO₂ is turned into 4-nitro-1-ethynylbenzene acid after the addition of 4-nitro-1-ethynylbenzene and, finally, the loss of R1.

3.2 Pathway II (PW-II): insertion process of activated CO₂ by NHC

As is well known, NHC is an efficient catalyst to activate CO₂;^2a,37 therefore, most of the time, it is easy to activate the involved CO₂ by the free carbene. Fig. 3 presents four potential energy profiles of the insertion process of activated CO₂ by NHC, which can be divided into three stages: CO₂ activation, CO₂ insertion and CO₂ deactivation. The CO₂ activation energy barrier that has to be overcome is only 1.3 kcal mol⁻¹.

Fig. 3 Potential energy profiles of PW-II. All energies listed are Gibbs free energies (kcal mol⁻¹).

After CO₂ activation by NHC, II-M1 (Fig. 4) is formed. In II-M1, one NHC is contributed to form copper–NHC–acetylide, and the other is contributed to form NHC-activated CO₂. Due to the strong electrophilicity of copper (0.602e) (Table S2†), it is easy to attract the adjacent electronegative oxygen atom (−0.518e) to generate the VDW complex II-M2 (Fig. S2†), heightening the energy barrier a little, by about 5.4 kcal mol⁻¹.

Fig. 4 Structures of II-M2, II-M3, A-M1 in PW-II. All the hydrogen atoms are omitted. Bond distances and bond angles are measured by Å and degree, respectively.

During the stage of CO₂ insertion, by reference to the charge decomposition analysis results, the electronic interaction energy between fragment R1 and fragment CO₂ becomes increasingly larger, which shows the reaction process between them (Table S6†). It is the strong donation interaction between fragment L-Cu and fragment CO₂ that induces the formation of intermediate II-M3 (Fig. 4) via II-TS2 (Fig. S2†), with an energy barrier of 23.0 kcal mol⁻¹. The Cu–O1 distance continues shortening to 2.47 Å.

Starting from II-M3, four channels were designed and calculated. Channels A and B will go through the same transition state A-TS1 (Fig. S4†) to obtain the intermediate A-M1 (Fig. 4), with a 20.2 kcal mol⁻¹ energy barrier. In this step, the Cu–O1 bond is formed, along with the disassociation of fragment L-Cu and the acetylide moiety. The bond distance of Cu–O1 shortens to 1.83 Å in A-M1 (1.87 Å in A-TS1). Second order perturbation theory analysis shows that Cu–O bond has been formed in A-TS1, and the acetylide moiety begins to be far away from the copper center. The charge of the C3 atom in A-M1 also decreases greatly, to −0.047e (Table S2†).

From here on, channels A and B are divided. In channel A, a half-encircled structure complex A-M2 (Fig. S4†) is formed via A-TS2, with an energy barrier of only 9.2 kcal mol⁻¹. Table S6† shows the strong back donation interaction between fragment L-Cu–(CO₂) and fragment (C₆H₅NO₂)C≡C in the HOMO+2 orbital of A-TS2, the performance on the strong virtual frequency vibration between C1 and C2 atoms. In A-M2, the bond distances of the Cu and O1 atoms and the Cu and O2 atoms are 1.88 Å and 2.05 Å, respectively.

Until now, CO₂ has been successfully inserted into the metal-carbon bond. Due to the instability of NHC complex, it is quite easy to break this C–C bond to vacate this NHC to recover its capacity for activating CO₂ again via A-TS3 (Fig. S4†) finally. The energy barrier of this step is very small, only 0.5 kcal mol⁻¹.

For channel B, a similar reaction mechanism is found, except for the intermediate B-M1 (Fig. S4†) and production P1 replacing their corresponding isomerides A-M2 and P2 in channel A, respectively. The Cu–O1–C1–O2 dihedral angle is −24.3° while that in B-M1 is 90.3°. This means the acetylide moiety attacks the rest moiety from two different orientations. The energy barrier is also a little higher compared to that of channel A, about 11.8 kcal mol⁻¹.

In channel C, formations of Cu–O1 and C2–C1 bonds synchronously occur during the process of CO₂ insertion. A-M2 is generated after a much higher energy barrier of 34.9 kcal mol⁻¹ via C-TS (Fig. S5).† In channel D, the insertion process of CO₂ moiety into the Cu–C bond and the process of CO₂ deactivation are realized in one step after D-TS (Fig. S5),† generating the same product, P2. Therefore, channel D has the highest energy barrier, 35.6 kcal mol⁻¹.

Compared with these four possible reaction channels, it is obvious that channel A is optimal. For the reaction mechanism, a universal speculation is that the transfer of the CO₂ unit from the carbene center to the copper center²¹ is induced by the formation of a new C–C bond. However, our calculations show that actually, the Cu–O bond is formed first, then inducing the formation of a new C–C bond to finish the transfer of CO₂.

It is worth noting that the reaction process is indeed promoted after the formation of CO₂–NHC–Cu cocatalyst (II-M2), corresponding to the lower maximum energy barrier that needs to be overcome. On the contrary, because the energy of CO₂–NHC–Cu cocatalyst is much smaller, this situation instead has a negative effect on the whole reaction process.

3.3 Pathway III (PW-III): insertion process of CO₂ catalyzed by III-M1 initially

In PW-II, it is obvious that the reaction process is promoted after the formation of CO₂–NHC–Cu cocatalyst. Then, what will an associative path in which the NHC fragment act as bidentate ligand happen? Therefore, pathway III is investigated. Herein, the energy of starting reactant III-M1, the structure containing two NHCs connected by the same Cu metal center, is 11.2 kcal mol⁻¹ lower than that of its isomeride R1. Also, an interesting phenomenon was found during this calculation. It is that the largest energy barrier of this classical associative pathway is much lower than that of PW-I and PW-II, only 10.3 kcal mol⁻¹. The C1 and C2 atoms in III-TS (Fig. 5) have a strong interaction to form a bond, with charges of 0.543e and −0.634e, respectively (Table S4†). From III-M2 (Fig. S6†) to III-TS, the C1–C2 distance greatly shortens, from 3.32 Å to 1.93 Å. This interaction motivates the reaction to go further and gives a chance for the Cu atom to interact with the O1 atom deserved. In III-M3 (Fig. S6†), the distance of Cu–O1 elongates to 2.14 Å. The Cu–O1–C1–O2 atoms can only form a half-encircled structure due to the steric hindrance of chelation. The distances of the Cu–O1 and Cu–O2 bonds are 2.14 Å and 3.14 Å, respectively.

Fig. 5 Potential energy profiles of PW-III (kcal mol⁻¹). For structure of III-TS, all hydrogen atoms are omitted. Bond distances and bond angles are measured by Å and degree, respectively. All energies listed are Gibbs free energies.

3.4 DMF solvent effect

DMF solvent effect is considered and added to the optimal channels of these three pathways (Fig. 5 and 6), and then the maximum energy barriers of each step are compared. In Fig. 6, it is found that the maximum energy barrier among each elementary reaction of the stepwise channel (optimal channel in PW-I) (23.2 kcal mol⁻¹) is still smaller than that of channel A (optimal channel in PW-II) (23.8 kcal mol⁻¹). This result is consistent with that in gas phase. Similarly, the potential energy profile of PW-III in DMF is shown in Fig. 5. The sum of relative energies of III-M1 and CO₂ in DMF (−0.9 kcal mol⁻¹) is a little lower than that in gas phase (1.6 kcal mol⁻¹), maybe because of the steric hindrance of chelate.³⁸ Seemingly, this DMF solvent situation is more helpful for the actual reaction process.

Fig. 6 Potential energy profiles of optimal channels of PW-I and PW-II in DMF. All energies listed are Gibbs free energies (kcal mol⁻¹).

3.5 Roles of NHC

3.5.1 Compared with pathway I and II. In order to explore the role of NHC, the reaction mechanisms of PW-I and PW-II are compared first. It is demonstrated that NHC plays a dual role in PW-II, not only to form NHC–Cu cocatalyst as both the ligand and catalyst, but also as an activating catalyst to activate CO₂.

The charge decomposition analysis (CDA) and extended CDA (ECDA) were analyzed. (Table 2) ECDA results show that the total net electrons always transfer from fragment NHC-included to fragment copper-included, which is one of the typical characters of a Fischer-type NHC complex.³⁹ However, the carbon atom of Fischer-type NHC linked to Cu center is always positively charged, which is not consistent with the Mulliken charge analysis (Table S5†). Analyzing the Mulliken charges changes the optimal channel of PW-I and PW-II; it is found that the Mulliken charges of carbon atom of NHC change from positive to negative a total of three times, which would not be the typical character of a Fischer-type NHC complex.

Table 2 CDA and ECDA analysis of related complexes and transition states in PW-I and PW-II

	Complex/transition state	Net electrons transfer^a	dmax MO number^b	dmax^b
a For PW-I, net electron transfer is between (C6H5NO2)C≡C–Cu(CO2) to fragment L; for PW-II, net electron transfer is between L-CO2 and (C6H5NO2)C≡CCu.b dmax is max number of electrons donated from fragment NHC-included to fragment of copper-included among all MO.
PW-I	R2	0.1368	HOMO + 22	0.0443
	I-TS1	0.1055	HOMO + 22	0.0199
	I-M1	0.1303	HOMO + 1	0.0312
PW-II	II-M2	0.0782	HOMO + 20	0.0310
	II-TS2	0.0423	HOMO + 2	0.0206
	II-M3	0.0306	HOMO + 7	0.0277
	A-TS1	0.5668	HOMO + 15	0.0145

---

Therefore, the orbital diagram's donated maximum electron number was compared (Fig. S8† and Fig. 7). It was found that none of the orbital diagrams of complex and transition state charges changed into negative (I-M1, II-TS2 and A-TS1) revealed the corresponding Fischer-type shape. In Fig. S7† and Fig. 7, the d_max orbital diagrams of R2, I-TS1, II-M2 and II-M3 illustrate that their NHCs are all the typical Fischer-type. (Fig. 7) However, for II-TS2 and A-TS1, they show approximate temporary Schorck-type⁴⁰ shape NHCs, while for I-M1, it shows the approximate σ bond shape (Fig. 7).

Fig. 7 d_max orbital diagrams of related complexes and transition states in PW-I and PW-II (the part of C atom of NHC and Cu atom that it linked).

These abnormal phenomena result in the different reaction activity of NHC linked to copper center. Meanwhile, we find the energy barrier from II-M3 to A-TS1 in PW-II (20.2 kcal mol⁻¹) is lower than that from R2 to I-TS1 in PW-I (26.5 kcal mol⁻¹) in gas phase. This shows that CO₂ activated by NHC can really promote the reaction process after the formation of CO₂–NHC–Cu cocatalyst. However, due to the dramatic dropping of the energy of the cocatalyst II-M1, the maximum energy barrier of elementary reactions in PW-II goes 28.4 kcal mol⁻¹ (from II-M1 to II-TS2), which is larger than that in PW-I. Therefore, this abnormal overstable activated CO₂–NHC–Cu cocatalyst results in the maximum energy barriers of the total reaction process increasing but not decreasing in PW-II, finally.

3.5.2 Compared with pathway I and III. For the direct CO₂ insertion processes, the reaction energy barrier of PW-III is much lower than that of PW-I. It is obvious that PW-III has much higher reaction activity than that in PW-I. So what role does NHC play here?

The two catalysts of PW-I and PW-III (R1 and III-M1) have different structures. R1 is a chelate via one pyridine ring associated with two NHCs, with only half of the carbene species coordinated with copper, the other half remaining free carbenes. While in III-M1, two NHCs are associated with the same copper center. This can be very helpful for increasing the electropositivity of Cu (from 0.534e in R1 to 0.841e in III-M1) in PW-III.

Based on the NBO calculations, we compared the second-order stabilization energy “E(2)” of key stationary points (SP) in the two pathways. For PW-I, the C4 atom of NHC always has a strong interaction to donate its lone pair electrons to the unoccupied orbital of Cu atom in R2, I-TS1, I-TS3, with relatively big “E(2)” values (64.3, 59.7, 54.7 kcal mol⁻¹) (Table S1†). Taking I-TS1 as an example, the orbital imagery of the favorable NBO donor–acceptor interactions of LP(C4) → LP*(Cu) in I-TS1 is shown in Fig. 8.³⁹ Similarly, for PW-III, the LP(C4) → LP*(Cu), LP(C5) → LP*(Cu) interactions are both always responsible for the high degree of electron delocalization (Table 3) in all corresponding relevant complexes and transition states.

Fig. 8 The orbital imagery of the favorable NBO donor-acceptor interactions of LP(C4) → LP*(Cu) in I-TS1. Blue and red represent the plus and minus isosurfaces of LP(C4), respectively; orange and green represent the plus and minus isosurfaces of LP*(Cu), respectively.

Table 3 Second order perturbation theory analysis of Fock matrix in NBO basis of relevant complexes and transition states in PW-III^a

	Donor NBO (I)	Acceptor NBO (j)	E(2)/kcal mol^−1	E(j) − E(i)/a.u.	F(i, j)/a.u.
a LP, valence lone pairs; BD, valence bond; BD*, valence antibond; LP*, low occupancy lone pairs with no exact physical significance as interpreted in ref. 36; E(2), E(j)–E(i), F(i, j) refer to formula in Scheme S1.
III-M1	LP(1)C4	LP*(6)Cu	42.7	0.55	0.204
	LP(1)C5	LP*(6)Cu	43.8	0.55	0.207
III-M2	LP(1)C4	LP*(6)Cu	41.1	0.54	0.199
	LP(1)C5	LP*(6)Cu	44.4	0.54	0.207
	LP(1)C2	LP*(6)Cu	48.3	0.54	0.214
III-TS	LP(1)C4	LP*(6)Cu	43.7	0.51	0.198
	LP(1)C5	LP*(6)Cu	45.6	0.51	0.201
	LP(1)C2	LP*(6)Cu	35.7	0.48	0.165
	BD(3)C2–C3	LP*(6)Cu	33.5	0.25	0.127
III-M3	LP(1)C4	LP*(6)Cu	50.2	0.46	0.202
	LP(1)C5	LP*(6)Cu	45.8	0.47	0.194

---

To sum up, the additional interaction of NHC to the same metal atom increases the electrophilicity of the metal center. This special design promotes the reaction process greatly. This is the reason why PW-III is the optimal pathway.

4. Conclusions

In conclusion, DFT calculations had been carried out to study the reaction mechanisms of C–H carboxylation of terminal alkynes with CO₂ by copper(I)–NHC complexes. Three types of reaction mechanisms were explored and compared. For the direct insertion process of CO₂ starting from R1, the asynchronous channel was superior to the concerted channel. For the insertion process of activated CO₂ by NHC, four channels were calculated and channel A was selected as the optimal one. Also, the functions of NHC were figured out to help provide insight into how NHC influences the reaction process. The insertion process of CO₂ starting from III-M1 was also designed to auxiliary assist the research of the role of NHC.

For the reaction mechanisms, an unexpected finding is that an abnormal ultrastable cocatalyst with quite low energy is generated in the insertion process of activated CO₂ by NHC, which is one of the main reasons to increase the difficulty of the energy barriers' dropping in the whole reaction of PW-II, even though the special difunctional roles of NHC (not only to be used to form NHC–Cu cocatalyst as both ligand and catalyst, but also used as an activating catalyst to activate CO₂) can indeed facilitate the reaction process after its actual generation. Also, the Mulliken charges of carbon atom of NHC change from positive to negative a total of three times, which indicates that this NHC is not a typical Fischer-type carbene in the actual reaction process. Compared to the experimental proposed mechanism, in PW-II, we also found that the formation of a new C–C bond was induced by the formation of a Cu–O bond, but not the formation of a new C–C bond as the universal speculation. Besides, the comparison results of PW-II and PW-III point out that the metal center of copper–NHC–acetylide is much more helpful than only activating CO₂ by the other NHC for the reaction to proceed. In addition, in PW-III, the additional interaction of NHC to the same metal atom increases the electrophilicity of the copper center, which thereby greatly promotes the reaction process. These works are expected to help researchers to devise more efficient synthetic strategies and more efficient catalysts for carboxylation reactions utilizing CO₂.

Acknowledgements

The support of the National Natural Science Foundation of China (10947171 and 21103082) and the Special Graduate Student Innovation Foundation of Jiangxi Province (YC2013-B019) are gratefully acknowledged.

Notes and references

1. (a) J. D. Figueroa, T. Fout, S. Plasynski, H. McIlvried and R. D. Srivastava, Int. J. Greenhouse Gas Control, 2008, 2(1), 9–20; (b) S. Chunshan, Catal. Today, 2006, 115(1), 2–32; (c) M. Peters, B. Köhler, W. Kuckshinrichs, W. Leitner, P. Markewitz and T. E. Muller, ChemSusChem, 2011,(9), 1216–1240; (d) Carbon Dioxide as Chemical Feedstock, ed. Aresta, M., Wiley-VCH, Weinheim, 2010.
2. (a) S. N. Riduan, Y. Zhang and J. Y. Ying, Angew. Chem., Int. Ed., 2009, 48(18), 3322–3325; (b) D. M. D'Alessandro, B. Smit and J. R. Long, Angew. Chem., Int. Ed., 2010, 49(35), 6058–6082; (c) M. Mikkelsen, M. Jorgensen and F. C. Krebs, Energy Environ. Sci., 2010, 3(1), 43–81.
3. (a) S. Choi, J. H. Drese and C. W. Jones, ChemSusChem, 2009, 2(9), 796–854; (b) R. Zevenhoven, S. Eloneva and S. Teir, Catal. Today, 2006, 115(1–4), 73–79; (c) Z. Z. Yang, Y. N. Zhao and L. N. He, RSC Adv., 2011, 1(4), 545–567.
4. (a) G. A. Olah, K. S. A. Prakash and A. Goeppert, J. Am. Chem. Soc., 2011, 133(33), 12881–12898; (b) Z. Z. Yang, L. N. He, J. Gao, A. H. Liu and B. Yu, Energy Environ. Sci., 2012, 5, 6602–6639.
5. (a) S. N. Riduan and Y. Zhang, Dalton Trans., 2010, 39(14), 3347–3357; (b) I. Omae, Catal. Today, 2006, 115(1–4), 33–52; (c) Y. G. Zhang and J. Y. G. Chan, Energy Environ. Sci., 2010, 3(4), 408–417; (d) W. L. Dai, S. L. Luo, S. F. Yin and C. T. Au, Appl. Catal., A, 2009, 366(1), 2–12.
6. (a) A. Correa and R. Martin, Angew. Chem., Int. Ed., 2009, 48(34), 6201–6204; (b) S. P. Bew, in ComprehensiveOrganic Functional Groups Transformation II, ed. A. R. Katritzky and R. J. K. Taylor, Elsevier, Oxford, 2005, p. 19.
7. (a) K. Ukai, M. Aoki, J. Takaya and N. Iwasawa, J. Am. Chem. Soc., 2006, 128(27), 8706; (b) F. Lehmann, L. Lake, E. A. Currier, R. Olsson, U. Hacksell and K. Luthman, Eur. J. Med. Chem., 2007, 429(2), 276–285; (c) D. Bonne, M. Dekhane and J. Zhu, Angew. Chem., 2007, 119(14), 2537–2540 (Angew. Chem. Int., 2007, 46, 2485–2488); (d) C. S. Yeung and V. M. Dong, J. Am. Chem. Soc., 2008, 130(25), 7826; (e) G. Ebert, W. L. Juda, R. H. Kosakowski, B. Ma, L. Dong, K. E. Cummings, M. V. B. Phelps, A. E. Mostafa and J. Luo, J. Org. Chem., 2005, 70(11), 4314–4317; K. Park, T. Palani, A. Pyo and S. Lee, Tetrahedron Lett., 2012, 53(7), 733–737.
8. L. A. Zhang, J. H. Cheng, T. Ohishi and Z. M. Hou, Angew. Chem., Int. Ed., 2010, 49(46), 8670–8673.
9. M. Shi and K. M. Nicholas, J. Am. Chem. Soc., 1997, 119(21), 5057.
10. (a) T. Ohishi, M. Nishiura and Z. Hou, Angew. Chem., Int. Ed., 2008, 47, 5792; (b) J. Takaya, S. Tadami, K. Ukai and N. Iwasawa, Org. Lett., 2008, 10, 2697; (c) H. Ohmiya, M. Tanabe and M. Sawamura, Org. Lett., 2011, 13, 1086.
11. C. M. Williams, J. B. Johnson and T. Rovis, J. Am. Chem. Soc., 2008, 130, 14936.
12. J. Takaya and N. Iwasawa, J. Am. Chem. Soc., 2008, 130, 15254.
13. M. Tokimoto, Y. Nakamura, K. Kimura and M. Mori, J. Am. Chem. Soc., 2004, 126, 5956.
14. M. Aoki, M. Kaneko, S. Izumi, K. Ukai and N. Iwasawa, Chem. Commun., 2004, 2568.
15. (a) C. S. Yeung and V. M. Dong, J. Am. Chem. Soc., 2008, 130, 7826; (b) H. Ochiai, M. Jang, K. Hirano, H. Yorimitsu and K. Oshima, Org. Lett., 2008, 10, 2681; (c) A. Metzger, S. Bernhardt, G. Manolikakes and P. Knochel, Angew. Chem., Int. Ed., 2010, 49, 466.
16. (a) R. Johansson and O. F. Wendt, Dalton Trans., 2007, 488; (b) J. G. Wu and N. Hazari, Chem. Commun., 2011, 47, 1069.
17. I. I. F. Boogaerts and S. P. Nolan, J. Am. Chem. Soc., 2010, 132(26), 8858–8859.
18. I. I. F. Boogaerts, G. C. Fortman, M. R. L. Furst, C. S. J. Cazin and S. P. Nolan, Angew. Chem., Int. Ed., 2010, 49(46), 8674–8677.
19. M. Arndt, E. Risto, T. Krause and L. Goossen, ChemCatChem, 2012, 4(4), 484–487.
20. H. Mizuno, J. Takaya and N. Iwasawa, J. Am. Chem. Soc., 2011, 133(5), 1251–1253.
21. D. Y. Yu and Y. G. Zhang, Proc. Natl. Acad. Sci. U. S. A., 2010, 107(47), 20184–20189.
22. A. Uhe, M. Hölscher and W. Leitner, Chem.–Eur. J., 2012, 18(1), 170–177.
23. L. Dang, Z. Y. Lin and T. B. Marder, Organometallics, 2010, 29(4), 917–927.
24. A. Ariafard, F. Zarkoob, H. Batebi, R. Stranger and B. F. Yates, Organometallics, 2011, 30(22), 6218–6224.
25. S. Diez-Gonzalez, N. Marion and S. P. Nolan, Chem. Rev., 2009, 109, 3612–3676.
26. Organometallic Oxidation Catalysis in Topics in Organometallic Chemistry.ed. F. Meyer and C. Limberg, Springer, 2007.
27. (a) C. T. Lee, W. T. Yang and R. G. Parr, Phys. Rev. B: Condens. Matter Mater. Phys., 1988, 37, 785; (b) A. D. Becke, J. Chem. Phys., 1993, 98, 5648; (c) B. Miehlich, A. Savin, H. Stoll and H. Preuss, Chem. Phys. Lett., 1989, 157, 200.
28. 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, T. Keith, 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, Revision A02, Gaussian 09, Wallingford CT, 2009.
29. (a) A. J. H. Wachters, J. Chem. Phys., 1970, 52, 1033; (b) P. J. Hay, J. Chem. Phys., 1977, 66, 4377.
30. (a) K. Fukui, J. Phys. Chem., 1970, 74, 4161; (b) K. Fukui, Acc. Chem. Res., 1981, 14, 363.
31. V. Barone and M. Cossi, J. Phys. Chem. A, 1998, 102, 1995.
32. M. Cossi, N. Rega, G. Scalmani and V. Barone, J. Comput. Chem., 2003, 24, 669.
33. (a) A. E. Reed, L. A. Curtiss and F. Weinhold, Chem. Rev., 1988, 88, 849; (b) J. P. Foster and F. Weinhold, J. Am. Chem. Soc., 1980, 102, 7211.
34. (a) C. H. Suresh and M. J. Ajitha, J. Org. Chem., 2013, 78(8), 3918–3924; (b) J. J. Lippstreu and B. F. Straub, J. Am. Chem. Soc., 2005, 127(20), 7444; (c) H. Jacobsen, A. Correa, A. Poater, C. Costabile and L. Cavallo, Coord. Chem. Rev., 2009, 253(5–6), 687; (d) G. Frison and A. Sevin, J. Phys. Chem. A, 1999, 103(50), 10998; (e) A. R. Rod and E. Vessally, Asian J. Chem., 2007, 19(3), 1709; (f) C. Trindle, J. Org. Chem., 2003, 68(25), 9669.
35. I. Hypercube, HyperChem 8.0, Hypercube Inc., USA, 2007.
36. T. Lu, Multiwfn, version 2.4, http://multiwfn.codeplex.com.
37. L. Q. Gu and Y. G. Zhang, J. Am. Chem. Soc., 2010, 132(3), 914.
38. X. H. Liu, R. Pattacini, P. Deglmann and P. Braunstein, Organometallics, 2011, 30(12), 3302–3310.
39. E. O. Fischer and A. Maasbol, Angew. Chem., Int. Ed. Engl., 1964, 3, 58.
40. R. R. Schrock, Acc. Chem. Rev., 1979, 12, 98.

---

Footnote

† Electronic supplementary information (ESI) available: Cartesian coordinates and calculated energies for all structures used in the quantum chemical calculations. See DOI: 10.1039/c4ra00254g

---