Does the Co + -assisted decarbonylation of acetaldehyde occur via C-C or C-H activation? A theoretical investigation using density functional theory

Transcription

1 Does the -assisted decarbonylation of acetaldehyde occur via C-C or C- activation? A theoretical investigation using density functional theory Lianming Zhao, Rongrong Zhang, Wenyue Guo,* Shujuan Wu, and Xiaoqing Lu College of Physics Science and Technology, University of Petroleum of China Dongying, Shandong , PR China Abstract The decarbonylation of acetaldehyde assisted by, which was selected as a representative system of the transition metal ions-assisted decarbonylations of acetaldehyde, has been investigated using density functional theory at the B3LYP/6-31+G** level of theory. The geometries and energies of the reactants, intermediates, products and transition states relevant to the reaction were located in detail on the triplet ground potential energy surfaces of [Co,,, 4 ] +. ur calculations indicate the decarbonylation of acetaldehyde takes place through three steps, that is, C-C activation, aldehyde -shift and nonreactive dissociation, while C- activation by can t lead to the decarbonylation of acetaldehyde. The stable precursor of the decarbonylation products is determined to have a structure of (C 4 ) (C) rather than (C 3 ) (C) proposed by earlier researchers [refs. 1,2]. (Correspondence should be addressed to wyguo@mail.hdpu.edu.cn) - 1 -

2 1. Introduction The reaction of transition metal ions with hydrocarbons represents one of the most active research areas in contemporary gas-phase ion chemistry. In this area, C- and C-C bond activations in small hydrocarbons by transition metal ions are of particular interest because of the enormous practical importance to the petroleum industry as well as the fundamental importance of σ bonds as among the simplest, strongest, and most ubiquitous of chemical bonds. The past three decades have thus seen a significant amount of experimental and theoretical studies concerning these aspects [1-10]. Experiments have provided tremendous information of intrinsic binding properties, substantial quantitative thermochemical data and reactivity of metal ions with organic molecules [1-6]. n the other hand, quantum chemical calculations in principle offer a complementary source of information on the structure and energy details of the potential energy surfaces (PES) for the reactions [7-11]. Gas-phase reactions of first-row transition-metal ions (M + ) with acetaldehyde have been studied by many experimental techniques, such as ion beam mass spectrometry and crossed-beam, etc. These experiments revealed that decarbonylation of acetaldehyde (reaction 1) by M + (M= Co, Fe or Cr) is a dominant process at low reaction energies [1,2,12]. M + + C 3 C C 4 + MC + (1) In order to verify the experimental findings, the earlier researchers presented a reaction mechanism for the decarbonylation [1,2], which is depicted in Scheme 1. owever, on the basis of the experimental results we can t say whether the reaction takes place via M + insertion into the -C or C-C bond (i. e., via I or II in Scheme 1), and whether the stable precursor of products C 4 and MC + does bear a structure as III (Scheme 1). ence theoretical study on the reaction is still needed. In this letter, we report a theoretical investigation on the -assisted decarbonylation of acetaldehyde using DFT method (B3LYP). This selected method has been widely applied to electronic structure calculations on systems containing transition metals, which combine reasonable computational costs with accuracy sufficient for describing open-shell metal systems [7-11]. 2. Computational method - 2 -

3 A hybrid density functional method, which is the unrestricted nonlocal Becke s three-parameter functionals [13] with the correlations due to Lee et al. (i. e., UB3LYP) [14], was used in the calculations. It was combined with 6-31+G** [15] basis sets. We optimized the structures of all the reactants, products, intermediates, and transition states involved in the reaction of with C 3 C at the B3LYP/6-31+G** level of theory. Frequency calculations at the same theoretical level were then followed to obtain the zero-point energy (ZPE) correction and confirm whether the optimized species is a minimum or a saddle point structure. The pathways between the transition states and their connected minima have been identified by intrinsic reaction coordinate (IRC) calculations [16]. All calculations were carried out using GAUSSIAN 03 package [17]. 3. Results and Discussion The optimized geometries and structural parameters for the reactants, intermediates, products, and saddle points along the decarbonylation pathway are shown in Fig. 1, relevant energies are tabulated in Table 1, and the sketch of the PES is given in Fig. 2. We inspected the values of <S 2 > for all species involved in the reaction, and found the deviation of <S 2 > is less than 5%. The fact suggests that spin contamination was small in all of the calculations. As shown in Fig. 1, three minima (noted as 1, 2, and 3, respectively) and two first-order saddle point (noted as TS 1-2 and TS 2-3, respectively) have been located along the decarbonylation coordinate on the triplet surface of [Co,,, 4 ] +. The encounter complex -CC 3 (1) is featured by the linkage of to the atom of acetaldehyde. Both cis- and trans- structures were located on the ground PES, suggesting the metal ion interacts with the two different lone pair orbitals of, respectively. Energetically, the stabilities of both the isomeric structures of 1 are nearly the same (trans-form is computed to be 0.6 kcal mol -1 more stable than the corresponding cis-conformer), indicating the coexistence of the isomers in the gas phase. The overall symmetries of the encounter complexes are nearly C s with the symmetry plane defined by Co-- -. Upon binding with, the largest change in C 3 C is the increase in the - bond length. This is a result of oxygen polarizing charge toward, weakening the - bond. The equilibrium - distances are almost identical for the two isomers (1.866 and Å for trans- and cis- forms, respectively). Correspondingly, the - binding energies are very close for the trans- and cisforms (44.9 and 44.3 kcal mol -1, respectively), which is in perfect agreement with the experimental result (47 kcal mol -1 [1]). This fairly good agreement at least partially validates our theoretical approach to the electronic and geometric structures of the species

4 The inserted species 3 C- -C (2) may be formed by the metal ion insertion into the C-C bond of acetaldehyde. As shown in Fig. 1, complex 2 has no symmetry element other than due to the slight out-of-plane rotation of the entity in the complex. The -Co- angle of 2 is calculated to be 96.8º, slightly larger than 90º as expected for a perfect sd hybridization at. The Co- 2, Co-, Co-, and Co- distances are 2.317, 1.876, 1.876, and Å, respectively, indicating the center interacts with all these atoms. Whereas the - distance is Å, suggesting rupture of the C-C bond. The energy of 2 is 13.8 kcal mol -1 more stable than that of the initial reactants ( 3 F) + C 3 C. It should be noted that 2 is 31 kcal mol -1 less stable than 1. It can be seen that the direct break of ( 3 C) -C bond gives rise to C 3 and C. The formation energy of these products from the separated reactants is computed to be 26.6 kcal mol -1 (Fig. 2), which qualitatively agrees with the experimental result [1]. TS 1-2 is the first-order saddle point connecting the two species 1 and 2 on the ground PES. Structurally, the - bond is substantially stretched to Å, while the Co- 2, Co-, Co-, and Co- distances are very close to the corresponding bond lengths found in 2, respectively. In other words, the geometry of TS 1-2 takes large similarity to that of 2, so it is a late transition state on the PES. The energy of this transition state is 5.2 kcal mol -1 below that of the separated reactants, and 39.7 and 8.6 kcal mol -1 above those of the connected intermediates (1 and 2, respectively). Along the decarbonylation coordinate, complex 2 can be rearranged to an -shifted species 4 C- (C) (3), which can be a direct precursor of the decarbonylation products (C 4 and (C)). It can be seen that complex 3 has no symmetry elements other than symmetry (Fig. 1). The methane moiety in 3 is η 2 coordinated to the cobalt ion, which is consistent with the calculated structures of both the -C 4 [18] and 4 C- (C 2 ) [10] complexes. An η 3 coordinated isomer of 3 was also located 25.3 kcal mol -1 less stable than 3 but was identified as a transition state with an imaginary frequency of 135i cm -1. As has been discussed in isolated Co(C 4 ) + [18], the reason for these coordinated situations may be that the donor-acceptor interaction and sd hybridization, which result in the strengthening of -C 4 bond and the stabilization of (C) (C 4 ), are more effective in η 2 structure than in η 3 structure. Energetically, species 3 is computed to be 63.3 kcal mol -1 below the separated reactants, which constitutes the deepest energy well along the whole reaction pathway (see Fig. 2). As shown in Fig. 1, the structure of (C) group in 3 is almost identical to that of free (C), and the Co-, Co- and Co- 2 distances of 3 are calculated to be 2.223, and Å, respectively, suggesting relatively weak interaction between C 4 and (C). Correspondingly, the binding energy of - 4 -

5 4 C- (C) bond is computed to be 19.8 kcal mol -1, thus the overall decarbonylation process of ( 3 F) + C 3 C C 4 + (C) is exothermic by 43.5 kcal mol -1. This explains why at low energies only the decarbonylation products were observed in the reaction of cobalt ions with acetaldehyde [1]. It is interesting to note that we also locate the channel for producing C 4 and C to be exothermic by 25.1 kcal mol -1, though it was not detected in the experiment [1]. 2 and 3 are connected through a direct, one-step shift occurring via saddle point TS 2-3. Structurally, the transition state is featured by one elongated and one shortened C-Co bond as well as the both elongated - and - bonds, suggesting it may be a saddle point that connects 3 to either 2 via insertion into C-C bond (i. e., II Scheme 1) or the C- inserted minimum (I). In order to reveal the real minima that TS 2-3 may connect, Fig. 3 depicts the potential energy curve (PEC) from the IRC calculation starting with the transition state, also shown in Fig. 3 are some selected structures along the reaction pathway. It is clear that TS 2-3 is indeed the saddle point connecting minima 2 and 3. The concerted nature of TS 2-3 contrasts the suggested mechanistic scenario of alle et al., who proposed a stepwise reaction profile for the reductive elimination of methane, in which the C- inserted species (C 3 ) (C) (III in Scheme 1) is involved as a stable species [1]. All our attempts to locate such an inserted minimum on the [Co,,, 4 ] + PES lead instead to 2 or 3, suggesting that only species 3 can be the precursor of products C 4 and (C). Analogous situation has been found in -assisted C-C activation in ethane leading to products C 4 and =C 2 [10]. With respect to its relative energy (Fig. 2), TS 2-3 is found to be the most demanding point along the whole decarbonylation pathway, lying 8.6 and 22.4 kcal mol -1 above the separated reactants and 2, respectively. The other possible decarbonylation pathway proposed by alle et al., i. e., via C- inserted minimum I as shown in Scheme 1 was also considered. We did locate a low energy route for the formation of the C- inserted mimimum (TS 1-4 and 4, see Figs. 1 and 2). owever, we didn t locate a transition state that connects 4 and the decarbonylation precursor 3 in spit of our careful searches. We have thus to exclude this route as a possible pathway for the decarbonylation of acetaldehyde at center, and instead insertion into the C- bond leads only to both + C 3 C and Co + C 3 C +, which are calculated to be endothermic by 43.8 and 31.2 kcal mol -1, respectively. This is also qualitatively consistent with the experimental observations, that is, the endothermic products and C 3 C + are produced at higher energies [1]. The potential energy surface profile of the decarbonylation of acetaldehyde at the center has been generalized in Fig. 2. It is very clear that the reaction occurs through insertion into C-C - 5 -

6 bond followed by an aldehyde -shift and then a nonreactive-dissociation of the 4 C- (C) bond. The rate determining step of the reaction is the aldehyde -shift, which has an energy barrier of 8.6 kcal mol -1 with respect to the separated reactants. The stable precursor for the decarbonylation products is 4 C- (C) rather than (C 3 ) (C) as proposed by alle et al. [1]. According to present calculations, we conclude the C- activation by (via I in Scheme 1) lead not to the decarbonylation of acetaldehyde but to the eliminations of both Mg and Mg +. We also anticipate to find the mechanism differences between + C 3 C and Fe + (or Cr + ) + C 3 C systems in order to find a general profile of the reactions of first-row transition metal ions with acetaldehyde [1,2,12]. An analogous decarbonylation pathway has also been located for the later two systems. From the similarity of the decarbonylations of C 3 C with, Fe + and Cr +, we proposed that the reactions of first-row transition metal ions with C 3 C could proceed in a similar mechanism, containing three elementary steps of C-C activation, -shift and nonreactive dissociation. A more detailed description of the reaction mechanism of Fe + and Cr + with C 3 C will be presented in another separated paper. 4. Conclusions In this work we report a DFT investigation on the decarbonylation of acetaldehyde with, which is the representative system selected for the reactions of transition metal ions with C 3 C. The following conclusions can be drawn from this work. 1. The decarbonylation of C 3 C with follows C-C activation rather than aldehyde C- activation. This reaction proceeds through three elementary steps, i. e., C-C activation, aldehyde -shift and nonreactive dissociation. 2. The stable precursor of the decarbonylation products C 4 and (C) has a structure of (C 4 ) (C) rather than (C 3 ) (C), while the later is located as a transition structure. 5. Acknowledgment This work was supported by CNPC Innovation Fund, the Excellent Young Teachers Program and Key Project (N ) of ME, PRC. We are also grateful to the financial support of National Natural Science Foundation of China (No )

7 References [1] L.F. alle, W.E. Crowe, P.B. Armentrout, J.L. Beauchamp, rganometallics 3 (1984) [2] D.M. Sonnenfroh, J.M. Farrar, J. Am. Chem. Soc. 108 (1986) [3] K. Eller,. Schwarz, Chem. Rev. 91 (1991) [4] C.L. aynes, E.R. Fisher, P.B. Armentrout, J. Am. Chem. Soc. 118 (1996) [5] M.R. Sievers, L.M. Jarvis, P.B. Armentrout, J. Am. Chem. Soc. 120 (1998) [6] Q. Zhang, M.T. Bowers, J. Phys. Chem. A 108 (2004) [7] N. Jiang, D. Zhang, Chem. Phys. Lett. 366 (2002) 253. [8] D. Zhang, C. Liu, W. Bian, J. Phys. Chem. A 107 (2003) [9] D. Zhang, C. Liu, S. Bi, S. Yuan, Chem. Eur. J. 9 (2003) 484. [10] M.C. olthausen, W. Koch, J. Am. Chem. Soc. 118 (1996) [11] M.C. olthausen, A. Fiedler,. Schwarz, W. Koch, J. Phys. Chem. 100 (1996) [12] R.C. Burnier, G.D. Byrd, B.S. Freiser, J. Am. Chem. Soc. 103 (1981) [13] A.D. Becke, J. Chem. Phys. 98 (1993) [14] C. Lee, W. Yang, R.G. Parr, Phys. Rev. B 37 (1988) 785. [15] M.J. Frisch, J.A. Pople, J.S. Binkley, J. Chem. Phys. 80 (1984) [16] C. Gonzalez,.B. Schlegel, J. Phys. Chem. 94 (1990) [17] Gaussian 03, Revision B.05, M.J. Frisch, G.W. Trucks,.B. Schlegel, G.E. Scuseria, M.A. Robb, J.R. Cheeseman, J.A. Montgomery, Jr.T. Vreven, K.N. Kudin, J.C. Burant, J.M. Millam, S.S. Iyengar, J. Tomasi, V. Barone, B. Mennucci, M. Cossi, G. Scalmani, N. Rega, G.A. Petersson,. Nakatsuji, M. ada, M. Ehara, K. Toyota, R. Fukuda, J. asegawa, M. Ishida, T. Nakajima, Y. onda,. Kitao,. Nakai, M. Klene, X. Li, J.E. Knox,.P. ratchian, J.B. Cross, C. Adamo, J. Jaramillo, R. Gomperts, R.E. Stratmann,. Yazyev, A.J. Austin, R. Cammi, C. Pomelli, J.W. chterski, P.Y. Ayala, K. Morokuma, G.A. Voth, P. Salvador, J.J. Dannenberg, V.G. Zakrzewski, S. Dapprich, A.D. Daniels, M.C. Strain,. Farkas, D.K. Malick, A.D. Rabuck, K. Raghavachari, J.B. Foresman, J.V. rtiz, Q. Cui, A.G. Baboul, S. Clifford, J. Cioslowski, B.B. Stefanov, G. Liu, A. Liashenko, P. Piskorz, I. Komaromi, R.L. Martin, D.J. Fox, T. Keith, M.A. Al-Laham, C.Y. Peng, A. Nanayakkara, M. Challacombe, P.M.W. Gill, B. Johnson, W. Chen, M.W. Wong, C. Gonzalez, J.A. Pople, Gaussian, Inc., Pittsburgh PA, [18] J.K. Perry, G. hanessian, W.A. Goddard III, J. Phys. Chem. 97 (1993)

8 Figure captions Figure 1 ptimized geometries and selected structural parameters (in Å and deg) for the reactants, products, intermediates and saddle points involved in the decarbonylation of acetaldehyde with. Figure 2 Calculated reaction branches of the [Co,,, 4 ] + PES. The scaling factor for ZPE is Figure 3 Some selected structures on the potential energy curve (PEC) from the IRC calculation starting with transition state TS 2-3. Table caption Table 1 Total energies E and zero point energies ZPE (in hartrees) calculated at the B3LYP/6-31+G** level for all the species involved in the reaction of C 3 C with. Scheme caption Scheme 1 Schematic decarbonylation mechanism proposed by earlier researchers for C 3 C with transition metal ions [1,2]

9 trans Co cis-1 1 TS TS Co TS C Co TS C 3 C C C C C C 2 C C C C 3 C Co C 3 C + Co Fig

10 60 Energy (kcal mol -1 ) ( 3 F) + C 3 C TS TS C 3 + C 26.6 TS C 3 C 31.2 Co + C 3 C C 4 + C C + C 4-80 Reaction Coordinate Fig. 2 Energy (artree) TS distance (Angstrom) Fig

11 Table 1 Total energies E and zero point energies ZPE (in hartrees) calculated at the B3LYP/6-31+G** level for all the species involved in the reaction of C 3 C with. Species E ZPE Species E ZPE C C 3 C C Trans Cis C 3 C C C C TS C TS Co TS C 3 C Scheme 1 M + + C 3 C C 3 M + C + M C C 3 + M C (I) C 3 (II) C + M (III) C 3 MC + + C