Oxidative addition of methane and benzene C H bonds to rhodium center: A DFT study

Transcription

1 Chemical hysics Letters 43 (006) Oxidative addition of methane and benzene C bonds to rhodium center: A DFT study Siwei Bi *, Zhenwei Zhang, Shufen Zhu College of Chemistry Science, Qufu Normal University, Jingxuan West Road, Qufu, Shandong 7365, China Received July 006; in final form 3 September 006 Available online 3 September 006 Abstract A density functional theory study on mechanisms of the oxidative addition of methane and benzene C bonds to the rhodium center containing Cp and Me 3 ligands has been performed. Our calculated results confirm that the C bond cleavage from a sigma complex to a hydride alkyl complex is the rate-determining step. Compared with the case of methane C bond, the oxidative addition of benzene C bond is more favorable kinetically and thermodynamically. Stronger backdonation from metal center to the r * antibonding orbital of benzene C bond is responsible for the observations. Ó 006 Elsevier B.V. All rights reserved.. Introduction Alkanes and arenes can coordinate to transitional metal undergoing oxidative addition reaction (Scheme ). The fact that a large number of transition metal complexes are able to activate arenes but not alkanes is surprising, on the basis of the observation that the bond energy of benzene C bond (around 0 kcal/mol) is larger than that of alkane C bond (around 95 kcal/mol) []. And, some reactions of arene C activation proceed faster than those of alkane C activation []. Using theoretical and computational methods to study the activation of unreactive carbon hydrogen bonds has aroused great interests ever since the first presence of the intermolecular oxidative addition of alkane C bonds to transition-metal centers [3].With great efforts to investigate the nature of the activation process in the past years [4,5], scientists explored the unsaturated transition-metal fragments of different structural types, employing methods such as semiempirical methods [6], various ab initio techniques [7] and density functional theory [8]. The calculation results have provided detailed and reasonable explanations * Corresponding author. Fax: address: siweibi@6.com (S. Bi). for a large range of experimental results [9], which in turn demonstrates that quantum chemical calculations are reliable. For example, (C 5 Me 5 )(Me 3 ) was irradiated in a mixed solvent system of propane/benzene at low temperature to afford (C 5 Me 5 )(Me 3 )(C 6 5 ) and (C 5 Me 5 )- (Me 3 )(C C C 3 ). Experimental observations confirmed the reactions for benzene C activation were slightly faster than those for propane C activation. This paradox arouses our interest to explore the potential energy surfaces of such reactions using density functional theory (DFT). For the convenience of calculations, we used (C 5 5 )(Me 3 ) instead of (C 5 Me 5 )(Me 3 ), and methane instead of propane in the model reactions studied here (see Scheme ).. Computational methods Density functional theory (DFT) calculations at the B3LY level [0] are carried out in order to explore the potential energy surfaces (ES) of the model reactions. The LANLDZ effective core potential [] is chosen for and, while the standard 6-3G(d, p) basis sets [] are employed for the rest of atoms. olarization functions [3] are added for phosphorus (f(d) = 0.34). Experience has /$ - see front matter Ó 006 Elsevier B.V. All rights reserved. doi:0.06/j.cplett

2 386 S. Bi et al. / Chemical hysics Letters 43 (006) [M] + Alkyl [M] [M] + shown that this level of theory provides reasonable predictions for transition-metal-containing compounds [4]. Intrinsic reaction coordinate (IRC) [5] calculations allow us to interconnect the different structures located on the ES and then construct the energy profiles. All the calculations are performed with the GAUSSIAN 98 software package [6]. Considering the involvement of gas molecules (,C 4 ) in the reaction mechanisms, we used free energy to describe the potential energy profile. The enthalpic energy was given in bracket. 3. Results and discussion 3.. Mechanism on the activation of C 3 hv 0 kcal/mol 0 kcal/mol Scheme. Scheme. + C 6 6 [M] + Alkyl + C 4 C 3 (0.00) 0.00 C (35.68) 6.70 (30.46) C 4 3 C 3 (36.) 39.8 TS C 3 + (.69) The energy profile on the reaction of C 4 with (C 5 5 )(Me 3 ) is shown in Fig.. Three steps are involved in the reaction. The first step is the reductive elimination of, the second one is the r-bonding coordination of C 3 to center, and the final one is formation of a metal alkyl hydride complex through C 3 oxidative addition. The free energy change DG and the enthalpy change D for the first step ( ) are calculated to be 6.70 and kcal/mol, respectively, indicating that this step is strongly unfavorable thermodynamically. Clearly, the conversion from an 8e species to a 6e species is responsible for the great thermodynamic data. As experiments confirmed, such kind of reductive elimination could be achieved through irradiation. The small DG compared with the D is a result of entropy increase. Theoretical calculations predict that the step is a non-barrier process and no non-classic hydride intermediate is located. This may be resulted from the strong backdonation from center to the antibonding orbital of. The ability of backdonation of is enhanced with the presence of the two strong electron-donating ligands, Cp and Me 3. In the second step ( 3), one C bond of C 4 is bonded to the metal center. The free energy change and the enthalpy change are calculated to be 5.0 and 5. kcal/mol, indicating that this step is unfavorable in free energy and favorable enthalpically. The free energy increase results from the entropy decrease from to 3, and the enthalpy decrease is due to the transformation from a 6e species to an 8e species. From 3 to 4, the activation free energy and activation enthalpy are calculated to be 7.38 and 5.75 kcal/mol, respectively. The C 3 bond distance in TS is.45 Å, larger than that in 3 (.3 Å), and smaller than that in 4 (.4 Å), meaning the C 3 bond is breaking. This step is thermodynamically favored with DG and D being 6.95 and 7.77 kcal/mol, respectively. roduct 4 is more stable than the intermediate 3. This is because strong backdonation from center to the antibonding orbital of C 3 strongly weakens the C bond and hence leads to the formation of strong and C 3 bonds. Also, product 4 is less stable than the reactant. This is a result of the fact that the bond is stronger than the Me bond. It can be seen from Fig. that the C 3 bond cleavage is the rate-determining step. Compared oxidative addition of ( ) with that of C 4 ( 3), it can be seen that the former reaction is a non-barrier process and no intermediate exists, while the latter reaction experiences an intermediate 3 and a transition state TS. The reason for the phenomena is that s orbital of atom is dense and nondirectional, favoring effective overlap with transition metal d orbitals and hence enhancing the interaction between and atom. The hybridized orbital of C atom is directional in space, which makes against effective overlap with transition metal d orbitals and hence weakens the interaction between 4 C 3 Fig.. Energy profile for the reaction of Cp Me 3 with methane. Corresponding enthalpy was given in parentheses.

3 S. Bi et al. / Chemical hysics Letters 43 (006) and C atom, relative to the case of M. Thus, relatively stronger interaction of with makes the molecular complex (intermediate) non-existent, while relatively weak interaction of with C 4 results in the existence of intermediate 3 (see Fig. ). 3.. Mechanism on the activation of C 6 5 Fig. 3 shows the energy profile for the reaction of Cp Me 3 with benzene. Fig. 4 illustrates the B3LY optimized structures involved in the reaction. The first step ( ) has been discussed above. The second step ( 3 0 )is the coordination of benzene to center by using one pair of its p electrons, which has been confirmed to be a nonbarrier process. In other words, no transition state is located. The free energy change and enthalpy change for the step are calculated to be.55, 5.68 kcal/mol, respectively, indicating this step is favorable both in free energy and in enthalpy. In contrast to the step ( 3), the free energy decreases as well. This is because the intermediate -.6 -l C -.5 -C C.3 C -.7 -C C.45 3 TS C -.7 -C C.4 4 Fig.. Selected B3LY optimized structures involved in the reaction of Cp Me 3 with methane, together with selected bond distances. The bond distances are given in Å (0.00) + C (35.68) C 6 6 (0.00) 4.5 3' + (7.80) TS' (7.67) ' (9.3) TS' + (6.6) 0.0 Fig. 3. Free energy profile for the reaction of Cp Me 3 with benzene. Corresponding enthalpy was given in parentheses. 5'

4 388 S. Bi et al. / Chemical hysics Letters 43 (006) Cl ' Cl ' 5' Cl TS' Cl TS' -.8 -Cl Fig. 4. B3LY optimized structures involved in the reaction of Cp Me 3 with benzene, together with selected bond distances. The bond distances are given in Å. 3 0 is more stable than 3 as a result of the stronger backdonation from the d orbital to the antibonding orbital of C 6 5 bond. The third step is the bonding mode conversion. The C@C p electrons coordinate to the metal in 3 0, while the C r bonding electrons coordinate to the metal in 4 0. A transition state, TS 0, is located. The activation free energy and activation enthalpy are calculated to be 6.7, 7.80 kcal/mol. Comparing the structural data of 3 0 and TS 0 (see Fig. 4), we can see that and bond distances increase, indicating that the p-electron bonding is weakened. The 3 bond distance becomes smaller (.73.4 Å) and 3 bond is elongated (.08.0 Å), implying that the 3 r-bonding electrons are effectively coordinating to the metal center. In 4 0, is far away from atom, while is slightly close to. The 3 bond distance becomes.3 Å, implying the bond is further activated. The r-donation of 3 to the metal and the backdonation from metal to the antibonding orbital of 3 are responsible for the 3 bond elongation. Correspondingly, the 3 distance becomes smaller (.9 Å). Based on the data above, we can predict that the 3 has bonded to the metal center. The free energy change and enthalpy change from are calculated to be 5.73, 7.67 kcal/mol, respectively, indicating the step is thermodynamically unfavored. This is because the C bonding is weaker than the C@C bonding to the metal. The last step is the C bond breaking to form a hydride phenyl product. The transition state, TS 0,is located. The activation free energy and activation enthalpy are calculated to be 3.8,.65 kcal/mol. Compared with 4 0, the 3 bond in TS 0 is further activated, supported by the bond length change from.3 Å in 4 0 to.46 Å in TS 0. In the product 5 0, and h r bonds are formed. Our calculations show that the free energy change and enthalpy change from 3 0 to 4 0 are 9.87,.4 kcal/ mol, indicating that this step is much more favored. Examining the overall energy profile shown in Fig. 3, we can see that the rate-determining step is still the C bond breaking Comparison between oxidative additions of C 4 and C 6 6 bond cleavage As seen from Figs. and 3 that the first step ( ) is the same in both model reactions. The free energy change from to 3 increases while that from to 3 0 decreases. This is because 3 0 is more thermodynamically stable than 3 due to stronger backdonation from metal center to the antibonding p * orbital of C@C double bond involved in 3 0, which overcomes the entropy decrease. 4 0 and 3 have the same bonding mode where C bond is bonded to metal center by using its r-bonding electrons, but 4 0 is more stable than 3. This is a result of the fact that the phenyl group is much more electron-withdrawing than the methyl group, making the level of the antibonding r * orbital of C 5 6 to be lower than that of C 3. Therefore, the backdonation from metal center to the C r * orbital is stronger in 4 0 than in 3. As has been mentioned above, the C bond cleavage is the rate-determining step for both reactions. The highest free energy for the reaction of C 3 oxidative addition relative to is 39.8 kcal/mol (TS), while that for the reaction of C 6 5 oxidative addition relative to is kcal/mol (TS 0 ), indicating that the former is less favored kinetically than the latter. One reason is that 4 0 is more stable than 3. What is more, the activation free energy (7.38 kcal/mol) for the cleavage of C 3 bond (3 4) is greater than that (3.80 kcal/mol) for the cleavage of C 6 5 bond (4 5 0 ). The lower activation free energy from 4 to 5 0 results from stronger backdonation from center to the C 6 5 r * antibonding orbital. The C bond distance (.45 Å) ints is shorter than the 3 bond distance (.46 Å) ints 0 although the uncoordinated C 3 bond is longer than the uncoordinated C 6 5 bond, implying that the C 6 5 bond is more strongly activated than the C 3 bond. The small C bond dis-

5 S. Bi et al. / Chemical hysics Letters 43 (006) tance (.3 Å) ints compared with that (. Å) ints 0 also indicates that stronger interaction between and the C atom of benzene. This is understandable because the phenyl group is more electron-withdrawing than methyl group. 5 0 (0.0 kcal/mol relative to ) is more thermodynamically stable than 4 (4.95 kcal/mol relative to ). Clearly, stronger C 6 5 interaction compared with C 3 interaction is responsible for the higher stability of Conclusions Two reaction mechanisms on oxidative addition of C 3 and C 6 5 to the fragment CpMe 3 were studied by using density functional theory. Reductive elimination of from Cp Me 3 was calculated to be much kinetically and thermodynamically unfavorable. Experimentally, this step could be achieved with irradiation. For C 3 oxidative addition a sigma complex (3) is directly obtained, while for C 6 5 oxidative addition a more stable g -arene complex (3 0 ) is first formed and then the corresponding sigma complex (4 0 ) is obtained via the transition state TS 0. Our results of calculations indicate that the rate-determining step is the C bond cleavage for both the reactions. The C 6 5 oxidative addition is more favored than the C 3 one both kinetically and thermodynamically, for which stronger electron-withdrawing of phenyl group compared with methyl group is responsible. Acknowledgement This work was supported by the National Science Foundation of China (No ). References [] S.W. Benson, Thermochemical Kinetics, Wiley, New York, 968, p [] W.D. Jones, F.J. Feher, Acc. Chem. Res. (989) 9. [3] (a) A.. Janowicz, R.G. Bergman, J. Am. Chem. Soc. 04 (98) 35; (b) A.. Janowicz, R.G. Bergman, J. Am. Chem. Soc. 05 (983) 399. [4] (a) For reviews, see: G.W. arshall, Acc. Chem. Res. 8 (975) 3; (b) R.G. Bergman, Science 3 (984) 90; (c) A.. Janowicz, R.A. erima, J.M. Buchanan, C.A. Kovac, J.M. Struker, M.J. Wax, R.G. Bergman, ure. Appl. Chem. 56 (984) 3; (d) C.L. ill, Activation and Functionalization of Alkanes, Wiley, New York, 989; (e) J. alpern, Inorg. Chim. Acta 00 (985) 4; (f) M. Ephritikhine, New. J. Chem. 0 (986) 9; (g) W.D. Jones, F.J. Feher, Acc. Chem. Res. (989) 9; (h) A.D. Ryabov, Chem. Rev. 90 (990) 403; (i) J.A. Davies,.L. Watson, J.F. Liebman, A. Greenberg, Selective ydrocarbon Activation, rinciples and rogress, VC ublishers, Inc., New York, 990; (j) R.G. Bergman, J. Organomet. Chem. 400 (990) 73; (k) N. Koga, K. Morokuma, Chem. Rev. 9 (99) 83; (l) T. Ziegler, Chem. Rev. 9 (99) 65; (m) R.G. Bergman, Adv. Chem. Series 30 (99) ; (n) E.. Wasserman, C.B. Moore, R.G. Bergman, Science 55 (99) 35; (o) R.. Crabtree, Angew. Chem., Int. Ed. Engl. 3 (993) 789; (p) D. Schroder,. Schwarz, Angew. Chem., Int. Ed. Engl. 34 (995) 937; (q) A.J. Lees, A.A. urwoko, Coord. Chem. Rev. 3 (994) 55; (r) B.A. Amdtsen, R.G. Bergman, T.A. Mobley, T.. eterson, Acc. Chem. Res. 8 (995) 54; (s) T. Ziegler, Can. J. Chem. 73 (995) 743; (t) B.A. Arndtsen, R.G. Bergman, Science 70 (995) 970; (u) J.C. Lohrenz,. Jacobsen, Angew. Chem., Int. Ed. Engl. 35 (996) 30. [5] (a) S. Niu, M.B. all, Chem. Rev. 00 (000) 353; (b).e.m. Siegbahn, R.A. Blomberg, Chem. Rev. 00 (000) 4. [6] J.Y. Saillard, R. offmann, J. Am. Chem. Soc. 06 (984) 006. [7] A. Dedieu, Chem. Rev. 00 (000) 543. [8] (a) M. Torrent, M. Sola, G. Frenking, Chem. Rev. 00 (000) 439; (b) G. Frenking, N. Frohlich, Chem. Rev. 00 (000) 77. [9] (a) G.. Loew, D.L. arris, Chem. Rev. 00 (000) 407; (b) J. Alonso, Chem. Rev. 00 (000) 637; (c) J.F. arrison, Chem. Rev. 00 (000) 679. [0] C.S. Cramer, Essentials of Computational Chemistry. Theories and Models, Wiley, New York, 00. [] (a).j. ay, W.R. Wadt, J. Chem. hys. 8 (985) 70; (b) W.R. Wadt,.J. ay, J. Chem. hys. 8 (985) 84; (c).j. ay, W.R. Wadt, J. Chem. hys. 8 (985) 99. [] W. ehre, J.L. Radom,.V.R. Schleyer, J.A. ople, Ab initio Molecular Orbital Theory, Wiley, New York, 986. [3] S. uzinaga, Gaussian Basis Sets for Molecular Calculations, Elsevier Science, Amsterdam, 984. [4] Chem. Rev. 00 (000) The whole issue is devoted to computational transition metal chemistry. [5] (a) C. Gonza lez,.b. Schlegel, J. Chem. hys. 90 (989) 54; (b) C. Gonza lez,.b. Schlegel, J. Chem. hys. 94 (990) 553. [6] M.J. Frisch et al., GAUSSIAN 98, Revision A.9, Gaussian, Inc., ittsburgh, A, 998.