Ps.uci.edu

Organometallics 2008, 27, 1384–1392
Theoretical Studies of the sp2 versus sp3 C-H Bond Activation
Chemistry of 2-Picoline by (C5Me5)2An(CH3)2 Complexes (An ) Th, U)
Ping Yang,* Ingolf Warnke,† Richard L. Martin, and P. Jeffrey Hay* Los Alamos National Laboratory, Los Alamos, New Mexico 87545 ReceiVed September 18, 2007 The C-H bond activation chemistry of (C5Me5)2Th(CH3)2 and (C5Me5)2U(CH3)2 with 2-picoline (2- methylpyridine) is examined with use of density functional techniques. In particular, the differences betweeninsertion into the ortho ring sp2 C-H bond and the methyl sp3 C-H bond are explored. The energies toform the η2-(N,C)-pyridyl products resulting from the activation of the aromatic ring sp2 C-H bond arecalculated as thermodynamic products for both thorium and uranium systems with similar reaction energiesof -15.8 kcal/mol. The products corresponding to insertion into the methyl sp3 C-H bond are found tobe higher in energy by 3.5 and 5.4 kcal/mol for Th and U, respectively. In the transition states the actinideatom mediates the hydrogen migration from 2-picoline to the leaving methyl group by forming an agosticfive-centered C-H complex. The relative activation energies between sp2 and sp3 C-H bond activationdiffer slightly between Th, ∆Eq(sp2) > ∆Eq(sp3) and U, ∆Eq(sp2) < ∆Eq(sp3). These results are inagreement with the experimental observations that the sp2 insertion product is the thermodynamic productin both cases, but that the sp3 insertion product is the kinetic product in the case of Th.
N-oxide, whereas the corresponding U(IV) complexes activateonly the sp2 C-H bonds in pyridine N-oxide (Scheme 1).1 Later studies7,8 were carried out on 2-picoline (2-methylpy- (C5Me5)2AnR2 (where An ) Th, U; R ) CH3, CH2Ph, Ph) have ridine), which possesses both sp2 and sp3 hybridized C-H proven to be versatile starting materials for the synthesis of a bonds. Deuterium labeling studies demonstrated that the thorium diverse array of actinide organometallic systems containing and uranium (C5Me5)2An(CH3)2 complexes react with 2-picoline An-N bonds such as imido, hydrazonato, and ketimido by different mechanistic reaction pathways. The thorium alkyl complexes, which feature novel electronic properties.1–5 It has complex (C5Me5)2Th(CH3)2 selectively activates a sp3 C-H been observed that complexes of lanthanide, actinide, and bond on the 2-picoline methyl group to give the kinetic R-picolyl transition metal activate hydrocarbon substrates by different mechanisms.6 Recently, Kiplinger and co-workers reported that acts with additional 2-picoline to afford the thermodynamic η2- these actinide alkyl complexes undergo interesting C-H and C-N bond cleavage chemistry with N-heterocycles. For This is in marked contrast with the uranium system that only example, the Th(IV) complexes (C reacts with a sp2 C-H bond on the 2-picoline aromatic ring to give the η2-pyridyl product (C 5Me5)2Th(CH2Ph)2 readily react with the sp2 C-H bonds in pyridine N-oxide and the sp3 C-H bonds in 2,6-lutidine NC5H3], see Scheme 2. The deuterium-labeling studies suggested that the 2-picoline C-H activation chemistry isproceeded by σ-bond metathesis for both the thorium and * Corresponding authors. E-mail: [email protected]; [email protected].
† uranium (C5Me5)2An(CH3)2 complexes.8 Current address: Institut für Physikalische Chemie, Universität Karl- sruhe (TH), Germany.
Density functional theory (DFT) methods have been em- (1) Pool, J. A.; Scott, B. L.; Kiplinger, J. L. J. Am. Chem. Soc. 2005,
ployed to explore actinide-ligand interactions in a variety of complexes, using the current generation of hybrid functionals.
(2) Jantunen, K. C.; Burns, C. J.; Rodriguez, I. C.; Da Re, R. E.; Golden, The structures, thermochemistry, and spectroscopic properties J. T.; Morris, D. E.; Scott, B. L.; Taw, F. L.; Kiplinger, J. L. Organometallics
2004, 23, 4682.
using such approaches provide information to compare with (3) Kiplinger, J. L.; Morris, D. E.; Scott, B. L.; Burns, C. J. Organo- available structural and spectroscopic data.9–16 By comparison, metallics 2002, 21, 3073.
(4) Wiedemann, S. H.; Lewis, J. C.; Ellman, J. A.; Bergman, R. G. J. Am. Chem. Soc. 2006, 128, 2452–2462.
(7) Pool, J. A.; Scott, B. L.; Kiplinger, J. L. J. Alloys Compd. 2006,
(5) Bruno, J. W.; Smith, G. M.; Marks, T. J.; Fair, C. K.; Schultz, A. J.; Williams, J. M. J. Am. Chem. Soc. 1986, 108, 40–56.
(8) Kiplinger, J. L.; Scott, B. L.; Schelter, E. J.; Pool Davis Tournear, (6) This field in transition metal has been extensively reviewed. For J. A. J. Alloys Compd. 2007, 444-445, 477–482.
example: (a) Davies, J. A. W. P. L.; Liebman, J. F.; Greenberg, A., Eds.
(9) Kaltsoyannis, N. Chem. Soc. ReV. 2003, 32, 9–16.
SelectiVe Hydrogencarbon ActiVation: Principles and Progress; VCH (10) Bursten, B. E.; Drummond, M. L.; Li, J. Faraday Discuss. 2003,
Publishers: New York, 1990. (b) Arndtsen, B. A.; Bergman, G. G.; Mobley, 124, 1–24, 457–8.
T. A.; Peterson, T. H. Acc. Chem. Res. 1995, 28, 154–162. (c) Shilov, A. E.;
(11) Clark, A. E.; Martin, R. L.; Hay, P. J.; Green, J. C.; Jantunen, K. C.; Shul'pin, G. B. ActiVation and Catalytic Reactions of Saturated Hydro- Kiplinger, J. L. J. Phys. Chem. A 2005, 109, 5481–5491.
carbons in the Presence of Metal Complexes; Kluwer: Boston, MA, 2000.
(12) Peralta, J. E.; Batista, E. R.; Scuseria, G. E.; Martin, R. L. J. Chem. (d) Crabtree, R. H. J. Chem. Soc., Dalton Trans. 2001, 2437–2450. (e)
Theory Comput. 2005, 1, 612–616.
Labinger, J. A.; Bercaw, J. E. Nature 2002, 417, 507–514. (f) Goldberg,
(13) Hayton, T. W.; Boncella, J. M.; Scott, B. L.; Palmer, P. D.; Batista, K. I.; Goldman, A. S., Eds. ActiVation and Functionalization of C-H Bonds; E. R.; Hay, P. J. Science 2006, 310, 1941–1943.
ACS Symp. Ser. No. 885; American Chemical Society: Washington, DC, (14) Gagliardi, L.; Handy, N. C.; Ioanou, A. G.; Skylaris, C.-K.; Spencer, S.; Willetts, A.; Simper, A. M. Chem. Phys. Lett. 1998, 283, 187–193.
10.1021/om700927n CCC: $40.75  2008 American Chemical Society Publication on Web 03/13/2008 Bond ActiVation Chemistry of (C5Me5)2An(CH3)2 Organometallics, Vol. 27, No. 7, 2008 1385 Scheme 1. (C5Me5)2Th(CH3)2 and (C5Me5)2Th(CH2Ph)2 Readily React with the sp2 C-H Bonds in Pyridine N-Oxide and the sp3
C-H Bonds in 2,6-Lutidine N-Oxide, whereas the Corresponding U(IV) Complexes Only Activate the sp2 C-H Bonds in
Scheme 2. The Thorium Alkyl Complex (C5Me5)2Th(CH3)2 and 2-Picoline React To Give Preferential sp3 C-H Bond Activation
in the Presence of a More Reactive sp2 C-H Bond, while the Analogous Uranium Complex, (C5Me5)2U(CH3)2, Reacts with only
the Ortho 2-Picoline sp2 C-H Bond
rather little information has been available on reaction mech- theory (DFT) techniques. The current approach chose a particular anisms of f-element complexes such as in the aforementioned DFT method that gave excellent thermochemical predictions for C-H activation processes. Most studies focus on the lanthanide actinide halide species20,21 where there are available experimental compounds.17–19 We have performed a computational study of quantities and applied it to the set of model Th and U reactions the competitive sp2 versus sp3 C-H bond activation chemistry discussed above. Our calculations employed the hybrid B3LYP with 2-picoline and (C functional.22,23 The Stuttgart RSC 1997 effective core potential 5Me5)2An(CH3)2 for An and Th as shown in Scheme 2, where the competition between sp2 and sp3 C-H (ECP) was used for the thorium and uranium atoms, where 60 core- activations within the same reactant molecule is examined.
electrons are replaced to account for scalar relativistic effects.24The valence electrons are represented by a contracted [8s/7p/6d/ Th(IV) complexes possess a closed shell electronic ground state 4f] basis; 6-31G* basis sets were used for carbon, hydrogen, and (5f 0), while U(IV) systems represents a high-spin (5f2) system nitrogen. Spin–orbit interactions have not been considered explicitly.
with two unpaired electrons in f-orbitals.
The effects of polarization functions on transition states were tested The products and resulting thermochemistry in these reactions by reoptimizing the transition states by using the 6-31G** basis are compared for the cases of Th(IV) and U(IV), and likely that contains p polarization functions on hydrogen atoms. The newly reaction precursors and transition states are also identified. We found transition states were essentially identical structurally with find that the actinide atom plays a fundamental role during the the ones found by using the 6-31G* basis sets, and thermochemistry hydrogen migration process from 2-picoline to the methyl and reaction barriers differed by less than 0.5 kcal. Harmonic leaving group. Agostic 5-centered transition structures for the vibrational analyses were performed to confirm that structures were actinide C-H activation reaction pathways are reported, to the minima or saddle points and to obtain the thermochemical correc- best of our knowledge, for the first time.
tions to the energy, entropy, and Gibbs free energy. Inclusion ofthese effects results in changes of ∼1 kcal/mol in reaction energiesand leads to the same relative ordering (see the Supporting Information). All thermodynamic data were calculated at the The equilibrium molecular structures as well as transition state standard state (298.15 K and 1 atm). The intrinsic reaction geometries of the actinide species involved in the C-H bond coordinate (IRC) method was used to follow the reaction path in activation reactions were computed with use of density functional (20) Batista, E. R.; Martin, R. L.; Hay, P. J. J. Chem. Phys. 2004, 121,
(15) Gagliardi, L.; Schimmelpfennig, B.; Maron, L.; Wahlgren, U.; (21) Batista, E. R.; Martin, R. L.; Peralta, J. E.; Scuseria, G. E. J. Chem. Willetts, A. Chem. Phys. Lett. 2001, 344, 207–212.
Phys. 2004, 121, 2144–2150.
(16) Kaltsoyannis N.; Hay, P. J.; Li, J ; Blaudeau, J. P.; Bursten, B. E.
(22) Becke, A. D. J. Chem. Phys. 1993, 98, 5648.
Theoretical Studies of the Electronic Structure of Compounds of the Actinide (23) We also tested a few functionals on the relative energies. Since Elements. In The Chemistry of the Actinide and Tranactinide Elements, the geometries are not sensitive to the functional used, the single point 3rd ed.; Morss, L. R., Edelstein, N. M., Fuger, J., Eds.; Springer: Dordrecht, energy difference was compared between thorium transition states TS5A-2A
The Netherlands, 2006; Vol. 3.
(sp3) and TS5A-1A (sp2): -0.54 kcal/mol for B3LYP, -0.71 kcal/mol for
(17) Sherer, E. C.; Cramer, C. J. Organometallics 2003, 22 (8), 1682–
B3PW91, -1.42 kcal/mol for TPSS, and -1.92 kcal/mol for BMK. The chosen functional B3LYP described the essential properties and fundamental (18) Barros, N.; Eisenstein, O.; Maron, L. Dalton Trans. 2006, (25),
chemistry of these compounds but slightly underestimated the energy difference compared to others.
(19) de Almeida, K. J.; Cesar, A. Organometallics 2006, 25, 3407–
(24) Dolg, M.; Stoll, H.; Preuss, H.; Pitzer, R. M. J. Phys. Chem. 1993,
97, 5852.
Organometallics, Vol. 27, No. 7, 2008 Yang et al. Figure 1. Possible isomers of the products resulting from the C-H
bond activation chemistry between (C5H5)2An(CH3)2 (An ) Th,
U) and 2-picoline. These isomers are labeled as 1A*, 1B*, 2A*,
2B*, 3A*, 3B*, 4A*, and 4B* when the ligands are Cp*, i.e.,
(C5Me5)2An(CH3)2 (An ) Th, U).
both directions from the transition state.25,26 Solvation effects wereinvestigated by using polarizable continuum models (PCM) in thesolvent toluene (
) 2.38) employing UFF atomic radii.27,28 Allcalculations were carried out with the Gaussian03 suite of codes.29Analysis of the electronic structure was carried out using the NaturalBond Orbital (NBO) approach.30 Figure 2. The calculated 3D structures for the possible sp2 and
Results and Discussion
sp3 C-H bond activation products. The geometric data obtainedfrom crystal structures (from ref 8) are shown above the calculated In the following sections the DFT results are presented for data, which are in parentheses. The bond lengths are given in the species involved in the reactions of the thorium and uranium angstroms and bond angles are given in degrees.
(C5Me5)2An(CH3)2 complexes shown in Scheme 2. First, thestructures and thermochemistries of the two classes of products products for the C-H activation chemistry of (C5Me5)2Th(CH3)2 arising from sp2 and sp3 C-H bond activation are examined with 2-picoline. Two isomers result from activation of the sp2 for both C5Me5 and model C5H5 ligands. These results help to C-H bond ortho to the ring nitrogen atom, (C5Me5)2Th(CH3)η2- focus the subsequent study of reaction pathways in which only (N,C)-6-Me-NC5H3] (1A* and 1B*). Two additional possibilities
C5H5 ligands were employed. In the second step, adducts of arise from activation of the sp3 C-H bond on the 2-picoline the actinide complexes with 2-picoline were examined as methyl group, which depend on the orientation of the methyl precursors for the reactions. Transition states from these adducts NC5H4] (2A*
to the products were then identified and characterized theoreti- and 2B*). We refer to the analogous species for the Cp ligand
cally, followed by reaction pathway analyses. Finally, the by omitting the asterisk as simply 1A, 1B, etc.
electronic properties of all these species are analyzed.
The optimized ground-state geometries are presented in Figure Products from Competitive sp2 versus sp3 C-H Bond
2 and Tables 1 and 2. Table 1 compares the geometric data for Activation Chemistry. Structures. The geometries of all
the thorium products obtained from available crystal structures possible products for the thorium and uranium sp2 and sp3 C-H with those data calculated for both Cp* and Cp complexes. Table bond activation reactions were fully optimized. Both the Cp* 2 shows the corresponding geometric data for uranium com- (C5Me5 or pentamethylcyclopentadienyl) compounds and sim- plexes. Figure 3 illustrates the numbering and labeling scheme plified model Cp (C5H5 or cyclopentadienyl) complexes were employed for the following discussions regarding the calculated examined. As illustrated in Figure 1, there are four possible adduct, transition state, and product structures. Excellent agree-ment is obtained between the experimental and calculated (25) Gonzalez, C.; Schlegel, H. B. J. Chem. Phys. 1989, 90, 2154.
structures, with An-C bond lengths deviating by 0.05 Å and (26) Gonzalez, C.; Schlegel, H. B. J. Phys. Chem. 1990, 94, 5523.
(27) Miertus, S.; Scrocco, E.; Tomasi, J. Chem. Phys. 1981, 55, 117.
the bond angles deviating by 3°. The largest variation is (28) Cammi, R.; Mennucci, B.; Tomasi, J. J. Phys. Chem. A 2000, 104,
observed for the An-N bond lengths, with deviations of only 0.05 Å. We need to point out that the only discrepancy is the (29) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb, Th1-C2 bond length of the methyl group bound to the Th in M. A.; Cheeseman, J. R.; Montgomery, J. A.; Vreven, J. T.; Kudin, K. N.;Burant, J. C.; Millam, J. M.; Iyengar, S. S.; Tomasi, J.; Barone, V.; the sp3 product. In this case the experimental bond length (2.902 Mennucci, B.; Cossi, M.; Scalmani, G.; Rega, N.; Petersson, G. A.; Å) is well out of the normal range of thorium carbon single Nakatsuji, H.; Hada, M.; Ehara, M.; Toyota, K.; Fukuda, R.; Hasegawa, J.; bond lengths of about 2.5 Å. Aside from this one exception, Ishida, M.; Nakajima, T.; Honda, Y.; Kitao, O.; Nakai, H.; Klene, M.; Li,X.; Knox, J. E.; Hratchian, J. P.; Cross, J. B.; Adamo, C.; Jaramillo, J.; considering that the calculations are for gas-phase geometries Gomperts, R.; Stratmann, R. E.; Yazyev, O.; Austin, A. J.; Cammi, R.; and the X-ray crystal data are for solid-state geometries, the Pomelli, C.; Ochterski, J. W.; Ayala, P. Y.; Morokuma, K.; Voth, G. A.; agreement between the two is quite remarkable.
Salvador, P.; Dannenberg, J. J.; Zakrzewski, V. G.; Dapprich, S.; Daniels,A. D.; Strain, M. C.; Farkas, O.; Malick, D. K.; Rabuck, A. D.; The Th1-C2 bond lengths in the products are slightly Raghavachari, K.; Foresman, J. B.; Ortiz, J. V.; Cui, Q.; Baboul, A. G.; lengthened (∼0.02 Å) compared to the Th-Cmethyl bond distance Clifford, S.; Cioslowski, J.; Stefanov, B. B.; Liu, G.; Liashenko, A.; Piskorz, of 2.483 Å calculated for the starting material model complex P.; Komaromi, I.; Martin, R. L.; Fox, D. J.; Keith, T.; Al-Laham, M. A.;Peng, C. Y.; Nanayakkara, A.; Challacombe, M.; Gill, P. M. W.; Johnson, (C5H5)2ThMe2. This result is not too surprising since the small B.; Chen, W.; Wong, M. W.; Gonzalez, C.; Pople, J. A. Gaussian03, methyl group is being replaced by the more bulky picolyl ligand.
Revision C.02; Gaussian, Inc.: Wallingford, CT, 2004.
The insertion of the picolyl group also causes the angles between (30) Glendening, E. D.; Badenhoop, J., K.; Reed, A. E.; Carpenter, J. E.; the two C5H5 rings to increase marginally by 2–5 Bohmann, J. A.; Morales, C. M.; Weinhold, F. NBO 5.0; Theoretical Chemistry Institute, Univeristy of Wisconsin: Madison, WI, 2001.
metallocene framework of the reactant does not change and both Bond ActiVation Chemistry of (C5Me5)2An(CH3)2 Organometallics, Vol. 27, No. 7, 2008 1387 Table 1. Experimental Data Obtained from Crystal Structures and Calculated Structural Data for the Possible Products Resulting from sp2
and sp3 C-H Bond Activation Chemistry between (C5Me5)2Th(CH3)2 and 2-Picoline
exptl8 1A*
calcd 1A* (1A)
calcd 1B* (1B)
exptl8 2A*
calcd 2A* (2A)
a The labeling scheme is shown in Figure 3, and is adopted by Tables 2-6. b The experimental and optimized geometric parameters with C5Me5 ligands (1A*, 1B*, 2A*, 2B*) are shown. c The optimized geometric parameter with C5H5 ligands (1A, 1B, 2A, 2B) are shown in parentheses.
Table 2. Experimental Data Obtained from Crystal Structures and Calculated Structural Data for the Possible Products Resulting from sp2
and sp3 C-H Bond Activation Chemistry between (C5Me5)2U(CH3)2 and 2-Picoline
exptl8 3A*
calcd 3A* (3A)
calcd. 3B* (3B)
calcd 4A* (4A)
calcd 4B* (4B)
a The labeling scheme is shown in Figure 3, and is adopted by Tables 2-6. b The experimental and optimized geometric parameters with C5Me5 ligands (1A*, 1B*, 2A*, 2B*) are shown. c The optimized geometric parameter with C5H5 ligands (1A, 1B, 2A, 2B) are shown in parentheses.
Table 3. Summary of the Calculated Binding Energies (all values in
kcal/mol) for the Possible Adducts and Reaction Energies Relative
to the Sum of Reactants for the C-H Bond Activation Chemistry
between 2-Picoline and (C5Me5)2An(CH3)2 or (C5H5)2An(CH3)2
(An ) Th, U)
Figure 3. The labeling scheme for calculated adducts, transition
and methyl ligands in the model C5H5 complexes are identical states, and products in the sp2 and sp3 C-H bond activation with the C5Me5 structures with bond lengths within 0.02 Å and chemistry between (C5H5)2An(CH3)2 (An ) Th, U) and 2-picoline.
bond angles within 2 In particular, the structures along the reaction pathways for products °. These errors are within the errors of A series are shown as examples.
experimental measurement. The only noticeable difference isthe plane angle between the two C5H5 rings, which is about 10 ligands stay within the metallocene wedge. The sp2 C-H bond larger than that observed for the more sterically congested activation products (1A and 1B) are η2-(N,C)-pyridyl complexes
bis(pentamethylcyclopentadienyl) system. These results are with the actinide atom lying in the plane of the coordinated provided in tables in the Supporting Information.
picolyl ligand. On the other hand, the sp3 C-H bond activation Thermochemistry. The computed reaction energies showing
products (2A and 2B) have a bent configuration with the actinide
the relative stability of the products for the two insertion center being folded out of the pyridine ring plane. The calculated pathways for Th and U complexes are given in Table 3 and structures for the uranium products are essentially identical with Figure 4. Results are given for both the model C5H5 complexes the thorium complexes with slightly shorter bond lengths, which as well as the C5Me5 complexes studied experimentally. As is consistent with the contraction of ionic radius as one proceeds shown in Table 3, the reaction energies are exothermic in all down the actinide series.
cases and range from -14.70 to -18.25 kcal/mol for the model In general, the model C5H5 complexes perform well in C5H5 complexes and from -7.74 to -15.81 kcal/mol for the reproducing the geometric parameters of the C5Me5 complexes, C5Me5 complexes. The reaction energies for the sp2 and sp3 including reactants and products for both the observed thorium C-H bond activation pathways corresponding to the lowest and uranium chemistry. The bonds for the coordinated picolyl isomer for the thorium and uranium complexes are plotted in Organometallics, Vol. 27, No. 7, 2008 Yang et al. Figure 4. Reaction energies calculated for the sp2 and sp3 C-H
bond activation products from the reaction between (C5Me5)2-
An(CH3)2 or (C5H5)2An(CH3)2 (An ) Th, U) and 2-picoline. The
numbers on bars are the reaction energies for corresponding
compounds.
Figure 6. The calculated 3D structures for the possible adducts
involved in the C-H bond activation chemistry between
(C5H5)2Th(CH3)2 and 2-picoline.
Table 4. Calculated Geometric Parameters for 2-Picoline Adducts
with (C5H5)2An(CH3)2 (An ) Th, U)
Figure 5. The possible adducts involved along the reaction
pathways in the sp2 and sp3 C-H bond activation chemistry 5H5)2An(CH3)2 (An ) Th, U) and 2-picoline.
the Figure 4 for both C5Me5 and C5H5 ligand frameworks. The product 1A* corresponding to the Th sp2 C-H bond activation
product is the most stable structure among all four products considered. It lies 3.34 kcal/mol lower in energy than the isomer 2A* corresponding to the sp3 C-H bond activation product.
The same ordering is obtained for the thorium model C5H5
group encounters unfavorable steric interaction with the methyl complexes, where 1A is 1.67 kcal/mol lower in energy than
groups on C5Me5 ligands resulting in greater congestion and a higher energy configuration. For the uranium complexes, the For the uranium C5Me5 complexes, the same trend is observed relative differences between the alternate products mirror their with 3A* being more stable in energy than 4A* by 5.36 kcal/
thorium counterparts, with 1B* and 2B* higher by 3.11 and
mol. For the C5H5 model compounds, 3A is 2.26 kcal/mol lower
2.61 kcal/mol compared to 3A* and 4A*, respectively. These
in energy than 4A. The energy difference between two products
for uranium is larger than that for thorium (Table 3). Overall,
these results are consistent with the experimental observations
in that the sp2 C-H bond activation products (1A* and 3A*)
are the thermodynamically stable products. Since the same trends
in reaction energies are reproduced by the model C5H5 ligands,
the reaction pathways examined in later sections have employed
C5H5 ligands. Geometrical analysis regarding the differences
between the possible products provides insight into the regi-
oselectivity observed for these reactions. As illustrated in Figure
1, for the thorium complexes, other possible products are 1B*
and 2B*, but each is less stable than 1A* and 2A* by 2.92 and
2.36 kcal/mol, respectively. These numbers reduce to 2.00 and
1.26 kcal/mol for the model C5H5 complexes. In comparison to
1A*, 1B* has longer bond distances of Th-N and average
Th-C to Cp rings, but a shorter Th-C bond length. This is
Figure 7. The calculated 3D structures of the transition states
likely due to steric interactions introduced by the orientation of involved in the sp2 and sp3 C-H bond activation of 2-picoline from the exo methyl group on the picolyl ligand. The picolyl methyl the thorium adduct 5A.
Bond ActiVation Chemistry of (C5Me5)2An(CH3)2 Organometallics, Vol. 27, No. 7, 2008 1389 Figure 8. The calculated 3D structures of the transition states
involved in the sp2 and sp3 C-H bond activation of 2-picoline from
the uranium adduct 6A.
Table 5. Metrical Data from the Calculated Transition States
Structures for the sp2 and sp3 C-H Bond Activation Chemistry
Figure 9.
between (C
Reaction energy diagram for the sp2 and sp3 C-H bond 5Me5)2An(CH3)2 (An ) Th, U) and 2-Picoline
activation chemistry between (C5H5)2Th(CH3)2 and 2-picoline.
Table 6. Activation Energies (all values are in kcal/mol) Relative to
sp3 TS5A-2A
sp2 TS6A-3A
sp3 TS6A-4A
the Sum of Reactants
The optimized three-dimensional structures of the possible numbers reduce to 0.85 and 0.74 kcal/mol for the C adducts for thorium (5A-D), all of which are local minima,
5H5 systems.
Overall, the model C are shown in Figure 6. The summarized geometric parameters 5H5 systems accurately reflect the relative thermochemical relationships among the C for the thorium 2-picoline adducts 5A and 5D and the uranium
5Me5 products, but with smaller energy difference.
2-picoline adducts 6A and 6D are presented in Table 4. For
Adducts of (C
both thorium and uranium systems, the most stable adducts are 5H5)2An(CH3)2 (An ) Th, U) with 2-Picoline.
Although the existence of stable adducts is not required along 5A and 6A, respectively, which are the only adducts calculated
reaction paths leading to C-H bond activation, the possible to be exothermic relative to the (C5H5)2An(CH3)2 and 2-picoline 2-picoline adducts with the model complexes (C (An ) Th, U) were examined to gain mechanistic insights into For all the calculated adducts, the An1-N3 distances are in the C-H bond activation reaction processes as well as to aid the range of 2.8–3.2 Å, which are longer than the An-N in the identification of transition states for the C-H activation distances of the products. The coordination interactions for these reactions. As illustrated in Figure 5, the 2-picoline ligand can adducts are fairly weak resulting in very small binding energies approach the metal center in one of three ways, leading to four and slightly longer (by 0.05 Å) An1-C2 bond lengths. The possible adduct structures: distances between the actinide metal center and the o-H (1) A frontal approach between two methyl groups. There (An · · · H-Csp2) and the closest H atom on the methyl group are two possible orientations for the 2-picoline ring relative to (An · · · H-Csp3) on the coordinated 2-picoline are in the range the plane formed by the actinide metal center and the carbon of 3.3–3.7 Å. Therefore, no substantial interactions exist between atoms on two methyl groups, parallel or perpendicular. The the actinide center and these hydrogen atoms in the adduct perpendicular conformation is preferred (5A and 6A), while the
parallel configuration is unstable due to large unfavorable steric However, it is worth pointing out that for the most stable interactions between the 2-picoline methyl group and the adduct 5A, the Th · · · H-Csp3 distance is 3.299 Å, which is
actinide methyl groups.
significantly shorter (by ∼0.5 Å) than the Th · · · H-Csp2 distance (2) A lateral approach from the side of the metallocene. In of 3.736 Å. In contrast, for 6A the U · · · H-Csp2 and
this case, two stable conformations can also be formed as a U · · · H-Csp3 distances are about the same at 3.455 and 3.401 result of different orientations of the 2-picoline methyl group.
Å, respectively. This might be an indicator that the H atom on Adducts 5B and 6B are formed when the 2-picoline methyl
the 2-picoline methyl group for sp3 activation is more easily group is pointing toward the actinide methyl groups and accessed in the thorium system to produce the kinetic product, complexes 5C and 6C have the 2-picoline methyl group directed
while uranium system does not.
away from the actinide methyl groups.
The energies of the four possible adducts for both thorium (3) A backside approach of the 2-picoline forming the linear and uranium are summarized in Table 3. The bent metallocene metallocene complexes 5D and 6D with the two C5H5 rings
configurations (5A-C, 6A-C) are more favorable than the
being parallel.
parallel or linear metallocene configuration (5D and 6D).
Organometallics, Vol. 27, No. 7, 2008 Yang et al. Complexes 5D and 6D are the most unstable adducts with large
longer than the U-H bond length in (C5Me5)2U(H)(CH3), which positive binding energies, which are 4.7 and 10.9 kcal/mol is 2.044 Å. The migrated H atom sits closer to picoline in the higher in energy, respectively, than the most stable adducts 5A
sp2 transition state and is located at the midpoint of two carbon and 6A. In comparison, the formation of adducts 5B, 5C, 6B,
atoms in the transition state for sp3 activation. The leaving and 6C shows small positive reaction energies (<1.50 kcal/
methyl group binds very weakly with a long coordination bond mol). Since the difference between 5B, 5C, 6B, and 6C is
greater than 2.71 Å.
merely due to the orientation of the 2-picoline ring, the steric We refer to this reaction mechanism as "agostic migration", interactions do not have any significant effects on the formation which is a synergistic nonadditive H-migration process. This of these adducts.
reaction mechanism is very stereospecific. A similar reaction Attempts to find stable adducts for the C5Me5 systems were mechanism has been recently reported for cyclometalation of not successful. Although local minima might not exist, one palladium and iridium.31,32 These combined data suggest that would expect that the configuration along the reaction pathways an "agostic migration mechanism" is operative for the C-H of the two reactant molecules ((C5Me5)2An(CH3)2 and 2-pi- activation in the Th(IV) and U(IV) centers involving a five- coline) would closely resemble structures of the adducts center transition state. This differs slightly from the conventional calculated for the pathways involving the model C5H5 ligand.
metathesis mechanism, which usually involves a four-center Transition States and Reaction Pathways Involved in
transition state and a H center that does not coordinate with its the C-H Activation Chemistry. As the deuterium labeling
diagonal atom, in this case the actinide metal center. It has been studies suggested that the 2-picoline C-H activation chemistry reported for more than 30 years that agostic structures play an proceeded by σ-bond metathesis for both the thorium and important role in transition metal complexes and catalytic uranium (C5Me5)2An(CH3)2 complexes,8 we examined the role reaction pathways.33–44 There are only a few reports regarding of the actinide metal center in the hydrogen atom migration. A agostic structures in actinides,45–49 and practically no mention search for transition states was carried out for both sp2 and sp3 involving actinide reaction mechanisms.
C-H bond activation pathways starting from each adduct toverify if the proposed σ-bond metathesis mechanism was Reaction Pathways for Thorium and Uranium. The
operative. In the discussion that follows, TS refers to the reaction diagram for the sp2 and sp3 C-H activation pathways transition state and the subscript refers to the reactant and involving the (C5H5)2Th(CH3)2 complex is shown in Figure 9.
product. For example, TS
In the initial step, the reactants combine to form the weakly bound 5A-1A represents the transition state that
occurs along the sp2 C-H bond activation pathway starting from (1.80 kcal/mol) adduct complex 5A. Following the sp3 insertion
adduct 5A and giving product 1A.
pathway to form the transition state TS5A-2A (∆Eq ) 21.46 kcal/
Transition States. Figures 7 and 8 show the transition state
mol), the product 2A is produced as a kinetic product and product.
structures involved in the sp2 and sp3 C-H bond activation of Along the sp2 insertion pathway a slightly higher transition state 2-picoline from the most stable adduct structures for thorium TS5A-1A is surmounted (∆Eq ) 22.00 kcal/mol)50 on the way to
(5A) and uranium (6A), respectively. The geometric data for
formation of the thermodynamic product 1A (Table 6).
both systems are listed in Table 5. The most important and As mentioned earlier, it is unlikely that structures resembling common characteristic for these transition states are the agostic 5B, 5C, or 5D are involved in the reaction pathway, since their
structures whereby the actinide center mediates the migration energies are 2–4 kcal/mol higher than those of 5A and also give
of the activated H atom from 2-picoline to leaving methyl group.
rise to higher activation energies for C-H insertion. The For the thorium sp2 and sp3 C-H bond activation chemistry, the Th · · · H-Csp2 and Th · · · H-Csp3 Th-H distances are 2.328 (31) Davies, D. L.; Donald, S. M. A.; Macgregor, S. A. J. Am. Chem. and 2.352 Å, respectively. This can be compared with the Th-H Soc. 2005, 127, 13754–13755.
bond length of 2.116 Å determined for the model complex (32) Davies, D. L.; Donald, S. M. A.; Duaij, O. A.; Macgregor, S. A.; Plleth, M. J. Am. Chem. Soc. 2006, 128, 4210–4211.
5Me5)2Th(H)(CH3). This indicates that a significant bonding interaction exists between the thorium metal center and the (33) Li, J. L.; Geng, C. Y.; Huang, X. R.; Zhang, X.; Sun, C. C.
Organometallics 2007, 26, 2203–2210.
migrating H atom. As illustrated in Figure 7, the H atom is not (34) Brookhart, M.; Green, M. L. H.; Parkin, G. Proc. Natl. Acad. Sci. completely transferred from the carbon atom to the thorium atom U.S.A. 2007, 104, 6908–6914.
and the activated carbon-hydrogen bonds are significantly length- (35) Cotton, F. A. Inorg. Chem. 2002, 41, 643–658.
(36) Paul, A.; Musgrave, C. Organometallics 2007, 26, 793–809.
ened in all the transition states. In the sp2 C-H bond activation (37) Bi, S.; Lin, Z.; Jordan, R. F. Organometallics 2004, 23, 4882–
reaction profile, the bond distance between the thorium metal center and the carbon atom of the leaving methyl group (Th-C5) is (38) Sakaki, S.; Takayama, T.; Sumimoto, M.; Sugimoto, M. J. Am. Chem. Soc. 2004, 126, 3332–3348.
strongly deformed from 2.538 Å in the adduct 5A to greater than
(39) Ben-Air, E.; Cohen, R.; Gandelman, M.; Shimon, L. J. W.; Martin, 2.77 Å in the transition state (TS5A-1A). In TS5A-1A, the H atom is
J. M. L.; Milstein, D. Organometallics 2006, 25, 3190–3210.
also slightly closer to the 2-picoline (H · · · C (40) Niu, S. Q.; Hall, M. B. Chem. ReV. 2000, 100, 353–405.
sp2 ) 1.340 Å) than to the leaving Th-methyl group (H · · · C (41) Lersch, M.; Tilset, M. Chem. ReV. 2005, 105, 2471–2526.
sp3 ) 1.517 Å) (Figure 7). To (42) Ritleng, V.; Sirlin, C.; Pfeffer, M. Chem. ReV. 2002, 102, 1731–
compare, in the thorium-mediated sp3 C-H bond activation chemistry, the calculated transition state TS5A-2A shows that the H
(43) Rybtchinski, B.; Cohen, R.; Ben-David, Y.; Martin, J. M. L.; atom is approximately located equally between the 2-picoline Milstein, D. J. Am. Chem. Soc. 2003, 125, 11041–11050.
(44) Gilbert, T. M.; Hristov, I.; Ziegler, T. Organometallics 2001, 20,
methyl carbon (H · · · Csp3 ) 1.431 Å) and the leaving Th-methyl group (H · · · Csp3 ) 1.450 Å).
(45) Cruz, C.; Emslie, D.; Harrington, L.; Britten, J.; Robertson, C.
A similar situation is observed for the uranium system.
Organometallics 2007, 26, 692–701.
(46) Summerscales, O.; Cloke, F.; Hitchcock, P.; Green, J. C.; Hazari, Starting from 6A, the lowest energy transition states for both
N. Science 2006, 311, 829–831.
sp2 and sp3 C-H bond activation of 2-picoline are the five- (47) Andrews, L.; Lyon, J. T. Inorg. Chem. 2006, 45, 1847–1852.
member agostic structures, TS
(48) Clark, D. L.; Grumbine, S. K.; Scott, B. L.; Watkin, J. G.
6A-3A and TS6A-4A, respectively
Organometallics 1996, 15, 949–957.
(Figure 8). The U · · · H-Csp2 and U · · · H-Csp3 U-H distances (49) Barnhart, D.; Clark, D. L.; Grumbine, S. K.; Watkin, J. G. Inorg. are 2.247 and 2.258 Å, respectively, in transition states and Chem. 1995, 34, 1695–1699.
Bond ActiVation Chemistry of (C5Me5)2An(CH3)2 Organometallics, Vol. 27, No. 7, 2008 1391 Figure 11. NPA charges on the actinide metal center.
Figure 10. Reaction energy diagram for the sp2 and sp3 C-H bond
activation chemistry between (C5H5)2U(CH3)2 and 2-picoline.
transition state structures and reaction energy profile for thesespecies are given in Figure S2 in the Supporting Information.
The calculated reaction pathway for insertion of the adduct 6A of (C5H5)2U(CH3)2 into the picoline C-H bonds is illustrated
in Figure 10. In contrast to what was found in the thorium
reaction, the sp3 activation (TS6A-4A) is associated with a higher
activation energy than sp2 activation (TS6A-3A) by 1.24 kcal/
mol. In the case of uranium, the thermodynamically favored
product is also the kinetically favored one, leading to the sp2
product as the only observed product. We also note a larger
difference between the sp2 and sp3 activation barrier (1.24 kcal/
mol) in uranium compared with thorium (0.5 kcal/mol). Despite
these small differences, the result is nevertheless consistent with
experimental data since no kinetic sp3 product is observed in
the uranium reaction.
The overall results regarding the trends in sp2 vs sp3 activation Figure 12. Orbital population on the actinide metal center.
pathways are encouraging in that they delineate the experimentalobservations to date regarding the difference between Th and structures and reaction energy profile for these species are given U complexes. With respect to the conclusions from previous in Figure S3 in the Supporting Information.
studies12,20,21 we first note that these activation energies are Electronic Properties and Solvent Effects. In this section
calculated for the model C5H5 systems. Larger differences might we examine the electronic properties of the reactants, products, be anticipated between the activation barriers for sp2 versus sp3 and transition states in order to understand the similarities and for C5Me5 systems. Indeed we observed larger energy differ- differences found between the Th and U complexes. Although ences between product isomers using C5Me5 than with the model the differences in thermochemistry and kinetic barriers are C5H5 ligand.
consistent with experiment, the differences are subtle involving We also investigated alternative pathways involving the very small energy changes. We have found no single dominant adducts 6B, 6C, and 6D. These were found to give higher
factor that explains these results.
barriers, the same as in the thorium system. The transition state We begin by discussing the charges determined from the NBO analysis. The NPA charges for the reactants, adducts, (50) We have used electronic energies here to examine the reaction transition states, and products for the sp2 and sp3 insertion pathways. The quantitative results and the trends between Th and U would pathways are shown in Figure 11 for both Th and U cases. For have been equally well given by the enthalpies at 298 K (see Tables S5 Th the natural charges are found to be 2.11, 1.93, and 1.95 and S6 in the Supporting Information). The Gibbs free energies at 298 Kalso give consistent predictions as given above with the exception of the respectively for the reactant, the sp2 product (1A), and the sp3
relative ordering of the Th sp2 and sp3 activation barriers (-1.4 kcal/mol) product (2A). As is typically the case, these charges are much
compared to ZPE corrected energies (+0.7 kcal/mol) The quantities in smaller than those for the formal +4 oxidation state because of context use electronic energies without zero-point corrections. Because thelow rotation mode of the methyl group on the picoline ring is constrained ligand donation of electron density into primarily metal 6d and in the sp3 activation path, the frequencies of TS5A-2A(sp3) are higher than
5f orbitals. The 6d and 5f populations are depicted in Figure that of TS5A-1A(sp2). As a result, the lower entropy contribution of
12 (actual values may be found in Table S4 in the Supporting TS5A-2A(sp3) leads to a higher Gibbs free energy (G ) H - TS) compared
to TS
Information). The thorium 6d populations are 1.29 and 1.38 5A-1A(sp2). This observation also holds for the uranium system as well.
However, the entropy contributions are calculated for gas-phase species.
electrons for the TST and product along the sp2 pathway and For reactions in solution, as in the present calculations, the translational, 1.22 and 1.26 electrons for TST and product along the sp3 rotational, and low-frequency vibrational contributions to the entropy can pathway. Corresponding values for 5f populations are 0.28 and be reduced. And as a result, the gas-phase free energies may overestimatethe reaction barrier. For a recent discussion, see ref 38.
0.32 along sp2 and 0.16 and 0.15 along sp3 pathways. As seen Organometallics, Vol. 27, No. 7, 2008 Yang et al. in Figure 11, the 6d and 5f populations are slightly higher for spectively, by 2-picoline are shown in Figures 9 and 10, the Th complexes along the sp2 compared to the sp3 pathway.
respectively. In summary, the reaction begins with the formation For U similar trends can also be seen (Figure 11) in the NPA of a weakly bound or slightly unbound adduct in the case of charge distributions of 1.91, 1.69, and 1.75 in reactant, sp2 the model C5H5 and the C5Me5 complexes, respectively. This product, and sp3 product, respectively, resulting in roughly 0.2 is followed by the activation of adjacent C-H bonds by the higher electron population on the metal center when compared An-center leading to an agostic transition state. The origin of to Th. The U 6d populations are comparable to Th, but with the regioselectivity rests in the highly ordered nature of the comparable values for both transition states and slightly higher transition states. The calculations find the sp2-activated structure 6d population for the sp2 product. The higher 5f populations to be the thermodynamically favored one for both actinide for U reflect the two unpaired electrons in 5f orbitals. After metals. Despite many common features found between thorium subtracting the population for these two electrons, the remaining and uranium systems, including similar geometries of products, 5f populations of approximately 0.7 are higher than the 5f adducts, and the agostic transition states, the relative activation populations of 0.3 in the Th analogues.
energies between sp2and sp3 activation differ slightly: Th, As noted earlier, the sp2 product is more stable than the sp3 ∆Eq(sp2) > ∆Eq(sp3); U, ∆Eq(sp2) < ∆Eq(sp3).
product for both the thorium and uranium systems. From a These results are in agreement with the experimental observa- simply geometric perspective (Figure 2), in the sp2 product (1A
tions that the sp2 insertion product is the thermodynamic product and 3A) the actinide atom is in the picoline ring plane but in
in both cases, but that the sp3 insertion product is the kinetic the sp3 product (2A and 4A) it does not lie in the plane. This
product in the case of Th.
planar configuration can result in stronger donor-acceptorinteractions between the CdN bond on the picolyl ring and the The differences in the competition between sp2 and sp3 5f and 6d orbitals on the actinide center. This could account pathways for the two An-centers are subtle. Electronic property for the fact that the sp2 product 1A is the thermodynamically
analyses confirm that both 5f and 6d orbitals are involved in favored pathway.
the product bonding and that oxidative addition is not associated The possible influence of solvation effects on the reaction with the reaction process. There does not seem to be a single mechanisms of Th and U complexes was examined by using electronic factor accounting for the differences in reactivity. On the polarizable Continuum Models (PCM).27,28 One does not the basis of the combination of labeling and structural and anticipate large changes as the reactions were carried out computational information, we propose an "agostic migration" experimentally in the nonpolar solvent toluene with a dielectric cyclomelatation mechanism for C-H activation of N-hetero- constant of 2.38. The solvent effects on the overall reaction cycles by actinocene complexes.
thermochemistry and on the reaction barrier were examined.
The difference with (and without) solvent effects on thermo- Acknowledgment. We gratefully thank Jaqueline L.
chemistry for the sp2 vs sp3 pathways was 1.6 (1.7) kcal/mol Kiplinger for her helpful discussions and assistance in for Th and 3.5 (2.3) kcal/mol for U. The difference between manuscript publication. This work at Los Alamos National barriers for the sp2 and sp3 pathways at their respective transition Laboratory was supported by the LANL Glenn T. Seaborg states was -0.64 (-0.54) kcal/mol for Th and +0.9 (+1.23) Institute for Transactinium Science (Postdoctoral fellowship kcal/mol for U, and the overall barriers changed by less than 1 to P.Y. and summer research fellowship to I.W.), the LANL kcal/mol. The solvation effects are of the same degree because Laboratory Directed Research and Development Program, the complexes along the reaction pathway have similar dipole and the Division of Chemical Sciences, Office of Basic moments. The solvent-induced dipole moments are about 15% Energy Sciences, U.S. Department of Energy under the uniformly larger for reactants, adducts, transition states, and Heavy Element Chemistry Program at Los Alamos National products for both Th and U systems. Overall the role of solvent Laboratory. Los Alamos National Laboratory is operated by effects does not alter any of the previous conclusions in the Los Alamos National Security, LLC, for the National Nuclear previous sections regarding the thermodynamic or kinetic Security Administration of U.S. Department of Energy under pathways for the C-H insertion reactions.
Supporting Information Available: Tables of experimental data
The current theoretical study provides some insight into the obtained from crystal structures and calculated structural data, aspects of C-H activation chemistry involving the actinide charge distributions, natural electron configurations, reaction ener-gies, and HOMO-LUMO data, as well as figures showing transition metal center. The results of density functional theoretical states and reaction energy surfaces. This material is available free exploration are consistent with reported experimental results.
of charge via the Internet at http://pubs.acs.org.
The plausible reaction pathways for the insertion into the sp2and sp3 bonds of (C5H5)2Th(CH3)2 and (C5H5)2U(CH3)2, re-

Source: http://ps.uci.edu/~ffgroup/lib/exe/fetch.php/pubs/publications/yang08a.pdf