Intermolecular C-H bond activation reactions promoted by transient titanium alkylidynes. Synthesis, reactivity, kinetic, and theoretical studies of the Ti=C linkage

Article abstract of DOI:10.1021/ja070989q

The neopentylidene-neopentyl complex (PNP)Ti=CH^tBu(CH ₂^tBu) (2; PNP^- = N[2-P(CHMe₂) ₂-4-methylphenyl]₂), prepared from the precursor (PNP)Ti=CH^tBu(OTf) (1) and LiCH₂^tBu, extrudes neopentane in neat benzene under mild conditions (25 °C) to generate the transient titanium alkylidyne, (PNP)Ti=C^tBu (A), which subsequently undergoes 1,2-CH bond addition of benzene across the Ti=C linkage to generate (PNP)Ti=CH^tBu(C₆H₅) (3). Kinetic, mechanistic, and theoretical studies suggest the C-H activation process to obey pseudo-first-order in titanium, the α-hydrogen abstraction to be the rate-determining step (KIE for 2/2-d₃ conversion to 3/3-d₃ = 3.9(5) at 40 °C) with activation parametete ΔH^? = 24(7) kcal/mol and ΔS^? = -2(3) cal/mol·K, and the post-rate-determining step to be C-H bond activation of benzene (primary KIE = 1.03(7) at 25 °C for the intermolecular C-H activation reaction in C ₆H₆ vs C₆D₆). A KIE of 1.33(3) at 25 °C arose when the intramolecular C-H activation reaction was monitored with 1,3,5-C₆H₃D₃. For the activation of aromatic C-H bonds, however, the formation of the σ-complex becomes rate-determining via a hypothetical intermediate (PNP)Ti=C^tBu(C ₆H₅), and C-H bond rupture is promoted in a heterolytic fashion by applying standard Lewis acid/base chemistry. Thermolysis of 3 in C₆D₆ at 95 °C over 48 h generates 3-d₆, thereby implying that 3 can slowly equilibrate with A under elevated temperatures with k = 1.2(2) x 10^-5 s^-1, and with activation parameters ΔH^? = 31(16) kcal/mol and ΔS^? = 3(9) cal/mol·K. At 95 °C for one week, the EIE for the 2 → 3 reaction in 1,3,5-C₆H₃D ₃ was found to be 1.36(7). When 1 is alkylated with LiCH ₂SiMe₃ and KCH₂Ph, the complexes (PNP)Ti=CH^tBu(CH₂SiMe₃) (4) and (PNP)Ti=CH ^tBu(CH₂Ph) (6) are formed, respectively, along with their corresponding tautomers (PNP)Ti=CHSiMe₃(CH₂^tBu) (5) and (PNP)Ti=CHPh(CH₂^tBu) (7). By means of similar alkylations of (PNP)Ti=CHSiMe₃(OTf) (8), the de-generate complex (PNP)Ti=CHSiMe₃(CH₂SiMe₃) (9) or the non-degenerate alkylidene-alkyl complex (PNP)Ti=CHPh(CH₂SiMe ₃) (11) can also be obtained, the latter of which results from a tautomerization process. Compounds 4/5 and 9, or 6/7 and 11, also activate benzene to afford (PNP)Ti=CHR(C₆H₅) (R = SiMe₃ (10), Ph (12)). Substrates such as FC₆H₅, 1,2-F ₂C₆H₄, and 1,4-F₂C₆H ₄ react at the aryl C-H bond with intermediate A, in some cases regioselectively, to form the neopentylidene-aryl derivatives (PNP)Ti=CH ^tBu(aryl). Intermediate A can also perform stepwise alkylidene-alkyl metatheses with 1,3,5-Me₃C₆H₃, SiMe ₄, 1,2-bis(trimethylsilyl)alkyne, and bis(trimethylsilyl)ether to afford the titanium alkylidene-alkyls (PNP)Ti=CHR(R′) (R = 3,5-Me ₂C₆H₂, R′ = CH₂-3,5-Me ₂C₆H₂; R = SiMe₃, R′ = CH ₂SiMe₃; R = SiMe₂C=CSiMe₃, R′ = CH₂SiMe₂C=CSiMe₃; R = SiMe ₂OSiMe₃, R′ = CH₂SiMe ₂OSiMe₃).

Full text of DOI:10.1021/ja070989q

Published on Web 06/26/2007
Intermolecular C-H Bond Activation Reactions Promoted by
Transient Titanium Alkylidynes. Synthesis, Reactivity, Kinetic,
and Theoretical Studies of the TitC Linkage
Brad C. Bailey, Hongjun Fan, John C. Huffman, Mu-Hyun Baik,* and
Daniel J. Mindiola*
Contribution from the Molecular Structure Center and School of Informatics, Department of
Chemistry, Indiana UniVersity, Bloomington, Indiana 47405
Received February 11, 2007; E-mail: mbaik@indiana.edu; mindiola@indiana.edu
t
t
-
Abstract: The neopentylidene-neopentyl complex (PNP)TidCH Bu(CH
2
Bu) (2; PNP ) N[2-P(CHMe
2
2
) -
t
t
4
-methylphenyl]
2
), prepared from the precursor (PNP)TidCH Bu(OTf) (1) and LiCH
in neat benzene under mild conditions (25 °C) to generate the transient titanium alkylidyne, (PNP)TitC Bu
2
Bu, extrudes neopentane
t
(
A), which subsequently undergoes 1,2-CH bond addition of benzene across the TitC linkage to generate
t
(PNP)TidCH Bu(C
6 5
H ) (3). Kinetic, mechanistic, and theoretical studies suggest the C-H activation process
to obey pseudo-first-order in titanium, the R-hydrogen abstraction to be the rate-determining step (KIE for
q
q
2
-
)
/2-d
3
conversion to 3/3-d
2(3) cal/mol‚K, and the post-rate-determining step to be C-H bond activation of benzene (primary KIE
1.03(7) at 25 °C for the intermolecular C-H activation reaction in C vs C ). A KIE of 1.33(3)
. For the
3
) 3.9(5) at 40 °C) with activation parameters ∆H ) 24(7) kcal/mol and ∆S )
6
H
6
6 6
D
6 3 3
at 25 °C arose when the intramolecular C-H activation reaction was monitored with 1,3,5-C H D
activation of aromatic C-H bonds, however, the formation of the σ-complex becomes rate-determining via
t
a hypothetical intermediate (PNP)TitC Bu(C
6
H
5
), and C-H bond rupture is promoted in a heterolytic fashion
at 95 °C over 48 h generates
, thereby implying that 3 can slowly equilibrate with A under elevated temperatures with k ) 1.2(2) ×
6 6
by applying standard Lewis acid/base chemistry. Thermolysis of 3 in C D
3
1
-d
6
0
-
5
-1
q
q
s
, and with activation parameters ∆H ) 31(16) kcal/mol and ∆S ) 3(9) cal/mol‚K. At 95 °C for one
was found to be 1.36(7). When 1 is alkylated with
week, the EIE for the 2 f 3 reaction in 1,3,5-C
6 3 3
H D
t
t
LiCH SiMe and KCH Ph, the complexes (PNP)TidCH Bu(CH
2
3
2
2
SiMe
3
) (4) and (PNP)TidCH Bu(CH
2
Ph) (6)
tBu) (5) and
(OTf) (8), the de-
) (9) or the non-degenerate alkylidene-alkyl complex
) (11) can also be obtained, the latter of which results from a tautomerization
process. Compounds 4/5 and 9, or 6/7 and 11, also activate benzene to afford (PNP)TidCHR(C ) (R )
SiMe (10), Ph (12)). Substrates such as FC , 1,2-F , and 1,4-F react at the aryl C-H bond
are formed, respectively, along with their corresponding tautomers (PNP)TidCHSiMe
3
(CH
2
t
(
PNP)TidCHPh(CH
2
Bu) (7). By means of similar alkylations of (PNP)TidCHSiMe
(CH SiMe
3
generate complex (PNP)TidCHSiMe
PNP)TidCHPh(CH SiMe
3
2
3
(
2
3
6 5
H
3
6
H
5
C H
2 6 4
2 6 4
C H
with intermediate A, in some cases regioselectively, to form the neopentylidene-aryl derivatives
t
(
PNP)TidCH Bu(aryl). Intermediate A can also perform stepwise alkylidene-alkyl metatheses with 1,3,5-
Me
alkyls (PNP)TidCHR(R′) (R ) 3,5-Me
SiMe CtCSiMe , R′ ) CH SiMe CtCSiMe
3
C
6
H
3
, SiMe
4
, 1,2-bis(trimethylsilyl)alkyne, and bis(trimethylsilyl)ether to afford the titanium alkylidene-
, R′ ) CH -3,5-Me ; R ) SiMe , R′ ) CH SiMe ; R )
; R ) SiMe OSiMe , R′ ) CH SiMe OSiMe ).
2 6
C
H
2
2
2
C
H
6 2
3
2
3
2
3
2
2
3
2
3
2
2
3
1
. Introduction
that lack energetically accessible orbitals or binding sites to
facilitate the activation process, unlike most substrates that are
easier to activate, given their inherent binding affinity. This
problem is often circumvented by preparing highly unsaturated
metal complexes that are capable of interacting with the C-H
bonds of the substrate. The most common routes to C-H
The controlled and selective activation of inert C-H bonds
of alkanes and arenes, and their functionalization to commodity
products, is one of the major challenges that organotransition
metal chemistry faces today, and constitutes an area of intense
1
-4
study. Part of the challenge lies within the inert C-H bonds
5
activation involve either discrete oxidative addition or σ-bond
6
-9
metathesis.
The former reaction typically implicates late
(
1) (a) Crabtree, R. H. Chem. ReV. 1985, 85, 245-69. (b) Bergman, R. G.
Science 1984, 223, 902-8. (c) Graham, W. A. G. J. Organomet. Chem.
7
transition metal complexes, while the latter often applies
1
986, 300, 81-91. (d) Halpern, J. Inorg. Chim. Acta 1985, 100, 41-8. (e)
Crabtree, R. H.; Hamilton, D. G. AdV. Organomet. Chem. 1988, 28, 299-
3
38. (f) Jones, W. D.; Feher, F. J. Acc. Chem. Res. 1989, 22, 91-100. (g)
(4) Arndtsen, B. A.; Bergman, R. G.; Mobley, T. A.; Peterson, T. H. Acc.
Chem. Res. 1995, 28, 154-62.
(5) (a) Jones, W. D.; Hessell, E. T. J. Am. Chem. Soc. 1993, 115, 554-62. (b)
Wang, C.; Ziller, J. W.; Flood, T. C. J. Am. Chem. Soc. 1995, 117, 1647-
8.
Shilov, A. E.; Shul’pin, G. B. Chem. ReV. 1997, 97, 2879-932. (h) Crabtree,
R. H. Dalton Trans. 2001, 2437-50.
(
(
2) Duncan, A. P.; Bergman, R. G. Chem. ReV. 2002, 2, 431-45.
3) Labinger, J. A.; Bercaw, J. E. Nature 2002, 417, 507-14.
10.1021/ja070989q CCC: $37.00 © 2007 American Chemical Society
J. AM. CHEM. SOC. 2007, 129, 8781-8793
9
8781

A R T I C L E S
Bailey et al.
6
lanthanides, actinides, or early transition metals. Radical bond
catalysts, such processes are often dominated by radical
pathways encompassing H-atom removal, thus rendering the
C-H activation step dependent on the C-H bond strength
10
11
homolysis, electrophilic activation, and 1,2-additions across
a highly polarized metal-ligand multiple bond2
,4,12-29
encom-
3
pass the remaining forms of C-H bond activation reactions
itself. As a result, exploring C-H activation reactions with
known to date.
systems that avoid radical pathways would constitute an
attractive paradigm, since selectivity would not be entirely
dominated by the strength of the C-H bond. In a closed-shell
formalism, transition metal alkylidenes and imides are an
important class of ligands, given their implication not only in
While a common goal of C-H activation chemistry is to
functionalize the arene or alkane substrate to a synthetically
useful product, finding a species that can initiate the C-H
activation under controlled conditions can be regarded as one
of the most difficult steps, especially if such a reaction can be
performed selectively. Even though most existing transforma-
tions of alkane C-H bonds can be done by applying metal oxide
2
,30
catalysis
but also in intermolecular alkane/arene C-H
2
,21
activation, in some cases regioselectively. The latter type of
transformation can be perhaps traced back to Bercaw and
9
,18,31
Rothwell,
who found that transient early transition metal
alkylidenes could not only participate in metathesis-type
chemistry but also engage in the intramolecular C-H activation
(
6) (a) Guram, A. S.; Jordan, R. F.; Taylor, D. F. J. Am. Chem. Soc. 1991,
1
13, 1833-5. (b) Thompson, M. E.; Baxter, S. M.; Bulls, A. R.; Burger,
B. J.; Nolan, M. C.; Santarsiero, B. D.; Schaefer, W. P.; Bercaw, J. E. J.
Am. Chem. Soc. 1987, 109, 203-19. (c) Watson, P. L.; Parshall, G. W.
Acc. Chem. Res. 1985, 18, 51-6. (d) Fendrick, C. M.; Marks, T. J. J. Am.
Chem. Soc. 1986, 108, 425-37.
of the ancillary ligand via 1,2-addition reactions across the Md
CHR linkage (M ) Ti, R ) H; M ) Ta, R ) SiMe3).1
8,19
To
our knowledge, these results represents the first reported cases
(
7) Burger, P.; Bergman, R. G. J. Am. Chem. Soc. 1993, 115, 10462-3.
8) (a) Watson, P. L. J. Am. Chem. Soc. 1983, 105, 6491-3. (b) Vidal, V.;
Theolier, A.; Thivolle-Cazat, J.; Basset, J.-M.; Corker, J. J. Am. Chem.
Soc. 1996, 118, 4595-602. (c) Niccolai, G. P.; Basset, J.-M. Appl. Catal.
A 1996, 146, 145-56.
(
of an alkylidene endorsing a C-H activation step. Contrary to
15,19,20,32-34
intramolecular 1,2-CH activation reactions,
intermo-
lecular 1,2-addition of a C-H bond across a metal-ligand
multiple bond is a far more scant but desired phenomenon,
and such a transformation has been narrowly confined to
(
9) Rothwell, I. P. Acc. Chem. Res. 1988, 21, 153-9.
(
10) (a) Sherry, A. E.; Wayland, B. B. J. Am. Chem. Soc. 1990, 112, 1259-61.
(
5
b) Wayland, B. B.; Ba, S.; Sherry, A. E. J. Am. Chem. Soc. 1991, 113,
305-11.
11) Stahl, S.; Labinger, J. A.; Bercaw, J. E. Angew. Chem., Int. Ed. 1998, 37,
2,13,15-17,22,24,27
12,21,23,25,26,28
MdNR
and MdCHR
functionalities.
(
2
181-92.
The first examples of such a process were reported by
(
(
12) van der Heijden, H.; Hessen, B. Chem. Commun. 1995, 145-6.
13) Cummins, C. C.; Baxter, S. M.; Wolczanski, P. T. J. Am. Chem. Soc. 1988,
1
3
22
Wolczanski and Bergman, who demonstrated that un-
saturated group 4 imides can trigger the activation of a
variety of hydrocarbons, including methane in the case of
Wolczanski’s system. C-H activation across polarized
metal-ligand multiple bonds was further expanded to the
1
10, 8731-3.
(
14) (a) Howard, W. A.; Waters, M.; Parkin, G. J. Am. Chem. Soc. 1993, 115,
4
5
3
917-18. (b) Lee, S. Y.; Bergman, R. G. J. Am. Chem. Soc. 1995, 117,
877-8. (c) Parkin, G.; Bercaw, J. E. J. Am. Chem. Soc. 1989, 111, 391-
. (d) Cook, G. K.; Mayer, J. M. J. Am. Chem. Soc. 1994, 116, 1855-68.
(
e) Legzdins, P.; Veltheer, J. E.; Young, M. A.; Batchelor, R. J.; Einstein,
F. W. B. Organometallics 1995, 14, 407-17.
25
23
12
alkylidene series by Girolami, Gibson, Hessen, and
Legzdins.
(
15) (a) Cummins, C. C.; Schaller, C. P.; Van Duyne, G. D.; Wolczanski, P.
T.; Chan, A. W. E.; Hoffmann, R. J. Am. Chem. Soc. 1991, 113, 2985-
2
1,26,28
In this respect, the latter type of systems are
9
4. (b) Schaller, C. P.; Wolczanski, P. T. Inorg. Chem. 1993, 32, 131-44.
unique in the sense that they provide an opportunity to
mechanistically study new and globally important reactions
(
c) de With, J.; Horton, A. D. Angew. Chem., Int. Ed. 1993, 32, 903-5.
(d) Walsh, P. J.; Hollander, F. J.; Bergman, R. G. Organometallics 1993,
1
2
1
2, 3705-23. (e) Royo, P.; Sanchez-Nieves, J. J. Organomet. Chem.
000, 597, 61-8. (f) Lee, S. Y.; Bergman, R. G. J. Am. Chem. Soc. 1995,
17, 5877-8. (g) Schaller, C. P.; Cummins, C. C.; Wolczanski, P. T. J.
35
such as alkene and alkane metathesis. From another
applied perspective, C-H activation reactions involving
metal-ligand multiply bonded functionalities such as the
terminal oxo have also been studied employing bioinorganic
Am. Chem. Soc. 1996, 118, 591-611. (h) Polse, J. L.; Andersen, R. A.;
Bergman, R. G. J. Am. Chem. Soc. 1998, 120, 13405-14. (i) Schafer, D.
F., II; Wolczanski, P. T. J. Am. Chem. Soc. 1998, 120, 4881-2. (j) Hoyt,
H. M.; Forrest, M. E.; Bergman, R. G. J. Am. Chem. Soc. 2004, 126,
3
6
4,37
mimics, and have been implicated in monooxygenases.
1
018-9.
(
(
16) Bennett, J. L.; Wolczanski, P. T. J. Am. Chem. Soc. 1994, 116, 2179-80.
17) (a) Schaller, C. P.; Bonanno, J. B.; Wolczanski, P. T. J. Am. Chem. Soc.
(30) Schrock, R. R. Chem. ReV. 2002, 102, 145-79.
1
994, 116, 4133-4. (b) Cundari, T. R.; Klinckman, T. R.; Wolczanski, P.
(31) Chamberlain, L.; Rothwell, I. P.; Huffman, J. C. J. Am. Chem. Soc. 1982,
104, 7338-40.
T. J. Am. Chem. Soc. 2002, 124, 1481-7.
(
(
(
18) McDade, C.; Green, J. C.; Bercaw, J. E. Organometallics 1982, 1, 1629-
(32) (a) van Doorn, J. A.; van der Heijden, H.; Orpen, A. G. Organometallics
1994, 13, 4271-7. (b) van Doorn, J. A.; van der Heijden, H.; Orpen, A.
G. Organometallics 1995, 14, 1278-83.
3
4.
19) Chamberlain, L. R.; Rothwell, I. P.; Huffman, J. C. J. Am. Chem. Soc.
1
986, 108, 1502-9.
(33) (a) Duncalf, D. J.; Harrison, R. J.; McCamley, A.; Royan, B. W. Chem.
Commun. 1995, 2421-2. (b) Minhas, R. K.; Scoles, L.; Wong, S.;
Gambarotta, S. Organometallics 1996, 15, 1113-21. (c) Deckers, P. J.
W.; Hessen, B. Organometallics 2002, 21, 5564-75. (d) Kickham, J. E.;
Guerin, F.; Stephan, D. W. J. Am. Chem. Soc. 2002, 124, 11486-94. (e)
Hanna, T. E.; Keresztes, I.; Lobkovsky, E.; Bernskoetter, W. H.; Chirik,
P. J. Organometallics 2004, 23, 3448-58.
20) Couturier, J. L.; Paillet, C.; Leconte, M.; Basset, J. M.; Weiss, K. Angew.
Chem. 1992, 31, 628-31.
21) Pamplin, C. B.; Legzdins, P. Acc. Chem. Res. 2003, 36, 223-33.
22) Walsh, P. J.; Hollander, F. J.; Bergman, R. G. J. Am. Chem. Soc. 1988,
(
(
1
10, 8729-31.
(
(
(
(
23) Coles, M. P.; Gibson, V. C.; Clegg, W.; Elsegood, M. R. J.; Porrelli, P. A.
Chem. Commun. 1996, 1963-4.
(34) Basuli, F.; Bailey, B. C.; Huffman, J. C.; Mindiola, D. J. Organometallics
2005, 24, 3321-34.
24) Bennett, J. L.; Wolczanski, P. T. J. Am. Chem. Soc. 1997, 119, 10696-
7
19.
(35) (a) Goldman, A. S.; Roy, A. H.; Huang, Z.; Ahuja, R.; Schinski, W.;
Brookhart, M. Science 2006, 312, 257-61. (b) Le Roux, E.; Taoufik,
M.; Coperet, C.; de Mallmann, A.; Thivolle-Cazat, J.; Basset,
J.-M.; Maunders, B. M.; Sunley, G. J. Angew. Chem., Int. Ed. 2005,
44, 6755-8. (c) Soulivong, D.; Coperet, C.; Thivolle-Cazat, J.; Basset,
J.-M.; Maunders, B. M.; Pardy, R. B. A.; Sunley, G. J. Angew. Chem.,
Int. Ed. 2004, 43, 5366-9. (d) Coperet, C.; Maury, O.; Thivolle-
Cazat, J.; Basset, J.-M. Angew. Chem., Int. Ed. 2001, 40, 2331-4. (e)
Basset, J. M.; Coperet, C.; Lefort, L.; Maunders, B. M.; Maury, O.;
Le Roux, E.; Saggio, G.; Soignier, S.; Soulivong, D.; Sunley, G. J.;
Taoufik, M.; Thivolle-Cazat, J. J. Am. Chem. Soc. 2005, 127, 8604-5. (f)
Basset, J.-M.; Coperet, C.; Soulivong, D.; Taoufik, M.; Thivolle-Cazat, J.
Angew. Chem., Int. Ed. 2006, 45, 6082-5. (g) Blanc, F.; Coperet, C.;
Thivolle-Cazat, J.; Basset, J.-M. Angew. Chem., Int. Ed. 2006, 45, 6201-
3. (h) Blanc, F.; Coperet, C.; Thivolle-Cazat, J.; Basset, J.-M.; Lesage, A.;
Emsley, L.; Sinha, A.; Schrock, R. R. Angew. Chem., Int. Ed. 2006, 45,
1216-20.
25) Cheon, J.; Rogers, D. M.; Girolami, G. S. J. Am. Chem. Soc. 1997, 119,
6
804-13.
26) (a) Tran, E.; Legzdins, P. J. Am. Chem. Soc. 1997, 119, 5071-2. (b) Adams,
C. S.; Legzdins, P.; McNeil, W. S. Organometallics 2001, 20, 4939-55.
(
c) Adams, C. S.; Legzdins, P.; Tran, E. Organometallics 2002, 21, 1474-
8
6. (d) Wada, K.; Pamplin, C. B.; Legzdins, P. J. Am. Chem. Soc. 2002,
24, 9680-1. (e) Wada, K.; Pamplin, C. B.; Legzdins, P.; Patrick, B. O.;
1
Tsyba, I.; Bau, R. J. Am. Chem. Soc. 2003, 125, 7035-48. (f) Tsang, J. Y.
K.; Buschhaus, M. S. A.; Legzdins, P.; Patrick, B. O. Organometallics
2
006, 25, 4215-25.
(
(
(
27) Slaughter, L. M.; Wolczanski, P. T.; Klinckman, T. R.; Cundari, T. R. J.
Am. Chem. Soc. 2000, 122, 7953-75.
28) Adams, C. S.; Legzdins, P.; Tran, E. J. Am. Chem. Soc. 2001, 123, 612-
2
4.
29) Bailey, B. C.; Fan, H.; Baum, E. W.; Huffman, J. C.; Baik, M.-H.; Mindiola,
D. J. J. Am. Chem. Soc. 2005, 127, 16016-7.
8782 J. AM. CHEM. SOC.
9
VOL. 129, NO. 28, 2007

C−H Bond Activation Involving a TitC Linkage
A R T I C L E S
Scheme 1
Scheme 2
Given the remarkable reactivity observed with alkylidenes,
we postulated that generation of an early transition metal
alkylidyne may afford an even more potent and polarized M-C
multiple bond, which could enable us to perform unique
transformations under non-forcing conditions. We have recently
shown that transient TitC linkages29 can, in fact, be generated
in situ, and can now be added to the short list of metal-ligand
multiply bonded systems that activate the C-H bonds of
hydrocarbons under mild conditions. In this work, we thoroughly
investigate the steps leading to and involved in the C-H
activation of benzene by applying a combination of experimental
kinetic studies and molecular reaction modeling based on density
functional theory. In addition, we report regioselective aryl C-H
activation reactions with a variety of substrates and propose
that the binding event of the substrate can direct the C-H
activation process. The product formation in these reactions
appears to be under kinetic more than thermodynamic control.
Among our attempts at regioselective C-H activation, we also
discovered that the TitC linkage can engage in alkylidene-
alkyl metathesis via multiple C-H activation reactions of
tetramethylsilane and hydrocarbons by employing a combination
of 1,2-CH addition, R-hydrogen migration, and R-hydrogen
abstraction steps.
drogen migration in 2 is rapid on the chemical time scale (min-
utes) in solution phase but slow on the NMR time scale. Other
alkyl groups can be readily incorporated with retention of the
neopentylidene unit. For example, 1 equiv of LiPh reacts cleanly
with 1 in Et2O to afford the alkylidene-phenyl derivative
t
(
PNP)TidCH Bu(C6H5) (3) (Scheme 1). Tautomers also arise
when the alkyl group deviates from neopentyl and contains
R-hydrogens. Consequently, when 1 was treated with 1 equiv
40
of LiCH2SiMe3 or KCH2Ph, a mixture of the alkylidenes
t
29
t
(
(
PNP)TidCH Bu(CH2SiMe3) (4) /(PNP)TidCHSiMe3(CH2 Bu)
29
t
t
5) or (PNP)TidCH Bu(CH2Ph) (6)/(PNP)TidCHPh(CH2 Bu)
7) was observed at 25 °C via a combination of H, C, and
P NMR spectra (Scheme 2). Mixtures of 4/5 or 6/7 are formed
1
13
(
3
1
2. Results and Discussion
by R-hydrogen migration from the alkyl ligand to the neopen-
2
.1. Synthesis and Characterization of Titanium Alkyli-
30
tylidene R-carbon. Stronger evidence for the tautomerization
t
dene-Alkyl Complexes. LiCH2 Bu reacts cleanly with
PNP)TidCH Bu(OTf) (1; PNP ) N[2-P(CHMe2)2-4-methyl-
phenyl]2) to afford the neopentylidene-neopentyl complex
PNP)TidCH Bu(CH2 Bu) (2) in 88% yield (Scheme 1). The
of 4/5 or 6/7 can be collected by an independent synthetic route
t
38
-
(
forming the species in question. This conjecture is shown to be
3
9
29
correct since transmetalation of (PNP)TidCHSiMe3(OTf) (8)
t
t
29
(
t
with LiCH2 Bu generates identical ratios of 4/5, thereby sug-
transmetalation step must be performed quickly in pentane (or
C6H12) to avoid rapid decomposition of 2 (vide infra). Solids
of 2, however, are relatively stable when stored at -35 °C (t1/2
gesting a 4 T 5 equilibrium scenario (Scheme 2). While R-hy-
drogen migration among complexes bearing alkyl, alkylidene,
30,41-45
and alkylidyne substituents has been documented,
R-hy-
≈
2 months). Evidence for R-hydrogen migration between the
drogen migration amid non-degenerate alkyl, alkylidene, and
neopentylidene and neopentyl fragments in 2 was supported by
alkylidyne ligands is an exceedingly rare phenomenon, given
t
reaction of 1 with LiCD2 Bu to rapidly provide 2-d2, which exists
as a 1:2 mixture of isotopomers (PNP)TidCH Bu(CD2 Bu) and
30,34,46
the low number of known archetypical systems.
In the
t
t
case of degenerate systems, similar tautomerization processes
t
t
1
(
PNP)TidCD Bu(CDH Bu), respectively, as judged by H or
P NMR spectra (see Supporting Information). Thus, R-hy-
44
42
45
involving Ta, W, and Os have been documented by
Schrock and Xue, and this phenomenom appears to occur at
much slower chemical time scales than in our systems. The
3
1
(
(
36) (a) Costas, M.; Chen, K.; Que, L. Coord. Chem. ReV. 2000, 200-202,
5
17-44. (b) Hirao, H.; Kumar, D.; Que, L., Jr.; Shaik, S. J. Am. Chem.
Soc. 2006, 128, 8590-606. (c) Kaizer, J.; Klinker, E. J.; Oh, N. Y.; Rohde,
J.-U.; Song, W. J.; Stubna, A.; Kim, J.; Muenck, E.; Nam, W.; Que, L., Jr.
J. Am. Chem. Soc. 2004, 126, 472-3.
(40) Schlosser, M.; Hartmann, J. Angew. Chem. 1973, 85, 544-5.
(41) Liu, X.; Li, L.; Diminnie, J. B.; Yap, G. P. A.; Rheingold, A. L.; Xue, Z.
Organometallics 1998, 17, 4597-606.
(42) Chen, T.; Wu, Z.; Li, L.; Sorasaenee, K. R.; Diminnie, J. B.; Pan, H.; Guzei,
I. A.; Rheingold, A. L.; Xue, Z. J. Am. Chem. Soc. 1998, 120, 13519-20.
(43) (a) Morton, L. A.; Zhang, X.-H.; Wang, R.; Lin, Z.; Wu, Y.-D.; Xue, Z.-
L. J. Am. Chem. Soc. 2004, 126, 10208-9. (b) Morton, L. A.; Wang, R.;
Yu, X.; Campana, C. F.; Guzei, I. A.; Yap, G. P. A.; Xue, Z.-L.
Organometallics 2006, 25, 427-34.
(44) Schrock, R. R.; Fellmann, J. D. J. Am. Chem. Soc. 1978, 100, 3359-70.
(45) LaPointe, A. M.; Schrock, R. R.; Davis, W. M. J. Am. Chem. Soc. 1995,
117, 4802-13.
(46) (a) Caulton, K. G.; Chisholm, M. H.; Streib, W. E.; Xue, Z. J. Am. Chem.
Soc. 1991, 113, 6082-90. (b) Xue, Z.; Chuang, S.-H.; Caulton, K. G.;
Chisholm, M. H. Chem. Mater. 1998, 10, 2365-70. (c) Xue, Z.; Caulton,
K. G.; Chisholm, M. H. Chem. Mater. 1991, 3, 384-6.
37) (a) Que, L., Jr.; Dong, Y. Acc. Chem. Res. 1996, 29, 190-6. (b)
Rosenzweig, A. C.; Lippard, S. J. Acc. Chem. Res. 1994, 27, 229-36. (c)
Baik, M.-H.; Newcomb, M.; Friesner, R. A.; Lippard, S. J. Chem. ReV.
2
003, 103, 2385-419. (d) Du Bois, J.; Mizoguchi, T. J.; Lippard, S. J.
Coord. Chem. ReV. 2000, 200-202, 443-85.
(38) Bailey, B. C.; Huffman, J. C.; Mindiola, D. J.; Weng, W.; Ozerov, O. V.
Organometallics 2005, 24, 1390-3.
(39) (a) Fan, L.; Foxman, B. M.; Ozerov, O. V. Organometallics 2004, 23, 326-
8
. (b) Fan, L.; Yang, L.; Guo, C.; Foxman, B. M.; Ozerov, O. V.
Organometallics 2004, 23, 4778-87. (c) Ozerov, O. V.; Guo, C.; Fan, L.;
Foxman, B. M. Organometallics 2004, 23, 5573-80. (d) Ozerov, O. V.;
Guo, C.; Papkov, V. A.; Foxman, B. M. J. Am. Chem. Soc. 2004, 126,
4
792-3.
J. AM. CHEM. SOC.
9
VOL. 129, NO. 28, 2007 8783

A R T I C L E S
Scheme 3
Bailey et al.
difference in the rate of tautomerization might be a consequence
of the meridional constraint imposed by the PNP framework,
resulting in the alkylidene and alkyl groups being more proximal
to one another, thus rendering the H atom more susceptible to
the rate of reactions involving C6H6 and the alkylidene-alkyl
precursor or mixture. Whereas complex 2 transforms to 3 with
t1/2 ) 3.1 h, tautomeric mixtures 4/5 and 6/7 react much slower
with benzene. Complex 9, on the other hand, appears to
transform in C6H6 to 10 at a rate that is comparably slower
relative to that for the 4/5 mixture. Likewise, 11 also transforms
very slowly, which may give rise to undesirable competing side
reactions. In accord with these observations, we infer that the
stability of the trimethylmethylidene motif arises from dissipa-
tion of electron density away from the alkylidene R-C via
strengthening of the C-Si bond by σ-hyperconjugation, which
decreases the nucleophilicity of the R-C. In the case of the Tid
CHPh functionality, the negative charge at the R-C is likely
minimized by delocalization onto the phenyl group (vide infra).
Steric effects may also play a role in the stability of each system,
since complex 2 has the most crowded environment. As it will
be described below, elimination of alkane is one of the key steps
along the transformation of complexes such as 2.
1,3-migration.
Analogous to the degenerate R-hydrogen migration reaction
in 2, treatment of 8 with LiCH2SiMe3 yields the complex
(
PNP)TidCHSiMe3(CH2SiMe3) (9) (Scheme 2). The triflato
-
group in 8 can also be replaced with Ph to afford a close
relative of 3, namely (PNP)TidCHSiMe3(C6H5) (10)
29
(
Scheme 2). Treatment of 8 with KCH2Ph results in essentially
complete tautomerization to the benzylidene derivative,
PNP)TidCHPh(CH2SiMe3) (11), via the intermediate
PNP)TidCHSiMe3(CH2Ph) (inferred from NMR spectra, Scheme
). All alkylidene-alkyl systems reported here are far less
(
(
3
vulnerable to intramolecular ancillary ligand degradation when
compared to other titanium derivatives previously prepared in
our laboratory using the sterically encumbering â-diketiminate
ligand.34
Alkylidene-alkyl compounds 2-7 and 9-11 display 13C
NMR resonances between 253 and 310 ppm, indicative of each
system bearing a terminal alkylidene functionality. These
complexes also reveal relatively low JC-H coupling constants
of 86-99 Hz for the alkylidene carbon, in addition to a low-
2.2.2. Structural Characterization of Titanium Alky-
lidene-Alkyl Complexes. In order to establish the degree of
R-H agostic interaction in the alkylidene-alkyl complexes, we
collected single-crystal X-ray diffraction data for compounds
2, 3, 11, and 12 (Figure 1). Crystal structure data for 2 and 3
30
field R-H resonance (6.76-13.00 ppm). The latter two
parameters suggest strongly that an agostic R-hydrogen interac-
3
0
29
tion with the electron-deficient metal center is present. The
have been reported in an earlier communication. Although the
C2-symmetric nature of the pincer framework renders complexes
molecular structure of 2 lies with the Ti and N atoms on the
2-fold axis at / , y, / , where the TidCH Bu and TisCH Bu
4 4 2
1
3
t
t
2
-7 and 9-11 chiral, which is clearly manifested by each
31
system having two sets of doublets in its P NMR spectrum.
Furthermore, the inequivalent phosphines display low JP-P
coupling constants of 32-53 Hz, an indication that the PNP
framework is significantly skewed from a rigid trans geometry.
groups are disordered about this axis, these motifs are well
resolved, with none of the atomic positions overlapping. The
partial occupancy hydrogen atoms associated with the disordered
ligands were clearly visible in a difference Fourier map, and a
slightly better model was obtained when the methyl hydrogen
atoms were placed in idealized riding positions. Similar
structural disorders have been recently observed with tungsten
2
.2. C-H Activation of Benzene. 2.2.1. r-Hydrogen
Abstraction Reactions and C-H Activation. Compounds 2,
-7, 9, and 11 are all kinetic products under N2 inasmuch as
4
4
1
they activate the C-H bond of benzene (eqs 1-3). While the
bis(alkylidene)/alkylidyne-alkyl tautomers. In all cases, the
molecular structures reveal short TidC bonds (1.790(5) Å, 2;
neopentylidene-neopentyl complex 2 reacts with C6H6 at 27
t
12,29,30,32,34,38,47,48
°
(
C over ∼12 h to afford the phenyl derivative 3 and CH3 Bu
1.873(5) Å, 3; 1.876(2) Å, 11; 1.889(2) Å, 12)
eq 1), the 4/5 mixture also transforms smoothly (15 h at 30
and indicate a metal center residing in a pseudo-trigonal-
bipyramidal geometry, where the two phosphine groups occupy
°
C) in C6H6 to form only one alkylidene product, namely
compound 10 (eq 2). Likewise, the 6/7 mixture of alkyl-
idenes gradually decay in C6H6 at 30 °C over 48 h to also
produce one compound, the benzylidene-phenyl derivative
(
47) (a) Krueger, C.; Mynott, R.; Siedenbiedel, C.; Stehling, L.; Wilke, G. Angew.
Chem. 1991, 103, 1714-15. (b) Kahlert, S.; Gorls, H.; Scholz, J. Angew.
Chem., Int. Ed. 1998, 37, 1857-61. (c) Van de Heisteeg, B. J. J.; Schat,
G.; Akkerman, O. S.; Bickelhaupt, F. J. Organomet. Chem. 1986, 310,
C25-8. (d) Baumann, R.; Stumpf, R.; Davis, W. M.; Liang, L.-C.; Schrock,
R. R. J. Am. Chem. Soc. 1999, 121, 7822-36. (e) Hartner, F. W., Jr.;
Schwartz, J.; Clift, S. M. J. Am. Chem. Soc. 1983, 105, 640-1. (f) Scholz,
J.; Goerls, H. Organometallics 2004, 23, 320-2. (g) Sinnema, P.-J.; van
der Veen, L.; Spek, A. L.; Veldman, N.; Teuben, J. H. Organometallics
(PNP)TidCHPh(Ph) (12) (eq 3). Complex 12 can also be
generated, albeit slowly, from C-H activation of benzene via
the thermolysis of 11 (∼80% conversion at 60 °C over 3 days),
but unfortunately other products are formed in the reaction
mixture. Complex 11 clearly shows that the benzylidene
complex is the preferred tautomer compared to the correspond-
ing (trimethylsilyl)methylidene. Hence, the reactivity of the
1
997, 16, 4245-7. (h) Bailey, B. C.; Basuli, F.; Huffman, J. C.; Mindiola,
D. J. Organometallics 2006, 25, 3963-8. (i) Mindiola, D. J. Acc. Chem.
Res. 2006, 39, 813-21. (j) Mindiola, D. J.; Bailey, B. C.; Basuli, F. Eur.
J. Inorg. Chem. 2006, 3335-46.
(48) (a) Basuli, F.; Bailey, B. C.; Tomaszewski, J.; Huffman, J. C.; Mindiola,
D. J. J. Am. Chem. Soc. 2003, 125, 6052-3. (b) Basuli, F.; Bailey, B. C.;
Watson, L. A.; Tomaszewski, J.; Huffman, J. C.; Mindiola, D. J.
Organometallics 2005, 24, 1886-906.
t
alkylidene unit decreases in the sequence TidCH Bu . Tid
CHSiMe3 g TidCHPh. This trend is further corroborated by
8784 J. AM. CHEM. SOC.
9
VOL. 129, NO. 28, 2007

C−H Bond Activation Involving a TitC Linkage
A R T I C L E S
Figure 1. Molecular structures of complexes 2, 3, 11, and 12, depicting thermal ellipsoids at the 50% probability level. Solvent and H-atoms have been
excluded for clarity. Selected metrical parameters (distances in angstroms, angles in degrees): For 2, Ti1-N2, 2.124(3); Ti1-P10, 2.6003(6); Ti1-C17,
1
.790(5); Ti1-C22, 2.206(5); Ti1-C17-C18, 166.3(4); Ti1-C22-C23, 133.4(4); P10A-Ti1-P10, 148.59(3). For 3, Ti1-N10, 2.089(4); Ti1-P2, 2.580-
(
2
8); Ti1-P18, 2.612(8); Ti1-C31, 1.873(5); Ti1-C36, 2.141(5); Ti1-C31-C32, 174.7(5); P2-Ti1-P18, 149.70(6). For 11: Ti1-N10, 2.061(6); Ti1-P2,
.561(1); Ti1-P18, 2.607(1); Ti1-C31, 1.876(2); Ti-C38, 2.166(2); Ti1-C31-C32, 164.0(8); Ti1-C38-Si39, 126.2(3); P2-Ti1-P18, 147.75(3). For
12: Ti1-N10, 2.029(5); Ti1-P2, 2.5360(6); Ti1-P18, 2.5840(6); Ti1-C31, 1.889(2); Ti1-C38, 2.146(8); Ti1-C31-C32, 159.2(7); P2-Ti1-P18, 150.70(2).
Figure 2. Computed reaction energy profile for 2 f 3 conversion in benzene (ꢀ ) 2.284). Relative energies in kcal/mol are given in parentheses. The
transition states marked with asterisks cannot be located with standard methods and are estimated; see text for details.
the axial positions. As expected, the obtuse TidCsC angles
Figure 1) coupled with the low R-C JC-H (vide supra) are
consistent with R-H agostic interactions being present in both
solution and solid phases. The molecular structure of 12 is
unique, as it constitutes a rare example of a terminal titanium
benzylidene complex.30 Strong evidence for tautomerization in
some of our systems originates from the molecular structure of
considering a number of reasonable mechanistic scenarios.
Figure 2 summarizes the most plausible reaction pathway, and
the computed structures of the key intermediates and transition
states are shown in Figure 3. Complex 2 first undergoes
R-hydrogen abstraction, traversing the transition state 2-TS
at 27.8 kcal/mol to afford the hypothetical σ-complex
(
t
t
(PNP)TitC Bu(CH Bu) (13), which is 15.8 kcal/mol higher
3
1
1 (Figure 1), exposing a rare example of a non-degenerate
in solution-phase free energy than 2. Rapid entropy-assisted loss
alkylidene-alkyl system bearing R-hydrogens at both the
While Ti-CH2SiMe3
formation is evident from the acute Ti-C-Si angle
of neopentane yields the reactive titanium alkylidyne intermedi-
ate A, which is only slightly higher in energy than 2, with a
relative free energy of 4.6 kcal/mol. Locating the transition state
for loss of neopentane (13-TS) is, unfortunately, impossible with
standard quantum mechanical methods, because entropy
plays an important and non-trivial role in determining this
transition-state energy. Our simplistic approach only probes the
electronic surface, however, and projects onto the free energy
space by adding the entropy after the minima and first-order
saddle points are located. We assume, therefore, that the
electronic and free energy surfaces are sufficiently similar.
alkylidene and alkyl functionalities.3
4,46
(
(
126.2(3)°) and Ti-C distance (2.166(2) Å), the TidC distance
vide supra) and distorted TidCsC angle (164.0(8)°) for the
benzylidene unit in 11 are also consistent with R-hydrogen
migration from the formal benzyl to the former trimethylsilyl-
methylidene motif (Figure 1).
2
.2.3. Theoretical Studies. To better understand the C-H
activation mechanism, we turned to density functional theory
DFT) calculations and simulated the reaction of 2 f 3
(
J. AM. CHEM. SOC.
9
VOL. 129, NO. 28, 2007 8785

A R T I C L E S
Bailey et al.
than aromatic C-H bonds using the titanium alkylidyne
24
complex. The activation energy for cleaving the aromatic C-H
bond is, in fact, so low that the formation of the aromatic
σ-complex becomes rate-determining. This result is not intuitive,
since aromatic C-H bonds are typically stronger than aliphatic
49
C-H bonds by ∼10-15 kcal/mol. However, this observation
is not unprecedented, since the kinetic preference for the stronger
2
3
sp C-H bond activation versus the sp C-H bond vis- a` -vis
q
2
q
1
16,24
∆
G addition(R H) - ∆G addition(R H) values
has been exten-
sively investigated and discussed by Wolczanski, and such
selectivity has been ascribed to the relative free energies of three-
coordinate titanium imide precursors by applying a more
2
4
compressed reaction coordinate. Legzdins has also observed
2
28
similar selectivities for sp C-H bond activation reactions.
2
π-Orbital involvement of the sp C-H substrate has been
repeatedly suggested as an explanation to why there is more
stabilization at the TS. By attaining shorter bond lengths with
2
the sp -hybridized substituent, H-transfer occurs over a much
shorter distance and with minimal reorganization energy, hence
resulting in an earlier TS for 14 versus late for 2 (vide infra).
Consequently, the 1,2-CH bond addition step for the latter type
3
16,24
will be prompted over sp C-H substrate activation.
As
mentioned above, our computed reaction energy profile places
the titanium alkylidyne intermediate A at a relative energy of
4.6 kcal/mol compared to 2, suggesting that A may be accessible
at room temperature under equilibrium conditions and in the
absence of reactive substrates. Experimental efforts toward
exploiting this theoretical insight and detecting A directly are
currently underway.
Figure 3. Computed structures of 2-TS, A, and 14-TS. Bond lengths are
given in angstroms.
There are, of course, a few plausible mechanistic alternatives
that deserve attention. For example, given the precedence for
Whereas this assumption is plausible, it is not expected to be
valid for ligand dissociation transition states that are dominated
by entropy changes. Intuitively, we do not expect this transition
state to be competitive with 2-TS, as it is unlikely that the simple
dissociation of neopentane to afford A is kinetically less facile
than the C-H activation step to re-form 2, which our calcula-
tions suggest to be associated with a barrier of 12.0 kcal/mol.
Upon addition of benzene, the titanium alkylidyne intermedi-
unsaturated MdC linkages to undergo intramolecular 1,2-CH
12,21,23,25,26,28
addition of arenes,
we inquired whether or not
complex 2 could engage in C-H activation events prior to the
R-hydrogen abstraction event (eq 4). Similar processes have
2,13,15-17,22,24,27
been documented for imide analogues.
Using a
t
-
slightly simplified model (PNP′)TidCH Bu(CH3) (2′; PNP′ )
N[2-P(CH3)2phenyl]2) instead of 2, we explored this alternative
mechanistic scenario and found this reaction path to be both
thermodynamically and kinetically unreasonable. Our calcula-
tions show that the putative six-coordinate intermediate (PNP′)-
t
ate A first forms a σ-complex (PNP)TitC Bu(C6H6) (14), which
undergoes 1,2-addition of the C-H bond of the arene across
the reactive TitC linkage to produce the final product 3.
Although we are unable to locate the transition state A-TS,
which connects complex A with the σ-complex 14, as we
explained for 13-TS (vide supra), we have some experimental
evidence for A-TS being slightly higher in energy than 14-TS,
which must be traversed to form the final phenyl-alkylidene
product 3. This asymmetry of the reaction energy profile is
interesting and highlights the significant difference between the
activation of aliphatic and aromatic C-H bonds, with the
t
Ti(CH2 Bu)(CH3)(C6H5) is 30.3 kcal/mol higher in solution-
phase free energy than the titanium alkylidene complex 2′,
indicating that the alkylidene moiety is thermodynamically
unable to perform the C-H activation (even in a less congested
environment provided by our simplified models). In addition,
the lowest energy transition state was found at 53.3 kcal/mol,
further supporting our conclusion that this reaction path is not
accessible. There are a few other possible reaction channels to
consider; e.g., the long Ti-P bond (>2.5 Å) could break to
release some of the steric crowding and electronic strain that
we found for the putative mechanism outlined in eq 4. Despite
much effort, we were unable to locate another reaction mech-
anism that was competitive to our proposed pathway involving
the in situ formation of the titanium alkylidyne intermediate A.
The most notable mechanistic alternatives that we considered
are highlighted in the Supporting Information.
q
computed ∆∆G being 6.8 kcal/mol, which translates to several
orders of magnitude difference in the rate.16,24 When the titanium
alkylidyne complex A reacts with an alkane, such as neopentane,
the C-H activation process is rate-determining. For the activa-
tion of aromatic C-H bonds, however, the formation of the
σ-complex becomes rate-determining. This difference is il-
lustrated in Figure 2 by the computed energy differences
between the σ-complexes and their respective C-H activation
transition states. The energy difference between 13 and 2-TS
is 12.0 kcal/mol, whereas the energy difference between 14 and
A more detailed examination of the computed transition-state
structures of 2-TS and 14-TS (Figure 3) revealed structural
1
4-TS is only 5.1 kcal/mol. Thus, our calculations suggest that
aliphatic C-H bonds are significantly more difficult to activate
(49) Jones, W. D.; Feher, F. J. J. Am. Chem. Soc. 1982, 104, 4240-2.
8786 J. AM. CHEM. SOC.
9
VOL. 129, NO. 28, 2007

C−H Bond Activation Involving a TitC Linkage
A R T I C L E S
elements that are intuitively expected for a σ-bond metathesis
invoking transient metallacycles. In 2-TS, the C1-H1-C2
arrangement is not linear per se (151.7°), and the C1-H1 bond
is nearly broken at a bond length of 1.625 Å, whereas the C2-
H1 for the leaving neopentane is nearly formed at a distance of
1.460 Å. Thus, the transition state 2-TS can be considered “late”
with respect to the titanium alkylidyne formation. In the case
of 14-TS, the C1-H1-C2 arrangement is similar to that in 2-TS
(152.7°), but the C1-H1 is even longer, at 1.704 Å, when
compared to the C2-H1 linkage (1.305 Å). This feature strongly
implies that the computed transition-state structure of 14-TS
can be regarded as “early” with respect to formation of 3. On
the basis of these observations, the C-H-C array in 2-TS and
1
4-TS can also be referred to as a three-center, four-electron
(
3c-4e) system undergoing a proton transfer promoted by the
50
Ti(IV) center. For both 2-TS and 14-TS, there is small
perturbation of the TitC linkage (1.763 and 1.776 Å, respec-
tively) and metrical parameters concerning the (PNP)Ti sur-
rogate (see Supporting Information). Titanium alkylidyne
character is fully developed at the transition state, with a Ti-
C1 distance of 1.763 Å. The Ti-C distance becomes 1.749 Å
in intermediate A, which we assign to a TitC triple bond. As
expected, the TitC bond is significantly shorter than TidC and
TisC moieties that were found for systems such as 2, 3, and
Figure 4. Most important molecular orbitals of A (isodensity ) 0.05 au).
strongly support the presence of a Ti(IV)tC fragment in A with
a Mayer bond order of 2.49, which is the projected value for a
54
TisC triple bond. For comparison, bond orders of 1.74 and
0
.85 were computed in 2 for the TidC and TisC bonds,
11. For example, TidC and TisC distances in 2 were computed
respectively.
to be 1.847 and 2.137 Å, respectively. Although it is reasonable
to interpret the unusually short Ti-C distance obtained in our
computer model of A as a strong indicator of a triple bond, a
few different interpretations are equally plausible. Most interest-
ing among these alternatives for the bonding in the TitC
fragment is formally invoking intramolecular ligand-to-metal
Figure 4 shows the most important molecular orbitals of
intermediate A. The HOMO-5, HOMO-1, and HOMO are
easily recognized as the σ-orbital and the two orthogonal
π-oribtals establishing the TitC bond, respectively (Figure
4
a-c). The LUMO (Figure 4d) is a M-L nonbonding orbital,
2
2
mainly consisting of the metal d(x -y ) orbital and significantly
extending into the empty coordination site to facilitate binding
of substrate. As this MO can be considered to be purely metal-
based and fully decoupled from the ligand orbitals, we expect
that utilization of the MO for binding additional ligands will
have very little effect on both the geometry and the electronic
structure of the titanium complex. Clearly, this is one of the
key features that the structurally rigid (PNP)Ti framework
provides. Being able to bind substrate to afford the σ-complex
in this case, and doing so without causing significant electronic
disruptions, is an important characteristic for the unparalleled
reactivity of such a system.
•
charge transfer to afford a Ti(III)dC moiety, which may adopt
an open-shell singlet state, where the two unpaired electrons
on Ti and C reside in spatially different orbitals but with opposite
spins. These non-classical electronic structure patterns have
recently received much attention,51 and there is growing
awareness that intramolecular M-L redox events may play a
52
pivotal role in catalysis. The question of whether a Ti(IV)tC
•
or a Ti(III)dC fragment is responsible for the remarkable
reactivity of an intermediate such as A is crucial for designing
more optimized reagents, because it has immediate consequences
for the underlying driving force in the C-H activation mech-
anism. Despite significant efforts, including the use of manually
constructed Kohn-Sham functions that were consistent with
the open-shell singlet state as initial guesses,53 we were unable
to obtain a properly converged electronic structure that is
Once the σ-complex (13 and 14 in Figure 2) is formed
between the reactant and the alkylidyne intermediate A, there
are two possible formal mechanisms by which the C-H bond
activation can take place. First, the C-H bond could be broken
in a homolytic fashion at the transition state, which must be
matched by a similarly homolytic cleavage of the triple bond
of the titanium alkylidyne moiety to formally afford a
•
expected for a Ti(III)dC fragment. Instead, our calculations
(
(
(
50) Werkema, E. L.; Maron, L.; Eisenstein, O.; Andersen, R. A. J. Am. Chem.
Soc. 2007, 129, 2529-41.
51) Ray, K.; Bill, E.; Weyhermuller, T.; Wieghardt, K. J. Am. Chem. Soc. 2005,
•
Ti(III)dC fragment. In this case, the σ-bond metathesis reaction
1
27, 5641-54.
52) (a) Bart, S. C.; Chlopek, K.; Bill, E.; Bouwkamp, M. W.; Lobkovsky, E.;
is driven by the radical recombination energy. Alternatively,
the C-H bond can also be envisioned to break heterolytically,
Neese, F.; Wieghardt, K.; Chirik, P. J. J. Am. Chem. Soc. 2006, 128, 13901-
1
2. (b) Scott, J.; Gambarotta, S.; Korobkov, I.; Budzelaar, P. H. M. J. Am.
Chem. Soc. 2005, 127, 13019-29.
(53) Yang, X.; Baik, M.-H. J. Am. Chem. Soc. 2006, 128, 7476-85.
(54) Baik, M.-H.; Friesner, R. A.; Parkin, G. Polyhedron 2004, 23, 2879-900.
J. AM. CHEM. SOC.
9
VOL. 129, NO. 28, 2007 8787

A R T I C L E S
Bailey et al.
afford a negatively polarized C, whereas the H becomes
positively polarized. The titanium alkylidyne moiety is already
highly polarized and undergoes only a slight charge adjustment
to match the induced dipole of the C-H of the substrate. The
C-H carbon acts as a Lewis base and begins to form the C-Ti
bond by interacting with the metal-based LUMO via the adduct
13 or 14. At the same time, the Lewis-acidic C-H proton begins
to interact with the alkylidyne carbon to formally complete the
cleavage of one of the Ti-C triple bonds. The σ-bond metathesis
is therefore driven by a closed-shell mechanism whereby the
HOMO and LUMO of the metal-complex fragment interact in
a cooperative fashion to activate and then break the C-H bond
of the alkane or arene. Our calculations do not support a radical-
recombination-type mechanism, which would be expected to
show much less polarized transition states.
This analysis implies that electron-donating substituents on
the alkylidyne moiety are important for its reactivity. Electron-
withdrawing groups will diminish the charge polarization of the
titanium alkylidyne carbon and therefore reduce the overall
reactivity toward σ-bond metathesis. This conclusion is in good
agreement with the experimental observation that the titanium
alkylidyne precursors 4/5, 6/7, 9, and 11 are less reactive toward
C-H activation of benzene than the neopentylidyne precursor
2
. Based on these arguments, our theoretical analysis predicts
that negative charge polarization at the alkylidyne carbon should
follow the decaying order TitC Bu . TitCSiMe3 g TitCPh,
which would also explain their trend in reactivity.
Figure 5. (a) ESP-fit charges of the titanium alkylidyne, the σ-complex,
and the transition states. (b) Conceptualized MO interaction patterns leading
to the heterolytic cleavage of the C-H bond.
t
2
.2.4. Kinetic Studies of the C-H Activation of Benzene.
matched by the corresponding heterolytic cleavage of the Ti-C
triple bond. In this case, the C-H activation reaction is promoted
by standard Lewis acid/base binding; viz., the titanium center
serves as a strong Lewis acid, and the C-H bond of benzene
attacks in a nucleophilic fashion. The electronic structure
described in Figure 4 favors the latter scenario in principle, and
our calculated transition states show highly charge-polarized
transition states. Figure 5 summarizes the charge distribution
at the titanium alkylidyne motif during the reaction with
neopentane and benzene, monitored using charges that are fit
to the electrostatic potentials (ESPs).55 Our calculations assign
charges of +1.37 to the titanium and -1.38 to the alkylidyne
carbon in A. These ESP-fit charges must not be interpreted as
an indication that a “carbanion” has been formed and/or the
TitC linkage is ionic, as the absolute values of quantum
mechanically computed charges are not physically meaningful
per se. The ESP charges merely indicate that the alkylidyne
carbon is relatively electron-rich, resulting from the three main
M-L bond MOs (Figure 4) being essentially dominated by
Complex 2 was reacted with C6D6 to confirm that (PNP)Tid
t
1
CD Bu(C6D5) (3)-d6 was quantitatively formed. Both H and
2
2
H NMR spectra of 3-d6 reveal >99+% H incorporation into
the R-carbon without H scrambling into the PNP ligand
framework. In C6D6, conversion of 2 f 3-d6 generates only
2
t
CH3 Bu, supporting our hypothesis that R-abstraction in 2 should
precede the C-H activation of benzene and that the overall
slow step of the reaction is formation of a high-energy
intermediate species. The fact that the PNP ancillary ligand was
not involved in the C-H activation pathway narrowed our list
of mechanistic possibilities.
2
.2.5. Kinetic Studies of the Rate-Determining Step.
Kinetic parameters for the conversion of 2 f 3 were assessed
experimentally by monitoring the decay of 2 using P NMR
31
-5 -1
spectroscopy in C6D6 to reveal a kaverage ) 6.5(4) × 10
s
at
2
7 °C. The reaction was found to obey pseudo-first-order
kinetics in titanium, which is in good agreement with the
reaction energy profile suggested by our DFT calculations
Figure 2). Unfortunately, under a depleted amount of benzene
(
(
t
carbon-based atomic orbitals. In addition, the Bu group supplies
e.g., C6D12 as a medium), formation of 2 is not clean, given
extra electron density to lower-lying MOs, making the carbon
even more electron-rich. More meaningful are the differential
charges during the reaction. They serve as reporters of the
electronic reorganizations that promote the reaction. Most
importantly, our calculations indicate that the transition states
the promiscuity of intermediate A to react rapidly with both
benzene and other media, such as cyclohexane, hexane, and
pentane (vide infra). Temperature dependence studies of the 2
f 3 conversion between 15 and 40 °C allowed for extraction
q
q
of the activation parameters ∆H ) 24(7) kcal/mol and ∆S )
2-TS and 14-TS are both highly polarized (Figure 5). The carbon
q
-2(3) cal/mol‚K from the Eyring plot, thereby yielding a ∆G
of the C-H bond that undergoes activation becomes negatively
polarized at the transition states, with computed differential
charges of -0.26 and -0.48 for 2-TS and 14-TS, respectively,
compared to the σ-complexes 13 and 14 (Figure 5).
Figure 5b illustrates the electronic changes that are most
consistent with our MO and charge analyses. At the transition
state, the C-H bond breaks in a formally heterolytic fashion to
value of ∼24.7 kcal/mol at 298.15 K (Figure 6), which is in
reasonable agreement with the DFT prediction of 27.8 kcal/
mol. The intermolecular kH/kD ratio is remarkably close to unity
(1.03(3) at 25 °C) when the C-H activation reaction is carried
out in C6H6 vs C6D6, and remains essentially the same in a 1:1
(55) Breneman, C. M.; Wiberg, K. B. J. Comput. Chem. 1990, 11, 361-73.
8788 J. AM. CHEM. SOC.
9
VOL. 129, NO. 28, 2007

C−H Bond Activation Involving a TitC Linkage
A R T I C L E S
Figure 7. Eyring plot for the 3 f 3-d6 transformation in C6D6.
Figure 6. Eyring plot for the 2 f 3-d6 transformation in C6D6.
of the 3 f 3-d6 conversion between 95 and 120 °C allowed for
extraction of the activation parameters ∆H ) 31(16) kcal/
mixture of C6H6 and C6D6 (1.03(7) at 25 °C). In fact, only the
latter competitive intermolecular KIE experiment in C6D6 and
C6H6 will provide information about the substrate binding event,
but, given the value close to unity, we conclude that the overall
9
q
q
q
mol and ∆S ) 3(9) cal/mol‚K, which yielded a ∆G value of
30.1 kcal/mol at 298.15 K (Figure 7). The latter parameter is
∼
in excellent agreement with the DFT prediction of at least 32.8
kcal/mol for the reverse reaction, 3 f 14 (Figure 2). For the 2
f 3 and 3 f 3-d6 conversions, the errors affiliated with the
activation parameters were determined by error propagation
methods that include uncertainties for both rates and tempera-
ture, and which were derived from the Erying equation (see
rate of the reaction has minor contributions from the C-H bond
cleavage of benzene.56 However, the rates are substantially
t
different when the decays of 2 and 2-d3, (PNP)TidCD Bu-
t
(
CD2 Bu), in benzene are measured, yielding a kH/kD ratio of
3
.9(5) at 40 °C (at 27 °C, the kH/kD ratio was 3.7). Although
this primary KIE deviates from unity, the value is not as large
5
8
1
6,24
Supporting Information).
as the KIEs reported by Wolczanski (KIE ) ∼5-14).
This
As suggested by Wolczanski and co-workers, 1,2-CH bond
difference in KIE might be rationalized on the basis of a more
asymmetric transfer of H from the Calkylidene to the Calkyl group
+
addition across the MdN functionality could involve a short-
1
7
lived σ-bonded alkane intermediate. To examine the C-H
activation pathway stemming from 2 in greater detail and to
determine whether or not an intermediate is formed, we resorted
to the isotopically enriched benzene-1,3,5-d3 reagent, previously
used by Jones and co-workers to separate isotope effects that
originate from intramolecular C-H activation processes from
the intermolecular effects that are present when a 1:1 mixture
proposed in our four-centered TS of 2. As described earlier,
our computed transition-state structure for 2-TS reveals a
C-H-C angle of 151.7° (vide supra), which could account for
the lower KIE. In summary, these data provide strong evidence
for the rate-determining step being the R-abstraction to generate
a transient titanium alkylidyne linkage. The 1,2-CH bond
t
addition of benzene across the TitC Bu functionality is most
56,59
of C6H6/C6D6 is used.
Having only two chemically identical,
likely not rate-determining. Since the highest barrier in the C-H
bond activation process is R-hydrogen abstraction, one would
expect only a small contribution from secondary isotope effects
isotopically marked reaction sites in this substrate provides a
means of discerning whether or not secondary isotope effects
play a significant role in the post-rate-determining step. In
addition, the intramolecular kinetic isotope effect should give
details about other intermediates that may be present along the
activation and breaking of the C-H bond.
and/or equilibrium isotope effects.5
6,57
As it will turn out, the
latter process does not contribute significantly to the KIE values
obtained at 25 or 40 °C (vide infra).
2
.2.6. Kinetic Studies of the Post-Rate-Determining Step.
1
When 2 was treated with benzene-1,3,5-d3, H NMR spectral
Our DFT calculations suggest that both the C-H and C-D
activation steps should be reversible, since the backward reaction
for the C-H activation, 3 f A or 3 f 14, is associated with a
barrier g33 kcal/mol. Therefore, conversion of 3 f A should
be overcome at a slow but reasonable rate under elevated
temperature conditions (Figure 2). Gratifyingly, we determined
that 3-d6 can slowly regenerate the alkylidyne A in C6H6 at
analysis revealed a kH/kD ratio of 1.33(3), quantified by
integrating the aromatic regions of the NMR spectra. Since 3
does not exchange with C6D6 to 3-d6 under these conditions
(25 °C), these results indicate that this kH/kD ratio reflects mostly
contributions from both the substrate binding event (post-rate-
determining) and the C-H bond activation step. This value does
not originate from an equilibrium isotope effect (EIE), given
the high barrier for the 3 f A conversion under these
1
00 °C within 48 h to afford 3, while the latter (after
purification) converts in neat C6D6 back to 3-d6, thus linking
2
7
9
conditions. Independent measurements of the EIE (2 f 3
conversion in benzene-1,3,5-d3 at 95 °C for one week) do, in
fact, reveal this value to be minor but comparable to the
intramolecular KIE (1.36(7), see Supporting Information). Our
KIE and EIE values therefore suggest that contributions from
the alkylidyne intermediate A to the final product. Kinetic
parameters for the conversion of 3 to 3-d6 were assessed by
1
monitoring the decay of 3 using H NMR spectroscopy in C6D6,
-
5
-1
to reveal a kaverage ) 1.2(2) × 10
s
at 95 °C. Albeit using
a narrow temperature range, our temperature-dependent studies
(
56) Jones, W. D. Acc. Chem. Res. 2003, 36, 140-6.
57) Carpenter, B. K. Determination of Organic Reaction Mechanisms; Wiley-
Interscience: New York 1984.
(58) Morse, P. M.; Spencer, M. D.; Wilson, S. R.; Girolami, G. S. Organome-
tallics 1994, 13, 1646-55.
(59) Jones, W. D.; Feher, F. J. J. Am. Chem. Soc. 1986, 108, 4814-19.
(
J. AM. CHEM. SOC.
9
VOL. 129, NO. 28, 2007 8789

A R T I C L E S
Bailey et al.
t
Figure 8. Computed structure for the proposed intermediates 14 (left) and (PNP)TitC Bu(FC6H5) (right), depicting selected bond lengths (angstroms) and
angles (degrees): For 14, Ti1-C7, 2.739; Ti1-H10, 2.412; Ti1-N2, 2.097; Ti1-C5, 1.751; H10-C7, 1.092; 1.751; Ti1-C5-C6, 169.2; P3-Ti1-P4,
t
1
1
47.9. For (PNP)TitC Bu(FC6H5), Ti1-F7, 2.269; F7-C8, 1.398; Ti1-N2, 2.094; Ti1-C5, 1.753; Ti1-C5-C6, 165.9; Ti1-F7-C8, 136.2; P3-Ti1-P4,
47.7.
the C-H bond activation step are likely operative. Although
the magnitude is small between the inter- and intramolecular
KIEs (∆ ) 0.33), and could be the net result of more
complicated secondary isotopes effects, we propose that this
discrepancy in the intra- vs intermolecular KIE originates from
the substrate binding event being rate-determining as well as
from contributions from C-H and C-D activations; i.e., A-TS
must be higher in energy than 14-TS, as illustrated in Figure 2.
activation transition state (Figure 8). A π-complex, if formed,
may rapidly rearrange or equilibrate to the σ-complex. As
expected, the computed structure for the σ-complex of 14 reveals
t
an essentially intact (PNP)TitC Bu fragment with a weak
interaction with the C-H bond of benzene. The metrical
parameters reflect subtle changes from that of free benzene, as
enumerated in Figure 8.
2.3. Intermolecular C-H Activation Reactions of Sub-
stituted Arenes. In addition to benzene, complex 2 reacts
q
This result also implies that the ∆∆G between A-TS and
2
1
4-TS should be small, albeit detectable via the intramolecular
cleanly with other sp C-H bonds in substituted arenes. When
KIE experiment. As a result, our data suggest that the barrier
from 14 to A should be small and in the range of ∼5.1 kcal/
mol (Figure 2, vide supra). Alternatively, our intramolecular
KIE value could also reflect a fast equilibrium between C-H
and C-D bound σ-complexes of benzene-1,3,5-d3 from a
π-complex prior to the C-H/C-D bond activation process.
Intermediate 14, which we identify as a σ-adduct, could of
course be in equilibrium with a corresponding π-complex,
treated with neat solvent, complex 2 generates A, which
subsequently reacts with the arene C-H bond of the solvent to
form the corresponding titanium alkylidene complex, having a
substituted aryl group ligand. Accordingly, when complex 2 is
dissolved in solvents such as fluorobenzene, 1,2-difluoroben-
zene, and 1,4-difluorobenzene, the alkylidene derivatives (PNP)-
t
TidCH Bu(Ar′) (Ar′ ) o-FC H (15), o-1,2-F C H (16), 1,4-
6
4
2
6
2
1
F C H (17)) are quantitatively generated when assayed via H,
2
13
6
3
t
x
31
(PNP)TitC Bu(η -C6H6) (x ) 2, 4, 6), and we have no method
C, and P NMR spectroscopy (Figure 9). Unfortunately,
toluene and m-xylene afforded at least three products when
judged by P{ H} NMR spectroscopy. When 1,3-difluoroben-
zene was used, two compounds were evident by P{ H} NMR
spectra (see Supporting Information), thus suggesting that aryl
C-H activation is not regioselective for the former and latter
to discard these possible intermediates during formation of the
5
6,59
31
1
σ-complex prior to the C-H bond-breaking event.
In fact,
0
31
1
given the ability of d metal complexes to readily form
π-complexes in general,3
4,48,60
the existence of a π-complex is
plausible to propose. Despite substantial efforts, we were unable
to find any stable adduct consistent with a π-complex in our
2
-6
computer models. All attempts at enforcing a Ti-η (C6H6)
binding mode converged to the σ-adduct 14 in our geometry
optimizations. Apparently, the titanium center in the Ti alky-
lidyne complex is too hard (or too sterically crowded) to
accommodate the soft π-base and prefers the relatively hard
σ-base instead.
(60) (a) Carpentier, J.-F.; Wu, Z.; Lee, C. W.; Stroemberg, S.; Christopher, J.
N.; Jordan, R. F. J. Am. Chem. Soc. 2000, 122, 7750-67. (b) Casey, C. P.;
Carpenetti, D. W., II; Sakurai, H. J. Am. Chem. Soc. 1999, 121, 9483-4.
(c) Casey, C. P.; Hallenbeck, S. L.; Pollock, D. W.; Landis, C. R. J. Am.
Chem. Soc. 1995, 117, 9770-1. (d) Casey, C. P.; Hallenbeck, S. L.; Wright,
J. M.; Landis, C. R. J. Am. Chem. Soc. 1997, 119, 9680-90. (e) Casey, C.
P.; Klein, J. F.; Fagan, M. A. J. Am. Chem. Soc. 2000, 122, 4320-30. (f)
Casey, C. P.; Lee, T.-Y.; Tunge, J. A.; Carpenetti, D. W., II. J. Am. Chem.
Soc. 2001, 123, 10762-3. (g) Casey, C. P.; Tunge, J. A.; Lee, T.-Y.; Fagan,
M. A. J. Am. Chem. Soc. 2003, 125, 2641-51. (h) Stoebenau, E. J., III;
Jordan, R. F. J. Am. Chem. Soc. 2003, 125, 3222-3. (i) Stoebenau, E. J.,
III; Jordan, R. F. J. Am. Chem. Soc. 2004, 126, 11170-1. (j) Stoebenau,
E. J., III; Jordan, R. F. J. Am. Chem. Soc. 2006, 128, 8638-50. (k)
Stoebenau, E. J., III; Jordan, R. F. J. Am. Chem. Soc. 2006, 128, 8162-
75. (l) Casey, C. P.; Fisher, J. J. Inorg. Chim. Acta 1998, 270, 5-7. (m)
Casey, C. P.; Fagan, M. A.; Hallenbeck, S. L. Organometallics 1998, 17,
Since we did not detect any intermolecular KIE (1.03(3)-
.07(3)), we also propose that our intramolecular KIE can be
1
interpreted to implicate an intermediate 14 that does not have
2
57
significant rehybridization at the sp carbon. Unfortunately,
our noncompetitive KIE experiment does not provide conclusive
evidence for the binding event being rate-determining after
formation of A, since the R-hydrogen abstraction from 2 is the
overall rate-determining step. Taken together, our experimental
and theoretical work suggests that the σ-complex 14 is a likely
intermediate that connects the titanium alkylidyne and the C-H
2
87-9. (n) Crowther, D. J.; Swenson, D. C.; Jordan, R. F. J. Am. Chem.
Soc. 1995, 117, 10403-4. (o) Galakhov, M. V.; Heinz, G.; Royo, P. Chem.
Commun. 1998, 17-8.
(
61) (a) Bosque, R.; Clot, E.; Fantacci, S.; Maseras, F.; Eisenstein, O.; Perutz,
R. N.; Renkema, K. B.; Caulton, K. G. J. Am. Chem. Soc. 1998, 120,
1
2634-640. (b) Reinhold, M.; McGrady, J. E.; Perutz, R. N. J. Am. Chem.
Soc. 2004, 126, 5268-76. (c) Clot, E.; Oelckers, B.; Klahn, A. H.;
Eisenstein, O.; Perutz, R. N. Dalton Trans. 2003, 4065-74.
8790 J. AM. CHEM. SOC.
9
VOL. 129, NO. 28, 2007

C−H Bond Activation Involving a TitC Linkage
A R T I C L E S
Figure 9. Synthesis of complexes 15-17 from precursor 2. The molecular structures of each product are displayed with thermal ellipsoids at the 50%
probability level. Solvent, disordered aryls, and H-atoms have been excluded for clarity. Selected metrical parameters (distances in angstroms, angles in
degrees): For 15, Ti-N10, 2.038(2); Ti-P2, 2.563(1); Ti-P17, 2.6175(9); Ti-C30, 1.865(3); Ti-C35, 2.138(6); Ti-C30-C31, 166.7(2); P2-Ti1-P17,
150.28(3). For 16, Ti-N10, 2.040(4); Ti-P2, 2.6305(7); Ti-P18, 2.5692(8); Ti-C31, 1.866(7); Ti-C36, 2.173(7); Ti-C31-C32, 166.5(4); P2-Ti1-P18,
150.47(8). For 17, Ti-N10, 2.060(2); Ti-P2, 2.5865(5); Ti-P18, 2.5589(5); Ti-C31, 2.210(6); Ti-C39, 1.864(6); Ti-C39-C40, 163.1(2); P2-Ti1-P18,
147.15(7).
Scheme 4
cases. For all reactions, C-H bond activation of the arene
typically occurs at 25 °C over 24 h. With the exception of
toluene, m-xylene, and 1,3-difluorobenzene, activation of the
arene C-H bond takes place exclusively ortho with respect to
the F group. Given our proposed reaction profile (Figure 2),
this result implies that coordination of such a group to A likely
takes place prior to the C-H activation step. It is reasonable to
speculate that binding of the arene substrate places the aryl C-H
group proximal to the TitC bond, to form the hypothetical
t
intermediate “(PNP)TitC Bu(Ar′H)” displayed in Figure 9
by 7.27 kcal/mol, due to the entropy and solvation penalties.
This result implies that an adduct such as (PNP)TitC Bu-
(
Ar′H ) fluoroarene). This might explain why the aryl C-H
t
bond proximal to the substituent is more amenable to activa-
tion.61 However, given our mechanistic C-H activation study
involving benzene, a species such as (PNP)TitC Bu(Ar′H)
might be a side intermediate that does not lie directly on the
C-H activation pathway, but rather rearranges to a π- and/or
σ-complex. We are currently examining the basis for regiose-
lectivity in compounds 15-17 by addressing the role of the
fluoride in the C-H activation process since incipient negative
charge at the ortho-carbon might be playing a crucial role in
these types of reactions.
(FC H ) might be in a pre-equilibrium with intermediate A, and
6
5
t
might explain why regioselectivity in the activation of the aryl
C-H bond is not always observed.
Although NMR spectroscopic data for all cases were con-
sistent with the proposed connectivity, we collected single-
crystal X-ray diffraction data of compounds 15-17 to further
confirm that activation of one aryl C-H bond is generally ortho
to the F substituent. Although the molecular structures of 15
and 16 suffered from co-crystallization of the two aryl isomers
(disorder from F pointing in and out of the CdTisCipso groove),
Our DFT studies suggest a dative fluorobenzene adduct,
t
all structures are consistent with the proposed connectivity and
clearly depict a short TidC bond and an obtuse TidCsC angle
(PNP)TitC Bu(FC6H5), as a viable intermediate along the C-H
bond activation step. Binding of FC6H5 to A slightly perturbs
the metrical parameters when compared to the computed
structure of A (Figure 8). The long Ti-F distance of 2.269 Å
and the F-C6H5 distance of 1.398 Å are indicative of a weak
interaction: in free fluorobenzene, the F-C6H5 distance is
(Figure 9). In each molecular structure, the interaction between
the Ti and the ortho group of the arene is negligible (Ti-F (15-
1
7) > ∼3.04 Å). Other pertinent features for the molecular
structures of 15-17 are given in Figure 9.
6
2
2.4. Intermolecular C-H Activation of Aliphatic Groups.
Complex 2 remains the most reactive alkylidene-alkyl spe-
∼
1.36 Å. In addition, the Ti-F distance observed in (PNP)-
t
TitC Bu(FC6H5) is much longer when compared to the only
structural example of a Ti(IV)-FC6H5 adduct (2.113(7) Å) for
the cationic titanium imide salt [(nacnac)TidNAr(FC6H5)]-
cies from our series and readily undergoes R-H abstraction
t
(
CH3 Bu is observed spectroscopically) in the presence of
-
t
i
62
alkanes such as cyclohexane-d12, hexane-d14, and pentane-d12
at 25 °C over 24 h. Unfortunately, reaction of 2 with these
solvents is not clean, hampering characterization of the products.
Despite these characterization problems, the most concrete
evidence for an aliphatic C-H group stems from the clean
transformation of 2 in neat SiMe4 (90 °C, 12 h) to provide
[B(C6F5)4] (nacnac ) [ArNC( Bu)]2CH, Ar ) 2,6- Pr2C6H3)).
t
Although the gas-phase electronic energy of (PNP)TitC Bu-
(
FC6H5) is downhill by 9.78 kcal/mol, the free energy is uphill
(62) Basuli, F.; Aneetha, H.; Huffman, J. C.; Mindiola, D. J. J. Am. Chem. Soc.
2
005, 127, 17992-3.
J. AM. CHEM. SOC.
9
VOL. 129, NO. 28, 2007 8791

A R T I C L E S
Bailey et al.
Figure 10. Multiple C-H activation reactions to prepare complexes 18-21. The molecular structures of 20 and 21 are displayed, depicting thermal ellipsoids
at the 50% probability level, with H-atoms excluded for clarity. Selected metrical parameters (distances in angstroms, angles in degrees): For 20, Ti-N10,
2
1
.072(1); Ti-P2, 2.5946(4); Ti-P18, 2.6129(5); Ti-C40, 1.883(3); Ti-C31, 2.112(4); Ti-C31-Si32, 129.94(8); Ti-C40-Si41, 143.07(8); P2-Ti1-P18,
49.30(4). For 21, Ti-N10, 2.015(9); Ti-P2, 2.6098(8); Ti-P18, 2.6181(8); Ti-C43, 1.891(3); Ti-C31, 2.191(3); Ti-C43-C44, 156.9(2); Ti-C31-
C32, 115.4(7); P2-Ti1-P18, 153.02(3).
Scheme 5
3
9
(Scheme 4). The latter process indicates that R-hydrogen
Moreover, the net activation of three sp C-H bonds of
abstraction, R-hydrogen migration, and 1,2-CH bond addition
solvent is not limited to SiMe alone, inasmuch as hydrocar-
4
processes are all playing a role along the reaction profile
bons can be similarly cleaved under mild conditions. For ex-
(
Scheme 5). We propose that formation of 9 proceeds first via
ample, 2 can also break three benzylic C-H bonds of 1,3,5-
trimethylbenzene to yield (PNP)TidCH{3,5-Me2C6H3}(CH2-
the rate-determining step, R-hydrogen abstraction to generate
intermediate A, which subsequently undergoes 1,2-CH bond
addition of tetramethylsilane across the TitC Bu linkage to
furnish the non-degenerate alkylidene-alkyl complex 4. As
noted previously with non-degenerate titanium alkylidene-alkyls,
complex 4 is amenable to R-hydrogen migration, thus equili-
brating to 5. By virtue of a second R-hydrogen abstraction,
{
3,5-Me2C6H3}) (18) quantitatively (Figure 10; see Supporting
t
Information for spectroscopic details). Monitoring the reaction
31
1
by P{ H} NMR identifies two intermediates, likely the
t
tautomers (PNP)TidCH Bu(CH2{3,5-Me2C6H3})/(PNP)TidCH-
t
{
4
3,5-Me2C6H3}(CH2 Bu), which form and decay over 48 h at
5 °C to ultimately produce 18 (Figure 10). Other reagents, such
t
intermediate 5 generates a second equivalent of CH3 Bu and
the (trimethylsilyl)methylidyne intermediate (PNP)TitCSiMe3
as Me3SiCtCSiMe3, (Me3Si)2O, and MeC6F5, also react cleanly
with 2 to generate a sole product resulting from triple C-H
activation of the methyl groups of the substrate, to produce the
compounds (PNP)TidCHSiMe2CCSiMe3(CH2SiMe2CCSiMe3)
(B), and the latter undergoes activation of another C-H bond
of tetramethylsilane (which is present in large excess) to
ultimately produce 9 (Schemes 4 and 5). In fact, independent
experiments confirm that the 4/5 mixture (isolated from 2 and
LiCH2SiMe3 in C6H12, vide supra) cleanly reacts with tetra-
methylsilane to afford 9, therefore suggesting that these are
viable intermediates along the 2 f 9 transformation. Formation
(
(
19), (PNP)TidCHSiMe2OSiMe3(CH2SiMe2OSiMe3) (20), and
PNP)TidCHC6F5(CH2C6F5) (21) (Figure 10). We believe that
formation of 19-21 also proceeds by a combination of two
R-hydrogen abstraction steps, one R-hydrogen migration, and
two 1,2-CH bond additions across two different transient
titanium alkylidyne intermediates, the first of which entails A
(Scheme 5). Complexes 18-21 have also been fully character-
3
of 9 from 2 inVolVes an unprecedented actiVation of three sp
C-H bonds of solVent under mild conditions by a transition
metal complex,63 thereby insinuating that the TitCSiMe3
1
13
31
ized by H, C, and P NMR spectra. Most remarkably, the
molecular structures of 20 and 21 portray unique examples of
transition metal complexes containing a trimethylsilyl ether or
perfluoroaryl-substituted alkylidene moiety (TidC, 1.883(3) Å,
20; TidC, 1.891(3) Å, 21, Figure 10).
functionality in B is far more within reach thermodynamically
than the prototypical TitC Bu ligand in A.
t
(
63) (a) Lee, J.-H.; Pink, M.; Caulton, K. G. Organometallics 2006, 25, 802-
. (b) Ozerov, O. V.; Pink, M.; Watson, L. A.; Caulton, K. G. J. Am. Chem.
Soc. 2004, 126, 2105-13.
4
8792 J. AM. CHEM. SOC.
9
VOL. 129, NO. 28, 2007

C−H Bond Activation Involving a TitC Linkage
A R T I C L E S
3
. Conclusions
governed kinetically, since activation of the least hindered C-H
group is typically observed. In some cases, however, the
combination of two R-hydrogen abstractions, an R-hydrogen
migration, and two 1,2-CH bond additions can result in the net
cleavage of five C-H bonds, three of which derive from solvent
molecules. This series of well-understood reactions that include
multiple C-H bond cleavages, two of which were intermo-
lecular while three were intramolecular, has allowed us to
incorporate novel substituted alkylidene moieties onto the
titanium center. Therefore, this new process can be best defined
as an alkylidene-alkyl metathetical reaction with hydrocarbons,
whereby alkylidene and alkyl moieties exchange with two
equivalents of hydrocarbon species. The search for systems
properly tuned for the stabilization of a terminal titanium
alkylidyne is an ongoing project in our laboratory since this
parameter will allow us to selectively harness the intrinsic power
stored in the TitC linkage. Although the work discussed in
this paper encompasses stoichiometric reactions, a mechanistic
understanding of the intermolecular binding and activation of
C-H bonds represents an important step toward the rational
and selective functionalization of inert substrates with low
binding affinities. This parameter could, in principle, aid future
efforts for the systematic design of more efficient catalysts
involved in processes such as hydrocracking and alkane
metathesis.
In this work, we have demonstrated that titanium alkylidene-
alkyl complexes can undergo R-hydrogen abstraction reactions
smoothly to generate transient titanium alkylidyne intermediates,
which display a remarkably versatile reactivity toward unacti-
vated C-H bonds in a number of common solvents. While
intramolecular C-H activation of an aryl moiety by a tungsten
neopentylidyne has been documented,64 our work clearly shows
the first example of intermolecular C-H bond-breaking
processes promoted by TitC linkages.64 We have utilized an
integrated approach combining experimental and theoretical
tools at our disposal to examine the mechanistic details of these
novel chemical transformations. The R-hydrogen abstraction step
leading to the reactive intermediate that contains an unprec-
edented TitC linkage is rate-determining in all cases that we
considered. The titanium alkylidyne complex undergoes facile
1
,2-CH bond addition of hydrocarbons or close derivatives
thereof across the polarized TitC bond to yield new alky-
lidene-alkyl products. When benzene is used as a reactant, the
thermodynamic product can slowly equilibrate back, at tem-
peratures above 95 °C, to the alkylidyne intermediate, thereby
creating an independent route to the alkylidyne functionality in
benzene from the post-rate-determining step. The notion that
the titanium alkylidyne functionality can be formed from a
starting material such as 2 or a thermodynamic product such as
3
offers an excellent opportunity to not only study C-H
Acknowledgment. We thank Indiana University-Blooming-
activation reactions in neat solvent using 2 as the precursor,
but also conduct stoichiometric reactions in camouflaging
solvents such as benzene by employing 3 as the precursor. Our
kinetic studies, from both theoretical and experimental stand-
points, suggest that a σ-complex (14) is formed along the C-H
bond activation step, which takes place in a heterolytic fashion
to afford a highly charge-polarized transition state. C-H bond
activation and rupture are promoted by the LUMO of A in
concert with the HOMO orbital, the latter serving as a potent
nucleophile in the 1,2-CH addition process. We established that
the reactivity of the alkylidyne functionality follows a decreasing
ton, the Dreyfus Foundation, the Sloan Foundation, NSF (CHE-
0
348941, PECASE Award to D.J.M.; CHE-0645381 to M.-
H.B.), and NIH (HG003894 to M.-H.B.) for financial support
of this research. M.-H.B. is a Cottrell Scholar of the Research
Corporation. B.C.B. acknowledges the Department of Education
and the Lubrizol Foundation for Fellowships. The authors thank
Prof. P. T. Wolczanski for advice and insightful suggestions,
and all of the reviewers for providing invaluable feedback on
this work.
Supporting Information Available: Complete kinetic data
(conversions of 2 to 3 and 3 to 3-d6, KIEs, EIEs), synthetic and
spectroscopic NMR data for all compounds, and computational
details (2, 3, relevant intermediates along the 2 f 3 coordinate,
t
trend in the order TitC Bu . TitCSiMe3 g TitCPh. In most
cases, aryl C-H activation of substituted arenes is preferred
over aliphatic C-H bonds since the group of the aryl motif
heavily directs orthometalation via coordination to the low-
t
other viable pathways and intermediates, and (PNP)TitC Bu-
t
(σ-FC6H5)) (PDF). X-ray crystallographic data for compounds
coordinate and transient intermediate (PNP)TitC Bu. In general,
11, 12, 15-17, 20, and 21 (CIF). This material is available free
it appears that intermolecular C-H activation reactions are
of charge via the Internet at http://pubs.acs.org.
(
64) Couturier, J. L.; Paillet, C.; Leconte, M.; Basset, J. M.; Weiss, K. Angew.
Chem. 1992, 104, 622-4.
JA070989Q
J. AM. CHEM. SOC.
9
VOL. 129, NO. 28, 2007 8793