1,2-CF bond activation of perfluoroarenes and alkylidene isomers of titanium. DFT analysis of the C-F bond activation pathway and rotation of the titanium alkylidene moiety

Article abstract of DOI:10.1016/j.jorganchem.2011.07.037

Isomeric alkylidene complexes syn- and anti-(PNP)Ti[C^tBu(C ₆F₅)](F) (1) and (PNP)Ti[C^tBu(C ₇F₇)](F) (2) have been generated from C-F bond addition of hexafluorobenzene (C₆F₆) and octafluorotoluene (C ₇F₈) across the alkylidyne ligand of transient (PNP)Ti≡C^tBu (A) (PNP^-N[2-P(CHMe₂) ₂-4-methylphenyl]₂), which was generated from the precursor (PNP)TiCH^tBu(CH₂^tBu). Two mechanistic scenarios for the activation of the C-F bond by A are considered: 1,2-CF addition and [2 + 2]-cycloaddition/β-fluoride elimination. Upon formation of the alkylidenes 1 and 2, the kinetic and thermodynamic alkylidene product is the syn isomer, which gradually isomerizes to the corresponding anti isomer to ultimately establish an equilibrium mixture (when using 1, 65/35) if the solution is heated in benzene to 105 °C for 1 h. Single crystal X-Ray crystallographic data obtained for the two isomers of 2 (and syn isomer of 1) are in good agreement with computed DFT-optimized models. Our calculations suggest convincingly that the isomerization process proceeds via a concerted rotation involving a heterolytic bond cleavage about the alkylidene bond. The two rotamers are thermodynamically very close in energy and interconvert with an estimated barrier of ～26 kcal/mol. The electronic reason for this unexpectedly low barrier is investigated.

Full text of DOI:10.1016/j.jorganchem.2011.07.037

Journal of Organometallic Chemistry 696 (2011) 4138e4146
Contents lists available at SciVerse ScienceDirect
Journal of Organometallic Chemistry
journal homepage: www.elsevier.com/locate/jorganchem
1,2-CF bond activation of perfluoroarenes and alkylidene isomers of titanium.
DFT analysis of the CeF bond activation pathway and rotation of the titanium
alkylidene moiety
*
*
José G. Andino, Hongjun Fan, Alison R. Fout, Brad C. Bailey, Mu-Hyun Baik , Daniel J. Mindiola
Department of Chemistry, Indiana University, Bloomington, IN 47405, USA
a r t i c l e i n f o
a b s t r a c t
Isomeric alkylidene complexes syn- and anti-(PNP)Ti][CtBu(C6F5)](F) (1) and (PNP)Ti][CtBu(C7F7)](F)
(2) have been generated from CeF bond addition of hexafluorobenzene (C6F6) and octafluorotoluene
(C7F8) across the alkylidyne ligand of transient (PNP)TihCtBu (A) (PNPꢀ]N[2-P(CHMe2)2-4-
methylphenyl]2), which was generated from the precursor (PNP)Ti]CHtBu(CH2t Bu). Two mechanistic
scenarios for the activation of the CeF bond by A are considered: 1,2-CF addition and [2 þ 2]-cycload-
Article history:
Received 26 May 2011
Received in revised form
26 July 2011
Accepted 27 July 2011
dition/b-fluoride elimination. Upon formation of the alkylidenes 1 and 2, the kinetic and thermodynamic
Keywords:
alkylidene product is the syn isomer, which gradually isomerizes to the corresponding anti isomer to
ultimately establish an equilibrium mixture (when using 1, 65/35) if the solution is heated in benzene to
105 ꢁC for 1 h. Single crystal X-Ray crystallographic data obtained for the two isomers of 2 (and syn
isomer of 1) are in good agreement with computed DFT-optimized models. Our calculations suggest
convincingly that the isomerization process proceeds via a concerted rotation involving a heterolytic
bond cleavage about the alkylidene bond. The two rotamers are thermodynamically very close in energy
and interconvert with an estimated barrier of w26 kcal/mol. The electronic reason for this unexpectedly
low barrier is investigated.
Alkylidene
Titanium
CeF bond activation
1,2-CF bond addition
Rotation
Ó 2011 Elsevier B.V. All rights reserved.
1. Introduction
In both cases, mixtures of syn and anti alkylidene rotational isomers
are observed in solution [2]. 31P NMR spectroscopy allowed the
Carbonefluorine bond activation in aromatic substrates
continues to be a challenge in organotransition metal chemistry for
multiple reasons. With a typical bond strength of w154 kcal/mol the
carbon fluoride bond in molecules such as C6F6 tend to be chemically
inert and resistant to processes like oxidative degradation [1].
Generally, fluorocarbons display poor binding affinities, especially
observation of the equilibrium (65:35 ratio of 1 syn:anti) of these
samples when heated at 105 ꢁC for 1 h. Titaniumealkylidene
complexes are not known to undergo rotation about the MeL double
bond readily and the coexistence of two alkylidene rotamers in
equilibrium is an exceedingly rare phenomenon [2,3]. Even though
alkylidene isomers for Schrock type alkylidenes are known for the
heavier congeners of group 5, 6 and 7 transition metals [4e6],
examples of isomers for group 4 transition metals have not been
documented until recently [2,3]. Understanding the rotationprocess
is important since knowing howeach isomer is favoredordisfavored
may allow for better controlling the reactivity of the alkylidene
catalyst formed in reaction (e.g. alkene metathesis and the poly-
merization of cyclic olefins). This feature is especially true since it is
now well documented that alkylidene isomers have different reac-
tivity and selectivity [4e6]. For instance, Schrock found that the anti
alkylidene isomer of (RO)2Mo]NAr(CHCMe2Ph) (R ¼ OCMe3,
OCMe2(CF3), OCMe(CF3)2, OC(CF3)2(CF2CF2CF3); Ar ¼ 2,6-iPr2C6H3),
is at least two orders of magnitude more reactive for the ring-
opening metathesis of bis(trifluoromethyl)norbornadiene than the
corresponding syn isomer [5]. It has been argued that the barrier to
for saturated substrates given the lack of p-frameworks that may be
susceptible to bind to a metal center [1]. Activating CeF bonds by
metalecarbon multiple bonds is conceptually a plausible strategy,
but successful execution has proven difficult with the only reported
example being promoted by a transient titanium alkylidyne (PNP)
TihCtBu (A) (PNPꢀ
Intermediate
]
N[2-P(CHMe2)2-4-methylphenyl]2) [2].
A
can cleave the CeF linkage in perfluoro
substrates like C6F6 and CF3C6F5 under mild conditions to
afford rare examples of disubstituted titanium alkylidene-fluorides,
(PNP)Ti]C[tBu(ArF)](F) (ArF¼C6F5, 1; p-CF3C6F4, 2) depicted in
Scheme 1 (the drawing of the simplified cartoon represents PNPꢀ).
* Corresponding authors. Fax: þ1 812 855 8300.
E-mail address: mindiola@indiana.edu (D.J. Mindiola).
0022-328X/$ e see front matter Ó 2011 Elsevier B.V. All rights reserved.
doi:10.1016/j.jorganchem.2011.07.037

J.G. Andino et al. / Journal of Organometallic Chemistry 696 (2011) 4138e4146
4139
(which was stirred for 2 days prior to transfer). All other chemicals
were used as received. 1H, 13C, 19F, and 31P NMR spectra were
P
P
P
P
t
CH2 Bu
recorded on Varian 400 or 300 MHz NMR spectrometers. 1H and 13
C
N
Ti
N
Ti
Ti
t
CHtBu
X
CtBu
- CH3 Bu
NMR are reported with reference to residual solvent resonances
(7.16 and 128.0 ppm for C6H6 in C6D6). 19F NMR chemical shifts are
reported with respect to external HOCOCF3 (ꢀ78.5 ppm). 31P NMR
chemical shifts are reported with respect to external H3PO4
(aqueous solution, 0.0 ppm).
X
A
F4
P
P
P
tBu
C
F4
F5
C
2.2. Separation of (PNP)Ti][CtBu(C6F5)](F) (1-syn) from the
mixture
N
Ti
N
tBu
F
F
P
X
In
a
vial was dissolved (PNP)Ti]CHtBu(CH2t Bu) [118 mg,
anti isomer
X = F, (1); X = CF3, (2)
syn isomer
X = F, (1); X = CF3, (2)
0.191 mmol] in hexafluorobenzene (w2 mL, excess, passed through
a short column of alumina) at room temperature. The solution was
allowed to stand for 2 days at 50 ꢁC where the solution changed
from green to dark-red. The solution was dried under vacuum and
the residue was extracted with pentane and filtered. The filtrate
was reduced in volume under reduced pressure, and then cooled
to ꢀ35 ꢁC. Red and orange crystals of (PNP)Ti][CtBu(C6F5)](F) (1)
[108 mg, 0.148 mmol, 78% yield] were collected and the crude 31P
NMR spectrum showed a mixture of alkylidene isomers to be in
approximately a 65:35 ratio. Fractional crystallization of the solids
allowed for isolation of the 1-syn isomer. For (1-syn, 85% pure with
Scheme 1. Synthesis of complexes 1 and 2 (and their respective isomers) from CeF
bond activation of C6F6 and C7F8 via transient (PNP)TihCtBu (A). The PNPꢀ pincer
ligand has been simplified to a caricature.
rotation of these alkylidene isomers is sensitive to the availability of
-type d-orbital orthogonal to the M]C bond as well as an
interaction of an -CH agostic bond with the metal center [4]. To
prohibit the interconversion of syn and anti isomers in solution the
groups of Fürstner [7,8] and Grubbs [9,10] prepared tethered
ruthenium carbene complexes with carbene ligands connected to an
N-heterocyclic carbene. Odom reported the only examples of early-
transition metal alkylidene complexes tethered to the adjacent
imido ligand (Schrock-type catalysts) [11].
We have shown previously that transient titanium
alkylidyne complexes supported by the tridentate PNP ligand
(PNPꢀ ] N[2-P(iPr)2-4-methylphenyl]2) can activate CeH bonds
present in aromatic and aliphatic substrates [3,12e14]. In addition
to promoting intermolecular CeH bonds activation, we have also
utilized the highly reactive alkylidyne moiety to activate other
strong bonds like the CeF bond of perfluoroaromatics, the CeO
bonds in ethers and the aromatic CeN bond in pyridine [2,15e17].
In the case of perfluoroaromatics, we can prepare the disubstituted
alkylidene ligands carrying an aliphatic group such as tBu, as well as
a perfluoroaromatic group on titanium (Scheme 1). These unusual
alkylidene complexes exhibit very long Ti]C distances and exist as
alkylidene rotational isomers in solution [2]. The goal of this
study is to understand how a titanium alkylidyne, A, activates the
CeF bond of perfluoroarenes (hexafluorobenzene and octa-
fluorotoluene) but also to elucidate the mechanism behind
formation of the two titanium alkylidene rotational isomers.
a
p
a
15% of the anti isomer): 1H NMR (23 ꢁC, 399.8 MHz, C6D6):
d 7.18 (br
d, 1H, JHeH ¼ 6.04 Hz, C6H3), 6.86e6.88 (m, 1H, C6H3), 6.70e6.77 (m,
3H, C6H3), 6.62 (m, 1H, C6H3), 2.64e2.76 (m, 1H, CHMe2), 2.41e2.52
(m, 1H, CHMe2), 2.28 (s, 3H, C6H3eCH3), 2.22 (m, 1H, CHMe2), 2.05
(s, 3H, C6H3eCH3), 1.73e1.84 (m, 1H, CHMe2), 1.48e1.59 (m, 6H,
CHMe2), 1.47 (s, 9H, Ti]CCMe3), 1.34 (dd, 3H, JPeH ¼ 14.70 Hz,
JHeH ¼ 7.14 Hz, CHMe2),1.26 (dd, 3H, JPeH ¼ 13.19 Hz, JHeH ¼ 6.87 Hz,
CHMe2), 1.14 (dd, 3H, JPeH ¼ 15.25 Hz, JHeH ¼ 7.15 Hz, CHMe2), 1.00
(dd, 3H, JPeH ¼ 14.56 Hz, JHeH ¼ 6.87 Hz, CHMe2), 0.86 (dd, 3H,
JPeH
¼
14.97 Hz, JHeH
¼
7.15 Hz, CHMe2), 0.72 (dd, 3H,
JPeH ¼ 12.91 Hz, JHeH ¼ 7.14 Hz, CHMe2). 13C NMR (23 ꢁC,100.6 MHz,
C6D6): d
317.0 (Ti]CtBuC6F5), 160.2 (d, C6H3), 155.9 (dd, C6H3), 133.1
(C6H3),132.9 (C6H3),132.8 (C6H3),132.0 (C6H3),131.8 (d, C6H3),126.7
(d, C6H3), 124.4 (d, C6H3), 121.5 (d, C6H3), 119.7 (d, C6H3), 114.1 (d,
C6H3), 48.2 (Ti]CCMe3C6F5), 31.1 (Ti]CCMe3C6F5), 25.3 (d,
CHMe2), 23.4 (d, CHMe2), 22.1 (d, CHMe2), 21.0 (CHMe2), 20.4
(CHMe2), 20.3 (d, CHMe2), 19.5 (C6H3eCH3), 19.2 (C6H3eCH3), 18.8
(d, CHMe2), 18.0 (d, CHMe2), 17.9 (CHMe2), 17.7 (CHMe2), 17.4 (d,
CHMe2), 16.1 (d, CHMe2). The aryl carbon resonances of the C6F5
were broad and poorly defined and thus were not included. 31P
NMR (23 ꢁC, 121.5 MHz, C6D6):
d
33.1 (dd, JPeP ¼ 40 Hz,
JPeF ¼ 20 Hz), 18.7 (dd, JPeP ¼ 40 Hz, JPeF ¼ 20 Hz). 19F NMR (23 ꢁC,
2. Experimental details
282.3 MHz, C6D6):
d
195.3 (t, JPeF ¼ 29 Hz, TieF), ꢀ133.1 (br s, Ti]
CtBuC6F5), ꢀ163.3 (t, Ti]CtBuC6F5), ꢀ166.6 (t, 2F, Ti]CtBuC6F5).
2.1. General considerations
For the mixture of 1: 1H NMR (23 ꢁC, 399.8 MHz, C6D6):
d 7.15 (d),
7.00 (d), 6.97 (d), 6.87 (d), 6.84 (d), 6.76 (d), 6.69, 6.70, 6.65, 6.64,
6.63, 6.61, 6.60, 6.59, 6.58, 2.84 (septet), 2.71 (septet), 2.45 (septet),
2.28, 2.25, 2.23, 2.05, 2.01, 1.77 (septet), 1.58, 1.56, 1.54, 1.53, 1.52,
1.44, 1.43, 1.41, 1.40, 1.37, 1.355, 1.33, 1.31, 1.30, 1.28, 1.26, 1.24, 1.07,
1.06, 1.04, 0.98, 0.96, 0.89, 0.87, 0.85, 0.82, 0.79, 0.74, 0.72, 0.70, 0.69.
Unless otherwise stated, all operations were performed in an M.
Braun Lab Master double-dry box under an atmosphere of purified
nitrogen or using high vacuum standard Schlenk techniques under
an argon atmosphere. Non-protic solvents such as n-hexane, n-
pentane, toluene, benzene, and Et2O were dried according to
literature procedures [18]. Deutero solvents were purchased from
Cambridge Isotope Laboratory (CIL), degassed and vacuum trans-
ferred to 4 Å molecular sieves. Celite, alumina, and 4 Å molecular
sieves were activated under vacuum overnight at 200 ꢁC.
Compound (PNP)Ti]CHtBu(CH2t Bu), was prepared according to the
literature [13]. The non-optimized syntheses of complexes 1 and 2
have been previously communicated [2]. All other solvents used as
reagents were dried by passage through an activated alumina
column and if necessary, vacuum transferred from a CaH2 mixture
13C NMR (23 ꢁC, 100.6 MHz, C6D6):
d 323.5, 317.0, 160.3 (d), 159.0 (d),
155.9 (dd), 153.7 (dd), 133.1, 132.9, 132.8, 132.7, 132.0, 131.8 (d), 130.6
(d), 126.8 (d), 126.7 (d), 125.2 (d), 124.4, 123.2 (br t), 122.4 (d), 121.7
(d), 121.5 (d), 119.6 (d), 114.1 (d), 114.0 (d), 48.2, 45.6, 32.9, 31.1, 26.0
(d), 25.3 (d), 23.9 (d), 23.5, 23.4, 23.4, 23.3, 23.2, 23.0, 22.9, 22.7, 22.1
(d), 21.0, 20.9, 20.4, 20.3,19.5,19.5,19.4,19.3,19.2,19.0,18.7,18.0,17.7,
17.4 (d), 16.1 (d). 31P NMR (23 ꢁC, 121.5 MHz, C6D6):
d 33.1 (dd,
JPeP ¼ 39 Hz, JPeF ¼ 21 Hz), 32.0 (ddd, JPeP ¼ 42 Hz, JPeF ¼ 23 Hz,
0
JPeF ¼ 6 Hz), 19.6 (dd, JPeP ¼ 39 Hz, JPeF ¼ 21 Hz), 18.7 (ddd,
JPeP ¼ 42 Hz, JPeF ¼ 23 Hz, JPeF ¼ 6 Hz). 19F NMR (23 ꢁC, 282.3 MHz,
0

4140
J.G. Andino et al. / Journal of Organometallic Chemistry 696 (2011) 4138e4146
C6D6):
d
ꢀ131.4 (d), ꢀ133.1 (br s), ꢀ134.2 (d), ꢀ161.9 (t), ꢀ163.4
geometry using Dunning’s correlation-consistent triple-z basis set
[26], cc-pVTZ(-f). For all transition metals we used a modified
version of LACVP, designated LACV3P, in which the exponents were
(t), ꢀ165.5 to ꢀ165.3 (m), ꢀ166.7 (br t), ꢀ166.9, ꢀ166.8 (m).
2.3. Preparation of (PNP)Ti][CtBu(C7F7)](F) (2-syn/2-anti)
decontracted to match the effective core potential with the triple-
z
quality basis. Vibrational frequency calculations based on analytical
second derivatives at the B3LYP/6-31G** (LACVP) level of theory
were carried out to derive the zero-point-energy (ZPE) and entropy
corrections at room temperature (unless otherwise noted) utilizing
unscaled frequencies. Full models were used in our calculations and
for simplicity, the rotation study was conducted using the isomers of
1 only, as 2 differs only by having a CF3 para substituent on the
perfluorinated phenyl ring.
In
a
vial was dissolved (PNP)Ti]CHtBu(CH2t Bu) [101 mg,
0.164 mmol] in octafluorotoluene (w2 mL, excess, passed through
a short column of alumina) at room temperature. The solution was
allowed to stand for 2 days at 50 ꢁC where the solution changed from
green to dark-red. The solution was dried under vacuum and the
residue was extracted with pentane and filtered. The filtrate was
reducedinvolumeunderreducedpressure, andthencooledtoꢀ35 ꢁC.
Orange crystals of the syn isomer (PNP)Ti][CtBu(C7F7)](F) 2-syn
[78 mg, 0.100 mmol, 61% yield] were collected but these are always
marred with significant amount of the anti isomer (dark orange).
For 2-syn: Crystals were manually picked from the crude
mixture and analyzed by multinuclear NMR. 1H NMR (23 ꢁC,
3. Results and discussion
Activation of CeF bonds by metalecarbon multiple
bonds is exceedingly rare with the only reported example
being the transient titanium alkylidyne (PNP)TihCtBu (A)
(PNPꢀ ] N[2-P(CHMe2)2-4-methylphenyl]2) [2]. Species A can
cleave the CeF linkage in perfluoro substrates like C6F6 and C7F7
under mild conditions to afford disubstituted alkylidene-
fluorides, (PNP)Ti]C[tBu(ArF)](F) (ArF ¼ C6F5, 1; C7F7, 2) depic-
ted in Scheme 1 (the lines represents PNPꢀ) [2]. In both cases,
mixtures of syn and anti alkylidene isomers are observed in
solution. In the case of complex 1, fractional crystallization
allowed for crystallographic characterization of the syn rotamer,
(PNP)Ti]C[tBu(C6F5)](F), and in the case of complex 2, both the
syn and anti isomers of (PNP)Ti]C[tBu(C7F7)](F) could be
physically separated based on their distinctively different colors.
At various temperatures it was also confirmed that these rota-
tional isomers slowly interconvert [2]. To our knowledge, the
only other example of a 1,2-CF bond activation by a metal-
eligand multiple bond was reported by Bergman and co-
workers using the racemic ansa-metallocene imido rac-(ethyl-
enebis(tetrahydro)indenyl)Zr]NtBu (generated in solution) and
NC5F5, to yield the amide fluoride (ethylenebis(tetrahydro)
indenyl)Zr(NtBu[NC5F4])F (Eq. (1)) [27]. In such a reaction, the
CeF breaking event was proposed to be driven by a combination
of the inherent reactivity of the pyridine CeF motif as well as
formation of a very strong ZreF bond. Naturally, one would
predict Bergman’s activation of NC5F5 to obey an analogous
pathway for the CeF bond breaking event observed between
transient A and the perfluoroarene given that a metaleligand
multiple bond is responsible. However, the nature of the
substrates, perfluoropyridine versus perfluorobenzene, should
render these two reactions fundamentally different. For
example, N-coordination of the perfluoropyridine substrate to
the metal center most likely plays an important role along the
bond-breaking event, which cannot be present in the case of the
perfluoroarene.
399.8 MHz, C6D6):
d
7.19 (br d, 1H, JHeH ¼ 5.77 Hz, C6H3), 6.85e6.88
(br d, 2H, JHeH ¼ 7.42 Hz, C6H3), 6.66e6.72 (m, 2H, C6H3), (m, 1H,
C6H3), 2.62e2.74 (m, 1H, CHMe2), 2.37e2.49 (m, 1H, CHMe2), 2.31
(C6H3eCH3), 2.15e2.29 (m, 1H, CHMe2), 2.05 (C6H3eCH3), 1.71e1.83
(m, 1H, CHMe2), 1.44e1.55 (m, 6H, CHMe2), 1.43 (s, 9H, Ti]
CHCMe3C7F7), 1.32 (dd, 3H, JPeH ¼ 14.83 Hz, JHeH ¼ 7.14 Hz, CHMe2),
1.23 (dd, 3H, JPeH ¼ 13.46 Hz, JHeH ¼ 6.87 Hz, CHMe2), 1.12 (dd, 3H,
JPeH
JPeH
JPeH
¼
¼
¼
10.17 Hz, JHeH
14.70 Hz, JHeH
15.94 Hz, JHeH
¼
¼
¼
6.87 Hz, CHMe2), 0.99 (dd, 3H,
6.86 Hz, CHMe2), 0.86 (dd, 3H,
7.15 Hz, CHMe2), 0.70 (dd, 3H,
JPeH ¼ 12.91 Hz, JHeH ¼ 7.14 Hz, CHMe2). 31P NMR (23 ꢁC, 121.5 MHz,
C6D6):
d
33.3 (dd, JPeP ¼ 53 Hz, JPeF ¼ 27 Hz), 18.9 (dd, JPeP ¼ 53 Hz,
JPeF ¼ 27 Hz). 19F NMR (23 ꢁC, 282.3 MHz, C6D6):
3F), ꢀ131.8 (s, 2F), ꢀ145.1 (m, 2F).
d
ꢀ55.9 (m,
Assignment of the mixture of the two isomers was not possible,
therefore only resonances and multiplicity when known are given.
1H NMR (23 ꢁC, 399.8 MHz, C6D6):
d 7.18 (br d), 7.14 (d), 7.01, 6.99,
6.95e6.96 (m), 6.79e6.87 (m), 6.57e6.71 (m), 2.82 (septet), 2.67
(septet), 2.42 (septet), 2.31, 2.27, 2.22 (septet), 2.19, 2.06, 2.03, 2.00,
1.77 (septet), 1.46e1.55 (m), 1.42, 1.21e1.40 (m), 1.13 (dd), 1.03, 1.01,
0.98 (dd), 0.78e0.89 (m), 0.70 (dd). 13C NMR (23 ꢁC, 100.6 MHz,
C6D6): d 320.6, 315.7, 160.2 (d), 158.8 (d), 155.8 (d), 153.2 (dd), 133.7,
133.2, 133.1, 133.1, 132.9, 132.1, 132.0, 131.8 (d), 130.7, 127.2 (d), 126.9
(d), 124.8, 124.5, 124.4, 124.3, 122.7 (d), 122.1, 122.0, 121.8, 121.6,
121.3, 119.7 (d), 114.2, 114.1, 114.0, 113.9, 48.2, 45.6, 33.0, 31.2, 31.1,
26.0 (d), 25.3 (d), 25.1 (d), 24.5 (d), 23.9 (d), 23.5, 23.4, 22.9, 22.8,
22.7, 22.3, 22.0, 21.9, 21.0, 20.8, 20.6, 20.5, 20.4, 20.3, 20.2, 19.4, 19.3,
19.2, 19.0, 18.9, 18.7, 18.0, 17.9, 17.8, 17.7, 17.5, 17.3, 17.2, 17.1, 16.1 (d).
31P NMR (23 ꢁC, 121.5 MHz, C6D6):
d
33.3 (dd, JPeP ¼ 53 Hz,
JPeF ¼ 27 Hz), 31.5e32.9 (m), 19.4e20.2 (m), 18.9 (dd, JPeP ¼ 53 Hz,
JPeF ¼ 27 Hz). 19F NMR (23 ꢁC, 282.3 MHz, C6D6):
ꢀ55.7 (m), ꢀ55.9
d
(m), ꢀ108.3, ꢀ108.9 (br s), ꢀ117.5 (d), ꢀ117.9 (m), ꢀ118.9 (d), ꢀ140.5
(m), ꢀ141.3 to ꢀ141.0 (m), ꢀ142.7 (m), ꢀ144.4 (m), ꢀ145.5
to ꢀ145.2 (m), ꢀ147.9 (m), ꢀ160.3 (m), ꢀ164.6 (m), ꢀ165.4
(m), ꢀ166.2 (m).
To assess whether or not binding of the substrate could alter
the pathway along the CeF activation reaction we explored two
types of mechanistic scenarios for the formation of the alkyli-
dene fluoride, 1-syn, computationally using DFT and focusing on
the substrate C6F6. As noted above, Bergman and co-workers
reported the CeF activation of NC5F5 by racemic (ethyl-
enebis(tetrahydro)indenyl)Zr]NtBu, but the mechanism leading
to CeF bond cleavage in the substrate was not discussed [27]. If
binding of the pyridine followed by cycloaddition of the N-
heterocycle were the steps preceding the CeF bond activation
2.4. Computational details
All calculations were carried out using Density Functional
Theory as implemented in the Jaguar 7.0 suite [19] of ab initio
quantum chemistry programs. Geometry optimizations were per-
formed with the B3LYP [20e22] functional and the 6-31G** basis set
with no symmetry restrictions. This protocol has been shown to be
a reasonable compromise between model accuracy and computa-
tional cost for transition metal-containing systems [23]. All transi-
tion metals were represented using the Los Alamos basis set
(LACVP) [24,25]. Energies of the optimized structures were reeval-
uated by additional single-point calculations on each optimized
reaction, then a b-fluoride elimination process should be domi-
nating. On the other hand, formation of the alkylidene species 1
and 2 shown in Scheme 1 may proceed by a more direct 1,2-CF
addition across the Ti^C or may involve initial coordination of
the fluorine followed by migration of the aryl carbon to the

J.G. Andino et al. / Journal of Organometallic Chemistry 696 (2011) 4138e4146
4141
N
F4
t
u
N
NH B
F5
r
Z
.
E
q
(
r
Z
r
Z
1
)
-
t
u
N B
H
C
4
N
B
H
C
3
t
u
F
rac
rac
alkylidyne carbon. Experimentally discerning which pathway is
involved is virtually impossible due to the difficulty in isotopi-
cally labeling both the C or F sites among many other problems.
Previous studies established that formation of (PNP)TihCtBu is
often rate-determining based on our KIE values obtained for the
rates of (PNP)Ti]CHtBu(CH2t Bu)/(PNP)Ti]CDtBu(CDt2Bu) along
the CeH activation of benzene, hydrocarbons, and hydro-
fluoroarenes (KIE ¼ 3.7, 27 ꢁC) [3,12e14] as well as the activation
and ring-opening metathesis of pyridine (KIE ¼ 3.8(3), 25 ꢁC)
[15]. Therefore, experimentally investigating steps after the
elimination of CHt3Bu could yield little mechanistic information.
determining in this type of reaction, as we have observed in
many other reactions (the rates of these reactions are also similar
with a t1/2 in w3.1 h at room temperature).
The computed pathway depicted by a dashed line also presents
a coordination event (A10), followed by a stepwise process invoking
first a [2 þ 2]-cycloaddition of C6F6 to form the metallacyclobutene
intermediate (PNP)Ti(CtBu[C6F6]) (A2), which then
b-fluoride
eliminates to yield 1-syn. Although an intermediate invoking the
facial binding of the -system in C6F6 may exist, we were unable to
p
locate this intermediate labeled as A10 with our current theoretical
methods. We marked this putative reactant complex with a “?” in
Fig. 1 and placed it at a reasonable energy on the reaction coordi-
nate for illustration. The formation of A to A2 should therefore be
best viewed as a concerted [2 þ 2]-cycloaddition step with an
overall energy cost of 26.2 kcal/mol. Interestingly, A2 is almost
The solid line depicted in Fig.
1 represents the most
energetically accessible pathway for 1,2-CF addition of C6F6 across
the TihCtBu ligand of A. Transient A is 4.6 kcal/mol higher in
energy than (PNP)Ti]CHtBu(CH2t Bu) and coordination of C6F6 to
(PNP)TihCtBu can be accomplished at an energy cost of 14.1 kcal/
isoenergetic with A, but the
b-fluoride elimination step is
mol to form the putative adduct (PNP)TihCtBu(
s
-C6F6) (A1)
extremely facile with a barrier of only 4.3 kcal/mol. In this pathway
we cannot estimate the energy for A10, but we propose the cyclo-
addition barrier to be the slowest step along the CeF bond-
breaking event since the energy to traverse from A to A2 is at least
26.2 kcal/mol. Although the DDGz between 1,2-CF bond addition
(overcoming a simulated barrier, labeled with a star *, which we
were not able to compute with standard methods and is therefore
estimated), which then undergoes a virtually barrierless 1,2-CF
addition step with a penalty of w3.7 kcal/mol to produce the syn
isomer of 1. Although the overall CeF activation barrier is 22.4 kcal/
and [2 þ 2]-cycloaddition/
b-fluoride elimination is relatively small
mol, this value is considerably lower than the
a
-abstraction to form
(3.8 kcal/mol), we propose that 1,2 addition is the more facile
pathway since the energy difference between A10-TS, which we
were able to locate, and A1 is substantial at 7.5 kcal/mol.
A (27.8 kcal/mol). It is not obvious whether the binding event or the
activation of the CeF bond would be more difficult. However, given
that A-TS is much higher in energy than A1 by 9.1 kcal/mol we are
As illustrated in Fig. 1, the C6F6 substrate binds to the Ti(IV)
center in an electrostatic fashion using F as the contact atom.
confident to state that
a-abstraction to form A is overall rate-
Fig. 1. Two distinct CeF bond activation pathways of the substrate C6F6 via transient A.

4142
J.G. Andino et al. / Journal of Organometallic Chemistry 696 (2011) 4138e4146
Table 1
Table 1 summarizes the substantial distortion of the electrostatic
potential (ESP) that the adduct formation affords in terms of the
ESP-fit charge variations. Fig. 2 depicts the computed structures of
the transition state structure A1-TS. Structurally, A1-TS can be
considered early as the TihC2 and C1eF bonds are only slightly
elongated from that of A and free C6F6, respectively. Fig. 2 also
illustrates the proposed structure for the [2 þ 2]-cycloaddition
ESP charge variations for the 1,2-CF activation pathway involving the substrate C6F6.
Ti
0.91
1.16
F1
0.02
C1
C2
C6F6
0.10
Adduct (A1)
TS (A1-TS)
ꢀ0.12
ꢀ0.03
þ0.09
ꢀ1.25
ꢀ1.03
þ0.22
ꢀ0.14
ꢀ0.17
D
q
þ0.25
ꢀ0.12
ꢀ0.27
Fig. 2. Computed transition state in 1,2-CeF addition pathway leading to 1-syn (A1-TS) and computed structure of the intermediate A2 involved in the cycloaddition pathway
leading to CeF activation. Distances are reported in Å. For A1-TS: TieF1, 2.14; TieC2, 1.76; TieP1, 2.64; TieP2, 2.63; TieN, 2.06; F1eC1, 1.46; C1eC2, 2.56. For A2: TieF1, 2.43; TieC2,
1.75; TieP1, 2.62; TieP2, 2.59; TieN, 2.09.
Fig. 3. Molecular models of the syn and anti isomers of (PNP)Ti][CtBu(C6F5)](F) (1) and (PNP)Ti][CtBu(C7F7)](F) (2). Hydrogen atoms have been omitted for clarity.

J.G. Andino et al. / Journal of Organometallic Chemistry 696 (2011) 4138e4146
4143
Table 2
Comparison of selected bond distances (Å) and angles (ꢁ) in DFT-optimized and crystal structures of 2-syn and 2-anti.
Bond/angle
1-syn
1-anti
2-syn
2-anti
DFT
X-ray
DFT
DFT
X-ray
DFT
X-ray
2.5610 (5)
2.6251 (5)
2.0491 (14)
1.8166 (10)
1.9510 (17)
1.493 (2)
Ti1eP1
Ti1eP2
Ti1eN1
Ti1eF1
Ti1eC1
C1eC2
2.620
2.658
2.065
1.802
1.919
2.6049 (7)
2.6050 (7)
2.0566 (17)
1.8211 (12)
1.946 (2)
2.630
2.662
2.055
1.800
1.920
2.623
2.658
2.061
1.800
1.922
2.5986 (6)
2.6181 (5)
2.0517 (14)
1.8287 (10)
1.9487 (18)
1.496 (2)
2.624
2.657
2.053
1.800
1.925
1.500
1.503 (3)
1.497
1.495
1.492
C1eC21
1.547
1.541 (3)
1.551
1.549
1.544(2)
1.554
1.536 (2)
P1eTi1eP2
N1eTi1eF1
F1eTi1eC1
C2eC1eC3
N1eTi1eC1
F1eTi1eC1eC2
145.45
128.60
109.34
115.71
121.92
ꢀ5.15
144.80 (2)
133.27 (6)
106.89 (7)
114.30 (17)
119.73 (8)
5.43 (16)
143.10
130.37
113.11
115.32
116.62
ꢀ175.67
145.19
129.20
108.42
115.61
122.21
ꢀ8.08
145.294 (19)
132.88 (5)
105.64 (6)
115.97 (14)
121.28 (7)
ꢀ6.03 (13)
142.76
132.46
112.54
115.32
114.99
ꢀ173.75
142.304 (18)
137.48 (5)
111.89 (6)
114.39 (14)
110.63 (6)
ꢀ174.89 (11)
product for C6F6, A2, where the carbon C1 has been hybridized to
sp3, while the bond order calculations suggest the C1eF bond of
this intermediate to be much weaker relative to the value
computed for the free substrate, C6F6. As noted previously, we were
of 2 and 1-syn are comparable which also places them really close
in energy at
E(SCF) ¼ 2.0 kcal/mol. For structural comparison we
D
focused our attention on complex 2 since single crystals suitable for
X-ray diffraction studies could be grown for both isomeric forms
(unfortunately only the syn isomer of 1 could be structurally
confirmed) [2]. The computed structures of the syn isomers of 1 and
2 are quite similar. In complex 2-syn, the molecular structure
reveals the Ti]C distance (Ti1eC1) of 1.9487(18) Å and elongates
slightly to 1.9510(17) Å for 2-anti (Table 1), while in 2-syn, the Ti]C
distance in 1-syn is 1.946(2) Å. For 1, the alkylidene isomers have
a dihedral angle between F1eTi1eC1eC2 of 0ꢁ in 1-syn and
changes to 178.87ꢁ in 1-anti. The bond between Ti1eF1 decreases
negligibly from the syn to the anti isomer (1.806 Å to 1.799 Å,
respectively).
unable to locate a minimum for the structure of a p-bound adduct
of A with C6F6 (A10) which ultimately undergoes a [2 þ 2]-cyclo-
addition to form A2 followed by
1-syn.
b-fluoride elimination to generate
To better understand the electronic structure of 1 and investi-
gate the mechanism of rotation of the alkylidene ligands from syn
to anti, it is instructive to first compare the DFT-optimized struc-
tures of alkylidene complexes 1 and 2 in their isomeric forms syn
and anti, where syn is the isomer where the C6F5 fragment is
oriented to the same side as the fluoride ligand. The computed
structures are shown in Fig. 3. The computed isomeric forms of 2
could be compared to the corresponding crystallographic struc-
tures of 2-syn and 2-anti which have been published elsewhere [2].
Both systems were found to be in good agreement with the
metrical parameters obtained from their crystal structures
(Table 2). As anticipated, bond lengths and angles for both isomers
An experimental study of the isomerization process was per-
formed using complex 1, since the syn isomer could be more
conveniently separated from the reaction mixture, therefore
allowing us to enrich a solution with approximately 85/15 mixture
of syn and anti isomers, respectively [2]. After refluxing this sample
for 1 h in benzene at 105 ꢁC the ratio converted to 65/35 (syn/anti).
Fig. 4. HOMO of 1-syn and 1-anti with isodensity value of 0.05 au and schematic representation of the participating orbitals.

4144
J.G. Andino et al. / Journal of Organometallic Chemistry 696 (2011) 4138e4146
Fig. 5. Plot of the electronic energy versus rotation about the titaniumealkylidene ligand of 1 (top) and schematic representation of the mechanism for rotation of the alkylidene
ligand from 1-syn at 0ꢁ to 1-anti at 180ꢁ (phenylene backbone of the pincer ligand and hydrogen atoms have been eliminated for clarity).
The formation of the two isomers (syn/anti) is manifested by two
distinct alkylidene resonances in the 13C NMR spectrum (323.5 and
317.0 ppm). Likewise, the 31P NMR spectrum confirms two C1
symmetric species in solution with an average JPF value of 20 Hz.
From our theoretical study, the calculated difference in electronic
energy between 1-syn and 1-anti isomers is estimated to be
2.5 kcal/mol with 1-syn being slightly favored thermodynamically.
To understand the mechanism of rotation of the alkylidene
ligand in 1 and 2, an orbital analysis of the most important orbitals
was conducted. The MO analysis of the frontier orbitals reveals not
surprisingly that the HOMO of both 1-syn and 1-anti is involved in
the multiple bonding between Ti]C, and can be assigned to the in-
26 kcal/mol is lower than intuitively expected. A video of the
rotation from syn to anti isomerization is available in the support-
ing information.
Intuitively, the twisting motion that the Mealkylidene moiety
must undergo to convert 1-syn to 1-anti is expected to be associated
with a higher barrier, because the Mealkylidene double bond must
break at the transition state. This process can be envisioned to take
place either in a heterolytic or homolytic fashion. The heterolytic
Ti]C bond breaking may traverse an electronic structure that
formally resembles a TiIVeC1ꢀ fragment where the two electrons
from the Ti]C
p-orbital are localized on the carbon atom. The
alternative charge polarization affording a TiIIeCþ moiety where
the two electrons are placed on Ti is less appealing, as it would
formally constitute a reductive elimination process that is difficult
phase
p interaction between dxy and the corresponding p orbital of
the alkylidene carbon (Fig. 4). Only minor changes are observed on
common structural metric parameters of both isomers (Table 2). A
scan of the potential energy surface associated with the rotational
motion to transform 1-syn to 1-anti reveals that the electronic
energy of the complex reaches a maximum at 26 kcal/mol when the
dihedral angle between F1eTi1eC1eC2 is 90ꢁ (Fig. 5). Whereas the
position of the maximum is not surprising its relative energy of
to envision. In either case, the two
spatial orbital and the overall spin state will be a closed-shell
singlet throughout the twisting motion. The homolytic -bond
p-electrons remain in the same
p
cleavage, on the other hand, must invoke an open-shell singlet
involving an electronic structure that formally contains a TiIIIeCꢂ
Table 4
Changes in Mayer bond order of the Ti]C bond distance upon change in dihedral
angle F1eTi1eC1eCArF in complex 1.
Table 3
Comparison of selected bond lengths (Å) of DFT-models syn, anti and 90ꢁ-rotated
isomers of 1.
F1eTi1eC1eC (ꢁ)
Mayer bond order
R(Ti1eC1) (Å)
0 (syn)
30
60
90
120
1.50
1.51
1.51
1.51
1.50
1.50
1.51
1.913
1.913
1.919
1.886
1.895
1.909
1.918
Bond
1-syn
90ꢁ-rotated
1-anti
Ti1eN1
Ti1eF1
Ti1eP1
Ti1eP2
Ti1eC1
2.064
1.802
2.623
2.655
1.918
2.099
1.823
2.663
2.726
1.886
2.049
1.800
2.630
2.662
1.920
150
180 (anti)

J.G. Andino et al. / Journal of Organometallic Chemistry 696 (2011) 4138e4146
4145
Fig. 6. HOMO (ꢀ4.208 eV) of model rotated 90ꢁ around the Ti]C bond of 1-syn with isodensity value of 0.05 au.
moiety with the two unpaired electrons adopting opposite spins.
Structurally, we expect the Ti/C1 distance to increase at the
transition state reflecting on the loss of double bond character.
Surprisingly, our calculations indicate that the Ti/C1 decreases
upon twisting from 1.918 to 1.886 Å (Table 3). This is a tantalizing
result, as we typically correlate M/L distances with bond order
and, thus, we may speculate that the double bond character
between the Ti/C1 fragments is maintained throughout the
twisting motion, which makes little intuitive sense. One potential
reason for such structural anomaly may be that the Ti/C1 distance
is not governed by electronics, but by sterics, that is the double
bond is formally broken but the metal-ligand distance is enforced
by different factors. To test this plausible speculation, we calculated
the bond order between titanium and C1 as we varied the twist
Fig. 7. Molecular orbital redistribution upon 90ꢁ rotation of alkylidene ligand in 1-syn.

4146
J.G. Andino et al. / Journal of Organometallic Chemistry 696 (2011) 4138e4146
angle, as summarized in Table 4. Throughout the series, the bond
order remains identical at 1.5 indicating that the invariance of the
Ti/C1 distance arises from the electronic structure of the complex.
A careful inspection of the molecular orbitals reveals finally that
Acknowledgments
We dedicate this work to Professor Kenneth Caulton for the
occasion of his 70th birthday. Ken is a terrific colleague and a friend
who has advised and inspired us as scientists over the last decade at
Indiana University-Bloomington. We also thank the National
Science Foundation (CHE-0848248 & CHE-0645381) for support of
this research.
the TieC p-bonding orbital illustrated in Fig. 4 is indeed maintained
at the transition state. Fig. 6 shows the HOMO of the transition state
where a twist angle of 90ꢁ is adopted. Fig. 7 compares the MO
diagrams of 1-syn and the 90-deg rotated transition state. In the
transition state structure severe distortion of the PNP ligand
backbone characterized by an N1eTieC1 angle of 121.92ꢁ allows for
a different metal-d orbital to serve as the
p-acceptor. In 1-syn the
Appendix. Supplementary data
p-accepting orbital is dxy, as sketched in Fig. 4, whereas it is the dyz
orbital in the transition state geometry that engages and maintains
the Ti]C double bond character as illustrated in Fig. 6. This elec-
tronic rearrangement is moderated by the amide nitrogen that is
not trans to C1, but instead distorts to a cis disposition. This elec-
tronic flexibility is possible, as all d-orbitals of the Ti(IV)-d0 center
can act as electron acceptors e a flexibility that the organic
analogues do not possess. To allow for proper aligning of the
Supplementary data related to this article can be found online at
doi:10.1016/j.jorganchem.2011.07.037.
References
[1] (a) B.E. Smart, in: S. Patai, Z. Rappoport (Eds.), The Chemistry of Functional
Groups, Supplement D, John Wiley & Sons, New York, 1983 (Chapter 14);
(b) J.L. Kiplinger, T.G. Richmond, C.E. Osterberg, Chem. Rev. 94 (1994) 373;
(c) J. Burdeniuc, R.H. Crabtree, Chem. Ber./Recueil 130 (1997) 145;
(d) W.D. .Jones, Dalton Trans. (2003) 3991;
(e) T.G. Richmond, in: S. Murai (Ed.), Topics in Organometallic Chemistry,
vol. 3, Springer, New york, 1999, p. 243.
[2] B.C. Bailey, J.C. Huffman, D.J. Mindiola, J. Am. Chem. Soc. 129 (2007) 5302.
[3] J.A. Flores, V.N. Cavaliere, D. Buck, G. Chen, B. Pinter, M.G. Crestani, M.-H. Baik,
D.J. Mindiola, Chem, Science 2 (2011) 1457.
appropriate d-p orbital, the observing ligands on Ti must be suffi-
ciently flexible and the PNP backbone fulfills this requirement
perfectly. This structural flexibility is the direct result of the soft-
ness of the P-atoms. Since Ti(IV) is a hard acid, the TiePNP moiety
constitutes a soft/hard mismatched construct. As a result, the TieP
bonds are relatively weak and flexible, resulting in a relatively small
energy penalty for structural distortions.
[4] R.R. Schrock, H.H. Fox, M.H. Schofield, Organometallics 13 (1994) 2804.
[5] R.R. Schrock, J.H. Oskam, J. Am. Chem. 114 (1992) 7588.
[6] (a) R.R. Schrock, Polyhedron 14 (1995) 3177 Perfluoroalkylidene isomers of
iridium have been reported and recently observed;
4. Conclusions
In this study we have examined the mechanism to activation of
a CeF bond in a perfluoroarene promoted by a transient titanium
alkylidyne. Based on the reaction profile we propose the activation
of a perfluoroarene to proceed via 1,2-CF bond addition across the
Ti^C bond. We have also investigated the mechanism for the
rotation about the Ti]C bond which gives rise to two isomers, syn
and anti. The syn isomer was found to be slightly lower in energy.
Surprisingly, we found that the rotation around the Ti]C bond can
(b) J. Yuan, R.P. Hughes, J.A. Golen, A.L. Rheingold, Organometallics 29 (2010)
29 1942;
(c) R.P. Hughes, personal communication.
[7] A. Fürstner, L. Ackermann, B. Gabor, R. Goddard, C.W. Lehmann, R. Mynott,
F. Stelzer, O.R. Thiel, Chem. Eur. J. 7 (2001) 3236.
[8] A. Fürstner, H. Krause, L. Ackermann, C.W. Lehmann, Chem. Commun. (2001)
2240.
[9] R.H. Grubbs, C.W. Bielawski, D. Benitez, Science 297 (2002) 2041.
[10] R.H. Grubbs, C.W. Bielawski, D. Benitez, J. Am. Chem. Soc. 125 (2003) 8424.
[11] K.S. Lokare, R.J. Staples, A.L. Odom, Organometallics 27 (2008) 5130.
[12] B.C. Bailey, H. Fan, E.W. Baum, J.C. Huffman, M.-H. Baik, D.J. Mindiola, J. Am.
Chem. Soc. 127 (2005) 16016.
[13] B.C. Bailey, H. Fan, J.C. Huffman, M.-H. Baik, D.J. Mindiola, J. Am. Chem. Soc.
129 (2007) 8781.
[14] A.R. Fout, J.L. Scott, D.L. Miller, B.C. Bailey, J.C. Huffman, M. Pink, D.J. Mindiola,
Organometallics 28 (2009) 331.
[15] A.R. Fout, B.C. Bailey, D. Buck, H. Fan, J.C. Huffman, M.-H. Baik, D.J. Mindiola,
Organometallics 29 (2010) 5409.
[16] B.C. Bailey, H. Fan, J.C. Huffman, M.-H. Baik, D.J. Mindiola, J. Am. Chem. Soc.
128 (2006) 6798.
be accommodated without breaking the
p
-bond. Even at the
transition state, where the twist angle is 90ꢁ, a full
p-bond is
maintained. This non-intuitive electronic structure is possible,
because the Ti(IV)-d0 can utilize both the dxz and dyz orbitals as the
p-accepting orbital. This electronic feature is reminiscent of the
situation found in organic acetylene moieties, where rapid rotation
is also possible due to the presence of two orthogonal
this case, the alkylidene fragment that serves as the
p-orbitals. In
p-donor only
[17] A.R. Fout, B.C. Bailey, H. Fan, J.C. Huffman, M.-H. Baik, D.J. Mindiola, J. Am.
Chem. Soc. 129 (2007) 12640.
[18] A.B. Pangborn, M.A. Giardello, R.H. Grubbs, R.K. Rosen, F.J. Timmers, Organo-
metallics 15 (1996) 1518.
[19] Jaguar, Version 7.0. Schrödinger, L.L.C, New York, NY, 2007.
[20] A.D. Becke, Phys. Rev., A 38 (1988) 3098.
[21] C. Lee, W. Yang, R.G. Parr, Phys. Rev. B 37 (1988) 785.
[22] A.D. Becke, J. Chem. Phys. 98 (1993) 5648.
[23] E.M. Siegbahn, R.A. Blomberg, Chem. Rev. 100 (2000) 421.
[24] P.J. Hay, W.R. Wadt, J. Chem. Phys. 82 (1985) 270.
[25] W.R. Wadt, P.J. Hay, J. Chem. Phys. 82 (1985) 284.
[26] T.H. Dunning, J. Chem. Phys. 90 (1989) 1007.
possesses one
p-orbital, but the p-accepting fragment has two
empty d- orbitals that can accommodate the Ti]C twisting. Key
p
to this electronic flexibility is the ease of structural distortion of the
PNP backbone that must accommodate a varying ligand environ-
ment that in turn affects the metal-d orbital alignment. Although
the bulky substituents on both, PNP and alkylidene ligands at first
appear to render this rotation an unfavorable process given the
severe distortions observed on the chelate’s binding to Ti, we
conclude that this parameter does not represent a major influence
to the energetic penalty.
[27] R.G. Bergman, F.E. Michael, H.M. Hoyt, J. Am. Chem. Soc. 126 (2004) 1018.