Titanium(IV) trifluoromethyl complexes: New perspectives on bonding from organometallic fluorocarbon chemistry

Article abstract of DOI:10.1021/om201055e

Trifluoromethyltrimethylsilane, (CH₃)₃SiCF ₃, in the presence of CsF serves as an excellent CF₃ group-transfer reagent, and reaction with Cp₂TiF₂ in THF gives the titanocene trifluoromethyl fluoride complex Cp₂Ti(CF ₃)(F) (1; Cp = C₅H₅) in 60% isolated yield. Reaction of complex 1 with the trimethylsilyl reagents, (CH₃) ₃SiX (X = OTf = OSO₂CF₃, Cl, Br, I, N ₃, and OSO₂Ph), in a tetrahydrofuran or toluene solution affords the corresponding Ti-CF₃ derivatives Cp₂Ti(CF ₃)(X) (X = OTf (2), Cl (12), Br (13), I (14), N₃ (15), and OSO₂Ph (16)) in good isolated yields of 67-84%. These compounds have been characterized by a combination of reactivity studies, IR and ¹H/¹³C{¹H}/¹⁹F NMR spectroscopies, and single-crystal X-ray diffraction. The Ti-CF₃ linkage in these complexes is remarkably robust, and although the α-C-F bonds are elongated, there is no evidence of an α-fluoride (Ti???F- CF₂) between the electrophilic Ti(IV) metal center and any of the C-F bonds in the trifluoromethyl group in the solid-state or in solution. In the solid-state, these complexes are shock-sensitive; energetic decomposition of Cp₂Ti(CF₃)(F) (1) produces uniform spherical nanoparticles ranging from ～70 to 120 nm in size and porous fluorinated oligomers and polymers containing both -(CF₂-CF₂)- and -(CF ₂-CFH)- units, as determined by a combination of SEM, XRD, XRF, XPS, and ¹⁹F MAS NMR. Density functional theory results show good agreement with experimental structural data obtained for Cp ₂Ti(CF₃)(X) (X = F (1), OTf (2), Cl (12), N₃ (15)) and accurately predicts longer Ti-CF₃ distances than for each specific CH₃ analogue, and the trend extends to structurally related Zr and Hf analogues. Simpler model compounds from groups 4 and 8 (M(CH ₃)₄, M(CH₃)₃(CF₃), M(CH₃)₃(CCl₃), and M(CH₃) ₃(CF₂CF₂CF₂CF₃); M = Ti, Zr, Hf, Fe, Ru, Os)) were also examined and show that, for group 4 complexes, π-bonding is a significant factor in shortening the strongly ionic M-CH ₃ relative to M-CF₃, whereas for the predominantly covalent group 8 analogues, π-back-bonding helps to shorten the predominantly covalent M-CF₃ relative to M-CH₃. The bonding analysis suggests that the significant elongation of C-F bonds α to metals is mainly a consequence of the electropositivity of the group 4 metal centers, with minor, if any, contributions from π-effects; the bond-lengthening effect is most pronounced at the α-position and decays rapidly on moving away from the metal.

Full text of DOI:10.1021/om201055e

Article
pubs.acs.org/Organometallics
Titanium(IV) Trifluoromethyl Complexes: New Perspectives on
Bonding from Organometallic Fluorocarbon Chemistry
Felicia L. Taw,† Aurora E. Clark,† Alexander H. Mueller,† Michael T. Janicke,† Thibault Cantat,†
Brian L. Scott,† P. Jeffrey Hay,† Russell P. Hughes,*,‡ and Jaqueline L. Kiplinger*,†
†Los Alamos National Laboratory, Los Alamos, New Mexico 87545, United States
‡Dartmouth College, 6128 Burke Laboratories, Hanover, New Hampshire 03755, United States
S
* Supporting Information
ABSTRACT: Trifluoromethyltrimethylsilane, (CH3)3SiCF3, in the presence of CsF serves
as an excellent CF3 group-transfer reagent, and reaction with Cp2TiF2 in THF gives the
titanocene trifluoromethyl fluoride complex Cp2Ti(CF3)(F) (1; Cp = C5H5) in 60%
isolated yield. Reaction of complex 1 with the trimethylsilyl reagents, (CH3)3SiX (X =
OTf = OSO2CF3, Cl, Br, I, N3, and OSO2Ph), in a tetrahydrofuran or toluene solution
affords the corresponding Ti−CF3 derivatives Cp2Ti(CF3)(X) (X = OTf (2), Cl (12), Br
(13), I (14), N3 (15), and OSO2Ph (16)) in good isolated yields of 67−84%. These
compounds have been characterized by a combination of reactivity studies, IR and
1H/13C{1H}/19F NMR spectroscopies, and single-crystal X-ray diffraction. The Ti−CF3
linkage in these complexes is remarkably robust, and although the α-C−F bonds are
elongated, there is no evidence of an α-fluoride (Ti···F−CF2) between the electrophilic Ti(IV) metal center and any of the C−F
bonds in the trifluoromethyl group in the solid-state or in solution. In the solid-state, these complexes are shock-sensitive;
energetic decomposition of Cp2Ti(CF3)(F) (1) produces uniform spherical nanoparticles ranging from ∼70 to 120 nm in size
and porous fluorinated oligomers and polymers containing both −(CF2−CF2)− and −(CF2−CFH)− units, as determined by a
combination of SEM, XRD, XRF, XPS, and 19F MAS NMR. Density functional theory results show good agreement with
experimental structural data obtained for Cp2Ti(CF3)(X) (X = F (1), OTf (2), Cl (12), N3 (15)) and accurately predicts longer
Ti−CF3 distances than for each specific CH3 analogue, and the trend extends to structurally related Zr and Hf analogues. Simpler
model compounds from groups 4 and 8 (M(CH3)4, M(CH3)3(CF3), M(CH3)3(CCl3), and M(CH3)3(CF2CF2CF2CF3); M = Ti,
Zr, Hf, Fe, Ru, Os)) were also examined and show that, for group 4 complexes, π-bonding is a significant factor in shortening the
strongly ionic M−CH3 relative to M−CF3, whereas for the predominantly covalent group 8 analogues, π-back-bonding helps to
shorten the predominantly covalent M−CF3 relative to M−CH3. The bonding analysis suggests that the significant elongation
of C−F bonds α to metals is mainly a consequence of the electropositivity of the group 4 metal centers, with minor, if any,
contributions from π-effects; the bond-lengthening effect is most pronounced at the α-position and decays rapidly on moving
away from the metal.
decarbonylation of the parent trifluoroacetyl complex.4 Other
trifluoromethyl complexes have been prepared in an analogous
INTRODUCTION
Continued interest in fluorocarbon chemistry stems from the
■
unique properties that fluorinated materials exhibit.1 The high
manner, but this technique requires the presence of a tri-
fluoroacetyl ligand on a metal center that possesses a vacant
electronegativity of fluorine and great strength of C−F bonds
coordination site to which the CF3 group can migrate. More-
over, thermolysis or photolysis is often necessary to promote
decarbonylation.5 A second technique was introduced that
involved oxidative addition of trifluoromethyl iodide to a low-
valent transition metal; however, this method was typically
successful only for d8 or d10 metal substrates.5,6 A third route to
perfluoroalkyl complexes consisted of metal atom reactions and
radical chemistry, in which either the metal or the CF3 fragment
was activated by an external high-energy source.5,7 More recently,
the group 12 cadmium and mercury systems, Cd(CF3)2·2L
conspire to give compounds that have low chemical and thermal
reactivity, as compared with hydrocarbon analogues.2 As such,
it is inevitable that these same characteristics of fluorocarbons
present challenges to the preparation of fluorinated complexes.
Specifically, although a vast number of transition-metal catalysts
have been proven to affect the polymerization of hydrocarbon
olefins,3 similar catalytic platforms do not exist for the poly-
merization of fluorinated olefins. As such, the discovery of new
transition-metal perfluoroalkyl complexes and developing an
understanding of their unusual behavior are crucial steps in this
direction.
In contrast to the arsenal of choices that are available for
alkylation, only a handful of pathways exist to functionalize transi-
tion metals with perfluoroalkyl groups. The first transition-metal
CF3 complex, Mn(CF3)(CO)5, was synthesized by thermal
Special Issue: Fluorine in Organometallic Chemistry
Received: October 31, 2011
Published: January 27, 2012
© 2012 American Chemical Society
1484
dx.doi.org/10.1021/om201055e | Organometallics 2012, 31, 1484−1499

Organometallics
Article
(L = pyridine or 1,2-dimethoxyethane) and Hg(CF3)2, have been
employed as CF3-transfer reagents for a variety of transition-
metal and main-group compounds.5,8 These reagents demon-
strated high versatility and allowed for the preparation of many
previously unknown transition-metal trifluoromethyl complexes.
The techniques described above have allowed entry into the
synthesis of M−CF3 complexes for virtually all the middle- and
late-row transition metals. However, these same methods have
failed for the early transition metals (groups 3, 4, and 5). An
early report described the treatment of Ti(CH3)4 with
trifluoromethyl iodide, which led to formation of Ti(CH3)3I.9
No Ti−CF3 species or any fluorine-containing compound
could be detected. Researchers have also reported attempts to
synthesize Zr−CF3 complexes using the aforementioned
cadmium and mercury CF3 reagents, but only zirconium
fluorides were isolated as the end products.10
Several years ago, we demonstrated that stable early transition-
metal perfluoroalkyl could be prepared with the synthesis and
characterization of the titanium perfluoroalkyl complexes,
Cp2Ti(CF3)(F) (1; Cp = C5H5) and Cp2Ti(CF3)(OTf)
(2; OTf = OSO2CF3).11 Titanocene trifluoromethyl fluoride (1) was
accessed by the addition of Ruppert’s reagent, (CH3)3SiCF3,
to a tetrahydrofuran (THF) suspension of Cp2TiF2,12 and the
titanocene trifluoromethyl triflate complex (2) was synthesized
by treating 1 with the trimethylsilyl compound, (CH3)3SiOTf
(eq 1).11
complex (11) has also been formed from reaction of the
corresponding Ce−H and (CH3)3SiCF3.21 Subsequent
α-fluoride abstraction of a fluoride by the cerium metal center
gives the cerium fluoride complex (1,2,4-tBu3-C5H2)2Ce-F and
difluorocarbene.
The Ti−CF3 complexes 1 and 2 were characterized by X-ray
diffraction studies, which revealed that the Ti−C bond
distances (2.221(3) Å for 1; 2.222(5) Å for 2)11 in the Ti−
CF3 moiety were substantially longer than Ti−C bond
distances (1.988−2.181 Å)22 reported for structurally related
Ti−CH3 complexes. This directly contradicted existing tenets
regarding M−C (M = middle- or late-row transition metal)
bond length trends in M−CF3 complexes versus analogous
M−CH3 complexes, in which a shorter M−C bond length is
expected for the M−CF3 compounds.5 Indeed, structural com-
parisons of previously reported middle- and late-row transition-
metal perfluoroalkyl and corresponding alkyl compounds
revealed a significant contraction (by ∼0.05 Å)5f of the
metal−perfluoroalkyl bond length relative to the metal−alkyl
bond length.5 For example, the M−CF3 bond distance in
23
(C5Me5)Mo(CF3)(CO)3 was reported to be 2.248(5) Å,
whereas in the related molybdenum complex, [η5-C5H4C(
O)CH3]Mo(CH3)(CO)3,24 the M−CH3 bond distance was
reported to be 2.303(5) Å. In the titanium case, an unexpected
elongation of the metal−perfluoroalkyl (M−C) bond length
was observed. Herein, we describe the synthesis, structural
characterization, properties, and reactivity of a series of Ti−CF3
complexes and present density functional theory (DFT)
calculations to investigate the bonding and electronic structure
in these Ti−CF3 systems that gives rise to the unprecedented
bond length trend in comparison to mid- and late-row transition-
metal systems.
RESULTS AND DISCUSSION
■
Although widely used to trifluoromethylate organic com-
pounds,13 Ruppert’s reagent has also been used in a few cases
to transfer CF3 to late transition metals (Chart 1).14 For example,
Synthesis, Structural Characterization, and Reactivity.
As shown in eq 2, the synthesis of titanium trifluoromethyl com-
Chart 1. Examples of Metal Trifluoromethyl Complexes
Prepared Using the CF3 Delivery Agent (CH3)3SiCF3
plexes was achieved by the addition of trimethylsilyl reagents,
(CH3)3Si-X (X = OTf, Cl, Br, I, N3, and OSO2Ph), to a THF or
toluene solution of Cp2Ti(CF3)(F) (1) to form Cp2Ti(CF3)-
(X) (X = OTf (2), Cl (12), Br (13), I (14), N3 (15), and
OSO2Ph (16)). In all cases, the reactions were performed at
room temperature, and release of (CH3)3SiF was observed by
1H and 19F NMR spectroscopy as the reaction progressed. Com-
plex 12 may be prepared in high yield (85−90% isolated) by
addition of 5 equiv of (CH3)3SiCl to a THF solution of 1. After
stirring at ambient temperature for 3 h, removal of volatile
materials under reduced pressure gave an orange powder. Trace
amounts of 1 were present in this crude product, which can be
purified by crystallization from a toluene/hexanes solution at
−30 °C. Alternatively, 5 equiv of (CH3)3SiCl may be added to
Caulton and co-workers showed that addition of (CH3)3SiCF3 to
a ruthenium fluoride complex led to replacement of the fluoride
ligand by CF3, giving 3.14a,b Subsequent α-migration of the
fluoride from CF3 to the metal center resulted in formation of
the final product, a ruthenium difluorocarbene fluoride complex
(4). Since our initial communication, additional reports have
shown that Ruppert’s reagent is quite effective as a CF3 delivery
agent and gives a variety of stable first-row (Ni, 5;15 Cu, 616),
second-row (Rh, 7;17 Pd, 818), and third-row (Pt, 9;19 Au,
1020) transition-metal complexes. A reactive cerium trifluoromethyl
1485
dx.doi.org/10.1021/om201055e | Organometallics 2012, 31, 1484−1499

Organometallics
Article
a
1
Table 1. Selected H, 19F, and 13C{1H} NMR and IR Spectroscopic Data for Complexes 1, 2, and 12−16
Cp2Ti(CF3)X
1H [δ C5H5]
19F [δ Ti−CF3]
13C{1H} [δ CF3]
13C{1H} [δ C5H5]
ν
ν
Ti−CF3
692
689
691
690
C−Fsymm
1081
1065
1077
1074
1
X = F (1)
6.40
6.43
6.52
6.58
6.63
6.72
6.76
−24.0
−23.2
−21.2
−20.6
−24.5
−20.5
−25.6
153.1 (q, JC−F = 382.6 Hz)
117.4
116.8
118.6
118.6
119.3
118.3
120.7
1
X = N3 (15)
X = Cl (12)
153.9 (q, JC−F = 381.8 Hz)
1
153.2 (q, JC−F = 382.1 Hz)
1
X = Br (13)
155.2 (q, JC−F = 381.5 Hz)
1
X = OSO2Ph (16)
X = I (14)
152.4 (q, JC−F = 383.1 Hz)
1
159.3 (q, JC−F = 380.0 Hz)
689
694
1061
1082
1
X = OSO2CF3 (2)
151.8 (q, JC−F = 382.9 Hz)
a
Chemical shifts reported in parts per million; THF-d8 used as solvent. IR spectroscopic data are reported in cm−1.
a toluene solution of 1, allowed to stir for 3 h, and layered
directly with hexanes to produce orange crystals of 12 in
approximately 70% yield.
have been prepared by reaction of the corresponding metal−
trifluoromethyl complex with protic acids.5d However, addition
of protic acids (HCl, HOTf, or HBAr'4 (Ar' = 3,5-(CF3)2-C6H3)
or the methide-abstracting agent B(C6F5)3 to the Ti−CF3
complexes 1 and 2 led to immediate decomposition with the
formation of intractable products.
Other Ti−CF3 derivatives of 1 may be prepared in an anal-
ogous manner. For complexes 2, 13, 14, and 16, use of 1 equiv
of the appropriate trimethylsilyl reagent is sufficient to drive
the ligand substitution reaction to completion. The bromide-
substituted complex 13 was formed (>95% by NMR spectros-
copy) within 45 min at room temperature, whereas the triflate
(2), iodide (14), and benzenesulfonate (16) analogues were formed
within seconds under the same conditions. Synthesis of the azide
complex (13) required the addition of 10 equiv of (CH3)3SiN3 to a
THF solution of 1 and a reaction time of 2 h. Isolated yields for
compounds 2 and 13−16 varied between 67 and 84%.
1
Selected H, 19F, and 13C{1H} NMR spectroscopic data for
all Ti−CF3 complexes (1, 2, 12−16) are listed in Table 1. The
1H NMR resonances for the Cp ligands fall in the range of
δ 6.40−6.76 ppm. Within the series where X = halide, the Cp
resonances shift downfield when going from X = F→Cl→Br→I;
this is consistent with decreasing π donation from the halide to
the d0 Ti metal center [i.e., X(π)→Ti(dπ) donation], with the
fluoride ligand providing the greatest amount of donation to the
metal center. This effect has been measured in numerous systems
using NMR and IR spectroscopy, and the π-donor power of
the halides has been ranked in the order F > Cl > Br > I, which is
the opposite of what would be predicted based upon electro-
negativity arguments.26 The singlets that appear in the region of
δ −20 to −25 ppm in the 19F NMR spectra and the quartets with
Addition of other trimethylsilyl reagents (CH3)3Si-R (R =
CH3, C6H5, CO2Me, CCH, Si(CH3)3, CH2CN, NCO, and
1
C6F5) or Ph3SiH to 1 resulted in no reaction (by H and 19F
NMR spectroscopy), even upon heating the reaction mixture
to 60 °C for several days. Further, numerous attempts to pre-
pare the bis(trifluoromethyl) complex, Cp2Ti(CF3)2, by treating
Cp2TiF2 or 1 with (CH3)3SiCF3 or Zn(CF3)2·2(pyridine) were
also unsuccessful. This contrasts behavior of late metal iron(II)
and nickel(II) fluoride complexes, which display reaction chemistry
with these trimethylsilyl reagents to give complexes, such as
(LtBu)Fe-CCSiMe3 (LtBu = N,N′-bis(2,6-diisopropylphenyl)-
2,2-6,6-tetramethylheptane-3,5-diiminate)25 and most likely (dippe)-
Ni(CF3)2 (dippe = 1,2-bis(diisopropylphosphino)ethane).15
This marked difference in reactivity is likely a reflection of the
stronger Ti−F bond compared with the later transition metal−
fluoride bonds.
The Ti−CF3 complexes 1, 2, and 12−16 decompose to in-
tractable products at room temperature over the course of several
hours (1, 2, 12−16) to several days (12) and require storage
under cold conditions (−30 °C). In particular, complexes 1, 2,
and 12 are shock-sensitive in the solid state (vide infra) and
should be handled with great care. In solution, the stability of all
the Ti−CF3 compounds is enhanced by the addition of Lewis
bases, such as THF, pyridine, 4-(N,N-dimethylamino)pyridine,
P(OMe)3, or phosphines (PPh3, PMe3). Solutions of complexes
2, 14, and 16 require the presence of a Lewis base, such as
pyridine, to stabilize these species at room temperature and allow
characterization and reactivity studies. Because complex 1 is
stable, the reaction between Cp2Ti(CO)2 and CF4 was also
examined, but no reaction was observed at room temperature or
with heating at 50 °C even after 5 days.
1
large JC−F coupling constants (∼380−384 Hz) that appear in
the region of δ 150−160 ppm in the 13C{1H} NMR spectra are
diagnostic for the presence of CF3 groups. This is also consistent
with the 19F and 13C{1H} NMR resonances for CF3 groups in
other characterized middle- and late-row transition-metal
trifluoromethyl complexes.5
Solid-state IR studies of complexes 1, 2, and 12−16 showed
Ti−CCF stretching frequencies spanning 689−694 cm−1 and
3
C−F (of Ti−CF3) stretching frequencies in the range of 1061−
1082 cm−1 (Table 1). The observed C−F stretches for the
titanium compounds are characteristic of CF3 ligands bound
to middle- and late-row transition metals (e.g., (CO)5MnCF3,
νC−F = 1045, 1063 cm−1;27 Ru(CF3)Cl(CO)2(PPh3)2, νC−F = 1073,
1006 cm−1;28 Ir(CF3)(CO)2(PPh3)2, νC−F = 1088, 1005 cm−1;29
CpMo(CF3)(CO)3, νC−F = 1073, 1006 cm−1;30 CpFe(CF3)(CO)2,
νC−F = 1068, 1042, 1015 cm−1).30 Typically, these frequencies are
approximately 100 cm−1 lower than the mean C−F stretching
frequencies in CF3X (X = Cl, Br, I; νC−F ∼ 1140 cm−1),31 indicating
weaker C−F bonds in the metal complexes.27
X-ray diffraction studies confirmed the identity of the titanium
trifluoromethyl complexes. Representative structures are pro-
vided for complexes 12 and 15 in Figure 1, and selected geo-
metric parameters for complexes 1, 2, 12, and 15 are presented
in Table 2. The titanium trifluoromethyl complexes all exhibit a
bent-metallocene framework with a pseudotetrahedral coordina-
tion environment about the titanium metal center and the non-
Cp ligands lying within the metallocene wedge. For all structures,
no α-fluoride interaction (Ti···F−CF2) is observed between the
electrophilic titanium(IV) metal center and any of the C−F
bonds in the trifluoromethyl group. Variable-temperature NMR
spectroscopic studies (−80→+40 °C) on the complexes
No reaction was observed between any of the Ti−CF3 com-
plexes 1, 2, and 12−16 and nucleophilic reagents, such as
nitriles, isocyanides, amines, carbon monoxide, or benzophe-
none. Addition of olefins, fluorinated olefins, diphenyldiazo-
methane, and diphenylacetylene also resulted in no reaction. A
variety of middle- and late-row transition-metal dihalocarbenes
1486
dx.doi.org/10.1021/om201055e | Organometallics 2012, 31, 1484−1499

Organometallics
Article
Figure 1. Molecular structures of complexes 12 (left) and 15 (right) with thermal ellipsoids projected at the 25% probability level. Hydrogen atoms
have been omitted for clarity.
Table 2. Selected Metrical Parameters for Complexes 1, 2, 12, and 15
1 (X = F)
2 (X = OTf)
12 (X = Cl)
15 (X = N3)
Ti−X (Å)
Ti−CF3 (Å)
C−F (Å)
1.831(2)
2.221(3)
1.362(3)
1.365(4)
1.370(3)
89.35(9)
103.7(2)
102.3(2)
102.0(2)
114.0(2)
114.5(2)
118.4(2)
1.985(4)
2.222(5)
1.367(6)
1.362(4)
1.362(4)
88.3(2)
2.335(2)
2.301(6)
1.369(7)
1.306(6)
1.286(6)
89.27(13)
105.1(5)
104.7(5)
108.9(5)
110.2(4)
113.3(4)
114.0(4)
1.991(1)
2.239(1)
1.379(2)
1.382(2)
1.376(2)
89.59(5)
103.4(1)
102.5(1)
102.9(1)
113.3(1)
115.2(1)
117.7(1)
1.317(4)
(OTf)
(OTf)
(OTf)
1.314(4)
1.314(4)
C−Ti−X (deg)
F−C−F (deg)
103.3(3)
103.3(3)
102.8(4)
114.2(2)
114.2(2)
117.2(3)
Ti−C−F (deg)
corroborate the solid-state data and confirm the absence of
density from the Ti−CCF linkage, resulting in a corresponding
lengthening of the trifluoromethyl C−F bonds. The observed
3
α-fluoride interactions in solution as no change in the CF3 19F
1
NMR shift or JC−F was observed.
lengthening of both the Ti−CCF and the C−F bonds coupled
3
Comparisons for the metrical parameters of the Ti−CF3
moiety can only be made to known middle and late transition-
metal trifluoromethyl complexes. As in the case of complexes 1
(2.221(3) Å) and 2 (2.222(5) Å), the metal−carbon bond
distances in the Ti−CF3 moiety for 12 (2.301(6) Å) and 15
(2.239(1) Å) are longer than Ti−C bond distances (1.988−
2.181 Å) reported for structurally related Ti−CH3 complexes.22
For the trifluoromethyl group in complexes 1, 2, 12, and 15,
the average C−F bond length (1.357(4) Å) and the average F−
C−F bond angle (103.7(3)°) are quite similar to values found
in known middle- and late-row transition-metal CF3 complexes
that are thermally stable.5 Although direct quantitative com-
parisons are difficult, these observations are somewhat surprising
as the Ti−CF3 complexes display low thermal stability.
As complex 2 contains a triflate (OTf = OSO2CF3) group, it
provides a nice internal platform to compare the metrical
parameters of the C−F bonds on the M−CF3 group with those
on the coordinated triflate ligand. Examination of salient bond
lengths between the titanium complexes reveals statistically
relevant lengthening of the C−F bond distances for the
Ti−CF3 groups in 1, 2, 12, and 15 (C−Fave = 1.357(4) Å)
compared with the 1.314(4)−1.317(4) Å for the CF3 moiety in
the OTf ligand in complex 2. This suggests that the electro-
positive titanium(IV) metal center is siphoning away electron
with the IR data points to a weakening of these C−F bonds.
The fact that the d0 titanium metal center does not disassemble
the CF3 group is remarkable.
The Ti−F (1) and Ti−OOTf (2) bond distances of 1.831(2)
and 1.985(4) Å, respectively, are somewhat shorter than the
average Ti−F (1.848 Å) and Ti−OOTf (2.036 Å) bond dis-
tances reported for similar titanium(IV) terminal fluoride32 and
monotriflate33 complexes. Likewise, for the new structure 12
presented in this work, the Ti−Cl bond length (2.335(2) Å)
in complex 12 is slightly contracted when compared to other
Ti−Cl bond distances (2.346−2.379 Å) in similar titanocene
chloride complexes.34 A comparison of the Ti−N distances in
the titanium azide complex 15 versus the known bis(azide)
complex, Cp2Ti(N3)2 (17),35 also shows that the Ti−N dis-
tance (1.991(1) Å) in 15 is smaller than the Ti−N distance
(2.03(1) Å) reported for 17. This consistent bond shortening
across the series of Ti−X complexes (where X is the α-hetero-
atom F (1), O (2), Cl (12), or N (15)) is likely due to the
electron-withdrawing effect of the CF3 ligand, which
increases the electrophilicity of the Ti center. However,
put another way, this bond shortening also suggests Ti−X
multiple bond character in these complexes resulting from
intramolecular X→Ti π-donation, with consequent transfer
1487
dx.doi.org/10.1021/om201055e | Organometallics 2012, 31, 1484−1499

Organometallics
Article
Figure 2. SEM images of the two morphologies of the fluorinated material formed upon the energetic decomposition of Cp2Ti(CF3)(F) (1): (a)
spherical nanoparticles and (b) highly porous foam.
of electron density to the d0 titanium metal center to
stabilize the Ti−CF3 interaction.26b
−(CF2−CF2)− and −(CF2−CFH)− units (see Figures 3 and 4).
The XPS spectrum was scaled using the lowest binding energy
peak of carbon to be adventitious, with the higher binding energy
Finally, the N−N distances in the azide ligand of 15 are
1.219(2) and 1.142(2) Å for N(1)−N(2) and N(2)−N(3),
respectively. Corresponding N−N distances in the bis(azide)
complex (17) are slightly contracted (1.18(2) and 1.10(2) Å,
respectively) in comparison. Single-crystal X-ray diffraction
studies were also performed on the bromide (13) and iodide
(14) complexes, but these structures were plagued with signif-
icant disorder between the CF3 and halide ligands and only
allowed assignment of connectivity.36
Solid-State Stability of Titanocene Trifluoromethyl
Complexes. In the solid-state, complexes 1, 2, and 12 are
shock-sensitive and should be handled with great care in
amounts no larger than 0.150 g. Complex 1 is even shock-
sensitive in hexanes. The instability of the Ti−CF3 systems is
not entirely unexpected given the facile α- and β-fluoride
elimination pathways available to early metal fluoroalkyl
complexes to give high lattice energy metal fluorides. In fact,
such favorable decomposition pathways have been used to
explain previous failures to prepare perfluoroalkyl complexes
similar to 1, 2, and 12−16.
Energetic decomposition of Cp2Ti(CF3)(F) (1) was
accomplished by tapping the solid with a glass pipet tip (inside
a 20 mL scintillation vial). The resulting decomposition pro-
duces intractable solids. The volatiles were analyzed by 19F
NMR, which determined that no CF4 was produced in the
decomposition. The decomposition products were collected
on a glass slide using two different methods and viewed using
scanning electron microscopy (SEM), which showed the for-
mation of uniform spherical nanoparticles ranging from ∼70 to
120 nm in size (Figure 2a) and a highly porous foam (Figure 2b).
The spherical nanoparticles were obtained by collecting the volatile
products onto a slide mounted directly above the Cp2Ti(CF3)(F)
(1) as it energetically decomposed. The porous foam was generated
by allowing the decomposition of complex 1 to take place within
an opened scintillation vial, followed by pouring the resulting
vapor (decomposition products) onto a glass slide.
Figure 3. Wide-scan survey spectrum (XPS) of the porous foam
produced from the energetic decomposition of Cp2Ti(CF3)(F) (1)
detecting that the material contains fluorine and carbon.
X-ray fluorescence (XRF) measurements indicated that re-
sidual titanium was present in the decomposition products. The
absence of a diffraction pattern using powder X-ray diffraction
(XRD) revealed that the spheres are amorphous. The porous
foam was determined by both X-ray photoelectron spectros-
copy (XPS) and 19F magic-angle spinning (MAS) NMR to be a
mixture of fluorinated oligomers and polymers containing both
Figure 4. 19F magic-angle spinning (MAS) NMR spectra of the porous
foam produced from the energetic decomposition of Cp2Ti(CF3)(F) (1)
(top), polytetrafluoroethylene tape (middle), and a blank sample (bottom).
peaks being assigned to F-bonded carbon atoms. The F signal
also showed binding energy shifts consistent with fluorocarbon
1488
dx.doi.org/10.1021/om201055e | Organometallics 2012, 31, 1484−1499

Organometallics
Article
oligomers and polymers. Interestingly, a comparison of the 19F
MAS NMR spectrum obtained for the porous foam with that of
PTFE reveals that Teflon-like materials are also present in this
mixture. To the best of our knowledge, the energetic decomposi-
tion of a fluorinated organometallic complex represents a new
method for producing fluorinated oligomers and polymers.
negatively charged. When attached to electronegative (δ−)
elements, the bond to CH3 was shorter than that to CF3,
whereas for strongly δ+ elements, the reverse trend was
observed. Such arguments might also be applied to analogous
transition-metal compounds in which electronegativity in-
creases from left to right in a particular row and could explain
why bonds from CH3 and CF3 to the electropositive Ti described
here show the opposite trend compared with analogues to the
right in the periodic table.
Computational Study. Comparison of M−CCF and M−
3
CCH Bond Lengths. Several different bonding rationales have
3
been proposed to account for the M−CCF bond length contrac-
3
To probe this possibility, DFT calculations were carried out
on the compounds reported here, Cp2Ti(CF3)(X) [X = F (1),
Cl (12), Br (13), I (14), N3 (15), and OTf (2)]; their CH3
analogues, Cp2Ti(CH3)(X) [X = F (18), Cl (19), Br (20), I
(21), N3 (22), and OTf (23)]; and their zirconium and hafnium
halide analogues, Cp2Zr(CR3)(X) (R = H, X = F (26), Cl (27),
Br (28), I (29); R = F, X = F (30), Cl (31), Br (32), I (33)) and
Cp2Hf(CR3)(X) (R = H, X = F (36), Cl (37), Br (38), I (39);
R = F, X = F (40), Cl (41), Br (42), I (43)), respectively, in
order to compare and understand the differences in bonding as a
function of both the metal electronic structure and the choice of
CF3 or CH3 ligand. Some simpler model compounds from
groups 4 and 8 have also been examined to gain greater insight
(vide infra). In addition, properties of some selected titanium
and zirconium compounds containing CCl3 ligands, Cp2M-
(CCl3)(X) (M = Ti, X = F (24), Cl (25); M = Zr, X = F (34),
Cl (35)), were computed.
Density functional theory (DFT) calculations were carried
out using the B3LYP functional43 and the triple-ζ LACV3P**+
+ basis set,44 as implemented in the Jaguar suite of programs.
A comparison of the metrical parameters for the Cp2Ti(CF3)(X)
compounds (1, 2, 12, and 15) is presented in Table 3. Good
agreement is observed between the measurements obtained
experimentally and those calculated by DFT, allowing en-
hanced credence to the DFT-calculated metrics for the
additional hypothetical complexes discussed. No Cp2Ti(CH3)-
(X) compounds have been crystallographically characterized,45
so in order to further appraise the ability of our chosen method
to predict Ti−CH3 metrics, we calculated the structures of
Cp2M(CH3)2 (M = Ti (44), Zr (46), Hf (48)), the structures
of which have all been crystallographically determined.22c,46
Results are summarized in Table 4 and once again show
excellent agreement between DFT at this level and crystallo-
graphically determined values. Significantly, for the Cp2Ti-
tion observed in middle- and late-row transition-metal perfluoro-
alkyl complexes. One early suggestion invoked π-back-bonding
from the metal dπ electrons to the low-lying antibonding σ*
C−F orbitals, thus imparting some double-bond character to
the M−CCF bond and shortening it relative to the analogous
3
M−CCH bond in a metal alkyl complex.37 A completely
3
analogous description presented this phenomenon in terms of
hyperconjugation within the M−CF3 moiety, leading to a
shortened M−C bond due to partial contribution from doubly
bonded metal−carbon [MCF2]+[F−] resonance structures.30
Later studies seemed to rule out such π-effects and accounted
for the metal perfluoroalkyl bond length contraction using
arguments strictly within the σ-bonding regime. A carbonyl
force constant analysis on (CO)5MnR (R = CH3 and CF3)
separated the σ- and π-bonding effects and concluded that the
CH3 and CF3 ligands are equally poor π-electron acceptors.38
Calculations by Hall and Fenske on the same Mn compounds
indicated that the highly electropositive nature of the
fluorinated carbon, as well as the enhanced s character of the
CF3 group within the M−CCF σ-orbital, concentrates the
3
electron density closer to the central carbon and leads to the
observed M−C bond length shortening.39 Intuitively, this is
consistent with Bent’s Rule,40 which predicts a metal−
fluoroalkyl bond length contraction based upon the high
electronegativity of fluorine. This leads to bond formation with
carbon centers that deviate from sp3 hybridization such that
minimal s character is present in the C−F bond and a larger
amount of s character contributes to metal−carbon (M−CCF
)
3
bonding. The M−CCF3 bond is then contracted relative to its
alkyl counterparts because the spatial extent of the 2s orbital is
less than that of the 2p. These arguments appear reasonable
assuming predominantly covalent bonding, but it seems clear
that the s character in the carbon orbital should be less
important in determining bonding distance in a more ionic
interaction.
(CF3)(X) compounds (Table 3), DFT predicts longer Ti−CCF
3
For the last three decades, these explanations have ade-
quately accounted for the metal perfluoroalkyl bond length
contraction observed in middle- and late-row transition-metal
species. Certainly, M−C bond length shortening has been a
consistent trend in all of the previously reported middle- and
late-row M−CF3 complexes that have been structurally char-
acterized, from group 6 Cr41 and Mo23 to group 10 Ni15 and
Pt19 complexes. However, the early transition-metal Ti−CF3
complexes (1, 2, 12, 15) described above follow the opposite
trend, exhibiting metal perfluoroalkyl bond length elongation
when compared with analogous Ti−CH3 complexes.
Almost three decades ago, Oberhammer discussed the dif-
ferences in structural parameters between CH3 and CF3 groups
bound to a variety of main group elements E and pointed out
that the charge distribution within the CX3 group correlated
well with the bond length to E.42 At that time, simple charge
distribution calculations showed that, in a CH3 group, the
carbon bears a negative charge, with hydrogens partially positive,
whereas in CF3, the carbon is strongly positive and the fluorines
distances than for each specific CH3 analogue, and the trend
extends to the Zr and Hf analogues. However, it is noteworthy
that, within 0.01 Å, the Ti−CCH distances are essentially con-
3
stant at 2.18 Å, Ti−CCF at 2.25 Å, Zr−CCH at 2.30 Å, Zr−CCF
3
3
3
at 2.38 Å, Hf−C at 2.26 Å, and Hf−C at 2.35 Å, whereas
CH3
CF3
the M−X distances vary significantly. The crystallographic
distance values for the Ti complexes are likewise essentially the
same. Identical values (to within 0.01 Å) and trends between
metals are observed for Cp2M(CH3)2 (M = Ti, Zr, Hf) (Table 4).
There seems to be no significant effect of the ancillary ligand on
the M−C bond lengths for CH3 or CF3.
The quantum theory of atoms in molecules (QTAIM) has
been used to determine characteristics of the electron density at
the bond critical points (BCP) for these molecules.47 A number
of characteristic parameters can be calculated at the BCP (r)
and have been used to elucidate the nature of the bonds be-
tween atoms in molecules. The electron density [ρ(r)] and its
Laplacian [∇2ρ(r)] have been used extensively to characterize
1489
dx.doi.org/10.1021/om201055e | Organometallics 2012, 31, 1484−1499

Organometallics
Article
Table 3. Metric, NBO, NPA, and QTAIM Data for the Group 4 Complexes Cp2M(CR3)(X) (M = Ti; R = H, F; X = F, Cl, Br, I,
N3, CF3SO3. M = Ti, Zr; R = Cl; X = F, Cl. M = Zr, Hf; R = H, F; X = F, Cl, Br, I)
distance (Å)
angle (deg)
atomic NPA charges
group
charge
bond
ionicity
C → M
hybrid
QTAIM BCP data
(M−C)
compound
M−CR3 M−X
X−M−CR3
qM
qX
qCR3
qCR3
CR3
M−CR3
CR3
sp2.9
ρ(r)
H(r)
Cp2Ti(CH3)(F)
(18)
2.176
2.182
2.180
2.180
1.848
2.350
2.592
2.821
1.999
1.986
1.836
93.4
0.94
−0.52
−0.87
0.20
−0.27
28%
0.094
−0.038
Cp2Ti(CH3)(Cl)
(19)
92.1
91.9
91.7
92.3
91.5
0.57
0.52
0.50
0.71
0.84
0.86
0.47
0.41
0.38
0.61
0.76
0.93
0.53
1.82
1.42
1.32
1.25
1.74
1.33
1.22
1.14
1.81
1.39
2.01
1.63
1.52
1.45
1.93
1.53
1.42
1.32
−0.30
−0.27
−0.23
−0.45
−0.88
−0.50
−0.25
−0.20
−0.14
−0.44
−0.86
−0.49
−0.24
−0.64
−0.39
−0.31
−0.25
−0.63
−0.36
−0.27
−0.18
−0.63
−0.35
−0.65
−0.40
−0.32
−0.25
−0.64
−0.37
−0.28
−0.19
−0.85
−0.86
−0.87
−0.86
−0.85
0.80
0.20
0.21
−0.25
−0.23
−0.24
−0.26
−0.22
−0.34
−0.32
−0.33
−0.33
−0.33
−0.33
−0.39
−0.38
−0.47
−0.47
−0.47
−0.47
−0.48
−0.47
−0.47
−0.47
−0.54
−0.52
−0.52
−0.50
−0.50
−0.48
−0.52
−0.51
−0.51
−0.51
26%
32%
28%
28%
25%
36%
38%
38%
40%
36%
34%
50%
52%
52%
50%
50%
52%
55%
54%
54%
54%
69%
68%
58%
56%
56%
57%
60%
58%
58%
58%
sp2.9
sp2.8
sp2.8
sp2.8
sp2.9
sp1.6
sp1.6
sp1.6
sp1.5
sp1.4
sp1.6
sp1.5
sp1.5
sp2.4
sp2.4
sp2.3
sp2.3
sp1.4
sp1.4
sp1.4
sp1.3
sp1.2
sp1.2
sp2.2
sp2.1
sp2.0
sp2.0
sp1.4
sp1.3
sp1.3
sp1.2
0.094
0.094
0.094
0.093
0.095
0.080
0.079
0.079
0.078
0.079
0.081
0.069
0.067
0.088
0.089
0.089
0.089
0.073
0.073
0.073
0.073
0.067
0.067
0.094
0.095
0.095
0.095
0.078
0.079
0.079
0.078
−0.038
−0.038
−0.038
−0.037
−0.039
−0.026
−0.026
−0.026
−0.025
−0.025
−0.027
−0.022
−0.021
−0.024
−0.024
−0.024
−0.024
−0.014
−0.014
−0.014
−0.014
−0.012
−0.012
−0.028
−0.029
−0.029
−0.029
−0.018
−0.018
−0.018
−0.018
Cp2Ti(CH3)(Br)
(20)
Cp2Ti(CH3)(I)
(21)
0.21
Cp2Ti(CH3)(N3) 2.183
(22)
0.20
Cp2Ti(CH3)
(OTf) (23)
2.172
2.248
2.221(3)
2.257
2.301(6)
0.21
Cp2Ti(CF3)(F)
(1)
92.0
89.35(9)
−0.38
−0.38
−0.38
−0.38
−0.38
−0.38
0.01
a
b
a
b
a
1.831(2)
2.320
2.335(2)
Cp2Ti(CF3)(Cl)
(12)
91.3
89.27(13)
0.82
b
Cp2Ti(CF3)(Br)
(13)
2.257
2.263
2.252
2.553
2.777
1.972
1.991(1)
1.958
1.985(4)
91.5
0.81
Cp2Ti(CF3)(I)
(14)
91.6
90.4
89.59(5)
90.3
88.3(2)
0.81
Cp2Ti(CF3)(N3)
(15)
0.81
b
a
b
a
b
2.239(1)
Cp2Ti(CF3)(OTf) 2.245
(2)
0.81
a
2.222(5)
Cp2Ti(CCl3)(F)
(24)
2.319
1.834
2.322
1.992
2.466
2.674
2.896
1.982
2.443
2.646
2.858
1.980
2.445
1.953
2.434
2.645
2.864
1.944
2.410
2.613
2.826
90.9
91.4
98.5
97.0
97.2
97.6
98.6
96.5
96.9
97.2
98.0
96.8
97.2
95.9
95.7
95.9
96.7
95.2
95.4
95.8
−0.42
−0.41
−1.10
−1.10
−1.10
−1.10
0.66
Cp2Ti(CCl3)(Cl) 2.333
(25)
0.01
Cp2Zr(CH3)(F)
(26)
2.298
2.295
2.295
2.294
2.384
2.383
2.383
2.386
2.427
0.21
Cp2Zr(CH3)(Cl)
(27)
0.21
Cp2Zr(CH3)(Br)
(28)
0.21
Cp2Zr(CH3)(I)
(29)
0.21
Cp2Zr(CF3)(F)
(30)
−0.38
−0.38
−0.38
−0.38
0.01
Cp2Zr(CF3)(Cl)
(31)
0.67
Cp2Zr(CF3)(Br)
(32)
0.67
Cp2Zr(CF3)(I)
(33)
0.67
Cp2Zr(CCl3)(F)
(34)
−0.57
−0.55
−1.15
−1.13
−1.13
−1.14
0.62
Cp2Zr(CCl3)(Cl) 2.427
(35)
0.01
Cp2Hf(CH3)(F)
(36)
2.268
0.21
Cp2Hf(CH3)(Cl) 2.266
(37)
0.21
Cp2Hf(CH3)(Br) 2.265
(38)
0.21
Cp2Hf(CH3)(I)
(39)
2.263
2.355
2.354
2.354
2.356
0.22
Cp2Hf(CF3)(F)
(40)
−0.38
−0.38
−0.38
−0.38
Cp2Hf(CF3)(Cl)
(41)
0.63
Cp2Hf(CF3)(Br)
(42)
0.63
Cp2Hf(CF3)(I)
(43)
0.63
a
X-ray structures of Cp2Ti(CF3)(F) and Cp2Ti(CF3)(OTf): Taw, F. L.; Scott, B. L.; Kiplinger, J. L. J. Am. Chem. Soc. 2003, 125, 14712−14713.
This work.
b
1490
dx.doi.org/10.1021/om201055e | Organometallics 2012, 31, 1484−1499

Organometallics
Article
the concentration or depletion of electron density at a BCP as
being indicative of electron-sharing (covalency) or closed-shell
interaction (ionic or nonbonded) at this point.48 It has been
noted that Laplacians may not be the most reliable indicators
for some polar bonds,49 but Cremer and Kraka have suggested
that the sum [H(r)] of the kinetic energy density [G(r)], and
the potential energy density [V(r)] at the BCP is a more
reliable indicator.50 Dominance of the potential energy density
at the BCP gives a negative value for H(r) and is indicative that
electron density buildup at the BCP is stabilizing, suggestive of
covalency; a positive value of H(r) indicates the dominance of
the kinetic energy density, in which electron density accu-
mulation at the BCP is destabilizing, typical of a closed-shell
interaction. Typical values of H(r) for covalent bonds lie be-
tween −0.15 and −0.45 au, and clearly, the best direct com-
parisons are between like pairs of atoms.50a,b Thus, in
Cp2Ti(CH3)(F), the value of H(r) at the C−H BCPs of the
CH3 ligand averages −0.237, clearly indicative of covalency,
whereas that at the Ti−F bond is −0.008, indicative of a much
more ionic character in the latter; similarly, H(r) at the C−F
BCP in Cp2Ti(CF3)(F) is −0.280, whereas that at the Ti−F
BCP is −0.010. Values of ρ(r) and H(r) at the M−C bond
critical points are presented in Tables 3 and 4.
The natural bond orbital (NBO)51 method allows bond ionicities
to be calculated from the NBO polarization coefficients (c) for the
M−CX3 bonds according to the expression
cM2 − cC2
ionicity MC =
cM2 + cC2
and these appear in Tables 3 and 4. Once again, the ionicities
are reasonably constant within a given series, with each M−CF3
being slightly more ionic than the corresponding M−CH3
analogue. The greatest difference in bond ionicity between
M−CH3 and M−CF3 is observed for Ti, with a general increase
in bond ionicity for both ligands on descending the group. The
bond ionicity trends are consistent with those observed in the
H(r) values.
Charges (q) on the metal centers and the individual atoms of
the CR3 ligands were calculated using the NBO/NPA method,51b,c
which is suggested to be among the better methods for
apportioning charge within molecules,51c and are also presented
in Tables 3 and 4, along with group charges for the CR3 ligands.
For the sake of comparison, the corresponding atom NPA
charges and lone-pair hybridizations for the CR3− anions were
also calculated at the same DFT level: CH3− qC = −1.40, qH =
+0.13 (sp5.4 hybrid lone pair), CF3− qC = +0.47, qF = −0.49
(sp0.6 hybrid lone pair), and CCl3− qC = −0.34, qCl = −0.22
(sp0.3 hybrid lone pair). In agreement with Oberhammer’s
results,42 the carbon of the CH3− anion bears a strong negative
charge, whereas that of the CF3− anion bears a significant
positive charge, with large negative charges on the F atoms. The
lone pair of the CH3− anion resides in an orbital with large p
character, whereas that of the CF3− anion has considerably
greater s character, as predicted by Bent’s Rule. The CCl3−
anion is interesting in that its charges are distributed in the
same sense as the CH3− anion, with a negatively charged
carbon, but, like the CF3− anion, the lone pair is in a high
s-character orbital. The NPA charges in CF3− and CCl3− anions
agree closely with those reported by Schrobilgen.52
Examination of Tables 3 and 4 illustrates that the trends in
charge distribution within the CR3 ligands are maintained in all
Ti, Zr, and Hf compounds. Remarkably, the charges on the
1491
dx.doi.org/10.1021/om201055e | Organometallics 2012, 31, 1484−1499

Organometallics
Article
hydrogen atoms of every CH3 ligand remain essentially con-
stant at +0.21 for all compounds, independent of the metal or
the anionic ligand; similarly, the CF3 fluorine atoms bear an
essentially constant charge of −0.38 throughout and the Cl
atoms in the CCl3 examples are constant at ∼0.01. The idea of
the different charge capacities of atoms in molecules has been
discussed in detail by Politzer and Murray.53 Inspection of the
charges on the carbon atoms is also instructive. For the CH3
and CCl3 ligands, the charge on carbon becomes more negative
on going from Ti→Zr→Hf, but for each metal, it is
independent of the anionic ligand. Likewise, the CF3 carbon
becomes less positive on going from Ti→Zr→Hf, but again
remains constant for a particular metal. In contrast, the negative
charges on ancillary ligand X are significantly different and the
positive charge on the metal center also varies significantly
within each series.
The carbon hybrid orbital in the M−CR3 bond depends
strongly on R, in the same manner as in the CR3− anions, with
significantly higher s character for R = F and Cl; there is a small
dependence on M, with slight increases in s character in each
class on descending the group. The hybridization is essentially
independent of the ancillary ligand in each subgroup.
Bond ionicities for M−CR3 depend on R, decreasing as
Cl > F > H, and increase somewhat on descending the group,
with only a small dependency on ancillary ligand X for each M.
However, the electron densities ρ(r) and the total energy
densities, illustrated by the Cremer−Kraka parameter H(r), at
the M−C bond critical points are each independent of X and
are constant for a given R in CR3. The values of ρ(r) are con-
stant for a given M and ancillary ligand X and vary only slightly
with varying M; the variation between R = H, F, and Cl is more
pronounced with lower electron densities as R = H > F > Cl
for each case. Similarly, the values of H(r) show completely
analogous dependencies; the small negative values are indic-
ative of strong ionic components to the M−C bonds in each
case, with a greater ionic character when X = Cl > F > H, con-
sistent with bond ionicities.
giving a 12-valence electron maximum, after which additional
coordination must be accommodated by multicenter bonding
motifs.51c Therefore, for the complexes Cp2Ti(X)(Y) described
in Tables 3 and 4, the necessity of a 16-valence electron count
makes a detailed NBO analysis more complex. To reduce steric
effects and to gain more insight into σ- and π-effects, all of
which contribute to differences in bond lengths, a model sys-
tem designed to accommodate a maximum of 12 valence elec-
trons was studied. The tetrahedral d0 compounds M(CH3)4
(M = Ti (52), Zr (55), Hf (58)) are well known, and their
structures have been studied computationally.55 The NBO
description of these molecules involves a reference Lewis struc-
ture with 4sd3 hybrids accommodating the 2-center/2-electron
σ-bonds to CH3 in a valence saturated structure, leaving two
empty, degenerate d orbitals. Similarly, the hypothetical d4
compounds M(CH3)4 (M = Fe (61), Ru (64), Os (67)) repre-
sent a set of isostructural and isovalent analogues, with the
same NBO Lewis structure for their singlet states, except with
the two degenerate d orbitals filled. Thus, in the group 4 com-
pounds, the metals are only capable of being π-acceptors, where-
as in the group 8 analogues, they are only capable of being
π-donors. Accordingly, the structures of M(CH3)4 (M = Ti (52),
Zr (55), Hf (58), Fe (61), Ru (64), Os (67)), M(CH3)3(CF3)
(M = Ti (53), Zr (56), Hf (59), Fe (62), Ru (65), Os (68)),
and M(CH3)3(CCl3) (M = Ti (54), Zr (57), Hf (60), Fe (63),
Ru (66), Os (69)) were optimized, as singlets, using the identical
functional and basis set as before. The calculated Kohn−Sham
LUMO for Zr(CH3)4 (55) and HOMO Ru(CH3)4 (64) are
shown in Figure 5, along with the corresponding NBOs. All the
compounds were also subjected to NBO, NPA, and QTAIM
analyses. Results are presented in Table 5.
The overall picture is consistently one of strongly polar
M−CR3 bonds. For each metal, the order of M−CR3 bond
length is R = Cl > F > H. This correlates inversely with the s
character in the carbon orbital, indicating that any spatial
contraction of this orbital is not dominant in determining bond
lengths in these compounds. While the signs of the local charges
on M and C in M−CH3 (δ+/δ−) and M−CF3 (δ+/δ+) are
consistent with a longer M bond to CF3, they do not explain an
even longer bond to CCl3. While local charges on carbon are
constant for each M−CR3, significant variation in the magnitude
of the positive charge on M generates no significant difference in
the M−C distances, the electron density ρ(r), or Cremer−Kraka
parameter H(r) at the bond critical points, or the bond ionicities.
We considered that the longer bonds to CF3 and CCl3 might
be a result of steric crowding in the somewhat constrained space
in the “wedge” of the Cp2M system. It seemed sensible to design
a model system in which steric effects would be minimized. The
complexes in Tables 3 and 4 are pseudotetrahedral M(IV) com-
plexes with d0 electron counts and 16-valence electron
configurations. The NBO method has been widely used to
describe the bonding in transition-metal complexes, for which
the valence orbital set is ns and (n − 1)d, with essentially no
contributions from the energetically high-lying np orbitals.51c,54
As a consequence of a valence orbital set of six, the maximum
number of electron pairs, bonding and nonbonding, that can be
accommodated around a transition metal must also be six,
Figure 5. Top: one of a degenerate pair of Kohn−Sham LUMOs in
Zr(CH3)4 (55) and the corresponding HOMO in Ru(CH3)4 (64).
Bottom: the corresponding NBOs. Each orbital is depicted as a 0.05e
isosurface.
Considering, first, the group 4 complexes and comparing
data from Tables 4 and 5 in which two Cp ligands have been re-
placed by two CH3 groups, the results are remarkably consis-
tent. Replacement of Cp by CH3 causes a dramatic increase in
the NPA positive charge at the metal center, illustrating how
superior a donor Cp is compared with CH3. The negative
charge at the CH3 and CCl3 carbons increases and the positive
charge at the CF3 carbon decreases slightly, with charges on
the H, Cl, and F atoms remaining essentially constant. Bond
ionicities are slightly higher compared with those of Cp2M
analogues, particularly for Ti, but the values of ρ(r) and H(r)
1492
dx.doi.org/10.1021/om201055e | Organometallics 2012, 31, 1484−1499

Organometallics
Article
1493
dx.doi.org/10.1021/om201055e | Organometallics 2012, 31, 1484−1499

Organometallics
Article
remain very similar. Hybridizations for the carbon donor orbital
in CR3 show higher p character in all cases, but the trend of
significantly lower p character for R = F and Cl remains. The
metal orbital hybrids are very close to the expected51c sd3 in all
cases. We conclude that M(CH3)4 and M(CH3)3(CR3)
complexes are good models for the Cp2M analogues and that
the fundamental electronic characteristics of the M−CCR bonds are
3
not significantly disturbed.
Comparing M(CH3)4, M(CH3)3(CF3), and M(CH3)3(CCl3)
illustrates that, in these less sterically hindered systems, the
M−CCF bond is always the longest, with a shorter M−CCCl
Figure 6. NLMOs for a single C−H(σ) in Zr(CH3)4 (55) and in
Ru(CH3)4 (64) represented as 0.03e isosurfaces.
3
3
bond and an even shorter M−CCH distance. A CF3 or CCl3
3
Zr(4d) orbitals of 0.216e, or 0.054e per CH3 group. Figure 7
illustrates NLMOs of C−F(σ*) in Zr(CH3)3(CF3) (56) and
Ru(CH3)3(CF3) (65), illustrating significant delocalization
from the filled Ru(4d) → CF(σ*), but none, of course, from
the empty Zr(4d). In this case, the two Ru(4d) orbitals are
group slightly shortens the M−CCH bond in M(CH3)3(CR3)
3
compared with that in the corresponding M(CH3)4 analogue.
Thus, the trend in M−C distances now follows the trend in
relative local charges on the atoms in the M−C bonds, with
M−CH3 being shorter than M−CCCl , and with M−CCF being
3
3
the longest, presumably due to δ+/δ+ repulsion. In these
strongly polar bonds, the M−C distance does not correlate with
the s character in the carbon donor orbital.
Turning to the d4 group 8 analogues in Table 5, we note that
all bond lengths decrease compared with those of the
appropriate group 4 analogues, as expected for the smaller
atomic radii. While the Fe−CCF distance remains slightly
3
longer than the analogous Fe−CCH distances, the Fe−CCCl
3
3
distance is now the same as Fe−CCH . However, examination of
3
the Ru and Os analogues reveals that the M−CCF and M−CCCl
3
3
Figure 7. NLMOs for a single C−F(σ*) in Zr(CH3)3(CF3) (56) and
Ru(CH3)3(CF3) (65) represented as 0.03e isosurfaces.
distances both become shorter than M−CCH , with M−CCCl
3
3
even becoming slightly shorter than M−CCF ! The trend for
3
depleted by 0.116e. Figure 8 shows the NLMOs of one of the
Zr(4d) and Ru(4d) orbitals; clearly, the Zr(4d) is populated
exclusively by delocalization from CH(σ) orbitals with
negligible contributions from F lone pairs, whereas the Ru(4d)
CH3 and CF3 is in line with other previously discussed examples
from later in the periodic table. The NPA charges on the metals
decrease substantially compared with their more electropositive
group 4 counterparts; charges on the CH3 and CCl3 carbons
become less negative and those on the CF3 carbons more pos-
itive, with those on H, Cl, and F remaining essentially un-
changed from the values observed in group 4. Thus, once again,
most of the changes in charge distribution are accommodated by
M and C. The bond ionicities for the group 8 compounds are
substantially lower than those for group 4; this additional covalency
is also reflected in greater values for the electron densities ρ(r) and
more negative values of the Cremer−Kraka parameter H(r) at the
M−C bond critical points. In these more covalent compounds,
there is no correlation between charges on the metals or ligated
carbon atoms and the M−C bond length variations, but a more
sensible correlation with increased s character in the carbon orbital.
NBO analysis provides further insight and suggests that, for
group 4, additional π-components serve to enhance the relative
Figure 8. NLMOs for Zr(4d) and Ru(4d) in Zr(CH3)3(CF3) (56)
and Ru(CH3)3(CF3) (65) represented as 0.03e isosurfaces.
is depopulated almost exclusively by delocalization into CF(σ*)
orbitals.
It is well established56 that such delocalizations can be quan-
tified using second-order perturbative NBO analysis for esti-
mating metal−ligand π-bonding interaction energies (ΔE), obtained
from the off-diagonal Fock matrix element expressed in the
NBO basis.51c Values of ΔE reflect stabilization energies re-
sulting from delocalization from the reference structure, and
previous studies have generally involved delocalization of metal-
based lone pairs into π* antibonding orbitals of the ligands.
Delocalization energies for the groups 4 and 8 model com-
pounds are presented in Table 5; for each group 4 complex, the
π-donation is 9−10 kcal/mol/methyl group for Ti, Zr, and Hf,
whereas for the group 8 analogues, the π-acceptor properties
of the CF3 and CCl3 ligands dominate over the relatively small
shortening of M−CCH , whereas in group 8, π-components
3
serve to shorten both M−CCF and M−CCCl relative to M−CCH
.
3
3
3
As discussed above, the principal electronic structural difference
between the groups 4 and 8 complexes involves the occupancies
(vacant or filled, respectively) of the degenerate pair of d orbitals.
NBO recognizes delocalizations from the reference structure and
allows their depiction as natural localized molecular orbitals
(NLMOs).51c Figure 6 shows the NLMOs of C−H σ-bonds in
Zr(CH3)4 (55) and in Ru(CH3)4 (64). In complex 55, partial
delocalization from the CH(σ)→Zr(4d) is apparent, whereas
no such delocalization is observed in the analogous Ru complex
(64); this donation from all 12 available CH(s) orbitals
provides an overall occupancy of the two formally empty
1494
dx.doi.org/10.1021/om201055e | Organometallics 2012, 31, 1484−1499

Organometallics
Article
π-acceptance by each CH3 group, with CCl3 being as good a π-
acceptor as CF3. The π-donor properties of the metal increase
significantly from Fe→Ru→Os. All things considered, we
conclude that, for group 4 complexes, π-bonding is a significant
factor in shortening the strongly ionic M−CCH relative to M−CCR
3
3
(R = F, Cl), for the predominantly covalent group 8 analogues,
π-back-bonding helps to shorten the predominantly covalent
M− to M−CCR relative to M−CCH , at least for Ru and Os.
3
3
Despite the presence or absence of π-donation into the
CF(σ*), the C−F bond lengths remain identical to within
0.01 Å in all compounds. In fact, the Ru and Os compounds in
which this back-bonding is at its greatest actually have slightly
shorter C−F bonds than the group 4 analogues, for which no
back-bonding exists. It is not clear what is “normal” for a C−F
bond length in such compounds, and comparison between dif-
ferent molecules usually introduces multiple variables. Accord-
ingly, the molecules M(CH3)3(CF2CF2CF2CF3) (M = Ti (70),
Zr (71), Hf (73), Fe (74), Ru (75), Os (76)) were studied
computationally, using the same DFT methodology, and results
are presented in Table 6. An internal comparison between a
C−F bond α, β, and γ to a metal can now be made; this is a
more informative intramolecular comparison and avoids com-
parisons between CF3 groups, which necessarily must be inter-
molecular. The trends in M−CCH versus M−CCF distances are
3
2
identical to those in Table 5, with Ru and Os having shorter
bonds to the fluoroalkyl carbon than to CH3. The same relative
trends are also observed for the α-C−F bond distances when
compared to the α-C−F distances in Table 5, though it is
noteworthy that the C−F bonds α to the metals are all
significantly shorter than the C−F bonds in the CF3 group at
the end of the fluoroalkyl chain (Table 6). In the internal
comparison set, the α-CF2 distances are the longest and most
variable, with significant shortening observed in the β-CF2
groups; on reaching the γ-CF2 groups, convergence to a con-
stant value of 1.351 Å is achieved. Application of Bent’s Rule40a
predicts that replacement of a fluorine in a CF3 group with
anything less electronegative than F (i.e., anything) should
result in less competition for p character between the remaining
fluorines and hence longer CF bonds; comparison between the
γ-CF2 and the terminal CF3 distances illustrates this nicely. We
suggest that the significant elongation of C−F bonds α to
metals is mainly a consequence of the electropositivity of the
metals, with minor, if any, contributions from π-effects; the
bond-lengthening effect is most pronounced at the α-position
and decays rapidly on moving away from the metal.
CONCLUSIONS
■
The first series of titanocene trifluoromethyl halide and
pseudohalide complexes, Cp2Ti(CF3)(X) (where X = F (1),
OTf (2), Cl (12), Br (13), I (14), N3 (15), and OSO2Ph (16)),
has been prepared. Our approach capitalized on the latent reac-
tivity of titanium fluoride bonds; reaction between Cp2TiF2 and
(CH3)3SiCF3 provided access to Cp2Ti(CF3)(F) (1), which
served as a platform for preparing other Cp2Ti(CF3)(X) com-
plexes using similar (CH3)3Si−F elimination chemistry. This
collection of compounds represents a milestone in the history
of organometallic fluorocarbon chemistry. In contrast to earlier
beliefs, the Ti−CF3 linkage in these complexes is remarkably
robust, showing no evidence of an α-fluoride (Ti···F−CF2)
between the electrophilic Ti(IV) metal center and any of the
C−F bonds in the trifluoromethyl group. However, in the
solid-state, these complexes are shock-sensitive, and the
1495
dx.doi.org/10.1021/om201055e | Organometallics 2012, 31, 1484−1499

Organometallics
Article
(Note: The reaction also works with only THF as the solvent, but
slightly higher yields of 1 are achieved in a 1:1 mixture of THF and
Et2O). While stirring, Me3SiCF3 (1.0 mL, 0.99 g, 6.94 mmol) was
added dropwise, and the resulting yellow/brown slurry was stirred at
room temperature for 15 h. The mixture was then filtered through a
glass-fiber plug to remove insoluble materials. The orange/brown
filtrate was collected and the volatile materials were removed under
reduced pressure to produce an orange/brown waxy solid. (Note: Care
was taken to remove the source of vacuum pressure as soon as a waxy
solid was formed, as complete removal of solvent causes rapid de-
composition of the compound.) The solid was washed with cold
hexanes (3 × 3 mL), and any residual volatiles were removed under
reduced pressure (again, by exposing the mixture to reduced pressure
for a minimum amount of time to avoid decomposition) to produce 1
as a bright yellow solid (110 mg, 0.413 mmol, 60%). Complex 1
decomposes at room temperature and must be stored cold (−30 °C).
Safety Reminder! The quantities in which 1 is synthesized should be
limited to ∼150 mg, as 1 is susceptible to violent decomposition and
must be handled with care. 1H NMR (300 MHz, THF-d8, RT): δ 6.40
energetic decomposition of Cp2Ti(CF3)(F) (1) provides a new
entry into fluorinated materials containing −(CF2−CF2)− and
−(CF2−CFH)− units.
These Ti−CF3 complexes also provide an unprecedented
opportunity to understand the electronic structure and bonding
of early, middle, and late transition-metal trifluoromethyl, methyl,
fluoroalkyl, and alkyl complexes. In contrast to middle- and late-
row transition metals, the Ti−CF3 complexes featured a longer
titanium−carbon bond compared with their Ti−CH3 counter-
parts. Density functional theory (DFT) calculations were per-
formed to gain insight into this unusual bonding scenario. The
calculations accurately predicted longer Ti−CF3 distances than
for each specific CH3 analogue, and the trend extended to
structurally related Zr and Hf analogues. Simpler model com-
pounds from groups 4 and 8 showed that, for group 4 complexes,
π-bonding is a significant factor in shortening the strongly ionic
M−CH3 relative to M−CF3, whereas for the group 8 analogues,
π-back-bonding helps to shorten the predominantly covalent
M−CF3 relative to M−CH3. The bonding analysis suggests that
the significant elongation of C−F bonds α to metals is mainly a
consequence of the electropositivity of the group 4 metal
centers, with minor, if any, contributions from π-effects; the
bond-lengthening effect is most pronounced at the α-position
and decays rapidly on moving away from the metal.
3
(d, JF−H = 1.0 Hz, Cp). 19F NMR (282 MHz, THF-d8, RT): δ 206.3
(s, Ti-F), −24.0 (s, Ti-CF3). 13C{1H} NMR (75 MHz, THF-d8, RT): δ
153.1 (dq, 1JC−F = 382.6 Hz, 2JC−F = 4.8 Hz, Ti-CF3), 117.4 (s, Cp). IR
(cm−1, Nujol mull): 585 (m), 620 (w), 692 (w), 822 (s), 836 (s), 849
(s), 941 (m), 979 (s), 992 (s), 1017 (w), 1029 (w), 1069 (s), 1081 (s),
1133 (w), 1201 (w), 1274 (w), 1440 (m).
Synthesis of Cp2Ti(CF3)(OTf) (2). In the drybox, a 20 mL
scintillation vial equipped with a stir bar was charged with 1 (100 mg,
0.376 mmol). The minimum volume of toluene (∼5 mL) required to
completely dissolve 1 was added. While stirring, Me3SiOTf (68 μL,
84 mg, 0.376 mmol) was added as one aliquot, and pyridine (36 μL,
36 mg, 0.451 mmol) added immediately thereafter. The resulting
orange solution was layered with hexanes (∼2 mL) and placed in a
freezer (−30 °C) to crystallize. After 15 h, orange-red crystals were
observed, and the solvent was removed by pipet. The crystals were
washed with cold hexanes (∼5 mL), and residual hexanes were
removed by brief exposure to reduced pressure. To remove traces of 1
still present in the product mixture, these crystals were allowed to sit at
RT for 2 h and then dissolved in a pyridine (36 μL, 36 mg, 0.451
mmol)/toluene (∼5 mL) solution. Hexanes were added (∼2 mL), and
the mixture was placed again in the freezer to crystallize. After 15 h,
orange-red crystals were formed, and the solvent was removed by
pipet. After a wash with cold hexanes (∼5 mL), the crystals were
exposed briefly to reduced pressure to remove residual hexanes, and
2 (105 mg, 0.265 mmol, 70%) was isolated. Complex 2 decomposes
at room temperature within a few hours and should be stored cold
(−30 °C). Safety Reminder! The quantities in which this compound is
synthesized should be limited to ∼150 mg, as 2 is susceptible to
EXPERIMENTAL SECTION
■
General Considerations. All reactions and manipulations were
carried out using a Vacuum Atmospheres (MO 40-2 Dri-train) inert
atmosphere (He) drybox, or using standard Schlenk techniques. Glassware
was dried overnight at 150 °C before use. All solution NMR spectra
were obtained in C61D6 using a Bruker Avance 300 MHz spectrometer.
Chemical shifts for H and 13C{1H} NMR spectra were referenced to
solvent impurities. For 19F NMR spectra, CFCl3 was used as an
external reference at δ 0.00 ppm. All IR data were recorded on a
Thermo-Nicolet FT-IR spectrometer. The SEM experiments were
performed using a JEOL JSM-6460 scanning electron micrsocope. The
powder X-ray diffraction experiments were performed using a Bruker
D8 Discover with a GADDS diffractometer. The bonding config-
uration and the compositional analysis of the fluorocarbon materials
were carried out by X-ray photoelectron spectroscopy (SPECS,
Germany) using a hemispherical analyzer (HSA 3500). All solid-state
NMR spectra were obtained using a Bruker Avance 400 MHz spec-
trometer with a wide-bore superconducting magnet. Samples were spun
at ∼20 000−30 000 Hz in a 2.5 mm zirconia rotor at the magic angle.
The 1H, 19F, and 13C frequencies were 399.95, 376.32, and 100.58 MHz,
respectively. The 13C spectra were collected under cross-polarization
1
violent decomposition and must be handled with care. H NMR (300
MHz, THF-d8, 0 °C): δ 6.76 (s, Cp). 19F NMR (282 MHz, THF-d8,
0 °C): δ −25.6 (s, Ti-CF3), −80.0 (s, OTf). 13C{1H} NMR (75 MHz,
THF-d8, 0 °C): δ 151.8 (q, 1JC−F = 382.9 Hz, Ti-CF3), 120.7 (s, Cp),
120.2 (q, 1JC−F = 317.3 Hz, OTf). IR (cm−1, Nujol mull): 594 (w), 632
(m), 694 (w), 766 (w), 836 (s), 857 (w), 864 (w), 951 (sh, w), 990
(s), 1015 (s), 1026 (sh, w), 1069 (s), 1082 (s), 1132 (w), 1175 (s),
1184 (sh, m), 1204 (s), 1237 (s), 1298 (w), 1328 (w), 1345 (s).
Synthesis of Cp2Ti(CF3)(Cl) (12). Method A. In the drybox, a
20 mL scintillation vial equipped with a stir bar was charged with 1
(100 mg, 0.376 mmol) and 5 mL of THF. While stirring, (CH3)3SiCl
(0.24 mL, 1.88 mmol) was added by syringe. The reaction mixture
was allowed to stir for 3 h at room temperature, and then all volatile
materials were removed under reduced pressure to produce an orange
solid. This solid was dissolved in toluene (3−4 mL), layered with
hexanes (1−2 mL), and placed in a freezer (−30 °C) to crystallize.
After 15 h, orange crystals were observed, and the solvent was re-
moved by pipet. The crystals were washed with cold hexanes (3 × 1 mL)
and the residual hexanes were removed by brief exposure to reduced
pressure to give complex 12 (93 mg, 0.329 mmol, 88%). Complex 12
decomposes within a day at room temperature and should be
stored cold (−30 °C). Safety Reminder! The quantities in which 1 is
1
from H. The number of acquisitions for the 13C cross-polarization
magic-angle spinning (CPMAS) experiment was ∼20 000, and 800 data
acquisitions were needed for the 19F MAS results.
Materials. Unless otherwise noted, reagents were purchased from
commercial suppliers and used without further purification. THF-d8
(Cambridge Isotope Laboratories) was dried over a Na mirror prior to
use. Celite (Aldrich), 4 Å molecular sieves (Aldrich), and alumina
(Brockman I, Aldrich) were dried under dynamic vacuum at 250 °C
for 48 h prior to use. All solvents (Aldrich) were purchased anhydrous
and were dried over KH for 24 h, passed through a column of acti-
vated alumina, and stored over activated 4 Å molecular sieves prior
to use. Complexes 1 and 2 were prepared according to literature
procedures.11
Caution! The quantities in which Cp2Ti(CF3)(F) (1), Cp2Ti(CF3)-
(OTf) (2), and Cp2Ti (CF3)(Cl) (12) are synthesized should be
limited to ∼150 mg, as 1, 2, and 12 are susceptible to violent decomposi-
tion and must be handled with care.
Synthesis of Cp2Ti(CF3)(F) (1). In the drybox, a 20 mL
scintillation vial equipped with a stir bar was charged with Cp2TiF2
(150 mg, 0.694 mmol) and CsF (105 mg, 0.694 mmol). Tetra-
hydrofuran (5 mL) and Et2O (5 mL) were added to create a suspension
1496
dx.doi.org/10.1021/om201055e | Organometallics 2012, 31, 1484−1499

Organometallics
Article
synthesized should be limited to ∼150 mg, as 1 is susceptible to
violent decomposition and must be handled with care.
1131 (m), 1170 (w), 1201 (w), 1340 (s), 1964 (w), 2060 (s), 2663 (w),
2724 (w).
Method B. A 20 mL scintillation vial equipped with a stir bar was
charged with 1 (100 mg, 0.376 mmol) and toluene (5 mL). While
stirring, (CH3)3SiCl (0.24 mL, 1.88 mmol) was added by syringe. The
reaction mixture was allowed to stir for 3 h at room temperature,
layered with hexanes (1−2 mL), and placed in a freezer (−30 °C) to
Synthesis of Cp2Ti(CF3)(OSO2Ph) (16). In the drybox, a 20 mL
scintillation vial equipped with a stir bar was charged with 1 (66 mg,
0.247 mmol) and toluene (3 mL). While stirring, (CH3)3Si(OSO2Ph)
(50 μL, 0.247 mmol) was added as one aliquot, and pyridine (24 μL,
0.296 mmol) immediately added. (Note: Pyridine was added to help
stabilize the product that is formed. Without the presence of a Lewis
base, rapid decomposition occurred in solution to form intractable
solids.) The resulting orange solution was layered with hexanes
(∼1 mL) and placed in a freezer (−30 °C) to crystallize. After 15 h, a
fine orange solid was observed, and the solvent was removed by pipet.
The solid was washed with cold hexanes (3 × 1 mL); residual hexanes
were removed by brief exposure to reduced pressure to produce
complex 16 (73 mg, 0.181 mmol, 73%). Complex 16 decomposes
within a few hours at room temperature and should be stored cold
1
crystallize. After 24 h, X-ray quality orange crystals of 12 formed. H
NMR (300 MHz, THF-d8, RT): δ 6.52 (s, Cp). 19F NMR (282 MHz,
THF-d8, RT): δ −21.2 (s, Ti-CF3). 13C{1H} NMR (75 MHz, THF-d8,
1
RT): δ 153.2 (q, JC−F = 382.1 Hz, Ti-CF3), 118.6 (s, Cp). IR (cm−1,
Nujol mull): 691 (w), 826 (s), 843 (w), 854 (w), 873 (w), 933 (m),
976 (s), 1015 (m), 1026 (w), 1064 (s), 1077 (s).
Synthesis of Cp2Ti(CF3)(Br) (13). In the drybox, a 20 mL
scintillation vial equipped with a stir bar was charged with 1 (100 mg,
0.376 mmol) and THF (5 mL). While stirring, (CH3)3SiBr (51 μL,
0.395 mmol) was added by syringe. The reaction mixture was allowed
to stir for 45 min at room temperature, and then all volatile materials
were removed under reduced pressure to give an orange-red solid.
This solid was dissolved in toluene (3−4 mL), layered with hexanes
(1−2 mL), and placed in a freezer (−30 °C) to crystallize. After 15 h,
orange-red crystals formed, and the solvent was removed by pipet. The
crystals were washed with cold hexanes (3 × 1 mL); residual hexanes
were removed by brief exposure to reduced pressure to produce com-
plex 13 (103 mg, 0.315 mmol, 84%). Complex 13 decomposes within
a few hours at room temperature and should be stored cold (−30 °C).
1H NMR (300 MHz, THF-d8, RT): δ 6.58 (s, Cp). 19F NMR (282
MHz, THF-d8, RT): δ −20.6 (s, Ti-CF3). 13C{1H} NMR (75 MHz,
THF-d8, RT): δ 155.2 (q, 1JC−F = 381.5 Hz, Ti-CF3), 118.6 (s, Cp). IR
(cm−1, Nujol mull): 690 (w), 827 (s), 844 (w), 854 (w), 871 (w), 932
(m), 975 (s), 1015 (m), 1024 (w), 1064 (s), 1074 (s).
1
(−30 °C). H NMR (300 MHz, THF-d8, RT): δ 7.80 (m, Ph), 7.48
(m, Ph), 7.45 (m, Ph), 6.63 (s, Cp). 19F NMR (282 MHz, THF-d8,
RT): δ −24.5 (s, Ti-CF3). 13C{1H} NMR (75 MHz, THF-d8, RT): δ
152.4 (q, 1JC−F = 383.1 Hz, Ti-CF3), 143.1 (s, Ph), 132.3 (s, Ph), 129.4
(s, Ph), 127.3 (s, Ph), 119.3 (s, Cp). IR (cm−1, Nujol mull): 582 (w),
612 (m), 691 (s), 727 (s), 763 (m), 832 (s), 844 (m), 855 (sh, w), 871
(sh, w), 940 (w), 974 (s), 1004 (w), 1028 (w), 1080 (s), 1110 (m),
1128 (w), 1168 (s), 1208 (w), 1223 (w), 1239 (w), 1302 (s).
X-ray Crystallographic Details for Cp2Ti(CF3)(Cl) (12). A
crystal (0.24 × 0.20 × 0.14 mm3) of complex 12 was mounted in a
nylon cryoloop using Paratone-N oil under an argon gas flow. The
data were collected on a Bruker Platform diffractometer with a 1K
CCD and cooled using a Bruker KRYO-FLEX liquid nitrogen vapor
cooling device. The instrument was equipped with a sealed, graphite
monochromatized Mo Kα X-ray source (λ = 0.71073 Å) and a 0.5 mm
monocapillary. A hemisphere of data was collected using φ scans, with
30 s frame exposures and 0.3° frame widths. Data collection and initial
indexing and cell refinement were handled using SMART software.57
Frame integration, including Lorentz-polarization corrections, and final
cell parameter calculations were carried out using SAINT software.58
The data were corrected for absorption using the SADABS program.59
Decay of reflection intensity was monitored by analysis of redundant
frames. The structure was solved using direct methods and difference
Fourier techniques. A substitutional chloride impurity was modeled at
the CF3 site, with site occupancy factors tied to one. The site occupancy
factor on the chloride impurity refined to 0.14(1). A racemic twin was
also refined, and the batch scale factors converged to 0.54(7). Non-
hydrogen atoms were refined anisotropically, and hydrogen atoms were
treated as idealized contributions. Structure solution, refinement, graphics,
and creation of publication materials were performed using SHELXTL.60
Additional details regarding data collection are provided in the CIF file
(Supporting Information).
Synthesis of Cp2Ti(CF3)(I) (14). In the drybox, a 20 mL
scintillation vial equipped with a stir bar was charged with 1 (93 mg,
0.351 mmol) and toluene (5 mL). While stirring, (CH3)3SiI (50 μL,
0.351 mmol) was added as one aliquot, and pyridine (34 μL, 0.421
mmol) immediately added. (Note: Pyridine was added to help stabilize
the product that is formed. Without the presence of a Lewis base, rapid
decomposition occurred in solution to form intractable solids.) The
resulting green solution was layered with hexanes (∼2 mL) and placed
in a freezer (−30 °C) to crystallize. After 15 h, a fine green-black solid
was observed, and the solvent was removed by pipet. The solid was
washed with cold hexanes (3 × 1 mL); residual hexanes were removed
by brief exposure to reduced pressure to produce complex 14 (88 mg,
0.235 mmol, 67%). Complex 14 decomposes within a few hours at
1
room temperature and should be stored cold (−30 °C). H NMR
(300 MHz, THF-d8, −20 °C): δ 6.72 (s, Cp). 19F NMR (282 MHz,
THF-d8, −20 °C): δ −20.5 (s, Ti-CF3). 13C{1H} NMR (75 MHz,
1
THF-d8, −20 °C): δ 159.3 (q, JC−F = 380.0 Hz, Ti-CF3), 118.3
(s, Cp). IR (cm−1, Nujol mull): 689 (m), 822 (s), 857 (s), 869 (sh, w),
927 (w), 943 (m), 971 (s), 992 (m), 1015 (m), 1024 (m), 1061 (s),
1076 (sh, m), 1132 (m), 1202 (w).
X-ray Crystallographic Details for Cp2Ti(CF3)(N3) (15).
A crystal (0.30 × 0.10 × 0.10 mm3) of complex 15 was mounted in a
nylon cryoloop using Paratone-N oil under an argon gas flow. The
data were collected on a Bruker D8 APEX II charge-coupled device
(CCD) diffractometer with an American Cryoindustries Cryocool low-
temperature device. The instrument was equipped with a graphite
monochromatized Mo Kα X-ray source (λ = 0.71073 Å) and a 0.5 mm
monocapillary. A hemisphere of data was collected using ω scans, with
10 s frame exposures and 0.5° frame widths. Data collection and initial
indexing and cell refinement were handled using APEX II software.61
Frame integration, including Lorentz-polarization corrections, and
final cell parameter calculations were carried out using SAINT+ soft-
ware.62 The data were corrected for absorption using redundant
reflections and the SADABS program.59 Decay of reflection intensity
was not observed as monitored by analysis of redundant frames.
The structure was solved using direct methods and difference Fourier
techniques. Non-hydrogen atoms were refined anisotropically, and
hydrogen atoms were treated as idealized contributions. Structure
solution, refinement, graphics, and creation of publication materials
were performed using SHELXTL.60 Additional details regarding data
collection are provided in the CIF file (Supporting Information).
Synthesis of Cp2Ti(CF3)(N3) (15). In the drybox, a 20 mL
scintillation vial equipped with a stir bar was charged with 1 (100 mg,
0.376 mmol) and THF (5 mL). While stirring, (CH3)3SiN3 (0.49 mL,
3.76 mmol) was added by syringe. The reaction mixture was allowed
to stir for 2 h at room temperature, and then all volatile materials were
removed under reduced pressure to produce an orange-red solid.
This solid was dissolved in toluene (3−4 mL), layered with hexanes
(1−2 mL), and placed in a freezer (−30 °C) to crystallize. After 24 h,
X-ray quality, orange-red crystals formed, and the solvent was removed
by pipet. The crystals were washed with cold hexanes (3 × 1 mL);
residual hexanes were removed by brief exposure to reduced pressure
to produce complex 15 (77 mg, 0.266 mmol, 71%). Complex 15
decomposes within a few days at room temperature and should be
stored cold (−30 °C). 1H NMR (300 MHz, THF-d8, −20 °C): δ 6.43 (s,
Cp). 19F NMR (282 MHz, THF-d8, −20 °C): δ −23.2 (s, Ti-CF3).
13C{1H} NMR (75 MHz, THF-d8, −20 °C): δ 153.9 (q, 1JC−F = 381.8
Hz, Ti-CF3), 116.8 (s, Cp). IR (cm−1, Nujol mull): 618 (w), 689 (m),
827 (s), 856 (sh, w), 938 (s), 972 (s), 1018 (m), 1027 (w), 1065 (s),
1497
dx.doi.org/10.1021/om201055e | Organometallics 2012, 31, 1484−1499

Organometallics
Article
Computational Methods. All reported DFT calculations were
carried out, without any symmetry constraints, using the robust hybrid
B3LYP functional43 and the triple-ζ LACV3P**++ basis set,44 as
implemented in the Jaguar63 suite of programs. Vibrational frequencies
were calculated using analytic second derivatives, and all structures
were confirmed as minima by the absence of imaginary frequencies.
NBO/NPA calculations were done using the B3LYP-optimized
structures, as part of the Jaguar suite. QTAIM calculations were run
using the AIMAll program.64
Rheingold, A. L. J. Am. Chem. Soc. 2001, 123, 7443−7444. (g) Bowden,
A. A.; Hughes, R. P.; Lindner, D. C.; Incarvito, C. D.; Liable-Sands,
L. M.; Rheingold, A. L. J. Chem. Soc., Dalton Trans. 2002, 3245−3252.
(h) Hughes, R. P.; Ward, A. J.; Golen, J. A.; Incarvito, C. D.;
Rheingold, A. L.; Zakharov, L. N. Dalton Trans. 2004, 2720−2727.
(i) Bourgeois, C. J.; Hughes, R. P.; Husebo, T. L.; Smith, J. M.; Guzei,
I. M.; Liable-Sands, L. M.; Zakharov, L. N.; Rheingold, A. L.
Organometallics 2005, 24, 6431−6439. (j) Hughes, R. P.; Maddock, S. M.;
Guzei, I. A.; Liable-Sands, L. M.; Rheingold, A. L. J. Am. Chem. Soc. 2001,
123, 3279−3288.
ASSOCIATED CONTENT
* Supporting Information
(7) (a) Klabunde, K. J.; Key, M. S.; Low, J. Y. F. J. Am. Chem. Soc.
1972, 94, 999−1000. (b) Lagow, R. J.; Gerchman, L. L.; Jacob, R. A.;
Morrison, J. A. J. Am. Chem. Soc. 1975, 97, 518−522.
■
S
Full details of crystallographic data for complexes 12 and 15.
Optimized geometry coordinates and final ZPE-corrected
energies for all computed structures. This material is available
free of charge via the Internet at http://pubs.acs.org.
(8) (a) Morrison, J. A.; Gerchman, L. L.; Eujen, R.; Lagow, R. J.
J. Fluorine Chem. 1977, 10, 333−339. (b) Krause, L. J.; Morrison, J. A.
J. Am. Chem. Soc. 1981, 103, 2995−3001.
(9) Berthold, H. J.; Groh, G. Angew. Chem., Int. Ed. 1963, 2, 398−399.
(10) (a) Ontiveros, C. D. Ph.D. Thesis, Unversity of Chicago,
Chicago, IL, 1986. (b) Jones, W. D. Dalton Trans. 2003, 3991−3995.
(11) Taw, F. L.; Scott, B. L.; Kiplinger, J. L. J. Am. Chem. Soc. 2003,
125, 14712−14713.
(12) (a) Wilkinson, G.; Birmingham, J. M. J. Am. Chem. Soc. 1954,
76, 4281−4284. (b) Druce, P. M.; Kingston, B. M.; Lappert, M. F.;
Spalding, T. R.; Srivastava, R. C. J. Chem. Soc. A 1969, 14, 2106−2110.
(13) (a) Prakash, G. K. S.; Yudin, A. K. Chem. Rev. 1997, 97, 757−
786. (b) Singh, R. P.; Shreeve, J. M. Tetrahedron 2000, 56, 7613−7632.
(c) Prakash, G. K. S.; Hu, J. ACS Symp. Ser. 2005, 911, 16−56.
(14) (a) Huang, D.; Caulton, K. G. J. Am. Chem. Soc. 1997, 119,
3185−3186. (b) Huang, D.; Koren, P. R.; Folting, K.; Davidson, E. R.;
Caulton, K. G. J. Am. Chem. Soc. 2000, 122, 8916−8931. (c) Vicente,
J.; Gil-Rubio, J.; Bautista, D. Inorg. Chem. 2001, 40, 2636−2637.
(15) Dubinina, G. G.; Brennessel, W. W.; Miller, J. L.; Vicic, D. A.
Organometallics 2008, 27, 3933−3938.
(16) (a) In situ: Kim, J.; Shreeve, J. M. Org. Biomol. Chem. 2004, 2,
2728−2734. (b) Isolated: Dubinina, G. G.; Furutachi, H.; Vicic, D. A.
J. Am. Chem. Soc. 2008, 130, 8600−8601.
(17) (a) Vicente, J.; Gil-Rubio, J.; Guerrero-Leal, J.; Bautista, D.
Organometallics 2004, 23, 4871−4881. (b) Vicente, J.; Gil-Rubio, J.;
Guerrero-Leal, J.; Bautista, D. Organometallics 2005, 24, 5634−5643.
(18) Grushin, V. V.; Marshall, W. J. J. Am. Chem. Soc. 2006, 128,
12644−12645.
(19) (a) Naumann, D.; Kirij, N. V.; Maggiarosa, N.; Tyrra, W.;
Yagupolskii, Y. L.; Wickleder, M. S. Z. Anorg. Allg. Chem. 2004, 630,
746−751. (b) Menjon, B.; Martinez-Salvador, S.; Gomez-Saso, M. A.;
Fornies, J.; Falvello, L. R.; Martin, A.; Tsipis, A. Chem. Eur. J. 2009, 15,
6371−6382.
AUTHOR INFORMATION
Corresponding Author
*E-mail: rph@dartmouth.edu (R.P.H.), kiplinger@lanl.gov
(J.L.K.).
■
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
For financial support, we acknowledge Los Alamos National
Laboratory (Director’s PD Fellowships to F.L.T., A.E.C., A.H.M.,
and T.C.) and the LANL Laboratory Directed Research &
Development program (J.L.K.). R.P.H. acknowledges the
National Science Foundation for generous support, and
Professor Clark Landis for useful suggestions. Finally, we thank
Drs. Brady J. Gibbons and George J. Havrilla (both LANL) for
their assistance with the SEM, XRD, and XRF measurements,
and Dr. Marisa J. Monreal for her assistance with generating the
SEM images.
REFERENCES
■
(1) Drobny, J. G. Technology of Fluoropolymers; CRC Press LLC:
Boca Raton, FL, 2001.
(2) (a) Kiplinger, J. L.; Richmond, T. G.; Osterberg, C. E. Chem. Rev.
1994, 94, 373−431. (b) Richmond, T. G. Top. Organomet. Chem.
1999, 3, 243−269.
(3) Kaminsky, W.; Sinn, H. Transition Metals and Organometallics as
Catalysts for Olefin Polymerization; Springer-Verlag: New York, 1988.
(4) (a) Coffield, T. H.; Kozikowski, J.; Closson, R. D. Abstracts of the
Proceedings of the 5th International Conference on Coordination
Chemistry; Chemical Society: London, England, 1959; p 126.
(b) King, R. B. Organometallic Syntheses; Academic Press: New York,
1965; Vol. 1, pp 148−150.
(5) For general reviews, see: (a) Treichel, P. M.; Stone, F. G. A. Adv.
Organomet. Chem. 1964, 1, 143−220. (b) Bruce, M. I.; Stone, F. G. A.
Prep. Inorg. React. 1968, 4, 177−235. (c) Morrison, J. A. Adv. Inorg.
Chem. Radiochem. 1983, 27, 293−316. (d) Brothers, P. J.; Roper, W. R.
Chem. Rev. 1988, 88, 1293−1326. (e) Hughes, R. P. Adv. Organomet.
Chem. 1990, 31, 183−267. (f) Morrison, J. A. Adv. Organomet. Chem.
1993, 35, 211−239.
(20) Zopes, D.; Kremer, S.; Scherer, H.; Belkoura, L.; Pantenburg, I.;
Tyrra, W.; Mathur, S. Eur. J. Inorg. Chem. 2011, 273−280.
(21) Werkema, E. L.; Messines, E.; Perrin, L.; Maron, L.; Eisenstein,
O.; Andersen, R. A. J. Am. Chem. Soc. 2005, 127, 7781−7795.
(22) For examples, see: (a) [Cp*2Ti(CH3)(THF)]+: Bochmann,
M.; Jaggar, A. J.; Wilson, L. M.; Hursthouse, M. B.; Motevalli, M.
Polyhedron 1989, 8, 1838−1843. (b) [CpTi(CH3)2]2(C10H8):
Woehrle, T.; Thewalt, U. J. Organomet. Chem. 1993, 456, C21−C23.
(c) Cp2Ti(CH3)2: Thewalt, U.; Woehrle, T. J. Organomet. Chem.
1994, 464, C17−C19.
(23) Koola, J. D.; Roddick, D. M. Organometallics 1991, 10, 591−
597.
(24) Rogers, R. D.; Atwood, J. L.; Rausch, M. D.; Macomber, D. W.
(6) For examples of CF3I oxidative addition to d8 and d10 metals, see:
(a) King, R. B.; Treichel, P. M.; Stone, F. G. A. J. Am. Chem. Soc. 1961,
83, 3593−3597. (b) Clark, H. C.; Manzer, L. E. J. Organomet. Chem.
1973, 59, 411−428. (c) Appleton, T. G.; Berry, R. D.; Hall, J. R.;
Neale, D. W. J. Organomet. Chem. 1988, 342, 399−422. (d) Hughes,
R. P.; Overby, J. S.; Williamson, A.; Lam, K.-C.; Concolino, T. E.;
Rheingold, A. L. Organometallics 2000, 19, 5190−5201. (e) Hughes,
R. P.; Sweetser, J. T.; Tawa, M. D.; Williamson, A.; Incarvito, C. D.;
Rhatigan, B.; Rheingold, A. L.; Rossi, G. Organometallics 2001, 20,
3800−3810. (f) Hughes, R. P.; Williamson, A.; Sommer, R. D.;
J. Crystallogr. Spectrosc. Res. 1990, 20, 555−560.
(25) Vela, J.; Smith, J. M.; Yu, Y.; Ketterer, N. A.; Flaschenriem, C. J.;
Lachicotte, R. J.; Holland, P. L. J. Am. Chem. Soc. 2005, 127, 7857−
7870.
(26) (a) Doherty, N. M.; Hoffmann, N. W. Chem. Rev. 1991, 91,
553−573. (b) Caulton, K. G. New J. Chem. 1994, 18, 25−41.
(27) Cotton, F. A.; McCleverty, J. A. J. Organomet. Chem. 1965, 4,
490.
(28) Clark, G. R.; Hoskins, S. V.; Roper, W. R. J. Organomet. Chem.
1982, 234, C9−C12.
1498
dx.doi.org/10.1021/om201055e | Organometallics 2012, 31, 1484−1499

Organometallics
Article
(29) Brothers, P. J.; Burrell, A. K.; Clark, G. R.; Rickard, C. E. F.;
Roper, W. R. J. Organomet. Chem. 1990, 394, 615−642.
(30) King, R. B.; Bisnette, M. B. J. Organomet. Chem. 1964, 2, 15−37.
(31) Taylor, R. C. J. Chem. Phys. 1954, 22, 714−716.
not be determined due to significant disorder between the CH3 and
the Cl ligands.
(46) Hunter, W. E.; Hrncir, D. C.; Bynum, R. V.; Penttila, R. A.;
Atwood, J. L. Organometallics 1983, 2, 750−755.
(32) For representative examples of other titanium(IV) metallocene
complexes containing a terminal Ti−F linkage, see: (a) [1,3-(SiMe3)2-
C5H3]2TiF2, Ti−F = 1.832(2) Å; [1,2,4-(SiMe3)3-C5H2][(SiMe3)
C5H4]TiF2, Ti−F = 1.838(2) Å: Winter, C. H.; Zhou, X. X.; Heeg, M. J.
Inorg. Chem. 1992, 31, 1808−1815. (b) [1,3-(SiMe3)2-C5H3][(SiMe3)
C5H4]TiF2, Ti−F = 1.8357(9),1.855(2) Å: Winter, C. H.; Lewkebandara,
T. S.; Shui, X.; Zhou, X. X. J. Crystallogr. Spectrosc. Res. 1993, 23, 685−
687. (c) (C5Me5)2Ti(CHCH2)F, Ti−F = 1.847(3) Å: Beckhaus, R.;
Sang, J.; Oster, J.; Wagner, T. J. Organomet. Chem. 1994, 484, 179−190.
(d) [(C5H4)2-C2Me4]TiF2, Ti−F = 1.855(1) Å: Edwards, A. J.; Burke, N.
J.; Dobson, C. M.; Prout, K.; Heyes, S. J. J. Am. Chem. Soc. 1995, 117,
4637−4653. (e) [(C5H4)2-CMe2]TiF2, Ti−F = 1.8565(10), 1.8648(9)Å:
Picka, M.; Cisarova, I.; Vinklarek, J.; Erben, M. Acta Crystallogr., Sect. E:
Struct. Rep. Online 2005, E61, m1266−m1268.
(33) For representative examples of other titanium(IV) complexes
containing a single terminal Ti−OSO2CF3 linkage, see: (a) [(2-O-3,5-
Me2-C6H2CH2)3N]Ti(O3SCF3), Ti−OOTf = 2.002(2) Å: Bull, S. D.;
Davidson, M. G.; Johnson, A. L.; Robinson, D. E. J. E.; Mahon, M. F.
Chem. Commun. 2003, 1750−1751. (b) [ArNC(Me)]2CH]
Ti(Me)2(O3SCF3), Ti−OOTf = 2.044(1) Å: Basuli, F.; Bailey, B. C.;
Watson, L. A.; Tomaszewski, J.; Huffman, J. C.; Mindiola, D. J.
Organometallics 2005, 24, 1886−1906. (c) [{ArNC(Me)}2CH]Ti(
NSiMe3)(O3SCF3), Ti−OOTf = 2.0235(18) Å; [N{2-P(CHMe2)2-4-Me-
C6H3}2]Ti(NSiMe3)(O3SCF3), Ti−OOTf = 2.019(3) Å: Bailey, B. C.;
Basuli, F.; Huffman, J. C.; Mindiola, D. J. Organometallics 2006, 25,
2725−2728. (d) [{ArNC(Me)}2CH]Ti(CHSiMe3)(THF)(O3SCF3),
Ti−OOTf = 2.073(2) Å: Basuli, F.; Adhikari, D.; Huffman, J. C.; Mindiola,
D. J. J. Organomet. Chem. 2007, 692, 3115−3120.
(34) (a) (C5H4CH3)2Ti(CH2OCH3)(Cl): Erker, G.; Schlund, R.;
Krueger, C. J. Organomet. Chem. 1988, 338, C4−C6.
(b) [(CH3)2Si(C5H4)2]Ti(CH3)(Cl): Gomez, R.; Cuenca, T.; Royo,
P.; Hovestreydt, E. Organometallics 1991, 10, 2516−2518.
(c) Cp2Ti[CH2Si(CH3)3](Cl): Eisch, J. J.; Caldwell, K. R.; Werner,
S.; Krueger, C. Organometallics 1991, 10, 3417−3419.
(35) Rodulfo de Gil, E.; De Burguera, M.; Valentina Rivera, A.;
Maxfield, P. Acta Crystallogr., Sect. B 1977, B33, 578−579.
(36) Single-crystal X-ray diffraction studies were performed on
Cp2Ti(CF3)(Br) (13) and Cp2Ti(CF3)(I) (14), but accurate bond
lengths could not be determined due to significant disorder between
the CF3 and the Br/I ligands, respectively.
(47) (a) Bader, R. F. W. Atoms In Molecules: A Quantum Theory;
Oxford University Press: Oxford, England, 1994. (b) Matta, C. F.,
Boyd, R. J., Eds. The Quantum Theory of Atoms in Molecules; Wiley-
VCH: Weinheim, Germany, 2007.
(48) (a) Bader, R. F. W.; Gillespie, R. J.; MacDougall, P. J. Mol.
Struct. Energ. 1989, 11, 1−51. (b) Popelier, P. L. A. Coord. Chem. Rev.
2000, 197, 169−189.
(49) (a) Henn, J.; Ilge, D.; Leusser, D.; Stalke, D.; Engels, B. J. Phys.
́
Chem. A 2004, 108, 9442−9452. (b) Kovacs, A.; Esterhuysen, C.;
Frenking, G. Chem. Eur. J. 2005, 11, 1813−1825.
(50) (a) Cremer, D.; Kraka, E. Angew. Chem., Int. Ed. Engl. 1984, 23,
627−628. (b) Cremer, D.; Kraka, E. Croat. Chem. Acta 1985, 57,
1259−1281. (c) Kraka, E.; Cremer, D. Chemical Implication of Local
Features of the Electron Density Distribution. In Theoretical Models of
Chemical Bonding: The Concept of the Chemical Bond; Maksic, Z. B.,
Ed.; Springer-Verlag: New York, 1990.
(51) (a) Glendening, E. D.; Badenhoop, J. K.; Reed, A. K.; Carpenter,
J. E.; Bohmann, J. A.; Morales, C. M.; Weinhold, F. NBO 5.0;
Theoretical Chemistry Institute, University of Wisconsin: Madison,
WI, 2001. (b) Reed, A. E.; Curtiss, L. A.; Weinhold, F. Chem. Rev.
1988, 88, 899−926. (c) Weinhold, F.; Landis, C. R. Valency and
Bonding: A Natural Bond Order Donor-Acceptor Perspective; Cambridge
University Press: Cambridge, U.K., 2005.
(52) Mercier, H. P. A.; Moran, M. D.; Schrobilgen, G. J.; Suontamo,
R. J. J. Fluorine Chem. 2004, 125, 1563−1578.
(53) Politzer, P.; Murray, J. S. General and Theoretical Aspects of the
C−X Bonds (X = F, Cl, Br, I): Integration of Theory and Experiment.
In The Chemistry of Functional Groups (Supplement D2); Patai, S.,
Rappoport, Z., Eds.; John Wiley & Sons: New York, 1995.
(54) Landis, C. R.; Weinhold, F. J. Comput. Chem. 2007, 28, 198−
203.
(55) Wu, Y.-D.; Peng, Z.-H.; Chan, K. W. K.; Liu, X.; Tuinman, A. A.;
Xue, Z. Organometallics 1999, 18, 2081−2090.
(56) (a) Huang, H.; Hurubeanu, N. R.; Bourgeois, C. J.; Cheah, S.-M.;
Yuan, J.; Rheingold, A. L.; Hughes, R. P. Can. J. Chem. 2009, 87, 151−
160. (b) Leyssens, T.; Peeters, D. J. Org. Chem. 2008, 73, 2725−2730.
(c) Leyssens, T.; Peeters, D.; Orpen, A. G.; Harvey, J. N. New J. Chem.
2005, 29, 1424−1430. (d) Leyssens, T.; Peeters, D.; Orpen, A. G.;
Harvey, J. N. Organometallics 2007, 26, 2637−2645.
(57) SMART 5.632; Bruker AXS, Inc.: Madison, WI, 2005.
(58) SAINT 6.45; Bruker AXS, Inc.: Madison, WI, 2003.
(37) (a) Wilford, J. B.; Stone, F. G. A. Inorg. Chem. 1965, 4, 389−
393. (b) Clark, H. C.; Tsai, J.-H. J. Organomet. Chem. 1967, 2, 515−
517. (c) Cotton, F. A.; Wing, R. M. J. Organomet. Chem. 1967, 9, 511−
517.
(59) Sheldrick, G. SADABS 2.05; University of Gottingen: Gottingen,
̈
̈
Germany, 2002.
(60) SHELXTL 5.10; Bruker AXS, Inc.: Madison, WI, 1997.
(61) APEX II 1.08; Bruker AXS, Inc.: Madison, WI, 2004.
(62) SAINT+ 7.06; Bruker AXS, Inc.: Madison, WI, 2003.
(38) Graham, W. A. G. Inorg. Chem. 1968, 7, 315−321.
(39) Hall, M. B.; Fenske, R. F. Inorg. Chem. 1972, 11, 768−775.
(40) (a) Bent, H. A. Chem. Rev. 1961, 61, 275−311. (b) Kaupp, M.
Chem. Eur. J. 1999, 5, 3631−3643.
(63) Jaguar 7.0−7.7; Schrodinger, LLC: New York, 2007−2010.
̈
(64) Keith, T. A. AIMAll, version 11.10.16; TK Gristmill Software:
Overland Park, KS, 2011. aim@tkgristmill.com.
(41) Loizou, D. C.; Castillo, J.; Oki, A. R.; Hosmane, N. S.; Morrison,
J. A. Organometallics 1992, 11, 4189−4193.
(42) Oberhammer, H. J. Fluorine Chem. 1983, 23, 147−162.
(43) (a) Lee, C.; Yang, W.; Parr, R. G. Phys. Rev. B: Condens. Matter
Mater. Phys. 1988, 37, 785−789. (b) Becke, A. D. J. Chem. Phys. 1993,
98, 5648−5652. (c) Becke, A. D. J. Chem. Phys. 1993, 98, 1372−1377.
(d) Stephens, P. J.; Devlin, F. J.; Chabalowski, C. F.; Frisch, M. J.
J. Phys. Chem. 1994, 98, 11623−11627. (e) Andersson, M. P.; Uvdal,
P. J. Phys. Chem. A 2005, 109, 2937−2941.
(44) (a) Dunning, T. H.; Hay, P. J. In Applications of Electronic
Structure Theory; Schaefer, H. H., III, Ed.; Modern Theoretical
Chemistry; Plenum: New York, 1977; Vol. 4. (b) Hay, P. J.; Wadt, W. R.
J. Chem. Phys. 1985, 82, 299−310. (c) Hay, P. J.; Wadt, W. R.
J. Chem. Phys. 1985, 82, 270−283. (d) Wadt, W. R.; Hay, P. J. J. Chem.
Phys. 1985, 82, 284−298.
(45) Numerous single-crystal X-ray diffraction studies were
performed on Cp2Ti(CH3)(Cl), but accurate bond lengths could
1499
dx.doi.org/10.1021/om201055e | Organometallics 2012, 31, 1484−1499