The nature of aryloxide and arylsulfide ligand bonding in dimethyltitanium complexes containing cyclopentadienyl ligation

Article abstract of DOI:10.1039/b412455c

A series of dimethyltitanium compounds [CpTi(EAr)Me₂] (E = O, S) ligated by one cyclopentadienyl (Cp) and one aryloxide (OAr) or arylsulfide (SAr) have been structurally characterized in order to gain a better understanding of aryloxide and arylsulfide bonding in these systems. Experimental structures were compared to those predicted by density functional theory (DFT). Bonding in the arylsulfide systems was found to be significantly different from bonding in the aryloxide systems. The aryloxide ligands exhibited wide Ti-O-Ar angles (≥ 150°) with the Ar group oriented proximal to the Cp group. DFT computations revealed two conformers for the arylsulfide systems. Arylsulfides with the Ar group proximal to the Cp group had a predicted Ti-S-Ar angle of ～120° while those with the Ar group distal to the Cp group had a measured and predicted Ti-S-Ar angle of ～100°. Molecular and natural bond orbital (NBO) analyses were employed to explain the nature of ligand bonding in these systems. The Royal Society of Chemistry 2005.

Full text of DOI:10.1039/b412455c

View Article Online / Journal Homepage / Table of Contents for this issue
F U L L P A P E R
The nature of aryloxide and arylsulfide ligand bonding in
dimethyltitanium complexes containing cyclopentadienyl ligation†
Thomas A. Manz,a Andrew E. Fenwick,b Khamphee Phomphrai,b Ian P. Rothwell‡b and
Kendall T. Thomson*a
a School of Chemical Engineering, Purdue University, West Lafayette, IN 47907, USA.
E-mail: thomsonk@purdue.edu
b Department of Chemistry, Purdue University, West Lafayette, IN 47907, USA
Received 12th August 2004, Accepted 17th December 2004
First published as an Advance Article on the web 21st January 2005
A series of dimethyltitanium compounds [CpTi(EAr)Me2] (E = O, S) ligated by one cyclopentadienyl (Cp) and one
aryloxide (OAr) or arylsulfide (SAr) have been structurally characterized in order to gain a better understanding of
aryloxide and arylsulfide bonding in these systems. Experimental structures were compared to those predicted by
density functional theory (DFT). Bonding in the arylsulfide systems was found to be significantly different from
bonding in the aryloxide systems. The aryloxide ligands exhibited wide Ti–O–Ar angles (≥150◦) with the Ar group
oriented proximal to the Cp group. DFT computations revealed two conformers for the arylsulfide systems.
Arylsulfides with the Ar group proximal to the Cp group had a predicted Ti–S–Ar angle of ∼120◦ while those with
the Ar group distal to the Cp group had a measured and predicted Ti–S–Ar angle of ∼100◦. Molecular and natural
bond orbital (NBO) analyses were employed to explain the nature of ligand bonding in these systems.
type [CpTi(OAr)R2] (R = Me, CH2Ph) have recently been
Introduction
reported.5,9 A detailed study of the formation and decomposi-
tion of monomethyl cationic derivatives formed when the di-
methyl compounds were activated by a Lewis acidic borane
was included in the study. These cationic derivatives have been
demonstrated to function as single-site catalysts for the poly-
merization of olefins such as 1-hexene and styrene.10
Cyclopentadienyl (Cp) ligation is ubiquitous in group 4
organometallic chemistry. Since the introduction of Cp2TiCl2
in 1953,1,2 the broad applicability of titanocenes has been
studied for derivatives containing a range of alkyl, aryl, and
heteroatom-substituted Cp rings, as well as ansa-titanocenes
that have the two Cp rings linked by a bridging group.3 The
innovation of modified titanocenes has motivated research
into the development of mono-Cp derivatives that contain a
different anionic ligand. Non-bridged half-metallocene group
4 transition metal complexes of the type, CpꢀM(L)X2 (Cpꢀ =
cyclopentadienyl group; M = Ti, Zr, Hf; L = anionic ligand
such as OAr, NR2, NPR3, etc.; X = halogen, alkyl) have
been studied, and many complement metallocene-type and
“constrained geometry” systems.3 In addition, subtle changes
within specific ligand types have been demonstrated to effect a
significant change in the environment around the metal center,
as observed in the reactivity.4
Bis(aryloxide) titanium and zirconium compounds of the
type [(ArO)2MR2] (R = Me, CH2Ph) have been thoroughly
studied.5,6 The aryloxide ligand can adopt a bonding motif
isolobal with that of cyclopentadienyl, potentially bonding in a
r2p4 fashion. A key property of these ligands is their tunability,
as a wide collection of phenols offering a unique and diverse set
of steric and electronic alternatives are commercially available
or can be readily synthesized. This has provided both novel and
complementary chemistry to the well-studied metallocenes.
The introduction of these compounds has been fol-
lowed by the development of mixed-cyclopentadienyl/aryloxide
[CpTi(OAr)Cl2] systems, which have been structurally com-
pared to the related metallocene [Cp2TiCl2] and bis(aryloxide)
[(ArO)2TiCl2] systems.7,8 In addition, dialkyl derivatives of the
To extend our study of the effects of aryloxide on structure
and reactivity, we began investigating the related arylsulfide
chemistry.11 Many substituted arylthiols are known and can
function as ligands that meet a variety of steric and electronic
requirements.12–17 There are several prior studies involving tita-
nium compounds that contain arylsulfide ligands, and several
crystal structures have been reported. Given the abundance of
CpTi(OAr)R2 complexes, we are not aware of any examples
of CpTi(SAr)R2 analogues being synthesized or characterized.
Herein, some CpTi(SAr)Me2 complexes have been synthesized
and X-ray crystal structures have been reported. DFT calcula-
tions have been performed and Ti–S bonding is compared with
CpTi(OAr)Me2 complexes.
Experimental
General details
All operations were carried out under a dry nitrogen atmo-
sphere using standard Schlenk techniques. The hydrocarbon
solvents were purified employing Grubbs-type column systems
manufactured by Innovative Technologies, and were stored over
sodium ribbons under nitrogen until use. The synthesis and
X-ray structure of [CpTi(OC6H2Me2-2,6-Br-4)Me2] has been
previously reported,9 and a similar procedure was followed
for the preparation of the arylsulfide analogues. The syntheses
of [CpTiCl3]18 and 2,4,6-trialkylbenzenethiols13 were performed
according to literature procedures or slight variations thereof.
All other reagents were purchased from Aldrich and used
without further purification. The 1H and 13C NMR spectra
were obtained on a Varian Associates Inova-300 spectrometer
and referenced to the protio impurity of commercial benzene-
d6 (C6D6) as an internal standard. Elemental analyses and
X-ray diffraction studies were performed in-house at Purdue
University.
† Electronic supplementary information (ESI) available: (1) DFT op-
timized geometries. (2) Synthesis procedure and ORTEP drawing for
compound 2. (3) Comparison of structural parameters computed
from different basis sets (6-311+G and 6-311++G**) and exchange–
correlation functionals (OLYP and B3LYP). See http://www.rsc.org/
suppdata/dt/b4/b412455c/
‡ This paper is dedicated to the memory of Ian P. Rothwell, Richard
B. Moore Distinguished Professor of Chemistry.
6 6 8
D a l t o n T r a n s . , 2 0 0 5 , 6 6 8 – 6 7 4
T h i s j o u r n a l i s
T h e R o y a l S o c i e t y o f C h e m i s t r y 2 0 0 5
©

View Article Online
Table 1 Crystal data and data collection parameters
Compound
[CpTi(SC6H2Me3-2,4,6)Me2]
[CpTi(SC6H2Pri3-2,4,6)Me2]
Formula
Mr
C16H22STi
294.32
C22H34STi
378.48
¯
Space group
P1 (no. 2)
6.947(1)
7.887(1)
15.038(2)
75.487(8)
80.377(7)
80.912(9)
780.6(2)
2
1.252
150.
Mo-Ka (0.71073)
0.045
0.123
C2 (no. 5)
32.440(3)
8.3708(3)
15.978(1)
90
102.68(3)
90
4232.9(5)
8
1.188
150.
Mo-Ka (0.71073)
0.077
0.187
˚
a/A
˚
b/A
˚
c/A
a/◦
b/◦
c /◦
3
˚
V/A
Z
Dc/g cm−3
T/K
˚
Radiation (k/A)
R
Rw
[CpTi(SC6H2Me3-2,4,6)Me2] (1)
CCD. Lorentz and polarization corrections were applied to the
data.24 An empirical absorption correction using SCALEPACK
was applied.25 Intensities of equivalent reflections were averaged.
The structure was solved using the structure solution program
PATTY in DIRDIF92.26 The remaining atoms were located
in succeeding difference Fourier syntheses. Hydrogen atoms
were included in the refinement but restrained to ride on the
atom to which they are bonded. The structure was refined
in full-matrix least-squares where the function minimized was
A suspension of [CpTiCl3] (2.00 g, 9.12 mmol in 50 mL Et2O) was
cooled to −78 ◦C in a dry ice/acetone bath. To this suspension
was added [LiMe] (17.1 mL, 1.6 M in diethyl ether, 27.4 mmol)
via syringe under a flush of nitrogen. After stirring for approx-
imately 4 h, a solution of 2,4,6-trimethylbenzenethiol (1.39 g,
9.12 mmol in 20 mL Et2O) was added dropwise at −78 ◦C. The
mixture was allowed to slowly warm to room temperature and
was stirred overnight. The solvent was removed under vacuum,
and benzene was added to the residue. The suspension was
filtered through a plug of Celite over fritted glass to remove the
lithium salts. The filtrate was then evacuated to dryness yielding
a dark yellow oil. The oil was dissolved in 5 mL hexane and
cooled to −20 ◦C, yielding X-ray quality crystals (1.89 g, 71%).
Anal. Calc. for C16H22STi: C, 65.30; H, 7.54; S, 10.90. Found:
2
2
Rw(|Fo| − |Fc| )2 and the weight w is defined as w = 1/[r2(Fo2) +
2
2
(0.0585P)2 + 1.4064P] where P = (Fo + 2Fc )/3. Scattering
factors were taken from the International Tables for Crystallogra-
phy.27 Refinement was performed on an AlphaServer 2100 using
SHELX-97.28 Crystallographic drawings were created using
ORTEP-3 for Windows version 1.076.29 The crystal structure of
compound 2 is disordered at the para-isopropyl group. Default
SHELX restraints were used in the refinement. The largest
C, 65.06; H, 7.40; S, 10.62%. H NMR (C6D6, 25 ◦C): d 7.01
1
(s, m-H); 5.96 (s, Cp); 2.48 (s, o-Me); 2.20 (s, p-Me); 0.97
(s, Ti-Me). 13C NMR (C6D6, 25 ◦C): d 140.7, 137.0, 136.2, 129.7
(aromatics); 114.4 (Cp); 66.8 (Ti-Me); 23.5 (o-Me); 21.1 (p-Me).
A similar procedure was used to synthesize [CpTi(SC6H2Pri3-
2,4,6)Me2] (2). Full details are found in the ESI.†
˚
residual electron density peak is located 1.48 A from C(112).
CCDC reference numbers 247467 and 247468.
See http://www.rsc.org/suppdata/dt/b4/b412455c/ for cry-
stallographic data in CIF or other electronic format.
Results and discussion
Density functional theory calculations
Synthesis and characterization
DFT calculations were performed in Gaussian 0319 using
the B3LYP exchange–correlation functional, the 6-311++G**
basis set, auto density fitting, the gdiis geometry optimization
algorithm, and default convergence criteria.20 Dependence of
the calculated energies and geometries on the basis set size and
exchange–correlation functional were also investigated. Ener-
gies and geometries using the OLYP21–23 exchange–correlation
functional and 6-311+G basis set are reported in the ESI† for
comparison purposes. Use of the OLYP exchange–correlation
functional was found to provide significant speed-up compared
to the B3LYP functional while still providing good accuracy. For
the energy minimum and transition state structures, forces were
converged to better than 0.00045 a.u. and atom displacements
were converged to better than 0.0018 a.u. For the constrained
Ti–E–Ar angle calculations, forces were converged to better
than 0.0017 a.u. and atom displacements were converged to
better than 0.01 a.u. The QST3 method was utilized to find the
transition state for interconversion between proximal and distal
conformations of [CpTi(SC6H2Me3-2,4,6)Me2].
Synthesis of the mixed-cyclopentadienyl/arylsulfide com-
pounds was attempted using two methods employed suc-
cessfully for the aryloxide analogues. In the first method,
LiSAr was added to a solution of CpTiCl3 in benzene giving
CpTi(SAr)Cl2. A subsequent methylation of the corresponding
dichloride CpTi(SAr)Cl2 with two equivalents of methyllithium
or MeMgCl gave CpTi(SAr)Me2 (Scheme 1(a)). Although the
products typically could be observed in the reaction mixture
vide infra, attempts to isolate pure products from the reaction
mixture were unsuccessful, yielding dark brown, viscous residues
regardless of reaction solvent and temperature. In the second and
preferred method, the dimethyl complexes CpTi(SAr)Me2 were
synthesized directly from CpTiCl3 using a one-pot procedure
X-Ray data collection and reduction
Crystal data and data collection parameters are contained in
Table 1. A suitable crystal was mounted on a glass fiber in a
random orientation under a cold stream of dry nitrogen. Pre-
liminary examination and final data collection were performed
˚
with Mo-Ka radiation (k = 0.71073 A) on a Nonius Kappa
Scheme 1
D a l t o n T r a n s . , 2 0 0 5 , 6 6 8 – 6 7 4
6 6 9

View Article Online
Table 2 Structural parameters for [CpTi(EAr)X2]; EAr = OAr, SAr; X = Cl, Me
E–Ti–X/◦
X–Ti–X/◦ Cp–Ti–X/◦
Ti–E–C/◦
˚
˚
˚
Compound
Ti–X/A
Ti–Cp/A
Ti–E/A
[CpTi(OC6H2Me2-2,6-Br-4)Me2]a
[CpTi(OC6H3Pri2-2,6)Me2]a
[CpTi(SC6H2Me3-2,4,6)Cl2]b
[CpTi(SC6H2Pri3-2,4,6)Cl2]b
[CpTi(SC6H2Me3-2,4,6)Me2] (1)
[CpTi(SC6H2Pri3-2,4,6)Me2] (2)
2.088(2),
2.101(2)
2.103(3),
2.105(3)
2.2522(8), 2.017(2)
2.2557(7)
2.251(1),
2.240(1)
2.105(2),
2.142(2)
2.072(9),
2.061(8)
2.046(2)
2.045(3)
1.812(1)
1.803(2)
103.64(8),
103.72(7)
104.0(1),
103.2(1)
100.80(9)
101.8(1)
101.59(3)
102.92(5)
100.2(1)
97.3(4)
112.91(9),
113.21(9)
112.4(1),
112.6(1)
118.48(8),
117.09(7)
117.2(2),
115.9(2)
113.2(1),
114.3(1)
114.9(3),
115.3(3)
150.0(1)
157.5(1)
106.55(7)
96.9(2)
2.2912(7) 106.71(3),
107.60(3)
2.287(1)
2.026(5)
2.040(3)
2.032(8)
106.98(5),
106.01(5)
2.3242(8) 109.74(7),
107.62(6)
2.325(2)
99.78(7)
102.5(2)
109.3(3),
108.5(3)
a Data taken from ref. 9. b Data taken from ref. 11.
◦
in which [CpTiMe3] was generated in situ at −78 C, followed
by the addition of HSAr at −78 ◦C to form CpTi(SAr)Me2
(Scheme 1(b)). This method allowed for the isolation of pure
crystalline product in high yield. However, slow decomposition
of the isolated solid was observed over days at room temperature.
The 1H and 13C NMR spectra of 1 and 2 were uncomplicated,
with one Cp, one methyl, and a single set of arylsulfide
resonances in each case. The Cp protons were observed ap-
proximately 0.2 ppm upfield relative to their dichloro analogues,
reflecting the decrease in electrophilicity of the metal center as
chloride atoms are replaced by methyl groups.
Compounds 1 and 2 have been analyzed by X-ray crystal-
lography (Table 1) and the ORTEP drawing of compound 1 is
represented in Fig. 1. The dimethyl compounds exhibit a pseudo-
tetrahedral geometry about the metal center. The Ti–S–Ar bond
angle in 2 is bigger than in 1 likely due to the steric hindrance of
iPr groups. Selected structural parameters, along with those for
closely related [CpTi(EAr)X2] derivatives (EAr = OAr, SAr; X =
Cl, Me) are collected in Table 2. Very little difference is observed
in the structural parameters of these somewhat distinct systems.
Noteworthy is the shorter Ti–Cp distance in the dichlorides
relative to the aryloxo and arylsulfide dimethyl compounds. In
Fig.
1
ORTEP
CpTi(SC6H2Me3-2,4,6)Me2, 1.
drawing
(50%
thermal
ellipsoids)
of
addition, the Ti–S–C bond angles are significantly smaller than
the Ti–O–C bond angles in the analogous systems. The larger
Ti–S–C bond angle observed in these systems is under 107◦,
while the smaller Ti–O–C angle is 150◦.
DFT Calculation
Table 3 compares bond lengths and angles in the crystalline state
obtained by X-ray diffraction to those predicted theoretically
◦
˚
Table 3 Selected bond distances (A) and angles ( )
[CpTi(SC6H2Me3-2,4,6)Me2]b
[CpTi(OC6H2Me2-2,6-Br-4)Me2]a
1.812(1)d
[CpTi(OC6H3Me2-2,6)Me2]
Distal
proximal
Ti–Ec
2.3242(8)
[2.3506]
2.105(2)
[2.1003]
2.142(2)
[2.1003]
2.040(3)
[2.08]
[1.8153]e
2.088(2)
[2.1111]
2.101(2)
[2.1111]
2.046(2)
[2.09]
[1.8110]
[2.1121]
[2.1121]
[2.10]
[2.3431]
[2.1000]
[2.1012]
[2.06]
Ti–C(6)
Ti–C(7)
Ti–Cp
E–Ti–Cp
120.29(9)
[123.7]
103.64(8)
[103.3]
103.72(7)
[103.3]
150.0(1)
[167.0]
100.80(9)
[99.9]
112.91(9)
[111.8]
113.21(9)
[111.9]
111.1(1)
[111.5]
107.62(6)
[109.0]
109.74(7)
[109.0]
99.78(7)
[103.4]
100.2(1)
[99.5]
113.2(1)
[113.6]
114.3(1)
[113.6]
[123.4]
[103.5]
[103.5]
[165.7]
[99.9]
[120.8]
[103.1]
[103.2]
[117.1]
[100.3]
[113.4]
[113.4]
E–Ti–C(6)
E–Ti–C(7)
Ti–E–C(11)
C(6)-Ti–C(7)
C(6)-Ti–Cp
C(7)-Ti–Cp
[111.8]
[111.9]
a Data taken from ref. 9. b Conformer with ArS distal or proximal to Cp ring. c E = O or S. d Experimental values (X-ray crystallography). e DFT
calculation (B3LYP/6-311++G**).
6 7 0
D a l t o n T r a n s . , 2 0 0 5 , 6 6 8 – 6 7 4

View Article Online
from density functional theory (DFT) using the B3LYP func-
tional and 6-311++G** basis set. From the calculations, a
single configuration was observed for the aryloxide compounds,
whereas two configurations were observed for each of the
arylsulfide compounds (Scheme 2). In the proximal config-
uration, the arylsulfide group bends toward the Cp group.
In the distal configuration the arylsulfide group bends away
from the Cp group. According to the DFT calculations, the
distal conformation of compound 1 is lower in energy than
the proximal conformation by 5.25 kJ mol−1. The calculated
structural parameters for the distal conformation are in good
agreement with those found in the solid state.
2,4,6)Me2], and the aryloxide [CpTi(OC6H3Me2-2,6)Me2],
respectively. Positive values of the molecular orbital are
illustrated in red while negative values are illustrated in green
(the plots are 0.02 isosurfaces). In each of the compounds,
the z-axis is along the Ti–O or Ti–S bond, the x-axis is
perpendicular to the z-axis and lies in the plane of the phenyl
ring, and the y-axis is perpendicular to the xz-plane.
Fig. 2 HOMO of distal [CpTi(SC6H2Me3-2,4,6)Me2] showing overlap
2
of the sulfur pz orbital with the titanium dz orbital.
Scheme 2
The Ti–O–Ar bond angles were found to be very wide (≥150◦)
whereas the Ti–S–Ar bond angles were found to be slightly
less than tetrahedral (∼100◦) in the distal conformation and
greater than tetrahedral (∼120◦) in the proximal conformation.
The large difference between Ti–O–Ar and Ti–S–Ar bond angle
suggests a significant change in underlying bonding between
aryloxide and arylsulfide ligands in these systems.
˚
Additionally, the Ti–O distance (∼1.81 A) is much shorter
˚
than the Ti–S distance (∼2.33 A). This is primarily due to the
larger radius of the S atom compared to the O atom. The atomic
˚
radii of Ti, S, and O are approximately 1.40, 1.00, and 0.60 A,
respectively.30 The sum of the atomic radii of Ti and S gives 2.40,
˚
while that of Ti and O gives 2.00 A. The Ti–O and Ti–S bond
lengths were found to be shorter than would be predicted by
the sum of atomic radii. This effect is more pronounced in the
aryloxide compounds,31,32 and is believed to indicate the extent
of p-bonding to the metal center.32
A few computations were done to investigate the choice
of basis set and exchange–correlation functional on the ac-
curacy of the computed structural parameters. The OLYP
and B3LYP functionals and the 6-311+G and 6-311++G**
basis sets were used to compute the structural parameters
of distal [CpTi(SC6H2Me3-2,4,6)Me2] which were compared
to the experimental values. Use of the OLYP functional was
found to give significant speed-up compared to B3LYP with
a slight loss in accuracy. The largest change was for the Ti–S
bond distance which had values of 2.3966 (OLYP/6-311 + G),
2.3716 (OLYP/6-311++G**), 2.3717 (B3LYP/6-311+G), and
2.35058 (B3LYP/6-311++G**) compared to the experimental
Fig. 3 HOMO of proximal [CpTi(SC6H2Me3-2,4,6)Me2].
˚
bond distance of 2.3242 A. The Ti–S–C bond angle followed a
similar trend, having values of 107.4 (OLYP/6-311+G), 106.5
(OLYP/6-311++G**), 103.8 (B3LYP/6-311+G), and 103.4
(B3LYP/6-311++G**) compared to the experimental bond
angle of 99.8◦. All of the other bond distances and angles
changed by amounts smaller than these. The B3LYP functional
with 6-311++G** basis set was used for the molecular orbital
and bonding analysis presented below.
Frontier molecular orbitals
Figs. 2–4 show the HOMO of arylsulfides, distal
[CpTi(SC6H2Me3-2,4,6)Me2] and proximal [CpTi(SC6H2Me3-
Fig. 4 HOMO of [CpTi(OC6H3Me2-2,6)Me2] showing overlap of the
oxygen py orbital with the titanium dyz orbital.
D a l t o n T r a n s . , 2 0 0 5 , 6 6 8 – 6 7 4
6 7 1

View Article Online
In the distal arylsulfide HOMO (molecular orbital # 78)
shown in Fig. 2, the r-bonding between the sulfur pz orbital
2
and the coaxial dz orbital on titanium is evident. The HOMO
also includes anti-bonding interaction of the phenyl ring p2 (e1g)
orbital with s-orbitals of periphery methyl hydrogens bound
to the phenyl ring. The methyl groups bound to the titanium
center have no contribution to the HOMO. The HOMO of the
proximal arylsulfide complex (see Fig. 3) in general is similar to
the distal conformation except for a change in Ti–S–Ar bond
2
angle which causes the titanium dz orbital and the overlapping
sulfur p-orbital to have different axes.
In the aryloxide HOMO (molecular orbital #70, Fig. 4), the
interaction between the oxygen py orbital with the coplanar
titanium dyz orbital is clearly p-bonding. As in the case of the
arylsulfide, the HOMO also includes contributions from p2 (e1g)
phenyl ring orbital and periphery methyl s-orbitals bound to the
phenyl ring. The HOMO includes a very weak interaction with
the cyclopentadienyl ring and the methyl groups bound to the
titanium center.
Fig. 5 Molecular orbital #64 of [CpTi(OC6H3Me2-2,6)Me2] showing
overlap of the oxygen px orbital with the titanium dxz orbital.
For the aryloxide, additional O–Ti bonding interactions
manifest as one looks at lower energy Kohn–Sham orbitals.
In particular, a p-bonding orbital derived from the oxygen px
interaction with Ti dxz is observed slightly below the Fermi level
(orbital #64, Fig. 5), and much deeper in the eigenspectrum
(orbital #42, Fig. 6) r-bonding is evident by the oxygen pz
2
overlap with the coaxial Ti dz . On the other hand in the
2
arylsulfide, aside from the pz–dz r-bonding in the arylsulfide
HOMO, there is a very weak px–dxz p-bonding contribution in a
slightly lower energy orbital (orbital #76, Fig. 7) for which there
is a strong localization around the S atom, leading us to suspect
this state is better characterized as a lone pair state. Another S
lone pair orbital, py, is shown in Fig. 8 (orbital #73).
Bonding analysis
The molecular orbital analysis above provides a bonding picture
whereby Ti–arylsulfide bonds are best represented as a single r-
bond with two lone-pair orbitals. Ti–aryloxide bonds seem to
be best represented as a triple bond; however, there is a clear
preference for electron density to localize on oxygen, suggesting
an ionic nature to the interaction. There is still considerable
debate as to the representation of the aryloxide bond as covalent
or ionic.32,33 We can extend our DFT analysis to gain a more
fundamental insight into the relative importance of ionic vs.
covalent ligand bonding character in titanium aryloxide and
arylsulfide complexes. As a first step, we computed the ionic and
covalent dissociation energies as in Table 4:
Fig. 6 Molecular orbital #42 of [CpTi(OC6H3Me2-2,6)Me2] showing
2
overlap of the oxygen pz orbital with the titanium dz orbital.
Dissociation
DE/kJ mol−1
[CpTi(OC6H3Me2-2,6)Me2]
309
761
229
667
ꢀ [CpTiMe2] + [(OC6H3Me2-2,6)]
[CpTi(OC6H3Me2-2,6)Me2]
ꢀ [CpTiMe2]+ + [(OC6H3Me2-2,6)]−
[CpTi(SC6H2Me3-2,4,6)Me2]
ꢀ [CpTiMe2] + [(SC6H2Me3-2,4,6)]
[CpTi(SC6H2Me3-2,4,6)Me2]
ꢀ [CpTiMe2]+ + [(C6H2Me3-2,4,6)]−
These results show that covalent dissociation of the metal–
ligand bond into free radicals is much more favorable than ionic
dissociation, suggesting that the dissociation of the arylsulfide
or aryloxide ligand from the titanium center preferably produces
a Ti(III) rather than a Ti(IV) complex.
We also computed the variation of energy as a function
of Ti–E–Ar bond angle for [CpTi(SC6H2Me3-2,4,6)Me2] and
[CpTi(OC6H3Me2-2,6)Me2] (Figs. 9 and 10). These were com-
puted by constraining the Ti–E–Ar bond angle and allowing
all other structural parameters to relax. In the case of covalent
bonding, a Ti–E–C bond angle should reflect the underlying
molecular orbital hybridization: ∼109◦ for sp3 hybridization,
Fig. 7 Molecular orbital #76 of distal [CpTi(SC6H2Me3-2,4,6)Me2]
showing overlap of the sulfur px orbital with the titanium dxy orbital.
∼120◦ for sp2 hybridization, and ∼180◦ for sp hybridization.
The bond angles for the Ti–S–Ar and Ti–O–Ar compounds
suggest sulfur sp3 and oxygen sp hybridization.
We further computed the transition state for conversion
from the proximal to distal conformation of [CpTi(SC6H2Me3-
2,4,6)Me2]. The corresponding activation barrier was found to
be 29 kJ mol−1 with a Ti–S–Ar angle of 132.0◦ and a Cp–Ti–S–C
dihedral angle of −88.2◦. We note that conversion of proximal
to distal conformation along the lowest energy pathway involves
6 7 2
D a l t o n T r a n s . , 2 0 0 5 , 6 6 8 – 6 7 4

View Article Online
observed vide infra (1H NMR) even at temperatures as low as
−70 ◦C. Perhaps the energies of the two conformations are
not as close as the calculations suggest leading one form to
dominate in solution, or perhaps the interconversion between
the two conformations was too rapid for NMR to distinguish
them at −70 ◦C.
The overall order of each titanium-ligand bond was computed
in Gaussian 03 using the Wiberg bond index array,34 for which
we quote, “the elements of this array are the sums of squares
of off-diagonal density matrix elements between pairs of atoms
in the NAO basis, and are the NAO counterpart of the Wiberg
bond index.”35
The Wiberg method for [CpTi(OC6H3Me2-2,6)Me2] gives a
Ti–O bond index of 0.76, an O–Ar bond index of 1.00, a Ti–Me
bond index of 0.80, and a Ti–C bond index of 0.19–0.22 for each
of the cyclopentadienyl carbons giving a Ti–Cp bond index of
1.00. Overall, it gives a total bond index of 3.66 for Ti, which
is consistent with the observation that Ti has four electrons to
share.
Fig. 8 Non-bonding type interaction between sulfur py and titanium
in molecular orbital #73 of distal [CpTi(SC6H2Me3-2,4,6)Me2].
The Wiberg method for distal [CpTi(SC6H2Me3-2,4,6)Me2]
gives a Ti–S bond index of 1.05, a S–Ar bond index of 1.02, a
Ti–Me bond index of 0.79, and a Ti–C bond index of 0.20–0.24
for each of the cyclopentadienyl carbons giving a Ti–Cp bond
index of 1.09. Overall, we observe a total bond index of 4.02 for
Ti, which is also consistent with the observation that Ti has four
electrons to share.
Natural bond order (NBO)35 analysis was used to generate
Lewis structures for the above compounds. We generated the
Lewis structure of Fig. 11(a) for the arylsulfide compounds
while we generated the ionic form of Fig. 11(b) for the aryloxide
compound.
Taking all of the above information into account, a consistent
picture emerges for the differences in bonding between titanium
arylsulfide and aryloxide complexes. The arylsulfides can be
adequately represented by the Lewis structure of Fig. 11(a).
This structure predicts a Ti–S–Ar angle of approximately 109◦
and a Ti–S bond order of one. We also predicted the two
different configurations, proximal and distal, that the molecule
can adopt. On the other hand, the aryloxides cannot be
adequately represented by a single Lewis structure, but may be
represented by the resonance forms depicted in Fig. 11(b). Note
that this resonance predicts a nearly linear Ti–O–Ar bond angle,
interaction through all three oxygen p orbitals, and an overall
Ti–O covalent bond index of one.
Fig. 9 Variation of energy as a function of Ti–S–Ar angle.
These observations can be rationalized as following based on
the molecular orbitals. The sulfur and oxygen atoms each have
three p orbitals which may overlap with d orbitals on titanium.
In the case of sulfur, one of the interactions is strong (Figs. 2–
3), while two of the interactions are very weak (Figs. 7–8). In
the case of oxygen, all three of the interactions are moderately
weak (Figs. 4–6). As a result, the geometry in the arylsulfide
complexes changes to optimize the strength of the one strong
interaction at the expense of the two very weak interactions
giving a bent Ti–S–Ar. In the case of aryloxide, the geometry
optimizes to simultaneously optimize the strength of all three
moderately weak interactions.
Fig. 10 Variation of energy as a function of Ti–O–Ar angle.
rotation about the Ti–S bond and therefore does not pass
through a Ti–S–Ar angle of 180◦. Computed geometries for
the transition state and several other structures are available
Our experience with computations of the geometries of similar
titanium aryloxide and arylsulfide complexes have indicated
similar bonding to the compounds presented here. Cationic
complexes [Ti(OAr)2Me+] and [CpTi(OAr)Me+], cation-anion
1
in the ESI.† Variable temperature H NMR experiments were
performed in an attempt to observe the interconversion between
proximal and distal conformers. However, this process was not
Fig. 11 Lewis structures of (a) CpTi(SAr)Me2 and (b) CpTi(OAr)Me2.
D a l t o n T r a n s . , 2 0 0 5 , 6 6 8 – 6 7 4
6 7 3

View Article Online
complexes [Ti(OAr)2Me+] [MeB(C6F5)3−] and [CpTi(OAr)Me+]
[MeB(C6F5)3−], and neutral complexes [Ti(OAr)2Me2] and
[CpTi(OAr)Me2] have all been found to have similar metal ligand
bonding, though small differences occur in bond distances and
angles.
10 K. Pomphrai, A. E. Fenwick, S. Sharma, J. M. Caruthers, W. N.
Delgass and I. P. Rothwell, in preparation.
11 A. E. Fenwick, P. E. Fanwick and I. P. Rothwell, Organometallics,
2003, 22, 535.
12 P. T. Bishop, J. R. Dilworth, T. Nicholson and J. A. Zubieta, J. Chem.
Soc., Dalton Trans., 1991, 385.
13 P. J. Blower, J. R. Dilworth, J. Hutchinson, T. Nicholson and J. A.
Zubieta, J. Chem. Soc., Dalton Trans., 1985, 2639.
14 K. Ruhlandt-Senge and P. P. Power, Bull. Soc. Chim. Fr., 1992, 129,
594.
15 R. R. Schrock, M. Wesolek, A. H. Liu, K. C. Wallace and J. C.
Dewan, Inorg. Chem., 1988, 27, 2050.
16 J. J. Ellison, K. Ruhlandt-Senge and P. P. Power, Angew. Chem., Int.
Ed. Engl., 1994, 33, 1178.
17 K. Look and R. K. Norris, Aust. J. Chem., 1999, 52, 1077.
18 F. A. Cotton and W. A. Wojtczak, Gazz. Chim. Ital., 1993, 123, 499.
19 M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria, M. A.
Robb, J. R. Cheeseman, J. J. A. Montgomery, T. Vreven, K. N. Kudin,
J. C. Burant, J. M. Millam, S. S. Iyengar, J. Tomasi, V. Barone, B.
Mennucci, M. Cossi, G. Scalmani, N. Rega, G. A. Petersson, H.
Nakatsuji, M. Hada, M. Ehara, K. Toyota, R. Fukuda, J. Hasegawa,
M. Ishida, T. Nakajima, Y. Honda, O. Kitao, H. Nakai, M. Klene, X.
Li, J. E. Knox, H. P. Hratchian, J. B. Cross, C. Adamo, J. Jaramillo,
R. Gomperts, R. E. Stratmann, O. Yazyev, A. J. Austin, R. Cammi,
C. Pomelli, J. W. Ochterski, P. Y. Ayala, K. Morokuma, G. A. Voth,
P. Salvador, J. J. Dannenberg, V. G. Zakrzewski, S. Dapprich, A. D.
Daniels, M. C. Strain, O. Farkas, D. K. Malick, A. D. Rabuck, K.
Raghavachari, J. B. Foresman, J. V. Ortiz, Q. Cui, A. G. Baboul,
S. Clifford, J. Cioslowski, B. B. Stefanov, G. Liu, A. Liashenko, P.
Piskorz, I. Komaromi, R. L. Martin, D. J. Fox, T. Keith, M. A. Al-
Laham, C. Y. Peng, A. Nanayakkara, M. Challacombe, P. M. W. Gill,
B. Johnson, W. Chen, M. W. Wong, C. Gonzalez, and J. A. Pople, in
Gaussian 03, Pittsburgh PA, 2003.
Conclusions
The titanium arylsulfide ligand bond is primarily a covalent
sp3 hybridized single bond with two lone pairs residing on the
sulfur atom. DFT calculations suggest that [CpTi(SC6H2Me3-
2,4,6)Me2] can exist in two conformations as proximal and distal,
although only the distal conformation was observed in solid
state. Conversion between the two conformations occurs readily
with an activation energy barrier of approximately 29 kJ mol−1.
However, this interconversion was unobservable vide infra (1H
NMR) even at −70 ◦C.
In contrast to the titanium arylsulfide bond, the titanium
aryloxide bond is primarily sp hybridized, and three of the
oxygen lone pairs interact weakly with the titanium metal center.
This creates a bond which is a resonance between a covalent
triple bond and an ionic bond. In both the arylsulfide and
aryloxide compounds, the overall Ti–E bond order is nearly
one. And in both compounds, covalent dissociation of the bond
is energetically preferred over ionic dissociation.
Acknowledgements
Support for this research was provided by the U.S. Department
of Energy, Office of Basic Energy Sciences, through the Catal-
ysis Science Grant No. DE-FG02-03ER15466. Computational
resources were obtained in part from the Purdue IBM Super-
computing Cluster—part of the Information and Technology at
Purdue (ITaP)—and through a grant from the National Center
for Supercomputing Applications (NCSA).
20 A. Frisch, M. J. Frisch, and G. W. Trucks, Gaussian 03 User’s
Reference, Gaussian Inc., 2003.
21 N. C. Handy and A. J. Cohen, Mol. Phys., 2001, 99, 403.
22 C. T. Lee, W. T. Yang and R. G. Parr, Phys. Rev. B, 1988, 37, 785.
23 J. Baker and P. Pulay, J. Chem. Phys., 2002, 117, 1441.
24 P. C. McArdle, J. Appl. Crystallogr., 1996, 29, 306.
25 Z. Otwinowski and W. Minor, Processing of X-ray Diffraction
Data Collected in Oscillation Mode, Methods in Enzymology:
Macromolecular Crystallography, Part A, ed. C. W. Carter, Jr. and
R. M. Sweet, Academic Press, New York, 1997, vol. 276, pp. 307–
326.
References
1 G. Wilkinson and J. L. Birmingham, J. Am. Chem. Soc., 1954, 76,
4281.
2 G. Wilkinson, P. L. Puason, J. L. Birmingham and F. A. Cotton,
J. Am. Chem. Soc., 1953, 75, 15.
3 A. Togni and R. L. Halterman, in Metallocenes: Synthesis, Reactivity,
Applications, Wiley-VCH, Weinheim, Germany, 1998.
4 Y. Zhang, E. K. Reeder, R. J. Keaton and L. R. Sita, Organometallics,
2004, 23, 3512.
5 M. G. Thorn, Z. C. Etheridge, P. E. Fanwick and I. P. Rothwell,
J. Organomet. Chem., 1999, 591, 148.
26 P. T. Beurskens, G. Admiraal, G. Beurskens, W. P. Bosman, S. Garcia-
Granda, R. O. Gould, J. M. Smits and C. Smykalla, The DIRDIF92
Program System; Technical Report, Crystallography Laboratory,
University of Nijmegen, Nijmegen, The Netherlands, 1992.
27 International Tables for Crystallography, Kluwer Academic, Dor-
drecht, 1992.
28 G. M. Sheldrick, SHELXS-97, Program for solution of crystal
structures, University of Gottingen, Germany, 1997.
29 L. J. Farrugia, J. Appl. Crystallogr., 1997, 30, 565.
30 J. C. Slater, J. Chem. Phys., 1964, 39, 3199.
¨
6 M. G. Thorn, Z. C. Etheridge, P. E. Fanwick and I. P. Rothwell,
Organometallics, 1998, 17, 3636.
7 K. Nomura, N. Naga, M. Miki, K. Yanagi and A. Imai,
Organometallics, 1998, 17, 2152.
8 M. G. Thorn, J. S. Vilardo, J. Lee, B. Hanna, P. E. Fanwick and I. P.
Rothwell, Organometallics, 2000, 19, 5636.
31 G. D. Smith, P. E. Fanwick and I. P. Rothwell, Inorg. Chem., 1990,
29, 3221.
32 D. C. Bradley, I. P. Mehrotra, I. P. Rothwell and A. Singh, Alkoxo
and Aryloxo Derivatives of Metals, Academic Press, 2001.
33 M. R. Russo, N. Kaltsoyannis and A. Sella, Chem. Commun., 2002,
2458.
9 A. E. Fenwick, K. Phomphrai, M. G. Thorn, J. S. Vilardo, C. A.
Trefun, B. Hanna, P. E. Fanwick and I. P. Rothwell, Organometallics,
2004, 23, 2146.
34 K. Wiberg, Tetrahedron, 1968, 24, 1083.
35 E. D. Glendening, A. E. Reed, J. E. Carpenter, and F. Weinhold,
NBO 3.0 Program Manual, University of Wisconsin, 1996.
6 7 4
D a l t o n T r a n s . , 2 0 0 5 , 6 6 8 – 6 7 4