Molecular engineering of coordination pockets in chloro-tris-phenoxo complexes of titanium(IV)

Article abstract of DOI:10.1016/j.poly.2005.12.027

The chloro-tris-phenoxo complexes [TiCl(OAr)₃] (OAr = OC₆H₄CMe₃-4 (1), OC₆H₃Me₂-2,4 (2), OC₆H₂Me₃-2,4,6 (4), OC₆H₃(CHMe₂)₂-2,6 (5), OC₆H₃(CMe₃)₂-2,4 (6) and OC₆H₄Ph-2 (8) are prepared by heating 3 equivalents of the phenol and [TiCl₄] in toluene. X-ray crystal structure determinations show that 2 is a phenoxy-bridged dimer with the ortho-methyl groups making the beginning of a pocket about the terminal chloro ligand and 6 is a tetrahedral monomer in which the pocket is more well developed by the ortho-tert-butyl groups. Both 2 and 6 react with dmbipy to give [TiCl(OAr)₃(dmbipy)] [OAr = OC₆H₃Me₂-2,4 (3) and OC₆H₃(CMe₃)₂-2,4 (7)] in which the original pocket is destroyed. Reaction of TiCl₄ with 3 equivalents of LiOC₆H₄Ph-2 in diethyl ether gives [TiCl(OC₆H₄Ph-2)₃(diethyl ether)] (9) for which an X-ray crystal structure determination shows a trigonal bipyramidal coordination geometry with the diethyl ether lying trans to the chloro ligand. The three phenoxide ligands make up the equatorial plane which takes the 2-phenyl substituent on each phenoxo ligand away from the chloro ligand resulting in a partially collapsed cavity. The tied-back analogues of 2 and 6, [TiCl{(OC₆H₂Me₂-2,4-CH₂-6) ₃N}] (11) and [TiCl({OC₆H₂(CMe₃)₂-2,4-CH ₂-6}₃N)] · diethyl ether (12), are prepared by adding (HOArCH₂)₃N [Ar = C₆H₂Me₂-2,4 and C₆H₂(CMe₃)₂ 2,4] to [TiCl(OCHMe₂)₃] in diethyl ether. An X-ray crystal structure of 12 showed a trigonal bipyramidal structure with a coordination environment about the terminal chloro ligand similar to that found in 6. Complex 12 reacts with pyridine to form the 6-coordinate complex [TiCl({OC₆H₂(CMe₃)₂-2,4-CH ₂-6}₃N)(py)] (13).

Full text of DOI:10.1016/j.poly.2005.12.027

Polyhedron 25 (2006) 2039–2054
www.elsevier.com/locate/poly
Molecular engineering of coordination pockets in
chloro-tris-phenoxo complexes of titanium(IV)
*
Alastair J. Nielson , Chaohong Shen, Joyce M. Waters
Chemistry, Institute of Fundamental Sciences, Massey University at Albany, Private Bag 102904, North Shore Mail Centre, Auckland, New Zealand
Received 6 December 2005; accepted 29 December 2005
Available online 28 February 2006
Abstract
The chloro-tris-phenoxo complexes [TiCl(OAr)3] (OAr = OC6H4CMe3-4 (1), OC6H3Me2-2,4 (2), OC6H2Me3-2,4,6 (4),
OC6H3(CHMe2)2-2,6 (5), OC6H3(CMe3)2-2,4 (6) and OC6H4Ph-2 (8) are prepared by heating 3 equivalents of the phenol and [TiCl4]
in toluene. X-ray crystal structure determinations show that 2 is a phenoxy-bridged dimer with the ortho-methyl groups making the
beginning of a pocket about the terminal chloro ligand and 6 is a tetrahedral monomer in which the pocket is more well developed
by the ortho-tert-butyl groups. Both 2 and 6 react with dmbipy to give [TiCl(OAr)3(dmbipy)] [OAr = OC6H3Me2-2,4 (3) and
OC6H3(CMe3)2-2,4 (7)] in which the original pocket is destroyed. Reaction of TiCl4 with 3 equivalents of LiOC6H4Ph-2 in diethyl ether
gives [TiCl(OC6H4Ph-2)3(diethyl ether)] (9) for which an X-ray crystal structure determination shows a trigonal bipyramidal coordina-
tion geometry with the diethyl ether lying trans to the chloro ligand. The three phenoxide ligands make up the equatorial plane which
takes the 2-phenyl substituent on each phenoxo ligand away from the chloro ligand resulting in a partially collapsed cavity. The tied-back
analogues of 2 and 6, [TiCl{(OC6H2Me2-2,4-CH2-6)3N}] (11) and [TiCl({OC6H2(CMe3)2-2,4-CH2-6}3N)] Æ diethyl ether (12), are pre-
pared by adding (HOArCH2)3N [Ar = C6H2Me2-2,4 and C6H2(CMe3)2 2,4] to [TiCl(OCHMe2)3] in diethyl ether. An X-ray crystal struc-
ture of 12 showed a trigonal bipyramidal structure with a coordination environment about the terminal chloro ligand similar to that
found in 6. Complex 12 reacts with pyridine to form the 6-coordinate complex [TiCl({OC6H2(CMe3)2-2,4-CH2-6}3N)(py)] (13).
ꢀ 2006 Elsevier Ltd. All rights reserved.
Keywords: Ti(IV); Phenoxo ligands; Coordination pockets; Crystal structures
1. Introduction
cal field, tied-back tripodal ligands have been used to good
effect either where the tie-back atom is not involved in
bonding to the metal such as in the case of bulky tris-
(pyrazoyl)borate ligands [8] or where it coordinates to the
metal as in the case of tris(carbamoyl)methyl amine [9].
In the non-biological area, tied-back triamido ligands,
[(RNCH2CH2)3N]3ꢀ, have been used to make steric pres-
sure pockets useful in the catalytic reduction of dinitrogen
[10] and non-tied back triamido complexes are also show-
ing interesting chemistry [11].
There is growing evidence that chloro-tris-phenoxo
complexes of titanium(IV), [TiCl(OAr)3], have a coordina-
tion environment that allows selectivity in organic
reactions. [TiCl(OC6H5)3], which has been known for some
time [12] has been used in regioselective and stereoselective
ring allylation of epoxides [13,14] and diastereoselective
allylation of ketones [15]. Recently, [TiCl(OC6H3Ph2-2,6)3]
The formation of transition metal complexes where
ligand systems are designed to form cavities or pockets in
which another ligand sits, is an area of growing interest
[1]. The impetus for the preparation of these metallopock-
ets has come from the desire to form small molecule com-
plexes that resemble the metal-containing pockets present
in metalloproteins and enzymes [2,3]. Presently, there is
interest in molecular engineering of metallopockets that
might be important for organic reaction catalysis in non-
aqueous media [4]. A related area is the formation of met-
alloclefts [5] for catalysis [6] and molecular recognition [7].
Where non-collapsible pockets are required in the biologi-
*
Corresponding author.
E-mail address: a.j.nielson@massey.ac.nz (A.J. Nielson).
0277-5387/$ - see front matter ꢀ 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.poly.2005.12.027

2040
A.J. Nielson et al. / Polyhedron 25 (2006) 2039–2054
was prepared and used as a catalyst precursor for epoxide
ring opening [16]. This reaction makes use of bulky 2,6 sub-
stituents to create a pocket in a similar manner to the
designer Lewis acid catalysts based on aluminium tris-phe-
noxo complexes [17]. For these reactions the influence of
local coordination environment is important and this can
be varied more easily by changing substituents on the aro-
matic ring of phenoxo ligands than it can with the alkoxo
ligands in the complexes [TiCl(OR)3] which also have an
extensive use in organic synthesis [18]. However, chloro-
tris-phenoxo complexes containing ortho-phenyl ring sub-
stituents are not well known. [TiCl(OC6H2Me3-2,4,6)3]
has been prepared [19,20] and some bis-amine adducts
formed [20]. [TiCl{OC6H3(CHMe2)2-2,6}3] has been briefly
described [20,21] and the X-ray structure of
[TiCl{OC6H3(CMe3)2-2,6}3] obtained [22]. We report here
work that is based on identifying pockets in chloro-tris-
phenoxo complexes of titanium(IV) and investigate how
this site topology may be retained.
164.5 which is upfield of that found for the mono-phenoxo
complex [TiCl3(OC6H4CMe3-4)] (d 169.7) [23] and the bis-
phenoxide complex [TiCl2(OC6H4CMe3-4)2] (d 166.3) and
is likely to represent a decrease in the p-bonding from each
phenoxo ligand when the number of these ligands increases
[24].
When [TiCl4] was reacted with 2,4-dimethylphenol using
the standard conditions, a crystalline solid analysing as
[TiCl(OC6H3Me2-2,4)3] was obtained for which an X-ray
crystal structure determination showed the dimeric phe-
noxo bridged complex 2. The coordination geometry about
each Ti is that of a trigonal bipyramid consisting of two
bridging phenoxides, two terminal phenoxides and a termi-
nal chloro ligand (Fig. 1a). Bond lengths and angles are
˚
shown in Table 1. The Ti–O(3) bond length [1.757(1) A]
is similar to that found in the mono-phenoxo complex
[TiCl3{OC6H2(CMe3)2-2,6-Me-4}] [Ti–O bond length
˚
1.750(2) A] for which a theoretical study shows p-donation
of both oxygen atom lone pairs to titanium and some p-
bonding between the oxygen and the phenyl ring p-system
[24]. The Ti–O(3)–C(31) bond angle [171.4(1)ꢁ] in 2 is more
‘linear’ than that found in [TiCl3{OC6H2(CMe3)2-2,6-Me-
4}] [163.1(1)ꢁ] for which theoretical studies show significant
changes in energy (and thus decrease in Op–Tid donation)
occur only after the Ti–O–C bond angle moves below ca.
2. Results and discussion
We have found that the best method for preparing the
chloro-tris-phenoxo complexes even when there is a rela-
tively large ortho-substituent on the phenyl ring is to simply
thermalise the phenol and [TiCl4] in toluene (Eq. (1)).
˚
160ꢁ. The Ti–O(2) bond length in 2 [1.794(2) A] is signifi-
cantly longer than the Ti–O(3) bond length and the Ti–
O(2)–C(21) bond angle decreases to 138.8(1)ꢁ reflecting
the decreased p-donation from the oxygen of this ligand.
[TiCl4] + 3ArOH ! [TiCl(OAr)3] + 3HCl
ð1Þ
OAr ¼ OC6H4CMe3-4;
˚
The Ti–O(1) bond length of 1.927(1) A and Ti–O(1)–
OC6H3Me2-2; 4;
C(11) bond angle of 126.1(1)ꢁ suggest that there is minimal
p-donation to Ti with the bridging phenoxo ligand. The
OC6H2Me3-2; 4; 6;
OC6H3ðCHMe2Þ2-2; 6;
OC6H3ðCMe3Þ2-2; 4;
OC6H4Ph-2.
˚
bond length of the terminal chloro ligand [2.269(1) A] is
appreciably longer than that observed for the two terminal
chloro ligands in the bridging phenoxo complex
[{TiCl2(OC6H4CMe3-4)(l-OC6H4CMe3-4)2}2] [2.2186(11)
˚
Obtaining very high yields relies on reaching the reaction
end-point and this can be easily determined by monitoring
the HCl exhaust gas production using amine hydrochloride
cloud formation. This has been found to be successful for
the general preparation of mono and bis-phenoxo com-
plexes of titanium [23,24].
and 2.2256(11) A] [24] as a result of the lessened need for
Clp–Tid p-donation when there is a third phenoxo ligand
present.
For chloro-tris-phenoxo complexes, the local environ-
ment of the terminal chloro ligand is of importance when
development of a pocket is considered since it is this ligand
and any replacement that will be affected by local steric
influences. Inspection of 2 and the space-filled model
(Fig. 1b) shows that the orientation of the phenyl rings in
the solid state is such that the methyl substituents in the
ortho-positions of the three phenoxo ligands associated
with each of the titanium atoms point up towards the ter-
minal chloro ligand and as such start to build a ‘wall’
around one side of the upper Ti(1)–Cl bond but do not cre-
ate a significant pocket. In fact the phenyl rings attached to
the adjacent Ti atom create more of a wall but are further
away from the chloro ligand. In solution, the influences
playing on the chloro ligand would appear to be different
from the solid state since the NMR spectra of 2 in CDCl3
solution do not reflect the dimeric structure shown by the
X-ray analysis but point to the three phenoxides being
To ascertain the features associated with a chloro-tris-
phenoxo complex in which there are no ortho-substituents
on the phenyl ring, the reaction of 4-tert-butylphenol and
[TiCl4] in a 3:1 ratio was undertaken. The production of
HCl gas was vigorous at first but the reaction took 35 h
for the last traces of HCl to be expelled. After removing
the solvent, a non-crystalline material was obtained which
failed to crystallise. Analytical data and the NMR spectra
showed that even after extended periods of drying in vacuo,
the complex retained toluene solvate and was characterised
as [TiCl(OC6H4CMe3-4)3] Æ 0.5C7H8 (1). Both the 1H NMR
and 13C{1H} NMR spectra showed the toluene as well-
defined resonances whereas the 4-tert-butylphenoxo ligand
resonances were broad suggesting solution dynamics [24].
In particular, the ipso-carbon resonance occurred at d

A.J. Nielson et al. / Polyhedron 25 (2006) 2039–2054
2041
Fig. 1. (a) ORTEP diagram of [{TiCl(OC6H3Me2-2,4)2(l-OC6H3Me2-2,4)}2] (2) showing the molecular structure. Hydrogen atoms have been omitted for
clarity. Only the major coordination site for Cl is shown. (b) Space-filling model looking down the Cl–Ti bond showing the position of the o-methyl groups
on the Ti(1) phenoxo ligands and the phenoxo phenyl groups on Ti(2).
1
equivalent. The H NMR spectrum shows a single set of
sharp resonances for the ortho and para-methyls, a multi-
plet for the two meta-phenyl ring hydrogens and the
13C{1H} NMR spectrum a single set of resonances for
the relevant phenoxide group carbons. The ipso-carbon res-
onance appears at d 164.4 which is in a position similar to
that found for 1 (d 164.5). The absence of resonances for
bridging and terminal ligands and the sharpened nature
of the resonances suggest that solution dynamics are less
important than for 1 so that the complex is likely to be
monomeric in the NMR solvent and have a tetrahedral
geometry in which the ortho-methyl groups all point up
around the Ti–Cl bond. In comparison, the resonances of
complex 1, particularly those associated with the aromatic
ring, were broad which may suggest monomer–dimer equi-
libria. We have found previously that [TiCl2(OAr)2] com-
plexes without ortho-methyl substituents are dimeric in
the solid state but are monomeric in solution. DFT calcu-
lations show that there is only a small dimerisation energy
involved (ꢀ26.2 kJ molꢀ1) [24].
Of fundamental importance in these systems is whether
the structure of any pocket is maintained in the presence of
other coordinating ligands. Addition of one or two equiv-
alents of pyridine to 2 gave what appeared to be isomeric
mixtures, which were difficult to distinguish by NMR spec-
troscopy. 4,40-Dimethyl-2,20-bipyridine (dmbipy) was used
since isomeric mixtures can be more easily identified using
this ligand [23,24]. Addition to 2 expanded the coordina-
tion number giving [TiCl(OC6H3Me2-2,4)3(dmbipy)] (3)
for which the NMR spectra showed one isomer. Based
on a 2:1 ratio of phenoxo ligand resonances and a symmet-
rical dmbipy ligand shown by a singlet for the two methyl
groups and a single set of resonances for the ring protons in
the 1H NMR spectrum and the required number of carbon
resonances in the 13C{1H} NMR spectrum, dmbipy
symmetric structure (a) rather than either of asymmetric

2042
A.J. Nielson et al. / Polyhedron 25 (2006) 2039–2054
Table 1
di-iso-propylphenol were reacted with TiCl4 using the stan-
dard conditions to form complexes in which phenoxo
ligand C–O bond rotations are less likely to occur on
account of the other ortho-position of the phenyl ring hav-
ing a larger substituent than hydrogen. [TiCl(OC6H2Me3-
2,4,6)3] (4) [19] formed as a solid which, characterised by
NMR spectroscopy, showed a single set of sharp reso-
nances for the relevant hydrogens and carbons as found
for complex 2 and with an ipso-carbon resonance in the
13C{1H} NMR spectrum at d 163.9 suggesting a mono-
meric and thus tetrahedral complex. [TiCl{OC6H3-
(CHMe2)2-2,6}3] (5) [20,21] was characterised by NMR
spectral comparison with the product formed by thermali-
sation in benzene [21] as NMR analysis of the toluene reac-
tion crude material indicated there was a significant
amount of the phenol present as well as the tetra-phenoxide
[Ti{OC6H3(CHMe2)2-2,6}4] (identified by NMR spectral
comparison with an independently prepared sample).
Based on the single set of ligand resonances in the NMR
spectra (in particular the ipso-carbon 13C resonance at d
163.1) 5 is also a tetrahedral monomer.
˚
Selected bond distances (A) and angles (ꢁ) for [{TiCl(OC6H3Me2-2,4)2-
(l-OC6H3Me2-2,4)}2] (2)
Ti–Cl(1)
Ti–O(1)
Ti–O(2)
Ti–O(3)
Ti–O(1)0
2.269(1)
1.927(1)
1.794(2)
1.757(1)
2.138(1)
Ti–O(1)–C(11)
Ti–O(2)–C(21)
Ti–O(3)–C(31)
Ti–O(1)0–C(11)0
Cl(1)–Ti–O(1)
Cl(1)–Ti–O(2)
Cl(1)–Ti–O(3)
Cl(1)–Ti–O(1)0
O(1)–Ti–O(1)0
O(1)–Ti–O(2)
O(1)–Ti–O(3)
O(2)–Ti–O(1)0
O(2)–Ti–O(3)
O(3)–Ti–O(1)0
Ti–O(1)–Ti0
126.1(1)
138.8(1)
171.4(1)
125.6(1)
121.21(7)
115.90(8)
96.86(5)
89.75(4)
71.76(6)
118.23(7)
95.89(6)
87.41(6)
98.96(7)
167.65(6)
108.24(6)
Of more importance than methyl or iso-propyl groups
for pocket development are large ortho-substituents. The
X-ray structures of [TiI{OC6H3(CMe3)2-2,6}3] [22] and
[TiCl(OC6H3Ph2-2,6)3] [16] show cavity formation and
would appear to be ideal molecules for cavity collapse stud-
ies. However, the former complex is difficult to prepare and
the ligand for the latter is expensive. We therefore turned
our attention to ligands that had a tert-butyl or phenyl sub-
stituent in only one of the phenoxide aryl ring ortho-posi-
tions. Reaction of 2,4-di-tert-butylphenol with [TiCl4]
under the standard conditions gave [TiCl{OC6H3-
(CMe3)2-2,4}3] (6) which could also be prepared in good
yield by the reaction of 3 equivalents of LiOC6H3-
(CMe3)2-2,4 with TiCl4 dissolved in diethyl ether. The com-
plex forms a crystalline mass that is analytically pure when
the solvent is removed from the reaction mixture but may
be recrystallised from petroleum spirit or diethyl ether.
An X-ray crystal structure determination showed a dis-
torted tetrahedron consisting of three terminal phenoxo
ligands and a single terminal chloro ligand (Fig. 2a), the
metal and chloro ligand lying along a threefold axis. Bond
lengths and angle are shown in Table 2. The three identical
structures (b) and (c) is preferred. We have been unable to
obtain suitable crystals to identify the absolute structure in
the isomer but with one phenoxo ligand coordinating trans-
to the chloro ligand the cavity structure about Cl is cer-
tainly destroyed.
OAr
Cl
ArO
OAr
Ti
Ti
N
N
N
N
ArO
ArO
ArO
Cl
Symmetric structure (a)
Asymmetric structure (b)
Cl
Ti
N
OAr
ArO
ArO
N
˚
Ti–O bond lengths [1.757(2) A] are similar to the Ti–O(3)
Asymmetric structure (c)
bond length in 2 which involves the most oxygen atom lone
˚
pair p-donation. The Ti–Cl bond length in 6 [2.240(2) A] is
˚
The structural features shown by the solid state struc-
ture of 2 indicate that one side of the titanium has become
enclosed to some extent by the phenoxo ligand ortho-
methyl groups so that the complex has the beginnings of
a pocket with defined local structure. A more well-defined
topology could arise if the complex were a tetrahedral
monomer as suggested by the NMR spectra. However,
analysis of the structure indicates that some rotation of
the phenyl C–O bonds in the monomer or dimer forms
may occur in solution taking the methyl groups away from
the chloro ligand. Thus, 2,4,6-trimethylphenol and 2,6-
slightly shorter than that observed in 2 [2.269(1) A] result-
ing from the greater need for electron density at the metal
with the lower coordination number. It is considerably
shorter than the three terminal Ti–Cl bonds in
[TiCl3{OC6H2(CMe3)2-2,6-Me-4}] [Ti–Cl bond length
˚
range 2.1822(8)–2.1945(9) A] in which there is only one
p-donating phenoxo ligand. In [TiCl4] where each chloride
must p-donate 2 electron pairs to attain an electron count
˚
of 16, the Ti–Cl bond lengths are 2.170(2) A [25].
The Ti–O–C bond angles in 6 [158.8(2)ꢁ] are slightly
smaller than those found in [TiCl3{OC6H2(CMe3)2-2,6-

A.J. Nielson et al. / Polyhedron 25 (2006) 2039–2054
2043
Fig. 2. (a) ORTEP diagram of [TiCl{OC6H3(CMe3)2-2,4}3] (6) showing the molecular structure. Hydrogen atoms are omitted for clarity. Only one of the
disordered positions for the tert-butyl group is shown. (b) Space-filling model looking down the Cl–Ti bond showing the pocket arrangement about Cl.
(c and d) Space-filling model, side-on view of the Ti–Cl bond showing the extent to which the tert-butyl group methyls cover the Cl atom.
Me-4}] [163.1(2)ꢁ] [23] and the Ti–O(3)–C(31) bond angle
in complex 2 [171.4(1)ꢁ] but the angle in 6 is only a little less
than the 160ꢁ for which theoretical studies with this type of
ligand show a steep rise in energy when this value is
reduced [23]. In [TiI{OC6H3(CMe3)2-2,6}3] where both
phenyl ring ortho-positions have tert-butyl substituents
the Ti–O–C bond angles range from 155.2(4)ꢁ to
158.2(4)ꢁ [22]. In 6 the O–Ti–O0 and O–Ti–Cl bond angles
are 108.90(8)ꢁ and 110.03(8)ꢁ, respectively, hence the geom-
etry does not deviate significantly from a tetrahedral
arrangement. In [TiI(OC6H3(CMe3)2-2,6)3] the O–Ti–O
bond angles range from 110.7(3)ꢁ to 114.9(3)ꢁ and the
O–Ti–I bond angle range is 102.9(3)–108.6(3)ꢁ [22].
The phenyl rings in complex 6 orientate so that two of
the methyl groups of the 2-tert-butyl group straddle the
chloro ligand but lie slightly to the side of this ligand.
The two closest methyl group carbon contacts with the
˚
chloro ligand are 3.13 and 3.11 A which, taken with the

2044
A.J. Nielson et al. / Polyhedron 25 (2006) 2039–2054
Table 2
(9) which had nearly identical NMR spectral characteristics
in the aromatic region to those shown by complex 8 with
the ipso-carbon at d 163.2. Interestingly, when complex 8
was dissolved and refluxed in diethyl ether the NMR spec-
tra of the reaction product indicated that no diethyl ether
was incorporated. This is probably related to the retention
of coordinated ether during the reaction with the titanium
starting material [TiCl4(diethylether)2]. An X-ray crystal
structure determination of 9 showed a distorted trigonal
bipyramidal coordination geometry with the diethyl ether
ligand bound in a trans position to the chloro ligand and
the three phenoxide ligands forming the equatorial plane
(Fig. 3a). Bond lengths and angle are shown in Table 3.
The Ti–O bond lengths lie in the range 1.790(1)–
˚
Selected bond distances (A) and angles (ꢁ) for [TiCl{OC6H3(CMe3)2-2,4}3]
(6)
Ti–Cl
Ti–O(1)
2.240(2)
1.757(2)
Ti–O(1)–C(1)
Cl–Ti–O(1)
O(1)–Ti–O(1)0
158.8(2)
110.03(8)
108.90(8)
Ti–O–C(1) bond angles [158.8(2)ꢁ], and the near ideal tetra-
hedral geometry, indicate that the 2-tert-butyl substituent
exerts little steric pressure. However, it is more constrained
than its counterpart in the 4-position of the phenyl ring
which is seen to occupy two disordered sites. No such dis-
order was observed for the 2-substituent. It provides some-
thing of a steric shield about the chloro ligand and thus
creates a pocket for this fourth coordination site as the
space filled models show (Fig. 2b–d). Molecular models
indicate that if the Ti–O–C bond angles are maintained
in solution near the solid state value of ca. 159ꢁ then rota-
tion about the phenyl ring C–O bond is difficult. However,
since this angle can flatten with little energy consequence
[23] rotation may be possible although the C–O p-bond
would need to be broken in this process.
The 4-coordination of complex 6 can be broken by the
addition of dmbipy but whereas the ortho-methyl complex
2 gave one product which NMR spectroscopy showed to
have a symmetric structure, reaction with 6 gave a product
analysing as [TiCl{OC6H3(CMe3)2-2,4}3(dmbipy)] (7) but
the NMR spectra showed three products (ratio ca. 72:21:7)
based on the easily distinguished features of the dmbipy
˚
1.794(1) A and, together with the Ti–Cl bond length
˚
[2.3067(6) A], are much longer than the corresponding
bonds in [TiCl{OC6H3(CMe3)2-2,4}3] (6) [Ti–O and Ti–Cl
˚
are 1.757(2) and 2.240(2) A, respectively]. This is a result
of 9 having a reduced p-loading [26] requirement compared
to 8 with the addition of the 2-electon donor diethyl ether
ligand. Consequently, the Ti–O–C bond angles of
152.3(1)ꢁ, 155.0(1)ꢁ and 157.1(1)ꢁ relax a little in compari-
son with those in 6 where Ti–O–C bond angle is 158.8(2)ꢁ.
The phenoxide oxygen atoms push away from the chloro
ligand [O(1)–Ti–Cl, 96.12(4)ꢁ; O(2)–Ti–Cl, 95.90(5)ꢁ] but
the O–Ti–O bond angles are close to the ideal trigonal
angle [the range is 117.88(6)–119.99(6)ꢁ]. The phenyl rings
are rotated into the equatorial plane to remove interaction
with the diethyl ether ligand and this takes the 2-phenyl
substituent on each phenoxo ligand away from the chloro
ligand and destroys any substantial cavity present if the
substituents all lie upright. The space-filling model looking
down the Cl–Ti bond (Fig. 3b) shows the cavity effect
which is different from that observed in [TiCl(OC6H3Ph2-
2,6)3] [16] where the substituents form an open-ended
groove (Fig. 3c). This results from a face-to-face orienta-
tion of two of the phenyl substituents and an end wall in
which the substituent has moved aside slightly. The extent
to which the phenyl groups cover the chloro ligand in 9 is
much less than in [TiCl(OC6H3Ph2-2,6)3] (compare Fig. 3d
and e) being somewhat similar to that found in
[TiCl{OC6H3(CMe3)2-2,4}3] (6) (compare Fig. 3d with
Fig. 2c).
1
hydrogens in the H NMR spectrum. The major product,
however, had spectral characteristics similar to complex 3
[symmetric structure (a)]. Of the three possible structures
only (c) has the potential for three tert-butyl groups to sur-
round Cl. We have also looked at the possibility of coordi-
nating only one extra ligand to 6 by adding 1 equivalent of
pyridine or various aldehydes but these reactions give
gummy substances that are hard to purify. However, the
NMR spectra show mainly single resonances for the phe-
noxo ligands indicating the ligands are equivalent which
offers the possibility of a 5-coordinate structure (see later).
The structural features shown by complex 6 suggest that
the chloro ligand would sit in a more effective pocket if the
2-tert-butyl substituent of the phenoxo ligands were
replaced by a 2-phenyl substituent which, although less ste-
rically demanding than the former, would protrude further
out. Reaction of 2-phenylphenol with [TiCl4] under the
standard conditions gave an oil which failed to solidify
but was identified as [TiCl(OC6H4Ph-2)3] (8) on the basis
of its NMR spectra. In particular, the more easily identified
13C{1H} NMR spectrum showed a single set of resonances
for each of the carbons of the ligand and the ipso-carbon
resonance (d 163.6) was in a similar position to the other
chloro-tris-phenoxide complexes prepared.
The structural features shown by complexes 6 and 9
indicate that the local cavity geometry is easily effected
by the addition of another ligand. We therefore looked at
tying back one of the three phenoxo ligands with a nitrogen
donor using the benzylamine ligand structure (d) to form
complexes of structural type (e).
Cl
R
R
OAr
OAr
OH
O
Ti
N
N
R
R
Me
Me
Me
Structure (d) (R = CMe3)
Me
Reaction of [TiCl4] with 3 equivalents of LiOC6H4Ph-2
in diethyl ether gave [TiCl(OC6H4Ph-2)3(diethyl ether)]
Structural type (e) (R = CMe3)

A.J. Nielson et al. / Polyhedron 25 (2006) 2039–2054
2045
Fig. 3. (a) ORTEP diagram of [TiCl(OC6H4Ph-2)3(diethyl ether)] (9) showing the molecular structure. Hydrogen atoms are omitted for clarity. Only the
major site of the coordinated ether ligand is shown. (b) Space-filling model of 9 looking down the Cl–Ti bond showing the collapsed o-phenyl group
arrangement about Cl. (c) Space-filling model of [TiCl(OC6H4Ph2-2,6)3] [16] showing the upright o-phenyl groups and the resulting groove (the 6-phenyl
groups and the phenoxo ligand Hs have been removed for clarity). (d) Space-filling model, side-on view of 9 showing the extent to which the o-phenyl
groups cover Cl. (e) Space-filling model, side-on view of [TiCl(OC6H4Ph2-2,6)3] [16] showing the molecular groove (the 6-phenyl groups and the phenoxo
ligand Hs have been removed for clarity).
This approach is then expected to produce one rigid
phenoxo ligand and two more that are less rigid if the iso-
mer with Cl trans to N is formed, allowing the influences
made on the cavity ligand to be varied. Preliminary results
of this work indicate that reaction of [TiCl2(OC6H4CMe3-
4)2] [24] with LiOC6H2(CMe3)2-2,4-(CH2NMe2)-6 gives a
solid analysing as the complex [TiCl(OC6H4CMe3-4)2-
{OC6H2(CMe3)2-2,4-(CH2NMe2)-6}] but the NMR spectra
show the complex is made up of two isomers. Similarly,
reaction of [TiCl4] with two equivalents of LiOC6H3-
(CMe3)2-2,4 in diethyl ether followed by one equivalent
of LiOC6H2(CMe3)2-2,4-(CH2NMe2)-6 gives a complex
analysing as [TiCl{OC6H4(CMe3)2-2, 4}2(OC6H2(CMe3)2-
2,4-(CH2NMe2)-6)] which was also shown by the NMR
spectra to consist of two isomers. The complexes do not
appear to crystallise from solution so the structural details
of any cavity formed in any of the isomers cannot as yet be
assessed. The only success we have had with forming a sin-
gle isomer complex using this type of tie-back is from a
reaction of [TiCl2({OC6H2(CMe3)-2-Me-4}2CH2-6)] [27]
with one equivalent of LiOC6H2(CMe3)2-2,4-(CH2NMe2)-
6 which gives a product analysing as [TiCl({OC6H2(CMe3)-
2-Me-4}2CH2-6) (OC6H2(CMe3)2-2,4-(CH2NMe2)-6)]. The
NMR spectra indicate the presence of only one isomer but
crystals have not yet been obtained to identify the overall
geometry by X-ray crystallography.
In further pursuing a non-collapsible coordination
geometry about titanium, we have looked at the coordina-
tion properties of the tripodal ligand type tris(2-hydroxy-
phenyl)benzylamine, (HOArCH2)3N) for which the
methyl and tert-butyl substituted compounds (HOC6H2-
Me2-2,4-CH2-6)3N
[structure
(f),
R = Me]
and

2046
A.J. Nielson et al. / Polyhedron 25 (2006) 2039–2054
Table 3
[{N(ArO)3}Ti–O–Ti{(OAr)3N}] Æ 1.5C6H6 (10a) is formed
in good yield and this reacts with Me3SiCl to give chloro
complex 11. There is an initial report of this dimer formed
by the hydrolysis of [Ti(OCHMe2){(OC6H2Me2-2,4-CH2-
6)3N}] in air [28] and the crystal structure of
˚
Selected bond distances (A) and angles (ꢁ) for [TiCl(OC6H4Ph-2)3(diethyl
ether)] (9)
Ti–Cl
2.3067(6)
1.791(1)
1.794(1)
1.790(1)
2.312(3)
Ti–O(1)
Ti–O(2)
Ti–O(3)
Ti–O(4)
[{N(Ar0O)3}Ti–O–Ti{(OAr0)3N}].
[(OAr0)3N = OC6H2-
(CMe3)2-2,4-CH2-6]3N which is more resistant to hydroly-
sis has been obtained [30]. An X-ray crystal structure
determination of the oxo-bridged dimer prepared in this
work containing a pentane solvent molecule 10b shows a
distorted trigonal bipyramidal coordination geometry
(Fig. 4) with bond lengths and angles (Table 4) which are
similar to those found for the reported tert-butyl substi-
tuted analogue. The only significant difference between
the two is the smaller Ti–O–Ti bond angle in 10b
[147.5(1)ꢁ cf. 155.5(1)ꢁ] which is apparently related to the
smaller ortho-substituents providing less of a clash across
the two sides of the molecule. In the tert-butyl substituted
complex the tert-butyl groups are forced into close contact
in the dimer with eight inter-titanium methyl–methyl
Ti–O(1)–C(1)
Ti–O(2)–C(13)
Ti–O(3)–C(25)
Cl–Ti–O(1)
Cl–Ti–O(2)
Cl–Ti–O(3)
152.3(1)
155.0(1)
157.1(1)
96.12(4)
95.90(5)
95.80(6)
173.41(9)
118.95(6)
119.99(6)
79.72(9)
117.88(6)
81.83(9)
90.72(11)
Cl–Ti–O(4)
O(1)–Ti–O(2)
O(1)–Ti–O(3)
O(1)–Ti–O(4)
O(2)–Ti–O(3)
O(2)–Ti–O(4)
O(3)–Ti–O(4)
[HOC6H2(CMe3)2-2,4-CH2-6]3N [structure (f), R = CMe3],
respectively, should give the ‘‘tied back’’ versions of com-
plexes 2 and 6.
˚
contacts with Cꢁ ꢁ ꢁC contacts shorter than the 4.0 A sum
of the van der Waals radii [30] but in 10b there are no such
close contacts.
[TiCl{(OC6H2Me2-2,4-CH2-6)3N}] (11) was obtained
more conveniently and in virtually quantitative yield by
the reaction of the free ligand with [TiCl(OCHMe2)3] [18]
prepared in diethyl ether from [TiCl4] and 3 equivalents
of [Ti(OCHMe2)4]. Complex 11 is not particularly soluble
compared with the very soluble analogue [Ti(OCH-
Me2){(OC6H2Me2-2,4-CH2-6)3N}] [28]. The 1H NMR
spectrum shows a single set of resonances for the phenyl
rings and the substituents while the benzylic CH protons
form two very broad resonances. However, the
13C{1H}NMR spectrum shows only one CH2 carbon reso-
nance along with the other single set of carbon resonances.
The ipso-carbon resonance appears at d 160.2 which is
upfield to that found for the non-tied-back analogue 2
(d 164.4).
R
R
OH
R
HO
R
N
R
HO
R
Structure (f) (R = Me, CMe3)
Several alkoxo and phenoxo complexes of titanium con-
taining these ligands have been reported [28–30]. In
attempting to prepare chloro complexes, it was found that
reactions of [TiCl4] with either of these ligands in benzene
in the presence of Et3N gave mixtures which proved diffi-
cult to purify. This type of reaction is reported to give
[TiCl{(OC6H2Me2-2,4-CH2-6)3N}] in low yield but the
complex has not been well characterised [31]. The complex
was initially obtained via a circuitous route (Scheme 1).
We have noticed that when undried samples of
(HOC6H2Me2-2,4-CH2-6)3N, [(HOAr)3N], are used in the
reaction with [TiCl4] and Et3N, the oxo-bridged dimer
Addition of tris(3,5-di-tert-butyl-2-hydroxyphenyl)ben-
zylamine to a diethyl ether solution of [TiCl(OCHMe2)3]
gave the highly soluble complex [TiCl({OC6H2(CMe3)2-
2,4-CH2-6}3N)] Æ diethyl ether (12) in virtually quantitative
1
yield. The H NMR spectrum showed a single set of reso-
nances for the phenyl groups and the substituents and a
well-defined AB spin system for the methylene protons.
Although the structure has C3-symmetry (see later for the
X-ray structure) the methylene protons are non-equivalent
due to the methylene carbon being pushed to one side in
(a)
[TiCl4]
+
N(ArOH)3
[{N(ArO)3}Ti-O-Ti{(OAr)3N}]
(10)
(b)
[{N(ArO)3}TiCl]
(11)
Scheme 1. (a) NEt3 in benzene, reflux 14 h; (b) Me3SiCl in toluene, stir 8 h.

A.J. Nielson et al. / Polyhedron 25 (2006) 2039–2054
2047
Fig. 4. ORTEP diagram of [(Ti{(OC6H2Me2-2,4-CH2-6)3N})2O] Æ n-pentane (10b) showing the molecular structure. Hydrogen atoms are omitted for
clarity. The n-pentane solvent molecule is not shown.
the six-membered ring. The AB system also indicates that
rapid inversion of the structure does not occur on the
NMR time scale [28]. In the 13C{1H} NMR spectrum there
was a single set of resonances with the ipso-carbon reso-
nances appearing at d 160.8 which is similar to that found
in methyl substituted complex 11 but lies upfield to the
non-tied-back analogue 4-coordinate 6 (d 164.4). An X-
ray crystal structure of 12 showed a 5-coordinate distorted
trigonal bipyramidal structure (Fig. 5a) in which the
diethyl ether solvent molecule lies between the Ti complex
molecules in the unit cell and is not coordinated. Bond
lengths and angle are contained in Table 5.
In further comparisons of the relevant features of the
coordination geometries for the complexes reported here,
the O–Ti–Cl angles in the tied-back trigonal bipyramid
(12) are similar to those in the non-tied-back trigonal
bipyramid (9) [ranges 97.01(5)–97.18(5)ꢁ and 95.80(6)–
96.12(4)ꢁ, respectively] as are the O–Ti–O bond angles
[ranges 115.48(7)–120.72(7)ꢁ and 117.88(6)–119.99(6)ꢁ,
respectively] showing that tying back has little effect on
the trigonal bipyramidal geometry. However, for the tied-
back trigonal bipyramidal geometry there is a significant
effect for the Ti–O–C bond angles. In 12 the angles range
from 143.5(2)ꢁ to 144.0(1)ꢁ whereas the angle widens in
non-tied-back 9 [152.3(1)–157.1(1)ꢁ] (the complexes are
expected to have similar electronic environments). In 6
where there is no tie-back constraint, the Ti–O–C bond
angles are similar to those found for 9 [158.8(2)ꢁ]. There
is obvious flexibility in this angle and this is even more pro-
nounced in other complexes of this type of ligand. In the
zwitterionic complex [Zr{(OC6H2Me2-2,4-CH2-6)3NH}2]
which has a low requirement for p-donation with six phe-
noxo ligands present and in which there is also room for
a hydrogen atom on the inside of the cage at a non-agostic
˚
The Ti–Cl bond [2.3074(7) A] is similar to that found in
the non-tied back trigonal bipyramidal [TiCl(OC6H4Ph-2)3-
˚
(diethyl ether)] (9) [2.3067(6) A] and both these bond
lengths are longer than in tetrahedral [TiCl{OC6H3-
˚
(CMe3)22,4}3] (6) [2.240(2) A] which has a greater need
for electron density at the metal. This electronic situation
is further illustrated by the Ti–O bond lengths for the three
molecules that are longer for the 5-coordinate complexes
˚
[ranges 1.809(2)–1.817(2) and 1.790(1)–1.794(1) A for 12
˚
and 9 cf. 1.756(2) A for 6]. In the trigonal bipyramidal
˚
complexes [Ti(OCHMe2)({OC6H2(CMe3)2-2,4-CH2-6}3N)]
[28] and [Ti(OH)({OC6H2(CMe3)2-2,4-CH2-6}3N)] [30]
where there is an alkoxo or hydroxo ligand instead of Cl,
the Ti–O bond lengths [ranges, respectively, 1.825(6)–
distance [2.688(3) A], the angles range from 154(2)ꢁ to
160(2)ꢁ [32]. In [W(O)(OCH2CH2OH) ({OC6H2(CMe3)2-
2,4-CH2-6}3N)] where there is a similar low p-loading
requirement for the phenoxo ligands the angles range from
126.0(2)–141.9(2)ꢁ [33]. For 1r–2p donation from a phe-
noxo ligand, DFT calculations show there is a shallow min-
imum in the energy for the angle between about 160ꢁ and
180ꢁ, at 140ꢁ the energy is 20 kJ molꢀ1 and at 120ꢁ this
energy rises to 60 kJ molꢀ1 [23].
˚
˚
1.866 A and 1.783(2)–1.865(2) A] are mostly longer than
in 12 and 9 indicating that Cl is the poorer p-donor. This
effect is also illustrated by comparison of the relative Ti–
˚
N
bond lengths [2.216(2) A in 12; 2.334(5) and
˚
2.364(2) A, respectively, in the reported isopropoxo and
hydroxo complexes] where the bond is shorter when N is
trans to Cl. These features show there is a tightening of
the Ti–O and Ti–N bonds when there is a less effective p-
donor lying trans to the nitrogen atom.
Where the Ti–O–C bond angle is reduced it might be
expected that the tert-butyl group would move away from
the terminal Cl and create a wider but less well developed
pocket about this ligand. However, inspection of the inter-

2048
A.J. Nielson et al. / Polyhedron 25 (2006) 2039–2054
Table 4
Given the trigonal bipyramidal nature of complex 12
and the relevance of coordination expansion in the present
studies, an excess of pyridine was added to the complex
resulting in the formation of [(TiCl({OC6H2(CMe3)2-2,4-
˚
Selected bond distances (A) and angles (ꢁ) for [(Ti{(OC6H2Me2-2,4-CH2-
6)3N})2O] Æ n-pentane (10b)
Ti(1)–O(7)
Ti(1)–O(2)
Ti(1)–O(1)
Ti(1)–O(3)
Ti(1)–N(1)
Ti(2)–O(7)
Ti(2)–O(5)
Ti(2)–O(4)
Ti(2)–O(6)
Ti(2)–N(2)
O(1)–C(1)
O(2)–C(16)
O(3)–C(25)
O(4)–C(28)
O(5)–C(43)
O(6)–C(52)
1.8076(17)
1.8328(18)
1.8362(17)
1.8456(17)
2.379(2)
1.8219(17)
1.8304(18)
1.8348(18)
1.8387(19)
2.388(2)
1.357(3)
1.367(3)
1.355(3)
1.360(3)
1.360(3)
1.365(3)
1
CH2-6)3N})(NC5H5)] (13). The H NMR spectrum shows
the methylene protons as a broadened singlet for this 6-
coordinate compound compared with the AB quartet
found for the 5-coordinate parent. The pyridine resonances
1
in both the H and 13C{1H} NMR spectra were very well-
defined suggesting that fluxional processes are absent or are
fast on the NMR timescale. Suitable crystals of the com-
plex for X-ray crystallography have not yet been obtained
but the structure is expected to be octahedral with the pyr-
idine ligand coordinating trans to one of the phenoxo
ligands. For this geometry the chloro ligand would still
be surrounded by the tert-butyl groups but the pocket
would be opened up on one side by the pyridine ligand.
Recently, reported X-ray crystal structures show that this
type of coordination geometry can be obtained from biden-
tate oxygen and nitrogen ligands [34].
O(7)–Ti(1)–O(2)
O(7)–Ti(1)–O(1)
O(2)–Ti(1)–O(1)
O(7)–Ti(1)–O(3)
O(2)–Ti(1)–O(3)
O(1)–Ti(1)–O(3)
O(7)–Ti(1)–N(1)
O(2)–Ti(1)–N(1)
O(1)–Ti(1)–N(1)
O(3)–Ti(1)–N(1)
O(7)–Ti(2)–O(5)
O(7)–Ti(2)–O(4)
O(5)–Ti(2)–O(4)
O(7)–Ti(2)–O(6)
O(5)–Ti(2)–O(6)
O(4)–Ti(2)–O(6)
O(7)–Ti(2)–N(2)
O(5)–Ti(2)–N(2)
O(4)–Ti(2)–N(2)
O(6)–Ti(2)–N(2)
C(1)–O(1)–Ti(1)
C(16)–O(2)–Ti(1)
C(25)–O(3)–Ti(1)
C(28)–O(4)–Ti(2)
C(43)–O(5)–Ti(2)
C(52)–O(6)–Ti(2)
Ti(1)–O(7)–Ti(2)
99.36(8)
98.15(8)
116.97(8)
97.66(8)
118.80(9)
117.96(8)
179.41(8)
81.23(7)
81.54(7)
82.07(7)
97.24(8)
97.86(8)
115.75(9)
98.45(8)
122.94(9)
115.78(9)
178.52(8)
81.29(7)
3. Conclusion
The results of this work indicate that substituents in the
ortho-position of the phenoxo ligand ring in tetrahedral
chloro-tris-phenoxo complexes of titanium are positioned
so as to create a coordination pocket or cavity about the
chloro ligand. Where the substituent is a methyl group
the effect is minimal but a tert-butyl substituent gives rise
to a more substantial pocket. Coordination expansion
can occur when a further donor ligand is available which
removes this special environment. A phenyl substituent cre-
ates a better pocket but this can also be reduced in size
when a further ligand binds creating a trigonal bipyramidal
structure in which the pocket collapses somewhat. The col-
lapse is due to the need to remove interactions of the phe-
noxo ligand phenyl ring with the ligand lying trans to the
chloro ligand. Where an octahedral structure forms by fill-
ing two coordination sites, the possibility of isomers com-
plicates the issue producing structures in which not all
the ortho-substituents influence the chloro ligand. Where
the collapse needs to be prevented, the tied-back ligand sys-
tem gives rise to a trigonal bipyramidal structure of a more
robust nature. A further ligand may coordinate leading to
an octahedral structure which is expected to be more open
on one side but it is obvious that the nature of this ligand is
important given that diethyl ether does not coordinate but
pyridine does. Overall it is apparent that tert-butyl groups
in these complexes do not cover Cl completely forming
what is best described as a steric pressure pocket [10d]
whereas substituents that reach further outwards would
form a more enclosed pocket or cavity. Given the availabil-
ity of a wide variety of phenol ligands substituted in the
ortho-position, the chloro-tris-phenoxo approach to the
molecular engineering of coordination pockets should have
future potential. By varying the size of the substituent,
82.69(7)
82.51(8)
144.72(16)
142.42(17)
141.63(16)
139.95(17)
143.33(16)
140.04(17)
147.48(11)
atomic distances between the central tert-butyl group car-
bons and Cl in 6 and 12 shows that this effect is minimal
on changing from the tetrahedral structure (Cꢁ ꢁ ꢁCl dis-
˚
tance 4.87 A) to the trigonal bipyramid (Cꢁ ꢁ ꢁCl distances
˚
4.87, 4.91 and 4.93 A). Noticeable in the two structures is
˚
that the titanium atom lies 0.60 A above the plain made
by the three phenoxide oxygen atoms in the tetrahedron
but only 0.22 A above them in the trigonal bipyramid. This
is, of course, to be expected but it results in the chloro
ligand lying 2.15 A above the central carbon atoms of the
three tert-butyl groups in the tetrahedron and 1.69, 1.78
and 1.85 A in the trigonal bipyramid. The net effect is that
when the methyl groups are taken into account, the trigo-
nal bipyramidal structure creates a slightly better pocket
about Cl than does the tetrahedron. The pocket structure
in 12 is shown by the space-filling models in Fig. 5b and c.
˚
˚
˚

A.J. Nielson et al. / Polyhedron 25 (2006) 2039–2054
2049
Fig. 5. (a) ORTEP diagram of [TiCl({OC6H2(CMe3)2-2,4-CH2-6}3N)] Æ diethyl ether (12) showing the molecular structure. Hydrogen atoms have been
removed for clarity. The diethyl ether solvent is not shown. (b) Space-filling model looking down the Cl–Ti bond showing the pocket arrangement about
Cl. (c) Space-filling model, side-on view of the Ti–Cl bond showing the extent to which the tert-butyl group methyls cover the Cl atom.
collapsible through to non-collapsible complexes with dif-
ferent pocket characteristics may be realised.
13C{1H} NMR spectra were recorded at 400 and
100 MHz, respectively, on a Bruker AM400 spectrometer.
CDCl3 and CD2Cl2 were dried over, and distilled from,
freshly ground CaH2 and benzene-d6 from sodium wire.
C, H and N analyses were determined by Dr. A. Cunningh-
ame and associates, University of Otago, New Zealand.
The production of HCl gas in the thermalisation reactions
was monitored by passing the exhaust gases from the nitro-
gen bubbler over N,N,N0,N0-tetramethylethylenediamine
and observing the white cloud produced.
4. Experimental
4.1. General methods and materials
All preparations and manipulations were carried out
under dry oxygen-free nitrogen using standard bench-top
techniques for air-sensitive substances. [TiCl4] and the phe-
nols were used as received from commercial sources. 4,40-
Dimethyl-2,20-bibyridyl (dmbipy) was dried under vacuum
before use. Tris(2-hydroxy-4,6-di-tert-butylbenzyl)amine
and tris(2-hydroxy-4,6-dimethylbenzyl)amine were pre-
pared by the literature procedures [28]. Light petroleum
(b.p. 40–60 ꢁC) and toluene were distilled from sodium wire
4.2. Synthesis
4.2.1. [TiCl(OC6H4CMe3)3-4] Æ 0.5C7H8 (1)
4-tert-Butylphenol (4.75 g, 31.6 mmol) in toluene
(40 cm3) was added to [TiCl4] (2.0 g, 10.5 mmol) in toluene
(40 cm3) and the mixture was refluxed until no more HCl
1
and dichloromethane from freshly ground CaH2. H and

2050
A.J. Nielson et al. / Polyhedron 25 (2006) 2039–2054
2.15 mmol) in CH2Cl2 (20 cm3) and the mixture was stirred
for 1 h. The solution was filtered and the solvent removed
to give the complex as an orange flaky solid. Yield: 1.30 g,
96%. Anal. Calc. for C36H39ClN2O3Ti: C, 68.52; H, 6.23,
Table 5
˚
Selected bond distances (A) and angles (ꢁ) for [TiCl({OC6H2(CMe3)2-2,4-
CH2-6}3N)] Æ (OEt2) (12)
Ti–O1
Ti–O2
Ti–O3
Ti–N1
1.809(2)
1.817(2)
1.811(2)
2.216(2)
2.3074(7)
1.363(3)
1.370(3)
1.367(3)
1.501(3)
1.495(3)
1.497(3)
1
N, 4.44. Found: C, 67.80; H, 6.73, N, 4.68%. H NMR: d
1.67 (s, 3H, Me); 2.06 (s, 3H, Me); 2.24 (s, 6H, 2Me);
3
2.34 (s, 6H, 2Me); 2.40 (s, 6H, 2Me); 6.25 [d, J(HH) 8.1,
Ti–Cl
1H, o-H]; 6.46 [dd, 3J(HH) 8.1, 4J(HH) 1.6, 1H, m-H];
6.58 [d, 4J(HH) 1.6, 1H, m-H]; 6.83 [dd, 3J(HH) 8.1,
O1–C1
O2–C16
O3–C31
N1–C7
N1–C22
N1–C37
4
4J(HH) 1.5, 2H, m-Hs]; 6.91 [d, J(HH) 1.5, 2H, m-Hs];
3
3
7.20 [d, J(HH) 5.5, 2H, m-H dmbipy]; 7.38 [d, J(HH)
8.1, 2H, o-Hs]; 7.81 [bd, 2H, m-Hs dmbipy]; 8.95 [d,
3J(HH) 5.5, 2H, o-Hs dmbipy]. 13C{1H} NMR: d 16.1
(1Me); 17.1 (2Me); 20.4 (1Me); 20.7 (2Me); 21.5
(2Me, dmbipy); 118.5 (1CH); 120.1 (2CH); 122.2 (m-CH,
dmbipy); 126.3 (2C); 126.4 (1m-CH); 126.6 (2m-CHs,
dmbipy); 127.0 (2m-CHs); 128.7 (1C); 128.9 (1C); 130.1
(2C); 130.3 (1m-CH); 130.7 (2m-CHs); 149.0 (2o-CHs,
dmbipy); 150.7 (2Cs, dmbipy); 151.7 (2Cs, dmbipy); 162.7
(1 ipso-C); 163.51 (2 ipso-Cs).
O(1)–Ti–O(2)
O(1)–Ti–O(3)
O(3)–Ti–O(2)
O(1)–Ti–N(1)
O(2)–Ti–N(1)
O(3)–Ti–N(1)
O(1)–Ti–Cl(2)
O(3)–Ti–Cl(2)
O(2)–Ti–Cl(2)
N(1)–Ti–Cl(2)
C(1)–O(1)–Ti
C(16)–O(2)–Ti
C(31)–O(3)–Ti
119.29(7)
115.48(7)
120.72(7)
82.76(7)
82.66(7)
83.30(7)
97.10(5)
97.18(5)
97.01(5)
179.51(5)
143.54(15)
143.66(14)
143.97(14)
4.2.4. [TiCl(OC6H2Me3-2,4,6)3] (4)
2,4,6-Trimethylphenol (4.31 g, 31.7 mmol) in toluene
(40 cm3) was added to [TiCl4] (2.0 g, 10.5 mmol) in toluene
(40 cm3) and the mixture was refluxed until no more HCl
gas was produced (15.5 h). The cooled solution was filtered
from a small amount of solid, the solvent was removed and
the residue held under vacuum for 2 h giving a gum which
formed the complex as a deep-red crystalline mass on
standing overnight. Yield: 4.5 g, 87%. Anal. Calc. for
C27H33ClO3Ti: C, 66.33; H, 6.80. Found: C, 65.88; H,
gas was produced (12 h). NMR spectroscopy at this stage
confirmed that no 4-tert-butylphenol was present. The
cooled solution was filtered, the solvent was removed and
the residue held under vacuum for 2 h giving the complex
as a deep-red flaky solid. Yield: 5.82 g, 97%. Anal. Calc.
for C33.5H43ClO3Ti: C, 69.73; H, 7.51. Found: C, 69.48;
H, 7.58%. 1H NMR: d 1.27 (s, 27H, CMe3), 2.40 [s,
1.5H, 0.5CH3(toluene)], 7.11 (b, 12H, o and m-Hs), 7.22
and 7.30 (m, 2.5H, 0.5 aromatics (toluene). 13C{1H}
NMR: d 21.5 [Me(toluene)], 31.5 (CMe3), 34.2 (C), 118.7
(b, o-C) 124.7 (toluene), 124.6 (b, m-C), 128.2 (toluene),
129.0 (toluene), 137.9 [ipso-C(toluene)], 145.8 (b, C),
164.6 (ipso-C).
1
6.70%. H NMR: d 2.19 (s, 9H, Me), 2.26 (bs, 18H, Me),
6.72 (bs, 6H, m-Hs). 13C{1H} NMR: d 16.9 (o-Me), 20.6
(p-Me), 126.7 (o-C), 128.6 (CH), 132.4 (p-C), 163.9
(ipso-C).
4.2.5. [TiCl{OC6H3(CHMe2)2-2,6}3] (5)
2,6-Di-iso-propylphenol (3.0 g, 16.8 mmol) in toluene
(30 cm3) was added to [TiCl4] (1.0 g, 5.27 mmol) in toluene
(30 cm3) and the mixture was refluxed until no more HCl
gas was produced (19.5 h). The cooled solution was filtered
and the solvent removed giving an orange solid. NMR
spectroscopy at this stage indicated the presence of unre-
acted phenol and two coordination products. The crude
material was dissolved in toluene (50 cm3), the volume
reduced to ca. 30 cm3 and the solution stood at ꢀ20 ꢁC,
giving the complex as a non-crystalline orange solid con-
taining several large crystals of [Ti{OC6H3(CHMe2)2-
2,6}4] which were manually removed from the product.
Yield: 1.30 g, 40%. Anal. Calc. for C36H51ClO3Ti: C,
4.2.2. [{TiCl(OC6H3Me2-2,4)2(l-OC6H3Me2-2,4)}2] (2)
2,4-Dimethylphenol (3.87 g, 31.7 mmol) in toluene
(40 cm3) was added to [TiCl4] (2.0 g, 10.5 mmol) in toluene
(40 cm3) and the mixture was refluxed until no more HCl
gas was produced (12 h). The cooled solution was filtered,
the solvent was removed and the residue held under vac-
uum for 2 h giving a gum that formed the complex as a
deep-red crystalline mass on standing overnight. Yield:
4.5 g, 96%. Anal. Calc. for C24H27ClO3Ti: C, 64.51; H,
1
6.09. H NMR: d (s, 9H, Me), 2.20 (s, 9H, Me), 6.72–
6.89 (m, 9H, o and m-Hs). 13C{1H} NMR: d 16.6 (Me),
20.7 (Me), 119.5 (CH), 126.3 (C), 126.9 (CH), 131.0
(CH), 133.2 (C), 164.4 (ipso-C). Crystalline material was
cut from the mass and used for the X-ray analysis.
1
70.29; H, 8.36. Found: C, 70.78; H, 8.88%. H NMR: d
3
3
1.13 [d, J(HH) 6.9, 36H, CMe2], 3.42 [sept, J(HH) 6.9,
CH], 6.91 (t, 3H, p-Hs), 7.03 (m, 6H, m-Hs). 13C{1H}
NMR: d 23.2 (Me), 27.5 (CH), 123.0 (m-C), 123.8 (p-C),
137.4 (o-C), 163.1 (ipso-C). The NMR spectra are identical
to that found for the complex prepared in refluxing ben-
zene [20,21].
4.2.3. [TiCl(OC6H3Me2-2,4)3(dmbipy)] (3)
4,40-Dimethyl-2,20-bipyridine (0.4 g, 2.17 mmol) in
CH2Cl2 (20 cm3) was added to complex 2 (0.96 g,

A.J. Nielson et al. / Polyhedron 25 (2006) 2039–2054
2051
3
4
4.2.6. [TiCl{OC6H3(CMe3)2-2,4}3] (6)
(s, 6H, 2Me dmbipy); 6.96 [dd, J(HH) 8.2, J(HH) 2.4,
3
Method A: 2,4-Di-tert-butylphenol (9.8 g, 47.5 mmol) in
toluene (60 cm3) was added to [TiCl4] (3.0 g, 15.8 mmol) in
toluene (50 cm3) and the mixture was refluxed until no
more HCl gas was produced (14.5 h). After refluxing for
a further 2 h, the cooled solution was filtered and the sol-
vent removed giving a crystalline orange-red solid. Crude
yield: 11.0 g, 99%. This material has a slight red tinge
and NMR spectroscopy indicated the presence of a small
amount of unreacted phenol (ca. 3–5%) but it was suffi-
ciently pure for further use. The solid was washed with
petroleum spirit (40 cm3) giving the complex as an orange
crystalline solid. Yield: 9.21 g, 83%. (Further complex can
be obtained by cooling the petroleum spirit washings.)
Anal. Calc. for C42H63ClO3Ti: C, 72.13; H, 9.08. Found:
C, 71.96; H, 9.26%. 1H NMR: d 1.29 (s, 27H, CMe3),
1.46 (s, 27H, CMe3), 7.10 (s, 6H, o and m-Hs), 7.29 [d,
4J(HH) 1.3, 3H, m-H). 13C{1H} NMR: d 30.3 (Me), 31.6
(Me), 34.6 (C), 35.1 (C), 122.7 (CH), 123.4 (CH), 124.0
(CH), 135.6 (C), 146.3 (C), 163.2 (ipso-C). On a larger
scale, a reaction between 26.15 g of 2,4-di-tert-butylphenol
(0.127 mol) and 8 g of [TiCl4] (0.0422 mol) in 120 cm3 of
toluene produced 28.5 g (97%) of the crude product after
refluxing for 21 h.
2H, m-Hs]; 6.98 (obscured dd, 1H, m-H); 7.09 [d, J(HH)
3
4
5.6, 2H, m-Hs dmbipy]; 7.11 [d, J(HH) 8.4, J(HH) 2.3,
4
1H, m-H]; 7.23 [d, J(HH) 2.4, 2H, m-Hs]; 7.29 [d, J(HH)
3
2.3, 1H, m-H]; 7.55 [bd, J(HH) 8.2, 2H, o-Hs]; 7.91 (s,
2H, m-Hs dmbipy); 8.93 [d, 3J(HH) 5.6, 2H, o-Hs dmbipy].
Easily distinguishable ortho-hydrogen resonances for the
other isomers are at d 8.79 and 8.53.
4.2.8. [TiCl(OC6H4Ph-2)3] (8)
2-Hydroxybiphenyl (8.1 g, 47.6 mmol) in dichlorometh-
ane (50 cm3) was added to [TiCl4] (3.0 g, 15.8 mmol) in
dichloromethane (50 cm3) and the mixture was refluxed
until no more HCl gas was produced (30 h). The cooled
solution was filtered, the solvent was removed giving a dark
red gum which failed to solidify on extended pumping. The
complex was identified on the basis of NMR spectral char-
acteristics that were similar to complex 9. 1H NMR: d 6.58
[d, 3J(HH) 7.4, 3H, o-H], 7.01 [t, 3J(HH) 7.4, 3H, p-H], 7.07
[t, 3J(HH) 7.4, 3H, m-H], 7.12–7.18 [m, 9H, m and p-
Hs(phenyl substituent)], 7.20 [dd, 3J(HH) 7.4, 4J(HH)
1.4, 3H, m-H], 7.32 [d, 3J(HH) 7.4, 6H, o-Hs(phenyl substi-
tuent)]. 13C{1H} NMR: d 120.4 (CH), 124.0 (CH), 127.3
(CH), 128.2 (CH), 128.3 (CH), 129.2 (CH), 130.0 (CH),
130.6 (C), 137.3 (C), 163.6 (ipso-C).
Method B: n-Butyl lithium (9.8 cm3 of a 1.6 mol Lꢀ1
solution, 15.8 mmol) in hexane was added dropwise to a
solution of 2,4-di-tert-butylphenol (3.26 g, 15.8 mmol) in
diethyl ether (50 cm3) cooled to ꢀ30ꢁC and the solution
was stirred with the cooling bath removed for 2 h. [TiCl4]
(1.0 g, 5.27 mmol) was added to a flask which was then
cooled to ꢀ78 ꢁC and diethyl ether (50 cm3) was added.
The solution was heated until all the yellow solid dissolved
and then recooled to ꢀ78 ꢁC and the lithium phenoxide
solution added via a cannula. The cooling bath was
removed and the solution stirred for 20 h, filtered and the
solvent removed to give the crude complex as an orange
crystalline solid (yield: 3.08 g) which was dissolved in
diethyl ether (50 cm3). The solution was filtered and the
volume reduced to ca. 25 cm3 giving the complex as
yellow-orange crystals on standing. Yield: 1.24 g, 34%.
Anal. Calc. for C42H63ClO3Ti: C, 72.13; H, 9.08. Found:
C, 72.14; H, 8.72%. The complex has identical NMR
spectra to the sample prepared under method A. A single
crystal from the mass was used for the X-ray analysis.
4.2.9. [TiCl(OC6H4Ph-2)3(diethyl ether)] (9)
n-Butyl lithium (15.3 cm3 of a 1.6 mol Lꢀ1 solution,
24.4 mmol) in hexane was added dropwise to a solution of
2-hydroxybiphenyl (4.15 g, 24.4 mmol) in diethyl ether
(80 cm3) cooled to ꢀ30ꢁC and the solution was stirred with
the cooling bath removed for 2 h. [TiCl4] (1.54 g, 8.1 mmol)
was added to a flask which was then cooled to ꢀ78 ꢁC and
diethyl ether (80 cm3) was added. The solution was heated
until all the yellow solid dissolved and after cooling to room
temperature, it was added via a cannula to the suspension of
the lithium phenoxide cooled to ꢀ78 ꢁC. The cooling bath
was removed and the solution stirred for 20 h, filtered and
the remaining solid washed with diethyl ether until it was
no longer coloured orange. The combined filtrate and wash-
ings were kept hot while the volume was reduced to approx-
imately half and on standing the complex formed as large
ruby-red crystals. Yield: 1.8 g, 51%. Anal. Calc. for
C40H37ClO4Ti: C, 72.24; H, 5.61. Found: C, 71.91; H,
1
3
5.48%. H NMR: d 1.13 [t, J(HH) 7.0, 6H, Me], 3.45 [q,
3J(HH) 7.0, 4H, CH2], 6.62 (d, 3H, aromatic-H), 6.99 (t,
3H, aromatic-H), 7.05 (t, 3H, aromatic-H), 7.09–7.20 (m,
12H, aromatic-Hs), 7.32 (d, 6H, aromatic-Hs). 13C{1H}
NMR: d 15.0 (Me), 66.0 (CH2), 120.5 (CH), 123.6 (CH),
127.2 (CH), 128.2 (CH), 129.2 (CH), 130.0 (CH), 130.3
(C), 137.5 (o-C),163.2 (ipso-C). A single crystal from the
mass was used for the X-ray analysis.
4.2.7. [TiCl{OC6H3(CMe3)2-2,4}3(dmbipy)] (7)
4,40-Dimethyl-2,20-bipyridine (0.267 g, 1.45 mmol) in
CH2Cl2 (20 cm3) was added to complex 6 (1.014 g,
1.45 mmol) in CH2Cl2 (20 cm3) and the mixture was stirred
for 2 h. The solution was filtered and the solvent removed
to give the complex as an orange flaky solid. Yield: 1.28 g,
99%. Anal. Calc. for C54H75ClN2O3Ti: C, 73.40; H, 5.56,
N, 3.17. Found: C, 73.04; H, 5.76, N, 3.09%. The NMR
spectra are very complicated due to the presence of iso-
4.2.10. [(Ti{(OC6H2Me2-2,4-CH2-6)3N})2O] Æ 1.5C6H6
(10a)
1
mers. Only the H NMR of the distinguishable major iso-
mer is reported: d 1.28 (s, 18H, 2CMe3); 1.31 (s, 9H,
CMe3); 1.37 (s, 18H, 2CMe3); 1.51 (s, 9H, CMe3); 2.31
Undried tris(2-hydroxy-4,6-dimethylbenzyl)amine (1.77
g, 4.22 mmol) in benzene (40 cm3) was added dropwise to

2052
A.J. Nielson et al. / Polyhedron 25 (2006) 2039–2054
a stirred solution of [TiCl4] (0.8 g, 4.28 mmol) in benzene
(40 cm3) and triethylamine (1.8 cm3, 12.74 mmol) was
added dropwise. The mixture was refluxed for 14 h, the
cooled solution filtered and the solvent removed to give
the complex as yellow crystalline solid which was washed
with chilled diethyl ether (30 cm3). Yield: 2.0 g, 100%. Anal.
Calc. for C54H60N2O7Ti2 Æ 1.5C6H6: C, 71.25; H, 6.55; N,
Yield: 2.92 g, 96%. Anal. Calc. for C45H66ClNO3Ti (i.e.,
[{N(ArO)3}TiCl]): C, 71.81; H, 8.84, N, 1.86. Found: C,
72.99; H, 8.76; N, 2.01%. The solid was dissolved in diethyl
ether (120 cm3) and the solvent kept hot and removed in
vacuo until crystallisation could no longer be prevented.
On standing, well formed orange crystals were formed.
Further crystalline product was obtained by repeating this
process. Anal. Calc. for C49H76ClNO4Ti: C, 71.21; H, 9.27;
1
2.64. Found: C, 71.39; H, 6.76; N, 2.90%. H NMR: d
1
(C6D6) 2.15 (s, 9H, Me), 2.38 (s, 9H, Me), 2.8 (b, 3H,
CH), 3.8 (b, 3H, CH), 6.54 (bs, 3H, m-CH), 6.73 (bs, 3H,
m-CH), 7.28 (benzene). 13C{1H} NMR: d 16.1 (Me), 20.7
(Me), 58.8 (CH2),124.3 (C), 125.0 (C), 128.5 (CH), 129.3
(C), 131.0 (CH), 160.8 (ipso-C). A single crystal obtained
from the mass produced on recrystallisation from n-pen-
tane was used for the X-ray analysis [complex 10b].
N, 1.70. Found: C, 71.45; H, 9.36; N, 1.69%. H NMR: d
1.26 (t, 6H, diethyl ether),1.34 (s, 27H, CMe3), 1.52 (s,
27H, CMe3), 3.18 [d, 2J(AB) 13.9, 3H, CH], 3.53 [d,
2J(AB) 13.9, 3H, CH], 7.09 [d, 3J(HH) 2.1, 3H, m-H],
7.31 [d, 3J(HH) 2.1, 3H, m-H]. 13C{1H} NMR: d 15.3
(Me, diethyl ether), 29.6 (CMe3), 31.6 (CMe3), 34.5 (C),
34.9 (C), 58.7 (CH2), 65.8 (CH2,diethyl ether), 123.3 (o-
C), 123.5 (m-CH), 124.0 (m-CH), 135.7 (p-C), 144.3 (o-
C), 160.8 (ipso-C). A single crystal from the mass was used
for the X-ray analysis.
4.2.11. [TiCl{(OC6H2Me2-2,4-CH2-6)3N}] (11)
Method A: Chlorotrimethylsilane (2 cm3, 1.34 mmol)
was added via a syringe to [(Ti{(OC6H2Me2-2,4-CH2-
6)3N})]2O] (0.6 g, 0.67 mmol) in toluene (30 cm3) and the
mixture was stirred for 8 h. The solution was filtered and
the orange precipitate dried under vacuum. Yield: 0.49 g,
73%. Anal. Calc. for C27H30ClNO3Ti: C, 64.87; H, 6.05,
4.2.13. [TiCl{(OC6H2(CMe3)2-2,4-CH2-6)3N}(NC5H5)]
(13)
Pyridine (0.2 g, 2.5 mmol) in diethyl ether (10 cm3) was
added dropwise from a syringe to a rapidly stirred solution
of complex 12 (0.373 g, 0.45 mmol) in diethyl ether and the
mixture was stirred for 2 h. The solution was filtered and
the brown microcrystalline solid dried in vacuo. Yield:
0.364 g, 97%. Anal. Calc. for C50H71ClN2O3Ti: C, 72.23;
1
N, 2.80. Found: C, 64.82; H, 6.32; N, 2.97%. H NMR
(the complex is only slightly soluble in CDCl3): d 2.22 (s,
9H, Me), 2.25 (s, 9H, Me), 2.9 (b, 3H, CH), 3.9 (b, 3H,
CH), 6.73 (bs, 3H, m-H), 6.88 (bs, 3H, m-H). 13C{1H}
NMR: d 15.8 (Me), 20.7 (Me), 58.5 (CH2), 122.8 (C),
124.4 (C), 127.2 (CH), 131.1 (CH), 131.3 (C), 160.2 (ipso-
C).
Method B: [TiCl4] (0.29 g, 1.52 mmol) was cooled to
ꢀ78 ꢁC and diethyl ether (40 cm3) was added. The solution
was heated to dissolve the yellow solid, cooled to room
temperature and added dropwise to [Ti(OCHMe2)4]
(1.30 g, 4.57 mmol) in diethyl ether (30 cm3). The mixture
was stirred for 30 min [TiCl(OCHMe2)3] equivalent:
(1.59 g, 6.08 mmol) and dry tris(2-hydroxy-4,6-dimethylb-
enzyl)amine (2.55 g, 6.08 mmol) in diethyl ether (40 cm3)
added dropwise. The mixture was stirred for 2 h and the
orange precipitate filtered off and dried under vacuum.
Yield: 2.92 g, 96%. Anal. Calc. for C27H30ClNO3Ti: C,
64.87; H, 6.05, N, 2.80. Found: C, 65.31; H, 6.41; N,
2.81%. The complex had similar NMR spectra to the sam-
ple prepared under method A.
1
H, 8.61; N, 3.44. Found: C, 72.70; H, 9.27; N, 3.66%. H
NMR (CD2Cl2): d 1.99 (t, 6H, Me), 1.32 (s, 27H, CMe3),
1.49 (s, 27H, CMe3), 3.47 (q, 4H, CH2), 3.64 (bs, 6H,
3
3
CH2), 7.09 [d, J(HH) 2.2, 3H, m-H], 7.31 [d, J(HH) 2.2,
3
3H, m-H], 7.31–7.35 (obsm, 2H, m-Hpy), 7.73 (td, J(HH)
7.7, 1H, (p-Hpy)), 8.63 (bs, 2H, o-Hpy). 13C{1H} NMR: d
15.5 (Mediethyl ether), 29.7 (CMe3), 31.7 (CMe3), 34.8 (C),
35.2(C), 59.0 (CH2), 66.1 (CH2diethyl ether), 123.9 (CH),
123.95 (o-C), 124.2 (m-CHpy), 135.8 (p-C), 136.4 (p-
CHpy),144.8 (o-C), 150.1 (o-CHpy), 161.1 (ipso-C).
4.3. X-ray crystallography
Crystallographic and refinement data are given in Table
6. Data collected, with graphite monchromated Mo Ka X-
radiation, were measured on a Siemens SMART CCD area
detector diffractometer with the omega scan method. The
data were corrected for Lorentz and polarisation effects
and for absorption by the multi-scan method [35]. The
structures were solved by direct methods [36] and refined
by full-matrix least-squares methods [37] on F2. Hydrogen
atoms were placed in calculated positions and were refined
with a riding model (Uiso = 0.08). Diagrams were prepared
by ORTEP [38] and RASTOP [39]. Selected bond distances
and angles are listed in Tables 1–5. Complex 2 is dimeric
lying across a centre of symmetry. Some disorder was
noted for the Cl (with a site occupancy ratio of ca. 90:10)
and also for the H atoms in the Me group in the 4 positions
of both terminal phenoxo ligands. The occupancies of these
latter were fixed at 0.5. In 6 the tert-butyl group in the para
4.2.12. [TiCl({OC6H2(CMe3)2-2,4-CH2-6}3N)] Æ diethyl
ether (12)
[TiCl4] (0.234 g, 1.23 mmol) was cooled to ꢀ78 ꢁC and
diethyl ether (40 cm3) was added. The solution was heated
to dissolve the yellow solid, cooled to room temperature
and added dropwise to [Ti(OCHMe2)4] (1.05 g, 3.69 mmol)
in diethyl ether (30 cm3). The mixture was stirred for
30 min ([TiCl(OCHMe2)3] equivalent: 1.286 g, 4.95 mmol)
and
dry
tris(2-hydroxy-4,6-di-tert-butylbenzyl)amine
(3.33 g, 4.96 mmol) in diethyl ether (40 cm3) added drop-
wise. The mixture was stirred for 2 h, the solution filtered
and the solvent removed to give an orange crystalline solid.

A.J. Nielson et al. / Polyhedron 25 (2006) 2039–2054
2053
Table 6
Crystallographic data and structural information
Compound
2
6
9
10b
12
Molecular formula
Mr
C
446.81
24H27ClO3Ti
C42H63ClO3Ti
699.27
C40H37ClO4Ti
665.05
C39H72N2O7Ti2
1016.99
C40H37ClNO4Ti
826.46
T (K)
150(2)
293(2)
150(2)
150(2)
189(2)
Crystal size (mm)
Crystal system
Space group
0.40 · 0.30 · 01.6
monoclinic
P21/c
10.9316(1)
25.7193(1)
8.0853(1)
90.0
93.882(1)
90.0
2267.99(4)
4
0.22 · 0.24 · 0.38
hexagonal
0.50 · 0.40 · 0.40
monoclinic
P21/n
13.1474(2)
17.3038(2)
15.3222(2)
90.0
94.725(1)
90.0
3473.95(8)
4
0.42 · 0.30 · 0.22
triclinic
0.38 · 0.18 · 0.18
monoclinic
P21/n
15.0358(3)
15.8104(2)
20.6045(4)
90.0
92.630(1)
90.0
4892.98(15)
4
ꢀ
ꢀ
R3
P1
˚
a (A)
16.680(2)
16.680(2)
27.248(5)
90.0
90.0
120.0
6565.3(16)
6
1.061
0.289
2268
12.3309(2)
12.9833(2)
20.0930(2)
84.258(1)
72.168(1)
66.649(1)
2810.61(7)
2
˚
b (A)
˚
c (A)
a (ꢁ)
b (ꢁ)
c (ꢁ)
3
˚
V (A )
Z
Dcalc (mg cmꢀ3
)
1.309
0.516
936
1.272
0.363
1392
1.202
0.335
1080
1.122
0.270
1792
l (mmꢀ1
F (000)
)
h Range for data collection (ꢁ)
Reflections collected
1.58–23.32
10347
1.60–25.00
11450
1.78–26.30
19986
1.06–25.35
24227
1.62 to 26.43
28138
Unique reflections (Rint
Reflections with [I > 2r(I)]
Refined parameters
)
3267 (0.0172)
2983
275
2582 (0.0242)
2203
173
7008 (0.0193)
5712
443
10220 (0.0252)
8045
647
10004 (0.0440)
7179
526
Completeness to h = xꢁ (%)
Absorption correction Tmax, Tmin
Goodness-of-fit (all data)
23.32, 99.8
0.922, 0.820
1.045
25.0, 100.0
0.939, 0.898
1.124
26.30, 99.3
0.869, 0.839
1.056
25.35, 99.2
0.930, 0.872
1.051
26.43, 99.3
0.953, 0.905
1.051
Final R indices [I > 2r(I)] (R1,wR2)
0.0303, 0.0824
0.0338, 0.0843
0.396, ꢀ0.374
0.0613, 0.1895
0.0699, 0.1961
0.658, ꢀ0.576
0.0357, 0.0891
0.0492, 0.0974
0.240, ꢀ0.373
0.0451, 0.1143
0.0632, 0.1287
0.511, ꢀ0.437
0.0496, 0.1035
0.0833, 0.1204
0.533, ꢀ0.419
R indices (all data) (R1, wR2) ꢀ3
˚
Maximum, minimum Dq (e A
)
(b) D.V. Yandulov, R.R. Schrock, Inorg. Chem. 44 (2005) 1103;
(c) V. Ritleng, D.V. Yandulov, W.W. Weare, R.R. Schrock, A.S.
Hock, W.M. Davis, J. Am. Chem. Soc. 126 (2004) 6150;
(d) For general work related to this ligand type see: R.R. Schrock,
Acc. Chem. Res. 30 (1997) 9.
position is disordered (ratio ca. 80:20) and in 9 two sites
(ratio 65:35) have been distinguished for the co-ordinated
ether molecule. Solvent of crystallisation was present in
10b (n-pentane) and 12 (ether).
[11] (a) For titanium see: J.C. Peters, A.R. Johnson, A.L. Odom, P.W.
Wanandi, W.M. Davis, C.C. Cummins, J. Am. Chem. Soc. 118
(1996) 10175;
5. Supplementary data
(b) For molybdenum see: A.R. Johnson, P.W. Wanandi, C.C.
Cummins, Organometallics 13 (1994) 2907;
CCDC 281293–281297 contains the supplementary
crystallographic data for this paper. These data can be
obtained free of charge from The Cambridge
Crystallographic Data Centre via www.ccdc.cam.ac.uk/
datarequest/cif.
(c) E.L. Sceats, J.S. Figueroa, C.C. Cummins, M.M. Loening, P. van
der Wel, R.G. Griffen, Polyhedron 23 (2004) 2751;
(d) W. Zhang, S. Kraft, J.S. Moore, Chem. Commun. (2003) 832;
(e) T. Agapie, P.L. Diaconescu, C.C. Cummins, J. Am. Chem. Soc.
124 (2002) 2412;
(f) A. Fuerstner, C. Mathes, C.W. Lehmann, Chem. Eur. J. 7 (2001)
5299;
References
(g) J.B. Greco, J.C. Peters, T.A. Baker, W.M. Davis, C.C. Cummins,
G. Wu, J. Am. Chem. Soc. 123 (2001) 5003;
(h) A. Fuerstner, C. Mathes, C.W. Lehmann, J. Am. Chem. Soc. 121
(1999) 9453.
[1] N. Kitajima, W.B. Tolman, Prog. Inorg. Chem. (1995) 419.
[2] S.R. Colinson, D.A. Fenton, Coord. Chem. Rev. 148 (1996) 19.
[3] G. Parkin, Chem. Rev. 104 (2004) 702.
[4] B. Kersting, Z. Anorg. Allg. Chem. 630 (2004) 765.
[5] A.R. van Doorn, M. Bos, S. Harkema, J. van Eerden, W. Verboom,
D.N. Reinhoudt, J. Org. Chem. 56 (1991) 2371.
[12] D.C. Bradley, I.M. Thomas, J. Chem. Soc. (1960) 3857.
[13] T. Tanaka, T. Onoue, K. Kamei, K. Murakami, C. Iwata, J. Chem.
Soc., Chem. Commun. (1990) 906.
[6] V. van Axel-Castelli, A. Dalla-Cort, L. Mandolini, D.N. Reinhoudt,
L. Schiaffino, Chem. Eur. J. 6 (2000) 1193.
[7] V. van Axel-Castelli, A. Dalla-Cort, L. Mandolini, V. Pinto, D.N.
Reinhoudt, F. Ribaudo, C. Sanna, L. Schiaffino, B.H.M. Snellinkr-
uel, Supramol. Chem. 14 (2002) 211.
[8] R.R. Conry, G. Ji, A.A. Tipton, Inorg. Chem. 38 (1999) 906.
[9] B.S. Hammes, D. Ramos-Maldonado, G.P.A. Yap, L. Liable-Sands,
A.L. Rheingold, V.G. Young Jr., A.S. Borovik, Inorg. Chem. 36
(1997) 3210.
[10] (a) For recent work related to reduction of ammonia see: D.V.
Yandulov, R.R. Schrock, Can. J. Chem. 83 (2005) 341;
[14] H. Ohno, K. Hiramatsu, T. Tanaka, Tetrahedron Lett. 45 (2004) 75.
[15] D. Seebach, L. Widler, Helv. Chim. Acta 65 (1982) 1972.
[16] A. Sato, A. Hattori, K. Ishihara, S. Saito, H. Yamamoto, Chem.
Lett. 32 (2003) 1006.
[17] S. Saito, H. Yammoto, Chem. Commun. (1997) 1585.
[18] R. Imwinkelreid, D. Seebach, Org. Synth. 67 (1989) 180.
[19] A. Shah, A. Singh, R.C. Mehotra, Polyhedron 5 (1986) 1285.
[20] A. Shah, A. Singh, R.C. Mehotra, Indian J. Chem. A 32 (1993) 632.
[21] H. Noeth, M. Schmidt, Organometallics 14 (1995) 4601.
[22] S.L. Latesky, J. Keddington, A.K. McMullen, I.P. Rothwell, Inorg.
Chem. 24 (1985) 995.

2054
A.J. Nielson et al. / Polyhedron 25 (2006) 2039–2054
[23] A.J. Nielson, P. Schwerdtfeger, J.M. Waters, J. Chem. Soc., Dalton
Trans. (2000) 529.
[24] A.J. Nielson, C. Shen, P. Schwerdtfeger, J.M. Waters, Eur. J. Inorg.
Chem. (2005) 1343.
[25] Y. Morino, H. Uehara, J. Chem. Phys. 45 (1966) 4543.
[26] Y.E. Chao, P.M. Rodgers, D.E. Wigley, S.J. Alexander, A.L.
Reingold, J. Am. Chem. Soc. 113 (1991) 6326.
[30] V. Ugrinova, G.A. Ellis, S.N. Brown, Chem. Commun. (2004) 468.
[31] L. Michalczyk, S. de Gala, J.W. Bruno, Organometallics 20 (2001)
5547.
[32] (a) M.G. Davidson, C.H. Doherty, A.L. Johnson, M.F. Mahon,
Chem. Commun. (2003) 1832;
(b) A.J. Nielson, C. Shen, J.M. Waters, Acta Crystallogr., Sect. C 59
(2003) m494.
[27] (a) P.J. Toscano, E.J. Shermerhorn, E. Barren, S. Liu, J. Zubieta, J.
Coord. Chem. (1998) 169;
[33] A. Lehtonen, R. Sillanpaeae, Organometallics 24 (2005) 2795.
[34] K.C. Fortner, J.B. Bigi, S.N. Brown, Inorg. Chem. 44 (2005) 2803.
[35] R.H. Blessing, Acta Crystallogr., Sect. A 51 (1995) 33.
[36] G.M. Sheldrick, Acta Crystallogr., Sect. A 46 (1990) 467.
[37] G.M. Sheldrick, Program for the Refinement of Crystal Structures,
University of Go¨ttingen, Go¨ttingen, Germany, 1997.
[38] L.J. Farrugua, ORTEP-3 for Windows, Version 1.06, University of
Glasgow, Scotland, UK.
(b) M.H. Chisholm, J.-H. Huang, J.C. Huffman, W.E. Streib, D.
Tiedtke, Polyhedron 43 (1997) 2941;
(c) C. Floriani, F. Corazza, W. Lesueur, A. Chiesi-Villa, C. Guastini,
Angew. Chem. 101 (1989) 93.
[28] M. Kol, M. Shamis, I. Goldberg, Z. Goldschmidt, S. Alfi, E. Hayut-
Salant, Inorg. Chem. Commun. 4 (2001) 177.
[29] (a) Y. Kim, J.G. Verkade, Organometallics 21 (2002) 2395;
(b) W. Wang, M. Fujiki, K. Nomura, Macromol. Rapid Commun.
25 (2004) 504.
[39] H.J. Bernstein, C. Duque, G. Grossman, M. Molinaro, A. Mueller,
N. Salame, R. Sayle, P. Valadon, RASTOP, version 2.1 – September 14
(P. Valadon, P.O. Box 13014, La Jolla, CA 928064, USA), 2004.