Crystal structures and solution dynamics of monocyclopentadienyl titanium(IV) complexes bearing pendant ether and phosphanyl type functionalities

Article abstract of DOI:10.1016/S0277-5387(03)00407-8

Two novel half-sandwich Ti(IV) complexes, [η⁵:η ¹-O-C₅(CH₃)₄CH₂CH ₂OCH₃]TiCl₃ (3) and (η⁵:η ¹-P-C₅H₄CH₂CH₂PPh ₂)TiCl₃ (6), were prepared and structurally characterised. At elevated temperatures, complex 3 undergoes a conversion into [η ⁵:σ-C₅(CH₃)₄CH ₂CH₂O-]TiCl₂ (4). The dynamic behaviour of complexes 3 and 6 in solutions has been studied by variable-temperature ¹H, ¹³C and ³¹P NMR spectroscopy. Thermodynamic parameters of the intramolecular dissociation-coordination processes for 3 and 6 were elucidated by the numerical analysis of the δ (T) dependencies. The intramolecular M(IV) ← E (M = Ti, Zr; E = O, P) coordination in half-sandwich cyclopentadienyl complexes is discussed.

Full text of DOI:10.1016/S0277-5387(03)00407-8

Polyhedron 22 (2003) 2885ꢀ2894
/
www.elsevier.com/locate/poly
Crystal structures and solution dynamics of monocyclopentadienyl
titanium(IV) complexes bearing pendant ether and phosphanyl type
functionalities
Dmitrii P. Krut’ko a,*, Maxim V. Borzov a, Eduard N. Veksler a,
Andrei V. Churakov b, Karel Mach c
a Department of Chemistry, M.V. Lomonosov Moscow State University, Leninskie Gory, Moscow 119899, Russia
b N.S. Kurnakov Institute of General and Inorganic Chemistry, Russian Academy of Science, Russia
c J. Heyrovsky´ Institute of Physical Chemistry, Academy of Sciences of the Czech Republic, Czech Republic
Received 13 March 2003; accepted 23 May 2003
Abstract
Two novel half-sandwich Ti(IV) complexes, [h5:h1-O-C5(CH3)4CH2CH2OCH3]TiCl3 (3) and (h5:h1-P-C5H4CH2CH2PPh2)TiCl3
(6), were prepared and structurally characterised. At elevated temperatures, complex 3 undergoes a conversion into [h5:s-
C5(CH3)4CH2CH2Oꢀ
1H, 13C and 31P NMR spectroscopy. Thermodynamic parameters of the intramolecular dissociation-coordination processes for 3
E (MꢁTi, Zr; EꢁO, P)
/]TiCl2 (4). The dynamic behaviour of complexes 3 and 6 in solutions has been studied by variable-temperature
and 6 were elucidated by the numerical analysis of the d(T) dependencies. The intramolecular M(IV)1
/
/
/
coordination in half-sandwich cyclopentadienyl complexes is discussed.
# 2003 Elsevier Ltd. All rights reserved.
Keywords: Titanium; Cyclopentadienyl ligands; Intramolecular coordination; O ligands; P ligands; Variable-temperature NMR spectroscopy
1. Introduction
During recent years our research team performed
systematic studies of Zr(IV) complexes derived from
C5H4CH2CH2PPh2 [4], C5(CH3)4CH2CH2OCH3 [5],
C5(CH3)4CH2CH2SCH3 [6], C5(CH3)4CH2CH2N(CH3)2
[7], C5(CH3)4CH2CH2P(CH3)2 [7] and C5(CH3)4-
CH2CH2PPh2 [7,8] ligands. Here we report our results
on the synthesis, structural and dynamic behaviour
investigation of two novel mono-cyclopentadienyl com-
plexes of titanium(IV): (h5:h1-C5H4CH2CH2PPh2)TiCl3
and (h5:h1-C5(CH3)4CH2CH2OCH3)TiCl3.
The cyclopentadienyl complexes of transition metals
possessing chelating heterofunctional side-chain groups
form an important class of organometallic compounds
and are now under intense investigation (see recent
reviews [1ꢀ3]). Among these compounds, monocyclo-
/
pentadienyl-type complexes of the Group 4 metals
attract a particular interest, mostly due to their known
ability to be applied as precursors of catalysts for a-
olefin polymerisation.
Besides this practical interest, the Group 4 metal half-
sandwiches exhibit a rich coordination chemistry. In
many cases, peculiarities of the spatial organisation of
these complexes and details of their dynamic behaviour
in solutions are unpredictable and are of common
theoretical value.
2. Experimental
2.1. General remarks
All procedures were performed in sealed-off evacu-
ated glass vessels. The employed solvents (and their
perdeuterated analogues) were dried with and distilled
from conventional agents (namely: diethyl ether and
* Corresponding author. Tel.: ꢂ
8846.
E-mail address: kdp@org.chem.msu.su (D.P. Krut’ko).
/
7-095-939-1234; fax: ꢂ7-095-932-
/
THF*with sodium benzophenone ketyl; toluene, hep-
/
0277-5387/03/$ - see front matter # 2003 Elsevier Ltd. All rights reserved.
doi:10.1016/S0277-5387(03)00407-8

2886
D.P. Krut’ko et al. / Polyhedron 22 (2003) 2885ꢀ2894
/
tane and pentane*
/
with Naꢀ
/
K alloy; CH2Cl2*
/
with
m/z (%): 252 (18.5) [Mꢂ+], 237 (0.8) [Mꢂ+ꢃ
(4.6) [Mꢂ+ꢃCH2OCH3], 178 (4.0) [Mꢂ+ꢃ
HSi(CH3)3],
148 (16.6) [Mꢂ+ꢃ
CH3OSi(CH3)3], 133 (100)
/
CH3], 207
P2O5 and then with CaH2). When performing proce-
dures in evacuated vessels, the degassed solvents were
stored in evacuated reservoirs over the corresponding
drying agent and transferred on a high vacuum line
directly into reaction vessels by trapping with liq. N2.
Trimethylchlorosilane (Fluka) was refluxed with and
distilled from aluminium powder (high vacuum line);
titanium tetrachloride was refluxed over freshly reduced
copper powder, degassed and distilled on a high vacuum
/
/
/
[C7H4(CH3)3ꢂ], 119 (15.4) [C7H5(CH3)2ꢂ], 105 (7.0)
[C7H6(CH3)ꢂ], 91 (8.4) [C7H7ꢂ], 73 (49.9) [Si(CH3)3ꢂ],
59 (7.7) [CH2CH2OCH3ꢂ], 45 (20.1) [CH2OCH3ꢂ].
Anal. Found: C, 71.70; H 11.39. Calc. for C15H28OSi
(252.47): C, 71.36; H 11.18%. 1H NMR (400 MHz,
C6D6, 30 8C, d ppm): ꢃ
/
0.12 [br, Si(CH3)3], 1.16 [br,
CH3C(Si(CH3)3], 1.78, 1.82 (each br, ꢀ
/
CCH3), 2.28 [br,
line (Teflon grease must be applied). HC5(CH3)4ꢀ
CH2CH2OCH3 and (CH3)3SiꢀC5H4ꢀCH2CH2PPh2
/
OCH2CH2C(Si(CH3)3], 2.65, 2.83 (each br,
CCH2CH2O), 3.15 (br s, OCH3), 3.32 (br, OCH2).
ꢀ
/
/
/
were prepared accordingly to the reported procedures.
1H, 13C and 31P NMR spectra were recorded on a
Varian VXR-400 spectrometer at 400, 100 and 162
MHz, respectively. For 1H and 13C{1H} spectra, solvent
2.2.2. (h5:h1-O-C5H4CH2CH2OCH3)TiCl3 (3)
Silane 2 (2.34 g, 9.27 mmol) was dissolved in toluene
(50 ml), mixed rapidly with a solution of TiCl4 (1.76 g,
resonances [dHꢁ
/
7.15 and dCꢁ
/
128.0 (C6D6), dHꢁ
/
5.32
25.3 (THF-
and dCꢁ53.8 (CD2Cl2), dHꢁ
/
/
1.73 and dCꢁ
/
9.28 mmol) in toluene (10 ml) at ꢃ78 8C, then the
/
d8)] were used as internal reference standards. For
31P{1H} spectra, 85% H3PO4 was employed as an
external reference. For temperature calibration, the
standard methanol and ethyleneglycol samples were
used. The elemental analyses were performed on the
Carlo-Erba automated analyser. Mass spectra were
measured on Kratos-MS-890 and Varian MAT CH7a
Fa spectrometers.
mixture was allowed to warm up to r.t. and left
overnight. The mixture was then heated at 70 8C for 1
h, the solvent removed under high vacuum and the
residual solid recrystallised from ether. Yield of crude 3
(contaminated with unreacted TiCl4) 1.70 g (55.0%). A
part of the crude 3 was subjected to high-vacuum
sublimation (10ꢃ3 torr, 150ꢀ
/
170 8C). EI MS (70 eV)
Cl], 296 (8)
CH2OCH3], 179 (6)
m/z (%): 332 (1.5) [Mꢂ+], 297 (22) [Mꢂ+ꢃ
/
[Mꢂ+ꢃ
/
HCl], 252 (19) [Mꢂ+ꢃ
/
Clꢀ
/
2.2. Synthetic procedures
[C5(CH3)4CH2CH2OCH3ꢂ], 149 (6) [C7H4(CH3)2(O-
CH3)ꢂ], 148 (40) [C5(CH3)4CH2CH2ꢂ+], 147 (47)
[C7H3(CH3)4ꢂ], 135 (10) [C7H5(CH3)(OCH3)ꢂ], 134
2.2.1. (CH3)3SiꢀC5H4CH2CH2OCH3 (2)
/
A 1.6 M solution of n-BuLi in hexane (17.85 ml, 28.56
mmol) was added to a solution of cyclopentadiene
HC5(CH3)4CH2CH2OCH3 (5.15 g, 28.57 mmol) in
THF (60 ml) at 0 8C under stirring. The mixture was
allowed to warm up to room temperature (r.t.) and left
(54) [C5(CH3)4ꢀ
/
CH2ꢂ+], 133 (100) [C7H4(CH3)3ꢂ], 121
(33) [C7H6(OCH3)ꢂ], 119 (52) [C7H5(CH3)2ꢂ], 105 (18)
[C7H6(CH3)ꢂ], 91 (27) [C7H7ꢂ], 77 (14) [C6H5ꢂ], 45 (59)
[CH2OCH3ꢂ]. Anal. Found: C, 42.81; H, 5.83. Calc. for
1
C12H19Cl3OTi (333.52): C, 43.22; H, 5.74%. H NMR
(400 MHz, THF-d8, 30 8C, d ppm): 2.37, 2.39 [each s,
overnight. The resultant redꢀ/brown liquor was heated
3
6H, C5(CH3)4], 3.11 (t, 2H, JHH
on a water bath and concentrated by trapping of the
solvents into a vessel cooled with liq. N2 until the
precipitation of the solid from the hot solution started.
The mixture was allowed to crystallise for 1 day, the
white crystalline solid was filtered off, washed on a filter
two times with 60 ml portions of cold ether, dried on a
high-vacuum line and dissolved in 60 ml of THF (nearly
colourless solution). To this solution heated on a boiling
water bath, (CH3)3SiCl (3.60 g, 28.57 mmol) was added
ꢁ
/
6.1 Hz, CH2CH2O),
6.1 Hz, CH2O).
3
3.24 (s, 3H, OCH3), 3.51(t, 2H, JHH
ꢁ
/
13C NMR (100 MHz, THF-d8, 30 8C, d ppm): 14.44,
1
14.70 (each q, JCH
ꢁ129 Hz, C5(CH3)4), 30.37 (t,
/
1
130 Hz, CH2CH2O), 59.05 (q, JCH
1JCH
OCH3), 73.16 (t, JCH
ꢁ
/
ꢁ141 Hz,
/
1
ꢁ143 Hz, CH2O), 138.92, 139.26
/
(each s, C2ꢀ5), 140.31 (s, C1). 1H NMR (400 MHz,
CD2Cl2, 30 8C, d ppm): 2.39, 2.41 [each s, 6H,
3
C5(CH3)4], 3.10 (t, 2H, JHH
ꢁ6.4 Hz, CH2CH2O),
/
3
3.32 (s, 3H, OCH3), 3.61 (t, 2H, JHH
at once and the reaction mixture was heated at 90ꢀ
/
ꢁ6.4 Hz,
/
100 8C for 2 h. Formation of LiCl as well-formed
snow-white crystals was observed (LiCl exhibits the
inverted temperature dependence of solubility in THF;
the observed crystalline precipitate dissolves on cooling
the mixture back to the r.t.). The solvent and unreacted
trimethylchlorosilane was removed under high vacuum
and the residual yellow oil subjected to a bulb-to-bulb
high vacuum distillation, that gave 2.34 g of trimethyl-
silyl substituted cyclopentadiene 2 as a pale-yellow oil.
Yield 32.4%. A mixture of isomers. GC/MS EI (70 eV)
CH2O). 13C{1H} NMR (100 MHz, CD2Cl2, 30 8C, d
ppm): 14.75, 14.91 [C5(CH3)4], 29.41 (CH2CH2O), 59.71
(OCH3), 73.60 (CH2O), 139.02, 139.04 (C2ꢀ5), 140.67
1
(C1). H NMR (400 MHz, CD2Cl2, ꢃ
/
69 8C, d ppm):
6.4
Hz, CH2CH2O), 3.50 (s, 3H, OCH3), 4.27 (t, 2H,
3JHH 6.4 Hz, CH2O). 13C{1H} NMR (100 MHz,
CD2Cl2, ꢃ69 8C, d ppm): 14.76 [C5(CH3)4], 24.83
2.31, 2.38 (each s, 6H, C5(CH3)4), 2.93 (t, 2H, 3JHH
ꢁ
/
ꢁ
/
/
(CH2CH2O), 63.12 (OCH3), 80.60 (CH2O), 136.90,
137.50 (C2ꢀ5), 145.62 (C1).

D.P. Krut’ko et al. / Polyhedron 22 (2003) 2885ꢀ
/2894
2887
2.2.3. (h5:h1-P-C5H4CH2CH2PPh2)TiCl3 (6)
temperature (80ꢀ
/
90 8C) giving the intermediate silane 2.
Solutions of (3.39 g, 9.67 mmol) silane 5 (3.39 g, 9.67
mmol) in toluene (10 ml) and TiCl4 (1.47 g, 7.70 mmol)
in 20 ml of the same solvent were mixed at r.t. A deep
red (purple) coloration appeared immediately. In 15 min
the first orange crystals appeared. The reaction mixture
It is noteworthy that all attempts to perform the same
silylation reaction in diethyl ether failed. Silane 2 was
additionally purified by a high-vacuum distillation. A
moderate yield of silane 2 [32% with respect to
cyclopentadiene HC5(CH3)4CH2CH2OCH3] is due to
the fact that the cyclopentadiene itself, if prepared
according to literature procedure [9], is actually a crude
product. Attempts to purify it by high-vacuum distilla-
tions gave poor results. Thus, isolation and purification
of the lithium derivative 1 appeared to be inevitable.
Treatment of silane 2 with TiCl4 in toluene under mild
was left overnight, the bulky yellowꢀorange crystalline
/
precipitate was filtered off, washed with 30 ml of
toluene, recrystallised from 10 ml of THF (slow cooling
of the solution from 100 to 0 8C), washed with 30 ml of
ether, and dried under high vacuum. Yield 2.24 g,
55.2%. An adduct with one molecule of toluene (even
if recrystallised from THF). EI MS (70 eV) m/z (%): 430
conditions (ꢃ78 to 20 8C) produced the desired half-
/
(not observed) [Mꢂ+], 395 (86) [Mꢂ+ꢃ
/
Cl], 360 (6)
[Mꢂ+ꢃ2Cl], 277 (100) [C5H4CH2CH2PPh2ꢂ], 200 (7)
sandwich complex 3. Compound 3 was obtained as a
bright-red crystalline solid, readily soluble in polar
halocarbon solvents, THF and toluene. However, the
product 3 is contaminated by residual non-reacted
TiCl4. Thus, in CD2Cl2 or CDCl3 solutions, two sets
of proton signals, that correspond to [h5:h1-O-
C5(CH3)4CH2CH2OCH3]TiCl3 (3) and its adduct with
TiCl4 were observed. In THF-d8, however, the same
crude 3 exhibits only one set of signals, that is indicative
/
[C5H4CH2CH2PPhꢂ+], 199 (13) [CH2PPh2ꢂ], 185 (28)
[PPh2ꢂ], 183 (60) [C12H8Pꢂ, (9-phosphafluorene-H)],
121 (49) [CHꢀ
/
PPhꢂ], 108 (34) [PPhꢂ+], 91 (47)
[C7H7ꢂ+], 77 (21) [C6H5ꢂ]. Anal. Found: C, 59.51; H,
5.17. Calc. for C26H26Cl3PTi (523.70): C, 59.63; H,
1
5.00%. H NMR (400 MHz, CD2Cl2, 30 8C, d ppm;
toluene signals are omitted): 3.20ꢀ3.29 (m, 4H,
/
3ꢂ4
CH2CH2P), 6.67 (d of virtual t, 2H,
5JHP 1.2 Hz, H2,5), 6.83 (virtual t, 2H,
Hz, H3,4), 7.39ꢀ
J
HH
3ꢂ4
ꢁ
/
5.2 Hz,
of the decomposition of the 3×
/TiCl4 adduct in a
ꢁ
/
J
ꢁ5.2
/
solvating medium.
HH
/
7.48 (m, 6H, m- and p-H in PPh2), 7.66
The final purification of 3 from TiCl4 can be achieved
by a high-vacuum sublimation. However, if overheated
under lower vacuum conditions, complex 3 converts into
an ancillary cyclopentadienyl-alkoxide complex [h5:s-
O-C5(CH3)4CH2CH2O-]TiCl2 (4), with an equivalent of
CH3Cl evolved (see Scheme 2).1
A similar process of the thermal interconversion was
observed formerly by J.H. Teuben et al. [10] for a close
analogue of 3, [h5:h1-O-C5(CH3)4CH2CH2CH2-
OCH3]TiCl3, at 225 8C in toluene solution.
(m, 4H, o-H in PPh2). 13CNMR (100 MHz, CD2Cl2,
1
2
133 Hz, JCP
30 8C, d ppm): 24.98 (dt, JCH
Hz, CH2CH2P), 34.07 (dt, JCH
ꢁ
ꢁ
/
ꢁ
/
10.7
23.6
1
1
135 Hz, JCP
/
ꢁ
/
1
Hz, CH2P), 123.47, 123.50 (each d, JCH
ꢁ
/
180 Hz,
8.6 Hz, m-C in
161 Hz, p-C in PPh2), 133.13
8.1 Hz, o-C in PPh2), 133.95
C2ꢀ5), 129.51 (dd, 1JCH
ꢁ
/
161 Hz, 3JCP
ꢁ
/
1
PPh2), 130.92 (d, JCH
ꢁ
/
(dd, 1JCH
(d, 1JCP
ꢁ
/
160 Hz, 2JCP
ꢁ
/
ꢁ
/
28.0 Hz, ipso-C in PPh2), 144.16 (d, 3JCP
Hz, C1). 31P-{1H} NMR (162 MHz, CD2Cl2, 30 8C, d
ppm): 27.1 (s).
ꢁ8.4
/
3.1.2. X-ray structural investigation of 3
2.3. X-ray crystallographic study
A single crystal of 3 suitable for X-ray analysis was
obtained by a high-vacuum sublimation. Contrarily to
what observed for its Zr(IV) analogue, which was
proved to form a dimeric (m-Cl)2 bridged structures
with the metal centre possessing a distorted octahedral
coordination [5], in complex 3 the Ti(IV) centre co-
ordination polyhedron is a tetragonal pyramid assuming
Cp to occupy one coordination site (see Fig. 1).
Crystal data and data-collection information as well
as a description of the structural analysis and refinement
for compounds 3 and 6 are given in Table 1.
3. Discussion
3.1. O-Functionalized compounds
1
Analyses data for (h5:s-C5(CH3)4CH2CH2Oꢀ)TiCl2 (4): 1H
/
3.1.1. Synthesis
NMR (400 MHz, C6D6, 30 8C): dꢁ
/
1.89, 1.99 (both s, 12H,
The titanium(IV) half-sandwich complex (h5:h1-
C5(CH3)4CH2CH2OCH3)TiCl3 (3) was prepared in two
steps starting from lithium cyclopentadienide (1) via
trimethylsilylated cyclopentadiene (2) in a good overall
yield (see Scheme 1).
3
(CH3)4C5); 2.55 (t, 2H, JHH
3JHH
30 8C): dꢁ
(OCH2), 133.55 (H3Cꢀ
141.32 (ꢀH2Cꢀ
Cring). MS EI (70 eV) m/z (%): 282 (14) [Mꢂ+], 254
(71) [Mꢂ+ꢃCH2CH2], 252 (100) [Mꢂ+ꢃCH2O], 217 (7) [Mꢂ+ꢃ
CH2Oꢀ
ꢁ/
7.1 Hz, CH2CH2OTi), 5.17 (t,, 2H,
ꢁ/
7.1 Hz, CH2CH2OTi); 13Cꢀ{1H} NMR (100 MHz, C6D6,
/
/
12.78, 13.13 ((CH3)4C5); 28.76 (CH2CH2OTi), 94.65
/Cring, the second signal is not observed),
/
/
/
/
/
Initially cyclopentadiene [9] and its lithiated derivative
1 [5] were prepared as reported previously. Salt 1
smoothly reacts with (CH3)3SiCl in THF at an elevated
/
Cl], 216 (20) [Mꢂ+ꢃ
/
CH2Oꢀ
/
HCl], 134 (11) [(CH3)4C5 ꢀ
/
CH2ꢂ+], 133 (13) [C7H4(CH3)3ꢂ+], 119 (22) [C7H5(CH3)2ꢂ+], 91 (10)
[C7H6(CH3)ꢂ+].

2888
D.P. Krut’ko et al. / Polyhedron 22 (2003) 2885ꢀ2894
/
Table 1
Crystal data, data collection, structure solution and refinement parameters for complexes 3a and 6
Compound
[h5:h1-O-C5(CH3)4CH2CH2OCH3]TiCl3 (3a)
12H19Cl3OTi
[h5:h1-P-C5H4CH2CH2PPh2]TiCl3×
1/2CH2Cl2 (6)
C19.5H19Cl4PTi
/
Empirical formula
Formula weight
Color, habit
C
333.52
red needle
474.02
red block
Crystal size (mm)
Crystal system
Space group
0.5ꢄ
orthorhombic
Pna21
/
0.2ꢄ
/
0.1
0.5ꢄ
monoclinic
C2/c
/
0.4ꢄ0.2
/
Unit cell dimensions
˚
a (A)
˚
14.562(5)
8.267(4)
12.338(13)
15.793(4)
7.947(2)
32.77(3)
92.78(6)
4109(4)
8
b (A)
˚
c (A)
b (8)
ꢃ3
˚
V (A
)
1485(2)
4
Z
Dcalc (g cmꢃ3
Absorption coefficient
)
1.491
1.098
1.533
1.016
(mmꢃ1
F(000)
)
688
1928
Diffractometer
Temperature (K)
˚
Radiation (l (A))
Enrafꢀ
/
Nonius CAD-4
Enrafꢀ
/
Nonius CAD-4
293
graphite monochromated Mo Ka (0.71073)
2.80 to 24.94
293
graphite monochromated Mo Ka (0.71073)
2.49 to 24.97
u Range (8)
Index ranges
ꢃ
/
35
/
h 5
/
17, ꢃ
/
95/k 5
/
2, ꢃ
/
145
/
l 5/3
ꢃ
/
185
/
h 5
/
18, ꢃ
/
25/k 5
/
9, ꢃ
/
45
/
l 538
/
Reflections collected
Independent reflections
Absorption correction
Min. and max. transmission
Solution method
2793
1753 (Rint
none
5520
3614 (Rint
ꢁ/
0.0247)
ꢁ/0.0253)
empirical
0.5070 and 0.5731
direct methods (SHELX-86) [16]
full-matrix least-squares on F2
all H atoms were placed in calculated positions and refined all H atoms were placed in calculated positions and
direct methods (SHELX-86) [16]
full-matrix least-squares on F2
(SHELXL-93) [17]
Refinement method
Hydrogen treatment
(
SHELXL-93) [17]
using a riding model
1714/1/160
1.075
refined using a riding model
3423/0/307
1.054
Data/restraints/parameters
Goodness-of-fit on F2
Final R indices [Iꢀ
R indices (all data)
Absolute structure parameter 0.01(4)
Extinction coefficient 0.0057(11)
Largest difference peak and 0.479 and ꢃ
/
2s(I)]
R1ꢁ
/
0.0257, wR2ꢁ
/
0.0631
0.0711
R1ꢁ
/
0.0389, wR2ꢁ
/
0.1025
R1ꢁ
/
0.0410, wR2ꢁ
/
R1ꢁ
/
0.0788, wR2ꢁ
/
0.1231
ꢀ/
ꢀ/
/
0.253
0.561 and ꢃ0.386
/
ꢃ3
˚
hole (e A
)
Scheme 2.
C5H4CH2CH2OCH3]TiCl3 [11]. However, compara-
tively to [h5:h1-O-C5H4CH2CH2OCH3]TiCl3 all the
distances in the pseudo-five-membered metallocycle
Tiꢀ
[Tiꢀ
1.442(24) A, C(6)ꢀ
1.448(4) vs. 1.397(20) A, Oꢀ
/
Cpcent
ꢀ
/
C(6)ꢀ
/
C(7)ꢀ
/
Oꢀ(Ti) in 3 are somewhat longer
/
Scheme 1.
˚
/
Cpcent 2.046 vs. 2.028 A, C(1)ꢀC(6) 1.503(5) vs.
/
˚
˚
/
C(7) 1.510(6) vs. 1.477(27) A, C(7)ꢀO
/
˚
/Ti 2.295(2) vs. 2.214(10)
The oxygen moiety occupies one of the positions at
the base of the pyramid. In general features, the
molecular structure of 3 is close to that observed for
its ring non-permethylated analogue [h5:h1-O-
2
˚
A]. Meanwhile, the average TiꢀCl distance is nearly the
/
5
1
˚
˚
same [2.305(1) A in 3 vs. 2.308(2) A in (h :h -O-
C5H4CH2CH2OCH3)TiCl3]. The sum of the angles

D.P. Krut’ko et al. / Polyhedron 22 (2003) 2885ꢀ
/2894
2889
Scheme 3.
signals is observed, that is indicative of a fast exchange
process (NMR time-scale). However, the values of the
chemical shifts are strongly temperature dependent (see
Fig. 2).
The evaluation of the thermodynamic parameters was
performed using Eq. (1),
ꢀ
ꢁ
[free]
DS DH
dcoord ꢃ d(T)
d(T) ꢃ dfree
K ꢁ
ꢁexp
ꢃ
ꢁ
(1)
[coord]
R
RT
Fig. 1. Molecular structure of complex 3; displacement ellipsoids are
shown at 50% probability level; hydrogen atoms are omitted for
where [coord] and [free] are the molar concentrations of
chelate and non-chelate forms of 3 (3a and 3b, respec-
tively), dcoord and dfree are the values of the chemical
shifts of ‘pure’ 3a and 3b, and d(T) is the observed
chemical shift value at temperature T. To perform the
evaluation of DS and DH, the improved r.m.s. non-
linear curve fitting procedure for d(T) function was
applied (see Section 5 for the method modification and
computational details). In each series, all four para-
meters (dcoord, dfree, DS and DH) were varied. Both of
the values (d and T) were treated as measured with
finite accuracies sd and sT, respectively, that were
assumed to be one and the same within the entire region
of interest. The accuracies (r.m.s. deviations) sd and sT,
were also estimated directly from the empirical data.
Remarkably, these estimates are in excellent agreement
with the values of the corresponding instrumental errors
typical for variable-temperature NMR experiments. The
numerical results for the thermodynamic parameters for
3 are given in Table 2.
˚
clarity; selected bond lengths (A) and angles (8): Tiꢀ
/
Cl(3) 2.293(2), Oꢀ
C(7) 1.448(4), TiꢀCl(2) 2.327(1),
O 2.295(2), C(6)ꢀC(7) 1.510(6); Cl(3)ꢀTiꢀ
OꢀC(7) 110.7(3), Cl(3)ꢀTiꢀCl(2) 88.04(5),
Ti 120.5(2), Cl(1)ꢀTiꢀCl(2) 88.65(4), C(7)ꢀOꢀTi 114.9(2),
Tiꢀ 153.04(8), C(1)ꢀC(6)ꢀC(7) 110.2(3), Cl(3)ꢀTiꢀ
79.56(8), OꢀC(7)ꢀC(6) 106.1(3), Cl(1)ꢀTiꢀO 79.10(7), TiꢀCpcent
2.046.
/
C(8) 1.439(5), Tiꢀ
C(1)ꢀC(6) 1.503(5), Tiꢀ
Cl(1) 124.44(6), C(8)ꢀ
C(8)ꢀOꢀ
Cl(2)
/
Cl(1) 2.294(1), Oꢀ
/
/
/
/
/
/
/
/
/
/
/
/
/
/
/
/
/
ꢀ/
/
O
/
/
/
/
O
/
/
/
/
/
around the O atom in 3 is 346.1(4)8 which is somewhat
less than in (h5:h1-O-C5H4CH2CH2OCH3)TiCl3
(351.58).3 This means that the oxygen atom in 3 is
shifted more strongly out of the Tiꢀ
/
C(7)ꢀ
/
C(8) plane.
Along with the shorter Tiꢀ
/
O distance in (h5:h1-O-
C5H4CH2CH2OCH3)TiCl3, this is indicative of a weaker
effect of the back donation of the second lone-pair of the
oxygen atom to the Ti centre in 3 in comparison to its
ring non-permethylated analogue (vide infra).
3.1.3. Variable-temperature NMR spectroscopy
investigation of the dynamic behaviour of 3 in solutions
Similarly to results reported earlier by Zeijden et al.
[12], in a non-solvating medium, complex 3 exhibits a
dynamic behaviour due to the reversible intramolecular
ether group to metal centre dissociation/coordination
(see Scheme 3).
As one can see from Table 2, the found values for dfree
and dcoord are in good agreement with the independent
data available from the direct measurements. Thus, for a
non-coordinated ꢀ
THF-d8 can serve as a good comparison point [30 8C,
d(1H)ꢁ3.24 (OCH3), 3.51 (OCH2); d(13C)ꢁ
59.09
/CH2OCH3 moiety, chemical shifts in
/
/
In 1H and 13C{1H} NMR spectra, the lower-field
(OCH3), 73.16 (OCH2)]. As for the chelate form, the
chemical shifts of protons and carbon atoms of [h5:h1-
O-C5(CH3)4CH2CH2OCH3]ZrCl3 [5] measured in
shifts of the signals corresponding to the ꢀ
/
OCH2ꢀ
/
and
OCH3 groups are a good criterion for the coordination
ꢀ
/
CD2Cl2 were considered [30 8C, d(1H)ꢁ
/
3.70 (OCH3),
63.95 (OCH3), 82.90 (OCH2)].
of the oxygen atom to the titanium centre. Within the
whole temperature range studied, no broadening of the
4.33 (OCH2); d(13C)ꢁ
/
The values of parameters DH and DS drawn from the
d(13C) dependencies are in a better mutual agreement
than those obtained from the d(1H) ones. In practice,
the equilibrium process presented in Scheme 3 is more
complicated and one should consider that the solvent
molecules also participate in it. Unlike 13C chemical
shifts, the proton chemical shifts are markedly influ-
2
In the crystal, [h5:h1-O-C5H4CH2CH2OCH3]TiCl3 possesses two
crystallographically independent units in the cell, with C(6) and C(7)
carbons disordered in one of them. For comparison, the parameters
are taken only for the molecule that possesses no disordered atoms.
3
Hereinafter, the X-ray structural data not given in original papers
were retrieved from CCDC.

2890
D.P. Krut’ko et al. / Polyhedron 22 (2003) 2885ꢀ2894
/
Fig. 2. The optimised d(T) dependencies for compound 3. The plots correspond to the series for nuclei indicated in italic: (a) upper curve OCH3,
lower curve CH2OCH3; (b) upper curve OCH3, lower curve CH2OCH3.
Table 2
The thermodynamic parameters for 3a and 3b interconversion process
(in CD2Cl2). The calculated values dfree and dcoord are also presented
dicative of the lower Tiꢀ
that in 3. However, comparison of the Tiꢀ
lengths implies an opposite conclusion.
The point is, however, that the DH value corresponds
to the entire process and reflects not only the formation
/
O bond strength compared to
/
O bond
Curve
DH (kJ
molꢃ1
DS (J molꢃ1 dfree (ppm)
Kꢃ1
dcoord (ppm)
)
)
d(1H)
21.8 (0.9)
25.4 (0.9)
19.7 (1.0)
20.4 (1.0)
86.9 (3.8)
101.0 (3.7)
78.6 (4.2)
80.7 (4.2)
3.492 (0.014) 4.341 (0.010)
3.294 (0.003) 3.509 (0.002)
of the Tiꢀ
molecular geometry. These changes involving Tiꢀ
and TiꢀCl (averaged) distances are given in Table 3. As
/O coordination bond, but all changes in the
(OCH2)
d(1H)
/Cpcent
/
(OCH3)
d(13C)
(OCH2)
d(13C)
(OCH3)
the X-ray structural data for both 3b and the non-
chelated form of (h5-C5H4CH2CH2OCH3)TiCl3 are not
available, the analogous parameters for the structurally
similar (h5-C5H4CH3)TiCl3 and (h5-C5(CH3)4CH2-
CH3)TiCl3 were retrieved from the Cambridge Struc-
tural Database [13] and selected as the reference points.
The data in Table 3 show that the intramolecular
coordination of the ether functionality to the metal
72.02 (0.20)
58.95 (0.10)
81.58 (0.15)
63.55 (0.07)
enced by the magnetic anisotropy effects induced by the
solvent molecules. Thus, from the viewpoint of 1H
NMR spectroscopy, the equilibrium in Scheme 3 cannot
be treated as a strictly two-site exchange process and Eq.
(1) does not work well in this case. Contrarily, the 13C
NMR method allows treatment of the equilibrium in
question as the true two-site exchange process. Thus, the
DH and DS values obtained from the analysis of the
d(13C) series seem to be more reliable.
centre results in the elongation of both the Tiꢀ
/Cpcent
and TiꢀCl distances. However, while the Tiꢀ
/
/
Cpcent
distance elongations are approximately of the same
˚
value in both of the cases (ꢂ
/
0.024 and ꢂ0.026 A),
/
elongation of the TiꢀCl bonds are noticeably greater for
/
the non-methylated complex (h5:h1-O-C5H4CH2-
CH2OCH3)TiCl3 than for its ring-permethylated analo-
gue 3a (ꢂ
/
0.085 against ꢂ0.062, respectively). Thus, the
/
smaller absolute value DH measured for (h5:h1-O-
C5H4CH2CH2OCH3)TiCl3 comparatively to that for
Conclusions that one could draw from comparison of
the thermodynamic values for 3 with those reported by
Zejden et al. [12] for (h5:h1-O-C5H4CH2CH2O-
3a can be due to the greater weakening of the Tiꢀ
bonds in this first complex.
In conclusion, it should be emphasised that the
/
Cl
CH3)TiCl3 along with a comparison of the TiꢀO bond
/
lengths, at a first glance, are a crude contradiction. For
the ring non-permethylated half-sandwich, the enthal-
pies determined by the numerical analysis of the d(T)
dependencies are the following: 16.8 (2) kJ molꢃ1 for
the (OCH2) series and 17.6 (5) kJ molꢃ1 for the (OCH3)
one [12]. As one can conclude first, the lower enthalpy
values for (h5:h1-O-C5H4CH2CH2OCH3)TiCl3 are in-
stability of the chelated form 3a is lower compared to
zirconium
its
analogues
(h5:h1-O-C5R4CH2-
CH3 [5], H [12]); the latter
CH2OCH3)ZrCl3 (Rꢁ
/
compounds have their methoxy group always coordi-
nated to the metal atom both in crystalline state and in
solutions within the temperature range studied.

D.P. Krut’ko et al. / Polyhedron 22 (2003) 2885ꢀ
/2894
2891
Table 3
Comparison of the Tiꢀ
/
Cpcent and Tiꢀ
/
Cl distances for [h5:h1-O-C5R4-CH2CH2OCH3]TiCl3 (Rꢁ
H, CH3) and their non-chelate relatives
/
Distances
Ring non-permethylated
Ring-permethylated
[h5-C5H4CH3]-TiCl3
[h5-C5H4-CH2CH2OCH3]TiCl3
[h5-C5(CH3)4-CH2CH3]TiCl3
3a
0.024) a
0.085)
2.020
2.243
2.046 (ꢂ
2.305 (ꢂ0.062)
/
0.026)
˚
Cpcent (A)
Cl b (A)
˚
Tiꢀ
Tiꢀ
/
2.004
2.223
2.028 (ꢂ
2.308 (ꢂ
/
/
/
/
a
The bond length elongation caused by the intramolecular coordination are given in brackets.
Averaged distances.
b
3.2. P-Containing compounds
complexes exhibiting the same spatial configuration as
6. Comparison of the bond angles in the coordination
environment of the phosphorous atom in 6 and its
zirconium analogue (h5:h1-P-C5H4CH2CH2PPh2)-
3.2.1. Synthesis
Similarly to complex 3 the half-sandwich (h5:h1-P-
C5H4CH2CH2PPh2)TiCl3 (6) was prepared via the
trimethylsilyl substituted cyclopentadiene 5, as it is
depicted in Scheme 4.
ZrCl3×
/THF indicates that the pseudo five-membered
metallocycle in 6 is less constricted spatially. Indeed, the
range of the angles at the P-atom in the titanium
complex 6 is 198 (100.28ꢀ
/
119.28) while in the zirconium
Silane 5 smoothly reacts with TiCl4 in toluene at room
temperature to give a poorly-soluble product separating
from the reaction mixture as a finely-crystalline volu-
minous orange ‘wool’ sensitive to air and moisture. It is
remarkable that the composition of 6 as an adduct with
complex this range is 238 (101.98ꢀ
/
124.78) [4].
3.2.3. Variable-temperature NMR spectroscopy
investigation of the dynamic behaviour of 6 in solutions.
Comparison with that for its Zr-analogue
Complex 6 was studied by 31P{1H} NMR spectro-
scopy in both solvating (THF-d8) and non-solvating
(CD2Cl2) media within a broad range of temperatures.4
The lower field shifting of d(31P) values is a good
toluene (h5:h1-P-C5H4CH2CH2PPh2)TiCl3×
/
C6H5CH3
1
(analysis of the H and 13C{1H} NMR spectral data)
surprisingly does not change after crystallisation from a
minimum amount of hot THF. However, if crystallised
from CH2Cl2, no toluene is included into the crystal
lattice and its signals are not observed in the NMR
spectra.
3.2.2. X-ray structural investigation of 6
Single crystals of 6 were obtained by crystallisation
from dichloromethane as an adduct with 1/2 molecule of
CH2Cl2 lying on crystallographic two-fold axes. In the
crystalline state, complex 6 exhibits the same coordina-
tion polyhedron of the metal centre as that was observed
for 3, i.e. the ‘four-leg piano stool’ in which the
phosphorous atom occupies one of the positions at the
base of the pyramid. The molecular structure of 6 is
given in Fig. 3. Of interest, its formerly reported
zirconium analogue (h5:h1-P-C5H4CH2CH2PPh2)-
ZrCl3×
/
THF possesses the six-coordinated metal centre,
with the THF molecule occupying one of the apical
positions of the distorted tetragonal bipyramid [4].
˚
The TiꢀP bond length (2.627(2) A) is slightly longer
/
5
˚
than in (h -C5H5)TiCl3[P(CH3)3] (2.604(3) A) [14] and
Fig. 3. Molecular structure of complex 6; displacement ellipsoids are
shown at 50% probability level, hydrogen atoms and solvent dichlor-
nearly the same as in (h5-C5H5)TiCl2[h1:h1-O,P-
OC6H3(2-Bu )(6-PPh2)] (2.624(3) A) [15], with both
t
˚
˚
omethane molecule are omitted for clarity; selected bond lengths (A)
and angles (8): TiꢀCl(1) 2.310(2), PꢀC(7) 1.837(4), TiꢀCl(2) 2.326(1),
PꢀC(11) 1.817(3), TiꢀCl(3) 2.329(2), PꢀC(21) 1.819(3), TiꢀP 2.627(2);
Cl(1)ꢀTiꢀCl(2) 87.19(5), C(11)ꢀPꢀC(21) 100.2(2), Cl(1)ꢀTiꢀCl(3)
129.63(6), C(11)ꢀPꢀC(7) 106.7(2), Cl(2)ꢀTiꢀCl(3) 88.31(5), C(21)ꢀ
PꢀC(7) 107.5(2), Cl(1)ꢀTiꢀP 80.30(6), C(11)ꢀPꢀTi 118.6(1), Cl(2)ꢀ
TiꢀP 148.07(4), C(21)ꢀPꢀTi 119.2(1), Cl(3)ꢀTiꢀP 77.70(5), C(7)ꢀPꢀ
Ti 103.8(2), C(1)ꢀC(6)ꢀC(7) 114.9(3), C(6)ꢀC(7)ꢀP 110.8(3), Tiꢀ
Cpcent 2.035.
/
/
/
/
/
/
/
/
/
/
/
/
/
/
/
/
/
/
/
/
/
/
/
/
/
/
/
/
/
/
/
/
/
/
/
/
Scheme 4.

2892
D.P. Krut’ko et al. / Polyhedron 22 (2003) 2885ꢀ2894
/
criterion for the coordination of the phosphanyl group
towards the metal centre, while the chemical shift of a
free ꢀ
/
CH2PPh2 group [d(31P)$
/
ꢃ14] [4] was chosen as
/
a comparison point.
In non-solvating CD2Cl2, no evidence of any dynamic
behaviour was found. The chemical shift of the PPh2
Scheme 5.
group changes slightly (from d(31P)ꢁ
/
27.1 at ꢂ
78 8C), with no considerable broad-
ening of signal observed. This is indicative of the
retention of Ti1P coordination in a non-solvating
/
30 8C to
d(31P)ꢁ
/
32.1 at ꢃ
/
solvent concentration.5 Actually, the estimated value of
[THF]/([THF]ꢂ[coord]ꢂ[free]) for the experimental
/
/
/
conditions is only 0.99 which causes only approximately
0.1 J molꢃ1 Kꢃ1 correction for the DS parameter.
Taking into account the final r.m.s. deviations for DS,
this correction can be easily neglected. Thus, in this case,
the computation of the parameters DS, DH, dfree and
dcoord was performed similarly to that for 3. The found
medium within the entire temperature range studied.
In an n-donor solvent, however, the situation changes
dramatically. The observed variation of d(31P) exceeds
30 ppm (from d(31P)ꢁ
29.7 at ꢃ78 8C, see Fig. 4).
/
ꢃ
/
2.1 at ꢂ
/
60 8C to d(31P)ꢁ
/
ꢂ
/
/
It is apparent, that in contrast to what was observed
for 3, the n-donor solvent molecule assists the equili-
brium process of the pendant phosphanyl group in-
tramolecular dissociation-coordination (see Scheme 5).
Accordingly to this Scheme 5, Eq. (1) should be
changed to
values are the following: DHꢁ
DSꢁ
76.5 (7.3) J molꢃ1 Kꢃ1, dfree
dcoord 29.55 (0.22). The more accurate and trust-
/
23.4 (2.3) kJ molꢃ1 Kꢃ1
17.41 (5.99),
,
/
ꢁ
/
ꢃ
/
ꢁ
/
worthy value for dcoord than for dfree (compare with
d(31P) of complex 6 in CD2Cl2 and d(31P) of the free
PPh2 group, vide supra), evidently, is due to the fact that
the empirical points could be in practice measured only
for the lower-temperature branch of the computed
curve, with the higher-temperature part of the curve
remaining unavailable because of the boiling point of
the solvent used.
nfree
[free]([THF] ꢂ [coord] ꢂ [free])
K ꢁ
nTHFncoord
ꢀ
[coord][THF]
ꢁ
DS DH
R
dcoord ꢃ d(T)
d(T) ꢃ dfree
ꢁexp
ꢃ
ꢁ
(2)
RT
The obtained data are worthwhile to be discussed
from two viewpoints. Their comparison for complexes 6
and (h5:h1-O-C5H4CH2CH2OCH3)TiCl3 [11,12] reveals
that the ether group exerts a lower affinity towards the
Ti(IV) centre than the phosphanyl group. The dissocia-
tion of the ether group takes place in the non-solvating
medium (CD2Cl2), while the PPh2 group in the same
solvent remains coordinated to the Ti(IV) centre and
exhibits hemilability only in THF. In a THF solution of
3, the coordination of an intramolecular ether moiety to
the Ti(IV) centre is not observed at all. Most likely, the
where ncoord, nfree and nTHF are the molar fractions of
the corresponding components in the equilibrium mix-
ture, square bracketed values are the corresponding
molar equilibrium concentrations of the chelated and
non-chelated forms of 6 (6a and 6b, respectively) and the
same
would
be
observed
for
C5H4CH2CH2OCH3)TiCl3, if its spectra in THF-d8
(h5:h1-O-
were measured.
The great positive value of DS observed for 6 [76.5
(7.3) J molꢃ1 Kꢃ1] (compare to 79 J molꢃ1 Kꢃ1 for 3
and
66
J
molꢃ1
Kꢃ1
for
(h5:h1-O-
C5H4CH2CH2OCH3)TiCl3 [12]) is also noteworthy.
One should keep in mind, that in the case of 6, the
solvent molecule is captured by the metal centre along
with the dissociation of the PPh2 group. This fact could
be accounted for by a considerable loss in entropy due
to the loss of the rotational freedom degrees of the
Fig. 4. The optimised d(T) dependencies for compound 6 in THF.
phenyl rings in the PPh2ꢀ/group when coordinated.
4
Series of the variable-temperature 1H and 13C{1H} NMR spectra
were also measured. However, the appearance of the signals in the ꢀ
/
CH2CH2PPh2 region is complicated due to the strong coupling with
the phosphorous nucleus. This made it impossible to measure d(1H)
and d(13C) with the required accuracy.
5
Application of the molar fractions makes the value K non-
dimensional.

D.P. Krut’ko et al. / Polyhedron 22 (2003) 2885ꢀ
/2894
2893
Comparison of 6 with its Zr analogue (h5:h1-P-
C5H4CH2CH2PPh2)ZrCl3 [4] indicates their markedly
different dynamic behaviour in the solvating media.
Thus, while in the Zr complex the coordinated PPh2
group is replaced by the second THF molecule with a
decrease of temperature (see Scheme 6), the same
decrease of temperature in the case of 6 leads to a
different result (see Scheme 5). Thus, the enthalpy
changes for the intramolecular coordination-dissocia-
tion of the PPh2 group in 6 and in (h5:h1-P-
C5H4CH2CH2PPh2)ZrCl3 have opposite signs. This
˚
THF (2.473(5) A) is shorter than that in the ring non-
6
˚
permethylated analogue (2.4866(3) A). This resembles
strongly what is observed for Ti complexes 3 and (h5:h1-
P-C5H4CH2CH2OCH3)TiCl3. It is illustrative for the
fact that the thermodynamic stability of the chelate form
of a complex may not correlate (as it is considered
usually) with the length of the metalꢀheteroelement
/
coordination bond. However, this last parameter may be
considered a good one for the determination of the
relative metalꢀheteroatom bond strengths within a row
/
of similarly constructed compounds.
allows one to conclude that the intramolecular Ti1
/
P
bond in 6 is more stable than the analogous Zr1P bond
/
in (h5:h1-P-C5H4CH2CH2PPh2)ZrCl3. Apparently, one
should consider that this comparison is to be done with
care because of the different coordination polyhedrons
in all of the forms of 6 and in those of (h5-
C5H4CH2CH2PPh2)ZrCl3. Nevertheless, the opposite
trends of the coordination abilities for ether and
phosphanyl groups towards Ti(IV) and Zr(IV) centres
are rather clear. Thus, while the Zr(IV) metal centre
exhibits a greater affinity towards the ether-type oxygen
atom than to P-atom in the phosphanyl group, for the
Ti(IV) metal centre this tendency is opposite.
4. Conclusion
The performed investigation allows to confirm, that in
solution, the chelate form of complexes (h5-O-
C5R4CH2CH2OCH3)MCl3 (Rꢁ
/
H, CH3; MꢁTi, Zr)
/
is more stable for the Zr complexes. Comparison of O-
and P-side-chain functionalized Ti(IV) and Zr(IV) half-
sandwich complexes reveals the opposite tendencies in
the stability of the chelate forms in solutions. While the
Zr(IV) metal centre exhibits a greater affinity to the
ether type oxygen atom, for the Ti(IV) metal centre the
intramolecular coordination with the phosphanyl group
appears to be more preferable. Finally, comparison of
the X-ray structural analysis data along with the
thermodynamic parameters of the intramolecular co-
ordination-dissociation equilibrium processes in solu-
Here it is illustrative to present our previously
unpublished data on the dynamic behaviour of (h5:h1-
P-C5(CH3)4CH2CH2PPh2)ZrCl3×
/
THF reported earlier
[8]. Unlike (h5:h1-P-C5H4CH2CH2PPh2)ZrCl3, ring-
permethylated Zr(IV) half-sandwich complex (h5:h1-P-
tion for both (h5:h1-O-C5R4CH2CH2OCH3)TiCl3 (Rꢁ
H, CH3) and (h5:h1-P-C5R4CH2CH2PPh2)ZrCl3×
THF
(RꢁH, CH3) revealed that the thermodynamic stability
of the chelate forms of these compounds do not
correlate directly with the metalꢀheteroatom coordina-
/
C5(CH3)4CH2CH2PPh2)ZrCl3×/THF exhibits no dy-
namic behaviour in all suitable solvents tested. The
d(31P) value remains nearly unchanged with a decrease
/
/
of temperature from 30 8C (2.9) to ꢀ
for THF-d8; in CD2Cl2 d(31P)ꢁ
5.8 at 30 8C). Thus, in
(h5:h1-P-C5(CH3)4CH2CH2PPh2)ZrCl3×
THF no substi-
tution of the PPh2 group with a second THF molecule is
observed as takes place in
(h5:h1-P-
C5H4CH2CH2PPh2)ZrCl3×THF. Meanwhile, the Zrꢀ
distance in (h5:h1-P-C5(CH3)4CH2CH2PPh2)ZrCl3×
/
87 8C (3.9) (the data
/
/
/
tion bond length. From our viewpoint, this is due to the
fact that the thermodynamics of the processes is
contributed not only by the formation-cleavage of the
/
/P
metalꢀheteroatom coordination bond, but to a larger
/
/
extent, by the electron and structural reorganisation of
the complex in general.
5
1
˚
THF (2.8906(11) A) [8] is greater than that in (h :h -
˚
THF (2.8474(5) A) [4].
P-C5H4CH2CH2PPh2)ZrCl3×
The distance from the Zr atom to the mean plane of
the Cp ring and the Zrꢀ
O(THF) bond length in [h5:h1-
P-C5(CH3)4CH2CH2PPh2]ZrCl3×THF [2.256(1) and
2.419(2) A, respectively] are also longer than those in
(h5:h1-P-C5H4CH2CH2PPh2)ZrCl3×
THF (2.236(1) and
2.3613(12) A, respectively). However, the average ZrꢀCl
distance in (h5:h1-P-C5(CH3)4CH2CH2PPh2)ZrCl3×
/
/
5. Supplementary material
/
˚
Supplementary material for this article include the
details of the numerical processing of the variable-
temperature NMR data and is available from authors
on request.
/
˚
/
/
Crystallographic data for the structural analysis have
been deposited with the Cambridge Crystallographic
6
For (h5-C5H4CH2CH2PPh2)ZrCl3×
THF the data are given only
/
for one of the crystalline modifications. In the second modification
that possesses two crystallographically independent molecules, the
analogous parameters are nearly of the same value.
Scheme 6.

2894
D.P. Krut’ko et al. / Polyhedron 22 (2003) 2885ꢀ2894
/
[3] C. Muller, D. Vos, P. Jutzi, J. Organometal. Chem. 600 (2000)
¨
127.
Data Centre, CCDC Nos. 194203 and 194204. Copies of
this information may be obtained free of charge from
The Director, CCDC, 12 Union Road, Cambridge, CB2
[4] D.P. Krut’ko, M.V. Borzov, E.N. Veksler, A.V. Churakov,
J.A.K. Howard, Polyhedron 17 (1998) 3889.
1EZ, UK (fax: ꢂ44-1223-336033; e-mail: deposit@
ccdc.cam.ac.uk or www: http://www.ccdc.cam.ac.uk).
/
[5] D.P. Krut’ko, M.V. Borzov, V.S. Petrosyan, L.G. Kuz’mina,
A.V. Churakov, Russ. Chem. Bull. 45 (1996) 1740.
[6] D.P. Krut’ko, M.V. Borzov, V.S. Petrosyan, L.G. Kuz’mina,
A.V. Churakov, Russ. Chem. Bull. 45 (1996) 940.
[7] D.P. Krut’ko, M.V. Borzov, E.N. Veksler, R.S. Kirsanov, A.V.
Churakov, D.A. Lemenovskii, Pure and Appl. Chem. 73 (2001)
367.
Acknowledgements
[8] D.P. Krut’ko, M.V. Borzov, E.N. Veksler, R.S. Kirsanov, A.V.
Churakov, Eur. J. Inorg. Chem. (1999) 1973.
[9] P. Jutzi, J. Dahlhaus, Synthesis (1993) 684.
[10] E.E.C.G. Gielens, J.Y. Tiesnitsch, B. Hessen, J.H. Teuben,
Organometallics 17 (1998) 1652.
The authors of the paper are deeply thankful to Prof.
Georgievskii D.V. (Dept. of Mechanics and Mathe-
matics, M.V. Lomonosov Moscow State University) for
his kind advices and help in preparing of the computa-
tional part (Section 5) of the paper. The authors greatly
acknowledge the Russian Foundation for Basic Re-
search for CSD licence payment (grant 02-07-90322) and
financial support for A.V.C. (grant 01-03-32474).
[11] Q. Huang, Y. Qian, G. Li, Y. Tang, Transition Met. Chem. 15
(1990) 483.
[12] A.A.H. van der Zeijden, C. Mattheis, R. Frohlich, Organome-
¨
tallics 16 (1997) 2651.
[13] F.H. Allen, O. Kennard, Chem. Autom. Des. News 8 (1993)
31.
[14] T.T. Nadasdi, D.W. Stephan, Inorg. Chem. 32 (1993) 5933.
[15] C.A. Willoughby, R.R. Daff, Jr, W.M. Davis, S.L. Buchwald,
Organometallics 15 (1996) 472.
References
[16] G.M. Sheldrick, Acta Crystallogr., Sect. A 46 (1990) 467.
[17] G.M. Sheldrick, SHELXL-93: Program for the Refinement of
Crystal Structures, University of Gottingen, Germany, 1993.
¨
[1] U. Siemeling, Chem. Rev. 100 (2000) 1495.
[2] H. Butenschon, Chem. Rev. 100 (2000) 1527.
¨