Preparation and crystal structures of the mononuclear vanadium phenoxide complexes [VCl(OC6H3Pri2-2,6) 2(C4H8O)2] and [VO(OC6H3Pri2-2,6)3]: Procatalysts for ethylene polymerisation

Article abstract of DOI:10.1016/S0022-328X(97)00774-2

The complexes [VCl(OC₆H₃Prⁱ₂-2,6) ₂(thf)₂] 1 (thf, tetrahydrofuran) and [VO(OC₆H₃Prⁱ₂-2,6)₃] 2 have been prepared and structurally characterised. Compound 1 has a distorted trigonal bipyramidal structure with apical thf ligands [d(V-O_phenox) 1.865(2), d(V-O_thf) 2.120(3) and d(V-Cl) 2.277(2) A]. It reacts with Li[SC₆H₂Me₃-2,4,6] to give [V(SC₆H₂Me₃-2,4,6)(OC₆H ₃Prⁱ₂-2,6)₂(thf)₂]. Compound 2, which has trigonal pyramidal geometry, is disordered about a 2-fold crystallographic axis [d(V=O) 1.564(4), d(V-O_av) 1.761(1) A]. Compounds 1 and 2 are pro-catalysts for the polymerisation of C₂H₄.

Full text of DOI:10.1016/S0022-328X(97)00774-2

Journal of Organometallic Chemistry 554 (1998) 195–201
Preparation and crystal structures of the mononuclear vanadium
phenoxide complexes [VCl(OC6H3Pri2-2,6)2(C4H8O)2] and
[VO(OC6H3Pri2-2,6)3]: procatalysts for ethylene polymerisation
Richard A. Henderson a, David L. Hughes a, Zofia Janas b, Raymond L. Richards a,
Piotr Sobota b,*, Stawomir Szafert b
a John Innes Centre, Nitrogen Fixation Laboratory, Norwich Research Park, Norwich NR4 7UH, UK
b Faculty of Chemistry, Uni6ersity of Wroctaw, 14 F. Joliot-Curie Street, Wroctaw 50383, Poland
Received 13 October 1997
Abstract
The complexes [VCl(OC6H3Pri2-2,6)2(thf)2] 1 (thf, tetrahydrofuran) and [VO(OC6H3Pr2i -2,6)3] 2 have been prepared and
structurally characterised. Compound 1 has a distorted trigonal bipyramidal structure with apical thf ligands [d(V–Ophenox
1.865(2), d(V–Othf) 2.120(3) and d(V–Cl) 2.277(2) A]. It reacts with Li[SC6H2Me3-2,4,6] to give [V(SC6H2Me3-2,4,6)(OC6H3Pr2-
)
i
˚
2,6)2(thf)2]. Compound 2, which has trigonal pyramidal geometry, is disordered about a 2-fold crystallographic axis [d(V=O)
˚
1.564(4), d(V–Oav) 1.761(1) A]. Compounds 1 and 2 are pro-catalysts for the polymerisation of C2H4. © 1998 Elsevier Science
S.A. All rights reserved.
Keywords: Vanadium; Phenoxide; Crystal structure; Polymerisation
We are interested in the catalytic properties of vana-
dium centres, particularly relating to nitrogen fixation
[1] and to polymerisation of alkenes [2]. As an aid to
understanding the role of vanadium in vanadium nitro-
genase, we are studying the binding of hydrazines and
derived ligands at low-valent vanadium centres and
have prepared complexes of the bulky phenoxide ligand
[OC6H3Pri2-2,6]−, which was used to favour the forma-
tion of mononuclear vanadium centres, e.g. in such
complexes as [V(OC6H3Pri2-2,6)3(NH2NMeR)2] (R=
Me or Ph) [1]. Whilst engaged in this work we discov-
ered routes to two new monomeric phenoxide
complexes and here we describe their structures and
some of their reactions, principally in polymerisation
studies.
1. Results and discussion
1.1. Preparation, structure and reactions of [VCl
(OC6H3Pri2-2,6)2(thf)2] 1
Compound 1 was obtained by the straightforward
reaction of [VCl3(thf)3] with Li[OC6H3Pri2-2,6] in thf
(Eq. (1)] as dark red, paramagnetic crystals with a
magnetic moment (veff=2.6 vB) in the range expected
for VIII complexes of this type with a d2 high-spin
configuration [1,3].
[VCl3(thf)3]+2Li[OC6H3Pri2-2,6]
“[VCl(OC6H3Pri2-2,6)2(thf)2]+2LiCl+thf
(1)
The crystal structure of compound 1 is shown in Fig.
1, with bond dimensions in Table 1. It has essentially
trigonal bipyramidal geometry, with apical thf ligands
* Corresponding author. E-mail: PLAS@WCHUVE.CHEM.UNI.
WROC.PH
0022-328X/98/$19.00 © 1998 Elsevier Science S.A. All rights reserved.
PII S0022-328X(97)00774-2

196
R.A. Henderson et al. / Journal of Organometallic Chemistry 554 (1998) 195–201
and the two OC6H3Pri2-2,6 groups and the chloride
lying in the trigonal plane. The OC6H3Pri2-2,6 ligands
have their aromatic groups (related by a 2-fold symme-
try axis) in the ‘up–down’ configuration relative to the
trigonal plane. Compound 1 is an analogue of the
complexes [V(OC6H3Me2-2,6)3(C5H5N)2] [3] and
[V(OC6H3Pri2-2,6)3(NH2NMeR)2] (R=Me or Ph) [1]
which have trigonal bipyramidal geometry with the
phenoxide ligands in the trigonal plane.
Table 1
Selected molecular dimensions in [VCl(OC6H3Pri2-2,6)2(thf)2] 1. Bond
˚
lengths are in A and angles in degrees with estimated standard
deviations in parentheses
(a) About the vanadium atom
V–Cl
V–O(1)
Cl–V–O(1)
Cl–V–O(2)
O(1)–V–O(1%)
2.277(2) V–O(2)
2.120(3)
1.865(3)
92.2(1)
117.4(1)
175.5
O(1)–V–O(2)
O(1)–V–O(2%)
O(2)–V–O(2%)
92.5(1)
85.5(1)
125.2(1)
Compound 1 reacts with Li[SC6H2Me3-2,4,6] to give
(b) In the ligands
O(1)–C(1)
[V(SC6H2Me3-2,4,6)(OC6H3Pri2-2,6)2(thf)2] (Eq. (2))
1.447(6) V–O(1)–C(1)
1.451(7) V–O(1)–C(4)
C(1)–O(1)–C(4)
124.2(3)
126.7(3)
109.2(4)
133.2(3)
O(1)–C(4)
[VCl(OC6H3Pr2i -2,6)2(thf)2]+Li[SC6H2Me3-2,4,6]
O(2)–C(5)
1.352(5) V–O(2)–C(5)
“[V(SC6H2Me3-2,4,6)(OC6H3Pri2-2,6)2(thf)2]+LiCl
(2)
phy as described below. The origin of the VꢀO group in
2 has not been established, but is very likely a result of
hydrolysis by adventitious water of the presumably very
reactive species formed after loss of N2 from an inter-
mediate produced during the early stages of the reac-
tion. There also must be a step in the reaction sequence
which generates the third OC6H3Pri2-2,6 group found at
vanadium in the product 2. Having established the
structure of compound 2, we sought a rational synthesis
of it and found that it can be obtained in high yield by
treatment of [VO(OPri)3] with HOC6H3Pri2-2,6 or treat-
ment of [V(OC6H3Pri2-2,6)4Li(thf)] with propylene oxide
(Scheme 1 and Section 2).
We attempted to prepare a dinitrogen-bridged com-
plex by treatment of 1 with N2(SiMe3)4 but dinitrogen
was lost in the reaction, leading to the formation of
[VO(OC6H3Pri2-2,6)3] 2, as described below.
1.2. Preparation and structure of [VO(OC6H3Pri2-2,6)3]
2
Compound 2 was obtained by three routes, the first
being accidental, as mentioned above. Treatment of
compound 1 with N2(SiMe3)4 involved loss of dinitro-
gen from a transient intermediate which could not be
isolated. Work up of the oily reaction product and
standing of the resulting solution at low temperature
for several days (Section 2) gave dark-red, diamagnetic
crystals which have an IR band at 1000 cm−1
assignable to (VꢀO) and have been structurally charac-
terised as [VO(OC6H3Pr2i -2,6)3] 2 by X-ray crystallogra-
The structure of compound 2 is shown in Fig. 2, with
molecular dimensions in Table 2. It has trigonal pyra-
midal geometry about vanadium with the vanadium
˚
atom 0.633(1) A out of the plane of the three phenoxide
oxygen atoms. This is a rare geometry for VV which has
previously been described for the analogues
[VO(OCH2CH2Cl)3] [4] and [VO(OC5H9)3] [5], although
in these cases each molecule interacts, via an alkoxide
oxygen atom, with a second molecule in the solid state
to give dimers with essentially trigonal bipyramidal
environments for the vanadium atoms; in these com-
plexes, the fifth V–O distance (opposite the vanadyl
˚
oxygen) is much longer, at 2.262(2) and 2.297(2) A,
˚
respectively, and the V atoms are 0.32 and 0.35 A out
of the ‘equatorial’ planes of three alkoxide oxygen
atoms. Evidently, the bulk of the OC6H3Pri2-2,6 ligands
in compound 2 prevents close intermolecular approach.
A monomeric analogue, which also has bulky ligands,
is [NbO(OC6H3Ph2-2,6)3] which shows no distant, fifth
coordinating oxygen atom; here the Nb atom is dis-
˚
placed 0.70(2) A from the plane of the O3 trigonal base
[6].
The molecule of 2 shows disorder about a crystallo-
graphic 2-fold symmetry axis which passes through the
ligand binding through O(1) and relates the ligands
binding through O(2) and O(2%); the vanadyl group is
randomly disordered between sites related by this axis
Fig. 1. The structure of
a
single molecule of [VCl(OC6H3Pri2-
2,6)2(thf)2] 1, the H atoms are omitted for clarity. A 2-fold symmetry
axis passes through V and Cl and relates the two halves of the
molecule.

R.A. Henderson et al. / Journal of Organometallic Chemistry 554 (1998) 195–201
197
Scheme 1. (i) N2(SiMe3)4, hexane; (ii) propylene oxide, hexane; (iii) HOC6H3Pri2-2,6, hexane.
(Fig. 3). There is no disorder in the phenyl rings on
the bridging O atom).
O(2) and O(2%) but there is disorder in some of the
isopropyl groups of those ligands (Fig. 2). These two
phenoxide ligands are orientated almost perpendicu-
larly to the O3 base plane so that their isopropyl groups
lie well above and below the base plane; the third
phenoxide ligand, of O(1), lies with its phenyl ring
plane much closer to the base plane and essentially
normal to the other two phenyl rings. The V–Ophenox
In compound 2, the V–O(1)–C(11) angle, 136.3(4)°,
is in the normal range for this type of bonding, but the
other V–O–C angles, both ca. 165°, are quite unex-
pectedly large and presumably result from the bulkiness
of the ligands. In [{VO(OC5H9)3}2], the V–O–C angles
are all ca. 131° and the O–C bonds, viewed down the
direction of the vanadyl bond, project straight out from
the V atom and fold back from the base plane towards
the side of the vanadyl group; in complex 2, the three
O–C bonds project out similarly from the V atom and
the tendency of these bonds to fold back towards the
vanadyl side is again present but the large, 165° angles
at O(2) and O(2%) mean that these two bonds lie close to
the O3 base plane. The Ovanadyl–V–O–C torsion angles
in complex 2 are 0(1), −15(1) and +15(1)°, mean (of
absolute values) 9.9°; in [{VO(OCH2CH2Cl)3}2], the
corresponding mean (absolute) torsion angle is 8.2° and
in [{VO(OC5H9)3}2] 3.5°, showing that the orientations
about the oxygen atoms of the alkoxide and phenoxide
˚
distances, all 1.761(4) A, and the VꢀO distance, 1.564(4)
˚
A, are close to those observed in the dimeric analogues,
˚
viz. in [{VO(OCH2CH2Cl)3}2] 1.80(3)av and 1.584(2) A
˚
[4] and in [{VO(OC5H9)3}2] 1.79(2)av and 1.595(2) A [5].
In both dimers, there are two short V–Oalkoxide dis-
˚
˚
tances, ca. 1.76 A, and one longer one, ca. 1.85 A (to
Table 2
Selected molecular dimensions in [VO(OC6H3Pri2-2,6)3] 2. Bond
˚
lengths are in A and angles in degrees with estimated standard
deviations in parentheses
(a) About the vanadium atom
V–O(1)
1.761(4)
1.761(2)
V–O(2%)a
V–O(3)
1.761(2)
1.564(4)
V–O(2)
O(1)–V–O(2)
O(1)–V–O(2%)
O(1)–V–O(3)
108.1(2)
107.2(2)
107.3(2)
O(2)–V–O(2%)
O(2)–V–O(3)
O(2%)–V–O(3)
108.1(1)
112.8(2)
113.0(2)
(b) In the ligands
O(1)–C(11)
O(2)–C(21)
1.383(6)
1.360(3)
V–O(1)–C(11)
V–O(2)–C(21)
V(2%)–C(21%)
136.3(4)
165.9(2)
164.3(2)
Fig. 2. View of a single molecule of [VO(OC6H3Pri2-2,6)3] 2 showing
the disorder in the isopropyl ligands and the atom numbering scheme;
a 2-fold symmetry axis relates the ligand of O(2) with that of O(2%),
and passes through the ligand of O(1).
a The primed atoms are those in the other half of the molecule at
0.25−x, y, 0.25−z.

198
R.A. Henderson et al. / Journal of Organometallic Chemistry 554 (1998) 195–201
lar weight distribution and excellent morphology [8].
Although the 2–AlEt2Cl catalyst system polymerised
ethylene with lower activity than 1–AlEt2Cl, prelimi-
nary experiments show that compound 2 in presence of
methylaluminoxane (MAO) exhibits significantly higher
activity (45 kg polyethylene g−1
h
−1) than vanadi-
um(V) imido-ligand stabilised procatalysts (B1 kg
polyethylene g−1 −1), for which VV is suggested to be
h
the oxidation state of the catalyst [9]. We suppose that
in the 2–AlEt2Cl and 2–MAO catalysts the active
vanadium species has a lower oxidation state than VV
and we plan further studies to clarify this point. Gener-
ally, vanadium-based catalysts show configurations d2
and d3; d0 vanadium complexes have frequently been
used as catalysts but are readily reduced to lower
oxidation states on addition of cocatalysts [10]. It has
not been excluded, however, that vanadium(V) com-
plexes would also be effective catalysts if they could be
stabilised in their high oxidation states [11].
It is worthy of note that the catalysts of this study
show also high activity in ethylene copolymerisation
with hex-1-ene to produce, under the same conditions,
170 and 102 kg polymer g−1 h−1 for 1 and 2,
respectively.
Fig. 3. Showing the principal disorder of the molecules of complex 2
about a crystallographic 2-fold symmetry axis. There is no disorder in
the phenyl groups of O(2) {and O(2%)}. Resolution of C(12) and
C(13), from C(16%) and C(15%), respectively, was not possible.
ligands are similar in the three complexes. In the nio-
bium monomer with the bulky phenoxide ligands [6],
the Nb–O–C angles are also large, in the range 146.1–
155.0°, and the orientations about the Ophenox atoms are
again similar to those above with a mean (absolute)
torsion angle of 16.2°. However, in this complex, the
phenoxide ligands are tilted, propellor-wise, about a
pseudo 3-fold axis, quite differently from the arrange-
ment in complex 2.
In its NMR spectra (Section 2) compound 2 shows
the expected 1:2 relative intensity H patterns for the
1
2. Experimental
methyne protons (overlapping sextets) and the methyl
protons (overlapping doublets) and a 51V shift, (−
531.6 ppm rel. VOCl3), in the region expected for
VO(OR)3 compounds [4].
All operations were carried out under a dry dinitro-
gen or argon atmosphere using standard Schlenk and
glove box techniques. All solvents were distilled under
dinitrogen from the appropriate drying agents prior to
use. IR spectra were recorded on a Perkin-Elmer 882
instrument in Nujol mulls and NMR spectra on a
JEOL GSX-270 instrument. Analyses were performed
at the University of Wroclaw. Magnetic moments were
determined in solution using the NMR method of
Evans and James [12].
1.3. Polymerisation studies
The ethylene polymerisation catalysts were prepared
in n-hexane by milling a slurry of [MgCl2(thf)2] with the
vanadium compound and diethylaluminium chloride as
the cocatalysts. The results of ethylene polymerisation
are shown in Table 3. The catalytic activity of both
systems increases with temperature and is highest at 343
K to yield 243 and 160 kg polyethylene g−1 h−1 for 1
and 2, respectively. Addition of [MgCl2(thf)2] to the
vanadium catalyst precursors did not significantly raise
the activity and therefore it was a non-essential compo-
nent. However, when the vanadium catalysts were used
without [MgCl2(thf)2], polyethylene usually fouled the
reactor so badly that the results were difficult to repro-
duce. The activity of 1 is comparable to vanadium(III)
chloride-based catalysts [VCl3(thf)3] and [V2(m-
Cl)2Cl4(thf)4] which produced \200 kg polyethylene
g−1 h−1 at 323 K) [7] and is significantly lower than
that for [V2Mg2(v3,p2-thffo)2(v,p2-thffo)4Cl4]·2CH2Cl2
(thffo, 2-tetrahydrofurfuroxide) (410 kg polyethylene
Table 3
Polymerisationa of ethylene with V–MgCl2–AlEt2Cl catalyst
T (K)
Productivityb/kg PE g−1 V h−1
[VCl(OC6H3Pri2-2,6)2(thf)2] 1
303
323
343
71
185 (33)c
243 (98)c
[VO(OC6H3Pri2-2,6)3] 2
303
323
343
87
123 (65)c
160 (55)c
a Polymerisation conditions: [V]0=0.01 mmol dm−3, [Al]=5 mmol
dm−3, Mg: V=10, Pethylene=0.6 MPa, in hexene.
b Productivity is the mass in kg of polymer formed per g of vanadium
atom in 1 h.
g−1
h
−1) [2]. Vanadium alkoxide systems have the
c Precatalyst activated MAO (10% solution in toluene), [Al]=10
advantage of producing polymer with narrow molecu-
mmol dm−3, [V]=0.01 mmol dm−3
.

R.A. Henderson et al. / Journal of Organometallic Chemistry 554 (1998) 195–201
199
The compounds HOC6H3Pr2i -2,6, trimethylsilyl chlo-
2.3.2. Method 2
ride, hydrazine and vanadium trichloride were obtained
from Aldrich. The compounds [V(OC6H3Pri2-
2,6)4Li(thf)] [13], [VCl3(thf)3] [14], N2(SiMe3)4 [15] and
HSC6H2Me3-2,4,6 [16] were prepared by literature
methods.
To a green solution of [V(OC6H3Pri2-2,6)4Li(thf)] in
hexane (30 cm3) [generated from [VCl3(thf)3] (0.5 g, 1.34
mmol) and Li[OC6H3Pri2-2,6] (0.99 g, 5.36 mmol)] was
added propylene oxide (0.1 cm3, 1.43 mmol). After the
solution was stirred for 2 days, it had become dark red
and was filtered, concentrated in a vacuum to 10 cm3
and left at ca. −4°C to form red crystals which were
filtered off, washed with hexane and dried in a vacuum
(80%). (Found: C, 71.9; H, 8.6. C36H51O4V requires: C,
72.2; H, 8.6%). IR: w(VO) 1000 cm−1. NMR (C62H6):
1H l 7.16–6.94 (Ph multiplet), 3.96–3.78 [CH(CH3)2,
two overlapping sextets, 1:2 ratio], 1.29–1.24
[CH(CH3)2, two overlapping doublets, 1:2 ratio]; 51V l
−531.6 (s).
2.1. Preparation of [VCl(OC6H3Pri2-2,6)2(thf)2] 1
A solution of [VCl3(thf)3], (0.7 g, 1.87 mmol) and
Li(OC6H3Pri2-2,6) (0.69 g, 3.75 mmol) in thf (50 cm3)
was stirred for 12 h to give a dark red solution which
was taken to dryness in a vacuum and the residue
extracted with hexane (3×30 cm3). The resulting solu-
tion was concentrated in a vacuum to 15 cm3 and kept
at low temperature (ca. −4°C) when dark red crystals
deposited. These were filtered off, washed with hexane,
dried in a vacuum and used for X-ray studies. Yield
0.77g (70%). veff=2.6 vB. (Found: C, 65.7; H, 8.7; Cl,
5.9. C32H50O4ClV requires: C, 65.7; H, 8.6; Cl, 6.0%)
2.3.3. Method 3
To a solution of [VO(OPri)3] (0.51 g, 2.1 mmol) in
n-hexane (30 cm3) was added HOC6H3Pri2-2,6 (1.12 g,
6.3 mmol). The mixture turned immediately red/brown
and after stirring for 3 h and evaporation of solvent to
half volume, reddish, crystalline compound 2 precipi-
tated. Yield 1.06 g (85%).
2.2. Preparation of [V(SC6H2Me3-2,4,6)(OC6H3Pri2-
2,6)2(thf)2]
To a solution of 1 (1.2 g, 2.05 mmol) in thf (50 cm3)
was added Li[SC6H2Me3-2,4,6] (0.5 g, 2.05 mmol) and
the mixture stirred for 5 h. The reaction mixture turned
from brown/red to olive. The solvent was evaporated to
dryness, the residue extracted with hexane (3×30 cm3)
and the resulting solution concentrated to 10 cm3 and
left at low temperature (ca. −4°C) when an olive-
brown solid deposited which was filtered off, washed
with hexane and dried in a vacuum. Yield 50%. Al-
though the compound is too thermally unstable (chang-
ing to a red oil) to permit crystallisation, it was
sufficiently stable for characterisation by microanalysis.
veff=2.6 vB. (Found: C, 69.7; H, 8.3; S, 4.9.
C41H61O4SV requires: C, 70.3; H, 8.7; S, 4.6%).
2.4. Polymerisation tests
A slurry of [MgCl2(thf)2] (30 mmol) in n-hexane was
milled under argon in a glass mill (capacity 250 cm3,
with 20 balls of diameter 5–15 mm) at room tempera-
ture for 6 h. Then the vanadium compound (3 mmol)
and n-hexane (50 cm3) were added and the mixture was
milled for 24 h. The sample of precatalyst suspension
(containing 0.01% vanadium) was activated with
AlEt2Cl (20 mmol) or MAO (10% solution in toluene)
for 15 min at 323 K under argon to form the highly
active catalyst. The polymerisation of ethylene was
carried out at 303–343 K in a stainless-steel reactor (1
dm3) equipped with a stirrer, in n-hexane at an ethylene
pressure of 0.6 MPa. The polymerisation was quenched
with a 5% solution of HCl in methanol (150 cm3) and
the polymer was filtered off, washed with methanol and
dried under vacuum. Further details of polymerisation
conditions, yields, etc. are shown in Table 3.
2.3. Preparation of [VO(OC6H3Pri2-2,6)3] 2
2.3.1. Method 1
Compound 1 (0.5 g, 0.85 mmol) was dissolved in
hexane and N2(SiMe3)4 (0.27 g, 0.85 mmol) was added.
The reaction mixture was stirred for 24 h then the
solvent was removed in a vacuum, to leave an oily
residue which was heated at 60°C in a vacuum to
remove volatile silylated byproducts. The cooled
residue was then extracted with hexane and allowed to
stand at ca. −4°C for several days during which time
dark red crystals deposited which were filtered off and
used for X-ray structure determination. In a separate
experiment, [VCl(OC6H3Pri2-2,6)2(thf)2] 1 (1.06 g, 1.8
mmol) in hexane (60 cm3) was treated with N2(SiMe3)4
(0.58 g, 1.8 mmol) and N2 (22 cm3, 0.98 mmol, gas
chromatography) was evolved.
3. Crystallography
3.1. [VCl(OC6H3Pri2-2,6)2(thf)2] 1
Crystal data: C32H50ClO4V, M=585.11, monoclinic,
space group C2/c, a=21.187(4), b=9.853(2), c=
3
˚
˚
15.965(3) A, i=93.90(2)°, U=3325(2) A , Z=4,
Dc=1.169(1) g cm−3, Dm=1.179 g cm−3, F(000)=
1256, T=180.0(2) K, v(Mo–Kh)=0.41 mm−1
,
˚
u(Mo–Kh)=0.71069 A.

200
R.A. Henderson et al. / Journal of Organometallic Chemistry 554 (1998) 195–201
Crystals are red/brown needles and were cut down
groups were set to ride on the parent C atoms, while
the methyl hydrogen atoms and the major-occupancy
disordered CH atoms were refined with geometrical
constraints. Hydrogen atoms of the minor-occupancy
CH and methyl groups were not included, and further
resolution of the disordered groups was not possible.
The non-H atoms (except the disordered methyl C
atoms in the minor orientations) were refined with
anisotropic thermal parameters. The isotropic tempera-
ture factors of the minor occupation C atoms and the
hydrogen atoms were allowed to refine freely. At the
conclusion of the refinement, R=0.078 and Rg=0.075
for use. One, ca. 0.5×0.4×0.3 mm was sealed in a
glass capillary. On a Kuma KM4 four-circle diffrac-
tometer, cell parameters were calculated from the set-
tings of 25 reflections having q=7–12.5°, and
diffraction intensities were measured to qmax=23°.
There was no crystal decay and 3294 reflections were
measured of which 1510 were unique.
The structure was solved by the Patterson method
and refined by full-matrix least-squares calculations
using SHELXL-93 [17]. Neutral atom scattering factors
were from ref. [18] and real and imaginary components
of anomolous dispersion were included for all non-H
atoms. Absorption corrections following the DIFABS
[19] procedure were applied: minimum and maximum
absorption corrections were 0.861 and 1.129, respec-
tively. The hydrogen atoms of CH2 and CH3 groups
20
for 2444 reflections with I\|(I) and weighted w=
(|2F+0.00045F2)−1. In the final difference map, the
highest peak (at ca. 0.3 e A−3) was in one of the
˚
disordered isopropyl groups. Refinement in the corre-
sponding non-centrosymmetric space group, Fd (equiv-
alent to Cc, no. 9) did not improve the model nor the
results.
Scattering factors for neutral atoms were taken from
ref. [18]. Computer programs used in this analysis have
been noted above or in Table 4 of ref. [22], and were
run on a DEC-AlphaStation 200 4/100 in the Nitrogen
Fixation Laboratory, John Innes Centre.
˚
were put in calculated positions with d(C–H)=1.08 A
and introduced as fixed contributors in the final stage
of refinement. At the conclusion of refinement, the R1
and wR2 [17] values were 0.051 and 0.128, respectively
for the 1394 reflections with I\3|(I) using a weighting
scheme of the form w=1/{|2(Fo2) +(0.0826P)2+
2.79P} with P=(Fo2 +2F2c)/3. For the last cycle of
refinement the maximum value of the ratio D/| was
below 0.014 and the final difference map showed a
−3
˚
general background within −0.19 and 0.43 e A
.
Acknowledgements
3.2. [VO(OC6H3Pri2-2,6)3] 2
We thank the Polish State Committee for Scientific
Research, the BBSRC, the British Council and the EC
for support, and Dr Krzysztof Szczegot of the Univer-
stiy of Opole for preliminary polymerisation tests.
Crystal data: C36H51O4V, M=598.7. Monoclinic,
space group F2/d (equivalent to no. 15), a=31.787(5),
˚
b=10.092(2), c=22.599(5) A, i=107.83(3)°, U=
3
6901(2) A , Z=8, Dc=1.152 g cm−3, F(000)=2576,
˚
T=180(2) K, v(Mo–Kh)=3.1 cm−1, u(Mo–Kh)=
˚
References
0.71069 A.
Crystals are dark red and one, ca. 0.5×0.4×0.4
mm, was mounted in a glass capillary on the Kuma
diffractometer as above. A total of 3984 unique reflec-
tions to qmax=27.5°, 2106 of which were ‘observed’
with I\2|(I), were entered into the SHELX program
system [20].
During processing, corrections were applied for
Lorentz-polarisation effects; there was no crystal deteri-
oration. The structure was determined by direct meth-
ods in the SHELXS program [21] and refined by
full-matrix least-squares methods [20]. There is disorder
in the molecule: a 2-fold symmetry axis passes through
the molecule—it passes roughly through one phenoxide
ligand, the second and third phenoxide ligands are
related by the symmetry, and the vanadyl group is
randomly disordered in sites related by the axis. Some
of the isopropyl groups are also disordered, with the
methyl groups in the principal orientation in 70–80%
of the sites. Hydrogen atoms were included in idealised
positions; those on phenyl rings and in ordered CH
[1] C. Le Floc’h, R.A. Henderson, P.B. Hitchcock, et al., J. Chem.
Soc. Dalton Trans. (1996) 2755 and references therein.
[2] Z. Janas, P. Sobota, M. Kasprzak, T. Gtowiak, J. Chem. Soc.
Chem. Commun. (1996) 2727 and references therein.
[3] S. Gambarotta, F. van Bolus, M.Y. Chang, Inorg. Chem. 26
(1987) 4303.
[4] W. Priebsch, D. Rehder, Inorg. Chem. 29 (1990) 3013.
[5] F. Hillerns, F. Olbrich, U. Behrens, D. Rehder, Angew. Chem.
Int. Ed. Engl. 31 (1992) 447.
[6] J.S. Yu, P.E. Fanwick, I.P. Rothwell, Acta Crystallogr. C48
(1992) 1759.
[7] P. Sobota, J. Ejfler, S. Szafert, T. Gtowiak, I.O. Fritzky, K.
Szczegot, J. Chem. Soc. Dalton Trans. (1995) 1727.
[8] F.J. Karol, S.C. Kao, New J. Chem. 18 (1994) 97.
[9] S. Scheuer, J. Fischer, J. Kress, Organometallics 14 (1995) 2627.
[10] S. Jiao, D. Yu, Polym. J. 17 (1985) 899 and references therein.
[11] F.J. Feher, R.L. Blanski, Organometallics 12 (1993) 958.
[12] D.F. Evans, T.A. James, J. Chem. Soc. Dalton Trans. (1979)
723.
[13] W.C.A. Wilisch, M.J. Scott, W.H. Armstrong, Inorg. Chem. 27
(1988) 4333.
[14] L.E. Manzer, Inorg. Synth. 21 (1982) 138.
[15] K. Seppelt, W. Sundermeyer, Chem. Ber. 102 (1969) 1247.

R.A. Henderson et al. / Journal of Organometallic Chemistry 554 (1998) 195–201
201
[16] P.J. Blower, J.R. Dilworth, J.P. Hutchinson, J.A. Zubieta, J.
Chem. Soc. Dalton Trans. (1985) 1533.
[17] G.M. Sheldrick, SHELXL-93, Program for crystal structure
refinement, University of Go¨ttingen, Germany, 1993.
Birmingham, 1992, Tables 4.2.6.8 and 6.1.1.4.
[19] N. Walker, D. Stuart, Acta Crystallogr. Sect. A 39 (1983) 158.
[20] G.M. Sheldrick, SHELX, Program for crystal structure determi-
nation, University of Cambridge, 1976.
[18] (a) International Tables for X-ray Crystallography, vol. 4,
Kynoch Press, Birmingham, 1974, pp. 99 and 149. (b) Interna-
tional Tables for X-ray Crystallography, vol. C, Kynoch Press,
[21] G.M. Sheldrick, Acta Crystallogr. Sect. A A46 (1990) 467.
[22] S.N. Anderson, R.L. Richards, D.L. Hughes, J. Chem. Soc.
Dalton Trans. (1986) 245.
.