Vanadium(IV) and -(V) complexes with O,N-chelating aminophenolate and pyridylalkoxide ligands

Article abstract of DOI:10.1021/ic991415s

Two different monoanionic O,N-chelating ligand systems, i.e., [OC₆H₂(CH₂NMe₂)-2-Me₂-4,6]^- (1) and [OCMe₂-([2]-Py)]^- (2), have been applied in the synthesis of vanadium(V) complexes. The tertiary amine functionality in 1 caused reduction of the vanadium nucleus to the 4+ oxidation state with either [VOCl₃], [V(=NR)Cl₃], or [V(=NR)(NEt₂)₃] (R = Ph, (3a, 5a), R = p-Tol (3b, 5b)), and applying 1 as a reducing agent resulted in the synthesis of the vanadium(IV) complexes [VO(OC₆H₂(CH₂NMe₂)-2-Me₂-4,6)₂] (4) and [V(=NPh)(OC₆H₂(CH₂NMe₂)-2-Me₂-4,6)₂] (6). In the case of [V(=N-p-Tol)(NEt₂)(OC₆H₂(CH₂NMe₂)-2-Me₂-4,6)₂] (7b), the reduction was sufficiently slow to allow its characterization by ¹H NMR and variable-temperature studies showed it to be a five-coordinate species in solution. Although the reaction of 1 with [V(=N-p-Tol)(O-i-Pr)₃] (9b) did not result in reduction of the vanadium nucleus, vanadium(V) compounds could not be isolated. Mixtures of the vanadium-(V) (mono)phenolate, [V(=N-p-Tol)(O-i-Pr)₂(OC₆H₂(CH₂NMe₂)-2-Me₂-4,6)] (10), and the vanadium(V) (bis)phenolate, [V(=N-p-Tol)(O-i-Pr)(OC₆H₂(CH₂NMe₂)-2-Me₂-4,6)₂] (11), were obtained. With the pyridylalkoxide 2, no reduction was observed and the vanadium(V) compounds [VOCl₂(OCMe₂([2]-Py))] (12) and [V(=N-p-Tol)Cl₂(OCMe₂([2]-Py)] (13) were obtained. ⁵¹V NMR showed 7b and 12 to be five-coordinate in solution, whereas for 10, 11, and 13 a coordination number of 6 was found. Compounds 12 and 13 showed decreased activity compared to their nonchelated vanadium(V) analogues when applied as catalysts in ethene polymerization. Two polymorphic forms with a difference in the V-N-C angle of 12.5°have been found for 6. Crystal data: 6·Et₂O, triclinic, P1, a = 11.1557(6) A, b = 12.5744(12) A, c = 13.1051(14) A, α = 64.244(8)°, β = 70.472(7)°, γ = 87.950(6)°, V = 1547(3) A³, Z = 2; 6·C₆H₆, triclinic, P1, a = 8.6034(3) A, b = 13.3614(4) A, c = 15.1044(5) A, α = 98.182(3)°, β = 105.618(2)°, γ = 107.130(2)°, V = 1551.00(10) A³, Z = 2; 12, orthorhombic, Pbca, a = 11.8576(12) A, b = 12.6710(13) A, c = 14.722(2) A, V= 2211.9(4) A³, Z = 8.

Full text of DOI:10.1021/ic991415s

3970
Inorg. Chem. 2000, 39, 3970-3977
Vanadium(IV) and -(V) Complexes with O,N-Chelating Aminophenolate and
Pyridylalkoxide Ligands
Henk Hagen,† Chris Bezemer,† Jaap Boersma,† Huub Kooijman,‡ Martin Lutz,‡
Anthony L. Spek,‡,§ and Gerard van Koten*,†
Debye Institute, Department of Metal-Mediated Synthesis, Utrecht University, Padualaan 8,
3584 CH Utrecht, The Netherlands, and Bijvoet Center for Biomolecular Research, Crystal and
Structural Chemistry, Utrecht University, Padualaan 8, 3584 CH Utrecht, The Netherlands
ReceiVed December 8, 1999
Two different monoanionic O,N-chelating ligand systems, i.e., [OC6H2(CH2NMe2)-2-Me2-4,6]- (1) and [OCMe2-
([2]-Py)]- (2), have been applied in the synthesis of vanadium(V) complexes. The tertiary amine functionality in
1 caused reduction of the vanadium nucleus to the 4+ oxidation state with either [VOCl3], [V(dNR)Cl3], or
[V(dNR)(NEt2)3] (R ) Ph, (3a, 5a), R ) p-Tol (3b, 5b)), and applying 1 as a reducing agent resulted in the
synthesis of the vanadium(IV) complexes [VO(OC6H2(CH2NMe2)-2-Me2-4,6)2] (4) and [V(dNPh)(OC6H2(CH2-
NMe2)-2-Me2-4,6)2] (6). In the case of [V(dN-p-Tol)(NEt2)(OC6H2(CH2NMe2)-2-Me2-4,6)2] (7b), the reduction
1
was sufficiently slow to allow its characterization by H NMR and variable-temperature studies showed it to be
a five-coordinate species in solution. Although the reaction of 1 with [V(dN-p-Tol)(O-i-Pr)3] (9b) did not result
in reduction of the vanadium nucleus, vanadium(V) compounds could not be isolated. Mixtures of the vanadium-
(V) (mono)phenolate, [V(dN-p-Tol)(O-i-Pr)2(OC6H2(CH2NMe2)-2-Me2-4,6)] (10), and the vanadium(V) (bis)-
phenolate, [V(dN-p-Tol)(O-i-Pr)(OC6H2(CH2NMe2)-2-Me2-4,6)2] (11), were obtained. With the pyridylalkoxide
2, no reduction was observed and the vanadium(V) compounds [VOCl2(OCMe2([2]-Py))] (12) and [V(dN-p-
Tol)Cl2(OCMe2([2]-Py)] (13) were obtained. 51V NMR showed 7b and 12 to be five-coordinate in solution, whereas
for 10, 11, and 13 a coordination number of 6 was found. Compounds 12 and 13 showed decreased activity
compared to their nonchelated vanadium(V) analogues when applied as catalysts in ethene polymerization. Two
polymorphic forms with a difference in the V-N-C angle of 12.5° have been found for 6. Crystal data: 6‚Et2O,
triclinic, P1h, a ) 11.1557(6) Å, b ) 12.5744(12) Å, c ) 13.1051(14) Å, R ) 64.244(8)°, â ) 70.472(7)°, γ )
87.950(6)°, V ) 1547(3) Å3, Z ) 2; 6‚C6H6, triclinic, P1h, a ) 8.6034(3) Å, b ) 13.3614(4) Å, c ) 15.1044(5)
Å, R ) 98.182(3)°, â ) 105.618(2)°, γ ) 107.130(2)°, V ) 1551.00(10) Å3, Z ) 2; 12, orthorhombic, Pbca, a
) 11.8576(12) Å, b ) 12.6710(13) Å, c ) 14.722(2) Å, V ) 2211.9(4) Å3, Z ) 8.
Introduction
Recently we set out to study the use of monoanionic
potentially bidentate phenolate or alkoxide ligands, i.e.,
[OC6H2(CH2NMe2)-2-Me2-4,6]- (O-MDMBA) (1) and R,R-
dimethyl-2-pyridyl methoxide [OCMe2([2]-Py)]- (2) (see Figure
Figure 1. Selected potentially O,N-chelating monoanionic ligands.
Abbreviations used: 1, [O-MDMBA]-; 2, [OCMe2([2]-Py)]-.
1), in vanadium(V) chemistry. These ligands have been chosen
since vanadium catalyst systems with oxygen-containing ligands
are known to give polymers with narrow molecular weight
distributions,1 while the nitrogen donor atom in the ligands can
help in stabilizing the high oxidation state of the vanadium
center by V-N coordination. In addition to the latter effect, an
o-amino functionality itself represents a substituent with con-
siderable bulk, which can prevent bimolecular deactivation
pathways of the vanadium(V) catalyst.2-4 Finally, the electronic
properties of the vanadium center can be influenced considerably
by varying the substituents on the ring, as we have recently
shown for vanadium(IV) oxo (bis)phenolate complexes.5
Another way of stabilizing a high oxidation state vanadium
center is the use of imido ligands. These are more electron-
donating than an oxo group due to their reduced electronega-
tivity.6 Moreover, the vanadium center can be influenced both
electronically and sterically, when an organic group is present
on the imido nitrogen. For these studies we have used both the
* To whom correspondence should be addressed. E-mail: g.vankoten@
chem.uu.nl.
† Debye Institute, Department of Metal-Mediated Synthesis.
‡ Bijvoet Center for Biomolecular Research, Crystal and Structural
Chemistry.
(4) Buijink, J.-K. F.; Meetsma, A.; Teuben, J. H.; Kooijman, H.; Spek,
A. L. J. Organomet. Chem. 1995, 497, 161.
(5) Hagen, H.; Barbon, A.; van Faassen, E. E.; Lutz, B. T. G.; Boersma,
J.; Spek, A. L.; van Koten, G. Inorg. Chem. 1999, 38, 4079.
(6) Nugent, W. A.; Mayer, J. M. Metal-ligand Multiple Bonds; John Wiley
& Sons: New York, 1988.
§ To whom correspondence pertaining to crystallographic studies should
be addressed. E-mail: a.l.spek@chem.uu.nl.
(1) Karol, F. J.; Kao, S.-C. New J. Chem. 1994, 18, 97.
(2) Murphy, V. J.; Turner, H. Organometallics 1997, 16, 2495 and
references therein.
(3) Preuss, F.; Becker, H.; Kaub, J.; Sheldrick, W. S. Z. Naturforsch. 1988,
43b, 1195.
(7) Devore, D. D.; Lichtenhan, J. D.; Takusagawa, F.; Maatta, E. A. J.
Am. Chem. Soc. 1987, 109, 7408.
10.1021/ic991415s CCC: $19.00 © 2000 American Chemical Society
Published on Web 08/11/2000

Vanadium(IV) and -(V) Complexes
Inorganic Chemistry, Vol. 39, No. 18, 2000 3971
51V NMR (78.9 MHz, C6D6): δ -175 (partially resolved triplet,
phenyl- and the p-tolylimido ligands. The latter is slightly more
electron-donating7 and has the advantage that the p-methyl group
and the aromatic protons give very distinctive signals in the 1H
NMR.
In order to synthesize vanadium(V) complexes with the O,N-
chelating ligands, we have used three types of starting material,
viz., vanadium(V) chlorides, diethylamides, and isopropoxides.
The O,N-chelating vanadium(V) products as well as several
vanadium(V) starting materials were tested in ethene polym-
erization with Et2AlCl as cocatalyst.
1
51 14
J
) 99 Hz, ∆ν1/2 ) 275 Hz).
V,
N
Synthesis of [V(dNPh)(OC6H2(CH2NMe2)-2-Me2-4,6)2] (6). To a
stirred solution of [V(dNPh)(NEt2)3] (5a) (1.54 g, 4.3 mmol) in Et2O
(50 mL) was added a solution of HOC6H2(CH2NMe2)-2-Me2-4,6 (1)
(0.78 g, 4.4 mmol) in Et2O (10 mL) at 0 °C. The resulting mixture
was stirred for 2 days, after which it was filtered. Cooling the filtrate
to -20 °C gave dark-red crystals (0.86 g, 35%). Mp: 178 °C.
Anal. Calcd for C28H37N3O2V‚Et2O: C, 67.11; H, 8.27; N, 7.34.
Found: C, 66.96; H, 8.25; N, 7.37.
EPR (hexane, 7.3 mM): Aiso ) 85.28 × 10-4 cm-1, giso ) 1.9717.
Synthesis of [V(dNPh)(NEt2)(OC6H2(CH2NMe2)-2-Me2-4,6)2] (7a).
HOC6H2(CH2NMe2)-2-Me2-4,6 (1) (0.90 g, 5.0 mmol) was added slowly
to [V(dNPh)(NEt2)3] (5a) (0.90 g, 2.5 mmol) in hexane (50 mL) at 0
°C. After stirring for 2 h at room temperature the solvent was removed
in vacuo, leaving a brown-red solid in quantitative yield.
Experimental Section
General. All reactions were performed in an atmosphere of dry,
oxygen-free dinitrogen using standard Schlenk techniques. Solvents
were carefully dried and distilled prior to use. HOC6H2(CH2NMe2)-2-
Me2-4,6 (1),8 [V(dNPh)Cl3] (3a) and [V(dN-p-tol)Cl3] (3b),7
[V(dN-p-Tol)(O-i-Pr)3] (9b),9 [VO(O-i-Pr)3] (8),10 and LiOCMe2([2]-
Py)11 were prepared according to literature procedures. Et3N was
distilled from CaH2. All other chemicals were obtained from commercial
sources and used as received. Elemental analyses were performed by
H. Kolbe, Mikroanalytisches Laboratorium, Mu¨lheim, Germany. 1H and
13C NMR spectra were recorded on a Bruker AC200 or 300 spectrom-
eter or a Varian Inova 300 spectrometer. 51V NMR spectra were
recorded on a Bruker AC300 spectrometer at 78.9 MHz relative to an
external standard ([VOCl3]/CDCl3, 75/25 v/v).12 EPR spectra were
recorded on a Bruker ESP 300 spectrometer.
Synthesis of [VO(OC6H2(CH2NMe2)-2-Me2-4,6)2] (4). A mixture
of HOC6H2(CH2NMe2)-2-Me2-4,6 (1) (1.90 g, 10.6 mmol) and Et3N
(2 mL, excess) in C6H6 (50 mL) was added to a stirred solution of
[VOCl3] (0.74 g, 4.2 mmol) in C6H6 (10 mL) in 15 min. After stirring
for 16 h the solvent was removed in vacuo. The residue was washed
with nitrogen-flushed MeOH (2 × 150 mL) to give [VO(OC6H2(CH2-
NMe2)-2-Me2-4,6)2] as a purple solid (1.41 g; 80%). Crystallization
from MeNO2 at room temperature gave blue crystals. Mp: 205 °C.
Anal. Calcd for C22H32N2O3V: C, 62.40; H, 7.62; N, 6.62. Found:
C, 62.26; H, 7.58; N, 6.56.
1H NMR (200 MHz, C6D6): δ 0.88 (bs, 6 H, N(CH2CH3)2), 2.29 (s,
6 H, p-Me), 2.36 (s, 12 H, NMe2), 2.50 (s, 6 H, o-Me), 3.41 and 4.07
2
(AB, 4 H, JH,H ) 13.5 Hz, CH2NMe2), 4.48 (bs, 4 H, N(CH2CH3)2),
6.68-7.24 (m, 9 H, Ar-H).
Synthesis of [V(dN-p-Tol)(NEt2)(OC6H2(CH2NMe2)-2-Me2-4,6)2]
(7b). HOC6H2(CH2NMe2)-2-Me2-4,6 (1) (1.70 g, 9.5 mmol) in pentane
(20 mL) was added to [V(dN-p-Tol)(NEt2)3] (5b) (1.77 g, 4.7 mmol)
in pentane (40 mL) over a period of 30 min. After stirring for 2 h at
room temperature residual solid material was removed by centrifugation/
decantation. Subsequent removal of the solvent in vacuo gave an
orange-red solid in quantitative yield.
1H NMR (300 MHz, C6D6): δ 0.80 (bs, 6 H, N(CH2CH3)2), 2.01 (s,
3 H, NAr-Me), 2.29 (s, 6 H, p-Me), 2.38 (s, 12 H, NMe2), 2.50 (s, 6
H, o-Me), 3.44 and 4.11 (AB, 4 H, 2JH,H ) 13.5 Hz, CH2NMe2), 4.50
(bs, 4 H, N(CH2CH3)2), 6.74 and 7.17 (AB, 4 H, 3JH,H ) 8.2 Hz, N-Ar),
6.97 (s, 2 H, m-H), 6.99 (s, 2 H, m-H).
51V NMR (78.9 MHz, C6D6): δ -235 (∆ν1/2 ) 137 Hz).
Synthesis of [V(dNPh)(O-i-Pr)3] (9a). Phenyl isocyanate (2.11 mL,
19.4 mmol) was added to a stirred solution of [VO(O-i-Pr)3] (8) (4.73
g, 19.4 mmol) in octane (30 mL). After refluxing for 3 h the solvent
was removed in vacuo. Recrystallization from hexane at -30 °C yielded
orange-brown crystals (4.09 g, 66%). Mp: 53 °C.
Synthesis of [V(dNPh)(NEt2)3] (5a). A solution of LiNEt2 (1.43
g, 18.1 mmol) in Et2O (75 mL) was added dropwise to a stirred solution
of [V(dNPh)Cl3] (3a) (1.46 g, 5.9 mmol) in Et2O (75 mL) at 0 °C.
After stirring for 16 h at room temperature the white precipitate was
removed from the solution by centrifugation/decantation. The super-
natant solution was evaporated to dryness in vacuo, and the residue
was extracted with hexane (150 mL). Removal of the hexane in vacuo
afforded [V(dNPh)(NEt2)3] as a red oil (2.11 g, 100%).
Anal. Calcd for C15H26NO3V: C, 56.42; H, 8.21; N, 4.39. Found:
C, 56.58; H, 8.11; N, 4.36.
3
1H NMR (200 MHz, CDCl3): δ 1.38 (d, 18 H, JH,H ) 6.1 Hz,
OCHMe2), 5.15 (sept, 3 H, 3JH,H ) 6.1 Hz, OCHMe2), 7.05-7.29 (bm,
5 H, NAr).
13C NMR (50 MHz, CDCl3): δ 26.9 (OCHMe2), 80.2 (broad,
OCHMe2), 125.2 (NAr-C2 and NAr-C6), 125.5 (NAr-C3 and NAr-
C5) 128.5 (NAr-C4), NAr-C1 not observed.
1
51V NMR (78.9 MHz, CDCl3) δ -622 (t1:1:1, J
) 113 Hz).
3
1H NMR (300 MHz, C6D6): δ 1.16 (t, 18 H, JH,H ) 7.0 Hz,
51 14
V,
N
N(CH2CH3)2), 3.77 (q, 12 H, 3JH,H ) 7.0 Hz, N(CH2CH3)2) 6.80-7.30
(m, 5 H, NPh-H).
Attempted Synthesis of [V(dN-p-Tol)(O-i-Pr)2(OC6H2(CH2NMe2)-
2-Me2-4,6)] (10). To a solution of [V(dN-p-Tol)(O-i-Pr)3] (9b) (0.55
g, 1.7 mmol) in pentane (25 mL) was added HOC6H2(CH2NMe2)-2-
Me2-4,6 (1) (0.30 g, 1.7 mmol) in pentane (10 mL). After stirring for
24 h the solvent was removed in vacuo, leaving a dark-red oil. 1H NMR
analysis showed the oil to be a mixture of [V(dN-p-Tol)(O-i-Pr)3] (9b),
the monophenolate [V(dN-p-Tol)(O-i-Pr)2(OC6H2(CH2NMe2)-2-Me2-
4,6)] (10), and the (bis)phenolate [V(dN-p-Tol)(O-i-Pr)(OC6H2(CH2-
NMe2)-2-Me2-4,6)2] (11) in a 1:2:1 ratio.
51V NMR (78.9 MHz, C6D6): δ -172 (partially resolved triplet,
1
51 14
J
) 97 Hz, ∆ν1/2 ) 276 Hz).
V,
N
Synthesis of [V(dN-p-Tol)(NEt2)3] (5b). LiNEt2 (2.04 g, 25.8
mmol) in Et2O (40 mL) was added to a stirred suspension of [V(dN-
p-Tol)Cl3] (3b) (2.07 g, 7.9 mmol) in Et2O (30 mL) at 0 °C. After
stirring for 3 h at room temperature the solvent was removed in vacuo.
The remaining solid was extracted with pentane (50 mL). Removal of
the volatiles in vacuo yielded a red oil (2.74 g, 92%).
3
1H NMR (300 MHz, C6D6): for 10, δ 1.26 (d, 6 H, JH,H ) 6 Hz,
OCHMe2), 1.40 (d, 6 H, 3JH,H ) 6 Hz, OCHMe2), 2.05 (s, 3 H, NAr-
Me), 2.14 (s, 3 H, OAr-Me), 2.22 (s, 3 H, OAr-Me), 2.34 (s, 6 H,
3
1H NMR (300 MHz, C6D6): δ 1.18 (t, 18 H, JH,H ) 7.0 Hz,
N(CH2CH3)2), 2.11 (s, 3 H, NAr-Me), 3.78 (q, 12 H, 3JH,H ) 7.0 Hz,
3
NMe2), 3.71 (s, 2 H, CH2), 5.90 (sept, 2 H, JH,H ) 6 Hz, OCHMe2),
3
N(CH2CH3)2), 6.92 and 7.21 (AB, 4 H, JH,H ) 8.0 Hz, N-Ar).
6.83 and 7.44 (AB, 4 H, 3JH,H ) 8 Hz, N-Ar); for 11, δ 1.34 (d, 6 H,
3JH,H ) 6 Hz, OCHMe2), 1.89 (s, 3 H, NAr-Me), 2.24 (s, 6 H, OAr-
Me), 2.27 (s, 12 H, NMe2), 2.47 (s, 6 H, OAr-Me), 3.57 and 3.88
(AB, 4 H, 2JH,H ) 13 Hz, CH2), 5.96 (sept, 1 H, 3JH,H ) 6 Hz, OCHMe2),
(8) van der Schaaf, P. A.; Jastrzebski, J. T. B. H.; Hogerheide, M. P.;
Smeets, W. J. J.; Spek, A. L.; van Koten, G. Inorg. Chem. 1993, 32,
4111.
(9) Lutz, M.; Hagen, H.; Schreurs, A. M. M.; Spek, A. L.; van Koten, G.
Acta Crystallogr. 1999, C55, 1636.
(10) Prandtl, W.; Hess, L. Z. Anorg. Chem. 1913, 82, 103.
(11) van der Schaaf, P. A.; Boersma, J.; Smeets, W. J. J.; Spek, A. L.; van
Koten, G. Inorg. Chem. 1993, 32, 5108.
(12) Rehder, D. In Studies in Inorganic Chemistry 13: Transition Metal
NMR; Pregosin, P. S., Ed.; Elsevier: Amsterdam, The Netherlands,
1991.
3
6.59 and 6.71 (AB, 4 H, JH,H ) 8 Hz, N-Ar).
51V NMR (78.9 MHz, CDCl3): for 10, δ -444 (∆ν1/2 ) 450 Hz);
for 11, δ -358 (∆ν1/2 ) 808 Hz).
Attempted Synthesis of [V(dN-p-Tol)(O-i-Pr)(OC6H2(CH2NMe2)-
2-Me2-4,6)2] (11). To a solution of [V(dN-p-Tol)(O-i-Pr)3] (9b) (0.88
g, 2.6 mmol) in Et2O (25 mL) was added HOC6H2(CH2NMe2)-2-Me2-
4,6 (1) (0.96 g, 5.4 mmol) in Et2O (10 mL). After stirring for 16 h the

3972 Inorganic Chemistry, Vol. 39, No. 18, 2000
Hagen et al.
Table 1. Crystallographic Data for 6‚Et2O, 6‚C6H6, and 12
6‚Et2O
6‚C6H6
12
empirical formula
fw
cryst syst
space group
a, Å
b, Å
c, Å
R, deg
â, deg
γ, deg
V, Å3
C28H37N3O2V‚Et2Oa
C28H37N3O2V‚C6H6
576.65
C8H10Cl2NO2V
274.01
orthorhombic
Pbca (No. 61)
11.8576(12)
12.6710(13)
14.722(2)
90
90
90
2211.9(4)
8
1.646
572.69a
triclinic
P1h (No. 2)
11.1557(6)
12.5744(12)
13.1051(14)
64.244(8)
70.472(7)
87.950(6)
1547.4(3)
2
triclinic
P1h (No. 2)
8.6034(3)
13.3614(4)
15.1044(5)
98.182(3)
105.618(2)
107.130(2)
1551.0(1)
2
Z
D
calcd, g cm-3 a
1.229
0.4
614
1.235
0.4
614
µ(Mo KR), mm-1
1.4
1104
F(000)a
T, K
150
0.0666
0.1634
0.96
150
0.0575
0.1590
1.06
225
final R1b
0.0520
0.1378
1.04
final wR2c
goodness of fit
a With disordered solvent contribution. b R1 ) Σ||Fo| - |Fc||/Σ|Fo|. c wR2 ) [Σ[w(Fo - Fc2)2]/Σ[w(Fo )2]]1/2
.
2
2
solvent was removed in vacuo, leaving a dark-red oil. This oil was
extracted with pentane (50 mL). Removal of the solvent in vacuo gave
a dark-red sticky solid (1.53 g). 1H NMR analysis showed the product
to be a mixture of HOC6H2(CH2NMe2)-2-Me2-4,6 (1), the monophe-
nolate [V(dN-p-Tol)(O-i-Pr)2(OC6H2(CH2NMe2)-2-Me2-4,6)] (10), and
the (bis)phenolate [V(dN-p-Tol)(O-i-Pr)(OC6H2(CH2NMe2)-2-Me2-
4,6)2] (11) in a 1:1:2 ratio.
NAr-C6), 128.7 (NAr-C3 and NAr-C5), 139.9 (Py-C4), 140.4 (NAr-
C4), 148.7 (Py-C6), 162.6 (NAr-C1), 170.2 (Py-C2).
51V NMR (78.9 MHz, CDCl3): δ 33 (∆ν1/2 ) 1100 Hz).
Synthesis of [VOCl2(O-i-Pr)] (14). To a solution of [VO(O-i-Pr)3]
(8) (1.75 g, 7.2 mmol) in toluene (25 mL) was added dropwise [VOCl3]
(1.35 mL, 14.3 mmol). The color changed from colorless to dark-red.
The solvent was removed in vacuo, and the residue was stripped once
with pentane (25 mL), leaving a dark-red liquid. Flash distillation gave
a dark-orange liquid (3.61 g, 85%). Due to the volatile nature of this
compound a membrane pump (working pressure approximately 10
mbar) was used instead of an oil pump.
Synthesis of [VOCl2(OCMe2([2]-Py))] (12). A suspension of
LiOCMe2([2]-Py) (1.43 g, 8.6 mmol) in toluene (50 mL) was added to
a stirred solution of [VOCl3] (1.51 g, 8.7 mmol) in pentane (40 mL).
The solution turned dark-red immediately, and an orange solid
precipitated. The mixture was stirred for 16 h. The resulting pale-orange
solution was separated from the precipitate by centrifugation/decanta-
tion. The residue was extracted with CH2Cl2 (2 × 40 mL), and the
volume of the combined fractions was reduced to 40 mL in vacuo.
Addition of an equal amount of Et2O followed by reduction of the
volume to 15 mL in vacuo resulted in the precipitation of a yellow-
orange solid. Decantation of the supernatant solution followed by
washing with Et2O (25 mL) and drying in vacuo gave [VOCl2(OCMe2-
([2]-Py))] (0.90 g, 38%). Combining the supernatant solution with the
ethereal fraction and leaving this solution to stand for 2 days at room
temperature gave a second crop (0.47 g; 20%). Mp: 150 °C.
Anal. Calcd for C8H10NO2VCl2: C, 35.07; H, 3.68; N, 5.11. Found:
C, 35.02; H, 3.72; N, 5.11.
3
1H NMR (300 MHz, CDCl3): δ 1.68 (d, 6 H, JH,H ) 6.2 Hz,
OCHMe2), 6.96 (br, 1 H, OCHMe2).
X-ray Crystallographic Structure Determination and Refine-
ments. Dark-red crystals of 6 suitable for X-ray diffraction were grown
either from Et2O at -20 °C or from benzene at room temperature.
Bright-orange crystals of 12 were grown from a 1:1 mixture of CH2-
Cl2 and Et2O at -20 °C. Crystals suitable for X-ray structure
determination were glued to the top of a Lindemann-glass capillary
and transferred into a cold nitrogen stream on an Enraf-Nonius CAD4-T
diffractometer (6‚Et2O, 12) or on a Nonius KappaCCD diffractometer
(6‚C6H6, 9b), both on rotating anodes. Crystal data and details on data
collection are collected in Table 1. Data were collected using graphite-
monochromated Mo KR radiation (λ ) 0.71073 Å). Data were corrected
for Lp effects but not for absorption. The structure of 6‚C6H6 was solved
by automated direct methods (SIR-9713).The other structures were
solved by automated Patterson methods and subsequent difference
Fourier techniques (SHELXS-8614 for 6‚Et2O, DIRDIF-9215 for 12).
Refinement on F2 was carried out using full-matrix least-squares
techniques (SHELXL-9716); no observance criterion was applied during
1H NMR (300 MHz, CDCl3): δ 1.98 (s, 6 H, OCMe2), 7.39 (d, 1
3
3
H, JH,H ) 7.7 Hz, Py-H), 7.48 (t, 1 H, JH,H ) 6.5 Hz, Py-H), 8.05
(t, 1 H, 3JH,H ) 7.3 Hz, Py-H), 8.74 (d, 1 H, 3JH,H ) 5.4 Hz, Py-H).
51V NMR (78.9 MHz, CDCl3): δ -235 (∆ν1/2 ) 93 Hz).
Synthesis of [V(dN-p-Tol)Cl2(OCMe2([2]-Py))] (13). LiOCMe2-
([2]-Py) (1.10 g, 6.6 mmol) was added as a solid to a stirred solution
of [V(dN-p-Tol)Cl3] (3b) (1.72 g, 6.6 mmol) in C6H6 (100 mL), and
the mixture was stirred for 3 h. The greenish solution was separated
from the precipitate by centrifugation. The remaining solid was extracted
with C6H6 (50 mL). The volume of the combined fractions was reduced
to 50 mL in vacuo. Pentane (100 mL) was added, and the solution was
left to stand for 2 weeks, affording dark greenish-blue crystalline
material (1.40 g, 58%). Recrystallization from CHCl3 gave dark-green
needle-shaped crystals, which turned red-brownish upon drying. Mp:
>200 °C.
refinement. For all compounds refinement converged with ∆/σmax
<
0.01 and ∆/σav < 0.01. Hydrogen atoms were included in the
refinement. All non-hydrogen atoms were refined with anisotropic
displacement parameters. The hydrogen atoms were included in the
refinement with a fixed isotropic atomic displacement parameter related
to the value of the equivalent isotropic atomic displacement parameter
of their carrier atoms.
(13) Altomare, A.; Burla, M. C.; Camalli, M.; Cascarano, G. L.; Giaco-
vazzo, C.; Guagliardi, A.; Moliterni, A. G. G.; Polidori, G.; Spagna,
R. J. Appl. Crystallogr. 1999, 32, 115.
Anal. Calcd for C15H17N2OVCl2‚0.4CHCl3: C, 45.01; H, 4.27; N,
6.82. Found: C, 45.65; H, 4.04; N, 6.81.
(14) Sheldrick, G. M. SHELXS-86 Program for crystal structure determi-
nation; University of Go¨ttingen: Go¨ttingen, Germany, 1986.
(15) Beurskens, P. T.; Admiraal, G.; Beurskens, G.; Bosman, W. P.; Garc´ıa-
Granda, S.; Gould, R. O.; Smits, J. M. M.; Smykalla, C. The DIRDIF
program system, Technical report of the Crystallography Laboratory;
University of Nijmegen: Nijmegen, The Netherlands, 1992.
(16) Sheldrick, G. M. SHELXL-97 Program for crystal structure refinement;
University of Go¨ttingen: Go¨ttingen, Germany, 1997.
1H NMR (300 MHz, CDCl3): δ 1.77 (s, 6 H, OCMe2), 2.41 (s, 3 H,
NAr-Me), 7.05 and 7.26 (AB, 4 H, 3JH,H ) 8.4 Hz, N-Ar), 7.4 (m, 2
3
4
H, Py-H3 and Py-H5), 7.94 (double t, 1 H, JH,H ) 7.7 Hz, JH,H
)
1.1 Hz, Py-H4), 8.84 (d, 1 H, JH,H ) 5.4 Hz, Py-H6).
13C NMR (75.5 MHz, CDCl3): δ 21.6 (NAr-Me), 30.6 (OCMe2),
96.7 (OCMe2), 120.2, 123.8 (Py-C3 and Py-C5), 126.7 (NAr-C2 and
3

Vanadium(IV) and -(V) Complexes
Inorganic Chemistry, Vol. 39, No. 18, 2000 3973
Scheme 1
the reaction mixtures gave any evidence for the formation of
such products. Furthermore, p-benzoquinone is known to oxidize
vanadium(IV) compounds,19 and several vanadium(V) phenolate
complexes that are stable toward reduction are known.20
Therefore, it is unlikely that the reduction is caused by the
phenolic part of the ligand. It is more likely that the amine
function in the ligand is responsible for the reduction. The
reduction of vanadium(V) compounds in the presence of
aliphatic amines has been reported before: [VOCl2(NMe3)2] can
21,22
be synthesized by stirring [VOCl3] in NMe3.
The chloride
“lost” from the vanadium(V) starting material is recovered as
the ammonium salt, HNMe3Cl. The more protons are present
at the R-carbon atoms, the more facile the reduction becomes.23
Interestingly, the reduction of [VOCl3] by 1 can be used as
a tool for the synthesis of vanadium(IV) (bis)phenolates. By
addition of at least 2.5 equiv of the phenol to [VOCl3] the
already known vanadium(IV) oxo (bis)phenolate [VO(O-MD-
MBA)2] (4)5 was prepared in high yield on a multigram scale
(Scheme 1, ii).
(i) 1 equiv of RO-MDMBA (R ) H, SiMe3, Li, Na). (ii) 2.5 equiv
of HO-MDMBA, excess Et3N, C6H6, room temperature. (iii) R ) Ph
(a), p-Tol (b); 3.3 equiv of LiNEt2, Et2O, 0 °C. (iv) R ) Ph; 1 equiv
of HO-MDMBA, hexane, 0 °C. (v) R ) Ph (a), p-Tol (b); 2 equiv of
HO-MDMBA, hexane, room temperature.
To avoid reduction of the vanadium, a less electrophilic
vanadium starting material was used. Vanadium arylimido (tris)-
diethylamides, [V(dNR)(NEt2)3] (R ) Ph (5a), p-Tol (5b)), can
be prepared via a metathesis reaction between 3a or 3b and 3
equiv of lithium diethylamide (Scheme 1, iii). Compounds 5a
and 5b are obtained as dark-red oils. Attempts to prepare 5a
by refluxing 3a in neat diethylamine were not successful.
Although a reaction did take place, a mixture of unidentified
products was obtained.
For 6‚Et2O 7723 reflections were measured (1.8° < θ < 27.5°, -14
< h < 8, -16 < k < 16, -17 < l < 16), 7094 of which were unique
(Rint ) 0.0745) using a dark-red crystal of approximate dimensions
0.1 × 0.4 × 0.4 mm; 315 parameters were refined, and the final residual
density was in the range -0.37 < ∆F < 0.42 e/Å3. The structure
contained Et2O as disordered solvent. The associated contribution to
the structure factors was taken into account using the SQUEEZE
procedure as incorporated in PLATON.17 Eighty-one electrons were
found in a solvent area of 320 Å3, consistent with one solvent molecule
per vanadium complex.
For 6‚C6H6 18775 reflections were measured (1.6° < θ < 25.0°,
-10 < h < 10, -15 < k < 15, -17 < l < 17), 5418 of which were
unique (Rint ) 0.0530) using a brown crystal of approximate dimensions
0.03 × 0.21 × 0.27 mm; 361 parameters were refined, and the final
residual density was in the range -0.51 < ∆F < 0.54 e/Å3.
For 12 2992 reflections were measured (2.7° < θ < 27.5°, -15 <
h < 0, -16 < k < 0, -19 < l < 12), 2533 of which were unique (Rint
) 0.071) using a bright-orange crystal of approximate dimensions 0.1
× 0.5 × 0.6 mm; 129 parameters were refined, and the final residual
density was in the range -0.54 < ∆F < 0.39 e/Å3.
In order to prepare the vanadium(V) monophenolate [V(d
NPh)(NEt2)2(O-MDMBA)], 5a was reacted with 1 equiv of the
parent phenol of 1. Although quantitative conversion of 1
occurred, 1H NMR analysis showed that a complicated mixture
of products had formed. One of the reaction products that was
isolated by selective crystallization was the vanadium(IV)
species [V(dNPh)(O-MDMBA)2] (6) (Scheme 1, iv). EPR
analysis of 6 in solution at room temperature shows the expected
characteristic octet structure (Aiso ) 85.28(1) × 10-4 cm-1, giso
) 1.9717(6)) due to hyperfine interaction between the vanadium
nucleus (I ) 7/2) and the unpaired electron. The hyperfine
interactions of the nitrogens are not resolved. Recent EPR
studies on vanadium(IV) oxo (bis)phenolate complexes showed
that the value of the hyperfine couplings can be indicative for
the electronic influence of the ligands.5 The difference between
Neutral atom scattering factors and anomalous dispersion corrections
were taken from the International Tables for Crystallography.18
Geometrical calculations and illustrations were performed with PLA-
TON.17
the hyperfine couplings of 6 and 4 (Aiso ) 93.9 × 10-4 cm-1
)
is in accord with earlier observations that the phenylimido ligand
Results and Discussion
is a stronger electron donor than the oxo group.6
Initial attempts to prepare vanadium(V) complexes with the
o-aminophenolate anion 1 were carried out by adding the parent
phenol, the lithium phenolate or sodium phenolate, or the
corresponding trimethylsilyl ether to vanadium arylimido trichlo-
ride, [V(dNR)Cl3] (R ) Ph (3a), p-Tol (3b)) (Scheme 1, i).
These reactions were generally accompanied by a color change
from red/orange to very dark blue or green. On the basis of
In contrast to these findings, reaction of 2 equiv of 1 with
either 5a or 5b did not result in reduction products, but instead
gave the vanadium(V) (bis)phenolate complexes [V(dNR)-
(NEt2)(O-MDMBA)2] (R ) Ph (7a), p-Tol, (7b)), in quantitative
yields (Scheme 1, v). Unfortunately both compounds 7a and
7b have low thermal stability. In the solid state at -20 °C they
are stable for several months at least, but at room temperature
decomposition is observed within several days. In benzene
solution 7b decomposes into an unidentified paramagnetic
compound within 24 h. However, the latter reduction is
1
both the lack of signal in the H NMR and the presence of a
strong eight-line signal in the EPR, it was concluded that the
vanadium was no longer in the formal 5+ oxidation state and
that a paramagnetic vanadium(IV) species had been formed.
An explanation for the reduction of vanadium(V) to vanadium-
(IV) could be either the oxidation of the phenol to a quinone or
the oxidative coupling of the phenol to a dimer or higher
oligomer. However, neither 1H NMR nor GC-MS analysis of
(19) Pasquali, M.; Landi, A.; Floriani, C. Inorg. Chem. 1979, 18, 2397.
(20) Masthoff, R.; Ko¨hler, H.; Bo¨hland, H.; Schmeil, F. Z. Chem. 1965, 5,
122.
(21) Baker, K. L.; Edwards, D. A.; Fowles, G. W. A.; Williams, R. G. J.
Inorg. Nucl. Chem. 1967, 29, 1881.
(17) Spek, A. L. Acta Crystallogr. 1990, A46, C-34.
(18) International Tables for Crystallography; Wilson, A. J. C., Ed.; Kluwer
Academic Publishers: Dordrecht, The Netherlands, 1992; Vol C.
(22) Drake, J. E.; Vekris, J.; Wood, J. S. J. Chem. Soc. A 1968, 1000.
(23) Preuss, F.; Fuchslocher, E.; Leber, E.; Towae, W. Z. Naturforsch. 1989,
44b, 271.

3974 Inorganic Chemistry, Vol. 39, No. 18, 2000
Hagen et al.
Figure 2. Fluxional behavior of compound 7b in solution.
sufficiently slow to allow study of the structural features of 7b
in solution by NMR spectroscopy.
On the basis of these findings we propose that 7b is a five-
coordinate species. This conclusion is corroborated by 51V NMR
data (vide infra). The geometry around the vanadium is best
described starting from an octahedral geometry with the two
aminophenolate ligands in one plane with a trans N,N-
conformation. De-coordination of one of the NMe2 groups
creates a five-coordinate species with the second NMe2 group
at the top of a square pyramid. The basal plane is formed by
the two oxygen atoms of the phenolate ligands and the two
nitrogen atoms of the imido and amide ligand. Due to both steric
hindrance and the preferred bite angle of the aminephenolate
ligand of approximately 90°, this geometry distorts to a trigonal
bipyramidal one with the three nitrogen atoms in the equatorial
plane and the two oxygen atoms at the apical positions. This
geometry was confirmed by Molecular Mechanics calculations.
The 1H NMR spectra of 7a and 7b in C6D6 at room
temperature show one unique pattern of aminophenolate reso-
nances. While the NMe2 methyl groups appear as one singlet,
the benzylic protons give rise to a well-resolved AB pattern,
which indicates the presence of diastereotopic benzylic protons
(vide infra). Both the methylene and methyl signals of the amide
ligand appear as broad peaks.
To get more insight into the molecular geometry of 7b in
solution, a variable-temperature 1H NMR study was carried out.
The AB pattern for the benzylic protons remains the same, upon
raising the temperature to 100 °C in toluene-d8. Although the
signals for the methyl and methylene protons of the diethylamide
group get sharper, they remain poorly resolved. The methylene
protons show up as a broad signal with some fine structure,
whereas the methyl protons appear as a poorly resolved double
triplet. This indicates that the methylene protons are diaste-
reotopic as well. Lowering the temperature results in broadening
of both the methyl and the methylene signals, and they
decoalesce into two sets of two broad singlets at 15 and 0 °C,
respectively (∆G‡ ) 57 kJ‚mol-1).
Further lowering of the temperature causes a broadening of
the signals for the benzylic protons, and at -70 °C two poorly
resolved AB patterns are observed. Apart from these changes
also the signals of the NMe2 groups and those of the methylene
protons of the NEt2 group have changed. However, the
interpretation of the spectra at these low temperatures is severly
hampered by extensive broadening of all peaks. In fact only
the benzylic signals are separated sufficiently to give information
about the possible structure of 7b. The spectral features may
be interpreted as follows.
In the slow exchange limit (below -40 °C) V-N coordina-
tion exists, which is rigid on the NMR time scale. One of the
aminophenolate ligands is bonded in an η2-O,N motif, whereas
the other aminophenolate is η1-bonded via the oxygen atom
exclusively. This will result in two different AB patterns for
the benzylic protons. As a consequence of the rigid η2-O,N
coordination of one of the aminophenolate ligands, one singlet
for the free NMe2 group of the η1-O-bonded ligand and two
singlets for the coordinated NMe2 group are expected. Unfor-
tunately this was not observed due to the extensive broadening
of the signals in this region of the spectrum.
The exchange between the coordinated and noncoordinated
amine functionalities may happen via a facial attack of the
noncoordinated amine to give a six-coordinate octahedral
transition state and subsequent de-coordination of the second
amine (see Figure 2). Although an inversion at the vanadium
center takes place, due to the exchange between the η1-O- and
the η2-O,N-bonded aminophenolate ligands the configuration
at the vanadium center is retained as is observed by the AB
pattern for the benzylic protons in the fast exchange limit. If
the exchange would happen via a dissociative pathway with a
four-coordinate tetrahedral transtition state, an apparent molec-
ular plane of symmetry would be present in the fast exchange
limit and the benzylic protons would show up as singlets.
At temperatures below 0 °C the two ethyl groups of the
diethylamide moiety appear as two sets of broad singlets. For
the two ethyl groups to be chemically unequivalent, the rotation
around the V-N bond has to be slow on the NMR time scale.
We attribute this slow rotation around the V-N bond to the
presence of a vanadium-nitrogen double bond. A filled p-orbital
is present on the nitrogen atom, and this orbital may overlap
with an empty orbital on the vanadium atom and thus form a
vanadium-nitrogen double bond.
Since the presence of diethylamide ligands still leaves the
vanadium center too electrophilic to prevent its reduction by
the amine functionality, stronger electron-donating isopropoxide
ligands were used. Treatment of vanadium oxo (tris)isopro-
poxide (8) with phenyl or p-tolyl isocyanate in boiling octane
resulted in the formation of vanadium phenyl- or p-tolylimido
(tris)isopropoxides (9a,b) in high yields (see Scheme 2, i). Both
9a and 9b are stable, low-melting (53 and 43 °C) orange solids
which can be crystallized from hexane. They are soluble in
alkanes like pentane and hexane, in aromatics like benzene and
toluene, and in polar solvents like Et2O, CH2Cl2, and THF.
Cryoscopic measurements in benzene showed 9b to be mono-
meric in solution (n ) 1.08(2)).
At temperatures above -40 °C the exchange between the
coordinated and noncoordinated amine functionalities becomes
fast on the NMR time scale and the two states, both containing
one η1-O- and one η2-O,N-bonded aminophenolate ligand,
cannot be distinguished. Therefore only one set of aminophe-
nolate signals is observed. The prochiral benzylic protons show
up as an AB pattern, since there is no molecular plane of
symmetry containing the benzylic carbon atoms. The two methyl
groups of the NMe2 group appear as one singlet, because both
pyramidal inversion and rotation around the benzylic carbon-
nitrogen axis, which are taking place in the noncoordinated CH2-
NMe2 groups, are processes with very small activation barriers.
The reaction of 9b with 1 equiv of the aminophenol 1 yielded
1
a mixture of compounds. According to H NMR analysis a
mixture of unreacted 9b, the vanadium(V) (mono)phenolate
[V(dN-p-Tol)(O-MDMBA)(O-i-Pr)2] (10), and the vanadium-
(V) (bis)phenolate [V(dN-p-Tol)(O-MDMBA)2(O-i-Pr)] (11)

Vanadium(IV) and -(V) Complexes
Inorganic Chemistry, Vol. 39, No. 18, 2000 3975
Scheme 2
Figure 3. Displacement ellipsoid plot of 6‚Et2O, drawn at 50%
probability. Hydrogen atoms are omitted for clarity.
Scheme 3). In this way [VOCl2(OCMe2([2]-Py))] (12) and [V(d
N-p-Tol)Cl2(OCMe2([2]-Py))] (13) were obtained. Compound
12 is a stable orange solid which melts without decomposition
at 150 °C. It is insoluble in alkanes, poorly soluble in Et2O and
aromatic solvents like benzene and toluene, but soluble in CH2-
Cl2. Compound 13 is a green solid, which is poorly soluble in
most solvents. It does not melt at temperatures up to 200 °C.
The ease with which these compounds are prepared and their
stability toward reduction are remarkable in comparison to the
partly unsuccessful attempts to prepare vanadium(V) complexes
with the aminophenolate 1. This is particularly true when it is
taken into account that pyridine itself causes reduction of
[VOCl3].24 The reason for this difference is not known.
(i) R-NCO, reflux in octane (R ) Ph (a), p-Tol (b)). (ii) R ) p-Tol;
1 equiv of HO-MDMBA. (iii) R ) p-Tol; 2 equiv of HO-MDMBA.
Scheme 3. Synthesis of [VOCl2(OCMe2([2]-Py))] (12) and
[V(dN-p-Tol)Cl2(OCMe2([2]-Py))] (13)
1
The H NMR spectra of both 12 and 13 are consistent with
their expected structures. The chemical shift of the hydrogen
atom ortho to the pyridyl nitrogen indicates that the pyridine
of the pyridylalkoxide is involved in coordination.11,25,26 The
difference between the chemical shift of the pyridine-H(6)
proton in the metal complex and the parent alcohol (∆δ)
suggests that in both 12 (∆δ ) 0.49) and 13 (∆δ ) 0.59) the
ligand is η2-O,N-bonded to the vanadium center.
had formed (Scheme 2, ii). On the basis of the size of the signal
of the methyne protons of the isopropoxy groups, the ratio of
9b, 10, and 11 was approximately 1:2:1.
When 9b was reacted with 2 equiv of the aminophenol 1
again a mixture of compounds was obtained, viz., a 1:1:2
mixture of the aminophenol 1, the (mono)phenolate 10, and the
(bis)phenolate 11 (Scheme 2, iii). This ratio is shifted upon the
addition of more phenol. Reacting 9b with 3 or 4 equiv of 1
led to the formation of mixtures of 10 and 11 in the molar ratios
1:5 and 1:9, respectively. The formation of a tris-substituted
product was never observed.
Solid-State Structures of 6 and 12. Crystallization of the
vanadium imido (bis)phenolate 6 from Et2O gave crystals
suitable for X-ray diffraction analysis. Its molecular structure
is depicted in Figure 3, and pertinent bond lengths and bond
angles are given in Table 2. The structure of 6 is very similar
to that of another vanadium(IV) (bis)phenolate compound, [VO-
(OC6H4(CH2NMe2)-2)2].5 The vanadium is surrounded in a
trigonal bipyramidal fashion with a considerable distortion
toward a square pyramidal geometry (39% on the Berry
pseudorotation path between D3h and C4V 27 vs 37% for the oxo
compound). The oxygen atoms of the aminophenolate ligand
(O1 and O2) and the nitrogen atom of the imido group (N3)
form the equatorial plane. The amino nitrogen atoms (N1 and
N2) occupy the apical positions. The V-O, V-N, and O-C
distances as well as the bite angles of the phenolate ligands
(O-V-N) are comparable with those in the oxo compound
[VO(OC6H4(CH2NMe2)-2)2], whereas the N1-V1-N2 angle
is slightly smaller (163.72(10)° vs 165.13(11)° in the oxo
compound).
Although the vanadium (mono)phenolate 10 could not be
isolated as a pure compound, the characteristic signals for the
isopropoxy and aminophenolate ligands were well-separated
from those of the vanadium (bis)phenolate 11. 1H NMR analysis
of the mixture showed two doublets for the two isopropoxy
groups of 10. The benzylic and NMe2 signals of the phenolate
ligand, however, appeared as singlets. Assuming a monomeric
configuration for 10, Molecular Mechanics calculations showed
its geometry to be most likely trigonal bipyramidal. The
equatorial plane would then be formed by the p-tolylimido unit,
an isopropoxide, and the coordinating amine of the phenolate
ligand. The axial positions would be occupied by the second
isopropoxide and the oxygen of the phenolate ligand. However,
since in such a structure two different isopropoxide groups are
present, the benzylic protons would be diastereotopic and should
show up as an AB pattern in the 1H NMR spectrum. As this is
not the case, it is unlikely that 10 is a mononuclear complex,
and a dimeric structure with two bridging and two terminal
isopropoxy groups similar to the solid-state structure of 9b9
seems to be more likely.
A second pseudopolymorph of compound 6 has been found
after crystallization from benzene at room temperature. Although
both crystal structures are triclinic with a P1h space group and
(24) Funk, H.; Weiss, W.; Zeising, M. Z. Anorg. Chem. 1958, 296, 36.
(25) van der Schaaf, P. A.; Abbenhuis, R. A. T. M.; van der Noort, W. P.
A.; de Graaf, R.; Grove, D. M.; Smeets, W. J. J.; Spek, A. L.; van
Koten, G. Organometallics 1994, 13, 1433.
(26) Brandts, J. A. M.; Boersma, J.; Spek, A. L.; van Koten, G. Eur. J.
Inorg. Chem. 1999, 1727.
In order to obtain vanadium(V) complexes containing the
pyridylalkoxide [OCMe2([2]-Py)]- (2), [VOCl3] and [V(dN-
p-Tol)Cl3] (3b) were reacted with LiOCMe2([2]-Py) (see
(27) Holmes, R. R. Prog. Inorg. Chem. 1984, 32, 119 and references therein.

3976 Inorganic Chemistry, Vol. 39, No. 18, 2000
Hagen et al.
Table 2. Selected Bond Lengths and Bond Angles for 6‚Et2O,
6‚C6H6, and 12
6‚Et2O
6‚C6H6
12
Bond Lengths (Å)
V1-O1
V1-O2
V1-N1
V1-N2
O1-C1
O2-C12
V1-Cl1
V1-Cl2
V1-N3
N3-C23
O2-C1
1.887(2)
1.901(2)
1.911(2)
2.167(3)
2.181(3)
1.348(4)
1.349(4)
1.578(3)
1.763(3)
2.163(3)
1.911(2)
2.164(2)
2.168(2)
1.351(5)
1.340(4)
2.2282(14)
2.2195(14)
1.685(3)
1.370(5)
1.681(3)
1.376(4)
1.435(4)
Figure 4. Displacement ellipsoid plot of 12, drawn at 50% probability.
Hydrogen atoms are omitted for clarity.
Bond Angles (deg)
O1-V1-N1
O2-V1-N2
V1-N3-C23
V1-O1-C1
V1-O2-C12
O1-V1-O2
O2-V1-N3
O1-V1-N3
N1-V1-N2
O1-V1-Cl2
O2-V1-Cl2
O1-V1-Cl1
N1-V1-Cl1
N1-V1-O2
N1-V1-Cl2
V1-O2-C1
87.69(10)
87.16(10)
161.7(2)
87.90(11)
87.72(11)
174.2(3)
89.48(14)
other vanadium(V) alkoxo compounds (1.74-1.99 Å).28-30
However, the V-Cl and the VdO distances (av 2.23 and 1.578-
(3) Å, respectively) are among the shortest ever reported. Also
in view of the small bite angle of the η2-O,N-chelate bonded
pyridylalkoxide ligand of 75.39(12)°, this structure is probably
best described starting from a tetrahedral geometry. The
vanadium center is surrounded by two oxygen and two chlorine
atoms. Facial attack of the pyridyl nitrogen at the O1-Cl2-
O2 face creates a pseudo-five-coordinate structure following an
SN2 type of reaction coordinate. This view seems corroborated
by the fact that the V-Cl1 distance is slightly lengthened.
51V NMR Spectroscopy. The 51V nucleus is well suited for
NMR studies because of its high abundance (99.76%) and high
sensitivity (0.382 relative to that of the 1H nucleus at equal field
strength). The vanadium NMR data of the arylimido compounds
and those of related oxo vanadium(V) compounds are listed in
Table 3. The chemical shifts range from δ ) 305 ppm for [V(d
N-p-Tol)Cl3] to δ ) -624 ppm for [VO(O-i-Pr)3]. For
compounds 5a,b and 9a,b the signals are split up in a 1:1:1
triplet (partly resolved for 5a,b) with coupling constants of
approximately 100 Hz due to the coupling with the 14N atom (I
) 1) of the imido functionality. The chemical shifts of the
arylimido compounds follow the so-called “inverse halogen
dependence” in which the nucleus becomes more shielded with
increasing electronegativity of the substituents: δ [V(dNR)-
Cl3] > δ [V(dNR)(NEt2)3] > δ [V(dNR)(O-i-Pr)3].
133.6(2)
134.1(2)
129.83(12)
117.14(12)
113.03(12)
163.72(10)
133.0(2)
133.7(2)
125.37(11)
118.88(12)
115.73(12)
168.38(11)
113.17(13)
110.64(10)
131.29(9)
103.57(10)
166.31(10)
75.39(12)
84.85(11)
129.8(2)
the unit cells have approximately the same size, the unit cell
dimensions are different. Also the molecular structure of 6‚C6H6
is different. The geometry around the vanadium center is less
distorted toward a square pyramid (24% on the Berry pseu-
dorotation path), which is visible in the slightly larger N1-
V-N2 angle of 168.38(11)°. More remarkable is the much
larger V-N3-C23 angle of 174.2(3)° vs 161.7(2)°. Both
polymorphs contain 1 equiv of the solvent from which they were
crystallized, and the steric interference exerted by these solvent
molecules causes the observed geometrical differences. In the
polymorph obtained from benzene the solvent molecule is
located opposite to the imido functionality at the O1-N1-O2-
N2 face. This position can be considered as the sixth coordina-
tion side, and steric interference from the benzene molecule on
the imido functionality is expected to be negligible. In the
polymorph obtained from Et2O, however, the Et2O molecule is
located next to the imido functionality at the O1-N1-N3 face.
In this way it forces the imido functionality to bend away from
its ideal position, and as a result a smaller V-N3-C23 angle
is found. A smaller V-N-C angle could point to a decreased
V-N bond order. However, this is not reflected by a longer
V-N bond length in the molecular structure of 6‚Et2O. From
these observations it can be concluded that bond angles around
vanadium, as derived from molecular geometries of vanadium
complexes in the solid state, should be handled with care.
This inverse halogen dependence is also visible for the
arylimido functionality itself. Even though it is a stronger
electron donor than the oxo group (vide supra), due to the less
electronegative nitrogen atom it has a deshielding influence on
the vanadium nucleus as compared to the oxo functionality.
However, this effect weakens with the introduction of increas-
ingly more electronegative, π-donating ligands, something which
has also been observed in the series [V(dN-p-Tol)Cl3-x(O-t-
Bu)x] (x ) 0-3) in comparison with the analogous oxo
compounds.7
The 51V chemical shifts do not provide information about
the coordination number since the increase in the sum of
electronegativities of bonded atoms is counteracted by the
expansion of the coordination sphere.12 However, Crans and
Shin showed that a relation between the inverse line width and
the chemical shift might exist.31 Two distinct clusters were found
Crystals of 12 suitable for an X-ray structure determination
were obtained from a 1:1 mixture of CH2Cl2 and Et2O at -20
°C. The molecular structure is depicted in Figure 4 (see Table
2 for relevant bond lengths and angles). The geometry around
the vanadium center is a very distorted trigonal bipyramid (41%
on the Berry pseudorotation path). The equatorial plane is
formed by the two oxygen atoms O1 and O2 and by one of the
chlorine atoms, Cl2. The axial positions are occupied by the
pyridyl nitrogen atom N1 and the second chlorine atom Cl1.
The V-O distance (1.763(3) Å) falls in the range reported for
(28) Caughlan, C. N.; Smith, H. M.; Watenpaugh, K. Inorg. Chem. 1966,
5, 2131.
(29) Glas, H.; Herdtweck, E.; Artus, G. R. J.; Thiel, W. R. Inorg. Chem.
1998, 37, 3644.
(30) Crans, D. C.; Keramidas, A. D.; Amin, S. S.; Anderson, O. P.; Miller,
S. M. J. Chem. Soc., Dalton Trans. 1997, 2799.
(31) Crans, D. C.; Shin, P. K. J. Am. Chem. Soc. 1994, 116, 1305.

Vanadium(IV) and -(V) Complexes
Inorganic Chemistry, Vol. 39, No. 18, 2000 3977
Table 3. 51V NMR Data of Various Vanadium(V) Compoundsa
1
51 14
compound
[V(dNPh)Cl3] (3a)
[V(dN-p-Tol)Cl3] (3b)
δ (ppm)
∆ν1/2 (Hz)
J
N (Hz)
ref
V,
252
305
530
500
this work
7
[VO(NEt2)3]
-389
-172
-175
-624
-622
-611
-253
-444
-358
-309
-235
33
32
[V(dNPh)(NEt2)3] (5a)c
276
275
13
97b
this work
this work
this work
this work
this work
this work
this work
this work
32
[V(dN-p-Tol)(NEt2)3] (5b)c
[VO(O-i-Pr)3] (8)
[V(dNPh)(O-i-Pr)3] (9a)
[V(dN-p-Tol)(O-i-Pr)3] (9b)
[V(dN-p-Tol)(NEt2)(O-MDMBA)2] (7b)c
[V(dN-p-Tol)(O-MDMBA)(O-i-Pr)2] (10)
[V(dN-p-Tol)(O-MDMBA)2(O-i-Pr)] (11)
[VOCl2(O-i-Pr)] (14)
[VOCl2(OCMe2([2]-Py))] (12)
[V(dN-p-Tol)Cl2(OCMe2([2]-Py))] (13)
99b
113
114
137
450
808
93
1100
this work
this work
a CDCl3, 78.9 MHz, relative to external [VOCl3]/CDCl3 (75/25 v/v). b Partially resolved triplet. c In C6D6.
tively propose 7b to be five-coordinate as well, which was
1
already concluded from its dynamic behavior in the H NMR
(vide supra). For compounds 10, 11, and 13 a six-coordinate
structure is expected. On the basis of its 1H NMR data, a dimeric
structure with two bridging and two terminal isopropoxy groups
was proposed for 10. This implies that the amine function must
be coordinated to make it a six-coordinate species. For
compound 13 to be six-coordinate it has to be dimeric as well.
Whether the dimer is bridged via chlorine atoms or imido
functionalities remains inconclusive from these data.
Polymerization Tests. The compounds 12 and 13 were tested
in ethene polymerization together with several other nonchelated
vanadium(V) oxo and imido compounds. Whereas 12 displayed
a reasonable activity (88 kg PE/mol[V]‚h‚bar at 50 °C), 13
showed almost no activity under similar circumstances and for
both complexes the activities were much lower than those of
the nonchelated counterparts. We tentatively assign this decrease
in activity to a lower number of catalytically active species (for
more details see the Supporting Information).
Figure 5. Inverse line width, ∆ν1/2-1, plotted as a function of the 51
V
NMR chemical shifts for vanadium(V) complexes.
when the inverse line width was plotted against the chemical
shift of a series of vanadium(V) complexes. For five-coordinate
complexes a cluster was found in the quadrant defined by the
inverse line width values of 5-8 and chemical shift range -480
to -490 ppm. The second cluster containing six-coordinate
complexes occupied the area with inverse line width values of
1-4 and chemical shift range -490 to -526 ppm. The value
of 4 for the inverse line width appears to be the approximate
border value between five- and six-coordinate vanadium species.
The chemical shifts found for the complexes reported here do
not fall into these ranges, but since Crans and Shin examined
only cis-dioxo vanadium(V), complexes this is not surprising.
What is more important is that they found that coordination of
a second amine function does not significantly change the
chemical shift but only affects the value of the inverse line
width. In a plot of the inverse line width against the chemical
shift of the vanadium(V) complexes containing O,N-chelating
ligands (see Figure 5), two areas can be recognized: viz., one
with an inverse line width value smaller than 4 containing
compounds 10, 11, and 13 and one with values larger than 4
containing compounds 7b and 12. According to the X-ray
structure, 12 is a five-coordinate species. We therefore tenta-
Acknowledgment. This work was carried out in connection
with NIOK, The Netherlands Institute for Catalysis Research,
and was supported by the Department of Economic Affairs of
The Netherlands. Polymerization tests and polymer analysis
were carried out at the Center for Catalytic Olefin Polymeri-
zation in Groningen, The Netherlands. Thanks are due to E. A.
C. Brussee and A. Jekel for technical assistance. Crystallo-
graphical studies were supported in part (A.L.S., M.L.) by the
Councel for the Chemical Sciences of the Netherlands Orga-
nization for Scientific Research (CW-NWO). EPR analyses were
carried out at the Department of Atomic and Interface Physics,
Utrecht University, The Netherlands, by Dr. E. E. van Faassen.
Supporting Information Available: Detailed information on
polymerization activity and X-ray crystallographic files in CIF format
for the structures of 6‚Et2O, 6‚C6H6, and 12. This material is available
free of charge via the Internet at http://pubs.acs.org.
(32) Priebsch, W.; Rehder, D. Inorg. Chem. 1985, 24, 3058.
IC991415S