Imido-titanium/molybdenum heterobimetallic systems. Switching from η6-arene to fischer-type aminocarbene complexes by tuning reactivity conditions

Article abstract of DOI:10.1021/om900825a

Two types of bimetallic titanium/molybdenum systems are described: the first possesses unprecedented η⁶-arene-imido groups, while the second has Fischer-type aminocarbene ligands bridging the two metals. The (η⁶-arene)-Mo imido-Ti complex Ti[=N(η⁶-Ar)Mo(CO) ₃]Cl₂(NHMe₂)₂ (2; Ar = 2, 6-'Pr ₂-C₆H₃) has been prepared through the reaction of (η⁶-ArNH₂)Mo(CO)₃ (3) and TiCl ₂(NMe₂)₂. The alternative route from 2 and Ti(NMe₂)₄/Me₃SiCl also afforded 1, but it was contaminated with a small amount of Ti(=NAr)Cl₂(NHMe ₂)₂ (1). The reaction between the dimeric complexes {Ti(μ-NAr)(NMe₂)₂}₂ and Mo(CO)₆ gave complex mixtures of products, among which the Fischer-type aminocarbene complex [(CO)₅Mo{=C(NMeCH₂NMe₂)O}Ti(=N ₂,6-Prⁱ₂-C₆H₃)(NMe ₂)(NHMe₂)] (4) could be characterized. The formation of 4 resulted from the nucleophilic attack of the amido -NMe₂ to the carbonyl that afforded the titanoxy aminocarbene linker between the Mo and the Ti centers, followed by a subsequent C-H activation of a methyl group of the aminocarbene and C,N coupling with an amido ligand on titanium. Treatment of Ti(NMe₂)₄ with 1 equiv of Mo(CO)₆ produced the titanoxy aminocarbene [(CO)₅Mo{=C(NMe₂)O}Ti(NMe ₂)₃] (5), while in the presence of 1 equiv of ArNH ₂ the imido complex [(CO)₅Mo{=C(NMe₂)O} Ti(=N-2, 6Pr'₂-C₆H₃)(NMe₂)(NHMe ₂)₂] (6) is formed. The molecular structures of 1-6 have been determined by X-ray diffraction.

Full text of DOI:10.1021/om900825a

Organometallics 2010, 29, 1127–1136 1127
DOI: 10.1021/om900825a
Imido-Titanium/Molybdenum Heterobimetallic Systems. Switching
from η6-Arene to Fischer-Type Aminocarbene Complexes by Tuning
Reactivity Conditions
Christian Lorber*,†,‡ and Laure Vendier†,‡
†CNRS, LCC (Laboratoire de Chimie de Coordination), 205 route de Narbonne, F-31077 Toulouse, France,
‡
and Universite de Toulouse, UPS, INPT, LCC, F-31077 Toulouse, France
ꢀ
Received September 23, 2009
Two types of bimetallic titanium/molybdenum systems are described: the first possesses unprece-
dented η6-arene-imido groups, while the second has Fischer-type aminocarbene ligands bridging the
two metals. The (η6-arene)-Mo imido-Ti complex Ti[dN(η6-Ar)Mo(CO)3]Cl2(NHMe2)2 (2; Ar =
2,6-iPr2-C6H3) has been prepared through the reaction of (η6-ArNH2)Mo(CO)3 (3) and TiCl2-
(NMe2)2. The alternative route from 2 and Ti(NMe2)4/Me3SiCl also afforded 1, but it was
contaminated with a small amount of Ti(dNAr)Cl2(NHMe2)2 (1). The reaction between the dimeric
complexes {Ti(μ-NAr)(NMe2)2}2 and Mo(CO)6 gave complex mixtures of products, among which
the Fischer-type aminocarbene complex [(CO)5Mo{dC(NMeCH2NMe2)O}Ti(dN-2,6-Pri2-C6H3)-
(NMe2)(NHMe2)] (4) could be characterized. The formation of 4 resulted from the nucleophilic
attack of the amido -NMe2 to the carbonyl that afforded the titanoxy aminocarbene linker between
the Mo and the Ti centers, followed by a subsequent C-H activation of a methyl group of the
aminocarbene and C,N coupling with an amido ligand on titanium. Treatment of Ti(NMe2)4 with
1 equiv of Mo(CO)6 produced the titanoxy aminocarbene [(CO)5Mo{dC(NMe2)O}Ti(NMe2)3] (5),
while in the presence of 1 equiv of ArNH2 the imido complex [(CO)5Mo{dC(NMe2)O}Ti(dN-2,6-
Pri2-C6H3)(NMe2)(NHMe2)2] (6) is formed. The molecular structures of 1-6 have been determined
by X-ray diffraction.
Introduction
early-metal-amide bond,3 whereas such insertion is a
known reaction for late-transition-metal complexes4 and,
to a lesser extent, for some actinide compounds.5,6 It is worth
noting that alkali-metal salts of strongly basic nitrogen are
known to attack CO.7,8
As there is still a need for new catalysts, in particular those
that could operate under milder conditions, we have em-
barked on the study of bimetallic systems combining an early
transition metal (that could form M-NR bonds) with a
carbonyl complex (that could effect the carbonylation).
Herein we report our first results concerning the synthesis
The metal-catalyzed carbonylation of amines has been
intensively studied, as it gives access to important molecules
containing carbonyl-nitrogen bonds, such as ureas, ur-
ethanes, oxamides, formamides, and oxazolidinones. It also
represents a more environmentally friendly alternative to the
traditional and stoichiometric synthesis of ureas from
amines and toxic and/or corrosive phosgene or isocyanate
derivatives. A great number of mid- and late-transition-
metal complexes have been reported to catalyze the mono-
or dicarbonylation of amines.1 With the exception of tung-
sten,2 studies of early-transition-metal-catalyzed carbonyla-
tion of amines are unknown. One reason for that may
arise from the difficulty of insertion of CO into an
(3) For a rare CO insertion into a metal-amide bond in molybdenum
chemistry, see: Chisholm, M. H.; Hammond, C. E.; Huffman, J. C.
Organometallics 1987, 6, 210–211.
(4) For examples of CO insertion into late-metal-amide bonds, see:
(a) Holland, P. L.; Andersen, R. A.; Bergman, R. G. J. Am. Chem. Soc.
1996, 118, 1092–1104. (b) Belli Dell'Amico, D.; Calderazzo, F.; Pelizzi, G.
Inorg. Chem. 1979, 18, 1165–1168. (c) Bryndza, H. E.; Fultz, W. C.; Tam, W.
Organometallics 1985, 4, 939–940. (d) Beers, O. C. P.; Delis, J. G. P.; Mul,
W. P.; Vrieze, K.; Elsevier, C. J.; Smeets, W. J. J.; Spek, A. L. Inorg. Chem.
2002, 32, 3640–3647. (e) Cowan, R. L.; Trogler, W. C. Organometallics
2002, 6, 2451–2453.
*To whom correspondence should be addressed. E-mail: lorber@
lcc-toulouse.fr. Tel: (þ33) 5 61 33 31 44. Fax: (þ33) 5 61 55 30 03.
(1) For a recent review on transition-metal-catalyzed oxidative car-
bonylation of amines, see: Diaz, D. J.; Darko, A. K.; McElwee-White, L.
Eur. J. Org. Chem. 2007, 4453–4465and references cited therein.
(2) For tungsten-catalyzed carbonylations, see: (a) Jetz, W.; Angelici,
R. J. J. Am. Chem. Soc. 1972, 94, 3799–3802. (b) Doxsee, K. M.; Grubbs,
R. H. J. Am. Chem. Soc. 1981, 103, 7696–7698. (c) McCusker, J. D.;
Abboud, K. A.; McElwee-White, L. Organometallics 1997, 16, 3863–3866.
(d) McCusker, J. D.; Logan, J.; McElwee-White, L. Organometallics 1998,
17, 4037–4041. (e) McCusker, J. D.; Main, A. D.; Johnson, K. S.; Grasso,
C. A.; McElwee-White, L. J. Org. Chem. 2000, 65, 5216–5222. (f)
McCusker, J. D.; Grasso, C. A.; Main, A. D.; McElwee-White, L. Org. Lett.
1999, 1, 961–964.
(5) Boisson, C.; Berthet, J. C.; Lance, M.; Nierlich, M.; Ephritikhine,
M. J. Organomet. Chem. 1997, 548, 9–16.
(6) Fagan, P. J.; Manriquez, J. M.; Vollmer, S. H.; Day, C. S.; Day, V.
W.; Marks, T. J. J . Am. Chem. Soc. 1981, 103, 2206–2220.
(7) Rautenstrauch, V.; Joyeux, M. Angew. Chem., Int. Ed. 1979, 18,
85–86.
(8) Rautenstrauch, V.; Joyeux, M. Angew. Chem., Int. Ed. 1979, 18,
83–85.
r
2010 American Chemical Society
Published on Web 02/02/2010
pubs.acs.org/Organometallics

1128 Organometallics, Vol. 29, No. 5, 2010
Lorber and Vendier
Scheme 1. Attempted Synthesis of 2 Starting from 1 and
Mo(CO)6 or Mo(CO)3(CH3CN)3
Scheme 2. Two-Step Synthesis of 2 Starting via Formation of 3
Figure 1. Molecular structure of 3, showing 50% probability
ellipsoids and partial atom-labeling schemes. Hydrogen atoms
are omitted for clarity (except for NH2).
˚
Table 1. Comparison of Average Interatomic Distances (A) and
Angles (deg) in Complexes 1-3
1
2
3
Ti-Nimido
Cipso-N
Ti-Cl
1.7108(8)
1.3838(11)
1.7091(16)
1.380(2)
and structure of Ti/Mo heterobimetallic systems that led us
to characterize the first “Mo/η6-arene-imido/Ti” complex as
well as “Mo/aminocarbene/Ti” complexes.
2.3169(3) (Cl1) 2.3014(6) (Cl1)
2.3532(3) (Cl2) 2.3359(7) (Cl2)
2.2053(9) (N2) 2.1967(16) (N2)
2.2235(9) (N3) 2.2130(17) (N3)
2.4528(18)
Ti-NNHMe
2
Results and Discussion
Mo-Cipso
Mo-Cortho
2.518(3)
2.4142(19) (C4) 2.429(3) (C4)
2.4310(19) (C8) 2.415(3) (C6)
2.375(2) (C5)
2.352(2) (C7)
2.349(2)
The synthesis and proposed structures of the new com-
plexes are summarized in Schemes 1-6. Structures of six
complexes are set out in Figures 1-6 or are given in the
Supporting Information. Selected metric data are collected
in Tables 1 and 2.
1. Arene Complexes. The imide functionality is ubiquitous
in transition-metal chemistry9,10 and has found practical use
in multiple areas of research.11-15 In the past few years, our
group has developed new synthetic routes to early-transi-
tion-metal titanium-, zirconium-, and vanadium-imido
complexes, including those bearing functional, chiral, and
electron-deficient imido groups.16-23 As an extension of this
imido chemistry, we first examined the possibility of forming
Mo-Cmeta
Mo-Cpara
Mo-Ccentroid
Mo-CO
2.375(3) (C7)
2.354(3) (C9)
2.373(3)
1.950
1.931
1.950(2) (C1)
1.952(3) (C2)
1.953(2) (C3)
1.159(3) (C1)
1.160(3) (C2)
1.153(3) (C3)
175.39(14)
133.64(3)
165.26(6)
177.8(3) (C1)
179.5(3) (C2)
178.2(2) (C3)
1.935(3) (C1)
1.944(3) (C2)
1.946(4) (C3)
1.169(4) (C1)
1.164(4) (C2)
1.155(4) (C3)
C-O
Ti-Nimido-Cipso
Cl-Ti-Cl
179.49(7)
132.745(13)
NNHMe -Ti-NNHMe 166.34(3)
2
2
Mo-C-O
178.3(3) (C1)
177.7(3) (C2)
179.0(3) (C3)
(9) Nugent, W. A.; Haymore, B. L. Coord. Chem. Rev. 1980, 31, 123–
175.
(10) Nugent, W. A.; Mayer, J. M. In Metal-Ligand Multiple Bonds;
Wiley-Interscience: New York, 1998.
(11) Wigley, D. E. In Progress in Inorganic Chemistry; Karlin, K. D.,
Ed.; Interscience: New York, 1994; Vol. 42, pp 239-482.
(12) Mountford, P. Chem. Commun. 1997, 2127–2134.
(13) Duncan, A. P.; Bergman, R. G. Chem. Rec. 2002, 2, 431–445.
(14) Hazari, N.; Mountford, P. Acc. Chem. Res. 2005, 38, 839–849.
(15) Odom, A. L. Dalton Trans. 2005, 225–233.
(16) Lorber, C.; Donnadieu, B.; Choukroun, R. Dalton Trans. 2000,
4497–4498.
(17) Lorber, C.; Choukroun, R.; Donnadieu, B. Inorg. Chem. 2002,
41, 4217–4226.
(18) Lorber, C.; Choukroun, R.; Donnadieu, B. Inorg. Chem. 2003,
42, 673–675.
(19) Lorber, C.; Choukroun, R.; Vendier, L. Organometallics 2004,
23, 1845–1850.
(20) Lorber, C.; Choukroun, R.; Vendier, L. Eur. J. Inorg. Chem.
bimetallic complexes via an η6-arene-type coordination
of a metal carbonyl complex to the aryl substituent of an
arylimido-titanium synthon.
Our first attempts consisted of conducting the direct
reaction of [Ti(dNAr)Cl2(NHMe2)2] (1; Ar = 2,6-Pri2-
C6H3) with Mo(CO)6 or Mo(CO)3(CH3CN)3, as illustrated
in Scheme 1. However, this route met with limited success:
despite many attempts and varying the experimental condi-
tions (solvent, time, temperature), the reactions were slug-
gish and the conversions into the desired η6-arene-imido
complex Ti [(dN-η6-Ar)Mo(CO)3]Cl2(NHMe2)2 (2) were
always very low (<10%). One reason for that might be the
instability of 2 in solution under the thermal conditions
(refluxing octane) required to generate the η6-arene coordi-
nation to the Mo center (as shown later). Moreover, assisting
CO ligand release via UV irradiation of an octane solution of
1 and Mo(CO)6 did not lead to improved formation of the
expected 2. The extreme difficulty in synthesizing this com-
2006, 4503–4518.
(21) Lorber, C.; Choukroun, R.; Vendier, L. Inorg. Chem. 2007, 46,
3192–3202.
(22) Lorber, C.; Vendier, L. Organometallics 2008, 27, 2774–2783.
(23) Lorber, C.; Vendier, L. Dalton Trans. 2009, 6972–6984.

Article
Organometallics, Vol. 29, No. 5, 2010 1129
pound may also simply result from the presence of a strong
π-accepting substituent (the imido function, vide infra) that
is known to render the formation of M(CO)3(arene) more
difficult.24,25
away from the metal center is a common feature for arenes
having strong π-donor substituents, whereas π-acceptors are
found in the plane or present a small bend toward the metal
center.31-33
We then decided to use a different synthetic route, de-
picted in Scheme 2. In a first step, we prepared the molybde-
num η6-aniline tricarbonyl complex [Mo(CO)3(η6-ArNH2)]
(3) by the thermal replacement of three carbonyls in Mo-
(CO)6 by ArNH2.26 In order to achieve higher conversions
and to limit the purification required, the reaction was
conducted in refluxing octane under a stream of argon,27 in
the presence of a small amount of THF; it was also necessary
to use a slight excess of the aniline (1.2-1.4 equiv/Mo(CO)6).
Furthermore, the reaction and workup procedure were con-
ducted in the dark, as we noticed that the aniline complex
was light sensitive while in solution. Under these conditions,
3 was obtained in 52% yield as pale yellow crystals.28
Spectroscopic studies (IR, NMR), elemental analyses, and
X-ray studies confirmed these pale yellow crystals to be
complex 3. The IR spectrum displays the expected pairs of
strong carbonyl stretching frequencies of the Mo(CO)3 unit
In a second step, 3 was treated with an equimolar amount
of Ti(NMe2)4 in the presence of an excess of Me3SiCl
(6 equiv) (see Scheme 2). We have previously established this
one-pot synthesis as a very convenient and general way to
prepare a number of M(dNR)Cl2(NHMe2)2 complexes with
various terminal imido [NR]2- groups (M = Ti,20 V17). Never-
theless, in the present study, when using the Ti(NMe2)4/
Me3SiCl route, the expected η6-arene-imido complex 2 was
repeatedly obtained contaminated with ca. 10% of unreacted
molybdenum precursor and ca. 10% of Ti(dNAr)Cl2-
(NHMe2)2. Apparently in this reaction, Me3SiCl has a detri-
mental effect on the η6 coordination of the arene to the Mo
center by inducing decoordination of the arene moiety from
the Mo center. Indeed, using a larger amount of Me3SiCl
(8 equiv) induced a higher degree of decoordination (2:1
ratio ∼2).34
To prevent decoordination of the Mo(CO)3 moiety in-
duced by the presence of the chlorosilane during the forma-
tion of the arene-imido function, we turned instead to
TiCl2(NMe2)2 as a starting material. The transamination
reaction of TiCl2(NMe2)2 with complex 3 proceeded
smoothly at room temperature and afforded compound 2
in almost quantitative yields (92%) as red-orange crystals
(Scheme 2).
1
(1930 and 1835 cm-1), while the H and 13C NMR data
reveal that the diastereotopic methyl signals of the isopropyl
substituents and the aromatic protons of the η6-aniline are
shielded upfield. In the 13C NMR, the measure of the
difference between the chemical shifts of the para and
the meta aromatic carbon atoms is in agreement with the
π-donor character of the aniline ligand.24
The pale yellow crystals of 3 were suitable for an X-ray
structure determination. A thermal ellipsoid plot of the
molecular structure of 3 is presented in Figure 1, and Table 1
gives a comparison of its structural parameters. The overall
geometry around the molybdenum center is identical
with that of previously structurally characterized group
6 M(CO)3(η6-arene) complexes. The complex exhibits the
expected three-legged piano-stool structure having η6-
bonded arene rings. The relative (arene)Mo(CO)3 conforma-
tion is syn-eclipsed (torsion angle N1C2Mo1C5 6.1(2)°), in
agreement with theoretical calculations for donor-substi-
tuted arene.29,30 The bond distances are in the expected
The IR spectrum of 2 exhibits two strong absorption
bands in the carbonyl region at 1947 and 1841 cm-1 char-
acteristic of coordinated CO in fac-M(CO)3L3 complexes, as
well as a band at 3239 cm-1 for the NHMe2 ligands (νNH).
The increase of CO frequency in 2 vs 3 (with νCO 1930 and
1835 cm-1 in 3) reflects the lower electron-donating cap-
ability of the arene-imido fragment. As already seen in
precursor compound 3, the coordination of the Mo center
to the arene ring of the aryl-imido function is clearly visible in
1
the H and 13C NMR data of 2. This coordination renders
the two methyl groups inequivalent (-CHMeaMeb), as they
are in very different chemical environments. The aromatic
protons are also shifted considerably upfield and consist of a
doublet at 4.93 ppm and a triplet at 4.71 ppm (the two signals
are found at 6.99 and 6.85 ppm in the parent Ti(NAr)Cl2-
(NHMe2)2 compound).
X-ray-quality crystals were obtained for 2 and were stu-
died by diffraction analysis. As it was interesting to compare
directly the solid-state structure of 2 with that of its parent
complex 1, we have also determined the structure of 1 by
X-ray crystallography. Thermal ellipsoid plots or 1 and 2
are presented in Figures 2 and 3 along with selected bond
lengths and angles, and Table 1 gives a comparison of their
structural parameters. The two studies showed 2 to have a
structural arrangement around the titanium center analo-
˚
˚
range (Mo-CO(av) = 1.942(4) A, Mo-Ccentroid = 1.950 A,
˚
C-O(av) = 1.163(4) A).
The -NH2 in 3 is bent away from the Mo(CO)3 fragment
(see other views in the Supporting Information), which
induces a distortion of the arene planarity with an angle
between the least-squares plane defined by the ipso and ortho
carbon atoms of the aniline, and the least-squares plane
defined by the ortho, meta, and para carbon atoms of the
aniline is 9.4° away from Mo(CO)3. Such distortion
(24) Hunter, A. D.; Mozol, V.; Tsai, S. D. Organometallics 1992, 11,
2251–2262.
(25) Tamm, M.; Baker, R. J., Molybdenum compounds with CO or
isocyanides. In Comprehensive Organometallic Chemistry III; Crabtree,
R. H. M., Ed.; Elsevier: Oxford, U.K., 2007; Vol. 5, pp 391-512.
(26) Mo(CO)3(CH3CN)3 worked well too for this reaction but was
somewhat more difficult to separate from complex 3 during the puri-
fication step.
(31) Hunter, A. D.; Shilliday, L.; Furey, W. S.; Zaworotko, M. J.
Organometallics 1992, 11, 1550–1560.
(32) Djukic, J.-P.; Rose-Munch, F.; Rose, J.; Vaisserman, J. Eur. J.
(27) It is important to conduct the reaction under a stream of argon to
remove the liberated carbon monoxide (in a closed vial the yields
remained low).
Inorg. Chem. 2000, 1295–1306.
(33) Meyer, R.; Schindehutte, M.; van Rooyen, P. H.; Lotz, S. Inorg.
Chem. 1994, 33, 3605–3608.
(34) In theory, only 2 equiv of Me3SiCl is needed for this reaction.
However, the synthesis of M(NR)Cl2(NHMe2)2 complexes (M = Ti, V)
requires the use of at least 3-4 equiv of Me3SiCl in order to achieve
complete transformation, and in practice a larger excess (8-10 equiv) is
often used to speed up the reactions.17,20 In the present study, we did not
try to decrease the Me3SiCl concentration below 6 equiv.
(28) For a related Cr(CO)3(η6-aniline) complex, see: Mahaffy, C. A. L.
J. Organomet. Chem. 1984, 262, 33–37.
(29) Albright, T. A.; Hofmann, P.; Hoffmann, R. J. Am. Chem. Soc.
1977, 99, 7546–7557.
(30) Chinn, J. W.; Hall, M. B. J. Am. Chem. Soc. 1983, 105, 4930–
4941.

1130 Organometallics, Vol. 29, No. 5, 2010
Lorber and Vendier
related Ti and V compounds.17,20,36 Furthermore, this may
explain the unexpected longer bonds observed for Ti-Cl2
(vs Ti-Cl1) and shorter Ti-N2 (vs Ti-N3). The Mo(CO)3
group in 2 lies directly below the center of the arene ring plane
and shows the well-known piano-stool conformation found in
half-sandwich tricarbonyl complexes, with comparable bond
˚
distances and angles (Mo-CO(av) = 1.952(3) A, Mo-
˚
˚
Ccentroid = 1.931 A, C-O(av) = 1.157(4) A, CCO-Mo-CCO
-
(av) = 178.3(3)°). In contrast to the case for 3 (that has a syn-
eclipsed conformation for the relative (arene)Mo(CO)3 frag-
ment due to the donor -NH2 on the arene), the relative (arene-
imido)Mo(CO)3 conformation in 2 is now staggered with the
torsion angle N1C9Mo1C1 = 22.1(2)°, again reflecting
the π-accepting properties of the imido linkage (although we
cannot exclude that it could be due to crystal packing).
Consistently, the imido group in 2 is almost in the plane of
the arene ring with an angle between the least-squares plane
defined by the ipso and ortho carbon atoms of the aniline and
the least-squares plane defined by ortho, meta, and para
carbon atoms of the aniline only 1.5° away from Mo(CO)3
(see the Supporting Information for other Ortep views of 2).
In addition, compound 2 was shown not to be stable in
solution for extended periods of time. For example, in
benzene-d6 solutions at 80 °C, we noticed after 4 h only
20% of the initial product (2) remained, and considerable
decoordination of the Mo(CO)3 unit with formation of 80%
of 1. In contrast, in octane solutions, we observed only 15%
of the decoordination after 12 h at 80 °C.37 Thermal dis-
placement of the arene ligand in (η6-arene)metal species has
been reported.38-40 This observation will be crucial for
future comparative activity studies of 2 vs 1 in selected
transformations.
Figure 2. Molecular structure of 1, showing 50% probability
ellipsoids and partial atom-labeling schemes. Hydrogen atoms
are omitted for clarity (except for NHMe2).
2. Fischer-Type Aminocarbene Complexes. In the course of
our investigations on the preparation of the η6-arene-imido
complex 2 described above, and under certain experimental
conditions that quite obviously did not favor the formation of
the η6-arene complex (vide infra), we noticed the formation
of complex mixtures of products. Intrigued by this observation,
we decided to study in more detail the reactions between
the different components but now with the aim of avoiding
the formation of the η6-arene complex (room-temperature
reactions, closed vials to retain CO, no exclusion of light).
As a first example, the reaction between the imido-bridged
dimer complex {Ti(μ-NAr)(NMe2)2}2 and Mo(CO)6 or Mo-
(CO)3(CH3CN)3 afforded a complex mixture containing
several compounds (Scheme 3). Importantly, the reaction
proved to be time- and solvent-dependent (mixtures of
different products and compositions were obtained in
toluene and octane), and we also have indications that some
of the species involved may not be stable for long periods in
solution (see below). Therefore, this explains our extreme
difficulty in reproducing this synthesis and isolating pure
compounds.
Figure 3. Molecular structure of 2, showing 50% probability
ellipsoids and partial atom-labeling schemes. Hydrogen atoms
are omitted for clarity (except for NHMe2).
gous to that of the parent Mo-free precursor 1, i.e., between
a square pyramid and a trigonal bipyramid (τ = 0.56 (1), 0.53
(2)),35 containing an almost linear imido ligand in the apical
position, two trans chloride ligands, and two NHMe2 ligands.
The Ti-Nimido bond distance in the bimetallic complex 2
˚
(Ti-Nimido = 1.7091(16) A) is identical within error to that
˚
in the Mo-free complex 1 (Ti-Nimido = 1.7108(8) A).
However, all other bonds around the Ti center (Ti-Cl,
Ti-NHMe2) are shorter in 2(vs1), which reflects the decreased
donor capability of the aryl-imido function as a consequence of
the η6 coordination of the Mo center. In both imido com-
pounds 1 and 2, the supramolecular structure is dominated by
similar Me2N-H Cl hydrogen bonding that associate
3 3 3
pairs of molecules via intermolecular hydrogen bonds invol-
ving atoms N2, H2, and Cl2 (N-H Cl bond lengths of
However, from these mixtures, when conducting the
above reaction in octane, we have been very fortunate to
3 3 3
˚
˚
2.525 A in 1 and 2.485 A in 2; associated N-H Cl angles of
3 3 3
151.8° in 1 and 152.5° in 2), features that they share with other
(37) The fate of the decoordinated Mo(CO)3 fragment from 2 was not
determined, but we can theorize the concomitant formation of Mo(CO)6
or Mo(CO)3(C6D6).
(35) τ is the angular parameter commonly used to describe the
geometry around the metal center in pentacoordinate complexes and
is defined as τ = (R - β)/60 (R and β are the two largest L-M-L bond
angles, with R g β): Addison, A. W.; Rao, T. N.; Reedijk, J.; van Rijn,
J. V. J. Chem. Soc., Dalton Trans. 1984, 1349–1356.
(38) Davis, R.; Kane-Maguire, L. A. P. In Comprehensive Organo-
metallic Chemistry; Pergamon Press: New York, 1987; Vol. 3, Chapter 26.2,
pp 953-1076.
€
(39) Kundig, E. P.; Fabritius, C.-H.; Grossheimann, G.; Romanens,
(36) Adams, N.; Bigmore, H. R.; Blundell, T. L.; Boyd, C. L.;
Dubberley, S. R.; Sealey, A. J.; Cowley, A. R.; Skinner, M. E. G.;
Mountford, P. Inorg. Chem. 2005, 44, 2882–2894.
P. Organometallics 2004, 23, 3741–3744.
(40) Choukroun, R.; Lorber, C.; Vendier, L. Organometallics 2007,
26, 3604–3606.

Article
Organometallics, Vol. 29, No. 5, 2010 1131
Table 2. Comparison of Average Interatomic Distances (A) and Angles (deg) in Complexes 4-6
˚
4
5
6
Ti-Nimido
1.7343(14)
1.388(2)
1.8965(15)
1.7363(16)
1.385(2)
1.9050(17)
Cipso-Nimido
Ti-NNMe
1.8622(15) (N2), 1.8757(15)
(N3), 1.8792(15) (N4)
2
Ti-NNHMe
2.2302(15)
2.2844(15)
2.0020(12)
1.295(2)
2.2438(16) (N4), 2.2406(16) (N5)
2
Ti-NNMe CH
2
2
Ti-O
C-OTi
Mo-Ccarbene
Mo-CO
1.8975(12)
1.3125(19)
2.2655(16)
2.0477(19) (C4), 1.9947(19) (C5),
2.0270(19) (C6), 2.051(2) (C7),
2.0313(2) (C8)
2.0045(13)
1.288(2)
2.293(2)
2.040(3) (C4), 2.035(3) (C5),
2.018(3) (C6), 2.016(2) (C7),
1.984(3) (C26)
2.2616(18)
2.047(2) (C1), 2.046(2) (C2),
2.031(2) (C3), 2.040(2) (C4),
2.001(2) (C5)
C-Ocarbonyl
1.131(2) (C1), 1.135(2) (C2),
1.143(2) (C3), 1.140(3) (C4),
1.149(2) (C5)
1.134(2) (C4), 1.145(2) (C5),
1.138(2) (C6), 1.131(3) (C7),
1.139(2) (C8)
1.138(3) (C4), 1.144(3) (C5),
1.135(3) (C6), 1.147(3) (C7),
1.151(3) (C26)
C-NMe2/NMe
Ti-Nimido-Cipso
Mo-C-Ocarbonyl
1.345(2)
169.89(13)
177.03(18), 177.40(17),
174.56(19), 179.04(19), 178.64(17)
117.27(12)
129.20(12)
113.39(15)
141.74(11)
1.329(2)
1.344(3)
175.53(14)
177.12(19), 174.9(2),
175.7(3), 175.6(2), 176.9(2)
119.38(14)
127.09(14)
113.47(18)
164.88(13)
172.32(15), 178.10(18),
174.81(17), 178.49(19), 178.74(16)
117.79(11)
129.64(12)
112.32(15)
153.38(12)
Mo-Ccarbene-O
Mo-Ccarbene-N
N-C-O
Ti-O-C
be able to obtain a few crystals of the new compound 4. The
yellow crystals were suitable for an X-ray structure determi-
nation. Salient structural features for complex 4 are depicted
in Figure 4, along with selected bond lengths and angles in
Table 2 for a comparison of its structural parameters.
4 is a bimetallic titanium-molybdenum complex com-
posed of one Ti(dNAr)Cl2(NHMe2)2 and one Mo(CO)5
subunit linked by an unprecedented Fischer-type aminocar-
bene ligand. This ligand coordinates the Mo center via the
carbon atom of the carbene function, whereas the ligand
coordinates to the Ti center by both the oxygen atom coming
from the carbonyl and a dimethylamino group, the Ti atom
being part of a six-membered titanoxy ring.
center. This unusual coordination mode forms a six-membered
metallacycle ring with a distorted-boat conformation.
The formation of the Fischer aminocarbene function is
explained by the nucleophilic attack of one of the dimethyl-
amido group (on titanium) at the carbon atom of a coordi-
nated carbon monoxide ligand of Mo(CO)6 (path A in
Scheme 4). This would generate the titanoxy-aminocarbene
molybdenum pentacarbonyl intermediate species A with a Ti
center that functions as the electrophile.43 The next step
consists of a C-H bond activation at the R-position to an
amino group of the dimethylaminocarbene ligand (probably
via loss of a proton assisted by an amido ligand), and
intramolecular C,N coupling with a neighboring N atom of
a Ti-NMe2 ligand that transforms this group into a neutral
amino ligand. However, one cannot exclude a second path
(path B) in which the bidentate -N(Me)CH2NMe2 ligand is
formed first by the coupling of two dimethylamido ligands
on titanium followed by the nucleophilic attack of the new
amido ligand at the coordinated CO on molybdenum. Such a
coupling reaction, which might operate via the intermediacy
of azametallacyclopropane or η2-imine species, can be
viewed as a hydroaminoalkylation of an amido group. As
strong support for the involvement of -NMe2 ligands in the
formation of the fragment -N(Me)CH2NMe2, a similar
coupling reaction of two dimethylamido ligands, resulting
in the formation of a bidentate (dimethylamino)methylamide
ligand (-N(Me)CH2NMe2), was recently observed by Xue in
The coordination geometry of the pentacoordinated tita-
nium center is between square pyramidal and trigonal bipyr-
amidal (τ = 0.56), with the Ti atom surrounded by the
nitrogen atom of the arylimido group (Ti-Nimido
=
˚
1.7343(14) A, Ti-N-C = 169.89(3)°) in an apical position.
One nitrogen atom of the amido ligand -NMe2 (Ti-Namido
1.8665(15) A), the oxygen atom of the metalloxy amino-
=
˚
˚
carbene ligand (Ti-O = 2.0020(12) A), and the two respec-
tively trans-located nitrogen atoms of one NHMe2 ligand
˚
(Ti-N = 2.2302(15) A) and the dimethylamino group
˚
(Ti-N = 2.2844(15) A) of the Fischer carbene ligand form
the base of the pyramid. The molybdenum center has a
distorted-octahedral geometry, with five carbonyl ligands
with Mo-CO bond distances in the normal range
44,45
(Mo-CCO(av) = 2.033(3) A, C-O(av) = 1.140(3) A) and a
the reaction of TaX(NMe2)4 (X=NMe2,SiButPh2) withO2
˚
˚
and of MCl5 with LiNMe2.46
˚
carbon atom of the Fischer carbene at 2.2616(18) A. The
presence of a shorter than normal value for the C-O bond
The NMR studies for 4 were conducted on the same
crystalline sample as that used for the X-ray study (see the
˚
distance of the titanoxy-carbene ligand (C-OTi = 1.295(2) A)
is indicative of some π-delocalization through the titanoxy-
carbene bridging ligand. However, the Ti-O-C bond angle
(141.74(11)°) is not as large as in other alkoxycarbene41 or
aminocarbene42 compounds (also see below a comparison with
the examples of 5 and 6) in order to accommodate the
coordination of the dimethylamino nitrogen atom to the Ti
(43) For reviews on metalloxy Fischer carbene complexes, see: (a)
Erker, G. Angew. Chem., Int. Ed. Engl. 1989, 28, 397–412. (b) Barluenga,
~
J.; Fananas, F. J. Tetrahedron 2000, 56, 4597–4628.
(44) Qiu, H.; Chen, S.-J.; Wang, C.-S.; Wu, Y.-D.; Guzei, I. A.; Chen,
X.-T.; Xue, Z.-L. Inorg. Chem. 2009, 48, 3073–3079.
(45) Chen, S.-J.; Zhang, X.-H.; Yu, X.; Qiu, H.; Yap, G. P. A.; Guzei,
I. A.; Lin, Z.; Wu, Y.-D.; Xue, Z.-L. J. Am. Chem. Soc. 2007, 129, 14408–
14421.
(46) Zhang, X.-H.; Chen, S.-J.; Cai, H.; Im, H.-J.; Chen, T.; Yu, X.;
Chen, X.; Lin, Z.; Wu, Y.-D.; Xue, Z.-L. Organometallics 2008, 27,
1338–1341.
(41) Barger, P. T.; Santarsiero, B. D.; Armantrout, J.; Bercaw, J. E.
J. Am. Chem. Soc. 1984, 106, 5178–5186.
(42) Galakhov, M.; Martin, A.; Mena, M.; Palacios, F.; Yelamos, C.;
Raithby, P. R. Organometallics 2002, 14, 131–136.

1132 Organometallics, Vol. 29, No. 5, 2010
Scheme 3. Synthesis Leading to the Unexpected Compound 4
Lorber and Vendier
Scheme 4. Proposed Mechanism for the Formation of 4
Although the synthesis of 4 is fortuitous and very difficult
to reproduce, the observation of an unusual reactivity lead-
ing to the formation of compound 4, which also presents very
interesting features in relation to hydroaminoalkylation and
also potentially to aminocarbonylation, was a stimulating
factor that led us to investigate further the reaction between
similar components.
In the first instance, we came back to former studies,
initially reported by Bradley,47,48 on reactions between
M(NMe2)4 (M = group 4 metals) or CpTi(NMe2)3 and
M0(CO)n (M0 = Fe, Mo, W). Apparently, at that time, the
authors did not see the formation of the aminocarbene
complexes,49 but instead they reported adduct type com-
pounds with bridging dimethylamido ligands. Later, Petz
gave the correct interpretation of the formulation of the
species that are formed in the reaction of Ti(NMe2)4 and
metal carbonyls and characterized the aminocarbene func-
tion by IR and 1H NMR.50,51 To complete these studies, in
1995 Mena et al. reported an X-ray structure of the hetero-
bimetallic compound formed between Cp*Ti(NMe2)3 and
W(CO)6.42
In order to characterize without ambiguity the species that
might form in the reaction, we reproduced the reaction
between equimolar amounts of Ti(NMe2)4 and Mo(CO)6.
As shown in Scheme 5, the reaction proceeded smoothly in
toluene at room temperature to afford an orange solution
from which, after suitable workup, yellow-orange crystals of
5 were obtained in 79% yield. It is worth noting that, in the
solid state, 5 slowly decomposed into Mo(CO)6 (that sub-
limed as colorless crystals) and unknown Ti species. Mena et
Figure 4. Molecular structure of 4, showing 50% probability
ellipsoids and partial atom-labeling schemes. Hydrogen atoms
are omitted for clarity (except for NHMe2).
Supporting Information). However, the assignment was very
difficult due to the presence of two slightly different com-
pounds (complex 4 and a second compound that might be
[(CO)5Mo{dC(NMeCH2NMe2)O}Ti(dN-2,6-Pri2-C6H3)-
(NHAr)(NHMe2)] with a -NHAr amido group) that are
probably both present as two conformers (chair or boat
conformation of the metallacyclic bridging ligand) and
further complicated by dynamic equilibrium phenomena.
In any case, the data for the bridging aminocarbene fragment
compares particularly well with Xue’s data for the related
(dimethylamino)methylamide ligand on tantalum.44 The
two hydrogen atoms in the -CH2- group are diasterotopic,
giving rise to two broad signals at room temperature
(4.52 and 2.21 ppm) in the 1H NMR spectrum and one signal
at 71.5 ppm in the 13C NMR spectrum. The carbene function
is observed at 243.5 ppm in the 13C NMR spectrum, and
carbonyl ligands appear at 213.6 and 208.6 ppm.
(47) Bradley, D. C.; Charalamboud, J.; Jain, S. Chem. Ind. 1965,
1730–1730.
(48) Bradley, D. C.; Kasenally, A. S. Chem. Comm. 1968, 1430–1430.
(49) For a review on carbamoyl (aminocarbene) complexes, see:
Angelici, R. J. Acc. Chem. Res. 1972, 5, 335–341.
(50) Petz, W. J. Organomet. Chem. 1974, 72, 369–375.
(51) Petz, W. J. Organomet. Chem. 1993, 456, 85–88.

Article
Organometallics, Vol. 29, No. 5, 2010 1133
Scheme 5. Synthesis of 5
al. have also observed some decomposition in their related
Ti/W compounds.42
The spectral data (IR, NMR) and elemental analysis are in
agreement with the formulation [(CO)5Mo(dC(NMe2)O)-
Ti(NMe2)3] for 5, as suggested by Petz. In particular, the IR
spectrum shows three bands in the carbonyl region that are
characteristic of M(CO)5L complexes with pseudo-C4v sym-
metry.52 The 1H NMR presents a signal for three equivalent
-NMe2 groups at 2.96 ppm, and two signals for the two
diastereotopic methyl groups of the aminocarbene NMe2
group due to restricted rotation around the OC-NMe2 bond
(arising from a significant CN double-bond character typical
of aminocarbenes). The 13C NMR spectrum presents char-
acteristic signals of ModC at 238.8 ppm, as well as signals
for CO being located cis (207.9 ppm) and trans (213.2 ppm)
to the carbene ligand.
Figure 5. Molecular structure of 5, showing 50% probability
ellipsoids and partial atom-labeling schemes. Hydrogen atoms
are omitted for clarity.
In contrast, when the same reaction of 5 with 1 equiv of
ArNH2 was performed in cyclohexane-d12 and followed by
1H NMR, we observed a slow but complete transformation
of 5 in about 3 days (Scheme 6), leading to the imido complex
6, together with a second unidentified species X. In order to
isolate complex 6, we reproduced this experiment in a larger
scale and in pentane, but unfortunately we could not obtain 6
in a pure form due to similar solubility.55
This formulation was unambiguously confirmed by a
X-ray diffraction study. Thermal ellipsoid plots are pre-
sented in Figure 5 along with selected bond lengths and
angles, and Table 2 gives a comparison of their structural
parameters. As anticipated, 5 is a monomer in the solid state;
the bulky Mo(CO)5 unit apparently prevents the dimeriza-
tion of Ti(NMe2)3. The Ti center has a tetrahedral geometry
and is coordinated by three dimethylamido (Ti-N(av) =
We then turned our attention to the one-pot reaction
between equimolar amounts of Ti(NMe2)4 and Mo(CO)6,
now in the presence of 1 equiv of ArNH2 (Scheme 6). It was
difficult to obtain reproducible results from this reaction
(depending on the reaction time and the workup procedure),
˚
1.872(2) A) and an oxygen atom of the titanoxy aminocar-
˚
bene ligand (Ti-O = 1.8975(12) A). This Ti-O bond is
significantly shorter than in 4 (and 6, vide infra). The six-
coordinated Mo center is bonded to five CO (C-O(av) =
˚
˚
1.137(2) A, Mo-C(av) = 2.030(2) A) and a carbenic carbon
1
which prompted us to followed the reaction by H NMR.
˚
atom (Mo-C = 2.2655(16) A). As previously noted in 4,
the short C-O bond distance (1.3125(19) A) and the
The 1H NMR studies showed the formation of a single
product within the first hours, which after 6-12 h further
evolved to afford a complex mixture of unidentified pro-
ducts. Consequently, in order to get the compound in a pure
form, we reproduced the reaction in a larger scale and
stopped the reaction after 6 h of stirring by removing the
solvent or precipitation with pentane (as for the preparation
of 4). In this manner, we could produce the new complex 6 in
pure form, albeit in low yields (20%). Again, we attribute this
problem to instability in solution, as we noted that allowing a
benzene-d6 solution of 6 to stand at room temperature
showed slow decomposition (after 1 day) with almost com-
plete disappearance of the initial signals within 1 week and
formation of unidentified species. The decomposition pro-
duct of 6 seems identical with that obtained through the
reaction of 5 with ArNH2 when a C6D6 solution was left over
2 weeks (vide supra). Although uncharacterized, the
product(s) of decomposition is (are) apparently not complex
4, and further studies will be necessary to clarify its (their)
exact nature.56
˚
wide Ti-O-C bond angle (153.38(12)°) are indicative of
π-delocalization.
Next we examined the reaction of 5 with ArNH2 as a
possible route to 4 (or to 6). Such transamination reactions
between a primary amine and a group 4 and 5 metal
precursor with dialkylamido ligands is a well-known route
to imido complexes.14,17,19,21,22,36,53 However, the treatment
of a C6D6 solution of 5 with 1 equiv of ArNH2 proved to be
more difficult. In general, generation of an imido function by
transamination with an aniline is over within 1-2 h, whereas
in the case of 5 we observed a very slow reaction with
formation after 6 h of less than 5% of an intermediate
complex (namely 6). As this complex 6 is unstable in solution
(see later), we were unable to obtain this compound by this
route, and instead we noted after 2 weeks of reaction the total
conversion of 5 into a new compound, which separates from
the C6D6 solution as a brown insoluble oil. When separated
from the C6D6 supernatant solution, this oil eventually
solidified and was shown to contain Mo(CO)6,54 which
suggests that the decomposition observed in the C6D6 reac-
tion involves the loss of the Mo(CO)5 fragment with forma-
Compound 6 reveals a diagnostic carbene resonance in the
1
13C NMR spectrum centered at 233.9 ppm. The H NMR
tion of Mo(CO)6 and
compound.
a titanium-only unidentified
(55) Species X might be a coordination compound of the type
(ArNH2)Mo(CO)4(L) with a CO ligand that has been substituted by
an ArNH2 molecule. We believe the decomposition of our Ti/Mo
compounds might be inherent to the presence of external donors such
as the aniline that can further displace carbonyl ligands.
(56) Isolated compound 6 could be regarded as intermediate A in
Scheme 4 for formation of 4; however, attempted direct conversion of 6
into 4 has proved so far unsuccessful.
(52) Cotton, F. A.; Darensbourg, D. J.; Fang, A.; Kolthammer, B. W.
S.; Reed, D.; Thompson, J. L. Inorg. Chem. 1981, 20, 4090–4096.
(53) Thorn, D. L.; Nugent, W. A.; Harlow, R. L. J. Am. Chem. Soc.
1981, 103, 357–363.
(54) The crystals were confirmed to be Mo(CO)6 by determinatiion of
the cell parameters by an X-ray diffraction analysis.

1134 Organometallics, Vol. 29, No. 5, 2010
Lorber and Vendier
geometry, with five carbonyl ligands with Mo-CO bond dis-
˚
tances in the normal range (Mo-CCO(av) = 2.019(4) A,
˚
C-O(av) = 1.143(4) A) and a carbon atom of the Fischer
carbene at 2.293(2) A. Again, the presence of a shorter than
˚
normal value for the C-O bond distance of the titanoxycarbene
˚
ligand (C-OTi = 1.288(2) A) is indicative of some π-delocaliza-
tion through the titanoxycarbene bridging ligand. The
Ti-O-C bond angle (164.88(13)°) is larger than in related
aminocarbene compounds 4 and 5.
Conclusions
We have prepared the first bimetallic (η6-arene)moly-
bdenum imido-titanium complex 2 from a transamination
reaction between TiCl2(NMe2)2 and the corresponding
(η6-aniline)Mo(CO)3 precursor 3. When it was treated with
Mo(CO)6, we reported that the dimer {Ti(μ-NAr)(NMe2)2}2
afforded a complex mixture of products, from which complex
4 was characterized. Complex 4 was shown to contain an
unprecedented aminocarbene ligand doubly bridging the Ti
and the Mo centers, generated in part from the intramolecular
CNcoupling of a Ti-NMe2 group witha ModCNMe2 ligand.
We have also reported other examples of bimetallic Ti/Mo
complexes held together by a bridging dimethylaminocarbene
function (namely compounds 5 and 6).
Figure 6. Molecular structure of 6, showing 50% probability
ellipsoids and partial atom-labeling schemes. Hydrogen atoms
are omitted for clarity (except for NHMe2).
Scheme 6. Synthesis of 6
As most of these species suffer from instability in solution
and/or in the solid state, which may hamper their further use
in catalytic processes, we are also examining other bimetallic
systems composed of other pairs of complexes (group 4/5
metals and various metal carbonyl complexes). These studies
are currently underway and will be reported in due course, as
well as the reactivity studies toward carbonylation of amines,
as some of the reported compounds can be viewed as the first
step toward amine carbonylation.
Experimental Section
General Methods and Instrumentation. All manipulations
were carried out using standard Schlenk line or drybox techni-
ques under an atmosphere of argon. Solvents were refluxed and
dried over appropriate drying agents under an atmosphere of
argon and collected by distillation. NMR spectra were recorded
on Bruker ARX250, DPX300, and Avance500 spectrometers
and referenced internally to residual protio-solvent (1H) reso-
nances and are reported relative to tetramethylsilane (δ 0 ppm).
Chemical shifts are quoted in δ (ppm). Infrared spectra were
prepared as KBr pellets under argon in a glovebox and were
recorded on a Perkin-Elmer Spectrum GX FT-IR spectrometer.
Infrared data are quoted in wavenumbers (cm-1). Elemental
analyses were performed at the Laboratoire de Chimie de
Coordination (Toulouse, France) or by the Service Central de
Microanalyses du CNRS at Vernaison (France).
The Ti(NMe2)4 used in this study was prepared by a modi-
fication of a literature procedure57 or purchased from commer-
cial sources (Aldrich). TiCl2(NMe2)2 was prepared from the
reaction of equimolar amounts of TiCl4 and Ti(NMe2)4.22 The
imido precursor Ti(dN-2,6-Pri2-C6H3)Cl2(NHMe2)2 (1) was
prepared according to our published procedure.20 The imido-
bridged dimer {Ti(N-2,6-Pri2-C6H3)(NMe2)2}2 was prepared
by a synthesis similar to that used for analogous vanadium
compounds.17,21 Mo(CO)3(CH3CN)3 was prepared according
to a literature procedure.58
data consists of sets of signals attributed to one arylimido
group, two equivalent NHMe2 ligands, and one -NMe2
group, as well as two diastereotopic methyl groups of an
aminocarbene ligand. Other spectroscopic features for 6
include the observation in the carbonyl region of the three
characteristic bands for Mo(CO)5L complexes at 1968, 1897,
and 1876 cm-1
.
Altogether, the spectroscopic data, complemented by a
combustion analysis, suggests compound 6 can be formulated
as [(CO)5ModC(NMe2)O]Ti(dNAr)(NMe2)(NHMe2)2 with
(different from the case for compound 4) a “simple” dimethyla-
minocarbene ligand bridging the two metals.
This formulation was consolidated by a X-ray diffraction
analysis. The ORTEP diagram is shown in Figure 6 with
selected bond lengths and angles provided in Table 2. The
solid-state structure confirms the proposed connectivity and
displays structural features resembling those of 4. Complex 6 is
a bimetallic system composed of a five-coordinate titanium-
(IV) center and an octahedral molybdenum center linked by a
dimethylaminocarbene function. The Ti-Nimido bond distance
˚
˚
of 1.7363(16) A is comparable to that in 4 (1.7343(14) A)],
as is the Ti-N-Cipso angle, which moderately deviates from
180°. The remaining structural features around the Ti atom are
Crystal Structure Determination. The structures of six com-
pounds were determined. Crystal data collection and processing
˚
comparable to those in 4, including the Ti-O (2.0045(13) A)
and O-C (1.288(2) A) bond distances. As in complexes 4
and 5, the molybdenum center in 6 has a distorted-octahedral
˚
(57) Bradley, D. C.; Thomas, I. M. J. Chem. Soc. 1960, 3857–3861.
(58) King, R. B. J. Organomet. Chem. 1967, 8, 139–148.

Article
Organometallics, Vol. 29, No. 5, 2010 1135
Table 3. Crystallographic Data and Data Collection and Refinement Parameters for 1-6
3a
1
2
4
5
6
chem formula
formula wt
cryst syst
C16H31Cl2N3Ti C19H31Cl2MoN3O3Ti C15H19MoNO3 C26H41MoN5O6Ti C14H24MoN4O6Ti C26H43MoN5O6Ti
488.21
384.24
monoclinic
P21/n
564.21
triclinic
P1
357.25
663.48
triclinic
P1
665.49
monoclinic
P21/n
orthorhombic
P212121
7.8652(11)
13.455(2)
14.318(2)
90.0
90.0
90.0
1515.2(4)
4
1.566
monoclinic
P21/n
13.7140(4)
11.1816(3)
14.2280(4)
90.0
99.902(2)
90.0
2149.28(10)
4
1.509
space group
˚
a, A
9.8809(6)
22.4338(12)
10.2598(6)
90.0
113.557(3)
90.0
2084.7(2)
4
1.224
0.667
816
10.2058(3)
11.2671(3)
12.2010(4)
92.430(2)
108.6380(10)
108.1860(10)
1247.16(6)
2
1.502
1.060
576
2.69-29.80
26446
10.1513(8)
13.5008(11)
15.0634(13)
100.794(4)
108.827(4)
103.517(4)
1820.8(3)
2
1.21
0.602
688
1.62-29.89
42720
10.6408(5)
25.6907(10)
13.1933(7)
90.0
113.001(6)
90.0
3319.9(3)
4
1.331
0.66
1384
˚
b, A
˚
c, A
R, deg
β, deg
γ, deg
3
˚
V, A
Z
Dcalcd, g cm-3
μ(Mo KR), mm-1
F(000)
0.872
728
2.08-29.38
17696
0.988
992
2.64-26.37
16122
θ range (deg)
2.35-40.12
66700
2.62-26.37
25658
no. of measd rflns
no. of unique rflns/Rint
no. of params/restraints
12 957/0.0351
207/0
7009/0.0337
270/0
R1 = 0.0314
4182/0.0474
191/2
R1 = 0.0348
wR2 = 0.0651 wR2 = 0.0577
R1 = 0.051 R1 = 0.0507
wR2 = 0.0705 wR2 = 0.0635
10 352/0.0338
363/0
R1 = 0.0346
4383/0.0242
243/0
6789/0.0352
363/0
final R indices (I > σ2(I)) R1 = 0.0397
R1 = 0.0212
wR2 = 0.0862
R1 = 0.0255
wR2 = 0.0901
1.079
R1 = 0.0292
wR2 = 0.0555
R1 = 0.0479
wR2 = 0.0576
1.031
wR2 = 0.0658 wR2 = 0.0969
R1 = 0.0653 R1 = 0.0475
wR2 = 0.0698 wR2 = 0.11
1.013
0.652, -0.441
final R indices (all data)
goodness of fit
ΔFmax, ΔFmin
1.025
0.462, -0.483
1.027
0.466, -0.625
1.029
0.406, -0.347
0.368, -0.431
0.431, -0.500
a Absolute structure parameter (Flack): -0.02(4).
parameters are given in Table 3. Crystals of 1 (orange blocks), 2
(orange blocks), 3 (colorless parallelepipeds), 4 (yellow plates), 5
(yellow blocks), and 6 (yellow parallelepipeds) were obtained.
The selected crystals, sensitive to air and moisture, were
mounted on a glass fiber using perfluoropolyether oil and cooled
rapidly to 180 K under a stream of cold N2. For all the structures
data collection was performed at low temperature (T = 180 K)
on a Stoe Imaging Plate Diffraction System (IPDS), on an
Oxford Diffraction Kappa CCD Excalibur diffractometer, or
on a Bruker Kappa Apex II diffractometer, using graphite-
displacement parameters for all structures have been deposited
at the Cambridge Crystallographic Data Centre.
Synthesis of [Ti(dN-2,6-Pri2-C6H3)Cl2(NHMe2)2][η6-Mo-
(CO)3] (2). Method 1 (from TiCl2(NMe2)2 and Mo(CO)3(η6-
2,6-iPr2-C6H3NH2). To a toluene solution (2 mL) of 58 mg of
TiCl2(NMe2)2 (0.2803 mmol) was added by portions 1 equiv of
Mo(CO)3(η6-2,6-iPr2-C6H3NH2) (3; 100 mg, 0.2799 mmol) at
room temperature. The resulting red solution was stirred for 2 h
at room temperature. The volatiles were removed under va-
cuum, and the solid was washed with 2 mL of pentane. Yield:
145 mg (92%). 1H NMR (300 MHz, C6D6): δ 4.93 (d, 3J = 6.6
˚
monochromated Mo KR radiation (λ = 0.710 73 A) and
3
Hz, 2H, C6H3Pri2), 4.71 (t, J = 6.6 Hz, 1H, C6H3Pri2), 4.25
equipped with an Oxford Cryosystems Cryostream Cooler
Device. Final unit cell parameters were obtained by means of
a least-squares refinement of a set of 8000 well-measured reflec-
tions, and the crystal decay was monitored during data collec-
tion by measuring 200 reflections by image; no significant
fluctuation of intensities was observed. Structures were solved
by means of direct methods using the program SIR9259 and
subsequent difference Fourier maps, models were refined by
least-squares procedures on F2 by using SHELXL-9760 inte-
grated in the package WINGX version 1.64,61 and empirical
absorption corrections were applied on data.62 For 4, it was not
possible to solve diffuse electron-density residuals (enclosed
solvent molecules). Treatment with the SQUEEZE facility from
PLATON63 resulted in a smooth refinement. Since a few low-
order reflections are missing from the data set, the electron
count was underestimated. Thus, the values given for D(calc),
F(000), and the molecular weight are only valid for the ordered
part of the structure. Details of the structure solutions and
refinements are given in the Supporting Information. Full
listings of atomic coordinates, bond lengths and angles, and
3
(sept, J = 6.8 Hz, 2H, CHMe2), 3.05 (br sept, 2H, HNMe2),
3
3
2.26 (d, J = 6.6 Hz, 12H, NMe2), 1.30 (d, J = 6.6 Hz, 6H,
CHMeaMeb), 1.18 (d, 3J = 6.6 Hz, 6H, CHMeaMeb). 13C{1H}
NMR (75.47 MHz, C6D6): δ 224.3 (CO), 133.5 (ipso-C6H3),
122.8 (o-C6H3), 92.2 (p-C6H3), 91.1 (m-C6H3), 40.3 (NHMe2),
27.7 (CHMe2), 25.7 (CHMeaMeb), 22.2 (CHMeaMeb). IR: 3239
(w, NH), 2971 (m), 1947 (CO), 1841 (vs, CO), 1467 (m), 1396
(m), 1340 (m), 1019 (m), 895 (m), 622 (m), 501 (m), 442 (m).
Anal. Calcd for C19H31Cl2MoN3O3Ti (564.18): C, 40.45; H,
5.54; N, 7.45. Found: C, 40.35; H, 5.59; N, 7.45.
Method 2 (from Ti(NMe2)4/Me3SiCl and Mo(CO)3(η6-
2,6-iPr2-C6H3NH2)).
A
50 mg portion of Mo(CO)3(η6-
2,6-iPr2-C6H3NH2) (0.1400 mmol) was added to a toluene
(1 mL) solution of Ti(NMe2)4 (31.4 mg, 0.1401 mM), and the
resulting solution was stirred for 5 min at room temperature.
Me3SiCl (92 mg, 0.8468 mmol) was added dropwise to the
solution with stirring. After 3 h, the volatiles were pumped off
and the red-orange solid was washed with pentane (2 ꢀ 2 mL)
and dried under vacuum (67 mg). 1H NMR spectroscopy
showed this solid to be 2 contaminated by 10% of [Ti(dN-2,6-
Pri2-C6H3)Cl2(NHMe2)2].
(59) Altomare, A.; Cascarano, G.; Giacovazzo, G.; Guagliardi, A.;
Burla, M. C.; Polidori, G.; Camalli, M. J. Appl. Crystallogr. 1994, 27,
435.
(60) Sheldrick, G. M. SHELX97 [Includes SHELXS97, SHELXL97,
CIFTAB]-Programs for Crystal Structure Analysis (Release 97-2);
Synthesis of [Mo(CO)3(η6-2,6-iPr2-C6H3NH2)] (3). Mo(CO)6
(500 mg, 1.8939 mM), 2,6-iPr2-C6H3NH2 (470 mg, 2.6510 mM),
0.5 mL of THF, and 10 mL of octane were introduced into a
Schlenk flask equipped with a magnetic stir bar and a reflux
condenser connected to an oil bubbler and an argon inlet.
The glassware was covered with aluminum foil during the time
the compounds were in solution to prevent decomposition by
light. The mixture was heated to reflux under a stream of argon
for 7 h. Upon cooling to room temperature, pale yellow crystals
€
Instit€ut f€ur Anorganische Chemie der Universitat, Tammanstrasse 4,
€
D-3400 Gottingen, Germany, 1998.
(61) Farrugia, L. J. J. Appl. Crystallogr. 1999, 32, 837–838.
(62) Walker, N.; Stuart, D. Acta Crystallogr., Sect. A 1983, 39, 158–
166.
(63) Spek, A. L. Acta Crystallogr. 1990, A46, C–34.

1136 Organometallics, Vol. 29, No. 5, 2010
Lorber and Vendier
were formed. The crystals were collected by filtration (another
crop of crystals; about 30 mg can be obtained by cooling the
filtrate to -20 °C). Extraction with toluene (4 ꢀ 2 mL, to remove
small amounts of unreacted Mo(CO)6) afforded a yellow solu-
tion that was filtered under a bed of Celite, and the volatiles were
removed under vacuum to give 355 mg of pale yellow crystals of
3 (52%). 1H NMR (300 MHz, C6D6): δ 5.25 (d, 3J = 6.4 Hz, 2H,
C6H3Pri2), 4.43 (t, 3J = 6.5 Hz, 1H, C6H3Pri2), 2.96 (br s, 2H,
NH2), 1.95 (sept, 3J = 6.7 Hz, 2H, CHMe2), 1.01 (d, 3J = 6.6
Hz, 6H, CHMeaMeb), 0.71 (d, 3J = 6.6 Hz, 6H, CHMeaMeb).
13C{1H} NMR (75.47 MHz, C6D6): δ 224.4 (CO), 130.6 (ipso-
C6H3), 105.5 (o-C6H3), 95.6 (m-C6H3), 85.4 (p-C6H3), 27.9
(CHMe2), 22.5 (CHMeaMeb), 21.9 (CHMeaMeb). IR 3509 (w,
NH2), 3413 (w, NH2), 1930 (vs, CO), 1835 (vs, CO), 1629 (m),
1429 (m), 593 (m), 499 (m). Anal. Calcd for C15H19MoNO3
(357.26): C, 50.43; H, 5.36; N, 3.92. Found: C, 50.50; H, 5.38; N,
3.83.
18H, NMe2), 2.21 (s, 3H, dCNMeaMeb). 13C{1H} NMR
(75.47 MHz, C6D6): δ 238.8 (ModC), 213.2 (CO) 1 COtrans
,
207.9 (COcis), 44.2 (dCNMeaMeb), 43.8 (NMe2), 32.9
(dCNMeaMeb). IR: 2867 (m), 1975 (sh, CO), 1945 (vs, CO),
1904 (sh, CO), 1272 (m), 1229 (m), 942 (m), 729 (m), 591 (s).
Anal. Calcd for C14H24MoN4O6Ti (488.17): C, 34.44; H, 4.96;
N, 11.48. Found: C, 34.23; H, 5.00; N, 11.24.
Synthesis of [Ti(dN-2,6-Pri2-C6H3)(NMe2)(NHMe2)2(OC-
(NMe2)dMo(CO)5)] (6). Method 1 (from the Reaction between
5 and ArNH2). A cyclohexane-d12 solution (0.5 mL) of 50 mg of 5
(0.102 42 mmol) and 18 mg of ArNH2 (0.101 53 mmol) was
1
placed in an NMR tube, and the reaction was followed by H
NMR spectroscopy. The NMR studies showed the reaction to
be complete after 3 days and to lead to complex 6 and an
unidentified side product.
Method 2 (from Ti(NMe2)4, Mo(CO)6, and ArNH2). Ti-
(NMe2)4 (100 mg, 0.4460 mM) was added to a toluene (4 mL)
suspension of Mo(CO)6 (117 mg, 0.4432 mM) and 2,6-iPr2-
C6H3NH2 (79 mg, 0.4456 mM). The red solution was stirred for
6 h at room temperature. The solution was filtered, and the
solvent was removed under vacuum to afford a red oily residue.
Extraction of this oil with pentane (10 mL) afforded a red
solution that upon cooling to -20 °C produced 60 mg of red-
orange crystals of 6 (yield 20%) (note: the insoluble residue from
the pentane extraction was also confirmed to be complex 6, but
in a less pure form, and we did not try to further purify it to
Synthesis of [Ti(dN-2,6-Pri2-C6H3)(NMe2)(NHMe2)(OC-
(NMeCH2NMe2)dMo(CO)5)] (4). This compound was ob-
tained using a slight modification of the procedure used for
the synthesis of
6
(longer reaction time, solvent).
{Ti(NAr)(NMe2)2}2 was prepared in situ by addition of 79 mg
of ArNH2 (0.4456 mM) to a toluene solution (2 mL) of Ti-
(NMe2)4 (100 mg, 0.4460 mM) and left overnight. The solvent
was quickly removed under vacuum, and 10 mL of octane was
added followed by the addition of 117 mg of Mo(CO)6 (0.4432
mM). The solution was stirred for 2 h and left for 3 days without
stirring, during which time a few crystals of 4 were formed
(although glued in an oily material) from a complex mixture of
products. The crystals that were collected (about 40 mg) were of
a sufficient amount for X-ray diffraction and NMR studies.
Nevertheless, we were unable to fully assign the 1H and 13C
NMR spectra (given in the Supporting Information) due to the
presence of possible isomers or conformers of 4, contaminated
by another compound (see the text), further complicated by
dynamic exchange phenomena.
Synthesis of [(CO)5)Mo(dC(NMe2)O)Ti(NMe2)3] (5). To a
toluene solution (5 mL) of Ti(NMe2)4 (250 mg, 1.1151 mmol)
was added by portions 1 equiv of Mo(CO)6 (293 mg, 1.1100
mmol) at room temperature. The resulting orange suspension
was stirred overnight at room temperature, after which time the
volatiles were removed under vacuum to give an orange oil.
The oil was dissolved in the minimum amount of pentane
(ca. 5-7 mL), and the solution was filtered through a bed of
Celite and allowed to crystallize at -20 °C. Yellow crystals were
collected, and a second crop of crystals could be obtained from
cooling the concentrated filtrate. Yellow crystals were collected
and dried under vacuum. Yield: 433 mg (79%). This com-
pound slowly decomposed under argon in the solid state with
sublimation of Mo(CO)6 and uncharacterized Ti species. 1H
NMR (300 MHz, C6D6): δ 2.99 (s, 3H, dCNMeaMeb), 2.96 (s,
1
3
optimize yields). H NMR (300 MHz, C6D6): δ 7.03 (d, J =
7.5 Hz, 2H, C6H3Pri2), 6.88 (t, 3J = 7.5 Hz, 1H, C6H3Pri2), 4.19
(sept, J = 6.9 Hz, 2H, CHMe2), 3.28 (s, 6H, NMe2), 3.07 (s,
3
3H, dCNMeaMeb), 2.58 (s, 3H, dCNMeaMeb), 2.14 (d, J =
3
6.0 Hz, 12H, NHMe2), 1.23 (d, 3J = 6.9 Hz, 12H, CHMe2) (note
that (i) the peak at 2.14 ppm is a singlet at high concentration
and (ii) the signal for NHMe2 is too broad to be observed at
room temperature). 13C{1H} NMR (75.47 MHz, C6D6): δ 233.9
(ModC), 212.6 (COtrans), 208.7 (COcis), 155.5 (C6H3), 141.5
(C6H3), 122.8 (C6H3), 120.1 (C6H3), 47.1 (br, NMe2), 44.1
(dCNMeaMeb), 40.2 (NHMe2), 32.8 (dCNMeaMeb), 27.0
(CHMe2), 24.9 (CHMe2). IR 3296 (w, NHMe2), 2963 (m),
1968 (CO), 1897 (s, CO), 1876 (sh, CO), 1496 (m), 1420 (m),
1316 (m), 1277 (m), 983 (m), 950 (m), 896 (m), 755 (m), 613 (m),
594 (m). Anal. Calcd for C26H43MoN5O6Ti (665.46): C, 46.93;
H, 6.51; N, 10.52. Found: C, 46.67; H, 6.54; N, 10.30.
Acknowledgment. We are grateful to the CNRS for
financial support.
Supporting Information Available: Tables, figures, and CIF
files giving data for the X-ray crystal structures of complexes
1-6, other detailed Ortep views for 2 and 3, and NMR data (1H,
13C) for complex 4. This material is available free of charge via
the Internet at http://pubs.acs.org.