Oxygen-centered hexatantalum tetradecaimido cluster complexes

Article abstract of DOI:10.1021/ic701920v

The syntheses and characterization of several octahedral hexatantalum cluster compounds of formula (ArN)₁₄Ta₆O are described (Ar = Ph, p-MeC₆H₄, p-MeOC₆H₄, p-t-BuC6H₄, p-BrC₆H₄, m-ClC₆H ₄). Treatment of Bn₃Ta=N-t-Bu (Bn = CH₂C ₆H₅) or pentakis(dimethylamido)tantalum with an excess of the appropriate aniline and stoichiometric water or tantalum oxide afforded varying yields of arylimido clusters. The structures of two species were confirmed by X-ray diffraction (XRD), while the identity of the central oxygen atom was elucidated by electrospray mass spectrometry (MS) using ¹⁷O/¹⁸O-enriched material. The title species are very air- and moisture-sensitive but quite thermally stable in solution. Experimentally determined optical properties and oxidation/reduction potentials, as well as some computational results, indicate that they possess an electronic structure wherein the highest occupied molecular orbitale are ligand-centered, while the lowest unoccupied orbitals are metal-centered and delocalized throughout the tantalum cage. Whereas chemical oxidation resulted in cluster decomposition, reduction with decamethylcobaltocene yielded stable salts of formula [Cp*₂Co][(ArN)₁₄Ta₆O] (Ar = Ph, Ar = p-MeC₆H₄). Small-molecule reactivity studies on one of these clusters showed that its imido functionalities are moderately reactive toward oxide donors but inert with respect to metallaheterocycle-forming processes. Clean imido/oxo exchange was observed with aldehydes and ketones, leading cleanly to organic imines with no soluble byproducts being observed. This exchange was also observed with a rhenium oxo compound (generating an imidorhenium complex as the only soluble species). All 14 imido groups were transferred in these reactions, and no mixed-ligand cluster intermediates were ever observed.

Full text of DOI:10.1021/ic701920v

Inorg. Chem. 2008, 47, 1053−1066
Oxygen-Centered Hexatantalum Tetradecaimido Cluster Complexes
Jamin L. Krinsky, Laura L. Anderson,† John Arnold,* and Robert G. Bergman*
Department of Chemistry, UniVersity of California, Berkeley, California 94720
Received September 28, 2007
The syntheses and characterization of several octahedral hexatantalum cluster compounds of formula (ArN)14Ta6O
are described (Ar Ph, p-MeC6H4, p-MeOC6H4, p-t-BuC6H4, p-BrC6H4, m-ClC6H4). Treatment of Bn3Ta N-t-Bu
(Bn CH2C6H5) or pentakis(dimethylamido)tantalum with an excess of the appropriate aniline and stoichiometric
)
d
)
water or tantalum oxide afforded varying yields of arylimido clusters. The structures of two species were confirmed
by X-ray diffraction (XRD), while the identity of the central oxygen atom was elucidated by electrospray mass
spectrometry (MS) using 17O/18O-enriched material. The title species are very air- and moisture-sensitive but quite
thermally stable in solution. Experimentally determined optical properties and oxidation/reduction potentials, as well
as some computational results, indicate that they possess an electronic structure wherein the highest occupied
molecular orbitals are ligand-centered, while the lowest unoccupied orbitals are metal-centered and delocalized
throughout the tantalum cage. Whereas chemical oxidation resulted in cluster decomposition, reduction with
decamethylcobaltocene yielded stable salts of formula [Cp*2Co][(ArN)14Ta6O] (Ar
) Ph, Ar ) p-MeC6H4). Small-
molecule reactivity studies on one of these clusters showed that its imido functionalities are moderately reactive
toward oxide donors but inert with respect to metallaheterocycle-forming processes. Clean imido/oxo exchange
was observed with aldehydes and ketones, leading cleanly to organic imines with no soluble byproducts being
observed. This exchange was also observed with a rhenium oxo compound (generating an imidorhenium complex
as the only soluble species). All 14 imido groups were transferred in these reactions, and no mixed-ligand cluster
intermediates were ever observed.
Introduction
catalytic processes including aziridination4 and hydroami-
nation,5 and in the latter case, isolated group 4 metal-imido
compounds have been shown to be catalytically active.6
A thorough understanding of multiple bonding between
transition metals and nitrogen ligands is of great theoretical
and commercial interest. Imido complexes are known for
the majority of the d-block metals, and a wide range of
substitution on the imido ligands has been demonstrated.1
The properties of the metal-nitrogen bond vary greatly with
differences in the steric and electronic environment: while
the imido functionalities in Schrock’s molybdenum olefin
metathesis catalysts are chemically inert and serve as
supporting ligands,2 terminal imido bonds to many early and
late transition metals are extremely reactive.3 Imido com-
plexes have been postulated as intermediates in several
(3) a) Glueck, D. S.; Hollander, F. J.; Bergman, R. G. J. Am. Chem. Soc.
1989, 111, 2719-2721. (b) Glueck, D. S.; Wu, J. X.; Hollander, F.
J.; Bergman, R. G. J. Am. Chem. Soc. 1991, 113, 2041-2054. (c)
Mindiola, D. J.; Hillhouse, G. L. J. Am. Chem. Soc. 2001, 123, 4623-
4624. (d) Takemoto, S.; Morita, H.; Kamikawa, K.; Matsuzaka, H.
Chem. Commun. 2006, 1328-1330. (e) Kogut, E.; Wiencko, H. L.;
Zhang, L. B.; Cordeau, D. E.; Warren, T. H. J. Am. Chem. Soc. 2005,
127, 11248-11249. (f) Danopoulos, A. A.; Wilkinson, G.; Sweet, T.
K. N.; Hursthouse, M. B. J. Chem. Soc., Dalton Trans. 1996, 3771-
3778. (g) Dobbs, D. A.; Bergman, R. G. J. Am. Chem. Soc. 1993,
115, 3836-3837.
(4) (a) Li, Z.; Quan, R. W.; Jacobsen, E. N. J. Am. Chem. Soc. 1995,
117, 5889-5890. (b) Groves, J. T.; Takahashi, T. J. Am. Chem. Soc.
1983, 105, 2073-2074.
(5) (a) Walsh, P. J.; Baranger, A. M.; Bergman, R. G. J. Am. Chem. Soc.
1992, 114, 1708-1719. (b) Haak, E.; Bytschkov, I.; Doye, S. Angew.
Chem., Int. Ed. 1999, 38, 3389-3391. (c) Shi, Y. H.; Ciszewski, J.
T.; Odom, A. L. Organometallics 2001, 20, 3967-3969. (d) Tillack,
A.; Castro, I. G.; Hartung, C. G.; Beller, M. Angew. Chem., Int. Ed.
2002, 41, 2541-2543.
(6) (a) Johnson, J. S.; Bergman, R. G. J. Am. Chem. Soc. 2001, 123, 2923-
2924. (b) Li, Y. H.; Shi, Y. H.; Odom, A. L. J. Am. Chem. Soc. 2004,
126, 1794-1803. (c) Baranger, A. M.; Walsh, P. J.; Bergman, R. G.
J. Am. Chem. Soc. 1993, 115, 2753-2763.
* To whom correspondence should be addressed. E-mail: arnold@
berkeley.edu (J.A.), rbergman@berkeley.edu (R.G.B.).
† Current address: Department of Chemistry, University of California,
Irvine, CA 92697.
(1) a) Wigley, D. E. In Progress in Inorganic Chemistry; John Wiley &
Sons Inc.: New York, 1994; Vol. 42, pp 239-482. (b) Gade, L. H.;
Mountford, P. Coord. Chem. ReV. 2001, 216, 65-97.
(2) Schrock, R. R.; Hoveyda, A. H. Angew. Chem., Int. Ed. 2003, 42,
4592-4633.
10.1021/ic701920v CCC: $40.75
Published on Web 12/29/2007
© 2008 American Chemical Society
Inorganic Chemistry, Vol. 47, No. 3, 2008 1053

Krinsky et al.
Table 1. Screen of Possible Trace Oxygen Sources, Added in
Substoichiometric Quantities
entry
Ta source
additivea
solvent
C6H6
C6H6
C6H6
C6H6
C6H6
C6H6
C6H6
yield of 2 (%)b
1
2
3
4
5
6
7
8
9
1
1
1
trace
0
0
5
5
20
0
57
36
O2
O2
c
1d
1d
1
H2O/O2
H2O
1
1
1
H2Oc
H2O
C6H6/hexane
C6H6
Ta2O5
a 0.2 equiv except Ta2O5. b 0.2 mmol of 1, 5 equiv of aniline, 5 mL of
solvent, 135 °C, 48 h, N2 atmosphere. c Added after heating. d Partially
oxidized.
Figure 1. Preliminary structure of octahedral hexatantalum cluster 2 (X
) O or N). Other unidentified fragments are present in the unit cell.
That such a novel cluster complex appears to be a
thermodynamic sink under the conditions in which it forms
was intriguing to us, and thus we wished to gain an
understanding of its properties as well as its exact identity.
The characterization of 2 and a few substituted derivatives,
as well as preliminary results pertaining to their physical and
chemical properties, was recently communicated.14 Herein
we report a more complete account of the synthesis and
characterization of a family of cluster compounds of general
formula (ArN)14Ta6O. Experimentally and computationally
derived insights into their geometric and electronic config-
urations are presented in detail, in addition to new results
concerning their chemical reactivity.
Work in our research groups has extended hydroamination
chemistry to group 5 metals, using neutral and cationic
alkyltantalum imides as catalysts.7,8 During these studies it
was discovered that the reaction of the hydroamination
catalyst Bn3TadN-t-Bu (1) with aniline yielded a bright red,
polynuclear tantalum species of uncertain composition
(compound 2, eq 1). A partial structural determination by
X-ray diffraction9 showed that 2 appeared to consist of an
octahedron of six tantalum atoms, surrounded by six terminal
and eight facially bridging phenylimido ligands (Figure 1).
A central atom was located but its identity could not be
assigned on the basis of the available data, and a large
amount of residual election density was observed in the unit
cell suggesting the presence of additional species.10
Results and Discussion
Synthesis and Characterization of (PhN)14Ta6O (2). As
stated above, the reaction of Bn3TadN-t-Bu (1) with excess
aniline yielded a bright red, sparingly soluble crystalline
material possessing the initially tentative formula (PhN)14Ta6O
(2). If this assignment were correct, a source of oxygen would
be necessary for the formation of 2. Given the low and
irreproducible yields obtained in its synthesis, the identity
of this adventitious oxygen source was at first uncertain.
Organic oxygen sources such as peroxides, pyridine N-oxide,
or phosphine oxides failed to yield 2. Likewise, negligible
quantities of 2 were isolated when very pure starting
materials were used and the glassware was freshly washed
with HF (Table 1, entry 1). While addition of a substoichio-
metric quantity of oxygen after heating the reaction mixture
afforded 2 when the reaction was run on very small scale,
this method was not effective on a preparative scale (Table
1, entry 3). However, when partially oxidized (old) samples
of 1 were employed, a consistent yield was obtained (Table
1, entry 4). It was finally discovered that the addition of water
(0.2 equiv relative to tantalum) to the reaction mixture before
heating at 135 °C greatly increased the product yield (Table
1, entry 6). Because the alkyltantalum starting material 1 is
extremely moisture sensitive, the added water must generate
tantalum oxo species that subsequently donate an oxide to a
reactive intermediate during the cluster-formation process.
Indeed, a small amount of white precipitate was always
observed in the reaction mixture after the addition of water.
Bn3TadN-t-Bu + PhNH2 135 °C8 red crystals
(1)
C6H6
2
1
5 equiv
24 h
Many hexanuclear transition metal clusters are present in
the literature; however, such dense packing of organic ligands
on these species is quite uncommon.11 The closest analogues
are the polyimido hexamolybdates and hexatungstates pre-
pared by treatment of the corresponding oxo clusters with
isocyanates.12 In these cases only the terminal oxo moieties
were replaced by imido or hydrazido groups, and thus to
our knowledge no other hexanuclear clusters have been
reported which contain alkyl- or arylimido ligands at all
terminal and bridging sites.13
(7) Anderson, L. L.; Arnold, J.; Bergman, R. G. Org. Lett. 2004, 6, 2519-
2522.
(8) One example of a V(IV) system exists; see the following: Lorber,
C.; Choukroun, R.; Vendier, L. Organometallics 2004, 23, 1845-
1850.
(9) The preliminary structure was provided by Dr. Alexandr Shafir.
(10) While only a central O2- would result in a neutral complex, the charge
resulting from the presence of N3- could be balanced by an as-yet
unobserved cation in the lattice.
(11) (a) Pope, M. T.; Mueller, A. Angew. Chem., Int. Ed. Engl. 1991, 30,
34-48. (b) Gouzerh, P.; Proust, A. Chem. ReV. 1998, 98, 77-111.
(12) (a) Strong, J. B.; Yap, G. P. A.; Ostrander, R.; Liable-Sands, L. M.;
Rheingold, A. L.; Thouvenot, R.; Gouzerh, P.; Maatta, E. A. J. Am.
Chem. Soc. 2000, 122, 639-649. (b) Clegg, W.; Errington, R. J.;
Fraser, K. A.; Holmes, S. A.; Schaefer, A. J. Chem. Soc., Chem.
Commun. 1995, 455-456. (c) Kang, H.; Zubieta, J. J. Chem. Soc.,
Chem. Commun. 1988, 1192-1193.
(13) (a) Kessler, V. G.; Seisenbaeva, G. A. Inorg. Chem. Commun. 2000,
3, 203-204. (b) Decker, A.; Fenske, D.; Maczek, K. Angew. Chem.,
Int. Ed. Engl. 1996, 35, 2863-2866.
(14) Krinsky, J. L.; Anderson, L. L.; Arnold, J.; Bergman, R. G. Angew.
Chem., Int. Ed. 2007, 46, 369-372.
1054 Inorganic Chemistry, Vol. 47, No. 3, 2008

Title Running Head: Hexatantalum Tetradecaimido Cluster Complexes
Occasionally, a large amount of crystalline 2 was deposited
above the solution on the walls of the reaction vessel during
heating. When the vessel was agitated so that the crystals
redissolved, that quantity of product was not isolated upon
cooling the reaction mixture. Subsequent studies indicated
that in general the reaction side products greatly increase
the solubility of the desired product, making crystallization
from the reaction mixture very inefficient. Following this
lead, the highest yields of 2 (up to 57%, Table 1, entry 8)
were obtained by decreasing the solvent polarity such that
the product precipitated as it was formed at high temperature.
Compound 2 is much less soluble than the other products,
and thus, the only necessary purification step was the facile
mechanical separation of the white precipitate (formed after
water addition) from the microcrystalline product. Tantalum
oxide was also moderately effective as an oxygen source
(Table 1, entry 9). Because air and water were rigorously
excluded from the reaction conditions employed previously,
varying quantities of residual tantalum oxide remaining on
the walls of the reaction vessels must have been responsible
for the variable cluster yield obtained in those cases. It had
been assumed that Ta2O5 was no more reactive than the Pyrex
vessels themselves and thus was sometimes allowed to
accumulate from one run to the next. There is evidence that
even silica can act as an oxide donor, as a few crystals of
cluster were sometimes observed in reaction mixtures
contained in flame-sealed NMR tubes, wherein all other
sources of oxygen were carefully excluded.
Establishing conditions for the growth of high-quality
crystals of 2 was nontrivial; the only crystals of sufficient
size that could be grown during the initial stages of this
investigation were of moderate quality, resulting in a
problematic X-ray diffraction data set. More recently, a
single-crystal synchrotron15 X-ray diffraction apparatus
became available to us, enabling the analysis of very small
crystals of 2 grown from benzene. A high-quality data set
was obtained, and structural solution yielded the centrosym-
metric octahedral structure observed previously (Figure 2).
The crystal lattice has the space-group symmetry P1h, with
the central atom occupying the inversion center. Compound
Figure 2. ORTEP diagram of (PhN)14Ta6O (2) at 50% probability. H atoms
are omitted for clarity. Legend: C, black; N, blue; O, red; Ta, gold. Bond
lengths (Å): Ta-O 2.2079(4), 2.2035(5), 2.2119(4); Ta-Nterminal 1.790-
(3), 1.793(4), 1.791(4).
is unfortunately not of sufficient quality to determine whether
this subtle effect is also present in that system. In con-
trast to the Ta-O distances, the Ta-Nterminal distances of
1.790(3) (Ta1-N3), 1.793(4) (Ta2-N4), and 1.791(4) Å
(Ta3-N7) are all equivalent within error. One of the three
unique Ta-Nterminal-Cphenyl bond angles deviates significantly
from 180°, with a C31-N3-Ta1 angle of 172.1(4)°. This
appears to be the result of a close intermolecular contact
and is not indicative of any internal electronic effect.
The only additional species observed in the unit cell are
four molecules of benzene. If the assignment of the central
atom as oxygen were incorrect, the cluster should carry a
charge and thus a counterion would be present. These data,
taken together with the synthetic results, strongly indicated
that the central atom in 2 is indeed oxygen. However, a
stereochemically inactive proton (providing charge balance
for an anionic cluster) would not be observed by the methods
discussed above and so some form of direct detection was
needed for an unambiguous assignment.
Mass spectrometry was an attractive technique because it
can very accurately determine the mass of a substance, thus
distinguishing between compounds of very similar composi-
tion. While the observation of a molecular ion of the correct
mass would be quite convincing, a change in mass with
different isotopically labeled oxide centers would be superior.
To this end, samples of substituted analogs (p-tolN)14Ta6O
(3) (p-tol ) p-MeC6H4) and (p-MeOC6H4N)14Ta6O (4)
(syntheses are discussed below) were prepared by following
the procedure discussed above but using water enriched in
17O and 18O. Of the mass spectrometry methods available,
all presented difficulties when evaluated for their utility with
respect to this system. Simple electron impact MS is
generally not suitable for compounds of such size, and soft
ionization techniques usually require the use of a matrix
2 possesses the Ta6O core of hexatantalate (Ta6O19 )
8- 16 but
with all of the terminal and edge-bridging oxo ligands
replaced by terminal and facially bridging phenylimido
ligands.
In the solid state, the Ta6O cluster core is slightly distorted
from Oh symmetry and the phenylimido ligands pack so that
the potential D2d symmetry is not realized. The Ta-O
distances are shorter than those in hexatantalate, measuring
2.2079(4), 2.2035(5), and 2.2119(4) Å for Ta1, Ta2, and Ta3,
respectively. This significant, though small, anisotropy in
the cluster core is recurrent in these compounds. It is unclear
if this is the result of a second-order Jahn-Teller effect or
simply a geometric consequence of maximizing bonding
between the tantalum centers and the bridging imido ligands.
The published structure of hexatantalate mentioned above
(15) Advanced Light Source beamline 11.3.1, Lawrence Berkeley National
Laboratory, operated under DOE contract DE-AC03-76SF00098.
(16) Pickhard, F.; Hartl, H. Z. Anorg. Allg. Chem. 1997, 623, 1311-1316.
Inorganic Chemistry, Vol. 47, No. 3, 2008 1055

Krinsky et al.
Scheme 1
Figure 3. Selected region of ESI mass spectra of 3, 3(17,18O), 4, and
4(17,18O) (top to bottom, respectively). Additional peak intensities are shown
at (m + 1)/z and (m + 2)/z in spectra of isotopically labeled samples.
cially considering that the percent yields of product dimin-
ished when the reaction was scaled above 100 mg of 1. A
preliminary screen of alternate metal sources showed that
halide-containing complexes such as (pyridine)2Cl3Tad
N-t-Bu or TaCl5 did not afford the desired product. This is
probably due to the inability of aniline to fully replace the
tantalum-bound chlorides. As anticipated, it was confirmed
that tantalum oxide itself is ineffective as a tantalum source,
only functioning as an oxide source under the reaction
conditions.
Turning to species with more reactive ligands, it was found
that commercially available pentakis(dimethylamido)tantalum
(Ta(NMe2)5) provided acceptable yields of 2. Additionally,
this reagent enabled the development of a large-scale
synthesis of 2, producing useful quantities of product after
several days at reflux in toluene (eq 2). While the percent
yields were lower than the best obtained at small scale using
1, they did not drop off with increasing reaction scale.
containing functionality not compatible with reactive early
metal complexes. With this in mind, electrospray ionization
(ESI) mass spectrometry was attempted using only THF as
a matrix.
Due to extensive cluster fragmentation and subsequent
reaction with the THF matrix, no molecular ions were
observed. Indeed, none of the observed fragments cor-
responded to combinations of the cluster core and arylimido
fragments. However, many of the high molecular weight
fragments (up to ca. 2000 Da) possessed masses 16 Da
greater for 4 (MeO) than for 3 (Me) (Figure 3, additional
spectra available in the Supporting Information). Therefore,
these fragments apparently each retained one imido ligand.
Although informative, analysis of these nonlabeled samples
did not provide any evidence of the nature of the central
cluster atom. Samples of 3 and 4 enriched in 17O and 18O
(3(17,18O), 4(17,18O)), however, provided a convincing result.
As shown in Figure 3, each high-mass fragment from the
nonlabeled cluster was accompanied by fragments that were
1 and 2 Da heavier, including those that were inferred to
still possess an imido ligand.
Ta(NMe2)5 + aniline + H2O reflux, 4 d8
2
38%
(2)
toluene
6 equiv
1 g
1/6 equiv
These extra peaks can only be attributed to the presence
of heavier isotopes of oxygen, as the labeled and nonlabeled
samples were prepared in exactly the same manner with the
same batches of starting materials (excepting water). Both
labeled and nonlabeled samples were of high purity and
crystalline, ruling out the presence of significant amounts
of tantalum oxide that could also be isotopically enriched in
labeled samples. While this unconventional methodology
provided correspondingly unconventional results, the data
provide convincing evidence for assignment of these com-
pounds as oxide-centered clusters.
Reaction Scope in Tantalum Source and Amine. Given
that the only fragment of 1 retained in the product cluster is
the tantalum ion, any tantalum complex capable of exchang-
ing all of its ligands with arylimido (and oxo) ligands should
be an effective synthon. Moreover, a more synthetically
accessible starting material would be very desirable, espe-
Screening of a set of commonly encountered amines
revealed that certain substituted anilines also afforded the
corresponding imido clusters (Scheme 1). The arylamines
p-toluidine and p-anisidine provided (p-tolN)14Ta6O (3) and
(p-MeOC6H4N)14Ta6O (4) (introduced earlier) isolated in
48% and 49% yield, respectively, using 1 as starting material.
Compound 3 was also prepared on a larger scale (625 mg,
20%) using Ta(NMe2)5 as described above and proved to
be the most synthetically useful of the derivatives presented
here. In this case it was necessary to employ a mixture of
toluene and n-octane as solvent to force the product to
precipitate from the reaction mixture upon cooling. Under
these conditions tantalum oxide proved to be preferable to
water as an oxide source. The extremely soluble (p-t-
BuC6H4N)14Ta6O (5) was also successfully prepared but in
very poor isolated yield (2%). Product isolation required that
the reaction be performed in pure n-octane, which resulted
1056 Inorganic Chemistry, Vol. 47, No. 3, 2008

Title Running Head: Hexatantalum Tetradecaimido Cluster Complexes
isolated in reasonably pure form, but the purification method
1
proved irreproducible and only H NMR data supported its
formulation. Other halogenated anilines afforded intractable
mixtures that did not appear, by NMR, to contain the desired
products. In contrast to p-bromoaniline, attempted cluster
synthesis using p-iodoaniline resulted only in black tars. It
seems that cluster formation from electron-poor anilines is
generally disfavored, as 3,5-fluoroaniline, p-trifluoromethyl-
aniline, p-nitroaniline, 4-aminostyrene, 4-aminobenzonitrile,
and 4-aminopyridine were all ineffective amine sources under
the conditions investigated. Finally, only arylamines were
tolerated: neither benzylamine nor isopropylamine afforded
the desired alkylimido clusters.
When 1 was treated with excesses of both p-toluidine and
p-anisidine simultaneously, a material was obtained that
appeared by 1H NMR and APCIMS17 to consist of a mixture
of compounds possessing the formula (p-MeOC6H4N)n(p-
tolN)14-nTa6O. The ratios of p-anisyl to p-tolyl groups
corresponded to a Gaussian distribution centered at about n
) 9. Thus, the two anilines react at competitive rates, but it
appears that electron-rich anilines are preferred although
more competition experiments would be needed to confirm
Figure 4. ORTEP diagram of (m-ClC6H4N)14Ta6O (7) at 50% probability.
H atoms are omitted for clarity. Legend: C, black; Cl, green; N, blue; O,
red; Ta, gold. Bond lengths (Å): Ta-O 2.1741(2), 2.1808(2), 2.1658(2);
Ta-Nterminal 1.811(4), 1.806(4), 1.811(4).
1
this. By H NMR there appears to be no preference for
p-anisyl or p-tolyl substitution in the terminal or bridging
positions, as the overall ratios of terminal to bridging
resonances were still 6:8 for both the MeO and Me signals.
Separation of the species by crystallization was not pos-
sible: beautifully faceted, dark purple crystals were always
obtained which contained the same distribution of cluster
substitution present in the supernatant.
Physical Properties, Experimentally Determined. Com-
pounds 2-7 are thermally stable, extremely air- and moisture-
sensitive crystalline solids. In solution, the more electron-
rich 3-5 decomposed over several days if stored in
polypropylene-capped vials inside an inert-atmosphere glove-
box. Nonetheless, a sample of 3 in C6D6 showed no
decomposition after it was heated at 200 °C for 2 weeks in
a flame-sealed NMR tube. Thus, in the absence of a reaction
partner these compounds appear to exist in their thermody-
namically favored configuration. The symmetrical structure
allows each electropositive tantalum center to access the lone
available oxo ligand, and the arylimido ligands are arranged
to maximize bonding with the Ta6O core.
Compound 2 is sparingly soluble in aromatic solvents and
only moderately so in THF, which is not surprising consider-
ing the presence of a closely packed array of unsubstituted
phenyl rings. While 3 and 4 are considerably more soluble,
the 14 methyl groups only increase their solubility to about
10 mM in benzene and slightly more in THF. Substitution
of the methyl groups for tert-butyl groups in 5, however,
increases the compound’s solubility to the extent that it is
very soluble in Et2O and slightly soluble in pentane.
Halogenated compounds 6 and 7 differ greatly in their
solubility, with 6 exhibiting good solubility in THF, but not
in aromatic solvents, and 7 resisting dissolution in all
nonreactive organic solvents once isolated in pure form.
in a precipitate of complex composition. After the sample
was washed with pentane, this precipitate was extracted with
Et2O, affording a small amount of material that required
much further washing with pentane before it could be
successfully crystallized.
Halide-substituted anilines proved problematic, as their
corresponding imido clusters were either highly insoluble
or exhibited solubilities similar to those of the unwanted
byproducts. Of the ones investigated, only p-bromoaniline
and m-chloroaniline provided an isolable product, (p-
BrC6H4N)14Ta6O (6) and (m-ClC6H4N)14Ta6O (7) in 4% and
15% yield, respectively. In both cases the principal factor
responsible for the low yield was the difficulty with which
the product was separated from the reaction mixture.
While the insolubility of isolated 7 precluded spectroscopic
characterization, high-quality crystals were obtained from a
reaction mixture and subjected to X-ray diffraction analysis
(Figure 4). The structure is analogous to that of 2 but with
two molecules of t-BuNH2 and two molecules of benzene
present in the unit cell. The packing of the aryl substituents
is remarkably ordered, with only one of the seven unique
m-chlorophenyl groups disordered between two positions
related by an approximate 2-fold rotation about the aryl C-N
bond. The Ta-O distances in 7 are 2.1741(2), 2.1808(2),
and 2.1658(2) Å for Ta1, Ta2, and Ta3, respectively. These
bond distances are slightly contracted (∆ave ) 0.034 Å)
relative to those of 2, while the Ta-Nterminal bond dis-
tances are slightly longer than those of 2, measuring
1.811(4) (Ta1-N1), 1.806(4) (Ta2-N2), and 1.811(4) Å
(Ta3-N3).
Cluster compounds arising from p-Me2NC6H4NH2 and
4-aminobiphenyl were apparently observed in solution but
proved inseparable from the reaction byproducts. On one
occasion, a small quantity of (p-Me2NC6H4N)14Ta6O (8) was
(17) Atmospheric pressure chemical ionization mass spectrometry, THF
matrix.
Inorganic Chemistry, Vol. 47, No. 3, 2008 1057

Krinsky et al.
Figure 7. Cyclic voltammogram of 3 in THF with 0.1 M [Bu4N][PF6],
with scan rate ) 50 mV/s. Reductions: -1.44, -2.14 V. Oxidations: 0.01,
0.27 V (relative to Fc/Fc+ couple).
Figure 5. 1H NMR spectrum of 3 in C6D6, aryl region only. All bridging
or terminal phenylimido groups are spectroscopically equivalent.
Because the clusters are formally d0 tantalum compounds,
the charge-transfer excitations discussed above should be
ligand-to-metal charge transfers. The UV-vis data for 2-4
indicate that this transition is facile and thus the cluster
LUMO should be relatively low-lying. This could be the case
if the LUMO were a combination of d orbitals located on
more than one tantalum center, providing a delocalized and
thus stabilized molecular orbital.
Data obtained from analysis of 3 by cyclic voltammetry
are consistent with this bonding scenario (Figure 7). A
reversible reduction was observed at -1.44 V (vs Fc/Fc+
couple, Fc ) ferrocene), indicating that relatively mild
reducing agents should chemically reduce 3 in THF. An
additional reduction occurred at -2.14 V; however, its
reversibility is uncertain due to its position so near the
potential limit of the solvent. There were also two irreversible
oxidations, one at 0.01 V and one at 0.27 V. Therefore, 3
displays a first oxidation potential nearly identical with that
of ferrocene; however, the cations produced are apparently
very unstable.
Physical Properties, Computed by Hybrid DFT. The
physical properties discussed in the previous section suggest
a cluster bonding scheme wherein the high-lying populated
orbitals are ligand-centered, in analogy with mononuclear
d0 metal complexes. However, the ease with which 3 is
reduced and the intense visible-region absorption bands
indicate that the LUMOs of the clusters are stabilized relative
to their monomeric relatives. With the hope of a better
understanding concerning the nature of this stabilization,
electronic structure calculations at the X3LYP18/LACVP*
level of theory were performed using the crystallographically
determined coordinates of 2. Figure 8 depicts the HOMO of
2, one of three nearly degenerate molecular orbitals that
would possess tg symmetry were the cluster perfectly
octahedrally symmetric. As expected, these orbitals are purely
imido ligand-centered, consisting of linear combinations of
fragment orbitals analogous to the HOMO of aniline. The
majority of the electron density resides on the bridging imido
nitrogens, and each imido fragment orbital is weakly
antibonding with its closest neighbors.
Figure 6. UV-visible absorption spectra of 2-4 in THF. Legend: 2,
blue; 3, red; 4, black. Insert: visible-wavelength absorption onsets.
As stated above, the most-symmetric conformation of 2
would have D2d symmetry. While only inversion symmetry
was observed in the solid-state structure, NMR spectra
showed apparent Oh symmetry with two sets of aryl signals
corresponding to the terminal and facially bridging sites
(Figure 5). Thus, the imido aryl groups must rotate rapidly
relative to one another so that the effective electronic
environment on each side of each aryl group is equivalent.
From inspection of the solid-state structure, it is apparent
that the aryl groups are sufficiently close to one another that
they should not be free to rotate completely independently,
and so there should be some barrier to aryl group rotation.
This barrier must be very low, however, as 1H NMR spectra
showed no broadening or decoalescence of signals when a
sample of 5 in THF-d8 was cooled to -100 °C.
Compound hues vary from red/orange (7) to cherry red
for the alkyl-substituted 3 and 5, to red/black for the more
electron-rich 4. The very electron-rich 8 possesses a deep
brown/black color. Bromide substitution in the para position
of a phenyl ring presents an ambiguous situation electroni-
cally, but the deep crimson hue of 6 suggests that in this
case the bromide substituents are moderately electron-
donating. Compounds 2-4 exhibit intense UV absorptions
only slightly red-shifted from those of the anilines from
which they are derived (Figure 6). However, in contrast to
aniline these compounds also show charge-transfer bands in
the visible region, with onsets of about 600 nm (Figure 6
insert). Both features are more red-shifted with increasingly
electron-rich substitution. Also unlike aniline, fluorescence
measurements detected no emission between 300 and 1000
nm using a range of excitation wavelengths. The large
number and density of vibrational modes accessible to these
clusters probably makes internal conversion and subsequent
thermal relaxation the dominant excited-state decay pathway.
In analogy to mononuclear d0 metal complexes, the lowest
unoccupied MOs of 2 are metal-centered (Figure 9), with
no contributions from the imido fragments or central oxide.
However, in this case the LUMOs are linear combinations
(18) Extension of B3LYP by Xu and Goddard to include the Perdew-Wang
1991 gradient exchange correction.
1058 Inorganic Chemistry, Vol. 47, No. 3, 2008

Title Running Head: Hexatantalum Tetradecaimido Cluster Complexes
Table 2. MO Energies (eV) for X3LYP/LACVP*-Optimized 2-4
and 9
orbital
HOMO
LUMO
9
2
3
4
-7.06
-3.88
3.18
-5.46
-2.51
2.95
-5.18
-2.29
2.89
-4.81
-2.11
2.70
HOMO-LUMO gap
Scheme 2
performed on models of 3, 4, and a hypothetical complex
bearing electron-withdrawing 4-pyridylimido ligands,
(NC5H4N)14Ta6O (9). As a control experiment, 2 was first
optimized both under the constraint of inversion symmetry
and with no geometric constraints. While the Ta-Nterminal
bond distances in optimized 2 are essentially identical with
the experimentally determined values, the calculated Ta-O
distances average 0.016 Å longer. This discrepancy is pos-
sibly due to the lack of polarization functions in the basis
set used for tantalum, but larger basis sets proved prohibi-
tively expensive in this large system. Both optimized struc-
tures exhibited MO surfaces and energies very similar to
those calculated from the experimentally determined struc-
ture, and thus, the substituted models were subsequently opti-
mized using inversion symmetry as a constraint (greatly
reducing computational cost). Positional coordinates of the
optimized structures are available in the Supporting Informa-
tion.
Given the potential shortcomings of DFT in reproducing
meaningful energy values, the energies listed in Table 2
should not be directly compared with any observed values
of physical properties.19 However, because the four species
compared here are so similar, clear trends in energy level
changes with changing ligand substitution should be reason-
ably reliable. In the order from electron-poor 9 to electron-
rich 4, the HOMO energy increases by 2.25 eV while the
LUMO energy increases by only 1.77 eV. This results in a
decrease of the HOMO-LUMO energy gap with increasing
electron-donating ability of the arylimido substituents. In a
qualitative sense the antibonding, ligand-centered HOMO is
more destabilized than the metal-centered LUMO by the
presence of increasingly electron-rich ligands. These data
provide a theoretical rationale for the experimentally ob-
served red-shifts in the UV-visible absorption bands.
Chemical Reactivity Studies. Observations and calcula-
tions relating to electronic structure suggested that the title
compounds should be reactive toward chemical oxidants and
reductants, and treatment with the latter might yield stable
anionic cluster compounds. When compounds 2 and 3 were
treated with silver trifluoromethanesulfonate (OTf), dark tars
were obtained which displayed no observable signals in their
1H NMR spectra (Scheme 2). This result is consistent with
cluster oxidation by silver(I), which was predicted to yield
Figure 8. Orbital diagram of the HOMO of 2, calculated at the X3LYP/
LACVP* level of theory using crystallographically determined coordinates.
Figure 9. Orbital diagrams of the LUMO (left) and LUMO+1 (right) of
2, calculated at the X3LYP/LACVP* level of theory using crystallographi-
cally determined coordinates. The long axis of the cluster core is normal to
the plane of the page.
of d orbitals from all six tantalum centers and, thus, if
populated, the electron density would be much more delo-
calized than in mononuclear complexes. These two LUMOs
are nearly degenerate and would form a set of eg symmetry
if the cluster were truly octahedral. Therefore, the orbital
scheme is analogous to that of a d6 octahedral species
wherein the frontier orbitals consist of a filled, triply
degenerate set and a doubly degenerate virtual set. The
difference is that here the t set is ligand-centered and the
octahedral symmetry corresponds to the cluster as a whole
instead of a single metal center.
Closer inspection of the LUMOs of 2 reveals that the
LUMO possesses a slightly bonding combination of d orbitals
about two meridians defined by four tantalum centers each,
while the combination is slightly antibonding about the third
meridian. The LUMO+1 possesses a bonding combination
about one meridian only. Thus, population of the former
orbital might result in a Jahn-Teller distortion wherein two
opposite Ta-O distances are contracted, while population
of the latter would have the opposite effect.
The above discussion provides some explanation for the
observed properties of 2. We wished to extend this investiga-
tion to compounds for which no experimentally determined
structure was available, and so geometry optimizations were
(19) Schreiner, P. R. Angew. Chem., Int. Ed. 2007, 46, 4217-4219.
Inorganic Chemistry, Vol. 47, No. 3, 2008 1059

Krinsky et al.
(8) Å are in good agreement with those of other structurally
determined Cp*2Co+ salts.20 If electron transfer had not
occurred, much larger Co-C distances corresponding to the
neutral Cp*2Co would have been observed. This claim is
supported by the computed average dCo-C values of 2.091
and 2.168 Å for Cp*2Co+ and Cp*2Co, respectively (X3LYP/
LACVP* geometry optimizations, positional coordinates are
available in the Supporting Information). These data, together
with the observed changes in solubility and spectroscopic
properties upon treatment of 2 and 3 with Cp*2Co, provide
strong evidence for the assignment of 10 and 11 as ionic
compounds.
Because each tantalum center in these cluster compounds
is coordinatively saturated and embedded in what was
assumed to be a very bulky ligand set, it was not expected
that the cluster imido groups would exhibit reactivity often
observed in other early metal systems. A vacant cis coor-
dination site is usually necessary for metallacycle formation
by insertion of an unsaturated organic molecule across the
metal-imido bond.21 While metathesis reactions may be
feasible if metallacycle formation is only involved in the
transition state, the unfavorable steric situation should impede
any such reactivity. Accordingly, unsaturated molecules
including phenylacetylene, isothiocyanates, isocyanates, and
carbon monoxide did not react with 3 upon extended heating.
Although treatment of 3 with carbon dioxide did produce
p-tolNCO22 (eq 3), only trace quantities were observed by
1H NMR after the reaction mixture was heated at 135 °C
for 2 weeks. The observed reactivity is opposite to that of
the molybdate system reported by Maatta, in which imido-
molybdenum clusters were generated by treatment of hexa-
molybdate with isocyanates, generating carbon dioxide as a
byproduct.23
Figure 10. ORTEP diagram of [(PhN)14Ta6O][Cp*2Co] (10) at 50%
probability. H atoms are omitted for clarity. Legend: C, black; Co, green;
N, blue; O, red; Ta, gold. Bond lengths (Å): Ta-O 2.2018(4), 2.2099(4),
2.2120(4); Ta-Nterminal 1.785(6), 1.788(6), 1.803(6); Co-Cring 2.041(9),
2.043(9), 2.043(9), 2.054(9), 2.054(8).
reactive organic-centered radical cations. Chemical reduction
of 2 and 3 by decamethylcobaltocene (Cp*2Co), however,
provided isolable decamethylcobaltocenium salts of formula
[Cp*2Co][(ArN)14Ta6O] (Ar ) Ph (10), Ar ) p-tol (11)) in
good yield.
Compounds 10 and 11 are green/brown solids that form
brown solutions, although they are sparingly soluble in THF
and insoluble in aromatic solvents. Signals in their 1H NMR
spectra are broad but within the normal diamagnetic chemical
shift range. The absence of a signal at ca. 45 ppm corre-
sponding to neutral Cp*2Co confirmed that it had been
oxidized to its diamagnetic cation (the signal for which was
present at ca. 2 ppm). Magnetic moment measurements on
a solid sample of 11 yielded a µeff value of 1.78 µB, which
is very close to the theoretical spin-only value of 1.73
expected for a system containing one unpaired electron.
The solid-state structure of 10 is depicted in Figure 10.
As with the structure of 2, the central oxide is located on an
inversion center and thus the asymmetric unit contains only
half of the cluster anion. The cobalt of the Cp*2Co+ cation
is also situated on an inversion center. The Ta-O distances
in 10 are very similar to those in neutral 2, measuring 2.2018-
(4), 2.2099(4), and 2.2120(4) Å. To ensure accurate com-
parisons, larger crystals of 2 were successfully grown from
THF solution and its structure was redetermined using Mo
KR radiation, as was used for data collection on 10. The
redetermined Ta-O distances for 2 are 2.1949(4), 2.2096-
(4), and 2.2101(4) Å (CIF available in Supporting Informa-
tion).
CO2
excess
3 +
135 °C8 “TaxOy” + p-MeC6H4NCO
(3)
C6D6
In contrast to reactions with the π-ligands discussed above,
treatment of 3 with sterically nondemanding alcohols such
as ethanol and phenol resulted in evolution of p-toluidine
and the formation of tantalum alkoxide products. This
reactivity is not surprising given the high water sensitivity
of the title compounds. Multiple alkoxide complexes were
formed in reactions employing small alcohols, none of which
were isolated. Treatment with the more bulky triphenylsil-
anol, however, yielded only two siloxide-containing products,
1
as detected by H NMR spectroscopy. The major product
was easily crystallized from the reaction mixture and
identified as the homoleptic siloxide Ta(OSiPh3)5 (12)
(Scheme 3). The minor product, presumably containing the
(20) (a) Uhl, W.; Cuypers, L.; Kaim, W.; Schwederski, B.; Koch, R. Angew.
Chem., Int. Ed. 2003, 42, 2422-2423. (b) Heise, H.; Kohler, F. H.;
Herker, M.; Hiller, W. J. Am. Chem. Soc. 2002, 124, 10823-10832.
(c) Braga, D.; Eckert, M.; Fraccastoro, M.; Maini, L.; Grepioni, F.;
Caneschi, A.; Sessoli, R. New J. Chem. 2002, 26, 1280-1286.
(21) Duncan, A. P.; Bergman, R. G. Chem. Rec. 2002, 2, 431-445.
(22) Blake, R. E., Jr.; Antonelli, D. M.; Henling, L. M.; Schaefer, W. P.;
Hardcastle, K. I.; Bercaw, J. E. Organometallics 1998, 17, 718-725.
(23) Strong, J. B.; Ostrander, R.; Rheingold, A. L.; Maatta, E. A. J. Am.
Chem. Soc. 1994, 116, 3601-3602.
In a comparison of these two data sets, it is evident that
no significant change in Ta-O distance accompanied reduc-
tion. Therefore, either the metal-centered, singly occupied
molecular orbital (SOMO) in 10 is not sufficiently bonding
in character to have an observable effect or that orbital is in
fact nonbonding. The Cp*2Co+ cation’s Co-C bond dis-
tances of 2.041(9), 2.043(9), 2.043(9), 2.054(9), and 2.054-
1060 Inorganic Chemistry, Vol. 47, No. 3, 2008

Title Running Head: Hexatantalum Tetradecaimido Cluster Complexes
Scheme 3
central oxide from 3, was not successfully isolated in pure
form. Compound 12 was also prepared by treatment of
Ta(NMe2)5 with 5 equiv of triphenylsilanol. The crystallo-
graphically determined structure of 12 is shown in Figure
11. The geometry about the tantalum center is nearly
perfectly trigonal bipyramidal with a noncrystallographic
3-fold rotation axis defined by the axial ligands. The
bond angle between the axial siloxy ligands is 178.1(4)°,
and the sum of the equatorial O-Ta-O angles is 360.0(7)°.
The Ta-O bond lengths are slightly longer than those
of (t-Bu3SiO)3TadPPh24 and (t-Bu3SiO)3TadCdCdTa-
(t-Bu3SiO)3,25 possibly due to steric pressure in 12. Only a
few other homoleptic tantalum siloxides have been reported
previously, all containing only alkylsiloxy ligands, and none
were crystallographically characterized.26 To our knowledge,
none of the previously reported homoleptic alkoxy- or
mixed-alkoxy-tantalum species are monomeric.27
Figure 11. ORTEP diagrams showing two views of Ta(OSiPh3)5 (12) at
50% probability. H atoms are omitted for clarity. Legend: C, black; O,
red; Si, yellow; Ta, gold. Selected bond lengths (Å) and angles (deg):
Ta1-O1, 1.949(6); Ta1-O2, 1.914(7); Ta1-O3, 1.886(5); Ta1-O4, 1.884-
(5); Ta1-O5, 1.889(5); O1-Ta1-O2, 178.1(4); O3-Ta1-O4, 118.2(3);
O4-Ta1-O5, 119.1(2); O3-Ta1-O5, 122.7(2).
The reactivity discussed above indicated that ligand
substitution reactions at the cluster tantalum centers are
sterically feasible. Hence, the lack of reactivity with isocy-
anates must be of electronic origin, possibly a result of the
necessity of forming relatively high-energy carbodiimides.
Investigations employing simple organic carbonyls appeared
to confirm this postulate (Scheme 4). When compound 3
was treated with 14 equiv of benzaldehyde, quantitative
conversion to benzaldehyde p-tolylimine was observed after
2 d at ambient temperature. Benzophenone was also reactive,
providing the corresponding ketimine in quantitative yield
Scheme 4
1
(by H NMR) after heating at 135 °C overnight. The only
other product from these reactions was insoluble tantalum
oxide. The R-hydrogen-bearing acetophenone also provided
1
product but in decreased yield (75% by H NMR). While
the desired â-diketimine was observed in the reaction mixture
after treatment of 3 with acetylacetone, free p-toluidine was
present shortly after ketone addition and so the transformation
probably followed the classical condensation route. In this
case it appears that the cluster ligands were protonated by
the acidic diketone and the resulting tantalum species simply
acted as a Lewis acid and water scavenger. It has been known
for some time that certain early metal imides undergo imido/
oxo exchange with aldehydes and ketones.28 However, these
(24) Bonanno, J. B.; Wolczanski, P. T.; Lobkovsky, E. B. J. Am. Chem.
Soc. 1994, 116, 11159-11160.
(25) Neithamer, D. R.; Lapointe, R. E.; Wheeler, R. A.; Richeson, D. S.;
Vanduyne, G. D.; Wolczanski, P. T. J. Am. Chem. Soc. 1989, 111,
9056-9072.
(26) (a) Thomas, I. M. Can. J. Chem. 1961, 39, 1386-1389. (b) Bradley,
D. C. Coord. Chem. ReV. 1967, 2, 299-318.
(27) (a) Lewis, L. N.; Garbauskas, M. F. Inorg. Chem. 1985, 24, 363-
366. (b) Boyle, T. J.; Gallegos, J. J.; Pedrotty, D. M.; Mechenbier, E.
R.; Scott, B. L. J. Coord. Chem. 1999, 47, 155-171. (c) Johansson,
A.; Roman, M.; Seisenbaeva, G. A.; Kloo, L.; Szabo, Z.; Kessler, V.
G. J. Chem. Soc., Dalton Trans. 2000, 387-394.
species are generally coordinatively unsaturated and also
reactive toward isocyanates and carbon monoxide.
In addition to their reactivity with organic carbonyl com-
pounds, some imido complexes have been shown to undergo
imido/oxo exchange with metal oxo complexes.29 This
Inorganic Chemistry, Vol. 47, No. 3, 2008 1061

Krinsky et al.
Scheme 5
transiently formed in the ligand sphere; the clusters reversibly
fragment in solution.
A number of experiments were performed to test for cluster
lability. First, a solution of compound 3 was heated in the
presence of excess p-anisidine, and separately a solution of
4 was heated in the presence of excess p-toluidine. Depend-
ing on the thermodynamic preference for electron-rich or
electron-poor arylimido groups, one of these reaction mix-
tures should have produced a mixed-ligand species if the
imido groups of 3 or 4 were thermally labile. No exchange
was observed at temperatures at which the imido/oxo
exchange reactions were performed (Scheme 5). There is
evidence in the literature that acid can catalyze this type of
reaction,32 but here the addition of anilinium triflate had no
effect. An additional experiment involved heating an equimo-
lar mixture of 3 and 4. Again, no mixed-ligand clusters were
observed.
A recurring observation provides some evidence for the
ligand-compaction mechanism mentioned above: in all
reactions resulting in oxide transfer to the cluster tantalum
sites, no intermediate oxo/imido clusters were ever observed.
Thus, the remaining 13 imido groups must have reacted very
quickly after the first was exchanged. The only forthcoming
explanation for this phenomenon is that the first substitution
results in an area of little steric hindrance owing to the small
size of an oxo ligand relative to an arylimide. With this hole
opened in the ligand sphere, substrates could then access
the tantalum-nitrogen linkages much more easily and thus
each subsequent exchange would become progressively more
facile.
Figure 12. ORTEP diagram of Cp*Cl2RedN-p-C6H4Me (13) at 50%
probability. H atoms are omitted for clarity. Legend: C, black; Cl, green;
N, blue; Re, orange. Selected bond lengths (Å) and angles (deg): Re1-
N1, 1.726(4); Re1-Cl1, 2.3783(13); Re1-Cl2, 2.3845(13); Re1-C1, 2.217-
(5); Re1-C2, 2.178(5); Re1-C3, 2.224(5); Re1-C4, 2.415(5); Re1-C5,
2.405(4), N1-C11, 1.387(6); Re1-N1-C11, 171.5(3).
reaction is thermodynamically favored when the oxo complex
contains a metal that is more electronegative than that of
the imido species.30 Nonetheless, we were very surprised to
find that 3 reacted with 14 equiv of Cp*Cl2RedO at 75 °C,
forming Cp*Cl2RedN-p-tol (13) as the only soluble product
(eq 4). A black, insoluble precipitate was also obtained. As
no starting materials or other soluble products were present,
this material must contain Cp*Cl2Re and p-tolylimido frag-
ments as well as tantalum oxide. Compound 13 has been
fully characterized and its identity confirmed by X-ray dif-
fraction analysis (Figure 12). The structure is unremarkable,
being closely analogous to that of the known Cp*Cl2Red
N-t-Bu.31
The oxo ligand of Cp*Cl2RedO resides close to the bulky
Cp* ligand, and so there must be much more available space
between the arylimido groups of 3 than was originally
assumed. In fact, even the very sterically shielded 5 cleanly
exchanged its imido ligands to afford Cp*Cl2RedN-p-C6H4-
t-Bu (14). This reaction occurred very slowly at 75 °C but
was complete after 12 h at 135 °C. There are two scenarios
that would account for this reactivity: the arylimido groups
could compact in a concerted manner so that a large hole is
Conclusions
A new type of polyatomic imido complex was fully
characterized. The structure was confirmed by several X-ray
diffraction experiments, and the identity of the cluster-core
oxide was confirmed by mass spectrometry using isotopically
labeled samples. Reproducible, acceptably high-yielding
syntheses were developed for cluster species containing a
variety of substituted arylimido ligands. Optical properties
were found to vary with ligand substitution, resulting in red-
shifts in the UV-visible spectra of more electron-rich
species. Electronic structure calculations suggested that the
phenomenon primarily arises from the destabilization of
the ligand-centered HOMOs, while the tantalum-centered
LUMOs are relatively unperturbed. Reactivity studies using
one cluster derivative showed a surprising steric flexibility
in the clusters’ imido ligand sphere. Moderately bulky
(28) (a) Walsh, P. J.; Hollander, F. J.; Bergman, R. G. Organometallics
1993, 12, 3705-3723. (b) Royo, P.; Sanchez-Nieves, J. J. Organomet.
Chem. 2000, 597, 61-68. (c) Hanna, T. E.; Keresztes, I.; Lobkovsky,
E.; Bernskoetter, W. H.; Chirik, P. J. Organometallics 2004, 23, 3448-
3458. (d) Guiducci, A. E.; Boyd, C. L.; Mountford, P. Organometallics
2006, 25, 1167-1187.
(29) (a) Jolly, M.; Mitchell, J. P.; Gibson, V. C. J. Chem. Soc., Dalton
Trans. 1992, 1331-1332. (b) Wang, W. D.; Espenson, J. H. Inorg.
Chem. 2002, 41, 1782-1787. (c) Gibson, V. C.; Graham, A. J.; Jolly,
M.; Mitchell, J. P. Dalton Trans. 2003, 4457-4465.
(30) Holland, P. L.; Andersen, R. A.; Bergman, R. G. Organometallics
1998, 17, 433-437.
(31) Herrmann, W. A.; Weichselbaumer, G.; Paciello, R. A.; Fischer, R.
A.; Herdtweck, E.; Okuda, J.; Marz, D. W. Organometallics 1990, 9,
489-496.
(32) Burland, M. C.; Pontz, T. W.; Meyer, T. Y. Organometallics 2002,
21, 1933-1941.
1062 Inorganic Chemistry, Vol. 47, No. 3, 2008

Title Running Head: Hexatantalum Tetradecaimido Cluster Complexes
All starting materials obtained from commercial sources were used
without further purification, except the anilines (distilled or sublimed
prior to use). Compound 17 and Cp*Cl2RedO34 were prepared as
previously reported. Isotopically enriched water (30% 17O, 50%
18O) was purchased from Isotec and degassed prior to use.
organic molecules were found to react with 3 at surprisingly
low temperatures, provided the reaction was strongly ther-
modynamically favored (e.g., exchange of imido ligands for
oxo or alkoxy moieties).
Although this report provides a general overview of the
title compounds, several issues have yet to be resolved. For
example, evidence was presented indicating an energetic
preference for electron-donating arylimido ligands, even
while these substituents appear to be electronically destabi-
lizing. Whether this preference is kinetic or thermodynamic
is not currently known; indeed, the mechanism of cluster
formation is entirely mysterious at present. The heterogeneity
of the reaction mixture, as well as the polymeric nature of
the initial arylamide substitution product, has so far precluded
any thorough spectroscopic investigation. The fact that no
well-defined intermediates were observed in NMR spectra
during the course of the cluster-forming reaction suggests
that the closo-hexatantalum architecture forms rapidly after
the oxide moiety is trapped by some reversibly formed imido/
amido species. If conditions were developed under which
the kinetics of the reaction could be determined, some
information might be gained regarding the factors controlling
cluster assembly. In addition, elucidation of the mechanism
accounting for the quantitative exchange of imido and oxo
groups between the cluster and a cyclopentadienyl(oxo)-
rhenium complex may explain the absence of observable
partially exchanged cluster intermediates in this reaction.
Finally, the scope in metal source should be explored more
fully. The discovery that this reaction is general for tantalum
compounds containing only carbon and nitrogen ligands
would have disturbing implications for the study of this
element as a hydroamination or amidation catalyst, as these
types of donors (cyclopentadienides, amides, etc.) are cur-
rently considered the most promising supporting ligands.
Representative Procedure for X-ray Crystallography. A
crystal of appropriate size was mounted on a Kaptan loop using
Paratone-N hydrocarbon oil. The crystal was transferred to a
diffractometer/CCD area detector,35 centered in the beam, and
cooled by a nitrogen flow low-temperature apparatus that had been
previously calibrated by a thermocouple placed at the same position
as the crystal. An arbitrary hemisphere of data was collected, and
the raw data were integrated using SAINT.36 Cell dimensions
reported were calculated from all reflections with I > 10σ. The
data were corrected for Lorentz and polarization effects, but no
correction for crystal decay was applied. Data were analyzed for
agreement and possible absorption using XPREP.37 An empirical
absorption correction based on comparison of redundant and
equivalent reflections was applied using SADABS.38 The structure
was solved via direct methods, expanded using Fourier techniques,
and refined on F2 with the SHELXTL-97 suite of programs.39
ORTEP diagrams were created using the ORTEP-3 software
package40 and rendered in POV-ray.4141 and 4.
Computational Methods. Molecular orbital calculations were
performed using the Jaguar 6.5 package42 with Maestro 7.5 as the
graphical user interface43 at the UC, Berkeley, CA, Molecular
Graphics Facility. The hybrid DFT functional X3LYP44 was used
throughout, which consists of the Becke three-parameter functional45
and the correlation functional of Lee, Yang, and Par.46 It also
includes the Perdew-Wang 1991 gradient exchange functional,47
and the exchange has been parametrized to fit the Gaussian
exchange density. We found that for this system the X3LYP
functional gives results comparable to the well-known B3LYP48
but at slightly less computational cost. The LACVP* basis set was
employed, which uses an effective core potential and valence
double-ú contraction basis functions for metals.49 The remaining
atoms were treated with Pople’s 6-31G* basis set, which includes
Experimental Section
(34) Herrmann, W. A.; Floel, M.; Kulpe, J.; Felixberger, J. K.; Herdtweck,
E. J. Organomet. Chem. 1988, 355, 297-313.
(35) (a) Compound 2 (synchrotron): APEX2: Area-Detector Software
Package, version 1.0.27; Bruker AXS: Madison, WI, 1995-99. (b)
SMART: Area-Detector Software Package, version 5.631; Bruker
AXS: Madison, WI, 1995-99.
(36) a) Compound 2 (synchrotron): SAINT: SAX Area-Detector Integration
Program, version 7.06; Siemens Industrial Automation Inc: Madison,
WI, 2005. (b) SAINT: SAX Area Detector Integration Program,
version 7.07B; Siemens Industrial Automation Inc: Madison, WI,
2005.
(37) XPREP, version 6.12; Bruker AXS: Madison, WI, 2001 (part of the
SHELXTL Crystal Structure Determination Package).
(38) SADABS: Siemens Area Detector ABSorption Correction Program,
version 2.10; Bruker AXS: Madison, WI, 2005.
General Methods. All manipulations were performed using
standard Schlenk-line techniques or an N2-atmosphere glovebox.
Reactions followed by NMR spectroscopy were performed in 5
mm Pyrex NMR tubes which were flame-sealed under reduced
pressure. Solvents were dried by passage through a column of
activated alumina,33 degassed with nitrogen and stored over 4 Å
molecular sieves. Solvents were tested with Na/benzophenone
prior to use. Deuterated solvents were vacuum-transferred from
Na/benzophenone or CaH2 and stored in Teflon-sealed Kontes
vessels inside a glovebox. NMR spectra were obtained with a
Bruker DRX-500 spectrometer. Chemical shifts were calibrated
relative to residual solvent peaks and are reported relative to TMS.
UV-visible absorption spectra were recorded in THF on a Hewlett-
Packard 8453 spectrophotometer using a 2 mm path-length cell.
Elemental analyses were performed at the UC, Berkeley, CA,
Microanalytical Facility with a Perkin-Elmer 2400 Series II
CHNO/S analyzer. Electrospray mass spectrometry was performed
by Dr. John Greaves at the UC, Irvine, CA, Mass Spectrometry
Facility. Electrochemical investigations were performed on a BAS
100B electrochemical analyzer. Magnetic moment data were
collected on a Quantum Design MPMS XL SQUID magnetometer.
(39) Sheldrick, G. SHELXTL Crystal Structure Determination Package;
Bruker AXS: Madison, WI, 1995-1999.
(40) Farrugia, L. J. J. Appl. Crystallogr. 1997, 30, 565.
(41) POV-ray, v.3.5.icl.win32, copyright 1996-2002.
(42) Jaguar, version 6.5; Schrodinger, LLC: New York, 2005.
(43) Maestro, version 7.5; Schrodinger, LLC: New York, 2005.
(44) Xu, X.; Goddard, W. A., III. Proc. Natl. Acd. Sci. U.S.A. 2004, 101,
2673-2677.
(45) Becke, A. D. J. Chem. Phys. 1993, 98, 5648-5652.
(46) Miehlich, B.; Savin, A.; Stoll, H.; Preuss, H. Chem. Phys. Lett. 1989,
157, 200-206.
(47) Perdew, J. P.; Chevary, J. A.; Vosko, S. H.; Jackson, K. A.; Pederson,
M. R.; Singh, D. J.; Fiolhais, C. Phys. ReV. B 1992, 46, 6671-6687.
(48) Stephens, P. J.; Devlin, F. J.; Chabalowski, C. F.; Frisch, M. J. J.
Phys. Chem. 1994, 98, 11623-11627.
(33) Alaimo, P. J.; Peters, D. W.; Arnold, J.; Bergman, R. G. J. Chem. Ed.
2001, 78, 64.
(49) Hay, P. J.; Wadt, W. R. J. Chem. Phys. 1985, 82, 299-310.
Inorganic Chemistry, Vol. 47, No. 3, 2008 1063

Krinsky et al.
Table 3. Crystallographic Data for Compounds 2 (Synchrotron), 2 (Mo KR), 7, and 10
param 2 (synchro) 2 (Mo KR)
space group
7
10
P1h
P1h
P1h
C2/c
a, Å
b, Å
c, Å
R, deg
â, deg
γ, deg
V, Å3
Z
14.137(3)
14.262(2)
14.343(3)
76.192(4)
72.097(4)
60.491(4)
2381.5(9)
1
13.6126(17)
13.8895(17)
14.0722(17)
89.991(2)
76.981(2)
61.835(2)
2267.8(5)
1
13.148(1)
13.796(1)
16.005(1)
108.687(1)
96.277(1)
94.847(1)
2711.5(3)
1
24.989(2)
15.2762(14)
29.288(3)
90
99.8360(10)
90
11016.0(18)
4
F
calcd, g cm-3
1.875
1.952
1.936
6.43
1510
1.806
6.14
5820
µ(Mo KR), mm-1
F000
8.59a
7.27
1286
1278
cryst size, mm
T, K
0.05 × 0.05 × 0.04
0.23 × 0.16 × 0.05
0.22 × 0.14 × 0.06
0.20 × 0.17 × 0.08
193(2)
155(2)
168(2)
136(2)
θ range, deg
1.80-33.67
32 552
13 506
1.50-26.43
14 361
8933
1.36-26.39
15 509
10 598
3.32-24.72
27 405
9309
no. rflcns measd
no. indep rflcns
Rint
0.0600
0.0350
0.0196
0.0459
no. restraints, params
R1, wR2 b (all)
R1, wR2 (obsdc)
GOFd on F2
0, 583
0, 565
0, 617
0, 615
0.0520, 0.1384
0.0501, 0.1359
1.026
0.0726, 0.1069
0.0438, 0.0991
1.035
0.0422, 0.0741
0.0303, 0.0713
0.997
0.0673, 0.0939
0.0389, 0.0848
1.049
largest diff peak, hole, e Å-3
2.820, -2.150
3.041, -1.479
1.309, -1.065
1.809, -1.202
2
2
2
2
2
a λ ) 0.774 90 Å (synchrotron radiation). b R1 ) Σ||Fo| - |Fc||/Σ|Fo|; wR2 ) [Σw(Fo - Fc )2/ Σw(Fo )2]1/2, w ) 1/[σ2Fo + (ap)2 + bp], p ) [Fo
+
2Fc ]/3. c I > 2σ(I). d GOF ) [w(Fo - Fc )2/(n - p)]1/2
.
2
2
2
Table 4. Crystallographic Data for Compounds 12 and 13
param 12
space group
quency calculations performed on the optimized structures (same
level of theory) confirmed that they were true energetic minima.
Graphical representations of MOs were generated in and exported
from Maestro.
13
Cc
P21/n
a, Å
b, Å
c, Å
R, deg
â, deg
γ, deg
V, Å3
Z
22.873(3)
13.937(2)
26.999(4)
90
107.564(3)
90
8206(2)
4
1.324
8.4832(15)
17.131(3)
13.904(2)
90
102.723(2)
90
1971.0(5)
4
1.784
(PhN)14Ta6O (2). Pentakis(dimethylamido)tantalum (1.00 g, 2.49
mmol, 1 equiv) was dissolved in 5 mL of toluene. A solution of
aniline (1.41 g, 15.14 mmol, 6.08 equiv) in 10 mL of toluene was
added, followed by 7.5 µL (0.42 mmol, 0.16 equiv) of water. The
mixture was heated at reflux for 4 d, after which time the flask
was cooled to ambient temperature and the supernatant was
removed. The solid was washed with Et2O (2 × 10 mL) and then
extracted with THF (2 × 20 mL). The extracts were combined and
the solvent removed under vacuum, yielding 380 mg (0.158 mmol,
38%) of 2. Crystals suitable for X-ray diffraction analysis were
F
calcd, g cm-3
µ(Mo KR), mm-1
F000
1.47
3360
6.44
1024
cryst size, mm
T, K
0.22 × 0.19 × 0.16 0.12 × 0.08 × 0.05
165(2)
157(2)
θ range, deg
no. rflcns measd
no. indep rflcns
Rint
1.58-26.38
3.37-24.71
1
grown from concentrated solutions of benzene or THF. H NMR
23 045
8687
(500 MHz, THF-d8): δ 7.68 (dd, J ) 8.5 Hz, 1.0 Hz, 16H,
o-bridging Ar), 7.13 (m, 16H, m-bridging Ar), 6.98 (m, 12H,
m-terminal Ar), 6.92 (m, 8H, p-bridging Ar), 6.71 (m, 6H, p-terminal
Ar), 6.07 (dd, J ) 8.5 Hz, 1.0 Hz, 12H, o-terminal Ar). 13C{1H}
NMR (125.8 MHz, THF-d8): δ 157.84, 156.83, 129.80, 129.20,
128.35, 126.87, 125.27, 124.86. UV/vis (THF): λmax(onset) ) 280
nm (580 nm). Anal. Calcd for C84H70N14OTa6: C, 42.09; H, 2.94;
N, 8.18. Found: C, 42.23; H, 2.89; N, 8.86.
11 758
0.0651
3334
0.0257
no. restraints, params
R1, wR2 a (all)
R1, wR2 (obsdb)
GOFc on F2
2, 965
0, 223
0.0616, 0.0998
0.0461, 0.0941
0.912
0.0403, 0.0575
0.0256, 0.0544
1.114
largest diff peak, hole, e Å-3 1.660, -0.743
1.778, -0.981
a R1 ) Σ||Fo| - |Fc||/Σ|Fo|; wR2 ) [Σw(Fo2 - Fc )2/ Σw(Fo )2]1/2, w )
2
2
2
2
2
2
1/[σ2Fo + (ap)2 + bp], p ) [Fo + 2Fc ]/3. b I > 2σ(I). c GOF ) [w(Fo
2
- Fc )2/(n - p)]1/2
.
(p-MeC6H4N)14Ta6O (3). A Schlenk flask was charged with
pentakis(dimethylamido)tantalum (3.00 g, 7.47 mmol, 1.00 equiv),
Ta2O5 (200 mg, 0.452 mmol, 0.0605 equiv), and p-toluidine (2.41
g, 22.5 mmol, 3.01 equiv). Toluene (5 mL) was then added,
followed by 15 mL of hexamethyldisiloxane. The mixture was
heated at reflux for 4 d, after which time the flask was cooled to
ambient temperature and the supernatant was removed. The solid
was washed with Et2O (3 × 20 mL) and then extracted with 30
mL of hot THF. The THF solvent was removed under vacuum,
yielding 625 mg (0.243 mmol, 20%) of 3 in good purity. 1H NMR
(500 MHz, C6D6): δ 8.03 (d, J ) 8.5 Hz, 16H, o-bridging Ar),
6.87 (d, J ) 8.0 Hz, 12H, m-terminal Ar), 6.81 (d, J ) 8.5
Hz, 16H, m-bridging Ar), 6.47 (d, J ) 8.0 Hz, 12H, o-terminal
Ar), 2.08 (s, 18H, terminal ArMe), 1.96 (s, 24H, bridging ArMe).
13C{1H} NMR (125.8 MHz, C6D6): δ 155.49, 155.13, 134.25,
a set of d polarization functions for non-hydrogen atoms.50 Single
point energy calculations starting with the neutral cluster produced
excited-state electronic configurations. Thus, the wave function was
obtained by first calculating that of the dication using a level shift
of 0.5 au. This wave function was then used as an initial guess for
optimization of the neutral structure (level shift constraint removed).
Unrestricted wave functions were used for the computations
involving decamethylcobaltocene and its cation. Vibrational fre-
(50) (a) Ditchfield, R.; Hehre, W. J.; Pople, J. A. J. Chem. Phys. 1971, 54,
724-728. (b) Hehre, W. J.; Ditchfield, R.; Pople, J. A. J. Chem. Phys.
1972, 56, 2257-2261. (c) Hariharan, P. C.; Pople, J. A. Theor. Chim.
Acta 1973, 28, 213-222. (d) Francl, M. M.; Pietro, W. J.; Hehre, W.
J.; Binkley, J. S.; Gordon, M. S.; DeFrees, D. J.; Pople, J. A. J. Chem.
Phys. 1982, 77, 3654-3665.
1064 Inorganic Chemistry, Vol. 47, No. 3, 2008

Title Running Head: Hexatantalum Tetradecaimido Cluster Complexes
133.68, 130.18, 128.91, 126.78, 125.12, 21.17, 20.75. UV/vis
132.12, 128.19, 126.48, 119.94, 119.24. Anal. Calcd for
C84H56Br14N14OTa6: C, 28.98; H, 1.62; N, 5.63. Found: C, 29.14;
H, 1.51; N, 5.71.
(THF):
λmax(onset) ) 285 nm (600 nm). Anal. Calcd for
C98H98N14OTa6: C, 45.74; H, 3.84; N, 7.62. Found: C, 46.10; H,
4.08; N, 7.87.
(m-ClC6H4N)14Ta6O (7). Compound 1 (100 mg, 0.190 mmol,
1 equiv) was dissolved in 1 mL of benzene. m-Chloroaniline (129
mg, 0.950 mmol, 5 equiv) was treated with 0.6 µL (0.03 mmol,
0.16 equiv) of degassed water via microsyringe, and the resulting
mixture was dissolved in 0.5 mL benzene. The two solutions were
mixed, and the reaction mixture was transferred to a Kontes reaction
vessel. Hexane (4 mL) was then added, causing the formation of a
yellow precipitate. The vessel was sealed with a Teflon valve and
immersed in a 135 °C oil bath. After a few minutes, most of the
yellow solid had dissolved and the solution had developed a red
color. After being heated for ca. 24 h, the reaction vessel was
allowed to cool and was transferred to a glovebox where the solvent
was removed under vacuum. The orange oil was dissolved in 1
mL of toluene, into which pentane was allowed to diffuse. The
resulting crystalline material was insoluble in compatible (nonre-
active) solvents (13.4 mg, 0.0047 mmol, 15%). Anal. Calcd for
C84H56Cl14N14OTa6: C, 35.28; H, 1.97; N, 6.86. Found: C, 34.94;
H, 2.17; N, 6.81.
[Cp*2Co][(PhN)14Ta6O] (10). Compound 2 (50 mg, 0.021
mmol, 1.0 equiv) was dissolved in 5 mL of boiling THF, and a 2
mL THF solution of decamethylcobaltocene (8.2 mg, 0.025 mmol,
1.1 equiv) was added. The reaction mixture immediately developed
a deep, olive green color. Upon cooling to ambient temperature,
the supernatant was decanted from the crystals of 10‚4THF that
had formed (30 mg). Some of these crystals were of sufficient
quality for use in X-ray diffraction analysis. The supernatant was
heated to boiling, and 2 mL of pentane was slowly added. After
standing at -30 °C for 2 d, an additional crop of crystals was
recovered by decanting the supernatant and washing with Et2O (46
mg total, 0.015 mmol, 73%). 1H NMR (500 MHz, THF-d8): δ 8.33
(br, 6H, p-terminal Ar), 8.13 (br, 12H, o-terminal Ar), 6.15 (br,
16H, m-bridging Ar), 5.74, 12H, m-terminal Ar), 4.80 (br, 8H,
p-bridging Ar), 3.62 (v br, o-bridging Ar), 1.84 (v br, Cp*). Peak
assignments were determined by COSY NMR. Anal. Calcd for
(p-MeOC6H4N)14Ta6O (4). Compound 1 (100 mg, 0.190 mmol,
1 equiv) was dissolved in 1 mL of benzene. p-Anisidine (117 mg,
0.950 mmol, 5 equiv) was treated with 0.6 µL (0.03 mmol, 0.16
equiv) of degassed water via microsyringe, and the resulting mixture
was dissolved in 0.5 mL of benzene. The two solutions were mixed,
and the reaction mixture was transferred to a Kontes reaction vessel.
Hexane (4 mL) was then added, causing the formation of a yellow
precipitate. The vessel was sealed with a Teflon valve and immersed
in a 135 °C oil bath. After a few minutes, most of the yellow solid
had dissolved and the solution had developed a red color. After
being heated for ca. 24 h, the reaction vessel was allowed to cool
and was transferred to a glovebox. The supernatant was removed
by pipet, and the crystalline product was washed with Et2O (2 ×
2 mL) and then pentane (2 × 2 mL). Isotopically enriched samples
were prepared in an identical fashion using 17O/18O-enriched water.
1
Yield ) 43.4 mg, 0.016 mmol, 49%. H NMR (500 MHz, C6D6):
δ 8.19 (d, J ) 9.0 Hz, 16H), 6.74 (d, J ) 9.0 Hz, 16H), 6.63 (s,
24H), 3.12 (s, 18H), 3.09 (s, 24H). 13C{1H} NMR (125.8 MHz,
C6D6): δ 157.53, 157.45, 151.70, 151.63, 127.96, 126.22, 114.90,
113.75, 55.34, 55.17. UV/vis (THF): λmax(onset) ) 300 nm (630
nm). The product was recrystallized by pentane vapor diffusion
into a benzene solution of 4 prior to submission for EA. Anal. Calcd
for C98H98N14O15Ta6: C, 42.07; H, 3.53; N, 7.01. Found: C, 42.62;
H, 3.74; N, 6.90.
(p-t-BuC6H4N)14Ta6O (5). A Schlenk flask was charged with
pentakis(dimethylamido)tantalum (3.00 g, 7.47 mmol, 1.00 equiv)
and Ta2O5 (200 mg, 0.452 mmol, 0.0605 equiv). n-Octane (20 mL)
was then added, followed by p-tert-butylaniline (3.36 g, 22.5 mmol,
3.01 equiv). The mixture was heated at reflux for 4 d, during which
time the yellow suspension became red in color. The supernatant
was removed by filtration, and the orange solid was washed with
cold pentane (40 mL). The residue was then extracted with Et2O
(3 × 10 mL). The extracts were combined and evaporated under
reduced pressure, yielding a red solid. After being washed with 10
mL of pentane, the remaining solid was dissolved in 2 mL of Et2O
and placed in a vial which was nested in a larger vial containing
pentane. After 2 d, the supernatant was removed by pipet and the
pure 5 that had crystallized was washed with 5 mL of pentane (65
mg, 0.020 mmol, 1.6%). 1H NMR (500 MHz, THF-d8): δ 7.62 (d,
J ) 8.5 Hz, 16H), 7.18 (d, J ) 8.5 Hz, 16H), 6.96 (d, J ) 8.5 Hz,
12H), 5.98 (d, J ) 8.5 Hz, 12H), 1.25 (s, 72H), 1.17 (s, 54H).
13C{1H} NMR (125.8 MHz, THF-d8): δ 155.11, 154.93, 147.08,
146.82, 126.40, 126.19, 124.92, 124.56, 34.88, 34.68, 31.86, 31.84.
Anal. Calcd for C140H182N14OTa6: C, 53.17; H, 5.80; N, 6.20.
Found: C, 53.18; H, 5.98; N, 6.32.
C104H100CoN14OTa6‚4C4H8O: C, 47.80; H, 4.41; N, 6.50. Found:
C, 47.86; H, 4.36; N, 6.62.
[Cp*2Co][(p-MeC6H4N)14Ta6O] (11). Compound 3 (100 mg,
0.0388 mmol, 1.00 equiv) was dissolved in 10 mL of hot toluene.
A solution of decamethylcobaltocene (20 mg, 0.061 mmol, 1.5
equiv) in 3 mL of toluene was then added, resulting in a dark brown-
green solution from which 11 began to precipitate upon cooling to
ambient temperature. The reaction vessel was cooled at -30 °C
overnight. The green supernatant was removed by filtration, and
the amorphous product (93 mg, 0.032 mmol, 82%) was washed
with toluene (4 × 10 mL) and then 10 mL of pentane. Prior to
submission for elemental analysis, material was crystallized by
1
diffusion of pentane vapor into a THF solution of 11. H NMR
(p-BrC6H4N)14Ta6O (6). Pentakis(dimethylamido)tantalum (2.00
g, 4.98 mmol, 1.00 equiv) and p-bromoaniline (2.58 g, 15.0 mmol,
3.01 equiv) were mixed and dissolved in 10 mL of toluene. Water
(15 µL, 0.84 mmol, 0.17 equiv) was then added by microsyringe,
and the mixture was heated with stirring for 24 h. The reaction
mixture was stirred while cooling to ambient temperature, resulting
in a fine orange precipitate. The solid was separated by filtration
and washed with 100 mL of cold toluene. This residue was then
extracted with hot toluene (5 × 10 mL), and each extract was
cooled to -30 °C overnight, each yielding a small crop of crystal-
(500 MHz, THF-d8): δ 7.97 (br, 12H, o-terminal Ar), 6.00 (br,
16H, m-bridging Ar), 5.56 (br, 12H, m-terminal Ar), 5.35 (s, 24H,
bridging ArMe), 3.54 (v br, o-bridging Ar), 2.42 (v br, Cp*), 0.31
(s, 18 H, terminal ArMe). Assignments were determined by COSY
NMR and comparison with spectra of 10. µeff ) 1.786 µB (295 K).
Anal. Calcd for C118H128CoN14OTa6: C, 48.82; H, 4.44; N, 6.75.
Found: C, 49.02; H, 4.79; N, 6.81.
Reaction of 3 with Benzophenone. Compound 3 (100 mg,
0.0388 mmol, 1.00 equiv) was dissolved in 10 mL of benzene.
Benzophenone (99.0 mg, 0.543 mmol, 14.0 equiv) was then added,
and the reaction mixture was transferred to a Pyrex reaction vessel
fitted with a vacuum stopcock and heated at 135 °C for 24 h. Upon
cooling, the white precipitate that had formed was removed by
1
line 6 (127 mg combined, 0.0365 mmol, 4.4%). H NMR (500
MHz, THF-d8): δ 7.46 (d, J ) 9.0 Hz, 16H), 7.42 (d, J ) 9.0
Hz, 16H), 7.29 (d, J ) 9.0 Hz, 12H), 6.02 (d, J ) 9.0 Hz, 12H).
13C{1H} NMR (125.8 MHz, THF-d8): δ 155.58, 155.09, 133.58,
Inorganic Chemistry, Vol. 47, No. 3, 2008 1065

Krinsky et al.
filtration and the solvent was removed from the filtrate under
vacuum, yielding a yellow oil. A crystalline solid (127 mg, 0.468
mmol, 86%) was obtained after dissolution in and subsequent
removal of pentane. Analysis of this material by 1H NMR showed
only benzophenone p-tolylketimine; its NMR and EIMS data
matched those in the literature.51
visible in the 1H NMR spectrum. A fine white precipitate had also
formed at the bottom of the vessel. The reaction mixture was
filtered, and the solvent was removed under vacuum. 1H NMR and
EIMS spectra of the major product matched those in the literature
for acetophenone p-tolylimine.52
Acknowledgment. We are indebted to Dr. Fred Hollander
and Dr. Allen Oliver at the UC Berkeley CHEXRAY X-ray
crystallographic facility and Dr. Kathleen Durkin at the
Molecular Graphics Facility (NSF Grant CHE-0233882) for
crystallographic and computational consultation, respectively,
as well as Dr. John Greaves at UC, Irvine, CA, for mass
spectrometry data. We also thank Dr. Alexandr Shafir for
providing a preliminary structure of 2 using Mo KR radiation.
This work was supported by NIH Grant GM-25459 to R.G.B.
and DOE funding to J.A.
Reaction of 3 with Benzaldehyde. Compound 3 (10 mg, 0.0039
mmol, 1.0 equiv) was dissolved in 0.5 mL of C6D6, and 5.5 µL
(0.054 mmol, 14 equiv) of benzaldehyde was added by micro-
syringe. The mixture was transferred to an NMR tube and flame
sealed under reduced pressure. The solution became orange in color
1
after about 5 min, and the H NMR spectrum obtained after 24 h
showed that nearly all of the aldehyde had been converted to
benzaldehyde p-tolylimine. Heating the NMR tube at 75 °C for 2
h converted the last traces of aldehyde. Additionally, a fine white
precipitate had formed at the bottom of the vessel. The reaction
mixture was then filtered, and the solvent was removed under
1
vacuum. H NMR and EIMS data matched the literature values
Supporting Information Available: CIF-format files for all
crystallographically determined structures, experimental procedures
for alternate syntheses of 2, 3, and 7 as well as syntheses of
12-14, additional ESI mass spectra of 3, 3(17,18O), 4, and 4(17,18O),
and positional coordinates of all computed structures. This material
is available free of charge via the Internet at http:/pubs.acs.org.
for benzaldehyde p-tolylimine.51
Reaction of 3 with Acetophenone. Compound 3 (10 mg, 0.0039
mmol, 1.0 equiv) was dissolved in 0.5 mL of C6D6, and 6.3 µL
(0.054 mmol, 14 equiv) of acetophenone was added by micro-
syringe. The mixture was transferred to an NMR tube and flame
sealed under reduced pressure. After 3 d at ambient temperature,
only resonances corresponding to acetophenone p-tolylimine (75
( 5% yield by 1H NMR) and one other, unidentified species were
IC701920V
(52) Barluenga, J.; Fernandez, M. A.; Aznar, F.; Valdes, C. Chem.sEur.
J. 2004, 10, 494-507.
(51) Nongkunsarn, P.; Ramsden, C. A. Tetrahedron 1997, 53, 3805-3830.
1066 Inorganic Chemistry, Vol. 47, No. 3, 2008