Tris(3,5-di-tert-butylcatecholato)molybdenum(VI): Lewis acidity and nonclassical oxygen atom transfer reactions

Article abstract of DOI:10.1021/ic401736f

In the solid state, tris(3,5-di-tert-butylcatecholato)molybdenum(VI) forms a dimer with seven-coordinate molybdenum and bridging catecholates. NMR spectroscopy indicates that the dimeric structure is retained in solution. The molybdenum center has a high affinity for Lewis bases such as pyridine or pyridine-N-oxide, forming seven-coordinate monomers with a capped octahedral geometry, as illustrated by the solid-state structure of (3,5- ^tBu₂Cat)₃Mo(py). Structural data indicate that the complexes are best considered as Mo(VI) with substantial π donation from the nonbridging catecholates to molybdenum. Both the dimeric and the monomeric tris(catecholates) react rapidly with water to form free catechol and oxomolybdenum bis(catecholate) complexes. Monooxomolybdenum complexes are also obtained, more slowly, on reaction with dioxygen, with organic products consisting mostly of 3,5-di-tert-butyl-1,2-benzoquinone with minor amounts of the extradiol oxidation product 4,6-di-tert-butyl-1-oxacyclohepta-4,6-diene-2,3- dione. The pyridine-N-oxide complex reacts on heating (with excess pyO) to form initially (3,5-^tBu₂Cat)₂MoO(Opy) and ultimately MoO₃(Opy), with quinone and free pyridine as the only organic products. The decay of (3,5-^tBu₂Cat)₃Mo(Opy) shows an accelerated, autocatalytic profile because the oxidation of its product, (3,5-^tBu₂Cat)₂MoO(Opy), produces an oxo-rich, catecholate-poor intermediate which rapidly conproportionates with (3,5-^tBu₂Cat)₃Mo(Opy), providing an additional pathway for its conversion to the mono-oxo product. The tris(catecholate) fragment Mo(3,5-^tBu₂Cat)₃ deoxygenates Opy in this nonclassical oxygen atom transfer reaction slightly less rapidly than does its oxidized product, MoO(3,5-^tBu₂Cat)₂.

Full text of DOI:10.1021/ic401736f

Article
pubs.acs.org/IC
Tris(3,5-di-tert-butylcatecholato)molybdenum(VI): Lewis Acidity and
Nonclassical Oxygen Atom Transfer Reactions
Amanda H. Randolph, Nicholas J. Seewald, Karl Rickert, and and Seth N. Brown*
Department of Chemistry and Biochemistry, University of Notre Dame, 251 Nieuwland Science Hall, Notre Dame, Indiana
46556-5670, United States
S
* Supporting Information
ABSTRACT: In the solid state, tris(3,5-di-tert-butylcatecholato)-
molybdenum(VI) forms a dimer with seven-coordinate molybdenum
and bridging catecholates. NMR spectroscopy indicates that the dimeric
structure is retained in solution. The molybdenum center has a high
affinity for Lewis bases such as pyridine or pyridine-N-oxide, forming
seven-coordinate monomers with a capped octahedral geometry, as
illustrated by the solid-state structure of (3,5-tBu2Cat)3Mo(py).
Structural data indicate that the complexes are best considered as
Mo(VI) with substantial π donation from the nonbridging catecholates
to molybdenum. Both the dimeric and the monomeric tris(catecholates)
react rapidly with water to form free catechol and oxomolybdenum
bis(catecholate) complexes. Monooxomolybdenum complexes are also
obtained, more slowly, on reaction with dioxygen, with organic products consisting mostly of 3,5-di-tert-butyl-1,2-benzoquinone
with minor amounts of the extradiol oxidation product 4,6-di-tert-butyl-1-oxacyclohepta-4,6-diene-2,3-dione. The pyridine-N-
oxide complex reacts on heating (with excess pyO) to form initially (3,5-tBu2Cat)2MoO(Opy) and ultimately MoO3(Opy), with
quinone and free pyridine as the only organic products. The decay of (3,5-tBu2Cat)3Mo(Opy) shows an accelerated, autocatalytic
profile because the oxidation of its product, (3,5-tBu2Cat)2MoO(Opy), produces an oxo-rich, catecholate-poor intermediate
which rapidly conproportionates with (3,5-tBu2Cat)3Mo(Opy), providing an additional pathway for its conversion to the mono-
oxo product. The tris(catecholate) fragment Mo(3,5-tBu2Cat)3 deoxygenates Opy in this nonclassical oxygen atom transfer
reaction slightly less rapidly than does its oxidized product, MoO(3,5-tBu2Cat)2.
In contrast, Cass and Pierpont reported in 1986 that tris(3,5-
di-tert-butylcatecholato)molybdenum(VI) reacts avidly with
dioxygen to give 3,5-di-tert-butyl-1,2-benzoquinone and the
o x o b i s ( c a t e c h o l a t o ) m o l y b d e n u m ( V I ) d i m e r
Mo2O2(3,5-tBu2Cat)4.7 This represents a formal four-electron
nonclassical redox reaction, and we were thus interested in
investigating its mechanism. Here we report the results of our
study of the chemistry of tris(3,5-di-tert-butylcatecholato)-
molybdenum(VI). We find that this fragment is quite Lewis
acidic, with isolable compounds invariably displaying seven-
coordination at molybdenum. Interaction with a simple oxygen
atom transfer agent such as pyridine-N-oxide does result in
clean oxidation of catecholate to quinone, but reactions with
dioxygen are more complex.
INTRODUCTION
■
Catecholates are electron-rich ligands whose redox activity,
specifically oxidation to semiquinones and quinones, forms a
cornerstone of their chemistry.1 Further oxidation involving C−
C bond cleavage is also possible. For example, catechol-1,2-
dioxygenases cleave the bond either between the carbons
bonded to oxygen (intradiol oxidation) or adjacent to the
catechol moiety (extradiol oxidation).2 Poly(catecholate)
complexes can be efficient catalysts for the aerobic oxidation
of catechols, with mixtures of quinones as well as intradiol and
extradiol oxidation products being formed.3,4
Catecholate ligands can potentially be applied as electron
reservoirs to enable redox chemistry at normally redox-inert
metal centers.5 Such inner-sphere redox events, in which the
electrons originate in the ligands but the bonding changes take
place at the metal center, have been referred to as “non-
classical.” We recently elucidated an example of such a
nonclassical oxygen atom transfer reaction in the oxidation of
oxomolybdenum(VI) bis(3,5-di-tert-butylcatecholate) com-
plexes using pyridine-N-oxide.6 Dioxygen is of considerable
interest as an oxygen atom transfer reagent in these kinds of
reactions, but (3,5-tBu2Cat)2MoO(L) complexes are inert to
O2.6
EXPERIMENTAL SECTION
■
Unless otherwise noted, all procedures were carried out under an inert
atmosphere in a nitrogen-filled glovebox or on a vacuum line. When
dry solvents were needed, chlorinated solvents and acetonitrile were
dried over 4 Å molecular sieves, followed by CaH2. Benzene and
toluene were dried over sodium. Anhydrous pyridine was purchased
from Aldrich and stored in the drybox before use without further
Received: July 5, 2013
© XXXX American Chemical Society
A
dx.doi.org/10.1021/ic401736f | Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
drying. Deuterated solvents were obtained from Cambridge Isotope
Laboratories, dried using the same procedures as their protio
analogues, and stored in the drybox prior to use. Pyridine-N-oxide
(Acros) was sublimed under reduced pressure and stored in the
drybox. All other reagents were commercially available and used
without further purification. NMR spectra were measured on Varian
VXR-300, Varian VXR-500, or Bruker DPX-400 spectrometers.
Chemical shifts for 1H and 13C{1H} spectra are reported in ppm
downfield of TMS, with spectra referenced using the known chemical
shifts of the solvent residuals. Infrared spectra (4000−600 cm−1) were
recorded as Nujol mulls on a JASCO FT/IR-6300 spectrometer.
Elemental analyses were performed by Midwest Microlab, LLC
(Indianapolis, IN).
Generation of Alternative Isomer of Mo2O2(3,5-tBu2Cat)4
(1b) by Reaction of 3,5-tBu2CatH2 with MoO2(acac)2. In a
screw-cap NMR tube in the drybox, 19.0 mg of MoO2(acac)2 (Strem,
58.3 μmol) and 26.3 mg of 3,5-di-tert-butylcatechol (Aldrich, 118
μmol, 2.03 equiv) were dissolved in 0.6 mL of CDCl3. The dark purple
solution contained roughly equal quantities of the Buchanan−Pierpont
isomer of Mo2O2(3,5-tBuCat)4 (1a)8 (1H NMR of 1a: δ 0.90, 1.26,
t
1.29, 1.53 [s, 18H ea., Bu], 6.39, 6.77, 7.03, 7.19 [d, 2 Hz, 2H ea.,
4
1
ArH]) and a previously unobserved isomer of the dimer (1b). H
t
NMR of 1b (CDCl3): δ 0.98, 1.28, 1.31, 1.54 (s, 18H ea., Bu), 6.67,
6.80, 7.11, 7.20 (d, 2 Hz, 2H ea., ArH).
Variable-Temperature NMR Spectroscopy of
(3,5-tBu2Cat)3Mo(py). NMR spectra were recorded between −70°
and +50 °C in toluene-d8 on a Varian VXR-500 NMR spectrometer.
To calculate exchange rates, the pyridine resonances were simulated
using the program gNMR,9 and the calculated lineshapes were
superimposed on the observed spectra. Activation parameters for the
exchange reaction were calculated by plotting ln(kdiss/T) vs 1/T.
Kinetics of Reactions of (3,5-tBu2Cat)3Mo(Opy) in Toluene-
d8. In a typical reaction, 6.4 mg of Mo2(3,5-tBu2Cat)6 and 3.6 mg of
1,4-bis(trimethylsilyl)benzene (used as an internal standard) were
dissolved in 0.70 mL of a stock 0.102 M solution of pyridine-N-oxide
in C6D5CD3 in a screw-cap NMR tube in the drybox. (Use of a greater
excess of pyridine-N-oxide is precluded by its limited solubility in
toluene.) The tube was sealed with a Teflon-lined cap and stored in a
−20 °C freezer until ready for use. The Varian VXR-500 probe was
preheated to 70 °C with a dummy sample, then the actual sample tube
was inserted into the preheated probe. After locking and shimming,
spectra were acquired every 20 min for 24 h, and concentrations of the
various species were determined by integration of the tert-butyl
resonances against the trimethylsilyl peak of the internal standard.
Concentration vs time profiles were fit to the Supporting Information,
eq S5 (for [(3,5-tBu2Cat)3Mo(Opy)]) and Supporting Information,
eqs S6−S7 (for [(3,5-tBu2Cat)2MoO(Opy)]) using the program
Kaleidagraph (Synergy Software, v. 4.1). In the data fitting, the total
Mo concentration ([(3,5-tBu2Cat)3Mo(Opy)] + [(3,5-tBu2Cat)2MoO-
(Opy)] before the disappearance of (3,5-tBu2Cat)3Mo(Opy)) was
fixed at its average value over this time period, with the initial
concentration of (3,5-tBu2Cat)3Mo(Opy) (Mo0) and the rate
constants k1 and k2 the only adjustable parameters in the fits.
X-ray Crystallography. Crystals of Mo2(3,5-tBu2Cat)6·4 C6D6
were grown by slowly cooling the reaction mixture of 3,5-di-tert-butyl-
1,2-benzoquinone with Mo(CO)6 in C6D6 in a sealed NMR tube.
Crystals of (3,5-tBu2Cat)3Mo(py)·CH3CN were grown by diffusion of
acetonitrile into a benzene solution of the complex. Crystals were
placed in Paratone oil before being transferred to the cold N2 stream
of a Bruker Apex II CCD diffractometer in a nylon loop. Data were
reduced, correcting for absorption, using the program SADABS. The
structures were solved using a Patterson map to find the molybdenum
atoms and the atoms in the first coordination sphere, with other heavy
atoms found on subsequent difference Fourier maps. All nonhydrogen
atoms were refined anisotropically.
Hexakis(3,5-di-tert-butylcatecholato)dimolybdenum(VI),
Mo2(3,5-tBu2Cat)6. Method A. From Tricarbonyl(cycloheptatriene)-
molybdenum(0). In the drybox, (cycloheptatriene)molybdenum
tricarbonyl (0.325 g, 1.19 mmol, Aldrich) was dissolved in 18 mL of
benzene. In a separate vial, 3,5-di-tert-butyl-1,2-benzoquinone (0.821
g, 3.73 mmol, Aldrich) was dissolved in 4 mL of benzene. The
solutions were added to a 50 mL Erlenmeyer flask, and the solution
was stirred for 1 h. On mixing, the solution frothed vigorously for 5
min and turned dark blue, depositing a fine blue precipitate. The solid
was filtered on a glass frit and washed with hexanes to give 0.734 g of
1
Mo2(3,5-tBu2Cat)6 (81%) as fine blue microcrystals. H NMR (300
t
t
MHz, CDCl3, 20 °C): δ 1.29 (s, 54H, Bu), 1.38 (s, 54H, Bu), 6.77
(br s, 6H, ArH), 7.38 (br s, 6H, ArH). IR (cm−1): 1584 (s), 1549 (m),
1409 (w), 1395 (w), 1385 (m), 1362 (s), 1298 (m), 1254 (m), 1226
(s), 1201 (s), 1187 (m), 1175 (w), 1100 (m), 1029 (m), 993 (s), 978
(s), 915 (m), 875 (w), 856 (s), 837 (s), 813 (w), 774 (w), 767 (w),
754 (m), 745 (w), 692 (s), 655 (m), 631 (s). The analytical sample
retained two moles of benzene per dimer. Anal. Calcd for
C96H132Mo2O12: C, 69.04; H, 7.97. Found: C, 68.53; H, 7.50.
Method B. From Molybdenum Hexacarbonyl. In the drybox,
Mo(CO)6 (0.598 g, 2.27 mmol, Aldrich) and 3,5-di-tert-butyl-1,2-
benzoquinone (1.506 g, 6.84 mmol, Aldrich) were added to a glass
bomb with a Teflon valve. The valve was sealed and the bomb
removed from the drybox. Dry benzene (20 mL) was added to the
bomb by vacuum transfer. The solution was heated in an 85 °C oil
bath, with stirring, for 24 h. The glass bomb was allowed to cool to
room temperature and brought back into the drybox. The solution was
filtered through a glass frit, and the solid was washed with hexanes and
dried under vacuum 45 min to furnish large dark blue crystals of
Mo2(3,5-tBu2Cat)6 (0.613 g, 36%).
Tris(3,5-di-tert-butylcatecholato)(pyridine)molybdenum(VI),
(3,5-tBu2Cat)3Mo(py). In the drybox, Mo2(3,5-tBu2Cat)6 (98.5 mg,
0.130 mmol Mo) was dissolved in dry acetonitrile (2.0 mL) containing
excess pyridine (32.3 μL). After one week, large bronze crystals of
(3,5-tBu2Cat)3Mo(py) were isolated by filtration through a glass frit.
After washing with acetonitrile, the yield of the pyridine adduct was
t
42.0 mg (39%). 1H NMR (C6D6): δ 1.26 (s, 27H, Bu), 1.38 (s, 27H,
tBu), 6.53 (t, 7 Hz, 2H, py 3,5-H), 6.80 (t, 7 Hz, 1H, py 4-H), 7.09 (s,
3H, ArH), 7.16 (s, 3H, ArH), 9.15 (d, 5 Hz, 2H, py 2,6-H). 13C{1H}
NMR (C6D6): δ 30.40, 32.26 (C(CH3)3), 35.12, 35.39 (C(CH3)3),
111.35, 118.87, 124.53, 137.07, 139.21, 147.82, 151.08, 160.89, 163.43.
IR (cm−1): 1587 (s), 1552 (w), 1362 (s), 1298 (s), 1252 (w), 1232
(s), 1209 (s), 1100 (w), 1069 (m), 1022 (m), 983 (s), 938 (w), 913
(w), 850 (w), 835 (w), 761 (m), 722 (w). Anal. Calcd for
C47H65MoNO6: C, 67.53; H, 7.84; N, 1.68. Found: C, 67.37; H,
7.64; N, 1.90.
The tert-butyl group in (3,5-tBu2Cat)3Mo(py)·CH3CN centered on
C18 was disordered in two different orientations. The disorder was
modeled by constraining the thermal parameters of opposite carbons
to be equal and allowing the two orientations’ occupancies to refine
(the major component of the disorder was found to have an
occupancy of 80.0(6)%). Hydrogen atoms in the disordered tert-butyl
group of (3,5-tBu2Cat)3Mo(py)·CH3CN were placed in calculated
positions. All other hydrogen (and deuterium) atoms were found on
difference Fourier maps and refined isotropically. Calculations used
SHELXTL (Bruker AXS),10 with scattering factors and anomalous
dispersion terms taken from the literature.11 Further details about the
structures are given in Table 1.
Tris(3,5-di-tert-butylcatecholato)(pyridine-N-oxide)-
molybdenum(VI), (3,5-tBu2Cat)3Mo(Opy). This compound was
generated in situ by addition of pyridine-N-oxide (3.4 mg) to
Mo2(3,5-tBu2Cat)6 (20.8 mg) in 0.7 mL of deuterated solvent. 1H
t
t
NMR (C6D6): δ 1.29 (s, 27H, Bu), 1.43 (s, 27H, Bu), 6.21 (t, 7 Hz,
2H, Opy 3,5-H), 6.41 (t, 8 Hz, 1H, Opy 4-H), 6.93 (d, 2 Hz, 3H,
ArH), 7.12 (d, 2 Hz, 3H, ArH), 8.27 (d, 6 Hz, 2H, Opy 2,6-H).
13C{1H} NMR (C6D6): δ 30.52, 32.40 (C(CH3)3), 35.05, 35.43
(C(CH3)3), 110.45, 117.93, 125.38 (Opy 3,5-C), 128.68 (Opy 4-C),
136.65, 142.88 (Opy 2,6-C), 146.93, 161.02, 162.73.
RESULTS
■
Preparation and Structure of Homoleptic
Molybdenum(VI) Tris(3,5-di-tert-butylcatecholate). Cass
and Pierpont reported in 1986 that Mo(3,5-tBu2Cat)3 is formed
B
dx.doi.org/10.1021/ic401736f | Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
Table 1. Crystal data for Mo2(3,5-tBu2Cat)6·4C6D6 and
(3,5-tBu2Cat)3Mo(py)·CH3CN
Mo2(3,5-tBu2Cat)6·
(3,5-tBu2Cat)3Mo(py)·
4C6D6
CH3CN
empirical formula
temperature (K)
λ (Å)
C108H120D24Mo2O12
100(2)
C49H68MoN2O6
120(2)
0.71073 (Mo Kα)
0.71073 (Mo Kα)
space group
total data collected
no. of indep reflns.
Rint
P1
P1
̅
27516
̅
28023
9860
9535
0.0546
0.0556
obsd. refls. [I > 2σ(I)]
a (Å)
7680
7159
13.9365(12)
14.7478(13)
15.1427(14)
116.0609(17)
91.5854(16)
114.6777(15)
2454.1(4)
1
10.5873(17)
12.436(2)
18.282(3)
91.913(4)
95.778(4)
101.514(4)
2343.1(7)
2
b (Å)
c (Å)
α (deg)
β (deg)
γ (deg)
V (Å3)
Z
cryst size (mm)
no. refined params.
R indices [I > 2σ(I)]
0.13 × 0.10 × 0.03
838
0.41 × 0.12 × 0.10
805
Figure 1. Thermal ellipsoid plot (50% ellipsoids) of the metal complex
in Mo2(3,5-tBu2Cat)6·4C6D6. Hydrogen atoms are omitted for clarity.
R1 = 0.0417,
wR2 = 0.0784
R1 = 0.0634,
wR2 = 0.0864
1.017
R1 = 0.0419,
wR2 = 0.0834
R1 = 0.0701,
wR2 = 0.0929
1.012
O31 = 2.0932(17) Å, Mo−O31A = 2.1108(17) Å, Table 2).
The same connectivity is observed in the bridging catecholates
of the oxomolybdenum dimer Mo2O2(3,5-tBu2Cat)4, but in the
oxomolybdenum complex the bridging catecholates are steeply
inclined with respect to the Mo2O2 plane (124.2° angle
between planes).8 In contrast, the bridging catecholates of the
homoleptic dimer are roughly in the Mo2O2 plane (29.5° angle
between planes), a much more typical arrangement for metal
complexes containing a bis(μ-3,5-di-tert-butylcatecholate)
motif.13
R indices (all data)
goodness of fit
on heating Mo(CO)6 with 3,5-di-tert-butyl-1,2-benzoquinone
in toluene under anaerobic conditions.7 We find that this
reaction is conveniently carried out in benzene, from which
dark blue crystals of a benzene solvate of the compound deposit
on cooling (eq 1). The same compound forms immediately at
Seven-coordinate tris(chelate) complexes of molybdenum or
tungsten with small bite-angle chelates such as 2-mercaptopyr-
idine14 or dithiocarbamate15 adopt structures that are well
described as pentagonal bipyramids with two equatorial
chelates and one chelate that spans the equatorial and axial
positions. An oxomolybdenum catecholate complex MoO-
(C5H10NO)2(cat) with a nominally pentagonal bipyramidal
geometry, if the η2-piperidinolato ligands are counted as
occupying two coordination sites, has also been reported.16 The
molybdenum in Mo2(3,5-tBu2Cat)6 does not adopt a
pentagonal bipyramidal geometry, possibly because of the
somewhat larger bite angle of catecholate (∼76°) compared
with that of dithiocarbamate (∼70°) or 2-mercaptopyridine
(∼65°), or piperidinolate (∼35°). The geometry of the Mo
center in the dimer is rather irregular and is not well described
by any of the usual idealized seven-coordinate geometries
(capped octahedron, capped trigonal prism, 4:3 piano stool).17
High-valent molybdenum complexes usually have oxo or
other strong π-donor ligands, a class to which catecholates
potentially belong.18 Since π donation depletes electron density
in the catecholate highest occupied molecular orbital
(HOMO), just as ligand oxidation to the semiquinone does,
the extent of π donation in catecholates can be assessed by
calculating an apparent “metrical” oxidation state (MOS) of the
ligand based on the carbon−carbon and carbon−oxygen
distances.19 A noninteger MOS is particularly indicative of
partial transfer of electron density via π donation, rather than
outright electron transfer. Noninteger MOS values are observed
in the chelating ligands in Mo2(3,5-tBu2Cat)6, which have MOS
room temperature if the quinone is allowed to react with (η6-
cycloheptatriene)molybdenum tricarbonyl, and precipitates
from benzene as a fine powder. The low reactivity of Mo(CO)6
as a Mo(0) source has been documented previously in the
formation of a molybdenum tris(catecholate) complex from
1,10-phenanthroline-5,6-dione, where Mo(η6-C6H5CH3)2 was
found to be more reactive than Mo(CO)6.12
While it was originally formulated as a six-coordinate
monomer,7 the tris(catecholate) complex in fact exists in the
solid state as a seven-coordinate dimer with a pair of bridging
catecholates (Figure 1). Each bridging catecholate has the
oxygen ortho to the tert-butyl group (O32) bonded solely to
one molybdenum, with the less hindered oxygen bridging the
molybdenum atoms in a relatively symmetric fashion (Mo−
C
dx.doi.org/10.1021/ic401736f | Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
C3-symmetric octahedral isomer and three for the C1-symmetric
isomer), the large number of distinct environments observed
strongly suggests that a dimeric structure is largely retained in
solution. (While the solid-state structure would give only three
catecholate environments, isomerism involving different
orientations of the nonbridging catecholates is likely and
could give rise to up to 16 stereoisomers.) The spectrum in
CD2Cl2 is sharp at −30 °C and essentially unchanged upon
further cooling, which makes the presence of significant
amounts of monomer unlikely at these temperatures, since
the monomer−dimer equilibrium would be expected to be
highly temperature-dependent. The observed coalescence of
resonances into a single catecholate environment at temper-
atures above 30 °C (at 500 MHz) indicates that dissociation to
monomers is likely thermally accessible.
Table 2. Selected Bond Distances (Å) and Angles (deg) for
the First Coordination Spheres in Mo2(3,5-tBu2Cat)6·4C6D6
and (3,5-tBu2Cat)3Mo(py)·CH3CN
Mo2(3,5-tBu2Cat)6·4
(3,5-tBu2Cat)3Mo(py)·
C6D6
CH3CN
Mo−O11
Mo−O12
Mo−O21
Mo−O22
Mo−O31
Mo−O32
Mo−O31A
Mo−N
1.9729(17)
2.0017(16)
1.9876(17)
1.9827(16)
2.0932(17)
1.9617(17)
2.1113(17)
1.9767(19)
1.9792(18)
2.0030(18)
1.9731(18)
1.9893(18)
1.9987(18)
2.274(2)
O11−Mo−O12
O21−Mo−O22
O31−Mo−O32
O11−Mo−O21
O11−Mo−O22
O11−Mo−O31
O11−Mo−O32
O11−Mo−O31A
O11−Mo−N
O12−Mo−O21
O12−Mo−O22
O12−Mo−O31
O12−Mo−O32
O12−Mo−O31A
O12−Mo−N
O21−Mo−O31
O21−Mo−O32
O21−Mo−O31A
O21−Mo−N
O22−Mo−O31
O22−Mo−O32
O22−Mo−O31A
O22−Mo−N
76.76(7)
75.67(7)
75.53(7)
90.98(7)
84.19(7)
156.80(7)
81.75(7)
133.23(7)
77.06(7)
76.07(7)
75.98(7)
84.11(7)
110.86(8)
157.91(7)
109.32(7)
Reactions of Mo2(3,5-tBu2Cat)6 with Lewis Bases.
Addition of pyridine or pyridine-N-oxide to solutions of
Mo2(3,5-tBuCat)6 results in an immediate color change from
dark blue-violet to dark royal blue (pyridine) or electric blue
1
(pyridine-N-oxide). H NMR spectra reveal new signals due to
bound pyridine or pyridine-N-oxide (in a mole ratio of 1:3
relative to the catecholates), and aromatic doublets and tert-
butyl singlets due to a single type of 3,5-di-tert-butylcatecholate,
suggesting the formation of monomeric adducts
(3,5-tBu2Cat)3Mo(L) (eq 2). NMR spectra at low temperature
(−70 °C) show little broadening of the catecholate signals,
indicating that the adducts formed are either symmetrical or
highly fluxional.
76.86(8)
81.28(7)
154.86(7)
82.98(7)
80.15(7)
160.41(7)
117.39(7)
104.23(7)
81.90(7)
73.99(6)
131.08(8)
83.59(7)
153.82(7)
81.09(7)
81.25(7)
124.77(7)
135.39(8)
83.78(7)
117.01(7)
114.42(7)
152.67(7)
78.01(7)
73.83(8)
124.26(8)
70.62(7)
O31−Mo−O31A
O31−Mo−N
O32−Mo−O31A
O32−Mo−N
67.24(7)
128.08(7)
112.76(7)
Mo−O31−Mo0A
values of −1.66(12) and −1.67(14) for rings 1 and 2,
respectively. The bridging catecholate shows bond distances
typical of fully reduced catecholates with little π bonding, MOS
= −2.00(16). The nonbridging catecholates are similar to those
trans to oxo in MoO(cat)2(L) (cat = 3,6-tBu2Cat, L = ONC5H5,
OSMe2, OAsPh3, MOS = −1.63(7);20 cat = 3,5-tBu2Cat, L =
ONC5H4-4-CH3, MOS = −1.56(9)6), which are able to donate
into a single dπ orbital that is not engaged in π bonding with
the oxo group. Likewise, the seven-coordinate molybdenum in
Mo2(3,5-tBu2Cat)6 is expected to have two d orbitals available
for π bonding, allowing each of the nonbridging catecholates to
donate into one dπ orbital.
At room temperature in noncoordinating solvents such as
C6D5CD3 or CD2Cl2, the 1H NMR spectrum of
Mo2(3,5-tBu2Cat)6 shows broad resonances in both the tert-
butyl and the aromatic regions. Cooling solutions in CD2Cl2
causes the resonances to decoalesce into a very complex
spectrum with no fewer than nine distinct environments for the
catecholate ligands. Since monomeric Mo(3,5-tBu2Cat)3 would
exhibit no more than four distinct environments (one for the
Further insight into the nature of these adducts is provided
by the solid-state structure of (3,5-tBuCat)3Mo(py), which
shows a seven-coordinate monomer (Figure 2). Its geometry is
a nearly ideal capped octahedron, with the catecholates
spanning the vertices of the octahedron and the pyridine
binding along the pseudo-C3 axis. The capped face is
significantly expanded (O−Mo−O angles 112(4)° avg),
resembling tetrahedral geometry, while the opposite face is
slightly compressed compared to an ideal octahedral geometry
(O−Mo−O 82.6(12)°). The Mo−N distance (2.274(2) Å) is
typical of pyridines bonded to Mo(VI) that are not trans to an
oxo or imido group (2.27(4) Å avg, 15 examples).18,21
The three catecholates in (3,5-tBu2Cat)3Mo(py) also show
structural signs of noticeable π-donation (MOS = −1.87(11),
−1.73(11), and −1.72(14) for rings 1−3, respectively). These
values are indicative of somewhat less donation than observed
for the nonbridging catecholates in Mo2(3,5-tBu2Cat)6. This is
reasonable since the π donation in both seven-coordinate
complexes is into two dπ orbitals, but is shared among three
D
dx.doi.org/10.1021/ic401736f | Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
Figure 2. Thermal ellipsoid plot (50% ellipsoids) of the metal complex in (3,5-tBu2Cat)3Mo(py)·CH3CN, with hydrogen atoms omitted for clarity.
catecholates in (3,5-tBu2Cat)3Mo(py) as opposed to only two
in Mo2(3,5-tBu2Cat)6. The overall π donation, as judged from
the sum of the MOS values of the three catecholates, is very
similar in the two complexes (ΣMOS = −5.33 for
Mo2(3,5-tBu2Cat)6 and −5.32 for (3,5-tBu2Cat)3Mo(py)).
The dπ orbitals in the capped octahedral geometry, essentially
Pyridine-N-oxide binds to (3,5-tBu2Cat)3Mo somewhat more
tightly than does pyridine, with Keq = 2.5 for eq 3 in toluene-d8
2
2
dxy and dx −y with the Mo−N vector taken as the z axis, overlap
more strongly with the oxygen atoms on the capped face. This
may explain why the π interaction appears slightly weaker in
catecholate 1 which, uniquely, has its less electron-rich oxygen
at 0 °C. Analysis of the temperature-dependence of the
equilibrium constant K3 (−70−0 °C, Supporting Information,
Figure S2) gives ΔH°3 = −0.69 0.07 kcal/mol, ΔS°3 = −0.8
0.3 cal/mol·K.
t
(O11, meta to Bu) on the capped face. The observation of an
unsymmetrical arrangement of tert-butyl groups in the solid
state suggests that the compound must be fluxional in solution
to explain the high symmetry of its NMR spectra.
Hydrolysis of Tris(catecholate)molybdenum(VI) Complexs.
Contrary to preliminary reports,7 we find all of the
tris(catecholato)molybdenum(VI) complexes to be exception-
ally moisture-sensitive. In situ monitoring of reactions of
Mo2(3,5-tBu2Cat)6 with excess water by NMR spectroscopy
reveals that the complex reacts immediately to form free 3,5-di-
tert-butylcatechol. The two sets of catecholate resonances
characteristic of the oxomolybdenum dimer
Mo2O2(3,5-tBu2Cat)4 1a, originally reported by Buchanan and
Pierpont,8 and two other sets of catecholate resonances
assigned to a hitherto unreported isomer of it, 1b, are observed
(eq 4). Both 1a and 1b also form, in similar quantities, on
reaction of 3,5-di-tert-butylcatechol with MoO2(acac)2 in
CDCl3.22 We have been unable to find conditions to produce
1b in pure form or to separate it from 1a. The structure is
tentatively assigned to have bridging catecholates cis to the oxo
group by analogy to the quasi-isostructural
V2O2(3,5-tBu2Cat)2(3,5-tBu2SQ)2, where such a structure has
recently been observed13m in addition to the known isomer
analogous to 1a.23 If substoichiometric quantities of water are
used, a species that displays NMR signals for catecholate
ligands in five distinct environments can be observed, which is
tentatively assigned as a partially hydrolyzed monooxodimo-
While the catecholate resonances remain sharp and sym-
metrical in the NMR at all temperatures observed, resonances
due to bound pyridine in (3,5-tBu2Cat)3Mo(py) do show
broadening and coalescence with those of free pyridine (Figure
3). The linewidths of the bound pyridine peaks in the slow-
exchange regime are independent of the concentration of free
pyridine, indicating the mechanism of exchange involves
dissociation of pyridine from molybdenum. This is consistent
with the activation parameters obtained by Eyring analysis of
the exchange rate constants (0−50 °C, ΔH⧧ = 15.4 0.6 kcal/
mol, ΔS⧧ = +6.6
2.0 cal/mol·K, Supporting Information,
Figure S1). Judging from dissociation rates, the Mo-
(3,5-tBu2Cat)3 fragment appears to be less Lewis acidic than
the five-coordinate oxomolybdenum(VI) bis(amidophenoxide)
fragment (tBuClip)MoO, since (tBuClip)MoO(py) dissociates
pyridine about 10 times more slowly than does
(3,5-tBu2Cat)3Mo(py) at 22 °C.18 The oxobis(catecholate)
fragment (3,5-tBu2Cat)2MoO is qualitatively much more Lewis
acidic, since ligand exchange in (Cat)2MoO(L) is relatively
slow20 and does not involve ligand dissociation.6
E
dx.doi.org/10.1021/ic401736f | Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
1
Figure 3. Variable-temperature H NMR spectra of (3,5-tBu2Cat)3Mo(py) (500 MHz, toluene-d8). x = solvent residual peaks; B = C6H6;
=
●
t
○
(3,5- Bu2Cat)3Mo(py); = free py.
lybdenum dimer MoO(3,5-tBu2Cat)(μ-3,5-tBu2Cat)2Mo-
(3,5-tBu2Cat)2 (2).
Oxidations of Tris(catecholate)molybdenum(VI) Com-
plexes. As originally reported,7 Mo2(3,5-tBu2Cat)6 reacts with
O2, albeit more slowly than it reacts with water. The
molybdenum-containing product of the reaction with O2 (at
1 atm, added via vacuum line to avoid the addition of moisture)
is the Buchanan−Pierpont isomer of the oxomolybdenum(VI)
dimer Mo2O2(3,5-tBu2Cat)4 (1a); none of the second isomer
1b observed in the hydrolysis reaction is observed by 1H NMR
in the oxygenation reaction. The major (68% in CDCl3, 82% in
The monomeric pyridine-N-oxide adduct (3,5-tBu2Cat)3Mo-
(Opy) also reacts immediately with water to form the oxo
complex (3,5-tBu2Cat)2MoO(Opy)6 and free catechol (eq 5).
The pyridine adduct (3,5-tBu2Cat)3Mo(py) behaves similarly,
except that pyridinium salts of a variety of oxocatecholate
anions are formed in addition to the oxo complex
(3,5-tBu2Cat)2MoO(py).
F
dx.doi.org/10.1021/ic401736f | Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
toluene-d8) organic product from the reaction is 3,5-di-tert-
butyl-1,2-benzoquinone, with the residue consisting of the
extradiol oxidation product 4,6-di-tert-butyl-1-oxacyclohepta-
24
1
4,6-diene-2,3-dione (3), identified by its H NMR spectrum
(eq 6). At early times in the oxygenation reactions, the same
monooxo dimer 2 formed by incomplete hydrolysis of
1
Mo2(3,5-tBu2Cat)6 is observed by H NMR spectroscopy.
Figure 4. Typical time course of reaction of (3,5-tBu2Cat)3Mo(Opy)
with excess Opy (70 °C, C6D5CD3, total [Mo] = 0.0122 M, total
t
□
[Opy] = 0.103 M). Open squares ( ) represent [(3,5- Bu2Cat)3Mo-
t
In the presence of pyridine, (3,5-tBu2Cat)3Mo(py) reacts
with O2 to form (3,5-tBu2Cat)2MoO(py). This reaction is
substantially slower than the reaction with the homoleptic
dimer, suggesting that coordination of O2 takes place in the
initial stages of the reaction. Pyridine also suppresses formation
of the extradiol oxidation product 3. Quinone is nearly (∼90%)
the sole organic product of oxygenation in CDCl3; in toluene-
d8, quinone is the only organic product detectable by 1H NMR
spectroscopy.
○
(Opy)], open circles ( ) represent [(3,5- Bu2Cat)2MoO(Opy)]. Solid
lines represent the best fit to Supporting Information, eqs S5 and S6−
S7, respectively.
of the tris(catecholate) complex actually accelerates slightly as
the reactant is consumed, leading to an abrupt disappearance
time for the tris(catecholate). Both of these unusual features
can be explained if the initial product of reaction of the
oxobis(catecholate) complex, presumably a dioxomolybdenum
species (3,5-tBu2Cat)MoO2(Opy)n, conproportionates rapidly
with the tris(catecholate) complex (3,5-tBu2Cat)3Mo(Opy) to
form 2 equiv of oxobis(catecholate) (3,5-tBu2Cat)2MoO(Opy)
(Scheme 1). Qualitatively, capture of (3,5-tBu2Cat)-
MoO2(Opy)n by the starting tris(catecholate) complex explains
why MoO3(Opy) production is suppressed while any
(3,5-tBu2Cat)3Mo(Opy) remains. It also explains the accel-
erated rate of decomposition of (3,5-tBu2Cat)3Mo(Opy), since
the tris(catecholate) complex is consumed not only by its own
deoxygenation of pyridine-N-oxide but also through the
reaction of (3,5-tBu2Cat)2MoO(Opy), which is the product of
the oxidation of the tris(catecholate). Based on the good mass
balance of molybdenum observed by NMR (until MoO3(Opy)
begins forming), the proposed dioxo species must have a low
steady-state concentration; attempts to generate it by a different
route, such as addition of 1 equiv of 3,5-tBu2CatH2 to
MoO2(acac)2, have been unsuccessful.
The pyridine-N-oxide adduct (3,5-tBu2Cat)3Mo(Opy) reacts
under anaerobic conditions at elevated temperatures in toluene-
d8 in the presence of excess pyridine-N-oxide to form 3,5-di-
tert-butyl-1,2-benzoquinone, free pyridine, and
(3,5-tBu2Cat)2MoO(Opy) in quantitative yield (eq 7). Once
Since (3,5-tBu2Cat)2MoO(Opy) acts as a catalyst for the
consumption of (3,5-tBu2Cat)3Mo(Opy), it is reasonable that
the reaction shows the accelerated decay typical of autocatalytic
reactions. Most autocatalytic reactions retain a first-order
dependence on the reactant, giving rise to sigmoidal kinetics.25
Here, the autocatalytic step is zeroth order in the tris-
(catecholate) complex, so its decay continues to accelerate
until it is fully consumed, giving rise to an abrupt disappearance
of the reactant. Quantitatively, the rate equations for this
reaction scheme can be integrated exactly (see Supporting
Information for details), and the data are well described by this
kinetic model (Figure 4).
formed, (3,5-tBu2Cat)2MoO(Opy) is itself subject to oxidative
degradation as previously described.6 Thus, all three
catecholates of (3,5-tBu2Cat)3Mo(Opy) are ultimately con-
verted to quinone, and the molybdenum is mineralized to solid
MoO3(Opy).
The time course of the reaction (Figure 4) shows two
peculiarities. First, no MoO3(Opy) is formed until the
tris(catecholate) complex (3,5-tBu2Cat)3Mo(Opy) is com-
pletely consumed; until that point, 1H NMR integration against
an internal standard shows that the total amount of soluble
molybdenum species (the sum of (3,5-tBu2Cat)3Mo(Opy) and
(3,5-tBu2Cat)2MoO(Opy)) remains constant. Second, reaction
The kinetics would be equally well described if the
monocatecholate intermediate rapidly formed MoO3(Opy)
(either by disproportionation or further reaction with pyridine-
N-oxide), and it was the MoO3(Opy) that conproportionated
G
dx.doi.org/10.1021/ic401736f | Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
Scheme 1. Proposed Mechanism for Reaction of (3,5-tBu2Cat)3Mo(Opy) with Excess Opy
with the tris(catecholate) complex to form the monooxobis-
(catecholate) complex. However, control experiments indicate
that neither bulk MoO3 nor isolated MoO3(Opic)6 reacts with
(3,5-tBu2Cat)3Mo(Opy) (in the presence of Opy or Opic) at a
significant rate even at 70 °C. While it is possible that a nascent
form of MoO3 is produced in the reaction that is significantly
more reactive than bulk MoO3, it appears more likely that it is a
species such as MoO2(3,5-tBu2Cat)(Opy)n that is captured by
the tris(catecholate) complex (3,5-tBu2Cat)3Mo(Opy).
At the point where the tris(catecholate) complex
(3,5-tBu2Cat)3Mo(Opy) disappears, the fate of this dioxo
intermediate suddenly changes, from being converted to the
monooxo complex (3,5-tBu2Cat)2MoO(Opy) to being con-
verted to MoO3(Opy). At this cusp, almost all of the
molybdenum is in the form of the monooxo complex
(3,5-tBu2Cat)2MoO(Opy), so its rate of oxygen atom transfer
dictates the progress of the reaction. Just before the cusp, each
time a mole of (3,5-tBu2Cat)2MoO(Opy) undergoes oxygen
atom transfer, it produces one mole of benzoquinone and
actually makes an additional mole of itself (since subsequent
capture of the dioxomolybdenum intermediate by the tris-
(catecholate) reactant makes two moles of the monooxo
molybdenum complex).
Immediately after the cusp, the rate of oxygen atom transfer
within (3,5-tBu2Cat)2MoO(Opy) will be substantially un-
changed, since its concentration is almost unchanged, but the
rate of formation of products will depend on the route the
dioxo intermediate takes to form MoO3(Opy). If the dioxo
complex reacts by further deoxygenation of pyridine-N-oxide
(eq 8a), then each mole of (3,5-tBu2Cat)2MoO(Opy) that
undergoes oxygen atom transfer will produce two moles of
benzoquinone while consuming one mole of the oxo complex.
In contrast, if the dioxo intermediate forms MoO3(Opy) by
disproportionation (eq 8b), then each mole of
(3,5-tBu2Cat)2MoO(Opy) that undergoes oxygen atom transfer
will produce only one mole of benzoquinone (disproportiona-
tion produces no quinone) and will result in net consumption
of only one-half mole of monooxo complex (since each mole of
MoO3(Opy) produced is accompanied by restoration of a mole
of the monooxo complex). Experimentally, at the cusp, the rate
of production of benzoquinone is unchanged and the rate of
production of (3,5-tBu2Cat)2MoO(Opy) changes abruptly by a
factor of −0.5 (Supporting Information, Figure S3b). This is
exactly what is predicted if MoO3(Opy) is formed by
disproportionation (eq 8b) and inconsistent with eq 8a.
MoO2(3,5‐tBu2Cat)(Opy)n
→ MoO3(Opy) + py + 3,5‐tBu2BQ + (n − 2)Opy
(8a)
2MoO2(3,5‐tBu2Cat)(Opy)n
→ MoO3(Opy) + (3,5‐tBu2Cat)2MoO(Opy)
+ (n − 2)Opy
(8b)
Quantitative analysis of the experimental time course of the
concentration of (3,5-tBu2Cat)2MoO(Opy) in toluene at 70 °C
gives average values of the rate constant k1 = 3.6(5) × 10−5 s−1
and k2 = 9.4(3) × 10−5 s−1 (triplicate measurements). The
latter rate constant is in satisfactory agreement with the rate
constant determined independently by UV−visible spectro-
scopic monitoring of reactions of (3,5-tBu2Cat)2MoO(Opy)6
(e.g., k2 = 1.60(8) × 10−4 s−1 at 72.5 °C).26 In the regime where
tris(catecholate) is still present, the reactions of both tris- and
bis(catecholate) result in disappearance of tris(catecholate),
with its rate shifting from k1[Mototal] initially (when almost all
the molybdenum is in the form of (3,5-tBu2Cat)3Mo(Opy)) to
k2[Mototal] at the end of this phase of the reaction (when almost
all the molybdenum is in the form of (3,5-tBu2Cat)2MoO-
(Opy)). The fact that k2 is greater than k1 is consistent with the
qualitative observation that the decay of the tris(catecholate)
complex accelerates over time, and the fact that this
acceleration is only modest is consistent with the fact that k2
is only modestly greater than k1.
DISCUSSION
■
Origin of Lewis Acidity of Tris(3,5-di-tert-
butylcatecholato)molybdenum(VI). A prominent feature
of (3,5-tBu2Cat)3Mo is its Lewis acidity, resulting in a tendency
to form seven-coordinate adducts with Lewis bases. In the solid
state, it dimerizes via catecholate bridges, and the dimeric
structure appears to be retained at low temperature in solution.
Lewis bases such as pyridine or pyridine-N-oxide are bound
tightly, with no sign of free (3,5-tBu2Cat)3Mo even in the
H
dx.doi.org/10.1021/ic401736f | Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
absence of excess ligand. Monomeric (3,5-tBu2Cat)3Mo is likely
thermally accessible in solutions of the dimer in non-
coordinating solvent, based on the coalescence of all
catecholate resonances into a single environment at ∼30 °C,
and the monomeric complex is certainly thermally accessible
from (3,5-tBu2Cat)3Mo(py), based on its dissociative mecha-
nism of ligand substitution. Nevertheless, the enthalpy of
containing Mo(V)37 or Mo(IV)38) or to donation from the
ligand π orbitals enabled by ligand folding and reduction in
symmetry of the trigonal prismatic geometry to C3h.39 Again the
complexes are isolated as six-coordinate monomers, though
they are known to react under some conditions with
nucleophiles.40
Nonclassical Redox Reactions of (3,5-tBu2Cat)3Mo. In
an earlier study of (3,5-tBu2Cat)2Mo(O)(Opy), it was found
that the oxomolybdenum(VI) bis(catecholate) fragment
deoxygenated pyridine-N-oxide form 3,5-di-tert-butyl-1,2-ben-
zoquinone and, ultimately, MoO3(Opy).6 Because the oxygen
atom was delivered to the molybdenum, while the electrons
needed for reducing the N-oxide were drawn from coordinated
catecholate, this mode of reactivity was described as a
“nonclassical” oxygen atom transfer reaction. Since
(3,5-tBu2Cat)3Mo(Opy) also has coordinated pyridine-N-
oxide, coordinated catecholate, and molybdenum(VI) to
mediate between them, it was of interest to assess the ability
of the molybdenum tris(catecholate) fragment to undergo
nonclassical oxygen atom transfer as well.
activation for dissociation of pyridine (ΔH⧧ = 15.4
0.6
diss
kcal/mol) provides evidence for the appreciable Lewis acidity
of the six-coordinate tris(catecholate).
Electronically, monomeric octahedral tris(catecholato)-
molybdenum(VI) would best be considered a 16-electron
complex. Catecholate has one strongly π-donating orbital, the
in-phase combination of oxygen pπ lone pairs, which is raised in
energy because of its antibonding interaction with one of the
filled benzene π orbitals.27 Taken in addition to the six σ bonds,
π donation from the three ligands would saturate the
molybdenum center. However, the a2-symmetric combination
of ligand π donor orbitals finds no symmetry match among the
metal d orbitals (which transform as a1 + 2e in D3 symmetry).28
2
The tris(catecholate) complex (3,5-tBu2Cat)3Mo(Opy) does
react cleanly in the presence of excess Opy to form
(3,5-tBu2Cat)2Mo(O)(Opy), free pyridine, and 3,5-tBu2BQ. In
the absence of a detailed mechanistic study, we cannot
absolutely rule out the possibility of quinone dissociation
prior to N−O bond cleavage. However, since the reaction is not
inhibited by benzoquinone and is very sensitive to the nature of
the oxidant (N-methylmorpholine-N-oxide reacts in minutes at
room temperature), it seems likely that the tris(catecholate)
reacts by an analogous mechanism to the bis(catecholate)
species, that is, by initial N−O bond cleavage to form an
oxomolybdenum species with oxidized catecholates, followed
by subsequent dissociation of quinone. This nonclassical
oxygen atom transfer mechanism would require the formation
of a short-lived seven-coordinate oxotris(dioxolene) species.
The isolation of (3,5-tBu2Cat)3Mo(py) indicates that this is
sterically accessible, and while seven-coordinate molybdenum-
(VI) oxo complexes are uncommon (except for complexes with
peroxo or related ligands), they are precedented with small bite
angle ligands such as dithiocarbamates.41
The reactivity of the tris(catecholate)molybdenum fragment
toward pyridine-N-oxide is strikingly similar to that of the
oxobis(catecholate)molybdenum fragment, with rate constants
for the reaction of the former and latter species differing by less
than an order of magnitude at 70 °C in toluene. The reaction of
(3,5-tBu2Cat)3Mo(Opy) with excess pyridine-N-oxide consti-
tutes a net six-electron process, ultimately producing
MoO3(Opy) and 3 equiv of 3,5-di-tert-butyl-1,2-benzoquinone.
Not only is it unusual to find metal centers that can span such a
wide range of oxidation states, but if one does, then successive
one- or two-electron steps often occur at highly disparate rates
since successive redox potentials typically change dramatically
as the metal oxidation state changes. In the present system,
each successive two-electron transformation involves the same
catecholate to quinone oxidation, so it is perhaps not surprising
that the first and second oxidations occur at similar rates. (The
third oxidation step is obscured by fast disproportionation of
the putative dioxomonocatecholate species.) The ability
demonstrated here of multiple redox-active ligands to supply
large numbers of electrons at relatively similar rates illustrates
two potential advantages of a nonclassical oxidation strategy for
achieving multielectron redox chemistry.
This leaves the dz orbital (taking the z axis along the molecular
3-fold axis, this is a dπ orbital) as essentially nonbonding and
thus available for binding to a seventh ligand.
Previously prepared molybdenum tris(catecholate) com-
plexes have also shown signs of electron deficiency. For
example, in the solid state, one of the ligands in Mo-
(phenanthrenediolate)3 adopts an η4 geometry,29 with the
Mo−C distances (2.421−2.426 Å) similar to analogous ones in
Mo(η4-2,3-butadiene)3 (2.393−2.404 Å).30 This effectively
makes the molybdenum seven-coordinate. Molybdenum(VI)
tris(tetrachlorocatecholate) adopts a dimeric structure with two
bridging catecholates which bind in a κ1,κ1 fashion;31 the
tungsten analogue is isostructural.32 While the molybdenum
remains six-coordinate, the bridging catecholates are able to
bind with much more obtuse Mo−O−C angles than the
chelating ligands (134(5)° vs 118.4(10)° avg). Increasing the
M−O−C angle is known to allow for enhanced π donation,33
and indeed the bridging catecholates show significantly shorter
Mo−O distances than the chelating ones (1.861(11) Å vs
1.948(10) Å). A similar argument has been adduced to explain
the preference for monodentate aryloxide donation over
catecholate chelation in the (diketonate)2Ti(IV) fragment
isoelectronic with (Cl4C6O2)2Mo(VI).34 Thus, although
formation of adducts with exogenous bases has not been
previously reported, Lewis acidity appears to be a general
feature of tris(catecholate)molybdenum complexes.
The Lewis acidity of tris(catecholate)molybdenum contrasts
with the behavior of the lighter congener, chromium, where
(R4C6O2)3Cr complexes are invariably octahedral tris-
(chelates).35 This is undoubtedly due in part to the smaller
size of chromium, which makes seven-coordination less
favorable, but electronic factors also likely play a role. Because
of the lower d-orbital energy of Cr compared to Mo, neutral
tris(catecholato)chromium species are best described as Cr(III)
complexes of semiquinonate ligands, with strong antiferromag-
netic coupling giving rise to an S = 0 ground state.36 The
2
orbitals derived from the Cr dz are thus not empty, and the
metal is correspondingly not electrophilic. An analogous effect
is observed in tris(benzenedithiolate)molybdenum complexes,
where the orbital energy of the ligands is raised because of the
lower electronegativity of sulfur compared to oxygen. Again the
2
dz orbital has significant electron density, due either to formal
electron transfer to Mo (the complexes have been described as
I
dx.doi.org/10.1021/ic401736f | Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
In contrast to their similar reactivity toward pyridine-N-
oxide, the tris(catecholate)molybdenum and oxobis-
(catecholate)molybdenum fragments behave very differently
toward dioxygen. The tris(catecholate) complexes
(3,5-tBu2Cat)3Mo(L) or dimeric Mo2(3,5-tBu2Cat)6 are readily
oxidized by O2, with reactions proceeding rapidly even at room
temperature, while (3,5-tBu2Cat)2MoO(L) or dimeric
Mo2O2(3,5-tBu2Cat)4 are air-stable. When dioxygen reacts, it
affords significant amounts of the extradiol oxidation product
4,6-di-tert-butyl-1-oxacyclohepta-4,6-diene-2,3-dione (3), as
well as the quinone which is the exclusive product of oxidations
with pyridine-N-oxide. Ring-expanded products like 3 are
usually attributed to Baeyer−Villiger ring expansions of
peroxyquinone intermediates (e.g., Scheme 2). Similar
and dioxygen (which forms both quinone and the extradiol
oxidation product 4,6-di-tert-butyl-1-oxacyclohepta-4,6-diene-
2,3-dione). The metal center appears to play a critical role in
both cases in binding the oxidant as well as serving as the
ultimate acceptor of oxygen atoms from the oxidant, forming
oxomolybdenum(VI) species as products. The molybdenum
maintains its +6 oxidation state throughout the reactions. The
reaction with pyridine-N-oxide is thus a “nonclassical” oxygen
atom transfer reaction, in which the oxygen atom is delivered to
the metal center but the electrons needed to reduce the N-
oxide come from the catecholate. The reaction with dioxygen is
likely more complicated, as the presence of an extradiol
oxidation product strongly suggests that the catecholate serves
not only as a reservoir of electrons but also engages in
formation of new covalent bonds to the O2 substrate over the
course of the reaction.
Scheme 2. Possible Intermediacy of Peroxyquinones in
Formation of 3 and Quinone
ASSOCIATED CONTENT
■
S
* Supporting Information
Eyring plot for exchange of free and bound pyridine in
(3,5-tBu2Cat)3Mo(py), van’t Hoff plot of eq 3, plots of H
1
NMR spectra and time evolution of species in the thermolysis
of (3,5-tBu2Cat)3Mo(Opy), and derivation of kinetic equations
for the thermolysis of (3,5-tBu2Cat)3Mo(Opy) (PDF format)
and crystallographic information on Mo2(3,5-tBu2Cat)6·4C6D6
and (3,5-tBu2Cat)3Mo(py)·CH3CN in CIF format. This
material is available free of charge via the Internet at http://
pubs.acs.org.
AUTHOR INFORMATION
■
Corresponding Author
*E-mail: Seth.N.Brown.114@nd.edu.
intermediates derived from iminoquinones have been observed
directly in other reactions of dioxygen with catecholato or
amidophenoxide complexes of redox-inert metals.42 The
observation that donors such as pyridine retard the reaction
suggests that O2 must find an open coordination site on Mo to
react. The effect of pyridine on the product distribution
(suppressing formation of extradiol oxidation) is more difficult
to interpret. One possibility is that pyridine affects the
selectivity of fragmentation of the peroxyquinone intermediate,
enhancing quinone formation or inhibiting Baeyer−Villiger
rearrangement. Alternatively, the metal might also react with O2
by a pathway that does not involve C−O bond formation, and
pyridine might favor this pathway at the expense of forming the
peroxyquinone intermediate. Even with the limited data
available, it is clear that the reactivity of dioxygen does not
parallel that of simple two-electron oxygen atom donors such as
pyridine-N-oxide.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
We thank Dr. Allen Oliver for assistance with the X-ray
crystallography and Prof. Richard G. Finke for helpful
discussions about autocatalysis. This work was supported by
the National Science Foundation (CHE-1112356). K.R.
acknowledges summer fellowship support from the Notre
Dame RESACC program.
REFERENCES
■
(1) (a) Pierpont, C. G.; Lange, C. W. Prog. Inorg. Chem. 1994, 41,
331−442. (b) Pierpont, C. G. Coord. Chem. Rev. 2001, 219−221, 415−
433.
(2) (a) Costas, M.; Mehn, M. P.; Jensen, M. P.; Que, L., Jr. Chem.
Rev. 2004, 104, 939−986. (b) Vaillancourt, F. H.; Bolin, J. T.; Eltis, L.
D. Crit. Rev. Biochem. Mol. Biol. 2006, 41, 241−267.
(3) Yin, C.-X.; Finke, R. G. J. Am. Chem. Soc. 2005, 127, 9003−9013.
(4) Morris, A. M.; Pierpont, C. G.; Finke, R. G. J. Mol. Catal. A 2009,
309, 137−145.
(5) Munha, R. F.; Zarkesh, R. A.; Heyduk, A. F. Dalton Trans. 2013,
́
42, 3751−3766.
(6) Marshall-Roth, T.; Liebscher, S. C.; Rickert, K.; Seewald, N. J.;
Oliver, A. G.; Brown, S. N. Chem. Commun. 2012, 48, 7826−7828.
(7) Cass, M. E.; Pierpont, C. G. Inorg. Chem. 1986, 25, 122−123.
(8) Buchanan, R. M.; Pierpont, C. G. Inorg. Chem. 1979, 18, 1616−
1620.
CONCLUSIONS
■
The tris(catecholato)molybdenum(VI) fragment
[(3,5-tBu2Cat)3Mo] is Lewis acidic, existing in the solid state
and in noncoordinating solvents as a seven-coordinate dimer
with bridging catecholates. Lewis bases add to give seven-
coordinate monomers such as capped octahedral
(3,5-tBu2Cat)3Mo(py). These compounds are well described
as containing Mo(VI) and dianionic catecholates, with
significant π donation from the catecholates to the
molybdenum. The reducing power stored in the catecholate
ligands can be harnessed in reactions with pyridine-N-oxide
(which forms exclusively 3,5-di-tert-butyl-1,2-benzoquinone)
(9) Budzelaar, P. H. M. gNMR, v. 3.5.6; Cherwell Scientific
Publishing: Oxford, U.K., 1996.
(10) Sheldrick, G. M. Acta Crystallogr., Sect. A 2008, A64, 112−122.
J
dx.doi.org/10.1021/ic401736f | Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
(11) International Tables for Crystallography; Kluwer Academic
Publishers: Dordrecht, The Netherlands, 1992; Vol C.
(12) Calucci, L.; Pampaloni, G.; Pinzino, C.; Prescimone, A. Inorg.
(22) This reaction has been reported as a preparation of
Mo2O2(3,5-tBu2Cat)4, and the dimeric structure in noncoordinating
solvents was confirmed by osmometry, but NMR data were only
obtained in coordinating solvents in which the complex forms
monomeric MoO(3,5-tBu2Cat)2(L), which does not address the
question of isomerism in the dimer. Wilshire, J. P.; Leon, L.;
Bosserman, P.; Sawyer, D. T. J. Am. Chem. Soc. 1979, 101, 3379−3381.
(23) Cass, M. E.; Greene, D. L.; Buchanan, R. M.; Pierpont, C. G. J.
Am. Chem. Soc. 1983, 105, 2680−2686.
(24) (a) Funabiki, T.; Mizoguchi, A.; Sugimoto, T.; Tada, S.; Tsuji,
M.; Sakamoto, H.; Yoshida, S. J. Am. Chem. Soc. 1986, 108, 2921−
2932. (b) Weiner, H.; Finke, R. G. J. Am. Chem. Soc. 1999, 121, 9831−
9842.
Chim. Acta 2006, 359, 3911−3920.
(13) (a) Borgias, B. A.; Cooper, S. R.; Koh, Y. B.; Raymond, K. N.
Inorg. Chem. 1984, 23, 1009−1016. (b) Shoner, S. C.; Power, P. P.
Inorg. Chem. 1992, 31, 1001−1010. (c) Speier, G.; Tisza, S.; Tyeklar,
Z.; Lange, C. W.; Pierpont, C. G. Inorg. Chem. 1994, 33, 2041−2045.
(d) Chi, Y.; Lan, J.-W.; Ching, W.-L.; Peng, S.-M.; Lee, G.-H. J. Chem.
Soc., Dalton Trans. 2000, 2923−2927. (e) Manos, M. J.; Tasiopoulos,
A. J.; Raptopoulou, C.; Terzis, A.; Woollins, J. D.; Slawin, A. M. Z.;
Keramidas, A. D.; Kabanos, T. A. J. Chem. Soc., Dalton Trans. 2001,
1556−1558. (f) Fukin, G. K.; Zakharov, L. N.; Maslennikov, S. V.;
Piskunov, A. V.; Cherkasov, V. K. Acta Crystallogr., Sect. C 2001, 57,
(25) Watzky, M. A.; Finke, R. G. J. Am. Chem. Soc. 1997, 119,
1020−1021. (g) Kaizer, J.; Pap, J.; Speier, G.; Par
Rockenbauer, A. J. Inorg. Biochem. 2002, 91, 190−198. (h) Banse, F.;
Balland, V.; Philouze, C.; Riviere, E.; Tchertanova, L.; Girerd, J.-J.
́ ́
kanyi, L.; Korecz, L.;
10382−10400.
(26) The values reported in ref 6 are multiplied by two for
comparison with the value reported here to account for the fact that
oxidation of one mole of (3,5-tBu2Cat)2MoO(Opy) results in the
disappearance of only one-half a mole of the complex (since
disproportionation of the initial product as described in eq 8b restores
half a mole of monooxo complex). Thus the apparent decay constants
reported in ref 6 correspond to 0.5k2.
̀
Inorg. Chim. Acta 2003, 353, 223−230. (i) Boudalis, A. K.; Dahan, F.;
Bousseksou, A.; Tuchagues, J.-P.; Perlepes, S. P. Dalton Trans. 2003,
3411−3418. (j) Maslennikov, S. V.; Aivaz’yan, I. A.; Fukin, G. K.;
Cherkasov, V. K.; Kurskii, Y. A. Koord. Khim. 2004, 30, 667−672.
(k) Davidson, M. G.; Jones, M. D.; Lunn, M. D.; Mahon, M. F. Inorg.
Chem. 2006, 45, 2282−2287. (l) Wagner, M.; Limberg, C.; Tietz, T.
Chem.Eur. J. 2009, 15, 5567−5576. (m) Scales, E.; Sorace, L.; Dei,
A.; Caneschi, A.; Muryn, C. A.; Collison, D.; McInnes, E. J. L. Chem.
Sci. 2010, 1, 221−225. (n) Flogeras, J. C.; Allan, C. R.; Vogels, C. M.;
Decken, A.; Westcott, S. A. X-ray Struct. Anal. Online 2011, 27, 45−46.
(14) (a) Block, E.; Ofori-Okai, G.; Kang, H.; Zubieta, J. Inorg. Chim.
Acta 1991, 190, 179−184. (b) Hughes, D. L.; Lazarowych, N. J.;
Maguire, M. J.; Morris, R. H.; Richards, R. L. J. Chem. Soc., Dalton
Trans. 1995, 5−15. (c) Sukcharoenphon, K.; Capps, K. B.; Abboud, K.
A.; Hoff, C. D. Inorg. Chem. 2001, 40, 2402−2408. (d) Yonemura, T.;
Nakata, J.; Kadota, M.; Hasegawa, M.; Okamoto, K.; Ama, T.;
Kawaguchi, H.; Yasui, T. Inorg. Chem. Commun. 2001, 4, 661−663.
(e) Yonemura, T.; Hashimoto, T.; Hasegawa, M.; Ikenoue, T.; Ama,
T.; Kawaguchi, H. Inorg. Chem. Commun. 2006, 9, 183−186.
(f) Yonemura, T. Acta Crystallogr., Sect. E 2009, 65, m1463−m1464.
(15) (a) Seymore, S. B.; Brown, S. N. Inorg. Chem. 2006, 45, 9540−
9550. (b) Seymore, S. B.; Brown, S. N. Inorg. Chem. 2002, 41, 462−
469. (c) Seymore, S. B.; Brown, S. N. Inorg. Chem. 2001, 40, 6676−
6683. (d) Bishop, M. W.; Chatt, J.; Dilworth, J. R.; Dahlstrom, P.;
Hyde, J.; Zubieta, J. J. Organomet. Chem. 1981, 213, 109−124.
(e) Williams, G. A.; Smith, A. R. P. Aust. J. Chem. 1980, 33, 717−728.
(f) Hursthouse, M. B.; Motevalli, M. J. Chem. Soc., Dalton Trans. 1979,
1362−1366. (g) Butler, G.; Chatt, J.; Hussain, W.; Leigh, G. J.;
Hughes, D. L. Inorg. Chim. Acta 1978, 30, L287−L288.
(27) Gordon, D. J.; Fenske, R. F. Inorg. Chem. 1982, 21, 2907−2915.
(28) (a) Karpishin, T. B.; Gebhard, M. S.; Solomon, E. I.; Raymond,
K. N. J. Am. Chem. Soc. 1991, 113, 2977−2984. (b) Karpishin, T. B.;
Dewey, T. M.; Raymond, K. N. J. Am. Chem. Soc. 1993, 115, 1842−
1851.
(29) Pierpont, C. G.; Buchanan, R. M. J. Am. Chem. Soc. 1975, 97,
4912−4917.
́
(30) (a) Bogdanovic, B.; Bonnemann, H.; Goddard, R.; Startsev, A.;
̈
Wallis, J. M. J. Organomet. Chem. 1986, 299, 347−355. (b) Yun, S. S.;
Kang, S. K.; Suh, I.-H.; Choi, Y. D.; Chang, I. S. Organometallics 1991,
10, 2509−2512.
(31) Pierpont, C. G.; Downs, H. H. J. Am. Chem. Soc. 1975, 97,
2123−2127.
(32) deLearie, L. A.; Pierpont, C. G. Inorg. Chem. 1988, 27, 3842−
3845.
(33) Smith, G. D.; Fanwick, P. E.; Rothwell, I. P. Inorg. Chem. 1990,
29, 3221−3226.
(34) Dulatas, L. T.; Brown, S. N.; Ojomo, E.; Noll, B. C.; Cavo, M. J.;
Holt, P. B.; Wopperer, M. M. Inorg. Chem. 2009, 48, 10789−10799.
(35) (a) Pierpont, C. G.; Downs, H. H. J. Am. Chem. Soc. 1976, 98,
4834−4838. (b) Sofen, S. R.; Ware, D. C.; Cooper, S. R.; Raymond, K.
N. Inorg. Chem. 1979, 18, 234−239. (c) Chang, H.-C.; Kitagawa, S.
Mol. Cryst. Liq. Cryst. 2002, 376, 275−282.
(36) Kapre, R. R.; Bothe, E.; Weyhermuller, T.; George, S. D.;
Muresan, N.; Wieghardt, K. Inorg. Chem. 2007, 46, 7827−7839.
(37) Kapre, R. R.; Bothe, E.; Weyhermuller, T.; George, S. D.;
̈
(16) (a) Bristow, S.; Enemark, J. H.; Garner, C. D.; Minelli, M.;
Morris, G. A.; Ortega, R. B. Inorg. Chem. 1985, 24, 4070−4077.
(b) Bristow, S.; Garner, C. D.; Morris, G. A.; Enemark, J. H.; Minelli,
M.; Ortega, R. B. Polyhedron 1986, 5, 319−321.
̈
Wieghardt, K. Inorg. Chem. 2007, 46, 5642−5650.
(38) Tenderholt, A. L.; Szilagyi, R. K.; Holm, R. H.; Hodgson, K. O.;
Hedman, B.; Solomon, E. I. Inorg. Chem. 2008, 47, 6382−6392.
(39) Campbell, S.; Harris, S. Inorg. Chem. 1996, 35, 3285−3288.
(17) Drew, M. G. B. Prog. Inorg. Chem. 1977, 23, 67−210.
(18) Kopec, J. A.; Shekar, S.; Brown, S. N. Inorg. Chem. 2012, 51,
1239−1250.
(40) Cervilla, A.; Per
2004, 1461−1465.
́ ́
ez-Pla, F.; Llopis, E.; Piles, M. Dalton Trans.
(19) Brown, S. N. Inorg. Chem. 2012, 51, 1251−1260.
(41) (a) Dirand, J.; Ricard, L.; Weiss, R. Transition Met. Chem. 1975,
1, 2−5. (b) Dirand, J.; Ricard, L.; Weiss, R. J. Chem. Soc., Dalton Trans.
1976, 278−282. (c) Marabella, C. P.; Enemark, J. H.; Newton, W. E.;
McDonald, J. W. Inorg. Chem. 1982, 21, 623−627. (d) Arzoumanian,
H.; Corao, C.; Krentzien, H.; Lopez, R.; Teruel, H. Chem. Commun.
1992, 856−858. (e) Decoster, M.; Conan, F.; LeMest, Y.; Pala, J. S.;
Leblanc, A.; Molinie, P.; Faulques, E.; Toupet, L. New J. Chem. 1997,
21, 215−221.
(20) Liu, C.-M.; Nordlander, E.; Schmeh, D.; Shoemaker, R.;
Pierpont, C. G. Inorg. Chem. 2004, 43, 2114−2124.
(21) (a) Blatchford, T. P.; Chisholm, M. H.; Huffman, J. C. Inorg.
Chem. 1988, 27, 2059−2070. (b) Gosink, H.-J.; Roesky, H. W.;
Noltemeyer, M.; Schmidt, H.-G.; Freire-Erdbrugger, C.; Sheldrick, G.
M. Chem. Ber. 1993, 126, 279−283. (c) Feher, F. J.; Rahimian, K.;
Budzichowski, T. A.; Ziller, J. W. Organometallics 1995, 14, 3920−
3926. (d) Zhu, S. S.; Cefalo, D. R.; La, D. S.; Jamieson, J. Y.; Davis, W.
M.; Hoveyda, A. H.; Schrock, R. R. J. Am. Chem. Soc. 1999, 121,
8251−8259. (e) Alexander, J. B.; Schrock, R. R.; Davis, W. M.;
Hultzsch, K. C.; Hoveyda, A. H.; Houser, J. H. Organometallics 2000,
19, 3700−3715. (f) Hanna, T. A.; Ghosh, A. K.; Ibarra, C.; Mendez-
Rojas, M. A.; Rheingold, A. L.; Watson, W. H. Inorg. Chem. 2004, 43,
(42) (a) Abakumov, G. A.; Poddel’sky, A. I.; Grunova, E. V.;
Cherkasov, V. K.; Fukin, G. K.; Kurskii, Y. A.; Abakumova, L. G.
Angew. Chem., Int. Ed. 2005, 44, 2767−2771. (b) Cherkasov, V. K.;
Abakumov, G. A.; Grunova, E. V.; Poddel’sky, A. I.; Fukin, G. K.;
Baranov, E. V.; Kurskii, Y. V.; Abakumova, L. G. Chem.Eur. J. 2006,
12, 3916−3927. (c) Poddel’sky, A. I.; Vavilina, N. N.; Somov, N. V.;
Cherkasov, V. K.; Abakumov, G. A. J. Organomet. Chem. 2009, 694,
1511−1516. (g) Jiang, A. J.; Schrock, R. R.; Muller, P. Organometallics
̈
2008, 27, 4428−4438.
K
dx.doi.org/10.1021/ic401736f | Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
3462−3469. (d) Poddel’sky, A. I.; Kurskii, Y. A.; Piskunov, A. V.;
Somov, N. V.; Cherkasov, V. K.; Abakumov, G. A. Appl. Organomet.
Chem. 2011, 25, 180−189.
L
dx.doi.org/10.1021/ic401736f | Inorg. Chem. XXXX, XXX, XXX−XXX