Synthesis and Oxidation Catalysis of [Tris(oxazolinyl)borato]cobalt(II) Scorpionates

Article abstract of DOI:10.1002/ejic.201600237

The reaction of CoCl₂·THF and thallium tris(4,4-dimethyl-2-oxazolinyl)phenylborate (TlTo^M) in tetrahydrofuran (THF) provides To^MCoCl (1) in 95 % yield; however, appropriate solvents and starting materials are required to favor 1 over two other readily formed side-products, (To^M)₂Co (2) and {HTo^M}CoCl₂(3). ESR, NMR, FTIR, and UV/Vis spectroscopies were used to distinguish these cobalt(II) products and probe their electronic and structural properties. Even after the structures indicated by these methods were confirmed by X-ray crystallography, the spectroscopic identification of trace contaminants in the material was challenging. The recognition of possible contaminants in the synthesis of To^MCoCl in combination with the paramagnetic nature of these complexes provided impetus for the utilization of X-ray powder diffraction to measure the purity of the To^MCoCl bulk sample. The X-ray powder diffraction results provide support for the bulk-phase purity of To^MCoCl in preparations that avoid 2 and 3. Thus, 1 is a precursor for new [tris(oxazolinyl)borato]cobalt chemistry, as exemplified by its reactions with KOtBu and NaOAc to give To^MCoOtBu (4) and To^MCoOAc (5), respectively. Compound 5 is a catalyst for the oxidation of cyclohexane with meta-chloroperoxybenzoic acid (mCPBA), and the rate constants and selectivity for cyclohexanol versus cyclohexanone and ε-caprolactone were assessed.

Full text of DOI:10.1002/ejic.201600237

DOI: 10.1002/ejic.201600237
Full Paper
CLUSTER
ISSUE
Oxidation Catalysis
Synthesis and Oxidation Catalysis of
[Tris(oxazolinyl)borato]cobalt(II) Scorpionates
Regina R. Reinig,[a,b] Debabrata Mukherjee,[a] Zachary B. Weinstein,[a,b] Weiwei Xie,[a,b]
Toshia Albright,[a] Benjamin Baird,[a] Tristan S. Gray,[a] Arkady Ellern,[a] Gordon J. Miller,[a,b]
Arthur H. Winter,[a] Sergey L. Bud'ko,[b] and Aaron D. Sadow*[a,b]
Abstract: The reaction of CoCl2·THF and thallium tris(4,4-di- with the paramagnetic nature of these complexes provided im-
methyl-2-oxazolinyl)phenylborate (TlToM) in tetrahydrofuran petus for the utilization of X-ray powder diffraction to measure
(THF) provides ToMCoCl (1) in 95 % yield; however, appropriate the purity of the ToMCoCl bulk sample. The X-ray powder dif-
solvents and starting materials are required to favor 1 over two fraction results provide support for the bulk-phase purity of
other readily formed side-products, (ToM)2Co (2) and ToMCoCl in preparations that avoid 2 and 3. Thus, 1 is a precur-
{HToM}CoCl2 (3). ESR, NMR, FTIR, and UV/Vis spectroscopies sor for new [tris(oxazolinyl)borato]cobalt chemistry, as exempli-
were used to distinguish these cobalt(II) products and probe fied by its reactions with KOtBu and NaOAc to give ToMCoOtBu
their electronic and structural properties. Even after the struc- (4) and ToMCoOAc (5), respectively. Compound 5 is a catalyst
tures indicated by these methods were confirmed by X-ray crys- for the oxidation of cyclohexane with meta-chloroperoxy-
tallography, the spectroscopic identification of trace contami- benzoic acid (mCPBA), and the rate constants and selectivity for
nants in the material was challenging. The recognition of possi- cyclohexanol versus cyclohexanone and ε-caprolactone were
ble contaminants in the synthesis of ToMCoCl in combination assessed.
Introduction
azolyl)borate (Tp) or tris(3,5-dimethyl-pyrazolyl)borate (Tp*) li-
gands, which might provide more accessible metal centers, in-
stead form octahedral {κ3-Tp}2M compounds.[10] Even the bulk-
ier tris(3-phenyl-5-methylpyrazolyl)borate (TpPh,Me) ligand forms
octahedral TpPh,Me2Co upon isomerization of a pyrazole ring.[11]
Thus, the steric properties of non-pyrazolyl-based scorpionate
ligands, such as those involving nonplanar oxazoline donors,
may sufficiently stabilize reactive moieties and also allow new
metal-centered reactivity that involves switching between four-
Tridentate fac-coordinating monoanionic scorpionate-type li-
gands[1] support and stabilize first-row metal centers bonded
to reactive moieties including hydride,[2] alkyl groups contain-
ing ꢀ-hydrogen atoms,[3] imido and oxido ligands,[4] azides,[5]
and oxidizing moieties such as peroxides.[6] In addition, 3d
metal compounds coordinated by scorpionate ligands have
served as models for metal sites in enzymes and provided moti-
vation to study their spectroscopic and structural features in
detail.[7] Typically, sterically encumbered scorpionates are re-
quired to support reactive species such as peroxides and super-
oxides.
Variation of the steric and electronic properties of the ancil-
lary scorpionate donors can greatly influence the stability of
reactive moieties and the reactivity of the complexes. For exam-
ple, the steric encumbrance of tris(3-tert-butyl-5-methyl-
pyrazolyl)borate (TptBu,Me), known as the tetrahedral enforcer,
stabilizes reactive groups such as alkyl peroxides.[8] However,
this stabilization may also limit catalytic oxidation chemistry of
bulky scorpionates compared to the well-known reactivity of
other first-metal complexes.[6,9] The smaller parent tris(pyr-
and five-coordinate species in a way not accessible with TptBu
based compounds.
-
Recently, we discovered that tris(4,4-dimethyl-2-oxazolin-
yl)phenylborate (ToM) supports reactive main-group complexes
of mononuclear and tetrahedral zinc(II) and magnesium(II). For
example, a catalytically active mononuclear zinc hydride was
readily synthesized by the reaction of ToMZnOtBu and
PhSiH3.[12] The ToMZnX system also supports and stabilizes alkyl
peroxides, such that ToMZnOOEt is thermally persistent even at
temperatures above 100 °C.[13] Moreover, ToMZnEt is sufficiently
reactive to undergo selective oxidation upon treatment with O2
to give ToMZnOOEt. ToM gives a more open geometry and bowl-
like steric profile, in contrast to the tetrahedral enforcer TptBu,Me
that typically stabilizes reactive moieties.
These examples provide motivation for studying the synthe-
sis and reactivity of tetrahedral first-row transition-metal centers
supported by ToM, in which redox-active metal centers could
lead to new chemistry and catalysis. In particular, the electron-
donating properties and bowl-like steric environment[14] pro-
vided by the coordination pocket of ToM could stabilize reactive
[a] Department of Chemistry, Iowa State University,
1605 Gilman Hall, Ames, IA 50011, USA
E-mail: sadow@iastate.edu
http://www.chem.iastate.edu/faculty/Aaron_Sadow/sadow-group
[b] US Department of Energy Ames Lab,
Ames, IA 50011, USA
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/ejic.201600237.
Eur. J. Inorg. Chem. 2016, 2486–2494
2486
© 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
groups bonded to metal centers with accessible d electrons to of the rings suggested the tridentate coordination of the ToM
facilitate catalytic chemistry. Moreover, a series of optically ac- ligand to the cobalt center. A 11B NMR signal at δ = –29 ppm
tive tris(4-R-2-oxazolinyl)phenylborate ligands are also available was shifted significantly from the isotropic value observed for
for imposing chiral environments on tetrahedrally coordinated the ToM ligand in diamagnetic environments (e.g., the 11B NMR
first-row transition-metal sites.[15]
chemical shift of TlToM is δ = –16 ppm). This paramagnetically
shifted 11B NMR resonance further indicated the successful con-
version of diamagnetic TlToM to a new species, and we note
that the 11B NMR spectrum is useful for counting the number
of ToM species in the reaction mixture.[16]
The present study describes the synthesis and characteriza-
tion of ToMCoCl (1) as a precursor to reactive complexes. We
also provide syntheses and characterization data of ToM2Co (2)
and {HToM}CoCl2 (3), which are side-products discovered during
exploratory syntheses toward 1. The open steric profile of the
The infrared spectrum contained a band at ν = 1598 cm–1
˜
ToM ligand results in additional synthetic challenges compared (KBr), which was assigned to the C=N stretching mode (νCN) of
with syntheses of sterically encumbered tris(pyrazolyl)borate- the oxazoline groups. Typically, higher-energy νCN bands (ν >
˜
based first-row metal compounds, in terms of controlling the 1615 cm–1) are observed for dissociated oxazoline moieties in
formation of 1, 2, and 3. The chlorido ligand in 1 may be substi- oxazolinylborate ligands; therefore, this data further supports
tuted through salt-metathesis reactions to give ToMCoOtBu (4) tridentate ligand coordination. In addition, the observation of
and ToMCoOAc (5) without ToM transmetalation or ligand redis- only a single peak from the symmetric mode (the asymmetric
tribution demonstrating that 1 is a viable synthetic precursor νCN band had low intensity) in the solid-state and solution IR
for new cobalt(II) compounds. Complex 5 was studied as a cata- spectra (at ν = 1586 cm–1) suggested a similar configuration in
˜
lyst for the oxidation of cyclohexane. Although the selective the two phases. The tetrahedral geometry of 1 was also sup-
oxidation of cyclohexane remains a challenge, cobalt complexes ported by the electronic absorption spectrum (Figure 1), which
have shown great promise in this area.[9d]
contained a band from λ ≈ 550 to 700 nm with λmax at 568 (ε =
362
M
–1 cm–1) and 635 nm (ε = 641 –1 cm–1). The larger peak
M
was sandwiched by shoulders at λ ≈ 600 and 660 nm. The spec-
tra of high-spin pseudotetrahedral cobalt(II) complexes show
peaks in this range, and these signals are likely related to the
many-featured ν3 band [4T1(P)←4A2(F)] observed for [CoCl4]2–.[17]
The ToMCoCl spectrum is also similar to the spectra of re-
lated tris(pyrazolyl)borate (Tp) cobalt(II) complexes such as
TptBu,MeCoCl with maxima at λ = 526, 602, 636, and 660 nm[18]
or TpiPr2CoEt, which exhibited maxima at λ = 580, 610, and
690 nm.[19] For tetrahedral CoII complexes, two lower-energy
Results and Discussion
The reaction of TlToM and CoCl2·THF (THF = tetrahydrofuran)
affords ToMCoCl (1) as a bright blue solid in excellent yield
[Equation (1)]. ToMCoCl is paramagnetic, and its identity as a
pseudotetrahedral d7 center is supported by a host of charac-
terization data including 1H and 11B NMR spectroscopy, IR spec-
troscopy, UV/Vis spectroscopy, ESR spectroscopy, X-ray diffrac-
tion studies, elemental analysis, and magnetic measurements
that included both magnetometry and the Evans method. Many
of these tools were needed initially to interpret the spectro-
scopic data, as the identities of the paramagnetic products were
challenging to establish, and later to demonstrate the purities
of the isolated materials. As outlined below, the ratio of TlToM/
CoCl2, the use of CoCl2·THF, and the choice of solvent are crucial
to the high-yielding synthesis of 1 without contamination with
the readily formed side-products ToM2Co (2) and {HToM}CoCl2
(3).
4
transitions [4T1(F)←4A2(F) and T2(F)←4A2(F)] are also expected.
Indeed, a weak absorption was also observed at λ ≈ 1000 nm.
(1)
The signals in the 1H NMR spectrum of ToMCoCl dissolved in
[D6]benzene were broad and dispersed over a large chemical-
shift range, as expected for a paramagnetic compound. Despite
1
these spectroscopic effects, the number of H NMR signals and
Figure 1. UV/Vis spectra of ToMCoCl (1), ToM2Co (2), {HToM}CoCl2 (3), and
ToMCoOtBu (4).
their integrated ratio provided characteristic data associated
with a pseudo-C3v-symmetric oxazolinylborate species. Broad
resonances at δ = 8.38 (18 H) and 24.88 ppm (6 H) were as-
signed to the methyl and methylene groups, respectively, of
The paramagnetic nature of 1 was further investigated by
equivalent oxazoline rings in the ToM ligand. The equivalence magnetic susceptibility measurements and ESR spectroscopy.
Eur. J. Inorg. Chem. 2016, 2486–2494
www.eurjic.org
2487
© 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
The Evans method revealed a solution magnetic moment of
Although the reaction of TlToM and CoCl2·THF is reproduci-
4.5(2) μB. This data is consistent with a high-spin cobalt(II) com- ble, the choice of CoCl2·THF is critical to a successful and high-
plex (S = 3/2, for which spin-only μeff = 3.9 μB) with the 4A2 yielding synthesis. In micromolar-scale reactions, equimolar
ground state, as discussed above. This electronic configuration amounts of TlToM and CoCl2·THF provide 1 quantitatively. Ex-
was maintained at low temperature according to magnetome- cess CoCl2·THF (1.5 equiv.) is needed in larger-scale prepara-
try measurements, which showed μeff = 4.3(1) μB at 10.3 K. The tions to optimize the ToMCoCl yield (calculated with respect to
ESR spectrum (X-band, 10 K) of a neat point sample revealed TlToM), because the separation of ToMCoCl from CoCl2·THF is
the low, rhombic site symmetry. The ESR spectrum was simu- easier than that from (ToM)2Co (2). In contrast, the reaction of
lated for an S = 3/2 spin system with anisotropic g values of anhydrous CoCl2 and TlToM in THF provides a purple mixture of
gx = 5.96(2), gy = 3.50(2), and gz = 2.10(2).[20] Similar g values 1 and 2. During our initial synthetic studies, the appearance of
have been reported for other pseudotetrahedral cobalt(II) com- two sets of signals in the 1H NMR spectra and the formation
plexes featuring rhombic symmetry.[21,22] Hyperfine coupling to of purple material in these reactions made the isolation and
59Co (I = 7/2) was not detected under these conditions.[21]
Blue crystals of 1, obtained from a saturated toluene solution
cooled to –38 °C, were determined to be ToMCoCl through an
X-ray diffraction study (Figure 2). The X-ray crystal structure re-
vealed a pseudotetrahedral cobalt center coordinated by the
characterization of pure ToMCoCl challenging. The 1H NMR spec-
trum of the purple reaction mixture contained signals at δ =
15.1 and 12.5 ppm as well as the resonances later assigned to
1. The formation of two paramagnetic ToMCo species was also
suggested by the two 11B NMR signals at δ = –29 and 42 ppm;
the former was assigned to ToMCoCl, and the latter was as-
signed to (ToM)2Co on the basis of a single-crystal X-ray diffrac-
tion study and its independent synthesis, which is described
below. Repeated recrystallizations from saturated toluene solu-
tions at –38 °C yielded purple X-ray-quality crystals, and a sin-
gle-crystal diffraction study revealed that the substance was
(ToM)2Co (2; Figure 3).
tridentate tris(oxazolinyl)borate ligand. Compound
1 and
ToMZnCl[23] have similar structural features. For example, the
Co–N and Zn–N bond lengths range from 2.0091(9) to
2.040(1) Å; the zinc complex has both the highest and the low-
est values, and the cobalt compound has intermediate bond
lengths. The N–M–Cl angles range from 115.33(4) to 130.82(4)°
for the two compounds, and the cobalt compound has the ex-
treme angles. In addition, the B–M–Cl angles of 170.99(3) and
174.27(2)° for the cobalt and zinc complexes show a slight dis-
placement of the Cl atom from the anticipated position in a
pseudo-C3v structure, and this distortion is slightly larger for the
cobalt complex. The structural similarity of diamagnetic zinc(II)
and paramagnetic cobalt(II) compounds suggests that these
distortions are sterically controlled rather than the result of
electronic influences. The Co1–Cl1 bond length of 2.2025(4) Å
in 1 is similar to that observed in other four-coordinate co-
balt(II) scorpionate complexes [e.g., TpPh,MeCoCl, 2.2004(9) Å;
TptBu,HCoCl, 2.216(2) Å; and TptBu,MeCoCl, 2.2204(9) Å].[18,24]
Figure 3. Thermal ellipsoid plot of (ToM)2Co (2) with ellipsoids at 35 %
probability. H atoms and two cocrystallized toluene molecules are omitted
for clarity. Selected bond lengths [Å] and angles [°]: Co1–N1 2.010(4), Co1–
N3 2.003(4), Co1–N4 2.013(4), Co1–N5 2.000(4); N1–Co1–N3 97.2(2), N4–Co1–
N5 97.8(2), N1–Co1–N4 114.2(2), N3–Co1–N4 116.4(2), N5–Co1–N3 116.4(2),
N5–Co1–N1 116.0(2).
The cobalt center in 2, as in 1, has a distorted tetrahedral
geometry. Each of the tris(oxazolinyl)borate ligands in 2 is bi-
dentate with one non-coordinated oxazoline ring, and the
structure is similar to the solid-state structures of (ToM)2Mg and
(ToM)2Zn.[13,25] Although the (κ2-ToM)2Co structure forms readily,
the ToM ligand apparently does not support six-coordinate
pseudo-sandwich-type first-row metal compounds, in contrast
to (κ3-Tp*)2M compounds.[26] Moreover, we have not observed
disproportionation of ToMCoCl into ToM2Co and CoCl2; there-
fore, ToMCoCl may be a suitable precursor for the synthesis of
new inorganic compounds through chloride substitution. The
crystallographic data show that the metal centers are sterically
protected by the oxazoline methyl groups, and this likely pre-
vents the formation of the six-coordinate structure. Compound
2 has shorter Co–N interatomic distances than 1, and all of the
Figure 2. Thermal ellipsoid plot of ToMCoCl (1). Ellipsoids are plotted at 50 %
probability. H atoms are omitted for clarity. Selected bond lengths [Å] and
angles [°]: Co1–Cl1 2.2026(5), Co1–N1 2.011(1), Co1–N2 2.030(1), Co1–N3
2.026(1); Cl1–Co1–N1 115.31(4), Cl1–Co1–N2 130.80(4), Cl1–Co1–N3 121.73(4),
B1–Co1–Cl1 170.99(3).
Eur. J. Inorg. Chem. 2016, 2486–2494
www.eurjic.org
2488
© 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
N–Co–N angles in 2 are larger than those in 1. The interatomic perature was confirmed by magnetometry experiments, which
distances and angles of 2 fall between those of (ToM)2Mg and revealed μeff = 3.9(1) μB at 12.5 K. An ESR spectrum of 2 ac-
(ToM)2Zn and bear greater resemblance to those of the latter. quired at 5 K as a point sample or as a glassed solution in 2-
For example, the Co–N bonds in 2 are shorter than those in methyltetrahydrofuran contained only weak signals (see Fig-
(ToM)2Mg by ca. 0.05 Å but are equivalent within error to the ure S18 for the ESR spectrum of 2 as a point sample). One of the
Zn–N bonds encountered in (ToM)2Zn.
signals for the point sample showed evidence of the hyperfine
coupling of the cobalt center in the form of a seven-line pat-
tern. The discrepancy from the expected eight-line pattern for
I = 7/2 was presumably caused by the broadness at the onset
of the peak. In addition, the spectrum was consistent with a
low-symmetry species.
Unfortunately, purification by crystallization is not effective
for preparative-scale separations of mixtures of (ToM)2Co and
1
ToMCoCl. Moreover, the H NMR spectrum of the reaction mix-
ture did not contain features that easily identified inequivalent
oxazoline rings in the bidentate-coordinated structure. We pre-
pared 2 independently to confirm that the unidentified signals
Compound 1 and TlToM react overnight to give 2, and this
in the 1H NMR spectrum of the mixture of cobalt species corre- conversion occurs more rapidly than the reaction of TlToM and
sponded to 2. The reaction of CoCl2 and 2 equiv. of TlToM af- CoCl2 under equivalent conditions. In the attempted synthesis
fords purple (ToM)2Co in excellent yield [Equation (2)]. (ToM)2Co of 1 from anhydrous CoCl2, the poor selectivity is hypothesized
and ToMCoCl are similarly soluble in benzene, toluene, diethyl to come from the relative rates of transmetalation for CoCl2
ether, THF, and dichloromethane, and this similarity likely versus 1, and these rates are likely influenced by the heteroge-
causes the difficulties with the separation of mixtures of the neous nature of the reaction mixture. This contrasts with the
two compounds.
reactivity of bulky [tris(pyrazolyl)borato]cobalt compounds. For
example, TpPh,MeCoCl and TpPh,MeTl do not afford TpPh,Me2Co.[11]
In the course of our studies of the selective preparation of
1, polar solvents were tested to give a monophasic reaction
(i.e., to improve the kinetics). In fact, the reaction of TlToM and
CoCl2 in methanol gives a blue material, which was initially at-
1
tributed to ToMCoCl on the basis of its color and H NMR spec-
trum in [D6]benzene, as a result of the presence of 1 in the
mixture. Moreover, the electronic absorption spectra of 1 and
the substance obtained from the reaction in methanol are simi-
lar (see Figure 1). However, this material is less soluble in benz-
ene than 1, and the XRD powder pattern of this material did
not match that expected for 1 (see Figure 5 below). The recrys-
tallization of this material from dichloromethane afforded blue
X-ray-quality crystals, and a diffraction study proved that this
product is the protonated oxazoline derivative of 1, namely,
{HToM}CoCl2 (3, Figure 4).
(2)
The H and 11B NMR spectra of isolated 2 contained signals
that matched those detected in the crude mixture with 1; there-
fore, (ToM)2Co is indeed the second product obtained when an-
hydrous CoCl2 is the starting material. However, these NMR
spectroscopic data did not allow the direct quantitative estima-
tion of the purity and offered few further structural insights. In
1
contrast, the bands at ν = 1603 and 1554 (KBr) or 1609 and
˜
1555 cm–1 (CH2Cl2) in the infrared spectrum of 2 were structur-
ally informative. These signals were assigned to the νCN modes
of non-coordinated and coordinated oxazoline groups, respec-
tively, consistent with the bidentate bonding mode determined
by X-ray crystallography.
Furthermore, the d–d transitions in the UV/Vis spectrum of
2 again appeared in the expected region for a high-spin tetra-
hedral CoII complex at λmax = 562 (ε = 539
M
–1 cm–1) and
576 nm (ε = 552 M
–1 cm–1). These signals were similar to those
observed for the four-coordinate C2-symmetric Bp2Co species
–
[Bp = H2B(C3N2H3)2 ].[10a] The center of gravity of this band was
blueshifted with respect to that for the related transition in 1
(see the absorption spectra in Figure 1). Note that a similar
relationship was described between Bp2Co and TpCoCl, in
4
terms of a blueshifted absorption for the T1(P)←4A2(F) transi-
tion (for the tetrahedral site).[18] A second, lower-energy band
was observed at λ ≈ 1040 nm (ε = 120 M
–1 cm–1), and this band
was redshifted and more intense than the band for the transi-
tion in ToMCoCl. The high-spin nature of 2 was further sup-
ported by the solution magnetic moment of 4.2(2) μB (meas-
ured by the Evans method), which was slightly smaller than
that of Bp2Co but significantly reduced in comparison with that
of six-coordinate Tp2Co.[10a] The high-spin state of 2 at low tem-
Figure 4. Thermal ellipsoid plot of {HToM}CoCl2 (3) plotted at 50 % probability.
H atoms, with the exception of H1n, are omitted for clarity. Selected bond
lengths [Å] and angles [°]: Co1–N1 2.006(3), Co1–N2 2.005(4), Co1–Cl1
2.237(3), Co1–Cl2 2.261(2); N1–Co1–N2 96.2(1), N2–Co1–Cl1 109.77(9), N1–
Co1–Cl1 115.18(9), N2–Co1–Cl2 111.68(9), N1–Co1–Cl2 110.6(1), Cl1–Co1–Cl2
112.40(9).
Eur. J. Inorg. Chem. 2016, 2486–2494
www.eurjic.org
2489
© 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
The crystal structure of 3 shows a tetrahedral cobalt center showed paramagnetically shifted 1H NMR signals that were un-
featuring two chlorido ligands and a bidentate ToM–cobalt in- assignable yet characteristic (Figure S5). The 11B NMR signal ap-
teraction. The third oxazoline ring is protonated, and HToM may peared at δ = –7.5 ppm, and this peak was much less shifted
be viewed as an overall charge-neutral ligand coordinated to from the diamagnetic range than those of 1 and 2. The infrared
CoCl2. Thus, {HToM}CoCl2 differs from 1 and 2 by its zwitterionic spectrum was consistent with the crystallographically deter-
nature. Related iridium and rhodium compounds of protonated mined structure and provided structural insights. The bands at
or methylated tris(oxazolinyl)borate complexes have been re- ν = 1588 and 1598 cm–1 were assigned to νCN of cobalt-coordi-
˜
ported, as have zwitterionic cobalt(II) complexes stabilized by nated oxazoline and protonated oxazoline, respectively. Com-
bulky tris(carbene)borate ligands.[16,27] Compared with 1, com- pound 3 was also a high-spin CoII species, on the basis of mag-
pound 3 features slightly longer Co–Cl bonds (by more than netometry experiments that indicated a μeff of 4.2(1) μB at 5 K.
0.034 Å) and shorter Co–N bonds (by more than 0.025 Å).
The reaction of 1 and [HOEt2]Cl provides an independent
synthesis of 3 [Equation (3)]. Alternatively, LiToM and CoCl2 react
in wet dichloromethane to provide crystals of 3 directly from
the reaction mixture.
As all three compounds were formed from the same starting
materials and 1H NMR spectroscopy provided ambiguous re-
sults during the course of these synthetic studies, we instead
utilized powder XRD measurements to characterize the bulk
compositions of the samples. These experiments, along with
solution and solid-state IR spectroscopy, provide a connection
between the solid-state and molecular structures that usually
relies on solution-phase NMR spectroscopy [which was structur-
ally uninformative for these (oxazolinylborato)cobalt(II) com-
pounds]. Powder XRD is an appealing alternative characteriza-
tion technique, because it is not complicated by the electronic
structure or unpaired electrons and probes the composition of
the crystalline component of the bulk powder.
(3)
The experimental powder XRD patterns matched the corre-
sponding XRD patterns calculated from the single-crystal data
for samples of 1, 2, and 3 (Table S1). These results provided
additional confidence that the single-crystal diffraction experi-
ments corresponded to structures that describe the bulk sam-
ples. For example, the powder XRD experiments indicated that
the preparation from CoCl2·THF gives a reproducible and high-
yielding route to 1 (Figure 5).
Compound 3 is highly soluble in dichloromethane and meth-
anol but poorly soluble in benzene, toluene, tetrahydrofuran,
and diethyl ether. The poor solubility of 3 and the high solubil-
ity of 1 in [D6]benzene can result in erroneous interpretation of
the purity of ToMCoCl on the basis of the NMR spectra of crude
reactions mixtures. Thus, dry [D2]dichloromethane is the best
solvent for unambiguous NMR analysis of the purity of these
The viability of this method for the measurement of the pu-
compounds. The 1H NMR spectrum of 3 in dichloromethane rity was further tested with analytically pure 3/1 in a 50:50 ratio.
Figure 5. X-ray powder diffraction pattern of ToMCoCl (1) at 293 K with Cu-Kα radiation. The observed pattern (A) is shown in black, and the calculated intensity
pattern (B) is indicated by red solid lines.
Eur. J. Inorg. Chem. 2016, 2486–2494
www.eurjic.org
2490
© 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
The XRD pattern of this mixture matched the expected powder
pattern and verified that the presence of the two products
could be identified by the diffraction method.
We also attempted to synthesize ToMCoX compounds
using cobalt acetate as the starting material. However, the reac-
tion of cobalt acetate and TlToM generated a mixture of 2 and
another species, which was later identified as ToMCoOAc (see
below). Thus, CoCl2·THF is the preferred starting material for
the entry into the cobalt(II) chemistry of tris(oxazolinyl)borate
ligands.
With isolable, fully characterized, and spectroscopically and
analytically pure ToMCoCl in hand, we tested its reactivity in
halide substitutions to prepare ToMCoOtBu (4) and ToMCoOAc
(5). The reaction of ToMCoCl and KOtBu in tetrahydrofuran read-
ily provides ToMCoOtBu [Equation (4)].
Figure 6. Thermal ellipsoid plot of ToMCoOtBu (4). Ellipsoids are plotted at
50 % probability. H atoms are omitted for clarity. Selected bond lengths [Å]
and angles [°]: Co1–O4 1.821(2), Co1–N1 2.049(3), Co1–N2 2.055(3), Co1–N3
2.026(2); O4–Co1–N1 111.2(1), O4–Co1–N2 128.0(1), O4–Co1–N3 131.0(1), B1–
Co1–O4 166.2(1), Co1–O4–C22 134.2(2).
peared far downfield (δ = 171.25 ppm). The 11B NMR signal of
5 appeared at δ = 95 ppm, which is downfield relative to the
signals for 1 and 4.
(4)
1
In the H NMR spectrum of 4, all of the signals were readily
assigned by integration, including a new signal at δ =
11.16 ppm that integrated to 9 H relative to the ToM signals.
This resonance was assigned to the OtBu group. Moreover, a
new 11B NMR signal was detected at δ = 73 ppm, which was
further downfield than those of the other cobalt complexes.
The paramagnetically shifted 11B NMR spectrum rules out the
transmetalation of the ToM ligand to K.
(5)
In the IR spectrum, a single band corresponding to the ox-
azoline νCN stretching mode was observed at ν = 1591 cm–1.
˜
Unlike blue ToMCoCl, compound 4 is purple. Accordingly, the
UV/Vis absorption spectrum of 4 contained a broad peak with
a blueshifted onset edge at λ ≈ 500 nm, as might be expected
owing to the replacement of the weak-field chlorido ligand with
a tert-butoxide ligand. The tridentate coordination of the
tris(oxazolinyl)borate ligand to the cobalt center was suggested
Compound 5 is high spin [μeff = 4.8(2) μB], as determined by
the Evans method. X-ray-quality crystals were obtained by re-
crystallization from pentane at –35 °C (Figure 7). The Co–O
bonds [2.098(2) and 2.089(2) Å] are longer than the Co–O bond
by a single strong IR band at ν =1590 cm–1, which was assigned
˜
to the oxazoline νCN mode. Compound 4 is a high-spin cobalt(II)
complex from room temperature to 36 K, as determined by
magnetometry measurements [μeff = 4.6(1) μB at 36 K]. Unlike
compounds 1–3, ToMCoOtBu undergoes a broad, thermally in-
duced spin transition, and μeff is 1.3(1) μB at 2.5 K, consistent
with a low-spin state (S = 1/2).
The recrystallization of 4 from toluene at –35 °C provided X-
ray-quality crystals, and a single-crystal X-ray diffraction study
confirmed the identity of 4 as ToMCoOtBu (Figure 6). The Co–N
bonds in 4 are longer than those in 1. In addition, the B1–Co1–
O4 angle in 4 is smaller than the B1–Co1–Cl1 angle in 1. The
Co1–O4–C22 angle is 134.2(2)°, and this angle wedges the tBu
group between the methyl groups of the N2 and N3 oxazolines.
The reaction of ToMCoCl and NaOAc in THF provides
ToMCoOAc (5) as a light purple solid in high isolated yield [97 %;
Equation (5)]. Compound 5 was characterized by 1H and 11B
NMR spectroscopy. The 1H NMR spectrum contained new
tris(oxazolinyl)phenylborate signals, and the acetate signal ap-
Figure 7. Thermal ellipsoid plot of ToMCoOAc (5). Ellipsoids are plotted at
50 % probability. H atoms are omitted for clarity. Selected bond lengths [Å]
and angles [°]: Co1–O4 2.098(2), Co1–O5 2.089(1), Co1–N1 2.065(2), Co1–N2
2.064(2), Co1–N3 2.026(2); O5–Co1–N1 154.5(1), O5–Co1–N2 100.2(1), O5–
Co1–N3 112.4(1), O5–Co1–O4 61.8(1).
Eur. J. Inorg. Chem. 2016, 2486–2494
www.eurjic.org
2491
© 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
in 4 [1.821(2) Å], and the Co–N bonds are also longer than tration of cyclohexanol reaches a steady state after ca. 300 min.
those in 1 and 4.
The plot of [cyclohexanol] versus time in Figure 8 can be fitted
with a nonlinear least-squares regression (R = 0.98) to Equa-
tion (6).[30]
Cobalt(II) acetate is used as an industrial catalyst in the oxid-
ation of p-xylene to terephthalic acid,[28] and nitrogen-ligand-
based cobalt compounds have also been studied in cyclohex-
ane oxidation to cyclohexanol.[9d] In this oxidation catalysis, se-
lectivity for a single oxidation to cyclohexanol, rather than over-
oxidation to cyclohexanone or ε-caprolactone, is desired. De-
spite the important role of cobalt in oxidation chemistry,
TpCoX-based compounds are reported as not effective for cy-
clohexane oxidation,[29] so a study of ToMCoOAc as an oxidation
catalyst provides a comparison with the pyrazolylborate ana-
logues.
Complex 5 catalyzes the oxidation of cyclohexane to cyclo-
hexanol with meta-chloroperoxybenzoic acid (mCPBA) as the
terminal oxidant and affords 536 equiv. of cyclohexanol per
1 equiv. of catalyst after 7 h. At the early stages of the reaction
(3 h), the product ratio for cyclohexanol/cyclohexanone/ε-ca-
prolactone is 20:1.5:1 with 337 equiv. of cyclohexanol formed
per 1 equiv. of 5. This selectivity decreases as the reaction pro-
ceeds, and ε-caprolactone is formed from the reaction of
mCPBA and cyclohexanone. Cumene hydroperoxide and tert-
butyl hydroperoxide were tested as oxidants for this catalytic
transformation, but only mCPBA afforded the desired conver-
sion. A control experiment, in which cyclohexane and mCPBA
were mixed at room temperature, did not provide detectable
quantities of cyclohexanol, cyclohexanone, or ε-caprolactone.
Furthermore, anhydrous CoCl2 or anhydrous Co(OAc)2 [as
Co4(OAc)7OH] are poorly soluble under these reaction condi-
tions, and this limits comparisons with 5. However, the partially
soluble suspended materials give lower activities and poorer
selectivities than 5 under comparable conditions.
k1
1
2
[C6H12O]t = [C6H ]
{e–k t – e–k t
}
(6)
12 0 k2 – k1
From this analysis, the observed rate constant for the oxid-
ation of cyclohexane to cyclohexanol (k1) is 1.69(8) × 10–4 s–1,
and that for the oxidation of cyclohexanol is 3.9(1) × 10–3 s–1
(average of four experiments, see the Supporting Information
for individual curves). Thus, the selectivity for cyclohexanol pro-
duction comes largely from the high initial [C6H12]. On the basis
of these results, we are currently studying reaction conditions
(temperature, solvent polarity, and reagent concentrations) and
catalyst structures that favor higher k1 values and smaller k2
values.
Conclusions
The synthesis of 1 from TlToM and CoCl2·THF provides an entry
point into [tris(oxazolinyl)borato]cobalt chemistry. Like zinc and
magnesium compounds, four-coordinate pseudotetrahedral co-
balt complexes are supported by the tris(oxazolinyl)borate li-
gand, despite its reduced steric profile with respect to that of
the highly bulky tert-butyl-substituted tris(pyrazolyl)borate scor-
pionates. The starting materials and reaction conditions are crit-
ical for obtaining 1, rather than 2 or 3. However, these other
two compounds can be prepared by using 2 equiv. of TlToM
or polar protic solvents, respectively. Once the paramagnetic
compounds were fully characterized by X-ray diffraction as well
as IR, NMR, and UV/Vis spectroscopy, the characteristic spectro-
To better identify the parameters associated with the oxid-
ation process, the reaction mixture was monitored by GC to
determine the concentrations of cyclohexanol, cyclohexanone,
1
scopic signatures in the 11B NMR spectra, H NMR spectra, and
powder XRD patterns are useful for establishing the synthetic
reproducibility and product identities as well as the purities of
the materials.
and ε-caprolactone. In a reaction with 0.16 m 5, the concen-
M
ToMCoCl is reactive with KOtBu and NaOAc in salt-metathesis
reactions. In this context, we are currently working to synthe-
size, characterize, and study the reactivity of alkylcobalt(II) com-
pounds supported by tris(oxazolinyl)borate ligands. ToMCoOAc
is a catalyst for the selective oxidation of cyclohexane, in con-
trast to [tris(pyrazolyl)borato]cobalt compounds. We are cur-
rently preparing optically active analogues to study enantiose-
lective C–H bond functionalization.
Experimental Section
General Experimental Methods: All reactions were performed by
standard Schlenk techniques under dry argon. Tetrahydrofuran, di-
ethyl ether, and toluene were dried and deoxygenated with an IT
PureSolv system. [D6]Benzene was heated under reflux over Na/K
alloy and vacuum-transferred. CoCl2 (Strem) was used to prepare
CoCl2·THF by Soxhlet extraction of CoCl2 with THF. TlToM was syn-
thesized according to reported procedures.[16,31]
Caution! Thallium salts are highly toxic, and thallium-containing
compounds and waste should be handled appropriately.
Figure 8. Concentration versus time plots for the products of cyclohexane
oxidation with mCPBA. The experiment was performed four times.
Eur. J. Inorg. Chem. 2016, 2486–2494
www.eurjic.org
2492
© 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
KOtBu was purified by sublimation. mCPBA was purified by washing
with K2HPO4/KH2PO4 buffer solution (pH 7.5).[32] The 1H and 11B
NMR spectra were recorded with a Bruker Avance III 600 spectrome-
ter. The 11B NMR spectra were referenced to an external sample of
BF3·Et2O. The infrared spectra were measured with a Bruker Vertex
80 FTIR spectrometer. The electron paramagnetic resonance (EPR)
spectra were obtained with an X-band Elexsys 580 FT-EPR spectrom- 999 (130
eter in continuous-wave mode, and the spectra were simulated with (513.1): calcd. C 49.16, H 5.89, N 8.19; found C 49.35, H 6.16, N 7.72.
XSophe. The direct current (DC) magnetization was measured with M.p. 218–221 °C (dec.).
{HToM}CoCl2. 1H NMR ([D2]dichloromethane, 600 MHz): δ = 17.3,
14.3, 9.9, 8.6, 7.0, 6.6, –21.4, –23.3 ppm. 11B NMR ([D2]dichlorometh-
ane, 128 MHz): δ = –7.5 ppm. IR (KBr): ν = 3270 (m, νNH), 2976 (m),
˜
2932 (m), 2984 (m), 1598 (s, νCN), 1588 (s, νCN), 1462 (m), 1424 (m),
1371 (m), 1306 (m), 1274 (m), 1194 (m) 1171 (m), 965 (m), 937 (m)
cm–1. UV/Vis (CH2Cl2): λmax (ε) = 567 (360), 614 (479), 635 (528),
M
–1 cm–1) nm. μeff (CD2Cl2) = 4.0(2) μB. C21H30BCl2CoN3O3
ToMCoOtBu (4): Complex 1 (0.202 g, 0.423 mmol) was dissolved in
THF (10 mL), and the solution was added to KOtBu (0.053 g,
0.472 mmol) to instantly afford a purple solution. The reaction mix-
ture was stirred overnight, and then KCl was removed by filtration.
The volatiles were evaporated in vacuo to afford a bright purple
solid. The purple solid was washed with pentane (3 × 5 mL) and
then dried under vacuum to yield ToMCoOtBu (0.189 g, 0.367 mmol,
86.8 %). X-ray-quality crystals were obtained from pentane at
–40 °C. 1H NMR ([D6]benzene, 600 MHz): δ = 15.54 (s, 6 H,
CNCMe2CH2O), 11.33 (s, 2 H, C6H5), 11.16 (s, 9 H, CoOCMe3), 9.51 (s,
2 H, C6H5), 8.00 (s, 1 H, p-C6H5), –6.26 (s, 18 H, CNCMe2CH2O) ppm.
a Quantum Design MPMS-5 superconducting quantum interference
device (SQUID) magnetometer. The UV/Vis spectra were recorded
with an Agilent 8453 UV/Vis spectrophotometer with the analyte
(2 mM) in dichloromethane. Elemental analyses were performed
with a Perkin–Elmer 2400 Series II CHN/S analyzer. The catalytic
cyclohexane oxidation experiments and the kinetic studies were
performed with a Chemspeed Technologies SWING-XL automated
platform. GC–MS was performed with Agilent 7890A GC and Agilent
5975C MS instruments equipped with an HP-5MS column. The pow-
der XRD patterns were collected with a STOE WinXPOW powder
diffractometer. The single-crystal X-ray diffraction data were col-
lected with an APEX II diffractometer, as described in the Support-
ing Information. CCDC 1433255 (for 1), 1433256 (for 2), 1433257 (for
3), 1433258 (for 4), and 1457506 (for 5) contain the supplementary
crystallographic data for this paper. These data can be obtained
free of charge from The Cambridge Crystallographic Data Centre.
11B NMR ([D6]benzene, 128 MHz): δ = 72.9 ppm. IR (KBr): ν = 2966
˜
(m), 2930 (m), 2889 (m), 1590 (s, νCN), 1465 (m), 1433 (m), 1386 (m),
1364 (m), 1278 (m), 1251 (m), 1193 (s), 1158 (m), 1002 (m), 966 (m),
928 (m) cm–1. UV/Vis (CH2Cl2): λmax (ε) = 524 (245), 575 (327), 624
(179), 657 (148 M
–1 cm–1) nm. C25H38BCoN3O4 (514.3): calcd. C 58.38,
H 7.45, N 8.17; found C 58.04, H 7.48, N 8.29. M.p. 233–235 °C (dec.).
ToMCoCl (1): TlToM (2.015 g, 3.43 mmol) was dissolved in THF
(75 mL), and the solution was added dropwise to a stirred suspen-
sion of CoCl2·THF (1.01 g, 5.11 mmol) in THF (75 mL). The reaction
mixture was stirred overnight, and then the solvent was evaporated
to give a blue residue. The product was extracted with benzene
(100 mL) and dried under vacuum to afford ToMCoCl (1) as a bright
blue solid (1.55 g, 3.25 mmol, 94.8 % based on TlToM). Recrystalliza-
tion from toluene gave analytically pure ToMCoCl. 1H NMR ([D6]-
benzene, 600 MHz): δ = 24.88 (s, 6 H, CNCMe2CH2O), 8.38 (s, 18 H,
CNCMe2CH2O), 4.77 (s, 2 H, C6H5), 4.25 (s, 1 H, p-C6H5), –0.41 (s, 2
H, C6H5) ppm. 11B NMR ([D6]benzene, 128 MHz): δ = –29.6 ppm. IR
ToMCoOAc (5): Complex 1 (0.101 g, 0.211 mmol) was dissolved in
THF (5 mL), and the solution was added to NaOAc (0.017 g,
0.212 mmol). The reaction mixture was stirred for 24 h to afford a
purple solution and a white precipitate. The NaCl precipitate was
removed by filtration, and the volatiles were evaporated in vacuo
to afford a light purple solid. The resulting purple solid was washed
with pentane (3 × 5 mL) and dried under vacuum to yield
ToMCoOAc (0.102 g, 0.204 mmol, 96.7 %). X-ray-quality crystals were
1
obtained from pentane at –40 °C. H NMR ([D6]benzene, 600 MHz):
δ = 171.25 (s, 3 H, O2CMe), 33.36 (s, 2 H, C6H5), 17.86 (s, 2 H, C6H5),
15.06 (s, 1 H, p-C6H5), 12.40 (s, 6 H, CNCMe2CH2O), 42.84 (s, 18 H,
(KBr): ν = 2969 (m), 2930 (m), 2898 (m), 2871 (m), 1588 (s, νCN), 1461
˜
CNCMe2CH2O) ppm. 11B NMR ([D6]benzene, 128 MHz):
δ =
(m), 1388 (m), 1369 (m), 1351 (m), 1273 (m), 1193 (m) 1162 (m),
1107 (m), 988 (s) cm–1. UV/Vis (CH2Cl2): λmax (ε) = 568 (362), 635
˜
95.3 ppm. IR (KBr): ν = 2964 (m), 2930 (m), 2894 (m), 1591 (s, νCN),
(641), 940 (92 M
–1 cm–1) nm. Evans method: μeff (C6D6) = 4.5(2) μB.
1563 (m), 1463 (m), 1442 (m), 1368 (m), 1351 (m), 1275 (m), 1197
(m), 1162 (m), 1098 (m), 1024 (m) 960 (m) cm–1. UV/Vis (CH2Cl2)
C21H29BClCoN3O3 (476.7): calcd. C 52.91, H 6.13, N 8.82; found C
53.26, H 6.20, N 9.03. M.p. 194–196 °C (dec.).
λmax (ε) = 486 (51), 585 (104
M
–1 cm–1) nm. Evans method: μeff
(C6D6) = 4.8(2) μB. C23H32BCoN3O5 (500.3): calcd. C 55.22, H 6.45, N
8.40; found C 55.21, H 6.58, N 7.79. M.p. 161–164 °C (dec.).
(ToM)2Co (2): TlToM (0.102 g, 0.173 mmol) and CoCl2·THF (0.0169 g,
0.0855 mmol) were stirred in THF (5 mL) for 1 d. The reaction mix-
ture changed gradually from blue to purple. The reaction mixture
was filtered, and the solvent was evaporated to afford ToM2Co (2)
as a purple solid (0.0689 g, 0.0837 mmol, 97.8 %). Recrystallization
from toluene afforded analytically pure X-ray-quality single crystals
of (ToM)2Co. 1H NMR ([D6]benzene, 600 MHz): δ = 15.04, 12.41,
2.63 ppm. 11B NMR ([D6]benzene, 128 MHz): δ = 42.3 ppm. IR (KBr):
ToMCoOAc-Catalyzed Oxidation of Cyclohexane: The oxidation of
cyclohexane was performed under conditions similar to those previ-
ously reported by Hikichi et al.[9d] A reaction flask was charged with
cyclohexane (1.6 mL, 15 mmol, 2.45
0.16 m ) dissolved in dichloromethane/acetonitrile (95 %, 1 mL).
M
) and ToMCoOAc (1 μmol,
M
Upon the addition of 1,2-dichloroethane/acetonitrile (3.5 mL,
97.5 %) containing mCPBA (345 mg, 2.0 mmol) and nitrobenzene
(30 μL, 300 μmol), the reaction was initiated and maintained at
25 °C. At each time point, an aliquot (0.2 mL) was removed and
quenched with triphenylphosphine (10 mg, 38.1 μmol). The organic
products were identified and quantified by GC–MS through the in-
tegration of the peak areas with respect to a known amount of
the nitrobenzene standard. In this analysis, the quenched reaction
product (10 μL) was added to dichloromethane (1.5 mL) and ana-
lyzed by GC–MS. Each chromatogram was obtained under the fol-
lowing conditions: split, 25:1; inlet temperature, 250 °C; initial oven
ν = 2965 (m), 2928 (m), 2888 (m), 1602 (m, νCN), 1554 (s, νCN), 1463
˜
(m), 1369 (m), 1279 (m), 1261 (m), 1198 (m) 1153 (m), 1100 (m),
1002 (m), 969 (m) cm–1. UV/Vis (CH2Cl2): λmax (ε) = 562 (539), 576
(552), 1042 (120 M
–1 cm–1) nm. Evans method: μeff (C6D6) = 4.2(2) μB.
C42H58B2CoN6O6 (823.5): calcd. C 61.26, H 7.10, N 10.21; found C
61.21, H 7.20, N 9.78. M.p. 121–123 °C (dec.).
{HToM}CoCl2 (3): Complex 1 (0.199 g, 0.417 mmol) was dissolved
in THF (30 mL), and the solution was added dropwise to a solution
of HCl in diethyl ether (210 μL, 0.420 mmol). The blue solution was
stirred overnight, and the volatiles were removed in vacuo to afford temperature, 45 °C; temperature ramp, 15 °C min–1 to 150 °C. From
{HToM}CoCl2 (3) as a bright blue solid (0.144 g, 0.281 mmol, 67.4 %).
Recrystallization from dichloromethane afforded analytically pure
this data, the concentration of cyclohexanol versus time was ana-
lyzed with a nonlinear least-squares regression.
Eur. J. Inorg. Chem. 2016, 2486–2494
www.eurjic.org
2493
© 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
[12] D. Mukherjee, R. R. Thompson, A. Ellern, A. D. Sadow, ACS Catal. 2011, 1,
698–702.
Acknowledgments
This research was supported by the U.S. Department of Energy
through the Ames Laboratory (Contract No. DE-AC02-
07CH11358). Synthesis and catalysis work were supported by
the Office of Basic Energy Sciences, Division of Chemical Scien-
ces, Geosciences, and Biosciences. The powder XRD and SQUID
measurements were supported by the Division of Materials Sci-
ence. The Chemspeed SwingXL Automated Robotic Catalysis
Platform was supported by the Critical Materials Institute, an
Energy Innovation Hub funded by the U.S. Department of En-
ergy, Office of Energy Efficiency and Renewable Energy, Ad-
vanced Manufacturing Office.
[13] D. Mukherjee, A. Ellern, A. D. Sadow, J. Am. Chem. Soc. 2012, 134, 13018–
13026.
[14] K. Wu, D. Mukherjee, A. Ellern, A. D. Sadow, W. E. Geiger, New J. Chem.
2011, 35, 2169–2178.
[15] a) B. Baird, A. V. Pawlikowski, J. Su, J. W. Wiench, M. Pruski, A. D. Sadow,
Inorg. Chem. 2008, 47, 10208–10210; b) S. R. Neal, A. Ellern, A. D. Sadow,
J. Organomet. Chem. 2011, 696, 228–234; c) S. Xu, Y. Magoon, R. R. Reinig,
B. M. Schmidt, A. Ellern, A. D. Sadow, Organometallics 2015, 34, 3508–
3519.
[16] H.-A. Ho, J. F. Dunne, A. Ellern, A. D. Sadow, Organometallics 2010, 29,
4105–4114.
[17] F. A. Cotton, D. M. L. Goodgame, M. Goodgame, J. Am. Chem. Soc. 1961,
83, 4690–4699.
[18] J. Krzystek, D. C. Swenson, S. A. Zvyagin, D. Smirnov, A. Ozarowski, J.
Telser, J. Am. Chem. Soc. 2010, 132, 5241–5253.
Keywords: Scorpionates · Oxazolines · Cobalt · Structure
elucidation · Oxidation
[19] N. Shirasawa, T. T. Nguyet, S. Hikichi, Y. Moro-oka, M. Akita, Organometal-
lics 2001, 20, 3582–3598.
[20] G. R. Hanson, K. E. Gates, C. J. Noble, M. Griffin, A. Mitchell, S. Benson, J.
Inorg. Biochem. 2004, 98, 903–916.
[21] H. Drulis, K. Dyrek, K. P. Hoffmann, S. K. Hoffmann, A. Weselucha-Birczyn-
ska, Inorg. Chem. 1985, 24, 4009–4012.
[1] S. Trofimenko, Scorpionates: The Coordination Chemistry of Polypyrazolyl-
borate Ligands, Imperial College Press, London, 1999.
[22] a) A. Bencini, C. Benelli, D. Gatteschi, C. Zanchini, Inorg. Chem. 1979, 18,
2137–2140; b) A. Romerosa, C. Saraiba-Bello, M. Serrano-Ruiz, A. Canes-
chi, V. McKee, M. Peruzzini, L. Sorace, F. Zanobini, Dalton Trans. 2003,
3233–3239; c) J. M. Zadrozny, J. Liu, N. A. Piro, C. J. Chang, S. Hill, J. R.
Long, Chem. Commun. 2012, 48, 3927–3929.
[23] D. Mukherjee, A. Ellern, A. D. Sadow, J. Am. Chem. Soc. 2010, 132, 7582–
7583.
[24] a) K. Uehara, S. Hikichi, M. Akita, J. Chem. Soc., Dalton Trans. 2002, 3529–
3538; b) I. B. Gorrell, G. Parkin, Inorg. Chem. 1990, 29, 2452–2456.
[25] J. F. Dunne, D. B. Fulton, A. Ellern, A. D. Sadow, J. Am. Chem. Soc. 2010,
132, 17680–17683.
[26] S. Trofimenko, J. Am. Chem. Soc. 1967, 89, 6288–6294.
[27] a) A. V. Pawlikowski, T. S. Gray, G. Schoendorff, B. Baird, A. Ellern, T. L.
Windus, A. D. Sadow, Inorg. Chim. Acta 2009, 362, 4517–4525; b) R. E.
Cowley, R. P. Bontchev, E. N. Duesler, J. M. Smith, Inorg. Chem. 2006, 45,
9771–9779.
[2] a) J. D. Jewson, L. M. Liable-Sands, G. P. A. Yap, A. L. Rheingold, K. H.
Theopold, Organometallics 1999, 18, 300–305; b) R. Han, I. B. Gorrell,
A. G. Looney, G. Parkin, J. Chem. Soc., Chem. Commun. 1991, 717–719; c)
J. L. Kisko, T. Hascall, G. Parkin, J. Am. Chem. Soc. 1998, 120, 10561–
10562.
[3] a) A. Kunishita, T. L. Gianetti, J. Arnold, Organometallics 2012, 31, 372–
380; b) N. Shirasawa, M. Akita, S. Hikichi, Y. Moro-oka, Chem. Commun.
1999, 417–418; c) F. A. Jové, C. Pariya, M. Scoblete, G. P. A. Yap, K. H.
Theopold, Chem. Eur. J. 2011, 17, 1310–1318.
[4] a) D. M. Jenkins, T. A. Betley, J. C. Peters, J. Am. Chem. Soc. 2002, 124,
11238–11239; b) R. E. Cowley, R. P. Bontchev, J. Sorrell, O. Sarracino, Y.
Feng, H. Wang, J. M. Smith, J. Am. Chem. Soc. 2007, 129, 2424–2425.
[5] C. E. MacBeth, J. C. Thomas, T. A. Betley, J. C. Peters, Inorg. Chem. 2004,
43, 4645–4662.
[6] S. Hikichi, H. Komatsuzaki, M. Akita, Y. Moro-oka, J. Am. Chem. Soc. 1998,
120, 4699–4710.
[28] a) R. A. F. Tomás, J. C. M. Bordado, J. F. P. Gomes, Chem. Rev. 2013, 113,
7421–7469; b) J. K. Beattie, J. A. Klepetko, A. F. Masters, P. Turner, Polyhe-
dron 2003, 22, 947–965.
[29] S. Hikichi, K. Hanaue, T. Fujimura, H. Okuda, J. Nakazawa, Y. Ohzu, C.
Kobayashi, M. Akita, Dalton Trans. 2013, 42, 3346–3356.
[30] J. H. Espenson, Chemical Kinetics and Reaction Mechanisms, 2nd ed.,
McGraw-Hill, New York, 1995.
[31] J. F. Dunne, J. Su, A. Ellern, A. D. Sadow, Organometallics 2008, 27, 2399–
2401.
[32] V. K. Aggarwal, Z. Gultekin, R. S. Grainger, H. Adams, P. L. Spargo, J. Chem.
Soc. Perkin Trans. 1 1998, 2771–2782.
[7] a) N. Kitajima, W. B. Tolman, Prog. Inorg. Chem. 1995, 43, 419–531; b) G.
Parkin, Chem. Rev. 2004, 104, 699–767.
[8] J. W. Egan, B. S. Haggerty, A. L. Rheingold, S. C. Sendlinger, K. H. Theo-
pold, J. Am. Chem. Soc. 1990, 112, 2445–2446.
[9] a) G. Strukul, Catalytic Oxidations with Hydrogen Peroxide as Oxidant,
Kluwer Academic Publishers, Boston, 1992; b) S. Hikichi, M. Akita, Y.
Moro-oka, Coord. Chem. Rev. 2000, 198, 61–87; c) T. Nagataki, Y. Tachi, S.
Itoh, Chem. Commun. 2006, 4016–4018; d) J. Nakazawa, A. Yata, T. Hori,
T. D. P. Stack, Y. Naruta, S. Hikichi, Chem. Lett. 2013, 42, 1197–1199.
[10] a) J. P. Jesson, S. Trofimenko, D. R. Eaton, J. Am. Chem. Soc. 1967, 89,
3148–3158; b) A. L. Rheingold, L. M. Liable-Sands, S. Trofimenko, Inorg.
Chem. 2001, 40, 6509–6513.
[11] A. L. Rheingold, R. L. Ostrander, B. S. Haggerty, S. Trofimenko, Inorg.
Chem. 1994, 33, 3666–3676.
Received: March 3, 2016
Published Online: April 28, 2016
Eur. J. Inorg. Chem. 2016, 2486–2494
www.eurjic.org
2494
© 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim