Redox-active ligand-mediated Co-Cl bond-forming reactions at reducing square planar cobalt(III) centers

Article abstract of DOI:10.1016/j.poly.2009.06.046

Synthetic routes to new square planar cobalt complexes with redox-active amidophenolate chelates are presented. Contrary to previous reports, steric bulk on the ligands is not a prerequisite to formation of the low-coordinate materials. X-ray crystal structure metrical data of the neutral S = 1/2 complexes are most consistent with cobalt(III) bound to one iminobenzoseminonate(1-) radical and one amidophenolate(2-) ligand. Addition of 1e^- affords reduced congeners that are also square planar cobalt(III) because the redox-active ligand accepts an electron to generate bis(amidophenolate) species. The redox-activity of the ligands facilitates reactions with chlorine electrophiles to generate square pyramidal products containing new Co-Cl bonds. The bond-forming reactions all formally require oxidation of the metal fragment but there is no change in formal cobalt oxidation state. Instead, the reaction proceeds with oxidation of the amidophenolate ligands. Control of ligand oxidation state provides a mechanism for 1e^- versus 2e^- selectivity in the bond-forming redox reactions.

Full text of DOI:10.1016/j.poly.2009.06.046

Polyhedron 29 (2010) 164–169
Contents lists available at ScienceDirect
Polyhedron
journal homepage: www.elsevier.com/locate/poly
Redox-active ligand-mediated Co–Cl bond-forming reactions at reducing square
planar cobalt(III) centers
Aubrey L. Smith a, Laura A. Clapp a, Kenneth I. Hardcastle b, Jake D. Soper a,
*
a School of Chemistry and Biochemistry, Georgia Institute of Technology, Atlanta, GA 30332-0400, United States
b X-ray Crystallography Center, Department of Chemistry, Emory University, 1515 Dickey Drive, Atlanta, GA 30322, United States
a r t i c l e i n f o
a b s t r a c t
Article history:
Available online 26 June 2009
Synthetic routes to new square planar cobalt complexes with redox-active amidophenolate chelates are
presented. Contrary to previous reports, steric bulk on the ligands is not a prerequisite to formation of the
low-coordinate materials. X-ray crystal structure metrical data of the neutral S = ½ complexes are most
consistent with cobalt(III) bound to one iminobenzoseminonate(1–) radical and one amidophenolate(2–)
ligand. Addition of 1eÀ affords reduced congeners that are also square planar cobalt(III) because the
redox-active ligand accepts an electron to generate bis(amidophenolate) species. The redox-activity of
the ligands facilitates reactions with chlorine electrophiles to generate square pyramidal products con-
taining new Co–Cl bonds. The bond-forming reactions all formally require oxidation of the metal frag-
ment but there is no change in formal cobalt oxidation state. Instead, the reaction proceeds with
oxidation of the amidophenolate ligands. Control of ligand oxidation state provides a mechanism for
1eÀ versus 2eÀ selectivity in the bond-forming redox reactions.
Keywords:
Multielectron redox
Redox-active ligands
Square planar cobalt(III)
Ó 2009 Elsevier Ltd. All rights reserved.
1. Introduction
tionalization. The main challenge in translating these organometal-
lic-type transformations to mononuclear 3d metal catalysts is in
The ability to make and break bonds with a high degree of selec-
tivity has applications ranging from small-scale organic synthesis
to energy conversion and the production of clean chemical fuels
[1–5]. Selectivity in many synthetically useful methods for redox
transformations of small-molecule substrates is predicated on
the ability of platinum group inorganic and organometallic cata-
lysts to mediate the transfer of multiple electrons in a single step
[5–11]. The utility of these precious metal catalysts derives, in part,
from their thermodynamic propensity to deliver multiple electron
equivalents while avoiding odd-electron intermediates [12].
Accordingly, modification of supporting ligands is often employed
as a strategy to modulate reactivity by fine tuning steric and elec-
tronic properties, but the ligands are typically redox inert [13].
Although these catalysts are powerful and sophisticated, cost and
toxicity issues inherent to the precious metals can limit their
utility.
developing a multielectron capacity at metals that typically prefer
only single electron redox changes. One successful approach to this
problem is to stabilize metal ions in atypical, high-energy oxida-
tion states, often through novel ligand design strategies [14–21].
An alternative strategy is to utilize cooperative redox changes be-
tween a metal center in common oxidation states and a redox-ac-
tive ligand that can support charge localized structures [22–25].
This approach has been widely employed in reactions at metallo-
porphyrin complexes [26]. Recently, a number of other redox-ac-
tive ligands have been shown to facilitate metal-centered redox
reactions with small-molecules [27–39]. These ligands are more
modular and afford the benefits of a high degree of electronic tun-
ability and flexibility in coordination environment.
Reported herein are our initial efforts to utilize the close match
in frontier orbital energies between redox-active ligands and co-
balt(III) for bond-forming reactions at the metal center. Towards
this goal, we have prepared and characterized new square planar
cobalt complexes with amidophenolate ligands that were previ-
ously reported to be inaccessible [40]. X-ray crystallographic data
is used to rationalize the electronic structures of these species,
prompting a reevaluation of a previously reported assignment for
a related complex [41]. Finally, control of ligand oxidation state
is shown to facilitate redox 1eÀ versus 2eÀ selectivity in Co–Cl
bond-forming reactions, establishing the ability of redox-active li-
gands to engender multielectron reactivity at square planar
cobalt(III).
Catalysts featuring naturally abundant transition metal centers
offer the potential to address these issues. We are therefore pursu-
ing the development of later first-row transition metal complexes
as surrogates for platinum group catalysts, particularly those that
rely on multielectron oxidative addition and reductive elimination
steps in catalytic cycles for small-molecule activation and func-
* Corresponding author. Tel.: +1 404 894 4022; fax: +404 894 5909.
E-mail address: jake.soper@chemistry.gatech.edu (J.D. Soper).
0277-5387/$ - see front matter Ó 2009 Elsevier Ltd. All rights reserved.
doi:10.1016/j.poly.2009.06.046

A.L. Smith et al. / Polyhedron 29 (2010) 164–169
165
line solids suitable for single crystal X-ray diffraction were recov-
ered by slow diffusion of pentane into saturated CH2Cl2 solution
2. Experimental
in the absence of light. UV–vis (THF) nm (e
, MÀ1 cmÀ1): 280
2.1. General considerations
(15 700), 675 (11 500), 900 (16 200). FAB-MS (m/z): 649 [M]+. FTIR
(ATR): 3062 (w), 3041 (w), 3026 (w), 2956 (m), 2900 (m), 2863
(m), 1579 (w), 1536 (w), 1481 (m), 1432 (m), 1356 (m), 1302
(m), 1260 (m), 1215 (w), 1183 (w), 1138 (s), 1104 (s), 1025 (m),
1002 (w), 922 (m), 910 (m), 881(w), 859 (m), 824 (w), 766 (w),
738 (s), 692 (s), 659 (s), 609 (w), 571 (m), 540 (m), 510 (s), 467
(m), 424 (m) cmÀ1. The sample for elemental analysis was washed
in methanol, as described above. The reported analysis is for
CoIII(isqPh)(apPh)Á1.3MeOH, and the presence of the methanol in
the sample was confirmed by 1H NMR spectroscopy. Anal. Calc.
for C41.3H53.9N2CoO3.3: C, 71.74; H, 8.05; N, 4.05. Found C, 71.36;
H, 7.62; N, 4.16%.
Unless otherwise noted, all manipulations were performed un-
der anaerobic conditions using standard vacuum line techniques,
or in an inert atmosphere glove box under purified nitrogen. All
NMR spectra were acquired on a Varian Mercury 300 spectrometer
(300.323 MHz for 1H) at ambient temperature. Chemical shifts are
reported in parts per million (ppm) relative to TMS, with the resid-
ual solvent peak serving as an internal reference. Solution state
magnetic moments were determined by Evans’ NMR method
[42,43]. UV–visible absorption spectra were acquired using a Var-
ian Cary 50 spectrophotometer. Unless otherwise specified, all
electronic absorption spectra were recorded at ambient tempera-
tures in 1 cm quartz cells. IR spectra were obtained using attenu-
ated total reflection (ATR) with a diamond plate on a Thermo
Scientific Nicolet 4700 Fourier-transform infrared spectrophotom-
eter. All mass spectra were recorded in the Georgia Institute of
Technology Bioanalytical Mass Spectrometry Facility. Fast-atom
bombardment mass spectrometry (FAB-MS) was performed using
a VG Instruments 70-SE spectrometer. Cyclic voltammetric mea-
surements were made using a CH Instruments CHI620C potentio-
stat in a three component cell consisting of a platinum disk
working electrode, a platinum wire auxiliary electrode, and a
non-aqueous AgNO3/Ag reference electrode. All electrochemical
experiments were performed in CH3CN with 0.1 M [nBu4N][PF6]
as the supporting electrolyte. Electrochemical data are referenced
and reported to Fc+/Fc as an internal standard. Elemental analyses
were performed by Atlantic Microlab, Inc., Norcross, GA. All analy-
ses were performed in duplicate, and the reported compositions
are the average of the two runs.
2.4. Preparation of CoIII(isqiPr)(apiPr
)
In modification of a literature procedure [41], a five dram scin-
tillation vial was charged with a solution of Co2(CO)8 (0.069 g,
0.203 mmol) in 10 mL toluene and a stir bar. Slow addition of a
solution of ibqiPr (0.308 g, 0.811 mmol) in 5 mL toluene with vigor-
ous stirring immediately afforded a blue solution. The solution was
stirred for 1 h at ambient temperature and the volume was re-
duced to 4 mL to precipitate dark blue solids. The solids were col-
lected by vacuum filtration in air, washed with toluene
(2 Â 2.0 mL, ice cold) and dried in vacuo to give CoIII(isqiPr)(apiPr
)
(0.255 g, 0.312 mmol, 77%). Slow diffusion of toluene into a satu-
rated THF solution afforded single crystals of CoIII(isqiPr)(apiPr) suit-
able for assignment of connectivity by X-ray diffraction. UV–vis
(THF) nm
(e,
MÀ1 cmÀ1): 280 (15 500), 680 (10 400), 905
(15 400). FAB-MS (m/z): 817 [M]+. FTIR (ATR): 3061 (w), 2959
(m), 2925 (m), 2902 (m), 2864 (m), 1535 (w), 1462 (w), 1357
(m), 1327 (w), 1298 (m), 1257 (w), 1230 (w), 1215 (w), 1181
(w), 1133 (s), 1097 (s), 1019 (m), 911 (w), 897 (m), 859 (m), 825
(w), 737 (m), 680 (w), 657 (s), 586 (w), 565 (w), 514 (s), 506 (s),
2.2. Methods and materials
Anhydrous acetonitrile (CH3CN), tetrahydrofuran (THF),
toluene, dichloromethane, and pentane solvents for air- and mois-
ture-sensitive manipulations were purchased from Sigma–Aldrich,
further dried by passage through columns of activated alumina,
degassed by at least three freeze–pump–thaw cycles, and stored
under N2 prior to use. Methanol (anhydrous, 99.0%) was purchased
from Honeywell Burdick & Jackson, and used as received. Deuter-
ated acetonitrile (CD3CN) was purchased from Cambridge Isotope
Laboratories, degassed by three freeze–pump–thaw cycles,
vacuum distilled from CaH2, and stored under a dry N2 atmosphere
prior to use. The ligands 2,4-di-tert-butyl-6-(2,6-diisopropylphe-
nylimino)benzoquinone (ibqiPr) and 2,4-di-tert-butyl-6-(phenyla-
471 (m), 433 (m) cmÀ1
.
2.5. Preparation of Na[CoIII(apPh)2]
A five dram scintillation vial was charged with sodium-mercury
amalgam beads (5% Na, 0.250 g, 0.544 mmol) and CoIII(isqPh)(apPh
)
(0.350 g, 0.539 mmol) and 10 mL CH3CN. The reaction mixture was
fitted with a Teflon-lined cap and stirred for 12 h under N2. During
this time a color change was observed from dark blue to purple.
The solution was collected by vacuum filtration and the solvent
was removed in vacuo to afford Na[CoIII(apPh)2] (0.341 g,
mino)phenol (H2apPh
) were prepared by literature methods
0.504 mmol, 94%) as purple power. UV–vis (CH3CN) nm
M
(e,
À1 cmÀ1): 295 (20 000), 540 (4910), 850 (10 600). 1H NMR
[44,45]. All characterization data matched those referenced.
Co(ClO4)2Á6H2O and Co2(CO)8 were purchased from Strem
Chemical, Inc. All other chemicals were purchased from Sigma–
Aldrich and used as received.
(300 MHz, CD3CN, d): 58.89 (2H), 35.62 (2H), 26.82 (4H), 8.29
(18H), À0.17 (18H), À7.43 (2H), À24.62 (4H) (all br s). The sample
for elemental analysis was collected from a CH3CN solution. The re-
ported analysis is for Na[CoIII(apPh)2]Á2CH3CN, and the presence of
CH3CN in the sample was confirmed by 1H NMR spectroscopy. Anal.
Calc. for C44H56N4CoO2Na: C, 70.01; H, 7.48; N, 7.42. Found: C,
70.31; H, 7.53; N, 7.00%.
2.3. Preparation of CoIII(isqPh)(apPh
)
In modification of a literature procedure [46], Co(ClO4)2Á6H2O
(0.365 g, 1.00 mmol) and H2apPh (0.596 g, 2.00 mmol) were com-
bined in a 50 mL round bottom flask with a stirbar and dissolved
in 20 mL anhydrous methanol. Dropwise addition of Et3N
2.6. Preparation of Na[CoIII(apiPr)2]
(560 lL, 4.02 mmol) with vigorous stirring immediately afforded
This complex was prepared using sodium-mercury amalgam
beads (5% Na, 0.211 g,
0 )
459 mmol) and CoIII(isqiPr)(apiPr
a clear dark blue solution. The flask was fitted to a condenser
and the reaction mixture was heated to reflux in air for 1 h, then
cooled to ambient temperature for 2 h to deposit a dark blue pre-
cipitate. The solids were recovered by vacuum filtration, washed
with cold anhydrous methanol (3 Â 2.0 mL, ice cold), and dried in
vacuo to yield CoIII(isqPh)(apPh) (0.368 g, 0.566 mmol, 57%). Crystal-
(0.368 g, 0.449 mmol) in a procedure directly analogous to that de-
scribed above for Na[CoIII(apPh)2], to give Na[CoIII(apiPr)2] (0.351 g,
0.417 mmol, 93%). UV–vis (CH3CN) nm
(e,
MÀ1 cmÀ1): 290
(15 200), 535 (4430), 835 (9810). 1H NMR (300 MHz, CD3CN, d):
52.45 (br s, 2H), 34.18 (br s 2H), 24.71 (br d, J = 8 Hz, 4H), 10.11

166
A.L. Smith et al. / Polyhedron 29 (2010) 164–169
Table 1
(br t, J = 8 Hz, 2H), 7.94 (br s, 18H), 1.43 (br s, 12H), À0.32 (br s,
18H), À10.42 (br s, 4H), À13.91 (br s, 12H).
Crystallographic data and structure parameters for CoIII(apPh)(isqPh
)
and
(CpÃ2Co)[CoIII(apiPr)2]Á2CH3CN.
2.7. Preparation of (CpÃ2Co)[CoIII(apiPr)2]
Complex
CoIII(apPh)(isqPh
)
(CpÃ2Co)[CoIII(apiPr)2]Á2CH3CN
Empirical formula
Formula weight
T (K)
Crystal system
Space group
C40H50CoN2O2
649.75
173(2)
triclinic
p1
C76H110Co2N4O2
1229.54
173(2)
monoclinic
P21/c
In an adaption of a literature procedure [46], CpÃ2Co (0.040 g,
0.121 mmol) and CoIII(isqiPr)(apiPr) (0.098 g, 0.119 mmol) were com-
bined in a five dram scintillation vial and dissolved in 10 mL CH3CN
with vigorous stirring. The reaction mixture was fitted with a Teflon-
lined cap and stirred for 3 h under N2 to generate a dark-purple solu-
tion. The volume of the solution was reduced to 3 mL in vacuo and
stored at À20 °C for 48 h to yield (CpÃ2Co)[CoIII(apiPr)2] (0.111 g,
0.0965 mmol, 80%) as violet microcrystals. Purity was determined
by comparison of the UV–vis spectrum of the isolated product to that
reported above for the sodium salt. Slow cooling of a saturated
CH3CN solution afforded single crystals of (CpÃ2Co)[CoIII(apiPr)2] suit-
able for study by X-ray diffraction.
Unit cell dimensions
a (Å)
b (Å)
c (Å)
5.8590(2)
11.7285(4)
13.6999(6)
110.421(2)
93.205(2)
93.739(2)
877.26(6)
1
9.3811(1)
17.0552(4)
21.9136(5)
90
92.785(2)
90
3501.96(14)
2
1.166
a
(°)
b (°)
c
(°)
V (Å3)
Z
Dcalc (g cmÀ1
)
1.230
4.102
Absorption coefficient (mmÀ1
Crystal size (mm)
)
0.520
0.37 Â 0.12 Â 0.10 0.36 Â 0.11 Â 0.10
2.8. X-ray crystallography
h range for data collection (°)
Index ranges
4.29–66.89
À6 6 h 6 6
À13 6 k 6 13
À15 6 l 6 12
5817/2624
1.056
1.51–27.15
À10 6 h 6 12
À18 6 k 6 21
À27 6 l 6 28
23 193/7686
1.015
Single crystals of CoIII(apPh)(isqPh
)
and (CpÃ2Co)[CoIII(a-
p
iPr)2]Á2CH3CN suitable for X-ray diffraction analysis were coated
Reflections collected/unique
with Paratone N, suspended in a small fiber loop and placed in a
cooled nitrogen gas stream at 173 K on a Bruker D8 APEX II CCD
Goodness of fit on F2
R [I > 2
r(I)]
0.0338
0.0873
0.0542
0.1579
sealed tube diffractometer. Diffraction data for CoIII(apPh)(isqPh
)
wR2 (all data)
was collected using graphite monochromated Cu Ka (k = 1.54178 Å)
radiation. Data for (Cp2ÃCo)[CoIII(apiPr)2]Á2CH3CN was obtained with
graphite monochromated Mo Ka (k = 0.71073 Å) radiation. All other
data collection procedures, data processing and programs were the
same for both samples. Data were measured using a series of combi-
nations of phi and omega scans with 10 s frame exposures and 0.5°
frame widths. Data collection, indexing and initial cell refinements
were all carried out using APEX II software [47].Frame integration
and final cell refinements were done using SAINT software [48]. The
final cell parameters were determined from least-squares refine-
ment on 2789 reflections for CoIII(apPh)(isqPh) and 2968 reflections
for (Cp2ÃCo)[CoIII(apiPr)2]Á2CH3CN. The structures were solved using
direct methods and difference Fourier techniques using the SHELXTL
program package [49]. Hydrogen atoms were placed in their
expected chemical positions using the HFIX command and were
included in the final cycles of least-squares with isotropic Uij’s
related to the atoms ridden upon. All non-hydrogen atoms were
refined anisotropically. Details of data collection and structure
refinement are provided in Table 1.
Scheme 1.
the bond distances are crystallographically indistinguishable from
the previously reported structures of the Pr analog [41], and a
i
congener containing one ortho-triflouromethyl substituent on the
N-aryl ring [46]. The solution magnetic moment of 2.18 lB for
CoIII(apPh)(isqPh) at 25 °C in THF is in agreement with a previously
reported value for the trifluoromethyl analog [46]. This was pro-
posed to result from intramolecular antiferromagnetic coupling
between the intermediate spin S = 1 cobalt(III) ion and one imino-
benzoseminonate(1–) radical ligand to yield an S = ½ ground state.
3. Results
3.2. Preparation and characterization of [CoIII(apAr)2]À complexes
3.1. Preparation and characterization of CoIII(apAr)(isqAr) complexes
Addition of 1 equiv CpÃ2Co or Na (as 5% Na–Hg amalgam) to the
blue CoIII(apAr)(isqAr) complexes in CH3CN under N2 affords dark-
purple products over hours at ambient temperature. Recrystalliza-
Charge neutral, four-coordinate cobalt complexes were ob-
tained by modifications of literature procedures: via stoichiometric
reaction of Co2(CO)8 with ibqiPr in toluene [41], or by addition of
Co(ClO4)2Á6H2O to H2apPh in basic MeOH (ibqiPr = 2,4-di-tert-bu-
i
tion of the Pr derivative from CH3CN yielded single crystals of the
cobaltocenium salt suitable for analysis by X-ray diffraction. As
shown in Fig. 2a, the cobalt anion retains its square planar geometry.
Two CH3CN solvent molecules per anion are located in the crystal
structure but are not bound to cobalt. As in the neutral species,
the ligands in the anion are crystallographically indistinguishable,
and the N-aryl rings are nearly orthogonal to the plane defined by
the square planar cobalt center, with Co–N–C–C torsion angles of
À91.0(3)° and 82.8(3)°. However, the aromatic C–C bond distances
tyl-6-(2,6-diisopropylphenylimino)benzoquinone;
H2apPh = 2,4-
di-tert-butyl-6-(phenylamino)phenol; Scheme 1) [46]. A single
crystal X-ray structure of the Ph derivative is provided in Fig. 1a.
It contains cobalt bound to two coplanar ligands. The N-phenyl
substituents are rotated out of the plane defined by the square pla-
nar cobalt core, with Co–N–C–C torsion angles of À107.63(19)° and
68.5(2)°. The cobalt lies on an inversion center making the two
aminophenol-derived ligands crystallographically identical. The
C–O, C–N and C–C bond distances shown in Fig. 1b exhibit a small
but significant quinoid distortion that is in good agreement with
(Fig. 2b) are identical within 3
r (1.40 0.01 Å), and the ligand C–O
and C–N bond distances are elongated as compared to CoIII(apiPr)(is-
q
iPr). The reduced complexes are therefore best formulated as
the mean of those typical for [apPh 2À
gesting that the complex is best formulated as CoIII(apPh)(isqPh
(isqPh = 2,4-di-tert-butyl-6-(phenylimino)semiquinonate). All of
]
and [isqPh ꢀÀ
]
[25,46], sug-
[CoIII(apAr)2]À [25,41].
)
The [CoIII(apAr)2]À complexes exhibit diagnostic broad, para-
magnetically shifted 1H NMR spectra (d +59 to À24 ppm) at ambi-

A.L. Smith et al. / Polyhedron 29 (2010) 164–169
167
Fig. 3. Cyclic voltammograms of 5 mM Na[CoIII(apPh)2] (red) and Na[CoIII(apiPr)2]
(black) in CH3CN containing 0.1 M [nBu4N][PF6] at a 10 mm Pt electrode. Scan rate:
0.5 V sÀ1. Temperature: 25 °C.
blue-green in seconds at ambient temperature. Identical UV–vis
spectra are obtained by exposure of CH2Cl2 solutions of the CoIII(a-
p
Fig. 1. (a) Solid-state structure of CoIII(apPh)(isqPh) drawn with 50% probability
ellipsoids. Hydrogen atoms omitted for clarity. Selected bond angles (deg): O1–
Co1–N1 85.58(6), O1–Co1–N1A 94.42(6). (b) Schematic of selected bond lengths (Å)
in CoIII(apPh)(isqPh).
Ar)(isqAr) materials to ambient fluorescent light. The products of
all the reactions are the previously reported square pyramidal
CoIIICl(isqAr
)
2
species containing two iminobenzoseminonate(1–)
radical ligands [40,41,51].
As illustrated in Scheme 2, quantitative conversion of
CoIII(apAr)(isqAr) to CoIIICl(isqAr
) products occurs by net addition
2
of a chlorine radical. Because the Co–Cl bond-forming reactions
are formally a reduction of Clꢀ to ClÀ, they occur with 1eÀ oxidation
of the CoIII(ap)(isq) fragment, which is ligand centered. Addition of
1 equiv of the radical inhibitor 2,6-di-tert-butyl-4-methylphenol
(BHT) to THF solutions containing CoIII(apAr)(isqAr) results in a sig-
nificant decrease in its rate of reactions with CCl4: whereas addi-
tion of 5 equiv CCl4 to CoIII(apPh)(isqPh
) affords quantitative
conversion to CoIIICl(isqPh
)
in seconds, with added BHT the same
2
reaction takes hours to reach completion. The origin of this inhibi-
tion is apparently a reversible reaction of CoIII(apPh)(isqPh) with
BHT. Accordingly, addition of 1 equiv BHT to CoIII(apPh)(isqPh) in
THF or CH3CN under N2 affords an immediate color change from
blue to olive green. The dark green color is slowly discharged on
addition of CCl4 to afford CoIIICl(isqPh)2. The nature of the green
species is still under investigation.
Reaction of CH3CN solutions of the [CoIII(apAr)2]À complexes
with
1
equiv 2,3,4,5,6,6-hexachloro-2,4-cyclohexadien-1-one,
which serves as a source of electrophilic Cl+ [52], immediately
gives the same CoIIICl(isqAr)2 products. The reaction is quantitative
and proceeds without the accumulation of any observable inter-
mediates at 298 K, as evidenced by UV–vis spectroscopy. The rate
Fig. 2. (a) Solid-state structure of the anion in (Cp*2Co)[CoIII(apiPr)2]Á2CH3CN shown
with 50% probability ellipsoids. Hydrogen atoms, CH3CN solvate molecules and
countercation omitted for clarity. Selected bond angles (deg): O1–Co1–N1 85.75(8),
O1–Co1–N1A 94.25(8). (b) Schematic of selected bond lengths (Å) in [CoIII(apiPr)2]À.
of CoIIICl(isqPh
) formation is unaffected by addition of 1 equiv
2
BHT to the reaction mixture. Control reactions confirm that
[CoIII(apPh)2]À does not react with added BHT. The bond-forming
reaction with net Cl+ is a 2eÀ redox process, because the electro-
philic addition to [CoIII(apAr)2]À formally reduces Cl+ to
ClÀ (Scheme 2). However, the 2eÀ oxidation of the square planar
cobalt(III) fragment is not metal-centered because the redox-active
ligands each supply one electron for the reaction. Addition of
1 equiv CCl4 to CH3CN solutions of [CoIII(apPh)2]À gives no reaction
over days at ambient temperature, as evidenced by UV–vis
spectroscopy.
ent temperatures. The solution magnetic moment of 2.77 lB for
[CoIII(apPh)2]– at 25 °C in CH3CN is consistent with two unpaired
electrons as intermediate spin S = 1 cobalt(III) in a square planar li-
gand field. Cyclic voltammograms of both [CoIII(apAr)2]À complexes
display two quasi-reversible 1eÀ redox couples, centered at À0.350
and À0.770 V versus Fc+/Fc for the Ph derivative and À0.339 and
À0.809 V versus Fc+/Fc for the iPr congener (Fig. 3). Because the
CoIII(apAr)(isqAr) complexes are not reduced by CpÃ2Fe (with a reduc-
tion potential of À0.48 V versus Fc+/Fc in CH3CN) [50], the redox
event at more negative potential is assigned to interconversion of
CoIII(apAr)(isqAr) and [CoIII(apAr)2]À.
4. Discussion
The structural and electronic properties of the square planar co-
balt electron-transfer pairs reported here are notable in several re-
spects. First, our successful isolation of the Ph complexes
demonstrates that, contrary to a previous report [40], steric bulk
at the ortho-positions of the ligand N-aryl group is not a prerequi-
site for stabilization of a four-coordinate complex. This ability to
3.3. Redox-active ligand-mediated Co–Cl bond-forming reactions at
cobalt(III)
Addition of 1 equiv CCl4 or N-chlorosuccinimide (NCS) to the
blue CoIII(apAr)(isqAr) complexes in THF gives a color change to

168
A.L. Smith et al. / Polyhedron 29 (2010) 164–169
Scheme 2.
forego bulky ligands is essential for the development of reaction
chemistry at these complexes because it affords substrates better
access to the cobalt center.
Second, the X-ray crystallographic data for both the neutral
S = ½ and anionic S = 1 complexes suggest that they are all best for-
mulated as cobalt(III), so addition of an electron to CoIII(apAr)(isqAr
)
complexes occurs at a redox-active ligand. This proposed charge
distribution agrees with a recent assignment of a closely related
electron-transfer pair containing an ortho-trifluoromethyl substi-
tuent on the ligand N-aryl group [46]. It should be noted that an
alternative electronic structure assignment has been proposed for
the charge neutral S = ½ complexes. Poddel’sky and coworkers ar-
Scheme 3.
gue that these materials may be better formulated as CoII(isqAr
)
2
containing a low spin d7 cobalt(II) ion ligated by two iminobenz-
oseminonate(1–) radicals that couple antiferromagnetically
[25,41]. The basis for this claim are previously reported EPR spectra
for two CoIII(apAr)(isqAr) complexes that exhibit features expected
for an electron localized on the cobalt ion [41,46]. However, the co-
balt(III) assignment proposed by Wieghardt and coworkers (de-
scribed in Section 3.1 above), in fact invokes a metal-centered
radical that is consistent with the observed EPR data [46]. More-
over, Poddel’sky and coworkers suggest that the unpaired electrons
in an S = 1 square planar cobalt(III) center would occupy MOs that
without a formal change in oxidation state at the metal center
[27,30,31,37].
Two possible mechanistic pathways were considered for the
electrophilic addition reactions at [CoIII(apAr)2]À: (a) direct nucleo-
philic attack of the cobalt(III) center on the Cl+ electrophile,
resulting in a 2eÀ oxidation of the metal fragment; (b) initial
outer-sphere 1eÀ transfer (ET) to generate the radical pair,
CoIII(apAr)(isqAr) and Clꢀ, followed by coupling with bond formation
(Scheme 3). The facile reactions of CoIII(apAr)(isqAr) with sources of
derive from the metal dx2
and dz2 orbitals [25], but this is incor-
Ày2
rect. Calculations on a very closely related system clearly show that
the unpaired electrons are in orbitals of b2g and b3g symmetry that
are primarily of metal dxz and dyz parentage [46]. In this scheme,
chlorine radical Clꢀ to generate CoIIICl(isqAr
)2 demonstrate that the
stepwise mechanism b is viable, but experiments performed in the
presence of the radical inhibitor support mechanism a. Namely, the
reactions of CoIII(apPh)(isqPh) with CCl4 is retarded by addition of
BHT, but the reaction of [CoIII(apPh)2]À with 2,3,4,5,6,6-hexachlo-
ro-2,4-cyclohexadien-1-one is unaffected with added BHT.
Although the mechanism of this radical inhibition is ill defined,
control reactions demonstrate that CoIII(apPh)(isqPh) reacts rapidly
with BHT under the reaction conditions. Accordingly, mechanisms
that invoke formation of a long-lived CoIII(apPh)(isqPh) intermediate
in the reaction of [CoIII(apPh)2]À with net Cl+ (including mechanism
b) can likely be ruled out.
In sum, the reaction chemistry presented herein suggest that
the S = ½ CoIII(apAr)(isqAr) complexes behave as efficient radical
scavengers but the [CoIII(apAr)2]À complexes react as 2eÀ nucleo-
philes and do not react efficiently with sources of net Clꢀ. This is
somewhat surprising given the S = 1 diradical ground state of the
anionic complexes. The reactivity patterns of the [CoIII(apAr)2]À
species are therefore more reminiscent of the well-known cob-
aloxime(I) complexes that are closed-shell d8 nucleophiles [21].
However, the [CoIII(apAr)2]À complexes reported here are not
strong outer-sphere reductants, being oxidized at potentials
200–400 mV less negative than most cobalt(I) materials [21,53–55].
Control of 1eÀ versus 2eÀ reactivity derives from management of
ligand electron inventory. This ability to modulate the redox selec-
tivity of the metal center to achieve 2eÀ reactions under gentle
conditions is predicated entirely on the redox-activity of the li-
gands, and is a key design feature for the development of naturally
abundant transition metals as catalysts for selective bond-making
and bond-breaking reactions.
overlap with the aminophenolate
p-donor MOs raises the energy
of the metal b2g and b3g orbitals and provides a pathway for facile
intramolecular ligand-to-metal charge transfer. Because the ligand
metrical parameters for CoIII(apPh)(isqPh) are indistinguishable
from those of the previously reported iPr congener [41], we suggest
that all of these complexes are better described as CoIII(apAr)(isqAr
in the ground state.
)
Finally, the ability of the redox-active ligands to accept and
store electrons facilitates the preparation of coordinatively-unsat-
urated cobalt(III) complexes that can deliver multiple electrons at
modest reduction potentials. This forms a basis for oxidative
bond-forming reactions with chlorine electrophiles at cobalt(III)
centers. Identical CoIIICl(isqAr
) products are obtained from reac-
2
tions of CoIII(apAr)(isqAr) with sources of a net Clꢀ(via a 1eÀ reac-
tion) or by addition of electrophilic Cl+ to [CoIII(apAr)2]À (via a
2eÀ process). Both the radical addition and electrophilic addition
reactions are formally oxidations of the metal fragments, but nei-
ther oxidizes the cobalt(III) center because the redox-active ligand
manifold acts as a reservoir for electrons. As described above, facile
transfer of electrons from the ligand to the metal derives from a
high degree of covalency in the metal–ligand bonding that is a re-
sult of the close match of the frontier orbital energies of cobalt(III)
and the aminophenol-derived ligands [46]. In this way, the ligand-
mediated Co–Cl bond-forming reactions reported herein are very
similar to recent reports of pseudo-oxidative addition reactions
of X2 to redox-active ligand complexes that rely on ligand centered
changes in oxidation state to accommodate the transformation

A.L. Smith et al. / Polyhedron 29 (2010) 164–169
169
[14] M.P. Hendrich, W. Gunderson, R.K. Behan, M.T. Green, M.P. Mehn, T.A. Betley,
C.C. Lu, J.C. Peters, Proc. Natl. Acad. Sci. USA 103 (2006) 17107.
[15] M.J. Ingleson, M. Pink, H. Fan, K.G. Caulton, J. Am. Chem. Soc. 130 (2008) 4262.
[16] C. Vogel, F.W. Heinemann, J. Sutter, C. Anthon, K. Meyer, Angew. Chem., Int. Ed.
47 (2008) 2681.
[17] J.M. Smith, Comments Inorg. Chem. 29 (2008) 189.
[18] E. Kogut, H.L. Wiencko, L. Zhang, D.E. Cordeau, T.H. Warren, J. Am. Chem. Soc.
127 (2005) 11248.
[19] P.L. Holland, Acc. Chem. Res. 41 (2008) 905.
[20] K.D. Kitiachvili, D.J. Mindiola, G.L. Hillhouse, J. Am. Chem. Soc. 126 (2004)
10554.
[21] G.N. Schrauzer, E. Deutsch, J. Am. Chem. Soc. 91 (1969) 3341.
[22] C.G. Pierpont, C.W. Lange, Prog. Inorg. Chem. 41 (1994) 331.
[23] C.G. Pierpont, Coord. Chem. Rev. 216–217 (2001) 99.
[24] P. Chaudhuri, K. Wieghardt, Prog. Inorg. Chem. 50 (2001) 151.
[25] A.I. Poddel’sky, V.K. Cherkasov, G.A. Abakumov, Coord. Chem. Rev. 253 (2009)
291.
5. Conclusions
The redox-activity of amidophenolate ligands facilitates selec-
tive multielectron reactions at square planar cobalt(III) centers.
The cobalt(III) complexes are not strong outer-sphere reductants.
Instead bond-forming reactions with substrates occur at the metal
center through on-demand, intramolecular delivery of redox
equivalents from ligand-based electron reservoirs. Modulation of
the ligand oxidation state affords a unique mechanism to select
for 2eÀ redox reactivity over potentially competing 1eÀ pathways.
Future studies in this area are focused on exploiting this ligand-
mediated 2eÀ redox capacity for the development of new multi-
electron catalysis cycles for selective bond-making and bond-
breaking transformations of small-molecule organic substrates.
[26] K.M. Kadish, K.M. Smith, R. Guilard (Eds.), The Porphyrin Handbook, Academic
Press, 2000.
[27] K.J. Blackmore, J.W. Ziller, A.F. Heyduk, Inorg. Chem. 44 (2005) 5559.
[28] M.R. Haneline, A.F. Heyduk, J. Am. Chem. Soc. 128 (2006) 8410.
[29] R.A. Zarkesh, J.W. Ziller, A.F. Heyduk, Angew. Chem., Int. Ed. 47 (2008)
4715.
[30] N.A. Ketterer, H. Fan, K.J. Blackmore, X. Yang, J.W. Ziller, M.-H. Baik, A.F.
Heyduk, J. Am. Chem. Soc. 130 (2008) 4364.
[31] K.J. Blackmore, M.B. Sly, M.R. Haneline, J.W. Ziller, A.F. Heyduk, Inorg. Chem. 47
(2008) 10522.
[32] A.I. Nguyen, K.J. Blackmore, S.M. Carter, R.A. Zarkesh, A.F. Heyduk, J. Am. Chem.
Soc. 131 (2009) 3307.
Supplementary data
CCDC 730032 and 730033 contain the supplementary crystallo-
*
graphic data for CoIII(apPh)(isqPh
)
and (Cp 2Co)[CoIII(apiPr)2]Á
2CH3CN. These data can be obtained free of charge via http://
www.ccdc.cam.ac.uk/conts/retrieving.html, or from the Cambridge
Crystallographic Data Centre, 12 Union Road, Cambridge CB2 1EZ,
UK; fax: (+44) 1223-336-033; or e-mail: deposit@ccdc.cam.ac.uk.
[33] S.C. Bart, E. Lobkovsky, E. Bill, P.J. Chirik, J. Am. Chem. Soc. 128 (2006) 5302.
[34] M.W. Bouwkamp, A.C. Bowman, E. Lobkovsky, P.J. Chirik, J. Am. Chem. Soc. 128
(2006) 13340.
Acknowledgements
[35] S.C. Bart, A.C. Bowman, E. Lobkovsky, P.J. Chirik, J. Am. Chem. Soc. 129 (2007)
7212.
[36] C. Stanciu, M.E. Jones, P.E. Fanwick, M.M. Abu-Omar, J. Am. Chem. Soc. 129
(2007) 12400.
[37] C. Mukherjee, T. Weyhermueller, E. Bothe, P. Chaudhuri, Inorg. Chem. 47
(2008) 2740.
[38] C.J. Rolle, K.I. Hardcastle, J.D. Soper, Inorg. Chem. 47 (2008) 1892.
[39] C.C. Lu, T. Weyhermueller, E. Bill, K. Wieghardt, Inorg. Chem. 48 (2009) 6055.
[40] G.A. Abakumov, V.K. Cherkasov, M.P. Bubnov, L.G. Abakumova, V.N. Ikorskii,
G.V. Romanenko, A.I. Poddel’sky, Russ. Chem. Bull. 55 (2006) 44.
[41] A.I. Poddel’sky, V.K. Cherkasov, G.K. Fukin, M.P. Bubnov, L.G. Abakumova, G.A.
Abakumov, Inorg. Chim. Acta 357 (2004) 3632.
We gratefully acknowledge financial support from the ACS
Petroleum Research Fund (45130-G3) and the Georgia Institute
of Technology. We thank Dr. Les Gelbaum for assistance with
NMR spectroscopy, and David Bostwick for assistance with mass
spectrometry.
References
[42] D.F. Evans, J. Chem. Soc. (1959) 2003.
[43] D.H. Live, S.I. Chan, Anal. Chem. 42 (1970) 791.
[44] P. Chaudhuri, C.N. Verani, E. Bill, E. Bothe, T. Weyhermueller, K. Wieghardt, J.
Am. Chem. Soc. 123 (2001) 2213.
[45] G.A. Abakumov, N.O. Druzhkov, Y.A. Kurskii, A.S. Shavyrin, Russ. Chem. Bull. 52
(2003) 712.
[46] E. Bill, E. Bothe, P. Chaudhuri, K. Chlopek, D. Herebian, S. Kokatam, K. Ray, T.
Weyhermueller, F. Neese, K. Wieghardt, Chem. Eur. J. 11 (2005) 204.
[47] APEX II, Analytical X-Ray Systems, Brukers AXS, Inc., Madison, WI, 2005.
[48] SAINT Version 6.45A, Analytical X-ray Systems, Brukers AXS, Inc., Madison, WI,
2003.
[49] SHELXTL Version 6.12, Analytical X-ray Systems, Bruker AXS, Inc., Madison, WI,
2002.
[50] N.G. Connelly, W.E. Geiger, Chem. Rev. 96 (1996) 877.
[51] D. Herebian, P. Ghosh, H. Chun, E. Bothe, T. Weyhermuller, K. Wieghardt, Eur. J.
Inorg. Chem. (2002) 1957.
[1] N.S. Lewis, D.G. Nocera, Proc. Natl. Acad. Sci. USA 103 (2006) 15729.
[2] R. Eisenberg, D.G. Nocera, Inorg. Chem. 44 (2005) 6799.
[3] U.S. Department of Energy, Report of the Basic Energy Sciences Workshop on
Solar Energy Utilization, Washington, D.C., 2005.
[4] U.S. Department of Energy, Report from the Basic Energy Science Workshop
Washington, D.C., 2007.
[5] B.M. Trost, I. Fleming (Eds.), Comprehensive Organic Synthesis: Selectivity,
Strategy and Efficiency in Modern Organic Chemistry, vol. 9 Set, Pergamon
press, 1992.
[6] A. de Meijere, F. Diederich (Eds.), Metal-Catalyzed Cross-Coupling Reactions,
Second Completely Revised and Enlarged Edition, vol. 1, Wiley-VCH, 2004.
[7] J.F. Hartwig, in: Activation and Functionalization of C–H Bonds, vol. 885, 2004,
Oxford University Press, pp. 136.
[8] S.S. Stahl, Angew. Chem., Int. Ed. 43 (2004) 3400.
[9] S.S. Stahl, J.A. Labinger, J.E. Bercaw, Angew. Chem., Int. Ed. 37 (1998) 2181.
[10] K.I. Goldberg, A.S. Goldman (Eds.), Activation and Functionalization of C–H
Bonds, Oxford University Press, 2004.
[52] S. France, H. Wack, A.E. Taggi, A.M. Hafez, T.R. Wagerle, M.H. Shah, C.L. Dusich,
T. Lectka, J. Am. Chem. Soc. 126 (2004) 4245.
[53] D.G. Brown, Prog. Inorg. Chem. 18 (1973) 177.
[54] Y. Hisaeda, T. Nishioka, Y. Inoue, K. Asada, T. Hayashi, Coord. Chem. Rev. 198
(2000) 21.
[11] J.A. Labinger, J.E. Bercaw, Nature 417 (2002) 507.
[12] N.N. Greenwood, A. Earnshaw, Chemistry of the Elements, second ed.,
Butterworth Heinemann 1997.
[55] X.L. Hu, B.S. Brunschwig, J.C. Peters, J. Am. Chem. Soc. 129 (2007) 8988.
[13] A.B.P. Lever, Comprehensive Coordination Chemistry II, Elsevier Science 2004.