Hangman corroles: Efficient synthesis and oxygen reaction chemistry

Article abstract of DOI:10.1021/ja108904s

The construction of a new class of compounds - the hangman corroles - is provided efficiently by the modification of macrocyclic forming reactions from bilanes. Hangman cobalt corroles are furnished in good yields from a one-pot condensation of dipyrromethane with the aldehyde of a xanthene spacer followed by metal insertion using microwave irradiation. In high oxidation states, X-band EPR spectra and DFT calculations of cobalt corrole axially ligated by chloride are consistent with the description of a Co(III) center residing in the one-electron oxidized corrole macrocycle. These high oxidation states are likely accessed in the activation of O-O bonds. Along these lines, we show that the proton-donating group of the hangman platform works in concert with the redox properties of the corrole to enhance the catalytic activity of O-O bond activation. The hangman corroles show enhanced activity for the selective reduction of oxygen to water as compared to their unmodified counterparts. The oxygen adduct, prior to oxygen reduction, is characterized by EPR and absorption spectroscopy.

Full text of DOI:10.1021/ja108904s

Published on Web 12/10/2010
Hangman Corroles: Efficient Synthesis and Oxygen Reaction
Chemistry
Dilek K. Dogutan,† Sebastian A. Stoian,‡ Robert McGuire, Jr.,†
Matthias Schwalbe,§,† Thomas S. Teets,† and Daniel G. Nocera*,†
Department of Chemistry, 6-335, Massachusetts Institute of Technology 77 Massachusetts
AVenue, Cambridge, Massachusetts 02139-4307, United States, and Department of Chemistry,
Carnegie Mellon UniVersity Mellon Institute Pittsburgh, PennsylVania 15213, United States
Received October 2, 2010; E-mail: nocera@mit.edu
Abstract: The construction of a new class of compoundssthe hangman corrolessis provided efficiently
by the modification of macrocyclic forming reactions from bilanes. Hangman cobalt corroles are furnished
in good yields from a one-pot condensation of dipyrromethane with the aldehyde of a xanthene spacer
followed by metal insertion using microwave irradiation. In high oxidation states, X-band EPR spectra and
DFT calculations of cobalt corrole axially ligated by chloride are consistent with the description of a Co(III)
center residing in the one-electron oxidized corrole macrocycle. These high oxidation states are likely
accessed in the activation of O-O bonds. Along these lines, we show that the proton-donating group of
the hangman platform works in concert with the redox properties of the corrole to enhance the catalytic
activity of O-O bond activation. The hangman corroles show enhanced activity for the selective reduction
of oxygen to water as compared to their unmodified counterparts. The oxygen adduct, prior to oxygen
reduction, is characterized by EPR and absorption spectroscopy.
Introduction
by now reporting the synthesis of a new class of compoundss
hangman corroles.
The hangman construct allows for precise control of the
placement of a proton-donating or -accepting moiety over
the face of a macrocyclic redox site.1,2 Bond-making and
bond-breaking catalysis of small-molecule substrates residing
in the hangman cleft may be activated by proton-coupled
electron transfer (PCET)1-6 where proton transfer is delivered
from the hanging group and the redox equivalents and
electron transfer is to the macrocycle.7 The hangman motif
heretofore has emphasized porphyrin8-11 and salen12,13
macrocycles. We expand the scope of the hangman construct
Metallocorrole complexes are relatives of the corrin macro-
cycle found in coenzyme B12 and they are active catalysts for
a variety of transformations.14,15 Compared to porphyrins,
corroles lack one meso bridge carbon atom and thus have a
smaller macrocyclic cavity. Moreover, their four nitrogens bear
three protons, as compared with two protons in porphyrins. The
higher charge of the macrocyclic core, together with electronic
considerations, permit corroles to stabilize metals in higher
oxidation states,14,16 and thus open new avenues for oxidative
catalysis, especially involving substrates of energy consequence
such as carbon dioxide,17,18 oxygen19,20 and water.21 Many of
these transformations involve the coupling of the proton to the
electron and hence this class of catalysts can benefit from the
hangman motif.
† Massachusetts Institute of Technology.
‡ Carnegie Mellon University Mellon Institute Pittsburgh.
§ Current address: Department of Chemistry, Humboldt Universita¨t zu
Berlin, Brook-Taylor-Strasse 2, 12489, Berlin.
(1) Rosenthal, J.; Nocera, D. G. Acc. Chem. Res. 2007, 40, 543–553.
(2) Rosenthal, J.; Nocera, D. G. Prog. Inorg. Chem. 2007, 55, 483–544.
(3) Chang, C. J.; Chang, M. C. Y.; Damrauer, N. H.; Nocera, D. G.
Biochim. Biophys. Acta 2004, 1655, 13–28.
The synthesis of variously substituted corroles may proceed
along many routes.22 Recently, corroles have been introduced
into Pacman motifs, in which a difunctionalized polycyclic
(4) Nocera, D. G. Inorg. Chem. 2009, 48, 10001–10017.
(5) Dempsey, J. L.; Esswein, A. J.; Manke, D. R.; Rosenthal, J.; Soper,
J. D.; Nocera, D. G. Inorg. Chem. 2005, 44, 6879–6892.
(6) Reece, S. Y.; Hodgkiss, J. M.; Stubbe, J.; Nocera, D. G. Phil. Trans.
R. Soc. B 2006, 361, 1351–1364.
(14) Aviv-Harel, I.; Gross, Z. Chem.sEur. J. 2009, 15, 8382–8394.
(15) Aviv, I.; Gross, Z. Chem. Commun. 2007, 1987–1999.
(16) Gross, Z.; Gray, H. B. Comm. Inorg. Chem. 2006, 27, 61–72.
(17) Grodkowski, J.; Neta, P.; Fujita, E.; Mahammed, A.; Simkhovich, L.;
Gross, Z. J. Phys. Chem. A 2002, 106, 4772–4778.
(7) Chang, C. J.; Chng, L. L.; Nocera, D. G. J. Am. Chem. Soc. 2003,
125, 1866–1876.
(8) Chng, L. L.; Chang, C. J.; Nocera, D. G. Org. Lett. 2003, 5, 2421–
2424.
(18) Okamoto, A.; Nakamura, R.; Osawa, H.; Hashimoto, K. J. Phys. Chem.
C 2008, 112, 19777–19783.
(9) McGuire, R., Jr.; Dogutan, D. K.; Teets, T. S.; Suntivich, J.; Shao-
Horn, Y.; Nocera, D. G. Chem. Sci. 2010, 1, 411–414.
(10) Chang, C. J.; Yeh, C.-Y.; Nocera, D. G. J. Org. Chem. 2002, 67, 1403–
1406.
(19) Kadish, K. M.; Shen, J.; Fremond, L.; Chen, P.; Ojaimi, M. El.;
Chkounda, M.; Gros, C. P.; Barbe, J.-M.; Ohkubo, K.; Fukuzumi, S.;
Guilard, R. Inorg. Chem. 2008, 47, 6726–6737.
(11) Yeh, C.-Y.; Chang, C. J.; Nocera, D. G. J. Am. Chem. Soc. 2001,
123, 1513–1514.
(20) Kadish, K. M.; Fremond, L.; Shen, J.; Chen, P.; Ohkubo, K.; Fukuzumi,
S.; El Ojaimi, M.; Gros, C. P.; Barbe, J. M.; Gullard, R. Inorg. Chem.
2009, 48, 2571–2582.
(12) Liu, S.-Y.; Nocera, D. G. J. Am. Chem. Soc. 2005, 127, 5278–5279.
(13) Yang, J. Y.; Bachmann, J.; Nocera, D. G. J. Org. Chem. 2006, 71,
8706–8714.
(21) Gao, Y.; Akermark, T.; Liu, J. H.; Sun, L. C.; Akermark, B. J. Am.
Chem. Soc. 2009, 131, 8726–8727.
9
10.1021/ja108904s  2011 American Chemical Society
J. AM. CHEM. SOC. 2011, 133, 131–140 131

A R T I C L E S
Dogutan et al.
scaffold bear two corroles or a corrole and a porphyrin.23,24
Adapting Lindsey porphyrin25 and Gryko corrole forming
conditions,26 we now report synthetic methods for: (i) a one-
pot route to deliver 5,10,15-tris(pentafluorophenyl) corrole in
good yields and a new class of compounds, the (ii) hangman
corroles (HCX), which have been structurally characterized. We
go on to show by spectroscopy, augmented by density functional
theory (DFT), that the cobalt hangman corroles are capable of
accessing high oxidation states in which the corrole is a
noninnocent redox ligand. The PCET reactivity of the cobalt
hangman corroles is demonstrated for the four-electron, four-
proton reduction of oxygen, where a key intermediate for this
reaction, a cobalt corrole superoxide complex, has been observed
spectroscopically.
microwave irradiation at 75 °C. The protocol was as follows: (1)
heat the reaction vessel from room temperature to 75 °C, (2) hold
at 75 °C and irradiate for 10 min (temperature overshoots of 80-85
°C were permitted; temperature was re-established at 75 °C by using
open flow valve option), (3) allow the reaction mixture to cool to
room temperature, (4) check the reaction mixture by silica TLC
analysis, and (5) repeat steps 1 - 4 until all of the HCX-CO2Me
starting material was consumed. The reaction mixture was trans-
ferred to a flask, and diluted with 100 mL of CH2Cl2. The organic
phase was separated, washed with water, brine and concentrated.
The resulting crude product was dissolved in CH2Cl2 (100 mL),
and treated with 20% HCl (50 mL). The reaction mixture was stirred
at room temperature for 8 - 12 h. The organic phase was separated,
and washed with water until the solution changed from green to
purple, and then the solution was washed with brine, dried over
Na2SO4, and concentrated to dryness. The resulting crude product
was chromatographed (silica, CH2Cl2) to afford the corresponding
hangman corrole featuring the carboxylic acid hanging group.
Experimental Section
General Methods. 1H NMR spectra (500 MHz) were recorded
on samples in CDCl3 at room temperature unless noted otherwise.
Silica gel (60 µm average particle size) was used for column
chromatography. Compounds 4-710 and 8-1027 were prepared as
described in the literature. THF (anhydrous), methanol (anhydrous),
CH2Cl2 (anhydrous), CHCl3 and all other chemicals were reagent
grade and used as received. LD-MS data were collected in the
presence of a 1,4-di-(2-5-phenyl-oxazolyl)benzene matrix.
Microwave-assisted reactions (deprotection and metalation of
corroles) were performed inside the cavity of a CEM Discover
microwave synthesis system equipped with infrared, pressure and
temperature sensors to monitor synthesis conditions. The reaction
vessels were 10 mL crimp-sealed thick-wall glass tubes. The
contents of each vessel were stirred with a magnetic stirrer.
General Protocol for the Synthesis of Hangman Corroles
(Method A). The Gryko corrole forming reaction26 was optimized.
CH2Cl2 (20 mL) was placed in an oven-dried round-bottom flask
(250 mL). A sample of compound 710 (0.244 g, 0.600 mmol) and
the corresponding dipyrromethane (1.50 mmol) were added to the
reaction flask. The reaction mixture was stirred about 10 min to
obtain a homogeneous mixture. In a separate flask, 1.3 mM TFA
(10 µL TFA was mixed with 100 mL CH2Cl2) solution was
prepared. A sample of freshly prepared TFA solution (20 mL) was
slowly added to first flask via syringe. A septum was placed and
the reaction mixture was stirred for 7 h at room temperature. A
sample of 2,3-dichloro-5,6-dicyano-1,4-benzoquinone (DDQ, 0.340
g, 1.50 mmol) was added and the reaction mixture was stirred for
15 min. The reaction mixture was loaded directly on a silica gel
column and eluted with dichloromethane until the first fluorescent
band was collected. Further purification of the compound by flash
column chromatography (silica, hexanes/ethanol (9:1)) afforded a
dark violet solid.
General Protocol for Metalation of Corroles under
Microwave Irradiation (Method C). A microwave glass tube (10
mL) containing a magnetic stir bar was charged with CH3CN or
CHCl3 and the corrole. The solution was stirred at room temperature
for 10 min to obtain a homogeneous mixture. The metal salt was
added (5-10 mol equiv versus corresponding hangman corrole)
and the resulting mixture was stirred at room temperature for 10
min. The reaction vessel was sealed with a septum and subjected
to microwave irradiation at 65 °C, for reactions performed in CHCl3
and at 85 °C for reactions performed in CH3CN. The protocol was
as follows: (1) heat the reaction vessel from room temperature to
65 °C (85 °C for CH3CN), (2) hold at 65 °C (85 °C for CH3CN)
and irradiate for 20 min (temperature overshoots of 70-75 °C
(90-95 °C for CH3CN)) were permitted; temperature was re-
established at 70 °C (85 °C for CH3CN) by using open flow valve
option), (3) allow the reaction mixture to cool to room temperature,
(4) check the reaction mixture by silica TLC analysis, (5) repeat
steps 1-4 until all of the free base corrole was consumed. Upon
complete reaction, triethylamine (10 mol equiv to metal salt) was
added to the solution, which was washed with water and brine,
dried over Na2SO4, and concentrated to dryness. The resulting crude
product was chromatographed on silica to afford the corresponding
metal complex corrole.
5,10,15-Tris(pentafluorophenyl)corrole (1). By following modi-
fied Lindsey porphyrin forming reaction,25 CHCl3 (900 mL) was
placed in an oven-dried round-bottom flask (1000 mL) and purged
with high flow of argon for 40 min. Pyrrole (5.5 mL, 80 mmol),
paraformaldehyde (0.15 g, 5.0 mmol) and pentafluorobenzaldehyde
(9.2 mL, 75 mmol) were added to the reaction flask, which was
covered with aluminum foil. The solution was purged with argon
for 45 min. A sample of BF3 ·OEt2 (3.3 mL, 26 mmol) was added
to the reaction mixture dropwise via syringe and the solution stirred
under argon in the dark for 45 min. DDQ (4.5 g, 20 mmol) was
added, and the resulting mixture was stirred for an additional 1 h.
A sample of triethylamine (3.6 mL, 26 mmol) was added and the
reaction mixture was stirred for additional 20 min. The resulting
reaction mixture was concentrated to dryness and the crude product
was subjected to silica chromatography (hexanes f hexanes:
CH2Cl2 (6:1) f hexanes:CH2Cl2 (4:1)) to afford a purple solid
General Protocol for Basic Hydrolysis of Hangman
Corroles under Microwave Irradiation (Method B). A sample
of HCX-CO2Me (0.1 mmol) was placed in a microwave glass tube
(10 mL) containing a magnetic stir bar. THF (2 mL, HPLC grade
lacking stabilizer) was added. The resulting mixture was stirred to
obtain a homogeneous solution. A sample of 6 N NaOH (2 mL)
was added. The vessel was sealed with a septum and subjected to
1
(1.342 g, 7% based on pentafluorobenzaldehyde). H NMR, ESI-
(22) Paolesse, R. In The Porphrin Handbook; Kadish, K. M., Smith, K. M.,
Guilard, R., Eds.; Academic Press: New York, 2000; Vol. 2, Chapter
12.
MS and absorption spectral data are consistent with that reported
in the literature.281H NMR (500 MHz, CDCl3) δ/ppm: 8.52-8.63
(broad s, 2H), 8.68 (d, J ) 4.5 Hz, 2H), 8.85 (d, J ) 4.5 Hz, 2H),
9.03 (d, J ) 4.5 Hz, 2H), pyrrolic (3H) were not observed at room
temperature. HR(ESI)-MS (M+H) (M ) C37H11F15N4): Calcd for
m/z ) 797.0817, obsd 797.0834. LD-MS obsd 796.11; λmax,abs /nm
(CH2Cl2) ) 416, 561, 604. λmax,em(416 exc) /nm ) 643. Anal. calcd
(23) Guilard, R.; Jerome, F.; Barbe, J.-M.; Gros, C. P.; Ou, Z.; Shao, J.;
Fischer, J.; Weiss, R.; Kadish, K. M. Inorg. Chem. 2001, 40, 4856–
4865.
(24) Kadish, K. M.; Ou, Z.; Shao, J.; Gros, C. P.; Barbe, J.-M.; Jerome,
F.; Bolze, F.; Burdet, F.; Guilard, R. Inorg. Chem. 2002, 41, 3990–
4005.
(25) Lindsey, J. S.; Wagner, R. W. J. Org. Chem. 1989, 54, 828–836.
(26) Gryko, D. T.; Jadach, K. J. Org. Chem. 2001, 66, 4267–4375.
(27) Laha, J. K.; Dhanalekshmi, S.; Taniguchi, M.; Ambroise, A.; Lindsey,
J. S. Org. Process Res. DeV. 2003, 7, 799–812.
(28) Gross, Z.; Galili, N.; Saltsman, I. Angew. Chem. 1999, 38, 1427–
1429.
9
132 J. AM. CHEM. SOC. VOL. 133, NO. 1, 2011

Hangman Corroles
A R T I C L E S
for C37H11F15N4: C, 55.79; H, 1.39; N, 7.03. Found: C, 55.30; H,
0.96; N, 6.95.
calcd for C55H37CoF10N4O3 ·3H2O: C, 59.79; H, 3.92; N, 5.07.
Found: C, 59.77; H, 4.09; N, 5.26.
5,10,15-Tris(pentafluorophenylcorrolato)cobalt(III) (1-Co).
Method C was followed. A solution of 1 (79 mg, 0.099 mmol) in
CH3CN (2 mL) was treated with Co(OAc)2 (175 mg, 0.990 mmol).
The standard workup was followed by column chromatography
(silica, Hexanes/CH2Cl2 (1:1) f CH2Cl2) to afford a dark-brown
solid (79 mg, 95%). 1H NMR (500 MHz, CDCl3) δ/ppm: 6.48 (d,
J ) 4.5 Hz, 2H), 6.53 (d, J ) 4.5 Hz, 2H), 6.78-6.86 (m, 2H),
7.08-7.14 (m, 2H). HR(ESI)-MS in negative ion mode (M-) (M
) C37H8F15N4Co): Calcd for m/z ) 851.9847, obsd 851.9835. LD-
MS obsd 852.42, dimer 1073.68, trimer 2554.04; λmax,abs /nm
(CH2Cl2) ) 398, 534. Anal. calcd for C37H8F15N4Co: C, 52.13; H,
0.95; N, 6.57. Found: C, 52.16; H, 1.09; N, 7.07.
10-(4-(5-Methoxycarbonyl-2,7-ditert-butyl-9,9-dimethylxanthene))-
5,15-bis(4-tert-butylphenyl)corrole (HCX2-CO2Me). Method A was
followed. A mixture of 7 (0.244 g, 0.600 mmol), and 5-tert-butyl
dipyrromethane 9 (0.418 g, 1.50 mmol) in CH2Cl2 (20 mL) was
treated with 1.3 M freshly prepared TFA solution (20 mL). The
reaction mixture was stirred for 7 h and treated with DDQ (0.340
g, 1.50 mmol). The resulting crude product was chromatographed
to afford a purple solid (113 mg, 20%). 1H NMR (500 MHz, CDCl3)
δ/ppm: 0.85 (s, 3H), 1.29 (s, 9H), 1.50 (s, 9H), 1.59 (s, 18H), 1.91
(s, 6H), 7.34 (d, J ) 2.5 Hz, 1H), 7.65 (d, J ) 2.5 Hz, 1H), 7.81
(d, J ) 2.5 Hz, 1H), 7.84 (d, J ) 4.5 Hz, 4H), 7.94 (d, J ) 2.5 Hz,
1H), 8.32-8.34 (m, 4H), 8.48 (d, J ) 4.5 Hz, 2H), 8.62 -8.64 (m,
2H), 8.89 (d, J ) 4.5 Hz, 2H), 8.94-8.96 (m, 2H). The pyrrolic
protons (3H) were not observed at room temperature; HR(ESI)-
MS (M + H) (M ) C64H68N4O3): Calcd for m/z ) 941.5364, obsd
941.5368 LD-MS obsd 940.8. λmax,abs/nm (CH2Cl2) ) 407, 425,
559, 602, 634. λmax,em(407 exc) /nm ) 642. Anal. calcd for
C64H68N4O3: C, 81.67; H, 7.28; N, 5.95. Found: C, 81.72; H, 7.51;
N, 5.67.
10-(4-(5-Methoxycarbonyl-2,7-ditert-butyl-9,9-dimethylxanthene))-
5,15-bis(pentafluorophenyl)corrole (HCX1-CO2Me). Method A was
followed. A mixture of 7 (0.244 g, 0.600 mmol) and 5-pentafluo-
rophenyl dipyrromethane 8 (0.468 g, 1.50 mmol) in CH2Cl2 (20
mL) were treated with 1.3 M freshly prepared TFA solution (20
mL). The reaction mixture was stirred for 7 h and treated with DDQ
(0.340 g, 1.50 mmol). The resulting crude product was chromato-
1
graphed to afford a purple solid (139 mg, 23%). H NMR (500
10-(4-(5-Hydroxycarbonyl-2,7-ditert-butyl-9,9-dimethylxanthene))-
5,15-bis(4-tert-butylphenyl)corrole (HCX2-CO2H). Method B was
followed. A solution of HCX2-CO2Me (94 mg, 0.10 mmol) in THF
(1 mL) was treated with 6 N NaOH (2 mL). The reaction mixture
was subjected to microwave irradiation for 8 h. The aqueous work
up was performed and the resulting crude was treated with 20%
HCl. The biphasic reaction mixture was stirred overnight. The
organic phase was separated, and washed. The crude product was
chromatographed (silica, CH2Cl2) to afford a purple solid (65 mg,
MHz, CDCl3) δ/ppm: 1.03 (s, 3H), 1.27 (s, 9H), 1.50 (s, 9H), 1.91
(s, 6H), 7.29 (d, J ) 2.5 Hz, 1H), 7.65 (d, J ) 2.5 Hz, 1H), 7.85
(d, J ) 2.5 Hz, 1H), 7.91 (d, J ) 2.5 Hz, 1H), 8.52-8.56 (m, 2H),
8.65 (s, 4H), 9.12 (d, J ) 4.5 Hz, 2H), pyrrolic protons were not
observed at room temperature. HR(ESI)-MS (M + H) (M )
C56H42F10N4O3): Calcd for m/z ) 1009.3170, obsd 1009.3153. LD-
MS obsd 1009.17. λmax,abs/nm (CH2Cl2) ) 411, 558, 613. λmax,em(414
exc)/nm ) 648. Anal. calcd for C56H42F10N4O3 ·2H2O: C, 64.37;
H, 4.44; N, 5.36. Found: C, 64.24; H, 4.07; N, 5.61.
10-(4-(5-Methoxycarbonyl-2,7-ditert-butyl-9,9-dimethylxanthene))-
5,15-bis(pentafluorophenyl)cobaltcorrole (CoHXC1-CO2Me). Method
C was followed. A solution of HCX1-CO2Me (100 mg, 0.10 mmol)
in CH3CN (2 mL) was treated with Co(OAc)2 (88 mg, 0.50 mmol).
Standard workup was followed by column chromatography (silica,
CHCl3 f CHCl3:MeOH (100:1) f CHCl3:MeOH (10:1)) to afford
a dark brown solid (82 mg, 78%). HR(ESI)-MS (M-) (M )
C56H39CoF10N4O3): Calcd for m/z ) 1064.2200, obsd 1064.2242.
LD-MS obsd 1064.55. λmax,abs/nm (CH2Cl2) ) 310, 425, 620. Anal.
calcd for C56H39CoF10N4O3: C, 63.16; H, 3.69; N, 5.26. Found: C,
62.96; H, 3.78; N, 5.22.
1
70%). H NMR (500 MHz, CDCl3) δ/ppm: 1.28 (s, 9H), 1.53 (s,
9H), 1.59 (s, 18H), 1.98 (s, 6H), 7.73 (s, 1H), 7.76 (s, 1H),
7.80-7.83 (m, 4H), 7.88 (s, 1H), 8.02 (s, 1H), 8.30-8.32 (m, 4H),
8.42-8.45 (m, 2H), 8.62-8.65 (m, 2H), 8.90-8.02 (m, 2H),
8.97-8.99 (m, 2H). The pyrrolic (3H) and carboxylic acid protons
were not observed at room temperature. HR(ESI)-MS (M + H)
(M ) C63H66N4O3): Calcd for m/z ) 927.5313, obsd 927.5211.
LD-MS obsd 926.55. λmax,abs/nm (CH2Cl2) ) 407, 426, 565, 600,
634. λmax,em(407 exc) /nm ) 645. Anal. calcd for C63H66N4O3: C,
81.61; H, 7.17; N, 6.04. Found: C, 81.20; H, 6.90; N, 5.97.
10-(4-(5-Hydroxycarbonyl-2,7-ditert-butyl-9,9-dimethylxanthene))-
5,15-bis(4-tert-butylphenyl)cobaltcorrole (CoHCX2-CO2H). Method
C was followed. A solution of HCX2-CO2H (92 mg, 0.10 mmol)
in CH3CN (2 mL) was treated with Co(OAc)2 (88 mg, 0.50 mmol).
Standard workup was followed by column chromatography (silica,
CHCl3 f CHCl3: MeOH (100:1) f CHCl3:MeOH (10:1)) to afford
a dark brown solid (83 mg, 89%). HR(ESI)-MS (M + H) (M )
C63H63CoN4O3): Calcd for m/z ) 982.4232, obsd 982.4225. LD-
MS obsd 982.02, dimer 1964.59. λmax,abs/nm (CH2Cl2) ) 358, 438,
529. Anal. calcd for C63H63CoN4O3 ·3H2O. C, 72.96; H, 6.71; N,
5.40. Found: C, 73.34; H, 6.14; N, 5.77.
10-(4-(5-Hydroxycarbonyl-2,7-ditert-butyl-9,9-dimethylxanthene))-
5,15-bis(pentafluorophenyl)corrole (HCX1-CO2H). Method B was
followed. A solution of HCX1-CO2Me (100 mg, 0.0992 mmol) in
THF (1 mL) was treated with 6 N NaOH (2 mL). The reaction
mixture was subjected to microwave irradiation for 6 h. The aqueous
work up was performed and the resulting crude was treated with
20% HCl. The biphasic reaction mixture was stirred overnight. The
organic phase was separated, and washed. The crude product was
chromatographed (silica, hexanes/CH2Cl2 (5:3) f hexanes/CH2Cl2
1
(5:2) f CH2Cl2) to afford a purple solid (81 mg, 82%). H NMR
(500 MHz, CDCl3) δ/ppm: 1.27 (s, 9H), 1.54 (s, 9H), 1.97 (s, 6H),
7.60-7.62 (m, 1H), 7.77-7.78 (m, 1H), 7.96-7.97 (m, 1H),
8.0-8.04 (m, 1H), 8.65-8.68 (m, 4H), 8.75-8.76 (m, 2H), 9.21
(d, J ) 4.0 Hz, 2H), pyrrolic (3H) and carboxylic acid protons
were not observed at room temperature. HR(ESI)-MS (M - H)
(M ) C55H40F10N4O3): Calcd for m/z ) 993.2868, obsd 993.2859.
LD-MS obsd 994.3. λmax,abs/nm (CH2Cl2) ) 409, 563, 610.
10-(4-(5-Methoxycarbonyl-2,7-ditert-butyl-9,9-dimethylxanthene))-
5,15-bis(2,4,6-trimethylphenyl)corrole (HCX3-CO2Me). Method A
was followed. A mixture of 7 (0.244 g, 0.600 mmol), and 5-mesityl
dipyrromethane 10 (0.396 g, 1.50 mmol) in CH2Cl2 (20 mL) was
treated with 1.3 M freshly prepared TFA solution (20 mL). The
reaction mixture was stirred for 7 h and treated with DDQ (0.340
g, 1.50 mmol). The resulting crude product was chromatographed
to afford a purple solid (107 mg, 19%). 1H NMR (500 MHz, CDCl3)
δ/ppm: 0.55 (s, 3H), 1.26 (s, 9H), 1.47 (s, 9H), 1.87 (s, 6H), 1.89
(s, 6H), 2.04 (s, 6H), 2.59 (s, 6H), 7.22-7.28 (m, 2H), 7.26-7.27
(m, 2H, overlapped with CDCl3 peak), 7.33 (d, J ) 2.5 Hz, 1H),
7.64 (d, J ) 2.5 Hz, 1H), 7.78 (d, J ) 2.5 Hz, 1H), 7.86 (d, J )
2.5 Hz, 1H), 8.32 (d, J ) 4 Hz, 2H), 8.39 (d, J ) 4.5 Hz, 2H),
8.43 (d, J ) 4.5 Hz, 2H), 8.88 (d, J ) 4 Hz, 2H). The pyrrolic
protons (3H) were not observed at room temperature. HR(ESI)-
MS (M + H) (M ) C62H64N4O3): Calcd for m/z ) 913.5051, obsd
913.5033. LD-MS obsd 912.62. λmax,abs /nm (CH2Cl2) ) 406, 423,
λmax,em(409 exc) /nm ) 651.
10-(4-(5-Hydroxycarbonyl-2,7-ditert-butyl-9,9-dimethylxanthene))-
5,15-bis(pentafluorophenyl)cobaltcorrole (CoHCX1-CO2H). Method
C was followed. A solution of HCX1-CO2H (93 mg, 0.10 mmol)
in CH3CN (2 mL) was treated with Co(OAc)2 (88 mg, 0.50 mmol).
Standard workup was followed by column chromatography (silica,
CHCl3 f CHCl3:MeOH (100:1) f CHCl3:MeOH (10:1)) to afford
a dark brown solid (72 mg, 69%). HR(ESI)-MS (M + H) (M )
C55H37CoF10N4O3): Calcd for m/z ) 1050.2043, obsd 1050.2143.
LD-MS obsd 1050.78. λmax,abs/nm (CH2Cl2) ) 303, 430, 609. Anal.
9
J. AM. CHEM. SOC. VOL. 133, NO. 1, 2011 133

A R T I C L E S
Dogutan et al.
559, 601, 633. λmax,em(414 exc) /nm ) 645. Anal. calcd for
C63H66N4O3: C, 81.54; H, 7.06; N, 6.14. Found: C, 81.59; H, 6.94;
N, 6.18.
and solution statistics are summarized in Table S1 (Supporting
Information). All thermal ellipsoid plots are drawn at 50%
probability level, with solvent molecules omitted.
Cyclic Voltammetry. Cyclic voltammograms (CV) were re-
corded in CH3CN or CH2Cl2 solutions containing 0.1 M NBu4PF6
(tetrabutylammonium hexafluorophosphate) and the corrole com-
pound at 1.05 mM. A three compartment cell was employed
possessing a 0.07 cm2 glassy carbon button electrode as the working
electrode, Pt wire as the auxiliary electrode, and Ag/AgCl as a
reference electrode. CVs were collected with scan rates of 10-100
mV/s with iR compensation.
Spectroscopic Methods. UV-vis spectra were recorded on a
Varian Cary 5000 UV-vis-NIR spectrophotometer on solutions of
anhydrous CH2Cl2, THF or CH2Cl2:EtOH (3:1) at room temperature
contained in quartz cuvettes. Steady state emission spectra were
recorded on an automated Photon Technology International (PTI)
QM 4 fluorimeter equipped with a 150-W Xe arc lamp and a
Hamamatsu R928 photomultiplier tube. Excitation light was
wavelength selected with glass filters. Solution samples were
prepared under ambient atmosphere in anhydrous CH2Cl2 or THF
and contained in screw-cap quartz fluorescence cells.
Continuous wave X-band EPR spectra were recorded using a
Bruker ESP 300 spectrometer, fitted with an Oxford ESR 910
cryostat cooled with liquid helium. All spectra were recorded on
frozen solution samples at temperatures of 2-80 K. Data analysis
and spectral simulations were executed using the SpinCount
software package developed by Prof. Mike Hendrich at Carnegie
Mellon University.29 EPR spectra of the oxygen adduct were
simulated akin to the analysis of the EPR spectra recorded for single
crystals of the oxygen adduct of vitamin B12.30,31 As such, in order
to get a good representation of the experimental data we found
that the ACo tensor, A1 ) 91 ( 8 MHz, A2 ) 29 ( 4 MHz, and A3
) 35 ( 2 MHz, needs to be rotated from the g-tensor, g1 ) 2.11
( 0.01, g2 ) 2.02 ( 0.01, g3 ) 1.98 ( 0.01 by R ) 12° ( 5° and
ꢀ ) 9° ( 5°.
10-(4-(5-Hydroxycarbonyl-2,7-ditert-butyl-9,9-dimethylxanthene))-
5,15-bis(2,4,6-trimethylphenyl)corrole (HCX3-CO2H). Method B
was followed. A solution of HCX3-CO2Me (92 mg, 0.10 mmol)
in THF (1 mL) was treated with 6 N NaOH (2 mL). The reaction
mixture was subjected to microwave irradiation for 8 h. The aqueous
work up was performed and the resulting crude was treated with
20% HCl. The biphasic reaction mixture was stirred overnight. The
organic phase was separated, and washed. Silica chromatography
(CH2Cl2) afforded a purple solid (63 mg, 70%). 1H NMR (500 MHz,
CDCl3) δ/ppm: 1.26 (s, 9H), 1.52 (s, 9H), 1.87 (s, 6H), 1.94 (s,
6H), 2.03 (s, 6H), 2.59 (s, 6H), 7.23-7.24 (m, 2H), 7.26-7.27
(m, 2H, overlapped with CDCl3 peak), 7.69 (d, J ) 2.5 Hz, 1H),
7.72 (d, J ) 2.0 Hz, 1H), 7.85 (d, J ) 2.0 Hz, 1H), 8.05 (d, J )
2.0 Hz, 1H), 8.33 (d, J ) 4 Hz, 2H), 8.37 (d, J ) 4.5 Hz, 2H),
8.46 (d, J ) 4.5 Hz, 2H), 8.91 (d, J ) 4 Hz, 2H). The pyrrolic
(3H) and carboxylic acid protons were not observed at room
temperature. HR(ESI)-MS (M + H) (M ) C61H62N4O3): Calcd for
m/z ) 899.4895, obsd 899.4888. λmax,abs/nm (CH2Cl2) ) 408, 425,
566, 601, 635, λmax,em(408 exc) /nm ) 648. Anal. calcd for
C61H62N4O3: C, 81.48; H, 6.95; N, 6.23. Found: C, 80.95; H, 6.49;
N, 6.05.
10-(4-(5-Hydroxycarbonyl-2,7-ditert-butyl-9,9-dimethylxanthene))-
5,15-bis(2,4,6-trimethylphenyl)cobaltcorrole (CoHCX3-CO2H).
Method C was followed. A solution of HCX3-CO2H (95 mg, 0.11
mmol) in CH3CN (2 mL) was treated with Co(OAc)2 (88 mg, 0.50
mmol). Standard workup was followed by column chromatography
(silica,CHCl3 f CHCl3: MeOH (100:1) f CHCl3:MeOH (10:1))
to afford a dark brown solid (78 mg, 81%). HR(ESI)-MS (M - H)
(M ) C61H59CoN4O3): Calcd for m/z ) 953.3846 obsd 953.3839.
LD-MS obsd 954.63. λmax,abs /nm (CH2Cl2) ) 404, 534. Anal. calcd
for C61H59CON4O3. C, 76.71; H, 6.23; N, 5.87. Found: C, 76.39; H,
6.25; N, 5.27.
X-Ray Crystallographic Details. Single crystals of 1-CoCl,
HCX1-CO2H and HCX2-CO2Me were obtained by slow evapora-
tion of a hexanes and CH2Cl2 mixture. 1-CoCl was mounted on a
Bruker three circle goniometer platform equipped with an APEX
detector. A single crystal of HCX2-CO2Me was mounted on a
Bruker three circle goniometer platform equipped with an APEX
2 detector, whereas HCX1-CO2H was mounted on a Bruker four
circle goniometer equipped with an APEX 2 detector. A graphite
monochromator was employed for wavelength selection of the Cu
KR radiation (λ ) 1.54178 Å, HCX2-CO2Me and HCX1-CO2H)
or Mo KR radiation (λ ) 0.71073 Å, 1-CoCl). The data were
processed and refined using the program SAINT supplied by
Siemens Industrial Automation. Structures were solved by direct
methods in SHELXS and refined by standard difference Fourier
techniques in the SHELXTL program suite (6.10 v., Sheldrick
G. M., and Siemens Industrial Automation, 2000). For the solution
of the 1-CoCl structure, hydrogen atoms were placed in calculated
positions using the standard riding model and refined isotropically;
all other atoms were refined anisotropically. For the solution of
the HCX2-CO2Me and HCX1-CO2H structures, hydrogen atoms
bound to nitrogen or oxygen were located in the difference map
and restrained to a distance of 0.84 (O-H) or 0.88 (N-H) Å. The
isotropic displacement parameters of the hydrogen atoms were fixed
to 1.2 times the U value of the atoms to which they are bonded.
All non-hydrogen atoms were refined anisotropically. The structure
of 1-CoCl contained a dichloromethane solvate molecule that was
modeled as a two-part disorder. The structure of HCX1-CO2H
contained a disordered hexane molecule modeled in a similar
fashion. The (1,2) and (1,3) distances of all disordered parts were
restrained to be similar using the SADI command; the rigid-bond
restraints SIMU and DELU were also used on disordered parts. In
the disordered hexane of molecule of HCX1-CO2H, distance
restraints were used to fix carbon-carbon 1-2 and 1-3 distances
(1.54 Å and 2.52 Å, respectively). Unit cell parameters, morphology,
The DFT calculations were performed with Gaussian’09 (revision
A.2.) quantum mechanical software package. The B3LYP/6-311G
hybrid functional/basis set combination was used for all calculations.
For 1-CoCl, a structural geometric model with an initial Cs point
group symmetry was constructed and used as starting point for
geometry optimization. Subsequent optimizations by relieving the
symmetry constraints did not lead to a distortion from the stationary
point found for the Cs point group. The nature of the predicted
electronic state was assessed on the basis of time-dependent (TD)
DFT calculations, by analyzing the gross orbital population and
inspecting the calculated Mulliken charges, spin densities, and
values of the Fermi and spin-dipolar components of the hyperfine
coupling tensors.
Oxygen Reduction Measurements. A solution was prepared
consisting of 3.9 mL THF containing, 0.3 mM catalyst, 4 mg
MWCNT, and 45 µL 5% Nafion. For each experiment, the mixture
was sonicated for 20 min and then a 10 µL drop was applied to the
surface of a polished glassy carbon electrode and allowed to
evaporate to afford a film containing a MWCNT loading of 50 µgc/
cm2 and a catalyst loading of 3 nmol. The electrocatalytic activity
was measured in a 3-electrode cell in 0.5 M H2SO4 by rotating
ring disk electrode (RRDE) techniques. For oxygen reduction, the
counter electrode was a platinum wire and the reference electrode
was a saturated calomel electrode calibrated against reversible
hydrogen. All potentials in aqueous solution are given vs RHE. In
the RRDE experiments the platinum ring was held at 1.2 V vs SCE
(29) This program can be downloaded from http://www.chem.cmu.edu/
groups/hendrich/ facilities/index.html.
(30) Ramdhanie, B.; Telser, J.; Caneschi, A.; Zakharov, L. N.; Rheingold,
A. L.; Goldberg, D. P. J. Am. Chem. Soc. 2004, 126, 2515–2525.
(31) Jo¨rin, E.; Schweiger, A.; Gu¨nthard, Hs. H. Chem. Phys. Lett. 1979,
61, 228–232.
9
134 J. AM. CHEM. SOC. VOL. 133, NO. 1, 2011

Hangman Corroles
A R T I C L E S
Scheme 1
Figure 1. Crystal structure of 1-CoCl with ellipsoids shown at the 50%
probability level. The dichloromethane solvate is omitted for clarity, and
H-atoms are shown in idealized positions. Data were collected at 100(2)
K.
affords formation of the corresponding porphyrin instead of
target corrole. (3) The route described here does not require
presence of solid support such as, florisil, silica or alumina for
corrole synthesis. Finally, (4) the synthesis is scalable and
affords the corrole in a one-pot synthesis at much higher yields.
Microwave irradiation has been employed for the synthesis of
free base corrole monomers34 and for the metalation of porphyrin,35
though a similar metalation chemistry for corroles has yet to be
explored. We find that microwave irradiation of corroles in the
presence of a metal acetate drives metalation in high yields;
microwave irradiation of 1 in the presence of excess Co(OAc)2
affords 1-Co in 95% isolated yield. On the basis of the 1H NMR
spectrum and SQUID data, 1-Co is paramagnetic. The magnetic
moment is ca. 1 cm3K/mol at room temperature, which decreases
gradually to 0.1 cm3K/mol at 2 K.
Crystals of 1-CoCl were obtained by slow evaporation of a
dichloromethane solution of 1-Co; the source of Cl is likely
from trace impurities of HCl in the dichloromethane. The X-ray
crystal structure of 1-CoCl is shown in Figure 1; the central
cobalt atom resides in a square pyramidal coordination environ-
ment and the corrole macrocycle shows only slight deviations
from an ideal planar system. The Co-N distances are not
equally distributed; they are split into two groups. The Co-N
bonds on the side with the direct C-C pyrrole bridge are shorter
(1.888 and 1.887 Å) than those opposite to the methylene pyrrole
bridge (1.911 and 1.923 Å). One of the pentafluorophenyl
substituents is almost perpendicular to the macrocycle plane
(83°) whereas the other two are considerably less twisted (∼59°).
The Co-Cl distance is 2.243 Å and the displacement of the
cobalt atom out of the N4-plane is 0.413 Å. This displacement
is unusually large relative to that of known five-coordinate
Co(III) corroles, which exhibit displacements between 0.29 and
0.38 Å from the plane.36-40 These results contrast six coordinate
3-/4-
and the collection efficiency calibrated against the Fe(CN)6
reduction process.32
Results and Discussion
Synthesis. The observation of corrole as a minor side-product
with Lindsey porphyrin synthesis conditions prompted us to
investigate the reaction conditions in detail with the goal of
obtaining A3 corrole monomer as a major product. The synthesis
of the 5,10,15-tris(pentafluorophenyl)corrole (1) was undertaken
as an exemplar; the synthetic strategy is shown in Scheme 1.
Large scale (75 mmol) synthesis of 1 was performed by
condensation of commercially available pyrrole, pentafluo-
robenzaldehyde and paraformaldehyde under Lindsey condi-
tions.25 Paraformaldehyde has been utilized for the syntheses
of corrole monomer and corrole dimer.33 The role of paraform-
aldehyde in the corrole forming reaction, though not delineated,
is likely to tune the acid concentration for the formation of a
bilane that delivers corrole rather than corresponding porphyrin.
What is certain is that if paraformaldehyde is not used, the
formation of porphyrin is preferred. Thin layer chromatography
(TLC) and matrix-assisted laser desorption/ionization time of
flight (MALDI-TOF) analysis of the crude reaction mixture
revealed 1 as a major product. Trace amounts of porphyrins 2
and 3 were also observed as side products. The desired product
1 was isolated by flash column chromatography in 7% yield.
We note that the yield of the corrole diminishes when the
reaction is executed on small scale (3.75 mmol) and A4
porphyrin 2 is obtained in greater predominance.
The new synthesis of tris(pentafluorophenyl)corrole, 1, is an
advance over previous methods for the following reasons. (1)
The new route affording compound 1 does not require harsh
conditions (room temperature vs 100 °C) for intramolecular
cyclization. The corrole forming reaction at higher cyclization
temperature causes partial decomposition of the bilane inter-
mediate and the polymerization of the linear tetrapyrrolic species
which reduces the yield for corrole formation dramatically. (2)
The corrole forming reaction at lower concentration favors the
desired intramolecular cyclization of the bilane as opposed to
intermolecular reaction between dipyrromethane species at the
higher concentrations of previous methods. The former reaction
(34) Collman, J. P.; Decreau, R. A. Tetrahedron Lett. 2003, 44, 1207–
1210.
(35) Liu, M. O.; Hu, A. T. J. Organomet. Chem. 2004, 689, 2450–2455.
(36) Hitchcock, P. B.; McLaughlin, G. M. J. Chem. Soc., Dalton Trans.
1976, 1927–1930.
(37) Paolesse, R.; Licoccia, S.; Bandoli, G.; Dolmella, A.; Boschi, T. Inorg.
Chem. 1994, 33, 1171–1176.
(38) Will, S.; Lex, J.; Vogel, E.; Adamian, V. A.; Vancaemelbecke, E.;
Kadish, K. M. Inorg. Chem. 1996, 35, 5577–5583.
(39) Barbe, J. M.; Burdet, F.; Espinosa, E.; Guilard, R. Eur. J. Inorg. Chem.
2005, 6, 1032–1041.
(32) Myuller, L.; Nekrasov, L. Electrochim. Acta 1964, 9, 1015–1023.
(33) Koszarna, B.; Gryko, D. T. Chem. Commun. 2007, 2994–2996.
9
J. AM. CHEM. SOC. VOL. 133, NO. 1, 2011 135

A R T I C L E S
Dogutan et al.
Scheme 2a
a (a) (1) Phenyllithium in Hexanes-THF; (2) DMF, H2O; (3) neopentyl glycol, p-toluenesulfonic acid, toluene, under Ar, reflux. (b) (1) Phenyllithium in
hexanes-THF; (2) CO2(g), (3) trifluoroacetic acid, water. (c) Methanol, H2SO4, reflux. (d) (1) DPM 8 (HCX-1), 9 (HCX-2) and 10 (HCX-3), TFA, CH2Cl2;
(2) DDQ. (e) (1) THF/6 N NaOH, microwave irradiation; (2) 20% HCl, (3) Co(OAc)2, CH3CN, microwave irradiation.
Table 1. Statistical Synthesis of Hangman Corroles Xanthene
Co(III) corroles such as pyridine-coordinated, (Me4Ph5Cor)-
Co(py)2,41 which displays a cobalt ion that resides within the
plane of the corrole macrocycle. Similarly, a structurally
characterized five-coordinate triphenylsilylacetylide-ligated co-
balt(III) corrolazine displays a nearly planar geometry,30 in stark
contrast to the structure of 1-Co. Structurally characterized
Co(IV) corroles are rare, especially when axially coordinated
by Cl. To the best of our knowledge, the only other example of
a structurally characterized Co(IV) chloride corrole is a Pacman
structure42 bearing no aryl meso substituents. Here too the
displacement of Co from the corrole plane is small (0.19 Å)
relative to that observed in 1-CoCl, though the Co-Cl distance
is similar (d(Co-Cl) ) 2.275(2) Å. The structural metrics for
1-CoCl are quite similar to Fe(IV) chloride corrole complexes.43,44
under Microwave Irradiation
a Corrole formation was performed under Gryko conditions (ref 26).
The yields are based on aldehyde 7. b Overall yield is for free-base
corroles, steps d and e in Scheme 2.
The HCX compounds are delivered by following the proce-
dure outlined in Scheme 2. Compounds 4-7 were prepared by
following a published procedure.45 Hangman corroles were
obtained by modifying the conditions of Gryko for corrole
formation.26 Dipyrromethane was condensed with 7 under
Brønsted acid catalyzed conditions. Our investigations establish
that the macrocycle formation occurs with better yields (∼20%)
when the acid conditions are 0.6 mM TFA in CH2Cl2 (Table
1). The presumed intermediate, bilane,46,47 was oxidized in situ
by 4e- and 4H+ using DDQ. A crystal structure was obtained
for free base HCX2-OMe and it is displayed in Figure 2a. The
xanthene backbone is tilted considerably relative to the corrole
macrocycle, with an observed angle of 69° between the mean
N4 plane and the plane defined by the six-member aryl ring of
the xanthene bonded directly to the meso carbon. The macro-
cycle is puckered, with one of the pyrrolic nitrogens canted out
of the plane defined by the three remaining pyrrole nitrogens.
The three protons were found to be delocalized among all four
nitrogen atoms.
(40) Paolesse, R.; Nardis, S.; Sagone, F.; Khoury, R. G. J. Org. Chem.
2001, 66, 550–556.
(41) Guilard, R.; Gros, C. P.; Bolze, F. F.; Jerome, Z. P.; Ou, J. G.; Shao,
J.; Fischer, J.; Weiss, R.; Kadish, K. M. Inorg. Chem. 2001, 40, 4845–
4855.
Deprotection of the ester functionality on the xanthene
backbone is a difficult step. For hangman porphyrins, ester
deprotection has been attempted under both basic and acidic
conditions. The basic hydrolysis required refluxing the porphyrin
(42) Guilard, R.; Burde, F.; Barbe, J. M.; Gros, C. P.; Espinosa, E.; Shao,
J. G.; Ou, Z. P.; Zhan, R. Q.; Kadish, K. M. Inorg. Chem. 2005, 44,
3972–3983.
(43) Nardis, S.; Paolesse, R.; Licoccia, S.; Fronczek, F. R.; Vicente,
M. G. H.; Shokhireva, T. K.; Cai, S.; Walker, F. A. Inorg. Chem.
2005, 44, 7030–7046.
(44) Vogel, E.; Will, S.; Tilling, A. S.; Neumann, L.; Lex, J.; Bill, E.;
Trautwein, A. X.; Wieghardt, K. Angew. Chem., Int. Ed. Engl. 1994,
33, 731–735.
(46) Gossauer, A.; Engel, J. In The Porphyrins; Dolphin, D., Ed.; Academic
Press: New York, 1978; Vol. 2, pp 197-253.
(47) Falk, H. The Chemistry of Linear Oligopyrroles and Bile Pigments;
Springer-Verlag: Wien, 1989.
(45) Chng, L. L.; Chang, C. J.; Nocera, D. G. Org. Lett. 2003, 5, 2421–
2424.
9
136 J. AM. CHEM. SOC. VOL. 133, NO. 1, 2011

Hangman Corroles
A R T I C L E S
CO2H shown in Figure 2b is similar to the ester protected
hangman complex. The xanthene backbone is tilted relative to
the corrole macrocycle but not as much (53° for HCX1-CO2H).
The structure of HCX1-CO2H is distinguished by the proton
of the carboxylic acid hanging group; an intramolecular
hydrogen bond is observed between the carboxylic acid proton
and the oxygen of the xanthene backbone, with an observed
distance of 2.606(2) Å between the donor and the acceptor
oxygen atoms.
The resulting HCX-CO2H platforms were metalated ef-
ficiently (>95%) with Co(OAc)2 using the same conditions as
described for the metalation of 1. The cobalt complexes of HCX-
CO2H were obtained in overall yields of 15 - 20% based on
aldehyde 7.
Spectroscopy and Electronic Structure. The UV-vis absorp-
tion spectra of the cobalt hangman corroles offer an interesting
comparison to their porphyrin relatives. Figure S14 shows the
absorption spectra of the analogous hangman corrole and
porphyrin series. The Soret and Q bands of HCX compounds
are very broad and exhibit considerably attenuated extinction
coefficients as compared to their porphyrin congeners (ca.
50 000 M-1 cm-1 for CoHCX vs 200 000 M-1 cm-1 for
CoHPX). This asymmetry of the Soret bands is attributed to
the lower symmetry of the corrole macrocycle rings.56 The
absence of low-lying d-d states in free-base HCX compounds
removes efficient nonradiative decay pathways from the system;
consequently, the free base HCX corroles are strongly lumi-
nescent (Figures S3d-S6d, S7f and S8f, Supporting Informa-
tion) at λem,max ∼ 640 - 650 nm.
A formal electron count involving only the metal does not
often adequately account for the electronic structure of corrole
complexes, especially when they are highly oxidized. Spectro-
scopic measurements of highly oxidized corroles are suggestive
of a radical cationic state for the corrole macrocycle.51-56 EPR
can be especially helpful in discerning the electron distribution
between the metal center and the corrole ligand. The spectro-
scopic behavior of 1-CoCl was assessed by X-band EPR
spectroscopy using CH2Cl2 and THF solution samples, both at
cryogenic temperatures in frozen form as well as at room
temperature in liquid form. While no signal could be detected
at room temperature for fluid solution samples, Figure 3 shows
the EPR spectrum of 1-CoCl recorded at 18 K as a black trace
and a simulation of that spectrum is shown by the gray trace.
A rhombic signal with a rich hyperfine structure spanning 31
mT is observed. Spectra were recorded at temperatures from 2
to 80 K. At 2 K the observed signal saturates easily, even at 2
µW applied microwave power. For the temperature range where
this signal was investigated, its intensity (signal × temperature)
was found to be constant, indicating that indeed, the signal arises
Figure 2. Crystal structures of free base hangman corroles with ester and
carboxylic hanging groups (a) HCX2-CO2Me and (b) HCX1-CO2H,
respectively. For (a), the hexane solvate molecule and hydrogen atoms are
omitted for clarity. Ellipsoids for both structures are depicted at 50%
probability level.
for 3 days under nitrogen atmosphere48 whereas the acidic
hydrolysis is performed under even harsher conditions49 of reflux
in a strong acid mixture (sulfuric acid/acetic acid 4:1) for 7 days
under nitrogen. We first attempted ester deprotection by basic
hydrolysis of HCX-CO2Me. The reaction gave no product even
after the solution was refluxed for 2 weeks under nitrogen
atmosphere; only the starting HCX-CO2Me material was
recovered from decomposition products. In an attempt to
accelerate the sluggish reactivity associated with the HCX
platform, we turned our attention to the application of micro-
wave synthetic methods because we have found it to be effective
for the deprotection of hanging ester groups of porphyrins.50
Basic hydrolysis of the ester group on HCX-CO2Me under
microwave irradiation proceeds within 4-16 h as compared to
the several days of the aforementioned conventional thermal
hydrolysis conditions. The resulting product was treated with
20% HCl to furnish HCX-CO2H complexes in high yields for
such types of hangman macrocycles (Table 1). A crystal
structure of the hydrolyzed freebase hangman corrole HCX1-
1
from an isolated Kramers doublet, as expected for a S ) /2
(51) Cai, S.; Licoccia, S.; D’Ottavi, C.; Paolesse, R.; Nardis, S.; Bulach,
V.; Zimmer, B.; Kh; Shokhireva, T.; Walker, F. A. Inorg. Chim. Acta
2002, 339, 171–178.
(52) Zakharieva, O.; Schu¨nemann, V.; Gerdan, M.; Licoccia, S.; Cai, S.;
Walker, F. A.; Trautwein, A. X. J. Am. Chem. Soc. 2002, 124, 6636–
6648.
(53) Cai, S.; Walker, F. A.; Licoccia, S. Inorg. Chem. 2000, 39, 3466–
3478.
(54) Nardis, S.; Paolesse, R.; Licoccia, S.; Fronczek, F. R.; Vicente,
M. G. H.; Shokhireva, T. K.; Cai, S.; Walker, F. A. Inorg. Chem.
2005, 44, 7030–7046.
(48) Chang, C. J.; Yeh, C. Y.; Nocera, D. G. J. Org. Chem. 2002, 67,
1403–1406.
(49) Rosenthal, J.; Chng, L. L.; Fried, S. D.; Nocera, D. G. Chem. Commun.
2007, 2642–2644.
(55) Walker, F. A.; Licoccia, S.; Paolesse, R. J. Inorg. Biochem. 2006,
100, 810–837.
(50) Dogutan, D. K.; Bediako, D. K.; Teets, T. S.; Schwalbe, M.; Nocera,
D. G. Org. Lett. 2010, 12, 1036–1039.
(56) Ye, S.; Tuttle, T.; Bill, E.; Simkhovich, L.; Gross, Z.; Thiel, W.; Neese,
F. Chem.sEur. J. 2008, 14, 10839–10851.
9
J. AM. CHEM. SOC. VOL. 133, NO. 1, 2011 137

A R T I C L E S
Dogutan et al.
1
the presence of multiple S ) /2 species as it would be expected
for a signal arising from a mixture of cobalt oxygenated samples;
(iii) the O2 adducts are difficult to form in the absence of a reductant
(vide infra)sbubbling O2(g) through solutions of 1-Co for >20
min does not induce changes in the UV-vis and EPR spectra;
and finally (v) oxygenated products of 1-CoCl, [1-Co·O2] (vide
infra), display a signal that is distinctly different from that observed
for 1-CoCl, for example, the observed hyperfine splitting pattern
spans 31 mT for 1-CoCl and 46 mT for [1-Co·O2].
The electronic ground state predicted by DFT for the
geometry optimized structure of 1-CoCl is consistent with a
redox noninnocent corrole ligand. The ground state is best
described as a broken symmetry state with two spin-up, alpha
electrons, localized on the central Co(III) ion and one spin-
down, beta electron, delocalized on the one-electron oxidized
corrole ligand. This state is similar with that determined for
the analogous iron-corrole complex, [FeCl(tpfc)], for which the
total S ) 1 ground state is the result of an antiferromagnetic
Figure 3. EPR spectrum recorded for 1-CoCl (black) in THF at 18 K,
power 0.2 mW, frequency 9.617 GHz, modulation amplitude 10 G, overlaid
with a simulated spectrum (gray) generated using g1 ) 2.073, g2 ) 2.000,
g3 ) 1.977 (σg1 ) 0.0049, σg2 ) 0.0028, σg3 ) 0.0029), ACo1 ) 58.51
MHz, ACo2 ) 40.23 MHz, ACo3 ) 29.28 MHz and a Lorentzian line width
of 0.0018 mT.
species. Inspection of these spectra reveals the presence, for
the low field range, of a clearly defined eight line pattern
originating from the hyperfine coupling of the S ) 1/2 electronic
spin to that of the I ) 7/2 nuclear spin of the 59Co central metal
ion. The intricate pattern of the signal for H > 340 mT arises
from the overlap of the hyperfine splittings generated by the
remaining two components of the hyperfine coupling tensor ACo.
The two A tensor components affecting the high and low field
regions of the EPR signal can be estimated by evaluating the
peak-to-peak separation of the eight line patterns while the
remaining component is best determined from spectral simula-
tions. Least square fitting of these spectra using a typical S )
1/2 spin-Hamiltonian, eq 1, yielded our best estimates for g1 )
2.073, g2 ) 2.000, g3 ) 1.977 and A1 ) 59.0 ( 1.0 MHz, A2
) 40.0 ( 1.0 MHz, A3 ) 29.5 ( 0.5 MHz. Attempts to
determine the superhyperfine couplings of the coordinating
nitrogen atoms were not conclusive but even if present these
couplings would not exceed 3 MHz, a value consistent with
the 1.8 - 2.1 MHz range found for [(OEC)Co(C6H5)], vide infra.
Moreover, the EPR spectra recorded for 1-CoCl and
(TBP)8CzCo(CCSiPh3) are rather similar albeit, with a larger
hyperfine splitting pattern observed for the former.30
3
coupling between the S ) /2 of the intermediate-spin Fe(III)
56
1
site and the S ) /2 of the oxidized corrole ligand. When
compared to iron, cobalt complexes are typically found to have
a low-spin configuration due to their higher effective nuclear
charge, leading to the formation of stronger bonds. However, a
quantitative computational investigation of a four-coordinate
Co(III)-corrole model, using DFT, predicted a triplet ground
state (S ) 1), ∼2100 cm-1 lower in energy than the first excited
singlet (S ) 0) state.57 Indeed using NMR and SQUID
measurements (Figure S2f, Supporting Information), we observe
a paramagnetic ground state for 1-Co in agreement with the
theoretical prediction. Moreover, inspection of the magnetic
susceptibility data suggests that this complex exhibits a triplet
ground state with a large and positive zero field splitting.
Considering that for 1-CoCl, the chloride ion coordinated in
the apical position is a weak field ligand, we anticipate the
Co(III) ion in this complex to have an intermediate-spin
electronic configuration in agreement with the DFT predicted
ground state. The predicted values for the magnitudes of the
A-tensor components, 370, 108, and 225 MHz, are indicative
of two unpaired alpha electrons localized on the Co(III) ion.
However, these A values are in stark disagreement with the
experimentally observed values. Moreover, the components of
both g and A tensors determined for 1-CoCl are rather similar
to those observed for [(OEC)Co(C6H5)]58 (g ) 1.967, 2.112,
2.004; ACo ) 72, 8, and 10 MHz), a complex that has a single
unpaired alpha electron, which is found mainly on the ligand
with ∼70% of the spin density delocalized on the corrole ligand
and the remaining ∼30% localized on the cobalt ion. For this
complex the stabilization of the low-spin electronic configuration
for the Co(III) ion is easily rationalized on the basis of the strong
σ-donor ligand coordinated in the axial position. The predicted
Co-Cl distance for 1-CoCl of 2.353 Å is found to be 0.11 Å
longer than the experimental value of 2.243 Å; this difference
is much larger than the 0.04 Å error typically associated with
the functional/basis set employed here. An excited state corre-
sponding to a low-spin configuration for the Co(III) ion can be
obtained by shortening the Co-Cl distance. By scanning
adiabatically along the Co-Cl bond distance we find an energy
minimum for this state at 2.233 Å for which the predicted values
for the |A|-tensor components of 273, 72, and 33 MHz are
ˆ
b
b
b ˜k
b
S · A · I
k
H ) ꢀS · g˜ · B +
(1)
∑
k
In addition to [(TBP)8CzCo(CCSiPh3)], we note that the EPR
spectrum of 1-CoCl also shows a similarity to cobalt(III)-
superoxide complexes. We believe that these similarities are due
to the analogous electron distributions of the complexessnamely
the presence of an unpaired electron delocalized over a noninnocent
ligand coordinated to a low-spin cobalt(III) metal ion with the
distinction that the unpaired electron for 1-CoCl and
[(TBP)8CzCo(CCSiPh3)] is found mainly in the equatorial plane
whereas the unpaired electron is delocalized over the axially
coordinated dioxygen moiety of cobalt superoxide complexes.
Consequently, for both cases the observed EPR spectra are
characterized by g-values that are nearly 2 indicating a lack of
spin-orbit interactions and small cobalt hyperfine coupling con-
stants consistent with the assertion that only a small fraction of
the spin density is localized over the metal ion. Other observations
that do not support the assignment of the EPR spectrum of Figure
3 to an oxygenated cobalt adduct includes: (i) quantification of
the observed signal, using a copper spin standard, accounts for the
majority of the prepared 1-CoCl sample, (ii) power saturation and
the temperature dependence behaviors of the signal do not reveal
(57) Rovira, C.; Kunc, K.; Hutter, J.; Parrinello, M. Inorg. Chem. 2001,
40, 11–17.
(58) Harmer, J.; Van Doorslaer, S.; Gromov, J.; Bro¨ring, M.; Jeschke, G.;
Schweiger, A. J. Phys. Chem. B 2002, 106, 2801–2811.
9
138 J. AM. CHEM. SOC. VOL. 133, NO. 1, 2011

Hangman Corroles
A R T I C L E S
Table 2. Oxygen Reduction by Hangman Corroles Xanthene in
Acidic Solution
b
c
d
compound
E1/2 (V vs RHE)
navg
% H2Oavg
1-CoCl
0.56
0.54
0.58
0.52
0.56
2.7
2.9
2.6
2.5
2.8
40
55
30
25
41
CoHCX1-CO2H
CoHCX1-CO2Me
CoHCX2-CO2H
CoHCX3-CO2H
a Determined from data shown in Figures S15 to S19 (Supporting
b
Information). E1/2, calculated as the potential at which the current
equals half the limiting current at 100 rpm. c n ) number of electrons as
calculated from RRDE data according to 4Id/(Id + (Ir/N)). d %H2O
produced during ORR as calculated from RRDE data according to (2Ir/
N)/(Id + (Ir/N)) where Ir ) ring current, Id ) disk current, and N )
collection efficiency.
Figure 4. Relative energies of two states of 1-CoCl for which the Co(III) ion
has an intermediate (blue) and low (red) spin electronic configuration; the
corresponding surface plots of their spin densities are shown above each curve.
somewhat closer to those experimentally observed. For this
excited state we find a single unpaired electron that is delocal-
ized mainly on the corrole ligand (see Figure 4), with a
distribution of spin density that is similar to that predicted for
[(OEC)Co(C6H5)]. Regardless of the electronic configuration of
the central cobalt ion, for both states, intermediate and low-
spin configurations, we find that the corrole ligand is one-
electron oxidized and that the formal Co(IV) center is better
described as a Co(III) ion residing within the monocationic
radical corrole macrocycle.
Figure 5. RRDE of edge-plane-graphite disk (black), CoHCX1-CO2H on
EPG disk (red), and CoHCX1-CO2H on MWCNTs (blue) in 0.5 M H2SO4
at 100 rpm. Disk current (negative) and Pt ring current (positive). Scan
rate 20 mV/s.
Oxygen Reactivity. The hangman effect on the oxygen
reduction reaction (ORR) of the Co-HCX series was investi-
gated. Cobalt corroles have been shown to reduce oxygen but
the ORR primarily proceeds by the less desirable two-electron
energy conversion process to produce hydrogen peroxide; the
more desirable four-electron process to produce water is not
observed.19,59-62 Here, ORR activity measurements were made
on Co-HCX compounds that were supported on multiwall
carbon nanotubes (MWCNTs), which provide a high surface
area support for catalysts.63,64 Table 2 summarizes the electro-
catalytic activity as determined from RDE and RRDE measure-
ments of each of the Co-HCX catalysts supported on MWCNT
thin film electrodes in acidic solution. The onset for oxygen
reduction occurs between 650-700 mV for all of the compounds
investigated and the half wave potential for oxygen reduction,
E1/2, calculated as the potential at which the current equals half
the limiting current at 100 rpm, is ∼550 mV vs RHE.
suggesting reduction to hydrogen peroxide. However, a more
precise measure of the ratio between hydrogen peroxide and
water may be made based on the RRDE data (Table 2), which
directly measures the amount of peroxide generated. The parent
complex 1-CoCl reduces oxygen with lower selectivity than
the hangman corrole congeners. ORR by 1-CoCl is primarily
to hydrogen peroxide consistent with previous studies in acidic
and pH 7 solution where peroxide was also observed as the
major product for this molecule.19,61 The hangman effect is best
demonstrated by comparison of RRDE data for ORR activity
between CoHCX1-CO2Me and CoHCX1-CO2H; in the ab-
sence of a proton on the hanging group, selective reduction of
oxygen to water drops by 25%. This result highlights the
importance that the hanging group be able to transfer a proton
in order to promote O-O bond cleavage.
To ensure that immobilization on the MWCNT film did not
affect the ORR mechanism of the corrole ORR catalysis, we
undertook ORR measurements on CoHCX1-CO2H immobilized
on EPG and MWCNT electrodes. Figure 5 shows the ORR
activity at 100 rpm for EPG, CoHCX1-CO2H on EPG, and
CoHCX1-CO2H supported on MWCNTs. Under similar ex-
perimental conditions, the percent water formed for CoHCX1-
CO2H films on MWCNT and on EPG is the same. However,
ORR activity of the corrole catalyst on MWCNT occurs at a
lower onset potential with higher current density. This behavior
has been seen previously for porphyrins9 and is consistent with
a higher loading of catalyst per geometric surface area afforded
by the MWCNT support.
In addition to the potentials at which catalysis occurs, the
relative production of water (4 electron reduction) over hydrogen
peroxide (2 electron reduction) is also an important factor in
assessing activity. The number of electrons transferred during
reaction may be extracted from the slope of the Koutecky-Levich
plots. For each complex studied the number of electrons
transferred is approximately 2 within experimental error, thus
(59) Kadish, K. M.; Fremond, L.; Ou, Z. P.; Shao, J. G.; Shi, C. N.; Anson,
F. C.; Burdet, F.; Gros, C. P.; Barbe, J. M.; Guilard, R. J. Am. Chem.
Soc. 2005, 127, 5625–5631.
(60) Kadish, K. M.; Fre´mond, L.; Burdet, F.; Barbe, J.-M.; Gros, C. P.;
Guilard, R. J. Inorg. Biochem. 2006, 100, 858–868.
(61) Collman, J. P.; Kaplun, M.; Decreau, R. A. Dalton Trans. 2006, 554–
559.
Activity is highest for each complex at the potential correspond-
ing to the E1/2. Similar to results for asymmetrically substituted
corroles without a hanging group, the E1/2 does not shift signifi-
cantly among the CoHCX series, indicating that the substitution
pattern does not play a significant role in shifting the activity of
the molecules.19 Previous results with porphyrin macrocycles have
indicated that the activity is improved when the macrocycle is
(62) Givaja, G.; Volpe, M.; Edwards, M. A.; Blake, A. J.; Wilson, C.;
Schroder, M.; Love, J. B. Angew. Chem., Int. Ed. 2007, 46, 584–586.
(63) Choi, A.; Jeong, H.; Kim, S.; Jo, S.; Jeon, S. Electrochim. Acta 2008,
53, 2579–2584.
(64) Qu, J.; Shen, Y.; Qu, X.; Dong, S. Electroanalysis 2004, 16, 1444–
1450.
9
J. AM. CHEM. SOC. VOL. 133, NO. 1, 2011 139

A R T I C L E S
Dogutan et al.
for the dioxygen-[(corrolazine)Co(py)] adduct and of other
mononuclear cobalt(II)-O2 complexes.30,65-68
Overall, the CoHCX series operates under a mixed 2-electron/
4-electron mechanism. The activity of the series appears to be
dominated by the electronic properties of the macrocycle core
and peripheral substitution offers more modest changes in
reactivity as compared to porphyrin analogues. The installation
of a proton-donating moiety over the face of the macrocycle
results in enhanced ORR activity for CoHCX corrole complexes,
promoting the four-electron pathway.
Conclusion
Hangman cobalt corroles are a new ligand scaffold for PCET
catalysis that are synthesized in good yields from a one-pot
condensation of dipyrromethane with the aldehyde of a xanthene
spacer followed by metal insertion using microwave irradiation.
Spectroscopic investigation of the cobalt corroles axially ligated
by chloride using X-band EPR support the description of a
[Co(III)Cl-corrole+•] in which the corrole ligand is redox
noninnocent. These higher oxidation state species are important
in small molecule activation reactions involving O-O bond
cleavage.69 Along these lines, the hangman corroles promote
ORR. The hangman corroles are distinguished from their
unmodified counterparts by the presence of an intramolecular
proton transfer network, which is able to couple to redox
processes occurring at the corrole platform. We show here for
ORR that the proton-donating hanging group works in concert
with the corrole platform to activate the O-O bond.
Figure 6. CW X-band EPR spectrum of [1-Co•O2] recorded at 20 K,
microwave power 2 mW, modulation amplitude 10 G, frequency 9.631 GHz
(black) and simulation (gray) obtained for an S ) 1/2 using g1 ) 2.113, g2,
) 2.017, g3 ) 1.979, with g-strain such that σg1 ) 0.007, σg2 ) 0.004, σg2
) 0.006; an A tensor such that A1 ) 93.2 MHz, A2 ) 29.1 MHz, A3 ) 35.7
MHz and Euler angles that account for the rotation of the A vs g tensor R
) 12°, ꢀ ) 9° and γ ) 0°.
electron deficient, and within the CoHCX series, the highest activity
is indeed observed for CoHCX1-CO2H.
Insight into the mechanism for ORR comes from a compari-
son of the EPR signals from the Co corrole platform in the
presence and absence of oxygen under reducing conditions.
Although the NMR spectra and SQUID measurement (see
Figure S2f, Supporting Information) indicate that 1-Co has a
paramagnetic triplet ground state, since it is an integer spin,
not a Kramers system, we do not expect to observe an EPR
signal in transversal mode at X-band. Indeed, spectra recorded
at cryogenic temperatures from 2 to 100 K for a frozen THF
solution sample of 1-Co show only the presence of a hetero-
geneous, axial, S ) 1/2 signal that accounts for less than 1% of
the total cobalt amount in the sample (Figure S2g, Supporting
Information). Addition of an ethanol solution containing excess
NaBH4 to this 1-Co sample does not result in the appearance
of an EPR signal that could be associated with a homogeneous
Co(II) species (Figure S2g, Supporting Information). However,
electronic absorption spectra clearly show that the cobalt center
is reduced. The forest green solution turns yellow green upon
the addition of NaBH4; the Soret and Q-bands decrease in
intensity with the concomitant appearance of a new band at 990
nm (Figure S2h, Supporting Information). The reduction product
does not change even with the addition of up to a 10 equivalent
excess of reducing agent. Exposure of this solution to O2(g)
results in the restoration of the Soret and Q-band intensities
that are characteristic of an oxidized cobalt center. Fast freezing
of O2(g) exposed solutions of the reduced cobalt corrole
Acknowledgment. We acknowledge Lee-Ping Wang for useful
discussions on computational chemistry and Dr. Danna Freedman
for collecting the magnetic susceptibility data of 1-CoCl. The
developments of new synthetic methods and synthesis of the new
compounds were performed under the sole sponsorship of Eni S.p.A
under the Eni-MIT Alliance Solar Frontiers. EPR spectra were
recorded at Carnegie Mellon University, in Prof. Eckard Mu¨nck
laboratory; this research has been supported by the U.S. NIH
(EB001475 to E.M.). S.A.S. thanks Prof. Emile Bominaar for useful
discussions. Program and characterization work was executed under
the Division of Chemical Sciences, Geosciences, and Biosciences,
Office of Basic Energy Sciences of the U.S. Department of Energy
through Grant DE-FG02-05ER15745. M.S. thanks the German
Academic Exchange Service (DAAD) for fellowship support within
the Postdoc-Programme. T.S.T. acknowledges the Fannie and John
Hertz Foundation for a graduate research fellowship.
Supporting Information Available: Full experimental details
for synthesis, NMR, MALDI-MS and high resolution ESI-MS
spectral data of corrole compounds, crystallographic data,
electronic spectra, CVs, computational details, and ORR RDE
and RRDE traces. This material is available free of charge via
the Internet at http://pubs.acs.org.
macrocycle leads to the EPR spectrum displayed in Figure 6,
1
which exhibits an intense S )
/ rhombic signal with an
2
apparent 59Co hyperfine splitting. This finding indicates that in
the presence of a strong reductant such as NaBH4, 1-Co binds
O2 even though no stabilizing base, for example, pyridine, is
present. This signal was investigated at temperatures from 2 to
80 K. In addition to the g ≈ 2 signal associated with the
[1-Co•O2] adduct under various experimental conditions a very
broad signal with resonances at 4.6, 1.6, and 1.4 is observed.
The latter signal originates from a high-spin heterogeneous
Co(II) species, most likely, the result of a decomposition process.
Spin quantification of the g ≈ 2 signal using a copper standard
and spectral simulations indicate that this species accounts for
∼80% of the total metal content. The spectra recorded for
[1-Co•O2] are rather similar in appearance to those observed
JA108904S
(65) Hoffman, B. M.; Diemente, D. L.; Basolo, F. J. Am. Chem. Soc. 1970,
92, 61–65.
(66) Walker, F. A. J. Am. Chem. Soc. 1970, 92, 4235–4244.
(67) Van Doorslaer, S.; Schweifer, A. J. Phys. Chem. B. 2000, 104, 2919–
2927.
(68) Jo¨rin, E.; Schweiger, A.; Gu¨nthard, Hs. H. Chem. Phys. Lett. 1979,
61, 228–232.
(69) Kim, S. H.; Park, H.; Seo, M. S.; Kubo, M.; Ogura, T.; Klajn, J.;
Gryko, D. T.; Valentine, J. S.; Nam, W. J. Am. Chem. Soc. 2010,
132, 14030–14032.
9
140 J. AM. CHEM. SOC. VOL. 133, NO. 1, 2011