A fully delocalized mixed-valence bis-μ(Thiolato) dicopper complex: A structural and functional model of the biological CuA center

Article abstract of DOI:10.1002/anie.201100605

A {Cu₂S₂} diamond core has been stabilized thanks to the synthesis of a novel dinucleating ligand. The resulting dicopper complex has been isolated in the two formal oxidation states (see picture) that mimic most of the essential structural, spectroscopic, and functional properties of the Cu_A center. Copyright

Full text of DOI:10.1002/anie.201100605

Communications
DOI: 10.1002/anie.201100605
Bioinorganic Chemistry
A Fully Delocalized Mixed-Valence Bis-m(Thiolato) Dicopper
Complex: A Structural and Functional Model of the Biological CuA
Center**
Marcello Gennari, Jacques Pꢀcaut, Serena DeBeer, Frank Neese, Marie-Noꢁlle Collomb, and
Carole Duboc*
The biological CuA site, a fully delocalized mixed-valence
bis(m-thiolato) dicopper complex, is implicated in the respi-
ratory chain of all eukaryotes and some prokaryotes. It is
found in cytochrome c oxidase (COX)[1] and nitrous oxidase
reductase (N2OR),[2] where it acts as a very efficient electron-
transfer agent,[3] cycling between the Cu1.5Cu1.5 and CuICuI
states.[4] In the CuA center, the two copper ions bridged by two
thiolate sulfur atoms of two cysteine residues form an almost-
planar {Cu2S2} diamond core with a short Cu···Cu distance of
2.4–2.6 ꢀ.[5–8] Each copper is also bound to one histidine (Nd)
and a weakly coordinated axial ligand, which is either a
methionine or a backbone carbonyl group (Figure 1a). In its
oxidized state, CuA is a completely delocalized mixed-valence
class III Cu1.5Cu1.5 system,[9,10] with a direct Cu Cu bond,
ꢀ
[*] Dr. M. Gennari, Dr. M.-N. Collomb, Dr. C. Duboc
Universitꢀ Joseph Fourier Grenoble 1/CNRS
Dꢀpartement de Chimie Molꢀculaire, UMR-5250
Laboratoire de Chimie Inorganique Redox
Institut de Chimie Molꢀculaire de Grenoble FR-CNRS-2607
BP-53, 38041 Grenoble Cedex 9 (France)
E-mail: carole.duboc@ujf-grenoble.fr
Dr. J. Pꢀcaut
Laboratoire de Reconnaissance Ionique et Chimie de Coordination
Service de Chimie Inorganique et Biologique
(UMR E-3 CEA/UJF, FRE3200 CNRS), CEA-Grenoble, INAC
17 rue des Martyrs 38054 Grenoble cedex 9 (France)
Dr. S. DeBeer
Department of Chemistry and Chemical Biology, Cornell University
Ithaca, NY 14853 (USA)
Figure 1. a) Structure of the CuA center of COX issued from Thermus
thermophilus in the mixed-valence state;[5] b) the model complex
[Cu2(iPrdacoS)2]+.[13]
Prof. F. Neese
Institut fꢁr Physikalische und Theoretische Chemie
University of Bonn
which was the first metal–metal bond observed in nature.[11]
Apart a slight elongation of the ligand bonds at copper and
Wegelerstrasse 12, 53115 Bonn (Germany)
and
Max-Planck-Institut fꢁr Bioanorganische Chemie
Stiftstrasse 34-36, 45470 Mꢁlheim an der Ruhr (Germany)
the Cu···Cu distance, any change in the copper ligand
arrangement has been observed in the reduced state.[7] The
[**] M.G. thanks the CNRS for a post-doctoral fellowship. S.D. thanks
exceptional structure and properties of the CuA site have
the Department of Chemistry and Chemical Biology at Cornell
motivated considerable efforts to understand the structure–
function relationship in this highly unusual bimetallic center.
With peptide ligands, an important number of synthetic
models have been investigated.[11,12] In contrast, synthesizing a
model with conventional organic ligands is far from trivial and
remains a real challenge. A major synthetic problem is that
such metal–thiolate complexes are generally prone to metal-
reduction processes with concomitant formation of disulfides
or, in the presence of dioxygen, to oxygenation at sulfur. To
the best of our knowledge, a single example of a fully
University for generous financial support. Portions of this research
were carried out at the Stanford Synchrotron Radiation Lightsource,
a national user facility operated by Stanford University on behalf of
the DOE, BES. The SSRL SMB Program is supported by the DOE,
the BER, the NIH, and the NCRR. F.N. gratefully acknowledges the
University of Bonn and the Max-Planck-Gesellschaft (by a Max
Planck fellowship) for financial support of this work. Furthermore,
the special research unit SFB 813 (Chemistry at Spin Centers,
University of Bonn) is gratefully acknowledged.
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.201100605.
5662
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2011, 50, 5662 –5666

delocalized mixed valence bis(m-thiolato) dicopper complex,
[Cu2(iPrdacoS)2]+ (iPrdacoS = 1-isopropyl-5-ethylthiolato-
1,5-diazacyclooctane), has been isolated to date by Tolman
and co-workers (Figure 1b).[13] However, this complex cannot
be reduced to the CuICuI redox state and it does not
reproduce the complete set of structural and spectroscopic
features of CuA. The other class III mixed-valence dicopper
species are still rare and have been obtained by using either
octaazacryptands[14,15] or different kinds of bridges between
the two copper ions.[16–19] However, the lack of thiolate bridges
leads to drastically different electronic properties than those
of CuA. With regard to the reduced state, to date, the only
described synthetic dinuclear CuI complex with a {Cu2S2} core
is
[Cu(SC6H4CH3-o)(phen)]2
(phen = 1,10-phenanthro-
line).[20] Nevertheless, the aromatic thiolate ligands poorly
mimic the cysteine-bridged ligands in CuA and its electro-
chemical properties have not been reported.
We report herein the synthesis and characterization of a
new bis(m-thiolato) dicopper complex that mimics most of the
essential structural, spectroscopic, and functional properties
of the CuA center. This complex has been obtained in the two
formal oxidation states, [Cu21.5,1.5L’]+ and [Cu2I,IL’], thanks to
the dinucleating character of the ligand L’ (Scheme 1). The
Figure 2. Different representations of complexes 1 and 2: a) ORTEP
views with ellipsoids set at 30% probability (hydrogen atoms and
anion omitted for clarity). The same enantiomer of both complexes is
depicted, and only one crystallographically independent complex of
[Cu2L’]+ is shown. b) Molecular views (hydrogen atoms, phenyl groups,
and anion omitted for clarity) to enlighten the major conformational
modifications between the two redox states involving the methylene
linker (opposite orientations) and the {Cu2S2} core (more flattened in
the case of [Cu2L’]+). c) Representation with principal distances and
angles of the {Cu2S2} cores.
Scheme 1. Synthesis of complex 1.
structures of both complexes were resolved by X-ray dif-
fraction and investigated by EXAFS. The spectroscopic
properties of the oxidized state were further studied by Cu
K-edge XAS, CW X-band EPR optical and absorption
spectroscopy, and DFT calculations. More importantly, an
electrochemical investigation demonstrated the reversibility
of the Cu1.5Cu1.5/CuICuI redox couple, thus making this system
a functional model of the CuA site.
The L2ꢀ ligand (2,2’-(2,2’-bipyridine-6,6’-diyl)bis(1,1-
diphenylethanethiolate)), in the presence of CuCl, reacts
with CH2Cl2 under an inert atmosphere to form the green
[Cu2I,IL’] (1) complex (Scheme 1). The in situ synthesis of L’
results from the formation of a methylene linker between two
[CuL]ꢀ units by the reaction of one thiolate from each L
ligand with CH2Cl2 (see the Supporting Information). Such
reactivity has been reported with other thiolate ligands in
nickel and ruthenium complexes,[21,22] but not with L in the
cases of the mononuclear FeIII [23] and NiII [24] complexes. The
chemical oxidation of 1 by a mild oxidizing agent, such as
4-bromobenzenediazonium tetrafluoroborate (1 equiv), leads
to the formation of a purple-colored one-electron oxidized
form, which crystallizes in the presence of tetrakis(penta-
fluorophenyl)borate as [Cu21.5,1.5L’](B(C6F5)4) (2).
which is slightly more bent in 1 than in 2 (153.1 and 173.08).
The presence of a unique counteranion per complex in 2
confirms that it is the one-electron oxidized form of 1, and the
similarity of the structural properties of both copper ions in 2
is in agreement with a delocalized mixed valence Cu1.5Cu1.5
form. The most remarkable difference between 1 and 2 is the
Cu···Cu distance, which unexpectedly increases in the oxi-
dized form, from 2.6378(14) to 2.9349(11) ꢀ (as also verified
by EXAFS; see below). A twist of the ligand around the
methylene linker occurs in 2, accompanied by an increase of
the angle between the two CuN2 planes (66 vs. 568 in 2 and 1,
respectively; Figure 2b). Similarly, the expansion of the
Cu···Cu distance leads to an increase of the Cu-S-Cu angles
in 2 (70 vs. 818) and a decrease of the S-Cu-S angles (108 and
998). In both structures, each copper ion is in the center of a
distorted trigonal pyramid with the trigonal plane defined
with the equatorial N2(4) atom (Neq) and the two bridging
The X-ray structures of both 1 and 2 are very similar
(Figure 2) and contain a quasi planar {Cu2S2} diamond core,
Angew. Chem. Int. Ed. 2011, 50, 5662 –5666
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
5663

Communications
aliphatic thiolate sulfur atoms of L’, from which Cu1(2) is only
slightly displaced (less than 0.49 ꢀ). The N1(3) atom (Nax)
ꢀ
acts as axial ligand. The {Cu2S2} core of 2 presents Cu S/Neq
bond lengths that are slightly shorter than in 1, which is
consistent with an increase of the oxidation level of the
copper ion (4 and 2 pm, respectively). This is in excellent
agreement with EXAFS data reported on the native CuA
ꢀ
center, which indicate an average decrease of the Cu Scys and
ꢀ
Cu Nhis distances between the mixed valence and the reduced
states of 4 and 1 pm, respectively.[7]
The normalized Cu K-edge data for 1 and 2 are compared
in Figure 3a. Upon oxidation, the edge shifts up by about
1 eV in energy. A weak 1s to 3d pre-edge feature is present in
2, which is absent in 1. This is consistent with the presence of
d-hole character in the mixed valent species, and confirmed
by the X-band EPR spectrum of 2, which exhibits an axial
signal with discernible seven-line copper hyperfine lines in
both the g components (Figure 3b). The seven hyperfine lines
are associated with the coupling of the unpaired electron (S =
1
= ) to the two equivalent copper ions (I = 3/2).[9] The
2
observation that gk > g? > 2.0023 together with a seven-line
hyperfine pattern shows that the unpaired electron is
delocalized over both copper centers and that the singly
occupied molecular orbital (SOMO) contains contributions
from the local Cu dx2ꢀy2 orbitals. The temperature-independ-
ent behavior of the EPR spectrum between 5 and 100 K
confirms the fully delocalized class III nature of 2, and the
extremely small hyperfine splitting of 118 MHz on the gk
signal indicates a high degree of delocalization of the electron
onto the ligands, as previously discussed for CuA (117Ak <
123 MHz).[9] As shown in Figure 3b, the EPR properties of 2
closely resemble those obtained for the CuA center.
The absorption spectrum of 2 is shown in Figure 3c
together with those of the CuA mixed-valence state and of
Tolmanꢁs model complex [Cu2(iPrdacoS2)2]+. In contrast to
CuA, both synthetic models display the same intense tran-
sition in the near-infrared region (lmax = 1622 and 1466 nm in
2 and [Cu2(iPrdacoS2)2]+, respectively).[13] This feature has
been assigned to the mixed-valence y!y* transition asso-
ciated with spin delocalization between the two copper
ions.[25] In fact, in both complexes, the distance between the
two copper ions is too large for significant direct copper–
copper orbital overlap to occur and leads to a drastic decrease
of the y!y* transition energy relative to CuA (lmax = 780–
810 nm).[11]
As expected, the SOMO of 2 is completely delocalized
over the two copper ions and the bridging sulfur atoms of the
{Cu2S2} diamond core (Figure 4). The combined copper spin
population (Mulliken analysis) amounts to 48%, while the
sulfur ligands carry as much as 39% of the spin. These
numbers, while not representing physical observables, are
close to what has been interpreted for the native CuA site
(48% on Cu vs. 44% on S).[26] Noticeable spin density is also
found on the N atoms (5% in 2 and CuA). The SOMO analysis
shows that the pu ground state of 2 is of a similar nature as in
[Cu2(iPrdacoS2)2]+, which is consistent with a long Cu···Cu
distance. By contrast, the shorter Cu···Cu distance in CuA
leads to a ground state with the su* orbital that is singly
occupied.[27]
Figure 3. a) Comparison of the normalized Cu K-edge data for 1 (d)
and 2 (c); b) X-band EPR spectrum of 2 recorded in CH2Cl2 at 5 K.
Parameters used for the simulation: gk =2.215, g? =2.002,
Ak =118 MHz, A? =48 MHz. Inset: EPR spectrum of CuA of N2OR
from P. stutzeri;[34] c) absorption spectra of 2 (c), [Cu2(iPrdacoS2)2]+
(g),[13] and CuA of COX from B. subtilis (d).[35]
The electrochemical behavior of complexes 1 and 2 was
investigated. The cyclic voltammogram (CV) of 1, recorded in
CH2Cl2 under argon, displays two successive one-electron
oxidation processes (Figure 5a): a reversible Cu1.5Cu1.5/
CuICuI wave at E1/2 = ꢀ0.79 V vs. Fc+/Fc (DEp = 80 mV) and
a partially reversible wave at E1/2 = + 0.05 V (DEp = 100 mV)
that is assigned to the CuIICuII/Cu1.5Cu1.5 redox couple.
5664 www.angewandte.org
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2011, 50, 5662 –5666

core (ꢀ0.920 V vs. Fc+/Fc, CH2Cl2).[19] The electron self-
exchange rate constant was measured for the 2/1 redox couple
by means of electrochemical experiments (kel = 7.7(1) ꢂ
10ꢀ3 cmsꢀ1,
or
estimated
as
khom = 1.37(2) ꢂ
106 Lmolꢀ1 sꢀ1).[31] This khom value indicates that our model
can achieve electron transfer rates faster than mononuclear
CuII complexes, including plastocyanin (khom = 1.5 ꢂ
105 L molꢀ1 sꢀ1),[32] but more slowly than CuA (khom = 2.2 ꢂ
108 L molꢀ1 sꢀ1),[33] which is certainly due to the lack of the
ꢀ
Cu Cu bond in 2.
A bulk electrolysis experiment at + 0.64 V of a electro-
generated solution of 2 demonstrates that the CuIICuII form is
not stable. This is in parallel with the redox behavior observed
for the CuA center for which the fully oxidized CuIICuII state
has not been reported to date.[28]
In summary, we have isolated and characterized the first
dicopper system with a {Cu2S2} core that is stable in the two
oxidation states (I,I) and (1.5,1.5) and that can be related to
the biological CuA site. However, in contrast to CuA, the short
Cu···Cu distance of the reduced complex 1 notably increases
in the oxidized mixed valence form 2, despite the geometric
constraints imposed by L’. Investigations are underway in our
laboratories to understand this unexpected elongation of the
Cu···Cu distance upon the oxidation process.
Figure 4. Contour plot of the redox-active, half-filled molecular orbital
(SOMO) of complex 2.
Received: January 24, 2011
Published online: May 9, 2011
Keywords: bioinorganic chemistry · copper ·
.
mixed-valent compounds · S ligands
[1] B. G. Malmstroem, Chem. Rev. 1990, 90, 1247 – 1260.
[2] Suharti, M. J. F. Strampraad, I. Schrꢃder, S. de Vries, Biochem-
istry 2001, 40, 2632 – 2639.
[3] G. T. Babcock, M. Wikstrom, Nature 1992, 356, 301 – 309.
[4] D. W. Randall, D. R. Gamelin, L. B. LaCroix, E. I. Solomon,
J. Biol. Inorg. Chem. 2000, 5, 16 – 29.
[5] P. A. Williams, N. J. Blackburn, D. Sanders, H. Bellamy, E. A.
Stura, J. A. Fee, D. E. McRee, Nat. Struct. Biol. 1999, 6, 509 – 516.
[6] K. Brown, M. Tegoni, M. Prudencio, A. S. Pereira, S. Besson, J. J.
Moura, I. Moura, C. Cambillau, Nat. Struct. Biol. 2000, 7, 191 –
195.
Figure 5. a) Cyclic voltammogram of 1 (0.9 mm) in CH2Cl2 with 0.1m
Bu4NPF6 at a carbon vitreous electrode and b) after electrolysis at
ꢀ0.610 V (1 electron per molecule of 1 involved), scan rate of
100 mVsꢀ1. Potentials are given versus Fc+/Fc.
[7] N. J. Blackburn, S. de Vries, M. E. Barr, R. P. Houser, W. B.
Tolman, D. Sanders, J. A. Fee, J. Am. Chem. Soc. 1997, 119,
6135 – 6143.
[8] N. J. Blackburn, M. Ralle, E. Gomez, M. G. Hill, A. Pastuszyn,
D. Sanders, J. A. Fee, Biochemistry 1999, 38, 7075 – 7084.
[9] F. Neese, W. G. Zumft, W. E. Antholine, P. M. H. Kroneck,
J. Am. Chem. Soc. 1996, 118, 8692 – 8699.
[10] J. A. Farrar, F. Neese, P. Lappalainen, P. M. H. Kroneck, M.
Saraste, W. G. Zumft, A. J. Thomson, J. Am. Chem. Soc. 1996,
118, 11501 – 11514.
[11] M. G. Savelieff, Y. Lu, J. Biol. Inorg. Chem. 2010, 15, 461 – 483.
[12] Y. Lu, Angew. Chem. 2006, 118, 5714 – 5728; Angew. Chem. Int.
Ed. 2006, 45, 5588 – 5601.
[13] R. P. Houser, V. G. Young, W. B. Tolman, J. Am. Chem. Soc.
1996, 118, 2101 – 2102.
[14] C. Harding, J. Nelson, M. C. R. Symons, J. Wyatt, J. Chem. Soc.
Chem. Commun. 1994, 2499 – 2500.
[15] M. E. Barr, P. H. Smith, W. E. Antholine, B. Spencer, J. Chem.
Soc. Chem. Commun. 1993, 1649 – 1652.
[16] C. He, S. J. Lippard, Inorg. Chem. 2000, 39, 5225 – 5231.
Exhaustive electrolysis of 1 at ꢀ0.61 V results in the transfer
of one electron and quantitatively yields 2 (Figure 5b).
Compound 1 is fully regenerated by back electrolysis at
ꢀ0.91 V, demonstrating the perfect reversibility of this one-
electron process. The accessibility of the two Cu1.5Cu1.5 and
CuICuI redox states most likely results from the preorganiza-
tion of the system owing to the dinucleating character of L’
that allows the stabilization of the {Cu2S2} diamond core. The
potential of the Cu1.5Cu1.5/CuICuI redox system of 1 is lower
than that of the CuA site (+ 0.24 V vs. NHE,[28] or ꢀ0.44 V vs.
Fc+/Fc[29]), in accordance to the weaker axial bonding
interactions in the biological site.[30] The other fully delocal-
ized mixed valence dicopper complexes present higher
oxidation potentials than 1 and CuA (ꢀ0.390 < E1/2 vs. Fc+/
Fc, CH2Cl2 < + 0.025 V),[14,16,18] except for one with a {Cu2P2}
Angew. Chem. Int. Ed. 2011, 50, 5662 –5666
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
5665

Communications
[17] R. Gupta, Z. H. Zhang, D. Powell, M. P. Hendrich, A. S.
[27] S. I. Gorelsky, X. J. Xie, Y. Chen, J. A. Fee, E. I. Solomon, J. Am.
Chem. Soc. 2006, 128, 16452 – 16453.
Borovik, Inorg. Chem. 2002, 41, 5100 – 5106.
[18] S. B. Harkins, J. C. Peters, J. Am. Chem. Soc. 2004, 126, 2885 –
2893.
[19] N. P. Mankad, E. Rivard, S. B. Harkins, J. C. Peters, J. Am. Chem.
Soc. 2005, 127, 16032 – 16033.
[28] C. Immoos, M. G. Hill, D. Sanders, J. A. Fee, C. E. Slutter, J. H.
Richards, H. B. Gray, J. Biol. Inorg. Chem. 1996, 1, 529 – 531.
[29] N. G. Connelly, W. E. Geiger, Chem. Rev. 1996, 96, 877 – 910.
[30] G. N. Ledesma, D. H. Murgida, H. K. Ly, H. Wackerbarth, J.
Ulstrup, A. J. Costa-Filho, A. J. Vila, J. Am. Chem. Soc. 2007,
129, 11884 – 11885.
[20] R. K. Chadha, R. Kumar, D. G. Tuck, Can. J. Chem. 1987, 65,
1336 – 1342.
[21] Q. Wang, A. C. Marr, A. J. Blake, C. Wilson, M. Schroder, Chem.
Commun. 2003, 2776 – 2777.
[22] C. A. Grapperhaus, S. Poturovic, M. S. Mashuta, Inorg. Chem.
2002, 41, 4309 – 4311.
[23] M. A. Kopf, D. Varech, J. P. Tuchagues, D. Mansuy, I. Artaud,
J. Chem. Soc. Dalton Trans. 1998, 991 – 998.
[24] M. Gennari, M. Orio, J. Pꢄcaut, F. Neese, M.-N. Collomb, C.
Duboc, Inorg. Chem. 2010, 49, 6399 – 6401.
[25] K. R. Williams, D. R. Gamelin, L. B. LaCroix, R. P. Houser,
W. B. Tolman, T. C. Mulder, S. deVries, B. Hedman, K. O.
Hodgson, E. I. Solomon, J. Am. Chem. Soc. 1997, 119, 613 – 614.
[26] S. DeBeer George, M. Metz, R. K. Szilagyi, H. X. Wang, S. P.
Cramer, Y. Lu, W. B. Tolman, B. Hedman, K. O. Hodgson, E. I.
Solomon, J. Am. Chem. Soc. 2001, 123, 5757 – 5767.
[31] M. J. Weaver, J. Phys. Chem. 1990, 94, 8608 – 8613.
[32] M. R. b. Jensen, D. F. Hansen, J. J. Led, J. Am. Chem. Soc. 2002,
124, 4093 – 4096.
[33] C. E. Slutter, R. Langen, D. Sanders, S. M. Lawrence, P. Wittung,
A. J. Di Bilio, M. G. Hill, J. A. Fee, J. H. Richards, J. R. Winkler,
B. G. Malmstrꢃm, Inorg. Chim. Acta 1996, 243, 141 – 145.
[34] H. S. McHaourab, S. Pfenninger, W. E. Antholine, C. C. Felix,
J. S. Hyde, P. M. H. Kroneck, Biophys. J. 1993, 64, 1576 – 1579.
[35] D. R. Gamelin, D. W. Randall, M. T. Hay, R. P. Houser, T. C.
Mulder, G. W. Canters, S. de Vries, W. B. Tolman, Y. Lu, E. I.
Solomon, J. Am. Chem. Soc. 1998, 120, 5246 – 5263.
5666 www.angewandte.org
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2011, 50, 5662 –5666