Redox chemistry of a dimanganese(II,III) complex with an unsymmetric ligand: Water binding, deprotonation and accumulative light-induced oxidation

Article abstract of DOI:10.1002/ejic.200600676

A dinuclear manganese complex {[Mn₂^II,IIIL(μ-OAc) ₂]-ClO₄} has been synthesised, where L is the dianion of 2-{[bis-(pyrid-2-ylmethyl)amino]methyl}-6-{[(3,5-di-tert-butyl-2-hydroxybenzyl) (pyrid-2-ylmethyl)amino]

Full text of DOI:10.1002/ejic.200600676

FULL PAPER
DOI: 10.1002/ejic.200600676
Redox Chemistry of a Dimanganese(II,III) Complex with an Unsymmetric
Ligand: Water Binding, Deprotonation and Accumulative Light-Induced
Oxidation
Magnus F. Anderlund,[a] Joakim Högblom,[b] Wei Shi,[a] Ping Huang,[a] Lars Eriksson,[c]
Högni Weihe,[d] Stenbjörn Styring,[a] Björn Åkermark,[e] Reiner Lomoth,*[a] and
Ann Magnuson*[a]
Keywords: Manganese / Bioinorganic chemistry / Ligand exchange / EPR spectroscopy / IR spectroscopy
A
dinuclear manganese complex {[Mn2II,IIIL(µ-OAc)2]-
of about 100 ms. At high scan rates quasireversible voltam-
metric behaviour is found for all three electrode processes,
with particularly slow electron transfer for the II,III/II,II [k°(1)
= 0.002 cms–1] and III,III/II,III [k°(2) = 0.005 cms–1] couples,
which can be rationalised in terms of major distortions of the
MnIII centres. In aqueous media the bridging acetates are re-
placed by water-derived ligands. Deprotonation of these sta-
bilises higher valence states, and photo-induced oxidation
ClO4} has been synthesised, where L is the dianion of 2-{[bis-
(pyrid-2-ylmethyl)amino]methyl}-6-{[(3,5-di-tert-butyl-2-
hydroxybenzyl)(pyrid-2-ylmethyl)amino]methyl}-4-methyl-
phenol, an unsymmetric binucleating ligand with two coordi-
nating phenol groups. The two manganese ions, with a Mn–
Mn distance of 3.498 Å, are bridged by the two bidentate
acetate ligands and the 4-methylphenolate group of the li-
gand, resulting in a N3O3 and N2O4 donor set of MnII and
MnIII, respectively. Electrochemically [Mn2II,IIIL(µ-OAc)2]+ is
reduced to [Mn2II,IIL(µ-OAc)2] at E1/2(1) = –0.53 V versus
Fc+/0 and oxidised to [Mn2III,IIIL(µ-OAc)2]2+ at E1/2(2) = 0.38 V
versus Fc+/0. All three redox states have been characterised
by EPR, IR and UV/Vis spectroscopy. Subsequent oxidation
of [Mn2II,IIIL(µ-OAc)2]2+ [E1/2(3) = 0.75 V vs. Fc+/0] in dry ace-
tonitrile results in an unstable primary product with a lifetime
of the manganese complex results in
a
Mn2III,IVL
complex with oxo or hydroxo bridging ligands, which is
further oxidised to an EPR-silent product. These results dem-
onstrate that a larger number of metal-centred oxidations can
be compressed in a narrow potential range if build up of
charge is avoided by charge-compensating reactions.
(© Wiley-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim,
Germany, 2006)
of the water-oxidising complex (WOC), a catalyst consisting
of a protein-bound cluster with four manganese ions and
one calcium ion.[3,4] After four consecutive one-electron
transfer events, the WOC retrieves four electrons from water
and releases molecular oxygen.[3,5] During catalysis, the
WOC cycles through five redox states.[6,7] The manganese
ions are mostly in the MnIII oxidation state or higher, and
several of the intermediates involve manganese-centred oxi-
dation.[7–9]
Synthetic manganese complexes have been extensively
studied as model systems for the WOC but functional mim-
ics that are able to evolve molecular oxygen from water are
very few in number.[10,11] They require electrochemical oxi-
dation at highly anodic potentials or strong chemical oxi-
dants. In the latter case the formation of the important Mn-
oxo intermediate often relies on oxygen atom transfer from
the oxidant rather than from water.[11]
Introduction
The utilisation of solar energy for fuel production,
through a molecular system for artificial photosynthesis, is
an attractive solution to the need for renewable energy.[1]
In natural photosynthesis, solar energy is used to extract
electrons from water.[2] Photo-induced charge separation in
the Photosystem II (PSII) reaction centre leads to oxidation
[a] Department of Photochemistry and Molecular Science,
Uppsala University,
P. O. Box 523, 75120 Uppsala, Sweden
E-mail: ann.magnuson@fotomol.uu.se
reiner.lomoth@fotomol.uu.se
[b] Department of Biochemistry, Department of Chemistry, Lund
University,
P. O. Box 124, 22100 Lund, Sweden
[c] Division of Structural Chemistry, Arrhenius Laboratories,
Stockholm University,
10691 Stockholm, Sweden
[d] Department of Chemistry, University of Copenhagen,
Universitetsparken 5, 2100 Copenhagen, Denmark
[e] Department of Organic Chemistry, Arrhenius Laboratories,
Stockholm University,
In our work towards artificial photosynthesis, we pursue
supramolecular complexes in which the Mn portion un-
dergoes accumulative oxidation by light-induced charge sep-
aration.[12–14] With the ultimate goal of photochemical
water splitting, this requires Mn complexes that can un-
10691 Stockholm, Sweden
Supporting information for this article is available on the
WWW under http://www.eurjic.org or from the author.
Eur. J. Inorg. Chem. 2006, 5033–5047
© 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
5033

R. Lomoth, A. Magnuson et al.
FULL PAPER
III,IV
dergo four oxidation steps within the thermodynamic limit Mn2
level, which is not the case with [Mn2(bpmp)(µ-
of the photo-oxidant. On the other hand, all stored oxi- OAc)2], while the potentials are above those of
dation equivalents need to be sufficiently oxidising, with an [Mn2(bhpp)(µ-OAc)2]+. The favourable behaviour of the
average potential of 0.8 V (at pH7), to be capable of driving novel complex can be attributed to the combination of the
the catalytic oxidation of water.[15] In the WOC all four sin- flexible bridging motif that is inherent to this type of com-
gle-electron oxidations of the Ca(Mn)4 cluster occur within plex and the intermediate O/N ratio of the chelating ligand.
an interval of only 0.25–0.3 V at potentials close to the limit These findings point to the intricate balance of donor prop-
(ca. 1 V vs. NHE[15]) given by the oxidising TyrZ radical. erties in the set of permanent and flexible ligands that will
This compression of oxidation potentials is a result of be needed for functional mimics of the WOC suitable for
charge-compensating deprotonation reactions and it is evi- photo-induced water oxidation.
dent that a suitable mechanism for charge compensation is
a central feature in the design of synthetic mimics.
In previous studies we could show that oxidation of the
Results and Discussion
manganese dimer [Mn2(bpmp)(µ-OAc)2]+[16] in the presence
of water involves exchange of the bridging acetates for
Synthesis and Characterisation
water-derived ligands that are increasingly deprotonated in
higher oxidation states.[17,18] The resulting charge compen-
Synthesis
sation stabilises higher valence states and thereby allows for
II,II
The synthesis route for the preparation of compound 5 is
described in Scheme 1. The synthesis of the unsymmetrical
ligand H2L (4) is based on the unsymmetrical phenol 1.[27]
Compound 2 was obtained by letting phenol 1 react with
1 equiv. of (3,5-di-tert-butyl-2-hydroxybenzyl)(2-pyridyl-
methyl)amine[21] and 2 equiv. of Et3N in CH2Cl2 under re-
flux. Compound 3 was then obtained by reduction of 2 with
NaBH4 in methanol and thereafter chlorinated with SOCl2
in dry CH2Cl2.
The unsymmetrical ligand H2L (4) is produced by amin-
ation by nucleophilic attack on the benzylic chloride in 3
with bis-picolylamine[28] and Et3N in CH2Cl2 under reflux.
The yield was excellent in the first two steps (95.1 and
97.5%) but only moderate in the last step (67.4%), possibly
because of steric hindrance of the two tert-butyl groups on
the phenol.
three metal-centred oxidations, from the Mn2
a Mn2
dimer to
III,IV
species, to be driven by photo-generated RuII-
I(bpy)3.[13,17,19,20] The binding of water, deprotonation of
bridging ligands and accumulative, light-induced oxidation
of [Mn2(bpmp)(µ-OAc)2]+ mimic important aspects of the
catalytic complex in PSII and we therefore set out to investi-
gate modifications of the structural motif. With consider-
ation to the prevalence of high-valent manganese species
and the high O/N ratio in the ligand sphere of the Ca(Mn)
4 complex in PSII, we synthesised analogous dinuclear man-
ganese complexes, with an increased number of phenolic
ligand functions. In [Mn2(bhpp)(µ-OAc)2]+[19,21] two ad-
ditional phenol groups of the bhpp ligand result in an oxy-
gen-to-nitrogen ratio of 4:2. Compared to [Mn2(bpmp)(µ-
OAc)2]+ with O/N = 3:3, the potentials for the manganese
redox couples are lowered substantially and the oxidation
equivalents stored on [Mn2(bhpp)(µ-OAc)2]+ are on average
not sufficiently oxidising with respect to water oxidation.
In the present study we report a new di-µ-acetato-
The complexation of H2L with MnOAc3·2H2O in etha-
nol under argon affords the stable dimanganese(II,III) com-
plex 5 after addition of NaClO4·H2O in 52% yield. We have
III,III
not been able to isolate any Mn2
complex under these
bridged dinuclear manganese complex,
5 ([Mn2L(µ-
conditions. The reduction of MnIII to MnII was possibly
done by the solvent or by disproportionation of MnIII to
MnO2 and MnII in the presence of water. Though we have
not detected any MnO2, we cannot rule out that colloidal
MnO2 was present in the solution after the complexation.
However, the slow and well-behaved crystal growth, as well
as the purity of the crystals established by elemental analy-
sis, indicates that any impurities in the mother liquor were
successfully removed prior to characterisation of 5.
OAc)2]+), with an unsymmetric ligand molecule, L = 2-
{[bis(pyrid-2-ylmethyl)amino]methyl}-6-{[(3,5-di-tert-butyl-
2-hydroxybenzyl)(pyrid-2-ylmethyl)amino]methyl}-4-meth-
ylphenol. With a N3O3/N2O4 coordination sphere, the new
complex is an intermediate between the [Mn2(bpmp)(µ-
OAc)2]+ and [Mn2(bhpp)(µ-OAc)2]+ complexes. We chose
to synthesise an unsymmetric ligand, as we anticipated that
a compromise between the need for high valent Mn states
and sufficiently high redox potentials could be ac-
complished with the intermediate O/N ratio. Here we pres-
ent the synthesis and characterisation of the new manga-
nese complex and its redox properties in anhydrous and
X-ray Structure Determination
The crystal structure of 5 (Figure 1) is of a monovalent
partially aqueous solutions, including photo-oxidation cation, counterbalanced with a perchlorate ion (not shown).
studies. The displacement parameters of the perchlorate are heavily
Unsymmetric ligands have previously been used to a lim- anisotropic, in agreement with a weakly coordinated per-
ited extent for coordinating manganese, mainly for mimick- chlorate ion, and there are no ordered solvent molecules
ing the active site of manganese catalases,[22–25] or to sup- present in the structure. The manganese ions are bridged
port the formation of heterobinuclear metal com- by the oxygen in the central 4-methylphenolate and the two
plexes.[22,26] As a novelty in the present study, photo-oxi- bidentate acetates. The slightly distorted octahedral coordi-
dation of complex 5 generates products beyond the nation of Mn(1) is completed by the two pyridyl nitrogen
5034
www.eurjic.org
© 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Inorg. Chem. 2006, 5033–5047

Redox Chemistry of a Dimanganese(II,III) Complex
FULL PAPER
Scheme 1. Reaction conditions are as follows: (a) (3,5-di-tert-butylbenzyl-2-hydroxy)(2-pyridylmethyl)amine, CH2Cl2, reflux; (b) NaBH4,
MeOH, room temperature; (c) 1. SOCl2, CH2Cl2, 2. DPA, Et3N, CH2Cl2, reflux; (d) MnOAc3·2H2O, EtOH, 60 °C, NaClO4·H2O.
Figure 1. ORTEP view (30% probability ellipsoids) of 5 with atomic numbering for all non-hydrogen atoms. The hydrogen atoms were
geometrically placed and shown as small circles of arbitrary radii. Figure drawn by Diamond.
atoms and one amine nitrogen atom of one branch of the
Selected bond lengths and angles are found in Table 1.
ligand, resulting in a N3O3 ligand sphere. The oxygen atom Bond valence calculations gave 2.1 for Mn(1) and 3.2 for
from the tert-butyl-substituted phenolate coordinates Mn(2) in agreement with the different donor sets and coor-
Mn(2) in a trans position to the bridging phenolate group. dination symmetries of the two manganese ions.[29]
The N2O4 coordination sphere of Mn(2) is completed by
the pyridyl and amine nitrogens.
Mn(2) has a more distorted octahedral geometry, consis-
tent with its assignment as MnIII. For Mn(2), the shortest
Eur. J. Inorg. Chem. 2006, 5033–5047
© 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org
5035

R. Lomoth, A. Magnuson et al.
II,III
FULL PAPER
Table 1. Selected bond lengths [Å] and angles [°] for 5. The dis- antiferromagnetically coupled Mn2
tances and angles given describe the coordination polyhedron
around each Mn and the orientation of the two polyhedrons with
respect to each other.
complex with mod-
erate exchange coupling.[16,30] At the lowest temperatures
the magnetic moment was 1.79, which compares well with
the value 1.73 expected for an S = 1/2 ground state.
Mn1···Mn2
Mn1–O2b
Mn1–O2a
Mn1–O31
Mn1–N11
Mn1–N21
Mn1–N2
3.4985(15)
2.089(5)
2.127(4)
2.179(4)
2.255(5)
2.280(5)
2.317(5)
Mn1–O31–Mn2
Mn2–O51
Mn2–O31
Mn2–O1a
Mn2–N3
Mn2–O1b
Mn2–N41
O51–Mn2–O31
O51–Mn2–O1A
O31–Mn2–O1A
O51–Mn2–N
O31–Mn2–N3
O1A–Mn2–N3
O51–Mn2–O1B
O31–Mn2–O1B
116.95(18)
1.820(4)
1.922(4)
1.972(4)
2.114(4)
2.154(5)
2.266(6)
177.1(2)
86.47(18)
92.05(16)
89.93(18)
91.96(17)
169.45(18)
89.5(2)
O2B–Mn1–O2A 99.6(2)
O2B–Mn1–O31 100.98(19)
O2A–Mn1–O31 88.28(15)
O2B–Mn1–N11 102.2(2)
O2A–Mn1–N11 86.06(17)
O31–Mn1–N11 156.78(19)
O2B–Mn1–N21 91.0(2)
O2A–Mn1–N21 169.2(2)
O31–Mn1–N21 87.68(15)
N11–Mn1–N21 93.77(18)
88.16(19)
O1A–Mn2–O1B 97.3(2)
N3–Mn2–O1B
O51–Mn2–N41
O31–Mn2–N41
92.57(19)
96.5(2)
86.05(18)
O2B–Mn1–N2
164.6(2)
O2A–Mn1–N2 94.82(18)
O31–Mn1–N2
N11–Mn1–N2
N21–Mn1–N2
84.88(16)
73.20(19)
74.9(2)
O1A–Mn2–N41 92.22(19)
Figure 2. The magnetic susceptibility (+) and a fitted curve based
on Equation (1) (solid line) in CGS units, as well as the effective
magnetic moment as a function of temperature.
N3–Mn2–N41
O1B–Mn2–N41
78.33(19)
169.02(19)
II,III
For known Mn2
complexes the exchange coupling
between the two manganese ions is rather weak, compared
to that in Mn2III,IV complexes, and typically ranges between
about 0 and 10 cm–1.[25,30,31] When analysing the suscep-
tibility data for 5, it proved impossible to satisfactorily in-
terpret the magnetic susceptibilities in terms of an isotropic
spin Hamiltonian (Model 1 in Table 2). As the zero-field
splitting parameter D of MnIII is often of the same order
of magnitude as the exchange coupling J, the Hamiltonian
was augmented with a term accounting for the magnetic
anisotropy of the MnIII ion,
Mn–O bond length is that of the tert-butyl-substituted
phenoxyl (O51), and the bridging phenoxyl provides the
second shortest Mn–O bond length. Thus, in contrast to
the coordination sphere of Mn(1), the two bridging acetate
ligands provide the longest Mn–O bonds. A similar, al-
though less enhanced, effect was also observed for the MnIII
ion in the [Mn2II,III(bpmp)(µ-OAc)2]2+ structure.[16] The li-
gand sphere of Mn(2) is therefore very compressed along
the O–Mn–O axis, which is 0.21 Å shorter than the corre-
II,III
sponding molecular axis of the MnIII ion in [Mn2
-
(bpmp)(µ-OAc)2]2+. The O–Mn–N axis formed by the pyr-
idyl nitrogen and the carboxyl oxygen trans to it is quite
long, making the Mn(2) coordination even more distorted,
consistent with the Jahn–Teller effects that are common for
MnIII complexes. The Mn–Mn distance is 3.498 Å, and the
1
2
ˆ
ˆ
ˆ
ˆ
ˆ
H = J S2·S3 + µBB·g·(S2 + S3) + D3[S
–
S(S + 1)]
(1)
3z
3
Table 2. Results from fitting magnetic susceptibility data. Models
1 and 2 refer to Equation (1) without and with the third term,
respectively.
MnII–O–MnIII angle is 116.95°, which is in the range of
II,III
other Mn2
complexes.[16,25,30]
Model 1
Model 2
In conclusion, the crystal structure shows that the ter-
minal phenoxyl in one branch of the ligand favours coordi-
nation of MnIII over MnII, and that its axial position intro-
duces an unusually large distortion away from octahedral
symmetry.
J [cm–1]
11.0Ϯ0.05
–
–5.05Ϯ1.19
4.04Ϯ1.90
10.9Ϯ0.3
3.35Ϯ0.4
–3.95Ϯ4.97
25Ϯ4.5
D3 [cm–1]
TIP (ϫ10–12
)
Curie (ϫ10–11
χ2
)
131.9
–
3
137
140
100
0.07
4
136
140
f-test
No. of fit parameters
Degrees of freedom
No. of points
Magnetic Susceptibility
The experimentally determined magnetic susceptibilities
in the temperature range 4.45–301 K are shown in Figure 2.
At room temperature, the effective magnetic moment is In Equation (1) the subscripts refer to the oxidation num-
7.20, which is slightly lower than the value 7.68 expected bers of the metal ions. By inclusion of the last term in
for a MnIIMnIII dimer in which the metals are weakly anti- Equation (1), the following parameters were obtained: J =
ferromagnetically interacting. The magnetic moment de- 10.9Ϯ0.3 cm–1 and D3 = 3.35 cm–1 (Model 2 in Table 2).
creased smoothly from 298 K to 50 K, and then more The relatively large and positive zero-field parameter D3 re-
abruptly down to 3 K. This behaviour is expected for an flects to some degree the coordination geometry of the
5036
www.eurjic.org
© 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Inorg. Chem. 2006, 5033–5047

Redox Chemistry of a Dimanganese(II,III) Complex
FULL PAPER
MnIII ion. Six-coordinated MnIII complexes with distorted
octahedral coordination typically display D values of –3
to –4 cm–1. However, deviations occur, for example, when
the ligand field has low symmetry in the equatorial plane,
as in 5, and the increased rhombicity of the ZFS parameters
can lead to a change in sign to positive values of D.[32]
IR Spectroscopy
The IR spectrum of 5 in a KBr pellet was recorded be-
tween 4000 and 400 cm–1 (Supporting information). Based
on characteristic frequencies the following assignments are
suggested in agreement with the composition of the com-
plex: 3062 (w, νC–H, aryl); 2952, 2867 (m, νC–H, alkyl);
1590 (s, νaC–O, carboxylate); 1441 (s, νsC–O, carboxylate);
1477, 1435 (s, νC–C, aryl); 1094 (s, νaCl–O, perchlorate);
623 (m, δaCl–O, perchlorate) cm–1.
The IR spectrum of 5 in [D3]acetonitrile solution (Sup-
porting Information) is identical to the solid state spectrum,
which provides strong evidence that the di-µ-acetato bridg- tials. (a) After applying –0.08 V to the sample solution. The spec-
ing structure established for crystalline 5 is maintained in
anhydrous acetonitrile solution.
Figure 3. EPR spectra recorded by exposing 5 to controlled poten-
II,III
trum arises from the S = 1/2 ground spin state of the Mn2
complex. No current was recorded at this potential. (b) After re-
II,II
duction at –0.93 V a spectrum typical of a weakly coupled Mn2
complex is observed. (c) Recovery of the Mn2
II,III
complex after
reoxidation at –0.08 V of the sample giving spectrum b. (d) After
EPR Spectroscopy
oxidation at 0.57 V. The spectrum can be identified as a remaining
II,III
fraction (5%) of Mn2
complexes, still present after bulk elec-
The EPR spectrum of 5 in anhydrous acetonitrile is
shown in Figure 3, spectrum a. The spectrum recorded at
4 K shows the Mn2
centred at g = 2 and displaying 16–20 hyperfine-lines, with
a peak split of 110–140 gauss. The mixed-valence forms
trolysis. In the centre of the spectrum, a six-line signal from mono-
meric MnII can be distinguished, corresponding to less than 2% of
the manganese. Spectrometer settings: microwave frequency,
9.6 GHz; modulation amplitude, 10 G; microwave power spectrum
a and c, 20 mW; in spectrum b, 2 mW; temperature in spectra a, c
and d, 4 K; in spectrum b, 12 K. Insert: the signal at g = 5 arising
from the excited S = 3/2 spin state of 5, observed at T Ͼ 5 K.
II,III
S = 1/2 ground-state EPR signal,
II,III
III,IV
of dinuclear manganese complexes, Mn2
and Mn2
,
are known to display similar ground-state EPR signals in
the g = 2 region at cryogenic temperatures.[33] The acetato-
II,III
Redox Properties in Anhydrous Media
bridged Mn2
complexes normally show a weaker ex-
III,IV
change coupling than µ-oxo-bridged Mn2
complexes,
Electrochemistry
because of less orbital overlap between the Mn ions. The
excited S = 3/2 and S = 5/2 states in the Mn2II,III complexes
will then give rise to transitions at low field, at relatively
low temperatures (5–20 K). Thus, when the temperature
was raised from 4 K to 5–11 K, the first excited state (S =
3/2) gave rise to a signal at g = 5.4 (Figure 3, inset).
The spectrum from complex 5 (Figure 3, a) could be sim-
ulated assuming an effective spin of S = 1/2, and using the
spin Hamiltonian
The redox behaviour of 5 in acetonitrile was studied by
cyclic voltammetry (CV), differential pulse voltammetry
(DPV) and controlled potential electrolysis. The products
of the oxidation and reduction processes were characterised
by EPR spectroscopy and UV/Vis and IR spectroelectroch-
emisty. The electrochemical data are compiled in Table 3.
All potentials are referenced versus ferrocene as redox stan-
dard and about 0.4 V should be added to these values in
order to compare them with potentials in aqueous solution
versus NHE.[36]
ˆ
ˆ
ˆ
ˆ
H = β S·g·B + Î1·A1·S + Î2·A2·S
(2)
Voltammograms of 5 in acetonitrile solution (Figure 4)
showed reversible to quasireversible CV waves for reduction
(A full account of the EPR simulation parameters at at E1/2 = –0.53 V (1) and oxidation at E1/2 = 0.38 V versus
both X- and Q-bands, as well as simulated spectra, was Fc+/Fc0 (2) while further oxidation was irreversible on the
published by Huang et al.[34]) The obtained hyperfine ten- normal CV time scale (v = 0.100 Vs–1). At higher scan rates
sors showed considerable anisotropy,[35] which could be ex- on a microelectrode, the second oxidation gave rise to a
plained by a significant Jahn–Teller distortion of MnIII
,
quasireversible CV wave with the half-wave potential E1/2 =
taking into account the unexpected positive sign of the 0.75 V (3) in good agreement with the DPV peak potential.
zero-field parameter D. These spectroscopic properties are Chemical reversibility was obtained for scan rates of v Ն
a direct result of the rhombic distortion of the MnIII coor- 10 Vs–1 and a lower limit of the lifetime of the product of
dination sphere.
the second oxidation of about 100 ms (at room tempera-
Eur. J. Inorg. Chem. 2006, 5033–5047
© 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org
5037

R. Lomoth, A. Magnuson et al.
FULL PAPER
Table 3. Electrochemical data for [Mn2L(µ-OAc)2]+ in acetonitrile. ric data for the second oxidation because of the instability
of the product, but the voltammetric peak heights suggested
Redox process
(1)
(2)
(3)
the assignment to a one-electron process.[37]
[a]
Mn ox. states
II,III/II,II
–0.535
70
III,III/II,III
–
The electrode processes (1) and (2) exhibited much larger
CV peak splits at high scan rates than process (3) (Fig-
ure 4). This strongly non-Nernstian behaviour is indicative
of slow heterogeneous electron transfer. We could obtain
the standard rate constants k° to a close approximation by
digital simulations (Figure 4, c) using a model of three qua-
sireversible waves and the parameters given in Table 3. Ne-
glecting work terms, the standard rate constant for a hetero-
geneous electron transfer reaction is given by
[b]
E1/2 [V]
0.380
100
495
5
0.755
∆Ep(0.1)[c,d] [mV]
∆Ep(100)[c,f] [mV]
k°[f,g] [10–3 cms–1]
–
[e]
580
2
190
100
[a] Not assigned. [b] Half-wave potentialϮ0.02 V vs. ferrocene/
ferrocenium. [c] Peak split (scan rate in volts per second). [d] Glassy
carbon electrode. [e] Irreversible at 0.1 Vs–1. [f] Pt microdisk elec-
trode. [g] Apparent standard rate constant from digital simulation
of cyclic voltammograms taking α = 0.5 and D = 1ϫ10–5 cm2 s–1.
(3)
ture) could be estimated. The reduction and the first oxi-
dation of 5 are one-electron processes as evidenced by coul-
ometry during controlled potential electrolysis at –0.93 and
0.57 V, respectively. We could not obtain reliable coulomet-
where Zhet is the collision factor and λ is the reorganisation
energy.[38] For 5 we obtain
Zhet = (RT/2πM)1/2 = 2.2ϫ103 cms–1
(4)
(M = 883 gmol–1) and can estimate approximate reorganis-
ation energies of 1.4, 1.3 and 1.0 eV for the electrode pro-
cesses (1), (2) and (3), respectively. The contribution of the
solvent reorganisation can be estimated as λsolv Յ 0.3 eV.[39]
This would indicate that the electrode reactions, in particu-
lar (1) and (2), are associated with large inner reorganis-
ation energies of λi ≈ 1 eV. A similar estimate of λi based
on crystallographic data for the II,III/II,II couple of
[Mn2(bpmp)(µ-OAc)2]2+/+ has been presented and can ex-
plain the unusually slow electron transfer observed in elec-
tron transfer dyads and triads that involve manganese com-
plexes as donor units.[14,40] For the MnII/MnIII couples this
can be attributed to the Jahn–Teller distortion associated
with the high-spin d4 configuration of the MnIII centre. The
significantly smaller reorganisation energy for the second
oxidation implies that the coordination geometry changes
are less pronounced between MnIII and MnIV or that this
process is a ligand-based rather than a metal-centred oxi-
dation.
Spectroscopic Characterisation of the Oxidation and
Reduction Products
The one-electron reduction product was prepared by
bulk electrolysis at –0.93 V. Its EPR spectrum (Figure 3b) is
II,II
typical for a weakly antiferromagnetically coupled Mn2
complex and the reduction of 5 can be attributed to the
metal-centred process (5).
[Mn2II,IIIL(µ-OAc)2]+ + e– Ǟ [Mn2II,IIL(µ-OAc)2]
(5)
Mn2II,II is an integral spin system, where the diamagnetic
(S = 0) ground state dominates at liquid He temperatures
and accordingly we did not detect any EPR signals at 4 K.
II,II
Weakly coupled Mn2
complexes are, on the other hand,
Figure 4. Voltammograms of
5 (1 m) in acetonitrile with
known to display EPR-active excited states at temperatures
above about 7 K, similar to that in Figure 3, spectrum b.[41]
We therefore conclude that the spectrum in Figure 3 (b),
(n-C4H9)4N·ClO4 (0.1 ) as supporting electrolyte. (a) Differential
pulse voltammogram (step potential 5 mV, modulation amplitude
25 mV, modulation time 50 ms, interval time 100 ms, glassy carbon
disk, 3 mm). (b) Cyclic voltammograms (v = 0.100 Vs–1, glassy car-
bon disk, 3 mm). (c) Cyclic voltammogram (v = 100 Vs–1, Pt micro-
disk, 25 µm) and digital simulation (– · –) with the parameters from
Table 1.
which was recorded at 12 K, originates from a superposi-
II,II
tion of one or more excited spin states from Mn2
with S
= 1, 2, .... When [MnII2L(µ-OAc)2] was reoxidised by elec-
5038
www.eurjic.org
© 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Inorg. Chem. 2006, 5033–5047

Redox Chemistry of a Dimanganese(II,III) Complex
FULL PAPER
trolysis at –0.08 V, the Mn2II,III spectrum of 5 was quantita- ≈ 0.5 eV for the metal-to-metal charge transfer (8) and λ
tively recovered (Figure 3, c), confirming the stability of Ͼ 1 eV per metal centre, the intervalence charge transfer
[MnII2L(µ-OAc)2].
transitions might be superimposed on the LMCT bands in
The one-electron oxidation product was produced by the UV/Vis range.
bulk electrolysis at 0.57 V. The product is EPR-silent and
the spectrum only showed a residual (5%) of the original
II,III
Mn2
signal (Figure 3, d). Mn2III,III, which is an integral
spin system, might produce EPR-active states at higher
temperatures, but so far no EPR signals from this valence
state in either parallel or perpendicular mode EPR at X-
band frequencies have been found by us or other research-
ers. Therefore, the EPR-silent product of the first oxidation
can be assigned to[Mn2III,IIIL(µ-OAc)2]2+ according to
equation (6).
[Mn2II,IIIL(µ-OAc)2]+ Ǟ [Mn2III,IIIL(µ-OAc)2]2+ + e–
(6)
After reduction of [Mn2III,IIIL(µ-OAc)2]2+at –0.08 V
II,III
about 95% of the original Mn2
signal was recovered
Figure 5. Electronic spectra of [Mn2L(µ-OAc)2] (–), [Mn2L(µ-
OAc)2]+ (– –) and [Mn2L(µ-OAc)2]2+ (· · ·) in acetonitrile with (n-
C4H9)4N·ClO4 (0.1 ).
III,III
(not shown). A slow degradation reaction of [Mn2
-
L(µ-OAc)2]2+ resulted in the release of minor amounts of
monomeric MnII that could be observed in the spectra of
the oxidised and re-reduced material (2% and 5% of the
total Mn respectively), where it gave rise to the weak EPR
signal around g = 2 with six hyperfine lines at a separation
of about 100 gauss (Figure 4, d).
Table 4. UV/Vis absorption data of [Mn2L(µ-OAc)2]n+ in acetoni-
trile.
n
Mn oxidation state
λmax [nm] (ε [–1 cm–1])
0
1
2
II,II
II,III
III,III
202 (61200), 253 (20100), 315 (7350)
260 (18300), 393 (3200), 795 (420)
262 (22000), 417 (3900), 830 (650)
The short-lived primary product of the second oxidation
could not be prepared by bulk electrolysis. After electrolysis
at 0.82 V a significant amount of monomeric MnII was ob-
served by EPR, obscuring any other paramagnetic species
that might be present in the solution (not shown). The in-
stability of the oxidised complex on longer timescales can
be attributed to oxidative degradation of the ligand, as evi-
denced by the release of monomeric MnII ions. This reac-
tion might be triggered either by a transiently formed high-
valent metal centre (7a) or by direct ligand oxidation (7b).
[(N)3MnII(µ-O-φ)(µ-OAc)2MnIII(N)2(O)] Ǟ
[(N)3MnIII(µ-O-φ)(µ-OAc)2MnII(N)2(O)] (8)
Upon the second oxidation the optical absorption above
377 nm and below 310 nm was bleached and an increasing
absorption was initially observed between these wave-
lengths. The isosbestic points were, however, not main-
tained and a pure spectrum of the initial product could not
be obtained.
Scheme 2 compares the redox properties of 5 to those of
the symmetric analogues [Mn2(bpmp)(µ-OAc)2]+ and
[Mn2(bhpp)(µ-OAc)2]+ with (N3O3)2 and (N2O4)2 donor
sets, respectively.[17,21] It illustrates the stabilisation of high-
valent metal centres by the increasing number of oxygen
donors. For the unsymmetric complex the mixed (N3O3)/
[Mn2III,IIIL(µ-OAc)2]2+ Ǟ [Mn2III,IVL(µ-OAc)2]3+ + e–
[Mn2III,IIIL(µ-OAc)2]2+ Ǟ [Mn2III,IIIL·(µ-OAc)2]3+ + e–
(7a)
(7b)
UV/Vis Spectroelectrochemistry
The electronic absorption spectra of [Mn2L(µ-OAc)2]n+
(n = 0, 1, 2) in the UV and visible range are shown in Fig-
ure 5 and Table 4. The spectra after one-electron reduction
and one-electron oxidisation of 5 were obtained by spectro-
electrochemistry. The spectral changes induced by reduction
and the first oxidation were fully reversible and isosbestic
points (221, 242, 260, 306, 332 nm) were maintained in the
(N2O4) donor set results in localisation of and an extended
II,III
stability to the mixed-valence Mn2
state, with a free en-
ergy of comproportionation of ∆Gc = –88 kJmol–1 [Kc =
exp(∆E1/2F/RT) = 5.49ϫ1015 at 293 K] as compared to
–54.0 and –34.7 kJmol–1 in the (N3O3)2 and (N2O4)2 ana-
II,II
course of the reduction (Figure 5, left panel). In the Mn2
state the complex displayed basically no visible absorption
as metal-to-ligand charge transfer (MLCT) bands of MnII
complexes are usually of low intensity. The visible absorp-
II,III
III,III
tion bands of the Mn2
and Mn2
states (Figure 5,
right panel) can be attributed to ligand-to-metal charge
transfer (LMCT) transitions from multiple donor functions
to the MnIII centres. No additional band of the mixed-val-
ence complex could be detected including the near infrared
spectrum (Յ 2700 nm). With a free energy change of ∆G° Scheme 2.
Eur. J. Inorg. Chem. 2006, 5033–5047
© 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org
5039

R. Lomoth, A. Magnuson et al.
FULL PAPER
logues, respectively. For 5, as for the symmetric analogues, sharp absorption peak arising from the acetate ligands (νas,
no metal oxidation state higher than Mn2III,III could be ob- 1590 cm–1, ε = 4170 –1 cm–1) is depleted and replaced by
tained by bulk electrolysis in nonaqueous solvents, and the the broader and weaker νas band of free acetate ions that
second oxidation of 5 resulted in a short-lived product that can be observed at lower frequency (1574 cm–1, ε =
could not be characterised. This demonstrates the general 840 –1 cm–1).[18] The ratio of free versus bound acetate ap-
problem that several accumulative oxidations without proaches unity, indicating that on average only one acetate
charge compensation will be spread over a large potential per molecule of 5 is lost, even for the highest water concen-
range, placing the potentials for generating the higher metal tration (inset in Figure 6, a). Assuming that the absorption
oxidation states so high that oxidative degradation of the changes represent a uniform reaction, this indicates that
complexes is inevitable.
either all complexes have lost one acetate or half of them
have lost both ligands. The dependence of the acetate con-
centration on the water content (Figure 6, b) is in good
IR Spectroelectrochemistry
IR spectra of the isovalent complexes [Mn2II,IIL(µ-
OAc)2] and [Mn2III,IIIL(µ-OAc)2]2+ (see Supporting Infor-
mation) were obtained by spectroelectrochemistry that gave
fully reversible absorption changes for reduction and oxi-
dation of 5. For the νas mode of the acetate ligands only a
minor shift to higher frequency is observed upon reduction,
while oxidation results in a broader, less symmetric νas
band. The absorption between 1440 and 1440 cm–1 can be
assigned to the νs mode of the acetate ligands superimposed
on absorptions of the binucleating ligand. The spectral
changes induced by reduction and oxidation suggest a shift
towards higher frequency in the II,II state and to lower fre-
quency in the III,III state. The same trend has been found
for the Mn2(bpmp)(µ-OAc)2 complex[18] and the values of
∆ν = ν – ν of about 120, 170 and 220 cm–1 would be in
agreement with (9), with an equilibrium constant K9
7.6ϫ10–5 –1 (Figure 6, c), while other obvious possibilities
(10, 11) are not in agreement with the IR data.
=

˜
˜
˜
s
as
agreement with bridging coordination of the acetate in all
three oxidation states.
Redox Properties in Partially Aqueous Media
In previous studies of [Mn2(bpmp)(µ-OAc)2]+ and
[Mn2(bhpp)(µ-OAc)2]+, we have shown that complexes in
higher oxidation states can be generated in partially aque-
ous solutions.[17–19] For [Mn2(bpmp)(µ-OAc)2]+ we could
show that the exchange of acetate bridges for water-derived
ligands (H2O, OH–, O2–) accounts for the changes in redox
behaviour in the presence of water. We were able to identify
the products of the ligand exchange reactions and could
demonstrate that increasing deprotonation of water-derived
ligands provides the charge compensation needed for the
stabilisation of high metal oxidation states.[18] In this way a
Figure 6. (a) IR spectra of 5 (3.5 m) in [D3]acetonitrile and [D3]-
acetonitrile/water (D2O) mixtures. Arrows indicate direction of ab-
sorbance changes for increasing D2O concentration. Dashed line is
the spectrum of ionic acetate (3.5 m as NaOAc). Inset: ratio R =
[OAc–]/[M-OAc] of ionic vs. coordinated acetate as a function of
water concentration. (b) Absolute concentration of ionic acetate
and (c) plot according to Reaction (9) with a slope of K9
=
di-µ-oxo-bridged dimer [Mn2III,IV(bpmp)(µ-O)2]2+ can
7.6ϫ10–5 –1. [Lines in (a) inset and (b) are computed with this
value of K9.]
II,II
be obtained by three-electron oxidation of [Mn2
-
(bpmp)(µ-OAc)2]+ that cannot be oxidised beyond the
III,III
Mn2
state, in its original di-µ-acetato form.
[Mn2L(µ-OAc)2]+ + 2H2O p [Mn2L(µ-OAc)(H2O)2]2+ + OAc– (9)
It could be anticipated that similar ligand exchange pro-
cesses might take place in 5. We therefore wanted to investi-
gate the extent of these processes in 5, and if complexes
with oxidation states higher than Mn2III,III could be formed
from 5 as a result of ligand exchange in aqueous media.
[Mn2L(µ-OAc)2]+ + H2O p [Mn2L(µ-OAc)(µ-H2O)]2+ + OAc–
(10)
[Mn2L(µ-OAc)2]+ + 2H2O p [Mn2L(µ-H2O)2]3+ + 2OAc–
(11)
In summary, the IR data suggest that one of the acetate
bridges is lost more readily and is replaced by two terminal
IR Spectroscopy
The exchange of acetate ligands for water-derived ligands aquo ligands while the second acetate remains coordinated
was monitored by IR spectroscopy in CD3CN/D2O mix- to the Mn centres even at the highest water concentrations.
tures (Figure 6, a). With increasing water concentration the In this respect 5 behaves differently from the bpmp ana-
5040
www.eurjic.org
© 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Inorg. Chem. 2006, 5033–5047

Redox Chemistry of a Dimanganese(II,III) Complex
II,II
FULL PAPER
logues where the Mn2
complex, which has the same
With higher water concentration (10% v:v, 5.5 ) the
II,III
charge as 5, as well as the Mn2
complex, in the same first oxidation wave (2) is already depleted and fully
valence state as 5, release both acetate ligands under com- irreversible while the second oxidation wave (3) of the origi-
parable conditions.[18] Another difference between 5 and nal complex is absent (Figure 7, c). This shows that with
[Mn2(bpmp)(µ-OAc)2]2+ is the protonation state of the the increased water content more than half of the com-
II,III
water-derived ligands. For the bpmp complex it was ob- plexes react with water already in the initial Mn2
state,
II,III
III,III
served that the reaction of the Mn2
state with water and the more reactive Mn2
state is converted com-
results in free acetate as well as acetic acid, which can be pletely by the ligand exchange reactions. The additional an-
rationalised in terms of deprotonation of the water-derived odic peaks around 0.6 and 0.9 V can be attributed to the
ligands, resulting in the formation of oxo- and hydroxo oxidation of the products from the ligand exchange reac-
bridges. In the case of 5 the absence of acetic acid hence tions. The first attempts to investigate the ligand exchanged
implies that the aquo ligands are less acidic and remain oxidation products by EPR spectroscopy after bulk electrol-
protonated when bound to the Mn2II,III complex. It is there- ysis of 5 in partially aqueous solutions were unsuccessful
fore likely that the water-derived ligands are terminal and however. We therefore sought information on the oxidation
not bridging in this valence state, although the formation products from IR spectroscopy, which benefits from more
of bridging aqueous ligands cannot be ruled out.
rapid electrochemical conversion in the thin-layer cell em-
ployed for IR spectroelectrochemistry.
Electrochemistry
IR Spectroelectrochemistry
The effect of the ligand exchange reactions on the redox
behaviour of 5 was studied by cyclic voltammetry in dif-
ferent acetonitrile/water mixtures (Figure 7). With lower
water concentration (1% v:v, 0.55 ) the ligand exchange
equilibrium is largely on the left hand side and the magni-
tude of the first oxidation wave (2) is identical to dry aceto-
nitrile solution (Figure 7, a and b). Also the reversibility of
this wave (2) as well as the magnitude and reversibility of
the reduction wave (1) (not shown) are unaffected. This in-
Figure 8 shows the IR absorption changes induced by
oxidation of 5 in acetonitrile solution with 10% water. Elec-
trolysis at the potential of the first oxidation (0.47 V vs.
Fc+/0) generates acetic acid (1725, 1380 and 1300 cm–1) as
well as free acetate ions with their absorption band
(1575 cm–1) partly cancelled by the bleach of the acetate
ligand band (1590 cm–1). This is in line with the notion that
higher oxidation states are more reactive towards water (see
Electrochemistry) and demonstrates that the water-derived
ligands are at least partly deprotonated upon oxidation of
the Mn centres. Subsequent oxidation at 0.72 V versus
Fc+/0 transforms the remaining acetate ligands as well as
the acetate released in the previous oxidation step into ace-
tic acid. This shows that the product(s) generated in the
second oxidation wave peaking at 0.6 V do not carry any
acetate ligands and that the water-derived ligands are fur-
ther deprotonated.
dicates that under these conditions 5 does not react with
II,III
water to a sizeable extent neither in the initial Mn2
nor
III,III
II,II
in the Mn2
or Mn2
forms.
Therefore, it can be concluded that the most oxidised
products are oxo- ore hydroxo-bridged complexes. For
III,IV
[Mn2II,II(bpmp)(µ-OAc)2]+ we showed that [Mn2
-
(bpmp)(µ-O)2]2+ was formed close to the potential of the
III,III/II,III couple of the parent complex. The product
was identified by rapid on-line transfer of the electrolysis
products to a mass spectrometer.[18] The same product com-
plex was also detected by EPR spectroscopy after light-
induced oxidation of [Mn2II,II(bpmp)(µ-OAc)2]+ via
[RuIII(bpy)3]3+ and rapid freezing of the products.[17,19]
With the increased O/N ratio of complex 5 we anticipated
that product complexes at least at the Mn2III,IV level or even
higher might be formed with the RuIII oxidant (E1/2
=
0.90 V vs. Fc+/0) and be detected by a combination of
photo-oxidation and EPR spectroscopy.
EPR Spectroscopy of Photo-Oxidised Samples
The oxidant [RuIII(bpy)3]3+ was generated by oxidative
quenching of photoexcited [RuII(bpy)3]2+. The sacrificial
electron acceptor [CoIII(NH3)5Cl]2+ avoids electron recom-
bination, and thereby allows for accumulative oxidation of
the manganese complex. To dissolve a sufficient amount of
Figure 7. Cyclic voltammograms (v = 0.100 Vs–1, first scans) of 5
(1 m) with (n-C4H9)4N·ClO4 (0.1 ) as supporting electrolyte. (a)
Anhydrous acetonitrile. (b) Acetonitrile with 1% water (v:v). (c)
Acetonitrile with 10% water (v:v). The dashed and dotted line
shows the solvent background with 10% water.
Eur. J. Inorg. Chem. 2006, 5033–5047
© 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org
5041

R. Lomoth, A. Magnuson et al.
FULL PAPER
strates that even at this very high water content the major
II,III
part of the material prevails as coupled Mn2
dimers,
and the broadened hyperfine structure can be attributed to
strain effects due to accessibility to the bulk solution.
Figure 9. EPR spectra of 5 (0.5 m) exposed to a variable number
of laser flashes at 20 °C in the presence of [RuII(bpy)3]2+ (4 m)
and [CoIII(NH3)5Cl]2+ (8 m) in acetonitrile/aqueous solution (1:1
v/v). The spectra were recorded from the same sample, exposed to
0 (a), 20 (b), 100 (c), 180 (d) and 380 (e) laser flashes. In spectra d
and e the magnified parts on the low-field side show the remaining
III,IV
presence of the EPR signal from Mn2
in Figure 4.
. Spectrometer settings as
Upon illumination with an increasing number of laser
II,III
flashes, the initial EPR signal from Mn2
decreased suc-
cessively (Figure 10). To ensure accurate quantification of
Mn2II,III, all measurements were made at 4 K, where all
complexes are in the S = 1/2 ground state.[36] The simulta-
neously increasing EPR signal from the reduced electron
acceptor [CoII(NH3)x(H2O)y]2+ (g = 5, not shown but cf.
Figure 8. IR spectroelectrochemistry of 5 (3.5 m) in [D3]acetoni-
trile with 5.5  water (D2O). Absorbance changes upon oxidation
at 0.47 V vs. Fc+/0 (a) and subsequent oxidation at 0.72 V vs.
Fc+/0 (b). Spectra before oxidation (c) and after exhaustive oxi-
dation at 0.47 V (d) and 0.72 V (e).
II,III
Huang et al.[17]) showed that the depletion of the Mn2
signal is due to oxidation with the photogenerated [RuIII
-
(bpy)3]3+
After 40 flashes a new signal at g = 2 appeared (Figure 9,
.
the acceptor, a high water content (1:1 water/acetonitrile)
was needed for these measurements. The results from the spectrum c). It bears the typical signatures of a S = 1/2
IR spectra demonstrate that under these conditions each ground state of a strongly coupled Mn2III,IV complex, being
complex loses one of the acetate ligands already in the ini- about 2300 gauss wide and displaying 16 hyperfine lines
II,III
tial Mn2
state. When 5 was dissolved in the reaction with a spacing of 75–80 gauss.[34,42] The hyperfine splitting
II,III
mixture, the 17 hyperfine lines in the Mn2
EPR spec- arises because of strong antiferromagnetic coupling typical
III,IV
trum were broadened, and a signal with less-resolved hyper- of di-µ-oxo- or di-µ-hydroxo-bridged Mn2
dimers. For
fine structure dominated the spectrum of the starting mate- [Mn2(bpmp)(µ-OAc)2]+ we could demonstrate that the
rial (Figure 9, spectrum a). However, the integrated area of aquo and hydroxo complexes formed by ligand exchange in
the spectrum was almost the same as for the well-resolved the lower oxidation states were ultimately transformed into
II,III
III,IV
Mn2
spectrum in acetonitrile (the magnitude of the the di-µ-oxo complex upon oxidation to the Mn2
EPR signal corresponds to about 80% of the nominal con- level.[18] With the additional phenolate donor ligand in 5, a
centration), which indicates that essentially the entire sam- µ-oxo-µ-hydroxo complex is the most probable structure for
ple remained in the original valence state. This demon- this oxidation state.
5042
www.eurjic.org
© 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Inorg. Chem. 2006, 5033–5047

Redox Chemistry of a Dimanganese(II,III) Complex
FULL PAPER
complex or a Mn2III,IVL· form
IV,IV
could be either a Mn2
with a ligand-based radical strongly coupled to the metal.
Subsequent oxidation could also account for the relatively
low yield of Mn2III,IV species. When the Mn2III,IV signal was
almost fully depleted, a weak radical signal could be ob-
served (Figure 9, in the centre of spectrum d) that was
about 70 G wide (peak-to-trough) and had enhanced relax-
ation properties, as a consequence of being close to a para-
magnetic metal centre. It is similar to the radical spectrum
observed upon oxidation of the analogous [Mn2(bhpp)-
(OAc)2] complex[21] and most likely arises from the oxidised
tert-butyl-substituted phenol group. If it is the case that fur-
III,IV
ther oxidation of the Mn2
species results in formation
of Mn2III,IVL·, the detected radical signal might arise from
a minor fraction of the product where the oxidised phenol
ligand has been detached from the manganese, while the
major fraction with the radical coordinated to the metal
III,IV
centre is EPR-silent. The disappearance of the Mn2
II,III
Figure 10. Flash-dependent concentrations of Mn2
(–᭺–) and
EPR species cannot be explained by disintegration of the
complexes, as that usually results in the release of mono-
meric MnII. In contrast, only a small quantity of MnII
(about 2% of the total Mn concentration) was released ac-
cording to its EPR signal consisting of six lines at g = 2
(Figure 9, spectrum e).[43]
III,IV
Mn2II,III (–᭹–). The data are obtained after quantification of
III,IV
Mn2
ure 7. All Mn2
and Mn2
from EPR spectra similar to those in Fig-
II,III
is quantitatively oxidised after 50 flashes, and
III,IV
the maximum observed amount of Mn2
appears after 100
flashes. The intermediate, Mn2III,III, cannot be observed by EPR.
The inset shows a magnification of the data obtained between
flashes 0 and 50, to highlight the lag phase in formation of
III,IV
In summary, the results of photo-oxidation show that
Mn2
.
oxidation of 5 in partly aqueous solution generates a
III,IV
Mn2
complex, most likely with a di-µ-oxo or di-µ-oxo/
II,III
III,IV
Oxidation of the Mn2
complexes to the Mn2
hydroxo bridging motif replacing the acetate ligands. In ad-
product with the one-electron photo-oxidant RuIII(bpy)3 dition, our data indicate that further photoinduced oxi-
III,III
IV,IV
should proceed via intermediates at the Mn2
level. dation of this complex might result in a Mn2
or
These are expected to be EPR-silent and not observable by Mn2III,IV-L· product.
EPR, which explains the lag phase between complete disap-
II,III
pearance of the Mn2
intensity of the Mn2
EPR spectrum and the maximum
III,IV
EPR signal (Figure 10). After 40
Conclusions
II,III
flashes, approximately equal concentrations of the Mn2
III,IV
and Mn2
species coexist in the reaction solution mix-
The unsymmetric ligand in the mixed-valent manganese
ture, demonstrating that the high oxidation state in the complex 5 ([Mn2II,IIIL(µ-OAc)2]ClO4) provides, together
III,IV
Mn2
species possesses the desired stability. If we as- with the bridging acetates, a N3O3 and a N2O4 donor set
sume that the potential of the III,III/II,III couple of the to the MnII and MnIII centres respectively. Compared to
species we observe under the experimental conditions of the symmetric ligand analogues bpmp[16] and bhpp,[21] the
flash photolysis is similar to that of the di-µ-acetato com- unsymmetric ligand provides an intermediate O/N ratio and
plex (0.38 V vs. Fc+/0), we can assume a similar potential substantially stabilises the mixed-valence state, where the
III,IV
for the III,IV/III,III couple of the Mn2
species. We higher-valent metal ion is localised in the N3O3 environ-
made an analogous estimation of potentials for ment. The pseudo-octahedral coordination geometry is
[Mn2(bpmp)(µ-OAc)2]+, where the mass spectrometry data largely distorted and conversion to the isovalent states is
indicated formation of [Mn2III,IV(bpmp)(µ-O)2]2+ at poten- associated with large inner reorganisation energies resulting
tials close to the III,III/II,III couple of the parent com- in slow electron transfer reactions.
plex.[18] This compression of multiple redox reactions in a
Complex 5 displays two oxidation processes in nonaque-
narrow potential range, which is a consequence of the li- ous solvents, where charge accumulation precludes the for-
gand exchange reactions, indicates that disproportionation mation of stable oxidation states beyond the Mn2III,III state.
of III,III intermediates might contribute to the formation This situation changes in aqueous media where the acetate
III,IV
of the Mn2
The Mn2
species.
bridges of 5 are replaced by water-derived ligands. These
III,IV
species reached a maximum concentration provide charge compensation by deprotonation upon oxi-
II,III
of about 10% relative to the initial Mn2
started to decrease after about 100 flashes (Figure 10). The interval for accumulative oxidations becomes more com-
signal before it dation of the metal centres.[18,44] As a result, the potential
III,IV
Mn2
spectrum was not immediately replaced by any pressed, so that three metal-centred oxidation processes can
other EPR-active species, suggesting that the Mn2III,IV com- be driven by [RuIII(bpy)3]3+ (E1/2 = 1.27 V vs. NHE). More-
plex was further oxidised to an EPR-silent product. This over, the ability to acquire water-derived ligands is an essen-
Eur. J. Inorg. Chem. 2006, 5033–5047
© 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org
5043

R. Lomoth, A. Magnuson et al.
FULL PAPER
2-{[(3,5-Di-tert-butyl-2-hydroxybenzyl)(pyrid-2-ylmethyl)amino]-
methyl}-6-(hydroxymethyl)-4-methylphenol (3): Complex 2 (0.94 g,
2.0 mmol) was dissolved in methanol (20 mL) and treated with so-
dium borohydride (0.17 g, 4.4 mmol). The reaction was stirred at
room temperature overnight. The solution was evaporated under
reduced pressure and the residue was dissolved in water, acidified
to pH 2–3 and stirred for 1 h. The pH was adjusted to 7–8 with
sodium hydrogen carbonate and then extracted with dichlorometh-
ane. The combined organic phases were washed with brine, dried
with anhydrous sodium sulfate and filtered. Evaporation under re-
tial characteristic of any functional model for the water-
oxidising complex in PSII.
III,IV
The product of the photoinduced oxidation is a Mn2
complex, which can be further oxidised to, most likely, a
IV,IV
Mn2
complex. The high-valent complexes are oxo- or
hydroxo-bridged, which is supported by the EPR spectrum
observed for the Mn2III,IV complex, which is consistent with
a strongly antiferromagnetically coupled complex.
These results demonstrate that four, possibly five, manga-
nese valence states can be sustained in the same binucleat- duced pressure of the solvent afforded 0.92 g of 3 (97%). 1H NMR
(CDCl3): δ = 1.27 (s, 9 H), 1.41 (s, 9 H), 2.24 (s, 3 H), 3.77 (s, 2
H), 3.78 (s, 2 H), 3.85 (s, 2 H), 4.68 (s, 2 H), 6.84 (d, 1 H), 8.90 (d,
1 H), 6.96 (d, 1 H), 7.17 (d, 1 H), 7.22 (d, 1 H), 7.32 (dt, 1 H), 7.74
(dt, 1 H), 8.86 (dd, 1 H), 10.80 (s broad, 2 H) ppm.
ing ligand, as the build-up of charge can be balanced be-
cause of the flexibility of the bridging motif. Water oxi-
dation in PSII is believed to involve at least MnIV or even
MnV to provide sufficiently electrophilic oxo ligands that
can engage in O–O bond formation.[5,7,45] Compared to the
2-{[Bis(pyrid-2-ylmethyl)amino]methyl}-6-{[(3,5-di-tert-butyl-2-hy-
related [Mn2(bpmp)(OAc)2]+ we note that the increased O/ droxybenzyl)(pyrid-2-ylmethyl)amino]methyl}-4-methylphenol, H2L
(4): Complex 3 (2.27 g, 4.8 mmol) was dissolved in freshly distilled
thionyl chloride (5 mL) and CH2Cl2 (30 mL). The solution was
stirred at room temperature for 2.5 h and then distilled under re-
duced pressure. The residue was dissolved in CH2Cl2 and washed
with a cooled sodium hydrogen carbonate solution. The organic
phase was dried with anhydrous sodium sulfate, filtered and evapo-
rated under reduced pressure. The residue was dissolved in CH2Cl2
and added to a solution of bis-picolylamine (0.95 g, 4.8 mmol) and
triethylamine (1.03 g, 10 mmol) and refluxed overnight under ar-
gon. The solution was washed with brine and dried with anhydrous
N ratio of 5 stabilises higher valence states, which results
in oxidation of the ligand exchange products beyond the
III,IV
Mn2
level. This is a significant improvement in view of
prospective biomimetic water oxidation catalysts. It is how-
ever important to note that there is an unavoidable trade-
off between stabilising higher oxidation states and main-
taining sufficient oxidising potential. For the Mn2III,IV com-
plex derived from 5 the estimated oxidation potential of the
III,IV/III,III couple is probably comparable to the III,III/
II,III couple of the parent complex (≈0.4 V vs. Fc+/0, ≈0.8 V sodium sulfate. After filtration and evaporation under reduced
pressure the crude product was purified by column chromatog-
raphy on silica gel, eluting with CH2Cl2/MeOH, 95:5. Yield 2.10 g
vs. NHE). This suggests that all oxidation products derived
from 5 are within a desirable potential range with respect
to the average potential of 0.81 V versus NHE (at pH 7)
required for water oxidation. Thus, the O/N ratio in 5 seems
to create close to an optimal balance between oxidising po-
tential and valence state stabilisation.
1
(67%). H NMR (CDCl3): δ = 1.25 (s, 9 H), 1.38 (s, 9 H), 2.19 (s,
3 H), 3.73 (s, 2 H), 3.79 (s, 2 H), 3.82 (s, 2 H), 3.83 (s, 2 H), 3.86
(s, 4 H), 6.79 (d, 1 H), 6.84 (d, 1 H), 6.96 (d, 1 H), 7.06–7.15 (m,
4 H), 7.34 (d, 2 H), 7.42 (d, 2 H), 7.51–7.59 (m, 3 H), 8.49 (d, 1
H), 8.58 (d, 1 H), 10.90 (s broad, 2 H) ppm.
Mn2L(OAc)2·ClO4 (5): Mn(OAc)3·(H2O)2 (1.04 g, 3.9 mmol) was
added to an ethanol solution (9 mL) of 4 (1.05 g, 1.6 mmol) in one
portion. The dark red-brown solution was heated to 60 °C under
argon for 20 min. A solution of NaClO4·H2O (0.58 g, 4.1 mol) in
ethanol/water (3:1, 4 mL) was added dropwise and the reaction
solution was slowly cooled to room temperature and then stored
in a freezer (–18 °C) for 48 h. The dark red-brown microcrystalline
solid was filtered and washed with cooled ethanol and diethyl ether.
The solid was recrystallised from ethanol/hexane to give 0.79 g
dark red-brown rod-like crystals (yield 52%) suitable for X-ray dif-
Experimental Section
Materials: 2-(Chloromethyl)-6-formyl-4-methylphenol,[27] dipicol-
ylamine[28]
and
(3,5-di-tert-butyl-2-hydroxybenzyl)(2-pyridyl-
methyl)amine[21] were prepared according to the literature pro-
cedures. Dichloromethane and triethylamine were distilled under
argon over sodium/benzophenone and calcium hydride, respec-
tively. Column chromatography was performed using silica gel (60–
80 mesh). All other reagents were of reagent grade quality and used
as received. 1H NMR spectra were recorded with a Varian 400-
MHz spectrometer.
fraction. ESI-MS: (m/z) (%)
=
883.3354 [M
–
ClO4]+.
+
C46H55N5O6Mn2 (883.2913): found C 56.03, H 5.71, N 7.08, Cl
3.48, Mn 11.05. C46H55Mn2N5O6·ClO4: calcd. C 56.19, H 5.64, N
7.12, Cl 3.61, Mn 11.17.
2-{[(3,5-Di-tert-butyl-2-hydroxybenzyl)(pyrid-2-ylmethyl)amino]-
methyl}-6-formyl-4-methylphenol (2): (3,5-Di-tert-butyl-2-hydroxy-
benzyl)(pyrid-2-ylmethyl)amine (1.64 g, 5.0 mmol) and triethyl-
amine (0.62 g, 6.1 mmol) in CH2Cl2 (15 mL) under argon were
added to a solution of 2-(chloromethyl)-6-formyl-4-methylphenol
(1) (1.03 g, 5.6 mmol) in CH2Cl2 (15 mL). The reaction mixture
was refluxed overnight under argon. The reaction was poured into
water and extracted with CH2Cl2. The combined organic phases
were dried with anhydrous sodium sulfate, filtered and evaporated
under reduce pressure. The residue was purified by column
chromatography, eluting with CH2Cl2/MeOH, 97:3. Yield 1.84 g
X-ray Crystallography: Single crystal X-ray diffraction patterns
were recorded with a Stoe IPDS diffractometer on a rotating anode
Mo-radiation source (λ = 0.71073 Å) with &phis; scans of 1° width.
The total rotation range was 200°. The sample-detector distance
was 70 mm and with the diameter of the image plate being 180 mm
this gave max 2θ ≈ 52°. Measuring further out in 2θ proved to give
very little significant extra data. Suitable single crystals were chosen
from the crystallisation (see “Synthesis” above) and were mounted
and measured inside sealed glass capillaries with mother liquor sur-
rounding the crystals. Crystals that were simply glued to a glass
pin ceased to diffract after a few minutes. Indexing, cell refinements
and integration of reflection intensities were performed with STOE
IPDS software (Stoe & Cie GmbH, Darmstadt). Numerical ab-
1
(95%). H NMR (CDCl3): δ = 1.26 (s, 9 H), 1.42 (s, 9 H), 2.28 (s,
3 H), 3.77 (s, 2 H), 3.79 (s, 2 H), 3.84 (s, 2 H), 6.86 (d, 1 H), 7.16–
7.20 (m, 2 H), 7.28–7.34 (m, 3 H), 7.65 (dt, 1 H), 8.56 (dd, 1 H),
9.95 (s, 1 H), 10.9 (s broad, 2 H) ppm.
5044
www.eurjic.org
© 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Inorg. Chem. 2006, 5033–5047

Redox Chemistry of a Dimanganese(II,III) Complex
FULL PAPER
sorption correction was performed with the program X-RED
(Stoe & Cie GmbH, Darmstadt) and the crystal shape was verified
with the program X-SHAPE (Stoe & Cie GmbH, Darmstadt) using
multiple measurements of symmetry equivalent reflections. The
structure was solved by direct methods using SHELXS97 giving
electron density maps where most non-hydrogen atoms could be
resolved.[46] The rest of the non-hydrogen atoms were located from
difference electron density maps and the structure model was re-
fined with full-matrix least-squares calculations on F2 using the
program SHELXL97-2.[47] All non-hydrogen atoms were refined
with anisotropic displacement parameters and the hydrogens,
which were placed at geometrically calculated positions and let to
ride on the atoms they were bonded to, were given isotropic dis-
placement parameters calculated as ξ·Ueq. for the non-hydrogen
atoms with ξ = 1.5 for methyl H atoms (–CH3) and ξ = 1.2 for
methylenic (–CH2–) and aromatic H atoms.
2400) between 200 and 2700 nm. Spectroelectrochemical measure-
ments were made in a OTTLE-type quartz cell with an optical path
length of 1 mm. A platinum grid with a size of 10ϫ30 mm2 and
400 meshes per cm2 was used as the working electrode. The counter
and reference electrodes were of the same type as described in the
electrochemistry paragraph. Samples were bubbled for 20 min with
solvent-saturated argon and transferred to the argon-flushed cell
with the argon stream. The spectra were recorded with a UV/Vis
diode array spectrophotometer (Hewlett Packard 8435) with the
background collected on electrolyte solution in the potential free
OTTLE cell.
IR Spectroscopy: IR absorption spectra were recorded at a resolu-
tion of 1 cm–1 with a Bruker FTIR spectrometer (IFS 66v/S) with
the sample as a KBr pellet or as solution in a liquid-sample cell
(Bruker A140) between CaF2 windows and with a path length of
125 µm. The solutions were prepared in CD3CN (Aldrich,
99.8 atom-% D) or mixtures of CD3CN with D2O (Aldrich,
99 atom-% D).
Only 39.2% of the reflections fulfilled the significance criterion I
Ͼ 2σ(I), thus the residual values calculated from all reflection are
larger than those calculated from the significant reflections.
The ratio of free versus coordinated acetate (R = [OAc–]/[M-OAc])
was determined from the absorbance at 1590 cm–1 {Abs1590
=
CCDC-256953 contains the supplementary crystallographic data
for this paper. These data can be obtained free of charge from The
Cambridge Crystallographic Data Centre via www.ccdc.cam.ac.uk/
data_request/cif.
–
d(εOAc [OAc–] + εM-OAc [M-OAc]) = dε[OAc]} as
R = (εM-OAc – ε)/(ε – εOAc–
)
where ε is a formal extinction coefficient and [OAc] = 2ϫ[5] is the
total acetate concentration. The concentration of free acetate is
then given by
Magnetic Susceptibility: The magnetic susceptibility in the tem-
perature range 4–300 K was measured on a powder sample of 5 at
a magnetic field of 1.35 T using a Faraday balance. Correction for
diamagnetism was made using Pascal’s constants.
[OAc–] = {1 – 1/(R + 1)}[OAc]
Electrochemistry: All solutions were prepared from dry acetonitrile
(Merck, spectroscopy grade, dried with MS 3 Å) with 0.1  tetrabu-
tylammonium perchlorate that had been dried in a vacuum at
353 K (Fluka, electrochemical grade) as the supporting electrolyte.
Before all measurements, oxygen was removed by bubbling the
stirred solutions with solvent-saturated argon and the samples were
kept under argon during measurements. Cyclic voltammetry,
differential pulse voltammetry and controlled potential electrolysis
were carried out using an Autolab PGSTAT 100 potentiostat with
a GPES 4.9 electrochemical interface (Eco Chemie). The working
electrode was a glassy carbon disc (diameter 3 mm, freshly pol-
ished) for voltammetry, a Pt disk microelectrode (CH Instruments,
diameter 25 µm, freshly polished with alumina 0.3 µm) for fast vol-
tammetry or a platinum grid for bulk electrolysis. A platinum spiral
in a compartment separated from the bulk solution by a fritted
disk was used as the counter electrode. The reference electrode was
a nonaqueous Ag/Ag+ electrode (CH Instruments, 0.01  AgNO3
in acetonitrile) with a potential of –0.08 V versus the ferrocenium/
ferrocene (Fc+/Fc) couple in acetonitrile as an external standard.
All potentials reported here are versus the Fc+/Fc couple, which is
obtained by adding –0.08 V to the potentials measured versus the
Ag/Ag+ electrode. No compensation for iR drop has been made.
With a specific conductivity of κ Ͼ 0.01 Scm–1 of the electrolyte
the resistance for the microelectrode (r = 12.5 µm) is R = 1/(4κr)
Ͻ 20 kΩ. With voltammetric peak currents i Ͻ 100 nA, the iR drop
is negligible (Ͻ2 mV) in the microelectrode measurements. Digital
simulations of the cyclic voltammograms have been performed with
GPES 4.9 software (Eco Chemie). The simulated voltammograms
have been composed from the faradic current of a quasireversible
and a two-component quasireversible model superimposed on a
linear nonfaradic background. The heterogeneous electron transfer
rate constants are reported as apparent standard rate constants
without corrections for double-layer effects.
For Reaction (9) the equilibrium constant is defined as
K9 = {[Mn2L(OAc)(H2O)2] [OAc–])/([Mn2L(OAc)2] [H2O]2} =
[OAc–]2/{([5] – [OAc–]) [H2O]2}
and was determined from a plot of [OAc–]2/([5] – [OAc–]) over
[H2O]2.
IR spectroelectrochemical measurements were performed on solu-
tions of 5 in CD3CN or CD3CN/D2O with KPF6 (0.1 ). The spec-
tra were recorded in a thin-layer cell equipped with CaF2 windows
and an optical path length of 90 µm. The working electrode was a
carbon fibre mesh (49% open area) and the counter electrode was
a Pt cylinder concentric with the working electrode. The reference
electrode was of the same type as described for electrochemistry
and connected to the cell by a salt bridge (0.5  KPF6 in CD3CN).
CD3CN, D2O and KPF6 were purchased from Sigma Aldrich.
CD3CN was dried with molecular sieves (3 Å) prior to use and
KPF6 was dried overnight at 353 K in vacuo.
EPR Spectroscopy: The electrochemical preparation of EPR sam-
ples was performed by bulk electrolysis of 1-m solutions at con-
trolled potentials. After electrolysis, 250-µL samples were taken
with an argon-filled syringe via a septum from the electrolysis ves-
sel and transferred to argon-flushed EPR tubes. The samples were
immediately frozen and kept in liquid nitrogen. X-band EPR mea-
surements were performed at cryogenic temperatures with a Bruker
E500 ELEXSYS spectrometer equipped with a TE102 resonator, an
Oxford Instruments ESR 900 flow cryostat and an ITC 503 tem-
perature controller.
For flash photolysis, a 1-m solution of 5 in dry acetonitrile was
mixed at a 1:1 ratio with an aqueous solution containing tris(2,2Ј-
bipyridyl)dichloro-ruthenium(II) hexahydrate ([RuII(bpy)3]2+
)
(8 m), and penta-aminechlorocobalt(III) chloride ([CoIII(NH3)5-
Cl]2+) (16 m), as described previously.[32] The solution was trans-
ferred to EPR tubes and the samples were immediately frozen in
an ethanol/dry ice bath (T = 198 K) and then frozen at 77 K. All
UV/Vis Spectroscopy: The electronic absorption spectrum was re-
corded with a UV/Vis-NIR spectrophotometer (Varian CARY
Eur. J. Inorg. Chem. 2006, 5033–5047
© 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org
5045

R. Lomoth, A. Magnuson et al.
FULL PAPER
[9]
M. Haumann, C. Müller, P. Liebisch, L. Iuzzolino, J. Dittmer,
M. Grabolle, T. Neisius, W. Meyer-Klaucke, H. Dau, Biochem-
istry 2005, 44, 1894–1908.
a) Y. Naruta, M. Sasayama, T. Sasaki, Angew. Chem. Int. Ed.
Engl. 1994, 33, 1839–1841; b) M. Watkinson, A. Whiting, C. A.
McAuliffe, J. Chem. Soc., Chem. Commun. 1994, 18, 2141–
2142; c) W. Ruettinger, M. Yagi, K. Wolf, S. Bernasek, G. C.
Dismukes, J. Am. Chem. Soc. 2000, 122, 10353–10357.
a) J. Limburg, J. S. Vrettos, L. M. Liable-Sands, A. C. Rheing-
old, R. H. Crabtree, G. W. Brudvig, Science 1999, 283, 1524–
1527; b) A. K. Poulsen, A. Rompel, C. J. McKenzie, Angew.
Chem. Int. Ed. 2005, 44, 6916–6920.
a) L. Sun, H. Berglund, R. Davydov, T. Norrby, L. Ham-
marström, P. Korall, A. Börje, C. Philouze, K. Berg, A. Tran,
M. Andersson, G. Stenhagen, J. Mårtensson, M. Almgren, S.
Styring, B. Åkermark, J. Am. Chem. Soc. 1997, 119, 6996–
7004; b) A. Magnuson, Y. Frapart, M. Abrahamsson, O.
Horner, B. Åkermark, L. Sun, J.-J. Girerd, L. Hammarström,
S. Styring, J. Am. Chem. Soc. 1999, 121, 89–96; c) L. Sun, L.
Hammarström, B. Åkermark, S. Styring, Chem. Soc. Rev. 2001,
30, 36–49.
L. Sun, M. Raymond, A. Magnuson, D. LeGourrierec, M.
Tamm, M. Abrahamsson, P. Huang Kenéz, J. Mårtensson, G.
Stenhagen, L. Hammarström, S. Styring, S. Åkermark, J. In-
org. Biochem. 2000, 78, 15–22.
M. Borgström, N. Shaikh, O. Johansson, M. F. Anderlund, S.
Styring, B. Åkermark, A. Magnuson, L. Hammarström, J. Am.
Chem. Soc. 2005, 127, 17504–17515.
a) I. Vass, S. Styring, Biochemistry 1991, 30, 830–839; b) C.
Tommos, G. T. Babcock, BBA-Bioenergetics 2000, 1458, 199–
219; c) P. Geijer, F. Morvaridi, S. Styring, Biochemistry 2001,
40, 10881–10891; d) R. Radmer, G. Cheniae, Top. Photosynthe-
sis 1977, 2, 303–348.
a) bpmp = 2,6-bis{[N,N-bis(2-pyridinemethyl)amino]methyl}-
4-methylphenol; b) H. Diril, H. R. Chang, M. J. Nilges, X. H.
Zhang, J. A. Potenza, H. J. Schugar, S. S. Isied, D. N. Hend-
rickson, J. Am. Chem. Soc. 1989, 111, 5102–5114.
sample mixing and handling took place under dim red light and
inert atmosphere, and care was taken to minimise the interval be-
tween mixing and freezing of the samples (mix–freeze time Ͻ
1 min). Laser flashes from a frequency doubled Spectra Physics
DCR 3G Nd:YAG laser at 5 Hz and 532 nm (6 ns pulse width,
about 250 mJflash–1) were given at room temperature to the pre-
thawed EPR samples. The incident laser light was adjusted with
lenses to cover the entire EPR sample volume. After the designated
number of flashes, the sample was frozen in an ethanol/dry ice bath
within 2 s and then quickly transferred to liquid N2 until further
use. X-band EPR spectra were recorded before and after flash illu-
[10]
[11]
[12]
II,III
mination of the EPR samples. Evaluation of the change in Mn2
concentration was done by double integration of the ground-state
EPR signal at 4 K, and comparison with a reference Mn2II,III spec-
trum from a known concentration of 5 in acetonitrile. To quantify
the formation of Mn2(III,IV) we used Mn2III,IV(bpy)4(µ-O)2
as spectroscopic reference.[48] The concentration of Mn2
-
III,IV
(bpy)4(µ-O)2 in acetonitrile was determined by UV/Vis spectropho-
tometry, using the extinction coefficient ε = 900 –1 cm–1 at λ =
687 nm. Simulations of EPR spectra were made using the Xsophe-
Sophe-Xepr-View® software suite from Bruker GmbH.
[13]
Supporting Information (see footnote on the first page of this arti-
cle): Figure S1 shows IR spectroelectrochemistry of 5. IR spectra
of 5 in KBr, and of the isovalent products [Mn2II,IIL(µ-OAc)2] and
[Mn2III,IIIL(µ-OAc)2]2+ in [D3]acetonitrile solution. Selected crys-
tallographic data not shown in Table 1.
[14]
[15]
Acknowledgments
This work was supported by grants from the Swedish Energy Ag-
ency, The Knut and Alice Wallenberg Foundation, The Swedish
Research Council, NEST-STRP SOLAR-H (EU contract no.
516510), The Royal Swedish Academy of Sciences (A. M.), the Carl
Trygger Foundation (A. M.), the Swedish Foundation for Strategic
Research (M. F. A.) and Stiftelsen Bengt Lundqvists minne (P. H.).
[16]
[17]
P. Huang, A. Magnuson, R. Lomoth, M. Abrahamsson, M.
Tamm, L. Sun, B. van Rotterdam, J. Park, L. Hammarström,
B. Åkermark, S. Styring, J. Inorg. Biochem. 2002, 91, 159–172.
[18]
[19]
G. Eilers, C. Zettersten, L. Nyholm, L. Hammarström, R. Lo-
moth, Dalton Trans. 2005, 1033–1041.
P. Huang, J. Högblom, M. F. Anderlund, L. Sun, A. Magnu-
son, S. Styring, J. Inorg. Biochem. 2004, 98, 733–745.
[1] IR spectroelectrochemistry of 5: a) J. H. Alstrum-Acevedo, M.
Kyle Brennaman, T. J. Meyer, Inorg. Chem. 2005, 44, 6802–
6827; b) S. Chakraborty, T. J. Wadas, H. Hester, R. Schmehl,
R. Eisenberg, Inorg. Chem. 2005, 44, 6865–6878.
[2] a) J. H. A. Nugent (Ed.), Special Issue on Photosynthetic Water
Oxidation, BBA-Bioenergetics, Elsevier, Amsterdam, 2001, p.
1503; b) J. Barber, BBA-Bioenergetics 2004, 1655, 123–132.
[3] a) B. A. Diner, G. T. Babcock, in Oxygenic Photosynthesis: The
Light Reactions (Eds.: D. R. Ort, C. F. Yocum), Kluwer Aca-
demic Publishers, Dordrecht, 1996, pp. 213–247; b) J. Barber,
Curr. Opin. Struct. Biol. 2002, 12, 523–530; c) R. D. Britt, in
Oxygenic Photosynthesis: The Light Reactions (Eds.: D. R. Ort,
C. F. Yocum), Kluwer Academic Publishers, Dordrecht, 1996,
pp. 137–164.
[4] a) K. N. Ferreira, T. M. Iverson, K. Maglaoui, J. Barber, S.
Iwata, Science 2004, 303, 1831–1838; b) B. Loll, J. Kern, W.
Saenger, A. Zouni, J. Biesiadka, Nature 2005, 438, 1040–1044.
[5] W. C. Hoganson, G. T. Babcock, Science 1997, 277, 1953–1956.
[6] a) B. Kok, B. Forbush, M. McGloin, Photochem. Photobiol.
1970, 11, 457–475; b) B. Forbush, B. Kok, M. McGloin, Pho-
tochem. Photobiol. 1971, 14, 307–321.
[20]
[21]
bpy = 2,2-bipyridine.
a) bhpp = 2,6-bis{[(2-hydroxy-3,5-di-tert-butylbenzyl)(2-pyr-
idylmethyl)amino]methyl}-4-methylphenol; b) R. Lomoth, P.
Huang, J. Zheng, L. Sun, L. Hammarström, B. Åkermark, S.
Styring, Eur. J. Inorg. Chem. 2002, 2965–2974.
C. Higushi, H. Sakiyama, H. Okawa, R. Isobe, D. E. Fenton,
J. Chem. Soc., Dalton Trans. 1994, 7, 1097–1103.
B. C. Dave, R. S. Czernuszewicz, Inorg. Chim. Acta 1998, 281,
25–35.
P. Karsten, A. Neves, A. J. Bortoluzzi, J. Strahle, C. Maichle-
[22]
[23]
[24]
[25]
Mossmer, Inorg. Chem. Commun. 2002, 5, 434–438.
a) L. Dubois, D.-F. Xiang, X.-S. Tan, J. Pécault, P. Jones, S.
Baudron, L. LePape, J.-M. Latour, C. Baffert, S. Chardon-No-
blat, M.-N. Collomb, A. Deronzier, Inorg. Chem. 2003, 42,
750–760; b) L. Dubois, R. Caspar, L. Jacquamet, P.-E. Petit,
M. F. Charlot, C. Baffert, M.-N. Collomb, A. Deronzier, J.-M.
Latour, Inorg. Chem. 2003, 42, 4817–4827.
a) C. Fraser, L. Johnston, A. L. Rheingold, B. S. Haggerty,
J. K. Williams, J. Whelan, B. Bosnich, Inorg. Chem. 1992, 31,
1835–1844; b) R. Hotzelmann, K. Wieghardt, U. Florke, H. J.
Haupt, D. C. Weatherburn, J. Bonvoisin, G. Blondin, J.-J. Gir-
erd, J. Am. Chem. Soc. 1992, 114, 1681–1696; c) P. Karsten, A.
Neves, A. J. Bortoluzzi, M. Lanznaster, V. Drago, Inorg. Chem.
2002, 41, 4624–4626.
[7] M. Haumann, P. Liebisch, C. Müller, M. Barra, M. Grabolle,
H. Dau, Science 2005, 310, 1019–1021.
[26]
[8] a) S. Mukhopadhyay, S. K. Mandal, S. Bhaduri, W. H. Arm-
strong, Chem. Rev. 2004, 104, 3981–4026; b) J. H. Robblee,
R. M. Cinco, V. K. Yachandra, BBA-Bioenergetics 2001, 1503,
7–23; c) H. Dau, L. Iuzzolino, J. Dittmer, BBA-Bioenergetics
2001, 1503, 24–39; d) D. Kuzek, R. J. Pace, BBA-Bioenergetics
2001, 1503, 123–137.
5046
www.eurjic.org
© 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Inorg. Chem. 2006, 5033–5047

Redox Chemistry of a Dimanganese(II,III) Complex
FULL PAPER
[27] E. Lambert, B. Chabut, S. Chardon-Noblat, A. Deronzier, G.
Chottard, A. Bousseksou, J. Tuchagues, J. Laugier, M. Bardet,
J. M. Latour, J. Am. Chem. Soc. 1997, 119, 9424–9437.
[28] M. Lubben, B. L. Feringa, J. Org. Chem. 1994, 59, 2227–2233.
[29] I. D. Brown, D. Altermatt, Acta Crystallogr., Sect. B: Struct.
Sci. 1985, 41, 244–247.
[30] a) R. M. Buchanan, K. J. Oberhausen, J. F. Richardson, Inorg.
Chem. 1988, 27, 971–973; b) H.-R. Chang, S. K. Larsen,
P. D. W. Boyd, C. G. Pierpont, D. N. Hendrickson, J. Am.
Chem. Soc. 1988, 110, 4565–4576.
[37] Electrolysis at 0.92 V appears to be a multielectron process pre-
sumably because of the oxidation of degradation products.
[38] R. A. Marcus, J. Chem. Phys. 1965, 43, 679–701.
[39] a) The outer reorganisation energy of the electrode reaction in
acetonitrile solutions (n = 1.344, ε = 35.94) was estimated from
λo = [e2/(8πεo)](1/r – 1/R)(1/n2 – 1/ε) where r ≈ 5 Å is the molec-
ular radius and R(Ն2r) is the distance between the molecule
and its image charge; b) A. J. Bard, L. R. Faulkner, in Electro-
chemical Methods: Fundamentals and Applications, 2nd ed.,
Wiley, New York, 2001.
[31] H. Diril, H.-R. Chang, X. Zhang, S. K. Larsen, J. A. Potenza,
C. G. Pierpont, H. J. Schugar, S. S. Isied, D. N. Hendrickson,
J. Am. Chem. Soc. 1987, 109, 6207–6208.
[40] M. L. A. Abrahamsson, H. Berglund Baudin, A. Tran, C. Phil-
ouze, K. E. Berg, M. K. Raymond-Johansson, L. Sun, B. Åker-
mark, S. Styring, L. Hammarström, Inorg. Chem. 2002, 41,
1534–1544.
[41] a) P. J. Pessiki, S. V. Khangulov, G. C. Dismukes, J. Am. Chem.
Soc. 1994, 116, 891–897; b) S. V. Khangulov, P. J. Pessiki, V. V.
Barynin, D. E. Ash, G. C. Dismukes, Biochemistry 1995, 34,
2015–2025; c) T. Howard, J. Telser, V. J. DeRose, Inorg. Chem.
2000, 39, 3379–3385; d) S. Blanchard, G. Blondin, E. Rivière,
M. Nierlich, J.-J. Girerd, Inorg. Chem. 2003, 42, 4568–4578.
[42] K.-O. Schäfer, R. Bittl, W. Zweygart, F. Lendzian, G. Hasel-
horst, T. Weyhermüller, K. Wieghardt, W. Lubitz, J. Am. Chem.
Soc. 1998, 120, 13104–13120.
[32] G. Aromi, J. Telser, A. Ozarowski, L.-C. Brunel, H.-M.
Stoeckli-Evans, J. Krzystek, Inorg. Chem. 2005, 44, 187–196.
[33] a) S. R. Cooper, G. C. Dismukes, M. P. Klein, M. Calvin, J.
Am. Chem. Soc. 1978, 100, 7248–7252; b) M. Zheng, S. V.
Khangulov, G. C. Dismukes, V. V. Barynin, Inorg. Chem. 1994,
33, 382–387; c) J. S. Bashkin, A. R. Schake, J. B. Vincent, H.-
R. Chang, Q. Li, J. C. Huffman, G. Christou, D. N. Hendrick-
son, J. Chem. Soc., Chem. Commun. 1988, 11, 700–702; d) G. C.
Dismukes, J. E. Sheats, J. A. Smegal, J. Am. Chem. Soc. 1987,
109, 7202–7203.
[34] P. Huang, N. Shaikh, M. Anderlund, S. Styring, L. Ham-
marström, J. Inorg. Biochem. 2006, 100, 1139–1146.
[43] Formation of colloidal MnO2 was not observed by visual in-
spection but cannot be excluded.
[44] a) J. E. Sheats, R. S. Czernuszewitcz, G. C. Dismukes, A. L.
Rheingold, V. Petrouleas, J.-A. Stubbe, W. H. Armstrong, R. H.
Beer, S. J. Lippard, J. Am. Chem. Soc. 1987, 109, 1435–1444; b)
E. M. Boelrijk, S. V. Khangulov, C. G. Dismukes, Inorg. Chem.
2000, 39, 3009–3019.
[35] In ref.[34] the spin Hamiltonian H = β S·g·B + Î1·Â1·S + Î2·Â2·S
ˆ
ˆ
ˆ
ˆ
was used, resulting in the hyperfine parameters A1,x
=
171ϫ10–4 cm–1, A1,y = 148ϫ10–4 cm–1, A1,z = 261ϫ10–4 cm–1
and A2,x = 55ϫ10–4 cm–1, A2,y = 90ϫ10–4 cm–1, A2,z
122ϫ10–4 cm–1 for manganese 1 and 2 respectively.
=
[36] a) This is based on the potential of the [Ru(bpy)3]3/2+ couple
that is considered to be largely solvent independent and
amounts to 0.90 V versus Fc+/0 in acetonitrile (1.27 V vs. NHE
in water). Potentials in nonaqueous solvents are frequently
measured and reported versus the aqueous SCE. This intro-
duces unknown junction potentials, however, and a compari-
son of such data with potentials obtained in other solvents like
water is not meaningful even if the potentials are formally given
versus the same reference electrode like SCE or NHE. Indepen-
dent of this problem, electrochemical potentials are intrin-
sically solvent dependent and any comparison between data in
different solvents can only be approximate; b) I. Rubinstein,
A. J. Bard, J. Am. Chem. Soc. 1981, 103, 512–516.
[45] J. Limburg, J. S. Vrettos, H. Chen, J. C. de Paula, R. H.
Crabtree, G. W. Brudvig, J. Am. Chem. Soc. 2001, 123, 423–
430.
[46] G. M. Sheldrick, Acta Crystallogr., Sect. A: Found. Crystallogr.
1990, 46, 467–473.
[47] G. M. Sheldrick, Computer Program for the Refinement of
Crystal Structures, University of Göttingen, 1997.
[48] R. Manchanda, G. W. Brudvig, R. H. Crabtree, New J. Chem.
1994, 18, 561–568.
Received: July 19, 2006
Published Online: October 27, 2006
Eur. J. Inorg. Chem. 2006, 5033–5047
© 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org
5047