Too Many Cooks Spoil the Broth - Variable Potencies of Oxidizing Mn Complexes of a Hexadentate Carboxylato Ligand

Article abstract of DOI:10.1002/ejic.201500210

Given the hexadenticity of the monoanionic ligand in the procatalyst [Mn(tpena)(H₂O)](ClO₄) {tpena^- = N,N,N′-tris(2-pyridylmethyl)ethylenediamine-N′-acetate}, it is perhaps surprising that this complex can catalyze the epoxidation of alkenes. When peracetic acid is used as terminal oxidant, the selectivity and rates of reactions are comparable with those reported for the manganese complexes of the commonly employed neutral tetradentate N4 ligands under analogous conditions. Cyclooctene conversion rates are similar when tert-butyl hydroperoxide (TBHP) is used; however, the selectivity is greatly diminished. In the absence of organic substrates, [Mn^II(tpena)]⁺ catalyzes water oxidation by TBHP (initial rate ca. 23 mmol/h when [Mn] = 0.1 mM, at room temp.). To explain the variations in the selectivity of catalytic epoxidations and the observation of competing water oxidation, we propose that several metal-based oxidants (the "cooks") can be generated from [Mn^II(tpena)]⁺. These embody different potencies. The most powerful, and hence least selective, is proposed to be the isobaric isomer of [Mn^IV₂(O)₂(tpena)₂]²⁺, namely an oxylic radical complex, [(tpena)Mn^III(μ₂-O)Mn^IV(O^·)(tpena)]²⁺. The formation of this species depends on the catalyst concentration, and it is favoured when TBHP is used as the terminal oxidant. The generation of the less potent [Mn^IV(O)(tpena)]⁺, which we propose as the direct oxidant in epoxidation reactions, is favoured in non-aqueous solutions when peracetic acid is used as the terminal oxidant. A manganese(II) complex of a hexadentate ligand catalyzes epoxidation. However, reaction conditions are critical since several metal-based oxidants with varying potencies can be generated. This observation is probably general for related catalysts, and consequences include decreased selectivity and competing water oxidation.

Full text of DOI:10.1002/ejic.201500210

Job/Unit: I50210
/KAP1
Date: 28-04-15 16:07:13
Pages: 8
FULL PAPER
DOI:10.1002/ejic.201500210
Too Many Cooks Spoil the Broth – Variable Potencies of
Oxidizing Mn Complexes of a Hexadentate Carboxylato
Ligand
[
a]
[a]
[a]
Claire Deville, Maik Finsel, David P. de Sousa,
[a]
[a]
[a]
Barbara Szafranowska, Julian Behnken, Simon Svane,
Andrew D. Bond,[
a][‡]
Rune Kirk Seidler-Egdal, and
[a]
Christine J. McKenzie*[
a]
Keywords: Manganese / Epoxidation / Water oxidation / Carboxylate ligands / Hexadentate ligands
Given the hexadenticity of the monoanionic ligand in the
procatalyst [Mn(tpena)(H O)](ClO ) {tpena = N,N,NЈ-tris(2-
2 4
plain the variations in the selectivity of catalytic epoxidations
and the observation of competing water oxidation, we pro-
pose that several metal-based oxidants (the “cooks”) can be
generated from [Mn (tpena)] . These embody different po-
tencies. The most powerful, and hence least selective, is pro-
–
pyridylmethyl)ethylenediamine-NЈ-acetate}, it is perhaps
surprising that this complex can catalyze the epoxidation of
alkenes. When peracetic acid is used as terminal oxidant, the
selectivity and rates of reactions are comparable with those
reported for the manganese complexes of the commonly em-
ployed neutral tetradentate N4 ligands under analogous con-
ditions. Cyclooctene conversion rates are similar when tert-
butyl hydroperoxide (TBHP) is used; however, the selectivity
is greatly diminished. In the absence of organic substrates,
II
+
IV
2+
posed to be the isobaric isomer of [Mn
2 2 2
(O) (tpena) ] ,
III
namely an oxylic radical complex, [(tpena)Mn (μ
2
-O)-
IV
·
2+
Mn (O )(tpena)] . The formation of this species depends on
the catalyst concentration, and it is favoured when TBHP is
used as the terminal oxidant. The generation of the less po-
tent [Mn (O)(tpena)] , which we propose as the direct oxi-
dant in epoxidation reactions, is favoured in non-aqueous
solutions when peracetic acid is used as the terminal oxidant.
IV
+
II
+
[
Mn (tpena)] catalyzes water oxidation by TBHP (initial rate
ca. 23 mmol/h when [Mn] = 0.1 mM, at room temp.). To ex-
Introduction
can be envisaged in the usually requisite accompanying pro-
ton transfers. Coordination flexibility for carboxylato li-
One or more carboxylate donors (Glu, Asp) are a ubiqui-
tous presence in the first coordination spheres of most non-
1
2
gands in terms of monodentate, bidentate, μ -η -, μ -η -,
2
2
1
2
2 2
μ -η η -, μ -η η -bridging coordination modes is well
3
4
heme metalloenzymes capable of the O activation needed
2
known, and “carboxylate shifts” between some of these ex-
tremes are assumed to be important for the mechanisms of
metalloenzymes. Finally, transient percarboxylato functions
in biological catalytic oxidations.[ This is also the case for
1]
[
2]
the manganese-containing enzymes superoxide dismutase,
oxalate decarboxylase[ and the oxygen evolving centre
3]
[4]
generated by insertion of an O -derived oxygen atom into
2
where O evolution occurs. It seems therefore obvious to
2
M–Asp/Glu bonds are worth appraising as metal-based oxi-
dants for potential in vivo reactions. Speculations concern-
ing these species can be extended to the possibility that par-
allel mechanisms are at play when peracetic acid is em-
ployed as the terminal oxidant. Although it is not explicit
in much of the literature, the reason for the extensive use of
this oxidant in catalytic oxidations is likely to stem from
observations of comparatively better selectivity. It is plaus-
ible, but perhaps unconsidered, that improved selectivity
explore the use of supporting ligands that furnish mono-
dentate carboxylato donors in the context of discovering
efficient catalysts for biomimetic catalytic oxidation reac-
tions. Anionic weak-field carboxylato ligands are expected
to be suitable for accessing high-valent metal states, and
active roles involving non-coordinated carbonyl O atoms
[
a] Department of Physics, Chemistry and Pharmacy,
University of Southern Denmark,
over, for example, reactions using H O , might be due to
Campusvej 55, 5230 Odense M, Denmark
2
2
E-mail: mckenzie@sdu.dk
an active role of acetic acid. Could it be that in Nature
these groups can generate an “internal peracetic acid” con-
comitant with the activation or, conversely, with release of
O2?
Perhaps surprisingly, and arguably because of synthetic
bottlenecks, the manganese complexes of chelating ligands
http://www.sdu.dk/Om_SDU/Institutter_centre/fysik_kemi_og_
farmaci/Forskning/Forskningsgrupper/Christine_McKenzie
‡] Current address: Department of Chemistry,
University of Cambridge,
[
Lensfield Road, CB2 1 EW, United Kingdom
Supporting information for this article is available on the
WWW under http://dx.doi.org/10.1002/ejic.201500210.
Eur. J. Inorg. Chem. 0000, 0–0
1
© 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Job/Unit: I50210
/KAP1
Date: 28-04-15 16:07:13
Pages: 8
www.eurjic.org
FULL PAPER
furnishing monodentate carboxylato ligands have not been zation of a colourless air-stable manganese(II) complex,
as extensively examined as systems based on neutral amino- [Mn(tpena)(H O)]ClO ([1]ClO ). The ESI mass spectrum
2
4
4
pyridyl chelating ligands.[ Manganese complexes of neu- of this material shows that the [Mn (tpena)] ion at m/z
tral N4 tetradentate ligands are effective as catalysts in 445.13 is dominant in an otherwise clean spectrum. There
homogeneous epoxidation.[ The complexes of N,NЈ-bis(2- are no traces of ions with a higher metal oxidation state.
pyridylmethyl)ethylenediamine (bpmen)[ (Scheme 1) are The structure of the intercationic hydrogen-bonded pairs in
representative. In recent years, we have modified the ethyl- [1]ClO ·4.5H O, determined by single-crystal X-ray analy-
5]
II
+
6]
7]
4
2
enediamine scaffold of bpmen by the inclusion of carboxyl- sis, is shown in Figure 1. Selected parameters are listed in
ate groups in efforts towards the discovery of improved cat- Table 1. The manganese(II) atom is seven-coordinate, sim-
alysts. We discovered that the manganese complex of the ilar to the manganese(II) ions in complexes of pentadentate
–
[10c]
[11]
[10d]
pentadentate ligand N-methyl-N,NЈ-bis(2-pyridylmethyl)- mepena ,
and hexadentate edta
and tpen.
Seven-
–
ethylenediamine-NЈ-acetate (mepena ) and an analogue coordination is known to stabilize the +2 oxidation state of
containing a benzyl group in the place of the methyl group the manganese centre,[10e] and this structure will presumably
are sufficiently powerful to catalyze the oxidation of water prevent spontaneous oxidation of 1 in both solid and solu-
using tert-butyl hydroperoxide (TBHP) as the terminal oxi- tion states. [1]ClO ·4.5H O crystallizes in the Sohncke space
4
2
[
8]
[12]
dant. A 2-methylpyridyl arm in place of the methyl group group
P2 , and the same enantiomeric form in each
1
gives the hexadentate ligand N,N,NЈ-tris(2-pyridylmethyl)- monomer implies that it undergoes spontaneous resolu-
–
[13]
ethylenediamine-NЈ-acetate (tpena , Scheme 1). Using the tion.
The crystallization of [1]ClO ·4.5H O therefore
4 2
–
iron complex of tpena , we have achieved rapid and selec- constitutes a rare example where a seven-coordinate com-
tive catalysis of sulfoxidation by iodosylbenzene reactions.
[
9]
[14]
plex is isolated as a pure enantiomer. The two molecules
Since “coordinative unsaturation” is commonly perceived
as desirable for the construction of catalysts, this latter reac-
tion might seem surprising given the potential hexadenticity
–
of tpena . More commonly, procatalysts are based on che-
lating ligands with lower donor numbers (mostly four N
donors). Important roles as auxiliary ligands by solvents
are often ignored, and an increasing reliance on ESI mass
spectrometry for the identification of solution-based species
may have reinforced a misconception that solvent adducts,
which usually do not survive ionization, are irrelevant. Fur-
5
III
II
ther seven-coordination, especially for d Fe and Mn
and N/O donor sets, is reasonable.[
10]
Figure 1. The H-bonded [Mn(tpena)(H
asymmetric unit of [1]ClO ·4.5H O. All H atoms, except those at
O)]+ (1) cations in the
2
4
2
the aqua ligand are omitted for clarity. Hydrogen bonds are indi-
cated by dashed lines. Displacement ellipsoids are drawn at the
50% probability level.
Scheme 1. N4, N4O and N5O ligands based on an ethylenediamine
backbone.
Table 1. Average selected distances (Å) and bond angles (°) for
1]ClO ·4.5H O.
The use of chelating ligands with relatively high donor
[
4
2
–
numbers like the N4O and N5O ligands mepena and
–
Mn1–O1
Mn1–O3
Mn1–N1
Mn1–N2
2.2903(19)
2.1784(19)
2.394(2)
Mn1–N3
Mn1–N4
Mn1–N5
Mn1–Mn2
2.299(2)
2.394(2)
2.272(3)
5.482(2)
tpena , will limit the number of solution-state solvent-de-
rived species, potentially simplifying metal catalyst specia-
tion in reaction mixtures. Our studies show that this is not
counter-beneficiary, since the catalysis of oxidation reac-
tions is not blocked. Furthermore the less complicated solu-
tion-state metal-ion speciation has allowed for consider-
ation of the reaction mechanism.
2.399(2)
O1–Mn1–N3
O1–Mn1–N1
O1–Mn1–N4
O1–Mn1–N2
O3–Mn1–N5
O3–Mn1–O1
O3–Mn1–N3
O3–Mn1–N4
98.54(8)
129.21(8)
160.39(8)
68.61(8)
93.97(8)
78.85(7)
92.25(8)
81.56(8)
145.09(8)
141.14(8)
68.76(8)
N1–Mn1–N2
N3–Mn1–N1
N3–Mn1–N2
N3–Mn1–N4
N4–Mn1–N2
N5–Mn1–O1
N5–Mn1–N3
N5–Mn1–N1
N5–Mn1–N4
N5–Mn1–N2
73.78(9)
101.69(9)
72.86(9)
83.16 (9)
129.77(9)
83.51(8)
173.69(9)
72.63(9)
96.93(9)
102.60(9)
Results and Discussion
The reaction of manganese(II) perchlorate with O3–Mn1–N1
O3–Mn1–N2
N1–Mn1–N4
N,N,NЈ-tris(2-pyridylmethyl)ethylenediamine-NЈ-acetic acid
tpenaH) in a water/methanol mixture results in crystalli-
(
Eur. J. Inorg. Chem. 0000, 0–0
2
© 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Job/Unit: I50210
/KAP1
Date: 28-04-15 16:07:13
Pages: 8
www.eurjic.org
FULL PAPER
–
in the asymmetric unit have indistinguishable conforma-
We can exclude oxidative breakdown of tpena with for-
[
15]
tions and are associated by O–H···O interactions involv- mation of catalytically competent Mn–picolinate com-
ing the aqua ligands and the coordinated oxygen atoms of plexes during the reactions with both peracetic acid and
[
16]
the carboxylate groups with O···O distances of 2.646(3) and TBHP on the basis of the facts that the reactions start with-
2
.660(3) Å. These interactions give rise to a six-membered out any lag time and the post-catalysis ESI mass spectra of
II
+
ring including the two manganese atoms and containing reaction mixtures still show [Mn (tpena)] at m/z 445.13 as
two H-bonds, with a chiral twisted-boat conformation. The a dominant ion. No ions attributable to oxidative decompo-
structure comprises stacks of cations parallel to the a- and sition of the ligand can be discerned. More specifically, no
–
b-axes respectively, separated by ClO4 anions and H O peaks are found in common with the ESI mass spectrum of
2
molecules (Figure S1).
a mixture of manganese perchlorate and picolinic acid
II
Complex 1 catalyzes the epoxidation of alkenes by per- where Mn–picolinato complexes, for example [Mn (pic) +
acetic acid (PAA). Conversion and selectivities are listed in H] at m/z 299.99, and other derivatives are detected (Fig-
2
+
Table 2. No diol byproducts are detected. Catalytic activity ure S3).
drops with decreasing catalyst loadings, and at 0.15% the
ESI mass spectra of mixtures of 1 and 100 equiv. of
conversion and yield can be attributed to the background TBHP or peracetic acid recorded within 3 min of mixing
IV
reaction. Since bpmen (Scheme 1) possesses the same ethyl- are shown in Figure 2. A minor ion assignable to [Mn O-
+
enediamine backbone, for comparison, catalysis experi- (tpena)] at m/z 461.1246 can be observed in both spectra;
ments were repeated under the same conditions with the however, the proportion of this ion compared to that of
II
+
procatalyst [Mn (bpmen) (μ -CH CO ) ](ClO ) ([2]2ClO ) [Mn (tpena)] at m/z 445.1303 is much lower in the spec-
2
2
2
3
2 2
4 2
4
II
as an N4 representative. The conversions for catalytic oxid- trum of the solutions containing peracetic acid. [Mn -
ation of cyclooctene using PAA as terminal oxidant are not (tpena)] is the base peak in this latter spectrum, and a
strikingly different for the two systems, except that the selec- minor peak corresponding to [Mn (tpena)(HOAc)] is ob-
tivity in the reactions catalyzed by 1 is significantly higher served at m/z 505.1511. By contrast, the [Mn (tpena)] ion
+
II
+
II
+
under more dilute conditions (i.e. the concentration of all of is far less intense in the spectrum obtained with TBHP,
III
+
the catalyst, oxidant and substrate are decreased, but they where the base peak corresponds to [Mn (OH)(tpena)]
remain in the same proportion). Conversion rates for the at m/z 462.1322. The second highest peak corresponds to
III
IV
+
catalytic oxidation of cyclooctene using tert-butyl hydroper- [(tpena)Mn (O) Mn (tpena)] at m/z 922.2508. An ex-
2
oxide (TBHP) as the terminal oxidant were similar; how- pansion of the region m/z 460–462 shows that the peak for
ever, a significant loss of selectivity was observed, and sev- a doubly charged ion overlaps that for the singly charged
IV
+
eral unidentified products appear on the GC chromatogram [Mn O(tpena)] (m/z 461.1246, Figure S4). This is assigned
IV
IV
2+
(
Figure S2). Differences, over acceptable experimental loss, to the dimeric species [(tpena)Mn (O) Mn (tpena)] . A
2
IV
IV
2+
between alkene conversion and epoxide yield can be trans- favourable reduction of [(tpena)Mn (O) Mn (tpena)]
2
lated into a measure of selectivity; the smaller the differ- during ionization is a reasonable path for the formation of
III
IV
+
ence, the greater the selectivity. The Mn complex of the [(tpena)Mn (O) Mn (tpena)] {the opposite reaction,
2
III
IV
+
bpmen analogue N,NЈ-dimethyl-N,NЈ-bis(2-pyridylmethyl)- oxidation of [(tpena)Mn (O) Mn (tpena)] , is not ex-
2
(1R,2R)-(+)-cyclohexane-1,2-diamine has been reported to pected under MS conditions}. No trace of any oxo-bridged
give better results for epoxidation reactions in terms of cat- dinuclear manganese species is observed in the spectra with
alyst loadings (down to 0.1%), conversion and selectivity PAA.
compared to those we see with both 1 and 2.[ One could
7]
Bubbles can occasionally be seen in mixtures of 1 and
speculate that the non-sterically restricted ethylenediamine peracetic acid or TBHP in the presence of water. This sug-
backbones of 1 and 2 compared with the rigidifying cyclo- gests that water oxidation might be a competing reaction.
hexane backbone make the systems floppier and that this For this reason, 1 was tested as a water-oxidation catalyst.
may be one reason for their slightly lower efficiencies.
This was carried out in the absence of organic substrates
Table 2. Conversion and selectivity in the catalytic oxidation of alkenes. The catalyst loading was 1.5 mol-% with respect to substrate in
all reactions.
Substrate
Procatalyst [Mn]
Substrate concentration
(mm)
PAA concentration Conversion Selectivity
Byproducts
mm)[
a]
(mm)
(%)
(%)
(
Cyclooctene
Cyclooctene
Cyclooctene
Cyclooctene
Decene
Decene
cis-Stilbene
cis-Stilbene
[1]ClO
[2]2ClO
[1]ClO
[2]2ClO
[1]ClO
[2]2ClO
[1]ClO
[2]2ClO
4
3
3
0.3
0.3
3
3
3
3
200
200
20
280
280
28
96
100
98
89
84
84
30
31
75
96
92
84
80
83
80
86
not detected
not detected
not detected
not detected
not detected
not detected
trans-epoxide
trans-epoxide
4
4
4
4
4
20
28
4
200
200
200
200
280
280
280
280
4
[
a] In terms of [Mn] concentration. Dry acetonitrile, 0 °C (cyclooctene) and room temp. (decene and stilbene); reaction quenched after
one (cyclooctene and decene) and three (stilbene) hours with aluminium oxide.
Eur. J. Inorg. Chem. 0000, 0–0
3
© 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Job/Unit: I50210
/KAP1
Date: 28-04-15 16:07:13
Pages: 8
www.eurjic.org
FULL PAPER
4
Figure 2. (a) The ESI mass spectra of [1]ClO . (b) After addition of 100 equiv. of peracetic acid. (c) After addition of 100 equiv. of TBHP.
II
+
II
+
IV
+
IV
Assignments: m/z 445.1303 [Mn (tpena)] ; m/z 505.1511 [Mn (tpena)(HOAc)] ; m/z 461.1246 [Mn O(tpena)] and [(tpena)Mn -
IV
2+
III
+
IV
III
+
(O)
2
Mn (tpena)] , see text; m/z 462.1322 [Mn (OH)(tpena)] ; m/z 922.2508 [(tpena)Mn (O)
2
Mn (tpena)] . Solvent: acetonitrile/water
(9:1).
with TBHP as terminal oxidant under conditions similar
to those implemented for the catalysis of water oxidation
+
[8]
reactions using [Mn(mepena)] .
Figure 3 shows the
amount of dioxygen evolved as a function of time in the
absence of potential organic substrates, including solvent.
For comparison, the same experiment was performed using
[
(
Mn (mepena) (H O) ](ClO )
([3]2ClO4) and [Mn2-
bpmen) (O) ](ClO ) ([4]3ClO ) as the procatalysts. It was
2
2
2
2
4 2
2
2
4 3
4
interesting to observe that even the bpmen system catalyzes
water oxidation; however, the complexes with the carboxyl-
–
–
ato N4O and N5O ligands, mepena and tpena , show
higher activity. As mentioned, we have previously reported
3
as a catalyst for this reaction, and this system shows the
highest activity in the series reported here. A disproportion-
18
ation of TBHP is not expected, and was eliminated, by O-
labelling experiments using 3.[ In these experiments, we
also ascertained that one O atom from water and one from
8]
Figure 3. Amount of dioxygen evolved as function of time. [TBHP]
II
the TBHP ends in the product O . We have not performed = 2.56 m; [Mn] = 0.1 mm; (᭹) [Mn (tpena)(H O)]ClO , (1); (᭡)
2
2
4
III/IV
[
(
Mn
ClO
2
(mepena)
2
(H
2
O)
2
](ClO
4
)
2
(3); (♦) [Mn(
)
2
2 2
(bpmen) (O) ]-
these labelling experiments for reactions using 1 and 4;
however, it seems reasonable to assume that this would also
4 3
) (4). TON after 70 min: 3600 (᭹), 7400 (᭡) and 800 (♦).
be the case. The TONs (per [Mn]) measured after 70 min at
room temp. are 3600, 7400 and 800, with initial rates of 2, based oxidant in alkene epoxidation. The observation of
III
IV
+
2
3 and 0.5 mmol/h, for 1, 3 and 4 respectively.
ions assignable to [(tpena)Mn (O) Mn (tpena)] and
2
IV
IV
2+
On the basis of a catalyst concentration dependence in [(tpena)Mn (O) Mn (tpena)] in ESI mass spectra sug-
the rate of O evolution, the detection of Mn , Mn OH, gest that, despite the potential hexadenticity of tpena , oxo-
Mn O and dinuclear Mn(O) Mn species, and DFT calcu- bridged dinuclear species are present in solutions. On the
2
II
III
–
2
IV
2
lations, we have proposed a mechanism involving both mo- basis of our experience with both the N4O and N5O sys-
nonuclear and dinuclear Mn species in the catalysis of water tems, these appear in somewhat lower abundances com-
[
8a,10c]
Crucially, we suggested that a dinu- pared to the pentadentate mepena– system.
[10c]
oxidation by 3.
A lower
III
IV
IV
IV
2+
clear oxylic radical species [(mepena)Mn (μ -O)Mn - steady-state concentration of [(L)Mn (O) Mn (L)]
2
2
·
2+
(
O )(mepena)] was the oxidant capable of attacking water would seem then to be the likely explanation for the lower
in the critical O–O bond formation step. Significantly also, TONs for water oxidation by TBHP for the N5O system
IV
the calculations inferred that the mononuclear Mn O spe- relative to that of the N4O system. This situation is quite
cies were not capable of oxidizing water. Such a species may, different for the neutral N4 system, where the dioxo-
III
IV
3+
however, be sufficiently potent to oxidize weaker C–H/C–C bridged ([(L)Mn (O) Mn (L)] ) (4, L = bpmen) and
2
IV
IV
bonds and hence be capable of acting as a direct metal- homologues of its oxidized congener [(L)Mn (O) Mn -
2
Eur. J. Inorg. Chem. 0000, 0–0
4
© 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Job/Unit: I50210
/KAP1
Date: 28-04-15 16:07:13
Pages: 8
www.eurjic.org
FULL PAPER
L)]4+ can be isolated.[17] Isolation obviously suggests that steady-state concentrations of [Mn O(L)] are lower in
IV
n+
(
the oxo-bridged Mn (III/IV) and Mn (IV/IV) dimers of the peracetic acid reactions. Consequently, its condensation
2
2
II
IV
bpmen are less reactive (in other words, as we shall see, less with Mn and Mn species (steps v and vi) will be far less
readily form oxylic radical complexes). In the presence of favoured compared to mixtures with TBHP and no acetic
II
oxidants, the Mn complexes of bpmen cannot be detected, acid. Conversely, the steady-state concentration of
IV
n+
let alone isolated. This is, as mentioned, not the case for [Mn O(L)] will be significantly higher in the reactions
the N4O and N5O systems. using TBHP as terminal oxidant, thereby favouring forma-
By taking into account the observations reported here tion of [(L)Mn (μ -O)Mn (L)]
III
(2n+1)+
III
III
2n+
IV
or [(L)Mn (μ -
2
2
IV
IV
·
(2n+1)+
and previous results, we have set up a general mechanism O) Mn (L)]
↔ [(L)Mn (μ -O)Mn (O )(L)]
2
2
II
n+
for competing alkene epoxidation and water oxidation in by its reaction with [Mn (L)] or another molecule of
IV
n+
–
Scheme 2. Mononuclear and oxo-bridged dinuclear species [Mn O(L)] , respectively. In the case of tpena as the sup-
II
can be generated in solution with the ligands in Scheme 1. porting ligand, the Mn –oxidant adduct complexes,
II
n+
The concentrations of the various species will depend on [Mn (H)OOR(L)] , R = –CMe , –C(O)CH , are presum-
3
3
the concentration of water, the presence of organic substrate ably seven-coordinate, similar to the parent procatalyst 1.
and/or solvents, the nature of the terminal oxidant and the Decoordination of pyridine donors is likely in the structures
4
3
III
IV
number and type of donor atoms in the spectator ligands of the higher-valent d and d , Mn and Mn centres.
bound to manganese. A reasonable assumption might be This is entirely feasible, and several complexes in which the
that any (hydr)oxo species with oxidation state of +3 or pyridyl arm of multidentate aminopyridine ligands are de-
II
higher is potentially oxidizing, and Mn –oxidant adducts coordinated have been crystallographically characterized,
–
[9,18]
should not be excluded as potential direct oxidants. Direct including some with tpena .
oxidizing species for the alkenes could be both the mononu- philic dangling pyridine groups should in fact assist the
The presence of nucleo-
IV
n+
clear manganese(IV)oxo species [Mn O(L)] (step iii) and isomerization (step viii) towards the reactive oxylic radical
(step vii). form, furnishing a rationale for why water oxidation is fa-
Complex [Mn O(L)] is generated by heterolytic O–O voured for the N4O and N5O ligands.
III
IV
·
n+
the dinuclear [(L)Mn (μ -O)Mn (O )(L)]
2
IV
II
n+
n+
cleavage in [Mn (H)OOR(L)] (steps i and ii); however, in
Scheme 2 is entirely consistent with the facts that: (1)
the presence of excess acetic acid, as is the case for the reac- Epoxidation reactions using peracetic acid as terminal oxi-
tions in which peracetic acid is terminal oxidant, step i dant show the highest selectivity for all ligand systems we
shows greater reversibility than is possible for step ii. Thus, have tested; (2) while conversions are similar, the reactions
Scheme 2. Competing mechanisms for catalytic epoxidation of alkenes and oxidation of water by peracetic acid or TBHP as the terminal
–
–
oxidant showing possible Mn-based oxidants (in bold). L = tpena , mepena , bpmen or other similar ligand. Complex charges are not
drawn, since these will be dependent on the ligand, but they are related as indicated in the text. The species with bold lettering are
II
II
II
potential oxidants. A reaction arrow is not drawn between the Mn –peroxyacid or Mn –alkylperoxide complexes and the Mn –epoxide,
so that the scheme is not overcrowded. RH and ROH are general representations for an organic substrate and its oxidized product(s).
The latter does not need to be an alcohol (e.g. ketones, C–C cleavage products, etc. can be produced).
Eur. J. Inorg. Chem. 0000, 0–0
5
© 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Job/Unit: I50210
/KAP1
Date: 28-04-15 16:07:13
Pages: 8
www.eurjic.org
FULL PAPER
using TBHP as terminal oxidant are less selective; (3) scheme and the fact that the formation of stable dioxo
III/IV
decreasing the concentration of all species (catalyst and bridged Mn
substrates) improves selectivity in epoxidation (although conditions.
rates are expected to decrease to the point where the uncat-
species will be less favoured under these
alyzed reaction dominates); (4) incorporation of a mono-
dentate carboxylate donor into the ligand does not appear Experimental Section
to alter efficiencies in any obvious way; (5) if it is not re-
Elemental analysis was performed at the Chemistry Department at
moved completely from reaction mixtures, water oxidation
can be a competing reaction; this, however, may be compen-
sated for by the judicial consideration of the equilibria in
Scheme 2.
Copenhagen University. ATR-IR spectra were recorded as neat sol-
ids with a PerkinElmer Spectrum Two spectrometer. All spectra
have been ATR- and baseline-corrected. UV/Vis spectra were re-
corded with an Agilent 8453 spectrophotometer using 1 cm quartz
cuvettes. ESI mass spectra were recorded with a microspray LC-
MS Bruker micrOTOF-Q II or for high resolution MS with a nano-
spray ThermoFisher Orbitrap XL mass spectrometer at a resolu-
tion greater than 10000. X-ray diffraction data were collected at
Conclusions
1
80(2) K with a Bruker–Nonius X8 APEX-II instrument (Mo-K
α
A common assumption in catalyst design is that “vacant
sites” must be available for the coordination of substrates.
In this work, we have shown that even a potentially hexa-
dentate ligand does not quench catalysis of alkene epoxid-
ation or water oxidation by its Mn complexes. Another hy-
pothesis that we wanted to test was whether the presence
of a monodentate carboxylato arm might play the role of
an “internal peracetic acid” by allowing insertion of an O
atom of an oxidant-derived O atom between itself and the
manganese atom. We have found no evidence that this is
radiation). Structure solution and refinement were carried out by
using SHELXTL.[
19]
N,N,NЈ-tris(2-pyridylmethyl)ethylenediamine-NЈ-acetic
acid
tpenaH),[
Mn (mepena)
bpmen) ](ClO
20]
[Mn
(H O)
([4]3ClO
(bpmen)
(CH
CO
)
](ClO
)
([2]2ClO
and [Mn
),
[21]
(
[
(
2
2
3
2
2
4
2
4
[
10c]
2
2
2
2
](ClO
4
)
2
17]
([3]2ClO
4
)
2
2
(O) -
)
3
4
)[ were prepared according to pre-
2
4
viously reported procedures. Peracetic acid solution (39% in acetic
acid), tert-butyl hydroperoxide solution (5.5 m in decane) over mo-
lecular sieves 4 Å (catalysis experiments) and T-HYDRO solution
(
70 wt.-% in water for ESI-MS experiments) were purchased from
the case and can thus not infer any special role, over that Sigma–Aldrich and used without further purification. Acetonitrile
of spectator, for a carboxylate donor. We propose, however, was dried with activated 3 Å molecular sieves.
that under certain conditions the carboxylate-containing
4 2 2
Mn(ClO ) ·6H O
[
(
Mn(tpena)(OH
2
)]ClO
4
·4.5H
2
O
([1]ClO
4
):
N4O and N5O ligands favour formation of reactive diman-
502 mg, 1.4 mmol) was added to tpenaH (543 mg, 1.3 mmol) in
ganese oxylic radical species that are too oxidizing to methanol (6 mL) and water (5 mL). Slow evaporation of the sol-
achieve a good selectivity towards the desired product in vents afforded the product as a beige microcrystalline precipitate
alkene epoxidations. They can even oxidize water. It is, (98 mg, 12%), which was characterized by elemental analysis, IR
spectroscopy and ESI-MS. Overnight, a few colourless crystals of
however, possible to devise conditions for avoiding this
competing reaction.
X-ray quality were deposited in the filtrate. ESI-MS (positive mode,
+
CH
Mn(tpena)] . IR (KBr): ν˜ = 1605 (C=O, s), 1092 (ClO
32ClMnN 10 ([Mn(tpena)(OH )]ClO ·3H O) (616.91): calcd.
3
OH): m/z (%) = 545.1(3) (100) [Mn(tpena)]HClO
4
, 445.1
While the literature tends to focus on the choice of sup-
porting ligands, and in limiting cases this will be important,
we draw the conclusion that the choice of terminal oxidant,
rate of addition, catalyst concentration and the presence or
absence of water can be critical, but perhaps this has hith-
erto been less appreciated. The comparison of the results of
catalysis experiments from different labs is always difficult,
and this problem is only exacerbated if workers are not
aware of potential pitfalls like competing water oxidation.
These can be indicated by significant discrepancies in mass
balance and conversions that appear high compared to
product and by-product yields determined by traditional
methods like GC. Dioxygen is usually not specifically moni-
+
–
–1
[
C
4
, bs) cm .
22
H
5
O
2
4
2
C 42.83, H 5.23, N 11.35; found C 42.49, H 4.49, N 10.85.
CCDC-1050773 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.
Epoxidation of Cyclooctene: The reactions were performed in dry
acetonitrile in a total volume of 2.5 mL at 0 °C. Catalyst ([Mn] =
3
mm, 1.5 mol-%), substrate (200 mm) and oxidant (total concen-
tration 280 mm) were added over 5 min with a syringe-pump. The
reaction was quenched after one hour with aluminium oxide, and
the products were analyzed by GC with biphenyl as internal stan-
tored (and this is difficult), and these discrepancies appear dard. For the reactions run in diluted conditions, a similar mixture
often to be ignored. It is possible that Scheme 2 is relatively was prepared and diluted to a total volume of 25 mL.
general and the position of the various equilibria will, how-
ever, depend on the choice of capping ligand. While under
the right conditions the penta- and hexadentate N4O and
Epoxidation of Decene: The reactions were performed in dry aceto-
nitrile in a total volume of 2.5 mL at room temperature. Catalyst
(
[Mn] = 3 mm,1.5 mol-%), substrate (200 mm) and oxidant (total
N5O ligands can more easily promote oxo-bridge cleavage concentration 280 mm) were added over 50 min with a syringe-
IV
+
to form more selective oxidants [Mn O(L)] , they can also pump. The reaction was quenched after one hour with aluminium
generate the highly oxidizing dimanganese(III,IV) oxylic oxide, and the products were analyzed by GC with biphenyl as
internal standard.
radical species. That low catalyst loadings down to 0.1% or
even less have been reported without loss of activity for Epoxidation of cis-Stilbene: The reactions were performed in dry
the commonly used N4 catalysts is also consistent with our acetonitrile in a total volume of 2.5 mL at room temperature. Cata-
Eur. J. Inorg. Chem. 0000, 0–0
6
© 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Job/Unit: I50210
/KAP1
Date: 28-04-15 16:07:13
Pages: 8
www.eurjic.org
FULL PAPER
lyst ([Mn] = 3 mm, 1.5 mol-%), substrate (200 mm) and oxidant (to-
tal concentration 280 mm) were added over 30 min with a syringe-
[7] A. Murphy, G. Dubois, T. D. P. Stack, J. Am. Chem. Soc. 2003,
125, 5250–5251.
[
8] a) A. K. Poulsen, A. Rompel, C. J. McKenzie, Angew. Chem.
Int. Ed. 2005, 44, 6916–6920; Angew. Chem. 2005, 117, 7076;
b) R. K. Seidler-Edgal, A. Nielsen, A. D. Bond, M. J. Bjerrum,
C. J. McKenzie, Dalton Trans. 2011, 40, 3849–3858.
pump. The reaction was quenched after three hours with alumin-
ium oxide, and the products were analyzed by 1H NMR spec-
troscopy with 1,2-dichlorobenzene as internal standard.
All the experiments were replicated at least three times and the
values are within Ϯ3%.
[9]
A. Lennartson, C. J. McKenzie, Angew. Chem. Int. Ed. 2012,
51, 6767–6770; Angew. Chem. 2012, 124, 6871.
[
10] a) N. Ortega-Villar, V. M. Uglade-Saldivar, B. Flores-Pérez, M.
Flores-Alamo, J. A. Real, R. Moreno-Esparza, Inorg. Chim.
Acta 2011, 375, 213–219; b) A. Majumder, G. Pilet, M. Sa-
lah El Fallah, J. Ribas, S. Mitra, Inorg. Chim. Acta 2007, 360,
Measurement of Oxygen Evolution: An aqueous solution of the
complex (1 mL, 0.5 mm in [Mn]) and acetonitrile (2 mL) was im-
mersed in a water bath at 40 °C (controlled by using a Lauda eco-
line RE104 thermostat). After 10 min, cold TBHP (6.4 m in water,
2
307–2312; c) C. Baffert, M.-N. Collomb, A. Deronzier, S.
2
mL) was added. The test tube formed part of a closed system, and
Kjaergaard-Knudsen, J.-M. Latour, K. H. Lund, C. J. McKen-
zie, M. Mortensen, L. Preuss Nielsen, N. Thorup, Dalton
Trans. 2003, 1765–1772; d) S. Groni, C. Hureau, R. Guillot,
G. Blondin, G. Blain, E. Anxolabéhère-Mallart, Inorg. Chem.
the gas evolved was led through a reflux condenser and collected in
a low friction, air-tight glass syringe at room temperature. The
amount of oxygen evolved was calculated by volume at ambient air
pressure and temperature. All experiments were repeated at least
twice.
2
008, 47, 11783–11797; e) S. Brooker, V. McKee, J. Chem. Soc.,
Dalton Trans. 1990, 1990.
[11] S. Richards, B. Pedersen, J. V. Silverton, J. L. Hoard, Inorg.
Chem. 1964, 3, 27.
Acknowledgments
[12] H. D. Flack, Helv. Chim. Acta 2003, 86, 905.
[
[
[
13] L. Perez-Garcia, D. B. Amabilino, Chem. Soc. Rev. 2007, 36,
This work was supported by the Danish Council for Independent
Research | Natural Sciences (grant 12-124985 to C. J. McK.) and
the Villum Foundation (post-doctoral funding to C. D.). The
COST actions CM1003 and CM1202 are acknowledged for travel
funding.
941.
14] A. Lennartson, M. Vestergren, M. Håkansson, Chem. Eur. J.
2005, 11, 1757.
15] In fact, the structure displays pseudosymmetry, where the com-
plexes essentially adopt space group I2 (with the same unit
1
cell). The reduction to P2 symmetry arises from the positions
of the perchlorate anions. The diffraction data unambiguously
indicate the primitive lattice.
86–193; b) C. Leonarz, C. J. Schfield, Nat. Chem. Biol. 2008, [16] D. Pijper, P. Saisaha, J. W. de Boer, R. Hoen, C. Smit, A.
[
[
1] a) E. G. Kovaleva, J. D. Lipscomb, Nat. Chem. Biol. 2008, 4,
1
4.
Meetsma, R. Hage, R. P. van Summeren, P. L. Alsters, B. L.
Feringa, W. R. Browne, Dalton Trans. 2010, 39, 10375–10381.
[17] J. Glerup, P. A. Goodson, A. Hazell, R. Hazell, D. J. Hodgson,
C. J. McKenzie, K. Michelsen, U. Rychlewska, H. Toftlund,
Inorg. Chem. 1994, 33, 4105–4111.
[18] a) M. S. Vad, A. Lennartson, A. Nielsen, J. Harmer, J. E.
McGrady, C. Frandsen, S. Morup, C. J. McKenzie, Chem.
Commun. 2012, 48, 10880–10882; b) D. P. de Sousa, J. O. Bige-
low, J. Sundberg, L. Que Jr., C. J. McKenzie, Chem. Commun.
2015, 51, 2802–2805.
2] T. Haikarainen, C. Frioux, L.-Q. Zhnag, D.-C. Li, A. C. Papag-
eorgiou, Biochim. Biophys. Acta Proteins Proteomics 2014,
1844, 422–429.
[
[
[
[
3] R. Anand, P. C. Dorrestein, C. Kinsland, T. P. Begley, S. E.
Ealick, Biochemistry 2002, 41, 7659–7669.
4] Y. Umena, K. Kawakami, J.-R. Shen, N. Kamiya, Nature 2011,
473, 55–61.
5] E. P. Talsi, K. P. Bryliakov, Coord. Chem. Rev. 2012, 256, 1418–
434.
6] a) R. V. Ottenbacher, K. P. Bryliakov, E. P. Talsi, Inorg. Chem.
010, 49, 8620–8628; b) P. Saisaha, J. W. de Boer, W. R.
1
[19] G. M. Sheldrick, SHELXTL, Bruker, AXS, Madison, WI,
USA 2000.
2
Browne, Chem. Soc. Rev. 2013, 42, 2059–2074; c) I. Garcia- [20] M. S. Vad, A. Nielsen, A. Lennartson, A. D. Bond, J. E.
Bosch, L. Gomez, A. Polo, X. Ribas, M. Costas, Adv. Synth.
Catal. 2012, 354, 65–70; d) R. V. Ottenbacher, D. G. Sam-
sonenko, E. P. Talsi, K. P. Bryliakov, ACS Catal. 2014, 4, 1599–
McGrady, C. J. McKenzie, Dalton Trans. 2011, 40, 10698–
10707.
[21] C.-M. Che, W.-T. Tang, K.-Y. Wong, W.-T. Wong, T.-F. Lai, J.
Chem. Res. 1991, 30, 401.
1
606; e) B. Wang, C. Miao, S. Wang, C. Xia, W. Sun, Chem.
Eur. J. 2012, 18, 6750–6753; f) W. Dai, J. Li, G. Li, H. Yang,
L. Wang, S. Gao, Org. Lett. 2013, 15, 4138–4141.
Received: February 26, 2015
Published Online: ᭿
Eur. J. Inorg. Chem. 0000, 0–0
7
© 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Job/Unit: I50210
/KAP1
Date: 28-04-15 16:07:13
Pages: 8
www.eurjic.org
FULL PAPER
Mn–Carboxylate Catalysts
C. Deville, M. Finsel, D. P. de Sousa,
B. Szafranowska, J. Behnken, S. Svane,
A. D. Bond, R. K. Seidler-Egdal,
C. J. McKenzie* ................................ 1–8
Too Many Cooks Spoil the Broth – Varia-
ble Potencies of Oxidizing Mn Complexes
of a Hexadentate Carboxylato Ligand
Keywords: Manganese
/ Epoxidation /
Water oxidation / Carboxylate ligands /
Hexadentate ligands
A manganese(II) complex of a hexadentate
ligand catalyzes epoxidation. However, re-
action conditions are critical since several
metal-based oxidants with varying potenc-
ies can be generated. This observation is
probably general for related catalysts, and
consequences include decreased selectivity
and competing water oxidation.
Eur. J. Inorg. Chem. 0000, 0–0
8
© 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim