Tuning the metal-based redox potentials of manganese cis,cis-1,3,5-triaminocyclohexane complexes

Article abstract of DOI:10.1039/b102103f

The metal-based redox potentials of manganese complexes were studied. Salicylaldehyde and four of its 5-substituted derivates were used to prepare the symmetrical tri-amines. The cyclic voltammograms of the complexes in a DMF exhibited a quasi-reversible

Full text of DOI:10.1039/b102103f

View Article Online / Journal Homepage / Table of Contents for this issue
Tuning the metal-based redox potentials of manganese
cis,cis-1,3,5-triaminocyclohexane complexes
Elizabeth A. Lewis,a John R. Lindsay Smith,*a Paul H. Walton,*a Stephen J. Archibald,a
Simon P. Foxona and Gerard M. P. Giblinb
a Department of Chemistry, University of York, Heslington, York, UK YO10 5DD
b Medicinal Chemistry, Glaxo Wellcome Research and Development, Gunnels Wood Road,
Stevenage, Herts, UK SG1 2NY
Received 6th March 2001, Accepted 15th March 2001
First published as an Advance Article on the web 29th March 2001
A selection of 5-substituted salicylaldehydes have been
reacted with cis,cis-triaminocyclohexane to prepare a series
of tri-imine and tri-amine ligands; cyclic voltammetry
has been used to show that the electronic effect of the
substituent has a marked and systematic effect on the MnIII/
MnIV and MnIII/MnII redox potentials of their manganese
complexes.
uents were selected to encompass a wide range of electronic
effects from electron-donating methoxy- (σp Ϫ0.28) to strongly
electron-withdrawing nitro- (σp ϩ0.81).7 Each of the ligands
was reacted with equimolar amounts of manganese() acetate
dihydrate in methanol, in the presence of excess base, under an
inert atmosphere to give the mononuclear manganese() com-
plexes (3a–e and 4a–e).‡8 Ligands 1a and 2a were also reacted
with gallium() nitrate under the same conditions to give the
gallium complexes 5 and 6, respectively.8a
The cyclic voltammograms (CV) of the complexes (3a–e,
4a–e) in DMF exhibit a quasi-reversible one-electron MnIII/
MnIV oxidation and a less reversible MnIII/MnII reduction
process over the potential range Ϫ1.5 to 0.5 V vs. Fcϩ/Fc.§ To
confirm that the redox processes are metal- and not ligand-
centred, we carried out two experiments. In the first, we used
cyclic voltammetry to show that the gallium tri-imine and
tri-amine complexes, 5 and 6, do not undergo any redox
processes between Ϫ1.5 and 0.5 V, as would be expected of a
redox inactive metal. A poorly reversible process was observed
at ≈0.7 V (also seen with some of the manganese complexes)
which we attribute to oxidation of the ligand. In the second
experiment, we oxidised the manganese() complex 3a with
[FeCp2]PF6 in methanol, under an inert atmosphere, to give the
corresponding manganese() species 7 which was obtained as
intensely blue-coloured crystals.¶ Comparison of the structures
Much of the recent interest in Mn coordination chemistry
stems from the identification of manganese in metalloenzymes
such as the oxygen evolving centre (OEC) of photosystem II,1
catalases,2 and superoxide dismutases.3 Numerous manganese
complexes have been prepared with the aim of modelling these
enzymes. In this respect our recent synthesis of N3 face-capping
tri-imine ligands, from the condensation of aldehydes and
cis,cis-1,3,5-triaminocyclohexane (tach), provides
a useful
approach to a wide range of biomimetic metal complexes.4 A
further important target, which would aid the design of more
effective functional models of manganese containing active
sites,5,6 is the ability to control the metal-based redox potentials
of the complexes. This is the subject of the work described below.
We have used salicylaldehyde and four of its commercially
available 5-substituted derivatives to prepare the symmetrical
tri-imines (1a–e) and by reduction with sodium borohydride the
analogous tri-amine ligands (2a–e) (Scheme 1).† The substit-
Scheme 1 Reagents and conditions: (i) Et2O–H2O, 3 equiv. NaOH and appropriate salicylaldehyde, rt, 2 h, >90%; (ii) 1a–e, CH3OH, 10 equiv.
NaBH4, reflux 4 h, 50–70%; (iii) 1a–e, 2–e, dry CH3OH, N2, Mn() acetateؒ2H2O, Et3N, rt overnight, chromatography (neutral aluminium oxide,
85 : 15 v/v, CH2Cl2–CH3OH) 60–80%.
DOI: 10.1039/b102103f
J. Chem. Soc., Dalton Trans., 2001, 1159–1161
This journal is © The Royal Society of Chemistry 2001
1159

View Article Online
lighting the significant influence of the substituents on the
redox potentials of the two processes. These electronic effects
are larger than typically observed for reactions/equilibria
of phenols,14 although Borovik et al.15 have reported a ρ value
of 4.55 for the VOIV/VOV couple of some oxovanadium bis-
amidophenolate complexes. It seems likely that, as proposed by
Borovik et al., the substituent effect in these chelate complexes
is being relayed to the metal ion by two pathways, via the
phenolate and the imine groups, and the Hammett correlation
is measuring the combined effects of both. The excellent corre-
lation with σp, however, suggests that the dominant substituent
effect is via the phenolate group.
Controlling the redox potential of manganese in manganese
enzymes is crucial to ensure maximum activity in biological
systems. For iron and manganese superoxide dismutases,
for example, the optimum value (0.2–0.4 V vs. NHE)16 is
determined by the amino-acids in the second coordination
sphere which tune the proton coupled FeIII/FeII and MnIII/MnII
redox potentials.17 The present study shows that with model
systems the electronic effects of the substituents on the ligand
can be used, in an analogous way, to tune the redox potential of
a single manganese couple over a range of ≈0.6 V. Importantly,
for the MnIII/MnIV couple, this spans the redox potential of
many manganese enzymes. The complexes described in this
communication are currently under investigation as manganese
superoxide dismutase and catalase models.
Fig. 1 ORTEP9 representation (50% probability ellipsoids) of man-
ganese complex 3a, selected bond distances (Å): Mn–O(1) 1.906(8),
Mn–O(2) 1.906(7), Mn–O(3) 2.112(8), Mn–N(1) 2.016(9), Mn–N(2)
2.234(11), Mn–N(3) 2.124(10). For comparison the equivalent data for
7 are (Å): Mn–O(1) 1.875(6), Mn–O(2) 1.861(7), Mn–O(3) 1.857(6),
Mn–N(1) 2.011(8), Mn–N(2) 2.005(7), Mn–N(3) 2.007(7).
Acknowledgements
We acknowledge GlaxoWellcome and the EPSRC for providing
a CASE studentship (E. A. L.).
Notes and references
† Each ligand has been fully characterised by ESMS, NMR, IR and
UV-Vis techniques.
‡ Each complex has been fully characterised by ESMS, IR, UV-Vis
techniques and C,H,N elemental analyses.
§ Cyclic voltammetry was performed on an EG&G Princeton Applied
Research potentiostat/galvanostat model 273 using a standard three
electrode configuration with platinum working (0.5 mm diameter disk)
and counter electrodes and an Ag/AgCl reference electrode. All experi-
ments were carried out under N2 at ambient temperature with 0.2 M
[Bu4N]BF4 as the supporting electrolyte. No IR compensation was
applied to the recorded CVs. The Fcϩ/Fc couple was used to monitor
the reference electrode and was observed at 0.49 V.
Fig. 2 Hammet plot of redox potentials (vs. Fcϩ/Fc, in DMF at 298 K
with 0.2 M [Bu4N]BF4) against substituent parameter, σp; MnIII/MnIV
:
᭹ complexes 3a–e and ꢀ complexes 4a–e; MnIII/MnII: ᭿ complexes
3a–e.
Electrochemical data for complexes 3a–e and 4a–e. For MnIII/MnIV
processes, values of E1/2 were independent of scan speed (0.05–0.5 mV
s
Ϫ1). Also, plots of peak current vs. scan speed were linear. Relevant
data, E1/2 (peak to peak separation, ∆E, and ratio of current of anodic
and cathodic waves, ia/ic); Fcϩ/Fc, 0.49 V (78 mV, 0.98); 3a, Ϫ0.31 V
(98 mV, 1.02); 3b, Ϫ0.41 V (120 mV, 0.98); 3c, Ϫ0.37 V (82 mV, 1.02);
3d, Ϫ0.17 V (91 mV, 0.98); 3e, 0.15 V (88 mV, 1.00); 4a, Ϫ0.49 V (103
mV, 1.10); 4b, Ϫ0.56 V (143 mV, 0.67); 4c, Ϫ0.49 V (154 mV, 1.12); 4d,
Ϫ0.34 V (128 mV, 0.93); 4e, 0.07 V (189 mV, 0.92). For MnIII/MnII
processes, the return oxidation wave was clearly present but much
less distinguishable than the return reduction wave in the MnIII/MnIV
processes, ∆E > 290 mV in all cases. Relevant data E for the reduction
wave vs. Fcϩ/Fc: 3a, Ϫ0.96 V; 3b, Ϫ1.08 V; 3c, Ϫ0.99 V; 3d, Ϫ0.80 V;
3e, Ϫ0.44 V.
¶ Typical procedure for the synthesis of 7. Complex 3a (0.1 mmol,
5 mg) and [FeCp2]PF6 (0.1 mmol, 3.6 mg) were combined in a 1 : 1 v/v
mixture of dry CH3OH and CH2Cl2 (4 cm3) and stirred for 2 h. The
solvent was removed under vacuum and the residue was redissolved in
CH3OH. Layering with Et2O afforded 7 as dark blue crystals which
were characterised by ESMS, IR and UV-Vis techniques.
|| Crystal data for 3a: C28H26Cl2N3O3Mn, M = 578.36, a = 10.192(2),
b = 15.336(3), c = 16.322(5) Å, V = 2551.3(11) Å3, orthorhombic, space
group Pna21 (no. 33), Z = 4, µ(Mo-Kα) = 0.764 mmϪ1, 1379 (all unique),
T = 150 K. Final R1, wR2 on all data (1379 reflections) 0.0934, 0.1022.
R1, wR2 on [Io > 2σ(Io)] (916 reflections) 0.0371, 0.0823.
Crystal data for 7: C27H24N3O3MnPF6, M = 638.40, a = 36.598(13),
b = 7.890(1), c = 22.623(7) Å, β = 126.725(15)Њ, V = 5235(3) Å3, mono-
clinic, space group C2/c (no. 15), Z = 8, µ(Mo-Kα) = 0.645 mmϪ1, 4859/
4508 measured/unique (Rint = 0.117), T = 150 K. Final R1, wR2 on all
data (4508 reflections) 0.2395, 0.2525. R1, wR2 on [Io > 2σ(Io)] (1720
reflections) 0.0686, 0.1734.
(from single crystal X-ray diffraction) of 3a and 7 shows that
the Mn–O(3) and Mn–N(2) bonds of the former are signifi-
cantly elongated, whereas all the Mn–O and all the Mn–N bond
lengths of the latter are very similar (Fig. 1).|| These observ-
ations are consistent with a substantial tetragonal Jahn–Teller
distortion of the high-spin d4 manganese() complex and the
absence of such a distortion for the d3 manganese() species.12
This is further supported by the magnetic susceptibilities of 3a
and 7, using the Evans method,13 [295 K, (CD3)2SO)] as 5.0 and
4.0 µB respectively. When 7 was subjected to cyclic voltammetry,
under the conditions used for the tri-imine complexes, its CV
was identical to that of 3a.
The MnIII/MnIV and MnIII/MnII redox potentials (vs. Fcϩ/Fc)
of complexes 3a–e and 4a–e are strongly dependent on the
electronic properties of the substituent.§ The former for both
series of complexes give excellent linear correlations (r ≥ 0.990)
when plotted against the Hammett substituent constant,7 σp,
(Fig. 2).** For the MnIII/MnII redox potential, the equivalent
linear correlation is also very good for the tri-imines 3a–e
(r 0.996) but it is rather more scattered for the corresponding
tri-amine species 4a–e (r 0.808). Interestingly, the reaction con-
stants per substituent, ρ, for the redox processes (determined
from the slopes of the appropriate Hammett plots) are large
and very comparable (MnIII/MnIV: 3a–e, ρ = Ϫ2.97 0.14;
4a–e, ρ = 3.26 0.27; MnII/MnIII: 3a–e, ρ = Ϫ3.31 0.17) high-
1160
J. Chem. Soc., Dalton Trans., 2001, 1159–1161

View Article Online
The structures of 3a and 7 were solved by direct methods (SIR-92)10
and refined against F2 (SHELXL 93)11. 3a was refined with all non-
hydrogen atoms anisotropic. 7 was refined with only the manganese
atom having anisotropic thermal parameters (a full anisotropic refine-
ment on non-hydrogen atoms gave negative values for thermal param-
eters on some atoms; this is probably because crystals of 7 were weakly
diffracting and only a relatively small amount of data was collected). In
both 3a and 7 the overall structures of the complexes are clear. Hydro-
gen atoms were placed at calculated positions and refined as riding
atoms with isotropic displacement parameters. A dichloromethane
solvent molecule in 3a was disordered with the carbon atom occupying
two positions (0.60 and 0.40). CCDC reference numbers 155490 and
155491. See http://www.rsc.org/suppdata/dt/b1/b102103f/ for crystallo-
graphic data in CIF or other electronic format.
the first example of Mn() complexes: (a) J. E. Bollinger, J. T.
Mague and D. M. Roundhill, Inorg. Chem., 1994, 33, 1241; (b) J. E.
Bollinger, J. T. Mague, C. J. O’Connor, W. A. Banks and D. M.
Roundhill, J. Chem. Soc., Dalton Trans., 1995, 1677; (c) J. E.
Bollinger, J. T. Mague, W. A. Banks, A. J. Kastin and D. M.
Roundhill, Inorg. Chem., 1995, 34, 2143; (d) P. Caravan and
C. Orvig, Inorg. Chem., 1997, 36, 236.
9 C. K. Johnson, ORTEP, Report ORNL-5138, Oak Ridge National
Laboratory, Oak Ridge, TN, 1976.
10 A. Altomare, M. Cascarano, C. Giacovazzo and A. Guagliardi,
J. Appl. Crystallogr., 1993, 26, 343.
11 G. M. Sheldrick, SHELXL 93, Program for crystal structure refine-
ment, University of Göttingen, 1993.
12 Changes in geometry of Mn() and Mn() coordination spheres
are considered in detail by the following: (a) M. G. B. Drew, C. J.
Harding, V. McKee, G. G. Morgan and J. Nelson, J. Chem.
Soc., Chem. Commun., 1995, 1035; (b) S. K. Chandra and A.
Chakravorty, Inorg. Chem., 1991, 3795; (c) S. Dutta, P. Basu and
A. Chakravorty, Inorg. Chem., 1991, 30, 4031; (d) I. A. Fallis, L. J.
Farrugia, N. M. Macdonald and R. D. Peacock, J. Chem. Soc.,
Dalton Trans., 1993, 2759.
** To correct the gradient, in mV, to a conventional ρ value requires
dividing by 3, to take into account the three substituents per complex
and by Ϫ59.2 mV [Ϫ2.303(RT/zF)].
1 J. Limburg, V. A. Szalai and G. W. Brudvig, J. Chem. Soc., Dalton
Trans., 1999, 1351; V. K. Yachandra, K. Sauer and M. P. Klein,
Chem. Rev., 1996, 94, 807.
2 G. C. Dismukes, Chem. Rev., 1996, 96, 2909.
3 D. P. Riley, Chem. Rev., 1999, 99, 2573.
4 B. Greener, L. Cronin, G. D. Wilson and P. H. Walton, J. Chem.
Soc., Dalton Trans., 1996, 401.
5 K. Wieghardt, Angew. Chem., Int. Ed. Engl., 1989, 22, 1153.
6 I. Batinic-Harberle, I. Spasojevic, P. Hambright, L. Benov,
A. Crumbliss and I. Fridovich, Inorg. Chem., 1999, 38, 4011.
7 σp values are from O. Exner, Correlation Analysis of Chemical Data,
Plenum, New York, 1988 and H. H. Jaffé, Chem. Rev., 1953, 53,
191.
13 D. F. Evans, J. Chem. Soc., 1959, 2003.
14 The ionisation of phenols in water at 25 ЊC, ρ = 2.20, H. van
Bekkum, P. E. Verkade and B. M. Wepster, Recl. Trav. Chim.
Pays-Bas, 1959, 78, 815; the reaction of phenolate ions with
iodoethane in ethanol at 25 ЊC, ρ = Ϫ0.99, L. P. Hammett, J. Am.
Chem. Soc., 1937, 59, 97.
15 A. S. Borovik, T. M. Dewey and K. N. Raymond, Inorg. Chem.,
1993, 32, 413.
16 J. W. C. Barrette, D. T. Sawyer, J. A. Fee and K. Asada,
Biochemistry, 1983, 22, 624.
8 Al(), Ga(), In() and Fe() complexes of some of these
tri-amine ligand systems have previously been reported but this is
17 C. K. Vance and A.-F. Miller, J. Am. Chem. Soc., 1998, 120, 461;
C. K. Vance and A.-F. Miller, Biochemistry, 1998, 37, 5518.
J. Chem. Soc., Dalton Trans., 2001, 1159–1161
1161