Design, synthesis and redox properties of two ferrocene-containing iron chelators

Article abstract of DOI:10.1016/j.tetlet.2006.03.088

Two ferrocene-containing iron(III) chelators were synthesized from desferrioxamine B and kojic acid and their electronic absorption and electrochemical properties were studied in acetonitrile in the absence and presence of ferric ions. The results show a

Full text of DOI:10.1016/j.tetlet.2006.03.088

Tetrahedron Letters 47 (2006) 3371–3374
Design, synthesis and redox properties of two
ferrocene-containing iron chelators
a
b
b
Fabrice Moggia,a Hugues Brisset,a, Frederic Fages, Carole Chaix, Bernard Mandrand,
*
´ ´
c
c
`
Marylene Dias and Eric Levillain
aGroupe de Chimie Organique et Mate´riaux Mole´culaires, UMR CNRS 6114, Universite´ de la Me´diterrane´e, case 901,
163 avenue de Luminy, 13288 Marseille Cedex 9, France
b
`
´
Systemes Macromoleculaires et Immunovirologie Humaine, UMR CNRS 2714, Ecole Normale Superieure de Lyon,
´
46 alle´e d’Italie, 69007 Lyon, France
cCIMMA, UMR CNRS 6200, Universite´ d’Angers, 2 Bd. Lavoisier, 49045 Angers Cedex, France
Received 18 February 2006; accepted 14 March 2006
Available online 3 April 2006
Abstract—Two ferrocene-containing iron(III) chelators were synthesized from desferrioxamine B and kojic acid and their electronic
absorption and electrochemical properties were studied in acetonitrile in the absence and presence of ferric ions. The results show a
complex behavior arising from the occurrence of competing redox and complexation processes. Such systems are a first step toward
the generation of chemosensors for the electrochemical detection of iron(III) in a solution.
ꢀ 2006 Elsevier Ltd. All rights reserved.
Iron plays numerous essential roles in life processes.1
However, excess levels of this metal are deleterious to
living organisms.1 Therefore, Fe3+ sensing methods
capable of providing quantification of the available iron
pool in biological and environmental media are of con-
siderable interest. The methods of detection described so
far are not fully satisfactory because they require the
destruction of the sample, lack the requisite sensitivity,
or are not easily adaptable for high throughput assays
and in vivo measurements.2 A possibility to solve this
problem is the use of spectrofluorimetric methods that
offer valuable tools for real-time monitoring, sensitive
determination, and in situ imaging of chemical and bio-
chemical species. Fluorescently labeled siderophore-like
production cost, and miniaturization. These consider-
ations led us to investigate the potential of molecular
systems in which the Fe3+ chelating unit is instead
connected to a redox signaling unit, such as ferrocene,
for the electrochemical sensing of Fe3+
.
While ferrocene-based receptors were designed to allow
the electrochemical detection of a great variety of
metallic species,5 there is no example to our knowledge
of such systems that are capable to electrochemically
respond to the presence of ferric ions in a solution.
In this context, we have decided to synthesize two
ferrocene-containing siderophore-like prototype sys-
tems, 1 and 2, and investigate their physico-chemical
and electrochemical properties in the absence and
presence of Fe3+ in acetonitrile.
chelators were reported to act as versatile and effective
reporters of Fe3+
.
Recently, Hider and co-workers
3
described a series of iron-specific fluorescent probes in
which a fluorescent coumarin moiety forms a part of
bidentate hydroxypyridinone or hydroxypyranone
ligands.4 Another way is the use of electrochemical
methods. Compared to spectrofluorimetric methods,
electrochemical detections receive particular attention
due to their high sensitivity, easy instrumentation, low
O
O
O
HN
OH
N
N
O
H
Fe
OH
O
Fe
HO
O
OH
N
O
N
O
2
H
N
*
Corresponding author. Tel.: +33 491 829 587; fax: +33 491 829
580; e-mail: brisset@univmed.fr
1
0040-4039/$ - see front matter ꢀ 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetlet.2006.03.088

3372
F. Moggia et al. / Tetrahedron Letters 47 (2006) 3371–3374
Fe3+ is known to be a strong one-electron oxidant capa-
ble of oxidizing pristine ferrocene into ferricinium ion in
solution.6 We were, therefore, interested in examining
the interplay between chelation and redox phenomena
occurring in solutions of 1 and 2 containing gradual
amounts of ferric ions. In 1, the ferrocenyl moiety is
covalently bound to the terminal amino functionality
of the hexadentate ligand desferrioxamine B (DFOB),
a microbial trishydroxamate siderophore.7 Here, Fe3+
complexation was anticipated to lead to the formation
of a 1:1 metal/ligand complex, bearing a single ferrocene
unit attached to the coordination center via a flexible
link, similar to the case of the fluorescent DFOB deriv-
atives reported by Shanzer and co-workers.8 As a biden-
tate ligand, the hydroxypyranone9 derivative 2 is
expected to form a 1:3 metal/ligand complex featuring
three ferrocene peripheral units. Moreover, this ligand
was designed to provide a through bond communication
between the redox and metal centers owing to the
presence of the conjugated bridge.5a
the literature.5a After conversion of 7 to the chloride 6
with thionyl chloride and subsequent treatment of 6 with
trimethylphosphite, phosphonate 5 was reacted with
ferrocene carboxaldehyde under Wittig–Horner condi-
tions to form the double bond in the precursor 4 (80%
yield).11 Deprotection of 4 with boron trichloride
afforded 2 as a red solid (70% yield).12
The electronic absorption spectrum of free ligand 1 in
DMF showed a transition band at 440 nm, characteris-
tic of the ferrocene absorption (Fig. 1). Upon addition
of gradual amounts of Fe(ClO4)3, the absorbance at
440 nm was observed to increase and to level off at
one molar equivalent of added metal, which was attrib-
uted to the formation of the 1:1 Fe3+/DFOB complex.7
Beyond 1 equiv, a new absorption band was observed to
emerge at lower energy (620 nm), and was assigned to
the formation of the ferricinium ion.13
A similar behavior was noticed for 2 in the presence of
Fe3+ in acetonitrile. However, the electronic absorption
band (kmax = 482 nm) of the ferrocene unit was red-
shifted relative to that of 1 as a result of the presence
of the p-conjugating bridge between ferrocene and
hydroxypyranone subunits. Furthermore, addition of
Fe(ClO4)3 up to 0.3 mol equiv led to the formation of
the 1:3 metal/ligand complex (kmax = 517 nm), the
absorption of the ferricinium ion being observed only
when the metal-to-ligand ratio was more than 1:3. These
Ligand 1 was obtained by a classical two-step procedure
(Scheme 1). Ferrocene carboxylic acid was activated
via DCC-mediated condensation with N-hydroxy-
succinimide to give the activated ester 3. The latter was
reacted in DMF with DFOB in the presence of triethyl-
amine to give 1 as an orange solid (88% yield).10 Kojic
acid was reacted with benzyl bromide to afford 5-(benzyl-
oxy)-2-(hydroxymethyl)-4H-pyran-4-one 7 according to
O
O
O
N
O
OH
NHS, DCC
O
Fe
Fe
THF, r.t.
3
O
O
HN
N
N
O
H
Fe
OH
DFOB, Et3N
DMF, 50 C
HO
O
˚
OH
N
O
N
O
H
N
1
O
O
O
O
SOCl2
r.t.
P(OCH3)3
HO
O
Cl
O
O
7
6
O
O
O
O
O
Fe
O
O
P
O
Fe
C2H5OLi
O
O
4
5
OH
O
BCl3
r.t.
Fe
2
Scheme 1. Synthesis of 1 (top) and 2 (bottom).

F. Moggia et al. / Tetrahedron Letters 47 (2006) 3371–3374
3373
Figure 2. Open circuit potential (E) measured for a solution of 2
(5 · 10À4 M in 0.2 M nBu4NPF6/CH3CN) as a function of added
Fe(ClO4)3 (0–1 equiv).
Figure 1. UV–vis spectra of a DMF solution of 1 (5 · 10À4 M)
containing gradual amounts of Fe(ClO4)3 (0–2 equiv).
clearly indicate that Fe3+ coordination is the favored
process unless there is no excess of free metal in the
solution.
results indicate that, for both ligands, oxidation of the
pendant ferrocene moieties occurs only when an excess
of metal is present in the solution. This is consistent with
the high thermodynamic stability of ferric hydroxamates
and the reduced oxidizing ability of Fe3+ in such species.9
In conclusion, we have designed, synthesized, and char-
acterized two ferrocene-containing chelators that are
sensitive to Fe3+. Although CV proved to be poorly sen-
sitive to the binding events in the solution, these ligands
represent a first, simple approach toward potentiometric
chemosensors for Fe3+. To achieve this goal, more work
is to be directed to optimizing the coupling mechanism
between the redox active moiety and the guest binding
site. At this stage, ligands 1 and 2 allow to monitor
the OCP change in response to Fe3+ complexation. This
effect stems from the oxidation of the ferrocene probe
which occurs beyond a metal concentration threshold.
In this respect, ligands 1 and 2 could be useful for the
generation of alerting or dosimeter devices capable of
signaling Fe3+ concentration jumps.
Cyclic voltammetry (CV) of the free ligands shows a
reversible redox system at 0.04 and 0.10 V (vs Fc+/Fc)
for 1 in 0.2 M nBu4NPF6/DMF and 2 in 0.1 M Li-
ClO4/CH3CN, respectively. The positive shift observed
relative to pristine ferrocene could be attributed to the
electron-withdrawing effect of the carbonyl group.6 Sur-
prisingly, the addition of Fe3+ to the solution of 1 or 2
did not lead to any significant change in the position
of the ferrocene-centered redox wave. The absence of
electrochemical transduction in ligand 1 can be rational-
ized by the fact that the ferrocene probe is remote from
the coordination center. Moreover, the 1:1 metal/ligand
complex is neutral, which is expected not to induce any
electrostatic destabilization of the ferricium cation, in
contrast to the case of other ferrocene-containing recep-
tors.5a The behavior of the conjugated system 2 remains
yet not fully understood. It appears that the ferrocene
probe does not respond to a binding event at the
hydroxypyranone moiety, although both subunits are
connected via a conjugated bridge.
References and notes
1. (a) Crichton, R. R. Inorganic Biochemistry of Iron
Metabolism; E. Horwood: New York, 1991; (b) Mene-
ghini, R. Free Radical Biol. Med. 1997, 23, 783–792; (c)
Hentze, M. W.; Kuhn, L. C. Proc. Natl. Acad. Sci. U.S.A.
1996, 93, 8175–8182.
2. (a) Baliga, R.; Ueda, N.; Shah, S. V. Biochem. J. 1993,
291, 901–905; (b) Gower, J. D.; Healing, G.; Green, C. J.
Anal. Biochem. 1989, 180, 126–130; (c) Kozlov, A. V.;
Yegorov, D.; Vladimirov, Y. A.; Azizova, O. A. Free
Radical Biol. Med. 1992, 13, 9–16; (d) Rothman, R. J.;
Serroni, A.; Farber, J. L. Mol. Pharmacol. 1992, 42, 703–
710; (e) Cooper, C. E.; Lynagh, G. R.; Hoyes, K. P.;
Hider, R. C.; Cammack, R., et al. J. Biol. Chem. 1996,
271, 20291–20299.
These results led us to investigate the influence of the
addition of ferric ions on the open circuit potential
(OCP), that is, the potential at which there is no current.
The potential measured is a mixed potential, a function
of all species present in solution and their concentra-
tions. The measurement of the OCP can be used to elab-
orate a potentiometric sensor. Titration of a solution of
pristine ferrocene with Fe(ClO4)3 was observed to
induce a continuous positive shift of the OCP value
from the first metal solution aliquot, which is the result
of the Fe3+ promoted oxidation of ferrocene. In the case
of 1 and 2, a similar effect was observed only when the
metal-to-ligand ratio was more than 1:1 and 1:3, respec-
tively (Fig. 2). These results are in agreement with those
obtained by electronic absorption spectroscopy. They
´
3. (a) Thomas, F.; Serratrice, G.; Beguin, C.; Saint Aman, E.;
`
Pierre, J.-L.; Fontecave, M.; Laulhere, J.-P. J. Biol. Chem.
1999, 274, 13375–13383; (b) Fages, F.; Bodenant, B.; Weil,
T. J. Org. Chem. 1996, 61, 3956–3961; (c) Nudelman, R.;
Ardon, O.; Hadar, Y.; Chen, Y.; Libman, J.; Shanzer, A.
J. Med. Chem. 1998, 41, 1671–1678; (d) Barrero, J. M.;
´
´
Camara, C.; Perez-Conde, M. C.; San Jose, C.; Fernandez,
L. Analyst 1995, 120, 431–435.

3374
F. Moggia et al. / Tetrahedron Letters 47 (2006) 3371–3374
4. (a) Ma, Y.; Luo, W.; Quinn, P. J.; Liu, Z.; Hider, R. C.
3.47 (m, 6H, CH2NOH), 4.13 (s, 5H, HFc), 4.31
(s, 2H, HFc), 4.76 (s, 2H, HFc), 7.77 (s, 2H, NHCO),
7.94 (s, 1H, NHCOFc), 9.61 and 9.65 (2s, 2H and 1H,
NOH); 13C NMR (DMSO-d6) d: 20.2, 23.4, 23.5, 25.9,
26.0, 27.5, 28.7, 29.1, 29.8, 38.3, 38.4, 40.1, 40.2, 45.7,
46.7, 46.9, 47.0, 67.9, 69.1, 69.6, 76.8, 168.5, 169.9, 171.1,
171.8.
J. Med. Chem. 2004, 47, 6349–6362; (b) Ma, Y.; Luo, W.;
Camplo, M.; Liu, Z.; Hider, R. C. Bioorg. Med. Chem.
Lett. 2005, 15, 3450–3452.
5. (a) Beer, P. D.; Gale, P. A.; Chen, G. Z. Coord. Chem.
Rev. 1999, 185–186, 3–36; (b) Beer, P. D.; Hayes, E. J.
Coord. Chem. Rev. 2003, 240, 167–189; (c) Kaifer, A. E.;
Mendoza, S. In Comprehensive Supramolecular Chemistry;
Pergamon: Oxford, 1996; Vol. 1, pp 701–732; (d) Boulas,
11. Characterization of 4. Mp: 123–5 ꢁC. 1H NMR (CDCl3) d:
4.16 (s, 5H, HFc), 4.41 (d, 4H, HFc), 5.05 (s, 2H, CH2O),
´
P. L.; Gomez-Kaifer, M.; Echegoyen, L. Angew. Chem.,
6.20 (d, 1H, J = 16.00 Hz,
Hethylenic), 6.14 (d,
Int. Ed. 1998, 37, 216–247; (e) Bernhardt, P. V.; Moore, E.
G. Aust. J. Chem. 2003, 56, 239–258.
J = 15.80 Hz, 1H, Hethylenic), 7.35–7.41 (m, 5H, Harom),
7.52 (s, 1H). 13C NMR (CDCl3) d: 68.0, 69.6, 70.7, 71.9,
79.9, 111.9, 116.0, 127.8, 128.3, 128.7, 135.0, 137.0, 140.9,
161.7, 175.2. HRMS (FAB+): calculated for C24H20FeO3
(M+H)+ 413.0752, found 413.0756.
6. E1/2 (ferrocinium/ferrocene) = 0.55 V versus ENH; E1/2
(Fe3+/Fe2+) = 0.77 V versus ENH.
7. Raymond, K. N.; Mueller, G.; Matzanke, B. F. Top. Curr.
Chem. 1984, 123, 49–102.
8. Lytton, S. D.; Cabantchik, Z. I.; Libman, J.; Shanzer, A.
Mol. Pharmacol. 1991, 40, 584–590.
12. Characterization of 2. Mp: 192 ꢁC. 1H NMR (CDCl3)
d: 4.18 (s, 5H, HFc), 4.39 (d, 4H, HFc), 6.11 (d,
1H, J = 15.80 Hz, Hethylenic), 6.19 (d, J = 16.00 Hz, 2H,
9. Liu, Z. D.; Hider, R. C. Coord. Chem. Rev. 2002, 232,
151–171.
H
ethylenic and H), 7.50 (s, 1H). RMN 13C (CDCl3) d: 68.3,
69.5, 70.2, 80.2, 112.0, 116.0, 136.1, 137.0, 140.9, 161.2,
174.8. HRMS (FAB+): calculated for C24H20FeO3
(M+H)+ 323.0292; found 323.0301.
10. Characterization of 1. Mp: 181–184 ꢁC. 1H NMR
(DMSO-d6) d: 1.19–1.48 (m, 18H, 9*CH2), 1.95 (s, 3H,
CH3CO), 2.25 (4H, CH2CONH), 2.56 (4H, CH2CO-
NOH), 2.99 (4H, CH2NHCO), 3.14 (2H, CH2NHCOFc),
13. Sohn, Y. S.; Hendrickson, D. N.; Gray, H. B. J. Am.
Chem. Soc. 1971, 93, 3603–3612.