Iron(III) exchange process between hexadentate tripodal ligands: Models for the ternary complexes

Article abstract of DOI:10.1002/ejic.200700259

Using tripodal ligands as tools, the kinetics of iron removal from ferric complexes of the tripodal tris(hydroxyquinoline) ligands O-TRENSOX and of tris(catechol) TRENCAMS was investigated. Two hexadentate heterotripodal ligands were also examined (TRENSOX2CAMS with two hydroxyquinoline and one catechol subunits and TRENSOXCAMS2 with two catechol and one hydroxyquinoline subunits) as possible models of the ternary complexes involved in the iron exchange process. The main result of this kinetic study is the pertinence of the iron complexes of the mixed ligands as models for the transient and successive ternary complexes occurring in the iron exchange, which proceeds arm-by-arm between a homotopic ligand and the iron complex of another homotopic ligand. Wiley-VCH Verlag GmbH & Co. KGaA, 2007.

Full text of DOI:10.1002/ejic.200700259

FULL PAPER
DOI: 10.1002/ejic.200700259
Iron(III) Exchange Process between Hexadentate Tripodal Ligands: Models for
the Ternary Complexes
Guy Serratrice,*[a] Frédéric Biaso,[a] Jean-Louis Pierre,[a] Sylvie Blanc,[b] and
Anne-Marie Albrecht-Gary[c]
Keywords: Mixed ligands / Iron / Exchange interactions / Siderophores / Tripodal ligands
Using tripodal ligands as tools, the kinetics of iron removal
from ferric complexes of the tripodal tris(hydroxyquinoline)
ligands O-TRENSOX and of tris(catechol) TRENCAMS was
investigated. Two hexadentate heterotripodal ligands were
also examined (TRENSOX2CAMS with two hydroxyquinol-
ine and one catechol subunits and TRENSOXCAMS2 with
two catechol and one hydroxyquinoline subunits) as possible
models of the ternary complexes involved in the iron ex-
change process. The main result of this kinetic study is the
pertinence of the iron complexes of the mixed ligands as
models for the transient and successive ternary complexes
occurring in the iron exchange, which proceeds “arm-by-
arm” between a homotopic ligand and the iron complex of
another homotopic ligand.
(© Wiley-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim,
Germany, 2007)
discussed to the best of our knowledge. Recently, Crumbliss
et al. described the formation of a ternary complex in a
redox mechanism for iron release from a siderophore.[5]
Using tripodal ligands as tools and starting from the hy-
pothesis of a step-by-step (“arm-by-arm”) exchange, we in-
vestigated the kinetics of the iron removal from the ferric
complexes of the tripodal tris(hydroxyquinoline) ligands O-
TRENSOX (L0),[6] and LЈ0,[7] where L0 is anchored to a
tertiary amine and LЈ0 to a tertiary carbon (Figure 1), and
a tris(catechol) analogue built from a tertiary amine[8]
TRENCAMS (L3). Two hexadentate heterotripodal ligands
were also examined, TRENSOX2CAMS (L1) with two hy-
droxyquinoline units and one catechol[9] unit and TREN-
SOXCAMS2 (L2) with two catechol moieties and one hy-
droxyquinoline[9] building block (Figure 1). Our aim was to
gain insight into the step-by-step ligand exchange processes
via kinetic intermediates called “ternary complexes”. In ad-
dition, kinetic experiments on the iron removal from EDTA
by these tripodal ligands were carried out. All the collected
results allow the analysis of the influence of the structure
of the ligand and of the nature of the chelating groups on
the kinetic parameters.
Introduction
Microbial iron bioavailability is driven by the coordina-
tion chemistry of siderophores, which are iron-chelating
agents that are excreted by microorganisms to render iron
soluble in the environment and favor the uptake of this
metal. Iron chelation by some natural or synthetic chelators
can be applied to human diseases characterized by iron
overload, and water-soluble iron complexes can be used to
alleviate iron deficiency in plants, preventing and even re-
versing iron chlorosis. The coordination chemistry of a
great number of water-soluble natural or synthetic iron che-
lators was recently reviewed, and it is usually considered
that FeIII removal from ferrisiderophore or from other bio-
logical complexes (transferrin carriers or storage complexes,
receptor proteins, citrate, and low molecular-weight che-
lates) involves the formation of a ternary complex.[1] The
dynamic of Fe removal from various FeIII complexes was
studied and the formation of the ternary complex is often
consistent with the kinetic data.[2–4] Nevertheless, a molecu-
lar description of the “so-called ternary” complex was never
[a] Département de Chimie Moléculaire UMR-5250, ICMG FR-
2607 CNRS, Université Joseph Fourier,
Results
B. P. 53, 38041 Grenoble Cedex 9, France
Fax: +33-4-76514836
E-mail: guy.serratrice@ujf-grenoble.fr
We monitored the iron exchange reactions between two
ligands using different competing ligand concentrations in
pseudo first-order conditions. The global reaction of a li-
gand-exchange type is expressed by Equation (1) (charges
omitted for clarity).
[b] Institut Pluridisciplinaire de Recherche sur l’Environnement et
les Matériaux, UMR CNRS 5254, Université de Pau et des Pays
de l’Adour,
B. P. 27540, 64075 Pau Cedex, France
[c] Laboratoire de Physico-Chimie Bioinorganique, UMR CNRS
7509, ECPM,
25, rue Becquerel, 67200 Strasbourg, France
Supporting information for this article is available on the
WWW under http://www.eurjic.org or from the author.
FeL + LЈǞFeLЈ + L
(1)
Eur. J. Inorg. Chem. 2007, 3681–3685
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
3681

G. Serratrice, F. Biaso, J.-L. Pierre, S. Blanc, A.-M. Albrecht-Gary
FULL PAPER
Figure 1. Chemical formulae of tripodal ligands Ln (n refers to the number of catechol groups in the tripodal ligand).
Three FeL complexes were studied: Fe-EDTA, FeL0, and pseudo first-order rate constant kobs on [LЈ] with negligible
FeL3; LЈ was L0, LЈ0, L1, L2, and/or L3. In particular, this intercept was observed [Equation (2)]. The values of k1 are
series of exchanges allows the study of the kinetics of add- collected in Table 1.
ing a catechol or an 8-hydroxyquinoline arm into the iron
kobs = k1 [LЈ]tot
(2)
coordination sphere and the effect of the structure of the
tripodal ligand (i.e. homopodal or heteropodal, anchor
group). During the exchange reaction, monoexponential
growth of absorbance was observed in all cases at the se-
lected wavelength of the FeLЈ complex, indicating that the
reaction proceeds via a single rate-limiting step. No varia-
tion in the amplitude of the absorbance was observed dur-
ing the mixing time for all the studied systems, indicating
that a fast step, involving a change in the coordination
sphere around FeIII does not occur before the first detect-
able step. The UV/Vis spectra were recorded as a function
of the time and at the end of the kinetics. The structure of
the final complex was supported by its absorption spectrum
on the basis of our previous studies, particularly for the
mixed coordination around FeIII[9] related to one catechol
and two 8-hydroxyquinoline groups as in FeL1 or two cate-
chol and one 8-hydroxyquinoline groups as in FeL2 (see the
λmax and ε values in the Experimental Section).
Table 1. Kinetic data (k1 values in –1 s–1) for the first step of the
iron exchange reaction between FeL and LЈ.[a]
Competing
ligand
Ferric complex
[FeL0]3–
[Fe-EDTA]–
[FeL3]6–
L0
L1
L2
L3
LЈ0
4.0Ϯ0.5
14.2Ϯ1.5
97Ϯ2
145Ϯ3
5.1Ϯ0.4
–
1.9Ϯ0.1
5.5Ϯ0.8
1.1Ϯ0.15
–
–
23Ϯ1
77Ϯ4
81Ϯ4[b]
–
[a] In aqueous 50 m MOPS buffer, pH = 7.4, I = 0.1  (NaClO4)
and at temperature
T
=
25.0 °C. [b] Intercept
=
(6.6Ϯ0.6)ϫ10–2 s–1.
Iron Exchange from Fe-O-TRENSOX
The reaction was monitored with L1, L2, and L3 and was
found to proceed in one distinguishable step. The UV/Vis
spectra recorded as the reaction proceeded exhibited isosb-
estic points indicating that no intermediary species were
formed during the reaction. For FeL0 + L3, the reaction
Iron Exchange from Fe-EDTA
The reaction was monitored with L0, LЈ0, L1, L2, and L3. was complete (see Figure S1 in the Supporting Infor-
The reaction was found to be complete and to occur in one mation). For FeL0/L1, the spectrum recorded at the equilib-
distinguishable step. A linear dependence of the observed rium was found to be similar to that of the FeL1 complex,
3682
www.eurjic.org
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Inorg. Chem. 2007, 3681–3685

Iron(III) Exchange Process between Hexadentate Tripodal Ligands
FULL PAPER
indicating the substitution of an 8-hydroxyquinoline group according to Equations (3) and (4) and the rate law accord-
by a catechol group. However, the substitution by the two ing to Equation (5) by considering the two reactions occur
obs
obs
8-hydroxyquinoline groups of L1 does not change the spec- independently since k1 ϾϾ k2
.
trum and cannot be detected. For FeL0/L2 the spectrum
FeL3 + L0 h L0FeL3 k1, k–1
(3)
(4)
recorded at the end of the first step is similar to the spec-
trum of the FeL2 complex, indicating the substitution of
two 8-hydroxyquinoline groups by two catechol groups.
L0FeL3 ǞL0FeL3* k2
Figure 2 shows the UV/Vis spectra recorded as the reaction
proceeds.
k1 = k1[L0] + k–1
obs
(5)
=
Thus, the values k1
=
1.9Ϯ0.1 –1 s–1, k–1
(2.1Ϯ0.2)ϫ10–4 s–1, and k2 = (2.3Ϯ0.3)ϫ10–5 s–1 were ob-
tained.
Discussion
The general feature resulting from the kinetic studies is
that the iron exchange reactions proceed in one detectable
step except for the exchange FeL3/L0 (two detectable steps).
Hence, the formation of an intermediate ternary complex
was only shown for this last case. In a previous investigation
on the acid dissociation kinetics of the FeIII complexes with
these tripodal ligands, we observed a very fast dissociation
of one arm of the ligand which is maintained in proximity
Figure 2. UV/Vis spectra recorded as a function of time for the
exchange FeL0/L2; experimental conditions: [FeL0] = 0.04 m, [L2]
= 0.4 m, pH = 7.4 (50 m MOPS buffer), I = 0.1  (NaClO4), T
= 25 °C. (1) t = 0 s, (2) 260 s; the arrows indicate the direction of
change as the reaction proceeds.
to the first coordination sphere of FeIII [10]
This indicates
.
that the complex has one labile arm thus favoring the for-
mation of the precursor complex and then the substitution
by one arm of the competing ligand. So, it is reasonable
to propose that the kinetic rate constant k1 represents the
substitution rate of one arm of the initial complex by one
arm of the competing ligand. In addition, this study clearly
reveals the effect of the nature of the entering group, cate-
chol or an 8-hydroxyquinoline group, on the ligand substi-
tution rate constants.
The plots of kobs versus the ligand concentration show a
linear dependence with a negligible intercept except for
FeL0/L2. The values of k1 are reported in Table 1.
Iron Exchange from TRENCAMS
The spectrum recorded at the end of the reaction of FeL3
with L1 or L2 is similar to the spectrum of the FeL2 com-
plex, indicating the substitution of a catechol group by an
8-hydroxyquinoline group (see Figure S1 in the Supporting
Information). The reaction was found to proceed in one
detectable step. The plot of kobs versus the ligand concen-
tration shows a linear dependence with negligible intercept.
The values of k1 are reported in Table 1.
Iron Exchange with Homotopic Ligands L0, LЈ0, and L3
For the iron exchange from the EDTA complex, the final
spectrum indicates that the reaction is complete. The same
result is obtained with the heterotopic ligands. It is pro-
posed that when the first arm of the incoming ligand is
attached, EDTA dissociates rapidly and next two arms
quickly coordinate to FeIII via catechol or the 8-hydroxy-
quinoline groups. The values of k1 shows that the substitu-
tion rate of the catechol group (145 –1 s–1) is about one
order of magnitude faster than that of the 8-hydroxyquino-
line (4.0 and 5.1 –1 s–1). Similar values for k1 were obtained
for the iron exchanges between FeL0 and L3 (81 –1 s–1) or
between FeL3 and L0 (1.9 –1 s–1 for the first step).
For the reaction of FeL3 with L0, two steps were ob-
served as deduced from the two-exponential growth of ab-
sorbance measured at 600 nm. The spectrum recorded at
the end of the first step is similar to the spectrum of the
FeL2 complex, indicating the substitution of a catechol
group by an 8-hydroxyquinoline group (species noted
L0FeL3), and the spectrum recorded at the equilibrium after
the second step is similar to the spectrum of the FeL1 com-
plex, which indicates that the substitution of a second cate-
Iron Exchange with Heterotopic Ligands L1 and L2
chol group by an 8-hydroxyquinoline group (species noted
For the exchanges FeL0/L1 and FeL0/L2, the nature of
obs
L0FeL3*). The plot of k1
(first step) versus the ligand the entering group was clearly identified as a catechol group
concentration shows a linear dependence with a nonzero on the basis of the spectrum change. In a previous paper,[9]
obs
intercept while k2
was found to be independent of the we determined the UV/Vis parameters for the ferric com-
ligand concentration. This is consistent with the reactions plexes with a mixed catechol/8-hydroxyquinoline coordina-
Eur. J. Inorg. Chem. 2007, 3681–3685
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org
3683

G. Serratrice, F. Biaso, J.-L. Pierre, S. Blanc, A.-M. Albrecht-Gary
FULL PAPER
tion sphere. In addition, the values of k1 (23 and 77 –1 s–1 the incoming group is not a decisive factor for the rate con-
for FeL0/L1 and FeL0/L2, respectively) are of the same order stant.
of magnitude as those for the exchanges FeL0/L3 (81 –1 s–1)
It should be pointed out that the substitution rate by the
and Fe-EDTA/L3 (145 –1 s–1), which correspond to the catechol group is decreasing as the number of 8-hydroxy-
substitution by a catechol arm. For the exchanges FeL3/ quinoline groups increases (or inversely as the number of
L1 and FeL3/L2, the entering group was identified as an 8- catechol groups decreases) in the competing ligand: 81, 77,
hydroxyquinoline unit according to the spectral change. or 23 –1 s–1 for FeL0/L3, L2, or L1, respectively, and 145,
The values of k1 (5.5 and 1.1 –1 s–1 for FeL3/L1 and FeL3/ 97, or 14.2 –1 s–1 for Fe-EDTA/L3, L2, or L1, respectively.
L2, respectively) are of the same order of magnitude as In contrast, the same trend is not observed for the substitu-
those for the exchanges FeL3/L0 (1.9 –1 s–1), Fe-EDTA/L0 tion rate by the 8-hydroxyquinoline group (substitution of
(4.0 –1 s–1), or Fe-EDTA/LЈ0 (5.1 –1 s–1), which corre- FeL3 by L0, L1, or L2; 1.9, 5.5, or 1.1 –1 s–1, respectively).
spond to the substitution by an 8-hydroxyquinoline arm. This indicates that steric factors rather than statistical fac-
All these results indicate that the substitution rate is mainly tors occur: the 8-hydroxyquinoline group is assumed to
controlled by the nature of the entering group. Accordingly, slow down the entering of the tripodal ligand in the iron
the iron exchange rates from Fe-EDTA with L1 and L2, coordination sphere.
14.2 and 97 –1 s–1, respectively, can be attributed to the
Our results suggest that iron exchange occurs step-by-
substitution by the catechol arm. Table 2 summarizes the step, each arm of the incoming tripodal ligand substitutes
values of k1 for the three series of iron exchange collected successively the leaving ligand as the coordination bonds
according to the nature of the chelating group that enters are broken. This is clearly evidenced for the iron exchange
in the iron coordination sphere.
between FeL3 and L0 since the formation of an intermediate
The fact that close k1 values are obtained for one type ternary complex having a mixed coordination 8-hydroxy-
of entering group (catechol or 8-hydroxyquinoline) in the quinolinate/catecholate could be identified from the UV/Vis
different exchange systems suggests that the step is the sub- spectra as previously published.[9] A probable mechanism is
stitution of one arm of the initial complex by one arm of depicted in Figure 3. The same process is assumed to occur
the competing ligand, leading to the formation of an inter- for the exchange between FeL0 and L3, but the coordination
mediate ternary complex before the substitution of the two of the second and the third catechol groups is fast relative
other arms. The higher substitution rate measured for the to the first one. This can be inferred by the labilizing effect
catechol group can be attributed to its greater electron do- of strong electron pair donors as catechol ligands, which
nor ability and to its lower bulkiness relative to the 8-hy- increases the electron density of the metal, weakening the
droxyquinoline group. This might be also attributed to the bonds with the coordinated donor atoms. Replacement of
charge of the incoming chelating group that depends on the a coordinated group by a catechol group in the first coordi-
protonation state of the ortho hydroxy groups. Indeed, a nation shell enhances the ligand exchange of the remaining
negative deprotonated hydroxy is assumed to be more reac- arms. In contrast, in the case of the reaction between FeL3
tive than the protonated one. For the catechol groups in and L0, the substitution of the second arm of the tripodal
L1, L2, and L3 the lower pKa is 5.62,[9] 5.55,[9] and 5.57,[8] ligand is very slow (k2 = 2.3 10–5 s–1). It is suggested that
respectively, indicating that the hydroxy group of one arm is the substitution of catechol groups by poorer electron pair
deprotonated at pH 7.4, whereas for the 8-hydroxyquinoline donors as 8-hydroxyquinoline groups diminishes the elec-
groups the lower pKa is 7.44[6] for L0 and 6.68[7] for LЈ0. tron density at the iron center, slowing down the exchange
For the sake of comparison, the numbers of deprotonated of the remaining coordinating arms. In addition, this effect
OH groups in the catechol and 8-hydroxyquinoline moieties is enhanced by the steric effect of the incoming 8-hydroxy-
are reported in Table 2. Analysis of the data shows that k1 quinoline group. The sensitivity of the rate constant to the
cannot be directly related to the protonation state of the nature of the bidentate group of the incoming ligand indi-
OH group. It is thus suggested that the negative charge of cates a rather associative character of the mechanism for
Table 2. Kinetic data (k1 values in –1 s–1) for the first step of the iron exchange reaction between FeL and LЈ[a]collected according to the
entering group.
Competing
ligand
Catechol as an entering group
8-Hydroxyquinoline as an entering group
[FeL0]6–
[Fe-EDTA]–
Number of OH
[FeL3]6–
[Fe-EDTA]–
Number of OH
deprotonated[b]
deprotonated[b]
L0
L1
L2
L3
LЈ0
1.9
5.5
1.1
4.0
0.5
0
0
23
77
81
14.2
97
145
1
1.5
3
5.1
1.5
[a] In aqueous 50 m MOPS buffer, pH = 7.4, I = 0.1  (NaClO4) and at temperature T = 25.0 °C. [b] at pH = 7.4; half value indicates
a partially deprotonated group (pH ≈ pKa).
3684
www.eurjic.org
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Inorg. Chem. 2007, 3681–3685

Iron(III) Exchange Process between Hexadentate Tripodal Ligands
FULL PAPER
and L0, L1, L2, L3, or L4 were studied in aqueous 50 m MOPS
buffer, pH = 7.4, I = 0.1  (NaClO4), and at temperature T =
(25.0Ϯ0.2) °C. The kinetics were studied under pseudo first order
conditions by using a large excess of the entering ligand over the
complex. Initial concentration of the FeL complex was
2.0ϫ10–5 molL–1, initial concentration of ligand LЈ was varied in
a range from 2ϫ10–4 to 1.2ϫ10–3 molL–1. Fast kinetic measure-
ments were performed with a KINSPEC UV (BIO-LOGIC Com-
pany, Claix, France) stopped-flow spectrophotometer equipped
with a diode array detector (J & M) and connected to a microcom-
puter; slow kinetic measurements were performed with a Perkin–
Elmer Lambda2 or Varian Cary 50 UV/Vis spectrophotometer.
The kinetic data were treated on line with the commercial BIO-
KINE program (BIO-LOGIC Company, Claix, France). In each
run, equal volumes of solutions of the complex and of the ligand
were mixed. The absorbance changes with time were recorded at
the maximum of the charge-transfer band of the complex that is
formed, specifically, λmax = 490, 525, and 550 nm for FeL3,[8]
FeL2,[9] and FeL1,[9] respectively, and 595 nm for FeL0[6] and
FeLЈ0.[7] Some visible spectra from 400 to 650 nm were recorded in
time by using a diode array detector.
these series of iron exchange. In summary, the exchange rate
constants of the first step depend on the nature of the enter-
ing arm (catechol or 8-hydroxyquinoline) and on the nature
of second and third arm of the incoming tripodal ligand
(homotripodal or heterotripodal; Figure 3).
Supporting Information (see footnote on the first page of this arti-
cle): Figures S1 and S2: UV/Vis spectra recorded as a function of
time for the exchanges FeL0/L3 and FeL3/L2. Tables S1 to S8: rate
constants values for the different exchanges FeL/LЈ.
Figure 3. Successive steps for the ligand exchange at the iron center
via ternary complexes.
Acknowledgments
Conclusions
The main result of this study is the pertinence of the
FeL1 and FeL2 complexes as models for the transient and
successive ternary complexes occurring in the iron exchange
processes between FeL0 and L3 and between FeL3 and L0.
The results allow the first molecular description of the ter-
nary complex formed during the iron exchange between two
abiotic siderophores. Generally, ligand exchanges are slow,
but they are facilitated by the formation of the so-called
ternary ferric complexes with the two ligands bound to
FeIII. Our study supports this concept.
The authors thank Professor P. Baret and G. Gellon for managing
the syntheses of ligands Ln.
[1] A.-M. Albrecht-Gary, A. L. Crumbliss in Metals Ions in Bio-
logical Systems (Eds.: A.Sigel, H. Sigel), Marcel Dekker, New-
York, 1998, vol. 35, pp. 239–327.
[2] Y. Hara, M. Akiyama, J. Am. Chem. Soc. 2001, 123, 7247–
7256.
[3] A. Stinzi, C. Barnes, K. N. Raymond, Proc. Natl. Acad. Sci.
USA 2000, 97, 10691–10696.
[4] B. Faller, H. Nick, J. Am. Chem. Soc. 1994, 116, 3860–3865.
[5] K. A. Mies, J. I. Wirgau, A. L. Crumbliss, Biometals 2006, 19,
115–126.
[6] G. Serratrice, H. Boukhalfa, C. Béguin, P. Baret, C. Caris, J.-
L. Pierre, Inorg. Chem. 1997, 36, 3898–3910.
[7] F. Biaso, Thesis, Université Joseph Fourier Grenoble, France,
2002.
[8] F. Thomas, C. Béguin, J. L. Pierre, G. Serratrice, Inorg. Chim.
Acta 1999, 291, 148–157.
[9] A.-M. Albrecht-Gary, S. Blanc, F. Biaso, F. Thomas, P. Baret,
G. Gellon, J.-L. Pierre, G. Serratrice, Eur. J. Inorg. Chem. 2003,
2596–2605.
[10] G. Serratrice, F. Biaso, F. Thomas, C. Béguin, Eur. J. Inorg.
Chem. 2004, 1552–1565.
Experimental Section
Materials: All commercial reagents were of the highest purity grade
and used without further purification. Iron(III) stock solutions
were prepared by dissolving the appropriate amount of ferric per-
chlorate hydrate (Aldrich) in standardized HClO4 solution. The
solutions were standardized for ferric ion concentration spectro-
photomerically by using
a molar extinction coefficient of
4160 –1 cm–1 at 240 nm.[11] The ligands were synthesized according
to procedures previously described (L0,[6] LЈ0,[12] L1,[9] L2,[9] and
L3[8]).
[11] R. Bastian, R. Weberling, F. Palilla, Anal. Chem. 1956, 28, 459–
462.
Kinetics Studies: Iron(III) exchange kinetics between FeL0 and L1,
L2, or L3, between FeL3 and L0, L1, or L2, and between Fe-EDTA
[12] D. Imbert, Thesis, Université Joseph Fourier, Grenoble, France,
2000.
Received: March 1, 2007
Published Online: June 21, 2007
Eur. J. Inorg. Chem. 2007, 3681–3685
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org
3685