Functionalization of gold nanoparticles by iron(III) complexes derived from Schiff base ligands

Article abstract of DOI:10.1002/ejic.200800227

A series of iron(III) complexes derived from a variety of pentadentate Schiff base ligands, namely salten [H₂salten = bis(3- salicylideneaminopropyl)amine] derivatives, have been synthesized and further used to stabilize and functionalize gold nanoparticles (Au-NPs). The iron complexes have been specifically designed to interact with the Au-NPs through an ambidentate thiocyanate ligand or through a pendant thiol group carried by the salten derivative ligand. The gold nanocomposites have been synthesized either by direct reduction of Au^III in the presence of iron complexes or by coordination chemistry on prefunctionalized gold nanoparticles; in the latter case Au-NPs bearing pendant empty salten cages have been treated with Fe ^III to generate the iron complexes. Steric and electrostatic repulsions have been tuned to obtain stable colloidal solutions or to induce the precipitation of the gold nanocomposites. Wiley-VCH Verlag GmbH & Co. KGaA, 2008.

Full text of DOI:10.1002/ejic.200800227

FULL PAPER
DOI: 10.1002/ejic.200800227
Functionalization of Gold Nanoparticles by Iron(III) Complexes Derived from
Schiff Base Ligands
Cédric R. Mayer,[a] Gregory Cucchiaro,[a] Josseline Jullien,[a] Frédéric Dumur,[a]
Jérôme Marrot,[a] Eddy Dumas,*[a] and Francis Sécheresse[a]
Keywords: Gold nanoparticles / Schiff base ligands / Iron / Isothiocyanate
A series of iron(III) complexes derived from a variety of pen-
tadentate Schiff base ligands, namely salten [H2salten =
bis(3-salicylideneaminopropyl)amine] derivatives, have been
synthesized and further used to stabilize and functionalize
gold nanoparticles (Au-NPs). The iron complexes have been
specifically designed to interact with the Au-NPs through an
ambidentate thiocyanate ligand or through a pendant thiol
group carried by the salten derivative ligand. The gold nano-
composites have been synthesized either by direct reduction
of AuIII in the presence of iron complexes or by coordination
chemistry on prefunctionalized gold nanoparticles; in the lat-
ter case Au-NPs bearing pendant empty salten cages have
been treated with FeIII to generate the iron complexes. Steric
and electrostatic repulsions have been tuned to obtain stable
colloidal solutions or to induce the precipitation of the gold
nanocomposites.
(© Wiley-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim,
Germany, 2008)
–
–
triphosphate, H2PO4 and HSO4 anions.[14] If the interac-
tion of ferrocene derivatives with metal NPs has been well-
documented, the stabilization of metal NPs with other
types of iron complexes has only been scarcely described:
Au-NPs have been modified with pentacyanoferrates,
through suitable bridging ligands,[15,16] and with electro-
active Prussian blue.[17] Recently, Au-NPs functionalized by
the iron cluster [Fe4(µ3-CO)4(η5-C5H5)3{η5-C5H4CONH-
(CH2)11SH}] have also been used for the redox recognition
of oxo-anions.[18]
Introduction
Recent advances in the preparation of stable noble metal
nanoparticles (NPs) that exhibit tuneable shapes and a high
degree of size monodispersity have contributed to the rapid
development of nanoscience. The ability to synthesize
nanocomposites in aqueous medium or in organic solvents
led to the modification of the surface of metal NPs by a
large variety of stabilizing and functionalizing agents.
Among these are ionic liquids that can act as template, sta-
bilizer and solvent,[1] dendrimers[2] or biomolecules.[3] Com-
paratively, the interaction of metal complexes with metal
NPs has been much less investigated.[4] Metal ions have
been used, for example, to trigger the assembly/disassembly
of metal NPs bearing adapted functional pendant groups
such as carboxylic acid,[5,6] phenol,[7] bipyridine[8] or terpyr-
idine.[9] Such metal-ion-mediated assembly/disassembly of
metal NPs has been elegantly employed for the selective col-
orimetric sensing of metal ions.[5,7,8] Among the various
metal complexes anchored on metal NPs, ruthenium com-
plexes[10–12] and ferrocene derivatives[13] have been the most
employed. The aim was to study the influence of the metal
NPs on the photophysical and/or electrochemical properties
of these metal complexes. Astruc et al. also showed that Au-
NPs functionalized with electroactive ferrocenyl derivatives
could be used for the selective recognition of adenosine-5Ј-
On the other hand, pentadentate Schiff base ligands such
as salten2– [H2salten = bis(3-salicylideneaminopropyl)-
amine] and its derivatives have been extensively used to iso-
late transition-metal complexes with specific magnetic
properties. For example, salten2– was used for the synthesis
of the first non-heme diiron complex singly bridged by a
hydroxy group,[19] and mesalten2– [H2mesalten = bis(3-
salicylideneaminopropyl)methylamine] was employed as
blocking ligand to synthesize tetranuclear,[20] pentanu-
clear[21] and heptanuclear[21,22] Prussian-blue-like molecules.
A series of iron(III) compounds with the general formula
[(salten)Fe(L)]n+ (n = 0, 1) has also been characterized. The
choice of the ligand L occupying the sixth position around
the sixfold-coordinated iron centre is crucial to obtain light-
induced or thermally induced spin change in these com-
plexes.[23–26] The magnetic properties of spin-crossover di-
nuclear iron(III) complexes with the general formula
[(salten)Fe(LЈ)Fe(salten)]2+ have also been studied.[27]
We have recently reported the functionalization of noble
metal NPs with fully conjugated (polypyridyl)ruthenium
complexes.[11,28,29] The aim of this project is to induce a
synergistic effect between the intrinsic properties of the two
[a] Institut Lavoisier de Versailles, UMR 8180, Université de Ver-
sailles,
45 Avenue des Etats-Unis, 78035 Versailles, France
Fax: +33-1-39254381
E-mail: dumas@chimie.uvsq.fr
Supporting information for this article is available on the
WWW under http://www.eurjic.org or from the author.
3614
© 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Inorg. Chem. 2008, 3614–3623

Functionalization of Gold Nanoparticles by Iron(III) Complexes
Scheme 1. Ligands investigated in this work.
components of these nanocomposites. (Polypyridyl)ruthen-
ium complexes were initially chosen owing to their well-
known optical and electrochemical properties while the
fully π-conjugated bridge between the ruthenium centre and
the metal NP was introduced in order to facilitate electronic
transfers. Besides, this family of complexes can be “straight-
forwardly” modified with a variety of functional pendant
groups such as thiol,[12] pyridine,[11,28] phenanthroline,[11,28]
thiophene[11] and mercaptopyridine[29] to ensure their inter-
action with the metal NPs. In the present work, we ex-
tended this approach to another class of metal complexes,
namely salten derivatives, Schiff base complexes of
iron(III). Three new complexes with the general formula
[(L)Fe(NCS)], where L denotes the pentadentate ligands
salten2– (L1), mesalten2– (L2), tolylsalten2– (L3) [H2tolyl-
salten = bis(3-salicylideneaminopropyl)tolylamine], have
been structurally characterized and further used to func-
tionalize gold NPs. The anchoring point of the complexes
with the metal surface was ensured by the pendant sulfur
atom of the ambidentate thiocyanate ligand. The complex
Results and Discussion
Syntheses of the Ligands and Iron Complexes
Schiff base ligands H2L1 and H2L2 were prepared ac-
cording to procedures previously reported.[20,26] N-substi-
tuted Schiff base ligands H2L3 and H2L4 were synthesized
by treating H2L1 with 4-methylbenzyl bromide and bromo-
dodecane, respectively. The ligand H2L5 was isolated in its
disulfide form in two steps, as summarized in Scheme 2.
4,4Ј-dithiobis(benzyl chloride) was prepared from 4,4Ј-di-
thiobis(benzyl alcohol) in the first step, and was then
treated in the second step with two equivalents of H2L1 to
afford (H2L5)2. The disulfide bond was generated to protect
the thiol function during the second step. Disulfide ligands
have already been successfully used to stabilize metal NPs,
undergoing in situ reduction into two thiols capping the
NPs,[30] and in place-exchange reactions on metal NPs.[31]
All the iron complexes were obtained by following the
same procedure: An ethanol solution of one equivalent of
the appropriate salten derivative ligand and two equivalents
sodium methoxide was slowly added to an ethanol solution
containing one equivalent iron(III) nitrate and excess am-
monium thiocyanate. The ligands and iron complexes were
characterized by using electrospray ionization mass spec-
[(dodecanesalten)Fe(NCS)],
where
H2dodecanesalten
(H2L4) = bis(3-salicylideneaminopropyl)dodecanamine, has
then been employed to enhance the stability of the nano-
composites because of the presence of the long alkyl chain.
The fifth salten derivative ligand studied in this work, tolyl-
thiolsalten2– [(H2L5)2] [H2tolylthiolsalten = bis(3-salicyl-
ideneaminopropyl)-p-tolylthiolamine], was specifically de-
signed to strengthen the interaction of the iron complex
with the metal surface. L5 proved to be an effective reagent
for the synthesis of Au-NPs. The thiol end group of the
heteroditopic ligand L5 ensured the interaction with the
1
trometry, IR and H- and 13C NMR spectroscopy (see Ex-
perimental Section and Supporting Information).
Structures of the Iron Complexes
Suitable crystals for single-crystal X-ray diffraction
gold surface by leaving the salten2– pendant group available analysis were obtained for [(L1)Fe(NCS)] (1), [(L2)Fe(NCS)]
to complex Fe3+ ions (Scheme 1).
(2) and [(L3)Fe(NCS)] (3). The molecular structures of the
Scheme 2. Synthetic pathway of (H2L5)2.
Eur. J. Inorg. Chem. 2008, 3614–3623
© 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org
3615

E. Dumas et al.
FULL PAPER
Figure 1. ORTEP representation (ellipsoids drawn at the 30% probability level) of the molecular structures of [(L1)Fe(NCS)] (1), [(L2)-
Fe(NCS)] (2) and [(L3)Fe(NCS)] (3). Hydrogen atoms were omitted for clarity.
three complexes are shown in Figure 1. Selected bond The average iron–ligand bond lengths of 2.065, 2.067 and
lengths and angles are given in Table 1. In the three com- 2.073 Å for 1, 2 and 3, respectively, are consistent with high-
plexes, the iron atom adopts a highly distorted octahedral spin complexes.[36] The Fe–Ophenolato, Fe–Nimino and Fe–
geometry comprising three nitrogen atoms (one Namino and Namino distances in 1, 2 and 3 are well within the ranges
two Nimino) and two oxygen atoms (Ophenolato) of the pen- expected for a high-spin iron(III) centre.[36] Magnetic mea-
tadentate salten ligand. One isothiocyanato ligand com- surements confirmed the magnetic properties inferred from
pletes the coordination sphere to give a neutral molecule. the geometrical characteristics of the complexes (Figure S1,
In the monoiron complexes of general formula [(salten-de- Supporting Information). The only two significant struc-
rivative)Fe(L)] described in the literature, the ligand L is tural differences between the three complexes are the Fe–
trans either to the amino group [L = 4-styrylpyridine,[24] 1- Namino distance, indeed affected by the nature of the R
(pyridine-4-yl)-2-(N-methylpyrrol-2-yl)ethane,[25] 4-methyl- group in the Namino–R-substituted salten ligand (R being
pyridine,[26] 4-aminopyridine,[32] 3,4-dimethylpyridine[33]] or –H, –CH3 and –CH2–C6H4–CH3 in 1, 2 and 3, respec-
to one of the phenolato groups (L = chloride,[21] salicylalde- tively), and the slightly bent Fe–N–C(S) angle [166.9(3),
hyde[34]) of the salten-derivative ligand. The relative posi- 161.5(2) and 152.6(2) in 1, 2 and 3, respectively]. The NCS–
tion of L is crucial, as the steric hindrance induced by the ligands are almost linear in all three complexes.
two phenyl rings of the salten ligand around ligand L is
increased in the former case relative to the latter case. In
complexes 1, 2 and 3, the isothiocyanato ligand is trans to
a phenolato group, hence making the pendant S atom of
the isothiocyanato ligand fully available to interact with a
Preliminary Investigations: Preparation and
Characterization of Au-NPs with [(L1)Fe(NCS)] (1),
[(L2)Fe(NCS)] (2) and [(L3)Fe(NCS)] (3) as Capping Agents
NCS– was inserted in the coordination sphere of the iron
centre of 1, 2 and 3 to ensure the grafting of the metal
complex on the metal NP. NCS– has already been used as
(co)adsorbed probe for Surface Enhanced Raman Spec-
troscopy (SERS) characterization of various metal NPs,[37]
but the isothiocyanato group has only been scarcely em-
ployed as anchoring moiety for the functionalization of
metal NPs either with organic dyes[38] or with a ruthenium
complex.[39] The molecular structures of 1, 2 and 3 had a
similar geometry around the iron(III) centre and also
showed a limited steric hindrance of the salten derivative
around the isothiocyanato ligand. Given such favourable
geometric characteristics, the three complexes were used as
stabilizing agents for the synthesis of metal NPs. Very sim-
ilar findings were obtained with the three complexes and
only the results obtained with 1 are presented in this article.
The direct syntheses, in which AuIII was reduced in the pres-
metal surface. A closer examination of the molecular struc-
tures reveals that the steric hindrance around the isothiocy-
anato ligand is increased in 2, relative to 1 and 3, because
of the orientation of the phenyl ring of the phenolato group
cis to NCS– (Figure 1).
Table 1. Selected bond lengths [Å] and angles [°] in [(L1)Fe(NCS)]
(1), [(L2)Fe(NCS)] (2) and [(L3)Fe(NCS)] (3).
[(L1)Fe(NCS)] [(L2)Fe(NCS)] [(L3)Fe(NCS)]
Fe–O1
Fe–O25
Fe–N(CS)
Fe–N17imino
Fe–N9imino
Fe–N13amino 2.233(2)
Fe–N–C(S)
N–C–S
1.9242(16)
1.9368(18)
2.061(2)
2.108(2)
2.126(2)
1.9259(16)
1.9346(15)
2.0985(19)
2.0829(18)
2.0938(19)
2.2647(18)
161.5(2)
1.9161(13)
1.9319(14)
2.0881(18)
2.0827(16)
2.1090(15)
2.3113(15)
152.62(17)
178.47(19)
166.9(3)
179.1(3)
179.0(2)
It is now well established that, in ferric Schiff base com- ence of 1 by excess NaBH4, were performed in dmf (see
plexes (among others), the iron–ligand distances are notice- Experimental Section). The syntheses were realized by
ably affected by the spin state of the iron(III) centre.[24,35,36] using various [AuIII]/[1] ratios between 0.25 and 6. The sta-
3616
www.eurjic.org
© 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Inorg. Chem. 2008, 3614–3623

Functionalization of Gold Nanoparticles by Iron(III) Complexes
bility of the colloidal solutions and the size and shape of of the salten ligand. Stable colloidal solutions of Au-NPs
the final nanocomposites were not noticeably affected by functionalized by [(L4)Fe(NCS)] (4) were obtained, in dmf,
this ratio. The colloidal solutions were first studied by UV/ by reduction of HAuCl4 with excess NaBH4 in the presence
Vis spectroscopy. After reduction of AuIII with NaBH4, a of 4 (see Experimental Section). Excess iron complex was
maximum absorption peak was observed in the visible re- removed by repeated centrifugations of the colloidal solu-
gion at ca. 520 nm. The study was complicated by the pres- tions and redispersion of the nanocomposites in dmf. The
ence of the LMCT band of 1 centred at 513 nm and there- peak observed at ca. 520 nm in the absorption spectrum of
fore located in the region expected for the SPR (surface the colloidal solution was associated with the sum of the
plasmon resonance) band of Au-NPs with a diameter of plasmon resonance of the Au-NPs and the MLCT band of
few nanometres. Consequently, the peak at ca. 520 nm was 4 (Figure S3, Supporting Information). Nearly spherical
attributed to the sum of the absorption of the Au-NPs and monodisperse Au-NPs with an average diameter of
(SPR) and the LMCT of 1, explaining the slight redshift 4.6Ϯ0.7 nm were observed by TEM analysis. TEM images
and the slight increase in the intensity of the peak relative also showed regions in which Au-NPs adopt a hexagonal
to that of the peak of pure 1 in the same conditions (Fig- close-packed (hcp) arrangement (see Figure 2). In these lo-
ure S2, Supporting Information). The presence of Au-NPs cal superstructures, an average distance of 1.7Ϯ0.4 nm was
was further confirmed by TEM. TEM analysis showed that measured between Au-NPs. This distance is relatively short
Au-NPs polydisperse in size (between 3 and 12 nm) and in in view of the fact that the capping complex 4 contains a
shape were formed, but they aggregated or fused in a few dodecyl chain connected to a bulky {(NCS)Fe(salten)} moi-
minutes to give larger particles and aggregates. Aggregates ety. Such a short distance necessarily implies chain bundles
of several tens of nanometres were observed after 4 min.
of adjacent iron complexes, highly tilted and/or interdigi-
These first results clearly show that iron complexes 1, tated dodecyl chains of neighbouring Au-NPs.[41]
2 and 3 are not adapted for the functionalization of Au
nanocomposites. At first, the iron complexes perfectly play
their role of stabilizing agent as they clearly prevent further
growth of Au-NPs. Nevertheless, the stability of the Au
nanocomposites is only partial, as the Au-NPs rapidly ag-
gregate and fuse. Several phenomena could explain the ra-
pid aggregation of the NPs: (i) A weak interaction of the
pendant thiocyanato ligand with the gold surface, leading
to rapid leaching of the iron complexes. (ii) A rupture of
the SCN–Fe bond resulting from the interaction of the
thiocyanato ligand with the gold surface. Such behaviour
has already been observed with the labile [Fe(CN)5-
(2MPy)]3– complex (2MPy = 2-mercaptopyridine) that
acted as a source delivering 2MPy to the Au-NPs.[15] (iii)
The absence of sufficient steric repulsion by the surface-
bonded metal complexes to ensure dispersion stability. (iv)
The absence of electrostatic repulsion preventing aggrega-
tion of the nanocomposites, as the capping agents are neu-
tral iron complexes. The first two assumptions have been
ruled out by EDX analyses of the repeatedly washed pre-
cipitates. These analyses unambiguously showed the pres-
ence of iron complexes in these aggregates. We therefore
decided to modify the iron complexes either to enhance the
steric repulsion between nanocomposites or to generate an
electrostatic repulsion induced by charged iron complexes.
Figure 2. (a) TEM image of Au-NPs functionalized by [(L4)-
Fe(NCS)] (4) and (b) the corresponding histogram.
Functionalization of Au-NPs by [(L4)Fe(NCS)]
Capping agents bearing a long alkyl chain have often
been used to stabilize metal NPs. The role of this long alkyl
chain is to induce steric repulsion against the van der Waals
attraction between metal NPs and therefore to prevent them
from further growth and aggregation. Such an approach
Functionalization of Au-NPs by [(L5)Fe(X)]n+ (X = Solvent
Molecule, n = 1; X = NCS–, n = 0): Coordination
Chemistry on Prefunctionalized NPs
was elegantly employed for the well-known Brust–Schiffrin
A completely different approach was used to graft the
method.[40] The ligand H2L4 was specifically designed to iron complex [(L5)Fe(X)]n+ (X = solvent molecule, n = 1; X
carry a pendant dodecyl chain, grafted on the Namino atom = NCS–, n = 0) on Au-NPs. We first changed the type and
Eur. J. Inorg. Chem. 2008, 3614–3623
© 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org
3617

E. Dumas et al.
FULL PAPER
the position of the anchoring point, ensuring the connec- explained by the partial agglomeration of the Au-NPs in
tion between the iron complexes and the Au-NPs. The pen- EtOH/H2O. Figure 3 shows a TEM image of AuL5-NPs,
dant S atom of the isothiocyanato ligand used for com- purified and redispersed in dmf. Nearly spherical and mon-
plexes 1–4 was replaced by a thiol linker, well-known for its odisperse Au-NPs were observed with a mean diameter of
strong affinity for Au surfaces. In addition, this thiol linker 3.2Ϯ0.5 nm. The smaller size obtained for the Au-NPs sta-
was part of the salten derivative ligand L5, while in 1–4, the bilized with L5 relative to the ones stabilized with [(L4)-
isothiocyanato ligand was located on the sixth position left Fe(NCS)] (4) can be explained by the stronger interaction
available by the salten derivative ligand in the coordination of the thiol-terminated ligand L5 with the gold surface rela-
sphere of the iron centre. L5 was therefore perfectly adapted tive to the isothiocyanato pendant group of 4.
for a post-functionalization process. The idea was to first
AuL5-NPs in dmf and in EtOH/H2O were then treated
stabilize Au-NPs with L5 and then to insert FeIII into the with increasing amounts of FeCl3·6H2O in order to insert
pendant salten cage of the nanocomposites thus generated. iron(III) centres in the pendant salten cages of the AuL5-
Au-NPs stabilized by L5 (AuL5-NPs) were obtained NPs. The reaction was followed by UV/Vis spectroscopy.
through the direct reduction of a mixture of the disulfide When the reaction was carried out in dmf, addition of FeIII
(H2L5)2 ligand and AuIII with excess NaBH4 in dmf. The led to a slight blueshift from 523 to 517 nm; a slight nar-
Au-NPs were purified by repeated precipitations and redis- rowing and a slight increase in the intensity of the band
persions (see Experimental Section). The purification pro- were observed in the visible region of the spectra (Fig-
cess was facilitated by the solubility of the organic ligand ure S5, Supporting Information). The slight increase in the
in dichloromethane and the insolubility of the function- intensity of the peak was explained by the addition of the
alized Au-NPs in the same solvent. The purified Au-NPs MLCT band of [(L5)Fe(solv.)]+, centred at 519 nm in the
were finally suspended either in dmf or in a EtOH/H2O absorption spectrum of pure [(L5)Fe(solv.)]+, to the SPR
(v/v, 1:1) mixture. Stable colloidal solutions were obtained band of the Au-NPs. The position of this band led one to
in dmf, while partial agglomeration of the nanoparticles assume that the Au-NPs were kept basically unchanged af-
was observed in the EtOH/H2O mixture. These results were ter the insertion of FeIII in the pendant salten cages. This
confirmed by UV/Vis spectroscopy. The absorption spec- assumption was confirmed by TEM analysis (see Figure 4).
trum of dmf and EtOH/H2O solutions showed a single peak Nearly spherical Au-NPs with an average diameter of
in the visible region at 523 nm and 534 nm, respectively 3.4Ϯ0.5 nm were observed, very close to the size measured
(Figure S4, Supporting Information). This band was as- for AuL5-NPs. Lattice fringes observed in the high-resolu-
signed to the SPR of Au-NPs. The broadening and the red- tion TEM images showed that the majority of the Au-NPs
shift of this SPR band in EtOH/H2O relative to dmf were are single crystalline with only a small number of twinned
Figure 3. (a) TEM image of Au-NPs functionalized by L5 in dmf
and (b) the corresponding histogram.
Figure 4. (a) TEM image of AuL5-NPs after reaction with FeIII in
dmf and (b) the corresponding histogram.
3618
www.eurjic.org
© 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Inorg. Chem. 2008, 3614–3623

Functionalization of Gold Nanoparticles by Iron(III) Complexes
crystals (Figure S6, Supporting Information). When the re- ure 6). The redshift and the broadening of the SPR band
action was performed in EtOH/H2O, addition of FeIII to observed in the absorption spectrum were indeed explained
preformed AuL5-NPs significantly altered the electronic by the aggregation phenomenon and thus the delocalization
spectrum of the colloidal solution. As shown in Figure 5, of the plasmon wave through adjacent Au-NPs.[43] Alterna-
insertion of FeIII in the pendant salten cages induced a tive explanations could also justify the aggregation of the
slight increase in the intensity of the band in the visible Au-NPs observed after addition of excess ammonium thio-
region, once more explained by the superposition of the cyanate. (i) Thiocyanate ligands are known to bind to the
MLCT band of [(L5)Fe(solv.)]+ and the SPR band of Au- surface of Au-NPs and could therefore remove the posi-
NPs. Addition of increasing amounts of FeIII also induced tively charged iron complexes from the surface of the Au-
a marked narrowing of this band concomitant with a blue- NPs and thus induce the aggregation of the Au-NPs. Never-
shift from 534 nm (0 µL of FeIII solution) to 511 nm (45 µL theless, the affinity of thiocyanate for the surface of Au-
of FeIII solution, see Experimental Section). Further ad- NPs is much weaker than the affinity of thiol groups, such
dition of FeIII did not modify the intensity and the position as the one used in ligand L5, for the surface of Au-NPs.
of the band at 511 nm, implying that all the salten cages The following experiment was realized in order to rule out
were filled. The narrowing and the blueshift can be readily this conceivable explanation: excess ammonium thiocyanate
explained by the disassembly of the pristine agglomerated was added to a stable colloidal solution of Au-NPs func-
AuL5-NPs in EtOH/H2O. Insertion of FeIII in the pendant tionalized by ligand L5. No aggregation of the function-
salten cages of AuL5-NPs generated charged the capping alized AuL5-NPs was observed after addition of this excess
iron complexes [(L5)Fe(solv.)]+ grafted on the surface of the ammonium thiocyanate. The place-exchange reaction be-
Au-NPs, and therefore ensured sufficient electrostatic repul- tween [(L5)Fe(solv.)]+ and NCS– can not, therefore, account
sion between nanocomposites to disassemble the original for the aggregation observed in our initial experiment. (ii)
agglomerates. In the final nanocomposites, the sum of the A simple change in the ionic strength of the solution after
plasmon resonance of the Au-NPs and the MLCT band of addition of the excess ammonium thiocyanate could also
[(L5)Fe(solv.)]+ was observed at 511 nm and 517 nm in explain the aggregation observed. Excess ammonium ni-
EtOH/H2O and in dmf, respectively. This shift could be at- trate was therefore added to a stable colloidal solution of
tributed to the difference in the refractive index of the sol- Au-NPs functionalized by [(L5)Fe(solv.)]+ in EtOH/H2O. A
20
vent used in this study (nd = 1.333, 1.361 and 1.431 for small amount of precipitate was observed after several min-
H2O, EtOH and dmf, respectively). A redshift is actually utes, which showed that modification of the ionic strength
expected when the refractive index of the solvent in- partly explains the aggregation observed but can not ac-
creases.[42]
count for the rapid and complete aggregation observed in
our initial experiment, after addition of excess ammonium
thiocyanate.
Figure 5. Absorption spectra following the reaction of AuL5-NPs
with increasing amounts of FeIII (see Experimental Section for fur-
ther details) in EtOH/H2O. 0 µL (solid line), 15 µL (dashed line),
30 µL (dotted line), 45 µL (empty circles).
Figure 6. Absorption spectra of preformed Au-NPs functionalized
by [(L5)Fe(solv.)]+ in EtOH/H2O before (solid line) and after (dot-
ted line) addition of excess NH4NCS.
The crucial role of electrostatic repulsions in obtaining
stable colloidal solutions was further evidenced by the fol-
lowing experiment. Excess ammonium thiocyanate was
treated with a stable colloidal solution of Au-NPs function-
alized by [(L5)Fe(solv.)]+ in EtOH/H2O. Thiocyanate was
used to replace the solvent molecule in the coordination
Conclusions
We have successfully developed different approaches to
sphere of the iron complex to generate the neutral complex stabilize and functionalize gold nanoparticles by a new class
[(L5)Fe(NCS)]. Addition of thiocyanate rapidly induced ag- of iron species, namely salten derivatives of iron(III) com-
gregation and precipitation of the nanocomposites as a con- plexes. We have been able to vary the nature of the anchor-
sequence of the neutralization of the surface charge (Fig- ing point between the iron complex and the Au-NP (thiol
ure S7, Supporting Information). Aggregation of the Au- or thiocyanate), the relative position of this bridging unit
NPs was also monitored by UV/Vis spectroscopy (Fig- in the iron complex [directly connected to the iron(III) cen-
Eur. J. Inorg. Chem. 2008, 3614–3623
© 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org
3619

E. Dumas et al.
FULL PAPER
2 mmol) were solubilized in acetonitrile (60 mL). K2CO3 (1.65 g,
12 mmol) and KI (0.100 g, 0.6 mmol) were then added, and the
resulting mixture was heated at reflux overnight. After filtration,
the solvent was removed under vacuum. The crude ligand was iso-
lated as a yellow oil in 86% yield and was treated with iron(III)
tre or attached to the salten derivative ligand] and the route
used for grafting the iron complex on the Au-NPs (direct
synthesis or coordination chemistry on prefunctionalized
Au-NPs). Steric and electrostatic repulsions have been used
to obtain stable colloidal solutions in water and in organic
solvents. Such a variety of adaptable parameters of synthe-
sis gives great flexibility to envision the functionalization of
metal NPs by numerous iron and other metal complexes
derived from Schiff base ligands. This work will indeed be
1
without further purification. H NMR (300 MHz, CDCl3, 25 °C):
δ = 1.91 (quin, J = 6.7 Hz, 4 H, CH2–CH2–CH2), 2.37 (s, 3 H,
CH3), 2.57 [t, J = 6.7 Hz, 4 H, CH2–N(CH2)2], 3.58 (s, 2 H, N–
CH2–Ph), 3.65 (t, J = 6.6 Hz, 4 H, C=N–CH2), 6.75 (t, J = 7.5 Hz,
2 H, Harom HC=C–OH), 6.84 (d, J = 8.3 Hz, 2 H, Harom N=CH–
extended to metal Schiff base complexes with specific mag- C–CH), 6.99 (d, J = 7.7 Hz, 2 H, Harom tolyl), 7.07–7.11 (m, 4 H,
H
arom), 7.18 (d, J = 7.9 Hz, 2 H, Harom tolyl), 8.13 (s, 2 H, CH=N),
netic properties. For example, it is well-known that the spin-
crossover properties can be modulated by the 3D packing
of the individual complexes.[36] Spin-crossover metal com-
plexes will therefore be confined around metal NPs, as only
capping components or diluted by magnetically neutral alk-
13.60 (br. s, 2 H, OH) ppm. ESI-MS (negative ion mode): calcd.
for [H2L3 – H] 442.3; found 442.2.
Bis(3-salicylideneaminopropyl)dodecylamine,
H2L4): 1-Bromododecane (560 µL, 2 mmol), K2CO3 (1.65 g,
(H2dodecanesalten,
ylthiol, in order to analyze the influence of the metal NPs 12 mmol) and KI (0.100 g, 0.6 mmol) were added to a solution of
H2L1 (0.644 g, 1.9 mmol) in butanone (60 mL). The resulting mix-
ture was heated at reflux overnight. After removal of the solvent
under vacuum, the residue was purified by column chromatography
(SiO2): dichloromethane was used as first eluant to remove unre-
acted bromododecane, while AcOEt was used to recover the pure
product. H2L4 was isolated as a yellowish oil in 45% yield (0.460 g).
1H NMR (300 MHz, CDCl3, 25 °C): δ = 0.93 (t, J = 6.9 Hz, 3 H,
CH3), 1.30 (m, 20 H, CH2), 1.89 (quin, J = 6.9 Hz, 4 H, CH2–
CH2–CH2), 2.45 (t, J = 7.1 Hz, 2 H, N–CH2 from C12H25), 2.58 [t,
J = 6.9 Hz, 4 H, CH2–N(CH2)2], 3.66 (t, J = 6.7 Hz, 4 H, C=N–
CH2), 6.90 (t, J = 7.5 Hz, 2 H, Harom HC=C–OH), 7.00 (d, J =
5.6 Hz, 2 H, Harom N=CH–C–CH), 7.25 (dd, J = 1.3, J = 7.5 Hz,
2 H, Harom), 7.34 (td, J = 8.7, J = 1.5 Hz, 2 H, Harom), 8.37 (s, 2
H, CH=N), 13.60 (s, 2 H, OH) ppm. 13C NMR (75 MHz, CDCl3,
25 °C): δ = 14.2 (CH3), 22.7 (CH3–CH2), 27.2, 27.7, 28.5, 29.4,
29.7, 29.74, 32.0, 51.4 [N–(CH2)2], 54.1 (N–CH2), 57.5 (C=N–
CH2), 117.0 (CH=C–OH), 118.5 (C–C=N), 188.8 (CH=CH–C–
C=N), 131.1 (CH=CH–C–C=N), 132.1 (CH=CH–C–OH), 161.4
(C–OH), 164.9 (C=N) ppm. ESI-MS (positive ion mode): calcd.
for [H2L4 + H] 508.4; found 508.6.
on their magnetic properties.
Experimental Section
All reagents and solvents were purchased from Aldrich and used
without further purification. H2salten (H2L1),[26] H2mesalten
(H2L2),[44] 4,4Ј-dithiobis(benzyl alcohol)[45] and 4,4Ј-dithiobis(ben-
zyl chloride)[46] were synthesized by following procedures pre-
viously reported, without modification and with similar yields. ESI-
MS measurements were carried out with an API 3000 (ESI/MS/
MS) PE-SCIEX triple quadrupole mass spectrometer and a HP
5989B single quadrupole mass spectrometer equipped with an elec-
trospray source from Analytica of Branford. For the API 3000
(ESI/MS/MS) PE-SCIEX triple quadrupole mass spectrometer, the
experiments were performed either by direct infusion with a syringe
pump (with a flow rate of 10 µLmin–1) or by flow injection acqui-
sition (with a flow rate of 200 µLmin–1). Standard experimental
conditions were as follows: sample concentration: 10–3 to
10–5 molL–1, nebulizing gas N2: 7 units flow rate, ion spray voltage:
–5.00 kV, temperature: 200–400 °C, declustering potential: –20 V,
focusing potential: –200 V, entrance potential: 10 V. IR spectra
(4000–250 cm–1) were recorded with a Nicolet Magna 550 spec-
trometer (from KBr pellets). Magnetization studies were carried
out on a pellet of ground crystals by using a Quantum Design
MPMS 5 magnetometer working in the dc mode in the temperature
range 300 K–5 K within a magnetic field of 0.5 T and a sweeping
rate of 1 Kmin–1. UV/Vis spectra were recorded with a Perkin–
Elmer UV/Vis/NIR lambda 19 PC scanning spectrophotometer.
Each colloidal solution was analyzed similarly, being introduced
directly from the mother solution into the cell (optical path of
0.2 cm). 1H and 13C NMR spectra were obtained at room tempera-
ture in 5 mm-o.d. tubes with a Bruker Avance 300 spectrometer
equipped with a QNP probe head. The following standard abbrevi-
ations are used for the characterization of 1H NMR signals: s =
singlet, d = doublet, t = triplet, quin = quintet, m = multiplet, dd
= doublet of doublet, td = triplet of doublet. Elemental analyses
were realized by the “Service central d’analyse du CNRS”, Vernai-
son (France). Energy-dispersive spectrometry was performed with
a PGT-IMIX-PC, and the visualization of metal nanoparticles was
performed with a Transmission Electron Microscope (microscope
JEOL 2010 UHR and JEOL 100 CX2) in the “Centre Régional de
Mesures Physiques” Paris 6: the visualization was performed by
drying a drop of the mother solution on a copper grid.
Bis(3-salicylideneaminopropyl)-p-tolylthiolamine
[(H2tolylthiol-
salten)2, (H2L5)2]: K2CO3 (1.65 g, 12 mmol) and KI (0.100 g,
0.6 mmol) were added to a solution of 4,4Ј-dithiobis(benzyl chlo-
ride) (0.516 g, 1.5 mmol) in acetonitrile (240 mL). H2L1 (1 g,
3 mmol) was then added, and the resulting mixture was heated at
reflux overnight. After filtration, the solvent was removed under
vacuum and the product was purified by flash chromatography
(SiO2, AcOEt). (H2L5)2 was isolated as a yellow oil in 91% yield
1
(1.26 g). H NMR (300 MHz, CDCl3, 25 °C): δ = 1.87 (quin, J =
7.0 Hz, 8 H, CH2–CH2–CH2), 2.59 [t, J = 7.0 Hz, 8 H, CH2–
N(CH2)2], 3.64 (t, J = 6.1 Hz, 8 H, C=N–CH2), 3.69 (s, 4 H, CH2–
Ph), 6.82 (t, J = 7.9 Hz, 4 H, Harom CH=C–OH), 6.92 (d, J =
8.8 Hz, 4 H, Harom thiophenol), 7.11 (t, J = 7.9 Hz, 6 H) 7.20–7.31
(m, 6 H), 7.64 (dd, J = 7.0, J = 1.7 Hz, 4 H), 8.3 (s, 4 H, CH=N),
13.5 (s, 4 H, OH) ppm. 13C NMR (75 MHz, CDCl3, 25 °C): δ =
27.9 (CH2–CH2–CH2), 50.8 (N–CH2–CH2–CH2), 57.5 (CH2–CH2–
CH2–N=C), 58.4 (N–CH2–Ph), 117.0 (CH=C–OH), 117.7 (C–
C=N), 118.4 (CH=CH–C–C=N), 126.3 (S–C=CH–CH=C), 128.6
(S–C=CH–CH=C), 129.9 (CH–C–C=N), 131.2 (CH=CH–C–OH),
137.9 (S–C=CH–CH=C), 138.3 (C–S), 161.3 (C–OH), 164.9 (C=N)
ppm. ESI-MS (negative ion mode): calcd. for [(H2L5)2 – H] 920.2;
found 919.9, calcd. for [(H2L5)2 – 2H] 459.6; found 460.6.
[(L1)Fe(NCS)] (1), [(L2)Fe(NCS)] (2), [(L3)Fe(NCS)] (3), [(L4)-
Bis(3-salicylideneaminopropyl)tolylamine, (H2tolylsalten, H2L3): 4- Fe(NCS)] (4) and [(L5)Fe(NCS)]2: The five complexes were synthe-
methylbenzyl bromide (279 µL, 2 mmol) and H2L1 (0.679 g, sized by following similar procedures. Fe(NO3)3·9H2O (0.404 g,
3620
www.eurjic.org
© 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Inorg. Chem. 2008, 3614–3623

Functionalization of Gold Nanoparticles by Iron(III) Complexes
1 mmol) and NH4NCS (0.457 g, 6 mmol) were solubilized in etha-
nol (50 mL) to give solution S1. The appropriate salten derivative
ligand [H2L1: 0.339 g, 1 mmol; H2L2: 0.353 g, 1 mmol; H2L3:
0.444 g, 1 mmol; H2L4: 0.508 g, 1 mmol; (H2L5)2: 0.461 g,
332 (12110), 296 (sh, 9185) nm. ESI-MS (positive ion mode): calcd.
for [M – NCS]+ 497.4; found 497.3. [(L4)Fe(NCS)] (4): IR (KBr):
ν = 2922 (s), 2852 (s), 2052 (vs, CN thiocyanate), 1647 (m), 1619
˜
(s, CN imine), 1544 (s), 1469 (s), 1446 (s), 1309 (m, CO), 1150 (m),
0.5 mmol] and NaOCH3 (0.108 g, 2 mmol) were mixed together in 758 (m) cm–1. ESI-MS (positive ion mode): calcd. for [M – NCS]+
ethanol (50 mL), or in butanone for H2L4, to give solution S2. Solu-
tion S1 was then slowly added to solution S2, and the resulting
solution was stirred at room temperature for 30 min. Crystals of 1,
2 and 3 were obtained by slow evaporation of the solvent. Yields
of about 40% were obtained for the three complexes. For [(L4)-
Fe(NCS)] (4), the solution was concentrated to ca. 50 mL. Complex
4 was recovered after extraction with a mixture of diethyl ether
(2ϫ100 mL) and water (2ϫ20 mL). The organic phases were then
combined, dried with MgSO4, and the solvent was evaporated to
afford 4 (80% yield). [(L5)Fe(NCS)]2 precipitated from the solution
by slow evaporation of the solvent (95% yield). [(L1)Fe(NCS)] (1):
C21H23FeN4O2S (451.34): calcd. C 55.88, H 5.14, Fe 12.37, N
12.41, S 7.10; found C 55.27, H 5.09, Fe 12.15, N 12.44, S 6.72. IR
561.6; found 561.4. [(L5)Fe(NCS)]2: ESI-MS (positive ion mode):
calcd. for [M – NCS]+ 1087.0; found 1086.6.
Au-NPs Coated by 1, 2 and 3: Syntheses of the nanocomposites
were identical with those of the three iron complexes. The ratio
[AuIII]/[FeIII] was varied between 0.25 and 6 without noticeable in-
fluence on the final size and shape of the nanocomposites. In a
typical experiment, a dmf solution (2 mL) of HAuCl4·3H2O
(5.8 mg, 0.015 mmol) was added to a solution of 1 (6.8 mg,
0.015 mmol) in dmf (8 mL). An aqueous solution of NaBH4
(0.140 mL, 0.4 ) was then added at once to the previous mixture.
The resulting solution was stirred for 15 min at room temperature.
The solution was analyzed by UV/Vis spectroscopy and TEM.
Au-NPs Stabilized by 4: A dmf solution (2 mL) of HAuCl4·3H2O
(5.8 mg, 0.015 mmol) and an aqueous solution of NaBH4
(0.140 mL, 0.4 ) were successively added, at once, to a solution of
4 (9.3 mg, 0.015 mmol) in dmf (8 mL). The resulting solution was
stirred 15 min at room temperature. The dark red colloidal solution
was then repeatedly centrifuged for 20 min at 14000 rpm at 5 °C,
and the nanocomposites were redispersed in dmf (10 mL) in order
to remove excess 4.
(KBr): ν = 3236 (w, NH), 2053 (vs, CN thiocyanate), 1622 (vs, CN
˜
imine), 1597 (s), 1542 (s), 1468 (s), 1444 (s), 1401 (m), 1387 (m),
1335 (m), 1304 (s, CO), 1204 (m), 1149 (m), 1089 (m), 1048 (m),
887 (m), 787 (m), 778 (m), 754 (m), 431 (m), 276 (m) cm–1. UV/Vis
(dmf): λ (ε /Lmol–1 cm–1) = 513 (3200), 402 (sh, 2895), 329 (7870),
298 (sh, 5815) nm. ESI-MS (positive ion mode): calcd. for [M –
NCS]+ 393.2; found 393.3. [(L2)Fe(NCS)] (2): C22H25FeN4O2S
(465.4): calcd. C 56.78, H 5.41, Fe 12.00, N 12.04, S 6.89; found C
56.64, H 5.43, Fe 11.73, N 12.11, S 6.49. IR (KBr): ν = 2917 (w), Au-NPs Stabilized by L5: A solution of (H2L5)2 in dichloromethane
˜
2890 (w), 2865 (w), 2062 (vs, CN thiocyanate), 1620 (vs, CN imine), (0.130 mL, 0.11 ) was poured into dmf (8 mL). To this solution
1598 (s), 1540 (s), 1467 (s), 1443 (s), 1398 (m), 1305 (s, CO), 1148 were then successively added a solution of HAuCl4·3H2O (5.8 mg,
(m), 761 (s), 594 (m), 440 (m) cm–1. UV/Vis (dmf): λ (ε / 0.015 mmol) in dmf (2 mL) and an aqueous solution of NaBH4
Lmol–1 cm–1) = 512 (2645), 406 (sh, 3310), 323 (10425), 296 (sh,
8900) nm. ESI-MS (positive ion mode): calcd. for [M – NCS]+
407.3; found 407.3. [(L3)Fe(NCS)] (3): C29H31FeN4O2S (555.49):
calcd. C 62.71, H 5.63, Fe 10.05, N 10.09, S 5.77; found C 62.35,
(0.140 mL, 0.4 ). The resulting red solution was stirred for 15 min
at room temperature. The AuL5-NPs were then purified from free
organic ligands by using the following protocol. A mixture of
dichloromethane (60 mL) and water (5 mL) was added to the pre-
vious stable colloidal solution. AuL5-NPs precipitated, while the
H 5.71, Fe 9.98, N 9.93, S 5.31. IR (KBr): ν = 2921 (m), 2897 (m),
˜
2057 (vs, CN thiocyanate), 1620 (vs, CN imine), 1541 (vs), 1466 non-grafted organic ligands remained in solution. The precipitate
(s), 1444 (vs), 1395 (m), 1340 (m), 1301 (s, CO), 1201 (m), 1148 was then recovered by centrifugation and redispersed in dmf
(m), 1124 (m), 1090 (m), 1047 (m), 757 (s), 742 (m), 596 (s), 451 (10 mL). The above treatment was then repeated in order to remove
(m), 431 (m) cm–1. UV/Vis (dmf): λ (ε /Lmol–1 cm–1) = 508 (3940),
traces of free (H2L5)2. Purified AuL5-NPs were finally redispersed
Table 2. Crystallographic data for [(L1)Fe(NCS)] (1), [(L2)Fe(NCS)] (2) and [(L3)Fe(NCS)] (3).
[(L1)Fe(NCS)] (1)
[(L2)Fe(NCS)] (2)
[(L3)Fe(NCS)] (3)
Empirical formula
Formula weight
Crystal system
Space group
T [K]
a [Å]
b [Å]
C21H23FeN4O2S
451.34
orthorhombic
Pna21
C22H25FeN4O2S
465.37
orthorhombic
Pbca
C29H31FeN4O2S
555.49
monoclinic
P21/n
293(2)
296(2)
293(2)
13.5764(6)
16.6768(7)
9.3723(3)
90
12.9976(3)
16.7318(4)
20.2176(4)
90
10.1055(3)
15.1459(4)
17.5384(6)
90
c [Å]
α [°]
β [°]
90
90
90
90
92.623(2)
90
2681.56(14)
4
1.376
0.674
γ [°]
V [Å3]
2121.99(15)
4
1.413
4396.79(17)
8
1.406
Z
ρcalcd. [g cm–3]
µ [mm–1]
0.833
0.807
F(000)
940
19896
6026 (Rint = 0.0357)
1.022
0.0371, 0.0832
0.0702, 0.0962
0.379, –0.338
1944
83482
6446 (Rint = 0.0528)
1.095
0.0361, 0.1008
0.0830, 0.1365
0.649, –0.721
1164
67913
7847 (Rint = 0.0362)
1.079
0.0407, 0.1125
0.0576, 0.1233
0.467, –0.330
Reflections collected
Independent reflections
Goodness-of-fit on F2
R1, wR2 [IϾ2σ(I)]
R1, wR2 [all data]
∆ρ(max), ∆ρ(min) [eÅ3]
Eur. J. Inorg. Chem. 2008, 3614–3623
© 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org
3621

E. Dumas et al.
FULL PAPER
[10]
[11]
a) Y. Kuwahara, T. Akiyama, S. Yamada, Langmuir 2001, 17,
5714–5716; b) P. Pramod, P. K. Sudeep, K. G. Thomas, P. V.
Kamat, J. Phys. Chem. B 2006, 110, 20737–20741; c) W. Hu-
ang, G. Masuda, S. Maeda, H. Tanaka, T. Hino, T. Ogawa,
Inorg. Chem. 2008, 47, 468–480.
C. R. Mayer, E. Dumas, F. Miomandre, R. Méallet-Renault, F.
Warmont, J. Vigneron, R. Pansu, A. Etcheberry, F. Sécheresse,
New J. Chem. 2006, 30, 1628–1637.
either in dmf (10 mL), in which case the solution was denoted Sdmf
or in a mixture of ethanol (5 mL) and water (5 mL), in which case
the solution was denoted Saq.
Reactivity of AuL5-NPs with FeIII: Portions (0.015 mL) of solutions
of FeCl3·6H2O in dmf (0.04 ) and in EtOH/H2O (v/v, 1:1, 0.02 )
were added to Sdmf and Saq, respectively. The reaction of complex-
ation was followed by UV/Vis spectroscopy.
,
[12]
[13]
M. Jebb, P. K. Sudeep, P. Pramod, K. G. Thomas, P. V. Kamat,
J. Phys. Chem. B 2007, 111, 6839–6844.
X-ray Diffraction Analyses: X-ray intensity data were collected with
a Bruker X8-APEX2 CCD area-detector diffractometer by using
Mo-Kα radiation (λ = 0.71073 Å). Three sets of narrow data frames
(90 s per frame) for 1, four sets of narrow data frames (240 s per
frame) for 2, and six sets of narrow data frames (30 s per frame)
for 3 were collected at different values of θ, for two and one initial
values of φ and ω, respectively, for 1 (two and two initial values for
2 and three and three initial values for 3), by using 0.5° increments
of φ and ω. Data reduction was accomplished with SAINT,
v7.03.[47] The substantial redundancy in data (4.56 for 1, 9.37 for
2 and 6.46 for 3) allowed a semiempirical absorption correction
(SADABS v2.10)[47] to be applied on the basis of multiple measure-
ments of equivalent reflections. The structure was solved by direct
methods, developed by successive different Fourier syntheses and
refined by full-matrix least-squares on all F2 data by using
SHELXTL v6.14.[48] Hydrogen atoms were included in calculated
positions and allowed to ride on their parent atoms. The crystallo-
graphic data are summarized in Table 2.
a) J. Wang, J. Li, A. J. Baca, J. Hu, F. Zhou, W. Yan, D. W.
Pang, Anal. Chem. 2003, 75, 3941–3945; b) M. Yamada, H.
Nishihara, Langmuir 2003, 19, 8050–8056; c) R. L. Wolfe, R.
Balasubramanian, J. B. Tracy, R. W. Murray, Langmuir 2007,
23, 2247–2254.
D. Astruc, M.-C. Daniel, J. Ruiz, Chem. Commun. 2004, 2637–
2649.
F. S. Nunes, L. D. S. Bonifacio, K. Araki, H. E. Toma, Inorg.
Chem. 2006, 45, 94–101.
S. H. Toma, J. A. Bonacin, K. Araki, H. E. Toma, Eur. J. Inorg.
[14]
[15]
[16]
[17]
Chem. 2007, 3356–3364.
a) S. S. Kumar, J. Joseph, K. L. Phani, Chem. Mater. 2007, 19,
4722–4730; b) J. D. Qiu, H. Z. Peng, R. P. Liang, J. Li, X. H.
Xia, Langmuir 2007, 23, 2133–2137.
J. R. Aranzaes, C. Belin, D. Astruc, Chem. Commun. 2007,
3456–3458.
[18]
[19]
J. Jullien, G. Juhasz, P. Mialane, E. Dumas, C. R. Mayer, J.
Marrot, E. Riviere, E. L. Bominaar, E. Munck, F. Secheresse,
Inorg. Chem. 2006, 45, 6922–6927.
G. Rogez, A. Marvilliers, E. Rivière, J.-P. Audière, F. Lloret, F.
Varret, A. Goujon, N. Mendenez, J.-J. Girerd, T. Mallah, An-
gew. Chem. Int. Ed. 2000, 39, 2885–2887.
G. Rogez, A. Marvilliers, P. Sarr, S. Parsons, S. J. Teat, L.
Ricard, T. Mallah, Chem. Commun. 2002, 1460–1461.
G. Rogez, S. Parsons, C. Paulsen, V. Villar, T. Mallah, Inorg.
Chem. 2001, 40, 3836–3837.
a) A. Ohyoshi, J. Honbo, N. Matsumoto, S. Ohta, S. Sakamoto,
Bull. Chem. Soc. Jpn. 1986, 59, 1611–1613; b) Y. Maeda, M.
Suzuki, S. Hirose, S. Hayami, T. Oniki, S. Sugihara, Bull.
Chem. Soc. Jpn. 1998, 71, 2837–2843; c) A. Sour, M.-L. Boillot,
E. Rivière, P. Lesot, Eur. J. Inorg. Chem. 1999, 2117–2119.
S. Hirose, S. Hayami, Y. Maeda, Bull. Chem. Soc. Jpn. 2000,
73, 2059–2066.
C. Faulmann, S. Dorbes, B. Garreau de Bonneval, G. Molnár,
A. Bousseksou, C. J. Gomez-Garcia, E. Coronado, L. Valade,
Eur. J. Inorg. Chem. 2005, 3261–3270.
N. Matsumoto, S. Ohta, C. Yoshimura, A. Ohyoshi, S. Kohata,
H. Okawa, Y. Maeda, J. Chem. Soc., Dalton Trans. 1985, 2575–
2584.
a) S. Otha, C. Yoshimura, N. Matsumoto, H. Okawa, A.
Ohyoshi, Bull. Chem. Soc. Jpn. 1986, 59, 155–159; b) S. Hay-
ami, Y. Hosokoshi, K. Inoue, Y. Einaga, O. Sato, Y. Maeda,
Bull. Chem. Soc. Jpn. 2001, 74, 2361–2368.
CCDC-678513 (for 1), -678514 (for 2) and -678512 (for 3) contain
the supplementary crystallographic data for this paper. These data
can be obtained free of charge from The Cambridge Crystal-
logrpahic Data Centre via www.ccdc.cam.ac.uk/data_request/cif.
[20]
[21]
[22]
[23]
Supporting Information (see footnote on the first page of this arti-
cle): Thermal dependence curve of χMT for 3, absorption spectra of
functionalized Au-NPs, HRTEM image of AuL5-NPs and picture
showing the influence of the addition of NCS– to a stable colloidal
solution of Au-NPs functionalized by [(L5)Fe(solv.)]+.
Acknowledgments
[24]
[25]
We thank Dr. L. Catala for magnetic measurements. This work has
been supported by the Région Île-de-France in the framework of
CЈnano ÎdF. CЈNano-ÎdF is the nanoscience competence centre of
the Paris Region, supported by CNRS, CEA, MESR and Région
Île-de-France. This work was also supported by a grant from the
French Agence Nationale de la Recherche (ANR-07-JCJC-0040).
[26]
[27]
[1] P. Migowski, J. Dupont, Chem. Eur. J. 2007, 13, 32–39.
[2] R. W. J. Scott, O. M. Wilson, R. M. Crooks, J. Phys. Chem. B
2005, 109, 692–704.
[3] a) E. Katz, I. Willner, Angew. Chem. Int. Ed. 2004, 43, 6042–
6108; b) N. L. Rosi, C. A. Mirkin, Chem. Rev. 2005, 105, 1547–
1562.
[4] J. D. E. T. Wilton-Ely, Dalton Trans. 2008, 25–29.
[5] a) Y. Kim, R. C. Johnson, J. T. Hupp, Nano Lett. 2001, 1, 165–
167; b) C.-C. Huang, H.-T. Chang, Chem. Commun. 2007,
1215–1217.
[6] C.-C. Huang, Z. Yang, K.-H. Lee, H.-T. Chang, Angew. Chem.
Int. Ed. 2007, 46, 6824–6828.
[7] K. Yoosaf, B. I. Ipe, C. H. Suresh, K. G. Thomas, J. Phys.
Chem. C 2007, 111, 12839–12847.
[8] B. I. Ipe, K. Yoosaf, K. G. Thomas, J. Am. Chem. Soc. 2006,
128, 1907–1913.
[9] a) T. B. Norsten, B. L. Frankamp, V. M. Rotello, Nano Lett.
2002, 2, 1345–1348; b) X. Zhang, D. Li, X.-P. Zhou, New J.
Chem. 2006, 30, 706–711.
[28]
[29]
[30]
C. R. Mayer, E. Dumas, F. Sécheresse, Chem. Commun. 2005,
345–347.
C. R. Mayer, E. Dumas, A. Michel, F. Sécheresse, Chem. Com-
mun. 2006, 4183–4185.
a) T. Yonezawa, K. Yasui, N. Kimizuka, Langmuir 2001, 17,
271–273; b) C. S. Love, V. Chechik, D. K. Smith, C. Brennan,
J. Mater. Chem. 2004, 14, 919–923.
P. Ionita, A. Caragheorgheopol, B. C. Gilbert, V. Chechik, J.
Am. Chem. Soc. 2002, 124, 9048–9049.
K. Tanimura, R. Kitashima, N. Bréfuel, M. Nakamura, N.
Matsumoto, S. Shova, J.-P. Tuchagues, Bull. Chem. Soc. Jpn.
2005, 78, 1279–1282.
Y. Maeda, Y. Noda, H. Oshio, Y. Takashima, Bull. Chem. Soc.
Jpn. 1992, 65, 1825–1831.
X. Wang, M. E. Kotun, W. T. Pennington, J. C. Fanning, Inorg.
Chim. Acta 1988, 154, 189–199.
[31]
[32]
[33]
[34]
3622
www.eurjic.org
© 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Inorg. Chem. 2008, 3614–3623

Functionalization of Gold Nanoparticles by Iron(III) Complexes
[35]
[36]
[39] C. PerezLeon, L. Kador, B. Peng, M. Thelakkat, J. Phys. Chem.
B 2005, 109, 5783–5789.
E. Sinn, G. Sim, E. V. Dose, M. F. Tweedle, L. J. Wilson, J. Am.
Chem. Soc. 1978, 100, 3375–3390.
S. Floquet, M. C. Munoz, E. Riviere, R. Clement, J.-P. Audiere,
[40] M. Brust, M. Walker, D. Bethell, D. J. Schiffrin, R. Whyman,
J. Chem. Soc., Chem. Commun. 1994, 801–802.
[41] X. Wang, Y. Li, Chem. Commun. 2007, 2901–2910.
[42] S. Underwood, P. Mulvaney, Langmuir 1994, 10, 3427–3430.
[43] S. Zhang, G. Leem, L. Srisombat, T. R. Lee, J. Am. Chem. Soc.
2008, 130, 113–120.
[44] B. B. Corden, R. S. Drago, R. P. Perito, J. Am. Chem. Soc.
1985, 107, 2903–2907.
[45] R. Grice, L. N. Owen, J. Chem. Soc. 1963, 1947–1954.
[46] V. I. Dronov, R. F. Nigmatullina, Russ. Chem. Bull. 1976, 25,
2609–2611.
M.-L. Boillot, New J. Chem. 2004, 28, 535–541.
[37] a) J.-L. Yao, X. Xu, D.-Y. Wu, Y. Xie, B. Ren, Z.-Q. Tian, G.-
P. Pan, D.-M. Sun, K.-H. Xue, Chem. Commun. 2000, 1627–
1628; b) J.-Q. Hu, Q. Chen, Z. X. Xie, G. B. Han, R. H. Wang,
B. Ren, Y. Zhang, Z. L. Yang, Z. Q. Tian, Adv. Funct. Mater.
2004, 14, 183–189; c) B. Zhang, J. F. Li, Q. L. Zhong, B. Ren,
Z. Q. Tian, S. Z. Zou, Langmuir 2005, 21, 7449–7455; d) G.
Hu, Z. Feng, D. Han, J. Li, G. Jia, J. Shi, C. Li, J. Phys. Chem.
C 2007, 111, 8632–8637; e) G. Hu, Z. Feng, J. Li, G. Jia, D.
Han, Z. Liu, C. Li, J. Phys. Chem. C 2007, 111, 11267–11274.
[38] a) P. V. Kamat, J. Phys. Chem. B 2002, 106, 7729–7744; b) W. E.
Doering, S. Nie, Anal. Chem. 2003, 75, 6171–6176; c) D.
Cheng, Q.-H. Xu, Chem. Commun. 2007, 248–250.
[47] APEX2 version 1.0–8; Bruker AXS: Madison, WI, 2003.
[48] SHELXTL version 6.14; Bruker AXS: Madison, WI, 2001.
Received: March 4, 2008
Published Online: July 4, 2008
Eur. J. Inorg. Chem. 2008, 3614–3623
© 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org
3623