Supramolecular gold nanoparticles for the redox recognition of oxoanions: Syntheses, titrations, stereoelectronic effects, and selectivityt

Article abstract of DOI:10.1021/ja017015x

Stable, CH₂Cl₂-soluble mixed dodecanethiol/(amidoferrocenyl)alkanethiol (AFAT) gold colloids were synthesized by ligand substitution reactions from Brust's dodecanethiol gold colloids and the AFAT ligands to study the recognition and titration of oxoanions. Gold colloids were obtained with various chain lengths (C₁₁ and C₆ chains) of the AFAT ligand and different proportions of AFAT ligands in the colloids. Modification of the amidoferrocenyl structure [replacement of the free C₅H₅ ferrocene ring by C₅Me₅ (CP*) or C₅H₄COCH₃] has been achieved to investigate the stereoelectronic effects on the recognition. The cyclic voltammetry of these colloids in CH₂Cl₂ on Pt electrode shows a reversible Fe^II/IIII wave with some adsorption. With AFAT ligands, a new, less electrochemically reversible wave (with some adsorption) at a potential 220 ± 20 mV less positive than that of the initial wave appears upon titration of [n-Bu₄N][H₂PO₄], and the initial wave completely disappears after addition of 1 equiv of anion, which allows its titration. The potential shift does not depend on the AFAT proportion nor on its chain length but is reduced with Cp* and enhanced with C₅H₄COCH₃, showing the key role of the hydrogen bonding between the -NH- amido group and a terminal oxygen atom of the oxoanions. According to the Echegoyen-Kaifer model, the potential shift leads to the ratio K₍₊₎IK₍₀₎ of apparent association constants. In the presence of both [n-Bu₄N][HSO₄ and [n-Bu₄N]Cl, a shift of the initial wave (rather than its replacement) allows an easy titration, ideally with 20-Fc. Upon addition of [n-Bu₄N][HSO₄] alone, a weak wave shift (30 mV) of the colloids is also observed, allowing the titration of the HSO₄^- anion by the colloids containing a low percentage of AFAT ligand. The EchegoyenKaifer model provides access to the apparent association constant K₍₊₎ in this case for which the interaction between the anion and the neutral form of the host is not significant. With the C₅H₄COCH₃ modification of the amidoferrocenyl branch, a new wave appears at a potential 70 mV more positive than the initial wave, signifying a stronger interaction with this modified ligand than with the parent AFAT ligand. These colloids favorably compare with ferrocenyl dendrimers in terms of rapid synthesis and selectivity of H₂PO₄^- over HSO₄^- and with gold surfaces for the recognition of HSO₄^-.

Full text of DOI:10.1021/ja017015x

Published on Web 02/02/2002
Supramolecular Gold Nanoparticles for the Redox
Recognition of Oxoanions: Syntheses, Titrations,
†
Stereoelectronic Effects, and Selectivity
Agn e` s Labande, Jaime Ruiz, and Didier Astruc*
Contribution from Laboratoire de Chimie Organique et Organom e´ tallique,
UMR CNRS No. 5802, UniVersit e´ Bordeaux I, 33405 Talence Cedex, France
Received September 6, 2001
2 2
Abstract: Stable, CH Cl -soluble mixed dodecanethiol/(amidoferrocenyl)alkanethiol (AFAT) gold colloids
were synthesized by ligand substitution reactions from Brust’s dodecanethiol gold colloids and the AFAT
ligands to study the recognition and titration of oxoanions. Gold colloids were obtained with various chain
lengths (C11 and C
Modification of the amidoferrocenyl structure [replacement of the free C
or C COCH ] has been achieved to investigate the stereoelectronic effects on the recognition. The cyclic
voltammetry of these colloids in CH
With AFAT ligands, a new, less electrochemically reversible wave (with some adsorption) at a potential
20 ( 20 mV less positive than that of the initial wave appears upon titration of [n-Bu N][H PO ], and the
6
chains) of the AFAT ligand and different proportions of AFAT ligands in the colloids.
H
5 5
ferrocene ring by C Me (Cp*)
5
5
H
5 4
3
II/III
2
2
Cl on Pt electrode shows a reversible Fe wave with some adsorption.
2
4
2
4
initial wave completely disappears after addition of 1 equiv of anion, which allows its titration. The potential
shift does not depend on the AFAT proportion nor on its chain length but is reduced with Cp* and enhanced
with C
terminal oxygen atom of the oxoanions. According to the Echegoyen-Kaifer model, the potential shift leads
to the ratio K(+)/K(0) of apparent association constants. In the presence of both [n-Bu N][HSO ] and [n-Bu N]-
Cl, a shift of the initial wave (rather than its replacement) allows an easy titration, ideally with 20-Fc. Upon
5 4 3
H COCH , showing the key role of the hydrogen bonding between the -NH- amido group and a
4
4
4
addition of [n-Bu
titration of the HSO
4
N][HSO
4
] alone, a weak wave shift (30 mV) of the colloids is also observed, allowing the
-
4
anion by the colloids containing a low percentage of AFAT ligand. The Echegoyen-
Kaifer model provides access to the apparent association constant K(+) in this case for which the interaction
between the anion and the neutral form of the host is not significant. With the C COCH modification of
H
5 4
3
the amidoferrocenyl branch, a new wave appears at a potential 70 mV more positive than the initial wave,
signifying a stronger interaction with this modified ligand than with the parent AFAT ligand. These colloids
-
favorably compare with ferrocenyl dendrimers in terms of rapid synthesis and selectivity of H
2
PO
4
over
-
-
HSO
4
4
and with gold surfaces for the recognition of HSO .
3
4
Introduction
involved. The discovery by Brust of stable, very well defined
5
1
alkanethiol-stabilized gold nanoparticles with narrow polydis-
Among supramolecular materials, colloids are excellent
candidates for sensors especially when redox recognition is
2
persities has led to a new field of research. We wished to
combine the effect of the topography of alkanethiol-gold
nanoparticles and the supramolecular properties of their redox-
active termini for their use as exo-receptors for the recognition
of oxoanions. To reach this goal, we have synthesized new
*
Corresponding author. E-mail: d.astruc@lcoo.u-bordeaux.fr.
Abbreviations: AFAT, 11-(amidoferrocenyl)-1-undecanethiolate (1);
†
1a
Fc, ferrocenyl; 20-Fc, mixed gold colloids containing 80% dodecanethiolate
ligands and 20% AFAT ligands 1, etc.; 18-Fc (C6), mixed gold colloids
containing 82% dodecanethiolate ligands and 18% 6-(amidoferrocenyl)-1-
hexanethiolate ligands 6; Cp*, η -C5Me5; 25-Fp*, mixed gold colloids
containing 75% dodecanethiolate ligands and 25% AFAT ligands modified
6
5
(3) (a) Astruc, D. Electron-Transfer and Radical Processes in Transition-Metal
Chemistry; VCH: New York, 1995; Chapter 2. (b) Electron Transfer in
Chemistry; Balzani, V., Ed.; Wiley-VCH: Weinheim, Germany, 2001; Vol.
with Cp*, i.e. 11-(1-amido,1′-pentamethylferrocenyl)-1-undecanethiol, 7;
2
, Section II.
6
6
-FcAc, mixed gold colloids containing 94% dodecanethiolate ligands and
% AFAT ligand modified with η -C5H4COMe, i.e. 11-(1-amido,1′-
(
4) (a) Brust, M.; Walker, M.; Bethell, D.; Schiffrin, D. J.; Whyman, R. J.
Chem. Soc. Chem. Commun. 1994, 801. (b) Brust, M.; Fink, J.; Bethell,
D.; Schiffrin, D. J.; Kiely, C. Chem. Commun. 1995, 1655.
5
acetylferrocenyl)-1-undecanethiol, 15; CV, cyclic voltammetry or cyclic
voltammogram.
(5) For reviews on colloids previous to Brust’s discovery, see: (a) Schmid,
G. Chem. ReV. 1992, 92, 1709. (b) Bradley, J. S. In Clusters and Colloids;
Schmid, G., Ed.; VCH: Weinheim, Germany, 1995; Chapter 6. (c) See
also: Schmid, G.; Chi, L. F. AdV. Mater. 1998, 18, 515.
(6) (a) Val e´ rio, C.; Fillaut, J.-L.; Ruiz, J.; Guittard, J.; Blais, J.-C; Astruc, D.
J. Am. Chem. Soc. 1997, 119, 2588. (b) Val e´ rio, C.; Alonso, E.; Ruiz, J.;
Blais, J.-C.; Astruc, D. Angew. Chem., Int. Ed. 1999, 38, 1747. (c) Labande,
A.; Astruc, D. Chem. Commun. 2000, 1007. (d) Alonso, E.; Labande, A.;
Raehm, L.; Kern, J.-M.; Astruc, D. C. R. Acad. Sci., Ser. 2c 1999, 2, 209.
(e) Daniel, M.-C.; Ruiz, J.; Nlate, S.; Palumbo, J.; Blais, J.-C.; Astruc, D.
Chem. Commun. 2001, 2000.
(
1) Lehn, J.-M. Supramolecular Chemistry: Concepts and PerspectiVes;
VCH: Weinheim, Germany, 1995. (b) Kumar, A.; Arbott, N. L.; Kim, E.;
Biebuyck, A.; Whitesides, G. M. Acc. Chem. Res. 1995, 28, 219.
2) (a) Fitzmaurice, D.; Rao, S. N.; Preece, J.; Stoddart, J. F.; Wenger, S.
Zaccheroni, N. Angew. Chem., Int. Ed. 1999, 38, 1147. (b) Sampath, S.;
Lev, O. AdV. Mater. 1997, 9, 410. (d) Boal, A. K.; Rotello, V. M. J. Am.
Chem. Soc. 1999, 121, 4914. (e) Niemz, A.; Rotello, V. M. Acc. Chem.
Res. 1999, 32, 44. (f) Elghanian, R.; Storhoff, J. J.; Mucic, R. C.; Letsinger,
R. L.; Mirkin, C. A. Science 1997, 277, 1078. (g) Shenton, W.; Davis, D.
A.; Mann, S. AdV. Mater. 1999, 119, 11132.
(
1782 VOL. 124, NO. 8, 2002
9
J. AM. CHEM. SOC.
10.1021/ja017015x CCC: $22.00 © 2002 American Chemical Society

Redox Recognition of Oxoanions
A R T I C L E S
Scheme 1. Synthesis of the 11-(1-Amido,1′-pentamethylferro-
functional alkanethiol-gold colloids containing a mixture of
alkanethiol ligands and (amidoferrocenyl)alkanethiol (AFAT)
cenyl)-1-undecanethiol Ligand 12
+
-
ligands and monitored the titration of n-Bu4N salts of H2PO4
-
and HSO4 by these colloids using cycling voltammetry.
Recognition of anions is an important topic because a number
of anions are found in biology.7
,8
Beer has shown that amidoferrocenyl groups linked to various
8
endo-receptors are able to sense anions. We have also disclosed
-
- 6a
-
- 6b
the recognition of H2PO4 , HSO4 , and Cl and Br
by
metallodendrimers. From preliminary studies, we know that the
redox potential of the amidoferrocenyl group is sufficiently
perturbed by the synergy between the H-bonding, electrostatic
interaction, and topography in alkanethiol-gold nanoparticles
containing AFAT ligands (unlike in alkylamidoferrocene mono-
Scheme 2. Synthesis of the 11-(1-Amido,1′-acetylferrocenyl)-
1-undecanethiol Ligand 15
-
- 6c
mers) to recognize H2PO4 and HSO4 . The detailed results
of this study are reported here with thiol ligands containing
amidoferrocene termini (AFAT ligands). In particular, we have
also synthesized and stabilized on gold nanoparticles new thiol
ligands containing branches that were modified using a C5Me5
or a C5H4COMe ring in the ferrocenic frame to investigate the
role of the stereoelectronic factors on the recognition sites.
Various alkanethiols functionalized with ferrocenyl termini were
already synthesized, mainly for electron-transfer studies on self-
assembled monolayers. For instance, amidoferrocenyl-al-
kanethiol ligands are known from Creagers’ work.9 The thiols
containing functionalized ferrocenyl groups introduce the su-
pramolecular character that had not been exploited and allowed
us to develop the study of the molecular-recognition properties.
Ligand-substitution reactions in alkanethiol-gold nanoparticles
of Brust type are known with other alkanethiol ligands including
ferrocenylalkanethiol ligands, and it was shown that the gold
Table 1. Results of Ligand-Exchange Reactions between AFAT
and Dodecanethiol (DT) Gold Nanoparticles
,10
AFAT/DT ratio
AFAT chain
length
AFAT/DT nos.
in productc
% AFAT
in productd
reacna
productb
C11
C11
C11
C11
C11
C6
1:4
1:1
1.5:1
1.75:1
2:1
1:13
1:8
1:5
1:3.5
1:2.6
1:5.5
8/97
7-Fc
10/70
21/85
31/75
39/64
16/70
13-Fc
20-Fc
29-Fc
38-Fc
18-Fc
11
core was unchanged during such reactions. Recently, Nishihara
synthesized nanoparticles containing thiol ligands that were
2:1
a
terminated by mixed-valence biferrocenyl groups, bringing
Mole ratio of AFAT ligand to nanoparticle-bound dodecanethiolate (DT)
b
versatile redox properties to the nanoparticles.12
ligands in ligand-exchange reactions. c Average mole ratio AFAT/DT
ligands obtained in the nanoparticles. Average numbers of AFAT/DT
ligands obtained in the particles. d Average % of AFAT ligands obtained
in the nanoparticles.
Results and Discussions
Gold Colloid Syntheses. The thiol-ligand syntheses are
described in the Supporting Information. With the Cp* and
CpCOCH3 groups, however, they are shown in Schemes 1
and 2.
1
13
analysis, H NMR, and TEM according to Leff’s model
allowed us to determine an average number of 269 gold atoms/
core of 2-nm diameter and 103 dodecanethiolate chains (see
Experimental Section). These results are in agreement with those
obtained by Brust with gold nanoparticules stabilized by
dodecanethiol ligands. Then, a variable proportion of AFAT
ligands has been introduced by the ligand-exchange procedure
used by Murray for other ligands.
We have applied this technique with AFAT ligands of
variable length (11 and 6 C atoms) and structure (Cp, Cp*,
CpCOCH3) and verified using TEM that the core size was not
modified upon ligand substitution. Table 1 shows the proportion
of AFAT ligands (determined by H NMR) that is introduced
onto the nanoparticles as a function of the amount engaged in
the ligand-substitution reactions (Scheme 3).
In particular, with (amidoferrocenyl)undecanethiol ligands,
saturation is reached for 38 AFAT ligands/particle. Reactions
carried out in the presence of larger amounts of AFAT ligands
never gave a content higher than 38% AFAT ligands in the
mixed colloids, which is in accord with Murray’s investiga-
The gold colloids with dodecanethiol ligands were synthesized
4
according to Brust’s procedure. The combination of elemental
4
(
7) Seel, C.; de Mendoza, J. In ComprehensiVe Supramolecular Chemistry;
Atwood, J., Davies, J. E. D., McNichol, D. D., V o¨ gtle, F., Eds.; Elsevier:
New York, 1996; Vol. 2, Chapter 17, pp 519-552.
11
(8) (a) Beer, P. D. AdV. Inorg. Chem. 1992, 39, 79. (b) Beer, P. D. J. Chem.
Soc., Chem. Commun. 1996, 689. (c) Beer, P. D. Acc. Chem. Res. 1998,
3
1, 71. (d) Beer, P. D.; Gale, P. A.; Chen, Z. AdV. Phys. Org. Chem. 1998,
1, 1. (e) Beer, P. D.; Gale, P. A. Angew. Chem., Int. Ed. 2001, 40, 486.
6
c
3
(
9) (a) Weber, K.; Creager, S. E. Anal. Chem. 1994, 66, 3164. (b) Rowe, G.
K.; Creager, S. E. J. Phys. Chem. 1994, 98, 5500. (c) Weber, K.; Hockett,
L.; Creager, S. E. J. Phys. Chem. B 1997, 101, 8286. (d) Creager, S.; Yu,
C. J.; Bamdad, C.; O’Connor, S.; MacLean, T.; Lam, E.; Chong, Y.; Olsen,
G. T.; Luo, J.; Gozin, M.; Kayyem, J. F. J. Am. Chem. Soc. 1999, 121,
1
1
059.
(
10) For esters, see: Chidsey, C. E. D.; Bertozzi, C. R.; Putvinski, T.; Mujsce,
A. M. J. Am. Chem. Soc. 1990, 112, 4301.
(
11) (a) Hosteler, M. J.; Green, S. J.; Stockes, J. J.; Murray, R. W. J. Am. Chem.
Soc. 1996, 118, 4212. (b) Ingram, R. S.; Hostetler, M. J.; Murray, R. W.
J. Am. Chem. Soc. 1997, 119, 9175. (c) Hostetler, M. J.; Templeton, A.
C.; Murray, R. W. Langmuir 1999, 15, 3782. (d) Templeton, A. C.;
Wuelfing, W. P.; Murray, R. W. Acc. Chem. Res. 2000, 33, 27.
(
12) (a) Horikoshi, T.; Itoh, M.; Kurihara, M.; Kubo, K.; Nishihara, H. J.
Electroanal. Chem. 1999, 473, 113. (b) Yamada, M.; Tadera, T.; Kubo,
K.; Nishihara, H. Langmuir 2001, 17, 2263. (c) Yamada, M.; Quiros, I.;
Mizutani, J.; Kubo, K.; Nishihara, H. Phys. Chem. Chem. Phys. 2001, 3,
3
377. (d) Men Y., Kubo, K.; Kurihara, M.; Nishihara, H. Phys. Chem.
(13) Leff, D. V.; Ohara, P. C.; Heath, J. R.; Gelbart, W. M. J. Phys. Chem.
1995, 99, 7036.
Chem. Phys. 2001, 3, 3427.
J. AM. CHEM. SOC.
9
VOL. 124, NO. 8, 2002 1783

A R T I C L E S
Labande et al.
Scheme 3. Ligand Substitution Reactions (CH2Cl2, 2 d, RT) for the Syntheses of the Gold Colloids Containing Mixed Dodecanethiol and
AFAT-Type Ligands with Variation of the Chain Length (C11 vs C6) and Ring Structure of the Ferrocenyl Motif (Cp, Cp*, C5H4COMe)
tions.11 These nanoparticles are stable in air and thermally, and
they are soluble in CH2Cl2. They are not soluble in polar
solvents, however, which limits the range of conditions for the
recognition studies.
Cl2, 0.1 M [n-Bu4N][PF6]). As for previously reported nano-
1
1,12
particles bearing ferrocenyl termini,
a single reversible
ferrocene/ferrocenium wave was obtained with all the colloids
bearing a proportion of thiol ligands with an amidoferrocenyl
termini or one of its derivatives whose structure and syntheses
are described above. The shape of this wave varies and depends
on the exact nature of the ferrocene termini, but the following
common features are observed. The cathodic return wave is, in
general, higher and thinner than the anodic forward wave, which
can be attributed to some adsorption of the colloids onto the
electrode surface in the ferrocenium form. The difference
between the anodic and cathodic peak potentials (∆Ep) is always
lower than the 58-mV value expected for a reversible single-
electron wave of an electroactive species in solution at 20 °C:
Only 18% of 6 was introduced on the colloid starting from
2
1
-fold excess of this ligand vs dodecanethiol whereas 25% of
2 was introduced starting from an excess of 1.7/1.
The ligand substitution with (1-amido,1′-acetylferrocenyl)-
undecanethiol ligands was carried out using 1.1 equiv of this
ligand (carefully reduced from the disulfide) vs dodecanethiol
ligands. This resulted in the introduction of only 6% of ((1-
amido,1′-acetyl)ferrocenyl)undecanethiol ligands starting from
1.1 equiv of 15/dodecanethiol ligand. The polarity of the acetyl
group of 15 lowers the solubility of the colloids in nonpolar
solvents, which inhibits further ligand substitution. This propor-
tion of redox-active ligand is sufficient for electrochemistry and
recognition studies, however. The core size (2.1 ( 0.5 nm in
average measured by HRTEM; 270 ( 70 gold atoms) and ligand
number (103 ( 25) are essentially unchanged in these reactions.
Various proportions of AFAT ligands 1 have been introduced
depending of the amount engaged in the exchange reaction (see
Table 1; saturation is reached at 38-Fc for the C11 chain ligands
1
4
the ∆Ep values are in the range of 20-30 mV, which also
means that some adsorption occurs (the fully adsorbed redox
species would ideally have a nil ∆Ep value). The observation
of a single wave means that, in principle, all the redox sites are
equivalent on the electrochemical time scale. One may note that
since the rotation of the colloidal particles must be much faster
than the electrochemical time scale, the location of all the redox
sites vs the electrode is, on average, similar. In particular, there
is no redox site for which a slow electron transfer is disclosed.1
5-17
1
and 18-Fc for the C6-chain ligand 6). A 1.7 equiv amount of
the ligand 11-(1-amido,1′-pentamethylferrocenyl)-1-undecaneth-
iol ligand 12 yielded 25-Fc* colloids, i.e. containing 25% of
this ligand.
Cyclic Voltammetry of the Colloids. The cyclic voltam-
mograms (CVs) were recorded using classic conditions (Pt, CH2-
(
14) Bard, A. J.; Faulkner, R. L. Electrochemical Methods; Wiley: New York,
1980.
(15) (a) Gorman, C. B.; Parkhurst, B. L.; Su, W. Y.; Chen, K.-Y. J. Am. Chem.
Soc. 1997, 119, 1141. (b) Gorman, C. B. AdV. Mater. 1997, 9, 1117; 1998,
1
0, 295. (c) Gorman, C. B.; Smith, J. C.; Hager, M. W.; Parkhurst, B. L.;
Sierzputowska-Gracz, H.; Haney, C. A. J. Am. Chem. Soc. 1999, 121, 9958.
1784 J. AM. CHEM. SOC.
9
VOL. 124, NO. 8, 2002

Redox Recognition of Oxoanions
A R T I C L E S
Figure 2. Titration of [n-Bu4N][H2PO4] with the 20-Fc (AFAT) gold
nanoparticles monitored by CV: decrease of the intensity of the initial CV
wave (2) and increase of the intensity of the new CV wave (b) vs the
number of equivalents of [n-Bu4N][H2PO4] added/AFAT branch. Nano-
-
6
-2
particles: 5 × 10 M in CH2Cl2. [n-Bu4N][H2PO4]: 10 M in CH2Cl2.
Figure 1. Cyclic voltammograms of the gold nanoparticles 38-Fc: solvent,
[n-Bu4N][PF6]: 0.1 M. T: 20 °C. Reference electrode: SCE. Auxiliary
and working electrodes: Pt. Scan rate: 200 mV‚s .
-
1
CH2Cl2; reference electrode, aqueous SCE; working and counter electrode,
-
1
Pt; supporting electrolyte, 0.1 M n-Bu4NPF6; scan rate, 200 mV‚s . Key:
(
a) without [n-Bu4N][H2PO4]; (b) with 0.6 equiv of [n-Bu4N][H2PO4]/
Chart 1. Double Hydrogen-Bonding Interaction between
-
amidoferrocenyl branch; (c) with excess [n-Bu4N][H2PO4]. Independently,
the E1/2 values (see text) were determined using decamethylferrocene as an
internal reference: [FeCp*2]/[FeCp*2] (E°′ ) -0.070 V vs SCE on Pt in
CH2Cl2).
Amidoferrocenium and H PO4
2
Such a slow electron transfer could be caused by a location
remote from the electrode. On the other hand, a difference in
location of the various redox sites vs the electrode is awaited
for colloidal particles adsorbed on the electrode surface, but a
hoping mechanism among the ferrocenyl redox sites may then
establish this equivalence.
new wave, it is possible to carry out a titration using the
disappearance of the initial wave and the appearance of the new
wave. For this purpose, the variations of the intensities of the
initial wave and the new wave are recorded as a function of the
number of equivalents of [n-Bu4N][H2PO4] added/amidoferro-
cenyl branch (Figure 2). These results are obtained with gold
nanoparticles containing various proportions of AFAT ligands.
For instance, Figure S1 (Supporting Information) compares the
titration with various proportions of AFAT ligands.
The standard redox potential E° is estimated from the average
values of the anodic and cathodic waves [E1/2 ) (Epa + Epc)/
2
], although the adsorption causes a small error. Decamethyl-
1
8
ferrocene, actually a better reference than ferrocene, can be
used as universal reference. The E1/2 values are 0.665 V with
the AFAT ligands, 0.680 V with the C6-chain ligands, 0.365 V
with Cp*, and 0.750 V with CpCOCH3 (all values vs Cp2*Fe,
i.e. -0.545 V vs Cp2Fe in CH2Cl2).
It is remarkable that the potential shift resulting from the
+
-
addition of the n-Bu4N salt of the anion H2PO4 is large (∆E1/2
) 220 ( 20 mV). It is also almost constant upon variation of
the proportion (from 8 to 38%) and length (11 vs 6 CH2 units)
of AFAT ligands in the nanoparticles. This trend is different
from that observed with amidoferrocenyl dendrimers for which
the potential shift dramatically increases with the dendrimer
generation [potential shift in CH2Cl2 upon addition of 1 equiv
-
-
Recognition of H2PO4 and HSO4 by the Redox-Active
-
Colloids. (a) Recognition and Titration of H2PO4 . Addition
of [n-Bu4N][H2PO4] to a CH2Cl2 solution of these nanoparticles
provokes the appearance of a new, cathodically shifted wave
whose intensity increases at the expense of the initial wave. As
for the initial wave, the new wave shows some adsorption and
is characterized by a larger peak-to-peak separation than the
standard 60-mV value for a reversible wave. This trend is
characteristic of a relatively slow heterogeneous electron transfer
due to a structural reorganization of the amidoferrocenyl-H2-
PO4 complex in the course of the redox change. The replacement
of the initial wave by the new one is complete after addition of
approximately 1 equiv of [n-Bu4N][H2PO4]/AFAT branch
-
of H2PO4 anion: 45 mV (1-Fc); 110 mV (3-Fc); 220 mV (9-
6
,18b
Fc); 315 mv (18-Fc)].
Thus, the mixed alkanethiol/AFAT
colloids seem to be about as efficient as the nonakis(amidof-
errocenyl) dendrimer. The phenomenum is also comparable to
that obtained for the same C11 AFAT ligand alone at a gold
electrode in a preliminary study (coverage, τ ) 36%; potential
shift, 290 mV; Epa - Epc ) 150 mV for the new wave; the
6d
redox system is not stable more than a few cycles, however).
(Figure 1), indicating a one-to-one interaction between the anion
According to the seminal Echegoyen-Kaifer model,19 the
replacement of a wave by another one due to interaction with
one of the redox states corresponds to the square Scheme 4.
In this model, the interaction of the host is strong with one
of the redox forms and weaker (but significant) with the other,
the ratio of apparent association constants being
and the amidoferrocenium group, as proposed by Beer for
receptors containing amidoferrocenyl groups (Chart 1).
Despite the perturbation caused by the adsorption phenomena
and the slow heterogeneous electron transfer observed for the
(
16) Green, S. J.; Pietron, J. J.; Stockes, J. J.; Hosteler, M. J.; Vu, H.; Wuelfing,
W. P.; Murray, R. W. Langmuir 1998, 14, 5612.
(
17) (a) For a recent review of electron transfer in and supramolecular
electrochemistry of nanosized assemblies containing many ferrocenyl
centers, see: Kaifer, A. E.; Gomez-Kaifer, M. Supramolecular Electro-
chemistry; Wiley-VCH: Weinheim, Germany, 1999; Chapter 16, p 207.
E°free - E°bound ) ∆E° (V) ) 0.058 log (K(+)/K ) at 20 °C
(0)
This ∆E1/2 (value close to ∆E°) corresponds to the apparent
-
(b) Cardona, C. M.; Kaifer, A. E. J. Am. Chem. Soc. 1998, 120, 4023.
association constant K(+) between H2PO4 and the polyferro-
J. AM. CHEM. SOC.
9
VOL. 124, NO. 8, 2002 1785

A R T I C L E S
Labande et al.
Scheme 4. Square Scheme Relating the Apparent Association
Constants between the Oxoanionic Guest and Colloids under the
Neutral Amidoferrocenyl Form (K(0)) and the Amidoferrocenium
Form (K(+)) to the Standard Redox Potentials of the Free and
Bound Forms (E°free and E°bound, Respectively)
Table 2. Determination of the Average Number of AFAT Ligands
in the Nanoparticles in CH2Cl2
total no. of AFAT ligands
titration by CV
1H NMR
H PO
2 4
-
H PO
2 4
- a
HSO
-
4
%
AFAT
8
1
2
2
3
1
-Fc (C11)
3-Fc (C11)
0-Fc (C11)
9-Fc (C11)
8-Fc (C11)
8-Fc (C6)
8
10
21
31
39
16
8.3
9.5
22
29
32
14
9
19
25
10.25
18
b
17
13
b
29.5
b
9
a
Titration carried out in the presence of excess [n-Bu4N][HSO4] (5 ×
6
-6
1
0- mol) and n-Bu4NCl (2.5 × 10 mol). b Low value due to intramo-
lecular H-bonding at high content of AFAT ligand (see text and Experi-
mental Section).
cenium form of the nanoparticles that is 6210 ( 620 times larger
than that, K(0), related to the neutral polyferrocene form
(
by the larger value of the positive charge on the NH atoms of
the amido group. This analysis also shows that, within the
chelating hydrogen-bonding interactions, the hydrogen bond of
electrostatic effect).8,19 The fact that the interaction of H2PO4-
-
is so much stronger with the amidoferrocenium form than with
the amidoferrocenyl one signifies that the hydrogen bonding
between the negatively charged oxygen atom of this anion and
the positively charged -NH- group is the dominant one. The
reason thereof is that this positive charge is largely enhanced
upon oxidation of the ferrocenyl group.
The new wave always appears whatever the structure of the
amidoferrocene termini of the thiol ligand, i.e. whether it bears
electron-releasing or electron-withdrawing substituents. How-
ever, the potential shift very much depends on the detailed
stereoelectronic features of the amidoferrocene termini (Figure
S1 and S2, Supporting Information).
this NH group with the oxygen atom of the H2PO4 anion is
more significant than that between the oxygen atom of the amido
-
group and the OH group of H2PO4 . The remarkable influence
of the stereoelectronic effect shows the importance the geometric
and electronic conditions in the weak supramolecular interac-
tions leading to recognition. Thus, the channels defined between
the parallel ligands create sufficient steric strain to provide a
favorable particle effect comparable in some way to the dendritic
1
8b,20
effect obtained with metallodendrimers.
This titration technique also allows one to confirm the number
1
of AFAT ligands in the particles determined by H NMR for
the gold nanoparticles containing a proportion of AFAT ligands
in the range 13%-29% (Table 2).
By comparison with the parent AFAT ligand, the AFAT
ligand bearing a Cp* ring induces a smaller potential shift of
-
(b) Recognition and Titration of HSO4 . Addition of
[n-Bu4N][HSO4] to a CH2Cl2 solution of these gold nanopar-
125 mV, which corresponds to an apparent association constant
ticles containing dodecanethiol and AFAT ligands or AFAT
ligands modified with Cp* provokes the progressive shift of
the initial CV wave, a behavior completely different from that
observed upon addition of [n-Bu4N][H2PO4]. This progressive
shift stops when approximately 1 equiv of [n-Bu4N][HSO4]/
redox active ligand has been added in the case of nanoparticles
containing a low percentage of AFAT ligands (8-20%). The
potential shift is rather weak, and its value is 30 ( 5 mV in
average with the parent AFAT ligands. Using the Cp* ring
instead of the Cp one led to a potential shift of 25 ( 5 mV.
The potential shift is weaker than that recorded with the
amidoferrocenyl dendrimers [no significant shift (1-Fc), 30-mV
ratio K(+)/K(0) ) 143 ( 15 that is 50 times smaller than with
the parent AFAT ligands. The steric constraints brought about
by the ring permethylation disfavor the interaction between the
amido group and the dihydrogenophosphate anion. The elec-
tronic effect is also clear since electron-releasing groups lower
the positive charge at the NH group responsible for the main
H-bonding leading to the recognition. Thus, this dramatic
decrease of potential shift is due to the addition of negative
steric and electronic effects. On the other hand, the potential
shift recorded with the amidoferrocenyl termini bearing an acetyl
substituent in larger (275 ( 10 mV) than that observed with
the parent AFAT ligand. This potential shift means that the
apparent association constant K(+) of the anion with the colloid
bearing the cationic ferrocenium form of the electroactive (1-
acetyl,1′-amidoferrocenyl)dodecanethiolate ligand is 55 130 (
6a,b
(3-Fc), 65 mV (9-Fc), 130 mV (18-Fc)], whereas no potential
shift was observed with the gold electrode modified with the
6
d
C11-AFAT ligand. In the Kaifer-Echegoyen model, this
corresponds to the square scheme above but with weak
interaction between the host and the reduced form (K(0) , 1).
Under these conditions, there is, according to this model, a direct
access to the absolute value of the apparent association constant
K(+): ∆E° (V) ) 0.058 log K(+)c at 20 °C, c being the
concentration of added anion.
5
500 times larger than that (K(0)) of the anion with the colloid
II
bearing the neutral Fe form of the same ligand. This ratio is
nine times larger than with the parent AFAT ligand. Thus, the
compared effects of the ligands modified with the Cp* and
acetyl-Cp ligands respectively are in agreement given their
opposite electronic characters. These results may be taken into
account, with the electron-accepting acetyl substituent, by a
larger electrostatic interaction between the anion and the
ferrocenium centers when this center is more electron poor and
This corresponds to an apparent association constant K(+) )
00 ( 100 mol ‚L at a concentration of 10 M of AFAT
-
1
-2
4
(
20) (a) Newkome, G. R.; He, E.; Moorefield, C. N. Chem. ReV. 1999, 99, 1689.
(
b) Balzani, V.; Campagna, S.; Denti, G.; Juris, A.; Serroni, S.; Venturi,
(
18) (a) Ruiz, J.; Astruc, D. C. R. Acad. Sci. Paris, Ser. 2c 1998, 21. (b) Astruc,
D. In ref 3b, Chapter 4, p 728.
M. Acc. Chem. Res. 1998, 31, 26. (c) Cuadrado, I.; Mor a´ n, M.; Casado, C.
M.; Alonso, B.; Losada, J. Coord. Chem. ReV. 1999, 189, 123. (d)
Hearshaw, M. A.; Moss, J. R. Chem. Commun. 1999, 1. (e) Astruc, D.;
Chardac, F. Chem. ReV. 2001, 101, 2991.
(19) Seminal report: Miller, S. R.; Gustowski, D. A.; Chen, Z.-H.; Gokel, G.
W.; Echegoyen, L.; Kaifer, A. E. Anal. Chem. 1988, 60, 2021.
1786 J. AM. CHEM. SOC.
9
VOL. 124, NO. 8, 2002

Redox Recognition of Oxoanions
A R T I C L E S
Chart 2. Possible Interactions between Amidoferrocenium and
HSO4- Involving Various Amounts of Equivalents of HSO4-
Figure 3. Cyclic voltammograms of the gold nanoparticles 25-Fc* during
-
titration with H2PO4 : solvent, CH2Cl2; reference electrode, aqueous SCE;
-
1
ligand (with Cp*, ∆E1/2 ) 25 mV; K(+) ) 270 ( 70 mol ‚L).
working and counter electrode, Pt; supporting electrolyte, 0.1 M n-Bu4-
NPF ; scan rate, 200 mV‚s . Key: (a) without [n-Bu N][H PO ]; (b) with
4 2 4
0.75 equiv of [n-Bu N][H PO ]/1-amido,1′-pentamethylferrocenyl branch;
-
-1
It is noteworthy that the ∆E1/2 value is much smaller for HSO4
than for H2PO4 , which is consistent with the selectivity of the
6
4
2
4
-
(
c) with excess [n-Bu4N][H2PO4].
recognition by these nanoparticles (vide infra).
With the AFAT ligand modified with Cp*, the value of the
potential shift ∆E1/2 is even lower, only 25 mV, which
corresponds to an apparent association constant K(+) ) 270 (
-
1
7
0 mol ‚L. The titration was carried out only with the 25-Fc*
colloids for which ∆E1/2 ) 27.5 ( 5 mV, which is the limit of
the possibility given the high percentage of error. This very
low ∆E1/2 value is due to the decreased hydrogen-bonding
interaction, electrostatic effect, and steric constraints with the
Cp* modification compared to the parent AFAT ligand, already
-
mentioned above for H2PO4 .
The weaker interaction of the amidoferrocenium group with
-
-
the HSO4 anion than with the H2PO4 anion can be taken into
account by the lower charge density on the oxygen atom in the
former. This is due to the facts that the negative charge is
-
delocalized onto three oxygen atoms in HSO4 instead of only
two in H2PO4 and that sulfur is more electronegative than
-
phosphorus (Pauling electronegativity indices: 2.6 and 2.2,
-
-
respectively). Thus, H2PO4 is more basic than HSO4 in CH2-
-
-
Figure 4. Titration of [n-Bu4N][HSO4] by gold nanoparticles containing
AFAT ligands in CH2Cl2, showing shift of E1/2 toward positive potentials
as a function of the number of equivalents of [n-Bu4N][HSO4] added/
amidoferrocenyl branch of the colloids. See Figures 1 and 2 and Experi-
mental Section for conditions. Key: (a) titration by 13-Fc, where the
equivalence point is 1 equiv of [n-Bu4N][HSO4]/amidoferrocenyl branch;
Cl2. On the other hand, HSO4 is more acidic than H2PO4 ,
but this property does not dominate because the amidic oxygen
atom has little basicity in the amidoferrocenium form. The
compared recognition experiments for these two anions indeed
confirm that the dominant hydrogen-bonding interaction in-
volves the -NH- group of the amide and a terminal oxygen
(b) titration by 38-Fc, where the equivalence point is 0.4 equiv of [n-Bu4N]-
[HSO ]/amidoferrocenyl branch.
4
8
atom of the anions. When the proportion of AFAT ligands is
large (in 29-Fc and 38-Fc) or the chain short (C6), however,
the equivalence point is no longer reached after addition of 1
subject of perturbations due to any relatively low-energy event.
Figure 4 illustrates the fact that the titration is clear, however.
-
equiv of HSO4 anion/amidoferrocenyl branch but near 0.5
Despite the low ∆E value, it is thus possible to titrate [n-Bu N]-
4
-
equiv of HSO4 . This effect is even more marked with the Cp*
[HSO ] using the nanoparticles that have a low content of AFAT
4
modification of the AFAT ligand for which the equivalent point
was observed for only 0.7 equiv of HSO4 anion with the 25-
ligands (8-20%).
-
The accuracy is better with a higher content of AFAT ligand
than with a low one. Thus, the best choice is a compromise
with 13-Fc, i.e. a proportion of 13% AFAT ligands in the
colloids (see Table 4).
Fc colloids. It is possible that the proportion of amidoferrocenyl
dimers being higher given the parallelism of the ligands in the
-
nanoparticles, the HSO4 anion cannot compete with their
formation. Note along this line, in comparison, that the AFAT
ligand alone on a gold electrode does not sense the HSO4 anion
The use of the acetylcyclopentadienyl variation of the AFAT
ligand offers the possibility of a stronger interaction with the
-
6
d
-
-
at all. Another possibility is that the HSO4 anion bridges
two neighboring amido groups (Chart 2).
HSO4 anion. Interestingly, a new event is observed in this case
-
upon addition of the HSO4 anion to the 6-Fc . A new wave
Ac
-
Accordingly, this effect is also more marked for HSO4 than
for H2PO4 . The HSO4 anion-amidoferrocenyl branch inter-
appears at a peak potential Epc ) 170 mV more positive than
that of the original wave, but this new wave is not reversible;
i.e., the corresponding anodic wave is almost absent (weak
-
-
action is weak in the AFAT-containing colloids and is thus the
J. AM. CHEM. SOC.
9
VOL. 124, NO. 8, 2002 1787

A R T I C L E S
Labande et al.
Table 4. Determination of the Number of Equivalents of Anion/
Table 3. Potential Shift ∆E1/2 Observed for the FeII/III Redox
System upon Addition to the Colloids of 1 equiv of Anion/
Amidoferrocenyl Branch of the Colloids in the Titrations Using the
-
-
a
Amidoferrocenyl Branch during Titrations of H2PO4 , HSO4 , and
H2PO4 in the Presence of both [n-Bu4N][HSO4] and [n-Bu4N]Cla
Colloids with Various Proportions of AFAT Ligand (C11)
-
equiv of H
HSO4 and Cl
2 4
PO with
∆E1/2 (mV)
% AFAT
equiv of H PO4
-
-
-
equiv of HSO4
-
2
equiv of H
2
PO
4
with
8-Fc (C11)
13-Fc (C11)
1.25
1.0
1.0
0.94
0.75
1.10
1.0
0.88
0.55
0.35
-
-
andCl-
-
4
%
AFAT
H
2
PO
4
HSO
4
HSO
0.87
0.82
0.75
0.75
2
2
3
0-Fc (C11)
9-Fc (C11)
8-Fc (C11)
8
1
2
2
3
-Fc (C11)
200 ( 20
240 ( 20
215 ( 20
200 ( 20
232 ( 20
25 ( 10
32 ( 10
27 ( 10
37 ( 10
30 ( 10
3-Fc (C11)
0-Fc (C11)
9-Fc (C11)
8-Fc (C11)
112 ( 15
100 ( 15
125 ( 15
175 ( 20
a
See the caption to Figures 1 and 2 for conditions and text for dis-
crepancies with respect to the one-to-one interaction (1 equiv/branch).
a
See the captions to Figures 1 and 2 for conditions.
shoulder). This could be the indication that the interaction
between the HSO4- anion and the 6-FcAc colloids is strong in
the ferrocenium form and almost too weak to be significant in
the neutral form. The practical consequence is that the thermo-
dynamic potential corresponding to the new wave is not
accessible with a reasonable accuracy. Another problem is that
the percentage of 11-(amido(acetylferrocenyl))-1-undecanethiol
ligand in the colloid is very low (6%). This latter aspect
disfavors the accuracy in titrations. Such a titration has been
best carried out, however, by measurement of the variation of
intensity the new wave. The accuracy of the titration is only of
-
the order of 10-20% of the number of equivalents of HSO4
Figure 5. Titration of [n-Bu4N][H2PO4] in the presence of both [n-Bu4N]-
anion, which is confirmed by the slightly too large value of 1.1
4
[
HSO4] and [n-Bu4N]Cl (in addition to the 2 × 10 excess of supporting
-
equiv of HSO4 anion observed.
electrolyte [n-Bu4N][PF6]) by 20-Fc gold nanoparticles in CH2Cl2, showing
the shift of E1/2 toward positive potentials as a function of the number of
equiv of [n-Bu4N][H2PO4] added/amidoferrocenyl branch of the 20-Fc
colloids.
(
c) Other Anions. Finally, these particles do not recognize
the chloride, bromide, and nitrate anions, the potential shift being
nil or nonsignificant upon addition of a [n-Bu4N] salt of these
anions. Further work with other supramolecular systems and
techniques is underway for these anions.
d) Titration of [n-Bu4N][H2PO4] in the Presence of both
n-Bu4N][HSO4] and [n-Bu4N]Cl: Selectivity. It was of
interest to examine the behavior of the CV wave of the AFAT-
containing colloids as a function of the addition of [n-Bu4N][H2-
PO4] when other salts of anions were present in the solution.
This type of study allows investigating the selectivity for the
recognition of the H2PO4 anion. Thus, addition of [n-Bu4N][H2-
PO4], when the salts [n-Bu4N][HSO4] and [n-Bu4N]Cl (in
with the 13-Fc and 29-Fc colloids but reaches 175 mV with
the more highly AFAT-loaded 38-Fc colloids (see Table 3). It
is also noteworthy that the Epa - Epc value increases from 20
to 200 mV in the course of the addition, which is again a
characteristic of the relatively slow reorganization during the
redox change. The shift of potential of the wave (instead of
appearance of a new wave) when the other anions are present
shows, according to the Kaifer-Echegoyen model, a much
weaker interaction between the amidoferrocenyl groups and the
(
[
-
-
H PO anions in the presence of the other anions than in their
2
4
4
addition to the electrolyte [n-Bu4N][PF6] in 2 × 10 fold excess)
absence. This is fully confirmed by the reduction of potential
shift (120 mV instead of 220 mV). This shows that the apparent
association constant K(0) becomes much weaker in the presence
of the other anions, presumably because of the competition
among the different anions in their interactions with the neutral
amidoferrocenyl form of the AFAT ligands. The change of the
association constant may be only apparent, however, at constant
ionic strength. The observed change may in fact be the result
of a dynamic equilibrium among the anion complexes. Despite
this competition that considerably influences the recognition
are present in solution in concentrations equal to that of the
amidoferrocenyl ligands, does not lead to the appearance of a
new wave at a less positive potential than the initial one. On
the other hand, this addition leads to a progressive shift of the
initial wave toward more positive potentials. This means that,
contrary to the situation for which these salts are absent, it is
-
now possible to titrate the H2PO4 anions by plotting the
potential shift as a function of the addition of [n-Bu4N][H2-
PO4]. We have carried out these titrations with the 13-Fc, 20-
Fc, 29-Fc, and 38-Fc colloids, and the titration graphs led to
equivalence points reached for 0.87, 0.82, 0.75, and 0.75 equiv
of [n-Bu4N][H2PO4], respectively (see Table 4 and Figure 5).
Thus, this equivalence point is approaching unity when the
proportion of AFAT ligands becomes low. This difference of
behavior when the other anions are present shows that there is
competition and perturbation among the anions that interact with
the amidoferrocenyl groups of the colloids. Nevertheless, the
-
features, there is selectivity for the recognition of the H PO4
2
anion. Although the potential shift is precisely the event
observed with [n-Bu N][HSO ] alone, the magnitude is con-
4
4
siderably larger with [n-Bu4N][HSO4] + [n-Bu4N]Cl +
[n-Bu4N][H2PO4] (120 mV) than with [n-Bu4N][HSO4] alone
(30 mV). With the 38-Fc colloids, the significantly higher value
of the potential shift indicates that the competition with the
-
-
HSO4 anions is weaker, possibly because the HSO4 anions
cannot disrupt the dimers formed between AFAT neighbors
-
H2PO4 anion can be conveniently titrated. The shift of potential
at the equivalence point vs the initial wave is 120 ( 10 mV
-
whereas the H2PO4 anions essentially do. In addition, the
1788 J. AM. CHEM. SOC.
9
VOL. 124, NO. 8, 2002

Redox Recognition of Oxoanions
A R T I C L E S
increase of the Epa - Epc value along the titration shows that
the redox change of the amidoferrocenyl-branch anion involves
some structural rearrangements whose kinetics is of the order
of magnitude of the electrochemical time scale, i.e. rather slow.
Such slow heterogeneous electron transfers are not found with
the other anions alone, but they are characteristic of the rather
Cp), 68.2 (CH C
5.0 (CH CH ). Anal. Found for 8-Fc: Au, 69.42; S, 4.41. Calcd for
Au/S: 2.56/1. Found for 13-Fc: Au, 69.24; S, 3.34. Calcd for Au/S:
.37/1. Found for 20-Fc: Au, 63.37; S, 4.08. Calcd for Au/S: 2.53/1.
Found for 29-Fc: Au, 61.35; S, 3.93. Calcd for Au/S: 2.54/1. Found
for 38-Fc: Au, 57.37; S, 3.58. Calcd for Au/S: 2.61/1. Found for 18-
5
H
4
CO), 40.0 (CH
2
NH), 33.0-30.0 (CH
2 2
), 23.5 (CH ),
1
2
3
3
6
Fc (C ): Au, 69.19; S, 3.60. Calcd for Au/S: 3.13/1. Mean diameter
-
strong amidoferrocenium branch-H2PO4 anion interaction. In
of gold core for AFAT-functionalized gold nanoparticles (TEM.): D
the case of the shorter C6 chain, however, no increase of the
Epa - Epc value was observed in the course of the titration,
although the other features are as indicated above with the C11-
AFAT-containing chains. This could be the result of a weaker
13
3
)
3ν
2.06 nm. Calculation of the number of gold atoms:
N
Au ) 4πR /
3
g
) 4π(D/2) /51 ) 269. Calculation of the number of thiolate ligands
N ) N /(Au/S) ) 269/2.61 ) 103. Molecular weight calculation (38-
th
Au
-1
Fc): Au269(C12H25S)64(SC11H22NHCOFc)39. Average M ) 82035 g‚mol .
-
amidoferrocenium-H2PO4 interaction due to the buried char-
Gold Colloids Functionalized with 11-(1-Amido,1′-pentameth-
ylferrocenyl)-1-undecanethiol, 7. The nanoparticles were synthesized
according to the procedure described above. The black solid was dried
1
2c
acter of this shorter-chain AFAT ligand.
In summary, the titration of the H2PO4- anions can be carried
1
in vacuo. H NMR (CDCl
3
): δ (ppm) 4.33 (CH (C
NH), 1.83 (C CH ), 1.25 (CH
): δ (ppm) 169.2 (CO), 81.0 (C CO), 74.2 (C
CO), 39.8-39.2 (CH NH), 32.0-27.0 (CH ), 22.7 (CH
CH ), 10.6 (C CH ). Anal. Found: Au, 38.20; S, 2.32. Calcd
5 4
H CO)), 3.84 (CH
out in the presence of other anions, and the best choice is the
colloids containing a proportion of 13% C11 AFAT ligands (see
Table 4).
1
3
(C H
5 4
CO)), 3.35 (CH
NMR (CDCl
0.7 (C
14.1 (CH
2
5
3
2
), 0.88 (CH
3
).
CO),
),
C
3
5
H
4
5 4
H
7
5
H
4
2
2
2
Concluding Remarks
2
3
5
3
for Au/S: 2.68/1. Mean diameter of gold core (measured by TEM):
D ) 2.56 nm.
The stereoelectronic effects of the Cp permethylation and Cp
acetylation of the amidoferrocenyl group have a significant
influence on the recognition properties that are weakened by
Cp permethylation (electron releasing and sterically demanding)
and enhanced by Cp acetylation (electron withdrawing). These
results emphasize the key role, for the recognition by the
amidoferrocenium form, of the H-bonding between the -NH-
amido group and the charged terminal oxygen atom of the
oxoanion.
Gold Colloids Functionalized with 11-(1-Amido,1′-acetylferro-
cenyl)-1-undecanethiol, 15. The colloids were synthesized according
1
to the above procedure. The black solid was dried in vacuo. H NMR
(CDCl ): δ (ppm) 4.74 (CH(C H CO)), 4.53 (CH(C H CO)), 4.34 (CH-
3
5
4
5
4
(C H
5 4
CO)), 3.36 (CH
2
NHCO), 2.41 (CH
3 2
CO)),1.67-1.25 (CH ), 0.88
3
(CH ). Anal. Found: Au, 53.56; S, 3.79. Calcd for Au/S: 2.30/1.
Titrations. The calculation method used for the titrations (with
example of 38-Fc) is as follows. Quantity added into the CV cell:
-
1
m
)
c
particles ) 5 mg; M38-Fc particle ) 82035 g‚mol ; nparticles ) mparticles/M
One can titrate the oxoanions using either the decrease of
the initial CV wave and the increase of the new wave at less
-
8
6.095 × 10 mol. Quantity of anion added in 5 mL of solution:
-8
sol ) 0.01 M f nsol ) 5 × 10 mol. An addition of 5 mL of solution
allows the complexation of 0.82 branch (5 × 10 /6.095 × 10
.82). At the equivalence point (38-Fc), X/mL of solution of anion is
-
-
positive potential (H2PO4 ) or the potential shift (HSO4 ), the
-
8
-8
)
1
3-Fc and 20-Fc colloids being the best choices for good results
0
-
and accuracy. Moreover, it is also possible to titrate the H2PO4
anions in the presence of both [n-Bu4N][HSO4] and [n-Bu4N]-
Cl. This titration is characterized by a potential shift signifying
added. Y ) (0.82X)/5 amidoferrocenyl branches are complexed. It is
considered that 1 equiv is reached when the number of AFAT branches
1
found by
H NMR is reached. For 38-Fc, 1 equiv is reached for a
4
PO :
-
an apparently lower H2PO4 anion-AFAT interaction due to
number of AFAT ligands equal to 39. Data for the titration of H
2
the competition and perturbation by the other anions. The
potential shift allows a convenient titration, ideally with the 13-
Fc colloids.
number of branches at the equivalence point, Y ) (0.82 × 193)/5 )
-
3
2; Number of equivalents of H
PO
2 4
added, 32/39 ) 0.8.
Acknowledgment. We thank Michel Chambon (Bordeaux I)
for the HRTEM measurements, Professor Jean-Claude Blais
Paris VI) for mass spectra, Victor Martinez (LCOO) for helpful
Experimental Section
(
For general data, cyclic voltammetry, HRTEM, and ligand synthesis,
see the Supporting Information.
discussions, and the Institut Universitaire de France (D.A.), the
CNRS, the University Bordeaux I, and the R e´ gion Aquitaine
(grant to A.L.) for financial support.
Gold Colloids Functionalized with AFAT Ligands. Nanoparticles
stabilized with dodecanethiolate ligands were synthesized by following
4
the method described by Brust et al. In a typical procedure, these gold
Supporting Information Available: Ligand syntheses (Schemes
S1 and S2), additional references and comments on colloid
syntheses, CV, and recognition (Figures S1 and S2), general
nanoparticles (ca. 200 mg, 2.0 µmol) were dissolved in 100 mL of
CH
degassed and added under N
which was stored in a dry Schlenk tube. The mixture was stirred at
room temperature for 48 h under N . The solvent was evaporated to
Cl
2 2
(concentrated in nanoparticles ca. 2 mg/mL). The solution was
2
to the 11-(amidoferrocenyl)undecanethiol,
1
13
data including cyclic voltammetry, H and C NMR spectra of
the AFAT ligands and mixed alkanethiolate/AFAT gold colloids,
and HRTEM pictures of the mixed gold colloids (PDF). This
material is available free of charge via the Internet at
http//pubs.acs.org.
2
dryness, and the residue was washed thoroughly with methanol until
the methanolic phase was colorless. The black solid was dried in vacuo.
1
H NMR (CDCl
3
): δ (ppm) 4.68 (CH (C
NH), 1.26 (CH
): δ (ppm) 170.3 (CO), 70.3 (CH C
5
H
4
CO)), 4.32 (CH (C
), 0.88 (CH
). 13C
CO), 69.7 (CH
5 4
H -
CO)), 4.19 (CH (Cp)), 3.35 (CH
NMR (CDCl
2
2
3
3
5
H
4
JA017015X
J. AM. CHEM. SOC.
9
VOL. 124, NO. 8, 2002 1789