Ultrafine selective metal-complexing nanoparticles: Synthesis by microemulsion copolymerization, binding capacity, and ligand accessibility

Article abstract of DOI:10.1021/ma0113632

Stable translucent aqueous suspensions of ligand-functionalized nanoparticles, with average diameters ranging from 12 to 20 nm and containing from 0.25 to up to 0.75 mmol of ligand/g, are readily obtained by copolymerization in oil-in-water microemulsions

Full text of DOI:10.1021/ma0113632

1644
Macromolecules 2002, 35, 1644-1650
Ultrafine Selective Metal-Complexing Nanoparticles: Synthesis by
Microemulsion Copolymerization, Binding Capacity, and Ligand
Accessibility
Son ia Am igon i-Ger bier ,† Sylva in Deser t,‡ Th a d e´e Gu lik -Kr ysw ick i,§ a n d
Ch a n ta l La r p en t*,†
SIRCOB ESA CNRS 8086, LRC CEA DSM 95-1, Universite´ de Versailles St Quentin en Y.,
45 Avenue des Etats-Unis, 78035 Versailles Cedex, France; Service de Chimie Mole´culaire,
CEA-Saclay, 91191 Gif sur Yvette Cedex, France; and Centre de Ge´ne´tique Mole´culaire, CNRS,
91198 Gif sur Yvette Cedex, France
Received J uly 30, 2001; Revised Manuscript Received November 28, 2001
ABSTRACT: Stable translucent aqueous suspensions of ligand-functionalized nanoparticles, with average
diameters ranging from 12 to 20 nm and containing from 0.25 to up to 0.75 mmol of ligand/g, are readily
obtained by copolymerization in oil-in-water microemulsions of a polymerizable tetraaza macrocycle,
vinylbenzylcyclam (cyclam: 1,4,8,11-tetraazacyclotetradecane). These nanoparticles exhibit a very high
selectivity for cupric ions with binding capacities of up to 0.65 mmol of Cu/g as well as a remarkable
ligand accessibility, deduced from spectrophotometric titration, with about 75% of the cyclam residues
located near the surface for nanoparticles in the 12-13 nm range. Core-shell bimetallic nanoparticles
are obtained by cation exchange. The cyclam-functionalized nanoparticles are comparable to high-
generation PAMAM dendrimers in their ability to complex metals.
In tr od u ction
variety of purposes as diverse as separation and recov-
ery of metal ions,2,20-24,27 catalysis,2,25,26,28,29 dioxygen
transport,30 or sensors.31 In this field the use of micro-
emulsions as polymerization media offers new prospect
for the development of efficient colloidal supports with
improved specific area and ligand accessibilty. Two
pioneer studies have shown that metal complexing
nanoparticles in the 15-35 nm diameter range can be
readily synthesized by a one-step copolymerization
process with monomers containing a macrocycle14 or a
bipyridine unit.16
Polymer supports have been used in organic chemis-
try and catalysis for many years.1-3 This technique has
revolutionized organic synthesis in the past decade and
created a new field known as combinatorial chemistry.4,5
Since the activity of supported ligands, catalysts, or
reagents depends on the accessibility of the active sites
and is often limited by diffusion, considerable efforts
have been made to develop new polymer supports with
improved capacity, accessibility, and selectivity as well
as suitable solvent affinity.5-8 The past decade has
witnessed a significant transformation in the area of
particle science and technology with the emergence of
nanoparticulate materials. The ability to control the
surface characteristics of such nanoparticulate systems
assumes paramount importance because of the high
surface-to-volume ratio of the particles.7
In this context, the use of microemulsions as reaction
media for the synthesis of polymers offers new op-
portunities. Indeed, the polymerization of oil component
in oil-in-water microemulsions leads to polymer nano-
particles with diameters smaller than 30 nm and
specific areas larger than 300 m2/g.9-11 Moreover, be-
cause of the well-defined structure of microemulsions,
polymerization inside these systems allows one to
control the structural properties of the resulting polymer
and to synthesize special polymer materials, such as
copolymers with high degrees of chemical functional-
ization, which are not accessible by other techniques.9-18
On the other hand, resin beads or latex-supported
ligands as well as soluble ligand-functionalized macro-
molecules and dendrimers have been used as metal ion
complexing agents2,19-31 and proposed for an extensive
In this paper, we show that microemulsion copolym-
erization is a very useful method to produce stable
aqueous suspensions of ultrafine selective metal-com-
plexing nanoparticles with controlled size and adjust-
able metal-binding capacity. Using tetraazacyclotet-
radecane, so-called cyclam, as a ligand and its Cu(II)
complex as a sensor, the complexation process can be
monitored spectrophotometrically in the translucent
aqueous suspensions of nanoparticles. The well-known
cyclam ligand exhibits high affinities for transition
metal cations32 with thermodynamic stability constants
in the order Cu(II) . Ni(II) > other cations, an order
that remains unchanged after anchoring to a polymer
backbone or a resin.22,23 We take advantage of the
binding of cupric ions, which gives a very stable deep
violet copper complex (K ) 1027), to study the capacity,
accessibility, and selectivity of cyclam-functionalized
nanoparticles.
Exp er im en ta l Section
Syn th esis of Vin ylben zylcycla m (1).24 A mixture of
cyclam (2 g, 10 mmol) and 100 mL of pure DMF was stirred
at 100 °C for 20 min until complete dissolution. Vinylbenzyl
chloride (70/30 mixture of meta and para isomers purchased
from Fluka) (0.3 g, 2 mmol) in 4 mL of pure DMF was then
added dropwise, and the mixture was stirred at 100 °C for 2
h. After cooling at room temperature, unreacted cyclam that
precipitated was removed by filtration and the filtrate con-
† Universite´ de Versailles St Quentin en Y.
‡ CEA-Saclay.
§ Centre de Ge´ne´tique Mole´culaire, CNRS.
* To whom correspondence should be addressed. Tel (33)-
139254413; Fax (33)139254452; e-mail Larpent@chimie.uvsq.fr.
10.1021/ma0113632 CCC: $22.00 © 2002 American Chemical Society
Published on Web 01/31/2002

Macromolecules, Vol. 35, No. 5, 2002
Metal-Complexing Nanoparticles 1645
centrated under vacuum (0.2 mmHg). The crude product was
dissolved in light petroleum (50 mL) and washed with water
(3 × 50 mL). The organic layer was separated, dried over
MgSO4, and concentrated under vacuum leading to 1 (0.5 g,
80%) as a white sticky liquid. TLC, Rf ) 0.8 (CHCl3/MeOH/
NH3, 2/2/1). GC/MS-EI (capillary column OV1, 260 °C; HP
engine, 70 eV): tr ) 29.1 and 30.8 min; (m/z): 99.2 (21), 117.1
(94), 160.2 (100), 317.2 (1). 1H NMR (CDCl3, 300 MHz) δ
(ppm): 1.67 (m, 2H); 1.8 (m, 2H); 2.73 (broad, 20H); 5.25 (dd,
compositions of the isolated polymers are as follows: competi-
tion Ni/Cu: % C 75.22, % H 7.41, % N 3.99, % Ni 2.30, % Cu
0.51; competition Zn/Cu: % C 76.42, % H 7.54, % N 4.43, %
Zn 2.39, % Cu 0.33; competition Co/Cu: % C 76.92, % H 7.42,
% N 3.43, % Co 2.33, % Cu 0.30.
Meta l Bin d in g Ca p a city-Sp ectr op h otom etr ic Mea -
su r em en ts. UV/vis analyses were performed on a Perkin-
Elmer spectrophotometer UV/vis/NIR Lambda 19 equipped
with
a reflection sphere. Measurements were performed
2
3
2
3J ) 11 Hz, J ) 1 Hz, 1H); 5.78 (dd, J ) 17.6 Hz, J ) 1 Hz,
directly on the translucent suspensions resulting from polym-
erization. The spectra were recorded from 350 to 800 nm. In
every case, the maximum absorption wavelength of the copper
complex was observed at 536 ( 2 nm.
The ligand accessibility was studied by adding small ali-
quots (100-300 µL) of a 0.01 mol/L aqueous solution of copper
nitrate to 2 g of suspension placed in a 1 cm quartz cell. The
absorbance was measured 3 min after each addition. For the
study of diffusion-limited complexation processes, similar
titrations have been performed with 0.1 and 0.5 mol/L solu-
tions of copper nitrate.
The metal binding capacity was deduced from the absor-
bance of the Cu complex at the equilibrium in the presence of
stoichiometric amounts of copper nitrate (1 equiv of Cu per
cyclam residue in the polymer as deduced from elemental
analysis, copper concentration in the suspension ∼ (2-3) ×
10-2 mol/L). The same absorbance values were obtained when
the complexation were performed in the presence of an excess
of copper (10 equiv/cyclam) followed by removal of the unre-
acted cupric ions by dialysis (cellulose membrane, MWCO
3500) toward aqueous solutions of DTAB (15 wt %).
The spectroscopic study of the replacement of Zn(II) by Cu-
(II) was performed on a colorless suspension of Zn-containing
20 nm nanoparticles prepared by reacting an excess of Zn-
(NO3)2 with suspension C followed by removal of the unreacted
Zn(II) salt by extensive dialysis toward an aqueous solution
of DTAB (15%). The absorbance of the Cu-cyclam complex
was measured until the equilibrium was reached.
Qu a si-Ela stic Ligh t Sca tter in g. The analyses were per-
formed with a Brookhaven (BI2030AT correlator) equipped
with an 2016 Ar laser (514.5 nm); the data were analyzed by
the exponential sampling method or by the nonnegatively
constrained least-squares multiple pass method (BI30atn
software). The samples were diluted in pure water (0.2 mL of
suspension in 100 mL of water) before analysis.
F r eeze-F r a ctu r e Tr a n sm ission Electr on Micr oscop y.
A 20-30 µm thick layer of the sample was deposited on a thin
copper holder and then rapidly quenched in liquid propane.
The frozen samples were fractured in vacuo (about 10-7 Torr)
at liquid nitrogen temperature with the liquid nitrogen cooled
knife inside a Balzers 301 freeze-etching unit. The replication
was done using unidirectional shadowing with platinum-
carbon at the angle of 35°. The mean thickness of the metal
deposit was 1-1.5 nm. The replicas were washed with organic
solvents and distilled water and then observed in a Philips
EM 410 electron microscope. The contrast in images is related
to the depth fluctuations of the metal deposit. The electron
microscopy analyses have been performed on previously
dialyzed suspensions: cyclam-functionalized nanoparticles
were dialyzed toward DTAB (0.15 wt %), and metal-containing
nanoparticles were dialyzed toward pure water.
3
2
1H); 6.70 (dd, J ) 17.6 Hz, J ) 11 Hz, 1H); 7.30 (m, 4H).
P r ep a r a tion of th e Cycla m -F u n ction a lized Na n op a r -
ticles: Micr oem u lsion Cop olym er iza tion . Styrene (2) and
divinylbenzene (3) were purified by flash chromatography on
silica gel (eluent cyclohexane) before use. The surfactant
dodecyltrimethylammoniuim bromide, DTAB (purum, >98%),
was purchased from Fluka and used without further purifica-
tion.
P r ep a r a tion of th e Micr oem u lsion s. The microemulsions
were prepared by adding under gentle magnetic stirring the
mixture of monomers (0.4-0.83 g) to 20 g of a 15 wt % aqueous
solution of DTAB. In every case a clear transparent micro-
emulsion was obtained. When DMPA was used as the radical
initiator, it was introduced in solution in the monomers blend
before the preparation of the microemulsion. In sample C, 1.2
mL of NaOH (0.1 N) was added to the microemulsion before
polymerization. The compositions of the starting microemul-
sions (wt %) are as follows: A, C, and F: water 81.7%, DTAB
14,4%, 1 1%, 2 1.3%, 3 1.6%; B: water 81.9%, DTAB 14,4%, 1
0.5%, 2 1.6%, 3 2%; D: water 81.6%, DTAB 14,4%, 1 1%, 2
3%; E: water 82.5%, DTAB 14,6%, 1 1%, 2 1.9%.
P olym er iza tion . Oil-Soluble Radical Initiator, 2,2-Dimeth-
oxy-2-phenylacetophenone (DMPA) (0.05 mol/mol of monomers,
previously solubilized in the mixture of monomers): the freshly
prepared microemulsion was transferred in a double-jacketed
refrigerated 100 mL schlenck tube and degassed with nitrogen
for 20 min. The polymerization was performed at 20 °C under
white light irradiation using two 60 W lamps for 15 h.
Water-Soluble Radical Initiator, Ammonium Persulfate/
Tetramethyldiaminomethane:12,13 the freshly prepared micro-
emulsion was transferred in a three-necked flask and degassed
with nitrogen for 20 min. A solution of ammonium persulfate
(0.06 mol/mol of monomers) in the minimum amount of water
and pure tetramethyldiaminomethane (0.12 mol/mol of mono-
mers) were then successively added. The polymerization was
carried out at 20 °C for 15 h.
Gas chromatography12,13 and thin-layer chromatography
(light petroleum) indicated complete polymerization of all
monomers. In every case, translucent stable suspensions are
obtained.
Sep a r a tion a n d Ch em ica l Com p osition of th e Resu lt-
in g P olym er s. 20 mL of methanol was added to 10 mL of
suspension; the flocculated polymer was then separated by
centrifugation (4000 rpm, 15 min). The resulting white paste
was dispersed in demineralized water (100 mL) and heated
at 60 °C overnight. After centrifugation, the crude polymer
was washed twice again with water (2 × 100 mL) and dried
at 50 °C until constant weight. The absence of remaining water
was checked by ATG experiment (from 20 to 350 °C). The
1
complete removal of the surfactant was checked by H NMR
and confirmed by the absence of bromine in elemental analysis.
Elemental analyses have been obtained from the “Service
Central d’Analyse” (CNRS, Vernaison, France). A: % C 82.80,
% H 7.84, % N 3.52; C: % C 81.64, % H 8.48, % N 4.16; D: %
C 85.16, % H 8.05, % N 4.01; E: % C 70.25, % H 6.61, % N
3.43. F: % C 84.67, % H 8.07, % N 2.97.
Com p etition Exp er im en ts were performed on suspension
C according to the previously described procedure.14 Briefly,
about 10-2 mol of Ni(NO3)2, Zn(NO3)2, or CoCl2 and 10-5 mol
of Cu(NO3)2 were added to 5 g of suspension C, and the mixture
was allowed to equilibrate. The resulting suspensions were
then purified by dialysis (cellulose membranes, MWCO 3500).
The metal-containing polymers were obtained after concentra-
tion to dryness and extensive washings with water followed
by final drying with a purification yield of about 90%. The
Resu lts a n d Discu ssion
I. P r ep a r a tion a n d Ch a r a cter iza tion of Cycla m -
F u n ction a lized Na n op a r ticles. a . P r ep a r a tion by
Cop olym er iza tion in Micr oem u lsion . Cyclam-func-
tionalized nanoparticles are obtained by copolymeriza-
tion of the polymerizable tetraaza macrocyclic derivative
1, vinylbenzylcyclam, in oil in water microemulsions
(Scheme 1). Microemulsions of mixture of monomer 1
and styrene 2, with or without a cross-linking agent 3,
are prepared using a cationic surfactant DTAB (dode-
cyltrimethylammonium bromide)14 and polymerized at
room temperature using either an oil-soluble radical

1646 Amigoni-Gerbier et al.
Macromolecules, Vol. 35, No. 5, 2002
Sch em e 1. Syn th esis of th e Cycla m -F u n ction a lized
Na n op a r ticles a n d Cop p er Bin d in g: (i)
Cop olym er iza tion in Oil-in -Wa ter Micr oem u lsion , 3-4
w t %; (ii) Ad d ition of Cu p r ic Ion (or Oth er Meta llic
Ca tion ) to th e Aqu eou s Su sp en sion
Ta ble 1. Cycla m -F u n ction a lized Na n op a r ticles P r ep a r ed
by Micr oem u lsion Cop olym er iza tion
nanoparticles
starting microemulsion
ligand contentc,d
% 1
D,b mol mmol/g mmol/g
monomers
mol inita
nm
%
totc
surfd
A
B
C
D
E
F
1 + 2 + 3
1 + 2 + 3
1 + 2 + 3e
1 + 2
10
5
11
10
15
11
D
D
D
D
D
P
13
12
20
15
15
22
9
5
11
10
10
7
0.65
0.35
0.73
0.71
0.61
0.53
0.48
0.25
0.31
0.35
0.35
0.23
1 + 2
1 + 2 + 3
a
b
Initiator: P ) (NH4)2S2O8/TMDAM; D ) DMPA. Average
diameter determined by QELS and TEM ((10%). c Determined
d
by elemental analysis. Amount of cyclam surface residues de-
duced from spectrophotometric titration. e Polymerization per-
formed in the presence of 2 equiv of NaOH per 1.
photoinitiator (DMPA: 2,2-dimethoxy-2-phenylacetophe-
none) or a water-soluble redox system (ammonium
persulfate/tetramethyldiaminomethane).12-14 In every
case, polymerization readily takes place, reaching 100%
conversion of all monomers within a few hours, and
leads to stable translucent aqueous suspensions of
nanoparticles: the average diameters, determined by
quasi-elastic light scattering and electron microscopy,
range from 12 to 22 nm (Table 1). The content of cyclam
incorporated in the particles is deduced from elemental
analysis and can be easily modulated from the amount
of polymerizable macrocycle 1 initially introduced.
b. Ch a r a cter iza tion . Whatever the composition of
the starting mixture of monomers (molar ratio, presence
or absence of a cross-linking agent), DMPA-initiated
copolymerizations in cationic surfactant-based micro-
emulsions afford stable translucent aqueous suspen-
sions of very small nanoparticles in the 12-15 nm
diameter range, among the smallest ever described
(Table 1; samples A, B, D, E). As illustrated by freeze-
fracture electron microscopy in Figure 1a, these ul-
trafine particles are remarkly monodispersed with a
very narrow size distribution. Slightly larger particles,
with a mean diameter of 20 nm, are obtained in basic
medium, i.e., when the macrocycle is not protonated
(sample C). A pH-dependent location of the polymeriz-
able ligand 1 in between the oil droplet core and the
interfacial layer may account for these variations of size.
In basic medium, the neutral aza macrocycle is prefer-
entially solubilized within the oil droplet core and does
F igu r e 1. Freeze-fracture electron microscopy (×102 000): (a)
suspension of cyclam-functionalized nanoparticles A (DTAB
0.15 wt %); (b) surfactant-free suspension of Cu-cyclam-
functionalized nanoparticles A-Cu(II) (0.6 mmol of Cu/g of
polymer).
not significantly take part in the interfacial stabiliza-
tion: in this case, the particle size is quite similar to
those obtained by polymerization of styrene alone under
the same experimental conditions. On the other hand,
in a neutral medium, the aza macrocyclic comonomer 1
is protonated and get enriched at the oil-water inter-
face: it contributes to the droplet surface coverage, thus
giving rise to a decrease of the particle size according
to the previously proposed geometrical Antionietti-Wu
model.10,33
The molar ratio of polymerizable cyclam derivative 1
incorporated in the nanoparticles reaches up to 11%,

Macromolecules, Vol. 35, No. 5, 2002
Metal-Complexing Nanoparticles 1647
Ta ble 2. Liga n d Accessibility a n d Meta l Bin d in g Ca p a city
Ded u ced fr om Cu (II) Bin d in g Exp er im en ts a n d
Sp ectr oscop ic Stu d ies
i.e., 0.7-0.75 mmol of cyclam residues per gram of
polymer (Table 1). When higher concentrations are
introduced in the starting microemulsions, 1 is only
partly incorporated within the particles (sample E: 15%
introduced, 10% incorporated). In the latter case, po-
lymerization of 1 partly takes place in water, giving rise
to water-soluble low-molecular-weight polymers which
are removed during the purification.
Remarkably, the amount of cyclam ligand that can
be incorporated without increasing the particle size is
comparatively much higher than in previously reported
bipyridine-functionalized nanoparticles (only 2.5-3.5
mol %, i.e., 0.23-0.3 mmol of ligand/g for 25-35 nm
particles).16
nanolatex
A
B
C
D
cyclam content (mmol/g)a
accessible cyclam (mmol/g)b
% cyclam on the surface
Cu-cyclam max (mmol/g)c
diameter (nm)d
0.65
0.48
74
0.35
0.25
72
0.73
0.31
42
0.71
0.35
50
0.62
0.32
0.61
0.58
starting particles
13
12
20
15
+ Cu(II)
13
12
22
17
n cyclam/particlee
450
335
430
460
230
190
140
175
500
125
1840
780
1540
300
150
760
375
615
400
170
n cyclam surf/particlee
n Cu-cyclam max/particlee
specific area (m2/g)
cyclam surf (m2/g)f
c. Colloid a l Sta bility. The aqueous suspensions of
cyclam-functionalized nanoparticles prepared in the
cationic surfactant DTAB-based microemulsions using
the nonionic photoinitiator DMPA are remarkably
stable: the as-prepared suspensions can be stored for
more than 6 months without particle aggregation.
Furthermore, these colloidal suspensions tolerate changes
of pH, from pH 2 to 8, as well as changes of ionic
strength. They are stable in the presence of added
electrolytes, giving no flocculation up to 5 mol/L chloride
or nitrate salts of monovalent or divalent cations
(sodium, calcium, copper, zinc, etc.) and up to 0.5 mol/L
of sodium sulfate. Interestingly, these colloidal suspen-
sions remain stable without particle aggregation when
the surfactant concentration is reduced to 0.15 wt % by
dialysis, as illustrated by the TEM picture in Figure 1.
It is worth noticing that, considering the copolymer-
ization of the aza macrocyclic monomer 1, cationic
surfactant-based microemulsions should be used since
copolymerization in “classical SDS microemulsions”
prepared with the well-known sodium dodecyl sulfate
anionic surfactant12,13,34 was found to give unstable
suspensions of larger particles. The formation of ion
pairs between the macrocycle 1, protonated at neutral
pH, and the anionic surfactant may account for these
results. In the same way, similar ionic interactions may
explained that nanoparticles resulting from persulfate-
initiated polymerization (sample F) slowly aggregate
and finally give rise to large particles with a broad size
distribution after a few weeks. Consequently, nonionic
radical initiators, like DMPA used in the present work,
should be preferred for the synthesis of cyclam-func-
tionalized nanoparticles.
II. Cop p er Bin d in g Ca p a city of Cycla m -F u n c-
tion a lized Na n op a r ticles. a . Bin d in g Ca p a city a n d
Ch a r a cter iza tion of th e Cu (II)-Cycla m Na n op a r -
ticles. Owing to the high affinity of cyclam for cupric
ions,22,23,32 copper elemental analysis has been used to
estimate the metal binding capacity of the cyclam-
functionalized nanoparticles. Complexation experi-
ments, performed at room temperature in the presence
of stoichiometric amounts of copper nitrate, clearly
illustrate the high binding capacity and ligand acces-
sibility (Table 2). Whatever the starting suspension, the
complexation of more than 85% of the cyclam residues
is reached with copper contents of up to 0.6-0.65
mmol/g in the isolated particles. For 12-15 nm par-
ticles, complete complexation is achieved under sto-
ichiometric conditions (complexation yield >95%).
a
b
From elemental analysis. Outer cyclam residues deduced
from spectroscopic titration in dilute medium. c Under stoichio-
metric conditions, deduced from Cu-cyclam complexe absorption
d
and Cu elemental analysis. Mean diameter of the particles before
and after copper complexation, determined by QELS and TEM (
10%. e Number of cyclam and Cu-cyclam moieties per particle
calculated assuming a density of 1. f Specific area of surface-
cyclam residues assuming an area of 0.8 nm2 per cyclam residue.
tion, well dispersed all over the sample and the absence
of aggregates.
Particles as small as 12-13 nm and containing up to
400-450 Cu(II) per particle are thus readily obtained
(Table 2). The cation binding capacity of these cyclam-
functionalized nanoparticles, prepared by a straight-
forward one-step polymerization procedure, can be
compared to those of the more sophisticated dendrim-
ers: for example, generation six28 and generation eight27
poly(amidoamine) dendrimers, PAMAM, bind respec-
tively 64 and 153 Cu(II) ions per molecule, i.e., 1.1 and
0.65 mmol/g. It is worth noticing that the copper
concentration in our 12-13 nm particles, i.e., 350-450
mmol per nm3, is very close to those obtained in the
dendritic structures (320-420 mmol per nm3, calculated
from the respective dendrimers diameters: 6.7 and 9.7
nm).27,28
b. Sta bility of th e Cu (II)-Cycla m Na n op a r ticles.
The deep violet translucent suspensions of Cu-cyclam
functionalized nanoparticles exhibit a remarkable col-
loidal stability: they can be stored for months and do
not aggregate in the presence of up to 2 mol/L chloride
or nitrate salts. Moreover, neither flocculation nor
copper decomplexation is observed when the pH is
lowered down to pH 3. Furthermore, the surfactant can
be entirely removed by dialysis without destabilization
affording surfactant-free suspensions of nanoparticles
(TEM picture, Figure 1b). Electrostatic stabilization
arising from ionic repulsions between the positively
charged nanoparticles may account for the colloidal
stability of the aqueous suspensions of Cu-cyclam
nanoparticles.
III. Liga n d Accessibility. a . Sp ectr op h otom etr ic
Stu d ies of th e Com p lexa tion : Discr im in a tion be-
tw een Ou ter a n d In n er Liga n d s. Taking advantage
of the transparency of the suspensions, the amount of
Cu(II)-cyclam complex in the nanoparticles is easily
determined spectrophotometrically from its character-
istic absorbance: quantitative and reproducible mea-
surements, in good agreement with copper elemental
analyses, are obtained by using an integrating sphere
to collect and integrate the scattered light. These
spectroscopic studies give useful information on the
ligand accessibility. Thus, upon progressive addition of
a dilute 0.01 M solution of Cu(II), the suspensions of
Remarkably, the particle size remains unchanged
after complexation (Table 2). The freeze-fracture elec-
tron microscopy picture given in Figure 1b shows
individual particles, with a very narrow size distribu-

1648 Amigoni-Gerbier et al.
Macromolecules, Vol. 35, No. 5, 2002
F igu r e 2. Spectrophotometric titration of cyclam upon pro-
gressive addition of dilute Cu(NO3)2 (0.01 M) to suspensions
A, B, and D. Instantaneous absorbance of Cu-cyclam complex
at 536 nm vs copper concentration.
cyclam-functionalized particles instantaneously turn
violet, indicating that complexation readily takes place
even at very low copper concentrations. The suspensions
exhibit a maximum absorption wavelength at 536 nm
(ꢀ ) 134 L mol-1 cm-1) very close to those of the
monomeric Cu(II)/1 complex (537 nm, 134 L mol-1
cm-1), indicating that the binding to the nanoparticles
does not modify the structure of the complex. As can be
seen in Figure 2, the absorbance of the copper complex
linearly increases up to a maximum value that corre-
sponds to the instantaneous complexation of all the
accessible cyclam moieties in dilute medium. The amount
of cyclam residues on the surface can thus be readily
deduced from spectrophotometric titration and com-
pared for various suspensions (Table 2, Figure 2).
Furthermore, in the case of nanoparticles containing
the lowest amounts of accessible cyclam moieties, the
spectrophotometric studies clearly indicate that two
complexation processes take place: a rapid “solution-
like” complexation process involving the outer cyclam
residues located near the surface, which occurs at the
minute time scale in dilute medium, and a slow diffu-
sion-limited complexation process involving the inner
cyclam residues entrapped within the core of the
particles. Thus, as illustrated in Figure 3a, in the case
of 20 nm particles, about 42% of the cyclam moieties
are accessible for complexation at the minute time scale
in dilute medium (first plateau). A second complexation
step then proceeds much slowly and complete Cu(II)
binding is achieved after several hours. Upon further
stepwise additions of dilute copper, the same scenario
is observed: no instantaneous complexation (second and
third plateau) and then progressive complexation for
longer reaction times; the higher the concentration of
Cu-cyclam in the particles, the longer the complexation
time. Moreover, in agreement with a diffusion-limited
complexation process for the inner cyclam moieties, the
higher the copper concentration in solution, the higher
F igu r e 3. Spectrophotometric titration of cyclam upon pro-
gressive addition of Cu(NO3)2 to suspension C (nanoparticles
average diameter: 20 nm). Relative absorbance (per gram of
suspension) vs equivalents of Cu added per cyclam. (a)
Variation of absorbance with time upon addition of dilute 0.01
M Cu(NO3)2. (b) Instantaneous absorbance upon addition of
Cu(NO3)2 at various concentrations.
the extent of instantaneous complexation at a given Cu-
(II) per cyclam ratio (Figure 3b). It is worth noticing
that the maximum complexation yield, 85%, is reached
at the minute time scale upon addition of stoichiometric
amounts of Cu(II) as a 0.5 M solution, i.e., about 2.5 ×
10-2 mol of Cu(II)/L in the suspensions.
Similar behaviors have been observed for ultrafine
12-15 nm nanoparticles containing high densities of
outer cyclam residues (e.g., suspensions A and B) with
a much minor contribution of the diffusion-limited
process: the overall maximum complexation yield in
dilute medium is about 85-90% including 70-75% of
outer cyclam residues involved in a rapid Cu(II) binding
process and only 10-15% of inner cyclam residues
involved in a diffusion-limited complexation process. A
complexation yield of up to 95% is achieved at the
minute time scale upon addition of stoichiometric
amounts of Cu(II) as a 0.5 M solution.
These results clearly shed light on the huge improve-
ment of the ligand accessibility brought by decreasing

Macromolecules, Vol. 35, No. 5, 2002
Metal-Complexing Nanoparticles 1649
Ta ble 3. Selectivity for Cu p r ic Ion s fr om Com p etition
Exp er im en tsa ,b
the particle size with a clear correlation between the
percentage of accessible ligands and the surface-to-
volume ratio.
competing cation (M)
Ni(II)
Zn(II)
Co(II)
It is worth noticing that, in a previously reported
study of bipyridine-functionalized nanoparticles,16 the
authors observed “a slow approach to the finite color in
the presence of divalent metallic cations”: considering
the size of their particles (25-35 nm), the diffusion-
limited process evidenced above, involving inner bipy-
ridine moieties entrapped within the microgel particles
core, may well account for these observations.
M (mmol/g)
Cu (mmol/g)
0.39
0.08
>98
55
65
985
200
0.37
0.05
>98
52
60
935
125
0.39
0.05
>98
55
60
985
125
Cu complexation yield (%)
M/cyclam (%)
M + Cu/cyclam (%)
n M(II) per particlec
n Cu(II) per particlec
a
Starting suspension C: 20 nm. b Experimental conditions: Cu/
cyclam ) 0.1-0.15, M/cyclam ) 100, M/Cu ) 1000. c Number of
Cu or M residues per particle calculated assuming a density of 1.
b. Rela tion sh ip betw een th e Liga n d Accessibil-
ity a n d th e P olym er iza tion Con d ition s. The results
summarized in Tables 1 and 2 clearly demonstrate that
the density of ligand on the particle surface depends on
the copolymerization conditions and can be readily
modulated. The cyclam accessibility is remarkably
enhanced when the particle size decreases: thus, for
similar compositions of the starting microemulsions, the
percentage of cyclam residues located on the surface is
much higher on 12-15 nm particles resulting from an
“oil-initiated” polymerization at neutral pH (sample A,
75% of the whole cyclam residues) than on 20 nm
particles resulting from a “water-initiated” polymeriza-
tion at neutral pH or an “oil-initiated” polymerization
at basic pH (samples C and F, only 40-45%). Further-
more, as expected, for comparable sizes (13-15 nm) and
comparable cyclam contents, the amount of accessible
ligands is significantly increased by cross-linking (sus-
pensions D and A: 0.35 and 0.48 mmol/g prepared
respectively in the absence and in the presence of 40%
of cross-linking agent 3). Whatever the overall content
of cyclam, highly cross-linked microgel-type nanopar-
ticles exhibit a remarkable ligand accessibility since the
complexation of about 75% of the whole cyclam residues
is achieved in dilute medium at the minute time scale
(Table 2, suspensions A and B).
complexation of Cu(II) does take place even in the
presence of a very large excess of competing ion (M/Cu
) 1000; M/cyclam ) 100; Table 3). These results
demonstrate the remarkable selectivity of the cyclam-
functionalized nanoparticles for cupric ions: the ligand
selectivity is retained on particles. The nanoparticles
also bind Ni(II), Co(II), and Zn(II) with complexation
yields, reaching about 55% of the remaining free-cyclam
moieties. Moreover, the surfactant and the excess of
unreacted competing ions are easily removed by dialysis
without destabilization or decomplexation. Competition
experiments thus give access to stable surfactant-free
colloidal suspensions of “bimetallic” nanoparticles in
which the molar ratio of the two metallic cations can
readily be modulated.
b. P r ep a r a t ion of Cor e-Sh ell Typ e Bim et a llic
Nan opar ticles by Cation Exch an ge. Bimetallic nano-
particles with adjustable ratios of two cations, M2+ and
Cu2+, can alternatively be prepared by exchanging a
given cation M2+ (Zn2+, Co2+, Ni2+) for cupric ion. The
cation exchange is readily monitored by spectrophoto-
metric titration of the copper-cyclam complex. As
illustrated in Figure 4, in the case of 20 nm particles,
the displacement of Zn(II) by Cu(II) readily takes place
upon addition of dilute copper nitrate to a colorless
aqueous suspension of Zn-cyclam nanoparticles. In
dilute medium, copper complexation is quantitative
within a few hours (2-5 h) up to the substitution of 65%
of Zn and then reaches a plateau with longer equilib-
rium times. Nanoparticles containing adjustable Cu/Zn
ratios, from 0 to 2, are thus readily accessible by
addition of controlled Cu aliquots. Remarkably, in dilute
medium, the maximum substitution yield corresponds
to the complexation of about 45% of the whole cyclam
moieties by Cu(II): this value corresponds to the
amount of accessible ligands located on the surface for
the 20 nm particles studied (Table 2, suspension C).
These results clearly indicate that, in dilute medium,
the substitution is limited to the outer cyclam residues
and therefore results in the production of core-shell
type nanoparticles with a Cu-rich shell and a Zn-rich
core.
In agreement with previously reported studies, these
observations can be explained by considering that the
mechanism of copolymerization in microemulsion de-
pends on the microenvironment and especially on the
relative local concentrations of the monomers at the
region where the initiation step takes place.9,10-13,17
When the comonomer is preferentially located at the
oil-water interface, like 1 in neutral medium, and when
the free radicals are produced within the oil droplets
core (oil-soluble initiator DMPA), the polymerization
proceeds first with styrene until the comonomer con-
centration at the reaction loci becomes sufficient and
finally leads to a high density of hydrophilic ligand
residues in the particle shell.12 The effect of the pref-
erential orientation is much pronounced in the presence
of a lipophilic cross-linking agent since kinetic prevails
over thermodynamic.10 In basic medium, the nonproto-
nated comonomer 1 is preferentially located within the
oil droplets core: the copolymerization takes place
randomly and leads to a random distribution of the
cyclam residues over the whole particle volume, thus
resulting in a restricted ligand accessibility.12
Con clu sion
Microemulsion copolymerization is a useful technique
to produce ultrafine ligand-functionalized nanoparticles
with high density of ligands and up to ca. 300 m2
cyclam-functionalized surface per gram of polymer. One
of the most outstanding features of copolymerization
using microemulsion as reaction medium is the control
of the surface characteristics. The spectroscopic study
of the complexation process clearly indicates that the
ligand accessibility is closely correlated with the particle
size. Moreover, the properties of a selective ligand like
IV. Selectivity of Cycla m -F u n ction a lized Na n o-
p a r ticles. a . Com p etition Exp er im en ts. The selec-
tivity of the cyclam-functionalized nanoparticles for
Cu(II) and their binding capacity for other divalent
cations were studied by performing competition experi-
ments with various metallic cations (M: Ni(II), Co(II),
Zn(II)). In every case, elemental analyses of the metals
content in the isolated particles show that complete

1650 Amigoni-Gerbier et al.
Macromolecules, Vol. 35, No. 5, 2002
R. A.; Pritchard, M. C.; Reid, A. M. Angew. Chem., Int. Ed.
2001, 40, 1053-1055.
(5) Haag, R. Chem. Eur. 2001, 7, 327-335.
(6) Li, W.; Yan, B. J . Org. Chem. 1998, 63, 4092-4097.
(7) Hodges, I. C.; Hearn, J . Langmuir 2001, 17, 3419-3422.
(8) Guyot, A.; Hodge, P.; Sherrington, D. C.; Widdecke, H. React.
Polym. 1991/1992, 16, 233-259.
(9) (a) Candau, F. In Handbook of Microemulsion Science and
Technology; Kumar, P., Mittal, K. L., Eds.; Marcel Dekker:
New York, 1999; Chapter 22, pp 679-712. (b) Candau, F. In
Polymerization in Organized Media; Paleos, C. M., Ed.;
Gordon and Breach Publishers: Philadelphia, 1992; Chapter
4, pp 215-282.
(10) Antonietti, M.; Basten, B.; Lohmann, S. Macromol. Chem.
Phys. 1995, 196, 441-466.
(11) Capek, I. Adv. Colloid Interface Sci. 1999, 80, 85-149.
(12) Larpent, C.; Bernard, E.; Richard, J .; Vaslin, S. Macromol-
ecules 1997, 30, 354-362.
(13) Larpent, C.; Bernard, E.; Richard, J .; Vaslin, S. React. Funct.
Polym. 1997, 33, 49-59.
(14) Larpent, C.; Amigoni-Gerbier, S. Macromolecules 1999, 33,
9071-9073.
(15) Antonietti, M.; Lohmann, S.; Van Niel, C. Macromolecules
1992, 25, 1139-43.
(16) Antonietti, M.; Lohmann, S.; Eisenbach, C. D.; Schubert, U.
S. Macromol. Rapid Commun. 1995, 16, 283-289.
(17) (a) Dreja, M.; Pyckhout-Hintzen, W.; Tieke, B. Macromol-
ecules 1998, 31, 272-280. (b) Pyrasch, M.; Tieke, B. Colloid
Polym. Sci. 2000, 278, 375-379.
(18) Favero, C.; Graillat, C.; Guyot, A. Macromol. Symp. 2000,
150, 235-240.
F igu r e 4. Replacement of Zn(II) by Cu(II) upon addition of a
dilute solution of Cu(NO3)2 (0.01 M) to Zn(II)-loaded nanopar-
ticles. Relative absorbance (per gram of suspension) at 536 nm
at the equilibrium, molar ratios Cu-cyclam complex/cyclam
and Zn-cyclam complex/cyclam vs copper concentration.
Starting suspension C-Zn(II) [20 nm, 0.37 mmol Zn/g, molar
ratio Zn/cyclam ) 0.62] obtained by reaction of an excess of
zinc nitrate with suspension C.
(19) Beauvais, R. A.; Alexandratos, S. D. React. Funct. Polym.
1998, 36, 113-123.
(20) Alexandratos, S. D.; Crick, D. W. Ind. Eng. Chem. Res. 1996,
35, 635-644.
(21) Bergbreiter, D. E.; Koshti, N.; Franchina, J . G.; Frels, J . D.
Angew. Chem., Int. Ed. 2000, 39, 1039-1042.
(22) Altava, B.; Burgette, M. I.; Frias, J . C.; Garcia-Espana, E.;
Luis, S. V.; Miravet, J . F. Ind. Eng. Chem. Res. 2000, 39,
3589-3595.
cyclam are not influenced by the binding to the particles
so that a real “solution-like” chemistry, very close to
classical supramolecular chemistry, becomes accessible
on colloidal polymer nanoparticles. A comparably simple
technique of microemulsion polymerization results in
polymer nanomaterials with properties very similar to
the more sophisticated dendrimers.27,28 Considering
their very large specific area up to ca. 450 m2/g, the
ligand accessibility and selectivity, metal-complexing
nanoparticles are very attractive supports for a variety
of applications like specific recovery, sensors, and the
development of new immobilized catalysts including
atom transfer polymerization catalysts.35 Moreover, as
illustrated in this paper, such selective ligand-function-
alized nanoparticles may also find promising applica-
tions in materials science for the development of new
polymer composites.
(23) Puanik, D. B.; David, V. A.; Morris, R. E.; Chang, E. L. Energy
Fuels 1997, 11, 1311-1312.
(24) Blain, S.; Appriou, P.; Chaumeil, H.; Handel, H. Anal. Chim.
Acta 1990, 232, 331-336.
(25) Blacker, N. C.; Findlay, P. H.; Sherrington, D. C. Polym. Adv.
Technol. 2001, 12, 183-196.
(26) Angelino, M. D.; Laibinis, P. E. J . Polym. Sci., Part A: Polym.
Chem. 1999, 37, 3888-3898.
(27) Diallo, M. S.; Balogh, L.; Shafagati, A.; J ohnson, J . H., J r.;
Goddard, W. A., III; Tomalia, D. A. Environ. Sci. Technol.
1999, 33, 820-824 and references therein.
(28) Crooks, R. M.; Zhao, M.; Sun, L.; Chechik, V.; Yeung, L. K.
Acc. Chem. Res. 2001, 34, 181-190 and references therein.
(29) Vassilev, K.; Ford, W. T. J . Polym. Sci., Part A: Polym. Chem.
1999, 37, 2727-2736.
(30) Cameron, J . H.; Graham, S.; Harvey, B. H.; Liggat, J . J .;
McKee, A.; Soutar, I.; Scott, E. L. React. Funct. Polym. 1998,
36, 173-183.
(31) Wang, B.; Wasielewski, M. R. J . Am. Chem. Soc. 1997, 119,
12-21.
(32) Tho¨m, V. J .; Hosken, G. D.; Hancock, R. D. Inorg. Chem.
1985, 24, 3378-3381. (b) Alexander, V. Chem. Rev. 1995, 95,
273-275. (c) Christensen, J . J .; Eatough, D. J .; Izatt, R. M.
Chem. Rev. 1974, 74, 351-384.
(33) Antonietti, M.; Bremser, W.; Mu¨schenborn, D.; Rosenauer,
C.; Schupp, B.; Schmidt, M. Macromolecules 1991, 24, 6636-
6643.
(34) Guo, J . S.; Sudol, E. D.; Vanderhoff, J . W.; El-Aasser, M. S.
J . Polym. Sci., Part A: Polym. Chem. 1992, 30, 691-702,
703-712.
(35) Hong, S. C.; Paik, H. J .; Matyjaszewski, K. Macromolecules
2001, 34, 5099-5102.
Refer en ces a n d Notes
(1) Sherrington, D. C. J . Polym. Sci., Part A: Polym. Chem. 2001,
39, 2364-2377.
(2) Bergbreiter, D. E. J . Polym. Sci., Part A: Polym. Chem. 2001,
39, 2351-2363.
(3) Ford, W. T. React. Funct. Polym. 1997, 33, 147-158.
(4) (a) Kirschning, A.; Monenschein, H.; Wittenberg, R. Angew.
Chem., Int. Ed. 2001, 40, 650-679. (b) Blackburn, C. Biopoly-
mers (Peptide Science) 1998, 47, 311-353. (c) Eames, J .;
Watkinson, M. Eur. J . Org. Chem. 2001, 1213-1224. (d) Ley,
S. V.; Massi, A.; Rodriguez, F.; Horwell, D. C.; Lewthwaite,
MA0113632