Luminescent nanocomposites containing CdS nanoparticles dispersed into vinyl alcohol based polymers

Reactive & Functional Polymers 68 (2008) 1144–1151

Contents lists available at ScienceDirect

Reactive & Functional Polymers

journal homepage: www.elsevier .com/locate / react

Luminescent nanocomposites containing CdS nanoparticles dispersedinto vinyl alcohol based polymers

Andrea Pucci a,*, Massimiliano Boccia a, Fernando Galembeck b, Carlos Alberto de Paula Leite b,Nicola Tirelli c,*, Giacomo Ruggeri a,d,e

a Department of Chemistry and Industrial Chemistry, University of Pisa, Via Risorgimento 35, I-56126 Pisa, Italyb Institute of Chemistry, Universidade Estadual de Campinas, P.O. Box 6154, CEP 13084-862, SP, Brazilc School of Pharmacy, University of Manchester, Stopford building, Oxford Road, M13 9PT Manchester, United Kingdomd INSTM, Pisa Research Unit, Via Risorgimento 35, 56126 Pisa, Italye PolyLab-CNR, c/o DCCI, University of Pisa, via Risorgimento 35, I-56126 Pisa, Italy

a r t i c l e i n f o a b s t r a c t

Article history:Received 14 December 2007Received in revised form 17 March 2008Accepted 31 March 2008Available online 7 April 2008

Keywords:Cadmium sulphide nanoparticlesPolymer nanocompositesLuminescence(nano-)dispersionOptical responsiveness

1381-5148/$ - see front matter � 2008 Elsevier Ltddoi:10.1016/j.reactfunctpolym.2008.03.007

* Corresponding authors.E-mail addresses: [email protected] (A. Pucc

chester.ac.uk (N. Tirelli).

Size-controlled cadmium sulphide nanoparticles (CdS) stabilized by mercaptoethanollayers were prepared in solution and successively dispersed into different poly(vinylalcohol)-based polymer matrices. The absorption and the emission features of the CdSnanocomposites were found to differ from those of colloidals dispersions and to be mainlyaffected by the particle (loose) aggregation in phase-separated, particle-rich regions of thematerials. It was also found that the optical behaviour of the nanocomposites can be mod-ulated through uniaxial orientation, which is presumed to partially destroy theaggregations.

� 2008 Elsevier Ltd. All rights reserved.

1. Introduction

Size, morphology and aggregation of inorganic nanopar-ticles dramatically influence their optical and electronicproperties, therefore methods to control these variablesoffer ideal means for modulating the physical propertiesof such materials [1–5]. For example, only clusters of noblemetals and not the corresponding macroscopic materials,such as smooth surfaces or powders, assume a real colourdue to the absorption of visible light at the surface plasmonresonance (SPR) frequency. Also colloidal oxide or sulfide(II–VI groups) nanocrystals made of a few hundreds up toa few thousands of atoms (quantum dots, QDs) are receiv-ing considerable attention, due to their appealing proper-

. All rights reserved.

i), nicola.tirelli@man-

ties derived from the zero-dimensional quantumconfined characteristics [6–10]; in particular attentionhas been focused on the size-dependence of absorptionand emission features, and, more generally, of opto-elec-tronic properties and charge-transfer phenomena [11]. Inthis area, however, control of size alone may not be enoughand one must take into account that surface chemicalderivatization plays a major role too, e.g. because of thepresence of surface defect sites, or because of the controlover surface interactions leading to agglomeration. Thecombined control over particle size and distribution, sur-face properties and aggregation behaviour can then opennew applications in optics, electronics, catalysis and biol-ogy [12–15].

In this area, much effort has been devoted to the fabri-cation of nanocomposites containing nanostructured andcrystalline inorganic semiconductors dispersed in poly-meric matrices, which confer them with superior process-ability and possibly prevent agglomeration of the

A. Pucci et al. / Reactive & Functional Polymers 68 (2008) 1144–1151 1145

nanomaterials (kinetic stability) [16–24]. For example,nanocomposites containing highly dispersed QDs havebeen used to develop optically functional materials in or-der to confer or enhance the photoconducibility of hostpolymers or to modify their refractive index [25,26].

In this work, we report the preparation and the opticalproperties of new luminescent nanocomposites based onthe dispersion of cadmium sulphide (CdS) nanoparticleswithin different vinyl alcohol-containing polymer matri-ces. CdS nanoparticles are inorganic semiconductors,widely used for light-emitting diodes (LED) and as quan-tum dots [15]; they combine interesting opto-electronicproperties, simplicity of preparation, possibility to tailortheir surface chemistry and therefore to provide them withmiscibility in a dispersing matrix. Indeed, the dispersion ofindividual nanostructured inorganic semiconducting parti-cles within a mechanically supporting polymer host wouldbe an easy method to prepare new, advanced but low costoptical materials.

We have prepared nano-sized CdS in solution at roomtemperature controlling the size of the aggregates throughthe use of a surface-capping agent, mercaptoethanol,which at the same time provides surface functionality[27,28]. We have then investigated the dispersability ofthe OH-covered nanoparticles in polar and protic polymermatrices. In particular, we have compared poly(vinyl alco-hol) (PVA) with a poly(ethylene-co-vinyl alcohol) (EVAl)copolymer (with 0.44 ethylene molar fraction), whose par-tially hydrophobic character can reduce some of the well-known drawbacks of PVA films, i.e., low thermal stabilityand moisture sensitivity.

We here discuss the influence of the nature of the poly-mer matrix on the size, morphology and optical propertiesof the CdS nanoparticles, and how the optical behaviour ofthese materials may be influenced by the nature of thepolymer, aggregation behaviour and mechanical stress-induced deformation.

2. Experimental

2.1. Materials

Cadmium chloride (CdCl2 � 2.5H2O, >99.5%) was pur-chased from Carlo Erba (Italy) and was used as received.All the other chemicals were purchased from Aldrich andwere used without further purification. For the preparationof CdS nanoparticles, deionised and distilled water wasused.

2.2. Sample nomenclature

Two polymer matrices were used:Poly(vinyl alcohol) (PVA, 99+% hydrolyzed, Mw =

146,000–186,000, supplied by Aldrich),Poly(ethylene-co-vinyl alcohol) with 44% by mol of eth-

ylene content (EVAl44, 60.7% of non-hydrolyzed vinylacetate units, melt index (210 �C, ASTM D 1238) = 3.5 g/10 min, density (25 �C) = 1.14 g/mL, supplied by Aldrich).

Samples were named by polymer, nanoparticle, concen-tration and draw ratio, e.g. EVAl44CdS_0.5_2.

2.3. CdS nanoparticles synthesis

CdS nanoparticles were synthesized with mercap-toethanol as a capping stabilizing layer following a modi-fied kinetic trapping method [27]. Briefly, 2.28 g(10 mmol) of CdCl2 � 2.5H2O was dissolved in 800 mL ofdeionised water previously degassed by passing N2 undervigorous stirring. After dissolution, 30 mL of degassedwater containing 100 mmol of 2-mercaptoethanol wasadded and the pH was raised by the addition of dilutedNaOH solution to pH 6.8. Then, 50 mL of a 0.2 M solutionof Na2S � 9H2O was added dropwise in the dark to the cad-mium solution with rapid stirring and the pH incrementedfurther to the final value of 8.20. The reaction mixture wasstirred for 2 h under nitrogen atmosphere in the dark andthe solution was then concentrated to 50 mL and purifiedby dialysis (SpectrumLabs, Cellulose Ester, 10 mL, MWCO:500) against water. After purification the CdS particleswere separated by size-selective precipitation: 200 mL ofTHF was added to the solution and a solid yellow precipi-tate was recovered after centrifugation and freeze-drying.Thermogravimetric analysis performed on the dry powderindicated a product composition of 70% CdS and 30% or-ganic thiol (onset: 258 �C, under nitrogen atmosphere).

2.4. Nanocomposite preparation

The typical procedure for the preparation of CdS/poly-mer nanocomposites is reported as follows:

0.335 g of the polymer (PVA or EVAl) was dissolved in20 mL of solvent (respectively deionised water for PVA,dimethylsulfoxide for EVAl matrix) under stirring at110 �C. After cooling to room temperature the desiredamount of CdS was added and dissolved under gentle stir-ring. The resultant yellow dispersions were cast into poly-tetrafluoroethylene (PTFE) Petri dish and kept in the darkduring solvent evaporation (2 days under hood at roomtemperature for PVA and at 45 �C for EVAl solutions).

Oriented composites were obtained by uniaxial tensiledrawing of the polymer matrix on thermostatically con-trolled hot stage at 110 �C for PVA mixtures and 90 �C forEVAl films. The draw ratio, defined as the ratio betweenthe final and the initial length of the sample, respectively,was determined by measuring the displacement of ink-marks printed onto the films before stretching.

2.5. Physico-chemical characterization

FT-IR spectra were recorded with a Perkin–Elmer Spec-trum One spectrometer as dispersions in KBr.

X-ray diffraction (XRD) patterns were obtained in Bragg-Brentano geometry with a Siemens D500 KRISTALLOFLEX810 (CT: 1.0 s; SS: 0.050 dg and Cu Ka, k = 1.541 Å) diffrac-tometer. Data were acquired at room temperature. Curve-fitting error estimates for the peak widths were calculatedusing the Origin data analysis software.

Thermogravimetric scans were obtained with a Perkin–Elmer TGA-7 under nitrogen flux, at a scan rate of 20 �C/min.

Bright field transmission electron microscopy (TEM)pictures were obtained on polymer composites by using

Fig. 1. XRD pattern of CdS nanoparticles capped by mercaptoethanol.

1146 A. Pucci et al. / Reactive & Functional Polymers 68 (2008) 1144–1151

a Carl Zeiss CEM-902 transmission electron microscope,equipped with a Castaing–Henry–Ottensmeyer energy fil-ter spectrometer within the column. Other experimentaldetails regarding TEM and sample preparation are pre-sented elsewhere [29]. Particle analysis was performedusing the public domain Image Tool 3.00 version imageanalyzer program developed at the University of TexasHealth Science Center in San Antonio and is available onInternet at http://ddsdx.uthscsa.edu/dig/itdesc.html.

UV–Vis absorption spectra of the polymer films wererecorded under isotropic conditions with a Perkin–ElmerLambda 650.

Steady-state fluorescence spectra of the polymer nano-composites were acquired at room temperature under iso-tropic excitation with the help of a Perkin–ElmerLuminescence spectrometer LS55 controlled by FL Winlabsoftware, fitted with motor-driven linear polarizers andequipped with the Front Surface Accessory: i.e., the posi-tion of the sample was adjusted in the direction of the exci-tation beam in such a way that the optical axis ofexcitation and emission crossed in the film plane.

The film roughness was diminished using ultra-pure sil-icon oil (poly(methylphenylsiloxane), 710� fluid, Aldrich)to reduce surface scattering between the polymeric filmsand the quartz slides used to keep them planar.

Origin 7.5, software by Microcal Origin�, was used inthe analysis of the XRD and spectroscopic data.

Digital images were obtained by using a Canon Power-Shot Pro1 camera exposing the films under a Camag UV-Cabinet II equipped with Sylvania 8W long-range lamps(366 nm).

3. Results and discussion

3.1. Characterization of CdS nanoparticles

Yellow coloured, readily water-dispersable CdS nano-particles were produced in water at slightly basic pH. Thecontrol of the initial pH (with the literature values rangingfrom 6.5 to 10) is well known to be a crucial factor forobtaining size-controlled CdS sols [27].

In thermogravimetric analysis the nanoparticles exhib-ited a thermally-induced weight loss due to desorption oforganic materials starting at 258 �C; this accounts forroughly 30% of the total weight, indicating therefore thatthere are slightly more sulphur atoms from mercap-toethanol (66% of the total S atoms) than from sulphideions. This is not unexpected, since the nanoparticles arevery small and therefore exhibit a large surface area, whichcorresponds to an even larger number of cadmium ligandsites (cadmium is tetracoordinated) available there. Thehigh temperature necessary for observing this desorp-tion/degradation ensures that the conditions later usedduring nanocomposite preparation and drawing experi-ments (T < 110 �C) would not harm the chemical integrityof the nanoparticles.

The X-ray diffractogram of the CdS nanoparticles inpowder form is reported in Fig. 1.

Three diffraction peaks can be found at about 27�, 44�and 51� and attributed to the (111), (220) and (311)

planes of cubic CdS phase [22,30,31]. Compared with bulkCdS, the diffraction peaks of the nanoparticles appearedbroadened due to the reduced particle size and surface de-fects. Actually, according to the Scherrer’s equation [29]0.9 � k/D(2h) � cosh on the (111) peak XRD pattern sug-gested an average crystallite size range of 1.5–2.4 nm.

Bright-field TEM of the CdS nanoparticles from colloidaldispersions (Fig. 2) seems to indicate them to be presentas objects with a (irregular) spherical shape – possibly H-bonded aggregates – but also possibly as individual nano-meter-sized nanoparticles (grey spots on whitebackground).

CdS nanoparticles within the aggregates appeared toexhibit a monomodal distribution in size with an averagediameter range of 2.33 ± 0.84 nm, which is in a good agree-ment with the XRD measurements.

Fig. 3 shows the absorption and luminescence behav-iour of CdS nanoparticles dispersed in dimethylsulfoxide(DMSO). A strong absorption peak at 346 nm is assignedto the optical transition of the first excitonic state of theCdS nanoparticles and its rather narrow shape is an evi-dence of the very small size of the dispersed particles[17,32]. The mean size of the CdS nanoparticles in DMSOcan be calculated from the onset absorption (or absorptionedge) of the UV–Vis spectrum (in this case pointed atke = 385 nm) and it generally represents the large-sizeend of the size distribution [17,19,33]. In our case one ob-tains 2.3 nm, which is in a good agreement with the valuesfrom XRD and TEM; however, since this measurement pro-vides a limit value, the real average dimensions may besmaller than that.

CdS nanoparticles typically show two characteristicemission bands: one at 470 nm, attributed to recombina-tions from the excitonic state in the crystalline interior,and one at higher wavelengths at about 540 nm, assignedto hole–electron recombinations at surface traps [34]. Inour case (Fig. 3), a strong emission band appeared at470 nm with a small shoulder above 500 nm and maybe attributed to recombinations from the excitonic stateonly.

Fig. 2. Bright-field transmission electron micrograph (A) and particle size distribution (B) of CdS nanoparticles.

Fig. 3. UV–Vis and emission spectra of a dilute DMSO dispersion (1 wt.%)of CdS nanoparticles, showing the absorption edge (ke) at 385 nm. For theemission spectrum, kexc. = 390 nm.

Fig. 4. Bright-field transmission electron micrograph of the cross-sectionof a PVACdS_1 composite film.

A. Pucci et al. / Reactive & Functional Polymers 68 (2008) 1144–1151 1147

3.2. Preparation and structural characterization of CdS/polymer nanocomposites

Composites were prepared by dispersing the CdS nano-particles in a 1.6 wt.% polymer solution, in water for poly-(vinyl alcohol) and in DMSO for poly(ethylene-co-vinylalcohol), which were then cast as a film.

All TEM measurements performed on the cross-sectionsof the films showed that the particles tend to accumulateclose to the film surface exposed to air during casting(Fig. 4). This would suggest that during solvent evapora-tion a liquid–liquid phase separation in a polymer-richand a nanoparticle-rich component can occur, similar towhat often happens for dispersions of low MW compoundsin polymers [35] and despite the presence of OH groupson both nanoparticles and polymers; the interactionsbetween these groups are evidently not sufficient forensuring good miscibility of the two systems.

An extensive analysis of TEM images taken on differentfilm portions of both PVA and EVAl nanocomposites showsthat the nanoparticle size distribution has a main peak(that can be easily fit with a Gaussian curve. See Supple-mentary material) flanked by a more or less conspicuouscomponent at larger sizes.

In PVA nanocomposites (Fig. 5A) the average size of thedispersed objects is roughly analogous to that observed bydrying the nanoparticle dispersions in DMSO, although theaverage value is shifted to somewhat lower values (1.7 nmcompared to 2.0 nm); on the other hand, EVAl matricesseem to disperse distinctly smaller particles (Fig. 5B), withan average dimension close to 1 nm and with little pres-ence of large aggregates.

This behaviour is not due to a better dispersion of theCdS nanoparticles in EVAl; on the contrary, the separationbetween particle-rich and particle-poor regions in EVAl iseven more evident there than in PVA (pictures on the rightin Fig. 5). In EVAl dispersions we face therefore the appar-ent oxymoron of having smaller nanoparticles in regionswhere they are more densely present.

We can, however, hypothesize that (a) the OH-coveredCdS nanoparticles also aggregate in the absence of amatrix: there is no reason why they should remain

Fig. 5. Particle size distributions (left) and bright-field transmission electron micrographs (right) for CdS nanoparticle dispersions in PVA (A) and EVAl (B) asa function of concentration and compared to the original nanoparticles (in grey). TEM micrographs are referred to polymer films containing the 0.5 wt.% ofCdS nanoparticles.

1148 A. Pucci et al. / Reactive & Functional Polymers 68 (2008) 1144–1151

separated during solvent removal; (b) in PVA a similaraggregation takes place, but it may be mediated by the‘‘bridging” polymer chains that interact with surfacegroups of different nanoparticles; we suppose that because(c) in the EVAl matrix the polymer chains, due the presenceof apolar sections, could bridge less efficiently, allowingtherefore for a larger distance between the nanoparticlesand therefore permitting to recognize individual objects.In such a condition it would seem counterintuitive thatnanoparticles coated with EVAl chain and less stronglyassociated cannot dispersed better in the matrix: however,phase separation can originate not only because of poorsurface interactions, but possibly also because of the over-all polarity of the nanoparticles, which have a considerableenergetic gain in forming a polar (high dielectric constant)particle-rich phase rather than being dispersed in a lowdielectric constant matrix.

The question arising at this point is whether the opticalproperties of the nanocomposites will be more influencedby the small-distance physical separation between theindividual nanoparticles (proximity at a <1 nm scale,which should proceed in the order solution < EVAl < PVAdry state) or by their longer-range phase segregation at>1 nm scale (concentration in the particle-rich phase,

which seems to be PVA > EVAl). In the first case, EVAlnanocomposites should have a behaviour intermediate be-tween PVA samples and DMSO dispersion; in the secondcase, PVA nanocomposites should be mid-way betweenEVAl ones and DMSO dispersion.

3.3. Optical characterization of CdS/polymer nanocomposites

The absorption band of CdS is indeed considerablyaffected by the dispersion of the nanoparticles in polymermatrices (Fig. 6).

It is apparent that the band appearing at 350 nm forsamples in DMSO dispersion is considerably red-shiftedfor both PVA and EVAl dispersions; the extent of the shiftdoes not depend on the concentration. Indeed, since thisphenomenon is caused by a quantum-confined effect[18], it depends on the aggregation of the nanoparticles,which TEM has shown not to be influenced by theirconcentration.

PVA has produced a shift of 35 nm, while EVAl has pro-duced a shift of 50 nm, whereas the absorption edge (ke)moves from 385 nm (CdS in DMSO) to 428 nm forPVACdS_0.5 and 453 nm for EVAl44CdS_0.5. Additionally,EVAl composites showed a considerably higher increase

Fig. 7. Comparison between the fluorescence spectra of PVA and EVAlnanocomposites containing 2 wt.% of CdS particles with pure CdS sols(kexc. = 390 nm).

Fig. 6. Comparison between the UV–Vis spectra (after removing thescattering background through an exponential fitting; the original spectraare in the inserts) of PVA (a) and EVAl (b) nanocomposites containingdifferent concentrations of CdS particles with a 1 wt.% CdS dispersion inwater.

1 For interpretation of color in Fig. 8, the reader is referred to the webversion of this article.

A. Pucci et al. / Reactive & Functional Polymers 68 (2008) 1144–1151 1149

of the difference between absorption edge and absorptionpeak values, i.e., 68 nm for EVAl and 43 nm for PVA. Utiliz-ing the generally accepted correlation between ke and theaverage CdS size, the diameter of the dispersed particlesin PVA and EVAl can be estimated to be 2.9 and 3.6 nm,respectively. The discrepancy with the values obtainedfrom TEM pictures is, however, only apparent: indeed theEVAl dispersions show more dense large aggregates(although smaller isolated nanoparticles) and thereforethe optical behaviour may be ascribed mainly to thislarge-scale aggregation rather to the clustering of isolatednanoparticles.

The fluorescence spectra of PVA and EVAl44 nanocom-posites containing the 2 wt.% of CdS particles are comparedin Fig. 7 with the emission behaviour of the CdS dispersionin DMSO.

It is noteworthy that, differently from DMSO or waterdispersions, where degradation [33,36,37] or dynamic

equilibrium [38] between free or bound capping ligandsleads to a clear modification of the emission features ofthe CdS nanoparticles (usually a quenching of the lumines-cence intensity), the PVA- or EVAl-nanocompositesshowed optical stability exceeding 6 months, a phenome-non that may be ascribed to the much slower dynamicsin the polymer matrices.

The emission spectra of the two nanocomposites aresubstantially analogous, with a �50 nm red-shift with re-spect to the emission of the particles in DMSO, conferringto PVA and EVAl films a typical orange-red colour1 whenexcited with a long-range UV radiation (Fig. 8).

This shift, which confines CdS emission in the region>470 nm, does not appear to have a strong dependenceeither on the nanoparticle concentration or on the natureof the matrix. It seems reasonable to ascribe it to autoab-sorption of the emitted radiation, due to the increase inthe nanoparticle concentration following phase segrega-tion. However, it cannot be excluded that the presence ofbound polymer chains on the nanoparticle surface may in-crease the emission from the interface (the band at 470 nmis often attributed to luminescence from excitonic states inthe nanocrystal interior and that at about 400 nm torecombinations of excitons and/or shallowly trapped elec-tron hole pairs [22,39]). These hypotheses can be verifiedby reducing the extent of phase segregation without dras-tically changing the degree of interactions with the poly-mer chains, e.g. via high temperature drawing of thematerial.

Indeed the uniaxial orientation of PVA and EVAl44 nano-composites (respectively performed at 110 �C and 90 �Cand at a draw ratio of 2 and 4) caused a substantial changein the optical properties: (a) a blue-shift is observed in theabsorption peaks, with the absorption edges being shiftedto 10 (PVA) and 25 nm (EVAl) with respect to the pristinenanocomposites (inset in Fig. 9); (b) analogously, the emis-

Fig. 8. Digital images of EVAl44CdS_2 film under visible light (a) and under excitation with a long-range UV lamp (b, k = 366 nm).

Fig. 9. Normalised emission (kexc. = 390 nm) and UV–Vis absorption (inset) spectra of EVAl44CdS_2 film before and after uniaxial orientation at draw ratioof 2 and 4, respectively.

1150 A. Pucci et al. / Reactive & Functional Polymers 68 (2008) 1144–1151

sion (EVAl shown in Fig. 9) resulted increasingly similar tothat of the CdS particles dispersed in DMSO with increas-ing drawing ratio (=increasing forced dispersion in the ma-trix, as indicated by the position of the absorption edge). Inparticular, luminescent experiments performed by varyingthe excitation wavelength did not produce any significantchange of the emission of both oriented and unorientedfilms (i.e., emission maximum and band shape, seeFig. S3 in the Supplementary material section).

This aggregation-dependent shift of the emission spec-trum has been observed beforehand [20,21,31,34]. Nodefinitive explanation has been provided but it was specu-lated that the luminescence of nanocrystals is assigned astrap-induced electron–hole recombination and/or recom-bination from an excitonic state. However, it seems thatthe nature of the emitting states can change as the sizeand the aggregation extent of the clusters change.

It is noteworthy that uniaxial orientation could leavenanoparticle aggregates oriented along the drawing direc-tions; in the present case, however, we record a complete

lack of linear dichroism (see Supplementary material),which indicates that no anisotropic structure was formed.

4. Conclusions

Very small (Ø � 2.3 nm) luminescent CdS nanoparticles,capped with a mercaptoethanol stabilizing layer, were effi-ciently prepared in water under controlled pH conditionsand dispersed by solution-casting into vinyl alcohol con-taining polymer matrices, i.e., poly(vinyl alcohol) (PVA)and poly(ethylene)-co-(vinyl alcohol) with 44% by mol ofethylene content (EVAl44). The nanoparticles clearlyphase-separate, but our results seem to indicate that theinteractions between the hydroxyl groups of the host ma-trix and mercaptoethanol residues allow the significantpresence of polymer chains in the particle-rich phase,which may influence the nanoparticle short-distance sepa-ration. Additionally, we have shown that the features ofthe phase segregation seem to be the main factors in regu-lating the optical properties of the nanocomposites and,

A. Pucci et al. / Reactive & Functional Polymers 68 (2008) 1144–1151 1151

additionally, that these features can be controlled e.g.through mechanical orientation.

Acknowledgements

The authors wish to thank Prof. Francesco Ciardelli(DCCI, Pisa) for the very helpful discussions. Financial sup-port by MIUR-FIRB 2003 D.D.2186 grant RBNE03R78E iskindly acknowledged. A.P. acknowledges the Royal Societyof Chemistry for the journals grant for internationalauthors (application no.: 07 04 584).

Appendix A. Supplementary material

Supplementary data associated with this article can befound, in the online version, at doi:10.1016/j.reactfunctp-olym.2008.03.007.

References

[1] U. Kreibig, M. Vollmer, Optical Properties of Metal Clusters, Springer-Verlag, Berlin, 1995.

[2] K.J. Klabunde, Nanoscale Materials in Chemistry, Wiley Interscience,New York, 2001.

[3] K.L. Kelly, E. Coronado, L.L. Zhao, C. Schatz George, J. Phys. Chem. B107 (2003) 668–677.

[4] H.H. Huang, X.P. Ni, G.L. Loy, C.H. Chew, K.L. Tan, F.C. Loh, J.F. Deng,G.Q. Xu, Langmuir 12 (1996) 909–912.

[5] J.P. Wilcoxon, B.L. Abrams, Chem. Soc. Rev. 35 (2006) 1162–1194.[6] H. Weller, Adv. Mater. 5 (1993) 88–95.[7] C.J. Murphy, Anal. Chem. 74 (2002) 520A–526A.[8] H. Sakurai, Organomet. News (2004) 100.[9] I. Willner, B. Willner, Pure Appl. Chem. 74 (2002) 1773–1783.

[10] T. Murakata, Y. Higashi, N. Yasui, T. Higuchi, S. Sato, J. Chem. Eng. Jpn.35 (2002) 1270–1276.

[11] P.D. Cozzoli, T. Pellegrino, L. Manna, Chem. Soc. Rev. 35 (2006) 1195–1208.

[12] T. Trindade, P. O’Brien, N.L. Pickett, Chem. Mater. 13 (2001) 3843–3858.

[13] G. Schmid, B. Corain, Eur. J. Inorg. Chem. (2003) 3081–3098.[14] X. Michalet, F.F. Pinaud, L.A. Bentolila, J.M. Tsay, S. Doose, J.J. Li, G.

Sundaresan, A.M. Wu, S.S. Gambhir, S. Weiss, Science 307 (2005)538–544.

[15] G. Ozin, A. Arsenault, Nanochemistry: A Chemistry Approach toNanomaterials, Royal Society of Chemistry, Cambridge, 2005.

[16] V.S. Gurin, M.V. Artemyev, J. Cryst. Growth 138 (1994) 993–997.[17] M. Moffitt, H. Vali, A. Eisenberg, Chem. Mater. 10 (1998) 1021–1028.[18] Y. Lin, A. Boeker, J. He, K. Sill, H. Xiang, C. Abetz, X. Li, J. Wang, T.

Emrick, S. Long, Q. Wang, A. Balazs, T.P. Russell, Nature 434 (2005)55–59.

[19] S.-W. Yeh, K.-H. Wei, Y.-S. Sun, U.S. Jeng, K.S. Liang, Macromolecules36 (2003) 7903–7907.

[20] S.-W. Yeh, T.-L. Wu, K.-H. Wei, Nanotechnology 16 (2005) 683–687.[21] S.-W. Yeh, T.-L. Wu, K.-H. Wei, Y.-S. Sun, K.S. Liang, J. Polym. Sci. Part

B: Polym. Phys. 43 (2005) 1220–1229.[22] Y. Ni, H. Hao, X. Cao, S. Su, Y. Zhang, X. Wei, J. Phys. Chem. 110B

(2006) 17347–17352.[23] W. Caseri, Hybrid Mater. (2007) 49–86.[24] W.R. Caseri, Mater. Sci. Technol. 22 (2006) 807–817.[25] D.Y. Godovsky, Adv. Polym. Sci. 153 (2000) 163–205.[26] G. Carotenuto, A. Longo, P. Repetto, P. Perlo, L. Ambrosio, Sens.

Actuators B 125 (2007) 202–206.[27] T. Vossmeyer, L. Katsikas, M. Giersig, I.G. Popovic, K. Diesner, A.

Chemseddine, A. Eychmueller, H. Weller, J. Phys. Chem. 98 (1994)7665–7673.

[28] C. Barglik-Chory, D. Buchold, M. Schmitt, W. Kiefer, C. Heske, C.Kumpf, O. Fuchs, L. Weinhardt, A. Stahl, E. Umbach, M. Lentze, J.Geurts, G. Muller, Chem. Phys. Lett. 379 (2003) 443–451.

[29] A. Pucci, N. Tirelli, E.A. Willneff, S.L.M. Schroeder, F. Galembeck, G.Ruggeri, J. Mater. Chem. 14 (2004) 3495–3502.

[30] M.Z. Rong, M.Q. Zhang, H.C. Liang, H.M. Zeng, Appl. Surf. Sci. 228(2004) 176–190.

[31] N. Herron, Y. Wang, H. Eckert, J. Am. Chem. Soc. 112 (1990) 1322–1326.

[32] M. Tamborra, M. Striccoli, R. Comparelli, M.L. Curri, A. Petrella, A.Agostiano, Nanotechnology 15 (2004) S240–S244.

[33] L. Spanhel, M. Haase, H. Weller, A. Henglein, J. Am. Chem. Soc. 109(1987) 5649–5655.

[34] R. Premachandran, S. Banerjee, V.T. John, G.L. McPherson, J.A. Akkara,D.L. Kaplan, Chem. Mater. 9 (1997) 1342–1347.

[35] N. Tirelli, S. Amabile, C. Cellai, A. Pucci, L. Regoli, G. Ruggeri, F.Ciardelli, Macromolecules 34 (2001) 2129–2137.

[36] V. Biju, R. Kanemoto, Y. Matsumoto, S. Ishii, S. Nakanishi, T. Itoh, Y.Baba, M. Ishikawa, J. Phys. Chem. C 111 (2007) 7924–7932.

[37] D.E. Henneke, G. Malyavanatham, D. Kovar, D.T. O’Brien, M.F. Becker,W.T. Nichols, J.W. Keto, J. Chem. Phys. 119 (2003) 6802–6809.

[38] X. Cao, C.M. Li, H. Bao, Q. Bao, H. Dong, Chem. Mater. 19 (2007)3773–3779.

[39] W. Chen, Z.G. Wang, Z.J. Lin, L.Y. Lin, Solid State Commun. 101 (1996)371–375.