Biomolecule induced nanoparticle aggregation: Effect of particle size on interparticle coupling

Journal of Colloid and Interface Science 313 (2007) 724–734www.elsevier.com/locate/jcis

Biomolecule induced nanoparticle aggregation:Effect of particle size on interparticle coupling

Soumen Basu, Sujit Kumar Ghosh, Subrata Kundu, Sudipa Panigrahi, Snigdhamayee Praharaj,Surojit Pande, Subhra Jana, Tarasankar Pal ∗

Department of Chemistry, Indian Institute of Technology, Kharagpur 721302, India

Received 20 February 2007; accepted 28 April 2007

Available online 3 May 2007

Abstract

Gold nanoparticles of variable sizes have been prepared by reducing HAuCl4 with trisodium citrate by Frens’ method. It has been found thatthe gold particles under consideration produce well-ordered aggregates upon interaction with a biomolecule, glutathione in variable acidic pHcondition and exhibit pronounced changes in their optical properties arising due to electromagnetic interaction in the close-packed assembly. Theeffect of nanoparticle size on the nature of aggregation as well as the variation in the optical response due to variable degree of interparticlecoupling effects amongst the gold particles have been investigated. The optical properties of the gold aggregates have been accounted in the lightof Maxwell-Garnett effective medium theory considering the changes in the filling factor in different aggregates produced by variable sizes ofgold colloids. The aggregates have been characterized by UV–vis spectroscopy, FTIR, Raman, XRD and TEM studies. It has been observed thata new peak appearing at a longer wavelength intensifies and shifts further to the red from the original peak position depends on the particle size,concentration of glutathione and pH of the solution. On the basis of the first appearance of a clearly defined new peak at longer wavelength,a higher sensitivity of glutathione detection has been achieved with gold nanoparticles of larger dimension.© 2007 Elsevier Inc. All rights reserved.

Keywords: Biomolecule; Surface plasmon; Nanoparticle assembly; Filling factor

1. Introduction

A burst of research activities has been seen in recent yearsfor the synthesis and characterization of metal nanoparti-cles, which arise from their numerous possible applicationsin physics, chemistry, biology, materials science and their dif-ferent interdisciplinary fields [1–8]. Due to their versatility inapplication, while evolution of the dispersion of small metal-lic particles with a tight size distribution is important, assemblyof individual nanoparticles into well-defined aggregates has re-cently become a widely pursued objective [9]. Most recently,emphasis has been given on organizing or assembling metalnanoparticles into defined architectures, mainly for two rea-sons. First, such metal nanoparticle aggregates can display

* Corresponding author.E-mail address: [email protected] (T. Pal).

0021-9797/$ – see front matter © 2007 Elsevier Inc. All rights reserved.doi:10.1016/j.jcis.2007.04.069

rich optical and electrical characteristics that are distinctlydifferent from a simple collection of individual particles orthe extended solid. Second, in relation to emerging electronictechnologies, more sophisticated nanostructures are in demand(e.g., nanowires, nanotubes, and their two dimensional (2D) andthree dimensional (3D) nanoparticle assemblages). Whereas at-tention is focused primarily on the assembly of macroscopiccrystals and films of dense cluster matter, there is also interest indeveloping methods for assembling small controlled aggregatesof nanoparticles in a solution [10–15]. Due to the sphericalsymmetry and uniform reactivity of individual nanoparticle sur-faces, the synthesis of small controlled nanoparticle assembliesis a significant challenge. To meet the practical challenges thatlie ahead, there still exists the need of suitable procedure for theevolution of metal nanoclusters assembled into an aggregatedstructure.

The optical properties of noble metal nanoclusters have fas-cinated scientists since recent past because of their applications

http://www.elsevier.com/locate/jcis

mailto:[email protected]

http://dx.doi.org/10.1016/j.jcis.2007.04.069

S. Basu et al. / Journal of Colloid and Interface Science 313 (2007) 724–734 725

as functional materials in optical devices [16,17], optical en-ergy transport [18,19], near-field scanning optical microscopy(NSOM) [20,21], surface enhanced Raman scattering spec-troscopy [22,23], and chemical and biological sensors [24,25].Characteristically, noble metal nanoparticles exhibit a strongabsorption band in the visible region and this is indeed a smallparticle effect, since they are absent in the individual atomas well as in the bulk [3,6]. The physical origin of the lightabsorption by metal nanoparticles is the coherent oscillationof the conduction band electrons induced by the interactingelectromagnetic field. The absorption band results when theincident photon frequency is resonant with the collective os-cillation of the conduction band electrons and is known as thesurface plasmon resonance (SPR). The resonance frequency ofthis SPR is strongly dependent upon the size, shape, dielec-tric properties, and local environment of the nanoparticles [3,26–30]. When a cluster of metal nanoparticles are placed inclose proximity to one another, the interparticle coupling ef-fect become very important to study the particle plasmon res-onances. Since the interparticle coupling is stronger than thecoupling within the surrounding medium, the Mie theory devel-oped for very dilute solutions and isolated particles fails to de-scribe the optical absorption spectrum. However, the effectivemedium theories, dating back to 1904, predicted by Maxwell-Garnett [31] have been successfully applied to this problemto account for this optical absorbance behavior of the metalnanoparticles present in a closed-packed assembly. Therefore,it is interesting to produce and study the interparticle interac-tions while the particles are held by weak forces in the aggre-gate.

The interesting colors observed in gold sols have led to ex-tensive study of their optical spectroscopic properties in aneffort to correlate their behavior under different microenviron-mental conditions [32–40]. Furthermore, gold is one of thevery few metals noble enough to survive as a nanoparticle un-der atmospheric conditions. This serendipitous combination ofproperties has encouraged its application in a diverse rangeof applications. In particular, with their excellent compatibil-ity with biomolecules [41–44] and distance-dependent opticalabsorbance [41,45], colloidal gold nanoparticles have drawnintense scientific and technological interest. In colloidal solu-tions, the color of gold nanoparticles may range from red topurple, to blue and almost black, due to the formation of ag-gregates. This is attributable to electric dipole–dipole interac-tion and coupling between plasmons of neighboring particlesin the aggregates. Exploitation of this unusual phenomenon hasgiven rise to new analytical and sensing techniques. The ag-gregation of gold nanoparticles induced by specific biologicalinteractions has attracted significant interest as a self-assemblyprocess for the construction of complex nanostructures that ex-hibit new collective properties. The interaction between organicligands and the surface of an inorganic nanoparticle paves theway for the coupling of biomolecular recognition systems togenerate novel materials. Although inorganic nanoparticles canbe prepared from various materials by several methods, thecoupling and functionalization with biological components hasonly been carried out with a limited number of chemical meth-

ods. For example, Kimura and co-workers have reported theformation of a three dimensional superlattice of gold nanopar-ticles with mercaptosuccinic acid [46]. Mirkin and Ratner [26]and Alivisatos et al. [47] have demonstrated the formation ofaggregated metal clusters, using DNA as the recognition ele-ment in aqueous medium. Some of the amino acids can induceaggregation for gold nanoparticles [48,49] and so on.

In this article, citrate-capped gold nanoparticles prepared byFrens’ method have been induced to aggregate by the additionof biomolecule, glutathione which can bind with gold nanopar-ticles by its amine group adjacent (α) to the carboxylic acidmoiety (–COOH) and –SH group. We have studied the interac-tion between GSH and gold colloids at different pH and it hasbeen found that GSH can bind with gold nanoparticles only atrelatively low pH, but suppressed at intermediate and high pH.We have investigated the nanoparticle size effect on the natureof aggregation amongst the gold particles. This nanoparticleaggregation process occurs with a concomitant color changefrom red (dispersed gold nanoparticles) to blue (aggregatednetworks), which can be monitored spectrophotometrically insolution. The modification in the optical response of the goldaggregates has been accounted in the light of Maxwell-Garnetteffective medium theory. From the kinetic experiments, it isclear that the interparticle spacing between the nanoparticlesdecreases with the progress of the reaction. The aggregates havealso been characterized by TEM, XRD, FTIR and Raman stud-ies. On the basis of the first appearance of a clearly-defined newpeak at longer wavelength, a higher detection limit (10−6 M) forglutathione is found for large gold particles.

2. Experimental

2.1. Reagents and instruments

All the reagents used were of AR grade. Chloroauric acid(HAuCl4·3H2O) was purchased from Aldrich. Trisodium cit-rate (Mallinckrodt) and glutathione (Aldrich) were used as re-ceived. Double distilled water was used throughout the courseof this investigation.

The absorption spectrum of each solution was recorded in aSpectrascan UV 2600 digital spectrophotometer (Chemito, In-dia) in a 1 cm well-stoppered quartz cuvette. Transmission elec-tron microscopy (TEM) was carried out on a Hitachi H-9000NAR transmission electron microscope, operating at 200 kV.Samples were prepared by placing a drop of solution on acarbon coated copper grid. FTIR spectral characteristics ofthe samples were collected in reflectance mode with Nexus870 Thermo-Nicolet instrument coupled with a Thermo-NicoletContinuum FTIR Microscope. One drop of the test solution wasplaced on a KBr pellet and was dried under vacuum for 6 h be-fore analysis. Raman spectra were obtained with a RenishawRaman Microscope, equipped with a He–Ne laser excitationsource emitting at a wavelength of 633 nm, and a Peltier cooled(−70 ◦C) charge coupled device (CCD) camera. The X-ray dif-fraction (XRD) pattern was recorded in an X’pert pro diffrac-tometer with Co (Kα = 1.78891) radiation.

726 S. Basu et al. / Journal of Colloid and Interface Science 313 (2007) 724–734

Table 1Details for the size-selective synthesis of gold nanoparticles by Frens’ method

SetNo.

Volume ofHAuCl410−2 M (µL)

Volume ofcitrate (1 wt%)(µL)

Color λmax(nm)

Particlesize(nm)

A 1250 2000 Dark red 518 8B 1250 1300 Red 520 13C 1250 1000 Red 522 16D 1250 875 Red 526 20E 1250 625 Pinkish red 529 32F 1250 400 Pink 534 55

2.2. Preparation of gold nanoparticles

To study the size effect on the process of aggregation ofthe metallic gold particles in the nanometer size regime, it isimportant to have sets of nanoparticles within the range of 1–100 nm with a tight size distribution. The well-documentedFrens’ method has been used to obtain monodispersed gold col-loids over a wide size range [50]. In a typical preparation, analiquot of 50 mL aqueous solution of HAuCl4 (0.25 M) washeated to boiling and 2 mL of trisodium citrate (1%) was addedfor 8 nm gold colloid. In about 25 s, the boiling solution turnsfaintly blue. After ∼70 s, the blue color changed to deep red.The solution was set aside to cool down to room temperature.The UV–vis spectral characteristics of variable sizes for differ-ent sets of gold particles are summarized in Table 1.

2.3. Synthesis of glutathione/Au assemblies

A stock solution of glutathione (10 mM) was prepared indouble distilled water. Dilute HCl solution (10 mM, 100 µL)was added to the gold colloids (3 mL) to lower the pH value(pH ∼4). Then, an aliquot of GSH solution (10 mM, 30 µL) wasadded to that deep red acidic gold colloidal solution. The colorchange of the gold sol from red to blue indicates the formationof the aggregates amongst the gold particles and the changes inthe surface plasmon resonance was measured with the UV–visspectrophotometer.

3. Results and discussion

3.1. Evolution of gold nanoparticle aggregates: measuringsurface plasmon oscillation

Six different sizes of gold nanoparticles have been employedto investigate the size effect on the aggregation behavior (i.e.,size, shape and morphology of the produced aggregates) metal-lic particles in the nanometer size regime. The particle sizewas varied within the size range of 8 to 55 nm (set A–F)where the concentrations of the gold are same in all cases. Thegold particles have been prepared by Frens’ method (employ-ing HAuCl4 as the precursor salt and trisodium citrate as thesurface capping agent) which offers us the ease of achievingmonodispersed gold colloids over a wide size range [50]. Inthis method, it is possible to control the size of the particles byvarying [Au(III)]/[citrate] ratio during the reduction step as has

Fig. 1. UV–vis spectral dependence on Au particle size, before (—) and after(· · ·) addition of GSH (0.1 mM) at pH ∼4.

Fig. 2. A plot of �λmax as a function of particle size.

been listed in Table 1. The color of the solution varies from redto pink depending on the size of the particles. The surface plas-mon resonance of the gold particles is red shifted with increasein particle size in accordance to Mie theory [51]. Now, uponaddition of glutathione to the citrate-stabilized gold particles,the color of all the solution becomes blue indicating the forma-tion of aggregates amongst the gold particles. The changes inthe UV–vis spectra of the resultant colloids were measured tostudy the size effect of metal nanoparticles on the surface plas-mon resonance due to aggregation. It is observed that differentsets of gold colloids show distinctly different absorption profilein the presence of glutathione molecules at a lower pH value.The salient feature of physical significance is that an extendedplasmon band is developed at longer wavelength and a clearbathochromic shift in λmax is observed with increasing particlesize from 8 to 55 nm, which is shown in Fig. 1. An almost linearincrease in peak displacement against particle size is observedin the range of 5–20 nm, but deviating slightly for 20–55 nm(Fig. 2).

The optical absorption behavior of the gold particles withsizes in the nanometer regime upon aggregation and its depen-dence on the individual particle size in the closely-packed as-sembly could be accounted within the framework of Maxwell-Garnett theory [31]. The optical properties of isolated colloidalparticles and in particular, their dependence on particle sizehave been extensively investigated through Mie scattering the-ory. In particular, the Mie theory is a mathematical–physical


description of the scattering of electromagnetic radiation byspherical particles immersed in a continuous medium. Since theinterparticle coupling in the nanoscale assemblies is strongerthan the coupling with the surrounding medium, the Mie’stheory developed for very dilute solutions and isolated par-ticles fails to describe the optical absorption spectrum. TheMaxwell-Garnet theory is an effective medium theory. It is nowknown theoretically and experimentally that when individualspherical gold particles come into close proximately to one an-other, electromagnetic coupling of clusters becomes effectivefor cluster–cluster distances smaller than five times the clusterradius (d � 5R) and may lead to complicated extinction spectradepending on the size and shape of the formed cluster aggregateby splitting of the plasma resonance. The optical properties ofthe metallic nanoparticles are mainly determined by two contri-butions: (1) the properties of the particles acting as well-isolatedindividuals and (2) the collective properties of the whole en-semble. Thus, in an ensemble of large number of particles, ifthe particles come close together, the oscillating electrons inone particle feel the electric field due to oscillation of the elec-trons in the surrounding particles and this leads to a collectiveplasmon oscillation of the aggregated systems. Under such situ-ation the isolated-particles approximation breaks down and theelectromagnetic interactions between the particles play a de-termining role to offer a satisfactory description of the surfaceplasmon oscillations.

The optical properties of an ensemble of metal nanoparticlesdepend both on the material properties of the constituents andon the structural parameters of the aggregate. For a selected par-ticle in an aggregate of metal particles, the dielectric function ofthe particle is the sum of all contributions of the electrical polar-ization including retarded electrodynamics in the neighboringparticles. Under the quasi-static approximation, the aggregatemay be assumed to be an effective medium and can be describedby an effective dielectric function, εeff = ε1,eff + iε2,eff, that ex-presses the linear response of the whole sample to the externalfield. To describe the aggregate geometry, the averaged volumefraction of the sample i.e., filling factor can be introduced as [3]

(1)f = Vcluster

Vsample= NR3

r3,

where R is the mean radius of the nanoparticles and r is the ra-dius of the aggregate in which both the individual particles andaggregates are assumed to be spherical. The Maxwell-Garnettrelation describes the link between the effective dielectric func-tion εeff and the volume fraction of the inclusion (f ) and canbe expressed as

(2)εeff − εm

εeff + 2εm= f

ε − εm

ε + 2εm,

where εm and ε are the dielectric functions of the surround-ing medium and the material itself. There is only one resonanceat ε = −2εm corresponding to the surface plasmon resonanceat ω = ωp/

√1 + 2εm of an isolated metal particle where ωp is

the frequency of plasmon oscillation of free electrons. Thus,the assembly of nanoparticles in various states of aggregationinfluences the plasma resonance, of which the frequency and

intensity depend on the degree of aggregation as well as orien-tation of the individual particles within the aggregate [52].

In the present experiment, the aggregation between thecitrate-stabilized gold nanoparticles of different size was in-duced by the short-chain molecular linker, glutathione. Glu-tathione acts as a mediator to direct the self-assembly of goldclusters into controlled ensembles with varied functional re-sponse. Thus, the interparticle distance corresponding to themolecular length of glutathione remains fixed for all sets ofgold nanoparticles. However, the assembly process providesmodular collective optical behavior, as examined through UV–vis spectroscopic measurements. As the particle size increasesfrom 8 to 55 nm, the extended plasmon band at longer wave-length shows a clear bathochromic shift in the absorption max-ima (Fig. 1). Thus, the magnitude of the spectral shift couldbe ascertained to the proximity effects due to nearest neighborinteractions between the particles. When gold nanoparticles as-semble into aggregated structures, there is an increase in thedielectric constant of the medium, shifting the plasmon peakto a lower energy. Both of these effects contribute to the redshift and broadening of the plasmon band to longer wavelengthcorresponding to the aggregated particles. As the particle size(individual) increases, the particles come at a closer distances inmaking the nanoscale aggregates. It was also noted that the timerequired for aggregation for larger particles is sufficiently smallthan that of the smaller particles. As the time of coalescence isessentially proportional to the T R4/(σsurfDself) with σsurf beingthe surface tension and Dself is the temperature self-diffusioncoefficient, at a particular temperature with increase in particlesize rate of aggregation between the gold particles increases.With increase in particle size, the averaged volume fraction ofthe aggregates (f ) increases and as a result, the single plas-mon band of isolated spheres starts to red-shift and broaden dueto dipole–dipole interactions in accordance to Maxwell-Garnetteffective medium theory.

3.2. Characterization of the gold aggregates

3.2.1. X-ray diffraction patternThe as prepared sample was centrifuged and the precipitate

was dried under vacuum and taken for XRD analysis. XRD pat-tern for the GSH-capped gold aggregates is shown in Fig. 3.Several peaks are observed, these being at 38.2◦, 44.5◦, 64.7◦,77.6◦ and 82.1◦, which correspond to the {111}, {200}, {220},{311} and {222} facets of the fcc crystal structure of gold, re-spectively. The observation of diffraction peaks for the goldnanoparticles indicates that these are crystalline in this sizerange while its broadening is related to the particles in thenanometer size regime. Although X-ray diffraction patterns ofthe 8 and 20 nm gold particles were measured to examine a dif-ference in their crystallinity, their diffraction patterns showedsimilar structure, as shown in Fig. 3.

3.2.2. FTIR spectroscopyFTIR and Raman spectra revealed the characteristic bands

of glutathione moiety after gold nanoparticle conjugation.Fig. 4 shows the FTIR spectra recorded for the gold aggre-


gates (curve 1) in the spectral window 400–4000 cm−1 alongwith the spectrum recorded from GSH powder (curve 2). Theevidence for the presence of surface-bound GSH is providedby FTIR measurements of the gold particles in the range of3200–3500 cm−1. The N–H stretch of the GSH molecules isobserved at 3250 cm−1. But upon complexation of GSH withgold nanoparticles, this band becomes sharper and shifts to3075 cm−1. This clearly indicates that GSH molecules bindto gold nanoparticles through the nitrogen atom. The positionof the carboxylate stretching vibration in the GSH capped goldremain unaltered at 1713 cm−1 which suggests that there is nohydrogen bonding amongst the GSH molecules. Inset of Fig. 4shows the Raman spectra of pure glutathione and glutathioneinduced gold aggregates. The free glutathione exhibits a charac-teristic band at ∼2530 cm−1 due to the presence of –SH group;in contrast, the nanoparticle aggregate sample exhibits no de-tectable bands in this region of the spectrum. This is a strongevidence of surface binding of GSH to the gold particles viathiolate linkage (soft–soft interaction) which agrees with ear-

Fig. 3. X-ray diffraction pattern of (a) 8 nm and (b) 20 nm gold aggregates.

lier studies on alkanethiol modification of gold nanoparticlesby Murray and co-workers [53].

3.2.3. TEM studiesBecause of the random nature of aggregate formation, the

synthesized gold nanoparticle aggregates have a broad distri-bution of sizes and shapes. This variety of sizes and shapesis apparent in the TEM images. Figs. 5d–5h show the TEMimages of large aggregates of 8, 13, 20, 32 and 55 nm goldnanoparticles with GSH, respectively. The aggregate formationis understandable while we compare the TEM images of nonag-gregated Au particles (Figs. 5a–5c) of three representative (8,20 and 55 nm) sizes with their aggregates. TEM images showthat dispersed gold particles become aggregated upon additionof glutathione molecules without any significant alteration inthe morphology of the particles. From the TEM images, it isevident that the variation in particle size are unable to pro-duce samples containing aggregates with a controlled geometry,rather the sizes and shapes of the aggregates varied over ex-tended ranges.

3.3. Kinetic study

3.3.1. Effect of reaction timeThe kinetic experiment was performed measuring the SPR

shifts during the aggregation process. Solution of GSH (10 mM,30 µL) was added to the acidic solution (pH ∼4) of 20 nm goldparticles (3 mL) at room temperature and the spectra were ac-quired over different time intervals, which are shown in Fig. 6.In the absorption spectra, overall a distinctive red shift of the520 nm peak to ∼670 nm is seen accompanied by a generationof new absorption peak at 520 nm. According to Mie theory,monodispersed gold nanoparticles in colloidal solution shouldexhibit only a single peak which can be attributed as due to

Fig. 4. FTIR spectra of pure glutathione (curve 1) and glutathione induced gold aggregates at pH ∼4 (curve 2) in the spectral windows 400–4000 cm−1. Inset showsthe Raman spectra of glutathione (curve 1) and gold aggregates (curve 2).


(a) (b)

(c) (d)

(e) (f)

(g) (h)

Fig. 5. Typical TEM images for (a) 8, (b) 20 and (c) 55 nm nonaggregated Au nanoparticles and (d) 8, (e) 13, (f) 20, (g) 32 and (h) 55 nm Au aggregates.


Fig. 6. Time-dependent absorption spectra of gold nanoparticles at pH ∼4 recorded at various times after addition of glutathione (0.1 mM). Inset shows a plot of�λmax as a function of time.

a dipole plasmon oscillation of the gold colloids induced bythe external electric field. If other clusters are nearby, the totalelectric field results from superposition of the external incidentfield and the dipole fields of all other clusters and the effect ofpolarization has a great influence on the electrodynamic calcu-lations of absorption spectra of the gold colloids. For sphericalparticles with polarization P , the next higher order correctionsinvolve rewriting the equation as

(3)P = α[Eloc + Erad],where α is the polarizability and the radiative correction field,Erad is given by

(4)Erad = 2/3ik3P + k2/RP,

where R is the particle radius. The first term in this expressiondescribes radiative damping which arises from spontaneousemission of radiation by the induced dipole while the secondterm comes from depolarization of the radiation across the par-ticle surface due to the finite ratio of particle size to wavelength.Thus, the first peak, situated near the resonance peak for singleparticles, is attributed due to the quadrupole plasmon excitationin coupled spheres, while the second peak at a longer wave-length is attributed to the dipole plasmon resonance of the goldnanoparticles [54]. As the interparticle spacing decreases, thefirst peak becomes weaker while the second peak intensifiesand shifts to the longer wavelengths. The quadrupole resonancealso red shifts as the particle size is increased, but the effect ismuch smaller than for the dipole resonance. A plot of �λmaxas a function of time (inset, Fig. 6) indicates that the aggrega-tion process becomes completed after a certain period of thereaction. Nanoparticle aggregates of gold with interparticle dis-tances substantially greater than the average particle diameter

Fig. 7. UV–vis spectra for Au colloidal solution with 0.01 M HCl added pro-gressively as blank effect of pH.

appear red with λmax ∼ 520 nm, but as the interparticle dis-tances in the aggregates decrease to less than approximately theaverage particle diameter, the color becomes blue. Although theblue color of the nanoparticle aggregates in solution is a fac-tor of 1.5 less intense than the red color of the dispersed goldnanoparticles, the blue color of the solution indicates that theinterparticles distances in the aggregate are obviously less thanthe average particle diameter which is also evident from theTEM studies.

3.3.2. Effect of pHThe evolution of particle aggregates from the monometallic

constituents is found to be dependent on the pH of the reac-tion medium. Fig. 7 shows the optical spectra of gold colloid(3 mL) with sequential additions (100 µL aliquots after eachtime step) of 0.01 M HCl solution as a preliminary controlledexperiment. There is almost no shift in the maximum peak po-


Scheme 1. Molecular structure of glutathione (GSH).

sition until the pH value is below 2.4, indicating no aggregationeffect. But below pH ∼2.4 aggregation occurs with HCl dueto surface charge neutralization of the particles and eventuallyprecipitation occurs. However, red shifted peak does not appearwith HCl (Fig. 7). It has also been noted that the generation ofaggregates with a perfect blue color is observed at pH <5 withGSH. It can be anticipated from the fact that as glutathione con-sists of the amino acids viz. glutamic acid, cysteine and glycine,so it’s binding modes are pH sensitive. The molecular struc-ture of GSH is shown in Scheme 1. In GSH, several plausiblebinding or anchoring points exist, which make different bindingmodes to be possible. The availability of one or several anchor-ing points for binding on the gold surface depends primarily onthe pH of the medium. Some of the anchoring points in glu-tathione are the two carboxylic groups in the glutamic acid andglycine residues, three –NH2 groups in the three amino acidsand the –SH group in the cysteine residue. The pKa values cor-responding to the two carboxylic groups have been determinedas 2.56 and 3.50 [55] and for the cysteine residue a pKa of9.42 has been obtained [56]. However, judging solely from thepKa values of the different groups as to which binding modecould be expected is not always possible since the acid–basecharacter of a certain group can change upon binding onto themetal surface [57]. At lower pH, linking via the α-amines is ac-tivated but linking through –COOH group is suppressed. But atintermediate or higher pH, dissociation of the carboxyl groupwould be expected to hinder the binding of gold colloids viathe α-amine group, resulting in little or no cross-linking. Fig. 8shows the optical spectra of 20 nm size gold colloids at dif-ferent pH for the GSH solution. There is clearly a progressivesuppression of new color development and a red shift of sur-face plasmon resonance band with increasing pH. At pH ∼7, nospectral changes occur but as the pH decreases then the dipoleplasmon resonance shifts to the higher wavelength. At lowerpH, one –SH group of cysteine moiety and the primary –NH2group of glutamic acid moiety can bind effectively with the goldnanoparticles to form aggregates which is authenticated fromFTIR spectra. In this regard, it is necessary to mention that be-cause of the delocalization of electrons, the other two –NH–moiety of GSH remain ineffective for binding with the gold sur-face.

3.3.3. Effect of glutathione concentrationNow, we have studied the nature of aggregation of the gold

nanoparticles with the variation in concentration of the GSH.

Fig. 8. Absorption spectra for 20 nm Au aggregates with GSH (0.1 mM) atdifferent pH.

Fig. 9. UV–vis spectra of the aggregate at different concentrations of GSH:(a) 0, (b) 10−6, (c) 10−5, (d) 10−4 and (e) 10−3 M.

With variation in reagent concentration, it is possible to shift thedipole plasmon substantially. It is now well-established in theliterature that three main parameters affect the dipole plasmon:aggregate size, shape, and individual particle size. Alterationin any of these parameters or their combination, thereof, willresult in drastic changes in peak intensity and position. Gen-erally, as gold nanoparticle aggregates grow larger, the dipoleplasmon would be red shifted. However, aggregate shape alsoplays an important role in band position. Reagent concentra-tion undoubtedly affects not only shape but also size of theaggregates. The width of the dipole plasmon, however, is sim-ply dependent on the polydispersity of the size and shape of theaggregates. Fig. 9 shows the change in the absorption spectrumof the 20 nm citrate-stabilized gold nanoparticles with increas-ing concentrations of GSH. It is observed that with increasingthe concentration of GSH, the intensity of the 526 nm peakgradually decreases and the shoulder peak at ∼750 nm dom-inates. When the concentration of GSH was 10−6 M, then onepeak with a λmax at 526 nm was observed (trace b, Fig. 9).Increased concentration of GSH (10−5 M) produced a singlehump at about 650 nm along with a peak at 526 nm (trace c,Fig. 9). But a further increase in concentration (10−3 M) ofGSH indicated a broad peak (trace e, Fig. 9). At this highconcentration (10−3 M) of GSH, the gold particles were pre-cipitated.


3.4. Stabilization of the gold aggregate

The classical Derjaguin–Laudau–Verwey–Overbeck (DLVO)theory [58] has been widely employed in colloid science tostudy particle–particle interactions, colloidal stability, coagula-tion and the behavior of the electrolyte solutions. This theoryis based on the idea that pair-wise interaction forces domi-nate which arise from the interplay of attractive van der Waalsforces, Fattrac and repulsive Coulomb forces, Frep screened byDebye–Hückel ion clouds. Obviously, the dispersed colloid isstable for Frep � Fattrac whereas the condition of Frep � Fattracleads to aggregation. Therefore, in colloidal solution severalfactors, such as, particle size, surface potential, and electricdouble layer influence the stability of nanoparticles and theiraggregation. Nanoparticles are stable in solution due to elec-trostatic repulsion of their charged surface. Lack of sufficientsurface charge or stabilizing agent will cause the particles toaggregate or precipitate. In the present experiment, citrate ionsare adsorbed onto the surface of the as-prepared gold nanopar-ticles, creating a negative surface charge that stabilizes theparticles. Before the addition of glutathione molecules, the goldparticles are colloidally stable since the energy barrier is highenough to prevent aggregation. Addition of glutathione to thecitrate-stabilized gold nanoparticles disrupts the citrate layerscausing particle aggregation via ‘place exchange reaction.’ Theglutathione molecules can decrease the energy barrier by bothlowering the surface potential of the particles and increasingthe ionic strength. Since there are no species in solution togenerate sufficient surface charge on the gold particles, the par-ticles will aggregate. Upon addition of glutathione to a solutionof colloidal gold, one functional group reacted with the goldnanoparticles while the other functional groups remained ex-tended from the nanoparticles. Thus, the multifunctional linkermolecules position the colloidal nanoparticles in close proxim-ity to each other, causing the aggregation between the particles.

3.5. Analytical application of nanoscale gold aggregates

Glutathione (GSH) is a major cellular antioxidants that playsa crucial role in maintaining the balance between oxidationand antioxidation. This tripeptide also plays an important partin cellular detoxification and is required for many aspects ofimmune response. It is also an important factor in brain dam-age [59], glucose homeostasis [60], HIV expressions [61] andcancer therapy [62]. In the present experiment, glutathione hasbeen used as a short chain molecular linker because of its bi-ological and clinical significance. Glutathione has been chosento study the effect of gold particle size on the relative opticalabsorbance change, as a basis for molecular recognition andanalytical detection. As shown in Fig. 1, the smaller is the goldparticle size, the less is the shift and the lower is the relative in-tensity of the maxima of surface plasmon resonance. An almostlinear increase in peak displacement against particle size is seenin the range of 5–20 nm, but deviating slightly for 20–55 nm(Fig. 2). Thus, the general conclusion for analytical purposes isthat larger gold particles are more sensitive to the target (GSH)molecules. For comparison, the optical response of three differ-

(a)

(b)

(c)

Fig. 10. UV–vis spectra of Au aggregates at various concentrations of GSH atpH 4 using (a) 8, (b) 20 and (c) 55 nm Au colloidal particles.

ent sizes of gold particles (8, 20 and 55 nm) to a wide rangeof GSH concentration is shown in Figs. 10a, 10b and 10c, re-spectively. On the basis of the first appearance of a clearly de-fined new peak at longer wavelength, a detection limit for GSHwith 8 nm Au particles was found to be 10−3 M, as against10−5 M for 20 nm particles and 10−6 M for 55 nm particles.The gold nanoparticles have a negative surface charge due toweakly bound citrate molecules. The negatively charged goldnanoparticles repel each other and inhibit aggregation. Sincethe 55 nm gold particles have a less negative charge due to asmaller amount of citrate adsorbed onto the surface, a crosslinkbetween the gold nanoparticles should be more facile throughthe sulfur and nitrogen atom of the GSH moiety. Therefore, theconcentration of GSH required for monolayer coverage on gold


should decrease with increasing particle size. This may resultin more free –SH and –NH2 groups available to be bound to thegold surfaces. Therefore, the larger gold particles form aggre-gate with lower concentration of glutathione molecules.

4. Conclusion

We have demonstrated a very simple approach to synthesizenanoscale gold aggregates by linking individual Au colloidalparticles with glutathione. This facile synthesis has been shownto be particularly favorable for easy manipulations via place-exchange reaction. Interparticle coupling effects on the surfaceplasmon resonance of gold particles with variable sizes in thenanometer regime have been investigated and the optical ab-sorbance behavior of the resulting nanoscale aggregates hasbeen enlightened within the framework of Maxwell-Garnett ef-fective medium theory. This experiment reveals that metal solscan be induced to aggregate by replacing the charged surfacespecies by uncharged adsorbates containing various functionalgroups. Experiments demonstrate that the surface chemistry ofcolloidal gold is dominated by electrodynamic factors relatedto its surface negative charge. Several factors, such as, ligandconcentration, pH adjustment and reaction time of the judi-cious intermixing are seen to regulate or “tune” the interactionsbetween the gold nanoparticles. It has been seen that the in-terparticle interaction of gold nanoparticles prepared by citratereduction procedure could be manipulated to control the sizeand colloidal stability of nanoparticle aggregates and the re-sults have been treated with DLVO theory. The reactivity of thegold nanoparticles assembled into well-defined architectureshas been found to be useful in excellent sensory applicationsby tuning the electrochemical characteristics and the sensitiv-ity of surface plasmon resonance of gold nanoparticles with theglutathione molecules.

Acknowledgments

Authors are thankful to IIT Kharagpur, CSIR, UGC and DSTNew Delhi for financial assistance.

References

[1] G. Schmid, Clusters and Colloids: From Theory to Applications, VCH,Weinheim, 1994.

[2] C.F. Bohren, D.R. Huffman, Absorption and Scattering of Light by SmallParticles, Wiley, New York, 1983.

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

[4] (a) N. Pradhan, A. Pal, T. Pal, Langmuir 17 (2001) 1800;(b) N.R. Jana, T. Pal, Langmuir 15 (1999) 3458;(c) N.R. Jana, T.K. Sau, T. Pal, J. Phys. Chem. B 103 (1999) 115.

[5] (a) H.D. Robert, P. Rao, J. Mater. Res. 11 (1996) 2834;(b) A. Henglein, Isr. J. Chem. 33 (1993) 77;(c) N.R. Jana, L. Gearhert, C.J. Murphy, Chem. Commun. (2001) 617.

[6] S. Link, M.A. El-Sayed, J. Phys. Chem. B 103 (1999) 8410.[7] (a) D. Falkenhagen, Artif. Organs. 19 (1995) 792;

(b) L. Palato, L.M. Benedetti, L. Callegaro, J. Drug Targeting 2 (1994) 53.[8] N.P. Gaponik, D.V. Talapin, A.L. Rogach, J. Mater. Chem. 10 (2000) 2163.[9] (a) C.B. Murray, C.R. Kagan, M.G. Bawendi, Annu. Rev. Mater. Sci. 30

(2000) 545;(b) J.H. Fendler, Chem. Mater. 13 (2001) 3196.

[10] M.P. Pileni, J. Phys. Chem. 97 (1993) 6961.[11] (a) A.M. Schwartzberg, C.D. Grant, A. Wolcott, C.E. Talley, T.R. Huser,

R. Bogomolni, J.Z. Zhang, J. Phys. Chem. B 108 (2004) 19191;(b) Z. Peng, T. Walther, K. Kleinermanns, Langmuir 21 (2005) 4249;(c) P.D. Cozzoli, E. Fanizza, M.L. Curri, D. Laub, A. Agostiano, Chem.Commun. (2005) 942;(d) J. Fendler, F.C. Meldrum, Adv. Mater. 7 (1995) 607.

[12] K. Solecka-Cermakova, B. Vlckova, J. Phys. Chem. 100 (1996) 4954.[13] (a) M. Brust, D. Bethell, D.J. Schiffrin, C. Kiely, J. Adv. Mater. 7 (1995)

795;(b) J.P. Spatz, A. Roescher, M. Moller, Adv. Mater. 8 (1996) 337;(c) N. Kimizuka, T. Kunitake, Adv. Mater. 8 (1996) 89.

[14] (a) A.S. Tse, Z. Wu, S.A. Asher, Macromolecules 28 (1995) 6533;(b) S.A. Asher, J. Holtz, L. Liu, Z. Wu, J. Am. Chem. Soc. 116 (1994)4997.

[15] (a) A. Pal, S.K. Ghosh, K. Esumi, T. Pal, Langmuir 20 (2004) 575;(b) S. Panigrahi, S. Praharaj, S. Basu, S.K. Ghosh, S. Jana, S. Pande, T. Vo-Dinh, H.J. Jiang, T. Pal, J. Phys. Chem. B 110 (2006) 13436;(c) S. Basu, S. Panigrahi, S. Praharaj, S.K. Ghosh, S. Pande, S. Jana, T. Pal,New J. Chem. 30 (2006) 1333;(d) S.Y. Chang, L. Liu, S.A. Asher, J. Am. Chem. Soc. 116 (1994) 6739.

[16] Y. Dirix, C. Bastiaansen, W. Caseri, P. Smith, Adv. Mater. 11 (1999) 223.[17] M. Quinten, A. Leitner, A. Krenn, F.R. Aussenegg, Opt. Lett. 23 (1998)

1331.[18] J.R. Krenn, A. Dereux, J.C. Weeber, E. Bourillot, Y. Lacroute, J.P. Coad-

onnet, Phys. Rev. Lett. 82 (1999) 2590.[19] J.B. Pendry, Science 285 (1999) 1687.[20] E.J. Sanchez, L. Novotny, X.S. Xie, Phys. Rev. Lett. 82 (1999) 4014.[21] B. Knoll, F. Keilmann, Nature 399 (1999) 134.[22] (a) T. Wadayama, O. Suzuki, K. Takeuchi, H. Seki, T. Tanabe, Y. Suzuki,

A. Hatta, Appl. Phys. A 69 (1999) 77;(b) S. Nie, S.R. Emory, Science 275 (1997) 1102.

[23] P.J. Tarcha, J. Desaja-Gonzalez, S. Rodriguez-Llorente, R. Aroca, Appl.Spectrosc. 53 (1993) 43.

[24] J.J. Storhoff, R. Elghanian, R.C. Mucic, C.A. Mirkin, R.L. Letsinger,J. Am. Chem. Soc. 120 (1998) 1959.

[25] R. Elghanian, J.J. Storhoff, R.C. Mucic, R.L. Letsinger, C.A. Mirkin, Sci-ence 277 (1997) 1078.

[26] C.A. Mirkin, M.A. Ratner, Annu. Rev. Phys. Chem. 101 (1997) 1593.[27] M.A. Rampi, O.J.A. Schueller, G.M. Whitesides, Appl. Phys. Lett. 72

(1998) 1781.[28] D.L. Klein, R. Roth, A.K.L. Lim, A.P. Alivisatos, P.L. McEuen, Na-

ture 389 (1997) 699.[29] L.A. Bumm, J.J. Arnold, M.T. Cygan, T.D. Dunbar, T.P. Burgin, L. Jones,

D.L. Allara, J.M. Tour, P.S. Weiss, Science 271 (1996) 1705.[30] S.K. Ghosh, S. Kundu, M. Mandal, S. Nath, T. Pal, J. Nanopart. Res. 5

(2003) 577.[31] J.C. Maxwell-Garnett, Philos. Trans. R. Soc. 203 (1904) 805.[32] S.W. Chen, R.J. Pei, J. Am. Chem. Soc. 123 (2001) 10607.[33] A. Henglein, J. Phys. Chem. B 104 (2000) 6683.[34] A. Henglein, M. Giersig, J. Phys. Chem. B 104 (2000) 5056.[35] J.I. Park, J. Cheon, J. Am. Chem. Soc. 123 (2001) 5743.[36] S. Park, P.X. Yang, P. Corredor, M.J. Weaver, J. Am. Chem. Soc. 124

(2002) 2428.[37] Z.L. Wang, J. Phys. Chem. B 104 (2000) 1153.[38] J.T. Krug, G.D. Wang, S.R. Emory, S.M. Nie, J. Am. Chem. Soc. 121

(1999) 9208.[39] A.C. Templeton, M.J. Hostetler, E.K. Warmoth, S.W. Chen, C.M.

Hartshorn, V.M. Krishnamurthy, M.D.E. Forbes, R.W. Murray, J. Am.Chem. Soc. 120 (1998) 4845.

[40] M.P. Pileni, A. Taleb, C.J. Petit, Dispersion Sci. Technol. 19 (1998) 185.[41] M.A. Hayat, Colloidal Gold: Principles, Methods, and Applications, Aca-

demic Press, San Diego, CA, 1989.[42] S. Mann, M.S. Li, S. Connolly, D. Fitzmaurice, Adv. Mater. 12 (2000)

147.[43] M. Mrksich, Chem. Soc. Rev. 29 (2000) 267.[44] R.M. Bright, D.G. Waletr, M.D. Musick, M.A. Jackson, K.J. Allison, M.J.

Natan, Langmuir 12 (1996) 810.


[45] M.M. Alvarez, J.T. Khoury, T.G. Schaaff, M.N. Shafigullin, I. Vezmar,R.L. Whetten, J. Phys. Chem. B 101 (1997) 3706.

[46] S. Wang, H. Yao, S. Sato, K. Kimura, J. Am. Chem. Soc. 126 (2004) 7438.[47] A.P. Alivisatos, K.P. Johnsson, X. Peng, T.E. Wilson, C.J. Loweth, M.P.

Bruchez jr, P.G. Schultz, Nature 382 (1996) 609.[48] Z. Zhong, S. Patskovskyy, P. Bouvrette, J.H.T. Luong, A. Gedanken,

J. Phys. Chem. B 108 (2004) 4046.[49] S. Mandal, A. Gole, N. Lala, R. Gonnade, V. Ganvir, M. Sastry, Lang-

muir 17 (2001) 6262.[50] G. Frens, Nature 20 (1973) 241.[51] G. Mie, Ann. Phys. 25 (1908) 377.[52] J. Belloni, M. Mostafavi, H. Remita, J.L. Marignier, M.O. Delcourt, New

J. Chem. 22 (1998) 1239.[53] A.C. Templeton, S. Chen, S.M. Gross, R.W. Murray, Langmuir 15 (1999)

66.

[54] T. Jensen, L. Kelly, A. Lazarides, G.C. Schatz, J. Cluster Sci. 10 (1999)295.

[55] K. Takehara, Y. Ide, Bioelectrochem. Bioenerg. 29 (1992) 103.[56] S.G. Tajc, B.S. Tolbert, R. Basavappa, B.L. Miller, J. Am. Chem. Soc. 126

(2004) 10508.[57] S.B. Lee, K. Kim, M.S. Kim, J. Raman Spectrosc. 22 (1991) 811.[58] J. Isrealachvili, Intermolecular and Surface Forces, second ed., Academic

Press, San Diego, CA, 1992.[59] S. Tsakiris, K.H. Schulpis, K. Marinou, P. Behrakis, Pharmacol. Res. 49

(2004) 475.[60] M. Barbagallo, L.J. Dominguez, M.R. Tagliamonte, L.M. Resnick, G. Pa-

olisso, Hypertension 34 (1999) 1002.[61] T. Yamaguchi, I. Katoh, S. Kurata, Eur. J. Biochem. 269 (2002) 2782.[62] G.K. Balendiran, R. Dabur, D. Fraser, Cell Biochem. Funct. 22 (2004)

343.

Biomolecule induced nanoparticle aggregation: Effect of particle size on interparticle coupling

Documents

Transcript of Biomolecule induced nanoparticle aggregation: Effect of particle size on interparticle coupling