A

eawogabt1©

K

1

msb(toabotslpata

NT

0d

Materials Science and Engineering B 140 (2007) 172–176

Controlling the aggregation behavior of gold nanoparticles

Yong Yang a,b,∗, Shigemasa Matsubara a, Masayuki Nogami a, Jianlin Shi b

a Department of Materials Science and Engineering, Nagoya Institute of Technology, Showa, Nagoya 466-8555, Japanb School of Materials Science and Engineering, East China University of Science and Technology, Shanghai 200237, PR China

bstract

The aggregation of Au nanoparticles (NPs) in solution is influenced by cationic and oligocationic species. The polarization of the conductionlectron oscillations in adjacent gold nanoparticles causes a new red-shifted plasmon absorbance attributed to the coupling of the plasmonbsorbance of the particles. This appearance of an additional plasmon band is of particular interest to the field of SERS and has led to researchorks directed at the stabilization of small colloid aggregates in solution. The surface plasmon coupling can be tuned by controlling the aggregationf gold nanoparticles by the addition of some “cross-linking” agent. Here we develop a simple method to fabricate linear-chainlike aggregates ofold nanoparticles (so-called nanochains), tuning the linear optical properties in a wide wavelength range from visible to the near-infrared. Theggregation behavior and linear self-assembly mechanism of citrate-stabilized gold colloids as provoked by the addition of cetyltrimethylammonium

romide (CTAB) and 11-mercaptoundecanoic acid (MUA) are also analyzed. The line-assembly mechanism of gold nanochain is attributed tohe preferential binding of CTAB molecules on a certain facet of gold NPs and the Au NP electrostatic interactions. We also found that the1-mercaptoundecanoic acid was effective to prevent the further aggregation of CTAB-modified gold colloids. 2007 Elsevier B.V. All rights reserved.

scnHatveamaacteg

eywords: Gold nanoparticles; Self-assembly; Optical properties

. Introduction

Nanoparticles assemblies are of particular interest for funda-ental research and applications for surface enhanced Raman

pectroscopy (SERS), second harmonic generation (SHG) andio-photonics [1]. Especially, 1D assembly of nanoparticlesNPs) by controlling the aggregation of nanoparticles is likelyo play important roles in the improvement of the efficienciesf various electronic, optical, magnetic and other devices [2–4],nd can significantly help in the understanding of a number ofiological processes and fundamental aggregation mechanismf nano-system [5]. Generally, the aggregation of Au nanopar-icles in solution is influenced by cationic and oligocationicpecies [6]. The polarization of the conduction electron oscil-ations in adjacent gold nanoparticles causes a new red-shiftedlasmon absorbance attributed to the coupling of the plasmon

bsorbance of the particles [7]. This appearance of an addi-ional plasmon band is of particular interest to the field of SERSnd has led to research works directed at the stabilization of

∗ Corresponding author at: Department of Materials Science and Engineering,agoya Institute of Technology, Showa, Nagoya 466-8555, Japan.el.: +81 52 735 5285; fax: +81 52 735 5285.

E-mail address: yang.yong@nitech.ac.jp (Y. Yang).

2

pbwA

921-5107/$ – see front matter © 2007 Elsevier B.V. All rights reserved.oi:10.1016/j.mseb.2007.03.021

mall colloid aggregates in solution [8]. The surface plasmonoupling can be tuned by controlling the aggregation of goldanoparticles by the addition of some “cross-linking” agent.ere we develop a simple method to fabricate linear-chainlike

ggregates of gold nanoparticles (so-called nanochains), tuninghe linear optical properties in a wide wavelength range fromisible to the near-infrared. The aggregation behavior and lin-ar self-assembly mechanism of citrate-stabilized gold colloidss provoked by the addition of cetyltrimethylammonium bro-ide (CTAB) and 11-mercaptoundecanoic acid (MUA) are also

nalyzed. The line-assembly mechanism of gold nanochain isttributed to the preferential binding of CTAB molecules on aertain facet of gold NPs and the Au NP electrostatic interac-ions. We also found that the 11-mercaptoundecanoic acid wasffective to prevent the further aggregation of CTAB-modifiedold colloids.

. Experimental

Hydrogen tetrachloroaurate (HAuCl4·3H2O, 99.99%) was

urchased from Tokyo chemical Co. Cetyltrimethylammoniumromide (CTAB, 99%) and 11-mercaptoundecanoic (MUA)ere obtained from Sigma. All other reagents were fromldrich and were used as received. Ultrapure deionized water

nd En

(t

msstrwra0pwpMbp

rtTgmscTt

3

3co

iotescTeimtwTcCitttga

aestbttpunrttttm

(gt07twWwcbnt

tc(ecopedtoTatprFmCnni

Y. Yang et al. / Materials Science a

Narnstead Nanopure H2O purification system) was usedhroughout the experiments.

Gold colloids were prepared by citrate thermal reductionethod [9]. Typically in the process of thermal reduction, a gold

ol was prepared by adding 1 ml of 1 wt% HAuCl4 aqueousolution and 1.5 ml of 38.8 mM sodium citrate aqueous solu-ion into 90 ml boiling water. The citrate ion acted as both aeductant and stabilizer. After the solution had turned purple redithin 30 s, the solution was cooled quickly in the ice bath. This

esulted in a stable dispersion of gold particles with an aver-ge diameter of around 13.2 nm and 10% polydispersity [10]..2 ml of 0.01 mM, 0.1 mM, 1 mM, 0.01 M and 0.1 M freshlyrepared cetyltrimethylammonium bromide aqueous solutionere added into 20 ml as prepared gold colloid at room tem-erature for different time, respectively. Finally, 1 ml of 0.5 mMUA aqueous solution was added in the gold colloid modified

y 0.1 mM CTAB in order to restrain the overmuch aggregationrocess.

The absorption optical spectra of these gold colloids wereecorded using Jasco Ubest 570 UV–vis–NIR spectrophotome-er. All the spectra were recorded in air at room temperature.he microstructure and morphology of gold nanoparticles inold colloids was measured with a JEOL JEM-2000EXII trans-ission electron microscopy (TEM) operating at 200 kV. Those

amples were prepared by dropping the colloid onto a carbon-oated Cu grid underlying tissue paper, leaving behind a film.he ζ-potential was measured with Otsuka ELS-6000KS sys-

em.

. Results and discussion

.1. Aggregation behavior and optical properties ofitrate-stabilized gold colloids as provoked by the additionf CTAB

In the present work, we investigated the aggregation behav-or of citrate-stabilized gold colloids as provoked by the additionf CTAB. The prepared Au NP colloids were analyzed using aransmission electron microscopy (TEM) and their optical prop-rties were studied by absorbance spectrophotometer. Fig. 1hows the optical absorption spectra of citrate-stabilized goldolloids before and after addition of different volume of CTAB.he gold colloids without CTAB and with 0.01 mM CTABxhibit one narrow absorbance band at 520 nm (Fig. 1a), whichs attributed to the surface plasmon resonance (SPR) band of

onodisperse Au NP. The dipole plasmon resonance leads tohis absorption band for gold NPs [11]. Monodisperse Au NPsith a diameter of 13 nm were observed in this solution byEM and the ζ-potential was evaluated to be −27.76 mV. Inontrast, we found that gold colloids prepared with 0.1 mMTAB formed linear-chainlike aggregated Au NPs, as shown

n the inset of Fig. 1b. When 0.1 mM CTAB was added inhe citrate-stabilized gold colloids solution, first we observed

he formation of a new absorbance band near 585 nm, besideshe absorption band at 520 nm (Fig. 1b), which indicated theseold NPs begun to form linear-chainlike aggregates. The secondbsorbance band is attributed to the SPR coupling band of linear-

tico

gineering B 140 (2007) 172–176 173

ggregated gold NPs [12]. The polarization of the conductionlectron oscillations in adjacent gold NPs causes a new red-hifted plasmon absorbance band attributed to the coupling ofhe plasmon absorbance of the particles. The second absorbanceand red-shifted to 660 nm after 60 min, and the intensity relatedo the band at 520 nm slowly increased with increasing reactionime. These changes were thought to be that more monodis-erse gold NPs self-assembled into linear-aggregated gold NPnits. While 1 mM CTAB aqueous solution was added, the goldanoparticles aggregated into very big particles and precipitatedapidly from the solution. No absorbance band could be found inheir absorbance spectra shown in Fig. 1c, which also indicatedhe majority of gold nanoparticles precipitated from the solu-ion. The ζ-potential changed to be 10.46 mV, and it indicatedhat the citrate ions were completely counteracted by CTAB

olecules.Interestingly, the second SPR band appeared again at 740 nm

Fig. 1d) when 0.2 ml of 0.01 M CTAB (0.1 mM) was added intoold colloids. Their TEM image shows that nanochains of two,hree and four Au NPs are obtained in this gold colloid. While.3 ml of 0.01 M CTAB was added, the second SPR band at40 nm blue-shifted to 680 nm, as shown in Fig. 1e. It indicatedhe particle number in every aggregated cluster decreased, whichas proved by their TEM image shown in the inset of Fig. 1e.hat is more, only one SPR band at 520 nm (Fig. 1f) was foundhen 0.1 M CTAB was added. It indicated that the surface citrate

apping layer of gold nanoparticles were completely displacedy CTAB bilayers, and these absorbed CTAB bilayers formedew protecting layers of gold nanoparticles, which could makehese gold nanoparticles remain monodisperse again [4].

The new absorption band at longer wavelength observed inhe present system is explained on the basis of interplasmonoupling effects [13]. The localized surface plasmon resonanceLSPR) is a collective oscillation of the nanoparticle conductionlectrons. The oscillation can be localized on a single nanoparti-le, or it may involve many coupled nanoparticles. The electronscillation frequency depends on several parameters, includingarticle shape and size, and nanoparticles arrangement. Thestablished electric field alignment description of the point-ipole model [4,14] for a metal chain suggests the presence ofwo SPR modes: a longitudinal and a transverse plasmon res-nance along and perpendicular to the chain-axis, respectively.he transverse modes are located around the SPR position ofsingle-particle dipole mode (520 nm), and the positions of

he longitudinal modes depend on the number N of aggregatedarticles. Then the aggregation extent of gold colloids can beepresented by the position of their longitudinal SPR modes.ig. 2 shows that the peak position changes of longitudinal SPRodes with the different quantity of CTAB. Hence, it is clear thatTAB concentration plays an important role in determining theature of the aggregation. A very low concentration of CTAB hasot enough influence on gold nanoparticles capped with citrateons and these gold nanoparticles remain monodisperse. With

he concentration of CTAB increases, the aggregation extentncreases and the chain-like aggregates of gold nanoparticlesan form by the “glue” function of CTAB [15]. When 0.2 mlf 1 mM CTAB was added, the gold nanoparticles aggregated

174 Y. Yang et al. / Materials Science and Engineering B 140 (2007) 172–176

F olar o1 ter the

ittζ

tcbtatheaeT

ic

3n

gota

ig. 1. UV–vis–NIR absorption spectra of gold colloids modified by different m0 mM, (e) 0.3 ml, 10 mM, (f) 0.2 ml, 100 mM; TEM images of gold colloids af

nto very big particles which could be observed by eyes, dueo the cross-linking effect of more CTAB molecules [4]. Whilehe concentration of CTAB continually increased, the value of-potential changed to be above 0 mV, which indicated the quan-ity of cation is more than that of anion. The majority of surfaceitrate capping layer of gold nanoparticles would be displacedy CTAB bilayers, and it would lead to the anisotropic distribu-ion of the residual surface charge of the residual surface chargend this extrinsic electric dipole formation is responsible forhe linear organization of the gold nanoparticle [8]. In a veryigh concentration of CTAB, the surface citrate capping lay-

rs of Au NPs are completely displaced by CTAB molecules,nd these absorbed CTAB molecules form new protecting lay-rs of Au NPs, thus these Au NPs keep monodisperse again.hus, our simple method using different CTAB concentrations

siaa

f CTAB: (a) 0.2 ml, 0.01 mM, (b) 0.2 ml, 0.1 mM, (c) 0.2 ml, 1 mM, (d) 0.2 ml,addition of different quantity of CTAB for 60 min are shown in related insets.

s appropriate for controlling the aggregation extent of goldolloids.

.2. Reduce the excessive aggregation of goldanoparticles by MUA

When 0.1 mM CTAB was added into the citrate-stabilizedold colloid, firstly two SP bands around 520 and 600 nm werebserved, as shown in Fig. 3a. With the aging time increasedo 75 min, three SPR bands around 520, 590 and 692 nmppeared, which indicated that more monodisperse gold NPs

elf-assembled into linear-aggregated gold NP units. Interest-ngly, other SPR bands except the SPR band at 520 nm broadenednd red-shifted to the near-infrared region above 1100 nm afterging for 150 min. It indicated that the CTAB-modified gold col-

Y. Yang et al. / Materials Science and En

Fc

lw

M5

FaC

srpebrh(vgac[ttc

4

ig. 2. The ζ-potential and peak position of the longitudinal SPR mode for goldolloids modified by different quantity of CTAB.

oid was not stable and the aggregation extent would increase

ith the increased aging time.However, we found that this solution was very stable after

UA was added into this solution, and two SPR bands at 520 and90 nm did not show any change until the aging time of 96 h, as

ig. 3. UV–vis–NIR absorption spectra of gold colloid at different time intervalsfter (a) addition of 0.2 ml of 0.1 mM CTAB, (b) addition of 0.2 ml of 0.1 mMTAB and 1 ml of 0.5 mM MUA.

oClwaaAuisgwo

A

(SpC

R

gineering B 140 (2007) 172–176 175

hown in Fig. 3b. It indicated that the excessive aggregation waseduced by MUA. As reported by literature, the CTAB surfactantreferentially binds to the {1 0 0} facets rather than the {1 1 1}nd facets of gold NPs, and the thiol derivatives preferentiallyind to the {1 1 1} facets of gold NPs [7,12]. Su et al. [16]eported that 3D close-packed gold aggregates could form by theydrogen-bonding interactions of MUA under acidic conditionspH 3). In our solution, the potential was −27.76 mV and pHalue was greater than 7, the thiol derivates would bind to theold nanoparticles by thiol agent and the carboxylate-terminatedgent could interact with the polar head of one CTAB bilayer, andould lead to aggregates formation via electrostatic interactions12]. While the cross-linking function of CTAB as the “glue”o link adjacent gold nanoparticles was thus diminished. Thenhe excessive aggregation in this CTAB-modified gold colloidould be prevented by the addition of MUA.

. Conclusions

In summary, the aggregation behavior and linear organizationf citrate-stabilized gold colloids as provoked by the addition ofTAB has been studied. The concentration of CTAB in gold col-

oids is critical for self-assembling linear-chainlike aggregatesith different interconnecting particle number and network-like

ggregates. The CTAB with appropriate concentration servess the “glue” that can link the {1 0 0} facets of two adjacentu NPs, which leads to an anisotropic distribution of the resid-al surface charge; and this extrinsic electric dipole formations responsible for the linear organization of the gold NPs intohort chains. The excessive aggregation in this CTAB-modifiedold colloid could be prevented by the addition of MUA,hich could diminish the overmuch cross-linking functionf CTAB.

cknowledgements

Financial support and granting of the postdoctoral fellowshipP04416) of this work from the Japan Society for Promotion ofcience is gratefully acknowledged. The work was also sup-orted by the NITECH 21st Century COE Program “Worlderamics Center for Environmental Harmony”.

eferences

[1] (a) W.L. Barnes, A. Dereux, T.W. Ebbesen, Nature 424 (2003) 824;(b) Y. Yang, L. Xiong, J. Shi, M. Nogami, Nanotechnology 17 (2006) 2670.

[2] U. Kreibig, M. Vollmer, Optical Properties of Metal Cluster, Springer,Berlin, 1995.

[3] E. Prodan, C. Radloff, N.J. Halas, P. Nordlander, Science 302 (2003) 419.[4] (a) Y. Yang, M. Nogami, J. Shi, H. Chen, G. Ma, S. Tang, Appl. Phys. Lett.

88 (2006) 081110;(b) Y. Yang, S. Matsubara, M. Nogami, J. Shi, W. Huang, Nanotechnology17 (2006) 2821.

[5] Z.Y. Tang, N.A. Kotov, Adv. Mater. 17 (2005) 951.

[6] A.N. Shipway, M. Lahav, R. Gabai, I. Willner, Langmuir 16 (2000) 8789.[7] K.G. Thomas, S. Barazzouk, B.I. Ipe, S.T.S. Joseph, P.V. Kamat, J. Phys.

Chem. B 108 (2004) 13066.[8] (a) G. Freeman, K.C. Garbar, et al., Science 267 (1995) 1629;

(b) K. Kneipp, H. Kneipp, J. Kneipp, Acc. Chem. Res. 39 (2006) 443;

1 and E

[

[[

76 Y. Yang et al. / Materials Science

(c) S. Lin, M. Li, E. Dujardin, G. Christian, S. Mann, Adv. Mater. 17 (2005)2553.

[9] (a) Y. Yang, M. Nogami, J. Shi, H. Chen, G. Ma, S. Tang, J. Phys. Chem.B 109 (2005) 4865;

(b) Y. Yang, M. Nogami, J. Shi, H. Chen, Y. Liu, S. Qian, J. Mater. Chem.13 (2003) 3026.

10] (a) Y. Yang, M. Hori, T. Hayakawa, M. Nogami, Surf. Sci. 579 (2005)215;(b) Y. Yang, J. Shi, H. Chen, S. Dai, Y. Liu, Chem. Phys. Lett. 370 (2003) 1.

[[[[

ngineering B 140 (2007) 172–176

11] Y.N. Xia, N.J. Halas, MRS Bull. 30 (2005) 338.12] (a) T. Ung, L.M. Liz-Marzan, P. Mulvaney, J. Phys. Chem. B 105 (2001)

3441;(b) C.J. Murphy, T.K. Sau, A.M. Gole, C.J. Orendorff, J. Gao, L. Gou, S.E.

Hunyadi, T. Li, J. Phys. Chem. B 109 (2005) 13857.

13] L.M. Liz-Marzan, P. Mulvaney, J. Phys. Chem. B 107 (2003) 7312.14] S.A. Maier, P.G. Kik, H.A. Atwater, Phys. Rev. B 67 (2003) 205402.15] T.K. Sau, C.J. Murphy, Langmuir 21 (2005) 2923.16] C. Shu, P. Wu, C. Yeh, Bull. Chem. Soc. Jpn. 77 (2004) 189.