Probing real time gold nanostar formation process using two-photon scattering spectroscopy

Chemical Physics Letters 504 (2011) 46–51

Contents lists available at ScienceDirect

Chemical Physics Letters

journal homepage: www.elsevier .com/locate /cplet t

Probing real time gold nanostar formation process using two-photonscattering spectroscopy

Dulal Senapati, Anant K. Singh, Sadia A. Khan, Tapas Senapati, Paresh Chandra Ray ⇑Department of Chemistry, Jackson State University, Jackson, MS, USA

a r t i c l e i n f o

Article history:Received 27 October 2010In final form 14 January 2011Available online 19 January 2011

0009-2614/$ - see front matter � 2011 Elsevier B.V. Adoi:10.1016/j.cplett.2011.01.046

⇑ Corresponding author. Fax: +1 601 979 3674.E-mail address: [email protected] (P.C. Ray)

a b s t r a c t

We report the growth mechanisms for the formation of 40 nm star shape gold nanocrystals using realtime TEM and in situ two-photon scattering technique. The overall process consists of several interme-diate steps and these are: the nucleation process for the formation and accumulation of nanoseeds, Ost-wald ripening process for the formation of nanoflower and intraparticle ripening process for theformation of the star sharp nanoparticles via nanocrowns. We demonstrates that the real time shape evo-lution of intermediate colloidal nanoparticles and their number density change can be fully accessed dur-ing the synthesis of star shaped gold nanoparticles using time dependent in situ TPS experiment.

� 2011 Elsevier B.V. All rights reserved.

1. Introduction

Establishing control over the size and shape of nanoparticles isone of the most important goals of materials science, which re-quires a detailed understanding of the mechanism of particlegrowth [1–5]. Due to the intense electromagnetic field locatedaround the pods, star shape gold nanocrystals are highly SERS ac-tive, which makes these particles very attractive for various reallife applications [6–10]. Recently we and other groups have shownthat, due to the strong absorption in near IR, asymmetric nanopar-ticles are potential candidates for novel cancer therapeutic agentsthrough photo-thermal mechanism [11–15]. In spite of the goodprogress over the past decade in the synthesis of gold nanomateri-als of various sizes and shapes, an important challenge nanomate-rials are facing in technological applications are due to the fact thatthe mechanisms underpinning these synthetic methods haveevolved empirically and there is no accepted coherent mechanisticexplanation to explain how shape control works [1–6]. The de-tailed understanding of the design and control parameters shouldhave a profound impact on the engineering of the growth of nano-particles to the desired size and shape, which is necessary forexploration of gold nanoparticles in a wide range of technologicalapplications [6–18]. Usually nucleation and growth processes oftenoccur on a relatively short time scales, as a result in situ time-resolved measurements are helpful to understand the growthmechanism. Driven by the need, in this Letter we report the growthmechanisms for the formation of star shape gold nanocrystalsusing real time TEM and in situ two-photon scattering technique.

In noble metal gold nanoparticles, the coherent collective oscil-lation of electrons in the conduction band induces large surface

ll rights reserved.

.

electric fields, which greatly enhances the radiative properties ofthese nanoparticles [6–19]. As a result, the light scatteringcross-section of metal nanoparticles are several orders of magni-tude more intense than that of organic dyes, which makes it asan excellent candidate for novel contrast agents for optical detec-tion [3–19]. As the particle size increases and interparticle distancedecreases, coherent photoinduced nonlinear optical effect whichshows a marked dependence on the interparticle distance and den-sity can play a very important role.[9] Hyper-Rayleigh scattering(HRS) [15,20–28] or two-photon scattering (TPS), is a nonlinearoptical effect observed in isotropic solutions due to the fluctuationsin symmetry, caused by rotational fluctuations, where scatteringby a fundamental laser beam can be detected at the second-har-monic wavelength [15,20–28]. In the last ten years several publica-tions have shown that [15,20–28] two-photon scattering (TPS)properties can be greatly enhanced (�104–106) for molecules ona roughened metal surface. Also, like other scattering techniques,TPS has been known to be highly sensitive toward small changesin size, structure, or shape [15,21–28]. As a result, TPS can be afar superior technique for monitoring nanoparticle size evaluationin liquids. In this Letter, for the first time we have demonstratedthat in situ two-photon scattering spectroscopy [15,20–28] canbe used for real time monitoring of intermediate colloidal nanopar-ticles and their number density change during the synthesis of starshaped gold nanoparticles.

2. Experimental methods

Hydrogen tetrachloroaurate (HAuCl4�3H2O), NaBH4, silver ni-trate, cetyltrimethylammonium bromide (CTAB), glutaraldehydeand sodium citrate were purchased from Sigma–Aldrich and usedwithout further purification.

D. Senapati et al. / Chemical Physics Letters 504 (2011) 46–51 47

2.1. Synthesis of gold spherical seeds

Gold spherical seeds were synthesized by mixing aqueous solu-tions of hydrogen tetrachloroaurate (III) hydrate with trisodiumcitrate in 20 ml double distilled deionized water (18 MX) wherethe final concentration of HAuCl4.3H2O was at 2.5 � 10�4M andthe concentration of tri-sodium citrate was 10�4M. An ice cooled,freshly prepared aqueous solution of sodium borohydride, NaBH4

(0.1 M, 60 lL) was then added under vigorous stirring. Solutionturned pink immediately after the addition of NaBH4, which becamered after keeping the solution in dark for overnight. Nano-seedsexhibit absorption spectra with a maximum at 510 nm, which corre-sponds to 4.3 nm seed, which has been confirmed by TEM.

2.2. Synthesis of gold nano star

Our star shape gold nanomaterial synthesis was achievedthrough a two-step process, using seed-mediated growth, as re-ported before [3,12] . In the first step, very small, reasonably uni-form, spherical seed particles are generated using tri-sodiumcitrate as stabilizer and sodium borohydride as strong nucleatingagent, as we discussed before. In the second step, we have usedascorbic acid as weak reductant and CTAB as shape templating sur-factant so that the seeds can grow into larger particles of particularmorphology we desired. We have also added AgNO3 to CTAB inaqueous solution, for immediate formation of AgBr, which can playan important role in the shape control mechanism. It has been sug-gested that underpotential deposition of metallic silver occurs onthe crystal facets of gold seed, leading to symmetry-breaking andmulti-shape branched gold nanoparticle formation [3–4]. Theascorbic anions transfer electrons to the seed particles, which willreduce gold ions to form gold shell, which grows into star shapenanoparticles in the presence of CTAB. For nanostar preparation,we have dissolved 0.05 gm CTAB in 46.88 ml H2O by sonicationin a small vial and then we added 2 ml 0.01 M HAuCl4.3H2O underconstant stirring. Next, 0.3 ml 0.01 M AgNO3 was added to thesolution and mixed properly. After that, we added 0.32 ml 0.1 Mascorbic acid drop wise as a reducing agent. The solution turnedcolorless from yellow. To this colorless solution, we instantlyadded 0.5 ml gold nano-seed at a time and mixed the solution. Col-or changed immediately and become blue within 2 min, whichindicates the formation of star shape nanostructures. This starshape gold nanoparticle has only one plasmon band like sphericalgold nanoparticle, but their kmax shifted about 60 nm in compari-son to the spherical gold nanoparticle.

2.3. Real time nano-structure monitoring

At certain time interval, one drop of mixture sample was placedon a mesh for TEM experiment. Sample was dried immediatelyusing blow air. TEM experiment was performed using JEM-2100Fadvanced field emission electron microscope, operating at 100–200 kV.

2.4. Hyper rayleigh scattering spectrosocpy

We have monitored the two photon scattering intensity usingHyper Rayleigh scattering (HRS) technique [15,20–28]. HRS is asecond harmonic generation experiment in which the light is scat-tered in all directions rather than as a narrow coherent beam. De-tails of experimental set up has been described before [15,21–22,24]. We performed TEM data before and after exposure of about5 min to the laser and we have not noted any photo-thermaldamage of nanostructures within our HRS data collecting time.The HRS light was separated from its linear counterpart by a3 nm bandwidth interference filter and a monochromator and then

detected with a cooled photomultiplier tube, and the pulses werecounted with a photon counter. The fundamental input beamwas linearly polarized, and the input angle of polarization was se-lected with a rotating half-wave plate. Since horizontally polarizedsecond harmonic output light is known [15,20–28] to be indepen-dent of the polarization of the excitation, for HRS measurement,the polarization state of the harmonic light was horizontal. Forall the polarization experiments, the polarization state of the har-monic light was vertical. Since gold nanostructure, do not exhibitany absorption band at 410 nm, we can easily avoid the two-pho-ton fluorescence contribution in our experiment.

3. Results and discussions

We have performed our star shape gold nanomaterial synthesisusing two-step process, as we discussed before. In absence of CTAB,just by using trisodium citrate as surfactant and NaBH4 as strongreducing agent, a fast nucleation generates very small sized(4.3 nm) spherical gold nanoparticles [34] from HAuCl4�3H2O. Thisis the step one. Depending on the ratio of Au:Tri-sodium citrate, wecan control the size of the seed. In the second step we used thesesmall seeds as nucleation centre to generate larger particles of par-ticular morphology we desired. So, in that sense gold seeds startthe second nucleation step. This is true that like the nanostar for-mation, a nano seed also goes through Ostwald ripening process[34] before it reaches to uniform �4.3 nm nanoparticles which isnanoseed in our case.

To monitor the shape evaluation, we have performed real timeTEM image during the synthesis procedure. TEM experiment wasperformed using JEM-2100F advanced field emission electronmicroscope, operating at 200 kV. TEM images obtained from ali-quots taken at different time intervals demonstrate that initiallygold nanoseeds grew like nano-flower configuration from whichsubsequently crown shape and at last star shape structure evolved(as shown in Figure 1).

Our result shows clearly that in the early stage, it is likely thatOstwald ripening [29–31] of nanoseeds is operative here for theformation of nanoflower. As shown in Figure 1, that within 40 s,adjacent particles orient so as to share a common orientationand the formation of a larger particle results, which is nanoflower.And then as CTAB starts controlling the growth kinetics, flowerstructure breaks and starts to form stars shape nanoparticle viacrown shape. So, our experiment clearly shows that Ostwald ripen-ing is prevented by the presence of CTAB surfactant. Now withtime, when more CTAB is available, intraparticle ripening[3,6,30–31] phenomenon controls the shape evolution. As a result,flower shape bigger nanostructure breaks and forms crown shapestructure and which evolves into star shape nanostructure, asshown in Figure 1. When the crystals reach their equilibrium shaperequired by the Gibbs�Curie�Wulff law [3,30–31], the intraparti-cle ripening should stop and as a result, we did not see any shapeevolution from 90 s to 30 min time. To understand how CTAB con-trols the reaction kinetics, we have also measured the kinetics ofstar shape gold nanostructure formation using time dependentabsorption spectroscopy. As shown in Figure 2, after the additionof CTAB, within 40 s, kmax shifted to red by about 80 nm, which isdue to the formation of nanoflower. Since nanoflower formeddue to the Ostwald ripening of nanoseeds, nanoflower is much big-ger size than gold nanoseeds and as a result, one can expect redshift of plasmon band. And then, after 40 s, since nanoflowerbreaks and forms nanocrown due to intraparticle ripening, kmax

for plasmon band shows blue shift of about 20 nm for nanocrowns.(see Figures 3 and 4)

Next 40 s, kmax shows very slight blue shift and it is due to thesimilar size of nanocrown and nanostar. After that, kmax remains

Figure 1. TEM images and photographs demonstrating real time shape evolution during gold nanostar formation in the presence of 0.003 M CTAB.

400 500 600 700 800 9000.0

0.5

1.0

1.5

2.0

2.5

Nano Star

Nano Crown

Nano Flower

Nano Seed

Abs

orba

nce

Excitation Wavelength (nm)

Figure 2. Demonstrating time dependent absorption spectra at 40 s time intervaland showing how the absorption profile change with time during the formation ofstar shape nanoparticle from nanoseeds in the presence of CTAB.

48 D. Senapati et al. / Chemical Physics Letters 504 (2011) 46–51

same even after 30 min, which indicates that nanoparticlesreached their equilibrium shape. Though TEM and absorption spec-tra clearly demonstrate the reaction pathway for the formation ofstar shape gold nanoparticle from gold nanoseeds via the path,nanoseeds ? nanoflowers ? nanocrowns ? nanostars, we willnot be able to find the time dependent concentration for interme-diates or final product during the shape evaluation using our TEMor absorption data. As a result, we employed two-photon scatter-ing spectroscopy to monitor the real time concentration of eachshapes during star shape nanoparticle formation from nanoseedsin the presence of CTAB.

We have measured the two-photon scattering intensity usinghyper-Rayleigh scattering (HRS) technique [21–28]. HRS techniqueis a second harmonic generation experiment in which the light isscattered in all directions rather than as a narrow coherent beam.The intensity ITPS of the two-photon Rayleigh scattering signal froman aqueous solution of gold nanoparticle can be expressed as[15,20–28],

IHRS ¼ GhNWb2W þ N2

nanoiI2xe½�Nnano1ð�þ2�2xþ2�xÞ� ð1Þ

where G is a geometric factor which include call shape, scatteringfactor and experimental correction terms, Nw and Nnano the numberof water molecules and gold nanoparticle per unit volume, bx andbnano are the quadratic hyperpolarizabilities of a single water mole-cule and a single nanoparticle, ex and e2x are the molar extinctioncoefficient of the gold nanoparticle at x and 2x respectively, l is thepath length and Ix the fundamental intensity. The exponential fac-tor accounts for the losses through absorption at the harmonic andfundamental frequencies. The brackets imply orientation averaging.The TPS signal intensities were normalized to the square of the inci-dent light intensity. To understand whether two-photon scatteringintensity is due to second harmonic generation, we performedpower dependent as well as concentration dependent study. Ourpower dependent study shows a linear nature plot between I2x

and Ix [2], which implies that the doubled light is indeed due tothe TPS signal. To extract absolute values for the hyperpolarizabili-ties, the TPS intensities were normalized again with para-nitro ani-line (pNA) in methanol as we have discussed before [15,20–28].

We and other groups have demonstrated that TPS intensity canbe varied by more than two orders of magnitudes by simplychange the size and shape of nanoparticles [15,20–28]. As we havedescribed in earlier section that during nanostar formation fromnano seeds, the particle size and shape both change tremendouslyand as a result, one can monitor the shape evaluation process dur-ing nanostar formation using two-photon scattering process. Forthis purpose, first we have measured the bnano for each shape

Figure 3. TEM Images and corresponding photographs demonstrating the formation of different shaped branched gold nanomaterials by varying the CTAB concentration.

0 150 300 450 600 750 900

0

200

400

600

800

1000

1200

1400

TP

S In

tens

ity

Cha

nge

Time (seconds)

Figure 4. Time dependent TPS spectra demonstrating shape evaluation and growthkinetics during the formation of star shape gold nanoparticle.

Table 1Values for gold nanoparticles of different shapesmeasured by HRS technique.

NLO Systems (10�24 esu)

0.17

182

508

1290

D. Senapati et al. / Chemical Physics Letters 504 (2011) 46–51 49

separately, as reported in Table 1. To measure bnano for each shapeseparately, we have synthesized them separately just by varyingCTAB concentration as shown in Figure 2. As we have reported be-fore, these experimental results also suggest that CTAB turned outto have key features for controlling gold nanoparticle shape [3].The formation of various shapes is likely due to interplay betweenthe adsorption of surface capping agent and growth kinetics in-duced by the reducing agent.

It is interesting to note that, bnano = 1290 � 10�24 esu for 40 nmgold nanostar, which is about 5 orders of magnitude higher thanthe b values reported for the best available molecular chromoph-ores [15,20]. This very high b value can be due to several contribu-tions and these are [15,20–28] as follows. (1) First one is theelectric dipole approximation, which may arise due to the imper-fect structure in nanoparticle. This contribution is actually identi-cal to the one observed for any non-centrosymmetrical point-likeobjects such as efficient rod-like push–pull molecules. (2) The sec-ond contribution is multipolar contribution like electric quadru-pole contribution. This contribution is very important when thesize of the particle is no longer negligible when compared to thewavelength of light (d � k/10). (3) Due to the presence of multipod,the intense electromagnetic field located around the pods, whichcan help to increase the TPS intensity as we and other groups have

0 10 20 30 400

20

40

60

80

100

120%

Num

ber

Den

sity

Cha

nge

Time (seconds)

NanoseedsNanoflowers

40 50 60 70 800

20

40

60

80

100

120

% N

umbe

r D

ensi

ty C

hang

e

Time (seconds)

nanoflowernanocrown

150 300 450 600 750 900

0

20

40

60

80

100

120

% N

umbe

r D

ensi

ty C

hang

e

Time (seconds)

Nanocrowns

Nanostar

ba

c

Figure 5. Demonstration of growth kinetics for transient nanoparticles of different shapes during the synthesis of nanostar. (A) Showing number density change of nanoseedsand nanoflowers. (B) Showing number density change of nanoflowers and nanocrowns. (C) Showing number density change of nanocrowns and nanostar.

50 D. Senapati et al. / Chemical Physics Letters 504 (2011) 46–51

observed very high SERS intensity in the presence of nano-star. (4)Another possibility for high TPS intensity is due to increase innano-confined effect, which can enhance NLO properties as re-ported recently [32–33]. It is also interesting to note from Table1 that b value for nanostar is much higher than even nanoflower,though the size of nanoflower is bigger than nanostar and alsokmax shifted slightly towards blue for nanostar than that of nano-flower. Our experimental results clearly demonstrate that thepresence of intense electromagnetic field located around the podscan increase the TPS intensity tremendously and as a result,bnanostar > bcrown > bnanoflower. During shape evaluation from nano-seed ? nanoflower ? nanocrown ? nanostar, since minimumtwo different nanoparticles can exist together, ITPS can be ex-pressed as,

ITPS ¼ GhNwb2w þ ðNnano1 � CpNnano2Þb2

nano1

þ Nnano2b2nano2iI

2xe½�Nnano11ð�2xþ2�xÞþNnano21ð�2xþ2�x Þ� ð2Þ

where Nnano1 and Nnano2 are the number of nanoparticle1 and nano-particle2 per unit volume, bnano1 and bnano2 are the quadratic hyper-polarizabilities of a single nanoparticle1 and a single nanoparticle2.Cp is the ratio of number of gold atoms for single nanoparticle2 andsingle nanoparticle1 measured by ICPMS experiment, ex and e2x

are the molar extinction coefficient of the gold nanoparticle at xand 2x respectively, l is the path length and Ix the fundamentalintensity.

Here CpNnano2 denote the loss of number density of nanoparti-cle1 during the formation of nanoparticle2. Since bnano1 and bnano2

are known as reported in Table 1, from TPS intensity one can mea-sure Nnano2, since Nnano1 is known. Figure 5 reports the time depen-dent % number density change of nanoparticles of different shapesduring shape evaluation from nanoseed ? nanoflower ? nano-crown ? nanostar, measured by TPS experimental data.

4. Conclusions

In conclusion, in this Letter, for the first time we have demon-strated that the real time shape evolution of intermediates of col-loidal nanoparticles and their number density change can be fullyaccessed during the synthesis of star shaped gold nanoparticlesusing time dependent TEM and in situ TPS experiment. Our exper-imental data provide unique and convincing information forunderstanding the growth mechanisms of the shape evaluationfor gold nanocrystals. The overall process of the size and shape dis-tribution control process consists of several steps and these are:the nucleation process for the formation and accumulation ofnanoseeds, Ostwald ripening process for the formation of nano-flower and intraparticle ripening process for the formation of thestar sharp nanoparticles via nanocrowns. Our study demonstratesthe great potential of TPS experiment to study the real-time tran-sient concentration during star shape nanoparticle synthesis. Withthe high degree of synthetic control and techniques to probe shape

D. Senapati et al. / Chemical Physics Letters 504 (2011) 46–51 51

evaluation, colloidal gold nanocrystals will have a profound impactin scientific arenas as molecular biology, medicine and catalysis foroffering a highly tailored material.

Acknowledgements

Dr. Ray thanks NSF-PREM grant # DMR-0611539, NSF-CRESTgrant # HRD-0833178 and DOD grant # W 912HZ-06-C-0057 fortheir generous funding.

References

[1] H. Zheng, R.K. Smith, Y.-W. Jun, C. Kisielowski, U. Dahmen, A.P. Alivisatos,Science 324 (2009) 1309.

[2] G. Renaud et al., Science 300 (2003) 1416.[3] D. Senapati, A.K. Sing, P.C. Ray, Chem. Phys. Lett. 487 (2010) 88.[4] M. Grzelczak, J. Pérez-Juste, P. Mulvaney, L.M. Liz-Marzán, Chem. Soc. Rev.

(2008) 1783.[5] J. Polte, T.T. Ahner, F. Delissen, S. Sokolov, F. Emmerling, A.F.T. Qnemann, R.

Kraehnert, J. Am. Chem. Soc. 132 (2010) 1296.[6] B. Yan, Br.M. Reinhard, J. Phys. Chem. Lett. 1 (2010) 1595.[7] L.R. Lorenzo, F. Javier, G. Abajo, L.M. Liz-Marzn, J. Phys. Chem. C 114 (2010)

7336.[8] C.G. Khoury, T. Vo-Dinh, J. Phys. Chem. C 112 (2008) 18849.[9] I.V. Kityk, J. Ebothé, I. Fuks-Janczarek, A. Ali Umar, K. Kobayashi, M. Oyama, B.

Sahraoui, Nanotechnology 16 (2005) 1687.[10] A.J. Bonham, G. Braun, I. Pavel, M. Moskovits, N.O.J. Reich, Am. Chem. Soc. 129

(2007) 14572.[11] W. Lu et al., ACS Nano 4 (2010) 1739.

[12] W. Lu, A.K. Singh, S.A. Khan, D. Senapati, H. Yu, P.C. Ray, J. Am. Chem. Soc. 132(2010) 18103.

[13] S.D. Brown et al., J. Am. Chem. Soc. 132 (2010) 4678.[14] A.K. Singh, D. Senapati, A. Neely, H. Yu, P.C. Ray, Chem. A Eur. J. 16 (2010) 5600.[15] P.C. Ray, Chem. Rev. 110 (2010) 5332.[16] A.E. Prigodich, O.-S. Lee, W.L. Daniel, D.S. Seferos, G.C. Schatz, C.A. Mirkin, J.

Am. Chem. Soc. 132 (2010) 10638.[17] J.R. Kalluri et al., Angew. Chem. Int. Ed 48 (2009) 9668.[18] A. Louie, Chem. Rev. 110 (2010) 3146.[19] J. Griffin et al., Chem. Eur. J. 15 (2009) 342.[20] K. Clays, A. Persoons, Phys. Rev. Lett. 66 (1991) 2980.[21] P.C. Ray, Angew. Chem. 45 (2006) 1151.[22] G.K. Darbha, A.K. Singh, U.S. Rai, E. Yu, H. Yu, P.C. Ray, J. Am. Chem. Soc. 130

(2008) 8038.[23] J. Duboisset, I. Russier-Antoine, E. Benichou, G. Bachelier, C. Jonin, P.F. Brevet, J.

Phys. Chem. C 113 (2009) 13477.[24] M. Chandra, S.S. Indi, P.K. Das, J. Phys. Chem. C. 111 (2007) 10652.[25] G.K. Darbha, U.S. Rai, A.K. Singh, P.C. Ray, Chem. Eur. J. 14 (2008) 3896.[26] J.P. Novak, L.C. Brousseau, F.W. Vance, R.C. Johnson, B.I. Lemon, J.T. Hupp, D.L.

Feldheim, J. Am. Chem. Soc. 122 (2000) 12029.[27] M. Chandra, P.K. Das, Chem. Phys. 358 (2009) 203.[28] D. Segets, L. Martinez, J. Gradi, W. Peukert, J. Phys. Chem. C 113 (2009) 11995.[29] W. Ostwald, Lehrbuch der Allgemeinen Chemie, vol. 2, part 1, Leipzig,

Germany, 1896.[30] W.L. Noorduin, E. Vlieg, R.M. Kellogg, B. Kaptein, Angew. Chem. Int. Ed. 48

(2009) 9600.[31] Jongnam Park, Jin Joo, Soon Gu Kwon, Youngjin Jang, Taeghwan Hyeon, Angew.

Chem. Int. Ed. 46 (2007) 4630.[32] A.H. Reshak, A. Wojciechowski, W. Kuznik, A. Popeda, Physica E 42 (2010)

1769.[33] M. Makowska-Janusik, S. Tkaczyk, I.V. Kityk, J. Phys. Chem. B. 110 (2006) 6492.[34] J. Polte, T.T. Ahner, F. Delissen, S. Sokolov, F. Emmerling, A.F. Thünemann, R.

Kraehnert, J. Am. Chem. Soc. 132 (2010) 1296.