INSTITUTE OF PHYSICS PUBLISHING Nanotechnology 17 (2006) 40194024

NANOTECHNOLOGY doi:10.1088/0957-4484/17/16/004

Synthesis of silver nanoparticles using the polyol process and the inuence of precursor injection
Dongjo Kim, Sunho Jeong and Jooho Moon
Department of Materials Science and Engineering, Yonsei University, 134 Shinchon-dong Seodaemun-gu, Seoul 120-749, Korea E-mail: jmoon@yonsei.ac.kr

Received 20 May 2006, in nal form 29 June 2006 Published 14 July 2006 Online at stacks.iop.org/Nano/17/4019 Abstract Spherical silver nanoparticles with various sizes and standard deviations were synthesized by the polyol process. Two different synthesis methods were compared in order to investigate the inuence of reaction parameters on the resulting particle size and its distribution. In the precursor heating method, wherein a solution containing silver nitrate was heated to the reaction temperature, the ramping rate was determined to be a critical parameter affecting the particle size. In contrast, in the precursor injection method, in which a silver nitrate aqueous solution was injected into hot ethylene glycol, because of rapid nucleation, the injection rate and the reaction temperature were important factors in terms of reducing the particle size and attaining monodispersity. Silver nanoparticles with a size of 17 2 nm were obtained at an injection rate of 2.5 ml s1 and a reaction temperature of 100 C. (Some gures in this article are in colour only in the electronic version)

1. Introduction
Metal nanoparticles play important roles in many different areas. For example, they can serve as a model system to experimentally probe the effects of quantum connement on electronic, magnetic, and other related properties [1]. They have also been widely exploited for use in photography, catalysis, biological labelling, photonics, optoelectronics, information storage, surface-enhanced Raman scattering (SERS), and the formulation of magnetic ferrouids [25]. The intrinsic properties of metal nanoparticles are mainly determined by their size, shape, composition, crystallinity, and structure. In principle, any one of these parameters could be controlled to ne-tune the properties of metal nanoparticles [6]. Among the various metal nanoparticles, silver nanoparticles have been widely investigated because they exhibit unusual optical, electronic, and chemical properties, depending on their size and shape, thus opening many possibilities with respect to technological applications [79]. During the last two decades, many synthesis methods have been reported for the preparation of silver nanoparticles with tailor-made size, shape, and size controllability.
0957-4484/06/164019+06$30.00

In general, silver nanoparticles can be produced by various methods including the chemical reduction of silver ions with [10] or without [11] stabilizing agents, thermal decomposition in organic solvents [12], and chemical and photoreduction in reverse micelles [1315]. Using these methods, silver nanoparticles with spherical, octahedral, tetrahedral, hexagonal, cubic, wire, coaxial cable, triangular prism, disc, triangular mark, belt, and shell shapes have been manufactured [1622]. All these advances have promoted the scientic knowledge on the nature of nanomaterials. The synthesis method used in the present study is the so-called polyol method, which is well suited for the preparation of nano-sized metal or oxide particles of various shapes [23, 24]. In particular, the synthesis of spherical silver nanoparticles with high monodispersity from the polyol method has been reported in various studies [2529]. The general polyol process involves the dissolution of a protecting agent or stabilizer in a polyol medium. The required silver precursor is then added to this solution. Although the synthesis process of monodisperse silver nanoparticles is well established, little is known about the inuence of precursor injection during the polyol synthesis. In this paper, we have 4019

2006 IOP Publishing Ltd Printed in the UK

D Kim et al

1.5 0 min 10 min 20 min 30 min 40 min 50 min Time Increasing 0.5

1.0 Absorbance 0.0 300

400

500 Wavelength (nm)

600

700

Figure 1. XRD diffraction patterns of the synthesized silver nanoparticles at a temperature of 120 C and an injection rate of 1 ml s1 for 30 min by the injection method (0.25 mol l1 AgNO3 and 0.027 mol l1 PVP).

Figure 2. UVvis spectra during the formation of silver nanoparticles at a temperature of 120 C and an injection rate of 1 ml s1 with time intervals of 10 min after precursor injection. (0.25 mol l1 AgNO3 and 0.027 mol l1 PVP).

synthesized silver nanoparticles by either heating the precursor solution to the reaction temperature or injecting an aqueous silver salt solution into hot ethylene glycol. The inuences of the synthesis conditions including the reaction temperature, concentration of reactants, and injection rate were examined for both methods.

2.3. Apparatus The injection of the precursor was carried out using a syringe pump (KDS200, Kd Scientic) at a constant rate. The synthesized silver nanoparticles were characterized by scanning electron microscopy (SEM, JEOL JSM-6500). The mean particle size of silver nanoparticles and the standard deviation of particle population were determined from image analyses of SEM micrographs of each particle. The x-ray diffraction (XRD) patterns of silver nanoparticles were recorded using a D/max-Rint 2000 (Rigaku), and ultraviolet visible (UVvis) spectroscopy was conducted for the samples extracted in the course of the reaction using a quartz cuvette (6030-UV, Hellma) by a UVvis spectrophotometer (JASCO V-570).

2. Experimental details
2.1. Synthesis of silver particles by heating the precursor solution Silver nitrate (99.9%, Aldrich), used as a precursor of Ag, was dissolved in ethylene glycol (EG, 99.9%, Aldrich) together with polyvinylpyrrolidone (PVP, Mw = 10 000, Aldrich), which protects the synthesized silver nanoparticles from agglomeration. This solution was stirred vigorously in a reactor tted with a reux condenser, followed by heating the precursor solution at temperatures of 100150 C, at a constant heating rate of 17.5 C min1 . The reaction was maintained for 30 min at each reaction temperature. After the reaction was completed, the solution was cooled to room temperature, and the silver particles were separated from the liquid by centrifugation and then repeatedly washed with ethanol. The resulting particles were dried at room temperature. 2.2. Synthesis of silver particles by injecting precursor solution All the chemicals used in the precursor injection method are identical to those described above, except for the injection of silver nitrate aqueous solution. First, PVP was completely dissolved in ethylene glycol and this solution was heated to the reaction temperature. Deionized water containing 40 wt% silver nitrate was injected into a hot PVPethylene glycol solution maintained at the reaction temperature. The particle recovery procedures are also identical to those of for the previous method. 4020

3. Results and discussion

The x-ray diffraction patterns and UVvis spectroscopy reveal that the synthesized silver particles have good crystallinity and high purity, despite the silver nanoparticles being prepared by the reduction of a precursor at low temperature. The x-ray diffraction pattern of typical silver nanoparticles synthesized by the precursor injection method under the conditions of a 1 ml s1 injection rate and 120 C reaction temperature for 30 min is presented in gure 1. The x-ray diffraction patterns of all synthesized silver particles in this study are almost identical and reveal high crystallinity. The spectroscopic analysis results of the synthesized silver nanoparticles, whose reaction conditions corresponded with those of the samples used for the XRD analysis, are shown in gure 2 as a function of time at intervals of 10 min. The silver particles were generated gradually, as is clearly indicated by the increasing intensity of the surface plasmon band at 420 nm in the UV vis spectra. Furthermore, we observed no signicant change in the absorption peak intensities within 1020 min after the precursor was injected. This implies that the nucleation and growth reactions reach completion within such a time period in which the specic reaction kinetics depend on

Synthesis of silver nanoparticles using the polyol process and the inuence of precursor injection

Table 1. Synthesis conditions for the preparation of silver particles using the method of heating the dissolved solution. AgNO3 (mol l1 ) 0.25 0.25 0.25 0.25 0.25 0.004 0.016 PVP (mol l1 ) 0.027 0.027 0.027 0.027 0.027 0.002 0.093 Temperature ( C) 120 120 120 100 150 120 120 Time (min) 30 30 30 30 30 30 30 Heating rate ( C min1 ) 1 5 7.5 1 1 1 1 Mean size (nm) 42 27 18 27 33 36 37 Standard deviation (nm) 14 7 4 10 10 11 13 Figure ID 3(a) 3(b) 3(c) 4(a) 4(b) 4(c) 4(d)

Sample 1 2 3 4 5 6 7

the silver nitrate precursor concentrations. Based on these observations, we selected a reaction time of 30 min for further syntheses to assure complete reaction from silver ions to silver nanoparticles. 3.1. Synthesized silver particles by heating the precursor dissolved solution It is well known that the morphology and size distribution of metallic particles produced by the reduction of metallic salts in solution depends on various reaction conditions such as temperature, time, concentration, molar ratio of metallic salt/reducing agent, mode and order of addition of reagents, presence and type of protective agents, degree and type of agitation, and whether nucleation is homogeneous or heterogeneous [27]. The synthesis conditions used for producing the Ag particles by heating the precursor solution are summarized in table 1. The mean particle size and the standard deviation of the resulting Ag particles obtained at each set of synthesis conditions are also presented, together with gure number identications for the corresponding particles. The reaction scheme for producing ne and monodisperse silver particles using the polyol process involves the following successive reactions: reduction of the soluble silver nitrate by ethylene glycol, nucleation of metallic silver, and growth of individual nuclei in the presence of a protective agent, PVP. The fully reacted particle sizes synthesized from the polyol process depended strongly on the ramping rate of the precursor solution to the reaction temperature; at a lower heating rate larger particles were generated, most likely due to a slower nucleation rate, while at a higher rate faster nucleation produced smaller-sized particles. This inuence on the Ag particles is shown in gure 3. At a heating rate of 1 C min1 , the mean size of silver particles was 42 nm (gure 3(a) of sample 1), and increasing the heating rate to 7.5 C min1 yielded smaller and more monodisperse particles with a mean size of 18 nm (gure 3(c) of sample 3). The particle size dependence on the reaction temperature and the precursor concentration of the reactant are also illustrated in gure 4. The particle size of the silver decreased slightly when the reaction temperature was decreased from 150 to 100 C (from gures 4(a) to (b)). The particles prepared at concentrations of 0.004 and 0.016 M exhibited almost identical sizes of 36 nm (gures 4(c) and (d)). These ndings indicate that the particle size dependence on temperature and concentration is negligible; rather, the heating rate is a more critical factor in determining the particle size in the method of heating the precursor solution.

Figure 3. SEM images of the synthesized silver nanoparticles by the method of heating the dissolved solution at different heating rates of (a) 1 C min1 , (b) 5 C min1 and (c) 7.5 C min1 (details in table 1).

3.2. Synthesized silver particles using precursor injection method In order to obtain monodisperse metal particles, generally, rapid nucleation in a short period of time is important; that is, almost all ionic species have to be reduced rapidly to metallic species simultaneously, followed by conversion to stable nuclei so as to be grown [28]. In the method of heating a precursor solution, however, both nucleation and growth can 4021

D Kim et al

Table 2. Synthesis conditions for the preparation of silver particles using the injection method. AgNO3 (mol l1 ) 0.25 0.25 0.25 0.25 0.25 PVP (mol l1 ) 0.027 0.027 0.027 0.027 0.027 Temperature ( C) 100 120 150 100 120 Time (min) 30 30 30 30 30 Injection rate (ml s1 ) 1 1 1 2.5 0.2 Mean size (nm) 19 36 54 17 Standard deviation (nm) 3 8 13 2 Figure ID 5(a) 5(b) 5(c) 6(a) 6(b)

Sample 8 9 10 11 12

Figure 4. SEM images of the synthesized silver nanoparticles by the method of heating the dissolved solution at different temperatures of (a) 100 C and (b) 150 C and at different precursor concentrations of (c) 0.004 mol l1 AgNO3 , 0.002 mol l1 PVP and (d) 0.016 mol l1 AgNO3, 0.093 mol l1 PVP (details in table 1).

proceed gradually with increasing temperature. As such, it is difcult to synthesize particles with high monodispersity. Therefore, the rapid injection of silver nitrate aqueous solution into ethylene glycol maintained at the reaction temperature would guarantee a short burst of nucleation after which the nuclei would continue to grow without additional nucleation, thus ensuring monodispersity. Upon addition of the silver nitrate aqueous solution to hot ethylene glycol, the Ag+ species are reduced to metallic silver. The concentration of metallic silver in solution increases, reaching the supersaturation conditions and nally the critical concentration to nucleate. Spontaneous nucleation then takes place very rapidly and many nuclei are formed in a short time, lowering the silver concentration below the nucleation and supersaturation levels into the saturation concentration region. After a short period of nucleation, the nuclei grow by the deposition of metallic silver until the system reaches the saturation concentration. At the end of the growth period, all the metal particles have grown at almost the same rate and the system exhibits a narrow particle size distribution. Table 2 summarizes the synthesis conditions, each of which yielded silver nanoparticles of different size and morphological features, using the precursor injection method. 4022

The size of the resulting silver particles was strongly dependent upon the reaction temperature and the injection rate. The particle size dependence on the reaction temperature, which is dened as the temperature of ethylene glycol when the precursor solution is injected, is exhibited in gure 5. On decreasing the reaction temperature from 150 to 100 C, at an injection rate of 1 ml s1 , the mean size of the nal silver particles was reduced from 54 nm (gure 4(c) of sample 10) to 19 nm (gure 4(a) of sample 8). This temperature dependence on particle size can be explained as follows. Because of the relatively high temperature used in the synthesis of silver particles, the Brownian motion and mobility of surface atoms increase. This enhances the probability of particle collision, adhesion, and subsequent coalescence. However, PVP is added to protect the particles from agglomeration. Particle coalescence is the means by which the system tries to attain thermodynamic equilibrium by reducing its total surface area. The injection rate also exhibits a signicant inuence on the size and the morphology of the synthesized silver nanoparticles. The silver particles from the more rapid injection rate of 2.5 ml s1 are shown in gure 6(a) (sample 11). A further increase in the injection rate yielded silver particles with a slightly reduced size of 17 nm. However,

Synthesis of silver nanoparticles using the polyol process and the inuence of precursor injection

Figure 6. SEM images of the synthesized silver nanoparticles at different conditions: (a) a temperature of 100 C and an injection rate of 2.5 ml s1 ; (b) a temperature of 120 C and an injection rate of 0.2 ml s1 (details in table 2).

Figure 5. SEM images of the synthesized silver nanoparticles by the injection method at different temperatures of (a) 100 C, (b) 120 C and (c) 150 C (details in table 2).

the silver particles obtained at an injection rate of 0.2 ml s1 are polydisperse, with a size ranging from tens of nanometres to tens of micrometres. The slower injection rate results in larger silver particles with a broader distribution in size and shape. It is believed that the slower injection rate condition induces a long period of nucleation. Upon precursor injection, a few nuclei form and stable nuclei grow over the time that silver nitrate is continuously injected into the ethylene glycol. Some of these injected silver ions are converted to nuclei and some are used for growing pre-existing nuclei. Eventually, polydispese particles with a larger size are obtained. Based on these observations, we found that careful control of the precursor injection rate and the reaction temperate is important in terms of obtaining silver nanoparticles with higher monodispersity in the polyol method. 3.3. Comparison between two methods Figure 7 presents a schematic illustration of the silver nanoparticle synthesis methods of precursor heating and precursor injection. The notable difference between the two methods is the period of reduction of Ag+ species. When the precursor dissolved solution is heated, the temperature of

the reactants increases gradually and the reduction of Ag+ also proceeds slowly, followed by the concentration of Ag0 approaching a critical concentration for nucleation. When nucleation occurs, some Ag0 species convert to nuclei and some Ag+ reduce to Ag0 continuously; that is, the nucleation step is maintained for a relatively long period of time. The long period of the nucleation step yields synthesized particles that are larger-sized and polydispersive. For the injection method, in contrast, Ag+ species dissolved in water (or other solvents) reduce to Ag0 rapidly upon injection into the hot solution. As a result, the solution reaches a critical concentration for nucleation, and consequently the nucleation occurs in a short period of time. Therefore, the injection of precursor is a critical step for synthesizing nanoparticles with a reduced size and a narrow size distribution.

4. Conclusions
Spherical silver nanoparticles with a controllable size and high monodispersity were synthesized by the polyol method. Two different synthesis methods for producing the Ag nanoparticles were compared in terms of particle size and monodispersity. Silver nanoparticles with a size of 18 4 nm were obtained at a reaction temperature of 120 C and a heating rate of 7.5 C min1 in the precursor heating method, where the heating rate was a critical parameter affecting particle size. In the precursor injection method, on the other hand, the injection rate and reaction temperature were important factors for producing uniform-sized Ag with a reduced size. Silver nanoparitlces with a size of 17 2 nm were obtained at an injection rate of 2.5 ml s1 and a reaction temperature of 100 C. The injection of the precursor solution into a hot solution is an effective means to induce rapid nucleation 4023

D Kim et al

Figure 7. Schematic illustration of the two synthesis methods: heating the precursor dissolved solution and injection of the precursor.

in a short period of time, ensuring the fabrication of silver nanoparticles with a smaller size and a narrower size distribution.

Acknowledgment
This work was supported by the National Research Laboratory (NRL) Program of the Korea Science and Engineering Foundation.

References
[1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] El-Sayed M A 2001 Acc. Chem. Res. 34 257 Pileni M-P 2001 Adv. Funct. Mater. 11 323 Nie S and Emory S R 1997 Science 275 1102 Kamat P V 2002 J. Phys. Chem. B 106 7729 Maier S A, Brongersma M L, Kik P G, Meltzer S, Requicha A A G and Atwater H A 2001 Adv. Mater. 13 1501 Sun Y and Xia Y 2002 Science 298 2176 Sun Y, Yin Y, Mayers B T, Herricks T and Xia Y 2002 Chem. Mater. 14 4736 Zhang D, Qi L, Yang J, Ma J, Cheng H and Huang L 2004 Chem. Mater. 16 872 Suber L, Sondi I, Matijevic E and Goia D V 2005 J. Colloid Interface Sci. 288 489 Toshima N, Yonezawa T and Kushihashi K 1993 J. Chem. Soc. Faraday Trans. 89 2537 Liz-Marzan L M and Philipse A P 1995 J. Phys. Chem. 99 15120

[12] Esumi K, Tano T, Torigoe K and Meguro K 1990 Chem. Mater. 2 564 [13] Mafune F, Kohnok J Y, Takeda Y and Kondow T 2000 J. Phys. Chem. B 104 8333 [14] Huang H H, Ni X P, Loy G L, Chew C H, Tan K L, Loh F C, Deng J F and Xu G Q 1996 Langmuir 12 909 [15] Pileni M P, Lisiecki I, Motte L, Petit C, Cizeron J, Moumen N and Lixon P 1993 Prog. Colloid Polym. Sci. 93 1 [16] Rodrituez-Sanchez L, Blanco M C and Lopez-Quintela M A 2000 J. Phys. Chem. B 104 9683 [17] Salkar R A, Jeevanandam P, Aruna S T, Koltypin Y and Gedanken A 1999 J. Mater. Chem. 9 1333 [18] He R, Qian X, Yin J and Zhu Z 2002 J. Mater. Chem. 12 3783 [19] Chen S and Carroll D L 2002 Nano Lett. 2 1003 [20] Sun Y, Gates B, Mayers B and Xia Y 2002 Nano Lett. 2 165 [21] Pastoriza-Santos I and Liz-Marzan L M 2002 Nano Lett. 2 903 [22] Callegari A, Tonti D and Chergui M 2003 Nano Lett. 3 1565 [23] Feldmann C 2003 Adv. Funct. Mater. 13 101 [24] Feldmann C and Jungk H-O 2001 Angew. Chem. Int. Edn 40 359 [25] Slistan-Grijalva A, Gerrera-Urbina R, Rivas-Silver J F, Avalos-Borja M, Castillon-Barraza F F and Posada-Amarillas A 2005 Physica E 25 438 [26] Sun Y, Mater B, Herricks T and Xia Y 2003 Nano Lett. 3 955 [27] Sanguesa C D, Urbina R H and Figlarz M 1992 J. Solid State Chem. 100 272 [28] LaMer V K and Rober H D 1950 J. Am. Chem. Soc. 72 4847 [29] Silvert P-Y, Urbina R H, Duvauchelle N, Vijayakrishnan V and Elhsissen K T 1996 J. Mater. Chem. 6 573

4024