- Email: [email protected]
Highly charging of nanoparticles through electrospray of nanoparticle suspension
Journal of Colloid and Interface Science 287 (2005) 135–140 www.elsevier.com/locate/jcis
Highly charging of nanoparticles through electrospray of nanoparticle suspension Jeongsoo Suh a , Bangwoo Han b , Kikuo Okuyama c , Mansoo Choi a,∗ a National CRI Center for Nano Particle Control, School of Mechanical and Aerospace Engineering, Seoul National University, Seoul 151-742, South Korea b Eco-machinery Engineering Department, Korea Institute of Machinery and Materials (KIMM), Daejon 305-343, South Korea c Department of Chemical Engineering, Graduate School of Engineering, Hiroshima University, Higashi-Hiroshima 739-8527, Japan
Received 24 September 2004; accepted 24 January 2005 Available online 22 March 2005
Abstract Patterned deposition of nanoparticles is a prerequisite for the application of unique properties of nanoparticles in future nanodevices. Recent development of nanoxerography requires highly charged aerosol nanoparticles to avoid noise deposition due to random Brownian motion. However, it has been known that it is difficult to charge aerosol nanoparticles with more than two elementary charges. The goal of this work is to develop a simple technique for obtaining highly charged monodisperse aerosol nanoparticles by means of electrospray of colloidal suspension. Highly charged aerosol nanoparticles were produced by electrospraying (ES) and drying colloidal suspensions of monodisperse gold nanoparticles. Size and charge distributions of the resultant particles were measured. We demonstrate that this method successfully charges monodisperse nanoparticles very highly, e.g., 122 elementary charges for 25.0 nm, 23.5 for 10.5 nm, and 4.6 for 4.2 nm. The method described here constitutes a convenient, reliable, and continuous tool for preparing highly charged aerosol nanoparticles from suspensions of nanoparticles produced by either wet chemistry or gas-phase methods. 2005 Elsevier Inc. All rights reserved. Keywords: Charging; Patterning; Monodisperse particles; Electrospray
1. Introduction The application of nanoparticles in electronic, magnetic, or optical devices such as quantum devices, field emission display, single-electron transistors, and data storage devices requires the precise and accurate assembly of nanoparticles in the desired patterns. At the same time, parallel patterning of nanoparticles is needed for practical applications. The recently developed method of nanoxerography [1] utilizes charged particles and charge pattern transferred substrates. In this process, the charging of particles plays an important role in positioning aerosol nanoparticles onto a desired area [2–4]. The charging of aerosol nanoparticles has been also of great importance for the controlled synthesis [5,6] and * Corresponding author. Fax: +82-2-878-2465.
E-mail address: [email protected] (M. Choi). 0021-9797/$ – see front matter 2005 Elsevier Inc. All rights reserved. doi:10.1016/j.jcis.2005.01.078
deposition [7] of aerosol nanoparticles as well as in aerosol measurement systems such as differential mobility analyzers (DMA) [8,9]. Aerosol charging can generally be done by the collision of aerosol particles with ambient ions. Radioactive materials [10], corona discharge [11], ultraviolet rays [12], and soft X-rays [13,14] are used to generate ions and charge aerosol particles into the well-known stationary charging state. However, as particle size becomes smaller, the charging ratio of the particles becomes smaller. Even in a monopolar ion atmosphere, it is difficult to obtain charged nanoparticles carrying over ±2 charges. Therefore, it is very much needed to develop a method to prepare highly charged aerosol nanoparticles for enhancing the deposition efficiency and suppressing the diffusion effect during the nanoparticle patterning processes. It is noted that monodisperse nanoparticles can be effectively prepared by wet chemistry methods
136
J. Suh et al. / Journal of Colloid and Interface Science 287 (2005) 135–140
using surfactants [15]. If we combine the size-controlling capability of wet chemistry method with aerosol technology, we could develop a method to prepare highly charged monodisperse aerosol nanoparticles that are needed for the precise patterning. Electrospray, since it was first reported by Zeleny [16], has been widely used for a variety of applications such as electrostatic coating, fuel supply systems, and crop spraying. It has recently been used for the generation of standard nanometer particles [17] and the production of aerosol nanoparticles for the synthesis of materials [6]. Electrospray of colloidal suspension has been recently utilized for sizing the colloidal nanoparticles by combination of the use of a differential mobility analyzer [18]. It is well known that electrospray synthesis of nanoparticles can generate highly charged droplets and particles. However, this method has not been applied to charging already-made particles. In the present study, we aim to develop a method using the electrospray of colloidal suspension to prepare highly charged monodisperse aerosol nanoparticles and to determine the charge state of the nanoparticles after electrospraying and complete drying. Two issues are answered before the electrospray of colloidal suspension can be claimed as a reliable method for preparing highly charged aerosol nanoparticles. The first is to make sure that the original particle size remains even after electrospraying and drying of nanoparticle suspension. The second is whether the highly charged state of electrosprayed droplets maintains or disappears after the complete evaporation of liquid covered on the surface of the original nanoparticle. A couple of the previous studies on the electrical mobility analysis of the ions produced via electrospray indicated some unexpected peaks that were not corresponding to ion signals and the peaks were considered to be the ones for highly charged aerosol residues [19,20]. Loscertales and Fernández de la Mora [20] analyzed the mobility of residue particles using a differential mobility analyzer (DMA) and a hypersonic impactor in order to investigate ion field evaporation from highly charged droplets after electrospraying. They indirectly showed that aerosol particles were highly charged. Han et al. [21] also showed that impurity nanoparticles could be multiply charged in the process of analyzing the mobility of ion clusters when pure water, tap water, and aqueous solutions were electrosprayed. However, direct morphological and compositional measurements of the sampled particles have not been investigated. Electrospray charging of colloidal suspensions of nanoparticles has not been attempted. The charge state of the electrosprayed and dried nanoparticles needs to be determined depending on the particle size. Here, colloidal suspension of monodisperse gold nanoparticles were electrosprayed and then dried to prepare highly charged monodisperse aerosol nanoparticles. Their charge and size distributions were measured to examine whether the size of the original nanoparticles maintains and highly charged state could be attained.
2. Methods Figs. 1a and 1b show the experimental apparatus used in this study. Monodisperse gold nanoparticle suspensions (0.01% HAuCl4 suspended in water-based solution, G1, G2, and G3, Sigma–Aldrich Chemie Gmbh) were used and the specified nominal sizes of the particles were 5 ± 2 nm for G1, 10 ± 2 nm for G2, and 20 ± 3 nm for G3 with narrow size distributions. Mixtures of the colloidal suspension and methanol (50/50, v/v) were prepared to stabilize the electrospraying condition. It was supplied into a capillary by a highresolution syringe pump (Model 5M361F, ORION). A positive DC voltage of 2–3 kV (Model 205B-10R, BERTAN) was applied to the metal capillary (0.2 mm O.D. and 0.1 mm I.D.) located 1–3 cm from the counterelectrode. A proper voltage-generating straight cone-jet mode was found using a visualization system composed of a CCD camera and a TV monitor. Flow rates of a carrier gas (nitrogen, N2 ) and a liquid precursor are fixed at 2.0 l/min and 10 µl/h, respectively. The proper liquid feed rate was chosen to achieve one particle per one droplet by measuring the size distributions of the resultant particles and comparing these with the nominal size of particles contained in the colloid solution. Highly charged droplets containing gold nanoparticles were formed at the end of the liquid cone and complete solvent evaporation results in highly charged solid aerosol
(a)
(b) Fig. 1. (a) Electrospray of colloidal suspension contained monodisperse particles and drying for highly charged monodisperse aerosol nanoparticles. (b) Measurement system for size (A), electrical mobility (B), and charge (C) distributions.
J. Suh et al. / Journal of Colloid and Interface Science 287 (2005) 135–140
nanoparticles as shown in Fig. 1a. A nanodifferential mobility analyzer (nano-DMA, Model 3085, TSI Inc.), a scanning mobility particle sizer (SMPS, Model 3934, 3936, TSI Inc.), and a Faraday cup electrometer (FCE, Wyckoff Co. Ltd., Japan) were used to measure the sizes and charge distribution of the resultant nanoparticles. Transmission electron microscopy together with electron dispersive spectrometry was also used to confirm size, morphology, and composition of collected particles. All the transport stainless tubes connected to measurement systems were grounded to prevent the highly charged particles from sticking to the tube wall. This measurement system consists of three parts. As shown in Fig. 1b, the size, electrical mobility, and charge distributions of particles are measured through A, B, and C directions, respectively. A polonium (Po-210) neutralizer is added before reaching SMPS to minimize the possibility of incomplete neutralization. Tandem DMA (TDMA) has been chosen to analyze the charge distribution of particles. In this system, gold nanoparticles of equal electrical mobility are classified through a nano-DMA and then they are fed into the SMPS to measure particle size distribution of equal mobility particles. By means of the TDMA system, charge distribution of particles can be simply determined in case of monodisperse particles. Since electrical mobility of a singly charged particle is known, the charge distribution of particles, f (qi ), is calculated as qi =
Zp (Vi ) , Zp (q = 1, dp )
137
(a)
Ni (Vi ) f (qi ) = ∞ , i=1 Ni
where qi is the number of elementary charges, Vi is the voltage of nano-DMA (Model 3085, TSI Inc.), Zp is the electrical mobility at the voltage Vi , Zp (q = 1, dp ) is the electrical mobility of singly charged particles at size dp (cm2 /V s), Ni is the particle number concentration at voltage Vi (particles/cm3 ), and f (qi ) is the normalized number fraction of particles having qi charges.
(b)
3. Results and discussion Fig. 2 shows the size distributions of gold nanoparticles (G1, G2, G3) measured by a neutralizer and an SMPS when they are electrosprayed and dried. As shown in the figure, the geometric mean diameter of the particle is 4.2 nm for G1, 10.4 nm for G2, and 25.0 nm for G3, respectively. In all cases, they show very narrow distributions (σg = 1.1) that can be considered as monodisperse nanoparticles. We measured the electrical mobility distributions of these electrosprayed nanoparticles in order to investigate whether they were highly charged states. Fig. 3 shows the distributions of electrical mobility for G1, G2, and G3 obtained by a nano-DMA (Model 3085, TSI Inc.), FCE, and CPCs (Model 3022A and 3025, TSI Inc.). The abscissa means electrical mobility and the ordinate stands for arbitrary unit of the currents measured by FCE. As shown in the figure, the peak val-
(c) Fig. 2. Size distributions of the gold nanoparticles. (a) G3, (b) G2, (c) G1.
ues of electrical mobility (Zp ) are almost the same regardless of the particle size. It means that the number of elementary charges (q) increases as the particle size increases. The electrical mobility of the particles should be 0.080 cm2 /V s for 4.2 nm, 0.017 cm2 /V s for 10.5 nm, and 0.004 cm2 /V s
138
J. Suh et al. / Journal of Colloid and Interface Science 287 (2005) 135–140
(a) Fig. 3. Mobility distributions of gold nanoparticles.
(b)
Fig. 4. TEM photographs of gold nanoparticles. (a, b) for G3, (c) for G2, (d) for G1.
for 25.0 nm if they are singly charged. However, the mobility of gold nanoparticles prepared by the present method has a significantly higher value (Zp = 0.44 cm2 /V s) than those of the singly charged particles. These results confirm that the gold nanoparticles have been highly charged. To prove whether these particles are real gold particles or solvent residues, TEM analysis was done together with energy dispersive spectrometry. Fig. 4 shows the morphology of the resultant particles. Fig. 4a is the morphology of the particles directly collected after electrospraying G3 without classification and Figs. 4b–4d are the morphologies of
(c) Fig. 5. Charge distributions of the resultant nanoparticles. (a) G1, (b) G2, (c) G3.
the particles having equal mobility (Zp = 0.44 cm2 /V s), classified by nano-DMA, for G3, G2, and G1, respectively. It is clear that the electrosprayed gold particles are almost monodisperse and spherical. The mean sizes of the particles
J. Suh et al. / Journal of Colloid and Interface Science 287 (2005) 135–140
139
Table 1 Electrical properties of electrosprayed and dried gold nanoparticles at Q = 10 µl/h
Fig. 6. Variations of number of elementary charges depending on particle size.
measured by TEM (Dp = 4 nm for G1, 11 nm for G2, 23 nm for G3) are very close to those measured by SMPS. The sizes of gold nanoparticles were also examined directly from colloidal suspensions without electrospraying and they were almost the same as the ones obtained after electrospraying and drying. These particles were examined by the analysis of an energy dispersive spectrometry confirming gold particles. Charge distributions were also measured and the results were shown in Fig. 5. The abscissa describes the number of elementary charge (qi ) of particles and the ordinate is the normalized number concentration of the particles. It is found that the mean numbers of elementary charge are 4.6 for 4.2 nm, 23.5 for 10.5 nm, and 122 for 25.0 nm, respectively. For the monodisperse particles, average charge can also be simply calculated from mobility distribution (see Fig. 3). Dividing the peak mobility by that of singly charged particles gives qi = 4.8 for 4.2 nm, 24 for 10.5 nm, and 128 for 25.0 nm, respectively. These results are almost the same as the values shown in Fig. 5. These results demonstrate that we could easily prepare highly charged monodisperse aerosol nanoparticles from colloidal suspensions by electrospraying and drying. Also, they suggest that the number of elementary charges on the particle increases as the size of the particle becomes larger. The number of charges is approximately proportional to Dp2 . This can be explained by the fact that the electrical mobility, Zp , is proportional to the ratio q/Dp2 in the free molecular limit (Dp 40 nm) for ion evaporation from charged droplets [19,20]. Since the measured electrical mobilities of particles are shown to be independent of particle sizes (see Fig. 3), q would be proportional to Dp2 . Fig. 6 shows that the average number of charges in our experiment is proportional to Dp1.95 . The constant mobility value of our nanoparticles supports the ion evaporation model [19,20]. If the droplets solely ex−1/2 perience Rayleigh fission, their field would be E ∼ dp (here, dp is particle diameter), whereas if droplets emit ions through ion evaporation, the droplets would adjust their ion evaporation rate according to the rate of solvent evaporation; as a result, the electric field, E, could become rela-
Used colloid
Dg a (nm)
G1 G2 G3
4.2 10.5 25.0
Zp b (m2 /V s) 0.44 × 10−4 0.44 × 10−4 0.44 × 10−4
qc (–)
Ed (V/nm)
ER e (V/nm)
E/ER (–)
4.6 23.5 122
1.25 1.18 1.12
3.28 2.07 1.35
0.38 0.57 0.83
Note. Surface tension was measured by DuNouy ring method (digital tensiometer, K9, Kruss). a Geometric mean diameter of particles measured by SMPS (Model 3934, TSI Inc.). b Mobility of particles measured by DMA (Model 3085, TSI Inc.) and FCE (Wyckoff Co. Ltd., Japan). c Number of average charge of particles. d Electric field of particles calculated using E = qe/π ε d 2 (ε = 8.855× 0 p 0 10−12 F/m, e = 1.602 × 10−19 C). e Electric field at Rayleigh limit calculated from E = (8γ /ε d )1/2 and R 0 p based on ε0 = 8.855 × 10−12 F/m, γ = 0.050 N/m.
tively constant [20]. Table 1 shows the electrical properties of the resultant nanoparticles in our experiments. E is quite close to the earlier observation (∼1 V/m) for ion evaporation [19,20] independent of particle size, which is well below the Rayleigh limit (ER ). Decreasing the particle size causes more departure from the Rayleigh limit. Therefore, the gold nanoparticles in our experiments would be saturated with ions at the ion evaporation limit and thus would lose some ions at the surface while drying process occurs. However, final dried gold nanoparticles still attain high charge states that are difficult to obtain by conventional aerosol charging methods.
4. Conclusions We demonstrated that electrospraying of nanoparticle suspensions and subsequent drying produced highly charged aerosol nanoparticles having the same sizes as the original nanoparticles. Measurements of size and charge distributions by using a tandem DMA system and TEM analysis were done and confirmed the validity of the method. The geometric mean numbers of charge are 4.6 for 4.2 nm, 23.5 for 10.5 nm, and 122 for 25.0 nm, respectively. The number of charges is proportional to Dp1.95 . The method described here constitutes a convenient, reliable, and continuous tool for preparing highly charged aerosol nanoparticles from nanoparticles produced by either wet chemistry or gasphase methods.
Acknowledgments This work was funded by the Creative Research Initiatives Program supported by the Ministry of Science and Technology, Korea. J. Suh was partially supported by the BK 21 program funded by the Ministry of Education.
140
J. Suh et al. / Journal of Colloid and Interface Science 287 (2005) 135–140
References [1] H.O. Jacobs, G.M. Whitesides, Science 291 (2001) 1763–1766. [2] T.J. Krinke, H. Fissan, K. Deppert, M.H. Magnusson, L. Samuelson, Appl. Phys. Lett. 78 (2001) 3708–3710. [3] H.O. Jacobs, S.A. Campbell, M.G. Steward, Adv. Mater. 14 (2002) 1553–1557. [4] M. Kang, H. Kim, B. Han, J. Suh, J. Park, M. Choi, Microelectron. Eng. 71 (2004) 229–236. [5] M. Adachi, S. Tsukui, K. Okuyama, Jpn. J. Appl. Phys. 42 (2003) L77–L79. [6] K. Nakaso, B. Han, K.H. Ahn, M. Choi, K. Okuyama, J. Aerosol Sci. 34 (2003) 869–881. [7] T.J. Krinke, K. Deppert, M.H. Magnusson, H. Fissan, Part. Part. Syst. Charact. 19 (2002) 321–326. [8] E.O. Knutson, K.T. Whitby, J. Aerosol Sci. 6 (1975) 443–451. [9] K. Okuyama, M. Shimada, A. Okita, Y. Otani, S.J. Cho, J. Aerosol Res. Jpn. 13 (1998) 83–93. [10] M. Adachi, Y. Kousaka, K. Okuyama, J. Aerosol Sci. 16 (1985) 109– 123.
[11] P. Büscher, A. Schmidt-Ott, A. Wiedensohler, J. Aerosol Sci. 25 (1994) 651–663. [12] M. Shimada, S.-J. Cho, K. Okuyama, T. Tamura, M. Adachi, T. Fujii, J. Aerosol Sci. 28 (1997) 649–661. [13] M. Shimada, B. Han, K. Okuyama, Y. Otani, J. Chem. Eng. Jpn. 35 (2002) 786–793. [14] B. Han, M. Shimada, M. Choi, K. Okuyama, Aerosol Sci. Technol. 37 (2003) 330–341. [15] V.F. Puntes, K.M. Krishnan, A.P. Alivisatos, Science 291 (2001) 2115– 2117. [16] J. Zeleny, Phys. Rev. 3 (1914) 69–91. [17] D.-R. Chen, D.Y.H. Pui, S.L. Kaufman, J. Aerosol Sci. 26 (1995) 963– 977. [18] I.W. Lenggoro, B. Xia, K. Okuyama, Langmuir 18 (2002) 4584–4591. [19] V. Katta, A.L. Rockwood, M.L. Vestal, Int. J. Mass Spectrom. Ion Proc. 103 (1991) 129–148. [20] I.G. Loscertales, J. Fernández de la Mora, J. Chem. Phys. 103 (1995) 5041–5060. [21] B. Han, I.W. Lenggoro, M. Choi, K. Okuyama, Anal. Sci. 19 (2003) 843–851.
Highly charging of nanoparticles through electrospray of nanoparticle suspension Jeongsoo Suh a , Bangwoo Han b , Kikuo Okuyama c , Mansoo Choi a,∗ a National CRI Center for Nano Particle Control, School of Mechanical and Aerospace Engineering, Seoul National University, Seoul 151-742, South Korea b Eco-machinery Engineering Department, Korea Institute of Machinery and Materials (KIMM), Daejon 305-343, South Korea c Department of Chemical Engineering, Graduate School of Engineering, Hiroshima University, Higashi-Hiroshima 739-8527, Japan
Received 24 September 2004; accepted 24 January 2005 Available online 22 March 2005
Abstract Patterned deposition of nanoparticles is a prerequisite for the application of unique properties of nanoparticles in future nanodevices. Recent development of nanoxerography requires highly charged aerosol nanoparticles to avoid noise deposition due to random Brownian motion. However, it has been known that it is difficult to charge aerosol nanoparticles with more than two elementary charges. The goal of this work is to develop a simple technique for obtaining highly charged monodisperse aerosol nanoparticles by means of electrospray of colloidal suspension. Highly charged aerosol nanoparticles were produced by electrospraying (ES) and drying colloidal suspensions of monodisperse gold nanoparticles. Size and charge distributions of the resultant particles were measured. We demonstrate that this method successfully charges monodisperse nanoparticles very highly, e.g., 122 elementary charges for 25.0 nm, 23.5 for 10.5 nm, and 4.6 for 4.2 nm. The method described here constitutes a convenient, reliable, and continuous tool for preparing highly charged aerosol nanoparticles from suspensions of nanoparticles produced by either wet chemistry or gas-phase methods. 2005 Elsevier Inc. All rights reserved. Keywords: Charging; Patterning; Monodisperse particles; Electrospray
1. Introduction The application of nanoparticles in electronic, magnetic, or optical devices such as quantum devices, field emission display, single-electron transistors, and data storage devices requires the precise and accurate assembly of nanoparticles in the desired patterns. At the same time, parallel patterning of nanoparticles is needed for practical applications. The recently developed method of nanoxerography [1] utilizes charged particles and charge pattern transferred substrates. In this process, the charging of particles plays an important role in positioning aerosol nanoparticles onto a desired area [2–4]. The charging of aerosol nanoparticles has been also of great importance for the controlled synthesis [5,6] and * Corresponding author. Fax: +82-2-878-2465.
E-mail address: [email protected] (M. Choi). 0021-9797/$ – see front matter 2005 Elsevier Inc. All rights reserved. doi:10.1016/j.jcis.2005.01.078
deposition [7] of aerosol nanoparticles as well as in aerosol measurement systems such as differential mobility analyzers (DMA) [8,9]. Aerosol charging can generally be done by the collision of aerosol particles with ambient ions. Radioactive materials [10], corona discharge [11], ultraviolet rays [12], and soft X-rays [13,14] are used to generate ions and charge aerosol particles into the well-known stationary charging state. However, as particle size becomes smaller, the charging ratio of the particles becomes smaller. Even in a monopolar ion atmosphere, it is difficult to obtain charged nanoparticles carrying over ±2 charges. Therefore, it is very much needed to develop a method to prepare highly charged aerosol nanoparticles for enhancing the deposition efficiency and suppressing the diffusion effect during the nanoparticle patterning processes. It is noted that monodisperse nanoparticles can be effectively prepared by wet chemistry methods
136
J. Suh et al. / Journal of Colloid and Interface Science 287 (2005) 135–140
using surfactants [15]. If we combine the size-controlling capability of wet chemistry method with aerosol technology, we could develop a method to prepare highly charged monodisperse aerosol nanoparticles that are needed for the precise patterning. Electrospray, since it was first reported by Zeleny [16], has been widely used for a variety of applications such as electrostatic coating, fuel supply systems, and crop spraying. It has recently been used for the generation of standard nanometer particles [17] and the production of aerosol nanoparticles for the synthesis of materials [6]. Electrospray of colloidal suspension has been recently utilized for sizing the colloidal nanoparticles by combination of the use of a differential mobility analyzer [18]. It is well known that electrospray synthesis of nanoparticles can generate highly charged droplets and particles. However, this method has not been applied to charging already-made particles. In the present study, we aim to develop a method using the electrospray of colloidal suspension to prepare highly charged monodisperse aerosol nanoparticles and to determine the charge state of the nanoparticles after electrospraying and complete drying. Two issues are answered before the electrospray of colloidal suspension can be claimed as a reliable method for preparing highly charged aerosol nanoparticles. The first is to make sure that the original particle size remains even after electrospraying and drying of nanoparticle suspension. The second is whether the highly charged state of electrosprayed droplets maintains or disappears after the complete evaporation of liquid covered on the surface of the original nanoparticle. A couple of the previous studies on the electrical mobility analysis of the ions produced via electrospray indicated some unexpected peaks that were not corresponding to ion signals and the peaks were considered to be the ones for highly charged aerosol residues [19,20]. Loscertales and Fernández de la Mora [20] analyzed the mobility of residue particles using a differential mobility analyzer (DMA) and a hypersonic impactor in order to investigate ion field evaporation from highly charged droplets after electrospraying. They indirectly showed that aerosol particles were highly charged. Han et al. [21] also showed that impurity nanoparticles could be multiply charged in the process of analyzing the mobility of ion clusters when pure water, tap water, and aqueous solutions were electrosprayed. However, direct morphological and compositional measurements of the sampled particles have not been investigated. Electrospray charging of colloidal suspensions of nanoparticles has not been attempted. The charge state of the electrosprayed and dried nanoparticles needs to be determined depending on the particle size. Here, colloidal suspension of monodisperse gold nanoparticles were electrosprayed and then dried to prepare highly charged monodisperse aerosol nanoparticles. Their charge and size distributions were measured to examine whether the size of the original nanoparticles maintains and highly charged state could be attained.
2. Methods Figs. 1a and 1b show the experimental apparatus used in this study. Monodisperse gold nanoparticle suspensions (0.01% HAuCl4 suspended in water-based solution, G1, G2, and G3, Sigma–Aldrich Chemie Gmbh) were used and the specified nominal sizes of the particles were 5 ± 2 nm for G1, 10 ± 2 nm for G2, and 20 ± 3 nm for G3 with narrow size distributions. Mixtures of the colloidal suspension and methanol (50/50, v/v) were prepared to stabilize the electrospraying condition. It was supplied into a capillary by a highresolution syringe pump (Model 5M361F, ORION). A positive DC voltage of 2–3 kV (Model 205B-10R, BERTAN) was applied to the metal capillary (0.2 mm O.D. and 0.1 mm I.D.) located 1–3 cm from the counterelectrode. A proper voltage-generating straight cone-jet mode was found using a visualization system composed of a CCD camera and a TV monitor. Flow rates of a carrier gas (nitrogen, N2 ) and a liquid precursor are fixed at 2.0 l/min and 10 µl/h, respectively. The proper liquid feed rate was chosen to achieve one particle per one droplet by measuring the size distributions of the resultant particles and comparing these with the nominal size of particles contained in the colloid solution. Highly charged droplets containing gold nanoparticles were formed at the end of the liquid cone and complete solvent evaporation results in highly charged solid aerosol
(a)
(b) Fig. 1. (a) Electrospray of colloidal suspension contained monodisperse particles and drying for highly charged monodisperse aerosol nanoparticles. (b) Measurement system for size (A), electrical mobility (B), and charge (C) distributions.
J. Suh et al. / Journal of Colloid and Interface Science 287 (2005) 135–140
nanoparticles as shown in Fig. 1a. A nanodifferential mobility analyzer (nano-DMA, Model 3085, TSI Inc.), a scanning mobility particle sizer (SMPS, Model 3934, 3936, TSI Inc.), and a Faraday cup electrometer (FCE, Wyckoff Co. Ltd., Japan) were used to measure the sizes and charge distribution of the resultant nanoparticles. Transmission electron microscopy together with electron dispersive spectrometry was also used to confirm size, morphology, and composition of collected particles. All the transport stainless tubes connected to measurement systems were grounded to prevent the highly charged particles from sticking to the tube wall. This measurement system consists of three parts. As shown in Fig. 1b, the size, electrical mobility, and charge distributions of particles are measured through A, B, and C directions, respectively. A polonium (Po-210) neutralizer is added before reaching SMPS to minimize the possibility of incomplete neutralization. Tandem DMA (TDMA) has been chosen to analyze the charge distribution of particles. In this system, gold nanoparticles of equal electrical mobility are classified through a nano-DMA and then they are fed into the SMPS to measure particle size distribution of equal mobility particles. By means of the TDMA system, charge distribution of particles can be simply determined in case of monodisperse particles. Since electrical mobility of a singly charged particle is known, the charge distribution of particles, f (qi ), is calculated as qi =
Zp (Vi ) , Zp (q = 1, dp )
137
(a)
Ni (Vi ) f (qi ) = ∞ , i=1 Ni
where qi is the number of elementary charges, Vi is the voltage of nano-DMA (Model 3085, TSI Inc.), Zp is the electrical mobility at the voltage Vi , Zp (q = 1, dp ) is the electrical mobility of singly charged particles at size dp (cm2 /V s), Ni is the particle number concentration at voltage Vi (particles/cm3 ), and f (qi ) is the normalized number fraction of particles having qi charges.
(b)
3. Results and discussion Fig. 2 shows the size distributions of gold nanoparticles (G1, G2, G3) measured by a neutralizer and an SMPS when they are electrosprayed and dried. As shown in the figure, the geometric mean diameter of the particle is 4.2 nm for G1, 10.4 nm for G2, and 25.0 nm for G3, respectively. In all cases, they show very narrow distributions (σg = 1.1) that can be considered as monodisperse nanoparticles. We measured the electrical mobility distributions of these electrosprayed nanoparticles in order to investigate whether they were highly charged states. Fig. 3 shows the distributions of electrical mobility for G1, G2, and G3 obtained by a nano-DMA (Model 3085, TSI Inc.), FCE, and CPCs (Model 3022A and 3025, TSI Inc.). The abscissa means electrical mobility and the ordinate stands for arbitrary unit of the currents measured by FCE. As shown in the figure, the peak val-
(c) Fig. 2. Size distributions of the gold nanoparticles. (a) G3, (b) G2, (c) G1.
ues of electrical mobility (Zp ) are almost the same regardless of the particle size. It means that the number of elementary charges (q) increases as the particle size increases. The electrical mobility of the particles should be 0.080 cm2 /V s for 4.2 nm, 0.017 cm2 /V s for 10.5 nm, and 0.004 cm2 /V s
138
J. Suh et al. / Journal of Colloid and Interface Science 287 (2005) 135–140
(a) Fig. 3. Mobility distributions of gold nanoparticles.
(b)
Fig. 4. TEM photographs of gold nanoparticles. (a, b) for G3, (c) for G2, (d) for G1.
for 25.0 nm if they are singly charged. However, the mobility of gold nanoparticles prepared by the present method has a significantly higher value (Zp = 0.44 cm2 /V s) than those of the singly charged particles. These results confirm that the gold nanoparticles have been highly charged. To prove whether these particles are real gold particles or solvent residues, TEM analysis was done together with energy dispersive spectrometry. Fig. 4 shows the morphology of the resultant particles. Fig. 4a is the morphology of the particles directly collected after electrospraying G3 without classification and Figs. 4b–4d are the morphologies of
(c) Fig. 5. Charge distributions of the resultant nanoparticles. (a) G1, (b) G2, (c) G3.
the particles having equal mobility (Zp = 0.44 cm2 /V s), classified by nano-DMA, for G3, G2, and G1, respectively. It is clear that the electrosprayed gold particles are almost monodisperse and spherical. The mean sizes of the particles
J. Suh et al. / Journal of Colloid and Interface Science 287 (2005) 135–140
139
Table 1 Electrical properties of electrosprayed and dried gold nanoparticles at Q = 10 µl/h
Fig. 6. Variations of number of elementary charges depending on particle size.
measured by TEM (Dp = 4 nm for G1, 11 nm for G2, 23 nm for G3) are very close to those measured by SMPS. The sizes of gold nanoparticles were also examined directly from colloidal suspensions without electrospraying and they were almost the same as the ones obtained after electrospraying and drying. These particles were examined by the analysis of an energy dispersive spectrometry confirming gold particles. Charge distributions were also measured and the results were shown in Fig. 5. The abscissa describes the number of elementary charge (qi ) of particles and the ordinate is the normalized number concentration of the particles. It is found that the mean numbers of elementary charge are 4.6 for 4.2 nm, 23.5 for 10.5 nm, and 122 for 25.0 nm, respectively. For the monodisperse particles, average charge can also be simply calculated from mobility distribution (see Fig. 3). Dividing the peak mobility by that of singly charged particles gives qi = 4.8 for 4.2 nm, 24 for 10.5 nm, and 128 for 25.0 nm, respectively. These results are almost the same as the values shown in Fig. 5. These results demonstrate that we could easily prepare highly charged monodisperse aerosol nanoparticles from colloidal suspensions by electrospraying and drying. Also, they suggest that the number of elementary charges on the particle increases as the size of the particle becomes larger. The number of charges is approximately proportional to Dp2 . This can be explained by the fact that the electrical mobility, Zp , is proportional to the ratio q/Dp2 in the free molecular limit (Dp 40 nm) for ion evaporation from charged droplets [19,20]. Since the measured electrical mobilities of particles are shown to be independent of particle sizes (see Fig. 3), q would be proportional to Dp2 . Fig. 6 shows that the average number of charges in our experiment is proportional to Dp1.95 . The constant mobility value of our nanoparticles supports the ion evaporation model [19,20]. If the droplets solely ex−1/2 perience Rayleigh fission, their field would be E ∼ dp (here, dp is particle diameter), whereas if droplets emit ions through ion evaporation, the droplets would adjust their ion evaporation rate according to the rate of solvent evaporation; as a result, the electric field, E, could become rela-
Used colloid
Dg a (nm)
G1 G2 G3
4.2 10.5 25.0
Zp b (m2 /V s) 0.44 × 10−4 0.44 × 10−4 0.44 × 10−4
qc (–)
Ed (V/nm)
ER e (V/nm)
E/ER (–)
4.6 23.5 122
1.25 1.18 1.12
3.28 2.07 1.35
0.38 0.57 0.83
Note. Surface tension was measured by DuNouy ring method (digital tensiometer, K9, Kruss). a Geometric mean diameter of particles measured by SMPS (Model 3934, TSI Inc.). b Mobility of particles measured by DMA (Model 3085, TSI Inc.) and FCE (Wyckoff Co. Ltd., Japan). c Number of average charge of particles. d Electric field of particles calculated using E = qe/π ε d 2 (ε = 8.855× 0 p 0 10−12 F/m, e = 1.602 × 10−19 C). e Electric field at Rayleigh limit calculated from E = (8γ /ε d )1/2 and R 0 p based on ε0 = 8.855 × 10−12 F/m, γ = 0.050 N/m.
tively constant [20]. Table 1 shows the electrical properties of the resultant nanoparticles in our experiments. E is quite close to the earlier observation (∼1 V/m) for ion evaporation [19,20] independent of particle size, which is well below the Rayleigh limit (ER ). Decreasing the particle size causes more departure from the Rayleigh limit. Therefore, the gold nanoparticles in our experiments would be saturated with ions at the ion evaporation limit and thus would lose some ions at the surface while drying process occurs. However, final dried gold nanoparticles still attain high charge states that are difficult to obtain by conventional aerosol charging methods.
4. Conclusions We demonstrated that electrospraying of nanoparticle suspensions and subsequent drying produced highly charged aerosol nanoparticles having the same sizes as the original nanoparticles. Measurements of size and charge distributions by using a tandem DMA system and TEM analysis were done and confirmed the validity of the method. The geometric mean numbers of charge are 4.6 for 4.2 nm, 23.5 for 10.5 nm, and 122 for 25.0 nm, respectively. The number of charges is proportional to Dp1.95 . The method described here constitutes a convenient, reliable, and continuous tool for preparing highly charged aerosol nanoparticles from nanoparticles produced by either wet chemistry or gasphase methods.
Acknowledgments This work was funded by the Creative Research Initiatives Program supported by the Ministry of Science and Technology, Korea. J. Suh was partially supported by the BK 21 program funded by the Ministry of Education.
140
J. Suh et al. / Journal of Colloid and Interface Science 287 (2005) 135–140
References [1] H.O. Jacobs, G.M. Whitesides, Science 291 (2001) 1763–1766. [2] T.J. Krinke, H. Fissan, K. Deppert, M.H. Magnusson, L. Samuelson, Appl. Phys. Lett. 78 (2001) 3708–3710. [3] H.O. Jacobs, S.A. Campbell, M.G. Steward, Adv. Mater. 14 (2002) 1553–1557. [4] M. Kang, H. Kim, B. Han, J. Suh, J. Park, M. Choi, Microelectron. Eng. 71 (2004) 229–236. [5] M. Adachi, S. Tsukui, K. Okuyama, Jpn. J. Appl. Phys. 42 (2003) L77–L79. [6] K. Nakaso, B. Han, K.H. Ahn, M. Choi, K. Okuyama, J. Aerosol Sci. 34 (2003) 869–881. [7] T.J. Krinke, K. Deppert, M.H. Magnusson, H. Fissan, Part. Part. Syst. Charact. 19 (2002) 321–326. [8] E.O. Knutson, K.T. Whitby, J. Aerosol Sci. 6 (1975) 443–451. [9] K. Okuyama, M. Shimada, A. Okita, Y. Otani, S.J. Cho, J. Aerosol Res. Jpn. 13 (1998) 83–93. [10] M. Adachi, Y. Kousaka, K. Okuyama, J. Aerosol Sci. 16 (1985) 109– 123.
[11] P. Büscher, A. Schmidt-Ott, A. Wiedensohler, J. Aerosol Sci. 25 (1994) 651–663. [12] M. Shimada, S.-J. Cho, K. Okuyama, T. Tamura, M. Adachi, T. Fujii, J. Aerosol Sci. 28 (1997) 649–661. [13] M. Shimada, B. Han, K. Okuyama, Y. Otani, J. Chem. Eng. Jpn. 35 (2002) 786–793. [14] B. Han, M. Shimada, M. Choi, K. Okuyama, Aerosol Sci. Technol. 37 (2003) 330–341. [15] V.F. Puntes, K.M. Krishnan, A.P. Alivisatos, Science 291 (2001) 2115– 2117. [16] J. Zeleny, Phys. Rev. 3 (1914) 69–91. [17] D.-R. Chen, D.Y.H. Pui, S.L. Kaufman, J. Aerosol Sci. 26 (1995) 963– 977. [18] I.W. Lenggoro, B. Xia, K. Okuyama, Langmuir 18 (2002) 4584–4591. [19] V. Katta, A.L. Rockwood, M.L. Vestal, Int. J. Mass Spectrom. Ion Proc. 103 (1991) 129–148. [20] I.G. Loscertales, J. Fernández de la Mora, J. Chem. Phys. 103 (1995) 5041–5060. [21] B. Han, I.W. Lenggoro, M. Choi, K. Okuyama, Anal. Sci. 19 (2003) 843–851.