Professional Documents
Culture Documents
A R T I C L E I N F O A B S T R A C T
Article history:
Received 3 December 2017 A customized droplet-based microfluidic reactor was fabricated for the synthesis of silver nanoparticles
Received in revised form 25 February 2018 (Ag NPs) using silver nitrate (AgNO3) and branched polyethyleneimine (BPEI) as a precursor and a
Accepted 26 February 2018 reducing agent, respectively. The effects of static mixing, temperature, and the volumetric flow rates of
Available online 7 March 2018 AgNO3 and BPEI on the particle size were investigated. The use of a static mixer and the optimization of
the reaction temperature enhanced the monodispersity of the Ag NPs. In addition, the size of the Ag NPs
Keywords: was manipulated by changing the flow rate ratio of AgNO3 to BPEI at 60 C for 60 min.
Customized microfluidic device © 2018 The Korean Society of Industrial and Engineering Chemistry. Published by Elsevier B.V. All rights
Silver nanoparticle
reserved.
Droplet formation
T-junction
Static mixer
https://doi.org/10.1016/j.jiec.2018.02.040
1226-086X/© 2018 The Korean Society of Industrial and Engineering Chemistry. Published by Elsevier B.V. All rights reserved.
406 C.H. Kwak et al. / Journal of Industrial and Engineering Chemistry 63 (2018) 405–410
requirements, the customized microfluidic reactor, which has connector (T-junction 2, Tee, 1/1600 OD, 0.75 mm ID, VICI,
similar advantages to those of the aforementioned modular Switzerland) was used to generate the droplet. To mix the
microfluidic device, is proposed in this study. solutions, a precolumn filter (PreColumn PEEK 0.5 mm, IDEX,
Recently, a modular microfluidic device suggested the compo- USA) was filled with a polyester porous sponge and used as a static
sition of one microfluidic system through the production and mixer. All reagents were injected using microsyringe pumps
assembly of individual components with specific functions instead (Legato 200, KD Scientific, USA).
of the conventional PDMS-based microfluidic device [13,18,19]. The
advantage of a modular microfluidic device is a simple reconfig- Synthesis of Ag NPs using a droplet-based microfluidic reactor
urable device that enables the rapid assembly and surface-
modification of a specific area. In a typical experiment, silver nitrate (AgNO3, 99.0%) and
In this study, silver nanoparticles (Ag NPs) were synthesized branched polyethyleneimine (BPEI, MW = 750,000, 50 wt% in H2O)
using the customized microfluidic reactor to evaluate the effective were purchased from Sigma–Aldrich Korea (Seoul, Korea), which
synthesis and size control of the nanoparticles in this microfluidic were used as the precursor and a reducing agent, respectively.
system. The customized microfluidic reactor was fabricated easily Aqueous AgNO3 (33.8 mg mL1) and BPEI (5.07 mg mL1) solutions
while allowing the simple customization of the system by were prepared by dissolution in deionized water. For droplet
assembling single functional microfluidic units that are commer- formation, n-decane (Daejung, Korea) was used as a continuous oil
cially available. This microfluidic system overcame the limitations phase. The microfluidic crystallization of Ag NPs was carried out at
of the device reconfiguration in the PDMS-based microfluidic 60 C for 60 min. The reaction was quenched by dropping the
platform. The static mixer and T-junction in this microfluidic reaction product into a 50 mL conical tube filled with isopropyl
reactor were adjusted for mixing and droplet generation, alcohol (IPA). Finally, Ag NPs were collected by centrifugation at
respectively. In addition, fluoropolymer-based micro tubing was 12,000 relative centrifugal force for 40 min and washed three times
used as a microfluidic channel to prevent the deformation induced with IPA to remove the unreacted reactants.
by the organic solvent. The effect of the static mixer, temperature,
and the volumetric flow rates of the reactants on the particle size Characterization
were investigated.
X-ray diffraction (XRD, D/Max-2500, Rigaku, Japan) was carried
Materials and methods out using Cu Ka radiation. The Ag NPs for XRD analysis were
prepared by freeze-drying the colloidal Ag NPs overnight using a
Fabrication of customized droplet-based microfluidic reactor lyophilzer (FDS 8508, Ilshin Biobase, Korea) after removing the
unreacted reactants using centrifugation. Thermogravimetric
The customized microfluidic reactor consisted of three func- analysis (TGA, STA 409 PC/NETZSCH, Germany) was conducted
tional elements: micro tubing (DuPont fluorinated ethylene at 25–800 C at a rate of 10 C min1 using N2 as the carrier gas. The
propylene (FEP) tubing, 1/1600 OD, 0.03000 ID, IDEX, USA), two Tee morphology of the Ag NPs was examined by transmission electron
connectors, and a static mixer. The first Tee connector (T-junction 1, microscopy (TEM, JEM-2100F JEOL, Japan). The purified colloidal
Tee Assembly for 1/1600 OD, 0.02000 thru hole, IDEX, USA) was used Ag NPs (40 mL) were dropped onto a TEM grid (CF200-CU, Electron
for rapid mixing by introducing each of two solutions, an Ag Microscopy Sciences) and dried overnight in a desiccator for TEM
precursor and reducing agent, to the static mixer. The second Tee analysis. The optical properties of Ag NPs were analyzed using a
Fig. 1. Customized droplet-based microfluidic reactor fully integrated by T-junction, static mixer, and heating zone. (a) Schematic diagram illustrating the microfluidic reactor
setup. Photograph of (b) droplets formed using the T-junction 2, and (c) static mixer for homogeneous mixing (scale bars are 1 cm).
C.H. Kwak et al. / Journal of Industrial and Engineering Chemistry 63 (2018) 405–410 407
UV–vis/NIR spectrometer (V-770, Jasco, Japan). The Ag NPs for UV– have synthesized Ag NPs using a microfluidic device based on
vis analysis were collected every 4 min during the reaction up to micro tubing [20–22]. On the other hand, the micro tubing was
28 min. The tubing length was controlled from 55 to 340 cm using used only as a substitute for the PDMS-based device in previous
tubing connectors according to the reaction time. The details of the studies. In contrast, the advantage of using micro tubing for simple
tubing length are provided in Table S1 in Supporting information. reconfigurability was obvious in the present study.
Fig. 2. Characteristics of the Ag NPs synthesized at 60 C for 60 min as an optimal reaction condition. (a) XRD, (b) UV–vis spectra, (c) TEM, and (d) HR-TEM of the as-
synthesized Ag NPs.
408 C.H. Kwak et al. / Journal of Industrial and Engineering Chemistry 63 (2018) 405–410
fouling of Ag NPs in the channel between the T-junction 1 and T- showed non-uniform and aggregated particles. In contrast, the Ag
junction 2 [22]. NPs synthesized using a static mixer (Fig. 3b) were uniform and
well dispersed. Fig. 3c and d shows the corresponding size
Function of static mixer distribution curves. The particle size decreased from
6.79 3.20 nm to 4.37 0.73 nm. The only different reaction
The homogeneous mixing of reactants is a key factor for the condition was the concentration gradient resulted from the mixing
synthesis of Ag NPs with a narrow size distribution. The droplet- efficiency on the use of static mixer. It was obvious that the size of
based microfluidic system enabled the reactants to be mixed Ag NPs was affected by the concentration ratio of reactants. The
homogeneously inside the droplet by not only the diffusion but effect of concentration ratio of AgNO3 to BPEI on the size of Ag NPs
also the convection [28–30]. A shear stress induced a velocity will be explained later. In addition, the coefficient of variation
gradient of the fluid across the channel due to the contact of the decreased from 47.1% to 16.7%. These results suggest that the static
moving droplet with rigid channel wall [31]. Since the droplet was mixer plays a significant role in the synthesis of uniform and well-
immiscible with the continuous phase, the locally fast fluid near dispersed Ag NPs.
the channel axis could not move through the interface. Convective
flows, therefore, were formed to improve mixing efficiency due to Effect of temperature on the crystallization rate
fluids recirculating from the channel axis toward the channel wall
inside droplets. The droplet formation, however, was not enough to The crystallization rate and size uniformity of the Ag NPs were
achieve the monodispersity of the Ag NPs. As discussed above, the investigated according to the temperature (50, 60, and 70 C) and
aqueous solutions of reactants (AgNO3 and BPEI) could form the reaction time from 4 to 28 min at a constant AgNO3 and BPEI flow
stable laminar flow after introduction through T-junction 1 and rate (20 mL min1). The tubing length for the reaction time control
diffuse through the fluid–fluid interface, resulting in a concentra- was adjusted from 55 to 340 cm using a tubing connector. Fig. 4a–c
tion gradient of the reactants. Unfortunately, the concentration presents the UV–vis spectra of the as-synthesized Ag NPs at 50, 60,
gradient is maintained even after the formation of droplets to and 70 C respectively, which were recorded every 4 min for up to
discourage the homogeneous mixing. A mixing efficiency is 28 min. The crystallization rate of Ag NPs at 50 C was too slow to
generally an important factor in chemical reaction because the take one of the advantages of the microfluidic device, which is a
formation of concentration gradient causes non-homogeneity of fast reaction resulting from the convection in a single droplet due
product. In case of a synthesis of nanoparticles, the homogeneity of to the high surface area to volume ratio of the droplet [34]. On the
particles, such as the uniformity of particle size and morphology is other hand, the crystallization rate of the Ag NPs at 70 C was faster,
dependent on the homogeneity of the nucleation, that is, the early so that the reduction of Ag+ reached a maximum at 24 min and the
stage of nanoparticle synthesis [32,33]. For this reason, a static Ag NPs were ripened over 28 min. The optimal crystallization rate
mixer was equipped immediately after the T-junction 1 to provide of the Ag NPs was at 60 C. Crystallization at 50 and 70 C resulted
the homogeneous reaction environment with fast mixing for a in Ag NPs with a wide size distribution. In contrast, the Ag NPs
short time in the nucleation reaction. Fig. 3 presents TEM images of produced at 60 C were monodispersed, as shown in the TEM
the as-synthesized Ag NPs with and without the use of a static images (Fig. 4d–f) and corresponding particle size histograms
mixer. The Ag NPs synthesized without a static mixer (Fig. 3a) (Fig. S4, Supporting information).
Fig. 3. Effect of the static mixer on the size distribution of Ag NPs synthesized at 60 C for 60 min when the ratio of injection flow rate of AgNO3:BPEI was 20:20 mL min1. TEM
images of the as-synthesized Ag NPs (a) without the use of a static mixer, and (b) using a static mixer. (c, d) Corresponding size distribution curves.
C.H. Kwak et al. / Journal of Industrial and Engineering Chemistry 63 (2018) 405–410 409
Fig. 4. Effects of temperature on the crystallization of Ag NPs. UV–vis spectra of the as-synthesized Ag NPs at (a) 50 C, (b) 60 C, and (c) 70 C for every 4 min up to 28 min. TEM
images of Ag NPs synthesized at (d) 50 C, (e) 60 C, and (f) 70 C for 28 min.
Fig. 5. Manipulation of various sizes of Ag NPs using different injection flow rate ratios of AgNO3:BPEI. (a) UV–vis spectra and (b–f) corresponding TEM images when the
injection flow rates were (b) 30:10 (mL min1), (c) 25:15 (mL min1), (d) 20:20 (mL min1), (e) 15:25 (mL min1), and (f) 10:30 (mL min1) (Fig. 5b–d insets show the size
distribution curves).
Effect of injection flow rate on the particle size (Ag0); the mechanism is discussed in detail elsewhere [27,36]. BPEI
also acts as a stabilizer for the synthesis of Ag NPs and prevent
Typically, the high supersaturation of a metal precursor aggregation due to an electrostatic interaction [37]. The synthetic
concentration leads to the formation of smaller nanoparticles condition was fixed to 60 C for 60 min, and the ratios of the
[34]. In addition, a difference in the composition of the reactants volumetric flow rate of AgNO3 and BPEI were 30:10 (mL min1),
causes a change in the shape and size of the nanoparticles [35]. 25:15 (mL min1), 20:20 (mL min1), 15:25 (mL min1), and 10:30
Therefore, the precursor and a reducing agent ratio could be (mL min1). As shown in Fig. 5a, a sharp plasmonic absorption peak
optimized to control the particle shape and size. Fig. 5a presents was observed at 427 nm when the injection flow rate ratios of
the UV–vis spectra and TEM images of the as-synthesized Ag NPs at AgNO3 and BPEI were 30:10 (mL min1), 25:15 (mL min1), and
different AgNO3 and BPEI ratios. Generally, the primary and 20:20 (mL min1). TEM (Fig. 5b–d) revealed uniform spherical Ag
secondary amine groups of BPEI reduced Ag+ to metallic silver NPs with a controlled size. This indicates that uniform-sized Ag
410 C.H. Kwak et al. / Journal of Industrial and Engineering Chemistry 63 (2018) 405–410
NPs were synthesized in an Ag-rich environment. The Ag NPs were The customized microfluidic reactor will be a promising platform
11.45 2.00 nm, 9.96 1.59 nm, and 4.37 0.73 nm in size at for the synthesis of several noble metal nanoparticles providing
AgNO3 and BPEI injection flow rate ratios of 30:10, 25:15, and easy control of the device, low cost, and high chemical resistance.
20:20 (mL min1), respectively. On the other hand, when the ratios
of injection flow rate were 15:25 and 10:30 (mL min1), a broad
Acknowledgment
shoulder peak appeared at approximately 500–700 nm, and
aggregated Ag NPs were observed, as shown in Fig. 5e and f. This
This work was supported by the National Research Foundation
means that non-uniform Ag NPs were generated in the Ag-starved
of Korea (NRF) grant funded by the Korea government (MSIT) (No.
environment because the relatively high concentration of BPEI
2014R1A5A1009799).
allows the Ag NPs to aggregate and grow to larger sizes [38]. It
should be noted that a stable Ag-BPEI complex was formed by the
reaction of AgNO3 with BPEI. It was already mentioned that BPEI Appendix A. Supplementary data
was used as a stabilizer, however, the amount of Ag0 monomer was
insufficient to reach the high level of supersaturation due to the Supplementary data associated with this article can be found, in
formation of Ag-BPEI complex in the Ag-starved environment. the online version, at https://doi.org/10.1016/j.jiec.2018.02.040.
Subsequently, the nucleation rate decreased and the slow
nucleation process was promoted for the long reaction time in References
the Ag-starved environment, resulting in the formation of Ag NPs
with broad size distribution [39,40]. These results clearly show [1] A. Huebner, S. Sharma, M. Srisa-Art, F. Hollfelder, J.B. Edel, A.J. deMello, Lab
Chip 8 (2008) 1244.
that the size of the Ag NPs was controlled easily by changing the [2] J.Q. Boedicker, L. Li, T.R. Kline, R.F. Ismagilov, Lab Chip 8 (2008) 1265.
flow rate of AgNO3 and BPEI. [3] M.D. Tarn, N. Pamme, Expert Rev. Mol. Diagn. 11 (2011) 711.
[4] H.M. Xia, S.Y.M. Wan, C. Shu, Y.T. Chew, Lab Chip 5 (2005) 748.
[5] J.D. Tice, H. Song, A.D. Lyon, R.F. Ismagilov, Langmuir 19 (2003) 9127.
Comparison of the microfluidic and batch systems on the synthesis of [6] C.-X. Zhao, L. He, S.Z. Qiao, A.P.J. Middelberg, Chem. Eng. Sci. 66 (2011) 1463.
Ag NPs [7] T. Thorsen, R.W. Roberts, F.H. Arnold, S.R. Quake, Phys. Rev. Lett. 86 (2001)
4163.
[8] S.L. Anna, N. Bontoux, H.A. Stone, Appl. Phys. Lett. 82 (2003) 364.
The microfluidic reactor system was compared with a batch [9] S. Zeng, B. Li, X.o. Su, J. Qin, B. Lin, Lab Chip 9 (2009) 1340.
system on the synthesis of Ag NPs in terms of the reaction time [10] A.S. Utada, E. Lorenceau, D.R. Link, P.D. Kaplan, H.A. Stone, D.A. Weitz, Science
using UV–vis analysis (Fig. S5a, Supporting information). The 308 (2005) 537.
[11] D. Ogonczyk, J. Wegrzyn, P. Jankowski, B. Dabrowski, P. Garstecki, Lab Chip 10
yellow-colored suspensions of the Ag NPs synthesized using the
(2010) 1324.
customized microfluidic reactor exhibited relatively narrow [12] N.S.G.K. Devaraju, M.A. Unger, Lab Chip 11 (2011) 1962.
absorption peaks centered at l = 427 nm; the peak intensity [13] K.C. Bhargava, B. Thompson, N. Malmstadt, Proc. Natl. Acad. Sci. U. S. A. 111
(2014) 15013.
increased until 60 min and reached a plateau at 70 min. This
[14] K. Ren, J. Zhou, H. Wu, Acc. Chem. Res. 46 (2013) 2396.
suggests that there were no changes in the size and concentration [15] P.H. Hoang, C.T. Nguyen, J. Perumal, D.-P. Kim, Lab Chip 11 (2011) 329.
of Ag NPs after 60 min [41]. On the other hand, the Ag NPs [16] S. Marre, K.F. Jensen, Chem. Soc. Rev. 39 (2010) 1183.
synthesized in the batch system took approximately 10 h (Fig. S5b, [17] S.B. Ali Asgar, J. Preetha, P. Andrea, P. Ian, J. Micromech. Microeng. 17 (2007) 42.
[18] K.G. Lee, K.J. Park, S. Seok, S. Shin, D.H. Kim, J.Y. Park, Y.S. Heo, S.J. Lee, T.J. Lee,
Supporting Information). In particular, the rapid and efficient RSC Adv. 4 (2014) 32876.
mixing of individual droplets in the microfluidic reactor promotes [19] F. Benito-Lopez, M. Antonana-Diez, V.F. Curto, D. Diamond, V. Castro-Lopez,
a short nucleation time and a homogeneous reaction in the limited Lab Chip 14 (2014) 3530.
[20] X.Z. Lin, A.D. Terepka, H. Yang, Nano Lett. 4 (2004) 2227.
space, compared to a batch reaction with stirring, where a longer [21] S.T. He, Y.L. Liu, H. Maeda, J. Nanopart. Res. 10 (2008) 209.
time was needed for a homogeneous reaction [42,43]. Therefore, [22] R. Baber, L. Mazzei, N.T.K. Thanh, A. Gavriilidis, RSC Adv. 5 (2015) 95585.
the synthesis process using the customized microfluidic reactor [23] C. Wang, J. Li, S. Sun, X. Li, G. Wu, Y. Wang, F. Xie, Y. Huang, RSC Adv. 6 (2016)
14016.
system showed a shorter reaction time compared to the batch [24] J. Siegel, O. Kvítek, P. Ulbrich, Z. Kolská, P. Slepi9 cka, V. Švor9
cík, Mater. Lett. 89
system. (2012) 47.
[25] Z. Wang, C. Xu, M. Zhao, C. Zhao, RSC Adv. 4 (2014) 47021.
[26] M.R. Dousti, R.J. Amjad, J. Nanophotonics 9 (2015) 093068.
Conclusion [27] A. Shahzad, W.-S. Kim, T. Yu, RSC Adv. 5 (2015) 28652.
[28] Z.B. Stone, H.A. Stone, Phys. Fluids 17 (2005) 063103.
In this study, Ag NPs were synthesized successfully using a [29] M.R. Bringer, C.J. Gerdts, H. Song, J.D. Tice, R.F. Ismagilov, Philos. Trans. R. Soc. A
362 (2004) 1087.
customized microfluidic reactor. The microfluidic reactor consisted
[30] N. Lorber, F. Sarrazin, P. Guillot, P. Panizza, A. Colin, B. Pavageau, C. Hany, P.
of micro tubing as microchannels, Tee connectors for the Maestro, S. Marre, T. Delclos, C. Aymonier, P. Subra, L. Prat, C. Gourdon, E.
construction of T-junctions, and a static mixer for homogeneous Mignard, Lab Chip 11 (2011) 779.
mixing in a moment. The Ag NPs were typically monodispersed [31] Y. Li, R.K. Reddy, C.S.S.R. Kumar, K. Nandakumar, Biomicrofluidics 8 (2014)
054125.
and their sizes were manipulated from 4.37 to 11.45 nm by [32] A. Knauer, A. Thete, S. Li, H. Romanus, A. Csáki, W. Fritzsche, J.M. Köhler, Chem.
changing the volumetric flow rate ratios of AgNO3:BPEI from 20:20 Eng. J. 166 (2011) 1164.
to 30:10 mL min1 at an optimized temperature of 60 C for 60 min [33] Y. Song, J. Hormes, C.S.S.R. Kumar, Small 4 (2008) 698.
[34] N. Cathcart, A.J. Frank, V. Kitaev, Chem. Commun. (2009) 7170.
of reaction. The customized microfluidic reactor had the following [35] K. Dongjo, J. Sunho, M. Jooho, Nanotechnology 17 (2006) 4019.
advantages: (i) the reactor was very easy and flexible to fabricate [36] M. Chen, Y.-G. Feng, X. Wang, T.-C. Li, J.-Y. Zhang, D.-J. Qian, Langmuir 23 (2007)
because all components were removable and reconstructible, and 5296.
[37] Z. Liu, Y. Wang, Y. Zu, Y. Fu, N. Li, N. Guo, R. Liu, Y. Zhang, Mater. Sci. Eng. C 42
(ii) the size of the Ag NPs was controlled precisely by simply (2014) 31.
changing the injection flow rate of the precursor and reducing [38] D.L. Van Hyning, W.G. Klemperer, C.F. Zukoski, Langmuir 17 (2001) 3120.
agent. A device with complicated 3D assembled structure, [39] N.T.K. Thanh, N. Maclean, S. Mahiddine, Chem. Rev. 114 (2014) 7610.
[40] T. Pedro, M. María del Puerto, V.-V. Sabino, G.-C. Teresita, J.S. Carlos, J. Phys. D:
however, is difficult to fabricate using commercial products Appl. Phys. 36 (2003) R182.
because most of them were designed to connect linearly to each [41] S. Agnihotri, S. Mukherji, S. Mukherji, RSC Adv. 4 (2014) 3974.
other. Therefore, the microfluidic reactor proposed in this study is [42] S. Sharma, R.A. Maurya, K.-I. Min, G.-Y. Jeong, D.-P. Kim, Angew Chem. Int. Ed.
52 (2013) 7564.
suitable for material synthesis applications requiring temperature
[43] Y. Wang, X. Zhang, A. Wang, X. Li, G. Wang, L. Zhao, Chem. Eng. J. 235 (2014) 191.
control and mixing with a range of concentrations of chemicals.