Size-Controlled Flow Synthesis of Gold Nanoparticles Using a

Apr 4, 2012 - Segmented flow is often used in the synthesis of nanomaterials to achieve narrow particle size distribution. The narrowness of the distr...
1 downloads 9 Views 4MB Size
Download PDF
Article pubs.acs.org/Langmuir

Size-Controlled Flow Synthesis of Gold Nanoparticles Using a Segmented Flow Microfluidic Platform Victor Sebastian Cabeza,†,‡ Simon Kuhn,† Amol A. Kulkarni,§ and Klavs F. Jensen* Department of Chemical Engineering, Massachusetts Institute of Technology, 77 Massachusetts Avenue, Cambridge, Massachusetts, United States S Supporting Information *

ABSTRACT: Segmented flow is often used in the synthesis of nanomaterials to achieve narrow particle size distribution. The narrowness of the distribution is commonly attributed to the reduced dispersion associated with segmented flows. On the basis of the analysis of flow fields and the resulting particle size distribution, we demonstrate that it is the slip velocity between the two fluids and internal mixing in the continuous-phase slugs that govern the nature of the particle size distribution. The reduction in the axial dispersion has less impact on particle growth and hence on the particle size distribution. Synthesis of gold nanoparticles from HAuCl4 with rapid reduction by NaBH4 serves as a model system. Rapid reduction yields gold nuclei, which grow by agglomeration, and it is controlled by the interaction of the nuclei with local flow. Thus, the difference in the physical properties of the two phases and the inlet flow rates ultimately control the particle growth. Hence, a careful choice of continuous and dispersed phases is necessary to control the nanoparticle size and size distribution.

■

INTRODUCTION The difficulty of preparing nanoparticles in a controlled and reproducible manner represents an obstacle to the exploitation of many nanoscale phenomena.1,2 As novel applications for nanoparticles continue to emerge, there is an increasing need for approaches to manufacture these materials using inexpensive, rapid, and reproducible methods that have minimal environmental impact.3−5 Among the many metal nanoparticles, gold nanoparticles (AuNPs) possess size- and shape-dependent properties rendering them useful in a wide range of applications from bionanotechnology to chemical catalysis.6−8 However, the spatiotemporal variability of reaction conditions (e.g., local mixing rate) in conventional synthesis devices challenges the reproducible preparation of monodisperse gold clusters.9 Conventionally, protective ligand-stabilized metal clusters are prepared by reducing the corresponding metal ion precursor in the presence of stabilizing ligands in batch reactors.6 Under such conditions, any variation in local conditions (e.g., temperature and precursor concentration) across the vessels influences the kinetics of cluster growth and results in polydispersity in the size distribution. Consequently, when a strong reducing agent, such as NaBH4, is used to prepare small clusters, inhomogeneous mixing of the reactants will be the main cause of polydispersity. One approach to diminish the size distribution of the clusters generated under such diffusion-controlled conditions © 2012 American Chemical Society

is to mix the metal ions and reductants homogeneously on a molecular level using microfluidic technology,10,11 as demonstrated in several studies with control over the size distribution and crystal phase of the resulting nanocrystals,3−5,12 taking advantage of the improved heat and mass transport within the microfluidic channels.13 Although fast reactions can be easily accomplished in a microfluidic reactor (MFR) due to the reduced characteristic time of the transport processes at the micrometer length scale,14 the relative values of the kinetic rate and the rate of mixing, together with axial dispersion in singlephase flow, decide the extent of polydispersity in the resulting nanoparticle sizes.13,15,16 Flow segmentation was proposed to generate uniformly sized nanoparticles because of the rapid mixing and efficient mass transfer in discrete slugs by internal circulation.4,17 Nevertheless, a microscale study of the nanofluidic conditions promoting the monodispersed syntheses of metal nanocrystals has not been reported yet. In this context, here we probe the efficiency of using a segmented continuous MFR platform to precisely tune the mixing-reaction conditions in gold nanocrystal flow syntheses. We found that the growth and particle Received: December 28, 2011 Revised: April 3, 2012 Published: April 4, 2012 7007

dx.doi.org/10.1021/la205131e | Langmuir 2012, 28, 7007−7013

Langmuir

Article

size distribution of gold nanocrystals is strongly dominated by the mixing conditions induced by internal recirculation, and not only by controlling the axial dispersion in segmented flow. This is consistent with a fast kinetic chemical reaction where there is a rapid and complete conversion of the gold precursor into Au(0) in the form of gold nuclei, followed by particle growth exclusively via coalescence of the nuclei into bigger particles.18 By injecting a suitable inert fluid of lower viscosity as a disperse phase, we can precisely control the particle size distribution of gold NPs to obtain monodispersed gold colloids within the time scale of 10 s.

■

EXPERIMENTAL SECTION

Silicon Fabricated Microreactors. Figure 1 illustrates the silicon/Pyrex MFR selected in this work to synthesize monodisperse gold nanocrystals, which was prepared as previously reported.10 The reactor consisted of two zones (mixing-cold; reaction-hot) separated by a thermally isolating halo etch that allowed for a temperature

Figure 3. TEM image of gold NPs synthesized in toluene−aqueous segmented flow, [Au] = 1 mM: (a) Rt = 40 s, (b) Rt = 20 s, (c) Rt = 10 s. (d) Particle size distribution diagram from AuNPs obtained in toluene−aqueous segemented flow at Rt = 40, 20, and 10 s, [Au] = 1 mM. Inset: gold NPs obtained in the toluene−aqueous phase, Rt = 10 s. Bottom: segmented slugs generated at Rt = 10 and 40 s (fluorescein was added to improve the optical resolution).

Table 1. Fluid Properties of Air, Toluene, and Silicone Oil at Room Temperaturea

Figure 1. (a) Schematic of the segmented flow generation in a hydrophilic MFR. (b) Detail of segmented slugs generated at a residence time of 10 s (hydrophilic MFR). The aqueous slug (continuous phase) contains fluorescein (green color); the disperse phase is toluene. (c) Spiral silicon/Pyrex microfluidic reactor designed for gold nanocrystal synthesis (400 μm channel width and depth, 100 μL reaction zone volume).

air toluene silicone oil

density (kg/m3)

viscosity (Pa·s)

surface tension (N/m)

1.205 866.9 920.0

1.88 × 10−5 5.90 × 10−4 4.87 × 10−2

0.0288 0.0021 0.0089

a

The surface tension denotes the interfacial tension between the respective phase and the water−surfactant mixture.

gradient of over 25 °C mm−1. The mixing zone was maintained at room temperature using a recirculating coolant flow to avoid nonhomogenous nucleation. The reaction zone (volume 100 μL, channel cross-section 400 μm × 400 μm) was heated to a temperature of 100 °C. Three inert fluids with different physical properties (air, toluene, and silicone oil) were injected as discrete droplets into a flowing carrier aqueous phase composed of the gold precursor and the reductant. Three syringe pumps (Harvard Apparatus HP 2000) and a high-pressure pump (Teledine Isco) were used for fluid injection. All connections and tubes were made of poly(ether ether ketone) (PEEK) and Teflon. The microchannel walls were covered by a hydrophilic SiO2 layer. Therefore, the aqueous phase used as the carrier preferentially wets the microchannel walls. We investigated the effect of the inert fluid properties and residence time (Rt) on gold nanoparticle synthesis. The effect of axial dispersion on the gold nanoparticle size distribution was studied by coating the microchannels with a poly(tetrafluoroethylene)

Figure 2. Segmented slugs generated in (a) the hydrophilic microchannel (water is the carrier phase) and (b) the hydrophobic microchannel (water is the dispersed phase). Fluorescein was added to the aqueous phase to improve the optical resolution. 7008

dx.doi.org/10.1021/la205131e | Langmuir 2012, 28, 7007−7013

Langmuir

Article

Figure 4. Velocity and internal recirculation of the water phase for toluene−water two-phase flow at residence times of 40, 20, and 10 s. The flow direction is from right to left. (PTFE) hydrophobic layer,19 resulting in the aqueous solution being the dispersed phase and thus limiting the axial dispersion (see Figure 2). Reagents and Analysis. The reactants were introduced in three separated currents (Figure 1). The first stream was the reducing agent (aqueous phase), which consisted of a 3.8 mM aqueous solution of sodium borohydride (NaBH4; Aldrich). The second stream was the gold precursor (aqueous phase), which was prepared by mixing a 1 mM aqueous solution of chloroauric acid (HAuCl4; Aldrich) with tetradecyltrimethylammonium bromide (TTABr, C14H29N(CH3)3+Br−; Aldrich), 20 mM. The third inlet was for the segmenting fluid, which was air, toluene (Aldrich), or silicone oil (Alfa Aesar). A gold precursor/ reducing agent flow ratio of 10 was maintained constant for all experimental conditions. Slug flow was obtained by the injection of equal flows of the aqueous and segmented fluids. For the synthesis of gold nanoparticles at a residence time of 10 s, flows of 27, 273, and 300 μL/min were selected for the reducing agent, gold precursor, and segmented fluid,

respectively. Sample collection was performed using an ice bath and a highpressure regulator. In the case of toluene or silicone oil representing the inert phase, separation of the aqueous phase was achieved by decantation. Transmission electron microscopy (TEM) images were taken with a JEOL JEM 200-CX transmision electron microscope. Particle size distribution diagrams were obtained by counting all particles on each TEM picture, considering a particle population higher than 300. Air, toluene, silicone oil, and the aqueous phase (TTAB + water) were used to create the multiphase flow for the residence time distribution measurements using 0.1 wt % sodium benzoate dissolved in water as a pulse-injected tracer. The absorption of this tracer at a wavelength of 269 nm was recorded via an Ocean Optics HR2000CG detector as a function of time. This measurement was made possible by a Z-type flow cell downstream of the microreactor. Particle image velocimetry measurements were realized using an inverted fluorescence microscope (Zeiss Axiovert 200). The light source was provided by a frequency-doubled Nd:YAG laser (BigSky Ultra CFR, 30 mJ, 532 nm), 7009

dx.doi.org/10.1021/la205131e | Langmuir 2012, 28, 7007−7013

Langmuir

Article

and the images were recorded using a dual-frame charge-coupled device (CCD) camera (PCO Sensicam QE, 1376 × 1024 pixels2, 8 bits). The interrogation area for the cross-correlation was 64 × 64 pixels2 with 50% overlap.

■

RESULTS AND DISCUSSION Particle Polydispersity. The influence of the flow rate (residence time) on the particle size distribution is shown in Figure 3. An increase in the residence time was found to broaden the particle size distribution independent of the inert fluid dispersed in the aqueous phase. With toluene as the inert fluid, particle sizes of 3.8 ± 0.3, 4.6 ± 2.1, and 4.9 ± 3.0 nm are obtained at residence times of 10, 20, and 40 s, respectively. The slug (in segmented flows it is conventional that the segments of the fluids are called slugs; the continuous-phase slugs are connected through a thin film, while the dispersed-phase slugs are discrete) length of the continuous phase is influenced by the residence time, obtaining small slugs at smaller residence times (Figure 3, bottom). The slug length is an important hydrodynamic parameter, along with the slip velocity; they have a very significant effect on interslug mass transfer in segmented flow.20 It can be inferred from these results that good control of the mass transfer between the aqueous slugs is a strict requirement to decrease the particle polydispersity. Film Thickness and Slug Length Calculations. On the basis of these results, we investigated the influence of the dispersed-phase fluid on the resulting two-phase flow properties. Air, toluene, and silicone oil (aqueous immiscible fluids) were selected for proof of concept because of their different physical properties, especially viscosity and surface tension (see Table 1). The slug flow observed in microchannel reactors strongly depends upon the relative flow rates of the two phases, the channel dimension, the surface wettability, and the physicochemical properties of the fluids. Typically, the size of the dispersed-phase slugs (which are disconnected) depends upon the dispersed-phase detachment dynamics, while the continuous phase is characterized by slugs connected with each other through a thin liquid film. This liquid film connects the continuous-phase slugs and hence also leads to interslug mass transfer. The film thickness is a function of the fluid velocity, interfacial tension, and viscosity of the fluid. A smaller film thickness is achieved at very low values of velocity and/or viscosity and/ or at higher interfacial tension. The extent of fluid transport that can happen through a smaller film will be much less than that of a larger film. The interslug mass transfer (when the reacting phase is the continuous phase) decides the extent of axial dispersion in the reactor. The higher the extent of interslug mass transfer, the wider the particle size distribution. For the systems where the reactant phase is dispersed in discontinuous slugs (nonwetting phase), the film thickness of the inert continuous phase will decide the length and diameter of the reactant-phase slug. The velocity difference between the two phases and the slug length govern the nature of internal circulation in the slug and hence the particle size. Thus, a higher slip velocity will yield a narrow particle size distribution, while a lower slip velocity will lead to poor mixing inside the slugs, which as a consequence leads to a wider particle size distribution. In the following, we will discuss the influence of the wall wettability on the film thickness and the dispersed slug length (for details of the calculation, see the Supporting Information).

Figure 5. TEM image of gold NPs synthesized at Rt = 10 s in a hydrophilic reactor: (a) silicone oil−aqueous segmented flow, (b) toluene−aqueous segmented flow, (c) air−aqueous segmented flow. (d) Particle size distribution diagram from AuNPs obtained in silicone oil/toluene/air−aqueous segmented flow at Rt = 10 s, [Au] = 1 mM.

For the case of the hydrophilic microchannel, the estimated film thickness based on the viscosity of the continuous phase (i.e., the aqueous reacting phase) and the interfacial tension between the different phases was seen to increase with the Weber number (We, the relative effect of convective and surface tension forces) (see the Supporting Information, text and Figure S1). The trend indicated that, at identical We, the film thickness would increase in the order of air−water, Si oil−water, and toluene−water. On the other hand, for the hydrophobic channel, the film thickness for the silicone oil−water system was maximum (Figure S1). Upon estimation of the dispersed-phase slug lengths (L) based on the film thickness and the phase holdup, while the value of L continued to increase with the residence time in the reactor, it was seen that there was no effect of the surface wettability on the slug length for the air−water system (Figure S2, Supporting Information). The hydrophobic reactor always yielded longer slugs of the dispersed aqueous reacting phase. This was primarily due to the higher film thickness compared to that of the aqueous continuous phase in the hydrophilic reactor. Particle Image Velocimetry. As discussed above, segmented multiphase flows in microchannels are characterized by an internal circulation in the dispersed phase and a recirculation zone in the continuous phase, induced by the slip between the phases.21 Using particle image velocimetry (PIV), we characterize the recirculation zone in the two-phase flow of toluene− water at residence times of 40, 20, and 10 s.22,23 The strength of the recirculation, characterized by the vorticity, is found to increase with decreasing residence time and slug length (Figure 4). The vorticity increases more in the lower portion of the slug since the measurements are not performed in a straight channel (Figure 1). The local channel curvature induces centrifugal forces which result in stronger vorticity toward the wall. These differences in vorticity subject to residence time correspond to the findings for the particle size distributions (Figure 3d). At a residence time of 40 s hardly any internal mixing is visible in the 7010

dx.doi.org/10.1021/la205131e | Langmuir 2012, 28, 7007−7013

Langmuir

Article

Figure 6. Velocity and internal recirculation (charactericed by the vorticity) of the water phase for the air−water, toluene−water, and silicone oil− water two-phase flow at a residence time of 10 s. The flow direction is from right to left.

silicone oil segmented the aqueous flow (Figure 5a,d). In the case of toluene the distribution is monodisperse (3.8 ± 0.3 nm, Figure 5b,d), and air provides the best internal mixing within the continuous phase, resulting in a narrow monodisperse distribution of gold NPs (2.8 ± 0.2 nm, Figure 5c,d). Considering the practical aspects of monodispersed gold nanoparticle production, air is more convenient than toluene due to the straightforward phase separation, and additionally, particle loss, by capture at the phase interface, is reduced. To further elucidate the differences in flow patterns depending on the inert fluid, we investigated the two-phase flow of air−water and silicone oil−water at a residence time of 10 s

PIV results, and consequently, a broad size distribution is obtained. For a residence time of 10 s distinct regions of increased internal mixing are found in the continuous water phase between the dispersed toluene slugs, resulting in a narrow particle distribution of the nanoparticles. On the basis of these results, we further investigated the influence of the dispersed-phase fluid on the resulting fluid circulation within the continuous phase. The resulting particle size distributions of the gold nanoparticles obtained at a residence time of 10 s show a significant dependence on the choice of the inert fluid (Figure 5d). A bimodal particle size distribution (7.8 ± 6.5 and 15.5 ± 3.1 nm) was obtained when 7011

dx.doi.org/10.1021/la205131e | Langmuir 2012, 28, 7007−7013

Langmuir

Article

with PIV (Figure 6). At identical flow rate and for a given continuous phase (the reacting aqueous phase in this case), an increase in the viscosity of the dispersed phase leads to a reduction in the slip velocity, which reduces the magnitude of recirculation and hence yields a broader particle size distribution, which is observed for silicone oil−water (see Figures 5d and 6c). For air−water, where the mobility of the phase interface is relatively increased, higher circulation rates help to achieve smaller particles as well as a narrow particle size distribution (see Figures 5d and 6a). For the case of toluene− water, the interfacial tension is the lowest. Together with the comparably higher interface rigidity and the relatively smaller viscosity, this case falls between the two extreme cases discussed above, and an intermediate internal circulation rate is observed (see Figures 5d and 6b). Though the vorticity distribution between toluene−water and silicone oil−water looks similar, it has to be noted that for silicone oil−water (Figure 6c) the maximum of vorticity is concentrated close to the walls in the vicinity of the slug, whereas for toluene−water (Figure 6b) increased vorticity is also found at the phase interface. Furthermore, the region affected by elevated vorticity extends well into the continuous slug and thus influences the mixing in the bulk, which is also verified by the fact that the spatially averaged vorticity is higher for the toluene−water case. In conclusion, the higher viscosity of the dispersed phase yields a reduced slip velocity and reduced circulation rates, leading to a wider size distribution. On the other hand, the relatively better surface mobility of air−water helps to enhance the internal circulation rates and thereby helps in attaining smaller particle sizes and a narrow particle size distribution. Residence Time Distributions. To further elucidate the role of axial dispersion resulting from different fluids in segmented flow, we measured the residence time distribution (RTD) of the three multiphase systems in the microreactor for a nominal residence time of 20 s.24,25 Therefore, we detect a pulse-injected tracer (sodium benzoate dissolved in water) by applying UV−vis spectroscopy, and the postprocessed RTD curves are depicted in Figure 7. It is observed that the RTD curves do only change marginally with a change of the multiphase flow system. To further analyze the experimental RTD data, we apply a one-dimensional axial dispersion model of the form E(θ) =

⎡ (1 − θ)2 ⎤ 1 exp⎢ ⎥ ⎣ 4θD/uL ⎦ 4πθD/uL

Figure 7. Residence time distribution curve in air−water, silicone oil− water, and toluene−water systems.

Table 2. Mean Residence Times and Vessel Dispersion Numbers for the Three Multiphase Flow Systems air−water toluene−water silicone oil−water

mean residence time (s)

vessel dispersion number

38.34 36.41 32.76

0.028 0.022 0.014

gold nanoparticles in the hydrophobic reactor, where organic fluids are the continuous phase, with the aqueous phase being dispersed. Thus, axial dispersion is greatly reduced, and each individual aqueous segment acts like a disconnected reacting zone (similar to a stand-alone stirred vessel) with no communication between different aqueous slugs.19 The capillary number for silicone oil as the continuous phase is much higher than that of toluene as the continuous phase, resulting in a much higher film thickness for silicone oil compared to toluene (see Figure S1, Supporting Information). Thus, the higher film thickness for the hydrophobic reactor along with the relatively higher viscosity of the fluids will reduce the slip velocity and hence also the internal circulation in the aqueous-phase slug, thereby yielding a wider size distribution (see Figure S3, Supporting Information). On the contrary, for the air−water system, water is still the continuous phase. However, with a hydrophobic wall, water will experience a certain nonzero slip at the wall, which strongly depends upon the extent of the hydrophobic surface roughness.26 This slip at the wall is accompanied by a reduced shear rate close to the wall, which then reduces the extent of circulation in the continuous phase. Consequently, using the hydrophobic reactor, the mean particle size is greater than in the hydrophilic case. In addition, when using silicone oil or toluene, the internal circulation in the reacting dispersed aqueous phase is poor, and particles following different streamlines in the slug will experience different circulation rates, leading to a polydispersed particle size distribution.

(1)

where D/uL is the vessel dispersion number, with the dispersion coefficient D, the superficial velocity of the water segments u, and the length of the reactor L. The nondimensional time θ is obtained by dividing the measurement time by the mean residence time. For a given flow rate and microreactor geometry, the only parameter in eq 1 is the dispersion coefficient D, and thus, using this value, the equation can be fitted to the experimental RTD curves. The resulting dispersion coefficients are tabulated in Table 2, and again no major differences in axial dispersion are found. Thus, the observed findings are purely related to the internal mixing efficiency, and axial dispersion plays only a minor role in the synthesis of nanoparticles in segmented flow. The above discussion implies that although for a segmented flow the axial dispersion is low, the effective contribution of the segmentation to the particle sizes and the size distribution depends on the slip velocity between the phases. Role of Dispersion in Hydrophobic Microchannels. The effect of dispersion is investigated further by synthesizing

■

CONCLUSIONS These results conclusively show that for nanoparticle synthesis in two-phase flow systems, where the segmentation helps in reducing the axial dispersion, the mean particle diameter and the width of the particle size distribution are primarily governed by the slip velocity (between the two fluids for the no-slip boundary as well as the boundary with slip) and the internal mixing in the slugs. In addition, our results are in full agreement 7012

dx.doi.org/10.1021/la205131e | Langmuir 2012, 28, 7007−7013

Langmuir

Article

(5) Chan, E. M.; Mathies, R. A.; Alivisatos, A. P. Size-Controlled Growth of CdSe Nanocrystals in Microfluidic Reactors. Nano Lett. 2003, 3, 199−201. (6) Daniel, M. C.; Astruc, D. Gold Nanoparticles: Assembly, Supramolecular Chemistry, Quantum-Size-Related Properties, and Applications toward Biology, Catalysis, and Nanotechnology. Chem. Rev. 2004, 104, 293−346. (7) El-Sayed, M. A. Some Interesting Properties of Metals Confined in Time and Nanometer Space of Different Shapes. Acc. Chem. Res. 2001, 34, 257−264. (8) Murphy, C. J.; San, T. K.; Gole, A. M.; Orendorff, C. J.; Gao, J. X.; Gou, L.; Hunyadi, S. E.; Li, T. Anisotropic Metal Nanoparticles: Synthesis, Assembly, and Optical Applications. J. Phys. Chem. B 2005, 109, 13857−13870. (9) Shalom, D.; Wootton, R. C. R.; Winkle, R. F.; Cottam, B. F.; Vilar, R.; deMello, A. J.; Wilde, C. P. Synthesis of Thiol Functionalized Gold Nanoparticles Using a Continuous Flow Microfluidic Reactor. Mater. Lett. 2007, 61, 1146−1150. (10) Marre, S.; Jensen, K. F. Synthesis of Micro and Nanostructures in Microfluidic Systems. Chem. Soc. Rev. 2010, 39, 1183−1202. (11) Abou-Hassan, A.; Sandre, O.; Cabuil, V. Microfluidics in Inorganic Chemistry. Angew. Chem., Int. Ed. 2010, 49, 6268−6286. (12) Zhao, C.-X.; He, L.; Qiao, S. Z.; Middelberg, A. P. J. Nanoparticle Synthesis in Microreactors. Chem. Eng. Sci. 2011, 66, 1463−1479. (13) Wagner, J.; Kohler, J. M. Continuous Synthesis of Gold Nanoparticles in a Microreactor. Nano Lett. 2005, 5, 685−691. (14) Jensen, K. F. Microreaction EngineeringIs Small Better? Chem. Eng. Sci. 2001, 56, 293−303. (15) Wagner, J.; Tshikhudo, T. R.; Koehler, J. M. Microfluidic Generation of Metal Nanoparticles by Borohydride Reduction. Chem. Eng. J. 2008, 135, S104−S109. (16) Tsunoyama, H.; Ichikuni, N.; Tsukuda, T. Microfluidic Synthesis and Catalytic Application of PVP-Stabilized, ∼1 nm Gold Clusters. Langmuir 2008, 24, 11327−11330. (17) Kreutzer, M. T.; Gunther, A.; Jensen, K. F. Sample Dispersion for Segmented Flow in Microchannels with Rectangular Cross Section. Anal. Chem. 2008, 80, 1558−1567. (18) Polte, J.; Erler, R.; Thunemann, A. F.; Sokolov, S.; Ahner, T. T.; Rademann, K.; Emmerling, F.; Kraehnert, R. Nucleation and Growth of Gold Nanoparticles Studied via in Situ Small Angle X-ray Scattering at Millisecond Time Resolution. ACS Nano 2010, 4, 1076−1082. (19) Kuhn, S.; Hartman, R. L.; Sultana, M.; Nagy, K. D.; Jensen, K. F. Teflon Coated Silicon Microreactors: Impact on Segmented LiquidLiquid Multiphase Flows. Langmuir 2011, 27, 6519−6527. (20) Liu, H.; Vandu, C. O.; Krishna, R. Hydrodynamics of Taylor Flow in Vertical Capillaries: Flow Regimes, Bubble Rise Velocity, Liquid Slug Length, and Pressure Drop. Ind. Eng. Chem. Res. 2005, 44, 4884−4897. (21) Gunther, A.; Jhunjhunwala, M.; Thalmann, M.; Schmidt, M. A.; Jensen, K. F. Micromixing of Miscible Liquids in Segmented GasLiquid Flow. Langmuir 2005, 21, 1547−1555. (22) Meinhart, C. D.; Wereley, S. T.; Santiago, J. G. PIV Measurements of a Microchannel Flow. Exp. Fluids 1999, 27, 414−419. (23) Santiago, J. G.; Wereley, S. T.; Meinhart, C. D.; Beebe, D. J.; Adrian, R. J. A Particle Image Velocimetry System for Microfluidics. Exp. Fluids 1998, 25, 316−319. (24) Trachsel, F.; Gunther, A.; Khan, S.; Jensen, K. F. Measurement of Residence Time Distribution in Microfluidic Systems. Chem. Eng. Sci. 2005, 60, 5729−5737. (25) Kulkarni, A. A.; Kalyani, V. S. Two-Phase Flow in Minichannels: Hydrodynamics, Pressure Drop, and Residence Time Distribution. Ind. Eng. Chem. Res. 2009, 48, 8193−8204. (26) Rothstein, J. P. Slip on Superhydrophobic Surfaces. Annu. Rev. Fluid Mech. 2010, 42, 89−109.

with the in situ small-angle X-ray scattering measurements already reported,18 where the first step in gold NP synthesis is the initial and rapid reduction of the gold precursor into gold nuclei within less than 200 ms. In the second step, the nuclei grow due to coalescence, decreasing the number of particles but increasing the particle size. In the present study we found that, with increasing residence time at a constant temperature, the size and polydispersity of gold nanoparticles increased as well. Thus, we conclude that it is reasonable to propose that the gold NP growth in our system follows a nucleation−growth by coalescence process. This assumes that the slug mixing and the interaction of gold nuclei between different flow streamlines govern the gold nuclei coalescence and therefore the particle size distribution. Segmented flow microfluidic reactors are a valuable platform to improve the mixing in fast kinetic chemical reactions such as nanocrystal growth, where an excellent control of reactant diffusion is required to obtain monodiperse distributions of nanoparticles; however, a careful choice of continuous and dispersed phases is useful.

■

ASSOCIATED CONTENT

S Supporting Information *

Details about two-phase Taylor flow characterization, film thickness and slug length calculations, and dispersion in hydrophobic microchannels. This material is available free of charge via the Internet at http://pubs.acs.org/.

■

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. Present Addresses ‡

Instituto Universitario de Nanociencia de Aragón, Universidad de Zaragoza, Zaragoza, Spain. § Chemical Engineering Division, National Chemical Laboratory, Pune, India. Author Contributions †

These authors contributed equally to this work.

Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS This work was supported in part by CHE-0714189. V.S.C. acknowledges the support of the Fulbright Commission and the Ministry of Education in Spain (Programa Nacional de Movilidad de RR.HH. del Plan Nacional de I+D+I 20082011). S.K. acknowledges funding from the Swiss National Science Foundation (SNF). A.A.K. acknowledges funding from the Indo-US Science and Technology Forum.

■

REFERENCES

(1) Dushkin, C. D.; Saita, S.; Yoshie, K.; Yamaguchi, Y. The Kinetics of Growth of Semiconductor Nanocrystals in a Hot Amphiphile Matrix. Adv. Colloid Interface Sci. 2000, 88, 37−78. (2) Donega, C. D.; Hickey, S. G.; Wuister, S. F.; Vanmaekelbergh, D.; Meijerink, A. Single-Step Synthesis to Control the Photoluminescence Quantum Yield and Size Dispersion of CdSe Nanocrystals. J. Phys. Chem. B 2003, 107, 489−496. (3) Krishnadasan, S.; Brown, R. J. C.; deMello, A. J.; deMello, J. C. Intelligent Routes to the Controlled Synthesis of Nanoparticles. Lab Chip 2007, 7, 1434−1441. (4) Khan, S. A.; Gunther, A.; Schmidt, M. A.; Jensen, K. F. Microfluidic Synthesis of Colloidal Silica. Langmuir 2004, 20, 8604−8611. 7013

dx.doi.org/10.1021/la205131e | Langmuir 2012, 28, 7007−7013