Microfluidic System Enabling Multistep Tuning of ... - ACS Publications

May 6, 2016 - continuous aqueous phase to droplets of 1-octanol with a diameter of ∼40 μm. ..... aqueous solution of 0.05 mM RB (containing 0, 0.2,...
0 downloads 0 Views 4MB Size
Download PDF
Technical Note pubs.acs.org/ac

Microfluidic System Enabling Multistep Tuning of Extraction Time Periods for Kinetic Analysis of Droplet-Based Liquid−Liquid Extraction Natsuki Nakajima, Masumi Yamada,* Shunta Kakegawa, and Minoru Seki Department of Applied Chemistry and Biotechnology, Graduate School of Engineering, Chiba University, 1-33 Yayoi-cho, Inage-ku, Chiba 263-8522, Japan S Supporting Information *

ABSTRACT: When analyzing the kinetics of liquid−liquid extraction (LLE), the change in the concentration of extracted target molecules over time should be monitored for a known interfacial area. Herein, we developed a microfluidic system for precisely analyzing the kinetics of LLE using droplets of a constant size even in the presence of additives. Extraction is initiated by exchanging the carrier fluid for the droplets with a target solution and then terminated by phase separation, based on the principle of hydrodynamic filtration. By using one out of several pairs of outlet/buffer inlet at a given time, the extraction time period is tuned stepwise without changing the flow rate condition. We successfully demonstrated droplet-based LLE by controlling the extraction period from ∼0.03 to ∼1.2 s and evaluated the extraction kinetics of rhodamine B from the continuous aqueous phase to droplets of 1-octanol with a diameter of ∼40 μm. In addition, the effect of additives on extraction efficiency was evaluated. The system presented in this study would be useful for determining rate constants for interfacial mass transfer in general LLE kinetic studies as well as for developing new extraction chemistries and optimizing microfluidic chemical/ biochemical analysis systems that involve an LLE process.

L

relatively small amount of reagents required in the case of microfluidic devices are desirable for various applications such as continuous environmental monitoring.20−25 In addition, such small-scale extraction processes are suitable for handling extremely small volume samples such as during the analysis of single cell contents.26,27 In particular, microfluidic devices are suitable for precisely analyzing the extraction kinetics of LLE because the extraction period and interfacial area can be rigidly controlled.28−30 Among the various types of microfluidic schemes, droplet microfluidics-based extraction processes have become a feasible means for studying the LLE kinetics.31−34 The relatively large interface-to-volume ratio of the droplets in addition to the presence of circulating flow inside and outside the droplets significantly accelerates the extraction process.35−37 The known interfacial area along with the rapid mixing allows the evaluation of interfacial mass transfer constants. In addition, compared to parallel two-phase flow-based LLE, droplet microfluidics can minimize the influence of molecular diffusion in the lateral direction. To evaluate the extraction kinetics of LLE, it is necessary to monitor the concentration change of the target molecule in the continuous and/or dispersed phases over time. In previously developed microfluidic droplet-based LLE systems, image

iquid−liquid extraction (LLE) is an essential and widely employed process in various chemical/biochemical operations including the detection of environmental substances, biomolecular analysis for clinical diagnosis and cell biology, purification of synthetic chemical compounds, and recycling of rare metals.1−3 A number of studies have been reported on the development of new techniques and processes for efficiently performing LLE4,5 as well as on the investigation of extraction mechanisms and design of new extraction solvents. In particular, analysis of extraction kinetics is highly important because it can help one understand the extraction phenomenon at the molecular level and hence leads to the development of new LLE chemistries and processes.6−9 However, conventional bulk-scale LLE techniques that employ stirred tanks are not suitable for obtaining true extraction constants because of the uncontrolled interfacial area and the lack of proper techniques for monitoring the amount of extracted compounds under complex flow conditions in the presence of convection and diffusion phenomena. While kinetic analyses using single droplets10−14 or Lewis cells15 have been demonstrated, these techniques do not permit the analysis of fast extraction phenomena. In addition, the effect of diffusion as a confounding factor cannot be eliminated for extraction systems with relatively large diffusion distances. Recently, microfluidic devices have been employed as a practical platform for efficiently conducting LLE by utilizing the inherent laminar flow.16−19 The continuous-flow process and © XXXX American Chemical Society

Received: January 14, 2016 Accepted: May 6, 2016

A

DOI: 10.1021/acs.analchem.6b00176 Anal. Chem. XXXX, XXX, XXX−XXX

Technical Note

Analytical Chemistry

Figure 1. (a) Schematic diagrams showing the microfluidic droplet-based liquid−liquid extraction system. One pair of outlet/buffer inlet is used at a given time to significantly change the extraction time period. (b) Schematic image showing the relationship between the filtered droplet size and flow rate distribution at a branch point based on “hydrodynamic filtration”, together with a corresponding resistive circuit.

analysis has been used to investigate the concentration change of target molecules.31,38,39 However, such an analysis process has limitations in terms of the types of target molecules to which it can be applied; that is, it is not suitable for nonlabeled molecules. A system for off-chip analysis of the extracted molecules has also been reported, in which the continuous phase was split at a microchannel bifurcation.32 The extraction time period was controlled by changing the flow rate in the microchannel. However, the size of the generated droplets and the interfacial area are subject to alteration when the retention time is significantly changed, making precise kinetic analysis highly difficult. We consider a microfluidic device having multiple outlet ports to be ideal for measuring the extraction behavior over time, since the continuous phase is collected at

different time periods of extraction and examined for off-chip analysis. Another important requirement for droplet microfluidicsbased LLE is the ability to investigate the effect of coexisting substances on the extraction process.40,41 While such an investigation is essential for developing efficient additives that act as extraction agents, it is also highly difficult to perform because the size of the generated droplets is also affected when the compositions of the continuous and dispersed phases are altered. If droplets of a constant size are generated and LLE is initiated by exchanging the carrier fluid of the droplets (continuous phase) with another fluid containing the target and coexisting molecules, it is possible to eliminate the influence of droplet-size variations. Such a system would be especially beneficial for evaluating the effect of additives. B

DOI: 10.1021/acs.analchem.6b00176 Anal. Chem. XXXX, XXX, XXX−XXX

Technical Note

Analytical Chemistry

channel, the continuous phase is split through the side channels connected to one of the outlets. This indicates that the extraction is terminated at this point. We have defined the “extraction region” as the main channel segment between the first branch point to Drain 1 and the center of the branch points to the outlet (Figure 1a). The concentration of the target molecule in the medium recovered from each outlet is measured in order to determine the amount of target molecule extracted by subtraction. Tuning of the extraction period (the retention time of droplets in the extraction region) is performed by using different pairs of outlets and buffer inlets to change the volume of the extraction region. For example, the upstream pair of outlet/buffer inlet is used for short extraction periods, whereas the downstream pair is used for increasing the extraction period. The introduction of buffer flow downstream of the branch points to one of the outlets is mandatory for maintaining the pressure drop through the entire microchannel at a constant value, regardless of the choice of the pair of outlets/buffer inlets. Extraction has already been terminated at the buffer inlet, and hence, droplet extraction was not affected by the introduction of the buffer flow. The volumetric flow rate of this buffer introduced from the buffer inlet should be equal to that of the continuous phase distributed to the outlet. In this operation, the total retention time of droplets from the first junction to Drain 2 remains unchanged even when different pairs of outlets/buffer inlets are used. If the buffer solution is not introduced, the relative flow rates distributed to each drain/ outlet are changed, and thus, the volumetric ratio of the continuous/dispersed phases is changed, rendering precise kinetic analysis impossible. Microdevice Design and Fabrication. The microchannel design used in this study, together with the design strategy of the microchannel network is shown in Figure 2 and the Supporting Information. Five pairs of outlets and buffer inlets were available, as a result of which the extraction period could be controlled in five steps from ∼0.03 to ∼1 s. The width of the main channel was 80−1000 μm, and the channel depth was uniform at ∼35 μm. Wide regions of the main channel were essential for exponentially increasing the retention time. The main channel was turned at a right angle to make the entire microchannel compact. A total of 20 side channels was connected to Drain 1 and the outlets, and each side channel consisted of narrow (20 μm) and broad (30 μm) segments. The microchannel was so designed that theoretically 3.4% of the volumetric flow rate passing through each separation point was split into each side channel (Qsplit), when the ratio of the flow rates from Inlets 1 and 3 and the buffer inlet was 2:18:5 and a parabolic flow rate profile was assumed. This value corresponds to the virtual width wv of 8.85 μm, indicating that spherical droplets with a diameter larger than ∼18 μm would flow through the main channel and would be recovered from Drain 2, not flowing into the side channels. In addition, 50% (= 1 − (1 − 0.034)20) of the volumetric flow rate was estimated to be split at the 20 branch points. We fabricated microfluidic devices made of polydimethylsiloxane (PDMS) using soft-lithography and replica molding techniques as previously described.45 After inserting and gluing silicone tubes to the inlets/outlets on a PDMS replica, the replica was bonded with a flat PDMS plate via O2 plasma treatment (Figure 2b). To generate oil-in-water droplets, microfluidic devices were used immediately after the bonding process to ensure a hydrophilic surface.

Herein, we propose a new microfluidic network capable of rigidly controlling and significantly changing the extraction time period from milliseconds to seconds and quantitatively analyzing the extraction kinetics of microfluidic droplet-based LLE. In all experiments, droplets with a constant size were generated by using same continuous/dispersed phases under the same flow rate conditions. The continuous phase, i.e., the carrier fluid for the droplets, was then exchanged with a solution containing the target molecules based on the principle of “hydrodynamic filtration”,42−44 in order to initiate the droplet-based LLE. In addition, the microdevice was equipped with several pairs of outlets and buffer inlets. By using one of these pairs at a given time, we were able to collect the continuous phase after a certain time period of extraction. This device is therefore capable of (i) tuning the extraction period stepwise, (ii) using droplets having exactly the same size, and (iii) maintaining a constant flow rate and constant interface-tovolume ratio of the two phases, even when the physicochemical properties of the target solution are changed. In the experiments, we used rhodamine B (RB) as the model target molecule. This molecule was extracted from a continuous aqueous phase to the dispersed phase of 1-octanol. In addition, we analyzed the extraction kinetics and evaluated the thicknesses of the fluid films when different amounts of coexisting substances were added to the continuous phase.

■

EXPERIMENTAL SECTION Principle. The concept of the microfluidic system used for tuning the time period of droplet-based LLE is illustrated in Figure 1a. In this system, the target molecule is extracted from the continuous aqueous phase to the droplets of an oil phase. This microfluidic network has an inlet channel for the distilled water as the continuous phase (Inlet 1), inlet channel for the dispersed phase (Inlet 2), inlet for the continuous phase containing target molecules and additives (Inlet 3), main channel for LLE, several inlets for the buffer solution (buffer inlets), and several sets of multiple side channels connected to Drain 1 or outlets. It is important to note that only one pair of outlets and buffer inlets is used at a given time, while the other pairs are closed. Aqueous solutions and the oil phase are continuously introduced from the individual inlets. Droplets are generated at the first junction, which flow through the main channel. At the branch points connected to Drain 1, a small amount of the continuous phase is repeatedly split into the side channels whereas droplets do not enter the side channels because only the thin flow region near the sidewall is split based on the concept of hydrodynamic filtration (Figure 1b). The amount of split flow (Qsplit) and the width of virtual flow (wv) can be precisely controlled by properly designing the entire microchannel network based on a resistive circuit model.42,43 The carrier fluid of the droplets is, therefore, exchanged with the target solution introduced from Inlet 3. The droplet-based LLE is initiated at this point where the initial carrier fluid is completely removed from the main channel. By employing this fluid exchange-based initiation of LLE, we can use preformed droplets of a constant size even when target solutions with different compositions and characteristics are examined. Hence, this strategy is highly advantageous for precisely analyzing extraction kinetics. As the droplets flow through the main channel, the target molecules are extracted from the continuous phase to the droplets. After a certain period of droplet retention in the main C

DOI: 10.1021/acs.analchem.6b00176 Anal. Chem. XXXX, XXX, XXX−XXX

Analytical Chemistry

■

Technical Note

RESULTS AND DISCUSSION Evaluation of the Microdevice. First, we examined if the fabricated microdevices functioned as we designed, in terms of the volumetric flow rates distributed to each outlet/drain. One of the five pairs of outlet/buffer inlet was used at a given time, and all the five pairs were tested. Distilled water was introduced from Inlet 1, Inlet 3, and one of the five buffer inlets with flow rates of 20, 180, and 50 μL/min, respectively. The output samples were recovered from each outlet/drain for 2 min, and the volumes were measured by weighing. Theoretically, the ratio of volumetric flow rates distributed to Drains 1 and 2 and one of the five outlets was constant, regardless of the outlet/ buffer inlet pair used (a detailed explanation for this reason is described in the Supporting Information). Figure 3 shows the measured output volumetric flow rates. We confirmed that the measured values were in good

Figure 2. (a) Schematic design of the microchannel having five pairs of outlets and buffer inlets. (b) Photograph of the PDMS microdevice.

Extraction Experiments. All the experiments were conducted at room temperature (24 °C). We used rhodamine B (RB) (Wako PureChemical, Osaka, Japan) as the model target molecule. RB was extracted from the aqueous continuous phase to the droplets of 1-octanol (Wako). Distilled water was introduced from Inlet 1 at 20 μL/min, whereas 1-octanol and aqueous solution of 0.05 mM RB (containing 0, 0.2, and 0.8 M NaCl) were introduced from Inlet 2 and Inlet 3 at flow rates of 0.2 and 180 μL/min, respectively. Distilled water was introduced from one of the buffer inlets at 50 μL/min using syringe pumps (KDS-200, KD Scientific, MA). The extraction behavior was observed with an inverted optical microscope (IX73, Olympus, Tokyo, Japan) and a CCD camera (DP72, Olympus). Approximately 70 μL of the continuous phase was recovered from one of the five outlets for each condition. Other outlets/buffer inlets were closed by inserting metal wires to the attached silicone tubes. The RB concentration in the recovered continuous phase was determined by measuring the absorbance of RB at λ = 554 nm using a spectrophotometer (NanoDrop, Thermo Fisher Scientific, MA). In addition, the equilibrium values of RB concentrations in the aqueous and oil phases were obtained from bulk-scale experiments. An aqueous solution of 0.05 mM RB and 1octanol were introduced in a 50 mL centrifuge tube at a volume ratio equal to the volumetric flow rates of the continuous and the dispersed phases in the microfluidic experiment and vigorously stirred for 5 min to ensure the completed partition of RB. After completely separating the two phases via centrifugation, the aqueous phase was recovered and the RB concentration was determined by measuring the absorbance at λ = 554 nm. The RB concentration in the oil phase was then obtained via calculation.

Figure 3. Distribution of the volumetric flow rates to Drains 1 and 2 and one of the five outlets. The total input flow rate was maintained at 250 μL/min. Each data shows the average ± SD value from 3 independent experiments.

agreement with the theoretical values regardless of the pair of outlet/buffer inlet used. There was a slight difference in the theoretical and measured values, especially when outlets/buffer inlets 1−3 were used. This difference might have been caused by the discrepancy between the theoretical and experimental hydrodynamic resistances of the microchannel. For example, a slight increase in the width and depth of a microchannel segment results in a larger amount of flow distributed to that segment. When all the buffer inlets were closed and the buffer solution was not used, the ratio of the flow rate distributed to each outlet/drain was affected by the outlet used (Figure S7 in Supporting Information). This result indicates that the introduction of the buffer was mandatory for maintaining the constant volumetric ratio of the continuous and the dispersed phases. On the basis of the experimental values of output flow rates, we calculated the average retention times of the fluid in the extraction region. The retention times were 0.03, 0.10, 0.28, 0.55, and 1.19 s, when outlets/buffer inlets 1, 2, 3, 4, and 5, respectively, were used. These values were used as the extraction periods in the subsequent droplet-based LLE experiments. Droplet Generation, Fluid Exchange, and Phase Separation. Next, we generated droplets of 1-octanol and performed LLE of RB from the continuous phase. When the flow rate of the dispersed phase was 0.2 μL/min, droplets with an average diameter ± SD of 40.6 ± 0.5 μm were generated at D

DOI: 10.1021/acs.analchem.6b00176 Anal. Chem. XXXX, XXX, XXX−XXX

Technical Note

Analytical Chemistry the first junction, with a generation frequency of ∼110 Hz. Considering that the microchannel depth (∼35 μm) was slightly smaller than the droplet diameter, these droplets were not completely spherical but were slightly deformed into an ellipsoidal shape. The surface area and volume of the droplets were calculated to be 4.7 × 10−3 mm2 and 3.0 × 10−5 mm3, respectively, by assuming an ellipsoidal droplet shape. The flow behavior of droplets of 1-octanol flowing in the microchannel is shown in Figure 4. The generated droplets flowed through the main channel in a stable manner and did not coalesce with each other. At the branch points to Drain 1 and to one of the five outlets, the droplets were not divided and

only the continuous phase was drained from the main channel (Figure 4a,e). This high stability of the droplets was due to the relatively small droplet size, sufficiently low ratio of flow rate split into the side channels, and the relatively high interfacial tension between the two phases.46,47 At the first branch points to Drain 1, the initial carrier fluid (distilled water) for the droplets was almost completely drained through the first 3−4 side channels (Figure 4c), indicating that the carrier fluid exchange was complete and extraction was initiated. In the wide channel segment of the extraction region, the droplets flowed near the center of the main channel but slightly close to one sidewall, possibly because the shift of the droplet position during flowing through the branch points to Drain 1 was maintained (Figure 4d). As the droplets flowed toward Drain 2, extraction progressed and the droplet color gradually turned into red (Figure 4f) because of the extracted RB. When an additive (NaCl) was added to the RB solution, the viscosity of the continuous phase was slightly increased (10% increase when the NaCl concentration was 0.8 M), whereas the droplet behavior was unaffected. In the usual droplet formation process using microfluidic devices, the droplet diameter is changed when the composition of the fluids is changed. The microfluidic device developed in this study is advantageous because droplets of a constant size can be employed and droplet generation and LLE are completely independent. Measurement of Extraction Kinetics. Next, we attempted to analyze the extraction kinetics of RB in the presence of additives. Using all the five outlet/buffer inlet pairs, we first measured the increase in the RB concentration in the droplets over time. Figure 5 presents the RB concentrations in the

Figure 5. Increase in the RB concentrations in the droplets over time in the presence and absence of coexisting substances (NaCl) in the target solution. The lines are the fitting curves obtained by using the nonlinear least-squares method with eq 1.

droplets as a function of the extraction period. The NaCl concentration was set at values of 0, 0.2, and 0.8 M and equilibrium RB concentration values were measured from the bulk-scale experiments. At all the three NaCl concentrations, we observed that the RB concentration in the droplets increased with an increase in the extraction period. When the NaCl concentration was 0.8 M, the amount of RB extracted into the droplets was the highest, mainly due to the salting out effect48 under such high salt concentration conditions. On the basis of this result, we calculated several kinetic parameters in the presence and absence of NaCl in the continuous phase. First, the observed kinetic constant, k, was

Figure 4. Micrographs showing the droplet and fluid behaviors. Outlet 3 and buffer inlet 3 were used, while the other outlets/buffer inlets were closed. (a, b) Droplets of 1-octanol generated at the first junction, (c) fluorescence micrograph showing the flow of RB aq. and the carrier-fluid exchange at the first branch points to Drain 1, (d) droplets flowing through the extraction region between Outlets 2 and 3, (e) droplets flowing through the branch points to Outlet 3, and (f) droplets flowing near Drain 2. Scale bar: 200 μm. E

DOI: 10.1021/acs.analchem.6b00176 Anal. Chem. XXXX, XXX, XXX−XXX

Technical Note

Analytical Chemistry Table 1. Values of Kinetic Parameters as a Function of Additive (NaCl) Concentration NaCl concn (M)

Co, eq (mM)

μ (mPa s)

Daq (10−10 m2/s)

k (/s)

κ (−)

βo′ (10−6 m/s)

δaq (μm)

0 0.2 0.8

3.29 3.66 6.20

0.98 1.00 1.07

4.22 4.14 3.87

0.489 0.479 0.429

75.3 85.2 162.7

2.74 2.64 2.11

2.05 1.84 1.13

determined by analyzing the fitting curves for the three sets of plots shown in Figure 5 using the nonlinear least-squares method with the following equation:10 Co = Co,eq(1 − e−kt )

that the viscosity of the continuous phase increased with increase in the NaCl concentration. Table 1 shows the values of equilibrium and kinetic parameters obtained. The values of k decreased with increase in the NaCl concentration, probably because of the slight increase in the viscosity of the continuous phase with the addition of NaCl. The fluid film thicknesses, δaq, were in the range of ∼1.1 to ∼2.1 μm, and the value decreased slightly with increase in the NaCl concentration in the continuous phase. This decrease was mainly caused by the significantly increased values of κ with the addition of NaCl. In a previous report on microfluidic droplets for LLE in which RB was extracted from aqueous droplets (Φ = ∼200 μm) to the continuous oil phase and the fluid film thickness was measured via fluorescence imaging, values of δaq as high as 10−40 μm were measured.31 However, image analysis cannot be applied for the evaluation of fluid film thicknesses as thin as ∼1 μm for droplets as small as several tens of micrometers. In contrast, the microdevice developed in this study enables the quantitative measurement of molecule extraction outside the microdevice over time. Therefore, the fluid film thickness can be estimated even for significantly small-sized droplets. In addition, the system presented in this study is advantageous because it is capable of precisely estimating the kinetics for rapid extraction processes even in the presence of additives. These features are realized by utilizing droplets of constant size, which were affected by the operating parameters in previously developed droplet microfluidics-based LLE.

(1)

where Co and Co,eq are the concentrations of RB in the droplets at time point t and at equilibrium, respectively. Next, coefficient of the overall mass transfer based on the oil (dispersed) phase concentration, βo′, was calculated using the following equation: k = (A i /Vo)βo′{κ × (Vo/Vaq) + 1}

(2)

where κ is the partition coefficient, Vo and Vaq are the volumes of the dispersed oil phase and continuous aqueous phase, and Ai is the interfacial area (surface area of the droplet). In this experiment, the values of Ai/V0 and V0/Vaq were constant at 1.56 × 105 m−1 and 1.92 × 10−3, respectively, owing to the remarkable characteristic that the volume of the generated droplets is always constant. In the wide extraction areas, a region of the continuous phase at a large distance from the droplets might have not been involved in LLE; however, in this study the value of Vaq was defined as the entire volume of the continuous phase. In addition, we examined the thickness of the fluid film, which is a parameter closely associated with the LLE efficiency. According to the double film theory, fluid films are formed on both sides of the interface between the two phases and the target concentrations in the two phases are regarded as homogeneous except for the fluid film regions. Assuming rapid transfer of RB molecules across the liquid−liquid interface, βo′ is expressed by the following equation: 1/βo′ = (1/βo) + (κ /βaq )

■

CONCLUSIONS A new microfluidic system has been developed for evaluating the kinetics of droplet-based LLE. The continuous phase and the droplets were separated by utilizing the principle of hydrodynamic filtration, and the carrier fluid for the droplets was exchanged to initiate LLE. In this process, precise control of the extraction period in the subsecond time scale and quantitative off-chip analysis of target molecules were achieved, indicating that various techniques would be available for determining the amount of extracted molecules in the droplets/ continuous phase, such as electrophoresis 50 and mass spectrometry.51 The presence of multiple pairs of inlet/outlet in the microdevice enabled multistep tuning of the extraction period without changing the extraction conditions including flow rate, droplet volume, and interfacial area. We successfully evaluated the extraction kinetics of RB including the kinetic constant and the fluid film thickness, both in the presence and absence of NaCl in the continuous phase. Although we only demonstrated extraction from the continuous phase to the droplets, reverse extraction would also be possible depending on the nature of the fluids. LLE using microfluidic droplets has recently become increasingly popular and important because of the ability to analyze small amounts of samples such as single cells. Therefore, the system presented in this study would be useful not only for accurately evaluating the rapid extraction kinetics in picoliter-scale droplets but also for developing new

(3)

where βo and βaq are the individual mass transfer coefficients for the oil and aqueous phases, respectively. For both phases, the mass transfer coefficient β is presented by the following equation: β = D/δ

(4)

where D is the diffusion coefficient of RB and δ is the fluid film thickness in each phase. Therefore, eq 3 can be written as 1/βo′ = (δo/Do) + (κδaq /Daq )

(5)

where D0 and Daq are the diffusion coefficients of RB in the oil and aqueous phases, respectively, and δo and δaq are the fluid film thicknesses in the oil and aqueous phases, respectively. In the present study, the first term on the right side in eq 5, which indicates the mass transfer resistance of the oil phase, was neglected because κ was sufficiently large.10,34 The thickness of the fluid film in the aqueous phase (i.e., outside the droplets), δaq, can be therefore calculated using the following equation: δaq = (1/βo′) × (Daq /κ )

(6)

In this study, the values of Daq of RB in the presence of additives were calculated from reported values,49 by considering F

DOI: 10.1021/acs.analchem.6b00176 Anal. Chem. XXXX, XXX, XXX−XXX

Technical Note

Analytical Chemistry

(24) Wijethunga, P. A.; Nanayakkara, Y. S.; Kunchala, P.; Armstrong, D. W.; Moon, H. Anal. Chem. 2011, 83, 1658−1664. (25) Chao, L.; Richards, M. J.; Hsia, C. Y.; Daniel, S. Anal. Chem. 2013, 85, 6696−6702. (26) Guo, M. T.; Rotem, A.; Heyman, J. A.; Weitz, D. A. Lab Chip 2012, 12, 2146−2155. (27) Rosenfeld, L.; Lin, T.; Derda, R.; Tang, S. K. Y. Microfluid. Nanofluid. 2014, 16, 921−939. (28) Surmeian, M.; Hibara, A.; Slyadnev, M.; Uchiyama, K.; Hisamoto, H.; Kitamori, T. Anal. Lett. 2001, 34, 1421−1429. (29) Priest, C.; Zhou, J.; Sedev, R.; Ralston, J.; Aota, A.; Mawatari, K.; Kitamori, T. Int. J. Miner. Process. 2011, 98, 168−173. (30) Goyal, S.; Desai, A. V.; Lewis, R. W.; Ranganathan, D. R.; Li, H.; Zeng, D.; Reichert, D. E.; Kenis, P. J. Sens. Actuators, B 2014, 190, 634−644. (31) Mary, P.; Studer, V.; Tabeling, P. Anal. Chem. 2008, 80, 2680− 2687. (32) Nichols, K. P.; Pompano, R. R.; Li, L.; Gelis, A. V.; Ismagilov, R. F. J. Am. Chem. Soc. 2011, 133, 15721−15729. (33) Lubej, M.; Novak, U.; Liu, M.; Martelanc, M.; Franko, M.; Plazl, I. Lab Chip 2015, 15, 2233−2239. (34) Wang, W. T.; Sang, F. N.; Xu, J. H.; Wang, Y. D.; Luo, G. S. RSC Adv. 2015, 5, 82056−82064. (35) Song, H.; Chen, D. L.; Ismagilov, R. F. Angew. Chem., Int. Ed. 2006, 45, 7336−7356. (36) Teh, S. Y.; Lin, R.; Hung, L. H.; Lee, A. P. Lab Chip 2008, 8, 198−220. (37) Zhu, Y.; Fang, Q. Anal. Chim. Acta 2013, 787, 24−35. (38) Kumemura, M.; Korenaga, T. Anal. Chim. Acta 2006, 558, 75− 79. (39) Chen, Y. Y.; Chen, Z. M.; Wang, H. Y. Lab Chip 2012, 12, 4569−4575. (40) Wu, H. F.; Yen, J. H.; Chin, C. C. Anal. Chem. 2006, 78, 1707− 1712. (41) Amid, M.; Manap, Y.; Azmira, F.; Hussin, M.; Sarker, Z. I. J. Chromatogr. B: Anal. Technol. Biomed. Life Sci. 2015, 993−994, 1−8. (42) Yamada, M.; Seki, M. Lab Chip 2005, 5, 1233−1239. (43) Yamada, M.; Seki, M. Anal. Chem. 2006, 78, 1357−1362. (44) Yamada, M.; Kobayashi, J.; Yamato, M.; Seki, M.; Okano, T. Lab Chip 2008, 8, 772−778. (45) Duffy, D. C.; McDonald, J. C.; Schueller, O. J. A.; Whitesides, G. M. Anal. Chem. 1998, 70, 4974−4984. (46) Yamada, M.; Doi, S.; Maenaka, H.; Yasuda, M.; Seki, M. J. Colloid Interface Sci. 2008, 321, 401−407. (47) Moritani, T.; Yamada, M.; Seki, M. Microfluid. Nanofluid. 2011, 11, 601−610. (48) Pelit, L.; Bagatir, I.; Pelit, F. O.; Ertas, F. N. RSC Adv. 2014, 4, 32189−32196. (49) Gendron, P. O.; Avaltroni, F.; Wilkinson, K. J. J. Fluoresc. 2008, 18, 1093−1101. (50) Chen, F.; Rang, Y.; Weng, Y.; Lin, L.; Zeng, H.; Nakajim, H.; Lin, J. M.; Uchiyama, K. Analyst 2015, 140, 3953−3959. (51) Zhang, Y.; Li, H.; Ma, Y.; Lin, J. M. Analyst 2014, 139, 1023− 1029.

extraction devices and chemistries for analytical and preparative applications.

■

ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.analchem.6b00176. Design and design process of the microfluidic device in detail and a result of flow rate distributed to each drain/ outlet when the buffer flow was not introduced (PDF)

■

AUTHOR INFORMATION

Corresponding Author

*Fax/phone: +81-43-290-3398. E-mail: m-yamada@faculty. chiba-u.jp. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS This study was supported in part by Grants-in-Aid for Scientific Research (Grants 23655063 and 23106007) from the Ministry of Education, Culture, Sports, Science, and Technology, Japan.

■

REFERENCES

(1) Pedersen-Bjergaard, S.; Rasmussen, K. E. J. Chromatogr. B: Anal. Technol. Biomed. Life Sci. 2005, 817, 3−12. (2) Han, D.; Row, K. H. Microchim. Acta 2012, 176, 1−22. (3) Yan, Y.; Chen, X.; Hu, S.; Bai, X. J. Chromatogr. A 2014, 1368, 1− 17. (4) Berthod, A.; Carda-Broch, S. J. Chromatogr. A 2004, 1037, 3−14. (5) Silvestre, C. I.; Santos, J. L.; Lima, J. L.; Zagatto, E. A. Anal. Chim. Acta 2009, 652, 54−65. (6) Burns, J. R.; Ramshaw, C. Lab Chip 2001, 1, 10−15. (7) Kashid, M. N.; Harshe, Y. M.; Agar, D. W. Ind. Eng. Chem. Res. 2007, 46, 8420−8430. (8) Kashid, M. N.; Renken, A.; Kiwi-Minsker, L. Chem. Eng. Sci. 2011, 66, 3876−3897. (9) Xie, Y.; Chindam, C.; Nama, N.; Yang, S.; Lu, M.; Zhao, Y.; Mai, J. D.; Costanzo, F.; Huang, T. J. Sci. Rep. 2015, 5, 12572. (10) Jeannot, M. A.; Cantwell, F. F. Anal. Chem. 1996, 68, 2236− 2240. (11) Jeannot, M. A.; Cantwell, F. F. Anal. Chem. 1997, 69, 235−239. (12) Ma, M. H.; Cantwell, F. F. Anal. Chem. 1998, 70, 3912−3919. (13) Fiamegos, Y. C.; Stalikas, C. D. Anal. Chim. Acta 2007, 599, 76− 83. (14) Jain, A.; Verma, K. K. Anal. Chim. Acta 2011, 706, 37−65. (15) Danesi, P. R.; Vandegrift, G. F.; Horwitz, E. P.; Chiarizia, R. J. Phys. Chem. 1980, 84, 3582−3587. (16) Lin, C. C.; Hsu, J. L.; Lee, G. B. Microfluid. Nanofluid. 2011, 10, 481−511. (17) Noel, T.; Kuhn, S.; Musacchio, A. J.; Jensen, K. F.; Buchwald, S. L. Angew. Chem., Int. Ed. 2011, 50, 5943−5946. (18) Assmann, N.; Ładosz, A.; Rudolf von Rohr, P. Chem. Eng. Technol. 2013, 36, 921−936. (19) Ciceri, D.; Perera, J. M.; Stevens, G. W. J. Chem. Technol. Biotechnol. 2014, 89, 771−786. (20) Tokeshi, M.; Minagawa, T.; Uchiyama, K.; Hibara, A.; Sato, K.; Hisamoto, H.; Kitamori, T. Anal. Chem. 2002, 74, 1565−1571. (21) Sato, K.; Hibara, A.; Tokeshi, M.; Hisamoto, H.; Kitamori, T. Adv. Drug Delivery Rev. 2003, 55, 379−391. (22) Tokeshi, M.; Kikutani, Y.; Hibara, A.; Sato, K.; Hisamoto, H.; Kitamori, T. Electrophoresis 2003, 24, 3583−3594. (23) Hiki, S.; Mawatari, K.; Aota, A.; Saito, M.; Kitamori, T. Anal. Chem. 2011, 83, 5017−5022. G

DOI: 10.1021/acs.analchem.6b00176 Anal. Chem. XXXX, XXX, XXX−XXX