Development of a Measurement Technique for Ion Distribution in an

ARTICLE pubs.acs.org/ac

Development of a Measurement Technique for Ion Distribution in an Extended Nanochannel by Super-Resolution-Laser-Induced Fluorescence Yutaka Kazoe, Kazuma Mawatari, Yasuhiko Sugii, and Takehiko Kitamori* Department of Applied Chemistry, School of Engineering, The University of Tokyo, 7-3-1 Hongo, Bunkyo, Tokyo 113-8656, Japan ABSTRACT: Ion behavior confined in extended nanospace (101
103 nm) is important for nanofluidics and nanochemistry with dominant surface effects. In this paper, we developed a new measurement technique of ion distribution in the nanochannel by super-resolution-laser-induced fluorescence. Stimulated emission depletion microscopy was used to achieve a spatial resolution of 87 nm higher than the diffraction limit. Fluorescein was used for ratiometric measurement of pH with two excitation wavelengths. The pH profile in a 2D nanochannel of 410 nm width and 405 nm depth was successfully measured at an uncertainty of 0.05. The excess protons, showing lower pH than the bulk, nonuniformly distributed in the nanochannel to cancel the negative charge of glass wall, especially when the electric double layer is thick compared to the channel size. The present study first revealed the ion distribution near the surface or in the nanochannel, which is directly related to the electric double layer. In addition, the obtained proton distribution is important to understand the nanoscale water structure between single molecules and continuum phase. This technique will greatly contribute to understanding the basic science in nanoscale and interfacial dynamics, which are strongly required to develop novel miniaturized systems for biochemical analysis and further applications.

A

miniaturized chemical system is a rapidly expanding engineering field for integration and automation of chemical processing in pharmaceutics, biomedical and chemical analysis, and chemical synthesis. High-efficiency and high-throughput chemical processing with a small volume of sample has been achieved based on microfluidics with accurate surface control for multiphase flows, short diffusion time, and stable laminar flows. Now the microchemical system is moving toward practical application, and a number of applications have been reported.1
3 Recently, the size of integration is downscaling to 101
103 nm, referred to as extended nanospace, for more efficient and novel applications of chemical analysis and synthesis.4 This regime is between single molecules and continuum phase under dominant surface effects due to the large surface to volume ratio and provides specific liquid properties. Our experimental work by capillary injection and time-resolved fluorescent measurements has reported higher viscosity and lower permittivity of water confined in extended nanospace compared to the bulk.5 In addition, nuclear magnetic resonance relaxation measurements revealed the slower intermolecular motion and higher proton mobility of water.6,7 Consequently, we proposed a model where the water molecules are loosely coupled within 50 nm of the wall to yield these liquid properties. On the other hand, another main reason for the specific liquid properties is well known as the electric double layer, i.e., a layer of counterions attracted to a charged wall surface.8,9 Since the thickness of the double layer is considered to be 1
102 nm and comparable to the space size, in some cases the double layer fills the whole region of the nanochannel to be r 2011 American Chemical Society

electrically polarized. Many researchers have highlighted that the liquid confined in a small space appears to be more viscous due to ion drags by the electric potential in the double layer, termed the electroviscousity effect.10,11 Therefore, investigating the behavior and distribution of ions, especially those of protons related to both the water molecules and the double layer, is strongly required to establish nanofluidics and nanochemistry. The ion behavior near the surface is a classical issue from the early 20th century. In the field of colloid science, a lot of efforts have been made to establish a theoretical model of the electric double layer.8 Experimental and theoretical work to validate the model have been reported, also for nanochannel flows.12
15 However, these studies were based on indirect measurements using particle and fluid motion (e.g., sedimentation potential, streaming potential, electrophoresis, electroosmosis). To date, measurement of the ion distribution near the surface has been difficult owing to the lack of a method. One approach to reveal the ion distribution near the surface is using an optical method of a resolution higher than the Abbe’s optical diffraction limit. Previous studies have developed a method using fluorescent ions and the evanescent wave, which can illuminate the near-wall region within 400 nm.16,17 The near-wall ion concentration related to the double layer was extracted as a fluorescent intensity, and the method was applied to obtain the Received: June 28, 2011 Accepted: September 26, 2011 Published: September 26, 2011 8152

dx.doi.org/10.1021/ac201654r | Anal. Chem. 2011, 83, 8152–8157

Analytical Chemistry

ARTICLE

distribution of the electrostatic potential at the wall, i.e., the zeta potential.18 Furthermore, in recent advances, several kinds of super-resolution microscopy of 10
100 nm resolution have been developed.19
22 In particular, stimulated emission depletion (STED) microscopy, which enables one to achieve a fluorescence excitation spot smaller than 100 nm, is consider to be most applicable to realize visualization of the ion behavior, with a similar principle as the previous method. In this paper, we developed a new method by super-resolution-laser-induced fluorescence (LIF) based on STED microscopy in order to measure the ion distribution in the extended nanochannel. A fluorescent dye was selected as a pH indicator with two excitation wavelengths and used for measurement of the proton distribution in a 400 nm fused silica nanochannel. The spatial resolution was verified experimentally, and measurements using the indicator fluorescent dye and STED microscopy were validated by obtaining the calibration curve between the pH of the liquid and the fluorescent intensity. The obtained results of the proton distribution were compared to the double-layer scale. The effect of nanoconfinement on the proton behavior in extended nanospace was also discussed.

’ EXPERIMENTAL SECTION Materials and Chemicals. A 8.6  10
6 mol/L fluorescein

(C20H12O5) was used as a fluorescent pH probe and dissolved into electrolyte solutions. For pH measurements in nanochannels, three kinds of electrolyte solution of different Debye lengths λD, deionized water (λD = 311 nm), potassium chloride (KCl) 10
4 mol/L (λD = 30 nm), and KCl 10
2 mol/L (λD = 3 nm), were prepared. These Debye lengths were calculated from the bulk ion concentration ci as follows  λD ¼

εkT 2NA e2 I

Figure 1. Schematic of (a) the measurement system and (b) fluorescence excitation with STED.

1=2 ð1Þ

where ε is the permittivity, k is the Boltzmann constant, T is the temperature, NA is the Avogadro constant, e is the elementary electric charge, and I = 1/2Σ(cizi2) is the ionic strength (zi = the ion valence). It is noted that the Debye length of the water was calculated using a pH measured in a microchannel as mentioned in a later section. For calibration experiments of the pH, nine kinds of buffer solution were prepared for pH 2.92, 3.61, 4.46, 5.38, and 6.26 from citric acid (C6H8O7) and disodium hydrogenphosphate (Na2HPO4), pH 6.78, 7.60, and 8.18 from sodium dihydrogenphosphate (NaH2PO4) and Na2HPO4, and pH 9.09 from sodium carbonate (Na2CO3) and sodium hydrogen carbonate (NaHCO3). The pH of the buffer solution was measured by a pH meter of a measurement error of 0.02. Ionic strengths were adjusted to be O(10
2 mol/L). For the experiments to verify a spatial resolution of the STED microscopy, fluorescent calboxylate-modified polystyrene particles of different diameters of 36, 110, and 200 nm were used (Invitrogen; absorption peak, 505 nm; emission peak, 515 nm). The particles were suspended in a 10
2 mol/L calcium chloride (CaCl2) to stick particles on a fused silica surface. Measurement System. Measurements were conducted using a STED microscope system (Leica, TCS STED CW), as illustrated in Figure 1a. The system has two laser beams: one is an Ar laser of 458 nm/488 nm wavelengths with a Gaussian intensity profile for fluorescence excitation, and another is a fiber

Figure 2. (a) Schematics of a pressure-driven flow control system, and (b) cross-sectional view of a 2D nanochannel (A
A0 ). Five nanochannels were fabricated on a fused silica plate to verify the repeatability.

laser of 592 nm wavelength of a doughnut shape intensity profile for STED, i.e., the STED beam formed by an optical phase filter. The two laser beams were completely aligned and focused to fluorescent molecules through dichroic mirrors and an oil immersion objective lens (100, NA = 1.4, refractive index of immersion oil n = 1.52), as shown in Figure 1b. The fluorescent molecules in the center of the focal spot are excited to the singlet state and emit fluorescence with de-excitation to the ground state. On the other hand, those in the periphery of the focal spot, where the STED beam is illuminated, are also excited to the singlet state but forced to return to the ground state without fluorescence emission.21 Hence, only the fluorescence emitted from the molecules in the doughnut hole is detected. Therefore, 8153

dx.doi.org/10.1021/ac201654r |Anal. Chem. 2011, 83, 8152–8157

Analytical Chemistry the spatial resolution of O(10
100 nm) can be achieved, which is smaller than the width of the Gaussian focal spot under the optical diffraction limit. In the present system, fluorescence was detected by an avalanche photodiode (APD) through the objective lens, the dichroic mirrors, a confocal pinhole of a diameter corresponding to 1 Airy unit (AU) = 1.22λ/NA by assuming the emission wavelength λ = 500 nm, and an emission filter passing wavelengths of 500
550 nm. It is noted that the linear sensitivity of APD was validated prior to measurements. Image scanning at an interval of 33.7 nm was conducted by galvanometer mirrors. Figure 2a shows a schematic of the pressure-driven flow control system for extended nanochannels. Fabrication of a microchip containing nanochannels has been described previously.6,7,23,24 A rectangular nanochannel of 410 nm width and 405 nm depth (2D nanochannel) was fabricated on a fused silica plate by electron beam lithography and plasma etching as illustrated in Figure 2b. The width and depth of the nanochannels were measured by scanning electron microscopy and atomic force microscopy, respectively. After fabricating the nanochannels, microchannels for sample injection were fabricated on the same glass plate. Then holes for inlets and outlets were made, and the channels were sealed with another fused silica plate of a 0.17 mm thickness by thermal fusion bonding. Prior to the experiments, the channel was cleaned by flashing acetone, ethanol, deionized water, and 10
1 mol/L potassium hydroxide (KOH) and then rinsed by deionized water. The sample solution was driven and injected into the nanochannels by an air pressure of 0.4 MPa. Measurements were conducted under the flow condition to avoid photobleaching of fluorescein. An effect of CO2 dissolution using compressed air on the pH was confirmed to be negligible, because the pHs of liquid injected by the air and nitrogen gas were almost similar within measurement errors.

’ RESULTS AND DISCUSSION Evaluation of Spatial Resolution. The present study used the fused silica microchip with a refractive index n = 1.46 for the measurements. However, since the commercial objective lens is optimally designed for a 0.17 mm cover glass of n = 1.52, it is possible that super-resolution imaging cannot be achieved owing to optical aberrations. Hence, the spatial resolution of the STED system was verified by imaging nanoparticles stuck on the 0.17 mm fused silica plate due to the charge inversion effect25 by suspension in 10
2 mol/L CaCl2 solution. Results were compared to those using the commercial cover glass. Figure 3a shows a fluorescence image of 110 nm particles stuck on the fused silica plate, obtained from a 20 frame average. Then the particle image diameter was determined by the cross correlation between the particle image and a 2D Gaussian function assumed as an identical image.26 Figure 3b shows the estimated particle image diameters as a function of the actual particle diameter. Vertical error bars show the standard deviation from 10
20 particles. When the diameter is 200 nm, the measured diameters are approximately equal to the actual size. For the 36 and 110 nm particles, the values are almost constant and larger than the actual diameters: approximately 134 nm for the cover glass and 174 nm for the fused silica. These values indicate the spatial resolution limited by the width of the fluorescence excitation spot. Therefore, the spatial resolution, i.e., the minimal distance to resolve two separate points, was estimated to be 134/ 2 = 66 nm for the cover glass and 174/2 = 87 nm for the fused silica plate. Although the spatial resolution using the fused silica is

ARTICLE

Figure 3. (a) Fluorescence image of 110 nm particles stuck on the fused silica plate. Image was obtained from a 20 frame average. (b) Diameter of the particle image estimated from cross-correlation analysis as a function of the actual diameter of the particle.

lower than that using the cover glass mainly due to the optical aberration, the STED microscope system could be applied also for the fused silica. On the other hand, the depthwise resolution for imaging (z direction) is under the diffraction limit. The depthwise resolution was estimated to be 655 nm based on the same principle as confocal microscopy.27 Calibration Experiments for pH Measurement. For the pH measurements using fluorescence, calibration was conducted to obtain a relationship between the pH of the liquid and the fluorescent intensity emitted from fluorescein molecules. Fluorescein changes its chemical structure as an ion and the fluorescent intensity If depending on the pH of the surrounding liquid given by the following equation28
30 If ¼ Ie cϕεðpHÞ

ð2Þ

where Ie is the excitation intensity, c is the concentration, ϕ is the quantum yield, and ε is the molar absorption coefficient. The dependence of the molar absorption coefficient on pH enables pH measurement. However, the fluorescent intensity is also affected by the excitation intensity and concentration. In particular, a nonuniform distribution of fluorescein ions due to the wall electrostatic potential in the nanochannel causes significant errors.31 Therefore, the measurement technique of pH in nanochannels was established based on ratiometric imaging using two excitation wavelengths. The ratiometric measurements utilized the property of the molar absorption coefficient of fluorescein, which is strongly dependent on the pH at excitation wavelengths around 490 nm while almost independent at wavelengths around 460 nm.29,30 The fluorescent intensities at pH 3
9 were detected by 488 and 458 nm excitations, which were provided by changing the wavelength of the Ar laser. A fused silica microchannel of 70 μm width and 30 μm depth was used for the calibration experiments in stable pH conditions without wall effects. Figure 4a shows the fluorescent intensity as a function of the pH of the buffer solution for the 488 and 458 nm excitations. Vertical error bars show the 8154

dx.doi.org/10.1021/ac201654r |Anal. Chem. 2011, 83, 8152–8157

Analytical Chemistry

ARTICLE

Figure 4. (a) Fluorescent intensity as a function of the pH of the buffer solution. The intensity is normalized by the maximum value of the grayscale intensity of the 488 nm excitation. (b) Ratio of fluorescent intensity by the 488 nm excitation to that by the 458 nm excitation, and calibration curve obtained by a nonlinear regression.

standard deviation obtained from three measurements. The fluorescent intensity varies significantly with pH in the 488 nm excitation compared to that in the 458 nm excitation. Then a calibration curve for pH measurement was obtained from the ratio of fluorescent intensities by the 488 and 458 nm excitations (Figure 4b) given using eq 2 If ðe488nmÞ Ie488nm ϕ488nm ε488nm ¼ If ðe458nmÞ Ie458nm ϕ458nm ε458nm

ð3Þ

The fluorescent intensity ratio varies at the pH 3.0
7.5 range, and this pH range is suitable for measurement of the pH of aqueous solutions. These results suggest that measurement using fluorescent indicator molecules is also applicable for STED microscopy. Distribution of Proton Concentration in the Nanochannel. Using the calibration curve, the pH in the 400 nm fused silica nanochannel was measured. The five nanochannels were imaged from a bottom-wall side in Figure 2b at a scanning rate of 20 μs/ point for 17.3 μm  2.2 μm (512  64 points). Owing to the extremely small volume of the fluorescence and the low concentration, the signal-to-noise ratio (S/N) of a single-point detection was 0.5 and not sufficient to obtain the fluorescence image. Hence, the S/N was increased to 7.4 by averaging 320 frames to successfully image the nanochannels as shown in Figure 5a. Assuming a flow velocity based on the Hagen
Poiseuille law of 6 mm/s, O(103) fluorescein molecules were excited to detect the fluorescent intensity. In this excitation process, heating of water by the laser probes can occur and affect the measurement results. From an approximate estimate based on an assumption of the laser probe with a cylinder shape of 400 nm diameter, the probe can generate a temperature increase of 0.001 K for 458 nm (1 mW), 0.005 K for 488 nm (6 mW), and 1.1 K for 592 nm

Figure 5. (a) Fluorescence image of the 400 nm 2D nanochannels. Image was obtained from 320 frame averaging. (b) Profiles of the pH and (c) the proton concentration in the 400 nm 2D nanochannel for water, KCl 10
4 and 10
2 mol/L, compared to the bulk pH measured in the microchannel.

(300 mW) within 20 μs exposure, which are dependent on the light absorption coefficients of water. However, considering the thermal diffusion and flow velocity, the effect of laser heating is negligible in the experiments. The pH profile in the 2D nanochannel was estimated from the fluorescence images. The coordinate in the nanochannel was determined by locating the center axis (x direction) of the channel by Gaussian fitting to the image. The (x, y) coordinate of each measurement point in the image was assumed to be the geometrical center of the STED focal spot (87 nm width, 655 nm depth). However, there still can be an error of the data position owing to a nonuniform distribution of fluorescein ions within the STED focal spot. Considering the fluorescent intensity was obtained by the integral of intensities within the focal volume, the maximum error of the position was one-half the width of the spot of 44 nm, which may be induced mainly near the wall. After determining the coordinate, the fluorescent intensities were averaged over the x direction to further increase the S/N to 15.6. The calibration curve was applied to the ratio of fluorescent intensities by 488 and 458 nm excitations, and the pH profile in the y direction was obtained. Since the depth of the focal spot is larger than the channel depth, the obtained results were considered to be the pH averaged over the z direction. The measurement uncertainty of pH at the 95% confidence level was 0.05. Figure 5b and 5c shows y-direction profiles of the pH and proton concentration in the 400 nm nanochannel, respectively. Vertical error bars show the standard deviation of the measurement of five nanochannels. The results were compared to the 8155

dx.doi.org/10.1021/ac201654r |Anal. Chem. 2011, 83, 8152–8157

Analytical Chemistry bulk value of pH 6.04 ( 0.12 (0.93 ( 0.24  10
6 mol/L) measured in the microchannel, obtained by averaging the results for the cases of water (pH 6.22), KCl 10
4 mol/L (pH 6.03), and KCl 10
2 mol/L (pH 5.92), plotted as a dashed line. When the water flowed through the nanochannel, the whole region of the channel is filled by the electric double layers of λD = 311 nm formed by protons to cancel the negative charge of the glass surface, and the average proton concentration is 19.1 times higher than the bulk. The concentration profile is significantly nonuniform: the concentration in the near-wall region of 30 nm is 6 times larger than the channel center. For the case of KCl 10
4 mol/L with λD = 30 nm, the pH is still lower than the bulk but the maximum difference of pH is approximately 0.5 since the double layer is mainly formed by potassium ions (K+). The proton concentration shows a profile relatively uniform compared to the case of water because only the near-wall region is electrically polarized. In contrast, when the KCl, 10
2 mol/L, was injected into the channel, the result shows an approximately similar pH to the bulk and the uniform concentration profile since the double layer of λD = 3 nm is thin and negligible compared to the channel size. These results for the proton distribution in the nanochannel, which were first observed by experiments, suggest that the ions distribute in the channel to cancel the wall charge strongly depending on the Debye length and form the electric double layer, as expected in the classical theory. However, for the water case, several specific effects are considered in formation of the electric double layer. Since the electric double layer is formed only by protons, the proton distribution is determined by dissociation of silanol groups at the surface (tSiOH T tSiO
+ H+) and thermal diffusion of the proton. Then the thickness of the electric double layer can be different from the Debye length estimated using the bulk pH. The data in the water case (Figure 5c) showed an obviously thinner electric double layer of approximately 100 nm than the expected value λD = 311 nm. In addition, it is considered that the protons to form the electric double layer are provided only by the silanol groups, not by the water molecules and the fluorescein acid. If the water molecules produce protons, the pH in the center of the nanochannel should be higher than the bulk, since the water molecules also produce hydroxide ions. If the 8.6  10
6 mol/L fluorescein acid is completely ionized in water, the bulk pH should be 5.07, which is much lower than the experimental results (5.93 e pH e 6.22). Further research will reveal specific effects of nanoconfinement on formation of the electric double layer, which yield different liquid properties from the bulk.

’ CONCLUSIONS A novel fluorescence-based measurement technique of ion concentration in an extended nanochannel was developed for the first time. The technique was based on the STED microscopy, which enables nanometer-order spatial resolution higher than the Abbe’s optical diffraction limit. The spatial resolution for use of a fused silica channel was evaluated to be 87 nm by imaging fluorescent nanoparticles stuck on the glass surface. Fluorescein was used as a probe for pH measurement using the ratio of fluorescent intensities by 488 and 458 nm excitations. The calibration curve between the pH of liquid and the fluorescent intensity ratio was successfully obtained, suggesting that measurements using a fluorescent indicator are available even for STED microscopy. The profile of pH in a 400 nm 2D nanochannel was first measured at an uncertainty at the 95% confidence level of 0.05.

ARTICLE

The results show excess protons distribute in the nanochannel to cancel the negative charge of fused silica, strongly related to the electric double-layer thickness. However, for the water case, formation of the electric double layer can be different from other cases, since the double layer is formed only by protons, which affects dissociation of the silanol groups. In this paper, we experimentally observed the proton distribution in an extended nanochannel. The developed technique also can be applied to measurements of the distribution of other ion species by selecting an optimal fluorescence indicator. The experimental observation of ions directly related to the electric double layer was first achieved, while previous work was based on indirect measurements of surface conductivity, streaming potential, and electroosmotic flows or numerical simulation based on the classical model of the double layer. Therefore, this technique will greatly contribute to the basic interfacial science and the novel research field of nanofluidics and nanochemistry with the specific properties of fluid and ion transport.

’ AUTHOR INFORMATION Corresponding Author

*Phone: +81-3-5841-7231. Fax: +81-3-5841-6039. E-mail: [email protected].

’ ACKNOWLEDGMENT This work was supported by a Grant-in-Aid for Specially Promoted Research from the Japan Society for the Promotion of Science (JSPS). The facility of the STED system was provided from the Academic Consortium for Nano and Micro Fabrication of four Universities (Keio University, Waseda University, Tokyo Institute of Technology, and The University of Tokyo, Japan). ’ REFERENCES (1) West, J.; Becker, M.; Tombrink, S.; Manz, A. Anal. Chem. 2008, 80, 4403–4419. (2) Hartman, R. L.; Jensen, K. F. Lab Chip 2009, 9, 2495–2507. (3) Tokeshi, M.; Minagawa, T.; Uchiyama, K.; Hibara, A.; Sato, K.; Hisamoto, H.; Kitamori, T. Anal. Chem. 2002, 74, 1565–1571. (4) Tsukahara, T.; Mawatari, K.; Kitamori, T. Chem. Soc. Rev. 2010, 39, 1000–1013. (5) Hibara, A.; Saito, T.; Kim, H. B.; Tokeshi, M.; Ooi, T.; Nakao, M.; Kitamori, T. Anal. Chem. 2002, 74, 6170–6176. (6) Tsukahara, T.; Hibara, A.; Ikeda, Y.; Kitamori, T. Angew. Chem., Int. Ed. 2007, 46, 1180–1183. (7) Tsukahara, T.; Mizutani, W.; Mawatari, K.; Kitamori, T. J. Phys. Chem. B 2009, 113, 10808–10816. (8) Ren, C. L.; Li, D. J. Colloid Interface Sci. 2004, 274, 319–330. (9) Wang, M.; Chang, C.-C.; Yang, R.-J. J. Chem. Phys. 2010, 132, 024701(6pp). (10) Hunter, R. J. Zeta Potential in Colloid Science; Academic Press: London, 1981. (11) Probstein, R. F. Physicochemical Hydrodynamics: an introduction, 2nd ed.; John Willy & Sons: New York, 1994. (12) Scales, P. J.; Grieser, F.; Healy, W. Langmuir 1992, 8, 965–974. (13) Revil, A.; Pezard, P. A.; Glover, P. W. J. J. Geophys. Res. 1999, 104, 20021–20031. (14) Pennathur, S.; Santiago, J. G. Anal. Chem. 2005, 77, 6772–6781. (15) Pennathur, S.; Santiago, J. G. Anal. Chem. 2005, 77, 6782–6789. (16) Kazoe, Y.; Sato, Y. Anal. Chem. 2007, 79, 6727–6733. (17) Kazoe, Y.; Sato, Y. J. Fluid Sci. Technol. 2007, 2, 429–440. (18) Kazoe, Y.; Miyakawa, S.; Miki, N.; Sato, Y. Appl. Phys. Lett. 2009, 95, 234104(3pp). 8156

dx.doi.org/10.1021/ac201654r |Anal. Chem. 2011, 83, 8152–8157

Analytical Chemistry

ARTICLE

(19) Gustafsson, M. G. L. Proc. Natl. Acad. Sci. U.S.A. 2005, 102, 13081–13086. (20) Rust, M. J.; Bates, M.; Zhuang, X. Nat. Methods 2006, 3, 793–795. (21) Hell, S. W. Nat. Biotechnol. 2003, 21, 1347–1355. (22) Rittweger, E.; Han, K. Y.; Irvine, S. E.; Eggeling, C.; Hell, S. W. Nat. Photonics 2009, 3, 144–147. (23) Tsukahara, T.; Mawatari, K.; Hibara, A.; Kitamori, T. Anal. Bioanal. Chem. 2008, 391, 2745–2752. (24) Tamaki, E.; Hibara, A.; Kim, H. B.; Kitamori, T. J. Chromatogr., A 2006, 1137, 256–262. (25) ven der Heyden, F. H. J.; Stein, D.; Besteman, K.; Lemay, S. G.; Dekker, C. Phys. Rev. Lett. 2006, 96, 224502(4pp). (26) Adrian, R. J. Annu. Rev. Fluid Mech. 1991, 23, 261–304. (27) Park, J. S.; Choi, C. K.; Kihm, K. D. Exp. Fluids 2004, 37, 105–119. (28) Walker, D. A. J. Phys. (Paris) 1987, 20, 217–224. (29) Coppeta, J.; Rogers, C. Exp. Fluids 1998, 25, 1–15. (30) Sj€oback, R.; Nygren, J.; Kubista, M. Spectrochim. Acta, Part A 1995, 51, L7–L21. (31) Bottenus, D.; Oh, Y.-J.; Han, S. M.; Ivory, C. F. Lab Chip 2009, 9, 219–231.

8157

dx.doi.org/10.1021/ac201654r |Anal. Chem. 2011, 83, 8152–8157