Simple and Precise Size-Separation of Microparticles by a Nano-Gap Method

Transcription

1 ANALYTICAL SCIENCES MAY 2009, VOL The Japan Society for Analytical Chemistry Simple and Precise Size-Separation of Microparticles by a Nano-Gap Method Yukiko ENOMOTO, Hideaki MONJUSHIRO, and Hitoshi WATARAI Department of Chemistry, Graduate School of Science, Osaka University, Machikaneyama, Toyonaka, Osaka , Japan A simple and precise size measurement device of microparticles in liquid has been developed by using a narrow gap between a cylindrical lens and a flat glass wall, in which the distance of the gap is precisely determined by the diffraction Moiré pattern of irradiated light. Two different methods to transfer the particles toward the center line of the lens, a hydrodynamic flow mode and an electrophoretic mode, have been established. By using these methods, simultaneous size-determination and size-separation of small amounts of micrometer and sub-micrometer sized polystyrene particles in water have been attained with high precision. Novel advantages of the cylindrical nano-gap method were demonstrated for different surface types of polystyrene particles. (Received February 17, 2009; Accepted March 30, 2009; Published May 10, 2009) The separation of nano- and micro-particles is a crucial subject in recent nano-biochemistry, environmental analysis and nanotechnology. However, in comparison with the separation methods developed for ions and molecules, such as CE and HPLC, those for nano- and micro-particles are extremely limited, 1 because of their mesoscopic properties, such as low diffusivity, high adsorptivity and aggregation ability. Various kinds of external fields have been investigated for the development of separation methods of microparticles in liquids, including laser-photophoresis, 2,3 dielectrophoresis, 4,5 sonophoresis, 6 thermophoresis, 7 magnetophoresis 8,9 and electromagnetophoresis. 10 The field flow fractionation, invented by Giddings, 1 has been applied to the size-separation of microparticles in liquids, but the resolution is still unsatisfactory. 11 The limitation in the resolution of previous size-separation methods is due to the inevitable fluctuation of fluid molecules around a particle, or the Brownian motion of the particle. On the other hand, the repulsive interaction between a wall and a particle is highly sensitive to the distance between them, as formulated by the Lenard Jones potential, which is proportional to the 12th power of the distance, regardless of the nature of the particle. 12 This gave us a simple idea that a wedge gap of two smooth walls should be used as a size-fractionation device as well as for measuring the particle size, provided that the particle is properly trapped in the gap and the gap distance can be calibrated precisely. Fortunately, such a gap distance can be measured by the diffraction of visible light between the surfaces of the walls. We have recently reported on the possibility to use a nano-gap between a flat glass and a convex lens for simultaneous size-determination and size-fractionation of microparticles, 13 in which the distance of the gap is precisely determined by the light diffraction pattern, so-called Newton ring. 14,15 When a sample solution including particles was introduced into the gap, the meniscus force carried the sample To whom correspondence should be addressed. watarai@chem.sci.osaka-u.ac.jp particles toward the center of the lens, accompanied by solvent evaporation. 13 After evaporation of solvent, the smaller particles were trapped closer to the center of the nano-gap device, just at the position where the gap distance equaled to the diameter of each particle. By this method, the size-determination of yeast cells and DNA in aqueous solution was tried. 16 The nano-gap micro-fractionation method using a convex lens demonstrated its definite superiority to other conventional fractionation methods in simplicity. The required sample amount was very small (below 1 ml), and the time for fractionation was rather short (<10 min). However, in the convex lens-type device utilizing the solvent evaporation, all solutes in the initial sample solution were concentrated during the fractionation, and undesirable phenomena of crystallization and precipitation were induced. Solvent water hardly evaporated from a concentrated electrolyte solution in the central region, where the gap distance became smaller to sub-micrometer. The capillary condensation effect also prevented the evaporation of water solvent near the lens center, and made it difficult to migrate and trap submicrometer particles near the central region. Thus, it was difficult to apply this method to particles smaller than 1 mm in diameter. To improve these drawbacks of the solvent evaporation method and to establish a novel nano-gap method, we fabricated in the present study a sealed nano-gap cell using a flat glass and a cylindrical lens, and combined it with a hydrodynamic flow method and an electrophoretic method. Experimental Nano-gap method Figure 1(A) shows a schematic drawing of the nano-gap microparticle fractionation system. The nano-gap device was constructed by a flat cover glass (Microslide glass, Matsunami) and a cylindrical lens (10 mm 10 mm with 3 mm maximum height, 7.78 or mm in curvature radius, Edmund Optics, Japan), which was fixed in a cell frame made by PTFE and

2 606 ANALYTICAL SCIENCES MAY 2009, VOL. 25 Fig. 1 (A) Schematic drawing of the nano-gap microparticle fractionation system using a cylindrical lens and a flat glass. Sample particles are introduced into the gap by the hydrodynamic flow of a solvent. (B) Top view of the nano-gap cell used for the electrophoretic mode. A sample solution including particles was introduced into both positions shown in the figure. silicon rubber. The gap distance between the lens and the flat glass was controlled with nanometer order precision by a piezoelectric element and/or fine-pitched screws attached under the cylindrical lens. At the center of the cylindrical lens, the gap was controlled to become larger than 50 nm by using finepitched screws, in order to allow the aqueous medium to flow through the gap. In this device, an inlet and outlet comprising stainless-steel tubes (510 mm i.d., 810 mm o.d., and 3 cm length) were attached oppositely, and a sample solution was introduced by a siphon flow in the hydrodynamic flow method. The nanogap cell used for the electrophoresis method was similar to the flow-type cell, as shown in Fig. 1(B); besides, two silver-silver chloride electrodes were set instead of the inlet and outlet tubes in the flow-type cell, and a dc electric field of V/cm was applied to the electrodes. The distance between the electrodes was set to be 10 mm. The sample solution was put at both sides of the cylindrical lens, as shown in Fig. 1(B), before being sealed with a flat cover glass. The behaviors of the sample particles and the trapped positions were measured by reflection light, or a fluorescence image observed by a microscope (BX51 WI, Olympus, Japan) with a CCD camera (U-DP1XC, Olympus, Japan). CCD images captured with a personal computer were analyzed by using the image-processing software ImageJ. 17 The gap distance was calibrated by the interference Moiré pattern of light (520 nm) as a function of the distance from the center line of the cylindrical lens (Fig. 2). The diameters of the particles were determined from the trapped positions, which were on straight lines of the iso-gap-distance. Scanning electron microscopy (SEM) images of polystyrene Fig. 2 Top illustration shows the interference at the gap of a flat glass and a cylindrical lens. The interference Moiré patterns of 520 ± 5 nm light is shown in the middle potion of this figure. The bottom figure shows the relation between the gap distance, h, and the lateral distance from the center line of the cylindrical lens (7.78 mm in curvature radius) with a minimum gap distance of h 0 = 0.10 mm. The resolution in h, which depends on the curvature radius of lens, R, and magnification of the objective, was larger than 5 nm under the present conditions. particles were obtained by an SEM instrument (JSM-5800, JEOL, Japan) under 10 kv. Materials Sample polystyrene particles (PS) used in the hydrodynamic flow method were fluorescent carboxylated particles of 0.5 mm (excitation wavelength (ex), 441 nm; emission wavelength (em), 486 nm), 1.75 mm (ex, 360 nm; em, 407 nm), and 3.0 mm (ex, 529 nm; em, 546 nm) in diameter (Fluoresbrite Carboxylate Microspheres, Polysciences, Inc.). These particles were dispersed into an aqueous solution of 0.01 wt% Triton X-100 (Wako Pure Chemicals, Japan) to prepare the final concentration of ml 1. In the nano-gap experiment using the electrophoretic method, amine-modified fluorescent polystyrene particles (ex, 505 nm; em, 515 nm) with 1 mm in diameter (FluoSpheres Fluorescent Microspheres Amine-Modified Microspheres, Molecular Probes, Inc.) were also used as well as carboxylated particles of 1.75 and 3.0 mm in diameter. These particles were dispersed in an aqueous solution of 50 mm KCl

3 ANALYTICAL SCIENCES MAY 2009, VOL and 10 mm Tris HCl buffer (ph 7.9, S m 1 in conductivity) at a concentration of ml 1. Water was purified by a Milli-Q system (Millipore, Bedford, UK). Determination of particle diameter Interference was generated between the lights reflected at the upper surface of the lens and at the bottom of the flat glass, as shown in Fig. 2. Therefore, the gap distance, h, can be expressed as the enhancement (bright line) and the cancellation (dark line) of the interference light intensity, h = (2m + 1)l/4n and h = 2ml/4n, respectively, where l is the light wavelength (520 nm), m an integer (m = 0, 1, 2, ) and n the refractive index of the medium at the wavelength. Because the curvature of the lens was spherical, the gap distance, h, could be correlated with the lateral distance, r, from the center line of the lens by the equation h = h 0 + R R 2 r 2, (1) where R is the curvature radius of the lens and h 0 is the minimum gap distance at the center line of the lens. The values of h 0 and R could be determined experimentally by fitting the gap distance obtained from the interference Moiré pattern. Thus, one can precisely determine the gap, h, from the lateral distance, r (Fig. 2). It is also noted that the gap, h, can be changed by using different curvatures of the lens, R. For a spherical particle, the diameter of the trapped particle, d p, does not agree with the gap distance, h. The diameter of the trapped particle can be expressed as 16 d p = r2 2R + h0. (2) In this study, the diameter of each particle was determined by using Eq. (2) from the trapped position, r, which was obtained by microscope image analysis. Measurement of zeta potentials of PS particles Zeta potentials of the PS particles were determined from the electrophoretic mobility measurements by using a Briggs cell 18 (Mitamura Riken Kogyo Inc.) and an optical microscope (Labophot YF, Nikon) with a CCD system (WAT-100N, Watec, Japan). In this measurement, the PS particles were dispersed in an aqueous solution of 50 mm KCl, 1.0 wt% sucrose, and 10 mm Tris HCl buffer (ph 7.9, S m 1 in conductivity). Results and Discussion Hydrodynamic flow nano-gap method Figure 3(A) shows a typical flowing behavior of fluorescent polystyrene particles, where the velocities are increasing while approaching to the center line of the lens due to a decrease in the gap distance. Figure 3(B) demonstrates that the flow method can well trap 0.5 mm PS particles, which was not attained by the solvent evaporation method. In order to examine the sizedetermination performance of the hydrodynamic flow method, the three kinds of fluorescent polystyrene particles (0.5, 1.7, and 3.0 mm in diameter) were examined, as shown in Fig. 4. The particle diameters were independently determined from the scanning electron microscopy (SEM) images. It is clear that the histograms determined by the nano-gap method are in good agreement with those obtained by SEM images on the 0.5 and 1.75 mm particles; 0.58 ± 0.05 mm (N = 377) and 1.66 ± 0.15 mm (N = 352) by the nano-gap method agreed with 0.55 ± 0.03 Fig. 3 Hydrodynamic flow mode fractionation. (A) Sequential video images of the migration of fluorescent PS particles during the hydrodynamic injection flow. The velocity of the particles is accelerating in the center region of the nano-gap cell. (B) Interference Moiré pattern (R = mm) and the fluorescent image of the trapped 0.5 mm PS particles in 0.01 wt% Triton X-100 in water. mm (N = 197) and 1.69 ± 0.08 mm (N = 234) by SEM analysis, respectively. However, on the 3.0 mm particles, the size distribution obtained by the nano-gap method, 2.93 ± 0.10 mm (N = 134), was smaller than that from the SEM images, 3.10 ± 0.36 mm (N = 273). This is due to the fact that the 3.0 mm particles used are not absolutely spherical. Indeed, relatively large ellipticity was recognized for the 3.0 mm particles in SEM images. The diameter for the 3.0 mm polystyrene particles obtained by the nano-gap method, 2.93 ± 0.10 mm, was in good agreement with the averaged length of the minor axis of polystyrene particles, 2.95 ± 0.35 mm, obtained from the SEM images. Therefore, it was concluded that the polystyrene particles were trapped along the minor axis, as in the case of yeast cells. 16 In the present study, the utility of the hydrodynamic injection flow mode was confirmed. The accuracy of the observed diameters was estimated to be on the order of a few tens of nanometer.

4 608 ANALYTICAL SCIENCES MAY 2009, VOL. 25 Fig. 4 Comparison between the results measured by the nano-gap method (hydrodynamic flow mode) and the scanning electron microscope (SEM) method in a size analysis of fluorescent PS particles of (A) 0.5, (B) 1.7, and (C) 3.0 mm in diameters. The particle diameter distributions determined by SEM images are shown as black histograms, the average values by black figures, and the gray ones (colored in the color version availble on the web) by the nano-gap method. The observed average diameters by the two methods are shown in each figure. The average diameter measured by the nano-gap method for 3.0 mm PS is consistent with the diameter of the minor axis of PS (2.95 ± 0.35 mm) determined by SEM images. Electrophoretic nano-gap method A second particle transfer mode was developed in the present study by using the electrophoresis. Mixed particles suspended in a 50 mm KCl, 10 mm Tris HCl (ph 7.9) aqueous solution were injected to both sides of the nano-gap device, and then a dc voltage of 10 to 30 V was applied to the electrodes. A fluorescence image of the trapped particles in the gap is shown in Fig. 5(A). The 1.0 and 3.0 mm particles migrated from the anode (+) to the cathode ( ), and were trapped at the righthand side of the gap. On the contrary, 1.75 mm particles migrated from the cathode to the anode, and were trapped on the left-hand side of the gap. The migration directions of the particles are shown by arrows. The diameters of the three kinds of particle were determined to be 1.08 ± 0.05, 1.58 ± 0.08, and 2.95 ± 0.26 mm. It was interesting to note that the particles migrated faster as they moved closer to the center of the cylindrical lens. The dependence of the velocities of the three kinds of particles on the nano-gap distance is shown in Fig. 5(B). Each particle showed accelerated migration to the center line of the gap from both sides, and was suddenly trapped at a position corresponding to the diameter. The migration direction and the migration velocity of the particle were analyzed in terms of the zetapotential of the particle, and that of the nano-gap walls. Since the ph value of the sample solution was 7.9, the surface of the glass in the nano-gap device should be negatively charged, which results in an electro-osmotic flow of the aqueous medium in the direction from the anode to the cathode. The observed velocity, n obs, of the particle is considered to be a linear superposition of the electrophoretic and electroosmotic components of the system, expressed as n obs = n EP + n EOF, where n EP is the electrophoretic velocity of a particle and n EOF the electroosmotic velocity of the medium. This superposition can be expressed by using Smoluchowski equation, n obs = (z particle z wall) e h E = (zparticle zwall) e h A h, (3) where E is the electric field, z particle the zeta-potential of a particle, z wall the zeta-potential of the glass surface, e the dielectric constant of the medium, and h the viscosity of the medium. The electric field, E, in the nano-gap is inversely proportional to the gap distance, h, with the proportional constant A. By using Eq. (1), the following relation is obtained for the applied voltage, V: V = q b b Edr = Aq b b 1 h dr = Aq b b 1 Ih 0 + R R 2 r 2 J dr, (4) where the integral is calculated from b to b, which corresponds to the width of the cylindrical lens, since the voltage is applied between electrodes put on both sides of the cylindrical lens in the cell. Thus, the value of A can be calculated as a function of the applied voltage, V, the curvature of the cylindrical lens, R, the gap distance at the center of the lens, h 0, and the distance from the electrode to the center of the lens, b, as A = t + V 2k k 2 1 tan 1 O k2 1 k 1 where t = sin 1 (r/r) and k = h0 + R R. tani t 2 JP t = sin 1 (b/r) t = sin 1 ( b/r), (5)

5 ANALYTICAL SCIENCES MAY 2009, VOL particle), 75.8 ± 13.5 mv (1.75 mm particle), and 38.0 ± 9.9 mv (3.0 mm particle), respectively. Although the zeta-potential values obtained by the Briggs method are somewhat scattered, the observed zeta-potentials from the migration velocity in the nano-gap are in agreement with those independently measured by using a Briggs cell in the present study. Conclusions In conclusion, we have developed a simple and precise sizedetermination and size-fractionation method by using a controllable nanometer-sized wedge gap, and demonstrated their applicability for the analysis of micrometer to sub-micrometer sized particles in solution. The gap distance was precisely calibrated with the interference Moiré pattern, and an accurate real physical diameter of micro-particles in solution was determined. By combining the electrophoretic migration and the electro-osmotic flow, the size and the zeta-potential of the particles were also measured simultaneously. In principle, the proposed nano-gap method could be applied to nanometer-sized particles by using a larger curvature cylindrical lens. The size determination ability of this method is only limited by the precision of the surface roughness of the lens and the flat glass plate used for constructing the nano-gap. Therefore, the nanogap method will be useful as a microscale sample-preparation device for any kind of particles, including aerosol and bioparticles, which might be further combined with sensitive detection methods, like Raman microscopy, TOF/MS, ICP-MS and X-ray spectrometry. Acknowledgements This work was in part financially supported by Special Coordination Funds for Promoting Science and Technology: Yuragi Project of Osaka University from the Ministry of Education, Culture, Sports, Science and Technology of Japan. Supporting Information Fig. 5 (A) Fluorescence image of trapped polystyrene particles of 1.0, 1.75, and 3.0 mm in diameter in 50 mm KCl and 10 mm Tris HCl (ph 7.9) with the electrophoresis mode. (B) Dependence of the velocities of the particles on the nano-gap distance with h 0 = 0.67 mm in 20 V/cm. (C) Linear relationships between the velocity and the inverse nano-gap distance, h 1. For example, the value of A under the conditions of V = 20 V, R = 7.78 mm, h 0 = 1.35 mm, and b = 5 mm was V. The values of A are given in the supporting information as well as the derivation of Eq. (5). The slopes of the straight lines of Fig. 5(C) correspond to (z particle z wall)ea/h. Therefore, by using the literature value on the zeta-potential of the glass wall, z wall = 43.8 mv, 18 the zeta-potentials of the particles were determined to be 36.3 ± 4.9 mv (1.0 mm particle modified by amino group), 60.0 ± 3.4 mv (1.75 mm particle modified by carboxyl group), and 38.4 ± 1.3 mv (3.0 mm particle modified by carboxyl group). It is obvious that the reversed migration of the 1.75 mm particle is due to its largest negative zeta-potential. In this case, the electrophoretic velocity exceeded the electroosmotic velocity of the medium. The zeta-potentials obtained by using the conventional Briggs cell were 40.0 ± 6.4 mv (1.0 mm Supporting Information is available free of charge on the Web at concerning the value of A and the derivation of Eq. (5). References 1. J. C. Giddings, United Separation Science, 1991, John Wiley & Sons, Inc., NY. 2. T. Imasaka, Y. Kawabata, T. Kaneta, and Y. Ishizu, Anal. Chem., 1995, 67, A. Hirai, H. Monjushiro, and H. Watarai, Langmuir, 1996, 12, M. Washizu, S. Suzuki, O. Kurosawa, T. Nishizawa, and T. Shinohara, IEEE Trans, Ind. Appl., 1994, 30, S. Tsukahara, K. Yamanaka, and H. Watarai, Chem. Lett., 2001, T. Masuda and T. Okada, Anal. Chem., 2001, 73, M. E. Hovingh, G. H. Thompson, and J. C. Giddings, Anal. Chem., 1970, 42, H. Watarai and M. Namba, Anal. Sci., 2001, 17, i C. B. Fuh, Anal. Chem., 2000, 72, 266A. 10. Y. Iiguni and H. Watarai, J. Chromatogr., A, 2005, 1073, 93.

6 610 ANALYTICAL SCIENCES MAY 2009, VOL K.-G. Wahlund and A. Zattoni, Anal. Chem., 2002, 74, J. E. Lennard-Jones, Proc. Phys. Soc., 1931, 43, M. Hatta, H. Monjushiro, and H. Watarai, Chem. Commun., 2004, R. Hooke, Micrographia, or, Some Physiological Descriptions of Minute Bodies Made by Magnifying Glasses, 1st ed., 1665, J. Martyn and J. Allestry, London. 15. I. Newton, Optics, or a Treatise of Reflection, Refractions, Inflection, and Colours of Light, 1704, The Royal Society, London. 16. H. Monjushiro, M. Hatta, and H. Watarai, J. Chromatogr., A, 2006, 1106, ImageJ software, US Institutes of Health, Bethesda, MD, 2008; freely available on the website D. R. Briggs, Anal. Chem., 1940, 12, 703.