We report a separation of two different size particles in fluidic flow by an optical fiber. With a light of 1.55 μm launched into the fiber, particles in stationary water were massively trapped and assembled around the fiber by a negative photophoretic force. By introducing a fluidic flow, the assembled particles were separated into two different downstream positions according to their sizes by the negative photophoretic force and the dragging force acted on the particles. The intensity distribution of light leaked from the fiber and the asymmetry factor of energy distribution have been analysed as crucial factors in this separation. Poly(methyl methacrylate) particles (5-/10-μm diameter), SiO2 particles (2.08-/5.65-μm diameter), and SiO2 particles (2.08-μm diameter) mixed with yeast cells were used to demonstrate the effectiveness of the separation. The separation mechanism has also been numerical simulated and theoretical interpreted.
©2012 Optical Society of America
Particle and/or microbe separation from a mixture is an interesting work and has potential for further functional on-chip integration, such as sample preparation , cell incubation , biomedical and chemical analysis [3,4]. Efficient physical methods have been developed for separation of particles in fluidic flow via electrophoresis , dielectrophoresis [6–9], magnetophoresis [10,11], and acoustophoresis [12,13] based on homogeneous electric fields, inhomogeneous electric fields, inhomogeneous magnetic fields, and acoustic standing waves, respectively. In these methods, different size particles flow into different channels and thus are spatially separated . Another approach is termed as optical chromatography [15–17], which is an optical separation technique. In this technique, when an optical force induced by radiation pressure resulting from a mildly focused laser beam and the dragging force induced by a fluidic flow in the opposite direction are equal, different size particles remain stationary at different equilibrium position, and thus are separated. Recent works demonstrate that particles can also be trapped in water via negative photophoresis [18–20]. Photophoresis is induced by the asymmetry distribution of energy in volumes of particles, which can generate a local heat gradient in liquid environment near the surface of the particle. For low absorptive particles and high absorptive water at a given wavelength, the water close to the rear surface of the particle becomes hotter than that to the front surface, which results in a negative photophoretic force. Compared to conventional optical force, photophoretic force is usually orders of magnitude larger , so that it is considered as an effective technique for massive particle trapping and manipulation over a large scale region in liquids, with an advantage of less damage to the particles [22–25]. According to the previous works undertaken in stationary droplets of solutions [18–20], it is known that both the negative photophoretic force and the dragging force of liquid surrounding depend on the particle size. Therefore, in a fluidic flow, particles can also be trapped to downstream where the negative photophoretic force and the dragging force of flow keep in equilibrium and thus be separated. In this work, we report a use of negative photophoresis induced by light leaked from an optical fiber for separation of particles in fluidic flow.
Figure 1(a) shows a schematic for particle separation. The fluidic channel (width 500 μm and depth 120 μm) was fabricated in a quartz slide by femtosecond (fs) laser writing based on an amplified fs laser system (Spectra Physics, Hurricane-1K-X, pulse center wavelength 800 nm) with a writing speed of 160 µm/s. An optical fiber (1.2 μm in diameter, as shown in the inset of Fig. 1(a)), which was drawn from a commercial single-mode fiber by the flame heated technique, was placed in the U-groove channel (width 125 μm and depth 120 μm). The fabrication of the U-groove channel is the same as that of the fluidic channel. Three suspensions were prepared by diluting different size particles into deionized water (volume ratio of particles to water ~1:1,000) with the assistance of ultrasonic. One suspension is poly(methyl methacrylate) (PMMA) mixture in which 5- and 10-μm-diameter PMMA particles were suspended in water (Fig. 1(b)). Another is SiO2 mixture in which 2.08- and 5.65-μm-diameter SiO2 particles were suspended in water (Fig. 1(c)). The third suspension is a mixture in which 2.08-μm-diameter SiO2 particles and yeast cells were suspended in water (Fig. 1(d)). The suspension was pumped into the fluidic channel by a syringe pump with a push-pull configuration. To avoid or reduce the fluidic flow instability, relatively low flow velocities (less 14 μm/s) were applied in the fluidic channel by tuning the propelling and drawing speeds of the syringes.
In experiment, first, the suspension of 5- and 10-μm-diameter PMMA particles was used as the fluidic flow and was pumped into the fluidic channel by the syringe pump. Without light in the fiber, the two PMMA particles are randomly dispersed in the fluidic flow and moved with the flow. Figure 2(a) , as an example, shows the optical microscope image of the two size of PMMA particles in the fluidic flow (velocity vf = 4.5 μm/s). Once a light of 1.55 μm is launched into the fiber with 1.2-μm diameter, the light will be leaked out the fiber. In this case, the fiber acts as a line-shaped “light source” and a light intensity distribution region is formed in water. Therefore, the two particles with a very low absorption coefficient (≤ 1 × 10−3) to the 1.55-μm light will move towards the light region because of negative photophoresis . Each particle experiences two forces. One is the negative photophoretic force (Fph) induced by the light intensity distribution and the other one is the dragging force (Fd) induced by the fluidic flow. Therefore, each particle is trapped to a downstream position where the Fph and Fd are equal. Since the numerical value of the Fph and Fd depend on particle size, the trapped particles will be separated into different equilibrium positions according to their sizes, typically with larger retention distance (d) to the fiber for larger size particles. Figure 2(b), as an example, shows the optical microscope image of a separation of the 5- and 10-μm-diameter PMMA particles in the fluidic flow (vf = 4.5 μm/s) with optical power P = 120 mW launched and remained for 30 s. From Fig. 2(b), it can be seen that, the retention distances of the separated particles to the fiber are d1 = 46 μm and d2 = 72 μm for the 5-μm-diameter PMMA particles (yellow dashed line circle) and the 10-μm-diameter PMMA particles (crimson dashed line circle), respectively. Further experiment indicates that, with an increase of the flow velocity and the optical power unchanged, both the two distances d1 and d2 increase and the separation distance (d2 − d1) increase too. Figure 2(c) shows the optical microscope image of a separation of the 5- and 10-μm-diameter PMMA particles in a higher velocity of the fluidic flow (vf = 8 μm/s) with P = 120 mW launched and remained for 45 s. A larger separation distance (d2 − d1) = 44 μm was obtained. Detailed separation process with the flow velocity increased from vf = 4.5 to 8 μm/s is shown in Media 1. By further increasing the flow velocity over the peak of the photophoretic velocity, the Fph on the particles was totally counteracted by the Fd and the particles were flushed away by the fluidic flow. Of course, the limitation of the flow velocity and the peak of the photophoretic velocity can be overcome by increasing the launched optical power. For comparison, the trapping of both the 5- and 10-μm-diameter PMMA particles around the fiber in stationary water at the input optical power of 120 mW was also added (Fig. 2(d)). From Fig. 2(d) we can see that, due to the larger photophoretic velocity with larger size particle , some PMMA particles with 10-μm diameter were firstly assembled in the inner region (closer to the fiber) while some PMMA particles with 5-μm diameter were assembled in the outer region. However, the particle separation in stationary water was still not efficient.
To further investigate the impact of the flow velocity on the particles separation, the retention distances of the 5- and 10-μm-diameter PMMA particles to the fiber have been measured as a function of flow velocity and shown in Fig. 3 . It can be seen that, the retention distance exhibits an ascending trend with the increasing flow velocity and the separation becomes more obvious. As mentioned above, the fluidic flow was achieved via a syringe pump. Fluidic flow instability exists in such systems. Due to the non-uniformity of the flow velocities in the fluidic channel, the spread of a single species in the fluidic flow direction also exists in the system. Generally, at a higher flow velocity, the non-uniformity of flow velocities in the fluidic channel becomes larger, leading to a more obvious spread of particles. To show the spread of a single species in separation process, we measured the standard deviation (σ) to the retention distance of the 5- and 10-μm-diameter PMMA particles as a function of flow velocity, as also shown in Fig. 3. It can be seen that σ became larger with the increasing flow velocity. Further experiments show that at the same flow velocity, the number of particles in the retention distance also affects σ. At the beginning of trapping process, the number of trapped particles is small and the trapped particles are easily affected by the non-uniformity of the flow velocities in the fluidic channel, leading to an obvious spread of particles. With the increase of the particle number, the trapped particles are gradually concentrating at the retention distance and the impact of non-uniformity of the flow velocities becomes less, leading to a weakened effect of the spread of particles. After that, with a further increase of the particle number, the trapping region becomes larger and the spread of trapped particles will gradually become obvious again.
3. Analyses and discussions
For an isotropic homogenous PMMA particle in water radiated by the light of 1.55 µm, a negative photophoretic force Fph will be acted on it and can be expressed as 22]Fig. 4 .
According to Eq. (2), J is only governed by the normalized electric field amplitude within the particle and the size of particle. To calculate the J values, three-dimensional (3D) finite-difference time-domain (FDTD) simulations were performed. Figure 5 shows the simulated normalized electric field distribution within 1- to 12.5-µm-diameter particles radiated by the light of 1.55 µm. It can be seen that, each particle acts as a microlens and focuses the radiation to the apheliotropic side of the particle, leaving a hotter back surface and leading to a negative photophoretic force. Moreover, the electric field distribution on the apheliotropic side of the particle with larger diameter is wider. Figure 6 shows the calculated J according to the simulations. It can be seen that, the J value increases with increasing of the particle diameter. The corresponding values of J for the 5- and 10-µm-diameter PMMA particles are 14.0 × 10−3 and 29.0 × 10−3, respectively.
In fluidic flow, each particle experiences a resultant force of Fph and Fd. Since the Fph and the Fd are in parallel but in reverse direction, so the particles can be trapped at a downstream position (named equilibrium position) where the two forces are equal. At the equilibrium position, Fd can be expressed as Fd = 6πμRvf using the Stokes law, where μ is the viscosity of water. So Fph can be rewritten as
The intensity I at a distance d to the fiber can be expressed as I = I0 e–αd, where I0 is the optical intensity at the surface of the fiber, α is attenuation coefficient along the radial direction of the fiber. The two values can be estimated by simulating its optical field distribution based on 3D FDTD. According to Eqs. (1) and (3), at the equilibrium position, the retention distance d can be derived as
According to Eq. (4), it is known that as long as the material of particle, the water solution, and the flow velocity are determined, d is only governed by the factor J. As shown in Fig. 6, a larger particle has a larger J, so the d is larger and thus the mixed particles can be separated. By substituting the parameters into Eq. (4), we estimated that the theoretical retention distances of the 5- and 10-μm-diameter PMMA particles to the fiber as a function of flow velocity, as shown in Fig. 7 . It should be pointed out that the value of J gradually becomes larger with the cross-section of the field profile of the leaking light changed from closing to the one of a spherical wave to the one of a plane wave . For comparison, experimental data were also shown in Fig. 7. It can be seen that the theoretical retention distance also exhibits an ascending trend with the increasing flow velocity, implying that the separation becomes more obvious. The experimental results agree with the theoretical estimation.
To better understand the impact of the flow velocity on particle separation, schematic illustrations are given in Fig. 8 . The red area indicates the radiation of the leaked light from the fiber. The green and blue solid lines indicate the Fph distributions on the 5-and 10-μm-diameter PMMA particles , respectively, while the green and blue dashed lines indicate the distributions of resultant force Fr (Fph + Fd) on the 5- and 10-μm-diameter PMMA particles, respectively. Figure 8(a) schematically shows the separation of the two particles with lager retention distance d2 for the larger particle. The separation distance is (d2 − d1). With an increase of the flow velocity while keeping the optical power unchanged, the resultant force decreases accordingly but the decrement is larger for larger particles. Therefore, the two retention distances d1 and d2 increase, and the separation distance of (d2 − d1) increases too as schematically shown in Fig. 8(b).
4. Separation of other particles and microbes
The separation has also been demonstrated using other particles and microbes. Figure 9(a) shows the optical microscope image for the separation of 2.08- and 5.65-μm-diameter SiO2 particles (refractive index of 1.44 at wavelength of 1.55 μm) in fluidic flow (vf = 4.5 μm/s) by a 4.2-μm-diameter fiber with the light of 1.55 μm launched for 40 s at P = 150 mW. It can be seen that, the retention distances of the separated particles are d1 = 70 μm and d2 = 95 μm for the 2.08-μm-diameter SiO2 particles (yellow dashed line circle) and the 5.65-μm-diameter SiO2 particles (crimson dashed line circle), respectively. Figure 9(b) shows that, by increasing the flow velocity to vf = 8 μm/s, a larger separation distance (d2 − d1) = 40 μm was obtained for the 2.08- and 5.65-μm-diameter SiO2 particles. Figure 9(c) shows the optical microscope image for the separation of 2.08-μm-diameter SiO2 particles and yeast cells in fluidic flow (vf = 4.5 μm/s) by a 5.8-μm-diameter fiber with the light of 1.55 μm launched for 40 s at P = 160 mW. It can be seen that, the retention distances of the separation are d1 = 85 μm and d2 = 115 μm for the 2.08-μm-diameter SiO2 particles (yellow dashed line circle) and the yeast cells (crimson dashed line circle), respectively. Figure 9(d) shows that, by increasing the flow velocity to vf = 8 μm/s, a larger separation distance (d2 − d1) = 50 μm was obtained.
We have demonstrated a separation of two different size particles in fluidic flow by an optical fiber with the assistance of the light of 1.55 μm. With the combination of negative photophoretic force generated by the light leaked from the optical fiber and the dragging force induced by the fluidic flow, particles and/or microbes mixtures (5-/10-µm PMMA, 2.08-/5.65-µm SiO2, 2.08-µm SiO2/yeast cell) were successfully separated. The intensity distribution of light leaked from the fiber and the asymmetry factor of energy distribution were analysed. The separation mechanism was simulated and interpreted. The separation brings a potential application for the combination of photophoretic trapping with microfluidic systems, and provides a useful tool to separate micro/submicro-particles and/or microbes for lab-on-a-chip analysis and preparation.
This work was supported by the National Natural Science Foundation of China (Grants 61007038, 60625404 and 10974261).
References and links
3. E. A. Schilling, A. E. Kamholz, and P. Yager, “Cell lysis and protein extraction in a microfluidic device with detection by a fluorogenic enzyme assay,” Anal. Chem. 74(8), 1798–1804 (2002). [CrossRef] [PubMed]
4. M. A. McClain, C. T. Culbertson, S. C. Jacobson, N. L. Allbritton, C. E. Sims, and J. M. Ramsey, “Microfluidic devices for the high-throughput chemical analysis of cells,” Anal. Chem. 75(21), 5646–5655 (2003). [CrossRef] [PubMed]
6. B. G. Hawkins, A. E. Smith, Y. A. Syed, and B. J. Kirby, “Continuous-flow particle separation by 3D Insulative dielectrophoresis using coherently shaped, dc-biased, ac electric fields,” Anal. Chem. 79(19), 7291–7300 (2007). [CrossRef] [PubMed]
7. J. J. Zhu, R. C. Canter, G. Keten, P. Vedantam, T. R. J. Tzeng, and X. C. Xuan, “Continuous-flow particle and cell separations in a serpentine microchannel via curvature-induced dielectrophoresis,” Microfluid. Nanofluid. 11(6), 743–752 (2011). [CrossRef]
8. C. Zhang, K. Khoshmanesh, F. J. Tovar-Lopez, A. Mitchell, W. Wlodarski, and K. Klantar-Zadeh, “Dielectrophoretic separation of carbon nanotubes and polystyrene microparticles,” Microfluid. Nanofluid. 7(5), 633–645 (2009). [CrossRef]
10. S. H. S. Lee, T. A. Hatton, and S. A. Khan, “Microfluidic continuous magnetophoretic protein separation using nanoparticle aggregates,” Microfluid. Nanofluid. 11(4), 429–438 (2011). [CrossRef]
11. T. T. Zhu, F. Marrero, and L. Mao, “Continuous separation of non-magnetic particles inside ferrofluids,” Microfluid. Nanofluid. 9(4-5), 1003–1009 (2010). [CrossRef]
12. F. Petersson, A. Nilsson, C. Holm, H. Jönsson, and T. Laurell, “Continuous separation of lipid particles from erythrocytes by means of laminar flow and acoustic standing wave forces,” Lab Chip 5(1), 20–22 (2005). [CrossRef] [PubMed]
15. T. Imasaka, Y. Kawabata, T. Kaneta, and Y. Ishidzu, “Optical chromatography,” Anal. Chem. 67(11), 1763–1765 (1995). [CrossRef]
16. S. J. Hart and A. V. Terray, “Refractive-index-driven separation of colloidal polymer particles using optical chromatography,” Appl. Phys. Lett. 83(25), 5316–5318 (2003). [CrossRef]
17. S. J. Hart, A. V. Terray, T. A. Leski, J. Arnold, and R. Stroud, “Discovery of a significant optical chromatographic difference between spores of Bacillus anthracis and its close relative, Bacillus thuringiensis,” Anal. Chem. 78(9), 3221–3225 (2006). [CrossRef] [PubMed]
18. H. X. Lei, Y. Zhang, X. M. Li, and B. J. Li, “Photophoretic assembly and migration of dielectric particles and Escherichia coli in liquids using a subwavelength diameter optical fiber,” Lab Chip 11(13), 2241–2246 (2011). [CrossRef] [PubMed]
22. C. Y. Soong, W. K. Li, C. H. Liu, and P. Y. Tzeng, “Theoretical analysis for photophoresis of a microscale hydrophobic particle in liquids,” Opt. Express 18(3), 2168–2182 (2010). [CrossRef] [PubMed]
23. A. Hirai, H. Monjushiro, and H. Watarai, “Laser photophoresis of a single droplet in oil in water emulsions,” Langmuir 12(23), 5570–5575 (1996). [CrossRef]
24. H. Monjushiro, M. Tanaka, and H. Watarai, “Periodic expansion-contraction motion of photoabsorbing organic droplets during laser photophoretic migration in water,” Chem. Lett. 32(3), 254–255 (2003). [CrossRef]
25. Y. L. Xu, B. Å. S. Gustafson, F. Giovane, J. Blum, and S. Tehranian, “Calculation of the heat-source function in photophoresis of aggregated spheres,” Phys. Rev. E Stat. Phys. Plasmas Fluids Relat. Interdiscip. Topics 60(22 Pt B), 2347–2365 (1999). [CrossRef] [PubMed]