The purely refractive index driven separation of uniformly sized polystyrene, n = 1.59 and poly(methylmethacrylate), n = 1.49 in an optical chromatography system has been enhanced through the incorporation of a custom poly(dimethysiloxane) (PDMS) microfluidic system. A customized channel geometry was used to create separate regions with different linear flow velocities tailored to the specific application. These separate flow regions were then used to expose the entities in the separation to different linear flow velocities thus enhancing their separation relative to the same separation in a constant velocity flow environment. A microbiological sample containing spores of the biological warfare agent, Bacillus anthracis, and a common environmental interferent, mulberry pollen, was investigated to test the use of tailored velocity regions. These very different samples were analyzed simultaneously only through the use of tailored velocity regions.
©2005 Optical Society of America
Since the discovery that momentum transferred from laser radiation to microscopic entities could be used as a method to directly control and manipulate them, a growing number of techniques, devices and applications have followed.1 Currently visible and infrared laser radiation has and is being applied as an elegant link from the macroscopic to the microscopic world for the manipulation of individual microscopic particles suspended in a fluid solvent. This method of optical manipulation is most commonly termed optical trapping.2 The evolution of optical trapping research from its beginnings in 1970 has yielded several advanced methods for multi-particle control and actuation.3, 4 These methods represent significant advancement in direct micro- and nano-manipulation and control for biological and chemical research on smaller and smaller length scales.5–7 Although optical trapping has been applied as a technique to investigate and interrogate single and multi-particle suspensions, only recently has optical trapping been directed to particles within microfluidic flow.8–10
Microfluidic research has seen an explosion of attention and progress in the past few years owing to the extraordinary advantages fluidic system miniaturization promises.11, 12 Miniaturized systems boast faster analysis of less sample volume with greater resolution than corresponding macroscopicsystems resulting in faster, cheaper, and better analysis platforms. Add to this the ability to rapidly replicate and parallelize multiple system processes and clearly microfluidics has great potential for progress in chemical and biological research, chemical and biological sensors,13 DNA analysis,14, 15 and protein crystal growth.16 These are just some examples of the areas in which microfluidics research is already making great strides.
The work presented here involved the combination of an optical manipulation technique and a microfluidic system to achieve a hybrid device capable of separating two identically sized polymer microspheres differing only in refractive index. This separation was enhanced by changing the fluidic design to create tailored velocity environments. The system was also used for the simultaneous separation of samples in different optical separation regimes (i.e. normally not separable under the identical conditions).
In our case, a laser was lightly focused by a long focal length lens (i.e. 75mm) rather than a high numerical aperture objective (NA > 1.3), as used in optical trapping. The resulting effect on a particle of higher refractive index than the surrounding medium near the focal point is to draw the particle toward the beam center and then push it in the direction of laser propagation. If this laser is focused into a fluid flow traveling in the opposite direction to laser propagation, particles in the fluid flow that encounter the beam near the focal point become trapped. Trapped particles are pushed against the fluid flow and stop when the optical and fluidic forces on the particle are balanced. The distance traveled from the focal point is termed the retention distance (Z). This method of optical separation is termed Optical Chromatography (OC) and has been well described elsewhere.17–19 Using this technique it has been shown that particles can be separated by not only size but also by refractive index, a physical property which can be related to their chemical composition.20
A diagram of our OC system, shown in Fig. 1, consists of either a 1064nm ytterbium fiber laser (IPG Photonics, Oxford, MA, USA) or an argon ion laser (Innova 100, Coherent, Palo Alto, CA) operating at 515 nm focused by a 1 in. diameter plano-convex 75mm lens (Thorlabs, Inc., Newton, NJ, USA) into a microfluidic system. The aligned system was placed underneath a microscope and mounted (Leica Microsystems, Model DMRX, Wetzlar, Germany,) on a custom x-y-z translation platform. This platform allowed for movement of the aligned OC system under the microscope used for observation and image data collection through a CCD mounted on the top of the microscope. Fluid flow was achieved using gravity feed and a syringe pump. The syringe pump was used for the occasional increase in flow required to rinse the flowcell where the gravity feed system was used to obtain extremely stable, pulseless flow. The flow rate from gravity feed was adjusted by changing the height difference between the inlet and exit reservoirs. Two 1 mL syringes were also connected to the microfluidic system for sample injection. The entire microfluidic system, shown in Fig. 2, consisted of a polycarbonate flow introduction base allowing for inlet, outlet and injection tubing connections, a poly(dimethylsiloxane) (PDMS) channel network adhered to the base and a glass cover slip bonded to enclose the PDMS fluidic network.
The PDMS fluidic network was fabricated using Soft Lithography.21 In our case, a thick (> 500 μm) layer of SU-8 negative photoresist (NANO™ SU-8 2100, MicroChem Corp., Newton, MA, USA) was used.22 Our photomask consisted of a design printed to film (Linotronic 530) and the exposure source was 365 nm ultra-violet (UV) light from a lamp (SpectroLine SB-100PC, Spectronics Corp., Westbury, NY, USA). The resulting PDMS replicas, cast from the developed photoresist relief, were irreversibly bonded to a glass coverslip by oxidizing the exposed glass and PDMS surfaces with a Tesla coil before bringing them into contact. The entire unit was then heated on a hot plate at 70° C for two minutes to strengthen the bond.23 A permanently bonded unit was then placed in contact with the polycarbonate base to achieve a reversible bond holding the unit together. This complete unit was then placed in the OC system, filled with a scattering fluid (≈ 0.1% Glycogen) and the laser aligned into the fluidic channel.
The PDMS channel network design is shown in Fig. 2. The design incorporates a linear channel with two smaller injection channels near where the laser enters the flowcell. This design accommodates the laser, incorporates a channel several millimeters in length for separations and two fluidic channels for a maximum of two isolated sample injections. The injection ports are located near the turn in the flow channel such that an injection is introduced near or below the focal point of the laser. This method of sample injection allows for a sample to be introduced to the OC beam in a more direct manner rather than relying on upstream injections that may or may not pass through the focal point. The resulting trapped particles are directed into the clean, particulate free, flow above the focal point, and are able to reside in the OC beam for an extended period of time undisturbed. With this design, either pure injections or mixture injections from either of the two ports can be made. To identify the location of new analytes, pure injections are made, their resulting retention distances are recorded and when a mixture sample is injected the pure injection Z values are used to identify the particulates in a mixture separation.
The ultimate capacity of this system to separate dense colloidal suspensions will be limited by degradation of the beam when large numbers of particles are downstream of the focal point. This may cause retention distance instability and if excessive numbers are present, an inability to trap particles. The solution would be to position the focal point close to where the flow turns to exit the flowcell, thus preventing the optical disruption of the beam. In addition, a high density of particles may also result in band broadening (many particles competing for similar Z location), thus limiting the resolution of similar types of particles; this limits the capacity for separating large groups simultaneously.
The modification of the OC system makes it possible for flow velocities along the diverging laser beam to be tailored to affect changes in existing separations. Because the soft lithography fabrication method relies primarily on the mask for the design, variations from the common straight edge designs are easily created. Fig. 3 is a movie which shows a ripple channel with a 10 μm polystyrene tracer particle in the higher velocity region. The movie shows how linear velocity changes as the particle travels through the narrow and wide channel regions. In this example the minimum channel width of 500 μm and maximum channel width of 800 μm resulted in a minimum and maximum linear velocity for the given flow rate of 176 and 225 μm/s respectively, as seen in the plot in Fig. 4 of linear velocity versus time for the particle traveling between the rippled regions of the channel.
The separation of individual 2.2 μm diameter polystyrene (PS) and polymethylmethacrylate (PMMA) microspheres (Magsphere, Inc., Pasadena, CA, USA) with refractive indexes of 1.49 and 1.59 respectively was achieved using 1.00 W of incident laser power at 1064 nm. This initial separation was conducted in a channel 500 μm in width having a linear flow velocity of 58 μm/s which resulted in a separation of 708 μm ± 30 μm. An enhancement of this initial separation under identical flow and laser power conditions was achieved by positioning the separation across a sudden increase in the channel width. This was done by moving the focusing lens with a linear translator and moving the focal point and hence the particles through the rippled regions of the flowcell. The increase in channel width results in a lower fluid linear velocity in that region. In this experiment the PS particle is placed in the lower linear flow velocity region upstream, which allows the PS particle in the separation to increase its retention distance (to maintain the optic-fluidic force balance) and thus enhance its separation from the PMMA particle remaining in the higher linear flow velocity region.
As seen in Fig. 5, three different width increases from the initial 500 μm channel size were tested to enhance the PS and PMMA separation. Enhanced separation distances (ΔZ) of 97, 257 and 407 μm were observed for width increases from 500 μm to 630, 750 and 870 μm respectively. The corresponding linear flow velocities for these width increases were 47, 40 and 36 μm/s respectively. It is important to note that the separation distance for a straight (non-enhanced) 870 μm wide channel is comparable to the enhanced separation for the 630 μm width increase; Data are given in Table 1. This demonstrates that a separation due to an overall decrease in linear flow velocity affecting both particles from 58 μm/s, found in the 500 μm wide channel, to 36 μm/s, found in the 870 μm wide channel, can also be achieved at overall higher flow rates when the particles are subject to the different individual linear flow velocities. A video showing the increasing and decreasing separation of PS and PMMA particles as they are translated through a rippled flowcell is given in Fig. 6.
While tailored velocity environments are useful for enhancing separations, in some cases it may be advantageous to reduce the separation between very different analytes. Optical chromatography and trapping of particles with greatly differing refractive indices and/or size can prove difficult. Good, stable retention of each particle dictates optimum conditions; when the particles being studied are too different neither can be trapped under the perfect conditions (i.e. laser power and flow rate). Additionally, the separation may not be achievable due to the divergence of the laser beam. The expanding beam results in decreasing photon density causing the upstream particle to eventually “fall out” or wander out of the beam at extreme retention distances. In such a situation, creating a tailored flow environment enables the simultaneous trapping of both separated species. A PDMS flowcell was created with a wide (low velocity) region that tapers into a narrow (high velocity) region. The beam travels from the wide region to the narrow region as it propagates upstream against the liquid flow. This increases the velocity in the upstream region while maintaining a lower fluid velocity in the wide channel area.
The tailored velocity flowcell was used to separate two very different particles: Bacillus anthracis (B.a.) spores (Sterne strain, or vaccine strain, grown and sporulated in house) and a common environmental interference, Mulberry pollen (Duke Scientific, Palo Alto, CA). The flowcell with the separated particles can be seen in Fig. 7. The size difference between these particles is significant, B.a. has a width near 1 μm and a length of approximately 2 μm; Mulberry pollen is approximately 14 μm in diameter. Under the same conditions it is difficult to retain them simultaneously in a non-tailored flowcell. Performing the separation in a tailored velocity flowcell resulted in a stable separation of only 1.11 mm (B.a. Z = 0.57 mm and pollen Z = 1.68 mm). This is despite that fact that the linear velocity experienced by the pollen particle (182 μm/s) was 1.9× that of the linear velocity acting on the B.a. spores (97 μm/s). Theoretical estimates17,20 of retention distance for mulberry pollen and B.a. under the lower linear velocity (i.e. 97 μm/s in a straight channel) are 2.9 mm and 0.5 mm respectively. This indicates that a substantial reduction in the separation (ΔZ = 1.1 mm with the tailored velocity flowcell versus ΔZ = 2.4 mm in a straight channel) between these particles was achieved through manipulation of the flow profile.
In conclusion, it has been demonstrated here that PDMS microfluidic devices can be coupled with a mildly focused laser to create an OC system. The fabrication flexibility, speed and quality that the soft lithography technique makes possible has allowed for the creation of an OC system with enhanced capabilities. This is demonstrated by the fabrication of a flowcell incorporating custom injection channels and a tailored velocity environment. This environment was used to create an enhanced separation between PS and PMMA particles and to augment the normal separation between B. anthracis and mulberry pollen. The ease with which a flowcell can be tailored to a specific separation is significant and should allow for future enhancement of much smaller separations as the ability to tailor finer linear flow velocity environments will increase as the channel dimensions decrease.
The authors would like to acknowledge the Office of Naval Research (ONR), the Naval Research Laboratory (NRL) and the Defense Threat Reduction Agency (DTRA) for support of this research.
References and links
1 . A. Ashkin , “ History of optical trapping and manipulation of small-neutral particle, atoms, and molecules ,” IEEE J. Sel. Top. Quantum Electron. 6 , 841 – 856 ( 2000 ). [CrossRef]
2 . A. Ashkin , “ Acceleration and Trapping of Particles by Radiation Pressure ,” Phys. Rev. Lett. 24 , 156 – 159 ( 1970 ). [CrossRef]
3 . H. Melville , G. F. Milne , G. C. Spalding , W. Sibbett , K. Dholakia , and D. McGloin , “ Optical trapping of three-dimensional structures using dynamic holograms ,” Opt. Express 11 , 3562 – 3567 ( 2003 ), http://www.opticsexpress.org/abstract.cfm?URI=OPEX-11-26-3562 [CrossRef] [PubMed]
4 . C. Mio , T. Gong , A. Terray , and D. W. M. Marr , “ Design of a scanning laser optical trap for multiparticle manipulation ,” Rev. Sci. Instrum. 71 , 2196 – 2200 ( 2000 ). [CrossRef]
5 . N. Kitamura and F. Kitagawa , “ Optical trapping - chemical analysis of single microparticles in solution ,” J. Photochem. Photobiol. C 4 , 227 – 247 ( 2003 ). [CrossRef]
6 . T. E. Bridges , M. P. Houlne , and J. M. Harris , “ Spatially resolved analysis of small particles by confocal Raman microscopy: Depth profiling and optical trapping ,” Anal. Chem. 76 , 576 – 584 ( 2004 ). [CrossRef] [PubMed]
7 . J. Plewa , E. Tanner , D. M. Mueth , and D. G. Grier , “ Processing carbon nanotubes with holographic optical tweezers ,” Opt. Express 12 , 1978 – 1981 ( 2004 ), http://www.opticsexpress.org/abstract.cfm?URI=OPEX-12-9-1978 [CrossRef] [PubMed]
9 . A. Terray , J. Oakey , and D. W. M. Marr , “ Fabrication of linear colloidal structures for microfluidic applications ,” Appl. Phys. Lett. 81 , 1555 – 1557 ( 2002 ). [CrossRef]
10 . P. A. Prentice , M. P. MacDonald , T. G. Frank , A. Cuschieri , G. C. Spalding , W. Sibbett , P. A. Campbell , and K. Dholakia , “ Manipulation and filtration of low index particles with holographic Laguerre-Gaussian optical trap arrays ,” Opt. Express 12 , 593 – 600 ( 2004 ), http://www.opticsexpress.org/abstract.cfm?URI=OPEX-12-4-593 [CrossRef] [PubMed]
11 . S. C. Jakeway , A. J. de Mello , and E. L Russell , “ Miniaturized total analysis systems for biological analysis ,” Fresenius’ J. Anal. Chem. 366 , 525 – 539 ( 2000 ). [CrossRef]
13 . C. Jensen-McMullin , A. Au , J. Quinsaat , E. R. Lyons , and H. P. Lee , “ Fiber-optic-based optical trapping and detection for lab-on-a-chip (LOC) applications ,” Proc. SPIE - Int. Soc. Opt. Eng. (USA) 4622 , 188 – 194 ( 2002 ).
15 . Y. J. Liu , C. B. Rauch , R. L. Stevens , R. Lenigk , J. N. Yang , D. B. Rhine , and P. Grodzinski , “ DNA amplification and hybridization assays in integrated plastic monolithic devices ,” Anal. Chem. 74 , 3063 – 3070 ( 2002 ). [CrossRef] [PubMed]
18 . T. Imasaka , Y. Kawabata , T. Kaneta , and Y. Ishidzu , “ Optical Chromatography ,” Anal. Chem. 67 , 1763 – 1765 ( 1995 ). [CrossRef]
19 . T. Imasaka , “ Optical chromatography. A new tool for separation of particles ,” Analusis 26 , M53 – M55 ( 1998 ). [CrossRef]
20 . S. J. Hart and A. V. Terray , “ Refractive-index-driven separation of colloidal polymer particles using optical chromatography ,” Appl. Phys. Lett. 83 , 5316 – 5318 ( 2003 ). [CrossRef]
21 . D. C. Duffy , J. C. McDonald , O. J. A. Schueller , and G. M. Whitesides , “ Rapid prototyping of microfluidic systems in poly(dimethylsiloxane) ,” Anal. Chem. 70 , 4974 – 4984 ( 1998 ). [CrossRef] [PubMed]
22 . C. H. Lin , G. B. Lee , B. W. Chang , and G. L. Chang , “ A new fabrication process for ultra-thick microfluidic microstructures utilizing SU-8 photoresist ,” J. Micromech. Microeng. 12 , 590 – 597 ( 2002 ). [CrossRef]