Delivered by Publishing Technology to: Korea Advanced Institute of Science & Technology (KAIST)IP: 143.248.30.96 On: Fri, 06 Sep 2013 06:49:51

Copyright: American Scientific Publishers

RESEARCH

ARTIC

LE

Copyright © 2013 American Scientific PublishersAll rights reservedPrinted in the United States of America

Journal ofNanoscience and Nanotechnology

Vol. 13, 7993–7997, 2013

Dielectrophoresis in a Slanted Microchannel forSeparation of Microparticles and Bacteria

Seong-Won Nam1, So Hyun Kim1, Je-Kyun Park2, and Sungsu Park1�∗1Department of Chemistry and Nano Sciences (BK21), Ewha Womans University, Seoul 120-750, Korea

2Department of Bio and Brain Engineering, Korea Advanced Institute of Science and Technology(KAIST), Daejeon 305-701, Korea

Dielectrophoresis (DEP) is an effective method to trap, manipulate and separate various dielectricparticles. To generate a DEP force, a spatially nonuniform electrical field has been generated by anarray of electrodes, while electrodeless DEP has been accomplished by placing an insulating mate-rial between two electrodes. Here, we describe a new DEP method for generating a nonuniformelectrical field using a slanted microchannel. The electric field gradient is induced due to a slope inthe channel and can be used to move and separate particles. Based on the gradual electric fieldinduced by three dimensional structure of the microchannel, our method enables particles of differ-ent sizes to be separated solely by DEP force without flow. The slanted microchannel was easilyfabricated by a replica molding technique using a commercial UV-cured photopolymer (NOA 63) andbonded as an insulating layer between two indium-tin-oxide films. By applying the electrical field,polystyrene beads of different sizes (6–45 �m in diameter) were trapped and separated depend-ing on the applied electric strength and frequency. Using this method, the opportunistic pathogenPseudomonas aeruginosa attached to antibody-conjugated microbeads was successfully separatedfrom Escherichia coli in a slanted microchannel.

Keywords: Dielectrophoresis, Slanted Microchannel, PDMS Stamp, NOA, Polystyrene Beads,Pseudomonas aeruginosa.

1. INTRODUCTION

Characterization and manipulation of micron particles areimportant for analyzing biomolecules. Laborious proce-dures such as filtration, analyzing and separation can bereduced by trapping particles. Moreover, reaction time anddetection limit can be lowered by trapping particles. In thisregard, dielectrophoresis (DEP) is a powerful tool to trapand separate particles.1 DEP refers to the force exerted onthe induced dipole moment of an uncharged and conduc-tive particle by a nonuniform electric field.2–5 Dependingon the conductivity of the particles and the fluid, particlesare attracted to higher electric field strength (called posi-tive DEP) or to lower electric field (called negative DEP).The particles react to the electric field differently depend-ing on the spatial distributions of the electric field, andthe electric properties of the particles and of surround-ing medium.1 In DEP migration, particles with differentdielectric properties migrate differently to strong or weakelectric field regions. In DEP retention, the DEP forcescompete with other fluid-flow forces, and thus particles

∗Author to whom correspondence should be addressed.

with large positive DEP forces are trapped. Meanwhile,particles are exposed to the field with different char-acteristic concentration profiles by field-flow fractiona-tion (FFF), and moved at different velocities under theinfluence of fluid flow. Therefore, DEP force can alsobe generated from different types of electrodes, togetherwith gravitational, magnetic and thermal fluids, etc. Sev-eral DEP-based separation techniques including DEP-FFF have been used to analyze various particles such aspolystyrene beads,6–8 viruses,9 bacteria,10 malaria-infectedblood cells11 and other cells.12�13

The standard way to generate DEP is to generatean electric field gradient with an array of planar elec-trodes in the presence of an alternating current (AC)field. Although this method has been well characterizedand commonly used, there have been some limitations inarranging electrodes; the fabrication of such electrodes iscostly and technically challenging. Recently, electrodelessor insulator-based dielectrophoresis (IDEP) without pat-terned metal electrodes has been introduced as an alter-native to conventional DEP.14–18 Electrodeless is definedas dielectrophoresis without metal surface at the positionwhere the sample is manipulated, although electrodes are

J. Nanosci. Nanotechnol. 2013, Vol. 13, No. 12 1533-4880/2013/13/7993/005 doi:10.1166/jnn.2013.8158 7993

Delivered by Publishing Technology to: Korea Advanced Institute of Science & Technology (KAIST)IP: 143.248.30.96 On: Fri, 06 Sep 2013 06:49:51

Copyright: American Scientific Publishers

RESEARCH

ARTIC

LE

Dielectrophoresis in a Slanted Microchannel for Separation of Microparticles and Bacteria Nam et al.

needed to apply the electric fields.14�15 Meanwhile, theinsulating posts induced a gradual distribution of electricfield in a channel, and a flow can be produced throughthe channel by applying of direct current (DC) field.15�16

Other IDEP methods exploit topographical structures suchas ridges etched in glass or silicon substrates in orderto generate a nonuniform electric field.14�17 Packing insu-lating materials such as oil in a microchannel induced alocally nonuniform DC field, which allows the particles tobe separated by size.18

In this paper, we report a novel method to generate anonuniform electric field in a microfluidic channel withgradually increasing channel depths. The slanted chan-nel was fabricated by a simple printing method in whicha polydimethyl siloxane (PDMS) master mold fabricatedby soft lithography was pressed against an UV-curablepolymer such as Norland Optical Adhesive 63 (NOA 63).The method has many advantages over other DEP andIDEP methods. First, the printing method does not requireany cumbersome steps for fabricating electrodes. Second,the channel is chemically inert and would not cause anyproblems, including troublesome steps caused by electrodefouling, filling materials and posts installed in microchan-nels. Third, particles can be separated without fluid flow,which can help to distinguish properties between parti-cles. To demonstrate the feasibility of the method for bio-analysis, the method was used to separate microbeads andbacteria.

2. EXPERIMENTAL DETAILS

2.1. Design and Fabrication of a Microfluidic Device

PDMS was obtained from Dow-Corning (Cortland, NY,USA) and used to fabricate a stamp with a microchan-nel pattern by soft lithography.19 In detail, SU-8 2035(MicroChem, Newton, MA, USA) was first coated ontoa silicon wafer. The wafer was inclined at 11� whileexposed to ultraviolet (UV) through a film mask with themicrochannel pattern. As a result, the wafer has a chan-nel with different depths of 119�8± 6�2 �m and 95�6±5�8 �m at the ends, as shown in Figure 1(a). The masterwas then passivated with vapor of tridecafluoro-1,1,2,2-tetrahydroctyl)-1-1-trichlorosilane for 15 min. Next, a 10:1mixture of PDMS prepolymer and curing agent (Sylgard184®� was casted onto the Si master and cured for 2 h at80 �C. The cured PDMS layer was peeled from the SU-8master, producing a PDMS stamp with the microchannelpattern.The pattern on the PDMS stamp was transferred onto

NOA 63 (Norland Company, Cranbury, NJ, USA) by areplica molding technique20–23 as shown in Figure 1(b).First, NOA 63 was poured over an indium tin oxide(ITO)-coated polyethylene terephthalate (PET; Sigma) film(60 �/sq.ft.). The PDMS stamp was then pressed into theNOA 63 layer. After the pressing step, both the PDMS

Fig. 1. Fabrication of a slanted microchannel. (a) The deepest part ofthe channel is 119.8 �m ± 6.2 �m and the shallowest part is 95.6 �m±5.8 �m. (b) Fabrication of a slanted microchannel using NOA 63 andITO films. Channel pattern was transferred from the PDMS mold ontoNOA 63 prepolymer. After UV curing (�= 365 nm) for 8 min, the PDMSmold released from the patterned NOA63 substrate. The second ITO filmwith inlet and outlet holes was attached to NOA 63 substrate by UVcuring (�= 365 nm) for at least 2 h.

stamp and NOA layer were cured under UV (�= 365 nm,135 mW/cm2� for 8 min. The cured NOA 63 layer wasthen peeled off the PDMS stamp. This short curing timeis enough to prevent the patterned structure from collaps-ing but still enables the layer to remain sticky enough forbonding to another ITO film. The second ITO film withholes for the inlet and outlet was first attached to the NOAlayer, and their bonding was then completed by curingthem under UV-light for 2 h (Fig. 1(b)).

2.2. Polystyrene Beads and Experimental Setup

Bare polystyrene beads (Polybead®� with different sizes(6, 15 and 45 �m) were purchased from PolySciences(Warrington, PA, USA). Phosphate buffered saline (PBS)was made of 0.1 M phosphate and 0.15 M NaCl andits electrical conductivity was adjusted to 22 �S/cm withdeionized (DI) water to be used as a separation buffer(pH 7.2). 5% bovine serum albumin (Sigma-Aldrich,St. Louis, MO, USA) was added to PBS for inhibitingnon-specific adsorption.A slurry-bead mixture was introduced into a microchan-

nel using a syringe. The electric field was producedfrom a function generator (MXG-9802A; METEX, Seoul,Korea). The movement of beads in microchannels wasobserved under a stereomicroscope (Nikon, Tokyo, Japan)and recorded by a CCD camera (SPOT INSIGHT™, Diag-nostic Instruments, Sterling Heights, MI, USA).

2.3. Bacterial Strain and Growth Condition

Escherichia coli O157:H7 43895 was obtained from theAmerican Type Culture Collection (Manassas, VA, USA),and Pseudomonas aeruginosa PAO1 was obtained fromProfessor You-Hee Cho (Cha University, Korea). Thestrains were grown in LB (BD, CA, USA) at 37 �C ina shaker at 220 r.p.m. Their Growth was monitored by

7994 J. Nanosci. Nanotechnol. 13, 7993–7997, 2013

Delivered by Publishing Technology to: Korea Advanced Institute of Science & Technology (KAIST)IP: 143.248.30.96 On: Fri, 06 Sep 2013 06:49:51

Copyright: American Scientific Publishers

RESEARCH

ARTIC

LE

Nam et al. Dielectrophoresis in a Slanted Microchannel for Separation of Microparticles and Bacteria

measuring the optical density at 600 nm using a spec-trophotometer (Biowave CO8000, WPA Ltd., Cambridge,England).

2.4. Conjugaton of Antibody to PolystyreneMicrobeads and Separation of Bacteria

For capturing bacteria, 15 �m bare polystyrene beads in0.1 M sodium phosphate buffer (pH 7.4) were reacted withanti-rabbit IgG (Sigma) in the same buffer by gentle shak-ing at RT overnight. P. aeruginosa PAO1 (O.D.= 0.44) inPBS was incubated with 20 �L of rabbit anti-P. aerug-inosa IgG (AbCam, Cambridge, UK) at 1:100 dilutionfor 1 h by end-over-end rocking and then washed threetimes with PBS by centrifugation at 574× g for 1 min.Anti-rabbit IgG-conjugated beads were then added into themixture and incubated for 1 h. The beads were then col-lected by centrifugation, and anti-rabbit IgG-TRITC (tetra-methyl rhodamine isothiocyanate, Sigma) was added andincubated for 1 h. Then, the beads were washed threetimes with PBS by centrifugation and resuspended in PBS(22 �S/cm).To test if the dielectrophoretic characteristics of the

beads depend on the binding of antibodies, the 15-�mpolystyrene beads conjugated with anti-mouse IgG fromgoat (0�54× 107/mL∼1�5× 104 IgG/particle) (SpherotechInc., Lake Forest, IL, USA) were incubated with diverseratios of rabbit anti-goat IgG-FITC (fluorescein isoth-iocyane) (3.4 mg/mL, Sigma) for 1 h by end-over-end rocking at RT. Fluorescence was observed underan inverted fluorescence microscope (Nikon EclipseTE2000-U) equipped with the CCD camera and analyzedby Image J Software (NIH, Bethesda, MD, USA).

3. RESULTS AND DISCUSSION

3.1. Simulation of Electric Field Distributionin the Slanted Channel

As shown in Figure 2(a), the channel was 47 mm long and400 �m wide, and the depth difference between both ends

Fig. 2. Characterization of a slanted microchannel. (a) The overall sizeof microchannel. (b) The simulated electric field strength in the chan-nel and vector diagram of the forces caused by dielectrophoretic effects(dotted lines indicate equilibrium lines depending on beads size whileapplying 7.4 V at 10 kHz; FDEP = DEP force, FINT = force interactionsbetween beads).

was 24.2 �m. The electric field distribution in the chan-nel was simulated with a commercial computational fluiddynamics program (CFD-ACE+; ESI Group, Huntsville,AL, USA) (Fig. 2(b)). The electric field strength dependson the channel depth, and decreases from the lower partto the higher part of the channel. The gradual distributionof the electric field strength is the driving force for move-ment and separation of particles. When the electric field isgenerated between the two ITO films, the DEP force canbe generated, and particles are affected by the DEP force,which is defined for the spherical particles as:2

FDEP = 2�r3mReK��� �E2 (1)

where r is a radius of the particle, m is the medium’spermittivity, E is the electrical field and K is the Clausius-Mossotti factor.To move particles, the DEP force must overcome the

drag force in a microchannel. For the special case ofsmall spherical objects moving through a fluid with smallReynolds number,

FDrag = 6��av (2)

where a is the stokes radius of a particle and � is the fluidviscosity.Therefore, the velocity by DEP force is proportional to

the square of the particle radius and electric field:

� ∝ r2�E2 (3)

Since E = V /d, where V is voltage and d is the distanceof the conductive layers, the velocity of the particle isproportional to the radius of particles and inversely pro-portional to the distance of the conductive layers. Thus,the beads move differently in an electric field dependingon their radius.5

Another crucial factor for DEP force is the frequencyapplied to the conductive layers, as shown in Eq. (1).In this case, the DEP force can be modulated by changingthe Clausius-Mossotti factor K, which can be adjusted bychanging the frequency of signals applied on the conduc-tive layers.20 Thus, the movement of beads depends on theapplied frequency as well.

� ∝ ReK��� �E2 (4)

3.2. Separation of Micron Beads with DEP Force

Polystyrene beads with two different diameters (15 �mand 45 �m) were mixed in a test tube containing PBSbuffer (permittivity = 22 �S/cm), and the mixture wasloaded in the slanted channel. When 4 V and 100 kHzwere applied to the ITO film, the 15 �m beads movedtoward the deep side of the microchannel while the 45 �mbeads did not move (Fig. 3(a)). When the applied volt-age was increased slowly, the velocity of the 15 �m beads

J. Nanosci. Nanotechnol. 13, 7993–7997, 2013 7995

Delivered by Publishing Technology to: Korea Advanced Institute of Science & Technology (KAIST)IP: 143.248.30.96 On: Fri, 06 Sep 2013 06:49:51

Copyright: American Scientific Publishers

RESEARCH

ARTIC

LE

Dielectrophoresis in a Slanted Microchannel for Separation of Microparticles and Bacteria Nam et al.

Fig. 3. Effect of voltage on bead movement. (a) After applying 4 V at100 Hz, 15 �m beads moved to the deepest end of the channel, while45 �m beads did not move. (b) After applying 7.4 V at 10 kHz, 6 �mbeads moved and aggregated to a heap. When 10 MHz was applied, thebeads dispersed again and moved to the deepest part of the channel.

increased. To move the 45 �m beads, higher voltage andlower frequency below 10 kHz was required (data notshown).When 7.4 V was applied at 10 kHz, the 6 �m beads

congregated in the channel (Fig. 3(b)). When the samevoltage was applied at a higher frequency (10 MHz), theheaps of beads dispersed and flowed through the channel(Fig. 3(b)). Dielectric interaction between two sphericalparticles in an electrolyte medium when applying a uni-form electric field was described in previous reports.2�24

When particles were aligned parallel to the electric field,the lowest electric field occurred between the particles,and thus the interaction between the particles becameattractive.24 The temporary aggregation of particles wasanalyzed as well, as shown in Figure 3(b), and the two-particle model could be adopted for explanation of aggre-gation of the particles.2

To increase the separation efficiency, diluted PBS(permittivity = 11 �S/cm) was used. When 2.1 V wasapplied at 1 kHz, the 6 �m beads and 15 �m beadsin the diluted PBS were successfully separated (Fig. 4).The separated equilibrium line depended on each bead’s

Fig. 4. Effect of voltage on bead separation. (a) After applying 2.1 Vat 1 kHz, 15 �m beads moved to the deepest part of the channel, while6 �m beads moved and aggregated to a heap. (b) The relation betweenthe particle position and velocity of 6 �m and 15 �m beads. The particleY -position corresponds to the channel width of (a).

radius, due to the nonuniformity of the electric field aty-position (Fig. 4(b)). In this condition, the dielectric forcewas exerted in three different directions, and the forcesbalanced each other depending on the radius of beads,applied field strength and frequency. When the dielectricforce moved the beads in the direction of the x-axis, the15 �m beads moved to the y-direction at the same time upto the lowest position of the gradually changing electricfield strength of the y-axis. On the other hand, the 6 �mbeads moved in the x-direction, were brought close to eachother, and then aggregated.

3.3. Separation of Bacteria Cells with DEP Force

To demonstrate the feasibility of the DEP technique forbioanalysis, the technique was used to separate pathogenicbacteria such as P. aeruginosa and E. coli. When bothP. aeruginosa and E. coli were incubated with anti-P. aeruginosa IgG (from rabbit), anti-rabbit IgG conju-gated beads and TRITC-anti-rabbit IgG in a sequentialmanner, red fluorescence from TRITC was observed onlyon the beads. This result suggests that P. aeruginosa cellswere specifically bound to the beads. When an electricfield at 10.8 V at 11.8 kHz was applied into the electrodeto separate P. aeruginosa-bound beads from E. coli cells,P. aeruginosa cells bound to the beads moved in the direc-tion of higher electric field (Fig. 5(a)) while E. coli labeledwith green fluorescent protein (E. coli-GFP) was confinedto the channel wall (Fig. 5(a)). These results indicate thatthe DEP technique can be used to specifically separateP. aeruginosa from E. coli.

Fig. 5. Separation of E. coli from P. aeruginosa attached to microbeadsusing DEP and the effect of the antibody on the movement of beads.(a) Rabbit anti-P. aeruginosa IgGs were first bound to P. aeruginosa andthe antibody molecules on P. aeruginosa were later bound to goat anti-rabbit immobilized microbeads. For fluorescent confirmation of bacterialbinding to the beads, anti-rabbit IgGs conjugated with TRITC (red fluo-rescence) were added into the P. aeruginosa–bead complex. After apply-ing 10.8 V at 11.8 kHz, E. coli cells (green) were confined to the channelwall while P. aeruginosa attached to microbeads were concentrated in themiddle of the channel. (b) The relation between the secondary antibodybound to the primary antibodies on beads and the frequency required formoving beads. 15 �m polystyrene beads with goat IgG were incubatedwith anti-goat IgG-FITC. While 7.4 V was applied at 1 Hz, beads movedto the shallowest end of channel.

7996 J. Nanosci. Nanotechnol. 13, 7993–7997, 2013

Delivered by Publishing Technology to: Korea Advanced Institute of Science & Technology (KAIST)IP: 143.248.30.96 On: Fri, 06 Sep 2013 06:49:51

Copyright: American Scientific Publishers

RESEARCH

ARTIC

LE

Nam et al. Dielectrophoresis in a Slanted Microchannel for Separation of Microparticles and Bacteria

The dielectric characteristics of the particles can bealtered by the surface properties. When the polystyrenebeads were conjugated with IgG, the beads moved in thedirection of higher electric field, while the beads withoutantibody did not move in any direction in the same con-ditions. This implies that the conjugated antibody affectsthe dielectric property of the beads. To test the effectof antibody on the mobility of beads, the beads wereincubated with the secondary antibody (developed from arabbit anti goat IgG) conjugated with fluorescein isothio-cyanate (FITC) molecules. After incubation with the sec-ondary antibody, the beads recovered the dielectrophoreticproperty, which was correlated to the number of secondaryantibodies added. When the beads were incubated with100- or 1000-fold secondary antibodies, the beads weremoved by 7.4 V and 1 Hz (Fig. 5(b)). Meanwhile, thebeads incubated with 104- or 105-fold secondary antibod-ies needed 7.4 V with 10 Hz or 100 Hz to move. Theincreased molecules of the secondary antibody relative tothe first antibodies on the bead needed higher frequencyto produce movement, and relatively less electric strength.This implies that the dielectric properties of the beads arenot directly correlated to the size of particles including theantibodies on the beads, but might depend on the surfaceproperties of the beads, because more than one secondaryantibody can bind on molecules of the first antibodies.In this range of frequency, the beads moved in the oppositedirection, since positive DEP was exerted on the beads.1

Our separation method using a slanted microchannnelhas several advantages in separating particles over otherDEP-based methods. The slanted channel can be easilyfabricated, and the method does not need patterned elec-trodes at the location of dielectrophoretic manipulation.Microbeads of different sizes can be separated from eachother without the use of flow. Our results suggest that themethod can be used for separation of micron-sized biopar-ticles, including microorganisms.

4. CONCLUSIONS

A slanted microchannel has been fabricated and tested fordielectrophoretic manipulation. In the channel, a gradualelectric field was generated in parallel and perpendicularto the channel in which DEP forces were balanced to eachother. Under this regime, beads with different size movedwith different velocities and equilibrium positions andaggregated in the microchannel. Since the gradual electricfield is distributed throughout the slanted microchannel,beads of different sizes can be separated without exter-nal flow, which offers certain advantages in mixing and

separating bioparticles such as cells and microorganisms.Thus, the slanted microchannel-based DEP method haspotential for bioapplications.

Acknowledgments: This research was supported bygrants (Public welfare and Safety research program#2012M3A2A1051681; Korea Science and EngineeringFoundation grant #2012-0001138; #2012012221) throughthe National Research Foundation of Korea (NRF). Thefirst author was supported by a Research Program Grantfor 2012 from Ewha Womans University.

References and Notes

1. P. R. C. Gascoyne and J. Vykoukal, Electrophoresis 23, 1973 (2002).2. T. B. Jones, Electromechanics of Particles, Cambridge University

Press, Cambridge (1995).3. T. R. Lee, Y. S. Chang, J. B. Choi, and Y. J. Kim, J. Nanosci.

Nanotechnol. 11, 281 (2011).4. D. F. Chen, W. H. Li, H. Du, and M. Li, J. Nanosci. Nanotechnol.

12, 3035 (2012).5. A. Lee, S. J. Lim, and D. J. Kang, J. Nanosci. Nanotechnol. 12, 1152

(2012).6. X.-B. Wang, J. Vykoukal, F. F. Becker, and P. R. C. Gascoyne, Bio-

phys. J. 74, 2689 (1998).7. M. Durr, J. Kentsch, T. Muller, T. Schnelle, and M. Stelzle, Elec-

trophoresis 24, 722 (2003).8. S. Choi and J.-K. Park, Lab Chip 5, 1161 (2005).9. N. G. Green, H. Morgan, and J. J. Milner, J. Biochem. Biophys.

Methods 35, 89 (1997).10. R. Holzel, Biochim. Biophys. Acta 1450, 53 (1999).11. P. Gascoyne, C. Mahidol, M. Ruchirawat, J. Satayavivad,

P. Watcharasit, and F. F. Becker, Lab Chip 2, 70 (2002).12. F. F. Becker, X.-B. Wang, Y. Huang, R. Pethig, J. Vykoukal, and

P. R. C. Gascoyne, Proc. Natl. Acad. Sci. USA 92, 860 (1995).13. S. M. Ha, W. Cho, Y. A, and S. Y. Hwang, J. Nanosci. Nanotechnol.

11, 4250 (2011).14. C.-F. Chou, J. O. Tegenfeldt, O. Bakajin, S. S. Chan, E. C. Cox,

N. Darnton, T. Duke, and R. H. Austin, Biophys. J. 83, 2170 (2002).15. E. B. Cummings and A. K. Singh, Anal. Chem. 75, 4724 (2003).16. B. H. Lapizco-Encinas, B. A. Simmons, E. B. Cummings, and

Y. Fintschenko, Anal. Chem. 76, 1571 (2004).17. L. M. Barrett, A. J. Skulan, A. K. Singh, E. B. Cummings, and G. J.

Fiechtner, Anal. Chem. 77, 6798 (2005).18. I. Barbulovic-Nad, X. Xuan, J. S. H. Lee, and D. Li, Lab Chip 6, 274

(2006).19. S.-W. Nam, D. Van Noort, Y. Yang, and S. Park, Lab Chip 7, 638

(2007).20. Y. Xia and G. M. Whitesides, Angew. Chem. Int. Ed. 37, 550 (1998).21. N. Y. Lee, J. R. Lim, M. J. Lee, S. Park, and Y. S. Kim, Langmuir

22, 7689 (2006).22. M. J. Lee, N. Y. Lee, J. R. Lim, J. B. Kim, M. Kim, H. K. Baik,

and Y. S. Kim, Adv. Mater. 18, 3115 (2006).23. L. Wang, L. A. Flanagan, N. L. Jeon, E. Monuki, and A. P. Lee, Lab

Chip 7, 1114 (2007).24. K. H. Kang and D. Li, Langmuir 22, 1602 (2006).

Received: 14 May 2012. Accepted: 20 December 2012.

J. Nanosci. Nanotechnol. 13, 7993–7997, 2013 7997