Article pubs.acs.org/ac
Exploiting Particle Mutual Interactions To Enable Challenging Dielectrophoretic Processes Mario A. Saucedo-Espinosa and Blanca H. Lapizco-Encinas* Microscale Bioseparations Laboratory, Rochester Institute of Technology, Rochester, New York 14623, United States S Supporting Information *
ABSTRACT: Dielectrophoresis (DEP) is the motion of particles under the influence of a nonuniform electric field. In insulator-based dielectrophoresis (iDEP), the required nonuniform electric fields are generated with insulating structures embedded in a microchannel. These structures distort the electric field distribution when an electric potential is applied. This contribution presents an experimental characterization of the electrokinetic (EK) and DEP velocities of a set of target particles, under DC potentials, when additional innocuous particles are used as fillers. Streak-based particle velocimetry in a tapered channel was used to assess particle motion. Filler particles of various sizes were added at different volume fractions (ϕ) to suspending media containing the target particles/cells. The presence of the filler particles resulted in electric field distortions and dissimilar particle behaviors caused by particle−particle interactions. These particle mutual interactions were exploited to improve the enrichment of low-abundance yeast cells in an iDEP channel. It was shown that the smallest studied filler particles (500 nm) have the potential to aid the enrichment of low-abundance yeast cells when filler volume fractions ∼1 × 10−5 v/v are used. Enrichment factors of ∼115 were achieved by applying electric potentials as low as 500 V. icrofluidics has revolutionized the field of analytical assessments.1,2 Miniaturized electric-field-driven techniques have proved to be robust and efficient methods for particle manipulation.3 Insulator-based DEP (iDEP) is a dielectrophoretic mode that employs insulating structures to distort the electric field distribution, creating electric field gradients.4 Under the effect of DC potentials, particles in iDEP devices are also subjected to electrophoresis (EP) and electroosmotic (EO) flow. These phenomena have a linear dependency on the electric field, and their superposition can be considered as a single electrokinetic (EK) transport term. The motion of particles in iDEP systems results from the interplay between EK and DEP, and is a function of the electric potential.5 Particles are captured or trapped in specific regions in an iDEP device when DEP overcomes EK.6 Several studies have shown the capabilities of DEP for the sorting, enrichment, and isolation of nonbiological7−9 and biological particles.10−14,17−19 Many reports have investigated the trapping capacity of iDEP systems, but most of these have been qualitative in nature.15,16 Uncertainty in the local electric field gradients and particles dielectric properties and deformability, pose a challenging derivation of an accurate estimation of the DEP force.16 Further, the effect of particle size/shape in the DEP force is often simplified,17 and most approaches use the point-dipole approximation, which is valid only when the particle size is much smaller than the characteristic length scale of the electric field.18 This condition may breakdown when particles approach the channel walls, or when particles are too closely spaced, because particles themselves perturb the electric field distribution in their vicinity. In particular, the gap between neighboring particles has a large effect on the net DEP force.
M
© XXXX American Chemical Society
When two particles are close to each other, the resulting perturbation of the electric field exerts an attractive mutual DEP force on each other (Figure 1a),19 and particles experience an initial acceleration. As the particles get closer to each other, however, a repulsive hydrodynamic pressure force arises between them, and they decelerate (Figure 1b).20 Exploiting these mutual forces can improve the particle enrichment in iDEP devices by causing nearby particles to come together.21 These forces are usually not considered in numerical models because the simulation time scales exponentially with the number of particles.15 Numerous reports have focused on matching numerical simulations to experimental observations, without incurring expensive particle−particle computations. Baylon-Cardiel et al.22 studied the regions where 1-μm polystyrene particles were trapped in an iDEP device with circular insulating structures. Weiss et al.16 determined experimentally the DEP mobility of 1-μm polystyrene particles in a tapered microchannel using streak-based particle velocimetry. The experimental mobility found (in the range of −2 × 10−4 mm4 V−2 s−2) was observed to be 4 orders of magnitude larger than the theoretical DEP mobility. This observation highlighted the necessity of using experimental observations to fine-tune numerical simulations. A common approach among the DEP community is to use heuristic correction factors to scale the DEP force experienced by particles. For instance, Kang et al. used a correction factor to Received: May 25, 2017 Accepted: July 7, 2017 Published: July 7, 2017 A
DOI: 10.1021/acs.analchem.7b02008 Anal. Chem. XXXX, XXX, XXX−XXX
Article
Analytical Chemistry
Figure 1. (a,b) Distribution of the electric field in the vicinity of two particles. Gradients in the electric field are generated because the electric field lines (shown in black) need to travel around the particles. (a) A mutually attractive DEP force initially accelerates the particles in the direction of each other. (b) A repulsive hydrodynamic pressure force arises when these particles are close enough, decelerating them. (c) Schematic of the tapered microfluidic channel. The channel has three sections: an initial wide uniform region, followed by a tapered region and a narrow uniform region. The image shows the region of interest for the streak-based velocimetry technique. (d) Schematic of the iDEP device with arrays of cylindrical posts.
that reduce particle−particle interactions. However, there are applications that can benefit from exploiting particle mutual interactions. For instance, increasing particle enrichment in iDEP devices. An important example is the enrichment of diluted biomarkers from clinical samples. A successful enrichment method should be able to trap and retain the low biomarker particles from a sample, but it should also be gentle enough to avoid denaturation. The later usually implies a reduction in the applied electric potential.29 However, higher electric fields are needed when dealing with dilute cell suspensions.30 In solutions with high particle concentration, similar particles experience attractive dipole−dipole interactions, causing them to cluster and facilitating their DEP manipulation.31 To mimic these conditions and achieve the enrichment of low-abundance target particles, additional particles that behave as innocuous fillers can be added to a solution. These filler particles should increase the particle mutual interactions and aid the enrichment process, but they should not be enriched along with the particles/cells of interest. A previous report of our group32 studied the stability of particle enrichment during the isolation of low-abundance particles. It was found that iDEP systems can effectively and selectively capture rare particles/cells in a sample, despite the presence of highly concentrated nontarget particles of different sizes. Filler particles with different sizes and at different concentrations promote dissimilar mutual interactions and alter the DEP enrichment mechanism to different extents. This contribution
simulate the trajectory of particles around insulator hurdles when DC23 and AC24 electric potentials were applied. The authors noted that any corrections made to account for particle size should be local, and should be a function of both electric field and particle size.23 However, they assumed the correction factor for a particular particle size to be constant, reporting that the dependence on the local electric field and particle size is complicated and unknown.23 Zhu et al. employed correction factors to predict the focusing of particles in curved25 and serpentine26 channels with no insulating structures, as well as on straight channels with a single constriction.27 Similarly, Li et al. used correction factors to match predicted particle trajectories with experimental results in waved38 and curved39 channels. Lewpiriyawong et al.28 demonstrated the enrichment of polystyrene particles and bacterial cells by negative DEP using a tapered channel. In their simulations, they used a constant correction factor for each particle size to account for unobserved/unknown phenomena.28 Most of these studies reveal that the perturbations in the electric field caused by particle−wall interactions are important. Further, when nondiluted solutions are employed, significant perturbations in the electric field and increasing mutual DEP forces may be caused by a large number of particles closely spaced together. Several applications exist in which particle mutual interactions are not desirable and should be kept to a minimum (e.g., single-cell manipulation). For these applications, Aubry and Singh21 presented a comprehensive study on the factors B
DOI: 10.1021/acs.analchem.7b02008 Anal. Chem. XXXX, XXX, XXX−XXX
Article
Analytical Chemistry Table 1. Characteristics of Particle Samples Used in This Studya filler particle concentration (beads/mL) filler particle volume fraction (v/v) 1 5 1 5 1 5 1 5 1
× × × × × × × × ×
10−7 10−7 10−6 10−6 10−5 10−5 10−4 10−4 10−3
500-nm 1.5 7.6 1.5 7.6 1.5 7.6 1.5 7.6 1.5
× × × × × × × × ×
106 106 107 107 108 108 109 109 1010
1-μm (red) 1.9 9.6 1.9 9.6 1.9 9.6 1.9 9.6 1.9
× × × × × × × × ×
105 105 106 106 107 107 108 108 109
interparticle spacing (μm)
2-μm
500-nm
1-μm
2-μm
× × × × × × × × ×
3.8 4.9 8.8 11.2 19.2 24.1 38.1 44.3 54.1
7.6 9.9 17.5 22.3 38.5 48.1 76.2 88.5 108.1
15.2 19.7 35.1 44.7 77.0 96.3 152.3 177.0 216.2
2.4 1.9 2.4 1.9 2.4 1.9 2.4 1.9 2.4
104 105 105 106 106 107 107 108 108
In all samples, green 1-μm were the target particles (7.5 × 105 beads/mL). These target particles were mixed with filler particles of 500-nm, 1-μm and 2-μm particles. Column 1 lists the volume fraction of the filler particles. Columns 2−4 list the concentration of filler particles. Columns 5−7 show the interparticle spacing34 for each solution. a
linearly increasing electric field, producing a constant field gradient that generates the least complex DEP force.16 If the charge density of the channel is considered, particle velocity along the centerline can be redefined as16
aims to identify a particle size and concentration range where polystyrene particles act as innocuous fillers, thereby increasing particle mutual interactions and improving the enrichment efficiency in iDEP devices to isolate low-abundance cells in suspension.
■
vp = (γ 2k 2μdep + γkμek )x + (γ 2kμdep + γkμek )
INTRODUCTION Under the influence of a DC electric field, and in the absence of particle mutual interactions, the net particle velocity of a spherical particle along the longitudinal direction in a microchannel is33 μdep E2 ⎛ δE ⎞⎟ ⎜ sin θ Ecos θ − vp = μek + μdep ⎝ R δs ⎠
Hence, if particle velocity is measured along the centerline of the tapered region, then the DEP contribution can be estimated by comparing the resulting slope when fitting the particle velocity to a linear model (as a function of the position) with the corresponding term of eq 4. In this equation, γ represents the electric field magnitude in region 1, which is a function of the electrical current flowing through the channel (i), the bulk conductivity of the suspending medium (σ) and the crosssectional area of the channel in this region:
(1)
where μek and μdep are the EK and DEP mobilities, respectively, E denotes the magnitude of the electric field, s is the arc length along the field line, θ is the angle between the tangent of the field line and the longitudinal axis, and R is the radius of curvature of the field line. If the particle velocity is analyzed at the centerline of a symmetrical channel, the field lines are parallel to the longitudinal axis (θ = 0) and the particle velocity simplifies to16 ⎛ δE ⎞ vp = ⎜μek + μdep ⎟E ⎝ δx ⎠
(4)
γ=
i σhw1
(5)
Polystyrene particle solutions with volumetric fractions of the filler particles ranging from 1 × 10−7 to 1 × 10−3 v/v were studied (Table 1). No significant changes in pH and conductivity (
