micromachines

Article
Dielectrophoresis Separation of Platelets Using
a Novel Zigzag Microchannel
Yanfang Guan * , Yansheng Liu, Hui Lei, Shihua Liu, Fengqian Xu, Xiangxin Meng,
Mingyang Bai, Xiaoliang Wang and Gexuan Yang
School of Electromechanical Engineering, Henan University of Technology, Zhengzhou 450001, China;
[email protected] (Y.L.); [email protected] (H.L.); [email protected] (S.L.);
[email protected] (F.X.); [email protected] (X.M.); [email protected] (M.B.);
[email protected] (X.W.); [email protected] (G.Y.)
* Correspondence: [email protected]




Received: 2 September 2020; Accepted: 22 September 2020; Published: 25 September 2020


Abstract: Platelet separation and purification are required in many applications including in the
detection and treatment of hemorrhagic and thrombotic diseases, in addition to transfusions and
in medical research. In this study, platelet separation was evaluated using a novel zigzag microchannel
fluidic device while leveraging a dielectrophoresis (DEP) electric field using the COMSOL multiphysics
software package and additional experimentation. The zigzag-shaped microchannel was superior to
straight channel devices for cell separation because the sharp corners reduced the required horizontal
distance needed for separation and also contributed to an asymmetric DEP electric field. A perfect
linear relationship was observed between the separation distance and the corner angles. A quadratic
relationship (R2 = 0.99) was observed between the driving voltage and the width and the lengths of
the channel, allowing for optimization of these properties. In addition, the voltage was inversely
proportional to the channel width and proportional to the channel length. An optimal velocity ratio
of 1:4 was identified for the velocities of the two device inlets. The proposed device was fabricated
using laser engraving and lithography with optimized structures including a 0.5 mm channel width,
a 120◦ corner angle, a 0.3 mm channel depth, and a 17 mm channel length. A separation efficiency of
99.4% was achieved using a voltage of 20 V and a velocity ratio of 1:4. The easy fabrication, lower
required voltage, label-free detection, high efficiency, and environmental friendliness of this device
make it suitable for point-of-care medicine and biological applications. Moreover, it can be used for
the separation of other types of compounds including lipids.

Keywords: zigzag-shape microchannel; Red blood cells (RBCs); platelets; dielectrophoresis

1. Introduction
The rapid separation and purification of biological particles have attracted considerable research
attention recently due to their value in research, especially in medical diagnosis and environmental
detection applications [1–3]. Centrifugation [4,5] is the most frequently used technology for separating
biological particles. However, centrifugation requires professional equipment, laboratory facilities,
and operators, while also being expensive and not suitable for point-of-care testing (POCT) [6,7].
Consequently, methods to separate biological particles have been extensively explored, with microfluidic
technology [8–10] emerging as a promising means for separating particles.
Microfluidics is a simple, low-cost, and easy-to-operate technology compared with traditional
macroscale technology that requires large-scale operation equipment and a suitable laboratory facility.
The principle of separation in microfluidic systems operates via well-designed microchannels and
precisely controlled external physical separation methods that utilize acoustic fields [11,12], fluorescence

Micromachines 2020, 11, 890; doi:10.3390/mi11100890 www.mdpi.com/journal/micromachines

Micromachines 2020, 11, 890 2 of 15

activation [13,14], optical filing [15,16], magnetic activation [17,18], and electrical fields [19–21].
Shamloo et al. [11] simulated the separation of blood cells with acoustic fields, while evaluating the
most efficient and low-cost separation devices by altering factors of acoustic field devices. In addition,
Rosental et al. [14] demonstrated and developed a method to isolate different cell populations in corals
and other cnidarians using fluorescent activated cell sorting (FACS) technology, which is an advanced
cell sorting technique. Further, Tamura et al. [15] used optical cell separation technology to separate
and classify heterogeneous cancerous cells based on morphology, confirming the feasibility of optical
technology in cancer cell manipulation.
Dielectrophoresis (DEP) utilizes different charge characteristics among cells to achieve cell sorting
and differential identification without relying on immunochemistry, making it useful for most cell sorting
applications and rendering it different than other previously developed methods [21]. DEP leverages
the differential dielectric responses in electric fields of particles with different structures and materials
to achieve separation without destroying particle structures. The principle of DEP relies on different
particles being separated by different forces in an inhomogeneous electric field [22]. Positive and
negative electrodes are first applied to the sidewalls of separation devices. Different particles or cells
then move toward different areas based on dielectrophoretic forces resultant from the asymmetric
electric field (i.e., DEP). The unique features of the method make it simple, low-cost, and highly
efficient [23]. Huang et al. [24] used DEP to separate monocytic cells from T cells, achieving separation
efficiencies as high as 95%. In addition, Rahmani et al. [25] enriched biological particles by using
a contactless DEP device, achieving enrichment rates of about 86%. Further, Liao et al. [26] separated
plasma and blood cells from a whole blood sample (1 µL) using DEP, achieving a separation efficiency
of 90%.
The measurement of platelet concentrations in blood is important for detecting many types of
diseases. Low concentrations of platelets can lead to hemorrhage, and high concentrations can lead to
thrombosis, stroke, and other diseases [27]. Thus, platelet separation using DEP devices have been
proposed. For instance, Pommer et al. [28] used a two-stage dielectrophoresis-activated cell sorter
(DACS) chip to separate platelets from blood. The electrodes were arranged on both sides of the device,
and the researchers achieved 95% platelet purity. Piacentini et al. [29] also used rectangular electrodes
arranged on a single side to achieve highly efficient platelet separation (98%) in microfluidic channels
with lower driving voltages.
In this study, platelets and red blood cells (RBCs) were separated by DEP using a novel
zigzag-shaped microchannel. Moreover, a unilateral zigzag-shaped arrangement was proposed
in order to achieve a uniform distribution of potential in microfluidic chip channels, which represents
a novel means to replace traditional DEP devices by arranging electrodes on double-layers with
rectangular or semicircular configurations [28–31], leading to more efficient separation. Compared with
electrode configurations in other channels, zigzag channels require electrode configurations around
the corners of the channel, minimizing the negative effects of electrode slots on particle movement.
Thus, particles can flow more smoothly and efficiently in the new devices, reducing particle loss rates.
Moreover, zigzag channels represent deformed serpentine channels that more efficiently separate
particles due to the influence of inertial forces [32–34]. An interlaced zigzag channel structure was
implemented in a fabricated device and the COMSOL multiphysics simulation software package was
used to arrange the electrodes on one side of the channel to form an uneven electric field distribution
and achieve better particle separation. Lastly, the structure of the zigzag channel was further evaluated
to compare particle separation characteristics for different structures and the effects of various factors
on particle separation in order to achieve optimal cell separation.
Micromachines 2020, 11, 890 3 of 15

2. Materials and Methods

2.1. Dielectrophoresis Theory

In this study, platelets and RBCs were regarded as spherical particles since the concentrations of
white blood cells were very low and could be considered negligible. The time-averaged dielectrophoretic
force (Fdep ) of spherical particles in an electric field is derived from the following equations [35]:

Fdep = πr3 ε1 Re[K(w)]∇|E|2 (1)

where r is the radius of the particle, ε1 is the permittivity of the medium, and ∇|E|2 is the gradient of
the electric field intensity squared. K is the Clausius-grope factor, namely CM factor. K is calculated
from Equation (2) in the AC electric field:

εe2 − εe1
K= (2)
εe2 + 2εe1

where εe2 and εe1 are the complex permittivity of the particle and medium with the AC electric field,
which may be calculated with the follow Equation (3):

εi − jσ1
εei = (3)
w
where σ is the conductivity, ω is the angular frequency, and i = 1, 2. When the particle is more prone to
polarization than the medium (Re (K) > 0), the particle will generate a positive dielectric force (pDEP).
If the medium is more susceptible to polarization than the particles (Re (K) < 0), a negative dielectric
force (nDEP) will be produced.

2.2. Initial and Boundary Conditions

The zigzag-shaped microchannel used for platelet separation is shown in Figure 1a. A no-slip
boundary condition was imposed between the fluids and channel walls. Fully developed flow
conditions were applied at the D1 and E1 outlets using the COMSOL software package. In addition,
alternating-current electric fields (1 × 10−3 Hz of driving frequency), creeping flow, and particle tracing
for fluid flow models were used to promote separation. Released particle density was set as 1050 kg/m3 .
Inlet D was filled with RBCs and platelet microparticles. Inlet E was filled with phosphate buffer saline
diluted in a sucrose solution to achieve a conductivity of 55 mS/m. Inlets D and E were set with different
flow velocity ratios (1:4, 1:3, 1:2, and 1:1). The potential, as indicated in blue in Figure 1a, was loaded
at the corner of the zigzag microchannel, where 1, 3, 5, 7, and 9 indicate the positive electrodes (+V)
and 2, 4, 6, and 8 indicate the negative electrodes (−V). A rebound condition was used between the
particles and the wall. To complete the simulation, the velocity field and DEP modules were added to
the finite element analysis (FEA) module. The simulation mesh is shown in Figure 1b. The calculated
electric field distribution in the zigzag microchannel was non-uniform, as shown in Figure 1c, making
it suitable for the separation of particles with different sizes by DEP.

2.3. Fabrication of the Microfluidic Chip

The fabricated microchip included two layers (15 mm × 30 mm × 1 mm), as shown in Figure 2.
The upper layer was equipped with two inlet/outlet pipes with 2 mm outside diameters and was
fabricated using polydimethylsiloxane (PDMS) using reverse molding technology. The lower layer
was designed with a zigzag-shaped microchannel (corner angle = 120 gz, channel width = 0.5 mm,
channel depth = 0.3 mm, and channel length = 17 mm) and inlet/outlet holes using the CAD software
package. The microchannel and holes were subsequently carved to a depth of 0.3 mm using a laser
engraving machine. The silver electrode pattern (Ag, 300 nm) was deposited on the lower glass wafer
using traditional lithography. Wires were arranged at the sidewall positions. Finally, the two layers
Micromachines 2020, 11, 890 4 of 15

were irreversibly bonded after placing the upper layer of PDMS under a mercury lamp for 30 min,
as shown in Figure 2020,
Micromachines 2b. 11, x 4 of 17

Figure 1. Initial and boundary condition settings for the zig-zag microchannel. (a) Inlet, outlet, and
position of the potential in the microchannel. (b) The 2D finite element analysis (FEA) mesh. (c)
Distribution of the non-uniform electric field by dielectrophoresis.

2.3. Fabrication of the Microfluidic Chip

The fabricated microchip included two layers (15 mm × 30 mm × 1 mm), as shown in Figure 2.
The upper layer was equipped with two inlet/outlet pipes with 2 mm outside diameters and was
fabricated using polydimethylsiloxane (PDMS) using reverse molding technology. The lower layer
was designed with a zigzag-shaped microchannel (corner angle = 120 gz, channel width = 0.5 mm,
channel depth = 0.3 mm, and channel length = 17 mm) and inlet/outlet holes using the CAD software
package. The microchannel and holes were subsequently carved to a depth of 0.3 mm using a laser
engravingFigure
machine. The silver electrode pattern (Ag, 300 nm) was deposited on the lower glass wafer
1. Initial and boundary condition settings for the zig-zag microchannel. (a) Inlet, outlet, and
Figure 1. Initial
using traditional and boundary condition
lithography. settings at for thesidewall
zig-zagpositions.
microchannel. (a) Inlet, outlet,
position of the potential Wires were arranged
in the microchannel. (b) The the
2D finite element Finally,
analysis (FEA) the
mesh. (c)two layers
and position of
were irreversibly the potential
bonded
Distribution in the
after placing
of the non-uniform microchannel.
thefield
electric upper layer of PDMS under a mercury lamp for 30mesh.
(b) The 2D
by dielectrophoresis. finite element analysis (FEA) min,
(c) Distribution of
as shown in Figure 2b.the non-uniform electric field by dielectrophoresis.
2.3. Fabrication of the Microfluidic Chip
The fabricated microchip included two layers (15 mm × 30 mm × 1 mm), as shown in Figure 2.
The upper layer was equipped with two inlet/outlet pipes with 2 mm outside diameters and was
fabricated using polydimethylsiloxane (PDMS) using reverse molding technology. The lower layer
was designed with a zigzag-shaped microchannel (corner angle = 120 gz, channel width = 0.5 mm,
channel depth = 0.3 mm, and channel length = 17 mm) and inlet/outlet holes using the CAD software
package. The microchannel and holes were subsequently carved to a depth of 0.3 mm using a laser
engraving machine. The silver electrode pattern (Ag, 300 nm) was deposited on the lower glass wafer
using traditional lithography. Wires were arranged at the sidewall positions. Finally, the two layers
were irreversibly bonded after placing the upper layer of PDMS under a mercury lamp for 30 min,
as shown in Figure 2b.

Figure 2. Diagram of the dielectrophoresis-based microfluidic chip for red blood cell (RBC) and
Figure 2. Diagram of the dielectrophoresis-based microfluidic chip for red blood cell (RBC) and platelet
platelet separation. (a) Structural parameters of the microfluidic chip and (b) image of the microfluidic
separation. (a) Structural parameters of the microfluidic chip and (b) image of the microfluidic chip
chip after sealing.
after sealing.
2.4. Materials and Experimental Setup
2.4. Materials and Experimental Setup
The experiments utilized a laser engraving machine (MUV-E-A, Yueming Laser, Guangdong,
The experiments
China), an electron utilized a laser engraving
beam evaporator machineULVAC,
(E-beam VT1-10CE, (MUV-E-A, Yueming
Shanghai, China),Laser,
an ABM Guangdong,
semi-
China), an electron beam evaporator (E-beam VT1-10CE, ULVAC, Shanghai, China), an ABM semi-auto
mask aligner and
FigureUV exposure
2. Diagram of thesystem (ABM/6/350/NUV/DCCD/BSV/SA,
dielectrophoresis-based microfluidic chip for red blood ABM, Chicago,
cell (RBC) and IL, USA),
a syringe pump (LSP01,
platelet Baoding,
separation. China),
(a) Structural a signal
parameters generator
of the microfluidic chip (DG1022,
and (b) image Rigol, Zhengzhou, China),
of the microfluidic
chip after sealing.
a Charge-coupled Device (CCD)and microscope (Obvious Ltd., Co., Guangzhou, China), PDMS (Sylgard
184, Dow 2.4.
Corning,
Materials Midland, MI,Setup
and Experimental USA), and a flow cytometer (FACS CALIBAR, BD Biosciences,
Shanghai, China).
The experiments utilized a laser engraving machine (MUV-E-A, Yueming Laser, Guangdong,
BloodChina),
was collected
an electronfrom
beamhealthy
evaporator volunteers and a sodium
(E-beam VT1-10CE, ULVAC,citrate
Shanghai,solution
China), anwasABMadded
semi-to prevent
blood clotting. The blood was subsequently placed in a centrifuge tube and centrifuged for 10 min
at 1000 rpm. The upper supernatant in the tube (containing mostly platelets) and the lower portion
(primarily containing RBCs) were mixed and diluted at a 1:10 ratio with normal saline to obtain the
experimental samples. Flow cytometry was used for cell selection, with 10,000 RBCs and 10,000 platelets
used for the experimental samples. Phosphate buffer saline (PBS) was used as the working solution to
adjust the conductivity of the samples to 55 mS/m in order to maintain the osmolarity. All experiments
 Charge-coupled Device (CCD)and microscope (Obvious Ltd., Co., Guangzhou, China), PDMS
(Sylgard 184, Dow Corning, Midland, MI, USA), and a flow cytometer (FACS CALIBAR, BD
Biosciences, Shanghai, China).
Blood was collected from healthy volunteers and a sodium citrate solution was added to prevent
blood clotting. The blood was subsequently placed in a centrifuge tube and centrifuged for 10 min at
1000 rpm. The upper supernatant in the tube (containing mostly platelets) and the lower portion
Micromachines 2020, 11, containing
(primarily 890 RBCs) were mixed and diluted at a 1:10 ratio with normal saline to obtain the 5 of 15
experimental samples. Flow cytometry was used for cell selection, with 10,000 RBCs and 10,000
platelets used for the experimental samples. Phosphate buffer saline (PBS) was used as the working
were approved
solutionbyto the ethical
adjust committee
the conductivity at samples
of the the Henan University
to 55 mS/m in orderof
to Technology, and were
maintain the osmolarity. Allperformed
experiments
in compliance with the were approved
ethical by the
policy forethical
use ofcommittee
human at the Henan
subjects University
according toofthe
Technology,
national and
guidelines of
were performed in compliance with the ethical policy for use of human subjects according to the
China. Informed consent was obtained for all experimentation with human subjects.
national guidelines of China. Informed consent was obtained for all experimentation with human
The experimental
subjects. setup was divided into three zones for liquid supply, separation, and recovery
(Figure 3). The Thesample andsetup
experimental buffer
wassolution
divided into(PBS) wereforsupplied
three zones by separation,
liquid supply, an injection pump at speeds
and recovery
of 200 and 800 µm/s, respectively, from inlets D and E, respectively. The fluid fieldofwas
(Figure 3). The sample and buffer solution (PBS) were supplied by an injection pump at speeds 200 observed
and 800 μm/s, respectively, from inlets D and E, respectively. The fluid field was observed
microscopically and with the CCD camera. Samples in the recovery zone were evaluated with flow
microscopically and with the CCD camera. Samples in the recovery zone were evaluated with flow
cytometrycytometry
to determine separation
to determine efficiency.
separation efficiency.

Figure 3. Experimental platform for platelet separation.

Figure 3. Experimental platform for platelet separation.
3. Results and Discussion
3. Results and Discussion
3.1. Structural Parameters and Analysis of Influential Parameters
3.1. Structural Parameters and Analysis of Influential Parameters
DEP occurs when dielectric particles are subjected to an uneven electric field and is widely used
DEP in biomedical
occurs when equipment
dielectric including biosensors,
particles diagnostictomachines,
are subjected an unevenparticle manipulation
electric and is widely
field and
filtration (sorting) devices, and particle assembly. Here, we evaluated the use of a zigzag flow channel
used in biomedical equipment including biosensors, diagnostic machines, particle manipulation and
filtration (sorting) devices, and particle assembly. Here, we evaluated the use of a zigzag flow channel
for DEP applications. Several putative influential factors are involved in DEP separation and were
evaluated individually.

3.1.1. Influence of Corner Angles

To evaluate the influence of corner angles on platelet sorting, other structural parameters of the
zigzag microchannel were fixed (Figure 4a). The channel width (A) was set to 0.5 mm, the corner
length (C) to 2 mm, and corner angles of 60◦ , 90◦ , 120◦ , and 150◦ were evaluated (B). The velocities at
inlets D and E were set at 200 and 800 µm/s, respectively, and the driving voltage and frequency as
20 V and 1 × 10−3 Hz.
The particles were separated in the zigzag channel (Figure 4b), although this was not possible
when the corner angle was 60◦ (Figure 4c). Further, some of the RBCs were lost with the platelets
because of small angles, indicating that particles are more likely to mix rather than separate in passages
with too small of angles [36]. In contrast, the separation of platelets and blood cells was achieved
when the corner angle was increased to 90◦ (Figure 4c). Moreover, larger corner angles produced
better separations based on the separation distance, wherein the distance between the platelets and
RBCs was equal to d, the distance between the RBCs and the voltage-applied wall side was equal to
Micromachines 2020, 11, 890 6 of 15

d1 , and d = d1 − d2 in Figure 4b. Larger angles changed the distribution of the electric field, and the
different DEP forces induced by the non-uniform electric field then acted on the platelets and RBCs,
leading to total separation (Figure 4c–e). Indeed, a perfect linear relationship (R2 = 1) was observed
between the corner 2020,
Micromachines angle
11, xand the separation distance d, as shown in Figure 4f. 7 of 17

Figure 4. Structural parameters and platelet separation analysis. (a) The structural parameters of the
Figure 4. Structural parameters and platelet separation analysis. (a) The structural parameters of the
zigzag-shaped microchannel. (b) Cell separation resulting from the zigzag channel. (c–f) Analysis of
zigzag-shaped
the separation distance at(b)
microchannel. Cellangles
corner separation resulting
of 60°, 90°, from
120°, and 150°,the zigzag channel.
respectively, blue color(c–f) Analysis of
arrows
the separation
meandistance at corner
the flow trajectory angles
of the of 60
platelet ◦ ,red
and 90◦color ◦ , andmean
, 120arrows 150◦the
, respectively, blue color arrows mean
RBC flow trajectory.
the flow trajectory of the platelet and red color arrows mean the RBC flow trajectory.
Larger corner angles did not, however, guarantee better separation (Figure 5a,b) and minimizing
space is one of the main advantages of the zigzag microchannel. For example, total platelet separation
Larger corner angles did not, however, guarantee better separation (Figure 5a,b) and minimizing
was not achieved using a straight microchannel with an L = 12.314 mm (Figure 5c). However,
space is one of the main
separation advantages
was achieved of corner
when the the zigzag microchannel.
angle (B) was 90° with an For
LH = example, total platelet
12.314 mm (Figure 5a). Noteseparation
was not achieved using ainstraight
that the arrows Figure 5bmicrochannel with
point to Figure 5c, an L = 12.314
indicating that the mm
same(Figure
horizontal5c). However,
distance with aseparation
was achieved when the corner angle (B) was 90 with an LH = 12.314 mm (Figure 5a).
90° corner angle was equivalent to the straight channel
◦ in Figure 5b in order to compare the two Note that
scenarios. Importantly, the zigzag microchannel conserved space compared with the straight
the arrowsmicrochannel,
in Figure 5b point to Figure 5c, indicating that the same horizontal distance with a 90◦
while achieving complete separation, with the latter requiring a longer distance to
corner angle was equivalent
separate to the straight
the platelets (Figure channel
5a). Moreover, in Figure
an electric field 5b
caninbeorder
easily to compare
applied to thethe two scenarios.
zigzag-
Importantly, the zigzag microchannel conserved space compared with the straight microchannel,
shaped microchannel, and cells do not become trapped in the channel due to construction
considerations.
while achieving complete Rather, the cellswith
separation, movetheuplatter
and requiring
down, accelerating
a longercell separation,
distance shortening
to separate the platelets
separation time, and enhancing separation efficiency (discussed below).
(Figure 5a). Moreover, an electric field can be easily applied to the zigzag-shaped microchannel,
and cells do not become trapped in the channel due to construction considerations. Rather, the cells
move up and down, accelerating cell separation, shortening separation time, and enhancing separation
efficiency (discussed below).
Micromachines 2020, 11, 890 7 of 15

Micromachines 2020, 11, x 8 of 17

Figure 5. Comparison between straight and zigzag-shaped microchannel. (a) Linear relationship
Comparison
Figure 5. between between
separation straight
distance and corner and zigzag-shaped
angles. (b) Comparison ofmicrochannel.
the horizontal and(a) Linear
spread relationship
lengths
between separation
for differentdistance and (c)
corner angles. corner angles.
Diagram (b) the
showing Comparison ofofthe
inseparability horizontal
platelets and in
from RBCs spread
the lengths
for differentstraight
cornermicrochannel
angles. (c)for an LH = showing
Diagram 12.314 mmthe (compared against B
inseparability of=platelets
90°). All experiments
from RBCs were
in the straight
replicated with n = 5 and values are shown as mean ± SD.
microchannel for an LH = 12.314 mm (compared against B = 90◦ ). All experiments were replicated with
n = 5 and
3.1.2.values areofshown
Influence Channel asWidth
mean and± SD.
Length
The
3.1.2. Influence ofcorner angleWidth
Channel affectedand
the distribution
Length of the electric field, while channel width and length
affected the separation distance of platelets and RBCs. To compare the effects of the width and length
The corner
relative angle
to the affected thecorner
voltage, the distribution
angle andofinlet
the velocity
electricratios
field,were
while channel
fixed at 120°width and length
and 1:4,
respectively. Linear, quadratic, exponential, logarithmic, and power function fitting
affected the separation distance of platelets and RBCs. To compare the effects of the width and length of the driving
voltage were calculated separately as a function of the channel width and length (Figure 6). The
relative to driving
the voltage, the corner angle and inlet velocity ratios were fixed at 120◦ and 1:4, respectively.
voltage was considered as the minimum separation voltage, wherein the separation of the
Linear, quadratic, exponential,
platelets and logarithmic,
RBCs was obtained and power
once the voltage function
was larger fitting
than this value of theondriving
based voltage were
simulations.
calculated Comparison
separately of ascorrelation
a functioncoefficient
of the channel width
(R2) values and length
for various models(Figure
indicated6).that
Thethedriving voltage was
quadratic
consideredfunctions best expressed
as the minimum the relationships
separation between
voltage, width,the
wherein length, and voltage,
separation as exemplified
of the by R2RBCs was
platelets and
values that were nearly all > 0.99 (Figure 6b,d). Furthermore, the length of the channel was inversely
obtained once the voltage was larger than this value based on simulations. Comparison of correlation
proportional to the voltage (Figure 6a,c) and the width was proportional to the voltage. Specifically,
coefficienta(R 2 ) values for various models indicated that the quadratic functions best expressed the
larger width and smaller length corresponded to a greater driving voltage. Thus, when fabricating
relationships between
microfluidic width,
chips, length, should
the channels and voltage, as exemplified
be as narrow as possible by R2 satisfying
while values that were nearly
process
all > 0.99 (Figure 6b,d). Furthermore, the length of the channel was inversely proportional to the
requirements, which ultimately saves space and reduces energy losses while also achieving better
separation.
voltage (Figure 6a,c) and the width was proportional to the voltage. Specifically, a larger width and
smaller length corresponded to a greater driving voltage. Thus, when fabricating microfluidic chips,
the channels should be as narrow as possible while satisfying process requirements, which ultimately
saves space and reduces energy losses while also achieving better separation.

3.1.3. Influence of the Velocity Ratio

To examine the influence of inlet velocity on particle flow within the channel, velocity ratios of
1:1, 1:2, 1:3, and 1:4 were used between inlets D and E, and the simulated separation effects were
analyzed (Figure 7a–d). In the simulations, the structural parameters were fixed to ensure that the
particles were separated with a corner angle = 120◦ , length = 8C, width = 0.5 mm, and driving voltage
= 20 V. The velocity at inlet D was set as 200 µm/s and the velocity at inlet E was varied to achieve
the desired ratio. When the velocity ratio was 1:1, platelets were not fully separated (Figure 7a).
When the ratio was increased to 1:2, the platelets and RBCs separately flowed out from the two outlets.
Moreover, the separation distance d2 decreased as the velocity ratio increased (Figure 7e), as modeled
by a quadratic function that captured distance variation. Thus, the velocity ratio significantly affected
the distance d1, but had a small influence on the distance d (Figure 7f). Consequently, to achieve
Micromachines 2020, 11, 890 8 of 15

complete separation of platelets from blood samples, the buffer velocity should be higher than that of
the blood sample when using the same driving voltage in order to ensure that the sample is closer to
the side of theMicromachines
electrode. If the above is not achieved, particle separation will be affected.
2020, 11, x 9 of 17

Figure 6. Influence of channel length and width on separation voltage. (a) The relationship between
Figure 6. Influence
channelof channel
length length
and driving and(b)
voltage. width on separation
Correlation coefficients (Rvoltage.
2 values) of(a) The
fitted relationship
functions for between
channel lengthdifferent
and driving voltage.
channel lengths. (b) Correlation
(c) Proportional coefficients
relationships (R2 values)
between channel widths and of driving
fitted functions for
voltages. (d) Correlation coefficients (R2 values) for different fitted functions for different channel
different channel lengths. (c) Proportional relationships between channel widths and driving voltages.
widths. All experiments were repeated with n = 5, and values are shown as mean ± SD.
(d) Correlation coefficients (R2 values) for different fitted functions for different channel widths.
3.1.3. Influence
All experiments were2020,
Micromachines
of the Velocity
repeated
11, x n = 5, and values are shown as mean ± SD.
withRatio 11 of 18
To examine the influence of inlet velocity on particle flow within the channel, velocity ratios of
1:1, 1:2, 1:3, and 1:4 were used between inlets D and E, and the simulated separation effects were
analyzed (Figure 7a–d). In the simulations, the structural parameters were fixed to ensure that the
particles were separated with a corner angle = 120°, length = 8C, width = 0.5 mm, and driving voltage
= 20 V. The velocity at inlet D was set as 200 μm/s and the velocity at inlet E was varied to achieve
the desired ratio. When the velocity ratio was 1:1, platelets were not fully separated (Figure 7a). When
the ratio was increased to 1:2, the platelets and RBCs separately flowed out from the two outlets.
Moreover, the separation distance d2 decreased as the velocity ratio increased (Figure 7e), as modeled
by a quadratic function that captured distance variation. Thus, the velocity ratio significantly affected
the distance d1, but had a small influence on the distance d (Figure 7f). Consequently, to achieve
complete separation of platelets from blood samples, the buffer velocity should be higher than that
of the blood sample when using the same driving voltage in order to ensure that the sample is closer
to the side of the electrode. If the above is not achieved, particle separation will be affected.

(revised)Figure 7. Separation conditions in response to the velocity ratio. Separation results are
Figure 7. Separation conditions
shown when in response
using velocity to (b)
ratios of (a) 1:1, the1:2,velocity
(c) 1:3, andratio.
(d) 1:4,Separation
blue color arrows results are shown when
mean the
using velocity ratios of (a)of1:1,
flow trajectory (b) 1:2,
the platelet and(c)
red1:3,
colorand
arrows(d) 1:4,theblue
mean color
RBC flow arrows
trajectory. mean
(e) The the flow trajectory of
relationship
between separation distances d1 and d2 and different velocity ratios. (f) Comparison of platelet and
the platelet and red color arrows mean the RBC flow trajectory. (e) The relationship between separation
RBC separation distances. All experiments were conducted with n = 5 and values are shown as mean
distances d1 and± SD.d2 and different velocity ratios. (f) Comparison of platelet and RBC separation
distances. All experiments were conducted with n = 5 and values are shown as mean ± SD.
3.1.4. Influence of Driving Voltage
As shown previously, the corner angle, channel width, and length of separation devices can
directly affect the distribution of electric fields, thereby affecting the separation of particles with
different DEP forces. Consequently, the driving voltage is critical for platelet or other types of
separation, as shown by Equation (1). The platelets were separated normally when the driving
voltage ranged between 11 and 22 V (Figure 8a,b) and the structural parameter values of the zigzag
microchannel were the same as described in Section 3.1.3. However, when the voltage was higher
than 22 V or lower than 11 V, platelets were not separated from the blood samples and the platelets
Micromachines 2020, 11, 890 9 of 15

3.1.4. Influence of Driving Voltage

As shown previously, the corner angle, channel width, and length of separation devices can directly
affect the distribution of electric fields, thereby affecting the separation of particles with different DEP
forces. Consequently, the driving voltage is critical for platelet or other types of separation, as shown
by Equation (1). The platelets were separated normally when the driving voltage ranged between
11 and 22 V (Figure 8a,b) and the structural parameter values of the zigzag microchannel were the
same as described in Section 3.1.3. However, when the voltage was higher than 22 V or lower than
11 V, platelets were not separated from the blood samples and the platelets flowed out from the upper
outlet (Figure 8c). Further, there was no RBC flow out of both outlets when using a 23 V driving
voltage, wherein the RBC remained on the lower wall when moving along the zigzag channel to the
second corner angle. These activities can be explained by the relationship between DEP forces and
the electric field, as given by Equation (1). Thus, the distances d1 and d2 increased with the driving
voltage (Figure 8e). The higher the voltage, the greater the separation effect on separation distance,
wherein a voltage of 22 V was the critical threshold value for platelet separation from blood samples.
Overall, these results indicate that voltage had the most significant effect on particle separation in the
dielectric field. Consequently, the critical separation point of particles can be achieved by adjusting the
voltage range
Micromachines for11,different
2020, x microchannel structures. 11 of 17

Figure 8. Effects of the driving voltage on platelet separation. (a–d) Separation under different driving
Figure 8. Effects of the driving voltage on platelet separation. (a–d) Separation under different driving
voltages
voltages of
of 11,
11, 22,
22, and
and 23
23 V,
V, Blue
Blue color
color arrows
arrows mean
mean the
the flow
flow trajectory
trajectory of
of the
the platelet
platelet and
and red
red color
color
arrows
arrows mean the RBC flow trajectory. (e) Comparison of platelet and RBC separation distances with
mean the RBC flow trajectory. (e) Comparison of platelet and RBC separation distances with
different
different driving
drivingvoltages.
voltages.(f)(f)
The relationship
The relationship between
betweenseparation distance
separation andand
distance driving voltages.
driving All
voltages.
experiments were conducted with an n = 5, and values are shown as mean
All experiments were conducted with an n = 5, and values are shown as mean ± SD. ± SD.

3.1.5. Influence of Driving Frequency

Figure 9 shows the separation performance with respect to the driving frequency at a constant
driving voltage of 12 V and a 1:4 velocity ratio between RBCs and platelets. The results indicated that
the separation performance was inversely proportional to the driving frequency (Figure 9f). That is,
lower driving frequency like 1 × 10−3 Hz led to a better separation effect as indicated by a larger
separation distance and the effect diminished as frequency increased until a failure was obtained at
1000 kHz.
Micromachines 2020, 11, 890 10 of 15

3.1.5. Influence of Driving Frequency

Figure 9 shows the separation performance with respect to the driving frequency at a constant
driving voltage of 12 V and a 1:4 velocity ratio between RBCs and platelets. The results indicated that
the separation performance was inversely proportional to the driving frequency (Figure 9f). That is,
lower driving frequency like 1 × 10−3 Hz led to a better separation effect as indicated by a larger
separation distance and the effect diminished as frequency increased until a failure was obtained at
1000 kHz. 2020, 11, x
Micromachines 12 of 17

Figure 9.
Figure 9. Effects of the
Effects of the driving
driving frequency
frequency onon platelet
platelet separation.
separation. (a–d)
(a–d) Separation
Separation under
under different
different
driving frequencies of 1 × 10−3 Hz,
−3 1 Hz, 1 × 10 +3 Hz,
3 and 1 × 10 +6 Hz,
6 blue
driving frequencies of 1 × 10 Hz, 1 Hz, 1 × 10 Hz, and 1 × 10 Hz, blue color arrows mean color arrows mean the flow
the
trajectory of the platelet and red color arrows mean the RBC flow trajectory. (e) Comparison
flow trajectory of the platelet and red color arrows mean the RBC flow trajectory. (e) Comparison of platelet
of
and RBC
platelet andseparation distances
RBC separation with with
distances different driving
different frequencies.
driving frequencies. (f)(f)The
Therelationship
relationship between
between
separation distance
separation distanceand
anddriving
driving frequencies.
frequencies. AllAll experiments
experiments werewere conducted
conducted with
with an n =an n = 5,
5, and and
values
values
are are shown
shown as meanas±mean
SD. ± SD.

3.1.6. Influence of Cell Size

When the RBCs and platelets flowed in the microchannel,
microchannel, cell deformation could occur due to
the electric field and capillary forces [37]. To account for this, possible cell sizes with minimum and
maximum
maximum values
valuesofof4 and 15 µm
4 and for RBC
15 μm diameters,
for RBC respectively,
diameters, and 1 and
respectively, and5 µm for 5platelet
1 and μm fordiameters,
platelet
respectively, were used towere
diameters, respectively, model separation
used to modelwhile keeping
separation the other
while keepingparameters
the otherthe same as described
parameters the same
in
as Section 3.1.4.
described Platelet3.1.4.
in Section separation
Plateletwith variouswith
separation cell sizes
variouscould
cellbe achieved
sizes could bebyachieved
adjustingby the driving
adjusting
voltage (Figure 10). Thus, if the cell diameters are large, a lower voltage should be used.
the driving voltage (Figure 10). Thus, if the cell diameters are large, a lower voltage should be used. In contrast,
aInhigher voltage
contrast, should
a higher be used
voltage for the
should be separation
used for theofseparation
small-sizedofcells. For example,
small-sized the example,
cells. For cells become
the
cells become larger and smaller in the case of the two cell sizes, with their respective voltage ranges
being 4–8 V and 13–24 V, respectively (Figure 10a,b). The large cells were more easily separated with
low voltage and low voltage regulation. When the deformation of particles was small, the voltage
needed to be adjusted within the normal range (Figure 10a). Larger particle deformations required
smaller voltage values (Figure 10b). Thus, these results also show that relatively large particles were
Micromachines 2020, 11, 890 11 of 15

larger and smaller in the case of the two cell sizes, with their respective voltage ranges being 4–8 V
and 13–24 V, respectively (Figure 10a,b). The large cells were more easily separated with low voltage
and low voltage regulation. When the deformation of particles was small, the voltage needed to be
adjusted within the normal range (Figure 10a). Larger particle deformations required smaller voltage
values (Figure 10b). Thus, these results also show that relatively large particles were more suitable for
DEP separation
Micromachines when
2020, 11, x the channel structure was consistent and energy loss was reduced. 13 of 17

Figure 10. Separation conditions when using different cell sizes. (a) Separation of RBCs and platelets
Figure 10. Separation conditions when using different cell sizes. (a) Separation of RBCs and platelets
with maximum sizes of 15 and 5 μm, respectively. (b) Separation of RBCs and platelets with minimum
with maximum sizes of 15 and 5 µm, respectively. (b) Separation of RBCs and platelets with minimum
sizes of 4 and 1 μm, respectively. All experiments were conducted with an n = 5, and values are shown
sizes of 4 and 1 µm, respectively. All experiments were conducted with an n = 5, and values are shown
as mean ± SD.
as mean ± SD.

3.2.3.2. ExperimentalAnalysis
Experimental AnalysisofofPlatelet
PlateletSeparation
Separation Efficiency
Efficiency

ToTo further
further verify
verify thethe
useuse of zigzag
of zigzag microfluidic
microfluidic chips
chips onon
cellcell separation,
separation, a practical
a practical experiment
experiment was
was conducted. Samples containing platelets, RBCs, and PBS solutions were injected into a
conducted. Samples containing platelets, RBCs, and PBS solutions were injected into a microchannel
microchannel from inlets D and E at a velocity ratio of 1:4 (200 and 800 μm/s, respectively) using a
from inlets D and E at a velocity ratio of 1:4 (200 and 800 µm/s, respectively) using a syringe pump.
syringe pump. To evaluate the effect of different driving voltages on separation, several voltages (0,
To evaluate the effect of different driving voltages on separation, several voltages (0, 10, 15, and 20 V)
10, 15, and 20 V) and velocity ratios (1:1, 1:2, 1:3, and 1:4) were used (Figure 11a–g). No separation of
and velocity ratios (1:1, 1:2, 1:3, and 1:4) were used (Figure 11a–g). No separation of platelets and RBCs
platelets and RBCs occurred at a voltage of 0 V, because a DEP force was not present in the
occurred at a voltage
microchannel (Figureof11a).
0 V,As because
voltageaincreased
DEP force to was not
10 and 15present in and
V, platelet the microchannel
RBC separation(Figure
occurred,11a).
Asandvoltage increased to 10 and 15 V, platelet and RBC separation occurred, and
many RBCs flowed out far from the electrode and the exit E1 (Figure 10b,c). Increased voltage to many RBCs flowed
out20farV from the in
resulted electrode and the separation
perfect platelet exit E1 (Figure 10b,c).
(Figure 11d),Increased
with almost voltage to 20observed
no RBCs V resulted in1.perfect
in D The
platelet separation (Figure 11d), with almost no RBCs observed
same result occurred with velocity ratios of 1:1, 1:2, 1:3, and 1:4. Further, in D 1 . The same result occurred
when the inlet velocity ratio with
velocity ratios of 1:1,
was increased from1:2,1:11:3,
toand
1:4,1:4. Further,
better when
platelet the inletoccurred
separation velocity ratio was11e–g).
(Figure increased
Thefrom 1:1 to
platelet
1:4,separation
better platelet separation
efficiency occurred from
was calculated (Figure 11e–g).the
counting The plateletofseparation
number cells in theefficiency
samples andwasfrom
calculated
the
from counting
outlet, the number of cells in the samples and from the outlet, as follows:
as follows:
Separation efficiency = (Cell numbers in outlet D1/Cell numbers in sample) × 100% (4)
Separation efficiency = (Cell numbers in outlet D1/Cell numbers in sample) × 100% (4)
A maximum separation efficiency of 99.4% was achieved using a driving voltage of 20 V and a
A maximum
velocity separation
ratio of 1:4 efficiency
(Figure 11h). Thus, of 99.4% was
complete achieved
separation using a and
of platelets driving
RBCs voltage of 20 V A
was achieved. and
a velocity ratio of
small number of 1:4
false(Figure 11h).may
detections Thus, complete
have occurredseparation
due to cellof platelets and
deformation, cellRBCs
death,was achieved.
or parts of
cells sticking
A small numbertoofthe walls
false during may
detections separation. These results
have occurred due toagreed with the simulation
cell deformation, cell death,analysis
or parts
described in Sections 3.1.3 and 3.1.4. Higher driving voltages and velocity ratios
of cells sticking to the walls during separation. These results agreed with the simulation analysisresulted in better
separation
described in efficiency. Thus,
Sections 3.1.3 the3.1.4.
and zigzag-shaped microchannel
Higher driving achieved
voltages better separation
and velocity effectsinif better
ratios resulted the
driving conditions were adjusted to optimal values. No electrode slot was needed to
separation efficiency. Thus, the zigzag-shaped microchannel achieved better separation effects if theachieve the DEP
force conditions
driving because thewere special zigzag-shaped
adjusted structure
to optimal values.was
No complete
electrodeand
slot smooth,
was neededpreventing a possible
to achieve the DEP
loss of particles trapped in the slot. Smaller channel widths resulted in smaller required
force because the special zigzag-shaped structure was complete and smooth, preventing a possible voltages andloss
better separation effects.
of particles trapped in the slot. Smaller channel widths resulted in smaller required voltages and better
separation effects.
Micromachines 2020, 11, 890 12 of 15

Micromachines 2020, 11, x 14 of 17

Figure 11. Experimental platelet separation using different driving voltages and velocity ratios based
Figure 11. Experimental platelet separation using different driving voltages and velocity ratios based
on dielectrophoresis (DEP)
on dielectrophoresis (DEP) forces
forces in
in zigzag-shaped
zigzag-shaped microchannels.
microchannels. (a–d)
(a–d) Platelet separation under
Platelet separation under
voltages of
voltages of 00V,
V,10
10V,
V,15
15V,V,and
and2020V,V,respectively.
respectively.(e–g) Separation
(e–g) Separationeffects at velocity
effects ratios
at velocity of 1:1,
ratios of 1:2,
1:1,
and 1:3. (h) Comparison of separation efficiency using different voltages and velocity
1:2, and 1:3. (h) Comparison of separation efficiency using different voltages and velocity ratios. ratios. All
experiments were conducted with an n = 5, and values are shown as mean
All experiments were conducted with an n = 5, and values are shown as mean ± SD. ± SD.

In addition to
In addition being used
to being used toto achieve
achieve high-efficiency
high-efficiency platelet
platelet separation,
separation, the
the new
new zigzag-shaped
zigzag-shaped
DEP
DEP device
device can
canalso
alsobe
beused
usedtotoseparate other
separate types
other of cells
types andand
of cells non-cellular particles.
non-cellular Moreover,
particles. the
Moreover,
device could be used for particle purification via altering the structural parameters to proper values.
the device could be used for particle purification via altering the structural parameters to proper values.
Furthermore,
Furthermore, thethe lower
lower required driving voltage,
required driving voltage, lack
lack of
of pollution
pollution output,
output, and
and simplicity
simplicity ofof operation
operation
will render
will render this
this new
new device
device popular
popular for
for point-of-care
point-of-care medicine
medicine and
and biology.
biology.

4. Conclusions
In this study, we proposed a new zigzag-shaped microchannel configuration to achieve platelet
separation through asymmetrical electric fields created by DEP forces that influenced cell motion.
Micromachines 2020, 11, 890 13 of 15

4. Conclusions
In this study, we proposed a new zigzag-shaped microchannel configuration to achieve
platelet separation through asymmetrical electric fields created by DEP forces that influenced cell
motion. Structural parameters that influenced separation including the corner angle, channel width,
channel length, driving voltage, velocity ratio, and cell size were analyzed using the COMSOL
software package, leading to the identification of optimal structural parameters. In contrast to the
electric fields distributed by two electrodes in traditional DEP, the proposed device used a zigzag
scattered structure to arrange the electrodes, allowing optimal separation of RBCs and platelets. In the
configuration, RBCs and platelets were allowed to flow smoothly throughout the whole channel,
unlike with traditional electrode slots that can trap particles. When the corner angles were between
90 and 150◦ , the separation distance of the particles and the sizes of the corners were perfectly linearly
related (R2 = 1), while particles could not be separated at angles of 60◦ or less. Moreover, the length
and width of the channel significantly affected the DEP electric field distribution, with a quadratic
relationship best describing their relationships. The voltages and velocity ratios of the inlets affected
the movement of particles in the channel, wherein velocity ratios were required to be in a certain
range, otherwise particles could not normally flow from the inlet and outlet. A suitable velocity ratio
range was observed as 1:2–1:4, and the velocity of the buffer should be greater than that of the sample.
The voltage also affected the separation of particles and was proportional to the separation distance.
The voltage range at which separation occurs can be determined based on the separation structure.
Finally, perfect separation of platelets and RBCs (separation efficiency of almost 99.4%) was achieved on
a microfluidic chip using a voltage of 20 V. Thus, the DEP device using a zigzag microchannel exhibited
high efficiency and practical operability. Furthermore, this device can be used to separate other types
of cells or particles and demonstrates a new conceptual model for particle separation applications
in medical biology due to its convenience, high efficiency, and lack of environmental pollution.

Author Contributions: Y.G. designed the study. Y.L., H.L., S.L., M.B., and X.M. performed the experiments.
G.Y., F.X., and X.W. performed the simulations. Y.G. wrote the paper. Y.G. reviewed and edited the manuscript.
The manuscript was written with contributions of all authors. All authors have read and agreed to the published
version of the manuscript.
Funding: This research was funded by NSFC, grant number 51505128 and the Cultivation Program for Young
Backbone Teachers at the Henan University of Technology. It is also supported by the science and technology
scholarship program for students studying abroad in Henan Province.
Acknowledgments: The authors acknowledge undergraduate students Shuai Zhang and Chenyang Shao at the
Henan University of Technology for their assistance with experimental preparation.
Conflicts of Interest: The authors declare no conflict of interest.

References
1. Lee, J.; Rhyou, C.; Kang, B.; Lee, H. Continuously Phase-Modulated Standing Surface Acoustic Waves for
Separation of Particles and Cells in Microfluidic Channels Containing Multiple Pressure Nodes. J. Phys.
D Appl. Phys. 2016, 50, 165401. [CrossRef]
2. Antfolk, M.; Laurell, T.; Antfolk, M.; Laurell, T. Continuous flow microfluidic separation and processing of
rare cells and bioparticles found in blood—A review. Anal. Chem. Acta 2017, 965, 9–35. [CrossRef]
3. Imhof, H.K.; Schmid, J.; Niessner, R.; Ivleva, N.P.; Laforsch, C. A novel highly efficient method
for the separation and quantification of plastic particles in sediments of aquatic environments.
Limnol. Oceanogr. Methods 2012, 10, 524–537. [CrossRef]
4. Spieler, V.; Valldorf, B.; Spieler, F.; Valldorf, V.; Maaß, B.; Hüttenhain, S.H.; Kolmar, K. Coupled reactions on
bioparticles: Stereoselective reduction with cofactor regeneration on PhaC inclusion bodies. Biotechnol. J.
2016, 11, 890–898. [CrossRef] [PubMed]
5. Islinger, M.; Wildgruber, R.; Voelkl, A. Preparative free-flow electrophoresis, a versatile technology
complementing gradient centrifugation in the isolation of highly purified cell organelles. Electrophoresis 2018,
39, 2288–2299. [CrossRef] [PubMed]
Micromachines 2020, 11, 890 14 of 15

6. Sorsa, T.; Gieselmann, D.; Arweiler, N.B.; Hernández, M. A quantitative point-of-care test for periodontal
and dental peri-implant diseases. Nat. Rev. Dis. Primers 2017, 3, 17069. [CrossRef] [PubMed]
7. Gish, R.G.; Gutierrez, J.A.; Navarro-Cazarez, N.; Giang, K.; Adler, D.; Tran, B.; Locarnini, S.; Hammond, R.;
Bowden, S. A simple and inexpensive point-of-care test for hepatitis B surface antigen detection: Serological
and molecular evaluation. J. Viral Hepat. 2014, 21, 905–908. [CrossRef]
8. Syverud, B.C.; Lin, E.; Nagrath, S.; Larkin, L.M. Label-Free, High-Throughput Purification of Satellite Cells
Using Microfluidic Inertial Separation. Tissue Eng. Part C Methods 2018, 24, 32–41. [CrossRef]
9. Hisey, C.L.; Dorayappan, K.D.P.; Cohn, D.E.; Karuppaiyah, S.; Hansford, D.J. Microfluidic Affinity Separation
Chip for Selective Capture and Release of Label-free Ovarian Cancer Exosomes. Lab Chip 2018, 18, 15.
[CrossRef]
10. Ahmad, I.L.; Ahmad, M.R.; Takeuchi, M.; Nakajima, M.; Hasegawa, Y. Tapered Microfluidic for Continuous
Micro-Object Separation Based on Hydrodynamic Principle. IEEE Trans. Biomed. Circuits Syst. 2017,
11, 1413–1421. [CrossRef]
11. Shamloo, A.; Boodaghi, M. Design and simulation of a microfluidic device for acoustic cell separation.
Ultrasonics 2018, 84, 234–243. [CrossRef] [PubMed]
12. Savage, W.J.; Burns, J.R.; Fiering, J. Safety of acoustic separation in plastic devices for extracorporeal blood
processing. Transfusion 2017, 57, 1818–1826. [CrossRef] [PubMed]
13. Wu, T.H.; Chen, Y.; Park, S.Y.; Hong, J.; Teslaa, T.; Zhong, J.F.; Carlo, D.D.; Teitell, M.A.; Chiou, P.Y. Pulsed laser
triggered high speed microfluidic fluorescence activated cell sorter. Lab Chip 2012, 12, 1378–1383. [CrossRef]
[PubMed]
14. Rosental, B.; Kozhekbaeva, Z.; Fernhoff, N.; Tsai, J.M.; Traylor-Knowles, N. Coral cell separation and isolation
by fluorescence-activated cell sorting (FACS). BMC Cell Biol. 2017, 18, 30. [CrossRef] [PubMed]
15. Tamura, M.; Sugiura, S.; Takagi, T.; Satoh, T.; Sumaru, K.; Kanamori, T.; Okada, T.; Matsui, H.
Morphology-based optical separation of subpopulations from a heterogeneous murine breast cancer
cell line. PLoS ONE 2017, 12, e0179372. [CrossRef]
16. Strokin, N.A.; Bardakov, V.M. Development of Idea of Plasma-Optical Mass Separation. Plasma Phys. Rep.
2019, 45, 46–56. [CrossRef]
17. Pamme, N.; Wilhelm, C. Continuous sorting of magnetic cells via on-chip free-flflow magnetophoresis.
Lab Chip 2006, 6, 974–980. [CrossRef]
18. Oh, S.; Jung, S.H.; Seo, H.; Min, M.-K.; Kim, B.; Hahn, Y.K.; Kang, J.H.; Choi, S. Magnetic activated cell sorting
(MACS) pipette tip for immunomagnetic bacteria separation. Sens. Actuators B Chem. 2018, 272, 324–330.
[CrossRef]
19. Bayati, P.; Najafi, A. Electrophoresis of active Janus particles. J. Chem. Phys. 2019, 150, 234902. [CrossRef]
20. Maurya, S.K.; Gopmandal, P.P.; Ohshima, H. Electrophoresis of concentrated suspension of soft particles
with volumetrically charged inner core. Colloid Polym. Sci. 2018, 296, 721–732. [CrossRef]
21. Lin, C.-C.; Ho, Y.-H.; Yang, Y.-M.; Chang, H.-C. Combing Reman scattering technique with dielectrophoresis
chip for clinical isolates Helicobacter pylori analysis. In Proceedings of the 2008 3rd IEEE International
Conference on Nano/Micro Engineered and Molecular Systems, Sanya, China, 6–9 January 2008.
22. Pohl, H.A.; Pohl, H.A. Dielectrophosis: The Behavior of Neutral Matter in Nonuniform Electric Fields.
Q. Rev. Biol. 1980, 55, 6–69.
23. Pething, R. Dielectrophoresis: Status of the theory, technology, and applications. Biomicrofluidics 2010,
4, 022811. [CrossRef] [PubMed]
24. Huang, Y. Dielectrophoretic cell separation and gene expression profifiling on microelectronic chip arrays.
Anal. Chem. 2002, 74, 3362–3371. [CrossRef]
25. Rahmani, A.; Mohammadi, A.; Kalhor, H.R. A continuous flow microfluidic device based on contactless
dielectrophoresis for bioparticles enrichment. Electrophoresis 2017, 39, 445–455. [CrossRef]
26. Liao, S.-H.; Chang, H.-C.; Chang, H.-C. A capillary dielectrophoretic chip for real-time blood cell separation
from a drop of whole blood. Biomicrofluidics 2013, 7, 024110. [CrossRef] [PubMed]
27. Stroncek, D.F.; Rebulla, P. Platelet transfusions. Lancet 2007, 370, 427–438. [CrossRef]
28. Pommer, M.S.; Zhang, Y.; Keerthi, N.; Chen, N.; Thomson, J.A.; Meinhart, C.D.; Soh, H.T. Dielectrophoretic
separation of platelets from diluted whole blood in microfluidic channels. Electrophoresis 2008, 29, 1213–1218.
[CrossRef]
Micromachines 2020, 11, 890 15 of 15

29. Piacentini, N.; Mernier, G.; Tornay, R.; Renaud, P. Separation of platelets from other blood cells
in continuous-flow by dielectrophoresis field-flow-fractionation. Biomicrofluidics 2011, 5, 034122. [CrossRef]
30. Oblak, J.; Krizaj, D.; Amon, S.; Lebar, A.M.; Miklavčič, D. Feasibility study for cell electroporation detection
and separation by means of dielectrophoresis. Bioelectrochemistry 2007, 71, 164–17171. [CrossRef]
31. Shwetha, M.; Narayan, K. Mach-Zehnder interferometer for separation of platelets from red blood cells using
dielectrophoretics. Am. Inst. Phys. 2016, 1715, 020068.
32. Zhang, J.; Yan, S.; Sluyter, R.; Li, W.; Alici, G.; Nguyen, N.-T. Inertial particle separation by differential
equilibrium positions in a symmetrical serpentine micro-channel. Sci. Rep. 2014, 4, 4527. [CrossRef]
[PubMed]
33. Di Carlo, D. Inertial microfluidics. Lab Chip 2009, 9, 3038–3046. [CrossRef] [PubMed]
34. Zhang, J.; Yan, S.; Yuan, D.; Alici, G.; Nguyen, N.-T.; Ebrahimi Warkiani, E.; Li, W. Fundamentals and
applications of inertial microfluidics: A review. Lab Chip 2016, 16, 10–34. [CrossRef]
35. Feng, Q.L. Continuous Cells Manipulation and Separation Chip Using Dielectrophoresis Technology.
Master’s Thesis, Taiyuan University of Technology, Taiyuan, China, 2016.
36. Chen, J.K.; Yang, R.J. Electroosmotic flow mixing in zigzag microchannels. Vacuum 2012, 86, 1778–1782.
[CrossRef] [PubMed]
37. Qiang, Y.; Liu, J.; Yang, F.; Dieujuste, D.; Du, E. Modeling erythrocyte electrodeformation in response to
amplitude modulated electric waveforms. Sci. Rep. 2018, 8, 10224. [CrossRef] [PubMed]

© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access
article distributed under the terms and conditions of the Creative Commons Attribution
(CC BY) license (http://creativecommons.org/licenses/by/4.0/).