Download as pdf or txt
Download as pdf or txt
You are on page 1of 17

A STUDY ON APPLYING ION CONCENTRATION POLARIZATION IN MICROPUMP DESIGN

Khai H. Nguyen, Dung T. Nguyen, Van-Sang Pham*


School of Transportation Engineering, Hanoi University of Science and Technology. No1. Daicoviet
Road, Hai Ba Trung Dist, Hanoi, Vietnam, 100000.
*Corresponding author: sang.phamvan@hust.edu.vn

Abstract

This work focuses on applying the phenomenon of ion concentration polarization (ICP) to design and
investigate a micropump model that is a kind of electroosmotic micropump (EOP). An accurate
numerical solver was developed to solve the ion transport in the system. Via numerical simulation, we
conducted a detailed study on the manner in that ion transport generates electro-osmosis flow and
the influences of the ion transport on the pumping effect. The numerical simulation results show that
with increasing the applied voltage between the inlet and outlet of the micropump the ion
concentration polarization phenomenon next to the ion exchange membrane (IEM) and between both
end sides of the system occurs more observably. The thickness of the space charge layer also increases
in this tendency. The formation of the extended space charge layer and the actions of the electric field
upon this layer lead to electro-osmosis flow of the second kind. The instability of fluid flow that is a
large vortex in this layer at the high value of voltage enhances the system’s flow rate. Other obtained
results indicate the pumping effect can be enhanced by improving the geometry configurations.

Keywords: Concentration Polarization, Micropump, Poisson-Nernst-Planck-Navier-Stokes equations

1.0 INTRODUCTION

Micropumps are devices that can control and manipulate small fluid volumes. Notwithstanding that
any kind of small pump is often referred to as a micropump, a more accurate definition restricts this
term to pumps with functional dimensions in the micrometer range. They are normally categorized
into displacement pumps and dynamic pumps[1]. Displacement pumps, so-called mechanical pumps,
work by exerting pressure forces on the working fluid via one or more movable boundaries like a
piston, microvalve, or flaps. These types of pumps are commonly known as piezoelectric (PZT) pumps,
electrostatic pumps, and thermal-pneumatic pumps. Dynamic micropumps are classified into non-
mechanical micropump devices, such as centrifugal pumps, magnetohydrodynamic pumps,
electrohydrodynamic pumps, electroosmotic pumps, electrowetting, electrochemical pumps, etc. The
working fluid is provided energy continuously in a manner that increases either its momentum
(centrifugal pumps) or its pressure directly (electroosmotic, electrohydrodynamic pumps). When the
working fluid contains cells or other materials prone to damage or clogging dynamic micropumps are
advantageous.
Due to their non-mechanical feature, this kind of micropump has been reported to reduce erosion
flow effects and physical resistances while also optimizing micropump size for optimal pumping
efficiency.Electroosmotic pumps are capable of generating constant and pulse-free flows, controlling
the flow rate, the magnitude of flow velocity, and the direction of an EOP are convenient to control.
Since they have no moving parts and operate on the electrokinetic phenomenon, they can be
fabricated quite easily using microfabrication technologies and thus is readily integrable with lab-on-
chip (LOC) devices. Therefore, they have so far been used in various areas including high-performance
liquid chromatography (HPLC), separations, LOC assays, microelectronic equipment cooling, drug
delivery, and device actuation.Operating based on creating the electro-osmosis flow, so the flow rate
2

of EOPs is highly dependent on the cross-sectional dimensions of the structure in which the
electroosmotic flow is formed which is characterized by Debye length. Most EOPs are simply capillaries
or microchannel sections with flow rates produced that are typically very small
(𝑄𝑚𝑎𝑥 < 1𝜇𝑙 𝑚𝑖𝑛−1 ).
Although production of higher flow rates usig EO pumping generally requires structures with larger
dimensions in the directions normal to the flow, the effects of the electrical double layer (EDL)
decrease. This decrease leads to the flow rate of EOPs being small. In this work, we are introducing a
novel electroosmotic pump possessing architecture that operates on ion concentration polarization
(ICP)-a phenomenon that occurs when ions selectively pass through an ion exchange membrane (IEM).
The IEM is a functional membrane that has the permselectivity of ions. There are two types of IEMs:
anion exchange membrane (AEM) and cation exchange membrane (CEM) which are considered anion
and cation filters respectively. While the electrical field is applied across an ion exchange membrane,
an ion-enriched boundary layer and an ion-depleted boundary layer are formed next to the
membrane. One of the most interesting subjects of the ICP phenomenon is electroconvection (EC) as a
source of overlimiting conduction (OLC). In the overlimiting regimes which are defined in
characterizing the operation of the electromembrane system, the formation of the large vortex is the
consequence of the EC near the membrane. Unlike the electromembrane system which uses a
dominant normal electric field to induce the ICP phenomenon, our model is provided an electric field
where the dominant component is the tangential one intended to have the electroconvection at the
small applied bias voltage. Due to the action of the strong lateral electric field which incorporates that
in vertical direction upon the net space charge in the electrical double layer (EDL), the large vortex is
formed in the region near the IEM. The magnitude of flow velocity in the vortex increases when the
tangential electric field is stronger and the net charge in EDL is larger in the trend of increasing the bias
voltage between two sides of the IEM. To apply this alluring phenomenon in micropump design and
investigation, we use numerical modeling in theoretical consideration. Through studying the ion
transport characteristics across the IEMs specifically in the case of the diffusivity coefficient of sodium
ion and chloride ion, we have demonstrated that the depletion zone formed on the AEM is weaker
than that on the CEM. Therefore, CEM is applied to the numerical model to intend to generate a
strong ion depletion region from which, under the action of electric body force, the vortex formed in
this region has a high velocity. This improves the flow of this micropump model. Various applied
voltages are examined to elucidate their impact on the outcome which is the flow rate of the pump.
The disadvantage of the pump is that most of the vortex-based electroosmosis flows in the ion
depletion region cycle instead of going to the outlet. This detriment has been overcome by using a
vortex-breaking wall. Investigation of the dimension of the wall as well as its position relative to the
position of the membrane is performed to optimize the output flow of the micropump.
2.0 MICROPUMP DESIGN AND NUMERICAL METHOD
2.1. Micropump design
A schematic illustration of the micropump device is shown in Figure 1, where the system is restricted
by two insulating walls which are l0 (5µm) apart. The total length of the system is 3l0+Lm (31.25µm) in
which a cation exchange membrane (CEM) used and positioned in the direction of the bottom
insulating wall at its center has the length of Lm=0.25l0 (1.25 µm). A direct current (DC) voltage is
applied between two device ends, and a membrane electrode is set to zero, which generates an
electrical field including the tangential (ET) and normal (EN) components, respectively.
3

Figure 1. The schematic figure for the micropump design model.

2.2. Mathematical model


In order to bring out the detailed numerical modeling for the transport processes in the micropump,
we consider the computational domain as sketched in figure 1. In the domain, the appropriate
mathematical nature of the ion transport equations and physical boundary conditions are involved in
modeling the electrical and mechanical behavior of ion species, potential, and fluid flow.
In the initial operation condition, a salt solution is almost static in the channel of the micropump (Fig.
1b). When a voltage drop is applied between the inlet and outlet boundaries the positive and negative
ions in the system transport undergo the effect of the electric field, diffusion, and fluid flow
convection. The transportation is governed by mass conservation law represented via the Nernst-
Planck equations (1) and (2); the mutual effect of electric potential field and ion concentrations is
demonstrated by Poisson equation (3) and (4), and the motion of the fluid is described by the Navier-
Stokes equations (5) and (6). The dimensionless form of these equations is as follows:

1 𝐶̃± (1)
𝜕 = −∇̃ ∙ 𝐉̃± ,
𝜆̃𝐷 𝜕𝑡̃
𝐉̃± = −𝐷
̃ (∇
̃𝐶̃± + 𝑍± ∇
̃Φ̃ ) + 𝑃𝑒 𝐔
̃ 𝐶̃± , (2)
2
𝜆̃𝐷 ∇ ̃ ∙ (∇
̃Φ̃ ) = −𝜌̃𝑒 , (3)
𝜌̃𝑒 = 𝑍+ 𝐶̃+ + 𝑍− 𝐶̃− , (4)
1 1 𝜕𝐔 ̃ 1 (5)
̃𝑃̃ + ∇
= −∇ ̃2 𝐔
̃ − 𝑅𝑒(𝐔
̃∙∇
̃)𝐔
̃− ̃Φ
𝜌̃𝑒 ∇ ̃,
Sc 𝜆̃𝐷 𝜕𝑡̃ ̃𝜆𝐷 2
̃∙𝐔
∇ ̃ = 0 (6)
Where t̃, C̃± , Φ
̃, U
̃ and P
̃ denote the dimensionless time, concentration of cations (+) and anions (-),
electric potential, vector of fluid velocity, and pressure, respectively. These quantities are normalized
by the corresponding reference values of time, ionic concentration, electric potential, velocity, and
pressure, respectively as follows:
4

𝑙𝑜 2 𝑘𝐵 .𝑇 𝜀.𝛷𝑜 𝜂.𝑈𝑜 (7)


𝜏𝑜 = ; 𝐶𝑜 = 𝐶𝐵𝑢𝑙𝑘 ; Φo = ; 𝑈𝑜 = ; 𝑃𝑜 =
𝐷𝑜 𝑍𝑒 𝜂.𝑙𝑜 𝑙𝑜

Where C0 is the concentration scale, U0 is the velocity scale, l0 is the characteristic length scale, 𝐷0 =
(𝐷+ + 𝐷− )/2 is the average diffusion, k B is the Boltzmann constant, T is the absolute temperature, e
is the elementary charge, 𝑍 = |𝑍± | is ion valence, η is the dynamic viscosity of the solution, and ε is
the permittivity of the solvent. Parameters D ̃ ± = D±, λ ρ
λ̃D = D, and ρ̃e = e are dimensionless
D0 l0 C0
ε𝑘𝐵 𝑇
diffusion coefficient, the Debye length (𝜆𝐷 = √2Z2𝑒 2 𝐶 ), and the space charge, respectively. Pe =
0
l0 η ρm
U0 D , Sc = ρ D0, and Re = U0 l0 are the Péclet number, the Schmidt number, and the Reynolds
0 m η
number, respectively. The system is characterized by the dimensionless Debye length in this study,
λ̃D = 0.0028 corresponds to the characteristic length 𝑙0 = 5 𝜇m. The concentration of ions in bulk
space C0 = CBulk = 0.5 mM.
The action of electric field E generates an electric force upon the ions in the electric double layer
along to electric field line, the motion of ion causing the bulk fluid motion by viscous drag force and
produces an electroosmosis flow 𝑈𝐸𝑂 (m/s):

𝜀0 𝜀𝑟 𝜁𝑤 (8)
𝑈𝐸𝑂 = −
𝜇

𝛿𝑤 𝜆𝐷
For the Ohmic regime (I ≤ Ilim). Herein, 𝜁𝑤 = is the zeta potential of a polydimethylsiloxane
𝜀0 𝜀𝑟
(PDMS) material (V), 𝛿𝑤 = -0.0065 is the charge density on the solid surface (C/m2), and 𝜇 is dynamic
viscosity (Ns/m2). Or,

𝜕 2𝑐
1 𝜕𝑥𝜕𝑦 (9)
𝑈𝐸𝑂 = − 𝐸2
8 𝜕𝑐
𝜕𝑦
2.3. Numerical method
The Poisson-Nernst-Planck (PNP) and Navier-Stoke (NS) equations are nonlinearly coupled. The
relationship between these equations is described via the convection term in the PNP equations and
the electric body force in the NS equation. The finite volume method [15-17] which is a locally
conservative method is used for the discretization of the equationss. To avoid solving the large system
of linear equations and guarantee the strong coupling of the PNP and NS equations, in this work, we
employed the coupled method proposed by Pham to solve the sets of equations [15]. By starting with
a velocity field from the previous iteration or initial condition, the potential and concentrations are
simultaneously solved from the PNP equations. Then, electric body force is calculated and substituted
into the NS equations. The velocity field obtained by solving the NS equations is substituted back into
the PNP equations. The process is repeated until convergence is reached. The nonlinear discrete PNP
equations are solved using the Newton-Raphson method [18]. In the finite volume method, the solving
Navier-Stoke equations can be conducted separately or simultaneously. In the flow found next to the
permselective membrane, the electric body force varying dramatically in the membrane’s EDL causes a
large pressure gradient in the surface's normal direction. To ensure the unity coupling of pressure and
5

velocity. The Navier-Stoke equations are solved simultaneously. The Rhie-Chow interpolation [16] is
used to derive an explicit equation for the pressure.
To resolve the rapid variations of the ion concentrations and electric potential near charged surfaces,
the mesh near the membrane is extremely refined toward the surfaces by the GMSH software [19]
with a total volume of 40348 cells, and the smallest cell size is 9.38 × 10−12 for the cell next to the
intersection of walls and both ends of the membrane, while the biggest cell size is 5.93 × 10−3 , found
in the cell at the membrane location of the model. All simulations were conducted using the Asus
Vivobook laptop with Intel i5 processors running at 1.60GHz and 8GB of RAM. The total time required
to solve a pair of PNP-NS equations is 40 seconds, with most of the time spent computing the Poisson-
Nernst-Planck equations system.
2.4. Boundary conditions
At two ends of the micropump, the boundary condition for fluid flow is the free-stream condition; the
concentration of both cations and anions are also given and remains constant during system operation
that representing the actual configuration in which two connect to two electrolyte solution reservoirs;
the electric potential is given at two ends of micropump to enforce a rightward electric field that
almost defines the ions transport in the system:

∂𝐔
= 0; C+ = C− = C0 ;
∂n
Φ = 0 : at the inlet
Φ = Φapplied : at the outlet

At the wall, the no-slip condition is applied for fluid flow; the no-ionic flux condition is applied for both
anions and cations; the electric potential is also assumed to be floating.

∂Φ
U = 0; J− ∙ n = J+ ∙ n = 0; =0
∂n

On the CEM boundaries, to mimic the permselective feature of the membrane, a fixed value condition
is applied for cation concentration, while no ionic flux condition is enforced for anions; the no-slip
condition is applied for fluid flow; the electric potential is given at the membranes to enforce a
downward electric field that contributes to the process of ions transport in the system.

C+ = 1.5C0 ; J− ∙ n = 0; U = 0;
Φ = 0 : at the CEM

3.0 RESULTS AND DISCUSSION

One of the most critical factors affecting the results of this micropump model is the magnitude of the
vortex velocity. Since this factor is mainly determined by the applied electric field and its consequence
which is the distribution of ion concentrations in the system (i.e the increasing net space charge in EDL
which is next to the membrane and the electric body force acting upon it ), we investigated the model
with the voltage of 2Φ0 , 4Φ0 , 6Φ0 (Φ0 = k B T/e = 25.85 mV). The analysis of the distribution of ion
concentration along the horizontal and vertical directions with a detailed one at the special points of
6

the model is performed. From these results, the discussions on the outcome of the investigation will
be provided and then we will suggest an improved solution for the flow rate of the micropump using a
prevent-flow structure (PFS).
3.1. The ion distribution.
In the initial condition, the fluid in the micropump seems static. Then, when the electrical field is
applied to the system, the solution almost undergoes the electrostatic force. Therefore, the cations
tend to be attracted to the cathodic side of the system while the anions tend to move toward the
opposite side. During the process of going to the cathodic side of the micropump, under the action of a
vertical electrical field and the, a fairly large number of cations pass through the CEM. The remainder is
positioned near the anode to neutralize the charge with the anions attracted to this electrode. As a
result of this process, the ions in the micropump are concentrated in the area next to the anodic side
and diluted on the cathodic side along the horizontal direction of the system as shown in Figure 2a.
Although the figure displays the cation concentration, by the existence of electroneutrality, the
anion concentration at most points in the system is almost equal to the cation concentration except
for points in the space charge layer (SCL) which will be discussed in the next paragraph. From the
figure, we can see the larger the applied bias voltage the steeper the ion concentration gradient goes
leftwards and the change of ion concentration along the horizontal direction mostly occurs on the left-
hand side of the model. At points in the system with the same abscissa, the cation concentration
values seem equal. This does not happen at points where the abscissa ranges from 2.9 to 3.35 due to
the formation of an ion-depleted layer next to the CEM. Therefore, we plot the concentration of cation
in the horizontal direction at the system’s midpoint for the three voltages of 2Φ0 , 4Φ0 , and 6Φ0 which
are shown in Figure 2b. The three curves have almost the same form, where the cation concentration
is highest at the anode and then gradually decreases as the value of x increases from 0 to 3.25, while
this value ranges from 3.25 to 6.25 then the cation concentration does not seem to change. However,
when the value of x ranges from 0 to 3.25 the cation concentration decreases with an increased slope
coefficient corresponding to an increase in the applied voltage. Specifically, in the case of the voltage
of 2Φ0 , the cation concentration at x of 3.25 is ~0.71, a 30% reduction from the maximum value of
the cation concentration of 1 which is obtained at the position near the anode. Meanwhile, these
reductions in the case of the voltage of 4Φ0 and 6Φ0 are 60% and 85% respectively. This happens
because when increasing the voltage at the two ends of the system, the effect of the electric field on
the ions will be stronger and the ion transport must comply with the requirement of the conservation
of mass.
7

(a)

(b)
Figure 2. The ions distribution inside the micropump channel. (a) The contour of cation in micropump
at three of dimensionless voltage, 2ϕ0, 4ϕ0, 6ϕ0 respectively. (b) The ion concentration distribution
plotted in horizontal direction at midpoint of model.

By applying a voltage of 0 on the CEM boundary, the normal component of the electric field (𝑬𝑁 ) has a
significant effect on ion transport in the system. Under the action of the electric field in which an
important role is played by the normal component of the electric field, the cations pass through the
CEM. By the requirements of electroneutrality and the permselectivity of the CEM, an ion-depleted
region is formed close to the CEM. This region expands in the direction perpendicular to the
membrane, and the ion concentration of this region decreases in the increasing trend of electric field
applied to the system as shown in Figure. In the ion-depleted, the ion concentration in the direction
8

perpendicular to CEM at its midpoint in the range of y from 0 to 0.04 is plotted. Unlike the system
where the lateral electric field is weak, the thickness of space charge layer increases because of the
formation of the extended space charge layer in condition of critically strong vertical electric field. The
space charge layer in our model in three cases (0.012 for voltage of 4Φ0 , 0.017 for voltage of 2Φ0 ,
0.023 for voltage of 6Φ0 ) are much thicker than the initial EDL (0.0028) due to the effects of the strong
external lateral electric field.

Figure 3. The extended space charge layer (SCL) thickness. The examination at 2𝛟0, 4𝛟0, 6𝛟0 voltage
values present the surge in thickness of the EDL layer proportional with the electric field value
generated

3.2. The affects of magnitude of fluid flow in vortex on the velocity of output flow.

The action of the electric body force upon the space charge layer leads to the formulation of the fluid
convection caused by electro-osmotic slip of second kind. In the typical electro-membrane systems
where the action of the lateral electric field is weaker than that of the vertical and small net charge in
EDL, the fluid convection so is formed in the condition in which the bias voltage exceeds the critical
value. Meanwhile, in this micropump design, the strong lateral electric fields act upon the large space
charge, so the vortex is formed at much lower voltages than in conventional electromembrane
systems as shown in Figure. As can be seen, the large vortex is even formed at the low voltage of 2Φ0 .
When the bias voltage between two ends of the pump increase, the stepper concentration gradient
between two ends of membrane produces a stronger amplification to electric field. The combined
effect of electric field amplification and pressure drop accelerating the rotation in the vortex is
essentially positive feedback and promotes electroconvective instability in the system. Therefore, the
magnitude of the flow velocity in the large vortex in direction perpendicular to the membrane at it’s
middle point increases in the increasing trend of applied bias voltage. The maximum magnitude of flow
velocity in the vortex is ~1.2𝑈0 for the applied voltage of 6Φ0 while this value in the case of the
voltages of 4Φ0 , 2Φ0 are ~0.4𝑈0 and ~0.1𝑈0 , respectively.
9

Figure 4. The electroosmosis flow in the pumps at 2𝛟0, 4𝛟0, and 6𝛟0 voltage, respectively.

Due to depending on the magnitude of flow velocity in the vortex of the wanted outcome, which is
the flow rate banking on the velocity of output flow, we considered the magnitudes of flow fluid in the
vertical direction at the middle point and the outlet boundary of the model as shown Figure 5. In all
three voltages, the fluid flow velocities in vortex have a ~10 times larger magnitude than that at the
outlet boundary.

(a) (b)
Figure 5. Examining the pumping effect of the micropump device through a cutting line. (a) the flow
rate plot over the output vertical line and (b) the vortices speed induced by the tangent electric field
near the membrane location.

The advantage is the pumping flow rate can manipulate via voltage control. The obtained result in
figure 5 demonstrates the increase in the electric field promotes an increase in flow velocity in the
10

micropump channel. However, a defect is also revealed thanks to the simulation result. The disparity
between the vortices velocity in the middle of the channel and the output velocity observed is
extremely considerable. For instance, at 6𝜙0, the highest vortices velocity is 1.2U0, but the output
velocity is just only 0.1U0 - near a 10-fold reduction. This figures out the fact that the design is still not
fully optimized yet.

(a) (b)

(c)
Figure 6. The vortex flows in the micropump channel induced by three different voltage, increasing
from 2𝚽𝟎 , 4𝚽𝟎 , to 6𝚽𝟎 , respectively with a, b, and c. The value comparative with collor code (right
hand side).

The development of vortices flow in the middle of the channel is the result of the combination of
electroosmotic flow and vertical vicious drag fluid flow. These vortices tend to amplify themselves
(Fig.6a, b, c) without increasing the pumping flowrate (Fig.5a), resulting in the creation of a high-speed
zone near the membrane, which has an unexpectation results in adverse effect on micropump
performance.

3.3. Improving pumping flow rate

In this part, a second micropump arrangement will be provided to enhance the pumping flow rate and
eliminate the above-mentioned drawbacks. A prevent-flow structure (PFS) is inserted in the region
above the membrane, in this second configuration as shown in Figure 7, to block the fluid flow, the
11

whole of the first-design system's features remains identical. The system will be analyzed using two
analytical variables: (a) the PFS location and (b) the PFS dimensions.

Figure 7. The sketch estimating the second configuration in our micropump design. In this
configuration, three prevent-flow-structure positions (i), (ii), and (iii) are placed at the end left, center,
and end right of the membrane, respectively. Forth of the first design remained.

a) The PFS location


During estimating process, the crucial is to determine the impact of PFS placement on micropump
actuation ability. Here, we propose three locations based on the formation of vortices, as shown in Fig
7, where (i) the membrane’s left-hand side end, (ii) the membrane’s right-hand side end, and (iii) the
membrane’s center.
The simulation results show the pumping efficiency varies depending on the PFS where it is
positioned. The pumping flow rate is maximum when it is placed at the (i) - position of the membrane;
by contrast, it is reduced in the two positions left, (ii) and (iii). Figure 8 is a result that indicates the
position of the vortices in the microchannel affected by the membrane and the PFS placement. All
vortices appear at the membrane's first edge and gradually destabilize backward to the membrane's
ends. Basically, determining the location and stability of the vortices is a requirement to execute the
PFS design.
12

Figure 8. Estimating the micropump performance based on the PFS position at the same voltage for
the three consideration cases

In the cases of Figure 8-ii, iii, obviously, the vortices tend to grade to the LHS corner of the PFS, while
in the case of Figure 8-i the vortices lie in the middle of the PFS placement region. Thus, case 8i is a
positive candidate for manipulating vortices by enhancing the depth of the PFS way. This method is
possible to remove the vortices away from the devices. Additionally, the flow rate value in three
positions also indicates that the 8i-case has the highest flow rate (Fig.9). For all these reasons, case 8i
is the ideal place to construct a PFS design.

Figure 9. The flowrate of the micropump at three PFS locations

b) The PFS dimensions


13

Utilizing the PFS position that is determined in the previous part. Now, we will investigate several
cases of the PFS dimension. To insight into the effect of PFS on the prevent eddy flow target, as well as
optimize the micropump system, this task we will build and simulate 42 cases for the dimensions of
the 2D - PFS model, where h0 = 5 µm, and PFS dimension, width x depth, range in [0.3, …,0.8] x [0.3,
...,0.9] reference by h0, besides that whole former micropump design size unchanged.

(a) (b)
Figure 9. The PFS dimension simulation in multiple test cases. (a) the collected data in one block
describes the micro-pumping output flow rate and (b) extracts three of the highest pumping flowrate
obtained -the plot output velocity over the line graph.

Figure 10 depicts the result of 42 simulation instances. Where the PFS dimension is manipulated to
obtain the highest pumping flow rate. When the PFS is sufficiently deep, as in the previous mechanism,
the tangential electric field (ET) and the zeta potential forms an electroosmosis flow (EO) along the
microchannel's length. At the same time, due to the free space between PFS and membrane being
quite small, the fluid flow generated by ions passing through the CEM can be negligible. Therefore, by
this implementation, the fluid movement within the micropump is exaggerated, and no vortex
formation occurs.

Figure 10. The flow profile inside the micropump device at the PFS dimension 0.5h 0x0.9h0
14

The discovery of the best configuration of the micropump reaches as the PFS dimension is 0.5h0x0.9h0
(Fig.11).

3.4. Micropump performance

Following the design model above, this section will continue to consider an IEM-micropump model
with a 0.5h0x0.9h0 sample, integrating a prevent-flow structure (PFS). The results will report some
micro-pump characteristics and compare them with several available papers. The impacts of the
pumping performance such as pressure drop as well as pumping flow rate are also presented in this
section.

a) The pressure drops

Recently, the research on microfluidics is gradually increasing [20-25], which makes the understanding
of electro-osmosis no stranger definition. Following it, this study conveys the interaction between the
electric field and free ions in the EDL layer, the pumping effect thank that was produced.
In this system, the pressure drop is not the central driving force that induces EO flow. Instead, the EO
flow is more like a byproduct of fluid motion due to drag viscous force. This is the opposite in the
comparison with conventional pumps (Fig.12).
Figure 12 depicts the increase in pressure drop caused by increased fluid velocity in the pumps.
Obviously, the apparent pressure drop in the system is still negligible; as a result, pressure drop can be
ignored when building the real device

Figure 11. Pressure drop in the estimating micro-pumps follows the increase of electric field

b) Pumping Flowrate

Despite recent developments in microfluidic technology, micropump systems that employ this
technology still provide limited pumping flowrates. For instance, in Matthew's studies [26], the EO
micropump's maximum pumping flow rate is around 20-25 m/min, or in the AC-EO micropump,
15

Yoshida's device [27], is constructed with a dimension of 0.2x0.2x0.05 mm3, achieves a 1.6 mm/s
output velocity.
This work, based on a knowledge of the EO phenomena as well as inheritance of previous works, the
optimizing the pumping flow rate from the nanodevice using a simple construction of the micropump
was conducted. The acquired results demonstrated the importance of this finding. It's a significant
achievement in the field of micropump fabrication (Fig.13).

(a) (b)
Figure 12. Examination of the proposal device with (a) the pumping velocity and (b) the actual flow
rate of the device
The pumping flowrate is a significant finding in this work. Figure 13a is the value of velocity, and
figure 13b is the pumping flowrate gained following the voltage scale. Herein (Fig 13.b) the flowrate is
divided following the logarithmic scale in response to each voltage. Obviously, the data indicated a
dramatic change in the systems (change by logarithmic function). It is a demonstration that we can
obtain an arbitrary pumping flow rate by manipulating an electric field. In this instance (Fig 13), the
flow rate reached 0.098 ml/s at 8V0 (6600 V/m).

4.0 CONCLUSION

In this work, the fluid-accelerated phenomenon at the extend-space-charge layer is used to fabricate
a micropump. The numerical simulation method is used to investigate velocity, flow rate, and fluid
profile, as well as to explain a variety of physical flow acceleration phenomena. Based on the
simulation results, we propose a new micropumps design for flowrate optimization that includes an
ion-exchange membrane, salt liquid, electric supply, and an obstacle physical object. The simulation
findings reveal that the micropump flow rate can be increased by a factor of a hundred times. Also,
due to the scope of the study, we only simulated one microchannel; however, in practice, we can use a
variety of typical microchannel alignments to reach the expected flow rates. Such as mesh
microelectrodes alignment [28] or cascaded type [29]. This research also proposes a concept for
designing and fabricating dynamic-micropump micropumps that do not require mechanical structures,
are simple to manufacture and are simple to operate and maintain.

Acknowledgement

This research was funded by the Ministry of Education and Training under Project No. B2022-BKA-07.
16

References

[1] D. J. Laser and J. G. Santiago (2004). "A review of micropumps,". J. Micromechanics


Microengineering, vol. 14, no. 6. https://doi.org/10.1088/0960-1317/14/6/R01
[2] Byun, C. K., Abi-Samra, K., Cho, Y.-K., & Takayama, S. (2013). Pumps for microfluidic cell culture.
ELECTROPHORESIS, 35(2-3), 245-257. https://doi.org/10.1002/elps.201300205
[3] Zhou, C., Zhang, H., Li, Z., & Wang, W. (2016). Chemistry pumps: a review of chemically powered
micropumps. Lab on a Chip, 16(10), 1797-1811. https://doi.org/10.1039/C6LC00032K
[4] Sim, J. Y., Haney, M. P., Park, S. I., McCall, J. G., & Jeong, J.-W. (2017). Microfluidic neural probes: in
vivo tools for advancing neuroscience. Lab on a Chip, 17(8), 1406-1435.
https://doi.org/10.1039/C7LC00103G
[5] Yuan, F., & Isaac, K. M. (2017). A study of MHD-based chaotic advection to enhance mixing in
microfluidics using transient three dimensional CFD simulations. Sensors and Actuators B:
Chemical, 238, 226-238. https://doi.org/10.1016/j.snb.2016.07.063
[6] Rybkowska, N., Strzelać, K., & Koncki, R. (2018). A comparison of photometric methods for serum
iron determination under flow analysis conditions. Sensors and Actuators B: Chemical, 254, 307-
313. https://doi.org/10.1016/j.snb.2017.07.059
[7] V. S. Pham, Z. Li, K. M. Lim, J. K. White, and J. Han (2012). "Direct numerical simulation of
electroconvective instability and hysteretic current-voltage response of a permselective
membrane," Phys. Rev. E - Stat. Nonlinear, Soft Matter Phys, vol. 86, no. 4, pp. 1-11.
https://doi.org/10.1103/PhysRevE.86.046310
[8] Nguyen, V.-B., & Pham, V.-S. (2020). Study and modeling DNA-preconcentration microfluidic
device. Journal of Science and Technology, 143(1), 1-6.
[9] Kwon, Hyukjin J., Bumjoo Kim, Geunbae Lim, and Jongyoon Han (2018). "A Multiscale Pore Ion
Exchange Membrane for Better Energy Efficiency." Journal of Materials Chemistry A 6 (17). Royal
Society of Chemistry: 7714-23. https://doi.org/10.1039/C7TA10570C
[10] Rubinstein, I., & Zaltzman, B. (2000). Electro-osmotically induced convection at a permselective
membrane. Physical Review E - Statistical Physics, Plasmas, Fluids, and Related Interdisciplinary
Topics, 62(2), 2238-2251. https://doi.org/10.1103/PhysRevE.62.2238
[11] Dukhin, S. S. (1991). Electrokinetic phenomena of the second kind and their applications.
Advances in Colloid and Interface Science, 35, 173-196. https://doi.org/10.1016/0001-
8686(91)80022-C
[12] Primicerio, M., Rubinstein, I., & Zaltzman, B. (1999). Electrodiffusional free boundary problem, in a
bipolar membrane (semiconductor diode), at a reverse bias for constant current. Quarterly of
Applied Mathematics, 57(4), 637-659. https://doi.org/10.1090/qam/1724297
[13] J. Kim, I. Cho, H. Lee, and S. J. Kim (2017). "Ion Concentration Polarization by Bifurcated Current
Path" Sci. Rep., vol. 7, no. 1, pp. 1-12. https://doi.org/10.1038/s41598-017-04646-0
[14] B. J. Kirby, Micro- and Nanoscale Fluid Mechanics: Transport in Microfluidic Devices, 536,
Cambridge University Press, United States of America, first edition 2010.
[15] M. Darwish, I. Sraj, and F. Moukalled (2009). "A coupled finite volume solver for the solution of
incompressible flows on unstructured grids" J. Comput. Phys., vol. 228, no. 1, pp. 180 201.
https://doi.org/10.1016/j.jcp.2008.08.027
[16] Rhie, C.M. and Chow, W.L. (1983). "Numerical Study of the Turbulent Flow past an Airfoil with
Trailing Edge Separation". AIAA Journal, 21, 1525-1532. https://doi.org/10.2514/3.8284
[17] J. H. Ferziger and M. Peric (2001). Computational Methods for Fluid Dynamic 3rded. Springer.
17

[18] J. E. Dennis, Jr. and R. B. Schnabel (1983). Numerical Methods for Unconstrained Optimization and
Nonlinear Equations.Prentice-Hall, Englewood Cliffs, NJ.
[19] C. Geuzaine and J. F. Remacle (2009). Comput. Methods Appl. Mech. Eng. 79, 1309.
https://doi.org/10.1002/nme.2579
[20] Viefhues, Martina, and Ralf Eichhorn (2017). "DNA Dielectrophoresis: Theory and Applications a
Review." ELECTROPHORESIS 38 (11): 1483-1506. https://doi.org/10.1002/elps.201600482
[21] Nobari, M.R.H., S. Movahed, V. Nourian, and S. Kazemi (2016). "A Numerical Investigation of a
Novel Micro-Pump Based on the Induced Charged Electrokinetic Phenomenon in the Presence of a
Conducting Circular Obstacle." Journal of Electrostatics 83 (October): 97-107.
https://doi.org/10.1016/j.elstat.2016.08.007
[22] Hoshyargar, Vahid, Atieh Khorami, Seyed Nezameddin Ashrafizadeh, and Arman Sadeghi (2018).
"Solute Dispersion by Electroosmotic Flow through Soft Microchannels." Sensors and Actuators B:
Chemical 255 (February): 3585-3600. https://doi.org/10.1016/j.snb.2017.09.015
[23] Liu, Weiyu, Jinyou Shao, Yukun Ren, Yupan Wu, Chunhui Wang, Haitao Ding, Hongyuan Jiang, and
Yucheng Ding (2016). "Effects of Discrete-Electrode Arrangement on Traveling-Wave
Electroosmotic Pumping." Journal of Micromechanics and Microengineering 26 (9): 095003.
https://doi.org/10.1088/0960-1317/26/9/095003
[24] Lin, Shiang-Chi, Yu-Lung Sung, Chien-Chung Peng, Yi-Chung Tung, and Chih-Ting Lin (2017). "An In-
Situ Filtering Pump for Particle-Sample Filtration Based on Low-Voltage Electrokinetic
Mechanism." Sensors and Actuators B: Chemical 238 (January): 809-16.
https://doi.org/10.1016/j.snb.2016.07.147
[25] Ranjit, N.K., and G.C. Shit (2017). "Entropy Generation on Electro-Osmotic Flow Pumping by a
Uniform Peristaltic Wave under Magnetic Environment." Energy 128 (June): 649-60.
https://doi.org/10.1016/j.energy.2017.04.035
[26] Suss, Matthew E., Ali Mani, Thomas A. Zangle, and Juan G. Santiago (2011). "Electroosmotic Pump
Performance Is Affected by Concentration Polarizations of Both Electrodes and Pump." Sensors
and Actuators A: Physical 165 (2): 310-15. https://doi.org/10.1016/j.sna.2010.10.002
[27] Yoshida, Kazuhiro, Tomoyuki Sato, Sang In Eom, Joon-wan Kim, and Shinichi Yokota (2017). "A
Study on an AC Electroosmotic Micropump Using a Square Pole - Slit Electrode Array." Sensors and
Actuators A: Physical 265 (October): 152-60. https://doi.org/10.1016/j.sna.2017.08.026
[28] Gao, Meng, and Lin Gui (2016). "Electroosmotic Flow Pump." In Advances in Microfluidics - New
Applications in. https://doi.org/10.5772/64601
[29] Takamura, Y., Onoda, H., Inokuchi, H., Adachi, S., Oki, A., & Horiike, Y. (2003). Low-voltage
electroosmosis pump for stand-alone microfluidics devices. ELECTROPHORESIS, 24(12), 185-192.
https://doi.org/10.1002/elps.200390012

You might also like