You are on page 1of 11

International Journal of Engineering Science 94 (2015) 128–138

Contents lists available at ScienceDirect

International Journal of Engineering Science


journal homepage: www.elsevier.com/locate/ijengsci

Electroosmotic flow in the vicinity of a conducting obstacle


mounted on the surface of a wide microchannel
Subrata Bera 1, S. Bhattacharyya ⇑
Department of Mathematics, Indian Institute of Technology, Kharagpur, Kharagpur 721302, India

a r t i c l e i n f o a b s t r a c t

Article history: A numerical study is made on the electroosmotic flow (EOF) near a polarizable metallic
Received 28 July 2014 obstacle mounted on one of the non-conducting walls of a microchannel. The external elec-
Received in revised form 16 April 2015 tric field induces a Debye layer of non-uniform f-potential along the obstacle, which results
Accepted 26 April 2015
in a non-linear electroosmotic flow. The combined effect of surface roughness and
Available online 14 June 2015
non-uniform electric double layer on the polarizable obstacle creates a vortical flow. The
form of this vortical flow and its dependence on the bulk ionic concentration is analyzed.
Keywords:
Our numerical model is based on computation of the Navier–Stokes, Nernst–Planck and
Induced surface potential
Electroosmosis
Poisson equations. We have computed the governing non-linear coupled set of equations
Nernst–Planck equations by the control volume method over a staggered grid system. Our results show that the form
Micro-vortex of the vortical flow, which develops in the vicinity of the obstacle, depends on the thickness
of the Debye layer along the homogeneous part of the channel. The free-slip model, based
on zero charge density in the bulk region is found to over-predict the result based on the
present model. However, the present model approaches the free-slip model when the bulk
ionic concentration is considered to be sufficiently strong.
Ó 2015 Elsevier Ltd. All rights reserved.

1. Introduction

Electroosmosis offers an alternative means to pressure gradients to drive flow in microchannels. Electrokinetic phenom-
ena provides one of the most popular non-mechanical techniques in microfluidics. This has drawn wide interest due to its
robustness, no dynamic parts and easy to operate conditions. Electroosmosis is the preferred mode for manipulating fluids in
microdevices is gaining increased attention. One of the most promising applications of microfluidics is the lab-on-a-chip
device. EOF is the bulk fluid motion driven by the electrokinetic force acting on the net charged ions in the diffuse layer;
the outer part of an electrical double layer (EDL). For an EOF in a microchannel with thin EDL (linear EOF), the EOF is often
modeled using a simple slip velocity condition known as the Helmholtz–Smoluchowski equation. The free-slip condition on
the outer edge of the diffuse layer is based on the assumption that the electric field lines are tangent to the outer edge of the
diffuse layer and no transport of ions occur into or out of this diffuse layer. This boundary condition provides a linear rela-
tionship between the slip velocity and the local applied electric field. The outside flow is governed by the viscous diffusion
and the fluid is considered electrically neutral.

⇑ Corresponding author. Tel.: +91 3222283640.


E-mail address: somnath@maths.iitkgp.ernet.in (S. Bhattacharyya).
1
Present address: Department of Mathematics, National Institute of Technology Silchar, Silchar 788010, India.

http://dx.doi.org/10.1016/j.ijengsci.2015.04.005
0020-7225/Ó 2015 Elsevier Ltd. All rights reserved.
S. Bera, S. Bhattacharyya / International Journal of Engineering Science 94 (2015) 128–138 129

The non-linear electrokinetic phenomena provides a promising alternative mechanism for flow control in microfluidic
devices. The vortical flow and mixing through modulation of channel wall potential has been studied by several authors
namely, Ghosal (2002), Erickson and Li (2002), Yariv (2004),Fu, Lin, and Yang (2003), Bhattacharyya and Nayak (2009)
and Chen and Conlisk (2009), Lin and Chen (2013). It is established in those forgoing studies on EOF in a surface modulated
microchannel that the potential patch of opposite sign may induce a vortex, the bulk flow being governed by the induced
pressure gradient, viscous diffusion and electric body force. The concentration polarization near a discontinuous jump in sur-
face charge is studied by Khair and Squires (2008).
When an inert conducting metallic surface is embedded in a dielectric medium, an external field will cause the surface to
polarize and an induced surface charge density develops. Unlike the standard electrokinetic phenomena where the
surface-charge density is regarded as a material physicochemical property of the solid-electrolyte pair, the induced
surface-charge density is proportional to the external field. The electric field drives an ionic current in the electrolyte. These
ions can not penetrate the conducting surface and accumulate in a form of a charged cloud (Debye layer) near the surface.
This charge cloud grows and a diffusive current develops so as to oppose the charging. This process reaches a steady state,
typically within a fraction of a second (Squires & Bazant, 2004). At steady state the conductor behaves like an insulator and
expels the electric field lines. A non-uniform charge density develops around the polarizable surface, which results in a
non-constant f-potential. The current drives positive ions along the surface where the initial current enters and negative ions
where it leaves. Consequently the variable f-potential along the conducting surface have a change of sign. This is the reason,
the electrokinetics around a polarizable surface differ from the conventional EOF. Squires and Bazant (2004, 2006) referred
the electroosmosis above a polariazable surface as the induced charge electroosmosis (ICEO) and the electrophoresis of a
polarizable particle as induced charge electrophoresis (ICEP).
ICEO is a non-linear phenomena and many result into formation of vortices in a microchannel. The circulation and
enhancement of species mixing by introducing a conducting surface in a microchannel was investigated by Wu and Li
(2008a). Subsequently, Wu and Li (2008b) studied the ICEO around conducting hurdles which are embedded within a
microchannel. Eckstein, Yossifon, Seifert, and Miloh (2009) made a combined numerical and experimental study on vortex
generation phenomena around sharp corners in microfluidic devices through the ICEO mechanism. Sharp, Yazdi, and Davison
(2011) used the induced-charge electroosmotic flow combined with an applied bulk pressure driven flow for localized con-
trol near conducting obstacles. Zhao and Yang (2009) studied the ICEO flow within a long microchannel patterned with two
symmetric polarizable dielectric blocks. The recent progress on ICEO and its various applications have been discussed by
Bazant and Squires (2010).
Microvortices have an advantage in species mixing in microdevices. In other situations, the appearance of vortices need to
be suppressed so as to avoid aggregation of suspended particles leading to the eventual jamming of the device. In any case,
the study on vortical flow in ICEO is important. The ICEO near a conducting surface has been used by several authors (e.g.,
Zhang et al., 2011) for separation, translocation and enrichment of nanoparticles. Sharp et al. (2011) applied the ICEO near a
conducting obstacle for the flow control and manipulation of suspended particles in a microdevice.
If the Debye layer thickness is much less than the geometric length scale of interest, the flows can be captured by pre-
scribing a slip velocity at the outer edge of the Debye layer. The slip velocity is governed by the Smoluchowski velocity
i.e., U HS ¼ e fE0 =l, where e is the permittivity of the medium, E0 is the applied field, l is the viscosity and f is the potential
drop across the Debye layer. In the free-slip model, the electric field lines is assumed to be tangential to the outer edge of the
Debye layer and zero net charge density out side the Debye layer. The slip velocity model may not hold if the Debye length is
finite or the applied field is strong or both. For a highly charged surface, a concentration gradient may develop between the
diffuse layer and the bulk electrolyte. The electric body force which governs the fluid flow and the convective transport of
ions are significant if either of the above conditions hold. In those cases, momentum equations and ion transport equations
are coupled and the distribution of ions are not governed by the equilibrium Boltzmann distribution. It may be noted that the
Boltzmann distribution of ions is based on the assumption of thermodynamic equilibrium where convective transport of ions
and the effect of imposed electric field are neglected.
In this paper, the electroosmotic flow near a polarizable rectangular block mounted on one wall of an otherwise homo-
geneous micro-channel is studied. The present configuration involves a geometric modulation as well as non-uniform EDL
along the obstacle. The non-uniform axial flow rate will generate an induced pressure field. The non-uniform Debye layer
along with the induced pressure gradient may lead to a non-zero charge density outside the Debye layer. This makes the
equations for ion transport and fluid flow coupled. The present model is based on the Nernst-Planck (NP) equations for
ion transport coupled with the Navier–Stokes equations for fluid flow and a Poisson equation for electric field. It may be
noted that all the previous studies on EOF near a conducting obstacle (Eckstein et al., 2009; Sharp et al., 2011; Wu & Li,
2008a, 2008b; Zhao & Yang, 2009) are based on the free-slip model valid for a thin Debye length. The impact of the finite
Debye length on the induced-charge EOF over a polarizable electrode was made by Gregersen, Andersen, Soni, Meinhart,
and Bruus (2009). Zhang et al. (2011) considered the NP model to study the ICEO due to the presence of a conducting elec-
trode. Under the assumption of a small Dukin number and thin double layer, the obstacle and its double layer can be treated
as an insulator. For this condition, a slip-driven electroosmosis along the outer edge of the double layer is established. How-
ever, a no-slip boundary condition is imposed on the homogeneous walls of the channel. The effects of ionic concentration of
the electrolyte on EOF is analyzed. A comparison with the free-slip model is also made in this study.
130 S. Bera, S. Bhattacharyya / International Journal of Engineering Science 94 (2015) 128–138

2. Mathematical model

We consider a long rectangular channel of height h and width W with h  W filled with an incompressible Newtonian
electrolyte of uniform permittivity (e ) and viscosity (l). Permittivity e ¼ 0 r , where 0 is the dielectric constant of the solu-
tion and r is the permittivity of the vacuum. A polarizable obstacle in the form of a rectangular block of lengths l and height
dð< hÞ is mounted on the lower wall of the channel (Fig. 1). We consider the Cartesian coordinate system with x-axis is along
the length, y-axis along the height and z-axis along the width of the channel. External electric field E0 is applied along the
x-axis and is generated by placing the electrodes at the inlet and outlet of the channel. The surface of the obstacle polarizes
under the action of the external field. At steady state, which is achieved with in a fraction of second, the induced double layer
on the obstacle surface repel the local electric field lines in such a way that the obstacle behaves like an insulator and the
electric field lines become tangential to the surface.
The transport equation of the ionic species i is governed by the Nernst–Planck equation as
@ ni
þ r  Ni ¼ 0 ð1Þ
@t
where N i ð¼ Di rni þ ni xi zi FE þ ni u
 Þ is the net flux of ith ionic species. Here, u
 is the velocity field, Di is the diffusivity,
xi ð¼ Di =RTÞ is the mobility of i type species, R is the gas constant, F is the Faraday constant and T is the absolute temperature
of the solution. Here zi and ni are respectively, the valance and number concentration of the i type ion. We considered sym-
metric monovalent ions, thus i ¼ 1, and i ¼ 1 denote respectively, the cations and anions. The dimensional variables are
denoted by an overline. Eq. (1) in non-dimension form can be written as
@ni
Pe  r2 ni þ zi r  ðni EÞ þ Per  ðni uÞ ¼ 0 ð2Þ
@t
 by the Smoluchowski velocity U HS ¼ e E0 /0 =l; n
We scale the velocity field u  i by the bulk number concentration n0 , Cartesian
coordinates by ðh; h; WÞ, time t by h=U HS and electric potential is scaled by the thermal voltage /0 ð¼ kB T=eÞ. The
non-dimensional parameters governing the EOF are the Reynolds number Re ¼ U HS h=m, Schmidt number Sc ¼ m=Di , Peclet
1=2
number Pe ¼ Re:Sc. The parameter j ¼ ½ð2e2 n0 Þ=ðe kB TÞ is reciprocal of the electric double layer(EDL) thickness (kd ) and
the Debye–Huckel parameter is jh. Here kB is the Boltzmann constant, e is the elementary electric charge, m ¼ l=q is the
kinematic viscosity, q is the density of the liquid and l is the viscosity. We denote the non-dimensional concentration of
cation by g and anion by f. For the sake of simplicity, we have taken the diffusivity of cations and anions are equal.
Based on the Gauss law, the electric field E ¼ r/ is determined by solving the following non-dimensional Poisson
equation
2
ðjhÞ
r2 / ¼  qe ð3Þ
2
P
where the net charge density qe ¼ i zi eni is scaled by en0 i.e., qe ¼ qe =en0 .
The equation for transport of electrolyte is governed by the Navier–Stokes equations for a constant property of Newtonian
fluid and can be written in non-dimension form as

ru¼0 ð4Þ
2
@u ðjhÞ
Re þ Reðu  rÞu ¼ rp þ r2 u þ qE ð5Þ
@t 2K e
where uð¼ u; v ; wÞ is the velocity field of the fluid with u; v and w are the velocity components in the x; y and z directions,
respectively. We scale pressure p by lU HS =h. We denote K ¼ E0 h=/0 as the strength of the external field. The width (W) of the

Fig. 1. Schematic diagram of the microchannel with wall mounted conducting obstacle.
S. Bera, S. Bhattacharyya / International Journal of Engineering Science 94 (2015) 128–138 131

channel is considered to be large i.e., h=W  1. Thus, all gradients with respect to z can be neglected and the EOF can be
treated as two-dimensional.
The scaled electric potential at the edge of the Debye layer on the surface of the perfectly conducting obstacle is deter-
mined as
fi ¼ w þ ws ð6Þ
@w
where w satisfies the Laplace equation with @x
¼ K as x ! 1 and rw  n ¼ 0 on the obstacle surface, where n is the unit
outward normal to the surface. The external electric field is considered as 104 V/m, which corresponds to the
non-dimensional parameter K ¼ 4. The second term of r.h.s of Eq. (6) is due to electro-neutrality of the obstacle, which is
obtained as
R
B
wds
ws ¼ ð7Þ
S
Here B is the surface of the obstacle and S is the surface area.
Along the surface of the conducting obstacle (0:5 6 x 6 0:5 and 0 6 y 6 0:2) we imposed the impermeability of ions and
slip velocity condition i.e.,
N i :n ¼ 0; / ¼ fi ; u:t ¼ fi ðx; yÞrw  t ð8Þ
Here n and t are respectively, the normal and tangential direction to the obstacle surface.
Along the non-conducting walls of the channel, we imposed a no-slip condition along with the condition for the imper-
meability of ions. The constant charge density along the non-conducting walls lead to a constant fpotential. Thus,
u ¼ v ¼ 0; N i :n ¼ 0; /¼f ð9Þ
Here n is the outward unit normal to the wall. The surface potential of non-conducting walls are taken to be 256 mV which
corresponds to f ¼ 1. The conditions at the inlet and out let of the computational domain are obtained by solving the 1-D
fully developed electroosmotic flow over a homogeneous microchannel with wall potential f ¼ 1. The governing equations
and computational details for this fully-developed EOF is described in Bera and Bhattacharyya (2013).

3. Numerical methods

We solved the coupled set of governing non-linear equations for fluid flow and ionic species concentration through a
finite volume method on a staggered grid system. In the staggered grid system, the scalar quantities are evaluated at each
cell center and the velocity components are evaluated at the midpoint of the cell sides to which they are normal. The dis-
cretized form of the governing equations is obtained by integrating the governing equations over each control volumes. Dif-
ferent control volumes are used to integrate different equations. We have used the upwind-biased scheme, Quadratic
Upwind Interpolation Convective Kinematics (QUICK) (Leonard, 1979), to discretize the convective and electromigration
terms in both concentration and Navier Stokes equations. The QUICK scheme uses a quadratic interpolation/exterpolation
between three nodal values to estimate the variables at the interface of the control volume. The upwind scheme imparts
stability to the numerical solution in the region where a steep gradient in variables occur. An implicit first-order scheme
is used for discretising the time derivatives. The resulting discretized equations are solved iteratively through the pressure
correction based iterative algorithm SIMPLE Fletcher (1991). The iteration starts by assuming the induced electric potential /
at every cell center.
In this algorithm, the pressure link between the continuity and momentum equations is accomplished by transforming
the continuity equation into the Poisson equation for pressure. The Poisson equation implements a pressure correction for a
divergent velocity field. The pressure Poisson equation is given by
  
p0iþ1;j  2p0i;j þ p0i1;j p0i;jþ1  2p0i;j þ p0i;j1 1 ui;j  ui1;j v i;j  v i;j1
  
þ ¼  ð10Þ
ðDxÞ2 ðDyÞ2 Dt Dx Dy

The variable p0i;j denotes the pressure correction, and ui;j and v i;j denote the velocity components obtained by solving the
momentum equations. At each iteration, the Poisson equation for electric potential is solved through a central difference
scheme. Thus, at any time step, a single iteration in this algorithm follows these sequential steps:

1. Implicit calculations of the momentum and ion transport equations are performed through a quasi-linear approximation.
Due to coupling of equations, we solved the resulting system of linear algebraic equations through a block elimination
method Varga (1962).
2. The Poisson equation for pressure correction is solved using the successive under relaxation method.
3. The velocity field at each cell is updated using the pressure correction.

Iteration at each time step is continued until the divergence-free velocity field is obtained. The divergence in each cell is
towed below a preassigned small quantity ( 105 ). The iteration starts by computing the Poisson Eq. (3) for electric poten-
132 S. Bera, S. Bhattacharyya / International Journal of Engineering Science 94 (2015) 128–138

tial (/). The computer code developed based on the algorithm as described above has been validated by comparing with ana-
lytical, numerical and experimental results. The comparison of our computed results with the results due to other authors is
discussed in Appendix.

4. Results and discussions

We consider the height of the channel h ¼ 10 lm, viscosity l ¼ 0:001 kg/m s, density q ¼ 1000 kg/m3 , diffusion coeffi-
9 2
cient Dþ ¼ D ¼ 2:0 10 m /s and the thermal voltage /0 ¼ 0:0256 mv. The external electric field is assumed to be
104 V/m, thus the non-dimensional parameter K ¼ 4:0 when h ¼ 10 lm. In all the computations presented below the
fpotential of the homogenous part of the microchannel is taken to be 1. The Reynolds number based on the
Helmholtz-Smoluchowski velocity U HS ð¼ 1:788 104 m=sÞ is Re = 1:78 103 when f ¼ 1, Schmidt number, Sc = 500
and the Peclet number Pe = 0:89. The EDL thickness ðkd Þ varies from 0:96 lm to 96:28 nm when the ionic concentration
of the aqueous solution varies from 104 mM to 102 mM. This implies that the Debye-Hückel parameter jh varies between
10.38 and 103.85 when channel height h ¼ 10 lm.
Fig. 2 shows the distribution of the induced fpotential along the surface of the polarizable obstacle. When an electric
field is applied, the current drives positive ions into a thin diffuse layer on one side of the obstacle and negative ions into
the other side. After the polarization of the obstacle, it starts behaving like an insulator and an induced dipolar double layer
is formed. The distribution of fi shows a symmetry pattern around the vertical line y ¼ 0, this implies a zero net charge den-
sity on the block. The variation of fi on the top face BC shows a linear dependence with x-coordinate.
We first illustrate the electroosmotic flow within the channel when the obstacle is considered to be insulated and having
constant f-potential which is taken to be the same as the f-potential of the homogeneous part of the channel walls. Figs. 3(a)
and (b) show the streamlines and distribution of the scaled net charged density near an insulated obstacle when the ionic
concentration is 103 mM for a channel of height 10 lm i.e., jh ¼ 32:84. In the presence of an imposed electric field, the sur-
plus counter ions in the diffused double layer will move towards the electrolyte. The motion of ions drag the advancement

4
3 B C

2 A D

1
ζi

0
-1
-2
-3
-4
A B C D

Fig. 2. Distribution of induced scaled zeta-potential ðfi Þ along the surface of the polarizable obstacle when jh ¼ 32:84; h ¼ 10 lm and E0 ¼ 104 V/m (i.e.,
K ¼ 4).

0.8

0.6
y

0.4

0.2

0
-2 0 2
x

Fig. 3. (a) Streamline patterns and (b) scaled net charge density ðg  f Þ in the vicinity of the non-polarizable obstacle when h ¼ 10 lm, scaled surface
potential f ¼ 1; E0 ¼ 104 V/m (i.e., K ¼ 4) and kd ¼ 0:30 lm (jh = 32.8).
S. Bera, S. Bhattacharyya / International Journal of Engineering Science 94 (2015) 128–138 133

liquids and an electroosmotic flow sets-in. The streamline patterns near the obstacle show that the flow is two-dimensional
but without the appearance of any vortex. An induced pressure field develops as the fluid flow rate becomes non-uniform
due to the presence of the obstacle. However, the flow field becomes uni-directional as we move away from the obstacle.
The distribution of ions show that the charge density is non-zero near the obstacle. A non-zero charge density in the bulk
region implies that the electric body force outside the Debye layer has an impact in driving the fluid motion. The linear the-
ory under the thin Debye layer limit assumes the equilibrium Boltzmann distribution of ions, where convective transport
and electromigration of ions due to an imposed field are neglected. Though the effects of convection on ion transport are
small, the effect due to external field is significant. The advection, diffusion and electromigration of ions generate an induced
electric field. The induced electric field is mostly along the vertical direction over the homogeneous part of the channel. The
equation for induced electric field i.e., Eq. (3) suggests that a small value of net charge density ðg  f Þ can lead to a significant
2
change in potential as the factor ðjhÞ =2 is as great as 2157. For a thin Debye layer ðjh
1Þ, the net charged density outside
the Debye layer is almost zero and ions are convected parallel to the Debye layer under the action of imposed electric field E0 .
This causes a negligible charge redistribution and the fluid transport equation is decoupled from the equation of ion trans-
port. Our results suggest that the linear theory used to analyze the EOF near the obstacle becomes questionable. In this
region, the electric body force on fluid outside the EDL is non-zero, which makes the governing equations for fluid flow
and ion transport coupled. Unlike the typical pressure-driven flow, here there is no flow separation from the flat surface near
the upstream or downstream corners of the obstacle. The strong electric body force near the wall compensate the induced
adverse pressure gradient which occurs near the wall roughness and hence no flow separation takes place in the present
case.
The streamline patterns near the polarizable obstacle are presented in Fig. 4(a)–(c). The results in Fig. 4(a) and (b) are
based on the present model with Debye layer thickness jh = 32.84 and 73.4, respectively. The corresponding result based
on the free-slip model is presented in Fig. 4(c). In the free-slip model the bulk flow is assumed to be electrically neutral
and the slip velocity along the channel walls are obtained by the Smoluchowski velocity U HS . It may be noted that the
f-potential along the obstacle is non-uniform and is governed by the Eq. (6), as described before. The streamlines of the liq-
uid flow near the obstacle surface are distorted and a micro-vortex is generated because of the non-uniform induced charge
distribution on the conducting obstacle. By comparing Fig. 4(a) and (b) we find that the size of the vortex expands as the
ionic concentration of the electrolyte is raised. At high value of jh i.e., jh ¼ 73:4 (Fig. 4b) the thickness of the Debye layer
is much smaller than the channel height and the effect of the Debye layer on the bulk flow is also not significant. Because of
this, we find from Fig. 4(b) and (c) that the form of the vortex and streamline patterns for the present model are somewhat
similar with the free-slip model. We find that the EOF away from the obstacle is close to the Smoluchowski velocity U HS .
The dependence of the vortex size, which is developed over the obstacle, on the ionic strength is illustrated in Fig. 5(a).
We consider the flow in a channel of height 10 lm and external electric field as 104 V/m at different EDL thickness. The
fpotential on the non-conducting walls is 256 mV. The size of the vortex expands as the ionic concentration of the elec-
trolyte is increased (i.e., increase of jh) and approaches a constant value at large values of jh. Thus, the formation of vortical
flow can be regulated by changing the ionic concentration of the electrolyte. At large values of jh, the result based on the
present model approaches the free-slip model.
Fig. 5(b) presents the variation of the circulation strength (C) with the ionic concentration for a fixed value of the external
electric field and channel height. The circulation strength C is obtained as
ZZ
C¼ xdxdy ð11Þ
S0

1 1 1

0.8 0.8 0.8

0.6 0.6 0.6


y

0.4 0.4 0.4

0.2 0.2 0.2

0 0 0
-2 0 2 -2 0 2 -2 0 2
x x x

Fig. 4. Streamline patterns in the vicinity of the conducting obstacle when h ¼ 10 lm, f ¼ 1 and E0 ¼ 104 V/m. (a) Present model with jh = 32.84; (b)
Present model with jh = 73.4 and (c) Slip model.
134 S. Bera, S. Bhattacharyya / International Journal of Engineering Science 94 (2015) 128–138

0.325 0.41

0.4
0.32 0.39


hp 0.38
Present model
0.315 Present model Slip model
Slip model 0.37

0.36
0.31
0.35
20 40 60 80 100 20 40 60 80 100
κh κh

Fig. 5. Variation of the center of the vortex and circulation strength with jh when h ¼ 10 lm, f ¼ 1 and E0 ¼ 104 V/m. (a) Height of the vortex center (hp )
from the lower wall; (b) Circulation strength ðCÞ. The results corresponding to slip model is also indicated.

where
 
@ v @u
x¼  ð12Þ
@x @y

Here S0 is the region enclosed by a circuit drawn above the obstacle. The circulation strength gradually increases with the
increase of ionic concentration and approaches a constant value at large values of the ionic concentration i.e., large jh at
a fixed h. The dependence of the vortical flow near the obstacle on the ionic concentration for low to moderate values of
jh is evident from these results i.e., Fig. 5(a) and (b). A slip velocity assumption, which is a common practice in microchan-
nel, may not be correct when Debye layer is not thin. The charge density outside the EDL is not zero for low to moderate
values of jh, which makes the fluid flow equations coupled with the equations for ion transport. For a thin EDL, the electric
force is zero outside the EDL and the bulk flow is governed by the induced pressure gradient and viscous dissipation.
In Fig. 6(a) and (b) we present the axial velocity profile and ion distribution far upstream of the obstacle i.e., at x ¼ 2:5
for different values of the EDL thickness (kd ¼ 1=j) when E0 ¼ 104 V/m and h ¼ 10 lm. Along the flat surface, the Debye layer
is homogeneous and the flow is primarily along the direction of the electric field. The u-velocity profile is plug-like and inde-
pendent of the variation of Debye layer thickness for large jh. Outside the EDL on the plane surface, the flow is determined
by the Smoluchowski velocity U HS . However, at jh ¼ 10:3, the EDL is relatively thick and the velocity profile assumes a para-
bolic shape even at x ¼ 2:5. For thick EDL case, the electric body force in this region influences the flow, which makes the
Navier–Stokes equations and the Nernst–Planck equations for ion transport coupled. The velocity distribution over the flat
surface ðx ¼ 2:5Þ is in agreement with the analytic solution based on the Debye-Hu € ckel approximation i.e.,
 
sinhfð1  yÞjhg þ sinhfyðjhÞg
u ¼ f 1  ð13Þ
sinhðjhÞ
Here, f is the constant zeta potential of the homogeneous walls. The ion distribution shows that the bulk region has non-zero
charge density when the Debye layer is thick. At large value of jh (thin EDL), the ion distribution resembles the

3
1
2.5
0.8 κh
g
2 f
0.6
κh
f,g
u

1.5
0.4
1

0.2 0.5

0 0
0 0.2 0.4 0.6 0.8 1 0 0.2 0.4 0.6 0.8 1
y y

Fig. 6. (a) Streamwise velocity profile and ion distribution far upstream of the conducting block i.e., x ¼ 2:5 when h ¼ 10 lm, f ¼ 1; E0 ¼ 104 V/m. Arrow
indicates the increasing direction of jh (=10.3, 23.2, 32.8). Symbols indicate the analytic solution obtained by Eq. (13). (a) u velocity; (b) g, f.
S. Bera, S. Bhattacharyya / International Journal of Engineering Science 94 (2015) 128–138 135

Guy-Chapman type of distribution and a plug-like flow profile is obtained. Outside the thin Debye layer, the fluid is electri-
cally neutral i.e., ðg  f Þ ¼ 0. The ion distribution within the Debye layer is governed by the balance of electromigration and
diffusion. However, for small values of jh (thick EDL), the bulk fluid is not electrically neutral and hence, the electric body
force outside the EDL is not negligible.
In Fig. 7(a) and (b) we provide the velocity profiles at different jh (= 10.3, 32.8, 73.4) in the region above the obstacle i.e.,
at x ¼ 0. This result clearly suggests that the flow field is two dimensional above the block. We have also included the result
when the block is considered to be insulated with constant fpotential which is the same as that of the homogeneous parts
of the wall i.e., f ¼ 1. We find that the stream-wise velocity is low when block is considered to be conducting. A strong
transverse velocity component is found for the present case compared to the case when the block is insulated.
As we move towards the obstacle, the fluid encounters an adverse pressure gradient due to the geometric modification of
the channel wall along with the effect of non-homogenious EDL. The induced f-potential on the upstream half of the obstacle
is non-uniform but of the same sign as that of the plane surface and the non-uniform f-potential is of opposite sign along the
downstream half of the obstacle. The non-uniformity of the streamwise velocity across the cross-section in the region near
the obstacle results in a two-dimensional flow. In Fig. 8(a) and (b) we present the velocity vectors near the obstacle for thick
(jh ¼ 10:3) and thin EDL (jh ¼ 73:4) cases. Along the downstream half, the fpotential of the obstacle is in opposite sign
and a flow reversal close to the upper face of the obstacle is evident. The electric body force, which is proportional to
g  f , drives the fluid flow near the surface and it then becomes strong with the increase of the ionic concentration of elec-
trolyte i.e., increase of jh from 10:3 to 73:4. This leads to an expansion of the induced vortex with the rise of jh.
The average flow rate (U m ) is determined as
ZZ
U HS
Um ¼ uðx; yÞdxdy ð14Þ
A A

2 0.5

0.4
1.5

0.3
v
u

1
κh 0.2
κh
0.5 0.1

0
0
0.2 0.4 0.6 0.8 1 0.2 0.4 0.6 0.8 1
y y

Fig. 7. Velocity profiles above the conducting block i.e., x ¼ 0 for different values of jh when h ¼ 10 lm, f ¼ 1; E0 ¼ 104 V/m. Here, jh varies between 10.3
to 73.4. Arrow indicates the increasing direction of jh. Dashed lines correspond to non-conducting block. (a) Streamwise velocity (u); (b) Transverse
velocity (v).

Fig. 8. Velocity vectors in the vicinity of the conducting obstacle for (a) jh ¼ 10:38; (b) jh ¼ 73:4when h ¼ 10 lm, f ¼ 1 and E0 ¼ 104 V/m.
136 S. Bera, S. Bhattacharyya / International Journal of Engineering Science 94 (2015) 128–138

where A is the area of the undertaken computational domain. We determine the dimensional average flow rate (U m ) for dif-
ferent values of jh when E0 ¼ 104 V/m and the f-potential for the homogenous part of the channel is 1:0 (Fig. 9). Our
results shows that when jh is large i.e., Debye layer is thin compared to the channel height, the average rate of flow
approaches a constant value, the value determined by the free-slip model.
The electric current density (j) is defined as
X
j¼F zi Ni ð15Þ
i

where Ni is the net ionic flux of ith ionic species. The average current density (im ) is
ZZ
1 jx
im ¼ dxdy ð16Þ
A j
A 0

where jx is the x-component of the vector j and j0 ð¼ FDi n0 =hÞ is the characteristic electric current density. The total electro-
static body force ðF m Þ is obtained by integrating over the entire channel as
ZZ
Fm ¼ qe Ex dxdy ð17Þ
A

Here, F m is scaled by F 0 ð¼ Fn0 /0 =hÞ. We find from Fig. 10(a) and (b) that the average non-dimensional current density and
net electric body force approach a constant value with the increase of concentration of the ionic species.

200

150
U m(μm/sec)

Present model
100
Non-conducting block
Slip model

50
20 40 60 80 100
κh
Fig. 9. Variation of average flow (U m ) in dimensional form due to the variation of jh when E0 ¼ 104 V/m. Here, channel height h ¼ 10 lm and zeta potential
of homogeneous part of the wall f ¼ 1.

3 2.5

2.5
2
Presentmodel
2
Slipmodel 1.5
Fm/F0
im

1.5
1
1

0.5 0.5

20 40 60 80 100 0
20 40 60 80 100
κh κh

Fig. 10. Variation of the non-dimensional average current intensity (im ) and net electric body force (F m =F 0 ) with jh when h ¼ 10 lm, f ¼ 1 and
E0 ¼ 104 V/m. The solution for im based on slip model is indicated by a dashed line in (a).
S. Bera, S. Bhattacharyya / International Journal of Engineering Science 94 (2015) 128–138 137

0.8

y 0.6

0.4

0.2

0
-1 0 1
x
Fig. 11. Streamline patterns near the conducting obstacle for low f ¼ 0:084 when jh ¼ 73:4; h ¼ 10 lm and E0 ¼ 104 V/m.

5. Conclusions

The electroosmotic flow near a polarizable rectangular block mounted on one wall of a wide microchannel is studied. The
homogeneous walls of the channel have a constant fpotential. The surface of the block polarizes and a Debye layer of
non-uniform fpotential develops and the EOF in its vicinity does not resemble the classical plug-like form. A recirculation
vortex develops over the block and the strength of which depends on the ionic concentration of the electrolyte when the
Debye layer is considered finite. Our computed result matches with the result due to a free-slip model in the limit of the
thin EDL. The free-slip model overpredicts the EOF near the block when EDL is finite. For finite values of Debye layer thick-
ness, a non-zero charge density outside the EDL occurs, which makes the ion transport equation and equation for fluid flow
coupled in the bulk region of the channel.

Acknowledgements

Authors are grateful to the Department of Science & Technology, Govt. India for providing the financial support through a
research project grant. S. Bera acknowledges the post-doctoral fellowship received from NBHM, DAE.

Appendix A. Validation of numerical scheme

We have already shown in Fig. 6(a) that our computed solutions far upstream and downstream of the obstacle compares
well with the analytic solution for the fully develop EOF as given by Eq. (13). In Fig. 11, we have presented the streamlines
near the conducting obstacle when the surface potential of the homogeneous parts of the wall is low i.e., f ¼ 0:08. The
intention of this result is to make a qualitative comparison of our numerical scheme with previous experimental studies.
The experimental as well as the numerical study based on a free-slip model of Wu and Li (2008b) for induced-charge elec-
troosmotic flow in a rectangular channel with embedded conducting obstacles shows the presence of micro-vortices. They
have considered the electroosmotic mobility of non-conducting channel wall as 5:9 104 cm2 =v s which corresponds to
f ¼ 0:08 when ee ¼ 695:4074 1012 . The streamline patterns based on our computed results for the configuration involv-
ing a single rectangular block mounted on the lower wall of the channel is qualitatively similar to the result presented by Wu
and Li (2008b). The occurrence of vortices near the conducting surface has also been demonstrated in several experimental
studies e.g., Canpolat, Zhang, Rosen, Qian, and Beskok (2013), Daghighi et al. (2013). It may be noted that the streamline pat-
terns in Fig. 11 are different from the patterns as presented in Fig. 4. In Fig. 4, the f-potential along the homogeneous walls is
taken to be 1. In that case (Fig. 4), a relatively stronger electroosmotic flow prevents the formation of the vortex on the
upstream side of the obstacle and the size of the vortex appearing on the downstream side is reduced compared to Fig. 11.

References

Bazant, M. Z., & Squires, T. M. (2010). Induced-charge electrokinetic phenomena. Current Opinion in Colloid and Interface Science, 15, 203–213.
Bera, S., & Bhattacharyya, S. (2013). On mixed electroosmotic-pressure driven flow and mass transport in microchannels. International Journal of Engineering
Science, 62, 165–176.
Bhattacharyya, S., & Nayak, A. K. (2009). Electroosmotic flow in micro/nanochannels with surface potential heterogeneity: An analysis through the Nernst-
Planck model with convection effect. Colloids and Surfaces A: Physicochemical and Engineering Aspects, 339, 167–177.
Canpolat, C., Zhang, M., Rosen, W., Qian, S., & Beskok, A. (2013). Induced-charge electroosmosis around touching metal rods. Journal of Fluids Engineering-T
ASME., 135, 021103–0211010.
Chen, L., & Conlisk, A. T. (2009). Effect of nonuniform surface potential on electroosmotic flow at large applied electric field strength. Biomedical Microdevices,
11, 251–258.
Daghighi, Y., Sinn, I., Kopelman, R., & Li, D. (2013). Experimental validation of induced-charge electrokinetic motion of electrically conducting particles.
Electrochimica Acta, 87, 270–276.
Eckstein, Y., Yossifon, G., Seifert, A., & Miloh, T. (2009). Nonlinear electrokinetic phenomena around nearly insulated sharp tips in microflows. Journal of
Colloid and Interfaces Science, 338, 243–249.
Erickson, D., & Li, D. (2002). Influence of surface heterogeneity on electrokinetically driven microfluidic mixing. Langmuir, 18, 1883–1892.
138 S. Bera, S. Bhattacharyya / International Journal of Engineering Science 94 (2015) 128–138

Fletcher, C. A. J. (1991). Computational techniques for fluid dynamics. Springer series computational phyics (2nd ed., Vols. I and II). Berlin. Heidelberg, New York:
Springer.
Fu, L.-M., Lin, J.-Y., & Yang, R.-J. (2003). Analysis of electroosmotic flow with step change in zeta potential. Journal of Colloid and Interfaces Science, 258,
266–275.
Ghosal, S. (2002). Lubrication theory for electro-osmotic flow in a microfluidic channel of slowly varying cross-section and wall charge. Journal of Fluid
Mechanics, 459, 103–128.
Gregersen, M. M., Andersen, M. B., Soni, G., Meinhart, C., & Bruus, H. (2009). Numerical analysis of finite Debye-length effects in induced-charge electro-
osmosis. Physical Review E, 79, 066316.
Khair, A., & Squires, T. M. (2008). Surprising consequences of ion conservation in electro-osmosis over a surface charge discontinuity. Journal Fluid
Mechanics, 615, 323–334.
Leonard, B. P. (1979). A stable and accurate convective modelling procedure based on quadratic upstream interpolation. Computer Methods in Applied
Mechanics and Engineering,19, 59–98.
Lin, T.-Y., & Chen, C.-L. (2013). Analysis of electroosmotic flow with periodic electric and pressure fields via the lattice Poisson-Boltzmann method. Applied
Mathematical Modeling, 37, 2816–2829.
Sharp, K. V., Yazdi, S. H., & Davison, S. M. (2011). Localized flow control in microchannels using induced-charge electroosmosis near conductive obstacles.
Microfluidics and Nanofluidics, 10, 1257–1267.
Squires, T. M., & Bazant, M. Z. (2004). Induced-charge electro-osmosis. Journal of Fluid Mechanics, 509, 217–252.
Squires, T. M., & Bazant, M. Z. (2006). Breaking symmetries in induced-charge electro-osmosis and electrophoresis. Journal of Fluid Mechanics, 560, 65–101.
Varga, R. S. (1962). Matrix iterative numerical analysis. New York: Wiley.
Wu, Z., & Li, D. (2008a). Mixing and flow regulating by induced-charge electrokinetic flow in a microchannel with a pair of conducting triangle hurdles.
Microfluidics and Nanofluidics, 5, 65–76.
Wu, Z., & Li, D. (2008b). Micromixing using induced-charge electrokinetic flow. Electrochimica Acta, 53, 5827–5835.
Yariv, E. (2004). Electro-osmotic flow near a surface charge discontinuity. Journal of Fluid Mechanics, 521, 181–189.
Zhang, M., Ai, Ye., Sharma, A., Joo, S. W., Kim, D.-S., & Qian, S. (2011). Electrokinetic particle translocation through a nanopore containing a floating electrode.
Electrophoresis, 32, 1864–1874.
Zhao, C., & Yang, C. (2009). Analysis of induced-charge electro-osmotic flow in a microchannel embedded with polarizable dielectric blocks. Physical Review
E, 80, 046312.

You might also like