a

r
X
i
v
:
0
7
1
2
.
0
4
5
3
v
1


[
c
o
n
d
-
m
a
t
.
m
t
r
l
-
s
c
i
]


4

D
e
c

2
0
0
7
Induced-charge electrophoresis near an insulating wall
Mustafa Sabri Kilic
1
and Martin Z. Bazant
1,2
1
Department of Mathematics, Massachusetts Institute of Technology, Cambridge, MA 02139,USA and
2
Physico-Chimie Theorique, Gulliver-CNRS, ESPCI, 10 rue Vauquelin, Paris 75005, France
(Dated: August 4, 2008)
Induced-charge electrophoresis (ICEP) has mostly been analyzed for asymmetric particles in an
innite uid, but channel walls in real systems further break symmetry and lead to dielectrophoresis
(DEP) in local eld gradients. Zhao and Bau (Langmuir, 23, 2007, pp 4053) recently predicted that
a metal (ideally polarizable) cylinder is repelled from an insulating wall in a DC eld. We revisit this
problem with an AC eld and show that attraction to the wall sets in at high frequency and leads to
an equilibrium distance, where DEP balances ICEP, although, in three dimensions, a metal sphere
is repelled from the wall at all frequencies. This conclusion, however, does not apply to asymmetric
particles. Consistent with the recent experiments of Gangwal et al. (arXiv:0708.2417), we show that
a metal/insulator Janus particle is always attracted to the wall in an AC eld. The Janus particle
tends to move toward its insulating end, perpendicular to the eld, but ICEP torque rotates this end
toward the wall. Under some conditions, the theory predicts steady translation along the wall with
an equilibrium tilt angle, as seen in experiments, although more detailed modeling of the contact
region of overlapping double layers is required.
DRAFT
I. INTRODUCTION
Most theoretical work on electrophoresis has focused on
spherical particles moving in an innite uid in response
to a uniform applied electric eld [2, 17, 27, 35]. Of
course, experiments always involve nite geometries, and
in some cases walls play a crucial role in electrophore-
sis. The linear electrophoretic motion of symmetric (spher-
ical or cylindrical) particles near insulating or dielectric
walls [8, 12, 18, 19, 20, 28] and in bounded cavities or chan-
nels [6, 7, 13, 14, 15, 21, 22, 24, 37, 48] has been analyzed ex-
tensively. Depending on the geometry and the double-layer
thickness, walls can either reduce or enhance the transla-
tional velocity, and the rotational velocity can be opposite
to the rolling typical of sedimention near a wall. The classi-
cal analysis for thin double layers assumes force-free mo-
tion driven by electro-osmotic slip alone, but recent work has
shown that electrostatic forces can also be important near
walls [16, 45]. Heterogeneous particles with non-uniform
shape and/or zeta potential exhibit more complicated bulk
motion [1, 9, 25, 26], which can also aect boundary inter-
actions [42], especially if the particles are deformable, as in
the case of chain-like biological molecules [31].
In this article, we focus on the eect of nonlinear induced-
charge electro-osmotic (ICEO) ows at polarizable surfaces,
which are nding many new applications in microuidics and
colloids [4, 39, 41]. The canonical example of quadrupolar
ICEO ow around a polarizable particle, rst described by
Murtsovkin [10, 30], involves uid drawn in along eld axis
and expelled radially in the equatorial plane in an AC or
DC eld. Broken symmetries in this problem can gener-
ally lead to hydrodynamic forces and motion induced-charge
electrophoresis (ICEP), as well as electrical forces and mo-
(a) Insulating wall
E
u u
E
U
U
(b) Conducting wall
FIG. 1: Hydrodynamic forces on polarizable particles near (a)
insulating and (b) unscreened conducting walls due to ICEO ows
tion by dielectrophoresis (DEP). Such phenomena have only
been analyzed for isolated asymmetric particles in an in-
nite uid [4, 40, 44] or in a dilute solution far from the
walls [34, 36]. In contrast, experiments demonstrating trans-
lational ICEP motion have involved strong interactions with
walls [11, 29], which remain to be explained.
As shown in Figure 1, it is easy to see that the quadrupo-
lar ICEO ow around a polarizable particle typically causes
attraction to unscreened conducting walls (perpendicular to
the eld) and repulsion from insulating walls (parallel to the
eld). The former eect of ICEP attraction to conduct-
ing walls has not yet been analyzed; it may play a role in
colloidal self assembly on electrodes applying AC voltages
[32, 33, 38, 43, 46]. This phenomenon is mainly understood
in terms of electrohydrodynamic ows (what we would term
ICEO) induced on the electrodes, not the particles (typi-
cally latex spheres), but ICEP could be important for more
polarizable particles.
The latter eect of ICEP repulsion from insulating walls
has recently been analyzed by Zhao and Bau [47] in the case
2
of a two-dimensional ideally polarizable cylinder in a DC
eld. However, this phenomenon has not yet been conrmed
experimentally. On the contrary, Gangwal et al [11] have
recently observed that metallo-dielectric Janus particles are
attracted to a glass wall, while undergoing ICEP motion par-
allel to the wall and perpendicular to an applied AC eld.
It is not clear that the existing theory of ICEP can explain
this surprising behavior.
The objective of this work is to analyze the motion of
three-dimensional polarizable particles near insulating walls
in AC elds. As summarized in section II, we employ the
standard low-voltage model in the thin double-layer approx-
imation, following many authors [3, 23, 39, 40, 44], including
Zhao and Bau [47]. In section III, we rst analyze ideally po-
larizable cylinders and spheres near a non-polarizable wall,
which only experience forces perpendicular to the wall. In
section IV we then study spherical metal/insulator Janus
particles, which are half ideally polarizable and half non-
polarizable. Due to their broken symmetry, the Janus parti-
cles also experience ICEP and DEP torques, which strongly
aect their dynamics near the wall.
II. MATHEMATICAL MODEL
A. Low Voltage Theory
In this paper, we will consider either a cylindrical or a
spherical particle of radius a in a semi-innite electrolyte
bounded by a plane. The distance between the center of the
particle and the plane is denoted by h. In the absence an ap-
plied electric eld, we assume that the particle and the wall
surfaces are uncharged. In addition, we will assume the elec-
trolyte has a low Reynolds number, and impose Stokes equa-
tions. We will assume that the thin double layer approxima-
tion holds and the bulk electrolyte remains electroneutral,
which is the case when the Debye length

D
=
_
kT
2z
2
e
2
c
0
is much smaller than the characteristic length scale (in our
case, a). The Debye length is typically ranges between 1
100nm, and colloidal particles are usually in the range,
therefore thin double layer approximation holds for most of
the time.
Then the general equations consist of the Laplaces

2
= 0
and Stokes equations

2
u = p
u = 0 (1)
where is the electrostatic potential and the permittivity,
the viscosity of the electrolyte, u the velocity eld and
p the pressure. The wall boundary z = 0 is an insulator,
satisfying
n = 0
whereas the particle surface, being polarizable, acts as a ca-
pacitor in the thin double layer limit
dq
dt
= (n) ()
with q being the surface charge density on the particle,
the conductivity of the bulk electrolyte. Far away from the
particle, an electric eld
E

= E

x, |x|
is applied. In general, the magnitude E

= E

(t) may be
time dependent.
The Stokes equations are supplied by the no-slip condi-
tions on the wall and the Smoluchowskis electrokinetic slip
formula on the particle
u = u
slip
=

where is the potential dierence between the surface and

the bulk. Later, we will take into account the steric eects of
ions and show how this formula needs to be modied. Lastly,
we assume ow vanishes at the innities.
So far, the equations are complete except for a constitutive
relation between and q. The linear theory asserts that
q =

although the complete problem is still nonlinear in this case

because of the quadratic slip formula (5). In this paper, we
will study only the linear theory, but the calculations can be
repeated with (sometimes more accurate) nonlinear theories
like that of Poisson-Boltzmann.
B. Force and Torque on the Particle
Throughout this paper, we will assume that the particle is
xed and calculate the forces on the particle. For the case of
a moving particle, the slip velocity needs to be modied to
account for the motion of the particle surface.
The total force and torque on any volume of the uid are
conveniently given in terms of the stress tensor, , by
F =
_

n dA (2)
T =
_

r (n )dA (3)
The stress tensor contains from electrical and viscous stresses
on the uid, =
M
+
H
, where

M
= [EE
1
2
E
2
I]

H
= pI +
_
u + (u)
T
_
3
are the Maxwell and hydrodynamic stress tensors, respec-
tively.
C. Particle Dynamics
In order to calculate the movement of a colloidal particle,
we need to nd a translational velocity U, and a rotational
velocity such that the net force on the particle is zero,
when the slip velocity is modied by taking into account the
velocities U and . In other words, we are seeking U and
such that the problem (1) with boundary condition
u = u
slip
+U+r
yields F = 0 and T = 0.
Since Stokes problem is linear, there is a linear relation-
ship between the translational and rotational motion of the
particle and the resulting force and torque exerted on it by
the uid. Let us denote this relationship by
_
F
T
_
= M
_
U

_
The viscous hydrodynamic tensor M comes from solving for
the Stokes ow around a particle moving with translational
velocity U and and rotational velocity , assuming no slip
on all particle and wall surfaces.
If we then solve the electrokinetic problem for a xed par-
ticle in the applied eld, we obtain the ICEO slip velocity
u
slip
as well as the total (hydrodynamic + electrostatic) force
F
slip
and torque T
slip
needed to hold the particle in place,
thereby preventing ICEP and DEP motion.
Using these calculations and invoking linearity, the condi-
tion of zero total force and torque on the particle,
_
F
T
_
+
_
F
slip
T
slip
_
= 0, determines the motion of the particle
_
U

_
= M
1
_
F
slip
T
slip
_
(4)
The particle trajectory is then described by the solution to
the dierential equation
dx
dt
= U
together with the equations for the particles angular orien-
tation.
This angular orientation does not matter for a full polar-
izable or insulating particle. For the Janus particle, we will
argue that only the rotations about xaxis are important,
thus we will focus on the dynamics of just a single angle. In
this case, the equation of motion is simply
d
dt
=
x
D. Nondimensional Equations
We nondimensionalize the variables by
x

=
x
a
,

=

E

a
,

=

E

a
q

=
E

D
q
t

=
_

D
a
D
_
1
t, t

=
_

E
2

_
1
t
u

=u
_
E
2

_
1
, p

=
p
E
2

Note that there are two time scales in the problem,

=
Da
D
,
the charging time, and

=

E
2

, the time scale for particle

motion.
Plugging in the equations, we obtain (after dropping the
primes except for t)

2
= 0

2
u = p
u = 0
with the boundary conditions
dq
dt

= n
u =
s
(5)
on the particle surface, where =
surface

bulk
, is the zeta
potential. For a polarizable particle, we have
surface
= 0
by symmetry, therefore we are left with =
bulk
= .
In addition, we have
x, |x|
just as before, now with the dimensionless variables. The
remaining equations are the no-slip boundary condition on
the planar wall, and the requirement that the ow vanishes
at innities.
The constitutive relation between q and takes the simple
form
q = =
for the linear theory that we are going to analyze in this
paper.
The dimensionless force and torque on the particle are
given by the formulae (2) and (3), where the stress tensors
are replaced by their dimensionless counterparts

M
= EE
1
2
E
2
I

H
= pI +
_
u + (u)
T
_
4
The force, angular momentum and stress tensors are scaled
to
F
ref
= E
2

a, T
ref
= E
2

a
2
,
ref
= E
2

Finally, the particle motion will be governed by

dx
dt

= U,
d
dt

=
x
E. Simplications
1. Steady Problems
If a DC voltage is applied, then the system reaches a steady
state after a while and the time derivatives drop out. This is
the case when Neumann boundary conditions are valid also
on the cylinder or sphere. In that case
F
E
=
_

(EE
1
2
E
2
I)ndA
=
_

(EE n
1
2
E
2
n)dA =
1
2
_

E
2
ndA
because E n =0 on the surface. As a consequence, the elec-
trostatic torque induced on the particle is zero.
2. Symmetry
For the full cylinder problem, the electrostatic problem has
an odd symmetry in xdirection, that is
(x, z) = (x, z)
and for the full sphere problem, it has an odd symmetry in
xdirection and an even symmetry in y-direction:
(x, y, z) = (x, y, z) = (x, y, z)
As a result, E
2
has even symmetry in x and y. Therefore,
in the steady case, the electrostatic forces vanish in those
directions, and there can only be a vertical force. In general
time-dependent cases, E has an even symmetry in x and and
odd symmetry in y, and therefore the electrostatic force may
not vanish in x direction.
As for the Stokes problem, a glance at the slip formula
shows that the slip, just like the potential eld, has an odd
symmetry in both x and y, and therefore so does the ow
eld. Consequently, there are no hydrodynamic forces in
those directions.
Needless to say, all these symmetry arguments disappear
for the Janus particles.
3. AC elds
In the linear model we are considering, the time-periodic
electrostatic forcing problem can be solved by letting
= Re(

e
it
)
and solving for the complex potential

using the equations

= 0
with the boundary conditions
n

= i

(polarizable)
n

= 0 (insulator)

= E

x (at innity)
In the high frequency limit, the electrostatic problem ap-
proaches to the solution of the dirichlet problem, that is, the
rst boundary condition is replaced by

= 0 (polarizable) (6)
This is because

= n

/i 0. Physically, this means

that the double layers do not have enough time to charge
when the forcing frequency is too high.
Note that in cases of AC elds below, it is the eld ampli-
tude E
0
that enters into the scalings above, and the U is the
time-averaged velocity.
Once the electrostatic potential is calculated, the time-
averaged slip velocity can be obtained by the formula
u
s
=
1
2
Re[

//
] (7)
where

is the (complex) surface zeta potential, which is
equal to

in the linear theory, and

E

//
is the complex
conjugate of the tangential component of

E =

, the com-
plex electric eld.
In the DC limit as 0, the imaginary parts of the
solutions go to zero, and we are left with u
s
=
1
2
E
//
, which
is the standard Smoluchowskis formula with a factor of 1/2.
F. Numerical Methods
We have solved the equations using the nite element soft-
ware COMSOL for various geometries which have been ap-
proximated by the cylindrical or spherical colloidal particle
being enclosed by a nite rectangle. The equations are rst
converted to their weak forms, and entered into the general
weak PDE module of COMSOL. None of the COMSOLs spe-
cial modules are used.
5
FIG. 2: Streamlines to the Stokes ow problem for a cylinder
near a wall. The surface color indicates the pressure. In this case,
the increased pressure between the wall and the cylinder leads to
repulsion away from the wall.
For linear and nonlinear models alike, the computational
eciency is improved by rst solving the electrostatic prob-
lem, and then the hydrodynamic problem. In time dependent
cases, the uid slip can be averaged and the Stokes problem
is solved only once using this averaged slip.
For reasons of completeness and easy reference, here we
list the weak forms of the equations solved. This system
of equations are converted to weak form by multiplying by
corresponding test functions and integrating over the spatial
domain.
The electrical problem turns into
0 =
_

2
dr =
_

dr+
_

(n )ds
=
_

dr+
_

t
qdr
which is satised for all test functions

. The boundary con-
dition for is imposed in the form
0 =
_
q (V
col
) dr
to be satised for all test functions q.
The weak form for the stokes ow is similarly obtained as
0 =
_

[ u ( )dr+ p u]dr
=
_

[ u : p u]dr+
_

u (n ) ds
Since we do not have a simple expression for n , it is best to
introduce the new variable (Lagrangian multiplier) f = n .
This is also convenient for calculation of hydrodynamic forces
at the surface. Then we obtain
0 =
_

[ u : p u]dr+
_

[ u f +

f (u u
s
)]ds
III. ISOTROPIC PARTICLES NEAR A WALL
A. Cylinder in a DC Field
For isotropic particles near a wall, by symmetry,

cylinder
= 0, therefore = . Moreover, there is no net
horizontal force exerted on the particle, so the only forces of
interest is in the vertical direction. Another consequence of
symmetry is the absence of net torque on the cylinder.
The DC cylinder problem has been solved analytically by
Zhao and Bau [47] in the linear case in bipolar coordinates.
The mapping between the bipolar and the Cartesian coordi-
nates is given by
x =
c sin
cosh cos
, y =
c sinh
cosh cos
where
0
< < , and < < denes the region
outside the cylinder. The geometric constants
0
and c are
dened as

0
= sech
1
(a/h)
c =
h
coth
0
(note that there is a typo in the expression for
0
in [47]).
The hydrodynamic and electrostatic forces on the cylinder
are calculated to be
F
H
=
2 sinh
0
E
2

c
(
0
cosh
0
sinh
0
) coth
0
{
1
2 sinh
2

0
+

n=1
_
cosh
0
sinh(n + 1)
0
sinh
0

1
sinh(n + 2)
0
sinh n
0
_
} y
F
E
=
2E
2

h
coth
0

n=1
_
n
2
sinh
2
n
0

n(n + 1) cosh
0
sinhn
0
sinh(n + 1)
0
_
y
Because of symmetry, there is no force in the horizontal di-
rection.
We are going to use this solution to gain some condence
in our numerical simulations. In Fig.3, you can see the com-
parison of COMSOL results with the analytical expression.
The match is especially good when the particle is close to the
wall. It gets worse as this distance increases, because the ef-
fects from the other walls also kick in. The simulation in
Fig.3 has run with a box of size 20x20, and maximum mesh
size 1, with ner mesh on the particle, specically a maxi-
mum size of 0.1. Experimentation with Comsol shows that
the hydrodynamic error is sensitive to the size of the box,
while the error in the electrostatic force is more sensitive to
6
1 1.5 2 2.5 3 3.5 4
0
100
200
300
400
h/a
F
H
/
(

E
2

a
)


Numerical
Analytic
1 2 3 4
0
10
20
R
e
l
.

E
r
r
o
r

%
h/a
1 1.5 2 2.5 3 3.5 4
0
1
2
3
4
h/a
F
E
/
(

E
2

a
)


Numerical
Analytic
1 2 3 4
0
5
10
15
R
e
l
.

E
r
r
o
r

%
h/a
W, H = 20
FIG. 3: The Comsol numerical solution is compared to the ana-
lytical one given by Zhao and Bau. Although the absolute errors
tend to remain small, and the curves look identical, the relative
error grows fast as the particle is located at larger distances from
the planar wall.
the mesh size. For a larger box, 40x40, and twice as ner
mesh, we have obtained a similar picture with errors cut to
about one third of their values in Fig.3 (results not shown).
B. Cylinder in an AC eld
As the electric elds are screened quickly by the elec-
trolyte, an AC eld is usually preferred. Use of an AC elec-
tric eld also prevents harmful reactions on electrodes, and
enables experimentalists to go to higher applied voltage dif-
ferences. Such higher voltages may be desirable if they lead
to stronger electrokinetic eects of interest.
Far from the wall, the ICEO slip velocity around an ideally
polarizable cylinder in an AC eld was derived by Squires
and Bazant [39], which takes the dimensionless form
u

=
sin 2
1 +
2
. (8)
We use this expression to calibrate our numerical code, and
nd excellent agreement far from the wall. This result shows
that ICEO ow decays algebraically as
2
above the RC
charging frequency. Since electrostatic forces do not decay in
this limit, we may expect a change in behavior near the wall.
At high frequency, there is not enough time for double-layer
10
1
10
0
10
1
10
2
0.5
0
0.5
1
1.5
2
2.5
/(
D
a/D)
F
E
/
(

E
2
)
a
h = 0.1a
h = 0.2a
0.3a
h = a
FIG. 4: The total electrostatic force on the cylinder changes sign
as the frequency is increased. As the frequency approaches inn-
ity, this force has a nite nonzero limit.
log
10

l
o
g
1
0
(
h
/
a
)


1 0 1 2
1
0.8
0.6
0.4
0.2
0
0.2
0.4
0
50
100
150
F
total
/ ( E

2
a)
FIG. 5: Contour plot of total force on the ideally polarizable cylin-
der. There is an equilibrium distance between the cylinder and
the wall at high enough frequencies, indicated by the yellow con-
tour line. As the frequency is increased, this distance decreases.
relaxation, so the electric eld ressembles that of a conductor
in a uniform dielectric medium.
An important observation is that the total hydrodynamic
forces vanishes at higher frequencies whereas the total elec-
trostatic force changes sign. As a result, if the frequency is
high enough, there is an equilibrium distance from the wall.
This distance decreases as the frequency is increased.
In the high frequency limit, the electrostatic problem ap-
proaches to the solution of the dirichlet problem, that is, the
7
laplace equation

= 0
combined with the boundary condition

= 0 (on the cylinder)

n

= 0 (on the wall)

= E

x (at innity)
In this case, the solution is given by
= Re(

e
it
) =

cos t
This problem can be solved analytically, and the solution is
given by

= 2cE

n=1
e
n0
coshn
0
coshnsin n
c sin
cosh cos
= 2c

n=1
_
e
n0
coshn
0
coshn e
n
_
sin n
Plugging this into the electrostatic force leads to the formula
F
E,
= 2cE

n=1
(
n
2
cosh
2
n
0
+
n(n + 1) cosh
0
sinh n
0
sinh (n + 1)
0
)
with the same notation as in Zhao and Bau.
C. Sphere in an AC eld
ICEO ow around a sphere was rst considered by
Gamayunov et al. [10]. Following the cylinder analysis of
Squires and Bazant [39], it is straightforward to derive the
(dimensionless) ICEO slip velocity around an ideally polar-
izable sphere in an AC eld, far from the wall,
u

=
9
16
sin 2
1 + (/2)
2
(9)
Note that since cos
2
t = 1/2 the ICEO ow in a constant
DC eld E
0
is twice as large as the time-averaged ow in an
AC eld E
0
cos t of the same amplitude: u
DC

= 2u
AC

.
For a sphere near a wall, the results are qualitatively the
same as for a cylinder near a wall. That is, in the DC case,
both the hydrodynamic and the electrostatic forces are re-
pulsive. Moreover, the magnitude of hydrodynamic forces
are about 2 orders of magnitude larger than the electrical
forces.
Note that the steady and time-periodic plots are consis-
tent: At h/a = 1.5, the steady plots show forces about
F
H
8 and F
E
0.12. The time-periodic plots, at the low
1 2 3 4
0
10
20
30
40
50
h/a
F
H
/
(

E
2
)
a
1 2 3 4
0
0.2
0.4
0.6
0.8
h/a
F
E
/
(

E
2
)
a
FIG. 6: The hydrodynamic and electrostatic forces on a full metal
sphere.
10
0
10
2
0.05
0.1
0.15
0.2
0.25
/(
D
a/D)
F
E
/
(

E
2
)
a


10
1
10
0
10
1
0
1
2
3
4
5
/(
D
a/D)
F
H
/
(

E
2
)
a


FIG. 7: Both the hydrodynamic and the electrostatic forces on
an isotropic sphere are repulsive (away from the wall). While
hydrodynamic forces on the sphere decline as a function of forcing
frequency, electrostatic forces get stronger.
frequency limit, start o at values F
H
4 and F
E
0.06,
which are half of their steady counterparts.
Unlike the cylinder problem, however, the electrostatic
forces always remain repulsive, and therefore there is no equi-
librium plane for the spherical particle, Instead, it is repelled
to innity by the wall regardless of the forcing frequency.
IV. JANUS SPHERE NEAR A WALL
A. Broken symmetries
Without a nearby wall, a Janus sphere would align itself
perpendicular to the electric eld, that is, some of the elec-
tric eld lines would be included in the plane dividing the
Janus particles metal and insulating sides. This eect has
been demonstrated in [40], and is presumably stronger than
the wall eects, at least when the particle is suciently away
from the wall. That being said, we will assume that the par-
ticle always stays in the described conguration, that is, its
8
dividing plane aligned with the electric eld. This is not to
say that the particle has no room for dierent rotational con-
gurations, it can still rotate around x and y axis. Rotations
about the y axis (if existed) leave the particle unchanged,
so we are left with rotations only around the x axis. This is
much easier to deal with than the original problem though,
as only one angle is enough the describe the particles orien-
tation.
Far from the wall, the bulk velocity perpendicular to a
DC eld in the stable orientation is given by the formula of
Squires and Bazant [40] (Eq. 3.16), which takes the dimen-
sionless form,
U
DC
=
9
64
= 2U
AC
( 0) (10)
neglecting compact-layer surface capacitance. As noted
above, the time-averaged velocity in a sinusoidal AC eld
is smaller by a factor of two in the limit of zero frequency.
Even in the bulk, without a wall, it is dicult to solve an-
alytically for the ICEO ow at nite AC frequency around
a Janus particle, since the electrical response is not simply
an induced dipole, due to the broken symmetry. Neverthe-
less, we will see that the frequency dependence of the ow is
similar to that around a sphere (9), constant below the RC
charging time and decaying above it.
For a Janus sphere aligned perpendicular to the elec-
tric eld near a wall, a crucial observation is that the
ysymmetry breaks down. As a result, there is a net force in
the ydirection, as well as a net torque in xdirection. The
former leads to translation parallel to the wall, while the lat-
ter causes rotation of the dielectric face toward the wall. We
shall see that these eects of broken symmetry completely
change the behavior near wall in an AC or DC eld: Al-
though a polarizable sphere is always repelled to innity by
an insulating wall, a Janus particle is always (eventually)
attracted to it.
B. Basic mechanism for wall attraction
The key new eect is rotation due to hydrodynamic torque
caused by asymmetric ICEO ow near the wall. This gen-
erally causes the Janus particle to be attracted to the wall,
as shown in gure 8. The physical mechanism can be un-
derstood as follows. When the eld is rst turned on, the
Janus particle quickly rotates, by ICEP and DEP, to align
its metal/insulator interface with the eld axis, but with an
arbitrary azimuthal angle, mainly set by the initial condi-
tion. As described by Squires and Bazant [40], the ICEO
ow around the particle draws in uid along the eld axis
and ejects it radially at the equator but only on the polariz-
able hemisphere, which acts like a jet engine drives ICEP
motion in the direction of the non-polarizable hemisphere,
which leads the way like a nose.
Near a wall, as shown in the gure, the outward ICEO ow
pushes down on the wall harder on the side of the polarizable
U
E
T
u
FIG. 8: Sketch of ICEO ows u and resulting ICEP torques T
which cause a Janus particle to tilt its less polarizable end to-
ward a wall, while translating toward the wall (until stopped by
double-layer overlap) and perpendicular to the applied AC eld
E (directed into the page and parallel to the wall). This physi-
cal mechanism may explain why the transverse ICEP motion of
Janus particles was observable over the surface of a glass wall in
the experiments of Gangwal et al. [11].
engine than on that of the non-polarizable nose, which
produces a hydrodynamic torque tilting the nose toward the
wall. A second cause of this rotation is the hydrodynamic
coupling between ICEP translation parallel to the wall and
rotation by shear stresses to cause rolling past the wall. Re-
gardless of the initial position, these two sources of ICEP
rotation cause the nose to eventually face the wall, so that
the translational engine drives it toward the wall. This is
likely the origin of the counter-intuitive attraction of Janus
particles to a glass wall in the experiments of Gangwal et
al [11].
What happens next depends on the details of the particle-
wall interaction at very close distances. We will see that the
bulk model with thin double layers must eventually break
down, since the particle eventually collides with the wall,
leading to overlapping particle and wall double layers. It is
beyond the scope of this work to accurately treat the nonlin-
ear and time-dependent behavior of these overlapping dou-
ble layers, so we will explore two models: (i) innitely thin
double layers, i.e. using the bulk model to arbitrarily small
heights, and (ii) a cuto collision height, where overlap-
ping double layers stop any further motion toward the wall,
while still allowing transverse motion. The latter case as-
sumes, as in the experiments [11], that the particles and walls
have equilibrium surface charge of the same sign. For con-
creteness, we will simulate Model (ii) with a cuto height
h = = 0.05a, e.g. corresponding to a double-layer thick-
ness (screening length) of = 50nm with particles of size
a = 1m.
Based on the simple examples above, we expect a subtle
dependence on the AC frequency. Electrostatic DEP motion
will always begin to dominate the hydrodynamic ICEP mo-
tion at high frequency. Therefore, we now consider the low
and high frequency cases separately.
9
0 30 60 90
0.1
0
0.1

U
y
0 30 60 90
0.05
0
0.05

U
z


0 30 60 90
0
0.5
1
1.5

x
h = 0.5a
h = 0.05a
FIG. 9: In the DC limit ( 0), we plot (a) horizontal velocity
(b) vertical velocity and (c) tilting speed (degrees/charging time)
as a function of the tilt angle for the janus particle at distances
h = 0.5a and h = 0.05a from the wall.
C. Dynamics as a function of AC frequency
As shown in Fig. 9, in the low frequency limit, the Janus
particle experiences a rotational velocity turning its non-
polarizable side toward the wall, as explained above. The hy-
drodynamic ICEP torque is orders of magnitude larger than
the electrostatic DEP torque, until the particle gets quite
close to the wall. The magnitude of the horizontal ICEP
velocity U
y
parallel to the surface and perpendicular to the
eld is close to its bulk value U
y
= 9/128 0.07 even fairly
close to the wall at a height h = 0.5a at zero tilt, but reduces
with the tilt angle. For small tilt angles and close to the wall
at h = 0.05a, the horizontal velocity increases to U
y
0.10,
but it drops below the bulk value at larger tilt angles, e.g.
to U
y
0.05 at = 45 degrees. Below we will see that this
velocity is further reduced at higher forcing frequencies, due
to the reduction of ICEO ow (since DEP cannot contribute
to motion perpendicular to a uniform eld).
Regardless of the orientation, in the DC limit the particle
moves ever closer to the wall in Model (i) since U
z
< 0 for
any tilting of the nose toward the wall. Even if the the ver-
tical motion is stopped at a critical height in Model (ii), the
rotation continues in the DC limit until the particle points
its non-polarizable nose directly at the wall ( = 90) and the
motion stops, although this can take a long time, since the
rotation slows down substantially for tilt angles larger than
45 degrees. As discussed below, a number of eects might
lead to such a stabilization of the tilt angle, thus allowing
0 1000 2000 3000 4000 5000
0
0.05
0.1
t/(/ E

2
)
U
y
/
a

(
1
/
s
)
(a)
0 1000 2000 3000 4000 5000
0
2
4
t/(/ E

2
)
z
/
a
(b)
0 1000 2000 3000 4000 5000
0
45
90
t/(/ E

2
)

(c)
0 20 40 60 80
0
5
y/a
z
/
a
(d)
FIG. 10: Typical trajectory of a janus particle under the DC limit
0 interacting with the wall: As a function of time, plotted
are (a) The horizontal speed (b) Distance from the wall (c) Tilt
angle. Also, we plot the distance from the wall as a function of
horizontal position in (d).
steady translation along the wall.
As shown in Fig. 10, a typical simulated trajectory of the
Janus particle shows it translating perpendicular to the eld
while rotating and attracting to the wall, until eventually
coming to rest facing the wall. Even when the particles
motion stops, however, its polarizable hemisphere (engine)
continues driving a steady ICEO ow, which can lead to
long-range hydrodynamic interactions with other particles.
This is an interesting theoretical prediction which should be
checked in experiments. Such immobilized Janus particles
may have interesting applications in microuidics.
Similar behavior is predicted for nite AC frequencies in
many cases. In particular, if a particle is initially mostly
facing its non-polarizable hemisphere toward the wall ( near
90

), it will swim toward the wall and come to rest, as in the

DC limit of Figure 10.
There are some new eects in AC elds, however, since
ICEO ows are suppressed with increasing frequency. The
competing eect of DEP can prevent the Janus particle from
fully rotating and coming to rest on the surface, at least
in Model (ii) where the collision is prevented articially, as
shown in Figure 12. At = 1 (the characteristic RC fre-
quency of the particle), the rotation slows down substantially
beyond 45

but does not appear to stop. In this regime the

horizontal velocity decays to U
y
0.015. For = 10 the
particle appears to settles down to an equilibrium tilt angle
around 45

, while steadily translating over the wall. The

limiting horizontal velocity is roughly U
y
0.009.
10
0 30 60 90
0
0.05
0.1

U
y
0 30 60 90
0.05
0
0.05

U
z


0 30 60 90
0
1
2

x
h = 0.5a
h = 0.05a
FIG. 11: For AC frequency c = 1, we plot (a) horizontal ve-
locity (b) vertical velocity and (c) tilting speed (degrees/charging
time) as a function of the tilt angle for the janus particle at
distances h = 0.5a and h = 0.05a from the wall.
0 1000 2000 3000 4000 5000
0
0.02
0.04
t/(/ E

2
)
U
y
/
a

(
1
/
s
)
(a)
0 1000 2000 3000 4000 5000
0
1
2
t/(/ E

2
)
z
/
a
(b)
0 1000 2000 3000 4000 5000
0
45
90
t/(/ E

2
)

(c)
0 20 40 60
0
5
y/a
z
/
a
(d)
FIG. 12: Typical trajectory of a janus particle under AC fre-
quency c = 10 interacting with the wall: As a function of time,
plotted are (a) The horizontal speed (b) Distance from the wall (c)
Tilt angle. Also, we plot the distance from the wall as a function
of horizontal position in (d).
0 30 60 90
0
0.01
0.02

U
y
0 30 60 90
0.02
0.01
0
0.01

U
z


0 30 60 90
0
0.1
0.2

x
h = 0.5a
h = 0.05a
FIG. 13: For AC frequency (c = 10), we plot (a) horizontal ve-
locity (b) vertical velocity and (c) tilting speed (degrees/charging
time) as a function of the tilt angle for the janus particle at
distances h = 0.5a and h = 0.05a from the wall.
D. Comparison to experiment
The simulations with Model (ii) are in reasonable agree-
ment with the experimental observations of Gangwal et
al [11] for metallo-dielectric Janus particles in dilute NaCl so-
lutions in the low-frequency regime < 1. The bulk theory
of Squires and Bazant (10) accurately ts the experimental
velocity as a function of the eld strength (Fig. 3 of Ref. [11])
and the particle size (Fig. 4), if a ratio = 10 for the
compact-layer to diuse-layer capacitance is used to obtain
the dimensionless formula, U
expt
= (9/128)/(1+10) = 0.006.
However, the ICEP motion is observed only very close to the
walls.
Our simulations predict that the particles are quickly at-
tracted to the walls over a time of order the channel width
(60 m) divided by the typical ICEP velocity (10 m/s),
which is roughly one minute, consistent with experimental
observations. The particles are also predicted to tilt, and
moderate tilt angles can also be inferred from experimental
images, although more accurate measurements are needed. If
the tilt angle stabilizes around 45

(see below), then the sim-

ulations (Fig. 9) predict that the ICEP translational velocity
should be only 0.05/0.07 = 70% of the bulk value close to
the wall, which would imply = 7. This value is somewhat
larger than that inferred from prior experiments on ICEO
ow in dilute KCl around a larger (100m radius) platinum
cylinder [23], but it is also observed that the ICEP velocity
is slower than predicted at larger sizes (Fig. 4 of Ref. [11]).
11
Apart from the rotational dynamics, therefore, the theory is
able to predict the ICEP velocity fairly well.
Without stopping the rotation articially, we are able to
predict the experimentally observed steady motion along the
wall only at moderate to large . The reduction of ICEO
ow in this regime reduces hydrodynamic torque (see below)
and also enhances the eect of stabilizing electrostatic forces.
Although U
expt
= 0.006 is measured in the low-frequency
plateau < 1, this behavior otherwise seems quite consis-
tent, since the slower ICEP velocity can also t the exper-
imental data using smaller (and perhaps more reasonable)
values of . For example, the predicted velocity of U = 0.015
at = 1 implies = 1.5, while the velocity U = 0.009 at
= 10 implies = 0.5.
The diculty in predicting the stable tilt angle at low
frequency may be due to our use of the low-voltage, dilute-
solution theory, which generally overpredicts the magnitude
of ICEO ows, especially with increasing salt concentration.
For example, the electrophoretic mobility can saturate at
large induced voltages, and the charging dynamics can also
be altered signicantly when crowding eects are taken into
account [5]. As a result, our simulation results at moder-
ate frequencies = 1 10, which exhibit reduced ICEO
ow due to incomplete double-layer charging, may ressemble
the predictions of more accurate large-voltage, concentrated-
solution theories at low frequency < 1, where ow is re-
duced instead by ion crowding in the double layer. This will
be the subject of future work.
E. Contact mechanics
Another source of error in the model is our inaccurate
treatment of the contact region, where double-layers overlap.
We have simply imposed a small cuto height in Model (ii) to
prevent the wall collision, but there may be more complicated
mechanical eects of the contact region. In particular, there
may be enhanced hydrodynamic slip, due to the repulsion of
overlapping (equilibrium) double layers of the same sign, as
in the experiments.
By examining the forces and torques close to the wall, we
can infer to some degree what mechanical properties of the
contact region might lead to the observed ICEP sliding along
the wall and smaller tilt angles at lower frequencies (and thus
also somewhat larger velocities). As shown in Fig. 14, before
the particle gets very close to the wall, the (mostly hydro-
dynamic) torque acts to completely tilt the non-polarizable
face toward the wall leading to collision. As noted above in
Fig. 8, this can be understood as a result of the downward
component of ICEO ow on the polarizable hemisphere rais-
ing the pressure by pushing on the wall on that side.
The situation changes when the particle gets very close
to the wall. As shown in Fig. 14, the torque changes sign
at a tilt angle which is roughly 45

. This again can be un-

derstood from Fig. 8, since the ICEO ow between the par-
ticle and the wall on the polarizable side, which drives the
0 30 60 90
10
0
10
20
30
40
50
60

x
(a)
0 30 60 90
4
2
0
2
4
6

x
(b)


h = 0.5a
h = 0.05a
0

c
= 10
FIG. 14: Torque on a xed Janus sphere versus tilt angle at
heights h = 0.5a and 0.05a when (a) 0 (b) c = 10.
0 30 60 90
0
1
2
3
4
5

F
y
(a)


0 30 60 90
0
0.2
0.4
0.6
0.8
1

F
y
(b)
h = 0.5a
h = 0.05a
0
c
= 10
FIG. 15: Horizontal force on a xed Janus sphere versus tilt angle
at heights h = 0.5a and 0.05a when (a) 0 (b) c = 10.
torque, is mostly absent. It would thus seem that even in a
DC eld, the particle would not rotate any farther, but this
thinking neglects the hydrodynamic coupling between trans-
lational force and rotational velocity near the wall, Eq. (4).
In Fig. 15, we see that the force on the particle parallel to the
wall F
y
remains strong, and this leads to a rolling eect over
the wall due to shear stresses. For this reason, the rotational
velocity persists in Fig. 9 even when the torque goes to zero
in Fig. 14.
The model assumes no slip on all non-polarizable surfaces,
but this may not be a good approximation near the contact
point when double layers overlap. If the equilibrium sur-
face charges (or zeta potentials) on the non-polarizable hemi-
sphere and the wall have opposite signs, then the overlapping
double layers lead to a strong attraction, which would only
stien the eective contact with the surface, and thus only
increase the viscous rolling eect during motion along the
surface. If the equilibrium surface charges (or zeta poten-
tials) have the same sign, however, as in the experiments on
12
gold-coated latex Janus particles near glass walls [11], then
there is a strong repulsion at the contact point. This repul-
sion stops the collision with the wall in Model (ii), but it may
also lubricate the contact and allow for some sliding. This
eective slip over the wall near the contact point could re-
duce the viscous rolling, and, in the absence of torque, cause
the rotation to stop, or at least be reduced for tile angles
above 45

. In that case, we might expect a more accurate

model of the contact region to predict to the experimentally
observed motion, sliding over the surface by ICEP with a
small tilt angle ( < 45

), for a wider range of conditions,

including lower AC frequency, perhaps even in the DC limit.
V. CONCLUSION
We have use the existing low-voltage theory of ICEP to
predict the motion of polarizable particles near an insulating
wall. Our results for symmetric spheres and cylinders con-
rm the expected repulsion from the wall due to ICEO ow,
sketched in Figure 1(a). In the case of the cylinder we show
that attraction is also possible at high frequency, where DEP
from electrostatic forces dominates slip-driven ICEP motion.
Our results for asymmetric Janus particles reveal an un-
expected attraction to the wall by a novel mechanism illus-
trated in Figure 8, which involves tilting of the less po-
larizable face toward the wall. Once it reaches the wall, if
double-layer repulsion prevents further collision, the particle
either rotates completely and ceases to move, while driving
steady ICEO ow, or reaches an equilibrium tilt angle around
45

while steadily translating along the surface, perpendic-

ular to the elecric eld. The latter motion only arises at
moderate frequencies in our model, above the characteristic
charging frequency for the double layers, while in experi-
ments it is also observed at low frequencies. More accurate
models taking into account reduced ICEO ow at large volt-
age in non-dilute solutions and more accurate models of the
contact region may improve the agreement with experiments.
In any case, we have shown that polarizable particles can
display complex interactions with walls due to broken sym-
metries in ICEO ows. Attractive and repulsive interactions
can be tuned by varying the geometry of the particles (and
the walls), as well as the AC frequency and voltage. These
phenomena may nd applications in separations and self-
assembly of colloids or in local ow generation in microuidic
devices.
[1] J. L. Anderson. Eect of non-uniform zeta potential on par-
ticle movement in electric elds. J. Colloid Interface Science,
105:4554, 1984.
[2] J. L. Anderson. Colloid transport by interfacial forces. Annu.
Rev. Fluid Mech., 21:6199, 1989.
[3] M. Z. Bazant. Electrokinetic motion of polarizable particles.
In Dongqing Li, editor, Encyclopedia of Microuidics and
Nanouidics. Springer, 2008.
[4] M. Z. Bazant and T. M. Squires. Induced-charge electro-
kinetic phenomena: Theory and microuidic applications.
Phys. Rev. Lett., 92:066101, 2004.
[5] Martin Z. Bazant, Mustafa Sabri Kilic, Brian Story, and
Armand Ajdari. Nonlinear electrokinetics at large voltages.
arXiv:cond-mat/0703035v2.
[6] M.-H. Chih and J.-P. Hsu. Electrophoresis of a sphere normal
to a plane at arbitrary electrical potential and double layer
thickness. J. Colloid Interface Science, 248:383388, 2002.
[7] S. M. Davison and K. V. Sharp. Boundary eects on the
electrophoretic motion of cylindrical particles: Concentri-
cally and eccentrically positioned particles in a capillary. J.
Colloid Interface Science, 303:288297, 2006.
[8] J. Ennis and J. L. Anderson. Boundary eects on elec-
trophoretic motion of spherical particles for thick double
layers and low zeta potential. J. Colloid Interface Science,
185:497514, 1997.
[9] M. C. Fair and J. L. Anderson. Electrophoresis of heteroge-
neous colloids doublets of dissimilar particles. Langmuir,
8:28502854, 1992.
[10] N. I. Gamayunov, V. A. Murtsovkin, and A. S. Dukhin. Pair
interaction of particles in electric eld. 1. features of hydro-
dynamic interaction of polarized particles. Colloid J. USSR,
48:197203, 1986.
[11] S. Gangwal, O. J. Cayre, M. Z. Bazant, and O. D. Velev.
Induced-charge electrophoresis of metallo-dielectric particles.
arXiv:0708.2417v1, 2007.
[12] K. D. Horng H. J. Keh and J. Kuo. Boundary eects on elec-
trophoresis of colloidal cylinders. J. Fluid Mech., 231:211
228, 1991.
[13] J.-P. Hsu, Z.-S. Chen, M.-H. Ku, and L.-H. Yeh. Eect of
charged boundary on electrophoresis: Sphere in a spherical
cavity at arbitrary potential and double-layer thickness. J.
Colloid Interface Sci., 314:256263, 2007.
[14] J.-P. Hsu and M.-H. Ku. Boundary eect on electrophoresis:
nite cylinder in a cylndrical pore. J. Colloid Interface Sci.,
283:592600, 2005.
[15] J.-P. Hsu, L.-H. Yeh, and Z.-S. Chen. Eect of a charged
boundary on electrophoresis: A sphere at an arbitrary posi-
tion in a spherical cavity. J. Colloid Interface Sci., 310:281
291, 2007.
[16] J.-P. Hsu, L.-H. Yeh, and M.-H. Ku. Evaluation of th electric
force in electrophoresis. J. Colloid Interface Sci., 305:324
329, 2007.
[17] R. J. Hunter. Foundations of Colloid Science. Oxford Uni-
versity Press, Oxford, 2001.
[18] H. J. Keh and J. L. Anderson. Boundary eects on the
electrophoretic motion of colloidal spheres. J. Fluid Mech.,
153:417439, 1985.
[19] H. J. Keh and S. B. Chen. Electrophoresis of a colloidal
sphere parallel to a dielectric plane. J. Fluid Mech., 194:377
390, 1988.
[20] H. J. Keh and J. S. Jan. Boundary eects on diusiophoresis
and electrophoresis: Motion of of a colloidal sphere normal
to a plane wall. J. Colloid Interface Science, 183:458475,
1996.
13
[21] E. Lee, J.-W. Chu, and J.-P. Hsu. Electrophoretic mobility of
a spherical particle in a spherical cavity. J. Colloid Interface
Science, 196:316320, 1997.
[22] Eric Lee, Jhih-Wei Chu, and Jyh-Ping Hsu. Electrophoretic
mobility of a sphere in a spherical cavity. J. Colloid Interface
Science, 205:6576, 1998.
[23] J. A. Levitan, S. Devasenathipathy, V. Studer, Y. Ben,
T. Thorsen, T. M. Squires, and M. Z. Bazant. Experimental
observation of induced-charge electro-osmosis around a metal
wire in a microchannel. Colloids and Surfaces A, 267:122
132, 2005.
[24] Hui Liu, Haim H. Bau, and Howard H. Hu. Electrophore-
sis of concentrically and eccentrically positioned cylindrical
particles in a long tube. Langmuir, 20:26282639, 2004.
[25] D. Long and A. Ajdari. Electrophoretic mobility of compos-
ite objects in free solution: Application to dna separation.
Electrophoresis, 17:11611166, 1996.
[26] D. Long and A. Ajdari. Symmetry properties of the elec-
trophoretic motion of patterned colloidal particles. Phys.
Rev. Lett., 81:15291532, 1998.
[27] J. Lyklema. Fundamentals of Interface and Colloid Science.
Volume II: Solid-Liquid Interfaces. Academic Press Limited,
San Diego, CA, 1995.
[28] F. A. Morrison and J. J. Stukel. Electrophoresis of an insu-
lating sphere normal to a conducting plane. J. Colloid and
Interface Science, 33:88, 1970.
[29] V. A. Mursovkin and G. I. Mantrov. Study of the motion
of anisometric particles in a uniform variable electric eld.
Colloid J. USSR, 52:933936, 1990.
[30] V. A. Murtsovkin. Nonlinear ows near polarized disperse
particles. Colloid Journal, 58:341349, 1996.
[31] G. C. Randall and P. S. Doyle. Dna deformation in electric
elds: Dna driven past a cylindrical obstruction. Macro-
molecules, 38:24102418, 2005.
[32] W. D. Ristenpart, I. A. Aksay, and D. A. Saville. Electrically
guided assembly of planar superlattices in binary colloidal
suspensions. Phys. Rev. Lett., 90, 2003.
[33] W. D. Ristenpart, I. A. Aksay, and D. A. Saville. Ehd ow
around a colloidal particle near an electrode with an oscillat-
ing potential. J. Fluid Mech., 575, 2007.
[34] Klint Rose and Juan G. Santiago. Rotational electrophoresis
of striped metallic microrods. Physical Review E, 75:197203,
2006.
[35] W. B. Russel, D. Saville, and W. R. Schowalter. Colloidal
Dispersions. Cambridge University Press, Cambridge, Eng-
land, 1989.
[36] D. Saintillan, E. Darve, and E. S. G. Shaqfeh. Hydrodynamic
interactions in the induced-charge electrophoresis of colloidal
rod dispersions. Journal of Fluid Mechanics, 563:223259,
2006.
[37] Alexander A. Shugai and Steven L. Carnie. Electrophoretic
motion of a spherical particle with a thick double layer in
bounded ows. J. Colloid Interface Science, 213:298315,
1999.
[38] Paul J. Sides. Electrohydrodynamic particle aggregation on
an electrode driven by an alternating electric eld normal to
it. Langmuir, 17:57915800, 2001.
[39] T. M. Squires and M. Z. Bazant. Induced-charge electro-
osmosis. J. Fluid Mech., 509:217252, 2004.
[40] T. M. Squires and M. Z. Bazant. Breaking symmetries in
induced-charge electro-osmosis and electrophoresis. J. Fluid
Mech., 560:65101, 2006.
[41] T. M. Squires and S. R. Quake. Microuidics: uid physics
on the nanoliter scale. Rev. Mod. Phys., 77, 2005.
[42] Y.-P. Tang, M.-H. Chih, Eric Lee, and J.-P. Hsu. Elec-
trophoretic motion of a charge-regulated sphere normal to
a plane. J. Colloid Interface Sci., 242:121126, 2001.
[43] M. Trau, D. A. Saville, and I. A. Aksay. Assembly of colloidal
crystals at electrode interfaces. Langmuir, 13:6375, 1997.
[44] E. Yariv. Induced-charge electrophoresis of nonspherical par-
ticles. Phys. Fluids, 17:051702, 2005.
[45] Ehud Yariv. Force-free electrophoresis? Phys. Fluids,
18:031702, 2006.
[46] S. Yeh, M. Seul, and B. Shraiman. Assembly of ordered col-
loidal aggregrates by electric-eld-induced uid ow. Nature
(London), 386:57, 1997.
[47] H. Zhao and H. H. Bau. On the eect of induced electro-
osmosis on a cylindrical particle next to a surface. Langmuir,
23:40534063, 2007.
[48] Andrew L. Zydney. Boundary eects on electrophoretic mo-
tion of a charged particle in a spherical cavity. J. Colloid
Interface Science, 169:476485, 1995.
0 1000 2000 3000 4000 5000
0
0.05
0.1
t/(/ E

2
)
U
y
/
a

(
1
/
s
)
0 1000 2000 3000 4000 5000
0
2
4
t/(/ E

2
)
z
/
a
0 1000 2000 3000 4000 5000
0
45
90
t/(/ E

2
)

0 20 40 60 80
0
5
y/a
z
/
a