Electrophoresis 2012, 33, 419–425 419

Jeevanjyoti Chakraborty1 Research Article

Subhashis Ray2
Suman Chakraborty1,3
Role of streaming potential on pulsating
1
Advanced Technology
Development Center, Indian
mass flow rate control in combined
Institute of Technology,
Kharagpur, India
electroosmotic and pressure-driven
2
Institute of Thermal
Engineering, TU-Bergakademie,
microfluidic devices
Freiberg, Freiberg, Germany
3
Department of Mechanical In the present study, we investigate the implications of streaming potential on the mass
Engineering, Indian Institute of flow rate control in a microfluidic device actuated by the combined application of a
Technology, Kharagpur, India
pulsating pressure gradient and a pulsating, externally applied, electric field. We
demonstrate that the temporal dynamics due to streaming potential effects may lead to
Received April 27, 2011 interesting non-trivial aspects of the resultant transport characteristics. Our results
Revised September 19, 2011 highlight the importance of an adequate accounting of the streaming potential effects for
Accepted September 25, 2011
temporally tunable mass flow rate control strategies, which may act as a useful design
artifice to augment mass flow rates in practical scenarios.

Keywords:
Electroosmosis / Microfluidics / Nanofluidics / Streaming potential
DOI 10.1002/elps.201100414

1 Introduction better exploitation of the device performance, indeed,

demand an improved understanding of the interactions
Transport processes pertinent to devices and systems, created between the concerned transiences in the system dynamics
in the wake of the tremendous advancements in micro- and the issues of precise flow control, several aspects of
fabrication techniques, are becoming increasingly important which are yet to be fundamentally understood, critically
in several emerging applications. This is due to their inherent analyzed and emphatically addressed.
advantages such as high transport coefficients (on account of Flow manipulation in many fluidic devices is traditionally
large surface area-to-volume ratios), efficient process manage- achieved by the application of a pressure gradient. However,
ment, excellent durability, reduction of required sample in the microfluidic [2] and nanofluidic [3] regime, flow can
volumes and processing times, effective scaling down for also be induced and controlled precisely taking advantage of
achieving desired functionalities, and addressability of length various other interfacial phenomena like electrokinetics
scales typically encountered in many intricate biophysical which is contingent on the development of the electrical
systems. The concerned applications are truly in diverse high- double layer (EDL) [33]. On application of an external electric
technology areas, including biotechnology and biomedical field on the EDL, the mobile ions in the diffuse zone of this
engineering, inkjet printing, microactuators, microsensors, EDL are actuated to move. This, in turn, induces a flow of the
microreactors, thermo-mechanical data storage systems and liquid – this is usually referred to as EOF. Another electro-
thermal management of electronic devices/systems [1–3]. kinetic flow phenomenon, which manifests itself even with-
Design and optimization of many of these devices/systems out the application of any external electric field, is the
necessitate a comprehensive understanding and analysis of streaming potential effect [33], which is the development of a
the underlying flow characteristics. The device or system back potential due to the migration and build-up of ions due
characteristics may be further complicated by a time- to a pure pressure gradient-driven flow. As this back potential
modulated pulsating nature of the pertinent flow actuation inhibits the very flow that induces it, it results in what is
mechanisms, which may be common to many biophysical as usually called the electroviscous effect (an enhanced effective
well as artificially fabricated fluidic pathways [4–32]. Further flow resistance).
miniaturization of microfluidic device components and a Although the streaming potential by itself is not a flow
actuator, it may hold the capability of dramatically altering
the flow [34]. As such, the streaming potential can signifi-
Correspondence: Dr. Suman Chakraborty, Department of
Mechanical Engineering and Advanced Technology Develop-
cantly alter a flow induced by the combined application of a
ment Center, Indian Institute of Technology, Kharagpur 721302, pressure gradient and an external electric field, provided that
India the induced streaming potential is of comparable strength
E-mail: [email protected] as that of the electrical potential due to the applied field, or
Fax: 191-3222-282278

Abbreviation: EDL, electrical double layer Colour Online: See the article online to view Figs. 1 and 3 in colour.

& 2011 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.electrophoresis-journal.com
420 J. Chakraborty et al. Electrophoresis 2012, 33, 419–425

even stronger. The underlying consequences may get axial direction, and y is the transverse direction (HryrH
further involved in case the applied driving fields possess a such that the channel centre-line is given by y 5 0).
temporal variation over and above their steady components. The pulsating pressure gradient, in the most general
These temporal variations of the driving agents together case, is given by:
with the consequent temporally varying streaming potential  
qp qp  X 1
qp 
may have far ranging implications, dictating the subtle ¼  1 sinðo1n tÞ ð2Þ
qx qx 0 n¼1
qx  n
intricacies of temporally dependent mass flow rate control.
From an alternative viewpoint, the existence of the stream- The electric field is a combined consequence of the electric
ing potential offers a delicate scheme to manipulate the flow field (ES) due to the induced streaming potential and the
by tuning the externally applied electric field in various externally applied electric field (Ex)
proportions of the generated streaming potential. In order to X1 X1
explore a broader regime of control, it is imperative that the E ¼ ES0 1Ex0 1 ESn sinðo1n tÞ1 Exn sinðo2n t1fÞ ð3Þ
n¼1 n¼1
requirements of the Debye–Hückel small zeta potential
limit be relaxed [27, 34]. where f is the phase difference between the applied pres-
From a practical viewpoint, it is the volumetric flow rate sure gradient oscillation and the externally applied electric
and not the velocity profile per se that bears a more signif- field oscillation, the subscript ‘x’ denotes the
icant consequence in relation to flow control aspects of externally applied field component along the axial direction
fluidic devices. From this consideration, a research study and the subscript ‘S’ denotes the induced streaming
has been reported in the literature on the pulsating nature of component. The EDL potential distribution is given
the driving electrical field for achieving desired mass flow by the Poisson equation [33], which, in one dimension,
rate characteristics in time periodic pure electroosmotic reduces to
flows [24]. Motivated by this initial investigation and the d2 c r
possibilities of flow alteration as mentioned earlier, we study ¼ e ð4Þ
dy2 e
here the effect of the streaming potential on the mass flow
rate control characteristics of a flow actuated by the where e is the permittivity of the medium. Assuming the
combined application of a pulsating pressure gradient and a electrolyte to be symmetric, the charge density, re, is given
pulsating external electric field. We also investigate the by
implications of the near-wall conductivities relative to the re ¼ ezðn1  n Þ ð5Þ
bulk conductivities on the mass flow rate control. We apply a
where e is the protonic charge, z is the valence of the positive
recently established theory [34] for the accurate calculation
ions and n7 represents the number density of the positive/
of the streaming potential, together with requisite additions
negative ion. This number density is given, using the
[27] to take into account the distinctive effect of the surface
Boltzmann distribution, as
and bulk conductivities. Using this framework, we establish  
certain interesting results that are direct manifestations of ezc
n ¼ n0 exp  ð6Þ
the temporal dynamics of the induced streaming potential kB T
together with the applied electric field and the applied where n0 is the bulk ionic concentration, c is the potential
pressure gradient, and their combined interplay. These distribution in the EDL, kB is the Boltzmann constant and T
results are likely to act as an important design basis for is the absolute temperature. Combining Eqs. (5) and (6), the
microflow manipulation and control in a time-pulsating celebrated Poisson–Boltzmann equation, describing the
environment. potential distribution in the EDL, which screens the chan-
nel-wall surface charge, is obtained as:
 
d2 c 2ezn0 ezc
2 Mathematical modeling ¼ sinh ð7Þ
dy2 e kB T
We consider transient flow through a slit channel The above equation may be solved analytically in its
(with half-height H). The flow is actuated by the combined nonlinear form, subjected to the boundary conditions: (i) at
application of a pulsating pressure gradient and a y 5 H and y 5 H, c 5 z, and (ii) y 5 0, dc=dy ¼ 0, to yield
pulsating electric field. In the case of low Reynolds number (considering identical surface charging effects at both the
flow, the general form of the momentum conservation walls) [35, 36]
equation reduces to the unsteady version of the Stokes’
8    
equation >
> 4kB T 1 zez
>
> tanh tanh exp f k ð H  yÞ g ; 0yH
< ze 4kB T
qu q2 u qp c¼
rf ¼ m 2  1re E ð1Þ >    
qt qy qx >
> 4k T zez
> B tanh1 tanh
: expfkðH1yÞg ; H  y  0
ze 4kB T
where rf is the fluid density, u is the fluid velocity in the
axial direction, m is the dynamic viscosity of the fluid, re is ð8Þ
the charge density and E is the electric field. Here, x is the 2
where k ¼ 1=l, l ¼ ekB T=2n0 e z .2 2

& 2011 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.electrophoresis-journal.com
Electrophoresis 2012, 33, 419–425 General 421

It is important to mention here that although in the only. With this understanding, the electric field due to the
most general case the applied pressure gradient and the streaming potential is calculated by equating the ionic
electric field are expressed as in Eqs. (2) and (3) in the form current to zero. The fluid velocity considered for
of a Fourier series, a significant physical insight can be the sole purpose of calculation of the streaming
obtained by considering the special case of a sinusoidal potential is due to a combination of the external pressure
variation, so that n 5 1, which gives a signature of the gradient and the induced streaming electric field, and is
system response against fields with a generic periodic given by
  
nature. Further, for the sake of notational convenience, we 1 qp eðES0 Þ
u0 ¼ ðy2  H 2 Þ1 ðc  zÞ
write o11o1 and o21o2, in Eq. (3). Considering these, 2m qx 0 m
Eqs. (1)–(3) lead to:     
1 qp coshðB1 yÞ
     1Im expðio1 tÞ 1
qu q2 u qp  qp 2
mB1 qx 1 coshðB1 HÞ
rf ¼m 2  1 sinðo1 tÞ
qt qy qx 0 qx 1 ð9Þ
2 2
coshðB1 yÞ
Z H 33
6 6 coshðB HÞ re sinh½B1 ðH  ZÞdZ 77
1re fES0 1Ex0 1ES1 sinðo1 tÞ1Ex1 sinðo2 t1fÞg 6 ES1 6 1 0 77
1Im6 6mB1 expðio1 tÞ6
6 77
77
4 4 Z y 55
This equation is solved subjected to the following boundary  re sinh½B1 ðy  ZÞdZ
0
conditions: (i) at y 5 0, qu=qy ¼ 0, and (ii) at y 5 H, u 5 0,
together with the already mentioned boundary conditions on ¼u00 1Im½u01 expðio1 tÞ
c. The solution is obtained as a superposition of (i) the ð11Þ
steady part u0 of the velocity field (in response to the steady
where
pressure gradient, the steady electric field due to the induced   
1 qp 
2  eES0
streaming potential, and the steady externally applied electric u00 ¼ y  H2 1 ðc  zÞ
field), (ii) periodic velocity field (in response to the periodic 2m qx 0 m
pressure gradient and the periodic electric field due to the and
induced streaming potential) and (iii) the periodic velocity    
1 qp coshðB1 yÞ
u01 ¼ 1
field (in response to the periodic externally applied electric mB21 qx 1 coshðB1 HÞ
 Z H Z y 
field). Thus, the complete solution is obtained as ES1 coshðB1 yÞ
1 re sinh½B1 ðH  ZÞdZ  re sinh½B1 ðy  ZÞdZ
    mB1 coshðB1 H Þ 0 0
1 qp  eðES0 1Ex0 Þ
u¼ ðy2  H2 Þ1 ðc  zÞ
2m qx 0 m The ionic current is given by
|fflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflffl{zfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflffl}
u0 Z H
z2 e2 Esi
8  
1 qp
 
coshðB1 yÞ
 9 iion;i ¼ 2ze ðn1  n Þu0i dy12
>
>
> expðio1 tÞ Im 1 > >
> 0 f
>
> 2 qx
mB1 coshðB1 HÞ >
>
>
>
>
1 >
>
> Z H
>
>
> 2 2 Z H 33 >
>
> ð12Þ
< coshðB1 yÞ =  ðn1 1n Þdy12sStern Esi
1 6 6 re sinh½B1 ðH  ZÞdZ 77 0
>
> 6 6 coshðB1 HÞ 0 77 >
>
> 6 ES1 6 77 >
>
>
>
>
1Im6 expðio1 tÞ6 77 >
>
>
> with i ¼ 0; 1
>
> 6mB 1 6 Z y 77 >
>
>
> 4 4 55 >
>
:  r sinh½B1 ðy  ZÞdZ ; e
0
|fflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflffl{zfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflffl
ffl} where f is the friction coefficient of the ions, and sstern is the
u1
Stern layer conductivity.
8 2 2 Z H 339
>
> coshðB2 yÞ >
>
> 6 6 re sinh½B2 ðH  ZÞdZ 77> >
>
< 6E
x1 6 coshðB2 HÞ 0 77= We use the following non-dimensionalization scheme:
1 Im6 6 expfiðo2 t1fÞg6 6 7 7
> Z 7 7
>
> 4 2
mB 4 y 55> >
> y c ez sl
>
:  re sinh½B2 ðy  ZÞdZ > ; y ¼ ; c ¼ ; t¼ot; a ¼
0
|fflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflffl{zfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflffl} H sl=e kB T e
u2

ð10Þ ESi Exi

ESi ¼ ; Exi ¼with ði ¼ 0; 1Þ
mU=sl mU=sl
where, B2m ¼ iom = n; m51; 2 and n ¼ m=r is the kinematic  
viscosity. l 1 H2 qp
r e ¼  re ¼ sinhðac Þ; U ¼ 
In Eq. (10), values of the steady and time-periodic s a 2m qx 0 ð13Þ
components of the electric field due to the induced uj
streaming potential, Esi (i 5 0,1) are unknown. In order to u j ¼with j ¼ 0; 1; 2
U
determine the same, one may note that the velocity field in 
qp
present situation can be viewed as the superposition of the qx 1 H ok H 2

velocity fields due to the following effects acting in tandem: P ¼ ; K ¼ ; R2k ¼ with k ¼ 1; 2
qp l n
(i) pressure gradient in combination with its streaming qx 0

potential field, and (ii) the externally applied electric field. It Using the aforementioned non-dimensional scheme, we
is important to mention in this context that the streaming obtain the net non-dimensional fluid velocity in the channel as:
potential, by definition, is a consequence of the ionic
u ¼ u 0 1u 1 1u 2 ð14Þ
advection due to the pressure-driven flow component

& 2011 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.electrophoresis-journal.com
422 J. Chakraborty et al. Electrophoresis 2012, 33, 419–425

In particular The non-dimensional mass flow rate is given by



u 0 ¼ 1  y 2 1 ES0 1Ex0
ðc  z Þ ð15aÞ m

  m ¼ R1 ð20Þ
u 1 ¼ Im 2P u^ p  ES1
K 2 u^ ES expðio1 tÞ ð15bÞ 2rfHU ¼ 0 u dy

 
u 2 ¼ Im Ex1
K 2 u^ Ex expfiðo2 t1fÞg ð15cÞ
In the expressions of the above equation

pffi  ! 3 Results and discussion
1 cosh iR1 y
u^ p ðy Þ ¼ 2
pffi   1 ð16aÞ
iR1 cosh iR1 We perform a comparative study of the non-dimensional
8
pffi ! Z 1 9 mass flow rate with and without considering the streaming
> cosh iR1 y h pffi i >
>
> > potential effects. mSP
denotes the non-dimensional
>
pffi 
re sinh iR1 ð1  Z >
 Þ dZ >
1 < cosh iR1 0
=
mass flow rate considering streaming potential effects,
u^ ES ðy Þ ¼ pffi Z y
iR1 >
>
>
h pffi i > >
> while m denotes the non-dimensional mass flow rate
>
:  r e sinh iR1 ðy  Z >
 Þ dZ ;
0
without considering streaming potential effects. The rele-
ð16bÞ vant parameter values used in the present calculation
8 9 are: e 5 1.6  1019 C, kB 5 1.38  1023 J/K, T 5 300 K,

pffi ! Z 1 h pffi i >
>
> cosh iR2 y > e 5 80  8.85  1012 C V1 m1, m 5 1  103 N m2 s,
>
>
pffi 
re sinh iR2 ð1  Z >
 Þ dZ > 3
1 < cosh iR2 0
= sb 5 7.2  10 S/m, l 5 20 nm, z 5 1. Unless otherwise
u^ Ex ðy Þ ¼ pffi Z y mentioned, we use P 5 1, Ex0
¼ 1, Ex1 ¼ 1. The values
iR2 >
>
>
h pffi i > >
>
>
:  r e sinh iR2 ðy  Z >
 Þ dZ ; of a ¼ ðez=kB TÞðsl=eÞ, K ¼ H=l, Ri ¼ oi H2 =n, (where
2
0
i 5 1, 2) and Du ¼ sStern =sb H are mentioned in the relevant
ð16cÞ
contexts.
It is interesting to note that if R1 5 R2 (i.e. o1 5 o2), Figure 1 shows the variation in the ratio of the ampli-
then u^ ES ¼ u^ EX . tudes of the mass flow rate with and without considering the
We have, until now, expressed the velocity in terms of streaming potential effect, with the phase difference f
the as yet unknown streaming potential electric fields (the between the applied pressure gradient and the externally
steady component ES0 and the time periodic component applied electric field for various values of the ratio

ES1). In order to find explicit expressions for these compo- f1 ¼ Ex1 =ES1 . The values of a, R( 5 R1 5 R2) and K are kept
nents, we set the ionic current (from Eq. 12) equal to 0, i.e. fixed at 4, 0.1 and 5, respectively. This amplitude is half
iion,i 5 0. In addition, we employ the aforementioned non- the difference between the maximum and the minimum
dimensional scheme. Thus values of the oscillating mass flow rates. The amplitude has
1 1
R the potential of being non-trivially altered by the interplay
a 1ð1  y 2 Þ sinhðac Þdy
ES0 ¼ hR i R1 between the streaming potential and the phase difference
1
msb
s2 y12Du 1 a1 1 ðz  c Þ sinhðac Þdy
1 coshðac Þd between the applied driving agents – the pressure gradient
ð17Þ
 sffiffiffiffiffiffiffiffiffiffiffiffi   
 2P I 2 1I 2 
 3 4 1 I1 I4  I2 I3
ES1 ¼  2  exp i tan ð18Þ
 K I12 1I22  I1 I3 1I2 I4

Further, we denote the bulk ionic conductivity as:

sb ¼ n0 z2 e2 =f , and the surface charge density on the wall
 
 
as: s ¼ edc dy  ¼ edc
dy  . The Stern layer conductivity
y¼H y¼H
may be expressed relative to the diffuse layer conductivity by
means of the Dukhin number, given by: Du ¼ sStern =sb H.
In the expression of Eq. (18)
Z 1 Z 1 
1 msb
I1 ¼ Refu^ ES ðy Þg sinhðac Þdy1 coshðac Þdy12Du
a 1 s2 K 2 1

ð19aÞ
Z 1
1
I2 ¼ Imfu^ ES ðy Þg sinhðac Þdy ð19bÞ
a 1
Z 1
1  
I3 ¼ Re u^ p ðy Þ sinhðc Þdy ð19cÞ Figure 1. Variation of the ratio of amplitudes

of mass flow rate
a 1 with and without streaming potential amp m SP =ampðm Þ with
Z 1 the phase f for various values of the ratio f1 ¼ Ex1
=ES1 .
1  
I4 ¼ Im u^ p ðy Þ sinhðac Þdy ð19dÞ ‘Amplitude’ means half the difference between the maximum
a 1 and the minimum values of the oscillating mass flow rate.

& 2011 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.electrophoresis-journal.com
Electrophoresis 2012, 33, 419–425 General 423

and the external electric field. This alteration of the ampli- the streaming potential dependent tuning of the externally
tude is of immediate significance to mass flow rate control applied electric field.
because it offers an extra degree of maneuverability to Figure 2 shows the variation of the ratio of the steady
tuning mechanisms. Indeed, it can be seen that the mass components of the mass flow rate with  and
 without
flow rate amplitude ratio shows a peak (or a trench) when F considering the streaming potential effect mSP =hm i with

is around p. These peaks (or trenches) show extremely the ratio f0 ¼ Ex0 =ES0 . The values of a and K are kept fixed
interesting trends with variation of the f1 ratio. Indeed, when at 5 and 5, respectively. The non-dimensional number R
f1r0.1, the amplitude ratio has a constant value just below 1 has no effect on this variation. It can be seen that the mass
throughout; when 10rf1r20, the amplitude ratio peaks flow rate ratio is a constant slightly below unity for low
decrease and show a very deep trench corresponding to values (o1) of the parameter f0, whereas the same approa-
f1 5 20; when f1Z23, the amplitude ratio shows peaks ches a value marginally greater than unity for high values
corresponding to fEp and decreases with increase in the (450) of f0. The physical justification for this is in line with
value of f1. For very high values of f1, the amplitude ratio the explanation for Fig. 1 for the amplitude ratio. When f0 is
settles down to a value of almost 1 for all values of f. We low, the ratio is below 1 because in this case, the effect of the
now look at the physical justification of these trends. When applied electric field is practically negligible. The flow takes
f1 magnitude is very small (E0.1) the effect of the applied place primarily due to the contribution of the applied pres-
electric field is practically negligible. In this case, the sure-gradient. Naturally, consideration of the streaming
amplitude ratio is, practically, just the ratio between the potential results in a slightly lower value of the mass flow
amplitudes of a pressure-gradient-driven flow with stream- rate which, in turn, results in the ratio being slightlyo1.
ing potential and of a pressure-gradient-driven flow without This is consistent with the explanation given in the context
streaming potential. Since streaming potential effects of Fig. 1. Again, as the magnitude of f0 increases, the
always inhibit the pressure-driven flow, the ratio is consis- contribution of the steady component of the applied electric
tently below 1 (as seen in Fig. 1) for this case. As the field increases. The steady component of the streaming
magnitude of f1 is increased, the contribution from the potential (arising in response to the steady component of the
applied electric field becomes progressively more important. pressure-gradient) reinforces this contribution of the applied
For relatively lower values of f1, the applied electric field is electric field, which results in a dip in the values of the ratio.
not yet the dominant driving force so that it basically acts However, beyond a critical value of f0, the contribution of the
only to modulate the primary pressure-driven flow. This is applied electric field becomes predominant. Of course, the
vindicated by the observation that deep trenches are streaming potential field continues to ‘‘reinforce’’ this
observed when the phase difference between the applied contribution of the applied electric field. But, as the domi-
electric field and the applied pressure-gradient is p nance of the applied electric field continues to increase,
(i.e. when the streaming potential and the applied electric the relative strength of this ‘‘reinforcement’’ continues
field are in phase). However, beyond a critical value of f1, the to dwindle which results in a tapering off of the peak.
contribution of the applied electric field becomes dominant Ultimately, the effect of the externally applied electric field
and it is now the pressure gradient contribution, which may totally outweighs the effect of the field due to the streaming
be viewed as a modulator. As the value of f1 is increased, the potential, so that there is virtually inconsequential influence
relative contribution of the pressure gradient (on which the of the streaming potential on the mass flow rate. This is
generation of the streaming potential is dependent) gradu-
ally decreases. Of course, just beyond the critical point, the
streaming potential continues to ‘‘reinforce’’ the effect of the
applied electric field. This is consistent with the behavior
below the critical value; the only difference being that the
same phenomenon is now manifested in the form of peaks
(instead of trenches) due to the dominance of the applied
electric field. As the relative contribution of the pressure-
gradient decreases more and more, the disparity between
the presence and absence of this ‘‘reinforcing’’ streaming
potential becomes less and less, which is immediately
manifested in the amplitude ratio tending to unity for high
values of f1. Physically, this means that for these particular
values of a, K and R, it is only in the relatively narrow
regime of 1rf1r50 that the streaming potential effects have
to be considered carefully and non-trivially, in tandem with
both the applied driving agents of pressure gradient and
electric field for the oscillating components of the mass flow Figure 2. Variation of the ratio of the steady components
 of the
rate. These hitherto uninvestigated results bring to light the mass flow rate with and without streaming potential m SP =hm i
sensitive dependencies of the mass flow rate oscillations on with the ratio f0 ¼ Ex 0 =ES0

.

& 2011 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.electrophoresis-journal.com
424 J. Chakraborty et al. Electrophoresis 2012, 33, 419–425

Figure 3. Variation in the

ratio of the amplitudes of the mass
flow rate oscillations amp m SP =ampðm Þ with the phase f for Figure 4. Variation of the ratio of the steady components
 of the
various values of the Dukhin number Du keeping a, K and R mass flow rates with and without streaming potential m SP =hm i
constant. with K for various values of Du.

reflected in the mass flow rate ratio value being close to

unity. Thus, Fig. 2 shows that even in the steady case and can be seen that with increase in the value of K, the ratio
consistent with the observations of Fig. 1, the mass flow rate tends to unity. This can be explained on the basis of the fact
has a very high sensitivity to the streaming potential influ- that as the relative thickness of the EDL compared to the
enced resultant electric field particularly in the regime channel height becomes less, the electrokinetic effects
1rf0r50. become negligible and the mass flow rates are dominated by
Figure 3 shows the variation in the ratio
ofthe ampli- the applied pressure gradient. Furthermore, it is also
tudes of the mass flow rate oscillations amp mSP =ampðm Þ observed that the ratio tends to unity with increase in the
with the phase difference f for various values of the Dukhin value of Du. The physical justification is that as the Dukhin
number Du, keeping a 5 4, K 5 5, R 5 0.1 and the number increases, the bulk mass flow rate response solely
frequency ratio 5 1 (such that R1 5 R2 5 R). The trenches are due to the streaming potential effect decreases because it is
manifested when the streaming potential effects are in phase the conductivity in the near-wall region, where the immobile
with the externally applied electric field. The variations in the Stern layer is located, that brings about the predominant
values of these trenches for different values of the Dukhin response to the back potential.
number can be understood from the fact that the Dukhin
number basically denotes the relative strength of the surface
conductivity to the bulk conductivity. For higher values of the 4 Concluding remarks
Dukhin number, there will be lesser tendency for the bulk
ions to respond to the back potential (the streaming potential) In the present study, we investigate the effect of the
and surface conduction will contribute more towards equili- streaming potential on the mass flow rate response under
bration, resulting in a reduced electroviscous effect. In the the combined driving influence of a pulsating pressure
present context, higher values of Dukhin number result in gradient and a pulsating externally applied electric field. In
lower mass flow rate in response to the streaming potential particular, we compare the amplitudes of the mass flow rate
(this component of the mass flow rate opposes that due to the with and without the influence of the streaming potential
purely pressure-driven component) and thus less augmenta- effects. We observe that the mass flow rate decreases with
tion of the trench values due to the externally applied electric consideration of the streaming potential effect when there is
field. It is important to understand that for values of the no phase difference between the applied pressure gradient
phase difference close to p, the flow component due to the and the applied electric field. We find sensitive dependen-
externally applied electric field opposes that due to the purely cies of the mass flow rate on the streaming potential
pressure-driven component, which also means that it is in dependent tuning of the applied electric field. We also
phase with the streaming potential-driven component of establish that the phase difference between the pressure
the flow. gradient pulsations and the externally applied electric field
Figure 4 shows the variations in the ratios of the mean pulsations can be strategically tuned in order to achieve
values of the mass flow rates, with and without streaming significant variations in the amplitudes of the mass flow rate
potential effects, with variations in the value of K for pulsations especially in the case of thick EDLs. These
different values of the Dukhin number (Du) with a 5 4. It amplitudes also show interesting trends with increase in the

& 2011 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.electrophoresis-journal.com
Electrophoresis 2012, 33, 419–425 General 425

value of the oscillation frequencies due to the combined [10] Yang, J., Masliyah, J. H., Kwok, D., Langmuir 2004, 20,
effects of decreasing streaming potential amplitudes and 3863–3871.
variations in the phase difference between the applied [11] Marcos, Yang, C., Ooi, K. T., Wong, T. N., Masliyah,
pressure gradient and the induced streaming potential field. J. H., J. Colloid Interface Sci. 2004, 275, 679–698.
We also find that for lower values of the Dukhin number [12] Luo, W.-J., J. Colloid Interface Sci. 2004, 278, 497–507.
corresponding to the case where the bulk ionic conductivity [13] Campisi, M., Accoto, D., Paolo, D., J. Chem. Phys. 2005,
predominates over the near-wall conductivity, the streaming 123, 204724.
potential dependent variations in the mass flow rate [14] Xuan, X., Li, D., J. Colloid Interface Sci. 2005, 289,
amplitudes are noticeably higher. Furthermore, we observe 291–303.
that a decrease in the relative thickness of the EDL [15] Marcos, Kang, Y. J., Ooi, K. T., Yang, C., Wong, T. N.,
compared to the channel height results in an expected J. Micromech. Microeng. 2005, 15, 301.
decrease in the mass flow rate due to reduced significance [16] Mishchuk, N. A., Gonzalez-Caballero, F., Electrophoresis
of the electrokinetic effects; however, the interesting 2006, 27, 650–661; ibid. 661–671.
interplay between the phase difference of the induced [17] Huang, H. -F., Lai, C. -L., Proc. R. Soc. A 2006, 462,
streaming potential and the tuned phase difference of the 2017–2038.
externally applied electric field brings about non-trivial [18] Zhang, Y., Wong, T. N., Yang, C., Ooi, K. T., Microfluid.
implications in the amplitude response. These interesting Nanofluid. 2006, 2, 205–214.
results are significant from the practical viewpoint of [19] Chakraborty, S., Srivastava, A. K., Langmuir 2007, 23,
achieving augmented mass flow rates within specific 12421–12428.
regimes of operations. These also indicate that even [20] Wang, X., Chen, B., Wu, J., Phys. Fluids 2007, 19,
in theoretical frameworks, the effect of streaming 127101.
potential effects under the combined influence of a pressure [21] Bhattacharyya, S., Nayak, A. K., Colloids Surf. A 2008,
gradient and an electric field cannot be trivially overlooked, 325, 152–159.
especially when the driving influences are time-periodic in [22] Talapatra, S., Chakraborty, S., Eur. J. Mech – B/Fluids
nature. 2008, 27, 297–308.
[23] Kuo, C.-Y., Wang, C.-Y., Chang, C.-C., Electrophoresis
The authors have declared no conflict of interest. 2008, 29, 4386–4390.
[24] Chakraborty, S., Ray, S., Phys. Fluids 2008, 20, 083602.
[25] Gong, L., Wu, J., Wang, L., Cao, K., Phys. Fluids 2008,
20, 063603.
5 References
[26] Chang, C. C., Wang, C.-Y., Phys. Fluids 2009, 21, 042002.

[1] Li, D., Encyclopedia of Microfluidics and Nanofluidics, [27] Goswami, P., Chakraborty, S., Langmuir 2010, 26,
Springer, New York 2008. 581–590.

[2] Squires, T. M., Quake, S. R., Rev. Mod. Phys. 2005, 77, [28] Jian, Y., Yang, L., Liu, Q., Phys. Fluids 2010, 22, 042001.
977–1026. [29] Srivastava, A., Chakraborty, S., Int. J. Adv. Eng. Sci.
Appl. Math. 2010, 2, 61–73.
[3] Schoch, R., Han, J., Renaud, P., Rev. Mod. Phys. 2008,
80, 839–883. [30] Zhao, C., Yang, C., Electrophoresis 2011, 32, 629–637.
[4] Dutta, P., Beskok, A., Anal. Chem. 2001, 73, 1979–1986. [31] Zhao, C., Yang, C., Phys. Rev. E 2011, 83, 066304.
[5] Dutta, P., Beskok, A., Anal. Chem. 2001, 73, [32] Haiwang, L., Wong, T. N., Nguyen, N. -T., Int. J. Heat
5097–5102. Mass Transfer 2010, 53, 772–785.
[6] Reppert, P., Morgan, F. D., J. Colloid Interface Sci. 2003, [33] Hunter, R. J., Zeta Potential in Colloid Science,
254, 372–383. Academic Press, San Diego 1981.
[7] Bhattacharyya, A., Masliyah, J. H., Yang, J., J. Colloid [34] Chakraborty, S., Das, S., Phys. Rev. E 2008, 77, 037303.
Interface Sci. 2003, 261, 12–20. [35] Chakraborty, J., Chakraborty, S., Phys. Fluids 2010, 22,
[8] Yang, J., Bhattacharyya, A., Masliyah, J. H., Kwok, D. Y., 122002.
J. Colloid Interface Sci. 2003, 261, 21–31. [36] Chakraborty, J., Chakraborty, S., Phys. Fluids 2011, 23,
[9] Erickson, D., Li, D., Langmuir 2003, 19, 5421–5430. 082004.

& 2011 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.electrophoresis-journal.com