ARTICLE

Received 16 May 2011 | Accepted 5 Aug 2011 | Published 6 Sep 2011 DOI: 10.1038/ncomms1471

Microfluidic quadrupole and floating

concentration gradient
Mohammad A. Qasaimeh1,2, Thomas Gervais3 & David Juncker1,2,4

The concept of fluidic multipoles, in analogy to electrostatics, has long been known as a
particular class of solutions of the Navier-Stokes equation in potential flows; however,
experimental observations of fluidic multipoles and of their characteristics have not been
reported yet. Here we present a two-dimensional microfluidic quadrupole and a theoretical
analysis consistent with the experimental observations. The microfluidic quadrupole was
formed by simultaneously injecting and aspirating fluids from two pairs of opposing apertures
in a narrow gap formed between a microfluidic probe and a substrate. A stagnation point was
formed at the centre of the microfluidic quadrupole, and its position could be rapidly adjusted
hydrodynamically. Following the injection of a solute through one of the poles, a stationary,
tunable, and movable—that is, ‘floating’—concentration gradient was formed at the stagnation
point. Our results lay the foundation for future combined experimental and theoretical
exploration of microfluidic planar multipoles including convective–diffusive phenomena.

1
Biomedical Engineering Department, McGill University, Montréal, Quebec, H3A 1A4 Canada. 2 Genome Quebec Innovation Centre, McGill University,
Montréal, Quebec, H3A 1A4 Canada. 3 Department of Engineering Physics, École Polytechnique de Montréal, Montreal, Quebec, H3C 3A7 Canada.
4
Department of Neurology and Neurosurgery, McGill University, Montréal, Quebec, H3A 1A4 Canada. Correspondence and requests for materials should
be addressed to D.J. (email: [email protected]).

nature communications | 2:464 | DOI: 10.1038/ncomms1471 | www.nature.com/naturecommunications 

© 2011 Macmillan Publishers Limited. All rights reserved.
ARTICLE nature communications | DOI: 10.1038/ncomms1471

Q
uadrupoles are used in numerous physical and engineering a b
applications and arise under many forms, such as electri-
cal charges and currents1, magnetic poles2, acoustic poles3,4,
and gravitational masses5. Two popular applications of quadrupoles
include the quadrupole mass spectrometer that uses four parallel
– + – +
metal rods with opposing alternating electrical currents to filter
ions based on their mass-to-charge ratio6, and the quadrupole
magnets that are used to focus beams of charged particles in par- + – + –
ticle accelerators2. In fluid mechanics, flow dipoles, or ‘doublets’
have been studied in porous rocks notably in the context of oil
extraction7,8. To increase the recovery of oil pumped from a well, a
solution can be injected in another well to displace the oil trapped
within the rocks, according to a dipolar flow profile. We recently
c e f
introduced the microfluidic probe (MFP)9 that uses both injec- Stagnation point
Y
tion and aspiration aperture to flush a stream across a substrate (SP)

′
Y

X
Asp Inj
surface, and that in fact represents a two-dimensional fluidic QA QI QA
dipole, although this analogy has not been developed. Planar SP
multipoles with more than two flow poles have been investigated Asp
Inj
theoretically by Koplik et al. with increasing numbers of sources QI QA
QA

and sinks (such as quadrupoles, octopoles etc.)10. Bazant et al. have

Y
X
SP Y′

′
also demonstrated the use of conformal mapping to solve a broad d QI QI MFP
Immersion
X X′
class of advection–diffusion problems in 2D irrotational flows11 by medium
Gap
drawing an analogy with the pioneering work of Burgers on out- Transparent substrate
of-plane flow velocity in 2D vortex sheets12. However, experimental
validation of these theoretical analyses is still outstanding as flow Figure 1 | Two-dimensional MQ formed in gap between a MFP and
multipoles have not, to the best of our knowledge, been produced a substrate. (a) Schematics of the field lines in a 2D electrostatic
experimentally. quadrupole consisting of two pairs of opposing positive and negative
In this work, we present an experimental planar microfluidic charges. (b) Experimental image of a MQ when formed by injecting
quadrupole (MQ) in an open space confined between two paral- and aspirating liquids with tracer beads via two injections (plus sign)
lel plates (Fig. 1). We extend the theoretical framework to describe and two aspirations (minus sign) apertures. Path lines were captured
hydrodynamic and convective–diffusive properties of the MQ in by the long-term exposure of the flowing fluorescent tracer beads.
Hele–Shaw flows and derive the equations for the key parameters Whereas we adopt the same symbolism than for electrical charges,
characteristic of the MQ such as the size of the flow confinement the ‘ + ’ and ‘ − ’ signs correspond to purely fluidic sources and sinks.
and length of the gradient formed within the MQ. Finite element Scale bar is 400 µm. (c) Schematic of a MQ generated between
simulations complete the analytical results and both are compared two parallel plates formed by a MFP with four apertures on top and
with experimental results. Finally, we present a first application of a transparent substrate at the bottom. The flow lines between the
the MQ that can be used as a gradient generator and which, unlike apertures represent a scenario where the aspiration flow rates are larger
prior gradient generators, affords low shear stress, rapid spatiotem- than the injection flow rates and thus all the injected liquid is confined
poral tuning of the gradient either hydrodynamically by adjusting and re-aspirated. A SP is formed at the centre. (d) X–X′ and (e) Y–Y′
the flow rates, or physically by moving the MFP, and can be applied cross-sectional views of the MQ under the MFP (not to scale). (f)
on any planar substrate and displaced as well. Fluorescence micrograph showing path lines in a MQ using 2 µm red
and green tracer microbeads injected (Qinj = 10 nl s − 1) through the top
Results right and bottom left apertures, respectively, into a 50 µm gap filled
Microfluidic quadrupole. MQs were formed between a MFP9 with liquid, and aspirated back (Qasp = 100 nl s − 1) through the other two
with 4 apertures arranged in a classical symmetric quadrupole apertures. Scale bar is 200 µm.
configuration (Fig. 1b) and a flat bottom substrate. The MFP was
microfabricated into Si and diced into squares with four openings
constituting pairs of opposing apertures used for injection and
aspiration. A Polydimethylsiloxane (PDMS) adapter block was (drains) (Supplementary Figure S1; Supplementary Methods).
bonded to the die, and capillaries were plugged into the block and The MFP used here featured a centre-to-centre separation of
connected to syringe pumps (Fig. 2a,b). 1,075 µm between pairs of opposing apertures, each 360 µm in
The MFP was immersed in fluid and brought close to a planar diameter, and was positioned above, parallel to the bottom substrate
substrate so as to form a narrow gap (see Methods). A glass slide so as to form a 50 µm gap for all experiments. This setup represents
was used as the bottom substrate permitting visualization of the a parallel-plates configuration, and the MQ formed in the gap is
flow within the gap using an inverted microscope (Fig. 2c). Simul- quasi-two-dimensional. Using the Hele–Shaw approximation, it can
taneous injection and aspiration of fluid generates the quadrupo- thus be conveniently described as a two-dimensional Stokes flow of
lar flow, and the two injected fluids meet head-on at the centre of the form13:
the MQ, generating a stagnation point (SP) with zero flow veloc-
 H2 z  z 
ity (Fig. 1c–f). When the aspiration and injection flow rates are n H − S (x , y , z ) = − 1 −  ∇p(x , y )
identical, a large fraction of the injected stream is re-aspirated; 2h H  H (1)
but the fraction of fluid that radially flows outward is not (Fig. 1b)

and thus leaks into the fluid surrounding the MFP. However, if the where n H − S is the velocity field of the MQ, η is the viscosity,
ratio of aspiration/injection flow rates is > 1, all the injected liquid is p is the hydrostatic pressure, H is the height of the gap between the
captured and aspirated back into the MFP (Fig. 1f). plates (H = 50 µm), and z is the vertical coordinate whereas z = 0 cor-
We arbitrarily define the two injection apertures as positive responds to the bottom plate (substrate) and z = H to the top plate
poles (sources) and the aspiration apertures as negative poles (MFP). The resulting flow velocities in the MQ are on the order
 nature communications | 2:464 | DOI: 10.1038/ncomms1471 | www.nature.com/naturecommunications
© 2011 Macmillan Publishers Limited. All rights reserved.
nature communications | DOI: 10.1038/ncomms1471 ARTICLE
of the MQ by using the superposition principle as in the case of
a c
electrical charges and current quadrupoles (see Supplementary
PDMS chip ­Methods):
4 4  
Glass capillaries   Qi (r − r ′i ) (3)
nave = ∑ ni = ∑   2
i =1 i =1 2p H | r − r ′ i |
MFP
Gap 

3 mm
Silicon chip where r is the position vector and r ′i is the position of the sources
Immersion
medium
and drains, both with respect to the centre of the probe. Qi is the
b Transparent substrate
flow rate from the ith inlet corresponding to Qinj and Qasp for injected
and aspirated flow rate, respectively.
z
Glass capillaries
x Microscope Flow confinements of the MQ. As mentioned earlier, the flow rates
Clamping rod
y
ratio Qasp/Qinj needs to be sufficiently high to hydrodynamically
confine and capture all the injected fluid (Fig. 1b versus Fig. 1f).
MFP
The confinement of the injected streams as function of the flow
rates ratio were visualized using fluorescent tracer particles
Figure 2 | Experimental MFP apparatus. (a) A 3D illustration of the MFP (Fig. 3a–c). The area flushed by the injected streams is an important
along with a micrograph shown in the inset. (b) A MFP connected to the concept for fluidic applications, because chemicals injected in the
syringe pumps using capillaries and fixed within the clamping rod and streams may be used to process selectively the underlying surface9,
mounted on a XYZ micropositioner that together form the MFP holder. and we thus establish an analytical solution. The fluid that radiates
(c) Using the MFP holder (not shown), the MFP is aligned parallel to the outwards, relative to the centre of the MQ, will travel the greatest
substrate and positioned above it with micrometer accuracy thus forming distance until the flow velocity becomes zero, at which point the
a microscopic gap. The MFP is immersed in a liquid that thus fills the gap. flow direction is reversed, and it is eventually recaptured by an aspi-
The bottom substrate, usually a glass slide or a cover slip, is placed on ration aperture. The confinement radius R of the MQ was defined as
the motorized stage of a microscope (not shown) and can be moved with the distance between the SP and the outermost point of zero velocity
respect to the MFP (and the objective) using a joystick or controlled using for the radial streams, and using equation (3) yields (Supplementary
a computer program. Schematics are not to scale. Methods):

d Qasp / Qinj + 1 (4)

R=
2 Qasp / Qinj − 1
Table 1 | Fundamental equations and equivalences for
electrical charge (electrostatics), electrical current
(electrokinetics), and fluid velocity (fluid mechanics) under where Qasp is the aspiration flow rate, Qinj is the injection flow rate,
the Hele–Shaw approximation. and d is the centre-to-centre distance between the two inlets. We
found excellent agreement between the calculated and experimen-
Electrostatics Electrokinetics Fluid mechanics tally measured R (Fig. 3). These results confirm the validity of our
(charge q) (current I) (flow rate Q) theoretical assumptions and analysis.
Potential Electrostatic Electrostatic Velocity potential
(scalar) potential φ potential φ (x, y) Fluid flow around the SP. A two-variable (x,y) Taylor expansion of
H2 p(x , y)
(x, y) f( x , y ) = − equation (3) yields the velocity profile in the very close vicinity of
12h
the SP (valid for x,y << d), see Supplementary Methods for details:
Field Electric field Current
  Density Velocity profile  8〈Q 〉
(vector)   J = −s ∇f nave = − (xxˆ − yyˆ )
E = −∇f   H2  p Hd 2 (5)
σ: electrical nave = ∇f = − ∇p
12h
conductivity where 〈Q〉 = (Qasp + Qinj)/2 is the average flow rate between the inlet
Gauss’s   q   dq   and outlet, and the origin (0,0) corresponds to the position of the
theorem ∫ E ⋅ dS = e0 ∫ J ⋅ dS = dt = I ∫ n ⋅ dS = Q
S S S SP. The flow profile can also be expressed in the form of a velocity
(continuity potential11, (Table 1):
equation)
4〈Q 〉 2
f(x , y ) = − (x − y 2 )
p Hd 2 (6)
of 2 mm s or less for aspiration flow rates of 100 nl s and the flow
−1 −1 Interestingly, this velocity potential is identical to the one describing
is purely viscous (Re  1), Supplementary Methods for details. two-dimensional flow along a right-angle corner14. In the MQ, the
Averaging equation (1) over the gap height, the velocity profile can four quadrants defined by the symmetry lines XX′ and YY′ in
be simplified as: Figure 1d each constitute a virtual corner with a corresponding
corner flow. Path lines formed by green tracer beads and a ratio
H Qasp/Qinj = 10 outline the SP at the centre (Fig. 1f; Supplementary
 1  H2  (2)
nave (x , y ) =
H ∫ n H −Sdz = − 12h ∇p(x, y) Movie 1), and are consistent with the velocity profile and stream-
lines obtained using FEM simulations (Fig. 4a). As the flow velocity
0
tends to zero when approaching the SP (Fig. 4b), the calculated wall
The height-averaged velocity profile given by this expression is shear stress was found to converge to zero as well, Supplementary
irrotational13, and can therefore be represented by a scalar poten- Figure S2a. The shear stress increases linearly when moving away
tial akin to an electric field or a current distribution that can also from the SP and reaches a maxima near the inner edge of each aspi-
be described based on an electrostatic potential (Table 1). It thus ration aperture. Within a circle with a diameter of 100 µm (~10%
becomes possible to calculate the flow velocity for any coordinate of the distance d) centred at the SP, shear stresses are less than

nature communications | 2:464 | DOI: 10.1038/ncomms1471 | www.nature.com/naturecommunications 

© 2011 Macmillan Publishers Limited. All rights reserved.
ARTICLE nature communications | DOI: 10.1038/ncomms1471

a d a b
2.5
2R
1.5

0.5
d
0
mm s–1
c 0.25
b e Outer edges of the aspiration apertures

0.2 Inner edges of the

aspiration apertures

Shear stress (Pa)

0.15

0.1
Stagnation
point
0.05
c f

0
–1,000 –500 0 500 1,000
Distance along the line Y–Y′ (µm)

Figure 4 | Fluid flow and shear stress in the MQ. FEM simulation of the MQ
with 100 and 10 nl s − 1 aspiration and injection flow rates, respectively.
(a) Streamlines revealing the SP at the centre. (b) Flow velocity profile at the
mid-plane (z = 25 µm) in the MQ revealing the zero fluid flow at the SP and
the low flow in the areas surrounding the SP as well as at the centre of the
g 3,500 poles. (c) Total shear stress profile at the substrate along a line connecting
the two aspiration apertures and crossing the SP (line Y–Y′ shown in
Supplementary Fig. S2a). The shear stress between the four apertures
3,000 increases almost linearly from zero at the SP to a first maxima at the inner
edge of each aspiration aperture (τ = 0.15 Pa), drops back to almost zero at
the centre, and reaches the maxima at the outer edge of the aperture.
2,500
2R (µm)

1 + Qasp / Qinj
2R = d
2,000 1 – Qasp / Qinj
is τmax = 0.15 Pa, and hence < 0.015 Pa inside the circle. The shear
stress also varies linearly with the aspiration flow rates that may
1,500 be reduced to further minimize stress (Supplementary Fig. S2b;
Supplementary Methods).
1,000 Convection–diffusion and gradients at the SP. When a solute is
1 2 3 4 5 6 7 8 9 10
Flow ratio (Q asp /Q inj) added to one of the injected streams, diffusive mass transport grad-
ually dominates when approaching the SP as the convective flow
Figure 3 | Flow confinement in a MQ for varying ratio of aspiration to converges towards zero. The solute is continuously replenished by
injection flow rates. 2 µm fluorescent tracer microbeads were injected at one stream acting as a source, while it is also continuously trans-
50 nl s − 1 through the top right and bottom left apertures ( + ) and aspirated ported away by the other stream that acts as a sink. Along the axis of
back through the other two apertures ( − ). For all experimental data, the two injection apertures (XX′), the solute diffuses into the oppos-
fluorescent paths of microbeads revealing the streamlines were captured ing stream at the SP while it is simultaneously pushed back by the
with a 1 s exposure. (a–c) Flow pattern for a ratio Qasp/Qinj = 1.75, 2, and inward convection of the opposing stream, and a steady-state, sta-
4, respectively. (d–f) Streamlines obtained by FEM simulations for the tionary concentration gradient is formed. To characterize this gra-
flow ratios used in (a–c), respectively. (g) Comparison of theoretical and dient mathematically, we first consider the general time-dependent
experimental (shown as red dots) 2R as a function of the flow ratio reveals diffusion–convection equation15:
a good agreement. Error bars represent the standard error of four separate
∂C   
experiments. Scale bar is 400 µm. = D∇2C − (∇f ) ⋅∇C (7)
∂t 

10% of that local maxima as shown in Figure 4c (Supplementary where D is the diffusivity of the species studied, n = ∇f is the
Methods). For a MQ with injection and aspiration flow rates of height-averaged in plane velocity profile and C is the concentra-
10 and 100 nl s − 1, respectively, the local maxima of shear stresses tion of the solute (normalized between 0 and 1). The steady-state

 nature communications | 2:464 | DOI: 10.1038/ncomms1471 | www.nature.com/naturecommunications

© 2011 Macmillan Publishers Limited. All rights reserved.
nature communications | DOI: 10.1038/ncomms1471 ARTICLE

a b 1 velocity profile around the SP of equation (5), and combining it with

Y X′ 0.9
Line a–b
equation (7), the convection–diffusion equation relevant to the flow
0.8 around the SP formed within the MQ can be re-written as:

Normalized intensity (l/l0)

Line c–d
0.7 Line e–f
Theoretical ∂C ( x , t ) ∂2C(x , t ) 8〈Q 〉 ∂C(x , t )
0.6
=D + x (8)
0.5 ∂t ∂x 2 p Hd 2 ∂x
0.4
X Y′ 0.3 where D is the diffusion constant of the solute. The derivatives
0.2 with respect to y have been neglected in equations (7) and (8) as
c a e
0.1 the concentration interface is invariant in y (∂C/∂y = 0 everywhere)
d b f 0 under Hele–Shaw assumption; see Supplementary Methods for
–80 –60 –40 –20 0 20 40 60 80
Position across gradient (µm) details. Using the dimensionless numbers x = x / x0 = x Pé / d ,
t = t / t0 = Pé ⋅ Dt / d 2 , where Pé = 8〈Q 〉/p HD is the global Péclet
c d e f number for the MQ, we find the simple dimensionless expression
for this transport problem to be:

∂C(x, t ) ∂2C(x, t ) ∂C(x, t )

= + x
∂t ∂x2 ∂x (9)

Therefore x0 = d / Pé and t0 = d 2/(Pé ⋅ D) are the natural scales of

g 1 the MQ. They represent, respectively, the length over which diffu-
90
0.9 QA = 20 nl s–1 sion takes place at the interface and the characteristic time before
Normalized intensity (l /l0)

QA = 50 nl s–1
80
0.8
QA = 100 nl s–1 steady state is achieved. Typical conditions in our experiments were
0.7
0.6
QA = 150 nl s–1 an average flow rate of 55 nl s − 1, and D = 500 µm2 s − 1 as we used
QA = 200 nl s–1
Gradient length (µm)

70
0.5 fluorescein as the diffusive species. Using these values, we obtain
60
0.4 x0 = 15 µm and t0 = 0.4 s. Thus, after a few seconds, the gradient will
0.3 have reached a steady –state, and the time-dependent term of equa-
0.2
50 tion (9) can be neglected. When performing this approximation,
0.1
0 equation (9) reduces to a one-dimensional problem with C(x,t)
40 –80 –60 –40 –20 0 20 40 60 80 being the concentration profile at the SP perpendicularly to the
HD Position across gradient (µm)
30
Lgrad = 1,606.d
〈Q 〉 interface where diffusion takes place.
For the experimental conditions used here, the concentration
20
0 25 50 75 100 125 150 175 200 225 250 275 300 gradients at the SP were much shorter than the distance d separat-
Aspiration flow rate (nl s–1) ing the two inlets, d may thus be considered infinite and thus the
boundary conditions become:
Figure 5 | Tunable concentration gradients formed at the SP of a MQ.
(a) 50 and 10 nl s − 1 were used as aspiration (Qasp) and injection (Qinj) C(∞) = 1, C(−∞) = 0 (10)
flow rates, respectively, and fluorescein was added to the solution
injected through the top right aperture. A close-up view of the gradient This allows simplifying the calculations and solving equations (9)
corresponding to the white rectangle is shown as inset. (b) Fluorescence and (10) for the concentration under these conditions gives:
intensity profiles of the gradients across the lines a–b, c–d, and e–f in
(a) show good agreement between theory and experiment and that the x −  x2
1 1  Pé x  
gradient shape does not change along the Y–Y′ axis. (c–f) Images of the
fluorescein distribution and concentration gradients within the MQ for
C(x) = ∫
2p −∞
e 2 dx =  1 + erf 
2  2 d  
(11)

an injection flow rate Qinj = 10 nl/s and for a flow rate ratio Qasp/Qinj = 2, 10,
15, and 20, respectively (insets are close-up views). Scale bar is 200 µm where erf is the error function17. Once again, the solution here is
in the overviews and 100 µm in the close-up views. (g) The measured similar to that of the out-of-plane velocity in a Burgers vortex sheet,
length of the gradient with different aspiration flow rates are shown as red but with the Péclet number replacing the Reynolds number of the
dots. Error bars represent the standard error of five separate experiments. classical analysis16. The gradient length can be calculated using
The analytical solution is shown as a dashed line and shows excellent equation (11) and expressed as:
agreement with the experiment. Fluorescence intensity profiles along the
central line across each gradient for different aspiration flow rates are d2 HD
shown in the inset. Lgrad = 2, 563 = 1, 606 ⋅ d (12)
Pé 〈Q 〉
The gradient length is proportional to the average flow rate and
obeys a simple power law.
version of equation (7) (with the left-hand term set to zero) is The analytical solution for the gradient length was compared
reminiscent of the flow in a ‘Burgers vortex sheet’12 where C is with experimental measurements using fluorescein sodium salts
analogous to the out of plane velocity, as was pointed out by Bazant (376 Da, diffusion coefficient in water D = 500 µm2 s − 1 (refs 18–20))
and Moffatt16. diluted in water. Fluorescein was injected through the top-right
On the basis of the experimental observations shown below, the aperture, and the gradient formation at the SP within the MQ was
gradient length, which here is defined as the distance between the modelled numerically (Supplementary Fig. S3a), and measured
points with a concentration of 10 and 90% of the original solute experimentally, (Fig. 5a). Theoretical and experiment gradient
concentration, is small when compared with the distance between profiles are in excellent agreement, (Fig. 5b; Supplementary
the inlets along the XX′ axis (Fig. 5a). Using the approximated Figure S3b).

nature communications | 2:464 | DOI: 10.1038/ncomms1471 | www.nature.com/naturecommunications 

© 2011 Macmillan Publishers Limited. All rights reserved.
ARTICLE nature communications | DOI: 10.1038/ncomms1471

a b c d

QA1 QI1
y x

QI2 QA2

Figure 6 | Hydrodynamic positioning of the SP. The SP and the gradient floating between the streams were tuned by independently adjusting the flow
rate of each pole of the MQ. (a) Streamlines of the flow showing the SP at the centre (upper) of the MQ for QI1 = QI2 = 10 nl s − 1 and QA1 = QA2 = 100 nl s − 1,
and shifted in xy-plane (lower) for QI1 = 10, QI2 = 70, QA1 = 170, and QA2 = 100. (b) Fluorescence micrograph of green and red tracer microbeads revealing the
pathlines following a 2 s exposure. Position of the SP (indicated by the white arrow) can be deduced from the diverging path lines. In the lower micrograph,
the SP moved from the centre towards the top-right and top-left apertures by ~97 and ~102 µm, respectively. (c) FEM modelization and (d) experimental
results showing the asymmetric flow confinement and the gradient when the interface is straight (upper) and curved (lower) as a result of moving the SP.
The gradient can be moved and adjusted rapidly as illustrated in the Supplementary Movie 4.

We observed that the gradient was constant along the YY′ axis of the MQ. The gradient is formed along a straight interface when
connecting the two aspiration apertures(Fig. 5a,b). The flow per- the SP is centred, and along a curved interface when moving the SP
pendicular to the gradient carries it towards each of the aspiration towards any of the injection apertures (Fig. 6c,d). By using a pro-
apertures, and convection rapidly dominates diffusion when mov- grammable flow control system, or by manually changing the flow
ing away from the SP (see Supplementary Methods for details) rates, oscillating and rapidly moving gradients can be generated
while the concentric flow around the aspiration apertures focuses (Supplementary Movie 4). Complex spatiotemporal gradient land-
the stream when approaching the inlet. Collectively, these effects scapes may thus be formed around a particular point by changing
counter diffusive broadening of the gradient and explain the both the gradient slope and the gradient position. The gradient may
quasi-constant gradient profile along the YY′ line connecting both be adjusted either in a preprogrammed manner, or in real time in
aspiration apertures. response to observations.
We compared the calculated gradient length with experimen- Alternatively, for scanning larger areas and reach positions
tal measurements for a Qasp/Qinj, varying from 2 to 15 (Fig. 5c–g), outside the MQ’s central area, the gradient at a point may also be
and found them to be in good agreement. For Qasp increasing from changed by displacing the substrate relative to the MFP, either by
20 to 250 nl s − 1, the gradient length decreased from 69 to 25.7 µm. moving the bottom substrate that is clamped to the motorized xy-
This range of accessible gradient slopes may be further increased stage of the microscope, or by manually moving the MFP using the
by exploring a wider range of flow rates from pl s − 1 to µl s − 1. The micrometer screws of the probe holder (Fig. 2c)21. The MQ field lines
gradient changes rapidly (Supplementary Movie 2), and, therefore, and the concentration gradient are temporarily perturbed while the
dynamic gradients may readily be formed simply by pre-program- MFP moves relative to the substrate proportionally to the speed of
ming the aspiration flow rates. Whereas the confinement of the MQ displacement. For example, a speed of 300 µm s − 1 lead to a small but
is governed by Qasp/Qinj, equation (12) indicates that the gradient visible disturbance of the gradient (Supplementary Movie 5). The
slope and length depend on 〈Q〉 and, thus, primarily on Qasp for high results indicate that the gradient re-equilibrated within few seconds
ratios of Qasp/Qinj. This theoretical result is supported by the experi- after stopping the movement. Faster movements are possible but at
ments where only Qasp was changed, Figure 5, or where only Qinj was the expense of greater disturbance of the gradient. Supplementary
tuned (Supplementary Figure S4; Supplementary Movie 3). It is thus Movie 6 shows the gradient while moving the substrate at speed of
possible to vary the confinement and the gradient slope independ- 7 mm s − 1, and Supplementary Movie 7 shows the gradient while the
ently by keeping one of them constant by simultaneously tuning MFP is being manually displaced. Using these two approaches, dif-
Qasp and Qinj. ferent areas of the substrates can be exposed to the gradient one
after the other.
Asymmetric MQ flow profiles and floating gradient. For all
experiments described up to this point, each pair of injection and Discussion
of aspiration flow rates were adjusted in synchrony and kept equal, We have presented an experimental demonstration of fluidic quad-
which fixed the SP at the centre of the MQ. If the flow rate of only rupoles and found excellent agreement between theoretical predic-
one of the two injection or aspiration apertures is changed, the sym- tion and experimental results, thus lending support to the general
metry is broken and the position of the SP will change (Fig. 6a,b). analytical framework developed to describe multipolar flow. Using
The gradient, which originates at the SP, moves along with the SP the Hele–Shaw formalism, we established a close analogy between
and can thus be displaced hydrodynamically within the central area MQs, electrical quadrupoles, and vortex sheets in fluid dynamics.

 nature communications | 2:464 | DOI: 10.1038/ncomms1471 | www.nature.com/naturecommunications

© 2011 Macmillan Publishers Limited. All rights reserved.
nature communications | DOI: 10.1038/ncomms1471 ARTICLE

Our analysis was extended to include phenomena that were not fluidic connection holes. The PDMS was cured in an oven at 60 °C for at least 3 h
studied before and the governing equations for a Hele–Shaw geom- (usually over night). The PDMS block was bonded to the Si chip by activating both
parts in air plasma (Plasmaline 415 Plasma Asher, Tegal Corporation), at 0.2 mbar
etry derived for flow confinement and for the convection–diffusion for 45 s at 75 W, and joining the two together using a home-made mechanical
of solutes at the SP. In analogy to electrical poles, and as was devel- alignment aid and placed in an oven at 90 °C for 20 min. Operation of this MFP is
oped theoretically11,16, the number of poles may be further expanded similar to the use of the original MFP9,21,44, except that four independent syringe
and inspire combined theoretical and experimental studies of con- pumps are required to control the flow rate of each aperture independently. Briefly,
vection–diffusion phenomena with multiple poles and multiple the MFP was secured inside the clamping rode (Fig. 2b) and mounted on the xyz
micropositioner (7600-XYZL, Siskiyou), what we call the probe holder. Using the
chemicals. probe holder, the MFP was positioned parallel to the transparent substrate (glass
We demonstrated a first application of MQs as concentration gra- slide) atop an inverted microscope (TE2000, Nikon) so as to form a microscopic
dient generators. The gradient is generated around the SP that expe- gap while immersed under the surrounding medium (Fig. 2c). Glass capillaries
riences no flow and no shear stresses similar to stationary source- (Polymicro Tech) connect the MFP apertures to glass syringes (Hamilton) that are
sink gradients22,23, while being continuously fed by the streams operated using a computer-controlled syringe pumps (Nemesys Cetoni).
of the MQ and thus formed within seconds and could be rapidly Numerical simulations. The three-dimensional simulations were carried out using
adjusted, akin to microfluidic gradients with active flow24–27. Several the commercially available finite element simulation software Comsol Multiphysics
approaches for forming gradients using hybrid approaches combin- 3.5 (Comsol) and run on an eight-core, 64-bit computer (Xeon, Dell) with 26 GB
ing flow and no-flow zones have been proposed, but required porous of RAM. The MFP geometry was modelled with the same shape and dimensions
elements28,29 or nanochannels30,31. Gradients in open chambers as the microfabricated MFP used in the experiments. Simulations coupled the
solution of Navier–Stokes equation and convection–diffusion equation. Sample
have also been proposed32,33 to allow user access (that is, pipetting) solutions were assumed to be water (incompressible Newtonian fluid with a density
to the cell culture. These devices provide low shear stress cultures, of 998.2 kg m−3, and a dynamic viscosity of 0.001 N·s m−2), and the diffusion coef-
while allowing to adjust the concentration gradients. However, ficient of the solute was 500 µm2 s − 1, which corresponds to the diffusion constant of
these devices entail either closed chambers, or complicated fabri- fluorescein in water. The simulations were run under steady-state conditions and
assumed no-slip boundary conditions on the substrate and the MFP surface, with
cation processes, or both, and the time to adjust the gradient was
the flow boundary conditions at the MQ’s perimeter sides set as open boundaries
typically longer than with the MQ. Another recent technique34 used (equal to atmospheric pressure). As boundary conditions for the mass transfer
a laser beam to open and close pores in an integrated impermeable equation, the MFP surface and substrate were defined as an insulating boundary,
membrane to expose the cell culture to reagents selectively; how- and a zero inward flux was set at the perimeter of the MQ model. The experimental
ever, this device involves closed cell culture chamber and a sophisti- values were used for injection and aspiration flow rates. The concentration of the
solute at the source injection aperture was arbitrarily set to 1 mol m−3, 0 at the other
cated experimental setup. injection aperture, and convective flux was set for both aspiration apertures.
The floating gradients formed within the MQ are remarkable
because of a series of individual characteristics, and owing to the Image acquisition and analysis. Images were recorded using a cooled CCD
properties and possibilities that arise by the combination of these camera (Photometrics CoolSNAP HQ2) connected to the microscope. Black-and-
characteristics. They include, being open and brought down onto white images were analysed and coloured using the software ImageJ 1.42 (National
Institutes of Health Maryland). Images of green and red beads pathlines were
the target areas, the combination of active flow and shear-free zone joined together digitally using the freeware software GIMP 2.6.7 (Free Software
obtained using hydrodynamic effects alone and without need for Foundation). Videos were recorded using a digital consumer camcorder camera
nanochannels or membranes, the possibility to rapidly adjust and (HDR-SR7, Sony Electronics) connected to the microscope through a C-mount
move the gradient hydrodynamically within a few seconds, while adaptor. The gradient lengths shown in Figure 5g were defined as the distance
also allowing to move it physically by displacing the MFP, and, between 90 percent and 10 percent of the maximum fluorescence intensity signal,
as shown in Supplementary Figure S5.
finally, the possibility of repeated use of the device for many differ-
ent experiments. Chemicals and experimental procedures. Fluorescein sodium salt (C20H10Na2O5)
This setup should be well suited for applications in cell biology was purchased from Sigma-Aldrich, and fluorescein solutions were prepared by
such as studying cellular migrations26,35 and neuronal navigation36,37, dissolving 0.03 g of the salt with 100 ml of distilled water. Red and yellow-green
or stem cell differentiation38, because chemical cues can be applied carboxylate-modified fluorescent microspheres (fluosphere, 2 µm diameters) were
purchased from Invitrogen. The solutions used in experiments were prepared by
and modulated rapidly to individual cells at the SP with minimal diluting 0.1 ml of the beads solution with 13 ml of distilled water.
shear stress. It will be interesting to find out whether these proper-
ties can be exploited and the MFP–MQ used to study the response References
of neurons and stem cells to gradients; indeed neurons and stem 1. Jackson, J. D. Classical Electrodynamics 3rd edn, 832 (John Wiley & Sons, 1999).
cells are particularly difficult to culture and maintain inside closed 2. Wiedemann, H. Particle Accelerator Physics. I. Basic Principles and Linear Beam
channels for long periods39,40, and that have been shown to be very Dynamics 2nd edn, 469 (Springer, 1999).
sensitive to shear stress41. Furthermore, the MQ and floating gradi- 3. Russell, D. A., Titlow, J. P. & Bemmen, Y. J. Acoustic monopoles, dipoles, and
quadrupoles: an experiment revisited. Am. J. Phys. 67, 660–664 (1999).
ents could be used to study large samples that are hard to cultivate 4. Meyer, E. & Neumann, E.- G. Physical and Applied Acoustics: An Introduction
inside closed channels, such as embryos or tissue slices42. Finally, it 412 (AcademicPress, 1972).
may be useful for surface patterning by positioning and scanning 5. Thorne, K. S. Multipole expansions of gravitational radiation. Rev. Mod. Phys.
the MFP atop surfaces9,43,44. 52, 299–339 (1980).
6. de Hoffmann, E. & Stroobant, V. Mass Spectrometry: Principles and Applications
3rd edn, 503 (John Wiley & Sons, 2007).
Methods 7. Scheidegger, A. E. The Physics of Flow through Porous Media (University of
The MFP fabrication and operation. The MFP comprises a square Si chip (3 mm TorontoPress, 1974).
in width) with 4 etched holes of 360 µm diameter, and a PDMS interface chip with 8. Bear, J. Dynamics of Fluids in Porous Media (American Elsevier, 1972).
holes for receiving the capillaries of 360 and 250 µm outer and inner diameters, 9. Juncker, D., Schmid, H. & Delamarche, E. Multipurpose microfluidic probe.
respectively, that connect it to the injection and aspiration syringe pumps Nat. Mater. 4, 622–628 (2005).
(Nemesys) (Fig. 2a). The Si chips were fabricated in 200 µm thick, double-side- 10. Koplik, J., Redner, S. & Hinch, E. J. Tracer dispersion in planar multipole flows.
polished silicon wafers (Silicon Quest), using photolithography and deep-reac- Phys. Rev. E 50, 4650–4671 (1994).
tive-ion etching. Etching masks were formed on silicon wafers by using standard 11. Bazant, M. Z. Conformal mapping of some non-harmonic functions in
photo­lithography and S1813 photoresist (Shipley) to pattern a 2 µm thick thermal transport theory. Proc. R. Soc. A 460, 1433–1452 (2004).
SiO2 layer. The openings were made by deep reactive ion etching (DRIE, ASE 12. Burgers, J. M. in Adv. Appl. Mech. (eds Mises Richard Von & Kármán Theodore
etcher, Surface Technology Systems). The PDMS interface block was fabricated by Von) Vol 1 171–199 (Elsevier, 1948).
casting poly(dimethylsiloxane) PDMS (Sylgard 184Dow Corning) into a home- 13. Batchelor, G. K. An Introduction to Fluid Dynamics (Cambridge University
made micromold composed of two structured steel plates, a polished steel plate Press, 1967).
with vias (access holes) forming the bottom, and four capillaries (each inserted 14. Batchelor, G. K. An Introduction to Fluid Dynamics (Cambridge University
into one of the four vias in the polished steel plate) serving as place holders for the Press, 2000).

nature communications | 2:464 | DOI: 10.1038/ncomms1471 | www.nature.com/naturecommunications 

© 2011 Macmillan Publishers Limited. All rights reserved.
ARTICLE nature communications | DOI: 10.1038/ncomms1471

15. Deen, W. M. Analysis of Transport Phenomena 624 (Oxford University Press, 35. Cosson, S., Kobel, S. A. & Lutolf, M. P. Capturing complex protein gradients on
1998). biomimetic hydrogels for cell-based assays. Adv. Funct. Mater. 19, 3411–3419
16. Bazant, M. Z. & Moffatt, H. K. Exact solutions of the Navier Stokes equations (2009).
having steady vortex structures. J. Fluid Mech. 541, 55–64 (2005). 36. Dertinger, S. K. W., Jiang, X. Y., Li, Z. Y., Murthy, V. N. & Whitesides, G. M.
17. Abramowitz, M. A. & Stegun, I. A. Handbook of Mathematical Functions: with Gradients of substrate-bound laminin orient axonal specification of neurons.
Formulas, Graphs, and Mathematical Tables 1st edn (Dover series, 1964). Proc. Natl Acad. Sci. USA 99, 12542–12547 (2002).
18. deBeer, D., Stoodley, P. & Lewandowski, Z. Measurement of local diffusion 37. Yam, P. T., Langlois, S. D., Morin, S. & Charron, F. Sonic hedgehog guides
coefficients in biofilms by microinjection and confocal microscopy. Biotechnol. axons through a noncanonical, Src-family-kinase-dependent signaling pathway.
Bioeng. 53, 151–158 (1997). Neuron 62, 349–362 (2009).
19. Wang, L. Y. et al. In situ measurement of solute transport in the bone lacunar- 38. van Noort, D. et al. Stem cells in microfluidics. Biotechnol. Prog. 25, 52–60 (2009).
canalicular system. Proc. Natl Acad. Sci. USA 102, 11911–11916 (2005). 39. Regehr, K. J. et al. Biological implications of polydimethylsiloxane-based
20. Rani, S. A., Pitts, B. & Stewart, P. S. Rapid diffusion of fluorescent tracers into microfluidic cell culture. Lab. Chip. 9, 2132–2139 (2009).
Staphylococcus epidermidis biofilms visualized by time lapse microscopy. 40. Millet, L. J., Stewart, M. E., Sweedler, J. V., Nuzzo, R. G. & Gillette, M. U.
Antimicrob. Agents Ch. 49, 728–732 (2005). Microfluidic devices for culturing primary mammalian neurons at low
21. Perrault, C. M., Qasaimeh, M. A. & Juncker, D. The microfluidic probe: densities. Lab. Chip. 7, 987–994 (2007).
operation and use for localized surface processing. J. Vis. Exp. e1418 (2009). 41. Kim, I. A. et al. Effects of mechanical stimuli and microfiber-based substrate on
22. Abhyankar, V. V., Lokuta, M. A., Huttenlocher, A. & Beebe, D. J. neurite outgrowth and guidance. J. Biosci. Bioeng. 101, 120–126 (2006).
Characterization of a membrane-based gradient generator for use in cell- 42. Queval, A. et al. Chamber and microfluidic probe for microperfusion of
signaling studies. Lab. Chip. 6, 389–393 (2006). organotypic brain slices. Lab. Chip. 10, 326–334 (2010).
23. Frevert, C. W., Boggy, G., Keenan, T. M. & Folch, A. Measurement of cell 43. Lovchik, R. D., Drechsler, U. & Delamarche, E. Multilayered microfluidic probe
migration in response to an evolving radial chemokine gradient triggered by a heads. J. Micromech. Microeng. 12, 115006–115014 (2009).
microvalve. Lab. Chip. 6, 849–856 (2006). 44. Perrault, C. M. et al. Integrated microfluidic probe station. Rev. Sci. Instrum. 81,
24. Keenan, T. M. & Folch, A. Biomolecular gradients in cell culture systems. 115107–115115 (2010).
Lab. Chip. 8, 34–57 (2008).
25. Toetsch, S., Olwell, P., Prina-Mello, A. & Volkov, Y. The evolution of chemotaxis
assays from static models to physiologically relevant platforms. Integr. Biol. 1, Acknowledgements
170–181 (2009). We acknowledge funding from NSERC, CIHR, CHRP, Genome Canada, Genome
26. Jeon, N. L. et al. Neutrophil chemotaxis in linear and complex gradients of Quebec, CFI, and the assistance of McGill Nanotools Microfab Laboratory (funded by
interleukin-8 formed in a microfabricated device. Nat. Biotechnol. 20, 826–830 CFI, NSERC and VRQ). M.A.Q. thanks Roozbeh Safavieh and Cécile Perrault for their
(2002). help and discussion. We acknowledge Michel Godin, Sebastien Ricoult, Gina Zhou,
27. Barkefors, I. et al. Endothelial cell migration in stable gradients of vascular Huiyan Li, Setareh Ghorbanian, and Nageswara R Ghattamaneni for critical reading the
endothelial growth factor a and fibroblast growth factor 2—Effects on manuscript. M.A.Q. acknowledges Alexander Graham Bell Canada NSERC Scholarship
chemotaxis and chemokinesis. J. Biol. Chem. 283, 13905–13912 (2008). and D.J. acknowledges support from a Canada Research Chair.
28. Kim, T., Pinelis, M. & Maharbiz, M. Generating steep, shear-free gradients of
small molecules for cell culture. Biomed. Microdevices 11, 65–73 (2009).
29. Diao, J. et al. A three-channel microfluidic device for generating static Author contributions
linear gradients and its application to the quantitative analysis of bacterial M.A.Q. designed the research, carried out the FEM analysis, performed the experiments,
chemotaxis. Lab. Chip. 6, 381–388 (2006). analysed the results and wrote the article; T.G. carried out the mathematical analysis and
30. Paliwal, S. et al. MAPK-mediated bimodal gene expression and adaptive wrote the article; D.J. designed the research, analysed the results and wrote the article.
gradient sensing in yeast. Nature 446, 46–51 (2007).
31. Shamloo, A., Ma, N., Poo, M.- M., Sohn, L. L. & Heilshorn, S. C. Endothelial
cell polarization and chemotaxis in a microfluidic device. Lab. Chip. 8,
1292–1299 (2008). Additional information
32. Keenan, T. M., Hsu, C.- H. & Folch, A. Microfluidic ‘jets’ for generating steady- Supplementary Information accompanies this paper at http://www.nature.com/
state gradients of soluble molecules on open surfaces. Appl. Phys. Lett. 89, naturecommunications
114103–114105 (2006). Competing financial interests: The authors declare no competing financial interests.
33. Cate, D. M., Sip, C. G. & Folch, A. A microfluidic platform for generation of sharp
Reprints and permission information is available online at http://npg.nature.com/
gradients in open-access culture. Biomicrofluidics 4, 044105–044110 (2010).
reprintsandpermissions/
34. Moorjani, S., Nielson, R., Chang, X. A. & Shear, J. B. Dynamic remodeling of
subcellular chemical gradients using a multi-directional flow device. Lab. Chip. How to cite this article: Qasaimeh, M. A. et al. Microfluidic quadrupole and floating
10, 2139–2146 (2010). concentration gradient. Nat. Commun. 2:464 doi: 10.1038/ncomms1471 (2011).

 nature communications | 2:464 | DOI: 10.1038/ncomms1471 | www.nature.com/naturecommunications

© 2011 Macmillan Publishers Limited. All rights reserved.