Professional Documents
Culture Documents
DOI 10.1007/s10404-014-1414-y
RESEARCH PAPER
Huai Z. Li
Received: 10 February 2014 / Accepted: 28 April 2014 / Published online: 9 May 2014
Ó Springer-Verlag Berlin Heidelberg 2014
Abstract This work aims at investigating the breakup BPO Breakup with permanent obstruction
dynamics of the ferrofluid droplet under magnetic fields in BTO Breakup with temporary obstruction
a microfluidic T-junction. The whole breakup processes of BNO Breakup without obstruction
the ferrofluid droplets under uniform magnetic field and NB Non-breakup
non-uniform magnetic field were exhaustively considered. NS Unstable breakup
The influence of both the flow rates and magnetic flux
density on the formation and breakup were extensively
studied. It was found that the type of breakup process can
be changed and the breakup frequency of ferrofluid drop-
lets can be actively modulated by applying a uniform 1 Introduction
magnetic field. While under non-uniform magnetic field,
the asymmetric breakup of ferrofluid droplets can be real- The droplet-based microfluidics is capable of precisely
ized. Both the magnetic field intensity and hydrodynamic handling minute amounts of fluidic elements (Belloul et al.
force affect the asymmetry of the breakup process. The 2011; Fu et al. 2009; Jullien et al. 2009; Leshansky et al.
non-uniform magnetic field can exert attraction on ferro- 2012). Under microscale, these systems possess high
fluid droplets in a dissymmetrical way to avoid breakup. interfacial areas and short diffusion distance which facili-
tate the heat and mass transfer (Shao et al. 2010; Tan et al.
Keywords Droplet-based microfluidics 2012). Thus, much attention has been paid to their appli-
Ferrofluid droplet Breakup Multiphase flow cations in chemical engineering (Abate and Weitz 2011;
Interfacial phenomenon Hou et al. 2011), biochemistry analysis (Engl et al. 2008),
materials synthesis (Duncanson et al. 2012) and so on. To
Abbreviations achieve successfully these applications, the control of
UM Uniform magnetic field droplets’ size is of prime importance. In the present studies,
NM No magnetic field the droplet size is controlled during droplet generation and
the breakup/coalescence processes downstream (Fu et al.
2012; Leshansky et al. 2012; Salkin et al. 2013; Wu et al.
Y. Wu T. Fu (&) Y. Ma 2014). Most of these studies take advantage of geometry-
State Key Laboratory of Chemical Engineering, School of based passive methods (Salkin et al. 2013). The formations
Chemical Engineering and Technology, Tianjin University,
of droplets take place usually in T-/Y-junction (Carrier
Tianjin 300072, People’s Republic of China
e-mail: ttfu@tju.edu.cn et al. 2014; Glawdel et al. 2012; Leshansky and Pismen
2009; Steegmans et al. 2010; Wang et al. 2011), co-flow
Y. Wu H. Z. Li (&) (Cramer et al. 2004; Ganan-Calvo and Riesco-Chueca
Laboratory of Reactions and Process Engineering, University of
2006) and flow-focusing (Fu et al. 2012; van Hoeve et al.
Lorraine, CNRS, 1, Rue Grandville, BP 20451,
54001 Nancy Cedex, France 2011; Ward et al. 2010) microdevices. The generations of
e-mail: Huai-Zhi.Li@univ-lorraine.fr droplets under different operating parameters and various
123
20 Microfluid Nanofluid (2015) 18:19–27
flow regimes such as squeezing, dripping and jetting were field (Liu et al. 2011a; Tan et al. 2010; Tan and Nguyen
studied (Christopher et al. 2008; Cordero et al. 2011; Fu 2011; Wu et al. 2013). The effects of uniform and non-
et al. 2012; Li et al. 2011; Steegmans et al. 2010; Utada uniform magnetic fields on the droplet formation of fer-
et al. 2007). Some physical models were proposed, and rofluids are studied (Liu et al. 2011a, b). The first work in
previous works show that the droplet size increases with T-junction concerns the active control of droplets’ size by
the flow rate ratio of dispersed phase to continuous phase. changing the magnet positions (Tan et al. 2010). When the
The breakup and coalescence process downstream can be magnet is placed at the downstream of the T-junction, the
regarded as competing mechanism: The former decreases formation of bigger droplets is favoured. Inversely, it forms
the droplet size, while the latter increases it. To breakup smaller droplet. Some similar observations were also made
droplets in a microfluidic device, various geometries were in flow-focusing device, in particular under a uniform
employed. Besides T-/Y-junction, obstacles can be placed magnetic field (Liu et al. 2011a, b; Tan and Nguyen 2011;
in microchannels (Bedram and Moosavi 2011; Leshansky Wu et al. 2013). Under a radial uniform magnetic field, the
et al. 2012; Salkin et al. 2013). In these devices, the size magnetic force acting on the droplet offsets some part of
ratio of daughter droplet can be controlled (Bedram and the impact force of the continuous phase during expanding
Moosavi 2011; Hoang et al. 2013; Leshansky et al. 2012; process of droplet. While under axial magnetic field, the
Salkin et al. 2013; Samie et al. 2013) by changing the magnetic force stretches the droplet to enlarge the flow
length ratio, cross-section ratio of the branches of the section for the continuous phase within the microchannel to
junction or by changing the position/length of obstacles. To expel. The impact force of the continuous phase on droplet
realize the coalescence between droplets (Bremond et al. is then reduced. To our best knowledge, only a very few
2011; Christopher et al. 2009; Wang et al. 2013), divergent studies focus on the breakup dynamics of ferro-droplets.
and convergent structures are mainly used to facilitate Our present work aims to study the breakup of ferro-
contact and collision between droplets. droplets in T-junction under uniform and non-uniform
Within geometry-based passive microdevices, the magnetic fields.
droplet size is mainly determined by the device geometry
(Bedram and Moosavi 2011; Christopher et al. 2008;
Cordero et al. 2011; Fu et al. 2012; Hoang et al. 2013; 2 Experimental
Leshansky et al. 2012; Li et al. 2011; Salkin et al. 2013;
Samie et al. 2013; Steegmans et al. 2010; Utada et al. 2.1 Details of microchannel
2007). For a given geometry, it is difficult to casually
modulate the size of droplets. It is then necessary to make Figure 1 shows the layout of the microchannel that consists
use of external factors to regulate the size of droplets. of a flow-focusing geometry to form droplet and a sym-
Several works have elaborated local control coupled with metric loop. The continuous phase (mineral oil) flows into
the flow rate ratio to extending the droplet size distribution the two identical lateral branches with the equal flow rate
(Kim et al. 2007; He et al. 2010; Ho et al. 2011; Cheung Qc/2, and the dispersed phase (ferrofluid) flows into main
and Qiu 2011; Liu et al. 2011a; Tan et al. 2010; Tan and channel at a flow rate of Qd; these three flow streams focus
Nguyen 2011; Wu et al. 2013). He et al. transformed the at the intersection downstream forming new droplets. And
flow front of the dispersed phase to a Taylor cone in turn then, the droplet flow toward the symmetric loop is illus-
causing Rayleigh capillary instability to control the trated in the right-hand side diagram of Fig. 1. This
breakup processes. Ho et al. controlled the viscosity and structure guarantees the same pressure drop in two sym-
interfacial tension by local heating to adjust the sizes of metric branches, which reduces maximally the influence of
droplets. Cheung and Qiu utilized the acoustic actuation to the distribution of continuous phase to the breakup of ferro-
bring periodic oscillation to control the sizes of droplet.
Recently, the magnetic manipulation has attracted much
attention (Liu et al. 2011a; Tan et al. 2010; Tan and
Nguyen 2011; Wu et al. 2013). The magnetic field will
neither induce thermal effect nor affect pH, surface charge
and ionic concentration (Nguyen 2012). In contrary to
laser, acoustic actuation and electric field, it is simple and
practicable to endow a fluid with the paramagnetic char-
acter by adding a minute amount of nanoscale magnetic
Fig. 1 Schematic diagram of the microfluidic channel. Ferrofluid
particles (Afkhami et al. 2010; Nguyen 2012; Wu et al. droplets breakup into two daughter droplet at the microfluidic
2013). Therefore, it is straightforward to manipulate the T-junction divergence. The cross-section of the formation part of
formation of ferrofluid droplets by an external magnetic the microchannel is 400 9 400 lm
123
Microfluid Nanofluid (2015) 18:19–27 21
droplets. At the T-junction divergence of the loop, the of the microdevice as light source. The images were ana-
microchannel is divided into two branches of equal length; lyzed by customized image processing software Matlab.
here, droplet could breakup into two daughter droplets. At
the T-junction convergence, the two streams recombine 2.3 Materials
and exit into the outlet channel. The microfluidic device
was fabricated in a plate (90 9 30 9 8 mm) of polymethyl In order to form ferrofluid droplets, water-based ferrofluid
methacrylate (PMMA) by precision milling and sealed (EMG 807, Ferrotech, Germany) and oil phase (mineral
with another PMMA plate (90 9 30 9 3 mm). The mi- oil) were, respectively, introduced as the dispersed phase
crochannel has a square cross-section of 400 lm wide and and continuous phase. At 27 °C, the density and viscosity
400 lm high. of the ferrofluid were qd = 1,100 kg m-3 and ld = 2 -
mPa s, respectively. As the magnetic particles had an
2.2 Experimental setup and procedure average diameter of 10 nm, the magnetoviscous effect
could be negligible (Pamme 2006). The particles volume
The scheme of the experimental setup is shown in Fig. 2. concentration was 1.8 % in this ferrofluid with the initial
The microfluidic device was placed under an inverted susceptibility v = 0.39. The mineral oil (CAS 8042-47-5,
microscope (Leica, Germany) which was connected to a Sigma-Aldrich, Germany) with 4 % (wt) surfactant sorbi-
high-speed camera up to 100,000 frames per second tan lauric acid ester (Span20, Sigma-Aldrich, Germany)
(CamRecord600, Optronis GMBH, Germany). Three frame was used as the continuous phase. The density and vis-
rates (500 fps with the resolution 1,280 9 1,024 pixels, cosity of the continuous phase were qc = 838 kg.m-3 and
1,000 fps with the resolution 800 9 500 pixels and 2,000 lc = 35.2 ± 0.2 mPa s, respectively. The interfacial ten-
fps with the resolution 400 9 248 pixels) and a shutter sion between the ferrofluid and mineral oil measured by a
speed of 1/6,000 s were used. The dispersed phase (ferro- dynamical tensiometer (IT Concept, France) was
fluid) and the continuous phase were injected to the mi- r = 1.5 ± 0.1 mN m-1.
crochannel from two independent syringes (5 and 10 mL,
Hamilton, Germany) driven by syringe pumps (Harvard
Apparatus, USA). The uniform magnetic field was realized 3 Results and discussion
by placing two neodymium–iron–boron (NdFeB) magnet
(50 9 15 9 15 mm, Supermagnete, France) at each side of 3.1 Droplet breakup and non-breakup in the T-junction
the loop as shown in Fig. 2. The T-junction divergence under uniform magnetic fields
where the droplets fragment was placed in the middle
height of magnets. The direction of the uniform magnetic Previous studies indicate the behaviors of droplets at the
field is parallel to the branches of the T-junction diver- T-junction depend on the droplets’ size L or the dimen-
gence. The non-uniform magnetic field is achieved by sionless length e = L/pWc (where Wc denotes the width of
placing only one boron (NdFeB) magnet at one side of the the channel) and the capillary number Ca (Bedram and
T-junction divergence. The magnetic flux density could be Moosavi 2011; Leshansky et al. 2012; Salkin et al. 2013).
easily adjusted by changing the separation distance In our experiments, the influence of these two parameters
between the T-junction divergence and magnets. The and the magnetic flux densities will be addressed. To gain a
magnetic flux densities were measured using a Gaussmeter comprehensive insight into the effect of a uniform mag-
(GM07, Hirst, UK) with an accuracy of 0.01 mT. A cold netic field (UM), it is suitable to study different breakup
fiber illumination (Jeulin SA, France) was placed on the top regimes at T-junction.
There are four main breakup regimes for ferro-droplet at
the T-junction both under uniform magnetic field (UM) and
no magnetic field (NM) as illustrated in Fig. 3a–d. The
breakup processes without magnetic field (NM) are intro-
duced as a reference for comparison. These different
regimes provide a visual description of the influence of a
magnetic field.
1. Breakup with permanent obstruction (BPO) as shown
in Fig. 3a
Fig. 2 Scheme of the experimental setup. The magnets are added in Droplets penetrate T-junction with two branches
the two sides and the left side to form uniform and non-uniform obstructed by the droplet under UM and NM. The differ-
magnetic field, respectively ence between the breakup processes under UM and NM is
123
22 Microfluid Nanofluid (2015) 18:19–27
123
Microfluid Nanofluid (2015) 18:19–27 23
Fig. 4 a–d Operating diagram organizing droplet behavior as a represents the critical capillary number for the transformed from
function of capillary number Ca and the dimensionless bubble size e BPO to BTO region. e The displacement of the critical capillary
under four different magnetic flux densities. The dotted line number as the increase in the magnetic flux densities
The ferrofluid droplet is stretched in the direction of the (B = l0H), Bo ranges from 0, 47.7 and 222.8–661.9. The
uniform magnetic field. This direction is parallel to the droplet aspect ratio b/a (b is the major axis of droplet and
branches of T-junction; thus, the two dumbbell-like parts a the minor axis) increases with the magnetic Bond number
detach from the walls offering larger flow sections of the Bo. Therefore, the increase in magnetic intensity enlarges
continuous phase in two branches as shown in Fig. 3b. the area of BTO region toward BPO region. As shown in
According to the previous works (Afkhami et al. 2010; Liu Fig. 4, the area for BNO is barely changed. This is due to
et al. 2011b; Wu et al. 2013), the magnetic force that the small droplet size and the tunnels are inherent. The
stretches the droplet is proportional to the square of the external magnetic field expands only the opening of the
magnetic field strength H. The stable shape of the droplet is tunnels.
determined by the ratio of interfacial tension and magnetic The uniform magnetic field affects not only the type of
force in unbounded flow. The magnetic Bond number was breakup, but also the breakup processes. As illustrated in
used to represent the ratio between magnetic force and Fig. 5, the temporal evolution of the minimum width of the
surface tension Bo = l0vWcH2/r (Nguyen 2012). Here, thinning neck Wn in BPO (Fig. 5a), BTO (Fig. 5b) and
l0 = 4p 9 10-7 N/A2 is the permeability of the free BNO (Fig. 5c) regions. The time zero is set at the moment
space. For B increase from 0 to 24.1, 52.1 and 89.8 mT of the droplet entirely penetrates into the branches.
123
24 Microfluid Nanofluid (2015) 18:19–27
123
Microfluid Nanofluid (2015) 18:19–27 25
time the shear of continuous phase acting on droplet of 630 ± 30 lm. Figure 8a reveals that the ratio VR/VL
increases. This implies that in BTO region, the pressure decreases obviously with the increase in the magnetic flux
difference is the primary factor to determine the breakup density. In general, the volume ratio VR/VL increases with
process. Qt. This implies that the hydrodynamic force plays an
increasing role. To assess this point, we use the velocity
3.2 Droplet breakup and non-breakup in the T-junction ratio of droplets in the right and left branch UR/UL to
under non-uniform magnetic fields represent the flow rate ratio in two branches. As shown in
Fig. 8b, the velocity ratio UR/UL was measured as a func-
Under a non-uniform magnetic field, the magnetic force tion of Qt. With the increase in Qt, the ratio UR/UL
F acting on droplet is not symmetrical (Mefford et al. decreases gradually toward an asymptotic value of one.
2007). In our experiments, the magnet was placed at the
left side of the microchannel. Under such a magnetic field,
the breakup of the ferrofluid droplet becomes asymmetrical
due to the magnetic force applied in the left direction. Two
daughter droplets with different sizes flow into two oppo-
site branches as shown in Fig. 7. These daughter droplets
bring additional flow resistances to the branch which they
flow through (Salkin et al. 2013). And the additional flow
resistance due to a droplet increases with the increase in
droplet size (Fuerstman et al. 2007). This gives rise to
feedback effects and affects then the flow distribution of
continuous phase at the T-junction. In our experiments, the
breakup of droplet occurs in a symmetric loop; the breakup
process is mainly controlled by the magnetic field and the
continuous flow. But the T-junction divergence and the
T-junction convergence are nearly of the equipotential line
in the magnetic field. The magnetic field affects only the
breakup processes but does not provide additional energy
to the daughter droplets. The kinetic energy is only pro-
vided by the continuous phase. Thus, the influences of the
magnetic field and continuous phase are coupled to some
extent.
Figure 8a shows the volume ratio of the droplets flowing
in the right branch and in the left branch VR/VL (where
V denotes droplet volume, and subscripts R and L mean
‘right’ and ‘left’, respectively) as a function of the total
flow rate of both phases Qt = Qc ? Qd under three dif- Fig. 8 a Volume ratio of the droplets flowing in the right branch and
ferent magnetic flux densities. The size of mother droplets in the left branch VR/VL as a function of the total flow rate Qt under
three different of magnetic flux density. b The velocity ratio of
is controlled by a fine adjustment of continuous and dis- droplets in right and left branch UR/UL as a function of the total flow
persed phases to maintain the length within a narrow range rate Qt
123
26 Microfluid Nanofluid (2015) 18:19–27
123
Microfluid Nanofluid (2015) 18:19–27 27
Cybulski O, Garstecki P (2010) Dynamic memory in a microfluidic Mefford OT, Woodward RC, Goff JD, Vadala TP, St Pierre TG,
system of droplets traveling through a simple network of Dailey JP, Riffle JS (2007) Field-induced motion of ferrofluids
microchannels. Lab Chip 10(4):484–493 through immiscible viscous media: Testbed for restorative
Duncanson W, Lin T, Abate AR, Seiffert S, Shah RK, Weitz DA treatment of retinal detachment. J Magn Magn Mater
(2012) Microfluidic synthesis of advanced microparticles for 311:347–353
encapsulation and controlled release. Lab Chip 12:2135–2145 Nguyen NT (2012) Micro-magnetofluidics: interactions between
Engl W, Tachibana M, Colin A, Panizza P (2008) A droplet-based magnetism and fluid flow on the microscale. Microfluid Nano-
high-throughput tubular platform to extract rate constants of fluid 12:1–16
slow chemical reactions. Chem Eng Sci 63:1692–1695 Pamme N (2006) Magnetism and microfluidics. Lab Chip 6:24–38
Fu T, Ma Y, Funfschilling D, Li HZ (2009) Bubble formation and Salkin L, Schmit A, Courbin L, Panizza P (2013) Passive breakups of
breakup mechanism in a microfluidic flow-focusing device. isolated drops and one-dimensional assemblies of drops in
Chem Eng Sci 64:2392–2400 microfluidic geometries: experiments and models. Lab Chip
Fu T, Wu Y, Ma Y, Li HZ (2012) Droplet formation and breakup 13:3022–3032
dynamics in microfluidic flow-focusing devices: from dripping Samie M, Salari A, Shafii MB (2013) Breakup of microdroplets in
to jetting. Chem Eng Sci 84:207–217 asymmetric T junctions. Phys Rev E 87:053003
Fuerstman MJ, Lai A, Thurlow ME, Shevkoplyas SS, Stone HA, Shao N, Gavriilidis A, Angeli P (2010) Mass transfer during Taylor
Whitesides GM (2007) The pressure drop along rectangular flow in microchannels with and without chemical reaction. Chem
microchannels containing bubbles. Lab Chip 7:1479–1489 Eng J 160:873–881
Ganan-Calvo AM, Riesco-Chueca P (2006) Jetting-dripping transition Steegmans MLJ, de Ruiter J, Schroen KGPH, Boom RM (2010) A
of a liquid jet in a lower viscosity co-flowing immiscible liquid: descriptive force-balance model for droplet formation at micro-
the minimum flow rate in flow focusing. J Fluid Mech 553:75–84 fluidic Y-junctions. AIChE J 56:2641–2649
Glawdel T, Elbuken C, Ren CL (2012) Droplet formation in Tan SH, Nguyen NT (2011) Generation and manipulation of
microfluidic T-junction generators operating in the transitional monodispersed ferrofluid emulsions: the effect of a uniform
regime. I. Experimental observations. Phys Rev E 85:016322 magnetic field in flow-focusing and T-junction configurations.
He P, Kim H, Luo D, Marquez M, Cheng Z (2010) Low-frequency ac Phys Rev E 84:036317
electro-flow-focusing microfluidic emulsification. Appl Phys Tan S-H, Nguyen N-T, Yobas L, Kang TG (2010) Formation and
Lett 96(17):174103 manipulation of ferrofluid droplets at a microfluidic T-junction.
Ho PC, Yap YF, Nguyen NT, Chai JCK (2011) Thermally mediated J Micromech Microeng 20:045004
droplet formation at a microfluidic T-junction. Micro Nanosyst Tan J, Lu YC, Xu JH, Luo GS (2012) Mass transfer performance of
3(1):65–75 gas-liquid segmented flow in microchannels. Chem Eng J
Hoang DA, Portela LM, Kleijn CR, Kreutzer MT, van Steijn V (2013) 181–182:229–235
Dynamics of droplet breakup in a T-junction. J Fluid Mech Utada AS, Fernandez-Nieves A, Stone HA, Weitz DA (2007)
717:R4 Dripping to jetting transitions in coflowing liquid streams. Phys
Hou J, Qian G, Zhou X (2011) Gas-liquid mixing in a multi-scale Rev Lett 99:094502
micromixer with arborescence structure. Chem Eng J van Hoeve W, Dollet B, Versluis M, Lohse D (2011) Microbubble
167:475–482 formation and pinch-off scaling exponent in flow-focusing
Jullien MC, Tsang Mui Ching MJ, Cohen C, Menetrier L, Tabeling P devices. Phys Fluids 23:092001
(2009) Droplet breakup in microfluidic T-junctions at small Wang K, Lu YC, Xu JH, Tan J, Luo GS (2011) Generation of
capillary numbers. Phys Fluids 21:072001 micromonodispersed droplets and bubbles in the capillary
Kim H, Luo D, Link D, Weitz DA, Marquez M, Cheng Z (2007) embedded T-junction microfluidic devices. AIChE J 57:299–306
Controlled production of emulsion drops using an electric field Wang K, Lu Y, Yang L, Luo G (2013) Microdroplet coalescences at
in a flow-focusing microfluidic device. Appl Phys Lett microchannel junctions with different collision angles. AIChE J
91(13):133106 59:643–649
Leshansky AM, Pismen LM (2009) Breakup of drops in a microflu- Ward T, Faivre M, Stone HA (2010) Drop production and tip-
idic T junction. Phys Fluids 21:023303 streaming phenomenon in a microfluidic flow-focusing device
Leshansky AM, Afkhami S, Jullien MC, Tabeling P (2012) via an interfacial chemical reaction. Langmuir 26:9233–9239
Obstructed breakup of slender drops in a microfluidic T junction. Wu Y, Fu T, Zhu C, Lu Y, Ma Y, Li HZ (2012) Asymmetrical
Phys Rev Lett 108:264502 breakup of bubbles at a microfluidic T-junction divergence:
Li X-B, Li F-C, Yang J-C, Kinoshita H, Oishi M, Oshima M (2011) feedback effect of bubble collision. Microfluid Nanofluid
Study on the mechanism of droplet formation in T-junction 13:723–733
microchannel. Chem Eng Sci 69:340–351 Wu Y, Fu T, Ma Y, Li HZ (2013) Ferrofluid droplet formation and
Link DR, Anna SL, Weitz DA, Stone HA (2004) Geometrically breakup dynamics in a microfluidic flow-focusing device. Soft
mediated breakup of drops in microfluidic devices. Phys Rev Matter 9:9792–9798
Lett 92:054503 Wu Y, Fu T, Zhu C, Ma Y, Li HZ (2014) Bubble coalescence at a
Liu J, Tan S-H, Yap YF, Ng MY, Nguyen N-T (2011a) Numerical microfluidic T-junction convergence: from colliding to squeez-
and experimental investigations of the formation process of ing. Microfluid Nanofluid 16:275–286
ferrofluid droplets. Microfluid Nanofluid 11:177–187
Liu J, Yap YF, Nguyen NT (2011b) Numerical study of the formation
process of ferrofluid droplets. Phys Fluids 23:072008
123