RESEARCH PAPER

Experimental study of microbubble coalescence in a T-junctionmicrofluidic device

Lu Yang • Kai Wang • Jing Tan • Yangcheng Lu •

Guangsheng Luo

Received: 7 August 2011 / Accepted: 9 November 2011 / Published online: 29 November 2011

� Springer-Verlag 2011

Abstract This study presents the microbubble coalescence

process in a confined microchannel. Triple T-junction

microfluidic devices with different main channel size were

designed to generate monodispersed microbubble pairs with

air/n-butyl alcohol–glycerol solution as the working system.

The head-on collision of microbubble pair was realized in the

microfluidic devices. Three collision results including

absolute coalescence, probabilistic coalescence, and non-

coalescence were distinguished. The effects of liquid vis-

cosities and two-phase superficial velocities on the coales-

cence behavior were determined. The results showed that

microbubble coalescence process in the confined space was

slightly faster than in the free space. Increasing liquid vis-

cosity apparently prevents coalescence. In the probabilistic

coalescence region, higher two-phase superficial velocity

could reduce the percentage of coalescence events. Two

characteristic parameters representing the bubble contact

time and film drainage time have been introduced to analyze

the microbubble coalescence behaviors and a linear corre-

lation could clearly distinguish the coalescence and non-

coalescence region.

Keywords Microfluidic � Microbubble �Head-on collision � Coalescence

1 Introduction

Microfluidic technology has been widely applied in many

research areas, such as analysis, chemical synthesis,

separation, material preparation, and biological engineer-

ing (Huebner et al. 2008; Song et al. 2006). These appli-

cations are attributed to the advantages of microfluidic

devices, where monodispersed microdroplets or micro-

bubbles can be controllably prepared and manipulated,

with enhancement of mass and heat transfer due to the

large specific surface area (Baroud et al. 2010; Teh et al.

2008; Gordillo et al. 2004). Gas/liquid system is one of

most important systems in scientific and industrial appli-

cations including chemistry, food, environment, and phar-

maceutical process. Combining the advantages of

microfluidic technology, a number of gas/liquid processes

involving reaction and mass transfer have been realized in

microfluidic device and developed rapidly during recent

years (Leclerc et al. 2008; Su et al. 2010). However, bubble

coalescence is a common phenomenon in the above prac-

tical applications. For the controllable use of gas/liquid

microfluidic system, the gas phase is sometimes required to

be stably dispersed in the liquid phase to keep good mixing

condition (coalescence need to be prevented), and some-

times efficient coalescence is desirable to enhance the

phase separation (coalescence is favorable). It can be seen

that a lack of control over the bubble coalescence would

severely limit their usefulness. Therefore, it is essential to

understand the fundamental fluid dynamics of microbubble

coalescence and to further realize effective control of

coalescence process.

The investigation of bubble coalescence behavior under

the general conditions has a long history and many

experimental and theoretical analyses have been carried out

to study the coalescence mechanisms, effect factors, and

coalescence models (Liao and Lucas 2010; Lessard and

Zieminski 1971; Stover et al. 1997). Several experimental

studies have been conducted and various methods have

been designed to observe the coalescence process between

L. Yang � K. Wang � J. Tan � Y. Lu � G. Luo (&)

The State Key Laboratory of Chemical Engineering,

Department of Chemical Engineering,

Tsinghua University, Beijing 100084, China

e-mail: gsluo@tsinghua.edu.cn

123

Microfluid Nanofluid (2012) 12:715–722

DOI 10.1007/s10404-011-0912-4

bubble pairs. For example, Tsang et al. (2004) studied the

effect of electrolyte concentration on the bubble coales-

cence with adjacent capillaries. A critical electrolyte con-

centration, whose value increased with decreasing bubble

diameter from 4.2 to 1.4 mm, was found in their work. Ata

et al. (2010) observed the pH-induced coalescence of 2 mm

contacted air bubbles with parallel capillaries. A pH sen-

sitive latex was introduced to control the coalescence by

adjusting the latex structure on bubble surface. Martin et al.

(2007) investigated the bubble coalescence on sieve plates

with pore diameters varying from 1.5 to 2.5 mm and the

distance between two neighboring holes was controlled.

Kazakis et al. (2008) focused on the effect of liquid

properties and operating conditions on the bubble interac-

tion by using two adjacent micro-tubes and a porous

sparger. Their experimental results enhanced the compre-

hension of coalescence mechanism in the bubble genera-

tion (bubble diameter *1 mm). Tse et al. (1998) observed

the stages of coalescence process between two bubbles

formed in a stirred tank, where the bubbles get close and

contact each other under the action of buoyancy. In addi-

tion, some researchers also focus on the interaction and

force between coalesced bubbles from microcosmic point

of view and determined the coalescence conditions and

mechanisms by using atomic force microscope (Vakarelski

et al. 2010). Besides experimental researches, theoretical

studies have also been developed to understand the coa-

lescence phenomenon. Liao and Lucas (2010) reviewed

several theoretical models for bubble and droplet coales-

cence process including collision frequency and coales-

cence efficiency. Among those models, the film drainage

model is the most widely used, which describes the coa-

lescence efficiency with two time scales, i.e. the contact

time and the drainage time (the time required for the liquid

film to thin down to a critical thickness). This model states

that coalescence occurs when the liquid film ruptures

before bubbles separate.

According to the previous studies on bubble coalescence

process, researchers paid more attention to the ideal,

stagnant, and millimeter-sized bubbles in free spaces. In

contrast, few studies focus on the coalescence process of

microbubbles which flow in confined microchannels.

Therefore, it is valuable to explore the microbubble coa-

lescence in microfluidic device. In our previous works, the

generation of microbubbles has been investigated in

T-junction microfluidic devices with cross-flow and per-

pendicular-flow shearing methods (Tan et al. 2009; Wang

et al. 2011; Xu et al. 2006a, b). In this work, we used the

T-junction microfluidic device to give an in-depth study on

the coalescence behavior of microbubbles. Based on the

fusion studies of microdroplets in microchannels (Chris-

topher et al. 2009; Zhao and Middelberg 2011), the head-

on collision of fluid particles is a simple method to realize

coalescence of microdroplets in solutions. This method was

introduced into our experiment and coalescence process of

flowing microbubbles in confined T-junction microchannel

devices was investigated. The effects of liquid viscosity

and two-phase superficial velocity were carefully examined

and the fluid dynamic conditions of bubble coalescence

were analyzed based on the experimental results. To our

knowledge, the microbubble coalescence behavior in con-

fined microchannel space has not been reported.

2 Experimental section

2.1 Microfluidic device

A schematic figure of the microfluidic chip is shown in

Fig. 1. Two perpendicular rupture generators are used in

this device with T-shaped structure, and their section

width and height are 0.4 and 0.36 mm, respectively. The

main channel has a square section whose height and

width equal to 0.6 mm. The narrower side T-junctions

were designed to use the sudden expansion structure to

generate microbubbles, and the produced bubbles can be

more stable and well monodispersed. The advantages of

this perpendicular flow shearing method have been proved

by our previous work (Tan et al. 2009). For comparison,

another microfluidic device with main channel dimension

of 0.8 mm 9 0.8 mm and narrower channel dimension of

0.5 mm 9 0.48 mm was also used to study the micro-

bubble coalescence process in microchannel with different

widths. Microbubbles were generated in these specially

designed microfluidic devices and head-on collision of

microbubbles was realized at the T-shaped meeting

junction, as shown in Fig. 1. The microfluidic device was

fabricated on a polymethyl methacrylate (PMMA) plate

by precision milling, and sealed with another PMMA

plate by use of a high-pressure thermal sealing machine

(A274, Techson) at 0.4 MPa, 70�C.

Fig. 1 Schematic figure of microdevice, microbubbles generated

with perpendicular rupture method and then collided at meeting

junction

716 Microfluid Nanofluid (2012) 12:715–722

123

2.2 Operation and observation

Figure 2 shows the experimental set up. Experiments were

carried out with a microscope, which equipped with a high-

speed CCD camera (DK-2740, Dantec Dynamics) and set

over the microchannel chip. The collision processes of

microbubbles at the meeting junction were recorded with a

frame frequency of 2,500 fps. An optical-fiber light source

was placed on the bottom of the chip to illuminate the

microchannel. Both gas and liquid phases were delivered

and controlled by syringe pumps (LSP01-1A, Longer) with

gastight syringes. In the experiments, the gas phase was

pumped into the microchannel firstly and then the liquid

phase followed. After changing any operating conditions,

at least 2 min was allowed for the flows to reach stable

state. The gas flow rates (QG) were controlled at 50, 80,

100, 150, 160 lL/min, and the liquid flow rates (QL) were

varied from 120 to 800 lL/min. The average diameters of

generated bubbles were in the range of 362–543 and

492–626 lm, when using the microchannel with main

channel width of 0.6 and 0.8 mm, respectively. The poly-

dispersity index of bubble diameters at every operating

condition, which is defined as the ratio of the relative

deviation of bubble size to the average bubble diameter,

was less than 3%, so it indicates the microbubbles are

highly uniform (Xu et al. 2006a). Flows of both phases

were consistent in the two generators for equal-sized

bubbles in collision. In order to eliminate timing difference

between the arrivals of two bubbles at the meeting junc-

tion, the synchronization of the generators were strictly

controlled and all the results in this paper were chosen from

the tests that the head-on collision were taken place at the

meeting junction. At least 15 tests were repeated for one

operating condition to give parallel experimental results.

2.3 Working systems

Air and n-butyl alcohol solutions were introduced to form

the working systems in this work. Different concentrations

of glycerol (0, 7, 10, 15 and 20 wt%) were added into butyl

alcohol for varying the liquid phase viscosities. Analyti-

cally pure n-butyl alcohol and glycerol (C99.0%) used in

the experiment were provided from Beijing Chemical

Plant. The viscosities of the liquid phases were measured

with an Ubbelohde viscometer and the surface tension of

the working systems was measured by a pendent drop

interfacial tension meter (OCAH200, DataPhysics Instru-

ments GmbH) at 25�C. The physical properties of the

working systems are listed in Table 1.

3 Results and discussion

3.1 Observation of microbubble coalescence processes

Figure 3 shows a typical microbubble coalescence process

in the microchannel. The result shows the coalescence

process proceeds quickly at millisecond time scale in the

meeting junction, which can be divided into three distinct

stages. In the first stage, the opposite flowing bubbles

approach and contact each other forming a thin liquid film

between bubble surfaces, which has a typical thickness of

several micrometers (Ata et al. 2010). The second stage is

film thinning or drainage. With the drainage of liquid film,

the film thickness is gradually reduced. Thinning of the

film is primarily driven by capillary force which is sig-

nificantly affected by the physical properties. When the

film thickness reduces to about 100 nm, Van der Waals

attraction will accelerate the drainage process (Almatrou-

shi and Borhan 2006). Once the liquid film is thinning less

than a critical thickness (a few tens of nanometers) (Horn

et al. 2011), film rupture occurs instantaneously leading to

the bubble coalescence as the third stage. Following bub-

ble fusion process, the surface of coalesced bubble shows

strong oscillations due to the release of surface energy. The

bubble expands in the horizontal and vertical orientations

alternately (Ata et al. 2010), and then contracts to sphere.

The stages of microbubble coalescence almost are the

same as previous studies from a microcosmic aspect, but

the time scales are different. Figure 3 shows the time scale

for the microbubble coalescence in the confined space is

only within 1.6 ms. Comparing with the bubble coales-

cence in free space, where time scale for coalescence is

usually longer than 10 ms (Ata et al. 2010; Tse et al.

1998), the coalescence process of microbubble in the

confined space is slightly faster. It can be attributed to the

effects of small bubble size and microchannel confinement.

Because the small bubbles will result in the small diameter

of liquid film between two bubbles and it can reduce the

coalescence time. In addition, the channel wall will play an

important role in this process due to the microbubbles are

confined in microchannel. Especially at the meeting junc-

tion microchannel wall can push bubbles together to

facilitate them coalescence. Of course, the role of the

channel wall should be studied in our further research.Fig. 2 Experimental set up

Microfluid Nanofluid (2012) 12:715–722 717

123

Moreover, the film-thinning step is generally believed to

be the rate-controlling step in the coalescence process

(Almatroushi and Borhan 2006), so here the film-thinning

stage might be accelerated.

In contrast, we also observed bubble collision process

without coalescing, as shown in Fig. 4. At the meeting

junction, deformed bubbles lost their kinetic energy to

increase the surface energy with collision. For these bubble

pairs, the thin liquid film formation and film drainage

stages are the same as coalesced bubble pair in Fig. 3.

However, without film rupture occurring, the microbubbles

eventually slip past one another near the junction with

recovery of surface energy and travel downstream in a pair.

This result is quite different from Fig. 3, since the film

drainage stage is not completed in this collision process.

From Figs. 3 and 4, it can be seen that the collision

results of microbubble between different working systems

or operation conditions are different indeed. For all the

experimental observations, it could be concluded that the

coalescence behaviors of microbubble pair can be classi-

fied into three primary categories: (1) absolute coalescence,

(2) probabilistic coalescence and (3) non-coalescence as

shown in Fig. 5. Absolute coalescence behavior is defined

as the coalescence being observed for all bubble pairs in

repeated tests for one operating condition. On the contrary,

non-coalescence behavior is defined as all bubble pairs

separating from each other in repeated tests. Between the

absolute coalescence and non-coalescence behaviors, a

transition region, where part of the repeated tests showed

coalescence behaviors, was observed with varying test

conditions. This behavior is named as probabilistic coa-

lescence since the occurrence of coalescence is random

with a certain coalescence probability.

3.2 Studies of influence factors on microbubble

coalescence process

3.2.1 Effect of liquid viscosities

The liquid viscosity (lc) plays an important role in the

microbubble coalescence process. In our experiments, five

continuous phases with different viscosities were used to

explore the effects of liquid viscosity on microbubble

coalescence. As shown in Table 1, the viscosities of the

liquid phases are ranged from 2.42 to 5.47 mPa s.

Although the viscosity variation is not in a very wide

range, the change of microbubble coalescence behavior is

obvious in the experiment. Figure 6 shows all the experi-

mental results of the five working systems at different flow

conditions. The horizontal axis is liquid phase viscosity,

and vertical axis is the two-phase superficial velocity uTP

which is defined as uTP = (QG ? QL)/a, where a is the

cross section area of the microchannel. It can be seen that

absolute coalescence behavior almost appears at liquid

viscosity less than 3.48 mPa s and non-coalescence

behavior of microbubbles at viscosity higher than

3.48 mPa s. The working system with lc = 3.48 mPa s

contains absolute coalescence and probabilistic coales-

cence behaviors for both microdevices, and the working

system with lc = 4.28 mPa s contains probabilistic coa-

lescence and non-coalescence behaviors only for mic-

rodevice with main channel width of 0.8 mm. Thus, a

transition region appeared in these working systems, and

microbubble becoming less likely to coalesce with

increasing liquid viscosity.

As mentioned before, the theoretical studies on coales-

cence process (Liao and Lucas 2010) pointed out that the

Table 1 Physical properties at

25�CDensity q (g/mL) Surface tension c (mN/m) Viscosity l (mPa s)

Air 1.185 9 10-3 1.83 9 10-2

1. n-butyl alcohol 0.8039 21.0 2.42

2. 7% glycerol solution 0.8241 22.1 3.05

3. 10% glycerol solution 0.8342 22.7 3.48

4. 15% glycerol solution 0.8504 23.1 4.28

5. 20% glycerol solution 0.8704 23.9 5.47

Fig. 3 The typical microbubble coalescence process in microfluidic

device (the main channel dimension is 0.6 mm 9 0.6 mm and the

working system is air/7 wt% glycerol–butyl alcohol solution with gas

flow rate at QG = 100 lL/min and liquid flow rate at QL = 250 lL/

min. The moment of bubble contact each other is identified as the zero

time and the numbers under the pictures give the proceeding time)

718 Microfluid Nanofluid (2012) 12:715–722

123

film drainage model is the most popular theoretical model.

It used the lubrication theory, which starts with the Navier

Stokes Equation and the continuity equation and derives a

thinning equation. The drainage time in the model can be

obtained by integrating the film thinning equation. This

model mainly concerned if the interface of bubbles is

deformable and mobile, and according to this, the detailed

forms of model are different. Even in confined space, the

bubble size is in microscales, which is much larger than

interfacial scale. Therefore, the film drainage model is a

fundamental theory and it can be thought to be still suitable

for confined space. According to the film drainage model,

the film drainage stage can be significantly affected by

liquid viscosity, so microbubble obeys a similar rule as the

bubble in millimeter scale. Due to the film drainage time

increase with the increase of liquid viscosity, and the

contact time is not enough for film drainage, therefore the

bubble coalescing become gradually difficult.

3.2.2 Effect of two-phase superficial velocities

The experimental results showed that the two-phase super-

ficial velocities could affect the coalescence behavior mostly

for the working system with lc = 3.48 mPa s and

lc = 4.28 mPa s at the transition region. Figure 6a and b

indicates that in the working system with lc = 3.48 mPa s

for both microdevices, the coalescence behavior transforms

from absolute coalescence to probabilistic coalescence with

the increase of the two phase superficial velocities up to 18.5

and 14.6 mm/s, respectively. These superficial velocities

Fig. 4 The typical process indicating microbubble pair collision without coalescing (the main channel dimension is 0.6 mm 9 0.6 mm and the

working system is air/20 wt% glycerol–butyl alcohol solution with gas flow rate at QG = 100 lL/min and liquid flow rate at QL = 250 lL/min)

Fig. 5 The primary categories

of microbubble coalescence

behaviors: 1 absolute

coalescence, 2 probabilistic

coalescence, and 3 non-

coalescence

Fig. 6 Effect of liquid viscosities on the coalescence behavior of

microbubbles in the T-junction microchannel. The symbol (opensquare) is representing absolute coalescence, open circle is

probabilistic coalescence, and open triangle is non-coalescence.

a The main channel dimension is 0.6 mm 9 0.6 mm; b The main

channel dimension is 0.8 mm 9 0.8 mm

Microfluid Nanofluid (2012) 12:715–722 719

123

correspond to the transition capillary numbers whose values

are Ca = 0.0028 and 0.0022 (Ca = lcuTP/c). From Fig. 6b,

it can be seen that especially in the working system with

lc = 4.28 mPa s for the microdevice with main channel

width w = 0.8 mm, the coalescence behavior transforms

from probabilistic coalescence to non-coalescence with the

increase of superficial velocity up to 10.4 mm/s, and it

corresponds to transition capillary number Ca = 0.0019.

These results mean that the higher superficial velocity does

not facilitate the microbubble coalescence.

It is notable that the transition region appeared when the

working system with lc = 3.48 mPa s was applied. The

results showed in this transition region the percentage of

coalescence events is influenced by two phase superficial

velocity seriously, as shown in Fig. 7a. The percentage of

coalescence events is defined as the ratio of the accumu-

lated coalescence events to the total head-on collision

events in more than 15 repeated tests for one operating

condition. It can be found the percentage of coalescence

decreases with the increase of the two phase superficial

velocity for all gas flow rate conditions in the experiments

for microdevice with main channel width w = 0.6 mm.

Namely the higher two phase superficial velocity would

hinder the coalescence process. According to film drainage

model, the effects of two-phase superficial velocities could

be explained by the contact time of bubble pairs decreasing

with the increase of superficial velocity. The condition with

higher superficial velocity cannot provide sufficient time

for liquid film drainage, and then the bubbles move apart

from each other without coalescing.

Moreover, Fig. 7a also shows that at the same two-phase

superficial velocity condition (especially, when

uTP C 16.2 mm/s) the higher gas flow rate can result in the

lower percentage of coalescence. The microbubble diame-

ters d under different superficial velocity conditions are

shown in Fig. 7b. It can be found that the higher gas flow

rate leads to the larger bubble size. Thus, this result also

indicated that under the same liquid viscosity and two-phase

superficial velocity conditions, the small microbubbles are

more likely to coalesce than big bubbles. Figure 7c and d

shows the same tendency by comparison of two micro-

channels with different dimensions. Due to the microbubble

size in w = 0.8 mm microchannel is larger than in

w = 0.6 mm microchannel, under the same superficial

velocity condition the smaller microbubbles more tend to

coalesce, and the transition superficial velocity and capil-

lary number value in w = 0.8 mm microchannel mentioned

above are lower than in w = 0.6 mm microchannel. The

reason can be explained that the small bubbles will result in

the smaller diameter of liquid film, so the small bubbles

tend to coalesce due to requiring shorter film drainage time.

It is generally thought that small bubbles are more difficult

to coalesce than big bubbles, often because of the lower

collision frequency of small bubbles. Tsang et al. (2004)

also found that the smaller bubbles tend to coalesce in

electrolyte solution with the same concentration.

Fig. 7 a Effect of two phase

superficial velocities on

percentage of microbubble

coalescence events (the main

channel dimension is

0.6 mm 9 0.6 mm and the

continuous phase is 10wt%

glycerol–butyl alcohol

solution); b Microbubble

diameters under different

superficial velocity conditions

in the microdevice with main

channel width w = 0.6 mm.

The symbol with dot center is

representing absolute

coalescence, open symbol is

probabilistic coalescence;

c Percentage of microbubble

coalescence events in different

microchannels; d Comparison

of microbubble diameters in

different microchannels

720 Microfluid Nanofluid (2012) 12:715–722

123

3.3 Fluid dynamic rule for the microbubble

coalescence

As the film drainage model mentioned before, the contact

time and drainage time are the crucial parameters for

bubble coalescence process. This concept is introduced in

this work and two characteristic parameters representing

the bubble contact time and film drainage time are defined

to analyze the fluid dynamic conditions of microbubble

coalescence. According to the theoretical model (Liao and

Lucas 2010), we define the characteristic drainage time as

tdrainage * lcd/c, where lc is the continuous phase vis-

cosity, c is the surface tension, and d is the bubble diam-

eter. The characteristic contact time is in inverse proportion

to uTP which is the two-phase superficial velocity, and

characteristic contact time tcontact can be estimated as w3/

2(QG ? QL), where w is the microchannel width. It is

representing the residence time scale of external flow,

which is starting from the moment that bubbles contact

each other at meeting junction and until they separate.

Therefore, the area of microbubble coalescence behavior in

the experiment could be distinguished by comparing the

bubble contact time with the film drainage time, as shown

in Fig. 8a, b. It can be found that the probabilistic coa-

lescence area is between the absolute coalescence and non-

coalescence regions. The straight solid lines in Fig. 8a and

b are representing the relations as w3/2(QG ? QL) = 600lc

d/c-25 and w3/2(QG ? QL) = 900lcd/c-50, for each

microfluidic device respectively. The area on the left side

of the line, i.e. with longer contact time, is absolute coa-

lescence region. In contrast, the area on the right side of the

line is non-coalescence region. The relative deviation

between solid line and dash line is ±10%, and nearly all

the probabilistic coalescence points can fall into this band

region.

4 Conclusion

In this work, the coalescence process of flowing micro-

bubbles in confined microchannels with different channel

size was observed, and different influence factors on the

coalescence process were investigated. The results showed

that the microbubble coalescence process in confined space

was slightly faster than in free space. Three behaviors

including absolute coalescence, probabilistic coalescence,

and non-coalescence were observed. The coalescence

behavior was mainly influenced by the liquid viscosity and

two phase superficial velocity. Increasing liquid viscosity

could apparently prevent the coalescence process. In the

probabilistic coalescence region, the higher two-phase

superficial velocity could reduce the percentage of coa-

lescence events. The fluid dynamic conditions of micro-

bubble coalescence were analyzed by defining the

characteristic contact time and film drainage time, and a

linear relation was developed to distinguish the coales-

cence and non-coalescence region. To our knowledge, it is

a first quantitative study on the microbubble coalescence in

confined microchannels.

With the aid of microfluidic device, the controllable

coalescence process of microbubbles was realized. This

method could offer a good tool for investigation of

microbubble coalescence behavior. In order to provide a

general equation for predicting the microbubble coales-

cence process, further studies on the process with different

working systems and much wider range of microbubble

size will be carried on in our research.

Acknowledgments We would like to acknowledge the support of

the National Natural Science Foundation of China (21036002,

20876084) and SRFDP (20090002110070) for this work.

Fig. 8 Comparison of characteristic contact time and film drainage

time. The symbol (open square) is representing absolute coalescence,

open circle is probabilistic coalescence, and open triangle is non-

coalescence. a The main channel dimension is 0.6 mm 9 0.6 mm

and the straight solid line represent w3/2(QG ? QL) = 600lcd/c-25;

b The main channel dimension is 0.8 mm 9 0.8 mm and the straightsolid line represent w3/2(QG ? QL) = 900lcd/c-50

Microfluid Nanofluid (2012) 12:715–722 721

123

References

Almatroushi E, Borhan A (2006) Coalescence of bubbles translating

through a tube. Blackwell, Oxford, pp 508–526

Ata S, Davis ES, Dupin D, Armes SP, Wanless EJ (2010) Direct

observation of pH-induced coalescence of latex-stabilized bubbles

using high-speed video imaging. Langmuir 26(11):7865–7874

Baroud CN, Gallaire F, Dangla R (2010) Dynamics of microfluidic

droplets. Lab Chip 10(16):2032–2045

Christopher GF, Bergstein J, End NB, Poon M, Nguyen C, Anna SL

(2009) Coalescence and splitting of confined droplets at

microfluidic junctions. Lab Chip 9(8):1102–1109

Gordillo JM, Cheng ZD, Ganan-Calvo AM, Marques M, Weitz DA

(2004) A new device for the generation of microbubbles. Phys

Fluids 16(8):2828–2834

Horn RG, Del Castillo LA, Ohnishi Satomi (2011) Coalescence map

for bubbles in surfactant-free aqueous electrolyte solutions. Adv

Colloid Interface Sci. doi:10.1016/j.cis.2011.05.006

Huebner A, Sharma S, Srisa-Art M, Hollfelder F, Edel JB, Demello

AJ (2008) Microdroplets: a sea of applications? Lab Chip

8(8):1244–1254

Kazakis NA, Mouza AA, Paras SV (2008) Coalescence during bubble

formation at two neighbouring pores: an experimental study in

microscopic scale. Chem Eng Sci 63(21):5160–5178

Leclerc A, Alame M, Schweich D, Pouteau P, Delattre C, Bellefon C

(2008) Gas–liquid selective oxidations with oxygen under

explosive conditions in a micro-structured reactor. Lab Chip

8:814–817

Lessard RR, Zieminski SA (1971) Bubble coalescence and gas

transfer in aqueous electrolytic solutions. Ind Eng Chem Fundam

10(2):260–269

Liao YX, Lucas D (2010) A literature review on mechanisms and

models for the coalescence process of fluid particles. Chem Eng

Sci 65(10):2851–2864

Martin M, Montes FJ, Galan MA (2007) Bubble coalescence at sieve

plates: II. Effect of coalescence on mass transfer. Superficial area

versus bubble oscillations. Chem Eng Sci 62(6):1741–1752

Song H, Chen DL, Ismagilov RF (2006) Reactions in droplets in

microflulidic channels. Angewandte Chem Int Edn 45(44):7336–

7356

Stover RL, Tobias CW, Denn MM (1997) Bubble coalescence

dynamics. AIChE J 43(10):2385–2392

Su HJ, Wang SD, Niu HN, Pan LW, Wang AJ, Hu YK (2010) Mass

transfer characteristics of H2S absorption from gaseous mixture

into methyldiethanolamine solution in a T-junction microchan-

nel. Separat Purif Technol 72:326–334

Tan J, Li SW, Wang K, Luo GS (2009) Gas-liquid flow in t-junction

microfluidic devices with a new perpendicular rupturing flow

route. Chem Eng J 146(3):428–433

Teh SY, Lin R, Hung LH, Lee AP (2008) Droplet microfluidics. Lab

Chip 8:198–220

Tsang YH, Koh YH, Koch DL (2004) Bubble-size dependence of the

critical electrolyte concentration for inhibition of coalescence.

J Colloid Interface Sci 275(1):290–297

Tse K, Martin T, McFarlane CM, Nienow AW (1998) Visualisation

of bubble coalescence in a coalescence cell, a stirred tank and a

bubble column. Chem Eng Sci 53(23):4031–4036

Vakarelski IU, Manica R, Tang X, O’Shea SJ, Stevens GW, Grieser F,

Dagastine RR, Chan DYC (2010) Dynamic interactions between

microbubbles in water. Proc Natl Acad Sci USA 107(25):11177–

11182

Wang K, Lu YC, Xu JH, Tan J, Luo GS (2011) Generation of

micromonodispersed droplets and bubbles in the capillary

embedded t-junction microfluidic devices. AIChE J 57(2):299–

306

Xu JH, Li S, Chen GG, Luo GS (2006a) Formation of monodisperse

microbubbles in a microfluidic device. AIChE J 52(6):2254–

2259

Xu JH, Li SW, Wang YJ, Luo GS (2006b) Controllable gas–liquid

phase flow patterns and monodisperse microbubbles in a

microfluidic t-junction device. Appl Phys Lett 88:13350613

Zhao CX, Middelberg A (2011) Two-phase microfluidic flows. Chem

Eng Sci 66(7):1394–1411

722 Microfluid Nanofluid (2012) 12:715–722

123