BioChip J. (2014) 8(2): 122-128DOI 10.1007/s13206-014-8207-y

Abstract The extent to which the carrier fluid wetsthe walls of a microchannel is crucial in the dropletformation process for segmented flow microfluidicapplications and can be influenced by the use of surfac-tants. Surfactants dynamically modify the microchan-nel surface leading to stabilization of the two phaseinterface, affecting the droplet formation process. Anexperimental study of the influence of hydrophobicsurfactant (Span 80) during the formation of water-in-oil droplets in a T-shaped microchannel geometry ispresented and the wetting properties of the micro-channel walls were characterized. The range of data tobe analyzed on the microscale is estimated from themacroscopic interfacial tension and contact angle mea-surements. The critical micelle concentration (CMC)level at the microscale was estimated by observingthe trend of droplet length variation with concentrationof surfactant in a microchannel. Microchannels used inthis work were fabricated using softlithography meth-ods and bonded using a custom-made plasma bondingsetup that does not require an ultra high vacuum cham-ber and hence saves the fabrication cost.

Keywords: Microdroplets, Microchannel, Double emul-sions, Wettability, Surfactants

Introduction

Technology based on droplet microfluidics offers anovel approach for the controlled preparation of mono-dispersed microemulsions. These monodispersed mic-roemulsions have been widely used to obtain high qual-ity pharmaceuticals1, food products2 and cosmetics3.They also have potential uses in a number of otherchemical and biological applications4,5. Precise controlof droplets containing nanoliter volumes of fluids isrequired for potential utilization of micro emulsions,coupled with an in-depth knowledge of the dynamicalproperties of multiphase flows. Many studies have in-vestigated droplet controlling parameters such as flowrate and viscosity ratios between the two immisciblefluids6,7 however improved understanding of the effectsof interfacial tension and wetting properties is stillrequired to fully control and explain the dynamics ofdroplet formation. This provides the motivation forthe present study.

T-junction microfluidic devices are among the mostpopular droplet generation devices. Droplets are form-ed when the dispersed phase stream is sheared off bythe carrier fluid at the junction8-11. It has been suggest-ed quantitatively that for the stable formation of W/Oemulsions, the contact angle of a water droplet withthe microchannel surface should be 180�12. Typically,the water droplet contact angle with a hydrophobicsurface is between 90�and 120�13 where the oil phaseis the surrounding media. If a surfactant is added tothe continuous oil phase, the contact angle graduallyincreases and the surface properties change from hydro-phobic to superhydrophobic12. Dreyfus and co-work-ers12 studied the wetting characteristics of droplet for-mation of deionized water/tetradecane-Span 80 sys-tems in a microfluidic device and identified different

Original Article

Dynamic Wetting in Microfluidic Droplet Formation

Shazia Bashir1,2,*, Xavier Casadevall i Solvas3, Muhammad Bashir4, Julia Margaret Rees1

& William Bauer Jay Zimmerman4

Received: 24 February 2014 / Accepted: 11 May 2014 / Published online: 20 June 2014�The Korean BioChip Society and Springer 2014

1School of Mathematics and Statistics, Hicks Building, HounsfieldRoad, University of Sheffield, S3 7RH, U.K.2Department of Physics and Applied Mathematics, Pakistan Instituteof Engineering and Applied Sciences, Islamabad, Pakistan3Institute of Chemical and Bioengineering, Department of Chemistryand Applied Biosciences, HC1 F109, ETH Zurich4Department of Chemical and Biological Engineering, University ofSheffield, Sheffield S1 3JD, U.K.*Correspondence and requests for materials should be addressed to S. Bashir ( shazia@pieas.edu.pk, shaziapieas@gmail.com)

flow regimes for various flow rate ratios between oiland water that were similar to those identified by8.They concluded that to achieve stable formation ofdroplets, the surfactant concentration should be wellabove the CMC. These investigations were furtherconfirmed by14,15 who emphasized the importance ofwetting properties by generating W/O and O/W emul-sions using a single microfluidic device.

The current work reports on an experimental inves-tigation comprising a systematic study of the effectsof wettability of microchannel walls on droplet forma-tion. Wettability in a microfluidic T-junction devicewas controlled by the addition of a surfactant into theoil phase. We have considered relatively low flow ratesof the dispersed phase fluid. The flow rate ratio wasaltered by varying the cross shear flow of the contin-uous oil phase in order to closely monitor the relativeeffects of interfacial tension and shear force on thedroplet break up process. The effect of surfactant con-centration on droplet lengths in microchannel withinthe contact angle transition region was studied quanti-tatively and CMC level was estimated and comparedwith that obtained from macroscale analysis for thefirst time. Furthermore, dynamic contact angles (advanc-ing and receding) formed by the aqueous phase withthe microchannel surface were also measured in orderto characterize the dynamic wetting properties of twophase microchannel flows.

Results and Discussion

Macroscale analysis

The measured values of water contact angles on thePDMS surface are shown in Figure 1(a) along withthe interfacial tension data. It can be seen that the con-tact angle of water with the PDMS surface increasesfrom hydrophobic to superhydrophobic as the con-centration of surfactant in the surrounding oil mediaincreases. The corresponding interfacial tension bet-ween water and oil decreases with surfactant concentra-tion and does not change significantly when the CMCof �0.5% w/w is exceeded.

Influence of surfactant on droplet length

Experiments of droplet formation in the microfluidicdevice were performed with surfactant concentrationsin mineral oil of above 0.1 wt.% as ordered flow pat-terns were obtained above the transition concentrationof Span 80 at 0.1 wt.% that were similar to those obtain-ed by14 for a system comprising DI-water/octadecane-Span 80. In Figure 1(a) the droplet length is presentedfor different surfactant concentrations ranging from

0.1 to 4% w/w at three different continuous phase flowrates of 0.1 μL/min, 0.2 μL/min and 0.8 μL/min, anda constant aqueous phase flow rate of 0.2 μL/min. Itwas observed that the droplet length decreases with anincrease in surfactant concentration. This is primarilydue to a decrease in the interfacial tension betweenthe two immiscible phases. The influence of the flowrate of the continuous phase on the length of dropletsproduced at surfactant concentrations above 0.8% wasinvestigated. Figure 2(a) shows the effect on dropletlength of surfactant concentrations in the range 0.8%to 4% by weight for a range of flow rates of continu-ous phase. The purpose of this study was to focus onthe transition of wettability of PDMS.

BioChip J. (2014) 8(2): 122-128 123

Figure 1. (a) The interfacial tension between DI water andmineral oil (shown by circles) and contact angle of a waterdroplet on PDMS surface in an oil (shown by squares) at vary-ing surfactant concentrations of Span 80. (b) Droplet lengthplotted against surfactant concentration for three different oilflow rates. A constant dispersed phase flow rate of Qd==0.2μL/min was used alongwith a photographic image showingthe variation of the droplet length with concentration of sur-factant in the oil phase. Dispersed phase flow rate was Qd==0.2 μL/min and the continuous phase flow rate was Qd==0.8μL/min.

Con

tact

ang

le(d

egre

es)

Dro

plet

leng

th(m

icro

ns)

Inte

rfac

ial t

ensi

on(m

N/m

)

180

160

140

120

100

20

15

10

5

0

1200

1000

800

600

400

200

0

(a) Macroscale

(b) Macroscale

0.01 0.1 1 10

0 1 2 3 4

Surfactant concentration (% w/w)

Sessile drop

Transition region

Pendent drop

Q==0.25 CS==0.2%

CS==0.8%

CS==4%

Q==0.25Q==1Q==0.5

The droplet length was found to be approximatelyindependent of the surfactant concentration for con-centration levels in excess of 2 wt.%. This is due tothe fact that the interfacial tension reaches its mini-mum value when the surfactant concentration is higherthan the CMC. The effect is more prominent as wemove towards the higher oil flow rate (Qc¤0.4 μL/min). Thus, we can also qualitatively describe the CMCfor the microchannel which is otherwise difficult todetermine accurately as the interfacial tension measure-ment between the two phases in a microchannel is stillnot well developed compared to those for macroscaleflows. It was observed that the effects of surfactant aremore prominent at high flow rate ratios (Q==Qd/Qc

¤1) or lower oil flow rates than for high oil flow rates.This is due to the fact that for lower oil flow ratesinterfacial forces dominate over shear forces and there-fore the change in droplet size achieved by altering the

surfactant concentration is greater. The same effect isnot significant for high oil flow rates or for low flowrate ratios (Q==Qd /Qc⁄1) due to the dominance ofshear forces over interfacial forces. Therefore, thegeneration of uniform, stable droplets at low oil flowrates without surfactant is difficult.

Effect of flow rate ratio on droplet length fordifferent surfactant concentrations

Figure 2(b) presents the droplet lengths for differentflow rate ratios. The concentration of surfactant inmineral oil was kept fixed at 4%. Figure 3(b) showsthe micrographs of droplets generated in mineral oilcontaining 4% surfactant with different flow rate ratios.It was observed from both figures that the droplet lengthdecreases as the flow rate of the continuous oil phaseis increased. The curve fitting analysis in Figure 2(b)shows that the length of droplets scales with Ca-0.25

which corroborates with the scaling obtained in experi-ments performed by16.

The ease of movement of droplets due to the increas-ed thickness of the film between the droplet interfaceand the channel wall as a consequence of increasingsurfactant concentration reduces the distance betweentwo droplets at a given fixed flow rate ratio as shownin Figure 4(a).

Surface wetting

Dynamical processes occur at the region of intersectionof the main channel with the lateral channel of themicrodevice due to interfacial instabilities. The adsorp-tion of surfactants and their mixtures on solid surfacescan promote their wettability. Thus the addition ofsurfactant to a mineral oil can enhance the capabilityof the oil to wet the hydrophobic solid surface of aPDMS microchannel or promote the dewetting of DIwater on the surface. The wettability of microfluidicchannel walls can be quantified by the ‘contact angle’that the liquid forms with the channel surface. Duringthe motion of a droplet the contact angle it forms withthe channel wall on the upstream, or de-wetting side,is referred to as the receding contact angle. The angleformed on the downstream, or wetting side, is referredto as the advancing contact angle17. The advancingand receding contact angles are usually denoted as θa

and θr respectively. The advancing contact angle isalways greater than or equal to the receding contactangle i.e., θa≥θr The difference θa-θr is known asthe contact angle hysteresis18. The underlying mecha-nisms of hysteresis and of dynamic contact angles arenot yet fully understood and depend on a number ofparameters including surface roughness and surfacechemistry19. However, in experiments incorporating a

124 BioChip J. (2014) 8(2): 122-128

Figure 2. (a) Length of droplets versus flow rate of mineraloil with added higher concentrations of Span 80 with constantdispersed phase flow rate of Qd==0.2 μL/min. (b) Length ofdroplets plotted against flow rate of mineral oil containing4% w/w of Span 80 surfactant for three different values offlow rate ratio, Q.

Dro

plet

leng

th(μ

m)

Dro

plet

leng

th(μ

m)

800

700

600

500

400

300

200

100

220

200

180

160

140

120

100

80

0.0 0.5 1.0 1.5 2.0 2.5

0 2 4 6 8 10 12

Qc (μL/min)

0.8% Span801% Span802% Span803% Span804% Span80

Q==0.5Q==0.25Q==0.125y==210x-0.25

y==178x-0.25

y==149x-0.25

(a)

(b)

moving contact line, it is possible to observe the exis-tence of these two contact angles. Figure 4(b) showsthe receding and advancing contact angles of detacheddroplets in a microchannel corresponding to carrieroils containing different surfactant concentrations atconstant flow rate ratios of Q==1 and Q==0.125 respec-tively. Both receding and advancing contact anglesincrease as the surfactant concentration is increased. Ata given surfactant concentration level lower values ofθr and θa were observed at a flow rate ratio of Q==1 ascompared to those at Q==0.125. Beyond the transitionsurfactant concentration which was observed at a 1%w/w of surfactant concentration, stable and uniformwater droplets were observed. This might be due to thedynamic surfactant adsorption on both liquid-liquidand liquid-solid interfaces in the microfluidic channel.Also, droplets exhibit less uniformity beyond Q¤2 atall surfactant concentrations and eventually the twophases flow parallel to each other and the droplet stopsto form. This regime is usually known as the stratifiedregime and results as the shear forces from the carrierfluid acting on the emerging interface are low andhence fail to rupture the water phase to form droplets.In this case, when the aqueous phase exhibits stratifi-ed wetting instead of confined dispersion, the dynamiccontact angles are assumed to be 0�.

Conclusions

In this paper we have presented direct measurementsof interfacial tensions between DI-water and mineral

oil and static contact angles of DI-water on a PDMSsurface surrounding an oil medium, for a range of con-centrations of the surfactant Span 80. The data providea basis for the dynamic flow experiments that wereconducted in a T-shaped PDMS microfluidic device.The effects of wetting properties between the mineraloil and DI-water on droplet formation were investigat-ed. Droplet lengths were measured for different levelsof surfactant concentration (Cs) and it was observedthat droplet length decreases as Cs increases and even-tually becomes almost independent of Cs at the CMClevel. The value of the CMC level at microscales is�1% w/w which is slightly higher than that measur-ed at macroscale level due to large surface to volumeratio associated with microscale flows. Therefore, itcan be concluded that at microscales where accurateinterfacial tension measurements are difficult to obtain,an alternative way to estimate CMC is from the behav-ior of changing droplet length with Cs. This is an im-portant original finding which can be applied to differ-ent combinations of surfactants and oils.

It was found that dynamic contact angles decreaseas the flow rate ratio between the dispersed and con-tinuous phase fluids is increased. Dynamic contactangles were observed to increase as the level of surfac-tant concentration in the carrier oil increased. Stableand uniform water droplets were observed in the mic-rochannel at concentrations in excess of the transitionsurfactant concentration level. The wetting dynamicsof droplets in the presence of a surfactant mixturedepend on a number of factors, such as the propertiesof the microchannel surface, viscous effects, interfa-

BioChip J. (2014) 8(2): 122-128 125

Figure 3. (a) Schematic of microfluidic chip, (b) Droplet formation images using Cs==4% and for a fixed flow rate Qd==0.2 μL/minin a T-junction where droplets of water are sheared off by the oil phase (c) flow focusing section where droplets relax to a sphericalshape.

Dispersed phase

Continuous phase

Qc==0.2μL/minQ==1

Qc==0.4μL/minQ==0.5

Qc==0.8μL/minQ==0.25

Qc==1.6μL/minQ==0.125

Outflow

(a)

(b) (c)

2.3 mm

cial tension between liquid-liquid interfaces and theadsorption of surfactant at solid-liquid interfaces.

Materials and Methods

Interfacial tension and contact angle measurements

DI-water was used as the aqueous phase in all of theexperiments. The non-aqueous phase used was a lightwhite mineral oil (M5904) obtained from Sigma-Ald-rich. The density of the mineral oil measured at roomtemperature (25�C) was found to be 843.1 kg/m3 byweighing the solution in a specific gravity bottle usinga sensitive mass balance with precision up to 5 signi-ficant figures. Viscosity was measured to be 24.1 mPas using an Ostwald viscometer. The Ostwald viscome-

ter was calibrated by measuring the flow time of DIwater thrice. The error was found to be ±0.06 mPa s.Before each measurement the viscometer was cleanedwith acetone and dried with compressed air in a cleanroom. All measurements were made in a controlledtemperature room with variation of less than ±1�C.Values of fluid viscosity were determined from theviscometer measurements using relationship:

δt ρμ==μwmmm mmmδtw

ρw

where μ is the viscosity of the mineral oil, μw is the ρ

viscosity of DI-water. The ratio mmm is the mineral oil ρw

density relative to that of water. The flow time δt is thetime that elapses as the meniscus of mineral oil passesbetween two marked locations on the viscometer andδtw represents the elapsed time in the case of water.

The surface active chemical tested was Span 80 (sor-bitan mono-oleate: C22H44O2) obtained from Sigma-Aldrich(S6760). Span 80 is one of the most commonlyused surfactants in microfluidics. It is a hydrophobicnonionic surfactant that contains a mixture of fattyacid components in addition to the predominant oleicacid component. Its hydrophilic lipophilic balance(HLB) is 4.3. Span 80 has a nominal molecular weightof 428 g/mol and its density was measured to be 986kg/m3 at 25�C. Span 80 was dissolved in the mineraloil and was found to be readily soluble at all concen-tration levels considered. The surfactant was used inits original form without further purification. To makesolutions of a desired concentration, a known volumeof surfactant was added with a micropipette to a knownvolume of solvent. The oil-surfactant solutions wereallowed to equilibrate in the laboratory for 24 h.

The interfacial tension of DI water and mineral oilwas determined using a fully automated commercialtensiometer (FTA200, First Ten Angstrom) using thedrop shape method. The calibrations were made usinga water droplet emerging from a needle of known dia-meter for which the interfacial tension is known. Al-though a range of syringe sizes can be used in thepump, it was found that small volume syringes (‹1mL) produced data with the lowest relative standarddeviation. Therefore, for all the experiments in thisstudy, a 1 mL syringe is used. The equilibrium valueof the surface tension of DI water after calibration wasfound to be 72.89 mN/m and that of mineral oil (with-out the addition of surfactant) was found to be 34 mN/mwith a standard deviation of 0.03 mN/m. Droplets wereformed at the tip of a flat needle using an inflow at rateof 1μL/s which was maintained using a programmablesyringe pump. The software provided with the tensio-meter used the Laplace equation to calculate the inter-

126 BioChip J. (2014) 8(2): 122-128

Figure 4. (a) Distance between two consecutive droplets plott-ed as a function of the concentration of surfactant. Dispersedand continuous phase flow rates were kept fixed at Qd==0.2μL/min and Qc==0.8 μL/min respectively. The flow rate ratiowas Q==0.25. (b) Advancing and receding contact angles plott-ed against surfactant concentration for flow rate ratios of Q==1 and Q==0.0125.

Dis

tanc

e be

twee

n tw

o dr

ople

ts(μ

m)

Dyn

amic

con

tact

ang

le(d

egre

es)

800

750

700

650

600

550

500

450

400

180

170

160

150

140

130

120

110

0 1 2 3 4

0 1 2 3 4

Surfactant concentration (% w/w)

Q==1

Q==0.125

(a)

(b)

θa

θa

θr

θr

facial tension from the droplet shape. The interfacialtensions between sample solutions and water weremeasured using the pendant drop method where thependant drop of oil floating up from the dispense tipinto the water phase. The values of interfacial tensionwere measured after the system achieved equilibriumstate. Each experiment was repeated 3 times. The stan-dard deviation in each case remained below 1.0 mN/mwhich confirmed the accuracy of the measurements.

The viscosity of mineral oil is sufficiently high toensure the likelihood of droplet formation in the mic-rochannel due to the existence of a high viscous shear.The viscosity of the oil did not change significantly atdifferent surfactant concentration levels. All the rea-gents were filtered through millipore microfilters withpore size of 0.45 μm before being injected into themicrochannel to minimize contamination, e.g. by for-eign particles. Droplet images obtained from the micro-fluidic experiments were processed using the imageprocessing software packages FTA32 and ImageJ toobtain measurements of the advancing and recedingcontact angles.

The degree of wettability of a PDMS surface caus-ed by the addition of surfactant to a mineral oil wascharacterized through contact angles measurementsmade using a tensiometer. A small piece of PDMS slabwas placed in a flat and optically transparent cuvette.The cuvette was filled with mineral oil containingdifferent concentrations of surfactant ranging from 0to 10% w/w. A small volume of water was pumpedthrough a syringe needle, which was already immersedin an oil, onto the PDMS surface at very slow flowrate of 0.1 μL/s. All contact angles were measuredusing the sessile drop method.

Microchip fabrication and design

The fabrication process of the PDMS microchip con-sisted of the standard photolithography method usingSU-8 negative photoresist mould20-22. The PDMS cur-ing agent and base (Sylgard 184; Dow Corning) weremixed rigorously in the ratio of 1 : 10 for 5 minutes.The mixture was degassed using a vacuum desiccatorand then poured onto the patterned SU-8 master. It wasthen cured in an oven at 65�C for 1 hr and then peeledoff from the master. The inlet and outlet holes werepunched using a 1 mm biopsy punch. The resultingchannel structure was then bonded onto another flatPDMS slab. The bonding of a PDMS replica with a 2mm thick PDMS slab was carried out using the labora-tory made plasma chamber. Both the PDMS replica(with feature side up) and the slab were placed into airplasma for 6 min and then were brought in contact witheach other. After 5 min the sealed microchip was plac-

ed in an oven for 1 hr to enhance the bonding strength.The microfluidic chip design consists of a T-junction

followed by a diverging channel with 180�return bendsas shown in Figure 3(a). The channel dimensions areapproximately 100μm wide×50μm high. The purposeof the return bends is to elongate the channel thus pro-viding increased residence times for long time chemi-cal reactions. Practical examples of such reactionswithin droplets include incubation of cells and proteincrystallization. Droplet breakup occurs by a cross shearflow method and then emerging droplets pass througha divergence which increases the surface to volumeratio. The cross-section of the outlet channel was dou-ble than that of the main channel. This diverging cham-ber was not made so wide as to disrupt the originalsequence of droplets. If the channel is too wide thenthe precise control of the position each droplet is com-promised. When a droplet enters into the divergentsection, the shear stress decreases and drop deforms.

Experimental setup

Mineral oil containing surfactant and DI-water, keptin 1 mL syringes, were loaded into the two inlets ofmicrofluidic chip via PTFE tubing for the purpose ofdroplet formation. Each of the syringes was driven byan individual syringe pump in order to control the flowrate ratio as required. Small volume syringes (‹1 mL)were used in order to attain tight control over the accu-racy of the volumes of liquids dispensed. The outletwas connected to a waste container by a section of tub-ing. The pumps were connected in series and controll-ed via the use of a personal computer (PC) over a ser-ial cable RS232. An inverted microscope (Zeiss Axio-vert 100) and a 1280×1024 resolution charge coupl-ed device (CCD) camera (Sensicam QE, PCO Co.,Ltd.) were used for acquiring droplet images.

An objective lens with a 2.5 times magnificationpower was used to capture the flow field near the wallsof the microchannel. The channel dimensions anddroplet lengths were measured by counting pixels. Thesystem was calibrated against a true length scale. Theimage of a graticule with 0.05 mm graduation markswas captured onto the CCD array with a 2.5× objec-tive lens. By comparing the number of pixels in hori-zontal direction on an imaged graticule with its cor-responding mm scale units, 1 pixel was calculated tobe equivalent to 0.002577 mm==2.577 μm. In order tofully wet the microchannel with the continuous phase,we first introduced oil into the channel. Once the con-tinuous phase filled the entire microchannel, the dis-persed phase was introduced from the lateral channelwhilst maintaining the flow of the continuous phasefrom the main channel in order to facilitate the genera-

BioChip J. (2014) 8(2): 122-128 127

tion of microdroplets. The addition of a surfactant suchas Span 80 to a continuous phase (oil) changes the wet-ting properties of the microchannel walls. Therefore,in order to achieve stabilized initial conditions, foreach of our experiments, the carrier fluid of oil contain-ing a known concentration of surfactant was pumpedinto the main channel at least 1 hr before the dispersedphase was released. Also, before each experiment thechannels were cleaned with ethanol and dried in orderto reduce the risk of contamination. Changes to anyof the operating parameters necessitated a minimumdelay time of about 5 min for the purpose of equilibrat-ing the system.

Acknowledgements Authors acknowledge supportfrom the EPSRC grant No. EP/I019790/1.

References

1. Zheng, B., Roach, L.S. & Ismagilov, R.F. Screeningof protein crystallization conditions on a microuidicchip using nanoliter-size droplets. J. Am. Chem. Soc.125, 11170-11171(2003).

2. Skurtys, O. & Aguilera, J.M. Applications of microui-dic devices in food engineering. Food Biophysics 3,115 (2008).

3. Boonme, P. Applications of microemulsions. J. Cosmet.Dermatol. 6, 223-228 (2007).

4. Theberge, A.B. et al. Microdroplets in microuidics:an evolving platform for discoveries in chemistry andbiology. Angew. Chem. Int. Ed. 49, 5846-5868 (2010).

5. Solvas, X.C.I. & deMello, A. Droplet microuidics:recent developments and future applications. Chem.Commun. 47, 1936-1942 (2011).

6. Lee, W., Walker, L.M. & Anna, S.L. Role of geometryand uid properties in droplet and thread formation pro-cesses in planar ow focusing. Phys. Fluids 21, 032103(2009).

7. Baroud, C.N., Gallaire, F. & Dangla, R. Dynamics ofmicrouidic droplets. Lab Chip 10, 2032-2045 (2010).

8. Thorsen, T., Roberts, R.W., Arnold, F.H. & Quake,S.R. Dynamic pattern formation in a vesicle-genera-ting microuidic device. Phys. Rev. Lett. 86, 4163-4166(2001).

9. Nisisako, T., Tori, T. & Higuchi, T. Droplet formationin a microchannel network. Lab Chip 2, 24-26 (2002).

10. Garstecki, P., Fuerstman, M.J., Stone, H.A. & White-sides G.M. Formation of droplets and bubbles in amicrouidic T-junction scaling and mechanism ofbreak-up and mechanism of break-up. Lab Chip 6,437-446 (2006).

11. Bashir, S., Rees, J.M. & Zimmerman, W.B. Investi-gation of pressure profile evolution during confinedmicrodroplet formation using a two-phase level setmethod. Int. J. Multiphase Flow 60, 40-49 (2014).

12. Dreyfus, R., Tabeling, P. & Willaime, H. Orderedand disordered patterns in two-phase flows in micro-channels. Phys. Rev. Lett. 90, 144505 (2003).

13. Armani, D., Liu, C. & Aluru, N. Re-configurable fluidcircuits by PDMS elastomer micromachining, MEMS’99: 12th IEEE International conference on Micro Elec-tro Mechanical Systems, USA. 222-227 (1999).

14. Xu, J.H., Luo, G.S., Li, S.W. & Chen, G.G. Shearforce induced monodisperse droplet formation in amicrouidic device by controlling wetting properties.Lab Chip 6, 131-136 (2006).

15. Xu, J.H., Li, S.W., Tan, J., Wang, Y.J. & Luo, G.S.Controllable preparation of monodisperse O/W andW/O emulsions in the same microuidic device. Lang-muir 22, 7943-7946 (2006).

16. Van der Graaf, S., Nisisako, T., Schroen, C.G.P.H., vander Sman, R.G.M., Boom, R.M. Lattice Boltzmannsimulations of droplet formation in a T-shaped micro-channel. Langmuir 22, 4144-4152 (2006).

17. Dussan, E.B. On the spreading of liquids on solid sur-faces: static and dynamic contact angles. Ann. Rev.Fluid Mech. 11, 371-400 (1979).

18. Dorrer, C. & Ruhe, J. Advancing and receding motionof droplets on ultrahydrophobic post surfaces. Langmuir22, 7652-7657 (2006).

19. Extrand, C.W. Model for contact angles and hysteresison rough and ultraphobic surfaces. Langmuir 18, 7991-7999 (2002).

20. Duffy, D.C., McDonald, J.C., Schueller, O.J.A. &Whitesides, G.M. Rapid prototyping of microfluidicsystems in poly(dimethylsiloxane). Anal. Chem. 70,4974-4984 (1998).

21. McDonald, J.C., Du_y, D.C., Anderson, J.R., Chiu,D.T., Wu, H., Schueller, O.J.A. & Whitesides, G.M.Fabrication of microuidic systems in poly(dimethyl-siloxane), Electrophoresis 21, 27-40 (2000).

22. Sia, S.K. & Whitesides, G.M. Microuidic devicesfabricated in poly(dimethylsiloxane) for biologicalstudies. Electrophoresis 24, 3563-3576 (2003).

128 BioChip J. (2014) 8(2): 122-128