Wh

CMa

b

c

d

a

ARRAA

KHHHW

1

hpepscms[

(

0h

Applied Surface Science 290 (2014) 59– 65

Contents lists available at ScienceDirect

Applied Surface Science

j ourna l ho me page: www.elsev ier .com/ locate /apsusc

etting behavior on hybrid surfaces with hydrophobic andydrophilic properties

hun-Wei Yaoa, Jorge L. Alvaradob,∗, Charles P. Marshc,d, Barclay G. Jonesd,ichael K. Collinsc,d

Dept. of Mechanical Engineering, Texas A&M University, College Station, TX 77843, USADept. of Engineering Technology and Industrial Distribution, Texas A&M University, College Station, TX 77843, USAERDC – Construction Engineering Research Laboratory, 2902 Newmark Dr., Champaign, IL 61826, USADept. of Nuclear, Plasma, and Radiological Engineering, University of Illinois at Urbana-Champaign, Champaign, IL 61801, USA

r t i c l e i n f o

rticle history:eceived 2 September 2013eceived in revised form 28 October 2013ccepted 30 October 2013vailable online 8 November 2013

eywords:ybrid surfaceydrophobicydrophilicater condensation

a b s t r a c t

Hybrid surfaces consisting of a micropillar array of hydrophobic and hydrophilic sites were designedand fabricated to understand the effects of their unique surface morphology and chemistry on dropletcondensation. Droplet impingement experiments have revealed that hybrid surfaces exhibit high contactangles, which is characteristic of purely hydrophobic surfaces. However, little is known about the wettingbehavior of droplets that nucleate and grow on hybrid surfaces during condensation. In fact, condenseddroplets display a distinct wetting behavior during the droplet growth phase which cannot be reproducedby simply impinging droplets on hybrid surfaces. In this study, hybrid surfaces with three different spacingratios were subjected to condensation tests using an environmental scanning electron microscopy (ESEM)and a condensation cell under ambient conditions. For hybrid surfaces with spacing ratio below 2, dropletswere observed to form on top and sides of the micropillars, where they grew, coalesced with adjacentdroplets, and shed after reaching a given size. After shedding, the top surface remained partially dry,which allowed for immediate droplet growth. For hybrid surfaces with spacing ratio equal to 2, a differentwetting behavior was observed, where droplets basically coalesced and formed a thin liquid film which

was ultimately driven into the valleys of the microstructure. The liquid shedding process led to therenucleation of droplets primarily on top of the dry hydrophilic sites. To better understand the nature ofdroplet wetting on hybrid surfaces, a surface energy-based model was developed to predict the transitionbetween the two observed wetting behaviors at different spacing ratios. The experimental and analyticalresults indicate that micropillar spacing ratio is the key factor for promoting different wetting behaviorof condensed droplets on hybrid surfaces.

. Introduction

Dropwise condensation is characterized by significantly highereat transfer, even by an order of magnitude greater when com-ared to filmwise condensation [1,2]. In the past few years,ngineered surfaces have been designed and characterized toromote dropwise condensation by making use of nano- or micro-cale features [3–13]. Most of the recent studies about dropwiseondensation have focused on superhydrophobic surfaces with

icro-scale features. In those cases, it has been found that conden-

ate droplets may partially penetrate the texture of the surfaces3,12,13]. Under those conditions, droplets do not fully exhibit a

∗ Corresponding author. Tel.: +1 9794581900.E-mail addresses: Alvarado@entc.tamu.edu, jalvaradvega@yahoo.com

J.L. Alvarado).

169-4332/$ – see front matter © 2013 Elsevier B.V. All rights reserved.ttp://dx.doi.org/10.1016/j.apsusc.2013.10.188

© 2013 Elsevier B.V. All rights reserved.

hydrophobic behavior, which could inhibit proper droplet sheddingneeded for enhanced dropwise condensation. Condensate dropletsunder those conditions tend to grow to a large size (∼2 mm) untilthey roll off the surface by gravity [14–16]. However, such a sce-nario is detrimental to the overall condensation process since largedroplets have higher thermal resistance than smaller droplets.

Droplet size should also be reduced while increasing the dropletdeparture rate so more surface can become available for dropletrenucleation. This in turn enables greater heat transfer per contactarea [17]. Recent studies have focused on nanostructured surfacesfrom which suspended condensate droplets can spontaneouslyeject or jump when they reach a relatively small droplet diame-ter (less than 100 �m) [5,6,10,11,18]. However, such a movement

could be lost when droplets grow very fast and coalesce with neigh-boring droplets [19,20]. Also, the air entrained in the nanocavitiesfound in those surfaces could contribute to a decrease in heat trans-fer rate [11].

6 rface

bc(iiwtEodfHwfidttegt

2

2

pabawrauSifdF(

E

wistldt

ss

E

0 C.-W. Yao et al. / Applied Su

Hybrid surfaces consisting of a micropillar array of hydropho-ic and hydrophilic sites were designed to promote dropwiseondensation. Although studies have reported on heterogeneousdiscretized) nucleation behavior [7] and distinct wetting behav-or [21] of droplets on hybrid surfaces, they provide little insightnto the role of surface morphology on condensation. In this paper,

e report on the effects of surface morphology and chemistry onhe wetting behavior of condensate droplets on hybrid surfaces.xperimental observations reveal that condensate droplets formn top of each micropillar until they coalesce with neighboringroplets. Once they reach a given size, the coalesced droplets shedrom the surfaces when the micropillar spacing is less than 50 �m.owever, coalesced droplets fill the valleys of the hybrid surfaceshen the micropillar spacing is about 50 �m, forming a thin liquidlm which leads to a combination of dropwise and filmwise con-ensation mechanisms. Once the droplets have shed, the tops ofhe micropillars experience dropwise condensation again. In ordero understand condensation phenomena on hybrid surfaces, theffects of surface morphology and surface chemistry on dropletrowth dynamics were investigated analytically and experimen-ally.

. Surface design and methods

.1. Surface design criterion

It is known that with an increase in surface roughness, dropletsrefer to sit on top of heterogeneous interfaces consisting of peaksnd air gaps [22]. When droplets sit on a surface, the liquid coulde in the Cassie–Baxter wetting state, where droplets sit on top ofir-filled cavities, or in the Wenzel wetting state, where dropletset the space or cavities below them. In order to design and fab-

icate hybrid surfaces that can ensure the Cassie–Baxter condition, simple surface energy based analysis [4,8,13,23–25] has beenndertaken to prescribe the desired wetting state for droplets.urface energy based analysis of a predetermined hybrid surfacencluding its boundary conditions was undertaken by accountingor the surface energy of a characteristic unit cell. Fig. 1a shows aroplet sitting on a hybrid surface in the Cassie–Baxter state [19].or such a condition, the surface energy of the interfacial system 1IS1) within a unit cell is given by:

= �SL1 ×(

a2

(a + b)2(a + b)2

)+ �SV2

×(

4ah +(

1 − a2

(a + b)2

)(a + b)2

)+ �LV

×((

1 − a2

(a + b)2

)(a + b)2

)+ �LV × cos �h(a + b)2 (1)

here �SL1, �SV2, and �LV is the surface tension at the solid–liquidnterface of the hydrophilic surface, surface tension at theolid–vapor interfaces of the hydrophobic surface, and the surfaceension at the liquid–vapor interface, respectively. The micropil-ar width and gap size are denoted as a and b, respectively, and henotes the micropillar height. �h is the droplet contact angle onhe hybrid surface.

Fig. 1b shows a droplet sitting on a hybrid surface in the Wenzeltate [14]. For such a condition, the surface energy of the interfacialystem 2 (IS2) within a unit cell is given by:

= � ×(

a2(a + b)2

)+ � × (4ah) + �

SL1

(a + b)2 SL2 SL2

×((

1 − a2

(a + b)2

)(a + b)2

)+ �LV × cos �h(a + b)2 (2)

Science 290 (2014) 59– 65

where �SL2 is the surface tension at the solid-liquid interface of thehydrophobic surface.

To determine the critical condition corresponding to the tran-sition between interfacial system 1 and interfacial system 2, thesurface energies of both interfacial systems must be the same.Namely, Eqs. (1) and (2) are set equal to each other as follows:

�SV2 ×(

4ah +(

1 − a2

(a + b)2

)(a + b)2

)+ �LV

×((

1 − a2

(a + b)2

)(a + b)2

)

= �SL2 ×(

4ah +(

1 − a2

(a + b)2

)(a + b)2

)(3)

For a liquid droplet on a flat hydrophobic surface, the well-known Young’s equation [21] can be expressed as follows:

cos(�equ) = �SV2 − �SL2

�LV(4)

where �equ is the equilibrium contact angle of a liquid droplet on aflat hydrophobic surface

Eq. (3) can then be simplified by using Eq. (4) as follows:

�critical = cos−1(

A − 1R − A

)(5)

A = a2

(a + b)2

R = 1 + 4h

a(1 + b/a)2

where �critical is the critical contact angle that denotes the transitionbetween interfacial systems 1 and 2.

The critical angle (�critical) obtained by using Eq. (5) was used tocompare with the experimental contact angle data. Since it is diffi-cult to find the equilibrium contact angle under ambient conditions[4,12,26], the average contact angle (�ave) was defined as follows:

�ave = (�adv + �rec)2

(6)

where �adv and �rec are the advancing and receding angles of a liquiddroplet on a flat hydrophobic surface, respectively.

The desired interfacial system 1 or 2 can be predicted using Eq.(5) by specifying the appropriate micropillar width (a), gap size(b), and micropillar height (h). From Eq. (5), the calculated criticalangle (�critical) can be calculated, which can then be compared withthe average contact angle (�ave) obtained using Eq. (6). When themeasured �ave of a droplet on a flat hydrophobic surface is less thancalculated �critical, interfacial system 2 (as shown in Fig. 1b) becomesmore favorable because of its lower surface energy. This explainswhy the hydrophobicity of the hydrophobic materials is not enoughto promote and support interfacial system 1, since droplets wouldtend to fill the surface roughness (i.e. air cavities) to reach a lowersurface energy state. On the other hand, when the measured �ave isgreater than the calculated �critical, the interfacial system 1 becomesmore favorable in terms of surface energy resulting in the settling

of droplets on top of the micropillars. Therefore, Eq. (5) can beused to predict which of the two distinct wetting states may pre-vail on hybrid surfaces consisting of hydrophilic tops or spots andhydrophobic valleys exclusively.

C.-W. Yao et al. / Applied Surface Science 290 (2014) 59– 65 61

ssie–B

2

apsesswacnt

Fpo

Fig. 1. Schematic drawings of (a) the interfacial system 1 (a droplet in the Ca

.2. Fabrication and characterization

Hybrid surfaces consisting of a micropillar array of hydrophobicnd hydrophilic sites were fabricated through a photolithogra-hy process [21]. Three hybrid surfaces with various micropillarpacings were characterized by using techniques such as scanninglectron microscopy (SEM) and time-of-flight secondary ion masspectrometry (TOF-SIMS) accordingly. The fabrication process con-isted of dry-oxidizing a 10 cm SSP (single-side polished) siliconafer in an oxidation tube furnace for 4 h at 1100 ◦C, and growing

pproximately a 250 nm layer of SiO . Then, the desired pattern

2onsisting of an array of 25 × 25 �m squares was developed by spin-ing a SPR-220 photoresist with an AP8000 adhesion promoter. Aotal of three designs were exposed to the photoresist with a mask

ig. 2. (a) EDX spectrum analysis of tower tops (b) EDX spectrum analysis of tower sideresence of the hydrophobic (C2F4)n , and green color indicates the presence of the hydropf the references to color in this legend, the reader is referred to the web version of the ar

axter state) and (b) the interfacial system 2 (a droplet in the Wenzel state).

aligner using edge-to-edge spacings of 25 �m, 37.5 �m, and 50 �m.After submerging in the AZ 400 K solution to reveal the squarepattern, a buffered oxide etchant BOE solution (hydrofluoric acidsolution) was prepared to remove SiO2 around the photoresist-patterned squares. The wafer was loaded into an anisotropic dryetcher (Plasmatherm ICP-DRIE) for 50 min to etch around the pho-toresist squares and make towers at least 75 �m in height. After theetching process, C4F8 plasma was run for 1 min to deposit a 70 nmlayer of hydrophobic material ((C2F4)n) (Teflon-like polymer) onthe whole surface. The wafer was then placed into photoresist strip-per AZ 400 T for 24 h to remove the photoresist and exposed the

SiO2 hydrophilic layer sitting on top of the towers.

An Energy-Dispersive X-Ray Spectroscopy (EDX) instrumentwas used to verify the chemical composition of the hydrophilic and

walls (c) TOF-SIMS of the 50 �m spacing hybrid surfaces: red color indicates thehilic SiO2, (d) SEM picture of the 50 �m spacing hybrid surface. (For interpretationticle.)

62 C.-W. Yao et al. / Applied Surface

Table 1Dimensions and critical angles of hybrid surfaces.

Sample no. a (�m) b (�m) h (�m) Critical angles,�critical (◦) (Eq. (5))

1 22.8 24 75 1012 23.2 38 75 1083 24.6 49 75 113

abh

hoflstSSFaiat

2

seaobqPp

Flm

, micropillar width., gap size., micropillar height.

ydrophobic sites. EDX data verified the presence of silicon andxygen or SiO2 on the tower tops as shown in Fig. 2a. Carbon anduorine signatures were also detected on the tower sidewalls ashown in Fig. 2b, which verified the presence of (C2F4)n. In additiono EDX, Time-of-Flight Secondary Ion Mass Spectrometry (TOF-IMS) was performed to detect oxygen and fluorine signatures iniO2 (hydrophilic sites) and C2F4 (hydrophobic sites) as shown inig. 2c. Scanning electron microscopy was used to verify and char-cterize the morphology and dimensions of the surfaces as shownn Fig. 2d, and described in Table 1. All the hybrid surfaces exhibit

hydrophobic behavior when water droplets gently impinge onhem, as shown in Fig. 3.

.3. Experimental method

The wetting processes of hybrid surfaces during water conden-ation were visualized in situ using an environmental scanninglectron microscopy (ESEM). A 1 cm × 1 cm sample was placed on

holder with a tilt angle of 63–66◦. The sample holder was placedn a Peltier cooling stage mounted inside the ESEM. The electron

eam voltage was set at 15 or 20 keV in order to ensure better imageuality while minimizing beam heating. The temperature of theeltier cooling stage was fixed at −1 ◦C. To start the condensationrocess, the vapor pressure was slowly increased from 5 Torr to

ig. 3. Water droplet sitting on top of hybrid surfaces (a) Sample 1 (25 �m micropil-ar spacing), (b) Sample 2 (37.5 �m micropillar spacing), (c) Sample 3 (50 �m

icropillar spacing).

Science 290 (2014) 59– 65

8 Torr. For condensation under ambient conditions, the growth ofcondensed droplets was also observed using an optical microscopeand a custom-built condensation chamber. A 3 cm × 3 cm samplewas placed on top of a copper surface oriented at 90◦ with respectto the horizontal plane. The cooling copper surface was maintainedat 8 ◦C using a liquid chiller. The temperature of moist air was 23 ◦Cwith a relative humidity of 92%. For a flat hydrophobic surface, con-tact angle measurements were performed using a Goniometer (KSVInstruments CAM 200) with accuracy of 0.1◦. Advancing and reced-ing angles were measured using droplet volumes in the range from5 to 50 �l.

3. Results and discussion

Droplet impingement and condensation experiments were con-ducted to study the effects of surface morphology (i.e. spacing ratio)and surface chemistry of hybrid surfaces on droplet wetting andcondensation behavior. First, the average contact angle that waterdroplets make on a flat hydrophobic ((C2F4)n) surface was mea-sured. The measured average contact angle (�ave) was 112.1◦ withadvancing (�adv) and receding (�rec) contact angles of 125.6◦ and98.6◦, respectively. Table 1 shows analytical (calculated) criticalangles for all the samples. Only Sample 3 has an analytical criticalangle (�critical) of 113◦, which is larger than the measured intrinsicaverage contact angle (�ave) of 112.1◦. That is, the hydrophobicityof the hydrophobic material is not enough to promote interfacialsystem 1. Therefore, when droplets nucleate and grow on Sample 3,the wetting state should resemble interfacial system 2 (IS2) morethan interfacial system 1 (IS1). Namely, droplets should wet theentire surface. By using Eq. (5), water droplets are expected to siton top of the micropillars when using hybrid surfaces with spacingratio below 2 (Samples 1 and 2), but the liquid should fill the gapsof the hybrid surface when the spacing ratio is equal to 2 (Sample3).

However, when droplets gently impinge on Sample 3, the wet-ting behavior closely follows interfacial system 1 as shown inFig. 3(c). Notice that for Sample 3, the measured average contactangle is close to the calculated critical angle (�critical), which sug-gests that impinged droplets could be in a metastable state [27] fora finite amount of time. In a metastable state, droplets could spon-taneously fall into the substrate cavities to reach a lower energystate [27,28]. To validate the notion that a metastable state dropletcould spontaneously transition to a lower energy state, a dropletwas deposited on Sample 3 at ambient conditions as shown in Fig. 4.After few seconds, it was observed using a microscope that thedroplet transitioned from IS1 to IS2. Fig. 4 validates the applica-bility of Eq. (5) in predicting the final wetting state of a droplet ona hybrid surface.

When conducting condensation experiments using Samples 1and 2, tiny droplets were observed to nucleate and grow on the topsand sides of the micropillars, and on the bottom of each sample asshown in Fig. 5a and e (nucleation stage). Droplets located at thetop of the micropillars continued to grow in size and coalesced withneighboring droplets until larger spherical droplets were formed,as shown in Fig. 5b and f (coalescence stage). After coalescence,droplets popped up to the top surface, forming liquid bridges asshown in Fig. 5c and g (pop-up stage). Subsequent droplet coales-cence continued to take place, resulting in the formation of largedroplets with droplet diameter exceeding 100 �m as seen in Fig. 5dand h (large droplet stage).

When condensation was observed in Sample 3, the initial

droplet nucleation (Fig. 5i) followed the same behavior as seen onthe other hybrid surfaces. However, once coalescence took place,Sample 3 depicted different wetting dynamics as seen in Fig. 5j. Ascondensation proceeded, coalescence and droplet collapse led to

C.-W. Yao et al. / Applied Surface Science 290 (2014) 59– 65 63

F m 2 onv

tuahltoc

wccdc

FS

ig. 4. Sequence images of the transition from interfacial system 1 to interfacial syste1).

he formation of a thin liquid film. The formation of the thin liq-id film is attributed to the transition of the interfacial system to

lower energy state as seen in Fig. 5k. Once the thin liquid filmad formed, it pulled the droplets sitting on top of the micropil-

ars into it. In essence, this process triggered a cleaning effect ofhe top surfaces of the micropillars, which led to the renucleationf droplets at the surface as seen in Fig. 5l (dropwise and filmwiseondensation stage) and Video 3.

In order to validate the condensation wetting behavior observedhile using the ESEM, hybrid surfaces were placed on top of a

ondensation cell. An optical microscope was used at ambient

onditions to image the whole condensation process. Figs. 6 and 7epict time-sequence images of surfaces experiencing dropletondensation (top-view). The condensation experiments were

ig. 5. Environmental scanning electron microscopy (ESEM) time sequence images of waupplementary-video v2–v4).

Sample 3 (50 �m micropillar spacing hybrid surface) (see the Supplementary-video

conducted over an hour on all the hybrid surfaces. For the hybridsurfaces with spacing of 25 �m and 37.5 �m, the observed wettingbehavior corresponds fairly well with the dropwise condensationprocess observed when using the ESEM. The surfaces depict anidentifiable nucleation stage (Fig. 6a), droplet growth stage (Fig. 6b),coalescence stage (Fig. 6c), and large droplet stage (Fig. 6d). Similarcondensation behavior was observed in Sample 1 (25 �m spacing).

For the hybrid surface with a 50 �m-spacing, the condensationbehavior also compares fairly well with the dropwise and filmwisecondensation processes observed when using the ESEM. The sur-faces depict several condensation stages (or phases), including the

nucleation stage (Fig. 7a), coalescence stage (Fig. 7b), collapse stage(Fig. 7c), and the dropwise-filmwise condensation stage (Fig. 7d ande). Even though it is challenging to visualize the formation of a thin

ter droplets during the growing and coalescence stages on hybrid surfaces (see the

64 C.-W. Yao et al. / Applied Surface

Fh

lfifbau

a1w

Fh

ig. 6. (a–d) Sequential micrographs of condensation process on 37.5 �m spacingybrid surface (see the Supplementary-video v5).

iquid film when viewing Fig. 7c and d, Fig. 5k and l clearly depictlmwise condensation behavior on the 50 �m-spacing hybrid sur-

ace. Fig. 7e shows a large portion of the thin liquid film demarcatedy its own triple contact line (TCL). In summary, the condensationctivities depicted in Figs. 5 and 7 were characterized by a contin-ous process of droplet coalescence, renucleation and growth.

Sample 3 with a micropillar spacing of 50 �m can be regardeds a superhydrophobic surface since the droplet contact angle is50.7◦ ± 0.2◦. However, during the condensation process, liquidas observed to fill the cavities of it, and droplets simultaneously

ig. 7. (a–e) Sequential micrographs of condensation process on the 50 �m spacingybrid surface (see the Supplementary-video v6).

Science 290 (2014) 59– 65

sat on top of the hydrophilic sites. The wetting characteristics ofdropwise and filmwise condensation observed in Sample 3 are sim-ilar to the hemi-wicking [29] behavior, which is the intermediatecondition between spreading and imbibition as wetting mecha-nisms. In hemi-wicking [29], the liquid fully penetrates the textureof the engineered surface, leaving no droplets on the top side ofthe micropillars when hydrophilic spikes are present. However, onhybrid surfaces, the sides of the micropillars are hydrophobic innature (Sample 3), but the droplets still collapse vertically aroundthem and coalesce with other droplets. The observed behavioris similar to the hemi-wicking mechanism, which conventionallyapplies only to hydrophilic surfaces. However, visual observationsand surface energy based analysis presented in this article indicatethat the surface free energy of the interfacial system 2 (IS2) reachesa thermodynamically-favored state as shown in Figs. 1b and 5l,which resembles the hemi-wicking behavior.

Even though a well-design array of hydrophilic spots andhydrophobic surfaces has led to a combination of dropwise andfilmwise condensation mechanisms for removing droplets, morestudies are still needed to determine which wetting mechanism(s)enhances or inhibits the overall condensation heat transfer pro-cess. Under ideal circumstances, discrete minute droplets formingon top of and rolling off hybrid surfaces are still desired from theheat transfer point of view. However, impinging water droplets thatexhibit high contact angle on hybrid surfaces may not always resultin enhanced dropwise condensation since they are susceptible toinstabilities, which could lead to a rapid transition to filmwise con-densation. Therefore, when designing hybrid surfaces, the mainobjective is to ensure that interfacial system 1 (IS1) droplets arestable enough during the condensation process.

In this study, the effect of microscale patterned structures onwetting behavior has been thoroughly investigated for hybrid sur-faces. Other effects such as the level of hydrophobicity of thehydrophobic coatings on hybrid surfaces still remain to be stud-ied in the future. Future studies should consider the use of otherfabrication techniques and various hydrophobic materials. How-ever, Enright et al. [30] and Miljkovic et al. [31] have observedthat the wetting dynamics of condensed droplets are not nec-essarily sensitive to various hydrophobic coatings, which can bepredicted when using patterned structured surfaces. Therefore,similar droplet dynamics including droplet shedding could beexpected when using various hydrophobic materials.

4. Conclusions

In summary, we have reported on two different condensa-tion wetting mechanisms of hybrid surfaces, namely dropwise anddropwise-filmwise, which arise from the hydrophobic and super-hydrophobic properties of the hybrid surfaces. A surface energybased model has been formulated that is capable of predicting thetransition between droplet wetting states characteristic of drop-wise and dropwise-filmwise condensation modes. The model canalso be used to design hybrid surfaces with distinct droplet wet-ting characteristics in order to avoid or induce unfavorable orfavorable wetting states during condensation. Experimental resultsreveal that surface morphology has a strong influence over thefinal wetting state of condensing droplets which can lead to theself-removal of droplets from the hydrophilic sites, which in turnshould facilitate enhanced droplet renucleation and growth at thesame patterned sites.

Acknowledgements

The authors would like to thank Dr. David Garrett for his assis-tance during droplet condensation characterization experiments

rface

cl

A

f2

R

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

C.-W. Yao et al. / Applied Su

onducted using the environmental scanning electron microscopeocated at the University of North Texas.

ppendix A. Supplementary data

Supplementary data associated with this article can beound, in the online version, at http://dx.doi.org/10.1016/j.apsusc.013.10.188.

eferences

[1] J.W. Rose, Dropwise condensation theory and experiment: a review, Proceed-ings of the Institution of Mechanical Engineers, Part A: Journal of Power andEnergy 216 (2002) 115–128.

[2] J.W. Rose, Condensation heat transfer fundamentals, Chemical EngineeringResearch and Design 76 (1998) 143–152.

[3] K.A. Wier, T.J. McCarthy, Condensation on ultrahydrophobic surfaces and itseffect on droplet mobility: ultrahydrophobic surfaces are not always waterrepellant, Langmuir 22 (2006) 2433–2436.

[4] R.D. Narhe, D.A. Beysens, Growth dynamics of water drops on asquare-pattern rough hydrophobic surface, Langmuir 23 (2007)6486–6489.

[5] C.-H. Chen, Q. Cai, C. Tsai, C.-L. Chen, G. Xiong, Y. Yu, Z. Ren, Dropwise conden-sation on superhydrophobic surfaces with two-tier roughness, Applied PhysicsLetters 90 (2007) 173103–173108.

[6] J.B. Boreyko, C.-H. Chen, Self-propelled dropwise condensate on superhy-drophobic surfaces, Physical Review Letters 103 (2009) 184501.

[7] K.K. Varanasi, M. Hsu, N. Bhate, W. Yang, T. Deng, Spatial control in the hetero-geneous nucleation of water, Applied Physics Letters 95 (2009) 094101.

[8] R.D. Narhe, W. González-Vinas, D.A. Beysens, Water condensation on zinc sur-faces treated by chemical bath deposition, Applied Surface Science 256 (2010)4930–4933.

[9] C. Dietz, K. Rykaczewski, A.G. Fedorov, Y. Joshi, Visualization of dropletdeparture on a superhydrophobic surface and implications to heat transferenhancement during dropwise condensation, Applied Physics Letters 97 (2010)033104.

10] X. Chen, J. Wu, R. Ma, M. Hua, N. Koratkar, S. Yao, Z. Wang, Nanograssedmicropyramidal architectures for continuous dropwise condensation,Advanced Functional Materials 21 (2011) 4617–4623.

11] N. Miljkovic, R. Enright, E.N. Wang, Effect of droplet morphology on growthdynamics and heat transfer during condensation on superhydrophobic nanos-

tructured surfaces, ACS Nano 6 (2012) 1776–1785.

12] R.D. Narhe, D.A. Beysens, Water condensation on a super-hydrophobic spikesurface, EPL (Europhysics Letters) 75 (2006) 98.

13] C. Dorrer, J. Rühe, Condensation and wetting transitions on microstructuredultrahydrophobic surfaces, Langmuir 23 (2007) 3820–3824.

[

[

Science 290 (2014) 59– 65 65

14] P. Dimitrakopoulos, J.J.L. Higdon, On the gravitational displacement of three-dimensional fluid droplets from inclined solid surfaces, Journal of FluidMechanics 395 (1999) 181–209.

15] L.R. Glicksman, A.W. Hunt Jr., Numerical simulation of dropwise condensation,International Journal of Heat and Mass Transfer 15 (1972) 2251–2269.

16] C. Dietz, K. Rykaczewski, A.G. Fedorov, Y. Joshi, Visualization of dropletdeparture on a superhydrophobic surface and implications to heat transferenhancement during dropwise condensation, Applied Physics Letters 97 (2010)033103–033104.

17] S. Kim, K.J. Kim, Dropwise condensation modeling suitable for superhydropho-bic surfaces, Journal of Heat Transfer 133 (2011) 081502–081508.

18] N. Miljkovic, R. Enright, Y. Nam, K. Lopez, N. Dou, J. Sack, E.N. Wang, Jumping-droplet-enhanced condensation on scalable superhydrophobic nanostructuredsurfaces, Nano Letters 13 (2012) 179–187.

19] J. Cheng, A. Vandadi, C.-L. Chen, Condensation heat transfer on two-tier super-hydrophobic surfaces, Applied Physics Letters 101 (2012) 131904–131909.

20] N. Miljkovic, R. Enright, Y. Nam, K. Lopez, N. Dou, J. Sack, E.N. Wang, Jumping-droplet-enhanced condensation on scalable superhydrophobic nanostructuredsurfaces, Nano Letters, 13 (2012) 179–187.

21] C.W. Yao, T.P. Garvin, J.L. Alvarado, A.M. Jacobi, B.G. Jones, C.P. Marsh, Dropletcontact angle behavior on a hybrid surface with hydrophobic and hydrophilicproperties, Applied Physics Letters 101 (2012), 111605-111605.

22] A.B.D. Cassie, S. Baxter, Wettability of porous surfaces, Transactions of the Fara-day Society 40 (1944) 546–551.

23] W. Li, A. Amirfazli, A thermodynamic approach for determining the contactangle hysteresis for superhydrophobic surfaces, Journal of Colloid and InterfaceScience 292 (2005) 195–201.

24] T. Liu, W. Sun, X. Sun, H. Ai, Thermodynamic analysis of the effect of the hierar-chical architecture of a superhydrophobic surface on a condensed drop state,Langmuir 26 (2010) 14835–14841.

25] K. Rykaczewski, W.A. Osborn, J. Chinn, M.L. Walker, J.H.J. Scott, W. Jones, C. Hao,S. Yao, Z. Wang, How nanorough is rough enough to make a surface superhy-drophobic during water condensation? Soft Matter 8 (2012) 8786–8794.

26] E.L. Decker, B. Frank, Y. Suo, S. Garoff, Physics of contact angle measurement,Colloids and Surfaces A: Physicochemical and Engineering Aspects 156 (1999)177–189.

27] M. Sbragaglia, A.M. Peters, C. Pirat, B.M. Borkent, R.G.H. Lammertink, M.Wessling, D. Lohse, Spontaneous breakdown of superhydrophobicity, PhysicalReview Letters 99 (2007) 156001.

28] Q.S. Zheng, Y. Yu, Z.H. Zhao, Effects of hydraulic pressure on the stability andtransition of wetting modes of superhydrophobic surfaces, Langmuir 21 (2005)12207–12212.

29] J. Bico, Wetting of textured surfaces, Colloids and Surfaces A: Physicochemicaland Engineering Aspects 206 (2002) 41–43.

30] R. Enright, N. Miljkovic, A. Al-Obeidi, C.V. Thompson, E.N. Wang, Condensationon superhydrophobic surfaces: the role of local energy barriers and structurelength scale, Langmuir 28 (2012) 14424–14432.

31] N. Miljkovic, R. Enright, E.N. Wang, Modeling and optimization of superhy-drophobic condensation, Journal of Heat Transfer 135 (2013) 1–14.