0
Research Papers

# Condensation on Superhydrophobic Copper Oxide Nanostructures

[+] Author and Article Information
Ryan Enright

Department of Mechanical Engineering,
Massachusetts Institute of Technology,
77 Massachusetts Avenue, Cambridge, MA 02139;
Stokes Institute, University of Limerick,
Limerick, Ireland
e-mail: ryan.enright@alcatel-lucent.com

Nicholas Dou

Department of Mechanical Engineering,
Massachusetts Institute of Technology,
77 Massachusetts Avenue,
Cambridge, MA 02139

Youngsuk Nam

Department of Mechanical Engineering,
Massachusetts Institute of Technology,
77 Massachusetts Avenue,
Cambridge, MA 02139;
Kyung Hee University,
Yongin, Korea

Evelyn N. Wang

Department of Mechanical Engineering,
Massachusetts Institute of Technology,
77 Massachusetts Avenue,
Cambridge, MA 02139
e-mail: enwang@mit.edu

1Present address: Thermal Management Group, Efficient Energy Transfer (ηET) Dept., Bell Labs Ireland, Alcatel-Lucent Ireland Ltd., Blanchardstown Business and Technology Park, Snugborough Road, Dublin 15, Ireland.

2Corresponding authors.

Contributed by the Heat Transfer Division of ASME for publication in the Journal of Heat Transfer. Manuscript received July 2, 2012; final manuscript received May 5, 2013; published online July 26, 2013. Guest Editors: G. P. “Bud” Peterson and Zhuomin Zhang.

J. Heat Transfer 135(9), 091304 (Jul 26, 2013) (12 pages) Paper No: HT-12-1346; doi: 10.1115/1.4024424 History: Received July 02, 2012; Revised March 05, 2013

## Abstract

Condensation is an important process in both emerging and traditional power generation and water desalination technologies. Superhydrophobic nanostructures promise enhanced condensation heat transfer by reducing the characteristic size of departing droplets via coalescence-induced shedding. In this work, we investigated a scalable synthesis technique to produce functionalized oxide nanostructures on copper surfaces capable of sustaining superhydrophobic condensation and characterized the growth and departure behavior of the condensed droplets. Nanostructured copper oxide (CuO) films were formed via chemical oxidation in an alkaline solution resulting in dense arrays of sharp CuO nanostructures with characteristic heights and widths of ≈1 μm and ≈300 nm, respectively. To make the CuO surfaces superhydrophobic, they were functionalized by direct deposition of a fluorinated silane molecular film or by sputtering a thin gold film before depositing a fluorinated thiol molecular film. Condensation on these surfaces was characterized using optical microscopy and environmental scanning electron microscopy to quantify the distribution of nucleation sites and elucidate the growth behavior of individual droplets with characteristic radii of ≈1–10 μm at supersaturations ≤1.5. Comparison of the measured individual droplet growth behavior to our developed heat transfer model for condensation on superhydrophobic surfaces showed good agreement. Prediction of the overall heat transfer enhancement in comparison to a typical dropwise condensing surface having an identical nucleation density suggests a restricted regime of enhancement limited to droplet shedding radii $<~$2.5 μm due to the large apparent contact angles of condensed droplets on the fabricated CuO surfaces. The findings demonstrate that superhydrophobic condensation typified by coalescence-induced droplet shedding may not necessarily enhance heat transfer and highlights the need for further quantification of the effects of surface structure on nucleation density and careful surface design to minimize parasitic thermal resistances.

<>

## Figures

Fig. 1

SEM images (a) and (b) and XRD pattern (c) of the copper oxide nanostructures, after 1 min (a) and 5 min (b,c) of oxidation. (d) SEM image of a FIB milled sample showing a cross-section of the nanostructured copper surface after a 10 min oxidation step. The arrows indicate the approximate extent of the Cu2O and Cu2O + CuO regions.

Fig. 2

Schematic of the optical microscopy experimental set-up. Dry N2 was supplied from a cylinder with the flow rate measured using a rotometer. A three-way valve was used to route the N2 supply either directly to the enclosure or through a temperature-controlled water reservoir via a sparging head and then to the enclosure. The enclosure humidity was monitored using a humidity probe located ≈1 cm from the mounted sample. Images were captured at either 40× or 100× magnification using a CMOS camera mounted to an upright microscope.

Fig. 3

Droplet apparent contact angle as a function of the droplet diameter extracted from the ESEM data on the Au/thiol (red ○) and silane (blue □) functionalized CuO surfaces (pv = 1300 ± 75 Pa, Tw = 283 ± 1.5 K, S = 1.07 ± 0.1). The solid curve is defined as θapp=cos-1(rp/R)+90 deg with rp=1.5μm. The dashed dot curves represent the bounds of Eq. (5) for rp=1.5 ± 0.5μm. The horizontal dashed line represents the macroscopically measured apparent contact angle, θapp≈165 deg. The inset shows a typical ESEM image captured during the droplet growth process on the silane functionalized CuO surface. (b) Diagram showing the evolution of the droplet shape predicted by Eq. (5) for 90 deg≤θapp≤170 deg in steps of 20 deg (droplets bounded by solid curves). Once the advancing state is reached the droplet grows with constant θapp (droplet bounded by dashed curve).

Fig. 4

Emergent droplet morphology on the functionalized CuO surfaces in the coalescence-dominated growth stage. Partially wetting Cassie behavior with jumping droplets emerged on the Au/thiolated CuO surface where 〈L〉=0.5N-0.5≈8.1μm obtained by (a) focusing at the surface (top) and confirmed by (b) focusing through the droplets (below) to show the wetting state (S = 1.51 ± 0.05, N = 3.83 × 109 m−2). The arrow in the inset of (b) points to a light-absorbing region surrounded by a light-reflecting region indicative of the partial-wetting morphology (inset scale bar: 10 μm). Mixed-mode wetting behavior with pinned droplets on the silanated CuO surface where 〈L〉=0.5N-0.5< 2.2μm obtained by (c) focusing at the surface (top) and confirmed by (d) focusing through the droplets (below) to show the wetting state (S = 1.48 ± 0.05, N > 5 × 1010 m−2). (e) Time-lapse images of condensation on the silane-coated CuO surface during ESEM imaging. The dashed and solid circles indicate droplet groups before and after coalescence, respectively. ESEM conditions: pv = 800 ± 75 Pa and Tw = 276 ± 1.5 K (S = 1.07 ± 0.1).

Fig. 5

Nucleation site spatial distribution. (a) Coordinate map showing the position of the nucleation sites (●) and arrows indicating their nearest neighbor captured using optical microscopy at 100× magnification on the Au/thiol-functionalized CuO surface at t = 10 s after the start of the experiment (S = 1.51 ± 0.05, N = 3.83 × 109 m−2). (b) Cumulative probability distribution of the nucleated droplet nearest neighbors (○) compared to the predictions for a random distribution,P=1-e-NπL2 (solid line). The mean separation distance between nucleation sites is given by 2LN=1. The horizontal bars represent the bin width.

Fig. 6

Droplet jumping to pinning transition. (a) The excess liquid/vapor surface energy was estimated by considering the difference in energy between states E1 and E2. Energy is required to overcome the work of adhesion to form a liquid/vapor interface of area 2Ap for the two pinned necks of the coalescing droplets. (b) Excess surface energy compared to the work of adhesion, |ΔE/W|, as a function of the droplet separation distance, L, divided by the droplet pinned base diameter, 2rp. Three values of rp (=1 μm, 1.5 μm, and 2 μm) are shown for each surface. Increasing rp results in smaller values of |ΔE/W|. For rp = 1μm, the model predicts |ΔE/W| = 0.07 and |ΔE/W| = 6.27 for the silanated CuO (〈L〉/(2rp) = 1.1, □) and thiolated CuO (〈L〉/(2rp) = 4.05, ○), respectively. The shaded region (〈L〉/(2rp)≤1) marks the transition to the Wenzel state. The horizontal bars for each point show ±〈L〉.

Fig. 7

Individual droplet heat transfer model. (a) Schematic of the droplet on the condensing surface growing in the partially wetting morphology. (b) Droplet thermal resistance diagram showing the droplet curvature (ψc), liquid–vapor interface (ψi), droplet conduction (ψd), hydrophobic coating (ψhc), CuO nanostructure (ψCuO), liquid bridge (ψw) and Cu2O under layer (ψCu2O) thermal resistances.

Fig. 8

Model prediction of individual droplet growth rates averaged for 12 individual droplets. (a) The experimentally measured droplet diameters as a function of time (○) are compared to the individual droplet growth model (solid curve) with rp = 1.5 μm, δCuO = 1.5 μm. ESEM conditions: pv = 800 ± 75 Pa and Tw = 276 ± 1.5 K (S = 1.07 ± 0.1). Model solutions were obtained for ΔT = 0.034 K, which was within the uncertainty of the measured temperatures and pressures in the ESEM chamber. This value was chosen based on the best fit between the model and experimental growth rate data. The inset shows the experimental data, the model predictions and a fitted R∝t1/3 scaling (dashed curve) in log coordinates. The error bars correspond to uncertainty in the measured droplet radius. (b) The key thermal resistances normalized to the total thermal resistance corresponding to (a) as a function of droplet radius. The vertical line delineates the transition from radius-dependant apparent contact angle (θapp(R)) to a fixed contact angle equal to the macroscopically measured apparent advancing contact angle (θaCB) at 2 R = 11 μm. The thermal resistance components indicated in the plot are the conduction resistance of the droplet volume pinned within the nanostructures, ((ψhc+ψCuO)−1+(ψw+ψhc)−1)−1, the interface curvature resistance (ψc), the interfacial resistance (ψi), the Cu2O layer resistance (ψCu2O) and the droplet bulk resistance (ψd).

Fig. 9

Droplet number density as a function of droplet diameter for steady-state condensation averaged over several cycles of droplet growth, coalescence-induced jumping and regrowth for ESEM conditions: pv = 800 ± 75 Pa, Tw = 276 ± 1.5 K, S = 1.07 ± 0.1. Summing over the range of droplet diameters gives a total droplet number density n = 1.28 × 1010 m−2 corresponding to 〈L〉 = 4.42μm according to Eq. (6). The counting error associated with the droplet number density was estimated to be ∼10% at. each size range.

Fig. 10

Overall heat transfer behavior. (a) Predicted overall heat flux ratio qCuO"/qF" as a function of scaled droplet coalescence length L/(2rp) for the CuO nanostructured surface (δCuO = 1.5 μm) compared to a smooth hydrophobic surface for rp = 1 μm, 1.5 μm, and 2 μm (solid curves). The CuO surface shows an enhancement for L/(2rp)→1 and rp≥1.5μm. This behavior is compared to a similar, hypothetical surface with the CuO height reduced to δCuO = 100 nm (dashed curves). The hypothetical surface demonstrates a wider range of enhancement. Modeled conditions: ΔT = 0.034 K, pv = 800 Pa. The inset shows the predicted heat transfer behavior rp = 1 μm, 1.5 μm, and 2 μm with 〈L〉 = 4.42μm. Predicted overall heat flux ratio qCuO"/qF" as a function of droplet coalescence length L with (b) δCuO = 1.5 μm and (c) δCuO = 100 nm for a range of driving temperature differences (0.01 K ≤ ΔT ≤ 0.05 K in steps of 0.01 K) with constant pv = 800 Pa. (d) The values of qCuO"/qF" at L/(2rp) = 1 (solid curves) and |qCuO"/qF"|max (dashed curves) for δCuO = 1.5 μm (□) and δCuO = 100 nm (○) obtained from (b) and (c), respectively.

## Discussions

Some tools below are only available to our subscribers or users with an online account.

### Related Content

Customize your page view by dragging and repositioning the boxes below.

Related Proceedings Articles
Related eBook Content
Topic Collections