Enhanced condensation heat transfer using porous silica inverse opal coatings on copper tubes

Abstract

Phase-change condensation is commonplace in nature and industry. Since the 1930s, it is well understood that vapor condenses in filmwise mode on clean metallic surfaces whereas it condenses by forming discrete droplets on surfaces coated with a promoter material. In both filmwise and dropwise modes, the condensate is removed when gravity overcomes pinning forces. In this work, we show rapid condensate transport through cracks that formed due to material shrinkage when a copper tube is coated with silica inverse opal structures. Importantly, the high hydraulic conductivity of the cracks promote axial condensate transport that is beneficial for condensation heat transfer. In our experiments, the cracks improved the heat transfer coefficient from ≈ 12 kW/m² K for laminar filmwise condensation on smooth clean copper tubes to ≈ 80 kW/m² K for inverse opal coated copper tubes; nearly a sevenfold increase from filmwise condensation and identical enhancement with state-of-the-art dropwise condensation. Furthermore, our results show that impregnating the porous structure with oil further improves the heat transfer coefficient by an additional 30% to ≈ 103 kW/m² K. Importantly, compared to the fast-degrading dropwise condensation, the inverse opal coated copper tubes maintained high heat transfer rates when the experiments were repeated > 20 times; each experiment lasting 3-4 h. In addition to the new coating approach, the insights gained from this work present a strategy to minimize oil depletion during condensation from lubricated surfaces.

Figure 1

Schematic depictions of different modes of condensation. (a) Filmwise condensation (FWC) where the condensate forms a continuous liquid film on a chemically clean (high surface energy) surface. The wall is maintained at a lower temperature ($T_{\text{w}}$) than the surrounding saturated vapor ($T_{\text{sat}}$). The temperature and velocity profiles are shown for laminar film condensation where the flow-rate based Reynolds number is < 30. (b) Dropwise condensation (DWC) on hydrophobic (low surface energy) surface. It is essential for dropwise condensation that some promoter material, for example organic compound, be present on the surface. Periodic removal of large droplets clears the surface for renewed droplet nucleation and growth. (c) Inverse opal condensation (IOC) on a porous inverse opal structure. Condensate seeps into the porous structure by displacing the air in the pores. Preferential condensate transport through high hydraulic conductivity micro cracks improves the heat transfer rate. (d) Condensation on a slippery liquid-infused porous surface (SLIPS). The porous interstices are impregnated with a chemically matched oil. Compared to DWC, droplets depart with smaller radius and higher frequency in SLIPS condensation.

Figure 2

Schematic depiction of the fabrication procedure to produce inverse-opal-coated copper tubes. (a) Colloidal solution is prepared by adding 0.5 wt% colloidal polystyrene particles (395 nm diameter) into a solution of 99.4 wt% deionized water and 0.1 wt% tetraethyl orthosilicate (TEOS). (b) A copper tube coated with silica via atomic layer deposition (ALD) and oxygen plasma cleaned is inserted vertically in the colloidal solution to allow colloidal self-assembly as the solution evaporates. (c) The polystyrene beads are dissolved by immersing the copper tube in toluene. (d) A porous inverse opal structure composed of a network of voids is created when the sacrificial polystyrene beads dissolve in toluene. (e) During the drying stage, interconnected micro cracks featuring large hydraulic conductivity form between islands of inverse opals.

Figure 3

Scanning electron microscope (SEM) images of an inverse opal coating at different magnifications. The low magnification SEM images in (a) and (b) show fabrication defects (cracks) that formed when the supporting matrix shrank during drying. The inset in (a) shows a vanishing contact angle for a water droplet deposited on the porous structure. The high magnification images in (c) and (d) show the pores which resulted from dissolution of the sacrificial polystyrene beads (≈ 395 nm diameter). The contact points between the polystyrene beads that was not wetted by the solution during co-assembly created interconnected pores (≈ 215 nm diameter).

Figure 4

Schematic of the experimental setup constructed for phase change condensation study. The environmental chamber maintains saturation condition by eliminating NCGs from the system. The vapor that enters the chamber condenses on the copper tube that is maintained below the saturation temperature by circulating chilled water from a water bath heat exchanger. To minimize heat loss to the surrounding environment, both the chamber and the vapor generator are insulated using a fiberglass sheet. The chamber has a view port to observe droplet nucleation, growth, and departure.

Figure 5

Time-lapse images of water condensation on copper tubes subject to varying surface treatments, showing different modes of condensation. (a) FWC on a smooth plasma treated copper tube, (b) DWC on smooth hydrophobized copper tube, (c) IOC on silica inverse opal-coated copper tube, and (d) SLIPS condensation on oil-impregnated porous structure. In all cases except IOC, condensate was drained vertically downwards when gravity overcame pinning forces. During IOC, however, the condensate was transported preferentially in the axial direction through the cracks. Depending on the geometry, size, and orientation of the crack, the condensate moved axially to the right or to the left, but never vertically.

Figure 6

Heat transfer measurement. (a) Heat flux and (b) heat transfer coefficient as a function of logarithmic mean temperature difference. The slope in (a) is the heat transfer coefficient ($h_c$). The heat transfer coefficient for DWC and IOC are nearly identical at ≈ 80 kW/m² K; a sevenfold increase from FWC (≈ 12 kW/m² K). The heat transfer coefficient increased further by ≈ 30% to ≈ 103 kW/m² K when the silica pores were impregnated with oil (SLIPS condensation). The classical Nusselt model for laminar film condensation is used to validate our heat transfer measurements.

Figure 7

Coating durability. (a) The silane coating on the copper tube degraded after 5–8 experimental runs. The heat transfer coefficient for DWC decreased from ≈ 79 kW/m² K (first experiment) to ≈ 13 kW/m² K (eighth experiment). The enhanced heat transfer coefficient for the inverse opal coated copper tube (b) and the oil-impregnated copper tube (c) remained nearly the same at ≈ 80 kW/m² K (IOC) and ≈ 102 kW/m² K (SLIPS) when condensation experiments were conducted for the first, tenth, and twentieth times. This result shows the durability of our coating.

Figure 8

Environmental SEM. The test sample was attached to a cold stage that was subcooled by 2 °C. Time-lapse images show that vapor condensed preferentially near cracks (t = 32 s and 44 s). When condensation continued, the condensate overflowed (t = 74 s) the porous structure. The condensate was transported through the cracks (t = 88 s) and the surface dried out (t = 92 s) when vapor supply was discontinued.