Persistent reduction of boiling incipience of ethanol on biphilic porous textured surfaces

https://doi.org/10.1016/j.ijmultiphaseflow.2021.103739Get rights and content

Highlights

  • Enhanced ethanol boiling by combining surface texturing and wettability patterning.

  • Over-twofold higher heat transfer rates obtained on micro- and nanoporous surfaces.

  • Facilitated heterogeneous nucleation suggestive of a vital role by surface nanobubbles.

Abstract

Boiling of highly wetting fluids is of great interest to thermal management of high-powered electronic devices, enhancement of which conventionally relies on functional modifications of surface shape and structure so as to promote bubble generation and growth. In this paper we attempt to combine the approach of porous texturing with a novel technique of surface wettability engineering to enhance saturated pool boiling of ethanol. Thin layers of microporous and nanoporous surface topographies were chemically deposited on an aluminum substrate by anodization in phosphoric acid and sulfuric acid, respectively. Boiling on the textured surfaces recorded over-twofold increases in terms of the heat transfer coefficient compared with that on a plain smooth surface, which can be attributed to a proliferation of active nucleation sites. The boiling incipience point on both porous surfaces, however, exhibited an interesting dependence on the initial wetting state of the surface. With the application of amphiphobic coatings of fluoropolymer modified with halloysite nanotubes, we managed to engender a biphilic (i.e., spatially alternating hydrophobicity and hydrophilicity) pattern on the porous-textured surfaces. The repeat control experiments on the hybrid surfaces showed an equally efficient mode of nucleate boiling, whose inception, by contrast, seemed to occur at consistently low superheats. The remarkable reduction (by more than 80%) of the minimum superheat at the boiling onset hints at an alternative nanobubble-related mechanism for heterogeneous bubble nucleation, which is notably less affected by incondensable gas, to the classic vapor-trapping-cavity model.

Introduction

Driven by emerging technologies such as artificial intelligence, Big Data analytics, and IoT (Internet of Things), a demand for energy-intensive large-scale data infrastructure was widely projected in the early 2000s to cause a runaway rise in electricity use and to potentially exacerbate the ongoing crisis of climate change (Dayarathna et al., 2016). In a rare positive turn of events, however, such worst-case scenario failed to materialize as the world only saw modest growth in energy consumption related to data centers despite the nearly six-fold increase in the aggregate workload in the intervening decades (Masanet et al., 2020), thanks in part to recent innovations in thermal management strategies for IT equipment. Key amongst those is a rapidly maturing technology of two-phase immersion liquid cooling (2PILC), in which phase-change heat transfer schemes of boiling and condensation are relied on to safely and efficiently remove massive amounts of heat generated by increasingly powerful semiconductor computer chips. Adoption of 2PILC promises to sharply improve power usage effectiveness (PUE), which is used to measure data-center energy efficiency, from the present 1.3–1.5 for a standard air cooling system to a level approaching 1.0 (i.e., the ideal case) (Alkharabsheh et al., 2015; Li and Kandlikar, 2015). In order to continue the current trend of energy efficiency improvement, further advances in novel cooling technologies such as 2PILC are needed, which entails the task of significant enhancement of boiling heat transfer (BHF).

In a typical pool boiling configuration where a heating surface is immersed in initially convection-free liquid, heat transfer occurs in successive stages of (with ascending surface heat flux) natural convection, nucleate boiling, transient boiling, and finally film boiling (Rohsenow, 1971). From the perspective of thermal engineering, the mode of nucleate boiling is of particular interest because it can provide significantly higher heat transfer rates than either single-phase heat conduction or convection. In the isolated bubble regime of saturated boiling, a bubble emerges from a pre-existing vapor nucleus embedded in the surface and continues to expand by drawing heat from both the heating wall (largely through evaporation of a thin liquid microlayer trapped underneath the bubble (Moore and Mesler, 1961; Yabuki and Nakabeppu, 2014)) and from a surrounding layer of superheated liquid (through evaporation at the bubble surface (Han and Griffith, 1965; Liao et al., 2004)). Upon reaching a critical size (Cole, 1967) (which itself depends on subtle interplay between buoyancy, surface tension (Jo et al., 2016), and hydrodynamic drag (Zeng et al., 1993)), the growing bubble begins to lift off the wall, which is responsible for enhanced microconvection (Kim, 2009). Along with the departing bubble is pumped away superheated liquid from the surface. The void formed in its aftermath is then filled by replenishing cool liquid. The ensuing regeneration of the thermal boundary layer adjacent to the heating wall—which features highly efficient transient heat conduction (Han and Griffith, 1965; Mikic and Rohsenow, 1969)—helps keep the surface temperature low during the ‘quiet’ waiting period between cycles of bubble nucleation and growth. As the surface heat flux continues to rise, more bubble nucleation sites on the surface become activated. The consequently shrinking distance between neighboring nucleation sites gives rise to more complex thermal and hydrodynamic interactions (Zhang and Shoji, 2003), which, when combined with the conjugate nature of heat transfer at the solid-liquid interface, introduce extra nonlinearity to the boiling behavior that is akin to chaos dynamics (Shoji, 2004). As a result, a comprehensive deterministic description of BHF mechanisms still remains elusive despite decades’ worth of research efforts (Dhir, 2006; Kim, 2009).

In contrast to the halting progress towards a refined theoretical understanding of boiling behavior, empirical endeavors to induce more efficient BHF through surface engineering have yielded significant improvement over the past half century (Liang and Mudawar, 2018; Lu and Kandlikar, 2011). Since the early days of Jakob and Fritz (Rohsenow, 1971; Webb, 1981), surface finish has long been recognized as one of the crucial factors in deciding nucleate boiling performance. More than simply increasing the surface area, the “pits and scratches” in a sandblasted surface were shown to engender more vigorous boiling compared with an untreated plain surface, which can be attributed to a sudden proliferation of viable nucleation sites. However, such roughness-induced enhancement turned out to be short-lived. The observation of subsequent performance decline (often within a day)—known as the “aging” effect—was associated with cavity flooding of those artificial nucleation sites. Thereafter pioneering mechanistic models of heterogeneous bubble nucleation (Hsu, 1962) inspired the development of novel surface geometries and shapes that are capable of providing more stable vapor traps. Notable examples include reentry grooves (Deng et al., 2016; Sun et al., 2017), porous coatings (Godinez et al., 2019; Jun et al., 2016), and micro-nano hierarchical structures (Kim et al., 2015; Rioux et al., 2014), whose activation as bubble nucleation sites is greatly facilitated by constantly maintaining liquid menisci in the subsurface structure. Especially noteworthy is the commercially successful “pore-and-tunnel” three-dimensional surface geometry (THERMOEXCEL-E type, Hitachi) (Pastuszko, 2012). Under the “suction-evaporation” mode unique to such a microstructured surface (Nakayama et al., 1980), liquid can be directly sucked into the tunnel cavity through the open pores in the perforated top surface by the pressure difference created in the aftermath of bubble departure, which is capable of generating extremely efficient heat transfer.

One alternative approach to enhancing BHF focuses on active manipulation of bubble behavior and contact-line dynamics. By means of microchannels (Jaikumar and Kandlikar, 2016), selective deposition of low-thermal-conductivity materials (Rahman et al., 2015; Rahman and McCarthy, 2017), and laser texturing (Zakšek et al., 2020), separate pathways for escaping vapor and replenishing liquid (Kandlikar, 2013) have been shown to form on the resulting heterogeneous surfaces with periodic in-plane variations of surface shape, thermophysical properties, and topography, respectively. Furthermore, tuning the geometry of surface heterogeneities appears to maximize the heat transfer coefficient in a manner that mimics a resonance effect dependent on the reduced spacing between neighboring nucleation sites (Rahman et al., 2015). On the other hand, both the suppression of stick-and-slip motion of the triple-phase contact line on chemically heterogeneous substrates (Jaikumar and Kandlikar, 2017) and the creation of additional contact-line regions on a microgrooved surface with variable shape (Raghupathi and Kandlikar, 2017) were demonstrated to be particularly effective in delivering a remarkable augmentation of the critical heat flux (CHF) (i.e., the maximum heat flux that is allowed under the regime of nucleate boiling).

It is interesting to note that modification of surface wettability can satisfy both the design principles above. Not a new concept by any means, early attempts at promoting nucleate boiling by depositing a thin coating of nonwetting material date back to the 1960s (Webb, 1981). Surface cavities with hydrophobic-coated interior walls were found to be able to sustain more stable bubble nucleation at a lower surface superheat (which is defined as the excess temperature relative to the saturation temperature of the working fluid). Furthermore, significant heat transfer enhancement, in addition to the exceptionally early onset of boiling, was reported in the case of an otherwise hydrophilic stainless steel surface spray-coated with hydrophobic Teflon spots (Young and Hummel, 1964), which incidentally represented the very first biphilic (namely, mixed-wettability) surface feature (Betz et al., 2013). The enhancement was claimed to result from altered bubble dynamics following contact line crossing the wetting boundary (Shen et al., 2020). However, application of hydrophobic coatings to the entire surface is counterproductive because the resulting extensive vapor coverage over the surface could lead to deteriorating heat transport instead. (New evidence suggests that as long as contact-line motion is pinned under the initial Wenzel state achieved by adopting a proper degassing procedure (Allred et al., 2019, 2018), enhanced boiling can be realized even on homogeneously superhydrophobic surfaces.) The recent advances in superhydrophobic materials (Attinger et al., 2014) have led to a revival of the idea of wettability-enhanced boiling surfaces. Specifically, the current iteration of biphilic surface design employs state-of-the-art surface engineering techniques including nanostructuring (Kim et al., 2017), photo-induced superhydrophilicity (Takata et al., 2003), hydrophobic fluoropolymer coatings and silanization (Allred et al., 2018; Kim et al., 2018), and photolithography (Betz et al., 2010) to generate more pronounced wettability contrast as well as a precise control of the pattern geometry and shape that improves manipulation of bubble dynamics (Shen et al., 2017) and optimizes BHF (Jo et al., 2014).

Wettability engineering, especially surface hydrophobization, has long been considered ineffective for highly wetting fluids (Webb, 1981), enhanced boiling of which relies on roughness-based methods instead (Kalani and Kandlikar, 2013). In our previous study (Shen et al., 2019), a novel amphiphobic coating was used to fabricate a first-of-its-kind biphilic boiling surface designed for low-surface-tension ethanol, which achieved three-fold heat transfer enhancement compared with a plain copper surface. In the present work, we aim to expand upon that study to explore the possibility of further enhancement of ethanol boiling through combining the treatment of wettability patterning and the more conventional approach of surface texturing. The results show that proliferating bubble populations benefit from the deposition of porous structures—which in turn gives rise to significantly higher heat transfer rates—while the deposition of patterned amphiphobic coatings atop such micro- and nanoporous surfaces leads to consistently low superheats at the onset of nucleate boiling (ONB) by endowing the hybrid surfaces with more stable bubble nucleation sites.

The remainder of the paper is organized as follows. In Sec. 2 we will outline the technical aspects of the experimental setup and procedures. The experimental data of saturated ethanol boiling under the effects of enhanced surface porosity and wettability modifications are presented in Sec. 3, which is to be followed by a discussion with regard to the divergent routes to heterogeneous bubble nucleation in boiling on the biphilic textured surface in Sec. 4. Finally, the paper is concluded in Sec. 5.

Section snippets

Test rig

As shown schematically in Fig. 1, the boiling test facility used in the present study consisted of a 5-liter transparent Plexiglass cylindrical chamber (with a height of 450 mm and an inner diameter of 120 mm) which was filled with ethanol through a filling port at the bottom. A K-type thermocouple (with a measurement uncertainty of ±0.12 K) was placed in a fixed position below the liquid surface to monitor the bulk temperature, which was maintained at the saturation value Tsat (corresponding

Experimental results

In this section, we present the results of the saturated ethanol boiling experiments using various boiling surfaces, which show remarkable heat transfer enhancement on the microporous and nanoporous surfaces. Owing to its reliance on the vapor-trapping-cavity mechanism, however, bubble nucleation on the textured surfaces is found to be wildly unstable between the experimental runs under the ‘dry’ and ‘wet’ initial conditions. By contrast, a more consistent mode of boiling incipience appears to

Discussion

We summarize in Table 3 the ONB superheats for the various test surfaces under the ‘dry’ and ‘wet’ initial conditions, respectively. It thus becomes clear that a significant delay in initiating nucleate boiling of ethanol took place on a porous-structured surface—dependent on the ‘dryness’ of the surface—while consistently early boiling onset became possible if the textured surface was further modified by (partial) hydrophobic functionalization. It is argued in (Lee et al., 2012) that a

Conclusions

In this study, we have investigated the influences of surface porous texturing and heterogeneous wettability modifications on saturated pool boiling of ethanol. Microporous and nanoporous structures were first deposited on an aluminum substrate using phosphoric-acid anodization and sulfuric-acid anodization, respectively. The enhanced surfaces were then partially hydrophobized by coatings of fluoropolymer modified by halloysite nanotubes. To study the effect of surface ‘dryness’ on the boiling

CRediT authorship contribution statement

Biao Shen: Formal analysis, Visualization, Writing – original draft. Takeshi Hamazaki: Investigation. Kohei Kamiya: Investigation. Sumitomo Hidaka: Methodology, Resources. Koji Takahashi: Conceptualization, Funding acquisition. Yasuyuki Takata: Conceptualization, Funding acquisition, Writing – review & editing, Supervision. Junji Nunomura: Resources, Methodology, Writing – review & editing. Akihiro Fukatsu: Resources, Methodology. Yoichiro Betsuki: Resources, Methodology.

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgement

This paper is based on results obtained from a project, JPNP14004, commissioned by the New Energy and Industrial Technology Development Organization (NEDO) of Japan. We are grateful to Prof. Atsushi Takahara and Dr. Wei Ma of Kyushu University, Japan for their assistance with preparing the amphiphobic coating solution.

References (57)

  • J.S. Kim et al.

    Effect of surface roughness on pool boiling heat transfer of water on hydrophobic surfaces

    Int. J. Heat Mass Transf.

    (2018)
  • C.Y. Lee et al.

    Morphological change of plain and nano-porous surfaces during boiling and its effect on nucleate pool boiling heat transfer

    Exp. Therm. Fluid Sci.

    (2012)
  • G. Liang et al.

    Pool boiling critical heat flux (CHF) – Part 1: review of mechanisms, models, and correlations

    Int. J. Heat Mass Transf.

    (2018)
  • J. Liao et al.

    The influence of the bulk liquid thermal boundary layer on saturated nucleate boiling

    Int. J. Heat Fluid Flow

    (2004)
  • R. Pastuszko

    Pool boiling for extended surfaces with narrow tunnels - Visualization and a simplified model

    Exp. Therm. Fluid Sci.

    (2012)
  • I. Pioro

    Experimental evaluation of constants for the Rohsenow pool boiling correlation

    Int. J. Heat Mass Transf.

    (1999)
  • B. Shen et al.

    Enhanced pool boiling of ethanol on wettability-patterned surfaces

    Appl. Therm. Eng.

    (2019)
  • M. Shoji

    Studies of boiling chaos: a review

    Int. J. Heat Mass Transf.

    (2004)
  • Y. Sun et al.

    Pool boiling performance and bubble dynamics on microgrooved surfaces with reentrant cavities

    Appl. Therm. Eng.

    (2017)
  • T. Yabuki et al.

    Heat transfer mechanisms in isolated bubble boiling of water observed with MEMS sensor

    Int. J. Heat Mass Transf.

    (2014)
  • L.Z. Zeng et al.

    A unified model for the prediction of bubble detachment diameters in boiling systems—I. Pool boiling

    Int. J. Heat Mass Transf.

    (1993)
  • L. Zhang et al.

    Nucleation site interaction in pool boiling on the artificial surface

    Int. J. Heat Mass Transf.

    (2003)
  • S. Alkharabsheh et al.

    A brief overview of recent developments in thermal management in data centers

    J. Electron. Packag.

    (2015)
  • T.P. Allred et al.

    Enabling highly effective boiling from superhydrophobic surfaces

    Phys. Rev. Lett.

    (2018)
  • D. Attinger et al.

    Surface engineering for phase change heat transfer: a review

    MRS Energy Sustain.

    (2014)
  • A.R. Betz et al.

    Do surfaces with mixed hydrophilic and hydrophobic areas enhance pool boiling?

    Appl. Phys. Lett.

    (2010)
  • R. Cole

    Bubble frequencies and departure volumes at subatmospheric pressures

    AIChE J.

    (1967)
  • M. Dayarathna et al.

    Data center energy consumption modeling: a survey

    IEEE Commun. Surv. Tutorials

    (2016)
  • Cited by (8)

    • An efficient aluminum gradient mesh wick for enhancing boiling heat transfer performance

      2024, International Communications in Heat and Mass Transfer
    View all citing articles on Scopus
    View full text