Aqueous ionic liquid solutions for boiling heat transfer enhancement in the absence of buoyancy induced bubble departure

International Journal of Heat and Mass Transfer 122 (2018) 354–363

Contents lists available at ScienceDirect

International Journal of Heat and Mass Transfer journal homepage: www.elsevier.com/locate/ijhmt

Aqueous ionic liquid solutions for boiling heat transfer enhancement in the absence of buoyancy induced bubble departure Nirbhay Kumar a, Md. Qaisar Raza a, Debabrata Seth b, Rishi Raj a,⇑ a b

Thermal and Fluid Transport Laboratory, Department of Mechanical Engineering, Indian Institute of Technology Patna, Bihar 801103, India Department of Chemistry, Indian Institute of Technology Patna, Bihar 801103, India

a r t i c l e

i n f o

Article history: Received 22 October 2017 Received in revised form 16 December 2017 Accepted 22 January 2018

Keywords: Ionic liquid Pool boiling Inverted heater Bubble departure Bubble coalescence Critical heat flux

a b s t r a c t Advantage of high heat transfer during boiling is lost due to the absence of buoyancy induced bubble departure in space. Absence of bubble departure with pure water on an inverted heater in earth gravity ð1gÞ resembles boiling behavior in space and is often used to mimic zero-gravity ð0gÞ boiling experiments on earth. Here we perform experiments on an inverted heater to show that unlike water, the aqueous solution of surface active ionic liquid (IL) avoids coalescence to form multiple small bubbles with significantly large wet area on the heater surface. The force of repulsion due to the interaction of ILs adsorbed at the liquid–vapor interface of neighboring bubbles induces a completely passive bubble departure away from the inverted heater surface against the combined effect of buoyancy and surfacetension. Resulting rewetting of the heater surface increases the critical heat flux (CHF) to  950 kW=m2 , which is an enhancement of  4:5 in comparison to pure water. Effect of bulk liquid subcooling and concentration on pool boiling CHF are extensively investigated. The mechanism of CHF is explained with the help of the adsorption dynamics of IL at the solid–liquid and liquid–vapor interface of bubbles. Ó 2018 Elsevier Ltd. All rights reserved.

1. Introduction Boiling heat transfer is fundamental to many household processes and industrial applications such as cooking, water purification, chemical processing, thermal management of electronics, refrigeration and air conditioning, and cryogenic fuel storage [1– 4]. The relatively high heat transfer during boiling is usually attributed to the ebullition cycle which comprises of bubble nucleation, growth, buoyancy-induced bubble departure, and rewetting of the heater surface in a cyclic manner (Fig. 1a) [5,6]. Ebullition cycle, which strongly depends on gravity, is absent in space ðg  0Þ [7– 12], and, on an inverted heater in earth gravity ð1gÞ [13–15] where buoyancy opposes bubble departure. In the absence of bubble departure, nucleating bubbles coalesce together to form a big primary bubble and cover the entire heater surface (dry patch in Fig. 1b and c) which abruptly increases the heater surface temperature and results in a premature critical heat flux (CHF). A summary of the studies in literature with pool boiling data in microgravity ðg  0Þ and on an inverted heater in earth gravity ð1gÞ is shown in Table 1. The CHF data in 0g and 1g are compa-

⇑ Corresponding author. E-mail address: [email protected] (R. Raj). https://doi.org/10.1016/j.ijheatmasstransfer.2018.01.101 0017-9310/Ó 2018 Elsevier Ltd. All rights reserved.

rable and significantly lower than the CHF ðP 1000 kW=m2 [1]) in earth gravity ð1gÞ. Accordingly, much of the research in this field is focused on devising innovative strategies to induce the bubble departure to resemble the typical ebullition cycle in 1g and improve boiling heat transfer in 0g and 1g. For example, use of active approaches such as acoustic field [16] and electric field [17] have been proposed in the literature to facilitate the bubble departure by applying external body forces on the bubble. However, such strategies make the overall system complex, energy intensive, and less reliable. It is probably due to these limitations that despite the high latent heat of vaporization, thermal management of spacecraft, except for the heat pipe [18], is usually based on the single-phase cooling systems which rely on the low sensible heat of the fluid. However, the size and power requirements of typical space platforms are on an ever-increasing trend and the need for developing advanced cooling strategies based on boiling are increasingly realized [19]. We recently demonstrated (S.N. 6, Table 1) a completely passive mechanism of bubble departure against buoyancy on an inverted heater setup where both surface-tension and buoyancy forces act upwards ("Þ and push bubbles towards the heater surface [15]. In comparison to water where bubbles gradually coalesce together to form a large dry patch on the heater surface, surfactant additives were shown to avoid bubble coalescence and form multiple smal-

N. Kumar et al. / International Journal of Heat and Mass Transfer 122 (2018) 354–363

355

Nomenclature C D D FB FS h q00 T

concentration (mol/l) bubble diameter (m) diffusion coefficient (m2/s) buoyancy force (N) surface-tension force (N) heat transfer coefficient (kW/m2 K) heat flux (kW/m2) temperature (°C)

Greek symbols r surface-tension (N/m) C surface concentration (mol/m2) sD time-scale of diffusion (ms)

Fig. 1. (a) Boiling with water in earth gravity ð1gÞ. Buoyancy induced bubble departure is observed. Ebullition cycle is present. (b) Boiling with water in space ð0gÞ, (c) Boiling with water on an inverted heater in earth gravity ð1gÞ. Ebullition cycle is absent, in the absence of buoyancy induced bubble departure, coalesced big vapor-bubble covered the entire heater area.

ler non-coalescing bubbles. These bubbles not only reduced the dry patches on the heater surface but were also observed to depart downwards away from the inverted heater surface (#), even against the combined effect of buoyancy (") and surface-tension force ("). Bubble departure frequency of  15  16 Hz resulted in  2:4 enhancement in both the CHF and the heat transfer coefficient (HTC), in comparison to the pure water. In this work, we use surface active ionic liquid (SAIL) 1-butyl-3methylimidazolium octyl sulfate ([BMIM][OS]) as additive in the base fluid (water) for further heat transfer enhancements in the absence of buoyancy induced bubble departure. Ionic liquids (ILs) are mostly non-volatile, non-flammable, and thermally stable organic salts entirely composed of ions. Some ILs are also biodegradable [20,21]. Molten potassium chloride is an IL; whereas a solution of potassium chloride in water is an ionic solution. These salts (e.g. KCl or NaCl) need to heat to several hundred degrees to melt. The term ‘ionic liquid’ was selected with due care since the phrase ‘molten salt’ made an imperfect image of these solvents

Subscripts sat saturation sub subcooling sup superheat Abbreviation CHF critical heat flux HTC heat transfer coefficient IL ionic liquid ppm parts per million SAIL surface active ionic liquid

as being high temperature and corrosive media. However, in reality, ILs show wide liquidous temperature range [21,22]. They have many applications in areas of science and technology, such as, in the absorption refrigeration system [23,24], heat exchangers [25], lubrication [26], space propulsion [27], stabilization of metal nanoparticles [28,29], chemical industry [21], electrochemistry [30,31], nanotribology [32], and in separation science [33,34], among others. ILs with surface active properties are called SAILs. Many of these SAIL molecules are generally amphiphilic in nature and researched as a suitable alternative to surfactants for various applications [35]. For example, bubbles in an aqueous solution of such ILs avoid coalescence to form vapor foam, very similar to the aqueous surfactant solutions. Here we use an amphiphilic SAIL as an additive in water to avoid bubble coalescence and improve the boiling heat transfer performance in the absence of bubble departure during pool boiling on an inverted heater in earth gravity. Subcooled pool boiling experiments were performed with aqueous IL ([BMIM][OS]) solution, aqueous surfactant (SDS) solutions, and pure water on an inverted heater in earth gravity. We illustrate that [BMIM][OS] avoids coalescence significantly to induce bubble departure in excess of 30 Hz which is almost twice in comparison to SDS. The twofold enhancement in departure frequency is attributed to the relatively small value of the time-scale of diffusion in comparison to aqueous SDS solution. Experimental results on the effect of concentration of IL/surfactant on CHF reveal that the mechanism of CHF is dictated by the adsorption dynamics at both, the solid–liquid, and, the liquid–vapor interfaces. A maximum heat flux of  950 kW=m2 was dissipated at an optimum concentration of CMC/64 of [BMIM][OS].

2. Experimental facility 2.1. Solution preparation As received SAIL 1-butyl-3-methylimidazolium octyl sulfate ([BMIM][OS]) (Sigma Aldrich, purity P 95%) and surfactant Sodium Dodecyl Sulphate (SDS) (Sigma Aldrich, purity P 98:5%) were used in this work. The critical micelle concentration (CMC) of [BMIM][OS] and SDS are  10; 000 ppm and  2500 ppm, respectively. The test fluids (aqueous solution of [BMIM][OS] and aqueous solution of SDS) were prepared in a separate container (other than the test container) by adding the required concentration of [BMIM][OS] and SDS in pure water (Milli-Q, Merck). The test fluids were then stirred properly with the help of a magnetic stirrer for two hours to allow proper mixing of [BMIM][OS]/SDS in water.

356

N. Kumar et al. / International Journal of Heat and Mass Transfer 122 (2018) 354–363

15 36 208* 500* 50 Water SDS (2500 ppm)

1g

13.8 – Water

1g

340*

Saturated pool boiling was performed with different heater orientations Pool boiling was performed on a downward facing heater with varying gap Experiments were performed on an inverted heater. Bubble departure against buoyancy was observed with surfactants 1g

1.9 1.1 1.8 18.8 86.4 58.5 124.4 230 17–21 13-15 14–16 0 0g

n-pentane CFC-113 Water Water

5.7 229* 33 0g FC-72

4.7 134.2 26 0g FC-72 Raj et al. [7]

Kim et al. [8]

Oka et al. [9]

Guo et al. [13]

Su et al. [14]

Raza et al. [15]

1.

2.

3.

4.

5.

6.

Critical heat flux (CHF). *

DTsub (°C) Gravity Fluid Heater surface Reference S.N.

averaged steady–state temperature of thermocouples T 1 , T 2 , and T 3 were used to estimate the experimental heat flux using Fourier’s law of heat conduction (one-dimension, q00 ¼ k½dT=dx, where k is the thermal conductivity of the copper, dT is the temperature difference between the thermocouples, and dx is the spacing between the thermocouples). The time-averaged temperature T 4 measured at a height of 15 mm far from the heater surface was used to find the actual temperature at the heater surface ðT w Þ by using Fourier’s law of heat conduction (one-dimension). Simulations were performed for estimating the heat loss to the ambient and Teflon insulation. The predicted value of heat flux was found to be in good agreement with the heat flux values estimated using the temperature gradients measured with thermocouples T 1 , T 2 , and T 3 as

Table 1 Pool boiling literature data in microgravity ðg  0Þ and on an inverted heater in earth gravity ð1gÞ.

The copper heater surface was polished with a 600 grit size sand paper to ensure uniform surface roughness prior to the start of each set of experiments. The complete heater assembly and the glass container were then cleaned with acetone followed by DI water wash. Test fluid was poured in the glass container up to a height of 150 mm. The heater assembly was then immersed in the test fluid with the help of a fixture such that boiling surface was facing down. Heating of the heater assembly was initiated using a DC power source (300 V, Agilent, N5751A) and the temperature readings were monitored to ensure steady state boiling conditions. Visualization of the bubble behavior from the side and the bottom was performed using a high-speed (5000 fps) camera (Vision Research, Phantom v7.3) and the temperature and the power data were recorded using a data acquisition system (Agilent, 34972A). After each data run, the voltage was increased in steps and the steady-state conditions were ensured before next round of data acquisition. A COMSOLTM simulation was performed to confirm onedimensional heat conduction near the location of thermocouples T 1 , T 2 , and T 3 . Power input to the cartridge heater was simulated as uniform volumetric heat generation in our model. HTC of 10 kW=m2 K and 1000 kW=m2 K [15], for natural convection with air and water, respectively, were provided as boundary conditions to the model (heater assembly). HTC of the boiling surface was varied until the surface temperature T w from simulation matched with the experimental value. The corresponding theoretical heat flux through the boiling surface (q00predicted ) was estimated. Time-

q00 max (kW/m2)

2.3. Experimental procedure and data reduction

Microheater array 4.2 mm  4.2 mm Microheater array 0.27 mm  0.27 mm Transparent indium oxide film plated on a glass plate 50 mm  50 mm Flat surface (Copper) 50.8 mm diameter Flat surface (Stainless steel) 300 mm diameter Flat surface (Aluminium) 21.5 mm diameter

hmax (kW/m2-K)

Remarks

The sectional view of the three-dimensional computer-aided design model of the experimental setup is shown in Fig. 2. Experimental setup consists of a transparent cylindrical glass enclosure (diameter 130 mm, height 185 mm) which contained the test fluid and the inverted heater block assembly. Silicon cloth heater was wrapped around the enclosure to maintain the subcooling of the bulk fluid with the help of a thermocouple T 5 and a PID (Proportional integral derivative) temperature controller. While a similar experimental setup was used in our previous work [15], here we used a stepped copper block with a flat circular boiling surface (14 mm in diameter) and two cartridge heaters (250 W, 110 V, Make: Marathon) to reach high heat flux values. The copper block was insulated by a Teflon bush to ensure one-dimensional heat flow in the neck region above the boiling surface. Four thermocouples T 1 , T 2 , T 3 , and T 4 were embedded within the heater assembly to estimate the wall temperature and heat flux through the boiling surface. Thermocouples T 5 and T 6 were used to monitor the temperature of the bulk fluid. We used J-type thermocouples with 0:5 mm probe diameter. The thermocouples were calibrated and the uncertainty was estimated to be 0:5  C.

Zero gravity experiment was conducted on the International Space Station to investigate the effect of gravity Effects of subcooling and gravity level were investigated. CHF increased with increasing subcooling Heat transfer reduction was more dominant in case of water

2.2. Experimental setup

357

N. Kumar et al. / International Journal of Heat and Mass Transfer 122 (2018) 354–363

Fig. 2. Pool boiling experimental setup. Heater surface is facing down (inverted) to mimic the zero-gravity condition.

Table 2 Maximum uncertainty in different parameters in fully nucleate pool boiling regime on an inverted heater. Boiling fluid

Aqueous [BMIM][OS] Solution Aqueous SDS Solution Water

Maximum uncertainty

DT sup ð CÞ

q00 ðkW=m2 Þ

h ðkW=m2 KÞ

±0.9 ±0.9 ±0.9

±35 ±39 ±26

±6 ±5 ±6

high speed camera (these videos were played at lower frame rates for better visualization of the bubble departure cycle). The uncertainty in f was based on the statistical deviation between 20 measurements at each data points. Maximum uncertainties in different parameters during pool boiling experiments are listed in Table 2. Please refer our previous work [15] for further details on data reduction and uncertainty analysis. Fig. 3. Comparison of experimentally measured and predicted heat flux values for water and aqueous solutions of [BMIM][OS], and, SDS during pool boiling on an inverted heater.

3. Results and discussion 3.1. Heat transfer

shown in Fig. 3. The good agreement (±10% variation) suggests that the heat loss to the insulation and the fixture have been estimated accurately in our experiments. We estimated the bubble departure frequency ðf Þ by taking the average of 20 departure events, each recorded from the bottom view and the side view visualization of boiling process using the

Pool boiling curves (plot of experimental heat flux q00 versus surface superheat, DT sup ¼ T w  T sat , where T sat is the saturation temperature of the test fluid corresponding to atmospheric pressure) for pure water, aqueous solutions of [BMIM][OS], and, SDS at CMC, at a subcooling of 50  1  C are shown in Fig. 4a. The horizon-

358

N. Kumar et al. / International Journal of Heat and Mass Transfer 122 (2018) 354–363

Fig. 4. (a) Pool boiling curves: Plots of heat flux versus wall superheat with aqueous [BMIM][OS] and SDS solutions, and water. Horizontal arrows indicate the CHF. (b) Corresponding plots of HTC versus heat flux. (c) Side view images of the bubbles with pure water and aqueous [BMIM][OS] solution at CMC during pool boiling. Bubbles depart (orange color dashed arrow) away from the inverted heater surface against the combined effect of buoyancy force (red arrow) and surface-tension force (blue arrow), inset image. (d) Plots of bubble departure frequency versus heat flux for aqueous solutions of [BMIM][OS] and SDS at CMC. Subcooling of the liquid pool was maintained at DT sub ¼ 50  1  C. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

tal arrows indicate the CHF. The corresponding plots of HTC versus heat flux are shown in Fig. 4b. A CHF value of  770 kW=m2 with aqueous [BMIM][OS] solution was obtained. This is an enhancement of  3:7 in comparison to boiling with pure water (CHF  210 kW=m2 ) on the same setup. Conversely, the same value of CHF with [BMIM][OS] suggests an enhancement of  1:6 in comparison to SDS ( 500 kW=m2 ; in present study, and, our previous work [15]). We performed additional experiments to check the repeatability in data without cleaning the heater surface after each run of experiments. The changes in CHF were insignificant, however, slight decrease in HTC similar to our earlier work [15] was observed. These tests confirm the robustness of the boiling data presented in this study. Bubble visualization was performed first to investigate this significant enhancement in heat transfer performance in comparison to boiling with pure water. In the absence of any bubble departure mechanism, nucleating bubbles coalesced together and covered the entire heater surface (dryout of the heater surface) during pool boiling with water (Fig. 4c). Ebullition cycle was absent and the mode of heat transfer changed from boiling/phase-change to conduction through small thermal conductivity vapor trapped within the bubble [15]. This dryout of the heater surface is highly undesir-

able as it severely deteriorates the HTC leading to the sudden rise in the temperature and often burnout of the heater surface (CHF). Bubble behavior changed significantly during boiling with aqueous [BMIM][OS] solution. Nucleation site density significantly increased in comparison to pure water. Bubble coalescence was reduced and multiple smaller bubbles were formed on the heater surface. Moreover, these smaller bubbles were observed to depart away from the heater surface (see supplementary movie) against the combined effect of surface-tension (blue arrow) and buoyancy force (red arrow) as shown in Fig. 4c. These results suggest that similar to our previous work with surfactants [15], the presence of [BMIM][OS] in the aqueous solution provides the required force of repulsion to induce a completely passive bubble departure against the combined effect of buoyancy and surface-tension forces. The maximum bubble departure frequency of 32 Hz with the aqueous [BMIM][OS] solution was significantly (2) higher than those with aqueous SDS solutions ( 16 Hz, Fig. 4d) and can primarily be attributed to the relatively high CHF with [BMIM] [OS] in comparison to SDS. The enhancement in HTC with [BMIM][OS] and SDS in comparison to water (Fig. 4b) can be attributed to the increased nucleation site density due to the adsorption of the respective monomers at the heater surface [36,37].

359

N. Kumar et al. / International Journal of Heat and Mass Transfer 122 (2018) 354–363

Supplementary movie. Side view of bubble departure on an inverted heater during pool boiling with aqueous ionic liquid [BMIM][OS] solution at a subcooling and heat flux of 50 ± 1 °C and 300 kW/m2, respectively.

3.2. Bubble departure mechanism We now discuss the adsorption dynamics of IL monomers on the liquid–vapor interface of bubbles to explain the bubble departure mechanism. [BMIM][OS] molecules are amphiphilic in nature with a hydrophilic head and a hydrophobic tail (Fig. 5a). When a bubble nucleates, [BMIM][OS] monomers gradually adsorb at the liquid–vapor interface of the growing bubbles from the bulk liquid, such that hydrophilic head resides in the liquid-phase and hydrophobic tail resides in the vapor-phase (Fig. 5b inset). When two vapor bubbles come closer to each other, electrostatic repulsion between the adsorbed amphiphilic monomers suppresses bubble coalescence. When a secondary bubble nucleates in the wedge shaped micro-layer region underneath a big primary bubble, it is accompanied with the formation of a thin-liquid film ðHÞ between them as shown in Fig. 5b (inset image). With the growth of this small secondary bubble, the thickness ðHÞ of the thin liquidfilm starts decreasing, resulting in an increase in pressure inside the thin liquid-film called disjoining pressure ðPÞ (Fig. 5c). When the downward component of this force of repulsion (Fig. 5b inset,

orange color dashed arrow) due to the disjoining pressure ðPÞ overcomes the combined surface-tension (blue arrow) and buoyancy force (red arrow), bubble departs away from the heater surface [15]. If this force of repulsion is not sufficient to depart the bubble, then the film thickness reduces to the critical value ðHcritical Þ and ruptures, resulting in bubble coalescence (Fig. 5c, bubble coalescence corresponding to Hcritical ). The value of disjoining pressure depends on the type and properties of ILs and is widely discussed in the foam stability studies [38,39]. The passive bubble departure mechanism discussed above is similar to our previous study in which we reported bubble departure against buoyancy with surfactants [15]. However, the timescale of diffusion ðsD Þ and surface concentration ðCÞ are significantly different with [BMIM][OS]. The time-scale of diffusion ðsD Þ determines the rate of adsorption of the additives at the interface. The lifetime of a bubble (ebullition cycle) during a typical boiling process is of the order of a few milliseconds, hence the relatively small time-scale of diffusion allows quick adsorption of monomers at the liquid–vapor interface, and, relatively larger force of repulsion. We next present the results of surface-tension measurement experiments (Force Tensiometer, Attension) for ionic liquid ([BMIM][OS]) and quantify the associated time-scales of diffusion to explain this improvement over surfactants. The value of ð@ r=@lnCÞT was obtained from the slope of r versus lnC curve to estimate the maximum surface concentration ðCm Þ using Eq. (1). For surfactant (SDS), the value of Cm was taken from literature Table 3 Properties of [BMIM][OS] and SDS at a temperature of 298.15 K. Properties 

DG ðkJ=molÞ D ðm2 =sÞ

Value

Ionic liquid/surfactant

References

28:2 54:4

[BMIM][OS] SDS [BMIM][OS] SDS

Pal et al. [42] Rosen et al. [43] Heintz et al. [44] Wu et al. [40]

[BMIM][OS] SDS

Present Work Wu et al. [40]

8:13  1010 * 4:6  1010

Cm ðmol=m2 Þ

1:21  106 6:27  106

*

Value of diffusion coefficient (D ) is for the highly diluted aqueous solution.

Fig. 5. (a) Chemical structure of the IL ([BMIM][OS]) and surfactant (SDS) used in present work. (b) Schematic representation of IL induced bubble departure mechanism against buoyancy. (c) Variation of disjoining pressure ðPÞ with liquid film thickness (H).

360

N. Kumar et al. / International Journal of Heat and Mass Transfer 122 (2018) 354–363

Fig. 6. (a) Vapor bubble condensation due to subcooled bulk liquid. (b) The CHF mechanism with ILs and surfactants on an inverted heater during pool boiling.

[40]. The value of C and sD for [BMIM][OS] and SDS were calculated using Eq. (2) [41] and (3), respectively. The constant a in Eq. (2) was calculated by using Eq. (4).

Cm ¼ 1=nRTð@ r=@lnCÞT

ð1Þ

C ¼ Cm ðC=C þ aÞ

ð2Þ

sD ¼ C2 =C 2 D

ð3Þ



a ¼ 55:3eðDG

=RTÞ

ð4Þ

where R is universal gas constant (8.314 J=mol:KÞ, @ r=@lnC is the slope of the r versus lnC plot, T is absolute temperature ðKÞ, 

n ¼ 2 (for ionic liquid), and DG is free energy of adsorption in kJ=mol. Values of properties used to calculate sD and C are shown in Table 3. The sD and C for aqueous [BMIM][OS] solution ( 0:002 ms and Fig. 7. Plots of CHF versus subcooling for aqueous solutions of [BMIM][OS] and SDS at CMC.

1:18  106 mol=m2 Þ are smaller in comparison to aqueous SDS

solution ( 1 ms and 6  106 mol=m2 Þ at CMC. The relatively small bubble departure frequency with SDS, despite the large value of C,

Fig. 8. Bottom view images of vapor bubbles at different subcooling and heat fluxes, for aqueous solutions of [BMIM][OS] and SDS at CMC, during pool boiling on an inverted heater.

N. Kumar et al. / International Journal of Heat and Mass Transfer 122 (2018) 354–363

can be attributed to the large value of sD for SDS in comparison to [BMIM][OS]. Moreover, each molecule of [BMIM][OS] consists of two amphiphilic monomers, whereas SDS only has one such monomer per molecule (Fig. 5a). The resulting strong force of repulsion reduces bubble coalescence effectively, in turn ensuring bubble departure even at higher heat fluxes and is primarily responsible for the increased CHF in comparison to SDS. 3.3. CHF mechanism Several models such as hydrodynamic [45], microlayer [46], and hot and dry spot [47,48], among others, have been developed to explain the mechanism of CHF during pool boiling. While these models successfully predict the CHF on an upward facing heater configuration where buoyancy induced bubble departure is the primary heat transfer mechanism, they are not suitable for boiling on an inverted heater where buoyancy induced bubble departure is absent. Fundamentally, CHF is observed when the vapor-generation rate exceeds the vapor-removal rate during boiling. The very low CHF values are observed during boiling with water on an inverted heater in absence of buoyancy induced bubble departure. However, IL and surfactant solutions do not allow coalescence and instead force bubble downwards away from the inverted heater surface. Non-coalescing bubbles departing downwards are exposed to the subcooled liquid wherein they release the associated thermal energy via condensation (Fig. 6a). Bubble removal allows rewetting of the heater surface by the surrounding colder fluids. While the number of departing bubbles increases with the heat flux (Fig. 4d), the rate at which these bubbles condense decreases due to the vapor-crowding underneath the inverted heater surface (Fig. 6b). At higher heat fluxes, the vapor crowding below the heater surface increases the residence time of the bubbles growing on the heater surface. Continued heating eventually evaporates the wedge-shaped thin-liquid film ðdÞ between the bubbles (blue color shaded region between the bubbles in

361

Fig. 6b. Evaporation of the liquid-film between the otherwise non-coalescing bubbles forces coalescence to form a big bubble, which covers the entire heater surface. Large bubble inhibits the supply of the fresh liquid to the heater surface. Hence, the CHF is observed when the heat flux increases to a critical value beyond which the vapor-generation rate exceeds the vapor-removal rate from the heater surface. Quick condensation of the departing bubbles at higher subcoolings weakens the vapor crowding below the heater surface to facilitate bubble departure and improve rewetting. Forced coalescence is delayed to higher heat fluxes eventually enhancing the CHF. In order to qualitatively confirm this dependence, we performed pool boiling experiments with aqueous [BMIM][OS] and SDS solutions (at CMC) at various subcooling of DT sub ¼ 30  1  C, 20  1  C, and 10  1  C (Fig. 7). The effective bubble size was found to decrease with the increase in subcooling (Fig. 8) due to enhanced condensation. Subsequently, vapor crowding was decreased and the CHF was found to increase with subcooling for IL and surfactant. 3.4. CHF enhancement The heat transfer and visualization results presented so far suggest that the IL/surfactant monomers in the bulk adsorb at the liquid–vapor interface of bubbles (Fig. 6b, right inset image) to promote bubble departure, however, at the cost of stronger vapor crowding which adversely affects the heat transfer. Conversely, it is well known that ILs/surfactants adsorb at the heater surface (solid–liquid interface, Fig. 6b, left inset image) to promote nucleation. The adsorption of the ILs/surfactants monomers at the solid– liquid interface changes the wetting behavior of the surface. The ILs/surfactants monomers present in the aqueous solution adsorb at the solid–liquid interface, such that hydrophilic head is in contact with solid surface and the hydrophobic tail in contact with the bulk liquid. The hydrophobic tail face the bulk liquid and acts as a hydrophobic coating (non-wetting), thus promoting nucle-

Fig. 9. (a) Plots of pool boiling CHF versus [BMIM][OS] and SDS concentration in bulk liquid. Inset images show the bubble behavior of [BMIM][OS] at a heat flux of  300 kW=m2 at different concentrations (green arrow). Subcooling of the liquid pool was maintained at DT sub ¼ 50  1  C ð50 SÞ and 30  1  C ð30 SÞ. (b) Plots of surface concentration (primary axis) and diffusion time-scale (secondary axis) versus [BMIM][OS] and SDS concentration in bulk liquid. (c) Side view visualization showing vaporcrowding effect of [BMIM][OS] with concentration at a heat flux of  300 kW=m2 . (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

362

N. Kumar et al. / International Journal of Heat and Mass Transfer 122 (2018) 354–363

ation sites [36,49,50]. As a result, while the decrease in the nucleation site density due to the decrease in IL/surfactant concentration would weaken vapor crowding and allow easy rewetting of the heater surface [36], it would also weaken the force of repulsion between bubbles. This would in turn reduce the bubble departure frequency eventually allowing coalescence and hence the formation of large bubbles. These opposing trends suggest that the CHF values can be further enhanced by optimizing the concentration of IL/surfactant. Accordingly, we performed experiments at various concentrations (up to CMC) of aqueous solutions of [BMIM][OS] and SDS at the subcoolings of 50  C and 30  C. The plots of CHF versus bulk concentration ðCÞ are shown in Fig. 9a. The plot of sD and C as a function of C for [BMIM][OS] and SDS in the water is shown in Fig. 9b. As expected, the observed CHF values were not a monotonic function of the bulk concentration of ILs/surfactants (Fig. 9a). CHF, which was adversely affected by vapor crowding (access of fresh liquid to the heater surface was inhibited), increased with decreasing concentration. However, significant reduction in the concentration of IL/surfactant increases the time-scale of diffusion ðsD Þ, resulting in insufficient adsorption of monomers at the liquid–vapor interface of bubbles (Fig. 9b). Consequently, the tendency to avoid the bubble coalescence (Fig. 9c, left) and facilitate the bubble departure decreases. Hence, upon decreasing the IL/surfactant concentration in our experiments, the CHF was observed to increase only up to a critical/optimum concentration of CMC/64 (Fig. 9c, middle), beyond which a decrease in CHF with decrease in IL/surfactant concentration was observed (Fig. 9c, right). For example, the highest value of CHF at the subcooling of 50  1  C was  950 kW=m2 for [BMIM][OS] while it was  890 kW=m2 for SDS (dotted vertical lines at CMC/64 in Fig. 9a). Moreover, the maximum CHF was observed at the same concentration of IL/surfactant at both subcoolings. The ionic liquid/surfactant concentration levels we tested are widely spaced and insufficient to confirm this trend. We feel additional tests at many intermediate concentrations are required to confirm this trend. The highest heat flux of  950 kW=m2 using aqueous [BMIM] [OS] solution in our work is considerably larger than the maximum heat flux values reported for 0g/1g hus far in literature (see Table 1). Such encouraging heat flux numbers suggest that the bubble departure mechanism demonstrated here can be adopted to design a gravity independent compact vapor chamber with very high heat dissipation capabilities. Such devices can replace the bulky single-phase based thermal management solutions employed in space, essentially increasing the energy-to-mass ratio of space based infrastructure, which in turn may allow heavy savings in launch and maintenance costs.

4. Conclusions In summary, we report the results from pool boiling experiments on an inverted heater with aqueous SAIL ([BMIM][OS]) solution to demonstrate a passive bubble departure mechanism similar to the surfactants. We show that the force of repulsion due to the interaction of ILs adsorbed at the liquid–vapor interface of neighboring bubbles is sufficient to overcome the combination of buoyancy and surface-tension to induce a completely passive bubble departure downwards away from the inverted heater surface. The mechanism of CHF was explained with the help of adsorption dynamics of [BMIM][OS] and SDS at the heater surface and liquid– vapor interface of the bubbles. Relatively small time-scale of diffusion of ionic liquid [BMIM][OS] in comparison to the surfactant SDS resulted in a twofold increase in bubble departure frequency. Effective condensation of departed vapor bubbles improves CHF with increase in bulk liquid subcooling. The effect of additive concentra-

tion on CHF was also investigated to increase CHF to a very high value of  950 kW=m2 at an optimum concentration of CMC/64 of [BMIM][OS]. We believe that the relative enhancements in CHF in comparison to water will further increase in the microgravity conditions where buoyancy is absent and does not oppose bubble departure unlike boiling on an inverted heater in earth gravity. Furthermore, minimal bubble coalescence with IL solutions may also be instrumental in delaying the dryout of the heater surface during boiling in narrow gaps. Conflict of interest The authors declare that there is no conflict of interest. Acknowledgment This work was supported by the Indian Space Research Organization through the RESPOND Programme (ISRO/ RES/3/674/2014-15). References [1] V.K. Dhir, Boiling heat transfer, Annu. Rev. Fluid Mech. 30 (1) (1998) 365–401, https://doi.org/10.1146/annurev.fluid.30.1.365. [2] A.E. Bergles, Enhancement of pool boiling, Int. J. Refrig. 20 (8) (1997) 545–551, https://doi.org/10.1016/S0140-7007(97)00063-7. [3] I. Mudawar, Assessment of high-heat-flux thermal management schemes, IEEE Trans. Components Packag. Technol. 24 (2) (2001) 122–141, https://doi.org/ 10.1109/6144.926375. [4] H. Merte, J.A. Clark, Boiling heat transfer with cryogenic fluids at standard, fractional and near-sero gravity, J. Heat Transfer 86 (1963) 351–358. [5] R. Raj, J. Kim, Heater size and gravity based pool boiling regime map: transition criteria between buoyancy and surface tension dominated boiling, J. Heat Transfer 132 (9) (2010) 91503, https://doi.org/10.1115/1.4001635. [6] T.J. Snyder, J.N. Chung, J.B. Schneider, Competing effects of dielectrophoresis and buoyancy on nucleate boiling and an analogy with variable gravity boiling results, J. Heat Transfer 120 (2) (1998) 371–379. [7] R. Raj, J. Kim, J. McQuillen, Pool boiling heat transfer on the international space station: experimental results and model verification, J. Heat Transfer 134 (10) (2012) 101504, https://doi.org/10.1115/1.4006846. [8] J. Kim, J.F. Benton, Highly subcooled pool boiling heat transfer at various gravity levels, Int. J. Heat Fluid Flow 23 (4) (2002) 497–508, https://doi.org/ 10.1016/S0142-727X(02)00139-X. [9] T. Oka, Y. Abe, Y.H. Mori, A. Nagashima, Pool boiling of n-pentane, CFC, and water under reduced gravity: parabolic flight experiments with a transparent heater, J. Heat Transfer 117 (2) (1995) 408–417, https://doi.org/10.1115/ 1.2822537. [10] R. Raj, J. Kim, J. McQuillen, On the scaling of pool boiling heat flux with gravity and heater size, J. Heat Transfer 134 (1) (2012) 11502, https://doi.org/10.1115/ 1.4004370. [11] R. Raj, J. Kim, J. McQuillen, Gravity scaling parameter for pool boiling heat transfer, J. Heat Transfer 132 (9) (2010) 91502, https://doi.org/10.1115/ 1.4001632. [12] R. Raj, J. Kim, J. McQuillen, Subcooled pool boiling in variable gravity environments, J. Heat Transfer 131 (9) (2009) 91502, https://doi.org/ 10.1115/1.3122782. [13] Z. Guo, M.S. El-Genk, An experimental study of saturated pool boiling from downward facing and inclined surfaces, Int. J. Heat Mass Transfer 35 (9) (1992) 2109–2117, https://doi.org/10.1016/0017-9310(92)90056-X. [14] G.H. Su, Y.W. Wu, K. Sugiyama, Subcooled pool boiling of water on a downward-facing stainless steel disk in a gap, Int. J. Multiphase Flow 34 (11) (2008) 1058–1066, https://doi.org/10.1016/j.ijmultiphaseflow.2008. 07.004. [15] M.Q. Raza, N. Kumar, R. Raj, Surfactants for bubble removal against buoyancy, Sci. Rep. 6 (2016) 19113, https://doi.org/10.1038/srep19113. [16] J.S. Sitter, T.J. Snyder, J.N. Chung, P.L. Marston, Terrestrial and microgravity pool boiling heat transfer from a wire in an acoustic field, Int. J. Heat Mass Transfer 41 (14) (1998) 2143–2155, https://doi.org/10.1016/S0017-9310(97) 00344-X. [17] P. Di Marco, R. Raj, J. Kim, Boiling in variable gravity under the action of an electric field: results of parabolic flight experiments, J. Phys. Conf. Ser. 327 (1) (2011) 12039, https://doi.org/10.1088/1742-6596/327/1/012039. [18] A. Faghri, Heat pipes: review, opportunities and challenges, Front. Heat Pipes 5 (1) (2014), https://doi.org/10.5098/fhp.5.1. [19] F.P. Chiaramonte, J.A. Joshi, Workshop on critical issues in microgravity fluids, transport, and reaction processes in advanced human support technology, NASA/TM—2004-212940 (2004). [20] D. Coleman, N. Gathergood, Biodegradation studies of ionic liquids, Chem. Soc. Rev. 39 (2) (2010) 600–637, https://doi.org/10.1039/b817717c.

N. Kumar et al. / International Journal of Heat and Mass Transfer 122 (2018) 354–363 [21] S. Keskin, D. Kayrak-Talay, U. Akman, Ö. Hortaçsu, A review of ionic liquids towards supercritical fluid applications, J. Supercrit. Fluids 43 (1) (2007) 150– 180, https://doi.org/10.1016/j.supflu.2007.05.013. [22] M. Petkovic, K.R. Seddon, L.P.N. Rebelo, C. Silva, Pereira, Ionic liquids: a pathway to environmental acceptability, Chem. Soc. Rev. 40 (3) (2011) 1383– 1403, https://doi.org/10.1039/C004968A. [23] S.K. Swarnkar, S.S. Murthy, R.L. Gardas, G. Venkatarathnam, Performance of a vapour absorption refrigeration system operating with ionic liquid-ammonia combination with water as cosolvent, Appl. Therm. Eng. 72 (2) (2014) 250– 257, https://doi.org/10.1016/j.applthermaleng.2014.06.020. [24] X. Liu, L. Bai, S. Liu, M. He, Vapor–liquid equilibrium of R1234yf/[HMIM][Tf 2 N] and R1234ze(E)/[HMIM][Tf 2 N] working pairs for the absorption refrigeration cycle, J. Chem. Eng. Data 61 (11) (2016) 3952–3957, https://doi. org/10.1021/acs.jced.6b00731. [25] V.V. Wadekar, Ionic liquids as heat transfer fluids – An assessment using industrial exchanger geometries, Appl. Therm. Eng. 111 (2017) 1581–1587, https://doi.org/10.1016/j.applthermaleng.2016.04.156. [26] F. Zhou, Y. Liang, W. Liu, Ionic liquid lubricants: designed chemistry for engineering applications, Chem. Soc. Rev. 38 (9) (2009) 2590–2599, https:// doi.org/10.1039/b817899m. [27] Q. Zhang, J.M. Shreeve, Energetic ionic liquids as explosives and propellant fuels: a new journey of ionic liquid chemistry, Chem. Rev. 114 (20) (2014) 10527–10574, https://doi.org/10.1021/cr500364t. [28] Z. He, P. Alexandridis, Nanoparticles in ionic liquids: interactions and organization, Phys. Chem. Chem. Phys. 17 (28) (2015) 18238–18261, https:// doi.org/10.1039/c5cp01620g. [29] A. Podgoršek, A.S. Pensado, C.C. Santini, M.F. Costa Gomes, A.A.H. Pádua, M.F.C. Gomes, A.H. Pa, Interaction energies of ionic liquids with metallic nanoparticles: solvation and stabilization effects, J. Phys. Chem. C. 117 (7) (2013) 3537–3547, https://doi.org/10.1021/jp309064u. [30] H. Li, L. Chen, H. Wu, H. He, Y. Jin, Ionic liquid-functionalized fluorescent carbon nanodots and their applications in electrocatalysis, biosensing, and cell imaging, Langmuir 30 (49) (2014) 15016–15021, https://doi.org/10.1021/ la503729v. [31] M.Y. Teo, N. Kim, S. Kee, B.S. Kim, G. Kim, S. Hong, S. Jung, K. Lee, Highly stretchable and highly conductive PEDOT:PSS/Ionic liquid composite transparent electrodes for solution-processed stretchable electronics, ACS Appl. Mater. Interfaces 9 (1) (2017) 819–826, https://doi.org/10.1021/ acsami.6b11988. [32] R.Á. Asencio, E.D. Cranston, R. Atkin, M.W. Rutland, Ionic liquid nanotribology: stiction suppression and surface induced shear thinning, Langmuir 28 (26) (2012) 9967–9976, https://doi.org/10.1021/la3010807. [33] M.D. Joshi, J.L. Anderson, Recent advances of ionic liquids in separation science and mass spectrometry, RSC Adv. 2 (13) (2012) 5470–5484, https://doi.org/ 10.1039/c2ra20142a. [34] T. Vander Hoogerstraete, B. Onghena, K. Binnemans, Homogeneous liquidliquid extraction of metal ions with a functionalized ionic liquid, J. Phys. Chem. Lett. 4 (10) (2013) 1659–1663, https://doi.org/10.1021/jz4005366.

363

[35] A. Zeinolabedini, S. Dorostkar, S. Ayatollahi, M. Nabipour, B. Hemmateenejad, Investigating the effect of ionic liquid (1-dodecyl-3-methylimidazolium chloride ([C12 mim] [Cl])) on the water/oil interfacial tension as a novel surfactant, Colloids Surfaces A Physicochem. Eng. Asp. 421 (2013) 63–71, https://doi.org/10.1016/j.colsurfa.2012.12.008. [36] H.J. Cho, J.P. Mizerak, E.N. Wang, Turning bubbles on and off during boiling using charged surfactants, Nat. Commun. 6 (2015) 8599, https://doi.org/ 10.1038/ncomms9599. [37] V.M. Wasekar, R.M. Manglik, Pool boiling heat transfer in aqueous solutions of an anionic surfactant, J. Heat Transfer 122 (4) (2000) 708–715. [38] V. Bergeron, Disjoining pressures and film stability of alkyltrimethylammonium bromide foam films, Langmuir 13 (13) (1997) 3474–3482, https://doi.org/10.1021/la970004q. [39] L. Wang, R.H. Yoon, Stability of foams and froths in the presence of ionic and non-ionic surfactants, Miner. Eng. 19 (6) (2006) 539–547, https://doi.org/ 10.1016/j.mineng.2005.09.006. [40] N. Wu, J. Dai, F. Micale, Dynamic surface tension measurement with a dynamic wilhelmy plate technique, J. Colloid Interface Sci. 215 (2) (1999) 258–269, https://doi.org/10.1006/jcis.1999.6270. [41] I. Langmuir, The adsorption of gases on plane surfaces of glass, mica and platinum, J. Am. Chem. Soc. 40 (9) (1918) 1361–1403, https://doi.org/ 10.1021/ja02242a004. [42] A. Pal, A. Yadav, Modulations in the aggregation behavior of ionic liquid 1butyl-3-methylimidazolium octylsulfate in aqueous alcohol solutions, J. Mol. Liq. 212 (2015) 569–575, https://doi.org/10.1016/j.molliq.2015.10.009. [43] M.J. Rosen, J.T. Kunjappu, Surfactants and interfacial phenomena, fourth ed., John Wiley & Sons, New Jersy, 2012. [44] A. Heintz, J.K. Lehmann, E. Schmidt, A. Wandschneider, Diffusion coefficients of imidazolium based ionic liquids in aqueous solutions, J. Solution Chem. 38 (8) (2009) 1079–1083, https://doi.org/10.1007/s10953-009-9431-2. [45] N. Zuber, Hydrodynamic aspects of boiling heat transfer Thesis, Univ. of California, Los Angeles, CA (United States), 1959. [46] Y. Haramura, Y. Katto, A new hydrodynamic model of critical heat flux, applicable widely to both pool and forced convection boiling on submerged bodies in saturated liquids, Int. J. Heat Mass Transfer 26 (3) (1983) 389–399, https://doi.org/10.1016/0017-9310(83)90043-1. [47] C. Unal, V. Daw, R.A. Nelson, Unifying the controlling mechanisms for the critical heat flux and quenching: the ability of liquid to contact the hot surface, J. Heat Transfer 114 (1992) 972, https://doi.org/10.1115/1.2911909. [48] T.G. Theofanous, J.P. Tu, A.T. Dinh, T.N. Dinh, The boiling crisis phenomenon: Part I: nucleation and nucleate boiling heat transfer, Exp. Therm. Fluid Sci. 26 (6) (2002) 775–792, https://doi.org/10.1016/S0894-1777(02)00193-0. [49] Y.M. Yang, J.R. Maa, On the criteria of nucleate pool boiling enhancement by surfactant addition to water, Chem. Eng. Res. Des. 79 (4) (2001) 409–416, https://doi.org/10.1205/026387601750282337. [50] J. Zhang, R.M. Manglik, Additive adsorption and interfacial characteristics of nucleate pool boiling in aqueous surfactant solutions, J. Heat Transfer 127 (4) (2005) 684–691, https://doi.org/10.1115/1.1924626.