International Journal of Heat and Mass Transfer 119 (2018) 333–342

Contents lists available at ScienceDirect

International Journal of Heat and Mass Transfer

journal homepage: www.elsevier.com/locate/ijhmt

Wettability modification to further enhance the pool boiling

performance of the micro nano bi-porous copper surface structure
Ya-Qiao Wang, Jia-Li Luo, Yi Heng, Dong-Chuan Mo ⇑, Shu-Shen Lyu
School of Chemical Engineering and Technology, Sun Yat-sen University, Guangzhou, Guangdong, China
Guangdong Engineering Technology Research Centre for Advanced Thermal Control Material and System Integration (ATCMSI), Guangzhou 510275, China

a r t i c l e i n f o a b s t r a c t

Article history: Boiling heat transfer is widely used in industry and in daily life and it can be enhanced by micro/nano
Received 18 July 2017 surface modification. Herein, we study the pool boiling characteristics of the micro nano bi-porous cop-
Received in revised form 30 September per surface with optimal cavity size from other researchers and present an efficient way to further
2017
enhance its boiling heat transfer performance by wettability modification of enlarging the particle size
Accepted 15 November 2017
Available online 28 November 2017
to lower the surface energy. In this work, two micro nano bi-porous copper surface samples were pre-
pared and compared with conventional surfaces. One is the original micro nano bi-porous copper surface
(Sample#O) prepared using the hydrogen bubble template deposition method to form abundant pores in
Keywords:
Pool boiling
optimal cavity size, and the other one is the modified micro nano bi-porous copper surface (Sample#M)
Micro/nano structure that is modified by applying a low current density on Sample#O for a few minutes. Scanning Electron
Bi-porous Microscope (SEM) images show that Sample#M keeps the pore size but enlarges the nano dendrite on
Surface energy the top of pore wall to micro balls. The conducted pool boiling experiments indicate that both micro nano
Bubble departure bi-porous copper surfaces have superior heat transfer coefficients than the plain copper surface, and the
Wettability high-speed camera shows that the micro nano bi-porous copper surfaces have shorter bubble growth
periods than those surfaces with pure nano structure or pure micro structure. At a heat flux of 90 W
cm
2, the heat transfer coefficient of Sample#O is 13 W cm
2 K
1, which is 2.8 times over that of the
plain surface. Compared to Sample#O, Sample#M can further enhance the pool boiling heat transfer.
At the same heat flux of about 90 W cm
2, the heat transfer coefficient of Sample#M is 23 W cm
2
K
1, which is 1.7 times over that of Sample#O and 4.8 times over that of the plain surface. The heat trans-
fer coefficient of Sample#M can be as high as 30 W cm
2 K
1 when it reaches the CHF. High-speed cam-
era images show that highest bubble growth period for Sample#M is just less than 20 ms, which is
shorter than that of Sample#O having a value in between 20 ms and 40 ms. It confirms that
Sample#M after wettability modification has a lower interface energy which can accelerate the bubbles
departure, and has an even more superior heat transfer performance than Sample#O.
Ó 2017 Elsevier Ltd. All rights reserved.

1. Introduction done by machining [5], sintered metal powders [6], covered mesh
[7], etc. Wei and Honda [5] used different silicon chips with micro-
Boiling heat transfer is widely used in industry and in daily life, pin-fins to enhance the pool boiling of FC-72. Li and Peterson [7]
such as heat exchangers, air conditioners, heat pumps, refrigera- used sintered isotropic copper mesh to enhance the pool boiling.
tion machinery and so on. Generally speaking, there are two com- Xu and Zhao [8] enhanced boiling heat transfer by gradient porous
mon ways to enhance pool boiling. One way is to modify the metals in saturated pure water.
working fluid, such as to add nano particles [1,2] or to add surfac- With the progress in nano technologies, surface modifications
tant [2–4]. The other way is to modify microgeometry or/and wet- with nano structure have been used to enhance boiling heat
tability of the boiling surfaces. transfer. Li et al. [9] found ‘‘unexpected enhancements” in boiling
The micro geometry of boiling surface strongly influences the performance by using copper nanorod structure on copper sub-
boiling heat transfer. Traditionally, surface modification can be strate. Yao et al. [10] prepared 35 lm-tall SiNW nanowire that
yielded a heat flux of 134 W cm
2 at 23 K wall superheat, about
⇑ Corresponding author at: School of Chemical Engineering and Technology, Sun 300% higher than that of a plain Si surface at the same wall super-
Yat-sen University, Guangzhou, Guangdong, China. heat. Chen et al. [11] fabricated a surface with TiO2 nanotube
E-mail address: [email protected] (D.-C. Mo).

https://doi.org/10.1016/j.ijheatmasstransfer.2017.11.080
0017-9310/Ó 2017 Elsevier Ltd. All rights reserved.
334 Y.-Q. Wang et al. / International Journal of Heat and Mass Transfer 119 (2018) 333–342

Nomenclature

A the projected area of porous structure (cm2) Greek symbol

h heat transfer coefficient (W cm
2 K
1) r liquid surface tension (N m
1)
Fs starting frame number n constant number
Fe ending frame number q density of pure copper (g/cm
3)
K permeability (m2) h contact angle (°)
kCu thermal conductivity of pure copper (W m
1 K
1) U porosity of porous structure
m mass of porous structure (mg) l viscosity of the liquid (Pa s)
q heat flux (W cm
2) d thickness of porous structure (lm)
Qsu total heat supplied by heaters (W) cs surface energy of solid surface (J m
2)
Qloss heat loss (W)
Reff effective pore radius (m) Subscripts
DT superheat (K) Cu copper
t time (s) su supplied heat
T bubble growth period (s) loss heat loss
T1 temperature of thermocouple (T1) (°C) eff effective
T2 temperature of thermocouple (T2) (°C) w wall
T3 temperature of thermocouple (T3) (°C)
Ts temperature of surrounding (°C) Abbreviation
Tw temperature of the bottom of porous coating (°C) CHF critical heat flux (W cm
2)
x distance between the thermocouple (T1) and the boiling HTC Heat transfer coefficient (W cm
2 K
1)
surface (m) FR frame rate (fps)
y liquid rising height (m)

arrays, and found it could enhance the pool boiling heat transfer heterogeneous nucleation in pool boiling and found that the effect
compared to the pure Ti metal plate. of microstructures greatly enhanced the nucleation of bubbles
Researchers recently found that surfaces with both micro and when the curvature radius of these microstructures is in the range
nano porous structure have superior boiling heat transfer perfor- of 5–100 times less than the bubble radius. Dong et al. [23] pre-
mance than those with only pure nano structure or pure micro sented that microstructures can enhance bubble nucleation to
structure. Nagato [12] found that the heat flux at the same super- reduce the wall superheat and enhance heat flux, while nano-
heat increased in the order flat surface < nanostructure < structures can accelerate bubble departure. Li et al. [9] calculated
microstructure < nano/micro composite structure. Li et al. [13] the range of active cavity size as a function of wall superheat,
reported 3-D porous interconnected foam-like structures using and indicated that the optimal range for cavity size, in which the
hydrogen bubble dynamic template method, and the HTC is about wall superheat can be even as low as 2 °C, is in the range between
17 times of plain copper at 1 W/cm2 when R134a is used as work- 10 lm and 100 lm.
ing fluid. Similar works were reported by different researchers. El- The surface wettability is also important for bubble dynamics in
Genk et al. [14] deposited micro-porous copper structure in copper nucleation boiling. Fritz et al. [24] and Wen et al. [3] found that the
sulfate solution which showed an excellent boiling heat transfer surfaces with smaller contact angle has smaller bubble diameter
coefficient of PF-5060. Patil et al. [15] developed a two-step and can enhance nucleation pool boiling. On the contrary, Phan
electro-deposition process to prepare cauliflower-like micro struc- et al. [25] observed that the hydrophilic surface with smaller con-
ture, the superheat of pool boiling of saturated distilled water is tact angle has larger departure bubble diameter. Recently,
only about 7 °C when CHF is about 1400 kW/m2. Xu et al. [16] pre- researchers found that wettability patterned surfaces (or the biphi-
pared composite porous surface and the highest CHF of the porous lic surfaces) have much better boiling heat transfer performance
surface was 239 W/cm2, which is 101% higher than that of plain than merely hydrophilic or hydrophobic surface [26–29]. Betz
surface. Wang et al. [17] fabricated micro nano-porous copper sur- et al. [26] observed that hydrophilic networks featuring hydropho-
faces, and found that the sample with multi-layer has lower wall bic islands has better heat transfer performance than the hydrophi-
superheat than that of the one with mono-layer, own to its better lic surface, due to the hydrophobic islands can enhance the
liquid accommodation from the morphology structure at high heat nucleation and the hydrophilic networks efficiently prevent the
flux. Ji et al. [18] fabricated several 3D porous coating surfaces with formation of an insulating vapor layer. Other researchers also
significant heat transfer enhancement, and the sample (TS#7) sig- found that the wettability patterned surface can enhance nucle-
nificantly increases the critical heat flux, which can be 3.7 times ation density, accelerate bubble growth rate and bubble departure
that of the plain surface. Kandlikar et al. [19] deposited coatings rate, such as Dai et al. [27] functionalized the hydrophobic nan-
at the top of micro channels to gain a heat transfer coefficient as otube with hydrophilic functional group, Jo et al. [28] spatially pat-
high as 2900 kW m
2 K
1. Li et al. [20] manufactured modulated terned self-assembled monolayers (contact angle 81°) on the head
porous structure and the highest heat transfer coefficient reached of Si micro posts (contact angle of plain Si is 57°), Zupancic et al.
450 W/cm2, over 3 times that of a plain copper surface. Wang [29] manufactured hydrophobic polydimethylsiloxane-silica coat-
et al. [21] fabricated copper vertical micro dendrite fin arrays ing on laser induced superhydrophilic steel surface, et al.
which can increase the heat transfer coefficient about 260% com- In this work, we try to fabricated micro/nano bi-porous surface
pared to that of the plain copper surface. combining both optimal cavity size and wettability modification.
In order to optimize the structure of micro nano-porous mate- Firstly, we fabricated a micro nano bi-porous copper surface with
rial to get a better heat transfer performance, it is worth revealing abundant micro pores whose diameters are in the optimal range
the effects of different structural scale on the boiling heat transfer. for cavity size according to previous study from other researchers
Dong et al. [22] analyzed the surface-microstructural effects on [9]. Secondly, a prior development by wettability modification is
 Y.-Q. Wang et al. / International Journal of Heat and Mass Transfer 119 (2018) 333–342 335

2Hþ þ 2e
¼ H2 " ð3Þ

2
With a current density of 1 A cm for 1 min, the deposition
process for Sample#O is finished. The Sample was washed with
deionized water and then dried using a hairdryer. Finally, the sam-
ple was sintered in reducing atmosphere at 710 °C for 30 min to
strengthen the structure [16]. Fig. 3 shows the SEM images of the
two micro nano bi-porous copper surfaces. There are lots of micro
pores with a diameter of approximately 100 lm on the surface of
Sample#O, and the wall of these pores is made up of nano den-
drites, as shown in Fig. 3(a) and (b).
Fig. 1. Setup for the sample preparation.
Sample#M was prepared by applying a lower current density of
0.1 A cm
2 to the as deposited Sample#O for 4 min. In such a small
made to further enhance the boiling heat transfer of the micro current, the hydrogen evolution reaction (Eq. (3)) disappeared on
nano bi-porous copper surface. The surface energy decreases with the cathode and only the copper deposition reaction (Eq. (2)) can
increasing particle size [30] and the bubble departure is much still occurred on the cathode. Thus, no more micro pores can be
easier on a surface with lower energy [26–29]. Hence, enlarging form, and the copper will deposit on the dendrites. After this pro-
the size of nano dendrites on the pore wall to micro balls is an effi- cess, the Sample#M was washed, dried and sintered in reducing
cient way to form wettability pattern. Pool boiling heat transfer atmosphere following the same procedure as Sample#O. Fig. 3(c)
performance of the samples fabricated in these two steps were and (d) clearly shows that, compared to Sample#O, the large pore
tested with water at atmosphere pressure respectively. size of Sample#M is still kept almost the same at about 100 lm,
and the structure in the bottom of the pore wall is still kept as den-
2. Sample preparation and pool boiling experimental set-up drites, but the dendrites at the top of the pore wall are enlarged to
micro balls which will lower the surface energy [30].
2.1. Sample preparation The porosity of the sample, U, was calculated as following:
m
/¼1
ð4Þ
The original micro nano bi-porous copper surface (Sample#O) qdA
was made as cylinders with a diameter of 19 mm and thickness
where m is mass of porous structure, q is density of pure copper, d is
of 8 mm. The sample was cleaned with dilute sulfuric acid, hot
the thickness of porous structure, and A is the projected area of por-
dilute caustic solution, and deionized water, and were then used
ous structure which is also the boiling surface area when applied to
as the cathode as shown in Fig. 1. A circle copper plate with diam-
pool boiling test.
eter of 6.5 cm was used as anode which was 3.5 cm above the cath-
According to Table 1, compared to Sample#O, the thickness of
ode. The electrodeposition process was performed in stationary
Sample#M was increased and the porosity was kept the same. As
solution in which the CuSO4 was 0.4 M and H2SO4 1.8 M. A DC
Sample#M was prepared by further depositing copper on Sam-
power supply (Maynuo 8852) was used for the deposition process.
ple#O, the thickness of Sample#M is higher than that of the Sam-
In this process, the copper in the anode was solved (Eq. (1)) and
ple#O. Li [32] found that the capillary wick thickness can strongly
copper ion was deposited into solid on the cathode (Eq. (2)). If
affect the CHF but has no significant effect on the HTC. The previ-
the current density is high enough, the copper deposition reaction
ous studies on similar structure of micro nano bi-porous copper
(Eq. (2)) and hydrogen evolution reaction (Eq. (3)) were occurred
surface [13,16] also show that the HTC is not sensitivity on the
simultaneously on the cathode, this is the so-called hydrogen bub-
thickness.
ble template electrodeposition method [17,31]. Fig. 2(a) shows the
Wickability of micro/nano porous structure has strong effects
sketch of using the dynamic bubble template electrodeposition
on boiling heat transfer performance [33–35]. K/Reff (permeability
method to prepare Sample#O, and Fig. 2(b) shows a photo taken
to effective pore radius) shown in Table 1 is an important param-
in this process. Thanks to abundant hydrogen bubble departure
eter to characterize liquid propagation in porous structure and it
from the cathode, porous copper surface can be easily formed.
can be measured by the capillary rate of rise experiment herein
In the anode:
absolute ethyl alcohol used as working fluid. According to the
Cu ¼ Cu2þ þ 2e
ð1Þ Lucas-Washburn equation [36] (Eq. (5)), the square of liquid rising
height (y2) has a linear correlation with rising time (t):
In the cathode:
4r cos h K
Cu2þ þ 2e
¼ Cu ð2Þ y2 ¼ t ð5Þ
/l Reff

Fig. 2. Using the dynamic bubble template electrodeposition method to prepare Micro nano bi-porous copper surface (Sample#O) (a) sketch; (b) photo taken in the sample
preparation.
336 Y.-Q. Wang et al. / International Journal of Heat and Mass Transfer 119 (2018) 333–342

Fig. 3. SEM of the micro nano bi-porous copper surfaces (a) Sample#O; (b) Enlarge the yellow rectangle part in (a); (c) Sample#M; (d) Enlarge the yellow rectangle part in (c).

Table 1 vapor. A cyclic annular auxiliary heater was used to make sure the
Sample properties. working fluid was in a saturated state. Four T-type thermocouples
Mass/mg Thickness/lm Porosity/% K/Reff/m  107 with an accuracy of ±0.3 °C were connected to the data acquisition
Sample#O 39 140 89 2.47 (Agilent 34970A). The first thermocouple (T1) was 6.5 mm below
Sample#M 50 170 89 3.37 the micro nano bi-porous copper surface. The second thermocou-
ple (T2) was located in the boiling vessel to measure pool water
temperature. The third thermocouple (T3) was on the surface of
where r is the liquid surface tension, h is the contact angle, U is the the heating copper block. The fourth thermocouple (Ts) was used
porosity of the porous structure, and l is the viscosity of the liquid. to test the surrounding temperature. Besides, a high-speed camera
The liquid rising height was tracked by a high-speed camera (Vision Research Phantom V211) with a cold light source system
(Vision Research Phantom V211), and Fig. 4 shows the liquid rising was installed to record the bubble dynamic images.
height image for Sample#M. Fig. 5 shows the linear fit of the A power meter was used to record the total heat supplied (Qsu)
square of the liquid rising height and the time (solid lines). The by ceramic rod heaters. Before the boiling test, the relationship of
square of liquid rising height evolved with rising time for different heat loss (Qloss) and temperature difference between the heat block
samples. It demonstrates that the liquid rising rate is increasing surface and surrounding (T3
Ts) was detected. Constant low
after modification. The slope of Sample#M is larger than Sam- power was supplied when no water was in pool and an aerogel
ple#O, and the K/Reff for Sample#M should be larger than Sam- blanket was placed on the boiling surface. When the temperature
ple#O. Combined with Eq. (5), we can finally get the K/Reff in was steady, heat loss at a defined temperature difference
Table 1. The K/Reff of Sample#M is about 36% higher than that of (T3
Ts) can be recorded and the results are shown in Fig. 7(a).
Sample#O. The heat loss is linear to the temperature difference of the heat
block surface and surroundings and can be calculated as:
2.2. Pool boiling experimental set-up Q loss ¼ 0:0538  ðT 3
T s Þ ð6Þ

Fig. 6 shows the set-up of the pool boiling experiment. Copper During the boiling test, the heat flux conducted to the boiling
block with 6 ceramic rod heaters inside was implemented as the surface (q) can be calculated as:
heat source. The test sample was welded to the copper block to
Q su
Q loss
decrease the thermal resistance between the sample and the cop- q¼ ð7Þ
A
per block. An aerogel blanket was applied to decrease the heat leak
to the environment. Ultra-pure water, boiled for 30 min, was used where A is the boiling surface area. According to the Fourier 1-D
as working fluid. A copper coiled cooler was used to cool down the conduction equation, q can be also calculated as:
 Y.-Q. Wang et al. / International Journal of Heat and Mass Transfer 119 (2018) 333–342 337

0.02 sec 0.32 sec 3.02 sec 6.27 sec

Fig. 4. Liquid rising from Sample#M by high-speed camera.

ðQ su
Q loss Þ  x
Tw ¼ T1
ð9Þ
A  kCu
Finally, the wall superheat, DT can be calculated:
DT ¼ T w
T 2 ð10Þ
The heat transfer coefficient (HTC) can be calculated by the fol-
lowing equation:
q
h¼ ð11Þ
DT
It was confirmed that the heat loss was no more than 3% of Qsu
in all the pool boiling tests. The reliability of the present
experimental set-up was also tested by a sample with plain copper
surface, and the boiling curve is shown in Fig. 7(b). The curve can
Fig. 5. The linear fit of the square of the liquid rising height of Sample#M and the
agree with other researchers’ experiments (Kim [37], Xu [8],
time.
Akbary [38] and Partil [15]), especially the Rohsenow’s model
[39]. The error of whose pool boiling curve was within ±5%
Tw
T1 compared to Rohsenow’s model [39].
q ¼
kCu ð8Þ The uncertainty of q, DT, and h can be calculated from
x
Eqs. (12)–(14).
where kCu is the thermal conductivity of copper, x is the distance
ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
s 2  2  2  2
between the thermocouple T1 and the boiling surface, and Tw is @q @q @q @q 2
the wall temperature of the boiling surface. Combining Eqs. (7) Dq ¼ ðr Q s Þ2 þ ðDT 3 Þ2 þ ðD T s Þ2 þ Dd
@Q s @T 3 @T s @d
and (8), Tw can be derived as: ð12Þ

Fig. 6. Pool boiling experimental set-up (1) copper block (6 ceramic rod heaters) (2) aerogel blanket (3) FR-4 epoxy glass fiber board (4) ultra-pure water (5) cyclic annular
auxiliary heater (6) cold light source system (7) high-speed camera (8) thermocouple (9) tested sample (10) copper coiled cooler (11) data acquisition.
338 Y.-Q. Wang et al. / International Journal of Heat and Mass Transfer 119 (2018) 333–342

Fig. 7. The reliability of the present experimental set-up (a) Heat loss relationship between temperature difference of heat block surface and surroundings; (b) Pool boiling
curve on the plain copper surface.

sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
 2  2  2  2 decreasing region (down) and the second heat flux increasing
@ DT @ DT @ DT @ DT
DðDTÞ ¼ ðrT 1 Þ2 þ ðDT 2 Þ2 þ ðDqÞ2 þ ðDxÞ2 region (up2).
@T 1 @T 2 @q @x
ð13Þ For Sample#O in the up1 region, the wall superheated temper-
ature increases with increasing heat flux at the very beginning.
sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi When the heat flux reaches about 14 W cm
2 and the wall super-
 2  2 heated temperature reaches about 10 °C, the wall superheated
@h @h
Dh ¼ ðDqÞ2 þ ðDðDTÞÞ2 ð14Þ temperature firstly suddenly drops and then increases slightly
@q @ DT
with increasing heat flux. This phenomenon demonstrates clearly
nucleation hysteresis [40] occurred on Sample#O. When the heat
where Dq is the uncertainty of heat flux, D(DT) is the uncertainty of
flux is low (here, <14 W cm
2), many cavities on the surface are
wall superheat, Dh is the uncertainty of heat transfer coefficient.
filled with liquid. If the heat flux is sufficiently high, the cavities
The uncertainty of direct parameters were listed in Table 2. The
maximum uncertainty of q, DT, h are less than 0.6 W cm
2, 0.5 °C
and 1.6 W cm
2 K
1, respectively.
Boling hysteresis phenomenon [40] is commonly occurred in
porous material. When this phenomenon happens, the wall super-
heated temperatures are quite different even if the same heat flux
is added to the surface. In order to get more information during the
boiling process, the following procedure was taken to get the boil-
ing curve:

(1) The first heat flux increasing process (up1): the heat flux was
increased from a small value to near CHF step by step.
(2) The heat flux decreasing process (down): the heat flux was
decreased from the last value in the up1 process to about
0.1 W cm
2 step by step.
(3) The second heat flux increasing process (up2): the heat flux
was increased from the last value in the down process to
CHF step by step.

3. Result and discussion

3.1. Boiling curve

Fig. 8(a) shows the pool boiling curve of different samples. For
the plain copper surface, there is a good overlap with Rohsenow’s
model [39] which ensures the accuracy of our pool boiling results.
For both Sample#O and Sample#M there exist three heat flux
regions: the first heat flux increasing region (up1), heat flux

Table 2
Uncertainty of direct parameters.

Parameters Uncertainty
Q su 0.4%
DT 1 ; DT 2 ; DT 3 0.3 °C
Dd 0.01 mm Fig. 8. Pool boiling curve of two micro-nano bi-porous samples and plain coppers at
Dx 0.05 mm atmospheric pressure (a) heat flux over wall superheated temperature (b) heat
transfer coefficient over heat flux in the second heat flux increasing.
 Y.-Q. Wang et al. / International Journal of Heat and Mass Transfer 119 (2018) 333–342 339

Fig. 9. Bubble dynamic images during the second heat flux increasing region.

are active. Because of the appearance of a large amount of cavities, as low as 1.7 °C at CHF of 420 W/cm2. On the other hand, wettability
the wall superheated temperature first suddenly drops and then patterned surface also can affect the wall superheat temperature.
increases slightly with increasing heat flux [17]. Hysteresis is also Betz [41] prepared superhydrophobic islands on the superhy-
found in the up1 region for Sample#M. When the heat flux reaches drophilic Si surface to achieve a wall superheat temperature of
about 14 W cm
2 and the wall superheated temperature reaches about 8 °C at CHF of 140 W/cm
2. Similar, compared to Sample#O,
about 7 °C, the wall superheat first suddenly drops and then Sample#M has a lower surface energy which can further lower the
increases slightly with increasing heat flux. It indicates that Sam- superheat. This will be further discussed in the section of boiling
ple#M needs a lower wall superheat to make a large number of enhancement mechanism.
cavities active. Fig. 8(b) shows the heat transfer coefficient in the up2 region for
When the curve of Sample#O enters the down region, the wall all surfaces. It can be found that the heat transfer coefficient
superheat decreases with decreasing heat flux, and the wall super- increases with heat flux. Both of the micro nano samples have a
heat temperature is much lower than that at the same heat flux in superior heat transfer coefficient than the plain copper surface at
the up1 region. After the heat flux reaches a low value, it increases the same heat flux. At a heat flux of about 90 W cm
2, the heat
again and the curve enters the up2 region. It can be found that, the transfer coefficient of Sample#O and Sample#M are 13 W cm
2
boiling curve in the up2 region repeats the down region quite well. K
1 and 23 W cm
2 K
1, respectively, which is about 2.8 times
Once the cavities are active, they can keep active during the heat and 4.8 times over that of the plain surface. It can also be seen
flux change. So, the wall superheat temperatures in the down clearly that Sample#M can further enhance the pool boiling heat
region and the up2 region are much lower than those at the same transfer and its heat transfer coefficient is about 1.7 times over that
heat flux in the up1 region. A similar feature is found for Sam- of Sample #O. Moreover, it has a maximum heat transfer coeffi-
ple#M. The CHF for Sample#O and Sample#M are both about cient of 30 W cm
2 K
1 when it reaches the CHF.
130 W cm
2, which are about 30% higher than the plain copper Usually, wickability of micro/nano porous structure can
surface. Compared to the wall superheat temperature (20 °C) enhance the CHF significant while has little effect on the HTC
for plain copper surface at CHF, Sample#O was much lower around [33–35]. Unfortunately, the CHF of Sample#O (130 W cm
2) is
9 °C, while Sample#M had an even lower value of 5 °C. The wall almost the same as that of Smaple#O (128 W cm
2), though Sam-
superheat temperature relies on the bubble nucleation which ple#M has a larger wickability value of K/Reff than Sample#O in
relate to cavities [37] and surface wettability [41]. As shown in Table 1. The wettability is also very important for the CHF. Xu
Fig. 2(b), there are abundant micro hydrogen bubble in the prepa- [16] and Ahn [34] both observed that the CHF will decreases as
ration of Sample#O and Sample#M, which can supply abundant the contact angle increase. The Sample#M has a lower surface
cavities for bubble nucleation. Besides the cavities density, the cav- energy and a larger contract angle which is negative for the CHF.
ities size is also very important. It is indicated that the optimal Combine the positive effect in wickability and the negative effect
range for cavity size, in which the wall superheat can be even as in wettability, the CHF of Sample#M is almost the same as the
low as 2 °C, is in the range between 10 lm and 100 lm [9]. Fig. 3 Sample#O.
shows that the diameters of the micro pore of both Sample#O
and Sample#M are just in this range. Some researchers also got 3.2. Bubble dynamics
very low wall superheat temperatures in CHF. Kim et al. [37]
achieved a wall superheat temperature of 6.8 °C at CHF due to Fig. 9 shows the bubble dynamic images in the up2 region. The
use the larger diameter particles to form cavities. Jaikumar and high-speed camera images show that there are three nucleate
Kandlikar [19] manufactured sintered copper powder on the wall boiling statuses for both Sample#O and Sample#M, which is
of the micro channel (300 lm in width) which has a wall superheat similar to the composite porous structure [16]. When the heat flux
340 Y.-Q. Wang et al. / International Journal of Heat and Mass Transfer 119 (2018) 333–342

Fig. 10. Bubble dynamics of Sample#M at a heat flux of 4 W cm
2 (a) in the first heat flux increasing (up1) region; (b) in the heat flux decreasing (down) region.

Fig. 11. Statistics of isolated bubble growth period at about 0.1 W/cm2 (a) a typical bubble growth period on Sample#M; (b) Sample#O; (c) Sample#M.

is small, it is in the status of isolated bubbles nucleate boiling. As ðF s
F e Þ

T¼ ð15Þ
the heat flux increases higher than 4 W cm
2, the bubble dynamic FR
status transforms into fully developed nucleate boiling. If the heat
flux is higher than 50 W cm
2, the status changes to bubble coales-
cence nucleate boiling. where FR is camera frame rate and equals to 10,000 frames per
The nucleation hysteresis can be visualized at the isolated bub- second.
ble nucleation boiling region. Fig. 10 shows the bubble dynamics of Fig. 11(b) and (c) shows the statistics of an isolated bubble
Sample#M at a heat flux of 4 W cm
2. Evidently, there are few growth period at about 0.1 W cm
2 in the second heat flux increas-
active cavities on the surface at 4 W cm
2 in the up1 region, as ing region. In Fig. 11(b), the bubble growth periods of Sample#O
shown in Fig. 10(a). Once the cavities are active, they can keep range from less than 10 ms to about 100 ms, while most of them
active during the heat flux change. So, there are a large amount (about 40%) are in the range of 20–40 ms. Dong et al. [23] presents
of active cavities on the surface at the same heat flux of 4 W that the bubble growth period on the surface with micro-cavities is
cm
2 in the down region, as shown in Fig. 10(b). about 70 ms, while the bubble growth period on the surface with
The bubble growth period was measured by the high-speed nano-cavities is about 50 ms. It indicates that the surfaces with
camera with a frame rate of 10000 frames per second. Fig. 11(a) both micro and nano porous structure have a shorter bubble
is a typical bubble growth period on Sample#M. Those bubbles growth period than those surfaces with pure nano structure or
without shielding from others were chosen to measure the bubble pure micro structure. Our research can agree with results from
growth period. The videos of 22 bubbles in different location for other researchers [12] that the surface with both micro and nano
each sample were analyzed. The videos were replay frame by porous structures have superior boiling heat transfer performance.
frame to find the nucleation site carefully. When the bubble just In Fig. 11(c), the bubble growth periods on Sample#M are all
came out of the cavity, the frame number was recorded as the less than 60 ms, and most of them are less than 20 ms. Obviously,
starting frame number (Fs). When the bubble departed from the the periods on Sample#M are much shorter than Sample#O. In
surface, the frame number was recorded as the ending frame num- other words, the bubbles escape more easily so that a better heat
ber (Fe). The bubble growth period (T) can be calculated by Eq. (15). transfer performance would be gained after further modification
 Y.-Q. Wang et al. / International Journal of Heat and Mass Transfer 119 (2018) 333–342 341

Fig. 12. Boiling enhancement mechanism and photos for a micro nano bi-porous copper surface (a) (b) mechanism for Sample#O and Sample#M respectively; (c) (d) Contact
angle for Sample#O and Sample#M respectively.

on the micro nano bi-porous copper surfaces. The boiling enhance- of lower surface energy sites, the surface can accelerate the bubble
ment mechanism will be discussed as follows. growth rate and the bubble departure rate [27–29,41]. Here, the
Sample#M has a lower surface energy than Sample#O, and it also
3.3. Boiling enhancement mechanism can make the bubbles easier to depart, lead to an even superior
heat transfer performance than Sample#O.
A proposed enhancement mechanism for micro nano bi-porous
copper surface is presented in Fig. 12. Both Sample#O and Sam-
4. Conclusion
ple#M have abundant pores with diameter of about 100 lm, which
is in the optimal range for cavity size [9]. These pores are suitable
In this work, we suggested to combine both optimal cavity size
for bubble initiation, growth and departure, and the micro nano bi-
and wettability modification to enhance pool boiling performance
porous surfaces have shorter bubble growth periods than those
of the micro nano bi-porous copper surface. To this end, different
surfaces with pure nano structure or pure micro structure. What’s
samples were prepared and compared. Sample#O has abundant
more, the wall of these pores on the micro nano bi-porous surfaces
pores with diameter in the optimal range for cavity size. Sample#M
are made up of nano dendrites, which enhances the liquid supply
keeps large pores sizes almost the same as Sample#O, but the den-
for boiling. Besides, both Sample#O and Sample#M has natural
drites on the wall of the large pores are enlarged to micro balls. This
separated liquid-vapor path that the vapor escapes through the
benefit the Sample#M a lower surface energy. After analyzing the
cavities and the liquid supplies through the dendrites walls which
pool boiling experiments, the following conclusions can be drawn:
will extremely enhance the heat transfer performance [13,19]. So,
the micro and nano composite structures lead to a superior heat
1. The order of heat transfer coefficient at the same heat flux is:
transfer performance to those surfaces with pure nano structure
Sample#M > Sample#O > the plain copper surface. When the
or pure micro structure.
heat flux is 90 W/cm2, Sample#M is about 1.7 times of the heat
The reason why Sample#M has superior heat transfer perfor-
transfer coefficient to Sample#O and 4.8 times of that on the
mance can be wettability modification. The main difference
plain copper surface. When it reaches the CHF, the heat transfer
between Sample#O and Sample#M is that the dendrites on the
coefficient of Sample#M can be as high as 30 W cm
2 K
1.
original pore wall are enlarged to micro balls for Sample#M. Previ-
2. The order of bubble growth period is: Sample#M < Sample#O <
ous study [30] shows that the surface energy decreases when the
surface with micro-cavities < surface with nano-cavities [23].
nano dendrites are enlarged into micro balls. For the liquid wetting
Most bubble growth periods on Sample#M are just less than
process on a solid surface, the relationship between surface energy
20 ms, which is shorter than that of Sample#O with quantities
cs and contact angle h can be described as Eq. (16) [42]:
between 20 ms and 40 ms.
c 1=2 3. Having large pores and nano structure on the pore wall, the
cos h ¼ 2n s
1 ð16Þ
r micro nano bi-porous surface has superior heat transfer perfor-
mance than the plain surface, surfaces with micro-cavities and
where, n is a constant (normally equates to 1), and r is the liquid
surfaces with nano-cavities.
surface tension.
4. Sample#M has superior boiling heat transfer performance than
As shown in Fig. 12(c) and (d), the contact angle of a typical
Sample#O. It confirms that enlarging the nano dendrites into
bubble on the Sample#M with growth period of 5.0 ms is 56°,
micro balls can lower the porous surface energy and is an effi-
while that on the Sample#O with growth period of 36.2 ms is
cient wettability modification to enhance pool boiling.
31°. In that case, the surface energy is 0.0358 J m
2 and 0.0508 J
m
2 for the Sample#M and Sample#O, respectively. For wettability
patterned surface which modified hydrophilic (or superhy- Conflicts of interest
drophilic) surface with hydrophobic (superhydrophobic) spots,
the hydrophobic sites have lower surface energy. With the effects None.
342 Y.-Q. Wang et al. / International Journal of Heat and Mass Transfer 119 (2018) 333–342

Acknowledgment [20] C.H. Li, T. Li, P. Hodgins, C.N. Hunter, A.A. Voevodin, J.G. Jones, G.P. Peterson,
Comparison study of liquid replenishing impacts on critical heat flux and heat
transfer coefficient of nucleate pool boiling on multiscale modulated porous
This work was supported by the National Science Foundation of structures, Int. J. Heat Mass Transf. 54 (15–16) (2011) 3146–3155.
China under Grant No. 51206193 and Pearl River S&T Nova Pro- [21] Y.-Q. Wang, S.-S. Lyu, J.-L. Luo, Z.-Y. Luo, Y.-X. Fu, Y. Heng, J.-H. Zhang, D.-C. Mo,
Copper vertical micro dendrite fin arrays and their superior boiling heat
gram of Guangzhou No. 201710010043. Yi Heng acknowledges
transfer capability, Appl. Surface Sci. 422 (2017) 388–393.
support from the Thousand Young Talents Program of China. We [22] L.N. Dong, X.J. Quan, P. Cheng, An analysis of surface-microstructures effects
thank Dr. Zhen-Cheng Huang for his constructive suggestion. on heterogeneous nucleation in pool boiling, Int. J. Heat Mass Transf. 55 (15–
16) (2012) 4376–4384.
[23] L.N. Dong, X.J. Quan, P. Cheng, An experimental investigation of enhanced pool
boiling heat transfer from surfaces with micro/nano-structures, Int. J. Heat
References Mass Transf. 71 (2014) 189–196.
[24] W. Fritz, Maximum volume of vapor bubbles, Phys. Z 36 (1935) 379–384.
[1] X.P. Fang, Y.M. Xuan, Q. Li, Experimental investigation on enhanced mass [25] H.T. Phan, N. Caney, P. Marty, S. Colasson, J. Gavillet, Surface wettability
transfer in nanofluids, Appl. Phys. Lett. 95 (20) (2009) 3. control by nanocoating: the effects on pool boiling heat transfer and
[2] Y.H. Jeong, W.J. Chang, S.H. Chang, Wettability of heated surfaces under pool nucleation mechanism, Int. J. Heat Mass Transf. 52 (23–24) (2009) 5459–5471.
boiling using surfactant solutions and nano-fluids, Int. J. Heat Mass Transf. 51 [26] A.R. Betz, J. Xu, H.H. Qiu, D. Attinger, Do surfaces with mixed hydrophilic and
(11–12) (2008) 3025–3031. hydrophobic areas enhance pool boiling?, Appl Phys. Lett. 97 (14) (2010) 3.
[3] D.S. Wen, B.X. Wang, Effects of surface wettability on nucleate pool boiling [27] X.M. Dai, X.Y. Huang, F.H. Yang, X.D. Li, J. Sightler, Y.C. Yang, C. Li, Enhanced
heat transfer for surfactant solutions, Int. J. Heat Mass Transf. 45 (8) (2002) nucleate boiling on horizontal hydrophobic-hydrophilic carbon nanotube
1739–1747. coatings, Appl. Phys. Lett. 102 (16) (2013) 5.
[4] V.M. Wasekar, R.M. Manglik, Pool boiling heat transfer in aqueous solutions of [28] H. Jo, D.I. Yu, H. Noh, H.S. Park, M.H. Kim, Boiling on spatially controlled
an anionic surfactant, J. Heat. Trans-T. ASME 122 (4) (2000) 708–715. heterogeneous surfaces: wettability patterns on microstructures, Appl. Phys.
[5] J.J. Wei, H. Honda, Effects of fin geometry on boiling heat transfer from silicon Lett. 106 (18) (2015) 5.
chips with micro-pin-fins immersed in FC-72, Int. J. Heat Mass Transf. 46 (21) [29] M. Zupancic, M. Steinbucher, P. Gregorcic, I. Golobic, Enhanced pool-boiling
(2003) 4059–4070. heat transfer on laser-made hydrophobic/superhydrophilic
[6] A.S. Kousalya, J.A. Weibel, S.V. Garimella, T.S. Fisher, Metal functionalization of polydimethylsiloxane-silica patterned surfaces, Appl. Therm. Eng. 91 (2015)
carbon nanotubes for enhanced sintered powder wicks, Int. J. Heat Mass 288–297.
Transf. 59 (2013) 372–383. [30] X.B. Chen, Y.C. Li, P.D. Hodgson, C. Wen, The importance of particle size in
[7] C. Li, G.P. Peterson, Parametric study of pool boiling on horizontal highly porous titanium and nonporous counterparts for surface energy and its impact
conductive microporous coated surfaces, J. Heat. Trans-T. ASME 129 (11) on apatite formation, Acta Biomater. 5 (6) (2009) 2290–2302.
(2007) 1465–1475. [31] H.-C. Shin, M. Liu, Copper foam structures with highly porous nanostructured
[8] Z.G. Xu, C.Y. Zhao, Experimental study on pool boiling heat transfer in gradient walls, Chem. Mater. 16 (25) (2004) 5460–5464.
metal foams, Int. J. Heat Mass Transf. 85 (2015) 824–829. [32] C. Li, G.P. Peterson, Y.X. Wang, Evaporation/boiling in thin capillary wicks (I) -
[9] C. Li, Z. Wang, P.-I. Wang, Y. Peles, N. Koratkar, G.P. Peterson, Nanostructured Wick thickness effects, J. Heat Transf.-Trans. ASME 128 (12) (2006) 1312–
copper interfaces for enhanced boiling, Small 4 (8) (2008) 1084–1088. 1319.
[10] Z. Yao, Y.W. Lu, S.G. Kandlikar, Effects of nanowire height on pool boiling [33] M.M. Rahman, E. Olceroglu, M. McCarthy, Role of wickability on the critical
performance of water on silicon chips, Int. J. Therm. Sci. 50 (11) (2011) 2084– heat flux of structured superhydrophilic surfaces, Langmuir 30 (37) (2014)
2090. 11225–11234.
[11] Y. Chen, D.C. Mo, H.B. Zhao, N. Ding, S.S. Lu, Pool boiling on the [34] H.S. Ahn, C. Lee, J. Kim, M.H. Kim, The effect of capillary wicking action of
superhydrophilic surface with TiO2 nanotube arrays, Sci. China Ser. E- micro/nano structures on pool boiling critical heat flux, Int. J. Heat Mass
Technol. Sci. 52 (6) (2009) 1596–1600. Transf. 55 (1–3) (2012) 89–92.
[12] K. Nagato, S. Miyazaki, S. Yamada, M. Nakao, Nano/microcomposite surface [35] C.Y. Lee, M.M.H. Bhuiya, K.J. Kim, Pool boiling heat transfer with nano-porous
fabricated by chemical treatment/microembossing for control of bubbles in surface, Int. J. Heat Mass Transf. 53 (19–20) (2010) 4274–4279.
boiling heat transfer, CIRP Ann-Manuf. Technol. 65 (1) (2016) 511–514. [36] N. Fries, K. Odic, M. Conrath, M. Dreyer, The effect of evaporation on the
[13] S.H. Li, R. Furberg, M.S. Toprak, B. Palm, M. Muhammed, Nature-inspired wicking of liquids into a metallic weave, J. Colloid Interface Sci. 321 (1) (2008)
boiling enhancement by novel nanostructured macroporous surfaces, Adv. 118–129.
Funct. Mater. 18 (15) (2008) 2215–2220. [37] J.H. Kim, A. Gurung, M. Amaya, S.M. Kwark, S.M. You, Microporous coatings to
[14] M.S. El-Genk, A.F. Ali, Enhanced nucleate boiling on copper micro-porous maximize pool boiling heat transfer of saturated R-123 and water, J. Heat.
surfaces, Int. J. Multiph. Flow 36 (10) (2010) 780–792. Trans-T. ASME 137 (8) (2015) 7.
[15] C.M. Patil, K.S.V. Santhanam, S.G. Kandlikar, Development of a two-step [38] E. Akbari, A.M. Gheitaghy, H. Saffari, S.M. Hosseinalipour, Effect of silver
electrodeposition process for enhancing pool boiling, Int. J. Heat Mass Transf. nanoparticle deposition in re-entrant inclined minichannel on bubble
79 (2014) 989–1001. dynamics for pool boiling enhancement, Exp. Therm. Fluid. Sci. 82 (2017)
[16] P.F. Xu, Q. Li, Y.M. Xuan, Enhanced boiling heat transfer on composite porous 390–401.
surface, Int. J. Heat Mass Transf. 80 (2015) 107–114. [39] W.M. Rohsenow, A method of correlating heat transfer data for surface boiling
[17] Y.Q. Wang, D.C. Mo, S.S. Lyu, Enhanced pool boiling heat transfer on mono and of liquids, Heat Transf. 74 (4) (1952) 969–973.
multi-layer micro-nano bi-porous copper surfaces, in: Proceedings of the [40] M.E. Poniewski, Peculiarities of boiling heat transfer on capillary-porous
Asme 5th International Conference on Micro/Nanoscale Heat and Mass coverings, Int. J. Therm. Sci. 43 (5) (2004) 431–442.
Transfer, Singapore, 2016. [41] A.R. Betz, J. Jenkins, C.-J.C.J. Kim, D. Attinger, Boiling heat transfer on
[18] X. Ji, J. Xu, Z. Zhao, W. Yang, Pool boiling heat transfer on uniform and non- superhydrophilic, superhydrophobic, and superbiphilic surfaces, Int. J. Heat
uniform porous coating surfaces, Exp. Therm. Fluid. Sci. 48 (2013) 198–212. Mass Transf. 57 (2) (2013) 733–741.
[19] A. Jaikumar, S.G. Kandlikar, Ultra-high pool boiling performance and effect of [42] N.T. Correia, J.J.M. Ramos, B.J.V. Saramago, J.C.G. Calado, Estimation of the
channel width with selectively coated open microchannels, Int. J. Heat Mass Surface tension of a solid: application to a liquid crystalline polymer, J. Colloid
Transf. 95 (2016) 795–805. Interface Sci. 189 (2) (1997) 361–369.