Thermal Bubble Nucleation in Graphene Nanochannels - The Journal

Subscriber access provided by EKU Libraries

C: Surfaces, Interfaces, Porous Materials, and Catalysis

Thermal Bubble Nucleation in Graphene Nanochannels Hongyang Yu, Zhongwu Li, Yi Tao, Jingjie Sha, and Yunfei Chen J. Phys. Chem. C, Just Accepted Manuscript • DOI: 10.1021/acs.jpcc.8b09038 • Publication Date (Web): 22 Jan 2019 Downloaded from http://pubs.acs.org on January 28, 2019

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 36 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

Thermal Bubble Nucleation in Graphene Nanochannels Hongyang Yu1, Zhongwu Li1, Yi Tao1, Jingjie Sha1*, Yunfei Chen1,2, 1. School of Mechanical Engineering , Jiangsu Key Laboratory for Design and Manufacture of Micro-nano Biomedical Instruments , Southeast University , Nanjing 211189, China 2. State Key Laboratory of Bioeletronics, Southeast University , Nanjing 211189, China

Abstract Molecular dynamics (MD) simulations are carried out to simulate thermal bubble nucleation processes in homogeneous and heterogeneous systems, respectively. It is found that the nucleation temperature of water confined in graphene nanochannels depends strongly on the channel height. Once the channel height is reduced below 3.1 nm, the nucleation temperature in the heterogeneous system is higher than that in the corresponding homogeneous systems, which violates the classical nucleation theory that homogeneous thermal bubble nucleation sets an upper limit for nucleation temperature under a given pressure. The abnormal phenomenon is attributed to the formation of solid-like structure for the water confined in nanochannels, in which the whole water system has a lower potential energy than that in the corresponding homogeneous system. Decreasing or increasing the solid-liquid interaction strength may reduce or increase the nucleation temperature, which can explain the nucleation temperature dependent on surface properties in various heterogeneous systems. Meanwhile, we also observe the nucleation sites for the different solid-liquid



Corresponding author. E-mail: [email protected]; [email protected] 1

ACS Paragon Plus Environment

The Journal of Physical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 36

interfacial wettability. Based on this finding, it is demonstrated that the site of thermal bubble nucleation can be actively controlled through adjusting the solid-liquid interaction strength.

Introduction Thermal bubble nucleation initiates the liquid vapor phase transition, which exists widely in natural world and engineering applications. With the growing interests in the development of smaller devices and micro/nano-fluidic systems, thermal bubble nucleation processes were widely applied in ink-jet printers, thermal-bubble-actuated

actuators

and

micro

pumps

in

various

microelectromechanical systems (MEMS)1-5. Meanwhile, thermal bubbles induced by the high density power in electronic devices may affect the transport of fluid through micro/nanochannels, which may pose a hazard to the performance of those devices and shorten their lifespan. Thermal bubble nucleation can be classified into two categories: homogeneous and heterogeneous nucleation. Homogeneous nucleation occurs within a superheated liquid and heterogeneous nucleation occurs at the interface of a superheated liquid and another material phase6-8. Homogeneous nucleation is widely believed to have higher nucleation temperature compared to heterogeneous nucleation using the classical nucleation theory8. This is due to the fact that the heterogeneous nucleation has a lower energy barrier than the homogeneous nucleation and the lower energy barrier allows the heterogeneous sites to nucleate with a smaller input of energy9. Lin et al. observed that thermal bubble nucleation process in a microchannel

2


Page 3 of 36 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


became much more difficult than that in bulk. The temperature for thermal bubble formation is close to the superheat limits for tested liquids. They concluded that the thermal bubble nucleation in the microchannel was more like as quasi homogeneous even though it occurred in a heterogeneous system10-11. They believed that the extremely smooth channel inner surface did not promote the nucleation of thermal bubbles. This important finding attracts massive attentions and many experimental and theoretical studies have been conducted to explain the results.11-12 In addition to experimental and theoretical approaches, molecular dynamics has also been used to study thermal bubble nucleation process in nanoscale. These simulations provide deep understanding of bubble nucleation mechanism and various effects on nucleation process are investigated including the impurities and interfaces13-15. Novak et al.15 carried out a series of studies on nanoscale homogeneous and heterogeneous nucleation. The results of Novak et al. showed that the solid surface in the system could enhance thermal bubble nucleation in heterogeneous systems. Up to now, there exists a contradictory conclusion between the classical nucleation theory and the nucleation phenomenon in micro/nanoscale. In nanochannels, the interactions between the solid wall and the fluid play a crucial role in the behavior of flow properties with small channel width16. Recent experiments and atomistic simulations suggested that the solid-liquid interfacial wettability directly affected liquid transport properties and heat transfer in nanochannels17-21. In this paper, molecular dynamics simulations are carried out to study thermal bubble nucleation processes inside a graphene nanochannel, from which the mechanism

3



behind thermal bubble nucleation is revealed. In a heterogeneous system, thermal bubble nucleation temperature strongly depends on the system size and the solid-liquid interfacial wettability. The underlying mechanism is attributed to that thermal bubble is more easily nucleated at the position where the local potential energy is relatively high in the whole system. Based on this finding, it is demonstrated that the nucleation site can be actively controlled through adjusting the channel surface properties.

Methods A nanochannel is constructed by two pieces of graphene plates as shown in Fig. 1a. Each graphene plate is composed of two layer graphene. Water molecules are confined in between the two graphene plates. A homogeneous simulation system is constructed by removing the two graphene plates from the heterogeneous simulation system. The homogeneous system has the same number of water molecules with the same temperature and pressure as shown in Fig. 1b. The carbon atoms of the top two layers and the bottom inner layer that locates near the fluid are kept thermal vibration. In order to facilitate further observation and statistics, the atoms of the outer layer in the bottom plate are fixed to their initial positions without thermal vibration. The inner layer of carbon atoms (orange atoms) in the bottom plate is set as a heat source. The NPZZT ensemble is used to model the thermal bubble nucleation in liquid confined between the two graphene plates, in which the normal stress (PZZ) in the vertical direction at a pressure of 1 bar is set in the simulation process (Fig. 1a). The top plate can move freely along the z direction with a small displacement in order to

4


Page 4 of 36

Page 5 of 36 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


maintain the pressure (Pzz) at 1 bar. For comparison, we also simulate the homogeneous system using the NPT ensemble under the same situation. The different type atoms interact with each other through the Lennard-Jones (LJ12-6) Potential.22 𝑈𝐿𝐽 = 4𝜀𝑖𝑗

𝜎𝑖𝑗 12

𝜎𝑖𝑗 6

𝑟𝑖𝑗

𝑟𝑖𝑗

{( ) ― ( ) }

(1)

Where ε and σ are the parameters corresponding to the depth of the potential well and the equilibrium distance at which the inter-particle interaction is zero, respectively, and r is the distance between the particles i and j. The notation i and j represent the atom types including carbon, hydrogen and oxygen atoms in Eq. (1). The interaction strength ε between the water molecules and the carbon atoms is adjusted to investigate the effect of surface properties on thermal bubble nucleation in a heterogeneous system. For all simulations, cut-off length is set at 12 Å in the calculation of the Lennard-Jones Potential and short-range electrostatic forces, and the particle mesh Ewald method23 for long-range electrostatics is computed over a 1.0 Å spaced grid. Molecular dynamics simulations are performed using the MD package NAMD.24 Periodical boundary conditions are applied along both the x and y directions, and time step is set as 2 fs. The CHARMM36 force field25 is used to describe carbon atoms (type CA)26 and TIP3P water. Type CA is one type of the carbon atoms in the CHARMM36 force field, which has been adopted to simulate graphene atoms in many researches27-31. The SETTLE algorithm32 is used to maintain the rigidity of water molecules. The temperature is maintained by using the Langevin thermostat to the carbon atoms in the heterogeneous system and to the oxygen atoms in the 5



homogeneous system with a time step of 0.1 ps, respectively. More details are described in Supporting Information. A leap-frog algorithm33 is used to integrate the equations of motion. As shown in Fig. 1, the geometrical size of the simulation system is set as 6 nm×6 nm along the x and y directions. Visualization and analysis are performed using the VMD 1.9.2.34

Figure 1. (a) Schematic of the heterogeneous simulation system. The blue and orange atomic layers represent graphene plates and between the two graphene plates are the water molecules. Periodical boundary conditions are applied along the x and y directions in the heterogeneous simulation system. (b) Periodical boundary conditions are applied along the x, y and z directions in the homogeneous simulation system.

For both heterogeneous and homogeneous simulations ， the system is first energy minimized for 1 ns, followed by 2 ns of NPT equilibration at 300 K under the external pressure of 1 bar. The channel height of each heterogeneous system along the z-dimension is taken as the averaged distance between the inner carbon centers on the 6


Page 6 of 36

Page 7 of 36 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


top and the bottom plates over the last 1 ns of equilibration, as shown in Fig. 2a. After equilibration, the system is heated until its temperature attains a preset value and the simulation is carried out for another 10 ns to observe whether the bubble nucleates. In the simulation ， the temperature for the phase transition is determined by observing whether the system volume increases continuously during the simulation. We can see from Fig. 2b that for both homogeneous and heterogeneous systems, as the temperature rise from 300 K to a given value, the channel height (or the system volume for the homogeneous system) firstly increases slightly due to thermal expansion. If there is no thermal bubble nucleation, the channel height (or the systems volume for homogeneous system) then remains almost constant for more than 10 ns; however, if thermal bubble nucleation occurs, the channel height (system volume) increases dramatically, as shown in Fig. 2c.

Figure 2. A schematic of the simulation system, not drawn to scale. (a) A schematic illustration of

7



a system at 300 K under 1 bar after equilibration. The channel height is defined as the distance between the mass center of the top and the bottom inner graphene sheet that locates near the water molecules. (b) When the temperature increases to a given value, if no thermal bubble nucleates, the channel height only increases slightly and then remains approximately constant during the simulation period. (c) When the temperature increases to a critical value, at which thermal bubble nucleation starts, the channel height of the system will increase continuously. The critical temperature is defined as the thermal bubble nucleation temperature for this system.

In our work, we propose a strategy to extract the bubble nucleation temperature by increasing the system temperature and monitoring whether thermal bubble nucleation occurs in the system in a certain time period. If thermal bubble nucleation does occur, we then reduce the temperature in the following simulation and observe whether thermal bubble nucleation occurs at the lower temperature. Our previous simulation results35 indicate the lowest temperature that thermal bubble nucleation occurs within 10 ns of simulation is very close to the theoretical superheat limit of the studied liquid. The theoretical superheat limit is the upper limit that a liquid phase can exist without experiencing a liquid-vapor phase transition. Thus, we set the simulation time of 10 ns to extract the thermal bubble nucleation temperature. Besides, the interaction strength β between the liquid water molecules and the carbon atoms in the top and bottom plates is adjusted to investigate the effect of the solid-liquid interfacial wettability on the thermal bubble nucleation process. A list of the potential parameters is given in Table I. TABLE I. The Lennard-Jones parameters for the MD model. The subscripts O, H, and C stand for

8


Page 8 of 36

Page 9 of 36 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


oxygen, hydrogen, and carbon atoms, respectively. A strength factor β is used to adjust the surface wettability from hydrophobic to hydrophilic. For cross interactions we use: 𝜎𝑖𝑗 = (𝜎𝑖𝑖 + 𝜎𝑗𝑗) 2, and 𝜀𝑖𝑗 = 𝜀𝑖𝑖𝜀𝑗𝑗, where i, j stand for the atom type of oxygen, hydrogen and carbon.

𝝈𝑶𝑶

𝝈𝑯𝑯

𝝈𝑪𝑪

𝜺𝑶𝑶

𝜺𝑯𝑯

𝜺𝑪𝑪

(Å)

(Å)

(Å)

kJ/mol

kJ/mol

kJ/mol

Homogeneous

3.152

0.400

0.636

0.192

Heterogeneous

3.152

0.400

0.636

0.192

3.552

0.293*β

Results and discussions 1. Heterogeneous thermal bubble nucleation in nanochannels. A series of heterogeneous systems with different channel heights are simulated. After 1ns for energy minimization, the simulation system runs another 2 ns for NPT equilibration at 300 K under the external pressure of 1 bar. Then, the temperature of the heat source is increased to a preset value. Heat will transport from the carbon atoms to the water molecules due to temperature gradient. Once the water is heated to a critical temperature, the liquid may shift to a metastable state. Because of the density fluctuation in the superheated liquid, the low-density region is formed gradually in several nanoscale regions, which implies the embryo of thermal bubble is formed. With the continue increase of the temperature, the accumulation energy will cause the liquid around the embryo to vaporize, leading to the growth of the bubble and then the phase transition occurs. The thermal bubble nucleation process of the homogeneous system is similar to that of the heterogeneous system. See the 9



Supporting Information for the details of the thermal bubble nucleation process of the homogeneous system. Figure 3 presents the thermal bubble nucleation temperature dependent on channel height in a heterogeneous system. The dashed line represents the nucleation temperature of the homogeneous system with the same water molecules in the corresponding heterogeneous system. For example, the nucleation temperature is 547 K for the homogeneous system containing 7684 water molecules. Compared with the homogeneous system, thermal bubble nucleation temperature increases sharply once the channel height decreases below 3.1 nm. When the channel height decreases from 6.6 nm to 0.7 nm, the nucleation temperature rises from 541 K to 627 K (the red curve). Figure 3 demonstrates that in a nanoscale confined space, the nucleation temperature of a heterogeneous system is significantly higher than that of the corresponding homogeneous system at the channel height below 3.1 nm with the same external pressure. This result is contrary to the common understanding8 that a homogeneous system always has a higher thermal bubble nucleation temperature than that in the corresponding heterogeneous nucleation under the same pressure.

Figure 3. The extracted nucleation temperature of a heterogeneous system as a function of the

10


Page 10 of 36

Page 11 of 36 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


channel height. For comparison, the nucleation temperature for the corresponding homogeneous system is also shown.

To explain this unexpected finding, we calculate the water molecule number density distribution as a function of channel height at 300 K. Water density distribution in nanochannels is one fundamental property of liquid water molecules, which has been extensively studied36-39. Figure 4 presents the water molecule number density distribution along the channel height (z-axis) for four nanochannels with height of 3.1, 1.9, 1.3 and 0.7 nm at 300 K, respectively. Figure 4a stands for the channel with height of 3.1 nm, in which the water density approaches bulk water except the neighboring three layers (peak 1, 2, and 3) near the solid surface. It is found from Fig. 4a that when the fluid molecules are confined in between two solid plates, the fluid can no longer be taken as homogeneous, and strong oscillations of fluid density occur near the solid-liquid interface. For the water molecules near the solid wall, the solid-like structure is formed due to the strong van der Waals forces. For the water molecules far away from the solid wall, the van der Waals forces between the solid wall and the water becomes weak, and the water density tends to the bulk water density. When the channel height is reduced to 1.9 nm as shown in Fig. 4b, the third peaks from the two side walls meet together, so five layers of water molecules are allowed in between the two solid plates. As the channel height is decreased to 1.3 nm in Fig. 4c, the second peaks overlap, so only three layers of water molecules are allowed in between the two solid plates. In this case, the density peak in the middle of the two plates is about 1.5 times larger than that in bulk water. As

11



Page 12 of 36

presented in Fig. 4d, when the channel height is 0.7 nm, only one layer of water molecules are allowed in between the two solid graphene plates with the peak of water density as high as 169 atom/nm3, which is about 5 times larger than that in bulk water of 33 atoms/nm338. It is well-known that solid walls will influence the arrangement of liquid atoms near the interface, leading to the near-wall layering structure5,

40

and if the channel

height is small, the layering structure penetrates through the entire liquid system, which means that all water molecules will form more ordered arrangement. It is believed that this more ordered arrangement will hinder thermal bubble nucleation. The smaller height the channel is, the stronger influence the solid channel surfaces enforce. At small channels, the effect of walls on fluid atoms extends practically to the whole fluid range, resulting in an important modification of the confined water behavior in comparison to the bulk water state where fluid atoms are embedded in the fluid environment41. Therefore, the nucleation temperature of the heterogeneous system depends strongly on the channel height.

12


Page 13 of 36 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 4. The number density of water molecules with different channel heights at 300 K and under the pressure of 1 bar: channel height of (a) 3.1 nm (3357 water molecules), (b) 1.9 nm (1866 water molecules), (c) 1.3 nm (1212 water molecules), and (d) 0.7 nm (337water molecules). The filled circles denote the channel wall atoms, and in between the two walls are the water molecules. The water molecules are highly structured near the wall (the channel wall atoms and the water molecules not to scale).

To further understand why the nucleation temperature strongly depends on the channel height, we analyze the system energy corresponding to different heterogeneous systems. The total energy of a system is the sum of the kinetic and the potential energy of all molecules. The kinetic energy of an equilibrium system is solely determined by the system temperature (i.e., the specific kinetic energy is the same for both homogeneous and heterogeneous systems at the same temperature). Therefore, we analyze the potential energy distribution of the water molecules along the z direction in a heterogeneous system as a function of channel height. In our 13



simulation, the potential energy per unit mole is defined as the specific potential energy. Figure 5 shows the potential energy distribution of the water molecules versus the distance from the bottom solid plate with channel height 4.9 nm. It can be seen that the potential energy demonstrates a layering distribution near the solid plates. For comparison, we also present the water number density in Fig. 5. Similar to the density profiles across the channel, the potential energy also demonstrates layering structure, while its valley corresponds to the peak of the water number density. It indicates that for heterogeneous systems, the liquid molecules near the solid walls experience strong confinement effects and has lower potential energy compared with bulk water. In addition, the confinement effect becomes strong with the decrease of the channel height.

Figure 5. The number density and the potential energy distributions of water molecules along the z direction with channel height of 4.9 nm at 300 K and under the pressure of 1 bar.

Figure 6 illustrates the average specific potential energy (normalized to unit mole) of the water molecules in the heterogeneous systems at 300 K corresponding to the channel height. As a comparison, the navy blue line in Fig. 6 shows the specific potential energy of the homogeneous system. As we know, thermal bubble nucleation 14


Page 14 of 36

Page 15 of 36 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


is the process during which the superheated liquid molecules overcome the energy barrier to become vapor. Therefore, the system with low potential energy has high energy barrier, which needs high kinetic energy to overcome the energy barrier for thermal bubble to nucleate. As demonstrated in Fig. 6, once the channel height is below 3.1 nm, the potential energy in the heterogeneous system is obvious smaller than that corresponding to the homogeneous system. Carefully comparing Fig. 3, Fig. 4 and Fig. 6, we can find the thermal nucleation temperature strongly depends on the potential energy. As shown in Fig. 4a, the near-wall layering structure of water molecules is about 1.1 nm thick away from the mass center of the inner graphene sheet. When the channel height is below 2.2 nm, all of the water molecules behave as solid-like structure, corresponding to that the specific potential energy decreases and the thermal bubble nucleation temperature increases synchronously. When the channel height is between 2.2 nm and 3.1 nm, the potential energy of water molecules in the heterogeneous channel is slightly lower than that in the homogeneous system. Thus, the nucleation temperature in the heterogeneous system is slightly higher than that in the homogeneous system. Once the channel height is higher than 3.1 nm, the ratio of the water molecules in the near wall region to the total water molecules in the heterogeneous system becomes small. Continually increasing the channel height, the central part of the water molecules behave the properties as a quasi-homogeneous system including water density and potential energy. Figure 6 demonstrates that the specific potential energy increases slowly and saturates to a limit with the increase of the channel height. The limit value of the specific potential energy in the

15



heterogeneous system is a little bit higher than that at the homogeneous system, which is attributed to the NPT system. In a NPT system, temperature is a control parameter. The system temperature will fluctuate around the given temperature T. It is found that the average temperature T at the central part of a heterogeneous system is always higher than that at the near wall region, which is also higher than that at the corresponding homogeneous system. The higher temperature at the central part of the heterogeneous system leads to the system has a higher potential energy compared to the homogeneous system, although the same temperature is set for the heterogeneous system and the homogeneous system. Here we note that this difference is due to the limited system size studied in our simulation systems. If the system size enlarges, this difference should reduce and even disappear when the system size is large enough. More details are presented in Supporting Information.

Figure 6. The specific potential energy of water molecules in a heterogeneous system as a function of the channel height (T = 300 K). For comparison, the potential energy of the corresponding homogeneous system under the same temperature is also shown.

2. The effect of the surface properties on nucleation temperature and site

16


Page 16 of 36

Page 17 of 36 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


As demonstrated above, the nucleation temperature strongly depends on the potential energy. Considering the potential energy is affected greatly by the surface properties of the channel wall, we investigate the effect of the solid-liquid interfacial wettability on nucleation temperature. The solid-liquid interfacial wettability can be adjusted by tuning the strength factor β, which stands for the interaction strength between the liquid and the solid wall. The strength factor β is set as 0.5, 1.0 and 1.5 to stand for the weak, medium and strong cases, respectively. Figure 7a shows that the nucleation temperature as a function of the channel height for the three different solid-liquid interaction cases. Clearly, the nucleation temperature as a function of the channel height follows the same trend in all the three cases, i.e. the thermal bubble nucleation temperature decreases with the channel height. At small channel heights, the effect of the solid-liquid interactions extends almost to the whole fluid, which raises the nucleation temperature. It can be seen that as the interaction strength increase, the thermal bubble nucleation temperature rises dramatically. Decreasing the solid-liquid interaction strength may decrease the nucleation temperature. In the weak interaction case, the increasing of nucleation temperature with the decreasing channel height is slight because of the influence of the solid phases on liquid molecules has been weakened. Weaker solid-liquid interaction strength reduces the nucleation temperature because it is easier for water molecules to escape from the weak bonding at the solid-liquid interface. Therefore, voids can be generated at lower temperature, which promotes phase transition possibility and reduces the nucleation temperature. When the solid-liquid interaction

17



strength increases, more ordered arrangement of the near-wall molecules tends to suppress void generation near the solid plate and prevent phase transition occurring. The inset in Fig. 7a shows the effect of the interaction strength between the liquid water molecules and the carbon atoms on the nucleation temperature with the channel height at 4.9 nm. We quantify the solid-liquid interaction strength over a broad range of surface chemistries from hydrophobic to hydrophilic. Given a channel height, it can be seen that as the interaction strength decreases, the nucleation temperature drops significantly, while the nucleation temperature may saturate at a certain value with the increase of the solid-liquid interaction strength. Figure 7b presents the potential energy as a function of the solid-liquid interaction strength at 300K with the channel height at 4.9 nm.

Figure 7. (a) The extracted nucleation temperature of a heterogeneous system as a function of the channel height for three cases with different strength interactions. For comparison, the inset shows the nucleation temperature as a function of the solid-liquid interaction strength in the heterogeneous system containing 5538 water molecules. (b) The potential energy as a function of the solid-liquid interaction strength in the heterogeneous system containing 5538 water molecules.

Meanwhile, we also observe the nucleation sites for the three cases with the 18


Page 18 of 36

Page 19 of 36 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


different interaction strength factor β set at 0.5, 1.0 and 1.5, which stand for the weak, medium and strong cases, respectively. Figure 8 presents the effect of the solid-liquid interfacial wettability on thermal bubble nucleation sites with the channel height at 4.9 nm. In our simulations, as shown in Fig. 8a, thermal bubble starts to form at the solid-liquid interface for the weak case. This result is in agreement with the results of previous literatures13,

15.

For weaker interatomic interactions, the solid wall shows

hydrophobicity. Accordingly, the bonding strength between the water molecules and solid plates is reduced and it is easier for water molecules to escape from the solid plate to form voids. This makes the formation of voids near the weak surface easily. Different from the weak case, thermal bubble nucleation occurs in the near center region for the medium and strong cases in the heterogeneous system, as shown in Fig. 8b, c. In the medium and strong cases, the solid wall shows hydrophilicity and the solid-like structure of the near-wall liquid molecules with the lower potential energy would hinder bubble nucleation. The water molecules in the near center region behave like the water molecules in the corresponding homogeneous system. This is the reason that thermal bubble starts to nucleate at the near center region in such heterogeneous systems. In fact, once the interaction strength increases to a certain value, the temperature of thermal bubble nucleation of the heterogeneous system saturates, which means the nucleation temperature is not dependent on β as demonstrated in the inset of Fig. 7a. The reason is attributed to the van der Waals interaction range. Increasing β does not affect the local potential energy far away from the solid wall. Accordingly, the thermal bubble nucleation temperatures in the heterogeneous

19



systems saturate and tend to a limit. Based on the above analysis, we can draw a conclusion that the smaller interaction strength between the channel surface and the liquid, the more easy for thermal bubble to nucleate.

Figure 8. Snapshots of thermal bubble nucleation for a heterogeneous system containing 5538 water molecules in (a) the weak case, (b) the medium case, and (c) the strong case.

Based on this finding, we divide the bottom inner graphene layer into two regions. One region is set as a weak solid-liquid interaction strength region with the strength factor β set at 0.1. The rest region is set as a medium solid-liquid interaction strength region with β set at 1.0. As shown in Fig. 9a, the weak region is on the left, β on the left part of the bottom plate is set as 0.1 and the thermal bubble nucleation does initiate on the left of the solid-liquid interface. In Fig. 9b, the weak region is located in the middle of the bottom plate. Thermal bubble nucleates first in this region. Once the weak region is set on the right part of the bottom plate, the site of the thermal bubble nucleation shifts to the right region as shown in Fig. 9c. In all the above three 20


Page 20 of 36

Page 21 of 36 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


simulations, the thermal bubble nucleation initiates at the solid-liquid interface, where the interaction between the water and the bottom plate is weak. By changing the strength of solid-liquid interaction in a certain region of the channel wall as shown in Fig. 9, the thermal bubble nucleation sites can be controlled accurately. This phenomenon provides us with a method for controlling and manipulating thermal bubble nucleation site in micro/nanoscale confined space.

Figure 9. Snapshots of thermal bubble nucleation for a heterogeneous system containing 5538 water molecules in three cases. The yellow atoms represent the regions where the solid-liquid interaction strength is weak. The strength factor β on the weak region of the bottom plate is set as 0.1. (a) The weak region is on the left. (b) The weak region is in the middle. (c) The weak region is on the right.

Conclusions In summary, we perform molecular dynamics simulations of thermal bubble nucleation processes in homogeneous and heterogeneous systems, respectively. It is found that for water confined in between two solid plates with several nanometers 21



apart, the heterogeneous thermal bubble nucleation temperature strongly depends on the system size and the solid-liquid interfacial wettability. More importantly, the heterogeneous bubble nucleation temperature can be significantly higher than the corresponding homogeneous bubble nucleation temperature when the channel height is below 3.1 nm, which is in contrary to the common understanding in the classical nucleation theory. This unexpected observation is attributed to the near-wall layering structure of the water molecules in nanoconfined systems. It is believed that this more ordered arrangement will hinder thermal bubble nucleation due to the lower potential energy. The results show that thermal bubble nucleation in nanoscale exhibits a significantly different behavior than that of the bulk fluid below a critical channel height. The nucleation site of the thermal bubble depends strongly on the interaction strength between the fluid and the channel surface. Our work demonstrates that it is possible to control the site of thermal nucleation through tuning the channel surface properties.

Supporting Information See Supporting Information for the details of thermal bubble nucleation in Si channels, the effects of boundary conditions on thermal nucleation temperature, temperature control in a heterogeneous system, thermal bubble nucleation in a homogeneous system, the potential energy in heterogeneous and homogeneous systems, and the effects of the channel height on the nucleation temperature when the channel height below 3.1 nm.

Author Information 22


Page 22 of 36

Page 23 of 36 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Corresponding Authors *E-mail: [email protected]. Tel: 86-13915968758. (J.S.) *E-mail: [email protected]. Tel: 86-13815888816. (Y.C.)

Acknowledgements This work was supported by the National Natural Science Foundation of China (Grant No.51435003, 51675101 and 51705074). Zhongwu Li is supported by the Fundamental Research Funds for the Central Universities and the Innovative Project for Graduate Students of Jiangsu Province (Grant No. KYCX18_0067).

Reference 1. Tsai, J.-H.; Lin, L., A Thermal-Bubble-Actuated Micronozzle-Diffuser Pump. J. Microelectromech. Syst. 2002, 11 (6), 665-671. 2. Shikida, M.; Imamura, T.; Ukai, S.; Miyaji, T.; Sato, K., Fabrication of a Bubble-Driven Arrayed Actuator for a Tactile Display. J. Micromech. Microeng. 2008, 18 (6), 065012. 3. Zhang, K.; Jian, A.; Zhang, X.; Wang, Y.; Li, Z.; Tam, H.-y., Laser-Induced Thermal Bubbles for Microfluidic Applications. Lab Chip 2011, 11 (7), 1389-1395. 4. Van Den Broek, D.; Elwenspoek, M., Bubble Nucleation in an Explosive Micro-Bubble Actuator. J. Micromech. Microeng. 2008, 18 (6), 064003. 5. Karniadakis, G.; Beskok, A.; Aluru, N., Simple Fluids in Nanochannels. Springer: 2005. 6. Carey, V. P., Liquid Vapor Phase Change Phenomena: an Introduction to the Thermophysics of Vaporization and Condensation Processes in Heat Transfer Equipment. CRC Press: 2018. 7. Debenedetti, P. G., Metastable Liquids: Concepts and Principles. Princeton University Press: 1996. 8. Blander, M.; Katz, J. L., Bubble Nucleation in Liquids. AICHE J. 1975, 21 (5), 833-848. 9. Colton, J.; Suh, N., The Nucleation of Microcellular Thermoplastic Foam with Additives: Part I: Theoretical Considerations. Polym. Eng. Sci. 1987, 27 (7), 485-492. 10. Lin, L., Microscale Thermal Bubble Formation: Thermophysical Phenomena and Applications. Nanoscale Microscale Thermophys. Eng. 1998, 2 (2), 71-85. 11. Peng, X. F.; Hu, H.; Wang, B., Boiling Nucleation during Liquid Flow in Microchannels. Int. J. Heat Mass Transf. 1998, 41 (1), 101-106. 12. Zhang, J.; Peng, X.; Peterson, G., Analysis of Phase-Change Mechanisms in Microchannels using Cluster Nucleation Theory. Nanoscale Microscale Thermophys. Eng. 2000, 4 (3), 177-187. 13. Nagayama, G.; Tsuruta, T.; Cheng, P., Molecular Dynamics Simulation on Bubble Formation in a Nanochannel. Int. J. Heat Mass Transf. 2006, 49 (23-24), 4437-4443. 14. Weijs, J. H.; Snoeijer, J. H.; Lohse, D., Formation of Surface Nanobubbles and the Universality of Their Contact Angles: a Molecular Dynamics Approach. Phys. Rev. Lett. 2012, 23



108 (10), 104501. 15. Novak, B. R.; Maginn, E. J.; McCready, M. J., Comparison of Heterogeneous and Homogeneous Bubble Nucleation using Molecular Simulations. Phys. Rev. B 2007, 75 (8), 085413. 16. Israelachvili, J., Solvation, Structural and Hydration Forces. Intermolecular and surface forces 2011, 260-288. 17. Ramos-Alvarado, B.; Kumar, S.; Peterson, G., Solid-Liquid Thermal Transport and Its Relationship with Wettability and the Interfacial Liquid Structure. J. Phys. Chem. Lett. 2016, 7 (17), 3497-3501. 18. Shenogina, N.; Godawat, R.; Keblinski, P.; Garde, S., How Wetting and Adhesion Affect Thermal Conductance of a Range of Hydrophobic to Hydrophilic Aqueous Interfaces. Phys. Rev. Lett. 2009, 102 (15), 156101. 19. Cottin-Bizonne, C.; Barrat, J.-L.; Bocquet, L.; Charlaix, E., Low-Friction Flows of Liquid at Nanopatterned Interfaces. Nat. Mater. 2003, 2 (4), 237. 20. Nagayama, G.; Cheng, P., Effects of Interface Wettability on Microscale Flow by Molecular Dynamics Simulation. Int. J. Heat Mass Transf. 2004, 47 (3), 501-513. 21. Liu, Q.; Xu, B., Actuating Water Droplets on Graphene via Surface Wettability Gradients. Langmuir 2015, 31 (33), 9070-9075. 22. MacKerell Jr, A. D.; Bashford, D.; Bellott, M.; Dunbrack Jr, R. L.; Evanseck, J. D.; Field, M. J.; Fischer, S.; Gao, J.; Guo, H.; Ha, S., All-Atom Empirical Potential for Molecular Modeling and Dynamics Studies of Proteins. J. Phys. Chem. B 1998, 102 (18), 3586-3616. 23. Essmann, U.; Perera, L.; Berkowitz, M. L.; Darden, T.; Lee, H.; Pedersen, L. G., A Smooth Particle Mesh Ewald Method. J. Chem. Phys. 1995, 103 (19), 8577-8593. 24. Phillips, J. C.; Braun, R.; Wang, W.; Gumbart, J.; Tajkhorshid, E.; Villa, E.; Chipot, C.; Skeel, R. D.; Kale, L.; Schulten, K., Scalable Molecular Dynamics with NAMD. J. Comput. Chem. 2005, 26 (16), 1781-1802. 25. Vanommeslaeghe, K.; Hatcher, E.; Acharya, C.; Kundu, S.; Zhong, S.; Shim, J.; Darian, E.; Guvench, O.; Lopes, P.; Vorobyov, I., CHARMM General Force Field: A Force Field for Drug-Like Molecules Compatible with the CHARMM All-Atom Additive Biological Force Fields. J. Comput. Chem. 2010, 31 (4), 671-690. 26. Wells, D. B.; Belkin, M.; Comer, J.; Aksimentiev, A., Assessing Graphene Nanopores for Sequencing DNA. Nano Lett. 2012, 12 (8), 4117-23. 27. Shankla, M.; Aksimentiev, A., Conformational Transitions and Stop-and-Go Nanopore Transport of Single-Stranded DNA on Charged Graphene. Nat. Commun. 2014, 5, 5171. 28. Titov, A. V.; Král, P.; Pearson, R., Sandwiched Graphene-Membrane Superstructures. ACS Nano 2009, 4 (1), 229-234. 29. Banerjee, S.; Wilson, J.; Shim, J.; Shankla, M.; Corbin, E. A.; Aksimentiev, A.; Bashir, R., Slowing DNA Transport using Graphene-DNA Interactions. Adv. Funct. Mater. 2015, 25 (6), 936-946. 30. Gracheva, M. E.; Xiong, A.; Aksimentiev, A.; Schulten, K.; Timp, G.; Leburton, J.-P., Simulation of the Electric Response of DNA Translocation through a Semiconductor Nanopore-Capacitor. Nanotechnology 2006, 17 (3), 622. 31. Wilson, J.; Sloman, L.; He, Z.; Aksimentiev, A., Graphene Nanopores for Protein Sequencing. Adv. Funct. Mater. 2016, 26 (27), 4830-4838. 24


Page 24 of 36

Page 25 of 36 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


32. Miyamoto, S.; Kollman, P. A., Settle: An Analytical Version of the SHAKE and RATTLE Algorithm for Rigid Water Models. J. Comput. Chem. 1992, 13 (8), 952-962. 33. Allen, M. P.; Tildesley, D. J., Computer Simulation of Liquids. Oxford University Press: 2017. 34. Humphrey, W.; Dalke, A.; Schulten, K., VMD: Visual Molecular Dynamics. J. Mol. Graph. 1996, 14 (1), 33-38. 35. Chen, M.; Yang, J.; Gao, Y.; Chen, Y.; Li, D., Molecular Dynamics Studies of Homogeneous and Heterogeneous Thermal Bubble Nucleation. J. Heat Transf.-Trans. ASME 2014, 136 (4), 041502. 36. Ho, T. A.; Striolo, A., Molecular Dynamics Simulation of the Graphene-Water Interface: Comparing Water Models. Mol. Simul. 2014, 40 (14), 1190-1200. 37. Chen, H.; Ruckenstein, E., Nanomembrane Containing a Nanopore in an Electrolyte Solution: a Molecular Dynamics Approach. J. Phys. Chem. Lett. 2014, 5 (17), 2979-2982. 38. Cheng, L.; Fenter, P.; Nagy, K.; Schlegel, M.; Sturchio, N., Molecular-Scale Density Oscillations in Water Adjacent to a Mica Surface. Phys. Rev. Lett. 2001, 87 (15), 156103. 39. Cicero, G.; Grossman, J. C.; Schwegler, E.; Gygi, F.; Galli, G., Water Confined in Nanotubes and between Graphene Sheets: A First Principle Study. J. Am. Chem. Soc. 2008, 130 (6), 1871-1878. 40. Xu, D.; Leng, Y.; Chen, Y.; Li, D., Water Structures near Charged (100) and (111) Silicon Surfaces. Appl. Phys. Lett. 2009, 94 (20), 201901. 41. Sofos, F.; Karakasidis, T.; Liakopoulos, A., Transport Properties of Liquid Argon in Krypton Nanochannels: Anisotropy and Non-Homogeneity Introduced by the Solid Walls. Int. J. Heat Mass Transf. 2009, 52 (3), 735-743.

25



TOC graphic

26


Page 26 of 36

Page 27 of 36 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 1. (a) Schematic of the heterogeneous simulation system. The blue and orange atomic layers represent graphene plates and between the two graphene plates are the water molecules. Periodical boundary conditions are applied along the x and y directions in the heterogeneous simulation system. (b) Periodical boundary conditions are applied along the x, y and z directions in the homogeneous simulation system. 631x338mm (300 x 300 DPI)



Figure 2. A schematic of the simulation system, not drawn to scale. (a) A schematic illustration of a system at 300 K under 1 bar after equilibration. The channel height is defined as the distance between the mass center of the top and the bottom inner graphene sheet that locates near the water molecules. (b) When the temperature increases to a given value, if no thermal bubble nucleates, the channel height only increases slightly and then remains approximately constant during the simulation period. (c) When the temperature increases to a critical value, at which thermal bubble nucleation starts, the channel height of the system will increase continuously. The critical temperature is defined as the thermal bubble nucleation temperature for this system. 170x109mm (300 x 300 DPI)


Page 28 of 36

Page 29 of 36 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 3. The extracted nucleation temperature of a heterogeneous system as a function of the channel height. For comparison, the nucleation temperature for the corresponding homogeneous system is also shown. 84x64mm (300 x 300 DPI)



Figure 4. The number density of water molecules with different channel heights at 300 K and under the pressure of 1 bar: channel height of (a) 3.1 nm (3357 water molecules), (b) 1.9 nm (1866 water molecules), (c) 1.3 nm (1212 water molecules), and (d) 0.7 nm (337water molecules). The filled circles denote the channel wall atoms, and in between the two walls are the water molecules. The water molecules are highly structured near the wall (the channel wall atoms and the water molecules not to scale). 390x237mm (300 x 300 DPI)


Page 30 of 36

Page 31 of 36 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 5. The number density and the potential energy distributions of water molecules along the z direction with channel height of 4.9 nm at 300 K and under the pressure of 1 bar. 84x56mm (300 x 300 DPI)



Figure 6. The specific potential energy of water molecules in a heterogeneous system as a function of the channel height (T = 300 K). For comparison, the potential energy of the corresponding homogeneous system under the same temperature is also shown. 84x64mm (300 x 300 DPI)


Page 32 of 36

Page 33 of 36 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 7. (a) The extracted nucleation temperature of a heterogeneous system as a function of the channel height for three cases with different strength interactions. For comparison, the inset shows the nucleation temperature as a function of the solid-liquid interaction strength in the heterogeneous system containing 5538 water molecules. (b) The potential energy as a function of the solid-liquid interaction strength in the heterogeneous system containing 5538 water molecules. 170x69mm (300 x 300 DPI)



Figure 8. Snapshots of thermal bubble nucleation for a heterogeneous system containing 5538 water molecules in (a) the weak case, (b) the medium case, and (c) the strong case. 170x107mm (300 x 300 DPI)


Page 34 of 36

Page 35 of 36 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 9. Snapshots of thermal bubble nucleation for a heterogeneous system containing 5538 water molecules in three cases. The yellow atoms represent the regions where the solid-liquid interaction strength is weak. The strength factor β on the weak region of the bottom plate is set as 0.1. (a) The weak region is on the left. (b) The weak region is in the middle. (c) The weak region is on the right. 170x95mm (300 x 300 DPI)



Table of Contents graphic showing the abnormal nucleation temperature of a heterogeneous system as a function of the channel height that arising from the solid-like water structure and the corresponding potential energy in the highly confined space. 59x44mm (300 x 300 DPI)


Page 36 of 36

Thermal Bubble Nucleation in Graphene Nanochannels - The Journal

Recommend Documents