Characterization and Mechanism of a New Superhydrophobic Deicing Coating Used for Road Pavement

Owing to its high efficiency and low environmental impact, superhydrophobic deicing coating material has a bright future for application on road pavements. In this paper, a heterogeneous nucleation ice crystal growth model is proposed, with particular focus on the effect of surface roughness and the contact angle. The ice suppression mechanism of superhydrophobic materials is determined by this model and experimentally verified. The experimental results of the water contact angle and anti-skid tests illustrated that the prepared TiO2–octadecanoic acid coating material has a contact angle greater than 150° and good skid resistance. The freezing test confirms that the applied coating on the surface can effectively delay the crystallization of water droplets and maintain the waterdrop’s semi-spherical shape after freezing. The microstructure observation demonstrates the TiO2–octadecanoic acid material has a good micro-nano mastoid structure. Consequently, the proposed coating materials could possibly be utilized for effectively enhancing the deicing performance of pavements.


Introduction
The bond formed between ice and the substrate has shown many drawbacks in the fields of transportation, power transmission, aviation, and coal mining, which could cause substantial economic losses and human safety risks. In terms of aviation, the ice accumulated on the surface of an aircraft can change its shape, resulting in stability loss [1][2][3]. In the coal mining industry, the cohesiveness of ice causes the wet coal to adhere to the tram and decreases the bearing capacity of the tram. In the field of transportation, the ice formed on the pavement significantly reduces skid resistance, which could lead to serious traffic accidents. It is noted that driving under snowfall and sleet with accumulated ice on road pavements has made conditions horrendous, increasing the traffic accident rate and casualty rate by 84% and 75%, respectively [4,5]. Therefore, reliable and cost-effective deicing methods need to be investigated to minimize the safety risks on driving on pavements under serve weather conditions.
Commonly used deicing methods include the application of chemical products, heating, and superhydrophobic material. The mechanisms of road deicing methods are mainly divided into passive and active deicing [3,[6][7][8][9] methods. The passive deicing method (e.g., using chemical products) is simple and of relatively low cost. However, it has low efficiency and is often labor-and material-intensive, and may also cause significant damages to pavement structure. In contrast, the active deicing method has shown high efficiency and is less detrimental to pavement structure. However, the current application of the active deicing method (e.g., heating) used in the market is usually costly. Using superhydrophobic materials, as another potential active deicing option, has drawn increasing attention due to its simple preparation, the fact that it is economic, and its pollution-free [3,[10][11][12] characteristics. Therefore, it is ideal to apply the superhydrophobic materials with the above advantages as ice suppression technology for road pavements.
Using a superhydrophobic surface to reduce the adhesion of ice is an emerging technology. To date, several studies have been conducted to investigate the adhesion mechanism of ice on superhydrophobic surfaces [13][14][15]. However, whether superhydrophobic surfaces can reduce the adhesion of ice is still not clearly demonstrated. In 2010, Mishchenko et al. [16] showed that the superhydrophobic surface could remain entirely ice-free, even when temperatures dropped to −25 • C. This is because the diffusion and the contraction time of water droplets on the superhydrophobic surface is shorter than the ice nucleation time. In 2013, Wang [17] et al. proved that under extreme condensation conditions (−10 • C, 90% relative humidity (RH)), both single and successive supercooled water droplets could be rebound to the surface and rolled off the surface at a tilt angle greater than 30 • . They also studied the effects of droplet velocity, size, and freezing conditions (temperature and RH). The results showed that small water droplets with a Weber number of 0.8 could roll down from these superhydrophobic surfaces at −10 • C and 85-90% RH with an angle of 10 • , and concluded that these surfaces might be ice proof.
However, some research showed that superhydrophobic material has no obvious ice-proofing effects. Varanasi [18] et al. studied the application of deicing on the superhydrophobic surface, and the results demonstrated that frost easily formed on the rough superhydrophobic surface at high RH, and then the ice adhesion on this surface was much greater than that on the smooth substrate, which indicated that the superhydrophobic surface maybe not be suitable for the deicing purpose. Chen [19] et al. prepared 13 silicon wafer surfaces with either superhydrophilic or superhydrophobic wettability and studied the influence of surface morphology and chemical properties on the adhesion of a wafer on a silicon surface. The experimental results showed that the superhydrophobic surface could not reduce the adhesion of ice, and the adhesion of ice on the superhydrophilic and superhydrophobic surface was basically the same.
The aforementioned results indicated that the property of the substrate surface is a critical factor in influencing the formation of ice. Regarding the pavement surface, the property of the top surface is critical to the formation of ice. The application of superhydrophobic coatings on the pavement surface may modify the deicing performance of existing roads. However, currently, superhydrophobic surfaces are rarely used for road deicing [8,20,21]. Therefore, fundamental research is required to investigate the theory and feasibility of using superhydrophobic materials for the purpose of pavement surface deicing.
This study aims to explore the mechanism of superhydrophobic materials on ice suppression and the experimental application of superhydrophobic materials on road surfaces. Based on the innovatively proposed growth model of "ice nucleation", the influencing factors of ice nucleation are theoretically analyzed, where the contact angle and surface roughness were investigated. The model hypothesized that the rough surface and large contact angle of a superhydrophobic surface could suppress the ice from forming. Meanwhile, experimental tests were conducted using the superhydrophobic coating material. The synthesized superhydrophobic material was coated on a pavement and characterized, and its effect on ice suppression was examined to support the proposed theory.

Theory of Ice Suppression Mechanism in Early Stage
At the end of the 19th century, Gibbs first proposed the theory of droplet nucleation. According to this theory [22,23], nucleation can generally be divided into two types: homogeneous and heterogeneous nucleation. In this section, the mechanism of delaying ice nucleation is analyzed from the perspectives of "air cushion" and "contact angle affecting ice nucleation", which provides a theoretical basis for the early stage of ice nucleation on the superhydrophobic coated surface. Generally, in the process of phase transformation, the medium follows the low energy path with the minimum resistance and the fastest speed, and reaches a new stage that has the most suitable environment for nucleation. According to the Second Law of Thermodynamics, the Gibbs free energy of the system is used to determine the direction of the phase transition path in the process of crystal nucleation: In Equation (1), ∆G V and ∆G S are the increment of volume free energy and interface free energy, respectively, ∆G is the energy barrier to overcome in the process of nucleation dynamics, which can be called the system difference of Gibbs free energy before and after phase transition. If the crystal nucleus wants to expand the surface area of the solid phase, it needs to overcome the work of the surface tension to increase the interface free energy. Therefore, when the volume of the crystal nucleus decreases, the phase transformation will be promoted, while the increase in its surface area will hinder the phase transformation.
From Equation (2), the reduction of ∆G V requires sufficient conditions as the system is supercooled. When homogeneous nucleation begins, ∆G increases. Consequently, ∆G > 0, and even in the state of threshold nucleation, the decrease in ∆G V is only two-thirds of that of the increase in ∆G S .
At the threshold nucleation time, the decrease in ∆G V is not enough to overcome the increase in ∆G S . Hence, other energies must be overcome to make the value of ∆G less than zero. From the thermodynamic point of view, nucleation cannot be formed in this case. However, for any system, there is local, small, and inhomogeneous wave energy, from which water droplets could obtain energy, and facilitate nucleation. The basis of the formation of a crystal embryo is the slight structure fluctuation and phase fluctuation. Therefore, the fluctuation of energy and structure in the system is a necessary condition for nucleation.
Under a normal pressure, the nucleation process of ice crystallization can be described as follows: when the ambient temperature gradually decreases, and the solution around the heterogeneous wall reaches a certain degree of supercooling, the crystal embryo will be formed. At this time, two scenarios can happen: the volume of crystal embryo increases within a certain time. If the supercooling degree of the solution around the heterogeneous wall continues to increase, the total free energy of the solution around the heterogeneous wall decreases. To a certain value, the melting rate of the embryo decreases, while its growth rate increases with the increase in the supercooling degree. When the volume of the embryo reaches a certain value, the ice core forms. The nucleation process begins when the embryo changes to the ice core. According to the different contact angles, ice nucleation can be divided into two processes: homogeneous nucleation and heterogeneous nucleation.

Homogeneous Nucleation
When homogeneous nucleation occurs, the surface tension of ice is isotropic, so the crystal nucleus is spherical. The nucleation structure is built into a sphere model with r (radius) as the fixed value, and the following results can be obtained: When the liquid water phase changes into solid ice, the change of unit volume free energy ∆G can be expressed as follows: The enthalpy and entropy changes are as follows (5) and (6).
The formula shows the solidification latent heat and phase transformation equilibrium temperature.
Substituting Equations (5) and (6) into Equation (4), Equation (7) is obtained: where ρ ice is the density of ice, T and T eq are the room temperature and the phase transition equilibrium temperature, respectively, and h f represents the latent heat of solidification. Surface tension (∆G S ) can be expressed in γ iw . If Equation (7) is substituted into Equation (3), resulting When the equilibrium temperature of the solution phase is lower than that of the system, ∆G does not decrease after nucleation. According to the nucleation determination conditions, the ice core does not grow in volume, so homogeneous nucleation can only occur in supercooled water. If the temperature of the system has reached the subcooling condition, a threshold radius r* exists. When r < r*, the ∆G increase with the increase in radius, so the ice core will not grow but melt; when r > r*, the ∆G will decrease with the increase in the radius, and it is only under this condition that the ice core can grow.
When ∂(∆G)/ ∂r = 0, ∆G can be taken as the minimum value, where r* is the critical radius. Take the derivative of r* in Equations (4)- (8), and then set ∂(∆G)/ ∂r = 0, the result can be expressed as: Therefore, the critical nucleation energy of crystal nucleus can be expressed as follows: It can be seen from Equations (9) and (10) that the threshold size and threshold nucleation energy are related to the surface tension of the ice crystal interface, solution equilibrium temperature, supercooling degree, and other factors.

Heterogeneous Nucleation
Heterogeneous nucleation usually occurs on the surface of suspended impurities or the surface of pipes, wires, solar panels, containers, and roads, etc. In most cases, the supercooling value of the impurity solution around the heterogeneous wall surface is much smaller than that of the homogeneous nucleation. These differences are mainly due to the differences in the external disturbance between the two systems, such as some small particles or heterogeneous walls in contact with aqueous solutions. In ice nucleation, these heterogeneous wall surfaces play an important role in the process of catalytic nucleation and provide the necessary surface for ice nucleation. The ratio of area to volume of crystal nucleus decreases, and some heterogeneous interfaces and crystal nucleus could replace the interface between the liquid and solid phases of crystal nucleus. Figure 1 shows a model of heterogeneous nucleation, where θ is the contact angle. Heterogeneous nucleation is determined by the contact state and surface tension of the wall, ice crystal, and aqueous solution.
differences in the external disturbance between the two systems, such as some small particles or heterogeneous walls in contact with aqueous solutions. In ice nucleation, these heterogeneous wall surfaces play an important role in the process of catalytic nucleation and provide the necessary surface for ice nucleation. The ratio of area to volume of crystal nucleus decreases, and some heterogeneous interfaces and crystal nucleus could replace the interface between the liquid and solid phases of crystal nucleus. Figure 1 shows a model of heterogeneous nucleation, where θ is the contact angle. Heterogeneous nucleation is determined by the contact state and surface tension of the wall, ice crystal, and aqueous solution. When heterogeneous nucleation occurs, the volume of the crystal nucleus (Vice) is expressed as: The contact areas (Aiw) of the crystal nucleus with the matrix and the solution are as follows: The contact area (Ail) between the crystal nucleus and solution is: (13) Before the nucleation, the area Alw is part of the interface between the solution and the heterogeneous wall. When the ice nucleation occurs, this part of the interface is replaced by the interface between the ice and the solution, so it can be expressed as: When the droplet is in equilibrium on the solid surface, the surface tension is: When heterogeneous nucleation occurs, the volume of the crystal nucleus (V ice ) is expressed as: The contact areas (A iw ) of the crystal nucleus with the matrix and the solution are as follows: The contact area (A il ) between the crystal nucleus and solution is: Before the nucleation, the area A lw is part of the interface between the solution and the heterogeneous wall. When the ice nucleation occurs, this part of the interface is replaced by the interface between the ice and the solution, so it can be expressed as: When the droplet is in equilibrium on the solid surface, the surface tension is: Change Equation (15) to: As shown in Figure 1, the following geometric relationships can be generated in the nucleation process: If the above equation is substituted into the calculation formula of the free energy of the crystal nucleus, it is as follows: The critical size of the crystal nucleus is related to interface energy and volume free energy, which can be written as follows: If the ratio of nucleation matrix radius to the critical radius of nucleation is R', it is as follows: Substituting Equations (21)-(23) into Equation (20) can be expressed as: ∆G ho in the formula: According to Equations (24) and (25), the threshold size and nucleation of heterogeneous nucleation are related to the surface tension of the ice crystal interface, solution equilibrium temperature, supercooling degree, contact angle, and other factors.
Comparing Equation (9) with Equation (15), it can be concluded that f (m, R ) is the main influencing factor causing the difference between heterogeneous and homogeneous nucleation. f (m, R ) is a function of θ and R , as shown in the following formula: When θ < 180 • , heterogeneous nucleation can occur, and the threshold nucleation decreases (i.e., f < 1). Thus, heterogeneous nucleation is more likely to occur during low subcooling, while homogeneous nucleation is possible only when the wetting angle is equal to 180 • .

Analysis of the Influencing Factors of "Ice Nucleation"
The contact morphology of water droplets can be changed significantly by altering the micromorphology of the substrate surface, consequently changing the value of the contact angle. In recent years, Bai et al. [24] also found that when the contact angle of a solid surface is less than the threshold value, ice crystals grow along the surface. In contrast, ice crystals grow off the surface when the contact angle of the solid surface is greater than this threshold value. In this paper, the contact angle and roughness are taken as the breakthrough points for characterization.
(1) Effect of contact angle on nucleation of ice crystal The ice crystal morphology on the cold surface is the result of the interaction between water molecules and the solid surface [24]. As the relationship between the contact angle and the direction of the ice crystals grows, as mentioned above, once the contact angle on the superhydrophobic material coated surface becomes larger, the Gibbs free energy will increase accordingly. Then, the liquid water droplets adhere to the coating surface at a certain contact angle, and they turn into solid ice crystals when the process of ice formation stops. Water freezing on the rough road surface is a heterogeneous nucleation process, while superhydrophobic ice suppression coating as a heterogeneous nucleation matrix provides conditions for ice nucleation. When superhydrophobic ice suppression coating is used as a nucleation matrix, the threshold size of the crystal nucleus can be ignored when compared with the size of the ice suppression coating (R > 1). In this case, Equation (26) can be simplified as: From Equation (28), the contact angle becomes the critical factor in determining the threshold nucleation difference between the ice inhibitor coating and the aqueous solution. Figure 2 shows the influence of the contact angle on threshold nuclear generation. The x-axis is represented by m = cosθ, and the y-axis is indicated by f (m), which is a function of m. When θ increased from 120 • to 180 • , m decreased from −0.5 to −1. At this point, f (m) presented a monotonically growing trend, and the maximum value is 1 when m = −1. Figure 3 showed that with the increase in the contact angle, water droplets reached hydrophobic and superhydrophobic status on the heterogeneous wall surface. At this point, the threshold nuclear generation is large, and consequently the heterogeneous nucleation does not occur easily in the wall. When the contact angle is in the range of 120-180 • , the heterogeneous walls become the main factors influencing the nucleation. When the contact angle is at 180 • , the critical nucleation reaches its maximum, where heterogeneous nucleation is extremely difficult to occur.     (2) Effect of roughness on ice nucleation Traditionally, the microscopic shape of the nucleating matrix surface has not been considered in the process of nucleation. However, the surface roughness of the nucleating matrix could be another major factor affecting ice nucleation besides the contact angle. Compared with the nucleating substrate surface, the size of the crystal nucleus is negligible. For the crystal nucleus, there is a certain roughness on the smooth wall surface. Therefore, the influence of roughness on ice crystal nucleation should be considered. The zexpression of the influence factor of roughness is shown in Equation (29).
In the formula, Ara represents the actual surface area, and Apa represents the projected area of the surface.
After introducing the roughness factor, the relationship between the contact angle and surface roughness can be expressed as: (2) Effect of roughness on ice nucleation Traditionally, the microscopic shape of the nucleating matrix surface has not been considered in the process of nucleation. However, the surface roughness of the nucleating matrix could be another major factor affecting ice nucleation besides the contact angle. Compared with the nucleating substrate surface, the size of the crystal nucleus is negligible. For the crystal nucleus, there is a certain roughness on the smooth wall surface. Therefore, the influence of roughness on ice crystal nucleation should be considered. The z-expression of the influence factor of roughness is shown in Equation (29).
In the formula, A ra represents the actual surface area, and A pa represents the projected area of the surface.
After introducing the roughness factor, the relationship between the contact angle and surface roughness can be expressed as: The relationship between the contact angle θ with roughness and contact angle θ without roughness can be expressed as: The surface roughness affects the nucleation of the solution by changing the droplet contact angle on the wall. Figure 4 revealed the relationship between the surface roughness and the contact angle. When the hydrophobic surface has a θ > 90 • , and has an increase in surface roughness, the contact angle increased, which further increased the hydrophobicity; if the hydrophilic surface has a θ < 90 • , with increased surface roughness, the contact angle decreases, and consequently its hydrophilicity increases. Moreover, the greater the surface roughness, the greater the influence of the docking contact will be. Especially for the wall surface with strong hydrophilic and hydrophobic materials, the impact of surface roughness on wall nucleation is far greater than that of the surface energy on wall nucleation. For the superhydrophobic contact surface, surface roughness makes it more difficult for wall nucleation to occur. In the actual road ice suppression, the impact of wall roughness on the contact is greater than that of surface tension, and the impact on wall nucleation is also greater.
According to the analysis results of roughness and contact angle, the ultimate determinant of ice crystal nucleation depends on the size of the contact angle. The greater the contact angle on the surface of the superhydrophobic and ice-suppressing materials, the more pronounced the ice-suppressing effect is. crease in surface roughness, the contact angle increased, which further increased the hydrophobicity; if the hydrophilic surface has a θ < 90°, with increased surface roughness, the contact angle decreases, and consequently its hydrophilicity increases. Moreover, the greater the surface roughness, the greater the influence of the docking contact will be. Especially for the wall surface with strong hydrophilic and hydrophobic materials, the impact of surface roughness on wall nucleation is far greater than that of the surface energy on wall nucleation. For the superhydrophobic contact surface, surface roughness makes it more difficult for wall nucleation to occur. In the actual road ice suppression, the impact of wall roughness on the contact is greater than that of surface tension, and the impact on wall nucleation is also greater.
According to the analysis results of roughness and contact angle, the ultimate determinant of ice crystal nucleation depends on the size of the contact angle. The greater the contact angle on the surface of the superhydrophobic and ice-suppressing materials, the more pronounced the ice-suppressing effect is.

Preparation of Superhydrophobic Materials
The superhydrophobic coating is usually prepared with a material that has a good inhibition performance [10][11][12]. In this paper, superhydrophobic materials were prepared by the hydrothermal method. The reaction between nano-sized titanium dioxide (TiO2) and stearic acid was synthesized in the media of anhydrous ethanol. The titanium dioxide has a purity of 99.8% and a particle size of 60 nm, which has a rutile structure and was supplied from McLean Co., Ltd., (Taipei, Taiwan) The stearic acid and anhydrous ethanol

Preparation of Superhydrophobic Materials
The superhydrophobic coating is usually prepared with a material that has a good inhibition performance [10][11][12]. In this paper, superhydrophobic materials were prepared by the hydrothermal method. The reaction between nano-sized titanium dioxide (TiO 2 ) and stearic acid was synthesized in the media of anhydrous ethanol. The titanium dioxide has a purity of 99.8% and a particle size of 60 nm, which has a rutile structure and was supplied from McLean Co., Ltd., (Taipei, Taiwan) The stearic acid and anhydrous ethanol are both analytical reagents, which are supplied from Shanghai Lingfeng Chemical Reagent Co., Ltd., (Shanghai, China), and Yaoxin Chemical Co., Ltd., (Mildenhall, Bury Saint Edmunds, UK), respectively. To prepare TiO 2 -octadecanoic acid composite material (noted as TiO 2 -OA sample), nano TiO 2 powder was first dispersed into pure ethanol by sonication for 1 h. Then, octadecanoic acid (C 18 H 36 O 2 , 0.9408 g/cm 3 ) was added to the previous solution. The mixture was homogenized by sonication for 3 h. The final solution was stirred using a DF-101S magnetic stirrer (Zhengzhou Ketai Experimental Equipment Co., Ltd., (Zhengzhou, China)) at 70 • C for 6~12 h. Afterward, an extra 60 mL ethanol was added at 25 • C, and the solution was centrifuged via a piece of TGL-16A centrifuge equipment (Hunan Xiangli Instrument Equipment Co., Ltd., (Hunan, China)) for 20 min to remove the top supernatant. Finally, the TiO 2 -OA sample was obtained after drying at 120 • C for 1 h, as shown in Figure 5. The obtained superhydrophobic materials were then sprayed on asphalt pavement or concrete pavement by manual brushing or mechanical spreading.
Co., Ltd., (Zhengzhou, China)) at 70 °C for 6 ~ 12 h. Afterward, an extra 60 mL ethanol was added at 25 °C, and the solution was centrifuged via a piece of TGL-16A centrifuge equipment (Hunan Xiangli Instrument Equipment Co., Ltd., (Hunan, China)) for 20 min to remove the top supernatant. Finally, the TiO2-OA sample was obtained after drying at 120 °C for 1 h, as shown in Figure 5. The obtained superhydrophobic materials were then sprayed on asphalt pavement or concrete pavement by manual brushing or mechanical spreading.

Image Characterization Using the Scanning Electron Microscope
The surface morphology and microstructure of the prepared superhydrophobic materials were observed and characterized by a JSM-7600F Scanning Electron Microscope (JEOL, Osaka, Japan). The superhydrophobic material coated samples are fixed on the SEM sample table using a conductive adhesive. The observation surface is facing up and firmly pasted. Once the conductive adhesive was air-dried, gold was sprayed on the sample surface to ensure good conductivity and a clear surface morphology image of the TiO2-OA material [25]. Its surface morphology and microstructure were observed up to ×50,000 magnification scales.

Contact Angle Test
The TiO2-OA coated pavement was placed at room temperature and glued on the coverlid. A syringe was used to drop the water droplet to the top surface of the sample. The volume of the droplet is 5 μL. Then the static contact angle of the water droplet on

Image Characterization Using the Scanning Electron Microscope
The surface morphology and microstructure of the prepared superhydrophobic materials were observed and characterized by a JSM-7600F Scanning Electron Microscope (JEOL, Osaka, Japan). The superhydrophobic material coated samples are fixed on the SEM sample table using a conductive adhesive. The observation surface is facing up and firmly pasted. Once the conductive adhesive was air-dried, gold was sprayed on the sample surface to ensure good conductivity and a clear surface morphology image of the TiO 2 -OA material [25]. Its surface morphology and microstructure were observed up to ×50,000 magnification scales.

Contact Angle Test
The TiO 2 -OA coated pavement was placed at room temperature and glued on the coverlid. A syringe was used to drop the water droplet to the top surface of the sample. The volume of the droplet is 5 µL. Then the static contact angle of the water droplet on the sample surface is measured by the water contact angle tester (Dataphysics OCA40Micro, Berlin, Germany). The contact angle from five different locations on the TiO 2 -OA surface is obtained and the average value is calculated.

The Water Droplet Icing Performance Test
The drip icing test on the TiO 2 -OA surface was tested under a constant RH of 70% but with varying temperatures, i.e., −2 • C, −5 • C, or 15 • C, via a programmable temperature control chamber (Shanghai Fengshi Experimental Instrument Co., Ltd., Shanghai, China). A set of samples of the same size were prepared with and without the superhydrophobic coating material; the samples were then placed into the environmentally controlled chamber with the desired temperature and RH. One mL of water was dropped on each sample surface. The time required for the water droplet to freeze was recorded. To avoid an operation error, the experiments were conducted with ten replicates, and their average value was taken.

Road Skid-Resisting Capability Evaluation
The anti-icing coating layer may affect the surface morphology of an asphalt pavement. Therefore, the friction on the road surface may be influenced. The skid resistance of the road pavement is critical to road safety [26]. The skid resistance of the road pavement with the applied superhydrophobic coating layer needs to be examined to ensure traffic safety. In this section, the application of the proposed coating layer was conducted on prepared asphalt concrete and cement concrete samples. The effects of using a superhydrophobic anti-icing layer on the road skid-resisting performance of the asphalt pavement were evaluated [27,28]. The test setup is shown in Figure 6, which used the pendulum instrument method, based on the China standards of JTJ073. 2-2001. ture control chamber (Shanghai Fengshi Experimental Instrument Co., Ltd., Shanghai, China). A set of samples of the same size were prepared with and without the superhydrophobic coating material; the samples were then placed into the environmentally controlled chamber with the desired temperature and RH. One mL of water was dropped on each sample surface. The time required for the water droplet to freeze was recorded. To avoid an operation error, the experiments were conducted with ten replicates, and their average value was taken.

Road Skid-Resisting Capability Evaluation
The anti-icing coating layer may affect the surface morphology of an asphalt pavement. Therefore, the friction on the road surface may be influenced. The skid resistance of the road pavement is critical to road safety [26]. The skid resistance of the road pavement with the applied superhydrophobic coating layer needs to be examined to ensure traffic safety. In this section, the application of the proposed coating layer was conducted on prepared asphalt concrete and cement concrete samples. The effects of using a superhydrophobic anti-icing layer on the road skid-resisting performance of the asphalt pavement were evaluated [27,28]. The test setup is shown in Figure 6, which used the pendulum instrument method, based on the China standards of JTJ073.2-2001.  Figure 7 revealed the micromorphology of TiO2-OA material characterized by ESEM [29,30]. During the hydrothermal reaction, nano titanium dioxide reacted with stearic acid, and the formed TiO2-OA product has a rough morphology and abundant mastoids in the nanostructure. The micro-nanometer mastoid structure is similar to the surface of the lotus leaf. Therefore, based on the SEM observation on the synthesized TiO2-OA material, it was characterized to have a low-surface energy.  [29,30]. During the hydrothermal reaction, nano titanium dioxide reacted with stearic acid, and the formed TiO 2 -OA product has a rough morphology and abundant mastoids in the nanostructure. The micro-nanometer mastoid structure is similar to the surface of the lotus leaf. Therefore, based on the SEM observation on the synthesized TiO 2 -OA material, it was characterized to have a low-surface energy.

Contact Angle Test
The contact angle test of water on the surface of the TiO2-OA material was captured at different positions on the coated pavement sample. When the contact angle of a water droplet on a surface is greater than 150°, the surface material can be defined as a superhy-

Contact Angle Test
The contact angle test of water on the surface of the TiO 2 -OA material was captured at different positions on the coated pavement sample. When the contact angle of a water droplet on a surface is greater than 150 • , the surface material can be defined as a superhydrophobic material. The contact angle results are illustrated in Figure 8. It was revealed that the contact angle of a water droplet on the prepared TiO 2 -OA product was 153 • (>150 • ), which confirmed that this coating material has a superhydrophobic property.

Contact Angle Test
The contact angle test of water on the surface of the TiO2-OA material was captured at different positions on the coated pavement sample. When the contact angle of a water droplet on a surface is greater than 150°, the surface material can be defined as a superhydrophobic material. The contact angle results are illustrated in Figure 8. It was revealed that the contact angle of a water droplet on the prepared TiO2-OA product was 153° (>150°), which confirmed that this coating material has a superhydrophobic property.

Water Droplet Freezing Test
The water droplet freezing test results were compared based on the recorded time when freezing was initiated. It was observed after 25 minutes sitting at −2 °C that the water droplets on the surface of ordinary (uncoated) pavement specimens started to freeze. In contrast, the droplet on the surface of the superhydrophobic TiO2-OA coating remained in the form of liquid at the same time. Until over 40 mins, the water droplets on the surface of the superhydrophobic layer started to freeze. A significant difference was found in terms of the required freezing time of the water droplets between the superhydrophobic surface and the ordinary specimens when the freezing temperature was set as −2 °C. The superhydrophobic coating surface largely extended the freezing time of the water droplets at this temperature. When the freezing temperature was adjusted to −5 °C, water droplets on the surface of ordinary specimens started to freeze after 10 min, while those on the superhydrophobic material coated surface started to freeze after 16 min. The results suggested that the superhydrophobic coating surface could delay ice crystal nucleation even when the temperature decreased from −2 °C to −5 °C. However, with the

Water Droplet Freezing Test
The water droplet freezing test results were compared based on the recorded time when freezing was initiated. It was observed after 25 min sitting at −2 • C that the water droplets on the surface of ordinary (uncoated) pavement specimens started to freeze. In contrast, the droplet on the surface of the superhydrophobic TiO 2 -OA coating remained in the form of liquid at the same time. Until over 40 mins, the water droplets on the surface of the superhydrophobic layer started to freeze. A significant difference was found in terms of the required freezing time of the water droplets between the superhydrophobic surface and the ordinary specimens when the freezing temperature was set as −2 • C. The superhydrophobic coating surface largely extended the freezing time of the water droplets at this temperature. When the freezing temperature was adjusted to −5 • C, water droplets on the surface of ordinary specimens started to freeze after 10 min, while those on the superhydrophobic material coated surface started to freeze after 16 min. The results suggested that the superhydrophobic coating surface could delay ice crystal nucleation even when the temperature decreased from −2 • C to −5 • C. However, with the further decrease in temperature, the suppression effect of the ice formation by superhydrophobic coating was weakened, since the time required for water droplet freezing was shortened. The state of a water droplet was examined at a certain time of the uncoated ( Figure 9A,C) and the superhydrophobic coated specimen ( Figure 9B,D), respectively. Under the condition of −2 • C, the water droplets on the surface of the uncoated specimen started to freeze, and a few sporadic ice crystals appeared at 25 mins ( Figure 9A). In contrast, the water droplet was still in liquid form on the superhydrophobic material coated surface until ice crystals appeared at 40 min, as shown in Figure 9B. Figure 9C,D showed the states of droplets at 10 min and 16 min under the condition of −5 • C, respectively. It is important to note that the water droplet maintained a spherical shape in Figure 9B,D, indicating that the suitable superhydrophobic property of TiO 2 -OA material can be found at under −2 • C and −5 • C. started to freeze, and a few sporadic ice crystals appeared at 25 mins ( Figure 9A). In contrast, the water droplet was still in liquid form on the superhydrophobic material coated surface until ice crystals appeared at 40 min, as shown in Figure 9B. Figure 9C,D showed the states of droplets at 10 min and 16 min under the condition of −5 °C, respectively. It is important to note that the water droplet maintained a spherical shape in Figure 9B,D, indicating that the suitable superhydrophobic property of TiO2-OA material can be found at under −2 °C and −5 °C. The superhydrophobic material coated on the substrate has a noticeable ice suppression effect in freezing weather. The superhydrophobic materials can delay the nucleation of ice crystals in the early stage of ice formation. The growth of ice crystals is inhibited in the middle stage of ice formation. Furthermore, considering the formed spherical shape of ice, it is expected that the superhydrophobic materials can isolate the ice layer from the road surface in the late stage of ice formation. This property reduces the adhesion between the ice layer and the road surface so that the ice layer can be easily removed. The superhydrophobic material coated on the substrate has a noticeable ice suppression effect in freezing weather. The superhydrophobic materials can delay the nucleation of ice crystals in the early stage of ice formation. The growth of ice crystals is inhibited in the middle stage of ice formation. Furthermore, considering the formed spherical shape of ice, it is expected that the superhydrophobic materials can isolate the ice layer from the road surface in the late stage of ice formation. This property reduces the adhesion between the ice layer and the road surface so that the ice layer can be easily removed.

Road Skid Resistance
Based on the China standards of JTJ073.2-2001, if the BPN is between 37-42, the skid-resisting performance is considered adequate. If the BPN is higher than 42, the skidresisting performance of the asphalt pavement is considered to be excellent. The laboratory testing results of the prepared samples are summarized in Table 1.

Conclusions
Superhydrophobic deicing coating materials for pavement surface applications can effectively delay the freezing of static water droplets, which is due to its unique micro-nano structure and excellent water-repellent properties. The major conclusions drawn from the theory of the nucleation mechanism of ice formation and the experimental works in this study include: (1) The influence of the contact angle and surface roughness of the superhydrophobic coating on the nucleation process of ice crystals and its law is theoretically discussed. It is concluded that the Gibbs free energy required for ice crystal nucleation increases with the increase in the contact angle, i.e., the larger the contact angle of the road superhydrophobic materials, the greater the critical nuclear generation is, resulting in less heterogeneous nucleation and a better surface icing inhibition effect. (2) The superhydrophobic coating surface has the structure of micro-nano mastoids. The super-large contact angle will "hide" the air in the gap of the rough structure, reducing the contact surface between water droplets and the superhydrophobic coated surface, resulting in good superhydrophobic characteristics. (3) The superhydrophobic material produces a good effect on inhibiting icing in the early stage of ice crystal nucleation. It is verified that the pavement surface coated with the superhydrophobic material can effectively delay the crystallization of water droplets. (4) The samples with a coated anti-icing layer could lead to a higher BPN. The improved BPN suggested that, in addition to the improvement of the anti-icing performance, the application of the proposed superhydrophobic anti-icing coating layer itself can also improve the skid-resisting performance of the pavement. Institutional Review Board Statement: Not applicable.
Informed Consent Statement: Not applicable.

Data Availability Statement:
The raw experimental data reported in this paper could be requested from the corresponding author.