1. Introduction

Black ice, also called clear ice, is a thin coating of glazed ice, on street surfaces. The ice itself is not black but is visually transparent, allowing one to see the black road below. The low level of visible ice pellets, snow or sleet that typically surrounds black ice means that these patches of ice are often invisible to drivers or people walking on them.

Therefore, there is a risk of slipping and subsequent accidents caused by unexpected loss of traction. Black ice can form even when the ambient temperature is several degrees above the freezing point of water (0 °C), when the air suddenly warms after a prolonged cold spell that has left road surfaces well below the freezing point. Black ice can also form on foggy roads in the early morning in winter. Bridges and overpasses are another risk point where temperatures can drop below zero because they do not absorb much heat from the ground and have areas that lose more heat into the air[1].

Many scientists have developed several methods to remove black ice. One of the cheapest and best- known agents is sodium chloride, but it promotes metal corrosion and damaging concrete. While applying sodium chloride to roads may ensure safe travel in the short term, it can also have adverse long-term environmental impacts[2-4].

Mechanical ice removal is a method that uses mechanical force such as snow blowers, ice makers, and abrasives, but residual ice can remain on the surface[5]. It is inefficient and uses considerable human resources. Furthermore, mechanical removal systems can damage the road surface and road structure.

Another way to remove black ice is to use a heated road sheet wire system. The heated systems have been applied in various applications, including sidewalks, roads, ramps, bridges, access ramps, disabled parking spaces, and runways. This has been proposed over the past decades as an alternative to salt spreading. Popular ice melting systems use electric heating wires buried beneath the pavement[6]. However, this has huge installation costs, maintenance costs, and power consumption.

Recently, the use of latent heat from phase change materials in heating systems has attracted recent attention. When the temperature is lower than the phase change temperature, a large amount of latent heat is released along with the phase change process[7]. Cement-asphalt composite concrete capable of controlling phase change temperature was manufactured by injecting cement paste containing phase change materials into porous asphalt concrete using a self-leveling method[8]. Overall, phase change materials have promising applications in asphalt pavement. On the other hand, water leakage and workability are reduced when mixed with hot mixed asphalt.

The number of traffic accidents in Korea between 2011 and 2020 was 2,209,962[9]. Among them, the number of traffic accidents when the road surface was ‘iced/frosted’ was 16,280, accounting for approximately 0.7% of the total accidents. Iced roads are approximately 14 times more slippery than regular roads and approximately six times more slippery than snowy roads, which exceeds the braking ability of the vehicle. Hence, real-time monitoring and guidance through VMS (electronic road sign) are necessary to convey the status of road ice to drivers, but real-time monitoring of the section where ice is expected to occur, and the installation of VMS throughout the entire section are virtually impossible. In the field of road traffic, demand for cutting-edge technology that can improve driving safety by visualizing road information in real-time through light or color changes in response to stimuli from the climatic environment is increasing. To solve them, research is needed to enable the real-time visualization of road icing caused by rapidly lowering temperatures and snow in winter by displaying black ice and slip risk areas using temperature stimulus response performance.

Smart responsive materials change their properties in response to external stimuli, such as temperature, light, pressure, pH, electric and magnetic fields, or exposure to moisture and chemical compounds [10-12]. Thermochromic materials change color with temperature variations and typically include materials made of leuco dye systems, liquid crystals, quantum dots, gold or silver nanoparticles, or dye-polymer mixtures[13-17]. The most commonly used types are leuco dyes and liquid crystals. The commercial applicability of liquid crystals for thermochromic products is greatly hampered by their weak color change effect and high cost[18]. On the other hand, leuco dyes are more cost-effective than liquid crystals, and microencapsulation of these systems is typically performed using oil-in-water (o/w) emulsions through cross-linking of the hydrophilic shell material by the condensation of urea-formaldehyde solution. Reversible thermochromic mechanisms include ring-opening reactions in organic molecules[19-22].

This study aims to develop a smart freezing moisture visualization mixtures as a basic material that can visually display road surface conditions in real time in response to temperature and humidity among major road climate environmental factors such as solar radiation, temperature, and humidity.

2. Materials and methods

2.1. Materials

Thermochromic microcapsule in powder form were used as thermochromic materials. The reversible thermochromic materials were selected, which lose their color when they reach a set temperature and return to their original color when falls below the set temperature. Red was selected as the color. Reversible hydrochromic material is a special wet exposure binder that alternates from white to transparent when wet and returns to its original white when dried. Hydrochromic material is white and opaque when dry and can be used to hide the image underneath. The area touched by water changes from a white, opaque state to a transparent state, revealing the background color. Four types of thermochromic microcapsules were used: R0 (red 0 °C) was consist of the mixtures of 6'-(diethylamino)-1',3'-dimethylfluoran and 6'-(diethylamino)- 1',2'-benzofluoran, methyl stearate and ethyl stearate. R5 (red 5 °C) was 6'-(diethylamino)-1',3'-dimethylfluoran and 6'-(diethylamino)-1',2'- benzofluoran, methyl stearate and ethyl stearate. Hydrochromic material (H) was the mixtures of acrylic resin paste and silica (white paste). All materials were purchased from Nano I&C company in Republic of Korea and used received.

2.2. Preparation of thermochromic coating film

Among the thermosensitive materials, materials that change at 0 °C and materials that change at 5 °C were selected, and an appropriate mixing weight ratio was selected. The sample were dried at room temperature for two days. The thickness of freezing visualization materials coated on the overhead projector film was 200 μm. A sample coated with water once with a brush was selected for the moisture experiment. When brushed with water at room temperature, the white-coated surface became transparent and showed the color of the thermosensitive material. The water on the specimen was allowed to soak into the specimen for one minute, and the color change was observed under freezing conditions. The sample names are as follows: R5 (white powder), Red, 5 °C; R0 (white powder), Red, 0 °C; Hydrochromic material (white paste), H. R5R0 is a mixture of R5 and R0. The sample name H is a hydrochromic material and a mixture binder. When mixing the two substances, a 1:9 ratio with H was tested to make the 0 °C color stand out. In addition, a 9:1 ratio, which has a high 5 °C color ratio, was assessed. The color became lighter as the amount of hydrochromic material was increased.

2.3. Characterization of materials

The reflectance spectra of the thermochromic materials were obtained using a UV–vis spectrophotometer (V-550, JASCO) over the wavelength range of 190~900 nm. Differential scanning calorimetry (DSC 25, Discovery) was carried out at a heating rate of 1 °C/min from -70 to 400 °C under N₂ gas. The tensile strength was measured using universal testing machine. Tensile stress rate was 10 mm/min and the thickness of freezing visualization materials coated on the paper was 400 μm. The sample size was 110 mm (L) × 20 mm (W) × 0.5 mm (T).

2.4. Surface morphological characterization

The surface morphology of the thermochromic materials and hydrochromic material was analyzed by field emission scanning electron microscope (JSM-IT800, Aztec Live Ultim Max). Before the analysis, the samples were sputter-coated with a thin gold and imaged at different magnifications.

2.5. Determination of colorimetric coordinates

The colorimetric coordinates (L^*, a^*, and b^*) of the thermochromic specimens were measured according to the CIELAB coordinate system using a CM-5 spectrometer provided by Konica Minolta. The CIELAB color system represents the quantitative relationship of color on three axes. L is defined on the vertical axis with values from 0 (pure black) to 100 (pure white), with the a^* coordinates representing red and green. The minimum value of a^* represents an intense green color, and the maximum value of a^* represents an intense red color. Similary, the b^* coordinates represent blue and yellow. The maximum value of b^* indicates intense yellow color, and the minimum value of b^* indicates an intense blue saturation. The total color difference (ΔE, eq. 1) was determined using observations at different periods to determine the color stability: Where lightness (ΔL, eq. 2), red chromaticity (Δa, eq. 3), and yellow chromaticity (Δb, eq. 4).

3. Results and discussion

3.1. The color of freezing moisture visualization mixture

Organic thermochromic materials are commonly used in various commercial applications such as smart packaging, security printing, textile coloring, toys, and marketing, owing to their low cost and low toxicity. They are also known as leuco dyes and are typically supplied as a three-component system consisting of a colorant, developer, and cosolvent. These systems are associated with color loss caused by the destruction of the dye-developer complex, which results in a color change when the cosolvent dissolves. Color developers are electron-donating compounds such as spirolactone, fluorane or spiropyran, and developers are electron acceptor (proton donor) compounds such as bisphenol A, alkyl gallates, phenols, hydroxybenzoates and hydroxycoumarins. The co-solvent is usually a long-chain alkyl alcohol, ester, or acid[23].

The road icing caused by rapidly dropping temperatures, moisture and snow in winter has highlights the need for a temperature-sensitive material visualize black ice in real time. Table 1 lists the color changes of frozen moisture visualization materials according to temperature changes. At room temperature, the raw material had an almost light unique color, but under freezing conditions, the vivid colors unique to the raw material, such as dark red for R0H, were observed. When the same color series was mixed at a mass ratio of 1:9, R5R0H appeared dark red. When the same color series was mixed at a mass ratio of 9:1, R5R0H appeared dark red. The higher ingredient content in terms of mass ratio, the higher its contribution to determining the final color. In the case of R5R0H, a more vivid red color was obtained at the 9:1 ratio than at the 1:9 ratio. When mixed alone or with colors of the same series, a bright and unique color was observed instead of a cloudy color. Since color is an intensive property of physical properties of a matter, there was no change in color when colors of the same series were mixed.

3.2. Intrinsic surface morphology and characteristics of the freezing visualization materials

Figure 1 presents the surface morphology of the freezing moisture visualization materials at ambient temperature. The particles sizes of the microcapsules varied from several nanometers to a few micrometers and exhibited a relatively uniform size distribution. The maximum sizes of R5 and R0 were 3.9 μm and 4.5 μm, respectively. R5 showed an even particle size distribution compared to R0. The shape of the microcapsules was maintained when mixed with the hydrochromic material. Figure 2 presents the ultraviolet-visible light reflection curve of the thermosensitive material at room temperature. When used alone, the reflectance was 58.5% for R5 and 66.9% for R0 at 526 nm in the visible light region, and when it was a mixture with hydrochromic material added, R5R0H was 82.9%. Since R5R0H is a mixture with polymer H, its absorbance is lower than that of pure substances with R5 and R0, so its reflectance increases.

The reflectance of R5 and R0 around 411 nm was 62.5% and 73.7%, respectively, and no peak was observed in the case of the hydrochromic mixture. The reflectance of R5, R0, and R5R0H in the UV region around 320 nm was 12.4, 14.3, and 37.5%, respectively. The hydrochromic material showed 42.3% reflection at 233 nm. The reflectance increased more significantly when it was a mixture than when used as a pure substance.

In order to measure the durability of the film, the reflectance of coating materials was measured after 120 hours under conditions below 0 °C. No significant results were obtained for reflectance under the experimental conditions, and the experimental time will be extended in the future to track the results.

The results of the tensile strength test by the content of the freezing visualization material are 9.91 (R0H18), 9.59 (R0H110), 7.21 (R5R0H19), 8.11 (R5R0H91), and 8.65 (H). In the case of pure substances, it increased by 11.0 ~ 14.5% compared to H, and in the case of mixtures, it decreased by 6.2~16.6%. This shows that the adhesion to polymer substances is evenly distributed and high in the case of pure substances, but when two pure substances are mixed and used, the applicability with polymer substances is not as good as that of pure substances, so the tensile strength decreases. This is because the composition ratio of the microcapsule has an effect. Adhesion of the mixtures can be improved by adjusting the composition ratio and mixing time of microcapsules.

3.3. Thermochromic performance of freezing visualization materials.

Table 2 lists the color change with time when the red material freezes. After one minute of freezing conditions, the R0H sample at a 1:8 ratio showed more water spreading than at a 1:10 ratio. The spreading phenomenon decreased as the amount of the hydrochromic material increased. When combined with moisture under freezing conditions, the color of the frozen area was a darker red than the background color, which the naked eye could the distinguish. The areas where black ice occurred were sufficiently distinguishable by the intensity of the color change. The time for the background color to completely change for R0H at 1: 8 was seven minutes, and the water in the specimen was frozen after being placed in freezing conditions for approximately nine minutes.

In R5R0H at 9:1, water froze after one minute of freezing condition, and the reaction speed was 11 times faster than R0H at 1: 10. The time for the background color to change completely according to the mixing ratio was observed to be three minutes in the case of 1:9 and one minute in the case of 9:1.

Under dry freezing conditions, a clear color difference was observed in red, and in moisture freezing conditions, a darker red color was observed than under the dry freezing conditions. The color change response speed of R5R0H at 1:9 was 3.7 times faster when a freezing visualization material was used as a mixture than when used alone (R0H 1:10). This means that moisture freezes quickly. When moisture combined under the conditions for moisture freezing, the color of the frozen area was a darker red than the background color that had high visibility. The areas where black ice occurred were sufficiently distinguishable by the intensity of the color change. The color change speed was 3 times faster in R5R0H at 9:1 than in R5R0H at 1:9. The R5R0H (9:1) mixture among the specimens had the fastest dry-freezing and moisture-freezing times.

The color change performance of the reversible moisture-freezing visualization material could be evaluated in three aspects:

The discoloration time is expressed as △t. The color sensitivity is excellent, good, and mediocre when △t is < 20 s, 20 ≤ △t v 30s, and 30 ≤ △t < 40s, respectively. If △t is 40 s ≤, the color sensitivity is poor.

Table 3 lists the discoloration performance of freezing visualization materials. The discoloration time at room temperature after dry freezing conditions was 10 s for all samples, making it an excellent material. The discoloration time at room temperature after moisture freezing conditions was each 38 s and 37 s for R0H 1: 8 and 1:10 in the case of a single moisture-freezing visualization material. R5R0H (1:9), and R5R0H (1:9) showed 42 s, and 46 s, respectively.

In the case of moisture freezing conditions, the discoloration time was delayed maximum 4.6 times compared to the dry freezing conditions. After black ice formed, it was clearly distinguishable from dry freezing conditions to room temperature. Considering safety, research on the appropriate discoloration time under moisture-freezing conditions needs to be improved, but this is not expected to be a major problem on the road because the discoloration time was within 46 seconds at room temperature.

Table 4 lists the experimental results using a spectrophotometer.

Under dry freezing conditions, the color difference of the single substance R0H (1:8) were 29.39, and R0H (1:10) 40.18, showing a larger color difference at the 1:10 ratio. R0H (1:10) had 37% larger color difference than R0H (1:8). In the case of the same color family, R5R0H exhibited a color difference of 39.98 and 38.61, at a 1:9 and 9:1 ratio, respectively, showing a 4% larger color difference at a 1:9 ratio.

Under moisture freezing conditions, the color difference of the single substance R0H (1:8) were 60.57, and R0H (1:10) 60.65, showing a larger color difference at the 1:10 ratio. The color difference of red was similar. In the case of the same color family, R5R0H exhibited a color difference of 58.66 and 60.05, at a 1:9 and 9:1 ratio, respectively, showing a 2% larger color difference at a 9:1 ratio. No clear difference was noted. They had similar value.

When changing from dry freezing conditions to moisture freezing conditions, the color difference was 31.18 for R0H (1:8), and 20.47 for R0H (1:10), respectively. R0H (1:8) had a 52% % greater color difference. In the case of the same color family, R5R0H had a color difference of 18.68 and 21.44 at a 1:9 and 9:1 ratio, respectively, showing a 15% larger color difference at a 9:1 ratio. In the case of pure substances, the higher the concentration, the higher the color development, and in the case of mixtures, the color development increased as the amount of R5 increased. As the amount of color development increased, the absorbance increased and the reflectance decreased, so the color difference was large.

When changing from moisture conditions to moisture freezing conditions, the color difference was 47.17 for R0H (1:8), and 47.37 for R0H (1:10), respectively. They had a similar value. In the case of the same color family, R5R0H had a color difference of 46.03 and 46.02 at a 1:9 and 9:1 ratio, respectively, showing a similar value.

Figure 3 shows the reflectance of R5R0H using spectrophotometry. Measurement of the reflectance difference of R5R0H from 360 to 740 nm, revealed a reflectance difference in the range of 23.83~82.78 in the case of dry-freezing conditions. The range was 48.64~53.16 under moisture conditions and 10.21~76.61 under moisture-freezing conditions.

A comparison of the reflectance difference according to the difference in conditions at 530 nm revealed, a reflectance difference of approximately 38.43 when the moisture condition was changed to moisture- freezing conditions, and a reflectance difference of approximately 13.62 when changing from dry-freezing conditions to moisture-freezing conditions. A higher reflectance difference appeared when moisture was present than when it was not present, and the color became darker as it changed from dry freezing conditions to moisture freezing conditions. A clear color change occurred when moisture was included. It exhibits visible thermochromic and hydrochromic behavior.

R5R0H showed a reflectance of 48.79% at 400 nm when exposed to moisture, which decreased to 17.24% under moisture-freezing conditions. The minimum reflectance was observed at approximately 530 nm, 48.64% and 10.21% under moisture and moisture-freezing conditions, respectively. Uniform values were obtained at 630 nm, 53.00% and 76.61%, under moisture and moisture-freezing conditions, respectively.

As the wavelength increased, the reflectance also increased. The minimum reflectance was observed in the 495-570 nm wavelength range, which is the color range resulting from the red dye reacting with the developer and breaking the lactone ring. In this case, the freezing moisture visualization material was white at room temperature, but red under freezing conditions. This was confirmed through photographs of the coating film of the freezing visualization materials. An increase in reflectance was observed even above 700 nm.

In the methyl stearate system, these R5, R0, and R5R0H samples showed color reversibility below 0 °C. The frozen visualization material turned dark red due to the leuco system, which was attributed to intramolecular bonding between the electron acceptor moiety and the leuco lactone unit. It is colorless (proton acceptor) because the carbon atom at the center of the spiro molecule forms a bond with the carbon atom through a sp³ hybrid orbital to form a colorless lactone. When the frozen material comes in contact with a proton donor, the lactone ring opens and the sp³ hybrid orbital of the central carbon atom transforms into a sp² hybrid orbital. This factor causes the π-conjunction of the fluorane moiety to expand and the λ_max to shift to longer wavelengths[ 24].

As shown in Figure 4, the DSC thermogram contains endothermic peaks, indicating the melting of pure and freezing visualization materials. The melting temperature (T_m) and final melting temperature (T_f) obtained through DSC are important parameters for evaluating the temperature-responsive color change performance of the freezing visualization materials. T_f is the final melting temperature, i.e. the solid- liquid transformation temperature range of the freezing visualization material. For those materials, the endothermic peak melting point ranged from 4.50 °C to 9.68 °C, respectively (Table 5). When the temperature of the solvent was increased to T_m, the color changed to red. The melting point of R5 was predicted to be around 5 °C and that of R0 around 0 °C. However, in the case of R0, it was not observed at 0 °C, so it was shown that its temperature-sensitive role works differently from R5. R0 and R5 have different developer composition ratios, so R0 developed color at a higher temperature.

The initial and final melting temperatures for the materials were -17.41 and 17.08 °C, respectively. The peak melting point increased when a hydrochromic material was introduced at R0. The peak melting point decreased when R5 was introduced into R0H at a 9:1 mass ratio. The color change temperature of the freezing visualization material is directly related to the melting enthalpy (ΔH_m), as shown in Figure 4 and Table 5. The enthalpy of melting (ΔH_m) is an important parameters for evaluating the heat storage capability of the freezing visualization material during the phase change process. The ΔH_m of R5 and R0 were high at 78.16 and 87.85 J/ g, respectively. For 1 g of R5 or R0 to completely change from a white solid to a red liquid, it must absorb 78.16 or 87.85 J of heat. When hydrochromic material was added, the ΔH_m for R0H110 decreased by 85% compared to R0. Introducing R5 into R0H resulted in a 4% and 0.4% increase for R5R0H (1:9) and for R5R0H (9:1), respectively. R5R0H (9:1) is expected to react faster because it absorbs less heat than R5R0H (1:9).

The phase transition enthalpies of the four mixtures have relatively low values, which can also store and release a small amount of latent heat and are the low energy storage materials compared to R5 and R0. As the amount of R5 increases, the phase transition temperature decreases, so to lower the phase transition temperature to low temperatures, the experiment should be designed to increase the ratio of R5 material.

The T_m range of all specimens was expected to be around 0 °C or 5 °C, but in actual application, it was less than 10 °C except R5R0H91. This is because the melting point of all specimens are affected by the composition of the microcapsules. The R5R0H91 specimen had the lowest melting point. As the amount of R5 increases, the phase transition temperature decreases, so the ratio of R5 material is important to lower the phase transition temperature to low temperatures. R5R0H91 among the specimens was obtained at the most suitable black ice indicator concentration at temperature point of view and also had the highest visibility among the mixtures.

Although more experiments should be conducted to lower the T_m in the future, all specimens showed clear visibility, so they can be used as a color indicator of black ice.

4. Conclusion

This study evaluated the color change performance of a smart freezing visualization material comprised of thermochromic and hydrochromic materials that can respond to various road environments according to temperature and moisture. The following conclusions were obtained as a result of experiments conducted at different concentrations.

The developed freezing visualization materials were able to visually distinguish the risk of black ice by changing the color to darker when black ice occurs, increasing visibility. The samples had excellent temperature sensing performance and showed clear color conversion in response to moisture.