Multi Self-Healable UV Shielding Polyurethane/CeO2 Protective Coating: The Effect of Low-Molecular-Weight Polyols

We prepared a series of polyurethane (PU) coatings with defined contents using poly(tetramethylene oxide)glycol (PTMG) with two different molecular weights (i.e., Mn = 2000 and 650), as well as polydimethyl siloxane (PDMS) with a molecular weight of Mn 550. For every coating, maximum adhesive strength and excellent self-healing character (three times) were found using 6.775 mol% mixed with low-molecular-weight-based polyols (PU-11-3-3). Defined 1.0 wt% CeO2 was also used for the PU-11-3-3 coating (i.e., PU-11-3-3-CeO2) to obtain UV shielding properties. Both the in situ polymerization and blending processes were separately applied during the preparation of the PU-11-3-3-CeO2 coating dispersion. The in situ polymerization-based coating (i.e., PU-11-3-3-CeO2-P) showed similar self-healing properties. The PU-11-3-3-CeO2-P coating also showed excellent UV shielding in real outdoor exposure conditions.


Introduction
An organic coating deposited on a substrate surface can enhance the protection of that surface, which can ultimately increase the lifespan of that substrate and its structures. These coatings can provide long-term performance, giving them priority in industry [1][2][3][4]. Coating longevity also depends on the geographical location and indoor or outdoor application conditions. Any coatings used in outdoor atmospheric conditions where sunlight, high temperatures, humidity, and marine aerosols are common need extended protective properties [2,4]. Otherwise, the coating might be damaged within a short amount of time. In such conditions, if there is any micro-damage in the coating, the damage is rapid due to the easy passage of electrolytes (i.e., marine aerosols and water vapor). Scientists have previously tried to obtain extended coating protective properties by using self-healing, crosslinking, and multifunctional techniques in organic coatings [4][5][6][7].
It is common for a coating to be scratched or damaged during its service life. In such conditions, the coating loses its protective property rapidly. To repair the scratch or damage, the coating should heal properly. To heal the coating, a common trend is to give the coating a self-healing property. Self-healing

PU Dispersion Preparation
The basic polymerization steps were kept the same during the preparation of all dispersions based on previous reports. In all cases, the pre-polymer was prepared by charging PTMG, DMPA solution with NMP and H 12 MDI. The MEK (10.0 wt%) was added into the pre-polymer to reduce viscosity. Later, the pre-polymer was neutralized, dispersed, and chain extended by using the proper amount of TEA, H 2 O (70.0 wt%), and EDA (mixed with 10 ml water). The pre-polymer formation and reaction termination were monitored via FT-IR analysis.

PU-CeO 2 Dispersion Preparation via In Situ Polymerization
All of the polymerization steps were kept exactly the same, except for the addition of CeO 2 . The CeO 2 (1.0 wt%) was premixed with acetone (15 mL) and ultrasonicated for 30 min. The CeO 2 -acetone mixture was added into a premixed DMPA-NMP solution and stirred further for 30 min at 45 • C. This solution was added into a polyol and then charged with H 12 MDI.

PU-CeO 2 Dispersion Preparation via a Blending Process
The CeO 2 (1.0 wt%) was premixed with acetone (15 mL) and ultrasonicated for 30 min. Then, the CeO 2 -acetone mixture was added directly into a PU dispersion. The mixture was stirred for 1 h at 30 • C.

Substrate Coating
The coating on the substrate was fixed at 100 µm under wet conditions. The coating was made using a bar auto-coater. All prepared wet coatings were dried at ambient conditions for 48 h. To remove the used solvents completely, the earlier dried coatings were further dried for 12 h at 65-70 • C in an oven.

Self-Healing Test
The self-healing test was conducted at 50 • C. Each coating was cut by a sharp blade. To heal the coating, the tested coating was kept in an oven for a maximum of 12 h. The healing was checked via optical microscopy after every 1 h.

Characterization
Different techniques were applied to identify coating properties. In all cases, the value was the average of five testing results. The reaction monitoring was performed and the final polymer was identified via FT-IR (Impact 400D, Nicolet, Madison, WI, USA) analysis. The photographs of coatings were obtained with a Hitachi S-4200 (Tokyo, Japan) field emission scanning electron microscope (FESEM) at 5 kV. A Theta Optical tensiometer instrument (Attension, Helsinki, Finland) was used to obtain the water contact angle. The ASTM D4541 method was followed to evaluate adhesive strength using the pull-off test. XPS (ESCA 250 XPS, Thermo Scientific, East Grinstead, UK) analysis was done to characterize the surface of the coating. The ASTM D2240-75 specification was followed to evaluate the Shore A hardness test using the LX-A rubber Hardness Tester (Shanghai Liuling Instrument Company, Shanghai, China). An exposure test was used to check the UV shielding property of the coating and was based on our previous report [31].

Results and Discussion
A series of 17 formulated coatings were prepared based on different polyol contents (see Table 1). A defined CeO 2 content was also used in PU-11 coatings. The different steps of PU preparation were Polymers 2020, 12, 1947 5 of 11 monitored via FT-IR analysis (see Figure 1). The polyol, DMPA, and H 12 MDI formed the pre-polymer. Usually, the appearance and disappearance of a peak at 2170 cm −1 for the NCO group is considered a tool for monitoring the reaction [4]. During the pre-polymer formation stage, a strong peak appeared at 2170 cm −1 due to the unreacted NCO group. The peak remained during the neutralization and dispersion steps. During chain extension in the last stage, the EDA reacted with the NCO group and the peak disappeared within 1 h, confirming the completion of the reaction. All the prepared coatings showed identical PU functional groups, such as urethane (NHCOO) and urea (NHCOO), as the peaks appeared at 3400 cm −1 (for NH groups) and 1706 cm −1 (for CO groups). A peak at 816 cm −1 for the CH 3 group attached with Si was also recorded for coatings using PDMS and thus confirmed the presence of PDMS in PU. Moreover, for PU-CeO 2 coatings, a new peak appeared at 560 cm −1 , which also confirmed the presence of CeO 2 in the PU coatings. The peak at 560 cm −1 was similar for the coatings prepared via blending or polymerization. It was not possible to identify the interaction between the CeO 2 and PU groups clearly, as the PU-CeO 2 coatings were prepared via two processes. XPS analysis was used to evaluate the effect of the blending and polymerization preparation techniques on CeO 2 and PU interactions (see Figure 2). As expected, peaks were recorded at 100, 285, 400 and 530 eV, which reflected Si, C, N and O, respectively. Two additional peaks around 887 and 889 eV were also recorded for samples containing CeO 2 . Both peaks had slightly higher values in the blended sample. Moreover, slightly higher-intensity peaks were recorded for coatings made via the polymerization method. Most importantly, a minor new peak also appeared at 897 eV for the blended coatings. This peak and the pristine CeO 2 peak were almost at the same position. Thus, the CeO 2 had comparatively less interaction with PU in the blended coating. On the other hand, the CeO 2 particles entered the polymer chain by interacting with carboxyl groups in the polymerized coating. However, the CeO 2 was homogeneously distributed in both PU-CeO 2 coatings. Neither precipitation nor phase separation was seen in PU-CeO 2 dispersions (see Figure 2). To confirm the CeO 2 distribution in coatings, SEM analysis was done. As a small content of CeO 2 was used, the different synthesis process had no noticeable effect on CeO 2 distribution. CeO 2 was homogeneously distributed mostly in both cases. A typical SEM photograph is shown in Figure 2.
UV shielding property of the coating and was based on our previous report [31].

Results and Discussion
A series of 17 formulated coatings were prepared based on different polyol contents (see Table  1). A defined CeO2 content was also used in PU-11 coatings. The different steps of PU preparation were monitored via FT-IR analysis (see Figure 1). The polyol, DMPA, and H12MDI formed the pre-polymer. Usually, the appearance and disappearance of a peak at 2170 cm −1 for the NCO group is considered a tool for monitoring the reaction [4]. During the pre-polymer formation stage, a strong peak appeared at 2170 cm −1 due to the unreacted NCO group. The peak remained during the neutralization and dispersion steps. During chain extension in the last stage, the EDA reacted with the NCO group and the peak disappeared within 1 h, confirming the completion of the reaction. All the prepared coatings showed identical PU functional groups, such as urethane (NHCOO) and urea (NHCOO), as the peaks appeared at 3400 cm −1 (for NH groups) and 1706 cm −1 (for CO groups). A peak at 816 cm −1 for the CH3 group attached with Si was also recorded for coatings using PDMS and thus confirmed the presence of PDMS in PU. Moreover, for PU-CeO2 coatings, a new peak appeared at 560 cm −1 , which also confirmed the presence of CeO2 in the PU coatings. The peak at 560 cm −1 was similar for the coatings prepared via blending or polymerization. It was not possible to identify the interaction between the CeO2 and PU groups clearly, as the PU-CeO2 coatings were prepared via two processes. XPS analysis was used to evaluate the effect of the blending and polymerization preparation techniques on CeO2 and PU interactions (see Figure 2). As expected, peaks were recorded at 100, 285, 400 and 530 eV, which reflected Si, C, N and O, respectively. Two additional peaks around 887 and 889 eV were also recorded for samples containing CeO2. Both peaks had slightly higher values in the blended sample. Moreover, slightly higher-intensity peaks were recorded for coatings made via the polymerization method. Most importantly, a minor new peak also appeared at 897 eV for the blended coatings. This peak and the pristine CeO2 peak were almost at the same position. Thus, the CeO2 had comparatively less interaction with PU in the blended coating. On the other hand, the CeO2 particles entered the polymer chain by interacting with carboxyl groups in the polymerized coating. However, the CeO2 was homogeneously distributed in both PU-CeO2 coatings. Neither precipitation nor phase separation was seen in PU-CeO2 dispersions (see Figure 2). To confirm the CeO2 distribution in coatings, SEM analysis was done. As a small content of CeO2 was used, the different synthesis process had no noticeable effect on CeO2 distribution. CeO2 was homogeneously distributed mostly in both cases. A typical SEM photograph is shown in Figure 2.  Coating hardness is an important criterion. As hardness implies polymer chain stiffness, low hardness means polymer chain deformation and fast entanglements. The respective hardness values are summarized in Table 2. As the low-molecular-weight polyol content decreased, hardness also increased. Among all of the pristine PU coatings, the maximum hardness value was recorded for the PU-18 coating. The lower hardness values were recorded for the coatings that have a higher content of low molecular weight-based polyol. The PU-13 and PU-15 coatings had hardness values closer to that of the PU-18 coating. Only 2.25 and 4.525 mol% low-molecular-weight polyol replaced high-molecular-weight polyol, and the polymer deformation structure only slightly deviated in the PU-13 and PU-15 coatings. A moderately decreased value was recorded for the PU-11 coatings. The PU-CeO 2 coatings showed higher values than their parent PU coating. The CeO 2 filler interacted with the PU groups, and these polymers had greater stiffness, which reflected higher hardness. At the same time, the blended coatings' hardness values were less than those of the polymerized coatings. This also implies that the CeO 2 interacted strongly with PU groups due to proper mixing in the polymerization method. Coating hardness is an important criterion. As hardness implies polymer chain stiffness, low hardness means polymer chain deformation and fast entanglements. The respective hardness values are summarized in Table 2. As the low-molecular-weight polyol content decreased, hardness also increased. Among all of the pristine PU coatings, the maximum hardness value was recorded for the PU-18 coating. The lower hardness values were recorded for the coatings that have a higher content of low molecular weight-based polyol. The PU-13 and PU-15 coatings had hardness values closer to that of the PU-18 coating. Only 2.25 and 4.525 mol% low-molecular-weight polyol replaced high-molecular-weight polyol, and the polymer deformation structure only slightly deviated in the PU-13 and PU-15 coatings. A moderately decreased value was recorded for the PU-11 coatings. The PU-CeO2 coatings showed higher values than their parent PU coating. The CeO2 filler interacted with the PU groups, and these polymers had greater stiffness, which reflected higher hardness. At the same time, the blended coatings' hardness values were less than those of the polymerized coatings. This also implies that the CeO2 interacted strongly with PU groups due to proper mixing in the polymerization method.
All of the formulated coatings were considered for outdoor exposure applications. Any coating should be considered for exposed conditions with sufficient hydrophobicity. Otherwise, moisture, air, and water can easily pass through the coating to delaminate it, and it can ultimately corrode. The most widely used method to evaluate coating hydrophobicity is the water contact angle test. The higher the contact angle value, the higher the hydrophobicity of the coating. The contact angle value highly depends on the polyol's molecular weight and contents. Polyester-based PU showed less hydrophobicity, whereas polyether-based PU showed higher hydrophobicity. All of the recorded water contact angle values for the tested coatings are summarized in Table 2. As expected, the PU coating using a higher content of lower-molecular-weight PTMG had a smaller water contact angle. The water contact angle value continuously decreased with increasing lower-molecular-weight PTMG. However, the water contact angle value significantly increased with the inclusion of PDMS, and the value continuously increased with increasing PDMS content. The water contact angle value further increased with CeO2 content. The polymerized coatings also showed higher values compared to the blended coatings. Coatings with well-mixed CeO2 had higher stiffness and also resisted water penetration; ultimately, the contact angle value increased. The highest water contact angle value was recorded for the PU-11-3-3-CeO2-P coating, which was prepared with two PTMG polyols, a PDMS-polyol, and CeO2. Further, polymerization was applied during CeO2 use.
All of the formulated coatings were considered for outdoor exposure applications. Any coating should be considered for exposed conditions with sufficient hydrophobicity. Otherwise, moisture, air, and water can easily pass through the coating to delaminate it, and it can ultimately corrode. The most widely used method to evaluate coating hydrophobicity is the water contact angle test. The higher the contact angle value, the higher the hydrophobicity of the coating. The contact angle value highly depends on the polyol's molecular weight and contents. Polyester-based PU showed less hydrophobicity, whereas polyether-based PU showed higher hydrophobicity. All of the recorded water contact angle values for the tested coatings are summarized in Table 2. As expected, the PU coating using a higher content of lower-molecular-weight PTMG had a smaller water contact angle. The water contact angle value continuously decreased with increasing lower-molecular-weight PTMG. However, the water contact angle value significantly increased with the inclusion of PDMS, and the value continuously increased with increasing PDMS content. The water contact angle value further increased with CeO 2 content. The polymerized coatings also showed higher values compared to the blended coatings. Coatings with well-mixed CeO2 had higher stiffness and also resisted water penetration; ultimately, the contact angle value increased. The highest water contact angle value was recorded for the PU-11-3-3-CeO 2 -P coating, which was prepared with two PTMG polyols, a PDMS-polyol, and CeO 2 . Further, polymerization was applied during CeO 2 use.
Any material to be considered for coating applications should have a minimum adhesive strength. The adhesive strength should remain at a minimum level throughout the coating's lifespan. PU coatings have good adhesive strength for fabrics, metals, and PVC surfaces. The adhesive strength values are summarized in Table 2. As expected, a good adhesive strength was recorded using a higher-molecular-weight PTMG. The adhesive strength was affected by adding low-molecular-weight-based polyols. Both merits and demerits in terms of adhesive strength were found when using low-molecular-weight-based polyols. When half of the high-molecular-weight polyol was replaced with low-molecular-weight polyol (9.05 mol%), the adhesive strength decreased significantly (PU-9 coatings). Using a huge amount (PU-9 coatings) of low-molecular-weight polyols resulted in less mechanical strength, which decreased the adhesive strength. However, adhesive strength increased with the addition of 2.25, 4.525, and 6.775 mol% low-molecular-weight PDMS-polyols (PU-11, PU-13, and PU-15 coatings). The continuous increase of low-molecular-weight PDMS-polyol content in the PU-11, PU-13, and PU-15 coatings slightly affected the hardness, which was maximal in the PU-11 coatings. Eventually, adhesive strength reached a maximum at 6.775 mol% low-molecular-weight polyols. The adhesive strength increased by almost 20%-22% when adding 6.775 mol% low-molecular-weight mixed polyols (PU-11-3-3 coating). In all cases in the PU-11, PU-13 and PU-15 coatings, the adhesive strength was slightly higher when using low-molecular-weight mixed polyols, especially when compared to a single low-molecular-weight polyol. The adhesive strength value also increased with the addition of CeO 2 . The CeO 2 acted as a filler in the coating, which was why the adhesive strength increased. The adhesive strength increased by almost 28%-40% with the addition of 1.0 wt% CeO 2 content. At the same time, the adhesive strength values for CeO 2 -based coatings were always higher for polymerized than blended coatings. A better interaction between the CeO 2 and polymer functional groups in the PU-CeO 2 polymerized coating provided an extra improvement in adhesive strength. Among all coatings, the maximum adhesive strength was recorded for the PU-11-3-3-CeO 2 -P coating. This enhanced adhesive strength was due to using a proper content of mixed polyols, as well as the proper mixing of CeO 2 during polymerization. The coatings were tested for self-healing via optical microscopy. All coatings showed self-healing characteristics, as all scratches healed within 12 h. However, the PU-9 coatings healed faster (9 h) than the other coatings. The fastest healed time was found to be 7 h for the PU-9-9-0 coating. As all coatings showed self-healing characteristics the first time, they were also considered for healing a second time. This time, the PU-18, PU-13, and PU-15 coatings failed to heal during the defined duration. The scratches in the PU-18, PU-13, and PU-15 coatings remained after that defined (12 h) duration. The PU-18, PU-13, and PU-15 coatings were comparatively rigid in nature. The rigid structure opposed chain mobility, and ultimately, the coating did not gain a self-healing nature. Interestingly, the PU-9 and PU-11 coatings showed self-healing characteristics the second time as the scratch disappeared within 12 h. In the PU-9 and PU-11 coatings, the flexibility was sufficient to regain the shape, as the scratch healed within that time. The proper balance of polyols made it easy to switch the polymer chain to repair the damage. The PU-9 and PU-11 coatings were also considered for a third healing test. The typical optical photographs are shown in Figure 3. These coatings showed a repairing tendency again and healed within 12 h. However, proper healing was not seen for a fourth time with the defined duration. When the coatings were scratched three times, the polymer chains lost their chain entanglement capacity.
However, although the PU-9 coatings performed positively in terms of self-healing characteristics, they also showed very poor adhesive strength. Thus, the PU-9 coatings were not considered for the UV shielding test. The PU-11-3-3 coating showed excellent water resistance, adhesive strength, and self-healing, and this coating was considered for the UV shielding test. To minimize the number of testing coating samples, only the PU-11-3-3 coating, which showed the highest adhesive strength, Polymers 2020, 12, 1947 8 of 11 was considered for mixing with CeO 2 . As mentioned before, blending and in situ polymerization were used to include the CeO 2 content (see Table 1). These coatings were also considered for the multi-time self-healing tests within the same duration of 12 h. The in situ polymerized coating healed three times within that duration (exactly 11 h). However, the blended coating only healed the first time. Properly interacted CeO 2 did not interfere during healing, and the healing was faster. The blended CeO 2 had less interaction with PU groups, and free CeO 2 may oppose polymer chain entanglement; ultimately, the healing stopped or slowed down.
coating. As all coatings showed self-healing characteristics the first time, they were also considered for healing a second time. This time, the PU-18, PU-13, and PU-15 coatings failed to heal during the defined duration. The scratches in the PU-18, PU-13, and PU-15 coatings remained after that defined (12 h) duration. The PU-18, PU-13, and PU-15 coatings were comparatively rigid in nature. The rigid structure opposed chain mobility, and ultimately, the coating did not gain a self-healing nature. Interestingly, the PU-9 and PU-11 coatings showed self-healing characteristics the second time as the scratch disappeared within 12 h. In the PU-9 and PU-11 coatings, the flexibility was sufficient to regain the shape, as the scratch healed within that time. The proper balance of polyols made it easy to switch the polymer chain to repair the damage. The PU-9 and PU-11 coatings were also considered for a third healing test. The typical optical photographs are shown in Figure 3. These coatings showed a repairing tendency again and healed within 12 h. However, proper healing was not seen for a fourth time with the defined duration. When the coatings were scratched three times, the polymer chains lost their chain entanglement capacity. However, although the PU-9 coatings performed positively in terms of self-healing characteristics, they also showed very poor adhesive strength. Thus, the PU-9 coatings were not considered for the UV shielding test. The PU-11-3-3 coating showed excellent water resistance, adhesive strength, and self-healing, and this coating was considered for the UV shielding test. To minimize the number of testing coating samples, only the PU-11-3-3 coating, which showed the highest adhesive strength, was considered for mixing with CeO2. As mentioned before, blending and in situ polymerization were used to include the CeO2 content (see Table 1). These coatings were also considered for the multi-time self-healing tests within the same duration of 12 h. The in situ polymerized coating healed three times within that duration (exactly 11 h). However, the blended coating only healed the first time. Properly interacted CeO2 did not interfere during healing, and the healing was faster. The blended CeO2 had less interaction with PU groups, and free CeO2 may oppose polymer chain entanglement; ultimately, the healing stopped or slowed down.
To evaluate the UV shielding resistance, only five coatings (i.e., PU-11-3-3, PU-11-3-3-Ce-P, 11-3-3-Ce-P-1 st -Healed, 11-3-3-Ce-P-2 nd -Healed, and 11-3-3-Ce-P-3 rd -Healed) were considered. Visually, it was not possible to verify the coatings' resistance against UV degradation in the very early stages. When a coating rapidly degraded, which thus initiated metal corrosion, the degradation impact was indicated by color change, pit appearance, and delaminating. The PU-11-3-3 coating was without CeO2 and was completely degraded within the mentioned duration (3 months), wherein the specimen hugely corroded. The coating color changed to yellowish. Some part of the coating was also delaminated. This implies that without a UV shielding material, the coatings could not be used for UV protection. The other coatings were almost free of discoloration, To evaluate the UV shielding resistance, only five coatings (i.e., PU-11-3-3, PU-11-3-3-Ce-P, 11-3-3-Ce-P-1st-Healed, 11-3-3-Ce-P-2nd-Healed, and 11-3-3-Ce-P-3rd-Healed) were considered. Visually, it was not possible to verify the coatings' resistance against UV degradation in the very early stages. When a coating rapidly degraded, which thus initiated metal corrosion, the degradation impact was indicated by color change, pit appearance, and delaminating. The PU-11-3-3 coating was without CeO 2 and was completely degraded within the mentioned duration (3 months), wherein the specimen hugely corroded. The coating color changed to yellowish. Some part of the coating was also delaminated. This implies that without a UV shielding material, the coatings could not be used for UV protection. The other coatings were almost free of discoloration, brittleness, delamination, and corrosion. Typical photos of healed exposed coatings are shown in Figure 4. Visual inspection also confirmed that the PU coating using CeO 2 had good UV shielding capacity. No significant difference was seen for these coatings. All coatings were considered for further XPS analysis. Almost the same peak appeared at the same position, confirming that they were least affected under exposure conditions. A typical deconvulated curve analysis was also done for a specific C1s area. It has been reported before that a peak appears at around 289 eV for the CO group due to chain scission [4]. As the carboxyl group also has a peak at the same position, the intensity of that peak usually increases if there is any chain scission. We found that all coatings had very slightly higher intensity peaks at that position (see Figure 5), which implies almost no degradation for all coatings. All healed coatings had similar UV shielding, as the peaks were of similar types. Although the PU-CeO 2 coating showed promising results in self-healing and UV-shielding, the operational temperature for self-healing is quite high. The self-healing time is also longer. Above 1.0 wt% CeO 2 content also can affect the self-healing and UV-shielding properties. Currently, our group is working to solve these issues. very slightly higher intensity peaks at that position (see Figure 5), which implies almost no degradation for all coatings. All healed coatings had similar UV shielding, as the peaks were of similar types. Although the PU-CeO2 coating showed promising results in self-healing and UV-shielding, the operational temperature for self-healing is quite high. The self-healing time is also longer. Above 1.0 wt % CeO2 content also can affect the self-healing and UV-shielding properties. Currently, our group is working to solve these issues.

Conclusions
A coating with multi-time self-healing and UV-shielding properties (i.e., PU-11-3-3-Ce-P) was prepared using a proper combination of polyols and CeO2 content, as well as an in situ polymerization application. The UV degradation resistance was improved by the addition of a small CeO2 content (1.0 wt %) without sacrificing the basic protective properties of hydrophobicity and adhesive strength. A small content (6.775 mol %) of mixed low-molecular-weight polyols contributed to 3 times better self-healing. The healed coating also showed excellent UV degradation resistance for open atmospheric conditions. This type of coating can be useful for outdoor applications.  degradation for all coatings. All healed coatings had similar UV shielding, as the peaks were of similar types. Although the PU-CeO2 coating showed promising results in self-healing and UV-shielding, the operational temperature for self-healing is quite high. The self-healing time is also longer. Above 1.0 wt % CeO2 content also can affect the self-healing and UV-shielding properties. Currently, our group is working to solve these issues.

Conclusions
A coating with multi-time self-healing and UV-shielding properties (i.e., PU-11-3-3-Ce-P) was prepared using a proper combination of polyols and CeO2 content, as well as an in situ polymerization application. The UV degradation resistance was improved by the addition of a small CeO2 content (1.0 wt %) without sacrificing the basic protective properties of hydrophobicity and adhesive strength. A small content (6.775 mol %) of mixed low-molecular-weight polyols contributed to 3 times better self-healing. The healed coating also showed excellent UV degradation resistance for open atmospheric conditions. This type of coating can be useful for outdoor applications.

Conclusions
A coating with multi-time self-healing and UV-shielding properties (i.e., PU-11-3-3-Ce-P) was prepared using a proper combination of polyols and CeO 2 content, as well as an in situ polymerization application. The UV degradation resistance was improved by the addition of a small CeO 2 content (1.0 wt%) without sacrificing the basic protective properties of hydrophobicity and adhesive strength. A small content (6.775 mol%) of mixed low-molecular-weight polyols contributed to 3 times better self-healing. The healed coating also showed excellent UV degradation resistance for open atmospheric conditions. This type of coating can be useful for outdoor applications.