Performance Prediction of Antiaging Aliphatic Polyurethane Topcoat

Abstract

Polyurethane coatings are widely used due to their excellent flexibility, abrasion resistance, Ultraviolet stability, and chemical durability. However, quantitative evaluation of their aging behavior remains a challenge. In this study, two aliphatic acrylic polyurethane topcoats were subjected to accelerated cyclic aging tests, and their degradation processes were characterized by infrared spectroscopy and glossiness measurements. The ratio values of CO/C–H band area with time for two aliphatic acrylic polyurethane topcoats were found to reflect the increasing of carbonyl content, which governs the transformation of surface microstructure and glossiness. A significant negative correlation between glossiness and the semiquantitative analysis of carbonyl groups was confirmed by Spearman correlation coefficients of −0.772 and −0.782, respectively. Furthermore, the relationship between glossiness and aging time was described by an exponential function, indicating that glossiness decreases progressively with extended exposure. These results provide a quantitative framework for assessing the aging performance of polyurethane coatings and offer a basis for predicting their long-term durability under ultraviolet exposure.

Introduction

It is reported that atmospheric corrosion causes greater economic losses than other types of corrosion on different materials. Currently, the most effective, economical, and widely applicable method of preventing atmospheric corrosion is using organic coatings.

Polyurethane (PU) has been in use for more than 90 years since its introduction in the 1930s. With the development, PU materials have been widely utilized in various sectors, including aerospace, construction, automobile manufacturing, petroleum, metallurgy, electronic engineering, medical equipment, and the light industry. In 2022, the global consumption of PU raw materials exceeded 25.18 million tons, with a value of production exceeding $708.9 million.

Polyurethane coatings are widely used to protect substrates, due to their excellent flexibility, abrasion resistance, adhesion, weather resistance, and chemical resistance. Depending on their structural systems, polyurethanes may be categorized as aliphatic or aromatic polyurethanes. Among these, aliphatic polyurethanes have been shown to exhibit superior weather resistance and light stability, rendering them more suitable for prolonged outdoor exposure. For instance, Wang et al. prepared amine-functionalized graphene oxide (AFGO) and used amine-functionalized graphene oxide (AFGO) to reduce the bandgap of graphene oxide and improve the aging resistance of the coating to prepare AFGO/flexible PU (FPU) nanocomposite coatings.

However, the molecular structure of the polyurethane coating is susceptible to damage from light, heat, and various corrosive media in the environment. This damage leads to aging phenomena such as loss of light, discoloration, chalking, cracking, blistering, and others. Polyurethane aging is a chemical reaction that cannot be reversed. It often has a fatal impact on the performance of the coating, resulting in coating failure and loss of the protective effect on the substrate. PU coatings, when applied as a topcoat and exposed directly to the atmosphere, are particularly vulnerable to bond breakage and oxidation reactions induced by UV radiation. These chemical transformations typically involve urethane bond cleavage, CO₂ release, and carbonyl formation, resulting in substantial microstructural and optical changes. ,

Numerous studies have investigated the weathering and aging mechanisms of polyurethane coatings through both natural exposure and accelerated aging tests. For instance, Santos evaluated four PU topcoats using condensation and salt spray tests, combining glossiness, color, and Fourier transform infrared spectroscopy (FTIR) analyses to assess degradation. Yang compared low-glossiness aircraft coatings. It was revealed that adhesion loss, driven by permeability, was the factor limiting the coating’s lifespan. Aznar conducted a study on waterborne PU/acrylate topcoats and determined that relative glossiness change was a pivotal indicator of weathering resistance. Similarly, Qu conducted an analysis of acrylic urethane coatings when subjected to cyclic wet–dry conditions, utilizing electronic impedance spectroscopy (EIS), Fourier-transform infrared spectroscopy (FT-IR), and scanning electron microscopy (SEM) to investigate coating failure mechanisms.

Despite extensive progress, most studies have remained qualitative in nature. The extant research primarily documents visible degradation, such as glossiness loss and surface roughening, without establishing a quantitative linkage between these macroscopic observations and molecular-level chemical changes. The absence of a clear correlation between surface optical behavior and chemical structure evolution prevents the reliable prediction of coating performance decay.

Furthermore, existing accelerated aging evaluations tend to analyze glossiness, carbonyl index, or color shift as isolated indicators. To the best of the present author’s knowledge, there have been few efforts to date that have integrated these parameters into a unified framework capable of quantitatively describing the dynamics of aging. Consequently, a substantial research gap remains in the field of modeling the quantitative relationship between glossiness, infrared spectral changes, and exposure time, which is imperative for the predictive evaluation of polyurethane coating durability.

In order to address this gap in the literature, the present study conducted an accelerated cyclic aging test on two aliphatic polyurethane topcoats. The characterization of the coatings was conducted through the utilization of glossiness measurements and infrared spectroscopy, which facilitated monitoring of both physical and chemical transformations during the degradation process. A mathematical model was developed to describe the exponential relationship between glossiness and aging time by correlating glossiness variation with semiquantitative infrared spectral data. The present study proposes a novel quantitative framework for elucidating the aging mechanism of polyurethane coatings and concomitantly offers a predictive tool for assessing their long-term weathering performance.

Experimental Section

Materials

The substrate used Q235 steel plates with dimensions of 15.20 cm × 7.65 cm. The coating is applied by two different coating companies, denoted as “ZDS” and “HLS”, and consists of a primer and topcoat with a dry film thickness ranging from 240 ± 5 μm, based on the average of five measurements. After coating application, all samples were cured at room temperature for 7 days in order to ensure adequate curing, surface uniformity, and consistency of the initial surface morphology. The edges of all samples were sealed with the corresponding coating systems. For ZDS samples, Jotun Jotaprime 510 and Jotun Hardtop XP were applied as the primer and topcoat, respectively. For HLS samples, Hilonuprimer 3100 was used as the primer and HilonDeco PU 9400 as the topcoat. All coatings were prepared and applied in accordance with the technical instructions provided by their respective manufacturers.

According to the supplier technical data sheets, Jotun Hardtop XP is a two-component chemically cured aliphatic acrylic polyurethane topcoat, whereas HilonDeco PU 9400 is a two-component high-gloss acrylic polyurethane topcoat cured with an aliphatic isocyanate. The recommended mixing ratio for Jotun Hardtop XP is 10:1 by volume (Component A/Component B), with a recommended dry film thickness of 50–100 μm. For HilonDeco PU 9400, the recommended mixing ratio is 7:1 by volume (Component A/Component B), with a target dry film thickness of 50–70 μm.

Both Jotun Hardtop XP and HilonDeco PU 9400 are commercial two-component aliphatic acrylic polyurethane topcoats. Although the precise proprietary formulations of these products are not publicly available, their curing mechanism generally involves the reaction between hydroxyl-functionalized acrylic resins and aliphatic polyisocyanate hardeners to form a cross-linked polyurethane network. Specifically, the hydroxyl groups (−OH) in the acrylic polyol component react with isocyanate groups (−NCO) in the curing agent to form urethane linkages (−NH–COO−), as illustrated in the following general reaction

$$\mathrm{R}-\mathrm{OH}+{\mathrm{R}}^{\prime }-\mathrm{NCO}\to \mathrm{R}-\mathrm{NH}-\mathrm{COO}-{\mathrm{R}}^{\prime }$$

This reaction forms a three-dimensional cross-linked coating structure, which confers excellent protective and mechanical properties to the polyurethane topcoat.

Accelerated Aging Experiment

An accelerated aging experiment was conducted following the ISO 12944 standard. The experiment consisted of 16 cycles, with each cycle lasting a total of 168 h. Each cycle included a 72-h UV accelerated aging procedure, a 72-h salt spray procedure, and a 24-h freezing procedure at −20 °C. The UV accelerated aging procedure was conducted using a QUV device under the conditions of 4 h at 60 °C irradiated by UVA340 at an intensity of 0.89 W/m²/nm, and 4 h at 50 °C condensation alternately. During the condensation stage, the humidity was set to 100% as specified by the standard, while relative humidity was not individually monitored during other stages of the cycle. The salt spray experiment was conducted according to ASTM B117-19 standard.

Coating Characterization

The characterization of the coating was performed for each aging cycle. The glossiness of the samples was measured by means of a 3nh glossiness meter at 20°, with the average of five measurements obtained. Additionally, the samples were analyzed using the PE200 series (FT-IR) with the attenuated total reflection (ATR) mode, which recorded the infrared spectra of the sample surface. Meanwhile, samples were analyzed with the EM30AX PLUS instrument (SEM) at 15 kV. The surface topography and height differences of the coatings were examined using laser confocal microscopy (OLS4100). Water contact angle measurements were also performed using an OCA25 instrument. Together, these characterization techniques provided information about the physical and chemical properties of the coatings throughout the aging test.

Results

Morphological Analysis

Figures and show the surface of the samples (ZDS and HLS) before and after the accelerated aging experiments. The results have shown that the glossiness of the test samples gradually decreased and the surface became rough.

1.

Appearance of the sample ZDS observed at 0 h (a), 1344 h (b), and 2688 h (c) of the accelerated aging experiment.

2.

Appearance of the sample HLS observed at 0 h (a), 1344 h (b), and 2688 h (c) of the accelerated aging experiment.

Figures and show the SEM images of ZDS and HLS after accelerated aging experiments at 0 h (a), 1344 h (b), and 2688 h (c), respectively.

3.

SEM images of ZDS polyurethane topcoat after 0 h (a), 1344 h (b), and 2688 h (c) accelerated aging experiment. The red box indicates surface defects that have arisen due to the aging process of the topcoat.

4.

SEM images of HLS polyurethane topcoat after 0 h (a), 1344 h (b), and 2688 h (c) accelerated aging experiment. The red box indicates surface defects that have arisen due to the aging process of the topcoat.

As can be seen from Figures and , with the accelerated aging experiment, the originally smooth surface of the topcoat was gradually blistered and many pits appeared, and the topcoat was gradually aging and failing.

Figures and present laser confocal microscopy images of the ZDS and HLS coatings at aging times of 0, 672, 1344, 2016, and 2688 h, showing the macroscopic surface morphology and corresponding height variations. The results reveal that surface height variations became increasingly pronounced as aging progressed. The accelerated aging process led to increased surface roughness of the coatings, as evidenced by the formation of surface pits. These observations are consistent with the findings from the macroscopic photographs and SEM images.

5.

Laser confocal microscopy images of HLS and ZDS polyurethane topcoat (macroscopic surface morphology) after 0, 672, 1344, 2016, and 2688 h accelerated the aging experiment.

6.

Laser confocal microscopy images of HLS and ZDS polyurethane topcoat (height variations) after 0, 672, 1344, 2016, and 2688 h accelerated the aging experiment.

Water Contact Angle

Water contact angle measurements were conducted on ZDS and HLS coating samples before and after the aging test to assess changes in surface wettability. The results are presented in Figure . The initial water contact angles of the unaged ZDS and HLS polyurethane topcoats were 93.7 and 94.0°, respectively, indicating that both coatings possessed relatively hydrophobic surfaces. Following 2688 h of accelerated aging, a decline in water contact angles was observed, with values decreasing to 86.8° for ZDS and 73.7° for HLS.

7.

Water contact angle measurements of ZDS and HLS polyurethane topcoat after 0, 672, 1344, 2016, and 2688 h accelerated the aging experiment.

As the aging test proceeded, the water contact angles of both coatings declined continuously, suggesting a gradual increase in surface wettability. These results demonstrate that the surface characteristics of the polyurethane topcoats were modified during aging. Combined with the aforementioned surface morphology observations, it can be inferred that the aging process induced an increase in surface roughness and the formation of surface defects, which collectively contributed to the reduction in water contact angle.

Infrared Analysis

The obtained infrared spectra of the samples were processed to identify vibrational peaks. The entire spectrum underwent baseline flattening and peak splitting, as shown in Figure .

8.

FT-IR spectrum of the sample ZDS (a) and HLS (b).

Figure illustrates the FT-IR spectra of the ZDS and HLS polyurethane topcoats after 0, 672, 1344, 2016, and 2688 h of accelerated aging, respectively. For both coating systems, the broad absorption band at approximately 3375–3382 cm^–1 is attributed to the stretching vibrations of N–H and/or O–H groups, while the absorption bands near 2920 and 2870 cm^–1 correspond to the stretching vibrations of aliphatic C–H groups. The characteristic absorption bands observed at 1726 and 1687 cm^–1 are assigned to carbonyl groups existing in different chemical environments. In contrast, the bands located at 1451–1454, 1239–1243, and 1158–1159 cm^–1 are associated with CH bending vibrations and C–O/C-N/C–O–C stretching vibrations within the polyurethane network.

With increasing aging time, gradual changes were observed in the characteristic absorption bands of both two coatings, particularly in the carbonyl region. These spectral changes indicate the progressive evolution of the chemical structure of the coatings during accelerated aging. Such alterations are consistent with the occurrence of photo-oxidation and hydrolysis reactions, which may induce chain scission and the formation of oxygen-containing degradation products. Compared with ZDS, HLS exhibited more significant spectral changes after prolonged aging, suggesting that the HLS coating underwent more substantial chemical degradation. This observation is consistent with the more obvious deterioration observed in the surface morphology and wettable measurements.

To obtain a reliable average value for the band area ratio, the raw spectra were first processed using PerkinElmer Spectrum IR software, including smoothing and baseline correction. To resolve the overlapping absorptions in the carbonyl region, deconvolution was performed over the range of 1600–1800 cm^–1.

The deconvoluted spectra were subsequently subjected to further analysis using Origin 2022. The 1600–1800 cm^–1 region was analyzed to decompose the carbonyl band into two sub-bands centered at approximately 1725 and 1688 cm^–1, respectively. The band area of each component was calculated individually.

The relative evolution of carbonyl groups was evaluated by comparing the band area ratio of the CO and C–H absorptions. The C–H band area was integrated over the range of 2800–3000 cm^–1, whereas the CO band area was integrated over the range of 1710–1800 cm^–1. The ratio of the CO band area to the C–H band area, denoted as σCO/C–H, was calculated using eq .

$${\sigma }_{\mathrm{C}\mathrm{O}/\mathrm{C}\text{−}\mathrm{H}}={\sigma }_{\mathrm{C}\mathrm{O}}/{\sigma }_{\mathrm{C}\text{−}\mathrm{H}}$$ 1

Here, the σ_CO stood for the band area of CO integrated in the infrared spectrum, while σ_C–H stood for the band area of C–H integrated in the same spectrum.

It was observed that the σ_CO/C–H values of the HLS and ZDS samples increased over time. Specifically, the σ value for HLS samples increased from 0.456 to 0.760, and that for ZDS samples increased from 0.752 to 1.10, as shown in the time-dependent images in Figure . This phenomenon could be attributed to the promotion of polyurethane polymer chain segment oxidation caused by UV radiation. The aging process of polyurethanes involved both chain breakage and cross-linking processes, resulting in the generation of a significant number of carbonyl groups.

9.

Ratio of CO/C–H band area with time for ZDS and HLS.

Glossiness Analysis

The specific glossiness of the samples was represented by eq

$$G_t=G_t/G_0\times 100\%$$ 2

here, G_t is the glossiness of the sample at time t, and G ₀ represents the initial glossiness of the sample.

To determine the relative glossiness of each sample for every cycle, the initial glossiness was assumed to be 100%, and eq was used to calculate the relative glossiness. Each sample was evaluated at 5 points per cycle, and the average of the calculated relative glossiness values was recorded as the relative glossiness for that cycle. A scatter plot was then created to compare the glossiness of the ZDS and HLS samples over time, as shown in Figure .

10.

Scatter plots of the glossiness with time for ZDS and HLS.

For the regression fitting of glossiness over time, linear regression is initially conducted, resulting in the eqs and for ZDS and HLS samples

$$G(\mathrm{ZDS})=100\%-2.71\times {10}^{-4}t$$ 3

$$G(\mathrm{HLS})=100\%-1.55\times {10}^{-4}t$$ 4

where, t was time (in hours), G(ZDS), and G(HLS) represented the corresponding glossiness.

The results showed that the linear regression coefficients for ZDS and HLS were 0.896 and 0.908, respectively.

However, it was observed that the overall trend of the scatter plots could be better fitted with an exponential regression form, suggesting that a nonlinear regression method might be more suitable. Nonlinear regression methods allowed us to choose appropriate models based on the specific data and situations. In this case, an exponential regression was used to obtain a fitting equation that better captures the actual trend. Equations and represented the nonlinear regression equations for the ZDS and HLS samples, respectively.

$$G(\mathrm{ZDS})=0.552\exp (-t/917.39)+0.404$$ 5

$$G(\mathrm{HLS})=0.683\exp (-t/4142.63)+0.271$$ 6

The nonlinear regression coefficients for the ZDS and HLS samples indicated the goodness of fit of the regression model to the data. A coefficient close to 1 indicates a better fit. In this case, the nonlinear regression coefficients for the ZDS and HLS samples were 0.980 and 0.969, respectively. Both the linear and nonlinear regression results are shown in Figure .

11.

Linear and exponential fitting of glossiness with time for ZDS and HLS.

Correlation Analysis

A normal Q–Q plot was generated using SPSS software for the data sets σ­(ZDS), σ­(HLS), G(ZDS), and G(HLS), as shown in Figures and . The results indicated that all of these data sets followed a normal distribution. Subsequently, the correlation coefficients between σ­(ZDS) and G(ZDS), as well as between σ­(HLS) and G(HLS), were calculated using SPSS software. The Spearman correlation coefficient between σ­(ZDS) and G(ZDS) was −0.772, and −0.782 between σ­(HLS) and G(HLS), as shown in Table and Table . These correlation coefficients were significant at the 0.01 level (two-tailed), indicating a significant correlation between them.

12.

Normal Q–Q plot of σ­(ZDS) and G(ZDS).

13.

Normal Q–Q plot of σ­(HLS) and G(HLS).

1. Spearman Correlations of σ­(ZDS) and G(ZDS).

			σ(ZDS)	G(ZDS)
Spearman’s rho	σ(ZDS)	correlation coefficient	1.000	–0.772**
		significance (two-tailed)	0	0.0002
		N	17	17
	G(ZDS)	correlation coefficient	–0.772**	1.000
		significance (two-tailed)	0.0002	0
		N	17	17

The double asterisks (**) indicate that the correlation is statistically significant at the 0.01 level (two-tailed).

2. Spearman Correlations of σ­(HLS) and G(HLS).

			σ(HLS)	G(HLS)
Spearman’s rho	σ(HLS)	correlation coefficient	1.000	–0.782**
		significance (two-tailed)	0	0.0003
		N	16	16
	G(HLS)	correlation coefficient	–0.782**	1.000
		significance (two-tailed)	0.0003	0
		N	16	16

The negative correlation coefficients suggested an inverse relationship between the glossiness and the aging level of the samples. As the aging level increases, the glossiness would decrease. This relationship was consistent with the findings of previous studies, which have shown that aging tests can lead to the degradation of the surface properties of materials, such as glossiness.

Discussion

The aging behavior of polyurethane (PU) topcoats under the influence of ultraviolet (UV) radiation, moisture, and oxygen involves a series of complex photo-oxidation reactions. , As stated by Bauer, the increase in carbonyl content during UV aging is proportional to the square root of the photoinitiation rate. This confirms that Fourier-transform infrared spectroscopy (FT-IR) is an effective tool for monitoring coating degradation; , however, caution should be exercised when extending this conclusion to other formulations, since differences in initial composition, cross-linking structure, and additive chemistry may influence both the original spectra and the degradation pathways. Bhargava utilized ATR-FTIR to monitor structural changes in organic coatings during UV and thermal aging. It was observed that UV irradiation caused degradation of polymer curing agents and the generation of hydroxyl radicals. This finding indicates that polyurethane topcoats are susceptible to radical-induced photo-oxidation, a process that is influenced by the combined effects of UV radiation, moisture, and oxygen. UV irradiation has been shown to induce bond cleavage and photo-oxidation, whereas moisture facilitates hydrolytic degradation of susceptible linkages. Moreover, oxygen promotes oxidative conversion to carbonyls and other oxygen-containing species.

As illustrated in Figure , the primary reactions occurring during the aging process are as follows: The urethane group has been observed to absorb water, undergoing hydrolysis to form amino acids and alcohol compounds. When PU is subjected to prolonged ultraviolet irradiation, a process of photoexcitation occurs, leading to the formation of free radicals. The excited state generated upon energy absorption has been shown to transfer energy to adjacent ether bonds or hydrogen atoms, yielding ether radicals or carbon radicals. It has been established that, under the influence of atmospheric oxygen, these radicals undergo oxidation reactions, which in turn lead to the further generation of aldehydes, ketones, and carboxylic acids. This process has been shown to result in an increase in the relative content of carbonyl groups, as measured by FT-IR.

14.

Schematic illustration of the aging mechanism of polyurethane topcoats under UV irradiation, oxygen, and moisture, showing the main photo-oxidation and hydrolysis pathways and their relationship to surface deterioration.

The present study employed FT-IR analysis, revealing that the oxidation reaction of ether bonds promotes the formation of aldehyde groups, resulting in a progressive increase in the relative carbonyl content alongside a sustained decrease in glossiness. Glossiness retention exhibited a significant negative correlation with carbonyl formation, indicating that glossiness loss partially reflects surface aging within the topcoats. However, research has demonstrated that cleavage and breakage of internal macromolecular structures and segments can compromise the molecular integrity and cohesion. The aforementioned factors ultimately result in an increased porosity of the topcoats, reduced adhesion, chalking, and embrittlement. The macroscopic manifestations of these phenomena are reflected in the study as diminished glossiness, an uneven surface finish of the topcoat, and a decrease in the water contact angle.

The experimental data further indicates that the glossiness declines exponentially with increasing duration of aging and eq was used to express the change of glossiness over time as follows

$$G_t=A\exp (-t/B)+C$$ 7

In eq , t represents the aging time, G_t represents the relative glossiness of the coating after aging for t hours, and the remaining terms are coefficients specific to the aging experiment.

Further experimental data indicate that glossiness exhibits an exponential decay trend with aging time. This suggests that the degradation of polyurethane topcoats is associated with rate-determining photochemical oxidation processes. Successful nonlinear regression fitting validates this interpretation, establishing a correlation between molecular-scale chemical reactions and the macroscopic optical appearance of the topcoats.

At present, there is an absence of a universally accepted portable quantitative standard for evaluating the weather resistance of the topcoats. The correlation established in this study between glossiness loss and semiquantitative FT-IR analysis provides a foundation for the development of predictive models of topcoat aging. The FT-IR method described herein shows promise for future predictive modeling, although its application requires preaging baseline spectra. The integration of glossiness measurements with infrared spectral parameters has enabled the development of a practical, portable, quantitative method for the assessment of the aging extent and service life of polyurethane topcoats. This approach offers promising prospects for predicting the lifespans of protective topcoats in future research and industrial applications.

However, it is imperative to acknowledge the potential influence of additional degradation mechanisms on the service life of protective topcoats, which were not considered in the present study. For instance, mechanical damage, including abrasion, impact, and cracking, has been shown to accelerate coating failure. , Furthermore, biological activity in certain service environments can also contribute to surface deterioration. Although these factors were not investigated in this work, they merit consideration in future research and practical durability evaluations.

Conclusions

Based on the accelerated cyclic aging tests conducted on two aliphatic polyurethane topcoats, combined with infrared spectroscopy and glossiness measurements, the following conclusions can be drawn:

• 1.The ratio values of σ­(ZDS) and σ­(HLS) effectively reflected the variation in carbonyl content on the topcoat surface, demonstrating that the increase in carbonyl groups is associated with changes in surface microstructure and glossiness.
• 2.The relationship between glossiness and aging time follows an exponential decay function, indicating that glossiness decreases progressively as aging time increases.
• 3.The Spearman correlation coefficients between glossiness and the semiquantitative infrared spectral analysis were −0.772 and −0.782 for the two coatings, respectively, confirming a significant negative correlation between glossiness loss and the degree of aging.

Overall, under the conditions of this test, this study establishes a quantitative approach to characterize the degradation of polyurethane coatings and provides a mathematical model to predict glossiness evolution during aging.

However, the present study is limited by a relatively small sample size and the lack of long-term field validation. Future work should include more samples and investigate coatings under different UV intensities and formulations to further verify and generalize the proposed model.

X.X.: writingreview and editing, writingoriginal draft, methodology, investigation, conceptualization. T.W.: investigation, conceptualization. J.L.: investigation. P.L.: investigation. W.W.: resource, conceptualization, writingoriginal draft, supervision, funding acquisition, project administration. H.W.: writingreview and editing, supervision, methodology, conceptualization, project administration. F.W.: resource, supervision, methodology, conceptualization, project administration.

The authors declare no competing financial interest.