Self-Cleaning Road Marking Paints for Improved Road Safety: Multi-Scale Characterization and Performance Evaluation Using Rhodamine B and Methylene Blue as Model Pollutants

Abstract

Throughout the lifetime, road markings (RMs) accumulate dirt, oils, and greases, which reduce visibility, shorten service life, and compromise road safety. If RMs could degrade these pollutants, their service life would increase. When exposed to UV light and humidity, semiconductors, such as titanium dioxide (TiO₂), can interact with contaminants and promote their chemical degradation. Semiconductors are commonly used on different types of substrates to achieve self-cleaning ability. In this study, 0.25–3 wt% TiO₂ was incorporated into a commercial RM paint for this purpose. After functionalization, the RM paint samples were contaminated with Methylene Blue and Rhodamine B. After pollution, the specimens were irradiated with a light source that simulates sunlight. To assess the self-cleaning capacity of the paints, visual analysis, color variation and discoloration by using CIELAB color coordinates, diffuse reflectance, and digital image processing techniques were applied. In both techniques, the samples with 2% and 3% of TiO₂ showed a greater capacity to degrade pollutants. Further, the chemical and morphological characteristics of the reference paint and the samples that showed the best self-cleaning results were analyzed by using Fourier Transform Infrared Spectroscopy (FTIR), Scanning Electron Microscopy (SEM), Energy-Dispersive X-ray Spectroscopy (EDS), and X-ray Diffraction (XRD). They identified the polymer, filler, and pigment in the commercial paint and confirmed the TiO₂ increase after functionalization. This study demonstrated the innovative potential of incorporating semiconductors to achieve a new capability (self-cleaning) for RM paints. This breakthrough not only has the potential to extend the RM service life, but also to improve road safety through greater visibility.

1. Introduction

Road markings (RMs) consist of two layers working together as a functional system and play a crucial role in ensuring road safety (Figure 1). They provide visual information about the road and guidance to drivers [1,2,3,4]. The first layer, also referred to as the paint layer or base layer, adheres to the road pavement surface and improves daytime visibility through contrasting colors. The use of waterborne and solvent-borne paints is preferable due to their ease of application, availability, and affordability [5]. The second layer, the retroreflective layer, comprises glass beads anchored to the paint layer to improve nighttime visibility and protect the underlying layer from wear and tear. The glass beads should be applied immediately after the base layer. Ideally, they have diameters below 2000 μm and must be embedded in the first layer about 60% of their diameter [2,6].

The visual properties of RM, particularly luminance and reflectance [7], are crucial factors in assessing their remaining service life. When these properties deteriorate, the RM should be renewed through the application of a new thin layer over the existing ones [8]. Early studies on these paints focused on improving these visual properties, especially the retroreflectance. Moreover, recent research advances have led to the exploration of smart materials in various aspects of road infrastructure. Consequently, new studies have shifted their focus towards enhancing paints with additional smart functionalities, such as photoluminescence, anti-aging, and thermochromism [4,9,10]. This entails the application of semiconductor nanoparticles, thermocapsules, photoluminescent materials, and other innovative technologies. These advancements have as their main goal the improvement of visibility conditions and the extension of the RM service life [11].

The visibility of RM is often compromised during rain, and abrasion and dirtiness due to traffic action [12,13]. The integration of semiconductor materials, such as titanium dioxide (TiO₂) and zinc oxide (ZnO), into RM, with the aim of enhancing their self-cleaning capabilities, presents an opportunity to enhance the visibility of RM and, consequently, road safety and service life. When exposed to UV light and humidity, these semiconductor materials generate superoxide ions that interact with contaminants, including dirt, organic compounds (such as oils and greases), and inorganic substances, causing their chemical degradation. Therefore, they provide a self-cleaning property due to their photocatalytic capacity [11,14,15,16].

The self-cleaning capability has already been studied in several substrates, such as asphalt pavements, mortars, glasses, and paints for architectural purposes [15,16,17]. There are a few examples that help illustrate the techniques the researchers used to obtain this function and the assessment methods. Van Hal et al. [18] investigated TiO₂- and ZnO-coated glass slides for soot degradation, assessing discoloration by digital image processing, and reported degradation rates above 90% under ultraviolet (UV) light. Reza Omranian et al. [16] functionalized asphalt mixtures with nano-TiO₂, simulated pollution by using soot particles, and assessed the self-cleaning potential through digital image processing. After UV exposure, they concluded that higher soot coverage degradation (from 65.3% to 7.2%) was observed at increased TiO₂ concentrations. Rocha Segundo et al. [15] sprayed a solution with nano-TiO₂ and micro-ZnO particles over asphalt mixtures and attested the ability to photodegrade organic materials (up to 92%) by the degradation of the model pollutant Rhodamine B.

Lessa et al. [14] compared cementitious façades functionalized with nano-TiO₂ and micro-ZnO, evaluating Rhodamine B degradation under UV–Vis via CIELAB colorimetry, and found higher discoloration than the reference, with rates up to 36% (ZnO) and 35% (TiO₂). Smits et al. [19] assessed the self-cleaning capacity in mortars coated with different TiO₂ contents. Soot was applied as a pollutant model, and the degradation rates were assessed by area of soot coverage via digital image processing, and absorbance. The results indicated a reduced soot coverage area to nearly zero after 420 h of UV irradiation, and a decrease in absorbance values over irradiation time. Half of the absorbance reduction occurred within the first 50 h of irradiation, while complete oxidation of the soot required about 15 days for all coating conditions.

Silva et al. [20] evaluated the self-cleaning capacity of protective coatings containing TiO₂ and ZnO in external thermal insulation through the assessment of Rhodamine B, Methylene Blue, and commercial aerosol spray paint degradation. The discoloration rates were evaluated by measuring changes in CIELAB color coordinates. The results indicated a remarkable reduction in pollutant color, reaching up to 97% discoloration after solar exposure. Pal et al. [21] investigated the functionalization of resin paints and silicate paints through the incorporation of TiO₂ for the self-cleaning of surfaces. They used Rhodamine B and Methylene Blue as pollutant models, and the dye degradation was assessed by using the diffuse reflectance technique. The results revealed a significantly faster degradation rate in functionalized silicate paints, resulting in nearly complete discoloration of the dyes after 30 min of exposure to UV light.

In the field of self-cleaning RM, Quan et al. [22] incorporated SiO₂ (8.97 wt.%) and TiO₂ (10.46 wt.%) into a two-component acrylic-based RM coating. Their study evaluated the photocatalytic activity and superhydrophobic characteristics of the modified formulations. The findings revealed that the functionalized RM paint exhibited notable catalytic performance, achieving degradation rates of 23.4%, 8.3%, 2.5%, and 2.9% for NOx, hydrocarbons (HC), CO, and CO₂, respectively. In addition, the surface of the coating displayed a water contact angle of 134.2°, confirming its hydrophobic nature and effective self-cleaning behavior.

Taheri et al. [23] investigated self-cleaning RM formulated from an acrylic resin modified with 5–10 wt.% TiO₂ nanoparticles in both anatase and rutile crystalline phases. The specimens containing 7.5–10 wt.% TiO₂ exhibited pronounced photocatalytic activity. Colorimetric analysis revealed a Δb value of approximately 3*, indicating satisfactory stability against yellowing after accelerated aging. Furthermore, the Qd and RL parameters remained above 130 and 150, respectively, during the initial three months of outdoor exposure. Nevertheless, these formulations also presented an increased degree of polymer matrix degradation, as evidenced by greater mass loss and ester group decomposition. Overall, the study demonstrated that controlled incorporation of TiO₂ nanoparticles can significantly enhance the self-cleaning efficiency and surface whiteness of RM; however, excessive TiO₂ loading may promote accelerated photodegradation of the acrylic matrix.

Previous studies have reported the application of semiconductors to improve the self-cleaning ability of different substrates, yielding significant advances in this field. However, their application to RM is still at initial stages. The main novelty of this research lies in demonstrating the achievement of self-cleaning properties in RM paints through TiO₂ incorporation. Beyond material development, this study provides a comprehensive multi-technique evaluation of self-cleaning performance, including visual analysis of standardized sample pictures, CIELAB colorimetry, diffuse reflectance, and digital image processing. In addition, it combines optical, chemical, and morphological techniques to evaluate the functionalization process and its effects. This represents a technological advance for this road element. The results obtained demonstrate the potential of the proposed technology and open new perspectives for the application of self-cleaning coatings in road elements, contributing to extending the service life of RM and enhancing road safety by maintaining their visibility and high contrast with the pavement surface for longer.

2. Methodology

The approach of this research involved the functionalization of a commercial water-based acrylic paint through the incorporation of nano-TiO₂ at different mass incorporation ratios. To evaluate the effectiveness of the self-cleaning ability, widely used pollutant models were applied, Methylene Blue and Rhodamine B, along with optical detection methods for quantifying the photocatalytic removal of pollutants, including visual analysis, CIELAB color space, diffuse reflectance, and digital image processing. Moreover, a chemical and morphological analysis of the reference paint and the samples showing the best self-cleaning results was performed by using Fourier Transform Infrared Spectroscopy (FTIR), Scanning Electron Microscopy (SEM), Energy-Dispersive X-ray Spectroscopy (EDS), and X-ray Diffraction (XRD).

2.1. Materials

A white commercial water-based acrylic RM paint (

Table 1. RM paint characteristics according to EN 12802 and EN 1871 from ref. [24].

RM Paint Characteristics	Declared Values
Luminance factor, β	β ≥ 0.85; class LF7
Chromatic coordinates (x, y)	x, y within the polygon defined for white color
Density	1.67 ± 0.04 g/cm3
Solid content	80.0% ± 2
Ash content	37.6% ± 3
Coverage capacity	cr ≥ 95%; class HP4
UVB aging	Δß ≤ 0.05; class UV1
Storage stability	≥4

), from Candela (Aveiro, Portugal), was used as the substrate to be functionalized. A semiconductor catalyst (

Table 2. Nano-TiO₂ characteristics from ref. [24].

Nano-TiO2 Characteristics	Declared Values
Particle size	23–28 nm
Appearance	White powder
Purity	>99.5%
Crystalline phases	80% anatase, 20% rutile

), nano- TiO₂ (Aeroxide TiO₂ P25) from Quimidroga (Barcelona, Spain), was used to endow self-cleaning ability to the RM paint. Two pollutant models widely adopted for self-cleaning assessment were explored: Methylene Blue and Rhodamine B, both from Merck.

2.2. Methods

The study was developed into four phases (Figure 2). In the first phase, a commercial RM paint was functionalized by mass incorporation of different TiO₂ ratios. In the second phase, the paints, placed in glass slides, were contaminated with a controlled amount of pollutant models commonly used for self-cleaning assessment. Subsequently, the specimens underwent controlled intervals of light irradiation using a light source that simulates daylight. The samples were evaluated before and after pollution. At the end of each irradiation period, non-destructive measurement techniques, including visual analysis, CIELAB Color Coordinates, Diffuse Reflectance, and Digital Image Processing, were carried out to assess the self-cleaning capacity of the samples. They provided a comprehensive evaluation of this capacity for the RM paints functionalized with different amounts of TiO₂, as well as a comparison among the techniques. As a form of negative control, samples under the same conditions of functionalization and pollution were kept in the dark, and the same self-cleaning evaluation techniques were executed.

In the final phase, a chemical and morphological analysis was conducted on both the reference paint and the functionalized paints that showed the best results for self-cleaning. This phase aimed to compare the samples and verify the effectiveness of the functionalization process, attesting the incorporation of the semiconductor. To this end, Fourier Transform Infrared Spectroscopy (FTIR), Scanning Electron Microscopy (SEM), Energy Dispersive X-ray Spectroscopy (EDS), and X-ray Diffraction (XRD) tests were performed.

2.2.1. Paint Functionalization and Sample Preparation

The commercial RM paint was functionalized by the mass incorporation of nano-TiO₂ in five percentages by weight, namely 0.25%, 0.5%, 1%, 2% and 3%. Two hundred grams of paint was functionalized for each situation. The nanoparticle incorporation was carried out at room temperature by a low-shear mixer at 1000 RPM for 15 min [9]. The paints with different amounts of nano-TiO₂, including the reference paint (RP), were applied on glass slides (2.5 × 7.5 cm²) according to the EN 1436 ratio (750 g/m²) and allowed to dry.

Eight specimens were prepared for each nano-TiO₂ mass incorporation condition, including RP. Of these eight specimens, four were polluted with Methylene Blue and four were polluted with Rhodamine B. Of the four samples polluted with each pollutant model, two were used for analysis under light irradiation (test results to be averaged) and two to be kept in the dark (negative control), also averaged.

2.2.2. Sample Pollution

For each nano-TiO₂ mass incorporation condition, half of the specimens were polluted with 500 µL of Methylene Blue solution (100 ppm); the other half with 500 µL of Rhodamine B solution (100 ppm). The polluted samples were allowed to dry for 24 h.

2.2.3. Light Exposure

A lamp with a spectral distribution similar to solar irradiation (Osram Ultra Vitalux 300 W) (Munich, Germany) was used. The spectral radiation flux of this lamp has been reported elsewhere [25,26]. The lamp was positioned equidistantly 25 cm away from the samples, resulting in an incident light intensity of 11 W/m² as measured by a Quantum Photo Radiometer HD9021 Delta Padova (Selvazzano Dentro, Padova, Italy). Under these conditions, the surface temperature of the samples reached 50 °C. The specimens were irradiated for 2880 min. At irradiation times 0, 15, 30, 60, 120, 180, 360, 720, 1440, and 2880 min, the self-cleaning ability was assessed by monitoring the degradation of the pollutant model applied to the sample surfaces.

2.2.4. Self-Cleaning Ability Assessment

Visual Analysis

The visual analysis was carried out to monitor the discoloration (self-cleaning) achieved by the specimens after each interval of light irradiation. Images of the samples were captured in a photobox, where external light was excluded [16,18,19], and the lighting conditions, distance from camera to samples, and camera settings were the same for all the pictures. The entire surface area of the samples $\left(2.5\times 7.5 {\mathrm{c}\mathrm{m}}^2\right)$ was cropped in the photographs and considered to this analysis. The camera model Canon PowerShot SX50 HS (Tokyo, Japan) was used and fitted in the photobox 20 cm above the samples. Camera settings were set to manual mode (exposure time 1/100 s, F-stop: f/8, and ISO 80) in raw image file format (.CR2).

The images captured and cropped under the same conditions were grouped according to the pollutant model used, the amounts of nano-TiO₂ incorporated, and the irradiation periods they were subjected to. Thus, it was possible to qualitatively identify (visually) which specimens presented the best self-cleaning results.

Cielab Color Space

A portable chromameter PCE-CSM 20 (Meschede, Germany) was used to measure the CIELAB color coordinates (L*, a*, b*) of the samples. The measurements were taken using a Ø 8 mm measuring aperture, in specular component inclusion (SCI) mode, and using the D65 illuminant (that corresponds to average daylight illumination).

In the CIELAB color system, L* is the lightness, ranging from black (L* = 0) to white (L* = 100). The parameters a* and b* correspond to the chromatic coordinates, in which a* is the variation from red (+) to green (−), and b* is the variation from yellow (+) to blue (−) [20,27].

After acquiring the color coordinates at the previously mentioned time periods, it was possible to obtain the chroma or color saturation (C) according to Equation (1), using the chromatic coordinates (a* and b*).

$${\mathrm{C}=\left[{\left({\mathrm{a}}^{\ast }\right)}^{\mathrm{2}}+{\left({\mathrm{b}}^{\ast }\right)}^{\mathrm{2}}\right]}^{\mathrm{0.5}}$$

(1)

To assess the discoloration caused by the degradation of the pollutant models on the sample surfaces, it was calculated the photocatalytic efficiency through CILEAB color coordinates ($\mathrm{P}\mathrm{E}\mathrm{c}\mathrm{i}\mathrm{e}\mathrm{l}\mathrm{a}\mathrm{b}$), according to Equation (2) [20,28]:

$$\mathrm{P}\mathrm{E}\mathrm{c}\mathrm{i}\mathrm{e}\mathrm{l}\mathrm{a}\mathrm{b}\left(\mathrm{\%}\right)=\frac{\mathrm{C}\left(\mathrm{0}\right)-\mathrm{C}\left(\mathrm{t}\right)}{\mathrm{C}\left(\mathrm{0}\right)-\mathrm{C}\left(\mathrm{b}\mathrm{p}\right)}\times \mathrm{100}$$

(2)

where $\mathrm{C}\left(\mathrm{b}\mathrm{p}\right)$ is the chroma value of the sample surface before pollution (totally clean), $\mathrm{C}\left(\mathrm{0}\right)$ represents the color saturation of the sample after pollution (before starting irradiation), and $\mathrm{C}\left(\mathrm{t}\right)$ is the color saturation after exposure to light over time.

Besides photocatalytic efficiency, it was also calculated the color variation $\left(∆\mathrm{E}\right)$ according to Equation (3), which works as a perceptibility factor and indicates the value of the surface color difference [14,20]:

$${∆\mathrm{E}=\left[{\left({∆\mathrm{L}}^{\ast }\right)}^{\mathrm{2}}+{\left({∆\mathrm{a}}^{\ast }\right)}^{\mathrm{2}}+{\left({∆\mathrm{b}}^{\ast }\right)}^{\mathrm{2}}\right]}^{\mathrm{0.5}}$$

(3)

where ${∆\mathrm{L}}^{\ast }$, ${∆\mathrm{a}}^{\ast }$, and ${∆\mathrm{b}}^{\ast }$ denote the difference of ${\mathrm{L}}^{\ast }$, ${\mathrm{a}}^{\ast }$ and ${\mathrm{b}}^{\ast }$ values at the known irradiation period and the polluted sample (before starting irradiation), respectively.

$\mathrm{P}\mathrm{E}\mathrm{c}\mathrm{i}\mathrm{e}\mathrm{l}\mathrm{a}\mathrm{b}$ and $∆\mathrm{E}$ provide colorimetric indications of the color change on the surface of the specimens. Higher values of $\mathrm{P}\mathrm{E}\mathrm{c}\mathrm{i}\mathrm{e}\mathrm{l}\mathrm{a}\mathrm{b}$ and $∆\mathrm{E}$ indicate a greater removal of the colored pollutant from the surface of the specimen.

Diffuse Reflectance Spectroscopy

The measurement of diffuse reflectance spectroscopy is a suitable technique for color assessment due to its capability to provide an accurate and representative evaluation of surface reflectance. The integrating sphere, included in the spectrophotometer, is designed to capture diffusely reflected light from an object in multiple directions, a crucial feature when measuring the color of materials. Furthermore, conducting measurements across a broad range of wavelengths allows for obtaining detailed information about the spectral properties of the analyzed process, which is essential for an in-depth analysis of color and optical characteristics [29,30,31].

The spectral measurements were conducted by directing light onto the sample surfaces using a spectrophotometer (Shimadzu UV-310 PC) (Kyoto, Japan) equipped with an integrating sphere, where the wavelength range to acquire diffuse reflectance extended from 250 nm to 850 nm and the characteristic reflectance peaks for each pollutant model were monitored. Their values can vary depending on the experimental conditions, environment, substrate, concentration of the solution, and equipment [32,33].

For the samples polluted with Methylene Blue, a main peak was observed at $\lambda =600 \mathrm{n}\mathrm{m}$ whereas for the ones polluted with Rhodamine B solution, the peak was detected at $554$ nm. In order to assess the photocatalytic efficiency through diffuse reflectance (PEdr), the reflectance peaks of the polluted samples were monitored as a function of time. It was calculated via Equation (4), adapted from Rocha Segundo et al. [34] and Carneiro et al. [35].

$$\mathrm{P}\mathrm{E}\mathrm{d}\mathrm{r}\left(\mathrm{\%}\right)=\frac{\mathrm{R}\left(\mathrm{t}\right)-\mathrm{R}\left(\mathrm{0}\right)}{\mathrm{R}\left(\mathrm{b}\mathrm{p}\right)-\mathrm{R}\left(\mathrm{0}\right)}\times \mathrm{100}$$

(4)

In Equation (4), $\mathrm{R}\left(\mathrm{b}\mathrm{p}\right)$ is the reflectance value of the sample before pollution (totally clean), $\mathrm{R}\left(\mathrm{0}\right)$ is the reflectance after pollution (before starting irradiation), and $\mathrm{R}\left(\mathrm{t}\right)$ is the reflectance after light exposure along the time, according to Figure 3.

More expressive peaks are indicative of the intact dye molecule in the electromagnetic spectrum. Over the irradiation time, the greater the removal of the pollutant from the surface of the specimens, the closer the spectra tended to be to the spectra of the specimen before the pollution.

Digital Image Processing

Digital image processing was carried out on the pictures captured and cropped in the visual analysis technique. The entire surface area of each photographed sample was analyzed, and the number of pixels of each image was the same, as the images were cropped to the same sample size. Given that all the photographs were taken under the same lighting conditions, distance from camera to samples, and camera settings, a grayscale histogram was generated for each image utilizing the ImageJ software (version 1.51k). In this histogram, each grayscale value ranging from 0 to 255 (0 = total black and 255 = total white) contained the frequency count of pixels in the image exhibiting that particular shade [16,18,19]. From the histogram, the average grayscale value (AL) was calculated according to Equation (5):

$$\mathrm{A}\mathrm{L}=\frac{\mathrm{\Sigma }(\mathrm{L} \times \mathrm{N}\mathrm{p}\mathrm{i}\mathrm{x})}{\mathrm{\Sigma } \left(\mathrm{N}\mathrm{p}\mathrm{i}\mathrm{x}\right)}$$

(5)

where L is the grayscale value ranging from 0 to 255, and (Npix) is the count of pixels for each value of L in the image.

In order to monitor the self-cleaning ability, the surface discoloration of the samples was observed by the shift in AL values along irradiation times. Once the unpolluted samples present a white color, pollutant removal in the decontamination process is detectable through the grayscale. Therefore, the photocatalytic efficiency through digital image processing (PEdip) can be expressed through Equation (6):

$$\mathrm{P}\mathrm{E}\mathrm{d}\mathrm{i}\mathrm{p} (\%)=\frac{\mathrm{A}\mathrm{L}\left(\mathrm{t}\right)-\mathrm{A}\mathrm{L}\left(\mathrm{0}\right)}{\mathrm{A}\mathrm{L}\left(\mathrm{b}\mathrm{p}\right)-\mathrm{A}\mathrm{L}\left(\mathrm{0}\right)} \times \mathrm{100}$$

(6)

where $\mathrm{A}\mathrm{L}\left(\mathrm{t}\right)$ is the average grayscale value at a specific time of irradiation, $\mathrm{A}\mathrm{L}\left(\mathrm{0}\right)$ is the average grayscale value after pollution (before starting irradiation), and $\mathrm{A}\mathrm{L}\left(\mathrm{b}\mathrm{p}\right)$ is the average grayscale value of the sample before pollution (totally clean), according to Figure 4.

As long as pollutants are photocatalyzed from sample surfaces, both the overall histogram and the average grayscale value undergo a shift to the right (i.e., to higher L values). Higher L values correspond to a brighter color on the grayscale and mean that the polluted samples are approaching the L values exhibited by the samples before pollution, which were initially white (self-cleaning effect) [18].

2.2.5. Chemical and Morphological Assessment

Fourier Transform Infrared Spectroscopy

The Fourier Transform Infrared Spectroscopy (FTIR) test was carried out to analyze the chemical bonds from the different RM paints. A Shimadzu IR-Prestige-21 spectrometer (Kyoto, Japan) was used with the spectral range spanning from $400 {\mathrm{cm}}^{-1}$ to $4000 {\mathrm{c}\mathrm{m}}^{-1}$. The spectra analysis and peak identification of different samples revealed relevant information concerning the chemical composition of the commercial paint, as well as the main changes caused by the functionalization process [36], such as the increased intensity of the peaks related to the material incorporated into the paint.

Scanning Electron Microscopy and Energy Dispersive Spectroscopy

The morphological and chemical characterization (elemental composition) of the RM paint samples (before and after functionalization) was carried out by using a Scanning Electron Microscope (SEM) coupled to Energy Dispersive X-ray Spectroscopy (EDS). SEM enabled the visualization of surface and cross-section morphology and microstructure of the samples at high magnification, providing detailed structural information. A FEI Nova 200 (FEG/SEM) (Houston, TX, USA) was used. In turn, EDS enabled elemental analysis by detecting characteristic X-rays emitted from the sample upon electron beam interaction. A Pegasus X4M (EDS/EBSD) (Austin, TX, USA) was used.

X-Ray Diffraction

X-ray diffraction (XRD) analysis was carried out to assess the structural characteristics of the RM paint samples. The XRD allowed the determination of the crystalline phase of the semiconductor material incorporated in the RM paint samples. A CuKα source from a Philips PW 1710 X-ray diffractometer (Billerica, MA, USA) was used. Besides determining which crystalline phases are present in the RM paint samples, the Scherrer Equation (7) was also used to calculate their crystallite sizes (d) [37,38].

$$\mathrm{d}=\frac{\mathrm{0.9}\mathrm{\lambda }}{\mathrm{F}\mathrm{W}\mathrm{H}\mathrm{M}\times \cos {\mathrm{\theta }}_{\mathrm{p}}}$$

(7)

where $\lambda$ is the X-ray wavelength; FWHM is the full width at half maximum height for the peak at (2θ); and ${\mathrm{\theta }}_{\mathrm{p}}$ is the Bragg’s diffraction angle for the same peak position.

3. Results and Discussion

3.1. Visual Analysis

The visual analysis technique (Figure 5) allowed the qualitative assessment of the discoloration results of the two pollutant models applied to RM paint with different nano-TiO₂ contents over the irradiation periods. Although it is not a quantitative technique, the grouping of photographs taken under the same conditions enabled the visual identification of the relationship between the increase in the amount of TiO₂ and the increase in the discoloration of the pollutants (self-cleaning capacity), that occurs more intensively for the samples 2% and 3%.

Discoloration in the specimens polluted with Methylene Blue became noticeable within the first few minutes of irradiation. After 180 min of irradiation, the samples 2% and 3% showed significantly more discoloration than the RP. However, even for the non-functionalized samples, some discoloration is evident due to the lower stability of the Methylene Blue solution, which degrades in the presence of light, a phenomenon commonly observed in dyes: photolysis. After 2880 min of light irradiation, all the specimens showed drastic discoloration, suggesting that this pollutant might be suitable for a rapid assessment test, such as 2 or 3 h, when the gains in the self-cleaning capacity of the functionalized paints compared to the reference are noticeable.

Specimens containing Rhodamine B as a pollutant model showed slightly different behavior. Up to 180 min of irradiation, there were still no significant differences in terms of discoloration between the functionalized and non-functionalized specimens, or between the irradiated and non-irradiated ones. This pollutant model proved to be stable to degradation in the presence of light. As the light irradiation time increased and up to 2880 min, the samples with 2% and 3% of nano-TiO₂ showed exceptionally pronounced discoloration compared to the samples with lower nano-TiO₂ content or the RP, revealing that the incorporation of nano-TiO₂ into RM paints led to the self-cleaning capacity.

3.2. CIELAB Color Space

The CIELAB color coordinates of the RM paints functionalized with different amounts of nano-TiO₂ were determined. Subsequently, the samples were contaminated with aqueous dye model solutions, and the CIELAB color coordinates were measured before and after exposure to simulated sunlight. The colorimetry test enabled the determination of lightness (L*) comparison between the functionalized paints and RP (Figure 6), as well as the calculation of PEcielab and ΔE values for the samples contaminated with Methylene Blue (Figure 7) and Rhodamine B (Figure 8).

The lightness, corresponding to the average L* coordinate of the eight samples per formulation, remained practically constant after the incorporation of TiO₂ up to 3%. The RP showed an L* value of 96.29, while the RM paints with 0.25%, 0.5%, 1%, 2%, and 3% TiO₂ exhibited values of 96.04, 96.10, 96.28, 96.36, and 96.26, respectively, that is, variations below 0.3 units. Such differences indicate that nano-TiO₂ addition did not significantly affect the paint whiteness and preserved the original visual appearance and optical properties of the RP.

For samples polluted with Methylene Blue, a pronounced increase in PEcielab and $\Delta \mathrm{E}$ was observed in the early irradiation intervals. In the first few minutes, it was evident that samples with higher concentrations of nano-TiO₂ exhibited a remarkable self-cleaning capacity. After 180 min of light irradiation, the specimens 1%, 2%, and 3% presented discoloration rates of 74%, 75%, and 80%, respectively, while the reference paint (RP) reached a discoloration rate of 63%. Throughout all irradiation periods, samples with higher nano-TiO₂ concentrations consistently showed the highest discoloration rates and, consequently, color variation. From 1440 min of irradiation, specimens with different amounts of nano-TiO₂ incorporation achieved a high and similar discoloration rate above 90%, and color change above 16. This occurs as a result of extended exposure of the samples to light and the low stability of the dye, leading to a photolysis phenomenon and its degradation. At the end of 2880 min of irradiation, the specimens exhibited a discoloration rate above 94%, suggesting almost complete removal of the pollutant from the sample surface.

Non-irradiated samples (negative control) showed substantially lower discoloration and color variation values compared to irradiated ones. At the end of 2880 min, discoloration values were around 35%, with a color difference below 8. These values were similar regardless of the percentage of semiconductors incorporated into the paint (Figure 7). This discoloration and color change, even in darkness, indicated the instability and ease of degradation over time of the pollutant, which was accentuated in the presence of light.

Despite the instability of Methylene Blue as a pollutant model, it proved to be an effective model for accelerated analyses, as within a few hours of irradiation testing, the degradation of this dye could be observed, providing evidence of the photocatalytic capacity of the substrates.

For the samples polluted with Rhodamine B (Figure 8), unlike those contaminated with Methylene Blue, during the first 180 min of exposure to light irradiation, low values of PEcielab and $\Delta \mathrm{E}$ were observed. However, in the subsequent irradiation periods, the RM paint with higher concentrations of nano-TiO₂ revealed the highest rates of photocatalytic efficiency and color change.

Upon reaching 2880 min of light irradiation, the samples with 2% and 3% of nano-TiO₂ exhibited a significant PEcielab of 50% and 72%, respectively, when compared to the RP, with only 22.7%. The color variation resulting from the removal of the pollutant from the surface of the samples supported the discoloration results: the samples 2% and 3% also demonstrated the most significant color variation, registering values of 22.35 and 30.12, respectively, in contrast to the result referring to the RP, which was 9.39.

Both the obtained results for Methylene Blue and Rhodamine B indicated increased photocatalytic efficiency after nano-TiO₂ incorporation into RM paint. However, Rhodamine B proved to be a notably stable pollutant model in this study, evidenced by low discoloration indices in irradiated non-functionalized samples, as well as in samples kept in the dark (close to zero). It confirms that Rhodamine B is particularly suitable for the analysis of self-cleaning capacity over extended periods of light irradiation, experiencing less influence from photolysis and presenting gradual degradation over time.

Although pollutant models such as Methylene Blue and Rhodamine B are commonly used to evaluate photocatalytic and self-cleaning performance under laboratory conditions, they do not fully represent the complexity of real conditions on RM. In field environments, contaminants such as soot, oils, and rubber dust exhibit different removal dynamics compared to aqueous dye models. Real pollutants tend to undergo a slower photo-oxidation process due to their low water solubility, the reduced contact with the active TiO₂ surface, and the presence of inorganic fractions in their composition. In addition, these contaminants adhere mechanically to the RM surface, making their removal dependent not only on photocatalytic activity but also on the wettability properties.

3.3. Diffuse Reflectance Spectroscopy

To improve the assessment of the self-cleaning capacity of RM paints, photocatalytic efficiency can be evaluated through diffuse reflectance measurements. The results were obtained from the analysis of the characteristic peak values of each pollutant model used. Figure 9 shows the photocatalytic efficiency of the RM paints with varying percentages of nano-TiO₂ considering each pollutant model. Regardless of the pollutant model chosen, greater photocatalytic efficiency was achieved for the samples 2% and 3%.

Considering the suitability of Methylene Blue for short tests, gains in the photocatalytic efficiency of the functionalized specimens can be noticed when analyzing the first minutes of light exposure. The samples 2% and 3% demonstrated efficiencies of 59% and 64%, respectively, compared to the RP, which presented 46% at 360 min of light irradiation. The samples were kept in dark conditions (PEcielab ranging from 16% to 24%) and all the irradiated samples (PEcielab greater than 70%) corroborated the instability of the Methylene Blue dye. However, the increase in photocatalytic efficiency with increasing incorporation of nano-TiO₂ into the RM was also evident with this model pollutant.

For the samples polluted with Rhodamine B, the results were also in line with the analysis of the CIELAB color space. There was a gradual increase in photocatalytic efficiency over the irradiation periods, with a significant difference in the final efficiencies of the samples 1%, 2% and 3%. These samples achieved photocatalytic efficiencies of 35%, 51% and 68%, respectively, after 2880 min of light irradiation compared to the RP, which only presented 12%. Even with a more stable pollutant model, the increased self-cleaning capacity of the RM paint with greater nano-TiO₂ incorporation was verified under these conditions.

3.4. Digital Image Processing

Digital image processing emerged as a suitable technique for evaluating the self-cleaning capability in RM, standing out for providing a comprehensive analysis of the entire element surface. The use of the greyscale histogram in RM paints proved to be an effective method, as the unpolluted samples are white, and the deposition of pollutants and subsequent self-cleaning of the surface of these specimens could be detected by analyzing this parameter.

The similarity between the pollutant degradation graphs after daylight simulation exposure, compared to those observed in the other two previous techniques—CIELAB color space and diffuse reflectance—reaffirmed the validity of these methods for studying self-cleaning ability in these specific applications. Analyzing Figure 10, it is possible to observe that the samples polluted with Methylene Blue presented fast depollution ratios in the first few minutes of light irradiation. However, samples polluted with Rhodamine B showed slower and more gradual pollutant degradation. After 2880 min of exposure to irradiation, the samples 2% and 3% reached 22% and 31% of PEdip, respectively, contrasting with the RP sample that only presented a value of 4%. Regardless of the type of pollutant used, it could be observed that samples with a higher concentration of nano-TiO₂ tended to have higher dye degradation rates.

Although the shapes of the photocatalytic efficiency curves appear similar for the three techniques, digital image processing, especially in samples polluted with Rhodamine B, indicated lower degradation rates compared to those obtained by the previous methods. This can be explained by the fact that digital image processing captured information from the entire surface of the samples, while measurements performed with spectrophotometers were carried out at specific points on the surface of the samples. This feature confirmed digital image processing as an extremely advantageous technique, as it offers a more comprehensive view of the samples and avoids the need to use specialized spectrophotometric equipment, some of which are very expensive.

3.5. Fourier Transform Infrared Spectroscopy

The analysis of the FTIR spectrum (Figure 11) provided the identification of three main components in the commercial acrylic paint under consideration: the binder, filler, and pigment. The results revealed the acrylic binder with poly(isobutyl methacrylate) (PIBMA) as the polymer, calcite as the filler, and TiO₂ as the pigment. PIBMA is a type of polymer belonging to the family of methacrylate polymers, widely employed in various applications as a binder, coating, and sealant [39,40].

Characteristic peaks observed for PIBMA include the stretching vibrations of the C-H bond at 2955 and 2878 cm⁻¹. The vibration of the ester bond (C=O) in the methacrylate groups was identified at 1728 cm⁻¹. Skeletal vibrations of the C-O/C-C bonds in the acrylic binder were observed in the range of 900–1250 cm⁻¹. The literature mentions stretching of the ether group in the ester at 1142 cm⁻¹ [9,41,42]. Regarding the filler, calcite was identified through distinct peaks at 1396 and 872 cm⁻¹ [41]. The pigment TiO₂ is responsible for the appearance of peaks at 517 and 417 cm⁻¹, attributed to the Ti-O stretching vibrations, typically observed for crystalline TiO₂ within the 400–700 cm⁻¹ spectral region [41,43,44].

It should be highlighted that both RP, 2% and 3% samples presented very similar spectra. The difference lies in the fact that the peaks at 517 and 417 cm⁻¹ show greater intensity due to the functionalization process, indicating the incorporation of additional semiconductor material in the RM paint. TiO₂ was also already present in the RP because it is commonly used as a pigment in white paints [45]. However, identifying the crystalline phases of TiO₂ before and after functionalization is an essential task in order to relate it to the increase in the self-cleaning capacity, as the anatase crystalline phase is the one that confers the greatest photocatalytic efficiency [38].

3.6. Scanning Electron Microscopy and Energy Dispersive Spectroscopy

The RP, 2%, and 3% samples were subjected to SEM and EDS analysis in order to observe their morphological and chemical changes caused by the functionalization process and to correlate them with the achieved self-cleaning capacity.

In Figure 12, the SEM micrographs provided a surface visualization of the paints, allowing the identification of the solids present, such as the filler materials and pigments, immersed in the acrylic polymer matrix. It is noteworthy that in the functionalized paint samples there was a significant additional number of particles resulting from the mass incorporation of nano-TiO₂.

The EDS spectra (Figure 12) and

Table 3. Chemical analysis by EDS technique.

Sample	Chemical Composition Wt (%)
	C	O	Ca	Ti
RP	35.88	44.34	13.98	5.80
2%	35.26	44.80	12.74	7.20
3%	34.00	44.63	12.96	8.41

, which detailed the chemical composition of RP, 2%, and 3% samples, corroborated the results from SEM micrographs. They presented the most representative chemical elements in the composition of the paint, such as calcium (Ca) derived from the filler material, calcite (CaCO₃); titanium (Ti) from the titanium dioxide used as pigment and in the functionalization process; carbon (C) from the acrylic binding agent; and oxygen present in both paint components. As expected, the incorporation of nano-TiO₂ into the samples led to an increase in the intensity of the Ti peaks in the EDS spectra.

Table 3 also showed an increase in the amount of Ti by mass in the functionalized samples, from 5.8% in the chemical composition of the RP to 7.2% and 8.41% after functionalization (samples 2% and 3%, respectively). This reinforced the evidence that TiO₂ was already used as a pigment in the RP, but its mass incorporation in the crystalline form of anatase phase led to a notable increase in the chemical composition. This increase in the amount of TiO₂ due to the incorporation of nano-TiO₂ (anatase phase) explained the achievement of the self-cleaning capacity of the functionalized samples.

Figure 13 showed additional SEM micrographs, including surface and cross-sectional analyses. In these micrographs, the components of the paint, such as the filler (calcite) and the pigment (titanium dioxide), are noticeable. The calcite exhibited larger and non-uniform sizes and irregular and angular morphology, especially when observed in cross-sectional view. Conversely, the TiO₂ particles are rounder and smaller than calcite. In the SEM micrographs of the functionalized samples, even smaller TiO₂ particles can be observed, approximately 25 nm in size, resulting from the functionalization process, where it was already known that the incorporated TiO₂ particles had a diameter of around $23–28$ nm.

SEM and EDS analyses confirmed that the nano-TiO₂ is embedded within the polymeric matrix of the paint. This morphological evidence suggests that, under the particle immobilization technique employed, the semiconductor presents a low probability of particle release under normal service conditions. However, prolonged exposure to traffic and weathering may gradually erode the surface layer, potentially reducing the photocatalytic activity or TiO₂ accessibility, as well as leading to the possible release of this material into the environment. Complementary studies involving mechanical abrasion and leaching tests are therefore essential to ensure both the durability and the environmental safety of the functionalized RM paints. These evaluations are consistent with international regulatory efforts concerning photocatalytic surfaces, which, although not yet regulated by a specific standard for TiO₂ particle release, can be conducted through the adaptation of existing standards such as ISO 2812-2, ISO 12457-3, and EN 16637-2.

3.7. X-Ray Diffraction

The RP sample, as observed in SEM, EDS, and FTIR tests, already exhibited the TiO₂ material in its composition, which came from its use as a pigment. However, as crucial as confirming the increased concentration of TiO₂ in the RM paint (after functionalization) is identifying the presence of its crystalline phases.

Similarly to the other techniques for assessing chemical and morphological characteristics, the presence of TiO₂ in the paint was also confirmed by the XRD test. TiO₂ is commonly found in three crystalline phases: anatase, rutile, and brookite, but anatase is the one that displays the best photocatalytic efficiency [46,47]. Through the XRD analysis (Figure 14), new crystalline phases (anatase) could be observed resulting from the functionalization process. However, only peaks associated with the rutile crystalline phase were identified in the RP sample, indicating that rutile is the crystalline phase used as a pigment.

The semiconductor (nano-TiO₂) used as a photocatalytic material in the functionalization process was composed of 80% anatase and 20% rutile. Consequently, after the mass incorporation of the semiconductor material, anatase peaks, namely A(101) and A(200), were identified in the X-ray diffractogram, demonstrating the incorporation of TiO₂ in the desired crystalline phase. This explains the increased self-cleaning capacity of these paints after the functionalization process. The other peaks related to the RP sample materials remained unchanged.

However, it should be noticed that the anatase phase of TiO₂, while responsible for the enhanced photocatalytic activity, is also known to promote the degradation of organic binders under UV exposure, especially in higher concentrations. The reactive oxygen species generated during photocatalysis may attack polymer chains and functional groups, leading to oxidation and mass loss in the polymeric matrix of the paint [23,48].

For the crystallite size calculation (

Table 4. Crystallite size of the crystalline phases of the RM, 2% and 3% samples.

Sample	Crystallite Size (nm)
	CaCO3	TiO2, Anatase	TiO2, Rutile
RP	39.3502	-	47.6597
2%	35.4689	23.7546	45.1217
3%	38.3040	22.3782	40.5887

), the C(104) and C(018) diffraction peaks were used for calcite, R(110) and R(211) for rutile, and A(101) and A(200) for anatase. The crystallite size of calcite ranged from 35.4689 to 39.3502 nm, for anatase from 22.3782 to 23.7546 nm, while for the rutile crystalline phase it ranged from 40.5887 to 47.6597 nm. As expected, the anatase phase presented a crystallite size considerably smaller than rutile, since the TiO₂ material incorporated in the samples was in the anatase phase, presenting a nanometric dimension with a particle size of around 23–28 nm.

The higher variations in crystallite sizes of calcite and rutile point to less uniformity of these materials in the manufacturing process of the RP. This is patent, as seen in SEM micrographs, which show non-uniformity in grain size and shape, affecting the variation in crystallite size calculations. In the case of anatase (nano-TiO₂), there was a greater uniformity in the characteristics of the semiconductor material incorporated during the functionalization process. Consequently, there was less variation in the calculated grain size for this crystalline phase.

4. Conclusions

This work was devoted to the development of self-cleaning capacity in RM paints through the incorporation of semiconductors (nano-TiO₂) with photocatalytic properties. Following the functionalization process, the study focused on evaluating and quantifying the paints’ self-cleaning ability, as well as analyzing the chemical and morphological changes in the functionalized samples compared to the reference sample. The following conclusions can be highlighted:

• The grouping of images from the visual analysis technique displayed that over the light irradiation periods, the samples functionalized with higher amounts of TiO₂ showed greater discoloration.
• CIELAB color space revealed that the samples 2% and 3% exhibited higher color variation and photocatalytic efficiency, regardless of pollutant model.
• Diffuse reflectance spectroscopy confirmed photocatalytic efficiency for the samples 2% and 3% up to 4.25 and 5.7 times higher in comparison with the reference sample, respectively.
• Digital image processing detected pollutant degradation for the samples 2% and 3% up to 4.5 and 7.7 times higher than the reference sample, respectively.
• CIELAB color space and diffuse reflectance spectroscopy provided measurements on specific points of the sample surface, while digital image processing offered comprehensive pollutant degradation information from the entire specimen surface.
• Methylene Blue demonstrated high degradation rates in the initial minutes of the irradiation, revealing a suitable pollutant model for short-term assessment tests.
• Rhodamine B proved to be a stable dye and a suitable pollutant model for long-term testing, with significantly reduced and similar discoloration rates for both samples non-functionalized or non-irradiated.
• The combined analysis of the FTIR, SEM and EDS tests revealed the presence of PIBMA in the binder, calcite as the filler, and TiO₂ as the pigment of the commercial paint, as well as the increase in TiO₂ in the chemical and morphological evaluation of the paint after functionalization.
• XRD analysis evidenced the presence of anatase crystalline phase of TiO₂ exclusively in functionalized paints.
• Based on these results, it was demonstrated that the mass incorporation of nano-TiO₂ into the RM paint conferred self-cleaning capacity. Furthermore, chemical and morphological analysis highlighted the increase in TiO₂ in the chemical composition, as well as the presence of TiO₂ in the desired crystalline state, anatase, confirmed the successful functionalization process.

The incorporation of self-cleaning properties in RM paints represents promising practical implications. Maintaining visibility and contrast of RM for longer periods can reduce the frequency of repainting, thus lowering maintenance costs and minimizing traffic disruption. Continuous preservation of optical performance also contributes to improved road safety. From an environmental perspective, fewer interventions and less material use reduce the emission of volatile compounds, promoting more sustainable road infrastructure.

Nevertheless, the photocatalytic activity that enables self-cleaning may induce unwanted side effects on the paint matrix. The anatase phase of TiO₂, especially in high concentrations, while enhancing the degradation of pollutants, can simultaneously promote photooxidation of the acrylic binder under UV exposure. This dual behavior (photocatalytic efficiency and polymer degradation) represents a critical trade-off in the development of self-cleaning RM paints. Therefore, comprehensive durability and accelerated aging tests are essential to assess the balance between photocatalytic performance and the long-term stability of the acrylic matrix after functionalization.

Future work arises from these results. This includes testing a broader range of anatase TiO₂ concentrations, analyzing the stability and integrity of the paint’s polymeric matrix after functionalization, potential changes in TiO₂ accessibility, the occurrence of surface chalking, applying the technology to other types of RM paints, and exploring possible improvements in Qd and RL when it is working as a system (paint and GB). Further studies should also examine different particle immobilization techniques and aging effects, as semiconductors have shown anti-aging benefits in other binders. In addition, future research should evaluate how the incorporation of anatase TiO₂ nanoparticles may affect the physical and mechanical properties of the paint. Furthermore, it is essential to evaluate the photocatalytic performance under more realistic conditions. Although dyes like Methylene Blue and Rhodamine B are common in laboratory assessments, they do not fully replicate real-world pollutants. Thus, future work should involve more representative and stable contaminants, including not only real field pollutants such as oils, rubber residues, and environmental grime, but also other types of organic dyes (e.g., oily-food colorants, solvent-based dyes), in order to assess the self-cleaning capacity under more realistic service conditions.