TiO₂ Supported on Kaolinite via Sol–Gel Method for Thermal Stability of Photoactivity in Ceramic Tile Produced by Single-Firing Process

Abstract

Anatase is well known for its photocatalytic properties. However, it can be irreversibly transformed into rutile at temperatures above 600–850 °C. This is a major limitation for ceramic tiles with self-cleaning properties, which are usually single-fired at 1100–1250 °C. To avoid this issue, functionalized tiles are often produced by double firing, where the second firing stays below 850 °C. Supporting TiO₂ on kaolinite helps to stabilize the anatase phase even at temperatures above 850 °C. In this study, a photocatalytic coating was specially developed to be suitable for the single-firing ceramic tile process. TiO₂ and TiO₂ with Nb₂O₅ (from 0 to 12 wt.%) were supported on kaolinite. This material was mixed with a glass frit to create a surface texture typical of ceramic tiles. The coated tiles were single-fired at 1185 °C. The self-cleaning performance was evaluated using contact angle (CA) measurements and methylene blue (MB) degradation under UV-A light, on both unpolished and polished surfaces. The polished sample containing 12 wt.% TiO₂ showed the best photocatalytic activity: it degraded 57% of MB and the contact angle decreased from 64° to 30° after UV-A exposure. XPS, FTIR, and FEG-SEM analyses confirmed the effective presence of TiO₂. The results demonstrate that kaolinite-supported TiO₂ is a promising approach for producing self-cleaning ceramic tiles using a single-firing process.

1. Introduction

Titanium dioxide (TiO₂) exists in three polymorphic structures: anatase, rutile, and brookite. Anatase and rutile are particularly significant in technological applications [1]. TiO₂ is the most widely used white pigment in paints, enhancing coverage, durability, brightness, and opacity [2]. Anatase, in particular, is widely used in photovoltaic, photocatalytic, photo/electrochromic, and sensor applications [3]. Compared to other photocatalytic materials, TiO₂ stands out for its relatively low cost, stability against photocorrosion, and low toxicity [4,5,6].

Photocatalytic materials are produced by various industries, as highlighted by the Iberian Photocatalysis Association (AIF), which connects stakeholders in Spain and Portugal [7]. Italcementi Group, Kronos Worldwide, Tronox Holdings, and Chemours are examples of companies exploring this market [8,9]. The commercialization of TiO₂-based photocatalytic materials started in the mid-1990s [10], primarily to create self-cleaning surfaces that counteract environmental pollution and pathogenic microorganisms [5,10]. Self-cleaning coatings are hydrophilic (water-attractive) or hydrophobic (water-repellent), each reducing maintenance costs and improving surface cleanliness [11,12]. Effective self-cleaning requires anatase, which provides photocatalytic action and super-hydrophilicity, although not always simultaneously [13,14]. For external surfaces, super-hydrophilicity is essential, as it enables rainwater to form a thin film that removes dirt [15,16].

Anatase exhibits superior photocatalytic properties, with a band gap of 3.0–3.2 eV, making it UV-active [17,18]. However, it is thermally stable only below 600 °C; above this temperature, it irreversibly transforms into rutile, significantly reducing photocatalytic efficiency [6,17]. This poses challenges for the functionalization of ceramic tiles within the regular single-firing process due to the temperature required, 1100–1250 °C. The current method to produce a photoactive ceramic tile is to apply the photocatalyst layer over fired tile and then a second firing at lower temperatures, ~850 °C, to improve the adhesion and other mechanical properties. This procedure increases costs and limits market availability.

Studies by Reidy et al. [19], Chen et al. [20], and Da Silva et al. [15] indicate that doping can inhibit crystal growth and, consequently, the anatase-to-rutile phase transformation (ART). Common dopants include transition metals such as copper, silver, gold, and platinum, which reduce the recombination of photogenerated electrons and holes [16]. These metals also decrease the band gap of TiO₂, enabling activity under solar light or UV light and enhancing the photocatalytic efficiency of anatase [13]. According to Khlyustova et al. [21], trivalent metals such as aluminum (Al) and niobium (Nb) are effective dopants for enhancing TiO₂ ceramic coatings’ properties and surface functionalization. Fe/Nb-based composite photocatalysts synthesized via coprecipitation exhibit high photosensitivity across the visible solar spectrum and maintain strong photocatalytic activity utilizing both UV and visible light [22]. Niobium, in particular, plays a key role in catalysis, with small additions significantly improving catalytic activity, selectivity, and chemical stability [23,24,25].

TiO₂ can be synthesized through various methods, including hydrothermal synthesis [26,27], deposition techniques such as sputtering and chemical vapor deposition [13], coprecipitation [28], and hydrolytic sol–gel [29,30]. Among these, the sol–gel method is widely favored to produce TiO₂ in powder form [31]. This technique utilizes titanium isopropoxide as a precursor, enabling uniform distribution of titania on kaolinite particles [32]. It also operates at low temperatures, offering precise control over the synthesis reaction and the photocatalyst morphology [33].

Feltrin et al. [34] investigated the use of quartz, cristobalite, and amorphous silica as stabilizers to delay ART at high temperatures. Quartz was the most effective, achieving only 50% conversion to rutile, while cristobalite and amorphous silica resulted in 96% and 95% conversion, respectively. Titania samples stabilized with quartz also exhibited higher relative densities. Other studies have explored the combined use of alumina and quartz as stabilizers for pure anatase [35]. The addition of these particles reduced anatase–anatase contact points, decreasing nucleation sites and enhancing the thermal stability of anatase. The use of a mixture of 9.2% alumina and 25.4% quartz inhibited ART at 1100 °C, demonstrating a greater stabilizing effect than quartz alone. Additionally, high heating rates were found to further suppress ART [36].

In the work of Chong et al. [37], unsatisfactory results were found when using the sol–gel method with titanium (IV) butoxide as a precursor. Despite anatase being the predominant phase, titania/kaolinite prepared with 0.28 M HNO₃ achieved only 30% Congo Red removal at a calcination temperature of 600 °C. Similarly, Feltrin [35] reported that ART inhibition did not enhance photocatalytic activity compared to rutile, highlighting the need for alternative approaches to improve photocatalytic efficiency [37].

In contrast, Barbosa et al. [29] achieved promising results at higher calcination temperatures. Using titanium (IV) isopropoxide as a precursor, the material was synthesized at room temperature, washed, centrifuged, dried, and calcined at 400 °C, 700 °C, and 1000 °C for 24 h at each temperature, demonstrating the potential to optimize photocatalytic performance through controlled calcination processes [29]. The photocatalytic activity of the titania/kaolinite samples was evaluated through the degradation of methylene blue (MB) and methyl orange (MO). All samples showed good performance, degrading the dye chromophore structures within 1 h. In comparison, the Degussa P25 TiO₂ sample achieved only 10% degradation with methyl orange and 50% with MB [29]. However, the calcination conditions used in this study differ from those employed in the production of ceramic tiles, which involve temperatures higher than 1000 °C and holding time less than 10 min at the maximum firing temperature. Dassoler et al. [38] have reported a good level of photocatalytic activity in samples submitted to 1185 °C and holding time ~5 min at the maximum firing temperature. However, the surface texture was not compatible with that expected for ceramic tile applications.

To preserve the photocatalytic activity of TiO₂, proper immobilization and thermal treatment are essential. Anchoring the particles on a substrate is a viable alternative for industrial processes, considering technology and safety. The scalability of TiO₂ production depends on factors such as cost, effectiveness, and the safe use of chemicals, which are important to ensure the feasibility and sustainability of large-scale applications [39].

This study aims to address a gap in the literature by hypothesizing that the combination of TiO₂ supported on kaolinite, or TiO₂ doped with Nb₂O₅ on kaolinite, applied to ceramic tiles in their raw form and subjected to regular firing cycle, will result in a product that meets both functionality and cost-efficiency requirements. The use of Nb₂O₅ as a dopant offers several advantages, including reducing the surface energy of TiO₂, stabilizing the grains, delaying the ART, and increasing the specific surface area, thereby creating more active sites [40]. Additionally, this study aims to provide technical insights that can support the expansion of photocatalytic ceramic coatings into the market by exploring a functional, cost-effective approach capable of enabling commercial-scale production.

2. Materials and Methods

This research was divided into three main steps: (1) synthesis of the photocatalyst; (2) ceramic tile application; (3) photocatalytic activity characterization. The materials, experimental design, and characterization techniques are described below.

2.1. Materials

For the photocatalyst synthesis, the following materials were used: Kaolin, Caulisa Minerals Trading and Processing Ltd.a. Juazeirinho, Paraíba, Brazil; Absolute ethyl alcohol, Neon. 99.8% P.A, São Paulo, Brazil; Aceticacid, LAFAN Fine Chemistry Ltd.a. São Paulo, Brazil. Glacial P. A; Titanium (IV) isopropoxide, Aldrich 97%, Darmstadt, Germany, Niobium n-butoxide, Alfa Aesar 99%; P25, Degussa, Essen, Germany.

For the ceramic tile application, 20 × 20 cm² unfired glazed porcelain tile pieces were supplied by a Brazilian ceramic tile company located in the Santa Catarina state. The same company has also supplied a crystal-clear screen-printing base, usually applied as a protective layer over the inkjet-printed surface of the tile composed of (Al₂O₃, SiO₂, CaO, MgO, K₂O, ZnO).

2.2. Experimental Procedure

Kaolin samples were subjected to high-energy milling using a attritor mill (Netzsch, PE075, Selb, Germany). The milling process was conducted in a 500 cm³ jar containing 140 g of zirconia spheres with a diameter of 0.4 mm. The rotation speed was set to 1300 min⁻¹, and milling times of 1, 2, and 3 h were applied. Each milling batch consisted of 60 g of kaolin and 200 cm³ of absolute ethyl alcohol. This procedure was intended to enhance the surface area and plasticity of the kaolin. After the mechanical activation, the processed kaolin samples were used for further analysis.

Table 1. Sample codes and description of synthesized photocatalysts.

Run	Photocatalyst	Description
0	PE	Reference commercial glazed ceramic tile, no self-cleaning effect
1	K0hT0SFV	Pure Kaolin, 0 h activation, 0%wt. TiO2, no glassy phase (SFV)
2	K0hT12SFV	Kaolin, 0 h activation, 12%wt. TiO2, no glassy phase
3	K1hT0SFV	Kaolin, 1 h activation, 0%wt. TiO2, no glassy phase
4	K3hT0SFV	Kaolin, 3 h activation, 0%wt. TiO2, no glassy phase
5	K3hT3SFV	Kaolin, 3 h activation, 3%wt. TiO2, no glassy phase
6	K3hT6SFV	Kaolin, 3 h activation, 6%wt. TiO2, no glassy phase
7	K3hT12SFV	Kaolin, 3 h activation, 12%wt. TiO2, no glassy phase
8	K1hT12	Kaolin, 1 h activation, 12%wt. TiO2, with glassy phase
9	K2hT12	Kaolin, 2 h activation, 12%wt. TiO2, with glassy phase
10	K3hT12	Kaolin, 3 h activation, 12%wt. TiO2, with glassy phase
11	K3hT12-1C	Kaolin, 3 h activation, 12%wt. TiO2, with glassy phase, 1× polishing
12	K3hT12-2C	Kaolin, 3 h activation, 12%wt. TiO2, with glassy phase, 2× polishing
13	K1hT12N5	Kaolin, 1 h activation, 12%wt. TiO2, 5% wt. Nb2O5, with glassy phase
14	K1hT12N5-1C	Kaolin, 1 h activation, 12%wt. TiO2, 5% wt. Nb2O5, with glassy phase, 1× polishing
15	K1hT12N5-2C	Kaolin, 1 h activation, 12%wt. TiO2, 5% wt. Nb2O5, with glassy phase, 2× polishing
16	P25	Commercial TiO2, Degussa P25

presents the sample codes and a brief description of the synthesized photocatalysts, including both samples without a glassy phase and those with a 50 wt.% glassy phase. The samples include various kaolin activation times (1 to 3 h) and different TiO₂ concentrations (0%, 3%, 6%, and 12%) supported on kaolinite. Some samples were also doped with 5% Nb₂O₅ to investigate the effect of niobium addition.

Metallic oxides TiO₂ and Nb₂O₅ were impregnated onto the kaolinite surface following the sol–gel procedure described by Barbosa et al. [29]. Each synthesis batch consisted of 60 g of kaolin, 600 cm³ of absolute ethanol, 3 cm³ of acetic acid, and titanium(IV) isopropoxide and niobium n-butoxide, added according to the desired final composition. The components were combined in the following order: kaolin was first dispersed in ethanol, followed by the addition of acetic acid, and finally, the titanium(IV) isopropoxide and niobium n-butoxide precursors. The resulting mixture was stirred at approximately 700 min⁻¹ for 24 h at room temperature using a magnetic stirrer. For batches 8 to 15, glassy frit was incorporated based on the target formulation to assess its effect on coating adhesion, stability, and photocatalytic performance.

The photocatalyst suspensions were uniformly applied to 4 × 4 cm² unfired ceramic tile glazed surface using a spray-coating technique with a Western single-action airbrush kit, Vonder, Curitiba, Brazil. This kit includes six components: a 160 cm hose (0.38 cm thick), a 20 cm³ paint reservoir, a 90 cm³ container, and a 6.35 mm nipple adapter for an air compressor, operated at 30 psi.

After application, the coated substrates were dried at room temperature for 48 h and then fired in a laboratory roller kiln (40 °C min⁻¹ of heating rate up to 1185 °C, 5 min of soaking time, and followed by roughly 30 °C min⁻¹ of cooling rate up to room temperature), without the addition of a glassy phase. Furthermore, polished samples (with one or two polishing cycles) were included to evaluate the influence of surface finishing on photocatalytic activity. According to the literature [41], a polishing step can make the TiO₂ sites available for the molecules to be degraded. The polishing procedure was performed manually using a metallographic polisher equipped with a 203 mm diameter disc operating at 200 min⁻¹, 1 μm diamond paste, and distilled water as lubricant. A light pressure of ~2 N was applied to hold the samples against the outer edge of the disc for 1 min per cycle.

2.3. Characterization Techniques

XRD tests were conducted with a Rigaku Mini Flex 600 XRD equipment, Tokyo, Japan, using a copper lamp (Cu Kα = 1.5406 Å) at a voltage of 40 kV and a current of 30 mA. Scans were of 2θ, from 0° to 80°, with a step size of 0.05° and a step time of 1 s. The diffractograms were analyzed using Profex 5.0.1 software. Kaolinite, quartz, rutile, anatase, and fluorite were identified and refined using the Rietveld method with the BGMN model available in Profex.

Raman Spectroscopy (FT-Raman) analyses were conducted using a Bruker RFS 100/S spectrometer, Ettlingen, Germany, with a 1064 nm laser and measured between 100 and 1000 cm⁻¹ with a spectral resolution of 4 cm⁻¹. For X-ray Photoelectron Spectroscopy (XPS), analyses were performed using a Thermo Scientific K-Alpha spectrometer, Waltham, MA, USA analyzing the samples in survey and elemental modes with spot scans.

Contact angle measurements were performed using a Ramé-Hart goniometer, 250-F1, Netcong, NJ, USA by the sessile drop technique with distilled water, in triplicate. Microstructure was analyzed with a Field Emission Scanning Electron Microscopy (FEG-SEM), Thermo Scientific Quattro ESEM, and energy dispersive X-ray spectrometer, model ANAX-60P-B, Eindhoven, Netherlands. The samples were prepared with a gold layer and analyzed in various configurations.

Fourier Transform Infrared (FTIR) spectroscopy analyses were performed to investigate changes in the chemical bonds of the samples, using a Thermo Scientific Nicolet 6700 spectrometer, Waltham, MA, USA.

Photocatalytic Activity in Ceramic Coating

Figure 1 shows the schematic diagram used to evaluate the photocatalytic activity of the ceramic coating. The procedure was adopted from the ISO 10678:2010 [42]. A dye solution (200 cm³) containing methylene blue (3.7 mg dm⁻³, equivalent to 10.0 ± 0.5 μmol dm⁻³) was placed in contact with the tile surface, simulating the exposure of a photocatalytic coating to organic contaminants in water. A UV-A lamp with a wavelength of 365 nm and 9 W (DULUX S BL UVA 9W/78, Osram, Munich, Germany) was employed, with UV radiation intensity in the UV range (315 nm < λ < 400 nm) according to ISO 10678:2010 [42], which was 10.0 ± 1.0 W m⁻².

For dye degradation, ceramic coatings with photocatalyst were used. The samples were cleaned with distilled water, dried at 100 °C for 30 min and immersed in distilled water in the dark for 24 h, as per ISO 10678:2010 [42]. Subsequently, they were placed in a reactor with 200 cm³ of the MB dye solution, and kept for 15 min without exposure to UV-A with constant stirring for equilibrium. Afterward, exposure to UV-A radiation was initiated, with the reactor preheated for 30 min, as shown in Figure 1.

The photocatalytic activity was evaluated by monitoring the variation in the concentration of the dye solution over time. Aliquots (3 cm³) were collected every 30 min, transferred to quartz cuvettes for analysis, and then returned to the reaction medium.

The concentration of MB dye in the supernatant liquids was determined by ultraviolet–visible spectroscopy (UV-vis, UV-1601, Kyoto, Japan), using a spectrophotometer to measure absorbance at a wavelength of 664 nm. Absorbance at this wavelength was analyzed, considering the initial concentration of 50 mg dm⁻³ of MB and the calibration curve. In this process, absorbances for concentrations ranging from 5.0 mg dm⁻³ to 0.5 mg dm⁻³ were determined. Distilled water was used as a reference sample and the total testing time was 390 min.

All tests were conducted in duplicate, and the averages were used for analysis of results. A standard method for determining the photocatalytic activity of surfaces in aqueous media by degradation of MB dye was published in 2010 (ISO 10678:2010). This standard uses artificial ultraviolet (UV) light and describes the parameters and equipment to be used. The three main steps are:

(1)

Pre-irradiation to decompose any potential residual organic contaminants through photocatalytic oxidation on the tested materials surface;

(2)

Pre-adsorption of the solution onto the surface of the ceramic coatings;

(3)

Photocatalytic test.

As the exposure time increases, a decrease in the concentration of MB dye was observed due to the photocatalytic process, depending on the efficiency of the photocatalyst and the experimental conditions used. The results are presented in terms of ε%, which is calculated considering the initial concentration of MB dye and the variation observed over time. This calculation allows the effectiveness of the photocatalyst in reducing the MB dye concentration during the experiment to be assessed, providing a quantitative measure of the photocatalytic process performance. The results are expressed as ε%, calculated by the Equation (1).

ε% = (Co − Ct)/Co × 100

(1)

where: Co, is the initial concentration of the MB dye; and Ct, is the concentration after a given irradiation time.

3. Results

3.1. FE-SEM Study

Figure 2 show micrographs of the synthesized powders, before mixing and applying them to the tile samples. They reveal that the sample K0hT0SFV exhibits plate-shaped particles with a predominantly hexagonal profile, as previously observed in studies on similar clay minerals [43,44]. Changes resulting from high-energy milling (samples 1 h, 2 h and 3 h) are evident through increasing irregularities in the hexagonal shape and the presence of small particles surrounding the larger agglomerates. Observing TiO₂ and Nb₂O₅ nano-deposits on the kaolinite of samples T12 and T12N5 was difficult. Nevertheless, TiO₂ contributes to a rough and irregular surface, in contrast to the smooth and uniform texture of particles without TiO₂. Previous studies indicate that this rough texture can enhance catalytic activity and pollutant adsorption capacity due to the increased surface area [45].

3.2. Powder Samples X-Ray Diffraction

XRD analysis was performed to identify the crystalline phases present in the photocatalyst, and the results are shown in Figure 3. Kaolinite peaks are clearly observed in the samples K1hT12, K2hT12, and K1hT12N5. After 3 h milling, sample K1hT12N5 showed a loss of kaolinite crystallinity, as expected [46]. Quartz is detected, attributed to residue from the kaolin washing process. Quartz particles retain their crystallinity even after 3 h of high-energy milling. In the P25 sample, anatase and rutile phases were identified. Rietveld refinement, using CaF₂ as an internal standard, indicated approximately 54% amorphous phase, 39% anatase, and 8% rutile. No anatase or rutile were detected in the synthesized photocatalysts, even with a TiO₂ content of 12 wt.%. These results indicate that TiO₂ remains in an amorphous state after precipitation onto kaolinite particles.

3.3. Powder Samples Raman Spectroscopy

FT-Raman analysis of P25, K1hT0SFV, K3hT12, and K1hT12N5 are shown in Figure 4. Anatase peaks were clearly observed in sample P25 at 399, 518, and 641 cm⁻¹ positions while rutile peaks, that should supposedly appear at 612 cm⁻¹ and between 440–450 cm⁻¹ position [47], may have been obscured by the anatase peaks due to its lower content in the sample. The vibrational features of kaolinite from 200 to 2000 cm⁻¹, where a multiplicity of peaks is expected due to reduced symmetry from substantial deviations from the ideal hexagonal structure [48], were not fully resolved, as spectral noise hindered more accurate interpretation [49,50,51].

The Raman spectra of TiO₂-added photocatalysts show no anatase or rutile peaks. Instead, new peaks around 1045 cm⁻¹, 1341 cm⁻¹, and 1402 cm⁻¹ emerge, indicating amorphous or pre-crystalline TiO₂. In the work of Kim [52], non-crystalline TiO₂ was identified, indicating non-crystalline TiO₂ or TiO₂ in pre-crystalline phase, which aligns with the peaks observed in the analyzed photocatalysts [52].

3.4. Ceramic Tile Contact Angle

The surface wettability was evaluated by measuring the contact angle formed between a water droplet and sample surface in air. This parameter is commonly used to assess the self-cleaning performance of surfaces. For non-porous and low-roughness surfaces, lower contact angles—associated with increased hydrophilicity—are considered desirable. Variations in contact angle, whether under light exposure or in the absence of light, may indicate photoinduced surface activity.

As illustrated in Figure 5, images of water droplets on the sample surfaces were obtained after 2 h of UV-A light exposure, allowing visual comparison of wettability across different materials. The corresponding contact angle values—including measurements under both UV-A light and dark conditions—are presented in Figure 6, providing a quantitative assessment of the photoinduced changes in surface wettability. The numbers are provided in Table S1 of the Supplementary Materials.

The contact angles measured for the standard glazed product (PE) under dark and UV-A conditions were 83° and 72°, respectively. Although these values classify the surfaces as hydrophilic, since the contact angle is below 90° [53], the results indicate low self-cleaning ability and minimal photoactivity. The contact angles observed for the P25 and K3hT12SFV samples were 47° and 26°, respectively. These low values can be attributed to the inherent surface roughness of the P25. In the case of K3hT12SFV sample, surface roughness and porosity. However, they are not indicative of a self-cleaning effect. Furthermore, no significant change in the contact angle of the P25 sample was observed after UV-A exposure, suggesting the absence of photoinduced activity. This behavior can be explained by the well-known anatase-to-rutile phase transformation, which reduces the photocatalytic activity of TiO₂. In the case of the K3hT12SFV sample, a slight reduction in the contact angle after UV-A exposure was observed, which may be associated with photoinduced activity. Nevertheless, as the contact angle was already low under dark conditions, the further decrease was not significant enough to be considered indicative of an effective self-cleaning effect.

The samples K3hT12, K3hT12-1C, and K3hT12-2C exhibited an interesting behavior. Under dark conditions, these samples presented contact angles ranging from 64° to 89°, similar to the reference product (PE). The K3hT12 sample showed no change in contact angle upon UV-A exposure, indicating the absence of photoactivity. This result can be attributed to the encapsulation effect promoted by the glassy phase formed over the kaolinite-TiO₂ particles. In contrast, the samples K3hT12-1C and K3hT12-2C exhibited a significant reduction in contact angle after UV-A exposure, decreasing from 89° to 48° and from 64° to 30°, respectively. These samples were subjected to a polishing process, which exposed TiO₂ particles on the surface and made them accessible to water molecules in the droplet. The positive effect of polishing procedures on enhancing surface photoactivity has been previously reported in the literature [41]. Therefore, the polishing step plays an important role in optimizing the photocatalytic activity of ceramic tile surfaces, contributing to a more efficient and effective photocatalytic performance.

3.5. Microstructure and Composition of the Ceramic Tile Surface

Figure 7 shows the microstructure and EDS of the ceramic tile surface of the reference product (PE), K3hT12, and K3hT12-2C samples. It is possible to observe that the photocatalytic coating has brought an increase in surface roughness. The surface has slight reduced its brightness and no visual changes in the whiteness. No significant differences were observed between samples K3hT12 and K3hT12-2C, which corresponds to the desired effect, since the objective of polishing was to make TiO₂ available for adsorption. The low intensity polishing procedure was not able to remove the thin photocatalytic layer, which was as desired. EDS reveals the usual elements in ceramic tile glassy surface: Si, Al, Ca, K, Mg, Zn, and O. Some of those elements can affect positively or negatively the photoactivity of Ti [50,51,52]. Ti was observed in all samples, but the intensity was much lower in the reference product PE than in K3h12T and K3h12T-2C. This result reveals the effective presence of TiO₂ particles in the prepared samples with the synthesized photocatalyst. No significant difference was observed in the Ti intensity between K3h12T and K3h12T-2C due EDS penetration capacity, which is higher than the thin layer of glass phase that cover the TiO₂ particles.

Figure 8 shows the XPS spectra from 450 to 475 eV of the ceramic tile surface. The main objective of this procedure was to identify differences between the unpolished and polished samples, K3h12T and K3h12T-2C. Initially, the Ti 2p and C 1s regions were examined to correct potential shifts in binding energies due to charging effects. To ensure accurate XPS data, the binding energy of titanium (Ti) was calibrated against the stable C 1s peak, set at 284.5 eV. This calibration method provides a reliable basis for analysis, ensuring consistency and comparability across the results.

The background noise was in the same order of magnitude of the Ti 2p sign and no significant difference could be identified due to the low amount of TiO₂. Nevertheless, a slight difference in the Ti 2p sign between those samples and the reference tile (PE) could be observed. Once again, confirming the presence of TiO₂ in the coated samples. The presence of the photocatalyst suggests an initial concentration of 12% TiO₂, equivalent to 7% Ti. However, the addition of frits, which constitute 50% of the photocatalyst composition, significantly reduces the Ti percentage to 3.5%. XPS samples present a surface with ~60% adventitious carbon, and Ti concentration remains in less than 1%, which renders quantification and identification challenging by XPS.

Fourier Transform Infrared Spectroscopy

The qualitative analysis of the spectra was based on the relationship between intensity and specific wavenumbers as can be seen in Figure 9. As observed by Bagheri et al. [54], the peak observed at 732 cm⁻¹ in the spectra of samples K3hT12-2C and K3hT12 indicates the presence of the O-Ti-O bond, which is characteristic of the anatase crystal phase of TiO₂; the PE sample is absent this peak. These results confirm once again the presence of TiO₂ at the prepared layer, suggesting anatase is the predominant arrangement, which was difficult to identify by XRD due to its the low amount.

Furthermore, a distinct peak at 925 cm⁻¹ was identified in the same samples, suggesting the presence of Ti–O bonds in a different structural configuration, possibly associated with the rutile phase of TiO₂. In this crystalline form, the TiO₆ octahedra exhibit a more compact and symmetric arrangement compared to the anatase phase, leading to characteristic vibrational modes observable in FTIR spectroscopy. Specifically, bands in the region of 920–930 cm⁻¹ have been attributed to Ti–O stretching vibrations in rutile-type structures. This distinction is relevant for understanding the crystalline composition of the samples, which in turn has a direct impact on their photocatalytic behavior due to differences in electronic structure and charge carrier dynamics [55,56].

The bands located at 1621 cm⁻¹ and 1627 cm⁻¹ can be attributed to the presence of adsorbed water and surface hydroxyl groups. The absorption observed at 2900 cm⁻¹ is associated with residual acetic acid and/or the presence of isopropyl groups. Additionally, the band at 3710 cm⁻¹ is indicative of the formation of terminal hydroxyl groups, evidencing O–H bonding. The slight variation between the peaks in the 1621–1627 cm⁻¹ region may be attributed to different types of interactions among surface functional groups or to the relative amount of adsorbed water [57].

3.6. Photocatalytic Activity

3.6.1. Reference and Trial Photocatalytic Experiments

Several prior photocatalytic experiments were carried out to serve as a reference for performance evaluation and validation. The experimental runs were divided into two main groups. The first group consisted of: (a) photolysis, in which the MB solution alone was exposed to UVA, serving as the baseline where minimal degradation was expected; (b) P25 calcined at 800 °C, where no anatase-to-rutile (ART) transformation was expected, representing the highest degradation performance; (c) P25 calcined at 1185 °C, where the ART transformation was assumed to be complete; and (d) PE 1185 °C, representing the standard firing condition of a glazed porcelain tile product, where no photoactivity is expected. The second group consisted of: (a) kaolin activated for 3 h without TiO₂ addition; and (b), (c), and (d) kaolin activated for 3 h with 3%, 6%, and 12% TiO₂ addition, respectively. The results are provided in Table S1 of the Supplementary Materials.

According to the results presented in Figure 10, photolysis accounted for approximately 15% of MB degradation after 390 min of exposure. The regular porcelain tile (PE 1185 °C) exhibited ~13% MB degradation. This lower value, compared to photolysis, is attributed to the shadowing effect produced by the test pieces, which reduces the illuminated area inside the reactor chamber. Therefore, PE 1185 °C is considered the baseline for comparing all other results, as it better represents the illumination profile. P25 calcined at 1185 °C resulted in ~19% MB degradation, while P25 calcined at 800 °C achieved 63%, in agreement with expectations based on the ART transformation, as anatase exhibits higher photocatalytic activity than rutile [15,29,35].

Figure 10 also presents the results for kaolin with increasing TiO₂ content from 0% to 12%. Kaolin alone exhibited 17% MB removal, which is 5% higher than PE 1185 °C. This result is likely attributed to adsorption rather than degradation. The degradation yield increased proportionally with the TiO₂ content, reaching 57% at a loading of 12% TiO₂. It is noteworthy that the equivalent TiO₂ loading (g·m⁻²) in K3hT12SFV matched that applied in P25 samples calcined at 800 °C and 1185 °C. For the other samples—K3hT0SFV, K3hT3SFV, and K3hT6SFV—the total mass of kaolin plus TiO₂ was maintained constant, equal to that of the K3hT12SFV specimen. These findings supported the selection of 12% TiO₂ loading for subsequent experimental stages. Nonetheless, TiO₂ concentrations higher than 12% were not explored and remain a potential avenue for future investigations.

3.6.2. Photocatalytic Experiment with 12% TiO₂ Glazed Products

The previous steps demonstrated the feasibility of achieving MB degradation yields by synthesizing 12% TiO₂ in a kaolin matrix, fired at 1185 °C, with performance in the same order of magnitude as P25—800 °C. The next step was to investigate the effect of the presence of a glassy phase alongside the photocatalyst. Figure 11 presents the results of MB degradation over time.

Samples coated with K3hT12 and K1hT12N5 showed ~32% MB degradation after 390 min of UV-A exposure. This result is lower than 57% where the coated layer had no glassy phase, K3hT12SFV. The presence of the glassy phase in these samples results in a physical barrier around the photocatalyst particles, formed during thermal treatment at 1185 °C. This barrier partially limits the exposure of the photocatalyst to the external environment, reducing its ability to interact with environmental contaminants and, consequently, its photocatalytic activity. Nevertheless, the vitreous phase contributes to the material’s structural mechanical stability against wear and other aesthetic properties. At this level no contribution could be observed by the addition of Nb₂O₅ to TiO₂.

Polishing exposed TiO₂ particles on the surface, increasing photocatalytic activity. These results were observed for both K3hT12-1C and K1hT12N5-1C, submitted to one polishing cycle, after 390 min of UV-A exposure. Those samples rise 49% and 41% of MB degradation respectively. The second polishing procedure have produced different results for both samples: K3hT12-2C, achieved a photocatalytic degradation 59%, similar to that observed in the K3hT12SFV sample and close to the 63% degradation level of the P25—800 °C. On the other hand, K1hT12N5-2C samples have decreased to 16% the MB degradation, close to PE, P25—1185 °C, and photolysis. The sample with Nb₂O₅ addition was easily removed during the polishing procedure. This result is probably due to a reduction in the wear resistance of the sample, although it is not clear whether this was caused by Nb₂O₅ or kaolinite, since the activation time was 1 h, making them less reactive and less likely to dissolve during firing compared to the 3 h activation time. It can be concluded that polishing after firing is a mandatory procedure in order to develop the maximum photocatalytic activity. Nevertheless, the control of the polishing needs to be very precise to avoid over polishing.

3.6.3. Surface Aspect Before and After MB Degradation

The images shown in Figure 12 correspond to the samples before and immediately after the photocatalytic activity test. Colorimetric data (Lab coordinates) was extracted directly from the images by using a Python algorithm (v. 3.11.13) at Google Colab (v. 1.0.0). It can be observed that the P25—800 °C sample exhibits the same appearance before and after the test, and the same Lab results, despite being a sample with porosity and roughness. This behavior was also observed for samples K3hT12, K1hT12N5, and K1hT12N5-1C.

All other samples showed residual stains of methylene blue to a greater or lesser extent. Among them, the most notable result was the K3hT12SFV condition, where the b-value dropped from 4.2 to −22.6. This sample does not contain frit in its composition and presents a high porous surface. Before degradation, methylene blue is adsorbed onto the metakaolinite particles. However, upon exposure to UVA light, the stain disappears due to the action of TiO₂, as the MB molecules diffuse to the photocatalytic active sites.

All other samples contained frit addition and, therefore, are less porous and rough, as shown in the micrograph in Figure 8. For this reason, the staining was of lesser intensity and was also eliminated after continued exposure to UVA light.

4. Conclusions

This paper presents original results on the development of a photocatalytic layer specifically formulated to be compatible with the single-firing process used in ceramic tile production. Titanium dioxide (TiO₂) was supported on kaolinite particles in concentrations ranging from 0 to 12 wt.%. The optimal TiO₂ content was combined with a glass frit to produce a surface texture characteristic of ceramic tiles. The effect of Nb₂O₅ addition was also investigated. Surface characterization techniques confirmed the presence of TiO₂ in the top layer of the prepared samples.

Kaolinite particles containing 12 wt.% TiO₂, fired at 1185 °C, exhibited a degradation performance comparable to that of Degussa P25 fired at 800 °C. In contrast, the unfunctionalized commercial ceramic tile reference fired at 1185 °C, the kaolinite layer without TiO₂, and the Degussa P25 layer fired at 1185 °C showed negligible photocatalytic activity—on the same order of magnitude as the photolysis of methylene blue (MB) under UVA light for 390 min (ε% ~15%). The highest MB degradation (ε% ~63%) was achieved by the Degussa P25 layer fired at 800 °C, followed closely by the kaolinite + 12 wt.% TiO₂ layer fired at 1185 °C (ε% ~57%).

The addition of a glassy phase enhanced particle anchoring but reduced MB degradation efficiency by more than 50%. A polishing procedure was essential to restore photocatalytic performance by removing the vitreous phase covering the active TiO₂ sites. However, excessive polishing could entirely remove the thin TiO₂ layer from the surface. Contact angle measurements were consistent with the MB degradation results, suggesting that a self-cleaning effect may be attainable.

The addition of Nb₂O₅ did not enhance photocatalytic performance in this study. On the contrary, it appeared to reduce wear resistance, as the TiO₂ + Nb₂O₅ layer was more easily removed during polishing.

These results demonstrate the potential of kaolinite-supported TiO₂ photocatalysts for producing self-cleaning ceramic tiles using standard single-firing industrial processes. This approach could lower production costs and help expand the application of self-cleaning functionality in this globally significant product.