Acesulfame K Photodegradation over Nitrogen-Doped TiO₂

Abstract

Acesulfame K is a zero-calorie alternative to sugar used worldwide. There is contradictory information on the toxicity of the compound, but its accumulation in the aquatic environment is undeniable. In this study, one-pot sol-gel synthesis was used to obtain nitrogen-doped TiO₂ photocatalysts. Doping up to 6.29 wt % of nitrogen caused an increase in the surface area of the catalysts (48.55–58.23 m²∙g⁻¹) and a reduction of the pH_PZC value (5.72–5.05). Acesulfame K photodegradation was tested at the initial concentration of 20–100 ppm and the catalyst concentration at the level of 1 g∙L⁻¹. Compared to the pure anatase, 4.83–6.29 wt % nitrogen-doped TiO₂ showed an effective photodegradation of Acesulfame K. Ninety percent molecule removal was obtained after ~100 min, ~90 min, and ~80 min for initial concentrations of 20 ppm, 50 ppm, and 100 ppm, respectively. The increased activity of the catalysts is due to the modification of the TiO₂ lattice structure and probably the limitation of the photogenerated electron/hole charge carrier recombination. It was shown that the electrostatic interactions between Acesulfame K and the catalyst surface play an important role in the photodegradation efficiency.

1. Introduction

Acesulfame potassium (Ace K) is a zero-calorie synthetic alternative to sugar used to sweeten beverages (since 1993) and other items since 2003 (except for meat and poultry) [1,2]. In the food industry, Ace K is often mixed with various sweeteners, e.g., sucralose or aspartame, to achieve a sugar-like taste. It has good thermal resistance; it does not decompose at high temperatures, so it can be used in baked goods, as well as in products with a long shelf life. It helps to mask the bitter aftertaste of aspartame and sucralose. It can be found in sweeteners, low-calorie drinks, alcohols, dry bases for coffee and instant teas, fruit preparations, confectionery and bakery products, chewing gums, and cosmetics for oral hygiene and as a supplement to drugs.

Ace K is one of the most stable artificial sweeteners [3]. It has been shown that Ace K does not undergo biotransformation in vitro, nor in vivo, in both humans and experimental animals, where it is excreted unchanged in urine [4]. Wastewater treatment plants are not able to completely remove Ace K from water; the concentration in surface and drinking water was found at the level of 48–88 µg/L, in rivers 0.12 ng/L to 2 mg/L, and in water reservoirs and groundwater between 4.7 and 34 µg/L, which is of concern and may pose a risk to aquatic organisms, e.g., producing oxidative stress in the blood, liver, gill, brain, and muscle of the common carp (Cyprinus carpio) [3,5,6]. The reported concentrations vary depending on the location of the test and sampling. There is no doubt, however, that this compound accumulates in the aquatic environment. For this reason, an effective technology for the elimination of Ace K must be found so that its degradation does not cause adverse side effects, e.g., decomposition into more harmful products.

Ace K photolysis in sunlight is very slow and ineffective [7,8] or does not occur at all [9]. UV light photolysis revealed low (LP-UV Hg lamps and MP-UV Hg lamps [10], UV-LED sources of light [11], and LP-UV Hg lamps 0.75 mW cm⁻² [12]), or no (22W low pressure mercury lamps [13]) degradation efficiency. Boron-doped diamond [14], PbO₂, and Pt [15] were found to be effective in Ace K electro-oxidation processes. However, it should be remembered that the use of heavy metals as catalysts on an industrial scale of the process is undesirable due to the risk of catalyst penetration into groundwater and/or soil. The price of platinum catalysts or subsidized diamonds is not insignificant.

Among advanced oxidation processes (AOPs), Ace K was tested in photodegradation and ozonation processes. Shao et al. showed the best degradation performance at a neutral pH in the presence of O₃/peroxymonosulfate [16]. A higher ozone input, alkaline conditions, and higher temperatures favored Ace K removal in Zhou et al.’s study [17]. Scavenging tests confirmed that direct ozonation accounted for degradation under acidic conditions, whereas the hydroxyl radicals produced by ozone under alkaline conditions gradually promoted Ace K degradation with increasing pH [17].

Ghosh et al. [9] used Evonik TiO₂ (80–20% anatase to rutile) as a photocatalyst for Ace K degradation. The authors showed approximately 84–100% degradation for initial Ace K concentrations between 10 and 50 mg/L at a pH of 6.0 and a light intensity of 100 mW/cm². Photocatalyst loading was shown to be a crucial parameter for degradation, and the optimum dose was found to be 1 g/L. Lopez-Munoz [18] showed that an enhanced photocatalytic activity for Ace K degradation might be related to the catalyst bicrystalline composition, able to promote the separation of photogenerated electron-holes. The authors tested TiO₂ sol-gel catalysts calcinated at 650 and 750 °C and Evonik TiO₂. The anatase crystal size was in a range from 23.7 (Evonik) to 53.5 nm (TiO₂ calcinated at 750 °C), and the BET surface area of the tested materials was in a range from 38 (TiO₂ calcinated at 750 °C) to 94 m²/g (TiO₂ calcinated at 650 °C). Catalyst loading was on the level of 1.0 g/L with an initial Ace K concentration of 10 ppm. The authors do not ascribe a significant role to photodegradation depending on the surface properties (such as BET or crystalline size) [18].

To the authors’ knowledge, to date, nitrogen-doped TiO₂ was not tested in Ace K photodegradation processes. The presence of nitrogen can alter the band structure or suppress the recombination efficiency of the photogenerated electron–hole pairs, resulting in an enhanced photocatalytic capacity of TiO₂. The paper aimed to evaluate the photocatalytic activity of N-doped TiO₂ in the photodegradation of Ace K under conditions of variable nitrogen content (0–5 wt %) and variable initial Ace K concentration (20–100 ppm). The impact of the catalyst structure (porous structure and surface properties), geometry, and properties of the molecule over the course of Ace K removal from aqueous solutions was also discussed.

2. Results and Discussion

2.1. Catalysts Characterization

Figure 1 shows the XRD patterns of the synthesized pure and N-doped TiO₂ powders. All peaks are corresponding to those of anatase, while no diffraction peaks are induced by nitrogen, which is owing to the low content of the doping element. The intensity of the peaks of N–TiO₂ is slightly weakened compared with that of undoped TiO₂. A slight shift of the peak maxima towards lower ° values was observed in the case of all examined nitrogen-doped catalysts; nevertheless, the peaks are sharp, so it can be assumed that the anatase structure is stable. This observation does not depend on the dopant amount, at least in terms of the analyzed amount of nitrogen (1.55–6.29 wt %). From the line broadening of the (1,0,1) diffraction peak by Scherrer’s method, the crystallite sizes of pure and N-doped TiO₂ were calculated (

Table 1. Physicochemical characterization of TiO₂ and nitrogen-doped TiO₂.

Catalyst	SBET	VTOT	ADP	Average Crystallite Size	Band Gap Energy	pHPZC	Nitrogen Content
	m2∙g−1	cm3∙g−1	nm	nm	eV	-	wt %
TiO2	48.55	0.063	3.71	15.43	3.02	5.72	-
1N_TiO2	49.41	0.081	6.86	15.38	3.18	5.69	1.55
3N_TiO2	58.23	0.111	6.36	15.00	3.18	5.60	4.83
5N_TiO2	56.92	0.092	6.61	15.17	3.17	5.05	6.29

). From the analysis of the crystallite size, it can be concluded that the amount of nitrogen introduced into the TiO₂ lattice had little effect on the reduction in crystallite size in the range of 0.05–0.43 nm. N-doped TiO₂ powder XRD results were found to be strongly dependent on the nitrogen precursor. In the case of ammonia (as in this study) [19], ammonium nitrate [20], and nitric acid [21] as nitrogen precursors, no visible changes in the TiO₂ crystal structure were observed; i.e., the N-doped TiO₂ crystalline phase was determined by the synthesis conditions (mainly the calcination temperature, which determines the share of anatase and rutile phases), as in the case of undoped TiO₂. As a result, no significant differences between N-doped and undoped TiO₂ can be seen in the XRD patterns. Different behavior was observed in the case of semicarbazide [22] and N,N-dimethyl urea [22] as nitrogen precursors. In these cases, the doping process was promoting the rutilization of TiO₂ even at calcination temperatures as low as 500 °C. Divergent results in the described range were noted in the case of urea as a nitrogen precursor: in calcination at 500 °C [22], 400 °C [23,24], and 300 °C [25], there were no significant XRD differences; in Yadav et al.’s case, in calcination at 400 °C, clear rutile signals were observed [26]. The above-mentioned materials were synthesized with the use of the sol-gel method.

Nitrogen content was measured with the method described by Gazulla et al. [27]. The actual nitrogen content determined by the EDS (energy dispersive X-ray spectroscopy) method is within the range of 1.55–6.29 wt % (Table 1). These amounts differ slightly from the assumed nitrogen doping amount at the synthesis stage.

The nitrogen sorption isotherms of TiO₂ and N-doped TiO₂ are presented in Figure 2A. According to the updated IUPAC classification [28], which distinguishes eight types of adsorption isotherms, all synthesized photocatalysts are type IVa—typical for mesoporous materials. The presence of hysteresis is evidence of cylindrical pores larger than ~4 nm (the value adopted for the measurement of nitrogen at 77 K [28]). Despite maintaining the shape of IVa isotherms, in the case of all analyzed N-doped TiO₂ powders, two hysteresis loops were observed. A TiO₂ hysteresis loop can be classified as an H2 type typical for complex mesoporous structures, in which the pore structure is connected. Nitrogen-doped catalysts are typical of two hysteresis loops that can be classified as the H4 type. This type of hysteresis is characteristic of slit-shaped pores. The nitrogen introduction into the crystalline structure of TiO₂ influences the adsorption/desorption isotherms (Figure 2A) and, as a consequence, the pore size distribution (Figure 2B). It can be observed that, in the case of all the obtained materials, the largest pore volume is represented by small mesopores with an average width of 3.5–4 nm. The introduction of nitrogen into the TiO₂ structure, by increasing the proportion of medium and large mesopores, affects both the total pore volume (V_TOT) and the pore size distribution (Figure 2B). In the case of N-doped TiO₂, the proportion of larger diameter mesopores increases up to ca. 7 nm. The above observation significantly influences the average pore diameter (ADP, Table 1) value. Nitrogen TiO₂ doping in the range of 1.55–6.29 wt % increases the ADP value by ca. 3 nm (Table 1). As is known, the increase in S_BET and V_TOT influences both the adsorption of the pollutant on the catalyst surface and the potential number of active sites for the photocatalytic process. Changes in ADP primarily affect the adsorption process, which will be also strongly affected by the geometrical spatial dimension of the adsorbed molecule.

The band-gap values were found to be in the range of 3.02–3.18 eV (Table 1; the graphical calculation is presented in Figure 3; diffuse reflectance spectra are presented in Figure S1). Nitrogen doping caused a slight increase in the band gap energy value, which is the effect of interstitial or substitutional doping [29]. The location of nitrogen is crucial to its behavior as a photocatalyst, but there are still conducted studies regarding which position is most beneficial. Regardless of the type of doping, nitrogen alters the band structure and, as reported before, possibly suppresses the recombination efficiency in the photocatalytic reaction [29,30]. The band gap value results obtained in this study differ from the results published so far. Earlier studies have shown that nitrogen doping, regardless of the interstitial or substitutional doping and regardless of the amount of the subsidized element, leads to a narrowing of the band gap [31,32,33]. In this study, excitation of electrons from valence band to conduction band of the N–TiO₂ catalysts needs higher energies while the absorption changes toward lower energies in the case of the un-doped TiO₂. In case of all the obtained catalysts, UV range is suitable for the catalysts’ activation. Further research is required, especially regarding band localization, which will show what the reason was for the little, but still, increase in the band gap value.

Figure 4 displays the SEM (scanning electron microscopy) and EDS spectra images of the pure TiO₂ and 5N_TiO₂ catalysts. It can be seen that the N element distribution on the TiO₂ is uniform (even the distribution of the colors of individual elements, Figure 4C). Microscopic studies show the reduction in the agglomeration of catalyst particles in the case of materials doped with nitrogen. In the case of the doped catalyst (Figure 4B), smaller particles/agglomerates can be noticed, compared to the undoped TiO₂ (Figure 4A). An increased contact surface (due to smaller agglomerates) strongly impacts the molecule adsorption (greater access of particles to the catalyst surface) and the photodegradation process. In the case of photodegradation, access to a larger surface results in the availability of more active sites of the catalyst. As is known, the increased adsorption is also important in the process of photodegradation [34].

For the determination of the catalysts’ thermal stability, thermogravimetric studies under the constant argon flow were performed. In the entire range of the analyzed temperature (20–900 °C), no mass loss was observed (Figure S2), which means N-doping had no impact on TiO₂ thermal stability.

2.2. Physical Interaction of the Ace K Molecule and Catalyst Surface

In the range of 20–100 ppm, regardless of the concentration, the pH of the solutions (

Table 2. Physicochemical properties of the Acesulfame K.

Characteristic
	Acesulfame K
Molecular weight, g mol−1	163.15
Solubility in water at 20 °C, g dm−3	270
The longest spatial dimension of the dissociated form of the molecule *, nm	0.574
pKa	2.0
100 ppm solution pH	6.019
50 ppm solution pH	6.202
20 ppm solution pH	6.420

* Calculated from the DFT method—PBE0 functional/Def2-TZVP basis set.

) indicates that Ace K is dissociated. The anionic form of the compound will therefore be preferably adsorbed on the surface of positively charged materials due to electrostatic attraction. As is known, the photodegradation process strongly depends on the adsorption. An activated photocatalyst surface will degrade organic contamination faster when the contamination is on or near the catalyst surface. This dependence is due to a short lifetime of the electrons knocked out of the activated catalyst surface and the probability of the electron contacting the contamination.

Under the acidic or alkaline condition, the surface of titania can be protonated or deprotonated, respectively, according to the following reactions [35]:

$$\mathrm{pH}<{\mathrm{pH}}_{\mathrm{PZC}} \mathrm{TiOH}+{\mathrm{H}}^{+}\to {\mathrm{TiOH}}_2^{+}$$

(1)

$$\mathrm{pH}>{\mathrm{pH}}_{\mathrm{PZC}} \mathrm{TiOH}+{\mathrm{OH}}^{-}\to {\mathrm{TiO}}^{-}+{\mathrm{H}}_2\mathrm{O}$$

(2)

The pH_PZC (point zero charge pH) of the materials synthesized in this study was in the range of 5.72–5.05 (Table 1). An increase in the nitrogen content in the doped TiO₂ structure increases the acidity of the catalyst surface. Considering solutions pH (Table 2) and pH_PZC of the analyzed materials, under the conditions of the research, catalysts are negatively charged. As the concentration of Ace K decreases, the pH of the solution increases. The increasing difference between pH and pHpzc causes a greater proportion of negative charge on the catalyst surface, thus reducing the possibility of organic compound adsorption on the surface and limiting the strength of physical adsorption bonds between the catalyst and Ace K. Considering the above electrostatic relationships, the most favored catalyst is undoped TiO₂ in a 100 ppm solution, while the most limited adsorption should be observed for the 5N_TiO₂ catalyst in a 20 ppm Ace K solution.

Another dependence affecting the course of the adsorption process and, consequently, the efficiency of the photodegradation process is the geometric size of the pollutant molecule and the porous characteristics of the surface of the photocatalyst. The most preferable pore width for adsorption is ca. 2 times greater than the longest special dimension of the molecule [36]. In the case of Ace K, small micropores approximately 1.2 nm wide would be preferable for adsorption. As pointed out in Section 2.1., the synthesized materials are mesoporous with ADP in the range of 3.71–6.86 nm (Table 1). In the assumed quantitative share, nitrogen doping resulted in widening the pore width of the photocatalyst. This is due to the disturbance of the surface structure of the material, which is confirmed by the band gap energy. Considering ADP (Table 1) and the geometrical spatial dimension of Ace K (Table 2), there is no catalyst that would be significantly favored for the adsorption of the molecule. An increased pore width will have a significant impact on the diffusion process of the reactants. The larger diameter of the pores affects the rate of diffusion of the reagents into the pore and the decomposition products outside the pore.

The last factor that will be described in this paper that affects the adsorption process is the catalyst surface size. Nitrogen doping caused an increase in surface area of approximately 10 m²/g. This is not a particularly significant value, but it may influence the physical deposition of the compound on the surface. Taking into account the obtained S_BET results (Table 1), the 3N_TiO₂ (58.23 m²/g) and 5N_TiO₂ (56.92 m²/g) catalysts will be privileged.

2.3. Acesulfame K Photodegradation

The course of the Ace K degradation in the dark and under UV irradiation is presented in Figure 5 (the variation of absorbance over the process time is presented for the example of the photodegradation of a 100 ppm initial Ace K concentration in the presence of 5N_TiO₂ catalyst, Figure S3). It can be seen that both the course of the process and the final concentration of Ace K strongly depend on the initial compound concentration and the applied catalyst.

Physical adsorption, conducted in the dark, depends on the factors described in Section 2.2. The use of the catalysts 3N_TiO₂ and 5N_TiO₂ due to the largest S_BET (58.23 and 56.92 m²∙g⁻¹ respectively, Table 1) led to the highest level of adsorption among other catalysts in all applied initial Ace K concentrations. Considering the electrostatic interactions, the strongest repulsion was found in the case of the 5N_TiO₂, while the weakest was found in the case of TiO₂ catalyst. Observing the course of Stage I (Figure 5), it can be concluded that S_BET can play a greater role than the electrostatic interactions between the anionic compound and negatively charged catalysts. The variable adsorption level of the component on the surface is clearly due to its initial concentration and is the result of the probability of the contact between the catalyst surface and the particle.

In the case of two initial concentrations (20 and 50 ppm, Figure 5A,B) and all applied catalysts, an increase in the Ace K concentration at the first stage of irradiation was observed. In the first 30 min of irradiation in the case of the 20-ppm initial concentration (Figure 5A(II)) and 15 min in the case of the 50 ppm initial concentration (Figure 5B(II)), we can observe Ace K desorption and photodegradation simultaneously. The parameter that influenced this process was the catalyst activation and electrostatic interaction between the catalysts and the Ace K solution. Activation of the catalyst surface causes knocking electrons out of the surface, so it is assumed that the adsorption process is inhibited. When UV irradiation falls on the semiconductor surface, electrons in the valence band gain energy and become excited and thus move to the conduction band. Electrons may flow to or from the adsorbate, allowing for the formation of both positively and negatively charged species on the surface [37]. Assuming that the amount of a negative charge on the catalyst surface was increased, it was possible to observe the desorption process resulting from the anionic nature of the dissociated Ace K molecule. The initial concentration of 100 ppm (pH 6.019, Table 2) is characterized by the weakest electrostatic repulsion of the molecule, so it is possible that the phenomenon of desorption was not noticed (Figure 5C). The desorption of Ace K has not been observed by other authors, probably due to the Ace K solution pH modification [9,11].

Photodegradation (Figure 5(III)) over anatase TiO₂ and N-doped TiO₂ catalysts is effective for all analyzed initial concentrations. It can be noticed that nitrogen doping in the amount of 4.83 wt % (sample 3N_TiO₂) and 6.29 wt % (sample 5N_TiO₂) significantly affects the acceleration of the process. Ninety percent removal was observed after ca. 100 min (initial concentration of 20 ppm), ca. 90 min (initial concentration of 50 ppm), and ca. 80 min (initial concentration of 100 ppm). The variable time of the 90% photodegradation is due to initial Ace K concentration and is the result of the probability of the contact between the catalyst surface and the Ace K molecule. The most effective photodegradation was observed in the case of the catalysts 3N_TiO₂ and 5N_TiO₂. Nitrogen-modified lattice structure probably limits the recombination of the photogenerated electron/hole charge carriers, increasing the photodegradation efficiency, which was described before [29,38].Increasing S_BET also plays an important role due to the potential increase in the number of active sites where the photodegradation process can take place. In case of the doping in the amount of ca. 1 wt % (theoretical assumption at the stage of synthesis was 1%, determination of the actual nitrogen content by EDS method resulted in 1.55 wt %) insignificantly modifies anatase structure. 1N_TiO₂ catalyst resulted in similar as pure anatase photodegradation efficiency.

The enhanced photocatalytic activity of N-doped TiO₂ can be attributed to the synergistic effects of many factors, such as morphology, optical response, nitrogen concentration, crystal phase. The morphology and nitrogen doping level were considered [29] to be more significant than the crystal phase and optical response. Conducting additional research, especially on the Ace K degradation mechanism and detailed gap location research will lead to narrow down the parameters that improve the N-doped TiO₂ properties compared to the pure TiO₂.

3. Materials and Methods

3.1. Un-Doped and N-Doped TiO₂ Catalysts Synthesis

One-pot sol-gel synthesis was applied in this study. Firstly, 30 mL of dry ethanol (99.8%, POCH, Gliwice, Poland) was heated to a temperature of 70 °C under constant stirring. TIPP (titanium (IV) isopropoxide, 97%, Sigma Aldrich, St. Louis, MO, USA) was added slowly (1–2 drops min⁻¹) to achieve the desired amount of TiO₂. During this process, the system was kept in agitation at a constant temperature of 70 ± 1 °C. To promote the hydrolysis–condensation process of pure TiO₂, immediately after TIPP addition, a certain amount of a mixture of water/acetic acid (99.9%, POCH, Gliwice, Poland) (55 wt % ratio, nH₂O/nC₁₂H₂₈O₄Ti = 4:1) was added. Sol was cooled to room temperature and aged for 24 h. Subsequently, it was dried at 90 °C and calcinated at 450 °C for 3 h (at a heating rate of 2 °C min⁻¹). Nitrogen-doped TiO₂ (1, 3, and 5 wt % nitrogen) was prepared by the same method using an ammonia hydrolysis–condensation mixture. The ammonia solution (25%, Sigma Aldrich, St. Louis, MO, USA) was used as a nitrogen precursor (the calculated amount of ammonia water was diluted to keep the ratio nH₂O/nC₁₂H₂₈O₄Ti = 4:1). The other synthesis conditions were the same as in the case of pure TiO₂.

3.2. Catalysts Characterization

The X-ray diffraction (XRD) patterns were recorded with the use of a MiniFlex diffractometer (Rigaku, Tokyo, Japan) with a copper anticathode (λ = 1.54 Å). In the measurements, powder samples were placed on glass holders and scanned in the range from 10° to 90.0° in 0.01° steps with a scan speed at the level of 0.15°/min. Crystallite sizes were determined according to the Scherrer equation [39] with the use of Diffrac.Eva V 3.2, Billerica, MA, USA software.

Nitrogen content was measured with the use of the EDS method, described by Gazulla et al. [27]. Calibration standards used in this study: AgNO₃ (Merck, Kenilworth, NJ, USA), L-Lysine monohydrochloride (Alfa AEstar, Kandel, Germany), and D-Leucine (Roanal, Budapest, Hungary). Calibration curve R² (Figure S4) was at the level of 0.9606. Pictures of the surface were taken with a Jeol JSM-6610LVnx Scanning Electron Microscope (SEM, Oxford Instruments, Abingdon, GB), equipped with an Energy Dispersive Spectrometer (EDS) (Oxford Aztec Energy, Oxford Instruments, Abingdon, GB, England). Analysis parameters: a high vacuum mode, 20 kV, SS 60, and a 300 s acquisition time of the spectrum. The measurement result is the average value of 5 analyses. The fluctuations in the values for individual catalysts were as follows: 1N_TiO₂ (value: 0.328 ± 0.025 cps/eV), 3N_TiO₂ (value: 1.02 ± 0.08 cps/eV), and 5N_TiO₂ (value: 1.33 ± 0.15 cps/eV).

To determine the textural properties of the obtained catalysts, the nitrogen adsorption at 77 K was measured using ASAP 2020 (Micrometrics, Gwinath County, GA, USA). Before the analysis, all samples were outgassed at 423 K for 12 h under high vacuum (3.75∙10⁻³ Torr). The specific surface area (S_BET) was calculated by the Brunauer–Emmett–Teller (BET) method. Total pore volume (V_TOT) was derived from the amount of nitrogen vapor adsorbed at a relative pressure close to unity, with the assumption that the pores were then filled with liquid adsorbate. The average pore diameter (ADP) was obtained from the BJH (Barrett–Joyner–Halenda) desorption average pore diameter. The distribution of pore sizes was calculated using the BJH method.

UV-VIS diffuse reflectance spectra were recorded with the use of a V-750 spectrophotometer (Jasco, Tokyo, Japan) equipped with a PIV-756 integrating zone for diffuse reflection spectra recording. Powder samples were placed in the sample holder and scanned in the range of 250–800 nm. As a standard for testing, barium sulfate was used. The measurements were performed three times to eliminate an eventual error. The observed differences in spectral deviations did not exceed 2%. Band gap energy values were calculated from diffuse reflectance measurements transformed with the modified Kubelka–Munk equation [40].

The pH at which the catalyst surface charge takes a zero value (pH_PZC) was determined according to the method described by Moreno-Castilla et al. [41].

Pictures of the surface of the photocatalysts were taken with a Scanning Electron Microscope (SEM)—Jeol JSM-6610LVnx (Oxford Instruments, Abingdon, GB). The distribution of individual elements in the catalyst mass was determined with an Energy Dispersive Spectrometer (EDS) (Oxford Aztec Energy, Oxford Instruments, Abingdon, GB).

For determination of the thermal stability, thermogravimetric studies were performed using a Thermogravimetric Analyzer CKiC 5E-MAC6710 (Changsha, Hunan, China). Thermogravimetric analyses were performed under a constant argon flow (10 L∙h⁻¹), a heating rate at the level of 5 °C∙min⁻¹, and a temperature range of 20–900 °C. At the final temperature of 900 °C, a sample was stabilized for 10 min. The results were presented as a function of weight loss (%) and temperature. The analysis was performed twice for each of the catalysts. No differences between repeated samples were observed.

3.3. Photodegradation Procedure

Photodegradation was performed using a closed system with an inner-irradiation-type quartz reactor, presented in our previous work [42]. A medium pressure mercury lamp (power of 450 W, Ace Glass Inc., Vineland, NJ, USA) was used as the light source. The lamp was immersed in a quartz well with a cooling medium. Coolant temperature was kept at the level of 10 °C with the use of a flow cryostat (Julabo FP-50, Seelbach, Germany) to prevent the thermal degradation of the Ace K. Each experiment was performed according to the following procedure: 300 mL of the dye solution (100, 50, and 20 ppm concentration) was placed in the reactor with 0.3 g of the catalyst (with a catalyst concentration of 1.0 g∙L⁻¹). The suspension was cooled and mixed in the dark with a magnetic stirrer for 30 min. After half an hour of the adsorption process, the first sample of the solution was taken from the reactor (zero time), the dye concentration was estimated. After the adsorption process, the UV lamp was turned on. The efficiency of the photocatalysts was verified based on the samples taken every 15 min over the entire 3 h lamp operation. After catalyst separation, the dye concentration in the taken sample was estimated with a U-2001 Spectrophotometer (Hitachi High Technologies America, Inc., Pleasanton, CA, USA) at 225 nm, which is the Ace K peak maximum [43]. Based on the results, the photocatalytic degradation process efficiency was calculated by the following equation:

$$\mathrm{Photodegradation} \mathrm{efficiency} (\%)=\frac{{\mathrm{C}}_{\mathrm{t}}}{{\mathrm{C}}_0}\times 100\%$$

(3)

where C₀ and C_t denote the initial concentration and the concentration at time t, respectively.

For each catalyst, the photodegradation process was performed twice for confirmation. Variation of the results did not exceed 3% in the case of the applied initial concentrations.

4. Conclusions

Nitrogen-doped and undoped anatase catalysts synthesized by the sol-gel method were found to be efficient catalysts for Ace K photodegradation under UV irradiation. It was found that doping in the amount of 1.55–6.29 wt % causes a moderate increase in the catalyst surface area (48.55–58.23 m²∙g⁻¹), an increase in the width of the pores (ADP 3.71–6.86), and a reduction of the pH_PZC value (5.72–5.05). Electrostatic interactions between Ace K and the catalyst surface can play an important role in the adsorption as well as in the photodegradation efficiency. The most effective nitrogen-doped TiO₂ concentrations in the photodegradation process were 4.83 and 6.29 wt %. Ninety percent molecule removal was observed after 80–100 min of irradiation, depending on the initial Ace K concentration. A modified lattice structure probably limits the recombination of photogenerated electron/hole charge carriers, increasing photodegradation efficiency. Increasing S_BET also plays an important role due to the potential increase in the number of active sites where the photodegradation process take place.