Mechanochemically Synthesized Solid Solutions La_1−xCe_xFeO_3+x/2 for Activation of Peroxydisulfate in Catalytical Reaction for Tetracycline Degradation

Abstract

The synthesis of orthoferrites of the type La_1−xCe_xFeO_3+x/2, x = 0.00, 0.01, 0.03, 0.05, and 0.07, by applying a simple and effective mechanochemical transformation from the constituent oxides is presented. Physicochemical methods such as powder X-ray diffraction (XRD), transmission electron microscopy (TEM), UV–Vis spectroscopy, and Brunauer–Emmett–Teller (BET) adsorption were applied to gain information about the effect of Ce⁴⁺ content on the structural, textural, and optical properties of the samples. The catalytic activity of the samples for water decontamination was determined in a photo-Fenton-like activation of persulfate for removal of tetracycline hydrochloride as model pollutant. The presence of persulfate, PDS, considerably increased the removal efficiency under visible light illumination.

1. Introduction

The well-known and exploited methods for purification of waters, polluted by organic molecules, are based on physical [1], biological [2,3], and chemical [4] processes. Some of the oxidation processes apply oxidants such as peroxymonosulfate (PMS, SO₅²⁻) and peroxydisulfate (PDS, S₂O₈²⁻). They both produce sulfate radicals with high redox potential (2.6–3.1 V), capable of functioning in solutions with a wide range of pH, with long lifetimes in comparison with the hydroxyl radical [5]. In spite of the fact that PMS and PDS are oxidants, their reaction with contaminants is not possible, which is why an activation by a catalyst is needed, such as semiconductor, for example [6]. The metal-based catalysts for activation of PDS, as well as PMS, are considered cost-effective, requiring less energy, and relevant for application in both homogeneous and heterogeneous systems [6]. The activation involves a series of oxidizing processes through breaking the O–O bond [2]. Because PMS/PDS in water environments degrade slowly, they are tested for in situ degradation of contaminants in soil and groundwaters [5].

Some of the metal-based materials investigated for PMS and PDS activation are catalysts, containing iron [6,7,8,9]. Among them is the lanthanum orthoferrite, LaFeO₃, belonging to the group of the transition metal oxides with a formula ABO₃, where A is a rare-earth ion with 12 as a coordination number and B is a 3d transition metal with coordination number of 6, a p-type semiconductor with electrical and magnetic properties [10]. The LaFeO₃ shows visible-light-absorption properties with a value of the band gap depending on the method for synthesis, for example, 1.98 eV by combustion method [10], about 2.0 eV by melt-growing in a floating zone furnace [11], 2.07 eV by sol–gel route under ultrasonic treatment [12], and 2.15–2.30 eV by microwave plasma and calcination [13].

For synthesis of LaFeO₃, different methods based on simple and efficient chemical reactions have been tested, such as thermal decomposition of lanthanum ferrioxalate precursor [14], polymerizable complex method [15], and autocombustion synthesis [16,17]; some of them are reviewed in [18]. In addition, it was found that the perovskite structure of LaFeO₃ can be modified by different doping ions; by varying both the nature and the content of the doping agent, the properties of the final product can be modified and controlled [19,20]. The doping ions could occupy A or B site of LaFeO₃ so the modification of the properties depends on the dopant [21]. In order to synthesize LaFeO₃-based perovskite solid solutions, different ions have been tested, such as Gd(III) by solid casting route [22], Cu(II) by sol–gel method [23], Ga(III) and Al(III) by polymerization complex [24,25], and Ca(II) by citrate method [26]. Calcination at 500 °C was included in the procedure of the microemulsion method for La_1−xTb_xFeO₃ preparation [27]. By the combination of hydrothermal and mechanochemical procedures, Ru³⁺-doped LaFeO₃ with the Ru³⁺ ions substituting Fe³⁺ ions at the octahedral sites of the perovskite structure was prepared [28]. Modification of LaFeO₃ with Ce³⁺ by solution combustion and coprecipitation method showed that the solution combustion method generated better impact on the physicochemical properties of Ce-doped LaFeO₃ nanoparticles than the coprecipitation method [29,30]. Both the combustion and the coprecipitation method need time- and energy-consuming calcination for synthesis of Ce-doped LaFeO₃, 500 °C for 5 h and 700 °C for 4 h, respectively [30]. It is important to mention that under the conditions reported in [30], Ce³⁺ should be oxidized to Ce⁴⁺, according to [31]. By calcination at 850 °C, substitution of La³⁺ by Ce⁴⁺ in LaFeO₃ was accomplished [32]. According [29,30,32], an improvement of the structural, morphological, magnetic, and catalytic properties of materials based on LaFeO₃ could be accomplished by modification with cerium ions.

The mechanochemical ball-milling is a method considered suitable for synthesis of AFeO₃ (A = La, Pr, Nd, Sm), tested as catalysts [33]. It is reported as a synthetic method in a wide range of applications, especially taking into account that the method is leading to cleaner, safer, and more efficient chemical conversion [1,33]. Besides the advantages mentioned, the method could reduce the particle size of materials and create more defects and vacancies in the crystal lattice of materials [34]. The latter are considered especially important for photocatalysts because of their capability to limit the electron–hole recombination and react as reaction sites [35]. By decreasing particle size, increasing surface area with active sites, and modifying the morphology, the method could promote the Fe³⁺/Fe²⁺ conversion in Fenton- and Fenton-like reactions [36,37]. The mechanochemical ball-milling is considered suitable for synthesis of catalysts for water purification because it could cause surface activation of the materials, leading to increasing the reactive species in water, increasing the active sites, and increasing the pollutant absorption [36].

In the work presented, the influence of Ce⁴⁺ on the structure and properties of LaFeO₃ was investigated by studying solid solutions La_1−xCe_xFeO_3+x/2, which were mechanochemically synthesized by the initial oxides (x = 0.00, 0.01, 0.03, 0.05, and 0.07). Cerium was used as a dopant due to its existence in two stable oxidation states Ce³⁺/Ce⁴⁺ [31], with a possibility to be an advantage for the activity of the solid solutions La_1−xCe_xFeO_3+x/2 in a reaction for activation of PDS. In spite of the immense number of iron-based catalysts for activation of PDS, research on Ce–modified LaFeO₃ is not found in the literature available. Tetracycline hydrochloride is among the most widely used antibiotics [38]; at the same time, it is one of the common antibiotic pollutants [39], known to cause environmental problems by its biological toxicity, chemical stability, and stimulation of antibiotic resistance genes [40,41]. It was selected as the model pollutant in water solution and its degradation by catalytic reaction is a part of the discussion presented.

2. Materials and Methods

2.1. Materials

The chemicals used were La₂O₃ (≥99.9%, Sigma-Aldrich Chemie GmbH, Taufkirchen, Germany), CeO₂ (≥99.0%, Sigma-Aldrich Chemie GmbH, Taufkirchen, Germany), FeOOH (p.a., Valerus, Bulgaria), Tetracycline hydrochloride (Ficher BioReagents, Waltham, MA, USA), K₂S₂O₈ (ACS reagent, ≥99.0%, Sigma-Aldrich Chemie GmbH, Taufkirchen, Germany), and Na₂S₂O₃ (99%, Sigma-Aldrich, Sigma-Aldrich Chemie GmbH, Taufkirchen, Germany). The scavengers NaN₃(≥99.0%, Honeywell, Charlotte, NC, USA), tret-butanol (ACS reagent, ≥99.0%, Sigma-Aldrich Chemie GmbH, Taufkirchen, Germany), ascorbic acid (ACS reagent, ≥99.0%, Sigma-Aldrich Chemie GmbH, Taufkirchen, Germany), and methanol (ACS reagent, ≥99.8%, Riedel-de Haën, Charlotte, NC, USA) were used as well as Na₂S₂O₃(ReagentPlus^®, 99.0%, Sigma-Aldrich Chemie GmbH, Taufkirchen, Germany)as a quencher. Concentration of the scavenger solutions was kept constant for every scavenger and it was in great excess to the persulfate, namely, 100 mM.

2.2. Mechanochemical Synthesis of the Samples

The synthesis of LaFeO₃ can be represented by the simple equation La₂O₃ + Fe₂O₃ → 2LaFeO₃. The oxide La₂O₃ was first calcined at 900 °C for 12 h in order to remove hydroxide and carbonate phases occasionally included in the sample. The Fe₂O₃ was obtained from dehydrated FeOOH at 200 °C for 6 h. The oxides were mixed and homogenized by hand-grinding followed by addition of 0.03 g stearic acid (1 mass.% of the reaction mixture), playing the role of the surfactant. The mixture was placed in a planetary ball-mill (FRITSCH Pulverisette 7) with zirconia vials (45 mL) and balls (10 mm) for 4 h at 700 rpm at BPR 10:1. The milling protocol was 5 min of milling followed by 10 min of rest rolling orientation changing every rest cycle.

The synthesis of La_1−xCe_xFeO_3+x/2 can be represented by the simple equation

(1 − x)La₂O₃ + Fe₂O₃ +2xCeO₂ → 2 La_1−xCe_xFeO_3+x/2.

The calculated amount of CeO₂ was added so that La_1−xCe_xFeO_3+x/2 could be obtained, where x = 0.00, 0.01, 0.03, 0.05, 0.07. The samples obtained, LaFeO₃, La_0.99Ce_0.01FeO_3+x/2, La_0.97Ce_0.03FeO_3+x/2, La_0.95Ce_0.05FeO_3+x/2, and La_0.93Ce_0.07FeO_3+x/2, were characterized and used in catalytic reactions for tetracycline hydrochloride degradation.

A mixture of the initial oxides, La₂O₃, Fe₂O₃, and CeO₂ (in amount, calculated for La_0.93Ce_0.07FeO_3+x/2), was treated for 2 h by ball-milling using the same milling protocol in order to gain information about the milling time influence on the samples’ formation.

2.3. Methods for Characterization

X-ray diffraction, used to determine the crystal structure of the samples, was performed using a PANalytical Empyrean X-ray diffractometer(Malvern PANalytical Empyrean, Almelo, Netherlands) in the 2θ range of 15–80° by CuKα radiation (λ = 0.15405 nm), with steps of 0.01° and 20 s exposure time at each step. The structural and microstructural (crystallite size and microstrains) information was extracted using the full profile Rietveld method using the FullProf Suite software (v01-2021, Grenoble, France) [42].

UV–Vis absorption spectroscopy was applied using an Evolution 300 UV–Vis spectrometer (Thermo Scientific, Waltham, MA, USA) for measuring the absorption of the samples in the range of 200–900 nm.

Band gap energies were calculated from the UV–Vis absorption spectra in the range 200 to 400 nm. The UV–Vis data were analyzed for the relation between the optical band gap, absorption coefficient, and energy (hν) of the incident photon for near-edge optical absorption in semiconductors. The band gap energy was calculated using the measured curves by fits according to Tauc’s equation, αhν = A(hν − E_g)^n/2 [43], where A is a constant independent of hν, E_g is the semiconductor band gap, and n depends on the type of transition. The value used for n was 2, reflecting an indirect transition.

Textural characteristics, such as specific surface area, total pore volume, and pore size distribution, were determined at −196 °C using a TriStar II 3020 apparatus (Micromeritics, Norcross, GA, USA). The Brunauer–Emmett–Teller (BET) method was applied for specific surface area calculations. The pore size distributions were derived from the desorption branch of the isotherms employing the Barrett–Joyner–Halenda (BJH) method. The total pore volume was estimated at a relative pressure of 0.989.

2.4. Tetracycline Hydrochloride (TCH) Degradation

A model solution of TCH with concentration 20 ppm and a volume of 200 mL was used. A double-jacked Pyrex reactor was used, similar to one used by us earlier. After adding 0.1 g catalyst, a 60 min ”dark” period was followed in order to establish the equilibrium of the sorption process. After 68 mg K₂S₂O₈ was added to the solution (concentration of 1 mM to the total volume), it was illuminated by a Schott KL-2500 LED, emitting 5600 K cold light with light intensity of 1100 mL, immersed into the reaction vessel. A sample of 2.5 mL was taken and filtered through a 0.22 µm membrane filter to remove the catalyst, and 150 µL 0.05 M solution of Na₂S₂O₃ was added to quench the reaction. The concentration of TCH was determined by UV/Vis absorption spectroscopy, monitoring the peak at 354 nm. The data obtained were plotted in coordinates (C/C₀)/t and −ln(C/C₀)/t (where C₀ is the concentration after the “dark” period, and C is the concentration after t min irradiation), and apparent rate constants of the degradation process were determined assuming first-order kinetics. The degradation at moment t is determined by the formula degradation, % = (A₀ − A_t)/A₀ × 100, where A₀ is the initial absorption of the TCH solutions at t = 0 min and At is the absorption at t min.

3. Results

3.1. Characterization of the Samples

3.1.1. Phase Homogeneity

The XRD patterns of the mixture of the initial La₂O₃, Fe₂O₃, and CeO₂ (calculated for 7% Ce) before ball-milling, as well as after 2 and 4 h of ball-milling, were recorded (Figure 1). In order to follow the changes, literature data for the XRD of the pure initial oxides as well as for LaFeO₃ are included in Figure 1. As can be seen, the two most intense diffraction reflexes for CeO₂, namely, (111) and (220), at 28.5° and 47.8° 2θ, respectively, are well visible in the mixture before milling (Figure 1a). After 2 h of high-energy ball-milling, a significant size reduction of the starting oxides occurs, leading to increased peak broadening, but the (111) reflex of CeO₂ is still clearly visible (Figure 1b). In addition, the (112) diffraction peak of the orthorhombic LaFeO₃ is well expressed in the XRD of the 4 h ball-milling sample, indicating the crystallization of the material (Figure 1c). Based on these XRD data, the reaction mechanism can be briefly commented as involving a significant size reduction (even amorphization) of the starting oxides, followed by crystallization of the targeted material.

All the samples crystallized in the typical for LaFeO₃ orthorhombic Pbnm crystal group (ICDD #74–2203), shown by the comparison of the samples’ XRD diffractograms (Figure 2). No additional reflexes were observed of nonreacted oxides, which proves the successful synthesis of the homogeneous solid solutions.

The results of the Rietveld refinement are presented in Figure 3a–f, and summarized in

Table 1. Unit cell parameters, unit cell volume, crystallite size, and microstrains of the samples obtained by the Rietveld refinement, including parameters provided by the Rietveld refinement, the weighted R profile (R_wp), and the goodness of fit (χ²). * GOF = goodness of fit.

Sample	Unit Cell Parameters, Å	Unit Cell Volume, Å3	Crystallite Size, nm	Microstrains, ×10−3 a.u.	Dislocation Density, ×10−3 nm−2	Rwp, %	GOF *
LaFeO3	a = 5.552(4) b = 5.577(4) c = 7.843(5)	242.9(3)	14.4(4)	1.1	4.8	12.0	1.14
La0.99Ce0.01FeO3+x/2	a = 5.550(4) b = 5.578(5) c = 7.843(5)	242.8(3)	14.9(3)	0.9	4.5	12.1	1.12
La0.97Ce0.03FeO3+x/2	a = 5.551(3) b = 5.578(3) c = 7.844(4)	242.9(3)	15.0(2)	0.8	4.4	12.3	1.17
La0.95Ce0.05FeO3+x/2	a = 5.550(5) b = 5.575(7) c = 7.846(9)	242.8(5)	16.8(4)	0.8	3.5	11.9	1.16
La0.93Ce0.07FeO3+x/2	a = 5.549(5) b = 5.576(6) c = 7.848(7)	242.8(4)	16.9(5)	0.7	3.5	12.6	1.21

. As can be seen (Figure 3a–e), the solid solutions show some intensity peaks deviation in Rietveld refinement for the diffraction at 32° 2θ as Ce⁴⁺ content increases. The deviation is due to not releasing the oxygen occupancy. Rietveld refinement with an oxygen occupancy freely refined is shown in Figure 3f. The same procedure was applied for the other samples with different Ce⁴⁺ content. The values of all samples, although being different, are within the standard deviation. Considering the high background level (due to fluorescence of the Fe under CuK_α radiation) and the high degree of microstructural effects (low crystallites size), which can all cause deviation of observed peaks intensity, the values of the oxygen content purely based on the Rietveld refinement of the samples are not reported.

The addition of cerium does not lead to any meaningful changes to the unit cell parameter, which is somehow expected considering the similar ionic radii of La³⁺ and Ce⁴⁺, 1.03 Å and 0.87 Å, for lanthanum and cerium, respectively, with coordination number of 6 [44]. Despite that, the introduction of cerium leads to increased crystallites size, which could be attributed to the stabilization of the orthorhombic phase with the introduction of the smaller Ce⁴⁺ ion. The latter leads to a lower value of the theoretical perovskite tolerance factor of 0.85 compared to the 0.87 for the pure CeFeO₃ and LaFeO₃, respectively. This causes the lower value of the microstrains and dislocation density and therefore lowers the defects concentration, which facilitates the crystal growth (Table 1).

3.1.2. Morphology of the Samples

The TEM images of the plain LaFeO₃ (Figure 4a) and of the sample with highest Ce content La_0.93Ce_0.07FeO_3+x/2 (Figure 4b) show samples made up of agglomerates of irregular-shaped crystallites with size between 10 and 20 nm. This is consistent with the XRD analysis. The calculated d-spacing from both SAED and HRTEM shows that the incorporation of Ce does not lead to any changes of the unit cell parameters. The values for d-spacing are integrated in the images in Figure 4a–d.

3.1.3. Textural Characteristics of the Samples

The adsorption/desorption isotherm of the host perovskite matrix of LaFeO₃ is of type II according to IUPAC classification with hysteresis loop type H3, showing the prepared material is nonporous or macroporous with nonrigid aggregates of plate-like particles (Figure 5). The introduction of small quantities of Ce⁴⁺ to the host changes neither the type of isotherm nor the hysteresis type, and preserves the structure of the materials. However, the increased Ce⁴⁺ concentration tends to slightly decrease the specific surface area as well as the total pore volume of the prepared perovskites (

Table 2. Textural characteristics of the samples.

Sample	Specific Surface Area SBET, m2/g	Total Pore Volume Vt, cm3/g	Average Pore Size Dav, nm
LaFeO3	9	0.04	19
La0.99Ce0.01FeO3+x/2	9	0.04	16
La0.97Ce0.03FeO3+x/2	8	0.04	18
La0.95Ce0.05FeO3+x/2	8	0.04	19
La0.93Ce0.07FeO3+x/2	7	0.03	15

). This result corresponds well with the data obtained from the XRD, which show increased crystallite size of the Ce-modified samples in comparison to the bare LaFeO₃. With regard to the SSA, LaFeO₃ is known for its low values of the specific surface area of bulk material [18]. The value of 9 m²/g for the mechanochemically synthesized LaFeO₃ is similar to the sol–gel method synthesized 9.5 m²/g [12]. In spite of the attempt to increase the SSA by doping LaFeO₃ with Ce⁴⁺, we did not gain any essential improvement; the values obtained were 7–9 m²/g.

The average pore size for the samples LaFeO₃, La_0.99Ce_0.01FeO_3+x/2, La_0.97Ce_0.03FeO_3+x/2, and La_0.95Ce_0.05FeO_3+x/2 has an equal value, while that of La_0.93Ce_0.07FeO_3+x/2 is a bit smaller (Table 2). The maximum in the pore size distribution is at about 30–40 nm (Figure 5b).

3.1.4. UV/Vis Spectroscopy

The absorption spectra of the samples are predominantly in the visible region, absorbing up to about 550 nm which is the green color of the visible light (Figure 6). Based on these UV/Vis spectra, the band gap energy Eg was calculated after Tauc equation [43] for all the samples (

Table 3. Energy of the band gap, refractive index, and the potentials of current band (CB) and valence band (VB).

Sample	Eg, eV/λ, nm	Refractive Index	EVB, eV	ECB, eV
LaFeO3	2.32/533	2.61	2.21	−0.11
La0.99Ce0.01FeO3+x/2	2.38/520	2.59	2.24	−0.14
La0.97Ce0.03FeO3+x/2	2.37/522	2.59	2.23	−0.14
La0.95Ce0.05FeO3+x/2	2.39/518	2.58	2.24	−0.15
La0.93Ce0.07FeO3+x/2	2.36/524	2.59	2.23	−0.13

). The values for the band gap energy of the solid solutions La_0.99Ce_0.01FeO_3+x/2, La_0.97Ce_0.03FeO_3+x/2, La_0.95Ce_0.05FeO_3+x/2, and La_0.93Ce_0.07FeO_3+x/2 are in the range 2.36–2.39 eV (corresponding to 524–520 nm), which is a bit higher than the value for the pure LaFeO₃, 2.32 eV (corresponding to 533 nm). A similar tendency of the band gap to increase with increasing Ce content is observed for La_1−xCe_xFeO₃ obtained by solid-state reaction of the oxides, namely, from 2.22 to 2.41 eV [45].

The values of E_g for the solid solutions change insignificantly with the Ce⁴⁺ addition, and the corresponding wavelengths 524–520 nm clearly illustrate that they are exactly in the visible range of the light. They prove that the solid solutions are capable to act as photocatalysts under visible light illumination, similar to the LaFeO₃ itself.

The potentials of current band (CB) and valence band (VB), E_CB = X − 4.5 − 0.5E_g and E_VB = E_CB + E_g, are calculated, where E_g is the band gap energy and X is the absolute electronegativity of the material (Table 3). The energy of the free electrons on the hydrogen scale E₀ of 4.5 eV is used in the potentials calculations [46]. The results for potentials show positive valence band, which is similar to the literature data [18]. The negative value for the CB for the pure LaFeO₃, −0.11 eV, differs from the literature data (positive 0.2 eV [18]), quite likely because of the difference in the synthesis procedure. The values for the solid solutions are close to those of the pure LaFeO₃, 2.23–2.24 eV for VB; those for CB are also negative, varying between −0.13 to −0.15 eV. The refractive index (Table 3) is calculated by the modification of the Dimitrov–Sakka equation [47]:

$$\mathrm{n}=\sqrt{\frac{3}{\sqrt{E_g/20)}}-2}$$

where n is the refractive index and E_g is the band gap energy calculated after Tauc equation [43]. Considering the refractive index of the samples, the pure LaFeO₃ with a higher value is optically denser, possessing lower absorbance.

3.2. Catalytic Decomposition of TCH

The results demonstrate that all the solid solutions of La_1−xCe_xFeO_3+x/2 characterized are active as catalysts for TCH decomposition (Figure 7a). They show monotonous increasing of the catalytical activity with increasing Ce⁴⁺ content, where the highest value for the rate constant for the most active sample, La_0.93Ce_0.07FeO_3+x/2, is 29.2 × 10⁻³ min⁻¹ (

Table 4. Rate constants, degradation of TCH and TOC removal under visible light with PDS activated.

Catalyst	Rate Constant, ×10−3 min−1	R2	Degradation, %	TOC Removal, %
LaFeO3	16.4 ± 0.9	0.972	62.0	41.2 ± 0.3
La0.99Ce0.01FeO+x/23	18.8 ± 0.6	0.989	64.6	45.1 ± 0.7
La0.97Ce0.03FeO3+x/2	20.0 ± 1.1	0.975	67.3	47.0 ± 0.5
La0.95Ce0.05FeO3+x/2	22.6 ± 1.4	0.969	69.1	51.2 ± 1.1
La0.93Ce0.07FeO3+x/2	29.2 ± 2.5	0.946	72.9	61.0 ± 0.9

). Taking into account the band gap energy values, it should be pointed out that they are not varying monotonous with increasing Ce⁴⁺ content, but are very similar, and the essential point is that all they correspond to the visible light range (Table 3). The E_g cannot be considered as a decisive reason for the different catalytic activity of the samples. A sol–gel-synthesized LaFeO₃ is tested in a similar process, even though the rate constant is not indicated, making the comparison of the catalytical activity not relevant [48].

The degradation of TCH, determined by UV/Vis absorption spectroscopy monitoring of the band at 354 nm, is increasing with the Ce⁴⁺ increasing in the range 62 to 73% (Table 4). Evidence for the oxidation of TCH to CO₂ and H₂O is the total organic carbon (TOC) removal up to 61% (Table 4). By UV–Vis spectroscopy, only the targeted pollutant can be followed, ignoring the intermediate products (in fact they can be even more toxic than the original molecule itself), while by TOC analysis, the complete mineralization can be followed. Although the data obtained by both analyses are close (for the most active sample La_0.93Ce_0.07FeO_3+x/2 of 73/61%, respectively), a difference can be expected since the oxidation kinetics of organic molecules is different [49]. For example, TCH degradation of 93.3% by UV–Vis spectroscopy, but only 58% after TOC analysis, was detected by [49]. According [50], by HPLC, almost a complete removal of TCH in persulfate Fenton-like reaction was observed, whereas by TOC-analysis, only 26.27% of the organic matter mineralization was detected. The generated intermediates (and therefore the degradation pathway) are strongly affected by the reactive oxygen species generated [51], but in almost all cases the first step is removal of the amino, hydroxyl, and methyl groups, followed by rings opening [52,53]. Despite the generation of various organic byproducts, the toxicity analysis performed both by Toxicity Evaluation Software Tool (TEST) [53] and by experiments with different biological species [52] prove that almost all of the intermediates generated by peroxydisulfate Fenton-like processes have higher LD₅₀ and lower mutagenicity than the TCH itself.

The influence both of the light and the catalyst, as well as the potassium peroxydisulfate, was evaluated. It can be seen (Figure 7b) that the oxidizing capability of K₂S₂O₈ is insignificant even in the presence of visible light. It essentially increased in the presence of the catalysts La_1−xCe_xFeO_3+x/2. When PDS is assisted both by a catalyst and visible light, it is activated and TCH is successfully degraded (Figure 7b).

Considering a radical-dominated catalytic mechanism, radical quenching experiments were applied to determine the types of active radical species generated during the activation of PDS so as to determine both the mechanism of the reaction and the active oxidizing radicals. The scavengers are supposed to react fast and explicitly with a radical, producing a stable species which does not affect the reaction, and by this, removing the effect of the radical in the degradation [54]. Alcohols, such as methanol and tert-butanol (TBA), are applied as hydroxyl radical scavengers, azide ions N₃⁻ as scavengers for singlet oxygen, even it is not a radical [55], and ascorbic acid as scavenger for superoxide anion radical O₂^•− [56] in photoassisted processes.

The molecules mentioned, known as good scavengers for SO₄^•− and OH^• (methanol), OH^• (tert-butanol), singlet ¹O₂ (sodium azide), and superoxide anion radical O₂^•− (ascorbic acid), were tested and the results are presented in Figure 8a. The results show the ascorbic acid had the strongest influence (O₂^•−), but in spite of that, all the radicals tested can be considered responsible for the activity of the catalysts. Superoxide radical and singlet oxygen may be the controlling reactive oxygen species.

4. Discussion

It can be supposed that the substitution of La³⁺ by Ce⁴⁺ in LaFeO₃ is causing deformation of the structure, including deformation of FeO₆ polyhedra. The data obtained based on XRD of the samples show bigger crystallites size, and lower degree of microstrain, leading to low concentration of defects and dislocations with Ce⁴⁺ content increasing in the solid solutions La_1−xCe_xFeO_3+x/2. The concentration of dislocations is connected with the defects available; the defects within the perovskites are oxygen vacancies. Both the oxygen vacancies and their concentrations are ruling the properties of the perovskites [57] and they can be generated by planetary ball-milling [58]. The specific surface area is not even high, but it decreases with the crystallites size increasing when the content of Ce⁴⁺ is increasing. At the same time, the data from the catalytic process show increasing of the catalytic activity of the solid solutions La_1−xCe_xFeO_3+x/2 with increasing Ce⁴⁺ content. According to the literature, both the defects as active centers [59] and the specific surface area [18] can be factors for the catalytic activity increasing, but for our solid solutions, the data from the characterization show that neither the defects as active centers nor the specific surface area could explain the increasing activity with increasing Ce⁴⁺ content. We should take into account that the partial replacement of La³⁺ by Ce⁴⁺ needs charge compensation. This could happen by Fe³⁺ → Fe²⁺ reduction. The higher concentration of the radicals can be explained by the higher content of Fe²⁺, i.e., by the reaction Fe³⁺ → Fe²⁺, the peroxide radicals are formed. An additional path for increasing of the peroxide radical formation could be the reduction of Ce⁴⁺ → Ce³⁺. This could be a result of a mechanochemical reaction because no heating was applied. According to [58], by planetary ball-milling of CeO₂ the concentration of Ce³⁺ significantly increased in comparison with Ce⁴⁺ by the grinding time (up to 720 min; for our samples, 240 min).

Based on the above presented results and considerations, the possible visible-light-assisted PDS activation mechanism is proposed by the following equations.

Catalyst + hν → Catalyst* + h⁺ + e⁻

≡Fe³⁺ + e⁻ → ≡Fe²⁺

≡Fe²⁺ + S₂O₈²⁻ → Fe³⁺ + SO₄²⁻ + SO₄^•−

Fe²⁺ + O₂ → Fe³⁺ + O₂^•−

≡Ce⁴⁺ + e⁻ → ≡Ce³⁺

≡Ce³⁺ + S₂O₈²⁻ → Ce⁴⁺ + SO₄²⁻ + SO₄^•−

Ce³⁺ + O₂ → Ce⁴⁺ + O₂^•−

S₂O₈²⁻ + e⁻ → SO₄^•− + SO₄²⁻

S₂O₈²⁻ + H₂O → 2SO₄²⁻ + HO₂⁻ + 3H⁺

S₂O₈²⁻ + HO₂⁻ → SO₄²⁻ + SO₄^•− + O₂^•− + H⁺

SO₄^•− + H₂O/OH⁻ → OH^• + SO₄²⁻ + H⁺

O₂^•− + e⁻ + 2H⁺ → H₂O₂

O₂^•− + OH^• → ¹O₂ + OH⁻

2O₂^•− + 2H⁺ → ¹O₂ + H₂O₂

2O₂^•− + H₂O → ¹O₂ + H₂O₂ + 2OH⁻

TCH + h^+/O₂^•−_/SO₄^•−/OH^•/1O₂ → … → CO₂ + H₂O

The solid solutions were excited by the visible light to produce electron–hole pairs. The photogenerated electrons, e⁻, were transferred to the conduction band from the valence band, creating positive holes, h⁺. The photogenerated holes can directly degrade TCH (Figure 8b). The photogenerated electrons can react with O₂ to produce O₂^•−, thereby generating OH^• and SO₄^•−. Additionally, the Fe³⁺ ions, by accepting electrons, make available Fe²⁺, and the latter activates PDS to produce SO₄^•− radicals and O₂ to produce O₂^•−. By the Fe³⁺/Fe²⁺ cycle, continuous PDS activation takes place. At the same time, the photogenerated electrons could accelerate the Ce⁴⁺/Ce³⁺ redox cycle, leading to SO₄^•− and O₂^•−. The singlet oxygen ¹O₂ production is a result of the peroxide ion reactions. All active species produced in the visible-light-assisted PDS activation play a positive role in the pollutant degradation to CO₂ and H₂O.

5. Conclusions

Homogeneous solid-state solutions with orthorhombic crystal structure, La_1−xCe_xFeO_3+x/2, were synthesized by very simple and effective mechanochemical treatment of the oxides La₂O₃, Fe₂O₃, and CeO₂, later added as a dopant agent in different ratio La/Ce. The successful application of the mechanochemical ball-milling for La_1−xCe_xFeO_3+x/2 synthesis can be considered as one of the good points of the work presented.

The Ce⁴⁺ content increasing leads to increase of the crystallite size and decrease of SSA and the total pore volume. In the reaction of decomposition of tetracycline hydrochloride, the solid solutions showed more increasing activity for PDS activation than the pure LaFeO₃, in spite of the very similar values of the band gap energy varying between 2.32 and 2.39 eV for all of them in the visible light range. The monotonous increasing of the rate constants with the Ce⁴⁺ content increasing illustrated that the catalytic activity is governed by the amount of Ce⁴⁺ in the solid solutions. The samples expressed excellent catalytic response with the best result for La_0.93Ce_0.07FeO_3+x/2, showing about 73% degradation of the TCH within 60 min of the reaction time and rate constant of 29.2 × 10⁻³ min⁻¹. By use of molecules–scavengers, the mechanism of the catalytic reaction was detailed. The experiments with the radical scavengers showed that active oxidant species such as O₂^.−, OH^•, and h⁺ were all involved in this catalytic system, although it seems that O₂⁻ played the major role in the catalytic degrading of TCH. The effective activation of PDS in a reaction of catalytic antibiotic degradation confirms that the solid solutions of La_1−xCe_xFeO_3+x/2 are also worthy of experimentation for decomposition of other organic molecules as water pollutants.