Role of SrCO₃ on Photocatalytic Performance of SrTiO₃-SrCO₃ Composites

Abstract

Perovskites such as SrTiO₃ are interesting for photocatalytic applications due to their structure-related and electronic properties. These properties are influenced by the presence of SrCO₃ which is often formed simultaneously during the hydrothermal synthesis of SrTiO₃. In this study, SrTiO₃-SrCO₃ composites with different contents of SrCO₃ (5–24 wt%) were synthesized. Their morphological, structural, and optical properties were investigated using complementary methods such as scanning electron microscopy (SEM), X-ray diffraction (XRD), nitrogen sorption, and diffuse reflectance spectroscopy (DRS). Their photocatalytic activity was assessed during the degradation of diclofenac (DCFNa) in aqueous solution and CO₂ photoreduction under Xe lamp irradiation. Improved photocatalytic efficiency in DCFNa degradation was observed for all the studied composites in comparison with SrTiO₃, and the highest mineralization efficiency was obtained for the sample with 21 wt% SrCO₃ content. The presence of SrCO₃ led to an increased concentration of active species, such as ^•OH radicals. Otherwise, its presence inhibits CH₄ and C₂H₆ production during CO₂ photoreduction compared with pure SrTiO₃.

1. Introduction

Titanium-based materials, such as titania, alkaline earth metal titanates, etc., are considered as suitable candidates for photocatalytic applications [1,2]. Among these, SrTiO₃ seems to be a promising candidate given its band-structure [3], considerable catalytic activity, and chemical and thermal stability, and the long lifetime of the photogenerated charge carriers [4]. One unique property of SrTiO₃ is directly related to its slightly different conduction band (CB) edge situated at a more negative potential in comparison with anatase (i.e., E_CB(anatase): −0.1 V, E_CB(SrTiO₃): −0.3 V vs. NHE at pH = 0 [5]), which has a direct influence on its water splitting ability. Although the CB position of TiO₂ (both anatase and rutile) does not allow efficient H₂ production (from H₂O) unless in the presence of co-catalysts, this limitation is overcome in the case of SrTiO₃ [6,7].

Moreover, SrTiO₃ is one such photocatalyst which exhibits high activity towards both reduction (Cr⁶⁺/Cr³⁺ [8], CO₂/solar fuels [9], H₂O/H₂ [10]), and oxidation (photodegradation of contaminants [11]). SrTiO₃ is a well-known ABO₃-type model perovskite with ideal cubic structure (geometric tolerance factor ~1) at room temperature [12], possessing mixed ionic-covalent bonding properties [13], which confer excellent electronic features. Having a closer look at its unit cell, along the crystallographic axis two different types of alternating atomic planes (SrO and TiO₂) can be distinguished, which exhibit different acid-base and electronic properties [9,14]. Although SrTiO₃ exhibits favorable photocatalytic features itself, coupling with noble metals (i.e., Ag [15], Pt [16]) or with semiconductors (i.e., TiO₂ [17], Bi₂S₃ [18]) or insulators (i.e., SrCO₃ [19,20,21,22,23]) leads to the enhancement of catalytic activity of such composites given by the manifested synergistic effect [24].

Several methods are reported in the literature for the synthesis of SrTiO₃, such as solid-state synthesis [25], solvothermal route [11,26], sol-gel [27], molten salt reaction [28], and electrospinning [29]. Among these, hydrothermal synthesis is one of the most promising not only because of mild reaction conditions and efficient control of crystal growth and morphology [11], but also because of simultaneous (inevitable) formation of SrCO₃ (as byproduct), thus resulting in SrTiO₃-SrCO₃ heterojunctions [30,31]. SrTiO₃-SrCO₃ is considered as one of the most useful composites for photocatalytic applications, such as NO oxidation [19], CH₄ oxidation [20], H₂ generation (H₂O splitting) [21], degradation of methylene blue [22], and photoelectrochemical CO₂ reduction [23], where the SrCO₃ content has considerable importance. The efficiency of SrTiO₃-SrCO₃ in photocatalytic applications can be explained mainly by the improved charge separation provided by SrCO₃, which acts as an electron trap. Since the photogenerated electrons from the CB of SrTiO₃ are trapped by SrCO₃ (and SrCO₃ is not excited given its relatively large band gap), the recombination rate of charge carriers is quite decreased in such n–n type heterojunctions [19].

Dissolution of airborne CO₂ in the reaction mixture [11,32,33,34], the initial molar ratio between the Sr²⁺:Ti⁴⁺ precursors [22], and the carbonate content of the applied base [35,36] are potential factors which can influence the amount of SrCO₃ formed during the hydrothermal synthesis of SrTiO₃. Furthermore, it should be highlighted that the control of the carbonate content of the alkaline reagent is challenging not only because of its high CO₂ absorptivity, but also because of the uniformity of the carbonate content. Beyond the previously mentioned one-step preparation of SrTiO₃-SrCO₃ via solvothermal synthesis [20,22], in-situ pyrolysis [19], sol-gel [20], and solid-state methods [20] are also reported in the literature as methods to obtain the mixed material.

In this work, SrTiO₃-SrCO₃ composites with different SrCO₃ contents were synthesized and applied for DCFNa photodegradation and CO₂ photoreduction. Preliminary purification of KOH in association with the conduction of synthesis in Schlenk line technique, excess of Sr²⁺ source, and different Ti⁴⁺ sources (i.e., anatase vs. P25) were applied, aiming for the control of SrCO₃ content of SrTiO₃-based catalysts. According to our best knowledge, the combination of preliminary purification of KOH with Schlenk line technique and the application of different Ti⁴⁺ precursors for the adjustment of SrCO₃ content of the SrTiO₃-based catalysts have not been reported yet in the scientific literature. It is demonstrated that the presence of SrCO₃ increases both the DCFNa degradation and its mineralization. Furthermore, some reasons for the higher activity of SrTiO₃-SrCO₃ compared with SrTiO₃ are discussed.

2. Results and Discussion

2.1. Structure, Morphology, and Textural Properties

The powder XRD patterns of the studied catalysts are presented in Figure 1. The commercially available SrTiO₃ (cSTO) was considered as the reference catalyst, showing only reflections of the SrTiO₃ phase (ICDD 00-035-0734). The notation used for the hydrothermally synthesized samples is STO_x_SCO, where x represents the SrCO₃ content (5, 15, 21 and 24 wt%) of the samples determined via Rietveld analysis based on the recorded XRD patterns. The presence of the cubic phase of SrTiO₃ is confirmed in the recorded XRD patterns. The additional reflections identified were assigned to the orthorhombic SrCO₃ phase (ICDD 01-084-1778). Considerable differences were revealed in the SrCO₃ content of the samples as a function of the applied strategy (

Table 1. Summary of the applied strategy aiming for the control of SrCO₃ content and the primary crystallite size (PCS) of components for all tested photocatalysts (STO_x_SCO, x = 5, 15, 21, and 24 wt%).

No.	Samples	Strategy/Observation	PCSSTO (nm)	PCSSCO (nm)
1	cSTO	—	28	—
2	STO_5_SCO	Preliminary purification of KOH and Schlenk line technique	18	49
3	STO_15_SCO	Starting synthesis	25	55
4	STO_21_SCO	rn(Sr2+:Ti4+) = 1.25	20	54
5	STO_24_SCO	Modification of Ti4+ source	19	57

The content of SrCO₃ calculated from the C content (determined by elemental analysis) was in concordance with the results obtained based on Rietveld analysis.

, see detailed description in the Experimental Section). As was expected, the lowest SrCO₃ content (i.e., 5 wt%) was assessed when the alkaline reagent had undergone preliminary purification, and the synthesis was conducted under inert conditions. Furthermore, when 25% excess Sr²⁺ source was introduced into the initial reaction mixture, increased SrCO₃ content (i.e., 21 wt%) was determined in the final product. Finally, the usage of P25 (instead of anatase) as the Ti⁴⁺ source had considerable influence on the SrCO₃ content (15 vs. 24 wt% when Ti⁴⁺ source was anatase vs. P25, respectively). Previous studies highlighted the influence of the crystal structure of the TiO₂ precursor (anatase, rutile, or amorphous) on the hydrothermal crystallization of SrTiO₃ [37,38,39] (given the different stability of the precursors [40]), although without referring to the formation and content of SrCO₃.

The primary crystallite size (PCS) of components is summarized in Table 1. The PCS of SrTiO₃ was in the range of 18–28 nm, whereas much higher values (49–57 nm) were obtained in the case of SrCO₃.

According to Figure 2, different SrCO₃ content, and more implicitly different synthesis parameters, have only moderate influence on the PCS of SrTiO₃ and SrCO₃. In each case the PCs of SrCO₃ was higher than that of SrTiO₃.

The SEM micrographs of the selected hydrothermally synthesized SrTiO₃-based samples, namely STO_15_SCO and STO_24_SCO, are presented in Figure 3. No considerable differences are revealed in the case of the selected STO-based samples regardless of the Ti⁴⁺ precursor type (P25 or anatase). Two characteristic morphologies were identified, namely nanocubes and microrods (marked with white dotted line). The length of these microrods was situated between 1–1.5 μm, and their width was identified in the 165–400 nm range. The characteristic dimension of the nanocubes varied in the range of 40–80 nm, which is in agreement with the results reported in the literature [4,38].

The lowest specific surface area (BET-SSA) was assigned to the commercial SrTiO₃ (

Table 2. Specific surface area and optical properties of the studied catalysts.

Samples	SSA-BET (m2∙g−1)	ΔEg (eV)	λthres (nm)	Ethres (eV)
cSTO	15	3.23	374	3.32
STO_5_SCO	16	3.18	371	3.34
STO_15_SCO	45	3.16	378	3.28
STO_21_SCO	33	3.18	379	3.27
STO_24_SCO	41	3.16	377	3.29

Abbreviations: ΔE_g—band gap, λ_thres and E_thres—wavelength and energy corresponding to the absorption threshold.

), because of its spherical morphology, which is characterized by the lowest surface area to volume ratio. Furthermore, relatively low SSA was determined in the case of STO_5_SCO in comparison with the other hydrothermally synthesized samples (i.e., STO_x_SCO, where x: 15, 21, 24). According to Table 2, there is no direct correlation between the SSA and SrCO₃ content of the hydrothermally synthesized SrTiO₃-based catalysts, but the presence of higher SrCO₃ content seems to enhance the SSA. Similar observations were presented in the work of Marquez-Herrera and coworkers [22].

2.2. Optical Properties

The absorption spectra of the studied samples were recorded in the UV-Vis range (Figure 4a). As expected, the STO-based samples absorb photons mainly from the UV-range (200–400 nm). The band gap energy values (ΔE_g) were calculated from the Tauc-plot (Figure 4b) considering the appropriate exponent (0.50) for indirect band gap semiconductors [4]. The well-known procedure (i.e., extrapolation of the linear region of the Tauc-plot) was applied, which led to ΔE_g values situated in the range of 3.1–3.2 eV (Table 2). The ΔE_g values determined for commercial and hydrothermally synthesized SrTiO₃ are in agreement with the values reported in literature [41]. No considerable shift in the absorption edge can be observed in the case of samples with different SrCO₃ content in the recorded range, since SrCO₃ absorbs electromagnetic radiation characterized with relatively high energy (4.9 eV [22]), which cannot be resolved within the recorded range. Furthermore, the absorption threshold (λ_thres, E_thres) of the studied catalysts was determined based on the first derivative reflectance spectra and summarized in Table 2. As was expected, slightly higher values were assigned for the absorption threshold energies in comparison with the ΔE_g.

2.3. Photocatalytic Results

2.3.1. Photocatalytic Diclofenac Degradation

The photocatalytic activity of the SrTiO₃-SrCO₃ composites was assessed by the degradation of DCFNa (C₀ = 25 mg∙L⁻¹) in aqueous solution under white light irradiation. The degradation curves are presented in Figure 5a. 30 min were allocated for the establishment of adsorption-desorption equilibrium (stirring, dark) before turning on the light. Neglectable DCFNa adsorption can be observed in case of all the SrTiO₃-based catalysts after the dark phase, which can be explained by the relatively low differences in SSA (15–45 m²∙g⁻¹). The degradation of DCFNa was almost complete after 1 h irradiation in all studied samples (inclusively the commercial SrTiO₃ and SrTiO₃-SrCO₃ composites). This indicates that the presence of SrCO₃ mainly affects the mineralization, i.e., total oxidation of DCFNa. The degradation curve can be described by pseudo first-order kinetics (see Figure S1, Supplementary Material) as previously reported [42]. The fastest DCFNa degradation was achieved for STO_5_SCO and STO_21_SCO. Here, the rate constant was four times higher than that of pure SrTiO₃ (cSTO). Beside DCFNa degradation, the effect of SrCO₃ content on mineralization was also investigated after 4 h. The mineralization of DCFNa during photolysis was below 3%, which was in agreement with previously reported results [43]. According to Figure 5b, higher mineralization efficiencies (X_min) were observed in case of samples containing tSrCO₃ phase in comparison with cSTO. Although quite similar mineralization efficiencies were assessed in the cases of STO_5_SCO and STO_15_SCO (~50%), the highest efficiency was reported in the case of STO_21_SCO (62%). Similar optimal SrCO₃ content (i.e., 19 wt%) of the SrTiO₃-SrCO₃ catalyst was determined by Marquez-Herrera during the photodegradation of methylene blue [22].

To obtain information about the role of active species (namely h⁺, ^•O₂⁻, ^•OH) in DCFNa degradation, experiments were conducted with the addition of scavengers (ammonium oxalate—AO, 1,4-benzoquinine—PBQ, and isopropanol-IPA) [44] using STO_15_SCO as catalyst (Figure 6). According to Figure 6., the rate of DCFNa degradation only slightly decreased when AO, PBQ, or IPA were added to the reaction mixture, which indicated that none of the active species had considerable influence on the transformation of DCFNa in this case. However, comparing the effects of scavengers, the involvement of ^•OH was the most pronounced in comparison with h⁺ and ^•O₂⁻.

To investigate whether the presence of SrCO₃ can influence the amount of reactive ^•OH formed during irradiation, the terephthalic acid hydroxylation reaction was carried out using STO_15_SCO and cSTO (Figure 7a). Considering the proportionality between concentration and photoluminescence (PL) intensity, the amount of ^•OH was 3.5 times higher in the reaction mixture using STO_15_SCO in comparison with cSTO. Since improved charge separation takes place in SrTiO₃-SrCO₃ composites [19] (Figure 8), more holes are left over SrTiO₃, which may lead to the formation of higher amounts of ^•OH. However, this is just evidence for better charge separation, and not for the fact that ^•OH are the main active species responsible for the photodegradation of DCFNa.

Furthermore, to find out whether beside ^•OH, further radicals were formed during the photocatalytic reaction, in-situ EPR measurements were conducted with DMPO as a trapping agent using cSTO and STO_15_SCO (Figure 7b). Analyzing the g value (2.007) and the hyperfine splitting constants (a_N/a_H = 1), the DMPO/^•OH adduct (a_N/a_H = 1 [45]) was identified in the reaction mixture when cSTO and STO_15_SCO were used (Figure 7b). When STO_15_SCO was used, signals of a second radical were additionally observed, which can be attributed to the formation of a DMPO/CO₃^•− adduct (a_N/a_H = 1.37 [45]). Analyzing the redox potentials of CO₃^•−/CO₃²⁻ (1.54 V vs. NHE) and H₂O/^•OH (2.8 V vs. NHE) redox pairs, it is obvious that CO₃^•– is a weaker oxidizing agent than ^•OH [46]. Even though the carbonate anion radical was detected in only small amounts when SrTiO₃-SrCO₃ was used (Figure 7b), and it is a thermodynamically weak oxidizing agent, these facts do not exclude its possible involvement in the surface reactions (i.e., it may be a kinetically fast reaction partner). Moreover, the previously presented scavenging experiments performed using STO_15_SCO (i.e., the addition of ^•OH scavenger—IPA has reduced effect on the degradation rate, Figure 6) can be possible evidence for the specific role of the carbonate anion radical when SrTiO₃-SrCO₃ was used.

According to the scientific literature, the formation of carbonate anion radical can take place (1) through the reaction between hydroxyl radicals and carbonate anions [46], or (2) through the recombination of h⁺ by e⁻ provided from CO₃²⁻ [47]. In the latter case, the recombination of the generated charge carrier might be reduced, and more electrons can react with the dissolved oxygen to form the superoxide radical. Regardless of either of the previously presented mechanisms, the direct involvement of carbonate anions is obvious. To elucidate the formation of carbonate anion radical, first the inorganic carbon content of the final reaction mixture (IC_final) was evaluated (Figure S2, Supplementary Material). Although considerably lower IC_final was assessed using cSTO (IC_final-cSTO ~ 700 μg∙L⁻¹) in comparison with the STO_x_SCO (IC_{final-STO_x_SCO} > 3 mg∙L⁻¹), this value is relatively high compared with the IC_final using other catalysts (e.g., P25; IC_{final, P25} ~ 120 μg∙L⁻¹). The relatively high IC_final using SrTiO₃-SrCO₃ catalysts clearly indicates the dissolution of carbonate into the reaction mixture, given by its moderate solubility in aqueous medium (0.01 g SrCO₃/L H₂O) [48], thus the carbonate anions mainly resulted from the catalyst. Moreover, the dependence of IC_final (and implicitly the concentration of carbonate anions in the reaction mixture) on the SrCO₃ content of the studied samples can be highlighted (Figure S2, Supplementary Material). Second, the Rietveld analysis performed based on the XRD pattern of the spent catalyst (STO_15_SCO, Figure S3, Supplementary Material) also confirmed carbonate dissolution via the decreased SrCO₃ content of the catalyst after reaction. In addition to this, the relatively high IC_final-cSTO (compared with IC_final-P25) can be explained considering the dissolution of airborne CO₂ into the reaction mixture, followed by the formation of carbonate layer on the surface of SrTiO₃ [49] (PZC(SrTiO₃): ~8.5–9.5 [50]), and its further dissolution into the reaction mixture.

Finally, the reusability of the most efficient catalyst (STO_21_SCO) was tested in three cycles. The transformation of DCFNa was complete after 1 h during the three cycles (Figure S4, Supplementary Material); however, the mineralization of DCFNa did not take place starting from the second cycle. The XRD patterns of the best performing catalyst before the photocatalytic test and after the third cycle are depicted in Figure 9. The deactivation of the catalyst after 4 h might be explained by the adsorption of certain intermediates/products or by the loss of the carbonate phase. This assumption was supported by the observed mass loss of the spent catalyst in the TGA experiments which was obvious higher than that of the fresh catalyst (Figure S5, Supplementary Material).

2.3.2. Photocatalytic CO₂ Reduction Experiments

As the efficiency of SrTiO₃ towards both oxidation and reduction is reported in the literature, our curiosity has led us to assess the efficiency of SrTiO₃-SrCO₃ in a gas-solid photocatalytic CO₂ reduction experiment. Furthermore, a relatively limited number of studies are available aiming for the assessment of SrTiO₃-SrCO₃ photocatalytic activity in CO₂ reduction experiments. Although promising selectivity was reported by Li and coworkers towards photocatalytic CO₂-to-CO activity using SrTiO₃-SrCO₃ [51], Gyulavari and coworkers concluded that the presence of SrCO₃ had a negligible influence on the photocatalytic CO₂ reduction activity of SrTiO₃- SrCO₃ vs. SrTiO₃ under the studied experimental conditions [52].

The STO_15_SCO and cSTO samples were evaluated within this section to get an overview on the photocatalytic CO₂ reduction activity of SrTiO₃-SrCO₃ and SrTiO₃ samples. The CO₂ photoreduction step (CRR) was preceded by batch cleaning (BC) to remove the carbon-containing impurities (e.g., precursors and solvents used for synthesis; adsorbed species, such as CO₂ from atmosphere), which can lead to the overestimation of the products resulting from CO₂ photoreduction [53]. In the presence of CO₂ after 6 h of irradiation, the formation of 4.6 ppm and 1.6 ppm of CH₄ could be observed over cSTO and STO_15_SCO, as well as traces of C₂H₆ (Figure 10). In the literature, besides CH₄, the formation of CO is reported during photocatalytic CO₂ reduction experiments over SrTiO₃ under different experimental conditions [9,52]. Since the detection limit of the GC employed for the identification of the reaction products is for CO at approximately 20 ppm, the formation of CO cannot be excluded under the selected reaction conditions. Other differences compared with published works may result from the difference in reaction conditions (here gas-solid, no liquid phase).

In addition to this, the unfavorable effect of the carbonate of studied SrTiO₃-based samples (i.e., STO_15_SCO) on CH₄ production can be highlighted (Figure 10). Similar findings were reported by Pougin and coworkers using titania with enhanced surface carbonate vs. “carbonate-free” titania during photocatalytic CO₂ reduction experiments [54]. The role of carbonates as hole traps on the surface of catalysts has been reported in several works, which is not favorable in the case of (photocatalytic) CO₂ reduction reactions [55,56,57]. Furthermore, the considerable stability of carbonates should also be considered, which inhibits their involvement in further processes [57].

Although low concentration of products was detected under the selected experimental conditions (CH₄: cSTO, 12.6 ppm∙g_cat⁻¹∙h⁻¹ in comparison with STO_15_SCO, 5.4 ppm∙g_cat⁻¹∙h⁻¹), the role of SrCO₃ as a carbon-containing source can be excluded as it does not seem to contribute to the formation products in the presence or absence of CO₂. As CO₂ photoreduction is a multiparametric reaction, further studies are required under different experimental conditions, e.g., different H₂O and/or CO₂ concentrations, light intensity, etc. For a definite proof of the origin of C-containing products, ¹³CO₂ reduction can be employed for the best performing photocatalysts [58].

3. Experimental Materials and Methods

3.1. Chemicals

Strontium nitrate (Sr(NO₃)₂, ≥99%, Sigma Aldrich, Germany), titanium (IV) oxide—anatase (TiO₂, 99.5%, IoliTec, Heilbronn, Germany), titanium (IV) oxide-P25 (TiO₂, Evonik, Germany), potassium hydroxide (KOH, ≥85%, Sigma Aldrich, Germany), sodium hydroxide (NaOH, 98%, Acros Organics, Morris Plains, NJ, USA), ethanol (C₂H₆O, >99%, Merck, Darmstadt, Germany), ammonium oxalate monohydrate ((NH₄)₂C₂O₄∙H₂O, ≥99%, Sigma Aldrich, Germany), p-benzoquinone (C₆H₄O₂, >98%, Sigma Aldrich, Germany), isopropanol (C₃H₈O, 99.9%, Merck, Germany), terephthalic acid (C₈H₆O₄, >98%, Merck, Germany), 5,5-dimethyl-1-pyrroline-N-oxide (C₃H₁₁NO, >98%, Enzo Life Sciences GmbH, Germany). Commercial strontium titanate (SrTiO₃, 99.9%, IoliTec, Germany) and strontium carbonate (SrCO₃, 99.9%, Sigma Aldrich, Hamburg, Germany) were used as reference catalysts. The chemicals for synthesis and analysis were used without any preliminary purification except for KOH.

3.2. Material Synthesis

The synthesis of SrTiO₃-based catalysts with different SrCO₃ content (5, 15, 21, 24 wt%) was carried out via hydrothermal treatment (180 °C, 12 h), followed by washing (once with ethanol, three times with distilled water) and drying (80 °C, 12 h). Only the preparation of the reaction mixture is described in detail, because only this step was carried out in a different way in the case of samples with different STO contents.

The ‘‘starting’’ synthesis (STO_15_SCO) involved the addition of Sr²⁺ and Ti⁴⁺ precursors into 70 mL H₂O in stoichiometric ratio, thus obtaining an initial reaction mixture with 0.083 M Sr(NO₃)₂ and anatase (i.e., 1.270 g Sr(NO₃)₂, 0.479 g TiO₂). To facilitate the dispersion of anatase, the as-prepared reaction mixture underwent ultrasonication for 3 min. The last steps involved the addition of KOH (thus achieving 8 M in the reaction mixture), and homogenization (30 min).

STO_5_SCO. Although the same initial precursor concentrations were achieved as in the previous case (i.e., 0.083 M Sr(NO₃)₂ and TiO₂ (anatase), 8 M KOH), in this case the hydrothermal treatment was conducted at a smaller scale (V_autoclave = 47 mL, V_{reaction mixture} = 30 mL), due to the complexity of this approach. To assess the influence of scaling-up, the previously presented ‘’starting’’ synthesis (STO_15_SCO) was also carried out at this scale. The effect of scaling-up on the morphological, structural, and photocatalytic properties was negligible. The requirements of Schlenk line technique were considered during the synthesis of STO_5_SCO, as presented (i–iv):

(i) 33.663 g KOH was added gradually and dissolved in 50 g degassed water under Ar atmosphere, thus obtaining a 14 M KOH solution. The removal of K₂CO₃ from the concentrated KOH solution was performed considering the method reported by P. Sipos and coworkers [59]. More precisely, 50 mg CaO was added to the concentrated KOH solution and stirred overnight under inert conditions. The separation of the clear KOH solution (C_KOH~14 M, solution A) from the formed CaCO₃ was realized by applying the cannula method.

(ii) 0.653 g Sr(NO₃)₂ was dissolved in 12 g H₂O in a Schlenk flask (2.57 × 10⁻⁴ M) under Ar atmosphere (solution B).

(iii) The autoclave (V_autoclave = 47 mL) was purged with Argon, followed by the addition of 0.205 g anatase, followed by its immediate coverage.

(iv) 10.544 g solution B was withdrawn with a syringe and transferred to the autoclave

(v) 33.465 g solution A was transferred to the autoclave in a similar manner to solution A. The reaction mixture was stirred for 30 min, followed by hydrothermal crystallization.

The synthesis of STO_21_SCO was performed as the “starting” synthesis (i.e., STO_15_SCO) with the exception that in this case the initial Sr²⁺:Ti⁴⁺ molar ratio was 1.25:1.

The difference in the case of STO_24_SCO (vs. “starting“ synthesis) was related to the usage of P25 as a Ti⁴⁺ precursor (instead of pure anatase).

3.3. Characterization

The structural and morphological features of the catalyst were assessed via scanning electron microscopy (SEM). The SEM micrographs were recorded using a Merlin VP compact device (Zeiss, Oberkochen, Germany).

The X-ray diffraction (XRD) patterns were recorded on an Xpert Pro diffractometer (Panalytical, Almelo, the Netherlands) using CuK_α1K_α2 radiation (λ₁ = 0.15406 nm, λ₁ = 0.15443 nm). The primary crystallite size (τ) was calculated via Scherrer equation considering the as-called shape factor (K = 0.90 used for spherical crystallites), X-ray wavelength (λ), full width at half maximum (β), and peak position (θ) [60]:

$$\tau =\frac{K\cdot \lambda }{\beta \cdot \cos \theta }$$

(1)

The phase composition was determined via Rietvield analysis using HighScore Plus.

The UV-Vis spectra of the solid catalysts were recorded via a Lambda 650 spectrophotometer (Perkin Elmer, Waltham, MA, USA). The band gap values were determined based on the Tauc-plot.

NOVAtouch (Quantachrome Instruments, Boynton Beach, FL, USA) was used for the collection of the nitrogen sorption data at 77 K. The (BET) multipoint method was considered for the determination of specific surface area (p/p₀: 0.05–0.3). The pre-treatment of the samples was performed via heating at 350 °C under vacuum for 5 h.

Thermogravimetric analysis (TGA) of the catalyst was performed via a Netzsch STA 449 F3 Jupiter device (Selb, Germany) in the temperature range of 25–600 °C in air atmosphere.

The photodegradation experiments took place in a double jacket glass cylindrical batch photoreactor (V = ~120 mL, Hassa labor, Lübeck, Germany). The suspension was irradiated with a Xe arc lamp (300 W) equipped with a reflector system (LOT Quantum Design, Darmstadt, Germany). The irradiation of the reaction mixture was carried out from the top, where the distance between the reactor and reflector system was 6 cm. The light intensity inside of the reactor (at 8 cm distance from the reflector system) was 625 mW∙cm⁻². The photocatalytic experiments were conducted at a constant temperature (25 °C). The mass-transfer was facilitated via assuring continuous synthetic air flow (15 mL∙min⁻¹) and stirring (500 rpm). 40 mg SrTiO₃-based catalyst was suspended in 40 mL aqueous solution of diclofenac (C₀ = 25 mg∙L⁻¹), followed by stirring for 30 min under air flow prior irradiation. The DCFNa transformation was followed by high-performance liquid chromatography (HPLC, Agilent Technologies, 1260 Infinity Series, Santa Clara, CA, USA). The mineralization efficiency was assessed after 4 h by evaluating the total organic carbon content (TOC) of the filtered (20 μm) reaction mixture. The dissolved carbonate content from the SrTiO₃-based catalyst was quantified via the inorganic carbon content of the reaction mixture (IC) using a TOC analyzer (multi N/C, 3100, Analytik Jena, Jena, Germany).

The same experimental conditions were applied during the scavenging experiments as during the degradation experiments, using ammonium oxalate (AO), p-benzoquinone (PBQ), and isopropanol (IPA). The initial concentration of scavengers in the reaction mixture was 393 μM (AO, PBQ) and 1.57 mM (IPA).

The trapping of the hydroxyl radicals in the final reaction mixture was carried out according to the method described by Marschall and coworkers [61]. In this step, 40 mg catalyst was dispersed in 40 mL NaOH aqueous solution (2 × 10⁻³ M) with 5 × 10⁻⁴ M terephthalic acid. Although similar conditions were maintained as described previously at the photodegradation section, the duration of this radical-trapping experiment was 2 h. 4 mL samples were withdrawn (at 30, 60, 90, 120 min) and filtered. The formation of hydroxylated terephthalic acid compound (TA-OH) was assessed by setting 318 nm as the excitation wavelength and recording the photoluminescence spectra in the range of 350–600 nm via a Varian Cary Eclipse Fluorescence Spectrometer (Agilent Technologies, Mulgrave, Australia).

The identification of the active species in the reaction mixture was performed via in-situ electron paramagnetic resonance (in-situ EPR) measurements using 5,5-dimethyl−1-pyrroline-N-oxide (DMPO) as a trapping agent. X-band-EPR spectra were recorded via a Bruker EMX CW-micro EPR spectrometer (Rheinstetten, Germany). The reaction mixture was irradiated for 10 min prior to the addition of the trapping agent (10 μL DMPO). The g value was calculated according to Equation (2), considering the frequency (υ), the resonance filed (B₀), Bohr magneton (β = 9.27 × 10⁻²⁷ J∙mT⁻¹) and Planck constant (h = 6.626 × 10⁻³⁴ J∙s):

$$g=\frac{h·v}{\beta ·B_0}$$

(2)

The photocatalytic CO₂ reduction activity of the selected SrTiO₃-based samples was assessed in a high-purity gas phase photoreactor system described in detail by Mei et al. [62]. The main parts of this system include the reaction chamber, a Hg/Xe lamp (200 W, light intensity: 200 mW∙cm⁻², Newport-Oriel, Stratford, CT, USA), mass flow controllers, water saturators to humidify the inlet gases, a gas chromatograph (Tracera-2010, Shimadzu; Berlin, Germany), and a vacuum pump. The high purity flow of He (99.9999%) and 1.5 CO₂ (99.9995%)/He (99.9999%) were controlled by the mass flow controllers. The experiments were performed in batch mode, consisting of a batch cleaning (0.6 vol% H₂O/He) and CO₂ reduction experiments (1.5 vol% CO₂/He). According to the Antoine equation, the adjustment of initial H₂O vapor concentration in the reactor was realized by controlling the initial pressure (1500 mbar) and the temperature in the water saturators (278 K). The batch-mode experiments were preceded by flow measurements (in other words, the system was purged either by humified He or by CO₂/He) until the O₂ was removed from the system. The next step was related to the filling of the reactor with the appropriate gas mixture, reaching 1500 mbar as initial pressure. Gaseous samples were withdrawn in 45 min, which induced a pressure drop after each sample withdrawal. The detection and analysis of the products (CH₄, C₂H₆, CO, H₂, and other products) was carried out by the Shimadzu Tracera GC equipped with a barrier ionization discharge (BID) and flame ionization (FID) detectors.

4. Conclusions

This study shows different methods which can be applied to adjust the SrCO₃ content of hydrothermally synthesized SrTiO₃. Similar primary crystallite size (of STO: 18–28 nm), band gap energy (3.16–3.23 eV) and specific surface area (15–45 m²∙g⁻¹) values were determined in the case of all studied SrTiO₃-based samples with 5, 15, 21, or 24 wt% SrCO₃ content.

Furthermore, the photocatalytic performance of SrTiO₃-SrCO₃ catalysts was assessed in both oxidation (diclofenac mineralization) and reduction processes (CO₂ reduction).

Improved diclofenac mineralization (after 4 h) was determined in the cases of all SrTiO₃-SrCO₃ catalysts (vs. SrTiO₃) with an optimal SrCO₃ content of 21 wt% (X_{mineralization} = 62%). The enhanced mineralization of diclofenac using SrTiO₃-SrCO₃ (vs. SrTiO₃) may be explained by (1) the involvement of CO₃^•– (in-situ EPR), (2) better charge separation (and implicitly higher concentration of formed ^•OH), and (3) favorable morpho-structural and textural properties (i.e., higher surface area). Finally, although higher CO₂ reduction activity was assessed for SrTiO₃ in comparison with SrTiO₃-SrCO₃, further investigations are required to clarify the involvement of SrCO₃.