1 Introduction

It is extremely important for new devices to evaluate how the defects present in a crystalline lattice of semiconductor oxides influence the most diverse properties (electrical, optical, sensor, magnetic). Studies have shown that the composition of perovskite doping applied to the solar cell improves its performance due to the trapping of defects on the surface of the materials [1, 2]. In the same sense, materials applied to the anti-reflective coating have shown their performance is related not only to the composition but also to the type of defect generated in the structure [3,4,5]. Among the numerous applications, the photocatalytic also has features closely related to the presence of defects. In the fight against increasing water pollution, environmental remediation by photocatalysis has been a focus of the scientific community [6,7,8]. Highly photoactive materials that act in water decontamination have structural defects such as oxygen vacancies [9,10,11,12,13]. One way to generate structural defects is through the formation of heterojunctions [14, 15] whose interfaces act as regions of charge carriers that alter the electrical, optical and magnetic properties of the material according to the band structures of the present phases [16]. Composition and modification are another strategy to the design of the photocatalysts with higher reactivity of active sites and the desired surface, and thus reducing the charge carriers recombination and enhancing the photocatalytic activity [17, 18]. The strategies of obtaining heterojunctions for the effective separation, transfer and suppression of the charges combination are in several applications to obtain clean and sustainable energy sources. Oxygen and hydrogen evolution reaction for the purpose of water splitting are reactions totally dependent on the electronic structure and the behavior of the charge carriers present in the materials interface region [19,20,21]. Such properties that depend on the semiconductors photoelectrochemical behavior show improvements due to synergistic effects of the interface and oxygen vacancies obtained in these materials [21].

For large-scale application, it is necessary that the photocatalyst be abundant, easily obtainable and inexpensive; one such material that has been studied for such characteristics is CaO [11, 22,23,24,25]. Such compounds have been doped with metal ions based on their heterostructures [11, 26,27,28,29,30]. During CaO synthesis, other phases, such as hydroxide (Ca(OH)2) and carbonate (CaCO3), which also exhibit significant photocatalytic activity, [11, 22,23,24] can be detected. Pure, doped or heterostructured CaO has high photocatalytic versatility and has been studied in the application of dye discoloration [11, 22,23,24], biodiesel production via palm oil transesterification [29] biodiesel production via free fatty acid esterification [26, 28], synthesis of organic compounds [28], antibiotic degradation [27], ammonia degradation [31] and palm oil degradation [30].

CaTiO3 is a material that can be used to form heterostructures with CaO, an n-type semiconductor [32] with an ABO3 perovskite structure, with Ti–O terminations and a band gap of 3.5 eV [32, 33]. CaTiO3 has also been prepared for photocatalytic purposes with the addition of lattice and surface modifiers, e.g., Au decoration. In this compound, Rhodamine B (RhB) was 99.6% degraded under simulated solar irradiation during 120 min of photocatalytic reaction due to efficient separation of the photoexcited electron/hole (e−/h+) pairs, owing to electron transfer from the CaTiO3 to the Au [34].

Compared to CaTiO3/Eu3+ without Ni0, Eu3+/Ni0-doped CaTiO3 showed a higher photocatalytic activity for toluene degradation attributed to the lower photocatalyst band gap energy value [35], that is, a direct relationship with the defects present in the material structure. The CaTiO3/TiO2 heterojunction was used for photocatalytic CO2 reduction, exhibiting five times more efficiency compared to undoped TiO2. This result was attributed to crystal band structures, efficient separation of e−/h+ pairs, and the surface basicity of the CaTiO3/TiO2 composites that facilitated CO2 adsorption and activation [36]. CdSe-modified CaTiO3 nanocomposites and g-C3N4 have been reported as photocatalysts in hydrogen production [37, 38]. As an extension of our previous work [9,10,11], in this study, we have developed calcium heterojunctions demonstrating the importance of the defects generated in the formation of heterojunctions and the activity of photogenerated species, studied using scavengers.

2 Experimental

2.1 Heterojunction production by sol–gel method

For the production of Ca-based heterojunctions, titanium isopropoxide (Sigma-Aldrich, 97%) and acetic acid (Qhemis, 99.7%) were used in a 1:4 molar ratio. The acid was added to the titanium isopropoxide at a temperature of 100 °C and remained under constant stirring for 3 h. The titanium acetate formed was standardized, and the number of moles of Ti4+/g of titanium acetate was calculated. Concomitantly, a suspension of CaO (Êxodo, 95%) and isopropanol (Vetec, 99.5%) was obtained using a 1:10 molar ratio. This suspension was added to the titanium acetate solution. The concentration of titanium acetate was calculated to generate heterojunctions with 0.25%, 0.50%, 2.0%, 10.0% and 50.0% by weight of the CaTiO3 (CTO) in the CaO matrix, resulting in samples Ca + 0.25%CTO, Ca + 0.5%CTO, Ca + 2%CTO, Ca + 10%CTO and Ca + 50%CTO, respectively. The solid formed was dried at 150 °C for 12 h in a muffle furnace under air flow. Thereafter, the samples were heat treated at 350 °C for 4 h to remove organic matter, followed by treatment at 500 °C for 4 h. All heat treatments were done with air flow. The CaO used was a commercial reagent, Exôdo®, 95%.

2.2 Heterojunctions characterization

The materials were investigated by Thermogravimetry and Differential Thermal Analysis (TG/DTA) in NETZSCH® equipment. The samples were heated to 1000 °C at a rate of 10 °C/min using a nitrogen atmosphere. X-ray Diffraction (XRD) analyses were obtained using a Rigaku® diffractometer, model RINT2000. Measurements were made with a monochromatic X-ray beam by copper Kα radiation (50 kV, 120 mA) with emission wavelengths of 1.54 Å in increments of Δ2θ = 0.02°. UV/Vis/NIR Diffuse Reflectance Spectroscopy (UV/Vis/NIR DRS) characterization was performed using the Perkin Elmer® Lambda 1050 model UV–Vis-NIR absorption spectrophotometer with an integrating sphere 150 mm in diameter with InGaAs (NIR) and PMT (UV/Vis) detectors. Analyses were performed in the 200–800 nm spectral range in diffuse reflectance mode. The spectra were used to calculate the band gap energy of the materials using the Kubelka–Munk algorithm.

Photoluminescence (PL) characterization of samples was performed using a 325 nm laser, and the spectra were recorded in the range of 350 nm to 850 nm. Field Emission Scanning Electron Microscopy (FE-SEM) and Energy Dispersion Spectroscopy (EDS) using FEG-MEV JEOL® Model 7500F equipment with secondary electron detectors was used to illustrate the morphology and indicate the qualitative chemical composition. High Resolution Transmission Electron Microscopy (HRTEM) with a FEI TECNAI G2 F20 HRTEM® transmission electron microscope was used to identify the structures of the crystalline phases that formed the heterojunction as well as to illustrate the interface regions and short range defects. The Ca + 0.25%CTO (w/w) and Ca + 10%CTO (w/w) samples were characterized by X-ray Excited Photoelectron Spectroscopy (XPS) using a krypton laser (Coherent Innova 200) as the excitation source with a wavelength of 350 nm on Scientia Omicron® equipment with AlKα X-ray source.

2.3 Photoactivity of catalysts for discoloration of RhB

To evaluate the photocatalytic activity of the samples, 700 mL of 10–5 mol/L RhB dye aqueous solution was sonicated for 20 min in a dark room. For the light source, the photocatalytic process included a UV lamp (λmax = 254 nm, 11 W, Osram, Puritec HNS 2G7) placed inside a quartz tube inside the reactor. Constant suspension agitation and air bubbling were used. Aliquots were collected from the photocatalytic reaction at several points in the reaction, centrifuged and analyzed on a Perkin Elmer® Lambda 1050 spectrophotometer. Photocatalytic efficiency was measured by the percentage of discoloration of Rhodamine B solution, according to Eq. (1) below, and RhB photodiscoloration half-life (t1/2) was calculated using Eq. (2) below:

$$\% {\text{D}} = \frac{{{\text{C}}_{0} - {\text{C}}_{{\text{t}}} }}{{{\text{C}}_{0} }} \cdot 100$$
(1)
$${\text{t}}_{1/2} = \frac{\ln 2}{{\text{k}}}$$
(2)

where: % D = percentage of discoloration, C0 = Dye concentration at time zero, Ct = Dye concentration at time t, t1/2 = half-life, k = reaction rate.

To study the reusability and stability of the photocatalysts, the photocatalysts was separated from the suspension by centrifugation, calcined at 500 °C, analyzed by XRD, weighed and reused in the photocatalysis. Three cycles of photocatalysis were performed using Ca + 0.25%CTO as a photocatalyst.

To study the mechanism of action of the photocatalysts, CaO, CaTiO3, and Ca + 0.25%CTO were used under the same conditions described above, but scavengers was added before the ultrasound. It was used as an electronic hole scavenger (h+) 1 mmol/ disodium EDTA, as a superoxide radical scavenger (O2·−) 1 mmol/L p-benzoquinone, and as a hydroxyl radical scavenger (OH·) 200 mmol/L of isopropanol.

3 Results and discussion

3.1 Structural, morphological and electronic characterization

X-ray diffraction pattern profiles of the heterojunctions are shown in Fig. 1a. Peaks were attributed to CaO, CaCO3, Ca(OH)2, and CaTiO3 phases, as identified by similarity to JCPDS standards no 60199, 20179, 73467, and 74213. The addition of up to 2% of Ti ions to the CaO did not allow for identification of the perovskite phase of CaTiO3. In turn, the XRD of the Ca + 10%CaTiO3, Fig. 1b, indicated the presence of a two theta peaks at 32.7 relative to the CaTiO3 plane (112).

Fig. 1
figure 1

X-ray diffraction of CaO/CaTiO3/CaCO3/Ca(OH)2 heterojunctions

Full size image

The Ca + 50%CTO showed peaks at 32.8, 40.5, 47.3, 58.7, 69.0, and 78.6 degrees, related to the CaTiO3 phase. The peak intensities of samples containing 10% and 50% of CaTiO3 were different, demonstrating that the concentration of the perovskite phase changed for each heterojunction. Results from TG/DTA, Fig. S-1, and Table S-1 demonstrated the constitution of the materials, in which the CaTiO3 phase formation increased proportionally to the Ti ions added in the synthesis.

Heterojunction formation was verified by HRTEM analysis with the Ca + 10%CTO sample, as seen in Fig. 2. The sample had distances of 2.78 Å of the (111) CaO plane, 3.43 Å of the (111) CaTiO3 plane, 3.86 Å of the (012) CaCO3 plane, and 3.11 Å of the (111) Ca(OH)2 plane. This result showed the intimate contact interface formation of a quaternary heterojunction [39], wherein the structures CaO cubic, CaTiO3 orthorhombic, and CaCO3, Ca(OH)2 trigonal were the constituents. The formation of this heterojunction was thermodynamically favorable since the lattice mismatches were similar. For instance, consider the CaO-cubic/CaCO3-trigonal system, in which the lattice mismatch was 4% (aCaO = 4.812 Ǻ and aCaCO3 = 4.994 Ǻ) or the CaCO3-trigonal/Ca(OH)2-trigonal system, with a mismatch of 2% (aCaCO3 = 4.994 Ǻ and cCa(OH)2 = 4.911 Ǻ). FE-SEM analysis, Fig. 3, shows that the CaO had a micrometric size particle and CaO/CaTiO3/CaCO3/Ca(OH)2 heterojunctions had nanometric, spherical, and agglomerate grains. It was found that the increase in CaTiO3 concentration resulted in a decrease in average particle size, Table 1. EDS spectra of the Ca + 10%CTO sample, Fig. 3e, revealed the distribution composition of the elements presented in the heterojunction. These results showed peaks related to the O, Ca, and Ti at 0.52 keV, 3.69 keV, and 4.51 keV at Kα, respectively, which represented sample compositions, and at Si, the Kα peak at 1.74 keV represented the sample holder for analysis. The wide region, region 1, illustrated high intensity peaks related to Ca and O, which referred to the sample matrix, and a low intensity of Ti peak, which referred to the heterojunction. Analyzing specific regions of the sample, points 2 and 3 showed grains formed exclusively by Ca, without Ti composition, and point 2 showed grains rich in Ti, with a Ti peak of intensity higher than those measured in a large region of the sample, point 3. It was concluded that CaTiO3 inserted in the sample was not a solid solution, which was in agreement with the formation of the heterojunction.

Fig. 2
figure 2

HRTEM of Ca + 10%CTO heterojunction interface with expansion of the central region

Full size image
Fig. 3
figure 3

FE-SEM micrographs of the samples: a CaO; b Ca + 0.25%CTO; c Ca + 10%CTO; d Ca + 50%CTO; and e FE-SEM–EDS analysis of Ca + 10%CTO heterojunction

Full size image
Table 1 Average particle size of samples
Full size table

Electronically, the samples were characterized by diffuse reflectance and photoluminescent spectroscopy, as illustrated in Figs. 4 and 5. Results of diffuse reflectance spectra (see Fig. 4 and Table 2) to the CaO, CaTiO3, and CaO/CaTiO3/CaCO3/Ca(OH)2 samples showed a direct band gap energy by the Kubelka–Munk algorithm [32, 33, 40,41,42]. Heterojunctions showed lower band gap energy with increasing CaTiO3 concentration. This was due to the increasing perovskite concentration and decreased dielectric Ca(OH)2 concentration [11], which promoted decreased band gap energy.

Fig. 4
figure 4

Direct band gap energy graphs by Kubelka–Munk algorithm obtained by diffuse reflectance of the CaO, CaTiO3, and CaO/CaTiO3/CaCO3/Ca(OH)2 samples

Full size image
Fig. 5
figure 5

Photoluminescence spectra obtained from CaO and CaTiO3 samples and CaO/CaTiO3/CaCO3/Ca(OH)2 heterojunctions

Full size image
Table 2 Direct band gap energy of the studied materials
Full size table

Figure 5 shows the PL emission spectra of CaO, CaTiO3, and the heterojunctions. It can be observed from the CaO emission spectrum that the band-to-band recombination, formed by an electron photoexcited in the conduction band (CB) and an electronic hole in the valence band (VB)–recombination, occurred at different energy levels in the band gap region. In this sample, the highest emission intensity occurred in three main regions (410 nm, 570 nm, and 780 nm). In the heterojunction sample, the PL intensity decreased, which induced suppression of the band-to-band recombination. Another characteristic of CaO/CaTiO3/CaCO3/Ca(OH)2 heterojunctions was the narrower spectra compared to CaO, with maximum intensity at approximately 550 nm. CaTiO3, as well as the heterojunctions containing a higher CaTiO3 concentration, had a low intensity of band-to-band recombination, a characteristic also found in the SrTiO3 [11]. Deconvolution of PL spectra, Fig. 6, was performed for a more detailed analysis of the emission regions, and, in follow-up, the types of defects present in the pure and heterojunctions samples are discussed.

Fig. 6
figure 6

Deconvolution of CaO, CaTiO3 photoluminescent spectra and CaO/CaTiO3/CaCO3/Ca(OH)2 heterojunctions

Full size image

CaO showed five emission bands, due to different types of intrinsic structural defects of the material, with energy ranging from 3.02 eV (violet region, 411 nm) up to 1.62 eV (infrared region, 764 nm). The violet region band indicated direct recombination of an e−/h+ pair between the CB and VB. In addition, the emission in this region is a characteristic of material defects, mainly characteristic of the presence of oxygen vacancies (\(V_{O}^{x}\), \(V_{O}^{ \cdot }\), \(V_{O}^{ \cdot \cdot }\)). In which the superscripts ‘x’, ‘·’, and ‘··’, indicate types of neutral, mono, and double ionized vacancies, respectively [9,10,11, 43, 44]. The band in the blue region indicated a \(V_{O}^{ \cdot }\) type and bands in the green and red region referred to \(V_{O}^{ \cdot \cdot }\) type vacancies [43, 45]. The infrared region band was caused by distortion of the crystal lattice by the formation of the heterojunction [9,10,11, 43, 44, 46, 47]. The photoluminescent deconvolution spectrum of CaTiO3 showed green/red and infrared bands, which represented double-ionized oxygen vacancies and crystal lattice distortion, respectively. The CaO/CaTiO3/CaCO3/Ca(OH)2 heterojunctions spectra showed two bands, one in the green and one in the red region. Heterojunctions with spectra in the green/red region are indicative of the presence of double-ionized oxygen vacancies that acted as electron receptors [9,10,11, 43, 44]. The increase in CaTiO3 concentration in the CaO matrix changed the main emission region, resulting in an increase in the red region emission band.

The chemical environments of the Ca + 0.25%CTO and Ca + 10%CTO heterojunctions were investigated by XPS Fig. 7. Survey spectra of Ca + 0.25%CTO and Ca + 10%CTO heterojunctions are shown in Fig. 7a, f. Ca and Ti elements could be detected in the spectra at peaks 2p (~347 eV and ~457 eV). O 1 s is present in the region of 531 eV, and the C 1 s appears at approximately 285 eV. According to the results, the Ca 2p peak indicated the +2 oxidation state and the Ti 2p peak indicated the +4 oxidation state, as seen in Fig. 7b, c, g, h, respectively. This indicated that obtaining heterojunction with low or high CaTiO3 concentration did not alter the oxidation state of the metal ions present in the crystal lattice. The presence of carbonate on the surface was confirmed by C 1 s peaks at 289.3 eV [48, 49], (Fig. 7d, i). The O 1 s peak was deconvoluted into two peaks (Fig. 7e, j): one with a binding energy of approximately 531.1 eV, reflecting the metal–oxygen bonding of the crystal lattice (Ca-O and Ti–O) [28, 50] and the presence of hydroxyls (Ca(OH)2 and -OH terminations) [51], and a peak at approximately 532.5 eV, reflecting the carbon–oxygen bonding (–CO3) [52, 53] and structural defects as oxygen vacancy [52, 54]. Thus, the peak at approximately 532.5 eV confirms the presence of oxygen vacancies, also verified by the PL analysis. O 1 s deconvolution to the Ca + 0.25%CTO heterojunction showed the peak relative to CaCO3 and oxygen vacancies (532.7 eV) (Fig. 7e), with 22.9% of the peak area, while in the Ca + 10%CTO heterojunction, this value decreased to 15.9%. The Ca + 10%CTO sample would be expected to have the highest peak area (~532.6 eV) as it has the highest proportion of carbonate (Fig. S1—42.5%). However, this result demonstrated that the highest peak area (~532.6 eV) of the Ca + 0.25%CTO sample reflected the oxygen vacancies found on the material surface. As the Ca + 0.25%CTO heterojunction presented a higher PL emission in the green region (Fig. 6), and this emission was mainly related to defects of the oxygen vacancy type (Fig. 7e), one could deduce the types of defects present in the crystalline structure clusters, Eqs. 3–8:

$$\left[ {{\text{CaO}}_{6} } \right]_{{\text{o}}}^{{\text{x}}} + \left[ {{\text{CaO}}_{5} \,{\text{V}}_{{\text{o}}}^{{\text{x}}} } \right]_{{\text{d}}} \to \left[ {{\text{CaO}}_{6} } \right]_{{\text{o}}}^{^{\prime}} + \left[ {{\text{CaO}}_{5} \,{\text{V}}_{{\text{o}}}^{ \cdot } } \right]_{{\text{d}}}$$
(3)
$$\left[ {{\text{CaO}}_{6} } \right]_{{\text{o}}}^{{\text{x}}} + \left[ {{\text{CaO}}_{5} \,{\text{V}}_{{\text{o}}}^{ \cdot } } \right]_{{\text{d}}} \to \left[ {{\text{CaO}}_{6} } \right]_{{\text{o}}}^{^{\prime}} + \left[ {{\text{CaO}}_{5} \,{\text{V}}_{{\text{o}}}^{ \cdot \cdot } } \right]_{{\text{d}}}$$
(4)
$$\left[ {{\text{TiO}}_{6} } \right]_{{\text{o}}}^{{\text{x}}} + \left[ {{\text{TiO}}_{5} \,{\text{V}}_{{\text{o}}}^{{\text{x}}} } \right]_{{\text{d}}} \to \left[ {{\text{TiO}}_{6} } \right]_{{\text{o}}}^{^{\prime}} + \left[ {{\text{TiO}}_{5} \,{\text{V}}_{{\text{o}}}^{ \cdot } } \right]_{{\text{d}}}$$
(5)
$$\left[ {{\text{TiO}}_{6} } \right]_{{\text{o}}}^{{\text{x}}} + \left[ {{\text{TiO}}_{5} \,{\text{V}}_{{\text{o}}}^{ \cdot } } \right]_{{\text{d}}} \to \left[ {{\text{TiO}}_{6} } \right]_{{\text{o}}}^{^{\prime}} + \left[ {{\text{TiO}}_{5} \,{\text{V}}_{{\text{o}}}^{ \cdot \cdot } } \right]_{{\text{d}}}$$
(6)
$$\left[ {{\text{CaO}}_{8} } \right]_{{\text{o}}}^{{\text{x}}} + \left[ {{\text{CaO}}_{7} \,{\text{V}}_{{\text{o}}}^{{\text{x}}} } \right]_{{\text{d}}} \to \left[ {{\text{CaO}}_{8} } \right]_{{\text{o}}}^{^{\prime}} + \left[ {{\text{CaO}}_{7} \,{\text{V}}_{{\text{o}}}^{ \cdot } } \right]_{{\text{d}}}$$
(7)
$$\left[ {{\text{CaO}}_{8} } \right]_{{\text{o}}}^{{\text{x}}} + \left[ {{\text{CaO}}_{7} \,{\text{V}}_{{\text{o}}}^{ \cdot } } \right]_{{\text{d}}} \to \left[ {{\text{CaO}}_{8} } \right]_{{\text{o}}}^{^{\prime}} + \left[ {{\text{CaO}}_{7} \,{\text{V}}_{{\text{o}}}^{ \cdot \cdot } } \right]_{{\text{d}}}$$
(8)
Fig. 7
figure 7

XPS spectra: a survey spectrum Ca + 0.25%CTO sample; b spectrum of Ca 2p species of Ca + 0.25%CTO; c high resolution spectrum of Ti 2p species of Ca + 0.25%CTO; d spectrum of C 1 s species of Ca + 0.25%CTO; e spectrum of the O 1 s species of Ca + 0.25%CTO; f survey spectrum of Ca + 10%CTO sample; g Ca 2p species spectrum of Ca + 10% CTO; h spectrum of Ti 2p species of Ca + 10%CTO; i spectrum of C 1 s species of Ca + 10%CTO; j spectrum of O 1 s species of Ca + 10%CTO

Full size image

According to the Kroger Vink notations, the symbol VO designates the oxygen vacancies, and the superscript x, dot (·), and comma (,), the vacancy types: neutral, hole, and electron, respectively. Equations (3) and (4) represent the formation of the mono- and double-ionized oxygen vacancy types in the [CaO6] clusters to CaO, Ca(OH)2, and CaCO3, and Eqs. (5) to (8), to CaTiO3.

3.2 Photoactivity of catalysts for RhB discoloration

Figure 8a illustrates the discoloration of RhB as a function of time and Fig. 8b illustrates the pseudo 1st order Langmuir–Hinshelwood kinetics graph, ln (A0/At) as a function of time. From these data, we obtained the reaction rate constants (k) and calculated the half-life times (t1/2) of the RhB photodiscoloration reaction, shown in Table 3.

Fig. 8
figure 8

Photocatalytic activity of CaO, CaTiO3, and CaO/CaTiO3/CaCO3/Ca(OH)2 heterojunction

Full size image
Table 3 Photocatalysis data of CaO/CaTiO3/CaCO3/Ca(OH)2 heterojunctions and their precursors
Full size table

The best photocatalytic activity was observed in RhB dye discoloration obtained with the Ca + 0.25%CTO sample, which had a calculated half-life of 38.5 min. The photocatalytic performance of the Ca + 0.25%CTO heterojunction investigated in this work was compared with the previous reports, as shown in the Table 4. It can be observed from these results that the heterojunction material obtained in this work has greater discoloration kinetics than its respective pure materials. The use of an electromagnetic radiation source with energy greater than or equal to band gap energy generated free electrons in the CB and electronic holes in the VB. Electrons could react with oxygen molecules to form superoxide radicals (O2·−), just as electronic holes (h+) could react with water or adsorbed hydroxyl ions to generate hydroxyl radicals (OH·). These three chemical species had photocatalytic action on RhB discoloration, according to Eqs. (9) to (16):

$${\text{Photocatalyst}} + {\text{h}}\upnu \to {\text{Photocatalyst}}\,\left( {{\text{e}}_{{{\text{CB}}}}^{ - } + {\text{h}}_{{{\text{VB}}}}^{ + } } \right)$$
(9)
$${\text{Photocatalyst}}\,\left( {{\text{e}}_{{{\text{CB}}}}^{ - } } \right) + {\text{O}}_{2} \to {\text{Photocatalyst}} + {\text{O}}_{2}^{ \cdot - }$$
(10)
$${\text{Photocatalyst}}\,\left( {{\text{h}}_{{{\text{VB}}}}^{ + } } \right) + {\text{H}}_{2} {\text{O}}_{{{\text{ads}}}} \to {\text{Photocatalyst}} + {\text{H}}^{ + } + {\text{OH}}_{{{\text{ads}}}}^{ \cdot }$$
(11)
$${\text{Photocatalyst}}\,\left( {{\text{h}}_{{{\text{VB}}}}^{ + } } \right) + {\text{OH}}_{{{\text{ads}}}}^{ - } \to {\text{Photocatalyst}} + {\text{OH}}_{{{\text{ads}}}}^{ \cdot }$$
(12)
$${\text{Photocatalyst}}\,\left( {{\text{h}}_{{{\text{VB}}}}^{ + } } \right) + {\text{RhB}} \to {\text{RhB}}^{ \cdot } + {\text{RhB}}\;{\text{oxidation}}$$
(13)
$${\text{OH}}_{{{\text{ads}}}}^{ \cdot } + {\text{RhB}} \to {\text{RhB}}\;{\text{discoloration}}$$
(14)
$${\text{O}}_{2}^{ \cdot - } + {\text{RhB}} \to {\text{RhB}}\;{\text{discoloration}}$$
(15)
$${\text{e}}_{{{\text{CB}}}}^{ - } + {\text{h}}_{{{\text{VB}}}}^{ + } \to {\text{heat}}$$
(16)
Table 4 Photocatalytic performance of precursors and heterojunctions using UV light in previous reports
Full size table

The study of reuse and stability of the Ca + 0.25% CTO photocatalyst demonstrates that the photocatalyst is stable and can be reused, Fig. 9a. After three cycles of reuse, the photocatalyst remained active, with a reduction in its photoactivity of 11.6%. The good photocatalytic performance even after the 3 cycles can be attributed to the robustness of the photocatalyst that kept its crystalline constitution practically unchanged after reuse, Fig. 9b.

Fig. 9
figure 9

Stability and Reusability of the heterojunction a Photocatalytic efficiency of the reused Ca + 0.25%CTO heterojunction for discoloration of the RhB dye for three cycles. b XRD analysis of the Ca + 0.25%CTO sample used in reuse

Full size image

Scavenger tests were performed in order to compare the species with photocatalytic action in the most photoactive heterojunction. The photocatalytic activity of CaO, CaTiO3, and Ca + 0.25%CTO samples when EDTA-Na2 was used as an electronic hole (h+) scavenger, isopropanol as a hydroxyl (OH·) radical scavenger and p-benzoquinone as a superoxide (O2·−) radical scavenger, are shown in Fig. 10. The percentage indicated in the figure is related to 120 min of reaction time. Similar behaviors were found in the three samples (Table 5). In the three photocatalysts studied, it can be seen that by using EDTA-Na2 and isopropanol, there was an approximate 40% reduction in dye discoloration, demonstrating that both h+ and OH· species had similar functions. Among the photogenerated species, the superoxide radical (O2·−) showed the most relevant discoloration of approximately 27% after 120 min of reaction. Our results were in agreement with the study by Liu that also found that the most photoactive species was the superoxide radical [55], unlike the findings in other materials [10, 56,57,58].

Fig. 10
figure 10

Photocatalytic activity of CaO, CaTiO3, and Ca + 0.25%CTO samples using scavengers

Full size image
Table 5 RhB discoloration after 120 min under UV light using CaO, CaTiO3 and Ca + 0.25%CTO samples with scavengers
Full size table

XRD and TG/DTA analyses indicated the presence of Ca(OH)2 and CaCO3, which were dielectric [59, 60]. Sanchez-Cantu verified that Ca(OH)2, with a 5.69 eV band gap, presented high photoactivity attributed to the indirect sensitization of Rhodamine 6G dye. In this case, there was an electron transfer from the dye to the CB of the dielectric material, which interacted with the adsorbed species on the catalytic surface, producing radical species that acted on dye degradation [22]. Recently Pei et al. [61] has also correlated the catalytic efficiency of these materials with the presence of hydroxyl radical from calcium hydroxide. From the electrons injected into the photocatalyst, superoxide radicals formed that could, in turn, form hydroxyl radicals, according to Eqs. (17) to (23):

$${\text{RhB}} + {\text{h}}\upnu \to {\text{RhB}}^{ + } + {\text{e}}^{ - }$$
(17)
$${\text{e}}^{ - } + {\text{Photocatalyst}} \to {\text{Photocatalyst}}\left( {{\text{e}}_{{{\text{CB}}}}^{ - } } \right)$$
(18)
$${\text{Photocatalyst}}\left( {2{\text{e}}_{{{\text{CB}}}}^{ - } } \right) + {\text{O}}_{2} + 2{\text{H}}^{ + } \to {\text{H}}_{2} {\text{O}}_{2} + {\text{Photocatalyst}}$$
(19)
$$2{\text{O}}_{2}^{ \cdot - } + 2{\text{H}}^{ + } \to {\text{H}}_{2} {\text{O}}_{2} + {\text{O}}_{2}$$
(20)
$${\text{H}}_{2} {\text{O}}_{2} + {\text{Photocatalyst}}\left( {{\text{e}}_{{{\text{CB}}}}^{ - } } \right) \to {\text{OH}}^{ \cdot } + {\text{OH}}^{ - } + {\text{Photocatalyst}}$$
(21)
$${\text{H}}_{2} {\text{O}}_{2} + {\text{O}}_{2}^{ \cdot - } \to {\text{OH}}^{ \cdot } + {\text{OH}}^{ - } + {\text{O}}_{2}$$
(22)
$${\text{H}}_{2} {\text{O}}_{2} + {\text{h}}\upnu \to 2{\text{OH}}^{ \cdot }$$
(23)

P-Benzoquinone captures the photogenerated free electron, stabilizing it through resonance, as seen in Eq. (24), preventing the formation of superoxide and hydroxyl radicals. This electron capture is well elucidated in the described scavengers experiments, Fig. 10.

(24)

The CB and VB energy levels presented in scheme shown in the Fig. 11 were calculated using the empirical Eqs. (25) and (26) [62]:

$${\text{E}}_{{{\text{CB}}}} = \chi - {\text{E}}_{{\text{e}}} - 0.5{\text{E}}_{{\text{g}}}$$
(25)
$${\text{E}}_{{{\text{VB}}}} = {\text{E}}_{{{\text{CB}}}} + {\text{E}}_{{\text{g}}}$$
(26)

where EVB and ECB are the potentials in VB and CB, respectively. Eg is the energy gap calculated and Ee is the free electric energy versus hydrogen (4.5 eV) [62]. Finally, χ is the semiconductor electronegativity, 3.1815 eV for CaO [63] and 5.105 eV for CaTiO3 [64]. Therefore, this is a type II heterojunction considering the alignment between the BV and CB of the materials.

Fig. 11
figure 11

Scheme showing the conduction and valence bands of the Ca2 +-based quaternary heterojunction

Full size image

The photocatalytic results indicated that formation of the heterojunction generated defects that acted as trappers of the charge carriers, increasing the recombination time of the photogenerated e−/h + pair. The TG/DTA results showed that there was no direct relationship between CaCO3 and Ca(OH)2 concentrations present in the heterojunctions and their photoactivities, indicating that there were other factors that influenced the photoactivity. The deconvolution of the PL spectra showed emission bands in the green and red region. Although both bands were generated by doubly-ionized oxygen vacancies, the increased red band resulted in a decrease in the photoactivity, shown in Table 6. Deconvolution spectrum PL to CaO resulted in five emission bands, indicating the possible paths of band-to-band recombination. Deconvolution spectrum PL to the heterojuntion showed only two emission bands, limiting the paths of recombination of the e−/h+ pair. Increasing the CaTiO3 concentration in heterojunctions decreases the recombination that occurs in the highest energy (green) region, and increases recombination in the lowest energy (red) region. Therefore, photocatalysts whose pairs are separated by longer energetic range are more photoactive, as recombination is thermodynamically more difficult, increasing recombination time.

Table 6 Photocatalysis data of CaO/CaTiO3/CaCO3/Ca(OH)2 heterojunctions and their respective spectral percentages in the green and red region generated by the deconvolution of the photoluminescence spectra
Full size table

4 Conclusion

Quaternary heterojunctions—CaO/CaTiO3/CaCO3/Ca(OH)2—were obtained. Crystalline structure, optical properties, and defects generated in the formation of the heterojunctions were correlated with the photocatalytic process. The increasing CaTiO3 content in the heterojunction reduced the band gap energy and changed the photoluminescent emission energy according to the dielectric and titanate phases present. Heterojunction with the lowest CaTiO3 concentration and better photocatalytic activity was found to present intense energy emission in the region related to defects in the double-ionized oxygen vacancy. The chemical environment was evaluated and it was proved that the observed emission in the PL spectra was related to defects in relation to the oxygen ions, and there was no alteration of the oxidation state of the metal ions. The RhB photodiscoloration study using scavengers showed that the most photoactive species among the Ca-based catalysts were superoxide radicals (O2·−). The stability and reuse of the heterojunction were evaluated, showing good performance. There was similarity in the photoactive species generated in the heterojunction and in the pure materials; however, it was demonstrated that the former presented a different energy decay region, which favored the charge separation process.