1. Introduction
Human societies are finally aware of the need to use new technologies to reduce the climate crisis and avoid environmental disasters that have been occurring quite often in recent years. Therefore, the problem of environmental remediation using efficient and environmentally friendly techniques is a hot topic today [
1]. The photocatalytic process conducted under solar irradiation is a promising candidate that satisfies the need for efficiency and safety. Its economic convenience is mainly based on the possibility of using nanomaterials not only with high performance, but also with good (photo)stability under the operating conditions [
2,
3]. Currently, only about 0.014% of solar energy (3.85∙10
24 J∙year
−1) im**ing on the Earth, and which is absolutely free, is exploited to meet the demand for renewable energy sources and for the solution of environmental problems [
4]. Therefore, the conversion of sunlight has become one of the major goals of scientists. The activation of semiconducting materials by sunlight results in the formation of electron-hole pairs, which in turn produce an electric current or trigger useful photocatalyzed chemical reactions on their surface. In this way it is also possible to carry out up-hill reactions that simulate those occurring in nature, which store solar energy in simple chemical bonds [
5,
6]. It is also possible to remove dangerous compounds through their total photo-mineralization or to synthesize organic substances with high added value in a green and safe way, starting from compounds of little economic interest [
7,
8].
Zinc oxide (ZnO) and titanium dioxide (TiO
2) in the form of nanomaterials are maybe the most investigated and applied photocatalysts traditionally proposed for wastewater remediation, due to their low cost and good optical, electronic, and structural properties [
9,
10,
11,
12,
13,
14]. However, both semiconductors scarcely absorb sunlight and have a high rate of electron/hole recombination (e
−/h
+). ZnO, exclusively, is far less stable than TiO
2 due to photocorrosion phenomena [
15,
16,
17]. To overcome these drawbacks, it has been variously modified. Usually, sensitization to sunlight and greater stability have been obtained by do** ZnO with cationic or anionic elements, combining with other semiconductors or carbon-based materials, or decorating with noble metal particles [
18,
19,
20]. Specifically, introducing alkaline earth metals into ZnO has emerged as a promising approach to develop new doped photocatalysts with distinctive optoelectronic properties [
21,
22]. Numerous studies have highlighted the potential of Mg, Ba, Ca, and Sr do** in enhancing ZnO photoactivity under both artificial UV light and visible light irradiation [
23,
24,
25,
26,
27]. However, investigations under natural sunlight irradiation have been very limited. To the best of our knowledge, no studies have specifically addressed the effects of the do** on the surface features of the doped ZnO photocatalysts, and specifically on the induced hydrophilicity. Understanding the water molecules’ affinity to the photocatalyst’s surface is vital for photodegradation reactions, as it significantly influences the generation of the reactive species such as hydroxyl radicals (OH
●) [
28].
Prior research efforts on alkaline earth metal-doped ZnO, especially with barium, have primarily focused on investigating optoelectronic properties, surface area, and morphology. In recent studies, Bhamare et al. discovered that Ba-doped ZnO samples exhibit remarkable efficiency in the photocatalytic mineralization of Linezolid antibiotics when irradiated under UV light. Various experimental parameters such as pH medium, light intensity, and photocatalyst quantity were explored, and the superior photocatalytic performance was primarily attributed to the enhanced separation of e
−/h
+ pairs [
29]. Similarly, in another study, Jayakrishnan et al. demonstrated the photocatalytic activity of Ba-doped ZnO in the photodegradation of Rhodamine B under visible light. This performance was attributed to the narrow band gap and the presence of oxygen vacancies [
30]. Behnaz et al. verified that the photocatalytic efficacy of Ba-doped ZnO in the photodegradation of Rhodamine B is mainly due to the inhibition of the recombination rate under UV light, while, under visible light irradiation, indirect photocatalytic mechanisms involving the excited adsorbed dye molecules had a pivotal role [
31]. Qiu et al. [
23] stated that the replacement of Zn
2+ ions with Mg
2+ contributes to e
−/h
+ separation, resulting in increased photodegradation of methylene blue under UV light irradiation. Opposite results were found, however, when Mg
2+ ions were inserted into interstitial sites with a do** percentage higher than 4%, because the higher conduction band (CB) potential reduced the light harvesting ability [
23]. Elhalil et al. [
24] reported that the photodegradation of caffeine occurred more rapidly using the 5% Ca–ZnO–Al
2O
3 system than bare ZnO and ZnO–Al
2O
3 systems. To explain this result, the better crystallinity of the photocatalyst was invoked, as well as its greater ability to absorb UV light with a consequent higher formation of e
−/h
+ pairs. Modvi et al. [
25] found that the insertion of Ba
2+ ions into the ZnO lattice shifted the absorption into the wavelength range of visible light, reducing the phenomenon of e
−/h
+ recombination and improving the photocatalytic efficiency. The introduction of Ca
2+ ions can also reduce the band gap energy (E
g), as described by Irshad et al. [
26]. Similarly, Oliveira et al. found that the photoactivity of ZnO can be boosted by do** with Ca
2+, due to the crystal lattice changes caused by its presence and the role of this species as electron trap [
27].
The above results demonstrate that the alkaline earth metals are promising as far as the improvement of the optoelectronic features of ZnO-based photocatalysts are concerned. However, to the best of our knowledge, the impact of this modification on the surface properties, and in particular on hydrophilicity, has not been adequately considered. In fact, the photocatalytic reactions take place on the semiconductor surface and the peculiar interaction between the surface and the substrate is a key parameter to be considered. In this investigation, some structural, morphological, optical, electronic, and surface features of Ba-doped ZnO are reported, and particular attention is devoted to the influence of the presence of barium on the surface properties of ZnO. The characterization results were correlated with the higher photocatalytic activity under solar irradiation and the higher stability of the doped material compared with the bare one against photocorrosion of ZnO, which often discourages the possible applications of this oxide especially in aqueous systems.
2. Results and Discussion
The XRD patterns of the synthesized bare ZnO and BZO photocatalysts are presented in
Figure 1. The diffraction peaks of the ZnO and BZO samples conform to the hexagonal phase of zincite, as verified by JCPDS data (Card No. 36-1451). No other barium oxide peaks or barium-containing species were detected, confirming the incorporation of Ba
2+ ions into the ZnO lattice. The radius of Ba
2+ (1.35 Å) prevents the substitution of the Zn
2+ ion, which has a smaller radius (0.74 Å). Consequently, as previously suggested [
32], Ba
2+ ions can occupy interstitial sites within the ZnO lattice. This hypothesis was verified by determining the lattice parameters. Indeed, the inset of
Figure 1 clearly reveals that the presence of Ba
2+ dopant led to a relative shift in the 2θ values of +0.093°. As a result, the parameters (a) and (c) of ZnO increased from 3.2501 and 5.2068 to 3.2519 and 5.2098 Å, respectively. Therefore, the volume (V) of the unit cell increased from 47.633 to 47.714 Å
3, indicating the inclusion of Ba
2+ ions (see
Table 1).
Table 1 displays the values of the mean crystallite size (D), along with the specific surface area (SSA) and the band gap energy values (E
g) of the samples. After the introduction of Ba dopant, the mean crystallite size (D) of ZnO decreased from 36.4 to 23.0 nm. This reduction in crystallite size due to Ba do** aligns with recent research [
30]. The error values for (a) and (c) are indicated by placing the last digit in brackets.
Figure 2 shows the SEM micrographs of the photocatalysts. It can be seen that both bare ZnO and BZO nanoparticles possess a nanowire shape and are quite homogeneous in size. However, the doped BZO particles are much smaller. The incorporation of Ba
2+ ions into the ZnO lattice prevents grain growth, in agreement with XRD results and related literature [
24,
33].
The adsorption and desorption isotherms and the pore size distribution of the bare ZnO and BZO samples are presented in
Figure 3. The isotherms of the synthesized samples are type II, as generally observed in non-porous solids. The H3 type hysteresis is typical of aggregates with formation of interparticle voids [
33]. This is in accordance with the small nanoparticle size retrieved by XRD and SEM analysis.
Diffuse reflectance spectra (DRS) of bare ZnO and BZO samples were recorded in order to explore how Ba do** could affect the optoelectronic properties of the ZnO photocatalyst (
Figure 4). The spectrum of BZO is slightly redshifted with respect to bare ZnO. The inset reports Tauc plots showing E
g values that are equal to 3.25 and 3.22 eV for bare ZnO and BZO samples, respectively.
It is evident that the insertion of Ba
2+ ions only slightly affects the optical properties of ZnO. The slight decrease in the E
g value has been explained in the relevant literature by the inclusion of intermediate energy levels below the CB of ZnO which led to better e
−/h
+ separation. These levels can be related to the influence of the 4d orbitals of Ba atoms that narrow the E
g [
30]. However, the slight narrowing of E
g observed upon Ba do** in accordance with the literature [
25,
30,
31] is unlikely to account for the relevant changes in the photocatalytic activity (see below). For this reason, we further investigated the surface properties of the samples, considering that the surface is where the photocatalytic reactions take place.
To study surface functional groups, FTIR spectra of ZnO and BZO samples were collected in the wavenumber range 450–4000 cm
−1. The spectra were normalized by considering the vibration of the Zn–O bond at about 500 cm
−1, to obtain semi-quantitative information. As illustrated in
Figure 5, the broad bands detected at 3340–3650 and 1640 cm
−1 can be attributed to vibrations of the O–H bond of adsorbed water molecules or surface hydroxyl groups [
34]. Notably, the bands are clearly smaller in the case of the ZnO sample [
35,
36,
37]. This observation suggests that the incorporation of Ba into the ZnO lattice increases the affinity of the surface for water molecules, thus endowing it with a pronounced hydrophilic character.
The Raman spectra of bare and Ba-doped ZnO samples are provided in
Figure 6. The band at 437 cm
−1 is characteristic of the hexagonal zincite structure of ZnO and it is due to the optical phonon E
2 (high) [
38]. The A
1 longitudinal optical (LO) mode appears at 331 cm
−1, the vibrational activity of A
1 transverse optical (TO) mode at 378 cm
−1, and the E
1 (TO) at 410 cm
−1. These findings are in good accordance with relevant literature [
37,
38].
BZO sample shows a broad signal at 580 cm
−1 which is slightly visible in the spectrum of bare ZnO. This band is attributed to multiphonon scattering processes and has been correlated in the literature with defectivity such as oxygen vacancies [
28].
The intensity of the peaks of ZnO is higher with respect to the BZO sample. However, the relative intensity of the peak at 330 cm
−1 with respect to the main band at 437 cm
−1 is higher, even if broader, in the BZO sample. Moreover, as previously observed, the peak at 580 cm
−1 appears clearly in the doped sample and is negligible in ZnO. The higher intensity of the latter bands, associated with longitudinal optical (LO) modes, indicates that the presence of Ba
2+ into the lattice of ZnO introduces novel defectivity. Oxygen vacancies have often been identified as responsible for enhanced water adsorption [
28,
39]. Moreover, the adsorption of hydroxyl groups into Zn
2+ sites has been demonstrated to be kinetically favored with respect to oxygen adsorption [
40]. Therefore, the defectivity induced by the presence of barium can be related with the higher hydrophilicity observed by FTIR spectroscopy.
The higher defectivity of the BZO sample can be further highlighted by comparing its fluorescence spectrum with the one of bare ZnO.
Figure 7 shows the spectra acquired on different ZnO and BZO suspensions, i.e., 0.5, 1.0, and 2.0 g∙L
−1, reported in Panels A, B, and C, respectively. Panels D and E report the deconvoluted components of the ZnO and BZO spectra, respectively, obtained for 2.0 g∙L
−1 suspensions.
The emissive behavior of ZnO and BZO samples is similar in the range between 360 and 430 nm, and the recorded intensity in this region only slightly changes upon increasing the amount of dispersed photocatalyst (A–C). Deconvolution in this region highlights three main contributions at 376, 394, and 413 nm. The first two bands can be attributed to band-to-band exciton radiative recombination and band edge emissions [
41], while the violet emission at 413 nm can be ascribed to electron transitions from shallow donor levels of neutral interstitial Zn defects (Zn
i) close to the conduction band [
42]. The luminescence spectra of ZnO and BZO become significantly different between 430 and 600 nm, where the emission of the BZO sample is stronger than the one of ZnO. Notably, the intensity of the bands in this region decreases with increasing amounts of sample dispersed in water (A–C), possibly due to scattering, trivial energy transfer, or quenching phenomena. Deconvolution in this region highlights four bands at 451, 469, 500, and 540 nm. The blue emission at 451 nm has been attributed to electron transitions to shallow levels of Zn vacancies defects (Zn
V) close to the valence band. The redshifted emission at 469 nm can be related to electron transitions from interstitial Zn (Zn
i) to Zn vacancy (Zn
V) defects levels [
42]. Finally, the two green emissions at 500 and 540 are typical fluorescence bands related to the presence of oxygen vacancies [
43] and interstitial oxygen sites [
41], respectively. Notably, these two components show higher relative intensity in the BZO sample with respect to the ZnO one (D–E), thus confirming the higher defectivity of BZO induced by the presence of barium, in agreement with XRD and Raman results.
To further confirm the increased hydrophilicity of the BZO sample, contact angle measurements were performed.
Figure 8 shows the static contact angles measured between the deposited water droplet and the surface of the ZnO and BZO photocatalysts in the dark (a and b) and after 10 min of exposure to UV light irradiation (c and d).
Regarding the BZO photocatalyst, it is evident that the water droplet exhibits a flatter shape in both dark and UV light conditions when compared to the bare ZnO sample. The images clearly demonstrate that the contact angles of ZnO are notably higher than those of BZO photocatalyst under identical conditions, particularly when exposed to UV light. Consequently, BZO sample exhibits a more pronounced hydrophilic character, which is consistent with FTIR and Raman analyses.
Similar results were reported for ZnO nanorods which showed higher wettability due to surface roughness [
44]. Mg-doped ZnO, in which, unlike barium, Mg ions were substitutionally positioned within the ZnO lattice, also exhibited an enhanced hydrophilic character [
45]. The authors accounted for the higher hydrophilicity of Mg-doped ZnO under dark conditions with the more pronounced roughness of the modified samples. This explanation, along with the novel defectivity hereby observed, may also hold for the barium-modified ZnO in the dark, considering the different surface topography shown in the SEM analyses (see
Figure 2).
It is important to mention that the heightened adsorption of water molecules significantly amplifies the generation of hydroxyl radicals via oxidation of water induced by holes (h
+) at the valence band (VB), as reported below (Equation (1)) [
46,
47].
For this reason, there is general consensus on the direct relationship between higher hydrophilicity and photocatalytic activity [
48,
49]. To confirm this hypothesis, photocatalytic degradation of 4-nitrophenol (4-NP) as a model pollutant were performed under solar radiation. The results are shown in
Figure 9.
Under solar light irradiation, the BZO sample demonstrated the highest efficiency compared to both bare ZnO and TiO
2 P25. In particular, both ZnO-based samples showed superior performances compared to P25 under sunlight irradiation, and the doped sample stood out for the higher performance. Observed kinetic constant values of 8∙10
−3 and 11.3∙10
−3 min
−1 were determined for the ZnO and BZO samples, respectively, by assuming first order kinetics. The higher activity of the BZO sample cannot be explained simply by considering its slightly redshifted absorption spectra and improved e
−/h
+ separation, as invoked in previous reports [
29,
30,
31]. The significantly higher SSA in the BZO sample suggests a greater number of catalytically active sites. However, as presented in
Figure S1, under dark conditions, the adsorbed quantity of 4-NP showed was similar for both photocatalysts. In fact, even though the BZO photocatalyst exhibited a SSA about 14 times larger than ZnO, a similar amount of 4-NP was adsorbed onto bare ZnO (ca. 11%) and BZO (ca. 15%). The adsorption results indicate that the SSA is not the primary factor influencing the BZO’s photocatalytic performance. On the other hand, it is important to note that the BZO sample exhibits a higher degree of defects, which can justify its remarkable surface hydrophilicity as observed by FTIR analyses and contact angle measurements. This aspect, often overlooked in previous studies, may have a notable impact on the photocatalytic activity. In fact, several previous studies reported the mechanism pathway of 4-NP photodegradation and showed that the degradation was mainly triggered by HO
• radicals (see Equation (2)), which induced poly-hydroxylation of the aromatic ring followed by its opening and, eventually, total mineralization [
50,
51].
To verify the effect of Ba do** on the stability of ZnO, reusability tests were performed (
Figure 10).
The progress of commercial and industrial-scale photocatalytic applications of ZnO remains hindered primarily by the adverse effects of photocorrosion, especially in aqueous systems. This fact strongly discourages the use of bare ZnO under both UV irradiation and natural solar irradiation [
52]. Zhang et al. [
53] observed a dramatic decrease in ZnO activity after a period of UV irradiation ranging from a few hours to a month. Yu et al. [
54] showed an inactivation of ZnO of about 50% after 60 min of irradiation, while, more recently, Le et al. [
55] demonstrated a ZnO weight loss of 22.3% at pH 3, 4.2% at pH 7, and 2.5% at pH 11 after five days of UV irradiation. Anodic photocorrosion of ZnO leads to the evolution of O
2 accompanied by the dissolution of Zn
2+ ions, according to the following reactions (Equations (2)–(4)) [
16,
56]:
The three tests reported in
Figure 9 consistently exhibited similar performances in successive runs, indicating the remarkable stability of barium-doped ZnO. Introducing the Ba element into ZnO clearly reduces, or even suppresses, photocorrosion phenomena, aligning with findings from previous studies on the effect of alkaline earth metals dopants [
52]. Furthermore, the enhanced surface hydrophilicity may also explain the stability of the BZO sample. It can be hypothesized that the oxidation of water favoured in the BZO sample may actually compete with the lattice oxidation of oxygen induced by the photogenerated holes in the bare ZnO sample.