Characterization of Synthesized Nanocomposites
In order to reveal the variation in the crystal structure between as-synthesized ZnPc/TiO2 nanocomposites and TiO2 nanoparticles, XRD analyses were carried out (Fig. 3). In the XRD patterns of both nanoparticles, the major characteristic peaks were localized at (2θ) of 25.4°, 37.8°, 48.3°, 53.9°, 55.1°, 62.8°, 68.9°, 70.3° and 75.1°, indexing to the anatase structure of TiO2 compatible with JCPDS Card no. 21-1272. These (2θ) values indicate the (101), (112), (200), (105), (210), (204), (116), (220), (215), and (303) crystal plane of anatase phase.
Moreover, it was not observed the diffraction peak of the rutile phase. It is worth noting that metallo phthalocyanines (ZnPc) are known to show unsharpened peaks in XRD analyses due to their amorphous nature [39]. The average crystallite size of pure TiO2 and ZnPc/TiO2 nanoparticles was calculated to be 16.81 nm and 18.76 nm using the Debye-Scherrer equation, d = 0.9λ/β cos θ where λ is the incident wavelength (Cu K α = 1.5406 Å), β = full width at half-maximum (FWHM) and θ = angle of reflection, respectively. It is understood from this XRD pattern that ZnPc molecules are dispersed into the lattice structure of TiO2, but it is shown that broadening at the major peaks of ZnPc/TiO2 nanocomposites is wider than that of TiO2 particles. This alteration may be a sign of morphing of the crystal structure of TiO2 coming from the immobilization of ZnPc molecules despite its lower concentration.
SEM analyses were evaluated aiming to investigate the morphology and surface properties of the pristine TiO2 and ZnPc/TiO2 photocatalysts. Comparing the SEM image of ZnPc/TiO2 with that of pure TiO2, it was found that ZnPc molecules with macrocyclic structure covered the TiO2 photocatalyst, as can be seen in Fig. 4.
The SEM micrographs showed that the morphology of fabricated ZnPc/TiO2 particles was almost broccoli-like structure because the calcination process improved the interaction between nanoparticles and agglomerated with Pc molecules on the particles’ surface. It was observed agglomeration of nanoparticles and increasing in particle sizes with ZnPc doping. The fact that the ZnPc molecules were used low amount and surrounded the surface of the TiO2 catalyst can result in negative data about crystallize size, which are in agreement with the XRD results. EDX analysis was performed on prefabricated nanocomposites to determine all of the atoms in the composite structure (Fig. 4), which demonstrated the existence of C, O, N, Zn, and Ti in the naked TiO2 and synthesized ZnPc/TiO2. More importantly, it is confirmed that the decline in the weight % of major element Ti associated with arising N and Zn elements is referred to as immobilization of ZnPc molecules on the TiO2 catalyst surfaces.
The FT-IR spectra of both catalysts, shown in presented in Fig. 5, reveal several differences between the ZnPc dye pigment that includes lots of N and C atoms as against pure TiO2.
Especially, it has been shown characteristic aliphatic –C-H stretching vibration (2988–2900 cm− 1), stretching vibrations of -N-H peaks (3676 − 3663 cm− 1), and -C = C/C = N stretching bands (1534–1408 cm− 1). Striking peaks appearing in the fingerprint region nearly at 1059 cm− 1 could have been the signal of –C-N stretching. FT-IR spectra of ZnPc/TiO2 exhibits around 1007 − 650 cm− 1, which can be associated with the Ti-O-Ti or O-Ti-O stretching vibration.
The UV–Vis diffuse reflectance spectroscopy is used to research the optical spectral properties of pure TiO2 and ZnPc doped TiO2 photocatalytic composite materials in the range of 200 nm to 800 nm. As shown Fig. 6 shows, the band gaps were calculated by transforming obtained data from the UV-DRS spectra with the Tauc plots equation (αhν = A(hν-Eg)1/2, Eg: Band gaps) [40].
The band gap values of TiO2 nanopowders and ZnPc/TiO2 (Fig. 3) were found to be 3.08 eV and 3.14 eV, respectively. These results revealed that Zn phthalocyanine molecules loading to TiO2 photocatalyst could slightly enlarge the photo-response of TiO2 into the UV-vis region due to the π-π transitions and ligand-to-metal charge transfer transitions from ZnPc to Ti atoms [41, 42]. Zinc phthalocyanine molecules in the ZnPc/TiO2 photocatalyst reduce the VB-CB band gap and thus the absorption spectrum of TiO2 shifts into the visible region. As a result, the band gap energies of produced ZnPc/TiO2 nanocomposite was found to be less than the band gap energy of naked TiO2. Reducing the band energy between the valence and conduction bands by performing phthalocyanine modifications facilitates the photocatalytic research in the visible region and enriches the photocatalytic performance of composite photocatalytic materials.
Photodegradation Tests of AMX Degradation at Different pH Values
It is known that the notable effect of pH on the degradation of AMX is not only due to the ionization of AMX molecules but also the charge of TiO2 surface at different pH values [43]. Since AMX have different ionizable functional groups, namely carboxyl (pKa1 = 2.68), amine (pKa2 = 7.49) and phenol (pKa3 = 8.94), the AMX molecule has different ionization phases [44, 45]. Therefore, the percent degradation of AMX depends on the pH of the solution, since the charge of AMX changes from positive in pH < pHpzc to negative state in pH > pHpzc of TiO2 [8, 46].
When Fig. 7 was evaluated, as the pH increases, the degradation efficiency of AMX increases under the different illumination sources. Noteworthy degradation was anticipated under UV-A illumination because standardized AMX has a maximum absorption at 229 [47], which means that AMX could be thought that occurred degradation can be because of hydrolysis of AMX without catalyst at lower than UV-A. In the case of pure TiO2 or ZnPc/TiO2 nanocatalyst, the degradation of AMX can be clarified by considering the factors of catalyst type, light source, and structure of AMX at different pH values. As shown in Fig. 7, it was found that the increase of pH accelerated the hydrolysis of AMX because the β-lactam ring of AMX did not exhibit rigid behavior at high pH [47, 48]. Nevertheless, it can be said that the degradation of AMX is lower, since both the AMX molecule and the surface of TiO2 nanoparticles are positively charged at pH 3, which paves the way for the electrostatic repulsion between the catalyst and the pollutant, which could restrain the adsorption of AMX on the surface of the photocatalyst types [49, 50]. It was appeared that pure TiO2 nanoparticles provided a weak degradation rate at both pH values under visible light. Moreover, both TiO2 (64.79%) and ZnPc/TiO2 (66.38%) displayed the same photocatalytic behavior under UV-C irradiation at pH 11, but TiO2 nanoparticles (63.08%) showed a higher degradation rate than ZnPc immobilized TiO2 composites (40.17%) when the AMX solution was exposed to UV-A light irradiation. This could be a reason that the higher catalytic activity of ZnPc/TiO2 compared to bare TiO2 is related to the proximity of the energy states belonging to the TiO2 conduction band and the LUMO level of ZnPc including the π-conjugated system. Both the migration of the electrons from VB of the TiO2 nanoparticle and from the LUMO of ZnPc to CB of the TiO2 nanoparticle excited by the UV-Vis light source could be sparked off production of more active radically species, which can oxidize AMX rapidly [51–53].
It was monitored that the basic pH compared to pH 3, which improved not only by the hydrolysis of pharmaceutical drug but also from densely electron transition, enhanced disintegrate of AMX on account of synergistic effect between ZnPc and TiO2. If the pH of the solution is kept at a lower acidic level, the TiO2 nanoparticles could converge to form an agglomerate, resulting in a decrease in the amount of photons absorbed on the catalyst, which can then affect the photodegradation yield [54]. When AMX is irradiated with UV-A and pH 3, either the light source could be damage the structure of Pc derivative, causing some disjunction outputs of the Pc molecule, or intermediates of AMX that cannot be degraded could be found in the irradiation solution. Therefore, LC/MS2 studies were performed to estimate the DPs of AMX.
3.3. LC/MS2 Analysis
To determine the possible DPs of AMX, LC/MS2 analysis was performed with an antibiotic concentration of 20 mg/L and 1 g/L ZnPc/TiO2 catalyst mass under UV-C at pH 11. Opening of the β-lactam ring of amoxicillin leads to the formation of its diastereoisomers known as penicilloic acid, AMX-S oxide, penilloic acid, diketopiperazine, and phenol hydroxypyrazine [1, 7, 12, 55, 56].
As displayed in Fig. 8–10, LC/MS2 analysis of 20 mg/L AMX (M = C16H19N3O5S; molecular weight = 365.384 g/mol) has revealed minor and major molecular and fragment ions formed through by these impurities. These are as follows: Fig. 8 the typical AMX fragmentation pattern, the m/z: 349 ion peak [C16H17N2O5S]+ could be associated with either the loss of an –NH3 group from AMX or NH3 and H2O groups from penicilloic acid ([M + H]+: 384), which is one of the diastereoisomers of AMX and has characteristic fragments at m/z 325 [C11H21N2O5S2]+, 234 [C12H13N2O3]+, 207 [C10H11N2O3]+/[C10H10NO2S]+, 180 [C5H10NO2S2]+, 137 [C7H9N2O]+, and 114 [C4H4NOS]+/[C5H8NS]+.
Even assuming that the ion peak observed at m/z: 366 with lower relative abundance refers to the molecular ion peak [M + H]+ of AMX, it could be that this ion peak originated with the splitting off H2O loose from the protonated molecular ion of penicilloic acid. In addition, the fragment ion with m/z: 160 [C6H10NO2S]+ was formed when the five-membered thiazolidine ring.
Moreover, the peak at m/z: 277 [C14H17N2O2S]+ peak could be associated with the cleavage of [MH-NH3-CO2-CO2]+ from AMX-S oxide. Figure 9 shows that no fragment ion peaks attributable to the hydrolysis of AMX were observed at m/z: 82, 102 [C5H12NO]+/[C4H8NS]+, 123 [C7H8NO + H]+, 148.9[C5H10NO2S]+, 202 [C8H12NO3S+], 273.1 and 384 ([M + H], M: penicilloic acid).
The detected degradation product after irradiation in the presence of ZnPc/TiO2 showed fragmen ions at m/z: 123 [C7H8NO + H]+, 261, 365 [C16H17N2O6S]+/[C17H21N2O5S]+, and 403.6, respectively, different from those of TiO2 nanoparticles. 365 [C16H17N2O6S]+/[C17H21N2O5S]+, 384 [C16H22N3O6S]+ and nearly 261 degradation ions appeared in addition to the characteristic peaks in seen both Fig. 9 and Fig. 10 after the photocatalytic treatment process, which identified all these ions qualified in previous research [1, 7, 12, 55, 56]. The peak at m/z: 261 could have resulted from the loss of water, carbon dioxide groups, and deamidation of the protonated penilloic acid ([M + H]: 340). In addition, the coupling of the bond that binds the ammonia group to the protonated AMX-S oxide could be thought to lead to the appearance the m/z: 365 fragment ion peak. Finally, the positively charged ion at m/z 403.9 could have been caused by a potassium adduct of the protonated diketopiperazine amoxicillin at m/z 366.0.
As displayed in Fig. 11, LC/MS2 analysis of 20 mg/L AMX under different conditions was carried out to analyze the effects of the catalyst and pH. It can be said that the type of catalyst and the pH of the aqueous solution of AMX affect the removal. Considering the number and intensity of fragment ions, neither the catalyst nor UV radiation alone may be very effective. To follow the degradation of AMX solution without any treatment (without catalyst, UV and pH) and with UV after 150 minutes, the suspensions were pipetted at regular intervals, then centrifuged and filtered. At the end of 150 minutes, the Uv-vis spectrum of the AMX solution was monitored with a spectrophotometer, and the specific peak related to AMX (235 nm) showed no significant changes.
The proposed degradation products formed by utilization of fragment ions at characteristic m/z were summarized in Table 2.
Degradation products, variously identified or not were observed as a result of cleavage of groups from the AMX molecule or cleavage of chemical bonds under UV light at pH 11 in the presence of the ZnPc/TiO2 catalyst. Therefore, it could be concluded that either 20 mg/L AMX solution was not completely degraded, in short not mineralized until CO2, H2O and mineral ions, or ZnPc molecules could have broken down during photocatalysis of AMX and found as decomposition products in the solution.
3.4. Total Organic Carbon (TOC) Analysis
To accurately determine the degree of mineralization of the AMX antibiotic in the presence of zinc doped TiO2 photocatalysts prepared by the dye sensitization technique, total organic carbon (TOC) analysis was also evaluated. An initial concentration of 20 mg/L AMX was used for TOC testing under conditions similar to the photodegradation process.
The data show in Fig. 12 that although the pure TiO2 samples (9.04%) slightly exhibited higher mineralization efficiency than of AMX (6.51%), the highest degree of mineralization of AMX was observed in the photocomposite sample with ZnPc sensitized TiO2 (10.43%) after 150 minutes of treatment. This indicates that AMX was successfully degraded during the photocatalytic process and it may have been presented in smaller organic structures in the medium. The degradation of AMX by ZnPc/TiO2 nanocomposite was based on dye sensitization to form the degradation products, while AMX was not markedly mineralized to carbon dioxide. It could be concluded that while pH affected the decomposition of AMX pharmaceuticals, both catalysts used in the photocatalytic oxidation of AMX did not generate the sufficiently radical species. These TOC results are compatible with the LC/MS2 data showing unidentifiable fragments as well as identifiable mid-products during the photocatalytic experiments. Kanakaraju et al. stated that nominal mineralization was pursued because DPs generated during during AMX removal competed with adsorption and photocatalytic degradation, also some may resist oxidation [45].
Service procurement for LC/MS2 analysis was carried out to forecast the DPs subsequent to AMX degradation. By virtue of the frequent lack of availability of the required standards and the cost of them to estimate the likely DPs in LC/MS2 analysis, the intermediate products were generally surmised by making use of pre-made literature data on the DPs of AMX.
Clarification of Photocatalytic Mechanism of AMX
A schematic representation of the mechanism for the degradation of AMX by the ZnPc/TiO2 photocatalyst is illustrated in Fig. 13. The phthalocyanine molecules behave as p-type semiconductors in the solid state with reducing the energy of the band gap at about 2 eV, while that of TiO2 corresponds to 3.2 eV [57, 58]. In other words, the conduction band of ZnPc is more negative than the CB of TiO2 semiconductor, whereas the VB of TiO2 is much more positive than the fitting band of ZnPc [57]. When sensitized ZnPc/TiO2 is irradiated with UV light, the excited ZnPc injects electrons from its LUMO level into the conduction band of the semiconductor, which should increase the formation of the most common reactive oxygen species such as electrons (e−), holes (h+) and reactive oxygen species such as superoxide anions (O2−), hydroxyl radicals (HO·), singlet oxygen (1O2). Moreover, ZnPc can capture holes from the valence band of the semiconductor and directly induce the formation of singlet oxygen (which can also AMX) by interacting with molecular oxygen [59].
By inter system crossing, the energy of the excited ZnPc is transferred to the adsorbed molecular O2 on the TiO2 surface, producing 1O2, which is in equilibrium with the hydroxyl radicals in accordance with Eqs. (1), (2), and (3). At the same time, the excited ZnPc injects its electron into the conduction band of TiO2, resulting in more electrons (e− CB) on the TiO2, which can transfer to O2 or 1O2 to generate the superoxide radicals (O2−.) (Eq. 4), which upon protonation yield hydroperoxy radicals (HO2.) (Eq. 5), produce hydroxyl radicals (OH.), which are responsible for the degradation of AMX in the ZnPc-TiO2 system (Eqs. 6–7). AMX molecule is readily decomposed by these reactive charges attack AMX molecule, and then it subsequently mineralizes until the CO2, H2O and mineral ions (Eqs. 8–9). The sequential mechanism by which TiO2 is sensitized for photocatalytic degradation of AMX by ZnPc can be expressed as follows [23, 29];
$${ZnPc}^{*}+{O}_{2}?ZnPc+{}^{1}{O}_{2}$$
2
$${ZnPc}^{*}+{TiO}_{2}?{ZnPc}^{*}+{TiO}_{2}\left({e}^{-}\right)$$
3
$${TiO}_{2}\left({e}^{-}\right)+\left({O}_{2} or {}^{1}{O}_{2}\right)?{TiO}_{2}+{O}_{2}^{.-}$$
4
$${O}_{2}^{.-}+{H}_{2}O?{HO}_{2}^{.}+{OH}^{-}$$
5
$${HO}_{2}^{.}+{H}_{2}O?{H}_{2}{O}_{2}+{OH}^{.}$$
6
$${H}_{2}{O}_{2}?{2OH}^{.}$$
7
$${OH}^{.}+AMX?{CO}_{2}+{H}_{2}O+{NO}_{3}^{-}+{SO}_{4}^{2-}$$
8
$${ZnPc}^{*}+AMX?ZnPc+{\left(AMX\right)}^{.+}$$
9