A series of experiments were carried out by varying the amount of catalyst (0.5 to 4.5 g/L) to establish the effect of photocatalyst loading and to avoid unnecessary excess photocatalyst. The percentage of photodegraded p-cresol as a function of irradiation time was plotted in Figure 2. The maximum percentage of photodegradation was obtained at 2.5 g/L of photocatalyst. This is due to the increase in number of active sites with increasing photocatalyst loading which consequently leads to enhanced production of ^•OH radicals. Moreover, the number of adsorbed p-cresol molecules increased with the increase in the number of photocatalyst particles, thus increase the percentage of photodegradation [27]. When the amount of photocatalyst was increased beyond optimum, the percentage degradation decreased. This can be attributed to the increase in the turbidity of the solution that reduces the light penetration through the solution known as light screening effect [7]. Furthermore, agglomeration and sedimentation of photocatalyst particles is also possible [28]. In such condition, a part of the photocatalyst surface probably becomes unavailable for photon absorption and p-cresol adsorption, thus reduces the photocatalytic reaction.

The photocatalytic degradation of various p-cresol concentrations was studied. The results (Figure 3) showed that the percentage degradation decreased with increasing p-cresol concentration. At a high p-cresol concentration (150 ppm), the presumption is that the active sites are covered by p-cresol and its intermediates that can cause reduced generation of electron-hole pair (e⁻-h⁺), which subsequently reduces the photodegradation efficiency [28]. Since the mass of photocatalyst, the intensity of light and illumination time were kept constant, as the initial p-cresol concentration increased, the ^•OH and O₂^−• species formed on the surface of photocatalyst would therefore remain constant. Thus the relative ratio of the ^•OH and O₂^−• for attacking p-cresol decreased which led to a reduction in percentage degradation of p-cresol[29]. Another factor which may be responsible for the reduction in photocatalytic degradation rate is the competition between adsorbed p-cresol and H₂O molecules for photodegraded h⁺ [25].

The rate of the photocatalytic degradation vs. p-cresol concentration exhibited a quadratic behaviour which has an optimum value at 100 ppm (Figure 4). The rate is a nonlinear function of p-cresol concentration (C_cresol). The reason for this behavior may be related to the probability of interaction between p-cresol and ZnO surface. At very low p-cresol concentration, the probability of interaction between p-cresol molecules and surface of ZnO decreases. Hence, photodegradation rate decreases. The probability of interaction between substrates molecules and oxidizing species is increased by increasing p-cresol concentration, leading to an enhancement in the degradation rate [29–31]. Therefore, according to Langmuir-Hinshelwood’s modeling [32], the suggested model is:

where, the estimated value of k_cresol (mg L⁻¹ min⁻¹ ppm⁻¹) and K_cresol (ppm⁻¹) were 0.0403 and 0.33 with R-squared value 0.98 respectively. Similar results were reported for photodegradation kinetics of aqueous sodium oxalate, sodium dodecyl sulfate, 4-nitrophenol solution using TiO₂ photocatalyst [33–36] and our previous work [25].

It is important to study the effect of pH in the photodegradation of p-cresol because the charge of ZnO surface and p-cresol vary with shift in pH. Primarily, ZnO will be hydroxylated in the presence of water to form hydroxide layers (Zn-OH) [37]. The zinc hydroxide surface (Zn-OH) can become charged by reacting with H⁺ (acidic environment) or OH⁻ (basic environment) ions due to surface amphoteric reactions (Equations 11,12) [38].

At low pH hydroxide surfaces adsorb protons to produce positively charged surfaces. At high pH (usually above pH 9) they lose protons to produce negatively charged surfaces [39]. The zero point charge of ZnO (pH_zpc) has been reported to be 9 [40]. Therefore, the surface functional groups of ZnO can be ZnOH₂⁺, ZnOH, and ZnO⁻ at pH < pH_zpc, pH_zpc and pH > pH_zpc, respectively.

On the other hand, since p-cresol is a strong base (pK_a ≈ 10), in acidic conditions, more p-cresol molecules would tend to be positively charged while the lone pair of electrons of the OH group will be more available for hydrogen bonding (undissociated p-cresol) [41]. At high pH (pH ≥ pK_a), p-cresol would exist as a negatively charged 4-methylphenolate species (Equation 13).

The optimum condition was used as in the previous experiment. As observed, the percentage photodegradation of p-cresol increased slightly with increasing pH from pH 6 to 9 (Figure 5). The increase in photodegradation percentage may be due to an increase adsorption of p-cresol on the ZnO surface (Figure 6).

It has been reported that in a slightly alkaline solution (pH 8–9), hydroxyl radicals are more easily generated by oxidizing the available OH⁻ on the photocatalyst surface [42]. Thus, generally, the photodegradation efficiency is expected to be enhanced with increasing pH due to the ready availability of hydroxyl radicals for the reaction. However, a decrease in photodegradation percentage was observed at pH 10 (Figure 5). This can be attributed to the reduction for p-cresol adsorbed on the catalyst surface at pH 10 (Figure 6). It should also be noted that hydroxyl radicals are rapidly scavenged in presence of excess concentrations of hydroxyl ions and therefore would not have the opportunity to react with the substrates [43]. Hence, a drastic drop in the amount of p-cresol photodegraded was observed at pH 10.

The details of the mechanism of p-cresol degradation in the presence of hydroxyl radical (^•OH) until mineralization have been reported in a previous study with boron-doped diamond [44]. It is well known that aromatic intermediates are formed from phenolic compounds, which result thereafter in the formation of carboxylic acids. These carboxylic acids are easily transformed into CO₂ and H₂O. The CO₂ is volatized as gas from the solution [23,45].

Some of the photoproducts of photodegraded p-cresol were determined using UPLC (Figure 7A). The chromatograms show many peaks which may be due to unidentified aromatic intermediates [23]. In this study, the intermediates detected during 120 min of reaction are 4-hydroxy-benzaldehyde (0.2 ppm) and 4-methyl-1,2-benzodiol (0.01 ppm) with retention time (R_t) = 2.145 and R_t = 3.005 min respectively (Figure 7B). Both intermediates were reported earlier [45]. The ability of semiconductor photocatalysts to remove pollutants is based on the active oxidizing species (HO^•, O₂^−•, H₂O₂ and .h⁺) which are produced by the irradiation of the semiconductor (Equations 1–8). The hole is produced as a result of photoinduced charge separation which may then cause the formation of hydroxyl radical by directing the hole transfer reaction with ⁻OH and H₂O. Hydrogen peroxide may be generated via chain reactions involving conduction band electron. Even though hole and hydrogen peroxide may aid the degradation of p-cresol on ZnO dispersion, experimental evidences have proved the main oxidant to be hydroxyl radical [46,47]. Therefore, we propose the mechanistic pathways shown in Figure 8 to account for the photoproducts of p-cresol degradation encountered in this study based on hydroxyl radical intervention. Thus, 4-hydroxy-benzaldehyde and 4-methyl-1,2-benzenediol can be produced by through reactions shown by Equation 14–15 and Equation 16, respectively. It is believed that these intermediates are on the brink of ring opening [23]. Thereafter carboxylic acids are formed (Equations 17 and 18). The chain of carboxylic acids is decreased with increasing irradiation time. The final carboxylic acid by product is oxalic acid (HOOC-COOH) which is easily transformed to CO₂ and H₂O (Equation 19) by taking two hydroxyl radicals.

Mineralization, the main aim of photodegradation of p-cresol, was followed by measuring the total organic carbon (TOC) [48] and total inorganic carbon (TIC). Figure 9 shows the TOC and TIC values of p-cresol. The amount of TOC steadily decreased with increasing irradiation time, which indicates the decline of p-cresol (or intermediates) when irradiation time is increased. On the other hand, the TIC curve shows that the amount of total inorganic carbon was generated in the first hour of reaction and was constant thereafter. This is most likely due to some organic carbon converting into inorganic carbon, for example, carbonate ions, and after a few minutes TIC remained constant. Based on the results, it can be concluded that 93% organic carbon is removed from p-cresol solution as CO₂. The residual TOC value (7%) indicates the presence of other photoproducts such as carboxylic acids at the end of the reaction.

The reusability of photocatalyst was investigated in order to establish the stability (Figure 10) while studying reuse of photocatalyst; all parameters including irradiation time, pH, p-cresol concentration, amount of photocatalyst and irradiation time were kept constant. The photocatalyst was separated from the solution mixture through filtration. The recovered photocatalyst was washed five times with deionized water, dried at 96 °C in oven and reused five times as in the previous degradation process. Results show no significant reduction in photocatalytic performance in photodegrading p-cresol, thus this indicates the stability of ZnO as a photocatalyst. Moreover, inductively coupled plasma (ICP) showed that the photocorrosion of ZnO is quite insignificant.