NPX (A-methyl-6-methoxy-2-naphthaleneacetic acid, 98% purity) was obtained from West Asia Reagent Company. Acetonitrile (CR), methanol (CR), and ethanol (CR) were obtained from USA ACS Enke Chemical. FeSO₄·7H₂O (AR), NH₃·H₂O (AR), EDTA (AR), NaOH (AR), H₂SO₄ (AR), KI (AR), IPA (Isopropanol, AR), NaN₃ (AR), and p-BQ (p-Benzoquinone, AR) were obtained from Shanghai Aladdin Bio-Chem Technology Co., Ltd. Ultrapure water employed in the experiments was obtained via an integrated Smart2 Pure ultrapure water system, obtained from TKA Wasseraufbereitungs system GmbH, Germany.

Freshly prepared under magnetic stirring, 10 ml of pure NH₃·H₂O was added to a 110 ml 0.3 mol·L⁻¹ FeSO₄ solution, which had a pH of 8.6. Subsequently, 10 ml of 0.015 mol·L⁻¹ EDTA and ultrapure water were added to a 150-ml volume. Then, 1 L·min⁻¹ O₂ was introduced into the solution for about 30 min until the precipitation color changed from blue-green to orange under a controlled system temperature of 20°C, after which the pH was maintained at 4.3. Once the orange precipitate was filtered and rinsed, it was placed in a vacuum drying oven for 24 h at 30°C. Thereafter, the sample was finely ground and screened (200 mesh) (He et al., 2005).

X-ray diffraction (XRD) was carried out with a Cu K(α) source (λ = 0.15406 nm) at 40 kV and 30 mA over the range of 2θ = 20–80°.

Scanning electron microscopy (SEM) was used for investigating the morphology and dispersion of the samples. Prior to measurements, the sample was affixed to an aluminum sheet and sprayed with gold.

BET surface area (BET) analysis was used to determine the pore structure, specific surface area, and porosity of the samples. The porosity and pore distribution were determined by a nitrogen adsorption–desorption isotherm and the Barrett–Joyner–Halenda (BJH) method.

The physical properties of metal oxides, such as their crystal structures and surface characteristics, may influence their photocatalytic activity. Eight peaks were observed in the XRD pattern of FeOOH that were attributed to the (120), (011), (031), (111), (060), (220), (151), and (080) planes, indicating the presence of the γ structure (Figure 2A). When compared with the standard diffraction peaks of γ-FeOOH, the prepared powder was in the form of a pure crystal phase.

Surface morphological studies of γ-FeOOH by SEM revealed that the prepared powder was in the form of a mixed crystal phase, which contained, for the most part, nanoparticles (~50 nm) and short nanorods (~200 nm in length) (Figure 2B), due to differences in pH during the preparation of γ-FeOOH (Farcasiu et al., 1991). The smooth and dispersible properties of the mixed crystal phase revealed that the morphology was relatively regular.

The adsorption capacity of metal oxides for organic pollutants is affected by the BET size. On the basis of the N₂ adsorption–desorption isotherm and pore size distribution, the specific surface area of γ-FeOOH was determined to be 125.7 m²·g⁻¹ (Figure 2C). As shown in Figure 2C, we found that there was a significant hysteresis loop in the adsorption–desorption curve, which means that the sample possessed a mesoporous structure. According to the BJH desorption curve method (Kruk et al., 1997), which was employed to calculate the pore size distribution, the pore size range of the sample was ~50 nm. In photocatalytic experiments, γ-FeOOH was recovered and washed with pure water three times before drying. The XRD pattern showed no obvious change (Figure 2D) after this test. The degradation of NPX was reduced by only 1% when performing photodegradation with the recovered γ-FeOOH. These results show that the γ-FeOOH photocatalyst is highly stable.

As shown in Figure 3A, γ-FeOOH displayed a typical absorption edge at ~650 nm, and a bandgap width of 1.94 eV was calculated (Figure 3B). In order to further study the bandgap position of the semiconductor γ-FeOOH, X-ray photoelectron spectroscopy (XPS) was used to probe the valence band (XPS-VB). This revealed that the valence band of γ-FeOOH was located at 1.80 eV, as shown in Figure 3C. Therefore, it can be deduced that the rewind position of γ-FeOOH was −0.14 eV and the band gap structure is shown in Figure 3D.

A suspension was formed in the multiphase photocatalytic reaction system, as the catalyst is insoluble in water. With increased catalyst dosages, the effective surface area of the solution was increased; hence, its reaction efficacy was enhanced proportionally. Excessive catalyst loading caused reflection and scattering, which reduced the transmittance of the solution and thus the catalytic efficiency. It was observed that the γ-FeOOH dosage played a very important role in the photodegradation of NPX. To investigate the effect of γ-FeOOH dosage on the photodegradation of NPX, γ-FeOOH solutions were prepared at concentrations of 0.05, 0.1, 0.2, 0.4, and 0.6 g·L⁻¹, which were then introduced into separate NPX solutions. As shown in Figure 4A, the data collected from the photodegradation of NPX following the addition of different concentrations of γ-FeOOH were fitted to a first-order kinetic equation. It was observed that the NPX photodegradation rate increased with increased γ-FeOOH loading in water.

When the dosage of γ-FeOOH was varied from 0.05 to 0.6 g·L⁻¹, the NPX photodegradation rate increased from 0.0344 to 0.0509 min⁻¹. The position and photogenic charge of photocatalytic reactions in the system were enhanced with increased γ-FeOOH loading; however, the shielding, reflection, and scattering of light were increased with higher γ-FeOOH loads. With appropriate loads of γ-FeOOH, the transmittance of light in the solution decreased and the reaction rate slowly increased.

During the process of photodegradation, the relationship between the reaction rate constant k and the concentration of γ-FeOOH was fitted to the following empirical formula (Galindo et al., 2001):

where n is the correlation index and [γ-FeOOH] is the concentration of γ-FeOOH (g·L⁻¹).

The kinetic constants of NPX photodegradation and the dosage of γ-FeOOH (0.05 to 0.6 g·L⁻¹) in this experiment were analyzed by linear regression, with the relationship between k and [γ-FeOOH] shown in Figure 4B:

To: k = 0.05577[γ − FeOOH]^0.1394

It has been suggested that the charge transfer process between the contaminants adsorbed to the catalyst surface and the light-generated active species (h⁺, ·OH, and O₂·) facilitates the photocatalytic oxidation of pollutants in solution. Therefore, the coverage of pollutants on the catalyst surface has an important influence on photocatalytic activity.

In this section, the photodegradation of NPX by 0.2 g·L⁻¹ γ-FeOOH was investigated at initial NPX concentrations of 5, 10, 20, 30, and 40 mg·L⁻¹, with the results shown in Figure 5A. These experiments revealed that the photodegradation of NPX followed first-order kinetics at different initial concentrations upon the addition of γ-FeOOH. The activities of semiconductor photocatalysts arise primarily from photogenic e⁻ and h⁺, where, in the competitive process of photocatalysis, they may be recombined very rapidly (generally at the nanosecond level) (Hoffmann et al., 1995). From the kinetics perspective, only the adsorbents on the surface of the catalyst may be oxidized by e⁻. However, our results revealed that the NPX photodegradation rate decreased with higher initial concentrations in solution.

Under a certain light intensity, higher initial NPX concentrations resulted in a lower population of photons available per NPX molecule; hence, a lower photodegradation rate was obtained. An identical result was reported in previous NPX research (Ma et al., 2013). Secondly, higher initial NPX concentrations, with additional particles adsorbed to the γ-FeOOH surface, acted to lower the number of photocatalytically active sites that were available at the surface. Hence, the population of photogenerated e⁻/h⁺ pairs per unit of time was correspondingly reduced. Simultaneously, prior to the photodegradation of NPX molecules, they were required to undergo charge exchange with the active species generated at the γ-FeOOH surface and diffuse into the solution. Finally, when the initial NPX concentration was increased, it was difficult to completely decompose the reaction-generated intermediate products in a timely manner. This increased the opposition against adsorption to the surface of the γ-FeOOH, where these intermediates could once again reform the NPX matrix. Therefore, the photodegradation rate was finally decreased.

We considered the derivative of the obtained first-order kinetic equation with respect to t and set t = 0 to obtain the photodegradation rate r₀ under different initial concentrations of C₀, as shown in Figure 5B. When the initial concentration of NPX was increased from 5 to 40 mg·L⁻¹, the initial photodegradation rate r₀ also increased gradually, from 0.1415 to 0.7997 mg·L⁻¹·min⁻¹. This indicated that the photocatalytic degradation of NPX occurred on the surface of γ-FeOOH, and the photodegradation rate was an increasing function of the level of surface adsorption. When the Metastable-Equilibrium Adsorption Theory (Pan and Liss, 1998) is regarded under certain thermodynamic conditions, the adsorption amount is related to the surface binding strength and the adsorption configuration, while being balanced with the concentration of the solute. In this section, when the initial NPX concentration was raised, the coverage rate of the NPX molecules on the surface of γ-FeOOH increased accordingly. Consequently, the electron transfer efficiency of the NPX molecules that was adsorbed to the surface and the photogenerated charge were increased, which led to an increase of the initial photodegradation rate r₀.

A large quantity of experimental data has indicated that the photocatalytic degradation of organic pollutants on the surface of semiconductors conforms to the Langmuir–Hinshelwood kinetic equation (Hoffmann et al., 1995; Houas et al., 2001; Andreozzi et al., 2003; Du et al., 2008; Li et al., 2008). The applicable premise of the L–H kinetic equation is that the organic pollutant molecules are adsorbed to a solid surface (Turchi and Ollis, 1990; Alfano et al., 1997). Although researchers have not clarified the photocatalytic mechanisms of FeOOH, surface complexes (Faust and Hoffmann, 1986) and semiconductor-initiated photocatalytic mechanisms (Bandana et al., 1999) have had their respective supporters. More recent studies have supported semiconductor photocatalytic mechanisms and highlighted the role of organic pollutant molecules adsorbed to the FeOOH surface (Bandana et al., 1999, 2001a,b). In examining the FeOOH-facilitated photocatalysis of orange II, Du et al. (2008) analyzed the initial reaction rate, amount of FeOOH surface adsorption, and the position of the FeOOH activity. Thus, Du considered that the FeOOH photocatalytic reaction takes place at the solid surface; therefore, the available L–H kinetic equation could be employed to describe FeOOH photocatalysis. Based on this, 1/C₀ and 1/r₀ were calculated according to Equation (2), and a plot was created for 1/C₀-1/r₀ (as shown in Figure 5C). A linear relationship was found between them within the experimental concentration range (R² = 0.9996), k_L−H = 2.1867 mg·L⁻¹·min⁻¹, K = 0.01377 L·mg⁻¹. This signified that the photocatalytic degradation of NPX on the surface of γ-FeOOH satisfies the L–H kinetic equation, and that the adsorption of NPX on γ-FeOOH is of importance to its photocatalytic degradation (Li et al., 2008).

Within the range of experimental concentrations, the photodegradation kinetic constant of NPX gradually decreased (from 0.0285 to 0.0200 min⁻¹), whereas the correlation coefficient R² decreased from 0.9949 to 0.9791. The relationship between the reaction rate constant k and the initial substrate concentration during the photocatalytic process could be generally described by the following empirical formula:

where n is the correlation index and [NPX] is the initial concentration of NPX (mg·L⁻¹).

Linear regression was used to analyze the relationship between the NPX photodegradation kinetic constant and its initial concentration (5–40 mg·L⁻¹) within the experimental range. It can be seen in Figure 5D that the relationship between the reaction rate constant k and NPX concentration was as follows:

It is understood that pH is a critical factor that influences the photocatalytic degradation kinetics during semiconductor multiphase photocatalysis. First, pH can change the charge properties of the catalyst surface and affect how organic molecules adsorb on the catalyst surface (Barnard et al., 2005). Secondly, photogenerated charge carriers can combine with H⁺/OH⁻ in solution to form active species, such as OH⁻, which can capture photogenerated holes H⁺ to form ·OH. Finally, pH may alter the electron cloud density distribution of organic molecules, thus affecting photocatalytic degradation.

As the effects of pH on photocatalytic degradation kinetics are relatively complex, definitive studies of γ-FeOOH are relatively rare. In this section, to investigate the effects of initial pH on the photodegradation of γ-FeOOH/NPX reaction systems, the initial NPX concentration was established as 10 mg·L⁻¹, whereas that of γ-FeOOH was 0.2 g·L⁻¹, and the initial pH of the photodegradation solution was set at 5, 7, and 9. As shown in Figure 6A, at a pH of 5, the γ-FeOOH /NPX system demonstrated the fastest photocatalytic rate with a pH of 9 in the second place, while the slowest rate was observed at a pH of 7, much the same as the photocatalytic rates observed in water (Figure 6B).

As pH_ZPC = 8.47 for γ-FeOOH, the hydroxylation of the γ-FeOOH surface in an alkaline solution could allow OH⁻ to react with h⁺ to produce ·OH as follows:

Concurrently, the ether bonds of the NPX molecules are less stable under acidic conditions (Chen et al., 2013). Hence, γ-FeOOH was favorable for the photocatalytic degradation of NPX at pH < 7.

Based on the above analysis, when the pH was low, the stability of the ether bonds within the NPX molecules was decreased, which enabled the γ-FeOOH-based photocatalytic degradation of NPX. When the pH was high, OH⁻ combined with h⁺ to form ·OH, which facilitated the photocatalytic degradation of NPX. Due to the combined effect of these two factors, the reaction rate was lowest when the pH was 7 within the range of our experiments.

Quenching experiments were carried out (Figure 7A) by measuring the generation of active species during the photodegradation of NPX in pure water. It can be seen that there was not only direct photodegradation caused by ³NPX^*, but also self-sensitized photodegradation involving hydroxyl radicals (·OH), singlet oxygen (¹O₂), and superoxide anions (O2·-) which were produced in the photodegradation process of NPX (Figure 7B) (Zhanyi et al., 2018).

To further investigate the active radicals that participate in the photodegradation of NPX, electron paramagnetic resonance (EPR) measurements were carried out. As shown in Figure 8A, there was no signal in the dark, while the signal 1:1:1:1 appeared after 5 min of illumination. It could thus be concluded that O2·- was present and its concentration would be increased by illumination. As shown in Figure 8B, it was observed that the signal 1:2:2:1 appeared. This suggests that ·OH appeared and increased in concentration with illumination. It was also confirmed that ¹O₂ was present by TEMP from Figure 8C, while the signal 1:1:1 was detected in the light. So, the active radicals in the γ-FeOOH/NPX system were evidenced by electron spin resonance (ESR).

Photocatalytic degradation typically generates a variety of active substances, such as h⁺, e⁻, ·OH, ¹O₂, and O2·- (Hao et al., 2016), with the production processes shown in Equations (8)–(14):

According to the quenching experiment described in the section “Active Species Analysis,” when KI, IPA, NaN₃, and BQ were added to the solution, the photocatalytic NPX degradation rate was reduced by different degrees, as shown in Figure 8D. It may be seen from Figure 8D that free radicals such as h⁺, e⁻, ·OH, ¹O₂, and O2·- were involved in the γ-FeOOH-mediated photocatalytic degradation of NPX.

On one hand, γ-FeOOH is a semiconductor material with an energy band structure, where the energy barrier (Eg) between the valence band (VB) and the conduction band (CB) is only 2.2 eV. When the γ-FeOOH surface was irradiated with photons of energy equal to or greater than the forbidden band, the e⁻ in the VB were excited and jumped to the CB, with h⁺ being generated in the VB to form e⁻/h⁺ pairs. Because of the discontinuous region between the energy bands, the resulting e⁻/h⁺ pairs had greater longevity; hence, they migrated to the particle surface in large quantities. The oxidizing properties of the nanoparticle surfaces were potent enough to oxidize the NPX molecules that were adsorbed to the γ-FeOOH surfaces. Additionally, the cavities reacted with H₂O molecules, which were also attached to the surface of γ-FeOOH, which then generated ·OH. Due to the strong oxidization ability of ·OH, the NPX molecules on the surface of γ-FeOOH could also be oxidized and degraded. Simultaneously, conducting electrons were combined with O₂ at the surface of the γ-FeOOH to generate O2·-, which could also facilitate the oxidative degradation of NPX.

On the other hand, under sunlight exposure, Fe (III) could accelerate the oxidation of carboxylic acid. As NPX contains a carboxyl group, it could form strong complexes with Fe (III), which rapidly photochemically reacted under light irradiation (Zuo and Hoigne, 1992; Faust and Zepp, 1993), thereby accelerating the oxidative degradation of NPX. Several studies have suggested that the photochemical reaction of these complexes follows the H₂O₂ production process in water.

For this photodegradation experiment, the effect of hydrokinetics must also be considered, as light exposure under stirring was first applied. With agitation, the mass transfer rate of NPX from the solution to the γ-FeOOH surface was increased, so additional NPX was oxidized prior to e⁻/h⁺ recombination and thus the photodegradation of NPX was increased. Moreover, DO present in the solution could capture the photogenerated electrons generated during the photocatalytic process, which reduced the probability of the recombination of photogenerated electrons and holes, and thus increased the probability of holes oxidizing the NPX. When exposed to UV light, the e⁻ at the surface of the γ-FeOOH could reduce O₂ to O2·-, as shown in Equation (20). Subsequently, O2·- reacted with photogenerated holes h⁺ to form ·OH or peroxide in the presence of organic capture agents, as in Equations (21)–(23). Each of these species contributed to the photodegradation of NPX.

The degradation by-products of NPX on the γ-FeOOH /NPX system were identified by Thermo Scientific Ultimate 3000 RSLC and Q Exactive Orbitrap (HRLC-MS-MS). As shown in Figure 9, seven intermediates were detected. From attacked by h⁺, e⁻, ·OH, ¹O₂, and O2·-, compounds were generated because of the losses of the CO₂, H₂O, and/or CH₃ group. According to the deduced structure of the compounds and the early study, we speculated the reasonable reaction approach as shown in Figure 10.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.