Effect of Fe₂O₃–ZrO₂ Catalyst Morphology on Sulfamethazine Degradation in the Fenton-Like Reaction

Abstract

Fe₂O₃–ZrO₂ catalysts with different morphologies (nanoplates (HZNPs), nanorods (HZNRs), nanocubes (HZNCs), and nanotubes (HZNTs)) were prepared by a hydrothermal method to investigate the effect of the morphology on the catalytic performance in the Fenton-like reaction for sulfamethazine (SMT) degradation. The Fe₂O₃–ZrO₂ catalysts were characterized by scanning electron microscope (SEM), X-ray diffraction (XRD), X-ray photoelectron spectroscopy (XPS), and Brunauer–Emmett–Teller (BET) analysis. The H₂O₂ adsorption and the Fe²⁺ density sites on the Fe₂O₃–ZrO₂ catalysts had a close relationship with the morphologies and exhibited an important effect on the ·OH formation in the Fenton-like reaction. Free ·OH radicals were the main oxidative species in the reaction, and the normalized ·OH concentration per surface area of the catalysts was 4.52, 2.24, 2.20, and 0.37 μmol/m² for HZNPs, HZNRs, HZNCs, and HZNTs, respectively. The Fe₂O₃–ZrO₂ catalysts with different morphologies showed good catalytic performance, and the order of SMT degradation was HZNPs > HZNRs > HZNCs > HZNTs. Total SMT removal was achieved in the Fenton-like reaction over HZNPs at pH 3.0 and 45 °C after 240 min.

1. Introduction

Advanced oxidation processes (AOPs), involving ozone, photocatalysis, electrochemical oxidation, Fenton reaction and wet air oxidation and so on, are efficient treatment technologies for degrading toxic and/or refractory organic compounds [1,2,3,4]. Among the AOPs, the Fenton reaction (Fe²⁺/H₂O₂), as a green, simple and efficient process, can remove bio-refractory organic compounds. However, the application of the traditional Fenton reaction is limited in wastewater treatment, due to a narrow pH range and sludge production [2]. The heterogeneous Fenton-like process can avoid these shortcomings to some extent. Iron-based oxides (such as Fe₃O₄, Fe₂O₃, FeOOH, Fe⁰, etc.), which are environmental-friendly and easily available materials, exhibit good performance in the heterogeneous Fenton-like reaction [3,4,5,6,7,8]. Compared to Fe₃O₄, hematite (α-Fe₂O₃) showed good chemical stability along with poor activity. The development of iron oxide catalysts with good activity and stability is a crucial challenge for the heterogeneous Fenton-like reaction.

The performance of heterogeneous catalysts not only depends on their chemical composition, size, and surface area but also on the morphology and arrangement manner of the surface atoms. Hematite with different morphologies (such as nanocubes, nanorods, nanobelts, nanoflowers, etc.) has been widely focused [9,10]. The literature shows that the morphology of α-Fe₂O₃ can affect the catalytic efficiency of the reactions [11,12]. For instance, Zhang et al. found that hematite nanorods with exposed {110} facets exhibited a better foreign ion confinement capacity (Fe²⁺) and a higher H₂O₂ adsorption than hematite nanoplates with exposed {001} facets [13]. The result indicated that different morphologies of iron oxides could improve the catalytic activity in the Fenton-like reaction. Moreover, it was reported that the addition of transition metals (such as Ti⁴⁺, Ce⁴⁺, V⁴⁺, Zr⁴⁺, etc.) into iron oxides enhanced the catalytic activity in the Fenton-like reaction [14,15,16]. According to the above literature, we proposed a brand-new idea, synthesis of shape-controlled metallic oxides, for improving the performance of heterogeneous Fenton-like catalysts. Unfortunately, the effect of the different morphologies of bimetallic oxide catalysts on the catalytic performance has not been investigated in detail, and the reaction mechanism was not clearly in the heterogeneous Fenton-like system.

Sulfamethazine (SMT), one of the pharmaceutically active compounds (PhACs), is widely used due to its use simplicity, high chemical stability, good antimicrobial activity, fast absorption, and low price [1]. So the SMT is often detected in the aquatic environment, and selected as a target pollutant. In the study, we prepared the shape-controlled Fe₂O₃–ZrO₂ catalysts, evaluated the H₂O₂ adsorption/decomposition process to produce ·OH radicals, and investigated the effect of the morphology on the catalytic activity of the Fe₂O₃–ZrO₂ catalysts in the Fenton-like reaction for degrading SMT. Moreover, the reaction mechanism was proposed for the Fenton-like reaction over the shape-controlled Fe₂O₃–ZrO₂ catalysts.

2. Results and Discussions

2.1. Characterization of the Catalysts

2.1.1. Scanning Electron Microscope (SEM) Analysis

Figure 1 shows the morphologies of Fe₂O₃ and Fe₂O₃–ZrO₂ catalysts in the SEM images. Pure Fe₂O₃ with regular morphologies and sizes, involving nanoplates (HNPs), nanorods (HNRs), nanocubes (HNCs), and nanotubes (HNTs), was consistent with the literature [17,18]. Appropriate reagents (such as surfactant, additives, etc.) can affect the growth rate of various facets of Fe₂O₃ in the growth process [11,19]. Moreover, the addition of metal ions effectively retained the aggregation of the catalysts, and led to the formation of a crystal particle with a smaller size [16]. The above results indicated that the smaller particle and similar morphology could be obtained with the Zr addition, in good agreement with our study (in Figure 1). For the Fe₂O₃–ZrO₂ catalysts with different morphologies, the nanoplates (HZNPs) exhibited an irregular plate-like structure, and the diameter decreased from ca. 214 nm to ca. 143 nm; the nanorods (HZNRs) became like short and thick sticks, and the length was obviously reduced from ca. 695 nm to ca. 217 nm; the nanocubes (HZNCs) evolved from cube to olive-shaped with a smaller size; the nanotubes (HZNTs) gradually lost the shape of nanotubes and formed a wrinkled sheet-like structure.

2.1.2. X-ray Diffraction (XRD) and Brunauer–Emmett-Teller (BET) Analysis

For the shape-controlled Fe₂O₃–ZrO₂ catalysts, the XRD patterns revealed that diffraction peaks of α-Fe₂O₃ with hexagonal crystal structure were observed, and weaker ZrO₂ diffraction peaks were also observed (in Figure 2). Compared with Fe₂O₃ (in Figure S1), the Zr addition weakened the intensity of α-Fe₂O₃ peaks, suggesting that the Zr addition inhibited the Fe₂O₃ agglomeration and reduced the Fe₂O₃ crystallite size. According to the Scherrer’s equation, the crystallite sizes of α-Fe₂O₃ for HZNTs, HZNPs, HZNRs and HZNCs were 35, 47, 66, and 75 nm, respectively. As shown in

Table 1. Physicochemical properties of the shape-controlled Fe₂O₃–ZrO₂ catalysts.

Shape	dFe [nm]	Surface Area [m2/g]	ωFe a [wt.%]	SDIAs b [Fe/nm2]	τ Fe2+ c [at.%]	Density Site d [Fe2+/nm2]
HZNPs	47	107	0.21	21.10	0.28	5.91
HZNRs	66	97	0.26	28.81	0.18	5.18
HZNCs	75	66	0.25	40.72	0.06	2.44
HZNTs	35	193	0.19	10.03	0.09	0.90

^a The mass ratio of Fe on the catalysts surface; ^b SDIAs = ω_Fe × N_A/(SSA × 56 × 10¹⁸); ^c The atomic ratio of Fe²⁺/(Fe²⁺ + Fe³⁺) on the catalysts surface; ^d Density Site = SDIAs × τ _Fe2+.

, the surface areas for HZNTs, HZNPs, HZNRs, and HZNCs were 193, 107, 97, and 66 m²/g, and obviously increased, compared with that of pure Fe₂O₃ (in Table S1). The result was consistent with that of SEM of the Fe₂O₃–ZrO₂ catalysts.

2.1.3. X-ray Photoelectron Spectroscopy (XPS) Analysis

XPS analysis was used to characterize the chemical states of the shape-controlled Fe₂O₃–ZrO₂ catalysts. The Zr, O, and Fe peaks were observed at ca. 182, 530, and 711 eV, respectively (in Figure 3a). For pure Fe₂O₃, the Fe 2p_3/2 peak appeared at 710.6 eV and was attributed to Fe³⁺-O. With the Zr addition, the binding energy of Fe 2p_3/2 shifted from 710.6 to 710.1 eV (in Figure 3b), indicating that the chemical state of Fe element changed in the Fe₂O₃–ZrO₂ catalysts. The Fe 2p_3/2 peaks were fitted into two peaks and assigned to Fe²⁺-O (ca. 708.5 eV) and Fe³⁺-O (ca. 710.6 eV) [20]. According to the fitting results, the different Fe²⁺ amounts for HNZPs, HNZRs, HNZCs, and HNZTs were 28, 18, 6, and 9 at. %, respectively. As iron oxides are used as catalysts in the Fenton-like reaction, Fe²⁺ favors the ·OH formation, and a high Fe²⁺ concentration is advantageous to improve the degradation of organic pollutants [16].

2.2. The Activity of the Fe₂O₃–ZrO₂ Catalysts

2.2.1. The Interaction of Hematite Facets with H₂O₂

The H₂O₂ adsorption and activation on the catalyst surface are the important steps to produce ·OH radicals and degrade the pollutants in the heterogeneous Fenton-like reaction. High H₂O₂ adsorption could favor the H₂O₂ activation to produce ·OH radicals. Figure 4 shows the H₂O₂ adsorption on the shape-controlled Fe₂O₃–ZrO₂ catalysts at pH 3.0 under 5 °C. The catalysts exhibited a rapid H₂O₂ adsorption process, and then reached adsorption equilibrium. The order of the H₂O₂ adsorption capacity was HZNPs > HZNRs > HZNCs > HZNTs, and the H₂O₂ adsorption amounts on the catalysts were 45.6, 24.5, 16.3, and 14.3 mg/g_cat, respectively. The normalized H₂O₂ adsorption amount per surface area of the Fe₂O₃–ZrO₂ catalysts was calculated, and is shown in

Table 2. H₂O₂ absorption and decomposition properties of the shape-controlled Fe₂O₃–ZrO₂ catalysts.

Sample	AH2O2a [mmol/L]	A’H2O2b [mg/m2]	BH2O2c [%]	kH2O2d [min−1]	kH2O2/ QH2O2 [g/(mg·min)]
HZNPs	0.67	0.43	80.1	6.70 × 10−3	1.47 × 10−4
HZNRs	0.36	0.25	49.8	3.00 × 10−3	1.23 × 10−4
HZNCs	0.21	0.25	37.2	2.00 × 10−3	1.23 × 10−4
HZNTs	0.24	0.07	23.5	1.20 × 10−3	8.40 × 10−5

^a H₂O₂ maximum adsorption amount; ^b Adsorption amount of H₂O₂ per unit area; ^c H₂O₂ decomposition rate at 240 min; ^d H₂O₂ decomposition rate constant.

. It was found that HZNPs showed the highest H₂O₂ adsorption capacity (0.43 mg/m²) which was ca. six times higher than that of HZNTs (0.07 mg/m²).

The H₂O₂ decomposition on the shape-controlled Fe₂O₃–ZrO₂ catalysts was evaluated at pH 3.0 and a temperature of 45 °C. As shown in Figure 5a, the H₂O₂ decomposition rates of HZNPs, HZNRs, HZNCs, and HZNTs after 240 min reaction were 80%, 50%, 37%, and 24%, respectively. Moreover, the H₂O₂ decomposition obeyed a pseudo-first kinetic equation (in Figure 5b). The apparent H₂O₂ decomposition rate constants (k_H2O2) were 6.70 × 10⁻³, 3.00 × 10⁻³, 2.00 × 10⁻³ and 1.20 × 10⁻³ min⁻¹, and the order was HZNPs > HZNRs > HZNCs > HZNTs, respectively. To further investigate the H₂O₂ decomposition efficiency on the Fe₂O₃–ZrO₂ catalysts, we normalized k_H2O2 with its maximum adsorption capacity of H₂O₂ (Q_H2O2), and the result is shown in Table 2. It was observed that the k_H2O2/Q_H2O2 of HZNPs with a lower surface area had the highest value which was ca. 1.8 times that of HZNTs with a higher surface area. The results indicated that: (1) the H₂O₂ adsorption/decomposition process was related to the morphology, not to the surface area of the Fe₂O₃–ZrO₂ catalysts; (2) a higher adsorption/decomposition on the catalyst surface could be helpful to generate ·OH radicals in the Fenton-like reaction.

In the heterogeneous Fenton-like reaction, the ·OH radicals are known as the main active species for degrading organic compounds. To obtain insight on the ·OH radicals generated on the shape-controlled Fe₂O₃–ZrO₂ catalysts, a quantitative determination was carried out in the presence of H₂O₂ at pH 3.0. For pure Fe₂O₃, there was almost no ·OH formation (not shown), despite a good H₂O₂ adsorption obtained as shown in Figure S2. For the Fe₂O₃–ZrO₂ catalysts, the generated ·OH radicals were detected and the result was shown in Figure 6. The ·OH radicals rapidly increased in the initial 120 min reaction, and the concentration was ca. 241.8, 108.8, 72.9, and 35.9 μmol/L for HZNPs, HZNRs, HZNCs, and HZNTs at 240 min, respectively. The HZNPs exhibited a higher H₂O₂ decomposition rate and a greater ·OH concentration than the other catalysts. The normalized ·OH concentration per surface area of the catalysts in 240 min reaction was 4.52, 2.24, 2.20, and 0.37 μmol/m² for HZNPs, HZNRs, HZNCs, and HZNTs, respectively. The normalized ·OH amount of HZNPs was ca. 12 times that of HZNTs. The results further confirmed that the ·OH generation was independent of the specific surface area and was mainly affected by the morphologies of the Fe₂O₃–ZrO₂ catalysts. In addition, the free ·OH radicals generated were the main chemical state for total ·OH radicals in the Fenton-like reaction (in Figure S3). The shape-controlled Fe₂O₃–ZrO₂ catalysts favored the free ·OH formation when they reacted with H₂O₂. It was found that the free ·OH concentration for all catalysts was obviously higher than that of the SMT in the solution, indicating that the desorption of the generated ·OH radicals from the catalyst surface was not a limiting step for the SMT degradation in our experiments.

2.2.2. The Activity of the Fe₂O₃–ZrO₂ Catalysts

The Fenton-like reaction degrading the SMT was conducted at pH 3.0 and with an initial SMT concentration of 10 mg/L over the shape-controlled Fe₂O₃ and Fe₂O₃–ZrO₂ catalysts (see Figure S4 and Figure 7a). In the H₂O₂ and Fe₂O₃/H₂O₂ systems, almost no SMT removal was obtained after 240 min, indicating that both systems were not active for removing SMT. In the heterogeneous Fe₂O₃–ZrO₂/H₂O₂ system, the SMT removal significantly increased, while the SMT adsorption on the Fe₂O₃–ZrO₂ catalysts was negative (<5%). The result indicated that the shape-controlled Fe₂O₃–ZrO₂ catalysts exhibited good activity in the Fenton-like reaction of the SMT degradation. Moreover, the activity order was as follows: HZNPs > HZNRs > HZNCs > HZNTs, and HZNPs exhibited the best catalytic activity, whose SMT removal reached 98% after 240 min.

As shown in Figure 7b, a pseudo-first kinetic equation was obtained in the Fenton-like reaction for the SMT degradation over the shape-controlled Fe₂O₃–ZrO₂ catalysts. The apparent rate constants (k) for HZNPs, HZNRs, HZNCs, and HZNTs were 1.5 × 10⁻², 6.8 × 10⁻³, 1.7 × 10⁻³, and 9.3 × 10⁻⁴ min⁻¹, respectively. We normalized the rate constant (k) with the surface area of the Fe₂O₃–ZrO₂ catalysts. The value of HZNPs (1.4 × 10⁻⁴ g × min⁻¹ × m⁻²) was the highest, which was ca. two times that of HZNRs (7.0 × 10⁻⁵ g × min⁻¹ × m⁻²) and ca. 29 times that of HZNTs (4.8 × 10⁻⁶ g × min⁻¹ × m⁻²). The results indicated that the activity of the shape-controlled Fe₂O₃–ZrO₂ catalysts was closely related to the morphology, not the surface area of the catalysts.

During the Fenton-like reaction, the Fe ion leaching was measured after 240 min, and a low concentration of Fe ion leaching was obtained (<0.2 mg/L). The results indicated that the Fe₂O₃–ZrO₂ catalysts had good stability in the Fenton-like reaction for degrading SMT. The intermediates for the SMT degradation over HZNPs were detected by liquid chromatography–mass spectrometry (LC–MS). Several main intermediate products were obtained, including 4-amino-N-carbamimidoyl-benzenesulfonamide, 4,6-dimethyl-2-hydroxypyrimidine, sulfanilamide, aniline, and carboxylic acid, in good agreement with the literature [21].

2.3. Mechanism Analysis

In the heterogeneous Fenton-like reaction, the H₂O₂ adsorption and the H₂O₂ activation on the catalyst surface were the key steps limiting ·OH generation. In the literature [22], the adsorption of atoms, ions, and molecules on a catalyst surface is related to their morphologies, that is, the intrinsic crystal structure. For iron-based catalysts, the substance adsorption was both related to the exposed facets and the surface densities of the iron atoms (SDIAs) [13,17]. In our study, the order of the H₂O₂ adsorption on pure Fe₂O₃ was nanorods > nanoplates > nanocubes > nanotubes (see Figure S2), in agreement with the above result. With the Zr addition, the SDIAs decreased (in Table 1), and led to a decrease of H₂O₂ adsorption on the Fe₂O₃–ZrO₂ catalysts (in Figure 4). It was noticed that HZNPs and HZNRs showed an obvious change in H₂O₂ adsorption. For Fe₂O₃ nanoplates and nanorods, the main {001} and exposed {110} facets could adsorb the substances on the catalyst surface, and {110} facets had a higher reactivity than {001} facets in the adsorption process [13]. This suggested that Fe₂O₃ nanorods exhibited a better adsorption capacity than Fe₂O₃ nanoplates. Therefore, Zr ions, as foreign substances, could be easily confined on the Fe₂O₃ nanorods, indicating that more Zr ions occupied the exposed facet of HZNRs. This result led to the decrease of the SDIAs on the HZNRs. HZNPs exhibited a better H₂O₂ adsorption capacity than HZNRs.

Another key step limiting the ·OH generation, H₂O₂ activation, had a close relationship with the Fe²⁺ amount on the surface for the iron-based catalysts in the Fenton-like reaction [16]. The Zr addition obviously increased the surface area and induced the presence of Fe²⁺ on the catalyst surface. Moreover, the Fe²⁺ density site (Fe²⁺/nm²) of the Fe₂O₃–ZrO₂ catalysts (see Table 1) showed a closed relationship with the morphologies of the catalysts, rather than the surface areas. HZNPs with a low surface area had the highest Fe²⁺ site density, suggesting that the morphology affected the Fe²⁺ density sites on the surface of the Fe₂O₃–ZrO₂ catalysts.

In the heterogeneous Fenton-like system for degrading organic compounds, a radical chain reaction is commonly involved [2,5,6,16]. In our study, there were important reaction processes for ·OH radical generation, as follows. First, the H₂O₂ adsorption was the first step for the H₂O₂ activation to ·OH radicals on the catalyst surface. A higher H₂O₂ adsorption was helpful to react with Fe²⁺ on the catalysts and produce ·OH radicals. Second, the H₂O₂ activation occurred according to the reaction between ≡Fe²⁺ with the adsorbed H₂O₂ by electron transfer (Equation (1)). The result suggested that the ·OH radical formation was attributed to the H₂O₂ adsorption and Fe²⁺ density sites on the Fe₂O₃–ZrO₂ catalysts. Then the ·OH radicals, adsorbed on the surface by weak chemical bonds, easily desorbed to the free ·OH radicals in the solution, and rapidly oxidized organic pollutants. In the Fenton-like reaction, ZrO₂ with an amorphous and small particle on the catalyst surface, enhanced the regeneration process of ≡Fe²⁺ from ≡Fe³⁺ in the presence of H₂O₂ at a low pH value (Equations (2)–(4)) [16]. Therefore, in the Fenton-like reaction process over the Fe₂O₃–ZrO₂ catalysts, the higher H₂O₂ adsorption and Fe²⁺ density sites favored greater ·OH generation. The result suggested that HZNPs exhibited the best SMT removal, due to the highest H₂O₂ adsorption and Fe²⁺ density sites.

$$\equiv {\mathrm{Fe}}^{2+}{+\mathrm{H}}_2{\mathrm{O}}_2\to \equiv {\mathrm{Fe}}^{2+}{-\mathrm{H}}_2{\mathrm{O}}_2\to \equiv {\mathrm{Fe}}^{3+}{+·\mathrm{OH}+\mathrm{OH}}^{-}$$

(1)

$$\equiv {\mathrm{Fe}}^{3+}{+\mathrm{H}}_2{\mathrm{O}}_2\to \equiv {\mathrm{Fe}}^{2+}{+\mathrm{O}}_2{·}^{-}/{\mathrm{HO}}_2{·+\mathrm{H}}^{+}$$

(2)

$$\equiv {\mathrm{Fe}}^{3+}{+\mathrm{O}}_2{·}^{-}/{\mathrm{HO}}_2·\to \equiv {\mathrm{Fe}}^{2+}{+\mathrm{O}}_2{+\mathrm{H}}^{+}$$

(3)

$$\equiv {\mathrm{Zr}-\mathrm{HO}}_2^{-}{+\mathrm{H}}_2{\mathrm{O}}_2\to \mathrm{OH}·+\equiv {\mathrm{Zr}}^{+}{-\mathrm{O}}_2{·}^{-}{+\mathrm{H}}_2\mathrm{O}$$

(4)

3. Materials and Methods

3.1. Materials and Chemicals

In the experiments, chemical reagents without further purification were obtained from Aladdin Industrial Corporation in China. Methanol, acetic acid, and H₂O₂ (30 wt.%) were purchased from the Sigma-Aldrich Company.

3.2. Synthesis Methods

Fe₂O₃ and Fe₂O₃–ZrO₂ catalysts were prepared by the hydrothermal method according to the literature [17,18]. The Fe to Zr atomic ratio was one.

3.2.1. Preparation of Fe₂O₃–ZrO₂ and Fe₂O₃ Nanoplates

Amounts of 1.092 g of FeCl₃ × 6H₂O and 1.289 g of ZrOCl₂ × 8H₂O were dissolved in ethanol (40 mL) with a trace addition of distilled water (2.8 mL) under vigorously stirring. Sodium acetate (3.2 g) was added under stirring for 1 h. Then, the mixture was sealed in a Teflon-lined stainless steel autoclave (Kenuo, Beijing, China, 100 mL) and maintained at 180 °C for 12 h. After natural cooling to room temperature, the resulting solid was washed with distilled water and ethanol several times to remove any impurities and dried at 100 °C overnight. Finally, the products were ground and calcined at 500 °C for 3 h under air to obtain HZNPs. Pure Fe₂O₃ nanoplates (HNPs) were prepared in a similar way.

3.2.2. Preparation of Fe₂O₃–ZrO₂ and Fe₂O₃ Nanorods

Amounts of 1.683 g of Fe(NO₃)₃ × 9H₂O and 1.788 g of Zr(NO₃)₄ × 5H₂O were dissolved in 45 mL of deionized water. After completely dissolving, NaOH solution (10 wt.%) was added dropwise to adjust the pH to 12 ± 0.1, and then the mixture was vigorously stirred for 30 min. The mixture was transferred into a Teflon-lined stainless steel autoclave (100 mL) and kept at 180 °C for 12 h. After washing and drying the precipitate as in the previous method, the solid was calcined at 500 °C for 3 h under air to obtain HZNRs. Pure Fe₂O₃ nanorods (HNRs) were prepared in a similar way.

3.2.3. Preparation of Fe₂O₃–ZrO₂ and Fe₂O₃ Nanocubes

Amounts of 0.811 g of FeCl₃ × 6H₂O and 1.288 g of Zr(NO₃)₄ × 5H₂O were dissolved in 50 mL of cetrimonium bromide (CTAB) aqueous solution (C_CTAB = 0.04 mol/L) under vigorous stirring for 30 min, and then transferred into a Teflon-lined stainless steel autoclave (100 mL) and maintained at 120 °C for 12 h. After washing and drying the precipitate as in the previous method, the solid was calcined at 500 °C for 3 h under air to obtain HZNCs. Pure Fe₂O₃ nanocubes (HNCs) were prepared in a similar way.

3.2.4. Preparation of Fe₂O₃–ZrO₂ and Fe₂O₃ Nanotubes

Amounts of 1.081 g of FeCl₃ × 6H₂O and 1.289 g of ZrOCl₂ × 8H₂O were dissolved in 60 mL of NH₄H₂PO₄ solution (C_NH4H2PO4 = 0.04 mol/L). After stirring for 30 min, the mixture was transferred into a Teflon-lined stainless steel autoclave (100 mL) and kept at 220 °C for 12 h. After washing and drying as in the previous method, the solid was calcined at 500 °C for 3 h under air to obtain HZNTs. Pure Fe₂O₃ nanotubes (HNTs) were prepared in a similar way.

3.3. Catalyst Characterization

The morphology of the catalysts was obtained by a Hitachi SU8010 field emission scanning electron microscope (FESEM) (Hitachi, Tokyo, Japan) using an accelerating voltage of 5 kV. The specific surface area of the catalysts was measured by N₂ adsorption at 77 K using an Autosorb iQ-MP system (Quantachrome Instruments, Boynton Beach, FL, USA). Before analysis, the sample was outgassed at 200 °C for 4 h. The X-ray diffraction (XRD) pattern of the catalysts was measured on a Bruker D8 Focus diffractometer (Bruker, Karlsruher, Germany) using Cu Kα radiation (λ = 0.15406 nm). X-ray photoelectron spectroscopy (XPS) analysis for the catalysts was performed on a PHI 5400 ESCA analyser with an Al Kα X-ray source (PHI, Chanhassen, MN, USA) (hν = 1486.60 eV). The C1s peak (E_b = 284.60 eV) was used as a reference for the calibration of the binding energy.

3.4. H₂O₂ and ·OH Concentration Measurement

Titanium oxalate spectrophotometry was used for detecting the H₂O₂ concentration in the Fenton reaction. The concentration of hydroxyl radicals (·OH) was analyzed with a modified methanol probe molecule method [23]. Hydroxyl radicals reacted with methanol to form formaldehyde, and then the formaldehyde was determined by the 2,4-dinitrophenylhydrazine (DNPH) derivation method. The concentration of DNPH–formaldehyde derivative was measured by high-performance liquid chromatography (HPLC, Agilent 1260, Agilent, Santa Clara, CA, USA) with a C18 column at 400 nm.

3.5. Fenton Degradation Experiments

The experiments were performed in a 500 mL glass flask in the dark, equipped with a stirrer and a thermocouple. First, the initial pH of the SMT solution was adjusted to 3.0 ± 0.05 by adding 5 wt.% HCl, and then a catalyst and an SMT solution (10 mg/L, 250 mL) were loaded into the reactor with stirring. The reactor was heated to 45 °C, and then maintained for 30 min to achieve adsorption equilibrium on the catalyst. Secondly, H₂O₂ was added into the reactor, and the point was defined as “zero” time of the Fenton-like reaction. Samples were periodically withdrawn from the reactor, quenched with methanol, and then filtered through a 0.22 μm filter. The SMT concentration was analyzed by HPLC. The mobile phase was composed of methanol and water (v/v = 60:40). The flow rate was 1.0 mL/min and the detection wavelength was 254 nm.

4. Conclusions

Fe₂O₃–ZrO₂ catalysts with different morphologies were produced successfully by a hydrothermal method. The H₂O₂ adsorption and the Fe²⁺ density sites had a combined effect on ·OH generation in the Fenton-like reaction over the Fe₂O₃–ZrO₂ catalysts. Moreover, the H₂O₂ adsorption and the Fe²⁺ density sites depended on the morphology, rather than the specific surface area. The normalized H₂O₂ adsorption capacity per surface area on the catalyst surface was 0.43 mg/m² for HZNPs, and ca. 1.68, 1.72 and 5.75 times that of HZNRs, HZNCs, and HZNTs, respectively. The Fe²⁺ density sites (Fe²⁺/nm²) were 5.91, 5.18, 2.44, and 0.90 for HZNPs, HZNRs, HZNCs, and HZNTs, respectively. The normalized ·OH concentration per surface area of the catalysts was 4.52, 2.24, 2.20, and 0.37 μmol/m² for HZNPs, HZNRs, HZNCs, and HZNTs, respectively. HZNPs and HZNRs showed a more effective catalytic performance for degrading SMT at pH 3.0 in the Fenton-like reaction. This work indicated that the shape-controlled bimetallic oxides effectively improved the catalytic performance in the heterogeneous Fenton-like reaction.