Metal–Organic-Framework-Derived Ball-Flower-like Porous Co3O4/Fe2O3 Heterostructure with Enhanced Visible-Light-Driven Photocatalytic Activity

A porous ball-flower-like Co3O4/Fe2O3 heterostructural photocatalyst was synthesized via a facile metal–organic-framework-templated method, and showed an excellent degradation performance in the model molecule rhodamine B under visible light irradiation. This enhanced photocatalytic activity can be attributed to abundant photo-generated holes and hydroxyl radicals, and the combined effects involving a porous structure, strong visible-light absorption, and improved interfacial charge separation. It is notable that the ecotoxicity of the treated reaction solution was also evaluated, confirming that an as-synthesized Co3O4/Fe2O3 catalyst could afford the sunlight-driven long-term recyclable degradation of dye-contaminated wastewater into non-toxic and colorless wastewater.


Introduction
Various pollutants in water environments can directly cause serious harm to the lives and health of human beings, animals and plants. Organic dyes, for example rhodamine B (RhB), methylene blue and methyl orange, as one of the most common industrial pollution sources at present, have attracted tremendous attention because of their geno-and ecotoxicity [1][2][3][4][5][6]. Therefore, the development of water treatment technologies regarding dye degradation has become a top priority. Among various methods, photocatalysis is recognized as one green and efficient alternative for organic pollutant degradation, where its key issue lies in the facile preparation of highly active and stable photocatalysts [7][8][9][10].
As one of the most promising multi-functional materials, metal-organic frameworks (MOFs) are often considered to be novel photocatalysts due to their abundant and editable active sites and large surface area. However, some of their defects, such as poor light absorption and metal ion leaching due to an unstable structure, may seriously limit their practical applications [11][12][13]. In order to solve these problems, in this study, a flower-like cobalt 2,5-thiophenedicarboxylic coordination polymer (Co-TDC) was used as a template to synthesize a novel Co 3 O 4 /Fe 2 O 3 heterostructural photocatalyst with improved light harvesting and photocatalytic performance. The facile preparation, structural versatility,

Photocatalysis Measurements
The adsorption and photocatalysis processes of as-prepared catalysts were evaluated by the degradation of RhB in an aqueous solution under visible light irradiation at room temperature (ca. 25 • C). A 500 W xenon lamp with a cut-off filter (λ > 420 nm) was used to generate visible light. Amounts of 0.1 g of catalyst powder and 50 mL RhB aqueous solution (initial solution pH ≈ 4) were added to a 100 mL quartz tube and continuously stirred during the degradation experiment. Before irradiation, the reaction solution was magnetically stirred in the dark for 30 min to reach complete adsorption/desorption equilibrium. During the photocatalytic experiment, 5 mL reaction solution was extracted every 10 min, and the concentration of residual RhB was determined by measuring its absorbance at 590 nm on a UV-visible spectrometer (UV-3600i Plus by Shimadzu, Kyoto, Japan). The 5 mL solution was added back into the reaction solution after measurement.

Results and Discussion
The chemical composition and crystal structure of CF were analyzed by XRD. As shown in Figure 1a, the characteristic diffraction peaks located at 19 (Figure 1b), the asymmetric peaks at around 780.7 eV and 796.8 eV, and shake-up type satellite peaks at 785.8 eV and 802.3 eV of Co-TDC, could be well-indexed to Co 2+ , implying that cobalt in Co-TDC was only in the form of Co (II). On the other hand, for CF, two new peaks could be identified at around 779.4 eV and 794.3 eV, which were both ascribed to Co 3+ [14][15][16][17][18]. This revealed that Co 2+ in Co-TDC was partially oxidized to Co 3+ during calcination, and thus Co 3 O 4 was obtained as a result. Meanwhile, Fe 2p spectrum of CF was also recorded, as shown in Figure S1. It was revealed that Fe 3+ ions were still dominant, which corresponded to the Fe 2 O 3 phase. However, the minor peak at around 732.2 eV suggests that a little Fe 3+ was reduced to Fe 2+ along with the oxidation of Co 2+ to Co 3+ [19][20][21][22][23].
However, the minor peak at around 732.2 eV suggests that a little Fe 3+ was reduced to Fe 2+ along with the oxidation of Co 2+ to Co 3+ [19][20][21][22][23].  Figure 2a,b that CF has a regular ball-flower-like morphology with a spherical size ranged at 10-20 μm, which was retained from the Co-TDC template, as shown in Figure  S2 of the Supplementary Materials. It is worth noting that the sheet-like fundamental units of Co3O4 in CF became much more porous after calcination, with large numbers of ~200 nm Fe2O3 nanoparticles ( Figure S3) embedded within the pores, as indicated by the yellow arrows in Figure 2c, which facilitate the adsorption and degradation of dye molecules on the surface. Moreover, the elemental mapping profiles in Figure S4 also help to verify that the distribution of Fe2O3 within highly porous Co3O4 is uniform while it is random. In order to further determine the chemical composition of the synthesized catalyst, HRTEM image was also recorded, as shown in Figure 2d. The identified two lattice fringes with an interval of 0.25 and 0.20 nm could be indexed to the (311) facet of Fe2O3 and (400) facet of Co3O4, respectively, which is in good agreement with the XRD result.  Figure 2a,b that CF has a regular ball-flower-like morphology with a spherical size ranged at 10-20 µm, which was retained from the Co-TDC template, as shown in Figure S2 of the Supplementary Materials. It is worth noting that the sheet-like fundamental units of Co 3 O 4 in CF became much more porous after calcination, with large numbers of~200 nm Fe 2 O 3 nanoparticles ( Figure S3) embedded within the pores, as indicated by the yellow arrows in Figure 2c, which facilitate the adsorption and degradation of dye molecules on the surface. Moreover, the elemental mapping profiles in Figure S4 also help to verify that the distribution of Fe 2 O 3 within highly porous Co 3 O 4 is uniform while it is random. In order to further determine the chemical composition of the synthesized catalyst, HRTEM image was also recorded, as shown in Figure 2d. The identified two lattice fringes with an interval of 0.25 and 0.20 nm could be indexed to the (311) facet of Fe 2 O 3 and (400) facet of Co 3 O 4 , respectively, which is in good agreement with the XRD result.
The degradation efficiencies of different samples for RhB are displayed in Figure 3a. When the catalyst was not present in the solution, RhB could hardly undergo self-degradation under visible light (i.e., black plots). The reaction solution was first stirred in the dark for 30 min for the catalyst-RhB interface to reach the adsorption/desorption equilibrium. Typically, the contribution of RhB removal by adsorption is lower than 20%, which is in proportion to the surface area of the catalyst. In photocatalysis systems, CF demonstrated a superior performance than Co-TDC and Fe 2 O 3 , indicating that CF possesses the highest photocatalytic activity. This could be explained by the following aspects: (i) The highly porous structure of CF provided abundant active sites, as revealed in Figure 2b,c [24][25][26]; (ii) The p-n heterojunction that formed between Fe 2 O 3 and Co 3 O 4 could promote the separation of photo-generated electron and hole pairs [27][28][29][30]. The promoted charge separation, and thus the inhibited charge recombination, was witnessed by the significantly decreased photoluminescence (PL) intensity of CF composites compared to pristine Fe 2 O 3 particles, as displayed in Figure S5 of the Supplementary Materials [31][32][33][34][35][36][37][38]. The variation in the RhB degradation efficiency of CF in different pH conditions is presented in Figure 3b, suggesting that the catalyst could maintain a superior photocatalytic degradation activity in the pH range of 4-10, despite the fact that the degradation rate decreased to a certain extent in a strong acid environment (pH ≤ 2). This may be due to the dissolving of oxides by strong acid, resulting in a loss of active material in the CF catalyst for the degradation of RhB. However, considering that the actual surface water or groundwater is mostly weakly acidic or weakly alkaline, the CF catalyst is still applicable to the oxidative degradation of organic pollutants in natural water bodies. The degradation efficiencies of different samples for RhB are displayed in Figure 3a. When the catalyst was not present in the solution, RhB could hardly undergo self-degradation under visible light (i.e., black plots). The reaction solution was first stirred in the dark for 30 min for the catalyst-RhB interface to reach the adsorption/desorption equilibrium. Typically, the contribution of RhB removal by adsorption is lower than 20%, which is in proportion to the surface area of the catalyst. In photocatalysis systems, CF demonstrated a superior performance than Co-TDC and Fe2O3, indicating that CF possesses the highest photocatalytic activity. This could be explained by the following aspects: (i) The highly porous structure of CF provided abundant active sites, as revealed in Figure 2b,c [24][25][26]; (ii) The p-n heterojunction that formed between Fe2O3 and Co3O4 could promote the separation of photo-generated electron and hole pairs [27][28][29][30]. The promoted charge separation, and thus the inhibited charge recombination, was witnessed by the significantly decreased photoluminescence (PL) intensity of CF composites compared to pristine Fe2O3 particles, as displayed in Figure S5 of the Supplementary Materials [31][32][33][34][35][36][37][38]. The variation in the RhB degradation efficiency of CF in different pH conditions is presented in Figure 3b, suggesting that the catalyst could maintain a superior photocatalytic degradation activity in the pH range of 4-10, despite the fact that the degradation rate decreased to a certain extent in a strong acid environment (pH ≤ 2). This may be due to the dissolving of oxides by strong acid, resulting in a loss of active material in the CF catalyst for the degradation of RhB. However, considering that the actual surface water or groundwater is mostly weakly acidic or weakly alkaline, the CF catalyst is still applicable to the oxidative degradation of organic pollutants in natural water bodies.  The service life of a catalyst is an important technical indicator for evaluating its potential for practical usage. After the reaction, the catalyst in the solution could be easily and quickly separated due to its magnetism, as revealed in Figure 3c. Then, the recycled CF catalyst was rinsed with ethanol solution to remove the residual organics on the surface. Afterwards, it could be reused for RhB removal under the same conditions, as presented in Figure 3d. An excellent degradation efficiency of >86% was achieved after the CF catalyst was recycled and reused for five cycles, which maintained about 97% efficiency of the initial cycle (i.e., ~89.1%), confirming the recyclability of CF for long-term The service life of a catalyst is an important technical indicator for evaluating its potential for practical usage. After the reaction, the catalyst in the solution could be easily and quickly separated due to its magnetism, as revealed in Figure 3c. Then, the recycled CF catalyst was rinsed with ethanol solution to remove the residual organics on the surface. Afterwards, it could be reused for RhB removal under the same conditions, as presented in Figure 3d. An excellent degradation efficiency of >86% was achieved after the CF catalyst was recycled and reused for five cycles, which maintained about 97% efficiency of the initial cycle (i.e.,~89.1%), confirming the recyclability of CF for long-term dye degradation in practical wastewater treatment. Figure 4a displays the optical absorption of samples. It is observed that the absorption of Co-TDC is far lower than CF in the visible-light band. The CF catalyst maintains a superior absorption in the range of 550-750 nm, suggesting its capability for a visible-lightdriven photocatalytic reaction. In addition, the threshold wavelengths of Co-TDC and CF are determined to be 619 nm and 685 nm, respectively. The corresponding bandgap and conduction band (CB)/valence band (VB) position can be calculated according to the following formulas [39][40][41]: where E g , λ g , E e , E CB , and E VB represent the bandgap, threshold wavelength, energy of free electrons on the hydrogen scale (~4.5 eV), and the CB and VB position, respectively. The values χ, n, and N represent the electronegativity of the constituent atom, number of species, and total number of atoms in the compound, respectively. The calculated E g of Co-TDC and CF are 1.72 eV and 1.57 eV, indicating that the CF hybrid possesses a narrower bandgap, and thus requires less excitation energy. Thereby, Figure 4b [42][43][44]. In order to further verify this perception, quenching experiments were carried out using tert-butyl alcohol (TBA), ammonium oxalate (AO) and L-ascorbic acid (L-AA) to quench the ·OH, photo-generated holes and ·O 2 − , respectively [45,46]. It can be observed from Figure S6 that the degradation efficiency of RhB clearly decreases in presence of TBA and AO. Therefore, it can be deduced that the main reactive species involved in the photocatalytic reaction are photo-generated holes and hydroxyl radicals (·OH), which consequently degrade RhB molecules to colorless small molecules.  [42][43][44]. In order to further verify this perception, quenching experiments were carried out using tert-butyl alcohol (TBA), ammonium oxalate (AO) and L-ascorbic acid (L-AA) to quench the ·OH, photo-generated holes and ·O2 − , respectively [45,46]. It can be observed from Figure S6 that the degradation efficiency of RhB clearly decreases in presence of TBA and AO. Therefore, it can be deduced that the main reactive species involved in the photocatalytic reaction are photo-generated holes and hydroxyl radicals (·OH), which consequently degrade RhB molecules to colorless small molecules. In order to evaluate the ecological toxicity of the RhB solution before and after treatment, Chlorella vulgaris (FACHB-8) was used as the model aquatic organism being tested, and the toxicity of the residual RhB after the photocatalytic reaction was assessed according to its growth inhibition rate to C. vulgaris. A detailed experimental method for algae density measurement is presented in the Supplementary Materials, which could be re- In order to evaluate the ecological toxicity of the RhB solution before and after treatment, Chlorella vulgaris (FACHB-8) was used as the model aquatic organism being tested, and the toxicity of the residual RhB after the photocatalytic reaction was assessed according to its growth inhibition rate to C. vulgaris. A detailed experimental method for algae density measurement is presented in the Supplementary Materials, which could be referred to as the standard GBT 21805-2008 [47]. As exhibited in Figure 5, the growth of C. vulgaris was significantly suppressed in the original RhB solution, and the inhibition rate doubled as time increases. In contrast, C. vulgaris could grow normally in the solution after reaction, and the remaining intermediate and final products showed a neglectable influence within 24 h. Even when the incubation time was extended to 96 h, the growth inhibition rate was still about 1%, which is only 15.6% of the original RhB solution. This demonstrates that the CF catalyst can effectively degrade and mineralize RhB molecules to nearly non-toxic products.

Conclusions
In summary, a highly active and stable Co3O4/Fe2O3 heterostructural photocatalyst was prepared by a facile MOF-templated method, with its structure, morphology and optical properties verified by XRD, XPS, SEM and UV-visible DRS methodology. The results indicate that the CF catalyst showed a strong visible-light absorption and high photocatalytic activity towards RhB degradation. By calculating the CB and VB position, it could be inferred that hydroxyl radicals and photo-generated holes were the dominant active species in the reaction. Furthermore, the 96 h growth inhibition rate of C. vulgaris by the treated RhB solution was 84.4% lower than the original solution, confirming the potential of the CF photocatalyst for the sunlight-driven long-term degradation of dye molecules into non-toxic and colorless ones.
Supplementary Materials: The followings are available online at www.mdpi.com/xxx/s1, Figure  S1: Fe 2p XPS spectrum of as-synthesized CF catalyst, Figure S2: SEM images of pure Co-TDC at different magnifications, Figure S3: SEM images of pure Fe2O3 at different magnifications, Figure  S4: Elemental mapping profiles of as-synthesized CF catalyst: (a) O Kα1, (b) Co Kα1, (c) Fe Kα1, Figure S5: PL spectra of CF catalyst comparing with pure Fe2O3 at excitation wavelength of 355 nm, Figure S6: Photocatalytic degradation efficiency of RhB with and without quenching agent, Figure  S7: Photocatalytic degradation efficiency of RhB by as-synthesized CF catalyst comparing with commercial Degussa/Evonik P25-TiO2 catalyst, Figure S8: Photocatalytic degradation efficiency of RhB by CF over a prolonged period of time; Experimental biota for toxicity test and algae density measurement.

Conclusions
In summary, a highly active and stable Co 3 O 4 /Fe 2 O 3 heterostructural photocatalyst was prepared by a facile MOF-templated method, with its structure, morphology and optical properties verified by XRD, XPS, SEM and UV-visible DRS methodology. The results indicate that the CF catalyst showed a strong visible-light absorption and high photocatalytic activity towards RhB degradation. By calculating the CB and VB position, it could be inferred that hydroxyl radicals and photo-generated holes were the dominant active species in the reaction. Furthermore, the 96 h growth inhibition rate of C. vulgaris by the treated RhB solution was 84.4% lower than the original solution, confirming the potential of the CF photocatalyst for the sunlight-driven long-term degradation of dye molecules into non-toxic and colorless ones.
Supplementary Materials: The followings are available online at https://www.mdpi.com/article/ 10.3390/nano12060904/s1, Figure S1: Fe 2p XPS spectrum of as-synthesized CF catalyst, Figure S2: SEM images of pure Co-TDC at different magnifications, Figure S3: SEM images of pure Fe 2 O 3 at different magnifications, Figure S4: Elemental mapping profiles of as-synthesized CF catalyst: (a) O Kα1, (b) Co Kα1, (c) Fe Kα1, Figure S5: PL spectra of CF catalyst comparing with pure Fe 2 O 3 at excitation wavelength of 355 nm, Figure S6: Photocatalytic degradation efficiency of RhB with and without quenching agent, Figure S7: Photocatalytic degradation efficiency of RhB by as-synthesized CF catalyst comparing with commercial Degussa/Evonik P25-TiO 2 catalyst, Figure S8: Photocatalytic degradation efficiency of RhB by CF over a prolonged period of time; Experimental biota for toxicity test and algae density measurement.

Data Availability Statement:
The data presented in this study are available on request from the corresponding author.