Fe 2 O 3 -Modiﬁed Porous BiVO 4 Nanoplates with Enhanced Photocatalytic Activity

As BiVO 4 is one of the most popular visible-light-responding photocatalysts, it has been widely used for visible-light-drivenwater splitting andenvironmentalpuriﬁcation.However,thetypicalphotocatalyticactivity ofunmodiﬁedBiVO 4 for the degradation of organic pollutants is still not impressive. To address this limitation, we studied Fe 2 O 3 -modiﬁed porous BiVO 4 nanoplates. Compared with unmodiﬁed BiVO 4 , the Fe 2 O 3 -modiﬁed porous BiVO 4 nanoplates showed signiﬁcantly enhanced photocatalytic activities in decomposing both dye and colorless pollutant models, such as rhodamine B (RhB) and phenol, respectively. The pseudo-ﬁrst-order reaction rate constants for the degradation of RhB and phenol on Fe 2 O 3 -modiﬁed BiVO 4 porous nanoplates are 27and 31times larger than thatofpristine BiVO 4 , respectively. Wealso found thatthe Fe 2 O 3 may act as an efﬁcient non-precious metal co-catalyst, which is responsible for the excellent photocatalytic activity of Fe 2 O 3 /BiVO 4 . Graphical Abstract Fe 2 O 3 , as a cheap and efﬁcient co-catalyst, could greatly enhance the photocatalytic activity of BiVO 4 porous nanoplates in decomposing organic pollutants.

However, until now, the photocatalytic activity of singlecomponent BiVO 4 is still not ideal yet for practical application. The rational design of composite photocatalysts could extend the spectral responsive range and promote the separation of photogenerated carriers and thus would improve photocatalytic activity dramatically compared to their host single-component materials [8,9]. Based on this strategy, we fabricated n-p core-shell BiVO 4 @Bi 2 O 3 and n-n Bi 2 S 3 /BiVO 4 composite microspheres previously [10,11]. Although these composite materials showed better photocatalytic activity than the pure BiVO 4 , their photocatalytic activity was still not impressive. The studies on photoelectrochemical water splitting of BiVO 4 as photoanodes have showed that BiVO 4 was poor catalyst for water oxidation. However, the appropriate modification of its surface with various oxygen evolution catalysts such as cobalt-phosphate, RhO 2 , and Pt could greatly improve its performance [12,13]. These findings suggest that rationally loading co-catalyst on the surface of BiVO 4 can improve its photocatalytic activity for the degradation of various organic pollutants. Recently, Li's group reported that the photocatalytic activity of BiVO 4 in oxidizing thiophene could be significantly enhanced through the modification with Pt and RuO 2 co-catalysts [14]. Nevertheless, Pt and Ru are noble metals and expensive reagents. Therefore, it would be highly desirable if we can achieve the enhanced photocatalytic activity of BiVO 4 using earth-abundant elements instead of the rare and precious ones [15]. Fe 2 O 3 consists of earth-abundant iron and oxygen elements and possesses the advantages of low cost, environmental friendliness and easy production, which has wide applications in many fields such as energy storage and conversion [16,17], catalysis [18], sensing and biomedicine [19,20]. As n-type semiconductor with a band gap of ca. 2.2 eV, Fe 2 O 3 is also a potential visible-light-driven photocatalyst [21]. Nevertheless, the photocatalytic activity of unmodified Fe 2 O 3 is poor because of its low carrier mobility, short minority carrier life time (ca. 10 ps) and diffusion length (ca. 2-4 nm) [22]. Therefore, there has been a lot of research focused on how to enhance photoelectrochemical and photocatalytic performance of Fe 2 O 3 [23,24]. On the other hand, the successful applications of Fe 2 O 3 in many important organic catalytic reactions suggested that it had good catalytic reaction activity [18,25]. However, to the best of our knowledge, Fe 2 O 3 as a role of co-catalyst in photocatalytic system has never been reported until now.
On the other hand, the fabrication of porous two-dimensional (2D) nanostructures has drawn much interest because the porous 2D structure of these materials can increase materials' surface area, facilitate the adsorption and diffusion of reactant molecules, and accelerate the transfer of photogenerated carriers from the interior to the surface of the material [26][27][28]. As a result, the photocatalysts with porous 2D nanostructures are expected to have good photocatalytic activity. Recently, Yu and coauthors have reported the synthesis of porous CuS/ZnS nanosheets with excellent photocatalytic activity through cation exchange reaction between performed inorganicorganic hybrid ZnS(en) 0.5 nanosheets and Cu 2? ions [29]. Similarly, nanoporous Cd x Zn 1-x S nanosheets have been prepared through the cation exchange reaction of ZnSamine nanosheets with cadmium ions [30]. However, the above-mentioned methods were limited in a few inorganicorganic hybrid 2D nanomaterials. Comparatively speaking, 2D metal complex nanostructures are more easily prepared through self-assembly of ligand at room temperature, which is driven by various non-covalent interactions including p-p stacking, van der Waals bonding, hydrophobic interactions, etc. [31][32][33]. Subsequently, 2D porous compounds nanostructures can be obtained by exchange reaction between ligands in 2D metal complexes nanostructures and anions in the desired compounds under certain conditions. However, this kind of ligand-anion exchange route to synthesize 2D porous nanostructured material has not, so far, been developed.
In this study, we present a new method for synthesizing porous BiVO 4 nanoplates via exchange reaction between ligand 1,4-benzenedicarboxylate (BDC) in 2D Bi 2 (BDC) 3 nanoplates derived from direct self-assembly and VO 4 3-anions under hydrothermal condition, schematically shown in Fig. 1. Subsequently, Fe 2 O 3 -modified porous BiVO 4 nanoplates were obtained through a simple hydrothermal deposition-precipitation route. The as-obtained Fe 2 O 3 / BiVO 4 products displayed excellent photocatalytic activities for degrading both dye and colorless organic pollutants. Based on UV-Visible diffuse-reflectance spectra (DRS), photoluminescence (PL) spectra, and a series of comparative experiments, Fe 2 O 3 as a role of non-precious metal co-catalyst in the present composite system was proposed.

Preparation of Porous BiVO 4 Nanoplates
Bi 2 (BDC) 3 nanoplates were prepared through our previous report [34]. For the preparation of porous BiVO 4 nanoplates, 0.1504 g of the as-obtained Bi 2 (BDC) 3 and 0.0732 g of NaVO 3 were dispersed into 40 mL of ultrapure water under stirring. Then the solution was put into a Teflon Ò lined stainless steel autoclave with 50 mL of capability and heated at 180°C for 10 h. After the autoclave was cooled to room temperature, the products were separated through centrifugation and washed three times with ultrapure water and absolute ethanol. Finally, the products were dried under vacuum at 60°C for 4 h.

Preparation of Fe 2 O 3 -Modified BiVO 4 Porous Nanoplates
In a typical procedure, 0.4 mmol of porous BiVO 4 nanoplates, 1.0 mmol of NaOH, and different amounts of Fe(NO 3 ) 3 (0.008, 0.02, 0.04 mmol) were added into 40 mL of ultrapure water in sequence under stirring. After that, the solution was put into a Teflon Ò lined stainless steel autoclave with 50 mL of capability and heated at 160°C for 12 h. After the autoclave was cooled to room temperature, the resultant products were separated via centrifugation and washed three times with ultrapure water and absolute ethanol, respectively. Finally, the products were dried under vacuum at 60°C for 4 h.

Characterizations
Powder X-ray diffraction (XRD) patterns were carried out on a Bruker D8 Advanced X-ray diffractometer using Cu Ka radiation (k = 0.15418 nm) at a scanning rate of 8°/min in the 2h range of 10°-70°. X-ray photoelectron spectroscopy (XPS) measurements were carried out with a Thermo ESCALAB 250 X-ray photoelectron spectrometer with an excitation source of Al Ka radiation (k = 1,253.6 eV). Field emission scanning electron microscopy (FE-SEM) images and energy dispersed X-ray (EDX) spectra were taken on a Nova NanoSEM 200 scanning electron microscope. Transmission electron microscopy (TEM), EDX, high-resolution TEM (HRTEM) images and mapping images were taken on a JEOL 2010 microscope, using an accelerating voltage of 200 kV. The UV-Visible DRS were recorded on a UV2450 (Shimadzu) using BaSO 4 as reference. The PL spectra were recorded on a Fluoromax-4 spectrofluorometer (HORIBA Jobin Yvon Inc.) equipped with a 150 W xenon lamp as the excitation source.

Photocatalytic Properties
The photocatalytic activity of Fe 2 O 3 /BiVO 4 nanoplates was evaluated by the degradation of RhB and phenol under visible-light irradiation from 500 W Xe light (CHF-XM500, purchased from Beijing Trusttech Co., Ltd) equipped with a 400 nm cutoff filter, and water splitting using a 300 W Xe lamp and optical cutoff filter (k [ 420 nm). In a typical process, 100 mg of photocatalysts was added to 100 mL of rhodamine B (RhB) solution (10 -5 mol L -1 ). Before illumination, the solution was magnetically stirred in the dark for 12 h to ensure an adsorption-desorption equilibrium between the photocatalysts and RhB. After that, the solution was exposed to visible light irradiation under magnetic stirring. At given time intervals, 3 mL of solution was sampled and centrifuged to remove the photocatalyst particles. Then, the filtrates were analyzed by recording variations of the absorption band maximum (553 nm) in the UV-Vis spectra of RhB by using a Shimadzu UV2501PC spectrophotometer. The studies of photocatalytic activities for other samples adopted the same measurement process. For photocatalytic degradation of phenol, the initial concentration of phenol solution was 1 9 10 -4 mol L -1 and kept other conditions unchanged. After visible light irradiation of different period time, the centrifugated solution was analyzed by recording variations of the absorption band maximum (270 nm) of phenol in the UV-Vis spectra by using a Shimadzu UV2450 PC spectrophotometer. As to photocatalytic water splitting, 0.1 g of photocatalysts was dispersed in 150 mL of 0.02 M aqueous AgNO 3 solution in a Pyrex reaction cell and thoroughly degassed by evacuation in order to drive off the air inside. The amount of evolved O 2 was measured by an online gas chromatograph.  composites under hydrothermal treatment, which is consistent with our previous report [10]. But when different amounts of Fe(NO 3 ) 3 (0.008, 0.02, 0.04 mmol) and 1 mmol of NaOH were simultaneously added into the reaction system, the resultant products contained no Bi 2 O 3 (Fig. 2b, c, d). The results showed that Fe 3? ions would preferentially react with OHions in solution. But, no diffraction peaks of Fe 2 O 3 could be observed because its content in the products is lower than the XRD limit of detection and/or poor crystalline degree. The existence of Fe 2 O 3 could be confirmed by subsequent XPS, EDX spectra, and HRTEM images.

Crystal Structure, Compositions and Morphology
The survey XPS spectrum reveals that the products are composed of Bi, V, O, Fe, and C elements (Supporting Information, Fig. S1). Figure 3a, b, c and d shows highresolution XPS spectra of Bi 4f, V 2p, Fe 2p, and O 1s of the products, respectively. In Fig. 3a, the peaks with binding energies of 159.2 and 164.4 eV correspond to Bi 4f7/2 and Bi 4f5/2 in BiVO 4 [11], respectively. The values (516.7 and 524.5 eV) of the binding energies for V 2p are in agreement with previous report on V 5? in BiVO 4 [35]. In the high-resolution Fe 2p spectrum (Fig. 3c), two distinct peaks located at 710.8 eV for Fe 2p3/2 and 724.2 eV for Fe 2p1/2 with a shake-up satellite at 719.3 eV were observed. This is characteristic of Fe 3? in Fe 2 O 3 [19]. The O 1s peak is asymmetric, indicating that different oxygen species were present in the near surface region. The major component is attributed to the lattice oxygen (529.8 eV) in BiVO 4 or Fe 2 O 3 . The second component centered at 532. 3 eV is associated with non-equivalent OH group on the surface [36]. Combined with the XRD results, it can be known that the products consist of BiVO 4 and possible Fe 2 O 3 .
We chose Bi 2 (BDC) 3 as precursor to synthesize BiVO 4 because the former was easily self-assembled into nanostructures through p-p stacking of benzene ring structure in BDC molecule at room temperature. Moreover, the flexible BDC ligand was helpful to alleviate the strain caused by crystal lattice mismatch between Bi 2 (BDC) 3 and BiVO 4 and maintain the framework of nanoplate during chemical conversion process. Figure 4a represents a typical FE-SEM image of the as-obtained Bi 2 (BDC) 3 precursor, which shows that the sample is nanoplates with average size of 100 nm in thickness, 500 nm in width, and 1.5 lm in length. After chemical transformation, Bi 2 (BDC) 3 nanoplates were converted into porous BiVO 4 nanoplates (Fig. 4b). These pores are linked together and distributed throughout the whole nanoplates. Magnified FE-SEM image shows that there are some tiny nanoparticles on the surfaces of the products obtained through hydrothermal deposition-precipitation route (Fig. S2). Further insight morphology and the microstructure of the products can be obtained from TEM and HRTEM images. In agreement with FE-SEM observations, TEM also shows that the products are porous nanoplates. Careful analyses show that there are some aggregates of tiny particles (marked with white arrowhead) closely integrating with the surfaces of nanoplates. Figure 4d represents HRTEM image recorded on the white rectangular area in Fig. 4c. The fringe spacing of 0.362 and 0.312 nm in the HRTEM image corresponds to the interplanar separation of (110) and (103) planes of monoclinic BiVO 4 , respectively. The HRTEM image of tiny particles displays the fringe spacing of 0.221 nm, which is consistent with the interplanar spacing of (113) plane of hexagonal-phase Fe 2 O 3 . The EDX spectroscopy recorded on a single nanoplate shows that the sample consists of Bi, V, Fe, Cu, C, and O elements (Fig. 4f) (Fig. S3). The quantitative calculation on the EDX spectrum recorded on SEM (Fig. S4) shows that the content of Fe 2 O 3 in the sample is ca. 3.8 mol %. All the above results indicate that the as-obtained products are Fe 2 O 3 -modified BiVO 4 porous nanoplates.

Photocatalytic Properties
The photocatalytic activities of the as-obtained Fe 2 O 3 / BiVO 4 porous nanoplates were evaluated by the degradation of RhB and phenol, common pollutant models as dye and colorless organic compounds, respectively, in water under visible light irradiation (k [ 400 nm). As shown in Fig. 5a, the photodegradation efficiency of RhB over Fe 2 O 3 /BiVO 4 is near 100 % within 80 min under visible light irradiation. However, RhB is stable and has a negligible degradation in the absence of any photocatalysts under visible light irradiation [37]. The photocatalytic activity of Fe 2 O 3 /BiVO 4 is significantly superior to pristine BiVO 4 and Bi 2 O 3 /BiVO 4 as well. Considering most of the photocatalytic reactions follow a pseudo-first-order kinetic equation: ln(C 0 /C) = kt (where C 0 and C are the equilibrium concentration of adsorption and the concentration of RhB at the irradiation time t, respectively, and k is the apparent rate constant), the plots of ln(C 0 /C) versus t were performed. The ln(C 0 /C) versus t displays a linear relationship (Fig. 5b), which shows that the photodegradation of RhB on these photocatalysts obeys the rules of firstorder reaction kinetics. From the fitting results, the apparent reaction rate constants are 0.0013, 0.0057, and 0.0358 min -1 for pristine BiVO 4 , Bi 2 O 3 /BiVO 4 , and Fe 2 O 3 /BiVO 4 , respectively. As a result, the photocatalytic reaction rate of Fe 2 O 3 /BiVO 4 is 27 and 6 times larger than that of pristine BiVO 4 and Bi 2 O 3 /BiVO 4 , respectively. Fe 2 O 3 /BiVO 4 also displayed excellent photocatalytic activity for degrading colorless phenol. As shown in Fig. 5c, the photocatalytic activity of BiVO 4 was greatly enhanced after the modification of Fe 2 O 3 , which was in good agreement with the results of photocatalytic decomposition of RhB. The apparent reaction rate constants are 0.0006, 0.0009, and 0.019 min -1 for BiVO 4 , Bi 2 O 3 / BiVO 4 , and Fe 2 O 3 /BiVO 4 , respectively. As a result, the photocatalytic reaction rate of Fe 2 O 3 /BiVO 4 is 31 and 21 times larger than that of unmodified BiVO 4 and Bi 2 O 3 / BiVO 4 , respectively. Whether the pollutant model is RhB or phenol, the amount of Fe 2 O 3 in the products has obvious influence on photocatalytic activity of Fe 2 O 3 / BiVO 4 . When too lower or too higher concentration of Fe(NO 3 ) 3 was used, less or more Fe 2 O 3 was produced. It was found that the photocatalytic activity of Fe 2 O 3 /BiVO 4 would decline in both cases. However, all the composites show higher photocatalytic activity than pure BiVO 4 . Therefore, finely adjusting the content of Fe 2 O 3 will further improve the photocatalytic activity of Fe 2 O 3 / BiVO 4 . In addition, Fe 2 O 3 -modified BiVO 4 porous nanoplates show good reusability (Fig. 6). After five times cycles for the photocatalytic degradation of RhB and phenol, the sample did not exhibit obvious loss of photocatalytic activity.

The Role of Fe 2 O 3 and Photocatalysis Mechanism
The doping of Fe 3? ions in the lattice of some photocatalysts or its graft on the surfaces of certain photocatalysts would enhance the photocatalytic activities of the photocatalysts [38,39]. In order to survey whether the enhancement of photocatalytic activity of BiVO 4 stems from the doping or grafting effect of Fe 3? ions, the control experiment without using NaOH was carried out. The photocatalytic activities of the resultant products have no obvious change for the decomposition of RhB (Fig. 5a) Fig. S5 and Fig. S6) [40]. As shown in Fig. 5a       The photocatalytic activity of the composite photocatalysts in decomposing organic pollutant is closely related to their adsorption capability, light energy utilization, separation efficiency of photogenerated carriers, and surface catalytic performance [42][43][44][45]. Fe 2 O 3 /BiVO 4 and pristine BiVO 4 have similar adsorption behavior to both RhB and phenol (Fig. 7a, b). The introduction of Fe 2 O 3 and Bi 2 O 3 do not noticeably improve adsorption ability of BiVO 4 to the organic pollutants (Fig. S10). Figure 7c represents UV-Vis DRS of Fe 2 O 3 /BiVO 4 and pristine BiVO 4 . There are no obvious differences between them except that Fe 2 O 3 /BiVO 4 displays slightly higher absorption in both UV and visible light region than BiVO 4 , which is possibly attributed to larger absorption coefficient of Fe 2 O 3 (ca. 10 5 cm -1 ) than that of BiVO 4 and its narrower band gap (2.2 eV) [46]. As we know, PL spectra can be used as an efficient means to evaluate the separation efficiency of the photogenerated carriers [11]. Figure 7d displays the room temperature PL emission spectra (excited at 400 nm) of pure BiVO 4  It is known that noble metals can act as co-catalysts on the surfaces of photocatalysts, which can greatly enhance the surface catalytic activity of the materials. Recent studies showed that some non-precious metal such as MoS 2 , graphene, and cobalt phosphate could serve as cocatalysts like noble metal as well [47][48][49]. In this study, the loading of small amounts of Fe 2 O 3 also enhanced surface catalytic activity of BiVO 4 . To further support the role of a co-catalyst of Fe 2 O 3 in our system, we compared the photocatalytic activity of Fe 2 O 3 /BiVO 4 and Co 3 O 4 /BiVO 4 in degrading phenol. Co 3 O 4 /BiVO 4 was prepared following reported procedures [50], where Co 3 O 4 /BiVO 4 possessed excellent photocatalytic activity for degrading phenol. As shown in Fig. 8, the photocatalytic activity of the present Fe 2 O 3 /BiVO 4 greatly surpassed that of Co 3 O 4 /BiVO 4 . The photocatalytic reaction rate constant of the former (1.9 9 10 -2 min -1 ) is six times larger than that of the latter (3 9 10 -3 min -1 ). Compared with Fe 2 O 3 /BiVO 4 , Co 3 O 4 /BiVO 4 shows much larger and wider absorption in both UV and visible light region, as shown in Fig. 7c. PL spectrum indicates that the emission intensity of BiVO 4 was greatly reduced after the loading of Co 3 O 4 (Fig. S11), which showed that photogenerated carriers were efficiently separated in Co 3 O 4 /BiVO 4 system as well. In addition, Co 3 O 4 /BiVO 4 and Fe 2 O 3 /BiVO 4 has similar adsorption ability to phenol (Fig. S12). Considering all above facts, it can be understood that higher surface catalytic activity of Fe 2 O 3 /BiVO 4 lead to its higher catalytic activity than Co 3 O 4 /BiVO 4 . Specifically, Fe 2 O 3 played a key role as cocatalyst for enhanced photocatalytic activity of Fe 2 O 3 / BiVO 4 due to its low content, efficient separation of photogenerated carriers and good surface catalytic activity. It should be highlighted here that the photocatalytic activity of the as-obtained Fe 2 O 3 /BiVO 4 is among one of the best photocatalysts reported to date, and is superior to noble metal Pt loaded the same porous BiVO 4 nanoplates (Only 40 % of phenol was degraded under the same conditions, Fig. S13).

Conclusions
In conclusion, Fe 2 O 3 /BiVO 4 nanoplates with porous structures have been prepared through a mild chemical conversion and subsequently hydrothermal depositionprecipitation routes. The as-prepared Fe 2 O 3 /BiVO 4 showed excellent photocatalytic activity in degrading both RhB and phenol. After the modification of Fe 2 O 3 , the photocatalytic activity of pristine BiVO 4 could be increased more than one order of magnitude. It is believed that Fe 2 O 3 acts as an The facile preparation method, low cost of raw materials, excellent photocatalytic activity, and good reusability of the Fe 2 O 3 /BiVO 4 porous nanoplates make the material a promising photocatalyst for the application in the field of environmental remedy.