Facile Formation of Bi2O2CO3/Bi2MoO6 Nanosheets for Visible Light-Driven Photocatalysis.

Bi2O2CO3/Bi2MoO6 heterojunction catalysts were prepared by treating Bi2MoO6 sheets with aqueous NaHCO3 solutions at room temperature. All the Bi2O2CO3/Bi2MoO6 heterojunctions exhibited higher activities than pristine Bi2MoO6 in the photocatalytic degradation of rhodamine B (RhB), methyl orange, and ciprofloxacin under visible-light irradiation, and the most active photocatalyst was found to be the one with a C/Bi molar ratio of ∼1/2.3. Relevant samples were characterized by X-ray diffraction, X-ray photoelectron spectroscopy, scanning electron microscopy, transmission electron microscopy, N2 adsorption-desorption, Fourier transform infrared spectroscopy, and UV-vis spectroscopy. The higher activity of Bi2O2CO3/Bi2MoO6 than pristine Bi2MoO6 is explained by the enhanced separation and transfer of photogenerated electron/hole pairs, as verified by transient photocurrent densities, photoluminescence spectroscopy, and electrochemical impedance spectroscopy. Photogenerated holes (h+) and superoxide radical anions (•O2 -) were found to be the main active species. The good reusability of Bi2O2CO3/Bi2MoO6 was testified by cycling degradation of RhB and tetracycline hydrochloride.

Introduction

class="Chemical">TiO2, a beclass="Chemical">nchmark photocatalyst, works efficieclass="Chemical">ntly uclass="Chemical">nder UV light that merely accouclass="Chemical">nts for a very small fractioclass="Chemical">n of suclass="Chemical">nlight. Seekiclass="Chemical">ng class="Chemical">new catalysts that caclass="Chemical">n make full use of suclass="Chemical">nlight, especially the visible light portioclass="Chemical">n, is a meaclass="Chemical">niclass="Chemical">ngful topic. class="Chemical">n class="Chemical">Bi2MoO6, a nontoxic and visible light-responsive semiconductor, is promising for making photocatalysts.[1−4] Bi2MoO6 photocatalysts with different morphologies or structures (e.g., nanoparticles,[5] nanotubes,[6] nanosheets,[7] flowerlike structures,[8,9] and hollow microspheres[10,11]) have been developed. However, the recombination of photogenerated carriers is still a serious problem when using pristine Bi2MoO6 as a photocatalyst.[12] It is of great interest to develop Bi2MoO6-based modified photocatalysts with enhanced photocatalytic performance.

class="Chemical">Bi2MoO6-based modified photocatalysts caclass="Chemical">n be prepared by depositiclass="Chemical">ng metals (e.g., class="Chemical">n class="Chemical">Pt,[13] Pd,[14] and Ag[15]) onto Bi2MoO6, doping Bi2MoO6 with metal ions (e.g., Er3+,[16] Gd3+,[17] Ho3+,[17] and Yb3+[17]), and integrating Bi2MoO6 with other substances such as TiO2,[18] RGO,[19] g-C3N4,[20] MoS2,[21] Ag/AgCl,[22] Ag2MoO4,[23,24] Ag3VO4,[25] Ag2O,[26−28] Ag–Ag2CO3,[29] Ag2CO3,[30] MO (M = Cu, Co, Ni),[31] Fe2O3,[32] ZnFe2O4,[33] Bi2S3,[34] BiOX (X = Cl, Br, I),[35−37] Bi2O2CO3,[38−40] BiVO4,[41] Ta3N5,[42] Bi4MoO9,[43] Bi2Mo2O9,[44] Bi3.64Mo0.36O6.55,[45,46] Bi4V2O11,[47] and Bi2Mo3O12.[48] The numerous references cited above indicate that research along this direction is popular. However, Bi2O2CO3/Bi2MoO6 photocatalysts have been seldom reported,[38−40] although Bi2O2CO3 has been combined with other substances (e.g., Cu2O,[49] Co3O4,[50] ZnFe2O4,[51] Bi,[52] Au–Bi2O3,[53] Bi2O4,[54,55] Bi4O7,[56] Bi2S3,[57] BiOCl,[58,59] BiOBr,[60] BiOI,[61] BiVO4,[62] Bi2WO6,[63] Ag2O,[64] Ag2CO3,[65] AgI,[66] carbon quantum dots,[67] and porphyrins[68]) to form heterojunction photocatalysts.[69] In general, the formation of heterojunction systems is an efficient way to design visible light-responsive photocatalysts.[70−72]

As for the class="Chemical">Bi2O2CO3/class="Chemical">n class="Chemical">Bi2MoO6 heterojunction system, Xu et al. prepared Bi2O2CO3/Bi2MoO6 by subjecting a mixture containing BiOCl, MoO42–, and g-C3N4 to hydrothermal treatment at 180 °C.[38] The formed sesame-biscuit-like Bi2O2CO3/Bi2MoO6 heterostructures showed enhanced performance in photocatalytic degradation of rhodamine B (RhB). Zhang and co-workers prepared microsphere-like Bi2O2CO3/Bi2MoO6 heterojunctions composed of nanoplatelets of Bi2O2CO3 and Bi2MoO6 via a template-free solvothermal process.[39] The obtained samples showed high photocatalytic activity in the degradation of RhB. Wang and co-workers prepared a meshlike Bi2O2CO3/Bi2MoO6 composite composed of many nanoparticles via a hydrothermal method using urea to tune the morphology.[40] They again demonstrated the superior photocatalytic activity of Bi2O2CO3/Bi2MoO6 than Bi2MoO6 in the degradation of RhB. A quaternary heterostructured Ag–Bi2O2CO3/Bi3.64Mo0.36O6.55/Bi2MoO6 composite was also synthesized and tested in visible light-driven photocatalysis.[73]

In these relevant papers,[38−40] the morphologies of the class="Chemical">Bi2O2CO3/class="Chemical">n class="Chemical">Bi2MoO6 materials are very interesting, but there is still room for developing Bi2O2CO3/Bi2MoO6 materials with different morphologies via facile preparation methods, demonstrating the advantages of the materials in different reactions and elucidating the reaction mechanisms.

Herein, class="Chemical">Bi2O2CO3/class="Chemical">n class="Chemical">Bi2MoO6 nanosheets were prepared via a facile route, that is, by reacting Bi2MoO6 with aqueous NaHCO3 at room temperature. These materials were characterized and tested in the photocatalytic degradation of industrial dyes [RhB and methyl orange (MO)], ciprofloxacin (CIP), and tetracycline hydrochloride (TC) under visible-light irradiation. Reasons for the enhanced photocatalytic activity were investigated. A possible photodegradation mechanism based on Bi2O2CO3/Bi2MoO6 was proposed. Even though Bi2O2CO3/Bi2MoO6 has been reported by other groups,[38−40] the current work differs from these previous studies, in terms of the preparation method, morphology of the heterojunctions, and scope of photocatalytic reactions. At the end of the paper, we will provide more discussions on what is learned from this work and what can be done in the future.

Results and Discussion

Basic Characterization of Samples

X-ray diffraction (XRD) was used to characterize class="Chemical">Bi2MoO6 aclass="Chemical">nd class="Chemical">n class="Chemical">Bi2O2CO3/Bi2MoO6 (S1–S5). As seen in Figure , the obtained pristine Bi2MoO6 is orthorhombic Bi2MoO6 (JCPDS no. 21-0102). Bi2O2CO3/Bi2MoO6 (S1–S5) samples mainly exhibit orthorhombic Bi2MoO6 peaks. The strongest (013) peak (2θ = 30.3°) of tetragonal Bi2O2CO3 (JCPDS no. 41-1488) appears for S3, S4, and S5 with theoretical C/Bi molar ratio of 1/2.3, 1/1.8, and 1/1.5, respectively (Figure ). This peak enhances when going from S3 to S5 because of the addition of more NaHCO3 during the synthesis. No other impurity peaks can be observed. The presence of residual NaHCO3 can be excluded because the Bi2O2CO3/Bi2MoO6 samples were thoroughly washed by water and our later X-ray photoelectron spectroscopy (XPS) characterization excludes the presence of residual Na.

Figure 1

XRD patterns of Bi2MoO6 and Bi2O2CO3/Bi2MoO6 (S1–S5), and standard XRD patterns of orthorhombic Bi2MoO6 (JCPDS no. 21-0102) and tetragonal Bi2O2CO3 (JCPDS no. 41-1488).

XRD patterns of class="Chemical">Bi2MoO6 aclass="Chemical">nd class="Chemical">n class="Chemical">Bi2O2CO3/Bi2MoO6 (S1–S5), and standard XRD patterns of orthorhombic Bi2MoO6 (JCPDS no. 21-0102) and tetragonal Bi2O2CO3 (JCPDS no. 41-1488).

According to scanning electron miclass="Chemical">croscopy (SEM) characterizatioclass="Chemical">n, pure class="Chemical">n class="Chemical">Bi2MoO6 manly consists of irregular nanosheets (Figure A). Bi2O2CO3/Bi2MoO6 (S1–S5) samples contain Bi2MoO6 nanosheets with slightly smaller sizes (Figures B and S1) probably because of the vigorous stirring (1200 rpm) of the system for 4 h, after mixing Bi2MoO6 nanosheets with NaHCO3 solution.

Figure 2

SEM images of (A) Bi2MoO6 and (B) S3 (Bi2O2CO3/Bi2MoO6).

SEM images of (A) class="Chemical">Bi2MoO6 aclass="Chemical">nd (B) S3 (class="Chemical">n class="Chemical">Bi2O2CO3/Bi2MoO6).

Transmission electron miclass="Chemical">croscopy (TEM) aclass="Chemical">nd high-resolutioclass="Chemical">n TEM (HRTEM) were used to characterize class="Chemical">n class="Chemical">Bi2MoO6 and typical Bi2O2CO3/Bi2MoO6 samples (S3 and S5). These samples show a 2D nanosheet structure with irregular sizes (Figure ), consistent with the SEM results (Figures and S1). Figure B shows an HRTEM image of Bi2MoO6. The lattice spacing is determined to be 0.313 nm, corresponding to the (131) plane of orthorhombic Bi2MoO6 (JCPDS no. 21-0102). The HRTEM image of S3 (Figure D) also shows the lattice stripe belonging to the (131) plane of orthorhombic Bi2MoO6. The surface is covered by a thin layer, most likely Bi2O2CO3 because the corresponding XRD pattern shows a peak corresponding to Bi2O2CO3 (Figure ). Figure S2 enables us to take a closer look at the interface. For S5, the HRTEM image shows the presence of tetragonal Bi2O2CO3 nanoparticles on the Bi2MoO6 surface (Figure F).

Figure 3

TEM and HRTEM images of Bi2MoO6 (A,B), S3 (C,D), and S5 (E,F).

TEM and HRTEM images of n class="Chemical">Bi2MoO6 (A,B), S3 (C,D), aclass="Chemical">nd S5 (E,F).

The elemental distribution and composition in S3 were analyzed by scanning TEM (STEM) mapping and TEM–energy-dispersive X-ray (EDX). The STEM image confirms the nanosheet structure of S3 (Figure A). class="Chemical">Bi, Mo, O, aclass="Chemical">nd C elemeclass="Chemical">nts distribute homogeclass="Chemical">neously (Figure B–E). The measured molar ratio of C/class="Chemical">n class="Chemical">Bi is 1/2.4 (Figure F), consistent with the theoretical value (1/2.3). Additionally, EDX element mapping and EDX data from S5 confirmed the existence of the related elements in Bi2O2CO3/Bi2MoO6 (Figure S3). The measured molar ratio of C/Bi is 1/1.99 (Figure S3F), whereas the theoretical value is 1/1.5. Note that Bi2O2CO3/Bi2MoO6 samples denoted as S1–S5 were prepared by adding a certain amount (3, 5, 7, 9, or 11 mL) of NaHCO3 solution (0.1 mol/L) to a suspension containing 0.5 g of Bi2MoO6 (0.82 mmol) and 50 mL of water under vigorous stirring (1200 rpm), followed by continuous stirring for 4 h. Thus, the EDX data indicate that NaHCO3 in the solution can not be completely consumed via reacting it with Bi2MoO6 when it is overdosed.

Figure 4

(A) STEM, (B–E) EDX elemental mapping, and (F) EDX data images of S3.

Figure presents Fourier transform infrared (FTIR) spectra of class="Chemical">Bi2MoO6 (red liclass="Chemical">ne) aclass="Chemical">nd S3 (blue liclass="Chemical">ne). class="Chemical">n class="Chemical">Bi2MoO6 shows absorption bands at 400–900 cm–1.[7] The bands at 843.0 and 796.3 cm–1 can be assigned to the asymmetric and symmetric stretching modes of MoO6, respectively, involving vibrations of the apical oxygen atoms.[74] The band at 731.2 cm–1 is attributed to the asymmetric stretching mode of MoO6 involving vibrations of the equatorial oxygen atoms, and the band at 573.4 cm–1 corresponds to the bending vibration of MoO6.[75] The band at 442.0 cm–1 is attributed to the stretching and bending vibrations of octahedral BiO6.[76] Two bands at 1628.6 and 1384.7 cm–1 are ascribed to the stretching and bending vibrations of adsorbed water, respectively. These absorption bands exist in both samples. Besides, S3 also shows a peak at 1478.4 cm–1 because of the internal vibration of CO32– for Bi2O2CO3.[77] In addition, compared to S3, pristine Bi2MoO6 shows a clear peak located at 934.0 cm–1, which can be ascribed to C–H bonds; this is because Bi2MoO6 was prepared in the presence of organic solvent (i.e., ethanol and ethylene glycol).[34]

Figure 5

FTIR spectra of Bi2MoO6 and S3 (Bi2O2CO3/Bi2MoO6).

FTIR spectra of class="Chemical">Bi2MoO6 aclass="Chemical">nd S3 (class="Chemical">n class="Chemical">Bi2O2CO3/Bi2MoO6).

The surface composition and chemical status of class="Chemical">Bi2MoO6 aclass="Chemical">nd S3 were probed by XPS. Iclass="Chemical">n the survey spectra for class="Chemical">n class="Chemical">Bi2MoO6 and S3 (Figure S4), characteristic C, O, Bi, and Mo peaks are detected; no impurities (such as Na) are detected. The data indicate that no NaHCO3 is left behind after thorough washing during the synthesis stage. Next, selected elements were picked for precise XPS scanning. For the C 1s orbit scan of S3 (Figure A), three peaks at 284.6, 286.1, and 288.5 eV (attributed to the C–C, C–O, and O–C=O groups, respectively) are found, indicating the existence of adventitious carbon and CO32– in S3.[78] S3 has apparently more carbonate species than Bi2MoO6.

Figure 6

High-resolution XPS spectra of (A) C 1s, (B) O 1s, (C) Bi 4f, and (D) Mo 3d from Bi2MoO6 and S3 (Bi2O2CO3/Bi2MoO6).

High-resolution XPS spectra of (A) C 1s, (B) O 1s, (C) class="Chemical">Bi 4f, aclass="Chemical">nd (D) Mo 3d from class="Chemical">n class="Chemical">Bi2MoO6 and S3 (Bi2O2CO3/Bi2MoO6).

The O 1s peaks are deconvoluted (Figure B). class="Chemical">Bi2MoO6 shows two decoclass="Chemical">nvoluted O 1s peaks assigclass="Chemical">ned to class="Chemical">n class="Chemical">Bi–O (529.7 eV) and Mo–O (531.8 eV). However, for S3, the carbonate peak (530.4 eV) increases greatly at the expense of the Mo–O peak (531.8 eV), whereas the Bi–O (529.5 eV) peak is still obvious. The data are consistent with the TEM observation for S3, that is, the surface of Bi2MoO6 is covered by a Bi2O2CO3 layer (Figure D). Figure C shows Bi 4f5/2 and Bi 4f7/2 peaks at 164.1 and 158.8 eV, respectively, confirming the presence of Bi3+ in S3.[11] From Figure D, two peaks at 232.0 and 235.1 eV can be attributed to Mo 3d5/2 and Mo 3d3/2, respectively, confirming the presence of Mo6+ in S3.[11] The positions of Bi 4f, Mo 3d, O 1s, and C 1s peaks of S3 are identical to the corresponding positions of Bi2MoO6.

From the class="Chemical">N2 sorclass="Chemical">n class="Chemical">ption isotherms of Bi2MoO6 and S3 (Figure S5A), Brunauer–Emmett–Teller (BET) surface areas of Bi2MoO6 and S3 are determined to be 7.1 and 7.8 m2/g, respectively. The pore volumes of Bi2MoO6 and S3 are 0.0089 and 0.0097 m2/g, respectively (Figure S5B). The low surface areas and small pore volumes of these materials are consistent with the bulk morphology of the materials (Figures and 3).

Photocatalytic Performance

Photocatalytic degradation of class="Chemical">RhB (50 mL, 10 mg/L) was studied. As showclass="Chemical">n iclass="Chemical">n Figure A, after 30 miclass="Chemical">n iclass="Chemical">n the dark aclass="Chemical">nd 150 miclass="Chemical">n uclass="Chemical">nder visible-light irradiatioclass="Chemical">n, class="Chemical">n class="Chemical">RhB does not degrade when no catalyst exists. The photodegradation efficiencies clearly show the sequence S1 < S2 < S3 > S4 > S5. All the S1–S5 samples are more active than Bi2MoO6. It is noted that the rigorous stirring during the treatment of Bi2MoO6 by an aqueous NaHCO3 solution may make the Bi2MoO6 sheets smaller (Figures  and S1). Thus, a control experiment was carried out, in which Bi2MoO6 was rigorously stirred in water for 4 h, dried, and subjected to photocatalytic testing. The sample (Bi2MoO6-S) adsorbs less RhB than Bi2MoO6 during the dark adsorption stage, and the adsorption amount is the same of those of S1–S5 samples. Thus, the activity trend mentioned above can be confirmed. We also repeated the catalytic experiments by using RhB (50 mL, 2.5 mg/L) with a concentration lower than before (10 mg/L). In this case, the reaction takes less time, and a similar activity trend can be observed (Figure S6).

Figure 7

Degradation curves of (A) RhB (50 mL, 10 mg/L), (B) MO (50 mL, 10 mg/L), and (C) CIP (50 mL, 10 mg/L) with using Bi2MoO6 or Bi2O2CO3/Bi2MoO6 (S1–S5) as a catalyst (30 mg).

Degradation class="Chemical">curves of (A) class="Chemical">n class="Chemical">RhB (50 mL, 10 mg/L), (B) MO (50 mL, 10 mg/L), and (C) CIP (50 mL, 10 mg/L) with using Bi2MoO6 or Bi2O2CO3/Bi2MoO6 (S1–S5) as a catalyst (30 mg).

S1–S5 samclass="Chemical">ples are still more active thaclass="Chemical">n class="Chemical">n class="Chemical">Bi2MoO6 in the photocatalytic degradation of MO (50 mL, 10 mg/L) and CIP (50 mL, 10 mg/L), as shown in Figure B,C. The activities of S1–S5 still show the sequence S1< S2 < S3 > S4 > S5. We additionally found that S3 is effective for the degradation of TC (50 mL, 5 mg/L, Figure S7).

The recyclaclass="Chemical">bility of S3 was tested. To miclass="Chemical">nimize the testiclass="Chemical">ng time, low-coclass="Chemical">nceclass="Chemical">ntratioclass="Chemical">n solutioclass="Chemical">ns of class="Chemical">n class="Chemical">RhB (50 mL, 2.5 mg/L) and TC (50 mL, 5 mg/L) were used. In each run, 30 min dark adsorption and 60 min reaction were involved. As shown in Figure S8, S3 can be reused without any loss of activity. By comparing the XRD patterns (Figure S9) and SEM images (Figure S10) of S3 before and after five runs, we can clearly see that S3 remains the stable phase composition and morphology/structure.

Reasons for the Enhanced Activity

The question now arises why S3 is more active than class="Chemical">Bi2MoO6 iclass="Chemical">n visible light-driveclass="Chemical">n photocatalysis. Oclass="Chemical">ne reasoclass="Chemical">n may be traced to its visible-light absorclass="Chemical">n class="Chemical">ption. However, our UV–vis–diffuse reflectance measurement (Figure S11) data indicate that the addition of Bi2O2CO3 to Bi2MoO6 (S1–S5) does not lead to enhanced visible-light absorption (λ > 400 nm). Thus, that effect is excluded as a reason. The difference in BET surface areas (Bi2MoO6: 7.1 m2/g; S3: 7.8 m2/g) is also small (Figure S5), whereas the difference in photocatalytic activity is obvious (Figure ).

The separation and migration of photogenerated carriers usually class="Chemical">play aclass="Chemical">n importaclass="Chemical">nt role iclass="Chemical">n determiclass="Chemical">niclass="Chemical">ng the activity of a photocatalyst.[79] The photoclass="Chemical">n class="Chemical">current response tests of Bi2MoO6 and S3 were conducted in the dark and under visible light in 0.1 M Na2SO4 (Figure ). Both Bi2MoO6 and S3 generated transient photocurrent when the light was on, and the photocurrent decreased sharply when the light was off. After four intermittent on–off irradiation cycles, the photocurrent intensity kept steady. S3 showed a remarkably enhanced photocurrent density compared to Bi2MoO6, demonstrating the effective separation of photoinduced electron/hole pairs in Bi2O2CO3/Bi2MoO6 upon exposure to visible light.

Figure 8

Transient photocurrent densities of Bi2MoO6 and S3.

Transient photoclass="Chemical">curreclass="Chemical">nt declass="Chemical">nsities of class="Chemical">n class="Chemical">Bi2MoO6 and S3.

To gain insight into the electron-transport and recomclass="Chemical">biclass="Chemical">natioclass="Chemical">n properties of the samclass="Chemical">n class="Chemical">ples, EIS analyses were performed. The smaller radius of the EIS Nyquist plot demonstrates the lower impedance and a higher efficiency of charge transfer.[80] As shown in Figure A, S3 shows a much smaller arc radius than Bi2MoO6, implying that the separation efficiency of photogenerated charge carriers is enhanced for Bi2O2CO3/Bi2MoO6, thus extending the lifetime of photoexcited electron–hole pairs.

Figure 9

(A) Electrochemical impedance spectroscopy (EIS) Nyquist plots and (B) photoluminescence (PL) spectra of Bi2MoO6 and S3.

(A) Electrochemical impedance spectroscopy (EIS) Nyquist class="Chemical">plots aclass="Chemical">nd (B) photolumiclass="Chemical">nesceclass="Chemical">nce (class="Chemical">n class="Chemical">PL) spectra of Bi2MoO6 and S3.

class="Chemical">PL spectra were emclass="Chemical">n class="Chemical">ployed to confirm the enhanced separation efficiency of photogenerated charge carriers. A lower PL emission intensity usually means that a semiconductor has a stronger ability to facilitate the separation and migration of photogenerated carriers.[81] From Figure B, both Bi2MoO6 and S3 show a strong emission peak centered at around 580 nm. S3 shows a lower peak intensity than Bi2MoO6, meaning that S3 (Bi2O2CO3/Bi2MoO6) can facilitate the separation and migration of photoexcited electron/hole pairs more efficiently.

class="Chemical">Radical-caclass="Chemical">n class="Chemical">pture experiments involving corresponding scavengers were performed to monitor the active radicals involved in the photoreaction (Figure ). For the reaction system involving Bi2MoO6 (Figure A) or S3 (Figure B), the addition of ammonium oxalate (1 mmol, scavenger of h+) could inhibit RhB degradation obviously. Adding benzoquinone (0.02 mmol, scavenger of •O2– species) also led to the decrease of RhB degradation obviously. The existence of isopropyl alcohol (1 mmol, scavenger of •OH species) almost has no effect on RhB degradation. The data imply that the main active species in photocatalytic degradation of RhB are h+ and •O2–, whereas •OH is not relevant.

Figure 10

Effects of scavengers on RhB degradation in presence of (A) Bi2MoO6 or (B) S3.

Effects of scavengers on n class="Chemical">RhB degradatioclass="Chemical">n iclass="Chemical">n preseclass="Chemical">nce of (A) class="Chemical">n class="Chemical">Bi2MoO6 or (B) S3.

An ESR spin-trap with the class="Chemical">DMPO techclass="Chemical">nique was emclass="Chemical">n class="Chemical">ployed to confirm the results of radical-capture experiments. Bi2MoO6 does not generate DMPO ESR spin-trapping signals of •O2– and •OH in the dark (Figure A,B). With light irradiation, Bi2MoO6 leads to the formation of •O2– but no •OH. Similar results are obtained when testing S3 (Figure C–D). However, S3 can lead to the generation of more •O2–, as seen from stronger DMPO ESR spin-trapping signals. This can be a reason for the enhanced activity of Bi2O2CO3/Bi2MoO6.

Figure 11

5,5-Dimethyl-pyrroline N-oxide (DMPO) electron spin resonance (ESR) spin-trapping signals of (A,B) Bi2MoO6 and (C,D) S3 (Bi2O2CO3/Bi2MoO6) for •O2– and •OH.

class="Chemical">5,5-Dimethyl-pyrroline N-oxide (class="Chemical">n class="Chemical">DMPO) electron spin resonance (ESR) spin-trapping signals of (A,B) Bi2MoO6 and (C,D) S3 (Bi2O2CO3/Bi2MoO6) for •O2– and •OH.

In reference to the energy band positions of class="Chemical">Bi2MoO6 (class="Chemical">n class="Chemical">ECB = −0.37 V, EVB = 2.27 V)[29] and Bi2O2CO3 (ECB = 0.16 V, EVB = 3.56 V),[49] we propose a possible photocatalytic mechanism for Bi2O2CO3/Bi2MoO6 (Figure ). Although Bi2O2CO3 cannot be excited by visible light because of its wide band gap (Eg = ∼3.4 eV),[49,82] Bi2MoO6 can be excited by visible light,[83] thus producing photogenerated electrons (e–) and holes (h+). Moreover, because the ECB of Bi2O2CO3 (0.16 V) is more positive than that of Bi2MoO6 (−0.37 V), the photogenerated electrons will be able to flow to the conduction band of Bi2O2CO3. Thus, the photogenerated e–/h+ pairs in the heterojunction system can be efficiently separated, which prolongs the lifetime of active species existing in the reaction system. This explanation is consistent with our mechanistic experiments (Figures –11). It should be mentioned that Bi2O2CO3 cannot be excited by visible light because of its wide band gap (Eg = ∼3.4 eV),[49,82] and it has poor photocatalytic activity under visible light.[50,56,84,85] Thus, the presence of more Bi2O2CO3 (exceeding the optimal content in S3) in the Bi2O2CO3/Bi2MoO6 heterojunction catalyst (for the cases of S4 and S5) may lead to decreased photocatalytic activity compared with that associated with S3.

Figure 12

Possible photocatalytic degradation mechanism based on Bi2O2CO3/Bi2MoO6.

Possible photocatalytic degradation mechanism based on n class="Chemical">Bi2O2CO3/class="Chemical">n class="Chemical">Bi2MoO6.

Discussion

It should be mentioned that the objective of this work was not to develop the most active photocatalyst in the world. Numerous new catalysts with different levels of activities have been reported. It is interesting to develop catalysts with new compositions, structures/morphologies, functions, or preparation methods and to understand the correlations among preparation methods/details, structures/morphologies, and catalytic performance.[86,87]

To put the class="Chemical">curreclass="Chemical">nt work iclass="Chemical">n perspective, we compare the activity of the most active S3 catalyst iclass="Chemical">n this work with those of class="Chemical">n class="Chemical">Ag2O/Bi2MoO6,[26] Ag2MoO4/Bi2MoO6,[23] and Ag–Ag2CO3/Bi2MoO6[29] studied under identical reaction conditions. As seen in Figure S12, the activity of the samples follows the sequence Ag2MoO4/Bi2MoO6 ∼ Ag–Ag2CO3/Bi2MoO6 ∼ Ag2O/Bi2MoO6 > Bi2O2CO3/Bi2MoO6.

No attemclass="Chemical">pt was made to compare the activity of our class="Chemical">n class="Chemical">Bi2O2CO3/Bi2MoO6 catalyst with those of Bi2O2CO3/Bi2MoO6 catalysts reported in the literature because the reaction conditions are all different.[38−40] Nevertheless, the trend seen in our study is in line with the trend seen in others’ work,[38−40] in the sense that the optimal Bi2O2CO3/Bi2MoO6 catalyst is more active than Bi2MoO6. Further research may be conducted to study the influence of the morphologies and sizes of Bi2MoO6 and the preparation details (e.g., synthesis and calcination temperatures) on the catalytic performance of Bi2O2CO3/Bi2MoO6 catalysts. In fact, the Bi2MoO6 supports used for the preparation of Ag2O/Bi2MoO6 and Ag2MoO4/Bi2MoO6 mentioned above are flowerlike microspheres,[23,26] whereas the Bi2MoO6 supports used for the preparation of Ag–Ag2CO3/Bi2MoO6[29] and Bi2O2CO3/Bi2MoO6 in this study are sheets. A recent study by others shows that Bi2MoO6 microspheres (instead of sheets) can also be used to make efficient Ag2CO3/Bi2MoO6 photocatalysts.[30] Thus, the morphology effect may be a subtle effect that needs more carful studies in the future.

This work also has important imclass="Chemical">plicatioclass="Chemical">ns to the syclass="Chemical">nthesis of class="Chemical">n class="Chemical">Bi2MoO6-based heterojunction catalysts. In some cases, NaHCO3 was used as a precipitation agent for the formation of Ag2CO3/Bi2MoO6 catalysts.[29] Our current work indicates that the reaction between Bi2MoO6 and aqueous NaHCO3 to form surface Bi2O2CO3 can happen even at room temperature. Thus, this side reaction should be considered when synthesizing Ag2CO3/Bi2MoO6 catalysts. Care should be taken to control the synthesis procedure and the amount of NaHCO3 to avoid the formation of Bi2O2CO3/Bi2MoO6 unless the formation of Ag2CO3–Bi2O2CO3/Bi2MoO6 is needed. In our previous work,[29] AgNO3 solution was first mixed with Bi2MoO6 powders to allow for the adsorption of Ag+ onto Bi2MoO6 powders. A stoichiometric amount of NaHCO3 solution was then added into the system to allow for complete reaction between Ag+ and NaHCO3 to form Ag2CO3 on Bi2MoO6. The reaction between Ag+ and aqueous NaHCO3 is supposed to be much faster than the conversion of solid-phase Bi2MoO6 to Bi2O2CO3/Bi2MoO6. Thus, under the right synthesis conditions, the unintentional formation of Bi2O2CO3 can be avoided.

Conclusions

class="Chemical">Bi2MoO6 class="Chemical">naclass="Chemical">nosheets modified with class="Chemical">n class="Chemical">Bi2O2CO3 were prepared by chemical deposition. The enhanced photocatalytic activities of Bi2O2CO3/Bi2MoO6 were identified in degrading RhB under visible light, and the optimal C/Bi theoretical molar ratio was determined to be ∼1/2.3. The optimal Bi2O2CO3/Bi2MoO6 catalyst (S3) was also found to be more active than Bi2MoO6 in degrading MO, CIP, or TC by visible-light photocatalysis. The cycling degradation of RhB and TC together with relevant XRD and SEM characterization proved the stability of Bi2MoO6/Bi2MoO6 in catalytic reactions. Transient photocurrent densities, PL, and EIS data demonstrated that S3 can separate and transfer photogenerated electrons and holes more efficiently than Bi2MoO6. Photogenerated holes (h+) and superoxide radical anions (•O2–) were confirmed as the main active species by radical-capture experiments, whereas •OH is not relevant in the reaction. ESR data indicated that S3 can generate more •O2– species than Bi2MoO6 upon visible-light irradiation, whereas •OH is not generated for both S3 and Bi2MoO6. This work provides a simple and effective way for developing Bi2MoO6-based photocatalysts that can work under visible light.

Experimental Section

Materials

class="Chemical">Bi(NO3)3·5H2O, Na2MoO4·2H2O, class="Chemical">n class="Chemical">HNO3, NaHCO3, RhB, MO, and CIP of analytical grade were purchased from Sinopharm Chemical Reagent. TC was purchased from Aladdin. All reagents were used as received.

Synthesis of Bi2O2CO3/Bi2MoO6

Typically, 0.5 g of class="Chemical">Bi2MoO6 (0.82 mmol) prepared accordiclass="Chemical">ng to our previous report[29] was dispersed iclass="Chemical">n 50 mL deioclass="Chemical">nized class="Chemical">n class="Chemical">water with the aid of ultrasonic treatment for 15 min. A certain amount (3, 5, 7, 9, or 11 mL) of NaHCO3 solution (0.1 mol/L) was dropped therein under vigorous stirring (1200 rpm). The mixture was further stirred continuously for 4 h. Bi2O2CO3/Bi2MoO6 samples (denoted as S1–S5) were collected by removing the supernatant, washed with deionized water three times, and dried at 60 °C for 24 h. The theoretical C/Bi molar ratios of S1–S5 samples, by simply assuming that NaHCO3 reacts with Bi2MoO6 completely, are 1/5.5, 1/3.3, 1/2.3, 1/1.8, and 1/1.5, respectively.

Characterization

XRD data were obtained on an MSAL XD2 X-ray diffractometer with class="Chemical">Cu Kα radiatioclass="Chemical">n at 40 kV aclass="Chemical">nd 30 mA at a scaclass="Chemical">nclass="Chemical">niclass="Chemical">ng speed of 8° miclass="Chemical">n–1. SEM experimeclass="Chemical">nts were coclass="Chemical">nducted oclass="Chemical">n a Shimadzu SUPERSCAN SSX-550 field emissioclass="Chemical">n scaclass="Chemical">nclass="Chemical">niclass="Chemical">ng electroclass="Chemical">n miclass="Chemical">n class="Chemical">croscope. TEM experiments were performed using a JEOL JEM-2100F high-resolution transmission electron microscope. FTIR spectra were recorded with a Thermo Fisher Nexus 670 FTIR spectrometer using a standard KBr disk method. XPS data were obtained on a multifunctional photoelectron spectroscopy instrument (Axis Ultra DLD, Kratos). Optical diffuse reflectance spectra were recorded on a UV–vis–near-infrared spectroscopy scanning spectrophotometer (LAMBDA 35, PerkinElmer) with an integrating sphere accessory. PL experiments were conducted using a LabRAM HR Evolution instrument (HORIBA JY Company, France) with an excitation wavelength of 355 nm. EIS data were obtained by an electrochemical analyzer (CHI 660B Chenhua Instrument Company). Electron paramagnetic resonance signals of paramagnetic species spin-trapped with DMPO were recorded on a Bruker ESR 300E spectrometer.

Evaluation of Photocatalytic Activity

Photocatalytic degradation of class="Chemical">RhB (10 mg/L), MO (10 mg/L), class="Chemical">n class="Chemical">CIP (10 mg/L), and TC (5 mg/L) was conducted using a Xe lamp (300 W, HSX-F300, Beijing NeT Technology Co., Ltd.) coupled with a UV-cutoff filter (420 nm) as the light source.[23,26,29,88−96] The distance between the lamp and the top of the beaker mouth was ca. 15 cm. The temperature of the reaction system was controlled at around 25 °C.

Before visible-light illumination, 30 mg of the catalyst was suspended in 50 mL of solution in a 100 mL beaker, and the suspension was stirred (800 rpm) for 30 min. Subsequently, visible-light irradiation was conducted for 60 min for each experiment. 4 mL of the suspension was samclass="Chemical">pled every 15 miclass="Chemical">n aclass="Chemical">nd separated by ceclass="Chemical">ntrifugatioclass="Chemical">n at a speed of 8000 rpm for 15 miclass="Chemical">n. The superclass="Chemical">nataclass="Chemical">nt was theclass="Chemical">n aclass="Chemical">nalyzed by usiclass="Chemical">ng a UV-5200PC spectrometer. Cycliclass="Chemical">ng degradatioclass="Chemical">n experimeclass="Chemical">nts iclass="Chemical">nvolviclass="Chemical">ng class="Chemical">n class="Chemical">RhB and TC were conducted according to the procedure reported before.[29]