Recently, photocatalysts as a kind of favorable and green media for environmental remediation have drawn much attention due to the important roles that they play. Among these photocatalysts, bismuth-based semiconductors are considered to be environmental friendly photocatalysts to purify pollutants. Some photocatalysts, such as Bi2O3 [1, 2], BiPO4 [3, 4], Bi2WO6 [5, 6], BiOM (M = Cl, Br or I) [7-9] and BiVO4 [10, 11], etc., are widely applied to degrade dyes, phenol, other organic pollutants, and even photocatalytic removal of NO [12-15].
BiOCl, as a p-type and novel ternary semiconductor, has a unique layer structure, whose internal [Bi2O2]2+ layers interleaves double Cl- slabs [16, 17]. This interaction structure induces the formation of a strong internal electric field, leading to promoting the transformation of photo-generated carriers and effectively suppressing their recombination [18-20]. Moreover, the Bi 6s, O 2p and Cl 2p orbitals hybridize and its conduction band (CB) was consisted of Bi 6s and Bi 6p orbitals. According to existing reports, the large dispersed Bi 6s orbital is beneficial to strengthening the mobility of photo-induced carries [7, 21].
The large band gap energy of BiOCl (~ 3.4 eV) merely adsorbs UV light, limiting the application of BiOCl. That seeking new ways to modify photocatalytic performance of BiOCl is one of main directions of photocatalytic research. Li et al. [22] synthesized ultrathin BiOCl nanosheet using ethanol and water as solvents and Polyvinylpyrrolidone as surfactant. Mi et al. prepared ultrathin nanosheet photocatalyst of Fe(Ⅲ) modified BiOCl with exposure of active (001) facet [23]. These ultrathin BiOCl nanosheets decrease the transferring path length of carriers from interior to surface and prolong the life of charge carries, leading to enhancing the absorption for vis-light and photocatalytic activity of photocatalysts. Both BiOCl/Ag2CO3 hybrids [24] and synthesized BiOCl/BiPO4 via electrochemical anodization method [25] improved photocatalytic activity of BiOCl. However, these photocatalysts have some shortcomings, for instance, high cost of raw materials, complicated operation, low efficiency and poor adaptability to environment and so on.
In addition, forming heterojunctions between different semiconductors has been certified to be a useful method to promote photocatalytic activity of BiOCl. For instance, BiOCl/SnO2 [26], BiOCl/Bi2S3 [27], Bi2O2CO3/BiOCl [28] and g-C3N4/BiOCl [29] performed good photocatalytic ability in degrading organic pollutants. Both closely linked interface in heterojunctions and well-matched band structure can effectively inhibit the recombination of photo-generated carriers and speed up photocatalytic degradation for organic pollutants [30-32]. However, photocatalytic efficiency of BiOCl and charge transfer rate of related heterojunctions need to be further enhanced.
Nearly two years, a novel photocatalyst of monoclinic dibismuth tetroxide (m-Bi2O4) responding to visible light was constructed successfully [32-34]. The excellent property of m-Bi2O4 and m-Bi2O4-based compounds immediately caught the researchers' interest and achieved some good results. For example, Bi2O4/Fe3O4 hybrid nanocomposites were applied to degrade ibuprofen and Z-scheme g-C3N4/m-Bi2O4 catalyst was utilized to inactivate E. coli K-12 [35, 36]. Heterostructure Bi2O2CO3/Bi2O4 photocatalyst was used to decompose dyes and phenol [37].
The combination of photocatalysts with semiconductors is a trend to greatly heighten the photocatalytic performance [38]. Furthermore, ion-etching chemical method has indicated to be an efficient technology to synthetize semiconductor photocatalysts, such as bismuth-based photocatalysts [38, 39], etc. Based on existing literature, no researches of m-Bi2O4/BiOCl compound are reported.
Given the superiorities and researches of BiOCl and m-Bi2O4, we synthetized m-Bi2O4/BiOCl photocatalyst by an ion-etching method. The degradation of methyl orange and antibiotic of tetracycline were examined under vis-light to characterize capability of novel m-Bi2O4/BiOCl photocatalyst. The aim is to enhance the photocatalytic performance of BiOCl and enlarge its absorbance range to sun light. Furthermore, the degradation mechanism of m-Bi2O4/BiOCl photocatalyst was also proposed.
Bismuth nitrate pentahydrate (Bi(NO3)3∙5H2O), methyl orange (MO) and isopropanol (IPA) were gained from Macklin Reagents (shanghai) Co., Ltd. Sodium bismuthate hydrate (NaBiO3∙2H2O), 4-hydroxy-2, 2, 6, 6-tetramethylpiperidinyloxy (TEMPOL), sodium oxalate (Na2C2O4), hydrochloric acid (HCl, 36 wt%–38 wt%) and tetracycline were obtained from Aladdin Reagents (Shanghai). The above reagents were all analytical grade.
The m-Bi2O4 sample was prepared via hydrothermal method based on reported literature [40]. Typically, certain amount of NaBiO3∙2H2O was distributed in 90 mL ultrapure water with sonication for 0.5 h. It was vigorously stirred for 1 h at room temperature, then was putted into a 150-mL autoclave and heated to 433 K for 12 h. The resulted products were gathered through centrifuging and washing. The m-Bi2O4 sample was dried at 333 K for 12 h under vacuum.
Ion-etching method to synthesize m-Bi2O4/BiOCl complex was based on previous literature [41, 42]. The synthetic route followed as: 1 mmol/L m-Bi2O4 was dispersed in 10 mL absolute ethanol with ultrasound. Some volume of diluted HCl solution (1 mol/L) was added into suspension dropwise with magnetic stirring. Then the above solution was stirred continuously for 3 hours and ultrasonically treated for 1 hour. Then the product was dried at 333 K for 20 h in a vacuum oven after filtration and washing.
In this experiment, the relative ratio of m-Bi2O4 and BiOCl were controlled by adjusting the volume of diluted HCl solution. Thus, the various products of m-Bi2O4/BiOCl were denoted as BiOCl-1, BiOCl-1.5 and BiOCl-2, respectively, based on the volumes of used HCl solution of 1, 1.5 and 2 mL. Pure BiOCl was synthetized by adding excessive HCl. Fig. 1 was the schematic diagram of prepared m-Bi2O4/BiOCl photocatalyst.
The structure properties of as-prepared photocatalysts were checked by powder X-ray diffraction (XRD, Bruker D8 Advance diffractometer, USA). Transmission electron microscopy (TEM, JEM-2100, Japan) and field emission scanning electron microscopy (FE-SEM, Phenom, USA) equipped with energy dispersive X-ray spectroscopy detector (EDX) were used to detect the morphology and elemental content.
The valence states and composition were conducted via X-ray photoelectron spectroscopy (XPS, Escalab 250Xi, USA). Specific surface area (BET) was measured through N2 adsorption-desorption instrument (Micromeritic TriStarII 3020, USA). Fourier transform infrared spectrometer (FT-IR, Thermo, Nicolet-6700, USA) was applied to detect the functional groups. UV-vis spectrum was obtained via an UV-vis-NIR Spectrophotometer (Shimadzu UV-3600, Japan). Photocurrent test was executed by using a CHI-660E electrochemical station equipped with a standard three-electrode.
MO dye and tetracycline solution were selected to investigate the capability of above photocatalysts. The visible light source came from a Xenon lamp (CEL-HXF-300W), which equipped with a cut-off filter to ensure wavelength larger than 420 nm and was placed at 14 cm from liquid surface. The experimental device was showed in Fig. S1 (Supporting Information). The process of photocatalytic degradation was made as following: 50 mg samples were poured into 100 mL solution containing MO (10 mg/L) and tetracycline (30 mg/L). Before irradiation, the solution was magnetically stirred in the dark for 0.5 hour to achieve adsorption-desorption equilibrium.
During the process of vis-light irradiation, reaction solution was continually withdrawn approximately 5 mL at given time interval. Then it was centrifuged to dislodge photocatalysts and the remaining solution was analyzed via an UV-vis-NIR Spectrophotometer. The characteristic peaks for MO and tetracycline solution were at 464 and 357 nm, separately. The calculated degradation ratio (η, %) used the following formula:
here, C0 (mg/L) and Ct (mg/L) were the equilibrium concentration and instantaneous concentration of pollutants at any time t (min), respectively.
To gain insight probable mechanism of photocatalytic degradation, the diverse scavengers were applied to competitive experiments for the purpose of investigating the role of active radicals during the photocatalytic process. Generally, the active radicals involve hole (h+), superoxide radical (•O2-) and hydroxyl radical (•OH) [43]. In the test, IPA, Na2C2O4 and TEMPOL were utilized as scavengers for •OH, h+ and •O2-, separately. MO solution was selected as pollutants and other conditions were the same as the degradation experiments.
Fig. 2 was the XRD results of m-Bi2O4, BiOCl and m-Bi2O4/BiOCl composites with different ratios and the three-dimensional map was displayed in Fig. S2 (Supporting Information). Pure m-Bi2O4 exhibited identical diffraction peaks corresponding to the monoclinic phase of Bi2O4 (JCPDS 83-0410, lattice constants: a = 12.367 Å, b = 5.118 Å and c = 5.567 Å) [33, 36]. With the increasing addition of HCl solution, new diffraction peaks appeared, indicating that new materials were produced. The peak intensities of 2θ of 26.86° and 29.52° indexed to (111) and (–311) planes were decreased. Besides, when the amount of HCl solution was excessive, the two peaks disappeared.
Simultaneously, comparing with the standard maps, the characteristic diffraction peaks indexed perfectly to the tetragonal phase of BiOCl (JCPDS 06-0249, lattice parameters: a = b = 3.891 Å, c = 7.369 Å) [44, 45]. These peaks were detected at 2θ of 11.97°, 24.16°, 25.92° and 33.41°, corresponding to the (001), (002), (101) and (102) crystal planes [46, 47]. The intensities of these peaks gradually became stronger when the amount of added HCl solution was varied from 1 mL to overdose. Moreover, no impure peaks were discovered, verifying high purity of products. The method of relative intensity ratio (RIR) was applied to quantitatively calculate the composition of multi-phase sample. When the sample and corundum (α-Al2O3) are mixed in 50:50 by weight, the intensity ratio of its strongest peaks is defined as RIR value for sample phase [48]. When the two phases of a and b constitute one sample, their weight ratio can be calculated for each phase in compounds using the following formula:
where, W and RIR represent the weight ratios and RIR values of phase a and phase b, respectively; I is the integrated intensity of strongest peak of phase a or phase b, respectively [49]. For m-Bi2O4 and BiOCl phase, the RIR values are 11.82 and 9.65 according to PDF database, respectively. The calculated results of as-prepared samples were list in Table 1.
SEM measure was employed to examine the surface morphology of obtained catalysts. In the Fig. 3(a), the m-Bi2O4 was square shape and had a rough surface. For BiOCl (Fig. 3(b)), its nanoplate stacked layer by layer and the thickness reached 125.8 nm. The increasing of nanosheet thickness of BiOCl led to a much long moving path for charge carriers [22]. This corresponded with the lowest photocatalytic activity in the experiment described later.
For m-Bi2O4/BiOCl, with the increasing amount of HCl solution, the morphology of the complex changed, as displayed in Fig. 3(c)–(e). When 1 mL of HCl solution was added, BiOCl-1 sample began to appear nanosheet structure with a thickness of only 34 nm. The more amount of HCl solution was added, the more obvious the nanosheet was observed. The BiOCl-2 sample (Fig. 3(e)) completely become nanosheets and aggregated forming intimate contact.
EDX spectrum for BiOCl-1 was shown in Fig. 3(f) and the inserted picture was the corresponding part. It existed three elements of Bi, O and Cl. TEM and HR-TEM were presented in Fig. 3(g) and (h). The lattice fringes of 0.204 and 0.331 nm were corresponded to lattice planes of (221) and (111) of m-Bi2O4 and lattice spacing of 0.152 nm was ascribed to (114) plane of BiOCl. Fig. S3 was the selected-area electron diffraction (SAED) and its regular highlights indicated a single crystalline structure of BiOCl-1 sample.
FT-IR measurement was conducted to further identify the chemical composition and bonding of composites. Fig. 4 presented the characteristic peaks of the samples and the broad band at 3448 cm–1 was caused by the stretching and deformation vibration of O–H group in physisorbed and/or chemisorbed H2O [24]. This peak could be seen in every sample owing to the water sorption from air.
The range of absorption peaks of m-Bi2O4 were between 500 and 1000 cm–1. The bands at 553 and 593 cm–1 were designated to stretching vibrations of Bi–O bond [50]. Upon treatment with HCl solution, the intensities of the peaks diminished. When m-Bi2O4 was completely converted to BiOCl, the location of Bi–O bond stretching vibrations shifted to 531 cm–1, which was valent symmetrical vibrations of Bi–O in BiOCl [51]. Based on the FT-IR and XRD results, it can be concluded that the compounds consisted of m-Bi2O4 and BiOCl.
Fig. 5 showed the structure information of m-Bi2O4/BiOCl. The Raman spectra of m-Bi2O4 had two bands at 122 and 308 cm–1. BiOCl has a tetragonal P4/mmm space group and its vibrational modes can be depicted as:
here, Ag, Bg, and Eg are Raman actives and others are infrared actives [52, 53].
BiOCl sample had two obvious peaks and one weak peak. Symmetric vibrations rather than asymmetric vibrations increased Raman bands. Therefore, the strong peak at 142 cm–1 was ascribed to A1g internal Bi-Cl stretching vibration and 197 cm–1 was assigned to Eg internal Bi-Cl stretching vibration. The not readily noticeable and weak band at 395 cm–1 corresponded to Eg and B1g band produced from oxygen motion atoms [54, 55].
BiOCl-1 displayed three main characteristic bands at 123, 146 and 316 cm–1. The peak at 197 cm–1 disappeared, suggesting that Cl– replaced some O2– in the lattice to form Bi–Cl bonds. It can be drawn a conclusion that the m-Bi2O4/BiOCl complex was successfully synthetized.
Fig. 6 was the isotherms of three samples, and from Fig. 6(a) and (b), it can be known that the average pore sizes of m-Bi2O4 and BiOCl were 52.01 and 13.99 nm, respectively. Besides, the isotherm of BiOCl-1 sample can be classified to type-Ⅳ (Fig. 6(c)) and the shape of hysteresis loop was type H3 at relative high pressure, indicating a mesoporous structure (2–50 nm) [56, 57]. The hysteresis loop shifted at P/P0 ≈ 1.0, suggesting that BiOCl-1 sample had a macroporous structure (large than 50 nm).
The BET specific surface areas of m-Bi2O4, BiOCl and BiOCl-1 were 17.87, 237.05 and 112.90 m2/g, respectively. The pore size were examined through BJH model from desorption branch of nitrogen adsorption-desorption isotherm. The average pore diameter was 18.96 nm for BiOCl-1, corresponding to the isotherm showed in the insert of Fig. 6(c). The small mesopores might be stem from the pores within the nanosheets, but large pores were owing to the existence of the space between the interlaced nanosheets [58, 59].
Fig. 7 revealed the valence states and chemical composition of BiOCl-1. The survey spectra (Fig. 7(a)) stated clearly that BiOCl-1 sample consisted of only Bi, O and Cl elements, excluding the adventitious carbon species from XPS measurement, whose bending energy was at 284.6 eV. The two high-resolution peaks of Bi 4f were corresponding to the positions of Bi 4f5/2 and Bi 4f7/2 (Fig. 7(b)). The Bi 4f5/2 can be resolved into one bimodal peak at 164.59 and 163.76 eV. And the Bi 4f7/2 peak was resolved into other bimodal peak at 159.35 and 158.6 eV, which should be attributed to Bi (Ⅲ) and Bi(Ⅴ) on the basis of previous researches [40, 60].
The banding energies of Cl 2p at 197.95 and 199.55 eV (Fig. 7(c)) were ascribed to the positions of Cl 2p3/2 and Cl 2p1/2 [61]. The peak of O 1s can be decomposed into two peaks at 529.84 and 530.86 eV (Fig. 7(d)), indicating that the two distinguishable models of O existed simultaneously in BiOCl-1. Between the two peaks of O 1s, one with low binding energy was the lattice oxygen and the other was attributed to the chemisorbed and physically adsorbed oxygen on the surface of catalysts [28, 51]. Above all, the m-Bi2O4/BiOCl photocatalysts were prepared successfully.
The optical performance of photocatalysts was displayed in Fig. 8. For pure m-Bi2O4, the absorption threshold wavelength was about 640 nm. In contrast, the BiOCl absorption cutoff wavelength was approximately 380 nm, demonstrating that the BiOCl sample only absorbed UV light. With the proportion increasing of m-Bi2O4 in the composite, the absorption edge redshifted monotonically and the absorption ability to visible light was enhanced. The colors of photocatalysts were changed with the increasing of used HCl solution. Initially, the color of m-Bi2O4 was dark brown, and then became yellow after adding HCl solution. Finally, BiOCl became white powder. The color change was as vivid as shown in Fig. 8(a).
The photocatalytic activity was deeply influenced by the band gap energy (Ebg) of semiconductor. The value of Ebg is calculated through following equation according to DRS data:
here, α, h, ν, and A are all constants and can be found in other literatures. According to previous research, n values of m-Bi2O4 and BiOCl were all 4 for indirect transition. Ebg were estimated through the plot of (αhν) 1/2 vs. hν that were extrapolated to (αhν)1/2 = 0 as illustrated in Fig. 8(b). The values of Ebg (Fig. 8(c)) were estimated approximately to 3.2, 2.0 and 1.87 eV for BiOCl, m-Bi2O4 and BiOCl-1, respectively.
The path of migration and redox capability of photo-generated carriers were significantly affected by the location of band edge potentials. Generally speaking, the more positive of valence band (VB) position is, the stronger oxidation capability of holes becomes. On the contrary, the CB position is more negative, causing stronger reduction power of electrons. The potentials of VB and CB in semiconductors were determined by applying the electronegativity theory based on following empirical equations:
here, X is the Mulliken electronegativity. Ee is free electron energy (~ 4.5 eV vs. NHE) [62]. The X value of BiOCl was 6.36 [27, 63] and the VB values were 3.46 and 1.63 eV for BiOCl and m-Bi2O4. Correspondingly, their CB values were 0.26 and -0.37 eV, separately.
The phenomenon of photocatalytic degradation was investigated by degrading MO and tetracycline under visible light. And the results were presented in Fig. 9(a) and (b). Blank experiment (without catalysts) showed that the performance of MO and tetracycline was stable and almost no degradation was occurred during the process of photolysis.
From Fig. 9(a) and (b), it can be seen that pure BiOCl photocatalyst had neglected degradation performance for MO. Although the concentration of tetracycline was decreased 30% by BiOCl, it might be attributed to the selective adsorption due to the largest surface area of BiOCl. And m-Bi2O4 had about 80% degradation for MO or tetracycline, owing to its narrow band gap.
Through adding HCl solution to acidify m-Bi2O4, the activity of m-Bi2O4/BiOCl got obvious enhancement, and efficiency was also improved compared with BiOCl. The degradation efficiencies of BiOCl-1 reached 95% for MO within 10 min and 85.5% for tetracycline within 150 min. The results indicated that the BiOCl-1 had the best degradation performance.
Nevertheless, when the ratio of HCl and m-Bi2O4 was 2:1, the effect was reduced, which might since the BiOCl accounted for too much proportion in the complex and led to bad consequence.
The intensities of absorption spectra of MO and tetracycline diminished with time got by as displayed in Fig. 9(c) and (d). And the color of MO solution changed from yellow to colorless, as vividly inserted in the graph of Fig. 9(c).
To quantitatively describe the photocatalytic ability, the degradation kinetics followed equation:
here, k (min-1) is kinetic constant specified by the plot versus t (min). The fitted results were shown in Fig. S4 and meanwhile listed in Table S1.
All of the k values for MO and tetracycline were positive, as shown in Table S1. As the picture of Fig. S4(a) and (b) showed, all samples were fitted well and had very high correlation coefficient (R2 > 0.90). Fig. S4(c) and (d) exhibited k values by histogram. The reaction rate constants of BiOCl-1, BiOCl-1.5 and BiOCl-2 were far greater than that of m-Bi2O4 and BiOCl, especially BiOCl-1 had the biggest rate constant, which was 52.28 times as much as BiOCl.
Compared with pure BiOCl, the composite photocatalyst of m-Bi2O4/BiOCl showed excellent photocatalytic degradation performance for MO and tetracycline. The excellent photocatalytic capability of m-Bi2O4/BiOCl might be ascribed to p-n heterogeneous structure, large BET values and favorable band gap energies.
In fact, good photocatalytic effects do not indicate complete degradation for organic pollutants, for some intermediate products can be produced [36]. So, to further versify the degradation of organic pollutants over as-prepared photocatalysts, the values of total organic carbon (TOC) were determined and the results are shown in Fig. 10.
From the data in Fig. 10, it can be seen that BiOCl had very low degradation efficiencies for MO and tetracycline under visible light. However, after combined with Bi2O4, m-Bi2O4/BiOCl (BiOCl-1) had much high photocatalytic effect and the degradation efficiencies for MO and tetracycline reached 96.6% and 93.9%, respectively, which indicated a high removal efficiencies of TOC to the two pollutants.
The above results also stated that there were a large number of intermediate products being produced during the process of photocatalysis. And these intermediate products were completely decomposed into H2O and CO2 within a given time.
The photocatalyst of BiOCl-1 after reaction was filtered and washed with ethanol and water for several times. Then it was put into an oven to dry. The XRD and FT-IR technologies were employed to investigate the stability of the as-prepared photocatalysts before and after the reaction. It is vividly seen from the XRD pattern (Fig. 11(a)) that the diffraction peaks have not be changed and no impure peaks appeared, suggesting a stable crystal structure. Moreover, the vibration peaks of photocatalyst BiOCl-1 after reaction (Fig. 11(b)) were the same as those before reaction, demonstrating an excellent chemical structure. These results indicated that the synthetic photocatalyst of BiOCl-1 had a good stability.
To explore the mechanism and value the roles of reactive oxygen species (ROS) in photocatalysis, the selective addition of scavengers' analysis was implemented. IPA was used as scavenger for •OH, Na2C2O4 for h+ and TEMPOL for •O2-. The amount of these scavengers was 1 mmol/L. As showed in Fig. 12, when adding IPA, the degradation efficiency of MO was little lower than no scavenger, revealing that •OH played a part of role during the process of degradation. The activity was inhibited obviously with addition of TEMPOL, which demonstrated that •O2- was also important for photocatalytic process [64, 65]. Besides, the addition of Na2C2O4 thoroughly suppressed the degradation performance for MO, suggesting that h+ played a more critical role.
The contribution of •O2- and h+ to the degradation of MO were 39.3% and 90.9%, respectively. The kinetic fit and data were showed in Fig. S5 and Table S2. Those data fitted well and the k values for no scavenger, IPA, TEMPOL and Na2C2O4 were 29.02 x 10-2, 17.32 × 10-2, 7.61 × 10-2 and 0.31 × 10-2 min-1, respectively. The additive scavengers suggested that the dominant active species were •O2- and h+ during the process of photocatalytic degradation. They could directly or indirectly oxidize MO to non-toxic substances.
Separated and migratory efficiency of photo-generated carriers as chief factors, greatly influence the photocatalytic performance of photocatalysts. The transient photocurrents were measured by photocurrent response experiments. Generally, the stronger intensity of current signal is, the higher separated efficiency of photo-generated carriers will be.
As revealed in Fig. 13, the photocurrents responded rapidly via light on-off cycles under visible light. BiOCl failed to produce photo-induce carriers under visible light, therefore the current signal was weak. The main reason was that its large band gap needed higher energy to be excited. The m-Bi2O4 sample with decent photocurrent density was not enough high. However, BiOCl-1 sample demonstrated the highest photocurrent density, corresponding to the best photocatalytic activity as revealed by the test.
In the view of results analysis, a possible photocatalytic mechanism for m-Bi2O4/BiOCl was put forward in Fig. 14. The type of semiconductor for BiOCl and m-Bi2O4 were p-type and n-type respectively. Thence the electronic concentration in BiOCl was lower than that in m-Bi2O4. Moreover, the Fermi level (Ef) of m-Bi2O4 was close to the CB and it lied closely to the VB for BiOCl. Before contacting, the CB and VB of BiOCl were lower than those of m-Bi2O4.
Once the two semiconductors contacted, the band alignment between the two semiconductors led to establishing a p-n heterojunction. Charge redistribution occurred and the electrons transferred from m-Bi2O4 to BiOCl, resulting in a negative region in the vicinal region of BiOCl in the heterojunction, while the holes transferred from BiOCl to m-Bi2O4, creating a positive region in the m-Bi2O4 nearby the junction. The Fermi level of m-Bi2O4 down-shifted and that of BiOCl moved up until the Ef equilibrated [25, 66, 67]. Simultaneously, the energy band position of BiOCl shifted upward whereas m-Bi2O4 shifted downward accompanying with the Fermi level.
Consequently, the band structures bending in the opposite direction took place in the charge region, forming a corresponding internal electric field (Einternal). Upon the semiconductors were excited, the internal electric field accelerated the separation of carries and drove it to different directions. The sites of oxidation and reduction were spatially separated [11, 68]. Thence, the photo-excited electrons and holes transferred across the interface and internal electric field, extending the mobile path and prolonging the electronic lifetime.
According to above analysis, when the m-Bi2O4/BiOCl was irradiated by visible light, m-Bi2O4 was motivated to produce electron-hole pairs and then the holes were left on VB since the electrons were transferred from VB to CB. However, BiOCl could not response because of large band gap. Through charge redistribution, the CB and VB of BiOCl were higher than that of m-Bi2O4. The photo-generated holes were streamed to VB of BiOCl under the internal electric field force. These holes directly captured organics and then oxidized them.
The electrons on CB of m-Bi2O4 reacted with surface adsorbed oxygen [O2]ad. and reduced them to •O2- species, which could reacted with pollutants. Some •O2- also further reacted with e- and H+ to generate •OH, following the formulas below:
Therefore, the redox reaction carried out separately. This p-n heterojunction mode enhanced separation of electron-hole pairs, prolonged lifetime of carries and facilitated photocatalytic degradation, causing an excellent photocatalytic activity over m-Bi2O4/BiOCl composite.
The heterostructure m-Bi2O4/BiOCl photocatalyst was successfully synthetized through an ion-etching method. The ratio of m-Bi2O4 to BiOCl was controlled by adjusting the volume of HCl solution. The experimental results certificated that as-prepared photocatalysts displayed high degradation rate and complete mineralization ability for MO and tetracycline solution under visible light. The degradation efficiencies for MO and tetracycline reached 96.6% and 93.9%, respectively. Moreover, the as-prepared photocatalysts showed good stability.
The p-n heterojunction separated carries effectively and enhanced photocatalytic activity. Furthermore, the reactive oxygen species trapping experiment revealed that h+ and •O2- were primary radicals responsible for photocatalysis. The study proposed a strategy to design a heterostructure photocatalyst for treating water environmental issues.
2018,
Vol. 39 Issue (11): 1792-1803 
PDF
2018,
Vol. 39 Issue (11): 1792-1803 PDF