Photocatalytic degradation of organic dye and tetracycline by ternary Ag2O/AgBr–CeO2 photocatalyst under visible-light irradiation

In this work, CeO2 nanosheets decorated with Ag2O and AgBr are successfully fabricated via a simple sediment-precipitation method. The as-prepared ternary Ag2O/AgBr–CeO2 composite with double Z-scheme construction was analyzed by various analytical techniques. Ag nanoparticles (NPs) used as the electron medium could reduce the recombination of photoelectrons and holes, thus leading to the improvement of photocatalytic performance of these catalysts. Due to the unique structure and composite advantages, the optimal Ag2O/AgBr–CeO2 photocatalysts exhibit the superior tetracycline (TC) degradation efficiency of 93.23% and favorable stability with near-initial capacity under visible light irradiation. This ternary Z-scheme structure materials will be the well-promising photocatalysts or the purification of antibiotic wastewater.

In recent years, more and more researchers have paid attention to the preparation of ternary composites. In the whole ternary catalytic reaction process, photo-generated carriers can be transferred in multiple steps through the induction mechanism to achieve the purpose of electron and hole separation, thereby achieving photocatalytic activity beyond the binary catalytic system 17,18 . And an increasing number of ternary photocatalytic materials have been reported, such as Fe 3 O 4 /Bi 2 S 3 /BiOBr 19 , Ag 3 PO 4 /TiO 2 /Fe 3 O 4 20 , Bi 2 WO 6 /Ag 2 S/ZnS 21 , and V 2 O 5 /BiVO 4 /TiO 2 22 . In spite of some ternary composites reported, there are few reports about cerium dioxide ternary materials. Therefore, it is expected to design a preeminent three-way CeO 2 -based photocatalyst with favorable catalytic properties.
It's necessary to select suitable doping materials which could directly affect the efficiency of whole ternary composites, to design the ternary CeO 2 -based photocatalysts. Ag 2 O has excellent visible light absorption properties, while AgBr has excellent photocatalytic activity 23,24 . Compared to the existing materials, the two may be the better choice. However, pure Ag 2 O and AgBr also have problems such as lower photo-generated carrier yield, unstable photocatalytic activity and high carrier recombination rate. To address these shortcomings, researchers often use them as co-catalysts to improve their stability. For example, Wen et al. fabricated Ag 2 O-CeO 2 photocatalyst to degrade enrofloxacin effectively by Ag 2 O nanoparticles embellishing CeO 2 spindles 25 . Huang et al. synthesized flower-like AgBr/Bi 2 WO 6 , which degraded 87.5% TC solution within 60 min under visible light irradiation 6 . Thus, we attempt to modify cerium dioxide with silver oxide and silver bromide to construct Ag 2 O/AgBr-CeO 2 ternary composites.

Results and discussion
Physical and chemical properties of samples. The SEM images of CeO 2 , Ag 2 O, and ACA-2 were shown in Fig. 1. In Fig. 1a, the length and the thickness of CeO 2 nanosheets are about 1-5 μm and 50-100 nm, respectively. The size of Ag 2 O NPs is about 206 nm ( Supplementary Fig. S1), but most of Ag 2 O NPs were agglomerated. In Fig. 1b, the particle size of Ag 2 O was 100-400 nm and relatively inhomogeneous. In Fig. 1c, after decoration with Ag 2 O and AgBr NPs, it was clearly to be seen that Ag 2 O spreads over the surface of CeO 2 with the lower agglomeration of Ag 2 O. Ag 2 O and AgBr NPs stick well to the surface of CeO 2 nanosheets. The grain sizes of Ag 2 O and AgBr NPs were more uniform. The main reason is that the corrosion of hydrobromic acid makes silver oxide release Ag. Subsequently, the Ag NPs are evenly distributed on the surface of CeO 2 to form homogeneous AgBr NPs 17 . Therefore, the above result confirmed that the preparation of the three-way Ag 2 O/ AgBr-CeO 2 catalyst is successful.
The TEM and HRTEM were applied to further analyse the structure of the ACA-2. In the picture of Fig. 1d, a large number of Ag 2 O and AgBr NPs with an average diameter below 30 nm were grown along the surface of CeO 2 microsheets. The (311) crystal plane of CeO 2 was 0.343 nm 26 Fig. S2). The photos of TEM-EDX mapping are exhibited in Fig. 1e. TEM-EDX mapping was used to analyze each element of the sample. As shown in Fig. 1e 25 . The diffraction peak of AgBr standard cards (JCPDS: 06-0438) 29 correspond to XRD patterns of the pure AgBr. Respecting the ACA-2 composites, the typical diffraction peaks of CeO 2 , AgBr, and Ag 2 O can be clearly observed. According to Supplementary Fig. S3, it could also be clearly observed that the peak of AgBr of each material become more obvious with the increase of HBr added amount. The unobvious peaks of Ag 2 O due to the shading effect of the close peaks of CeO 2 25 . According to the above analysis, the crystal phase of Ag 2 O/AgBr-CeO 2 was not affected by loading the Ag 2 O and AgBr.
X-ray photoelectron spectroscopy (XPS) tests were carried out to determine the elemental composition and the chemical state of ACA-2. The XPS survey spectrum in Fig. 2b shows that the product contains Ce, O, Ag, and Br elements. There are four peaks of Ce (III) spectra at 882.27, 888.93, 900.69, and 907.60 eV. And two peaks of Ce (IV) at 898.30 and 916.71 eV in Fig. 2c, which is in keeping with previous reports 30 . The positions of Ag 3d 5/2 and Ag 3d 3/2 peaks are at 367.44 and 373.45 eV (Fig. 2d), respectively, illustrating the monovalent chemical valence of Ag 5 . The peaks of Br 3d can be assigned to Br 3d 5/2 (67.87 eV) and Br 3d 3/2 (68.67 eV) in Fig. 2e 31 . As presented in Fig. 2f, it is observed that O 1 s has two peaks ( Fig. 2f) of Ce-O and Ag-O bonds. Besides, the other peak is related to absorbed oxygen and H 2 O 27 . Therefore, the consequence of XPS analysis indicates that Ag 2 O and AgBr connect with CeO 2 via chemically bound interface rather than the physical contact.
UV-Vis spectra were illustrated in Fig. 3a. CeO 2 and AgBr exhibited visible light absorption with absorption band edges at 436 and 519 nm, respectively 32,33 . Meanwhile, pure Ag 2 O revealed an evident absorption in completely visible light scope, which corresponds with the previous reports 34. As shown in Fig. 3b, the bandgap (E g ) could be obtained by Kubelka-Munk function: αhv = A (hv-E g ) n/2 , where E g , α, A, h, ν correspond to energy gap, absorption coefficient, a constant, Planck's constant and light frequency. CeO 2 , Ag 2 O and AgBr were indirect bandgap, and four was the value of n 35 . The E g values of CeO 2 , AgBr and Ag 2 O were about 2.98, 2.56 and 1.34 eV, respectively.
According to the empirical formulas E CB =X − E C − 0.5E g and E VB = E CB + E g , where X, E C and E g represented the electronegativity of crystalline semiconductors, the energy of free electrons on the hydrogen scale (~ 4.50 eV NHE) and the energy gap of semiconductors 36,37 . The E CB of CeO 2 , Ag 2 O and AgBr were calculated to be − 0.39, 0.13 and 0.02 eV. Then their E VB was corresponded to 2.59, 1.47, and 2.58 eV.
Photocatalyst behaviors. The photocatalytic performances of the samples over RhB were studied under visible light. The details of the dark reaction experiments were shown in Supplementary Fig. S4. From Fig. 4a, the degradation of RhB did not proceed in the blank experiment, which illustrates RhB can't be degraded under visible irradiation. The degradation rates of RhB for AOC and CAB were 58.48% and 67.55% within 60 min. And degradation rate of pure CeO 2 was only 16.99%. For Ag 2 O/AgBr-CeO 2 composites, the ACA-2 sample showed the best photocatalytic activity, which can remove 95.59% RhB. It could be discovered that the quantity of AgBr affected their ability of degradation. The degradation rate was increased because the purity of AgBr was from 7.92 to 15.37 wt%. When the content of AgBr was exceeded 15.37 wt%, its shading effect had an impact on the photocatalytic activity of composites. As shown from the photocatalytic degradation reaction kinetics of RhB in Fig. 4b,c, the kinetic constant of ACA-2 (0.04805 min −1 ) based on the muri-Hinshelwood model was higher than other samples 38 . However, the kinetic constant of CeO 2 was only 0.00182 min −1 . According to the analysis in Fig. 4b, the appropriate addition of AgBr could enhance the catalytic property.
Moreover, TC was selected as a typical antibiotic pollutant, and the obtained samples were degraded to eliminate the influence of dye self-sensitization. The details of the dark reaction experiments were shown in Supplementary Fig. S5. As shown in Fig. 4d, only 17.95% and 55.84% of TC could be removed by pure CeO 2 and AOC after 60 min of visible light irradiation, respectively. As expected, when AgBr nanoparticle put into composites, the photocatalytic activity of ACA-2 (93.68%) was more outstanding than other materials. The catalyst exhibited evidently degradation effect was analogous to its catalytic behavior of removal RhB. Meanwhile, the degradation effect kinetic of TC for all products were studied in Fig. 4e,f. Kinetic constants of ACA-1, ACA-2, ACA-3, ACA-4, CAB and AOC were to 0.02314, 0.04014, 0.02526, 0.02016, 0.01292 and 0.01122 min −1 , while CeO 2 was only 0.00149 min −1 . The information could prove that the photocatalytic activity of CeO 2 could be improved by Ag 2 O and AgBr modification.
As a matter of fact, the concentration of the degradant had an outstanding impact on the photocatalytic activity of Ag 2 O/AgBr-CeO 2 . In Fig. 5a, different concentration of TC was removed by ACA-2. It was obvious that the degradation effect of 10 mg/L TC was 93.84% which was higher than the TC solutions of 20, 30 and 40 mg/L. The main reasons for the attenuation of degradation activity could be the following influences: (1)  (2) As the concentration of pollutants increases, the number of intermediates produced in the reaction increases, which led to the competition with TC 39 . Meanwhile, for TOC degradation, ACA-2 also shows the best removal rate ( Supplementary Fig. S8).
It is well known that there are many kinds of anions in practical wastewater, which may affect the degradation efficiency of polluted solution by photocatalysts. Therefore, we studied the effects of various anions (SO 4 2− , Cl − , and HCO 3 − ) on photocatalytic degradation of TC solution using ACA-2. As shown in Fig. 5b, the degradation rate of TC solution with Na 2 SO 4 decreased slightly 40 . However, when NaCl appended to the solution, the removal efficiency of TC was reduced to some extent, which might put down to competitive adsorption of Clwith other substances 38 . The addition of NaHCO 3 has a significant effect on the degradation of TC because HCO 3 has the www.nature.com/scientificreports/ function of a free radical scavenger, which can consume some active free radical 27 . Therefore, the reduction of radicals decreased the photocatalytic degradation of TC by ACA-2.
As a matter of fact, the excellent recycle property and stability of photocatalysts can effectively cut the waste water treatment cost and avert secondary pollution. The stability of ACA-2 was conducted by 4 recycling experiments in Fig. 5c. Specifically, even in the fourth cycle, Ag 2 O/AgBr-CeO 2 composites could degrade 86.64% TC. According to the SEM and TEM images of used ACA-2 ( Supplementary Fig. S6), the micro-morphology of ACA-2 has not been destroyed after multiple reactions. From Fig. 5d, the XRD pattern of the sample after degradation reactions basically corresponded to the fresh sample. The consequence indicated that the stability of Ag 2 O/AgBr-CeO 2 was excellent.
In order to investigate the sample changes before and after the reaction, XPS analysis was performed on the samples after the photocatalytic reaction. According to Fig. 6a,b, there are two peaks of Ag in two valence states, which shows that silver nanoparticles appear on the surface of the material. And Ag NPs could reduce photo corrosion and inhibit the photolysis of AgBr and Ag 2 O. In addition, Ag NPs can serve as an electronic medium in the catalytic process.
Photocatalytic mechanism. To research the photocatalytic mechanism of Ag 2 O/AgBr-CeO 2 composites, free radical capture experiments were carried out. Three different trapping agents of isopropyl alcohol (IPA, ·OH scavenger), sodium oxalate (Na 2 C 2 O 4 , ·h + scavenger) and benzoquinone (BQ, ·O 2 − scavenger) 41,42 were added to the solution of TC before photodegradation. As presented in Fig. 6c,d, the degradation of TC hardly reduced after the addition of BQ, explaining that the ·O 2 − played an important role in reaction systems. Meanwhile, when Na 2 C 2 O 4 and IPA were added into the TC solution, the degradation rate of TC was 63.54% and 65.99% respectively, less than the removal of no scavenger. Therefore, holes and ·OH also played a pivotal role in the catalysis system. It can be deduced that the h + , ·OH and ·O 2 − were the main active radicals affecting the elimination of TC solutions.
As shown form the Mott-Schottky curve of ACA-2 (Fig. 6e), the ternary Ag 2 O/AgBr-CeO 2 catalyst is composed of p-type semiconductor Ag 2 O and n-type semiconductor CeO 2 and AgBr. The p-n heterojunction is effectively constructed in the composite catalyst, which can effectively separate the electron-hole 43 . On the side, photocurrent density could be an efficient method of evaluating the transfer properties of the photogenerated electrons as well. And the higher the photocurrent response the higher separation efficiency. From Fig. 6f, these curves of photocurrent density of CeO 2 , AOC, CAB and ACA-2 composites under illuminating visible light were measured. These composites and CeO 2 showed the stable photocurrents under visible light irradiation respectively generating electrons and holes 44 . And the result of EIS spectra of ACA-2 ( Supplementary Fig. S7a) was consistent with that of photocurrent curves. It was obvious that ACA-2 displayed the best excellent photocurrent density curve that illustrated the better separation efficiency of photogenerated electron-hole pairs than others 45 . Generally speaking, the lower the fluorescence intensity of PL spectrum, the better the electron hole separation efficiency of catalyst. According to Supplementary Fig. S7b, ACA-2 has the lowest fluorescence intensity which indicates that it has better catalytic capacity.
In order to verify the activity of free radicals, the samples were tested by Electron Paramagnetic Resonance (EPR) 46 . It can be seen from Fig. 7a that there is no EPR signal under dark conditions. However, the characteristic signal of ·O 2 − appears under visible light irradiation, and the signal intensity gradually increases with the increase of the irradiation time (1-8 min), indicating that ·O 2 − are generated in the reaction system and participate in the photocatalytic degradation reaction. According to Fig. 7b, the characteristic signal of ·OH is also no signal under dark conditions, and gradually increases with the prolongation of the light time (1-8 min), which indicates that ·OH participate in the photocatalyst reaction. It is consistent with the results of active species trapping experiments.
The mechanism of photodegradation of TC over the catalyst is shown in Fig. 8. According to the traditional type I heterojunction structure (Fig. 8a), CeO 2 is excited to produce photoelectrons under visible light irradiation.   39,47 . Therefore, photocatalyst doesn't produce ·O 2 − to oxidize organic matter that is inconsistent with those of free radical capture experiments. Based on the above analysis, the conjecture of Scheme 1 is unreasonable, so the second mechanism is proposed to explain the reaction. According to Fig. 6b, it is clearly obvious that the XPS spectrum of used samples has the peak of Ag 0 . Thus, the Z-scheme heterojunction system with Ag NPs as the bridge between Ag 2 O, AgBr and CeO 2 can be carried out under visible light. As presented in Fig. 8b 48 . And then hydroxyl radicals oxidize organic matter into degradation products. The Scheme 2 is consist of the free radical capture experiments of reactive species. In brief, the double Z-scheme heterojunction system with Ag 2 O/AgBr-CeO 2 composites could point out a particular mechanism guess, which not only improves the efficiency of carrier separations but also consolidates its stability.
In order to further understand the degradation process of TC, the spectroscopy of the catalytic process of the intermediate products was detected by liquid chromatography-mass spectrometry 48 (LC-MS, details are in the support material) (Scheme 1). Many intermediates were detected in the solution, and the corresponding mass spectra are shown in Supplementary Fig. S9. The product with a mass-charge ratio (m/z) of 445 is due to the TC molecule. In addition, the m/z values produced during the photocatalytic degradation of TC in the presence of ACA-2 were 450, 406, 362, 334, 318, 290, 274, 246, 230 and 202. Based on the previous reports and the analysis of the results, two possible degradation pathways were proposed. For the first route, the double bond on TC are reduced to TC-2 (m/z = 450). And then TC-2 (m/z = 450) deaminates to form TC-3 (m/z = 334). TC-4 (m/z = 290) was obtained by removing one of the hydroxyl groups from TC-3 (m/z = 334). The second pathway was from the TC (m/z = 445) to TC-5 (m/z = 406) via fracture of the double bond. Afterwards, TC-5 deaminates to form TC-6 (m/z = 362), and TC-6 removed methyl group to form TC-7 (m/z = 318). Hydroxide radical of TC-7 was changed into TC-8 (m/z = 274) after oxidation. As the same time, TC-8 removed a formyl group to form TC-9 (m/z = 246). TC-9 continued to break off a carbonyl group to get TC-10 (m/z = 230). Meanwhile, TC-10 break the ring and removed a hydroxymethyl to form TC-11 (m/z = 202). Finally, a part of intermediates can ultimately mineralized into H 2 O and CO 2 .  We believe that the novel three-way Ag 2 O/AgBr-CeO 2 catalyst could have great potential in the field of energy and environmental protection. This work could provide a different Z-scheme heterojunction system to construct ternary catalysts improving the degradation efficiency.     Characterization. It is convenient to analyze and study the crystal structure and material type of samples through X-Ray Diffraction (XRD, Rigaku Co., Japan, D/max 2500 VL/PC) and the field emission scanning electron microscopy (FESEM, JEOL-2100). Field Emission Transmission Electron Microscope (FETEM, JEM-2100F) achieved TEM and HRTEM photographs with an acceleration voltage of 200 kV. X-ray photoelectron spectroscopy (XPS, Thermos ESCALAB Xi+) performed an elemental analysis of the samples. UV-Vis diffuse reflectance spectra (DRS) were recorded on Cary 5000 using with BaSO 4 as reference material. Electrochemical measurements were performed at Zahner electrochemical workstation using a standard three-electrode system. Detailed instructions are attached to the supporting information. Electron paramagnetic resonance (EPR) was used to verify the existence of ·O 2 − and ·OH.
Activity test of photocatalyst. The photocatalytic activity of composites was studied by the degradation of Rhodamine B (RhB) and tetracycline (TC) under the visible light irradiation (λ > 420 nm). The catalyst (0.025 g) was dispersed in the 50 ml solution with RhB or TC (the concentration of RhB or TC was 10 mg/l). The mixed solution stirred for 30 min in a dark environment to reach adsorption equilibrium. Afterward, the suspension was illuminated in visible light (500 w X lamp, Shanghai Qiqian Technology Co.). 4.00 ml suspensions that were achieved with every 15 min were centrifuged to remove catalysts. The concentration of the RhB (554 nm) and TC (357 nm) was received by the UV-1200 spectrophotometer at its maximum absorption wavelength.
Received: 1 June 2020; Accepted: 13 October 2020 Scientific Reports | (2021) 11:85 | https://doi.org/10.1038/s41598-020-76997-0 www.nature.com/scientificreports/ Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creat iveco mmons .org/licen ses/by/4.0/.