The Roles of Graphene and Ag in the Hybrid Ag@Ag 2 O-Graphene for Sulfamethoxazole Degradation

: Ag@Ag 2 O-graphene (Ag@Ag 2 O-G) with different concentrations of graphene was synthesized using a facile in situ precipitation method. The photocatalysts were characterized by ﬁeld emission scanning electron microscopy (SEM), transmission electron microscopy (TEM), X-ray diffraction (XRD), and X-ray photoelectron spectroscopy (XPS), UV-vis diffuse reﬂectance spectra (DRS). The antibioticsulfamethoxazole (SMX) degradationunder simulated solar light and visible light irradiationwas investigated to evaluate photocatalytic performance. The composite photocatalyst Ag@Ag 2 O-G with 2.5 wt% graphene presented the highest activity among all the prepared composite photocatalysts. The coupling of graphene and Ag 0 increased the photocatalyticactivity and stability of pure Ag 2 O. Under higher SMX concentrations, the adsorption, not the photocatalytic ability, playeda crucial role during the SMX removal process. On the basis of the characterization and reactive oxygen species (ROS) scavenging experiments, a separation and transfer mechanism of photogenerated carriers was proposed. In the photocatalytic degradation of SMX, the major active species wereidentiﬁed as photogenerated holes; photogenerated electrons in the conduction band (CB) of Ag 2 O could not transfer to graphene through Ag 0 due to the more negative reduction potential of graphene. This is an important result regardinggraphene and Ag 0 roles which isdifferent from that for the photocatalytic degradation of dyes. This researchmay provide new insights into photocatalytic processes for the degradation of non-dye pollutants bycomposite materials to guidethe design of highly efﬁcient reaction systems.


Introduction
Semiconductor photocatalysis, one of the most promising technologies for solar energy utilization and environmental remediation, has attracted tremendous attention since the semiconductor TiO 2 was applied to the decomposition of organic contaminants [1,2]. One of the most important objectives in semiconductor photocatalysis research is the development of efficient photocatalysts that can harvest visible light under sunlight. Although TiO 2 -based photocatalysts have been demonstrated to be efficient photocatalytic materials under visible light irradiation [3], the variety of visible light photocatalysts is still very limited, and the exploration of new and efficient photocatalytic materials sensitive to visible light remains challenging. Presently, some Ag-based semiconductors, such as Ag 3 PO 4 [4], AgCl [5], and Ag 2 O have gained attention because of their characteristics of easy visible light excitation and low mammalian toxicity. Among them, Ag 2 O, a brown powder with a simple cubic structure anda lattice distance of 0.472 nm, has a narrow band-gap energy that can be easily excited by incident visible light [6]. However, after absorbing incident light, photons in Ag 2 O, the photogenerated electrons in the conduction band (CB) are captured by surface lattice Ag + ions to form Ag 0 clusters, whereas photogenerated holes in the valence band (VB) oxidize lattice O 2− to release O 2 , which results in the photodecomposition of Ag 2 O and the formation of Ag 0 . Therefore, pure Ag 2 O has been rarely used as a visible light-driven photocatalyst because of its high photosensitivity and low-stability properties under light irradiation. There have been a few attempts to use Ag 2 O-based materials as a photocatalyst under visible light irradiation. Ag@Ag 2 O, AgBr/Ag 2 O, Ag 2 O/TiO 2 heterostructure were introduced to decompose methyl orange (MO), and such attempts have exhibited high efficiency and stability [7][8][9]. These results indicate that Ag 2 O-based materials have greater potential as stable and highly efficient photocatalysts than pure Ag 2 O for photocatalytic decomposition of organic contaminants under visible light irradiation.
It is worthwhile to note that once a certain amount of Ag 0 forms on the surface of Ag 2 O, the following photogenerated electrons tend to transfer to the Ag 0 sites and are captured by O 2 [10]. To inhibit Ag 2 O corrosion, it is expected that the photogenerated holes could transfer to other electron donors (such as organic substances), and the photogenerated electrons could transfer to other electron captures (such as O 2 ) before reacting with lattice elements. It is possible that the produced Ag 0 works as an electron pool and transfers the photogenerated electrons to improve photocatalytic activities. Therefore, the produced Ag 0 could bring about an effective segregation capability of photo-excited electron-hole pairs and suppress Ag 2 O photo-reduction. In order to addressAg 2 O photocatalyst instability, a strategy of implementing hybrid material has been undertaken to determine an effective method to improve the catalytic activity and stability. Graphene, owing to its large specific area, flexible structure, high thermal and electrical conductivity, high chemical stability, and excellent electron capture and transport properties [11], could provide a supporting matrix to suppress the aggregation of nanoparticles and promote the effective separation of photogenerated carriers as an electron mediator for shuttling electrons [12]. Recently, combing graphene (G) with photocatalysts to enhance their photocatalytic performance has attracted considerable attention [13,14]. For example, the coupling of graphene with TiO 2 could extend the light absorption from UV wavelengths into the visible region [15] and improve adsorption abilities and photocatalytic activities [16]. Graphene has been one of the most promising materials to enhance the photocatalytic performance in the new generation of photocatalysts [17]. However, in previous reports about Ag@Ag 3 PO 4 -G composite materials, the coupling graphene and Ag 0 could not enhance its photocatalytic activity in organic pollutants degradation [4]. Therefore, the effects of graphene and Ag 0 in composite photocatalysts remain unclear and deserve further research.
It should be noted that most previous studies of Ag 2 O-based composite photocatalysts have focused on the decomposition of dyes, such as methyl orange (MO) [7], rhodamine B (RhB) [14], or methylene blue (MB) [18]. However, dyes used as photosensitization compounds could be excited by some wavelength lights to release electrons affecting the transfer of photogenerated carriers [19]. Other appropriate target pollutants should also be explored. The trace-level water contaminant sulfamethoxazole (SMX) can only absorb light at wavelengths lower than 320 nm [20], which makes SMX an ideal candidate to analyze photocatalytic activities and mechanisms of visible light-driven photocatalysts.
Herein, hybrid Ag@Ag 2 O-G photocatalysts with different graphene concentrations were synthesized through a precipitation method. Photocatalytic activity of Ag@Ag 2 O-G was evaluated by monitoring the degradation of SMX under simulated solar light (λ > 280 nm) and visible light irradiation (λ > 400 nm). In addition to photodegradation and adsorption abilities of Ag@Ag 2 O-G on SMX, the stability of one of the photocatalysts after several recycles was also investigated. A possible photocatalytic degradation mechanism was proposed through investigation of the influence of reactive oxygen species (ROS) scavengers on Ag@Ag 2 O-G photocatalytic activities. Through this research, the roles of graphene and Ag 0 in the composite Ag 2 O-based materials were investigated systemically using SMX (non-photosensitizer, as target pollutant). This will provide new insights into photocatalytic mechanism analysis for composite photocatalysts.

Catalyst Characterization
Morphology of the as-prepared Ag@Ag 2 O-G2 was elucidated by SEM (scanning electron microscopy) and TEM (transmission electron microscopy), as shown in Figure 1a,b. It can be clearly seen that the Ag 2 O nanoparticles were close to spherical in shape with an average diameter of approximately 100 nm and covered by graphene with a folder surface. From the different colors of Figure 1b, the Ag 2 O particles in the dark color and graphene in the light color are clearly identified. The production of Ag on the surface of Ag 2 O particles was confirmed by the following XRD (X-ray diffraction) and XPS analysis.  Figure 2b, the XPS peaks of Ag 3d also confirm the formation of Ag. The peaks centered at 368, and 374 eV corresponded to Ag 3d 5/2 and Ag 3d 3/2 , respectively. The Ag 3d 5/2 peak is divided into two different peaks at 368.6 and 368.1 eV, and the Ag 3d 3/2 peak is divided into 374.6 and 374.1 eV, respectively. The peaks at 368.6 and 374.6 eV are attributed to Ag, whereas the peaks at 368.1 and 374.1 eV are attributed to Ag + in Ag 2 O, and the two pairs of separated silver peaks show a difference of D = 6 eV [21]. The molar content of the surface metallic Ag is calculated to be 11.2 wt% of the total silver element in Ag@Ag 2 O-G2 according to the areas of divided peaks (as shown in Figure 2b). The XPS peak corresponded highly to the XRD analysis, proving the co-existence of Ag and Ag 2 O in the composite.

Photocatalytic Performance
As shown in Figure 3a under visible light, respectively. It was concluded that the light of the UV part ofthe simulated solar light could result in more Ag 2 O corrosion, which desorbsthe SMX more than visible light [9]. The relationship between ln(C 5 /C T-5 ) (C 5 , the initial concentration of SMX after 5 min irradiation and C T-5 , the concentration of SMX after (T-5) min irradiation) and irradiation time followedfirst-order kinetics. The basic kinetic parameters for the Ag 2 O seriesphotocatalysts under different light irradiations areshowed in Table S1. Obtained values of the apparent rate constants were 0.016, 0.0353, 0.045, and 0.027 min −1 under simulated solar light and 0.012, 0.0340, 0.038, and 0.028 min −1 under visible light (R 2 > 0.96). The difference between photocatalytic activities under simulated solar light and visible light was not significant because of the similar light intensity (37.7 mW/cm 2 for simulated solar light and 34.2 mW/cm 2 for visible light). The adsorption capacity increased and the desorption percentages decreased with the additional graphene and the Ag 0 produced in the photocatalysts. However, the photocatalytic activities were not in linear relation with the graphene addition, and the Ag@Ag 2 O-G2 with 2.5 wt% of graphene content showed the highest degradation efficiency for SMX removal. The photocatalytic performance deteriorated when the graphene dosage was increased above its optimal value. This phenomenon could be because graphene absorbs some light and, thus, might cause a light harvesting competition between Ag 2 O and graphene with the increasing graphene concentration [22], which was consistent with light absorption abilities of photocatalysts (as shown in Figure S1). On the other hand, thephotogenerated Ag 0 on the surface of Ag 2 O enhanced its structural stability due to a self-stability mechanism [10] that decreased the photocatalyst corrosion and SMX desorption percentages. In order to avoid photocatalyst corrosion, visible light was adopted to conduct further experiments. The results of adsorption and photocatalytic degradation in different SMX concentrations are shown in Figure 4. In the higher concentrations of 2.5, 5, and 10 mg/L, the SMX was mostly removed by adsorption. The adsorption capacities of Ag@Ag 2 O-G2 for 1, 2.5, 5, and 10 mg/L SMX were 5.0, 23.6, 47.4, and 98.4 mg/L, respectively (as shown in Figure 4a). As the SMX concentrations were increased, the adsorption capacity of the photocatalyst also obviously increased, and almost no degradation products were detected after 10 mg/L SMX was adsorbed and photocatalytically degraded (as shown in Figure 5b). However, with the exception of an SMX substance peak at 4.11 min for 1 mg/L SMX, there were some other peaks for degradation products identified by high-performance liquid chromatography (HPLC) (as shown in Figure 5a). The results indicate that the photocatalyst adsorption ability was not positively related to the photocatalytic activity, which is consistent with the results of previous research [23]. This may be ascribed to the competition of light absorption between the photocatalyst and pollutants. The higher SMX concentration beyond the value of 2.5 mg/L with multilayer SMX coverage on the photocatalyst weakens the radiant flux necessary for the Ag 2 O [24], which explains the phenomenon that no obvious desorption occurred after 5 min irradiation under a higher SMX concentration (≥2.5 mg/L).

Photocatalyst Stability
As shown in Figure 6, in the first two successive experimental runs (a total of four hours), SMX removal by adsorption and photocatalytic degradation of Ag@Ag 2 O-G2 remained almost the same. However, the SMX removal percentagedecreased from almost 100% to approximately 40% in the third run and almost all of it was removed by adsorption. When the SMX concentration was 2.5 mg/L, the adsorption ability was crucial to the removal of SMX. After the photocatalyst was recycled twice, the SMX removal efficiency decreased rapidly in the third recycle due to adsorption capacity saturation and more photocorrosion occurred in the photocatalyst (as shown in Figure S2). It could be concluded that the Ag@Ag 2 O-G2 has the ability to photocatalytically oxidize low-concentration SMX (≤1 mg/L) and adsorb high-concentration SMX (≥2.5 mg/L).

Mechanism Analysis
In the photocatalytic system (C SMX = 1 mg/L, C Ag@Ag2O-G2 = 0.05 g/L), the oxidative species were probed by scavenger addition into the solution. The results in Figure 7 show that benzoquinone and sodium azide addition had negligible effects, while isopropanol addition had a slight effect on SMX photodegradation. The results indicatedthat the main reactive oxygen species were not •OH, O 2 • − and 1 O 2 in this photocatalytic process. A rapid termination of photocatalysis was investigated after Na 2 -EDTA addition and only 5.2% of SMX was removed after 90 min of irradiation. One possible explanation is that, on the surface of the photocatalyst, the deposited EDTA anion as a hole trap blocked the reaction sites. This implies that the SMX oxidation process occurred on the surface of photocatalyst, and photogenerated holes were the main reactive species. The schematics of the possible charge transfer and oxidation processes under visible light irradiation are illustrated in Figure 8. Ag 2 O was excited to photogenerated electrons and holes in its CB and VB. The CB and VB potentials of Ag 2 O were +0.20 and +1.50 V [25] (as shown in Figure 8) versus a normal hydrogen electrode (NHE), respectively. The holes in the VB of the Ag 2 O (+1.50 V/NHE) could degrade the SMX (+0.9 V/NHE [26]) to CO 2 or to other intermediates, but they could not oxidize the OH − to •OH(+1.89 V/NHE [27]) or H 2 O to •OH (+2.68 V/NHE [28]). The electrons in the CB of the Ag 2 O (+0.20 V/NHE) could not reduce the O 2 to O 2 • − (−0.33 V/NHE) and HO 2 • (−0.55 V/NHE) [29]. However, the electrons could reduce the O 2 to H 2 O 2 (+0.695 V/NHE) [29] through a two-electron reduction [30] and then continue to react with electrons to produce active •OH [31], as verified by the scavenging experiments. The Ag 0 produced in the catalyst preparation process could act as electron capture centers [32], and transfer electrons in the CB of the Ag 2 O to oxygen through multi-electron-transfer routes. In other studies, graphene has been considered effective in reducing the combination rate of photogenerated electron-hole pairs when coupling with semiconductors owing to the electron shuttle from the conjugated carbon network [22]. However, the electrons in the CB of the Ag 2 O could not transfer to the graphene through Ag 0 because of the more negative graphene reduction potential (−0.11 V) [19,33], which indicated that graphene could increase photocatalytic activities only through higheradsorption capacity and more active sites to anchor the SMX. Therefore, in the Ag@Ag 2 O-G photocatalytic system, the major active species for photocatalytic degradation were identified as the holes, while graphene was identified as the adsorbent and Ag 0 as the electron capture center, which increased the Ag 2 O photocatalytic activities to remove the SMX and the stability of photocatalysts.

Catalyst Preparation
Silver nitrate (AgNO 3 ), sodium hydroxide (NaOH), ammonia solution (25-28%), ethanol (>99.7%) were purchased from Sinopharm Chemical Reagent Co., Ltd. (Shanghai, China). Graphene was purchased from Nanjing XFNANO Materials Tech Co., Ltd. (Nanjing, China). All chemicals used in this study were of analytical grade without further purification. In a typical synthesis of the Ag@Ag 2 O-G composite, solutions of 50 mL of 0.2 M AgNO 3 , 100 mL of 0.2 M ammonia water, and 100 mL of 2 M NaOH were prepared in advance. The ammonia water was added to the AgNO 3 solution slowly, changing the color of the solution from yellowish-brown to transparent-thus, creating a silver-ammonia solution. A certain amount of graphene powder was then dispersed into the silver-ammonia solution and sonicated for 2 h. Afterwards, the NaOH solution was dropped into the silver-ammonia solution with added graphene to form black Ag 2 O-G. To synthesize the Ag@Ag 2 O-G, the Ag 2 O-G suspension was irradiated for 3 h using a 300 W Xe lamp equipped with an optical cut-off filter (λ > 400 nm). The resultantAg@Ag 2 O-G was separated by centrifugation (1000 rpm for 30 min), and then washed with Milli-Q water and ethanol several times and freeze-dried under a vacuum. According to the graphene concentrations of 1.7, 2.5, and 3.4 wt%, the obtained samples were labeled as Ag@Ag 2 O-G1, Ag@Ag 2 O-G2, and Ag@Ag 2 O-G3. For comparison, the same procedure without graphene and irradiation was carried out to obtain pure Ag 2 O.

Analysis Method
The SMX concentration was quantified by high-performance liquid chromatography (HPLC) supplied by Agilent with a Teknokroma C-18 Tracer Extrasil ODS2 (Teknokroma, Barcelona, Spain) (250 mm × 4.6 mm). The mobile phase was a mixture of 40% acetonitrile and 60% water which was adjusted at pH = 3 with phosphoric acid. The analysis was performed under isocratic conditions at a flow rate of 1 mL/min, with an injection volume of 10 µL and an oven temperature of 25 • C. The wavelength of the UV detector was set at 270 nm.

Photocatalytic Degradation of SMX
The photocatalytic experiments for SMX removal were conducted with a 300 W xenon lamp with a cut-off filter (λ > 280 nm) as a simulated solar light source and a cut-off filter (λ > 400 nm) as the visible light source. The light intensity on the surface of the suspension was kept at 37.7 mW/cm 2 for simulated solar light and 34.2 mW/cm 2 for visible light. SMX photocatalytic oxidation was completed in a 250 mL beaker with a cooling system on the bottom to maintain a constant solution temperature of 25 ± 2 • C. In a typical experiment, 5 mg of photocatalysts were suspended in 100 mL of 1 mg/L SMX solution by ultrasonication for 5 min. Before irradiation, the mixture suspension was stirred thoroughly in the dark for 30 min to ensure an adsorption-desorption equilibrium between the SMX and the photocatalyst. The lamp was then switched on to initiate the photocatalytic reaction. 3 mL of solution was sampled at certain time intervals and filtered with 0.45 µm syringe filters to remove suspended photocatalysts for the SMX analysis. Experiments for different SMX concentrations were conducted (C catalyst = 0.05 g/L, C SMX = 1, 2.5, 5, 10 mg/L) using the same procedures. Investigation of the reactive oxidative species was similar to the photodegradation process, while the scavengers (10 mM of iso-propanol, 0.5 µM of benzoquinone, 0.5 µM of sodium azide (NaN 3 ) and 5 mM of disodium ethylene diamin tetraacetate (Na 2 -EDTA)) were added to the SMX solution prior to addition of the photocatalyst. In the recycling experiments, higher concentrations of photocatalysts and SMX (C catalyst = 0.25 g/L, C SMX = 5 mg/L) were used. The used photocatalysts were separated from the solution and then washed with Milli-Q water thoroughly before being re-dispersed in the fresh SMX solution for another cycle.

Conclusions
A hybrid Ag@Ag 2 O-graphene photocatalyst was synthesized with a facile chemical precipitation approach and reduction procedure under visible light. The Ag@Ag 2 O-G2 with 2.5 wt% of graphene content possessed a higher activity than the other prepared composite photocatalysts with 1.7 and 3.4 wt% of graphene content. The addition of graphene and produced Ag 0 not only increased the photocatalyst adsorption capacities, but also decreased desorption percentages after irradiation. Under a higher SMX concentration, the adsorption, but not the photocatalytic ability, played a crucial role during the SMX removal process. The Ag@Ag 2 O-G2 showed stable photocatalytic activities in at least two successive experimental runs (a total of four hours). During the photocatalytic processes of Ag@Ag 2 O-G2, the generated holes were dominant reactive oxygen species for the photocatalytic degradation of SMX and Ag 0 , as the electron capture center could not transfer electrons in the CB of the Ag 2 O to graphene due to the negative graphene reduction potential.