Benxia Li a,⇑ , Baoshan Zhang b , Shibin Nie c , Liangzhi Shao b , Luyang Hu b

aDepartment of Chemistry, College of Science, Zhejiang Sci-Tech University, Hangzhou 310018, China

b School of Materials Science and Engineering, Anhui University of Science and Technology, Huainan, Anhui 232001, China

c School of Energy Resources and Safety, Anhui University of Science and Technology, Huainan, Anhui 232001, China

abstract

Exploiting photocatalysts with improved properties for solar-driven chemical reactions is of great significance in developing green chemistry. Here, a series of Au/CeO2 hybrid nanofifibers with different Au loadings have been fabricated by a simple method of electrospinning followed by calcination in air. The particle size and plasmonic absorption of Au nanoparticles (NPs) loaded in the nanofifibers were analyzed and found to vary with the dosage of chloroauric acid in the precursor solution. The Au/CeO2 hybrid nano- fifibers were used as photocatalysts for selective oxidation of benzyl alcohol to benzaldehyde with O2 under simulated sunlight and visible light (>420 nm), respectively. The results showed that introducing Au NPs into CeO2 nanofifibers induced a great improvement in photocatalysis. The degree of improvement increased fifirst and then decreased with the increase in Au loading, reaching an optimal level over 0.5 wt.% Au-loaded CeO2 nanofifibers. The photocatalytic reaction presents a very high selectivity of 100% for benzaldehyde, which is important for organic synthesis. The transient photocurrent responses of the Au/CeO2 catalysts were also tested for corroborative evidence. After detailed discussion of various factors including plasmonic absorption, charge transfer, and surface activity of the photocatalyst, a possible mechanism for the photocatalytic oxidation of benzyl alcohol occurring at the Au–CeO2 interface was proposed. 

 2017 Elsevier Inc. All rights reserved

1. Introduction

Driving chemical reactions with sunlight instead of traditional heating methods is of great signifificance for energy conservation and environmental protection. It represents a kind of sustainable chemistry and can potentially be applied to various industrial chemical processes [1–3]. Thus, the photocatalytic technique is very attractive as one of the most important strategies of green chemistry for fuel production [4], chemical synthesis [5,6], and environmental remediation [7]. Exploiting photocatalysts with new and improved properties in terms of effificient solar harvesting and catalytic activity is an overarching concern in this fifield [8,9]. Among various oxide semiconductors, ceria (CeO2) is of special interest due to its many fascinating properties, such as abundant oxygen vacancy defects, a special Ce4+/Ce3+ redox cycle, high oxygen storage capability and oxygen mobility, good chemical stability, and biocompatibility [10]. In particular, nanoceria has exhibited outstanding catalysis in many reactions, such as water–gas shift reactions [11], CO oxidation [12], and selective oxidation [13], which benefifits from its capability of adsorbing and releasing oxygen by shuttling between Ce(IV) and Ce(III) redox cycles [14]. However, the photocatalytic activity of ceria is usually unsatisfying because the wide bandgap energy (
3.2 eV) and poor carrier conductivity restrict the effificiency of solar energy utilization and sunlight-driven chemical reactions [15]. Therefore, improving the photocatalytic activity of CeO2 is important for high effificiency of solar-driven chemical reactions.

In recent years, plasmonic photocatalysts based on the localized surface plasmon resonance (LSPR) of noble metal nanoparticles have gained increasing attention due to their great potential for improving the conversion of solar energy to chemical energy [16–18]. The unique LSPR properties of noble metal (e.g., Ag and Au) nanoparticles endow them with distinct advantages in absorbing and scattering light at specifific wavelengths across a wide range of the optical spectrum [19,20]. Moreover, noble metal nanoparticles have been used widely as catalysts in conventional heatdriven catalysis for organic transformations because of their well-known surface catalytic properties [21]. Specififically, gold nanoparticles (Au NPs) have been recognized as powerful catalysts for various oxidative reactions, as in the oxidation of alcohols, amines, and hydrocarbons and in the epoxidation of alkenes [22]. Sup ported Au NPs have been applied as effective catalysts, making them a highly popular research topic in heterogeneous catalysis [23]. As a result, Au NPs supported on semiconductors have raised great expectations for plasmon-enhanced photocatalysis, because of their strong light–matter interactions and well-known surface catalytic properties [24–26]. In Au nanocrystal–semiconductor coupling photocatalysts, the charge generation can be enhanced in the semiconductor by energy transfer from Au nanocrystals to the semiconductor, which is generally attributed to three mechanisms: LSPR-induced light focusing, hot electron/hole transfer, and plasmon-induced resonance energy transfer (PIRET) based on the near fifield [27,28]. Therefore, improved photocatalysis can be expected in Au/CeO2 hybrid nanostructures because of improvement in several key aspects, including increased carrier conductivity, visible light response triggered by plasmonic Au nanocrystals, and synergetic catalytic effects between CeO2 and Au NPs [29,30]. Some efforts have been made in preparing some Au/CeO2 composite photocatalysts [31,32], understanding their action mechanism [33], and exploring more applications [34].

Continuous one-dimensional (1D) nanofifibers of semiconductors are always attractive for use in solar energy conversion due to their large surface area, high charge carrier mobility, and easy assembly in devices [35]. Moreover, the photocatalysts of 1D nano- fifibers are easy to recycle in practical applications. Aiming at the preparation of CeO2 nanofifibers modifified with controllable Au NPs, we used the electrospinning technique, which plays an increasingly important role in fabricating various functional nano- fifibers [36]. Electrospun CeO2 nanofifibers have high surface area and porosity, through adding appropriate amounts of polymers that could subsequently be removed. Au NPs were uniformly incorporated into the mesoporous CeO2 fifibers by in situ conversion from the metallic precursors. The Au loading amount in the hybrid nano- fifibers could be regulated by varying the dosage of chloroauric acid in the precursor solution. The obtained Au/CeO2 hybrid nanofifibers were used as photocatalysts for selective oxidation of benzyl alcohol to benzaldehyde with O2 under simulated sunlight and visible (>420 nm) light, respectively. The photocatalytic activities of Au/ CeO2 hybrid nanofifibers with different loading amounts of Au NPs (i.e., 0.25, 0.5, 1.0, and 2.5 wt.%) were studied. The photoelectrochemical responses of the Au/CeO2 catalysts were also tested for corroborative evidence. Various inflfluencing factors including plasmonic absorption, charge transfer, and surface catalysis have been analyzed and discussed to provide a reasonable explanation about the photocatalytic performance of the Au/CeO2 nanofifiber catalysts.

2. Experimental

2.1. Materials

Cerous nitrate hexahydrate [Ce(NO3)3 6H2O], polyvinylpyrrolidone (PVP, Mw = 1,300,000), chloroauric acid tetrahydrate (HAuCl4-  4H2O), and N,N-dimethylformamide (DMF) were commercially obtained from Aladdin Industrial Corporation. All of them are analytical reagent. Acetonitrile, benzyl alcohol, and benzaldehyde were purchased from Shanghai Sinopharm Chemical Reagent Co., Ltd., and they are of chromatographic grade.

2.2. Materials preparation

The preparation of Au/CeO2 hybrid nanofifibers with different Au loadings were fabricated by a facile electrospinning process with following calcination in air. A precursor solution was prepared by dissolving 2.6054 g of Ce(NO3)3 6H2O and 1.6 g of PVP in 12 mL of DMF and adding a certain amount of HAuCl4 solution (0.1 M). The precursor solution was stirred for 24 h at room temperature and then loaded into a plastic syringe with a 21 gauge stainless steel needle. The syringe was placed in a syringe pump and the distance between the needle tip and the product collector was 15 cm. The positive voltage applied to the tip was 20 kV, and the solution feed rate was set to 0.5 mL/h. The ambient temperature and humidity are controlled at around 20  C and below 30%, respectively. The electrospun nanofifibers were then calcined in a mufflfle furnace that was kept at 200  C for 1 h and then heated at 500  C for 3 h to remove the organic components, with a heating rate of 2  C/min. Finally, the Au/CeO2 hybrid nanofifibers were obtained. The as-obtained Au/CeO2 samples containing different Au amounts were denominated as CeO2–xAu, where x represents the Au loading mass percent calculated from the experimental dosage of HAuCl4, as shown in Table S1.

2.3. Materials characterization

The phase composition and crystallinity were characterized by powder X-ray diffraction (XRD) patterns recorded on a Japan Shimadzu XRD-6000 diffractometer with monochromatized CuKa (k = 0.15418 nm) radiation. The morphologies and microstructures of the samples were observed by fifield emission scanning electron microscopy (FESEM, JEOL JSM-6700F, Japan) and transmission electron microscopy (TEM, JEOL-2010, Japan). HRTEM and HAADFSTEM images associated with the energy-dispersive X-ray (EDX) mapping spectra were carried out on a Hitachi S4800. The actual Au contents were measured by inductively coupled plasma optical emission spectra (ICP-OES) on a Perkin Elmer Optima 8300 optical emission spectrometer (USA). X-ray photoelectron spectroscopy (XPS) measurements were performed on a VGESCALAB MKII Xray photoelectron spectrometer with an exciting source of MgKa. The nitrogen adsorption/desorption measurements were performed using a Micromeritics ASAP 2020 V4.01 analyzer (USA) at 77 K. Diffuse reflflectance spectra (DRS) were recorded on a UV– vis Perkin Elmer Lambda 950 spectrophotometer using BaSO4 as the reference.

2.4. Photocatalytic aerobic oxidation of benzyl alcohol

The photocatalytic activity of the Au/CeO2 hybrid nanofifibers was evaluated by aerobic oxidation of benzyl alcohol under simulated sunlight (UV–visible) and visible light (k > 420 nm), respectively. The photocatalytic reactions were carried out in a transparent quartz test tube having an inner diameter of 1.5 cm and a length of 12 cm. The suspensions for photocatalytic reactions were prepared by dispersing the catalyst (1 mg) and injecting benzyl alcohol (10 lL) in acetonitrile (2 mL), sequentially. A balloon full of O2 at a pressure of
1 atm was used to seal the test tube and provide an oxygen atmosphere. Then the reaction solution was irradiated using a Xe lamp (CEL-HXF300, Beijing China Education Au-light Co., Ltd.) without or with a cutoff fifilter for experiments carried out under simulated sunlight (UV–vis) or visible light, respectively. After 5 h of irradiation, the reaction solution was centrifuged to remove the catalyst. The amounts of benzyl alcohol, benzaldehyde, and other byproducts in the solution were determined by GC analysis on a gas chromatograph (Jiedao TECH, GC1690A, China).

2.5. Photoelectrochemical tests

The photocurrent response of each sample was tested on a CHI 660D electrochemical workstation (Shanghai Chenhua, China) using a standard three-electrode quartz cell with a platinum wire as a counter electrode and a saturated calomel electrode as a reference electrode. The electrolyte was Na2SO4 solution (0.1 M) that was bubbled with nitrogen. A Xe lamp (CEL-HXF300) without and with a cutoff fifilter was used to provide the simulated sunlight and visible (k > 420 nm) light, respectively. To prepare a working electrode, 10 mg of the catalyst was ground and suspended in 1 mL of ethanol. The mixture was ultrasonically scattered for 30 min and then spin-coated on a pretreated 25  30 mm indium tin oxide (ITO) glass. After natural drying in air for 12 h, the catalyst fifilm on ITO glass was dried at 120  C for 5 h and fifinally used as the working electrode. The photocurrent–time curves were measured at the applied voltage of 0 V under intermittent irradiation with light and dark phases of 30 s.

3. Results and discussion

3.1. Au NP loadings in CeO2 nanofifibers

The Au/CeO2 hybrid nanofifibers used in this study were prepared by a simple electrospinning method followed by calcination in air. Various Au loading amounts (0.25–2.5 wt.%) in CeO2 nanofifi- bers were adjusted expediently by changing the dosage of chloroauric acid (Table S1) in the precursor solution during electrospinning. The actual numbers of loaded Au NPs in the hybrid nano- fifibers were measured by ICP analysis and are summarized in Table 1. The ICP-measured Au contents are very close to those calculated theoretically from the experimental dosages of chloroauric acid used in the preparation, indicating that almost all of the HAuCl4 in the precursor solution was completely transferred into Au NPs loaded in CeO2 nanofifibers. The loaded Au particle sizes also varied with their loading contents, which will be discussed below.

3.2. Compositions and microstructures of Au/CeO2 photocatalysts

The phase composition and crystallinity of all the catalysts were characterized by XRD patterns (Fig. 1). The CeO2–0.25Au and CeO2–0.5Au samples present the identical patterns to the pure CeO2 (CeO2–0Au), which can be indexed to a cubic flfluorite structure (JCPDS 04-0593). No peaks associated with Au species were detected in their XRD patterns, probably due to the low loading content as well as the high dispersion of Au NPs in CeO2 fifibers [37]. With increasing Au loading amount, the diffraction peaks of the face-centered cubic (fcc) Au phase (JCPDS No. 04-0784) are detectable in XRD patterns of CeO2–1Au and CeO2–2.5Au samples, besides those of the cubic CeO2 phase. Moreover, the diffraction peaks of metallic Au in the Au/CeO2 hybrid nanofifibers get stronger as Au loading content increases. Fig. 2a–e shows SEM images of the Au/CeO2 hybrid nanofifibers with various Au loading contents (0, 0.25, 0.5, 1, and 2.5 wt.%). Pure CeO2 and all of the Au/CeO2 hybrid nanofifibers present continuous 1D structure after calcination, with diameters in the range of 200–500 nm and lengths of several micrometers. Au NPs are well loaded on CeO2 nanofifibers, and they become more observable with increased loading from 0.25 to 2.5 wt.%. No obvious agglomeration of Au NPs is observed even after the 500  C high-temperature calcination, suggesting that CeO2 supports can prevent the agglomeration and particle growth of Au NPs. In fact, some Au NPs were incorporated into CeO2 nano- fifibers during the in situ formation process, but they are hardly observed by TEM (Fig. S1 in the Supporting Information) because of the large diameters of CeO2 fifibers. The Au NP size distribution was roughly estimated by inspection of a number of SEM images (Fig. 2f). Specififically, with an increase of Au loading from 0.25 to 2.5 wt.%, the average size of Au NPs grows from 37 to 107 nm. Obviously, the average particle sizes of Au NPs estimated from SEM images are much larger than the actual average sizes, because the Au NPs that are too small to observe have been ignored.

The crystalline phase and elemental composition of the Au/ CeO2 hybrid nanofifibers were further characterized by HRTEM imaging and EDX analysis on a CeO2–0.5Au nanofifiber (Fig. 3). Close observation by TEM imaging (Figs. 3a and S2 in the Supporting Information) reveals that the CeO2 support is constructed by random assembly of CeO2 crystallites with sizes of 5–10 nm and presents a mesoporous structure. Some darker nanoparticles with sizes of 10–40 nm are possibly Au NPs loaded in CeO2 nanofifibers. The HRTEM image (Fig. 3b) taken of this nanofifiber depicts two distinctive interplanar spacings of 0.32 and 0.24 nm, which correspond to the (111) planes of face-centered cubic CeO2 and Au, respectively. The EDX spectrum (Fig. 3c) recorded on this nanofifiber shows strong signals of Ce along with weak signals of Au. The elemental mappings (Fig. 3e–g) confifirm the distributions of O, Ce, and Au in the Au/CeO2 hybrid nanofifibers. Au NPs are well dispersed throughout the nanofifiber.

XPS measurements were performed typically on CeO2–0.5Au and CeO2–1Au catalysts to obtain further information about the surface atomic compositions and valence states of the Au/CeO2 catalysts. The survey spectra (Fig. S3 in the Supporting Information) exhibit obvious signals of Ce and O, and the signals of Au species are very weak due to the low loading contents. In Ce3d XPS spectra (Fig. 4a), the eight binding-energy positions denoted by ‘‘a” and ‘‘b” can be assigned to Ce3d3/2 and Ce3d5/2 states, respectively. The positions denoted as a1, a2, a4, b1, b2, and b4 are indexed to Ce4+, and the other two, denoted as a3 and b3, correspond to Ce3+, indicating that the Ce oxidation state is mainly Ce4+ in the samples [38]. The Au4f spectra (Fig. 4b) show two individual peaks at 83.8 eV for Au4f7/2 and 87.4 eV for Au4f5/2, respectively, verifying the nature of Au0 within both the samples. The XPS spectra of both CeO2–0.5Au and CeO2–1Au catalysts almost present identical Ce3d and Au4f signals. Therefore, the valence effect on catalytic activity of the present Au/CeO2 catalysts can be excluded.

3.3. Physicochemical properties of Au/CeO2 photocatalysts

The specifific surface areas and pore size distributions of the Au/ CeO2 hybrid nanofifibers are characterized by the N2 adsorption– desorption method, taking CeO2–0Au, CeO2–0.5Au, and CeO2–1Au as representative samples. The N2 adsorption–desorption isotherms and the pore size distributions of the three samples are shown in Fig. 5. All samples exhibit a type II isotherm [39] with a hysteresis loop typical of mesoporous materials (Fig. 5a, c, and e). The specifific surface area was calculated through the BET equation. Pure CeO2 nanofifibers (CeO2–0Au) present a surface area of 64.5 m2 g 1 . The surface areas of CeO2–0.5Au and CeO2–1Au are 63.6 and 60.4 m2 g 1 , respectively. The pore size distribution curves (Fig. 5b, d, and f) were obtained using the BJH method to the desorption branch of the isotherms. All the samples show the pore size distribution with peaks centered in the range of 2.0– 3.5 nm. These results indicate that textural properties of the nano- fifibers were not affected obviously by Au NP loading.

The optical absorption properties of the Au/CeO2 catalysts with different Au contents are characterized by UV–vis diffuse reflflectance spectra (DRS) (Fig. 6). Pure CeO2 nanofifibers show only the absorption band below 450 nm due to the intrinsic bandgap absorption of ceria. After Au NPs are loaded, a broad SPR absorption peak appears in 500–800 nm with its center at ca. 605 nm. The SPR absorption becomes more intensive as the Au loading increases from 0.25 to 2.5 wt.%. The increased visible light absorption from Au SPR should help enhance the harvesting of solar energy and promote related photocatalytic processes.

3.4. Photocatalytic selective oxidation of benzyl alcohol

Selective oxidation is an important reaction in organic synthesis and plays a signifificant role in the production of valuable chemicals. The application of heterogeneous photocatalysis and molecular oxygen to oxidation reactions promises to avoid the use of traditional, toxic chemical oxidants, offering a green and energyeffificient technology for organic synthesis [40,41]. Benzaldehyde is an important precursor for the production of perfumes, dyestuffs, and pharmaceuticals [13]. The selective aerobic oxidation of benzyl alcohol to benzaldehyde driven by visible light is of vital importance for being ecofriendly and economically effificient [42,43]. Therefore, we chose the selective oxidation of benzyl alcohol with molecular oxygen in acetonitrile as a model reaction (Fig. 7a), to examine the photocatalytic activity of the Au/CeO2 hybrid nanofifibers with various Au loadings. The reaction was carried out under either UV–visible light or visible light (>420 nm) from the Xe lamp, which has maximal emission intensity in the range from 500 to 600 nm. The conversion yields after 5 h of reactions over various photocatalysts under UV–visible and visible light are listed in Tables S2 and S3 of the Supporting Information. The photocatalytic reactions present a very high selectivity of 100% for benzaldehyde, which is important for organic synthesis. For a more intuitive comparison, the conversion yields have been normalized against the total mass of each photocatalyst and the reaction time and are illustrated by a histogram in Fig. 7b. Notably, under both UV–visible and visible light, the benzyl alcohol oxidation has been obviously promoted by introducing Au NPs. Under UV–visible light, the normalized conversion yield presents an increase fifirst as the loaded Au NPs increase. The highest value is achieved over 0.5 wt.% Au-loaded CeO2 nanofifibers (CeO2–0.5Au), which is nearly fourfold the value over pure CeO2 nanofifibers. When Au loading further increases to 1 wt.%, the normalized conversion yield decreases greatly, and decreases continuously with more Au loading in CeO2–2.5Au catalyst. Under visible light (>420 nm) irradiation, the normalized conversion also presents a humplike variation tendency with varied Au loading content. The CeO2–0.5Au catalyst shows the highest visible light photocatalytic activity, but pure CeO2 nanofifibers present very weak photocatalysis for benzyl alcohol oxidation under visible light. The normalized conversion yield increases with loaded Au amount and attains a maximum value of 392 lmol h 1 g 1 over CeO2–0.5Au catalyst, which is nearly 10-fold the value over pure CeO2 nanofifibers under visible light. For each catalyst, the normalized conversion yield obtained under UV–visible light is much higher than that under visible light because of the difference in available photon energy. Moreover, the TiO2 and SiO2 nanofifibers loaded with 0.5 wt.% Au NPs (TiO2–0.5Au and SiO2–0.5Au), which were prepared by the same method of electrospinning followed with calcination in air, have been used as photocatalysts for the selective oxidation of benzyl alcohol. By comparison, the photocatalysis of TiO2–0.5Au is obviously lower than that of CeO2–0.5Au, and SiO2–0.5Au shows the lowest photocatalysis among the Au-loaded nanofifibers. The results indicate that the combination of Au NPs and CeO2 support can generate a synergetic enhancement of the photocatalysis toward selective oxidation of benzyl alcohol. In addition, the recyclability of the CeO2–0.5Au photocatalyst was examined under the simulated sunlight and the visible light (k > 420 nm), respectively, in three cycles. The results are shown in Fig. 8, indicating that the photocatalyst has good stability and reusability in the photocatalytic reaction.

3.5. Photocurrent response of Au/CeO2 photocatalysts

To help understand the photo-response of Au/CeO2 hybrid nanofifibers, the transient photocurrent–time (Fig. 9) curves of various CeO2–xAu photoelectrodes were recorded under UV–visible and visible light. Fig. 9a indicates that the CeO2–xAu electrodes with lower Au loadings (x = 0.25–0.5) present current switches similar to that of pure CeO2 under intermittent irradiation with UV–visible light. When Au loading is increased to 1 and 2.5 wt.%, the CeO2–xAu electrodes exhibit a little spikelike shape of the transient current when the light is switched on or off, which may be due to charge storage and recombination in the system [44]. Fig. 9b shows the transient current–time curves of the CeO2–xAu (x = 0–2.5) electrodes under visible light (k > 420 nm) irradiation. As expected, pure CeO2 (CeO2–0Au) presents negligible photocurrent response because of its weak absorption in the visible light region. The CeO2–xAu (x = 0.25–2.5)-coated electrodes give obvious photocurrent responses under visible light, which is in accordance with the plasmonic absorption of loaded Au NPs (Fig. 6). However, the periodic on/off photocurrents of the CeO2–xAu (x = 1–2.5) electrodes with further increased Au loadings present a pronounced spikelike shape. After the initial rise of photocurrent immediately after the light is switched on, an obvious decay of photocurrent is observed due to recombination of surfacetrapped electrons with holes [44], suggesting that the charge recombination is more severe as Au loading increases.

3.6. Photocatalytic mechanism

From the above results, a further increase in Au loading from 0.5 to 2.5 wt.% does not induce continuous enhancement in the photocatalytic activity for benzyl alcohol oxidation over the Au/CeO2 catalysts, though the plasmonic intensity increases with Au loading. The declining photocatalytic performance upon excess Au loading (1–2.5 wt.%) is partly due to photon scattering and charge recombination, as suggested by transient photocurrents (Fig. 9). On the other hand, the size-determined surface catalysis of metal NPs should also be considered simultaneously [45]. Generally, smaller metal NPs have a higher fraction of coordinatively unsaturated surface atoms, which are highly catalytically active [46]. Therefore the smaller Au NPs would show higher surface activity toward catalytic oxidation of benzyl alcohol, which involves adsorption and activation of molecular oxygen as well as benzyl alcohol. To obtain a comprehensive understanding of the effect of Au loading on the photocatalytic performance for benzyl alcohol oxidation, the plasmonic absorption, average diameter, and specifific surface area of loaded Au NPs in CeO2 nanofifibers are shown in Fig. 10. The specifific surface areas of Au NPs were calculated by their average diameters (Table 1), taking the Au NPs as spheres (Eqs. S1–4 in the Supporting Information). As expected, the specifific surface areas of Au NPs greatly decrease with the continuous increase of Au loading from 0.25 to 2.5 wt.%, because of the increased Au NP sizes, leading to a decrease in the exposed active sites. However, the CeO2–0.25Au catalyst does not present the highest photocatalytic activity despite its having the largest Au surface area, because its plasmonic absorption is too weak. Therefore, the humplike variation in the photocatalytic performance of our Au/CeO2 catalysts is a result of the balance between plasmon resonance and surface catalytic activity of the loaded Au NPs.

As a consequence, a mechanism for the photocatalytic oxidation of benzyl alcohol over the Au/CeO2 nanofifibers is illustrated in Fig. 11 and explained as follows. First, the charge generation and transfer in the photocatalyst are shown in Fig. 11a. Under UV–vis irradiation, CeO2 can be excited to generate electron–hole pairs in its conduction band (CB) and valence band (VB). Meanwhile, the plasmonic Au NPs act as photon antennas that can effectively concentrate the visible light energy and transfer it to the adjacent CeO2 through near-fifield interaction, namely, the plasmon-induced near-fifield enhancement effect. The electron–hole pair formation in adjacent CeO2 will be strongly enhanced, generating improved photocatalysis. In this case, Au NPs also serve as electron trappers to improve the charge separation. When the Au/CeO2 photocatalyst is irradiated only visible light alone (k > 420 nm), CeO2 is barely excited to produce electron–hole pairs because of its wide band gap. The enhanced photocatalysis of Au/CeO2 under visible lightis mainly attributed to a plasmonic hot electron injection process [47], which allows the transfer of high-energy plasmonic electrons from Au NPs to CeO2 across the junction. Subsequently, the photocatalytic oxidation of benzyl alcohol occurred at the Au–CeO2 interface, shown in Fig. 11b. Both Au NPs and CeO2 specififically contributed to promote the photocatalytic reaction. Benzyl alcohol is ready to be adsorbed on Au NPs, forming a metal–alkoxide intermediate via O–H bond cleavage [48]. Oxygen molecules may be prone to be adsorbed on the CeO2 support because of its abundant oxygen vacancies [49] and then be activated by electron transfer. Finally, the activated oxygen captured b-hydrogen of the adsorbed benzyl alcohol, forming benzaldehyde and H2O as products. Thus, the photocatalytic oxidation of benzyl alcohol was completed via synergistic interactions between loaded Au NPs and CeO2 support. When Au loading and particle size increase, the plasmon-induced charge generation should be enhanced, but the surface active sites for catalytic reaction would greatly decrease because of the reduced Au surface area. As a result, the Au/CeO2 catalysts displayed a humplike variation tendency in their photocatalytic performance with increased Au loading amount.

4. Conclusions

Au/CeO2 hybrid nanofifibers with different amounts of Au NPs (0.25–2.5 wt.%) were prepared through electrospinning followed by calcinations in air. The particle size and plasmonic absorption of Au NPs loaded in the nanofifibers were determined by the dosage of chloroauric acid added in the precursor solution. The photocatalytic activity of Au/CeO2 hybrid nanofifibers were evaluated by selective oxidation of benzyl alcohol to benzaldehyde with O2 under simulated sunlight and visible light (>420 nm), respectively. Incorporation of Au NPs in CeO2 nanofifibers induced a great enhancement of photocatalysis for benzyl alcohol oxidation. The enhancement is related to the Au loading amount, getting an optimal level over 0.5 wt.% Au loaded CeO2 nanofifibers. The photocatalytic performance of Au/CeO2 hybrid nanofifibers is determined by multiple factors including plasmonic absorption, charge transfer, and surface catalysis. Finally, a mechanism for the photocatalytic oxidation of benzyl alcohol occurring at the Au–CeO2 interface was proposed on the basis of synergistic interaction between Au NPs and CeO2 support. Therefore, this work sheds some light on optimizing the photocatalysis of plasmonic metal/ semiconductor photocatalysts and provides a potentially impactful photocatalyst system for solar-driven chemical reactions as well.