Introduce oxygen vacancies into CeO2 catalyst for enhanced coke resistance during photothermocatalytic oxidation of typical VOCs

Journal Pre-proof Introduce oxygen vacancies into CeO2 catalyst for enhanced coke resistance during photothermocatalytic oxidation of typical VOCs Jiejing Kong (Methodology) (Writing - original draft), Ziwei Xiang (Data curation), Guiying Li (Writing - review and editing), Taicheng An (Conceptualization) (Supervision)

PII:

S0926-3373(20)30170-3

DOI:

https://doi.org/10.1016/j.apcatb.2020.118755

Reference:

APCATB 118755

To appear in:

Applied Catalysis B: Environmental

Received Date:

6 January 2020

Revised Date:

8 February 2020

Accepted Date:

11 February 2020

Please cite this article as: Kong J, Xiang Z, Li G, An T, Introduce oxygen vacancies into CeO2 catalyst for enhanced coke resistance during photothermocatalytic oxidation of typical VOCs, Applied Catalysis B: Environmental (2020), doi: https://doi.org/10.1016/j.apcatb.2020.118755

This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition of a cover page and metadata, and formatting for readability, but it is not yet the definitive version of record. This version will undergo additional copyediting, typesetting and review before it is published in its final form, but we are providing this version to give early visibility of the article. Please note that, during the production process, errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain. © 2020 Published by Elsevier.

Introduce oxygen vacancies into CeO2 catalyst for enhanced coke resistance during photothermocatalytic oxidation of typical VOCs

Jiejing Kong, Ziwei Xiang, Guiying Li, Taicheng An*

ro of

Guangdong Key Laboratory of Environmental Catalysis and Health Risk Control, Guangzhou Key Laboratory Environmental Catalysis and Pollution Control, School of Environmental

-p

Science and Engineering, Institute of Environmental Health and Pollution Control,

lP

re

Guangdong University of Technology, Guangzhou 510006, China

Corresponding author: Prof. Taicheng An

na

Tel: 86-20-39322298

Jo

ur

E-mail: [email protected]

Graphic Abstract

1

ro of -p

Highlights

Ordered porous ARCeO2 with rich oxygen vacancies and low acidity was synthesized



Oxygen vacancies accelerates Mars–van Krevelen redox cycle



Low acidity enhances coke resistance of ARCeO2 during photothermocatalytic process



ARCeO2 demonstrated excellent photothermocatalytic activity and stability

ur

ABSTRACT

na

lP

re



Carbon deposit during the catalytic oxidation of volatile organic compounds (VOCs) is a key

Jo

factor to restrict long-term stability of catalysts. Herein, a modified ordered porous catalyst with high coke-resistance performance was developed by introducing oxygen vacancies (OVs) into CeO2 through simple redox and steam treatment (ARCeO2), and applied in efficiently photothermocatalytic degradation of VOCs. Abundant OVs in ARCeO2 leads to enhanced light absorption, improved charge separation and increased reactive oxygen 2

generation, and then promote its photothermocatalytic performance. Moreover, no obvious decrease was observed in the photothermocatalytic activity of ARCeO2 at 200 °C for 25 h reaction. Its superior photothermocatalytic stability can be attributed to the less intermediates and limited coke accumulation onto ARCeO2. This work provides a facile and cost-effective strategy for the design and fabrication of CeO2 catalyst with strong coke-resistance potential and highlights the significance of OVs engineering for improving the photothermocatalytic

ro of

performance of VOCs degradation.

Keywords: Photothermocatalytic degradation; VOCs; ARCeO2; Oxygen vacancy; Coke

-p

resistance

1. Introduction

re

Volatile organic compounds (VOCs), such as aromatic hydrocarbons straight-chain

lP

alkane and cycloalkanes, are typical gaseous pollutants emitted from refinery and electronic waste dismantling processes, and then directly cause harm to human health and the

na

environment [1, 2]. Both photocatalytic oxidation and thermocatalysis reaction onto metal oxide catalysts are suggested to be the highly effective and economical techniques for VOCs

ur

removal [3-6]. However, most metal oxide catalysts suffer from low photocatalytic efficiency

Jo

in the abatement of aromatics and alkane, because benzene ring and C-C bond are hard to be cracked [5]. For example, only 34.2% of styrene [7] and 30% n-hexane [8] can be degraded over TiO2 in the continuous flow fixed bed reactor with full spectrum and UV illumination, respectively. Besides, high temperature (> 500 °C) is always required to effectively eliminate the saturated alkanes onto metal oxide catalysts (such as Fe2O3 [9]), leading to high energy costs. The development of metal oxide catalysts and the enhancement of their catalytic 3

reaction processes for VOCs removal at low temperature is still a challengeable topic. Ceria (CeO2) with a band gap of ∼3.2 eV has been widely investigated as a photo/thermocatalyst because of its superior properties, for instance, easy Ce3+/Ce4+ redox cycle, high oxygen mobility at moderate temperature, and high chemical/photochemical stability [10, 11]. However, the accumulation of intermediates and even carbon deposits always inevitably occur during the catalytic processes, which would poison the active site of

ro of

the catalyst for rapid decline in its catalytic activity [4, 12]. The main reason of the formation of intermediates and coke deposits is that the inadequate generation of reactive oxygen species causes incomplete oxidation of VOC molecule instantly [13]. Recently, combining

-p

thermocatalysis with photocatalysis to maximize synergistic effects has been proven to be an

re

effective method to promote the formation of more reactive oxygen species [14-17]. Primarily, the photothermocatalytic oxidation processes onto CeO2 for VOC removal

lP

comply with Mars-van Krevelen mechanism [17], which can be summarized as follows: (1) the adsorbed VOCs molecules are oxidized with reactive oxygen species (such as O2-, O-, h+,

na

•OH and •O2−) generated onto CeO2 by both thermal and photo excitation, along with the

ur

reduction of CeO2 to produce CeO2-x; (2) the reduced CeO2-x is re-oxidized to CeO2 by gaseous O2 and the reactive oxygen species generated under photoexcitation. It has been

Jo

widely accepted that from the activating oxygen to reactive oxygen species is the rate-determining step, and oxygen vacancies (OVs) can not only act as an electron donor to effectively capture gaseous oxygen and strengthen the mobility of lattice oxygen [18, 19], but also accelerate the separation and migration of photogenerated charge for facilitating the activation of oxygen [20]. Therefore, introduce OVs into CeO2 can further promote 4

photothermocatalytic degradation efficiencies of VOCs. Crafting OVs can be usually realized by partial reduction of CeO2 into the non-stoichiometric form CeO2-x with methods such as (1) thermal treatment at reductive atmosphere such as H2, CO, NH3, vacuum and NaBH4 (Reductant + Olattice → OVs + Oxidized reductant); (2) electrochemical reduction; and (3) photolysis by ultraviolet radiation (Light + Olattice → OV + 1/2O2) [21]. It is noted that OVs sites also serve as activated acid sites, where the imbalance of

ro of

atomic electronegativity leads to a greatly increase in acidity and then improves the catalytic activity of bond breaking, polymerization and isomerization [22]. As a result, carbon deposits are prone to take place at OVs sites with their strong acidity during the catalytic oxidation

-p

process of VOCs. Much research has suggested that doped with alkali/alkali-earth metal or

re

metal oxide [23], high-temperature calcination [24], steam pre-treatment [22] and pre-carbon-deposit treatment [25] could significantly reduce the catalyst surface acidity and

lP

amount of acid sites, but always accompanied by the decline of OVs. The trade-off between

na

acidity and OVs in catalyst is a serious concern for its subsequent catalytic performance during the catalytic reaction processes.

ur

Herein, in order to attain high catalytic activity and good stability performance for degradation of VOCs, a modified CeO2 ordered porous catalyst with high coke-resistance

Jo

potential was developed and applied to the photothermocatalytic degradation of typical VOCs under simulated solar irradiation. The abundant OVs with weak acidity was introduced into modified CeO2 and further characterized by XPS, EPR, Raman and NH3-TPD. The results found that the catalytic activity of this newly prepared catalyst can effectively promote light harvesting, charge separation, reactive oxygen generation and reducibility, which then 5

subsequently improved the photothermocatalytic degradation performance of VOCs. Characterization

of

intermediate

accumulation

and

carbon

deposit

during

the

photothermocatalytic degradation of VOCs have also been conducted to understand the high performance as well as good stability onto this modified CeO2 ordered porous catalyst.

2. Experimental

ro of

2.1. Preparation of catalysts CeO2 ordered porous catalyst was prepared by a colloidal crystal template method reported in our previous work [16], then treated under the three different conditions

-p

respectively: (I) Annealed it in the N2 atmosphere at 400 °C for 3 h, then turned the flow of

re

air/N2 (volume ratio = 1:1) for another 2 h and then get obtained product which was denoted as CeO2; (II) Reduced it in the H2/N2 (volume ratio = 1:1) atmosphere at 400 °C for 2 h and

lP

then annealed it in N2 atmosphere at 400 °C for 1 h, followed by calcining in the air/N2 (volume ratio = 1:1) atmosphere at 400 °C for another 2 h to obtain the final product denoted

na

as RCeO2; (III) Reduced it in the humid H2/N2 (volume ratio = 1:1) atmosphere at 400 °C for 2 h (humid condition was achieved with N2 flow into the ultrapure water at 30 °C, relative

ur

humidity = 40%), followed by annealing in the N2 atmosphere at 400 °C for 1 h, then calcined

Jo

it in the air/N2 (volume ratio = 1:1) atmosphere at 400 °C for another 2 h, and the product obtained was denoted as ARCeO2. 2.2. Catalysts characterization The morphology, structure, and composition of the materials were characterized by transmission electron microscopy (TEM, 200 kV, Talos-F200S, FEI), X-ray diffraction (XRD, 6

D8 ADVANCE, Bruker) and X-ray photoelectron spectroscopy (XPS, Escalab 250Xi, Thermo Fisher). The optical properties of the catalysts were measured on UV-vis spectrophotometer (DRS, Agilent Cary 300) and fluorescence spectrophotometer (PL, Fluorolog-3, HORIBA Instruments Incorporated) with a xenon lamp (excitation wavelength 280 nm) as light source. The electron paramagnetic resonance (EPR) tests were carried out in the X-band (9.84 GHz) with 4.00 G modulation amplitude and a magnetic field modulation of 100 kHz using a EPR

ro of

spectrometer (EMXPlus-10/12, Bruker) at 77 K. Raman scattering measurements were recorded on a confocal Raman Spectroscopy (LabRAM HR Evolution, HORIBA Jobin Yvon), and the wavelength was set as 532 nm. Temperature-programmed reduction of H2 (H2-TPR),

-p

temperature-programmed oxidation of O2 (O2-TPO) and temperature-programmed desorption

re

of NH3 (NH3-TPD) were investigated using TP-5078 Autochem-absorption analyzer, with approximately 20 mg of catalysts loaded in a quartz tube. Prior to each test for H2-TPR, the

lP

catalyst was pretreated under high-purity N2 at 30 °C for 1 h, then heated from 30 °C to

na

780 °C at 12 vol.% H2/N2 atmosphere with a flow rate of 30 mL min-1 in the dark or under simulated solar irradiation. The illumination condition was achieved by equipping the furnace

ur

with a small window on its one side and 300 W Xe lamps on the outside. The consumption of H2 was detected using a thermal conductivity detector (TCD) (Xianquan Industrial and

Jo

Trading Co., Ltd.); For O2-TPO experiment, 20 mg of the samples were pretreated at 30 °C for 1 h under high-purity Ar flow, then exposed to 40 vol.% O2/Ar with a flow rate of 30 mL min-1. O2-TPO profile was recorded from 30 °C to 800 °C at a heating rate of 10 °C min-1 and O2 consumption was measured using a TCD detector; Prior to each NH3-TPD run, the catalyst was pretreated under high-purity N2 at 90 °C for 1 h, then turned the flow of 10 vol.% NH3/N2 7

into the system with a flow rate of 30 mL min-1 for 1 h. After that, the catalyst was flushed with high-purity N2 to remove physically adsorbed NH3 on the catalyst surface. Finally, the system was heated to 700 °C at a heating rate of 10 °C·min-1, and the NH3 desorption profiles were recorded under N2 flow by TCD detector. 2.3. Activity test Photothermocatalytic degradation reactions were conducted in a tube (i.d. = 6 mm)

ro of

fixed-bed reactor at atmospheric pressure, and the catalyst can be irradiated through a small window on the one side of the furnace, as schematized in Figure S1 of Supporting

-p

Information. The simulated solar irradiation came from 300 W Xe lamp (λ = 300−780 nm, optical power density is 200 mW cm-2). Approximately 0.1 g catalyst was packed in the

re

reactor. The vapor of styrene, n-hexane or cyclohexane was generated by passing N2 at a

lP

certain flow rate through the corresponding solution (≥99%), and the vapor was mixed with dry air to achieve different VOCs concentration. The final VOCs concentration of ∼50 ppm

na

was fixed in the feeding gas by controlling the solution temperature and N 2 flow rate through a saturator. Gas hourly space velocity (GHSV) value of 30,000 mL h−1 g−1 was tested. The

ur

concentrations of VOCs and CO2 were respectively measured using the gas chromatography

Jo

GC-9800 (FID) and GC9800 (FID, equipped with a Ni catalyst based methanizer), and detail analysis process was given in Supporting Information. Three to five concentration data were measured at each temperature with around a 10 min interval. The concentration data adopted showed less than 10% difference from other measured data at each temperature. The mineralization/degradation of VOCs was assessed by the following equations: 𝑆𝑡𝑦𝑟𝑒𝑛𝑒 𝐷𝑒𝑔𝑟𝑎𝑑𝑎𝑡𝑖𝑜𝑛 % =

[𝐶8 𝐻8 ]𝑖𝑛 −[𝐶8 𝐻8 ]𝑜𝑢𝑡 [𝐶8 𝐻8 ]𝑖𝑛

8

× 100%

(1)

𝑆𝑡𝑦𝑟𝑒𝑛𝑒 𝑀𝑖𝑛𝑒𝑟𝑙𝑖𝑧𝑎𝑡𝑖𝑜𝑛 % =

[𝐶𝑂2 ]𝑜𝑢𝑡 8

[𝐶8 𝐻8 ]𝑖𝑛

𝑛 − 𝐻𝑒𝑥𝑎𝑛𝑒 𝐷𝑒𝑔𝑟𝑎𝑑𝑎𝑡𝑖𝑜𝑛 % =

× 100%

(2)

[𝐶6 𝐻14 ]𝑖𝑛 −[𝐶6 𝐻14 ]𝑜𝑢𝑡

𝐶𝑦𝑐𝑙𝑜ℎ𝑒𝑥𝑎𝑛𝑒 𝐷𝑒𝑔𝑟𝑎𝑑𝑎𝑡𝑖𝑜𝑛 % =

[𝐶6 𝐻14 ]𝑖𝑛

× 100%

[𝐶6 𝐻12 ]𝑖𝑛 −[𝐶6 𝐻12 ]𝑜𝑢𝑡 [𝐶6 𝐻12 ]𝑖𝑛

× 100%

(3)

(4)

where [𝐶8 𝐻8 ]𝑖𝑛 , [𝐶6 𝐻14 ]𝑖𝑛 and [𝐶6 𝐻12 ]𝑖𝑛 are the inlet molar quantity of vaporous styrene,

ro of

n-hexane and cyclohexane, respectively; [𝐶8 𝐻8 ]𝑜𝑢𝑡 , [𝐶6 𝐻14 ]𝑜𝑢𝑡 , [𝐶6 𝐻12 ]𝑜𝑢𝑡 and [𝐶𝑂2 ]𝑜𝑢𝑡 are the outlet molar quantity of vaporous styrene, n-hexane, cyclohexane and CO2, respectively.

-p

2.4. The identification of Intermediates

re

The trace intermediates deposited onto the spent catalyst surface were extracted ultrasonically with 60 mL chromatographic grade methanol for 5 min and then stored

lP

overnight in the fridge (-18 °C). The extract was filtered through a 0.22 μm filter membrane and concentrated with a vacuum rotatory evaporator, then completely dried with a gentle

na

stream of high-purity N2. Afterwards re-dissolved it with 1.0 mL ethyl acetate and injected directly into a gas chromatography-mass spectrometer (GC-MS, Agilent 7890B-5977B) for

ur

the identification of intermediates, and detail analysis process was given in Supporting

Jo

Information.

3. Results and discussion 3.1. Structural properties CeO2, RCeO2 and ARCeO2 catalysts were obtained by the modification of raw CeO2, including directly calcination, redox calcination, redox calcination with steam treatment, 9

respectively. As shown in Figure 1a-c, TEM images of the three samples exhibit the ordered texture with macropores (120 nm) and uniformly sized walls interconnected with small windows. Besides, XRD patterns in Figure 1d prove that CeO2, RCeO2 and ARCeO2 are all composed of cubic cerianite structure (JCPDS No. 34-0394). No significant difference has been observed in either TEM or XRD spectra for the three resultant catalysts, suggesting the similar morphology, crystal phase, crystallinity, and particle size of the three samples obtained

ro of

under different atmosphere-calcinations as described in the reference [26].

XPS experiment was conducted to explore the valence states of elements and the types of oxygen species over CeO2, RCeO2 and ARCeO2, as presented in Figure 2a and b. For Ce 3d

-p

spectra of these samples (Figure 2a), all of them are fitted by eight bands marked as α1, α2, α3,

re

α4 and β1, β2, β3, β4. Moreover, the bands of α2 and β2 are resulted from Ce3+ ions while the others are corresponded to Ce4+ species according to the references [15, 16]. It is reported that

lP

a rise in Ce3+ favors the increase in the amount of the OVs and the relatively high mobility of the bulk oxygen species, then facilitates catalytic oxidation reaction [27]. It is generally

na

accepted that the area ratio of α2+β2 bands to the total eight bands can reflect Ce3+ ratio on the

ur

surface of the samples [28]. Consequently, the relative amount of Ce3+ in the samples was estimated based on the above area ratio, and the results are listed in Table S1. The ratio of

Jo

Ce3+ over these samples was in the order of ARCeO2 (21.7%) > RCeO2 (20.6%) > CeO2 (18.4%). As the migration and transformation of oxygen species over catalysts are also important for the catalytic reaction, O 1s spectra of CeO2, RCeO2 and ARCeO2 have been investigated and are given in Figure 2b. All the O 1s XPS spectra of these samples can be fitted into three Gaussian peaks at 531.7 and 529.2 eV, assigned to surface adsorbed oxygen 10

species (Oα) and lattice oxygen species (Oβ), respectively [29, 30]. As the migration ability of surface adsorbed oxygen species (including chemisorbed oxygen O2-, O- and hydroxyl group) is remarkably stronger than that of lattice oxygen species, the adsorbed oxygen is viewed as the dominant reactive oxygen species and plays a dominant role in the redox reaction [31]. In other words, the ratio of Oα/(Oα+ Oβ) could reflect the content of reactive oxygen species of the reaction [30]. Table S1 shows that the Oα/(Oα+Oβ) ratio has a sequence of ARCeO2

ro of

(31.4%) > RCeO2 (28.6%) > CeO2 (18.0%). As the surface adsorbed oxygen species mainly exists on the defect sites (e.g., OVs) of catalysts, it can be suggested that the reduction of Ce 4+ to Ce3+ in ARCeO2 after redox and steam treatment facilitates the formation of OVs, further

-p

promoting the oxygen escaping from the lattice sites and leaving vacant sites with two

re

electrons at each vacancy, then generates more reactive oxygen.

To further analyze Ce species states and OVs in CeO2, RCeO2 and ARCeO2, EPR

lP

characterization was also performed and summarized in Figure 2c. The signals marked by asterisks correspond to Mn2+ ion impurity, which is due to the quartz tube of EPR equipment

na

[32]. The signal occurred at g=2.002 represents the superoxide anions (•O2−) attached with the

ur

Ce4+ ions [16, 32], and its intensity mainly associated with OVs concentration [33]. Besides, the g value of 1.963 corresponds to the Ce3+ species next to unpaired electrons (O2− and O−)

Jo

[32-34]. The intensity of Ce3+ and OVs were both in the order of ARCeO2 > RCeO2 > CeO2, demonstrating that the redox and steam treatment reduced Ce4+ to Ce3+ in the CeO2 lattice then generated more OVs. Furthermore, Raman spectroscopy was performed to further verify the lattice defects of these samples, as shown in Figure 2d. Over CeO2, the Raman spectra present a only 11

characteristic peak at around 464 cm-1, corresponding to the F2g vibration mode of octahedral local symmetry around cubic fluorite structure of CeO2 [33]. For RCeO2 and ARCeO2, two new Raman scattering peaks could be observed with a broad shoulder at 602 cm-1 and one weak band at 252 cm-1. Both peaks are ascribed to the defects-induced vibration mode, especially oxygen defects [28, 33]. The difference in wavenumber between the oxygen defect (602 cm-1) and F2g bands (464 cm-1) is ∼140 cm−1, suggesting that OVs of RCeO2/ARCeO2

ro of

catalyst are generated from intrinsic defects (Ce4+→Ce3+) [30]. It is generally considered that the intensity ratio between defects-induced vibration mode and F2g mode (i.e., I602/I464) is proportional to the concentration of OVs in the CeO2-based materials [35]. The value of

-p

I602/I464 for ARCeO2 (ca. 0.1723) is slightly higher than that for RCeO2 catalyst (ca. 0.1571),

re

while the characteristic peaks of OVs are absent over CeO2 due to its low OVs concentration. This phenomenon is consistent with XPS and EPR results above.

lP

It is well known that the surface acidity of catalysts is not only an indispensable factor

na

for carbon deposit, but also for catalytic performance of VOCs oxidation [36]. Thus, to characterize the catalyst surface acidity, the NH3-TPD of these three catalysts was illustrated.

ur

As presented in Figure 3, two NH3 desorption peaks were observed over all samples, with the peak appearing at low temperature (100-250 °C, α peak) corresponded to the weak acid sites

Jo

and the high temperature peak (>300 °C, β peak) assigned to strong acid sites [29]. Compared with CeO2, α peak of both RCeO2 and ARCeO2 shifted from 190 °C to 152 °C with stronger desorption intensity, indicating a decline of acid strength but increase of weak acid quantity. Besides, the desorption temperature of β peak over CeO2, RCeO2 and ARCeO2 is 438 °C, 438 °C and 385 °C, respectively, and the latter two showed stronger signal than that of CeO2. 12

All these results reveal that both RCeO2 and ARCeO2 possess more acid sites than that of CeO2. In comparison with RCeO2, steam treatment for ARCeO2 did not changed the amount of strong acid sites but weaken its acidity, which may have positive effects on the adsorption of VOCs

molecules

[29,

37], simultaneously enhanced coke

resistance during

photothermocatalytic oxidation of typical VOCs [38]. Based on the above results, we know that when CeO2 ordered porous catalyst was

ro of

modified by redox treatment, in which firstly reduced by calcination in H2 atmosphere at 400 °C, followed by oxidized in the air atmosphere at 400 °C, OVs could be successfully introduced into CeO2 catalyst (denoted as RCeO2) by partial reduction of Ce4+ to Ce3+ [21, 39].

-p

The as-introduced OVs would induce a greatly increase in the surface acid sites of catalyst

re

due to its imbalance of atomic electronegativity [22]. In this context, the rich OVs and abundant acid sites were observed on RCeO2 (Figures 2d and 3). It is reported that steam

lP

pre-treatment of catalyst could significantly reduce its acidity and amount of acid sites [22].

na

After modification of CeO2 by combining redox process with steam treatment, surface acidity of the obtained catalyst (ARCeO2) has weakened, as indicated by NH3-TPD profiles (Figure

ur

3). Moreover, as the modifying temperature (400 °C) is much lower than the preparation temperature (600 °C) of raw CeO2, the morphology and crystal phase of ARCeO2 show no

Jo

observed change (Figure 1c and d). Also, ARCeO2 catalyst maintains the rich OVs, which is confirmed by XPS and Raman characterization (Figure 2). All these phenomena imply that conducting the redox and steam treatment on simultaneously could successfully introduce OVs into CeO2 catalyst and synchronously weaken its surface acidity, demonstrated the advantages of our modification method. 13

3.2. Catalytic performance The influence of both OVs concentration and surface acidity of catalysts on its photothermocatalytic degradation activity of three typical VOCs (including styrene, n-hexane and cyclohexane) is investigated and compared in Figure 4. As shown in Figure 4a, the photothermocatalytic degradation of styrene under simulated solar irradiation without catalyst below 250 ºC can be negligible, and the activity profiles demonstrate that the catalytic activity

ro of

varies with both OVs concentration and acidity of catalysts. ARCeO2 exhibited the best initial photocatalytic performance (30 °C) and highest thermocatalytic activity among the three samples (Figures 4 and S2). When under photothermocatalytic conditions (> 30 °C), ARCeO2

-p

also demonstrates the best apparent performance with the 30% styrene degradation

re

temperature (T30) at 43 °C, 50% styrene degradation temperature (T50) at 136 °C and 90% styrene degradation temperature (T90) at 226 °C (Figure 4a and Table S2). Particularly, T30 of

lP

ARCeO2 shows much lower than that of RCeO2 (104 °C) and CeO2 (160 °C). Moreover,

na

compared with ARCeO2 in the dark (T30,dark at 176 °C, T50,dark at 206 °C and T90,dark at 245 °C), ARCeO2 under simulated solar irradiation exhibited much more improved catalytic

ur

activity, especially at the low temperature region (T < 200 °C). It is noted that thermocatalytic activity of ARCeO2 (11% styrene degradation at 100 °C) in the low temperature is much

Jo

weaker than the initial activity in photocatalysis (28% styrene degradation at 30 °C), so it can be speculated that photocatalysis plays a key role at a low reaction temperature (30−100 °C) of photothermocatalytic process (Figure 4a). When in the relatively high temperature (150−200 °C), a photothermocatalytic synergistic performance of ARCeO2 is demonstrated, which shows remarkably higher styrene degradation than the contribution addition from the 14

pristine photocatalysis and thermocatalysis (Figure 4a). Besides, by proper processing data in Figure 4a according to the reference [16], the Arrhenius plots could be drawn and Ea values could be obtained for each sample from the slope of the Arrhenius plots, as presented in Figure 4b. The estimated Ea value was around 15.6 kJ mol-1 for ARCeO2 under simulated solar irradiation, which was observably lower than those of ARCeO2 in the dark (48.2 kJ mol-1), RCeO2 (20.9 kJ mol-1) and CeO2 (36.8 kJ mol-1) under simulated solar irradiation. This

ro of

result indicates that the redox and steam treatment, together with the existence of a photothermocatalytic synergetic effect onto ARCeO2 can greatly decrease the activation energy of styrene oxidation. Lower activation energy remarkably reduces the reaction energy

-p

barrier and shows more conductive for the catalytic reaction to proceed smoothly [40].

re

To further explore the applicability of the as-developed catalyst with introduced OVs and surface acid sites, the photothermocatalytic degradation performance of ARCeO2 for n-hexane

lP

and cyclohexane was also checked (Figure 4c and d), and the comparison results of the three

na

samples were summarized in Table S2. Obviously, among all the samples, ARCeO2 also performs the highest apparent activity in photothermocatalytic degradation of both n-hexane

ur

and cyclohexane. The 20% n-hexane degradation temperature of ARCeO2 is obtained as 142 °C, showing 98 °C and 172 °C lower than that for RCeO2 and CeO2, respectively.

Jo

Moreover, the 20% cyclohexane degradation temperature of ARCeO2 at 213 °C, also exhibits significantly lower than that of RCeO2 (333 °C) and CeO2 (432 °C). All these results reveal that OVs and surface acid sites in ARCeO2 can greatly promote the photothermocatalytic degradation of various typical VOCs, such as benzene series, straight-chain alkane and cycloalkanes. 15

3.3. Photothermocatalytic mechanism of catalyst To gain insight into the photothermocatalytic mechanism enhanced by ARCeO2, its optical properties including light response and charge separation are evaluated and illustrated first in Figure 5. All the samples exhibit similar UV-vis diffuse reflectance spectra (DRS) and the UV-vis absorption intensity ranked by ARCeO2 > RCeO2 > CeO2 (Figure 5a). It is widely accepted that OVs can largely modify the optical properties of metal oxides as the

ro of

OVs-induced defect states are involved in new photoexcitation processes [41]. The higher OVs concentration in ARCeO2 exhibited the stronger light absorption both in UV and visible-light region (Figure 5a). Besides, two absorption bands at 241 and 320 nm for all the

-p

three catalysts can be observed and were ascribed to charge transfer transitions: O2−→Ce3+

re

and O2−→Ce4+, respectively [34]. The main absorption edges of the three samples are around 430 nm and the corresponding optical band gaps (Eg) are estimated to be 2.9 eV. The valence

lP

band edges obtained from XPS valence band spectra of CeO2, RCeO2 and ARCeO2 are

na

estimated to be ca. 2.5, 2.3 and 2.3 eV, respectively, implying that redox treatment slightly changed the valence band structure of CeO2 (Figure S3). In combination with the estimated Eg

ur

values, their conduction band values are estimated to be ca. -0.4 for CeO2 and -0.6 eV for both RCeO2 and ARCeO2. The result reveals that OVs can elevate the conduction band position,

Jo

and the more negative potentials of conduction band can supply a strong potential for adsorbed oxygen reduction to generate more reactive oxygen species, according to the reference [6, 42]. In addition, charge recombination was also analyzed by photoluminescence spectroscopy, as exhibited in Figure 5b. ARCeO2 shows the weakest PL intensity, indicating that ARCeO2 16

has lower charge recombination but higher charge separation than that of CeO2 and RCeO2. Since the surface OVs can prevent the charge carrier from recombination by trapping the electrons via their defect states, while it is more beneficial for these charges to transfer to the reactants on the catalyst surface efficiently [43, 44], ARCeO2 with OVs-rich demonstrated the highest charge separation capability among the three catalysts, which would effectively promote the photothermocatalytic reaction.

ro of

It is reported that photothermocatalytic oxidation mechanism of VOCs over CeO2-based catalysts can be interpreted in terms of the Mars–van Krevelen reduction–oxidation pathway [16], and H2-TPR analysis is an ideal tool for measuring the reduction behavior of the

-p

catalysts [45]. In this study, the H2-TPR profiles of the catalysts in the dark and under

re

illumination (λ = 300−780 nm) was performed and displayed in Figure 6. CeO2 under illumination shows two broad peaks, one at 418 ºC and another at 522 ºC. The first one is

lP

related to the reduction of the surface capping oxygen species of CeO2, and the second is

na

associated with the reduction of Ce4+ to Ce3+ ions in the outermost layers (surface lattice oxygen) [27, 45]. On ARCeO2 and RCeO2 under illumination, both peaks shift to lower

ur

temperatures in comparison with CeO2. Moreover, compared with ARCeO2 in the dark, the second reduction peak of ARCeO2 under illumination shifted to lower temperature range, and

Jo

the H2 consumption of the first peak apparently increased. Overall, the reducibility and mobility of surface oxygen species are generally improved over ARCeO2 under illumination, since the enrichment of OVs and photoexcitation prompt the adsorption/migration of gaseous/lattice oxygen, and then accelerate the activation of gaseous/lattice oxygen [20]. As the catalytic oxidation process could be facilitated by the high reducibility and diffusion of 17

surface gaseous/lattice oxygen, ARCeO2 under photothermocatalytic conditions gives better catalytic performance for the degradation of VOCs. As mentioned above, the oxidation of VOCs over CeO2-based catalysts under photothermocatalytic conditions has been generally interpreted in terms of the Mars-van Krevelen mechanism. According to this mechanism, the first step involves the reduction of CeO2 with the adsorbed VOC molecules, in which the adsorbed VOCs are simultaneously

ro of

oxidized with reactive oxygen species (such as O2-, O-, h+, •OH and •O2−) generated onto CeO2 by both thermal and photo excitation, and partial CeO2 is reduced to CeO2-x. The second step is to re-oxidize the reduced CeO2-x to CeO2 by both gaseous O2 and the photogenerated

-p

reactive oxygen species. Thus, the reducibility of CeO2 and capacity of reactive oxygen

re

species generation determine the photothermocatalytic activity of CeO2. As discussed above, OVs introduced into CeO2 catalyst can effectively promote the adsorption/migration of

lP

gaseous/lattice oxygen, leading to the generation of more reactive oxygen species and lower

na

H2-TPR temperature peak onto ARCeO2 as compared with that of CeO2 (Figures 2b and 6). In addition, photoexcitation also improved the reducibility of ARCeO2, as indicated by H2-TPR

ur

characterization (Figure 6). Moreover, ARCeO2 shows the strongest capacity of UV-vis absorption and the lowest charge recombination among the three samples, as confirmed by

Jo

DRS and PL spectra (Figure 5). The redox treatment for ARCeO2 induces a significant negative shift of the conduction band to enhance the reductive potential, and then supply a strong potential adsorbed oxygen reduction [42, 46]. Both the unique optical properties and band structure of ARCeO2 are conducive to the generation of reactive oxygen species under photoexcitation. All these aspects accelerate the Mars–van Krevelen redox cycle, then leading 18

to the remarkable enhanced photothermocatalytic activity of ARCeO2. 3.4. Stability test of catalyst In order to evaluate the stability of prepared catalysts, cyclic photocatalytic degradation stability test of styrene over CeO2, RCeO2 and ARCeO2 with the total reaction time of 25 h were first investigated (Figure 7 a-c). It showed that CeO2 was completely deactivated after the second cycle (Figure 7a), while the deactivation rate of both RCeO2 and ARCeO2 was

ro of

much slower, and it was even not completely deactivated in the fifth cycle (Figure 7b and c). In addition, during the photothermocatalytic degradation process of styrene over ARCeO2,

-p

styrene mineralization efficiencies shows apparently lower than the corresponding degradation efficiencies in the temperature range of 30−200 °C (Figure 7d), implying that lots

re

of intermediates even coke were formed at that stage. Therefore, the temperature of 200 °C

lP

was chosen for long-term stability tests of catalysts in the dark and under illumination. As shown in Figure 7e, an obvious decrease in the catalytic activity was observed onto all the

na

prepared samples after 25 h reaction in the dark. Besides, ARCeO2 exhibited superior stability under illumination, with initial styrene degradation efficiency of 86.6% at 200 °C deceasing

ur

slightly to 78.3% after 25 h reaction, while styrene degradation efficiency at 200 °C decreased

Jo

rapidly from 68.8% over RCeO2 and 62.3% over CeO2 to 44.3% and 30.4% respectively under the identical reaction conditions (Figure 7f). Results obtained suggest that the introduction of OVs and weak acidity onto ARCeO2, together with the synergic effect of photothermocatalysis promotes catalyst more resistant to deactivation. 3.5. Intermediates identification and coke resistance To understand progressive deactivation of catalysts, the intermediates and carbon 19

accumulation on catalyst surface were explored during photocatalytic (PC), thermocatalytic (TC) and photothermocatalytic (PTC) processes by temperature-programmed oxidation (TPO), GC-MS and XPS analysis. For simplification, the spent catalysts (CeO2, RCeO2 and ARCeO2) after cyclic stability test of PC degradation of styrene with the total reaction time of 25 h, long-term stability tests of TC and PTC degradation of styrene at 200 °C for 25 h, are respectively denoted as catalyst-PC (CeO2-PC, RCeO2-PC, ARCeO2-PC), catalyst-TC RCeO2-TC,

ARCeO2-TC)

and

catalyst-PTC

(CeO2-PTC,

RCeO2-PTC,

ro of

(CeO2-TC,

ARCeO2-PTC).

The TPO signals of all the above spent catalysts indicated three types of

-p

carbon-containing species deposited onto the catalyst surface: type I at 30−150 °C, type II at

re

150−250 °C and type III at 250−400 °C (Figure 8a-c). Carbon-I represents non-volatile organic intermediates [47], while carbon-II and carbon-III were respectively assigned to

lP

amorphous and soft coke, which have a much lower combustion temperature than the

na

polymeric or hard coke with high ratio of C/H [25, 48]. All of the spent catalysts except ARCeO2-PTC showed apparent signals of carbon-I, indicating that parts of non-volatile

ur

intermediates could be easily accumulated onto catalyst surface during the PC, TC and even PTC processes. GC-MS analysis is also further applied to qualitatively identify these

Jo

intermediates on the spent catalysts surface, and the corresponding results are all summarized in Table S3. It shows that the intermediates deposited onto CeO2-PC are all the O-containing benzene series, including benzaldehyde, phenol, benzyl alcohol, 2-phenyloxirane, 2,2-dihydroxy-1-phenylethan-1-one,

benzoic

acid,

2-hydroxy-1-phenylethan-1-one,

1-phenylethane-1,2-diol, and phenylethyl alcohol. Moreover, the species number of 20

intermediates ranked by CeO2-PC (9) > RCeO2-PC (7) > ARCeO2-PC (5) and CeO2-PTC (5) > CeO2-TC (4) > RCeO2-TC (3) and ARCeO2-TC (3) > RCeO2-PTC (2) > ARCeO2-PTC (0), further demonstrating that little intermediates have been accumulated onto ARCeO2-PTC. Besides, a small amount of carbon-II (amorphous coke) but no carbon-III (soft coke) could be detectable over CeO2-TC, RCeO2-TC and ARCeO2-TC. Moreover, the peak intensity of carbon-III (soft coke) over CeO2-PTC, RCeO2-PTC and ARCeO2-PTC showed obviously

ro of

weaker than that of the corresponding catalysts after 25 h photocatalytic stability test (CeO2-PC, RCeO2-PC and ARCeO2-PC), confirming the enhanced coke-resistance in photothermocatalysis, and it is observed that the signal intensity of carbon-III (soft coke)

-p

follows the sequence of RCeO2-PTC > CeO2-PTC > ARCeO2-PTC, indicating the highest

re

coke-resistance of ARCeO2. Finally, the coke resistance of CeO2, RCeO2 and ARCeO2 was semi-quantitatively compared by the C1s XPS spectra. As shown in Figure 9, the spent

lP

catalysts (CeO2-PC, RCeO2-PC, ARCeO2-PC, ARCeO2-TC and ARCeO2-PTC) have the same composition of peaks, which includes three peaks at 284.8, 285.9, and 289.1 eV, attributed to

na

sp2C (C-C, C=C), sp3 C (C-O), and oxidized C (C=O), respectively [49], and the content of

ur

assigned C1s is summarized in Table S4. For the PC process (Figure 9a), the total content of carbon species on the samples was in the order of RCeO2-PC (39.4%) > CeO2-PC (37.8%) >

Jo

ARCeO2-PC (32.0%). Moreover, C-C/C=C species have the same sequence of RCeO2-PC (27.8%) > CeO2-PC (25.9%) > ARCeO2-PC (20.4%). Higher C-C/C=C concentration reflects more coke accumulation on the surface of the samples [16], suggesting that the rate of carbon deposit ranked by RCeO2 > CeO2 > ARCeO2 during the PC process. This result is consistent with TPO analysis in Figure 8a. Besides, both the content of total carbon species and 21

C-C/C=C species of ARCeO2-PTC is slightly higher than that for ARCeO2-TC (Figure 9b and Table S4). 3.6. Discussion on the enhanced photothermocatalytic stability of ARCeO2 As mentioned above, there is more coke accumulating onto RCeO2-PTC compared with CeO2-PTC

(Figure

8c),

but

RCeO2

still

had

better

activity

and

stability in

photothermocatalytic degradation of styrene than CeO2 (Figures 4 and 7f). This phenomenon

ro of

further confirmed the promotion effect of OVs on photothermocatalytic degradation of VOCs. Since OVs in catalyst lead to enhanced light absorption, improved charge separation and

-p

increased reactive oxygen generation (Figures 5 and 6), and then promote the photothermocatalytic oxidation reaction (Figures 4), the trade-off between high OVs

re

concentration and high carbon deposit rate of RCeO2 results in the slower deactivation rate

lP

than CeO2. As ARCeO2 has more OVs than RCeO2 (Figures 2), and the low surface acidity of ARCeO2 induces its enhanced coke-resistance performance (Figures 8 and 9a), ARCeO2

three catalysts.

na

proves the best long-term photothermocatalytic stability for VOC degradation among the

ur

In addition, the coke amount of ARCeO2-PTC is slightly higher than that for

Jo

ARCeO2-TC (Figure 9b), while the stability of ARCeO2 in photothermocatalysis is better than that in thermocatalysis. This phenomenon implies that carbon deposit is not the cause of reduced thermocatalytic activity but the produced intermediates, and a little carbon deposit has ignorable negative impact on photothermocatalytic stability. Maybe it could be believed that though a small amount of coke on ARCeO2 surface would block a small part of its active sites, the increased light absorption and enhanced heat storage/release of coke (which can be 22

seen as carbon-based material) could synchronously facilitate photothermocatalytic reaction, therefore prevent the rapidly deactivation of ARCeO2. Given the above, the synergistic effect of the abundant OVs and low surface acidity of ARCeO2 improves the instant mineralization performance for VOCs removal, leading to less intermediates and limited coke accumulation onto ARCeO2 surface, and then gives the excellent long-term photothermocatalytic stability

ro of

for VOC degradation.

4. Conclusions

ARCeO2 ordered porous catalyst with rich OVs and low surface acidity was successfully

-p

developed by simple redox and steam treatment. ARCeO2 exhibits substantially higher photothermocatalytic degradation performance in typical VOCs such as styrene, n-hexane and

re

cyclohexane as compared with the directly calcined sample (CeO2) and the redox-treated

lP

sample (RCeO2), which is believed to due to enhanced light absorption, improved charge separation and increased reactive oxygen generation by OVs introduction. Additionally, the

na

abundant OVs and weak acidity of ARCeO2, together with synergetic effect of photothermocatalysis facilitated the oxidation of intermediates, and thus intensify coke

ur

resistance for high photothermocatalytic performance stability without any activity decline

Jo

after 25 h reaction. This study proves that combine introducing OVs with reducing acidity is a feasible approach to achieve highly efficient simulated solar driven photothermocatalysts with high coke-resistance potential.

Author statement Jiejing Kong: Methodology, Writing-original draft 23

Ziwei Xiang: Data curation Guiying Li: Writing- Reviewing and Editing Taicheng An: Conceptualization, Supervision

Declaration of interests

ro of

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgements

-p

This work was financially supported by Key-Area Research and Development Program of Guangdong Province (2019B110206002), National Natural Science Foundation of China

re

(41731279 and 41415025), Local Innovative and Research Teams Project of Guangdong Pearl

lP

River Talents Program (2017BT01Z032), Leading Scientific, Technical and Innovation Talents of Guangdong special support program (2016TX03Z094) as well as China

ur

na

Postdoctoral Science Foundation Funded Project (2018M630925).

Jo

References

[1] T.C. An, Y. Huang, G.Y. Li, Z.G. He, J.Y. Chen, C.S. Zhang, Environ. Int. 73 (2014) 186-194.

[2] C. He, J. Cheng, X. Zhang, M. Douthwaite, S. Pattisson, Z.P. Hao, Chem. Rev. 119 (2019) 4471-4568. [3] Y. Boyjoo, H.Q. Sun, J. Liu, V. Pareek, S.B. Wang, Chem. Eng. J. 310 (2017) 537-559. [4] C.T. Yang, G. Miao, Y.H. Pi, Q.B. Xia, J.L. Wu, Z. Li, J. Xiao, Chem. Eng. J. 370 (2019) 24

1128-1153. [5] C. Tsang, K. Li, Y.X. Zeng, W. Zhao, T. Zhang, Y.J. Zhan, R.J. Xie, D. Leung, H.B. Huang, Environ. Int. 125 (2019) 200-228. [6] J.J. Kong, T. Yang, Z.B. Rui, H.B. Ji, Catal. Today 327 (2019) 47-63. [7] W.P. Zhang, G.Y. Li, H.L. Liu, J.Y. Chen, S.T. Ma, T.C. An, Environ. Sci-Nano 6 (2019) 948-958. [8] X.Y. Deng, Y.H. Yue, Z. Gao, Appl. Catal. B-Environ. 39 (2002) 135-147. [9] B. Solsona, T. García, R. Sanchis, M.D. Soriano, M. Moreno, E. Rodríguez-Castellón, S.

ro of

Agouram, A. Dejoz, J.M. López Nieto, Chem. Eng. J. 290 (2016) 273-281. [10] T. Montini, M. Melchionna, M. Monai, P. Fornasiero, Chem. Rev. 116 (2016) 5987-6041.

[11] C.Z. Zhu, Y.T. Wang, Z.F. Jiang, F.C. Xu, Q.M. Xian, C. Sun, Q. Tong, W.X. Zou, X.G.

-p

Duan, S.B. Wang, Appl. Catal. B-Environ. 259 (2019) 118072.

[12] S. Weon, W. Choi, Environ. Sci. Technol. 50 (2016) 2556-2563.

re

[13] P. Yang, Z.N. Shi, F. Tao, S.S. Yang, R.X. Zhou, Chem. Eng. Sci. 134 (2015) 340-347. [14] W.K. Ji, T. Shen, J.J. Kong, Z.B. Rui, Y.X. Tong, Ind. Eng. Chem. Res. 57 (2018)

lP

12766-12773.

[15] Y.C. Huang, Y.R. Lu, Y. Lin, Y.C. Mao, G.F. Ouyang, H. Liu, S.Q. Zhang, Y.X. Tong, J.

na

Mater. Chem. A 6 (2018) 24740-24747.

[16] J.J. Kong, G.Y. Li, M.C. Wen, J.Y. Chen, H.L. Liu, T.C. An, J. Catal. 370 (2019) 88-96. [17] D.H. Xia, H.D. Liu, B.H. Xu, Y.C. Wang, Y.H. Liao, Y.J. Huang, L.Q. Ye, C. He, P.

ur

Wong, R.L. Qiu, Appl. Catal. B-Environ. 245 (2019) 177-189. [18] Y.W. Jiang, J.H. Gao, Q. Zhang, Z.Y. Liu, M.L. Fu, J.L. Wu, Y. Hu, D.Q. Ye, Chem. Eng.

Jo

J. 371 (2019) 78-87.

[19] J.L. Deng, W.Y. Song, L.L. Chen, L. Wang, M.Z. Jing, Y. Ren, Z. Zhao, J. Liu, Chem. Eng. J. 355 (2019) 540-550.

[20] Z.Q. Cui, W.K. Wang, C.J. Zhao, C. Chen, M.M. Han, G.Z. Wang, Y.X. Zhang, H.M. Zhang, H.J. Zhao, ACS Appl. Mater. Inter. 10 (2018) 31394-31403. [21] C.L. Mao, J.X. Wang, Y.J. Zou, H. Li, G.M. Zhan, J. Li, J.C. Zhao, L.Z. Zhang, Green Chem. 21 (2019) 2852-2867. 25

[22] I. Son, S. Kwon, J. Park, S. Lee, Nano Energy 19 (2016) 58-67. [23] L. Thomas, R. Tanner, P. Gill, R. Wells, J. Bailie, G. Kelly, S. Jackson, G. Hutchings, Phys. Chem. Chem. Phys. 4 (2002) 4555-4560. [24] S. Shylesh, D. Kim, A. Gokhale, C. Canlas, J. Struppe, C. Ho, D. Jadhav, A. Yeh, A. Bell, Ind. Eng. Chem. Res. 55 (2016) 10635-10644. [25] S. Müller, Y. Liu, M. Vishnuvarthan, X.Y. Sun, A. van Veen, G. Haller, M. Sanchez-Sanchez, J. Lercher, J. Catal. 325 (2015) 48-59. [26] M.J. Xu, Y. Chen, J.T. Qin, Y.W. Feng, W. Li, W. Chen, J. Zhu, H.X. Li, Z.F. Bian,

ro of

Environ. Sci. Technol. 52 (2018) 13879-13886. [27] Y. Cheng, J. Liu, Z. Zhao, W.Y. Song, Y.C. Wei, Chem. Eng. Sci. 167 (2017) 219-228.

[28] X.J. Yao, L. Chen, J. Cao, Y. Chen, M. Tian, F.M. Yang, J.F. Sun, C.J. Tang, L. Dong, Chem. Eng. J. 369 (2019) 46-56.

-p

[29] Z.Y. Fei, C. Cheng, H.W. Chen, L. Li, Y.R. Yang, Q. Liu, X. Chen, Z.X. Zhang, J.H. Tang, M.F. Cui, X. Qiao, Chem. Eng. J. 370 (2019) 916-924.

re

[30] X.H. Zheng, Y.L. Li, L.Y. Zhang, L.J. Shen, Y.H. Xiao, Y.F. Zhang, C. Au, L.L. Jiang, Appl. Catal. B-Environ. 252 (2019) 98-110.

lP

[31] H.Y. Chen, M.N. Tang, Z.B. Rui, X.Y. Wang, H.B. Ji, Catal. Today 264 (2016) 23-30. [32] D. Srinivas, C. Satyanarayana, H. Potdar, P. Ratnasamy, Appl. Catal. A-Gen. 246 (2003)

na

323-334.

[33] X.X. Cheng, X.Y. Zhang, D.X. Su, Z.Q. Wang, J.C. Chang, C.Y. Ma, Appl. Catal. B-Environ. 239 (2018) 485-501.

ur

[34] A. Marciniak, O. Alves, L. Appel, C. Mota, J. Catal. 371 (2019) 88-95. [35] Z.R. Xiao, Y.T. Li, F. Hou, C. Wu, L. Pan, J.J. Zou, L. Wang, X.W. Zhang, G.Z. Liu, G.Z.

Jo

Li, Appl. Catal. B-Environ. 258 (2019) 117940.

[36] F. Al-Sultan, S. Basahel, K. Narasimharao, Fuel 233 (2018) 796-804. [37] Y. Kuwahara, W. Kaburagi, K. Nemoto, T. Fujitani, Appl. Catal. A-Gen. 476 (2014) 186-196. [38] Q.G. Dai, W. Wang, X.Y. Wang, G.Z. Lu, Appl. Catal. B-Environ. 203 (2017) 31-42. [39] Y.C. Huang, H.B. Li, M. Balogun, W.Y. Liu, Y.X. Tong, X.H. Lu, H.B. Ji, ACS Appl. Mater. Inter. 6 (2014) 22920-22927. 26

[40] W. Jiang, Y.N. Feng, Y.Q. Zeng, Y. Yao, L.L. Gu, H.C. Sun, W.J. Ji, H. Arandiyan, C. Au, J. Catal. 375 (2019) 171-182. [41] J.J. Kong, F.L. Xian, Y.Q. Wang, K.H. Ye, M. Balogun, L. Gu, X.Z. Liu, Z.B. Rui, H.B. Ji, J. Phys. Chem. C 122 (2018) 6137-6143. [42] J.J. Kong, Z.B. Rui, H.B. Ji, Ind. Eng. Chem. Res. 55 (2016) 11923-11930. [43] K.H. Ye, K.S. Li, Y.R. Lu, Z.J. Guo, N. Ni, H. Liu, Y.C. Huang, H.B. Ji, P.S. Wang, TrAC-Trend. Anal. Chem. 116 (2019) 102-108. [44] Y. Pu, Y.D. Luo, X.Q. Wei, J.F. Sun, L.L. Li, W.X. Zou, L. Dong, Appl. Catal. B-Environ.

ro of

254 (2019) 580-586. [45] D.K. Chen, D.D. He, J.C. Lu, L.P. Zhong, F. Liu, J.P. Liu, J. Yu, G.P. Wan, S.F. He, Y.M. Luo, Appl. Catal. B-Environ. 218 (2017) 249-259.

[46] S. Zhao, X. Zhao, Appl. Catal. B-Environ. 250 (2019) 408-418.

-p

[47] S. Saqer, D. Kondarides, X. Verykios, Appl. Catal. B-Environ. 103 (2011) 275-286.

[48] S. Kanitkar, M. Abedin, S. Bhattar, J. Spivey, Appl. Catal. A-Gen. 575 (2019) 25-37.

Jo

ur

na

lP

re

[49] Q. Li, J. Shang, T. Zhu, Atmos. Environ. 81 (2013) 68-75.

27

Figure captions: Figure 1. (a-c) TEM images of CeO2, RCeO2 and ARCeO2; (d) XRD patterns of the catalysts Figure 2. (a,b) Ce3d and O1s XPS spectra, (c) EPR spectra, and (d) Raman spectra of the catalysts Figure 3. NH3-TPD profiles of the catalysts Figure 4. (a) Initial activity and (b) Arrhenius plots of different catalysts for degradation of 50

ro of

ppm styrene, and initial activity of catalysts for degradation of 50 ppm (c) n-hexane and (d) cyclohexane in the dark or under light irradiation (λ = 300−780 nm)

Figure 5. (a) UV-vis diffuse reflectance spectra, and (b) Photoluminescence spectra of the

-p

catalysts with an excitation of a Xe lamp (wavelength 280 nm)

re

Figure 6. H2-TPR of the samples in the darkness or under simulated solar irradiation (λ = 300−780 nm)

lP

Figure 7. Cyclic stability test of catalysts for photocatalytic degradation of styrene with an initial styrene concentration of 50 ppm at GHSV = 30,000 mL h-1 g-1 and T = 30 °C

na

under simulated solar irradiation (λ = 300−780 nm) with the total time of 25 h: (a)

ur

CeO2, (b) RCeO2 and (c) ARCeO2; (d) Initial activity of ARCeO2 for 50 ppm styrene degradation under simulated solar irradiation (λ = 300−780 nm); Long-term

Jo

stability test of catalysts for (e) thermocatalytic and (f) photothermocatalytic degradation of styrene at 200 °C with the total time of 25 h

Figure 8. TPO profiles of the spent samples after 25 h stability test for (a) photocatalytic, (b) thermocatalytic and (c) photothermocatalytic degradation of styrene Figure 9. (a, b) C1s XPS spectra of the spent samples 28

ro of

-p

re

lP

na

ur

Jo Figure 1.

29

Figure 2. (b)

(a)

Ob = 529.2

O 1s

Ce 3d

Oa = 531.7

3+

Ce

a1

880

a4

a3

a2

890

b1

b2

900

b4

b3

ARCeO2 RCeO2 CeO2 910

RCeO2

CeO2

920

526

528

530

CeO2

* Mn2+ 3550

252

602

ARCeO2 RCeO2 CeO2

-p

3500

536

ro of

RCeO2

3450

534

464

ARCeO2

Intensity (a.u.)

*

(d)

g=1.963 (Ce3+)

Intensity (a.u.)

*

g=2.002 (Ce4+-O-2 )

532

Binding Energy (eV)

Binding Energy (eV)

(c)

ARCeO2

Intensity (a.u.)

Intensity (a.u.)

Ce3+

200

3600

300

400

500

Wavenumber (cm-1)

Jo

ur

na

lP

re

Field (Gauss)

30

600

700

Figure 3.

b 385

NH3 desorption (a.u.)

a

ARCeO2

152 RCeO2

100

438

200

300

400

500 o

600

Jo

ur

na

lP

re

-p

Temperature ( C)

CeO2

ro of

190

31

700

Figure 4. (b) -21

80

ARCeO2

ARCeO2

RCeO2 CeO2

RCeO2 CeO2

ARCeO2, Dark

60

ARCeO2, Dark

no Catal.

lnY

Styrene Degradation (%)

(a) 100

-22

40

15.6 kJ/mol 48.2 kJ/mol

20

36.8 kJ/mol

-23 0

50

100

150

200

250

300

2.0

2.4

Temperature (oC)

RCeO2 CeO2

60

ARCeO2, Dark no Catal.

40 20

100

200

300

400

ARCeO2

80 60

ARCeO2, Dark no Catal.

40 20 0

500

RCeO2 CeO2

-p

80

ro of

(d) 100 ARCeO2

c-hex. Degradation (%)

n-hex. Degradation (%)

2.8

1000/T

(c) 100

0

20.9 kJ/mol

100

200

300

400

Temperature (oC)

o

Jo

ur

na

lP

re

Temperature ( C)

32

500

600

Figure 5. (b)

(a) ARCeO2

0.9

RCeO2 CeO2

0.6

0.3

0.0 200

300

400

500

600

CeO2 RCeO2

PL Intensity (a.u.)

Intensity (a.u.)

1.2

700

ARCeO2

400

800

450

500

550

600

650

700

Wavelength (nm)

Jo

ur

na

lP

re

-p

ro of

wavelength (nm)

33

Figure 6.

H2 consumption (a.u.)

521 350

ARCeO2 Dark 508

ARCeO2 340

RCeO2

418

522

0

100

200

300

400

500

ro of

CeO2 600

700

Jo

ur

na

lP

re

-p

Temperature (oC)

34

800

Figure 7. (b) 50

PC-30 oC

40 1st run

3rd run

2nd run

CeO2

Styrene Degradation (%)

Styrene Degradation (%)

(a) 50

5th run

4th run

30 20 10 0

0

50

50

50

50

5

PC-30 oC

40

2nd run

1st run

3rd run

20 10 0

0

50

50

10

50

50

50

50

5

(e) 100

60 40

Degradation

20

Mineralization

0 0

50

100

150

200

250

300

RCeO2

lP

CeO2

60 40 20

5

10

Styrene Degradation (%)

80

TC-200 oC

re

(f) 100

ARCeO2

na

Styrene Degradation (%)

80

Temperature (oC)

Time (h)

0 0

5

ro of

20

0

50

ARCeO2 Light

-p

5th run

4th run

30

0

50

(d) 100

ARCeO2

Styrene Conversion (%)

Styrene Degradation (%)

3rd run

5th run

Time (h)

PC-30 oC

40 1st run 2nd run

4th run

30

Time (h) (c) 50

RCeO2

15

20

25

80 60 40 20

0 0

ARCeO2 RCeO2 CeO2 5

10

15

Time (h)

Jo

ur

Time (h)

PTC-200 oC

35

20

25

Figure 8. (a)

Intensity (a.u.)

315 75

ARCeO2-PC

285

RCeO2-PC 110 315

100

200

300

400

CeO2-PC

500

600

700

800

(b)

75

ro of

Temperature (oC)

178

Intensity (a.u.)

ARCeO2-TC

RCeO2-TC

100

200

300

400

-p

CeO2-TC

500

600

700

800

re

Temperature (oC)

(c)

315

lP 100

RCeO2-PTC

195

200

315

300

400

CeO2-PTC

500

600

Temperature (oC)

Jo

ur

ARCeO2-PTC

285

na

Intensity (a.u.)

75

36

700

800

Figure 9 C-C, C=C

(b)

C1s C=O

Intensity (a.u.)

C-O

C1s

ARCeO2-PC

ARCeO2-PTC

Intensity (a.u.)

(a)

RCeO2-PC

C-C, C=C C-O C=O

ARCeO2-TC

CeO2-PC

282

284

286

288

290

292

294

296

282

284

286

288

290

292

294

296

Binding Energy (eV)

Jo

ur

na

lP

re

-p

ro of

Binding Energy (eV)

37