Mesoporous Cetisimcm 48 As Novel Photocatalyst For Degradation Of Organic Compounds

Mesoporous CeTiSiMCM-48 as novel photocatalyst for degradation of organic compounds

Mihaela Mureseanu1, Viorica Parvulescu2*, Teodora Radu3,4, Mihaela Filip2,

Gabriela Carja5

1Faculty of Chemistry, University of Craiova, 107 I Calea București, 200478, Craiova, Romania, E-mail: [anonimizat]

2”Ilie Murgulescu” Institute of Physical Chemistry, Romanian Academy, Spl. Independentei 202, 060021, Bucharest, Romania, E-mail:[anonimizat]

3National Institute for Research and Development of Isotopic and Molecular Technologies, 65-103 Donath Street, 400293 Cluj Napoca, Romania

4Institute of Interdisciplinary Research in Bio-Nano-Sciences, Babes Bolyai University, 400271, Cluj-Napoca, Romania E-mail: [anonimizat]

5Faculty of Chemical Engineering and Environmental Protection, Technical University of Iasi, 71 D. Mangeron, Iasi, Romania. E-mail: [anonimizat]

Abstract

This work presents novel photocatalysts containing Ti and/or Ce embedded in the mesoporous silica framework (TiSiMCM-48, CeSiMCM-48 and CeTiSiMCM-48) that were prepared via a facile sol-gel process in the presence of ionic structure directing agents. The structural properties of the obtained materials were analyzed by X-ray diffraction (XRD), nitrogen adsorption-desorption, scanning and transmission electron microscopy (SEM, TEM), EDAX analysis, X-ray photoelectron microscopy (XPS), ultraviolet–visible diffuse reflectance spectroscopy (UV–Vis DRS) and Fourier transformation infrared spectroscopy (FT-IR). The results indicated that Ce and Ti were highly dispersed or incorporated into the framework of the cubic SiMCM-48, with an enhanced light-trapping effect both in the UV and Vis regions. When applied to the photocatalytic degradation of phenol, the best results were obtained for the bimetallic hybrid. The best activity of CeTiSiMCM-48 photocatalyst was ascribed to improved electron-hole pair separation efficiency and formation of more reactive oxygen species due to the presence of Ce4+/Ce3+. The mesoporous support increases the dispersability of the photoactive Ti4+ or Ce4+/Ce3+ species on the catalyst surface and the accessibility of the substrate to the active sites. Furthermore, the catalysts can be easily recovered and reused for four cycles without significant loss of activity.

Keywords: TiSiMCM-48; CeSiMCM-48; TiCeSiMCM-48; Photodegradation; Phenol

1. Introduction

Phenolic compounds are one of the most biorecalcitrant, toxic, carcinogenic and mutagenic pollutants of the aquatic environment and the discharge limits of phenols in inland water must not exceed a concentration of 1 mg L-1 [1]. For this reason, photocatalytic degradation of phenol and substituted phenols in wastewater has been widely investigated [2, 3], the most used process being based on TiO2/UV [4].

TiO2 is a versatile catalyst due to its excellent degradation of organic pollutants, chemical and biological inertness, high stability against photocorrosion, non-toxicity and low cost [5, 6]. However, titania can only operate under UV high irradiation because of its wide band gap (3.2 eV for anatase) and the fast recombination of photogenerated conduction band electrons (e-cb) and valence band holes (h+vb) on a time scale of 10-9 to 10-12s. For these inconveniences, the efficiency of TiO2 photocatalysis for organic degradation is still not high enough for practical applications [7]. Therefore, a lot of research efforts have been made to dope transition and rare earth metal ions into TiO2 in order to extend its response to visible light [8-12].

Many paper presented improved photocatalytic activity of rare earth doped TiO2 for organic degradation due to extended light absorption range and a better charge separation [13-17].

Cerium is the most favored dopant among lanthanide ions due to its low cost and ability to shift between CeO2 and Ce2O3 during oxidizing and reducing conditions which can lead to different optical and catalytic properties [13, 14, 17-19]. It has been found that the redox couple Ce4+/Ce3+ facilitates the electron transfer and enhances absorption capability in near ultraviolet or ultraviolet [20]. The visible light absorption could be induced by Ce3+ rather than by Ce4+ [21]. At the same time, CeO2 has the ability to store oxygen and transfer this capacity to the TiO2 photocatalysis [22]. Despite these advantages, the thermal and mechanical stability of Ce/TiO2 photocatalysts was not good enough for practical applications. Hence, in order to address these problems, one approach is to disperse TiO2 on the high surface area supports such as silica and alumina. Among different types of mesoporous materials, the M41S series originally developed by Mobil researchers continues to be the most used support [23]. Mesoporous silica support provides a large surface area that helps dispersing catalytically active sites in a controlled chemical environment, facilitating diffusion of reactant molecules to active titania. Other advantage of titania dispersing is separation of the charge carriers due to spatial separation of the photocatalyst particles. Thus, charge carrier recombination is minimized due to the constrained small size of TiO2 and the increased surface acidity [24]. MCM-48, with its high surface area, interpenetrating 3D pores, ordered pore structure array and narrow pore size distribution favors mass transfer kinetics and high dispersion of different semiconductor oxides or organic molecules that can improve the photochemical properties. However, MCM-48 material has not been extensively explored as a support for photocatalysts and literature reports are limited [25-27].

In addition, the tetrahedral Si(IV) in the Si-MCM-48 framework can be replaced by other metal ions such as Al(III), Ti(IV), Ce(IV) and other transition metal ions [28-33] allowing to construct the catalytic sites on the mesoporous silica surface. Different from usual inorganic semiconductor materials, transition metals incorporated into mesoporous supports provide an alternative to design new heterogeneous catalyst for various photocatalytic applications. The photocatalytic mechanism of transition metals dispersed on silica supports is different from bulk semiconductors and it is based on ligand to metal charge transfer excitation. The isolated metal oxide chromophores are excited by UV or visible light with appropriate energy and form charge-transfer excited state [Mn+- O-]*. The photogenerated electrons and holes have a stronger reduction and oxidation ability compared to the charges generated in the corresponding bulk semiconductor photocatalyst. MCM-48 containing two or more metal atoms is very attractive since one metal can modify the structural and redox properties of the other. Consequently, bimetallic catalysts usually improve the catalytic activity, selectivity, and stability of the mono-metallic catalysts [34].

Considering the enhanced photocatalytic activity of rare earth doped mesoporous titania and the necessity for a stable photocatalyst, novel photocatalysts containing Ti or Ce embedded in the silica framework and a bimetallic one with Ce and Ti have been synthesized via a facile sol-gel process in the presence of ionic structure directing agents, at room temperature. In the new bimetallic system, the TiO2 photocatalyst was both doped with Ce and supported onto the MCM-48 mesoporous silica. The photocatalysts were tested in the photodegradation reaction of phenol under UV and Vis light irradiation. The bimetallic photocatalyst presented the best activity both in UV and Vis spectral domains. To the best of our knowledge, this is the first time the enhanced photoactivity of ordered bimetallic mesoporous CeTiSiMCM-48 hybrids for the degradation of organic compounds has been reported. We demonstrated that by simultaneous introduction of Ce and Ti in the synthesis gel in an optimized Si/Ce/Ti molar ratio of 50/0.5/1 the photoactivity reached its maximum value and the spectral response was extended from ultraviolet to visible range. This result could be explained by the decreased TiO2 band gap, inhibition of photo-induced carriers recombination, formation of more reactive oxygen species owing to the presence of Ce3+/Ce4+ couple and enhanced dispersion of active components on the silica surface. In addition, product identification of phenol degradation and detection of the amount of formed on irradiated photocatalyst surfaces allow us to explain phenol degradation through an advanced oxidation process that involve reactive radicalic oxygen species.

2. Experimental

2.1. Chemicals

All chemicals were commercially purchased and used without further purification. Tetraethyl orthosilicate (TEOS, 98%), ammonium hydroxide, sodium hydroxide and anhydrous ethanol were obtained from Fluka. Cetyltrimethylamonium bromide (CTAB, 98%), titanium (IV) isopropoxide (TIPO) and cerium nitrate hexahydrate were purchased from Aldrich. Deionized water was used throughout this study.

2.2. Synthesis of the photocatalysts

Synthesis of SiMCM-48: 1.2 g of Cetyltrimethylammonium bromide (CTAB) was dissolved in 50 mL deionized water and 25 mL ethanol (EtOH) under vigorous stirring in a 500 mL polypropylene beaker until it became a clear solution. Then, 6 mL of aqueous NH3 and 1.8 mL of tetraethyl orthosilicate (TEOS) were consecutively added to the surfactant solution. After stirring at 300 rpm for 4 h, the gel was recovered by filtration, washed with distilled water and dried in an oven at 80 – 90oC overnight. The dried powder was then finely ground and calcined at 550oC for 6 hours at a heating rate of 3oC min-1 to remove the template.

The molar ratio: 1TEOS: 0.41 CTAB: 11.25 NH3: 65.25 EtOH: 347 H2O.

Synthesis of TiSiMCM-48: The same procedure as for SiMCM-48 with the difference that the appropriate amount of titanium precursor from titanium (IV) isopropoxide (TIPO) was dissolved in 1 mL ethanol and was added 1h after the TEOS in order to obtain samples with Si/Ti ratio of 25 and 50, respectively. The samples with different Si/Ti ratio in the gel were named TiSiMCM-48(25) and TiSiMCM-48(50).

The molar ratio: 1TEOS: 0.04 (0.02) TIPO: 0.41 CTAB: 11.25 NH3: 65.25 EtOH: 347 H2O.

Synthesis of CeSiMCM-48: The same procedure as for SiMCM-48 with the difference that the appropriate amount of ceria precursor from Ce(NO3)3 dissolved in 1 mL ethanol was added 1h after the TEOS in order to obtain samples with Si/Ce ratio of 25 and 50, respectively. The samples with different Si/Ce ratio in the gel were named CeSiMCM-48(25) and CeSiMCM-48(50).

The molar ratio: 1TEOS: 0.04 (0.02) Ce(NO3)3: 0.41 CTAB: 11.25 NH3: 65.25 EtOH: 347 H2O.

Synthesis of CeTiSiMCM-48: The same procedure as for SiMCM-48 with the difference that the appropriate amount of ceria precursor from Ce(NO3)3 and titania precursor from TIPO were dissolved in 1 mL ethanol and were added 1h after TEOS in order to obtain samples with Si/Ti/Ce ratio of 50/1/1 and 50/1/0.5, respectively. The samples with different Si/Ti/Ce ratio in the gel were named CeTiSiMCM-48(1) and CeMCM-48(0.5).

The molar ratio: 1TEOS: 0.04 TIPO: 0.04 (0.02) Ce(NO3)3: 0.41 CTAB: 11.25 NH3: 65.25 EtOH: 347 H2O.

2.3. Photocatalyst characterization

Powder X-ray diffraction (XRD) patterns were obtained on a Bruker AXS D8 diffractometer by using Cu K radiation and Ni filter. The FT-IR spectra of the samples were recorded using a Bruker Alpha spectrometer in KBr matrix in the range of 4000–400 cm−1. The UV-Vis diffuse reflectance spectra were recorded using a Thermo Scientific (Evolution 600) spectrometer. N2 adsorption–desorption isotherms were measured at -196°C with a Micromeritics ASAP 2010 instrument. The samples were previously degassed under vacuum at 40°C for 12 h. Specific surface area was calculated by the Brunaur, Emmett, Teller Isotherm Model (BET) and the mesopore volume was determined from the isotherm at the end of capillary condensation. The pores size distribution was obtained from the desorption branch using the Barret–Joyner–Halenda (BJH) method and the Harkins-Jura standard isotherm. The chemical microanalyses by scanning electron microscopy (SEM) were performed on a Hitachi S-4500 I scanning microscope and transmission electron microscopy (TEM) characterization was recorded on a JEOL 1200 EX II operated at 100 kV. XPS measurements were performed with a SPECS PHOIBOS 150 MCD instrument, equipped with monochromatized Al Kα radiation (1486.69 eV) at 14 kV and 20 mA, and a pressure lower than 10−9 mbar. The binding energy scale was charge referenced to the C 1s photoelectron peak at 285eV. A low energy electron flood gun was used for all measurements to minimize sample charging. The elemental composition on the outermost layer of samples (about 5 nm deep from surface) was estimated from the areas of the characteristic photoelectron lines in the survey spectra assuming a Shirley type background. High-resolution spectra were recorded in steps of 0.05eV using the analyser pass energy of 30 eV. The spectra deconvolution was accomplished with Casa XPS (Casa Software Ltd., UK).

2.4. Photocatalytic degradation experiments

Photocatalytic experiments were carried out in batch with a known quantity of photocatalyst added to 250 mL reaction flasks containing 100 mL of a 20 mg L-1 aqueous solution of phenol. The suspension was stirred at a constant speed (300 rpm) throughout the experiment. The reaction mixture was first stirred for 30 min in the dark in order to adsorption-desorption equilibrium to be reached between the photocatalyst and phenol, and then it was irradiated in a closed box with a UV lamp (125 W) having a primary emission at the wavelength of 365 nm. Photocatalytic experiments were performed for 240 min. For the photocatalytic degradation experiments under visible light irradiation, a 400 W halogen lamp with a 400 nm filter that block UV light was used and the experiments were performed for 40h. The progress of the reaction was followed by withdrawing 1 mL of phenol solution at regular time intervals and filtered using a 0.45 μm Millipore film. The separation and quantification of phenol were achieved by using a Surveyor Thermo Electron HPLC system (Thermo Scientific). The Hypersil Gold C18 column with 5 m packing (150 mm 4.6 mm i.d.) was used for the analysis. The determinations were made in isocratic conditions, at 250C using a mixture of methanol/water (40:60 v/v) as mobile phase. The volume injected was 5 μL and the flow rate of the mobile phase was 1mL/min. The wavelength was set at 225 nm to integrate the peaks. A good resolution was obtained for phenol (3.14 min). Calibration curves were constructed in the range of 1-150 ppm. Linearity was achieved and the correlation coefficient was 0.999. All analyses were performed in triplicate always with reproducibility within 3 %. For separation and quantification of phenol degradation products, a reverse phase Hypersil Gold C18 column with 5 m packing (250 mm 4.6 mm i.d.) was used. A mobile phase composition of 15%MeOH/85%H2O/1% glacial acetic acid was utilized. The mobile phase was delivered at a flow rate of 1mL/min and the wavelength was set at 270 nm. The reaction intermediates were confirmed by co-injection of commercial standards under the same operation conditions.

To evaluate the photocatalytic stability, the photocatalysts have been tested in four consecutive cycles under UV light irradiation. After each cycle, photocatalyst sample was collected by centrifugation, then washing with deionized water and dried in an oven at 600C for 6h. The loss of samples was measured by gravimetry and was insignificant.

radicals were detected by fluorescence technique [35] that detects the amount of these species formed on photocatalyst surface. The radicals react with terephthalic acid (TPA) to produce highly fluorescent 2-hydroxyterephthalic acid. 20 mg of photocatalyst were added to a solution of 5×10-4M TPA and 2×10-3M NaOH and the obtained mixture was irradiated 120 min at 315 nm. 3 mL solution was filtrated by Millipore filter membrane and analyzed at 425 nm after each 20 min. of irradiation using JY Horiba fluorimeter.

3. Results and discussion

3.1. Characteristics of photocatalysts

The SiMCM-48, TiSiMCM-48 and CeSiMCM-48 mesoporous materials were prepared via a facile sol-gel process in the presence of an ionic structure directing agent using approaches reported by Peng et al. [36] with little modifications. In the initial synthesis gel, the ratio of Si/Ti and Si/Ce were 50 and 25, respectively. Titanium isopropoxide was chosen as titania source and was added 1h after the TEOS since the hydrolysis and condensation rates of the titanium alkoxide and the silica alkoxide precursors vary vastly and it is important to control the experimental parameters in order to obtain a mesoporous silica containing titanium into its framework. For the CeSiMCM-48 sample, cerium nitrate was used as a cerium source and was added after the TEOS prehydrolysis step. For the CeTiSiMCM-48 material, TIPO and Ce(NO3)3 were added together 1h after the TEOS, in order to obtain Si/Ti/Ce ratio of 50/1/1 and 50/1/0.5, respectively. High quality titanium and cerium containing MCM-48 mesoporous materials were formed with an optimum stirring rate of 300 rpm for 4h at room temperature.

3.1.1. XRD and BET analysis

The formation of ordered mesoporous CeTiSiMCM-48 nanohybrids was firstly provided by XRD analysis. The powder XRD patterns of the samples with the best photocatalytic activity are presented in Fig. 1. The results show a distinctive Bragg reflection and the corresponding d spacing due to the cubic Ia3d symmetry that are in good agreement with those reported for the cubic MCM-48 materials [37]. The most intense diffraction peak (211) appears in the range of 2.5-3.50 for all the samples and demonstrates that the cubic phased mesopore structure was preserved after Ti and Ce incorporation. There is a shift of the (211) peak from 2.60 to 2.80 for the Ce containing samples, probably due to the higher ionic radius of Ce3+/Ce4+ in comparison with those of Ti4+ and Si4+. Furthermore, the shoulder peak (220) at 2θ of 3.2-3.30 is most visible for Ce containing samples, proving that the cubic symmetry was stabilized in the heteroatom-substituted silicate framework. Other low intense diffraction peaks of higher order appear in the interval 4-60 indicating high order array of mesopores. The d spacing corresponding to (211) reflection and the values of cubic cell constant (a0) of the samples are given in Table 1. These results show that there are not significant changes after the Ti incorporation in the MCM-48 silica framework. However, CeSiMCM-48 and CeTiSiMCM-48 samples exhibited less intense peaks that could be ascribed to the addition of Ce species that cause the disruption and gradual loss of long-range ordering of the mesoporous silica. The high angle XRD patterns of the samples show a broad peak at 2 near 250 due to the bulk silica. The amorphous peak width for the cerium containing samples suggests a further disorder of the silica framework which is probably due to the different radius of cerium and silicium ionic species. The absence of peaks due to TiO2 or CeO2 suggests that these components are highly dispersed or incorporated into the framework of MCM-48. It is possible that the concentration of these oxides may be quite small to be detected by powder XRD instrument, as previously observed [24]. Oxides that could be uniformly adsorbed on the surface of Si-MCM-48 originate in the Ce salt or TIPO used as precursors, via the calcination process. The presence of these oxides would be beneficial to separate the charge carriers and efficiently hinder the recombination of the electron-hole pairs [38].

The N2 adsorption-desorption isotherms of SiMCM-48, TiSiMCM-48(50), CeSiMCM-48(50) and CeTiSiMCM-48(0.5) samples are shown in Fig.2. They have the typical characteristics of the mesoporous material isotherms and are of the IV type, according to the IUPAC classification. The BET surface areas, the average pore sizes calculated by the BJH model and the pore volumes of all samples are presented in Table 1. The textural properties of the TiSiMCM-48 are comparable with those of the mesoporous silica without Ti incorporated in its framework. TiSiMCM-48(50) sample shows even larger pore volume and surface area than SiMCM-48, results that may be explained by the “mineralizing” effect of Ti substitution, which is reported to be able to remove the defect sites in the pure silica structure [39]. The pore size of CeSiMCM-48 samples increases, for example, the pore size of SiMCM-48 is about 2.46 nm and that of CeSiMCM-48(50) is about 2.66 nm. It is possible that another structure may be formed in the cerium containing surface areas and, consequently, silica framework may be disrupted. These results are similar to the situation of Ce incorporation into the silicate framework of MCM-48 [30].

The unit cell parameters and the pore volume for CeTiSiMCM-48 are slightly smaller than that of SiMCM-48 and it is reasonable to suppose that a small portion of cerium ions may enter inside the TiSiMCM-48 framework and the rest may be uniformly distributed on the surface via forming Ti-O-Ce or Si-O-Ce bonds (Table 1).

3.1.2. FT-IR

Fig. 3 shows the FT-IR spectra of the Si-, TiSi-, CeSi- and CeTiSi-MCM-48 samples. For all four samples, the broad band at 3000-3800 cm-1 is the characteristic band of the water absorbed on the sample surface, while the peak at 1636 cm-1 is due to the bending vibration of the H-O-H bond of the chemisorbed water. The band at 1000-1250 cm-1 is ascribed to the asymmetric stretching vibration of the Si-O-Si bond; the band about 801 cm-1 is due to the corresponding symmetric vibration of the Si-O-Si bond, while the band at 466 cm-1 is assigned to the rocking vibration of the Si-O-Si bond. The existence of silanol bands at 957 cm-1 could refer to the stretching vibrations of free silanol (Si-OH) groups on the surface of the amorphous solid samples. After the Ti, Ce or both metals incorporation in the MCM-48 silica framework, the peak at 1133 cm-1 appears as a shoulder and there are new peaks at 1046 cm-1 for TiSiMCM-48(50) sample or at 1086 cm-1 for CeSiMCM-48(50) and CeTiSiMCM-48(0.5) that could be assigned to the new Si-O-Ti or Si-O-Ce bonds, suggesting perturbations in the silica framework.

3.1.3. TEM and SEM characterization

The typical TEM images of Ti and Ce embedded into SiMCM-48 samples are presented in Fig. 4 and 5. The images show the long-range ordered 3D cubic channels with uniform pore size, consistent with the XRD observations. The structure of the Ti and Ce containing mesoporous silica was almost identical to that of the SiMCM-48, indicating a low influence of cerium and titanium immobilization on the integrity of the porous structure.

A more significant influence on spherical morphology of MCM-48 particles and the presence of CeO2 on the surface of the hybrid bimetallic photocatalyst was evidenced by SEM microscopy (Fig. 5). Fig. S1a displayed EDAX results and SEM image and S1b the corresponding element mappings of CeTiSiMCM-48(0.5) sample confirming the presence of Ti and Ce and their homogeneous dispersion on the silica support.

3.1.4. XPS analysis

The sample surface was investigated by XPS analysis to characterize the chemical state of Ce, Ti and O with particular attention given to the Ce and Ti oxidation states. The XPS survey spectra presented in Fig. 6 confirmed the presence of these elements in the TiSiMCM-48(50), CeSiMCM-48(50) and CeTiSiMCM-48(0.5) samples. The binding energies of the identified photoemission lines match those reported in the literature [40]. The Si/Ti molar ratio in TiSiMCM-48(50) sample was 53/1, in good agreement with the ratio in the initial mixture (Table 2). For the CeSiMCM-48(50) sample, the Si/Ce molar ratio was 50/0.5, meaning that only 50% of the initial cerium concentration was on the mesoporous hybrid surface. The Si/Ti/Ce molar ratio in CeTiSiMCM-48(0.5) sample was 50/1.4/0.6, proving that both titanium and cerium were mostly present on the material surface and their concentrations for the bimetallic mesoporous silica were higher than for the monometallic samples.

XPS spectra of the Ti 2p spin-orbit-split 3/2 and 1/2 peaks are found in Figure 7; the Ti 2p3/2 and Ti 2p1/2 binding energies are located at 459.9 and 465.4 eV for TiSiMCM-48(50), and 457 and 462.7 eV for CeTiSiMCM-48(0.5), respectively. The peak separation between Ti 2p3/2 and Ti 2p1/2 was 5.5 eV and 5.7 eV, respectively, indicating the presence of titania phase. The tail-off of intensity toward higher BE may be attributed to inelastically scattered electrons.

The Ti 2p region of both samples can be deconvoluted using four peaks corresponding to Ti3+ species and Ti4+, respectively. For the TiSiMCM-48(50) sample, the Ti3+ binding energies are located at 457.2 (462.5) eV and those corresponding to Ti4+ at 459.8 (465.7) eV. In the case of CeTiSiMCM-48(0.5) sample, the binding energies are located at 455.2 (461.3) eV for Ti3+ species and 457.4 (463.1) eV for Ti4+ species [41]. The appearance of Ti3+ in the silica framework indicated that a part of Ti4+ ions were reduced to Ti3+ during heat treatment. Another reason for the reduction of Ti4+ to Ti3+ could be induced by Ce incorporation during calcination treatment, as it was previously observed for TiO2 nanoparticles doped with CeO2 [42].

Fig. 8 presents the O 1s XPS spectra and the obtained deconvolution results for all investigated samples. The O1s XPS spectra appear broad and can be decomposed into four distinct peaks: three of them are assigned to lattice oxygen in the metal oxides while the peak at around 531.4 eV could be assigned to hydroxyl group OH- on the surface [43]. For the CeTiSiMCM-48(0.5) sample, the peaks at 529.9 eV, 528.7 eV and 533.6 eV could be attributed to lattice oxygen in Ce-O-Si, Ce-O-Ti and Si-O-Si linkages, while the peak at 531.3 eV corresponds to OH-. Similarly, for TiSiMCM-48(50) the peaks at 524.9 eV, 529.7 eV, 532.8 eV are assigned to Ti-O-Ti, Ti-O-Si and Si-O-Si linkages and for CeSiMCM-48(50) the peaks at 527.8 eV, 530.2 eV, 532.6 eV correspond to Ce-O-Ce, Ce-O-Si and Si-O-Si linkages. It is interesting that, after cerium introduction in the hybrid mesoporous silica, more oxygen in hydroxyl groups is present on its surface.

Fig. 9 shows the high resolution Ce 3d spectra and the obtained deconvolution results. The Ce3d region is complicated because of the hybridization of Ce4f and O2p electrons and can be split into four pairs of spin-orbital bands. For identifying the Ce3d XPS peaks, there were used the labels established by Burroughs et al [44] where and u denote the spin-orbit coupling 3d5/2 and 3d3/2. Through the peak fitting process of the photoelectron spectra of Ce ions, it can be found that these ones exist in two forms, namely Ce4+ (,,’’, ,’’’, u, u’’, u’’’) and Ce3+ (,o ,’, uo, u’). The quantitative analysis results and the percentage of the Ce3+ state of the Ce3++Ce4+ total are summarized in Table 3. Considering that the ionic radii of Ce4+ and Ce3+ ions were 0.092 and 0.103 nm, respectively, much larger than that of Ti4+ (0.064 nm), it is clear that, when both metal ions are present together, the tendency for cerium ions is to rest on the surface as Ce oxides rather than enters in the lattice by substituting Ti4+ ions. Therefore, Ce is present in a mixed oxide structure, namely CeO2 and Ce2O3. This observation is supported by the Si/Ti/Ce molar ratio in CeTiSiMCM-48(0.5) sample when the amount of both Ti and Ce used for the synthesis was found on the surface of the hybrid mesoporous material. The presence of both Ce3+ and Ce4+ shows that the surface of the photocatalyst was not fully oxidized, suggesting some oxygen vacancies. For the CeTiSiMCM-48 (0.5) sample, the percentage of Ce3+ is greater than for CeSiMCM-48 (50), which was beneficial to the light absorption towards the visible region. Furthermore, the presence of Ti3+ and Ce3+ ions on the surface of the MCM-48 hybrid material could result in a charge imbalance, yielding oxygen vacancies and unsaturated chemical bonds. In this situation, the additional oxygen which is chemisorbed or weakly adsorbed on the catalyst surface, improved the activity of phenol photocatalytic degradation, as previously observed [42]. For these Ce-containing samples, the additional oxygen on the material surface was supported by the increased 531.4 eV peak in the O1s XPS spectra.

3.1.5. UV-Vis diffuse reflectance spectroscopy

In order to investigate the electronic state of the catalysts, UV–Vis DRS studies were performed on SiMCM-48, TiSiMCM-48(50), CeSiMCM-48(50) and CeTiSiMCM-48(0.5) (as shown in Fig.10). In TiSiMCM-48, the strong absorption at 250 nm is attributed to the ligand to metal charge transfer from O2- to Ti4+ in tetrahedral geometry. A broad shoulder at 290 nm indicates the presence of a fraction of Ti4+ in octahedral coordination. The peak absence near 330 nm suggests no TiO2 bulk is formed in this sample. The same effects were observed by Peng and collab. [24].

The UV-visible spectra of CeSiMCM-48 sample (Fig 10a) present a broad absorption band at 300 nm that could be associated with the charge transfer of ligand to metal (O2-Ce4+). It was proved that the electronic transitions from oxygen to cerium require higher energy for tetra-coordinated Ce4+ than for hexa-coordinated Ce4+ (as in CeO2) [25]. Therefore, the absorption band near 300 nm is probably due to the Ce4+ cations present almost with tetra-coordination in the silica framework. The ordered mesopores stabilized in a new structure by the cerium present into the silica framework improved the reflection or transmission of the scattered light, and a similar effect is mentioned in literature [45].

CeTiSiMCM-48(0.5) sample shows strong absorption in ultraviolet region and a broad adsorption in visible region (= 400-550 nm) owing to the photosensitizing effect of Ce3+. It is worthy of note that the nanohybride containing cerium and titanium exhibits a much stronger absorption in both ultraviolet and visible-light regions than TiSiMCM-48 sample. This enhanced light-trapping effect is the result of the increased amount of Ce3+ on the surface, as the XPS results revealed.

The band gap energy of the above samples could be calculated by using the equation (αh)n = k(h–Eg), where α is the absorption coefficient, k is the parameter that related to the effective masses associated with the valence and conduction bands, n is 1/2 for a direct transition, h is the absorption energy and Eg is the band gap energy [46]. Plotting (αh)1/2 versus h based on the spectral response in Fig. 10a gives the extrapolated intercept corresponding to the Eg value (Fig. 10b).

As shown in Table 1, the optical band gap energies of the CeSiMCM-48 (2.5 eV) display obvious red-shifts compared to that of TiSiMCM-48 (3.1 eV). The difference of band gap energy between these two samples may result in a different photocatalytic activity, which is connected with their electron configurations. As the band gap energy corresponding to longer wavelengths gets lower, the electron-hole is more easily produced and the photocatalytic activity is higher. Furthermore, the existence of Ce3+ detected by XPS analysis led to an occupation of the 4f band, resulting in additional energy state and band gap reduction. Visible light then can excite the electrons from the interband Ce 4f level to the conduction band, inhibiting the recombination of electron and hole pairs. The red-shifts in the band gap energy of CeTiSiMCM-48(0.5) in comparison with TiSiMCM-48(50) can be attributed to the presence of Ce3+ ions on the material surface, as previously mentioned. It is also possible that cerium ions may have produced lattice deformation and vacancy, which will probably result in an impurity state in the band gap of TiO2 [47]. This impurity can narrow the band gap and improve the adsorption.

Therefore, the results of this study indicate that the enhanced ability of these ordered mesoporous CeTiSiMCM-48 nanohybrids to absorb visible-light makes them a promising photocatalyst for solar-driven applications.

3.2. Photocatalytic activity and durability

Photocatalytic oxidation of phenol in aqueous media was selected as a model reaction in order to study the catalytic efficiency of different CeTiSiMCM-48 hybrids upon UV and Vis light excitation. For comparison purposes, a blank experiment, in the absence of any catalyst, was also performed. In that case, only 5% of phenol abatement was achieved after 4 h of UV and 40 h of Vis light irradiation, respectively. The photocatalytic activity of Ti and Ce modified samples was evaluated in comparison with unmodified SiMCM-48 under the same experimental conditions. The SiMCM-48 sample presented insignificant phenol degradation and was not further considered for comparison. Fig. 11a presents the decomposition ratio (C/C0) correlated with the reaction time (t) of 20 mg L-1 phenol in the presence of 1.0 g L-1 photocatalyst, under UV illumination. The maximum phenol photodegradation was 78% for the TiSiMCM-48(50) photocatalyst. Ce-substituted MCM-48 silica was a more efficient catalyst than Ti-substituted one, and the best result was 96% phenol degradation for the CeSiMCM-48(50) sample. The best Si/Ti and Si/Ce molar ratio for the phenol degradation efficiency was 50/1, both for Ti- and Ce-substituted samples. The simultaneous introduction of Ce and Ti in the synthesis gel of the bimetallic MCM-48 mesoporous hybrid was beneficial to the photocatalytic activity. After 240 min. irradiation with UV light, the CeTiSiMCM-48 (0.5) and CeTiSiMCM-48 (1) samples presented a phenol degradation of 99% and 96%, respectively. The best phenol photodegradation results were obtained for Ce/Ti molar ratio of 0.5/1. The results of the phenol photodegradation under Vis light irradiation (Fig. 11b) confirmed the best photocatalytic activity for the CeTiSiMCM-48(0.5) sample and it was consistent with the red shift in the solid state UV–Vis diffuse reflectance spectra in Fig. 10(a). In these conditions, the cerium addition enhanced the visible absorption of TiO2 leading to its band gap narrowing. The Ce4+ ions, as Lewis acid, had strong capability of trapping electrons, being beneficial to the separation of electron-hole pairs [48] and, therefore, improved the photocatalytic activity. Due to the additional oxygen present on the surface of Ce-containing samples, the photoinduced electron can combine them to yield superoxide free radical with high oxidation capability [49]. However, when Ce content increases, Ce4+ becomes the recombination centers of photo-induced electron-hole pair and it is reasonable that there is an optimum content of cerium. Interesting is that for the samples containing only TiO2 embedded in the silica framework, the phenol conversion under Vis light irradiation was only ~10%. Furthermore, the presence of the mesoporous SiMCM-48 support increases the dispensability of Ti4+/Ti3+ or Ce4+/Ce3+ that are present either in the silica framework or as oxides on the catalyst surface, the accessibility of the substrate to the active sites being consequently increased. In addition, the bimetallic samples have more surface hydroxyl groups that are beneficial to the adsorption of the organic compound, which improves the photocatalytic activity. The more abundant surface adsorbed hydroxyl groups could facilitate the formation of hydroxyl radicals which could be useful to the oxidation process of the adsorbed phenol.

It is well known that photocatalytic oxidation of organic pollutants, when their initial concentrations are below or equal to 30 mg L-1, follows the Langmuir-Hinshelwood pseudo-first-order kinetics model. Therefore, this kind of reaction could be represented by the equation [50]:

where kapp is an apparent first-order reaction rate constant, C is the phenol concentration and t is the reaction time. Plotting ln(C/C0) versus reaction time (t) yields a straight line, and the slope is the kapp.

The photocatalytic degradation of 20 mg L-1 phenol using 1.0 g L-1 of the new Ti or/and Ce substituted MCM-48 photocatalysts under UV light (Fig. 12) was successfully fitted using L-H model, and can be described by pseudo-first order kinetics as confirmed by the obtained straight line. The apparent pseudo first-order degradation rate constant of phenol over TiSiMCM-48(50) was 6.7 10-3 min-1. CeSiMCM-48(50) and CeTiSiMCM-48(0.5) samples gave an improved reaction rate of 12.8 10-3 min-1 and 16.8 10-3 min-1, respectively. This may be attributed to the optimum Si/Ce and Si/Ti/Ce molar ratios that allow the new photocatalysts to effectively capture the photo-induced electrons and holes and inhibit their recombination. Furthermore, the mesoporous structure with a good dispersability and accessibility of the photoactive sites, large surface area and more surface hydroxyl groups is favorable to greater reaction rates.

The photocatalytic stability of various samples under ultraviolet irradiation was evaluated by recycled experiments, and the results are shown in Fig.13. After the first run, there was an obvious decrease in the activity of TiSiMCM-48(50) which always presents an instability problem [43]. For CeSiMCM-48 (50) and CeTiSiMCM-48(0.5) samples, the degradation rate of phenol kept a relatively steady state of ca. 92-96% in the subsequent four runs, although there was a slight activity loss in the first run. These results proved that the large specific area, the high pore volume and the good dispersability of the photocatalytic active sites are beneficial to the catalyst durability. Furthermore, the relative concentration of the elements presented at the investigated sample surface after repeated use and determined by XPS analysis are the same as for the initial samples. This result confirms that it was no Ce and Ti leaching during consecutive photocatalyst use. These new photocatalysts are much cheaper than noble metal based photocatalysts [51] and show a great potential in practical applications due to their easy separation, while preserving their very good activity.

The photoactivity and durability of this new hybrid bimetallic MCM-48 photocatalyst for phenol degradation could be explained on the basis of advanced oxidation processes (AOPS) promoted by the heterogeneous photocatalyst.

The incorporation of Ce in the TiSiMCM-48 framework allows activation by both UV and visible light and, thus, more electrons and holes can be generated and they can participate in the photocatalytic redox reaction. On the other hand, electrons can be more excited from the Ce 4f interband level to the conduction band of titania. The excited electrons of the semiconductor can react with the adsorbed oxygen to form superoxide radicals which can bring about the phenol photo-degradation. Furthermore, the excited electrons can be trapped by Ce4+ions through the following process:

Then the electrons trapped in Ce4+/Ce3+ site are transferred to the adsorbed oxygen in the system. Therefore, the recombination rate of photo-induced electron-hole pairs decreases and the photocatalytic activity of the CeTiSiMCM-48 photocatalyst increases. Thus, the presence of Ce4+/Ce3+pair not only efficiently separated and altered the recombination rate of the photogenerated electrons and holes but also resulted in the generation of oxidizing species beneficial to efficient phenol degradation. The holes can react with H2O or HO- to yield hydroxyl radicals (). The radicalic species and can reduce and oxidize the organic pollutant phenol to the mineral end products such as CO2 and H2O. The formation of HO radicals under UV light was examined for solutions with dispersed CeTiSiMCM-48 and Degussa TiO2 powders and was evidenced by the increase of the fluorescence emission intensity in time (Fig. 14). The fluorescence intensity increase due the formation of 2-HTPA was in agreement with published results [35]. The commercial TiO2 photocatalyst shows lower levels of intensity than the new bimetallic mesoporous photocatalyst. Presence of cerium favors in this case the production of a larger amount of radicals. Figure 14 shows a significant increase of fluorescence intensity after 2 hours of UV irradiation, which is consistent with the photocatalytic test results (Fig. 11a).

The HPLC analysis (Fig. S2) carried out on supernatant obtained after 240 min UV irradiation revealed the presence of five compounds at 3.73, 4.18, 4.75, 6.97 and 18 min, corresponding to maleic acid, fumaric acid, hydroquinone, resorcinol and phenol, respectively. The formation of maleic acid as one of the major intermediate sustains the phenol degradation through an advanced oxidation process that involves reactive radicalic oxygen species.

4. Conclusions

Novel hybrids with Ti, Ce or both metals embedded in the MCM-48 mesoporous silica framework were synthesized by a facile sol-gel process and then, the photocatalytic properties were evaluated by phenol photodegradation under UV irradiation. The ordered bimetallic mesoporous CeTiSiMCM-48 hybrid presented 99% phenol degradation efficiency after 240 min. irradiation with UV light and it was used during four cycles without significant loss of activity. The enhanced activity and stability were attributed to the presence of Ce4+/Ce3+ pair and to the high dispersion of the photoactive species on the SiMCM-48 mesoporous support surface. Thus, these novel CeTiSiMCM-48 photocatalysts are expected to serve as potential catalysts for advanced oxidative degradation processes of organic pollutants.

Acknowledgment

The authors gratefully acknowledge the financial support from the Romanian National Authority for Scientific Research, CNCS-UEFISCDI; project number PN-II-IDPCE 75/2013

References

[1] S. Lathasree, A. Nageswara Rao, B. SivaSankar, V. Sadasivam, K. Rengaraj, J. Mol. Catal. A: Chem. 223 (2004) 101–105.

[2] G. Busca, S. Berardinelli, C. Resini, L. Arrigi, J. Hazard. Mater. 160 (2008) 265–288.

[3] L.F. Liotta, M. Gruttadauria, G. Di Carlo, G. Perrini, V. Librand, J. Hazard. Mater. 162 (2009) 588–606.

[4] A. Turki, C. Guillard, F. Dappozze, Z. Ksibi, G. Berhault, H. Kochkar, Appl. Catal. B: Environ. 163 (2015) 404–414.

[5] A.L. Linsebigler, G. Lu, J.T. Yates, Chem. Rev. 95 (1995) 735–758.

[6] A. Fujishima, T.N. Rao, D.A. Tryk, J. Photoch. Photobio. C: Photochemistry Reviews 1 (2000) 1–21.

[7] S. Malato, P. Fernandez-Ibanez, M.I. Maldonado, J. Blanco, W. Gernjak, Catal. Today 147 (2009) 1–59.

[8] A.D. Paola, E. Garcia-Lopez, G. Marci, C. Martin, L. Palmisano, V. Rives, A.M. Venezia, Appl. Catal. B: Environ. 48 (2004) 223–233.

[9] D. Dvoranova, V. Brezova, M. Mazur, M.A. Malati, Appl. Catal. B: Environ. 37 (2002) 91–105.

[10] S. Paul, P. Chetri, A. Choudhury, J. Alloys Compd. 583 (2014) 578–586.

[11] K. Umar, M. Haque, M. Muneer, T. Harada, M. Matsumura, J. Alloys Compd. 578 (2013) 32–36.

[12] S. Sood, A. Umar, S.K. Mehta, S.K. Kansal, J. Colloid Interface Sci. 450 (2015) 213–223.

[13] M.J. Munoz-Batista, M.N. Gomez-Cerezo, A. Kubacka, D. Tudela, M. Fernandez-Garcia, ACS Catal. 4 (2014) 63–72.

[14] S. Ghasemi, S. Rahman Setayesh, A. Habibi-Yangjeh, M.R. Hormozi-Nezhad, M.R. Gholami, J. Hazard. Mater. 199–200 (2012) 170–178.

[15] R.M. Mohamed, I.A. Mkhalid, J. Alloys Compd. 501 (2010) 143–147.

[16] J. Reszczynska, T. Grzyb, J.W. Sobczak, W. Lisowski, M. Gazda, B. Ohtani, A. Zaleska, Appl. Catal. B: Environ. 163 (2015) 40–49.

[17] N. Aman, P.K. Satapathy, T. Mishra, M. Mahato, N.N. Das, Mater. Res. Bull. 47 (2012) 179–183.

[18] F.B. Li, X.Z. Li, M.F. Hou, K.W. Cheah, W.C.H. Choy, Appl. Catal. A: General 285 (2005) 181–189.

[19] Y. Liu, P. Fang, Y. Cheng, Y. Gao, F. Chen, Z. Liu, Y. Dai, Chem. Eng. J. 219 (2013) 478–485.

[20] S. Pavasupreea, Y. Suzukia, S. Pivsa-Art, S. Yoshikawa, J. Solid State Chem. 178 (2005) 128–134.

[21] J.M. Coronado, A.J. Maira, A. Martinez-Arias, J.C. Conesa, J. Soria, J. Photochem. Photobiol. A 150 (2002) 213-221.

[22] G. Dutta, U.V. Waghmare, T. Baidya, M.S. Hegde, K.R. Priolkar, P.R. Sarode, Chem. Mater. 18 (2006) 3249–3256.

[23] C.T. Kresge, W.J. Roth, Chem. Soc. Rev. 42 (2013) 3663-3670.

[24] R. Peng, J. Baltrusaitis, C.-M. Wu, R. T. Koodali, Int. J. Hydrogen Energ. 40 (2015) 905-918.

[25] L.Z. Wang, L. Jiang, C.C. Xu, J.L. Zhang, J. Phys. Chem. C 116 (2012) 16454-16460.

[26] Z. Wang, X. Ci, H. Dai, L. Yin, H. Shi, Appl. Surf. Sci. 258 (2012) 8258-8263.

[27] V. Krishna, V.S. Kamble, N.M. Gupta, P. Selvam, J. Nanosci. Nanotechnol. 6 (2006) 1811-4.

[28] P.A. Russo, M.M.L. Ribeiro Carrott, P.J.M. Carrott, J.M. Lopes, F. Ramôa Ribeiro, J. Rocha, Micropor. Mesopor. Mat. 114 (2008) 293–302.

[29] Y. Shao, L. Wang, J. Zhang, M. Anpo, J. Photoch. Photobio. A 180 (2006) 59–64.

[30] W. Zhan, G. Lu, Y. Guo, Y. Guo, Y. Wang, Y. Wang, Z. Zhang, X. Liu, J. Rare Earths, 26 (2008) 515-522.

[31] S. Gomez, O. Giraldo, L.J. Garces, J. Villegas, S.L. Suib, Chem. Mat. 16 (2004) 2411–2417.

[32] T. Jiang, D. Wu, J. Song, X. Zhou, Q. Zhao, M. Ji, H. Yin, Powder Technol. 207 (2011) 422–427.

[33] D. Zhao, S. Budhi, A. Rodriguez, R.T. Koodali, Int. J. Hydrogen Energ. 35 (2010) 5276-5283.

[34] H. Maneesuwan, R. Longloilert, T. Chaisuwan, S. Wongkasemjit, Mat. Lett. 94 (2013) 65–68.

[35] S. Rasalingam, H. S. Kibombo, C.-M. Wu, R. Peng, J. Baltrusaitis, R. T. Koodali, Appl. Catal. B: Environ. 148-149 (2014) 394–405.

[36] R. Peng, C.M. Wu, J. Baltrusaitis, N.M. Dimitrijevic, T. Rajh, R. T. Koodali, Chem. Commun. 49 (2013) 3221–3223.

[37] A.A. Romero, M.D. Alba, W. Zhou, J. Klinowski, J. Phys. Chem. B 101 (1997) 5294-5300.

[38] R. Nakamura, Y. Nakato, J. Am. Chem. Soc. 126 (2004) 1290–1298.

[39] C. Lamberti, S. Bordiga, A. Zecchina, G. Artioli, G. Marra, G. Spano, J. Am. Chem. Soc. 123 (2001) 2204-12.

[40] Q. Chen, D. Jiang, W. Shi, D. Wu, Y. Xu, Appl. Surf. Sci. 255 (2009) 7918-7924.

[41] J.-G. Yu, H.-G. Yu, B. Cheng, X.-J. Zhao, J.C. Yu, W.-K. Ho, J. Phys. Chem. B 107 (2003) 13871–13879.

[42] C. Hao, J. Li, Z. Zhang, Y. Ji, H. Zhan, F. Xiao, D. Wang, B. Liu, F. Su, Appl. Surf. Sci. 331 (2015) 17-26.

[43] X. Fu, L.A. Clark, Q. Yang, M.A. Anderson, Environ. Sci. Technol. 30 (1996) 647–653.

[44] P. Burroughs, A. Hamnett, A.F. Orchard, G. Thornton, J. Chem. Soc., Dalton Trans. 17 (1976) 1686-1698.

[45] Y. Wang, B. Li, C. Zhang, L. Cui, S. Kang, X. Li, L. Zhou, Appl. Catal. B: Environ. 130–131 (2013) 277–284.

[46] M. Yoon, M. Seo, C. Jeong, J.H. Kang, K.S. Jeon, Chem. Mater. 17 (2005) 6069-6079.

[47] Y. Cong, B.Z. Tian, J.L. Zhang, Appl. Catal. B: Environ. 101 (2011) 376–381.

[48] Z.I. Liu, B. Guo, L. Hong, H. Jiang, J. Phys. Chem. Sol. 66 (2005) 161–167.

[49] V. Augugliaro, L. Palmisano, A. Sclafani, C. Minero, E. Pelizzetti, Toxicol. Environ. Chem. 16 (1988) 89–109.

[50] Y. Xu, H. Chen, Z. Zeng, B. Lei, Appl. Surf. Sci. 252 (2006) 8565–8570.

[51] Z. Zheng, B. Huang, X. Qin, X. Zhang, Y. Dai, M.-H. Whangbo, J. Mater.Chem. 21 (2011) 9079–9087.