- Email: [email protected]
Electrospun cerium-based TiO2 nanofibers for photocatalytic oxidation of elemental mercury in coal combustion flue gas
Chemosphere 185 (2017) 690e698
Contents lists available at ScienceDirect
Chemosphere journal homepage: www.elsevier.com/locate/chemosphere
Electrospun cerium-based TiO2 nanofibers for photocatalytic oxidation of elemental mercury in coal combustion flue gas Lulu Wang, Yongchun Zhao**, Junying Zhang* State Key Laboratory of Coal Combustion, Huazhong University of Science & Technology, Wuhan 430074, PR China
h i g h l i g h t s Cerium-based TiO2 nanofibers were successfully synthesized by electrospinning. Hg0 photocatalytic oxidation was investigated over cerium-based TiO2 nanofibers. Roles of interferential gas components on Hg0 removal were clarified. High removal efficiency of Hg0 were observed for sample with 0.3% Ce. High stability of samples was observed for long-term use.
a r t i c l e i n f o
a b s t r a c t
Article history: Received 13 December 2016 Received in revised form 6 July 2017 Accepted 11 July 2017 Available online 12 July 2017
Photocatalytic oxidation is an attractive method for Hg-rich flue gas treatment. In the present study, a novel cerium-based TiO2 nanofibers was prepared and selected as the catalyst to remove mercury in flue gas. Accordingly, physical/chemical properties of those nanofibers were clarified. The effects of some important parameters, such as calcination temperature, cerium dopant content and different illumination conditions on the removal of Hg0 using the photocatalysis process were investigated. In addition, the removal mechanism of Hg0 over cerium-based TiO2 nanofibers focused on UV irradiation was proposed. The results show that catalyst which was calcined at 400 C exhibited better performance. The addition of 0.3 wt% Ce into TiO2 led to the highest removal efficiency at 91% under UV irradiation. As-prepared samples showed promising stability for long-term use in the test. However, the photoluminescence intensity of nanofibers incorporating ceria was significantly lower than TiO2, which was attributed to better photoelectron-hole separation. Although UV and O2 are essential factors, the enhancement of Hg0 removal is more obviously related to the participation of catalyst. The coexistence of Ce3þ and Ce4þ, which leads to the efficient oxidation of Hg0, was detected on samples. Hg2þ is the final product in the reaction of Hg0 removal. As a consequence, the emissions of Hg0 from flue gas can be significantly suppressed. These indicate that combining photocatalysis technology with cerium-based TiO2 nanofibers is a promising strategy for reducing Hg0 efficiently. © 2017 Elsevier Ltd. All rights reserved.
Handling Editor: Jun Huang Keywords: Electrospinning Nanofiber Photocatalytic oxidation Elemental mercury
1. Introduction Mercury has been identified as one of the most hazardous heavy metals to public health and the environment because of its volatility, persistence, and bioaccumulation in the biosphere (Zhou et al., 2015). According to the 2013 global mercury assessment by
* Corresponding author. ** Corresponding author. E-mail addresses: [email protected] (J. Zhang).
(Y.
http://dx.doi.org/10.1016/j.chemosphere.2017.07.049 0045-6535/© 2017 Elsevier Ltd. All rights reserved.
Zhao),
[email protected]
the United Nations Environment Programme (UNEP), coal burning is one of the most significant anthropogenic sources of mercury emissions to the atmosphere. Mercury from coal-fired flue gas is present in the following three primary forms: elemental mercury (Hg0), oxidized mercury (Hg2þ) and particulate-bound mercury (Hgp). Among them, Hg0 is more difficult to capture with existing air pollution control devices because of its inertness, high volatility € rl et al., 2014). This arouses urgent and low solubility in water (Spo needs for efficient technologies to reduce the Hg0 content in the flue gas. Adsorption and oxidation are the two primary methods used to eliminate Hg0. The injection of activated carbon upstream of the electrostatic precipitators (ESPs) or fabric filters (FFs) has
L. Wang et al. / Chemosphere 185 (2017) 690e698
shown to be an efficient adsorption method (Jia et al., 2010). Unfortunately, the high cost hinders its widespread applications (Jones et al., 2007). Therefore, significant attention has been paid to Hg0 oxidation. Among the different alternative routes currently available for Hg0 oxidation, photocatalysis has been proven to be one of the most promising methods, which can meet all the requirements mentioned above (Chen et al., 2014; Lasek et al., 2013; Wang et al., 2011). TiO2 is the most promising material in photocatalysis process (Feng et al., 2015; Huang et al., 2015). It has been used extensively for its stability, low cost, lack of toxicity and excellent photocatalytic activity. However, the photogenerated electron-hole pair recombination occurs within nanoseconds. To overcome this problem, modification materials have been extensively investigated. Of all these investigated materials, it cannot be ignored that owing to their f electron and multi-electron configuration, rare earth oxides have been found to have polymorphs with strong adsorption and good thermal stability (Xu et al., 2006), which can help improve the performance of TiO2 (Devi and Kumar, 2012; Wang et al., 2014). In particular, cerium oxides have attracted much attention due to the optical and catalytic properties associated with the redox couple Ce3þ/Ce4þ (Silva et al., 2009). The Hg0 removal performance over CePO4 has been found superior to commercial SCR catalyst in the temperature range of 300e400 C (Weng et al., 2015). Wu et al. (2015) provided a fundamental insight into Hg0 oxidation over CoCe/AC catalyst-sorbents and believed that Co4.5Ce6/AC had a promising industrial application prospect. He et al. (2016) further demonstrated that CeO2/TiO2-PILC catalysts had high specific surface area and distinct interlamellar structure during Hg0 removal process. However, these studies only considered high reaction temperature and no light was involved. The performance of cerium-based TiO2 on Hg0 photocatalytic removal at low temperature is still enigmatic. As for the morphology of the photocatalyst, it should be borne in mind that compared with traditional powder catalysts, nanofibers have received great interest in terms of development and industrial applications (Celebioglu et al., 2016; Park and Kim, 2016). Various methods have been investigated to fabricate continuous onedimensional (1D) TiO2 nanofibers, including self-assembly, phase separation, attraction and electrospinning (Salehifar and Nikfarjam, 2017). Among them, electrospinning has been deemed as the most versatile and efficient technique due to its high throughput during a continuous process, mild conditions, and the flexibility in controlling the composition, diameter and porosity (Li et al., 2017). Nano titanosilicate fibers appeared to have high thermal stability and a large number of terminal oxygens, which make them suitable for Hg0 photocatalytic removal (Jeon et al., 2008). The photocatalytic reactivity of TiO2-based materials for Hg0 removal under near-UV irradiation has been systematically investigated at ambient temperature (Cho et al., 2012, 2013). Shen et al. (2016) further demonstrated the superior performance of TiO2-based photocatalysts on Hg0 removal at relatively higher temperatures (>100 C). However, there have been few reports on the application of UV irradiation to the photocatalytic oxidation of Hg0. Our previous study (Yuan et al., 2012) found that joint application of nanofibers and UV could effectively enhance Hg0 removal. Nevertheless, the simulated flue gas system in this study lacks certain gases, which are essential in practice and indicates more in-depth study is needed. If the photocatalysis process can also efficiently remove Hg0 over cerium-based TiO2 nanofibers under UV irradiation, it is expected that it may eventually be developed into an effective technology for the removal of mercury in flue gas. Considering the problems mentioned above, it is of importance and interest to investigate the performance of cerium-based TiO2 nanofibers on Hg0 removal at low temperature during the
691
photocatalysis process. To the best of our knowledge, little effort has been directed to it until now. Besides, it should not be ignored that the emissions of Hg0 pose great challenges to the environment and public health, which incites urgent needs for efficient technologies to control Hg0. Thus, the emphases of this study were placed on (a) fabricating thermally stable and ordered nanofibers through electrospinning; (b) paving the way for understanding the effects of cerium-based TiO2 nanofibers on Hg0 removal by using photocatalysis process; (c) proposing possible strategies for reducing Hg0 emissions. The results will provide some theoretical guidance for follow-up studies and applications of this technology. 2. Experimental 2.1. Sample preparation and analysis A series of samples with different mass percentages of cerium (0e5 wt%) were synthesized by an electrospinning method. Accordingly, cerium (III) nitrate was used as the source of cerium. The detailed sample preparation procedures and specific dosages of reagents have been described in our earlier report (Yuan et al., 2012). To ensure sample comparability, pure TiO2 nanofibers were also prepared without adding the cerium source. The crystalline structures of samples were identified through Xray powder diffractometer (XRD, PANalytical B.V.X'Pert PRO). The evolution of the surface morphologies was investigated by field emission scanning electron microscope (FE-SEM, FEI Sirion200) and high-resolution transmission electron microscopy (HRTEM, Tecnai G2 F30). The diffuse reflectance UVeVis absorption spectra were determined with a UVeVis spectrophotometer (PerkinElmer Lambda35). The BET surface of samples was obtained via N2 isothermal adsorption (Micromeritics ASAP 2000), while the differences in functionalities of them were assessed with X-ray photoelectron spectroscopy (XPS, AXIS-ULTRA DLD-600W). Further analysis regarding the recombination of photo-induced charge carriers was carried out with photoluminescent spectra (PL, LabRAM HR800). Going deeper into the bandgap energies of the obtained samples from UVeVis spectra, they could be evaluated according to the following formula: ahn ¼ K(hn - Eg)n, where hv is the absorption energy, K is a constant correlating to the material, a could be replaced with the absorbance A, and n is correlated with the transition type of the semiconductor (1/2 and 2 for direct and indirect transition, respectively). Thus, the bandgap energies of the above samples were calculated by using (ahn/K)2 ¼ hn - Eg in the present study. 2.2. Experimental methods As shown in Fig. 1, the experimental set-up consisted mainly of a simulated flue gas-generating system, a specific vertical quartz reactor and an ultraviolet VM 3000 (Mercury Instrument Inc., Germany). The measuring method of VM 3000 is CVAAS based on the principle that elemental Hg vapor can effectively absorb the spectrum with the wavelength of 253.7 nm. The response time is 1 s. And the detection limit is 1 mg m3. All individual flue gas components came from cylinder gases and were precisely controlled by mass flow controllers, with a total flow rate of 1 L min1. Water vapor was generated from a heated water bath. The mercury permeation tube was placed in a U-shaped glass tube that was immersed in a water bath at a specific temperature to provide a constant feed of Hg0 concentration (~50 mg m3). When the photocatalytic reaction takes place in a gas-solid reactor, it is important to achieve higher light throughputs and lower pressure drops. Thus, an annular reactor was adopted in the present study, since this configuration has good prospects of
692
L. Wang et al. / Chemosphere 185 (2017) 690e698
Fig. 1. Schematic diagram of the experimental system.
implementation due to its geometry and its reduced volume (Palau et al., 2012). The quartz reactor consisted of an inner tube and an outer tube, which was wrapped with heating tape to maintain a constant flue gas temperature. It is well known that at ambient temperature, commercial TiO2 could reach a high efficiency over 90% (Jung et al., 2005; Suriyawong, 2009), while above 120 C, the increasing temperature inhibits the photocatalytic reactivity of TiO2 and reduces the photocatalytic removal efficiency of Hg0. Furthermore, according to previous studies (Shen et al., 2015, 2016), around 90% of Hg0 was oxidized over TiO2 at 120 C under the irradiation of 254 nm. Thus, 120 C was applied as the reaction temperature in the present study. Subsequently, a lamp was placed in the inner quartz tube. Irradiations were performed with a homeused UV lamp (Philips, TUV PL-S 9 W, Japan), which is commonly used for disinfection and sterilization. The primary wavelength of light emitted by the lamp was 253.7 nm (UVC). Accordingly, the light intensity was 3 mW/cm2, whereas the UV energy density (Ed) was fixed at 540 J/L. The reactor was placed inside an opaque box in order to make maximum use of UV energy, protect the operator from UV irradiation and prevent interference from sunlight and fluorescent light. Meanwhile, a high-pressure sodium lamp (Philips, SON-T 100 W, Japan) was used for visible light. This set-up allowed for the independent control of the on/off status of each lamp. Cold air was generated by an air pump efficiently to cool the lamp and maintain a constant reactor temperature. Besides, the reactor also has efficient contact between the photocatalyst and reactant gas with good transmission of UV-light (Lim et al., 2000). Moreover, Moussavi et al. (Moussavi and Mohseni, (2007) demonstrated that with a UV irradiation of 7.79 103 Wcm2, which is 2.6 times higher than our light intensity, the pressure drop maintained less than 3 mm H2O m1. It is reasonable to conclude that the present reactor and light source have good prospects of application. Additionally, interferences in Hg0 measurements by the empty reactor were verified to be negligible. 2.3. Data processing Before testing Hg0 removal over the photocatalysts, blank experiments were conducted to eliminate any possible interference
from the system. Furthermore, during each test, the gas stream firstly bypassed the fixed-bed reactor until a stable inlet concentration of Hg0 was obtained. The loss of Hg0 over photocatalysts could be related to its conversion to an oxidized Hg2þ form that is adsorbed, or to the adsorption of Hg0, or a combination of both (Eswaran and Stenger, 2005). The removal efficiency is calculated using the following equation:
E¼
DHg0 0 Hgin
¼
0 Hg 0 Hgin out 0 Hgin
100%
(1)
0 and Hg 0 represent the Hg0 concentration at the inlet Where Hgin out and outlet of the reactor, respectively. Photon utilization efficiency is the primary factor which determines the effectiveness of photolytic and photocatalytic oxidation processes. Therefore, an appropriate analysis of the photocatalytic oxidation of Hg0 should be carried out in terms of quantum yield:
F¼
⋕molecules decomposed A ¼ ⋕photons absorbed B
In the present study;
B¼
StI hy
A¼
ECin;Hg0 Qt NA M
(2)
(3)
(4)
Where E is the removal efficiency of Hg0, C0in,Hg the inlet concentration of Hg0 (mg/m3), Q the flue gas flow (L/min), t the running time per each experiment (in minutes), M the atomic mass of Hg (g/mol), and NA the Avogadro constant. For B, S is the irradiated area (cm2), h the Planck constant, y the frequency of light irradiation, and I the light intensity. I could be measured by the light radiometer.
L. Wang et al. / Chemosphere 185 (2017) 690e698
3. Results and discussion 3.1. Sample characteristics Fig. S1 (Electronic Supplementary Material) shows the XRD patterns of samples at 400 C. For TiO2, the sharp peak at 25.3 and the small peak at 27.5 correspond to anatase (101) and rutile (110), respectively. This finding indicates that the phase transformation of TiO2 from anatase to the thermodynamically more stable rutile phase has already occurred at 400 C. Correspondingly, the samples exhibit a mixture-type crystalline phase. After doping Ce into TiO2, the rutile diffraction peaks decreased, indicating that Ce doping inhibited the formation of rutile. Because the ionic radii of Ce4þ and Ce3þ are larger (0.093 nm and 0.103 nm, respectively) than that of Ti4þ (0.068 nm), it is difficult for cerium ions to enter the TiO2 lattice. And these ions may be present as the so-called second phase on the TiO2 surface. This second phase suppresses the crystallite growth of the rutile phase. Thus the phase transformation of TiO2 is inhibited (Xue et al., 2011). In contrast, Ti4þ easily enters into the crystal lattice of cerium oxide and substitutes for the cerium atom in the lattice (El-Bahy et al., 2009), forming TieOeCe bonds and promoting crystal lattice aberrance. However, no characteristic peaks of cerium species were detected. We tentatively inferred that they were highly dispersed in the samples. The representative FE-SEM micrographs of the nanofibers are shown in Fig. S2 (Electronic Supplementary Material). It is not difficult to find that the nanofibers showed uniform diameter and a random orientation because of the bending instability associated with the spinning jet. To probe into the microstructures of samples, Fig. 2 shows some typical TEM images of cerium-based TiO2 nanofibers. Each
693
individual nanofiber was composed of a large number of compactly packed nanoparticles. Meanwhile, the average fiber diameter values were calculated. The diameter of the pristine titania was approximately 80 nm. After Ce doping, it increased and ranged between 90 and 233 nm, indicating Ce doping has a small boosting effect on the diameter of fibers. In addition, the selected area electron diffraction (SAED) pattern confirms the good crystallinity of the sample (inset of Fig. 2 (a)). An HR-TEM analysis of pristine and Ce-doped TiO2 was further conducted. The lattice space was determined from HR-TEM images (Fig. 2 b, c, and d) to be 0.35 nm, 0.32 nm, 0.31 nm and 0.27 nm between the adjacent lattice planes of the nanofibers. Of these spaces, 0.35 nm and 0.32 nm correspond to the distance between (101) crystal planes of the anatase phase and the distance between (110) crystal planes of the rutile phase, while 0.31 nm and 0.27 nm corresponding to the (111) and (200) crystal planes of CeO2, respectively. Moreover, the EDS mappings of nanofibers (Fig. 2(eeh)) obviously show Ti and the corresponding Ce were homogeneously distributed throughout the fiber overall. The analysis of oxygen distribution along the fiber was added for a semi-quantitative evaluation. It is reasonable to conclude that CeO2 nanoparticles are highly dispersed in the nanofibers, which is in satisfactory agreement with the inference in XRD results. Besides morphology, the optical absorption properties were also identified. As illustrated in Fig. S3 (Electronic Supplementary Material), the bandgap energy initially increased from 3.17 to 3.20 eV for lower Ce concentration and then decreased from 3.20 to 3.07 eV as Ce content increases from 0.3 to 5. The noticed slight red shift (DEg ¼ 0.1 eV) in the bandgap is due to Ce doping in TiO2. This finding is interpreted as a possible evidence of good contact between TiO2 and cerium species, which indicates the formation of TieOeCe bonds. For TiO2, the absorption in the ultraviolet range is
Fig. 2. TEM images (aed) and EDS mappings (eeh) of samples. The inset of (a) shows the SAED image.
694
L. Wang et al. / Chemosphere 185 (2017) 690e698
associated with the excitation of the O 2p electron to the Ti 3d level (El-Bahy et al., 2009). Generally, the presence of redshift is caused by the new energy level in the bandgap. It is attributed to a chargetransfer transition between the cerium ion 4f level and the conduction or valence band of the host TiO2 (Xie and Yuan, 2003). Given that the central absorption peak of CeO2 should be situated at 400 nm (Song et al., 2007), the absorption over 400 nm is possible because of nanofibers. Subtle difference was observed among 0e0.5% samples, the absorption capability of which were superior to the rest of samples in the visible region. Sample with 0.3% Ce proposed relatively higher intensity in the visible region, which indicates that the doping of 0.3% Ce could lead to the best photoresponse as well as the highest photocatalytic activity. Furthermore, the results also confirm that the enhanced capacity of these ordered, uniform nanofibers makes it a promising photocatalyst for solar-driven applications.
3.2. Roles of interferential gas components In the gas purification field, the TiO2 photocatalytic technique also has great potential to remove SO2 and NO. SO2 removal efficiency could be well maintained at 100% over TiO2 immobilized on glass beads after the UV lights were applied (Wang and You, 2016). Liu et al. (2014) further demonstrated that the optimally integrated desulfurization and denitration efficiencies were observed on 15% MWCNTs/CueTiO2, 62% and 43% respectively. However, SO2, NO, and Hg0 are common gaseous pollutants in coal combustion flue gas. During Hg0 oxidation, competing occupation of photons and catalytic sites from SO2 and NO may exist, thereby significantly lowering the selectivity of the system and wasting photons. Thus, it is of importance and interest to evaluate the extent of this competing occupation. Experiments were designed to clarify the roles of SO2 and NO during Hg0 removal. Fig. 3 (a) displays the photocatalytic removal of SO2 and NO over samples. It shows clearly that almost no NO was oxidized, while the removal efficiency of SO2 was about 10%. Firstly, the feature of reaction was the possible dominated reason. SO2 is polar gas with high solubility in water (94 g/L). Although NO is nonpolar gas and difficult to dissolve in water, nitrogen dioxide, which is the oxidation product of NO, hydrolysis in water. Correspondingly, SO2 and NO are much easier to be removed in a photochemical gas-liquid reactor (Liu et al., 2010). In contrast, the mechanism involved in the present study is a gas-solid reaction with low water vapor, which is entirely different from the gasliquid reaction. Besides, it should not be ignored that UV intensity was invariable in the present study, which is another possible reason inhibited SO2 and NO oxidation. According to the BeerLambert law, the photochemical reaction yield is proportional to the UV irradiation intensity, which means that higher light intensity could facilitate removal efficiency. In summary, the low removal efficiencies of the above interferential gases indicate that the present experimental system had good selectivity for Hg0 removal. Besides the above tests, additional experiments were conducted to further investigate the removal performance of Hg0 under various SO2 and NO concentrations. As shown in Fig. 3 (b), with the increase of concentrations, NO promoted Hg0 removal, while SO2 displayed adverse effect. The boosting interaction may originate from the formation and accumulation of active nitrogen species such as NOþ and NO2 on the surface of nanofibers. The presence of Ce4þ on nanofibers has been reported to accelerate the oxidation of NO to NO2 (Gao et al., 2010). However, SO2 competed with Hg0 for the active sites on the catalyst, which resulted in the decrease of Hg0 removal efficiency.
Fig. 3. Roles of interferential gas components: (a) photocatalytic removal of NO and SO2 over nanofibers; (b) Hg0 removal under various concentrations of SO2 and NO.
3.3. Effect of calcination temperature on photocatalytic activity There is no denying that calcination temperature is crucial to Hg0 removal. Many efforts have been made to study the influence of calcination (Fan et al., 2012; Wang and Duan, 2011). As a higher calcination temperature tends to lead to agglomeration of nanofibers and decrease BET surface area and pore volume, 400 C, 500 C and 600 C were chosen as the operational temperatures. Fig. 4 depicts the effect of calcination temperature on Hg0 removal over nanofibers. Only a slight change in photocatalytic activity was observed for samples calcined at 500 C and 600 C, which were significantly lower than the efficiency found at 400 C. The results indicate that although the involvement of Ce can significantly enhance Hg0 removal, the extent of this promotion and the interaction between Ce and TiO2 may be strongly dependent on the calcination temperature. This might be attributed to the effect of calcination temperature on the surface area, vacant oxygen sites, and lattice oxygen (Jin et al., 2011). The increase in calcination temperature accompanied with crystal growth, which leads to the anatase-rutile transformation. Thus, the photocatalytic properties were influenced.
3.4. Effect of Ce-doped content on photocatalytic activity The performance of nanofibers on Hg0 removal was first evaluated under N2 and N2/O2 atmospheres, as shown in Fig. 5 (a). In
L. Wang et al. / Chemosphere 185 (2017) 690e698
695
Fig. 4. Effect of calcination temperature of catalysts on Hg0 removal.
this scenario, Hg0 removal was conducted with and without the catalyst under UV irradiation. In the absence of catalyst, Hg0 removal efficiencies are 0 under N2 and 40.78% under N2/O2 atmosphere. However, E of higher than 90% was reached by adding 0.3 g of samples. It is worth noting that, E over samples under N2 was almost the same as that in the presence of O2 without catalyst, which is a consequence of oxygen storage capability from ceriabased oxides. The results demonstrated that although UV and O2 are essential factors, the enhancement of Hg0 removal was more obviously related to the participation of samples. To clarify the photocatalytic activities of nanofibers, a set of tests were performed by varying the dopant content. As shown in Fig. 5 (b), doping 0.5% Ce into TiO2 in the dark and visible light irradiation significantly improved the performance by 15%. Hg0 removal efficiency increases with the increasing of Ce content in the TiO2 up to 0.3%, and catalyst with 0.3% Ce shows the highest Hg0 removal efficiency under UV irradiation (~91%). However, the further increase in Ce content leads to a decrease in the performance. More encouragingly, the photocatalytic Hg0 removal efficiencies of all samples are higher than that of pure TiO2 under UV irradiation. Going deeper into removal activities, Fig. 5 (c) illustrates the photocatalytic removal of Hg0 over pure TiO2 and the catalyst with 0.3% Ce in simulated coal combustion flue gas. Firstly, E over TiO2 under visible light irradiation was about the same as that in the dark, indicating that the photocatalytic activity of TiO2 was negligible under visible light irradiation. Even with the aid of UV, Hg0 removal efficiency was still very low, which indicated that TiO2 was insensitive to Hg0 removal. Secondly, doping 0.3% Ce significantly improved the efficiency, reaching 91.2% under UV irradiation, which is a result of better photoelectron-hole separation. This finding also makes up for the deficiency of our previous study and has fundamental value for the practical industrial application. The crystalline phase of TiO2 is a major factor in determining its activity. Anatase has been reported to show higher reactivity than rutile (Li et al., 2015). Our XRD results show that cerium doping inhibited the formation of rutile, which is beneficial for photocatalytic activity. Another significant factor affecting the performance is the surface area of samples. Table S1 (Electronic Supplementary Material) tabulates the BET surface area (SA), pore volume and pore size of samples. The BET SA increased significantly after doping with 0.3% Ce, which is 4 times greater than TiO2. Subsequently, it decreased gradually as the doping ratio was further
Fig. 5. Hg0 photocatalytic removal properties over catalysts: (a) blank experiment; (b) effect of Ce-doped content; (c) Hg0 removal efficiencies in simulated coal combustion flue gas.
696
L. Wang et al. / Chemosphere 185 (2017) 690e698
increased, indicating cerium concentration also affected pore structure of the sample. The large surface area could supply more active sites for Hg0 removal, which could also promote the separation efficiency of electron-hole pairs, resulting in a higher quantum efficiency. Meanwhile, light is also harvested more efficiently because of the large surface area (Yu et al., 2004). Besides crystalline phase and surface area, the efficient charge trapping, migration and transfer were further clarified. The fate of electron-hole pairs was investigated through wide survey PL spectra (Fig. S4, Electronic Supplementary Material) with an excitation wavelength of 325 nm. The lower the PL intensity, the lower the recombination rate of photo-induced electron-hole pairs, which usually relates to good photocatalytic activity (Yu et al., 2003). Compared with TiO2, peak positions of cerium-based samples are basically the same. It is easy to find that the PL intensities of samples are much lower than TiO2. And 0.3% treatment was the optimum one with the lowest intensity, which increased with the addition of Ce content. This increase occurred because the excess amount of Ce contributed to recombination centers on the surface of the samples. 3.5. Mechanism of Hg0 photocatalytic removal Based on the above Hg0 removal efficiencies, the quantum yields (QY, the number of photons utilized for the desired chemical reaction divided by the number of photons adsorbed by the catalyst) were calculated. The number for TiO2 was 16.26%. In particular, the addition of 0.3% Ce into TiO2 achieved a higher quantum yield of Hg0 reduction (44.69%). When minor cerium was doped into TiO2, the electron scavenger was formed, which may reduce the recombination of photoinduced electron-hole couples and thus lead to higher quantum yield. That means the combination of the present experimental set-up and cerium-based TiO2 nanofibers is effective in the removal of Hg0. Because of the highest Hg0 removal efficiency, the sample with 0.3% Ce was used for an extended test to investigate its stability. Fig. S5 (Electronic Supplementary Material) depicts Hg0 removal over the catalyst for 6 cycles. In each cycle, it was firstly exposed to visible light irradiation for 25 min, then UV irradiation for 25 min. It is not difficult to find that Hg0 removal efficiency was 25% under visible light and 91% under UV irradiation. Nevertheless, no apparent change in the efficiency was observed among those cycles, indicating that the sample showed promising stability for long-term use in removing Hg0. After photocatalysis, the determination of mercury forms was conducted with XPS. It can be seen from Fig. 6 (a) that an obvious peak found at 101.48 eV was attributed to Si 2p electron, for Si is the ingredient element of quartz sieve plate. Subsequently, the binding energy value at 105.43 eV was assigned to HgO, which is the oxidation product from Hg0. Moreover, no peak for Hg0 was found. This may be owing to the low Hg0 content on the surface of the sample, which is much lower than the detection limit of XPS analysis. An additional test was conducted in our previous study (Wang et al., 2017), which demonstrated that the adsorption capacity of our sample was low. In other words, both adsorption and oxidation exist in the photocatalytic reaction, while oxidation dominated the removal of Hg0. To provide more specific information, analysis of wide survey XPS spectra was conducted by peak integration, and the results were depicted in Fig. 6 (b). It reveals that both samples contain C, Ti and O elements. An additional small peak of Ce 3d was found for the sample with Ce. The XPS of Ti 2p shows a symmetric Ti 2p peak. The spin-orbit component (2p3/2 and 2p1/2) of the peak shows two shoulder peaks at ca. 458.5 eV and ca. 464.1 eV, respectively, indicating that Ti element primarily exists as the chemical state of Ti4þ. This is in satisfactory agreement with the findings in
Fig. 6. XPS survey spectra of: (a) Hg 4f for spent sample with 0.3% Ce; (b) Ti 2p, O 1s and Ce 3d.
literature (Biener et al., 2000; Ma et al., 2009). The O 1s peak is resolved into three component peaks which achieved an acceptable fit in the XPS spectra. The binding energy values at ca. 529.85 eV, ca. 532.02 eV, and ca. 533.44 eV can be assigned to lattice oxygen in TiO2, surface hydroxyl groups, and adsorbed oxygen, respectively. The Ce 3d photoelectron spectrum is rather complex and can be assigned to 3d3/2 spin-orbit states (labeled u) and 3d5/2 states (labeled v). However, a straight interpretation of the Ce 3d spectrum is attainable by following the intensity of the marked peaks in Fig. 6 (b), which shows the Ce 3d spectra for fresh and spent photocatalysts. The u000 /v000 doublet is caused by the primary photoemission from Ce4þ-O2. The u/v and u00 /v00 doublets are shakedown features resulting from the transfer of one or two electrons from a filled O 2p orbital to an empty Ce 4f orbital. The u0 /v' doublet is caused by the photoemission from Ce3þ (Ho et al., 2005; Liu et al., 2005). The Ce 3d spectra of fresh and spent samples denotes a mixture of Ce3þ/4þ oxidation states giving rise to a myriad of peaks, implying the coexistence of Ce3þ and Ce4þ on the surface of the ordered samples. Under UV light irradiation, the mechanism of Hg0 removal over TiO2 is considered to arise from the excitation of photoelectrons from the valance band (VB) to conduction band (CB), leaving holes in the VB. In the presence of O2 and H2O, the photo-generated
L. Wang et al. / Chemosphere 185 (2017) 690e698
697
Fig. 7. Schematic diagram of a proposed mechanism for Hg0 photocatalytic removal over cerium-based TiO2 nanofibers.
electrons and holes react with them to produce active oxygen 0 species such as O 2 and OH radicals, which are responsible for Hg oxidation (Yuan et al., 2012). But the electron-hole pair recombination occurs within nanoseconds, which weakens its activity sharply. After Ce is doped into TiO2, Ce4þ is the electron scavenger and efficiently traps the photo-excited electron. The electrons trapped in the Ce4þ/Ce3þ site are subsequently transferred to adsorbed O2 through the oxidation process, which also reduced the electron-hole pair recombination rate. Going deeper into the reaction evolution, it might be connected with a series of reactions, such as those listed below:
Ce4þ þ e /Ce3þ
(5)
Ce3þ þ O2 /O2 þ Ce4þ
(6)
Hg0 þ O2 /HgO
(7)
Hg 0 þ OH/HgO
(8)
Based on the analysis above, a schematic diagram of the mechanism is proposed as depicted in Fig. 7. Additionally, the thickness of the space charge layer (W) decreases with the increasing dopant concentration. When the value of W approaches the penetration depth of the light into the solid, all the adsorbed photons generate electron-hole pairs that are efficiently separated (Kumar and Devi, 2011), and the trap center is converted into the combination center. 4. Conclusions Cerium-based TiO2 nanofibers have been successfully prepared by the electrospinning method, and they were applied to the photocatalytic removal of Hg0 at low temperature. Results reported herein demonstrate that the homogeneous nanofibers possess a mixture-type crystalline phase. Ce ions were confirmed to be highly dispersed primarily in the form of metal oxides with a coexistence of Ce3þ/Ce4þ states. Sample with 0.3% Ce at 400 C showed the highest Hg0 removal capacity, reaching 91% under UV irradiation. There was no significant change in Hg0 removal during the long-term experiment, indicating that the samples were sufficiently stable for removing Hg0 from flue gas. Hg0 removal was more obviously related to the participation of sample. The
enhancement of photocatalytic activity was primarily attributed to the large surface area and the better photoelectron-hole separation. Consequently, it is a promising method to reduce the emissions of Hg0 during photocatalysis process by using cerium-based TiO2 nanofibers. dAcknowledgments This project was supported by the National Key Basic Research and Development Program (No. 2014CB238904) and the National Natural Science Foundation of China (51376074). The authors would like to express their thanks to the China Scholarship Council (CSC) and the Department of Analytical and Testing Center at Huazhong University of Science and Technology for providing financial support and the sample analysis, respectively. Appendix A. Supplementary data Supplementary data related to this article can be found at http:// dx.doi.org/10.1016/j.chemosphere.2017.07.049. References €umer, M., Wang, J., Madix, R.J., 2000. Electronic structure and growth of Biener, J., Ba vanadium on TiO2(110). Surf. Sci. 450, 12e26. Celebioglu, A., Sen, H.S., Durgun, E., Uyar, T., 2016. Molecular entrapment of volatile organic compounds (VOCs) by electrospun cyclodextrin nanofibers. Chemosphere 144, 736e744. Chen, S.S., Hsi, H.C., Nian, S.H., Chiu, C.H., 2014. Synthesis of N-doped TiO2 photocatalyst for low-concentration elemental mercury removal under various gas conditions. Appl. Catal. B-Environ. 160e161, 558e565. Cho, J.H., Eom, Y., Jeon, S.H., Lee, T.G., 2013. A pilot-scale TiO2 photocatalytic system for removing gas-phase elemental mercury at Hg-emitting facilities. J. Ind. Eng. Chem. 19, 144e149. Cho, J.H., Lee, T.G., Eom, Y., 2012. Gas-phase elemental mercury removal in a simulated combustion flue gas using TiO2 with fluorescent light. J. Air Waste Manag. 62, 1208e1213. Devi, L.G., Kumar, S.G., 2012. Exploring the critical dependence of adsorption of various dyes on the degradation rate using Ln3þ-TiO2 surface under UV/solar light. Appl. Surf. Sci. 261, 137e146. El-Bahy, Z.M., Ismail, A.A., Mohamed, R.M., 2009. Enhancement of titania by doping rare earth for photodegradation of organic dye (Direct Blue). J. Hazard. Mater. 166, 138e143. Eswaran, S., Stenger, H.G., 2005. Understanding mercury conversion in selective catalytic reduction (SCR) catalysts. Energy Fuel 19, 2328e2334. Fan, X., Li, C., Zeng, G., Zhang, X., Tao, S., Lu, P., Tan, Y., Luo, D., 2012. Hg0 removal from simulated flue gas over CeO2/HZSM-5. Energy Fuel 26, 2082e2089. Feng, D., Xu, S., Liu, G., 2015. Application of immobilized TiO2 photocatalysis to improve the inactivation of Heterosigma akashiwo in ballast water by intense pulsed light. Chemosphere 125, 102e107.
698
L. Wang et al. / Chemosphere 185 (2017) 690e698
Gao, X., Jiang, Y., Zhong, Y., Luo, Z., Cen, K., 2010. The activity and characterization of CeO2-TiO2 catalysts prepared by the sol-gel method for selective catalytic reduction of NO with NH3. J. Hazard. Mater. 174, 734e739. He, C., Shen, B., Chi, G., Li, F., 2016. Elemental mercury removal by CeO2/TiO2-PILCs under simulated coal-fired flue gas. Chem. Eng. J. 300, 1e8. Ho, C., Yu, J.C., Kwong, T., Mak, A.C., Lai, S., 2005. Morphology-controllable synthesis of mesoporous CeO2 nano- and microstructures. Chem. Mater. 17, 4514e4522. Huang, X., Feng, Y., Hu, C., Xiao, X., Yu, D., Zou, X., 2015. Mechanistic QSAR models for interpreting degradation rates of sulfonamides in UV-photocatalysis systems. Chemosphere 138, 183e189. Jeon, S.H., Eom, Y., Lee, T.G., 2008. Photocatalytic oxidation of gas-phase elemental mercury by nanotitanosilicate fibers. Chemosphere 71, 969e974. Jia, L., Dureau, R., Ko, V., Anthony, E.J., 2010. Oxidation of mercury under ultraviolet (UV) irradiation. Energy Fuel 24, 4351e4356. Jin, M., Kim, J.W., Kim, J.M., Jurng, J., Bae, G.N., Jeon, J.K., Park, Y.K., 2011. Effect of calcination temperature on the oxidation of benzene with ozone at low temperature over mesoporous a-Mn2O3. Powder Technol. 214, 458e462. Jones, A.P., Hoffmann, J.W., Smith, D.N., Feeley, T.J., Murphy, J.T., 2007. DOE/NETL's phase II mercury control technology field testing program: preliminary economic analysis of activated carbon injection. Environ. Sci. Technol. 41, 1365e1371. Jung, K.Y., Park, S.B., Anpo, M., 2005. Photoluminescence and photoactivity of titania particles prepared by the solegel technique: effect of calcination temperature. J. Photochem. Photobiol. A 170, 247e252. Kumar, S.G., Devi, L.G., 2011. Review on modified TiO2 photocatalysis under UV/ visible light: selected results and related mechanisms on interfacial charge carrier transfer dynamics. J. Phys. Chem. A 115, 13211e13241. Lasek, J., Yu, Y.H., Wu, J.C., 2013. Removal of NOx by photocatalytic processes. J. Photochem. Photobiol. C 14, 29e52. Li, B., Hao, Y., Zhang, B., Shao, X., Hu, L., 2017. A multifunctional noble-metal-free catalyst of CuO/TiO2 hybrid nanofibers. Appl. Catal. A-Gen. 531, 1e12. Li, H., Wu, S., Wu, C.Y., Wang, J., Li, L., Shih, K., 2015. SCR atmosphere induced reduction of oxidized mercury over CuO-CeO2/TiO2 catalyst. Environ. Sci. Technol. 49, 7373e7379. Lim, T.H., Jeong, S.M., Kim, S.D., Gyenis, J., 2000. Photocatalytic decomposition of NO by TiO2 particles. J. Photochem. Photobiol. A 134, 209e217. Liu, H., Yu, X., Yang, H., 2014. The integrated photocatalytic removal of SO2 and NO using Cu doped titanium dioxide supported by multi-walled carbon nanotubes. Chem. Eng. J. 243, 465e472. Liu, Y., Zhang, J., Sheng, C., Zhang, Y., Zhao, L., 2010. Wet removal of sulfur dioxide and nitric oxide from simulated coal-fired flue gas by UV/H2O2 advanced oxidation process. Energy Fuel 24, 4931e4936. Liu, Z., Guo, B., Hong, L., Jiang, H., 2005. Preparation and characterization of cerium oxide doped TiO2 nanoparticles. J. Phys. Chem. Solids 6, 161e167. Ma, T.Y., Cao, J.L., Shao, G.S., Zhang, X.J., Yuan, Z.Y., 2009. Hierarchically structured squama-like cerium-doped titania: synthesis, photoactivity, and catalytic CO oxidation. J. Phys. Chem. C 113, 16658e16667. Moussavi, G., Mohseni, M., 2007. Using UV pretreatment to enhance biofiltration of mixtures of aromatic VOCs. J. Hazard. Mater. 144, 59e66. Palau, J., Colomer, M., Penya-Roja, J.M., Martínez-Soria, V., 2012. Photodegradation of toluene, m-xylene, and n-butyl acetate and their mixtures over TiO2 catalyst on glass fibers. Ind. Eng. Chem. Res. 51, 5986e5994. Park, J.A., Kim, S.B., 2016. Antimicrobial filtration with electrospun poly (vinyl alcohol) nanofibers containing benzyl triethylammonium chloride: immersion, leaching, toxicity, and filtration tests. Chemosphere 167, 469e477. Salehifar, N., Nikfarjam, A., 2017. Improvement the visible light photocatalytic activity of gold nanoparticle, Co2O3 and nitrogen doped TiO2 nanofibers. Mater. Lett. 188, 59e62.
Shen, H., Ie, I.R., Yuan, C.S., Hung, C.H., 2016. The enhancement of photo-oxidation efficiency of elemental mercury by immobilized WO3/TiO2 at high temperatures. Appl. Catal. B-Environ. 195, 90e103. Shen, H., Ie, I.R., Yuan, C.S., Hung, C.H., Chen, W.H., Luo, J., Jen, Y.H., 2015. Enhanced photocatalytic oxidation of gaseous elemental mercury by TiO2 in a high temperature environment. J. Hazard. Mater. 289, 235e243. Silva, A.M., Silva, C.G., Dra zi c, G., Faria, J.L., 2009. Ce-doped TiO2 for photocatalytic degradation of chlorophenol. Catal. Today 144, 13e18. Song, S., Xu, L., He, Z., Chen, J., Xiao, X., Yan, B., 2007. Mechanism of the photocatalytic degradation of C.I. reactive black 5 at pH 12.0 using SrTiO3/CeO2 as the catalyst. Environ. Sci. Technol. 41, 5846e5853. €rl, R., Belo, L., Shah, K., Stanger, R., Giniyatullin, R., Maier, J., Wall, T., Spo Scheffknecht, G., 2014. Mercury emissions and removal by ash in coal-fired oxyfuel combustion. Energy Fuel 28, 123e135. Suriyawong, A., 2009. Mercury capture by nano-structured titanium dioxide sorbent during coal combustion: lab-scale to pilot-scale studies. Aerosol Air Qual. Res. 9, 394e403. Wang, H., You, C., 2016. Photocatalytic removal of low concentration SO2 by titanium dioxide. Chem. Eng. J. 292, 199e206. Wang, H., Zhou, S., Xiao, L., Wang, Y., Liu, Y., Wu, Z., 2011. Titania nanotubesda unique photocatalyst and adsorbent for elemental mercury removal. Catal. Today 175, 202e208. Wang, L., Zhao, Y., Zhang, J., 2017. Comprehensive evaluation of mercury photocatalytic oxidation by cerium-based TiO2 nanofibers. Ind. Eng. Chem. Res. 56, 3804e3812. Wang, P., Su, S., Xiang, J., You, H., Cao, F., Sun, L., Hu, S., Zhang, Y., 2014. Catalytic oxidation of Hg0 by MnOx-CeO2/gamma-Al2O3 catalyst at low temperatures. Chemosphere 101, 49e54. Wang, Y., Duan, Y., 2011. Effect of manganese ions on the structure of Ca(OH)2 and mercury adsorption performance of Mnxþ/Ca(OH)2 composites. Energy Fuel 25, 1553e1558. Weng, X., Mei, R., Shi, M., Kong, Q., Liu, Y., Wu, Z., 2015. CePO4 catalyst for elemental mercury removal in simulated coal-fired flue gas. Energy Fuel 29, 3359e3365. Wu, H., Li, C., Zhao, L., Zhang, J., Zeng, G., Xie, Y., Zhang, X., Wang, Y., 2015. Removal of gaseous elemental mercury by cylindrical activated coke loaded with CoOxCeO2 from simulated coal combustion flue gas. Energy Fuel 29, 6747e6757. Xie, Y., Yuan, C., 2003. Visible-light responsive cerium ion modified titania sol and nanocrystallites for X-3B dye photodegradation. Appl. Catal. B-Environ. 46, 251e259. Xu, Y., Chen, H., Zeng, Z., Lei, B., 2006. Investigation on mechanism of photocatalytic activity enhancement of nanometer cerium-doped titania. Appl. Surf. Sci. 252, 8565e8570. Xue, W., Zhang, G., Xu, X., Yang, X., Liu, C., Xu, Y., 2011. Preparation of titania nanotubes doped with cerium and their photocatalytic activity for glyphosate. Chem. Eng. J. 167, 397e402. Yu, J.C., Wang, X., Fu, X., 2004. Pore-wall chemistry and photocatalytic activity of mesoporous titania molecular sieve films. Chem. Mater. 16, 1523e1530. Yu, J.G., Yu, H.G., Cheng, B., Zhao, X.J., Yu, J.C., Ho, W.K., 2003. The effect of calcination temperature on the surface microstructure and photocatalytic activity of TiO2 thin films prepared by liquid phase deposition. J. Phys. Chem. B 107, 13871e13879. Yuan, Y., Zhao, Y., Li, H., Li, Y., Gao, X., Zheng, C., Zhang, J., 2012. Electrospun metal oxide-TiO2 nanofibers for elemental mercury removal from flue gas. J. Hazard. Mater. 227e228, 427e435. Zhou, C., Sun, L., Zhang, A., Wu, X., Ma, C., Su, S., Hu, S., Xiang, J., 2015. Fe3-xCuxO4 as highly active heterogeneous Fenton-like catalysts toward elemental mercury removal. Chemosphere 125, 16e24.
Contents lists available at ScienceDirect
Chemosphere journal homepage: www.elsevier.com/locate/chemosphere
Electrospun cerium-based TiO2 nanofibers for photocatalytic oxidation of elemental mercury in coal combustion flue gas Lulu Wang, Yongchun Zhao**, Junying Zhang* State Key Laboratory of Coal Combustion, Huazhong University of Science & Technology, Wuhan 430074, PR China
h i g h l i g h t s Cerium-based TiO2 nanofibers were successfully synthesized by electrospinning. Hg0 photocatalytic oxidation was investigated over cerium-based TiO2 nanofibers. Roles of interferential gas components on Hg0 removal were clarified. High removal efficiency of Hg0 were observed for sample with 0.3% Ce. High stability of samples was observed for long-term use.
a r t i c l e i n f o
a b s t r a c t
Article history: Received 13 December 2016 Received in revised form 6 July 2017 Accepted 11 July 2017 Available online 12 July 2017
Photocatalytic oxidation is an attractive method for Hg-rich flue gas treatment. In the present study, a novel cerium-based TiO2 nanofibers was prepared and selected as the catalyst to remove mercury in flue gas. Accordingly, physical/chemical properties of those nanofibers were clarified. The effects of some important parameters, such as calcination temperature, cerium dopant content and different illumination conditions on the removal of Hg0 using the photocatalysis process were investigated. In addition, the removal mechanism of Hg0 over cerium-based TiO2 nanofibers focused on UV irradiation was proposed. The results show that catalyst which was calcined at 400 C exhibited better performance. The addition of 0.3 wt% Ce into TiO2 led to the highest removal efficiency at 91% under UV irradiation. As-prepared samples showed promising stability for long-term use in the test. However, the photoluminescence intensity of nanofibers incorporating ceria was significantly lower than TiO2, which was attributed to better photoelectron-hole separation. Although UV and O2 are essential factors, the enhancement of Hg0 removal is more obviously related to the participation of catalyst. The coexistence of Ce3þ and Ce4þ, which leads to the efficient oxidation of Hg0, was detected on samples. Hg2þ is the final product in the reaction of Hg0 removal. As a consequence, the emissions of Hg0 from flue gas can be significantly suppressed. These indicate that combining photocatalysis technology with cerium-based TiO2 nanofibers is a promising strategy for reducing Hg0 efficiently. © 2017 Elsevier Ltd. All rights reserved.
Handling Editor: Jun Huang Keywords: Electrospinning Nanofiber Photocatalytic oxidation Elemental mercury
1. Introduction Mercury has been identified as one of the most hazardous heavy metals to public health and the environment because of its volatility, persistence, and bioaccumulation in the biosphere (Zhou et al., 2015). According to the 2013 global mercury assessment by
* Corresponding author. ** Corresponding author. E-mail addresses: [email protected] (J. Zhang).
(Y.
http://dx.doi.org/10.1016/j.chemosphere.2017.07.049 0045-6535/© 2017 Elsevier Ltd. All rights reserved.
Zhao),
[email protected]
the United Nations Environment Programme (UNEP), coal burning is one of the most significant anthropogenic sources of mercury emissions to the atmosphere. Mercury from coal-fired flue gas is present in the following three primary forms: elemental mercury (Hg0), oxidized mercury (Hg2þ) and particulate-bound mercury (Hgp). Among them, Hg0 is more difficult to capture with existing air pollution control devices because of its inertness, high volatility € rl et al., 2014). This arouses urgent and low solubility in water (Spo needs for efficient technologies to reduce the Hg0 content in the flue gas. Adsorption and oxidation are the two primary methods used to eliminate Hg0. The injection of activated carbon upstream of the electrostatic precipitators (ESPs) or fabric filters (FFs) has
L. Wang et al. / Chemosphere 185 (2017) 690e698
shown to be an efficient adsorption method (Jia et al., 2010). Unfortunately, the high cost hinders its widespread applications (Jones et al., 2007). Therefore, significant attention has been paid to Hg0 oxidation. Among the different alternative routes currently available for Hg0 oxidation, photocatalysis has been proven to be one of the most promising methods, which can meet all the requirements mentioned above (Chen et al., 2014; Lasek et al., 2013; Wang et al., 2011). TiO2 is the most promising material in photocatalysis process (Feng et al., 2015; Huang et al., 2015). It has been used extensively for its stability, low cost, lack of toxicity and excellent photocatalytic activity. However, the photogenerated electron-hole pair recombination occurs within nanoseconds. To overcome this problem, modification materials have been extensively investigated. Of all these investigated materials, it cannot be ignored that owing to their f electron and multi-electron configuration, rare earth oxides have been found to have polymorphs with strong adsorption and good thermal stability (Xu et al., 2006), which can help improve the performance of TiO2 (Devi and Kumar, 2012; Wang et al., 2014). In particular, cerium oxides have attracted much attention due to the optical and catalytic properties associated with the redox couple Ce3þ/Ce4þ (Silva et al., 2009). The Hg0 removal performance over CePO4 has been found superior to commercial SCR catalyst in the temperature range of 300e400 C (Weng et al., 2015). Wu et al. (2015) provided a fundamental insight into Hg0 oxidation over CoCe/AC catalyst-sorbents and believed that Co4.5Ce6/AC had a promising industrial application prospect. He et al. (2016) further demonstrated that CeO2/TiO2-PILC catalysts had high specific surface area and distinct interlamellar structure during Hg0 removal process. However, these studies only considered high reaction temperature and no light was involved. The performance of cerium-based TiO2 on Hg0 photocatalytic removal at low temperature is still enigmatic. As for the morphology of the photocatalyst, it should be borne in mind that compared with traditional powder catalysts, nanofibers have received great interest in terms of development and industrial applications (Celebioglu et al., 2016; Park and Kim, 2016). Various methods have been investigated to fabricate continuous onedimensional (1D) TiO2 nanofibers, including self-assembly, phase separation, attraction and electrospinning (Salehifar and Nikfarjam, 2017). Among them, electrospinning has been deemed as the most versatile and efficient technique due to its high throughput during a continuous process, mild conditions, and the flexibility in controlling the composition, diameter and porosity (Li et al., 2017). Nano titanosilicate fibers appeared to have high thermal stability and a large number of terminal oxygens, which make them suitable for Hg0 photocatalytic removal (Jeon et al., 2008). The photocatalytic reactivity of TiO2-based materials for Hg0 removal under near-UV irradiation has been systematically investigated at ambient temperature (Cho et al., 2012, 2013). Shen et al. (2016) further demonstrated the superior performance of TiO2-based photocatalysts on Hg0 removal at relatively higher temperatures (>100 C). However, there have been few reports on the application of UV irradiation to the photocatalytic oxidation of Hg0. Our previous study (Yuan et al., 2012) found that joint application of nanofibers and UV could effectively enhance Hg0 removal. Nevertheless, the simulated flue gas system in this study lacks certain gases, which are essential in practice and indicates more in-depth study is needed. If the photocatalysis process can also efficiently remove Hg0 over cerium-based TiO2 nanofibers under UV irradiation, it is expected that it may eventually be developed into an effective technology for the removal of mercury in flue gas. Considering the problems mentioned above, it is of importance and interest to investigate the performance of cerium-based TiO2 nanofibers on Hg0 removal at low temperature during the
691
photocatalysis process. To the best of our knowledge, little effort has been directed to it until now. Besides, it should not be ignored that the emissions of Hg0 pose great challenges to the environment and public health, which incites urgent needs for efficient technologies to control Hg0. Thus, the emphases of this study were placed on (a) fabricating thermally stable and ordered nanofibers through electrospinning; (b) paving the way for understanding the effects of cerium-based TiO2 nanofibers on Hg0 removal by using photocatalysis process; (c) proposing possible strategies for reducing Hg0 emissions. The results will provide some theoretical guidance for follow-up studies and applications of this technology. 2. Experimental 2.1. Sample preparation and analysis A series of samples with different mass percentages of cerium (0e5 wt%) were synthesized by an electrospinning method. Accordingly, cerium (III) nitrate was used as the source of cerium. The detailed sample preparation procedures and specific dosages of reagents have been described in our earlier report (Yuan et al., 2012). To ensure sample comparability, pure TiO2 nanofibers were also prepared without adding the cerium source. The crystalline structures of samples were identified through Xray powder diffractometer (XRD, PANalytical B.V.X'Pert PRO). The evolution of the surface morphologies was investigated by field emission scanning electron microscope (FE-SEM, FEI Sirion200) and high-resolution transmission electron microscopy (HRTEM, Tecnai G2 F30). The diffuse reflectance UVeVis absorption spectra were determined with a UVeVis spectrophotometer (PerkinElmer Lambda35). The BET surface of samples was obtained via N2 isothermal adsorption (Micromeritics ASAP 2000), while the differences in functionalities of them were assessed with X-ray photoelectron spectroscopy (XPS, AXIS-ULTRA DLD-600W). Further analysis regarding the recombination of photo-induced charge carriers was carried out with photoluminescent spectra (PL, LabRAM HR800). Going deeper into the bandgap energies of the obtained samples from UVeVis spectra, they could be evaluated according to the following formula: ahn ¼ K(hn - Eg)n, where hv is the absorption energy, K is a constant correlating to the material, a could be replaced with the absorbance A, and n is correlated with the transition type of the semiconductor (1/2 and 2 for direct and indirect transition, respectively). Thus, the bandgap energies of the above samples were calculated by using (ahn/K)2 ¼ hn - Eg in the present study. 2.2. Experimental methods As shown in Fig. 1, the experimental set-up consisted mainly of a simulated flue gas-generating system, a specific vertical quartz reactor and an ultraviolet VM 3000 (Mercury Instrument Inc., Germany). The measuring method of VM 3000 is CVAAS based on the principle that elemental Hg vapor can effectively absorb the spectrum with the wavelength of 253.7 nm. The response time is 1 s. And the detection limit is 1 mg m3. All individual flue gas components came from cylinder gases and were precisely controlled by mass flow controllers, with a total flow rate of 1 L min1. Water vapor was generated from a heated water bath. The mercury permeation tube was placed in a U-shaped glass tube that was immersed in a water bath at a specific temperature to provide a constant feed of Hg0 concentration (~50 mg m3). When the photocatalytic reaction takes place in a gas-solid reactor, it is important to achieve higher light throughputs and lower pressure drops. Thus, an annular reactor was adopted in the present study, since this configuration has good prospects of
692
L. Wang et al. / Chemosphere 185 (2017) 690e698
Fig. 1. Schematic diagram of the experimental system.
implementation due to its geometry and its reduced volume (Palau et al., 2012). The quartz reactor consisted of an inner tube and an outer tube, which was wrapped with heating tape to maintain a constant flue gas temperature. It is well known that at ambient temperature, commercial TiO2 could reach a high efficiency over 90% (Jung et al., 2005; Suriyawong, 2009), while above 120 C, the increasing temperature inhibits the photocatalytic reactivity of TiO2 and reduces the photocatalytic removal efficiency of Hg0. Furthermore, according to previous studies (Shen et al., 2015, 2016), around 90% of Hg0 was oxidized over TiO2 at 120 C under the irradiation of 254 nm. Thus, 120 C was applied as the reaction temperature in the present study. Subsequently, a lamp was placed in the inner quartz tube. Irradiations were performed with a homeused UV lamp (Philips, TUV PL-S 9 W, Japan), which is commonly used for disinfection and sterilization. The primary wavelength of light emitted by the lamp was 253.7 nm (UVC). Accordingly, the light intensity was 3 mW/cm2, whereas the UV energy density (Ed) was fixed at 540 J/L. The reactor was placed inside an opaque box in order to make maximum use of UV energy, protect the operator from UV irradiation and prevent interference from sunlight and fluorescent light. Meanwhile, a high-pressure sodium lamp (Philips, SON-T 100 W, Japan) was used for visible light. This set-up allowed for the independent control of the on/off status of each lamp. Cold air was generated by an air pump efficiently to cool the lamp and maintain a constant reactor temperature. Besides, the reactor also has efficient contact between the photocatalyst and reactant gas with good transmission of UV-light (Lim et al., 2000). Moreover, Moussavi et al. (Moussavi and Mohseni, (2007) demonstrated that with a UV irradiation of 7.79 103 Wcm2, which is 2.6 times higher than our light intensity, the pressure drop maintained less than 3 mm H2O m1. It is reasonable to conclude that the present reactor and light source have good prospects of application. Additionally, interferences in Hg0 measurements by the empty reactor were verified to be negligible. 2.3. Data processing Before testing Hg0 removal over the photocatalysts, blank experiments were conducted to eliminate any possible interference
from the system. Furthermore, during each test, the gas stream firstly bypassed the fixed-bed reactor until a stable inlet concentration of Hg0 was obtained. The loss of Hg0 over photocatalysts could be related to its conversion to an oxidized Hg2þ form that is adsorbed, or to the adsorption of Hg0, or a combination of both (Eswaran and Stenger, 2005). The removal efficiency is calculated using the following equation:
E¼
DHg0 0 Hgin
¼
0 Hg 0 Hgin out 0 Hgin
100%
(1)
0 and Hg 0 represent the Hg0 concentration at the inlet Where Hgin out and outlet of the reactor, respectively. Photon utilization efficiency is the primary factor which determines the effectiveness of photolytic and photocatalytic oxidation processes. Therefore, an appropriate analysis of the photocatalytic oxidation of Hg0 should be carried out in terms of quantum yield:
F¼
⋕molecules decomposed A ¼ ⋕photons absorbed B
In the present study;
B¼
StI hy
A¼
ECin;Hg0 Qt NA M
(2)
(3)
(4)
Where E is the removal efficiency of Hg0, C0in,Hg the inlet concentration of Hg0 (mg/m3), Q the flue gas flow (L/min), t the running time per each experiment (in minutes), M the atomic mass of Hg (g/mol), and NA the Avogadro constant. For B, S is the irradiated area (cm2), h the Planck constant, y the frequency of light irradiation, and I the light intensity. I could be measured by the light radiometer.
L. Wang et al. / Chemosphere 185 (2017) 690e698
3. Results and discussion 3.1. Sample characteristics Fig. S1 (Electronic Supplementary Material) shows the XRD patterns of samples at 400 C. For TiO2, the sharp peak at 25.3 and the small peak at 27.5 correspond to anatase (101) and rutile (110), respectively. This finding indicates that the phase transformation of TiO2 from anatase to the thermodynamically more stable rutile phase has already occurred at 400 C. Correspondingly, the samples exhibit a mixture-type crystalline phase. After doping Ce into TiO2, the rutile diffraction peaks decreased, indicating that Ce doping inhibited the formation of rutile. Because the ionic radii of Ce4þ and Ce3þ are larger (0.093 nm and 0.103 nm, respectively) than that of Ti4þ (0.068 nm), it is difficult for cerium ions to enter the TiO2 lattice. And these ions may be present as the so-called second phase on the TiO2 surface. This second phase suppresses the crystallite growth of the rutile phase. Thus the phase transformation of TiO2 is inhibited (Xue et al., 2011). In contrast, Ti4þ easily enters into the crystal lattice of cerium oxide and substitutes for the cerium atom in the lattice (El-Bahy et al., 2009), forming TieOeCe bonds and promoting crystal lattice aberrance. However, no characteristic peaks of cerium species were detected. We tentatively inferred that they were highly dispersed in the samples. The representative FE-SEM micrographs of the nanofibers are shown in Fig. S2 (Electronic Supplementary Material). It is not difficult to find that the nanofibers showed uniform diameter and a random orientation because of the bending instability associated with the spinning jet. To probe into the microstructures of samples, Fig. 2 shows some typical TEM images of cerium-based TiO2 nanofibers. Each
693
individual nanofiber was composed of a large number of compactly packed nanoparticles. Meanwhile, the average fiber diameter values were calculated. The diameter of the pristine titania was approximately 80 nm. After Ce doping, it increased and ranged between 90 and 233 nm, indicating Ce doping has a small boosting effect on the diameter of fibers. In addition, the selected area electron diffraction (SAED) pattern confirms the good crystallinity of the sample (inset of Fig. 2 (a)). An HR-TEM analysis of pristine and Ce-doped TiO2 was further conducted. The lattice space was determined from HR-TEM images (Fig. 2 b, c, and d) to be 0.35 nm, 0.32 nm, 0.31 nm and 0.27 nm between the adjacent lattice planes of the nanofibers. Of these spaces, 0.35 nm and 0.32 nm correspond to the distance between (101) crystal planes of the anatase phase and the distance between (110) crystal planes of the rutile phase, while 0.31 nm and 0.27 nm corresponding to the (111) and (200) crystal planes of CeO2, respectively. Moreover, the EDS mappings of nanofibers (Fig. 2(eeh)) obviously show Ti and the corresponding Ce were homogeneously distributed throughout the fiber overall. The analysis of oxygen distribution along the fiber was added for a semi-quantitative evaluation. It is reasonable to conclude that CeO2 nanoparticles are highly dispersed in the nanofibers, which is in satisfactory agreement with the inference in XRD results. Besides morphology, the optical absorption properties were also identified. As illustrated in Fig. S3 (Electronic Supplementary Material), the bandgap energy initially increased from 3.17 to 3.20 eV for lower Ce concentration and then decreased from 3.20 to 3.07 eV as Ce content increases from 0.3 to 5. The noticed slight red shift (DEg ¼ 0.1 eV) in the bandgap is due to Ce doping in TiO2. This finding is interpreted as a possible evidence of good contact between TiO2 and cerium species, which indicates the formation of TieOeCe bonds. For TiO2, the absorption in the ultraviolet range is
Fig. 2. TEM images (aed) and EDS mappings (eeh) of samples. The inset of (a) shows the SAED image.
694
L. Wang et al. / Chemosphere 185 (2017) 690e698
associated with the excitation of the O 2p electron to the Ti 3d level (El-Bahy et al., 2009). Generally, the presence of redshift is caused by the new energy level in the bandgap. It is attributed to a chargetransfer transition between the cerium ion 4f level and the conduction or valence band of the host TiO2 (Xie and Yuan, 2003). Given that the central absorption peak of CeO2 should be situated at 400 nm (Song et al., 2007), the absorption over 400 nm is possible because of nanofibers. Subtle difference was observed among 0e0.5% samples, the absorption capability of which were superior to the rest of samples in the visible region. Sample with 0.3% Ce proposed relatively higher intensity in the visible region, which indicates that the doping of 0.3% Ce could lead to the best photoresponse as well as the highest photocatalytic activity. Furthermore, the results also confirm that the enhanced capacity of these ordered, uniform nanofibers makes it a promising photocatalyst for solar-driven applications.
3.2. Roles of interferential gas components In the gas purification field, the TiO2 photocatalytic technique also has great potential to remove SO2 and NO. SO2 removal efficiency could be well maintained at 100% over TiO2 immobilized on glass beads after the UV lights were applied (Wang and You, 2016). Liu et al. (2014) further demonstrated that the optimally integrated desulfurization and denitration efficiencies were observed on 15% MWCNTs/CueTiO2, 62% and 43% respectively. However, SO2, NO, and Hg0 are common gaseous pollutants in coal combustion flue gas. During Hg0 oxidation, competing occupation of photons and catalytic sites from SO2 and NO may exist, thereby significantly lowering the selectivity of the system and wasting photons. Thus, it is of importance and interest to evaluate the extent of this competing occupation. Experiments were designed to clarify the roles of SO2 and NO during Hg0 removal. Fig. 3 (a) displays the photocatalytic removal of SO2 and NO over samples. It shows clearly that almost no NO was oxidized, while the removal efficiency of SO2 was about 10%. Firstly, the feature of reaction was the possible dominated reason. SO2 is polar gas with high solubility in water (94 g/L). Although NO is nonpolar gas and difficult to dissolve in water, nitrogen dioxide, which is the oxidation product of NO, hydrolysis in water. Correspondingly, SO2 and NO are much easier to be removed in a photochemical gas-liquid reactor (Liu et al., 2010). In contrast, the mechanism involved in the present study is a gas-solid reaction with low water vapor, which is entirely different from the gasliquid reaction. Besides, it should not be ignored that UV intensity was invariable in the present study, which is another possible reason inhibited SO2 and NO oxidation. According to the BeerLambert law, the photochemical reaction yield is proportional to the UV irradiation intensity, which means that higher light intensity could facilitate removal efficiency. In summary, the low removal efficiencies of the above interferential gases indicate that the present experimental system had good selectivity for Hg0 removal. Besides the above tests, additional experiments were conducted to further investigate the removal performance of Hg0 under various SO2 and NO concentrations. As shown in Fig. 3 (b), with the increase of concentrations, NO promoted Hg0 removal, while SO2 displayed adverse effect. The boosting interaction may originate from the formation and accumulation of active nitrogen species such as NOþ and NO2 on the surface of nanofibers. The presence of Ce4þ on nanofibers has been reported to accelerate the oxidation of NO to NO2 (Gao et al., 2010). However, SO2 competed with Hg0 for the active sites on the catalyst, which resulted in the decrease of Hg0 removal efficiency.
Fig. 3. Roles of interferential gas components: (a) photocatalytic removal of NO and SO2 over nanofibers; (b) Hg0 removal under various concentrations of SO2 and NO.
3.3. Effect of calcination temperature on photocatalytic activity There is no denying that calcination temperature is crucial to Hg0 removal. Many efforts have been made to study the influence of calcination (Fan et al., 2012; Wang and Duan, 2011). As a higher calcination temperature tends to lead to agglomeration of nanofibers and decrease BET surface area and pore volume, 400 C, 500 C and 600 C were chosen as the operational temperatures. Fig. 4 depicts the effect of calcination temperature on Hg0 removal over nanofibers. Only a slight change in photocatalytic activity was observed for samples calcined at 500 C and 600 C, which were significantly lower than the efficiency found at 400 C. The results indicate that although the involvement of Ce can significantly enhance Hg0 removal, the extent of this promotion and the interaction between Ce and TiO2 may be strongly dependent on the calcination temperature. This might be attributed to the effect of calcination temperature on the surface area, vacant oxygen sites, and lattice oxygen (Jin et al., 2011). The increase in calcination temperature accompanied with crystal growth, which leads to the anatase-rutile transformation. Thus, the photocatalytic properties were influenced.
3.4. Effect of Ce-doped content on photocatalytic activity The performance of nanofibers on Hg0 removal was first evaluated under N2 and N2/O2 atmospheres, as shown in Fig. 5 (a). In
L. Wang et al. / Chemosphere 185 (2017) 690e698
695
Fig. 4. Effect of calcination temperature of catalysts on Hg0 removal.
this scenario, Hg0 removal was conducted with and without the catalyst under UV irradiation. In the absence of catalyst, Hg0 removal efficiencies are 0 under N2 and 40.78% under N2/O2 atmosphere. However, E of higher than 90% was reached by adding 0.3 g of samples. It is worth noting that, E over samples under N2 was almost the same as that in the presence of O2 without catalyst, which is a consequence of oxygen storage capability from ceriabased oxides. The results demonstrated that although UV and O2 are essential factors, the enhancement of Hg0 removal was more obviously related to the participation of samples. To clarify the photocatalytic activities of nanofibers, a set of tests were performed by varying the dopant content. As shown in Fig. 5 (b), doping 0.5% Ce into TiO2 in the dark and visible light irradiation significantly improved the performance by 15%. Hg0 removal efficiency increases with the increasing of Ce content in the TiO2 up to 0.3%, and catalyst with 0.3% Ce shows the highest Hg0 removal efficiency under UV irradiation (~91%). However, the further increase in Ce content leads to a decrease in the performance. More encouragingly, the photocatalytic Hg0 removal efficiencies of all samples are higher than that of pure TiO2 under UV irradiation. Going deeper into removal activities, Fig. 5 (c) illustrates the photocatalytic removal of Hg0 over pure TiO2 and the catalyst with 0.3% Ce in simulated coal combustion flue gas. Firstly, E over TiO2 under visible light irradiation was about the same as that in the dark, indicating that the photocatalytic activity of TiO2 was negligible under visible light irradiation. Even with the aid of UV, Hg0 removal efficiency was still very low, which indicated that TiO2 was insensitive to Hg0 removal. Secondly, doping 0.3% Ce significantly improved the efficiency, reaching 91.2% under UV irradiation, which is a result of better photoelectron-hole separation. This finding also makes up for the deficiency of our previous study and has fundamental value for the practical industrial application. The crystalline phase of TiO2 is a major factor in determining its activity. Anatase has been reported to show higher reactivity than rutile (Li et al., 2015). Our XRD results show that cerium doping inhibited the formation of rutile, which is beneficial for photocatalytic activity. Another significant factor affecting the performance is the surface area of samples. Table S1 (Electronic Supplementary Material) tabulates the BET surface area (SA), pore volume and pore size of samples. The BET SA increased significantly after doping with 0.3% Ce, which is 4 times greater than TiO2. Subsequently, it decreased gradually as the doping ratio was further
Fig. 5. Hg0 photocatalytic removal properties over catalysts: (a) blank experiment; (b) effect of Ce-doped content; (c) Hg0 removal efficiencies in simulated coal combustion flue gas.
696
L. Wang et al. / Chemosphere 185 (2017) 690e698
increased, indicating cerium concentration also affected pore structure of the sample. The large surface area could supply more active sites for Hg0 removal, which could also promote the separation efficiency of electron-hole pairs, resulting in a higher quantum efficiency. Meanwhile, light is also harvested more efficiently because of the large surface area (Yu et al., 2004). Besides crystalline phase and surface area, the efficient charge trapping, migration and transfer were further clarified. The fate of electron-hole pairs was investigated through wide survey PL spectra (Fig. S4, Electronic Supplementary Material) with an excitation wavelength of 325 nm. The lower the PL intensity, the lower the recombination rate of photo-induced electron-hole pairs, which usually relates to good photocatalytic activity (Yu et al., 2003). Compared with TiO2, peak positions of cerium-based samples are basically the same. It is easy to find that the PL intensities of samples are much lower than TiO2. And 0.3% treatment was the optimum one with the lowest intensity, which increased with the addition of Ce content. This increase occurred because the excess amount of Ce contributed to recombination centers on the surface of the samples. 3.5. Mechanism of Hg0 photocatalytic removal Based on the above Hg0 removal efficiencies, the quantum yields (QY, the number of photons utilized for the desired chemical reaction divided by the number of photons adsorbed by the catalyst) were calculated. The number for TiO2 was 16.26%. In particular, the addition of 0.3% Ce into TiO2 achieved a higher quantum yield of Hg0 reduction (44.69%). When minor cerium was doped into TiO2, the electron scavenger was formed, which may reduce the recombination of photoinduced electron-hole couples and thus lead to higher quantum yield. That means the combination of the present experimental set-up and cerium-based TiO2 nanofibers is effective in the removal of Hg0. Because of the highest Hg0 removal efficiency, the sample with 0.3% Ce was used for an extended test to investigate its stability. Fig. S5 (Electronic Supplementary Material) depicts Hg0 removal over the catalyst for 6 cycles. In each cycle, it was firstly exposed to visible light irradiation for 25 min, then UV irradiation for 25 min. It is not difficult to find that Hg0 removal efficiency was 25% under visible light and 91% under UV irradiation. Nevertheless, no apparent change in the efficiency was observed among those cycles, indicating that the sample showed promising stability for long-term use in removing Hg0. After photocatalysis, the determination of mercury forms was conducted with XPS. It can be seen from Fig. 6 (a) that an obvious peak found at 101.48 eV was attributed to Si 2p electron, for Si is the ingredient element of quartz sieve plate. Subsequently, the binding energy value at 105.43 eV was assigned to HgO, which is the oxidation product from Hg0. Moreover, no peak for Hg0 was found. This may be owing to the low Hg0 content on the surface of the sample, which is much lower than the detection limit of XPS analysis. An additional test was conducted in our previous study (Wang et al., 2017), which demonstrated that the adsorption capacity of our sample was low. In other words, both adsorption and oxidation exist in the photocatalytic reaction, while oxidation dominated the removal of Hg0. To provide more specific information, analysis of wide survey XPS spectra was conducted by peak integration, and the results were depicted in Fig. 6 (b). It reveals that both samples contain C, Ti and O elements. An additional small peak of Ce 3d was found for the sample with Ce. The XPS of Ti 2p shows a symmetric Ti 2p peak. The spin-orbit component (2p3/2 and 2p1/2) of the peak shows two shoulder peaks at ca. 458.5 eV and ca. 464.1 eV, respectively, indicating that Ti element primarily exists as the chemical state of Ti4þ. This is in satisfactory agreement with the findings in
Fig. 6. XPS survey spectra of: (a) Hg 4f for spent sample with 0.3% Ce; (b) Ti 2p, O 1s and Ce 3d.
literature (Biener et al., 2000; Ma et al., 2009). The O 1s peak is resolved into three component peaks which achieved an acceptable fit in the XPS spectra. The binding energy values at ca. 529.85 eV, ca. 532.02 eV, and ca. 533.44 eV can be assigned to lattice oxygen in TiO2, surface hydroxyl groups, and adsorbed oxygen, respectively. The Ce 3d photoelectron spectrum is rather complex and can be assigned to 3d3/2 spin-orbit states (labeled u) and 3d5/2 states (labeled v). However, a straight interpretation of the Ce 3d spectrum is attainable by following the intensity of the marked peaks in Fig. 6 (b), which shows the Ce 3d spectra for fresh and spent photocatalysts. The u000 /v000 doublet is caused by the primary photoemission from Ce4þ-O2. The u/v and u00 /v00 doublets are shakedown features resulting from the transfer of one or two electrons from a filled O 2p orbital to an empty Ce 4f orbital. The u0 /v' doublet is caused by the photoemission from Ce3þ (Ho et al., 2005; Liu et al., 2005). The Ce 3d spectra of fresh and spent samples denotes a mixture of Ce3þ/4þ oxidation states giving rise to a myriad of peaks, implying the coexistence of Ce3þ and Ce4þ on the surface of the ordered samples. Under UV light irradiation, the mechanism of Hg0 removal over TiO2 is considered to arise from the excitation of photoelectrons from the valance band (VB) to conduction band (CB), leaving holes in the VB. In the presence of O2 and H2O, the photo-generated
L. Wang et al. / Chemosphere 185 (2017) 690e698
697
Fig. 7. Schematic diagram of a proposed mechanism for Hg0 photocatalytic removal over cerium-based TiO2 nanofibers.
electrons and holes react with them to produce active oxygen 0 species such as O 2 and OH radicals, which are responsible for Hg oxidation (Yuan et al., 2012). But the electron-hole pair recombination occurs within nanoseconds, which weakens its activity sharply. After Ce is doped into TiO2, Ce4þ is the electron scavenger and efficiently traps the photo-excited electron. The electrons trapped in the Ce4þ/Ce3þ site are subsequently transferred to adsorbed O2 through the oxidation process, which also reduced the electron-hole pair recombination rate. Going deeper into the reaction evolution, it might be connected with a series of reactions, such as those listed below:
Ce4þ þ e /Ce3þ
(5)
Ce3þ þ O2 /O2 þ Ce4þ
(6)
Hg0 þ O2 /HgO
(7)
Hg 0 þ OH/HgO
(8)
Based on the analysis above, a schematic diagram of the mechanism is proposed as depicted in Fig. 7. Additionally, the thickness of the space charge layer (W) decreases with the increasing dopant concentration. When the value of W approaches the penetration depth of the light into the solid, all the adsorbed photons generate electron-hole pairs that are efficiently separated (Kumar and Devi, 2011), and the trap center is converted into the combination center. 4. Conclusions Cerium-based TiO2 nanofibers have been successfully prepared by the electrospinning method, and they were applied to the photocatalytic removal of Hg0 at low temperature. Results reported herein demonstrate that the homogeneous nanofibers possess a mixture-type crystalline phase. Ce ions were confirmed to be highly dispersed primarily in the form of metal oxides with a coexistence of Ce3þ/Ce4þ states. Sample with 0.3% Ce at 400 C showed the highest Hg0 removal capacity, reaching 91% under UV irradiation. There was no significant change in Hg0 removal during the long-term experiment, indicating that the samples were sufficiently stable for removing Hg0 from flue gas. Hg0 removal was more obviously related to the participation of sample. The
enhancement of photocatalytic activity was primarily attributed to the large surface area and the better photoelectron-hole separation. Consequently, it is a promising method to reduce the emissions of Hg0 during photocatalysis process by using cerium-based TiO2 nanofibers. dAcknowledgments This project was supported by the National Key Basic Research and Development Program (No. 2014CB238904) and the National Natural Science Foundation of China (51376074). The authors would like to express their thanks to the China Scholarship Council (CSC) and the Department of Analytical and Testing Center at Huazhong University of Science and Technology for providing financial support and the sample analysis, respectively. Appendix A. Supplementary data Supplementary data related to this article can be found at http:// dx.doi.org/10.1016/j.chemosphere.2017.07.049. References €umer, M., Wang, J., Madix, R.J., 2000. Electronic structure and growth of Biener, J., Ba vanadium on TiO2(110). Surf. Sci. 450, 12e26. Celebioglu, A., Sen, H.S., Durgun, E., Uyar, T., 2016. Molecular entrapment of volatile organic compounds (VOCs) by electrospun cyclodextrin nanofibers. Chemosphere 144, 736e744. Chen, S.S., Hsi, H.C., Nian, S.H., Chiu, C.H., 2014. Synthesis of N-doped TiO2 photocatalyst for low-concentration elemental mercury removal under various gas conditions. Appl. Catal. B-Environ. 160e161, 558e565. Cho, J.H., Eom, Y., Jeon, S.H., Lee, T.G., 2013. A pilot-scale TiO2 photocatalytic system for removing gas-phase elemental mercury at Hg-emitting facilities. J. Ind. Eng. Chem. 19, 144e149. Cho, J.H., Lee, T.G., Eom, Y., 2012. Gas-phase elemental mercury removal in a simulated combustion flue gas using TiO2 with fluorescent light. J. Air Waste Manag. 62, 1208e1213. Devi, L.G., Kumar, S.G., 2012. Exploring the critical dependence of adsorption of various dyes on the degradation rate using Ln3þ-TiO2 surface under UV/solar light. Appl. Surf. Sci. 261, 137e146. El-Bahy, Z.M., Ismail, A.A., Mohamed, R.M., 2009. Enhancement of titania by doping rare earth for photodegradation of organic dye (Direct Blue). J. Hazard. Mater. 166, 138e143. Eswaran, S., Stenger, H.G., 2005. Understanding mercury conversion in selective catalytic reduction (SCR) catalysts. Energy Fuel 19, 2328e2334. Fan, X., Li, C., Zeng, G., Zhang, X., Tao, S., Lu, P., Tan, Y., Luo, D., 2012. Hg0 removal from simulated flue gas over CeO2/HZSM-5. Energy Fuel 26, 2082e2089. Feng, D., Xu, S., Liu, G., 2015. Application of immobilized TiO2 photocatalysis to improve the inactivation of Heterosigma akashiwo in ballast water by intense pulsed light. Chemosphere 125, 102e107.
698
L. Wang et al. / Chemosphere 185 (2017) 690e698
Gao, X., Jiang, Y., Zhong, Y., Luo, Z., Cen, K., 2010. The activity and characterization of CeO2-TiO2 catalysts prepared by the sol-gel method for selective catalytic reduction of NO with NH3. J. Hazard. Mater. 174, 734e739. He, C., Shen, B., Chi, G., Li, F., 2016. Elemental mercury removal by CeO2/TiO2-PILCs under simulated coal-fired flue gas. Chem. Eng. J. 300, 1e8. Ho, C., Yu, J.C., Kwong, T., Mak, A.C., Lai, S., 2005. Morphology-controllable synthesis of mesoporous CeO2 nano- and microstructures. Chem. Mater. 17, 4514e4522. Huang, X., Feng, Y., Hu, C., Xiao, X., Yu, D., Zou, X., 2015. Mechanistic QSAR models for interpreting degradation rates of sulfonamides in UV-photocatalysis systems. Chemosphere 138, 183e189. Jeon, S.H., Eom, Y., Lee, T.G., 2008. Photocatalytic oxidation of gas-phase elemental mercury by nanotitanosilicate fibers. Chemosphere 71, 969e974. Jia, L., Dureau, R., Ko, V., Anthony, E.J., 2010. Oxidation of mercury under ultraviolet (UV) irradiation. Energy Fuel 24, 4351e4356. Jin, M., Kim, J.W., Kim, J.M., Jurng, J., Bae, G.N., Jeon, J.K., Park, Y.K., 2011. Effect of calcination temperature on the oxidation of benzene with ozone at low temperature over mesoporous a-Mn2O3. Powder Technol. 214, 458e462. Jones, A.P., Hoffmann, J.W., Smith, D.N., Feeley, T.J., Murphy, J.T., 2007. DOE/NETL's phase II mercury control technology field testing program: preliminary economic analysis of activated carbon injection. Environ. Sci. Technol. 41, 1365e1371. Jung, K.Y., Park, S.B., Anpo, M., 2005. Photoluminescence and photoactivity of titania particles prepared by the solegel technique: effect of calcination temperature. J. Photochem. Photobiol. A 170, 247e252. Kumar, S.G., Devi, L.G., 2011. Review on modified TiO2 photocatalysis under UV/ visible light: selected results and related mechanisms on interfacial charge carrier transfer dynamics. J. Phys. Chem. A 115, 13211e13241. Lasek, J., Yu, Y.H., Wu, J.C., 2013. Removal of NOx by photocatalytic processes. J. Photochem. Photobiol. C 14, 29e52. Li, B., Hao, Y., Zhang, B., Shao, X., Hu, L., 2017. A multifunctional noble-metal-free catalyst of CuO/TiO2 hybrid nanofibers. Appl. Catal. A-Gen. 531, 1e12. Li, H., Wu, S., Wu, C.Y., Wang, J., Li, L., Shih, K., 2015. SCR atmosphere induced reduction of oxidized mercury over CuO-CeO2/TiO2 catalyst. Environ. Sci. Technol. 49, 7373e7379. Lim, T.H., Jeong, S.M., Kim, S.D., Gyenis, J., 2000. Photocatalytic decomposition of NO by TiO2 particles. J. Photochem. Photobiol. A 134, 209e217. Liu, H., Yu, X., Yang, H., 2014. The integrated photocatalytic removal of SO2 and NO using Cu doped titanium dioxide supported by multi-walled carbon nanotubes. Chem. Eng. J. 243, 465e472. Liu, Y., Zhang, J., Sheng, C., Zhang, Y., Zhao, L., 2010. Wet removal of sulfur dioxide and nitric oxide from simulated coal-fired flue gas by UV/H2O2 advanced oxidation process. Energy Fuel 24, 4931e4936. Liu, Z., Guo, B., Hong, L., Jiang, H., 2005. Preparation and characterization of cerium oxide doped TiO2 nanoparticles. J. Phys. Chem. Solids 6, 161e167. Ma, T.Y., Cao, J.L., Shao, G.S., Zhang, X.J., Yuan, Z.Y., 2009. Hierarchically structured squama-like cerium-doped titania: synthesis, photoactivity, and catalytic CO oxidation. J. Phys. Chem. C 113, 16658e16667. Moussavi, G., Mohseni, M., 2007. Using UV pretreatment to enhance biofiltration of mixtures of aromatic VOCs. J. Hazard. Mater. 144, 59e66. Palau, J., Colomer, M., Penya-Roja, J.M., Martínez-Soria, V., 2012. Photodegradation of toluene, m-xylene, and n-butyl acetate and their mixtures over TiO2 catalyst on glass fibers. Ind. Eng. Chem. Res. 51, 5986e5994. Park, J.A., Kim, S.B., 2016. Antimicrobial filtration with electrospun poly (vinyl alcohol) nanofibers containing benzyl triethylammonium chloride: immersion, leaching, toxicity, and filtration tests. Chemosphere 167, 469e477. Salehifar, N., Nikfarjam, A., 2017. Improvement the visible light photocatalytic activity of gold nanoparticle, Co2O3 and nitrogen doped TiO2 nanofibers. Mater. Lett. 188, 59e62.
Shen, H., Ie, I.R., Yuan, C.S., Hung, C.H., 2016. The enhancement of photo-oxidation efficiency of elemental mercury by immobilized WO3/TiO2 at high temperatures. Appl. Catal. B-Environ. 195, 90e103. Shen, H., Ie, I.R., Yuan, C.S., Hung, C.H., Chen, W.H., Luo, J., Jen, Y.H., 2015. Enhanced photocatalytic oxidation of gaseous elemental mercury by TiO2 in a high temperature environment. J. Hazard. Mater. 289, 235e243. Silva, A.M., Silva, C.G., Dra zi c, G., Faria, J.L., 2009. Ce-doped TiO2 for photocatalytic degradation of chlorophenol. Catal. Today 144, 13e18. Song, S., Xu, L., He, Z., Chen, J., Xiao, X., Yan, B., 2007. Mechanism of the photocatalytic degradation of C.I. reactive black 5 at pH 12.0 using SrTiO3/CeO2 as the catalyst. Environ. Sci. Technol. 41, 5846e5853. €rl, R., Belo, L., Shah, K., Stanger, R., Giniyatullin, R., Maier, J., Wall, T., Spo Scheffknecht, G., 2014. Mercury emissions and removal by ash in coal-fired oxyfuel combustion. Energy Fuel 28, 123e135. Suriyawong, A., 2009. Mercury capture by nano-structured titanium dioxide sorbent during coal combustion: lab-scale to pilot-scale studies. Aerosol Air Qual. Res. 9, 394e403. Wang, H., You, C., 2016. Photocatalytic removal of low concentration SO2 by titanium dioxide. Chem. Eng. J. 292, 199e206. Wang, H., Zhou, S., Xiao, L., Wang, Y., Liu, Y., Wu, Z., 2011. Titania nanotubesda unique photocatalyst and adsorbent for elemental mercury removal. Catal. Today 175, 202e208. Wang, L., Zhao, Y., Zhang, J., 2017. Comprehensive evaluation of mercury photocatalytic oxidation by cerium-based TiO2 nanofibers. Ind. Eng. Chem. Res. 56, 3804e3812. Wang, P., Su, S., Xiang, J., You, H., Cao, F., Sun, L., Hu, S., Zhang, Y., 2014. Catalytic oxidation of Hg0 by MnOx-CeO2/gamma-Al2O3 catalyst at low temperatures. Chemosphere 101, 49e54. Wang, Y., Duan, Y., 2011. Effect of manganese ions on the structure of Ca(OH)2 and mercury adsorption performance of Mnxþ/Ca(OH)2 composites. Energy Fuel 25, 1553e1558. Weng, X., Mei, R., Shi, M., Kong, Q., Liu, Y., Wu, Z., 2015. CePO4 catalyst for elemental mercury removal in simulated coal-fired flue gas. Energy Fuel 29, 3359e3365. Wu, H., Li, C., Zhao, L., Zhang, J., Zeng, G., Xie, Y., Zhang, X., Wang, Y., 2015. Removal of gaseous elemental mercury by cylindrical activated coke loaded with CoOxCeO2 from simulated coal combustion flue gas. Energy Fuel 29, 6747e6757. Xie, Y., Yuan, C., 2003. Visible-light responsive cerium ion modified titania sol and nanocrystallites for X-3B dye photodegradation. Appl. Catal. B-Environ. 46, 251e259. Xu, Y., Chen, H., Zeng, Z., Lei, B., 2006. Investigation on mechanism of photocatalytic activity enhancement of nanometer cerium-doped titania. Appl. Surf. Sci. 252, 8565e8570. Xue, W., Zhang, G., Xu, X., Yang, X., Liu, C., Xu, Y., 2011. Preparation of titania nanotubes doped with cerium and their photocatalytic activity for glyphosate. Chem. Eng. J. 167, 397e402. Yu, J.C., Wang, X., Fu, X., 2004. Pore-wall chemistry and photocatalytic activity of mesoporous titania molecular sieve films. Chem. Mater. 16, 1523e1530. Yu, J.G., Yu, H.G., Cheng, B., Zhao, X.J., Yu, J.C., Ho, W.K., 2003. The effect of calcination temperature on the surface microstructure and photocatalytic activity of TiO2 thin films prepared by liquid phase deposition. J. Phys. Chem. B 107, 13871e13879. Yuan, Y., Zhao, Y., Li, H., Li, Y., Gao, X., Zheng, C., Zhang, J., 2012. Electrospun metal oxide-TiO2 nanofibers for elemental mercury removal from flue gas. J. Hazard. Mater. 227e228, 427e435. Zhou, C., Sun, L., Zhang, A., Wu, X., Ma, C., Su, S., Hu, S., Xiang, J., 2015. Fe3-xCuxO4 as highly active heterogeneous Fenton-like catalysts toward elemental mercury removal. Chemosphere 125, 16e24.