Efficient Visible Light Photocatalytic Oxidation of NO on Fand N-Codoped Spherical TiO 2 Synthesized via Ultrasonic Spray Pyrolysis

1 Environmental and Life Sciences Department, Putian University, Putian 351100, China 2 Nano and Advanced Materials Institute Limited, The Hong Kong University of Science and Technology, Hong Kong 3 Department of Civil and Structural Engineering, Research Center for Environmental Technology and Management, The Hong Kong Polytechnic University, Hong Kong 4 Department of Science and Environmental Studies, The Hong Kong Institute of Education, Hong Kong


Introduction
Growing knowledge about the indoor air quality (IAQ) has validated it as one of the major risks of the human health as people spend more than 80% of their time indoors in most countries [1].Thus, more and more attentions have been drawn into developing a green process of eliminating pollutants from indoor air.Several physical and chemical methods for treating the pollutant in the indoor air have been studied extensively, including adsorption [2,3], ozonation [4], plasma [5], and photocatalytic oxidation [6][7][8][9].Among these, semiconductor-mediated photocatalysis is one of the most efficient destructive technologies because of its high redox ability, nonselectivity, at-room temperature, and atmospheric pressure.For decades, titanium dioxide (TiO 2 ) has been accepted as one of the most promising photocatalysts owning to its intrinsic properties such as inexpensiveness, high chemical stability, and low toxicity.
Unfortunately, a major obstacle in the popularization of TiO 2 is its large band gap (3.2 eV) which work only under specific UV light region (λ < 380 nm), hence, practically ruling out most of the solar energy for the environmental remediation.Thus, considerable efforts have been devoted to modified TiO 2 in order to broaden the light response range of TiO 2 -base photocatalysts which is still a significant issue in both academic and practical applications.Doping TiO 2 with various elements has been proved as an effective method to optimize the band structure of TiO 2 , and thereby in resulting to be a visible-light sensitive photocatalyst.While the cations doping TiO 2 are often impaired by thermal instability and an increase in the population of recombination centers, thus no significant results were achieved [10,11].Doping TiO 2 with nonmetal such as N, C, S, B, and P seems to be more efficient to address this problem in certain extend [12][13][14][15][16][17][18].Among them, doping with nitrogen atom is considered to be one of the most effective way to obtain visiblelight-driven photocatalytic activity [19,20].However, the reactivity or quantum efficiency is still not high enough with only doping nitrogen.Thus many researches were designed to be codoped with other nonmetal element such as C, S, B, and P [21][22][23][24] via the synergistic effects between two dopants to further enhance the photocatalytic efficiency.Among them, intense interest has been shown on the F-and N-codoped TiO 2 [25][26][27][28][29][30].The doping of fluorine will have effect on the physicochemical properties of TiO 2 in many aspects [31].The study of Yu et al. indicated that the fluorine doping would result in the formation of reduced Ti 3+ ions and surface oxygen vacancies which may promote charge separation and improve the efficiency of photoinduced processes [32].Vijayabalan reported that the fluorination enhanced the absorbance of TiO 2 in the UV-visible region which is favorable to the visible-light photocatalytic activity of TiO 2 [33].The doping of fluorine was also found to have side effect, for example, it would lower the conduction band which leads to a lower photocatalytic activity in some photocatalytic reactions [34].Thus it is meaningful to further study the properties of N-and F-codoped TiO 2 and the synergistic effects between the N and F dopants in the catalyst.
Synthetic methods such as sol-gel and hydrothermal technique have been widely used to make codoped TiO 2 photocatalyst which have advantages including precise control of pore structure, dopant concentration, and chemical purity [35][36][37]; however, they require severe conditions such as long time for reaction or high-pressure resistant equipment which is not easy to produce in large scale.Thus, a high-yield synthetic method that is capable of producing nanostructure TiO 2 with controllable dopant concentration and crystallinity will undoubtedly benefit the design of highperformance materials in the catalyst industry.Ultrasonic spray pyrolysis is a convenient method for the facile mass production of nanostructured spherical materials because of its simple and inexpensive apparatus and easy control of product composition.Recently, carbon, SiO 2 , Bi 2 WO 6 , InVO 4 , and so forth have been synthesized by this method [38][39][40][41].Although ultrasonic spray pyrolysis has been widely used to produce inorganic powders, it is still a challenge to fabricate in the prepared material decorated with heteroatom especially the nonmetal atoms.
In this study, we fabricated fluorine-and nitrogencodoped TiO 2 using ultrasonic spray pyrolysis method with urea and titanium tetrafluoride as precursors.The as-prepared product had unique spherical structures with puckered surface and exhibited high photocatalytic activity on the oxidation of NO in air under the irradiation of visible light.The preparation method is simple and continuous which is easy for batch production.Moreover, the dopant of such TiO 2 sphere is easily controlled by regulating the experimental conditions.

Experimental Section
2.1.Catalysts Preparation.All of the chemicals were of commercially available analytical grade and used without further purification.Fluorine-and nitrogen-codoped TiO 2 photocatalysts were prepared by an ultrasonic spray pyrolysis method.In a typical process, TiF 4 (1.5 g) was added to 150 mL HCl solution (0.5 M) under magnetic stirring.After stirring for several minutes, a transparent colorless solution was obtained.Then, different quantity of urea (0, 1, 5, and 10 g) was added to the solution under magnetic stirring.After the urea was dissolved completely, the resulting solution was nebulized at 1.7 MHz ±10% (YUYUE402AI, Shanghai, China).The produced aerosol was carried with pumping air flow (5 L/min) through a quartz tube surrounded by a furnace thermostated.The pyrolysis proceeded quickly as aerosol passed through the high-temperature tube maintained at the temperature of 500 • C. The quartz reaction tube with the diameter of 3.5 cm was 1 m long.The products were collected in a percolator with distilled water, then separated by centrifugation, washed thoroughly with ethanol and distilled water, and finally dried in an oven at 100 • C for 1 h.A vivid yellow product was obtained.The as-prepared product was then further calcinated at 500 • C and held for 1 h (ramp rate of 1 • C/min).The samples prepared by addition of 1 g, 5 g, and, 10 g urea were designated FNT-1, FNT-5, and FNT-10, respectively.For comparison, the fluorinedoped TiO 2 was also prepared without adding urea and it was designated FT.

Characterization. The X-ray diffraction (XRD) patterns
were recorded on a Bruker D8 Advance X-ray diffractometer with Cu Ka radiation (λ = 1.54178A • ) at a scan rate of 0.05 • 2θ/s.The accelerating voltage and the applied current were 40 kV and 40 mA, respectively.Scanning electron microscopy (SEM, JSM-5600) was used to characterize the morphology of the obtained products.Transmission electron microscopy (TEM) study was carried out on a Philips CM-120 electron microscopy instrument.The samples for TEM were prepared by dispersing the final powders in methanol; the dispersion was then dropped on carbon copper grids.A Varian Cary 100 Scan UV-Visible system equipped with a labsphere diffuse reflectance accessory was used to obtain the reflectance spectra of the catalysts over a range of 200-800 nm.Labsphere USRS-99-010 was employed as a reflectance standard.The spectra were converted from reflection to absorbance by Kubelka-Munk method.Xray photoelectron spectroscopy (XPS) measurements were performed on a PHI Quantum 2000 XPS system with a monochromatic Al Kα source and a charge neutralizer.All the binding energies were calibrated to the C 1s peak at 284.8 eV of the surface adventitious carbon.

NO Removal Test.
The nitric oxide (NO) is the common pollutant of indoor air.It was reported by WHO that it could lead to respiratory symptoms, bronchoconstriction, increased bronchial reactivity, airway inflammation, and decreases in immune defence leading to increased susceptibility to respiratory infection.Thus, we used NO photocatalytic oxidation to evaluate the performance of photocatalysts in this study.The NO removal with the resulting samples was performed at ambient temperature in a continuous flow reactor.The volume of the rectangular reactor which was made of stainless steel and covered with Saint-Glass was 4.5 L (10 cm × 30 cm × 15 cm (H × L × W)).The sample to be tested was prepared by coating an aqueous suspension of the photocatalyst onto a dish with a diameter of 5.0 cm.The weight of the photocatalysts used for each experiment was kept at 0.3 g.The dishes containing the photocatalyst were pretreated at 70 • C for several hours until a complete removal of water in the suspension and then cooled to room temperature before use.The lamp was vertically placed outside the reactor above the sample dish, and a glass filter was placed to cutoff light below 400 nm.Four minifans were fixed around the lamp to avoid the temperature rise of the flow system.Furthermore, adequate distance was also kept from the lamp to the reactor for the same purpose to avoid temperature rise.The NO gas was acquired from compressed gas cylinder at a concentration of 48 ppm NO (N 2 balance, BOC gas) with traceable National Institute of Stands and Technology (NIST) standard.The initial concentration of NO was diluted to about 400 ppb by the air stream supplied by a zero air generator (Thermo Environmental Inc., model 111).The desired humidity level of the NO flow was controlled at 70% (2100 ppmv) by passing the zero air streams through a humidification chamber.The gas streams were premixed completely by a gas blender and the flow rate was controlled at 4 L min −1 by a mass flow controller.After the adsorption-desorption equilibrium among water vapor, gases and photocatalysts were achieved, and the lamp was turned on.The concentration of NO was continuously measured by a chemiluminescence NO analyzer (Thermo Environmental Instruments Inc., model 42c), which monitors NO, NO 2 , and NOx (NOx represents NO + NO 2 ) with a sampling rate of 0.7 L min −1 .The reaction of NO with air was ignorable when performing a control experiment with or without light in the absence of photocatalyst.

Result and Discussion
Figure 1 shows the XRD patterns of the prepared samples.For all the samples exhibit five distinct peaks (2θ = 25.4,38.1, 48.1, 53.9, and 55.6 • ) which represent the indices of (101), (004), ( 200), (105), and (211) planes of anatase TiO 2 .The crystalline sizes of the samples prepared by adding 0, 1, 5, and 10 g urea estimated by Scherrer formula from the (101) peaks of anatase phase were 14, 13, 10, and 9 nm, respectively.This indicates that TiO 2 particles in the prepared samples are nanocrystalline in nature.The result also evidences the addition of urea could suppress undesirable sintering and agglomeration of TiO 2 particle during the thermal treatment.In this process, TiO 2 nanoparticle possibly formed via the following two consecutive reaction steps: In the absence of urea, the rate of hydrolysis and condensation reactions ((1) and ( 2)) is rather slow due to the high concentration of H + .Consequently, fewer TiO 2 nanoclusters are formed which is beneficial to crystal growth, rather than aggregation and a product with large crystal size could be obtained.For the addition of urea in the reaction, the added urea is easily decomposed and the produced NH 3 •H 2 O will react with HCl as shown in ( 3) and ( 4).Thus the acidity of reaction system is decreased which would favor the formation of more TiO 2 nanoclusters and results in smaller crystalline sizes of product [42,43].Consider the following: Figure 2 depicts the SEM and TEM images of FNT-5. Figure 3(a) is the low-magnification image of the asprepared FNT-5.It could be found that the as-prepared codoped sample consists entirely of small spherical particles with an average size of 200 nm∼2 μm. Figure 3(b) is the magnified surface morphologies of typical codoped titania spheres showing the puckered surface of codoped titania, which will endow them with more active points and larger light-harvesting ability.The size and morphology of FNT-5 were further analyzed by TEM and HRTEM images.The TEM image confirms the puckered surface of the codoped sphere.Figure 3(d) is HRTEM images of the codoped TiO 2 sphere edge.Well-resolved lattice fringes are clearly visible, with an interplanar distance of 0.35 nm corresponding to the (101) d spacing of the anatase phase, further confirming the good crystallinity of as-prepared products.
Figure 3 shows the SEM images of the products fabricated under different conditions.Interestingly, it could be found that the presence of urea is determining the surface morphology of the product.In the case of the sample prepared without urea, it consists of spherical particles with a relative smooth surface morphology, while the solid spheres with puckered surface were obtained with the addition of urea and the quantity of added urea had not much effect on the morphology of final products.On the basis of the above discussion, we propose the possible mechanisms for the special morphology of codoped product by the ultrasonic spray pyrolysis method.It is widely accepted that the droplets produced form nebulizer serve as microreactors and yield one particle per droplet when sprayed into a tubular reactor under pyrolysis condition.The size of droplet determines the product dimension [44].The solvent was evaporated out of the droplets in the hot zone of the furnace, leading to the formation of microspheres.In this process, the hydrolysis velocity of Ti source plays a vital effect on the structure.For the samples without urea, the hydrolysis of TiF 4 was restrained due to the low pH value of reaction system, so the hydrolysis of TiF 4 happened after the volatilization of HCl in the aerosol (see Scheme 1(a)).Thus a product with hollow structure was obtained (see Figure S1(a) in Supplementary Material available online at doi:10.1155/2012/753429).While for the codoped samples, the formation process may follow the route shown in Scheme 1(b).The urea will react with HCl in the precursor solution, thus the TiF 4 was unstable and easily hydrolysis before the volatilization of liquid in the aerosol, hence getting the solid sphere (see Figure S1(b)).In addition, the surfaces of the products prepared under different condition have distinct difference.For the fluorine and nitrogen codoped samples, the urea in the precursor was decomposed easily and formed NH 4 Cl according to reaction of ( 3) and ( 4).Hence, there were some NH 4 Cl left in the sphere.The NH 4 Cl will be decomposed (5) at high temperature and cause the shrink of forming sphere with puckered surface.The NH 3 from the decomposed NH 4 Cl and the HF from the hydrolysis of TiF 4 would react with TiO 2 under high temperature (6), and consequently the codoped catalyst is obtained as follows: The UV/Vis spectra of the prepared samples were recorded through the diffuse reflectance technique with the pure TiO2 (P25 Degussa) for comparison.The result is shown in Figure 4. Comparing with P25, the doped TiO2 spheres reveal a new absorption shoulder at 400-600 nm in the UV/Vis spectrum.The single doping of fluorine only has limited influence to light absorption properties of titania which only have a small shoulder at 400-450 nm.The Fdoped-TiO 2 hollow spheres exhibit a slight tinge of light yellow color which could be due to the surface oxygen vacancies generated by the doping of fluorine [45,46].In comparison, the codoped samples have vivid yellow color and show a remarkable shift to visible-light region.Two optical absorption thresholds at about 535 and 440 nm in the visible range were observed for all the codoped TiO 2 .These two absorption thresholds also demonstrate that two possible compositions have been introduced to TiO 2 .The absorption threshold at about 440 nm is ascribed to the introduction of nitrogen to the TiO 2 lattice.The absorption edge of this band shifts towards higher wavelength on the whole with the increasing of urea quantity in precursor for all the codoped titania photocatalysts.The difference of the shift is due to the different amount of nitrogen doped into the TiO 2 framework as shown in the following XPS results.The absorption thresholds at about 535 nm are the superposition result of the nitrogen and fluorine doped into TiO 2 lattice.This enlarged light absorption range for the codoped samples in visible region will have great influence on their photocatalytic property.The assessment of the surface chemical composition and electronic state of the product was investigated by using XPS analysis.The surface concentrations of fluorine and nitrogen in these photocatalysts are summarized in Table 1.It could be found that the fluorine and nitrogen concentration increased with increasing urea in the precursor.It was obvious that the increased N dopant is due to more nitrogen resource from urea.While for the fluorine, the HF in the codoping system may react with the NH 3 which was produced by the urea decomposition and get NH 4 F. The boiling point of NH 4 F (64.6 • C) is much higher than that of HF (19.4 • C) which could reduce the fluorine specie losing by the evaporation in the reaction.Thus more fluorine will be doped into TiO 2 with the increased amount of urea.The high-resolution XPS spectra of N 1s and F 1s are shown in Figures 5 and 6.The N 1s regions of all samples are asymmetric, indicating that there are at least two kinds of chemical states.Using XPSPEAK software, N 1s region could be fitted to two peaks.The binding energy region near 399.5 eV is derived from the presence of O-Ti-N linkages in the crystalline TiO 2 lattice [47].It has been proved that this nitrogen state is responsible for the observed shift of the photochemical threshold of TiO 2 [48,49], while the N 1s peaks at 401.6 eV are arisen from the N atoms from adventitious N-N, N-O, or NHX adsorbed on the surface of the sample [50][51][52].Figure 6 gives the F 1s XPS spectra and a wide and symmetrical peaks at 685 eV were observed which could be assigned to the substitutional fluorine in titania lattice [53].
The oxidation of NO under the irradiation of visible light was used to evaluate the photocatalytic activity of the samples prepared under different condition.It is found that NO could be oxidized over the catalysts prepared under different conditions when the visible light was switched on.The control experiment found that the photolysis of NO without photocatalysts or use pure TiO 2 (P25) was negligible under the same condition.Consequently, it could be concluded that all the samples have visible-light-driven photocatalytic  activity.Table 2 shows the photocatalytic activity of codoped TiO 2 for NO oxidation irradiated by visible light with the fluorine single-doped sample for comparison.As expected, the codping greatly affects the photocatalytic activity of spherical TiO 2 toward oxidizing NO in gasphase.For the single fluorine doping sample, the NO photocatalytic oxidation rate is 190.7 μmol/mol•h.The oxygen vacancy created by the doped fluorine slightly extended the optical absorption of TiO 2 in the visible-light region to contribute to the visible-light photocatalytic activity of fluorine doping sample [54].However, the activity of single doping of fluorine is limited as the doping of fluorine could not alter the basic band structure of TiO 2 .For the codoped samples, they show high oxidation rate under optimal conditions.For the sample of FNT-10, the NO photocatalytic oxidation rate is 287.4 μmol/mol•h which is about 50% higher than that of FT.The enhanced activity of FNT-10 can be ascribed to the following two reasons.Firstly, the improved visiblelight absorption and red shift of codoped TiO 2 which would lead to increased number of photons that take part in the photocatalytic reaction.Second, the small crystal size of codoped TiO 2 is a benefit to its photocatalytic performance.With a smaller particle size, the number of active surface sites increases, and so does the surface charge carrier transfer rate in photocatalysis.Therefore the higher photocatalytic performance is obtained [55].
The results also show that the activities of codoped samples were greatly influenced by the quantity of added urea in the precursor solution.The photocatalytic activity  of samples decreased with the decreasing amount of added urea.This is due to more fluorine and nitrogen atoms that were doped into TiO 2 lattice and the smaller crystalline sizes for the samples prepared with more urea in the precursor.While if we further increased the amount of urea (>10 g) in the preparation, the precursor solution was not stable and some sedimentation was formed, thus we could not get the codoped TiO 2 with ultrasonic spray pyrolysis method.

Conclusion
In summary, the fluorine-and nitrogen-codoped TiO 2 spheres with puckered surface have been fabricated via an ultrasonic spray pyrolysis method.Increasing the addition of urea from 1 g to 10 g makes more nitrogen and fluorine doped into TiO 2 lattice and resulted in a new strong absorption band in the visible range.Activity test shows that the fluorine and nitrogen codoped TiO 2 could efficiently photocatalytically oxidize NO under visible-light irradiation and the activity was higher than that of fluorine-doped TiO 2 .This facile method may be easily scaled up for industrial production.Similar doping or codoping with other nonmetal and metal ions (S, Cl, Br, B, Pt, and Fe) with this method should lead to the development of new kind of spherical semiconductor composites, which are promising candidates as multifunctional materials as catalysts, photocatalysts, and optoelectronic materials.

Figure 1 :
Figure 1: The X-ray diffraction patterns of fluorine-doped and fluorine-and nitrogen-codoped TiO 2 prepared at different conditions.

Figure 4 :
Figure 4: UV/Vis diffuse reflectance spectra of fluorine-and nitrogen-codoped TiO 2 with different urea in the precursor solution with pure TiO 2 (P25) for comparison.

Figure 5 :
Figure 5: The N 1s spectra of fluorine and nitrogen codoped TiO 2 with different urea in the precursor solution.

Table 1 :
The quantitative XPS analysis result using the appropriate molar weight-correction factors.

Table 2 :
The photocatalytic oxidation rate of NO under the irradiation of visible light over the catalyst prepared under different conditions.
a Oxidation rate: n (nitric oxide) per mol of catalyst per hour (μmol/mol•h).