Promotion of SO2 resistance of Ce–La/TiO2 denitrification catalysts by V doping

Conventional cerium-based denitrification catalysts show good catalytic activity at moderate and high temperatures, but their denitrification performance may be decreased due to poisoning by SO2 in the flue gas. In this paper, V was introduced into Ce–La/TiO2 catalysts by a ball-milling method, and the effects of the V content on catalyst denitrification performance and SO2 resistance were investigated. Fourier-transform diffuse reflectance in situ infrared spectroscopy was used to examine the denitrification mechanism and evaluate the catalysts for surface acidity, redox characteristics, and SO2 adsorption. After introducing V, Brønsted acids played the dominant role in the catalytic reaction by increasing the number of acidic sites on the catalyst surface, adsorbing NH3 to participate in the reaction, and improving the sulfur resistance by inhibiting SO2 poisoning. The Ce3+ and O ratio on the catalyst surface were also enhanced by V doping, which reduced interactions between SO2 and the primary metal oxide active ingredients. The modified catalyst inhibited the formation of sulfate species on the catalyst surface and prevented the generation of additional nitrate species on the surface, which protected the main active sites. After V doping, the NH3-SCR reaction on the catalyst surface followed the Langmuir–Hinshelwood mechanism.


Introduction
NO x is a major atmospheric pollutant whose emissions cause acid rain, photochemical smog, and ozone layer depletion. 1elective catalytic reduction (SCR) has a wide range of denitri-cation efficiencies, does not produce secondary contamination due to NH 3 escape, and is widely used in various plants and municipal waste facilities. 2 Cerium-based catalysts are one of the commonly used catalysts in SCR, which is due to the fact that Ce atoms contain a special arrangement of electrons outside the nucleus and have excellent oxidizing properties and oxygen storage capacity.CeO 2 can be converted from Ce 4+ to Ce 3+ when cerium ions generate oxygen vacancies, 3 giving the catalyst excellent redox properties and thermal stability.Cerium promotes the conversion of nitrate species and ammonia species, which facilitates the conversion of NH 3 and NO to reactive intermediates on the catalyst surface, thus promoting the NH 3 -SCR reaction.
Usually, the ue gas in coal-red boilers contains SO 2 , which inevitably poisons the catalyst and reduces its denitrication performance.The reasons for SO 2 poisoning can be classied into three categories: (1) the competitive adsorption of SO 2 with reactants inhibits the formation of reactive intermediates; (2)  the reaction of SO 2 with NH 3 to generate (NH 4 ) 2 SO 4 and NH 4 HSO 4 , which covers active sites on the catalyst surface and reduces the specic surface area of the catalyst; (3) metal active sites are sulfated by SO 2 to create metal sulfates, which results in the loss of the active components and results in irreversible deactivation.For Ce-based catalysts, SO 2 can be adsorbed and oxidized to SO 3 by CeO 2 to generate Ce(SO 4 ) 2 or Ce 2 (SO 4 ) 3 or react with NH 3 to form stable (NH 4 ) 2 SO 4 or NH 4 HSO 4 , the NO conversion of CeO 2 /TiO 2 catalysts was close to 100% above 300 °C and remained stable for 12 h.However, the value decreased signicantly aer 12 h, which was below 40% aer 48 h. 4 For Mn-based catalysts, the NO conversion of fresh samples of MnO 2 -based catalysts was 100% at 130-150 °C.Aer SO 2 poisoning, the NO conversion decreased sharply from 100% to 53% at 130 °C and did not increase when the reaction temperature was increased to 150 °C. 5 Ma et al. 6 prepared Cu 0.02 Fe 0.2 W 0.02 TiO x by doping W to Cu 0.02 Fe 0.2 TiO x catalysts, the NO conversion of which maintained at 100% aer passing water and sulfur at 240 °C for 10 h.Jiang et al. 7 prepared Mn-Fe(0.4)-Ce(0.4)/ACNcatalysts, which showed a NO x conversions higher than 95% at 100-250 °C, and remained around 91% at 175 °C for 32 h under 100 ppm SO 2 and 10 vol% H 2 O.
Early studies on SO 2 poisoning catalysts mostly concentrated on vanadium-based catalysts.In the temperature range of 180 to 400 °C, Li et al.'s 8 investigation of the distribution of SO 2 oxidation products during the NH 3 -SCR process over V 2 O 5 /TiO 2 catalysts discovered that the amount of gaseous SO 3 products increased with temperature in this range.The presence of O 2 in ue gas promoted the oxidation of SO 2 and the formation of gaseous SO 3 on the V 2 O 5 /TiO 2 catalysts.On the catalyst surface, the major pathway for SO 2 oxidation comprises oxidizing SO 2 to generate adsorbed SO 3 , which either desorbs to form gaseous SO 3 or interacts with NH 3 to form sulfate. Sulte is created when a tiny amount of SO 2 interacts with NH 3 ; this product was then quickly oxidized to (NH 4 ) 2 SO 4 .At 180 °C, the amount of deposition products peaked, while below 230 °C, more NH 4 + emerged.Additives that also strengthened the water and sulfur resistance of the rare-earth-based catalysts helped them conduct denitrication even better.Many researchers have prepared binary-loaded metal oxide catalysts using transition metals and doping with rare earth elements such as CoO, 9 Cr 2 O 3 , 10 and Sm 2 O 3 . 11This method boosted the denitrication activity of the original Ce-based catalysts as well as their resistance to water and sulfur.
In this paper, we prepared a Ce-La-V/TiO 2 catalyst via a previously reported ball milling method. 12The effect of the V introduction on the denitrication performance of the catalyst as well as on the water and sulfur resistance was tested.It was found that the NO conversion still maintained 85% aer 52 h of water and sulfur feeding.Aer stopping the water and sulfur, the NO conversion recovered to 99%.In addition, the mechanism of the resistance to SO 2 toxicity at high temperatures was elucidated using various physicochemical characterization techniques and reaction kinetics analysis.

Catalysts preparation
According to the chemical formula of Ce 10 La 2 V x Ti 88−x (where x is the mass ratio of V 2 O 5 per 100 parts of catalyst), cerium carbonate, lanthanum carbonate, ammonium metavanadate and titanium dioxide were added and put in the nylon ball milling jar, appropriate amount of deionized water and zirconia balls were added.The volume of the ball milling jar was 530 mL, the inner diameter was 94.5 mm, the height was 93.5 mm, the weight of zirconia balls was 95 g, and the volume of water was 100 mL.Then the slurry was ball milled in KQM-Z/B planetary ball mill, the rotational speed and time were set (usually 500 rpm, 1 h), and the direction of rotation was changed once per 20 min, separation of the slurry from the zirconia balls at the end of the process, and the slurry was dried in an oven at 105 °C for 12 h.Aer drying, the slurry was calcined at 500 °C for 4 h to obtain the desired catalysts, and the corresponding catalysts were notated as 10Ce-2La-xV/Ti (x = 0.1, 0.3, 0.6, 0.9, 1.4 and 2.3).In addition, the catalyst was poisoned by passing 300 ppm SO 2 and 5% H 2 O on top of a mixture of 500 ppm NO, 500 ppm NH 3 , 5% H 2 O and N 2 , which was allowed to stabilize and then reacted continuously for 4 h to obtain a poisoned sample, which was labeled as 10Ce-2La-xV/Ti-S.

Catalysts activity test
First, the prepared catalysts were ball-milled and attached to porous cordierite cylinders (commercial grade, 4 20 × L 50).Aer drying in an oven, the catalyst-loaded cordierite cylinders were placed in a quartz tube furnace to test the catalytic activity of the catalyst.The catalyst was loaded with a weight of 1 g.The gas mixture consisted of 500 ppm NO, 500 ppm NH 3 , 300 ppm SO 2 , 5% H 2 O, and 3% O 2 , with N 2 as the equilibrium gas.The total gas ow rate was 2500 mL min −1 and the corresponding GHSV was 150 000 mL g −1 h −1 .The ow of all the gases was controlled by a mass owmeter.NO and NO 2 concentrations at the outlet were monitored in real time using an ECOM ue gas analyzer (Germany).NO conversion, N 2 selectivity and SO 2 conversion were calculated from eqn (1)-(3): where the subscript 'in' and 'out' represent the inlet and outlet concentrations of each gas at steady state, respectively.In order to visualize the differences in the catalytic activity of the samples more intuitively, the tests were conducted at a higher GHSV (150 000 mL g −1 h −1 ), keeping the NO conversion rate lower than 30%, and the surface activation energies of the different catalysts were calculated.Assuming that the NH 3 -SCR catalytic reaction is not controlled by diffusion, the SCR reaction rate normalized by the specic surface area of the catalyst can be calculated according to eqn (4) 13,14 Rate where X NO is the NO conversion (%) at different temperatures, Q is the volumetric ow (2500 mL min −1 ), V m is the molar volume of gas under standard conditions (22.4 L mol −1 ), C f is the fed NO concentration (500 ppm), W is the weight of the catalyst (g), and S BET is the specic surface area of the catalyst (m 2 g −1 ).Turn over frequency (TOF) can be calculated to compare the catalytic rate of different catalysts.Keeping the NO conversion below 30% (to ensure that all catalytically active sites are Paper RSC Advances operational), the TOF value of NO on the active center Ce is calculated from eqn (5): 15,16 TOF ¼ Pv=RTa m cat b Ce =M Ce (5)   where P is the standard atmospheric pressure (1.01 × 10 5 Pa), v is the ow of NO, R is the molar gas constant (8.314J mol −1 K −1 ), T is the test temperature (K), a is the NO conversion (%), m cat is the weight of the catalyst (g), b Ce is the loading ratio of Ce (%) calculated using XPS data, and M Ce is the molecular weight of cerium (140.1 g mol −1 ).

Characterization of catalysts
The physical composition of the samples was examined using a Thermo ARL SCINTAG X'TRA X-ray diffraction analyzer.Cu Krays (wavelength 0.154056 nm) were used as its X-ray source, with a tube voltage of 40 kV and a tube current of 40 mA.Sample surface area, pore volume and pore size were measured by N 2 adsorption and desorption using a Biaode SSA-7300 automatic pore size and surface area analyzer.X-ray photoelectron spectroscopy (XPS) was carried out using a Thermo Scientic K-Alpha electron spectrometer equipped with an Al Ka (hv = 1486.6eV) radiation at an operating voltage of 12 kV and a reference pressure of 3 × 10 −7 mbar.Temperature programmed desorption/ reduction (TPD/TPR) experiments were carried out on a chemisorption apparatus (Micromeritics, AutoChem II 2920).TG and DSC quantities were measured using a simultaneous thermal analyzer (PerkinElmer STA-8000).N 2 was used as the carrier gas, and the samples were heated up to 100 °C at a heating rate of 10 °C min −1 to pretreat moisture and other impurities on their surface, and then heated up to 800 °C at a heating rate of 10 °C min

Results and discussion
3.1 Characterization of catalysts physical properties 3.1.1XRD.The physical phases present in the catalysts were analyzed by X-ray diffraction.As shown in Fig. 1, all samples contained peaks near 25.2°, 37.7°, 48.0°, 53.9°, and 54.9°, which belonged to anatase TiO 2 (JCPDS no.21-1272).In addition, weak diffraction peaks of CeO 2 (JCPDS no.43-1002) were detected near 28.6°, 33.0°, which indicated that some CeO 2 was present on the catalyst surface.La 2 O 3 and V 2 O 5 diffraction peaks were not observed for all samples, indicating that they are poorly crystallized and highly dispersed on the surface of the catalysts, which allows the active components to interact more fully and thus enhance the catalytic activity. 17The absence of neither crystallographic changes nor the appearance of new diffraction peaks in the poisoned samples compared with the fresh samples may be due to the low content of sulfate species generated on the catalyst surface or that they were amorphous.
Fig. 2 shows the N 2 adsorption-desorption isotherms and BJH pore size distribution.As shown in Fig. 2(a), all samples showed typical H3 hysteresis loops at higher relative pressures in the latter half (P/P 0 = 0.7-1.0),which is usually associated with capillary coalescence in the catalyst. 18The curves were a type IV isotherm according to the IUPAC classication, indicating that the catalysts had mesoporous structures.From Fig. 2(b), it is clear that all samples showed a wide range of pore sizes, further conrming that the catalysts are mesoporous and are made up of a variety of irregularly shaped particles stacked together.In addition, the specic surface area, total pore volume, and average pore size obtained from N 2 adsorptiondesorption isotherms were calculated (Table 1).These values were slightly reduced aer adding V because it favored sintering.Aer poisoning, the catalyst's surface area, average pore volume, and total pore volume were drastically decreased, demonstrating that sulfate species had coated the catalyst's surface. 19.1.2TEM.Fig. 3 shows the TEM images, HR-TEM images and elemental distributions of Ce, La, Ti and V for 10Ce-2La/Ti and 10Ce-2La-1.4V/Ti.From Fig. 3(a) and (c), it can be seen that the catalysts are porous structures formed by the stacking of nanoparticles with uneven sizes, which can exhibit good catalytic activity.In the HR-TEM images in Fig. 3(b) and (d), the crystal spacing of 0.35 nm corresponds to the (101) facets of anatase TiO 2 , 0.19 nm corresponds to the (220) crystalline facets of CeO 2 , and 0.17 nm can be indexed to the (202) crystalline facets of TiO 2 .Lattice fringes of lanthanum oxide and vanadium oxides were not found in any of the catalysts, conrming that they existed in an amorphous structure.It can also be observed that the elemental distribution of Fig. 3(e-h) and (i-k) shows that the elements Ce, La, Ti, and V are all uniformly distributed in the catalyst. 202 Catalyst surface acidity NH 3 -TPD was used to investigate the acidic properties of the catalyst surface (Fig. 4).While the peaks of medium and strong acidic sites in the temperature range of 200-500 °C were associated with the desorption of NH 4 + from Brønsted acidic sites, the peaks of weak acidic sites below 200 °C corresponded to physisorbed NH 3 on the catalyst.The peaks of strong acidic sites >500 °C were attributed to the desorption of liganded NH 3 on Lewis acidic sites.21 By comparison, we found that the weak acidic site peaks did not change much aer the introduction of V.However, the peak intensities and peak areas of the strong acidic sites in the catalysts increased signicantly and the   Paper RSC Advances detachment peaks moved toward higher temperatures, which may be due to the interaction of the different active components, which in turn led to an increase in the length of the metal-oxygen coordination bonds and the surface acidity.22 Meanwhile the decrease in the peak area of the strong acidic sites indicates that the medium-strong and weak acidic sites play a greater role in the catalytic activity of the catalyst during the catalytic reaction.In addition, it was observed that the peak intensity of the weak acidic sites of the 10Ce-2La/Ti catalyst became smaller aer poisoning, which may be due to the reaction of SO 2 on the surface of the catalyst to generate sulfate to reduce the number of acidic sites, which led to the weakening of the catalyst's ability to adsorb NH 3 .In contrast, the peak position of catalyst 10Ce-2La-1.4V/Ti-Snot only shied toward high temperature, but also the peak intensity and peak area of the medium-strong acidic sites increased instead, indicating that the sulfation of the catalyst increased the NH 3 uptake by the medium, which is one of the reasons that the catalyst exhibits good water and sulfur.23

Catalyst redox properties
XPS was used to investigate the chemical environment of the catalyst surface elements.The XPS spectra of catalyst Ce 3d are shown in Fig. 5(a), in which ten peaks were tted to the spectra of all the samples.These were identied with "U" and "V" for the 3d 5/2 and 3d 3/2 spin-orbit components of Ce, respectively, which were located at 879 eV (V 0 ), 885 eV (V 2 ), 900 eV (U 0 ), and 905 eV (U 2 ).These belong to the 3d 10 4f 1 initial electronic state of Ce 3+ , and the remaining peaks were related to the 3d 10 4f 0 state of Ce 4+ .The existence of the double peaks of V 0 /U 0 and V 2 /U 2 suggests the presence of partially-reduced oxygen vacancies in the catalyst. 24The tted integrated peak area ratio of Ce 3+ and total Ce was used to determine the relative content of Ce 3+ on the catalyst surface, as shown in Table 2 by the formula Ce 3+ (%) = (S V 0 + S V 2 + S U 0 + S U 2 )/(S U + S V ) × 100%. 25Aer introducing V, Ce and La concentrations on the catalyst surface considerably rose, proving that V interacted with the active ingredients and had an impact on how many atoms were present there.The higher proportion of Ce 3+ , was attributed to an increase in the Ce concentration and also to the movement of free carriers on the catalyst surface towards the CeO 2 surface, which had a higher binding energy (eV). 26A large fraction of Ce 3+ on the catalyst surface enhanced the formation of oxygen vacancies, unsaturated chemical bonds, and charge imbalance, all of which boost the NO oxidation reaction. 27,28Fig. 5(b) depicts the catalyst's La 3d XPS spectra, in which La 3d 5/2 and La 3d 3/2 peaks emerged in all spectra at 834 eV and 37 eV, as well as at 852 eV and 855 eV, respectively. 29Fig. 5(c) shows the O 1s XPS spectrum of the catalyst, in which peaks belonging to lattice oxygen and surface unsaturated oxygen (labeled O b and O a ), can be observed at positions 528.7-529 eV and 530.1-530.7 eV.Because surface unsaturated oxygen O a has a higher migration rate than lattice oxygen O b , O a increased the number of surface oxygen vacancies, which were more favorable to the SCR reaction. 30The relative atomic percentages of these oxygen species on the catalyst surface were estimated from the areas of the tted peaks, as listed in Table 2.The O a /(O a + O b ) ratio on the catalyst surface signicantly increased aer adding V, which is consistent with the Ce 3d XPS analysis and is one of the reasons for the good activity of the catalyst.Fig. 5(d) showed the V 2p XPS spectra before and aer catalyst poisoning, in which the V 4+ peak was located at 515.2 eV, and the V 5+ peaks were located at 516.4 eV and 523.1 eV.This indicates that V on the catalyst surface mainly existed as V 5+ . 31The ratio of V 4+ /(V 4+ + V 5+ ) on the catalyst surface increased signicantly aer poisoning.Ce 3+ also increased signicantly, but the concentrations of Ce and V, the main active components on the surface of the catalyst, decreased.This was because SO 2 reacted with the active components to produce sulfate in the presence of O 2 . 7Similarly, the signicant increase in the O a /(O a + O b ) ratio on the surface aer catalyst poisoning was because the higher Ce 3+ ratio promoted the creation of more surface oxygen vacancies.This was because oxygen from sulfate species was formed on the sample surface aer sulfation, as well as chemisorbed oxygen. 32ig. 5(e) shows the S 2p XPS spectra of the catalyst before and aer poisoning, in which peaks belonging to the spin-orbit components of S 6+ 2p 3/2 and S 6+ 2p 1/2 can be observed at 167.7-168 eV and 168.8-169 eV aer tting.This implies that S in +6 valence state was mostly present on the catalyst surface as SO 4  2− . 33,34Aer poisoning, the XPS peaks of all elements on the catalyst changed to higher binding energies, showing that the electron cloud density around each element was lowered to varying degrees.All were involved in the sulfation reaction.Peaks of Ce, La, O, Ti, V, S and C were observed in the full spectrum of the catalyst in Fig. 5

(f).
As shown in Fig. 6, the redox properties of catalysts before and aer 10Ce-2La/Ti and 10Ce-2La-1.4V/Tipoisoning were tested using H 2 -TPR.Measurements were performed using CuO as the standard, and the hydrogen consumption for the reduction peaks between 400 °C and 700 °C was quantitatively calculated (Table 3).10Ce-2La/Ti displayed a large H 2 reduction peak, which was attributed to the reduction of surface-liganded unsaturated Ce 4+ , which played an important role in the oxidation reaction. 35Aer introducing V, the catalyst reduction peaks shied toward lower temperatures, and the H 2 consumption increased signicantly (1.37 / 1.51 mmol g −1 ).This indicates strong interactions between V oxides and Ce  Paper RSC Advances oxides on the catalyst, which enhanced the reducing properties of the catalyst, which promoted catalytic cycling during SCR reaction. 36A strong peak was observed at 603 °C in the poisoned 10Ce-2La/Ti-S sample, which was shied to a higher temperature relative to the fresh sample.This shi was primarily due to the reduction of S 6+ to S 4+ in sulfate, along with the reduction of certain metal oxides. 37To study the degree of metal oxide reduction in the catalyst 10Ce-2La-1.4V/Ti-S, the peaks were divided into aand b-peaks by tting, and the tted area ratio was used to calculate the amount of hydrogen consumed for each peak.The a peak was associated with the reduction of metal oxides on the catalyst surface, while the b peak was related to the reduction process of sulfate species. 36The position of the metal oxide reduction peak in 10Ce-2La-1.4V/Ti-S was essentially unchanged, and the hydrogen consumption of a 1 was larger than that of a 2 (1.54 > 0.89 mmol g −1 ).That of b 1 was less than that of b 2 (1.00 < 1.157 mmol g −1 ), indicating that co-doping V with 10Ce-2La/Ti limited the interactions between SO 2 and active metal oxide.

SO 2 desorption and reaction
SO 2 -TPD was used to investigate the adsorption capacity of the catalyst toward SO 2 .As shown in Fig. 7(a), the catalyst contained two main desorption peaks in the temperature ranges of 400-460 °C and 600-670 °C.The low-temperature peaks were related to the desorption of SO 2 from the catalyst surface, 38 while the high-temperature peaks were related to the desorption of SO 2   generated by the decomposition of sulfate species.The peak of 10Ce-2La/Ti at 665 °C represented the decomposition of cerium sulfate. 39In comparison, the desorption peak of 10Ce-2La/Ti at low temperature has a larger peak area, implying that more SO 2 molecules are absorbed on the surface than 10Ce-2La-1.4V/Ti, the addition of V reduced the amount of SO 2 molecules absorbed on the surface of the catalyst in order to alleviate competition between SO 2 and NO adsorption, and to prevent the occupation of more active sites.The addition of V enhanced the strength of the catalyst's high-temperature peak, indicating that SO 2 molecules were adsorbed onto V 2 O 5 species, shielding the catalyst's major active sites. 40o study the substances generated on the surface of the samples during the catalytic reaction, the poisoned samples were analyzed using thermogravimetric analysis, and the results are shown in Fig. 7(b) and (c).The two samples' weight reduction procedure was separated into three steps: step I (40-200 °C) was mostly ascribed to the desorption of water from the samples; step II (200-600 °C) was attributed to the breakdown of ammonium sulfate and ammonia bisulfate on the catalyst surface; and step III (600-800 °C) to the decomposition of sulfates in the samples. 36,4110Ce-2La-1.4V/Ti-Sshowed an absorption peak related to the decomposition of sulfate (511 °C) and an exothermic peak related to the phase transition of the composite metal oxide (719 °C) above 500 °C.The 10Ce-2La/Ti-S catalyst had a higher decomposition temperature (540 °C), which was attributed to the low decomposition temperature of VOSO 4 generated by SO 2 with V.The 10Ce-2La/1.4V/Ti-S catalyst had the highest decomposition temperature (540 °C).In addition, the weight loss of ammonium sulfate or ammonia bisulfate on the surface of the 10Ce-2La-1.4V/Ti-Scatalyst was larger (0.609%) than that of 10Ce-2La/Ti-S (0.485%).It is shown that the introduction of V protects the main active component of the catalyst from reacting with SO 2 and inhibits the formation of sulfate species on the catalyst surface, thus enhancing the water and sulfur resistance of the catalyst.Absorption peaks near 1312-1343 (1350) cm −1 and 1414 (1413) cm −1 belonged to NH 4 + at Brønsted acid sites. 44,45The asymmetric bending vibration of NH 3 coupled to Lewis acid sites was responsible for the lower absorption peak at 1601 (1607) cm −1 . 44The peak at 3253-3385 cm −1 was attributed to the stretching vibration of NH 3 at the Lewis acid sites. 4610Ce-2La-1.4V/Tishows a weak adsorption band near 1550 cm −1 , which is due to the oxidation of NH 3 on the catalyst surface by surface-active oxygen to generate an important intermediate product, -NH 2 , in the catalytic reaction -NH 2 facilitates the conversion of NO, and promotes the smooth progress of the SCR reaction. 47In comparison, the intensity of the peaks related to the Brønsted acidic sites increased aer the introduction of V, and NH 3 was more easily activated.This indicates that Brønsted acids dominated the reaction, and the addition of V increased the proportion of Brønsted acids.

In situ infrared spectroscopy of catalysts
The DRIFTS spectra of the catalysts 10Ce-2La/Ti and 10Ce-2La-1.4V/Tiaer 30 minutes of NO (500 ppm) + O 2 adsorption are shown in Fig. 8(c) and (d).Nitrate bridged ions and free nitrate ions, respectively, were responsible for the bands at 1355 (1352) cm −1 and 1600 (1602) cm −1 . 48The absorption peak belonging to the nitrite species trans-N 2 O 2 2− was located at 1415 cm −1 .Weakly adsorbed gas-phase NO 2 was thought to be active in the band at 1625-1627 cm −1 because it may react quickly with gaseous NH 3 . 45Fig. 9(a) and (b) show the DRIFTS proles of the catalysts 10Ce-2La/Ti and 10Ce-2La-1.4V/Tiaer 30 min of preadsorption of NH 3 (500 ppm) followed by NO (500 ppm) + O 2 (3%) adsorption for 30 min.In the NH 3 adsorption curve, the peaks near 1317 and 1415 (1417) cm −1 belonged to NH 4 + on the Brønsted acid sites, and the intensity of the peak at 1415 (1417) cm −1 did not change greatly with the passage of NO + O 2 .This was attributed to the coincidence of the peak at 1600 (1604) cm −1 for the NH 3 and Lewis acid sites with the stronger peaks of rapidly-generated nitrate at 1604 and 1625 cm −1 aer owing NO + O 2 .This indicates that pre-adsorbed NH 3 did not directly react with gas-phase NO and that the NH 3 -SCR did not involve gas-phase NO.With the passage of NO + O 2 , the intensity of the 1415 (1417) cm −1 peak did not change signicantly, because, at this time, the 1600 (1604) cm −1 peaks of NH 3 and Lewis acid sites overlapped with those of the stronger nitrate peaks at 1604 and 1625 cm −1 .This suggests that pre-adsorbed NH 3 did not react directly with gas-phase NO and that the NH 3 -SCR reaction did not proceed via the Eley-Rideal mechanism.Fig. 9(c) and (d) show the DRIFTS spectra of the catalysts 10Ce-2La/Ti and 10Ce-2La-1.4V/Tipre-adsorbed with NO (500 ppm) + O 2 (3%) for 30 min and then allowed to adsorb NH 3 (500 ppm) for 30 min.Prior to the passage of NH 3 , peaks attributed to nitrates were observed at 1349 (1351) cm −1 , 1624 (1625) cm −1 , 1415 (1418) cm −1 , and 1600 (1604) cm −1 , similar to those in Fig. 8(a) and (b). 49The nitrate peaks at 1602 and 1625 cm −1 disappeared rapidly upon the passage of NH 3 , it indicates that adsorbed NO can react with NH 3 in the adsorbed state, so the reaction mechanism is Langmuir-Hinshelwood mechanism.In an L-H reaction mechanism, the reaction orders of both NO and NH 3 are zero. 506 Catalyst denitrication performance testing Fig. 10(a) shows the NO conversion versus temperature curves of the catalysts aer adding different ratios of V.The overall activity of the catalyst was improved aer adding V, and the starting temperatures were all lower.As the weight percent of V increases, the catalyst denitrication efficiency improved and then became worse, and the temperature window gradually became wider.The optimal catalytic efficiency was obtained when the weight ratio of V 2 O 5 was 1.4, and the denitrication efficiency remained above 80% in the temperature range of 270-490 °C.The NO conversion reached a maximum of 99% at 330 °C, and the catalytic activity was greatly improved compared with the original rare-earth-based catalyst.In addition to catalytic activity, N 2 selectivity is another important index for evaluating a catalyst's performance.As the temperature of the reaction increased, N 2 O and NO 2 were produced, causing the N 2 selectivity to slowly decrease, as shown in Fig. 10(b) and (c).In the reaction temperature interval, the N 2 selectivity and N 2 O concentration of catalyst was slightly improved by the addition of V.The lower N 2 O generation resulted in better overall N 2 selectivity than the original catalyst.
The stability of the catalyst was tested by passing 300 ppm SO 2 under constant test conditions and the results were shown in Fig. 10(d).Aer 2.5 h of the water and sulfur feeding, the NO conversion of 10Ce-2La/Ti decreased from the initial 97% to 44%, which maintained almost unchanged for 52 h.If the feeding of water and sulfur was cut off, the NO conversion recovered to 78%.Compared with 10Ce-2La/TiO 2 , the V-doped catalyst had a higher NO conversion in the presence of water and sulfur, which remained above 85% and recovered to 99% aer shutting down water and sulfur.Fig. 10(e) shows the N 2 selectivity comparison in the water and sulfur resistance stabilization experiments.Both the two catalysts showed a high N 2 selectivity over than 99% and the N 2 selectivity was not inuenced by water and sulfur.It can also seen that the N 2 selectivity was slightly improved by the addition of V.The turnover frequency (TOF) value calculated from eqn (5) is shown in Fig. 10(f).The 10Ce-2La-1.4V/Ti catalyst displayed better TOF values at each temperature, demonstrating that single Ce atoms on the catalytic surface were the primary active sites and displayed the best intrinsic activity under identical circumstances.Additionally, the reaction rate of NO on each square meter of a catalyst within the investigated temperature range was calculated according to eqn (4).It was plotted as ln(rate) against 1000/T, and the Arrhenius activation energy of the reaction was determined by tting.As shown in Fig. 10(g), the surface activation energy of 10Ce-2La/Ti was 24.6 kJ mol −1 , which was higher than that of 10Ce-2La-1.4V/Ti(13.5 kJ mol −1 ).The lower activation energy suggests an improved denitrication efficiency.

Conclusion
V-doped 10Ce-2La-xV/Ti catalysts were prepared by a ball milling method, and it was found that the rare earth Ce exerted a synergistic effect with V.The composite carrier-type oxides exhibited a good activity window and denitrication performance, as well as excellent water and sulfur resistance when the weight ratio of V 2 O 5 was 1.4.The prepared catalysts showed mesoporous structures, and the ball milling method enhanced the interactions between different active components, resulting in good dispersion of the active components.V doping improved the surface acidity of the catalysts, and Brønsted acids on the surface of the catalysts played the dominant role.The greater number of acid sites adsorbed more NH 3 that participated in the reaction and also prevented SO 2 poisoning.The Ce 3+ and O ratios on the catalyst surface were also enhanced by the addition of V, which was signicant for the SCR-NH 3 reaction.V limited the interactions between SO 2 and the main active component (metal oxides), reduced the amount of SO 2 adsorbed on the catalyst surface, and inhibited the formation of sulfate species on the catalyst surface.These all improved the sulfur resistance of the catalyst.The NH 3 -SCR on the catalyst surface proceeded via the Langmuir-Hinshelwood mechanism, with Brønsted acid sites playing the dominant role.The better denitrication activity and sulfur resistance of 10Ce-2La-xV/Ti catalysts make them promising for various applications.

− 1 .
The in situ IR tests were performed on an in situ FTIR spectrometer (Tensor 27).The tests were consisted by four parts: NH 3 adsorption, NO + O 2 adsorption, NO + O 2 reaction with preadsorbed NH 3 and NH 3 reaction with pre-adsorbed NO + O 2 .

Fig. 10
Fig. 10 Catalytic performance versus temperature curves of 10Ce-2La-xV/Ti series catalysts.(a) NO removal efficiency versus temperature transformation curves at different weight ratios V; (b) N 2 selectivity at different temperatures; (c) effect of temperature on the concentration of NO 2 and N 2 O production; (d) stability test of the catalysts for water and sulfur resistance; (e) N 2 selectivity at different time; (f) TOF values (200-225 °C).(g) Arrhenius plots.

Table 3
Catalyst H 2 -TPR peak positions and corresponding H 2 consumption