Marys Medicine

Doi:10.1016/j.apcatb.2009.09.029

Applied Catalysis B: Environmental 93 (2009) 194–204 Contents lists available at Applied Catalysis B: Environmental Effect of manganese substitution on the structure and activity of iron titanatecatalyst for the selective catalytic reduction of NO with NH3 Fudong Liu, Hong He , Yun Ding, Changbin Zhang State Key Laboratory of Environmental Chemistry and Ecotoxicology, Research Center for Eco-Environmental Sciences, Chinese Academy of Sciences, Beijing 100085, PR China Selective catalytic reduction (SCR) of NO with NH3 over manganese substituted iron titanate catalysts Received 9 June 2009 was fully studied. The low temperature SCR activity was greatly enhanced when partial Fe was Received in revised form 10 September 2009 substituted by Mn, although the N2 selectivity showed some decrease to a certain extent. The Mn Accepted 22 September 2009 substitution amounts showed obvious influence on the catalyst structure, redox behavior and NH Available online 25 September 2009 adsorption ability of the catalysts. Among FeaMn1aTiOx (a = 1, 0.75, 0.5, 0.2, 0) serial catalysts,Fe0.5Mn0.5TiOx with the molar ratio of Fe:Mn = 1:1 showed the highest SCR activity, because the interaction of iron, manganese and titanium species in this catalyst led to the largest surface area and the Selective catalytic reduction highest porosity, the severest structural distortion and most appropriate structural disorder, the Iron titanate catalystMn substitution amounts enhanced oxidative ability of manganese species, the highest mobility of lattice oxygen, the proper ratio Low temperature activity of Brønsted acid sites and Lewis acid sites together with the enhanced NOx adsorption capacity.
ß 2009 Elsevier B.V. All rights reserved.
Al2O3/SiO2/AC (activated carbon) and Mn–Ce, Mn–Cumixed oxides . Previous studies showed that in Fe- Selective catalytic reduction (SCR) of NO with NH3 is one of the containing SCR catalysts, the introduction of Mn could obviously most efficient and economic technologies for the removal of enhance the low temperature activity probably due to the nitrogen oxides (NOx) from stationary and mobile sources, and the synergistic effect between iron and manganese species. It was most widely used catalyst system is V2O5–WO3 (MoO3)/TiO2 .
reported that the introduction of lanthanide elements (such as La, Because of some inevitable disadvantages in practical application, Ce and Pr) and the third main group element In could also improve such as the narrow operation temperature window , high the activity, stability or SO2 durability of the SCR catalysts using conversion of SO2 to SO3 at high temperatures and the toxicity NH3 or hydrocarbons as reducing agent . Therefore, based of vanadium pentoxide to environment and human health , on our iron titanate catalyst, we can also substitute partial Fe by more and more researchers are focusing on the development of other elements to adjust its physicochemical properties, expecting new SCR catalysts. In our previous study , we have developed to enhance the low temperature SCR activity.
an environmentally friendly novel iron titanate catalyst in In this paper, five kinds of elements including La, Ce, Pr, In and Mn crystallite phase with specific Fe–O–Ti structure, which showed were introduced into the iron titanate catalyst, among which Mn excellent SCR activity, N2 selectivity and H2O/SO2 durability in the showed the best promoting effect. Based on this result, we further medium temperature range. However, the catalytic activity was investigated the influence of Mn substitution amounts on the not high enough for the application in denitrogenation of exhaust catalyst structure and catalytic activity using various characteriza- gas with low temperature, such as the flue gas after dust removal tion methods. The structural properties were characterized using N2 and desulfurization from coal-fired power plants and the exhaust physisorption, powder X-ray diffraction (XRD) and X-ray absorption gas from diesel engines in cold-start process. Therefore, it is very fine structure (XAFS) methods. Then, X-ray photoelectron spectra necessary to modify this iron titanate catalyst to improve the low (XPS) and H2-temperature programmed reduction (H2-TPR) were temperature activity, which is crucial for the practical utilization.
conducted to evaluate the variation of redox properties during the Manganese oxides usually show good SCR activity in the low substitution process. Finally, temperature programmed desorption temperature range, such as pure MnOx , MnOx loaded on TiO2/ of NH3 and NOx (NH3-TPD and NOx-TPD) together with in situ diffusereflectance infrared Fourier transform spectroscopy (in situ DRIFTS)of NH3 and NOx adsorption was carried out to reveal the evolution ofadsorption ability of reactants, which is important for the SCR * Corresponding author at: P.O. Box 2871, 18 Shuangqing Road, Haidian District, reaction. The promoting mechanism of Mn on the low temperature Beijing 100085, PR China. Tel.: +86 10 62849123; fax: +86 10 62849123.
E-mail address: (H. He).
SCR activity of iron titanate catalyst was proposed accordingly.
0926-3373/$ – see front matter ß 2009 Elsevier B.V. All rights reserved.
doi: F. Liu et al. / Applied Catalysis B: Environmental 93 (2009) 194–204 procedures . During the EXAFS data processing procedure, theback-subtracted EXAFS function was firstly converted into k space 2.1. Catalyst synthesis and activity test and weighted by k3 in order to compensate for the diminishingamplitude because of the decay of the photoelectron wave. The Fe0.9M0.1TiOx (M = La, Ce, Pr, In, Mn) and FeaMn1aTiOx with Fourier transforming of the k3-weighted EXAFS data was performed different Mn substitution amounts (a = 1, 0.75, 0.5, 0.2, 0) were in the range of k = 2–11.01 A˚1 for both Fe-K-edge and Mn-K-edge prepared by co-precipitation method using Fe(NO3)39H2O, with a Hanning function window.
Ti(SO4)2, relevant metal nitrates as precursors and NH3H2O XPS were recorded on a Scanning X-ray Microprobe (PHI (25 wt%) as precipitator. The precipitate cake was filtrated and Quantera, ULVAC-PHI, Inc.) using Al Ka radiation (1486.7 eV).
washed using distilled water, followed by desiccation at 100 8C for Binding energies of Fe 2p, Mn 2p, Ti 2p and O 1s were calibrated 12 h and calcination at 400 8C for 6 h in air condition. Pure oxides using C 1s peak (BE = 284.8 eV) as standard.
Prior to H2-TPR experiment, the samples (100 mg) were Fe(NO3)39H2O, Mn(NO3)2 and Ti(SO4)2 using the same precipita- pretreated at 300 8C in a flow of 20 vol.% O2/He (30 ml/min) for tion method for the comparison of SCR activity. The state-of-the- 0.5 h and cooled down to the room temperature (30 8C). Then the art SCR catalyst 4.5 wt% V2O5–10 wt% WO3/TiO2 was also prepared temperature was raised linearly to 900 8C at the rate of 10 8C/min using conventional wet impregnation method as reference in the in a flow of 5 vol.% H2/Ar (30 ml/min). The H2 signal (m/z = 2) was SCR activity test.
monitored online using a quadrupole mass spectrometer (HPR20, The NH3-SCR, NO oxidation and NH3 oxidation tests were Hiden Analytical Ltd.).
carried out over 0.6 ml catalysts (ca. 200–350 mg due to the NH3-TPD and NOx-TPD were also performed using the same different catalyst densities) in a fixed-bed quartz tube reactor and quadrupole mass spectrometer to record the signals of NH3 (m/ the reaction conditions were as follows: 500 ppm NO and (or) z = 16 for NH2 and m/z = 15 for NH) and NOx (m/z = 30 for NO and 500 ppm NH3, 5 vol.% O2, 1000 ppm CO (when used), 5 vol.% CO2 m/z = 46 for NO2). Prior to TPD experiments, the samples (100 mg) (when used), 5 vol.% H2O (when used), N2 balance and gas hourly were also pretreated at 300 8C in a flow of 20 vol.% O2/He (30 ml/ space velocity (GHSV) = 50 000 h1. The effluent gas was analyzed min) for 0.5 h and cooled down to the room temperature (30 8C).
using an FTIR spectrometer (Nicolet Nexus 670) equipped with a Then the samples were exposed to a flow of 2500 ppm NH3/Ar or heated, low volume multiple-path gas cell (2 m). NOx conversion 2500 ppm NO + 10 vol.% O2/Ar (30 ml/min) at 30 8C for 1 h, ) were calculated as follows: following by Ar purge for another 1 h. Finally, the temperature 2 selectivity (SN2 was raised to 500 8C in Ar flow at the rate of 10 8C/min.
The in situ DRIFTS experiments of NH3/NOx adsorption over x ¼ ½NO þ ½NO2 FeaMn1aTiOx catalysts were performed on an FTIR spectrometer(Nicolet Nexus 670) equipped with an MCT/A detector cooled by liquid nitrogen. An in situ DRIFTS reactor cell with ZnSe window in þ ½NH3in  ½NO2out  2½N2Oout ½NOin þ ½NH3in (Nexus Smart Collector) connected to a purging/adsorption gascontrol system was used for the NH3/NOx in situ adsorptionexperiments. The temperature of the reactor cell was controlled 2.2. Characterizations precisely by an Omega programmable temperature controller.
Prior to NH3/NOx adsorption, the samples were pretreated at N2 adsorption–desorption isotherms were obtained at 77 K 400 8C in a flow of 20 vol.% O2/N2 for 0.5 h and cooled down to using a Quantachrome Autosorb-1C instrument. Prior to N2 30 8C. The spectra of different catalysts at 30 8C were collected in adsorption, the samples were degassed at 300 8C for 4 h. The flowing N2 and set as backgrounds, which were automatically surface areas were determined by BET equation in 0.05–0.35 subtracted from the final spectra after NH3/NOx adsorption. Then partial pressure range. The pore volumes, average pore diameters the samples were exposed to a flow of 500 ppm NH3/N2 or and pore size distributions were determined by BJH method from 500 ppm NO + 5 vol.% O2/N2 (300 ml/min) at 30 8C for 1 h, the desorption branches of the isotherms.
following by N2 purge for another 0.5 h. All spectra were recorded Powder XRD measurements were carried out on a computer- by accumulating 100 scans with a resolution of 4 cm1.
ized Rigaku D/max-RB Diffractometer (Japan, Cu Ka as radiationresource). The data of 2u from 108 to 908 were collected at 4 8/min 3. Results and discussion with the stepsize of 0.028.
XAFS experiments were implemented on U7C beamline of 3.1. Catalytic performance National Synchrotron Radiation Laboratory (NSRL), of which thestorage ring was operated at 0.8 GeV with a maximum current of 3.1.1. SCR activity of Fe0.9M0.1TiOx catalysts (M = La, Ce, Pr, In, Mn) 300 mA. The hard X-ray beam was from a three-pole super- shows the NOx conversion as a function of temperature in conducting Wiggler with a magnetic field intensity of 6 T. A fixed- the NH3-SCR reaction over Fe0.9M0.1TiOx catalysts (M = La, Ce, Pr, exit Si(1 1 1) double-crystal monochromator was used to reduce the In, Mn). From the results we can see that, the substitution of partial harmonic content of the monochrome beam. The incident and Fe with other elements could indeed influence the SCR activity of output beam intensities were monitored and recorded using iron titanate catalyst. At temperatures below 250 8C, the Mn and Ce ionization chambers filled by Ar/N2. A Keithley Model 6517 substitutions could obviously enhance the NOx conversions, while Electrometer was used to collect the electron charge directly. Before the La, Pr and In substitutions decreased the NOx conversions to a XAFS measurement, the catalyst samples were crushed into fine certain extent. Moreover, the Ce substitution lowered the NOx powder above 200 mesh and coated onto transparent adhesive conversion at relatively high temperatures above 300 8C, while the tapes. The XAFS spectra (X-ray absorption near-edge spectroscopy, Mn substitution did not show such an obvious negative influence.
XANES and extended X-ray absorption fine-structure spectroscopy, Therefore, we chose Mn as the substitution element to carry out EXAFS) of Fe-K-edge and Mn-K-edge were recorded in transmission our further investigations, such as the effect of Mn substitution mode at room temperature in air condition. The collected XAFS data amounts on NH3-SCR, NO oxidation and NH3 oxidation activities, were calibrated according to standard Fe2O3 and MnO2 samples and together with the relationship between catalyst structure and then analyzed using Viper software package according to standard catalytic activity.

F. Liu et al. / Applied Catalysis B: Environmental 93 (2009) 194–204 Fig. 1. NOx conversion as a function of temperature in the NH3-SCR reaction overFe0.9M0.1TiOx catalysts. Reaction conditions: [NO] = [NH3] = 500 ppm, [O2] = 5 vol.%,N2 balance, total flow rate 500 ml/min and GHSV = 50 000 h1.
3.1.2. SCR activity of FeaMn1aTiOx catalysts A shows the results of NOx conversion and N2 selectivity in the SCR reaction over FeaMn1aTiOx (a = 1, 0.75, 0.5, 0.2, 0) catalystsand pure oxides including Fe2O3, MnOx and TiO2 as a function oftemperature from 75 to 400 8C. As we can see, pure TiO2 showed noSCR activity below 350 8C and only 30% NOx conversion was obtainedat 400 8C. Pure Fe2O3 and MnOx showed very narrow operationtemperature windows in the high and low temperature ranges,respectively, and both of the maximum NOx conversions could notreach 100%. Furthermore, the N2 selectivity over these two sampleswas rather low as shown in the inserted figure. The coexistence of Feand Ti in FeTiOx greatly enlarged the operation temperature windowand the NOx could be completely reduced from 225 to 350 8C withhigh N2 selectivity. When partial Fe was substituted by Mn, the NOxconversions in the relatively low temperature range had an obvious x conversion and N2 selectivity (inserted) in the NH3-SCR reaction over 0.5Mn0.5TiOx with the molar ratio of Fe:Mn = 1:1 showed FeaMn1aTiOx catalysts and pure oxides including Fe2O3, MnOx and TiO2, reaction the best activity, over which NOx was completely reduced at about [NO] = [NH3] = 500 ppm, [O2] = 5 vol.%, N2 balance and GHSV = 175 8C. However, the continuing substitution of Fe by more Mn led to 50 000 h1. (B) Comparison of apparent SCR activity over Fe an activity decrease, and the NO (GHSV = 50 000 h1 without H (GHSV = 100 000 h1 with x conversions over Fe0.2Mn0.8TiOx 10 vol.% H2O), 4.5 wt% V2O5–10 wt% WO3/TiO2, Fe/ZSM-5 (NO conversion on x from 75 to 250 8C were even lower than that over Fe(58)-ZSM-5(10) from Ref. , reaction conditions: [NO] = [NH 0.75Mn0.25TiOx. The apparent SCR activity at low temperatures [O2] = 2 vol.% and GHSV = 46 000 h1), Fe/HBEA (NOx conversion on 0.25Fe/HBEA from Ref. reaction conditions: [NO] = [NH3] = 500 ppm, [O2] = 5 vol.% and < Fe0.2Mn0.8TiOx < Fe0.75Mn0.25TiOx < Fe0.5Mn0.5TiOx. In previous GHSV = 50 000 h1) and Cu/ZSM-5 (NO conversion on CuZSM-5-124-fresh from study by Qi and Yang , they also observed that Fe–Mn/TiO Ref. reaction conditions: [NO] = [NH 3] = 500 ppm, [O2] = 5 vol.%, [H2O] = 10 vol.% and GHSV = 100 000 h1).
catalyst with Fe:Mn = 1:1 showed the highest activity among theirloaded type catalysts in the SCR reaction. This could be attributed tothe strong interaction between Fe and Mn which led to high the temperature for the 50% NOx conversion over Fe0.75Mn0.25TiOx dispersion of active phases and thus the high SCR activity.
was ca. 50 8C lower than that over the traditional V2O5–WO3/TiO2 Although the substitution of Fe by Mn in iron titanate catalyst catalyst, making it possible to be utilized for the removal of NOx from could enhance the SCR activity, the N2 selectivity had an obvious actual flue gas with low exhaust temperature. To fully compare the decrease owing to the production of N2O, especially at high SCR activity of Fe0.75Mn0.25TiOx with that of Cu/ZSM-5 reported by temperatures above 200 8C. The N2 selectivity at high temperatures Park et al. , we also performed another activity test under the FeTiOx > Fe0.75Mn0.25 identical reaction conditions with those in literature, i.e.
TiOx > Fe0.5Mn0.5TiOx  Fe0.2Mn0.8TiOx  MnTiOx. Considering the GHSV = 100 000 h1 and 10 vol.% H2O. Under the high GHSV and NOx conversion and N2 selectivity, we chose Fe0.75Mn0.25TiOx as H2O concentration, the operation temperature window of model catalyst over which the N2 selectivity could maintain above Fe0.75Mn0.25TiOx greatly shifted towards high temperature range, 90% at temperatures below 300 8C, to compare with the state-of-the- thus resulting in lower SCR activity below 300 8C than that of Cu/ art SCR catalysts including the traditional V2O5–WO3/TiO2 catalyst ZSM-5. Therefore, GHSV is an important factor to be considered in and Fe, Cu exchanged zeolites. As shown in B, Fe/ZSM-5 catalyst design and the H2O durability of Fe0.75Mn0.25TiOx still needs and Fe/HBEA catalysts showed good SCR activity at relatively to be improved in our future work. Furthermore, the influences of high temperatures, over which the maximum NO/NOx conversions GHSV and O2 concentration on NOx conversions over Fe0.75Mn0.25- were obtained above 350 and 250 8C, respectively. As for our TiOx were also investigated (Figs. S1 and S2 in Supporting Fe0.75Mn0.25TiOx catalyst, the low temperature SCR activity was Information). The NOx conversions over this catalyst still could much better than that of Fe exchanged zeolites, although the high get 100% above 250 8C even at a high GHSV of 100 000 h1, which is temperature SCR activity above 300 8C showed sharp decrease beneficial to the actual industrial application. The influence of O2 on resulting in a narrow operation temperature window. Remarkably, NOx conversions was more obvious at low temperatures than that at

F. Liu et al. / Applied Catalysis B: Environmental 93 (2009) 194–204 Fig. 3. Influence of CO, CO2 and H2O on NOx conversion in the NH3-SCR reaction over Fig. 4. (A) NO conversion in separate NO oxidation reaction and (B) NH3 conversion Fe0.75Mn0.25TiOx catalyst. Reaction conditions: [NO] = [NH3] = 500 ppm, [O2] = in separate NH3 oxidation reaction over FeaMn1aTiOx catalysts. Reaction 5 vol.%, [CO] = 1000 ppm (when used), [CO2] = 5 vol.% (when used), [H2O] = 5 vol.% conditions: [NO] = 500 ppm or [NH3] = 500 ppm, [O2] = 5 vol.%, N2 balance, total (when used), N2 balance, total flow rate 500 ml/min and GHSV = 50 000 h1.
flow rate 500 ml/min and GHSV = 50 000 h1.
high temperatures, implying that different SCR reaction mechan- obtained when the substitution amount was above 0.5. Although isms might be followed in different temperature ranges, just as those the NO oxidation activities of Fe0.2Mn0.8TiOx and MnTiOx were over unsubstituted iron titanate catalyst much higher than that of Fe0.75Mn0.25TiOx, the SCR activities overthe former two catalysts were still a little lower than that over the 3.1.3. Influence of CO, CO2 and H2O on the SCR activity latter one, as shown in This implies that the SCR activity over For practical purposes, the Fe0.75Mn0.25TiOx catalyst was also these catalysts is not only related with the NO oxidation activity, tested in the presence of CO, CO2 and H2O, and the results are but also related with some other structural or redox properties, shown in As we can see, the existence of CO and CO2 in the which will be discussed later in this paper.
feeding gas did not obviously influence the SCR activity below Previous study by Roy et al. showed that the N2 selectivity 200 8C, which is similar as the results shown by Balle et al.
in the SCR reaction had a strong inverse correlation with the However, the NOx conversion above 300 8C showed a slight oxidation of NH3, therefore the separate NH3 oxidation experi- increase, probably due to the reduction of NOx by CO when the ments were also conducted over FeaMn1aTiOx catalysts. As shown unselective oxidation of NH3 severely happened at high tempera- in B, the NH3 conversions had an obvious enhancement with tures: 2CO + 2NO ! N2 + 2CO2. To confirm this speculation, addi- the increasing of Mn substitution amounts, and the highest NH3 tional experiment concerning the reaction between NO and CO was conversions were obtained when the substitution amount was performed, during which partial NO was indeed reduced to N2 by above 0.5. However, the N2 selectivity showed an obvious decrease CO above 300 8C (as shown in ). On the other hand, the in NH3 oxidation reactions at the same time (see Fig. S4 in presence of H2O in the feeding gas significantly decreased the NOx Supporting Information), which is in accordance with the changing conversion below 200 8C mainly due to the blocking of active sites trend of N2 selectivity in the SCR reactions. This implies that . However, at temperatures above 300 8C the NOx conversion although the NO oxidation activity is enhanced when partial Fe is showed an obvious increase, which was due to the inhibition effect substituted by Mn which is beneficial to the promotion of SCR of H2O on the unselective oxidation of NH3. This point of view activity, the unselective oxidation of NH3 to N2O, NO or NO2 in the could be verified by the decrease of NH3 conversion and SCR conditions is also enhanced, resulting in the production of a enhancement of N2 selectivity in the SCR reaction over large amount of by-products. There should be a compromise of the Fe0.75Mn0.25TiOx in the presence of H2O (see Fig. S3 in Supporting SCR activity and N2 selectivity when we determine on the Mn Information). For the reaction condition containing CO, CO2 and substitution amount in practical application.
H2O, the SCR activity was very similar as that in the presence ofH2O alone, with the NOx conversion keeping at 100% from 200 to 3.2. Structural properties 3.2.1. N2 physisorption 3.1.4. NO and NH3 oxidation activities of FeaMn1aTiOx catalysts shows the pore size distributions of FeaMn1aTiOx It was reported that the enhancement of NO oxidation to NO2 catalysts derived from the desorption branches of N2 adsorption– over SCR catalysts could significantly promote the low tempera- desorption isotherms. In the diameter range below 3 nm, the N2 ture activity due to the occurrence of the ‘‘fast SCR'': NO + N- adsorbed volume per gram per nm increased in the following O2 + 2NH3 ! 2N2 + 3H2O The effect of NO2 and the FeTiOx < MnTiOx < Fe0.2Mn0.8TiOx < Fe0.75Mn0.25TiOx detailed ‘‘fast SCR'' reaction mechanism have been studied < Fe0.5Mn0.5TiOx. This means that the Fe0.5Mn0.5TiOx sample has extensively over conventional V2O5–WO3/TiO2 and Fe-zeolite the most abundant micropores or mesopores, which can supply catalysts (such as Fe/HBEA and Fe/ZSM-5) by many researchers more inner surface area for the occurrence of SCR reaction. The BET . In this study, the effect of Mn substitution amounts on surface areas in also followed such a sequence, which is in NO oxidation activity of iron titanate catalyst was also investigated well harmony with the sequence of SCR activity. The results in C and the results are shown in A. With the increasing of Mn show that Fe0.5Mn0.5TiOx possesses the largest pore volume due to substitution amounts the NO conversion to NO2 showed an the coexistence of iron and manganese species with the molar ratio obvious enhancement, and the maximum conversions were of Fe:Mn being 1:1, which is beneficial to the enhancement of SCR

F. Liu et al. / Applied Catalysis B: Environmental 93 (2009) 194–204 Fig. 5. N2 adsorption–desorption results of FeaMn1aTiOx catalysts: (A) pore size distributions; (B) BET surface areas; (C) pore volumes; (D) average pore diameters.
activity. With the increasing of Mn substitution amounts, the structure which showed high SCR activity. The substitution of Fe by average pore diameter in D became larger, which might be one Mn did not destroy the crystallite structures; furthermore, some of the reasons for the activity decline over Fe0.2Mn0.8TiOx and crystallites with Mn–O–Ti and Fe–O–Mn structures might also be formed, because either FeMnTiO4 or MnTiO3 also has somediffraction peaks in the positions of the broad bumps. The formation of new active phases might be another reason for the shows the XRD results of FeaMn1aTiOx catalysts together with the standard cards of FeTiO3, MnTiO3 and FeMnTiO4 in JCPDS(vertical lines). All of the samples showed no obvious diffraction patterns besides some broad bumps, implying that all samples For our FeaMn1aTiOx catalysts in crystallite phases, XAFS is a were in poor crystallization. The interaction between iron, suitable tool to characterize the structural information because it manganese and titanium species led to a highly dispersive state can be used to determine the local environment around specific of active phases, without forming iron oxides, manganese oxides or atoms, irrespective of crystallinity or dimensionality of the target titanium oxide particles. In our previous study , we have materials. A presents the normalized XANES of Fe-K-edge in concluded that the FeTiOx catalyst was mainly in the form of Fe-containing catalysts. All samples showed characteristic pre- crystallite phases of Fe2TiO5 and FeTiO3 with specific Fe–O–Ti edge peaks at 7114 eV, which could be attributed to 1s–3d dipolarforbidden transition The peak position and the peak shapecorresponded well with those of the ferric compounds in fourfoldor fivefold coordination indicating that the iron species in ourcatalysts was mainly in Fe3+ oxidation state. It is reported that thispre-edge peak will get additional intensity if the iron center is in anoncentral symmetric environment or through mixing of 3d and4p orbitals, which is caused by the breakdown of inversionsymmetry because of the structure distortion (i.e. bond-angledisorder) As the XANES spectra in this study have beenshifted vertically for comparison, we can directly read the pre-edgepeak intensities by subtracting the base line values from the peakvalues. The inserted figure is the enlargement of the spectra regiondenoted by the dashed rectangle to better discriminate the pre-edge peak intensities. For Fe0.5Mn0.5TiOx catalyst, the intensity ofthe pre-edge peak was largest, implying that when the molar ratioof Fe:Mn is 1:1, the interaction between these two species led to aseverest structure distortion of Fe–O coordination. presentsthe normalized XANES of Mn-K-edge in Mn-containing catalysts Fig. 6. XRD results of FeaMn1aTiOx catalysts and standard cards in JCPDS.
and all samples showed characteristic pre-edge peaks at 6541 eV, F. Liu et al. / Applied Catalysis B: Environmental 93 (2009) 194–204 Fig. 7. (A) XANES, (B) FT-EXAFS results of Fe-K-edge and (C) XANES, (D) FT-EXAFS results of Mn-K-edge in FeaMn1aTiOx catalysts.
which could also be attributed to the crystal field transition from in R space will only be relevant with the structural disorder (i.e.
the core 1s levels to the empty 3d levels and more or less 4p bond-length disorder). Lower peak intensity indicates higher hybridized by manganese ligands Fe0.5Mn0.5TiOx catalyst also degree of structural disorder. It was reported that the more had the largest intensity of pre-edge peak which can be disordering of the structure, the higher catalytic activity would be distinguished from the inserted figure, suggesting a severest obtained over catalysts for various reactions, such as over the Cu/ structure distortion of Mn–O coordination. It was reported that ZrO2 catalyst for the steam reforming of methanol , the Ni–Co– amorphous or crystallite materials with enormous structure B amorphous catalyst for the hydrogenation of benzene and distortion would provide more active sites for catalytic reactions the mixed La–Sr–Co–Fe–O perovskite catalyst for the CO oxidation than crystalline materials, which was probably responsible for the Although no direct correlation between the structural high catalytic activity. The amorphous MnOx as electrocatalyst by disorder and catalytic activity over SCR catalysts was clearly Yang and Xu for oxygen reduction reaction is one of these proposed by other researchers, some experimental results from examples Shishido et al. also concluded that the isolated and literature showed that this empirical conclusion might also be tetrahedrally coordinated iron sites with higher degree of structure applicable to the catalysts for the SCR reaction. For example, the distortion in the framework of Fe-MCM-41 were responsible for iron species in Fe/ZSM-5 with more distorted tetrahedrally the high activity in oxidation reaction, while small iron oxide coordinated structure showed higher activity than that with clusters with lower degree of structure distortion were not regular octahedrally coordinated structure in the catalytic reduc- effective Therefore, we can deduce that the severest tion of NO with iso-butane . Moreover, the zeolite encapsulated distortion of Fe–O and Mn–O coordination structure in our vanadium oxo species and the highly isolated vanadium crystallite Fe0.5Mn0.5TiOx catalyst is also an important reason for species in mesoporous V2O5–TiO2–SiO2 catalyst with dis- its highest SCR activity.
torted tetrahedrally coordinated structure also showed high The radial distribution function (R space, phase shift uncor- activity in the SCR of NO with NH3. In this study, both of the rected) of Fe-K-edge and Mn-K-edge derived from the EXAFS data iron species and manganese species in our catalysts contributed to are shown in and D, respectively. For Fe-K-edge, a peak the SCR activity, and there was an inverse correlation between the centered at 1.41 A˚ showed up, which could be attributed to the first structure distortion of Fe–O shell and Mn–O shell during the Mn Fe–O shell. No obvious peak above 2 A˚ belonging to the second substitution process. Summarizing the results in B and D, coordination shell was observed, indicating that all samples were Fe0.5Mn0.5TiOx catalyst had the most appropriate structural in crystallite phase, which is in accordance with the XRD results.
disorder of these two active species, and this is another important For Mn-K-edge, the situation was similar. Only one obvious peak at reason for its highest SCR activity.
1.35 A˚ due to the first Mn–O shell was observed, and the secondcoordination shell was not well crystallized, either. With the 3.3. Redox properties increasing of Mn substitution amounts, the peak intensity of Fe–Oshell showed a monotonic increase, and at the same time the peak intensity of Mn–O shell showed a monotonic decrease. If we The XPS results of Fe 2p are shown in Two characteristic assume that the coordination numbers of Fe–O and Mn–O shells do peaks ascribed to Fe 2p3/2 at 711.4 eV and Fe 2p1/2 at 724.9 eV not change during the Mn substitution process, the peak intensity appeared for each Fe-containing sample, indicating that the iron F. Liu et al. / Applied Catalysis B: Environmental 93 (2009) 194–204 Fig. 8. XPS results of (A) Fe 2p, (B) Mn 2p, (C) Ti 2p and (D) O 1s in FeaMn1aTiOx catalysts.
species in these samples was in Fe3+ oxidation state It was unselective oxidation of NH3 would also get enhanced with the reported in our previous study that the iron species in iron titanate increasing of Mn4+/Mn3+ ratio resulting in low N2 selectivity in the catalyst possessed higher binding energies than that in pristine SCR reaction, which was caused by the higher degree of hydrogen Fe2O3 due to the strong interaction between iron and titanium abstraction from ammonia by manganese species with higher species and the substitution of partial Fe by Mn did not change oxidation state This is in accordance with the NH3 oxidation this situation. The iron species with enhanced oxidative ability results in and Fig. S4.
than that in Fe2O3 was still responsible for the high SCR activity.
shows the XPS results of Ti 2p in all catalysts. For each With the increasing of Mn substitution amounts, the intensities of sample, two characteristic peaks attributed to Ti 2p3/2 at 458.6 eV Fe 2p3/2 and Fe 2p1/2 peaks gradually decreased owing to the and Ti 2p1/2 at 464.4 eV showed up, indicating the presence of Ti4+ concentration reduction of surface iron species. However, the . As the XPS results shown in our previous study the corresponding binding energies did not show variation, implying binding energies of Ti 2p3/2 and Ti 2p1/2 in FeTiOx were smaller than that the differences of SCR, NO oxidation and NH3 oxidation those in pristine TiO2 due to the strong interaction between iron activities over these catalysts were not caused by the redox ability and titanium species. This phenomenon was probably caused by change of iron species.
the deviation of electronic cloud from Fe3+ to Ti4+, because Ti4+ The XPS results of Mn 2p in Mn-containing samples are shown shows stronger affinity of electrons comparing with that of Fe3+.
in For Fe0.75Mn0.25TiOx catalyst with low Mn substitution Similar phenomenon was also observed on other iron–titanium amount, the binding energies of Mn 2p3/2 and Mn 2p1/2 peaks were oxide composites prepared by other researchers . In this located at 641.6 and 653.3 eV, respectively, which indicated that study, the introduction of manganese species did not influence the the majority of manganese species in this sample was in Mn3+ redox behavior of titanium species, because the Ti 2p peak oxidation state, similar as that in Mn2O3 . Besides, a small positions showed no obvious change with the increasing of Mn fraction of manganese species in this sample was in Mn4+ oxidation state, thus the overall manganese species in Fe0.75Mn0.25TiOx Summarizing the XPS results of Fe 2p, Mn 2p and Ti 2p, we can showed a little higher binding energies than those in Mn2O3 which conclude that the enhanced oxidative ability of FeaMn1aTiOx was reported in literature (Mn 2p3/2 at 641.2  0.2 eV) With catalysts was mainly caused by the introduction of Mn, which the increasing of Mn substitution amounts, the intensities of Mn 2p3/2 showed higher oxidation state when the substitution amount was and Mn 2p1/2 peaks gradually enhanced due to the concentration larger. We can infer that the adsorption of NOx over these catalysts increase of surface manganese species. At the same time, the will get enhanced due to the higher surface concentration and corresponding binding energies also showed variation, with Mn stronger oxidative ability of manganese species. This point of view 2p3/2 shifting from 641.6 to 641.9 eV and Mn 2p1/2 shifting from 653.3 will be verified in the following experimental sections concerning to 653.6 eV. This result showed that the Mn4+/Mn3+ ratio in NOx adsorption abilities.
FeaMn1aTiOx serial catalysts became larger when the Mn substitu- As the XPS results shown in the O 1s peak was fitted into tion amount was higher. With the increasing of Mn4+/Mn3+ ratio, the two peaks by searching for the optimum combination of Gaussian oxidation of NO to NO2 would get enhanced, which was beneficial to bands with the correlation coefficients (r2) above 0.99. The peak at promote the low temperature SCR activity This is in accordance 530.2 eV corresponds to the lattice oxygen O2 (denoted as Ob), with the NO oxidation results in At the same time, the and the one at 531.6 eV corresponds to the surface adsorbed F. Liu et al. / Applied Catalysis B: Environmental 93 (2009) 194–204 oxygen (denoted as O or O belonging to defect- 554 or 608 8C). Previous studies showed that the reduction oxide or hydroxyl-like group . The surface chemisorbed of pristine MnOx or MnOx/TiO2 samples followed a two step oxygen Oa was reported to be highly active in oxidation reaction MnO2 ! Mn2O3 ! MnO due to its higher mobility than lattice oxygen Ob and the MnO2 ! Mn2O3 ! Mn3O4 ! MnO, during which the area ratio high relative concentration ratio of Oa/(Oa + Ob) on catalyst of H2 consumption peaks should be 1:1 or 3:1:2. However, the surface could be correlated with high SCR activity After the Fe reduction process of our MnTiOx sample was not similar as either was substituted by Mn, the Oa/(Oa + Ob) ratio had an obvious of these. The area ratio of T1/(T2 + T3) was calculated to be nearly increase, especially at high Mn substitution amounts. This implies 2:1, which implied that the T1 reduction peak was due to Mn4+–Ox– that comparing with FeTiOx, there are more oxide defects or Ti ! Mn3+/2+–Ox–Ti. Herein, the manganese species in Mn3+/2+– hydroxyl-like groups in Mn-containing catalysts. On one hand, the Ox–Ti intermediate had similar oxidation state as that in Mn3O4.
oxide defects can adsorb and activate gaseous O2 to form active Both of the T2 and T3 reduction peaks were ascribed to Mn3+/2+–Ox– oxygen species, which is beneficial to promote the NO oxidation to Ti ! Mn2+–Ox–Ti, because further reduction to metallic Mn0 does NO2 and thus the ‘‘fast SCR'' process. On the other hand, the NH3 not proceed until over 1200 8C . Over the Mn-MCM-41 sample adsorption in the form of NH + can also be enhanced due to the prepared by Reddy et al., the Mn4+ species was also only reduced to production of larger amount of surface hydroxyl groups, which act Mn2+ by H2 at around 600 8C . Therefore, in our MnTiOx sample as Brønsted acid sites. The formed NH + can react with adsorbed after the T2 reduction peak the majority of manganese species was NO2 to produce active intermediate species, and then further react in Mn2+ state, resulting in the formation of analogous pyrophanite with gaseous NO to produce N2 and H2O The (Mn2+TiO3) compound in the presence of Ti4+ species. This enhancement of NOx and NH3 adsorption over Mn substituted pyrophanite compound would possibly be sintered to form catalysts will be discussed later.
compact oxide layer on the catalyst surface and make it difficultfor the gaseous H2 to diffuse into the inner bulk phase, thus leading to the delayed appearance of the small T3 reduction peak. The total The above XPS results of O 1s could supply some information reduction process of MnTiOx is similar as the results over MnOx or about the surface chemisorbed oxygen Oa over these catalysts, and MnOx/Al2O3 in previous studies . Moreover, it is the H2-TPR results could supply some information about the total noteworthy that the T2 peak in MnTiOx delayed ca. 60 8C than reducible oxygen including Oa and partial Ob. All H2 consumption that over Fe0.2Mn0.8TiOx and ca. 100 8C than that over Fe0.5Mn0.5- peaks shown in could be attributed to the reduction of iron TiOx, which was mainly due to the absence of iron species. During and manganese species, because the pristine TiO2 sample showed the reduction process of iron and manganese containing catalysts, no reduction peaks during the whole temperature range that we a small fraction of iron species was firstly reduced to metallic Fe0 investigated .
nanoparticles, which could dissociate H2 into H atoms; in the In our previous study, we have concluded that the reduction of presence of the in situ formed water vapor, the dissociated H atoms iron species in FeTiOx followed a two step process: Fe3+–O– could be transferred effectively to further reduce the manganese Ti ! Fe2+/3+–O–Ti ! Fe2+–O–Ti with a Tmax peak locating at oxides through the so-called H-spillover effect . This H- 414 8C. After Mn substitution, the reduction process of iron species spillover effect could significantly lower the reduction tempera- did not change, and the whole TPR profiles of FeaMn1aTiOx ture of T2 peak, which was a new feature introduced by the catalysts were composed of Fe and Mn reduction peaks. With the coexistence of iron and manganese species.
increasing of Mn substitution amounts, a low temperaturereduction peak (T1) between 350 and 400 8C showed up, which 3.4. NH3 and NOx adsorption abilities must be caused by the reduction of manganese species. This meansthat the oxygen mobility was greatly enhanced due to the 3.4.1. NH3-TPD and NOx-TPD introduction of Mn, which was beneficial to the SCR reaction.
shows the NH3-TPD results over FeaMn1aTiOx Fe0.5Mn0.5TiOx exhibited the lowest temperature of T1 peak at catalysts using the fragments of m/z = 16 (NH2) and m/z = 15 352 8C and this in harmony with its highest SCR activity. For (NH) to identify NH3 due to the disturbance of m/z = 17 by H2O. In Fe0.2Mn0.8TiOx and MnTiOx samples with higher Mn substitution the temperature range from 30 to 500 8C, all of the samples showed amounts, another two well defined reduction peaks locating at three NH3 desorption peaks. Comparing with the in situ DRIFTS relative high temperatures also emerged (T2 at 480 or 544 8C, T3 at results of NH3-TPD in Fig. S5, we can assign these three peaks as Fig. 9. H2-TPR profiles of FeaMn1aTiOx catalysts.
Fig. 10. TPD profiles of (A) NH3 and (B) NOx over FeaMn1aTiOx catalysts.
F. Liu et al. / Applied Catalysis B: Environmental 93 (2009) 194–204 follows: the small sharp peaks below 100 8C were caused by thedesorption of physisorbed NH3; the medium-sized peaks between120 and 140 8C were caused by the desorption of NH + bound to surface hydroxyls; and the broadest peaks centered at200–230 8C were caused by the desorption of coordinated NH3bound to Lewis acid sites and residual NH + strongly bound to surface hydroxyls with enhanced acidity by sulfate species (fromTi(SO4)2 precursor). Besides all the desorption peaks slightlymoved to the low temperature edge, it seemed that the Mnsubstitution of Fe did not obviously influence the NH3 adsorptionability of these catalysts, especially the adsorption amount. Thisresult implied in an opposite way that NH3 also mainly adsorbedon titanium sites over Mn substituted catalysts, similar as thesituation that we described in previous study over FeTiOx catalyst. The detailed relationship between surface adsorbed NH3species and SCR activity will be discussed in the following section.
The NOx-TPD results over FeaMn1aTiOx catalysts are shown in Different from the NH3-TPD results, the NOx desorptionprofiles exhibited obvious change when partial Fe was substituted Fig. 11. In situ DRIFTS results of (A) NH by Mn. Over unsubstituted FeTiO 3 adsorption and (B) NO + O2 adsorption over x catalyst, only one obvious NOx FeaMn1aTiOx catalysts at 30 8C.
desorption band centered at 337 8C showed up. With theincreasing of Mn substitution amounts, there was a largerproportion of NOx desorption from 100 to 300 8C. Comparing with process. This means that the introduction of Mn resulted in more the in situ DRIFTS results of NOx-TPD in Fig. S6, we can have the Brønsted acid sites on the catalyst surface, which was favorable for peak assignments as follows: the peaks below 120 8C could be the promotion of SCR activity. Schwidder et al. also proposed a attributed to physisorbed NOx; the peaks centered between 140 promoting effect of Brønsted acidity on the low temperature SCR and 175 8C were due to the decomposition of monodentate nitrate activity over iron-based catalyst probably via an acid-catalyzed species; and the broad peaks above 175 8C were due to the decomposition of active intermediate. The possible reason for the decomposition of bridging nitrate species and bidentate nitrate Brønsted acidity enhancement in this study might be that more species with higher thermal stability. The dotted lines representing residual sulfate species from Ti(SO4)2 precursor was left on the the fragment of desorbed NO2 (m/z = 46) also confirmed this point catalyst surface due to higher coordination ability of Mn4+ than of view. The introduction of Mn resulted in the enhanced NO that of Fe3+. This point of view can be verified by the intensity oxidation to NO2 and thus the enhanced adsorption of NOx as increase of negative band around 1344 cm1 attributed to the nitrate species at lower temperatures. This means that comparing coverage of residual sulfate species (nas S5O) by adsorbed NH3 with the unsubstituted catalyst, more nitrate species on catalyst with the increasing of Mn substitution amounts. For coordinated surface could participate in the SCR reaction in the temperature NH3 bound to Lewis acid sites (das at 1603 and ds at 1192 cm1), the range that we investigated, which was beneficial to promote the band intensity firstly had an obvious increase, and then showed an SCR activity.
intense decrease when the substitution amount was higher than0.5. The bands at 3365, 3253 and 3160 cm1 ascribed to N–H 3.4.2. In situ DRIFTS of NH3 and NOx adsorption stretching vibration modes of coordinated NH3 also followed The in situ DRIFTS results of NH 3 adsorption at 30 8C are shown similar trend. It was reported that both ionic NH4 and coordinated in A. Over all samples, a weak and broad band centered at NH3 could take part in the SCR process through reaction with NO2 1805 cm1 was observed with similar intensity, which was adsorbed species to form active intermediates . Therefore, the difficult to be assigned due to the lack of literature support.
proper proportion of Brønsted acid sites and Lewis acid sites over Similar weak bands around 1800 cm1 were also found on Fe– Fe0.5Mn0.5TiOx catalyst was responsible for its highest SCR activity.
TiO2–PILC catalyst after NH3 adsorption at room temperature in B presents the in situ DRIFTS results of NOx adsorption at Long and Yang's work however, they did not assign these 30 8C. With the increasing of Mn substitution amounts, the bands bands probably due to the low surface concentration of these attributed to bridging nitrate species (n3 high at 1612 or species. Considering the in situ DRIFTS results of NH3-TPD in Fig. S5 1618 cm1 and n3 low at 1244 cm1) and bidentate nitrate species and NH3-TPD results in A, these NH3 species showed rather (n3 high at 1583 cm1) showed no obvious change, except that low thermal stability and disappeared at ca. 100 8C, which might be the band at 1244 cm1 in Fe0.2Mn0.8TiOx and MnTiOx was strongly ascribed to physisorbed NH3. Over the catalyst surface with large overlapped by other growing nitrate species. This monotonously surface area and strong acidity, these physisorbed NH3 molecules growing species was ascribed to monodentate nitrate (n3 high at might form ammonia clusters ([NH3]n) through the effect of 1543 or 1518 cm1 and n3 low at 1286 or 1296 cm1) which hydrogen bonding in which N acted as electron couple donator and was thought to be the real reactive species in the SCR condition H from another NH3 molecule nearby acted as electron couple The higher Mn substitution amounts resulted in the receptor, thus exhibiting higher vibration frequency than that of enhancement of NO oxidation to NO2, and thus the production gas phase NH3. With the increasing of Mn substitution amounts in of more monodentate nitrate species M–O–NO2 (M = Fe and Mn), aMn1aTiOx serial catalysts, the bands attributed to NH4 which was similar as the NO2 adsorbed species in the previous 1676 cm1 and das at 1458 cm1) showed an obvious study by Long and Yang In the SCR reaction condition, this M– increase in intensity. The bands at 3020 and 2806 cm1 attributed 2 species could rapidly react with adjacent adsorbed NH4 to N–H stretching vibration modes of NH + NH3 to produce more reactive intermediates M–O–NO2[NH4 ]2 or progressive increase in intensity. At the same time, the intensities M–O–NO2[NH3]2, which could further react with gaseous NO to of the negative bands at 3732 and 3676 cm1 ascribed to O–H form N2 and H2O . This reaction mechanism is very similar as stretching vibration modes due to the interaction of surface the one proposed by other researchers, in which the reduction of hydroxyls with NH3 also became larger during this Mn substitution ammonium nitrate by NO is an important step . The in situ F. Liu et al. / Applied Catalysis B: Environmental 93 (2009) 194–204 Research and Development Program of China (2006AA06A304,2009AA06Z301).
Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at .
[1] H. Bosch, F. Janssen, Catal. Today 2 (1988) 369.
[2] G. Busca, L. Lietti, G. Ramis, F. Berti, Appl. Catal. B: Environ. 18 (1998) 1.
[3] J.P. Dunn, P.R. Koppula, H.G. Stenger, I.E. Wachs, Appl. Catal. B: Environ. 19 (1998) [4] P. Balle, B. Geiger, S. Kureti, Appl. Catal. B: Environ. 85 (2009) 109.
[5] F. Liu, H. He, C. Zhang, Chem. Commun. (2008) 2043.
[6] F. Liu, H. He, C. Zhang, Z. Feng, L. Zheng, Y. Xie, T. Hu, Appl. Catal. B: Environ., submitted for publication.
Fig. 12. In situ DRIFTS results of SCR reaction over Fe [7] F. Kapteijn, L. Singoredjo, A. Andreini, J.A. Moulijn, Appl. Catal. B: Environ. 3 (1994) 0.75Mn0.25TiOx catalyst at 200 8C, during which NO + O2 was let in firstly and NH3 was let in subsequently.
[8] X. Tang, J. Hao, W. Xu, J. Li, Catal. Commun. 8 (2007) 329.
[9] P.R. Ettireddy, N. Ettireddy, S. Mamedov, P. Boolchand, P.G. Smirniotis, Appl. Catal.
B: Environ. 76 (2007) 123.
DRIFTS result of SCR reaction over Fe [10] Z. Wu, B. Jiang, Y. Liu, Appl. Catal. B: Environ. 79 (2008) 347.
0.75Mn0.25TiOx in also [11] P.G. Smirniotis, P.M. Sreekanth, D.A. Pen˜a, R.G. Jenkins, Ind. Eng. Chem. Res. 45 showed that the monodentate nitrate species (1552 cm1) could (2006) 6436.
not be detected on the catalyst surface due to its high reactivity.
[12] T. Grzybek, J. Pasel, H. Papp, Phys. Chem. Chem. Phys. 1 (1999) 341.
Under the SCR reaction condition at 200 8C, only NH [13] G. Qi, R.T. Yang, R. Chang, Appl. Catal. B: Environ. 51 (2004) 93.
species (ionic NH + [14] F. Eigenmann, M. Maciejewski, A. Baiker, Appl. Catal. B: Environ. 62 (2006) 311.
4 at 1676 cm1 and coordinated NH3 at 1603 and [15] M. Kang, E.D. Park, J.M. Kim, J.E. Yie, Catal. Today 111 (2006) 236.
1190 cm1) and inert bidentate nitrate species (1576 cm1) [16] G. Qi, R.T. Yang, Appl. Catal. B: Environ. 44 (2003) 217.
existed on the surface stably, implying the rapid consumption of [17] S. Roy, B. Viswanath, M.S. Hegde, G. Madras, J. Phys. Chem. C 112 (2008) 6002.
reactive intermediates. Summarizing the in situ DRIFTS results of [18] G. Qi, R.T. Yang, R. Chang, S. Cardoso, R.A. Smith, Appl. Catal. A: Gen. 275 (2004) NH3/NOx adsorption in and the SCR reaction in , on [19] G. Carja, G. Delahay, C. Signorile, B. Coq, Chem. Commun. (2004) 1404.
0.5Mn0.5TiOx catalyst the amount of NH4 [20] M. Casanova, E. Rocchini, A. Trovarelli, K. Schermanz, I. Begsteiger, J. Alloys most abundant, and the formation of reactive monodentate nitrate Compd. 408–412 (2006) 1108.
[21] S. Kieger, G. Delahay, B. Coq, Appl. Catal. B: Environ. 25 (2000) 1.
species was also greatly enhanced. Therefore, it was reasonable to [22] X. Wang, T. Zhang, X. Sun, W. Guan, D. Liang, L. Lin, Appl. Catal. B: Environ. 24 obtain the highest SCR activity over this catalyst.
[23] K.V. Klementev, J. Phys. D: Appl. Phys. 34 (2001) 209.
[24] R.Q. Long, R.T. Yang, J. Catal. 188 (1999) 332.
[25] J.-H. Park, H.J. Park, J.H. Baik, I.S. Nam, C.-H. Shin, J.-H. Lee, B.K. Cho, S.H. Oh, J.
Catal. 240 (2006) 47.
The substitution of partial Fe by Mn could significantly promote [26] F. Liu, H. He, C. Zhang, Appl. Catal. B: Environ., submitted for publication.
[27] G. Madia, M. Koebel, M. Elsener, A. Wokaun, Ind. Eng. Chem. Res. 41 (2002) 3512.
the SCR activity of iron titanate catalyst, especially in the low [28] R.Q. Long, R.T. Yang, J. Catal. 198 (2001) 20.
temperature range. Fe0.5Mn0.5TiOx with the molar ratio of [29] C. Ciardelli, I. Nova, E. Tronconi, D. Chatterjee, B. Bandl-Konrad, M. Weibel, B.
Fe:Mn = 1:1 showed the best activity, over which NO Krutzsch, Appl. Catal. B: Environ. 70 (2007) 80.
eliminated at 175 [30] M. Devadas, O. Kro¨cher, M. Elsener, A. Wokaun, N. So¨ger, M. Pfeifer, Y. Demel, L.
8C at GHSV = 50 000 h1. However, the N2 Mussmann, Appl. Catal. B: Environ. 67 (2006) 187.
selectivity showed an obvious decrease with the increasing of Mn [31] M. Koebel, G. Madia, M. Elsener, Catal. Today 73 (2002) 239.
substitution amounts, and there should be a compromise between [32] A. Grossale, I. Nova, E. Tronconi, D. Chatterjee, M. Weibel, J. Catal. 256 (2008) 312.
the SCR activity and N [33] M. Schwidder, S. Heikens, A. De Toni, S. Geisler, M. Berndt, A. Bru¨ckner, W.
2 selectivity when we determine on the Mn Gru¨nert, J. Catal. 259 (2008) 96.
substitution amount in practical industrial application.
[34] Y. Wang, Q. Zhang, T. Shishido, K. Takehira, J. Catal. 209 (2002) 186.
The active phases in Mn substituted catalysts were still in [35] P.-E. Petit, F. Farges, M. Wilke, V.A. Sole´, J. Synchrotron Radiat. 8 (2001) 952.
crystallite states, similar as those in iron titanate catalyst. The [36] T. Kawabata, N. Fujisaki, T. Shishido, K. Nomura, T. Sano, K. Takehira, J. Mol. Catal.
A: Chem. 253 (2006) 279.
strong interaction of iron, manganese and titanium species in [37] E. Chalmin, F. Farges, G.E. Brown Jr., Contrib. Mineral. Petrol. 157 (2009) 111.
Fe0.5Mn0.5TiOx resulted in the largest surface area and porosity, the [38] J. Yang, J.J. Xu, Electrochem. Commun. 5 (2003) 306.
severest structural distortion and most appropriate structural [39] T. Shishido, Q. Zhang, Y. Wang, T. Tanaka, K. Takehira, Phys. Scripta T115 (2005) disorder, the enhanced oxidative ability of manganese species, the [40] A. Szizybalski, F. Girgsdies, A. Rabis, Y. Wang, M. Niederberger, T. Ressler, J. Catal.
highest mobility of lattice oxygen, the proper ratio of Brønsted acid 233 (2005) 297.
sites and Lewis acid sites together with the enhanced NO [41] B. Shen, S. Wei, K. Fang, J.F. Deng, Appl. Phys. A 65 (1997) 295.
[42] L.A. Isupova, V.A. Sadykov, S.V. Tsybulya, G.N. Kryukova, V.P. Ivanov, A.N. Petrov, adsorption capacity, which were all responsible for its highest O.F. Kononchuk, React. Kinet. Catal. Lett. 62 (1997) 129.
SCR activity. Studies concerning the SCR reaction mechanism, [43] M.S. Batista, M. Wallau, E.A. Urquieta-Gonza´lez, Braz. J. Chem. Eng. 22 (2005) 341.
[44] R.C. Adams, L. Xu, K. Moller, T. Bein, W.N. Delgass, Catal. Today 33 (1997) 263.
2O by-product and H2O/SO2 inhibition effect over Mn substituted catalyst are under way.
[45] V.I. Paˆrvulescu, S. Boghosian, V. Paˆrvulescu, S.M. Jung, P. Grange, J. Catal. 217 [46] T. Grzybek, J. Klinik, B. Buczek, Surf. Interface Anal. 23 (1995) 815.
[47] D.A. Pen˜a, B.S. Uphade, P.G. Smirniotis, J. Catal. 221 (2004) 421.
[48] J. Li, J. Chen, R. Ke, C. Luo, J. Hao, Catal. Commun. 8 (2007) 1896.
[49] M. Kang, E.D. Park, J.M. Kim, J.E. Yie, Appl. Catal. A: Gen. 327 (2007) 261.
We sincerely appreciate the help from National Synchrotron [50] S. Yuan, Q. Sheng, J. Zhang, H. Yamashita, D. He, Micropor. Mesopor. Mater. 110 Radiation Laboratory, University of Science and Technology of China for supplying the beam time to carry out XAFS experiments.
[51] N. Perkas, O. Palchik, I. Brukental, I. Nowik, Y. Gofer, Y. Koltypin, A. Gedanken, J.
Phys. Chem. B 107 (2003) 8772.
This work was financially supported by Chinese Academy of [52] A. Glisenti, J. Mol. Catal. A: Chem. 153 (2000) 169.
Sciences (KZCX1-YW-06-04) and the National High Technology [53] Z. Wu, R. Jin, Y. Liu, H. Wang, Catal. Commun. 9 (2008) 2217.
F. Liu et al. / Applied Catalysis B: Environmental 93 (2009) 194–204 [54] H. Chen, A. Sayari, A. Adnot, F. Larachi, Appl. Catal. B: Environ. 32 (2001) 195.
[62] O.J. Wimmers, P. Arnoldy, J.A. Moulijn, J. Phys. Chem. 90 (1986) 1331.
[55] R.Q. Long, R.T. Yang, J. Catal. 190 (2000) 22.
[63] R.P. Viswanath, B. Viswanathan, M.V.C. Sastri, React. Kinet. Catal. Lett. 2 (1975) 51.
[56] Z. Wu, B. Jiang, Y. Liu, H. Wang, R. Jin, Environ. Sci. Technol. 41 (2007) 5812.
[64] W.K. Jozwiak, E. Kaczmarek, T.P. Maniecki, W. Ignaczak, W. Maniukiewicz, Appl.
[57] A. Khan, P.G. Smirniotis, J. Mol. Catal. A: Chem. 280 (2008) 43.
Catal. A: Gen. 326 (2007) 17.
[58] J. Carno¨, M. Ferrandon, E. Bjo¨rnbom, S. Ja¨ra˚s, Appl. Catal. A: Gen. 155 (1997) 265.
[65] R.Q. Long, R.T. Yang, J. Catal. 186 (1999) 254.
[59] E.P. Reddy, B. Sun, P.G. Smirniotis, J. Phys. Chem. B 108 (2004) 17198.
[66] N.Y. Topsøe, Science 265 (1994) 1217.
[60] X. Tang, Y. Li, X. Huang, Y. Xu, H. Zhu, J. Wang, W. Shen, Appl. Catal. B: Environ. 62 [67] G. Ramis, L. Yi, G. Busca, Catal. Today 28 (1996) 373.
[68] M. Schwidder, M.S. Kumar, U. Bentrup, J. Pe´rez-Ramı´rez, A. Bru¨ckner, W. Gru¨nert, [61] M.C. A´lvarez-Galva´n, V.A. de la Pen˜a O'Shea, J.L.G. Fierro, P.L. Arias, Catal. Com- Micropor. Mesopor. Mater. 111 (2008) 124.
mun. 4 (2003) 223.
[69] G.M. Underwood, T.M. Miller, V.H. Grassian, J. Phys. Chem. A 103 (1999) 6184.

Source: http://hehong.rcees.ac.cn/bookpic/200911201481111103.pdf