Dongqi An,Yuyao Yang,Weixin Zou,Yandi Cai,Qing Tong,Jingfang Sun,*,Lin Dong*

1 Key Laboratory of Mesoscopic Chemistry of MOE,School of Chemistry and Chemical Engineering and Jiangsu Key Laboratory of Vehicle Emissions Control,Nanjing University,Nanjing 210093,China

2 Center for Nanochemistry,College of Chemistry and Molecular Engineering,Peking University,Beijing 100871,China

3 Jiangsu Key Laboratory of Vehicle Emissions Control,School of Environment,Nanjing University,Nanjing 210093,China

4 Jiangsu Key Laboratory of Vehicle Emissions Control,Center of Modern Analysis,Nanjing University,Nanjing 210093,China

5 Key Laboratory of Mesoscopic Chemistry of MOE,School of Chemistry and Chemical Engineering and Jiangsu Key Laboratory of Vehicle Emissions Control,School of Environment,Center of Modern Analysis,Nanjing University,Nanjing 210093,China

Keywords:NH3-SCR NbCe catalyst Cu modification NO2 promoting effect Fast-SCR Flue gas

ABSTRACT A simple strategy of Cu modification was proposed to broaden the operation temperature window for NbCe catalyst.The best catalyst Cu0.010/Nb1Ce3 presented over 90%NO conversion in a wide temperature range of 200-400 °C and exhibited an excellent H2O or/and SO2 resistance at 275 °C.To understand the promotional mechanism of Cu modification,the correlation among the ‘‘a(chǎn)ctivity-structure-property”were tried to establish systematically.Cu species highly dispersed on NbCe catalyst to serve as the active component.The strong interaction among Cu,Nb and Ce promoted the emergence of NbO4 and induced more Br?nsted acid sites.And Cu modification obviously enhanced the redox behavior of the NbCe catalyst.Besides,EPR probed the Cu species exited in the form of monomeric and dimeric Cu2+,the isolated Cu2+acted as catalytic active sites to promote the reaction:Cu2+-+NO(g)→Cu2+-+NO2(g).Then the generated NO2 would accelerate the fast-SCR reaction process and thus facilitated the lowtemperature deNOx efficiency.Moreover,surface nitrates became unstable and easy to decompose after Cu modification,thus providing additional adsorption and activation sites for NH3,and ensuring the improvement of catalytic activity at high temperature.Since the NH3-SCR reaction followed by E-R reaction pathway efficaciously over Cu0.010/Nb1Ce3 catalyst,the excellent H2O and SO2 resistance was as expected.

1.Introduction

The abatement of nitrogen oxides (NOx),one of the main contributors of atmospheric pollution,is still a challenging task worldwide [1-4].Among the denitrification techniques developed in recent years,NH3-SCR has been adopted as the most promising and commercial employed procedure for NOxremoval from stationary boilers and coal-fired plants [5,6].And the TiO2-anatase supported V2O5-WO3or V2O5-MoO3oxides have been commercialized for flue gas (300-400 °C) emitted from abovementioned stationary applications.Nevertheless,there are still some shortcomings,such as the biological toxicity of V2O5to human and eco-environment,narrow operation temperature window,and unselective oxidation of NH3to N2O,which constrain their application for nonelectric plant [7-9].Therefore,it is compulsory to develop environmental-friendly NH3-SCR catalysts with exceptional low-temperature NO conversion,high N2selectivity and a broad operating temperature window as alternatives.

Some transition metal oxides,such as MnOx,CuO,Fe2O3and CeO2have been found to be active for the SCR of NO with NH3[10-13].Among them,CeO2has been extensively investigated due to(i)the excellent oxygen storage capacities(OSC)on the basis of the redox behavior between Ce3+and Ce4+;(ii) the relatively easy formation of labile oxygen vacancies and the high mobility of bulk oxygen species [14].Besides,the acid site and redox site are always highly valued in the NH3-SCR process[15],so it’s found that modifying CeO2-based materials with acidic metal oxides(WO3,Nb2O5and MoO3as examples) could dramatically promote their SCR performance [6,16-18].Among them,the Nb2O5-CeO2catalyst system was proven to show great potential for diesel NOxemission control due to adequate activity and high thermal stability[19].However,NbCe catalysts with desirable NOxelimination performance at a broad operation temperature window are still in urgent need,and the poisoning tolerance to water vapor and sulfur oxides ubiquitously found in flue gas also waited to be enhanced [20].The commonly reported Cu were widely used as active species,which exhibited superior low-temperature NH3-SCR activity.As reported in the literature,the redox capacity of CeO2could be promoted drastically by Cu modification [21-23].And numerous studies manifested that the surface area,redox behavior and reactivity of NH3together with absorbed nitrate were notably elevated for the binary Cu and Nb oxides [24-26].

In this work,Cu was adopted as active species to modify NbCe catalyst and the optimal molar ratio of Cu/NbCe was explored.Then the promotional mechanism of Cu modification was comprehensively elucidated through clarifying the relationship among the‘‘a(chǎn)ctivity -structure -property” in depth.

2.Materials and Methods

2.1.Catalyst preparation

NbCe binary oxides were prepared by co-precipitation method.Desired amount of C10H5NbO20and Ce(NO3)3·6H2O was adequately dissolved in deionized water,ammonia solution was then added dropwise to the mixture with continuous stirring until pH=10.The mixture was further stirred for 3 h and then aged for 1 h.Afterwards,the precipitates were collected by filtration for several times and dried at 100 °C overnight and calcined at 500 °C for 4 h under air conditions.The catalysts were crushed and sieved to 0.25-0.38 mm for activity tests.The molar ratios of Nb/Ce were 1/3,1/1 and 3/1.The obtained catalysts were denoted as Nb1Ce3,Nb1Ce1and Nb3Ce1.

Cu/NbCe ternary oxides were prepared by impregnation method.Desired amount of Cu(NO3)2·6H2O and the prepared NbCe powders were adequately dissolved in deionized water.The mixture was sufficiently stirred and heated at 80 °C until water was evaporated and a porous gel was formed.The gel was dried at 100 °C overnight and calcined at 500 °C for 4 h under air conditions.The catalysts were crushed and sieved to 0.25-0.38 mm for activity tests.The obtained catalysts were denoted as Cu0.005/Nb1Ce3,Cu0.010/Nb1Ce3and Cu0.030/Nb1Ce3(molar ratio).

2.2.Characterization

X-ray diffraction (XRD) patterns were recorded on a Philips X’pert Pro diffractometer using Ni-filtered Cu Kα radiation(λ=0.15418 nm) from 10° to 80°.The X-ray tube was operated at 40 kV and 30 mA.

X-ray fluorescence (XRF) measurements were performed on a ARL9800XP+spectrometer to quantify the actual Cu contents of the samples.

N2adsorption-desorption measurements were performed at-196 °C using a Micromeritics ASAP-3020 analyzer.All samples were firstly degassed in vacuum at 300 °C.The specific surface areas of samples were calculatedviaBrunauer-Emmett-Teller(BET) method.

Laser Raman spectra were collected on a LabRAM Aramis(Japan Horiba)Laser Raman spectrometer with an Ar+laser beam with an emission line at 532 nm.The laser power was 10 mW.

X-ray Photoelectron Spectroscopy(XPS)experiments were performed on a PHI 5000 Versa Probe system equipped with a monochromatic Al X-ray radiation source (1486.6 eV) with an accelerating power of 15 kW.Binding energies (BE) of all the elements were calibrated with C1s at 284.6 eV.

H2-temperature-programed reduction (H2-TPR) was performed in a quartz U-tube reactor connected to a thermal conductivity detector (TCD).An H2-Ar (7% H2) mixture was used as reductant.The 10 mg samples were pretreated with purified N2at 200 °C for 1 h.The range of measuring temperature was 20 °C to 800 °C at a rate of 10 °C·min-1.

Electron paramagnetic resonance (EPR) spectra were recorded by a Bruker EMX spectrometer using 100 kHz modulation and 4-G standard modulation width.

Temperature programmed desorption (TPD) measurements were carried out in a fixed-bed quartz reactor equipped with a FTIR spectrometer to monitor outlet gas signal and the interval of spectrum collecting was set at 30 s.For NH3/NO2-TPD experiment:the 100 mg samples were exposed to 0.5‰NH3or 0.5‰NO2(gas mixture of 50 ml·min-1,Ar balanced)until saturation at ambient temperature.Subsequently,purged the samples with the flowing high purified Ar to remove gaseous and weakly absorbed species.Finally,the samples were heated to 500/400 °C at a rate of 10°C·min-1to record the NH3or NO2signal.

in situDRIFTS experiments were conducted on a Nicolet Nexus 5700 FTIR spectrometer.The spectra were collected by an MCT detector and presented as a Kubelka-Munk function.The resolution was 4 cm-1and 32 scans were taken.Before collecting the background spectra,the powder samples were pretreated with high purified N2at 400°C.Background spectra were then collected in a flowing high purified N2stream as the powder samples cooled to each desired temperature.The sample spectra were then obtained after the subtraction of the background spectra at each desired temperature.The samples were exposed to a controlled stream of specific gas mixture for a certain amount of time: 3‰NH3;3‰NO+5%(volume)O2.All of the reaction gasses were balanced with N2at a flow rate of 50 ml·min-1.

2.3.SCR performance measurements

The NH3-SCR activity and N2selectivity on the as-prepared samples were measured in a fixed-bed quartz flow reactor with 200 mg of catalyst.The feed gases were composed of 0.5‰ NO,0.5‰ NH3,5 %(volume) O2,0.2‰ SO2(when used),5 %(volume)H2O (when used) and balanced with Ar.The total flow rate was 200 ml·min-1and the WHSV (weight hourly space velocity) was 60000 ml·g-1·h-1.Before starting the activity test,the catalysts were pretreated in Ar at 200 °C for 1 h.The outlet gas concentrations were measured by an online Thermofisher IS10 FTIR spectrometer.The NO conversion and N2selectivity of the NH3-SCR reaction were determined according to the following equations:

3.Results and Discussion

3.1.Catalytic performance

Fig.1.NO conversion and N2 selectivity of(a-b)NbCe and(c-d)Cu/Nb1Ce3 as a function of the reaction temperature.(The feeding gas contained 0.5‰of NH3,0.5‰of NO and 5 %(volume) O2 balanced with Ar.Weight hourly space velocity (WHSV)=60000 ml·g-1·h-1.).

The NO conversion and N2selectivity over Nb1Ce3,Nb1Ce1and Nb3Ce1were evaluated in broad temperature of 100 °C to 400 °C,and the corresponding results were shown in Fig.1(a-b).All the NbCe catalysts had the favorable N2selectivity and the Nb1Ce3catalyst displayed the best NO removal efficiency in the range of 100°C to 400 °C,so it was adopted as the support in the following study.Then the catalytic performance over Cu0.005/Nb1Ce3,Cu0.010/Nb1Ce3and Cu0.030/Nb1Ce3were reported in Fig.1(c-d).It could be clearly seen that the Cu modification efficaciously enhanced the denitration efficiency and N2selectivity in a wide temperature.As the Cu adding amount increased,the NH3-SCR activity on Cu/Nb1Ce3catalysts initially increased and then decreased.Cu0.010/Nb1Ce3performed the supreme NH3-SCR activity in the entire temperature range,over which exceeding 90% NO conversion was accomplished from 200 °C to 400 °C.It was concluded that the proper amount of Cu modification significantly boosted the DeNOxefficiency of Nb1Ce3catalyst.

In the practical NH3-SCR reaction,SO2and H2O are considered as main deactivating species hence the excellent SO2and H2O durability of catalyst is highly required [27].As illustrated in Fig.2,Cu0.010/Nb1Ce3and Nb1Ce3were implemented to estimate the SO2and/or H2O resistance.The SO2and H2O durability test was conducted at 275 °C and the reaction was operated for 5 h before exposing to H2O.The addition of 5 %(volume) H2O had a negligible inhibition effect on the SCR performance and the NO conversion restored to initial state once the H2O supply was turned off,which indicated that the deactivation of H2O was reversible.When 0.2‰ SO2was introduced into the feed for 7 h,the activity of Cu0.010/Nb1Ce3only declined from 100%to 95%in first one hour and then stabilized.However,the activity of Nb1Ce3continued to drop from 85% to 75% until the SO2was cut off.Then both SO2and H2O were introduced into the feed to examine the combined effect and the deNOxefficiency of Nb1Ce3exhibited sharply impaired.Its activity dropped from 80% to 45% within one hour,and then slowly recovered to 75%in the next 11 h.The corresponding NO conversion of Cu0.010/Nb1Ce3decreased to 90% after 12 h,and restored to almost 95% when the SO2and H2O supply was turned off,suggesting the deactivated Cu0.010/Nb1Ce3catalyst was easy to regenerate.Thus,it could be concluded that the Cu0.010/Nb1Ce3catalyst still displayed outstanding NH3-SCR of NO performance even in the presence of SO2and H2O.

Fig.2.Tolerance tests over Nb1Ce3 and Cu0.010/Nb1Ce3 as a function of the reaction time.(The feeding gas contained 0.5‰ of NH3,0.5‰ of NO,5 %(volume) O2,5 %(volume)H2O(when used)and 0.2‰of SO2(when used)balanced with Ar.Weight hourly space velocity (WHSV)=60000 ml·g-1·h-1.).

3.2.Physicochemical characterizations

The XRD patterns of the catalysts were presented in Fig.3.All the catalysts provided typical diffraction peaks for the CeO2cubic phase (JCPDS 34-0394) and no diffraction peaks attributed to Nb species were detected,indicating that Nb species existed as amorphous.After Cu modification,the diffraction peaks of Cu/Nb1Ce3catalysts were broader than Nb1Ce3catalysts (clearer in the enlarged pattern attached to and FWHM obtained by fitting a Gaussian function) and no bands intensity ascribed to CuO were observed.This indicated that Cu ions inserted into the lattice of the Nb1Ce3catalysts or was highly dispersed on the surface of Cu/Nb1Ce3catalysts [28].

The actual molar ratio of Cu to Nb1Ce3in the as prepared Cu/Nb1Ce3samples were detected by XRF analysis and the results were listed in Table 1.And they were basically consistent with the theoretical value,which proved that our samples were successfully prepared.

The BET surface area,pore volume,and average pore size derived from N2physisorption results were shown in Table 1.It was noticeably that theSBETof Cu/Nb1Ce3catalysts gradually decreased with the increase of Cu addition yet its catalytic activities significantly preceded that of the Nb1Ce3catalyst.This indicated that the SCR performance did not directly depend on the specific surface area in such catalyst system.In addition,the pore volume of all samples decreased from 0.10 to 0.08 cm-3·g-1while the pore size changed conversely with Cu modification.This manifested that the Cu species entered the pores of Nb1Ce3and caused the plugging to a certain extent.

Raman spectra were also collected to further explain the catalyst structure and depicted in Fig.4.An intensive band centered at 463 cm-1was observed on Nb1Ce3,which was attributed to the F2gmode characteristic of the fluorite structure of ceria [29].As for Cu/Nb1Ce3,the CuO bands were completely absent,illuminating the amorphous state of CuO species as the XRD patterns suggested.The band at 446 cm-1ascribed to F2gshowed an obvious shift toward lower wavenumber,signifying a decrease in the Ce-O bond order.This illustrated that the Cu-O-Ce solid solution may occur due to the Cu modification.In addition,two bands detected at 314 and 800 cm-1were ascribed to the Nb2O5phase[30],indicative of the existence of the Nb-O symmetric modes of the NbO4tetrahedral structure.NbO4was deemed as a highly active site,which was conducive to the adsorption and activation of NH3[19].It should be noted that the intensity of NbO4tetrahedral coordination was the most obvious in Cu0.010/Nb1Ce3compared with other samples and this further corroborated the above activity test results.And the weakening of Ce-O bond and the appearance of NbO4indicated that there was a strong interaction among Cu,Nb and Ce species.

Fig.3.XRD patterns of Nb1Ce3 and Cu/Nb1Ce3.

3.3.Surface elements properties

To better understand the surface chemical states of the catalysts,XPS measurements were conducted and the results were displayed in Fig.5.Ce 3d,Nb 3d and O 1s were analyzed and the main surface metal content was summarized in Table 2.

Table 1 Physicochemical properties of Nb1Ce3 and Cu/Nb1Ce3

Table 2 Surface atomic ratio of Nb1Ce3 and Cu/Nb1Ce3

Due to the low Cu content,its peaks in XPS were greatly affected by the signal-to-noise ratio,so the atomic concentration was analyzed.As the Cu species addition increased,the concentration of Cu atoms on the surface of the catalyst gradually increased as well.However,it changed slowly(2.20%vs.3.28%)when the Cu loading exceeded 1% (mass),which might be attributed to the occurrence of agglomerated Cu species.The agglomerated Cu species would lead to the oxidation of NH3[31],which explained the reason why the catalytic activity of the Cu0.030/Nb1Ce3catalyst degraded at high temperature.

The spectrum of Ce 3d was deconvoluted into ten peaks and the corresponding results were given in Fig.5(a).The deconvoluted peaks labeled as ‘‘u” symbolized the Ce 3d3/2spin-orbit components,and those donated as‘‘v”symbolized the Ce 3d5/2spin-orbit components.And the peak labeled as v0,v′,u0,and u′were assigned to the Ce3+species,peaks donated as v,v′′,v′′′,u,u′′,and u′′′belonged to the Ce4+species [26].Cu0.010/Nb1Ce3acquired the highest Ce3+/(Ce4++Ce3+) ratio compared with Nb1Ce3and other Cu modified catalysts.A higher concentration of surface Ce3+might be beneficial for the enhanced NH3-SCR performance.It demon-strated that the addition of Cu effectively transformed the Ce4+to Ce3+on the surface of Cu0.010/Nb1Ce3through the interaction between Cu and Ce species [22].

Fig.4.Raman spectra of Nb1Ce3 and Cu/Nb1Ce3.

Fig.5.XPS spectra of (a) Ce 3d,(b) Nb 3d and (c) O 1s for Nb1Ce3 and Cu/Nb1Ce3.

Nb 3d spectra were presented in Fig.5(b) and the binding energy of Nb 3d5/2located at 206.8 and 205.6 eV could be ascribed to the Nb5+and Nb4+species,indicating that Nb species on Nb1Ce3and Cu0.010/Nb1Ce3were mainly in the highest valence state [32].Nb5+was easier to enter the tetrahedral vacancy of oxygen due to its smaller radius than Nb4+and thus the active NbO4species appeared.The proper content of Cu modification facilitated Cu0.010/Nb1Ce3to acquire the most Nb5+/(Nb5++Nb4+)ratio through the interaction between Cu and Nb.

As depicted in Fig.5(c),the XPS spectra of O 1s were fitted into two peaks.The O 1s binding energy located at 530.4-531.1 eV was attributed to the chemisorbed oxygen species including O-and O2-(donated as Oα) and the binding energy at 529.4.4-529.9 eV was ascribed to the lattice oxygen species (denoted as Oβ) [33].Cu0.010/Nb1Ce3acquired the highest Oα/(Oα+Oβ) ratio,demonstrating that the Cu0.010/Nb1Ce3catalyst possessed a large amount of the surface chemisorbed oxygen species,which was favor of NO oxidation to NO2[24].Thus,the abundant amount of chemisorbed oxygen species on the surface of Cu0.010/Nb1Ce3could contribute to the fast-SCR reaction and consequently much better activity at low temperature.

A deeper understanding towards the existence state of Cu species was obtained by EPR,shown in Fig.6.Only Cu2+species with d9electronic structure gave EPR signal,whereas other Cu valence were EPR silent[34].The signal A was a typical peak of monomeric Cu2+in octahedral sites with a tetragonal distortion[35].The signal K at 3000 G was assigned to dimeric Cu2+generated from the coupled unpaired electrons of two Cu2+[36].It’s quite clear that both the Cu0.010/Nb1Ce3and Cu0.030/Nb1Ce3had much more monomeric Cu2+species.Besides,the distinct K signal of Cu0.030/Nb1Ce3provided evidence for the excessive Cu addition induced agglomeration of Cu species.

To further confirm the surface state and content of Cu species,in situDRIFTS was performed for NO chemisorption.The interaction of NO with Cu2+and Cu+would give specific bands,so it’s a reasonable semi-quantitative means to determine the Cu2+/(Cu2++-Cu+) ratio.Cu/Nb1Ce3catalysts were exposed to flowing NO/N2until saturated at room temperature and the result was shown in Fig.7.The band detected atca.1905 cm-1was ascribed to the Cu2+-NO andca.1847 cm-1was ascribed to the Cu+-NO[37].Later,the peak area was integrated respectively and the Cu2+/(Cu2++Cu+)ratio was further calculated as placed in Table 3.Obviously,the Cu0.010/Nb1Ce3catalyst had the highest Cu2+/(Cu2++Cu+) ratio.

Table 3 Surface Cu2+/(Cu2++Cu+) ratio of Cu/Nb1Ce3 derived from NO-in situ DRIFTS (%)

Fig.6.EPR spectra of Nb1Ce3 and Cu/Nb1Ce3.

Fig.7. in situ DRIFTS spectra of Cu/Nb1Ce3 saturated in NO at room temperature.

3.4.Clarification of Cu promoting mechanism

H2-TPR was performed to reveal the promotional effect of Cu modification on the redox properties of Nb1Ce3and the results were reported in Fig.8.A broad reduction peak occurred on the Nb1Ce3surface from 400 °C to 750 °C,which could be associated with the reduction of the surface capping oxygen [20,38].For Cu/Nb1Ce3,a new reduction peak centered at about 200-300 °C occurred,representing that the Cu2+was reduced to Cu+and then Cu+was converted to metallic Cu [39],suggesting the significant enhancement of redox ability in low-temperature range.It’s observed that the reduction peak of Cu species shifted to lower temperature with the Cu addition increased,proving that the gradually enhanced redox properties of Cu/Nb1Ce3.Besides,the peak of surface capping oxygen moved to the lower temperature due to the interaction among Cu,Nb and Ce species.Nevertheless,excessive redox capacity would lead to the occurrence of NH3-SCR side reactions at high temperature,resulting in diminishment for NO removal [31].Therefore,it could be inferred that the strong redox property generated from agglomerated Cu species on Cu0.030/Nb1-Ce3made the weakened denitration efficiency at high temperature[40].However,an appropriate redox property generated from highly dispersed isolated Cu species enabled Cu0.010/Nb1Ce3to acquire a promoted SCR performance over a wide temperature.

Fig.8.H2-TPR profiles of Nb1Ce3 and Cu/Nb1Ce3.

Cu modification efficaciously enhanced the denitration efficiency of Nb1Ce3and the Cu0.010/Nb1Ce3catalyst with the most Cu2+exhibited the best activity.Hence,the effect of Cu2+on the denitrification reaction was worthy of further investigation.Recently studies showed that for Cu loaded zeolite catalysts,Cu2+would induce the Cu2+-+NO(g) →Cu2+-+NO2(g) reaction,then the generated NO2promoted fast-SCR reaction and improved low-temperature NH3-SCR reaction [37].Herein,an experiment of NO reduction of surface nitrates was conducted to explore the promotional effect of Cu2+.NO+O2/N2were pre-adsorbed on Cu/Nb1-Ce3catalysts until saturated at 150°C so that nitrates were formed on the surface.Then the catalyst was exposed to the NO after purged in inert atmosphere and the NO2formation was recorded in Fig.9.It could be clearly seen that the NO2formation increased after Cu modification and the promoting effect was volcanic with the increase of Cu2+content.Cu0.010/Nb1Ce3catalyst with the most Cu2+activated and reduced more surface nitrates to generate the most NO2.This helped to reveal the promotional mechanism of Cu2+on the low-temperature deNOxperformance of Cu0.010/Nb1Ce3catalyst.

It’s known that a key step in the catalytic reaction was the desorption process of the product,so the NO2adsorption capacity was further evaluated by NO2-TPD.As shown in Fig.10,after Cu modification,the desorption temperature of ad-[41] over Cu0.010/Nb1Ce3catalyst decreased significantly,from 242 °C to 155 °C.The easier the NO2desorbed,the faster the SCR reaction carried out.Besides,the desorption peak at high temperature was related to the strongly adsorbed nitrates[42],their easy desorption would also be beneficial to more sites available for NH3adsorption.And the NO elimination at high temperature was further facilitated.

Fig.9.NO2 formation of Nb1Ce3 and Cu/Nb1Ce3.

Fig.10.NO2-TPD profiles of Nb1Ce3 and Cu0.010/Nb1Ce3.

Fig.11.NH3-TPD profiles of Nb1Ce3 and Cu/Nb1Ce3.

3.5.Adsorption properties

Surface acidity was another decisive factor for the NH3-SCR reaction.Therefore,NH3-TPD was performed to determine the amount of acid sites and acid strength.As shown in Fig.11,all samples showed a broad desorption peak in the range of 50-450°C and the profiles changed little after Cu modification.The integrated peak areas manifested that the Cu0.010/Nb1Ce3catalyst showed the highest NH3adsorption capacity among all Cu modified NbCe catalysts.And the decrease in total acid sites probably caused by the coverage of agglomerated Cu species.

To further distinguish the specific adsorption sites of NH3,Fig.12 illustrated the results ofin situDRIFTS spectra of NH3-TPD over (a) Nb1Ce3and (b) Cu0.010/Nb1Ce3.For Nb1Ce3catalyst,the bands centered atca.1604,1568,1297 and 1190 cm-1could be ascribed to bending vibrations of N-H bonds in NH3coordinately linked to Lewis acid sites [43,44],while the weak band atca.1612 and 1458 cm-1was assigned to bending vibrations of ionicon the Br?nsted acid sites [6,38].For Cu0.010/Nb1Ce3catalyst,bands ascribed to Br?nsted acid sites were distinctly enhanced on account of the Cu modification.This should be caused by the appearance of NbO4species due to the strong interaction between Nb and Cu species as Raman spectra indicated [26].

The NO+O2-TPD over (a) Nb1Ce3and (b) Cu0.010/Nb1Ce3were also carried out throughin situDRIFTS.As depicted in Fig.13,the bands of bridging bidentate nitrates (1625,1607 and 1208 cm-1),chelating bidentate nitrates (1580,1540 and 1238 cm-1),monodentate nitrates (1548 cm-1) andcis-(N2O2)2-(1026 and 1000 cm-1) were observed [38,44].It could be found that after Cu modification,the ad-nitrate species were almost unchanged,but the thermal stability of the nitrates on Cu0.010/Nb1Ce3catalyst was remarkably reduced.This manifested that the nitrates adsorbed on Cu0.010/Nb1Ce3was more easily activated to participate in the following reaction,and thus the adsorption and activation sites would be more accessible for NH3at higher temperature.

3.6.Reaction mechanism

To determine the reactivity of NH3on Cu0.010/Nb1Ce3,in situDRIFTS was conducted at 150 °C with a flowing stream of NO+O2reacting with preadsorbed NH3(Fig.14(a)).Upon introducing NO+O2to catalyst preadsorbed with NH3,the bands assigned to NH3on Lewis and Br?nsted acid sites consumed rapidly,suggesting that the SCR reaction on Cu0.010/Nb1Ce3could proceed through the Eley-Rideal (E-R) mechanism efficiently.In turn,the reactivity of adsorbed nitrates on Cu0.010/Nb1Ce3was also studied with NH3reacting with preadsorbed NO+O2(Fig.14(b)).The bands assigned to nitrate species (bridged nitrate at 1621 cm-1,monodentate nitrates at 1539,1442 and 1290 cm-1)were displayed.Afterwards,NH3were introduced,the intensity of bridge nitrate was weakened and the monodentate nitrate was reconstructed,indicating that the adsorbed nitrates could not react with NH3.Therefore,the SCR reaction followed Eley-Rideal (E-R)mechanism over Cu0.010/Nb1Ce3at 150 °C.

Fig.12. in situ DRIFTS spectra of (a) Nb1Ce3 and (b) Cu0.010/Nb1Ce3 saturated in NH3 at 100 °C and then purged by N2 from 100 °C to 350 °C.

Fig.13. in situ DRIFTS spectra of (a) Nb1Ce3 and (b) Cu0.010/Nb1Ce3 saturated in NO+O2 at 100 °C and then purged by N2 from 100 °C to 350 °C.

Fig.14. in situ DRIFTS spectra of NH3-SCR reaction on Cu0.010/Nb1Ce3 at 150 °C.

4.Conclusions

A series of Cu modified NbCe catalysts were prepared and applied in the NH3-SCR process.The ternary oxides catalysts presented promising NO conversion in a broad reaction temperature window compared with binary oxides NbCe catalyst.Among all the synthesized catalysts,the Cu0.010/Nb1Ce3catalyst showed the optimal NO abatement efficiency from 200 to 400°C,together with near 100% N2selectivity.Besides,H2O or/and SO2had a slightly reversible inhibition influence on the catalytic activity.Several conclusions can be drawn through an array of characterization:The proper amount of Cu modification made the active Cu species highly dispersed in the form of monomeric or dimeric Cu2+on Cu0.010/Nb1Ce3,facilitating the less stable nitrates,the formation/desorption of NO2and the fast-SCR reaction.Meanwhile,the interaction between Cu and Nb species induced the generation of highly active NbO4species and more surface acid sites,therefore the adsorption and activation of NH3were enhanced,which were all responsible for the excellent NH3-SCR activity.Nevertheless,excessive Cu addition would lead to the appearance of agglomerated Cu species,resulting in undesirable NH3oxidation and eventually impaired deNOxefficiency at high temperature.

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgements

Financial support from the National Natural Science Foundation of China,China(Nos.21972062,21976081,21976111)is gratefully acknowledged.