High catalytic performance of CuCe/Ti for CO oxidation and the role of TiO2

Tingting Chang ,Ziyan Wang ,Zhimiao Wang,3 ,Hualiang An,3 ,Fang Li,3,,Wei Xue,3,,Yanji Wang,3,4

1 Hebei Provincial Key Laboratory of Green Chemical Technology and High Efficient Energy Saving, School of Chemical Engineering and Technology, Hebei University of Technology, Tianjin 300130, China

2 Purification Equipment Research Institute of CSIC, Handan 056027, China

3 Tianjin Key Laboratory of Chemical Process Safety, Tianjin 300130, China

4 Hebei Industrial Technology Research Institute of Green Chemical Industry, Huanghua 061100, China

Keywords:CO oxidation TiO2 crystal phase CuCe/Ti Reaction mechanism

ABSTRACT CuCe/Ti-A and CuCe/Ti-R catalysts were prepared using anatase TiO2(TiO2-A)and rutile TiO2(TiO2-R)as supports using the incipient wetness impregnation method for the carbon monoxide(CO)oxidation reaction and were compared with a CuCe-C catalyst prepared using the co-precipitation method.The CuCe/Ti-A catalyst exhibited the highest activity,with complete CO conversion at 90 °C,when the gas hourly space velocity was 24000 ml∙g-1∙h-1 and the CO concentration was approximately 1% (vol).A series of characterizations of the catalysts revealed that the CuCe/Ti-A catalyst has a larger specific surface area,more Cu+species and oxygen vacancies,and the Cu species of CuCe/Ti-A catalyst is more readily reduced.In situ FT-IR results indicate that the bicarbonate species generated on the CuCe/Ti-A catalyst have lower thermal stability than the carbonate species on CuCe/Ti-R,and will decompose more readily to form CO2.Therefore,CuCe/Ti-A has excellent catalytic activity for CO oxidation.

1.Introduction

Carbon monoxide (CO) is a flammable,explosive and toxic air pollutant that seriously pollutes the environment and human health[1].Currently,one of the most cost-efficient and productive techniques for CO removal is catalytic oxidation [2].Noble metal catalysts have excellent activity for CO oxidation,such as Au[3–6],Pt [7,8],Pd [9,10] and Ag [11,12]etc.However,noble metal catalysts are scarce,expensive and have poor thermal stability,inhibiting their applications [13].As a result,the development of non-noble metal catalysts with high activity of CO oxidation is critical.

CuO-CeO2composite oxides have gotten a lot of interest due to their low cost and high catalytic performance for CO oxidation.Sunet al.[14] prepared Cu/Ce(SI) and Cu/Ce(CI) catalysts using surfactant-assisted impregnation and conventional impregnation methods,respectively,with the former exhibiting higher catalytic activity than the latter.For Cu/Ce(SI) catalyst,the CO conversion could reach 99%at 120°C,while the Cu/Ce(CI)catalyst could reach 99% conversion of CO at 140 °C.According to the research,adding hexadecyl trimethyl ammonium bromide can promote copper species dispersion,facilitate insertion of Cu2+into the CeO2lattice,and enhance the interaction of CuO and CeO2,which contribute the good catalytic activity of Cu/Ce(SI).Xueet al.[15] prepared the Cu-benzene-1,3,5-tricarboxylate (Cu-BTC) and Ce-BTC mixture through a similar co-precipitation method,and calcined that mixture to obtain the CuO-CeO2-M catalyst.CuO-CeO2-M achieved complete CO elimination at 80 °C.The CO conversion of CuOCeO2-C catalyst preparedviaco-precipitation method was only 81% at 80 °C.The dispersion of copper species can be enhanced by adding organic ligands during the preparation of CuO-CeO2-M catalyst.The highly dispersed copper species is an important reason for the excellent catalytic performance of CuO-CeO2-M.

For supported catalysts,the support provides a large specific surface area and a suitable pore structure,thus facilitating the dispersion of the active species.Moreover,the interaction with active species and support can also affect catalytic activity of the catalyst.TiO2is frequently employed as a support for catalytic oxidation reactions due to its benefits of nontoxicity,corrosion resistance,large specific surface area,chemical and mechanical stability[16–18].TiO2exists in three crystal phases,anatase,rutile and brookite.Different crystal phases of TiO2have various surface,physical,chemical and structural properties,which affect their performance as catalysts or supports.Diet al.[19] prepared Au/TiO2-C catalyst via deposition–precipitation method for CO oxidation.The Au/TiO2(S1)-C catalyst with anatase TiO2(TiO2-A) as a support exhibited high CO oxidation activity,providing complete oxidation of CO at 60 °C.The Au/TiO2(S5)-C catalyst with rutile TiO2(TiO2-R)as a support showed the lowest CO oxidation activity,providing complete oxidation of CO at 280°C.Kanget al.[20]prepared CuO/TiO2catalysts with different crystal phases by a deposition–precipitation method and evaluated their catalytic performance for CO oxidation.CuO/R with TiO2-R as support showed the highest activity for CO oxidation,achieving complete CO oxidation at 170°C.It can be seen that the crystal phase of the TiO2support has a large influence on the catalytic performance of the supported catalyst.

In this paper,CuCe/Ti catalysts were prepared with TiO2-A and TiO2-R as supports,their catalytic CO oxidation activity was examined and compared with that of the CuCe-C produced using the coprecipitation method,where C refers to co-precipitation.The effect of the TiO2crystal phase on the catalytic performance of CuCe/Ti was investigated,and the CO reaction mechanism was postulated byin situFT-IR characterization analysis.

2.Experimental

2.1.Materials

There was no additional purification necessary for use because all reagents are of analytical quality.Shanghai Macklin Biochemical Co.Ltd.supplied the TiO2-A and TiO2-R.Tianjin Kermel Chemical Reagent Co.Ltd.supplied sodium carbonate (Na2CO3),cupric nitrate trihydrate(Cu(NO3)2∙3H2O)and cerium nitrate hexahydrate(Ce(NO3)3∙6H2O).

2.2.Catalyst preparation

CuCe/Ti catalysts with different TiO2crystal phase as supports were preparedviathe incipient wetness impregnation method.The prepared solution of Ce(NO3)3∙6H2O and Cu(NO3)2∙3H2O withnCe/nCuof 1.6 and(mCu+mCe)/mTiO2of 30%(mass)was added dropwise to the TiO2-A and TiO2-R supports,respectively.The TiO2supports were impregnated 24 h.After impregnation,the CuCe/Ti-A and CuCe/Ti-R catalysts were dried for 90 min at 70°C by rotating vacuum evaporator and calcined for 4 h at 500 °C.

CuCe-C catalyst withnCe/nCuof 1.6 was preparedviaa coprecipitation method.A certain quantity of Ce(NO3)3∙6H2O and Cu(NO3)2∙3H2O were dissolved in distilled water under stirring at 60 °C.10% (mass) Na2CO3solution was dropped into the above solution until the pH reached 7,and the combination was continuously stirred for 1 h.The precipitates were collected by filtered,repeatedly washed in distilled water at 50–60 °C,and left to dry at 70 °C overnight under vacuum.The CuCe-C catalyst was produced by calcining the dried powders at 500 °C for 4 h.

2.3.Catalytic activity test

The catalytic oxidation activity of CO was evaluated in an atmospheric fixed-bed reactor with an inner diameter of 11 mm.The fixed-bed reactor was loaded with 0.5 g catalyst (20–40 mesh,equivalent to 0.42–0.85mm)that was commixed with quartz sand.The feed gas was composed of approximately 1%(vol)CO,9%(vol)O2,balanced by N2,with the total flow rate of 200 ml∙min-1,which corresponds to GHSV of 24000 ml∙g-1∙h-1.Online analysis of the reaction products was performed using a gas chromatograph SP-3420A equipped with a TCD detector.The following equation was used to calculated the CO conversion.

where [CO]inand [CO]outdenote the CO concentration in the input gas and the output gas,respectively.

2.4.Catalyst characterization

A Bruker D8 FOCUS X-ray diffractometer with a monochromator for Cu Kα radiation was used for the X-ray diffraction (XRD)experiments.In the 2θ range between 5° and 90°,the data was scanned at a rate of 1 (°)∙min-1.The Scherrer equation and TOPAS software were used to determine the crystallite size and lattice parameter of CeO2in the samples.

Utilizing a FEI Talos F200S transmission electron microscope,transmission electron microscope(TEM)and EDS-mapping characterization were carried out.

Micromeritics ASAP2020 M+C porosity analyzer was used to measure the textural properties of the catalysts.The Brunauer-Emmett-Teller (BET) method was used to determine the specific surface area of the samples.

A Micromeritics AutoChem II 2920 adsorption instrument was used to perform hydrogen temperature programmed reduction(H2-TPR).0.1 g samples were placed in a U-shaped quartz tube and pretreated with Ar at a flow rate of 50 ml∙min-1at 200 °C for 10 min.The samples were cooled to 50 °C,then exposed to the 10% (vol) H2/Ar and heated to 300 °C at a rate of 10 °C∙min-1.

X-ray photoelectron spectroscopy(XPS)experiments of the catalysts were measured using a Thermo Scientific Escalab 250Xi Xray photoelectron spectrometer.XPS spectra was used to analysis the elemental composition,elemental states and relative content of the catalyst surface.The C 1s peak at 284.8 eV served as the reference binding energy for all other binding energies.

Raman spectra of the samples were carried out on an inVia Reflex Raman spectrometer from Renishaw.The excitation wavelength and the spectral resolution was 532 nm and ≤1 cm.

Bruker A300 EPR spectrometer was used to collect the Electron paramagnetic resonance (EPR) spectra of samples.

On a Thermo NICOLET NEXUS-470 spectrometer with a DTGS detector,in situFT-IR measurements were performed.In situFTIR spectra were operated with a resolution of 4 cm-1for 32 scans during 400–4000 cm-1.A self-supporting disc containing 10 mg catalyst was pressed into shape before being put into the IR cell.To eliminate surface contaminants,all samples underwent a pretreatment with N2at 300 °C for 30 min before being cooled to a specific temperature.The background spectra of CO adsorption were recorded at 30°C.The feed gas(2%(vol)CO/N2)went through the catalyst in the IR cell at the flow rate of 50 ml∙min-1for CO adsorption.The IR spectra were collected every minute until adsorption process reached equilibrium.The background spectra of CO oxidation were recorded at 90 °C.The feed gas (1% (vol)CO,9% (vol) O2balanced by N2) was introduced the IR call at the flow rate of 50 ml∙min-1and the IR spectra were collected every minute.

3.Results and Discussion

3.1.Catalyst characterization

Fig.1. XRD patterns of CuCe-C,CuCe/Ti and TiO2 supports.

The results of the XRD characterization of TiO2supports,CuCe/Ti and CuCe-C catalysts are displayed in Fig.1.The diffraction peaks are at 2θ=25.4° (101),37.9° (004) and 48.1° (200) for TiO2-A,and 2θ=27.5° (110),36.1° (101) and 54.4° (211) for TiO2-R.In the XRD patterns of the CuCe/Ti-A and CuCe-Ti-R catalysts,the diffraction peaks can be observed not only for TiO2but also for the fluorite structure of CeO2(PDF-#43–1002).However,the diffraction peaks of CuO cannot be observed,indicating that CuO particles are highly dispersed on the support.The CuCe-C sample also exhibits the diffraction peaks of CeO2fluorite structure(PDF-#43–1002),and weak diffraction peaks are detected at 35.5°and 38.8°,which are consistent with the XRD diffraction peaks of the(002)and(111)planes of CuO(PDF-#45–0937).The crystallite size and lattice parameter of CeO2in the catalysts are listed in Table 1.Clearly,CuCe/Ti-A catalyst has the smallest crystallite size for CeO2,indicating the high dispersion of CeO2on this catalyst.The lattice parameter of all three CuCe catalysts is smaller than that of pure CeO2(0.5411 nm),suggesting that the Cu2+(0.074 nm) with a smaller ionic radius than Ce4+(0.097 nm) has been incorporated into the CeO2lattice resulting in lattice contraction[21].The incorporation of Cu species into the CeO2lattice will result in the production of more Cu+and oxygen vacancies [22],which will improve the catalytic activity of CO oxidation.

The CuCe-C and CuCe/Ti catalysts were characterized using TEM and EDS mapping.The TEM image of the CuCe-C catalyst is displayed in Fig.2(a).It can be observed that the CuCe-C catalyst is generatedviathe aggregation of nanoparticles with an average size of around 7 nm.Fig.2(b)and(c)show that the TiO2-A support is an irregular lumpy particle with a rough surface and the TiO2-R support is a spherical-like particle with a smooth surface.The particle size of TiO2supports is approximately 50 nm and CuCe nanoparticles supported on the TiO2can be observed.Fig.3 shows the EDS mapping of three CuCe catalysts.The figure clearly shows the uniform distribution of Cu species in the CuCe-C catalyst.After loading on the TiO2support,Cu and Ce still have good dispersion.There is some aggregation of Cu species on the CuCe/Ti-A catalyst.However,according to XRD characterization results,CuO grains did not increase in size.

Fig.4 displays N2adsorption–desorption isotherms and pore size distribution of samples.The capillary coalescence of the meso-porous substance causes type-IV characteristics with a type-H3 hysteresis loop in the N2adsorption–desorption isotherms of the samples [23].It can be seen from the pore size distribution diagram that the pore sizes of the three samples are in the mesopore.The specific surface areas of each sample are listed in Table 1.It can be seen from table that the specific surface areas of both forms of TiO2are greater than that of CuCe-C catalyst,with TiO2-A having a specific surface area that is roughly twice as large as TiO2-R.Compared with that of the support,the specific surface areas of the catalysts obtained by loading CuCe onto TiO2are significantly reduced,but CuCe/Ti-A still has a significantly bigger specific surface area than CuCe/Ti-R.Combined with the results of CeO2crystallite size obtained using XRD,it can be concluded that the dispersion of CuO and CeO2is facilitated by TiO2with a large specific surface area,which enhances the catalytic activity.

Table1 Physicochemical properties of different samples

3.2.Catalytic activity

Fig.5(a) compares the catalytic activities of CuCe-C,CuCe/Ti-A and CuCe/Ti-R for CO oxidation.The results show that both CuCe-C and CuCe/Ti-R catalysts had no activity at 50 °C,whereas CuCe/Ti-A was active for CO oxidation with a CO conversion of 23.5%.The activity of all three catalysts increased as the reaction temperature increased.When the reaction temperature reached 90 °C,complete oxidation of CO was achieved on CuCe/Ti-A,whereas the CO conversion on CuCe-C and CuCe/Ti-R was only 58.5% and 32.8%,respectively.By continuing to increase the temperature,CuCe-C completely converted CO at 120 °C.For CuCe/Ti-R,this temperature was 130 °C.The differences between these catalysts indicated that TiO2support and its crystal phase significantly affected the activity of CuCe catalyst for CO oxidation.The stability test of CuCe/Ti-A catalyst was carried out at 77 °C,when the CO conversion was about 90%,as shown in Fig.5(b).During the 30 h experiment,the CO conversion of CuCe/Ti-A catalyst has almost no change,indicating that CuCe/Ti-A catalyst has good stability.In addition,as shown in Table 2,we contrasted the activity of the catalysts used in this investigation with that of other CuCe catalysts found in the literature.Based on the evaluation results in the table,CuCe/Ti-A catalyst in our study showed high activity,although the reaction conditions were different in each study.

Fig.6 illustrates the results of testing the redox properties of the catalysts using H2-TPR to understand the influence of the TiO2support.According to the literatures[32,33],CuO has the H2reduction temperature of approximately 300 °C,and CeO2has the H2reduction temperature of above 400°C.As shown in Fig.6,CuO and CeO2have the reduction peaks are higher than those of CuCe-C,which have two peaks at 137°C and 153°C.This phenomenon is attributed to the oxidation–reduction equilibrium of Cu2++Ce3+→Cu++Ce4+and the synergistic effect between CeO2and CuO on the CuCe-C surface,which greatly increases the interaction between Cu and Ce[34].The introduction of Ce reduces the reduction peak temperature of the Cu species,making Cu species more susceptible to reduction.

Compared with that of CuCe-C,the reduction peak of CuCe/Ti-R shifts in the high-temperature direction,whereas the CuCe/Ti-A reduction peak shifts in the low-temperature direction.The crystal phases of TiO2influenced the interaction of Cu species with Ce or the support,which in turn influenced the reduction temperature of the Cu species.The two reduction peaks of CuCe/Ti-A have the lowest temperatures of the three CuCe catalysts,at 93 °C and 131 °C respectively,which are significantly lower than those of CuCe/Ti-R and CuCe-C.The results showed that the TiO2support influenced the interaction of Cu species with Ce or the support,which in turn made the Cu species of CuCe/Ti-A catalysts are more easily reduced.This may be one of the reasons why the catalytic performance of the CuCe/Ti-A catalyst is significantly better than that of the CuCe/Ti-R and CuCe-C catalysts.

Fig.2. TEM images of (a) CuCe-C,(b) CuCe/Ti-A and (c) CuCe/Ti-R catalysts.

Fig.3. HAADF-STEM and EDS mapping of the CuCe-C (a),CuCe/Ti-R (b) and CuCe/Ti-A (c) catalysts.

Fig.4. N2 adsorption–desorption isotherms (a) and pore size distribution (b) of CuCe-C,CuCe/Ti and TiO2 supports.

Fig.5. (a) Catalytic activity of CuCe-C,CuCe/Ti-A and CuCe/Ti-R catalysts in CO oxidation.(b) Stability of CuCe/Ti-A catalyst at 77 °C in CO oxidation.

Table2 Comparison of CO oxidation activity over different catalysts

Fig.6. H2-TPR curves of CuCe-C and CuCe/Ti with different TiO2 crystal phases.

The surface of the catalysts was characterized using XPS for the elemental composition and elemental states.The Ce 3d spectra of the catalysts are shown in Fig.7(a).The Ce 3d spectra are fitted to two types of spin–orbit components of eight peaks,namely,the Ce 3d5/2(marked asv,v′,v′′andv′′′) and the Ce 3d3/2(marked asu,u′,u′′andu′′′).The peaks ofv′andu′are related to the Ce3+species on the catalyst surface [35–37].The Ce3+content is calculated from the ratio ofv′andu′peak areas to the total peak area,using the following equation [38].

where ΣS(u+v)is the sum of the peak areas of the eight peaksv,v′,v′′,v′′′,u,u′,u′′andu′′′;Sv′andSu′correspond to the peak areas of thev′andu′peaks,respectively.As shown in Table 3,the CuCe/Ti-R catalyst has a higher Ce3+content.

Fig.7(b)shows the O 1s spectra of the catalysts.For the CuCe-C catalyst,the peaks at 529,531 and 533 eV are ascribed to the lattice oxygen (Olatt),surface oxygen (Oads) and hydroxyl oxygen (O—OH)[39,40],respectively.The O 1s spectra of the CuCe/Ti-R catalyst is similar to that of CuCe-C,but the O—OHpeak does not appear in the O 1s spectra of the CuCe/Ti-A catalyst.Table 3 lists the content of Olatt,which is calculated according to the peak area.The CuCe/Ti-A catalyst has the most Olattspecies,as shown by the results in Table 3.

The Cu 2p XPS spectra of the catalyst are shown in Fig.7(c),where the Cu 2p3/2was in the range of 937–931 eV and the satellite peak was in the range of 946–938 eV.The Cu 2p3/2could be fitted into two peaks approximately located at 934 and 932 eV,which attributable to the Cu2+and the low valent Cu,respectively[41,42].In addition,the reduced degree of Cu species can be determined by calculating the ratio between the intensity of the satellite peak and the principal peak (Isat/Ipp).There are more low-valence Cu species whenIsat/Ippis smaller [43,44].As shown in Table 3,theIsat/Ippof CuCe/Ti-A is smaller than that of CuCe/Ti-R and CuCe-C,suggesting the appearance of a greater amount of Cu+species.Cu+as the CO adsorption site affects the catalytic activity[22],CuCe/Ti-A has more Cu+,which is one of the important reasons for its superior activity.

Fig.7. XPS spectra of CuCe-C and CuCe/Ti with different TiO2 crystal phases.(a) Ce 3d,(b) O 1s,(c) Cu 2p,(d) Ti 2p.

Table3 XPS results of CuCe/Ti and CuCe-C catalysts (%)

The Ti 2p spectra of the CuCe/Ti catalysts are shown in Fig.7(d).Ti4+has two peaks at 464 eV and 458 eV,respectively,for Ti 2p1/2and Ti 2p3/2[45].No peaks of Ti3+or Ti2+are observed in the Ti 2p XPS spectra over CuCe/Ti catalysts.In these samples,Ti predominantly exists in the form of Ti4+.There is no significant difference in the Ti 2p spectra between the CuCe/Ti-A and CuCe/Ti-R catalysts.

Fig.8 displays the Raman spectra of CuCe-C,TiO2supports,and CuCe/Ti catalysts.The band at 285 cm-1for the CuCe-C catalyst is assigned to CuO[46],as illustrated in Fig.8.According to the literatures[31,34,47],the F2gvibrational mode of CeO2is at 463 cm-1.The F2gvibrational mode of the CuCe-C catalyst is red shifted to 450 cm-1,which may be due to the interaction between Cu and Ce [48].The oxygen vacancy is responsible for the band at 599 cm-1.TiO2-A has the vibrational bands at 142,197,396,516 and 639 cm-1and TiO2-R has bands at 236,447 and 610 cm-1[49].The vibrational bands of TiO2-R are clearly observed in the Raman spectra of the CuCe/Ti-R catalyst.The peak at 142 cm-1presented a blueshift to 146 cm-1in the Raman spectra of the CuCe/Ti-A catalyst.To exclude the effect of thermal treatment,the Raman spectra of TiO2-A after the calcination at 500 °C for 4 h(TiO2-A-500) were investigated,and the results were shown in Fig.8.Compared with TiO2-A,the peak of TiO2-A-500 was not blueshift,which indicated that thermal treatment was not the reason for blueshift in Raman spectra of CuCe/Ti-A catalyst.The interaction between the TiO2-A support and the CuCe species generating an oxygen vacancy probably caused the blueshift in Raman spectra of CuCe/Ti-A catalyst [50].CuCe/Ti catalyst shows the F2gvibrational mode of CeO2at 463 cm-1in Fig.8.The peak of the oxygen vacancy was not observed in Raman spectra of the CuCe/Ti catalysts,possibly because the band of the oxygen vacancy is closer to the vibrational band of TiO2.To further investigate the presence of oxygen vacancies on the CuCe/Ti catalysts,the CuCe/Ti catalysts were subsequently characterized using EPR.

The EPR spectra of the TiO2supports are shown in Fig.9(a).A signal atg=2.003 is detected in the TiO2-A and TiO2-R catalysts,indicating the presence of oxygen vacancies in both samples [51].The stronger EPR signal intensity of TiO2-A than TiO2-R indicates that the oxygen vacancy content of TiO2-A is higher than that of TiO2-R.As shown in Fig.9(b),the EPR signal of oxygen vacancies can still be found in CuCe/Ti-A.There is a new EPR signal atg=2.234,which indicated the Cu2+ions in CuO clusters [52].The EPR signal of the oxygen vacancy was not observed in CuCe/Ti-R,probably because the supported Cu and Ce occupied the oxygen vacancy[53].The support of CuCe had a large effect on the number of oxygen vacancies in the catalysts.The oxygen vacancies make adsorbed gaseous oxygen more likely to become lattice oxygen species,thus promoting CO oxidation.The higher oxygen vacancy concentration may be another significant factor for the excellent CO oxidation activity of the CuCe/Ti-A catalyst.

3.3.Reaction mechanism of CO oxidation

Fig.8. Raman spectra of CuCe-C,CuCe/Ti and TiO2 supports.

Fig.9. EPR spectra of CuCe/Ti and TiO2 supports.

To investigate CO oxidation over CuCe/Ti catalysts,in situFT-IR characterization of the CO adsorption and CO oxidation was carried out.Fig.10(a) shows the IR spectra of the CO adsorption on the CuCe/Ti-A catalyst at 30 °C.For CuCe/Ti-A,the peak at 2107 cm-1is associated to the linear CO adsorbed on Cu+sites (Cu+–CO)[54],indicating that Cu+is providing the CO adsorption sites.The peaks at 1341 and 1473 cm-1are associated to monodentate carbonate species [38],the peak at 1561 cm-1is associated to bidentate carbonate species [55],and the peak at 1393 cm-1is associated to bicarbonate species [56].The peak at 2170 cm-1is associated to gaseous CO [38],and the peaks at 2361 and 2335 cm-1are associated to gaseous CO2[55].There are two ways to produce CO2.First,CO can be oxidized to CO2by Cu2+,while Cu2+is reduced to Cu+[43,57].Second,as shown in Fig.10,when CO is adsorbed on the CuCe/Ti catalyst,carbonates and bicarbonates are formed,and the two can be decomposed into CO2.Fig.10(b)shows the IR spectra of CO adsorption on the CuCe/Ti-R catalyst.Compared with that of the CuCe/Ti-A catalyst,the IR spectra of the CuCe/Ti-R catalyst have no peak of bicarbonate.The intensity of the peaks on the CuCe/Ti-A catalyst stopped increasing during CO adsorption for 10 min,and reached adsorption saturation,whereas the CuCe/Ti-A catalyst reached adsorption saturation at 30 min.Comparing the intensity of the peaks of Cu+–CO (2107 cm-1) in the two samples,it can be seen that the intensity of the Cu+–CO peak on the CuCe/Ti-A catalyst is significantly higher than that on CuCe/Ti-R.Therefore,there are more Cu+species on the CuCe/Ti-A catalyst and thus more Cu+–CO was formed,which facilitated the CO oxidation reaction.

Fig.10. In situ FTIR spectra of CO adsorption over CuCe/Ti with different TiO2 crystal phases.(2%(vol)CO balanced by N2,50 ml∙min-1;30°C;from bottom to top:0,1,3,5,10,30,60 min).

Fig.11. In situ FTIR spectra of CO oxidation over CuCe/Ti with different TiO2 crystal phases.(1%(vol)CO,9%(vol)O2 balanced by N2,50 ml∙min-1;90°C;from bottom to top:0,1,3,5,10,30,60 min).

To further investigate the CO oxidation pathway,the CuCe/Ti catalysts were examined for CO and O2co-adsorption at 90 °C and the results are shown in Fig.11.Fig.11(a) shows that the intensity of the peak of Cu+–CO(2107 cm-1)on the CuCe/Ti-A catalyst during the CO and O2co-adsorption is relatively weak compared with the CO adsorption.The peaks of monodentate carbonate (1063,1343 and 1475 cm-1),bidentate carbonate(1543 cm-1) and bicarbonate species (1393 cm-1) appeared rapidly,the intensity of the peaks strengthened gradually with increasing adsorption time and were accompanied by the peaks of gaseous CO2(2335 and 2361 cm-1).That indicated that the CuCe/Ti-A catalyst exhibited rapid reaction performance for CO oxidation during the CO and O2co-adsorption.The peak of the monodentate carbonate species at 1343 cm-1shifts significantly with increasing adsorption time,which may be due to the peak of the bicarbonate species gradually increasing at 1393 cm-1with increasing time,causing a change in the shape of the peak at 1343 cm-1.As shown in Fig.11(b),the carbonate species are also observed for the CuCe/Ti-R catalyst during CO and O2coadsorption,and the intensities of the peaks are significantly weaker than those on the CuCe/Ti-A catalyst.The peak of the bicarbonate species (1393 cm-1) does not appear over the CuCe/Ti-R catalyst.During the CO and O2co-adsorption,the intermediate bicarbonate species on the CuCe/Ti-A catalyst are less thermally stable than carbonate species and would decompose more readily to form CO2and facilitate the CO oxidation reaction [24,56].

Based on these results,we propose a reaction mechanism for CO oxidation over CuCe/Ti catalysts.CO first adsorbed on the Cu+species to form Cu+–CO,which could react with lattice oxygen to form carbonate species or with surface hydroxyl groups to form bicarbonate species [58].The intermediate carbonate and bicarbonate were further decomposed to form CO2.The occupied lattice oxygen and surface hydroxyl groups were released after CO2desorption.The consumed lattice oxygen was also regenerated by the adsorbed gaseous O2on the oxygen vacancies[13].The reaction mechanism for CO oxidation was as follows.

where Cu+–CO denotes adsorbed surface species of CO in Cu+sites,*-O denotes chemisorption sites where surface carbonate species(*–CO3) are formed,*–OH denotes chemisorption sites where surface bicarbonates (*–CO2OH) are formed,*-γ denotes the oxygen vacancies.

4.Conclusions

TiO2-A and TiO2-R were used as supports in the preparation of CuCe/Ti catalysts,and the TiO2crystal phases significantly affected the catalytic performance of the CuCe/Ti catalysts.The superior catalytic activity of CuCe/Ti-A catalyst for CO oxidation may be attributed to factors such as its larger specific surface area,more Cu+and oxygen vacancies,and the Cu species being easily reduced by the CuCe/Ti-A catalyst.According toin situFT-IR results,the peak intensity of Cu+–CO on the CuCe/Ti-A catalyst was significantly higher than that of CuCe/Ti-R.More Cu+–CO species and the intermediate bicarbonate species were formed on the CuCe/Ti-A catalyst during CO oxidation.The intermediate bicarbonate species are less thermally stable than carbonate species and would decompose more readily to form CO2.

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

This work was supported by the National Natural Science Foundation of China (U21A20306,U20A20152) and Natural Science Foundation of Hebei Province (B2022202077).We thank Ian McNaught,Ph.D.,from Liwen Bianji,(Edanz) China (www.liwenbianji.cn/),for editing the English text of a draft of this manuscript.