Cerium phosphate-supported Au catalysts for CO oxidation☆

Yulin Wang,Huan Liu,Zhen Ma*

Shanghai Key Laboratory of Atmospheric Particle Pollution&Prevention(LAP 3),Department of Environmental Science&Engineering,Fudan University,Shanghai 200433,China

Keywords:Gold Catalysis Catalyst support CePO4 Carbon monoxide Oxidation

ABSTRACT LaPO4 and hydroxyapatite(Ca10(PO4)6(OH)2)are typical metal phosphates recently found to be useful for making supported metal or metal oxide catalysts,but CePO4(also belonging to the metal phosphate family)has been rarely used to make supported catalysts.It would be interesting to develop CePO4-supported catalysts and explore their catalytic applications.Herein,hexagonal CePO4 nanorods(denoted as CePO4-H),hexagonal CePO4 nanowires(CePO4-HNW),monoclinic CePO4 nanoparticles(CePO4-M),and monoclinic CePO4 nanowires(CePO4-MNW)prepared by different methods were used to support gold via deposition-precipitation with urea(DPU).The gold contents of these catalysts were all around 1 wt%.The catalytic activities of these Au/CePO4 catalysts in CO oxidation were found to follow the sequence of Au/CePO4-MNW N Au/CePO4-HNW N Au/CePO4-M N Au/CePO4-H.These catalysts were characterized by inductively coupled plasma-optical emission spectroscopy(ICP-OES),N2 adsorption-desorption,X-ray diffraction(XRD),transmission electron microscopy(TEM),X-ray photoelectron spectroscopy(XPS),oxygen temperature-programmed desorption(O2-TPD),and CO2 temperature-programmed desorption(CO2-TPD)to find possible correlations between the physicochemical properties and catalytic activities of these catalysts.

1.Introduction

Gold was rarely used as a catalyst component several decades ago,not only because gold is expensive,but also because heterogeneous gold catalysts were found to be barely active in catalysis at that time.However,since Haruta and co-workers discovered that gold loaded on a suitable metal oxide support(such as TiO2)could be active for CO oxidation below room temperature[1],research on the synthesis,characterization,and application of supported gold catalysts has been booming.Supported gold catalysts are useful for catalyzing a number of reactions related to environmental remediation and the synthesis of chemicals[2-6].

Supports typically used for the preparation of heterogeneous gold catalysts include TiO2,Fe2O3,Al2O3,and SiO2[2,3].Supports can in fluence the physicochemical properties and catalytic performance of supported gold catalysts in profound ways.For instance,it was found that reducible supports are more suitable for making active gold catalysts than non-reducible supports[7,8].Different solid supports(with different isoelectric points)can not only in fluence the amount of gold to be deposited onto these supports,but also in fluence the size and thermal stability of the supported gold nanoparticles.As a result,the activities of different supported gold catalysts in a catalytic reaction can differ widely.Oxide supports have often been modified by other metal oxides to tune the physicochemical properties and the catalytic performance of the resulting supported gold catalysts[9].In addition,supported gold catalysts based on metal oxides with different crystal phases or morphologies were found to show different catalytic performance[10-13].However,metal salts have been seldom used as supports for making supported gold catalysts.

LaPO4and hydroxyapatite(Ca10(PO4)6(OH)2)are non-reducible metal phosphates.They have different acid-base properties and good thermal stability,and thus have been used as supports for making supported catalysts such as Au/LaPO4[14-18],Pt/LaPO4[19,20],Pd/LaPO4[21],Rh/LaPO4[22-24],Au/hydroxyapatite[15,25-28],Pt/hydroxyapatite [19],Rh/hydroxyapatite [22,29,30],and Ru/hydroxyapatite[31-33]useful for CO oxidation[14-19,24,26,27],N2O decomposition[22,24,29,30,33],NO decomposition[21],three-way catalysis[23],photocatalytic reduction of CO2[20],the water-gas shift reaction[31],direct tandem synthesis of imines and oximes[25],aerobic oxidation of cyclohexane[28],and aerobic oxidation of benzyl alcohols[32].

Similar to LaPO4,CePO4also belongs to the metal phosphate family.CePO4has been used as a catalyst for the dehydration of isopropanol[34],the oxidative dehydrogenation of isobutene[35],the removal of mercury elements[36],and the selective catalytic reduction of NO with NH3[37].CePO4can also be used as a support to make supported catalysts such as Au/CePO4for CO oxidation[38]and electrochemical oxygen reduction reaction[39],Pt/CePO4for the electrochemical methanol oxidation[40],Ru/CePO4for the aerobic oxidation of alcohols[41],and Rh/CePO4for N2O decomposition[42,43].Nevertheless,examples on the synthesis and application ofCePO4-supported catalysts are still limited in number.It would be interesting to develop CePO4-supported catalysts and explore their catalytic applications.

The physicochemical properties of CePO4supports may in fluence the catalytic performance of CePO4-supported catalysts.CePO4materials prepared by different methods and with specific preparation parameters may have different morphologies and crystal phases.For instance,hexagonal CePO4nanorods can be prepared by precipitation[44],whereas hexagonal CePO4nanowires can be prepared by a hydrothermal method[45,46].Monoclinic CePO4materials can be prepared by heating hexagonal CePO4at high temperatures(N 600°C)[47,48].These preparation methods may provide new opportunities for us to investigate the in fluence of different CePO4supports on the catalytic performance of supported gold catalysts.

One issue in the preparation of supported gold catalysts via a deposition-precipitation(DP)method is that the gold complex in the synthesis mixture cannot be effectively deposited onto supports and the amount of gold deposited varies greatly when using different supports[15,49].A DPU(deposition-precipitation with urea)method[50,51]could circumvent this problem so that the gold precursor can be used economically and the catalytic activities of different supported gold catalysts can be compared at the same gold content level.

In this work,CePO4materials with different crystal phases and morphologies were used as supports to load gold via a DPU method.The catalytic activities of these catalysts in CO oxidation were investigated.CO oxidation is not only a reaction of environmental relevance,but also a sensitive probe reaction that can be used to compare the activities of different catalysts[52].Interestingly,the catalytic activities of different Au/CePO4catalysts prepared using different CePO4supports were found to differ obviously.Relevant characterization was conducted to find the reason for this observation.

2.Experimental

2.1.Catalyst preparation

To 80 ml deionized water,6.95 g Ce(NO3)3·6H2O(Aladdin,99.9%)and 1.84 g NH4H2PO4(Sinopharm,AR)were added.The mixture was magnetically stirred for 20 min to allow for the complete dissolution of these salts.

Hexagonal CePO4nanorods(CePO4-H)and monoclinic CePO4nanoparticles(CePO4-M)were prepared according to the literature[47,53].The pHvalue of the above solution was adjusted to 7 by adding aqueous ammonia(27 wt%)dropwise,and the mixture was then stirred at room temperature for 6 h.The product(CePO4-H)was filtered,washed with ethanol,and dried in an oven at 80°C for 4 h.CePO4-M was prepared by calcining CePO4-H at 900°C for 4 h.

Hexagonal CePO4nanowires(CePO4-HNW)and monoclinic CePO4nanowires(CePO4-MNW)were prepared according to the literature[45,54].The pH value of the above solution was adjusted to 1 by adding aqueous ammonia(27 wt%)dropwise.The suspension was transferred to a Te flon-lined stainless steel autoclave and He was bubbled through the suspension for 1 h to remove the air from the autoclave in order to preventthe oxidation ofCe3+to Ce4+.Then the autoclave was sealed,heated at 150°C for 12 h,and naturally cooled to room temperature.The product(CePO4-HNW)was filtered,washed with ethanol,and dried in an oven at 80°C for4 h.CePO4-MNW was obtained by calcining CePO4-HNW at 900°C for 4 h.

Supported gold catalysts were prepared via deposition-precipitation with urea(DPU)[18].Urea(1.89 g)was dissolved in 75 ml HAuCl4solution(2.03 × 10-3mol·L-1),and 3 g CePO4was added,and the mixture was stirred at 80°C for 4 h.The suspension was centrifuged,washed with deionized water,and dried at 80°C for 12 h.The collected powders were calcined in a muf fle furnace in static air at 350°C(or 500°C)for 3 h.

2.2.Catalytic activity tests

The catalytic tests were performed using a FINESORB-3010 instrument( fixed-bed continuous flow microreactor,FINETEC)at atmospheric pressure.Before each test,a catalyst(0.15 g)was packed within the U-shaped glass tube and then pretreated with flowing He(30 ml·min-1)at 200 °C for 2 h.The catalyst was completely cooled to room temperature or 0°C and subjected to CO oxidation testing.The reaction gas,containing 1 vol%CO/airata flow rate of50 ml·min-1,was introduced into the reactor.The reaction temperature at 0°C was achieved by using an ice-water bath in a Dewar.The reaction temperature between 0°C and room temperature was actualized by removing portions of ice water or cold water and adding almost the same amount of water maintained at room temperature.The actual temperature was measured by a thermal couple inserted in the reactor.The reaction temperature above room temperature was increased to 200°C at a heating rate of 0.5 °C·min-1using a furnace.The reaction products were analyzed by using gas chromatograph(Agilent 7890A)with a thermal conductivity detector.The CO conversion was calculated as([CO]in-[CO]out)/[CO]in×100%,where[CO]inrefers to the concentration of CO without any catalyst in the U-shaped glass tube,and[CO]outrefers to the concentration of CO exiting the catalyst bed.

2.3.Catalyst characterization

The Au contents of catalysts were determined by ICP-OES(Perkin-Elmer OPTIMA 2100 DV optical emission spectrometer).Gold of precisely weighed samples was dissolved in aqua regia,and the obtained solutions were diluted and analyzed.

The specific Brunauer-Emmett-Teller(BET)surface areas of samples were determined by using a Micromeritics Tristar 3000 instrument at the liquid nitrogen temperature(-196°C).

X-ray diffraction(XRD)data were obtained with a MSAL XD2 instrument using a graphite monochromator and Cu Kα radiation.The scanning angles(2θ)were 10°-80°,and the scanning rate was 8(°)·min-1.

Transmission electron microscopic(TEM)experiments were conducted using a JEM-2011F transmission electron microscope operated at 200 kV.The sample was dispersed in ethanol,loaded on Cu gridsupported carbon film,and dried under an infrared lamp before analysis.

X-ray photoelectron spectroscopic(XPS)experiments were conducted using a Perkin-Elmer PHI 5000 C spectrometer with Mg KαX-ray as the emission source.The binding energies were calibrated by using the C1s peak at 284.8 eV.

Temperature-programmed desorption of O2or CO2(O2-TPD or CO2-TPD)were conducted on a FINESORB-3010 instrument with a TCD.A sample(0.1 g)was loaded into a U-shaped quartz tube,and then pretreated with He(30 ml·min-1)at 200 °C for 2 h.The sample was cooled down to room temperature,and the sample was then heated from room temperature to 50 °C at a rate of 5 °C·min-1.At the same time,pure O2or 5 vol%CO2/air was fed at a flow rate of 50 ml·min-1for 1 h.The system was then purged with He(30 ml·min-1)for 3 h.Finally,the sample was heated to 550 °C at a ramping rate of 10 °C·min-1.

3.Results and Discussion

3.1.Catalytic activity of Au/CePO4 in CO oxidation

Fig.1.CO conversions on Au/CePO4 catalysts calcined at 350°C.

Fig.1 compares the catalytic activities ofdifferent Au/CePO4catalysts(calcined at 350°C)in CO oxidation.The catalytic activities were tested asa function ofreaction temperature.As shown in the figure,Au/CePO4-MNWcalcined at350°C is the mostactive,achieving 71%COconversion at 5 °C and complete CO conversion at 20 °C.Au/CePO4-HNW is the second most active,achieving complete CO conversion at 30°C.Au/CePO4-M is the third most active,achieving complete CO conversion at 55°C.Au/CePO4-H shows the lowest activity among these catalysts.It achieves 11%CO conversion at 25 °C and 100%CO conversion at 80 °C.Overall,the activities of these catalysts clearly follow the sequence of:Au/CePO4-MNW N Au/CePO4-HNW N Au/CePO4-M N Au/CePO4-H,i.e.,Au/CePO4-MNWand Au/CePO4-HNWprepared by using CePO4nanowires(either monoclinic or hexagonal)are more active than Au/CePO4-M and Au/CePO4-H prepared by using monoclinic CePO4nanoparticles or hexagonal CePO4nanorods.The most active Au/CePO4-MNW in this study is slightly lessactive than Au/LaPO4-NW(with 1 wt%Au,achieving ca.90%COconversion at0 °C and 100%COconversion at15 °C)reported previously[18].Note that in both sets of experiments(the current experiments and those in our previous work[18]),the reaction conditions(i.e.,catalyst amount,CO concentration,gas flow rate)are the same except that the catalysts are different.

To see whether these catalysts can withstand high-temperature calcination and whether the activity trend observed above is general,Au/CePO4catalysts calcined at 500°C were also tested in CO oxidation.Such a high-temperature calcination step is usually used to test the thermal stability of supported gold catalysts as the agglomeration of gold nanoparticles upon high-temperature treatmentmay be a problem for practical applications[55,56].

As shown in Fig.2,the activities of the catalysts calcined at 500°C still follow the sequence of Au/CePO4-MNW N Au/CePO4-HNW N Au/CePO4-M N Au/CePO4-H,but the activities of the former three catalysts decrease obviously compared with those of the corresponding Au/CePO4catalysts calcined at 350°C.This observation may be due to the sintering of gold nanoparticles to some extent,as will be proved by TEM data later.However,the activity of the least active Au/CePO4-H calcined at 500°C is almost the same as that of the same catalyst calcined at 350°C.This point will be discussed further using TEM data presented later.

As seen from Figs.1 and 2,the difference of catalytic activity of different Au/CePO4catalysts is more obvious for the catalysts calcined at 350 °C,and Au/CePO4catalysts calcined at 350 °C are in general more active than the corresponding catalysts calcined at 500°C.Thus,subsequent characterization was mainly focused on Au/CePO4calcined at 350°C.

3.2.Characterization of Au/CePO4

Fig.2.CO conversions on Au/CePO4 catalysts calcined at 500°C.

To understand why the activities of Au/CePO4catalysts are so different,we first resort to XRD characterization.As shown in Fig.3,CePO4-H and CePO4-HNW both exhibit the hexagonal phase of CePO4(in reference to a standard card:PDF#04-0632),whereas CePO4-M and CePO4-MNW(prepared by calcining CePO4-H and CePO4-HNW,respectively,at 900°C)show the monoclinic phase(PDF#32-0199).Their respective crystal phases are kept after loading gold,but no gold peaks can be observed,probably due to the low gold loading(theoretical value 1 wt%)and high dispersion of gold nanoparticles.Indeed,the gold contents of Au/CePO4-H,Au/CePO4-M,Au/CePO4-HNW,and Au/CePO4-MNW were determined by ICP-OES as 1.07 wt%,0.97 wt%,0.97 wt%,and 0.99 wt%(Table 1),all close to 1 wt%,and gold nanoparticles were found to be dispersed on CePO4supports(as presented by TEM data in Fig.4).

TEM data can provide direct information about the morphologies of CePO4supports as well as the dispersion and sizes ofgold nanoparticles.Fig.4 shows the TEM images of Au/CePO4catalysts calcined at 350°C.Au/CePO4-H mainly exhibits CePO4nanorods(packed together)with lengths on the order of 100 nm,and some relatively big gold nanoparticles are dispersed on the support.Au/CePO4-M is composed of large CePO4nanoparticles with sizes on the order of 50-100 nm.However,the morphologies of the CePO4supports change obviously after calcination at900°C.Both Au/CePO4-HNWand Au/CePO4-MNWexhibit uniform CePO4nanowires with lengths of several hundred nanometers.

The BET surface areas of Au/CePO4-H,Au/CePO4-M,Au/CePO4-HNW,and Au/CePO4-MNWare 114.0,10.1,31.3,and 18.1 m2·g-1,respectively(Table 1).Since the gold loading of each catalyst is close to 1 wt%,the surface areas of the supported gold catalysts should be mainly determined by those of the CePO4supports.The surface area data thus show that CePO4-H prepared by precipitation has a higher surface area than CePO4-HNW prepared by hydrothermal synthesis at 150°C for 12 h,whereas CePO4-M and CePO4-MNW prepared by calcining CePO4-H and CePO4-HNW,respectively,at900°C both have low surface areas.The surface area data seem to be consistent with the morphologies and sizes of the CePO4supports imaged by TEM,in the sense that supports with larger sizes usually have low surface areas.

From the TEM images in Fig.4,we can also see the presence of gold nanoparticles on CePO4supports.It should be mentioned that these are representative TEMimages,meaning thatother TEMphotos were taken and used in the statistics of gold particle sizes.Fig.5 summarizes the gold particle size distribution based on the statisticsof100 gold particles for each sample.The average sizes of gold nanoparticles were obtained using the DigitalMicrograph software.The average gold particle sizes of Au/CePO4-H,Au/CePO4-M,Au/CePO4-HNW,and Au/CePO4-MNW are 8.4,7.2,6.8,and 5.6 nm,respectively,correlating with the activities of these catalysts,i.e.,Au/CePO4-MNW with the smallest average gold particle size shows the highest activity in CO oxidation,whereas Au/CePO4-H with the biggest average gold particle size shows the lowest activity.

Fig.3.XRD patterns of Au/CePO4 catalysts and corresponding supports.

Table 1 ICP,BET surface areas,T50 values,and average gold particle size of Au/CePO4 catalysts

To know why the activities of Au/CePO4-MNW,Au/CePO4-HNW,and Au/CePO4-M decrease after calcination at 500°C whereas that of Au/CePO4-H calcined at 500°C is similar to that of Au/CePO4-H calcined at 350°C(Figs.1 and 2,Table 1),TEM experiments involving Au/CePO4catalystscalcined at500°C were also conducted.Asshown in Fig.S1,the morphologies of these CePO4supports are the same as those of the corresponding supports of Au/CePO4calcined at 350°C.Fig.6 summarizes the gold particle size distribution based on the statistics of 100 gold particles for each sample.The average gold particle sizes of Au/CePO4-MNW,Au/CePO4-HNW,and Au/CePO4-M increase to 6.3 nm(c.f.5.6 nm for the sample calcined at 350°C),7.2 nm(c.f.6.8 nm),7.5 nm(c.f.7.2 nm),respectively,whereas the average gold particle size of Au/CePO4-H is 8.4 nm(c.f.8.4 nm).The data thus show that the sizes of gold nanoparticles on CePO4support correlate to the activities of these Au/CePO4catalysts.In addition,the seemingly small increase of average particle size(e.g.,from 5.6 to 6.3 nm for Au/CePO4-MNW)when increasing the calcination temperature from 350 to 500°C can make the catalytic activity decrease obviously(Figs.1 and 2),indicating that catalytic activity is sensitively in fluenced by the gold particle size.It was found previously that the size ofgold nanoparticles on supports matters forachieving high activity in COoxidation,and the reason for a supported gold catalyst to keep its high activity after high-temperature treatment is the stabilization of gold nanoparticles by a suitable support[55,56].Of course,high-temperature calcination may also change the contents of Au+,Au0,and surface hydroxyls as well as the microscopic details and other physicochemical properties ofthe support.These factors may also exertsubtle in fluences on the catalytic performance.

As mentioned before,the BET surface areas of Au/CePO4-H,Au/CePO4-M,Au/CePO4-HNW,and Au/CePO4-MNW are 114.0,10.1,31.3,and 18.1 m2·g-1,respectively(Table 1).It seems that CePO4-H with the highest surface area cannot lead to the formation of very small gold nanoparticles,probably due to the intrinsic nature of the support,but at least it can mitigate the sintering of gold nanoparticles at 500°C since the hexagonal CePO4nanorods are packed together(Fig.4),creating rough surfaces and relatively high surface areas.

Fig.4.TEM images of(a)Au/CePO4-H,(b)Au/CePO4-M,(c)Au/CePO4-HNW,and(d)Au/CePO4-MNW catalysts calcined at 350°C.

Fig.5.The size distributions of gold nanoparticles of Au/CePO4 catalysts calcined at 350°C.

Above,we have shown thatthe size ofgold nanoparticlesmattersfor catalytic CO oxidation.However,gold species may exist on the support in different forms,i.e.,cationic or metallic.Fig.7 shows the Au 4f XPS spectra of Au/CePO4catalysts calcined at 350°C(without undergoing reaction testing).The peak at about 84.3 eV corresponds to Au0of supported gold catalysts[57,58],whereas the peak located at 84.8 eV[58]or 85.6 eV[59]corresponds to Au+.Detailed data are summarized in Table 2.In any case,Au0is the main gold species,and Au+is a minority gold species.Au/CePO4-MNW with the highest activity has a relatively low Au+content,whereas for Au/CePO4-H with the lowest catalytic activity,the Au+content is not the highest.For Au/CePO4catalysts collected after reaction,the Au+content of each catalyst is similar to that of the corresponding fresh catalyst(Fig.S2 and Table 2).These data indicate that probably Au+does not play a major role in determining the catalytic activity,and thatmetallic Au nanoparticles are more important for catalytic CO oxidation[2,3].

Fig.6.The size distributions of gold nanoparticles of Au/CePO4 catalysts calcined at 500°C.

Above,we have showed that the catalytic activities of catalysts follow the sequence of Au/CePO4-MNW N Au/CePO4-HNW N Au/CePO4-M N Au/CePO4-H,and ascribed the difference in activity to the size of gold nanoparticles.However,other factors(physicochemical properties)should be considered and discussed.Surface hydroxyls were claimed to be able to be involved in CO oxidation over certain gold catalysts[60].Thus,Fig.8 summarizes the O 1s XPS spectra of Au/CePO4catalysts before reaction.The O 1s peak can be divided into two main characteristic peaks.The peak at about 531.1 eV corresponds to lattice oxygen,whereas the peak centered at about 532.5 eV is due to surface hydroxyls[10,47].The relative percentages of the surface hydroxyls in oxygen species of Au/CePO4-H,Au/CePO4-M,Au/CePO4-HNW,and Au/CePO4-MNW calcined at 350°C(without catalyzing CO oxidation)are 21.6%,20.0%,19.2%,and 13.4%,respectively.The corresponding values of these catalysts collected after reaction testing are 22.0%,20.7%,20.6%,and 14.4%,respectively(Fig.S3).The data show that the surface hydroxyl concentration increases a little after the reaction,but there is no simple correlation between the surface hydroxyl concentration and catalytic activity.

Fig.7.Au 4f XPS spectra of(a)Au/CePO4-H,(b)Au/CePO4-M,(c)Au/CePO4-HNW,and(d)Au/CePO4-MNW before reaction.

Table 2 The binding energy of Au 4f and the corresponding proportions of Au+

The abilities for CePO4supports or Au/CePO4in adsorbing and releasing oxygen should also be considered.Fig.9 shows the O2-TPD data of Au/CePO4catalysts and the corresponding supports.For Au/CePO4-H and Au/CePO4-HNW prepared with hexagonal CePO4supports,there is an O2desorption peak at 180-300°C,indicating that oxygen can be adsorbed on these catalysts and released upon heating towards 300°C.On the other hand,there is no O2desorption peak for Au/CePO4-M and Au/CePO4-MNW prepared with monoclinic CePO4supports,indicating that there is virtually no O2adsorption on these catalysts orthe adsorption ofO2on these catalystsis reversible,i.e.,there is no strong adsorption of O2.In addition,it is clear from the figure that the O2-TPD profiles of these Au/CePO4catalysts are quite similar to those of their corresponding supports,meaning that the O2adsorption-desorption behavior of Au/CePO4catalysts is largely determined by their corresponding CePO4supports.There is a marked difference between hexagonalCePO4supports and monoclinic supports.However,such a difference in terms ofoxygen adsorption-desorption behaviorcannot account for the observed activity difference(Au/CePO4-MNW N Au/CePO4-HNW N Au/CePO4-M N Au/CePO4-H).Rather,the difference is merely intrinsic for different CePO4supports.

Fig.9.O2-TPD profiles of Au/CePO4 catalysts.

Fig.8.O 1s XPS spectra of(a)Au/CePO4-H,(b)Au/CePO4-M,(c)Au/CePO4-HNW,and(d)Au/CePO4-MNW before reaction.

Fig.10.CO2-TPD profiles of Au/CePO4 catalysts.

A similar argument can be made for the lack of correlation between the basicity of Au/CePO4catalysts and their catalytic activity.CO2can be adsorbed on basic sites,and itsdesorption upon heating can indicate the basicity of the solid sample.If the CO2desorption peak is big,then that means that the solid sample contains a lot of basic sites.On the other hand,if CO2molecules desorb at high temperatures,that means that the solid sample has strong basicity.Fig.10 shows the CO2-TPD profiles of Au/CePO4catalysts and their corresponding supports.It is found that while Au/CePO4-H and Au/CePO4-HNW show an obvious CO2desorption peak above 200°C,there is virtually no CO2desorption from Au/CePO4-M and Au/CePO4-MNW.The CO2desorption behaviors of Au/CePO4catalysts are similar to those of their corresponding CePO4supports,i.e.,hexagonal CePO4supports have an obvious CO2desorption peak whereas monoclinic CePO4supports virtually do not have CO2desorption peak.The data indicate that the basicity is an intrinsic property of the CePO4supports,and there is no clear correlation between the basicity of Au/CePO4catalysts and their catalytic activities in CO oxidation.

The overallconclusion ofthe work is thatmetallic gold nanoparticles with small sizes seem to play a major role in determining the catalytic activity of Au/CePO4in CO oxidation,whereas the size of gold nanoparticles is determined by the nature(i.e.,physicochemical properties such as sizes,morphologies,surface areas)of CePO4supports and by the calcination temperature.To put the work in perspective,it should be mentioned thatourgroup also found thatAu/LaPO4based on hexagonal LaPO4nanowires is more active in COoxidation than Au/LaPO4based on commercial(hexagonal)LaPO4composed of many short nanorods packed together,due to the smaller average gold particle(3.7 nm)for the former catalystthan that(5.9 nm)ofthe latter[18].Further theoretical input is needed for revealing the fundamental reasons of the stabilization of gold nanoparticles on supports with different crystal phases and morphologies.

4.Conclusions

Hexagonal CePO4nanorods(CePO4-H)were prepared by precipitation,and hexagonal CePO4nanowires(CePO4-HNW)were prepared by hydrothermal synthesis.Monoclinic CePO4nanoparticles(CePO4-M)and monoclinic CePO4nanowires(CePO4-MNW)were prepared by calcining CePO4-H and CePO4-HNW,respectively,at 900°C for 4 h.Gold was loaded onto these supports via DPU(deposition-precipitation with urea).The activities of catalysts(calcined at 350 or 500°C)in CO oxidation consistently follow the sequence of Au/CePO4-MNW N Au/CePO4-HNW N Au/CePO4-M N Au/CePO4-H.XRD and TEM results show the gold particles on supports are highly dispersed.The average size of Au nanoparticles correlates to the catalytic activity.The smaller the gold nanoparticles are,the higher the activity.XPS data indicate that Au0is a majority gold species,whereas Au+is a minority gold species.The O2adsorption-desorption behaviors and basicity of Au/CePO4catalysts are largely determined by those of their corresponding CePO4supports,i.e.,these properties are intrinsic to CePO4supports.There is no clear correlation between the O2adsorption-desorption behaviors(or basicity)of Au/CePO4catalysts and the catalytic activities in CO oxidation.Rather,metallic gold nanoparticles with small sizes seem to play a major role.This work provides a catalyst system for further theoretical work and fundamental application in catalytic CO oxidation and other reactions.

Supplementary Material

Supplementary data to this article can be found online athttps://doi.org/10.1016/j.cjche.2017.08.008.