Supported nanosized gold catalysi the influence of support morphology and reaction mechanism 5

In preferential oxidation of carbonmonoxide over Au/CuO catalysts testing, Au/CuO NP still showed the best activity interms of CO conversion, while gold supported samples on three kinds

Chapter 5 Oxidation of Carbon Monoxide over

Performances and Reaction

Mechanism-In this chapter, CuO with different morphologies and structures were utilized to supportgold nanoparticles Au/CuO(NP) showed better catalytic activity than the Au/CuO(CB)and Au/CuO(NF) samples, whereby there is no significant differences in surface area ofthese three kinds of supports More Au+species were found on the Au/CuO(NP) sample,indicating different interactions between the Au nanoparticles and the CuO supports.Also great differences in catalytic activity for samples pre-treated at different temperaturewere observed Catalysts pre-treated at 300oC showed much better catalytic activities for

CO oxidation than catalysts pre-treated at 200oC In preferential oxidation of carbonmonoxide over Au/CuO catalysts testing, Au/CuO (NP) still showed the best activity interms of CO conversion, while gold supported samples on three kinds of copper oxidesupports showed similar H2selectivity

Due to their unique catalytic activity in various reactions, metal oxide supportednanosized gold has become one of the most popular research topics, and has attracted theattentions of researchers all over the world Extensive works have been conductedconcerning both application and fundamental academic research Nevertheless, there are

still a lot of controversial and uncertain topics in this area of study For example, thereaction mechanism and factors that can affect the reaction remain debatable It has beengenerally agreed that particles size of gold and the nature of the metal oxide supportshave a great influence on catalyst’s activity Other factors such as preparation method,support surface area; pre-treatment conditions etc can also affect their catalytic activity.Many kinds of metal oxide have been used as supports, and the most widely usedsupports are TiO2, Fe2O3, Al2O3, MgO, CeO2 and Co3O4 etc Although gold supported

on reducible metal oxide supports showed promising results, and copper oxide can beclassified into easily reducible oxide support, not much research has been conductedusing copper oxide as support for nano gold catalysis CuO itself can catalyze COoxidation, 1,2 and copper-based catalysts are known to be active in several industrialchemical processes, such as the methanol synthesis, the water-gas shift reaction and thecatalytic oxidation of hydrocarbons Many of these catalytic reactions involveintermediate states showing that carbon monoxide directly interacts with the copper-based catalyst A recent study showed 100% CO conversion on Au/CuO at temperaturesbetween 95 and 125oC.3 G Hutchings compared nano-Au catalysts prepared by co-precipitation on CuO, CuO/ZnO and ZnO supports Au particles on CuO and CuO/ZnOsupports had large particles size and hence lower CO oxidation activity4 But in generalthe study of CuO- supported Au catalysts for CO oxidations is very limited

In this chapter three CuO samples with different morphologies and structures, includingbulk (commercial, denoted as CuO (CB)), nanoflake (self-made, denoted as CuO (NF))and nanoparticles (commercial, denoted as CuO (NP)), were utilized to prepare CuO-support gold catalysts The catalytic activities and kinetics of carbon monoxide oxidation

were investigated using on-line GC The mechanism of the low temperature CO

oxidation was carefully investigated using in-situ DRIFT and in-situ XPS.

5.2.1 Materials and catalysts preparation

Commercial bulk CuO purchased from Merck (particle size~150 nm, surface area7.9m2/g) and nanoparticle CuO from Sigma-Aldrich (particle size~80nm surface area27.1m2/g) were used in the experiment without further treatment CuO nano flakes used

in this work were prepared by precipitation: The precipitation was performed by adding0.5 M NaOH dropwise to the prepared 0.5M Cu(NO3)2∙2H2O (Merck, >98.5%) solutiontill pH 9.5 The resulting mixture was stirred at 80oC for 48 h Then the precipitatemixture was separated by centrifuge and washed by deionized water for four times Theobtained powder was dried at room temperature for 24 h and calcined at 400oC for 6 h instatic air The surface area (BET) of the as-prepared CuO nanoflakes is 17.8 m2/g

These three kinds of copper oxide samples were used as support for depositing gold nanoparticles Colloid-based (CB) method as described in Chapter 3 was selected as the bestmethod for the preparation of Au/CuO catalysts HAuCl4(1mM) was used as a precursor,NaBH4 (0.1M) as a reducing agent and lysine as a capping agent During the reductionperiod, sonication was applied The slurry was dried at 70ºC after centrifuge four timesusing DI water.5-9

5.2.2 Evaluation of catalysts

Catalytic evaluations were carried out at atmospheric pressure in a continuous-flowfixed-bed quartz micro-reactor (I.D 4 mm) packed with samples and quartz wool Before

testing, the catalysts were pre-treated in-situ with a flow of air (100 ml min-1) for 1 h at

200 and 300oC respectively For CO oxidation reactions, the feed gas was a mixture of90%He + 5%CO + 5%O2, which was introduced into the reactor at a gas hourly spacevelocity (GHSV) of 60,000 cm3 g-1h-1 For preferential oxidation of CO in the presence

of hydrogen, the feed gas was a 70%H2+ 1%CO + 2%O2 mixture, introduced into thereactor at a GHSV of 60,000 cm3 g-1h-1 For both reactions, the reaction products wereanalyzed on-line using Shimadzu GC-2010 gas chromatography equipped with a thermalconductivity detector (TCD) The catalysts were evaluated for activity (in terms of COconversion) and CO2 productivity in a temperature range of 25-300 oC We tookmeasurement readings after the system had stabilized for at least 15mins for everydesignated reaction temperature For kinetics study, details are given on Chapter 3 page55-56

5.2.3 Characterization of catalysts

Powder X-ray diffraction patterns were recorded at room temperature on a Bruker D8Advance Diffractometer using a Cu Kα radiation source Diffraction angles were

measured in steps of 0.015o at 1 s/step in the range of 10-80o (2θ)

Transition electron microscope measurements were performed on a Tecnai TF 20 S-twininstrument Before measurement, all samples were ultrasonically dispersed in ethanolsolvent and then were dried over a carbon grid The average size of Au particles and itsdistributions was estimated by counting about 300 Au particles JEOL JSM-6700F FieldEmission Scanning Electron Microscope was used to observe the particle shape, size andmorphology The Au and Cu contents of prepared catalysts were determined by X-ray

fluorescence multi-elemental analyses on a Bruker AXS S4 Explorer.

Temperature programmed reduction studies were performed in a continuous-flow bed quartz micro-reactor (I.D 4 mm) with 50 mg samples The catalyst was firstoutgassed by heating at 300oC under air flow for 60 min to make sure that the sampleswere tested under the same condition as CO oxidation reaction After cooling to roomtemperature, the feed gas was switched to 5%H2/Ar After the baseline had stabilized, thetemperature was increased to 600oC at a heating rate of 10oC /minute The amount of H2consumed was measured as a function of temperature by means of a thermal conductivitydetector (TCD)

fixed-The in-situ Diffusion Reflectance Infrared Fourier Transform spectroscopy (DRIFTS) of

CO adsorption study was carried out on a Bio Rad FTIR 3000 MX spectrometer equippedwith a reaction cell (modified Harricks model HV-DR2) The CuO or CuO-supported Ausample was loaded into the DRIFT cell with 1:1 weight ratio with KBr The spectra wereacquired with a resolution of 4cm-1 typically averaging 150 scans The sample waspurged with Helium flow (20 ml min-1) for 2 hours before exposure to reaction gas For

CO adsorption experiments, in the flow of various concentrations of CO (0.5%, 1% and2.0% CO in He) DRIFT spectra were taken after And as for DRIFT study on surfacespecies during CO oxidation reaction, the spectra were taken after the introduction of5%CO with 5%O2 in He balance Table 5-1 summarizes the experimental procedure

implemented for CO adsorption and CO oxidation DRIFT study

Table 5-1 Experimental procedure for CO adsorption and oxidation DRIFT study

Pre-treat catalyst in air (He) flow at 300oC

(573 K) for 1 hour and then cool down to

Purge in He flow for at least 30 mins

remove gas phase CO and physisorbed CO

Purge in He flow for at least 30 minsremove gas phase CO and physisorbed CO

Purge in He flow for at least 30 mins

remove gas phase CO and physisorbed CO

Purge in He flow for at least 30 minsremove gas phase CO and physisorbed CO

remove gas phase CO and physisorbed CO remove gas phase CO and physisorbed CO

Purge in He flow for at least 30 mins

remove gas phase CO and physisorbed CO

Purge in He flow for at least 30 minsremove gas phase CO and physisorbed CO

Take spectra of 2%CO adsorption Take spectra o 2% oxidation

X-ray photoelectron spectroscopy was performed on a VG ESCALAB XPS, ESCA MK

II using Mg Kα (1254.6 eV) source under UHV better than 3 × 10-9 torr XPS spectrawere recorded at  = 90° for the X-ray sources The in-situ XPS experiments were

performed in a UHV chamber at the SINS beamline in the Singapore synchrotron lightsource (SSLS) at National University of Singapore.10 XPS spectra were measured using ahemispherical electron energy analyzer (EA 125, Omicron NanoTechnology GmbH) TheXPS experiments were done at normal photoelectron emission conFigureuration, with thephoton energy resolution of 0.5 eV XPS measurements were done at constant pass

energy mode Table 5.2 summarizes the experimental procedure for CO oxidation in-situ

XPS study The same scan time on each sample were maintained, because in a Au/CuOsystem, not only might the gold species be reduced under x-ray, but the CuO supportmight also be reduced under x-ray scan

Table 5.2 Experimental procedure for CO oxidation in-situ XPS study

CO + O2does was pumped out and sample was outgas for 1 hour then

transferred back to analysis chamber



Scan for C1s, O1s, Cu 3p and Au4f

5.3.1 Characterization of the Au/CuO catalysts

Figure 5.1 shows the SEM micrograph of three kinds of CuO and three kinds of Au/CuO

samples after pre-treatment at 300oC in air for 1 hour No obvious change in sample’smorphology was observed for these three kinds of CuO samples after gold deposition

Figure 5.1 SEM micrograph for six samples.

5.1(A) CuO(CB) pre-treated in air for 1 hour at 300oC

5.1(B) CuO(NP) pre-treated in air for 1 hour at 300oC

5.1(C) CuO(NF) pre-treated in air for 1 hour at 300oC

5.1(D) Au/CuO (CB) pre-treated in air for 1 hour at 300oC

5.1(E) Au/CuO(NP) pre-treated in air for 1 hour at 300oC

5.1(F) Au/CuO(NF) pre-treated in air for 1 hour at 300oC

Figure 5.2 displays XRD patterns of three kinds of Au/CuO samples after heating in air

at 300oC for 1 hour All three CuO supports show crystalline structure of monoclinicTenorite CuO The Au <111> diffraction at 38.2 º is not detected because of itsoverlapping with strong (111) (200) reflections of CuO Gold signals may be too weakdue to the small crystallite size of gold and the low concentration of gold

A

E

C B

10 20 30 40 50 60 70 80 3000

3500 4000 4500 5000 5500

a:Au/CuO (CB) b:Au/CuO (NF) c:Au/CuO (NP)



a b c

Figure 5.2 XRD data for three kinds of Au/CuO catalysts.

Figure 5.3 shows the TEM images of the three kinds of Au/CuO samples heated in air

for 1 hour at 300oC Clearly Au particles are well dispersed in a few nanometer ranges

F E

D

Figure 5.3 TEM micrograph for gold supported on three kinds of CuO supports

5.3 (A) (D) TEM for Au/CuO (CB) sample pre-treated in air for 1 hour at 300oC

5.3 (B) (E) TEM for Au/CuO (NP) sample pre-treated in air for 1 hour at 300oC

5.3 (C) (F) TEM for AuCuO (NF) sample pre-treated in air for 1 hour at 300oC

The size distribution of Au nanoparticles is shown in Figure 5.4 for the three kinds of

Au/CuO samples after pre-treated at 300oC in air for 1 hour Detailed information for

XRF and BET results of gold supported CuO samples are listed in Table 5.3 The gold wt%

content of these Au/CuO samples were all around 2.2-2.0 % according to the x-rayfluorescence (XRF) results It is noticed that the surface area of samples before and afterpre-treatement in air at 200oC (or 300oC) for 1 hour are not distinct

Figure 5.4 Bar graph of three kinds of Au/CuO samples

Table 5.3 Au atom% in three kinds of Au/CuO samples from XRF,

BET results of three kinds of CuO, and three kinds of Au/CuO samples.

CuO (CB)

Au/CuO (CB)

CuO (NP)

Au/CuO (NP)

CuO (NF)

Au/CuO (NF) XRF

H2 temperature programmed reduction (TPR) was conducted to investigate the effect ofsurface lattice oxygen and adsorbed oxygen species on the CO oxidation reaction TPR

profiles of CuO (CB), CuO (NP), Au/CuO (CB) and Au/CuO (NP) are shown in Figure 5.5 All the four samples are basically stable between 25-200oC which is our reactiontemperature range CuO(NP) (curve b) is very difficult to reduce while Au/CuO (NP)sample is the most active among these four samples The area of the TPR peak of CuO(CB) is much smaller than that of other three samples, suggesting smaller amount ofmovable oxide ions The presence of Au greatly enhances the reducibility of CuO(CB).Under similar experimental conditions (amount of the sample etc) the peak area of theAu/CuO (CB) catalyst is several times that of CuO(CB) Nevertheless the oxide ions inAu/CuO(CB) is less movable than Au/CuO(NP)

0 50000 100000

150000

a: CuO (CB) b: CuO (NP) c: Au/CuO(CB) d: Au/CuO(NP)

c

d

Figure 5.5 TPR profiles of CuO (CB), CuO (NP), Au/CuO (CB) and Au/CuO (NP)

(a) CuO (CB); (b) CuO (NP); (c) Au/CuO (B); (d) Au/CuO (NP)

5.3.2 Catalytic study of CO oxidation reaction over Au/CuO catalyst

A few groups have studied the carbon monoxide oxidation behavior over powderedcopper oxide.11-14 It is generally agreed that CO oxidation over copper oxide catalystsobeys a Mars-van-Krenvelen redox mechanism Carbon monoxide reacts with surfacelattice oxygen to release carbon dioxide and leaves a surface oxygen vacancy on themetal oxide surface This is the step that determines reaction rate Molecular oxygen thenreacts with surface oxygen vacancy to form surface lattice oxygen.12Before the reaction,all our three samples were black/dark brown in color During the reaction, thetemperature was increased from room temperature to 350oC at steps of 50oC per testing,and held at every testing for at least half an hour, in the end held at 350oC for 2 hours.After the reaction, the sample became red color copper metal This may imply that underthe oxygen-lean conditions CuO can be reduced to lower oxidation states from II to 0.

Cu0 species was not able to activate CO oxidation at reaction temperatures 25-350oC Inorder to keep the catalyst active for CO oxidation, the CO : O2 ratio in feed gas, thereaction temperature and the space velocity of the feed gas were optimized to increase thelife span of gold on copper oxide catalysts and at the same time, increase their catalyticactivities In this chapter, our interests are gold catalyst supported on CuO COoxidation reactions over CuO without the presence of Au were studied as reference forcomparing the differences between activities of CuO and Au/CuO samples Therefore,the oxidation reaction conditions were optimized for the CO oxidation over Au/CuO,rather than over Copper oxide without Au

Figure 5.6 shows the conversion of CO over CuO (CB), CuO (NF), CuO (NP), Au/CuO

(CB), Au/CuO (NF) and Au/CuO (NP), as a function of reaction temperature afterpretreatment in air for 1 hour at 200 oC Without the presence of Au, CuO can still

catalyze the CO oxidation Among the three kinds of copper oxide, CuO (NP) is the bestcatalyst, reaching 100% CO conversion at 140oC (d curve) CuO (CB) is the poorestachieving 100% CO oxidation at 220oC (f curve) Nano-gold on copper oxide samplesshow much better CO oxidation activities than the copper oxide samples Au/CuO (NP) isthe best CO oxidation catalyst among all these samples It reaches 50% CO conversion at

75oC, and 100% at 100oC Au/CuO(CB) sample achieved 100% CO conversion at c.a

130oC Note that the catalytic activity of both Au/CuO and CuO is related to the specificsurface area, higher surface area corresponding to higher activity [a(28.3

m2/g)>b(17.8m2/g)>c(7.7m2/g); d(27.6)>e(17.8)>F(7.9)] All these samples are poor incatalytic activities at temperature below 50oC

0 20 40 60 80 100 120

Figure 5.6 Conversion of CO as a function of reaction temperature over the three kinds of CuO samples and the three kinds Au/CuO samples Reaction conditions: 5%CO+5%O 2 in He, GHSV: 60,000 cm 3 g -1 h -1 Sample pre-treated at 200 o C in air for 1 hour.

(a) Au/CuO (NP) (b) Au/CuO (NF) (c) Au/CuO (CB)

In order to study the effect of pre-treatment temperature on the catalytic activity Au/CuOsamples, the above three Au/CuO samples were pre-treated in air for 1 hour at 200, 300and 400oC individually Their CO oxidation activities are displayed in Figure 5.7.

Figure 5.7 Conversion of CO as a function of reaction temperature over three kinds Au/CuO

samples at different temperature.

a, d, g: Au/CuO (NP) pre-treated at 200 o C 300 o C and 400 o C;

b, e, h: Au/CuO (NF) pre-treated at 200 o C 300 o C and 400 o C;

c, f, i: Au/CuO (CB) pre-treated at 200 o C 300 o C and 400 o C;

Reaction conditions: 5%CO+5%O 2 in He, GHSV: 60,000 cm 3 g -1 h -1 .

For each Au/CuO sample, the trend of increasing pre-treating temperature is the same, i.e

A300oC> A200oC > A400oC.(A=activity and d>a>g ; e>b>h ; f>c>i) The samples that were

those that were pre-treated in air at 200 oC and 400 oC The Au/CuO (NP) samplepretreated at 300oC reaches 100% CO conversion at 50oC (curve d in Figure 5.7).

0 20 40 60 80

c

Figure 5.8 Conversion of CO as a function of reaction temperature over three kinds Au/CuO samples Reaction conditions: 5%CO+5%O 2 in He, GHSV: 60,000 cm 3 g -1 h -1 Sample pre-treated at 300 o C in air for 1 hour (a) Au/CuO (NP) (b) Au/CuO (NF) (c) Au/CuO (CB)

The Au/CuO(NF) and Au/CuO (CB) samples show 100% CO conversion at 75 and 90oC

respectively (Figure 5.8) There are two possible explanations for the strong effect of

pre-treatment temperature on the activity of nano-gold supported on copper oxide Thefirst one is related to the use of capping agent in colloid-based preparation of the Au/CuOcatalysts The capping agent, lysine decomposes at ca 250oC Au/CuO samplespretreated at 300oC can completely decompose lysine molecules bonded to Au cluster sothat the reactants CO or O2have higher access to the active sites Hence pretreatment at

300oC can enhance the activity Further increase of the pretreating temperature to 400oCmay result in Au particle sintering or agglomeration The Tammann temperature for bulkgold is around 422oC, which is derived based on the equation: Tammann temperature

(TTammann) equals to 0.52 TF, where TF is metal’s absolute melting temperature Themelting point of bulk Au is 1064oC,15 while the Tammann temperature of as-preparedgold nano particles can fall between 200-300oC The melting point of small gold particles

is known to decrease with their particle size According to Cortie, measured melting

point for gold nano particles with particle size around 5nm is about 250oC.16 Hence goldparticle sizes would increase with increasing pre-treatment temperature The catalyticactivity of nano-Au is extremely sensitive to the Au particle size At 400oC, althoughgold particles were dispersed quite evenly on the copper oxide surface, gold particlesaggregated to become larger particles, thus samples pre-treated at this temperatureshowed the worst CO oxidation activity among these three sets of catalysts Also, thesurface area of Au/CuO samples calcined at 400oC is lower than those calcined at 200oCand 300oC。

The results might indicate that the size of the gold particle has greater influence over COoxidation activity than the dispersion of the gold particles Long-term CO oxidationactivity tests for three kinds of Au/CuO samples pre-treated at 300oC in air for 1 hourwere also conducted All samples were able to maintain their CO oxidation activities atthe temperature where they hit 100% conversion of CO for 72 hours

Kinetics of CO Oxidation.

Figure 5.8 shows that the catalytic activity of the Au/CuO (NP) is much higher than that

of the Au/CuO (CB) in the temperature range of 25-100 oC For Au/CuO (NP), theconversion of CO was ca 50% at 37oC (T50) and reached 100% at 50oC In contrast, in

the Au/CuO (CB) sample, T50 was detected at 70oC, and 100% conversion was obtained

at 80oC The Arrhenius plot of CO oxidation rate (Figure 5.9) yields apparent activation

energy (Ea) of 3.8 kJ/mol for Au/CuO (NP), which is relatively low compared to 5.8kJ/mol for Au/CuO (CB) The published kinetic data for CO oxidation on various Au

supported catalysts together with our results are summarized in Table 5.4, including the

intrinsic reactivity around room temperature in terms of the mass-specific reaction rates[mmol/gAu·s], the turnover frequencies (TOFs) and apparent activation energies The

apparent activation energy is as seen in Figure 5.9.

-3 -2 -1 0 1

CB CB DP DP DP DP CVD CVD PD

298 298 263 263 343 343 273 273 295

5.8 3.8 17 32

0.07 0.22 2.2 x 10 -3 6.0 x 10 -3 9.4 x 10 -2 - 6

0.04 0.12 - - - -

2 10 -2 0.02 0.12

this study this study 17,18 19 20 20 21 9 9

Preparation methods: CB, colloid - based method; DP, deposition participation; CVD, chemical vapor deposition; PD, photodepostion, IMP: impregnation with HAuCl 4

The reaction rate for Au/CuO (NP) equals to 0.22 mmol/gAu·s, and 0.07 mmol/gAu·s forAu/CuO (CB) The 0.12 s-1 of TOF at 25oC (298 K) for Au/CuO (NP) is about threetimes that of Au/CuO (CB) The activation energies of Au/CuO (CB) and Au/CuO (NP)samples are also lower than the values obtained from the Al2O3 results, and the reactionrate is faster than CeO2, Al2O3 and ZrO2 supported Au catalysts prepared by othergroups.17-22

Selective oxidation of carbon monoxide in hydrogen

Figure 5.10 -5.13 show the CO conversion and O2 selectivity for selective oxidation ofcarbon monoxide in hydrogen over the three kinds of Au/CuO samples These sets ofAu/CuO samples were pre-treated at 300oC in air for 1 hour, and the reaction gas mixturehad 1%CO, 2% O2 and 70%H2in He, with a GHSV of 60,000 cm-3g-1h-1 Au/CuO (NP)still exhibited the best CO conversion and selectivity to CO2among the three samples At

80oC, this sample attained 86% CO conversion and 39% selectivity to CO2 As forAu/CuO (NF) sample, the conversion of carbon monoxide was 71% at 95oC and theselectivity to CO2at this temperature was 18.1% The Au/CuO (CB) sample reached 18%selectivity to CO2 at 125oC when the CO conversion was 91% The CO conversions ofthese three kinds of Au/CuO samples with H2presence were worse than those without H2presence, which may be related to the high reducibility of CuO CuO(CB) has lessmovable oxide ions, and hence Au/CuO(CB) can achieve 91% conversion though highertemperature (130oC) is required

20 40 60 80 100 120 0

20 40 60 80

a:Au/CuO(NP) b:Au/CuO(NF) c:Au/CuO (CB)

0 20 40 60 80

Figure 5.11 CO conversion and O 2 selectivity as a function of reaction temperature over Au/CuO (NP) samples Reaction conditions: 1%CO+2%O 2 +70%H 2 in He, GHSV: 60,000 cm 3 g -1 h -1 Sample pre- treated at 300 o C in air for 1 hour.

20 40 60 80 100 120 0

20 40 60

0 20 40 60 80