Available online at www.sciencedirect.com

International Journal of Hydrogen Energy 29 (2004) 429–435www.elsevier.com/locate/ijhydene

Selective catalytic oxidation of CO in the presence of H2 overgold catalyst

Apanee Luengnaruemitchaia, Somchai Osuwana ;∗, Erdogan GularibaThe Petroleum and Petrochemical College, Chulalongkorn University, Soi Chuia 12, Phyathai Road, Patumwan, Bangkok 10330,

ThailandbDepartment of Chemical Engineering, The University of Michigan, Ann Arbor, MI 48109, USA

Received 19 September 2003; received in revised form 25 September 2003

Abstract

The proton exchange membrane fuel cells with potentially much higher e2ciencies and almost zero emissions o4er anattractive alternative to the internal combustion engines for automotive applications. A critical issue in fuel cell system is thesmall (∼ 0:5%) amount of CO present from output of fuel reformer. This amount of CO can deteriorate the performanceof fuel cells. Consequently, an additional gas conditioning process is required to minimize CO content in reformed gas. Theproposed research study focuses on preferential oxidation of CO in a simulated reformed gas to CO2 by using selective COoxidation catalysts. In this work, the e4ects of preparation method, O2, water vapor, and CO2 concentration in feed streamon the selective CO oxidation over Au=CeO2 catalysts were investigated in the temperature range of 323–463 K. Catalyticstability test was also performed.We ;nd that the activity of Au catalyst depends very strongly upon the preparation method, with co-precipitation prepared

Au=CeO2 catalyst exhibiting the highest activities.? 2003 International Association for Hydrogen Energy. Published by Elsevier Ltd. All rights reserved.

Keywords: CO oxidation; Gold catalyst; Ceria; Co-precipitation; Fuel cell

1. Introduction

In the past, the world has widely relied on the energyfrom burning fossil fuels, which are expected to be de-pleted in the near future. Extensive uses of fossil fuels havecreated air pollution, acid rain, and the greenhouse e4ect.Scientists have been searching for alternative means ofpowering vehicles more e2ciently without creating any pol-lution. The proton exchange membrane (PEM) fuel cell hasbeen attracting much attention in improving fuel e2ciencyand cleanliness of automobiles [1]. This is due to low tem-perature of operation, fast cold start, perfect CO2 toleranceby the electrolyte and a combination of high power den-sity and high energy conversion e2ciency. Normally, pure

∗ Corresponding author. Tel.: +662-2184-141;fax: +662-611-7619.E-mail address: somchai.o@chula.ac.th (S. Osuwan).

hydrogen is the ideal fuel for the fuel cell system since itsimpli;es system integration, maximizes system e2ciency,and provides zero emission. However, since no e4ective andpotentially safe means of storing adequate amount of hy-drogen in a vehicle exists, on-board hydrogen generation bysteam reforming of methanol or partial oxidation of liquidhydrocarbons followed by water gas shift reaction are alter-native ways of giving a fuel cell powered vehicle adequaterange. These technologies not only produce hydrogen as amain product but also has side products with a small amountof 1% CO. Many studies have shown the negative e4ect ofCO on the performance of PEM fuel cell since the electrodesof the fuel cell are typically made of platinum which is verysensitive to CO poisoning [2,3]. Selective catalytic CO ox-idation is perhaps the most promising method of reducingCO in the feed to a 10 ppm or lower range with a minimalloss of hydrogen.The preferential catalytic oxidation of CO was ;rst

studied by Oh and Sinkevitch [4] with alumina

0360-3199/$ 30.00 ? 2003 International Association for Hydrogen Energy. Published by Elsevier Ltd. All rights reserved.doi:10.1016/j.ijhydene.2003.10.005

430 A. Luengnaruemitchai et al. / International Journal of Hydrogen Energy 29 (2004) 429–435

supported ruthenium (Ru), rhodium (Rh) and platinum(Pt). Ru and Rh were very selective compared to Pt foroxidizing CO in the presence of H2 diluted with N2 at403 K. However, a gas composition of 900 ppm CO,800 ppm O2 and 0.85% H2 in N2 background was verydi4erent from the reformed gas that is generated by re-forming a liquid hydrocarbon or methanol. The conceptof using a zeolite support on Pt catalysts was appliedfor this reaction by Watanabe et al. [5] and Igarashiet al. [6].Kinetic studies of selective CO oxidation on Pt=Al2O3

and Au=Fe2O3 catalysts in a simulated reformed gas (75%H2, the rest is N2) over a wide range of CO concentrations(0.02–1.5%) were reported by Kahlich et al. [7,8]. In ad-dition to the noble metal catalysts other catalysts have alsobeen investigated such as CoO catalyst [9], and Pt promoteddeposited on a monolith catalyst [10]. The supported Au cat-alysts have been found to be very active for CO oxidationreaction [11–13]. Recently, the metal oxide supports suchas Al2O3 [14], MnOx [15], Fe2O3 [16], MgO–Al2O3 [17]were found to promote the activity of Au on selective COoxidation and the catalytic combustion of volatile organiccompounds [18].Among the di4erent metal oxides used as the sup-

port of Au for CO oxidation, cerium oxide or ce-ria (CeO2) is one of the interesting metal oxides.CeO2 is the oxide of the rare-earth metal cerium,which may exist in several compositions, due to thecapacity of Ce to switch between the two oxida-tion states of Ce3+ and Ce4+. A number of functionshave been ascribed to ceria, including promoting wa-ter gas shift activity [19–21], maintaining the disper-sion of the catalytic metals [2,22] and stabilizing thesurface area of the support [23]. Extensive research,in particular during the past decade, has shown thatceria has bene;cial e4ects on a number of catalyticreactions.Due to a facile redox reaction cycle, ceria exhibits oxy-

gen storage capacity and improves the CO and hydrocarbonoxidation. Summers and Ausen [24] claimed that ceria do-nated oxygen to Pt in their study of the oxidation of CO onPt=CeO2 catalyst. For CO oxidation, ceria has been foundto lower the activation energy, increase the reaction rate andsuppress the usual CO inhibition e4ect [25]. The percent-age loading of Ce has been shown to a4ect CO oxidation onPt [26]. Ceria promotes the activity of Pt in both lean andrich reactant gases for CO oxidation [27]. Consequently, wechose CeO2 as the support for studying selective catalyticoxidation of CO in the presence of H2 based on the afore-mentioned properties of ceria.From preliminary results, we found that Au=CeO2 cata-

lysts also work very well for the selective CO oxidation.Therefore, the objectives of this work are to optimize theoperating conditions for preferential oxidation of CO by O2in the presence of H2 over the temperature range of 323–463 K, which are suitable for fuel cell applications, and to

study the e4ect of preparation method on the catalytic ac-tivity of Au=CeO2 catalysts.

2. Experimental

2.1. Catalyst preparation

Three methods of the catalyst preparation; impregnation,co-precipitation and sol–gel were used in this work.Impregnation method: The Au catalysts were obtained

by impregnation of the commercial ceria support with anaqueous solution of HAuCl4:xH2O (Alfa AESAR) contain-ing the appropriate amount of Au. The catalysts were driedovernight at 383 K and calcined at 773 K for 5 h.Co-precipitation method: An aqueous solution of

Na2CO3 (1 M) was added into an aqueous mixture ofHAuCl4:xH2O and Ce(NO3)3:6H2O (Fluka) and was keptat room temperature and constant pH of 8.0. The precipitatewas aged for an hour and then was washed several timeswith distilled water until there was no excess anions. Af-ter washing with deionized water, the catalysts were driedovernight at 383 K and calcined at 773 K for 5 h.Sol–gel method: The single step sol–gel catalysts were

prepared by hydrolyzing a solution of Ce acetate (AlfaAESAR) and HAuCl4:xH2O with NH4OH. The reactionmixture was kept at 353 K while the pH was maintainedbetween 9.0 and 9.5. Then, HNO3 was added until gelation,the catalysts were dried overnight at 383 K and calcined at773 K for 5 h.

2.2. Catalyst characterization

Powder X-ray di4raction (XRD) patterns were used toobtain information about the structure and composition ofcrystalline materials. A Rigaku X-ray di4ractometer systemequipped with a graphite monochromator and a Cu tube forgenerating a CuK radiation was used to obtain the XRDpatterns.The Brunauer–Emmett–Teller (BET) method was used to

determine the surface area and pore size of the catalysts byN2 adsorption/desorption at 77 K. Speci;c surface area wasdetermined using an Autosorb-1 surface area analyzer.The particle morphology of the catalysts was observed

by a Scanning Electron Microscope (SEM) using JEOLJSM -5410 LV scanning microscope operated at 15 kV. TheTransmission Electron Microscopy (TEM) images were at-tained using a JEM 2010 operating at 200 kV in bright anddark ;eld modes. Crystallinity and crystal structure of thesamples were evaluated from the selected area of electrondi4raction patterns.

2.3. Activity measurement

Catalytic activity studies were conducted in a pyrexglass U-tube reactor having an internal diameter of 6 mm

A. Luengnaruemitchai et al. / International Journal of Hydrogen Energy 29 (2004) 429–435 431

at atmospheric pressure. One hundred mg of catalyst waspacked between glass wool plugs in the middle of the re-actor. The reactant gas typically consisted of 1% CO, 0.5–2% O2; 2% CO2; 2:6% H2 O; 40% H2 and helium. Thespace velocity was 30; 000 ml g−1 h−1. The reactant gaswas passed through ;lters to remove particles and a checkvalve to prevent a reverse Now. The individual stream Nowrate was controlled by mass Now controllers to achieve thedesired composition.The reactant and product were analyzed by an on-line gas

chromatograph, (Hewlett Packard 5890 Series II equippedwith a thermal conductivity detector). The column utilized inthe chromatograph was a Carbosphere, 80/100 mesh, 10 ft×1=8 in stainless-steel packed column.The CO conversion was calculated based on the carbon

dioxide formation. The selectivity of CO oxidation was de-;ned as the oxygen consumption for CO oxidation dividedby the total oxygen consumption. There was no methaneformation observed under reaction conditions performed inthis study.

3. Results and discussion

3.1. Catalyst characterization

Table 1 lists the BET surface areas, ceria crystallite sizesand Au particle sizes of the Au=CeO2 catalysts studied. Thesurface area varies signi;cantly with the catalyst prepara-tion method. The CeO2 crystallite sizes of the catalysts weredetermined from X-ray line-broadening using the DebyeScherrer equation and the results are also shown in Table 1.The sol–gel method gave the largest CeO2 crystallite sizewhile the co-precipitation method gave the smallest CeO2crystallite size on Au catalysts. The sizes of CeO2 crystallitesare inversely correlated with the surface areas of the cata-lysts. The SEM pictures of Au=CeO2 catalysts for impreg-nation, co-precipitation and sol–gel are given in Fig. 1. Thetypical XRD patterns of the CeO2 in the Au=CeO2 samplesprepared by three di4erent methods are shown in Fig. 2. It isapparent that the patterns consist of eight main reNections,typical of a cubic, Nuorite structure of CeO2 correspondingto (1 1 1); (2 0 0); (2 2 0); (3 1 1); (2 2 2), and (4 0 0) planes.XRD patterns of Au=CeO2 co-precipitation catalyst showed

Table 1Catalyst surface area, ceria crystallite size and Au particle size

Catalyst Preparation method BET surface area (m2=g) Ceria crystallite size (nm) Au particle size (nm)

XRD TEM

1% Au=CeO2 Impregnation 104.2 17.4 29 a

Co-precipitation 124.1 14.5 ¡ 5 2–8Sol–gel 66.7 38.6 30 a

aNot measured.

no metal peaks indicating that the metallic particle size ofthe catalyst might be smaller than 5 nm. This observationwas also con;rmed by TEM investigations. The evidence ofAu in metallic form on impregnation and sol–gel catalystswas observed at 2� = 38:2◦ and the average Au crystallitesizes were 29 and 30 nm, respectively.The TEM picture of 1% Au=CeO2 co-precipitation cata-

lyst shown in the Fig. 3(a) reveals the presence of the goldparticles as dark spots in the catalyst. The existence of Auparticles on CeO2 support was veri;ed by using EDS fo-cusing on the regions containing highly contrast spots underTEM. For 1% Au=CeO2 co-precipitation catalyst, the Auparticles were seen as highly contrast spots both on and in-side the support and the average particle size of Au is about4:5 nm. Fig. 3(b) shows the Au particle size distribution ofco-precipitation catalyst.

3.2. Catalytic activity

3.2.1. Temperature dependenceFig. 4 illustrates the e4ect of catalyst preparation method

on the catalytic activity of 1% Au=CeO2. The Au=CeO2co-precipitation catalyst has the best performance among thethree methods of catalyst preparation. The CO conversionshowed a peak on all of the catalysts. The temperature atwhich the activity of the Au=CeO2 co-precipitation catalystwas maximumwas 383 K. The reaction at 383 K is the mostsuitable temperature for the fuel cell operating temperaturelevel (353–393 K). At higher temperature, there is a con-tinuous decrease in selectivity, indicating higher activationenergy for the H2 oxidation than for the CO oxidation. Thelow temperature behavior of the selectivity is in agreementwith the studies of Igarashi et al. [6] and Kahlich et al. [7].The morphologies of Au=CeO2 co-precipitation cat-

alyst by SEM and XRD are very di4erent from thoseof impregnation and sol–gel catalysts. It was foundthat the co-precipitation catalyst has a crystallite sizeof about 4:5 nm which is much smaller than those pre-pared by impregnation and sol–gel methods which haveAu crystallite sizes of 29 and 30 nm, respectively. Theresults obtained with the Au=CeO2 catalyst prepared byco-precipitation method were similar to those observed withthe co-precipitation catalyst reported by Haruta et al. [12]

432 A. Luengnaruemitchai et al. / International Journal of Hydrogen Energy 29 (2004) 429–435

Fig. 1. SEM pictures of the Au=CeO2 catalysts: (a) impregnation;(b) co-precipitation; (c) sol–gel.

and ScirPe et al. [18]. Besides, from SEM pictures, we ob-served only crystalline ceria on Au=CeO2 co-precipitationcatalyst which is also con;rmed by the XRD results. It isthen concluded that the Au=CeO2 co-precipitation catalyst

Fig. 2. XRD patterns of the Au=CeO2 catalysts.

exhibits much higher activity than other two catalysts due tothe aforementioned characteristics. It has been reported thatthe high activity of Au=CeO2 catalyst might be related tothe capacity of Au nanoparticles to weaken the Ce–O bond.Thus, the mobility/reactivity of the surface lattice oxygen isincreased [18].The performance of Au catalysts in this work can

be compared with Au=MnOx and Au=Fe2O3 catalystsavailable in the literature. Kahlich et al. [8] and Tor-res Sanchez et al. [15] reported that the high activityand selectivity at the considerably lower temperatureof 393 and 353 K were observed over Au=MnOx andAu=Fe2O3 catalysts, respectively. However, their reac-tant gas mixture used containing CO, O2 and H2 was nota typical simulated reformed gas. In addition, Au=CeO2co-precipitation catalyst is most active at 383 K whichis suitable for the system connected directly to PEMfuel cells. This catalyst shows much higher activity andselectivity when compared to the Ag-based compositeoxide catalysts reported by GQuldQur and BalikRci [28]. How-ever, recent work by Avgouropoulos et al. [29] at similarreaction conditions shows that CuO–CeO2 sol–gel cat-alyst exhibited very high activity with selectivity closeto 90%.

3.2.2. E8ects of O2 and water vapor concentrationThe CO conversion and selectivity as a function of O2

concentration in the feed, for the Au=CeO2 co-precipitationcatalyst is shown in Fig. 5. As expected, the CO conver-sion increases with increasing O2 concentration while theCO selectivity decreases with increasing O2 concentration.With 0.5%, 1% and 2% O2 concentration the maxima in

A. Luengnaruemitchai et al. / International Journal of Hydrogen Energy 29 (2004) 429–435 433

Fig. 3. (a) TEM picture of gold particles (dark spots) on theAu=CeO2 co-precipitation surface; (b) Gold particle size distribu-tion determined by TEM.

CO conversion are ∼ 70%, 92% and 98% at 383 K, respec-tively. The selectivities at the point of maximum conversionfor 0.5%, 1%, and 2% O2 concentration are 64%, 62% and48%, respectively. The optimal O2 concentration needed foroxidizing 1% CO in the feed is 1% with high selectivity andminimal loss of hydrogen.In general, it is believed that the catalytic activity of

the catalyst is liable to be suppressed in the presence ofwater vapor. The e4ect of water vapor in the feed streamon the CO oxidation activity and selectivity of Au=CeO2co-precipitation catalyst is shown in Fig. 6. The feedgas mixture was humidi;ed by bubbling through a con-tainer of temperature controlled water, yielding 10% watervapor in the reactant gas. Under the humidi;ed condition,water lowered CO conversion in the region of temperaturelower than 373 K compared to the unhumidi;ed condition.

Fig. 4. Temperature dependence of the CO conversion of the 1%Au=CeO2 catalysts. Reactant composition: 1% CO, 1% O2, 2.6%H2O, 2% CO2, 40% H2 and helium: (c) co-precipitation; (•)impregnation; (�) sol–gel; - - - - -, selectivity; —, CO conversion.

Fig. 5. E4ect of O2 concentration in reactant gas over 1% Au=CeO2co-precipitation catalyst. Reactant composition: 1% CO, 0.5–2%O2, 2.6% H2O, 2% CO2, 40% H2 and helium: (•) 0.5% O2; (c)1% O2; (�) 2% O2; - - - - -, selectivity; —, CO conversion.

This is due to the strong adsorption of water on the activesite [30]. However, at high temperatures water seemed to beslightly favorable to the catalyst activity since it provides

434 A. Luengnaruemitchai et al. / International Journal of Hydrogen Energy 29 (2004) 429–435

Fig. 6. E4ect of water vapor addition in reactant gas over 1%Au=CeO2 co-precipitation catalyst. Reactant composition: 1% CO,1% O2, 0–10% H2O, 2% CO2, 40% H2 and helium: (•) w/owater; (c) 10% water; - - - - -, selectivity; —, CO conversion.

a hydroxyl group which is necessary for reaction to takeplace [12,31,32]. The selectivity of the catalyst was slightlya4ected by the presence of water vapor which is similar toCuO–CeO2 catalyst [29].

3.2.3. E8ect of CO2 concentrationFig. 7 shows the e4ect of CO2 addition to the feed stream

on the activity of the Au=CeO2 co-precipitation catalyst. In-creasing CO2 concentration from 2% to 20% reduced theactivity of the Au=CeO2 catalyst. The maximum in conver-sion dropped from ∼ 92% to ∼ 85% with a shift of 30 K tohigher temperatures. The selectivity of the catalyst was notsigni;cantly impacted by the presence of CO2 in the feed.

3.2.4. Deactivation testCatalytic stability of Au=CeO2 co-precipitation catalyst

was tested at the temperature of 383 K. As can be seen fromFig. 8, the CO conversion and selectivity were maintainedfor 2 days with very slight loss of activity. The result showsgood stability of the Au=CeO2 co-precipitation catalyst ascompared with Au=Fe2O3; CuO–CeO2 and Au=TiO2 cata-lysts reported by Kahlich et al. [8], Avgouropoulos et al.[29] and Lin et al. [13], respectively.

4. Conclusions

In this study, the catalytic activity of 1% Au=CeO2 cat-alysts prepared by three di4erent methods was investigatedfor the selective CO oxidation over the temperature range

Fig. 7. E4ect of CO2 concentration in reactant gas over 1%Au=CeO2 co-precipitation catalyst. Reactant composition: 1% CO,1% O2, 10% H2O, 2–20% CO2, 40% H2 and helium: (c) 2%CO2; (•) 20% CO2; - - - - -, selectivity; —, CO conversion.

Fig. 8. Deactivation test over 1% Au=CeO2 co-precipitation catalystin a feedstream containing 1% CO, 1% O2, 2% CO2, 2.6% H2O,40% H2, and helium: (•) CO conversion; (c) selectivity.

of 323–463 K. We found that 1% Au=CeO2 co-precipitationcatalyst is the most active, exhibiting high activity and goodselectivity at 383 K. The presence of water vapor in the feedstream lowered the CO conversion only in the lower tem-perature region. Increasing CO2 concentration in the feed

A. Luengnaruemitchai et al. / International Journal of Hydrogen Energy 29 (2004) 429–435 435

stream signi;cantly reduced the CO conversion. However,both water vapor and CO2 showed no signi;cant e4ect onthe CO selectivity. The optimal amount of O2 required foroxidizing 1% CO in the feed is 1%. The catalyst was foundto be quite stable in the deactivation test.

Acknowledgements

The authors gratefully acknowledge both fellowship sup-port and research grant of Ms. Apanee Luengnaruemitchaiby the Royal Golden Jubilee Ph.D. Project, ThailandResearch Fund (TRF), Thailand.

References

[1] Gottesfeld S, Pa4ord J. A new approach to the problem ofcarbon monoxide poisoning in fuel cells operating at lowtemperatures. J Electrochem Soc 1988;135(10):2651–2.

[2] GQotz M, Wendt H. Binary and ternary anode catalystformulations including the elements W, Sn and Mofor PEMFCs operated on methanol or reformate gas.Electrochimica Acta 1998;43(24):3637–44.

[3] Rohland B, Plzak V. The PEMFC-integrated CO oxidation—a novel method of simplifying the fuel cell plant. J PowerSources 1999;84(2):183–6.

[4] Oh SH, Sinkevitch RM. Carbon monoxide removal fromhydrogen-rich fuel cell feedstreams by selective catalyticoxidation. J Catal 1993;142:254–62.

[5] Watanabe M, Uchida H, Igarashi H, Suzuki M. Pt catalystsupported on zeolite for selective oxidation of CO in reformedgases. Chem Lett 1995;1:21–2.

[6] Igarashi H, Uchida H, Suzuki M, Sasaki Y, Watanabe M.Removal of CO from H2-rich fuels by selective oxidation overplatinum catalyst supported on zeolite. Appl Catal A: Gen1997;159:159–69.

[7] Kahlich MJ, Gasteiger HA, Behm RJ. Kinetics of theselective CO oxidation in H2-rich gas on Pt=Al2O3. J Catal1997;171(1):93–105.

[8] Kahlich MJ, Gasteiger HA, Behm RJ. Kinetics of the selectivelow-temperature oxidation of CO in H2-rich gas over Au=�–Fe2O3. J Catal 1999;182(2):430–40.

[9] Teng Y, Sakurai H, Ueda A, Kobayashi T. Oxidative removalof CO contained in hydrogen by using metal oxide catalysts.Int J Hydrogen Energ 1999;24:355–8.

[10] Korotkikh O, Farrauto R. Selective catalytic oxidation of COin H2: fuel cell applications. Catal Today 2000;62:249–54.

[11] Haruta M, Yamada N, Kobayashi T, Iijima S. Gold catalystsprepared by coprecipitation for low-temperature oxidationof hydrogen and of carbon monoxide. J Catal 1989;115:301–9.

[12] Haruta M, Tsubota S, Kobayashi T, Kageyama H, GenetMJ, Delmon B. Low-temperature oxidation of CO overgold supported on TiO2; �–Fe2O3 and Co3O4. J Catal1993;144:175–92.

[13] Lin SD, Bollinger M, Vannice MA. Low-temperature COoxidation over Au=TiO2 and Au=SiO2 catalysts. Catal Lett1993;17(3–4):245–62.

[14] Bethke GK, Kung HH. Selective CO oxidation in ahydrogen-rich stream over Au=�–Al2O3 catalysts. Appl CatalA: Gen 2000;194:43–53.

[15] Torres Sanchez RM, Ueda A, Tanaka K, Haruta M. Selectiveoxidation of CO in hydrogen over gold supported onmanganese oxides source. J Catal 1997;168(1):125–9.

[16] Guczi L, Horvath D, Paszti Z, Peto G. E4ect of treatmentson gold nanoparticles, relation between morphology, electronstructure and catalytic activity in CO oxidation. Catal Today2002;72(1–2):101–5.

[17] Grisel RJH, Weststrate CJ, Goossens A, Craje MWJ,van der Kraan AM, Nieuwenhuys BE. Oxidation ofCO over Au=MnOx=Al2O3 multi-component catalysts ina hydrogen-rich environment. Catal Today 2002;72(1–2):123–32.

[18] ScirPe S, MinicPo S, Crisafulli C, Satriano C, Pistone A.Catalytic combustion of volatile organic compounds ongold/cerium catalysts. Appl Catal B: Environ 2003;40:43–9.

[19] Barbier Jr J, Duprez D. Steam e4ects in 3-way catalysis. ApplCatal B: Environ 1994;4(2–3):105–40.

[20] Bunluesin T, Gorte RJ, Graham GW. Studies of thewater-gas-shift reaction on ceria-supported Pt, Pd, and Rh:implications for oxygen-storage properties. Appl Catal B:Environ 1998;15:107–14.

[21] Whittington BI, Jiang CJ, Trimm DL. Vehicle exhaustcatalysis: I. The relative importance of catalytic oxidation,steam reforming and water-gas shift reactions. Catal Today1995;26:41–5.

[22] Diwell AF, Rajaram RR, Shaw HA, Truex TJ. Role of ceriain three-way catalysts. Stud Surf Sci Catal 1991;71:139–48.

[23] Ozawa O, Kimura M. E4ect of cerium addition on thethermal stability of gamma alumina support. J Mater Sci Lett1990;9(3):291–3.

[24] Summers JC, Ausen SA. Interaction of cerium oxide withnoble metals. J Catal 1979;58:131–43.

[25] Shalabi MA, Harji BH, Kenney CN. Kinetic modeling ofCO oxidation on Pt=CeO2 in a gradientless reactor. J ChemTechnol Biotechnol 1996;65:317–24.

[26] Tiernan MJ, Finlayson OE. E4ects of ceria on the combustionactivity and surface properties of Pt=Al2O3 catalysts. ApplCatal B: Environ 1998;19:23–35.

[27] Holmgren A, Andersson B, Duprez D. Interaction of CO withPt/ceria catalysts. Appl Catal B: Environ 1999;22(2):215–30.

[28] GQuldQur C, BalikRci F. Selective carbon monoxide overAg-based composite oxides. Int J Hydrogen Energ2002;27:219–24.

[29] Avgouropoulos G, Ioannides T, Papadopoulou Ch, BatistaJ, Hocevar S, Matralis HK. A comparative study of Pt=�–Al2O3; Au=�–Al2O3 and CuO–CeO2 catalysts for theselective oxidation of carbon monoxide in excess hydrogen.Catal Today 2002;75:157–67.

[30] Date M, Ichihashi Y, Yamashita T, Chiorino A, Boccuzzi F,Haruta M. Performance of Au=TiO2 catalyst under ambientconditions. Catal Today 2002;72(1–2):89–94.

[31] Haruta M. Size- and support-dependency in the catalysis ofgold. Catal Today 1997;36:153–66.

[32] Wang GY, Lian HL, Zhang WX, Jiang DZ, Wu TH. Stabilityand deactivation of Au=Fe2O3 catalysts for CO oxidationat ambient temperature and moisture. Kinet Catal 2002;43:433–42.