Indian Journal of ChemiStryVol. 16A, September 1978, pp. 751-753

Carbon Monoxide Oxidation on Nickel ChromiteR. PITCHAI & V. SRINIVASAN

Department of Chemistry, Indian Institute of Technology, Madras 600036

Received 20 December 1977; accepted 14 Apri; 1978

The kinetics of the oxidation of carbon monoxide has been studied with a stoichiometricmixture of CO and O2 at an initial pressure of 50 torr and in the temperature range 180-230°.The rate of the reaction depends only on oxygen partial pressure. In situ electrical conducti-vity and adsorption studies have been carried out to explain the break in kinetic plots. Thecompound nickel chrorntte has been found to be more active than the component oxides andthis may be due to the synergetic effect usually observed in the case of spinels.

MANY oxide catalysts like CuCr20(, LaCo03,

etc., have become potential substitutes forthe costly noble metal catalysts in the

catalytic reactors used in control of automotiveemissions. To optimize catalytic performance ofthese catalysts, a knowledge of the mechanism andthe ki~etic~ of the r.eaction is required. Carbonmonoxide IS the major pollutant of automotiveemissions. In an attempt to contribute to thewealth of literature on the oxidation of carbonmonoxide, we have chosen the chromites of transitionmetals with the general formula MCr.04 (whereM is a divalent transition metal ion like Ni2+ Mn2+Cu2+, ~tc.) as catalysts for the study. Though thes~chromites have been tested for their activity forCO oxidation", no systematic study has been madeto compare the activity and stability of these cata-lysts. In this paper the work on the spinel nickelchromite has been reported.Materials and Methods

NiCr204 has been prepared by firing the co-precipitated hydroxides (at PH 6) of nickel and chro-mium at 950° for 30 hr. The for-mation of singlespinel phase was checked by X-ray method. Theco~ponent oxides have also been prepared fromtheir respective hydroxides in a similar way.

The kin~tics of t~e reaction CO+l02-»C02 has~een .studled .followmg the pressure change withtime III a static reactor with a recirculation pump.The !5ases, oxygen and carbon dioxide, from cylinders(Indian Oxygen Limited), were purified in the usualway. Pure carbon monoxide was prepared by themethod of W einhouse-.

The surface of the catalyst was purified by con-tinuously evacuating the catalyst at a pressure of10-6 tor.r at 400° for 6 hr. The catalyst was thensoaked 111 100 torr of oxygen at the reaction tem-perature for nearly 10 hr. Before the reaction,the physically adsorbed molecules were removedby pumping for 2-3 min. This kind of treatmento! .t~e catalyst was found to give good reprodu-cibility.

The surface area of the catalyst as measured bythe BET (N2) method was found to be 7·8 m2/g.

Electrical conductivity measurements in situ havebeen done using a two-probe cell. Adsorptionmeasurements have been carried out in a volumetricset-up.

Results and Discussion

The kinetics of oxidation of CO by oxygen wasstudied at an initial total pressure of 50 torr of amixture of CO and O~ in the ratio 2: 1 in thetemperature range 180-230°. The total order ofthe reaction and the order with respect to oxygenpartial pressure were found to be the same, viz.0.5, by varying the pressures over a wide range.The rate was, however, found to be independent ofCO partial pressure. The product CO2 does notinhibit the reaction.

CO adsorption experiments have been tried inthe temperature range 25-100° as above this tem-perature CO might stat t reacting with the surfacethough in small amounts. The results showed thefailure of CO to get adsorbed on the surface. CO2

did not adsorb on the surface, even at reactiontemperat meso

Oxygen adsorption was found to be extremelysmall in the reaction temperature range. A three-fold decrease in resistance observed when oxygenwas admitted to a freshly evacuated catalyst con-firmed adsorption of oxygen on the surface(Fig. 1).

The kinetic data were analysed using the rateequations (1) and (2)

Rate=k'Pb. ... (1)

... (2)

where k and k' are the rate constants and b is theadsorption coefficient. Eq. (2) was found to givea better fit for the kinetic data obtained.

The plots of 1frate versus IfPb, at 182° and 196°show distinct breaks (Fig. 2) when the conversionis around 10%. In order to get a possible insightinto what really happens at and after the break

751

6'0

••••~ 5-0

t!~ 4~••••0::C!) '1-0o...J

INDIAN J. CHEM., VOL. 16A, SEPTEMBER 1978

• VACUUM

o OXYGEN

2·5~-,:'i;;-TI-"t?--;~--;:l;----.,!r'-TI----,-t-;---",:;----,r;;--~----'ri1'4 1·7 "8 1'9 2'0 2'1 2·2 2-3 2'4 2'5

1000/T

28

24

2

••.• 16•....c0::- t2

8

4

2'~

Fig. 1 - Arrhenius plots for electrical conduction

2'71/JP;

2

Fig. 2 - Kinetic plots for the oxidation of CO

2'5

point, in s£tu conductivity studies have been madeand the results are discussed below.

Admission of pure CO on the freshly evacuatedcatalyst in the reaction temperature range did notresult in any conductivity change thus indicatingthe absence of simple reduction-oxidation mecha-nism for the reaction. In this connection it maybe mentioned that Cr3+ usually imparts stabi-lity to the catalyst. It has been observed byRennard et al.3 that the reduction of MgCrxFez_x04catalyst becomes difficult as 'x' tends to 2.

When pure CO was admitted over an oxygenpretreated catalyst frem which oxygen has beenpumped out for 2 min., a very small change inconductivity was observed. However, when amixture of CO and O2 in the ratio 2:1 was admittedover the same catalyst, conductivity change couldbe followed for a longer time though the changebecomes small with time. All these observationsindicate that oxygen gas is essential for the oxidationof CO on NiCr204.

752

2·8 2·9 )0

In order to check whether the break point in thekinetic plot disappears by taking a gaseous mixtureof CO, O2 and CO2 with a composition correspondingto a point after the break, kinetics has been followedwith a mixture of CO, O2 and CO2 in the ratio 4:2:1.Rate Eq. (2) was used in this case also and thekinetic plots are shown in Fig. 3. As expectedthere was no break in the kinetic plots. However,the activation energy for the reaction (16·0 kcalmole"] in this case (total pressure SO torr) is nearlyequivalent to the activation energy value (14·9kcal mole-") before the break point in the case ofreaction (total pressure SO torr) with stoichiometricamounts of CO and Oz. This may simply meanthat the situation before break point in the case ofstoichiometric mixture is the same as that achievedwhen CO, 02 and CO2 is taken in the ratio 4:2:1.Since neither CO nor CO2 gets adsorbed on thesurface, it is quite possible that all the observationsreported above have to be attributed only to somechange in the nature of oxygen sp ecies.

PITCHAI & SRINIVASAN: CO OXIDATION ON METAL CHROMITE

0·28 0·29 0·30I//PO

2

0·320·31

Fig. 3 - Kinetic plots for the oxidation of a mixture containing CO, Os and CO. in the ratio 4: 2: 1

Initial ESR experiments showed the absence ofthe formation of paramagentic O2 on the surface.Further studies are in progress to identify thespecies.

The overall mechanism of the reaction maytherefore be written as follows:

slow02--+20:~

lastCO(g)+0:;' --+CO~(:dSI

lastCO~(:d3)--+C02(g) +ne-

The kinetics of the oxidation on NiO (surfacearea=1·5 m2/g) could not be followed below 350°.It is well known that the activity of NiO can bevery much varied by changing the method andtemperature of preparations. The surface area ofNiO has been reported! to fall from 30·5 to 2 m2/gwhen the temperature of preparation has been raisedfrom 500° to 900°. In the case of chromia, thereaction could be followed only above 290°. Itis also known that the activity of chromia for COoxidation and its surface area do not depend verymuch on the temperature of preparations. Themechanism of CO oxidation on these two oxidesis also known to be different from that on nickelchromites-". In order to have a meaningful com-parison of the activities of these oxides one has,therefore, to choose either the percentage conversion

in a fixed period at a common temperature or thetemperature at which a particular conversion isobtained during a fixed period as a measure of theactivity of the catalyst. In this paper the latterhas been chosen since the temperature at whichthese oxides have been found to be active are quitedifferent. Further the temperature of preparationof the catalysts was kept the same in all cases inorder- to ensure that the effect of this parameter iseliminated. It has been found that nickel chromiteis more active than the component oxides. Thiscould be due to the synergetic effect usually observedin the case of spinel compounds.

AcknowledgementOne of the authors (R.P.) wishes to thank the

CSIR, New Delhi, for the award of a junior researchfellowship.

References1. DWYER, FRANCIS G., Cat. Rev., (, (1972), 261.2. WEINHOUSE, S., J. Am. chem Soc., 70 (194S), 442.3. RENNARD, R. ]. & KEHL, W. L., J. Calal., 21 (1971).

282.4. KUTSEVA, L. N., Doklady, Akad. Nauk. SSSR, 138 (1961),

409.5. SULTANOV, M. Yu, RUBINER, 1. A. & l\1YASNIKOV, 1. A.,

Zh. fiz. xu«, 44 (1970), 1S06.6. PIERRE RUE & TEICHNER, S. ]., Bull. Soc. Chin? (Fr.).

(1964), 2797.7. GYULA, G., BELA, J. & LASLO, F., Magy. Kern. Foiy., 77

(1971), 229; Chem, Abstr., 75 SOS16g.

753