A Study on Catalysis by Ferrospinels for Preventing Atmospheric Pollution from Carbon Monoxide

doi:10.4236/ojpc.2011.13017

Paper Menu >>

Journal Menu >>

Open Journal of Physical Chemistry, 2011, 1, 124-130

doi:10.4236/ojpc.2011.13017 Published Online November 2011 (http://www.SciRP.org/journal/ojpc)

A Study on Catalysis by Ferrospinels for Preventing

Atmospheric Pollution from Carbon Monoxide

Chennampilly Ummer Aniz, Thengumthanam Damodaran Radhakrishnan Nair*

School of Chemical Sciences, Payyanur Campus, Kannur University, Kerala, India.

*E-mail: [email protected]

Received July 19, 2011; revised August 22, 2011; accepted October 9, 2011

Abstract

Ferrospinel catalyst samples containing Nickel, Cobalt and Copper have been synthesized by room tempera-

ture co-precipitation route and have been found to be effective for the oxidative removal of carbon monoxide

from automobile exhaust gases even at relatively lower temperatures (cold-start). These catalyst materials

have been characterized by modern physico-chemical techniques such as XRD, TG, BET-BJH and SEM etc.

Nitrogen adsorption studies shows the samples are mesoporous in nature with pore diameter of 5 - 10 nm.

The catalytic efficiencies of these materials of having various compositions have been tested in a series of

temperature programmed oxidation reactions involving carbon monoxide and the results discussed.

Keywords: Catalysis, Ferrospinels, Mesoporous, Atmospheric Pollution, Carbon Monoxide Oxidation

1. Introduction

Protection of the environment and prevention of atmos-

pheric pollution from gaseous pollutants continues to be

a challenging task for chemists. The present study is

aimed at providing measures to ensure protection of the

atmosphere from automobile exhaust pollution due to

carbon monoxide, using some ferrospinel preparations as

green catalysts for oxidative purification. These catalysts

are ferrite spinels containing varying proportions of Ni,

Co and Cu which have been synthesized by room tem-

perature co-precipitation [1] and characterized before use.

The catalytic efficiencies of these materials have been

tested at various temperatures for the oxidative removal

of carbon monoxide (CO), from automobile exhaust

gases, which constitute the major polluting factor in at-

mospheric pollution. The catalyst materials prepared

have been characterized by methods like BET-BJH, TG,

XRD, Ammonia TPD etc. The fact that these catalyst

materials are cheaper compared to the platinum group

metals being used in a big way in the field make the pre-

sent study quite significant.

Spinel ferrites containing Ni, Co, and Cu have tech-

nological significance as catalysts due to their special

electronic and magnetic properties [2]. It has earlier been

observed that the presence of these metals in ferrite

spinel lattices strongly modify their stability and redox

properties [3].

2. Experimental

2.1. Synthesis of the Spinel Catalysts

Three samples of simple spinel ferrites, AFe2O4 (A = Ni,

Co & Cu) and three samples of mixed ternary spinel fer-

rites, NixCoyCuzFe2O4 (x = 0.6; y = z, y = 0.6; x = z, z =

0.6; x = y, where x + y + z = 1) have been prepared.

Metal hydroxides were co precipitated from an aqueous

homogenous mixture containing 0.17 moles of ferric

nitrate and its half the moles of Cu/Ni/Co nitrate mix-

tures depends on the sample composition. The precipi-

tating alkali, aqueous NaOH of 5.3 M, was added drop

wise from a micro burette with continuous stirring to a

final pH of 10. These precipitates were aged for 12 hrs,

filtered, washed off nitrates and alkali and dried at 110˚C

for 10 hrs. These powders were calcined at 300˚C for 4

hrs in air muffle to allow spinel phase formation. The

mixed ternary ferrites are given the code “xyz” where

xyz refer to the mole ratio x:y:z in the formula, NixCoy-

CuzFe2O4. All the samples were calcined at 500˚C also

separately for specific use.

2.2. Characterization of Samples

The samples of ferrospinels have been characterized by

different physico-chemical methods like XRD, BET sur-

face area, Ammonia-Temperature Programmed Desorp-

125

C. U. ANIZ ET AL.

tion (TPD), ICP-AES, TG/DTG, etc. The results of the

BET surface area and ammonia-TPD data are given in

Tables 1 and 2 respectively. TPD studies were done us-

ing Pulse Chemisorb 2705 from Micromeritics. Samples

were degassed in helium in a quartz reactor, at 200˚C,

and then cooled to ambient temperature. 5% NH3 in he-

lium gas (from Bheruka gases, Bangalore) is admitted to

the sample for 15 min to saturate the sample. Physi-

sorbed ammonia was purged away, keeping the sample at

100˚C. Sample temperature was raised to 700˚C at a rate

of 10˚C/min measuring the outlet desorbed ammonia

using a calibrated TCD.

Nitrogen adsorption-desorption measurements were

done on a volumetric Micromeritics Tristar apparatus at

Liquid N2 temperature, 77.35 K. On an average 33

points were taken for each sample. The average mass of

the sample was 0.3 g. Pore size distributions were calcu-

lated using BJH method and surface areas were taken

from BET isotherms.

The temperature-programmed reduction studies (TPR)

were performed in Micromeritics PulseChemisorb-2705,

which incorporates a thermal conductivity detector

(TCD). Samples were activated/surface-cleaned in he-

lium at 200˚C for 30 minutes, then cooled to ambient

temperatures. The reactive gas composition was 10%

balance N2, its consumption was measured while heating

the sample up to 750˚C at a rate of 10˚C/min.

2.3. Kinetic and Catalytic Studies

Kinetic and catalytic reaction studies of carbon monox-

ide oxidation have been done with all the catalyst sam-

ples, granulated to 16 × 20 mesh sizes after shaping to

cylindrical tablets. 0.5 g of the activated catalyst samples

were supported between glass wool plugs and flanked by

inert porcelain beads in the middle of a specially de-

signed quartz reactor, the flow diagram is as given in

Figure 1.

The inlet gas was calibrated mixture gas of 6% V/V of

oxygen, 1% V/V of carbon monoxide and the rest nitro-

gen. Gas flow was adjusted to a space velocity of 28,800

h–1. The reaction has been studied at four reaction tem-

peratures of 80˚C, 100˚C, 120˚C and 140˚C. The catalyst

samples have been activated at 300˚ for 1 hour prior to

reaction studies. The details are given in Table 3. The

CO and CO2 in the outlet gas were separated using

“Pouropack” packed column followed by converting to

methane by the “methanator” containing the Sud Chemie

Ni-catalyst working at 400˚C and finally detected sepa-

rately by GC-FID. The percentage conversion was cal-

culated using the equation:

% Conversion = (Cco-C’co)/Cco × 100

where, Cco and C’co are the % inlet and outlet carbon

Figure 1. Flow diagram of the reactor & the analytical set

up for carbon monoxide oxidation studies. (1. Mass flow

controller for N2-O2 gas mixture. 2. Mass flow controller for

CO. 3. Loop for mixing the gases. 4. Sampling point for

inlet analysis. 5. U-tube glass reactor in the electric furnace

with PID controller. 6. Outlet for measuring total gas flow.

7. Gas sample injector. 8. GC column. 9. Methanator with

Sud-Chemie methanation catalyst. 10. GC-FID detector. 11.

Data processor.)

monoxide compositions respectively.

Activation energies for the CO oxidation reactions

were calculated using the conversion data at various re-

action temperatures. The catalytic activity of the selected

samples were tested for CO-oxidation reaction continu-

ously for around 60 hrs at a constant temperature of

125˚C, which is near to its 70% CO conversion tempera-

ture, T70.

3. Results and Discussion

The composition of a few samples has been checked by

Inductively Coupled Plasma―Atomic Emission Spec-

troscopy technique and the results agreed fairly well with

theoretically expected w/w percentage of the transition

elements in them.

TG traces of the catalyst samples are given in Figure

2. Initial weight loss up to 120˚C is due to the loss of

adsorbed water. The samples are thermally stable after

300˚C.

The loss in weight in the region 120˚ to 300˚C could be

due to loss of oxygen arising out of structural transforma-

tions, beyond that the samples are stable. Hence these

catalyst samples are convenient for use in the automobile

exhaust systems. The content of surface and total hy-

droxide groups were calculated from the TG/DTA data

[4,5] from the mass loss in two different temperature

ranges of 150˚C - 200˚C and 150˚C - 400˚C respectively.

3.1. XRD Analysis

X-ray crystallographic studies show that the first three

C. U. ANIZ ET AL.

126

Table 1. Catalyst composition, identification codes, and surface area values.

Surface area (m2/g)

No. Composition,

NixCoyCuzFe2O4 Code (xyz) Pore volume

(cm3/g)

Average Pore dia

(nm) Calcined at

300˚C

Calcined

at 500˚C

1 NiFe2O4 100 0.209 6.55 123 53

2 CoFe2O4 010 0.199 10.50 81 39

3 CuFe2O4 001 0.241 7.07 124 16

4 Ni0.6Co0.2Cu0.2Fe2O4 622 0.233 5.57 149 40

5 Ni0.2Co0.6Cu0.2Fe2O4 262 0.243 6.64 117 56

6 Ni0.2Co0.2Cu0.6Fe2O4 226 0.203 7.51 106 39

Table 2. Ammonia TPD data on the ferrospinels, NixCoyCuzFe2O4 (x + y + z =1) activated at 300C.

Ammonia desorbed at various temperatures (mmol/g)

No.

Composition

code:xyz

NixCoyCuzFe2O4 Weak

(100 - 200)

Medium

(200 - 400)

Strong

(400 - 700) Total

1 NiFe2O4 0.111 0.284 0.249 0.644

2 CoFe2O4 0.107 0.210 0.115 0.431

3 CuFe2O4 0.186 0.362 0.156 0.704

4 622 0.179 0.674 0.222 1.075

5 262 0.129 0.622 0.314 1.065

6 226 0.108 0.402 0.131 0.641

1.1

1.2

1.3

1.4

1.5

1.6

1.7

1.8

1.9

0100200 300 400500 600 700800 900

Tem perature

Weight, mg

-0. 00 1

-0. 00 08

-0. 00 06

-0. 00 04

-0. 00 02

0.0002

Derivative Weight, mg/min

(a)

2.1

2.2

2.3

2.4

2.5

2.6

2.7

2.8

2.9

0100 200300 400500 600700 800900

Temperature

Weight, mg

-0.0012

-0.001

-0.0008

-0.0006

-0.0004

-0.0002

0.0002

Derivative weight, mg/min

(b)

Figure 2. TG picture of samples (a) 622 and (b) 262.

prominent peaks in the XRD patterns of inverse spinel

structures correspond to the hkl values, 311 (100%), 440

(60%) and 220 (50%).

The experimental dhkl values found for the simple fer-

rites and the mixed ferrites agree very closely with the

standard values given in the JCPDS cards for the simple

ferrite systems, proving the crystallographic features of

the samples. The XRD patterns for the samples 622, 262

and 226 are given in Figure 3. The fact that XRD data of

all samples are exactly alike is an expected result and

this aspect has been discussed elsewhere [3,6].

3.2. Nitrogen Adsorption Analysis

The BET surface area studies of the samples activated at

300˚C and 500˚C respectively are given in Table 1.

Among the simple ferrites, nickel and copper ferrites

have greater surface area than cobalt ferrite. For the

mixed ferrites, the surface area values are more or less

the same at an average value of 120 m2/g. The sample

262 shows an exceptionally high value 149 m2/g. Calci-

nation to 500˚C causes decrease in surface area, the de-

crease being drastic for copper ferrite. The decrease in

surface area with increase in calcination temperature is

partly due to the completion of dehydration of the sur-

000

10 15 202530 35 40 455055 60 6570

2theta values

In t e n s it y /a rb .values

Sam ple 622

Sam ple 262

Sam ple 226

Figure 3. XRD patterns of samples coded 622, 262 and 226.

127

C. U. ANIZ ET AL.

face hydroxyl groups, concomitant with the completion

of crystallization and growth of crystallite size, and

partly due to sintering. Nickel ferrite is retaining the rela-

tively higher surface area with increase of calcination

temperature and this property of nickel is expected to

support for the higher specific catalytic activity for the

resultant mixed ferrites at the higher temperatures.

Among the mixed ferrites, the better surface area values

of high copper composition samples even after 500˚C

calcination shows the support of nickel and cobalt on

retaining the surface area. The high relative surface area

of all the samples depends on the preparative methods

and high surface area accounts for the increased catalytic

activity for the oxidation

Selected representative isotherms have been plotted in

Figure 4, the corresponding Barret-Joyner-Halenda (BJ-

H) pore size distributions were calculated as applied to

the adsorption branches. All the samples exhibit nitrogen

adsorption/desorption isotherms of the Type IV, showing

a well-defined step and hysteresis loop. The P/Po posi-

tion of reflection points clearly shows that the pore di-

ameter is in the mesopore range, while the broad hys-

teresis loop reflects some mesopores are disorder in

shape. The initial part of the Type IV isotherm is attrib-

uted to monolayer-multilayer adsorption since it follows

the same path as the corresponding part of a Type II iso-

therm obtained with the given adsorptive on the same

surface area of the adsorbent in a non-porous form. The

average pore diameter is estimated to be mainly in a

range of 5.0 - 10.0 nm, while a very negligible portion is

also found at relatively larger dimensions. The hystereses

appear in the multilayer range of physisorption isotherms

and the two branches are nearly parallel over an appre-

ciable range of gas uptake which confirms the H1 type of

mesoporous materials. These studies were useful in iden-

tifying the nature of the pore structure of the catalysts [7].

The Type H1 loop is often associated with aggregates of

nano particles or compacts of uniform spheres in fairly

100

200

300

400

500

600

700

00.20.4 0.6 0.811.2

Rela tive pressure,p/po

Quan tiy o f N2 ad so r p be d, cm3/g , ST P

1OO

O1 O

OO1

622

262

226

Figure 4. N2-adsorption isotherms (type IV) of spinel ferrite

samples―BJH adsorption method.

regular array for the obtained samples, since it shows big

hysteresis loops at relative pressure between 0.6 and 0.8,

a capillary condensation in nano pores originated from

secondary inter-nano particle voids. These observations

are in line with the observations from SEM images

(Figure 5).

3.3. Temperature Programmed Reduction

Analysis

In the hydrogen TPR (Figure 6), the first peak charac-

terizes the reduction of +2 valence metal oxides,

Ni/Co/Cu, along with the reduction of ferric iron to fer-

rous stage, hematite to magnetite. The second separated

peak probably corresponds to the start of Fe3+ to Fe2+

reduction, and the third separated peak stands for the

reduction to metallic Iron. In the case of NiFe2O4, hema-

tite to magnetite peak is observed at 380˚C, where as in

CuFe2O4, it is at around 210˚C. In case of CoFe2O4, it is

further shifted to around 520˚C, and it might only in-

clude the reduction of Co2+ to Co+, while the reduction to

metallic Co took place at further extreme conditions. It

should also be noted that these Cu2+ ions may probably

play a significant role in promoting the reduction, even

without forming an extended separate metallic copper

phase. For the mixed spinel samples, corresponding re-

duction peaks of Ni/Co/Cu could be seen at respective

temperatures depends on their composition. It can be

assumed that addition of easily reducible metallic ele-

Figure 5. SEM photographs of samples 622, 262 and 226.

-2

0100 200300 400500600700 800

Tem pe ra ture, (

TCD signal, (a.u)

sample 100

sample 010

sample 001

Figure 6. Temperature programmed reduction curves of

pure ferrites.

C. U. ANIZ ET AL.

128

ments like Cu, may extend the Fe3+  Fe2+/3+ reduction

to a larger proportion, beyond the stoichiometric ratio.

Incorporation of structure stabilizers, like Co, may stabi-

lize iron in the ferric state in an even higher proportion,

exceeding the nominal 1:2 ratio. The temperature of the

Fe3+ to Fe3+/2+ is significantly shifted to lower tempera-

tures in the copper containing samples.

3.4. Carbon Monoxide Oxidation Studies

Data in Table 3 shows that copper ferrite among simple

ferrites, and sample with code 622 among mixed ferrites,

have the maximum specific activities for CO oxidation at

the reaction temperature of 120˚C. However, if the spe-

cific activities are divided by the surface area to get in-

trinsic activities, the maximum activity is for CoFe2O4

and the sample 262. Figure 7 shows the graphical repre-

sentation of CO oxidation performance.

In industrial catalysis, specific activity is more impor-

tant and, in this sense, sample 622 and copper ferrite are

the best catalysts. Intrinsic activity is important in theo-

retical correlation of activities with the chemical compo-

sition, concluding that the best two catalysts are cobalt

ferrite (at both activation temperatures) and sample 262,

having 70% of cobalt as the divalent cation. Figure 8

depicts the CO conversion observed for selected samples

after calcination at 500˚C. It is our observation that co-

balt ferrite, copper ferrite and mixed ferrites having high

percentage of cobalt are excellent catalysts practically in

accordance with theoretical requirements. This observa-

tion is in line with the surface area values of these sam-

ples too. Nickel ferrite and mixed ferrites with high pro-

portion of nickel have generally low activities, owing to

their lesser reducible nature irrespective of better surface

area values.

It is observed that the catalyst samples prepared and

employed have very high CO conversion abilities amo-

unting to almost 100% at as low a temperature as 140˚C

which has tremendous possibilities of application

100

120

80100120 140

Reaction temperature,oC

% CO con vesrio n

001

010

100

(a)

100

120

80100120 140

Reaction Temperature,oC

% CO con ver sio

262 226

622

(b)

Figure 7. Carbon monoxide oxidation activity evaluation of

ferrite spinel samples calcined at 300˚C (a) & (b).

100

80100120 140

Reaction Temperature,oC

% CO conversio

010

001

100

(a)

100

120

80100 120 140

Reaction Temperature,oC

% CO c o n versi o

622 262

226

(b)

Figure 8. Carbon monoxide oxidation activity evaluation of

ferrite spinel samples calcined at 500˚C (a) & (b).

in automobile exhaust purification processes. After the

cold start of an automobile engine, a very large fraction of

the total emission is left un-reacted [8]. Ordinary exhaust

catalysts start functioning only when the temperature rises

to about 200˚C. The usual noble metal catalysts require

higher reaction temperatures to catalyze CO or hydrocar-

bon oxidation. Hence the results of the present investiga-

tion have great relevance in automobile exhaust gas puri-

C. U. ANIZ ET AL.

129

Table 3. Carbon monoxide oxidation activity for the spinel ferrites.

% (v/v) CO conversion at various temperatures & Ea values (KJ/mol–1)

Samples calcined at 300˚C Samples calcined at 500˚C

Composition

Code:xyz

NixCoyCuzFe2O4 80˚C 100 120 140 Ea 80˚C 100 120 140 Ea

NiFe2O4 1.5 1.9 6.3 17.1 50.9 1.0 1.8 4.1 8.4 43.5

CoFe2O4 8.8 22.8 67.2 95.9 50.3 21.7 47.1 71.9 90.9 28.9

CuFe2O4 12.7 31.1 71.4 98.4 42.6 3.9 7.0 10.9 15.6 28.0

622 16.0 54.0 93.6 99.7 37.3 32.6 66.3 87.4 93.4 32.6

262 14.3 41.0 81.7 99.7 40.0 29.1 54.8 75.1 98.5 29.1

226 10.8 32.9 69.0 96.1 44.7 32.4 66.0 93.1 91.8 32.4

Test conditions: 1% CO and 6% O2 in balance nitrogen, Activation temperature of catalysts at 300˚C, Weight of catalyst loaded: 0.500 g, Space ve-

locity: 28000 h–1.

261014 18 22 2630 34 38 4246 50 5458

Time on stream, hrs

% CO Conversion

fication, especially at cold start of an automobile engine.

Ea values are in the range of 25 - 50 KJ·mol–1 for all the

samples, given in Table 3. These values are relatively, low

for sample 262 in agreement with its performance, even

though these values alone cannot be taken as the direct

measure of better catalytic performance.

The CO oxidation activities of single ferrites, say

NiFe2O4, CoFe2O4 and CuFe2O4 are in the order

NiFe2O4 < CoFe2O4 < CuFe2O4, which is in good

agreement with the reported study [9].

CO adsorption probability is relatively higher for Cu

rich samples and it is attributed to their more reducible

nature provided with high surface area. Higher surface

area provides better adsorption and help in inducing the

lattice oxygen to take part in CO2 formation. Figure 9. Time on stream study for carbon monoxide oxi-

dation of ferrite spinel sample 262, Ni0.1Co0.7Cu0.7Fe2O4

calcined at 300˚C, reaction temp = 125˚C.

The oxidation over mixed metal oxides is proposed to

follow a redox mechanism in which the metal ion cha-

nges oxidation state on consuming lattice oxygen for

CO2 formation. The gaseous oxygen appears to rejuve-

nate the catalyst surface at octahedral sites and enhance-

ing oxidation of the adsorbed CO [10].

operations. Except the initial dip, the conversion values

(lined out performance) are almost steady. A drop in

initial activity is expected for all catalytic systems and it

is due to the deactivation of a few unstable highly reac-

tive sites.

COads + Ocat  CO2 ads + *

CO2 ads  CO2 (g)

4. Conclusions

½ O2 + *  Ocat

where * depicts an oxygen vacancy and Ocat is a surface

lattice oxygen, probably O2–. Study of the catalytic activity for the oxidation of CO

using the ferrospinel samples in the temperature range 80

to 140˚C showed increase in activity with temperature,

reaching a limiting value of nearly 100% conversion of

CO for some samples even at 140˚C. The best two cata-

lyst samples in terms of specific activity are the mixed

ferrites coded 622, while in terms of intrinsic activity,

CoFe2O4 and the sample coded 262 are better. It is in-

teresting to note that these spinel samples of mesoporous

nature having 5 - 10 nm pore diameter are effective for

furnishing maximum catalytic surface contact for reac-

tant gaseous reactants. Our results suggest that these

spinel ferrites can function as efficient pollution abate-

ment catalysts for the protection of the atmospheric en-

The relatively higher CO conversion for copper-rich

mixed ferrite samples is due to the influence of copper

ions in lowering the reduction temperature of hematite to

magnetite, which is more catalytically active than hema-

tite. Presence of Cobalt can stabilize the ferrite structure

even at high temperatures. Nickel incorporation helps in

protecting the surface area of the ferrites samples even

after high temperature treatments, thereby allows the

application of these systems for high temperature appli-

cations.

Time on stream (TOS) studies (Figure 9) shows the

stability of the catalyst samples upon continuous opera-

tion. It tells about the fitness of our sample for industrial

C. U. ANIZ ET AL.

130

vironment from automobile pollutants and are economi-

cally viable.

5. References

[1] P. S. Anilkumar, J. J. Shrotri, S. D. Kulkarni, C. E. Desh

pande and S. K. Date, “Low Temperature Synthesis of

Ni0.8Zn0.2Fe2O4 Powder and Its Characterization,” Mate-

rials Letters, Vol. 27, No. 6, 1996, p. 293-296.

doi:10.1016/0167-577X(96)00010-9

[2] B. S. Rao, K. Sreekumar and T. M. Jyothi, “An Improved

Process for the Preparation of a Mixture of Alkylphenols

(O-Cresol and 2,6-Xylenol),” Indian Patent No. 2707/98,

1998.

[3] C. G. Ramankutty, S. Sugunan and B. Thomas, “Study of

Cyclohexanol Decomposition Reaction over the Ferro-

spinels, A1-xCuxFe2O4 (A = Ni or Co and x = 0, 0.3, 0.5,

0.7 and 1), Prepared by ‘Soft’ Chemical Methods,”

Journal of Molecular Catalysis A: Chemical, Vol. 187,

No. 1, 2002, pp. 105-117.

doi:10.1016/S1381-1169(02)00121-8

[4] S. Ek, et al., “Determination of the Hydroxyl Group

Content in Silica by Thermogravimetry and a Comparison

with 1HMAS NMR Results,” Thermochimica Acta, Vol.

379, No. 1-2, 2002, pp. 201-211.

doi:10.1016/S0040-6031(01)00618-9

[5] N. D. Ivanova, S. V. Ivanov, et al., “High Performance

Manganese Oxide for CO Oxidation,” Russian Journal of

Applied Chemistry, Vol. 75, No. 9, 2002, pp. 1420-1423.

doi:10.1023/A:1022216626347

[6] L. S. D. Glasser, “Crystallography and Its Applications,”

Van Nostrand Reinhold Company Ltd., New York, 1977,

p. 114.

[7] J. C. Groen, et al., “Pore Size Determination in Modified

Micro- and Mesoporous Materials. Pitfalls and Limita-

tions in Gas Adsorption Data Analysis,” Microporous and

Mesoporous Materials, Vol. 60, No. 1-3, 2003, pp. 1-17.

doi:10.1016/S1387-1811(03)00339-1

[8] F, Kapteijn, S. Stegenga, N. J. J. Dekker, J. W. Bijster-

bosch and J. A. Moulijn, “Alternatives to Noble Metal

Catalysts for Automotive Exhaust Purification,” Catalysis

Today, Vol. 16, No. 2, 1993, p. 273-287.

doi:10.1016/0920-5861(93)85024-T

[9] K. R. Krishnamurthy, B. Viswanathan and M. V. C. Sas-

tri, “Catalytic Oxidation of Carbon Monoxide on Spinel

type Ferrites Role of Magnetic Exchange Interactions,”

Proceedings of Indian National Science Academy, Vol.

43, 1977, pp. 34-38.

[10] P. Mars and D. W. Van Krevelen, “Oxidations Carried

out by the Means of Vanadium Oxide Catalysts,” Chemi-

cal Engineering Science, Special Supplement 3, 1954, pp.

41-59.