Thermochemistry of manganese oxides in reactive gas atmospheres: Probing catalytic MnOx compositions in the atmosphere of CO+O2

Article abstract of DOI:10.1016/s0040-6031(97)00417-6

A series of Mn-oxides in the range MnO₂-MnO were subjected to thermogravimetry on heating up to 1100°C in the reactive gas atmosphere of O₂, CO and CO+O₂. Some selected solid residues were analyzed for the MnO_x composition by infrared spectroscopy, and for carbon deposits by CHNS-analyzer. The CO+O₂ gas stream was chromatographically analyzed, following heating the gas/solid interface to some selected temperatures. The results indicate that Mn-oxides in the range MnO₂-Mn_1.3 exhibit catalytic properties towards the CO oxidation in the gas phase. The higher thermal stability (up to ~1000°C) of compositions in the range MnO_1.5-1.3, i.e. in the Mn₂O₃-Mn₃O₄ system, may render them particularly interesting in the field of applied catalysis. The experimental approach adopted in the present investigation may help accounting for the interfacial chemistry of solid-state reactions involving the surrounding gas atmosphere.

Full text of DOI:10.1016/s0040-6031(97)00417-6

Thermochimica Acta 311 (1998) 97±103
Thermochemistry of manganese oxides in reactive gas atmospheres:
Probing catalytic MnOx compositions in the atmosphere of CO‡O2
M.I. Zaki*, M.A. Hasan, L. Pasupulety, K. Kumari
Chemistry Department, Faculty of Science, University of Kuwait, P.O. Box 5969, Safat 13060, Kuwait
Received 29 April 1997; received in revised form 7 September 1997; accepted 19 October 1997
Abstract
A series of Mn-oxides in the range MnO2±MnO were subjected to thermogravimetry on heating up to 11008C in the reactive
gas atmosphere of O2, CO and CO‡O2. Some selected solid residues were analyzed for the MnOx composition by infrared
spectroscopy, and for carbon deposits by CHNS-analyzer. The CO‡O2 gas stream was chromatographically analyzed,
following heating the gas/solid interface to some selected temperatures. The results indicate that Mn-oxides in the range
MnO2±Mn1.3 exhibit catalytic properties towards the CO oxidation in the gas phase. The higher thermal stability (up to
ꢀ10008C) of compositions in the range MnO1.5±1.3, i.e. in the Mn2O3±Mn3O4 system, may render them particularly interesting
in the ®eld of applied catalysis. The experimental approach adopted in the present investigation may help accounting for the
interfacial chemistry of solid-state reactions involving the surrounding gas atmosphere. # 1998 Elsevier Science B.V.
Keywords: Catalytic behaviour of Mn-oxides in CO‡O2 atmosphere; Manganese oxides; Thermal behaviour of Mn-oxides in
CO‡O2 atmosphere; Thermochemistry of Mn-oxides; Thermogravimetry of Mn-oxides
1. Introduction
and MnO (MnII), and the mixed oxides of Mn5O8
(MnII,MnIV) and Mn3O4 (MnII,MnIII).
The adequate chemical makeup for deep oxidation
(combustion) catalysts [1,2], which are increasingly
sought after for environmental urgencies [3,4], seem
to necessitate the inclusion of metal oxides capable of
undergoing reversible oxygenation±deoxygenation
cycles [4]. Mn-oxides have been highly rated
[1,2,5] amongst the best candidate materials. Thermo-
dynamically stable MnOx compositions occur in the
range xˆ2 1 [6,7], including the differently-struc-
tured simple oxides of MnO2 (MnIV), Mn2O3 (MnIII)
Therefore, we have very recently probed the ther-
mal decomposition course of MnO2!MnO [8], so as
to identify the `redox' MnOx compositions amongst
the products yielding. Cyclic and non-cyclic thermo-
gravimetries (in different reactive gas atmospheres)
were amongst the investigation techniques employed.
The results have revealed that the mixed-valent
Mn3O4 (ˆMnIIMn2IIIO4) undergoes perfectly reversi-
ble oxygenation±deoxygenation cycles upon heating
in air (or oxygen atmosphere) at 500±10508C. The
oxygenation product has been identi®ed to be the
univalent Mn2O3 [8]. Moreover, MnII/MnIV couples
in the mixed-valent Mn5O8 (ˆ MnI2IMn3IVO8) have
been found [8] to undergo a redox reaction during
*Corresponding author. Tel.: 00965 4811188/5606; fax: 00965
4816482; e-mail: zaki@kuc01.kuniv.edu.kw
0040-6031/98/$19.00 # 1998 Elsevier Science B.V. All rights reserved
P I I S 0 0 4 0 - 6 0 3 1 ( 9 7 ) 0 0 4 1 7 - 6

98
M.I. Zaki et al. / Thermochimica Acta 311 (1998) 97±103
deoxygenation at 600±7008C (in air or O2 atmos-
phere), giving rise to Mn2O3 (MnIII).
atmosphere of CO or CO‡O2 were analyzed for
carbon deposits, using LECO CHNS-932 analyzer
(USA), and found to contain hardly any detectable
amounts.
The electron-mobile environment thus established
in such mixed-valent metal oxides [9] may promise a
catalytic potential in redox reactions. In the past, a
pronounced catalytic activity of hopcalite (CuMnOx)
in the H2O2 decomposition reaction had been attrib-
uted [10] to electron-exchange interactions between
ionic couples in the system CuI/CuII/MnIII/MnII.
Recently, Mn3O4-like surfaces have been allocated
the catalytic responsibility in oxidative coupling reac-
tions of methane [11,12].
2.2. Thermogravimetry (TG)
TG curves were measured while heating (at
ꢀˆ108C/min) small portions (ꢀ20 mg) of test sam-
ples in a dynamic atmosphere (30 ml/min) of O2, CO
or CO‡O2 (1 : 2) up to 11008C. The measurements
were carried out, using a model TGA-50 Shimadzu
automatic analyzer (Japan) equipped with a TA-50WS
work station for data acquisition and handling. The O2
gas was also a 99.99% pure product of KOAC.
The present investigation was basically designed to
probe MnOx compositions occurring in the range
MnO2 1, that perform catalytically when heated in
a mixed gas atmosphere of CO and O2. The experi-
mental work was carried out using thermogravimetry
of test Mn-oxides (on heating up to 800±11008C) and
gaschromatographyofthereactionatmosphere.Acom-
parison with results obtained for the same test oxides,
but in separate gas atmospheres of CO and O2, has
been the base on which the catalytic behaviour has been
recognized. Similar experimental approaches, whereby
interfacial events occurring during solid-state reac-
tions in reactive gas atmospheres are thermogravime-
trically resolved, are infrequent in the literature.
2.3. Infrared (IR) absorption spectroscopy
IR spectra were taken from KBr-supported test
samples (<1 wt%), over the frequency range 400±
1
4000 cm and at a resolution of 4 cm 1, using a
model 2000 Perkin±Elmer FT spectrophotometer
(UK). An on-line data station facilitated spectra acqui-
sition and handling.
2.4. Gas chromatography (GC)
Gas samples were trapped during the TG measure-
ments of MnO2 in the mixed stream of CO‡O2 (vide
supra) at different temperatures, viz. room tempera-
ture (RT), 300 and 7008C. The high temperatures were
chosen on the basis of TG results (vide infra). The gas
samples were then subjected to GC analysis for CO2
content, using a model CP-9000 Chrompack chroma-
tograph (USA) equipped with a TCD detector (at
2608C) and a packed column of PORAPACK Q (at
608C). A 99.99% pure He (KOAC) was employed as
the carrier gas (20 ml/min).
2. Experimental
2.1. Test Mn-oxides
MnO2 was a 99.9% pure product of Fluka (Switzer-
land), Mn5O8 was prepared by calcination of synthetic
MnC2O4Á2H2O at 3008C for 3 h [13], and Mn2O3,
Mn3O4 and MnO were AR-grade products of Aldrich
(USA).
For identi®cation purposes, MnO2 was heated at a
rate (ꢀ) of 108C/min in a 30 ml/min stream of CO to
500, 700 and 8008C. A similar heat treatment of MnO2
was carried out in a mixed stream (30 ml/min) of CO
and O2 (1 : 2) to 6008C. The temperatures applied
were chosen on the basis of thermal analysis results
(vide infra). The gases were 99.99% pure products of
Kuwait Oxygen and Acetylene Company (KOAC),
and were used as supplied.
3. Results and discussion
3.1. TG curves in O2 atmosphere
Fig. 1 displays TG curves obtained for the various
MnOx materials in the atmosphere of dynamic oxygen
(30 ml/min). The results help to characterize the fol-
lowing three temperature ranges, in each of which the
The various oxide samples were kept dry over silica
gel (in vacuum) till further use. Those heated in the

M.I. Zaki et al. / Thermochimica Acta 311 (1998) 97±103
99
Fig. 1. TG curves recorded on heating of small portions (20Æ2 mg) of the Mn-oxides indicated at ꢀˆ108C/min in a dynamic atmosphere
(30 ml/min) of pure O2 gas.
TG behaviours of the test materials are either analo-
gous or similar. Range-I, RT-5208C, in which MnO
and Mn3O4 suffer nothing but insigni®cant weight
changes (ÁW). Range-II, 520±9708C, in which all
the test materials concede major weight changes due
either to weight gain (WG) processes (viz. MnO &
Mn3O4) or weight loss (WL) processes (viz. Mn5O8 &
MnO2), except for Mn2O3 which remains weight
invariant. Range-III, >9708C, in which all the test
materials, without exception, undergo WL process.
The weight change values conceded by test materi-
als in the temperature ranges (II) and (III), together
with the processes assigned are indicated in Fig. 1.
Earlier reports [7,8] can lend strong support for the
assigned processes. It is obvious that the atmosphere
of oxygen could not sustain the compositions of
MnO2, Mn5O8 and Mn2O3 against thermal decompo-
sition (deoxygenation) which is seen to start near 600,
500 and 10008C, respectively. The sequence of
decomposition (Fig. 1) is as follows: MnO2!Mn5O8
(Tmaxˆ6808C), Mn5O8!Mn2O3 (Tmaxˆ8008C) and,
then, Mn2O3!Mn3O4 (Tmaxˆ990±10008C).
It is also obvious from Fig. 1 that the atmosphere of
oxygen could not sustain the compositions of Mn3O4
and MnO against heating at 400±9708C. These mate-
rials are shown to suffer a major WG process, whereby
they are oxygenated (oxidized) to yield Mn2O3. The
product is then decomposed in the temperature range-
III (>9708C) to give Mn3O4, which is a high-tempera-
ture thermodynamically stable composition of Mn-
oxide in O2-rich atmosphere [6,7].
Thus, Fig. 1 helps to con®rm that the test Mn-
oxides (MnO2 1) are thermally quite stable to heating
in O2 in the temperature range-I (RT-5208C). However
in the higher temperature ranges (II) and (III), the
thermal stability is maintained only for Mn2O3 (up to
9708C) and Mn3O4 (at >9708C), respectively. These
®ndings are in line with our previous results [8], in
con®rming the critical temperature dependence of the
equilibrium position of the reversible transformation
3Mn2O3$2Mn3O4‡1/2O2. The equilibrium is con-
siderably displaced to the right side at >9708C, and
vice versa. It is interesting to realize that the presence
of MnII in both Mn3O4 and MnO triggers their sus-

100
M.I. Zaki et al. / Thermochimica Acta 311 (1998) 97±103
Fig. 2. TG curves recorded on heating the Mn-oxides indicated in a dynamic atmosphere of pure CO gas.
ceptibility to oxygenation reactions, while its presence
in Mn5O8 does not do likewise. Mn5O8 has been
shown [8] to deoxygenate into Mn2O3, but has not
been found to oxygenate into MnO2.
been observed earlier [8] for the test materials on
heating up to 5008C in N2 atmosphere. Accordingly,
the WL effected at <5208C (Fig. 2) is CO-induced, i.e.
chemically, rather than thermally, assisted. In other
words, the effected WL marks the CO-reduction of
MnO2, Mn5O8, Mn2O3 and Mn3O4, eventually giving
MnO as the reduction product of each of them. Con-
sistently, MnO is shown (Fig. 2) to exhibit a pro-
nounced stability to heating in the CO atmosphere
up to 11008C. The minute WGs, observed at >5208C,
are most likely due to surface interactions with the CO
atmosphere, rather than to bulk interactions.
Hence, when O2 is replaced by CO in the surround-
ing gas atmosphere, the Mn-oxides (except for MnO)
are chemically destabilized on heating in the tempera-
ture range (RT±5208C) over which they are otherwise
thermally stable (Fig. 1). The solid residues of heating
MnO2 in CO up to 500, 700 and 8008C were subjected
to IR analysis. The spectra obtained (not shown)
3.2. TG curves in CO atmosphere
TG curves obtained for the various MnOx materials
in a dynamic atmosphere of CO (30 ml/min) are
compared in Fig. 2. It is evident that heating to
5208C in the CO atmosphere forces the test materials
to suffer a signi®cant WL, except for MnO. The WL is
shown (Fig. 2) to commence at ꢁ3508C for MnO2,
Mn2O3 and Mn3O4, but at a considerably lower tem-
perature for Mn5O8. The WL values determined cor-
respond very well to the formation of MnO as the
eventual product in all the cases (Fig. 2). It is shown to
suffer no further weight changes on heating up to
11008 in CO, except for minute (<2%) WGs.
In the temperature range (<5208C) of the WL
process observed in the CO atmosphere (Fig. 2), the
test materials are shown to be weight-invariant in O2
(Fig. 1). A similar weight-invariant behaviour has
showed four detectable absorption bands (at
1
415(s,sh), 509(s,sp), 614(s,sp) and 694(w,sh) cm
)
for the residue at 5008C, and two bands only (at
469(vs) and 704(w,sh) cm 1) for the residues at 700

M.I. Zaki et al. / Thermochimica Acta 311 (1998) 97±103
101
and 8008C. The former four bands are assignable to a
mixture of Mn2O3 and Mn3O4 [14], whereas the latter
two bands are most likely due to highly distorted MnO
[15]. These results may imply that the TG-displayed
single WL step for the CO-reduction of test samples is
somewhat composite in nature and conceals a multi-
step course to the eventual reduction product (MnO).
The fact that the CO-reduction of the test MnOx
materials to MnO is completely suppressed, or com-
pensated for, in the mixed atmosphere of CO‡O2 may
imply either complete unavailability of CO, or simul-
taneous re-oxidation by the coexisting O2. The beha-
viour communicated for MnO2 and Mn5O8, in
particular, may rather sustain the CO unavailability
option. The corresponding TG curves (Fig. 3) do not
display the tendency of both the test materials towards
the required dissociative uptake of O2. Consequently,
the CO molecules have most likely been catalytically
oxidized in the oxygen-rich atmosphere over the
operational temperature range of the CO-reduction
of the test oxides, i.e. at 200±5008C (Fig. 2). The
subsequent WG observed detectably at 500±10008C
for Mn5O8 (Fig. 3(B)) and Mn2O3 (Fig. 3(C)), may be
ascribed tentatively to CO2 absorption by the test
materials.
3.3. TG curves in CO‡O2 atmosphere
TG curves obtained on heating the test Mn-oxides
in a mixed atmosphere of CO‡O2 (1 : 2) are displayed
in Fig. 3, which also insets, for comparison purposes,
the TG curves obtained for the given oxide in the
separate gas atmospheres of CO (Fig. 2) and O2
(Fig. 1). The results help to characterize either of
the following two behaviours: (i) a behaviour to which
both CO and O2 have comparably contributed, and (ii)
a behaviour to which O2 is the dominant contributor.
The former behaviour is the one re¯ected by the TG
curves of MnO2 (Fig. 3(A)), Mn5O8 (Fig. 3(B)) and
Mn2O3 (Fig. 3(C)), whereas the latter is the one
monitored by the TG curves of Mn3O4 (Fig. 3(D))
and MnO (Fig. 3(E)).
GC analysis results of the CO‡O2 stream at RT, and
following heating over MnO2 to 300 and 7008C are
shown in Fig. 4. These results show the formation of
CO2 at 300 and 7008C. IR analysis of the solid
residues of MnO2 characterizes Mn5O8 (bands at
1
424(w,sh), 493(s,sp) and 608(s,sp) cm
[7]) and
For MnO2, Fig. 3(A) reveals that the coexistence of
CO in the surrounding gas atmosphere accelerates the
thermal decomposition to Mn5O8 (possibly plus
Mn2O3) to commence near 5008C instead of 6308C
in pure O2. On the other hand, the coexistence of O2
may be seen to completely suppress, or compensate
for, the reductive in¯uence of CO; the reductive WL
step commencing near 3508C in the CO atmosphere is
not visible in the mixed atmosphere of CO‡O2
(Fig. 3(A)). A similar behaviour can be seen in
Fig. 3(B) and (C) for Mn5O8 and Mn2O3, respectively.
The results reveal, for both materials, a complete
suppression of the CO-reduction, due to the coexis-
tence of O2, and show alternatively a steady WG till
the decomposition temperature (near 10008C) of
Mn2O3 to Mn3O4 is reached.
For Mn3O4, the TG obtained in CO‡O2 (Fig. 3(D))
is largely similar to that obtained in pure O2. Thus, the
CO-reduction to MnO is completely prevented,
whereas the oxidation to Mn2O3 is maintained. The
same is shown (Fig. 3(E)) to apply to MnO. The sole
difference here lies in the fact that CO originally has
no reductive in¯uence on MnO over the temperature
range scanned.
Mn2O3 (bands at 429(w,sh), 494(s,sp), 604(s,sh) and
694(w,sh) cm 1 [14]) following heating in CO‡O2 to
400 and 6008C, respectively. Consequently, it is not
until the temperature reaches 4008C that MnO2 com-
mences to reduce in CO‡O2, and the reduction pro-
duct, even at 6008C, is not MnO. Thus, the catalytic
CO‡O2 reaction (at 3008C in particular) is the
only source left for CO2 production in the gas
phase [16].
The above results indicate that all of the MnOx
compositions tested, except for MnO, exhibit catalytic
activity towards CO oxidation in O2-rich atmosphere.
However, the higher thermal stabilities of composi-
tions in the range MnO1.5±1.3, i.e. in the Mn2O3Ð
Mn3O4 system, might render them particularly inter-
esting in the ®eld of applied catalysis. The failure in
detecting carbon deposits on the solid residues of the
heating of MnO2 in CO‡O2 atmosphere at 6008C may
exclude the Boudouart mechanism (2CO(g)!CO2(g)
‡
C(s) [17]) from the likely surface reaction pathways of
CO oxidation [16]. It may, alternatively, suggest Mn-
oxides as promising ingredients for the catalyst being
sought [4] for soot combustion in automobile exhaust
emissions.

102
M.I. Zaki et al. / Thermochimica Acta 311 (1998) 97±103
Fig. 3. TG curves recorded on heating MnO2 (A), Mn5O8 (B), Mn2O3 (C), Mn3O4 (D) and MnO (E) in a dynamic atmosphere of CO‡O2
(1 : 2). Corresponding curves obtained in separate atmospheres of pure O2 (Fig. 1) and CO (Fig. 2) are inset for comparison.

M.I. Zaki et al. / Thermochimica Acta 311 (1998) 97±103
103
analogous series of the test solids and gases, is carried
out.
Acknowledgements
The ®nancial support of Kuwait University via
projects Nos. SPC 076, SLC056, SLC063 and
SLD002, and the excellent technical assistance at
Analab of the Chemistry Department, are highly
appreciated.
References
[1] R. Prasad, L.A. Kennedy, E. Ruckenstein, Catal. Rev. Sci.
Eng. 26 (1984) 1.
[2] P.J. Gellings, H.J.M. Bouwmeester, Catalysis Today 12
(1992) 1.
[3] F. Nakajima, Catalysis Today 10 (1991) 1.
[4] G. Mul, J.P.A. Neeft, F. Kapteijn, M. Makkee, J.A. Moulijn,
Appl. Catal. B, Environ. 6 (1995) 339.
[5] H.H. Kung, Transition Metal Oxides: Their Surface Chem-
istry and Catalysis, Elsevier, Amsterdam, 1989.
[6] F.A. Cotton, G. Wilkinson, P.L. Gaus, Basic Inorganic
Chemistry, J. Wiley and Sons, Inc., Chichester, 1995, pp.
561±564.
Fig. 4. Gas chromatograms for samples taken from the gas phase
of CO‡O2 at room temperature and following heating over MnO2
to 300 and 7008C.
[7] A.K.H. Nohman, M.I. Zaki, S.A.A. Mansour, R.B. Fahim, C.
Kappenstein, Thermochim. Acta 210 (1992) 103.
[8] M.I. Zaki, M.A. Hasan, L. Pasupulety, K. Kumari, Thermo-
chim. Acta (1997), in print.
4. Conclusion
[9] C. Zener, Phys. Rev. 81 (1951) 440; 82 (1951) 403; 83 (1951)
299; 85 (1952) 324.
Mn-oxide compositions in the range MnO2±1.3, i.e.
in the MnO2±Mn3O4 system, do exhibit catalytic
activity towards CO oxidation in O2-rich atmosphere.
The higher thermal stabilities (up to near 10008C) of
compositions in the range MnO1.5±1.3, i.e. in the
Mn2O3±Mn3O4 system, may render them particularly
interesting in the ®eld of applied catalysis.
The experimental approach adopted in the present
investigation can help examining solid catalysts under
working conditions by thermogravimetry in the reac-
tive gas atmosphere. Resolution of the interfacial
events is quite likely, provided that systematic ther-
mogravimetry of the solid/gas system, including
[10] G.-M. Schwab, S.B. Kanungo, Z. Phys. Chem. NF 107 (1977)
109.
[11] G.D. Moggridge, J.P.S. Badyal, R.M. Lambert, J. Phys.
Chem. 94 (1990) 508.
[12] G.D. Morggridge, T. Rayment, R.M. Lambert, J. Catal. 134
(1992) 242.
[13] P. Dubois, Compt. Rend. 198 (1934) 1502.
[14] R.M. Potter, G.R. Rossman, Am. Mineral. 64 (1979) 1199.
[15] F. Vratny, M. Dilling, F. Gugliotta, C.N.R. Rao, J. Sci. Ind.
Res. 20B (1961) 590.
[16] G.C. Bond, Heterogeneous Catalysis: Principle and Applica-
tions, 2nd ed., Clarendon Press, Oxford, 1990, pp. 115±117.
[17] D.J.C. Yates, L.L. Murrell, E.B. Prestridge, J. Catal. 57
(1979) 41.