Adsorption and Microcalorimetric Measurements on the Interaction of CO and H2 with Polycrystalline Ru and Ru/TiO2 Catalyst

Article abstract of DOI:10.1006/jcat.1997.1717

A microcalorimeter equipped with gas circulation cells and coupled at outlet to a gas chromatograph was used for the simultaneous measurements of the uptake and the differential heat (q_d) evolved during the adsorption of CO and H₂ pulses over polycrystalline ruthenium metal and a RuTiO₂ catalyst in the temperature range 300-475 K and as a function of surface coverage. The initial differential heat for the adsorption of CO and H₂ over polycrystalline ruthenium at 300 K was 120 and 65 kJ mol^-1, respectively, the corresponding values in the case of RuTiO₂ being around 130 and 57 kJ mol^-1. With the rise in sample temperature, the q_d for CO adsorption over Ru metal remained almost constant, while in the case of Ru/TiO₂ it decreased substantially. The fraction of CO or H₂ adsorbed, conversion of CO_ad to CO₂, and the corresponding values of heat evolved showed different trends, when these samples were exposed to the successive CO or H₂ pulses at different temperatures. The H₂ adsorption is found to be suppressed on Ru/TiO₂, particularly at the low sample temperatures. Also, the CO adsorption over Ru/TiO₂ at temperatures above 400 K resulted in the partial reduction of the support, and this is facilitated by the heat evolved at the metal/support interfaces during CO chemisorption. On the other hand, the CO dissociation followed by CO_(ad) + O_(ad) reaction was a predominant step giving rise to CO₂ formation in the case of Ru metal. This study also confirms that, for both the samples, while the CO adsorption remains uninhibited by the preadsorbed H₂, the catalyst surface covered with the CO was completely inaccessible to subsequent H₂ adsorption.

Full text of DOI:10.1006/jcat.1997.1717

JOURNAL OF CATALYSIS 169, 415–422 (1997)
ARTICLE NO. CA971717
Adsorption and Microcalorimetric Measurements on the Interaction of CO
and H2 with Polycrystalline Ru and Ru/TiO2 Catalyst
V. P. Londhe and N. M. Gupta
Chemistry Division, Bhabha Atomic Research Centre, Trombay, Mumbai-400 085, India
Received September 9, 1996; revised April 7, 1997; accepted April 9, 1997
interfaces have been discerned. In continuation, we have
now employed similar methodology to study the mecha-
nistic aspects of CO methanation reaction over the bulk
and the titania-supported ruthenium catalysts. The choice
of Ru was based on its high activity and selectivity for
methane formation, particularly at low reaction tempera-
tures.
The hydrogenation of CO continues to draw wide atten-
tion, not only because of its industrial importance but also
because it serves as a model catalytic hydrogenation re-
action where the reactants, products, and transient species
at the catalyst surface could be characterized simultane-
ously. A large volume of work published in this area is
reviewed in various articles (4–9). Broadly speaking, two
distinct reaction mechanisms have been put forward. Ac-
cording to an earlier school of thought, the CO hydrogena-
tion is governed by the direct participation of CO and H2
molecules to form oxygenated intermediates. Thus, Storch
A microcalorimeter equipped with gas circulation cells and cou-
pled at outlet to a gas chromatograph was used for the simultaneous
measurements of the uptake and the differential heat (qd) evolved
during the adsorption of CO and H2 pulses over polycrystalline
ruthenium metal and a Ru/TiO2 catalyst in the temperature range
300–475 K and as a function of surface coverage. The initial dif-
ferential heat for the adsorption of CO and H2 over polycrystalline
ruthenium at 300 K was 120 and 65 kJ molꢀ1, respectively, the
corresponding values in the case of Ru/TiO2 being around 130 and
57 kJ molꢀ1. With the rise in sample temperature, the qd for CO
adsorption over Ru metal remained almost constant, while in the
case of Ru/TiO2 it decreased substantially. The fraction of CO or H2
adsorbed, conversion of COad to CO2, and the corresponding values
of heat evolved showed different trends, when these samples were
exposed to the successive CO or H2 pulses at different temperatures.
The H2 adsorption is found to be suppressed on Ru/TiO2, particu-
larly at the low sample temperatures. Also, the CO adsorption over
Ru/TiO2 at temperatures above 400 K resulted in the partial reduc-
tion of the support, and this is facilitated by the heat evolved at the
metal/support interfaces during CO chemisorption. On the other
hand, the CO dissociation followed by CO(ad) + O(ad) reaction was
a predominant step giving rise to CO2 formation in the case of Ru
metal. This study also confirms that, for both the samples, while
the CO adsorption remains uninhibited by the preadsorbed H2, the
catalyst surface covered with the CO was completely inaccessible
c
–
et al. (10) proposed the formation of M CH2O species as
intermediates which on surface condensation resulted in
chain growth. Pichler and Schulz (11) postulated that even
–
though M CH2O species are present on the metal surface,
chain growth is accomplished by the direct and stepwise
insertion of adsorbed CO molecules. Subsequent investi-
gations, however, brought out various flaws in this scheme,
such as that it does not explain why alcohols can initiate syn-
thesis but cannot contribute to chain growth by dehydro-
condensation. According to another model, which is widely
accepted now, while methanol is produced by direct hy-
to subsequent H2 adsorption.
ꢁ 1997 Academic Press
INTRODUCTION
In the earlier reported studies we used a heat-flow mi- drogenation of molecularly adsorbed CO, the synthesis of
crocalorimeter equipped with the gas circulation cells and hydrocarbons is initiated by the dissociation of adsorbed
coupled at the outlet to a gas chromatograph in order to CO to form an active state of carbon species, which trans-
investigate the CO oxidation reaction over the bulk and forms directly to hydrocarbons via formation of methyl
the supported forms of noble metal catalysts (1–3). It has and methene groups for chain growth. The formation of
been shown that the preponderance of one of the several various forms of carbon during CO dissociation has also
possible reaction routes depends on the temperature and been proposed (12, 13). A number of studies supporting
the chemical nature of the catalyst surface at the time of in- “active-carbon” intermediate theory are reviewed in Refs.
teraction with the reactants. By simultaneous measurement (4–6).
of the enthalpy changes, gas adsorption, and the reaction
In the present study we therefore evaluated the heat
products formed, the role played by the lattice oxygen and evolved during exposure of CO, H2, and CO + H2 (1 :4)
the localized temperature fluctuations at the metal/support over polycrystalline ruthenium metal and a Ru/TiO2
415
0021-9517/97 $25.00
c
Copyright ꢁ 1997 by Academic Press
All rights of reproduction in any form reserved.

416
LONDHE AND GUPTA
catalyst at different temperatures in range 300–470 K. Microcalorimetry
The effect of surface coverage was evaluated by dosing
Microcalorimetric measurements of the heats of adsorp-
successive pulses of these adsorbates. The fraction of the
gases adsorbed and the reaction products formed at dif-
ferent stages were analyzed simultaneously. These data
helped us in delineating the role played by the support and
the mechanistic routes involved at different stages of CO
methanation reaction. While the present paper deals with
the adsorption of CO and H2 the data on the coadsorption
of these two gases are included in the following paper.
tion were carried out using a Calvet-type heat flow mi-
crocalorimeter equipped with stainless-steel gas circulation
vessels (C-80 calorimeter, Setaram, France). The proce-
dure is described earlier in detail (1–3). A 100-mg sample
was taken in one of the vessels, whereas the reference cell
was kept empty. The cells were connected at one end to a
common vacuum system or a gas manifold so as to have
a provision for evacuation of the vessels or alternatively
to maintain a sample under desired gas flow. Fine control
needle valves were used to maintain the vessels under iden-
tical and steady gas flow conditions. The effluent from the
sample cell was analyzed with an assembly of Porapak-P
and molecular sieves 5 A colums, both connected in tan-
dem and each followed by a thermal conductivity detector.
Both the detectors and the GC columns were thermally
insulated and were maintained under ambient conditions.
Prior to calorimetric measurements, a sample was treated
in situ in flowing hydrogen (20 ml minꢀ1) at 475 K for 2 h,
followed by evacuating (475 K, 1 h) and heating the sample
for 1 h under helium flow at 475 K. While maintaining a
sample under helium flow (20 ml minꢀ1) several successive
100 ꢀl (4.1 ꢀmol) pulses of CO or H2 were dosed over cata-
lyst sample through an injection port at the entry point of
the sample cell. The amount of a gas adsorbed over catalyst
surface and the corresponding heat evolved were recorded
simultaneously using a multipen recorder/integrater. The
value of differential or molar heat of adsorption (qd) was
thus obtained by dividing the heat evolved during an in-
dividual pulse injection with the amount of gas adsorbed.
The values of heat at initial stages of adsorption (qi) were
obtained from the plots of log qd vs surface coverage by
extrapolating them to ꢁ = 0. Each set of experiments was
repeated at least two or three times in order to check for the
data reproducibility. As mentioned in our earlier publica-
tions, the accuracy of q values depended on the accuracy of
gas chromatographic data and the variations in the q values
were found to be the in range of 5–10% .
EXPERIMENTAL
Catalyst
The Ru/TiO2 sample used in this study was prepared by
coprecipitation method starting with the chlorides of ruthe-
nium and titanium and usinga 20% ammonium bicarbonate
solution for precipitation at 335 K. The precipitate was sep-
arated by centrifugation after 16 h of aging and was washed
repeatedly till free from chloride ions. The catalyst was
dried in air at 345 K (24 h) and then at 370 K (8 h) followed
by sintering at 525 K for 16 h and at 620 K for 8 h under
O2 + Ar (1 : 1, 60 ml minꢀ1) flow. A 40–80 mesh fraction of
this sample having BET surface area of ꢂ90 m2 gꢀ1 was used
for all the experiments after 3 h reduction in H2 + Ar stream
(1 : 1, 60 ml minꢀ1) at 575 K. The samples were cooled
and stored under argon atmosphere. The metal content of
the sample as evaluated by X-ray fluorescence method was
found to be 4.5 wt% . The X-ray diffraction pattern revealed
the presence of about 80% anatase and 20% rutile phases
of TiO2, while no XRD signals due to ruthenium were de-
tected indicating its fine dispersion. TEM analysis revealed
that the size of the majority of Ru particles ranged between
2 and 4 nm.
A fine powder of ruthenium metal used in this study was
prepared by complete reduction of ruthenium trichloride
in flowing hydrogen at 575 K. A 40–80 mesh fraction of
this sample employed in this study was porous in nature
and showed a N2 adsorption BET surface area of around
Carbon monoxide of 99.9% purity from Airco (U.S.A.)
was used after passing through a dry ice trap to remove
any carbonyl or moisture impurities. Helium from Airco
was also passed through deoxo and molecular sieves traps
before its use as a carrier gas. High-purity hydrogen was
obtained with an Elhygen H2-generator.
6 ꢃ 2 m2 gꢀ1
.
The number of exposed metal sites was evaluated by
volumetric hydrogen chemisorption method using the
pulse flow technique (Micromeritics, U.S.A., Model-Pulse
chemisorb 2700). For this purpose, the samples were pre-
treated in situ at 573K first in H2 and then in He. The amount
of H2 chemisorbed at room temperature was found to be
8.6 and 7.8 ꢀmol gꢀ1 for the Ru/TiO2 and Ru metal samples,
respectively. Measurements carried out using temperature-
programmed desorption of H2, however, gave marginally
higher values, particularlyin the case ofthe supported metal
sample. The two samples may thus be considered compara-
ble in terms of metal surface area even though their BET
areas were quite different.
RESULTS
Microcalorimetry
CO Adsorption
Ru/TiO2 catalyst. When a 100-ꢀl (ꢂ4.1 ꢀmol) pulse of
carbon monoxide was dosed over 100 mg catalyst sample

ADSORPTION OF CO AND H2 OVER Ru AND Ru/TiO2 CATALYST
417
at 300 K show that the qd reduced sharply from a value
of 112 kJ molꢀ1 for the first pulse injection to a value of
ꢂ78 kJ molꢀ1 for the third CO pulse, which was accompa-
nied by the drastic decreases in CO adsorption. A negligibly
small heat was evolved for further CO pulse injections. With
the rise in sample temperature to 370K, the adsorption from
the first CO pulse increased and the qd was ꢂ96 kJ molꢀ1
.
Smaller amounts of CO continued to adsorb during succes-
sive CO pulse injections giving rise to ꢂ70 kJ molꢀ1 heat
(Fig. 3b). At a reaction temperature of 470 K, almost con-
stant qd of ꢂ75–80 kJ molꢀ1 was observed and a large frac-
tion of CO continued to get adsorbed during the successive
six pulse exposures over catalyst surface (Fig. 3d).
Ruthenium metal. Almost a similar fraction (ꢂ60 ꢃ
2% ) adsorbed/reacted when a 100 mg ruthenium metal
sample was exposed to the first 4.1 ꢀmol CO pulse at dif-
ferent temperatures in the range 300–470 K. Also, a similar
amount of heat (ꢂ100 ꢃ 5 kJ molꢀ1) was evolved during this
FIG. 1. Percentage of CO adsorbed/reacted (a) and the conversion interaction, irrespective of the catalyst temperature. How-
of CO(ad) to CO2 (b) when a pulse of 4.1 ꢀmol CO was dosed over a
fresh 100 mg Ru/TiO2 catalyst at different temperatures. Curve c shows
the corresponding values of differential heat evolved.
ever, when several successive CO pulses were dosed at a
particular temperature, a progressively reduced amount of
CO was adsorbed. Figures 4a and 4b show the data obtained
for the reaction temperatures of 300 and 370 K. Figures 4c
and 4d show the fraction of CO adsorbed/reacted from the
under He flow (20 ml minꢀ1) a part of the CO was ad-
successive CO pulses dosed at the sample temperatures of
420 and 470 K, respectively. While no carbon dioxide was
produced during CO interaction at the sample tempera-
tures below 400 K, a part of the adsorbed CO gave rise
to CO2 formation at higher temperatures. Figures 4e and 4f
give the typicalCO2 yield from the five successive CO pulses
sorbed, a part reacted to form CO2 while the rest was eluted.
The amount of eluted CO or that of the product CO2 de-
pended upon the catalyst temperature. The amount of CO
adsorption from a pulse and its conversion to CO2 as a func-
tion of catalyst temperature are given in Figs. 1a and 1b.
Even though the fraction of CO adsorbed increased pro-
gressively with the rise in catalyst temperature (Fig. 1a),
CO2 was formed only at the sample temperatures above
370 K (Fig. 1b). Thus, while only ꢂ8% of adsorbed CO was
converted to CO2 at reaction temperature of 425 K, the
yield at 470 K was ꢂ32% .
Figure 1c gives the amount of heat evolved (qd) per
mole of adsorbed/reacted CO from the first pulse injection.
These data indicate that the qd decreased progressively with
the increase of sample temperature. Thus, compared to a
value of qd ꢂ 112 kJ molꢀ1 for the first pulse CO adsorption
at 300 K, the corresponding value for reaction temperature
of 475 K was around 75 kJ molꢀ1 (Fig. 1c).
The amount of adsorbed CO was found to reduce pro-
gressively when several successive CO pulse (4.1 ꢀmol
each) exposures were made at a particular sample temper-
ature, without any in-between pretreatment. The decrease
in CO2 yield during these CO pulse exposure was how-
ever only marginal, particularly at a sample temperature
of 470 K. These data for different sample temperatures are
given in Fig. 2.
Figures 3a–3d summarize the data on the differential heat
of CO adsorption as a function of CO adsorbed from the
successive 4.1 ꢀmol CO pulses dosed over Ru/TiO2 at dif-
FIG. 2. Percentage of CO adsorbed/reacted and conversion of COad
to CO2 when five successive CO pulses (4.1 ꢀmol each) were dosed over
Ru/TiO2 at different temperatures: (a) 300 K, (b) 370 K, (c) 420 K, (d)
ferent temperatures. The data in Fig. 3a for the experiment 470 K, (e) 420 K, and (f) 470 K.

418
LONDHE AND GUPTA
FIG. 3. Differential heat evolved as a function of surface coverage when a 100 mg Ru/TiO2 catalyst was exposed to successive 4.1 ꢀmol CO pulses
at different temperatures.
at the reaction temperatures of 420 and 470 K, respectively. ple. In each experiment the heat of adsorption from the first
It is of interest to note that no CO2 formed from the first CO pulse was ꢂ110 ꢃ 5 kJ molꢀ1. However, the qd decreased
pulse injection; an increasing amount of CO2 was produced progressively with the increasing surface coverage, partic-
during the next two pulse injections, the yield being higher ularly at lower exposure temperatures. Thus, for a cata-
at higher reaction temperatures (Figs. 4e and 4f). Further- lyst temperature of 300 K, the value of qd decreased to
more, no CO2 was produced during the fourth and the fifth ꢂ60 kJ molꢀ1 for the fifth pulse injection (Fig. 5a). On the
CO pulse exposures.
other hand, it remained almost constant at 110 kJ molꢀ1 for
Figure 5 presents the microcalorimetry data for the suc- all the dosed CO pulses at a sample temperature of 470 K
cessive CO pulse adsorption on the ruthenium metal sam- (Fig. 5c).
FIG. 4. Percentage of CO adsorbed/reacted and conversion of COad
to CO2 when five successive CO pulses (4.1 ꢀmol each) were dosed over
100 mg polycrystalline Ru at different temperatures. (a) 300 K, (b) 370 K, crystalline Ru sample to five successive 4.1-ꢀmol CO pulses at different
(c) 420 K, (d) 470 K, (e) 420 K, and (f) 470 K. temperatures: (a) 300 K, (b) 420 K, and (c) 470 K.
FIG. 5. Differential heat evolved during exposure of a 100-mg poly-

ADSORPTION OF CO AND H2 OVER Ru AND Ru/TiO2 CATALYST
419
Hydrogen Adsorption
Ru/TiO2 catalyst. An almost similar amount was ad-
sorbed at a particular temperature when five to six succes-
sive H2 pulses (4.1 ꢀmol each) were dosed over a 100 mg
sample of Ru/TiO2. However, the amount of adsorbed H2
increased progressively with the rise in catalyst tempera-
ture. Thus, about 12, 28, and 40% of the dosed H2 was
adsorbed from an individual pulse when the sample temper-
ature was 300, 370, and 470 K, respectively. Figures 6a–6c
present these data.
Figures 6d and 6e give the values of qd when the Ru/TiO2
sample was exposed to five successive H2 pulses as men-
tioned above. As seen in Fig. 6d, about 48 kJ heat was
evolved per mole of H2 adsorbed from the first pulse
dosed at a sample temperature of 300 K, while the sub-
sequent H2 pulse exposures gave rise to a smaller amount
of heat (ꢂ30 kJ molꢀ1). For the exposure temperature in
the 370–470 K range, an almost similar qd ꢂ 37 kJ molꢀ1
was observed for the first H2 injection, the value be-
ing ꢂ30 kJ molꢀ1 for the subsequent H2 pulse exposures
(Fig. 6e).
FIG. 7. Percentage of H2 adsorbed over 100 mg polycrystalline Ru
during exposure to successive 4.1-ꢀmol H2 pulses at different tempera-
tures.
Ru metal. The catalyst temperature had only a marginal
effect on the hydrogen adsorption from the first pulse in-
jection. Thus, while around 70% of the injected H2 was ad-
sorbed at 300 and 370 K, the amount adsorbed was ꢂ60%
for the sample temperature of 470 K. Furthermore, a pro-
gressively smaller fraction of H2 was adsorbed from the
successive pulses dosed at a particular temperature. These
data are plotted in Fig. 7.
Figure 8 gives the qd values for the first five H2 pulses
dosed over Ru metal at different temperatures. The first H2
pulse gave rise to ꢂ45 ꢃ 3 kJ molꢀ1 heat, irrespective of the
exposure temperature. The qd was lower for the successive
pulse injections made at the sample temperatures of 300
and 370 K (Figs. 8a and 8b). The heat value at higher sample
temperatures was, however, found to be almost constant for
all the injected H2 pulses (Fig. 8c).
FIG. 6. Percentage of H2 adsorbed and corresponding differential
heat evolved when a 100-mg Ru/TiO2 catalyst sample was exposed to five
successive 4.1-ꢀmol H2 pulses at different temperatures. (a) 300 K, (b)
370 K, (c) 470 K, (d) 300 K, and (e) 370 and 470 K.
FIG. 8. Differential heat evolved during exposure of a polycrystalline
Ru metal sample to successive 4.1-ꢀmol H2 pulses at different tempera-
tures.

420
LONDHE AND GUPTA
Effect of Preadsorbed Gases
metal, particularly in its supported form. Also, to the best
of our knowledge no such study has so far been carried
out on this catalyst system under the actual reaction con-
ditions. This becomes apparent from the recent reviews by
Cardona-Martinez and Dumesic (14) and Toyoshima and
Somorjai (15) which include most of the earlier studies
on calorimetric measurements of CO and H2 adsorption
over noble metals. The initial heat (qi) values obtained in
our study for CO and H2 adsorption at room temperature
(Table 1) are quite in agreement with the reported values.
Thus the qi ꢂ 120 kJ molꢀ1 for CO adsorption over poly-
crystalline Ru (Table 1) matches the value reported in Ref.
(15). Guerrero-Ruiz (16) reported an initial enthalpy of
ꢂ138 kJ molꢀ1 for the adsorption of carbon monoxide over
ruthenium graphitized carbon black catalysts, which is con-
sistent with the qi observed in our study over Ru/TiO2 at
300 K (Table 1).
A comparison of the data in Figs. 3 and 5 shows a large
difference in the CO chemisorption behavior of Ru metal
and Ru/TiO2, particularly at the higher exposure temper-
atures. Thus while an almost constant qd ꢂ 110 kJ molꢀ1
was evolved during the first pulse injection over ruthenium
metal at different temperatures (Fig. 5), a progressively de-
creasing qd was observed with the rise in temperature using
Ru/TiO2. Also, the data in Fig. 1a show that the amount of
CO adsorbed from the first pulse increased with the rise
in sample temperature, contrary to the data of Fig. 4. IR
spectroscopy studies have shown that the interaction of CO
with a Ru/TiO2 catalyst gave rise to the vibrational bands
at around 1223, 1360, 1428, 1529, and 1590 cmꢀ1 which
are identified with the oxygenate species such as bridged
carbonates and bicarbonates (17–20). In addition, bands
were observed due to chemisorbed CO (1900–2150 cmꢀ1),
physisorbed CO (2185 cmꢀ1) and the adsorbed CO2. No
such species were, however, formed when a metal-free TiO2
sample was exposed to CO. We can thus conclude that the
heat evolved during chemisorption of CO over Ru/TiO2
promotes the formation of carbonate type surface species at
metal/support interfaces which dissociate at elevated tem-
peratures to form CO2 (Fig. 1b). The over all process thus
amounts to reduction of TiO2, for example,
The tests performed on Ru/TiO2 showed that the catalyst
surface covered with carbon monoxide at room tempera-
ture was not accessible to hydrogen molecules. Thus, no
hydrogen was adsorbed over catalyst sample exposed to 10
successive CO pulse doses under He flow (20 ml minꢀ1
)
at 300 K followed by flushing for 30 min under He. Cor-
respondingly, no heat was evolved during these H2 pulse
injections. On the other hand, when a sample exposed simi-
larly to 10 successive pulses of 4.1 ꢀmol H2 and flushed with
He (30 min) was dosed thereafter with CO pulses at 300 K,
the amount of gas adsorbed was only marginally lower than
that shown in Fig. 1. Thus, about 28% of CO was adsorbed
over Ru/TiO2 from the first pulse, compared to a value of
30% in the case of a fresh sample. The amount adsorbed
from the three successive pulses was 13, 5, and 2% respec-
tively (cf. Fig. 1a). The heat evolved during this process was
around 92 ꢃ 3 kJ molꢀ1 compared to a value of 110 kJ molꢀ1
in Fig. 3a.
Similar results were obtained using a ruthenium metal
sample.
Initial Heat of Adsorption
The heat values corresponding to initial stages of ad-
sorption, as calculated from the coverage dependent mi-
crocalorimetry data in Figs. 3, 5, 6, and 8 are given in Table 1.
Adsorption over Titania
Only a negligible (1–2% ) fraction of CO was absorbed
over the metal-free titania sample, prepared and pretreated
under conditions similar to those used for Ru/TiO2, when
dosed with successive 4.1-ꢀmol pulses at the sample tem-
peratures in the range 300–470 K. Also, no CO2 was formed
and no heat was evolved in this process. Similarly, no H2 was
found to adsorb over titania under the experimental condi-
tions of this study.
DISCUSSION
Only a few calorimetric studies have so far been re-
ported on the adsorption of H2 and CO over ruthenium
2TiO2 + CO → Ti2O3 + CO2↑, 1H = +80 kJ molꢀ1
.
TABLE 1
Differential Heat at Initial Stage of Adsorption (qi) of CO and H2
over Polycrystalline Ru Metal and Ru/TiO2 Catalyst at Different
Temperatures
The increasing contribution of this endothermic step, in
addition to the exothermic adsorption of CO at Ru sites,
explains the observed decrease in qd values with the rise in
temperature (Fig. 1c). The data on CO adsorption and its
conversion to CO2 may thus be correlated with the avail-
ability of reducible sites at the metal/support interfaces. A
similar decrease in the differential heat with the increase in
CO coverage was observed by Sharma et al. (21) over vari-
ous supported Pt catalysts at 403 K and has been attributed
to the decrease in adsorption strength due to presence of
lateral repulsive interactions between adsorbed species.
Ru metal
(kJ molꢀ1
Ru/TiO2 catalyst
(kJ molꢀ1
Catalyst
temperature
(K)
)
)
CO
H2
CO
H2
300
370
420
470
120
121
120
110
65
60
42
37
130
120
94
57
40
—
32
75

ADSORPTION OF CO AND H2 OVER Ru AND Ru/TiO2 CATALYST
421
A comparison of CO2 yields in Figs. 2e and 2f and Fig. 4 ported palladium and platinum catalysts (14, 24). Chou and
for successive pulse exposures shows that while the CO2 Vannice (24) have demonstrated that the heat of hydrogen
formation commenced from first CO pulse injection in case adsorption was independent of the support used but it in-
of Ru/TiO2 and the CO2 yield was almost constant during creased sharply with the decrease in average particle size.
successive CO pulse injections, the formation of CO2 on Ru The lowering of qd to ꢂ30 kJ molꢀ1 with the increase in
metal was observed only for the second and the third pulse surface coverage, irrespective of the catalyst temperature,
exposures. Also, no CO2 was formed from the subsequent reflects on the weakening of hydrogen bonding presumably
pulse exposures. These data thus suggest that a different due to multilayer adsorption (Fig. 6). A similar observation
mechanism was responsible for the CO2 formation over has been reported by Sharma et al. (21).
Ru metal. The possible reaction routes in this case may be
identified as
The qi for the hydrogen adsorption over bulk ruthenium
(ꢂ65 kJ molꢀ1) is of similar magnitude to that reported
for the bulk Pd (14, 24). The lowering of qd values during
successive pulse injections (Fig. 8) may be attributed to the
multilayer adsorption of H2 as discussed above.
In conclusion, we may thus state that the chemisorption
behavior of Ru is changed considerably on dispersion over
a reducible oxide support, such as titania. The H2 adsorp-
tion wasfound to be suppressed considerablyover Ru/TiO2,
particularly at the low reaction temperatures. The CO ad-
sorption, on the other hand, resulted in the reduction of
the titania support to a lower oxidation state, the extent
of which depended on the catalyst temperature. The heat
COad → C(ad) + O(ad) 1H = ꢀ8 kJ molꢀ1 [1]
CO(ad) + CO(ad) → CO2(g) + C(ad) 1H = ꢀ180 kJ molꢀ1
[2]
CO(ad) + CO(g) → CO2(g) + C(ad) 1H = ꢀ175 kJ molꢀ1
[3]
CO(g) + O(ad) → CO2(g) 1H = ꢀ143 kJ molꢀ1
CO(ad) + O(ad) → CO2(g) 1H = ꢀ140 kJ molꢀ1
[4]
[5]
–
evolved at Ru TiO2 interfaces during CO chemisorption at
The heat values given above indicate the order of the
magnitude and are either quoted from the literature (14, 15,
22) or evaluated from the heats of adsorption and formation
of different molecules over platinum group metals (23).
An almost similar amount of CO adsorption during the
first CO pulse exposure (Fig. 4) suggests that the mode of
CO adsorption over bulk Ru was independent of the cata-
lyst temperature. The CO2 yields from the successive CO
pulse injections suggest the involvement of the Eley–
Rideal-type reaction mechanism, i.e., steps [1], [3], and [4].
The low exothermicity of step 1 may explain why the ob-
served qd values were lower than the qd values expected for
steps [3] or [4]. Also, the carbon deposition during this step
may account for the progressive blockage of the adsorption
sites, leading to the reduced CO2 yield and the lower heat
of adsorption from the successive CO pulse doses (Figs. 4
and 5). The estimation of the quantitative contribution of
the above-mentioned individual steps is, however, not fea-
sible now.
metal sites may facilitate this process since no reduction of
titania was observed in the absence of Ru. This is in agree-
ment with the conclusions reached in our earlier study on
the tin oxide-supported platinum (1). In the case of bulk Ru
metal, the dissociation of CO and the subsequent reaction
of the surface-adsorbed species give rise to CO2 formation.
The experiments performed with the preadsorption of one
of the gases have unequivocally shown that, for both Ru
metal and Ru/TiO2, the CO adsorption remains uninhib-
ited by the preadsorbed H2, but the catalyst surface covered
with the CO was completely inaccessible to subsequent H2
adsorption.
ACKNOWLEDGMENTS
The authors thank Dr. S. R. Dharwadkar, Applied Chemistry Division,
for helpful discussions. We also thank Dr. L. D. Sharma, Indian Institute
of Petroleum, Dehradun, and Dr. A. D. Belapurkar of Chemistry Division
for the metal surface area measurements on our samples.
The data in Figs. 6 and 7 show that the hydrogen adsorp-
tion also depends strongly on the dispersion state of Ru.
Thus only a small (ꢂ10% ) fraction of hydrogen was ad-
sorbed on Ru/TiO2 at 300 K and it increased progressively
with the rise in catalyst temperature, the adsorbed frac-
tion being ꢂ40% at a catalyst temperature of 470 K. The
heat value per mole of H2 adsorbed, however, remained
the same indicating similar mode of hydrogen adsorption.
The near consistency in the amount of hydrogen adsorbed
from successive pulses indicates a weak hydrogen bonding
which is also reflected in a lower qd value (Fig. 6). The initial
heat of hydrogen adsorption on Ru/TiO2 at 57 kJ molꢀ1 is
similar to a value observed for the H2 adsorption over sup-
REFERENCES
1. Gangal, N. D., Gupta, N. M., and Iyer, R. M., J. Catal. 126, 13
(1990).
2. Gangal, N. D., Gupta, N. M., and Iyer, R. M., J. Catal. 140, 443
(1993).
3. Tripathi, A. K., and Gupta, N. M., J. Catal. 153, 208 (1995).
4. Bell, A. T., in “Proceedings, 9th International Congress on Catalysis,
Calgary, 1988” (M. J. Phillips and M. Ternan, Eds.), Vol. 5, p. 134.
Chem. Institute of Canada, Ottawa, 1988.
5. Biloen, P., and Sachtler, W. M. H., Adv. Catal. 30, 165 (1981).
6. Vannice, M. A., in “Catalysis: Science and Technology” (J. R.
Anderson and M. Boudart, Eds.), Vol. 3, p. 139. Springer-Verlag,
Berlin, 1982.

422
LONDHE AND GUPTA
17. Londhe, V. P., Kamble, V. S., and Gupta, N. M., J. Mol. Catal., in press.
7. Ponec, V., Catal. Rev. Sci. Eng. 18, 151 (1978).
8. Vannice, M. A., Catal. Rev. Sci. Eng. 14, 153 (1976).
18. Gupta, N. M., Kamble, V. S., Kartha, V. B., Iyer, R. M., Thampi, K. R.,
9. Mills, G. A., and Steffgen, F. W., Catal. Rev. 8, 159 (1974).
and Gratzel, M., J. Catal. 146, 173 (1994).
10. Storch, H. H., Columbic, N., and Anderson, R. A., “The Fischer– 19. Gupta, N. M., Kamble, V. S., Iyer, R. M., Thampi, K. R., and Gratzel,
Tropsch and Related Synthesis.” Wiley, New York, 1951.
11. Pichler, H., and Schulz, H., Chem. Ing. Tech. 42, 1162 (1970).
12. Gupta, N. M., Kamble, V. S., and Iyer, R. M., J. Catal. 60, 57 (1979).
13. McCarty, J. G., and Wise, H., J. Catal. 57, 406 (1979).
14. Cardona-Martinez, N., and Dumesic, J. A., in “Advances in Catalysis”
(D. D. Eley, H. Pines, and P. B. Weisz, Eds.), Vol. 38, p. 149. Academic
Press, New York, 1992.
M., J. Catal. 137, 473 (1992).
20. Solymosi, F., and Kno¨zinger, H., J. Catal. 122, 166 (1990).
21. Sharma, S. B., Miller, J. T., and Dumesic, J. A., J. Catal. 148, 198
(1994).
22. Shustorovich, E., in “Advances in Catalysis” (D. D. Eley, H. Pines, and
P. B. Weisz, Eds.), Vol. 37, p. 138. Academic Press, San Diego, 1990.
23. Weast, R. C. (Ed.), “CRC Handbook of Chemistry and Physics,”
p. D77. CRC Press, Boca Raton, FL, 1979.
15. Toyoshima, I., and Somorjai, G. A., Catal. Rev. Sci. Eng. 19, 105 (1979).
16. Guerrero-Ruiz, A., Appl. Catal. 55, 21 (1989).
24. Chou, P., and Vannice, M. A., J. Catal. 104, 1 (1987).