Sei sulla pagina 1di 12

Applied Catalysis A: General 260 (2004) 7586

Ca-doped chromium oxide catalysts supported on alumina for the


oxidative dehydrogenation of isobutane
G. Neri a , A. Pistone a , S. De Rossi b , E. Rombi c , C. Milone a , S. Galvagno a,
a

Dipartimento di Chimica Industriale ed Ingegneria dei Materiali, Universit di Messina, Salita Sperone 31, Vill. S. Agata, I-98166 Messina, Italy
b IMIP CNR Sezione Materiali Inorganici e Catalisi Eterogenea c/o Dipartimento di Chimica, Universit La Sapienza,
P.le Aldo Moro 5, 00185 Rome, Italy
c Dipartimento di Scienze Chimiche, Universit di Cagliari, Complesso Universitario di Monserrato, S.S. 554 Bivio per Sestu,
09042 Monserrato, CA, Italy
Received in revised form 8 October 2003; accepted 9 October 2003

Abstract
The oxidative dehydrogenation (ODH) of isobutane has been investigated on Ca-doped chromium oxide catalysts supported on -Al2 O3 .
The effect of Ca loading on the micro-structural properties of chromia catalysts was investigated by chemical analysis, X-ray diffraction
(XRD), scanning electron microscopy with elemental mapping (SEM-EDX), UV-Vis diffuse reflectance spectroscopy (DRS), temperature
programmed reduction (TPR), and micro-calorimetry of adsorbed NH3 . Cr3+ and Cr6+ species dispersed on alumina, as well as -Cr2 O3
and CaCrO4 crystallites, were found on the catalysts surface. The relative amount of the chromium species depends on the Ca loading. The
Cr3+ /Cr6+ ratio decreases on increasing the Ca loading due to the preferred formation of bulk chromate species. The Ca loading affects the
reducibility of the Cr6+ species and the acid sites strength distribution of the catalysts.
The catalytic activity in the ODH reaction of isobutane is enhanced in the presence of amounts of calcium <2 wt.%, then it decreases
with a further increase in the Ca content. The selectivity to isobutene follows the same trend showing a maximum (57% at 7% isobutane
conversion) on the sample promoted with 2 wt.% of calcium. The reported data suggest that the activity and selectivity to isobutene are
due to well-dispersed Cr6+ species on the alumina surface, whereas bulk chromates are less active and give mainly COx . On the basis of
micro-structural and catalytic results we proposed that the alkali promoter plays different roles depending on the Ca/Cr ratio. Besides causing
changes in surface acidity, calcium increases the amount of well-dispersed Cr6+ species in the range of low Ca/Cr ratios, enhancing the
catalytic properties, whereas, at higher Ca/Cr ratios, it acts as a poison promoting the formation of the less active and selective bulk chromate
species.
2003 Elsevier B.V. All rights reserved.
Keywords: Oxidative dehydrogenation; Isobutane; Chromium oxide catalysts

1. Introduction
The low cost of light alkanes have provided a great
effort to use them as feedstock for large-scale chemicals
production. The conversion of light saturated hydrocarbons
to unsaturated ones represents a challenging pursuit in both
economic and scientific terms.
Catalytic or thermal dehydrogenation of paraffin shows
relevant disadvantages. The alkane dehydrogenation is limited by thermodynamic equilibrium and in order to shift
it towards the formation of the dehydrogenated products,

Corresponding author. Tel.: +39-090-393134; fax: +39-090-391518.


E-mail address: galvagno@ingegneria.unime.it (S. Galvagno).

0926-860X/$ see front matter 2003 Elsevier B.V. All rights reserved.
doi:10.1016/j.apcata.2003.10.002

the reaction is carried out at high temperature (820920 K)


which causes a rapid deactivation of catalyst by extensive
coking and the need of frequent regeneration treatments,
making the process energy and capital intensive.
An attractive and alternative method for the production of olefins is represented by dehydrogenation of the
corresponding alkanes under oxidative conditions. This
reaction provides some relevant advantages with respect
to the non-oxidative processes due to engineering and
economic considerations. The reaction is exothermic and
able to proceed at much lower temperatures at which the
formation of coke is insignificant; furthermore, the conversion is no longer limited by thermodynamic constraints.
On the other hand, the main limitation of the oxidative

76

G. Neri et al. / Applied Catalysis A: General 260 (2004) 7586

dehydrogenation (ODH) is represented by the relevant formation of by-products. The formation of carbon oxides is
thermodynamically more favored than the formation of the
corresponding olefin and a rapid decrease in the selectivity
to the desiderated products with increasing the alkane conversion is often observed. A non-selective mechanism can
exist in which oxygen, from the lattice or activated from the
gas phase, can be inserted into the hydrocarbon and several
reaction steps can advance ultimately to carbon oxides.
From this point of view, the main effort in the scientific
investigations of the ODH of lower alkanes is to improve
the olefin yields. The key aspect of the oxy-dehydrogenation
technologies is, therefore, the development of catalysts capable of activating the CH bonds of the alkane molecule
in a flow of oxygen and capable of desorbing quickly the
alkene formed in the dehydrogenation step in order to avoid
a further oxidation. Acidic and basic properties as well as
the redox characteristics of the catalytic system seem to be
critical factors that can affect the performance of a selective
oxy-dehydrogenation catalyst [17].
Among low alkanes, the ODH of isobutane has received
in the last years an increasing interest as a route to obtain
isobutene, the key reactant for production of a variety of
chemicals, such as MTBE used as additive for gasoline to
regulate the octane number. Compared to other alkanes, in
particular ethane and propane, the ODH of isobutane has
received much less attention. The development of catalysts
able to activate, at low temperature and in presence of oxygen, the alkane molecule, to selectively promote the alkene
formation with high yields and simultaneously to avoid deep
oxidation of the substrate, represents the main goal [8].
The catalytic systems reported as promising for the
ODH of isobutane include ZnO/TiO2 systems, MgOV2 O5 ,
molybdates, heteropolycompounds, pyrophosphates, etc.
[917]. The maximum yields for these catalysts were
811% and selectivity varied between 50 and 80% for an
isobutane conversion of about 1020% and reaction temperature higher than 673 K. Recently, chromia-based catalysts have been studied in the ODH of isobutane because
of their favorable performances at relatively lower reaction temperatures [1824]. Moriceau et al. [20,24] reported
60% isobutene selectivity with 10% isobutane conversion
at 543 K for a binary CrCeO catalyst. Hoang et al. [25]
reported 70% isobutene selectivity with 10% isobutane
conversion for a chromia supported on lanthanum carbonate catalyst. In the last years, Grzybowska et al. [18] have
studied the ODH of isobutane on chromia-based catalysts
at temperatures between 473 and 673 K and reported selectivities up to 73% of isobutene obtained at 5% isobutane
conversion on chromia supported on titania and on K-doped
chromia supported on alumina; the authors also studied the
ODH of C2 C4 alkanes on chromia/alumina catalysts and
showed that it strongly depends on the structure of the hydrocarbon, with the total oxidation to carbon oxides being
the main reaction for ethane, propane, and n-butane and
the ODH to isobutene the main reaction for isobutane [26].

The authors also observed a strong correlation between the


Cr6+ content in the Cr/Al2 O3 catalysts and the activity in
the ODH of isobutane. Furthermore, a correlation among
acidbase properties of support, metaloxygen energy bond
and selectivity to olefin was also observed; in particular,
unsupported chromia catalyst showed lower selectivities
to isobutene because of a higher acidic character, a lower
CrOCr oxygen bond energy and a higher rate of oxygen
chemisorption than the Cr/Al2 O3 -based catalysts [23].
This paper aims to study the catalytic ODH of isobutane
to isobutene over chromium oxide catalysts supported on
-Al2 O3 and doped with Ca as an alkaline promoter. Alkali
metals have been indicated to change the acidbase characteristics of the catalysts and this may have a strong influence on the activity and selectivity in ODH reaction. High
selectivities to isobutene on K-doped Cr/Al2 O3 have been
explained with the increasing basic character of the surface
and, consequently, by weakening and facilitating the desorption of isobutene from the less acidic surface which prevents
further deep oxidation of the olefin to carbon oxides. However, contradictory results are reported in literature. In fact,
the modification of the acidbasic properties of a catalyst
can strongly influence not only the rate of desorption of the
alkene molecules, but also the nature and the dispersion of
the active sites, so affecting the overall catalytic properties.
Grabowski et al. [18] observed a positive or negative effect
of potassium doping on the activity of chromia-based catalysts depending on the type of support; in particular, the presence of the basic K additives on chromia/alumina catalysts
enhances strongly the selectivity to isobutene with respect
to the undoped catalysts. ODH of propane on zeolites-based
catalysts containing Ca was studied by Kubacka et al. [27].
No literature data have been instead found on the use of
calcium as a promoter for ODH reaction on chromia catalysts. In principle, Ca should behaves as other alkaline metals (Li and K) which have been more extensively studied, but
its different physico-chemical (ionic radius and charge) and
acidbase characteristics may modify differently the properties of chromium-based ODH catalysts.

2. Experimental
2.1. Catalysts preparation
Chromium oxide catalysts supported on alumina were
prepared by the incipient wetness technique. The following catalysts were prepared: ACr10 nominally containing
10 wt.% of chromium on -alumina, ACr10Ca1, ACr10Ca2,
ACr10Ca4, and ACr10Ca8 nominally containing also 1, 2, 4,
and 8 wt.% of calcium, respectively. -Alumina (grain size
100500 m, kindly provided by Sd Chemie MT, Novara,
Italy) was obtained by calcination in air of pseudoboehmite
Versal 250 La Roche at 1223 K for 4 h. The support was impregnated with comparable volumes of aqueous solutions of
the appropriate amounts of CrO3 (and CaCO3 , solubilized

G. Neri et al. / Applied Catalysis A: General 260 (2004) 7586

77

Table 1
Main characteristics of the chromia and Ca-doped chromia catalysts investigated
Catalyst code

Cr (wt. %)

Ca (wt. %)

Ca/Cr

Cr6+ (wt. %)

Cr3+ (wt. %)

Cr3+ /Cr6+

SBET (m2 /g)

Vpores (cm3 /g)

ACr10
ACr10Ca1
ACr10Ca2
ACr10Ca4
ACr10Ca8

9.4
9.2
9.3
9.2
9.7

0
1.23
1.81
3.62
6.42

0
0.13
0.19
0.39
0.66

1.5
2.0
2.4
5.0
8.8

8.0
7.1
6.9
3.9
1.1

5.33
3.55
2.88
0.78
0.13

94
83
84
82
80

0.35
0.33
0.34
0.27
0.31

Support: -alumina; SBET : 121 m2 /g; Vpores : 0.48 cm3 /g.

in the chromic acid medium, for the ACr10Cax samples).


Both chemicals were Carlo Erba reagent grade. The carefully stirred paste was dried overnight at 383 K and finally
calcined at 973 K for 12 h.
Table 1 lists the main characteristics of the prepared
catalysts.
2.2. Characterization
Chemical analyses of the total Cr (Table 1, column 1)
in the samples were carried out by atomic absorption (AA,
Varian SpectrAA-30). Samples were previously dissolved
by fusion with a mixture of KNO3 and Na2 CO3 (1:1 by
weight). On a different portion of the sample, the Cr6+
content (Table 1, column 4) was determined by AA after
several extractions with 10 M NaOH solution heated to incipient boiling. The same sample with the residual Cr3+
(Table 1, column 5) was then dissolved by fusion with the
mixture of KNO3 and Na2 CO3 and analyzed by AA. The
correspondence of the total amount of Cr and the sum of
the amounts of Cr6+ and Cr3+ could then be checked with
satisfactory results (Table 1).
Chemical analysis of Ca in the samples was carried out
by contacting them with 1 M HNO3 at room temperature for
30 min. The resulting limpid solution was then analyzed by
means of a Varian Liberty 200 inductively coupled plasma
analysis (ICP) spectrometer. Excepts for the ACr10Ca1
sample, Ca concentrations so determined are slightly lower
compared to nominal values (see Table 1), indicating that a
fraction of Ca is loss during the catalysts preparation.
Phase analysis was performed by X-ray diffraction (XRD)
using a Philips PW 1729 diffractometer equipped with a PC
for data acquisition and analysis (software APD-Philips).
Scans were taken with a 2 step of 0.01 , using Ni-filtered
Cu K radiation.
UV-Vis diffuse reflectance spectra (DRS) were taken in
the wavelength range 200800 nm (50,00012,500 cm1 )
with a Varian CARY 5E spectrometer equipped with a
PC for data acquisition and analysis and using PTFE as a
reference.
TianCalvet heat flow equipment (Setaram) was used
for micro-calorimetric measurements. Each sample was
pre-treated overnight at 673 K under vacuum (103 Pa)
before the successive introduction of the probe gas (ammonia). The equilibrium pressure relative to each adsorbed

amount was measured by means of a differential pressure


gauge (Datametrics). The run was stopped at a final equilibrium pressure of 133.3 Pa. The adsorption temperature
was maintained at 353 K, in order to limit physisorption.
Temperature programmed reduction (TPR) profiles were
obtained on a TPD/R/O 1100 apparatus (ThermoQuest),
under the following conditions: sample weight 4045 mg,
heating rate (from 313 to 1173 K) 20 K/min, flow rate
30 cm3 /min, H2 5 vol.% in N2 ; the hydrogen consumption
was monitored by a thermal conductivity detector (TCD).
Textural analyses were carried out on a Sorptomatic 1990
System (Fisons Instruments), by determining the nitrogen
adsorptiondesorption isotherms at 77 K. Before analysis,
the samples were heated overnight under vacuum up to
473 K (heating rate = 1 K/min).
Scanning electron microscopy (SEM) with elemental
mapping images of powder samples mounted on an aluminum holder were obtained on a JEOL JSM-5600 LV
microscope equipped with an EDX (Oxford) analyzer. The
quantitative analysis was carried out at 20 kV by using
the SEMQUANT software applying the Z.A.F. correction
procedure.
2.3. Catalytic activity
The activity of the catalysts in the ODH of isobutane was
measured in a fixed bed down-flow apparatus operated at
atmospheric pressure equipped with a quartz U-tube and an
internal coaxial thermocouple connected with a PID temperature controller. The ODH of isobutane was studied by varying both the temperature range (523673 K) and the contact
time (16 s), catalytic runs were carried out using oxygen
and isobutane in the ratio 2:1 diluted in helium (isobutane =
6%, oxygen = 12%, total flow = 320 cm3 /min) and using
0.20.4 g of catalyst. The catalyst was mixed with an equal
amount of granular quartz and loaded into the micro-reactor;
additional amounts of granular quartz were placed upon the
catalyst bed to reduce the dead volume of the reactor. A gas
chromatograph (HP 6890) was used for an on-line analysis
of both the feed and the product streams. The hydrocarbons
were separated by HP-plot column (HP-Plot Alumina M
deactivated 50 m 0.53 mm 15 m) and analyzed with
a FID detector, while O2 , CO, and CO2 were separated
by a molecular sieves and Hayesep columns (13 molecular sieves: 10 ft 1.8 in., Hayesep Q: 10 ft 1.8 in.) and

78

G. Neri et al. / Applied Catalysis A: General 260 (2004) 7586

analyzed by TCD detector. Conversion and selectivity were


defined as follows:


moles of isobutane reacted
conversion = Y =
100;
moles of isobutane in the feed
selectivity

= Si =

moles of isobutane converted to product i


moles of isobutane reacted
100,

all catalysts are dominated by broad peaks of the -alumina


support. On the undoped catalyst, -Cr2 O3 was the only
chromium phase detected. No chromate phases were observed indicating that they are absent or below the detection
limit of XRD. The addition of Ca upto about 2 wt.% does

where i = i-C4 H8 , CO, CO2 , and others (CH4 , C2 H6 ,


C3 H8 , etc.). Isobutene, CO, and CO2 were found to be the
main reaction products, while the amounts of other degradation products and of oxygenates were negligible. Blank
tests without catalyst in the reactor showed no conversion of
isobutane in the reaction temperature range considered, allowing to rule out the occurence of homogeneous reactions
to a significant extent. No deactivation of the catalysts was
observed during the measurements.

3. Results
3.1. Catalysts characterization
Ca-promoted chromia catalysts were prepared by impregnation of the respective salt precursors of a -alumina support having a specific surface area (SA) of 121 m2 /g and a
pore volume of 0.48 cm3 /g. The SA and pore volume decrease significantly upon addition of 10 wt.% of chromium
(see Table 1). On the contrary, the loading of Ca, after an
initial small decrease, does not affects significantly these
parameters.
Fig. 1 shows XRD patterns of the investigated catalysts.
Before discussing them, it should be mentioned that many
peaks of -Cr2 O3 and CaCrO4 are coincident. The spectra of

Fig. 1. XRD of the investigated catalysts.

Fig. 2. SEM micro-graphs of the undoped ACr10 catalyst: (a) lower


magnification; (b) and (c) high magnification.

G. Neri et al. / Applied Catalysis A: General 260 (2004) 7586

79

Fig. 3. SEM micro-graphs and EDX elemental analysis of the surface of Ca-doped catalysts. The numbers indicate points where the corresponding EDX
spectrum was collected: (a) ACr10Ca2 catalyst and (b) ACr10Ca8 catalyst.

not modify the XRD spectra. On samples containing more


calcium, the -Cr2 O3 peaks disappear while new peaks, related to calcium chromate, CaCrO4 , increase with calcium
content.
A SEM analysis was carried out to investigate the morphology of the catalysts. Fig. 2a reports a low magnification
view showing the typical granulometric distribution of the
alumina used as support. Fig. 2b and c present micro-graphs
taken at higher magnification of the surface of the ACr10
sample. Elemental EDX-mapping has shown chromium to
be highly dispersed and distributed homogeneously on all

Fig. 4. TPR patterns of the undoped and Ca-doped catalysts. Heating rate,
20 K/min; reducing mixture, 5% H2 /N2 at 30 ml/min; mcat = 40 mg.

surface of alumina. Crystalline particles of -Cr2 O3 of about


1 m in size were also imaged and identified.
Samples with low promoter loading (<2 wt.%) have the
same morphological characteristics of the undoped chromia
catalyst (Fig. 3a). On samples with higher promoter loading,
large crystallites were observed (Fig. 3b). EDX analysis has
shown that they contain chromium, calcium, and oxygen in
the ratio expected for CaCrO4 . No -Cr2 O3 particles have
been instead observed.
TPR patterns of the investigated catalysts are shown in
Fig. 4. A single reduction peak, centered at 653 K, was registered on the undoped chromia catalyst. As the level of the

Fig. 5. DRS-UV-Vis analysis of the investigated samples.

80

G. Neri et al. / Applied Catalysis A: General 260 (2004) 7586

Fig. 6. Cr species content as a function of Ca loading.

Fig. 7. Micro-calorimetric analysis of the undoped and Ca-doped catalysts. The curve related to support was also reported for comparison.

promoter increases, the reduction peak shifts to higher temperature and at the same time decreases of intensity, while
a second peak at 763 K appears. On samples ACr10Ca4
and ACr10Ca8, these low temperature peaks gradually dis-

appear whereas new strong peaks at higher temperature


(873973 K) were observed. The peaks at 673 and 763 K can
be attributed to one step reduction of monochromate species,
Cr6+ , stabilized on alumina surface and/or interacting with

Table 2
Acid sites strength distribution
Catalyst code

Weak acid sites (70120 kJ/mol)

Medium acid sites (120150 kJ/mol)

Strong acid sites (>150 kJ/mol)

Total acid sites

Alumina
ACr10
ACr10Ca1
ACr10Ca2
ACr10Ca4
ACr10Ca8

1.11
1.28
1.37
1.22
0.46
1.10

0.47
0.82
0.27
0.28
0.08
0.10

0.19
0.57
0.21
0.23
0.23
0.21

1.79
2.68
1.86
1.75
0.78
1.41

Concentration of sites is expressed as amount of NH3 adsorbed per unit surface area (mol/m2 ).

G. Neri et al. / Applied Catalysis A: General 260 (2004) 7586

Ca-modified sites, respectively, to Cr3+ species. The high


temperature peaks (between 873 and 973 K) are instead related to the reduction of bulk chromate species.
A quantitative analysis of the TPR patterns, on the assumption that H2 is consumed only in the reduction of Cr6+
to Cr3+ species, has shown that the total hydrogen consumption increases with increasing the Ca content. This means
that overall Cr6+ species increase on addition of the alkali
promoter. The hydrogen consumption related to the low temperature peaks (653 and 763 K), due only to dispersed Cr6+
species, increase slightly with Ca loading up to 2 wt.%, then
decrease on samples ACr10Ca4 and ACr10Ca8. Hydrogen
consumed during the reduction has been quantitatively determined. An H/Cr6+ ratio of about 1.5 has been calculated
for all the investigated samples, indicating that only a fraction (50%) of the total Cr6+ was reduced to Cr3+ under
the adopted operating conditions. The same value for the
H/Cr6+ ratio has been obtained by Cherian et al. [36] for a
sample containing about 7 wt.% of Cr on alumina, at variance with Grzybowska et al. [23] who reported a complete
reduction for a Cr/Al catalysts (Cr 8 wt.%), indicating that
operative conditions strongly affect TPR results.
DRS spectra are shown in Fig. 5. The ACr10 and ACr10Ca
catalysts show adsorption bands centered at 260, 380, 470,
and 600 nm. On the basis of literature data [2831] the following attributions have been made: the bands at 260 and
380 nm are related to the charge transfer transitions O Cr
typical of Cr6+ , whereas the bands at 470 and 600 nm are
due to dd transitions of Cr3+ species in octahedral symmetry (A2g T1g and A2g T2g , respectively). A very
weak band around 700 nm was also noted. The attribution
of this band is of more difficult interpretation. We can speculate, it can be due to Cr5+ ions, as reported by Zecchina
et al. [28] for a silica-supported chromium oxide. In comparison to the spectrum of ACr10, no significant modifications
were observed on promoted catalysts with low (<2 wt.%)
Ca loading. A remarkable decrease of the Cr3+ bands (par-

81

ticularly that at 600 nm) was instead observed on ACr10Ca4


and ACr10Ca8 samples. Correspondingly, an increase in intensity was observed for the bands at 260 and 380 nm, interpreted as due to the formation of bulk chromate particles
on these latter samples.
The indication given by the above characterization techniques are fully in agreement with chemical analysis data
(see Table 1). The Cr3+ /Cr6+ ratio in fact was found to
decrease with the Ca loading. Plotting data of the different species versus the Ca content, two regions are clearly
evidenced (Fig. 6). In the first region, ranging from 0 to
<2 wt.% of Ca, variations in the amount of Cr3+ and Cr6+
species detected on catalysts are little, whereas at higher Ca
content Cr3+ species decrease strongly with a correspondent
strong increase of Cr6+ species.
In order to characterize the acidic properties of the surface, a micro-calorimetric study of adsorbed ammonia was
carried out on all the investigated samples. Results are presented in Fig. 7 where the differential heats of adsorption,
Qdiff , are plotted versus NH3 uptake. An increase in the
initial value of Qdiff can be noted when chromium is added
to the pure support (263 and 245 kJ/mol for ACr10 and alumina, respectively), on the contrary, a pronounced decrease
to lower values (in the range of 165185 kJ/mol) is observed
for Ca-doped catalysts. Table 2 lists the conventional acid
sites distribution as determined by the calorimetric results,
according to the differential heats of adsorption. Surface
acidity increases on addition of chromium, especially in
terms of strong and medium sites amounts, which are ca.
50% higher on ACr10 than on alumina. On the other hand,
the acidity of the chromia catalyst greatly decreases when
Ca is added as promoter, the acid sites strength distribution being influenced by the Ca content. It is interesting to
note that increasing Ca concentration to values higher than
1 wt.% does not affect the number of strong acid sites, which
markedly diminishes going from ACr10 to ACr10Ca1 sample and then remains fairly constant. A drastic decrease in

Fig. 8. Conversion of isobutane and selectivity to oxidation products as a function of temperature for the ACr10 catalyst.

82

G. Neri et al. / Applied Catalysis A: General 260 (2004) 7586

the number of both weak and medium sites can be instead


observed when Ca content is increased from 2 to 4 wt.%.
Finally, it can be noted that, compared to the ACr10Ca4
sample, the ACr10Ca8 catalyst exhibits a greater amount
of weak acidity.
3.2. Catalytic activity
The catalytic behavior of the Cr/Al2 O3 and Ca-doped
Cr/Al2 O3 catalysts in the ODH of isobutane was first
investigated as a function of reaction temperature. Preliminary experiments have shown that no deactivation of
the catalysts occurs during the measurements. Isobutane
and oxygen conversion, and products distribution over the

ACr10 catalyst as a function of the reaction temperature


are shown in Fig. 8. Experimental conditions were: Wcat =
0.4 g, He:O2 :i-C4 H10 (molar ratios) = 13:2:1, total flow =
320 cm3 /min, GHSV (based on isobutane) = 2300 h1 . Under these conditions, we observed an isobutane conversion
of 6% already at 523 K. Isobutene, propene, and carbon
oxides (CO and CO2 ) were the main products, while the
amount of other products such as CH4 , C2 H4 , acetone, etc.
was negligible. Fig. 9 shows the catalytic performance data
for the ACr10Ca2 and the ACr10Ca8 catalysts. It can be
observed a change in the products distribution with respect
to the undoped catalyst. In particular, a strong enhancement in the CO2 formation and correspondingly a decrease
in CO formation with Ca loading was noticed. Isobutene

Fig. 9. Conversion of isobutane and selectivity to oxidation products as a function of temperature: (a) Acr10Ca2 and (b) Acr10Ca8.

G. Neri et al. / Applied Catalysis A: General 260 (2004) 7586

83

Fig. 10. Effect of contact time on the conversion of isobutane and selectivity to oxidation products: () isobutane; ( ) isobutene; () CO; () CO2 .

formation decreases only slightly. As seen, for all catalysts,


the selectivity to isobutene decreases with increasing the
temperature (increasing isobutane conversion).
To compare the selectivities of the different catalysts at a
fixed temperature, a series of experiments were performed at
598 K by varying the residence time. Fig. 10 shows the effect
of residence time on the catalytic performance on ACr10
catalyst. The isobutane conversion increases on increasing
the residence time. The selectivity to CO and CO2 follow this
trend whereas the selectivity to isobutene decreases. Such
behavior was similar for all the investigated catalysts.
Fig. 11 shows the overall activity in the ODH of isobutane, and the formation rate of isobutene, CO, and CO2 , respectively, as a function of the alkali content. For low Ca
loading, an increase of the overall activity and formation rate
of isobutene and CO2 was observed, but a further increase
in the Ca content reverse this trend. The formation rate of
CO instead decreases monotonically with Ca loading.
The selectivityconversion plot reported in Fig. 12 seems
indicate that isobutene and CO2 are primary products of
ODH of isobutane on these catalysts while carbon monoxide shows, extrapolating the selectivityconversion curve to
low conversion, a zero intercept typical of secondary products. The data agree with the general mechanism of ODH of
isobutane over chromia and vanadia catalysts proposed by
other authors where the formation of CO was mainly due
to combustion of adsorbed olefinic intermediates, while the
formation of CO2 was due to both this step and also the direct combustion of the reactant [32,33]. However, a more
detailed investigation at lower isobutane conversion is necessary to clarify this point.
Fig. 13 reports the selectivity to reaction products at the
temperature of 698 K and at a 7% conversion of isobutane,
as a function of the alkali content. At low loading of Ca
(<2 wt.%) an enhancement in the isobutane selectivity, from
52% on the ACr10 catalyst to 57% on the ACr10Ca2 cata-

Fig. 11. Effect of Ca loading on the overall activity and formation rate
of oxidation products: ( ) isobutane; () isobutene; () CO; () CO2 .

lyst, was registered, whereas the selectivity to CO2 remains


almost constant. Increasing the Ca content the selectivity to
CO2 increases whereas that of isobutane decreases. The CO
selectivity shows a constant decrease with the Ca loading
for all the catalysts.

4. Discussion
The characterization data on the unpromoted ACr10 catalysts have shown a complex micro-structure. It is well

84

G. Neri et al. / Applied Catalysis A: General 260 (2004) 7586

Fig. 12. Selectivity to main oxidation products vs. the isobutane conversion: () isobutene; () CO; () CO2 .

recognized that structural properties of supported chromium


oxide catalysts depend on many variables such as chromium
loading, heat treatment, support used, etc. [18,20,2325].
The beneficial role of the support was related to stabilization of both low coordinated Cr3+ ions and highly oxidized
species Cr6+ . Increasing the Cr loading over the monolayer,
bulk phases are also found on supported chromia catalysts.
The relative amount of chromium species on the catalyst
affects strongly its catalytic properties.
Cr3+ , in both dispersed or bulk phases, and dispersed
Cr6+ are the main species detected on the ACr10 catalyst. The high value of the Cr3+ /Cr6+ ratio (>5) indicate
that the Cr3+ species prevail on the undoped catalyst and

exist on the catalyst surface under at least two forms: (i)


well-dispersed Cr3+ species anchored to the alumina support; (ii) Cr3+ species in amorphous or crystalline -Cr2 O3
[34].
The relative amount of chromium in different species
and/or oxidation states drastically changes on addition of
the alkali promoter. However, at Ca loading <2 wt.%, small
changes in the distribution of chromium species were registered. At higher loading, Ca favors the formation of CaCrO4 ,
leading to a drastic decrease of Cr3+ species and giving rise
to Cr6+ in the calcium chromate phase. Cavani et al. [34]
report that also addition of K leads to the formation of a
potassium chromate phase, the formation of which occurs

G. Neri et al. / Applied Catalysis A: General 260 (2004) 7586

Fig. 13. Effect of Ca loading on the selectivity to oxidation products


(isobutane conversion is 7%): () isobutene; () CO; () CO2 .

at the expense of the dispersed Cr3+ phase and crystalline


-Cr2 O3 .
The isobutane ODH reactivity studies have shown that
the activity and selectivity of the supported chromium oxide
catalysts depend on the calcium loading. As the Ca loading is increased the ODH activity goes through a maximum
for the ACr10Ca1 catalyst, then it decreases. Alkali metals
have been indicated as promoters of activity and selectivity
for ODH reaction [18]. However, so far little is known how
the alkali promoters affect the activity and selectivity. The
promotion of activity and selectivity to isobutene at low Ca
loading can be attributed to many factors: (i) an increase in
the number of the most active and selective sites; (ii) the
blocking of unselective strong acid sites that favor the formation of COx ; (iii) the decrease in the acidity and the increase
in basicity, thus facilitating the desorption of isobutene from
the catalyst surface and preventing it from further oxidation
to carbon oxides [35].
Also the exact nature of chromium species active in
ODH reaction is still a matter of discussion. Moriceau
et al. [24] found that the activity of Cr2 O3 /CeO2 catalysts increases linearly with the Cr content in the range
where only well-dispersed Cr6+ species are present. They
suggested then that these species are the active sites for
ODH of isobutane. Cherian et al. [36] on the assumption
that a redox mechanism occurs, indicate in the redox pair
Cr6+ Cr3+ the active sites for ODH of propane. In any
case, regardless of the oxidation state, chromium species in
crystalline phases are less active and address the reaction

85

Fig. 14. Activity and selectivity to isobutene as a function of the area


under the low temperature TPR peaks.

towards total oxidation than the two-dimensional surface


chromium sites which allow selective addition of oxygen in
the organic molecule [22,24,37,38].
Taking into account these contributions, we can suggest
that the increase of activity at low Ca/Cr ratios is due to an
increase of dispersed Cr6+ species. This is supported from
data shown in Fig. 14 where the formation rate and selectivity to isobutene of the catalysts is reported as a function of
the hydrogen, namely HLOW , consumed at low temperature,
<773 K, in TPR experiments. HLOW can be taken as a measure of the amount of only dispersed Cr6+ species, being
Cr6+ in bulk chromates reduced at higher temperature. The
quite linear correlation found confirms that dispersed Cr6+
species can be considered the active species in Ca-promoted
chromia ODH catalysts.
The decrease in the catalytic properties (lower activity and
selectivity to isobutene) at higher calcium loading indicates
a poisoning effect of the alkaline promoter on the active
centers for hydrocarbon activation. This overdoping effect
was also noted previously for other alkali-doped systems
for dehydrogenation and ODH reactions [21,39,40]. This
occurs through a decrease of the amount of the active sites
(dispersed Cr6+ ) due to the preferred formation of the less
active bulk chromate phases.
The redox and acidbase characteristics of the catalysts,
may also affect the catalytic performance in ODH reactions
[6]. TPR experiments have shown that the reducibility of
Cr6+ species are related to the Ca loading. The temperature

86

G. Neri et al. / Applied Catalysis A: General 260 (2004) 7586

of the maximum reduction peak, Tmax , increases and then


the reducibility decreases, with the Ca content. At low alkali
doping, a decrease of reducibility corresponds to an increase
of activity and selectivity to isobutene, whereas at higher Ca
loading, when substantial structure modifications occur, this
trend is reversed. The effect of Ca doping on the acidbase
properties is more complex. The alkaline promoter strongly
decreases the acidity of the chromia catalyst, particularly
decreasing the medium and strong acid sites. As regards
the acidic properties, no clear correlations have been found
between the surface acidity and the catalytic behavior of the
investigated samples.

5. Conclusions
Ca-promoted chromium oxide catalysts have been prepared, characterized, and tested in the ODH of isobutane.
The presence of calcium significantly alters the active sites
distribution, promoting the formation of dispersed Cr6+
species at low Ca content, whereas higher loading leads to
the formation of bulk Ca-chromates species. When tested
in isobutane ODH, the activity and selectivity to isobutene
show an increase at low alkali content followed by a sharper
decline. A linear relationship is obtained when the formation rate and the selectivity to isobutene is plotted versus
the concentration of dispersed Cr6+ active sites obtained
from H2 uptake at low temperature in TPR experiments.
On these basis, we suggest that Ca plays different roles
in the ODH reaction of isobutane:
(1) At loading upto 2 wt.% it increases the amount of dispersed Cr6+ species at expense of the Cr3+ species.
This explain the enhancement of activity on low-loaded
Ca-doped catalysts.
(2) It decreases the acidity and increases the basicity of
chromia catalysts. Consequently, the alkali weakens the
adsorption of formed isobutene, thus facilitating its desorption as a product. This effect being maximized at
low Ca/Cr ratios where an increase of selectivity was
registered.
(3) It favors the formation, at higher Ca/Cr ratios, of less
active and selective chromate species, so negatively affecting the catalytic properties.

References
[1] H.H. Kung, Adv. Catal. 40 (1994) 1.
[2] A.A. Lemonidou, L. Nalbandian, I.A. Vasalos, Catal. Today 61 (2000)
333.
[3] G.C. Bond, S.F. Tahir, Appl. Catal. A: Gen. 71 (1991) 1.
[4] J.C. Vedrine, J.M.M. Millet, J.C. Volta, Catal. Today 32 (1996) 115.
[5] B. Grzybowska-Swierkosz, Top. Catal. 21 (1-3) (2002) 35.

[6] M.M. Bettahar, G. Costentin, L. Savary, J.C. Lavalley, Appl. Catal.


A: Gen. 145 (1996) 1.
[7] F. Cavani, F. Trifir, Catal. Today 36 (1997) 431.
[8] E.A. Mamedov, V. Cortes Corberan, Appl. Catal. A: Gen. 127 (1995)
1.
[9] Y. Takita, K. Sano, T. Muraya, H. Nishiguchi, N. Kawata, M. Ito,
T. Akbay, T. Ishihara, Appl. Catal. A: Gen. 170 (1998) 23.
[10] C.R. Dias, R. Zavoianu, M.F. Portela, Catal. Comm. 3 (2002) 85.
[11] Y. Takita, K. Sano, K. Kurosaki, N. Kawata, H. Nishiguchi, M. Ito,
T. Ishihara, Appl. Catal. A: Gen. 167 (1998) 49.
[12] S.M. Al-Zaharani, N.O. Elbashir, A.E. Abasaeed, M. Abdulwahed,
Catal. Lett. 69 (2000) 65.
[13] Y.J. Zhang, I. Rodriguez-Ramos, A. Guerrero-Ruiz, Catal. Today 61
(2000) 377.
[14] B. Sulikowski, Z. Olejniczak, E. Wloch, J. Rakoczy, R.X. Valenzuela,
V. Cortes-Corberan, Appl. Catal. A: Gen. 232 (2002) 189.
[15] F. Cavani, C. Comuzzi, G. Dolcetti, E. Etienne, R.G. Finke, G.
Selleri, F. Trifir, A. Trovarelli, J. Catal. 140 (1993) 226.
[16] Y. Takita, K. Kurosaki, Y. Mizuhara, Chem. Lett. 2 (1993) 335.
[17] E. Tempesti, A. Kaddouri, C. Mazzocchia, Appl. Catal. A: Gen. 169
(1998) L3.
[18] R. Grabowski, B. Grzybowska, J. Sloczynski, K. Wcislo, Appl. Catal.
A: Gen. 144 (1996) 335.
[19] S.M. Al-Zahrani, N.O. Elbaashir, A.E. Abasaeed, M. Abdulwahed,
Ind. Eng. Chem. Res. 40 (2001) 781.
[20] P. Moriceau, B. Grzybowska, Y. Barbaux, G. Wrobel, G. Hecquet,
Appl. Catal. A: Gen. 168 (1998) 269.
[21] G. Karamullaoglu, S. Onen, T. Dogu, Chem. Eng. Proc. 41 (2002)
337.
[22] M. Cherian, M.S. Rao, W.T. Yang, J.M. Jehng, A.M. Hirt, G. Deo,
Appl. Catal. A: Gen. 233 (2002) 21.
[23] B. Grzybowska, J. Sloczynski, R. Grabowski, K. Wcislo, A. Kozlowska, J. Stoch, J. Zielinski, J. Catal. 178 (1998) 687.
[24] P. Moriceau, B. Grzybowska, L. Gengembre, Y. Barbaux, Appl.
Catal. A: Gen. 199 (2000) 73.
[25] M. Hoang, J.F. Mathews, K.C. Pratt, J. Catal. 171 (1997) 320.
[26] B. Grzybowska, J. Sloczynski, R. Grabowski, L. Keromnes, K.
Wcislo, T. Bobinska, Appl. Catal. A: Gen. 209 (2001) 279.
[27] A. Kubacka, E. Wloch, B. Sulikowski, R.X. Valenzuela, V. Corts
Corbern, Catal. Today 61 (2000) 343.
[28] A. Zecchina, E. Garrone, G. Ghiotti, C. Morterra, F. Borrello, J.
Phys. Chem. 79 (1975) 966.
[29] A. Cimino, B.A. De Angelis, A. Luchetti, G. Minelli, J. Catal. 45
(1976) 316.
[30] S. Khaddar-Zine, A. Ghorbel, C. Naccache, J. Mol. Catal. A: Chem.
150 (1999) 223.
[31] R.L. Puurunen, B.M. Weckhuysen, J. Catal. 210 (2002) 418.
[32] J. Sloczynski, B. Grzybowska, R. Grabowski, A. Kozlowska, K.
Wcislo, Phys. Chem. Chem. Phys. 1 (1999) 333.
[33] K. Chen, A. Khodakov, J. Young, A.T. Bell, E. Iglesia, J. Catal. 186
(1999) 325.
[34] F. Cavani, M. Koutyrev, F. Trifir, A. Bartolini, D. Ghisletti, R. Iezzi,
A. Santucci, G. Del Piero, J. Catal. 158 (1996) 236.
[35] R. Grabowski, B. Grzybowska, K. Samson, J. Sloczynski, J. Stoch,
K. Wcislo, Appl. Catal. A: Gen. 125 (1995) 1299.
[36] M. Cherian, M.S. Rao, A.M. Hirt, I.E. Wachs, G. Deo, J. Catal. 211
(2002) 482.
[37] C.M. Pradier, F. Rodrigues, P. Marcus, M.V. Landau, M.L. Kaliya,
A. Gutman, M. Herskowitz, Appl. Catal. B: Env. 27 (2000) 73.
[38] S. Khaddar-Zine, A. Ghorbel, C. Naccache, J. Sol-Gel Sci. Technol.
19 (2000) 637.
[39] R.B. Watson, U.S. Ozkan, J. Catal. 191 (2000) 12.
[40] R.B. Watson, U.S. Ozkan, J. Mol. Catal. 194 (2003) 115.

Potrebbero piacerti anche