The role of Rh on a substituted Al Anderson heteropolymolybdate: Thermal and hydrotreating catalytic behavior

Article abstract of DOI:10.1016/j.tca.2006.04.014

The influence of Rh heteroatom on the molybdenum reducibility in the Anderson-type heteropolyoxomolybdate structure of formula (NH₄)₃[RhMo₆O₂₄H₆]·7H₂O was investigated by means of TPR technique. With comparative purposes, the thermal behavior in non-reducing conditions was also carried out by means of TGA-DTA studies. The study was performed by XRD, SEM, EDAX and FTIR-Raman techniques. Results were related to those preliminary measurements over other XMo₆ Anderson phases. Likewise, Rh(III)-Al(III) formal replacement in the RhMo₆ structural arrangement was proved. The formation of a solid solution in a limited range of composition (up to 0.25 Rh) was established in order to explore the catalytic performance of γ-Al₂O₃ supported planar heteropolyoxomolybdate, aiming at optimizing the noble metal content in the catalytic system. Preliminary measurements of RhMo₆/γ-Al₂O₃ and (Rh, Al)Mo₆/γ-Al₂O₃ activity for HDS and HYD processes were also performed. These results were compared to those obtained for CoMo₆/γ-Al₂O₃ system in similar operating conditions and other conventional catalytic systems. The potentiality and scope of RhMo₆ catalytic system for the HDS and HYD processes were analyzed.

Full text of DOI:10.1016/j.tca.2006.04.014

Thermochimica Acta 447 (2006) 22–29
The role of Rh on a substituted Al Anderson heteropolymolybdate:
Thermal and hydrotreating catalytic behavior
Carmen I. Cabelloa,∗,1, Mercedes Munoz , Irma L. Botto , Edmond Payen
a
b
c
˜
a Centro de Investigacio´n y Desarrollo en Ciencias Aplicadas Dr. Jorge J. Ronco,
CINDECA (CONICET-Universidad Nacional de La Plata), Calle 47 N◦ 257, (1900) La Plata, Buenos Aires, Argentina
b Centro de Qu´ımica Inorga´nica CEQUINOR (CONICET-Universidad Nacional de La Plata), (1900) La Plata, Buenos Aires, Argentina
c Unite´ de Catalyse et de Chimie du Solide, UCCS UMR CNRS 8181,
Universite´ des Sciences et Technologies de Lille, Baˆt. C3, 59655 Villeneuve d’Ascq Cedex, France
Received 11 January 2006; received in revised form 21 April 2006; accepted 25 April 2006
Available online 4 May 2006
Abstract
The influence of Rh heteroatom on the molybdenum reducibility in the Anderson-type heteropolyoxomolybdate structure of formula
(NH4)3[RhMo6O24H6]·7H2O was investigated by means of TPR technique. With comparative purposes, the thermal behavior in non-reducing
conditions was also carried out by means of TGA–DTA studies. The study was performed by XRD, SEM, EDAX and FTIR–Raman techniques.
Results were related to those preliminary measurements over other XMo6 Anderson phases. Likewise, Rh(III)–Al(III) formal replacement in the
RhMo6 structural arrangement was proved. The formation of a solid solution in a limited range of composition (up to 0.25 Rh) was established
in order to explore the catalytic performance of ␥-Al2O3 supported planar heteropolyoxomolybdate, aiming at optimizing the noble metal content
in the catalytic system. Preliminary measurements of RhMo6/␥-Al2O3 and (Rh, Al)Mo6/␥-Al2O3 activity for HDS and HYD processes were also
performed. These results were compared to those obtained for CoMo6/␥-Al2O3 system in similar operating conditions and other conventional
catalytic systems. The potentiality and scope of RhMo6 catalytic system for the HDS and HYD processes were analyzed.
© 2006 Elsevier B.V. All rights reserved.
Keywords: Thermal behavior of heteropolyoxomolybdates; Rh and Al Anderson phases; Rh–Mo/␥-Al2O3; Rh promoting effect; Hydrotreating catalysts
1. Introduction
to traditional systems. The advantages are particularly related to
the topology, well-defined chemical composition and electronic
properties of these polymeric arrays [5–9].
Bimetallic systems containing molybdenum and transition
metals such as Co, Cr, Ni or Cu have proved to be interesting cat-
alytic materials in a great amount of industrial processes [1]. The
addition of noble metals (belonging to VIII B group) enhances
the catalytic performance. So, the reducibility of metal compo-
nents in the mixture is increased by synergetic effect, leading to
a better activity and selectivity [2,3]. The preparation of these
catalysts is usually based on a support co-impregnation process
by using ammonium heptamolybdate (AHM) and metal nitrate
solutions, respectively [4].
Anderson-type heteropolyoxomolybdates of general formula
[XMo6O24Hx]n− (namely XMo6), where X is a di-, tri-, penta-
or hexavalent metal and x = 6 or 0, were characterized by the
presence of the X heteroatom in a central octahedral cavity
of the crown constituted by six edge-sharing MoO6 octahedra,
showing a D3d planar configuration [10]. Thermal and spectro-
scopic behavior of different Anderson phases were studied and
the catalytic applications of the CoMo6 were analyzed in HDS
processes [11–18].
For the last 10 years, the use of heteropolyoxomolybdates
(HPOMs) in catalysis has proved to be an interesting alternative
Although the design of catalysts based on noble metals
must be carefully considered because of their high cost, the
adequate combination with other common metals seems to
be a reasonable solution. In this sense, Anderson structure
allows a good alternative to obtain a desirable combination.
Particularly Rh(III) and Al(III) as heteroatoms in this struc-
ture form isomorphous heteropolyanions. In this work, the
∗
Corresponding author. Fax: +54 2214220288.
E-mail address: ccabello@quimica.unlp.edu.ar (C.I. Cabello).
Member of the research staff of CICPBA.
1
0040-6031/$ – see front matter © 2006 Elsevier B.V. All rights reserved.
doi:10.1016/j.tca.2006.04.014

C.I. Cabello et al. / Thermochimica Acta 447 (2006) 22–29
23
thermal behavior of the (NH4)3[Rh(III)Mo6O24H6]·7H2O salt,
Anderson-type (identified as RhMo6), was studied in oxidiz-
ing and reducing conditions, by means of DTA, TGA and TPR
techniques. On the other hand, in order to analyze the pro-
moting effect of Rh content in the planar configuration, the
(NH4)3[Rh(III)xAl(III)1−xMo6O24H6]·7H2O (or [Rh, Al]Mo6)
solid solution was also prepared. The study was carried out
by means of XRD, FTIR and Raman and SEM–EDAX tech-
niques. Thermal investigation results were compared with those
reported for other related phases as well as ammonium hepta-
molybdate whereas preliminary measurements of the catalytic
activity for HDS–HYD processes were obtained by RhMo6 and
(Rh, Al)Mo6 supported on ␥-Al2O3. The comparison with the
results for other catalytic systems allowed us to analyze the
potentiality and scope of the Rh–Mo based system.
lent to a monolayer with 6 wt% Mo and 1 wt% metal, following
the procedure given in previous works [14–18]. In addition, the
[RhMo6 + AlMo6]/␥-Al2O3 combined system was prepared by
using a mixed solution of both Rh and Al heteropolyoxomolyb-
dates.
Preliminary measurements of the thiophene hydrodesulfur-
ization and cyclohexene hydrogenation activity were done in a
high pressure reactor (at 280 ◦C, 4 h). The experimental con-
ditions for the activity tests were a feedstock of thiophene
(15,000 ppm), cyclohexane (90%) and cyclohexene (10%) flow
rate 0.353 ml/min, total pressure 26 kg/cm2 and LHSV = 52 l/h.
Operative conditions for the hydrotreating test were selected
according to those used for CoMo6 [15–18].
3. Results and discussion
2. Experimental
3.1. Synthesis and characterization of Anderson phases
involving Rh
RhMo6 phase was obtained by reaction in solution from
AHM and Rh(III) chloride hydrate in the stoichiometric ratio,
according to the method previously described [11–13]. The syn-
thesis of (Rh, Al)Mo6 solid solutions were obtained by adding
Al(NO3)3 solution to the respective AHM and Rh(III) chloride
solutions, according to the desired stoichiometry. Characteriza-
tion of the samples was checked by XRD using a Philips PW
1714 diffractometer (Cu K␣ radiation, Ni filtered), FTIR spec-
troscopy, SEM–EDAX microscopy and chemical analysis by
AAS. Cell parameters were refined by means of the Full Prof
3.0 Program [19].
Vibrational analyses were carried out by means of an FTIR
spectrometer Bruker IFSS 66 (KBr pellet technique) and a
Raman equipment Spex Ramalog 1403 double monochroma-
tor equipped with an SCAMP data processor (excitation line
514.5 nm of an Ar ion laser). Additional measurements were
carried out with a Raman LabRAM Infinity microprobe (Jobin
Yvon) equipped with a liquid nitrogen detector and a frequency-
doubled Nd:YAG laser supplying the excitation line at 532 nm.
The power at the sample was below 5 mW.
The RhMo6 phase was identified by XRD and SEM
microscopy. The chemical analysis data on the basis of Rh and
Mo contents of the Anderson stoichiometry by AAS yielded
Rh 15.91% and Mo 84.08%, in agreement with the theoretical
values (15.16% Rh and 84.83% Mo, respectively).
The replacement of Rh by Al was analyzed for catalytic pur-
poses in order to decrease the Rh content in a structural arrange-
ment. So, the synthesis of [Rh0.25Al0.75], [Rh0.5Al0.5] and
[Rh0.75Al0.25] solid solutions was performed. The [Rh0.25Al0.75
]
XRD pattern was similar to that of RhMo6, with the expected
slight shift of the diffraction lines upward 2θ angles (lower
cell parameters) as it is observed in Fig. 1. However, the lower
resolution of the [Rh0.5Al0.5]Mo6 and [Rh0.75Al0.25]Mo6 XRD
patterns did not permit to clearly identify the formation of a solid
solution with these stoichiometries. Likewise, SEM microscopy
revealed similar nanocrystal morphologies in all cases. The
[Rh0.25Al0.75]Mo6 and RhMo6 micrographs and EDAX profiles
are shown in Fig. 2. EDAX data were in acceptable agreement
with the theoretical Rh/Mo ratios for all stoichiometries but
TGA and DTA measurements were carried out by using a
Shimadzu Thermoanalyzer (TGA 50 and DTA 50), heating rate
of 10 ◦C/min and ␣-Al2O3 as reference. The temperature was
increased up to 1000 ◦C. Additional thermal studies were carried
out in a furnace, using similar conditions to those of TGA–DTA
measurements.
The reduction studies were carried out by using TPR tech-
nique in a home-made reactor fed with 10% H2 reducing agent
in N2 (100 cm3/min) from 20 to 1000 ◦C. The heating rate was
5 ◦C/min. The consumed H2 was detected by a thermal conduc-
tivity cell.
Samples were also investigated by scanning electron micro-
scope Philips SEM 505 with an EDAX 9100 energy disper-
sive system for X-ray microanalysis. The probe size used was
200 nm.
Supported catalysts were prepared by equilibrium impreg-
nation of ␥-Al2O3 (226 m2/g, pore volume of 0.65 cm3/g and a
grain size of 200 ␮m) with aqueous solutions of RhMo6 or (Rh,
Al)Mo6 ammonium salts at RT for 24 h to obtain the equiva-
Fig. 1. XRD diagrams of (a) RhMo6 and (b) [Rh0.25AL0.75]Mo6 ammonium
salts.

24
C.I. Cabello et al. / Thermochimica Acta 447 (2006) 22–29
Fig. 2. SEM micrograph and EDAX signals of: (A) RhMo6 (magnification: ×500, scale bar: 100 ␮m) and (B) [Rh0.25AL0.75]Mo6 (magnification: ×500, scale bar:
100 ␮m).
the Rh distribution for [Rh0.5] and [Rh0.75] was not uniform,
evidencing the co-existence of both AlMo6 and RhMo6 end
members of the solid solution separately. Thus, the formation
of a solid solution was only possible in a limited range of com-
position due to the great difference between the trivalent ionic
inductive effect [11–16]. Hypothetical [RhxAl1−x]Mo6 [x = 0.75
and 0.5] solid solution showed lines located at 947, 901 and near
565 cm−1 like the [Rh0.25Al0.75]Mo6 phase. The most sensitive
band corresponds to Mo–Ob stretching which is useful to reveal
the limit of the solid solution corroborating the previously sug-
gested inhibition to form a continuous solid solution between
RhMo6 and AlMo6. The discontinuity, attributed to the notice-
able difference between the heteroatom ionic radii (∼20%), led
˚
˚
radii (rAl(III) = 0.535 A and rRh(III) = 0.665 A, respectively).
It is well known that HPOMs are predominantly charac-
terized by the presence of Mo–O bridged and Mo O2t ter-
minal bonds, which can be considered as “closed” moieties
[10,12]. Such bonds directed outwards seem to be the most
sensible to heteroatom exchange. So, vibrational spectroscopy
can provide valuable information about this effect [11,12]. The
[Rh0.25Al0.75]Mo6 FTIR spectrum (not shown here) is typical
of Anderson structure, with Mo–O2t terminal bonds located at
940 cm−1 (νs) and 889 and 852 cm−1 (νas), Mo–Ob and Mo–Oc
bridge bonds at 636 and 341 cm−1, the Rh–O stretching at
575 cm−1 and water librations at 429 cm−1 [11,12]. However,
the sharp lines of Raman spectroscopy, unlike the FTIR high
band-widths of the primary Mo-O units, which did not per-
mit to reveal accurately the Al–Rh substitution, are useful to
analyze the problem [11,12]. Fig. 3 shows Raman spectra of
[RhxAl1−x]Mo6 (with x = 1, 0.25 and 0) and Table 1 compares
the main Mo–O Raman modes for the RhMo6, AlMo6 and real
and hypothetical intermediate members of the solid solution.
The highest nuclear charge of Rh(III) is responsible for the rein-
forcement of both Mo–O terminal bonds and Mo–Ob ones by
Fig. 3. Raman spectra of XMo6 ammonium salts (between 1200 and 150 cm−1).

C.I. Cabello et al. / Thermochimica Acta 447 (2006) 22–29
25
Table 1
obtained at this temperature was amorphous to XRD, further
heating allowed to reveal the Rh2MoO6 rutile-type phase which
was clearly detected by XRD at 600 ◦C (plus MoO3 diffraction
lines). The RhMo6 thermal behavior showed differences from
that observed for CoMo6, AlMo6 and CuMo6 [13,20]. Whereas
thefirstonestabilizedtheCo(II)speciessuchasCo(II)MoO4, the
secondoneformedAl2(MoO4)3 andthelastyieldedCuOastem-
perature increased. For the RhMo6 phase, the evolution of MoO3
from ∼750 ◦C, according to TGA curve, led to Rh2MoO6 like
the unique oxide phase, which remained stable up to ∼950 ◦C.
The calculated cell parameters deduced from XRD data by
Raman data of RhMo6, AlMo6 and [RhxAl(1−x)]Mo6 (between 1200 and
550 cm−1
Phase
)
νs Mo–O2t
νas Mo–O2t
ν Mo–Ob
(cm−1
)
(cm−1
)
(cm−1
)
RhMo6
949
947
947
944
898
901
901
902
549
565–569
565
a
[RhxAl(1−x)]Mo6
[Rh0.25Al0.75]Mo6
AlMo6
575
a
Hypothetical phase with x = 0.5 and 0.75.
˚
using a tetragonal cell (a = 4.61 A and c = 3.03 ) are in agreement
with those given in PDF pattern 30–0848. The Mo–Rh rutile-
type oxide decomposed at ∼950 ◦C into MoO3 and Rh2O3,
respectively. SEM micrographs and EDAX profiles shown in
Fig. 5 correspond to samples thermally treated and enable us
to observe the morphological and chemical data changes during
the RhMo6 → Rh2MoO6 + MoO3 → Rh2O3 sequence.
Comparative EDAX data also contributed to analyze the
oxidative thermal effect. The original Rh/Mo ratio (0.18)
increased to 2.20 by heating at 850 ◦C (68.77% Rh, 31.23%
Mo) in agreement with Rh2MoO6 theoretical value of 2.15
(Rh = 68.20%, Mo = 31.80%).
Regarding the formation of the trirutile phase, it should be
pointed out that Rh and Mo octahedral environment in the orig-
inal Anderson-type heteropolyoxoanion should be structurally
displaced and re-arranged to form MoO3 and the mixed oxide,
both with the metals in octahedral coordination. The variability
of EDAX results at 750 ◦C (from 36.43% Rh and 63.57% Mo to
100% Mo) can be attributed to the intermediate step to obtain the
mixed oxide and the MoO3 segregation. The different thermal
stability of both oxides enabled to establish the existence of the
mixed oxide, as unique phase, in the temperature range between
∼800 and ∼950 ◦C.
Fig. 4. (a) TGA and (b) DTA curves of RhMo6.
to the formation of Al-rich members (up to x ∼ 0.25) and thus
only [Rh0.25Al0.75]Mo6 HPOMs will be considered hereafter.
3.2. Thermal treatment of RhMo6 in oxidizing atmosphere
3.3. Thermal treatment in reducing atmosphere
TGA–DTA curves, recorded between RT and 1000 ◦C, are
shown in Fig. 4(a and b) and the results are summarized in
Table 2. Thermal analysis up to 450 ◦C showed the typical
behavior of the Anderson phases with three mass-losses at 126,
255 and 404 ◦C, respectively [12,13]. The first one is reversible
and attributed to the dehydration process. The resolution of the
second one, assigned to the release of NH3 and H2O (from
the countercation and partially from intermolecular H) was not
well defined, whereas the third mass-loss, associated to the col-
lapse of the anionic structure with the removal of the remaining
water molecules, occurred at ∼420 ◦C. Although the product
Fig. 6(a) shows the TPR pattern of RhMo6 phase. This is
characterized by an intense signal at 285 ◦C. Several signals
of lower intensity are observed in the 390–500 ◦C region as
well as at 732 ◦C. The pattern was appreciably different from
those corresponding to the CoMo6 studied phase (with signals
at 610 and 848 ◦C) and from AlMo6 identified in the figure as
pattern (c). The mentioned phases as well as ammonium hep-
tamolybdate present two sharp signals, basically attributed to
Mo(VI)–Mo(IV) and Mo(IV)–Mo◦ reduction steps which are
observed in the ranges of 580–730 and 850–990 ◦C, respec-
Table 2
TGA–DTA data of (NH4)3[RhMo6O24H6]·7H2O (RhMo6)
T (◦C) (exo ⇑, endo ⇓)
TGA data mass-loss (wt%)
Process
Theoretical mass-loss (wt%)
Solid residue (according to XRD)
126
255
404
800
946
9.60
13.65
18.48
83.88
89.58
Dehydration
−H2O, −NH3
−H2O, −NH3
−MoO3
10
Amorphous
Amorphous
14.42
18.87
84.07
89.83
Amorphous, Rh2MoO6 and MoO3 from 600 ◦C
Rh2MoO6
Rh2O3
−MoO3

26
C.I. Cabello et al. / Thermochimica Acta 447 (2006) 22–29
Fig. 5. SEM micrographs and EDAX signals of RhMo6 samples corresponding to several stages of thermal treatment (in air atmosphere). Sample heated at: (A)
(magnification: × 680, scale bar = 100 ␮m) and (B) magnification: × 650, scale bar = 100 ␮m), 700 ◦ C; (C) (magnification: × 2500, scale bar = 10 ␮m) 800 ◦C and
(D) magnification: × 2500, scale bar = 10 ␮m) 1000 ◦C.

C.I. Cabello et al. / Thermochimica Acta 447 (2006) 22–29
27
peaks of the monoclinic MoO2 oxide (PDF 86–0135). XRD
features of Rh and Mo metals were weak at this temperature,
suggesting a high degree of metallic dispersion (PDF of Rh
87–0714, PDF of Mo 42–1120).
EDAX data of TPR samples showed an increase of Mo
content (89%) at 400 ◦C and a subsequent decrease (85%) at
1000 ◦C. This effect can be associated to the reduction progress
of Mo(VI)–Mo(IV) and partially to Mo(IV)–Mo◦ processes,
both included in a stronger low temperature signal. Therefore,
the outer skin of the system was enriched by molybdenum [13].
The complete reduction of remaining MoO2 was observed at
732 ◦C, yielding a Mo/Rh ratio similar to that of the original
sample.
The high reducibility of RhMo6 phase can be related to
the oxidizing character of the heteroatom (ER◦ h(III)–Rh◦ = 0.44 V)
which affects the stability of Mo(VI). The AlMo6 pattern, use-
ful as reference, clearly presents a double area for Mo(IV)–Mo◦
(second stage at ∼990 ◦C) in relation to Mo(VI)–Mo(IV) (first
stage at 586 ◦C). The area of RhMo6 TPR second signal was
very weak with respect to the first one, thus showing the Rh◦
promoting effect. Therefore, the higher heteroatom reducibil-
ity, the higher molybdenum reducibility and the lower reduction
temperature. The characteristics of this type of reduction (high
rate and low temperature) inhibit the adequate atomic arrange-
ment affecting not only the size and dispersion of the metal
particles, but also the nucleation and growth of the oxidic inter-
mediate product. Consequently, the MoO2 crystallinity is very
poor. Typical XRD patterns of Mo and Rh were observed after
further heating (from ∼1000 ◦C).
Fig. 6. TPR patterns of: (a) RhMo6, (b) [Rh0.25Al0.75]Mo6 and (c) AlMo6.
tively (although the low temperature signal can also include the
heteroatom reduction, as in the CoMo6 phase) [13,20]. Pattern
(b) corresponds to [Rh0.25Al0.75]Mo6 solid solution, which par-
tially resembles RhMo6 and AlMo6 as well. In this case, the
strongest signal is shifted to higher temperatures (∼500 ◦C). On
the other hand, the Rh(III)–Rh◦ reduction in Rh2O3 oxide occurs
at 135 ◦C [21], temperature which is lower than that observed for
the RhMo6 phase (286 ◦C). This effect can be related to a higher
metal stability in the condensed structure. TPR data of RhMo6,
AlMo6 and CoMo6 Anderson phases are given in Table 3 where
AHM values are also included for comparative purposes. The
products of the reduction by XRD are also given.
From Fig. 6, it is also interesting to point out the inter-
mediate feature of the [Rh0.25Al0.75]Mo6 TPR pattern which
corroborates the existence of a solid solution. Hence, the temper-
atures and areas of TPR peaks are appreciably changed among
end and intermediate members of the solid solution. In fact,
the Mo(VI)–Mo(IV) signal for AlMo6 phase occurs at about
∼300 ◦C above that of RhMo6. Besides, the weak peak at 732◦C
for RhMo6 differs significantly from the strongest signal at
∼1000 ◦C for the AlMo6 phase. Regarding the solid solution,
the formation of Mo◦ in the first step is associated to the Rh
content in the structure. Likewise, the shift of the TPR signal
from 285 ◦C (for RhMo6) up to 586 ◦C (for AlMo6) is in agree-
ment with the decrease of Rh amount. The correlation between
Samples of RhMo6 after TPR at 400 ◦C were amorphous to
XRD and the presence of MoO2 can be suggested from ∼650 ◦C
bytheveryweakX-raysignalscorrespondingtothemostintense
Table 3
Comparison among the TPR results of different Anderson phases and ammonium heptamolybdate (AHM) (Effects of heteroatom reducibility on the molybdenum
reduction)
Phase
Heteroatom reduction
potential (V)
Temperature of
Mo(VI)–Mo(IV) process (◦C)
Products
Temperature of Mo(IV)–Mo
process (◦C)
Products
RhMo6
CuMo6
CoMo6
AlMo6
AHM
0.44
0.34
285
374–385
610
586
730
Rh, MoO2, Mo
Cu, MoO2
732
783
848
990
850
Rh, Mo
Cu, Mo
Co, Mo
Mo, Al2O3
Mo
1.80a, −0.28b
CoMoO4, MoO2
c
−1.66
MoO2
MoO2
a
Co(III)–Co(II).
b
c
Co(II)–Co.
“(Mo, Al)O2” reduced non-stoichiometric rutile-type phase [13].

28
C.I. Cabello et al. / Thermochimica Acta 447 (2006) 22–29
Fig. 7. RhMo6 adsorption isotherm on ␥-Al2O3.
Fig. 8. Raman spectrum of ␥-Al2O3 supported RhMo6.
the reducibility of the system and the proportion of Rh to pro-
mote the hydrogen spillover allows us to design the appropriate
combination in a suitable structure to ensure the reducibility of
molybdenum at a given temperature.
this Raman line, revealing some structural modifications of the
RhMo6 starting material as previously observed for the other
Anderson phases [16]. Indeed, during the impregnation stage, it
is known that electrostatic ion–support interactions, coordina-
tive ion–surface bonding, Al(III) dissolution with formation of
AlMo6; Rh(III) diffusion reactions on the support surface, etc.,
may intervene [14,22–24]. So, in this case, the support dissolu-
tion, the subsequent AlMo6 phase formation, its re-precipitation
on the support as well as the Rh(III) diffusion to form [Rh,
Al]Mo6 solid solution could be suggested.
Table 4 shows chemical data and catalytic conversion of some
of RhMo6 based catalysts. Data of CoMo6/␥-Al2O3, CoMo
commercial [14–18], Rh commercial catalysts and a TPR treated
RhMo6/␥-Al2O3 at 200 ◦C catalysts are included for compara-
tive purposes. This table clearly shows the role that Rh plays
in the enhancement of the Mo catalytic system. Whereas the
reducibility of Rh increases the hydrogenation ability, the cat-
alytic performances in HDS are similar to those observed in
the CoMo6 supported system. The synergetic effect observed in
the heteropolyanion array with a low Rh/Mo ratio seems to be
promising.
3.4. Preliminary data of catalytic activity
The potentiality of RhMo6 phase in catalysis was analyzed
by means of the results of the above-mentioned thermal behav-
ior. Preliminary data of thiophene hydrodesulfurization (HDS)
and cyclohexene hydrogenation (HYD) activities for the cata-
lysts based on RhMo6 and [Rh, Al]Mo6 supported on ␥-Al2O3
are reported. They were prepared by equilibrium adsorption of
the support given in the experimental item, following the opera-
tive conditions given in previous works [14–18]. AAS chemical
analysis and the isotherm adsorption curve (shown in Fig. 7) for
the RhMo6/␥-Al2O3 system revealed a monolayer of 6% Mo
and 1% Rh (g metal/100 g ␥-Al2O3). The corresponding data
for [Rh0.25, Al0.75]Mo6 solid solution were 5% Mo and 0.23%
Rh [14].
The RhMo6 phase supported on ␥-Al2O3 was characterized
by Raman spectroscopy. Fig. 8 shows the Raman spectrum,
which exhibits the main line characteristic of the MoO2t stretch-
ing mode at 947 cm−1. However, the comparison between sup-
ported and unsupported RhMo6 phase showed a broadening of
However, the decrease of Rh content, observed in the solid
solution, affects HYD as well as HDS activities [2].
Table 4
Chemical data and conversion of thiophene and cyclohexene for different RhMo6 catalytic and reference systems
␥-Al2O3 supported catalysts
X [%]
Mo [%]
HYD [%]
SHYD
HDS [%]
SHDS
CoMo6
0.80
1.80
1.00
0.23
0.50
1.12
5.00
8.00
9.30
6.00
5.00
6.00
6.80
–
30.00
15.55
84.68
10.00
55.00
66.80
65.00
37.50
8.63
70.81
71.00
72.40
21.00
50.14
13.70
22.38
8.85
7.63
12.06
4.20
8.35
2.01
–
CoMoa
RhMo6
[Rh0.25Al0.75]Mo6
[RhMo6 + AlMo6]b
RhMo6
Rha
84.68
43.50
100.00
59.64
13.00
c
Note: X [%] and Mo [%] are total metal concentration in g/100 g support. SHYD and SHDS = HYD/X [%] (Co or Rh) and HDS/Mo [%].
a
Commercial catalysts.
Catalyst prepared by RhMo6/AlMo6 combined solutions.
Catalyst after TPR condition treatment at 200 ◦C.
b
c

C.I. Cabello et al. / Thermochimica Acta 447 (2006) 22–29
29
Both catalysts [solid solution [Rh0.25, Al0.75]Mo6 and
[RhMo6 + AlMo6]/␥-Al2O3 combined system (0.5RhMo6/
0.5AlMo6)] reveal a catalytic effect depending on Rh content.
The reducing thermal pre-treatment of RhMo6/␥-Al2O3
induced a substantial decrease of HDS activity whereas HYD
activity was not so affected. The presence of Rh induced the
Mo(VI)–Mo(IV) reduction at lower temperature.
These Mo(IV) species were then poorly sulfided upon acti-
vation, which affected the catalytic performance as it is widely
described in literature [25,26].
gratefultoMrs. PascaleBaranek, Mrs. GracielaValle, Lic. Diego
˜
Pena and to Lic. Norberto Firpo for their contribution in experi-
mental measurements. C.I. Cabello is a member of the research
staff of CICPBA, Argentina.
References
[1] G. Ertl, H. Knozinger, J. Weitkamp (Eds.), Handbook of Heterogeneous
Catalysis, vol. 1, VHC, Weinheim, 1997, p. 278.
[2] Z. V´ıt, D. Gulkova´, L. Kaluza, M. Zdrazil, J. Catal. 232 (2005) 447.
[3] C. Bianchi, Catal. Lett. 76 (2001) 155.
[4] M. Shelef, G.W. Graham, Catal. Rev. Sci. Eng. 36 (6) (1994) 433.
[5] C. Ca´ceres, M.N. Blanco, H.J. Thomas, in: G. Poncelet, P. Grange, P.A.
Jacobs (Eds.), Preparation of Catalysts III, Elsevier, Amsterdam, 1983,
p. 333.
[6] M.T. Pope, A. Mu¨ller (Eds.), Polyoxometalate chemistry from topology
via self-assembly to applications, Kluwer Academic Publishers, London,
2001.
4. Conclusions
The main results of this work dealing with RhMo6 system
can be summarized as follows:
(i) Upon thermal treatment in oxidative conditions of bulk
RhMo6, the formation of the trirutile RhMo2O6 phase,
stable in the range between 700 and 950 ◦C is observed,
whereas in reducing conditions Rh acts as promoter of
the Mo reduction. The reduction behavior is different
from the one observed for the other Anderson phases.
Indeed, the presence of Rh metal induces the complete
Mo(VI)–Mo(IV)–Mo◦ reductionatlowtemperatures(from
∼250 ◦C). However, the Rh(III) reducibility to form Rh◦
occurs at higher temperature than that observed in Rh–O
binary systems due to the stability of the heteropolyanion
structure.
(ii) The formation of a bulk solid solution between RhMo6
and AlMo6 phases is not complete and it is observed
in a small range of Rh content (up to 0.25 M ratio or
[Rh0.25,Al0.75]Mo6). This fact can be attributed to the
differences between the trivalent ionic radii of the het-
eroatoms.
[7] M. Misono, Catal. Rev. Sci. Eng. 29 (1987) 269.
[8] N. Mizuno, M. Misono, Chem. Rev. 98 (1) (1998) 199.
[9] D. He, D. Pang, L. Wei, Y. Chen, T. Wang, Z. Tang, J. Liu, Y. Liu, Q.
Zhu, Catal. Commun. 3 (2002) 429.
[10] M.T. Pope, Heteropoly and Isopolyoxometallates, Springer-Verlag,
Berlin, 1983.
[11] I.L. Botto, A.C. Garcia, H.J. Thomas, J. Phys. Chem. Solids 53 (8)
(1992) 1075.
[12] C. Martin, C. Lamonier, M. Fournier, O. Mentre´, V. Harle´, D. Guillaume,
E. Payen, Inorg. Chem. 43 (2004) 4636.
[13] C.I. Cabello, I.L. Botto, H.J. Thomas, Thermochim. Acta 232 (1994)
183.
[14] C.I. Cabello, I.L. Botto, F. Cabrerizo, M.G. Gonza´lez, H.J. Thomas,
Ads. Sci. Tech. 18 (7) (2000) 591.
[15] C.I. Cabello, I.L. Botto, H.J. Thomas, Appl. Catal. A 197 (2000) 79.
[16] I. Pettiti, I.L. Botto, C.I. Cabello, S. Colonna, M. Faticanti, G. Minelli,
P. Porta, H.J. Thomas, Appl. Catal., A: Gen. 220 (1–2) (2001) 113.
[17] C.I. Cabello, F.M. Cabrerizo, A. Alvarez, H.J. Thomas, J. Mol. Catal.
A Chem. 186 (2002) 89.
[18] C.I. Cabello, M. Mun˜oz, E. Payen, H.J. Thomas, Catal. Lett. 92 (2004)
69.
[19] J. Rodriguez-Carvajal, Phys. B 192 (1993) 55.
[20] I.L. Botto, C.I. Cabello, G. Minelli, M. Occhiuzzi, Mater. Chem. Phys.
39 (1994) 21.
[21] A. Jones, B. Mc Nicol, Temperature Programmed Reduction for Solid
Materials Characterization, Marcel Dekker, New York, 1986, p. 109.
[22] I.L. Le Bihan, P. Blanchard, M. Fournier, J. Grimblot, E. Payen, J.
Chem. Soc., Faraday Trans. 94 (1998) 937.
[23] X. Carrier, J.F. Lambert, M. Che, J. Am. Chem. Soc. 119 (1997) 10137.
[24] X. Carrier, J.B. D’Espinose de la Caillerie, J.F. Lambert, M. Che, J.
Am. Chem. Soc. 121 (1999) 3377.
(iii) HDS activity of RhMo6/␥-Al2O3 catalyst is similar to that
observed for CoMo6/␥-Al2O3 and Co–Mo/␥-Al2O3 sys-
tems. Likewise, its HYD activity is considerably higher
than that for the commercial Rh/␥-Al2O3 catalyst.
Catalytic performance of this bifunctional system is a promising
target to be explored more accurately.
[25] D. Pirotte, J.M. Zabala, P. Grange, B. Delmon, Bull. Soc. Chim. Belg.
90 (1981) 1239.
Acknowledgements
[26] H. Topsøe, B.S. Clausen, F.E. Massoth, in: J.R. Anderson, M. Boudart
(Eds.), Catalysis Science and Technology, vol. 11, Springer-Verlag,
Berlin, 1996.
This work was supported by CONICET; CICPBA Argentina
and COCAFAL Program (CNRS France and TOTAL). We are