Amorphous oxide as a novel efficient catalyst for direct selective oxidation of methanol to dimethoxymethane

Article abstract of DOI:10.1039/b714260a

We report for the first time the use of an amorphous oxide catalyst for the selective oxidation of methanol in the gas phase, leading at 553 K to the production of dimethoxymethane with a selectivity as high as 90% at high methanol conversion (68%). The Royal Society of Chemistry.

Full text of DOI:10.1039/b714260a

View Article Online / Journal Homepage / Table of Contents for this issue
COMMUNICATION
www.rsc.org/chemcomm | ChemComm
Amorphous oxide as a novel efficient catalyst for direct selective
oxidation of methanol to dimethoxymethanew
a
a
b
a
´
´
Sebastien Royer,z Xavier Secordel, Markus Brandhorst, Franck Dumeignil,*
a
a
a
a
¨
Sylvain Cristol, Christophe Dujardin, Mickael Capron, Edmond Payen and
Jean-Luc Dubois*b
Received (in Cambridge, UK) 17th September 2007, Accepted 3rd December 2007
First published as an Advance Article on the web 20th December 2007
DOI: 10.1039/b714260a
We report for the first time the use of an amorphous oxide
catalyst for the selective oxidation of methanol in the gas phase,
leading at 553 K to the production of dimethoxymethane with a
selectivity as high as 90% at high methanol conversion (68%).
obtained at low conversion (o20%). Liu et al.5 obtained high
DMM production rates over bulk and supported H3+nVn-
Mo12ꢀnPO40 Keggin structures. Partial incorporation of va-
nadium in the Keggin structure (n = 1, 2) yielded the highest
DMM selectivities: SDMM = 58.1% at a methanol conversion
of 68.2% over 9.2 wt% H4PVMo11O40 supported on SiO2.
The same authors further used a different catalytic system,
namely RuO2 supported on TiO2, Al2O3 and TiO2–Al2O3.
Lower performances were obtained with a maximum DMM
selectivity of 66.8% at a methanol conversion of only 20% (4.4
wt% Ru/Al2O3).6 Promising results were obtained by the
Iwasawa’s group7 who observed high DMM yields over Re
oxide and Re-based mixed oxides supported on various oxide
supports (TiO2, V2O5, ZrO2, a-Fe2O3, g-Fe2O3, a-Al2O3...).
The maximum DMM selectivity varied between 60% (Re/
SiO2) and 90–93% (Re/a-Al2O3, V2O5, ZrO2) for a methanol
conversion comprised of between 15% (Re/SiO2, a-Al2O3, a-
Fe2O3) and 60% (Re/TiO2). These catalysts were patented8
but the low thermal stability of Re at intermediate reaction
temperatures (4573 K) and its prohibitive cost unfortunately
limit their industrial perspectives. Recently, Fu et al.9 reported
excellent results over Ti(SO4)2-modified V2O5/TiO2 catalysts
with DMM selectivities of 89–92% for 48–60% methanol
conversions. Irrespective of the catalytic system, all the studies
suggest a dual mechanism involving redox and acidic sites
(Brønsted in the case of Keggin structures4,5 or Lewis for Re-
based catalysts7). Achieving an adequate balance between the
two kinds of active sites is thus a crucial parameter for
optimizing the DMM production.
During the 20 past years numerous efforts have been devoted
to the development of selective catalysts for the partial oxida-
tion of methanol, with important industrial applications as a
target. It is now well admitted that this reaction is strongly
sensitive to the nature of the active sites.1 Different products
can be obtained as shown in Scheme 1. Redox sites enable the
production of partially oxidized species [formaldehyde (F),
formic acid (FA)] or totally oxidized species (CO and CO2).
Acidic sites enable condensation reactions, which can give
dimethyl ether (DME), dimethoxymethane [or ‘methylal’
(DMM)] and methyl formate (MF). As proposed by Tatibouet
et al.,1 this dual behavior enables probing of structural
(cristallographic plane reactivity) and chemical (acid–base
and redox) properties of oxide catalysts.
Among the aforementioned products, DMM is especially
interesting for industrial applications, since it is suitable as a
fuel additive with a high chemical stability. Industrially,
DMM is conventionally produced through catalytic distilla-
tion by acetalization of methanol and formaldehyde.2 Hagen
and Spangler extensively described the use of DMM for
polyoxymethylenedialkyl ether synthesis.3 Gas phase DMM
production patents are rare because of the difficulty of obtain-
ing high DMM yields in a one-step gas-phase methanol-
selective oxidation procedure. In the academic literature,
many authors reported attempts to use molybdenum-based
catalysts for this reaction. In studies reporting the use of bulk
and supported 12-molybdophosphoric acid catalysts,4 it was
shown that b-MoO3 was the active phase for direct formation
of DMM. High DMM selectivity (SDMM; up to 55%) was
We report in this work the first successful attempt of using a
purely amorphous mixed oxide for the direct selective synth-
esis of DMM from methanol. ARKEMA recently patented a
process for the production of DMM using this catalyst, called
hereafter AR01.10 Such a catalyst is currently produced at the
industrial scale. We compared the performances of AR01 to
those of two Re-based catalysts using the same evaluation
a
´
Unite de Catalyse et de Chimie du Solide, UMR CNRS 8181,
Universite des Sciences et Technologies de Lille, 59650 Villeneuve
´
d’Ascq Cedex, France. E-mail: Franck.Dumeignil@univ-lille1.fr;
Fax: +33 (0)3.20.43.65.61; Tel: +33 (0)3.20.43.45.38
b Arkema, 420 Rue d’Estienne d’Orves, 92705 Colombes France.
E-mail: jean-luc.dubois@arkema.com; Fax: +33 (0)4.72.39.88.61;
Tel: +33 (0)4.72.39.85.11
w Electronic supplementary information (ESI) available: experimental.
See DOI: 10.1039/b714260a
z Present address: Universite de Poitiers, Ecole Superieure d’Inge-
nieurs de Poitiers, LACCO UMR 6503 CNRS, 40, avenue du recteur
Pineau, 86022 Poitiers, France
Scheme 1 Reaction pathways for the reaction of catalytic partial
oxidation of methanol (adapted from ref. 1).
ꢁc
This journal is The Royal Society of Chemistry 2008
Chem. Commun., 2008, 865–867 | 865

View Article Online
procedure. AR01 is a bulk catalyst of general formula
Mo12V3W1.2Cu1.2Sb0.5Ox obtained by a simple coprecipitation
method. The synthesis procedure of AR01, as well as those of
point), which was identified as Re2O7. This is a well known
issue when using Re based solids.13 Further, it should be noted
that the given Re amounts in the catalysts (i.e., 7.4 wt% and 20
wt%) corresponds to the quantities of Re used during their
preparation. However, the calcination procedure, followed by
the necessary preactivation procedure and the subsequent
catalytic test progressively depleted the Re content of the
catalysts, especially when operating at high temperatures,
which considerably altered the expected performances when
increasing temperature. On the other hand, over AR01 it was
possible to increase the methanol conversion to 63% while
maintaining the DMM selectivity at a high value of 89.2% at
553 K (Table 1). All the other possible reaction products (F,
DME, MF, and COx; Scheme 1) were also formed in low
amount. With the increase in temperature, we observed an
increase in the selectivity of oxidation products (F and COx).
At the same time, a decrease in the selectivity in DME and MF
(condensation products) was observed. This trend can be
explained assuming a slight increase in the quantity of active
redox site, and/or the deactivation of a part of the acidic sites
when increasing the working temperature. Nevertheless, redox
sites present on the surface of AR01 are not too strong because
they clearly favor partial oxidation rather than complete
11
the reference catalysts (a sol–gel g-Al2O3 and a TiO2 sup-
ported Re catalyst12), are described in the ESI file.w The three
catalysts have been tested for the selective oxidation of metha-
nol. A full description of the catalytic test procedure is given in
the ESI file.w
Due to the simple straightforward coprecipitation method
used for its synthesis, AR01 unsurprisingly exhibits a low
specific surface area (o10 m2 gꢀ1). Moreover, because of the
intermediate/low calcination temperature (613 K), AR01 is
amorphous, which was confirmed by XRD analysis (Fig. S1w).
Optimal preactivation conditions were determined for the
lab-scale test. Especially, we found that the oxygen content in
the preactivation feed played an important role on the cata-
lytic performances of AR01 (Table S1 and Fig. S2w). The best
result was obtained for preactivation under pure O2 with a
DMM selectivity of 90.1% at a methanol conversion of 68.3%
(Table S1w), which is the best result ever reported in the
literature. Lowering the O2 percentage in the preactivation
feed led to a slight decrease in both DMM selectivity and
methanol conversion (Table S1 and Fig. S2w), which however
kept high levels. A test performed on a non-activated catalyst
oxidation (SCO remained low and the CO always represented
x
led to an unacceptable decrease in DMM selectivity (SDMM
=
more than 95% of the COx). Further, the amorphous char-
acter of AR01 is an important parameter. Methanol conver-
sion to DMM needs cooperative work of two types of sites
(acid ones and redox ones). Inadequate catalyst treatment
leads to phase segregation and improper sites distribution on
the surface with a strong detrimental effect on DMM selectiv-
ity. With properly handled amorphous AR01, DMM selectiv-
ity was not altered up to 553 K (Table 1).
2.0%). For further experiments, a 20 vol% O2 in He preacti-
vation feed was retained, because it simulates air flowing,
which is preferred for practical applications, while maintaining
very high catalytic performances, close to those obtained over
the catalyst preactivated in pure oxygen.
Table 1 compares the catalytic performances of AR01 with
those of the two Re-based catalysts. The g-Al2O3-supported
Re catalyst yielded mainly DME due to the acidobasic func-
tion of the support. The performances of the Re catalyst
supported TiO2 catalyst were better than that of AR01 at
low reaction temperatures (Re/TiO2: 473 K; Conversion =
10.3%, SDMM = 86.3%; AR01: 473 K; Conversion = 8.3%,
These catalytic results obtained over AR01 are clearly
superior to those claimed by Liu et al. who obtained their
highest selectivities in DMM in the range of 50–60% for
conversions in the range of 60–68% over H3+nPVnMo12ꢀnO40
supported SiO2 samples.5 In contrast to our results on AR01,
these authors observed a drastic decrease in DMM selectivity
when increasing the reaction temperature. While over the
S
DMM = 85.9%). The slightly higher activity at low tempera-
ture of the Re samples can be attributed, as proposed by
Iwasawa’s group,7 to a high reactivity (i.e., high reducibility)
of Re-oxide species, which enables lower temperature metha-
nol oxidation when compared to AR01. Nevertheless, increase
in temperature irreversibly damaged the Re/TiO2 catalyst by
volatilization of Re oxo-species. Over 513 K, we observed the
condensation of a green solid at the outlet of the reactor (cold
H5PV2Mo10O40/SiO2 catalyst
a methanol conversion of
39.9% with a DMM selectivity of 61.8% was obtained at
453 K, at 513 K the methanol conversion slightly increased to
42.4% but meanwhile the DMM selectivity dropped at 32.7%.
Note that the same authors found a maximum DMM selec-
tivity over a supported Ru catalyst (4.4 wt%Ru/Al2O3) of
Table 1 Comparison between the catalytic performances of AR01 and two rhenium reference samplesa
Selectivity (mol%)
Sample
AR01
SBET/m2 gꢀ1
o10
Temperature/K
Methanol conversion (%)
DMM
F
DME
MF
CO + CO2
473
523
553
473
533
473
513
8.3
31.3
63
10.3
52.1
11.6
29.5
85.9
89.7
89.2
86.3
78
Trace
1.1
4.0
0
4.3
0
6.1
5.3
3.2
12
7.7
8.0
3.1
1.7
1.7
5.4
2.8
4.1
0
0.7
1.8
0
4.6
0
b
7.4 wt%Re/TiO2
101
398
b
20 wt%Re/g-Al2O3
33
27.8
64.2
65.6
0
2.5
a
b
7.5 vol% CH3OH; 8.5 vol% O2 in He; Total flow rate of 54.4 mL minꢀ1 (GHSV = 22 000 mL hꢀ1 gꢀ1). 7.5 vol% CH3OH; 15.0 vol% O2 in He;
Total flow rate of 64.3 mL minꢀ1 (GHSV = 26 000 mL hꢀ1 ꢀ1).
g
ꢁc
This journal is The Royal Society of Chemistry 2007
866 | Chem. Commun., 2008, 865–867

View Article Online
O2 in He at a GHSV of 22 000 mL hꢀ1
g
ꢀ1), we observed
only a slight decrease in methanol conversion of less than 5
points, which occurred in the first 4 h of the run. Then,
conversion remained stable. This lag time thus corresponded
to the time required for catalyst stabilization. In contrast, no
change in DMM selectivity was observed (SDMM B 90 ꢃ 2%).
In summary, the amorphous Mo12V3W1.2Cu1.2Sb0.5Ox
(AR01) sample synthesized by a simple straightforward co-
precipitation procedure exhibits both high activity and high
selectivity in DMM. Results clearly showed superior perfor-
mances when considering practical industrial applications
among all the references already published in the field of the
one-step gas-phase methanol oxidation in DMM. Moreover, it
was clearly established that the catalyst can work in a wide
range of reaction temperatures and methanol concentrations
without any drastic loss in DMM selectivity, which is inter-
esting considering possible parameters fluctuation in real
operating conditions.
Fig. 1 Effect of the methanol partial pressure in the feed on the
catalytic performances of AR01 at 553 K.
66.8% at a methanol conversion of B20%.6 Superior DMM
selectivity (93.7% at 21.5% methanol conversion over 10 wt%
Re on V2O5) was only found by the Iwasawa’s group.7
Unfortunatly, no information on the influence of temperature
on SDMM was given.
The authors thank the Arkema company for the financial
support for this work.
The performances of AR01 preactivated in air flow are
substantially similar to those recently claimed by Fu et al.9
over a V2O5/TiO2–Ti(SO4)2 catalysts with DMM selectivities
of 89–92% for 48–60% methanol conversions at 433 K.
However, preactivation of AR01 in pure oxygen led to even
better performances with DMM selectivities of 90% for a
methanol conversion of 68% at 553 K. Note that this reaction
is exothermic. Design of an energy efficient industrial process
involves heat recovery through high pressure steam utilization
as an energy carrier, which can be optimized if the process
temperature is sufficiently high. This is enabled by the use of
AR01, which can work at higher temperatures while main-
taining very high SDMM (i.e., 90%).
Notes and references
1 (a) J.-M. Tatibouet and J. E. Germain, J. Catal., 1981, 72, 375; (b)
J.-M. Tatibouet, Appl. Catal., A, 1997, 148, 213; (c) J.-M. Tatibouet
and H. Lauron-Pernot, J. Mol. Catal. A: Chem., 2001, 171, 205.
2 G. Lambiotte, CH Pat., 688041, 1997.
3 (a) G. P. Hagen and M. J. Spangler, US Pat., US, 5959456, 1999;
(b) G. P. Hagen and M. J. Spangler, US Pat. 6160174, 2000; (c) G.
P. Hagen and M. J. Spangler, US Pat., 6160186, 2000; (d) G. P.
Hagen and M. J. Spangler, US Pat., 6166266, 2000; (e) G. P. Hagen
and M. J. Spangler, US Pat., 6265528, 2001; (f) G. P. Hagen and
M. J. Spangler, US Pat., 6350919, 2002; (g) G. P. Hagen and M. J.
Spangler, US Pat., 6392102, 2002; (h) G. P. Hagen and M. J.
Spangler, US Pat., 6437195, 2002.
4 (a) M. Fournier, A. Aouissi and C. Rocchiccioli-Deltcheff, J.
Chem. Soc., Chem. Commun., 1994, 307; (b) C. Rocchiccioli-
Deltcheff, A. Aouissi, M. M. Bettahar, S. Launay and M. Four-
nier, J. Catal., 1996, 164, 26; (c) C. Rocchiccioli-Deltcheff, A.
Aouissi, S. Launay and M. Fournier, J. Mol. Catal. A: Chem.,
1996, 114, 331.
We then studied the effect of methanol partial pressure on
the performances of AR01. With the increase in methanol
partial pressure, methanol conversion progressively decreased
from 78.2% for 5 vol% CH3OH to 25.1% for 38 vol%
CH3OH (Table S2w). The DMM selectivity only moderately
decreased with the increase in methanol partial pressure but
always remained larger than 70% (Fig. 1). As a result, DMM
productivity increased with a maximum of 126 ꢂ 10ꢀ5 mol
minꢀ1 gꢀ1 at a methanol concentration of 28 vol% (Table
S1w), which is much larger than the largest one obtained by Fu
et al.,9 i.e., 21.8 ꢂ 10ꢀ5 mol minꢀ1 gꢀ1 (calculated by applying
to their numerical results the formula given as a footnote in
Table S2w). Note that the observed decrease in DMM selec-
tivity led to an increase in the selectivities in F and DME,
which are valuable compounds. The sum of the DMM, F, and
DME selectivities remained around 95%, irrespective of the
methanol concentration (Fig. 1), which is thus satisfactory.
The last point concerns the stability of the AR01 sample,
which is critical when one considers industrial perspectives. As
aforementioned, Re-based catalysts performances are hin-
dered by metal loss occurring over B513 K, which makes
them unusable for practical applications. In contrast, AR01
was found to be stable with time on stream for said high
processing temperatures. After 28 h of reaction at 553 K
(preactivation in simulated air; 5 vol% CH3OH–8.5 vol%
5 (a) H. Liu and E. Iglesia, J. Phys. Chem. B, 2003, 107, 10840; (b) H.
Liu and E. Iglesia, J. Catal., 2004, 223, 161.
6 H. Liu and E. Iglesia, J. Phys. Chem. B, 2005, 109, 2155.
7 (a) Y. Yuan, H. Liu, H. Imoto, T. Shido and Y. Iwasawa, Chem.
Lett., 2000, 674; (b) Y. Yuan, H. Liu, H. Imoto, T. Shido and Y.
Iwasawa, J. Catal., 2000, 195, 51; (c) Y. Yuan, T. Shido and Y.
Iwasawa, Chem. Commun., 2000, 1421; (d) Y. Yuan and Y.
Iwasawa, J. Phys. Chem. B, 2002, 106, 4441; (e) Y. Yuan, K. Tsai,
H. Liu and Y. Iwasawa, Top. Catal., 2003, 22, 9.
8 (a) Y. Iwasawa, Y. Yuan, T. Shido and N. Nobuhiro, US Pat.,
1132366, 2001; (b) Y. Iwazawa, Y. En and S. Takafumi, Jpn. Pat.,
2000-177972, 2000.
9 Y. Fu and J. Shen, Chem. Commun., 2007, 2172.
10 J.-L. Dubois, M. Brandhorst, M. Capron and C. Dujardin, World
Pat., 2007/034264, 2007.
11 F. Dumeignil, K. Sato, M. Imamura, N. Matsubayashi, E. Payen
and H. Shimada, Appl. Catal., A, 2003, 241, 319.
12 (a) H. Balcar, R. Hamtil, N. Zilkova and J. Cejka, Catal. Lett.,
2004, 97(1–2), 25; (b) J. Okal, Appl. Catal., A, 2005, 287, 214.
13 (a) M. Valigi, D. Cordischi, D. Gazzoli, Cornelius P. Keijzers and
A. A. K. Klaasser, J. Chem. Soc., Faraday Trans., 1981, 77, 1871;
(b) F. D. Hardcastle, I. E. Wachs, J. A. Horsley and G. H. Via, J.
Mol. Catal., 1988, 46, 15; (c) M. A. Vuurman, D. J. Stufkens, A.
Oskam and I. E. Wachs, J. Mol. Catal., 1992, 76, 263; (d) B. Mitra,
X. Gao, I. E. Wachs, A. M. Hirt and G. Deo, Phys. Chem. Chem.
Phys., 2001, 3, 1144; (e) S. V. Kotov, E. Sh. Finkel’shtein,
S. P. Chernykh, T. N. Shabalina, V. A. Tyshchenko and V. I.
Milovantseva, Kinet. Catal., 2006, 47, 460.
ꢁc
This journal is The Royal Society of Chemistry 2008
Chem. Commun., 2008, 865–867 | 867