Bio-ethanol catalytic steam reforming over supported

metal catalysts
Fabien Aupr^eetre, Claude Descorme
*
, Daniel Duprez
Laboratoire de Catalyse en Chimie Organique (LACCO) UMR 6503 CNRS, Universit ee de Poitiers, Facult ee des Sciences,
40 avenue du Recteur Pineau, F-86022 Poitiers Cedex, France
Received 4 April 2002; accepted 29 April 2002
Abstract
Considering both the inuence of the nature of the metal (Rh, Pt, Ni, Cu, Zn, Fe) and the role of the support
(c-Al
2
O
3
, 12%CeO
2
Al
2
O
3
; CeO
2
; Ce
0:63
Zr
0:37
O
2
), CO
2
is presented as a primary product in the bio-ethanol steam-re-
forming catalytic reaction (SRR) over some supported metal catalysts. Based on this unexpected observation, a new
strategy for maximizing the hydrogen production and minimizing the CO formation is proposed. Any highly selective
catalytic formulation should be free of any promoter in the water gas shift reaction (WGSR) which tends to equilibrate
the SRR gas towards higher CO concentrations. 2002 Elsevier Science B.V. All rights reserved.
1. Introduction
Since a few decades, increasing attention is be-
ing paid to pollution-related environmental and
public health problems. Particularly, as one of the
major contributor to the atmospheric pollution,
the automotive sector had to work hard on pol-
lution control. Both catalytic converters on gaso-
line-fueled engines and lters for particulates on
diesel-fueled engines were implemented.
The next step, to overpass the forthcoming
regulations on CO
2
emissions according to the
Kyotos Protocol (in Europe, 8% reduction com-
pared to the 1990s emissions by 20082012), re-
quires the development of both new engines and
alternative fuels. . .
Fuel cells recently appeared as one solution in
the development of zero-emission electrical vehi-
cle. Technical options range from batteries to
liquid H
2
tanks, high pressure H
2
storage, me-
dium pressure H
2
storage on hydrides and nano-
structured carbons and on-board H
2
production
[14].
This study is in keeping with this last approach.
Bio-ethanol, presented as an environmentally
friendly fuel (renewable, CO
2
neutral, non
toxic. . .), was used as a fuel and the catalytic steam
reforming reaction (SRR) was studied for H
2
on-
board production (1).
C
2
H
5
OH3H
2
O !6H
2
2CO
2
DH
0
r
173:1 kJ mol
1
1
Catalysis Communications 3 (2002) 263267
www.elsevier.com/locate/catcom
*
Corresponding author. Tel.: +33-5-49-45-39-97; fax: +33-5-
49-45-34-99.
E-mail address: claude.descorme@univ-poitiers.fr (C. Desc-
orme).
1566-7367/02/$ - see front matter 2002 Elsevier Science B.V. All rights reserved.
PII: S1566- 7367( 02) 00118- 8
The goal was a high yield in H
2
and a high selec-
tivity towards CO
2
. In fact, CO, possibly produced
in the SRR, is commonly described as a poison of
the fuel cell electrodes made of noble metals.
A wide range of catalysts were tested using
dierent metals (Rh, Pt, Pd, Ru, Ni, Cu, Zn, Fe)
and various supports (Al
2
O
3
, 12%CeO
2
Al
2
O
3
;
CeO
2
; CeO
2
ZrO
2
; ZrO
2
). The reason for such a
screening was to obtain a catalyst both (i) highly
active in the SRR to maximize the ethanol con-
version to hydrogen, that is a high yield in H
2
and
(ii) highly selective towards the formation of CO
2
(low CO outlet concentration) both in the SRR(1)
and in the water gas shift reaction (WGSR):
COH
2
O !CO
2
H
2
2
2. Experimental
A 12%CeO
2
Al
2
O
3
was prepared by impreg-
nation of a c-Al
2
O
3
support commercialized by
Procatalyse (130 m
2
g
1
) with an aqueous solution
of CeNO
3

6
NH
4

2
, dried overnight at 120 C and
calcined at 500 C under owing dry air
(30 cm
3
min
1
) for 5 h.
The catalysts were prepared by impregnation
of the dierent supports using aqueous solutions
of the corresponding metal precursor salts
(RhNO
3

3
; PtNH
3

2
NO
2

2
; PdNO
3

2
; NiNO
3

2
;
CuNO
3

2
; ZnNO
3

2
; FeCl
3
). The as-prepared
catalysts were pre-calcined at 700 C under owing
dry air (30 cm
3
min
1
) for 5 h.
Reactivity testing was carried out at atmo-
spheric pressure in a quartz xed-bed reactor
(down ow). 40 mg catalyst (grain size: 250315
lm) diluted in 360 mg cordierite (grain size: 100
200 lm) were used for each test. Before testing, the
catalyst was reduced under owing hydrogen
(50 cm
3
min
1
) at 300 C for 1 h. A stoichiometric
reaction mixture was used (12.8 vol%C
2
H
5
OH +
38.4 vol% H
2
O + 48.8 vol%N
2
) water:etha-
nol 3:1 mol/mol). Total ow rate was
100 cm
3
min
1
. Ethanol and water were intro-
duced as liquids using automated syringes, va-
porized at 120 C and further eluted by N
2
(48 cm
3
min
1
). Each test was performed at con-
stant temperature, between 500 and 800 C. Before
the gas phase chromatographs, condensable va-
pors (ethanol, water, acetaldehyde, acetone. . .)
were trapped using a condenser which temperature
was set at 0.5 C. The liquid volume was used to
approximate the ethanol conversion. Gas analysis
was performed on-line using a combination of
three gas chromatographs: two of them equipped
with a thermal conductivity detector (TCD) and
the other one equipped with a ame ionization
detector (FID). H
2
was analyzed using a TCD,
equipped with a 5A molecular sieve column, with
N
2
as both the carrier gas and the reference gas.
N
2
, CO, CO
2
and CH
4
were analyzed using a
TCD, equipped with a CTR column (Alltech),
with H
2
as both the carrier gas and the reference
gas. N
2
was not only a dilutant but also a tracer
and the corresponding GC signal was used to es-
timate the outlet ow rate to determine the hy-
drogen yield. Finally, hydrocarbons (CH
4
; C
2
H
6
and C
2
H
4
essentially) were separated on a Porapak
Q column and analyzed with the FID detector.
3. Results and discussion
A wide range of catalysts was tested, changing
both the active phase (Rh, Pt, Pd, Ru, Ni, Cu, Zn,
Fe) and the oxide support (Al
2
O
3
, 12%CeO
2

Al
2
O
3
; CeO
2
; CeO
2
ZrO
2
; ZrO
2
). As the SRR was
thought to produce CO as a primary product, the
overall idea was to couple both a highly active SR
catalyst and a highly active catalyst in the WGSR
(2) for the subsequent transformation of CO
produced upon bio-ethanol SR to CO
2
. On
purpose, metals were selected either for their
activity in the SRR (Rh, Ni) [5] or/and for their
activity in the WGSR (Pt, Cu, Zn, Fe) [6]. Addi-
tionally, supports were tested either as promoters
of the WGSR (ceria-based supports) [7] or/and the
SRR (2) [8]. In fact, in the alkylated aromatics
selective reforming reaction, Duprez et al. [9]
proposed a bi-functional mechanism where the
hydrocarbon to be reformed would be activated on
the metal particle and the water would be activated
on the support as hydroxyl groups. In such a
scheme, oxide supports with high OH groups
surface mobility should be considered as promot-
ers in the SRR. In that sense, a Ce
0:63
Zr
0:37
O
2
264 F. Aupr^eetre et al. / Catalysis Communications 3 (2002) 263267
mixed oxide with enhanced oxygen surface mo-
bility [10] was tested as support.
In a rst part, the inuence of the metal particle
chemical nature was checked. Catalytic perfor-
mances in the bio-ethanol SRR of c-Al
2
O
3
(Pro-
catalyse, S
BET
130 m
2
g
1
) supported metal
catalysts (Rh, Pt, Pd, Ru, Ni, Cu, Zn, Fe) were
studied. The results are presented in Table 1.
At 700 C under atmospheric pressure, c-Al
2
O
3
supported Rh and Ni catalysts clearly appeared as
the most active and selective catalysts in the ethanol
SRR. Compared to Rh/c-Al
2
O
3
, Ni/c-Al
2
O
3
gave a
higher yield in hydrogen but a lower selectivity to-
wards CO
2
. In fact, earlier studies had demon-
strated that Rh and Ni are the best catalysts in the
hydrocarbons SRR. Additionally, Ni and especially
Rh were poor candidates in the WGSR [11,12].
Lower activities and lower selectivities towards
CO
2
are observed using c-Al
2
O
3
supported Pt, Cu,
Zn or Fe catalysts. Such metals are known as e-
cient catalysts in the WGSR while they have only
limited activity in the SRR. As a result, such cata-
lytic formulations lead to an equilibrated CO=CO
2
mixture, according to the thermodynamic.
Additionally, Ru catalyst was almost inactive in
the SRR but essentially active in the ethanol de-
hydration reaction leading to the formation of
ethylene (38%). Such a catalyst deactivate rapidly
due to coke formation by ethylene polymerization.
Dehydration is also observed on Cu, Pt, Zn and
Fe. Furthermore, one can note that Pd, Cu and Zn
catalysts were very poorly selective towards CO
2
and active in the hydrogenolysis reaction towards
methane formation.
Consequently, (i) metals highly active in the
SRR and poorly ecient in the WGSR would give
active and selective catalysts in the ethanol SRR
and (ii) active metals in the WGSR would result in
poorly selective catalysts towards the formation of
CO
2
upon bio-ethanol steam reforming.
In the second part, concentrating on Rh and Ni
catalysts, the role of the oxide support (c-Al
2
O
3
,
12%CeO
2
Al
2
O
3
; CeO
2
; Ce
0:63
Zr
0:37
O
2
and ZrO
2
)
was investigated. To improve the performances of
the catalyst in the SRR, ceria-containing supports
(12%CeO
2
Al
2
O
3
; CeO
2
; CeO
2
ZrO
2
), with en-
hanced OH surface mobility [10,13,14], were used
as supports. Furthermore, ceria is known as a
promoter in the WGSR [7]. The results obtained
upon reaction at 600 C under atmospheric pres-
sure are presented in Table 2 for Rh-based cata-
lysts and Table 3 for Ni-based catalysts.
Looking at Rh catalysts (Table 2), 1%Rh/
c-Al
2
O
3
showed the highest selectivity while
1%Rh=CeO
2
ZrO
2
exhibited the highest yield in
hydrogen. 1%Rh/ZrO
2
activity could not be prop-
erly measured due to rapid deactivation. In fact, H
2
production is initially low and large amounts of
ethylene are detected. In descending order, the
following sequences could be establish for both the
activity: 1%Rh=Ce
0:63
Zr
0:37
O
2
> 1%Rh=12%CeO
2
c-Al
2
O
3
> 1%Rh=CeO
2
> 1%Rh=c-Al
2
O
3
and
the selectivity towards CO
2
:1%Rh=c-Al
2
O
3
> 1%
Rh=12%CeO
2
c-Al
2
O
3
> 1%Rh=CeO
2
> 1%Rh=
Table 1
Ethanol steam reforming at 700 C under stoichiometric reaction conditions (nH
2
O 3, nC
2
H
5
OH 1) at atmospheric pressure
on c-Al
2
O
3
supported Rh, Pt, Pd, Ru, Ni, Cu, Zn and Fe catalysts
Gas phase composition (%) H
2
yield
(g h
1
g
1
catalyst)
Selectivity
towards
CO
2
(%)
H
2
CO
2
CO CH
4
C
2
H
4
C
2
H
6
Thermodynamic equilibrium 69.2 10 20 0.8 0 0 33
1%Rh/c-Al
2
O
3
72 21 7 0 0 0 2.3 75
1%Pt/c-Al
2
O
3
46 7 13 12 21 1 0.6 35
0.75%Pd/c-Al
2
O
3
55 2 18 15 9 1 1.1 10
0.67%Ru/c-Al
2
O
3
38 2 9 12 38 1 0.3 18
9.7%Ni/c-Al
2
O
3
70.5 18 11 0.5 0 0 3.1 62
9.1%Cu/c-Al
2
O
3
40 1 12 21 23 1 0.4 8
9.8%Zn/c-Al
2
O
3
42 0 16 21 20 1 0.4 0
8.7%Fe/c-Al
2
O
3
44 5 10 20 20 1 0.3 33
Variation of the gas phase composition, the H
2
yield and the selectivity towards CO
2
.
F. Aupr^eetre et al. / Catalysis Communications 3 (2002) 263267 265
Ce
0:63
Zr
0:37
O
2
. So that: (i) in agreement with Du-
prezs bifunctional mechanism, the activity of the
catalyst in the SRR increases as the OH groups
mobility at the catalyst surface increases and (ii) the
selectivity of the catalyst towards CO
2
decreases as
the eciency of the catalyst in the WGSR in-
creases. The last observation tends to indicate that
CO
2
is produced as a primary product in the eth-
anol SRR and transformed into CO via the WGSR
to reach the thermodynamic equilibrium. In fact,
looking at the results obtained on ceria-containing
catalysts, one can observe that the thermodynamic
equilibrium is almost reached. The dierent prod-
uct fractions are comparable to the thermodynamic
equilibrium concentrations obtained from the cal-
culations performed using an homemade program.
Almost the same conclusions can be drawn from
the results obtained with Ni catalysts (Table 3).
The activity in the SRR directly varies as the
degree of mobility of surface OH groups 9:7%
Ni=Ce
0:63
Zr
0:37
O
2
> 9:7%Ni=CeO
2
> 9:7%Ni=12%
CeO
2
c-Al
2
O
3
> 9:7%Ni=c-Al
2
O
3
and the selec-
tivity towards CO
2
is controlled by the activity in
the WGSR 9:7%Ni=c-Al
2
O
3
> 9:7%Ni=12%CeO
2
c-Al
2
O
3
> 9:7%Ni=CeO
2
> 9:7%Ni=Ce
0:63
Zr
0:37
O
2
. Use of ceria-based supports, highly active in
the WGSR, immediately lead to the thermody-
namic CO/CO
2
equilibrium. As a result, activity in
the SRR and selectivity towards CO
2
vary in the
opposite direction.
In conclusion, these unexpected and unprece-
dented results clearly evidence the formation of
CO
2
as a primary product in the ethanol SRR on
some metal supported catalysts. This is a totally
dierent picture compared to the results obtained
from earlier studies on hydrocarbons steam re-
forming. Such results extensively showed that CO
was produced as a primary product. For that
reason, high temperature WGS and low tempera-
ture WGS units are required on methane steam
reforming industrial plants for the synthesis of
ammonia. Nevertheless, recent results on dimethyl
ether catalytic SR also suggested CO
2
and H
2
as
primary products [15]. As a result, the most se-
Table 3
Ethanol steam reforming at 600 C under stoichiometric reaction conditions (nH
2
O 3, nC
2
H
5
OH 1, P1 atm.) on Rh-based
supported c-Al
2
O
3
, 12%CeO
2
Al
2
O
3
; CeO
2
; CeO
2
ZrO
2
and ZrO
2
catalysts
Gas phase composition (%) H
2
yield
(g h
1
g
1
catalyst)
Selectivity
towards
CO
2
(%)
H
2
CO
2
CO CH
4
C
2
H
4
C
2
H
6
Thermodynamic equilibrium 64.7 14.9 13.4 7 0 0 53
9.7%Ni/c-Al
2
O
3
76 24 0 0 0 0 0.6 100
10%Ni/12%CeO
2
c-Al
2
O
3
65 15.5 3 3 13 0.5 2.5 84
10%Ni/CeO
2
63 21 8 8 0 0 4.1 72
10%Ni/Ce
0:63
Zr
0:37
O
2
62 21 9 8 0 0 4.4 70
10%Ni/ZrO
2
68.5 20 7 3 1.5 0 2.5 74
Table 2
Ethanol steam reforming at 600 C under stoichiometric reaction conditions (nH
2
O 3, nC
2
H
5
OH 1, P 1 atm.) on Rh-based
supported c-Al
2
O
3
, 12%CeO
2
Al
2
O
3
; CeO
2
; CeO
2
ZrO
2
and ZrO
2
catalysts
Gas phase composition (%) H
2
yield
(g h
1
g
1
catalyst)
Selectivity
towards
CO
2
(%)
H
2
CO
2
CO CH
4
C
2
H
4
C
2
H
6
Thermodynamic equilibrium 64.7 14.9 13.4 7 0 0 53
1%Rh/c-Al
2
O
3
73.5 22 3 1 0.5 0 2.2 88
1%Rh/12%CeO
2
c-Al
2
O
3
63.5 17 7.5 12 0 0 4.3 69
1%Rh/CeO
2
63 16 13.5 5.5 2 0 4.0 54
1%Rh/Ce
0:63
Zr
0:37
O
2
62 17 13 7.5 0.5 0 5.1 57
1%Rh/ZrO
2
57 17 5.5 3.5 16 0 0.5 74
266 F. Aupr^eetre et al. / Catalysis Communications 3 (2002) 263267
lective catalyst towards CO
2
in the bio-ethanol
SRR should contain no WGSR promoters.
References
[1] Y. Jamal, M.L. Wyszinski, Int. J. Hydrogen Energy 19 (7)
(1994) 557.
[2] M.A. Pe~ nna, J.P. Gomez, J.L.G. Fierro, Appl. Catal. A:
General 144 (1996) 7.
[3] G. Maggio, S. Freni, S. Cavallaro, J. Power Sources 74
(1998) 17.
[4] A.N. Fatsikostas, D.I. Kondarides, X.E. Verykios, Chem.
Commun. (2001) 851.
[5] D. Duprez, Appl. Catal. A: General 82 (1992) 111.
[6] D.C. Grenoble, M.M. Estadt, D.F. Ollis, J. Catal. 67
(1981) 90.
[7] A. Trovarelli, Cat. Rev. Sci. Eng. 38 (4) (1996) 439.
[8] J. Barbier Jr., D. Duprez, Appl. Catal. B: Environ. 4 (1994)
105.
[9] D. Duprez, P. Peireira, A. Miloudi, R. Maurel, J. Catal. 75
(1982) 151.
[10] R.G. Delosh, M.E. Heyde, SAE Technical Paper Series
No. 770166 (1977).
[11] C. Descorme, Y. Madier, D. Duprez, T. Birchem, Stud.
Surf. Sci. Catal. 130 (2000) 347.
[12] J.C. Schlatter, SAE Technical Paper Series No. 780199
(1978).
[13] D. Martin, D. Duprez, J. Phys. Chem. 100 (1996) 9429.
[14] Y. Madier, C. Descorme, A.M. Le Govic, D. Duprez,
J. Phys. Chem. B 103 (50) (1999) 10999.
[15] V.V. Galtica, G.L. Semin, V.D. Belyaev, T.M. Yurieva,
V.A. Sobyanin, Appl. Catal. A: General 216 (2001)
85.
F. Aupr^eetre et al. / Catalysis Communications 3 (2002) 263267 267