Mechanism of di(methyl)ether (DME) electrooxidation at platinum electrodes in acid
medium
G. KE
´
RANGUE
´
VEN, C. COUTANCEAU*, E. SIBERT, F. HAHN, J- M. LE
´
GER and C. LAMY
UMR-CNRS 6503, Universite
´
de Poitiers, Equipe Electrocatalyse, 40 avenue du Recteur Pineau, 86022, Poitiers
cedex, France
(*author for correspondence, tel:+33-5-49-45-48-95, fax: +33-5-49-45-35-80, e-mail: christophe.coutanceau
@univ-poitiers.fr)
Received 28 February 2005; accepted in revised form 31 October 2005
Key words: dimethylether, electrooxidation, fuel cell, in situ Infrared Reflectance Spectroscopy, platinum
Abstract
The electrooxidation of DME was studied at a bulk platinum electrode. It was shown that the DME adsorption was
a slow step in the overall oxidation reaction. The DME adsorption is potential dependent in the hydrogen region of
platinum and independent in the double layer region. From low potential scan rate voltammetry and DME
stripping experiments, it was shown that the DME oxidation mechanism occurred via several reaction paths. At low
potentials, DME oxidation leads to the existence of a positive current plateau. ‘‘In situ’’ Infrared Reflectance
Spectroscopy experiments were carried out to identify the intermediate and reaction products of DME adsorption
and oxidation at different potentials. CO
L
(linearly bonded CO), CO
B
(bridge bonded CO), adsorbed COOH species
and CO
2
were detected. From these electrochemical and spectro-electrochemical results, it was proposed that some
adsorbed DME was hydrolysed and directly oxidized to CO
2
or HCOOH species and some partially blocked
platinum sites at the surface forming Pt–CHO and/or Pt–CO. Then, as soon as platinum becomes able to activate
water, a bifunctionnal mechanism occurs to form either HCOOH or CO
2
again following two different reaction
paths.
1. Introduction
Because of the difficulty of handling and storing hydro-
gen, many research agencies tend to develop fuel cells
with the direct combustion of liquid fuels: the drastic
decrease in the overall mass energy density when hydro-
gen is stored (Table 1) makes the use of liquid fuels such
as alcohols very promising. Methanol and ethanol are the
most studied fuels. With methanol, which only contains
one carbon atom, power densities close to 180–200 mW
cm
)2
can be achieved at 90 °C [1–4], which makes
methanol a good alternative liquid fuel to hydrogen in
terms of electrical performance. But, because of its high
toxicity, methanol is not a good candidate, particularly
for automobile applications. Other alcohols are envis-
aged and studied for this purpose. Ethanol is of course
the most studied [5, 6], because it is the simplest alcohol
after methanol; it has little toxicity and can be produced
from biomass, whereas methanol is produced from
natural gas. Another alternative fuel for direct fuel cell
applications is DME (dimethylether, also called methoxy
methane) [7–10]. This compound, which has no C–C
bond to break, has a good theoretical energy density of
8.2 kWh kg
)1
, which is higher than that of methanol and
close to that of ethanol (Table 1). Moreover, according
to the thermodynamic data of DME combustion with
oxygen (DG
reac
=)1362 kJ mol
)1
) [7], the standard
reversible cell voltage related to the following equation
(1) is 1.18 V:
CH
3
ÀOÀCH
3
þ3H
2
O!2CO
2
þ12H
þ
þ12e
À
ð1Þ
3O
2
þ 12H
þ
þ 12e
À
! 6H
2
O ð2Þ
CH
3
ÀOÀCH
3
þ 3O
2
! 2CO
2
þ 3H
2
O ð3Þ
with a total number of exchanged electrons of 12, as for
the electrooxidation of ethanol. DME is also an
interesting compound for fuel cell applications because
of its physical properties: it is a gas at room tempera-
ture, but its boiling point is only )23 °C and liquid
phase storage as for LPG is easy. It can also be used as a
liquid fuel because of its solubility in water close to
76 g l
)1
(i.e. 1.65 mol l
)1
)at20°C under a pressure of
1 bar [8]. This solubility corresponds to the concentra-
tion ranges generally used in a Direct Methanol Fuel
Cell [11–13] or a Direct Ethanol Fuel Cell (DMFC) [5,
14, 15].
Journal of Applied Electrochemistry (2006) 36:441–448
Ó
Springer 2005
DOI 10.1007/s10800-005-9095-6