230
B. Matas Güell et al. / Journal of Catalysis 257 (2008) 229–231
loaded in a fix-bed reactor and held by quartz wool plugs. The cat-
−1
alyst was first reduced in situ under 5% H2/Ar flow (50 ml min
)
◦
at 500 C for 30 min. After purging the reactor with Ar for 15 min,
the temperature was lowered to 320 C [2], in order to prevent any
◦
homogeneous decomposition of acetic acid as well as Pt metal sin-
tering. CH3COOD (Aldrich, 99 at% D) or CH4 pulses (17.5 μmol each)
were injected into the reactor, using a microsyringe. Outlet com-
position was measured online with a mass spectrometer (Balzers
QMS 200 F). Unconverted acetic acid was delayed from the gas-
mixture using a cold trap. This helped to avoid overlapping of the
MS-signals of the gaseous products formed, with those from acetic
acid. Blank experiments indicated no decomposition of acetic acid
at this temperature. Correspondingly, delayed broad signal corre-
sponding to acetic acid was only observed. All gaseous products
were determined semi quantitatively.
3. Results and discussion
Fig. 1 shows the spectra obtained during three subsequent
◦
pulses (I, II and III) of CH3COOD over Pt/C (graphite) at 320 C.
Contacting CH3COOD (m/z 43) over the catalyst resulted in the for-
mation of the following species: CO2 (m/z 44), CH4 (m/z 16), CH3D
(m/z 17), H2 (m/z 2), CH2D2 (m/z 18), HD (m/z 3), CHD3 (m/z 19)
and D2 (m/z 4).
CO2 is formed when acetic acid decomposes over Pt, as shown
by us earlier [1] and here it should come from CH3COOD according
to Eq. (1),
∗
∗
∗
CH3COOD + 2 → CH3 + CO2(g) + D ,
(1)
where the symbol ∗ denotes a metal site. CH3D and D2 (Fig. 1)
can be primary desorption products assuming recombination on
∗
∗
the catalyst surface of the species formed in Eq. (1) (CH3 and D ).
The other gaseous products observed (CH4, CH2D2, CHD3, H2 and
HD (see Fig. 1) cannot originate directly from the decomposition of
CH3COOD according to Eq. (1). Further, if the direct recombination
∗
∗
of CH3 with D was facile, these deuterated gas products would
not have been observed.
The following routes can be proposed, in agreement with Za-
era [11] and Kemball [12], concerning methane formation, i.e.,
(i) hydrogen redistribution—via dehydrogenation/hydrogenation
reactions—among CHx (1 ꢀ x ꢀ 3) species (e.g., CH3(ads) ꢁ CH2(ads)
+
H
; CH3(ads) + H(ads) → CH4(g), etc.) or (ii) hydrogen/deuterium
(ads)
∗
∗
redistribution among CH4−xDx and surface H or D species (e.g.,
CH3D(g) +H(ads) ꢁ CH4(g) + D
). In the former case a Langmuir–
(ads)
Hinshelwood and in the latter Eley–Rideal-type mechanisms are
involved as at least one of the species is in the adsorbed state. The
possibility of this Eley–Rideal-type exchange is also reported by
Guczi et al. [13] who observed exchange of D2 with CH4 over pure
Fig. 1. Typical product distribution for CH3COOD pulses over 5 wt% Pt/C (solid line)
◦
and over empty reactor (dash line) at 320 C.
◦
platinum in the temperature range from 300 to 400 C. In both
cases the presence of H atoms on the platinum is essential, sup-
porting partial dehydrogenation of CHx species formed in Eq. (1).
Further, it is important to recall that gas phase (noncatalytic) H–D
exchange (e.g., between CH3D and D2) is not probable under the
experimental conditions reported here [14–17].
From a pool of surface species such as CHx (0 ꢀ x ꢀ 3), H and D,
all species such as CH4−xDx (e.g. CH3D) can be formed. Compari-
son of the amounts of “C” based products based on elementary
statistical probability calculation with those from our experiments
showed that the amount of CH4 formed experimentally was larger
than would be expected if H–D redistribution was the only route
for CH4 formation. CH4 (Expt.—45%, Stat.—32%), CH3D (33% vs
42%) and CH2D2 (16% vs 21%) were observed in different amounts
than predicted by statistical calculations. Deviations from statis-
tics in our kinetic experiments may indicate that scrambling is
not the only pathway for the formation of CH4−xDx (0 ꢀ x ꢀ 4)
species.
Observation so far supports our earlier proposition [1,2] that
formation of CHx type species on Pt was involved in the steam
reforming of acetic acid over Pt/ZrO2. In the steam reforming of
methane, activation of CH4 on metal based catalyst is a crucial
step. It is known [18] that methane activation, leading to the reac-
tive CHx species for steam reforming, requires higher temperatures
◦
(800 C) since it involves rupture of C–H bond (435 kJ/mol [19]).
◦
Under our experimental conditions (320 C), pulsing of CH4 over
Pt/C (graphite) resulted in no detectable methane conversion as
well as no hydrogen and coke formation, confirming that methane
indeed cannot be activated at such low temperature. However, in
the case of acetic acid, fragmentation of the C–C bond, which is
energetically more favorable (368 kJ/mol [19]) directly results in
the reactive CHx type species. Thus, steam reforming of acetic acid
follows a route similar to that proposed over Pt/ZrO2 for methane
steam reforming but can occur at milder temperatures.
Matas Gueell
