1158
Helvetica Chimica Acta ± Vol. 84 (2001)
Experimental Part
Materials. The natural zeolite clinoptilolite used in this study was a volcanic tuff from Horseshoe Dam,
Arizona, USA. A clinoptilolite content of 96 wt-% was determined from the adsorption capacity for CO2 [6].
The Si/Al ratio was 4.5, the cation-exchange capacity 2.37 mequiv./g, and the ignition loss 14.1 wt-%. A
powdered sample (30 g) was treated by ion-exchange 5 times, (8 h each) under reflux in the presence of 200-ml
portions of 0.1m CdCl2. The ion-exchange degree, (Cd/2Al) ´ 100, was 58%, i.e., the sample contained 7.7 wt-%
cadmium. The Cd-exchanged material was recovered by filtration, dried, and compressed into pellets without
binder. The pellets were crushed and sieved. The fraction of 0.6 ± 1.0 mm was used.
Pure acetylene was obtained from a flow of ꢀDissous gasꢁ (product of ODV, Hungary) by trapping acetone
at 195 K (the term ꢀDissous gasꢁ refers to the gas obtained from a tank containing acetylene absorbed in acetone
supported on pumice.)
The acetaldehyde (ꢀzur Syntheseꢁ grade) of Merck, Germany, was used.
Methods. A conventional flow-through reactor was used at 100 kPa total pressure. The reaction system was
equipped with feeding accessories for acetylene, water, acetaldehyde, and N2. The unconverted water,
containing dissolved acetaldehyde product, was condensed by cooling the reactor output to room temperature
and separated from the gas phase. The gaseous effluent of the condenser consisted of carrier N2, unconverted
acetylene, and acetaldehyde vapor. The amount and composition of the liquid and gaseous streams were
determined. A perfect material balance was found between the feed and the reactor effluents. Conversions (c)
were calculated by dividing the amount of product acetaldehyde (in mol) leaving the reactor by the amount of
acetylene (in mol) fed during the same period of time.
The initial partial pressures of acetylene (pAo), water (pWo), and acetaldehyde (pAAo) in the feed and the
space-time of acetylene, t wcat/F (wcat weight of catalyst/gcat; F feed rate of acetylene/molA
s
1) were varied.
Conversions at given initial partial pressures and temperature were plotted against space-time. The slopes of the
conversion curves are equal to the rates of reaction under the applied conditions. The actual partial pressures pA,
pW, and pAA can be calculated from the initial partial pressures pAo, pWo, the actual conversion c, and the mol-
fraction xAo of acetylene in the feed by Eqns. 1 and 2, and, provided that acetaldehyde is not fed (pAA 0), by
o
Eqn. 3, respectively.
o
pA (1 c)pAo/(1 c xA
)
(1)
(2)
(3)
pW (pW c pAo)/(1 c xAo),
o
o
pAA c pAo/(1 c xA
)
Results and Discussion. ± Two series of measurements were carried out. In the first
o
series, the temperature was 413 K; pAo was increased to 33 kPa at constant pW 67 kPa
o
(Conditions A), pWo was varied between 17 and 83 kPa, while pA 17 kPa (Conditions
o
o
B), and, at pA 17 kPa and pW 67 kPa, pAAo was varied in the range 0.15 to 2.5 kPa
(Conditions C). The total pressure was maintained at 100 kPa by admixing N2. Initial
reaction rates (ro) were determined at t 0 with a reliability of Æ 2%. Results are
shown in Fig. 1,a ± c.
In the second series of measurements, experiments were carried out at 393, 413, 433,
o
o
o
and 453 K at pA 33 kPa, pW 67 kPa, and pAA 0. Conversion curves are shown in
Fig. 2. For the sake of brevity, only the reaction rates determined at different
conversions at 413 K are summarized in the Table.
Fig. 1 demonstrates that the initial reaction rate ro is proportional to the partial
pressure of acetylene and shows negative order in water and acetaldehyde. The
experimental data were evaluated with different tentative kinetic models taken into
account. The following cases were considered:
Case 1: Langmuir-Hinshelwood kinetics, where it is assumed that the rate-determin-
ing process step is the adsorption of a) the acetylene or b) the water, c) is the surface reac-
tion between the adsorbed reactants, or d) is the desorption of the acetaldehyde product;
Kallo, Denes
