Characterization of the active sites on the surface of alumina ethanol dehydration catalysts by EPR using spin probes, страница 5

The formation of the primary radical cations on the surface acceptor sites is known to be a fast process taking place without substantial activation energy. So, the concentration of the radical cations observed immediately after adsorption can be attributed to the concentration of the sites capable of ionizing anthracene molecules directly. The subsequent growth in the concentration of the observed radical species can be attributed to the gradual polycondensation of the adsorbed aromatic molecules. Previously it was shown that such polycondensation is readily observed after benzene adsorption on the strong acceptor sites of H-ZSM-5 zeolites. The difference of the ionization is the driving force of such polycondensation following the electrophilic substitution mechanism. The aromatic radicals and neutral molecules adsorbed on the surface acceptor sites that are not sufficiently strong to abstract the electron from the adsorbed aromatic molecules area strong electrophiles that can participate in such process. The radical species are generated when the ionization potential of the adsorbed polycondensed structure becomes sufficiently low for the molecule to be ionized. Note, that such gradual increase in the concentration of the observed radical species was not observed in the previous studies using zeolites as the adsorbents due to the limited size of the zeolite channels where the radical cations are generated.

Obviously, this process should be characterized by some activation energy. Indeed, the radical generation rate in the first 100 h after anthracene adsorption on Al2O3 at room temperature was approximately constant, being equal to 3.7 x 1015 g-1 h-1 (Fig. 3). Then, it decreases as the concentration approaches the maximum between 100 and 200 h after adsorption. At 80°C the radical are formed substantially faster. The concentration of the radicals generated in 12 h at this temperature was about 4 x 1017 g-1 that is about 80% of maximum value obtained at room temperature.

As the process is complex and involves sites of different strengths it is not possible to determine precisely the activation energy. Note that the minimum strength (electron affinity) of the acceptor sites tested by this method is not exactly equal to the ionization potential of the adsorbed donor molecules. Actually, the method tests somewhat weaker sites as well, and their exact strength is unknown. Still such procedure involving heating of the sample at 80°C for 12 hours appears to be a simple, reasonably fast and readily reproducible procedure for characterization of the weak acceptor sites.

As the studied redox process involved only a small number of the surface sites, it was natural to analyze the possible role of transition metal impurities in this process. The main sample used by us was rather pure and contained about 100 ppm Fe2O3 and no other transition metal ions in this concentration range (> 10 ppm). The total Fe concentration in the sample slightly exceeded the measured concentration of the generated radical cations. However, it is very unlikely that all Fe ions in this sample are located on the surface and are active in the studied process. We specially studied several other alumina samples with higher Fe concentrations. Still, the surface-normalized concentrations of the generated radical cations were about the same. Overall, we believe that transition metal ions do not account for the formation of the radical cations from anthracene on Al2O3.

Based on the results of these preliminary experiments we selected the optimum conditions yielding quantitative data on the concentration of the weak acceptor sites with good reproducibility (about 10%) and minimum effort that could be used in routine experiments. In al the following experiments the catalysts were activated in air at 400°C. This temperature is close to the one used in the catalytic experiments on the ethanol dehydration. Anthracene was adsorbed from 4 x 10-2 M solution in toluene. The concentration of the weak acceptor was assumed to be equal to the concentration of radicals generated at 80°C in 12 hours after adsorption.