Propylene adsorption, active site

The spectrum of chemisorbed propylene in the CH and O—H stretching region is shown in Fig. 15. The band at 3593 cm-1 is clearly due to an OH frequency hence, dissociation accompanies propylene adsorption. Careful scrutiny of the region from 1500 to 2000 cm-1 reveals no band assignable to a ZnH band. Since the presence of adsorbed propylene has been found to block out the infrared active hydrogen chemisorption on the ZnO pair sites, we may assume that propylene adsorption occurs as follows ... 

In the previous sections of this review, it has been shown that most effective catalysts for the selective oxidation of propylene contain at least two types of metal oxides—an amphoteric or low-valence oxide, such as bismuth, tin, iron, or cobalt, and an oxide of a high valence metal, such as molybdenum or antimony. Moreover, it has been suggested several times that each of these metal oxide components may give rise to an active site for example, propylene may adsorb on an active site associated with one of the metal oxide components, and oxygen may adsorb on an active site associated with another metal oxide component. This problem has been studied using spectroscopic, adsorption, and kinetic techniques. It now seems appropriate to consider some of these studies in detail, attempting to relate the solid structure of the catalyst to the active sites wherever possible. 

Kr adsorption coefficient for propylene, atm-l k surface reaction rate constant, s l Lc concentration of total active sites, mole/g m constant in equation (22)... 

Senchenya et al. (96) have treated the adsorption of ethanol on a structural hydroxyl group (Fig. 14) using a CTP scheme and the CNDO/BW method. The separation of a molecule and cluster with respect to the z axis was optimized, the optimal values being r = 1.19 A and R = 1.28 A The adsorption energy was 23.2 kcal/mol, which was close to the experimental value (97). Note that this was essentially the two-point adsorption involving both acid and base sites. This case is quite similar to the above propylene adsorption (90). There is also no definite trend toward proton transfer from the hydroxyl group of a zeolite to the alcohol molecule. The carbocation state is also predicted to be activated. This, in turn, increases relative efficiency of the synchronous mechanism (with the same recommendation for its experimental examination). The estimation (96) of the energetics of the intermediate structures of the synchronous mechanism showed that such a mechanism is quite realistic. 

This article therefore seeks to examine in depth just one mixed oxide catalyst, tin-antimony oxide, which has been commercially developed (2-5) for the oxidation of propylene to acrolein as well as for the ammoxidation of propylene to acrylonitrile and the oxidative dehydrogenation of butenes to 1,3-butadiene. A recent book (6) and a subsequent review (7) have shown how little unanimity has been established about the fundamental properties of the material. In particular there seems to be much confusion as to the phase composition, the nature of the cationic oxidation states, the chemical environment of the cations, the charge compensation mechanism, the nature of the active sites, the distortion of the host tin(IV) oxide lattice by the dopant antimony atoms and whether any changes in the catalyst result from the adsorption and catalytic processes. 

One of the first comprehensive kinetic analyses over Au/TS-1 concerned the oxidation of H2 to form water [65]. DPT calculations [63,65] showed the RDS to be the formation of adsorbed H2O2 which ultimately decomposes to form water [65]. In the presence of propylene, the generated H2O2 is expected to perform the epoxidation, therefore similarities between the production of PO and water can likely be drawn. Traditional kinetic analysis [65] produced a power rate law for water production of rn o = h exp[-(37.1 1.1 kj mol- )/RT][H2]° 0.02jqj0.17 0.02 Development of a series of elementary steps [Eqs. (11.1-11.5)] capable of reproducing the observed experimental orders and consistent with DPT calculations proved to require two active sites one capable of nondissociative adsorption of O2 and dissociative adsorption of H2 and a second available for only dissociative adsorption of H2 [65]. The resulting rate expression [65] is presented as Eq. (11.6). 

Although the mechanism is similar to the mechanistic steps presented in reference [65], Taylor et al. [55] used enhanced H2 dissociation in the presence of O2, consisfenf wifh bofh previous experimenfal observation [91] and DFT calculations [63,64], Moreover, the rate determining step (Eq. 11.12) is unique because the fractional propylene order required the inclusion of adsorbed propylene in fhe RDS [55]. This mechanism and ifs associated rate expression (Eq. 11.15) were the simplest means of reproducing fhe observed reaction orders (n = 0.18, m = 0.14). It does imply a relation between H2 and O2 orders m + 1/2, 2rti), however. The addition of a third active site for dissociative H2 adsorption [65] would provide independent control over all three reaction orders. 

It has been suggested that the Incorporation of alkali metals on Ti02-vanadia catalysts decreases both the V=0 stretching frequencies and their polarizing power, while the incorporation of acid anions produces an opposite trend [20]. In addition, the presence of alkali ions decreases the heat of the propylene adsorption [17,18, 21]. Thus the different catalytic behavior of doped alumina supported vanadia catalysts, could be explained on the bases of the influence of the acid-base character of catalysts on the adsorption/desorption of propane and propene. In any case, the redox properties must be also considered. In this way, it will be interesting to study if, realy, a lower reducibility of the active sites could favor a lower rate of the consecutive reactions, as it has been observed in the case of K-doped catalysts. 

Fig.6 shows PO yield over 8 wt % Au/TiOj/SiO, as a function of time. The catalytic activity of Au/TiOj/SiOj catalyst is not stable. Water is continuously formed during the oxidation of propylene and the oxygenated intermediates may block the active sites and depress the adsorption of propylene on the surface of the catalyst. MCM materials have hydrophobic character and Ti-MCM preferentially adsorbs less polar olefin molecules. This decreases the competition from water and probably avoid the accumulation of the oxygenated intermediates to lead to more stable catalytic activity. 

At this point it is interesting to compare the evolution of propylene adsorption over catalysts with different surface acid characteristics, i.e. a MoVTeNbO catalyst (active and selective in the partial oxidation of propane to acrylic acid), an alumina-supported vanadium oxide (active in the ODH of propane to propylene), or a MoVNbO mixed oxide (active in the oxidative transformation of propane to propylene and acetic acid). The final products observed in each case were related to the characteristics of the adsorbed intermediates (Fig. 24.7) (i) a ir-allylic compound, interacting with a redox site intermediate in the selective oxidation of... 

Maximum number of active sites for adsorption/reaction of isopropanol calculated from the addition of die amount of propylene produced during chemisorption and TPSR experiments. 

Post-synthesis alumination using A1(N03)3 as the precursor improves the acidity of siliceous MCM-41 materials significantly. FTIR results show that both Bronsted and Lewis acid sites are increased upon alumination. The number of acid sites increases with the Al content on MCM-41. NH3-TPD reveals the mild strength of these created acid sites. Due to the improved acidity, the catalytic activity for dehydration of isopropanol to propylene over these alumina-modified MCM-41 materials is considerably promoted by post-synthesis alumination. The results of XRD and N2 adsorption show that the enhancement of acidity for siliceous MCM-41 by postsynthesis alumination does not cause any serious structural deformation of the resulting material. 

Unfortunately, general agreement concerning the adsorption site for propylene has not been reached. One school of thought has developed which postulates the low-valence cation as a site of propylene activation. Another school of thought proposes the high-valence cation as the site of propylene activation. 

Further evidence supporting the bismuth center as a site of propylene activation comes from the analysis of the rates of formation and product distribution of propylene oxidation over bismuth oxide, bismuth molybdate, and molybdenum oxide. Bismuth molybdate is highly active and selective for the conversion of propylene to acrolein. However, the interaction of propylene with its component oxides yields very different results. Haber and Grzybowska (//. ), Swift et al. 114), and Solymosi and Bozso 115) showed that in the absence of oxygen, propylene is converted to 1,5-hexadiene over bismuth oxide with good selectivity and at a high rate, whereas molybdenum oxide is known to be a fairly selective but a nonactive catalyst for acrolein formation. The formation of 1,5-hexadiene over bismuth oxide can be explained if the adsorption of propylene on a bismuth site yields a ir-allylic species. Two of these allylic intermediates can then combine to give 1,5-hexadiene. 

A somewhat different picture emerges from the adsorption studies of Matsuura and Schuit 117). They have attempted to elucidate the active adsorption sites on -y-bismuth molybdate by measuring the adsorption equilibria of butene, propylene, oxygen, water, butadiene, acrolein, and... 

---