Platinum/alumina catalyst oxidation reactions

An earlier study using this same compound, DMMP, led to a mathematical model of the deactivation process. Graven et a/. studied the oxidation of DMMP vapor in a stream of air, or nitrogen, over platinum-alumina catalysts. A commercial catalyst and a number of laboratory-prepared catalysts were investigated over a range of temperatures from 573-773 K, residence times from 0.15 to 2.7 seconds. The average catalyst particle sizes varied from 0.31 to 2.4 mm. They found that the fresh catalyst showed a very high activity, but after a few hours on stream it deactivated to the point that measurable quantities of DMMP vapor appeared in the effluent.. The reaction products over the deactivated catalyst were methanol and phosphorus acid. 

The catalytic dehydrogenation of lower alkanes was first developed more than fifty years ago using chromia/alumina systems [1]. Although there has been development of new processes [2 - 6], the catalyst technology has tended to remain with either modified chromia/alumina or modified platinum/alumina catalysts. Therefore it seemed appropriate to re-examine the possibility of using oxide systems other than chromia to effect the alkane to alkene transition. Supported vanadium pentoxide has been extensively studied for the oxidative dehydrogenation of propane to propene [7-10] but rarely for the direct dehydrogenation reaction [6]. 

A silicon microreactor for preferential oxidation was designed by Srinivas et al. [163], which was 6 cm x 6 cm wide and long, while the flow path was only 400 pm high. Instead of microchannels, pillars were chosen for the flow distribution in the reactor. The reactor was coated with 2 wt.% platinum/alumina catalyst with a thickness of 10 pm. Tests were performed at an O/CO ratio of 2.0 and a high VHSV of 120 l/(h gcat)- Not more than 90% conversion of carbon monoxide could be achieved in the reactor at 210°C reaction temperature, while similar results were obtained for a small fixed catalyst bed. 

Some metallic compounds present in trace level (ppm) in petroleum feed, adsorbed to the active site of the catalyst, act and change the selectivity of the reaction by producing more and more unwanted products. When water vapour is present in the sulphur dioxide-air mixture supplied to a platinum-alumina catalyst, a decrease in oxidation activity occurs. This type of poisoning is due to the effect of wafer on fhe sfrucfure of the alumina carrier and is known as stability poisoning. The resulting increase in diffusional resistance may dramatically increase the Thiele modulus, and reduce the effectiveness factor for the reaction. In extreme cases, the pressure drop through a catalyst bed may also increase dramatically. 

The reaction rate of preferential oxidation of carbon monoxide increases with increasing pressure. For a platinum/alumina catalyst Kahlich et al. determined that the reaction rate was proportional to p [114]. 

Figure 4.29 Effect of 21 ppm sulfur dioxide on the activity of a platinum/alumina catalyst for the preferential oxidation of carbon monoxide the catalyst contained 3.1 wt.% platinum on alumina reaction temperature 175°C [338].

Platinum combustion catalysts are probably more tolerant to sulfur poisoning than palladium catalysts [350,355]. Corro et al. reported that sulfur dioxide may even have a promoting effect on propane oxidation [356]. They claimed that aluminium sulfate needs to be present on the catalyst surface to promote the reaction in the low temperature range below 300 °C. The sulfate formation was assumed to start at temperatures exceeding 500 °C. Therefore, the catalyst must have previously been exposed to such a temperature in the presence of sulfur dioxide. Over a pre-sulfated platinum/alumina catalyst, 50% methane conversion was achieved by 530 °C, while 560 °C was required for the sulfur-free counterpart [357]. However, no promotion effect is to be expected over non-sulfating carrier materials, such as sflica, according to Gelin and Primet [351]. 

Purely parallel reactions are e.g. competitive reactions which are frequently carried out purposefully, with the aim of estimating relative reactivities of reactants these will be discussed elsewhere (Section IV.E). Several kinetic studies have been made of noncompetitive parallel reactions. The examples may be parallel formation of benzene and methylcyclo-pentane by simultaneous dehydrogenation and isomerization of cyclohexane on rhenium-paladium or on platinum catalysts on suitable supports (88, 89), parallel formation of mesityl oxide, acetone, and phorone from diacetone alcohol on an acidic ion exchanger (41), disproportionation of amines on alumina, accompanied by olefin-forming elimination (20), dehydrogenation of butane coupled with hydrogenation of ethylene or propylene on a chromia-alumina catalyst (24), or parallel formation of ethyl-, methylethyl-, and vinylethylbenzene from diethylbenzene on faujasite (89a). 

More than three decades ago, skeletal rearrangement processes using alkane or cycloalkane reactants were observed on platinum/charcoal catalysts (105) inasmuch as the charcoal support is inert, this can be taken as probably the first demonstration of the activity of metallic platinum as a catalyst for this type of reaction. At about the same time, similar types of catalytic conversions over chromium oxide catalysts were discovered (106, 107). Distinct from these reactions was the use of various types of acidic catalysts (including the well-known silica-alumina) for effecting skeletal reactions via carbonium ion mechanisms, and these led... 

As a result of the studies discussed above, a reasonably consistent picture of the kinetics and mechanism of the dehydrocyclization reaction over oxide catalysts has evolved. However, as pointed out earlier in this section, relatively few kinetic data have been reported for dehydrocyclization over platinum-alumina reforming-type catalysts. The data which have been reported include those of Hettinger and co-workers (H7), who studied the dehydrocyclization of re-heptane over platinum catalysts. These investigators found that the rate of dehydrocyclization decreased with increasing total pressure at a constant hydrogen-to-hydrocarbon ratio (Fig. 9). They also reported that the extent of dehydrocyclization was substantially greater for re-nonane than for re-heptane, which is consistent with the results obtained on oxide catalysts. In a later study of the kinetics... 

Partial oxidation runs at 700-1000 °C, typically on a platinum or rhodium catalyst supported on alumina or other oxides and c) Autothermal Reforming (ATR) which combines steam reforming and partial oxidation reactions to produce a roughly thermo-neutral reaction ... 

Similarly the fast oxidation reaction of carbon monoxide proves to be amenable to the concept of opposing-reactant geometry [Veldsink et al., 1992]. In this case, alpha-alumina membrane pores are deposited with platinum as the catalyst for the reaction to proceed at about 250°C. 

Clearly, the best catalyst for the reduction reactions may not be the best for the oxidation reactions, so two catalysts are combined. The noble metals, although expensive, are particularly useful. Typically, platinum and rhodium are deposited on a fine honeycomb mesh of alumina (AI2O3) to give a large surface area that increases the contact time of the exhaust gas with the catalysts. The platinum serves primarily as an oxidation catalyst and the rhodium as a reduction catalyst. Catalytic converters can be poisoned with certain metals that block their active sites and reduce their effectiveness. Because lead is one of the most serious such poisons, automobiles with catalytic converters must use unleaded fuel. 

Example 9-2 Olson and Schuler determined reaction rates for the oxidation of sulfur dioxide, using a packed bed of platinum-on-alumina catalyst pellets. A differential reactor was employed, and the partial pressures as measured from bulk-stream compositions were corrected to fluid-phase values at the catalyst surface by the methods described in Chap. 10 (see Example 10-1). The total pressure was about 790 mm Hg. 

Two aspects of these results will be discussed here, namely the low platinum dispersion and the presence of significant amounts of coke on alumina and zeolite catalysts after testing. The latter finding is prohahly due to a combination of test tenq>eratures below 2S0°C and the propensity for toluene to form coke. In any event, coke formation did not seem to inhibit the oxidation reaction to any significant extent. The low platinum diversions, confirmed by XRD measurements, indicate that the active reside outside the zeolite micropores on... 

In order to evaluate the interaction between platinum and rhodium deposited on the alumina-lanthanum oxide, the different catalysts were characterized by temperature programmed reduction and measure of the activity for the reaction of propane-propene oxidation. 

Figure 7.15. NOx to N2 activity with octane as a reducing agent over silver on alumina catalyst and a platinum oxidation catalyst depending on the distance between the catalysts (K. Eranen, L.-E.Lindfors, F. Klingstedt, D.Yu.Murzin, Continuous reduction of NOx with octane over a silver/alumina catalyst in oxygen-rich exhaust gases combined heterogeneous and surface mediated homogeneous reactions, Journal of Catalysis, 219 (2003) 25).

Ammonia selectivity of platinum and platinum-nickel catalysts for NOx reduction varies with the nature of the supporting oxide. Silica, alumina, and silica-alumina supports on monolithic substrates were studied using synthetic automotive exhaust mixtures at 427°-593°C. The findings are explained by a mechanism whereby the reaction of nitric oxide with adsorbed ammonia is in competition with ammonia desorption. The ease of this desorption is affected by the chemistry of the support. Ammonia decomposition is not an important reaction on these catalysts when water vapor is present. 

The phosphorus deactivation curve is typical type C, and, according to the Wheeler model, this is associated with selective poisoning of pore mouths. Phosphorus distribution on the poisoned catalyst is near the gas-solid interface, i.e. at pore mouths, which confirms the Wheeler model of pore mouth poisoning for type C deactivation curves. Thus we may propose that in the fast oxidative reactions with which we are dealing, transport processes within pores will control the effectiveness of the catalyst. Active sites at the gas-solid interface will be controlled by relatively fast bulk diffusional processes, whereas active sites within pores of 20-100 A present in the washcoat aluminas on which the platinum is deposited will be controlled by the slower Knudsen diffusion process. Thus phosphorus poisoning of active sites at pore mouths will result in a serious loss in catalyst activity since reactant molecules must diffuse deeper into the pore structure by the slower Knudsen mass transport process to find progressively fewer active sites. 

---