Kinetics of ethylene oxidations

The catalyst modified with selenium is most suitable for the studies of reaction kinetics since this element, in contrast to chlorine usually used as promoter in the commercial processes, does not volatilize from the surface of silver under the reaction conditions. We studied the kinetics of ethylene oxidation under gradientless conditions (Section II) using a circulation flow system in the experiments at atmospheric pressure (59-61) and a reactor with rotating baskets for the catalyst (5) at elevated pressures (62). 

F. Kinetics of Ethylene Oxidation to Ethylene Oxide over Silver. 474... 

Todes and Adrianova (113) studied the kinetics of ethylene oxidation over pumice-supported silver and suggested that ethylene was converted to ethylene oxide, which yielded C02 and H20. In other words two consecutive reactions took place... 

The data on kinetics of ethylene oxidation to ethylene oxide are summarized in Table IX. 

Temkin et al. (159) studied the kinetics of ethylene oxidation over a stationary silver surface. It was shown by means of the flow-circulating method that the rate of ethylene oxide and carbon dioxide formation was proportional to ethylene concentration in the gas phase, and that there was inhibition with reaction products. 

The curve in Fig. 1 is drawn for n = 0.08, m = 0.40. Thus the parallel oxidation to COg is more significant than the consecutive oxidation of ethylene oxide. Most studies of kinetics of ethylene oxidation agree on the predominance of parallel oxidation to ethylene oxide and CO2. It is likely that the few studies which show important consecutive oxidation of ethylene oxide, such as Twigg (54) and Andrianova and Todes (35), are not at all typical of normal flow-system oxidation over silver. 

There is a bewildering amount of information on the kinetics of ethylene oxidation over silver. Static, flow, differential, and recycle systems have been used with various catalysts, feed ratios, and additives. From this work two conclusions are clear (1) it is not possible to cover all conditions with a simple rate law and (2) attempts to determine mechanisms by means of kinetics are quite risky. Earlier studies have been reviewed by Margolis (f), Sampson and Shooter (2), and Dixon and Longfield (S7). The classic paper was that of Twigg (34), who used mainly a static system. By now it is evident that, although Twigg s results contain much of value, they are not directly applicable to flow systems using modem catalysts. Under practical conditions the... 

A comprehensive study of the kinetics of ethylene oxidation was made by Kurilenko and co-workers 58), using a recirculating-flow reactor, with and without product removal. Concentrations were varied within the following limits, at a total pressure of 1 atm ... 

Table VII surveys other papers on kinetics of ethylene oxidation...

Example 2.2-2 Kinetics of Ethylene Oxidation on a Supported Silver Catalyst... 

Mugherz, P.D. and Harriott, P. (1971) Kinetics of ethylene oxidation on a supported silver catalyst AIChEf, 17, 856-866. 

Ethylene oxide secondary oxidation with C-tagged ethylene oxide, to clarify the source of CO2, was done at Union Carbide but not published. This was about 10 years before the publication of Happel (1977). With very limited radioactive supply only a semi-quantitative result could be gained but it helped the kinetic modeling work. It became clear that most CO2 comes from ethylene directly and only about 20% from the secondary oxidation of ethylene oxide. 

Figure 8.79. Steady-state kinetics of C2H4 oxidation on Pt/Ce02 as a function of catalyst potential, UWR, and ethylene partial pressure (a) catalyst A, T=500°C, pO2=5.0 kPa (b) catalyst C, T=510°C, Pc2h4=4-8 kPa.71 Reprinted by permission of The Electrochemical Society.

A packed-bed nonpermselective membrane reactor (PBNMR) is presented by Diakov et al. [31], who increased the operational stability in the partial oxidation of methanol by feeding oxygen directly and methanol through a macroporous stainless steel membrane to the PB. Al-Juaied et al. [32] used an inert membrane to distribute either oxygen or ethylene in the selective ethylene oxidation. By accounting for the proper kinetics of the reaction, the selectivity and yield of ethylene oxide could be enhanced over the fixed-bed reactor operation. 

The free amino group of the amino ester may then react analogously with another molecule of the monomer, etc. The kinetics of the polymerization are in harmony with a mechanism of this sort. The final polypeptide may contain up to 300 or more structural units. While the polymerization of N-carboxyanhydrides is closely analogous to the addition polymerizations of ethylene oxide and of other cyclic substances, definition unfortunately classifies it as a condensation polymerization inasmuch as carbon dioxide is eliminated in the process. 

Quite often in the ring-opening polymerization, the polymer is only the kinetic product and later is transformed to thermodynamically stable cycles. The cationic polymerization of ethylene oxide leads to a mixture of poly(ethylene oxide) and 1,4-dioxane. In the presence of a cationic initiator poly(ethylene oxide) can be almost quantitatively transformed to this cyclic dimer. On the other hand, anionic polymerization is not accompanied by cyclization due to the lower affinity of the alkoxide anion towards linear ethers only strained (and more electrophilic) monomers can react with the anion. 

Freeder, B. G. et al., J. Loss Prev. Process Ind., 1988, 1, 164-168 Accidental contamination of a 90 kg cylinder of ethylene oxide with a little sodium hydroxide solution led to explosive failure of the cylinder over 8 hours later [1], Based on later studies of the kinetics and heat release of the poly condensation reaction, it was estimated that after 8 hours and 1 min, some 12.7% of the oxide had condensed with an increase in temperature from 20 to 100°C. At this point the heat release rate was calculated to be 2.1 MJ/min, and 100 s later the temperature and heat release rate would be 160° and 1.67 MJ/s respectively, with 28% condensation. Complete reaction would have been attained some 16 s later at a temperature of 700°C [2], Precautions designed to prevent explosive polymerisation of ethylene oxide are discussed, including rigid exclusion of acids covalent halides, such as aluminium chloride, iron(III) chloride, tin(IV) chloride basic materials like alkali hydroxides, ammonia, amines, metallic potassium and catalytically active solids such as aluminium oxide, iron oxide, or rust [1] A comparative study of the runaway exothermic polymerisation of ethylene oxide and of propylene oxide by 10 wt% of solutions of sodium hydroxide of various concentrations has been done using ARC. Results below show onset temperatures/corrected adiabatic exotherm/maximum pressure attained and heat of polymerisation for the least (0.125 M) and most (1 M) concentrated alkali solutions used as catalysts. 

It is tempting to associate directly the absence of ethylene oxide over catalysts with more than 40% Pd with the appearance of holes in the d-band. It could be assumed that ethylene is chemisorbed directly on Pd-rich alloys and rapidly decomposed, whereas on Ag-rich alloys ethylene is only adsorbed on top of an oxygen-covered surface leading to selective oxidation. However, the general conclusion from earlier kinetic studies (143) is that the rate-determining step over pure palladium also involves the latter mode of ethylene chemisorption. 

Further evidence for the Aa11 mechanism was obtained from a solvent kinetic isotope study. The theoretical kinetic isotope effects for intermediates in the three reaction pathways as derived from fractionation factors are indicated in parentheses in Scheme 6.143,144 For the Aa11 mechanism (pathway (iii)) a solvent KIE (/ch2o A d2o) between 0.48 and 0.33 is predicted while both bimolecular processes (pathways (i) and (ii)) would have greater values of between 0.48 and 0.69. Acid-catalysed hydrolysis of ethylene oxide derivatives and acetals, which follow an A1 mechanism, display KIEs in the region of 0.5 or less while normal acid-catalysed ester hydrolyses (AAc2 mechanism) have values between 0.6 and 0.7.145,146... 

Figure 17.17 Transient absorbance decay kinetics of the oxidized Ru(dcbpy)2(NCS)2 at 620 nm in the presence and in the absence of 0.1 M Co(II) in acetonitrile/ethylene carbonate 40 60 v/v. Reprinted with permission from Ref. 55. Copyright 2001 American Chemical Society.

It is clear from the kinetics that both ethylene and oxygen adsorption are important since both compounds appear in the rate equations with non-zero orders. Moreover, it is well known that ethylene is not adsorbed on pure silver, but that it does adsorb on a surface that is partially covered with oxygen. This implies that ethylene is either adsorbed on top of pre-adsorbed oxygen or on silver sites that are activated by the presence of oxygen (i.e. by formation of surface oxides, or another form of electron transfer or polarization). Consequently, two different mechanisms arise for the formation of ethylene oxide. The (direct) combustion of ethylene is another point of discussion. Although many favour the idea that different oxygen species are involved, others assume the same oxygen species, but different forms of ethylene adsorption. 

Eastham and Derwent474 have also studied the kinetics of the perchloric acid-catalyzed reaction of ethylene oxide with pyridine. In excess of pyridine the rate was found to be dependent on the Conor Titrations of ethylene oxide and perchloric add. Addition of stronger bases,. g. ammonia, triethylamine, or benzylamiae, depressed the vum of cleavage, presumably by competing with ethylene oxide for thr-available proton source, believed to be pyridinium perchlorate in this case. Other acids examined included nitric acid and hydroiodie irireaction rate depended to a certain extent... 

Addition of an acid catalyst allows hydration of ethylene oxide to be accomplished under much milder conditions and at a markedly faster rate.21 f ° Numerous kinetic studies have been conducted for the reactions of ethylene oxide and several simple homologs with water, in the presence as well as the absence of acid catalysts. 4 - -1W7-13C1-Ml , U17. HU. UTS, son. ZOltt... 

The kinetics of the reaction between ethylene oxide and various alcohols has been a subject of some interest as an application of the theory of consecutive reactions. 8 1,M 12l7>1SW With ethanol, for example, a sequence of steps may be written in the manner depicted in Rq. (650), and a complex kinetic expression derived for the rate of ethylene oxide disappearance. 

Two years later Thompson and Hinshelwood (40) after studying the kinetics of the oxidation of ethylene in silica vessels at temperatures between 400° and 500° and finding that the rate is affected by the total pressure approximately in a reaction of the third order, the effects depending very much more on the partial pressure of the ethvlenes than on that of oxygen, suggested as a via media that while there is no doubt that Bone s interpretation of the course of the oxidation as a process of successive hydroxylations is essentially correct... the first stage in the reaction is the formation of an unstable peroxide if this reacts with more oxygen the chain ends but if it reacts with ethylene, unstable hydroxylated molecules are formed which continues the chain. It should be noted, however, that they adduced no experimental proof of the actual initial formation of the assumed peroxide. 

In view of the evidence cited, it is clear that reactions of formaldehyde make a substantial contribution to the kinetics of the high-temperature oxidation of ethylene. However, attention should be drawn to one further observation of Harding and Norrish 24). The maximum pressure of formaldehyde developed at 400° C. approximates 6 mm., whereas the amount just necessary to eliminate the induction period at this temperature is close to 18 mm. This result suggests that another intermediate may play a part without affecting the kinetics of the reaction. In this connection it was observed that ethylene oxide was formed and built up to a small steady-state concentration during the induction period. The relative ineffectiveness of prior addition of ethylene oxide in reducing the induction period is consistent with the conclusion that formaldehyde plays the dominant role. 

The kinetics of the oxidation of styrene and 3,3,3-trimethylpropene are presently under investigation to see if they follow the kinetics reported by the earlier workers for ethylene. These studies are quite complicated because they must take into account the 7r-complex equilibria discussed previously and which the earlier workers did not consider. Preliminary results indicate that the initial increase in rate with increasing [NaOAc] found for exchange (Figure 1) is also operative for oxidation. However the shape of the rate [NaOAc] profile at higher [NaOAc] is uncertain. 

In several experiments, in particular the study by Temkin and co-workers [224] of the kinetics in ethylene oxidation, slow relaxations, i.e. the extremely slow achievement of a steady-state reaction rate, were found. As a rule, the existence of such slow relaxations is ascribed to some "side reasons rather than to the purely kinetic ("proper ) factors. The terms "proper and "side were first introduced by Temkin [225], As usual, we classify as slow "side processes variations in the chemical or phase composition of the surface under the effect of reaction media, catalyst deactivation, substance diffusion into its bulk, etc. These processes are usually considered to require significantly longer times to achieve a steady state compared with those characterizing the performance of chemical reactions. The above numerical experiment, however, shows that, when the system parameters attain their bifurcation values, the time to achieve a steady state, tr, sharply increases. 

---