Cumene thermal decomposition

The thermal decomposition of thia2ol-2-yl-carbonyl peroxide in benzene, bromobenzene, or cumene affords thiazole together with good yields of 2-arylthiazoles but negligible amounts of esters. Thiazol-4-ylcarbonyl peroxide gives fair yields of 4-arylthiazoles, but the phenyl ester is also a major product in benzene, indicating reactions of both thiazol-4-yl radicals and thiazol-4-carbonyloxy radicals. Thiazole-5-carbonyl peroxide gives... 

In oxidation studies it has usually been assumed that thermal decomposition of alkyl hydroperoxides leads to the formation of alcohols. However, carbonyl-forming eliminations of hydroperoxides, usually under the influence of base, are well known. Of more interest, nucleophlic rearrangements, generally acid-catalyzed, have been shown to produce a mixture of carbonyl and alcohol products by fission of the molecule (6). For l-butene-3-hydroperoxide it might have been expected that a rearrangement (Reaction 1) similar to that which occurs with cumene hydroperoxide could produce two molecules of acetaldehyde. 

Singer and Chen demonstrated the inability of an a-chlorine substituent to stabilize the configuration of a cyclopropyl radical. They showed that the photochemical decomposition of both exo- (20) and endo-t-buty 6-chlorobicyclo[3.1.0]hexane-6-percarboxylate (21) in diisopropylbenzene resulted in an identical mixture of exo- (22) and endo-6-chlorobicyclo[3.1.0]hexane (23). A similar result" was obtained in the thermal decomposition of both exo- (24) and endo-t-buty 7-chlorobicyclo[4.1.0]heptane-7-percarboxylate (25). In solvents such as toluene, cumene or bromotrichloromethane the same ratio (20 80) of exo-26 and endo-21 products was formed within experimental error. 

Surprisingly, the Hunsdiecker reaction using the silver salts of exo- and endo-7-chlorobicyclo[4.1.0]heptanecarboxylic acids and bromine at 0°C did not result in the same ratio of products but instead showed a high retention to inversion ratio of 88 12 for the exo acid and 88 12 for the endo acid". This anomalous result may be a reflection of the bromine radical s ability to trap the cyclopropyl radical but this is unlikely. Altman and Baldwin as well as Ando and coworkers found that the reduction of each of the isomers of 7-bromo-7-chlorobicyclo[4.1.0]heptane, 30 and 31, respectively, by the excellent radical scavenger triphenyltin hydride resulted in an identical mixture (21 79) of exo-(32) and en io-7-chlorobicyclo[4.1.0]heptane (33). This ratio of products is, within experimental error, identical with that found in the thermal decomposition of exo- and endo-t-buiy 7-chlorobicyclo[4.1.0]heptane-7-percarboxylate in cumene. 

In the derivation of the kinetic relations it was assumed that free radicals enter the particles one by one the initiation process just described satisfies this condition. This is not the case when radicals are formed by thermal decomposition of an oil-soluble initiator. Such decomposition produces pairs of radicals in the hydrocarbon phase. One would expect a pair of radicals, confined to the extremely small volume of a latex particle, to recombine rapidly. The kinetics of this type of polymerization have been described above. It is recalled here that the subdivision factor, z, and hence rate and degree of polymerization are smaller than 1 and decrease with a. These predictions from kinetic theory are in contradiction to experimental observations. Although some oil-soluble initiators, which are good catalysts in solution systems, are poor initiators in emulsion polymerizations—e.g., benzoyl peroxide—other thermally decomposing peroxides and azo compounds produce polymer in emulsion at rates comparable to those observed in polymerization initiated by water-soluble catalysts, where the radicals enter the particles one by one. Such is the case for cumene hydroperoxide, which at low concentrations yields a rate of polymerization per particle equal to that of a persulfate-initiated reaction. It must therefore be concluded that, although oil-soluble initiators may decompose into radical pairs within the particles, polymer radicals are formed one by one. The following mechanisms are consistent with formation of polymer radicals singly. 

When cumene hydroperoxide is heated to cumene cracking temperatures, at least partial decomposition probably occurs. Thus, it is desirable to measure the inhibitor action of the decomposition products of cumene hydroperoxide. Some low-temperature thermal decomposition products which were identified (by chemical and mass spectroscopic analyses) are acetophenone, phenyl-dimethyl-carbinol, a-methylstyrene, phenol, acetone, and methyl alcohol. Kharasch, Fono, and Nudenberg (7) obtained similar results. According to them, the chief decomposition products at 158° are acetophenone and phenyl-dimethyl-carbinol, with acetophenone becoming relatively more important at higher temperatures. Since the temperature used for the cracking reaction is above 300°, acetophenone is probably the most important decomposition product. The equilibrium constant for the adsorption of cumene hydroperoxide and some of the individual decomposition products on catalytic sites are included in Table I. 

An overview of the cumene route will be given first, followed by a detailed description of the thermal decomposition behavior of CHP, and finally the mechanism of the liquid phase oxidation of cumene. The oxidation of cumene to CHP is one of three major reaction steps within the cumene route. 

When storing and handling CHP at medium or even elevated temperatures, the heat release from the thermal decomposition must be efficiently removed in order to avoid any hazards from thermal explosion (runaway). Especially, in large reactors for cumene oxidation, the exothermic decomposition of CHP has to be taken into account during a shutdown process when there is no more mixing by aeration, so only limited heat removal to ambient takes place. The heat evolved from thermal decomposition is 270 kj/mol [8,9]. From process safety point of view, and also to understand the auto-catalyzed mechanism in the cumene oxidation, it is necessary to describe and quantify the thermal decomposition characteristics of CHP. 

Twigg [10] was the first to present a kinetic study for the thermal decomposition of CHP in cumene. Experiments were performed at temperatures between 110 and 160 °C. Pure CHP (98-99.5%) was diluted in purified cumene down to concentrations below 10 wt%. 

Hattori et cd. [11] investigated the thermal decomposition in the following way Cumene was oxidized in isothermal batch experiments to a certain CHP concentration between approximately 25 and 55 wt%. The temperature was varied between 110 and 130 C. Then, the oxygen supply was stopped and the CHP decomposition was monitored over time. For the above-mentioned temperature and concentration range, a first-order kinetics was determined ... 

Duh etal. [12,13] used an 80 wt% CHP, which was diluted with cumene to lower concentrations down to minimum 15 wt% for various tests on the thermal decomposition behavior. In Taiwan, CHP is widely used as an initiator in polymerization. As described in [14], cumene is oxidized to 80wt% CHP for this purpose. It is therefore assumed that an 80 wt% CHP was sampled from such a production process. The reaction order was determined to be 0.5. 

As described above, DMBA is the major by-product from a moderate thermal decomposition of CHP in the oxidation process, due to the presence of (excess) cumene. It is obvious that the radical mechanism must change tov rards higher CHP concentrations as the cumene concentration largely decreases. Thus, it seems implausible to describe the thermal decomposition of CHP v rith only one kinetic model for the whole concentration range between 0 and 100 wt%. 

In commercial cumene oxidation processes, the radicals from the thermal decomposition of CHP are the initiators for the free-radical chain reaction. Therefore, the reaction towards CHP in the cumene oxidation is called autooxidation. 

Besides the formation of the major by-products DMBA and ACP and, in minor concentrations DCP, a lot of micro-impurities are formed in cumene oxidation. A key role is played by the methyl radical CHg from the thermal decomposition of CHP. In the presence of oxygen, methyl hydroperoxide (MHP) and formaldehyde are formed (Figure 2.2). Formaldehyde is further oxidized to formic acid, which can catalyze the acidic decomposition of CHP into phenol and acetone. As... 

The second mechanism worthy of consideration involves initial abstraction of one hydrogen by triplet nitrene to form an anilino radical which may subsequently either dimerize or abstract another hydrogen. This is illustrated for the thermal decomposition of p-anisyl azide in cumene. Indirect support for the hydrazo intermediate 37 comes from the observation that thermolysis of phenyl azide in decalin affords hydrazobenzene and benzidine as well as azobenzene and aniline. Anilino radical intermediates also are implicated in the oxidation... 

The kinetics of thermal decomposition of nitrogen trichloride have been monitored in solutions in carbon tetrachloride containing electron-donor co-solvents such as benzene, toluene, cumene, and mesitylene. ... 

DeSimone and coworkers (340) investigated the dimerization of a-methyl-styrene using DuPont Nation catalysts. They observed a rate enhancement over conventional liquid solvents such as cumene and o-cresol, which they attributed in part to plasticization of the perfluorinated catalyst resin with the SCCO2 combined with the enhanced mass transfer characteristics afforded by the SCF solvent. In a subsequent study, DeSimone and coworkers (341) measured the thermal decomposition rates of two perfluoroalkyl diacyl peroxides [bis(trifluoro-acetyl) and [bis(perfluoro-2-n-propoxyprionyl) peroxides] in liquid and SCCO2 and compared rates with similar measurements made in Freon-113. Both peroxides displayed activation energies approximately 5-6 kcal/mol lower than that obtained in Freon-113, which the authors attribute to differences in solvent viscosity. 

Ed, the activation energy for thermal initiator decomposition, is in the range 120-150 kJ mol-1 for most of the commonly used initiators (Table 3-13). The Ep and Et values for most monomers are in the ranges 20-40 and 8-20 kJ mol-1, respectively (Tables 3-11 and 3-12). The overall activation energy Er for most polymerizations initiated by thermal initiator decomposition is about 80-90 kJ mol-1. This corresponds to a two- or threefold rate increase for a 10°C temperature increase. The situation is different for other modes of initiation. Thus redox initiation (e.g., Fe2+ with thiosulfate or cumene hydroperoxide) has been discussed as taking place at lower temperatures compared to the thermal polymerizations. One indication of the difference between the two different initiation modes is the differences in activation energies. Redox initiation will have an Ed value of only about 40-60 kJ mol-1, which is about 80 kJ mol-1 less than for the thermal initiator decomposition [Barb et al., 1951], This leads to an Er for redox polymerization of about 40 kJ mol-1, which is about one half the value for nonredox initiators. 

The use of chemical sensitizers such as benzoyl peroxide, cumene hydroperoxide, or azo-bis-isobutyronitrile, which decompose thermally to give free radicals in a convenient temperature range (i.e., 60 C to 150 C), makes it possible to study polymerizations over an extended temperature range. The form of the rate law with chemical initiations would be given by setting III = 2k (ln)< >i in Eq. (XVI.10.4). Here (In) is the initiator concentration, k I its specific rate constant of decomposition which can usually be measured independently, and is the efficiency with which its radicals initiate chains. The measure of t is subject to the difficulties already indicated in connection with the photolysis systems. ... 

---