OCR Text |
Show The results in Figure 3 are for 1200 K and 950 K. In the first case the fuel is rapidly destroyed, while in the second case degradation is very slow. In both cases conversion to the chlonnated compounds are low. However one notes that dioxin yields are also very low. For case A, where ethylene is largely destroyed, this is due to the competition from the oxidation process. As soon as the chlonne abstracts a hydrogen from the ethylene, the vinyl radical is intercepted by the oxygen and is prevented from reacting with the chlonne in the system. Thus the decrease in concentration with increasing chlorination. However, note the decrease in the differences in the concentration as the chlonnation is increased. This is a reflection of the greater resistance to oxidation of the more highly chlorinated compounds. There is considerable structure in case B. The initial spike is a manifestation of the large atomic chlorine that is present. This adds to the ethylene and tnggers a whole chain of reactions. However after the initial perturbation, the steady state behavior is due to the high barrier for the abstraction of a vinylic hydrogen by chlorine. It is interesting that the total chlorine concentrations stay at the 10 p p m level. This is because under the present conditions large quantities of HC1 are present and these lower the concentration of OH. Since under the present conditions chlorine atom reacts very slowly with ethylene it will instead recombine to form chlorine molecule. The present analysis clearly highlights the importance of the reactivity of the radicals that are produced with respect to their reaction with oxygen. It should be emphasized that the yields obtained here are very much dependent on the reaction conditions. Presumably with a richer mixture there will be less oxygen. However chlorine concentrations will also decrease. Thus some compensation of effects can be expected. It is obvious that a whole series of simulations similar to those earned out here would be needed. O n the other hand this also calls into question the accuracy of the rate constants that are being used. On the basis of the present results, one would expect that the most likely chlorinated compounds are those whose precursors are less reactive towards oxygen. Thus the rate expression for vinyl radical reaction with oxygen has been found to be 4xl012exp(124/T) cm3morls"l[21] and can be compared with that for phenyl + 0 2 of 2.6xl0I3exp(-3080/T)+3xl013exp(-4520/T) cm3mor1s"l[22]. At 1250 this leads to a difference in rate constants of factors of 1.7 at 1250 K and 3.5 at 1000 K. One would therefore expect increased yields of chlorinated aromatics in comparison to chlorinated ethylenes. This is also consistent with the presence of carbon tetrachloride in combustion systems containing chlorine. Similarly, since resonance stabilized radicals are generally least reactive towards oxygen, this line of reasoning suggest that chlorinated compounds deriving from such radicals may be present in larger concentrations. SUMMARY AND CONCLUSIONS We have examined the possibility that incomplete mixing may lead to the formation of polychlorinated unsaturates during the combustion of hydrocarbon fuels containing small amounts of chlorine compounds. Although the simulation results did not |