Initiators that form diradicals are of interest due to their potential for increasing polymerization rate. Spiroperketals made from 2,5-dihydroperoxy-2,5-dimethylhexane and cyclic ketones theoretically can form two diradical fragments during decomposition. However, thermolyses in ethylbenzene and styrene showed that they mainly undergo in-cage decomposition, resulting in poor performance as both hydrogen atom abstractors and polymerization initiators.Recently, we reported the decomposition chemistry of 1,l-bis(tert-buty1peroxy)cyclohexane (11.l The decomposition chemistry of I involves three pathways (Scheme I). The use of difunctional peroxides as initiators for styrene polymerization leads to the formation of high molecular weight polystyrene a t faster rates than achievable using monoperoxides. This rate increase is generally viewed to arise from diradical fragments (11). However, for every diradical fragment formed, there are two tertbutylperoxy monoradicals also formed, which attenuates the polymerization rate actually gained by the use of typical difunctional initiators. A class of peroxides that should theoretically form only diradicals are cyclic peroxides. However, anomalous results have been reported. For example, cyclic peroxides 111-V decompose without initiating polymerization of styrene.2 U I11 IV VThe earliest work3 aimed at forming diradicals for styrene polymerization was performed in an effort to test the Flory diradical mechanism (Scheme IIIs4 However, earlyon, there was controversy over the ability of diradical initiation to produce higher molecular weight polymer^.^*^ Later w~r k ,~,~ however, showed successes in producing higher molecular weight vinyl polymers using biradical initiators. For example, when comparing the monoradical initiator VI and the diradical initiator VII, Borsig et found (polymerization of methyl methacrylate at 60 "C) that significantly higher molecular weight was formed at the same monomer conversion using the diradical initiator. However, these data are complicated by the fact that at the same molar initiator concentration, the polymerization rate was slower for VII. The difference in initiating efficiency of the two initiators was explained in terms of cage reactions. The monoradicals formed from initiator VI diffuse quite efficiently from the cage and react with monomer. The diradicals, on the other hand, are in a permanent cage in that they cannot diffuse away from each other. Therefore, significantly disproportionation of the diradical (VIII) takes place to compete with initiation.