Te@&e&m Lcum. Vol. 34. No. 20. Pp. 3219-3222.1993 Priited in Gnat Britain 0040-4039,93 56.00+ .oO Pcrgamon Press Lid Thermal Cleavage of the 3,4 Bond in a Cyclobutane-l&dione Thomas I-I. Peterson and Barry K. Carpenter* Dqmlment of Clmis~ ry, Baker Laboramy. Cornell University, IIhaca, NY 14853-1301.USA Abstrad: Pyrolysis of rrans-3,~chyl-3,4dipbenylcycl~~-l.2-~~ nxults in an isomerization interprti to involve fomation of a dioxatemethyleo~ thane biradicaL ‘his bimdical would arise from a hitbextounpmc&nted 3.4 bond cleavage. The biradical tetramethyleneethane. in which there has been much recent experimental1 and theoretical2 interest, can be generated revefiibly by thermal cleavage of the weakest bond -- the C3-C4 bond -- of 1.2- bismethylenecyclobutane.3 Generation of the dioxa- analog of tetramethyleneethane by the corresponding C3- C4 homolysis of cyclobutane-1,Zdione is not possible, however (at least under normal thermal reaction conditions), because this bond is the strongest in the ring, according to Benson group additivity calculations4 (vide infiu). The weakest is the Cl-C2 bond. One could imagine changing the relative bond strengths of cyclobutane-1,Zdione by substitution, and in fact, evidence has been found for both Cl-C2 and Cl-C4 homolysis in certain derivatives,5 but, to our knowledge, there has hitherto been no report of C3-C4 cleavage. We now describe such a diction, occurring in wa-3,4-dimethyl-3,4-diphenylcyclobutanedione (1). Its synthesis is summarized in Scheme 1. CO&H, A a_ Ph CH, 80% Scheme 1. Synthesis of trans-3,4-dimethyl-3,4-diphenylcyclobutane-1.2-dione. Reagents are as follows: (a) (i) NBSICC4; (ii) ZnKuC12, cat. HgBr2. 6 (b) (i) LiAlQ; (ii) ClCOCOCYDMSO, Et3N; (iii) TiCls(DME)z, Z&u. (c) ClCOCOCl/DMSO. Et3N. 3219