The Triplet Biradical State of a Cyclobutadiene – an Intermediate of 2+2 Retro-Cycloaddition

Arseni Kostenko, Schulich Faculty of Chemistry and the Lise Meitner-Minerva Center for Computational Quantum Chemistry, Technion - Israel Institute of Technology, Haifa, Israel
Boris Tumanskii, Schulich Faculty of Chemistry and the Lise Meitner-Minerva Center for Computational Quantum Chemistry, Technion - Israel Institute of Technology, Haifa, Israel
Akira Sekiguchi, Department of Chemistry, Graduate School of Pure and Applied Sciences, University of Tsukuba, Tsukuba, Japan
Yitzhak Apeloig, Schulich Faculty of Chemistry and the Lise Meitner-Minerva Center for Computational Quantum Chemistry, Technion - Israel Institute of Technology, Haifa, Israel

Ever since Kekulé first attempted the preparation of cyclobutadiene in 1872, this fascinating molecule presented a challenge to experimental and theoretic chemists. It was not until the mid-1970th that cyclobutadiene derivatives could be synthesized and characterized. The anti-aromatic character of cyclobutadiene and its geometric rigidity result in an unusual situation that the D_4h triplet excited state is energetically close to the D_2h singlet ground state. The existence of such triplet biradical intermediate is intriguing in terms of studying cycloaddition reactions - key reactions in organic synthesis.

Herein, we report the first observation by EPR spectroscopy of a triplet biradical state of a cyclobutadiene derivative and demonstrate that the retro-cycloaddition of 1,2,3,4-tetrakis(trimethylsilyl)cyclobuta-1,3-diene (1) to bis(trimethylsilyl)acetylene (observed by NMR) proceeds via a triplet biradical intermediate (2) (Scheme 1). To our best knowledge, this is the first example in which a triplet biradical state of a cyclobutadiene derivative was observed spectroscopically. An EPR study of 1, at 300K–395K, shows three lines, which are typical for triplet systems with a large zero-field splitting (ZFS) (|D|_exp=0.17cm^-1) (Figure 1). From the temperature dependent absorption area we derive a singlet→triplet (1→2) energy gap dE_ST=14.2kcal·mol^-1. This value is supported by quantum chemical calculations of model 1 and 2 (in which the SiMe₃ substituents were replaced by SiH₃) at the CCSD(T)/def2-TZVPP//B3LYP/6-311+G(d,p) level of theory, that give dE_ST=12.4kcal·mol^-1. To estimate the D value of 2,  D values of eleven delocalized organic triplet biradicals (Figure 2) were calculated at the B3LYP/TZVP level giving a good linear correlation between the theoretical and experimental values (R²=0.9392, D_exp=1.5814·D_calc). Using this relation we calculate that |D(2)|_calc=0.165cm^-1, in excellent agreement with |D|_exp=0.17cm^-1. In conclusion, we have provided an experimental and theoretical evidence that the retro-cycloaddition of 1 proceeds via a thermally induced triplet biradical state.