The Triplet–Singlet Gap in the m‑Xylylene Radical: A Not So Simple One

• Published in: Journal of chemical theory and computation 2014-01, Vol.10 (1), p.335-345
• Main Authors: Reta Mañeru, Daniel, Pal, Arun K, Moreira, Ibério de P. R, Datta, Sambhu N, Illas, Francesc
• Format: Article
• Language: English
• ISSN: 1549-9618; 1549-9626
• DOI: 10.1021/ct400883m

Meta-benzoquinodimethane (MBQDM) or m-xylylene provides a model for larger organic diradicals, the triplet–singlet gap being the key property. In the present work this energy difference has been the object of a systematic study by means of several density functional theory-based methods including B3LYP, M06, M06-2X, HSE and LC-ωPBE potentials and a variety of wave function-based methods such as complete active space self consistent field (CASSCF), Multireference second-order Møller–Plesset (MRMP), difference dedicated configuration interaction (DDCI), and Multireference configuration interaction (MRCI). In each case various basis sets of increasing quality have been explored, and the effect of the molecular geometry is also analyzed. The use of the triplet and broken symmetry (BS) solutions for the corresponding optimized geometries obtained from B3LYP and especially M06-2X functionals provide the value of the adiabatic triplet–singlet gap closer to experiment when compared to the reported value of Wenthold, Kim, and Lineberger, (J. Am. Chem. Soc. 1997, 119, 1354) and also for the electron affinity. The agreement further improves using the full π-valence CASSCF(8,8) optimized geometry as an attempt to correct for the spin contamination effects on the geometry of the BS state. The CASSCF, MRMP, and MRCI, even with the full π valence CAS(8,8) as reference and relatively large basis set, systematically overestimate the experimental value indicating either that an accurate description must go beyond this level of theory, including σ electrons and higher order polarization functions, or perhaps that the measured value is affected by the experimental conditions.