Atropisomerization in N-aryl-2(1H)-pyrimidin-(thi)ones: a ring-opening/rotation/ring-closure process in place of a classical rotation around the pivot bond.

Uncatalyzed racemization processes in atropisomeric diphenyl-like frameworks are classically described as the result of the rotation around the pivotal single bond linking two planar frameworks. Severe constraints leading to more or less distorted transition states account for the experimental barrier to atropenantiomerization. In 1988, one of us hypothesized that, in N-aryl-2(1H)-pyrimidin-(thi)ones, a ring-opening/ring-closure process was contributing to the observed racemization process accounting for the lower barriers in the sulfur analogues than in oxygen analogues. Now, a series of six novel 6-amino-5-cyano-1,4-disubstituted-2(1H)-pyrimidinones 5a-5f and two 6-amino-5-cyano-4-p-tolyl-1-substituted-2(1H)-pyrimidinethiones 6a and 6b were synthesized and characterized through spectroscopic and X-ray diffraction studies. Semipreparative HPLC chiral separation was achieved, and enantiomerization barriers were obtained by thermal racemization. The rotational barriers of 6-amino-5-cyano-1-o-tolyl-4-p-tolyl-2(1H)-pyrimidinone (5b) and 6-amino-5-cyano-1-(naphthalen-1-yl)-4-p-tolyl-2(1H)-pyrimidinone (5e) were found to be 120.4 and 125.1 kJ·mol(-1) (n-BuOH, 117 °C), respectively, and those of the corresponding thiones were 116.8 and 109.6 kJ·mol(-1) (EtOH, 78 °C), respectively. DFT calculations of the rotational barriers clearly ruled out the classical rotation around the pivotal bond with distorted transition states in the case of the sulfur derivatives. Instead, the ranking of the experimental barriers (sulfur versus oxygen, and o-tolyl versus 1-naphthyl in both series) was nicely reproduced by calculations when the rotation occurred via a ring-opened form in N-aryl-2(1H)-pyrimidinethiones.