Towards molecular alloys: computational and experimental studies on ( p -NCC 6 F 4 CNSeSeN) x ( p -NCC 6 F 4 CNSSN) 1- x .

The β-phase of the radical p -NCC 6 F 4 CNSSN (1β) crystallizes in the orthorhombic space group F d d 2 and orders as a canted antiferromagnet with T N = 36 K. Computational studies (B3LYP or M06-2X functional with the cc-pVTZ-PP(-F)+basis set) of the microscopic nearest-neighbour magnetic exchange coupling in 1β and in the hypothetical isomorphous phase of the selenium radical p -NCC 6 F 4 CNSeSeN (2β) revealed that replacement of S by Se should lead to a significant enhancement in the magnetic ordering temperature by ca. 20% (B3LYP) - 30% (M06-2X). Recrystallization of 2 from solution or via vacuum sublimation afforded only the known diamagnetic, dimeric phase, 2α. Computational studies indicated that both the molecular geometry and charge distribution for 1 and 2 are extremely similar and experimental approaches to form alloys of the general form 1 1- x 2 x were explored: attempts to cosublime 1 and 2 in vacuo were unsuccessful, forming only 1β due to the low volatility of 2. Crystallization of pure 1 by solution evaporation was found to afford polymorph 1α (triclinic, P 1̄) selectively, irrespective of the solvent employed (CH 2 Cl 2 , MeCN, PhMe or THF) but 1α transformed to 1β upon subsequent vacuum sublimation. Crystallization of 1 in the presence of 2 (up to 20 mol%) from solution evaporation was examined. At 20 mol% there was clear evidence for formation of both 1α and 2α as distinct crystallographic phases by powder X-ray diffraction (PXRD) but some evidence for doping of 2 into 1α at low concentration (≤15 mol percent) was observed. Attempts to sublime a sample of 1 0.9 2 0.1 led to phase separation with the isolation of needle-shaped crystals of pure 1β characterized by X-ray diffraction.