Solvent Polarity Predictably Tunes Spin Crossover T1/2 in Isomeric Iron(II) Pyrimidine Triazoles

Abstract

Two isomeric pyrimidine-based Rdpt-type triazole ligands were made: 4-(4-methylphenyl)-3-(2-pyrimidyl)-5-phenyl-4H-1,2,4-triazole (L2pyrimidine) and 4-(4-methylphenyl)-3-(4-pyrimidyl)-5-phenyl-4H-1,2,4-triazole (L4pyrimidine). When reacted with [FeII(pyridine)4(NCE)2], where E = S, Se, or BH3, two families of mononuclear iron(II) complexes are obtained, including six solvatomorphs, giving a total of 12 compounds: [FeII(L2pyrimidine)2(NCS)2] (1), [FeII(L2pyrimidine)2(NCSe)2] (2), 2·1.5H2O, [FeII(L2pyrimidine)2(NCBH3)2]·2CHCl3 (3·2CHCl3), 3 and 3·2H2O, [FeII(L4pyrimidine)2(NCS)2] (4), 4·H2O, [FeII(L4pyrimidine)2(NCSe)2] (5), 5·2CH3OH, 5·1.5H2O, and [FeII(L4pyrimidine)2(NCBH3)2]·2.5H2O (6·2.5H2O). Single-crystal X-ray diffraction reveals that the N6-coordinated iron(II) centers in 1, 2, 3·2CHCl3, 4, 5, and 5·2CH3OH have two bidentate triazole ligands equatorially bound and two axial NCE co-ligands trans-coordinated. All structures are high spin (HS) at 100 K, except 3·2CHCl3, which is low spin (LS). Solid-state magnetic measurements show that only 3·2CHCl3 (T1/2 above 400 K) and 5·1.5H2O (T1/2 = 110 K) undergo spin crossover (SCO); the others remain HS at 300-50 K. When 3·2CHCl3 is heated at 400 K it desorbs CHCl3 becoming 3, which remains HS at 400-50 K. UV-Vis studies in CH2Cl2, CHCl3, (CH3)2CO, CH3CN, and CH3NO2 solutions for the BH3 analogues 3 and 6 led to a 6:1 ratio of Lnpyrimidine/Fe(II) being employed for the solution studies. These revealed SCO activity in all five solvents, with T1/2 values for the 2-pyrimidine complex (247-396 K) that were consistently higher than for the 4-pyrimidine complex (216-367 K), regardless of solvent choice, consistent with the 2-pyrimidine ring providing a stronger ligand field than the 4-pyrimidine ring. Strong correlations of solvent polarity index with the T1/2 values in those solvents are observed for each complex, enabling predictable T1/2 tuning by up to 150 K. While this correlation is tantalizing, here it may also be reflecting solvent-dependent speciation - so future tests of this concept should employ more stable complexes. Differences between solid-state (ligand field; crystal packing; solvent content) and solution (ligand field; solvation; speciation) effects on SCO are highlighted.