Towards predictable tuning of spin crossover
Spin crossover (SCO) active metal complexes are highly versatile materials thanks to their sensitivity to tiny physical or chemical environmental changes. This property makes them very useful for a wide range of applications: employable for experimental studies as molecular switches or for theoretical studies investigating the M-L bonds. In both cases, these studies aim to develop strategies of predictably tuning them. Chapter One. An introduction to the SCO phenomenon: from gradual to cooperative SCO; various methods of monitoring the SCO transition. A summary of some literature examples of SCO-active systems is given. An overview of the published achievements in predicting the SCO phenomenon, including an introduction to the computational models deployed across the years. The EDA-NOCV model, employed in this field for the first time in this PhD thesis, is introduced. Finally, the aims of this study are presented. Chapter Two. The synthesis and characterisation of four new non-symmetrical ligands, 3-(2-(5-Z-pyridyl))-4-tolyl-5-phenyl-1,2,4-triazole (LpytZ, Z = CF3, Br, F, Me), and the corresponding [FeII(LpytZ)2(NCBH3)2] complexes are presented. All four of these new complexes are SCO-active in the solid state and in CDCl3 solution, but T1/2 tuning by the meta-Z substituents was very modest. Three literature families were also tested, successfully extending the generality of using the 15N NMR chemical shift δNA of the coordinated nitrogen atom of the free ligand as measure of the T1/2 in the resulting Fe(II) complex. Chapter Three. Theoretical study of a family of five iron(II) SCO-active [Fe(Lazine)2(NCBH3)2] (Lazine = 3-(2-azinyl)-4-tolyl-5-phenyl-1,2,4-triazole) and of the related five LS [Fe(Lazine)3(BF4)2]. The EDA-NOCV model was applied to molecular fragments to provide quantitative assessment of the σ- and π-bonding. A new corrected [Mn+ + L6] fragmentation was implemented which promises to enable a general approach suitable for any ML6 system. Finally, the σ- and π-bonding character is strongly correlated with the experimental T1/2 of the SCO-active [Fe(Lazine)2(NCBH3)2] complexes. Chapter Four. Theoretical study of the M-L bond in a family of sixteen SCO-active differently para-X substituted [Fe(bppX)2]2+ complexes (bppX is 2,6-di(pyrazol-1-yl)-4-X-pyridine). Results of the EDA-NOCV revealed the σ-strength of the bppX ligand is correlated with σp+(X), δNA(bppX), experimental T1/2([Fe(bppX)2]2+) and calculated AILFT ΔO([Fe(bppX)2]2+). Results are explained at the molecular level by investigating the orbital population of the valence orbitals of the coordinating nitrogen involved in the aromatic π-system (pπ) and in the Fe-N bond (sp2(Fe)). From the observed correlations, the unknown σp+ parameter for two substituents (X = SOMe, SO2Me) is predicted. Chapter Five. First theoretical study on [CoII(dpzca)2] SCO in the solid state, aiming to establish a computational protocol able to predict experimental T1/2 in pressure-activated SCO. The first part of the study validated a DFT protocol at p = 1 bar. The protocol was then extended and trialled up to 4300 bar. Results revealed good reproduction of the experimental results up to 2100 bar; but beyond this pressure, the theoretical and experimental findings diverge. Theoretical data suggest a possible phase change for the crystalline structure of HS [CoII(dpzca)2] at higher pressures than 2100 bar; this would explain why the implemented computational protocol lost validity.
Advisor: Brooker, Sally; Garden, Anna; Totti, Federico
Degree Name: Doctor of Philosophy
Degree Discipline: Chemistry
Publisher: University of Otago
Keywords: SCO; spincrossover; correlation; N-NMR; M-L; substituent; prediction; DFT+U
Research Type: Thesis