Bulk optical properties and magnetic properties of monolayers of Er(III) cyclooctatetraenide-based single-molecule magnets

Fundamental investigation in the field of molecular complexes displaying magnetic hysteresis opening, called single-molecule magnets (SMMs), has been very active in the past few decades for the outstanding magnetic properties and potential technological applications displayed by these compounds. Sandwich lanthanide organometallic complexes have recently been reported displaying magnetic bistability up to 80 K, which is very interesting for potential applications in technology featuring spintronics systems and data storage devices. However, the study of these systems deposited on metal substrates has shown that the hysteresis opening tends to be lost due to the interaction of the complexes with the surface in the monolayer regime. SMMs with planar ligands like cyclooctatetraenide (COT$^{2-}$) dianion and pentamethyl-cyclopentadienide (Cp*) anion are good candidates to form ordered layers on surfaces, although the self-assembly and the complex-substrate interaction can play a fundamental role in the properties of those systems. Moreover, complexes containing the trivalent Er ion have potential use in spintronic technology through the optical manipulation of the electronic states in the 4f shell. The details of the ligand field effect on the intra-4f transitions in the near-infrared range often remain elusive, relying on simulations or indirect techniques.
Here a study of the molecular self-assembly and the magnetic properties of two sandwich Er$^{3+}$ COT$^{2-}$-based SMMs deposited on the (100) surface of an Ag crystal in the monolayer range and a comparison of the optical properties of four structurally similar sandwich Er$^{3+}$ SMMs in the infrared region is presented. Combining low-temperature scanning-tunneling microscopy, X-ray photoemission spectroscopy and polarized X-ray absorption spectroscopy it is shown that the heteroleptic Cp*ErCOT complex self-assembles in a mixed standing-up and lying-down configuration forming alternating compact parallel rows, oriented parallel to the crystallographic [010] and [001] directions of the surface. On the other side, the homoleptic K[Er(COT)$_{2}$] complex, produced from the sublimation of [K(18-c-6)][Er(COT)$_{2}$]∙2THF precursor (18-c-6 = 18-crown-6 ether, THF = tetrahydrofuran), forms ordered domains with the COT rings standing perpendicular to the substrate and the complexes aligned along specific directions.
The polarization-dependent X-ray absorption spectra are compared to the simulations performed by multiX, an ab-initio model based on point charges emulating the electrostatic effect acting onto the central Er ion. The simulations suggest that the strong in-plane magnetic anisotropy of K[Er(COT)$_{2}$]/Ag(100) and the mixed in-plane and out-of-plane magnetization of Cp*ErCOT/Ag(100) can be attributed to the strikingly different surface ordering of these two complexes. With comparison to the magnetic properties of the bulk precursors, the surface-supported K[Er(COT)$_{2}$] shows a similar large hysteresis opening, while Cp*ErCOT shows a rather small opening. The different net magnetic properties on the metal substrate are attributed to the different orientations/interactions of the ligand rings to the surface.
At the same time, Fourier-transform infrared spectroscopy in the range of the transitions between the Er$^{3+}$ ground state ($^{4}I_{15/2}$) and the first excited state ($^{4}I_{13/2}$) manifolds unveils the ligand-field induced differences of four Er$^{3+}$-based SMMs. While [K(18-c-6)][Er(COT)$_{2}$]∙2THF displays few intra-4f transitions attributed to the low admixing of different m$_{J}$ states, [Li(DME)$_{3}$][Er(COT”)$_{2}$] (COT” = bis(trimethylsilyl)-cyclooctatetraenide; DME = dimethoxyethane), Cp*ErCOT and Cp$^{ttt}$ErCOT (Cp$^{ttt}$ = tris(tert-butyl)cyclopentadienide) complexes show multiple transitions due to a stronger admixing of states. The differences are attributed to the details of the ligand field induced by the different aromatic rings in the different complexes.
Moreover, the splitting of 4f-4f transition peaks at low temperatures is attributed to the presence of static disorder of the complexes in the form of different stacking configurations (rotamers) of the ligands. Temperature-dependent spectra reveal thermally activated vibrations of the ligand rings, with activation energies in the range of those found for FeCp$_{2}$ and CpTiCOT. The results show how the temperature affects the dynamics of the complexes featuring COT and/or Cp* ligands and how this influences the intra-4f transitions.
In conclusion, the study of how the substitution of one or both ligand rings of structurally similar Er$^{3+}$-based sandwich SMMs can drastically affect the magnetic properties in the monolayer range on Ag(100) and the optical properties of those compounds in bulk is reported. Specifically, the research sheds light on the behavior and suggests the key parameters affecting the SMM properties on metal surfaces and the intra-4f transitions in the NIR window of Er$^{3+}$-based organometallic sandwich complexes, giving a novel insight into these technologically promising complexes and lying the base to improve the design and understanding of similar systems.