Final Activity Report Summary - DFTPREDICTANDCHARCT (New generation magnetic materials - a synthetic methodology derived from computational predictions)
Research activities were focussed as planned in the original proposal with some appropriate turnings taken in this direction as we accumulated our results. The fellow started his studies on some simple building units (BU). The initial interest was on understanding the coupling between 4f-3d systems, thus he chose to study in detail the Gd(III)-Cu(II) complex.
Despite the fact that these was the simplest example of the 3d-4f type, many issues related to the mechanism of magnetic coupling and the calculation of J values were not clearly established. He addressed the above issues with density functional theory (DFT), yielding several important results particularly on the mechanism of the Gd(III)-Cu(II) coupling. After a comprehensive testing, suitable theoretical tools to compute Js in this type of systems were proposed, such B3LYP with a combination of relativistic all electron or effective core potential basis set.
Additionally, the study suggested unambiguously that the empty Gd(III) 5d orbitals had a prominent role on magnetic coupling. An exponential dependence of J on dihedral angle of CuGdO2 was established, as mentioned in steps one and two of the proposal. The studies were then extended to Gd-3d (3d=Ni(II), Cr(III) , Mn(II) and V(III)) and then to larger BUs, such as trinuclear complexes Cu(II)-Gd(III)-Cu(II), in order to understand the 1,2 (Cu-Gd) and 1,3 (Cu-CU) interactions. Studies on several trinuclear complexes revealed a correlation of 1,3 interaction to Cu-Gd-Cu angle and this eventually suggested some clues regarding the possible involvement of Gd(III) empty 6s orbital in mediating the 1,3 interaction.
Moreover, in order to obtain some support from the experiment, synthetic chemists prepared diamagnetic substituted compounds, as proposed in step five of the proposal, with which the nature of 1,3 interactions was understood and some predictions in this regard were made.
Another interesting motif of the extended systems were the Gd-nitronyl nitroxide radical (NitR) chains, as mentioned in the proposal step four. The initial attempt here was to extrapolate the knowledge gained with GdCu systems to GdNitR BUs. This was important because, unlike the GdCu interaction, the Gd-NitR interaction could either be ferromagnetic or antiferromagnetic. Calculations and magneto-structural studies revealed that the nature of interaction was due to the difference in the Gd-O-N(Nit) angle. The magneto-structural correlations on this system revealed that the empty 6s orbital of Gd(III) was participating in the 1,3 interaction. The calculations predicted that the structural changes due to the large substitutions resulted in the observation of different magnetic properties. To have more support to these predictions, the crystal structure of the Gd(hfac)3NITR (R=methoxy phenyl), synthesised in LAMM according to step five of the proposal, was solved and the magnetic properties of this compound were measured, as suggested in steps six and seven. The preliminary data supported the computed predictions.
The current state-of-the-art in the magnetism area was related to studies of molecules on the surface. As such studies paved forward with several potential applications, such as spintronic devices and quantum computing, we also decided to test our prediction of magnetic properties on surfaces. Since this was a relatively new area, several methodological developments were needed to make good prediction. Therefore we started our studies on simple thiol molecules on the surface to understand the nature of the adsorbate and the structure of the self-assembled monolayers (SAMs) formed. Periodic DFT studies on the simple thiol molecules on the surface revealed that thiol radicals were the most stable species. Computed kinetic energy barrier was also consistent with these predictions. Further studies on nitronyl nitroxide radical functionalised with thiols on Au(111) to understand its structure and magnetic properties revealed that the Au(111) surface was non-innocent, and in fact revealed a strong intermolecular interaction between NitR radicals that was mediated by gold atoms.
Despite the fact that these was the simplest example of the 3d-4f type, many issues related to the mechanism of magnetic coupling and the calculation of J values were not clearly established. He addressed the above issues with density functional theory (DFT), yielding several important results particularly on the mechanism of the Gd(III)-Cu(II) coupling. After a comprehensive testing, suitable theoretical tools to compute Js in this type of systems were proposed, such B3LYP with a combination of relativistic all electron or effective core potential basis set.
Additionally, the study suggested unambiguously that the empty Gd(III) 5d orbitals had a prominent role on magnetic coupling. An exponential dependence of J on dihedral angle of CuGdO2 was established, as mentioned in steps one and two of the proposal. The studies were then extended to Gd-3d (3d=Ni(II), Cr(III) , Mn(II) and V(III)) and then to larger BUs, such as trinuclear complexes Cu(II)-Gd(III)-Cu(II), in order to understand the 1,2 (Cu-Gd) and 1,3 (Cu-CU) interactions. Studies on several trinuclear complexes revealed a correlation of 1,3 interaction to Cu-Gd-Cu angle and this eventually suggested some clues regarding the possible involvement of Gd(III) empty 6s orbital in mediating the 1,3 interaction.
Moreover, in order to obtain some support from the experiment, synthetic chemists prepared diamagnetic substituted compounds, as proposed in step five of the proposal, with which the nature of 1,3 interactions was understood and some predictions in this regard were made.
Another interesting motif of the extended systems were the Gd-nitronyl nitroxide radical (NitR) chains, as mentioned in the proposal step four. The initial attempt here was to extrapolate the knowledge gained with GdCu systems to GdNitR BUs. This was important because, unlike the GdCu interaction, the Gd-NitR interaction could either be ferromagnetic or antiferromagnetic. Calculations and magneto-structural studies revealed that the nature of interaction was due to the difference in the Gd-O-N(Nit) angle. The magneto-structural correlations on this system revealed that the empty 6s orbital of Gd(III) was participating in the 1,3 interaction. The calculations predicted that the structural changes due to the large substitutions resulted in the observation of different magnetic properties. To have more support to these predictions, the crystal structure of the Gd(hfac)3NITR (R=methoxy phenyl), synthesised in LAMM according to step five of the proposal, was solved and the magnetic properties of this compound were measured, as suggested in steps six and seven. The preliminary data supported the computed predictions.
The current state-of-the-art in the magnetism area was related to studies of molecules on the surface. As such studies paved forward with several potential applications, such as spintronic devices and quantum computing, we also decided to test our prediction of magnetic properties on surfaces. Since this was a relatively new area, several methodological developments were needed to make good prediction. Therefore we started our studies on simple thiol molecules on the surface to understand the nature of the adsorbate and the structure of the self-assembled monolayers (SAMs) formed. Periodic DFT studies on the simple thiol molecules on the surface revealed that thiol radicals were the most stable species. Computed kinetic energy barrier was also consistent with these predictions. Further studies on nitronyl nitroxide radical functionalised with thiols on Au(111) to understand its structure and magnetic properties revealed that the Au(111) surface was non-innocent, and in fact revealed a strong intermolecular interaction between NitR radicals that was mediated by gold atoms.