Ab initio model potential embedded cluster calculation of the absorption spectrum of Cs2GeF6:Mn4+. Large discrepancies between theory and experiment

Alternative configuration interaction expansions for transition metal ions with intermediate oxidation states in crystals: The structure and absorption spectrum of Cs2GeF6:Mn4+

“…4 2Ϫ calculations where all 81 Mn and F electrons were treated explicitly and described by a much larger atomic natural orbital basis set: Mn(21s15p10d6 f 4g), 17,18 F(14s9 p4d), 19 contracted as Mn͓8s7 p6d2 f 1g͔, F͓4s3 p2d͔. In the end the test calculations indicated that the errors in the electronic transition energies associated with basis set limitations and use of core and embedding AIMP operators are smaller than 400 cm Ϫ1 .…”

“…2 The embedding potential we use for Cs 2 GeF 6 was obtained in Ref. 4. The spin free relativistic Cowan-Griffin-AIMP we use for Mn was obtained in Ref.…”

“…Along this line, a series of investigations conducted in our laboratory has pointed out that considerably different multiconfigurational expansions of the defect cluster wave functions are needed depending, basically, on the formal oxidation state of the impurity. 3,4 This is related to the fact that the interactions between d n and ligand-to-metal charge transfer configurations increase dramatically with formal oxidation state: They are small in V͑II͒, 5,6 considerable in Mn͑IV͒, 4 and very large in Fe͑VI͒ 3 impurities, for instance. Accordingly, the truncations of the configurational space needed to properly account for electron correlation cannot be the same.…”

“…The origins of the breakdown of the usual multireference treatment remained unrevealed and this, together with the instability of the results with the truncated space, was considered to be a significant methodological failure which needed to be communicated and further studied. 4 In effect, if the wave function based ab initio theoretical methods are meant to share with experimental techniques a significant role as tools for understanding the properties of these materials, it is necessary to find reasonable systematic configuration interaction truncation criteria which ultimately lead to uniform quality of both structural and spectroscopic properties irrespective of the oxidation state of the impurity. The work we present in this paper contributes in this line by proposing an alternative CI truncation for intermediate oxidation state Mn͑IV͒ which leads to the same high accuracy for the structure and spectroscopy of Cs 2 GeF 6 :Mn 4ϩ as that obtained so far for low oxidation state impurities doped in ionic hosts.…”

5 f → 5 f transitions of U4+ ions in high-field, octahedral fluoride coordination: The Cs2GeF6:U4+ crystal

“…21 They accurately reproduce the quantummechanical interactions between the ͑frozen͒ Hartree-Fock descriptions of the external crystal ions ͑e.g., Cs + , Ge 4+ , and F − in Cs 2 GeF 6 ͒ and the ͑multiconfigurational͒ wave functions associated with the point defect cluster ͑UF 6 ͒ 2− . They include ͑i͒ a long-range Coulomb term, which in ionic crystals is the corresponding Madelung potential, ͑ii͒ a shortrange Coulomb term, which corrects the latter taking into account that the lattice ions are not point charges but charge densities associated with Hartree Fock wave functions, ͑iii͒ an exchange term, which stems from the fact that the generalized antisymmetric product of the cluster and the external ion wave functions fulfil the first-principles requirement of antisymmetry with respect to interchange of electrons between the cluster and lattice group functions, and ͑iv͒ a projection term, which guarantees that the defect cluster and external ion wave functions are built using linearly independent sets of orbitals; this term actually prevents variational collapse of the cluster wave functions on the lattice ions.…”