Notorious discrepancies in crystal field analyses eliminated by configuration interaction

Some conspicuous discrepancies between experimental and calculated energy levels in the crystal field analysis of rare earth (4f²) or actinide (5f²) electronic spectra are eliminated if one includes in the interaction matrix the participation of the 4f¹6p¹ (or 5f¹7p¹) excited configuration in addition to the ground configuration. This was recently demonstrated in the case of the tetravalent uranium U⁴⁺ in the cubic centrosymmetric sites of elpasolithe compounds (M.D. Faucher, O.K. Moune, D. Garcia and P. Tanner, Phys. Rev. B, 53 (1996) 9501). The present work is a step towards the generalization of the proposition. It is shown that it is also true for the 4f³ configuration of Nd³⁺ in non-centrosymmetric sites. Indeed, the notorious discrepancy between experimental and calculated sublevels of the ²H(2)_11/2 group at 15000 cm⁻¹ is eliminated by configuration interaction with the excited 4f²6p¹ configuration. The experimental case of Nd₂O₂S is quoted as an example, but the demonstration is valid for all neodymium compounds.