Accurate Prediction of Large Antiferromagnetic Interactions in High-TcHgBa2Ca<

Abstract

The in-plane nearest-neighbor Heisenberg magnetic coupling constant, J, of La 2 CuO 4 , Nd 2 CuO 4 , Sr 2 CuO 2 Cl 2 , YBa 2 Cu 3 O 6 , and undoped HgBa 2 Ca n21 Cu n O 2n121d (n 1, 2, 3) is calculated from accurate ab initio configuration interaction calculations. For the first four compounds, the theoretical J values are in quantitative agreement with experiment. For the Hg-based compounds the predicted values are 2135 meV (n 1) and ϳ2160 meV (n 2, 3), the latter being much larger than in previous cases and, for n 3, increasing with pressure. Nevertheless, the physics governing J in all these layered cuprates appears to be the same. Moreover, calculations suggest a possible relationship between J and T c . PACS numbers: 74.25.Jb, 74.25.Ha, 75.30.Et More than ten years have passed since the discovery of the high-T c superconductivity in several cuprate compounds and, up to now, there has been no agreement on a complete theory capable of explaining their anomalous physical properties [1]. The layered crystal structure, the strong antiferromagnetic coupling between neighbor Cu 21 ions in these layers, and the strongly correlated nature of their electronic structure have been pointed out to play a very important role in the fundamental mechanisms of superconductivity [1,2]. The high-T c superconductors are produced by moderate doping of "parent compounds" such as La 2 CuO 4 , Nd 2 CuO 4 , or YBa 2 Cu 3 O 6 . It is widely accepted that proximity to the insulating phase and the interaction of dopant carriers with magnetic degrees of freedom are fundamental aspects for the existence of high-T c superconductivity. Recent theories strongly suggest that magnetic coupling is at the very origin of the superconducting state [2]. Thus, the magnetic coupling constants defining the magnetic structure of these compounds are key magnitudes in establishing trends and theories based on model Hamiltonians.Recently, several families of layered cuprates, containing Bi, Tl, or Hg, have been synthesized. The HgBa 2 Ca n21 Cu n O 2n121d (n 1, 2, 3) family [3][4][5] includes HgBa 2 Ca 2 Cu 3 O 81d , the compound with the highest transition temperature observed to date (133 K at ambient pressure and 164 K under 31 GPa) [4,6,7]. The difficult synthesis and the layered structure of those materials limits the capability of obtaining pure crystals large enough to perform accurate measurements of their magnetic properties. This restricts the theoretical understanding due to the difficulty to obtain reliable estimates of the model Hamiltonian parameters. Modern theoretical approaches to electronic structure could provide accurate values of these parameters, such as the nearest-neighbor Heisenberg magnetic coupling constant J, thus providing a more detailed description of these materials.Unfortunately, because of the strong correlated nature of the electronic structure and the antiferromagnetic insulating character of these cuprates, band structure calculations based on the local density approximation (LDA) or on the generalized gradient app...

Spin Symmetry Requirements in Density Functional Theory: The Proper Way to Predict Magnetic Coupling Constants in Molecules and Solids

“…The U-HF values (not shown) are close to CASCI or CASSCF results, provided that appropriate spin projection is carried out [4,5,[54][55][56][57][58][59]. This is because the symmetry breaking introduces certain amount of non-dynamical electron-correlation [85,86], which is largely responsible for the superexchange contribution of the magnetic coupling parameter [15,79,87]. Here, it is important to remark that the U-HF projected J values are always consistent with the mapping procedure.…”

“…In all cases, the calculated J values are comparable to experimental available data and for H-He-H nicely reproduce the exact FCI values. The CASSCF and CASCI methods tend to systematically underestimate the superexchange contribution of J, because the dynamic electron-electron correlation effects are neglected [15,79]. In some cases, as in [Cu 2 Cl 6 ] 2− , this lack of dynamical correlation leads even to qualitatively incorrect results [4].…”

“…Still, one can use the non-relativistic Hamiltonian and handle magnetic interactions through a proper introduction of spin coordinates and spin symmetry. In this formalism, magnetic interactions in systems with localized magnetic moments result from important electronelectron correlation effects [1,8,15] which, to provide accurate theoretical estimates of J i j values, must be appropriately described. This requires going beyond the usual mean field theories and therefore, the electronic structure of magnetic systems provide severe test cases for theoretical methods.…”

Neutron Scattering Studies of Antiferromagnetic Correlations in Cuprates

“…To calculate the values of J from spectroscopically determined parameters, one must consider at least a 3-band Hubbard model [98]. Recent ab initio cluster calculations [99,100] have been able to achieve reasonable agreement with experiment. While the magnitude of J in layered cuprates is rather large, it is not extreme; a value of J = 226(12) meV has been measured for Cu-O chains in SrCuO 2 [101].…”

Ab initioperiodic approach to electronic structure and magnetic exchange inA2CuO2X2Moreira1, Dovesi2 2003Phys. Rev. B10010View full textAdd to dashboardCite

“…Recently, embedded cluster model calculations of magnetic exchange in HTSC parent compounds and other TM compounds have been reported by Illas and collaborators [61][62][63][64][65][66] by using UHF and highly sophisticated configuration interaction ͑CI͒ schemes to account for the correlation effects on the relevant electronic states defining J 1 . Small cluster models have been used which are, however, large enough for a careful determination of the first neighbor superexchange constants that are in good agreement with available experimental data.…”