Previous abstract Next abstract

Session Q12 - DCMP: DENSITY FUNCTIONAL THEORY
Mixed session, Friday morning, March 24, 8:00
Room C1, San Jose Convention Center

[Q12.06] Escaping the Symmetry Dilemma through a Re-Interpretation of Spin-Density Functional Theory

Routine predictions for the properties of atoms, molecules, and solids rely upon Kohn-Sham electron spin-density functional calculations within the local density or generalized gradient approximations. In the standard interpretation, these calculations predict the ground-state energy E and spin densities n_\uparrow (\bf r) and n_\downarrow(r) . For stretched H _2 and other molecules, such calculations make a severe error in the spin magnetization, m = n_\uparrow -n_\downarrow , while yielding reasonable values for both E and n = n_\uparrow + n_\downarrow . On the other hand, if the correct spin magnetization is imposed on the system, the calculated value of E is seriously in error. This epitomizes the symmetry dilemma. But the same approximations can apply to more than one underlying theory. An alternative and more tenable interpretation, based upon a different underlying theory, will be presented. J. P. Perdew, A. Savin, and K. Burke, submitted to Phys. Rev. A. In this new interpretation, the physical predictions are E , n(r) , and the on-top electron pair density P(r,r) , instead of m(r) , and thus no symmetry dilemma arises. (Work supported by NSF grant DMR92-13755.)

Part Q of program listing