The number of operations required for conventional density-functional
algorithms grows as the cube of the number of atoms, N. For large syst
ems the computing requirements are unattainable. To overcome this limi
tation it is acceptable to approximate those variables with respect to
which the free energy is stationary. We show that the stationarity of
the free energy with respect to electron density, one-electron potent
ial, chemical potential, occupation function, and temperature allows f
or very useful approximations leading to rapid and accurate determinat
ion of the free energy. Here we discuss approximations involved in cal
culating the finite temperature electron density needed to evaluate th
e Harris-Foulkes free energy. Of particular importance are (1) an elec
tron density at each site that is based on exact solution of the Poiss
on equation combined with a solution of the multiple-scattering proble
m in which only scattering from a small cluster of sites surrounding t
he site in question is retained and (2) an approximate occupation func
tion having a finite number of poles in the complex energy plane. The
intention is to develop, within density-functional theory, an O(N) sca
lable first-principles scheme, based on spatially local multiple scatt
ering methods, for calculating free energies of large systems.