On the Existence of Antiferromagnetism in Hubbard's Approach to the Hubbard Model

Abstract We investigate the possibility of selfconsistent solutions for antiferromagnetism in the Hubbard model in the decoupling of the Greens functions introduced by Hubbard in his first paper. On the base of this approximation Arai has calculated the band splitting for antiferromagnetism, but, as will be shown in this paper, Hubbard's approach fails to yield antiferromagnetism for nearest neighbour hopping in the same way as it does not yield ferromagnetism, and no selfconsistent solutions of the problem beyond the well known paramagnetic solution do exist.