Coloumb interaction and instability of CE-structure in half doped manganites

In their Letter (Phys. Rev. Lett. 83, 5118 (1999)), den Brink, Khaliullin, and Khomskii proposed theoretically that the one-dimensional ferromagnetic zigzag chains in CE phase in half-doped manganites play an essential role in forming the orbital ordering, and, more surprisingly, the on-site Coulomb interaction U between electrons with different orbitals leads to experimentally observed charge ordering. In this Comment, I point out that the strong U will destroy the stability of CE-type phase, which is stable in a very narrow regime in the parameter space for electronic model.To solve this issue finally, we have to take into account other interactions, such as the long-range Coulomb interaction, Jahn-Teller distortion, and physics of topological berry phase. For example, the effect of finite large J$_{H}$ leads to an attractive particle-hole interaction, which favors to stabilize the charge ordering.