Evolution of the stripe phase as a function of doping from a theoretical analysis of angle-resolved photoemission data

By comparing single-particle spectral functions of $t\ensuremath{-}J$ and Hubbard models with recent angle-resolved photoemission results for ${\mathrm{La}}_{2\ensuremath{-}x}{\mathrm{Sr}}_{x}{\mathrm{CuO}}_{4}$ (LSCO) and Nd-LSCO, we can decide where holes go as a function of doping, and more specifically, which type of stripe (bond-, site-centered) is present in these materials at a given doping. For dopings greater than about $12%$ our calculation shows that the holes prefer to proliferate out of the metallic stripes into the neighboring antiferromagnetic domains. The spectra were calculated by a cluster-perturbation technique, for which we present an alternative formulation. Implications for the theory for $\mathrm{high}\ensuremath{-}{T}_{c}$ superconductivity are discussed.