Localized Bonds and Loge Theory

The various localization methods, which exploit the fact that the determinantal wave function is invariant to a unitary transformation, have been discussed in details by Surjan [13]. These methods lead to two-electron bonds for localizable systems, and 

The failure of efforts to localize certain systems to two-electron fragments calls our attention to the necessity of using larger bulding blocks than just a simple two-electron bond. The difficulties in the proper treatment of aromatic systems by the PCILO method, which uses exclusively such two-electron bonds (e.g. Kekule structures for aromatic systems) are well known [38], and lead to the development of the Extended PCILO Method , where this constraint has been released [39]. 

---