Excitons are bound electron-hole pairs, i.e., atomic-H like Bosonic quasiparticles, that determine many optical and optoelectronic properties of solid materials. Exciton formation and dissociation play decisive roles in next generation solar cells. In a conventional p-n junction solar cell, the built-in potential separates the photoexcited electron and hole. In contrast, separating the electron and the hole in an excitonic solar cell requires an energetic driving force at a donor/acceptor (D/A) materials interface. Here, photon absorption creates a localized Frenkel exciton or a delocalized Mott-Wannier exciton in the donor material. Such an exciton migrates to the D/A interface and decays into a charge transfer (CT) exciton: the Coulombically-bound electron and hole are located in spatially separate regions across the interface. Subsequent dissociation of the CT exciton leads to charge carriers and photocurrent. In this talk, I will present our understanding on the exciton dissociation problem from recent experiments and discuss challenges in theoretical/computation treatment of this problem. These challenges arise because one must simultaneous take into account translational symmetry of the donor and acceptor (when the donor and/or acceptor are crystalline materials) and the spatial correlation of the e-h pair.