About the charge dynamics at the perovksite-TiO2 interface, still in debate

V. Roiati, et al. suggested a double path for charge transport allowing electron percolation in both TiO2 and Perovskite (CH3NH3PbI3-xClx) through a detailed spectroscopy study.

see link: http://pubs.rsc.org/en/Content/ArticleLanding/2014/EE/C3EE43991G#!divAbstract

C.S. Ponseca, et al. concluded that when the perovskite (CH3NH3PbI3) is introduced into a TiO2 mesoporous structure, electron injection from perovskite to the metal oxide is efficient in less than a picosecond.

see link: http://pubs.acs.org/doi/abs/10.1021/ja412583t  

J. Bisquert, et al. suggested that  (using impedance spectroscopy)

"At the beginning of its discovery, the perovskite solar cell was a TiO2 sensitized cell, so it was natural to believe that electrons would be injected to the wide bandgap metal oxide as  the main transport pathway. However, the observation of significant efficiencies in perovskite solar cells with no electron transporting material, or with no hole transporting material or even in TF configuration,indicates that perovskite solar cell can work in a sensibly different configuration than DSC. These results constitute a strong indication that photogenerated electrons and holes coexist in CH3NH3PbX3 absorber material and travel to the selective contacts where they are separately collected.

see link:

http://pubs.acs.org/doi/abs/10.1021/nl404252e

So what is the real story?

Would be happy to see your opinion and even the experimental observation.