Imaging photoinduced surface potentials on hybrid perovskites by real-time Scanning Electron Microscopy.

We introduce laser-assisted Time-Resolved SEM (TR-SEM), joining Scanning Electron Microscopy and laser light excitation, to probe the long-term temporal evolution of optically excited charge distributions at the surface of Metal Ammonium Lead Triiodide (MAPbI3 ) hybrid perovskite thin films. Laser-assisted TR-SEM relies on the optically induced local modification of Secondary Electron (SE) detection yield to provide mapping of photoexcited potentials and charge dynamics at surfaces, and qualifies as a complementary approach to near-field probe microscopies and nonlinear photoemission spectroscopies for photovoltage measurements. Real-time imaging of evolving field patterns are provided on timescales compatible with SEM scanning rates, so that temporal resolution in the millisecond range can be ultimately envisaged. MAPbI3 is an outstanding light-sensitive material candidate for applications in solar light harvesting and photovoltaics, also appealing as an active system for light generation. In this work, the real time temporal evolution of optically induced SE contrast patterns in MAPbI3 is experimentally recorded, both under illumination by a 405 nm blue laser and after light removal, showing the occurrence of modifications related to photoinduced positive charge fields at surface. The long term evolution of these surface fields are tentatively attributed to ion migration within the film, under the action of the illumination gradient and the hole collecting substrate. This optical excitation is fully reversible in MAPbI3 over timescales of hours and a complete recovery of the system occurs within days. Permanent irradiation damage of the material is avoided by operating the SEM at 5 keV of energy and 1-10 pA of primary current. Optical excitation is provided by intense above-bandgap illumination (up to 50 W/cm2 ). TR-SEM patterns show a strong dependence on the geometry of SE collection. Measurements are taken at different axial orientations of the sample with respect to the entrance of the in-column detection system of the SEM and compared with numerical modeling of the SE detection process. This enables to single out the information regarding the local potential distribution. Results are interpreted by combining data about the spectral distribution of emitted SEs with the configuration of the electric and magnetic fields in the specimen chamber. The present modeling sets a robust basis for the understanding of photoinduced SE electron contrast.