Abstract
Ultrathin layers of native oxides on the surface of photovoltaic materials may act as very efficient carrier trapping/recombination centers, thus significantly affecting their interfacial photophysical properties. How ultrathin oxide layers affect the surface and interface carrier dynamics cannot be selectively accessed by conventional ultrafast transient spectroscopic methods due to the deep penetration depth (tens to thousands of nanometers) of the pump/probe laser pulses. Herein, scanning ultrafast electron microscopy (S-UEM) at a low voltage of 1 keV electrons was recently developed at KAUST to selectively map the ultrafast charge carrier dynamics of a few layers on the top surfaces of photovoltaic materials. Unlike high-voltage 30 keV experiments, at 1 keV, the depth of detected secondary electrons (SEs) underneath the surface is significantly reduced 5 times, thus making it possible to visualize the dynamics of charge carrier from the uppermost surface of photoactive layers. More specifically, this new approach has been employed to explore and decipher the tremendous impact of native oxide layers and oxygen defect states on charge carrier dynamics in space and time simultaneously at sub-nanometer levels on several photovoltaic material surfaces, including Si, GaAs, CdTe, and CdZnTe single crystals. Interestingly, the contrast in the SEs time-resolved difference images switched from a dark "energy-loss mechanism" to a bright "energy-gain mechanism" only by removing the layers of surface oxides. Moreover, the charge carrier recombination was estimated and found to be dramatically affected by the native oxide layers. The density functional theory (DFT) calculations demonstrate that the work function of oxygen-terminated Si surface also increases slightly upon optical excitation and makes for less SE intensity, providing another piece of evidence that the origin of the dark contrast observed on these material surfaces should be assigned to the surface oxide formation, mainly oxygen defect states in the band gap and/or work function increment. Our findings provide a new method and pave the way for evaluating the effect of surface morphology and defects, including but not limited to native oxide, on charge carrier dynamics, and interfacial properties of photovoltaic absorber layers.