R. Giridharagopal, M.D. Breshears, D.S. Ginger
University of Washington,
United States
Keywords: photovoltaics, electrostatic force microscopy, solar cells
Summary:
Next-generation energy materials, from organic photovoltaics to hybrid organic-inorganic perovskites, exhibit a wide range of dynamic processes that govern their operation. For example, minority carrier lifetimes (which affect quantum efficiency) and ion motion processes (which affect hysteresis and stability) span timescales ranging from sub-microseconds up to milliseconds or even seconds. More importantly, the timescales for these events have strong dependence on local structure and morphology, thus requiring a technique that combines both microsecond time resolution and nanometer spatial resolution. Atomic force microscopy (AFM) has been used with great success to image materials at such lengthscales. However, AFM is generally unable to resolve dynamics faster than milliseconds. On the other hand, more recent “big data” approaches to analyzing AFM cantilever motion have yielded new insight into dynamic processes. Changes in the AFM cantilever’s oscillating motion reflects information about the sample due to the electrostatic effects between the AFM tip and the substrate. Thus, by recording the cantilever position and then applying signal processing to the data, it is possible to extract new temporal insights from the measurement with sub-microsecond time resolution. Here we use an AFM-based method, time-resolved electrostatic force microscopy (trEFM), to analyze state-of-the-art perovskite solar cell materials in realistic device structures. trEFM measures dynamic information in sub-microsecond regimes by applying time-frequency analysis to the cantilever’s motion in response to a transient stimulus, in this case photoexcitation of the solar cell. This method allows us to extract dynamics from the AFM, which provides valuable insight into the transient phenomena that govern the behavior in many photovoltaics by imaging effects like ion motion and carrier recombination. We show that trEFM can be used to image these dynamics across different timescales. For example, trEFM data on a model layered perovskite system show that grain centers exhibit both faster charging and lower charge density, implying a small density of trap states at grain boundaries. More recently, we have used trEFM to explore microsecond-scale carrier recombination in perovskites. Under illumination, carriers in a solar cell will naturally recombine; the average time for this event is critical to determining the efficiency of the material. We demonstrate that trEFM can be used to image this recombination with high-spatial resolution. Further, we show that systematic ways to affect the recombination rate in device measurements are indeed replicated at the nanoscale in trEFM. Lastly, it is possible to correlate the trEFM measurements directly to recombination lifetimes in a solar cell through application of advanced data processing methods, such as Fourier mode decomposition or via neural network-based analysis. Dynamic electrical AFM methods provide a powerful platform for analyzing functional materials. While we present these approaches in the context of analyzing photovoltaic materials, the general principle of measuring time constants in AFM is applicable to a wide range of candidate materials, from fast motion in bio-active materials to ion diffusion in polymer electrolytes.