John and Ali both gave talks the Fall 2017 MRS meeting in Boston this week.
Ali spoke about his recent paper in the following talk: Tirmzi, A. M; Dwyer, R. P.; Hanrath, T. & Marohn, J. A.“An Improved Treatment of the Scanning Kelvin Probe Microscopy Experiment to Track Coupled Slow and Fast Dynamics in Cesium Lead Bromide Perovskite”, Materials Research Society Fall Meeting; Boston, Massachusetts; November 26 – December 1, 2017 [2017-11-28 talk TC01.04.05].
To interpret Scanning Kelvin Probe Microscopy (SKPM) and Electric Force Microscopy (EFM) measurements, it is standard to assume that the tip-sample charge adjusts instantaneously, i.e. on a timescale fast compared to cantilever period, to changes in the applied tip voltage and the cantilever position. While this assumption holds for most organic and inorganic samples, the breakdown of this assumption can have a profound effect on the interpretation of experimental observations. Here we show the breakdown of this assumption on a technologically relevant photovoltaic sample, cesium lead bromide perovskite.
In our experiment we applied light to the sample. As we increased the light intensity, the apparent capacitance increased nonlinearly while the sample induced dissipation increased nonlinearly, reached a maximum, and then decreased. To understand these extremely puzzling observations, we developed a more rigorous treatment of the SKPM experiment in which Lagrangian mechanics was used to model the coupled motion of the cantilever tip and charges in the cantilever and the sample. In this treatment, the measured frequency shift and sample-induced dissipation depend explicitly on the complex sample impedance [1]. This treatment revealed that the apparently puzzling photocapacitance increase and dissipation had a common origin — a monotonic decrease in the sample RC time with light intensity, from sub-millisecond in the dark to sub-microsecond at high intensity. We independently validated this sample-RC finding by performing a variant of Broadband Local Dielectric Spectroscopy (BDLS) as a function of illumination intensity. The resulting data confirmed that we were indeed causing a transition from the fast-response limit to slow-response limit by increasing the background light intensity applied to the sample [2].
Surprisingly, we found that sample’s RC time constant relaxed on the seconds time scale (at room temperature) when the light was turned off. We measured this relaxation time versus temperature and obtained an activation energy for the relaxation process. To explain the anomalously slow relaxation, we propose a picture in which ion-associated trapping centers cause this slow relaxation. This slow relaxation of sample conductivity in the dark provides direct evidence for coupled slow and fast dynamics in cesium lead bromide perovskite.
[1] Dwyer, R. P , Tirmzi, A. M., Harrell, L. E.; Marohn, J. A.(unpublished); [2] Tirmzi, A. M.; Dwyer, R. P.; Hanrath, T. & Marohn, J. A. ACS Energy Lett., 2017, 2, 488 - 496.
John spoke about Ryan’s recent paper in the following talk: Dwyer, R. P.; Nathan, S. R.; & Marohn, J. A.“Achieving Sub-cycle Measurements of Capacitance and Photocapacitance in a Scanning Probe Microscope Experiment by Using a Microcantilever as a Gated Mechanical Integrator”, Materials Research Society Fall Meeting; Boston, Massachusetts; November 26 – December 1, 2017 [2017-11-28 talk TC01.04.03].
Electric force microscopy (EFM) and Kelvin probe force microscopy have revealed an enormous amount of useful information about organic semiconductor materials by enabling measurements of charge generation, injection, transport, and trapping with high spatial resolution. To test theories of charge generation and recombination, however, one also needs to follow sample properties with high temporal resolution. Light-induced charge generation and recombination is typically studied using nanosecond-resolution time-resolved microwave conductivity, but the technique has no spatial resolution [1]. Ginger and coworkers introduced time-resolved EFM (tr-EFM), in which transient photocapacitance is followed by observing the cantilever’s oscillation frequency with microsecond temporal resolution, and showed that the photocapacitance charging rate measured by tr-EFM was proportional to the external quantum efficiency in benchmark organic photovoltaic systems [2]. For charging rates faster than half a cantilever oscillation period (e.g., 1.5 microseconds), however, the demodulated cantilever frequency cannot be clearly interpreted because the cantilever’s oscillation spectrum violates the requirements of Bedrosian’s product theorem for analytic signals [3].
We introduce a new method for measuring photocapacitance transients with a scanning-probe microscope that sidesteps this seemingly fundamental limit. In our experiment [4], a voltage pulse is applied to charge the cantilever while a light pulse is applied to generate free charges in the sample. These sample charges shift the cantilever’s frequency and phase of oscillation. Snapshots of the sample’s evolving photocapacitance are obtained by recording the net change in cantilever phase as a function of the time delay between the light and voltage pulses. The cantilever in this experiment is essentially acting as a gated mechanical integrator. We demonstrate the method by using it to reveal a biexponential photocapacitance buildup in a polymer-blend solar-cell film, PFB:F8BT on ITO, with the fast component having a risetime of 40 microseconds at high light intensity. We demonstrate the superior signal-to-noise and time resolution of the new method by using it to record the 10’s of nanoseconds probe-wiring time constant of our apparatus in ca. 100 ms of total acquisition time.
[1] (a) Coffey, D. C.; et al. & Rumbles, G. J. Phys. Chem. C, 2012, 116:8916; (b) Ward, A. J.; et al. & Samuel, I. D. W. Adv. Mater., 2015, 27:2496; and (c) Ihly, R.; et al.; Rumbles G. & Blackburn, J. L. Nature Chem., 2016, 8:603; [2] (a) Coffey, D. C. & Ginger, D. S Nat. Mater., 2006, 5:735; (b) Giridharagopal, R.; et al. & Ginger, D. S Nano Lett., 2012, 12:893; and (c) Karatay, D.U; et al. & Ginger, D.S. Rev. Sci. Instrum., 2016, 87:053702; [3] (a) Boashash, B. Proc. IEEE, 1992, 80:520 and (b) Rihaczek, A. & Bedrosian, E. Proc. IEEE, 1966, 54:434; and [4] Dwyer, R. P.; Nathan, S. R. & Marohn, J. A. Sci. Adv., 2017, 3:e1602951.
This work was funded by the U.S. National Science Foundation.