(581h) Elucidation and Mitigation of Radiation Damage in Hybrid Organic-Inorganic Perovskites Using Pulsed Electron Beams

Hybrid organic-inorganic perovskites (HOIPs) have achieved lab-scale photovoltaic (PV) power conversion efficiencies of over 25% using low-temperature solution-based syntheses, but the development of a thorough understanding of the fundamental science responsible for HOIPs ideal optoelectronic properties is ongoing [1,2]. Transmission electron microscopy (TEM) is one tool employed to advance this area of knowledge via high resolution characterization of structure and morphology, chemical analysis, beam-induced current mapping, or even ultrafast structural dynamics at high spatial resolution [3]. However, the organic components and van der Waals bonding critical to the hybrid nature of the material renders HOIPs susceptible to degradation under various environmental conditions, including electron beam irradiation. The sensitivity of HOIPs to electron beam damage in the TEM – even under dose rates as low as 1 eA^-2s^-1 – is such that insight into technologically-important features such as grain boundaries and defects is limited due to modulation of these features during study [3-5]. During electron beam irradiation, a combination of charging, ionic excitation, and heating leads to ion migration and eventually dissolution into organic and inorganic components [6]. Importantly, this damage mechanism is similar to those which cause instability in the presence of heat and solar irradiation, which limits commercial adoption of HOIP PVs [3,7]. Indeed, there is both technological and fundamental significance in electron-beam damage studies on HOIPs – mitigating damage may lead to the use of TEM to gain insight into fundamental structural and optoelectronic properties, and understanding the damage mechanisms themselves may lead to the ability to mitigate them in circumstances where they cause degradation of PV devices beyond acceptable levels for commercial implementation.

In light of the relevance of electron-beam damage, myriad mitigation methods have been employed in the study of HOIPs. One which demonstrates potential on both aspects of the above-stated goal – mitigating damage and understanding it – is the use of temporally modulated electron beams, which deliver electron doses with prescribed regularity in place of a thermionic beam in which electron emission is randomly distributed in time. Such beams have been used to illuminate polymer crystals for inordinately prolonged exposures with minimal radiation damage, to suppress a commonly-encountered damage-induced phase change in MgCl₂, and to reduce damage to model paraffin crystals (C₃₆H₇₄) by nearly a factor of two as compared to conventional thermionic (random) emission for the same total dose [8-10]. In the case of this work, the temporally-modulated electron beam is created by training a femtosecond pulsed laser on the source material in the TEM gun region. Importantly, the range of pulsed-electron beam parameters that can be covered in this way is directly dependent upon the pulsed-laser properties; time between electron packets is directly tied to laser repetition rate, and fine control over the number of electrons per packet is controlled with laser fluence. This in turn amounts to control over both the specimen relaxation time between packets and the instantaneous dose rate, respectively, enabling exploration of factors such as specimen heating and cumulative, spatially-common energy-deposition effects.

HOIPs are prime candidates for study using pulsed beams because not only is mitigation of damage possible, but fundamental mechanistic studies are also enabled. Specifically, during electron beam irradiation, volatilization and migration of iodide ions results in a superlattice with halide vacancies, the formation of which is highly dose-rate dependent [6]. This intermediate phase subsequently decomposes further with the loss of additional halide ions along with organic cations, ultimately transforming into pure metal halide. The timescales associated with this are not yet known, but the dose-rate dependence even at dose rates as low as 0.5-1.0 e·Å^-2·s^-1 suggests that energy deposition, for example by multiple electrons arriving rapidly relative to the timescale of energy dissipation, can lead to enhancement of the rate of damage. Similarly, HOIPs have a low thermal conductivity of 0.5 W·m^-1·K^-1 and have been shown to degrade more rapidly under thermal stresses, leading to the hypothesis that electron beam heating could enhance damage rates as well [11,12]. The use of the above-described fine control over specimen relaxation time and dose rate offers insights into these hypotheses. Here, we demonstrate the ways in which pulsed beams consisting of packets of electrons delivered at precise times can be used to reduce damage to polycrystalline methylammonium lead iodide (MAPbI₃) specimens, demonstrated quantitatively by using diffraction peak attenuation to track damage. We discuss insights given by these experiments into damage mechanisms of high-energy (200 kV) electron beams on HOIPs, and report real-space imaging studies on the material using temporal modulation techniques. Because of the two-fold aspect of this work, in which damage is both reduced and studied, our results provide insight into both the fundamental mechanisms and practical mitigation strategies for HOIPs towards their implementation in commercial devices.

References:

[1] Best Research-Cell Efficiency Chart, https://www.nrel.gov/pv/cell-efficiency.html (accessed February 17th, 2020).

[2] D. A. Egger, et al., Adv. Mater. 30, (2018), 1800691.

[3] J. Ran, et al., Adv. Energy Mater., (2020), 1903191.

[4] M. U. Rothmann, et al., Adv. Mater. 30, (2018), 1800629.

[5] J. M. Ball and A. Petrozza, Nature Energy 1, (2016), 16149.

[6] S. Chen, et al., Nat. Commun. 9, (2018).

[7] Z. Chu, et al., Nat. Commun. 8, (2017).

[8] E. J. VandenBussche and D. J. Flannigan, Nano Lett. 19, (2019), 6687.

[9] C. Kisielowski, et al., Adv. Funct. Mater. 29, (2019), 1807818.

[10] O. H. Kwon, et al., Proc. Natl. Acad. Sci. U.S.A. 108, (2011), 6026.

[11] A. Pisoni, et al., J. Phys. Chem. Lett. 5, (2014), 2488.

[12] G. Divitini, et al., Nature Energy 1, (2016), 15012.