Machine Learning for Perovskites' Reap-Rest-Recovery Cycle

Slides:

Advertisements

Similar presentations

Volume 1, Issue 1, Pages (September 2017)

Advertisements

Ian Marius Peters, Haohui Liu, Thomas Reindl, Tonio Buonassisi Joule

Ian Marius Peters, Haohui Liu, Thomas Reindl, Tonio Buonassisi Joule

Volume 1, Issue 1, Pages (September 2017)

Volume 3, Issue 2, Pages (August 2017)

A Reliability Look at Energy Development

Volume 2, Issue 2, Pages (February 2017)

“Seeing” the Invisibles at the Single-Molecule Level

Volume 2, Issue 1, Pages (January 2018)

by Dane W. deQuilettes, Sarah M. Vorpahl, Samuel D

Volume 1, Issue 4, Pages (December 2017)

Volume 1, Issue 3, Pages (November 2017)

Volume 1, Issue 4, Pages (December 2017)

Fig. 2 Nonlinearities in a cavity-embedded perovskite single crystal.

Fig. 2 Absolute recovery of species richness and relative recovery of species richness and composition in relation to stand age for Neotropical secondary.

Fig. 2 Transport properties of a BP transistor at low temperature.

Fig. 5 Metafluorophores with different photostability.

Fig. 3 Near-field maps of a multiply cracked α-(BEDT-TTF)2I3 crystal.

Fig. 3 Illustration of inter- and intrasegment dynamics contributing to NMR relaxation. Illustration of inter- and intrasegment dynamics contributing to.

Fig. 2 Zeeman splitting of the plateau and associated Kondo feature.

Fig. 5 Analytical performance of the wearable cortisol sensor.

Fig. 3 Scan rate effects on the layer edge current.

Fig. 4 Characterization of nanowood.

Fig. 1 Bioinspired design of AAD for promoting wound contraction.

Fig. 6 Solar utilization efficiency and SEC in thermal and nonthermal desalination systems. Solar utilization efficiency and SEC in thermal and nonthermal.

Fig. 1 Fluctuation-induced dynamic response patterns.

Fig. 5 Wearable closed-loop HMI.

Fig. 3 Characteristics of UV and temperature sensors.

Volume 1, Issue 2, Pages (October 2017)

Fig. 4 Model of the average SSE.

Fig. 2 EUV TG signal. EUV TG signal. Black lines in (A), (B), and (C) are the EUV TG signals from Si3N4 membranes at LTG = 110, 85, and 28 nm, respectively,

Fig. 4 EUV TG signal from Si.

Fig. 3 Thermal management strategies in STD systems.

Fig. 5 Schematic phase diagrams of Ising spin systems and Mott transition systems. Schematic phase diagrams of Ising spin systems and Mott transition systems.

Fig. 3 GIWAXS pattern of perovskite films with varied ligands.

Fig. 1 Structure and absorption spectra of the P

Fig. 3 Characterization of the current-induced effective fields.

Fig. 2 Results of the learning and testing phases.

Fig. 4 SPICE simulation of stochasticity.

by Lijian Zuo, Hexia Guo, Dane W

Fig. 2 NH3, NOx, SO2, and NMVOC emission changes triggered by the JJJ clean air policy. NH3, NOx, SO2, and NMVOC emission changes triggered by the JJJ.

by Alan She, Shuyan Zhang, Samuel Shian, David R

Fig. 1 Experiment description.

Fig. 4 Relationships between light and economic parameters.

Enzyme activity of PE-GA/Pt tuned by changing temperature or NIR light

by Weijun Ke, Constantinos C

Schematic of the proposed brain-controlled assistive hearing device

Fig. 4 Large-area solution-processed CdSe TFT arrays on a Si wafer and on glass substrates. Large-area solution-processed CdSe TFT arrays on a Si wafer.

Fig. 4 SOT-driven perpendicular magnetization switching in the FGT/Pt bilayer device. SOT-driven perpendicular magnetization switching in the FGT/Pt bilayer.

Fig. 4 CO2 emission changes triggered by the JJJ clean air policy.

Fig. 1 Structural and electrical properties of Bi2Se3/BaFe12O19.

Fig. 3 Maximal energy intake.

Multiplexed four- and eight-channel devices for rapid processing

Fig. 3 Comparisons of NDVI trends over the globally vegetated areas from 1982 to Comparisons of NDVI trends over the globally vegetated areas from.

Fig. 3 Depth-resolved structural characterization of perovskite nanocrystals in npSi films. Depth-resolved structural characterization of perovskite nanocrystals.

Fig. 3 Electrochemical performances of symmetric cells using control Li and composite Li electrodes. Electrochemical performances of symmetric cells using.

Fig. 3 Device architecture, photovoltaic performance, and operational stability of 3D/2D bilayer PSCs. Device architecture, photovoltaic performance, and.

Fig. 4 Polarizer tunable color multiplexing.

Fig. 6 Energetics of the CaL methane reforming process.

Fig. 2 Supraballs and films from binary SPs.

Organic Photovoltaics: Focus on Its Strengths

Fig. 3 High-tide flood extent at water levels of 1. 73, 2. 03, 2

Fig. 2 Comparison between the different reflective metasurface proposals when θi = 0° and θr = 70°. Comparison between the different reflective metasurface.

Fig. 1 STEM data: Original and processed by neural networks.

Height-dependent swarm response for a fixed amplitude of AM= 84 mm

Fig. 3 Rubbery strain, pressure, and temperature sensors.

Fig. 5 Schematics illustrating enhancement in April tornado activity due to SST. Schematics illustrating enhancement in April tornado activity due to SST.

Potential use of liquid membranes as selective solid/gas filters

Fig. 3 Spatial distribution of the shoot density (high densities are represented in dark green and low ones in bright yellow) in a simulation of a P. oceanica.

Fig. 5 Modeling of the ASE threshold using the kinetic equations and experimental parameter inputs. Modeling of the ASE threshold using the kinetic equations.

Presentation transcript:

Machine Learning for Perovskites' Reap-Rest-Recovery Cycle John M. Howard, Elizabeth M. Tennyson, Bernardo R.A. Neves, Marina S. Leite Joule Volume 3, Issue 2, Pages 325-337 (February 2019) DOI: 10.1016/j.joule.2018.11.010 Copyright © 2018 Elsevier Inc. Terms and Conditions

Joule 2019 3, 325-337DOI: (10.1016/j.joule.2018.11.010) Copyright © 2018 Elsevier Inc. Terms and Conditions

Figure 1 Perovskite Photovoltaic Route to Reliability (A) Clockwise: extrinsic (H2O and O2) and intrinsic (bias, temperature, and light) factors governing dynamics in perovskite solar cells. (B) The reap-rest-recovery (3R) cycle for obtaining long-term power output from hybrid perovskite solar cells. Operating conditions previously assumed to promote irreversible deterioration need to be reevaluated within the framework of the 3R cycle. Joule 2019 3, 325-337DOI: (10.1016/j.joule.2018.11.010) Copyright © 2018 Elsevier Inc. Terms and Conditions

Figure 2 3R Cycle in Perovskite Solar Cells (A) In the reap phase, devices provide power. There is an initial exponential decay in performance regardless of ambient gas composition or relative humidity (rH) level. The subsequent performance trend depends on the O2 and H2O levels (extrinsic parameters), where devices aged in (1) N2 (green) shows a clear linear regime, (2) dry air with 5% rH (blue) and with 0% rH (red) continue the decay, and (3) the dry air with 100% rH (black) experiences immediate performance deterioration. In all cases, the shaded area represents the standard deviation. Adapted by permission from Domanski et al.,12 Springer Nature: Nature Energy, Copyright 2018. (B) During the rest phase of the cycle, the solar cells are not operational. Here, the rest duration corresponds to the time it takes for the residual voltage under dark conditions to stabilize. Adapted with permission from Hu et al.,17 copyright 2017 American Chemical Society. (C) The recovery phase of this cycle takes place when all figures-of-merit return to a substantial fraction of their original values. In this example, greater than 70%, regardless of whether the device was used through its T80 or T50 lifetime. Adapted with permission from Khenkin et al.,18 Copyright 2018 American Chemical Society. Joule 2019 3, 325-337DOI: (10.1016/j.joule.2018.11.010) Copyright © 2018 Elsevier Inc. Terms and Conditions

Figure 3 Capturing the Microscopic Optoelectronic Dynamics of Perovskites (A) The effect of illumination duration (light ON = green and light OFF = white) on the photoluminescence (PL) intensity of different crystal facets. MAPbI3 grains varying in size show radically different dynamics under illumination cycling. Adapted with permission from Tian et al.,39 published by The Royal Society of Chemistry. (B) Illuminated-Kelvin-probe force microscopy is used to determine the time-dependent changes in local Voc (1 ms/pixel with 128 × 128 frames). Even within a single grain, the MAPbI3 perovskite exhibits ion motion. Reprinted with permission from Garrett et al.,10 Copyright 2017 American Chemical Society. (C) PL imaging establishes the intergrain heterogeneity in MAPbI3 films, and time-dependent measurements reveal the light-emission stabilization. The relative brightness of the grain provides indication of trap state density. Reprinted from deQuilettes et al.,43 Copyright the authors, some rights reserved; exclusive licensee to Macmillan Publishers Ltd. Distributed under a Creative Commons Attribution Noncommercial License 4.0 (CC BY-NC) http://creativecommons.org/licenses/by-nc/4.0/. Joule 2019 3, 325-337DOI: (10.1016/j.joule.2018.11.010) Copyright © 2018 Elsevier Inc. Terms and Conditions

Figure 4 A Machine Learning Framework for a Perovskite 3R Cycle (A) Time-series laboratory data including the effect of each intrinsic and extrinsic parameter (H2O, O2, bias, temperature, and illumination) on device efficiency is used for training the algorithm. (B) A feature vector is extracted out of the environmental sensor output, together with the solar cell efficiency, η. (C) An artificial neural network (ANN) is tuned to maximize long-term stability and overall power output. (D) After the neural network weights are optimized, data from solar modules can be used to determine the rest phase conditions that will lead to recovery and sustained reap. (E) The future PV module's real-time conditions and performance are displayed onto an internet-connected device, enabling consumers to monitor its 3R cycle. Joule 2019 3, 325-337DOI: (10.1016/j.joule.2018.11.010) Copyright © 2018 Elsevier Inc. Terms and Conditions