Exploring Physics of Ferroelectric Domain Walls in Real Time: Deep Learning Enabled Scanning Probe Microscopy

Yongtao Liu; Kyle P. Kelley; Hiroshi Funakubo; Sergei V. Kalinin; Maxim Ziatdinov

doi:10.1002/advs.202203957

Exploring Physics of Ferroelectric Domain Walls in Real Time: Deep Learning Enabled Scanning Probe Microscopy

Yongtao Liu, Kyle P. Kelley, Hiroshi Funakubo, Sergei V. Kalinin, Maxim Ziatdinov

Functional Atomic Force Microscopy Group

Research output: Contribution to journal › Article › peer-review

13 Scopus citations

Abstract

The functionality of ferroelastic domain walls in ferroelectric materials is explored in real-time via the in situ implementation of computer vision algorithms in scanning probe microscopy (SPM) experiment. The robust deep convolutional neural network (DCNN) is implemented based on a deep residual learning framework (Res) and holistically nested edge detection (Hed), and ensembled to minimize the out-of-distribution drift effects. The DCNN is implemented for real-time operations on SPM, converting the data stream into the semantically segmented image of domain walls and the corresponding uncertainty. Further the pre-defined experimental workflows perform piezoresponse spectroscopy measurement on thus discovered domain walls, and alternating high- and low-polarization dynamic (out-of-plane) ferroelastic domain walls in a PbTiO₃ (PTO) thin film and high polarization dynamic (out-of-plane) at short ferroelastic walls (compared with long ferroelastic walls) in a lead zirconate titanate (PZT) thin film is reported. This work establishes the framework for real-time DCNN analysis of data streams in scanning probe and other microscopies and highlights the role of out-of-distribution effects and strategies to ameliorate them in real time analytics.

Original language	English
Article number	2203957
Journal	Advanced Science
Volume	9
Issue number	31
DOIs	https://doi.org/10.1002/advs.202203957
State	Published - Nov 3 2022

Funding

This effort (implementation in SPM, measurement, data analysis) was primarily supported by the Center for 3D Ferroelectric Microelectronics (3DFeM), an Energy Frontier Research Center funded by the U.S. Department of Energy (DOE), Office of Science, Basic Energy Sciences under Award Number DE‐SC0021118. This research (ensemble‐ResHedNet) was sponsored by the INTERSECT Initiative as part of the Laboratory Directed Research and Development Program of Oak Ridge National Laboratory, managed by UT‐Battelle, LLC for the U.S. Department of Energy under contract DE‐AC05‐00OR22725. The research was partially supported (piezoresponse force microscopy) at Oak Ridge National Laboratory's Center for Nanophase Materials Sciences (CNMS), a U.S. Department of Energy, Office of Science User Facility. This effort (implementation in SPM, measurement, data analysis) was primarily supported by the Center for 3D Ferroelectric Microelectronics (3DFeM), an Energy Frontier Research Center funded by the U.S. Department of Energy (DOE), Office of Science, Basic Energy Sciences under Award Number DE-SC0021118. This research (ensemble-ResHedNet) was sponsored by the INTERSECT Initiative as part of the Laboratory Directed Research and Development Program of Oak Ridge National Laboratory, managed by UT-Battelle, LLC for the U.S. Department of Energy under contract DE-AC05-00OR22725. The research was partially supported (piezoresponse force microscopy) at Oak Ridge National Laboratory's Center for Nanophase Materials Sciences (CNMS), a U.S. Department of Energy, Office of Science User Facility.

Funders	Funder number
CNMS
center for 3D Ferroelectric Microelectronics
Oak Ridge National Laboratory's Center for Nanophase Materials Sciences
U.S. Department of Energy
Office of Science
Basic Energy Sciences	DE‐SC0021118
Basic Energy Sciences
Oak Ridge National Laboratory	DE‐AC05‐00OR22725
Oak Ridge National Laboratory
UT-Battelle

Keywords

automated experiments
deep convolutional neural network
ferroelastic domain walls
piezoresponse force microscopy

Access to Document

10.1002/advs.202203957

Cite this

@article{6224c835958a4c7db04dfb594fc0c9cb,

title = "Exploring Physics of Ferroelectric Domain Walls in Real Time: Deep Learning Enabled Scanning Probe Microscopy",

abstract = "The functionality of ferroelastic domain walls in ferroelectric materials is explored in real-time via the in situ implementation of computer vision algorithms in scanning probe microscopy (SPM) experiment. The robust deep convolutional neural network (DCNN) is implemented based on a deep residual learning framework (Res) and holistically nested edge detection (Hed), and ensembled to minimize the out-of-distribution drift effects. The DCNN is implemented for real-time operations on SPM, converting the data stream into the semantically segmented image of domain walls and the corresponding uncertainty. Further the pre-defined experimental workflows perform piezoresponse spectroscopy measurement on thus discovered domain walls, and alternating high- and low-polarization dynamic (out-of-plane) ferroelastic domain walls in a PbTiO3 (PTO) thin film and high polarization dynamic (out-of-plane) at short ferroelastic walls (compared with long ferroelastic walls) in a lead zirconate titanate (PZT) thin film is reported. This work establishes the framework for real-time DCNN analysis of data streams in scanning probe and other microscopies and highlights the role of out-of-distribution effects and strategies to ameliorate them in real time analytics.",

keywords = "automated experiments, deep convolutional neural network, ferroelastic domain walls, piezoresponse force microscopy",

author = "Yongtao Liu and Kelley, {Kyle P.} and Hiroshi Funakubo and Kalinin, {Sergei V.} and Maxim Ziatdinov",

note = "Publisher Copyright: {\textcopyright} 2022 The Authors. Advanced Science published by Wiley-VCH GmbH.",

year = "2022",

month = nov,

day = "3",

doi = "10.1002/advs.202203957",

language = "English",

volume = "9",

journal = "Advanced Science",

issn = "2198-3844",

publisher = "Wiley Open Access",

number = "31",

}

TY - JOUR

T1 - Exploring Physics of Ferroelectric Domain Walls in Real Time

T2 - Deep Learning Enabled Scanning Probe Microscopy

AU - Liu, Yongtao

AU - Kelley, Kyle P.

AU - Funakubo, Hiroshi

AU - Kalinin, Sergei V.

AU - Ziatdinov, Maxim

PY - 2022/11/3

Y1 - 2022/11/3

N2 - The functionality of ferroelastic domain walls in ferroelectric materials is explored in real-time via the in situ implementation of computer vision algorithms in scanning probe microscopy (SPM) experiment. The robust deep convolutional neural network (DCNN) is implemented based on a deep residual learning framework (Res) and holistically nested edge detection (Hed), and ensembled to minimize the out-of-distribution drift effects. The DCNN is implemented for real-time operations on SPM, converting the data stream into the semantically segmented image of domain walls and the corresponding uncertainty. Further the pre-defined experimental workflows perform piezoresponse spectroscopy measurement on thus discovered domain walls, and alternating high- and low-polarization dynamic (out-of-plane) ferroelastic domain walls in a PbTiO3 (PTO) thin film and high polarization dynamic (out-of-plane) at short ferroelastic walls (compared with long ferroelastic walls) in a lead zirconate titanate (PZT) thin film is reported. This work establishes the framework for real-time DCNN analysis of data streams in scanning probe and other microscopies and highlights the role of out-of-distribution effects and strategies to ameliorate them in real time analytics.

AB - The functionality of ferroelastic domain walls in ferroelectric materials is explored in real-time via the in situ implementation of computer vision algorithms in scanning probe microscopy (SPM) experiment. The robust deep convolutional neural network (DCNN) is implemented based on a deep residual learning framework (Res) and holistically nested edge detection (Hed), and ensembled to minimize the out-of-distribution drift effects. The DCNN is implemented for real-time operations on SPM, converting the data stream into the semantically segmented image of domain walls and the corresponding uncertainty. Further the pre-defined experimental workflows perform piezoresponse spectroscopy measurement on thus discovered domain walls, and alternating high- and low-polarization dynamic (out-of-plane) ferroelastic domain walls in a PbTiO3 (PTO) thin film and high polarization dynamic (out-of-plane) at short ferroelastic walls (compared with long ferroelastic walls) in a lead zirconate titanate (PZT) thin film is reported. This work establishes the framework for real-time DCNN analysis of data streams in scanning probe and other microscopies and highlights the role of out-of-distribution effects and strategies to ameliorate them in real time analytics.

KW - automated experiments

KW - deep convolutional neural network

KW - ferroelastic domain walls

KW - piezoresponse force microscopy

UR - http://www.scopus.com/inward/record.url?scp=85137343656&partnerID=8YFLogxK

U2 - 10.1002/advs.202203957

DO - 10.1002/advs.202203957

M3 - Article

C2 - 36065001

AN - SCOPUS:85137343656

SN - 2198-3844

VL - 9

JO - Advanced Science

JF - Advanced Science

IS - 31

M1 - 2203957

ER -

Exploring Physics of Ferroelectric Domain Walls in Real Time: Deep Learning Enabled Scanning Probe Microscopy

Abstract

Funding

Keywords

Access to Document

Other files and links

Fingerprint

Cite this