Bidirectional Non-Filamentary RRAM as an Analog Neuromorphic Synapse, Part I: Al/Mo/Pr<sub>0.7</sub>Ca<sub>0.3</sub>MnO<sub>3</sub> Material Improvements and Device Measurements
We report on material improvements to non-filamentary RRAM devices based on Pr<sub>0.7</sub>Ca<sub>0.3</sub>MnO<sub>3</sub> by introducing an MoOx buffer layer together with a reactive Al electrode, and on device measurements designed to help gauge the performance...
Main Authors: | , , , , , , , |
---|---|
Format: | Article |
Language: | English |
Published: |
IEEE
2018-01-01
|
Series: | IEEE Journal of the Electron Devices Society |
Subjects: | |
Online Access: | https://ieeexplore.ieee.org/document/8168326/ |
id |
doaj-5fc8bb6568104ebfaee6c2b4c7bf4cd8 |
---|---|
record_format |
Article |
spelling |
doaj-5fc8bb6568104ebfaee6c2b4c7bf4cd82021-03-29T18:45:48ZengIEEEIEEE Journal of the Electron Devices Society2168-67342018-01-01614615510.1109/JEDS.2017.27802758168326Bidirectional Non-Filamentary RRAM as an Analog Neuromorphic Synapse, Part I: Al/Mo/Pr<sub>0.7</sub>Ca<sub>0.3</sub>MnO<sub>3</sub> Material Improvements and Device MeasurementsKibong Moon0https://orcid.org/0000-0003-3229-0009Alessandro Fumarola1https://orcid.org/0000-0002-3397-7462Severin Sidler2https://orcid.org/0000-0003-3585-5191Junwoo Jang3https://orcid.org/0000-0001-5773-0192Pritish Narayanan4Robert M. Shelby5Geoffrey W. Burr6https://orcid.org/0000-0001-5717-2549Hyunsang Hwang7https://orcid.org/0000-0003-3229-0009Department of Materials Science and Engineering, Pohang University of Science and Technology, Pohang, South KoreaIBM Research–Almaden, San Jose, CA, USAIBM Research–Almaden, San Jose, CA, USASamsung, Suwon, South KoreaIBM Research–Almaden, San Jose, CA, USAIBM Research–Almaden, San Jose, CA, USAIBM Research–Almaden, San Jose, CA, USADepartment of Materials Science and Engineering, Pohang University of Science and Technology, Pohang, South KoreaWe report on material improvements to non-filamentary RRAM devices based on Pr<sub>0.7</sub>Ca<sub>0.3</sub>MnO<sub>3</sub> by introducing an MoOx buffer layer together with a reactive Al electrode, and on device measurements designed to help gauge the performance of these devices as bidirectional analog synapses for on-chip acceleration of the backpropagation algorithm. Previous Al/PCMO devices exhibited degraded LRS retention due to the low activation energy for oxidation of the Al electrode, and Mo/PCMO devices showed low conductance contrast. To control the redox reaction at the metal/PCMO interface, we introduce a 4-nm interfacial layer of conducting MoOx as an oxygen buffer layer. Due to the controlled redox reaction within this Al/Mo/PCMO device, we observed improvements in both retention and conductance on/off ratio. We confirm bidirectional analog synapse characteristics and measure “jump-tables” suitable for large scale neural network simulations that attempt to capture complex and stochastic device behavior [see companion paper]. Finally, switching energy measurements are shown, illustrating a path for future device research toward smaller devices, shorter pulses and lower programming voltages.https://ieeexplore.ieee.org/document/8168326/Resistive RAMneural network hardwarenonvolatile memory |
collection |
DOAJ |
language |
English |
format |
Article |
sources |
DOAJ |
author |
Kibong Moon Alessandro Fumarola Severin Sidler Junwoo Jang Pritish Narayanan Robert M. Shelby Geoffrey W. Burr Hyunsang Hwang |
spellingShingle |
Kibong Moon Alessandro Fumarola Severin Sidler Junwoo Jang Pritish Narayanan Robert M. Shelby Geoffrey W. Burr Hyunsang Hwang Bidirectional Non-Filamentary RRAM as an Analog Neuromorphic Synapse, Part I: Al/Mo/Pr<sub>0.7</sub>Ca<sub>0.3</sub>MnO<sub>3</sub> Material Improvements and Device Measurements IEEE Journal of the Electron Devices Society Resistive RAM neural network hardware nonvolatile memory |
author_facet |
Kibong Moon Alessandro Fumarola Severin Sidler Junwoo Jang Pritish Narayanan Robert M. Shelby Geoffrey W. Burr Hyunsang Hwang |
author_sort |
Kibong Moon |
title |
Bidirectional Non-Filamentary RRAM as an Analog Neuromorphic Synapse, Part I: Al/Mo/Pr<sub>0.7</sub>Ca<sub>0.3</sub>MnO<sub>3</sub> Material Improvements and Device Measurements |
title_short |
Bidirectional Non-Filamentary RRAM as an Analog Neuromorphic Synapse, Part I: Al/Mo/Pr<sub>0.7</sub>Ca<sub>0.3</sub>MnO<sub>3</sub> Material Improvements and Device Measurements |
title_full |
Bidirectional Non-Filamentary RRAM as an Analog Neuromorphic Synapse, Part I: Al/Mo/Pr<sub>0.7</sub>Ca<sub>0.3</sub>MnO<sub>3</sub> Material Improvements and Device Measurements |
title_fullStr |
Bidirectional Non-Filamentary RRAM as an Analog Neuromorphic Synapse, Part I: Al/Mo/Pr<sub>0.7</sub>Ca<sub>0.3</sub>MnO<sub>3</sub> Material Improvements and Device Measurements |
title_full_unstemmed |
Bidirectional Non-Filamentary RRAM as an Analog Neuromorphic Synapse, Part I: Al/Mo/Pr<sub>0.7</sub>Ca<sub>0.3</sub>MnO<sub>3</sub> Material Improvements and Device Measurements |
title_sort |
bidirectional non-filamentary rram as an analog neuromorphic synapse, part i: al/mo/pr<sub>0.7</sub>ca<sub>0.3</sub>mno<sub>3</sub> material improvements and device measurements |
publisher |
IEEE |
series |
IEEE Journal of the Electron Devices Society |
issn |
2168-6734 |
publishDate |
2018-01-01 |
description |
We report on material improvements to non-filamentary RRAM devices based on Pr<sub>0.7</sub>Ca<sub>0.3</sub>MnO<sub>3</sub> by introducing an MoOx buffer layer together with a reactive Al electrode, and on device measurements designed to help gauge the performance of these devices as bidirectional analog synapses for on-chip acceleration of the backpropagation algorithm. Previous Al/PCMO devices exhibited degraded LRS retention due to the low activation energy for oxidation of the Al electrode, and Mo/PCMO devices showed low conductance contrast. To control the redox reaction at the metal/PCMO interface, we introduce a 4-nm interfacial layer of conducting MoOx as an oxygen buffer layer. Due to the controlled redox reaction within this Al/Mo/PCMO device, we observed improvements in both retention and conductance on/off ratio. We confirm bidirectional analog synapse characteristics and measure “jump-tables” suitable for large scale neural network simulations that attempt to capture complex and stochastic device behavior [see companion paper]. Finally, switching energy measurements are shown, illustrating a path for future device research toward smaller devices, shorter pulses and lower programming voltages. |
topic |
Resistive RAM neural network hardware nonvolatile memory |
url |
https://ieeexplore.ieee.org/document/8168326/ |
work_keys_str_mv |
AT kibongmoon bidirectionalnonfilamentaryrramasananalogneuromorphicsynapsepartialmoprsub07subcasub03submnosub3submaterialimprovementsanddevicemeasurements AT alessandrofumarola bidirectionalnonfilamentaryrramasananalogneuromorphicsynapsepartialmoprsub07subcasub03submnosub3submaterialimprovementsanddevicemeasurements AT severinsidler bidirectionalnonfilamentaryrramasananalogneuromorphicsynapsepartialmoprsub07subcasub03submnosub3submaterialimprovementsanddevicemeasurements AT junwoojang bidirectionalnonfilamentaryrramasananalogneuromorphicsynapsepartialmoprsub07subcasub03submnosub3submaterialimprovementsanddevicemeasurements AT pritishnarayanan bidirectionalnonfilamentaryrramasananalogneuromorphicsynapsepartialmoprsub07subcasub03submnosub3submaterialimprovementsanddevicemeasurements AT robertmshelby bidirectionalnonfilamentaryrramasananalogneuromorphicsynapsepartialmoprsub07subcasub03submnosub3submaterialimprovementsanddevicemeasurements AT geoffreywburr bidirectionalnonfilamentaryrramasananalogneuromorphicsynapsepartialmoprsub07subcasub03submnosub3submaterialimprovementsanddevicemeasurements AT hyunsanghwang bidirectionalnonfilamentaryrramasananalogneuromorphicsynapsepartialmoprsub07subcasub03submnosub3submaterialimprovementsanddevicemeasurements |
_version_ |
1724196486108938240 |