logo
ResearchBunny Logo
Teaching a neural network to attach and detach electrons from molecules

Chemistry

Teaching a neural network to attach and detach electrons from molecules

R. Zubatyuk, J. S. Smith, et al.

This groundbreaking research introduces an enhanced machine learning framework, AIMNet-NSE, for accurately simulating open-shell anions and cations without the need for quantum mechanical calculations. Conducted by Roman Zubatyuk, Justin S. Smith, Benjamin T. Nebgen, Sergei Tretiak, and Olexandr Isayev, it offers insights into ionization potential and electron affinity, paving the way for innovative chemical reactivity modeling.

00:00
00:00
Playback language: English
Abstract
This research proposes an improved machine learning framework for simulating open-shell anions and cations using the AIMNet-NSE (Neural Spin Equilibration) architecture. This architecture predicts molecular energies for various charge and spin multiplicities with high accuracy (around 2–3 kcal/mol error for energy and 0.01e for spin-charges). It bypasses QM calculations to derive ionization potential, electron affinity, and conceptual DFT quantities, enabling chemical reactivity modeling.
Publisher
Nature Communications
Published On
Aug 11, 2021
Authors
Roman Zubatyuk, Justin S. Smith, Benjamin T. Nebgen, Sergei Tretiak, Olexandr Isayev
Tags
machine learning
open-shell ions
AIMNet-NSE
chemical reactivity
ionization potential
electron affinity
DFT quantities
Listen, Learn & Level Up
Over 10,000 hours of research content in 25+ fields, available in 12+ languages.
No more digging through PDFs, just hit play and absorb the world's latest research in your language, on your time.
listen to research audio papers with researchbunny