Density functional theory and machine learning for electrochemical square-scheme prediction: an application to quinone-type molecules relevant to redox flow batteries

Abstract

Computational high-throughput is used to evaluate proton–electron transfer reactions of quinone-type compounds that are potentially useful for energy storage.

A Unified Charge Storage Mechanism to Rationalize the Electrochemical Behavior of Quinone-Based Organic Electrodes in Aqueous Rechargeable Batteries

“…In light of these results and taking into account the fact that many p K a s of reduced n-type organic compounds are quite basic, − it is clear that a large number of works are questionable when they report that the charge storage mechanism of n-type organic electrodes in aqueous Zn/organic batteries is based on either the reversible insertion of Zn 2+ or coinsertion of Zn 2+ and H + . Among the few examples of n-type organic compounds that a priori allow for reversible insertion of multivalent cations are the naphthalene diimide derivatives. , Their theoretical p K a s values calculated by DFT were indeed found to be quite acidic, leading to p K a of QH – /Q 2– < 4.0, which is low enough to support charge compensation by cations such as Ca 2+ or Zn 2+ .…”

Basic guidelines of first-principles calculations for suitable selection of electrochemical Li storage materials: a review

“…There are several investigations with ML and articial intelligence (AI) have been conducted over the last decade. [107][108][109][110][111][112][113] Proton-electron transfer (PET) reactions, crucial in energy storage applications, exhibit varying mechanisms depending on the molecule and pH. Quantum chemical methods compute redox potential (E red. )…”

“…and acidity constant (pK a ) values but are time-consuming. In a recent study, 107 supervised machine learning (ML) models predict PET reactions and analyze molecular space. Using data generated through density functional theory (DFT) calculations, the authors train and test random forest regression models on a dataset with over 8200 organic molecules, each experiencing a two-proton two-electron transfer process, incorporating structural and chemical descriptors.…”