ChemInform Abstract: Selected Topics in ab initio Computational Chemistry in Both Very Small and Very Large Chemical Systems

Ab initio molecular dymamics (AIMD) simulation studies are a direct way to visualize chemical reactions and help elucidate non-statistical dynamics that does not follow the intrinsic reaction coordinate. However, due to the enormous amount of the ab initio energy gradient calculations needed for AIMD, it has been largely restrained to limited sampling and low level of theory (i.e., density functional theory with small basis sets). To overcome this issue, a number of machine learning (ML) methods have been employed to predict the energy gradient of the system of interest. In this manuscript, we outline the theoretical foundations of a novel ML method which trains from a varying set of atomic positions and their energy gradients, called interpolating moving ridge regression (IMRR), and directly predicts the energy gradient of a new set of atomic positions. Several key theoretical findings are presented regarding the inputs used to train IMRR and the predicted energy gradient. A hyperparameter used to guide IMRR is rigorously examined as well. The method is then applied to three bimolecular reactions studied with AIMD, including HBr+ + CO2, H2S + CH, and C4H2 + CH, to demonstrate IMRR’s performance on different chemical systems of different sizes. This manuscript also compares the computational cost of the energy gradient calculation with IMRR vs. ab initio, and the results highlight IMRR as a viable option to greatly increase the efficiency of AIMD.

Download Full-text

Identification of promising chemical systems for the synthesis of new materials structure types: An ab initio minimization data mining approach

Processing and Application of Ceramics ◽

10.2298/pac1301037z ◽

2013 ◽

Vol 7 (1) ◽

pp. 37-41 ◽

Cited By ~ 6

Author(s):

Dejan Zagorac ◽

Johann Schön ◽

Martin Jansen

Keyword(s):

Ab Initio ◽

Theoretical Work ◽

Structure Type ◽

Alkaline Metal ◽

New Materials ◽

Chemical Systems ◽

Data Mining Approach ◽

The Earth ◽

Alkali Metal Halides ◽

Binary Compounds

In this research we performed data exploring for binary compounds with elements from groups V, IV-VI, and III-VII, with the goal to identify chemical systems where the recently proposed ?5-5? structure type might be experimentally accessible. Among others, TlF, SnO, SnS, SnSe, GeS, GeSe, PbO, PbS, ZnO and ZnS, were chosen for the study. For each of these systems, a local optimization on ab initio level with the LDA functional was performed for the 5-5 structure type, plus other experimentally observed and theoretically proposed structure types, for comparison. Afterwards, the results were combined with earlier theoretical work involving the 5-5 structure in the earth alkaline metal oxides and the alkali metal halides. As a result, we suggest the GeSe and the ZnO systems as the most suitable ones for synthesizing the 5-5 structure type.

Download Full-text

The Crystal Structure of Triphendioxazine as Solved by a New Ab Initio Method Utilising High Resolution Powder Diffraction and Computational Chemistry Techniques

Molecular Crystals and Liquid Crystals Science and Technology Section A Molecular Crystals and Liquid Crystals ◽

10.1080/10587259408027183 ◽

1994 ◽

Vol 248 (1) ◽

pp. 277-289 ◽

Cited By ~ 2

Author(s):

P G Fagan ◽

K J Roberts ◽

R Docherty ◽

A P Chorlton ◽

G D Potts ◽

...

Keyword(s):

Crystal Structure ◽

High Resolution ◽

Ab Initio ◽

Computational Chemistry ◽

Powder Diffraction ◽

Ab Initio Method

Download Full-text

Active Machine Learning for Chemical Dynamics Simulations. I. Estimating the Energy Gradient

10.26434/chemrxiv.14371331.v2 ◽

2021 ◽

Author(s):

Kazuumi Fujioka ◽

Rui Sun

Keyword(s):

Machine Learning ◽

Ab Initio ◽

Density Functional ◽

Computational Cost ◽

Basis Sets ◽

Intrinsic Reaction Coordinate ◽

Chemical Systems ◽

Energy Gradient ◽

Gradient Calculation ◽

Dynamics Simulations

Ab initio molecular dymamics (AIMD) simulation studies are a direct way to visualize chemical reactions and help elucidate non-statistical dynamics that does not follow the intrinsic reaction coordinate. However, due to the enormous amount of the ab initio energy gradient calculations needed for AIMD, it has been largely restrained to limited sampling and low level of theory (i.e., density functional theory with small basis sets). To overcome this issue, a number of machine learning (ML) methods have been employed to predict the energy gradient of the system of interest. In this manuscript, we outline the theoretical foundations of a novel ML method which trains from a varying set of atomic positions and their energy gradients, called interpolating moving ridge regression (IMRR), and directly predicts the energy gradient of a new set of atomic positions. Several key theoretical findings are presented regarding the inputs used to train IMRR and the predicted energy gradient. A hyperparameter used to guide IMRR is rigorously examined as well. The method is then applied to three bimolecular reactions studied with AIMD, including HBr+ + CO2, H2S + CH, and C4H2 + CH, to demonstrate IMRR’s performance on different chemical systems of different sizes. This manuscript also compares the computational cost of the energy gradient calculation with IMRR vs. ab initio, and the results highlight IMRR as a viable option to greatly increase the efficiency of AIMD.

Download Full-text

Molecular Modeling of An Analog Of Curcumin Compounds Pentagamavunon-0 (PGV-0) And Pentagamavunon-1 (PGV-1) Through Computational Chemistry Methods Ab-Initio HF/4-31G

Indonesian Journal of Chemical Research ◽

10.20885/ijcr.vol2.iss1.art4 ◽

2017 ◽

Vol 3 (1) ◽

pp. 28-39

Author(s):

Nurcahyo Iman Prakoso ◽

Lukman Hakim ◽

Nuri Hidayati

Keyword(s):

Breast Cancer ◽

Cervical Cancer ◽

Ab Initio ◽

Computational Chemistry ◽

Structure Elucidation ◽

Anticancer Compounds ◽

Activity Modeling ◽

Experimental Geometry ◽

Ab Initio Hf ◽

Cancer Activity

Breast cancer is the second largest number of cancer cases in Indonesia, after cervical cancer. The growth of these cancer cells can be prevented with compounds Pentagamavunon-0 (PGV-0) and Pentagamavunon-1 (PGV-1). This compound is an analog of curcumin compounds that have anti breast cancer activity. Modeling the structure of compound PGV-0 and PGV-1 through computational chemistry methods Ab-initio HF/4-31G could be used to predict the geometry and structure elucidation spectra associated with pharmacological activity such as anticancer compounds theoretically.This research involves modeling the structures and spectra prediction calculation compounds PGV-0 and PGV-1 by computational chemistry methods Ab-initio HF/4-31G, using Gaussian03W. The result using Ab-initio HF/4-31G method then compared with data from experimental geometry and the results of calculations with AM1.The results showed that computational chemistry methods Ab-initio HF/4-31G calculations give better results for modeling the structure compared semiempirik method AM1.

Download Full-text