月別: 2019年10月

[Published online Journal of Computer Chemistry, Japan Vol.18, A6-A11, by J-STAGE]
<Title:> 相対論的量子化学計算プログラムRAQETの公開
<Author(s):> 五十幡　康弘, 吉川　武司, 中井　浩巳
<Corresponding author E-Mill:> nakai(at)waseda.jp
<Abstract:> The “Relativistic And Quantum Electronic Theory (RAQET)” program was released for academic use. One-component (spin-free) and two-component (spin-dependent) relativistic calculations are possible based on the infinite-order Douglas-Kroll-Hess transformation and local unitary transformation. This review describes available features, registration/installation, and input/output of the RAQET program.
<Keywords:>
<URL:> https://www.jstage.jst.go.jp/article/jccj/18/3/18_2019-0022/_article/-char/ja/

[Published online Journal of Computer Chemistry, Japan Vol.18, 123-125, by J-STAGE]
<Title:> 機械学習によるテキスト処理ツールの開発と応用
<Author(s):> 亘理　結香, 奥脇　弘次, 望月　祐志
<Corresponding author E-Mill:> fullmoon(at)rikkyo.ac.jp
<Abstract:> We have developed a text processing tool consisting of several machine learning-related programs such as TreeTagger and Doc2Vec. This system was applied to a number of abstracts of a couple of physical chemistry journals (Chemical Physics Letters and American Journal of Physics). The degree of similarity of contents was then made evaluatable as a guide for papers to be checked for a certain topic.
<Keywords:> Text processing, Machine learning, Doc2Vec, Degree of similarity
<URL:> https://www.jstage.jst.go.jp/article/jccj/18/3/18_2019-0012/_article/-char/ja/

[Published online Journal of Computer Chemistry, Japan Vol.18, 129-131, by J-STAGE]
<Title:> ABINIT-MP Openシリーズの最新の開発状況について
<Author(s):> 望月　祐志, 秋永　宜伸, 坂倉　耕太, 渡邊　啓正, 加藤　幸一郎, 渡辺　尚貴, 奥脇　弘次, 中野　達也, 福澤　薫
<Corresponding author E-Mill:> fullmoon(at)rikkyo.ac.jp
<Abstract:> We have been developing an original program named ABINIT-MP for fragment molecular orbital (FMO) calculations. Various functionalities for interaction analyses, e.g. PIEDA (Pair Interaction Energy Decomposition Analysis), have been augmented in Revision 10 and Revision 15 of the ABINIT-MP Open Series Version 1. In this Letter, we briefly describe such new features.
<Keywords:> キーワード：Fragment molecular orbital, FMO, ABINIT-MP, Interaction analysis, Correlation energy correction
<URL:> https://www.jstage.jst.go.jp/article/jccj/18/3/18_2019-0016/_article/-char/ja/

[Published online Journal of Computer Chemistry, Japan Vol.18, 126-128, by J-STAGE]
<Title:> Scratchを経由する機械学習教材の開発の試み
<Author(s):> 満野　仁美, 奥脇　弘次, 伊藤　雅仁, 望月　祐志
<Corresponding author E-Mill:> fullmoon(at)rikkyo.ac.jp
<Abstract:> We have been developing a Scratch-based educational program for machine learning in the chemistry context. This system has an interface with the DeepChem library set of Python. The solubility prediction of several molecules with SMILES notation was demonstrated as a preliminary application.
<Keywords:> キーワード：Scratch, Educational program, Machine learning, DeepChem, Solubility prediction
<URL:> https://www.jstage.jst.go.jp/article/jccj/18/3/18_2019-0013/_article/-char/ja/

[Published online Journal of Computer Chemistry, Japan Vol.18, 132-135, by J-STAGE]
<Title:> 転移学習と生成ネットワークの試行事例
<Author(s):> 伊藤　雅仁, 篠嶋　友也, 望月　祐志, 秋永　宜伸, 小杉　範仁
<Corresponding author E-Mill:> fullmoon(at)rikkyo.ac.jp
<Abstract:> Transfer learning has attracted interests because of the reduction of training costs. We applied this technique to the analyses for visualized results of computational fluid dynamics (CFD) simulations on 2-dimensional wing models. The accuracy and cost reduction were addressed. Preliminary studies of generative network have been made as well.
<Keywords:> Transfer learning, Generative network, Keras, 2-dimensional wing model, Computational fluid dynamics
<URL:> https://www.jstage.jst.go.jp/article/jccj/18/3/18_2019-0023/_article/-char/ja/

[Published online Journal of Computer Chemistry, Japan Vol.18, 136-138, by J-STAGE]
<Title:> カチオン性イリジウム触媒を用いた均一系触媒反応における相対論効果
<Author(s):> 髙島　千波, 五十幡　康弘, 栗田　久樹, 高野　秀明, 柴田　高範, 中井　浩巳
<Corresponding author E-Mill:> nakai(at)waseda.jp
<Abstract:> A cationic Ir complex was reported to show specific catalytic activity in C-H bond activation reaction of benzanilides. The present study examined reaction energy profiles of the C-H bond activation with Ir and Rh catalysts based on non-relativistic and relativistic quantum chemical calculations. We found that the relativistic effect is essential to demonstrate the difference in the catalytic activity. In particular, the activation of the d orbital of Ir, which is caused by the s- and p- orbital contraction followed by the self-consistent d-orbital expansion, leads to stabilization of the transition state and product of the C-H bond activation.
<Keywords:> C-H Bond activation, Transition metal catalyst, Relativistic effect, Oxidative addition, Iridium
<URL:> https://www.jstage.jst.go.jp/article/jccj/18/3/18_2019-0021/_article/-char/ja/