Direct design of active catalysts for low temperature oxidative coupling of methane via machine learning and data mining,Catalysis Science & Technology

当前位置： X-MOL 学术 › Catal. Sci. Technol. › 论文详情

Our official English website, www.x-mol.net, welcomes your feedback! (Note: you will need to create a separate account there.)

Direct design of active catalysts for low temperature oxidative coupling of methane via machine learning and data mining
Catalysis Science & Technology ( IF 4.4 ) Pub Date : 2020-10-30 , DOI: 10.1039/d0cy01751e
Junya Ohyama _{1,

2,

3,

4} , Takaaki Kinoshita _{2,

3,

5,

6} , Eri Funada _{2,

3,

5,

6} , Hiroshi Yoshida _{1,

2,

3,

4} , Masato Machida _{1,

2,

3,

4} , Shun Nishimura _{7,

8,

9,

10} , Takeaki Uno _{11,

12,

13} , Jun Fujima _{6,

14,

15,

16} , Itsuki Miyazato _{6,

14,

15,

16} , Lauren Takahashi _{6,

14,

15,

16} , Keisuke Takahashi _{6,

14,

15,

16}

Affiliation

Faculty of Advanced Science and Technology
Kumamoto University
Kumamoto
860-8555 Japan
Graduate School of Science and Technology
Japan
Graduate School of Advanced Science and Technology
Japan Advanced Institute of Science and Technology (JAIST)
Nomi
923-1292 Japan
National Institute of Informatics (NII)
Chiyoda-ku
101-8430 Japan
Department of Chemistry
Hokkaido University
Sapporo 060-0815

Direct design of low temperature oxidative coupling of methane (OCM) catalysts is proposed via machine learning and data mining. 58 OCM catalysts are experimentally synthesized and evaluated. The collected 58 sets of data are then classified by unsupervised machine learning in a multi-dimensional space where an active catalyst group for low temperature OCM is identified. Data mining then identifies the physical rule within the group. Catalysts satisfying such a physical rule are designed where 2 undiscovered low temperature OCM catalysts are found and experimentally validated. Thus, machine learning and data mining reveal the hidden physical rule behind the catalysis leading to the direct design of catalysts. Hence, machine learning and data mining open up the insight on a powerful strategy for designing catalysts.

中文翻译：

通过机器学习和数据挖掘直接设计用于甲烷低温氧化偶联的活性催化剂

建议通过以下方法直接设计甲烷（OCM）催化剂的低温氧化偶联机器学习和数据挖掘。实验合成和评估了58种OCM催化剂。然后，通过多维空间中的无监督机器学习对收集到的58组数据进行分类，在多维空间中识别出用于低温OCM的活性催化剂基团。然后，数据挖掘将识别组内的物理规则。设计满足这种物理规则的催化剂，其中发现了2种未发现的低温OCM催化剂并进行了实验验证。因此，机器学习和数据挖掘揭示了催化背后隐藏的物理规则，从而导致了催化剂的直接设计。因此，机器学习和数据挖掘为设计催化剂的强大策略开辟了见识。

更新日期：2020-11-18

点击分享查看原文

点击收藏

阅读更多本刊最新论文本刊介绍/投稿指南11