China News Service, Beijing, October 5th (Reporter Sun Zifa) Can robots not only become scientific research assistants of scientists, but also scientists?

The young scientific research team of the University of Science and Technology of China (USTC) gave a positive answer through the latest research and development results.

  The latest news from the Chinese Academy of Sciences (CAS) says that under the academy's "Data-Driven Machine Scientists in Chemistry, Materials and Biological Sciences" Young Team Program and the National Natural Science Foundation of China, Luo Yi from the School of Chemistry and Materials Science, University of Science and Technology of China. , Professor Jiang Jun's team cooperated with Shang Weiwei of the Department of Automation, etc., through the development and integration of mobile robots, chemical workstations, intelligent operating systems, and scientific databases, to develop a full-process machine chemist driven by data intelligence, and has initially realized the paradigm of intelligent chemistry. .

Mobile robots and intelligent workstations complete the whole process of high-throughput synthesis, characterization, and testing of chemical experiments.

Photo courtesy of China University of Science and Technology

  The research paper on the "Data Intelligence-Driven Whole-Process Artificial Intelligence Machine Chemist" has been published in the latest issue of the National Science Review (NSR) academic journal.

The international reviewers commented that the "robot system, workstation and intelligent chemical brain are all state-of-the-art" and "will have a huge impact on chemical science."

Industry experts believe that the research work of machine chemists breaks away from the limitations of the traditional trial-and-error research paradigm, showing the great advantages of the new paradigm of intelligence guided by the "strongest chemical brain", leading chemical research to digitize knowledge understanding, operational instruction, and The future trend of creating templates is moving forward, establishing China's global leading position in the field of intelligent chemical innovation.

  According to the research team of the University of Science and Technology of China, the machine chemist platform realizes the development of the whole process of chemical synthesis, characterization and testing driven by big data and intelligent models. Model and an intelligent platform with an open operating system, it has stronger chemical intelligence and extensive chemical development capabilities. It currently covers photocatalytic and electrocatalytic materials, luminescent molecules, optical thin film materials, etc., and the scope of application will be upgraded with the platform. And expansion continues to expand.

The mechanochemist platform realizes the efficient creation of high-entropy non-precious metal catalysts for oxygen evolution reaction.

Photo courtesy of China University of Science and Technology

  The machine chemist platform can use machine intelligence to find and read literature, draw expert experience from massive research data, put forward scientific hypotheses and formulate experimental plans on the basis of previous knowledge and data; dispatch 2 mobile robots and 15 self-developed The intelligent chemical workstation completes the whole process of high-throughput synthesis, characterization, and testing of chemical experiments, and reserves standard interfaces, with scalability; And load the cloud database, which can call and update database information in real time; the unique computing brain integrates the intelligent model into the underlying theoretical laws and complex chemical experiment evolution by invoking physical models, theoretical calculations, machine learning and Bayesian optimization. Machine scientists understand chemistry better and are better at chemical creation.

  According to the popular science interpretation of the research team of the University of Science and Technology of China, the objects of chemical research are increasingly complex and high-dimensional, and the traditional research paradigm mainly relies on the methods of "exhaustive" and "trial and error".

Faced with a huge chemical space, the search for formulations and processes often stops at local optima, and global exploration cannot be carried out.

Taking the high-entropy (highly complex, highly disordered) compound catalysts with great potential as an example, the highly disordered mixing of various elements brings high stability, and it also brings great challenges to finding the optimal ratio by manual experiments.

Obtaining the optimal formulation requires traversing and testing extremely large stoichiometric combinations, which are currently limited to optimizing up to 3 metal combinations.

  The newly developed machine chemist, taking advantage of its data-driven and intelligent optimization, intelligently reads 16,000 papers and independently selects 5 non-precious metal elements, integrates 20,000 sets of theoretical calculation data and 207 sets of full-process machine experimental data, and establishes The intelligent model that blends theory and reality guides the Bayesian optimization program to find the optimal high-entropy catalyst from 550,000 possible metal ratios, shortening the 1,400 years required for the traditional "frying-vegetable" traversal search to 5 weeks.

(Finish)