Carbon atoms can form many different structures due to their unique and changeable bonding methods. Organic matters contained carbon constitute countless life on Earth and is one of the most basic life elements. With the discovery and synthesis of fullerenes, nanotubes, graphene and other materials, carbon-based materials have also attracted wide attention. Among them, sp³ hybrid carbon materials, with high thermal conductivity, toughness, Young's modulus and wide band gap properties, are expected to be used in high-strength carbon fiber materials, sensors and photoelectric devices and other fields.
Experimentally, through top-down synthesis, a series of sp³ carbon materials can be obtained, but it is difficult to obtain controllable ordered materials with atomic scale. The bottom-up synthesis method starting from small molecules is also an important idea. Under pressure, small molecules can form sp³ carbon materials through pressure-induced polymerization. However, due to the complex and diverse reactions of aromatic compounds, it is difficult to accurately synthesize a single structure from the bottom up. In order to obtain sp³ carbon materials with excellent properties using aromatic compounds as substrates, we need to quickly and efficiently screen the aromatic molecular precursors to help us find suitable precursors according to the properties and functions of the products. In the experiment, due to time and labor cost, it is difficult to efficiently and accurately synthesize expected carbon materials. However, through theoretical calculation and machine learning, preliminary judgment and predictive screening can effectively reduce experimental costs and make accurate synthesis feasible and effective.
This project will explore the mechanism and general rule of high-pressure induced polymerization of aromatic compounds at molecular scale by combining first-principle calculation, molecular dynamics, transition state theory, machine learning, structure prediction and other methods as well as existing experimental basis. Then, based on multiple databases, organic structures under atmospheric pressure will be screened and a structural database under high pressure will be established. The crystal structure of the product was predicted by the reaction rule.
 

高压,理论预测,机器学习