首届世界科学智能大赛：材料科学赛道——金属有机框架材料的预测合成 初赛数据集

首届世界科学智能大赛：材料科学赛道——金属有机框架材料的预测合成 初赛数据集
MOF材料的预测合成 金属有机框架（Metal-Organic Frameworks，简称MOF）是一类由金属离子（或金属簇）和有机配体通过配位键连接而成的晶体材料。MOF材料具有多孔结构、高度可调性和巨大的比表面积，这使其在吸附、储气、分离、催化等领域具有广泛的应用潜力。预测合成是指通过计算机模拟和机器学习等方法，预测和设计新的MOF材料的合成路线和条件。

MOF材料的结构信息（输入）：MOF是由金属、有机配体、连接方式（拓扑结构）确定，可由CIF（Crystallographic Information File）文件经过算法（我们会提供）给出代表结构信息的输入数据。本次比赛初赛阶段将提供Fingerprint和RAC两种数据，选手需要依据该两种数据分别进行相应的任务预测。

MOF材料的合成条件（输出）：根据合成MOF材料的相应文章进行数据挖掘，得出四种有代表性的材料合成条件：温度（T），时间（t），使用的有机溶剂，添加剂。

The 1st World Science Intelligence Competition: Materials Science Track — Predictive Synthesis of Metal-Organic Framework Materials (Preliminary Contest Dataset)

Predictive Synthesis of MOF Materials
Metal-Organic Frameworks (MOFs) are a class of crystalline materials formed by metal ions (or metal clusters) and organic ligands connected via coordination bonds. MOFs feature porous structures, high tunability, and large specific surface areas, endowing them with broad application potential in fields such as adsorption, gas storage, separation, and catalysis. Predictive synthesis refers to the use of methods like computer simulation and machine learning to predict and design synthetic routes and conditions for new MOF materials.

Structural Information of MOF Materials (Input): The structure of a MOF is determined by its metal components, organic ligands, and connection modes (topology). The input data representing the structural information can be derived from CIF (Crystallographic Information File) files using an algorithm provided by the organizers. Two types of data, Fingerprint and RAC, will be provided during the preliminary contest, and participants are required to conduct corresponding task predictions based on each of these two datasets respectively.

Synthetic Conditions of MOF Materials (Output): Based on data mining of relevant papers on MOF synthesis, four representative synthetic conditions of the materials are obtained: temperature (T), time (t), the organic solvent used, and additives.

搜集汇总

背景与挑战

背景概述

该数据集来自首届世界科学智能大赛的材料科学赛道，专注于金属有机框架材料（MOF）的预测合成。它提供了Fingerprint和RAC两种特征数据，用于训练和测试模型，以预测MOF的合成条件，包括温度、时间、溶剂和添加剂等参数。