Machine learning-driven exploratory syntheses in molten salts of copper-based compounds for electrocatalytic reduction of carbon dioxide

This proposal aims at discovering new electrocatalyst materials by probing in situ chemical reactions in inorganic liquids, ie. molten salts. These reactions will be driven by machine learning to identify a priori the composition ranges most likely to provide new compounds of interest for the electrocatalytic valorization of CO2. Molten salts enable to trigger reactions at 300-1000 °C in conditions prone to yield metastable phases, hence new materials compared to traditional synthesis methods. We want to perform in situ time resolved X-ray diffraction and scattering (PDF analysis) in order to identify the reaction intermediates, including amorphous phases, that will form during the reactions. The reaction conditions identified in situ will then be used in our laboratory to isolate these intermediates, which will deliver new materials for electrocatalysis.

本研究提案旨在通过探测无机液体（即熔盐）中的原位（in situ）化学反应，发现新型电催化剂材料。研究将借助机器学习驱动上述反应，先验地确定最有可能生成适用于CO₂电催化增值转化的目标新化合物的组分范围。熔盐可在300~1000℃的反应条件下触发化学反应，这类环境易于生成亚稳相，因此相较于传统合成方法，能够获得新型材料。我们将开展原位时间分辨X射线衍射与散射测试（含对分布函数（Pair Distribution Function，PDF）分析），以识别反应过程中形成的反应中间体，包括非晶相。后续我们将基于原位表征得到的反应条件，在本实验室中分离这些中间体，最终获得可用于电催化领域的新型材料。