Posted in: information

“Ai + synchrotron radiation CT imaging technology” helps the application research of lithium battery materials

With the wide popularity of lithium-ion batteries in digital products and electric vehicles, how to further improve the performance of batteries has become the focus of attention. Among them, the mechanical performance degradation of lithium-ion batteries caused by battery particle cracks is one of the bottlenecks affecting battery performance, so how to avoid the generation of battery particle cracks has become a hot spot of scientists’ attention. Recently,4W1A imaging experimental station of Beijing Synchrotron Radiation FacilityAndResearch group of researcher Liu Yijin of Stanford synchrotron radiation light source in the United Statesas well asEuropean Synchrotron radiation source Peter cloetens research groupThe crack generation mechanism of the cathode material of commercial 18650 battery was studied by using deep learning technology and synchrotron radiation nano resolution CT imaging technology.  

研究人员首先通过纳米分辨X射线CT成像技术对商用18650型电池正极中的数万颗电极颗粒进行了三维成像。为了对电池的破损程度进行定量分析,研究人员提出了一种基于深度学习的图像分割算法,该算法克服了传统的基于图像灰度的图像分割算法的不足,实现了对数万颗电池颗粒裂纹信息的自动检测and提取(图1)。研究人员通过对“海量”电池裂纹数据的量化统计And分析,研究了电池能量密度And颗粒碎裂之间的关系as well as锂电池中电极的碎裂程度And颗粒的堆积密度之间的相互关系(图2)。研究结果表明为了减少电池颗粒的裂纹,应考虑不同深度颗粒堆积的自适应策略,即对于靠近隔膜的一侧,可以使用较大的颗粒堆积密度来减少整体颗粒裂纹。 

上述研究成果将对整个电极材料在深度方向的结构化梯度设计as well as电池材料性能的改善起到重要的指导作用。相关工作发表在Advanced Functional Materials(。Fu Tianyu (Institute of high energy, Chinese Academy of Sciences)andFederico Monaco (European Synchrotron radiation source)He is the co first author of the paper.

Fig. 1 Comparison and analysis of crack detection results in battery cathode materials

(a) 实验获取的电池颗粒的断层切片图像(b)断层切片图像(a)中选取的三个选定区域,(c)And(d)依次为深度学习方法and传统的基于图像灰度的分割方法获得的相应裂纹检测结果


(a)(d)靠近隔膜断层的破损程度and颗粒密度图。(b)(e)靠近集流体的破损程度and颗粒密度图。(c) 整个电池样品的破损程度and颗粒密度的相关图。白色拟合线表示总体趋势。(f) 逐层绘制了颗粒密度and损伤密度之间的关系