材料科学
钝化
光电子学
异质结
硅
覆盖层
带材弯曲
非晶硅
氟化锂
工作职能
肖特基势垒
聚合物太阳能电池
太阳能电池
二极管
晶体硅
能量转换效率
纳米技术
载流子
无定形固体
半导体
等离子太阳电池
肖特基二极管
氟化镁
阳极
锂(药物)
电接点
太阳能
电阻式触摸屏
接触面积
作者
Zhiyuan Xu,Wei Li,Yu Yan,Qianfeng Gao,Taiqiang Wang,Hongyu Dun,Shengzhi Xu,Jian He,Qian Huang,Xiaodan Zhang,Ying Zhao,Guofu Hou
出处
期刊:Small
[Wiley]
日期:2026-01-19
卷期号:22 (15): e13992-e13992
标识
DOI:10.1002/smll.202513992
摘要
ABSTRACT Dopant‐free materials with wide‐bandgap are promising alternatives to n‐type or p‐type amorphous silicon (a‐Si) contacts in silicon heterojunction (SHJ) solar cells. They offering strong carrier selectivity through favorable band alignment, negligible parasitic absorption, and full compatibility with low‐temperature processing. However, implementing such contacts on the front side remains challenging because optical losses, contact resistivity, and gird patterns should be balanced carefully. Here, these constraints are overcome through an interfacial design that integrates a lithium phosphate (Li 3 PO 4 ) electron‐selective layer, a low‐work function magnesium fluoride (MgF 2 ) anti‐reflection coating, and a grid with buried Al/Li 3 PO 4 contact. Upon annealing, the patterned metal grid confines Al diffusion to localized contact regions, while the MgF 2 overlayer promotes downward band bending at the c‐Si interface. Meanwhile, phosphorus self‐diffusion within Li 3 PO 4 reduces the contact resistivity, and the recovery of hydrogen passivation in intrinsic a‐Si:H further suppresses interfacial defect states. These effects yield an interface that combines excellent passivation with low resistive loss. As a result, the devices achieve a power conversion efficiency of 23.81% (V OC = 715.2 mV, J SC = 40.65 mA/cm 2 , FF = 81.9%), representing the highest reported efficiency for SHJ solar cells with a front contact by dopant‐free materials. This work highlights the potential of wide‐bandgap compounds on the illuminated side and provides a viable route toward next‐generation high‐efficiency silicon solar technologies.
科研通智能强力驱动
Strongly Powered by AbleSci AI