材料科学
薄脆饼
硅
残余应力
太阳能电池
拉曼光谱
微观结构
压力(语言学)
晶界
复合材料
表征(材料科学)
冶金
光电子学
光学
纳米技术
哲学
物理
语言学
作者
Vera Popovich,J.M. Westra,R.A.C.M.M. van Swaaij,M. Janssen,I.J. Bennett,I. M. Richardson
标识
DOI:10.1109/pvsc.2011.6186276
摘要
Stress in multicrystalline silicon (mc-Si) is a critical issue for the mechanical stability of the material and it has become a problem of growing importance, especially in view of silicon wafer thickness reduction. Without increasing the wafer strength, the high fracture rate during handling and subsequent processing steps leads to excessive losses. A non-uniform stress distribution could be expected in critical areas such as grain boundaries, wire-saw-damaged layer and areas near the metallization and soldered contacts. Therefore, a non-destructive method to locally determine stress in mc-Si solar cells is of technological importance. In this paper stress characterization based on a combination of Raman spectroscopy, electroluminescence imaging, cell bowing measured with a laser scanning device, confocal microscopy and ex-situ bending tests will be presented. The most critical processing steps during the manufacture of screen-printed solar cells are wafer cutting, firing of metallic contacts and the soldering process. In this work the development of mechanical stress in silicon wafers as a result of different processing steps will be evaluated. Furthermore, residual stress and stress developing during silicon cell bending are measured in relation to microstructure and defect density. It was found that there is an inhomogeneous distribution of stress along grain boundaries and metallic inclusions. It was found that at a certain load grain boundaries in mc-Si wafers experience a higher stress (~50 MPa) than grains themselves (~30 MPa). A significant Raman shift was observed in samples with a wire-saw-damaged layer and in the areas close to Ag fingers and Al/Ag bus bars. Raman scanning was also performed along the solar cell cross section with different metallization patterns.
科研通智能强力驱动
Strongly Powered by AbleSci AI