The Characterization Detection Application Of OCT For Crystalline Silicon Solar Cells

Optical coherence tomography (OCT) is a newly developed optical detection technologywhich emerged in the end of20th century. Its axial resolution and detecting depth havereached to the levels of micron and millimeter, respectively. It offers the advantage ofcontact-less and non-destructive detection with high image resolution and acquisition rate.Optical coherence tomography with its excellent performance has been widely used inbiomedical sciences, especially in the examination and treatment of eye diseases. Now, thistechnique is being used in a number of other applications gradually. OCT technology alreadyhas a lot of exploratory researches in materials testing, micro device manufacturing, siliconmaterial and plastic material of MEMS components quality control, polymer equipmentinspection as well as the pearl.In this thesis, OCT technology is attempted to apply in crystalline silicon solar cellquality inspection and to provide a new kind of efficient, real-time and three-dimensionaldetection technology. The following items were studied in this thesis:First we think OCT may be the most promising metrology candidate to succeed in thecrystalline silicon solar cell manufacturing by analyzing and comparing the advantages anddisadvantages of several kinds of microscopic imaging instruments which have already beenwidely applied in crystalline silicon solar cell inspection.In the experiments, a swept-source OCT system was used to image silicon samples anddigital image processing technology was widely used in the process of OCTthree-dimensional image data. The intensity of OCT signal varies dramatically with thedetecting depth which is caused by swept source and galvanometer to lead to blur imaging inswept source optical coherence tomography (SS-OCT) system. This phenomenon can affectthe measurement result of crystalline silicon solar cells. According to this problem, acompensation method used for different detecting depth was advanced in this paper. Takingthe SS-OCT system with multilayer special tape as the sample, the structure image wascompensated by data fitting and more real gray level of the image was obtained to guaranteethe accuracy of crystalline silicon solar cell detection by OCT.Crystalline silicon almost does not absorb light which wavelength is more than1200nm.Currently, OCT mainly works in1300nm and the imaging depth is a few millimeters. Since its200microns thick, crystalline photovoltaic cells can be imaged completely by OCT. OCT isused of backscattering light to image. The roughening textured surface impurities, defects andcracks scatter light strongly, and it can be displayed clearly in OCT3D images. Throughreasonable data analysis and image processing methods, the detection of crystalline siliconsolar cells was completed by OCT.Tested samples involve monocrystalline wafer, polycrystalline wafer and finished cell.The main content of experiment includes two parts. The first one is the test of quality oftexture and the second is the detection of impurities and defects in cells. The main content ofthe quality test of texture is that: Choose the quality consistent silicon wafers in the samebatch which were divided into four groups. In the same experimental conditions, the fourgroups of monocrystalline silicon wafer were textured in different time. Though processing the OCT acquiring date, the best texturing time can be found and the results can be proved bythe reflectance of samples. The main content of detecting impurities and defects in cells is that:Pure internal part of cells cannot be measured by OCT. But impurities, defects and cracks willabsorb or scatter light and they all can be detected by OCT. OCT3D images can be observedin arbitrary depth profile along x, y, z directions, impurities and defects are easily detected, nomatter on surface or inside the samples. Silicon has poor absorption ability at1300nm. Cleanwafer does not have any peaks in OCT image. The tiny impurity which is not cleaned up toleave on the cell surface may absorb the light at1300nm and cause the bright peaks in OCTimage. Comparing the number of absorption peaks in polycrystalline cells with differentefficiency in the same batch can find that the peaks are likely to affect the efficiency of cellsand the lower efficiency cells have more peaks. The experimental results show that OCT caneasily and nondestructively detect defects and impurities of crystalline silicon solar cells.