【作者】 罗翀；

【导师】 孟志国； 熊绍珍；

【作者基本信息】 南开大学 ， 微电子学与固体电子学， 2010， 博士

【摘要】 光伏材料需要兼顾较高的光电转换效率和较低成本的工业化生产的要求,而多晶硅薄膜兼具体硅的优异光电性能及非晶硅薄膜的低成本的优点,而且资源丰富没有毒性,又拥有高的转化效率的潜力；CdTe为直接能隙半导体,与太阳光谱接近理想的匹配,因此有较高的转化效率,且由于CdTe太阳电池结构简单,制备方法多样,也具低成本、便于大面积生产的特点。因此多晶硅薄膜太阳电池和碲化镉薄膜太阳电池很有可能成为未来市场的主导产品。但是,多晶硅薄膜太阳电池和碲化镉薄膜太阳电池某些关键工艺还存在一些问题,使得这两种薄膜太阳电池距离满足廉价光伏材料的目标还有一定差距。本论文就针对多晶硅薄膜太阳电池后氢化处理工艺、铝诱导晶化工艺和碲化镉薄膜太阳电池中的背接触等关键工艺进行研究,试图从材料研究角度,使其满足光伏产业高转换效率和低成本的要求。文中具有创新性的研究包括以下内容：(1)针对SPC法制备的CSG多晶硅薄膜太阳电池,其晶界和晶粒内部常含有的大量的缺陷态,严重影响其制备的器件性能和稳定性,需要进行后氢等离子钝化处理。当前使用该方法的多,但是对其微观机制深入揭示的较少。我们利用OES谱的直接观察将H等离子基元行为与被钝化物质关联起来,并对钝化后材料进行红外光谱、霍尔迁移率、拉曼光谱、吸收谱等手段的测试与详细分析,对其发挥关键作用的后氢化处理的等离子基元的表现及其与工艺(气压,温度,功率等)相结合,考察其微观效果,以期对氢化处理的物理机制进行探讨。发现氢化处理的开始阶段,氢等离子体的钝化作用占主导地位,H有能力进入薄膜内与硅键合,薄膜内的氢多以Si-H键形式存在的时候,其霍尔迁移率得到提高；而过长的氢化时间会使刻蚀和轰击作用更加明显,薄膜内Si-H2的数目明显增多,使薄膜的电学性能下降。氢化与多晶硅材料本身及其缺陷态类型有关。多晶硅材料氢化前所含的缺陷态数目越多,能够钝化的缺陷态就越多,霍尔迁移率提高的幅度就越大,氢化效果越明显。PE-SPC晶化的样品其缺陷态类型主要为晶界等处存在的大量悬挂键所造成的缺陷态,只需较低能量的H。实现钝化；而LP-SPC晶化样品以晶格缺陷为主,需要较高能量的Hβ和Hγ才能有效钝化其缺陷态；LP-MIC晶化样品主要含Ni杂质有关的缺陷态,能量适中的H*对其钝化效果更为明显。利用这样的机理我们优化衬底温度、辉光功率和腔体压强等工艺参数,550℃10W 500 m Torr条件下PE-SPC poly-Si霍尔迁移率提高了84.9%；550℃10W 300m Torr条件下LP-MIC poly-Si霍尔迁移率提高了50.1%；550℃10W1200m Torr条件下LP-SPC poly-Si霍尔迁移率提高了56.7%。(2)针对于使用Al诱导晶化的多晶硅电池,其实际效率低于理论值的主要原因是晶化多晶硅的内部含有残留的金属Al,成为复合中心,从而影响其制备的器件性能和稳定性。为此我们开发了溶液铝诱导晶化技术及氢等离子体铝诱导晶化技术,以期降低晶化工艺存在残留金属Al的浓度：前者利用控制铝溶液的浓度来控制非晶硅表面覆盖的铝含量,达到既能有效晶化又能减少晶化硅中金属残留的目的；后者将铝诱导晶化的退火过程在氢等离子体的气氛中进行来达到降低晶化工艺存在残留金属Al浓度的目的。另外这两种技术都试图以开辟更为廉价的电池制造工艺。前者省去真空过程,简化工艺。后者将铝诱导晶化的退火过程在氢等离子体的气氛中进行,这样可以将传统的退火与后氢化处理工艺合二为一。这样可以简化工艺,降低成本。通过实验发现它还能降低退火时间,克服铝诱导晶化低温时退火时间较长的缺点。利用这两种晶化方法制备多晶硅薄膜都尚属首次。(3)针对CdTe薄膜太阳电池,其背接触问题成为目前实际制备的CdTe太阳电池效率与理论预期值相比有较大差距的主要原因之一。另外,CdCl2后处理在CdTe太阳电池的制备中起着重要作用。对此我们利用碳糊成膜法,将含Cu、Te的CdCl2浆状悬浊液涂覆在CdTe表面,进行一次后退火,将传统的CdCl2后处理和形成CuxTe缓冲层工艺合二为一。实验发现碳糊成膜法能达到使CdTe薄膜结晶、CdS和CdTe之间互溶的目的,实现CdCl2后处理的作用；形成CuxTe缓冲层,达到改善背接触的目的；此外,其制备的CdTe太阳电池含较少的Cu,能提高电池性能。碳糊成膜工艺制备CdTe太阳电池没有使用强酸也不包含真空过程,简单易行,可以较显著地降低成本,适合大面积生产。此种方法制备CdTe太阳电池背接触也尚属首次。更多还原

【Abstract】 Photovoltaics has the potential to become a major source of energy and to have a significant and beneficial effect on the global environment. In order for photovoltaics to realize that potential, PV must become standardized products that are inexpensive, durable, and efficient. Until today only two kinds of material have shown a definite potential as the primary material used for PV power generation:poly-Si thin film and CdTe thin film. Poly-Si thin film has both advantages of amorphous silicon and mono-crystalline silicon. Its low cost of and high mobility and great stability give its possibility of application in large area solar cell. And CdTe has an energy gap of 1.45 eV, very well suited to absorb the solar light spectrum. The energy gap is direct, resulting in an absorption coefficient for visible light of>10-5 cm-1, so that the absorber layer need only be a few micon thick to absorb 90% of light above the band gap. By now numerous low-cost deposition technologies could prepare large area CdTe solar cells.However, there are some limit in preparing poly-si thin film and CdTe thin film solar cells.To begin with, as for CSG poly-si solar cells, poly-Si thin films always have grain boundary defects and intra-grain defects after they crystillized, which severely affect the performances and stabilities of the devices made of such poly-Si. The hydrogen passivation is one of the most effective methods to passivate the defects in poly-Si devices. However, by now, although Hydrogen Passivation was applied abroad, the passivation mechanism was not clear completely yet. In this dissertation, we studied the hydrogen passivation of different poly-Si thin films which were prepared by metal induced crystallization (MIC) or solid phase crystallization (SPC) using different amorphous silicon thin films as precursors. We investigated in-situ optical emission spectroscopy (OES) of hydrogen plasma during passivation, and characterized the corresponding electrical and optical properties of passivated poly-Si thin films. Based on the measurement results of OES and the electrical and optical properties of poly-Si, it was found that different hydrogen plasma radicals acted in different roles in hydrogen passivation. We found that the reaction between hydrogen radicals in plasma and poly-Si depend on the defect type in crystallized poly-Si. For LP-SPC poly-Si, in which the major defects were intra-grain defects, higher energy radicals Hp and Hγwere needed to perform the passivation effectively. For PE-SPC poly-Si, in which the major defects were the dangling bond defects in grain boundaries, low energy radicals Ha can passivate them effectively. For LP-MIC poly-Si, which contains many defects relative to Ni impurity, H* radicals with middle energy were suitable to passivate these kind of defects.We also optimized the hydrogen passivation process. After passivation 20 minutes at reaction pressure of 800 millitorr, RF active power of 10w and substrate temperature of 550℃, the hall mobility of PE-SPC poly-Si increased by 84.9%; At reaction pressure of 300 millitorr, RF active power of 10w and substrate temperature of 550℃, the hall mobility of LP-MIC poly-Si increased by 50.1%; At reaction pressure of 1200 millitorr, RF active power of lOw and substrate temperature of 550℃, the hall mobility of LP-SPC poly-Si increased by 50.1%.In addition, as for AIC poly-si solar cells, poly-Si thin films always have Al impurity as recombination center after crystillization, which affect the performances and stabilities of the devices made of such poly-Si. So we developed solution-based aluminum-induced crystallization technology and aluminum induced crystallization assisted by hydrogen plasma technology which had never been reported. The former technology could control Al content by controling aluminum concentration in salt solution, and the latter put the annealing process of AIC in hydogen plasma to reduce Al impurity. Furthermore, both technology were low-cost. The former technology could aviod vacuum pocess, and the latter could integrate the crystallization and passivation into one process. What’s more, analyzed by Raman spectroscopy, SIMS and Hall mobility, we found that it could not only reduce the annealing time of AIC but also enhance the performance of crystallized poly-Si.Finally, as for CdTe solar cells, the difficulty in making ohmic contact to p-CdTe is one of the greatest obstacles to achieve high efficiency. And maximum performance of CdTe solar cells can often be achieved only after CdCl2 treament. So we developed graphite-forming back contact technology. Analyzed by XRD and SIMS, we found this technology could not only form CuxTe layer but also lead to interdiffusion of CdTe and CdS simultaneously. In addition, this technology could reduce Cu impurity in CdTe, improving the performance of CdTe solar cells. We did’t use acid and vaccum in the process. This simple technology has never been reported before either.更多还原