【作者】 苦史伟；

【导师】 叶庆好； U.Rau；

【作者基本信息】 上海交通大学 ， 光学工程， 2012， 博士

【摘要】 薄膜硅太阳电池单片内部通常是由许多小的子电池片串联构成,单片内部串联是通过划线步骤和膜层沉积步骤交替进行而得到。通常需要三个划线步骤：即所谓的P1、P2和P3划线来形成电池内部的串联结构,激光刻蚀则通常被用来完成划线工作。一般情况下,激光束从衬底(透明衬底)面入射将目标薄膜材料层选择性地去除。然而,对于使用不透明衬底的太阳电池组件却无法使用激光束透过衬底层去除薄膜层的划线方法,同时对于一些使用透明塑料衬底的电池,由于塑料衬底通常熔点较低,也不适合使用激光束从衬底面入射的划线方法。在这些情况下,激光束必须从薄膜面入射划线,但是膜面入射很难得到高电学性能激光线。首先,激光束从薄膜面入射很难获得边缘陡峭的激光线;再者,激光对材料的消融深度不容易控制,因为仅由热效应把材料去除降低了能量的利用效率,而不能有效去除材料的那部份能量以热的形式注入材料中,往往会改变材料的性能；另外,材料蒸汽对后续激光会产生屏蔽作用。截至目前,对于薄膜面入射划线的文献报道,大多数是关于激光线的结构特性的研究,或是使用了机械划线或印刷技术,这样增加了划线区域造成的死区或使工艺变得繁琐。而本文则更重点研究了激光线的电学性能,这也是与子电池串联结构直接相关的激光线性能；进一步地,做了激光线的电学性能及其结构特性的相关性研究。本文将对三个划线步骤做具体研究,划线选用波长为355nm的纳秒紫外脉冲激光从薄膜面入射,选择性地去除材料,而选择紫外激光是由于它在需要去除的各层材料中的吸收系数都很高。首先,本工作研究了在几种Glass/TCO结构上的P1划线。常规P1划线用红外光从玻璃面入射,透过玻璃将TCO层去除。研究结果表明,使用紫外光时材料的消融阈值很小,同时指出这是由于紫外光在TCO中的吸收系数高,这也是紫外光划的线表面熔融痕迹和再沉积现象都不明显的原因。在研究了P1激光线的结构特性后,对与电池串联相关的电学特性做了研究。由于P1线是为了把相邻的TCO分隔开,接下来研究了在P1划线后的衬底上沉积了薄膜电池中常用的导电性很强的μc-Si：H p-层后的漏电情况,结果证明了使用紫外光从薄膜面入射可以获得高质量的激光线,并展示了膜面P1划线在组件中能得到与标准P1划线相接近的JV特性。P2线是把吸收层选择性地去除,从而在相邻子电池的前后电极之间形成一个导电通道,最重要的是,在前后电极之间形成一个很小的欧姆接触。文中研究了纳秒紫外激光从薄膜面入射在Glass/TCO/α-Si:Ⅱ样品上划线的接触电阻,并用传输线测量方法研究了不同的膜面划线的方法,研究结果表明,只有在双次划线时才可以得到很小的接触电阻值,同时也发现,这种划线方法受两次划线光斑的重叠程度的影响很大。进一步的SEM,SIMS和XPS分析结果表明：激光划线过程中,蒸发的材料再次以SiO2的形式沉积在激光线上,从而导致接触电阻比较大。由于将两次划线的光斑完全重叠并不容易实现,为了消除激光线对其依赖,我们提出了改进激光线的方法,并应用于组件中。最后,本工作对于在α-Si:Ⅱ/μc-Si:Ⅱ电池上用紫外光从薄膜面入射对P3背电极划线做了深入研究。P3划线是3步划线中最后一步,也是非常关键的一步。为了研究P3线的电学性能,我们独创性地设计出了一套表征其电学性能的方法,它可以区分出由背电极残留以及P3线侧壁引起的漏电流,同时可以监测P3划线对TCO的电导率影响。通过系统地改变P3划线的激光参数,表征P3激光线的电学性能以及近一步的SEM分析结果表明,该分析方法可以区分不同通道引起的漏电流：当背电极有残留时,会有很大的漏电流直接通过P3线。而最优的P3线出现在背电极完全被去除,Si层则只有部分被去除的情况下。当Si层完全被去除后,漏电流会开始增加,这源于P3线侧壁的漏电流。文中我们获得的最优P3线的漏电流非常小,在偏压1V时只有1.5μAcm-1(单位长度激光线的电流),而标准工艺P3划线得到的漏电流要大约一个数量级。最后,我们展示了薄膜面划线的组件,即3步划线(P1、P2和P3)分别使用了本文研究的纳米紫外激光从薄膜面入射划线的研究结果,用全膜面入射划线的组件显示了与标准工艺划线的组件相近的JV特性。更多还原

【Abstract】 Thin-film silicon solar modules are usually monolithically series connected.To achieve monolithic interconnection the deposition processes are alternatedwith patterning processes. For the patterning of thin-film modules laser scribingis often used. In the commonly used laser scribing methods for thin-film siliconsolar modules the laser beam is incident through the (transparent) substrate toremove the films. However, for modules with opaque substrates the thin-filmlayers cannot be removed with a laser beam incident through the substrate. Alsofor some transparent plastic substrates a substrate incidence is undesirable. Inthese cases a laser beam incident directly from the iflm side is required. The laserscribing of thin-film solar modules from the iflm side faces several challenges likediiffculties of producing abrupt edges of scribing grooves, and plume shielding.The monolithic series connection is commonly achieved with three laser pattern?ing processes, PI, P2, and P3. In this work we will investigate the patterning ofeach of these three patterning processes from the iflm side with a nano-secondpulsed ultraviolet (UV) laser with a wavelength of355nm. This laser was chosenbecause light of this wavelength is strongly absorbed in all layers that need to beablated for the series connection.Most work on film-side scribing of thin-film solar cells focused on the struc?tural properties of the lines. However, in this work we also investigated therelevant electrical properties required for the series connection, and put these results in relation to the structural investigations.We investigated P1laser lines for various transparent conductive oxides (TCOs) on glass. The standard process for the PI was with an infrared (IR) laser. In this work we demonstrated that the ablation threshold for the UV laser is much lower. Furthermore, we show that the difference in ablation threshold originates in the higher absorption coefficient for the UV laser, which results in a cleaner ablation with less sign of molten and re-solidified material on the sur-face of the lines. In addition to these structural investigations we examined the electrical properties of the PI line which are relevant to the monolithic series con-nection. The PI line is an isolation line, meaning that the line must isolate two adjacent parts of the TCO electrode. In order to investigate the leakage currents induced by the P1line we deposited highly conductive μc-Si:H p-layers as they are used in state of the art thin-film silicon solar cells. From this we demonstrate that with the film-side processes we were able to achieve equally good P1lines as with conventional glass side processes. We demonstrated the high quality film-side P1lines by comparing mini-modules prepared with conventional and film-side P1lines.The function of the P2lines is to achieve a low-ohmic contact between the fron and back electrodes. We investigated the contact resistance of P2lines in a-Si:H layers from the film side with a ns pulsed UV laser for thin-film solar modules. We compared the contact resistance for several scribing methods and find that a low contact resistance is only achieved for double scribing methods (i.e. scribing the same line twice). Furthermore, we find that for such double scribing methods the alignment between the laser spots of the first and second sub-scribes is critical for good-quality contacts. In order to analyze these results in more detail, we examined the morphology and chemical composition at the surface of the laser lines using scanning electron microscopy, X-ray photoelec-tron spectroscopy and secondary ion mass spectrometry. From this analysis, we conclude that a good alignment between the first and second scribes results in less re-deposition of silicon in the form of SiO2on the surface, which explains differences in contact resistance found for the various scribing methods. As a good alignment between the two sub-scribes is difficult to obtain, these double scribing methods are not attractive for industrial application. We developed a new scribing method for which alignment between the two scribes is not critical, and demonstrate that we can obtain high-quality contacts with this method.We investigated laser scribing of P3isolation lines for a-Si:Ⅱ/μc-Si:H solar modules from the film side. In order to study the P3line electrically we developed a method characterize the relevant electrical properties. With our new method we can distinguish between leakage currents directly across the P3line, leakage currents along the laser line side walls and monitor changes in conductivity of the TCO underneath the line. We systematically varied scribing parameters. The properties of the P3lines where then electrically analyzed with the devel-oped method. Furthermore, the P3lines were analyzed using Scanning Electron Microscopy images. As with our characterization method we can distinguish be-tween various paths for the leakage current we could observe that a high leakage current directly across the P3lines is present when the back contact is not fully removed. The best P3lines have the back contact fully removed but an incom-plete removal of the silicon in the line. When the silicon layers are fully removed we observe an increase in the leakage current which we could ascribe to currents flowing along the sidewalls of the laser lines. The best P3laser line we obtained exhibits a very low leakage current density of only1.5μAcm-1at1V (current per unit laser line length). Finally, we demonstrate all three patterning steps from the film-side in a thin-film mini-module. The film side mini-module has an equal performance compared to a mini-module processed with well developed substrate side patterning processes.更多还原