【作者】 陈官璧；

【导师】 杨德仁； 阙端麟；

【作者基本信息】 浙江大学 ， 材料物理与化学， 2010， 博士

【摘要】 太阳能光伏作为一种无污染、寿命长且不受地理限制的绿色能源,被认为是解决未来能源问题的有效途径之一。近年来,世界光伏市场迅猛发展,但是,光伏发电的成本相对还比较高,因此降低发电成本就成为关键。由于薄膜电池技术具有原材料消耗少、能耗低、可大面积快速生产等优点,成为国际学术界和产业界研究的热点领域。CuInS2以其接近太阳光最佳匹配的禁带宽度(1.5 eV)、较大的吸收系数、较高的理论转换效率等优点,是一种非常有潜力的薄膜电池材料。但是,目前CuInS2薄膜制备技术多采用真空方法,制造成本相对较高。本文探索了应用金属颗粒墨水制备CuInS2薄膜及太阳电池的方法。首先采用多元醇还原法制备铜、铟及铜铟合金金属颗粒,然后进行金属颗粒墨水、前驱体薄膜的制备和硫化过程等工艺的研究,制备了CuInS2薄膜太阳电池。全文取得了如下具有创新意义的结果：(1)采用多元醇还原法制备了铜、铟及铜铟合金纳米颗粒。研究发现：反应温度和反应速度是制备铟颗粒的主要因素,NaBH4在溶剂中的分解速度决定了生长时间,高分子添加剂能起到空间位阻作用但也会影响到颗粒的均匀性。基于以上的认识,采用三乙醇胺(TEA)作为添加剂,结合快速热注入法,在140℃高温下制备了直径为10nm甚至更小的均匀的铟颗粒,反应温度及TEA的量会影响到颗粒的尺寸。研究还指出：快速注入能一次形成大量形核中心并迅速长大,高温条件可以加速了NaBH4的分解,缩短生长时间,使得还原性环境被破坏,导致铟颗粒被钝化从而限制其进一步生长。另外,采用快速热注入法也制备得到颗粒尺寸在100nm以下、分布较窄的铜铟合金颗粒,同铟颗粒的制备类似,温度和时间对颗粒的形貌没有影响,但会提高结晶性。(2)研究了金属颗粒的高温稳定性。实验指出：铜铟颗粒及铟颗粒即便是在充满还原性气氛的高真空CVD腔体中氧化也比较严重,说明制备得到的金属颗粒在高温条件下不稳定。通过形貌、结构分析,发现铜铟颗粒在高温热处理后转变为氧化铟纳米线,而铟颗粒则有向氧化铟空心球转变的趋势。对比在玻璃衬底及铜衬底上的铟颗粒经过热处理后的形貌差异,认为铜在氧化铟纳米线的生长过程中起到了催化作用。(3)利用制备的金属颗粒制备了前驱体薄膜,进而硫化烧结成CuInS2薄膜。利用CBD法制备的多孔的硫化铜或硫化铟薄膜进行补铟或补铜,最后硫化制备成致密的CuInS2薄膜。采用硫化铜补铟法制备的CuInS2薄膜制备的电池得到了2%的光电转换效率。应用铜颗粒墨水、铟颗粒墨水依次制备了铜/铟或铟/铜叠层前驱体薄膜,用铜铟颗粒制备了混合前驱体薄膜。通过硫化均能制备结构致密、晶粒较大的CuInS2薄膜。发现混合前驱体薄膜制备的CuInS2薄膜杂相相对较少,该薄膜制备的电池具有0.7%的光电转换效率。(4)研究了铜铟合金颗粒的硫化过程,指出杂质Na、中间相等对薄膜性能和结构的影响。发现随着在金属前驱体膜中铜/铟比的增大,CuInS2薄膜晶粒越大、结晶性越好。钠对富铟前驱体薄膜的作用要比富铜薄膜大：铜/铟比为0.6的前驱体薄膜,以含钠的玻璃为衬底,硫化后杂相为Na2In2S4；以不含钠的钼片为衬底时,其杂相为CuIn5S8。而在铜铟比在0.9和1.3的前驱体膜中,无论是衬底中的钠还是在铜铟合金中掺入的钠,都对其结构没有大的影响。研究还发现：在硫化过程中有CuIn5S8中间相参与反应,这是由于铜、铟的扩散速度及与硫反应的速度不同导致的富铜相(CuS)与富铟相(CuIn5S8)的暂时分相,在温度升高时两者作用生成CuInS2。另外,制备的CuInS2薄膜截面均为双层结构：上层薄膜由大晶粒组成,而下层薄膜的晶粒较小。这可能是由铜铟颗粒中的活性部分和非活性部分形成的。(5)成功利用金属颗粒墨水制备了CuInS2薄膜电池,并探讨了其中一些关键步骤对电池性能的影响。研究表明：合适厚度的CdS缓冲层及i-ZnO阻挡层能降低器件的暗电流,提高电池的开路电压；经过200℃的空气退火能有效提高器件性能,而光浴则对电池的开压有少许提高。基于以上对电池器件的认识,采用铜铟颗粒墨水法,制备出了转换效率为1.43%的CuInS2薄膜电池。更多还原

【Abstract】 Photovoltaic technology is considered as one of effective approaches to solve the energy problems, for it is a green energy with little pollution, long service time and no geographical limitations. Recent years have witnessed the booming development of world PV market, steady increment of installed capacity and vigorous stage of PV industry. However, there’s no doult that the cost of PV power is relatively so high that policy subsidies are needed to support it. Cost cutting plays the key role in the large-scale commercial application of PV power.It depends on technological innovations and developments to cut the cost. The thin film soalr cell technologies provide this potential. Compared with custom technologies for crystal silicon solar cells, the thin film soalr cell technologies have the following advantages:low consumption of both raw materials and manufacturing energies, rapid production on large scale, etc. Among the familiar materials for thin film solar cells, CuInS2 is endowed with nearly optimal bandgap (1.5 eV) for solar spectrum, high absorption coefficient and theoretical light-to-electricity conversion efficiency. Therefore, it can become a very promising material for thin film solar cells. Unfortunately, preparation methods for CuInS2 films are mostly based on vacuum, which certainly will increase the manufacturing cost.This thesis researched on the preparation of CuInS2 films and soalr cells based on metal inks. Firstly Cu, In and Cu-In nanoparticles were synthesized by the polyol-reduction method. Then metal inks and precursor films were fabricated from these nanoparticles. Sulfurization processes were investigated and CuInS2 film solar cells were prepared finally. The primary significant results were summarized as follows:(1) Cu, In and Cu-In nanoparticles were synthesized by the polyol-reduction method. It was found that the reaction temperature and rate are two dominant points in the synthesis of In nanoparticles, the decomposition rate of NaBH4 determined the growth time, and polymer additives acted as steric hindrance and influence the uniformity of In nanoparticles as well. Based on thes knowledge, In nanoparticle with diameter of 10 nm or even smaller could be synthesized at 140℃by combining the hot-injection method and using triethanolamine (TEA) as additives. The particle size was controlled by the reaction temperature and the amount of TEA. The hot-injection method were able to form numerous nuclei simultaneously and growth rapidly. However, the high reaction temperature could speed up the decomposition of NaBH4, shorten the growth time, disrupt the reductive environment, lead to the passivation of In nanoparticles and limit their further growth. By employing the similar hot-injection method, Cu-In nanoparticles with size of below 100 nm and narrow distributions could be synthesized. The reaction temperature and time had little effects on the morphology of Cu-In nanoparticles, but could enhance their crystallinty.(2) The Cu-In and In nanoparticles were ready to be oxidized enven in high vacuum CVD chamber filled with H2, indicating their instability at high temperature. By morphological and structural investigation, after high temperature annealing the Cu-In nanoparticles were turned to In2O3 nanowires, whereas the In nanoparticles were inclined to hollow spheres. By comparing the morphological differences of the products of In nanoprticle on glass and Cu substrates after annealing, it was proved that Cu acted as catalyst during the growth of In2O3 nanowires.(3) Metal nanoparticles were used to prepare precursor films and sulfurized to form CuInS2 films. By using the porous CuxS and InxSy films by Chemical Bath deposition (CBD), In or Cu nanoparticles were supplemented on copper or indium sulfide films respectively, then sulfurized to form compact CuInS2 films. CuInS2 film solar cells with an efficiency of about 2% were fabricated on CuInS2 films by In supplement on CuxS films. Metal precursor films with Cu/In or In/Cu stacking layers were prepared with Cu and In nanoparticle inks, those with Cu-In mixing layers were directly prepared with Cu-In nanoparticles. After sulfurization, these two kinds of precursor films were transformed to compact CuInS2 films with large grains. CuInS2 films from mixing layers had fewer impurity phases and solar cells from these films acquired an efficiency of 0.7%.(4) The sulfurization processes of Cu-In nanoparticles were investigated. With the increasing of Cu/In ratios in metal precursor films, the grain sizes of CuInS2 films were larger and crystallinity better. Na played a more significant role in the In-rich precursor films than in Cu-rich precursor films:For precursor films with a Cu/In ratio of 0.6, when they were sulfurized on Na-containing glass substrates, the impurity phases were Na2In2S4, on Mo substrates without Na, the impurity phases were CuIn5S8; for precursor films with Cu/In ratios of 0.9 and 1.3, Na had little influences on the structures of the sulfurized films, no matter Na were from the substrates or doped in metal inks. The intermediate phase of CuIn5S8 took part in the sulfurization process. This was due to the different diffusion rates and sulfurization rates of Cu and In, which resulting in the temporary phase segregation of Cu-rich phase (CuS) and In-rich phase (CuIn5S8). However, CuS and CuIn5S8 phases could react with each other to form CuInS2 phase. The cross section image of CuInS2 films were displayed as two-layer structures:the upper layer with large grains and bottom layer with small grains, which were probably sulfurized from active and inactive Cu-In nanparticles respectively.(5) CuInS2 thin film solar cells were fabricated and the effects of some key processes on the performances of solar cells were investigated. Appropriate thickness of CdS buffer layer and i-ZnO barrier layer could suppress dark current and increase open voltage of the devices. Device performances were able to be effectively improved by a 200℃air-annealing. The open voltage could be increased a bit with light-soaking. Based on the above-mentioned knowledge on devices, CuInS2 thin film solar cells with an efficiency of 1.43% were fabricated based on Cu-In metal inks.更多还原