開發與優化氫化非晶矽鍺薄膜太陽電池以及其在多接面太陽電池的應用

Abstract

在本研究中，我們開發並優化氫化非晶矽鍺薄膜、單接面氫化非晶矽鍺太陽電池、氫化非晶矽/氫化非晶矽鍺串疊型太陽電池以及改善太陽電池之光利用率。在非晶矽鍺薄膜的開發方面，我們進行實驗並討論改變氫流率、電漿功率以及製程氣體流率對於薄膜特性以及非晶矽鍺單接面太陽電池之轉換效率之影響。製程氣體為矽甲烷(SiH4)、鍺甲烷(GeH4)以及氫氣，利用射頻電漿輔助化學氣相沉積系統(PECVD)來沉積非晶矽鍺薄膜。本研究發現氫氣流率比對於非晶矽鍺薄膜品質的提升有重要幫助，增加氫氣流率比可降低沉積速率及薄膜之微結構參數(microstructural factor)並提升光電導，這是因為氫可提供能量至薄膜表面並改善原子之排列，近而提升薄膜品質。本研究中，當通入氫氣流率比為2左右時，非晶矽鍺薄膜之微結構參數及電性可被大幅提升。同時我們也發現氣態氫原子有助於提升薄膜中之鍺含量，而薄膜中鍺含量亦影響光學能隙以及電性，因此可提升對於長波長光子的吸收。此外，提升氫氣流率比使得單接面太陽電池之轉換效率由5.73%提升至6.33%。 在單接面太陽電池方面，我們發現非晶矽鍺吸收層的厚度大約在150 nm 至 300 nm 之間為佳，過厚或過薄都會對效率有不良影響。此外，我們研究在p/i 與 i/n 介面間加入能隙漸變層對於太陽電池之影響，本研究發現，與未使用能隙漸變層之太陽電池相比，在p/i間與在i/n間之能隙漸變層能有效提升開路電壓，同時i/n能隙漸變層能大幅提升載子收集效率近而提升FF，而p/i能隙漸變層則有相反之影響，綜合能隙漸變層對於原件效率之影響，較適當之p/i及i/n能隙漸變層厚度分別為20 nm及160 nm，使用能隙漸變層於非晶矽參雜層及非晶矽鍺吸收層間，太陽電池之效率可由5.73%大幅提升至8.59%。 在非晶矽/非晶矽鍺串疊型太陽電池部分，提升光利用率具有相當之重要性，這包含適當分配子電池之電流、降低原件之光吸收損失以及增加元件之入光量，在本研究中，我們開發n型微晶矽氧合金以做為串疊型太陽電池之中間反射層(intermediate reflecting layer)及背反射層(back reflecting layer)，在開發n型微晶矽氧合金方面，增加放電功率可增加薄膜中之氧含量並提升能隙，同時氧含量的增加亦會使得薄膜之導電性下降。將n型微晶矽氧合金應用於太陽電池中做為中間反射層及背反射層可大幅增進元件效率，當作為串疊型太陽電池之中間反射層可增加上電池的吸收進而增進短路電流，元件效率可因此提升12.9%；另一方面，使用n型微晶矽氧合金作為串疊型太陽電池之背反射層可增進長波長光子之吸收，進而增進轉換效率達9.7%，非晶矽/非晶矽鍺串疊型太陽電池效率可達10.03%。 在增加太陽電池之入光量方面，開發適用於非晶矽/非晶矽鍺串疊型太陽電池之光侷限結構是本研究之另一重點，得利於科技部項下計畫：台灣-美國加州理工學院能源跨國研究(Caltech-Taiwan Energy Exchange)之故，我們得以開發並更深入探討光侷限結構對於太陽電池之影響，我們使用奈米壓印技術開發具低成本、大面積、可重複使用並具良好效果之光侷限效果之結構，此結構藉由奈米壓印技術轉印製透明導電膜玻璃基板上，後續再成長非晶矽/非晶矽鍺串疊型太陽電池於此基板，藉由改變光侷限結構之高度、間距及深寬比我們發現短路電流可大幅提升7.1%並提升轉換效率，後續對於光侷限結構之優化期望可進一步提升太陽電池之轉換效率。

In this thesis, the development of hydrogenated amorphous silicon germanium (a-Si1-xGex:H) thin films, a-Si1-xGex:H single-junction solar cells, a-Si:H/ a-Si1-xGex:H tandem cell, and the optimization of the light management of tandem cells were reported. Regarding the film property, we investigated the effect of hydrogen dilution and rf power on the film quality of a-Si1-xGex:H alloys. The silicon thin films and solar cells were prepared by radio-frequency plasma-enhanced chemical vapor deposition (RF PECVD). We found that hydrogen played an important role in both gas phase and surface reaction, which impacted the film property. The increase in hydrogen dilution ratio reduced the deposition rate which may be due to the hydrogen etching. The FTIR analysis suggested that the formation of dihydride and polyhydride were suppressed as the hydrogen dilution was increased, indicating structural defects were reduced. With adequate hydrogen dilution ratio of 2, the microstructure factor and electrical characteristics of a-Si1-xGex:H alloys were significantly improved. Moreover, the hydrogen radical enhanced the incorporation of Ge atoms into the a-Si:H network, leading to enhanced photo sensitivity toward long-wavelength region. In sum, the efficiency of a-Si1-xGex:H single-junction solar cells was improved from 5.73% to 6.33%. Regarding the development of a-Si1-xGex:H single-junction cells, the effect of bandgap grading of a-Si1-xGex:H absorber near the p/i and the i/n interfaces was discussed. The a-Si1-xGex:H single-junction solar cells were improved by applying both p/i grading and i/n grading. Our results showed that both the p/i and the i/n grading increased the open-circuit voltage (VOC) as compared to the cell without grading. Combining the effects of VOC, JSC and FF, a suitable thickness of the p/i and the i/n grading was 20 nm and 160 nm, respectively. Finally, the grading structures accompanied with further optimization of doped layers were integrated to achieve a cell efficiency of 8.59 %. Light management is of great importance to further improve the efficiency of a-Si:H/ a-Si1-xGex:H tandem cells. This included distributing the current generated by sub-cells appropriately, reducing the parasitic absorption loss, and increasing the light incoupling. In this thesis, we developed the n-type microcrystalline silicon oxide (μc-SiOx:H(n)) and employed it as intermediate reflecting layer (IRL) and back reflecting layer (BRL) in a-Si:H/ a-Si1-xGex:H tandem cells. In the development of μc-SiOx:H(n), increasing RF power increased film oxygen content, which widened the bandgap while reducing dark conductivity. Applying the μc-SiOx:H to a-Si:H/ a-Si1-xGex:H tandem cells as IRL and BRL significantly improved the cell performance. The μc-SiOx:H(n) IRL increases the current of the top cell, thus improving the distribution of photo-generated current of component cells. On the other hand, the μc-SiOx:H(n) was used as the BR replacing the n-type a-Si:H and ITO layers. The μc-SiOx:H increased cell conversion efficiency by 12.9% as IRL, and by 9.7% as BR, achieving 10.03% efficiency. Development of low-cost, reproducible, and customized light trapping of the a-Si:H/ a-Si1-xGex:H tandem cells was another focus of this thesis. Thanks to the collaboration of Taiwan and California Institute of Technology (Caltech) Energy Exchange Program, we were able to explore more on the light-trapping structure. Specifically, we have developed large-area and reproducible SiO2 light-trapping structure on the textured SnO2:F-coated substrates. The a-Si:H/ a-Si1-xGex:H tandem cells prepared on this substrates exhibited improved JSC by 7.1%, leading to improved efficiency.