氫、氧與硫對銅銦鎵硒薄膜太陽能電池效能之影響以及AMPS-1D之元件模擬

Abstract

銅銦鎵硒薄膜太陽能電池屬於第二代太陽能電池，其低成本、高效率有望在未來取代目前市佔率最高之多晶矽太陽能電池。唯其因材料為四元合金，成長控制不易以及薄膜介面與內部缺陷較多(跟矽晶材料相比)，使得大面積生產良率受到很大的挑戰。因此了解此材料複雜的缺陷組成並且使用合適且低成本的硒化與鈍化方式為提升電池效率和降低電池成本之方法。 本論文旨在探討氫元素、氧元素的摻雜與硫元素的表面鈍化處理對銅銦鎵硒薄膜太陽能電池效能之影響，並且做一維光伏元件模擬針對銅銦鎵硒吸收層、銅銦鎵硒與硫化鎘介面缺陷密度以及二硒化鉬過渡層(在銅銦鎵硒與鉬背電極之間)作探討。 首先，藉由新穎式大氣電漿輔助化學沉積系統在次大氣壓力下(150 Torr)做硒薄膜前驅層之沉積以及對銅銦鎵金屬前驅層進行不同程度之氧化，藉此探討氧摻雜對元件之影響。大氣壓電漿技術具有低溫、低成本、適合大面積化等競爭優勢，可期望於未來商業化應用。大氣電漿產生的OH-物種在銅銦鎵金屬前驅層表面形成不同程度之氧化態，這在硒化過程中影響銅銦鎵硒吸收層氧的摻雜量與硒元素、鎵元素的擴散。適當的氧摻雜對硒空缺(施體缺陷)的鈍化有正面影響，可提升元件有效電洞載子密度、降低吸收層串聯電阻 ; 此外，硒空缺所造成的金屬空缺也影響鎵、銦元素擴散，最佳化條件下可在吸收層中形成一均勻的鎵分布。硒擴散速率則受到金屬氧化層之影響，適當條件下可形成較薄的二硒化鉬過渡層來降低背電極串聯電阻。較薄的二硒化鉬層也使得鈉元素容易由基板擴散入吸收層，對銅占銦之錯位缺陷鈍化而使元件有效電洞載子密度增加。氧摻雜的最佳元件參數: 電池效率為8.9 %、開路電壓為0.454 V、短路電流為35.7 mA/cm2、填充因子為0.547。 其次，研究氫氣/氮氣混合氣體作為硒蒸氣之載氣對銅銦鎵金屬前驅層進行硒化，氫氣作為鈍化材料內部缺陷的摻雜來源。混合載氣比例在氫氣/(氫氣+氮氣)為15%時，元件有效電洞載子密度提升近2倍達到5.461  1015 cm-3，此對於電池整體的串聯電阻降低以及短路電流密度增加有很大貢獻。此外，藉由氫氣對銅空缺(受體缺陷)鈍化，增加n型表層厚度，使得空乏區增大，則可產生更多載子。氫摻雜的最佳元件參數: 電池效率為10.7 %、開路電壓為0.507 V、短路電流為33.8 mA/cm2、填充因子為0.625。 最後，銅銦鎵硒吸收層表面鈍化使用低成本的化學水域法作硫化處理。藉由硫元素對吸收層表面的硒空缺進行鈍化，降低銅銦鎵硒與硫化鎘介面複合中心缺陷密度而提升載子生命週期。氧摻雜與氫摻雜之最佳試片進行硫元素表面鈍化後，元件效能分別提升7.9 % (電池效率為9.6 %)和10.3 % (電池效率為11.8 %)。 銅銦鎵硒薄膜光伏元件使用AMPS-1D的模擬結果顯示: (i) 吸收層的能隙分布對元件效率影響很大，效率要高的準則為吸收層表層能隙要大、能隙大於1.15eV、吸收層能隙分布為一均質分布或V形狀為最佳、降低深層缺陷密度。 (ii) 銅銦鎵硒與硫化鎘介面缺陷密度大於1018 cm-3時，電池效率開始降低，尤其是填充因子減少幅度最大。 (iii) 二硒化鉬過渡層的能隙低於1.3 eV時，電池背向電場效應減弱造成電池效率下降。

Cu(In,Ga)Se2 (CIGS) thin-film solar cell was second generation solar cell, it was an alternative to poly-Si solar cells due to its low-cost and high cell efficiency in the future. The mass production of large-area CIGS solar modules faced the challenge owing to However, the mass production of large-area CIGS solar modules faced the challenge owing to the deviation from stoichiometry in the CIGS thin film and there are a lot of non-stoichiometric defects at the interface of CIGS/CdS and inside CIGS bulk (as compared to poly-Si). Therefore, it is essential to understand complicated intrinsic defects at the CIGS surface and bulk, and then adopted corresponding low-cost passivation methods and employed low-cost selenization process to enhancement of cell efficiency and decreasing fabrication cost. In this thesis, we investigated effects of hydrogen, oxygen and sulfur on the performance of CIGS solar cells, and device modeling using AMPS-1D focused on the study of the impact on performance caused by various CIGS band-gap profile, interface defect density (CIGS/CdS) and band-gap of MoSe2 transition layers. First, novel atmospheric pressure plasma enhanced chemical vapor deposition (AP-PECVD) is proposed to deposit Se precursors (at 150 Torr) and oxidize the surface of Cu3Ga/In (CIG) precursors, this was employed to investigate the effect of oxygen on the conversion-efficiency of CIGS solar cells. The AP-PECVD offers several competitive advantages, such as low temperature process, low cost and suitable for large area application. It is expected for commercial applications in the future. CIG/Se precursors were oxidized by OH- species produced using atmospheric pressure plasma, and oxygen has an influence on Se and Ga diffusion in the CIGS absorber. Optimal oxygen inclusion of CIG/Se precursors was beneficial to passivate donor-like Se vacancies (VSe) by OSe during CIGS formation process, and this resulted in the enhancement of electrical properties (effective hole carrier concentration and series resistance) of devices. The existence of VSe also induced metal vacancies, which resulted in homogeneous Ga diffusion in CIGS thin films at optimal condition. A metal oxide formed at the interface of CIG/Se precursors has an effect on Se diffusion rate and MoSe2-layed thickness during selenization process. Thin MoSe2 layer led to low series resistance of devices and facilitated Na diffusion from soda-lime glass at high temperature (>550 degree). The best performance of solar cell made using oxygenation CIGS absorber: cell efficiency (Eff.) is 8.9 %, Voc is 0.454 V, Jsc is 35.7 mA/cm2 and FF is 0.547. Second, the effect of hydrogen on suppression of CIGS defects was investigated using selenization source of non-toxic Se vapor carried by N2/H2 mixed gas. Hydrogen could passivate acceptor-like Cu vacancies (VCu) to increase n-type region of CIGS thin films and widen depletion region. A two-fold increase in effective hole carrier concentration (5.461 x 10^15 cm-3) of the CIGS solar cell achieved at H2/(H2+N2) gas flow rate ratio of 15 %, and this led to lower series resistance and increasing Jsc of the solar cell. The best performance of solar cell made using hydrogenation CIGS absorber: cell efficiency is 10.7 %, Voc is 0.507 V, Jsc is 33.8 mA/cm2 and FF is 0.625. Third, low-cost sulfur-containing acid chemical bath deposition (CBD) method was used to passivate CIGS surface. Se vacancies at the CIGS surface were passivated by sulfur ions (e.g. S2-) to decrease defects at the interface of CIGS/CdS layers and increase lifetime of carrier. After sulfur treatment, cell efficiencies of solar cell fabricated using oxygenation and hydrogenation CIGS absorber enhanced by 7.9 % (Eff. is 9.6 %) and 10.3 % (Eff. is 11.8 %), respectively. Finally, the CIGS solar cell modeling using AMPS-1D revealed: (i) Large band-gap (Eg) of CIGS absorber (>1.15 eV), uniform or V-shape CIGS Eg profile and low deep-level defects of CIGS bulk could reaches high cell efficiency. (ii) The cell efficiency began to decreasing as interface defect density (at the interface of CIGS/CdS layers) was above 1018 cm-3. (iii) The cell efficiency began to decreasing as Eg of MoSe2 layer was below 1.3 eV.