使用原子層沉積法成長氧硫化鋅緩衝層的無鎘/無毒銅鋅錫硫硒太陽能電池
No Thumbnail Available
Date
2020
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
銅鋅錫硫硒(CZTSSe)薄膜太陽能電池是極具淺力且便宜的新一代太陽能電池，由地表豐富的且無毒的元素組成。但眾所皆知的CZTSSe太陽能電池應用的N型材料為有毒且能隙較小的硫化鎘(CdS),這不僅會造成環境汙染且因能隙小(2.4eV)會吸收到短波長的可見光。為了克服這問題，本論文用寬能隙(2.7-3.2eV)且無毒的N型材料氧硫化鋅Zn(O,S)來取代CdS。
在之前實驗室的研究中，我們用化學水浴法(CBD)沉積無毒的Zn(O,S)與吸收層CZTSSe形成p-n接面。然後，發現用CBD的方式會形成寬能隙的副產物氫氧化鋅(Zn(OH)2)。因此我們提出了原子層沉積法(ALD)來成長Zn(O,S)，看中它的優點如精確控制薄膜厚度，精確控制材料化學計量，最重要的是不會產生副產物Zn(OH)2。
首先我們自己架設一台原子層沉積機台(ALD)在實驗室裡，並設計了穩定的參數。產生Zn(O,S)薄膜需要H2S氣體，因此我們找尋一個相對安全的方法去產生H2S氣體，且證明其穩定性與鋼瓶一樣穩定。透過量測X-射線光電子光譜(XPS)證明我們能藉由調控氧硫比例使Zn(O,S)的導帶位置改變。利用紫外光-可見光光譜儀(UV-Vis)來得到不同氧硫比例的吸收光譜以及帶隙位置，我們發現不論比例為多少的Zn(O,S)帶隙都是高於CdS。此外，我們從XPS和UV-Vis分析中證明透過ALD成長的Zn(O,S)是沒有副產物產生的。此外，經調整適當的氧硫比例，我們發現Zn(O,S)最佳厚度約30nm。X射線衍射儀（XRD）結果提供了有利的證據，相轉變會發生在40%<S<70%的範圍內成為非結晶態，並且在S=20％時，結晶性變差。另一方面，從XRD發現，當S＝20％時，（002）的特徵峰與ZnO相比沒有任何偏移。據推測，在ALD過程中使用較高的溫度成長，我們沒有讓Zn（O，S）成長為理想的薄膜。透過ALD成長Zn(O,S)的CZTSSe太陽能電池上效能表現相對於CdS是有所改善的，光與暗電流間無交叉，表示 p-n 接面處有較少的缺陷。其開路電壓（Voc）可達470 mV，填充因子有58.06 ％，其短路電流密度 (Jsc) 更從原本28.70提升至36.07 mA/cm2，串聯電阻（Rs）也從38.01降至30.03(Ω. cm²），終而達到效率9.77 %的無毒 CZTSSe 太陽能電池。
CZTSSe thin-film solar cells are good potential and inexpensive new-generation solar cells, composed of earth abundant and non-toxic elements. However, the well-known n-type buffer material, CdS used in CZTSSe solar cells is toxic and has low bandgap. This not only causes environmental pollution but also absorbs short-wavelength of visible light due to the lower band gap (2.4eV) value. To overcome these issues, we propose to replace CdS with a wide band gap (2.7-3.2eV) and non-toxic material zinc oxysulfide (Zn(O,S)) as a buffer layer in CZTSSe solar cell. In previous laboratory research, we have deposited non-toxic Zn(O,S) by chemical water bath (CBD) to form a p-n junction with the absorber layer CZTSSe. However, it was found that the Zn(O,S) grown by CBD method form by-product (Zn(OH)2), which is wide bandgap material and the conduction band offset Zn(O,S) and CZTSSe not suitable due to improper oxygen-sulfur ratio. Therefore, we propose the atomic layer deposition (ALD) technique to grow Zn(O,S), using its advantages such as precise control of film thickness, precise control of material stoichiometry. The most importantly, the formation of by-product Zn(OH)2 could be avoid. First, we have set up a homemade ALD in the laboratory and designed stable and optimize parameters. In this system, we have designed a relatively safe method to generate H2S gas for Zn(O,S) deposition. We have proved the stability of our lab generated H2S similar as like a cylinder filled with H2S gas. X-ray photoelectron spectroscopy (XPS) confirmed the change of conduction band position of Zn(O,S) by tuning the ratio of oxygen and sulfur. UV-Vis spectrometer (UV-Vis) was used to obtain the absorption spectrum and bandgap of different S/(O+S) ratios, and we seen that the optical bandgap of Zn(O,S) is larger than that of CdS, irrespective of the S/(O+S) ratios. Further we confirmed from XPS and UV-Vis results that Zn (O,S) grown by ALD has not produced by-products. After adjusting the appropriate S/(O+S) ratio, we found the optimum suitable thickness of Zn(O,S) buffer layer is ~30 nm. Then we also determined the band position of CZTSSe, different S/(O+S) ratios Zn(O,S) film, and CdS. X-ray Diffractometer (XRD) spectra indicated the phase change in Zn(O,S) structure in composition range of 40%<S<70% and in this range structure became amorphous, whereas at S=20% the structure have poor crystalline. Addition to this, we found in XRD of S=20%, the peak position of (002) plane remains in similar position of like ZnO. From this observation, we speculated that we did not let Zn(O,S) grow into a perfect thin film due to the higher temperatures in the ALD process. From measured device performance with different S/(O+S) ratios of Zn(O,S), the most improved efficiency achieved with S=20%. Current density-voltage (J-V) shown there is no crossover between light and dark current, indicating that there are fewer defects at the p-n junction. Its open-circuit voltage can reach up to 470 mV, the fill factor 58.06 %, and short circuit current density has increased from 28.70 mA/cm2 (for pristine) to 36.07 mA/cm2 (with optimized Zn(O,S)) and the series resistance also has reduced from 38.01 to 30.03 (Ω.cm²), eventually achieving a non-toxic CZTSSe solar cell with an efficiency of 9.77%.
CZTSSe thin-film solar cells are good potential and inexpensive new-generation solar cells, composed of earth abundant and non-toxic elements. However, the well-known n-type buffer material, CdS used in CZTSSe solar cells is toxic and has low bandgap. This not only causes environmental pollution but also absorbs short-wavelength of visible light due to the lower band gap (2.4eV) value. To overcome these issues, we propose to replace CdS with a wide band gap (2.7-3.2eV) and non-toxic material zinc oxysulfide (Zn(O,S)) as a buffer layer in CZTSSe solar cell. In previous laboratory research, we have deposited non-toxic Zn(O,S) by chemical water bath (CBD) to form a p-n junction with the absorber layer CZTSSe. However, it was found that the Zn(O,S) grown by CBD method form by-product (Zn(OH)2), which is wide bandgap material and the conduction band offset Zn(O,S) and CZTSSe not suitable due to improper oxygen-sulfur ratio. Therefore, we propose the atomic layer deposition (ALD) technique to grow Zn(O,S), using its advantages such as precise control of film thickness, precise control of material stoichiometry. The most importantly, the formation of by-product Zn(OH)2 could be avoid. First, we have set up a homemade ALD in the laboratory and designed stable and optimize parameters. In this system, we have designed a relatively safe method to generate H2S gas for Zn(O,S) deposition. We have proved the stability of our lab generated H2S similar as like a cylinder filled with H2S gas. X-ray photoelectron spectroscopy (XPS) confirmed the change of conduction band position of Zn(O,S) by tuning the ratio of oxygen and sulfur. UV-Vis spectrometer (UV-Vis) was used to obtain the absorption spectrum and bandgap of different S/(O+S) ratios, and we seen that the optical bandgap of Zn(O,S) is larger than that of CdS, irrespective of the S/(O+S) ratios. Further we confirmed from XPS and UV-Vis results that Zn (O,S) grown by ALD has not produced by-products. After adjusting the appropriate S/(O+S) ratio, we found the optimum suitable thickness of Zn(O,S) buffer layer is ~30 nm. Then we also determined the band position of CZTSSe, different S/(O+S) ratios Zn(O,S) film, and CdS. X-ray Diffractometer (XRD) spectra indicated the phase change in Zn(O,S) structure in composition range of 40%<S<70% and in this range structure became amorphous, whereas at S=20% the structure have poor crystalline. Addition to this, we found in XRD of S=20%, the peak position of (002) plane remains in similar position of like ZnO. From this observation, we speculated that we did not let Zn(O,S) grow into a perfect thin film due to the higher temperatures in the ALD process. From measured device performance with different S/(O+S) ratios of Zn(O,S), the most improved efficiency achieved with S=20%. Current density-voltage (J-V) shown there is no crossover between light and dark current, indicating that there are fewer defects at the p-n junction. Its open-circuit voltage can reach up to 470 mV, the fill factor 58.06 %, and short circuit current density has increased from 28.70 mA/cm2 (for pristine) to 36.07 mA/cm2 (with optimized Zn(O,S)) and the series resistance also has reduced from 38.01 to 30.03 (Ω.cm²), eventually achieving a non-toxic CZTSSe solar cell with an efficiency of 9.77%.
Description
Keywords
銅鋅錫硫硒, 薄膜太陽能電池, 原子層沉積法, 無鎘緩衝層, 氧硫化鋅, 異質接面, 導帶位障差異, CZTSSe, Thin-film solar cells, Atomic layer deposition, Cadmium-free buffer layer, Zinc oxysulfide, Heterojunction, Conduction band offset