利用漸變式矽過多氧化矽多層膜結構製做高密度矽量子點薄膜於光伏元件之應用

全文

(1)國立交通大學顯示科技研究所碩士論文利用漸變式矽過多氧化矽多層膜結構製做高密度矽量子點薄膜於光伏元件之應用. High Density Si Quantum Dot Thin Films Using a Gradient Si-Rich Oxide Multilayer Structure For Photovoltaic Devices Application. 研究生：黃品睿指導教授：李柏璁教授. 中華民國 101 年 8 月.

(2) 利用漸變式矽過多氧化矽多層膜結構製做高密度矽量子點薄膜於光伏元件之應用 High Density Si Quantum Dot Thin Films Using a Gradient Si-Rich Oxide Multilayer Structure For Photovoltaic Devices Application 研究生：黃品睿指導教授：李柏璁. Student：Pin-Ruei Huang 博士. Advisor：Dr. Po-Tsung Lee. 國立交通大學顯示科技研究所碩士論文 A Thesis Submitted to Department of Photonic and Display Institute College of Electrical Engineering and Computer Science National Chiao Tung University In partial Fulfillment of the Requirements for the Degree of Master in Electro-Optical Engineering August 2012 Hsinchu, Taiwan, Republic of China. 中華民國 101 年 8 月 .

(3) 利用漸變式矽過多氧化矽多層膜結構製做高密度矽量子點薄膜於光伏元件之應用. 研究生：黃品睿指導教授：李柏璁博士. 國立交通大學顯示科技研究所. 摘要矽基太陽能電池是目前全球市佔率最高的太陽能電池種類，加上其原料充足與製程技術成熟等優勢，未來前景仍相當被看好；為了更進一步提升元件效率與降低製程成本以達到第三代太陽能電池的目標，全矽基多接面太陽能電池正被廣泛地研究與開發。而其中具奈米結晶態的矽量子點薄膜則被視為極具潛力能克服矽基材料的能隙限制，以解決短波長光子的嚴重損耗議題。截至目前，國際上的研究團隊製做的矽量子點薄膜以矽過多氧化矽(silicon-rich oxide)單層膜或二氧化矽(silicon dioxide)/矽過多氧化矽(silicon-rich oxide)多層膜結構為主，但前者不易調控量子點尺寸，後者則有量子點間距過大的問題，加上兩者皆無法達到高密度的矽量子點形成，導致元件的光激載子傳輸嚴重受限，整體效率仍有待改善。故本篇論文提出以漸變式矽氧濃度之氧化矽薄膜取代二氧化矽侷限層的構想，藉由在沉積時連續性調控矽氧組成比，使其每一週期的矽濃度呈現漸變式分佈(低高低)，以此漸變濃度結構輔助後續退火過程中的奈米結晶矽形成，期望能同時達到矽量子點的尺寸調控及間距縮短，大幅改善光激載子傳輸效益，以提升矽量子點型太陽能電池之工作效率。 I.

(4) High Density Si Quantum Dot Thin Films Using a Gradient Si-Rich Oxide Multilayer Structure For Photovoltaic Devices Application. Student: Pin-Ruei Huang Advisor: Dr. Po-Tsung Lee Display Institute National Chiao Tung University. Abstract So far, the Si-based solar cell is the highest global market share and the good development potential due to the plentiful materials and the well-developed fabrication technique. In order to achieve the goal of the third generation solar cells with high efficiency and low cost, all Si-based multiple-junction solar cell is widely investigated and developed nowadays. The nano-crystalline Si quantum dot (QD) thin film is one of the potential structures to overcome the bandgap limitation of Si-based materials. Silicon-rich oxide (SRO) single layer and [SRO/SiO 2 ] multilayer (ML) thin films are the most commonly used deposition structures for Si QD thin films. However, the former is hard to control the QD's size and density simultaneously, the latter exists the QD’s separation limitation due to the SiO 2 barrier layers inserted. Furthermore, the QD’s density of both structures is still not high enough for a better PV application. These result in the difficulty for good photo-generated carrier’s transportation and high conversion efficiency. Hence, to efficiently improve the carrier’s transportation properties is a critical issue for the high efficiency Si-based solar cells integrating Si QD thin film. In this study, we propose a more potential deposition structure by a gradient Si-rich oxide multilayer (GSRO-ML) structure for the QD size control and the high QD density. The nano-structure, crystalline, and optical properties of Si QD thin films using a GSRO-ML structure had been studied. It also shows the better photovoltaic properties than that using a [SRO/SiO 2 ]-ML structure. A higher conversion efficiency of Si QD thin films utilizing a GSRO-ML structure can be highly expected by using a heavy doping concentration in the near future. II.

(5) Acknowledgements 碩士班生活到了尾聲，一千多個日子隨著每天每天的實驗飛快的流逝，回首這三個寒暑的學習以及歷練讓我成長了許多。這一路上感謝很多人的照顧，首先要感謝的是我的指導教授—李柏璁老師，提供給我們完好的研究環境以及豐富的實驗室資源，並在研究上適時的給予寶貴的意見。還有我最最敬愛的老大—光揚學長，抱歉很常惹你生氣，謝謝你的耐心指導以及包容我的白目和遲鈍，我想我唯一的優點應該就是使命必達吧。還要感謝贊博、岳哥、狗勾、佐哥、佳裕、胖胖、G 隆、yoyo 哥、小源，不管是實驗上的討論還是生活瑣事，暢快的談天說地總是可以讓我忘掉實驗的疲憊；還有一起奮鬥的 98 同學們小邱邱、金剛、崴崴學長、文齡、呂紹平、邱哥，在那些日子，一起修課，一起做實驗，一起夜唱到天亮然後隔天繼續上課的爆肝生活我好懷念。雖然你們先走一步，但是我才不會忘記你們呢！以及一起打拚的 99 學弟們小智、嶢嶢、權政，尤其”感謝”權政，因為你總讓我”事倍功半” !加油!你們也快要可以寫誌謝了!還有 100 學弟酷哥和小朱，感謝你們平常總是幫我們跑腿處理一些瑣事，還有戲弄你們兩位真是我歡樂的來源，讓我忘了做實驗的辛苦。真的非常感謝實驗室的大家，每天生活在一起，不管是研究、運動、休閒總是有大家的記憶，謝謝你們，對我而言你們不只是學長/同學 /學弟，也是家人。另外要感謝冉老師實驗室的治寬、良豪、小胖學弟，對不起每次換 round 的時候都 delay，但你們從來沒有抱怨過我。還要感謝余老師實驗室的劉孝威，百忙之中還抽空幫我量效率，而且連一頓麥當當都不願意讓我請，都沒機會報答你。還有感謝爐管帥哥，謝謝你，讓我曾經有一年的時光，在登入奈中的時候都是帶著期待又愉悅的心情。最後，感謝你，因為有你才能讓我更有動力，我會一直努力朝你所在的地方前進......。最重要的，感謝老媽和老哥，我很少回家，應該是個不孝女無誤，不過感謝有你們，讓我可以無後顧之憂的專心在研究上。過幾天，我即將邁入人生下一個階段—博士班研究生涯，所以這份誌謝…… ........to be continued 黃品睿 2012 年 9 月于國立交通大學 401 室. III.

(6) Content 摘. 要 ........................................................................................................................I. Abstract ........................................................................................................................... II Acknowledgements .......................................................................................................... III Content .......................................................................................................................... IV List of Figures ................................................................................................................. VII List of Tables .................................................................................................................... XI Chapter 1 Introduction ....................................................................................................... 1 1.1. Background ..................................................................................................... 1. 1.2. Solar Cell ........................................................................................................ 2. 1.3. Silicon Based Tandem Solar Cells .................................................................. 4 1.3.1. Energy Loss paths in Single Junction Solar Cells .................................. 4. 1.3.2. Quantum Confinement Effect ................................................................. 5. 1.3.3 Tandem Solar Cell Using Si Quantum Dot Thin Films .......................... 6 1.4. 1.5. Paper Review .................................................................................................. 9 1.4.1. Si-Rich-Oxide Single Layer (SRO-SL) Thin Film ................................. 9. 1.4.2. [SRO/SiO 2 ] Multilayer (ML) Thin Film .............................................. 10. Motivation .................................................................................................... 15. Reference .......................................................................................................................... 16 Chapter 2 Fabrication of Gradient Si-Rich Oxide Multilayer (GSRO-ML) Structure for High Density Si QD Thin Film................................................................................................. 19 2.1. Substrate Clean ............................................................................................ 19. 2.2. GSRO-ML Thin film Deposition................................................................. 21. IV.

(7) 2.2.1. Principle of Radio Frequency (RF) Magnetron Sputtering Process ..... 21. 2.2.2. GSRO-ML Thin Film ........................................................................... 22. 2.3. Post-annealing Process ................................................................................ 23. 2.4. Thermal Oxide Layer Remove .................................................................... 23. 2.5. Electrode Layer Deposition ......................................................................... 24. Chapter 3 Experimental Equipments and Analyzed Methods .......................................... 25 3.1. Raman Scattering Spectrum ........................................................................ 25. 3.4. Photoluminescence (PL) Spectrum ............................................................. 28. 3.5. Ultraviolet/visible/Near-infrared (UV/VIS/NIR) Spectrum ........................ 29. 3.6. Current-Voltage (I-V) Curve ....................................................................... 29. Reference .......................................................................................................................... 30 Chapter 4 Results and Discussion .................................................................................... 31 4.1. GSRO-ML Thin Films without NL ............................................................. 31 4.1.1. 4.2-1. Crystalline Properties of GSRO-ML Thin Films without NL .............. 31. GSRO-ML Thin Films with NL ............................................................... 34. 4.2-1-1. Crystalline Properties of GSRO-ML Thin Films with NL ............... 35. 4.2-1-2. Optical Properties of GSRO-ML Thin Films with NL ..................... 36. 4.2-1-3. Electro-Optical Properties of GSRO-ML Thin Films with NL ........ 37. 4.2-2. Comparison of Using GSRO-ML, [SRO/SiO 2 ]-ML, and SRO-SL. Structures. .................................................................................................................. 39. 4.2-2-1 Structures 4.2-2-2. Crystalline Properties of GSRO-ML, [SRO/SiO 2 ]-ML, and SRO-SL .......................................................................................................... 39 Structural properties of GSRO-ML and [SRO/SiO 2 ]-ML Structures 41. 4.2-2-3. Optical Properties of GSRO-ML, [SRO/SiO 2 ]-ML, and SRO-SL V.

(8) Structures. 4.3. .......................................................................................................... 42. 4.2-2-4. Electrical Properties .......................................................................... 45. 4.2-2-5. Carrier’s Transportation Mechanism of NC-Si QD Thin Film ......... 46. GSRO-ML Thin Films with Highly Doping NL .......................................... 49 4.3.1. Electrical Properties of GSRO-ML Thin Films with Highly Doping NL .............................................................................................................. 49. 4.3.2. H 2 Passivation Effect ............................................................................ 50. Reference .......................................................................................................................... 54 Chapter 5 Conclusion and Future Work ........................................................................... 56 5.1. Conclusion ................................................................................................... 56. 5.2. Future Work ................................................................................................. 57 5.2.1. Heavy Doping Si QD Thin Film........................................................... 57. 5.2.2. High Efficiency p-i-n Si QD Thin Film SCs ........................................ 57. Reference .......................................................................................................................... 59. VI.

(9) List of Figures Chapter 1 Fig. 1-1 Conversion efficiency improvements in silicon solar cells (1954-2008). .................... 2 Fig. 1-2 Efficiency and cost projections for first-, second- and third-generation photovoltaic technology (wafers, thin-films, and advanced thin-films, respectively). ........................... 3 Fig. 1-3 Energy loss paths in a single junction solar cell: (1) high-energy photon loss, (2) junction loss, (3) contact loss, (4) recombination loss, (5) low-energy photon loss. ......... 5 Fig. 1-4 Illustration of 3-D time-independence Schrödinger’s equation and boundary conditions. .......................................................................................................................... 6 Fig. 1-5 (a) Illustration of Si QD embedded in materials with different QD size. (b) Experimental bandgaps of Si QDs embedded in SiO 2 and SiN x from other groups. ........ 7 Fig. 1-6 Scheme of Si-based tandem solar cell. ......................................................................... 8 Fig. 1-7 Cross-sectional TEM images of SiO/SiO 2 superlattices: (a) As-prepared SiO/SiO 2 superlattice. The darker layers represent the SiO sublayers. (b) The same film after annealing. The separation of the nanocrystals by a thin oxide shell is clearly visible. (c) High resolution TEM image of the film. For clarity, the visible nanocrystals are highlighted by circles. The crystals are only found in the former SiO layers, which are emphasized by the lines in the image. (d) TEM image of a film with even thinner SiO layers ~2 nm after annealing. ........................................................................................... 11 Fig. 1-8 (a) Schematic diagram of (n-type) Si QDs and (p-type) c-Si heterojunction solar cell. And transmission electron microscopy (TEM) images of Si quantum dots in SiO 2 matrix with (b) low-magnification and (c) high-resolution lattice images for 5 nm Si QDs. ...... 11 Fig. 1-9 Four different I–V characteristics of the (n-type) Si QDs/ (p-type) c-Si heterojunction. VII.

(10) devices. ............................................................................................................................. 12 Fig. 1-10 (a) TEM images and (b) SIMS depth profiles of [SiO 2 (8 nm)/B-doped Si(10 nm)] 5 ML film with a B doping level of 1.7 × 1020 atoms cm-3 after annealing at 1100°C for 20 min by rapid thermal annealing. ....................................................................................... 13 Fig. 1-11 Representative TEM micrographs of the (a) as-deposited and (b) annealed [B-doped SiO 1.0 /SiO 2 ] 25 superlattice films at 1100 ◦C for 10 min. ................................................. 13 Fig. 1-12 Current–voltage (I-V) characteristics of a p-Si QDs/n-Si heterojunction solar cell under air mass 1.5 (AM 1.5G) illumination of 100 mW cm−2. ........................................ 13 Fig. 1-13 SIMS data of as-deposited (solid line) and annealed (dotted line) p-i-n structure. Significant interdiffusion is observed after annealing for 1 hour at 1100 °C. .................. 14 Fig. 1-14 Dark I-V characteristics of a 0.12 cm2 diode. Inset shows the I-V characteristics under 1-sun illumination with V OC =373 mV. ............................................................................. 14. Chapter 2 Fig. 2-1 Fabrication process of GSRO thin films. .................................................................... 19 Fig. 2-2 Clean process chart of (a) Si wafer and (b) quartz substrates. .................................... 20 Fig. 2-3 Illustration of Operation of magnetron sputtering deposition method ....................... 21 Fig. 2-4 The variation of Si concentration. ............................................................................... 22 Fig. 2-5 Scheme of as-deposited GSRO-ML ........................................................................... 22 Fig. 2-6 Scheme of as-deposited GSRO-ML structure. ............................................................ 23 Fig. 2-7 Schemes of top electrodes of (a) square and (b) finger patterns deposited by thermal evaporation coater. ............................................................................................................ 24. Chapter 3 Fig. 3-1 Illustration of Rayleigh and Raman scattering............................................................ 26 VIII.

(11) Fig. 3-2 Three components decomposed from Raman spectra of Si QD thin films................. 26 Fig. 3-3 FWHM of the Raman peaks and Si QD size against the corresponding Raman shift .... .......................................................................................................................................... 27 Fig. 3-4 Schematic of electronic transition. ........................................................................... 28. Chapter 4 Fig. 4-1 Schemes of GSRO-ML thin films (a) without NL and (b) with NL ........................... 31 Fig. 4-2 Raman spectra of GSRO-ML thin films without NL under different SiO 2 sputtering powers. .............................................................................................................................. 33 Fig. 4-3 Variations of Si and SiO 2 sputtering powers for each GSRO layer. ........................... 34 Fig. 4-4 Scheme of the GSRO-ML thin film with NL. ............................................................ 34 Fig. 4-5 Raman spectra of G20 and G20-NL. .......................................................................... 35 Fig. 4-6 Raman spectra of G20-NL under different annealing time. ........................................ 35 Fig. 4-7 Absorption spectra in Tauc’s plot for G20-NL under different annealing time. ......... 36 Fig. 4-8 I-V curves of G20-NL on p-type Si(100) wafer with and without a halogen lamp illumination under an annealing duration time of (a) 20 and (b) 60 minutes. .................. 37 Fig. 4-9 Schemes of (a) [SRO/SiO 2 ]-ML and (b) SRO-SL deposition structures ................. 39 Fig. 4-10 Raman spectra of [SRO/SiO 2 ]-ML, G20-NL, and SRO-SL thin films. ................... 40 Fig. 4-11 High-resolution TEM images of (a) [SRO/SiO 2 ]-ML and (b) G20-NL thin films. .. 42 Fig. 4-12 QD’s size distribution of [SRO/SiO 2 ]-ML and G20-NL thin films. ......................... 42 Fig. 4-13 PL spectra of [SRO/SiO 2 ]-ML and G20-NL. ........................................................... 43 Fig. 4-14 I-V curves of [SRO/SiO 2 ]-ML and G20-NL under a 488 nm laser illumination. .... 43 Fig. 4-15 (αhν)1/2 versus hν plots of [SRO/SiO 2 ]-ML and G20-NL thin films. ....................... 44 Fig. 4-16 Dark and light I-V curves of (a) [SRO/SiO 2 ]-ML and (b) G20-NL on p-type Si wafers. .......................................................................................................................................... 45 IX.

(12) Fig. 4-17 Temperature-dependent current versus voltage (I–V–T) characteristics of (a) [SRO/SiO 2 ]-ML and (b) G20-NL-ML thin film structure measured in the temperature range 300–340K using 20K steps. .................................................................................... 46 Fig. 4-18 (a) I-V curve of G20-NL and (b) J-E curve of [SRO/SiO 2 ]-ML. ............................. 48 Fig. 4-19 Scheme of GSRO-ML with different P-doping concentration. ................................ 49 Fig. 4-20 Dark and light I-V curves of (a) G20-NL(n-Si) and (b) G20-NL(n+-Si). ................. 50 Fig. 4-21 (a) Dark and (b) light I-V curves of G20-NL(n+-Si) annealed in forming gas under different post-annealing temperature. ............................................................................... 51 Fig. 4-22 (a) Dark and (b) light I-V curves of G20-NL (n+-Si) annealed at 400°C for 1 hour in N 2 or N 2 +H 2 . ......................................................................................................................... 52. Chapter 5 Fig. 5-1 V OC and J SC as functions of B-doping concentration (n B ). The inset shows fill factor and energy-conversion efficiency as functions of n B . ...................................................... 57 Fig. 5-2 I-V curves of p-type (a) [SRO/SiO 2 ]-ML and (b) GSRO-ML thin films on Si wafer under a halogen lamp illumination. .................................................................................. 58. X.

(13) List of Tables Chapter. 1. Table 1-1 Theoretical efficiencies and corresponding band gap combination depending on the active cell layers. ............................................................................................................. 8 Table 1-2 One-sun illuminated cell parameters of four different (n-type) Si QDs/ (p-type) c-Si heteroface devices measured at 298 K. ......................................................................... 12. Chapter. 4. Table 4-1 Sputtering parameters of GSRO-ML thin films without NL under different SiO 2 sputtering powers. ......................................................................................................... 32 Table 4-2 Curve-fitting results of Fig. 4-2. ........................................................................... 33 Table 4-3 Sputtering parameters of G20-NL. ....................................................................... 34 Table 4-4 Fitting results of Raman spectra for G20-NL. ...................................................... 36 Table 4-5 Curve-fitting result from Fig. 4-7 for E g,opt and α................................................. 37 Table 4-6 Parameters of the photo-response properties for the G20-NL with different annealing time. A halogen lamp with 1 mW/cm2 of power density is used as the illumination source. ....................................................................................................... 38 Table 4-7 Curve-fitting results of Fig. 4-10. ......................................................................... 41 Table 4-8 Curve-fitting results of Fig. 4-13. ......................................................................... 43 Table 4-9 the fitting results of [SRO/SiO 2 ]-ML and G20-NL thin films. ............................. 44 Table 4-10 Parameters of the photo-response properties for [SRO/SiO 2 ]-ML and G20-NL structures under a halogen lamp illumination ............................................................... 45 Table 4-11 Four different transport mechanisms through a junction. ................................... 48. XI.

(14) Table 4-12 Parameters of the electro-optical characteristics for G20-NL (n-Si) and G20-NL (n+-Si) under a halogen lamp illumination .................................................................... 50 Table 4-13 Parameters of the electro-optical characteristics of G20-NL (n+-Si) for different annealing temperature in N 2 +H 2 (5%) environment. .................................................... 51 Table 4-14 Information about series resistances (R s ) extracted from illuminated I-V curves under a halogen lamp illumination. ............................................................................... 51 Table 4-15 Parameters of the electro-optical characteristics of G20-NL (n+-Si) for different annealing ambient under a halogen lamp illumination ................................................. 53 Table 4-16 Information about series resistances (R s ) extracted from illuminated I-V curves under a halogen lamp illumination. ............................................................................... 53. XII.

(15) Chapter 1. 1.1. Introduction. Background. Energy, environment, and economy are the three major problems that plagued the development of modern society. The consumption of resource becomes larger with the development of industrialization and the population increasing, so it can be predicted the occurrence of energy crisis. Besides, the increase on average temperature of the earth and the acid rain due to the green house effect resulted from arising emission of green house gases( like CO 2 and SO 2 , etc. …) emission after burning fossil fuel are two main issues of the environmental protection. The problems mentioned above lead us to find out the best substitute energies, which are renewable and pollution-free such as wind, tides, geothermal heat and solar. Among them, solar energy is vital in the present times considering the fact that the power demand in the world is a never ending process. It’s promising to use solar energy to replace fossil fuel completely in the future with advanced technologies.. 1.

(16) 1.2. Solar Cell. The first solar cell (SC) was developed at Bell laboratories in 1883 by Charles Fritts who coated the semiconductor selenium with an extremely thin layer of gold to form the junctions. However, the conversion efficiency (~1%) was too low to be applied. It had not drawn attention until the oil crisis broke out in following decades. Recently, there are more and more researchers involving in efficiency enhancement. Fig. 1-1[1-1] is a timeline of the silicon solar cell’s steady rise in efficiency. But for applications, not only efficiencies but also the cost of manufacturing are to be considered. According to the efficiency as a function of cost, solar cells can be classified into three generations as shown in Fig. 1-2 [1-2].. Fig. 1-1 Conversion efficiency improvements in silicon solar cells (1954-2008)[1-1].. The first generation SCs are silicon wafer-based SCs which are the most commonly used and well-developed ones. Although the SCs have high efficiencies, the production cost also high. In order to reduce the cost, the second generation SCs, such as amorphous silicon (a-Si) thin film SC, organic cells, and dye-sensitized SCs (DSSCs) are developed. The cost is quiet cheaper than that of the first generation SCs. On the other hand, the efficiency of the second generation SCs are lower than that of first generation SCs. Other main features of the second generation SCs are their flexibility and light-weight which make lots of application innovations 2.

(17) such as flexible solar panels. For achieving SCs with high efficiency which is potential to be larger than efficiency limit of SC with single bandgap (31%) and low producing cost, the third generation SCs are proposed like poly-Si SCs and nanocrystalline Si (NC-Si) SCs, etc. The strategies are abundant, non-toxic and durable, so the technologies can develop without pollutions. Even if the third generation SC is superior to the others, they are still in the research phase.. Fig. 1-2 Efficiency and cost projections for first-, second- and third-generation photovoltaic technology (wafers, thin-films, and advanced thin-films, respectively) [1-2].. 3.

(18) 1.3. Silicon Based Tandem Solar Cells. 1.3.1. Energy Loss paths in Single Junction Solar Cells There are five energy loss paths in standard single junction solar cells, including junction. loss, contact loss, recombination loss, high-energy photon loss (also called thermalization loss) and low-energy photon loss (also called non-absorption loss) as shown in Fig. 1-3[1-3]. Thermalization loss occurs when the incident energy is larger than the bandgap, and non-absorption loss occurs when the incident energy is less than the bandgap. These two losses are the two most important losses in single-junction photovoltaic cells. In order to overcome the losses mentioned above, there are some approaches like: (a) increasing the number of bandgap for photon absorption from a wide-range spectrum; (b) capturing carrier before they thermalize to bandgap; (c) multiple carrier pair generation per high energy photon or single carrier pair generation with multiple low energy photons. Tandem solar cell, stacking sub-cells from large to small bandgap in turn, is a much more promising structure to achieve a high efficiency SC. NC-Si can be made very small size, less than 7 nm in diameter, and they behave like quantum dots (QDs), e.g. bandgap control with nanocrystal size, very fast optical transition, and multiple carrier, generation, owing to the three-dimensional quantum confinement of carrier [1-4]. NC-Si embedded in dielectric material cascaded with silicon-based solar cells is one of the proposed solar cell structures to achieve super high conversion efficiency due to its ability in energy bandgap engineering.. 4.

(19) Fig. 1-3 Energy loss paths in a single junction solar cell: (1) high-energy photon loss, (2) junction loss, (3) contact loss, (4) recombination loss, (5) low-energy photon loss.. 1.3.2. Quantum Confinement Effect. Behavior of particle waves confined in an infinite quantum well can be explained by three-dimensional (3-D) time-dependence Schrödinger’s equation expressed as. −. (). (). ().  2 ∇ Ψ r + V (r )Ψ r = EΨ r 2m. (1-1). We can obtain that allowance energy states are discrete and they depend on the width of the quantum well. The phenomenon comes from the quantum confinement effect. For particle-waves confined in nanoparticles covered in materials with finite barrier height, similar energy states can be obtained, as expressed by Eq. (1-2) En =.  2π 2 2 n , n 2 = nx2 + n y2 + nz2 2 2ma. (. ). (1-2). where n x , n y , n z are integers and equal to 1 for the ground state square box. Discrete energy levels depend on the dimension of nanoparticles and barrier height between nanoparticles and materials which cover around them. Fig. 1-4 [1-5] shows the wave functions of different nanoparticles sizes with a fixed barrier height and those of different barrier with a fixed nanoparticle size. As a result, we can control effective E g by tuning the QD’s size and changing the matrix material. 5.

(20) (b). (a) (c) Fig. 1-4 Illustration of 3-D time-independence Schrödinger’s equation and boundary conditions [1-5].. 1.3.3. Tandem Solar Cell Using Si Quantum Dot Thin Films From quantum theory, we know, when the QD’s size is reduced to a few nanometers, the. quantum confinement effect will occur. It will relax the K-space conservation requirement and transform the Si bandgap properties from indirect to quasi-direct and modify the effective E g of Si. Fig. 1-5(a) shows that embedding Si QD in a wide bandgap material can lead to a highly –tunable effective bandgap. In addition, Fig. 1-5(b) shows that the effective E g of Si QD can be widely modified by tuning the QD’s size in SiO 2 or Si 3 N 4 matrix material, and a larger bandgap than c-Si or a-Si material is also feasible using Si QD [1-6]. Therefore, we can integrate Si QD into Si-based SCs to reduce energy losses from mismatch bandgap.. 6.

(21) (a). (b). Fig. 1-5 (a) Illustration of Si QD embedded in materials with different QD size. (b) Experimental bandgaps of Si QDs embedded in SiO 2 and SiN x from other groups [1-6].. In order to achieve high efficiency and low cost solar cell, the third generation solar cells have been studied. One of the promising candidates is the tandem solar cell using Si QD thin films as shown in Fig. 1-6 [1-7]. The concept of the tandem solar cell is stacking different E g of solar cells from large to small ones for the absorbing different energy of photons. The uppermost cell has the highest bandgap and lets the photon less than its bandgap passing through to lower bandgap cells underneath. Tandem solar cells using Si QD thin films stacking materials with different energy bandgaps can utilize the wide solar spectrum more effectively. Table 1-1 [1-8] shows the best bandgap combinations under different numbers and their theoretical efficiencies.. 7.

(22) Fig. 1-6 Scheme of Si-based tandem solar cell [1-7].. Table 1-1 Theoretical efficiencies and corresponding band gap combination depending on the active cell layers [1-8].. 8.

(23) 1.4. Paper Review Si QD embedded in dielectric mediums have been investigated due to the potential for. optoelectronic applications such as photovoltaic devices. Si QD thin films fabrication by various deposition techniques is preferable because of the greater potential of integration into conventional devices. Si QD precipitating from silicon-rich layers is one of fabrications of Si QD systems by means of vacuum deposition techniques. For Si precipitation from an Si-rich oxide layer, high temperature annealing of excess Si in an inert atmosphere is necessary to form Si QD with a few nanometers in diameter, for example, Si QD precipitation in oxide [1-2, 1-9, 1-10], nitride [1-11], and carbide [1-12, 1-13]. Eq. (1-3) describes this Si precipitation mechanism: 1. 1. Si(O, N, C)x → �2� Si�O2 , N3/4 , C� + �1 − 2� Si. 1.4.1. (1-3). Si-Rich-Oxide Single Layer (SRO-SL) Thin Film Si QDs have been synthesized by several techniques such as microwave-induced or. laser-induced decomposition of silane (SiH 4 )-like precursors [1-14, 1-15], ion implantation of Si+ [1-16, 1-17]. , electrochemical etching of Si wafers [1-18], low pressure chemical vapor deposition. [1-19]. , plasma-enhanced chemical vapor deposition (PECVD) [1-20, 1-21], pulsed-laser deposition. (PLD) of Si [1-22], and sputtering systems [1-8~1-12, 1-22]. In 2004, G.A. Kachurin et al. fabricated Si QD embedded in SiO 2 matrix by implantation of Si ions at a fluence of 1017 cm-2 in thermally SiO 2 layers and by subsequent annealing at 1000 or 1100°C for 2 h. Then P ions were implanted in the layers within the dose range of 1013–1016 cm-2 [1-17]. In 2002, A. A. Gonza´lez-Fernańdez et al.[1-21] fabricated Si QDs embedded in SiO 2 matrix by PECVD with SiH 4 and N 2 O as reactant gas sources. The silicon excess of the layers was controlled by modulating the ratio of the partial pressures produced by the precursor gases in the chamber (P [N2O] /P [SiH4] ). After 9.

(24) deposition, all the PECVD samples were annealed in N 2 atmosphere at 1250°C for 60 min to induce Si nucleation and the formation of Si QD. In 2010, Mota-Pineda et al. [1-23] proposed the SiO x /Si QDs heterolayers which were fabricated employing a radio frequency (RF) magnetron sputtering system. In this study, the SRO thin films are deposited by sputtering the Si target under different oxygen pressures; the nucleation of as-grown crystals is promoted by the rough topography of the oxide films acting as a template. 1.4.2. [SRO/SiO2] Multilayer (ML) Thin Film. In order to obtain a narrow size distribution and more accurate size control, M. Zacharias et al. proposed the SiO/SiO 2 superlattices in 2002 [1-24]. Superlattices have been known since 1970, and they can be manufactured with the epitaxial growth techniques available to III-V compound semiconductor technology [1-25]. In their study, amorphous SiO x /SiO 2 superlattices were prepared by reactive evaporation of SiO powders in oxygen atmosphere. After deposition the samples were annealed at 1100 °C for 1 hour under N 2 atmosphere to obtain the Si QDs. Phase separation and thermal crystallization of SiO/SiO 2 superlattices results in ordered arranged Si QDs. This deposition structure can promise the QD size control and the QD density by tuning the stoichiometry of SRO layers. Besides, in 2009, X. J. Hao et al. [1-26~ 1-28] made their Si QD thin films by sputtering. First, multiple alternative layers of amorphous SRO (SiO x , x<2), and stoichiometric SiO 2 as precursor thin films were deposited by cosputtering with SiO 2 and Si targets and sputtering with single SiO 2 . Then Si QDs were obtained after annealing at 1100°C for 1hr under N 2 atmosphere. The dimension and density of Si QD can be controlled by adjusting the thickness or Si content of amorphous SRO layer.. 10.

(25) (a). (b). (c). (d). Fig. 1- 7 Cross-sectional TEM images of SiO/SiO 2 superlattices: (a) As-prepared SiO/SiO 2 superlattice. The darker layers represent the SiO sublayers. (b) The same film after annealing. The separation of the nanocrystals by a thin oxide shell is clearly visible. (c) High resolution TEM image of the film. For clarity, the visible nanocrystals are highlighted by circles. The crystals are only found in the former SiO layers, which are emphasized by the lines in the image. (d) TEM image of a film with even thinner SiO layers ~2 nm after annealing [1-24].. (b). (c). (a) Fig. 1-8 (a) Schematic diagram of (n-type) Si QDs and (p-type) c-Si heterojunction solar cell. And transmission electron microscopy (TEM) images of Si quantum dots in SiO 2 matrix with (b) low-magnification and (c) high-resolution lattice images for 5 nm Si QDs.. 11.

(26) Fig. 1-9 Four different I–V characteristics of the (n-type) Si QDs/ (p-type) c-Si heterojunction devices.. Table 1-2 One-sun illuminated cell parameters of four different (n-type) Si QDs/ (p-type) c-Si heteroface devices measured at 298 K.. One of important applications of Si QD thin films is photovoltaic devices. In 2009, S. Park et al. demonstrated that the efficiency of solar cells using Si QDs embedded in SiO 2 matrix (shown as Fig. 1-8) is up to 10.58 % [1-29]. In addition, in 2011, S. H. Hong et al.[1-30] proposed that active doping of boron atoms in nanometer silicon layers confined in a SiO 2 matrix. In their study, they demonstrated a p-type Si quantum dot (QD) solar cell with a high energy conversion efficiency (η) of 13.4 % was realized from a [B-doped SiO 1.0 (2 nm)/SiO 2 (2 nm)] 25 superlattices film with a B doping concentration of 4.0 × 1020 atoms cm−3.. 12.

(27) (a). (b). (c). Fig. 1-10 (a) TEM images and (b) SIMS depth profiles of [SiO 2 (8 nm)/B-doped Si(10 nm)] 5 ML film with a B doping level of 1.7 × 1020 atoms cm-3 after annealing at 1100°C for 20 min by rapid thermal annealing.. (a). (b). Fig. 1-11 Representative TEM micrographs of the (a). Fig. 1-12 Current–voltage (I-V) characteristics of a. as-deposited. p-Si QDs/n-Si heterojunction solar cell under air mass. and. (b). annealed. [B-doped. SiO 1.0 /SiO 2 ] 25 superlattice films at 1100 ◦C for 10. 1.5 (AM 1.5G) illumination of 100 mW cm−2.. min.. In 2009, the all Si QD thin film solar cell has been fabricated on a quartz substrate by I. Perez-Wurfl et al. [1-31]. Though the P- and B-doping concentrations are heavy, the dark and illuminated I-V properties of this device are lower than Si QD thin film on Si wafer. The worse electrical properties may result from the barrier is low compared to a Schottky junction on bulk Si, and it is not assigned to a discontinuity in the conduction band but rather represents the. 13.

(28) barrier that the electrons need to overcome to get injected into the depletion region located in the p-type side. On the other hand, it may result from the wide separation between QDs so that carriers can’t transport smoothly.. Fig. 1-13 SIMS data of as-deposited (solid line) and. Fig. 1-14 Dark I-V characteristics of a 0.12 cm2. annealed. diode. Inset shows the I-V characteristics under. (dotted. line). p-i-n. structure.. Significant. interdiffusion is observed after annealing for 1 hour at 1100 °C.. 14. 1-sun illumination with V OC =373 mV..

(29) 1.5. Motivation The efficiency of Si-QD solar cell is still not high enough because there are still many. problems exist in the SL and ML structure. For SRO-SL, it can’t obtain the uniform QD’s size and high QD’s density simultaneously. For ML structure, such as the thickness of barrier layer (if barrier is too thin, it will lose the ability of confinement), interfacial defects at Si QD/SiO 2 matrix, built-in electrical field [1-26~ 1-28], etc. still need to settle. In this study, we propose the gradient Si/O concentration ML (GSRO-ML) structure. After annealing, Si QDs will precipitate at the high Si/O region within a period layer. We expect that, in this structure, the separation between QDs will be reduced and the density of QDs will be increased but not influence the ability of size control. On the other hand, we adopt the highly phosphorus doping Si target and expect the increased conductivity will improve the electrical properties. We will discuss the characteristics about structural, crystalline, optical and electrical properties in this thesis.. 15.

(30) Reference [1-1] B. Chameides,” Statistically Speaking: The Evolution of the Silicon Solar Cell,” 2008 [1-2] M. A. Green, “Third Generation Photovoltaics: Ultra-high Conversion Efficiency at Low Cost,” Prog. Photovolt: Res. Appl., 9, 123-135, 2001 [1-3] G. Conibeer , M. A. Green , R. Corkish , Y. Cho , E. C. Cho, C. W. Jiang, T. Fangsuwannarak, E. Pink, Y. Huang, T. Puzzer, T. Trupke, B. Richards, A. Shalav, K. L. Lin, “Silicon Nanostructures for Third Generation Photovoltaic Solar Cells,” Thin Solid Films, 511 – 512, 654 – 662, 2006 [1-4] Y. Kanemitsu, T. Ogawa, K. Shiraishi, and K. Takeda, “Visible PL from Oxidized Si Nanometer-Sized Spheres: Exciton Confinement on A Spherical Shell,” Phys. Rev. B, vol. 48, pp. 4883- 4886, 1993 [1-5] B. H. Bransden and C. J. Joachain, “Introduction to Quantum Mechanics, 2nd edition,” 2000 [1-6] G. Conibeer, M. A. Green, E.C. Cho, D. Konig, Y. H. Cho, T. Fangsuwannarak, G. Scardera, E. Pink, Y. Huang, T. Puzzer, S. Huang, D. Song, C. Flynn, S. Park, X. Hao, and D. Mansfield,” ,” Thin Solid Film, 516, 6748-6756, 2008. [1-7] M. A. Green, G. Conibeer, D. Konig, E.C. Cho, D. Song, Y. Cho, T. Fangsuwannarak, Y. Huang, G. Scardera, E. Pink, S. Huang, C. Jiang, T. Trupke, R. Corkish, T. Puzzer, “Progress with All Silicon Tandem Cells Based on Silicon Quantum Dots in A Dielectric Matrix.” Proceedings 21st EU PVSEC, Dresden, 2006 [1-8] C. Feser, J. Lacombe, K. V. Maydell and C. Agert," A Simulation Study Towards A New Concept for Realization of Thin Film Triple Junction Solar Cells Based on Group IV Elements," Prog. Photovolt: Res. Appl., 20, 1, 74-81, 2011 [1-9] E. C. Cho, Y. H. Cho, T. Trupke, R. Corkish, G. Conibeer, and M. A. Green, “Silicon Nanostructures for All-Silicon Tandem Solar Cells,” in Proceedings of the 19th European Photovoltaic Solar Energy Conference and Exhibition, 235, Paris, France, 2004. [1-10] S. Lombardo and S. U. Campisano, “Electrical And Optical Properties of Semi-insulating Polycrystalline Silicon Thin Films: The Role of Microstructure And Doping,” Materials Science and Engineering, 17, 8, 281–336, 1996. [1-11] Y. Cho, E. C. Cho, Y. Huang, T. Trupke, G. Conibeer, and M. A. Green, “Silicon Quantum Dots In SiN x Matrix for Third Generation Photovoltaics,” in Proceedings of the 20th European Photovoltaic Solar Energy Conference and Exhibition, p. 47, Barcelona, Spain, June 2005 16.

(31) [1-12] Y. Kurokawa, S. Miyajima, A. Yamada, and M. Konagai, “Preparation of Nanocrystalline Silicon in Amorphous Silicon Carbide Matrix,” Japanese Journal of Applied Physics, 45, 40, L1064–L1066, 2006. [1-13] D. Song, E. C. Cho, Y. H. Cho, G. Conibeer, Y. Huang, S. Huang, M. A. Green, “Evolution of Si (and SiC) nanocrystal precipitation in SiC matrix,” Thin Solid Films, 516, 12, 30, 3824–3830, 2008 [1-14] H. Takagi, H. Ogawa, Y. Yamazaki, A. Ishizaki, and T. Nakagiri,” Quantum Size Effects on Photoluminescence in Ultrafine Si Particles,” Appl. Phys. Lett. 56, 2379, 1990 [1-15] M. Ehbrecht, H. Ferkel, F. Huisken, L. Holz, Y. N. Polivanov, V. V. Smirnov, O. M. Stelmakh, and R. Schmidt,” Deposition And Analysis of Silicon Clusters Generated By Laser‐Induced Gas Phase Reaction,” J. Appl. Phys. 78, 5302, 1995 [1-16] T. S. Iwayama, S. Nakao, and K. Saitoh,” Visible Photoluminescence in Si+‐implanted Thermal Oxide Films on Crystalline Si,” Appl. Phys. Lett. 65, 1814, 1994 [1-17] G.A. Kachurin, S.G. Cherkova, V.A. Volodin, V.G. Kesler, A.K. Gutakovsky, A.G. Cherkov, A.V. Bublikov, and D.I. Tetelbaum,” Implantation of P Ions in SiO 2 Layers with Embedded Si Nanocrystals,” Phys. Res. B, 222(3-4), 497–504, 2004 [1-18] G. Belomoin, J. Therrien, A. Smith, S. Rao, and R. Twesten, S. Chaieb, M. H. Nayfeh, L. Wagner and L. Mitas," Observation of A Magic Discrete Family of Ultrabright Si Nanoparticles," Appl. Phys. Lett., 80, 841, 2002. [1-19] A. Nakajima, Y. Sugita, K. Kawamura, H. Tomita, and N. Yokoyama,” Si Quantum Dot Formation with Low-Pressure Chemical Vapor Deposition,” Jpn. J. Appl. Phys., 35, L189, 1996 [1-20] F. Iacona, G. Franzò, and C. Spinella,” Correlation between Luminescence and Structural Properties of Si Nanocrystals,” J. Appl. Phys. 87, 1295, 2000 [1-21] A. A. Gonza´lez-Fernańdez, J. Juvert, Alfredo Morales-Sańchez, Jorge Barreto, M. Aceves-Mijares, and C. Dominguez,” Comparison of Electrical and Electro-Optical Characteristics of Light-Emitting Capacitors Based on Silicon-Rich Si-Oxide Fabricated By Plasma-Enhanced Chemical Vapor Deposition and Ion Implantation,” J. Appl. Phys. 111, 053109, 2002 [1-22] E. Werwa, A. A. Seraphin, L. A. Chiu, C. Zhou, and K. D. Kolenbrander,” Synthesis And Processing of Silicon Nanocrystallites Using A Pulsed Laser Ablation Supersonic Expansion Method,” Appl. Phys. Lett. 64, 1821, 1994 [1-23] E. Mota-Pineda, M. Meléndez-Lira, M. Zapata-Torres, P. del Angel, A. Pérez-Centeno, S. Jiménez-Sandoval, and M. A. Santana-Aranda,” Photoluminescence of As-Grown And Thermal Annealed SiO x /Si-Nanocrystals Heterolayers Grown by Reactive RF Sputtering,” J. Appl. Phys. 108, 094323, 2010 [1-24] M. Zacharias, J. Heitmann, R. Scholz, U. Kahler, M. Schmidt, and J. Bläsing, 17.

(32) “Size-Controlled Highly Luminescent Silicon Nanocrystals: A SiO/SiO 2 Superlattice Approach,” Appl. Phys. Lett., 80, 4, 661–663, 2002. [1-25] L. L. Chang, L. Esaki, W. E. Howard, and R. Ludeke, “The Growth of A GaAs-GaAlAs Superlattice,” J. Vac. Sci. Technol, 10, 1, 11–16, 1973. [1-26] X. J. Hao, E. C.Cho, C. Flynn, Y. S. Shen, S. C. Park, G. Conibeer, and M. A. Green,” Synthesis and Characterization of Boron-Doped Si Quantum Dots for All-Si Quantum Dot Tandem Solar Cells,” Sol. Energy Mater. Sol. Cells, 93, 273-279, 2009 [1-27] X. J. Hao, E. C. Cho, G. Scardera, E. Bellet-Amalric, D. Bellet, Y. S. Shen, S. Huang, Y. D. Huang, G. Conibeer, and M. A. Green,” Effects of Phosphorus Doping on Structural and Optical Properties of Silicon Nanocrystals in A SiO 2 Matrix,” Thin Solid Films, 517, 5646, 2009 [1-28] X. J. Hao, E. C. Cho, G. Scardera, Y. S. Shen, E. Bellet-Amalric, D. Bellet, G. Conibeer, and M. A. Green,” Phosphorus-Doped Silicon Quantum Dots for All-Silicon Quantum Dot Tandem Solar Cells,” Sol. Energy Mater. Sol. Cells, 93, 1524, 2009 [1-29] S. Park , E. C. Cho, D. Song, G. Conibeer, M. A. Green, “N-Type Silicon Quantum Dots and P-Type Crystalline Silicon Heteroface Solar Cells.” Sol. Energy Mater. Sol. Cells, 93, 684–690, 2009 [1-30] S. H. Hong, Y. S. Kim, W. Lee, Y. H. Kim, J. Y. Song, J. S. Jang, J. H. Park, S. H. Choi and K. J. Kim,” Active Doping of B in Silicon Nanostructures and Development of A Si Quantum Dot Solar Cell,” Nanotechnology, 22, 425203, 2011 [1-31] I. Perez-Wurf, X. Hao, A. Gentle, D. Kim, G. Conibeer, and M. A. Green," Si Nanocrystal p-i-n Diodes Fabricated on Quartz Substrates for Third Generation Solar Cell Applications," Appl. Phys. Lett., 95, 153506, 2009. 18.

(33) Chapter 2. Fabrication of Gradient Si-Rich Oxide Multilayer. (GSRO-ML) Structure for High Density Si QD Thin Film In this chapter, we introduce the fabrication process of our samples. The process is shown in Fig. 2-1. In this study, we use two different substrates to analyze the characterization of GSRO thin films. One is Si (100) wafer for photoluminescence (PL) analysis, X-ray diffraction, transmission electron microscopy (TEM), and electrical measurements. The other one is quartz for Raman in order to avoid the crystalline Si signal from Si wafer at 520cm-1 in Raman spectra and for UV/Vis/NIR analysis.. Fig. 2-1 Fabrication process of GSRO thin films.. 2.1. Substrate Clean Before depositing thin films, the first step as shown in Fig. 2-1 is substrate clean. Si wafers. are cleaned by standard RCA cleaning, and the standard process is shown in Fig. 2-2(a). The main purpose of RCA cleaning is to remove the particle, organic contaminates and native oxide on the wafer. In the cleaning process of quartz, we clean quartz in acetone and ethanol for 10min by ultrasonic cleaner to remove organic contaminates on the quartz surface. The process is shown in Fig. 2-2(b).. 19.

(34) (a) (b) Fig. 2-2 Clean process chart of (a) Si wafer and (b) quartz substrates.. 20.

(35) 2.2. GSRO-ML Thin film Deposition. 2.2.1. Principle of Radio Frequency (RF) Magnetron Sputtering Process The basic principle of sputtering is to accelerate the ion to bombard the target surface, after. ion and atomic in solid surface exchange the momentum, atoms will spill from the solid surface, this phenomenon called sputtering. Sputtering mainly depend on the state of the plasma ions and free radicals. Plasma is also known as the fourth state of matter. Plasma’s creation is similar to phase change in matter. By applying enough energy (like RF or microwave), gas can be broken down into plasma. In this state the plasma contains charged atoms, particles, ions and free radicals. Plasma is very chemically reactive due to its high energy state, making it very useful for changing the properties of material. To ignite the plasma of the sputtering gas, cathode should be added to hundreds of volts. The bias added on cathode relative to the anode is negative, it shows when Ar atoms become Ar+ ions, and they will be accelerated and impacted target, after collision, the atoms on the target surface flight and deposited on the substrate, that’s the principle of sputtering (Fig. 2-3).. Fig. 2-3 Illustration of Operation of magnetron sputtering deposition method. 21.

(36) 2.2.2. GSRO-ML Thin Film To deposit GSRO-ML thin films, co-sputtering by Si and SiO 2 targets are needed. During. deposition, we fixed the power of SiO 2 in the low power, and tune the sputtering power of Si from 30 W to 110 W and to 30W by 1W/sec for 20 cycles, and the schematic diagram is shown in Fig. 2-4. By this method, we obtain a periodical gradient SRO distribution (Fig. 2-5).. Fig. 2-4 The variation of Si concentration.. Fig. 2-5 Scheme of as-deposited GSRO-ML. 22.

(37) 2.3. Post-annealing Process After deposition, the GSRO-ML thin films were treated with high temperature annealing. process to precipitate Si atoms due to phase separation of Si and SiO 2 according to the following equation: annealing x. x. SiOx �⎯⎯⎯⎯⎯� 2 SiO2 + �1 − 2� Si. (2- 1). Here we annealed all samples at 1100°C for different durations in quartz furnace to form Si QD (Si QD).. Fig. 2-6 Scheme of as-deposited GSRO-ML structure.. 2.4. Thermal Oxide Layer Remove After annealing process, a thin thermal oxide layer was formed on the top and bottom of. samples due to residual oxygen in the furnace. The thermal oxide layers make an influence on the collection efficiency of photo-generated carriers. In order to reduce the influence of the thermal oxide layers, we remove top side oxide layer by CHF 3 :O 2 reactive ion etching (RIE) to and bottom oxide layers by buffered oxide etch (BOE).. 23.

(38) 2.5. Electrode Layer Deposition Finally, contact electrodes were deposited on samples for the electrical properties. measurement. We deposited Al layers on both top and bottom sides to form ohmic contact. The top electrodes are designed to square, as shown in Fig. 2-7. And another pattern which is for efficiency measurement is composed of a 5 mm*0.1 mm rectangle, the bar is a 0.2 mm* 4 mm rectangle, the pad is a 1.2 mm* 0.7 mm rectangle and the spacing is 0.29 mm (as shown in Fig. 2-7(b)).. (a). (b). Fig. 2-7 Schemes of top electrodes of (a) square and (b) finger patterns deposited by thermal evaporation coater.. 24.

(39) Chapter 3. Experimental Equipments and Analyzed Methods. In order to understand the characterization of the Si QD thin films. Some analyzed equipments are used, and the details are explained as below.. 3.1. Raman Scattering Spectrum Confocal-Raman microscope is a powerful characterization technique to study materials’. vibration modes in a material. It’s based on the Raman Effect. [3-1]. , which is the inelastic. scattering of photons and molecules when the incident photon interact with the molecules, the photons transit from the ground state to a virtual excited state. If the energy isn’t absorbed by the molecules, it’s released through the scattering method. Here shows the three different signal produced by the incident light interaction with the specimen in the Fig. 3-1. The released energy which is equal to the energy of the incident photons is called Raman scattering [3-1], and the incident light which interact with acoustic phonon is called Brilliouim scatter. Therefore, Raman microscope is a powerful and non-destructive technique to understand the materials’ physical and chemical properties. We analyzed the samples by high-resolution confocal Raman microscope (Lab RAM HR Raman Microscope), and used a 488-nm diode-pumped solid-state (DPSS) laser. The illuminated spot size is about 10um in diameter and the power of laser is about 7mW. Si substrate was used to calibrate the crystalline Si signal at 520 cm-1 before measuring our samples.. 25.

(40) Fig. 3-1 Illustration of Rayleigh and Raman. Fig. 3-2 Three components decomposed from Raman. scattering. spectra of Si QD thin films[3-2]. Generally, three peaks can be detected in Si QD thin films, including amorphous phase (~480 cm-1), intermediate phase (500 - 510 cm-1) and crystalline phase (510 – 520 cm-1) (Fig. 3-2) [3-2]. The crystallinity (χ c ) can be estimated by the following equation: χC = I. Inc +Ii. nc +Ii +Ia. × 100%. (3- 1). where I nc is the intensity of crystalline phase, I i and Ia represent the intensities of intermediate phase and amorphous mode, respectively. In addition, peaks shift away from 520 cm-1 and the full width at half maximum (FWHM) of Raman spectra can be utilized to roughly estimate the dimension of Si NC because of the phonon confinement effect, as shown in Fig. 3-3. [3-3].. Phonon confinement mode can be. expressed by the following equation: sin[(qD/a)π]2. L(w, D) ∝ ∫ [1−(qD/a)2 ]2 [w. dq 2 +(Γ/2) (q)−w] opt. (3- 2). Where L(ω,D) is the lineshape of phonons confined to a hard sphere of diameter D, a is lattice parameter of Si, and Γ is the damping parameter [3-4].. 26.

(41) Fig. 3-3 FWHM of the Raman peaks and Si QD size against the corresponding Raman shift. 27.

(42) 3.4. Photoluminescence (PL) Spectrum Photoluminescence (PL) spectroscopy is a contactless, nondestructive method of probing. the electrical structure of materials. Light is incident directly onto samples where it’s absorbed and imparts excess energy into the materials in a process called photo-excitation as shown in Fig. 3-4.. Fig. 3-4 Schematic of electronic transition.. The samples were analyzed by micro-PL measured on high-resolution Raman microscope (Lab RAM HR Raman Microscope). We used a laser on diode-pumped solid-state (DPSS) with a 488nm wavelength. The illuminated spot size is about 10um in diameter and the power of laser is about 7 mW. Si substrate is used to calibrate the laser signal at 488nm before PL spectroscopy measurement. All PL spectra were measured at room temperature (RT).. 28.

(43) 3.5. Ultraviolet/visible/Near-infrared (UV/VIS/NIR) Spectrum In this section, we used UV/Vis/NIR spectrophotometer (Hitachi U-4100, Japan) to. measure the transmittance (%T) and reflection (%R) of our thin films. We also can calculate the absorbance by Absorbance (%) = 100-T-R (%). (3- 3). Thus, we can understand the optical properties of the thin films. And, the absorption coefficient α can be got by the following relation Α (cm-1) = A/t. (3- 4). where t is the thickness of thin films. The optical bandgap (E g,opt ) of the thin films are determined by the intercept of linear part of the absorption edge to αhν=0 in the relationship as [3-5]. (αhν)γ = B(hν – E g,opt ). (3- 5). Where h is Plank’s constant, ν is the frequency of the radiation, and B is the edge width parameter. The value of r is dependent on the E g,opt behavior, γ=1/2 for indirect E g,opt and γ=2 for direct E g,opt .. 3.6. Current-Voltage (I-V) Curve The I-V curves in this study were measurement by the E5270B 8-slot precision. measurement mainframe (Agilent Technologies) and a halogen lamp illumination with power density of 1 mW/cm2 was applied to photo-response measurements. From the electrical properties, we can understand the rectification ratio, conductivity, and photo-responsive properties of our samples.. 29.

(44) Reference [3-1] C. V. Raman ,and K. S. Krishna, “A New Type of Secondary Radiation,” Nature, 121, 501, 1928 [3-2] Qijin Cheng; Tam, E.; Shuyan Xu; Ostrikov, K.," Si Quantum Dots Embedded in an Amorphous SiC Matrix: Nanophase Control by Non-Equilibrium Plasma Hydrogenation," Nanoscale, 594–600, 2010 [3-3] G. Faracil, S. Gibiliscol, P. Russol, A. R. Pennisil, G. Gompagnini, S. Battiato, R. Puglisi, and S. La Rosa “Si/SiO 2 Core Shell Clusters Probed by Raman Spectroscopy” Eur. Phys. J. B, 46, 457-461 ,2005 [3-4] S. Hernández, A. Martínez, P. Pellegrino, Y. Lebour, B. Garrido, E. Jordana, and J. M. Fedeli, “Silicon Nanocluster Crystallization in SiO x Films Studied by Raman Scattering,” J. Appl. Phys., 104, 044304, 2008 [3-5] D. song, E. C. Cho, G. Conibeer, C. Flynn, Y. Huang, M. A. Green, “Structural Electrical And Photovoltaic Characterization of Si Nanocrystals Embedded SiC Matrix And Si Nanocrystals/C-Si Heterojunction Devices”, Sol. Energy Mater. Sol. Cells, 92, 474-480, 2008. 30.

(45) Chapter 4. Results and Discussion. In this chapter, we discuss the GSRO-ML thin films with and without nucleation layer (NL), shown as Fig. 4-1. The Raman, PL, and UV/Vis/NIR spectra and TEM images and I-V curves are investigated for understanding the characteristics of the GSRO-ML thin films.. (a). (b). Fig. 4-1 Schemes of GSRO-ML thin films (a) without NL and (b) with NL. 4.1. GSRO-ML Thin Films without NL. 4.1.1. Crystalline Properties of GSRO-ML Thin Films without NL. So far, a GSRO-ML structure for the Si QD thin film hasn’t been studied; hence, it’s important to tune the deposition parameters for uniform QD’s size and high QD’s density formations. In the beginning, the gradient O/Si ratio is modified by tuning the SiO2 sputtering power for the good NC-Si formation and suitable SRO composition. Table 4-1 shows the sputtering parameters of the GSRO-ML thin films without NL under different SiO2 sputtering powers. The n-type Si sputtering power (P n-Si ) is periodically tuned from 30 to 110 and back to 30 W by rate of 1 W/sec while P SiO2 is fixed at 10, 20, or 30 W for 20 periods (as shown in Fig. 2-4 and Fig. 2-5). Each GSRO thin layer thickness is about 2.5~3 nm. In order to understand 31.

(46) the NC-Si properties, the samples are annealed at 1100°C in N 2 ambient for 20 minutes after deposition. To know the suitable parameters for the obvious NC-Si formation, the Raman spectra of the GSRO-ML thin films without NL under different P SiO2 are measured, as shown in Fig. 4-2. The signals can be decomposed into three components, including amorphous- (a-), intermediate- (i-Si), and nano-crystalline (NC-) Si phases, the curve-fitting results and the O/Si ratio form XPS measurements are listed in Table 4-2. The NC-Si intensities and crystallinity (C Si ) are obviously increased when the P SiO2 is decreased from 30 to 10 W, the results are well matched with the average O/Si ratio from XPS measurements. Hence, it indicates a low P SiO2 is necessary for the good NC-Si properties in a GSRO-Ml structure. To avoid the over-diffused Si-rich atoms during annealing, the P SiO2 of 20 W with medium O/Si ratio and obvious NC-Si intensity is used in the next experiments for preserving the QD’s size control ability.. Table 4-1. Sputtering parameters of the GSRO-ML thin films without NL under different SiO 2 sputtering powers.. Sample. Sputtering power. ID. P n-Si (W). P SiO2 (W). G30. Min.: 30 W. 30. G20. Max.: 110 W. 20. G10. Rate: 1 W/sec. 10. 32.

(47) Fig. 4-2 Raman spectra of GSRO-ML thin films without NL under different SiO 2 sputtering powers.. Table 4-2 O/Si ratio from XPS measurements and the curve-fitting results from Fig. 4-2 for the NC-Si properties.. XPS Sample ID. NC-Si properties after annealing from Raman spectrum. Ave. O/Si. Peak Position. FWHM. C Si. Intensity. (%). (cm-1). (cm-1). (%). (a.u.). G30. 0.84. 517.4. 10.0. 42.5. 8.6×103. G20. 0.54. 515.4. 10.3. 67.4. 4.4×103. G10. 0.39. 515.0. 8.8. 82.3. 5.1×103. 33.

(48) 4.2-1. GSRO-ML Thin Films with NL. In order to obtain the high density Si QD thin films with uniform QD’s size, the NLs, which are co-sputtered by P n-S i of 110 W and P SiO 2 of 20 W for 1 nm thickness, are inserted into the GSRO-ML thin films in the centers of each GSRO thin-layer to enhance the localized Si-rich atoms aggregation ability during annealing, as shown in Fig. 4-3 and Fig. 4-4. The sputtering parameters of the GSRO-ML thin film with NL (Sample ID: G20-NL) are also listed in Table 4-3. Here a more suitable annealing duration time for a GSRO-ML structure is also studied for the Si QD thin films with better electro-optical properties. The Raman spectra of GSRO-ML with and without NL are shown in Fig. 4-5. The NC-Si intensity of G20-NL is increased by about two times compared with that of G20, it means the inserted NLs in GSRO-ML can efficiently enhance the NC-Si properties after annealing.. Fig. 4-3 Variations of Si and SiO 2 sputtering powers for. Fig. 4-4 Scheme of the GSRO-ML thin film with NL.. each GSRO layer.. Table 4-3 Sputtering parameters of the GSRO-ML thin film with NL (Sample ID: G20-NL).. Sputtering power (W) Gradient layer (GL) P n-Si (W). Nucleation layer (NL). P SiO2 (W). P n-Si (W). P SiO2 (W). 20. 110. 20. Min.: 30W Max.: 110W Rate: 1 W/sec. 34.

(49) Fig. 4-5 Raman spectra of G20 and G20-NL.. 4.2-1-1. Crystalline Properties of GSRO-ML Thin Films with NL. Fig. 4-6. shows the Raman spectra of G20-NL and the corresponding curve-fitting results. under different annealing time are listed in Table 4-4. The higher C Si and narrower FWHM are observed with increasing the annealing time; it means the larger average NC-Si QDs’ size with better Si crystal quality is obtained by a longer annealing time.. Fig. 4-6 Raman spectra of G20-NL under different annealing time.. 35.

(50) Table 4-4 Curve-fitting results of Fig. 4-6 for the NC-Si properties.. Annealing. Peak Position. FWHM. Intensity. C Si. time. (cm-1). (cm-1). (a.u.). (%). 10min. 511.8. 10.9. 9.4×103. 67.3. 20min. 512.1. 10.4. 1.0×104. 72.3. 30min. 512.8. 10.0. 1.1×104. 74.5. 60min. 514.0. 9.8. 1.1×104. 75.4. 4.2-1-2. Optical Properties of GSRO-ML Thin Films with NL. The absorption spectra in Tauc’s plot for G20-NL under different annealing time are examined for the optical properties confirmations, as shown in Fig. 4-7, and the corresponding optical bandgap (E g,opt ) and the absorption coefficient (α) are also listed in Table 4-6. The E g,opt is increased by the longer annealing time due to a larger average Si QD size formation as observed in Table 4-5, besides, the α values of G20-NL annealed for 60 minutes is also significantly larger than that for 20minutes. The results represent that a longer annealing time can improve the optical absorption properties of the Si QD thin films in a GSRO-ML structure.. Fig. 4-7 Absorption spectra in Tauc’s plot for G20-NL under different annealing time.. 36.

(51) Table 4-6 Curve-fitting results for optical bandgap (E g,opt ) and absorption coefficient (α) from Fig. 4-7.. 4.2-1-3. Annealing time. E g,opt (eV). α (cm-1). 20min. 2.02. 2.2×103. 60min. 1.94. 3.3×103. Electro-Optical Properties of GSRO-ML Thin Films with NL. To confirm the electro-optical properties of the GSRO-ML thin films under different annealing time, the I-V curves of G20-NL with and without a halogen lamp illumination about 1 mW/cm2 of power density are shown in Fig. 4-8. The corresponding parameters are also listed in Table 4-7. G20-NL annealed for 60 minutes clearly shows not only a higher dark conductivity but also better photo-response properties, including V OC and I SC values, than that for 20 minutes. It may be contributed from the better Si crystal quality and optical absorption properties in the longer annealing time. Therefore, our results indicate that a long annealing time such as 60 minutes for the Si QD thin films using a GSRO-ML structure is more suitable for the better electro-optical properties.. (a). (b). Fig. 4-9 I-V curves of G20-NL on p-type Si(100) wafer with and without a halogen lamp illumination under an annealing duration time of (a) 20 and (b) 60 minutes.. 37.

(52) Table 4-7 Parameters of the electro-optical properties of G20-NL under different annealing time. A halogen lamp with 1 mW/cm2 of power density is used as the illumination source. Annealing. Conductivity. V OC. I SC. time. (Ω-cm)-1. (mV). (A). 20min. 2.4×10-7. 269. 2.0×10-8. 60min. 4.4×10-6. 298. 4.2×10-7. 38.

(53) 4.2-2. Comparison of Using GSRO-ML, [SRO/SiO2]-ML, and SRO-SL. Structures In this section, the nano-structural and electro-optical properties of G20-NL using a GSRO-ML structure are compared to those of using [Si-rich oxide/SiO 2 ] multilayer ([SRO/SiO 2 ]-ML) and Si-rich oxide single layer (SRO-SL) structures, as shown in Fig. 4-4. The SRO-SL is deposited by co-sputtering n-Si target with P n-Si of 110 W and SiO 2 target with P SiO2 of 20 W for 80 nm of thickness equal to that of G20-NL. For [SRO/SiO 2 ]-ML, the SRO layers are deposited by co-sputtering n-Si target with P n-Si of 110 W and SiO 2 target with P SiO2 of 10 W, and each SRO and SiO 2 layer is individually fixed at 5 and 2.5 nm for 20 periods. All these samples are annealed at 1100°C for 60 minutes in N 2 ambient after deposition.. (a). (b). Fig. 4-10 Schemes of (a) [SRO/SiO 2 ]-ML and (b) SRO-SL deposition structures. 4.2-2-1. Crystalline Properties of GSRO-ML, [SRO/SiO2]-ML, and SRO-SL Structures. The nano-crystalline properties of using three different structures are examined by Raman spectra, as shown in Fig. 4-11, and the corresponding curve-fitting results are listed in 39.

(54) Table 4-8. Among these three different structures, [SRO/SiO 2 ]-ML shows the lowest NC-Si intensity due to the highest average O/Si ratio than that of G20-NL or SRO-SL. However, G20-NL reveals the highest NC-Si intensity although it has a higher average O/Si ratio than SRO-SL. It represents using a GSRO-ML structure can enhance the Si-rich atoms aggregation to more efficiently form NC-Si during annealing. Besides, we notice that SRO-SL with a highly Si-rich oxide composition shows two peaks for c-Si phase not including the a- and i-Si phases. It means the size distribution of Si QDs formed during annealing is quite wide and the QD’s size control ability is lost in a highly Si-rich SRO-SL structure. For [SRO/SiO 2 ]-ML and G20-NL, only one peak for nc-Si phase is obtained, hence, it represents that the good size control ability can be obtained by using both deposition structures. The close FWHM also means the similar average QD size in both samples, however, the C Si of G20-NL is obviously higher than that of [SRO/SiO 2 ]-ML. Hence, compared to SRO-SL and [SRO/SiO 2 ]-ML, using a GSRO-ML structure has not only the NC-Si size control ability but also a better Si crystal quality.. Fig. 4-11 Raman spectra of [SRO/SiO 2 ]-ML, G20-NL, and SRO-SL thin films.. 40.

(55) Table 4-8 Curve-fitting results of Fig. 4-11.. Sample ID. Peak position Peak FWHM Intensity Crystallinity (cm-1). (cm-1). (a.u.). (%). [SRO/SiO 2 ]-ML. 516.1. 9.5. 1.9×103. 53.1. G20-NL. 514.0. 9.8. 1.1×104. 75.4. 513.9. 11.2. 516.5. 5.9. 6.4×103. 97.3. SRO-SL. 4.2-2-2. NC-Si properties after annealing. Structural properties of GSRO-ML and [SRO/SiO2]-ML Structures. To further understand the difference in the nano-structural properties between [SRO/SiO 2 ]-ML and G20-NL thin films, the high-resolution transmission electron microscope (TEM) images were examined by a JEOL JEM-2010F transmission electron microscope. Fig. 4-12 shows the cross-sectional high-resolution TEM images of the annealed [SRO/SiO 2 ]-ML and G20-NL ML thin films. From TEM images, the QD’s size is similar but the QD’s density of G20-NL is clearly higher than [SRO/SiO 2 ]-ML, this result is well matched with Raman spectra. In order to observe the size control abilities of both structures, we gather the QD’s sizes from different regions shown in Fig. 4-13. The good size control ability can be observed in both samples since most QDs’ sizes are located at 5±1 nm. However, the QD’s density of G20-NL about 2.58×1012 cm-2 is significantly higher than [SRO/SiO 2 ]-ML about 1.04×1012 cm-2 by over 2 times. Therefore, the results demonstrate a GSRO-ML can also obviously increase the Si QD’s density under preserving the Si QD size control ability.. 41.

(56) (b). (a). Fig. 4-12 High-resolution TEM images of (a) [SRO/SiO 2 ]-ML and (b) G20-NL thin films.. Fig. 4-13 QD’s size distribution of [SRO/SiO 2 ]-ML and G20-NL thin films.. 4.2-2-3. Optical Properties of GSRO-ML, [SRO/SiO2]-ML, and SRO-SL Structures. In addition to crystalline properties, the optical properties are also investigated by PL and UV/Vis/NIR spectra. Fig. 4-14 shows the PL spectra of [SRO/SiO 2 ]-ML and G20-NL ML thin film. From literatures, the PL emission can be possibly contributed from three mechanisms, including interface states at the interfacial region between QDs and matrix [4-1, 4-2], quantum confinement (Q.C.) [4-3, 4-4] effect of Si QDs, and defect states inside SiO 2 matrix [4-5, 4-6]. The peak positions of the curve-fitting results from PL spectra are listed in Table 4-9. From Fig. 4-14, we observe that the emission intensity from Q. C. effect of G20-NL is lower than that of [SRO/SiO 2 ]-ML. In order to confirm the cause, the I-V curves of both samples under a 488. 42.

(57) nm laser illumination are measured, as shown in Fig. 4-15, G20-NL shows more obvious photovoltaic properties than [SRO/SiO 2 ]-ML. It means that more photo-generated carriers can transport through QDs rather than recombine inside QDs due to the reduced QD’s separation in G20-NL as observed in TEM images. On the other hand, the integrated intensity ratio of the emission from oxygen-related defect of G20-NL is higher than that of [SRO/SiO 2 ]-ML, it may be contributed from the all Si-rich oxide materials used in G20-NL. It’s helpful for the PV properties of the Si QD thin films since the oxygen-related defects in SiO 2 matrix can enhance the carrier’s transportation efficiency.. Fig. 4-14 PL spectra of [SRO/SiO 2 ]-ML and G20-NL.. Fig. 4-15 I-V curves of [SRO/SiO 2 ]-ML and G20-NL under a 488 nm laser illumination.. Table 4-9 Curve-fitting results of Fig. 4-14.. Oxygen-related defect Sample ID. Position (nm). Integrated intensity ratio (%). [SRO/SiO 2 ]-ML. 604. 3.1. G20-NL. 577. 62.3. Q.C. effect Position (nm) 739 (1.68 eV) 705 (1.76 eV). 43. Integrated intensity ratio (%). Interfacial state Position (nm). Integrated intensity ratio (%). 83.3. 849. 13.6. 35.9. 816. 1.2.

(58) Fig. 4-16. shows the (αhν)1/2 versus (hν) plots of [SRO/SiO 2 ]-ML and G20-NL, and the. corresponding E g,opt and α values are listed in Table 4-10. The E g,opt value of G20-NL is slightly larger than that of [SRO/SiO 2 ]-ML matches with the results of the PL signals from Q. C. effect, which are peaks located at 1.68 eV for [SRO/SiO 2 ]-ML and 1.76 eV for G20-NL. It represents the effective E g of G20-NL is surely higher than [SRO/SiO 2 ]-ML although both samples have the quite close distribution of QD’s size. It may be originated from the difference in QD’s surface structure or density. Besides, the α value of G20-NL is obviously higher than that of [SRO/SiO 2 ]-ML due to the higher QD’s density obtained. It indicates a thinner film thickness is required for SC application integrating Si QD thin films by using a GSRO-ML structure.. Fig. 4-16 (αhν)1/2 versus hν plots of [SRO/SiO 2 ]-ML and G20-NL thin films.. Table 4-10 Curve-fitting results of [SRO/SiO 2 ]-ML and G20-NL thin films.. Sample. E g,opt. α. ID. (eV). (cm-1). [SRO/SiO 2 ]-ML. 1.83. 3.05×102. G20-NL. 1.94. 3.28×103. 44.

(59) 4.2-2-4. Electrical Properties. To understand the difference of the electro-optical properties of G20-NL and [SRO/SiO 2 ]-ML, the dark and light I-V curves are measured under a halogen lamp illumination with power density of ~1 mW/cm2 as shown in Fig. 4-17. The corresponding parameters of electro-optical characteristic are listed in Table 4-11. Compared with [SRO/SiO 2 ]-ML, G20-NL shows the better photo-response behavior including the increased V OC by 180 mV and I SC by over ten times. As the previous results in TEM images, the improved photo-response properties in G20-NL can be attributed to the significantly higher QD’s density, which also means the reduced QD’s separation than that in [SRO/SiO 2 ]-ML.. (a). (b). Fig. 4-17 Dark and light I-V curves of (a) [SRO/SiO 2 ]-ML and (b) G20-NL on p-type Si wafers.. Table 4-11 Parameters of the photo-response properties of [SRO/SiO 2 ]-ML and G20-NL under a halogen lamp illumination. Sample ID. V OC (mV). I SC (mA). [SRO/SiO 2 ]-ML. 118. 3.5×10-8. G20-NL. 298. 4.2×10-7. 45.