激光限制结晶技术制备nc-Si/SiO2多层膜

在等离子体增强化学气相淀积系统中，采用aSi:H层淀积和原位等离子体氧化相结合的逐层生长技术制备了aSi:H/SiO_2多层膜.在激光诱导限制结晶原理基础上，使用KrF准分子脉冲激光为辐照源，对aSi:H/SiO_2多层膜进行辐照，使纳米级厚度的aSi:H子层晶化.Raman散射谱和电子衍射谱的结果表明，经过激光辐照后纳米Si颗粒在原始的aSi:H子层内形成，晶粒尺寸可以根据aSi:H层的厚度精确控制.还研究了样品的光致发光（PL）特性以及激光辐照能量密度对PL性质的影响. 关键词： 脉冲激光 多层膜 限制结晶     

尺寸可控的纳米硅的生长模型和实验验证

基于经典热力学理论，对a-SiN_x/a-Si:H/a-SiN_x三明治结构或a-Si:H/a-SiN_x多层膜结构中纳米硅成核，以及从球形到鼓形的生长过程进行了研究. 建立了限制性晶化理论模型：在纳米硅生长过程中，由于界面能增大将导致生长停止，给出限制性晶化条件——a-Si:H子层厚度小于34 nm. 在激光晶化和常规热退火两种方法形成的a-SiN_x/nc-Si/a-SiN_x三明治结构和nc-Si/a-SiN_x多层膜结构中验证了该理论模型. 关键词： 非晶硅 纳米硅 激光辐照 结晶     

Nickel-disilicide-assisted excimer laser crystallization of amorphous silicon

Polycrystalline silicon (poly-Si) thin film has been prepared by means of nickel-disilicide (NiSi多晶硅 受激准分子激光器结晶 结晶化 界面晶粒生长polycrystalline silicon, excimer laser crystallization,Ni-disilicide, Ni-metal-induced lateral crystallization, two-interface grain growthProject supported by the National High Technology Development Program of China (Grant No 2002AA303250) and by the National Natural Science Foundation of China (Grant No 60576056).9/7/2005 12:00:00 AM3/6/2006 12:00:00 AMPolycrystalline silicon （poly-Si） thin film has been prepared by means of nickel-disilicide （NiSi2） assisted excimer laser crystallization （ELC）. The process to prepare a sample includes two steps. One step consists of the formation of NiSi2 precipitates by heat-treating the dehydrogenated amorphous silicon （a-Si） coated with a thin layer of Ni. And the other step consists of the formation of poly-Si grains by means of ELC. According to the test results of scanning electron microscopy （SEM）, another grain growth model named two-interface grain growth has been proposed to contrast with the conventional Ni-metal-induced lateral crystallization （Ni-MILC） model and the ELC model. That is, an additional grain growth interface other than that in conventional ELC is formed, which consists of NiSi2 precipitates and a-Si. The processes for grain growth according to various excimer laser energy densities delivered to the a-Si film have been discussed. It is discovered that grains with needle shape and most of a uniform orientation are formed which grow up with NiSi2 precipitates as seeds. The reason for the formation of such grains which are different from that of Ni-MILC without migration of Ni atoms is not clear. Our model and analysis point out a method to prepare grains with needle shape and mostly of a uniform orientation. If such grains are utilized to make thin-film transistor, its characteristics may be improved.     

Ni-Si多层膜中固态非晶化反应的热力学与动力学过程

研究了Ni/非晶Si(a-Si)多层组分调制膜(ML)中界面上的固相反应。根据计算的亚稳自由能图,对界面上固相反应形成非晶态进行了热力学与动力学的分析。提出了Ni/a-Si多层膜中的固相反应的动力学模型。 关键词：     

Quantum confinement in amorphous silicon quantum dots embedded in silicon nitride

Amorphous silicon quantum dots (a-Si QDs) were grown in a silicon nitride film by plasma enhanced chemical vapor deposition. Transmission electron micrographs clearly demonstrated that a-Si QDs were formed in the silicon nitride. Photoluminescence and optical absorption energy measurement of a-Si QDs with various sizes revealed that tuning of the photoluminescence emission from 2.0 to 2.76 eV is possible by controlling the size of the a-Si QD. Analysis also showed that the photoluminescence peak energy E was related to the size of the a-Si QD, a (nm) by E(eV) = 1.56+2.40/a(2), which is a clear evidence for the quantum confinement effect in a-Si QDs.     

Polycrystalline silicon films prepared by metal-induced crystallisation using pre- and post-deposited aluminium on amorphous silicon

We report on the successful fabrication of polycrystalline silicon films by aluminium-induced crystallisation (AIC) of Radio frequency (rf) plasma-enhanced chemical vapour deposited (PECVD) a-Si films. The effects of annealing at different temperatures (300 and 400°C), below the eutectic temperature of the Si–Al binary system, on the crystallisation process have been studied. This work emphasises the important role of the position of the Al layer with respect to the Si layer on the crystallisation process. The properties of the crystallised films were characterised using X-ray diffraction, Raman spectroscopy, ellipsometry, field-emission scanning electron microscopy (FESEM) and atomic force microscopy (AFM). With an increase in the annealing temperature, it was found that the degree of crystallisation of annealed a-Si/Al and Al/a-Si films increased. The results showed that the arrangement where the Al was on top of the a-Si had a more prominent effect on crystallisation enhancement than when Al was below the a-Si. The interfacial layer between the Al and a-Si film is crucial because it influences the layer-exchange process during annealing. The oxide layer formed between the Al and the a-Si layers greatly retards the crystallisation process in the case of the Al/Si arrangement. Our investigations suggest that polycrystalline Si films formed by AIC can be used as a seed layer in solar cell fabrication.     

Experimental determination of the energy dependence of electron inelastic mean free path in silicon oxide and silicon nitride

The energy dependence of the electron inelastic mean free path, λ(E), in silicon oxide and silicon nitride is experimentally determined via Auger electron spectroscopy according to Auger signal attenuation with varying film thickness. Silicon-oxide- and silicon-nitride films are formed on different metal substrates by means of plasma-enhanced chemical vapor deposition. Analysis of the results and their comparison with theoretical data indicate that, in the chosen material, variations in the electron mean free path versus their energy are estimated more reliably by means of experiments than through the use of universal theoretical curves. It appears that the results obtained in this work can help in the more accurate determination of the width and location of interfaces in multilayer structures.     

Excitonic photoluminescence characteristics of amorphous silicon nanoparticles embedded in silicon nitride film

We have investigated the photoluminescence (PL) properties of amorphous silicon nanoparticles (a-Si NPs) embedded in silicon nitride film (Si-in-SiN_x) grown by helicon wave plasma-enhanced chemical vapor deposition (HWP-CVD) technique. The PL spectrum of the film exhibits a broad band constituted of two Gaussian components. From photoluminescence excitation (PLE) measurements, it is elucidated that the two PL bands are associated with the a-Si NPs and the silicon nitride matrix surrounding a-Si NPs, respectively. The existence of Stokes shift between PL and absorption edge indicates that radiative recombination of carriers occurs in the states at the surface of the Si NPs, whereas their generation takes place in the a-Si NPs cores and the silicon nitride matrix, respectively. The visible PL of the film originates from the radiative recombination of excitons trapped in the surface states. At decreasing excitation energy (E_ex), the PL peak energy was found to be redshifted, accompanied by a narrowing of the bandwidth. These results are explained by surface exciton recombination model taking into account there existing a size distribution of a-Si NPs in the silicon nitride matrix.     

Numerical analysis of Excimer laser assisted processing of multi-layers for the tailored dehydrogenation of amorphous and nano-crystalline silicon films

The application of the striking electrical and optical properties of amorphous and nano-crystalline silicon in photovoltaic, photonic and nano-electronic devices is attracting increasing attention. In particular, its use both on polymeric substrates and in Integrated Circuit technology for the development of enhanced new devices has shown that processing techniques to produce amorphous hydrogenated and nano-crystalline silicon films avoiding high substrate temperatures are of great importance. A promising strategy to achieve this purpose is the combination of Hot-Wire Chemical Vapor Deposition at 150 °C with Excimer Laser Annealing, thus maintaining the substrate at relatively low temperature during the complete process.In this work we present a numerical analysis of Excimer Laser Annealing, performed at room temperature, of a multilayer structure of thin alternating a-Si:H and nc-Si films deposited on glass and grown by Hot-Wire Chemical Vapor Deposition. A set of two different layer thicknesses a-Si:H (25 nm)/nc-Si (100 nm) and a-Si:H (30 nm)/nc-Si (60 nm) were analysed for a total structure dimension of 900 nm. The aim is to determine the probable temperature profile to achieve controlled localized in depth dehydrogenation.Temperature distribution has been calculated inside the multilayer during the irradiation by a 193 nm Excimer laser, 20 ns pulse length, with energy densities ranging from 50 to 300 mJ/cm². Calculations allowed us to estimate the dehydrogenation effect in the different layers as well as the structural modifications of the same layers as a function of the applied laser energy.The numerical results have been compared to the experimental ones obtained in similar multilayer structures that have been analysed through Raman spectroscopy and TOF-SIMS in depth profiling mode.     

Structural investigations of silicon nanostructures grown by self-organized island formation for photovoltaic applications

The self-organized growth of crystalline silicon nanodots and their structural characteristics are investigated. For the nanodot synthesis, thin amorphous silicon (a-Si) layers with different thicknesses have been deposited onto the ultrathin (2 nm) oxidized (111) surface of Si wafers by electron beam evaporation under ultrahigh vacuum conditions. The solid phase crystallization of the initial layer is induced by a subsequent in situ annealing step at 700 °C, which leads to the dewetting of the initial a-Si layer. This process results in the self-organized formation of highly crystalline Si nanodot islands. Scanning electron microscopy confirms that size, shape, and planar distribution of the nanodots depend on the thickness of the initial a-Si layer. Cross-sectional investigations reveal a single-crystalline structure of the nanodots. This characteristic is observed as long as the thickness of the initial a-Si layer remains under a certain threshold triggering coalescence. The underlying ultra-thin oxide is not structurally affected by the dewetting process. Furthermore, a method for the fabrication of close-packed stacks of nanodots is presented, in which each nanodot is covered by a 2 nm thick SiO₂ shell. The chemical composition of these ensembles exhibits an abrupt Si/SiO₂ interface with a low amount of suboxides. A minority charge carrier lifetime of 18 µs inside of the nanodots is determined.     

Structural, optical and electrical properties of amorphous silicon thin films prepared by sputtering with different targets

Amorphous silicon (a-Si) films were prepared by sputtering method with polycrystalline and monocrystalline silicon targets. Structural, optical and electrical properties of the a-Si films have been systematically studied. The deposition power is from 100 to 200 W. Compared with the a-Si films deposited by using monocrystalline silicon target, the a-Si films prepared with polycrystalline silicon target exhibit better growth property, similar optical band gap, and own the highest mobility of 1.658 cm²/Vs, which make a good match with the optimal window of optical band gap for a-Si solar cells. The results indicated that the polycrystalline silicon target is superior to the monocrystalline silicon target when used to prepare a-Si films as the intrinsic layer in a-Si solar cells.     

Modelling amorphous silicon with hydrogenated defects: GW treatment of the ST12 phase

The ST12 phase of silicon is investigated as a possible model for amorphous silicon (a-Si). The structure is studied both with and without hydrogenated hole defects to model the properties of hydrogenated amorphous silicon (a-Si:H) as well as a-Si. A density functional theory model of ST12 Si is structurally relaxed, and the radial correlation function and phonon density of states are used to compare the structural properties of the model to those of a-Si. One-shot GW self-energy corrections are used to generate the band structure, and the corrected electronic structure is found to reproduce the experimental energy gap of a-Si. Introducing hydrogenated defects to the ST12 structure leads to a slight decrease in the band gap and a shift in the density of states, as the breaking of symmetry results in band splitting. The dielectric functions are calculated for both a-Si and a-Si:H, using the GW corrected band structures, with a density functional perturbation theory approach. The model ST12 Si is found to absorb strongly at slightly lower energies than experimental a-Si, whereas the spectrum of the hydrogenated ST12 closely matches that of a-Si:H.     

后退火增强氢化非晶硅钝化效果的研究

在本征氢化非晶硅(a-Si:H(i))/晶体硅(c-Si)/a-Si:H(i)异质结构上溅射ITO时, 发现后退火可大幅增加ITO/a-Si:H(i)/c-Si/a-Si:H(i)的少子寿命(从1.7 ms到4 ms). 这一增强效应可能的三个原因是: ITO/a-Si:H(i)界面场效应作用、退火形成的表面反应层影响以及退火对a-Si:H(i)材料本身的优化, 但本文研究结果表明少子寿命增强效应与ITO和表面反应层无关; 对不同沉积温度制备的a-Si:H(i)/c-Si/a-Si:H(i)异质结后退火的研究表明: 较低的沉积温度(<175 ℃)后退火增强效应显著, 而较高的沉积温度(>200 ℃)后退火增强效应不明显, 可以确定“低温长高温后退火”是获得高质量钝化效果的一种有效方式; 采用傅里叶红外吸收谱(FTIR)研究不同沉积温度退火前后a-Si:H(i)材料本身的化学键构造, 发现退火后异质结少子寿命大幅提升是由于a-Si:H(i)材料本身的结构优化造成的, 其深层次的本质是通过材料的生长温度和退火温度的优化匹配来控制包括H含量、H键合情况以及Si原子无序性程度等微观因素主导作用的一种竞争性平衡, 对这一平衡点的最佳控制是少子寿命大幅提升的本质原因.     

Effect of surface Si-Si dimers on photoluminescence of silicon nanocrystals in the silicon dioxide matrix

The effect of surface states of silicon nanocrystals embedded in silicon dioxide on the photoluminescent properties of the nanocrystals is reported. We have investigated the time-resolved and stationary photoluminescence of silicon nanocrystals in the matrix of silicon dioxide in the visible and infrared spectral ranges at 77 and 300 K. The structures containing silicon nanocrystals were prepared by the high-temperature annealing of multilayer SiO _x /SiO₂ films. The understanding of the experimental results on photoluminescence is underlain by a model of autolocalized states arising on surface Si-Si dimers. The emission of autocatalized excitons is found for the first time, and the energy level of the autolocalized states is determined. The effect of these states on the mechanism of the excitation and the photoluminescence properties of nanocrystals is discussed for a wide range of their dimensions. It is reliably shown that the cause of the known blue boundary of photoluminescence of silicon nanocrystals in the silicon dioxide matrix is the capture of free excitons on autolocalized surface states.     

IR selective sensitivity of amorphous/crystalline silicon heterostructure

An IR detector based on a-Sin-type) /c-Si(p-type) heterostructure is reported to have selective sensitivity in the wavelength ranged for 1000–1100 nm with a maximum at 1080 nm. The energy band diagram of the heterostructure is considered taking into account the existence of the interfacial layer. The analysis of the photogeneration and carrier transport processes in the a-Si(n-type) / c-Si(p-type) heterostructure has been carried out. It is shown that the interfacial layer and defect on the a-Si/c-Si boundary control the spectral photosensitivity characteristic.     

Surface passivation in n-type silicon and its application in silicon drift detector

Based on the surface passivation of n-type silicon in a silicon drift detector(SDD), we propose a new passivation structure of SiO2/Al2O3/SiO2 passivation stacks. Since the SiO2 formed by the nitric-acid-oxidation-of-silicon(NAOS)method has good compactness and simple process, the first layer film is formed by the NAOS method. The Al2O3 film is also introduced into the passivation stacks owing to exceptional advantages such as good interface characteristic and simple process. In addition, for requirements of thickness and deposition temperature, the third layer of the SiO2 film is deposited by plasma enhanced chemical vapor deposition(PECVD). The deposition of the SiO2 film by PECVD is a low-temperature process and has a high deposition rate, which causes little damage to the device and makes the SiO2 film very suitable for serving as the third passivation layer. The passivation approach of stacks can saturate dangling bonds at the interface between stacks and the silicon substrate, and provide positive charge to optimize the field passivation of the n-type substrate.The passivation method ultimately achieves a good combination of chemical and field passivations. Experimental results show that with the passivation structure of SiO2/Al2O3/SiO2, the final minority carrier lifetime reaches 5223 μs at injection of 5×10¹⁵ cm^-3. When it is applied to the passivation of SDD, the leakage current is reduced to the order of nA.     

Nanosecond pulsed laser ablation of silicon in liquids

Laser fluence and laser shot number are important parameters for pulse laser based micromachining of silicon in liquids. This paper presents laser-induced ablation of silicon in liquids of the dimethyl sulfoxide (DMSO) and the water at different applied laser fluence levels and laser shot numbers. The experimental results are conducted using 15 ns pulsed laser irradiation at 532 nm. The silicon surface morphology of the irradiated spots has an appearance as one can see in porous formation. The surface morphology exhibits a large number of cavities which indicates as bubble nucleation sites. The observed surface morphology shows that the explosive melt expulsion could be a dominant process for the laser ablation of silicon in liquids using nanosecond pulsed laser irradiation at 532 nm. Silicon surface’s ablated diameter growth was measured at different applied laser fluences and shot numbers in both liquid interfaces. A theoretical analysis suggested investigating silicon surface etching in liquid by intense multiple nanosecond laser pulses. It has been assumed that the nanosecond pulsed laser-induced silicon surface modification is due to the process of explosive melt expulsion under the action of the confined plasma-induced pressure or shock wave trapped between the silicon target and the overlying liquid. This analysis allows us to determine the effective lateral interaction zone of ablated solid target related to nanosecond pulsed laser illumination. The theoretical analysis is found in excellent agreement with the experimental measurements of silicon ablated diameter growth in the DMSO and the water interfaces. Multiple-shot laser ablation threshold of silicon is determined. Pulsed energy accumulation model is used to obtain the single-shot ablation threshold of silicon. The smaller ablation threshold value is found in the DMSO, and the incubation effect is also found to be absent.     

载流子选择性接触:高效硅太阳电池的选择

太阳电池可看成由光子吸收层和接触层两个基本单元组成,接触层是高复合活性金属界面和光子吸收层之间的区域.为了进一步提高硅太阳电池的转换效率,关键是降低光子吸收层和接触之间的复合损失.近年来,载流子选择性接触引起了光伏界的研究兴趣,其被认为是接近硅太阳电池效率理论极限的最后的障碍之一.本文分析了三种类型的载流子选择性接触:在光子吸收层与金属界面之间引入薄的重掺杂层,即所谓的发射极或背面场;利用两种材料之间的导带或价带对齐;利用高功函数的金属氧化物与晶硅接触从而在晶硅中感应能带弯曲.基于一维太阳电池模拟软件wx AMPS,模拟了扩散同质结硅太阳电池[结构为(p~+)c-Si/(n)c-Si/(n~+)c-Si]、非晶硅薄膜硅异质结太阳电池[结构为(p~+)a-Si/(i)a-Si/(n)c-Si/(i)a-Si/(n~+)a-Si]和氧化物薄膜硅异质结太阳电池[结构为(n)MoO_x/(n)c-Si/(n)TiO_x]暗态下的能带结构和载流子浓度的空间分布,其中c-Si为晶硅;a-Si为非晶硅;(i),(n)和(p)分别表示本征、n型掺杂和p型掺杂.模拟结果表明:载流子选择性接触的核心是在接触处晶硅表面附近形成载流子浓度空间分布的不对称进而使得电导率的不对称,形成了对电子的高阻和空穴的低阻或者对空穴的高阻和电子的低阻,从而让空穴轻松通过同时阻挡电子,或者让电子轻松通过同时阻挡空穴,形成空穴选择性接触或者电子选择性接触.     

Wet-chemical passivation of atomically flat and structured silicon substrates for solar cell application

Special sequences of wet-chemical oxidation and etching steps were optimised with respect to the etching behaviour of differently oriented silicon to prepare very smooth silicon interfaces with excellent electronic properties on mono- and poly-crystalline substrates. Surface photovoltage (SPV) and photoluminescence (PL) measurements, atomic force microscopy (AFM) and scanning electron microscopy (SEM) investigations were utilised to develop wet-chemical smoothing procedures for atomically flat and structured surfaces, respectively. Hydrogen-termination as well as passivation by wet-chemical oxides were used to inhibit surface contamination and native oxidation during the technological processing. Compared to conventional pre-treatments, significantly lower micro-roughness and densities of surface states were achieved on mono-crystalline Si(100), on evenly distributed atomic steps, such as on vicinal Si(111), on silicon wafers with randomly distributed upside pyramids, and on poly-crystalline EFG (Edge-defined Film-fed-Growth) silicon substrates.The recombination loss at a-Si:H/c-Si interfaces prepared on c-Si substrates with randomly distributed upside pyramids was markedly reduced by an optimised wet-chemical smoothing procedure, as determined by PL measurements. For amorphous-crystalline hetero-junction solar cells (ZnO/a-Si:H(n)/c-Si(p)/Al) with textured c-Si substrates the smoothening procedure results in a significant increase of short circuit current I_sc, fill factor and efficiency η. The scatter in the cell parameters for measurements on different cells is much narrower, as compared to conventional pre-treatments, indicating more well-defined and reproducible surface conditions prior to a-Si:H emitter deposition and/or a higher stability of the c-Si surface against variations in the a-Si:H deposition conditions.     

Crystallization of amorphous silicon beyond the crystallized polycrystalline silicon region induced by metal nickel

Crystallization of amorphous silicon(a-Si) which starts from the middle of the a-Si region separating two adjacent metal-induced crystallization(MIC) polycrystalline silicon(poly-Si) regions is observed. The crystallization is found to be related to the distance between the neighboring nickel-introducing MIC windows. Trace nickel that diffuses from the MIC window into the a-Si matrix during the MIC heat-treatment is experimentally discovered, which is responsible for the crystallization of the a-Si beyond the MIC front. A minimum diffusion coefficient of 1.84×10~(-9)cm~2/s at 550℃ is estimated for the trace nickel diffusion in a-Si.