液态Li(或Li／Pb)作为主要的氚增值剂，应用于聚变反应堆包层结构中。但在高温下，液态锂对于金属结构材料具有较强的腐蚀作用，会严重影响材料的使用寿命。了解材料在液态锂中的腐蚀行为，探讨如何提高材料抗腐蚀能力，在聚变堆材料研究方面很有价值。 本文中采用磁控溅射法在316奥氏体不锈钢基底制各Er₂O₃防腐蚀涂层，研究激光诱导击穿光谱(Laser-Induced Breakdown Spectroscopy，LIBS)技术用于经液态锂腐蚀处理后Er₂O₃涂层和316奥氏体不锈钢基底表层元素的纵向分布分析。在获取腐蚀介质锂的特征谱线强度随烧蚀脉冲作用于样品次数变化趋势的基础上，评估了锂在试样表层的渗透深度，探索了涂层和基底的锂腐蚀行为与机理。由于锂的渗透深度与失重法获得的试样腐蚀层厚度非常接近，可以将其作为判断材料抗腐蚀能力强弱的一个重要依据。而对于腐蚀层中的过渡元素(Fe、Cr、Ni和Mn)，受腐蚀层表面形貌特征、微观结构和成分变化以及元素激发特性的可能影响，腐蚀层表层中这些元素的LIBS特征谱线峰强度大幅增强，且随纵向深度的增加谱线强度剧烈波动，元素分布信息被信号波动淹没。虽然实现了对腐蚀层成分的定性分析，但无法通过强度变化趋势反映这些元素含量的实际变化和分布。 在LIBS测试条件优化获取元素半定量分布特征的基础上，开展锂腐蚀机理的研究和分析，主要包括在(1)选择更合适的光谱采集实验参数和(2)对激光诱导等离子体进行局部热力学平衡(Local temperature equilibtium，LTE)分析的基础上，进行基质效应评估。在LTE条件基本满足的条件下，LIBS纵向深度分析的基质效应影响不能完全忽略。进一步采用内标法处理LIBS光谱数据，可有效降低光谱信号波动的影响，减小了基质效应，获得了腐蚀层中主要元素的分布。分析引起成分含量变化可能原因，认为在本文的实验条件下，溶解和晶间渗透是主要的腐蚀机制。 以EDX测试结果为参考，证实LIBS技术用于腐蚀层元素纵向深度分析的有效性。继续(1)对不同腐蚀时间试样的表层成分分布进行分析，发现元素分布存在明显不同； (2)借助Ar气氛对LIBS光谱信号的增强作用，获取了对试样中低含量和难激发元素(Si、 Mo、C)的分布表征；(3)以锂的渗透深度为评判标准，比较了不锈钢焊接件中焊缝和母材的抗液态锂腐蚀能力。 基于腐蚀后不锈钢基底LIBS分析的有效性，开展Er₂O₃涂层试样制备与抗腐蚀性能研究，包括涂层沉积、试样腐蚀和LIBS深度分析三部分。以直流磁控溅射作为沉积Er₂O₃涂层的主要方式，采用两种制备流程：(1)预溅射Er金属层后氧化处理；(2)反应溅射。通过改变沉积条件参数，获得不同的Er₂O₃涂层，并对不同条件下沉积速率，涂层的形貌结构和晶粒尺寸进行探讨。对不同的涂层试样进行液态锂腐蚀处理，并对腐蚀前后的涂层做形貌结构比较，并对涂层表面的腐蚀产物进行表征，分析其生成过程。 在Ar气氛下用LIBS深度分析，表征锂腐蚀后Er₂O₃涂层试样由涂层表面到涂层与基底间界面的元素分布情况。由于LIBS对于被测试样的材质非常敏感，而待测的涂层试样存在沿深度方向的成分渐变(涂层中包含Er₂O₃和未完全氧化的Er，两者在涂层中呈不均匀分布)，且涉及涂层与基底间的界面，这使基于LIBS的涂层试样成分表征，相比于被锂腐蚀的316SS基底试样更加复杂。因此，我们以纯Er、纯Er₂O₃和316SS分别作为标准试样，对LIBS测试条件进行优化，以便能同时兼顾涂层和基底。由于涂层与基底材质差异较大，基质效应无法完全忽略，因此同样采用内标法处理测试数据。在数据处理过程中比较了不同内标的处理结果，发现在区分涂层界面方面，Ar谱线作为内标时表现较好。由数据处理后获得的成分分布曲线中发现，在锂腐蚀过程中，不锈钢中的Ni存在向涂层扩散的趋势，这使得Ni的分布不同于Fe和Cr。 在涂层腐蚀实验中观察发现，涂层失效主要表现为涂层由基底剥离。而通过形貌观察发现，裸露出的基底表面相对平整，没有明显的腐蚀迹象，而残余的Er₂O₃呈片状或花状结构在基底表面不均匀分布。由此可见涂层在一定程度上起到了隔绝锂的作用，但因为与基底结合较差而剥落。因此改进氧化物涂层与不锈钢基的结合性能，是需要优先解决的问题。通过增加Er过渡层，交替沉积Er金属层和Er₂O₃层的方式，获得与基底结合较好，且具有一定厚度的Er／Er₂O₃复合涂层。 关键词：液态锂腐蚀，Er₂O₃涂层，被腐蚀层，腐蚀产物，金属过渡层，元素成分分布，LIBS

 As a main tritium breeder, liquid lithium (or Li/Pb alloy) was applicated in blanket structure of fusion device. However, the highly corroison of liquid lithium under the high temperature, liquid lithium would short the life time of metal material in blanket structure. Hence, it is useful for researching the corrosion behavior of material and enhancement corrosion resistance in the field of fusion engineer. In our work, Laser-Induced Breakdown Spectroscopy (LIBS) was applied to characterize elemental concentration distribution in the surface layer of Er₂O₃ coating and austenite stainless steel (type 316) substrate, which were corroded by liquid lithium. Base on the result, the corrosion behavior and mechanism of liquid lithium were deduced. The relationship of intensity of emission lines of corrosion medium lithium as a function of the number of ablation of laser pulse was obtained by LIBS in-depth profile analysis, and the penetration depth of Li was estimated in the surface layer of corroded sample. Owing to the depth of Li permeation is consistent with the thickness of corrosion layer which was calculated by weight loss, it could be employed as the criterion of liquid lithium corrosion resistance of material. The trends of characterization spectral line intensity of transition elements (such as Fe, Cr, Ni and Mn) versus the pulse number were also collected, yet that could not reflect the real variation and distribution of the element content directly. It was attributed the interference factors which induced intensity fluctuation of LIBS spectral signal corresponding to the surface layer of corroded specimen and concealed the information of real elemental distributions. As the inducements of LIBS signal fluctuation, the effect of surface roughness, microstructure and compositional variation of the corroded layer were analysis respectively. To further research on corrosion mechanism, it is necessary to carry out semi-quantitatively analysis for elemental distribution. The conditions of LIBS measurement were optimized, such as (1) selecting reasonable parameter settings; (2) implementation local thermodynamic equalization (LTE) analysis for laser-induced plasma and evaluation of matrix effect during the LIBS testing. The LTE conditions were satisfied during the testing, however, the influence of matrix effect could not be completely neglected. The internal standard method was applied to data treatment of LIBS spectra for reducing the effect of signal fluctuation and gaining component distribution in corroded layer. When it was confirmed effective of LIBS depth profile analysis on content characterization of corroded specimens by EDX testing results, the follow-up work of corrosion study were carried out by LIBS, such as:(l) The specimens of 316 SS with different corrosion times were characterized through depth profile analysis, and found out the elemental distributions have varied follow by the time. (2) Argon atmosphere was used for spectral signal enhancement, that would be realized the compositional characterization of low content or hard to be excited elements (Si, Mo, C) via LIBS measurement. (3) The corrosion resistance of welding joint and base metal were compared by the penetration depths of lithium which as the main criteria of corrosion degree. As a pre-work, the LIBS characterization of corrosion substrate laid the foundation for the study of resistance corrosion of Er₂O₃ coatings, while the major research works for coating included three parts, such as fabrication of coating samples, liquid lithium corrosion and components analysis for corroded coating by LIBS. Two processes of direct current (DC) magnetron sputtering were respectively carried out for deposition of Er₂O₃ coating: (1) pre-sputtering a metallic Erbium layer on the substrate, and then oxidization it by heating in the air, (2) reactive sputtering. The Er₂O₃ coatings with different topographic, structure and grain size were fabricated though variation the conditions of deposition, and the relationships between the conditions and the state of coating were discussed. The coating specimens were treated by liquid lithium corrosion, and then compared the surface morphology, crystal structure before and after corrosion, characterized the corrosion produces and deduced the process of the produces creating. LIBS depth profile analysis under argon atmosphere was utilized to characterize the elemental distributions and concentration variation trends from the boundary of Er₂O₃ coating surface to the interface between the coating and the substrate. It is very sensitive of LIBS testing for variation of material on composition, structure and state of testing zone, while there is concentration gradient in the coating samples, therefore, it will make the LIBS component distribution characterization for the coating sample more complex than the corroded 316SS substrate specimens. As the standard specimens, pure Erbium, pure Er₂O₃ and 316 SS were employed for optimization conditions of LIBS measurement. The material of oxide ceramic coating and metal substrate are different, hence the influence of matrix effect could not be neglected, and the internal standard method was still applied on spectral data treatment. The different internal standards were investigated, and found Ar emission lines as internal standard performed well on distinguishing the interface between the coating and the substrate. The elemental concentration distributions which were acquired by data treatment shown the diffusion trend of Ni from the substrate to the coating, and that made the distribution of Ni was different from the other transition elements of the substrate, such as Fe and Cr. From the results observation, the major mechanism of the coating failure during the liquid lithium corrosion is the peel-off of the Er₂O₃ layer from the substrate. Hence, enhancing the bond between the coating and the substrate became the priority for all involved problems for coating modifying. The experiment results confirmed that adding transition layer of Erbium, alternately deposition Erbium layer and Erbium oxide layer are effectively means to obtain thicker coating with better binding ability. Key words: liquid lithium corrosion, Er₂O₃ coating, corroded layer, corrosion product, metal transition layer, elemental distribution, LIBS