采用自主搭建的多光谱成像系统、 便携式X射线荧光光谱仪（X-ray fluorescence spectrometer, XRF）、 激光拉曼光谱仪（laser Raman spectrometer, LRS）对5件战国时期多色硅酸盐制品的表面物质的光谱特性、 化学成分和物相结构进行了无损分析。 根据多光谱图像中的光谱响应情况, 对样品表面物质进行了区域划分, 并发现样品表面的多数蓝、 绿、 紫色系区域出现荧光现象。 XRF检测结果表明不同区域的化学成分存在明显差异, 但多数区域均以SiO2, PbO, BaO为主要化学成分, 5件样品基本属于铅钡硅酸盐体系, 而且出现荧光区域皆是铜离子致色。 LRS精确检测出各个区域中的玻璃相、 中国蓝、 中国紫、 石英、 赤铁矿、 铅白、 无定形碳等物相, 并发现产生可见发光二极管致近红外荧光的物质为中国蓝和中国紫。 利用X射线衍射技术(X-ray diffraction, XRD)验证了LRS检测结果的正确性。 本研究将多用于书画分析的多光谱成像面检测技术与常用于硅酸盐文物成分分析的XRF和LRS技术进行有效结合, 提出了一套更为高效的无损的分析古代多色硅酸盐制品的研究方法。 光谱图像与X射线荧光光谱、 激光拉曼光谱数据的结合, 将样品的光谱特性与化学构成联系起来, 有助于提高硅酸盐文物的分析效率, 加强对硅酸盐文物的整体认识, 和降低硅酸盐文物的受损风险。

The spectral properties, chemical compositions and phases of materials constituting the surface of 5 ancient polychromatic silicate artifacts have been analyzed non-invasively with self-built multispectral imaging system, X-ray fluorescence spectrometer (XRF) and laser Raman spectrometer (LRS). Based on spectral response in multispectral images, materials constituting the surface of 5 samples can be divided into different areas, and most of blue, green, purple areas with fluorescence behavior are also mapped. The results of XRF indicate that the chemical compositions of areas are different, but the major compositions of them are SiO2, PbO, BaO. 5 samples mainly belong to PbO-BaO silicates. The coloring agents of all areas with fluorescence behavior are Cu ions. A variety of mineral phases including vitreous phase, Chinese blue, Chinese purple, quartz, hematite, lead carbonate, amorphous carbon and so on, are also identified by LRS. Chinese blue and Chinese purple can emit infrared radiation when excited by visible LED. The result of LRS is verified by X-ray diffraction (XRD). Combining the multispectral imaging area-measurement technique used to research paintings, and XRF, LRS which are usually used to analyze chemical composition of silicate artifacts, the present research proposes a more efficient and non-invasive research method to analyze ancient polychromatic silicate artifacts. Spectral characteristic and chemical composition of the sample are connected when spectral images, X-ray fluorescence spectra and Laser Raman spectra are combined. It has great significance for increasing efficiency of analysis, enhancing the overall understanding of silicate artifacts and decreasing risk of damage.