激光玻璃光学光谱特性预测:理论与实践  

Predicting spectroscopic properties of laser glasses:Theory and practice

作　　者：董双丽 姬瑶 明红 王伟超 张勤远 DONG ShuangLi;JI Yao;MING Hong;WANG WeiChao;ZHANG QinYuan(School of Physics and Optoelectronics,South China University of Technology,Guangzhou 510640,China;School of Materials Science and Engineering,South China University of Technology,Guangzhou 510640,China;State Key Laboratory of Luminescent Materials and Devices,Guangdong Provincial Key Laboratory of Fiber Laser Materials and Applied Techniques,South China University of Technology,Guangzhou 510640,China)

机构地区：[1]华南理工大学物理与光电学院,广州510640 [2]华南理工大学材料科学与工程学院,广州510640 [3]华南理工大学发光材料与器件国家重点实验室,广东省光纤激光材料与应用技术重点实验室,广州510640

出　　处：《中国科学：技术科学》2024年第11期2179-2192,共14页Scientia Sinica(Technologica)

基　　金：国家自然科学基金重点项目(52130201);中央高校基本科研业务费专项基金项目(2023ZYGXZR099)资助。

摘　　要：锗酸盐光学玻璃具有红外透过范围宽、声子能量低、机械性能优异等特点,是当前红外窗口及近中红外激光器核心增益介质.研制具有特定物理与光学光谱特性的新型锗酸盐光学玻璃是该领域的重要挑战之一.以往的研究工作大多采用“试错法”,通过大量实验才得到一些结果,消耗的人财物力巨大.本文利用玻璃结构的相图模型扩充了掺Tm^(3+)锗酸盐激光玻璃现有文献数据,获得了约500个高质量的光谱特性数据.在此基础上,使用12个物理和化学描述符替代了20余个玻璃组成作为特征值,采用随机森林算法构建模型,准确预测了锗酸盐激光玻璃的半高宽、发射截面、辐射寿命和辐射跃迁几率等多项光学光谱性质.与已报道的模型相比,本文通过优化和减少描述符数量降低了模型的复杂度,提高了对极值区域的预测能力(相对误差<10%)和外推能力,增强了模型的可解释性和通用性,对于研制新型激光玻璃具有重要指导意义.Germanate optical glass, due to its wide infrared transmission range, low phonon energy, and excellent mechanical properties, hasimportant applications in military platforms as infrared window materials and gain media for near and mid-infrared fiber lasers.Developing germanate glass compositions with specific physical and spectroscopic properties is one of the important challenges inthis field. In this study, approximately 500 high-quality data were first obtained by a phase diagram approach. Based on this, a modelwas constructed using the random forest algorithm, replacing more than 20 glass compositions with 12 physical and chemicaldescriptors as features, accurately predicting properties such as full width at half maximum, emission cross-section, radiative lifetime,and radiative transition probability of the glass. This study optimizes and reduces the complexity of the model by reducing the numberof descriptors, and enhances the forecasting and extrapolation abilities in extreme regions. This is of great significance for improvingthe interpretability and generality of the model and discovering novel laser glasses.

关 键 词：激光玻璃 玻璃结构相图模型 机器学习 锗酸盐玻璃 光谱性质 

分 类 号：TQ171.734[化学工程—玻璃工业]