The effect of surface plasmon resonance (SPR) from noble metal nanostructures such as gold nanoparticles (Au NPs) has been proposed to promote the generation of energetic hot electrons as well as boosting resonant energy transfer, thereby resulting in significantly enhancing solar-light harvesting and energy conversion efficiency. Herein, Au NPs decorated zinc oxide nanorods with plasmonic metal–semiconductor heterostructures have been synthesized through UV/Ozone treatment. Absorption, light-to-plasmon conversion efficiency, plasmon-to-hot electron conversion efficiency, and quality (Q)-factor of Au@ZnO nanocomposites are further characterized in order to understand the related SPR effect from various aspects. Simultaneously, the use of machine learning (ML) as an artificial intelligence data-driven method to derive an alternative predictive model for evaluating the relationship between synthesis and properties of materials has been adopted. In this regard, we collect only a limited supply of experimental dataset as training data to establish the predictive model with an artificial neural network incorporating genetic algorithm. According to the results from experimental datasets and the proposed predictive model, our analysis has revealed that the conversion efficiency and Q-factor associated with the SPR effect from Au@ZnO nanocomposites can be efficiently evaluated through ML, which has potential application in plasmon-sensitized solar cells and plasmonic lasers in the future.

中文翻译：

---

Au纳米粒子修饰的ZnO纳米棒的表面等离子体共振特性的材料基因组进化

来自贵金属纳米结构（如金纳米粒子 (Au NPs)）的表面等离子体共振 (SPR) 的作用已被提出可促进高能热电子的产生以及促进共振能量转移，从而显着增强太阳光收集和能量转换效率。在此，通过紫外线/臭氧处理合成了具有等离子体金属-半导体异质结构的 Au NPs 修饰的氧化锌纳米棒。进一步表征了 Au@ZnO 纳米复合材料的吸收、光到等离子体转换效率、等离子体到热电子转换效率和质量 (Q) 因子，以便从各个方面了解相关的 SPR 效应。同时地，使用机器学习 (ML) 作为人工智能数据驱动的方法来推导替代预测模型，以评估材料的合成与性能之间的关系。在这方面，我们仅收集有限供应的实验数据集作为训练数据，以使用包含遗传算法的人工神经网络建立预测模型。根据实验数据集的结果和提出的预测模型，我们的分析表明，可以通过 ML 有效地评估与 Au@ZnO 纳米复合材料的 SPR 效应相关的转换效率和 Q 因子，这在等离子体敏化中具有潜在的应用价值。未来的太阳能电池和等离子体激光器。