Fabrication of tungsten oxide photoanode by doctor blade technique and investigation on its photocatalytic operation mechanism

WO₃ is a potential material candidate for construction of photoanode for solar driven water splitting. In this work, μm-thick porous WO₃ photoanode is prepared by depositing a stable ink made of WO₃ nanoparticles and Aristoflex velvet polymer in water using the doctor blade technique, followed by a sintering in air. The nature of WO₃ nanoparticles, its loading mass on F-doped tin oxide electrode as well as sintering temperature are examined in order to optimize the photocatalytic activity of the resultant WO₃ photoanode. The operation of WO₃ photoanode is investigated by varying the light illumination direction and light incident intensity as well as changing the nature of the electrolyte. Dissolved tungsten in electrolyte is quantified by ICP-MS providing insights into the influences of electrolyte nature and operating conditions to the corrosion of WO₃. It is proposed that the H₂O₂ and OH^. radical generated as by-products of the photo-driven water oxidation on the photoanode surface are harmful species that accelerate the dissolution of WO₃.     

Distribution-free uncertainty quantification for kernel methods by gradient perturbations

Machine Learning - We propose a data-driven approach to quantify the uncertainty of models constructed by kernel methods. Our approach minimizes the needed distributional assumptions, hence,...     

基于桥连接的词典学习方法的语义解析

现阶段的语义解析方法大部分都基于组合语义,这类方法的核心就是词典。词典是词汇的集合,词汇定义了自然语言句子中词语到知识库本体中谓词的映射。语义解析一直面临着词典中词汇覆盖度不够的问题。针对此问题,该文在现有工作的基础上,提出了基于桥连接的词典学习方法,该方法能够在训练中自动引入新的词汇并加以学习,为了进一步提高新学习到的词汇的准确度,该文设计了新的词语—二元谓词的特征模板,并使用基于投票机制的核心词典获取方法。该文在两个公开数据集(WebQuestions和Free917)上进行了对比实验,实验结果表明,该文方法能够学习到新的词汇,提高词汇的覆盖度,进而提升语义解析系统的性能,特别是召回率。