• Clean Technology
• /
• v.20 no.3
• /
• pp.306-313
• /
• 2014

Nanoporous $TiO_2$ films are commonly used as working electrodes in dye-sensitized solar cells (DSSCs). So far, there have been attempts to synthesize films with various $TiO_2$ nanostructures to increase the power-conversion efficiency. In this work, vertically aligned rutile $TiO_2$ nanorods were grown on fluorinated tin oxide (FTO) glass by hydrothermal synthesis, followed by deposition of an anatase $TiO_2$ film. This new method of anatase $TiO_2$ growth avoided the use of a seed layer that is usually required in hydrothermal synthesis of $TiO_2$ electrodes. The dense anatase $TiO_2$ layer was designed to behave as the electron-generating layer, while the less dense rutile nanorods acted as electron-transfer pathwaysto the FTO glass. In order to facilitate the electron transfer, the rutile phase nanorods were treated with a $TiCl_4$ solution so that the nanorods were coated with the anatase $TiO_2$ film after heat treatment. Compared to the electrode consisting of only rutile $TiO_2$, the power-conversion efficiency of the rutile-anatase hybrid $TiO_2$ electrode was found to be much higher. The total thickness of the rutile-anatase hybrid $TiO_2$ structures were around $4.5-5.0{\mu}m$, and the highest power efficiency of the cell assembled with the structured $TiO_2$ electrode was around 3.94%.

• Journal of Intelligence and Information Systems
• /
• v.28 no.2
• /
• pp.127-146
• /
• 2022

Recently, as word embedding has shown excellent performance in various tasks of deep learning-based natural language processing, researches on the advancement and application of word, sentence, and document embedding are being actively conducted. Among them, cross-language transfer, which enables semantic exchange between different languages, is growing simultaneously with the development of embedding models. Academia's interests in vector alignment are growing with the expectation that it can be applied to various embedding-based analysis. In particular, vector alignment is expected to be applied to mapping between specialized domains and generalized domains. In other words, it is expected that it will be possible to map the vocabulary of specialized fields such as R&D, medicine, and law into the space of the pre-trained language model learned with huge volume of general-purpose documents, or provide a clue for mapping vocabulary between mutually different specialized fields. However, since linear-based vector alignment which has been mainly studied in academia basically assumes statistical linearity, it tends to simplify the vector space. This essentially assumes that different types of vector spaces are geometrically similar, which yields a limitation that it causes inevitable distortion in the alignment process. To overcome this limitation, we propose a deep learning-based vector alignment methodology that effectively learns the nonlinearity of data. The proposed methodology consists of sequential learning of a skip-connected autoencoder and a regression model to align the specialized word embedding expressed in each space to the general embedding space. Finally, through the inference of the two trained models, the specialized vocabulary can be aligned in the general space. To verify the performance of the proposed methodology, an experiment was performed on a total of 77,578 documents in the field of 'health care' among national R&D tasks performed from 2011 to 2020. As a result, it was confirmed that the proposed methodology showed superior performance in terms of cosine similarity compared to the existing linear vector alignment.