The physicochemical similarities of hydrogen isotopes have made their separation a challenging task. Conventional methods such as cryogenic distillation, Girdler sulfide process, chromatography, and thermal cycling absorption have low separation factors and are energy-intensive. To overcome these limitations, research has focused on kinetic quantum sieving (KQS) and chemical affinity quantum sieving (CAQS) effects for selective separation of hydrogen isotopes. Porous materials such as metal-organic frameworks (MOF), covalent organic frameworks (COF), zeolites, carbon, and organic cages have been studied for hydrogen separation. In this study, we focus the enhancement for CAQS to provide the cations due to the chemical affinity between hydrogen isotope and unsaturated sites by cations in zeolite beads. Cation exchanged zeolite beads was synthesized with cobalt, copper, nickel, iron and silver in zeolite 4A beads. Synthesized cation exchanged zeolite was analyzed for the surface area and pore size in N2 and adsorption behaviors of hydrogen isotopes (D2/H2) for various cation exchanged zeolite beads using BET at 77 K. The study predicts the D2/H2 adsorption selectivity based on the results obtained with BET. These hydrogen isotope adsorption results will provide a foundation for future processes for tritium separation.

Using recombinant DNA technology, the vector system containing minimal fragment of a bacterial hemoglobin gene (vgb) was constructed. When this vector was inserted into Escherichia coli, the growth of the host was significantly improved in both viable cell counts and absorbance measurement, compared to that of the wild type strain. In addition, by minimizing the size of bacterial hemoglobin in the vector, the ability of vgb in growth improvement was augmented, due to the reduction of metabolic burden from the maintenance and replication of the plasmid. By using this system, it is expected that the growth of microorganisms can be improved even in the hypoxic condition.

플레이어 모델링은 플레이어의 행동에 대한 선호도에 대한 특성들을 반영하는 프로파일을 만들고, 플레이어가 게임과 상호작용하는 동안 계속 프로파일의 특성을 학습하는 것이다. 학습된 선호도는 플레이어 행동들의 발생 빈도에 대한 통계적 수치로 기록되며, 이를 바탕으로 플레이어에 최적화된 유저 인터페이스를 제공할 수 있다. 이를 위해서 인터페이스에 대한 플레이어의 선호도를 효과적으로 학습할 수 있는 알고리즘이 필요하다. 베이지안 네트워크를 이용한 플레이어 모델링 기법을 사용하면 플레이어의 선호도에 따라 모델은 게임 디자인 단계에서 모델의 선호도에 대한 초기값을 쉽게 넣을 수 있기 때문에 초기 학습 시간을 줄여줄 수 있다. 본 논문은 플레이어의 선호도에 따라 적합한 인터페이스를 제공하기 위해 플레이어 모델링의 선호도 학습 알고리즘을 제안하였다.