Keywords: 
• 三角形石墨烯纳米片 /
• 一维量子点阵列 /
• 磁电子学性质 /
• 磁输运性质

Abstract: Graphene (GN), a monolayer two-dimensional (2D) system closely arranged into a benzene ring structure by C atoms, has so far aroused considerable research interest due to its novel electronic, magnetic, mechanical and thermal properties. But 2D GN is a semimetal with zero band gap, and the lowest conduction band touches the highest valence band at Fermi level, leading to the inability to achieve the "off" effect in the electronic device. Therefore, many researchers are searching the solutions. A simple and feasible method is to convert 2D GN into quasi-one-dimensional (1D) graphene nanoribbons, quantum-dot arrays (QDAs) and zero-dimensional (0D) quantum-dot by tailoring it along a specific single crystallographic direction. The QDAs, due to their structural diversity, have great potential applications in future nano-integrated circuit. In this work, first-principles method based on density functional theory is used to study the magneto-electronic and magnetic transport properties of four 1D quantum-dot arrays (1D QDAs) consisting of triangular graphene nanoflakes with different linking modes. The calculated binding energy suggests that these structures are very stable, and the arrays that are linked by the "bottom-side" are more stable than that only by the "vertex". In particular, it is found that the electronic and magnetic features are not only related to the different magnetic states, but also depend on linking modes. For example, in the non-magnetism state, different QDAs can be a metal or a narrowed band-gap semiconductor. In the ferromagnetic state, different QDAs can be half-metal materials or bipolar magnetic semiconductors with different gaps, and have greatly different magnetic moments from 1.985 to 7.994μB/unit cell, reaching a difference almost as large as four times. While in the antiferromagnetic state, all QDAs are semiconductors but with different gaps. These results imply that the linking modes play a crucial role in effectively tuning the electronic and magnetic features for nanostructures. The calculated atom-projected density of states indicates that the highest valence band and the lowest conduction band are determined by the edge C atoms. The half-metallic and bipolar magnetic semiconducting behaviors presented by 1D QDA are extremely important for developing magnetic devices, which is not found in the intrinsic graphene nanoribbons. And, we also investigate the magnetic device properties based on one kind of QDA, and the single or dual spin-filtering effect with the perfect (100%) spin polarization and a rectification ratio of about 104 can be predicted. Particularly, a giant magnetoresistance over 109% is found unambiguously, which is two orders of magnitude higher than the value predicted based on the zigzag graphene nanoribbons and five orders of magnitude higher than previously reported experimental values for the MgO tunnel junction. Our results thus provide strong evidence for the effectiveness of QDAs on the magneto-electronic properties.