Hu Ruijin, Liang Yifei, Qian Wentao, Gan Xin, Liang Lei, Wang Junzhuan, Liu Zongguang, Shi Yi, Xu Jun, Chen Kunji, Yu Linwei
School of Electronics Science and Engineering/National Laboratory of Solid-State Microstructures, Nanjing University, Nanjing, 210093, P. R. China.
Small. 2022 Oct;18(42):e2204390. doi: 10.1002/smll.202204390. Epub 2022 Sep 9.
Fabricating ultrathin silicon (Si) channels down to critical dimension (CD) <10 nm, a key capability to implementing cutting-edge microelectronics and quantum charge-qubits, has never been accomplished via an extremely low-cost catalytic growth. In this work, 3D stacked ultrathin Si nanowires (SiNWs) are demonstrated, with width and height of W = 9.9 ± 1.2 nm (down to 8 nm) and H = 18.8 ± 1.8 nm, that can be reliably grown into the ultrafine sidewall grooves, approaching to the CD of 10 nm technology node, thanks to a new self-delimited droplet control strategy. Interestingly, the cross-sections of the as-grown SiNW channels can also be easily tailored from fin-like to sheet-like geometries by tuning the groove profile, while a sharply folding guided growth indicates a unique capability to produce closely-packed multiple rows of stacked SiNWs, out of a single run growth, with the minimal use of catalyst metal. Prototype field effect transistors are also successfully fabricated, achieving I ratio and sub-threshold swing of >10 and 125 mV dec , respectively. These results highlight the unexplored potential of versatile catalytic growth to compete with, or complement, the advanced top-down etching technology in the exploitation of monolithic 3D integration of logic-in-memory, neuromorphic and charge-qubit applications.
制造尺寸低至关键尺寸(CD)<10纳米的超薄硅(Si)通道,这是实现前沿微电子和量子电荷量子比特的一项关键能力,此前从未通过极低的成本催化生长来实现。在这项工作中,展示了3D堆叠超薄硅纳米线(SiNWs),其宽度W = 9.9 ± 1.2纳米(低至8纳米),高度H = 18.8 ± 1.8纳米,由于一种新的自限性液滴控制策略,它们能够可靠地生长到超细的侧壁凹槽中,接近10纳米技术节点的关键尺寸。有趣的是,通过调整凹槽轮廓,生长后的SiNW通道的横截面也可以很容易地从鳍状调整为片状几何形状,而急剧折叠引导生长表明具有独特的能力,能够在单次生长中以最少的催化剂金属用量生产紧密排列的多排堆叠SiNWs。还成功制造了原型场效应晶体管,分别实现了>10的I比和125 mV dec的亚阈值摆幅。这些结果突出了通用催化生长在利用逻辑-内存、神经形态和电荷量子比特应用的单片3D集成方面与先进的自上而下蚀刻技术竞争或互补的未被探索的潜力。