- 學位論文
二維材料後段兼容製程開發及在互連中的應用
Back-end-of-line compatible process development of two-dimensional materials and its application in interconnects
若您是本文的作者，可授權文章由華藝線上圖書館中協助推廣。
摘要
對於後段製程（BEOL）互連，銅（Cu）具有低電阻率，因此被認為是優選材料。其中在銅互連中應使用氮化鉭（TaN）作為阻障層(barrier)，是為了防止銅因其高擴散性而擴散到介電層中。然而，由於氮化鉭作為阻障層與銅的附著力差，鉭（Ta）被引入作為襯墊(liner)以增加附著力，最終形成金屬/內襯/阻障層/介電層的結構。然而，隨著銅互連尺寸的不斷縮小，隨著Ta/TaN佔用的體積增加，電阻率會急劇增加。由於傳統阻障材料（TaN/Ta）的三維特性，很難實現原子級厚度的薄膜，且在厚度低於3納米時，阻擋銅擴散的能力逐漸喪失。因此隨著尺寸縮放需求的增加，TaN/Ta在保持阻障特性和增加電阻率的同時已達到縮放極限。相比之下，二維材料具有縮放至低於2納米以下的潛力，並同時能夠保持阻障特性。因此，迫切需要找到新的互連替代材料，其中2D材料的特性是重要考慮因素之一。關於下一代互連技術，科學家們必須在電阻率和可靠性之間做出權衡，因此提出了可以替代銅的金屬材料，如鈷（Co）和釕（Ru）。鈷和釕都具有較小的平均自由徑，這減少了因平均自由徑造成散射效應的影響，而且它們同時是具有高熔點的材料，因此被視為是取代銅互連的候選材料。然而，先前的研究表明，完全替換互連材料對研究人員來說是相當具有挑戰性的，還有許多技術問題需要克服。因此本研究將問題分為兩個階段來解決，短期目標是替換現行的diffusion barrier而長期目標則是以Co來取代Cu導線。其中短期目標主要是利用二維材料的厚度優勢取代現行的阻障材料以增加Cu導線的截面積去減少尺寸微縮所造成電阻值上升的問題；而長期目標則是考慮以Co來取代Cu導線，以解決Cu導線微縮下面臨的electromigration日益嚴峻的情形，透過二維材料包覆住金屬導線的架構可以利用Co抵抗electromigration效應較好的優勢將其作為interconnect材料，同時也可以利用二維材料抑制金屬導線的surface scattering 的現象使得Co與銅相比電阻值並不會高出太多，進而實現替換互連材料之目標。
並列摘要
For back-end-of-line (BEOL) interconnects, copper (Cu) is recognized as the preferred material due to its low resistivity. Due to Cu's high diffusivity, tantalum nitride (TaN) barrier layers in Cu lines are recommended to hinder Cu diffusing into the insulate layer. However, as TaN has poor adhesion to copper when used as a barrier layer, tantalum (Ta) is selected as a liner to enhance adhesion. This has resulted in the formation of a structure consisting of a metal/liner/barrier layer/dielectric layer. However, Ta/TaN occupies a larger volume, which causes the resistivity to rise dramatically as the size of Cu interconnects decreases. Because of the 3D characteristics of conventional barrier/liner materials (TaN/Ta), it is challenging to achieve films with atomic-level thickness, and the diffusion barrier's ability to block Cu diffusion gradually diminishes at thicknesses thinner than 3 nm. Therefore, to meet the requirement for the size scaling increases, TaN/Ta has increasing resistivity to reach its scaling limit while maintaining barrier properties. On the contrary, 2D materials have the great potential to scale below than 2 nm while maintaining barrier properties. Consequently, it still needs to identify new interconnect materials, with the properties of 2D materials being an important research topic. Regarding next-generation interconnect technology, semiconductor research must consider the importance between resistivity and reliability, leading to the proposal of metal materials such as cobalt (Co) and ruthenium (Ru) as potential replacements for copper. The mean free paths of Co and Ru are lower than Cu, which reduces the impact of surface and boundary scattering effects, and they are higher melting-point materials, making them candidates for replacing Cu interconnects. However, previous scientists indicate that completely replacing line materials poses significant challenges, with many technical problems yet to be resolved. Therefore, this study addresses the issue in two stages: the short-term goal is to replace the current diffusion barrier, and the long-term goal is to replace Cu interconnects with Co. The short-term goal involves using the thickness advantage of 2D materials to replace the current diffusion barrier materials, thereby increasing the cross-sectional area of Cu interconnects and reducing the resistivity increase caused by size scaling. The long-term goal is to consider replacing Cu wires with Co to address the increasing severity of electromigration issues faced by scaling Cu interconnects. By encapsulating the Co interconnect with graphene, the better electromigration resistance of Co can be utilized, making it a viable interconnect material. Additionally, the use of 2D materials can suppress surface scattering in metal wires, ensuring that the resistivity of Co does not significantly exceed that of Cu, thereby achieving the goal of replacing interconnect materials.
參考文獻
[1] Y. Liu, X. Duan, H.-J. Shin, S. Park, Y. Huang, and X. Duan, "Promises and prospects of two-dimensional transistors," Nature, vol. 591, no. 7848, pp. 43-53, 2021.
[2] W. Steinhögl, G. Schindler, G. Steinlesberger, and M. Engelhardt, "Size-dependent resistivity of metallic wires in the mesoscopic range," Physical Review B, vol. 66, no. 7, p. 075414, 2002.
[3] K. S. Novoselov, A. Mishchenko, A. Carvalho, and A. Castro Neto, "2D materials and van der Waals heterostructures," Science, vol. 353, no. 6298, p. aac9439, 2016.
[4] Y. Shi, H. Li, and L.-J. Li, "Recent advances in controlled synthesis of two-dimensional transition metal dichalcogenides via vapour deposition techniques," Chemical Society Reviews, vol. 44, no. 9, pp. 2744-2756, 2015.
[5] Y. Venkata Subbaiah, K. Saji, and A. Tiwari, "Atomically thin MoS2: a versatile nongraphene 2D material," Advanced Functional Materials, vol. 26, no. 13, pp. 2046-2069, 2016.