The step-by-step strain evolution in the channel during the SiGe nanosheet (NS) integration process flow for pFETs is demonstrated using finite element analysis (FEA). The effect of device dimensions and defective source/drain (S/D) is studied. After fin formation, the 0.77% compressive biaxial strain resulting from the lattice mismatch between Si0.8Ge0.2 and Si substrate is observed. However, the strain $_{\textit {xx}}$ is gradually relaxed during the S/D recess and the inner spacer cavity formation. A large strain $_{\textit {xx}}$ is obtained on the channel along the current direction after Si0.6Ge0.4 S/D regrowth, increasing from 0.21% to 1.50% for defect-free S/D epitaxy. The compressive strain along the channel remains similar for different shapes of the S/D regrowth, with small variations between every channel. Decreasing the NS width only leads to an insignificant increase in channel strain $_{\textit {xx}}$ until the nanowire structure is formed. Nevertheless, the scaled body thickness can enhance the channel strainxx substantially with the 21.2% compressive strain $_{\textit {xx}}$ increase from ${t}_{\text {body}}$ = 5 nm to ${t}_{\text {body}}$ = 1 nm. The defective S/D is also simulated with air gaps between the merged epitaxes, where compressive strain in the channel is totally relaxed and further turns into tensile-strained. The hole mobility is expected to have a $3.6\times $ enhancement with 1.5% compressive strain $_{\textit {xx}}$ .

中文翻译：

---

SiGe 纳米片晶体管工艺流程中的应变演化

使用有限元分析 (FEA) 演示了 pFET 的 SiGe 纳米片 (NS) 集成工艺流程中通道中的逐步应变演变。研究了器件尺寸和缺陷源极/漏极 (S/D) 的影响。鳍片形成后，观察到由 Si0.8Ge0.2 和 Si 衬底之间的晶格失配引起的 0.77% 压缩双轴应变。然而，应变 $_{\textit {xx}}$在S/D凹口和内部间隔腔形成期间逐渐松弛。大应变 $_{\textit {xx}}$Si0.6Ge0.4 S/D 再生长后，沿电流方向在沟道上获得了电流方向的增益，对于无缺陷 S/D 外延，增益从 0.21% 增加到 1.50%。对于不同形状的 S/D 再生长，沿通道的压缩应变保持相似，每个通道之间的变化很小。减小 NS 宽度只会导致沟道应变的微不足道的增加 $_{\textit {xx}}$直至形成纳米线结构。尽管如此，缩放体厚度可以显着增强通道应变xx，压缩应变为21.2% $_{\textit {xx}}$增加自 ${t}_{\text {正文}}$= 5 纳米至 ${t}_{\text {正文}}$= 1 纳米。有缺陷的 S/D 还通过合并外延之间的气隙进行模拟，其中通道中的压缩应变完全松弛并进一步转变为拉伸应变。预计空穴迁移率 $3.6\次$1.5% 压缩应变增强 $_{\textit {xx}}$ 。