鍺通道環繞式閘極電晶體與薄膜體聲波共振器

Study of Germanium Gate-All-Around nFETs and Film Bulk Acoustic Wave Resonator

沈育誠(Yu-Cheng Shen)

指導教授 ： 劉致為

國立臺灣大學/電機資訊學院/電子工程學研究所/碩士(2017年)

https://doi.org/10.6342/NTU201702132

摘要

在過去四十年，半導體產業隨著摩爾定律的腳步，將元件尺寸不斷微縮，以製作更高效能的電晶體，造就了智慧型手機與電腦的蓬勃發展與延伸應用，使人類的生活更加的便利。在22nm 結點以前的許多研究都注重在傳統平面式的金氧半場效電晶體，隨著業界三維電晶體的量產，必須將新的模組技術應用在此架構上，我們利用磊晶技術將鍺薄膜成長在絕緣層上矽晶圓上，在使用反應式離子蝕刻技術，可以將鍺與矽之間的缺陷與錯位去除，並形成只有鍺的懸空奈米線，加上使用雙束聚焦離子束系統，在鍺通道奈米線沉積碳的保護層以阻擋通道受到離子佈植，之後使用原子層磊晶技術，成長環繞式的氧化層與閘極堆疊，並鍍上閘極電極，最後再使用選擇性雷射熱退火技術，將源極汲極的參雜載子濃度與活化率提升。使鍺通道奈米線形成環繞式閘極電晶體。
我們分析了電晶體的電性量測結果與在低溫下的特性，並外加拉伸應變給此電晶體且分析其電性，我們製作出通道長度為一百奈米之鍺通道環繞閘極式N 型場效電晶體，有著次臨限擺幅129mV/dec 和導通電流等於14 微安培，在外加拉伸應變下，導通電流可以進一步提升百分之六。
隨著摩爾定律的腳步，我們有許多高效能的手機可以使用，預期在五年內，手機與物聯網的成長仍是蓬勃發展的，這些產品背後的推手，除了高效能的數位晶片外，更重要的是有無線通訊技術，傳遞這些巨量的資料。
在手機載體中，必須有類比電路組成的雙工器或是多工器，處理高頻的類比訊號，在2G 與3G 時代，這些雙工器多是以表面聲波共振器組成，隨著無線通訊技術進入4G 時代，因為表面聲波共振器開始無法有效處理這些高頻訊號，他逐漸被薄膜體聲波共振器取代，薄膜體聲波共振器在處理高頻訊號上有著低插入損耗與高品質因子的優勢，是做成未來手機濾波器的主力技術，隨著5G 世代的到來與物聯網的崛起，他的成長仍是不斷被看好的。
在此研究中，我們探討了薄膜體聲波共振器的基本物理性質，利用MATLAB 科學計算軟體實作Mason 模型，並使用多層理論模型以方便快速設計共振器，再搭配有限元素模擬軟體驗證理論模型，最後，分別從材料選擇上與厚度的調整，我們得出氮化鋁作為共振器的壓電材料是最佳選擇，在電極的選擇上鉬元素會優於鋁，在厚度方面，越薄的電極厚度與越厚的壓電層可以做出高頻寬的共振器。

關鍵字

鍺 ； 金氧半場效電晶體 ； 閘極環繞式電晶體 ； 壓電材料 ； 薄膜體聲波共振器 ； Mason模型

並列摘要

In the past 40 years, the semiconductor industry follows the path of Moore’s law to continuously scale the physics dimension of the devices. As a result, high performance transistors enable the smartphone and computer’s application and growing. Human beings are benefit from the technology products and have a convenient life. Before 22nm technology node, most research focus on the traditional planar MOSFETs. With the mass production of 3-D transistor in industry, we need to incorporate new module technology into this architecture. High quality Ge is deposited on SOI wafer using epitaxy. Combining reactive ion etching technology, the misfit dislocations and defects can be easily removed, resulting in the floating germanium nanowire. The carbon mask gate pattering is performed by focus ion beam system before ion implantation. The conformal gate stack was deposited by ALD system. At the end of the fabrication, selective laser annealing is used to activate the dopant in S/D region.
The electrical properties and strain response of the device is measured and analyzed. The gate length of 100nm device is achieved with drive current = 14 A at Vov/VDS=1V/1V. The drive current can be enhanced for 6% under external uniaxial strain.
Following the path of the Moore’s law, we have many high performance smartphones in our daily life. The cellphone market is estimated to continuously grow for the next five years and the IOT generation is coming behind. The key technology pushes those product is the wireless communication system to transfer the huge amount of the data.
In cellphone, the multiplexers and duplexers are used to compute the high frequency analog signal. In 2G and 3G generation, most of those devices are composed of surface acoustic wave resonator. As the approach of 4G generation, the surface acoustic wave resonator cannot compute the high radio frequency signal efficiently. Thus, it is gradually replaced by film bulk acoustic wave resonator. Film bulk acoustic wave resonator has the advantages of low insertion loss and high quality factor, which is the mainstream technology for future filter.
In this research, we study the fundamental physics of the resonator form basic piezoelectric material properties and acoustic wave equation to frequency response. Then we extend the model to accurate multi-layer model to design the resonator efficiently using the MATLAB. The model is verified by the commercial finite element modeling software. In summary, aluminum nitride is the suitable piezoelectric material for FBAR. For piezoelectric layer, the thicker is the better while the electrode layer is opposite. Based on the above, high performance FBAR can be constructed.

並列關鍵字

germanium ； metal-oxide-semiconductor field effect transistor ； gate-allaround transistor ； piezoelectric material ； thin film bulk acoustic waver resonator ； Mason model

參考文獻

[2] Shien-Yang Wu, C.Y. Lin, M.C. Chiang, J.J. Liaw, J.Y. Cheng, S.H. Yang, M. Liang, T. Miyashita, C. H. Tsai, B.C. Hsu, H.Y. Chen, T. Yamamoto, S.Y. Chang, V.S. Chang, C.H. Chang, J.H. Chen, H.F. Chen, K.C. Ting, Y.K. Wu,

Chapter 1

[2] G. E. Moore, “Cramming more components onto integrated circuits (Reprinted from Electronics, pg 114-117, April 19, 1965),” Proc. IEEE, vol. 86, no. 1, pp.82-85, Jan, 1998.

[4] Aaron Thean, “Options Beyond FinFETs at 5nm Node,” IEEE International electron Device Meeting, short course, 2016.

[5] C. W. Liu, Mikael Ö stling, and J. B. Hannon. "New materials for post-Si computing." MRS bulletin 39.08 (2014): 658-662. 000.