建立模擬非晶形矽薄膜沉積的化學動力學資料庫( I )

Abstract

近年來，由於高純度矽晶的大量供應及高能需求，導致太陽能電池工 業考慮以非晶體矽薄膜取代矽晶，以降低矽太陽電池的價格並提高其競爭 力。在新的研究期間(July 1, 2011-June 30, 2014)，本人將全力研發有關大面 積矽薄膜(a-Si:H)利用PECVD (plasma enhanced chemical vapor deposition)及 cat-CVD (catalytic CVD)沉積的重要化學反應機制及反應速度常數量子計 算。 SiHx及SixHy (x=1-3; y=1-8)之氣相及其在Si 表面之化學反應與a-Si:H 薄膜之生長直接相關，但這些反應幾乎完全沒有文獻資料，本研究計劃將 以三年時間，利用高階量子及統計化學動力學計算其與溫度及壓力相關之 化學反應速度常數，以供三維計算流體力學及電漿流體模擬之用；並優化 a-Si:H 薄膜之生成速率、建立PECVD 及Cat-CVD 生長之完整程式。本計劃 在計算流體力學及電漿流體模擬方面，將與交大機械系吳宗信教授合作， 吳教授乃目前世界上知名三維CFD 電漿模擬專家之一。

In this new funding period (July 1, 2011-June 30, 2014), our primary focus will be placed on studies of chemical processes directly relevant to the fabrication of large area amorphous silicon (a-Si:H) thin films by plasma enhanced chemical vapor deposition (PECVD) and catalytic chemical vapor deposition (Cat-CVD). The deposition of high quality homogeneous a-Si:H thin film over a large area is a critical technology for manufacturing of a new generation of cheaper Si-solar cells replacing today’s more energy and material intensive polycrystalline Si-solar cells. In order to control the homogeneous growth of an a-Si:H thin film over a large area, realistic 3-D kinetic modeling of the growth process using a state-of-the-art computer code, which includes a realistic reaction mechanism with reliable rate constants and complex hydrodynamics, is required. Although the latter physical processes involving diffusion, convection, electric field distribution, among others, can be simulated quite well with the modern computational fluid dynamics (CFD) and plasma fuid modeling codes, the detailed kinetics and mechanisms for chemical processes involving reactions of SixHy species in the gas phase and on Si-surfaces are very limited. In this work, we propose to investigate the kinetics and mechanisms of the SixHy reactions occurring in both phases under the T,P-conditions directly relevant to the deposition techniques employed using the state-of-the-art quantum chemical and statistical reaction-rate theories, which have been so successfully utilized by us recently to characterize atmospheric ClOx kinetics, hydrocarbon combustion kinetics and heterogeneous reaction kinetics occurring in fuel cells, etc. Our detailed ab initio chemical kinetics data will be provided for simulations of a-Si:H thin film growth by one of the world’s experts at NCTU employing both 3-D CFD and plasma fluid modeling codes.