甚高頻電容式耦合矽烷/氫氣電漿模擬研究 —電漿物理化學機制與操作參數關聯性之分析探討

Study of a VHF capacitively coupled silane/hydrogen discharge by computer simulation – physical/chemical mechanism and parametric analysis

連頌恩

指導教授 ： 陳金順 ； 柳克強

國立清華大學/原子科學院/工程與系統科學系/碩士(2013年)

https://doi.org/10.6843/NTHU.2013.00465

摘要

近年來，微晶矽薄膜用在薄膜式太陽能電池越來越受重視，而矽薄膜沉積採用的電漿輔助化學氣相沉積（PECVD）系統因為高電漿密度與低離子撞擊能量等優點，從傳統的射頻（RF）功率轉向甚高頻（VHF）。本研究主要目標為建立應用於沉積微晶矽薄膜的VHF-PECVD系統之模擬計算分析，探討相關電漿以及矽薄膜成長之物理與化學機制，以提供設計電漿設備與改變實驗參數作有效參考的工具。
根據模擬結果在頻率為80 MHz的VHF電漿環境基本模擬條件下，電子密度、電子溫度與H、SiH3的粒子密度在65 ms前就已全部到達穩態。H與SiH2主要是由電子碰撞SiH4生成，而SiH3則主要由H與SiH4反應產生，使得H與SiH2在電漿中成雙峰分佈，而SiH3呈現鐘形分佈。電漿沿徑向分佈在3.0 cm內呈均勻。一個電壓周期內，電漿的電位分佈會隨著功率電極電位而改變，電漿與電極之間的電位差，會在靠近兩電極處產生強電場，可以加速電子來激發其餘重粒子，並使電漿中的正離子受電場加速往基板移動。
隨著功率的提高，平均電位、電子密度與電子溫度隨之增加，而H、SiH2、SiH3的密度以及通往基板的通量密度也上升，H/SiH3通量密度的比值也增加。推論對成膜速率與結晶率有正向影響。
電極間距增加會使電漿密度減少、平均電位下降，H、SiH2、SiH3的粒子密度和通量密度也下降，同時讓H/SiH3的通量密度比降低。推論對成膜速率和結晶率不利。
改變壓力時，電子密度、平均電位與H、SiH2的粒子密度隨著壓力上升而下降，H、SiH2、SiH3的通量密度與H/SiH3通量密度的比值也下降。推論電極間距增加會使成膜速率與結晶率下降。
增加SiH4流量不影響基本放電特性，但會增加SiH3的粒子密度，同時使得H、SiH3、SiH2的通量密度上升，由於SiH3的通量密度上升較快，故H/SiH3通量密度比值下降。推論在增加SiH4的比例時，會使成膜速率上升，但會使結晶率下降。
80 MHz與 27.12 MHz做比較，發現在80 MHz時腔體中央與電極之間的電位差下降，使得電子溫度下降、電子密度上升。SiH3的通量密度隨功率上升的梯度上升，可推論在高頻下沉積速率比低頻下更容易受功率影響。H/SiH3通量密度比顯示80 MHz時薄膜結晶率上升。顯示頻率提高對成膜速率與結晶率都有好處。比較不同頻率下改變壓力的模擬，可發現幾乎所有電漿特性的變化皆相同，只有SiH3的粒子密度在80 MHz不隨壓力而改變，而在27.12 MHz時，隨著壓力上升而上升。原因應為SiH3主要生成反應式在27.12 MHz時，反應速率隨著壓力上升而上升，但在80 MHz時，則是隨著壓力上升而下降。
根據本研究模擬結果，提高輸入功率、縮小電極間距、降低壓力、提高電漿頻率皆有利於微晶矽薄膜的成長，而提高通入氣體中SiH4的比例，雖然會使成膜速率上升，卻會使得結晶率下降，故須尋找一個最佳的平衡點。

關鍵字

電漿輔助化學氣相沉積 ； 矽烷/氫氣電漿 ； 微晶矽

並列摘要

In the recent years, micro crystal silicon thin film is more popular on thin film solar cell. Because higher plasma density and lower ion energy, the frequency of PECVD is turning from RF region to VHF region. The goal of our study is to build a VHF-PECVD model and to find out how the plasma behavior change with control variables by simulation.
For the VHF plasma in 80 MHz, electron density, electron temperature, and number density of H and SiH3 go to steady state before 60 ms. H and SiH2 are produced by electron collision reaction with SiH4, and SiH3 is produced by the reaction of H and SiH4. Because the different production reaction, the distribution of the density of H, SiH2 and SiH3 is different: the density of H and SiH2 is two-peak distribution and the density of SiH3 is bell-shaped distribution. The plasma is uniform in the radial direction in the range of a radius of 3.0 cm. In one period, the plasma potential changes with the potential of power electrode. The electric field would be created by the voltage between the plasma and electrode, and accelerate electrons and ions.
When the power increase, electron density, electron temperature, number densities and fluxes of H, SiH2, and SiH3 are increased. The ratio of H and SiH3 flux is increased, too. By the result, we may say that the deposition rare and crystallization rate would increase with power. It's good for thin film deposition.
When the gap between two electrodes is enlarged, the plasma density, number densities and fluxes of H, SiH2, and SiH3 and ratio of H and SiH3 flux are decreased. The result is not good for the deposition rare and crystallization rate.
When the pressure is increased, the plasma density and number density of H and SiH2 are decreased. The flux of H, SiH3 and SiH2 and the ratio of H and SiH3 flux are also decreased. It means that the deposition rare and crystallization rate decrease with pressure increased. Only the number density of SiH3 doesn't change with pressure.
The plasma density and electron temperature do not change when the SiH¬4 flow rate increase. But the number density of SiH3 and flux of H, SiH3 and SiH2 increase with the SiH4 flow rate. Because the flux of SiH3 increase faster than H, the ratio of H and SiH3 flux is decreased with increased SiH4 flow rate. It makes good deposition rate, but bad crystallization rate.
Compare with the case in 27.12 MHz, the result shows that the voltage between plasma and electrode is much lower in 80 MHz. It makes the electron temperature lower, but higher electron density. The SiH3 density in 80 MHz grows faster with power than in 27.12 MHz. It means that the deposition rate is more sensitive with change of power. The ratio of H and SiH3 flux in 80 MHz is higher than in 27.12 MHz. We may say that crystallization rate in 80 MHz is better than in 27.12 MHz.
Compare the case with pressure changing in 27.12 MHz and 80 MHz, we can find out that most of the behaviors of plasma have same tendency, beside SiH3 density. The density of SiH3 is increased with pressure in 27.12 MHz, but constant in 80 MHz. The reason might be that the producing rate of SiH3 is increased with pressure in 27.12 MHz but decrease in 80 MHz.
From the simulation result, we can say that the deposition rate and crystallization rate are better with higher power, smaller gap, lower pressure and higher frequency. Increasing the flow rate of SiH4 would increase the deposition rate, but decrease the crystallization rate.

並列關鍵字

PECVD ； SiH4/H2 ； microcrystal silicon

參考文獻

28. 張雅嵐, 微晶矽薄膜電漿輔助化學氣相沉積製程之電漿放射光譜量測分析研究. 國立清華大學工程與系統科學系碩士論文, 2010.

1. Kushner, M.J., A MODEL FOR THE DISCHARGE KINETICS AND PLASMA CHEMISTRY DURING PLASMA ENHANCED CHEMICAL VAPOR-DEPOSITION OF AMORPHOUS-SILICON. Journal of Applied Physics, 1988. 63（8）: p. 2532-2551.

3. Chapman, B., ed. Glow Discharge Processes. 1980, Wiley: New York.

5. Kuo, Y., K. Okajima, and M. Takeichi, Plasma processing in the fabrication of amorphous silicon thin-film-transistor arrays. Ibm Journal of Research and Development, 1999. 43（1-2）: p. 73-88.

6. Fauchet, P.M., The integration of nanoscale porous silicon light emitters: materials science, properties, and integration with electronic circuitry. Journal of Luminescence, 1998. 80（1-4）: p. 53-64.