ECR Plasma CVD淀积光电器件介质膜的工艺模拟

Simulation based prediction of dielectric thin film deposition by ECR plasma CVD for optoelectronic devices

摘要: 以实验数据为基础,运用人工神经网络方法,建立了电子回旋共振等离子体化学气相沉积(ECRPlasmaCVD)淀积硅的氮、氧化物介质膜的折射率、速率与气流配比Q(N₂)/Q(SiH₄)和Q(O₂)/Q(SiH₄)关系的数学模型,此模型在给定气流配比Q(N₂)/Q(SiH₄)和Q(O₂)/Q(SiH₄)时所预测的成模折射率跟实验值符合得很好,为ECRPlasmaCVD淀积全介质光学膜的工艺打下坚实的基础。

Abstract: Based on experiment result, a mathematical model has been established in this paper for the relation between refractive index n and deposition ratio V of silicon oxide or nitride and gas ratio Q(N₂)/Q(SiH₄), Q(O₂)/Q(SiH₄) by artificial neural network. A good prediction is made for the refractive index n and depositon ratio V by the model when gas ratio Q(N₂)/Q(SiH₄), Q(O₂)/Q(SiH₄) are given. And the model is basic of simulatin for all dielectric optical coatings deposition by ECR plasma CVD.