PECVD制备微晶硅工艺.ppt

preparation of μc-Si:H 8.15项目组 outline ZnO surface pre-treatment ZnO surface pre-treatment ZnO surface pre-treatment ZnO surface pre-treatment p-layer p-layer p-layer p-layer p-layer p-layer p-layer p-layer p-i interface p-i interface p-i interface i-layer i-layer i-layer i-layer i-layer i-layer i-layer i-layer i-layer i-layer i-layer n-layer n-layer That’s all.Thanks!! * 让不可能成为可能 Making the IMPOSSIBLE possible ZnO surface pre-treatment p-layer p-i interface i-layer n-layer Data from Neuchatel, Julich, oerlikon TCO influences amorphous voids and cracks H2 CO2 Good φc: nucleation layer needed to be very thick(30~70 nm) φc0.5; hydrogen etching to inhomogeneity compare pre-deposition plasma of H2 or CO2 undoped nucleation layer Hydrogen recrystallisation way characteristic Way choice Good p-type μc-Si:H layer φc ≥ 0.5 (on ZnO) σdark (RT)1 S/cm ( thickness of ~20- 30 nm) Material Frequency Power TMB/ SiH4 H2 flow choice B(CH3)3 --TMB vs. B2H6--diborane higher thermal stability, (T≈ 200°C) source-gas B2H6 self-decomposition in gas bottle No self-decomposition of TMB (without plasma) Material Frequency Power TMB/ SiH4 H2 flow φc is too low loss in VOC ~0.55 to~0.65 for 70.0 and 110.0 MHz on ZnO 5.5 sccm TMB; 60sccm H2 TMB/SiH4 ratio= 0.25 % i-layer amorphous σdark is too small Material Frequency Power TMB/ SiH4 H2 flow 70.0 MHz , TMB/ SiH4= 0.45 %, 60sccm H2. “standard” power of 20 W Material Frequency Power TMB/ SiH4 H2 flow TMB/ SiH4= 0.75 % (~10 sccm TMB), φc~ 0.5 σdark ~2.6 S/cm 110.0 MHz, 60sccmH2 Material Frequency Power TMB/ SiH4 H2 flow too high –- decreases φc too low –- insufficient σdark p= 0.6 mbar , 20 W , 110.0 MHz, TMB/SiH4 = 0.5 %, 10sccm TMB. 80sccm H2 φc ~0.5 to ~0.7 σdark2 S/cm Material Frequency Power TMB/ SiH4 H2 flow p= 0.6mbar, 20 W, 110.0MHz, TMB/SiH4 = 0.75 %, 10sccm TMB,80sccm H2 Φc 0.5 ~0.6 Dark 2~3 S/cm on ZnO-coated substrates μc-Si: