Investigation of Fired and Non-fired Si-SiN x Interface Properties by Deep-level Transient Spectroscopy measurements
- PDF / 284,235 Bytes
- 6 Pages / 612 x 792 pts (letter) Page_size
- 60 Downloads / 192 Views
1210-Q05-04
Investigation of fired and non-fired Si-SiNx interface properties by deep-level transient spectroscopy measurements Chun Gong, Eddy Simoen, Rui Yang, Niels Posthuma, Emmanuel Van Kerschaver, Jef Poortmans and Robert Mertens IMEC, Kapeldreef 75, B-3001 Leuven, Belgium ABSTRACT In this paper, fired and non-fired direct PECVD deposited Si-SiNx interface properties with and without NH3 pretreatment on both n- and p-type mono-crystalline silicon samples were investigated with deep-level transient spectroscopy (DLTS) measurements. A and B defect states are identified at the Si-SiNx interface. Energy-dependent electron and hole capture cross sections were measured by small-pulse DLTS. Fired samples with NH3 pretreatment show the lowest DLTS signals, which suggests the lowest overall Dit. The combination of NH3 pretreatment and firing is also suggested for application in the solar cell fabrication. INTRODUCTION Silicon nitride (SiNx) films deposited by plasma-enhanced chemical vapor deposition (PECVD) are widely used in silicon solar cell fabrication as passivation layers [1]. The use of hydrogenated silicon nitride is one of the most significant technological evolutions that has taken place in the solar cell industry, due to its ability to act simultaneously as antireflective coating as well as a source of hydrogen for surface and bulk passivation. Moreover, the firing step after SiNx film deposition is assumed to improve the passivation quality by forming hydrogen terminated dangling bonds [2]. However, until now, most of the published data on the Si-SiNx interface have been based on C-V measurements of MIS structures which mainly give interface states density (Dit) information. The aim is to investigate both fired and non-fired Si-SiNx interface properties by deep-level transient spectroscopy measurements which can provide insight on both Dit distribution and capture cross sections for electrons and holes σn and σp [3]. THEORY AND EXPERIMENT Sample preparation Monocrystalline FZ 0.5 Ohm.cm p- and CZ 2 Ohm.cm n-type silicon wafers with a orientation were used. Before the plasma deposition, wafers received a standard RCA clean and an HF dip prior to SiNx deposition. The deposition of Si-SiNx films was performed by a direct-plasma PECVD reactor with two parallel plates. For lifetime measurement samples, SiNx films were deposited on both sides. Meanwhile, DLTS samples received single-side deposition. The thickness of the deposited films was around 80 nm. Two deposition recipes with and without NH3 pretreatment were applied on both p- and n-type samples. NH3 pretreatment is an additional treatment before SiNx film deposition and it is used to remove the native oxide which is formed in the time period between the chemical cleaning and the start of the SiNx film deposition. After deposition, samples were diced in two. One half received an extra firing step. Then circular gate
contacts consisting of 1 µm aluminum were deposited onto the SiNx films for the DLTS samples by e-beam evaporation through a shadow mask. I
Data Loading...
