A gas-phase reaction mechanism is proposed for the chemical vapor deposition (CVD) of amorphous silicon from silane or disilane at atmospheric pressure. The gas stream in the CVD reactor is populated by silanes, silylenes, and disilenes in a variety of sizes. Silylenes form by the decomposition of silanes, and they rapidly insert into other silanes to form larger silanes. Although silylenes are expected to stick to growth surfaces to which they diffuse, they are too reactive in the gas phase to deliver a large flux onto the growth surface. Larger silylenes (SiH3SiH and larger) also isomerize to form less reactive disilenes, which we propose to be principally responsible for film growth. Film profiles observed in depositions from silane and disilane are presented, and computed film profiles are compared to these observations. Deposition from silane is explained quite well by the mechanism, as are some qualitative features of deposition from disilane.

No abstract

The flat glass industry successfully commercialized pyrolytically deposited low emissivity glass based on color-suppressed fluorine-doped tin oxide in 1990. The current production capability of this product approaches 500,000 m2 of coated glass per year. Spin-off products for the photovoltaic (PV) industry were introduced by LOF in 1991 and enjoy considerable success in this emerging technology. Many PV researchers, however, still prefer to produce optimized tin oxide off-line in CVD furnaces as they can be produced on higher performing but higher cost substrates. In this paper, we discuss the characteristics of on-line production of PV transparent conductors (TCO), including the tradeoffs required when a coating process is incorporated onto a float glass line. We conclude by suggesting that eventual large scale commercialization of thin film photovoltaics dictates the use of on-line TCO based on economics, the excellent uniformity and repeatability, and ready availability. We believe that the PV researchers are best served by understanding the capabilities of on-line tin oxide production, and tailoring their programs towards the characteristics of these new cost effective TCOs.

No abstract

No abstract