The emitter saturation current density (J Oe) and surface recombination velocity (S p) of various high quality passivation schemes on phosphorus-diffused solar cell emitters have been determined and compared. The passivation schemes investigated were ͑i͒ stoichiometric plasma enhanced chemical vapor deposited ͑PECVD͒ silicon nitride ͑SiN͒, ͑ii͒ forming gas annealed thermally grown silicon oxide, and ͑iii͒ aluminum annealed ͑alnealed͒ thermal silicon oxide. Emitters with sheet resistances ranging from 30 to 430 and 50 to 380 ⍀/ᮀ were investigated for planar and random-pyramid textured silicon surfaces, which covers both industrial and laboratory emitters. The electronic surface passivation quality provided by PECVD SiN films was found to be good, with S p values ranging from 1400 to 25 000 cm/s for planar emitters. Thin thermal silicon oxides were found to provide superior passivation to PECVD SiN, with the best passivation provided by an alnealed thin oxide (S p values between 250 and 21 000 cm/s͒. The optimized PECVD SiN films are, nevertheless, sufficiently good for most silicon solar cell applications.

The current density–voltage (J–V) curve that characterises the performance of a solar cell is often extra rounded, resulting in reduced efficiency. When fitting to the standard one‐dimensional models, it is often found that the rounding cannot be fitted by the series resistance only. In these cases, the diode factor m or the depletion region saturation current density J0DR (depending on the model used) is increased. This behaviour could not be explained so far; this paper discusses if a nonuniform contact resistance of the front side metallisation leads to an increase of m or J0DR. The theoretical part of the investigation is the simulation of the curve for a cell with two regions with different contact resistance. It was found indeed that m or J0DR is increased, while the series resistance is not increased as much as expected. The experimental part was the calculation of the J–V curve of a high‐m solar cell with local contact resistances measured with the so‐called Corescan and the cell's resistanceless J–V curve as measured with the so‐called Suns‐Voc method. The calculated curve approximated the actual curve quite well, demonstrating in practice that high diode factors can be explained by nonuniform contact resistance. Copyright © 2004 John Wiley & Sons, Ltd.

A novel high efficiency solar cell and module technology, named PANDA, using crystalline n-type CZ Si wafers has moved into large-scale production at Yingli. The first commercial sales of the PANDA modules commenced in mid 2010. Up to 600MW of mass production capacity from crystal-Si growth, wafer slicing, cell processing and module assembly have been implemented by the end of 2011. The PANDA technology was developed specifically for high efficiency and low cost. In contrast to the existing n-type Si solar cell manufacturing methods in mass production, this new technology is largely compatible with a traditional p-type Si solar cell production line by conventional diffusion, SiNx coating and screen-printing technology. With optimizing all technologies, Yingli's PANDA solar cells on semi-square 6-inch n-type CZ wafers (cell size 239cm 2) have been improved to currently have an average efficiency on commercial production lines exceeding 19.0% and up to 20.0% in pilot production. The PANDA modules have been produced and were certified according to UL1703, IEC 61215 and IEC 61730 standards. Nearly two years of full production on scale-up lines show that the PANDA modules have a high efficiency and power density, superior high temperature performance, near zero initial light induced degradation, and excellent efficiency at low irradiance.

No abstract

This article will review our recent progress in development of high-efficiency cells on n-type monocrystalline Si wafers. With boron-doped front emitter, phosphorous BSF, and screen-printed metallisation, at this moment such cells reach an efficiency of over 19%. We describe recent results of processing with reduced front contact area, and improved BSF and improved rear surface passivation, which are key parameters that limit the cell efficiency. The improved processing leads to an efficiency of 20%. The cell process has also been adopted for fabrication of metal-wrap-through back-contact cells. Without the improved contact recombination and BSF, an MWT cell efficiency of 19.7% is reached, 0.3% higher than the corresponding 'standard' (non-back-contact) cells.