Gallium doping keeps p-type in the frame

For several years, the world’s largest silicon PV manufacturers have investigated n-type technologies as a route to higher-performance products. Typically, these have been more expensive to produce, but come with the potential for higher efficiency and lower susceptibility to various degradation mechanisms that hurt performance over time, and therefore higher energy yields.

And in 2022, with the cost gap between n-type and p-type modules narrowing, and further improvements to p-type PERC technology in mass production proving difficult to come by, large-scale PV manufacturers are beginning to make the switch to n-type technology.

However, new research from Germany’s Fraunhofer ISE suggests that with a small change to their production, p-type wafers can achieve similar performance to n-type when integrated into the latest cell architectures.

“Leading wafer manufacturers see fundamental reasons why n-type Cz-Si wafers will remain more expensive than their p-type counterparts,” the researchers observe. “This contribution indicates that switching the doping level and doping species of p-type Cz-Si wafers towards higher resistivities and Ga doping offers an alternative that might be economically interesting compared to n-type material.”

Gallium doping

PV manufacturers began to replace boron doping with gallium around 2019, as a solution to light-induced degradation caused by a reaction between boron and oxygen. But this switch has raised further questions over the effect of gallium, and there is plenty more research underway.

In this investigation, Fraunhofer ISE worked with a major silicon ingot producer to produce a full sized gallium doped silicon ingot, made to ISE’s exact specifications to achieve a resistivity range a full order of magnitude higher than usual commercial products. The group then selected wafers representing the complete length of this ingot, and used them to fabricate cells based on architectures similar to those moving toward the manufacturing mainstream with n-type wafers. These cells were then subjected to a range of performance and reliability tests, and compared to identically processed n-type TOPCon and heterojunction (HJT) cells.

The results of these tests were described in “High Lifetime Ga-doped Cz-Si for Carrier Selective Junction Solar Cells,” which was recently published in RRL Solar. For a tunnel oxide passivate contact rear emitter cell, similar to the TOPCon architecture gaining ground commercially, the best of the p-type cells showed a 0.2% efficiency advantage, while those used in HJT cells were 0.36% less efficient.

The researchers also noted that the cells were produced using processes optimized for n-type cells, and that adaptations could push the p-type efficiency higher. With respect to the TOPCoRE cells, these results already show a significant economic advantage thanks to lower wafer costs and higher efficiency.