The long read: The right time for TOPCon

From pv magazine 03/2022

An eventual switch to n-type technologies has been on the PV cell manufacturing roadmap for several years. Various n-type cell concepts have been in the running and seen some commercial development, including n-PERT, interdigitated back contact, and heterojunction. Work on these continues, however, one n-type cell now looks to be headed into mass production ahead of the rest, with PV InfoLink expecting more than 50GW of production capacity online this year.

Work on the tunnel oxide passivated contact (TOPCon) cell concept dates back to the 1980s and research published by UNSW’s Martin Green. The term TOPCon, however, was first coined in a 2013 paper from scientists at Germany’s Fraunhofer Institute for solar energy systems (ISE). “A TOPCon cell consists of an n-type silicon wafer, with a boron diffusion on the front, and a full area TOPCon at the rear. The rear TOPCon consists of a thin oxide layer and a phosphorus doped polysilicon layer with a thickness between 80 and 150 nm,” explained Martin Hermle, head of department for high-efficiency silicon solar cells at Fraunhofer ISE. “The front side is normally passivated with an aluminum oxide or silicon nitride stack, and the rear side is capped with a silicon nitride layer. The metallization is mostly realized with silver screen printing on both sides.”

Hermle worked with lead author Frank Feldmann on a 2013 paper, “Passivated rear contacts for high-efficiency n-type Si solar cells providing high interface passivation quality and excellent transport characteristics.” It demonstrated TOPCon cells with a maximum efficiency of 23%, and inspired much more research into the concept. “The early publications of Fraunhofer ISE and ISFH triggered a vivid renaissance of research on poly-Si/SiO2 based passivating contacts,” stated ISE researcher Stefan Glunz in a November 2021 paper published in Progress in Photovoltaics, that reviewed recent progress in TOPCon. “This led to many interesting technological and scientific results.”

Since then, TOPCon cell efficiency has been pushed further to 25.8% in ISE’s lab, and Longi set the current record for a “full size” cell at 25.21% in June 2021. In the past year Longi and several others have introduced TOPCon modules to the market.

Maxing out PERC

Also in June 2021, Longi announced the achievement of 25.02% efficiency for a p-type mono PERC cell. In production, however, PERC cell efficiencies much beyond 23% are difficult to obtain. And with the transition to new wafer formats and associated cost optimizations all but complete, manufacturers are again turning to improvements on the cell efficiency side in search of lower costs and better energy yields for the end user.

Added cost and complexity have kept n-type development limited. However, TOPCon is known for its potential for high efficiencies at a cost much closer to that of PERC. Current estimates place TOPCon between $0.01 and $0.02/W more expensive than PERC, and as the efficiency gap between the two widens, the price gap should close. On the technical side, developing and standardizing an industrial process for the doped polysilicon layer deposition has been the biggest challenge. While there is still debate over the most effective approach (see pp. 54-57), manufacturers are ready to move into large-scale production and continue to develop the technology.

JinkoSolar is the biggest mover so far, with its plans representing the bulk of the TOPCon module capacity expected this year. “The efficiency curve is becoming flat as we reach the physical limits of p-type technology,” said Roberto Murgioni, head of technical service and product management Europe at JinkoSolar. “We decided to move strongly into n-type products, and we developed our technology which we call ‘HOT 2.0”, where we use the tunneling layer in the middle to reduce the resistance losses, increase carrier generation and decrease carrier recombination.”

Murgioni says that Jinko is already producing TOPCon cells at 24.5% efficiency, and that the limit for this technology is somewhere around 28.7%, leaving plenty of room for further development.

For now, TOPCon looks to have achieved the best balance between investment cost and efficiency gains to be attractive to manufacturers. Jolywood is another Chinese player with big TOPCon plans this year. The company had initially, from its beginnings in 2008, investigated n-type passivated emitter rear totally diffused (nPERT) technology as a route to bring n-type modules into production. It found, however, that these came with limited efficiency gains at a very high cost. The company switched to producing TOPCon in 2017, and currently operates 3.6GW of cell capacity, with another 16GW fab under construction in China.

Meanwhile, heterojunction (HJT), which has long been TOPCon’s main rival for the title of PERC successor, will likely continue to see some commercial development. PV InfoLink expects HJT capacity to reach 24GW this year, a little under half that of TOPCon. HJT production comes with a considerably higher investment cost, however, alongside the question of very high silver consumption. And this means in the short term at least, TOPCon will see faster growth.

N-type advantage

For the moment TOPCon cells still come at a higher cost per watt than PERC, which might make them a hard sell for some projects, particularly combined with lower maturity and little track record in the field. Beyond higher efficiency potential, however, n-type has a few other tricks up its sleeve that manufacturers are confident will translate into a lower levelized cost of electricity at project level.

First among these is a lower degradation rate, thanks to much lower susceptibility to both light induced degradation (LID) and light elevated temperature induced degradation (LETID). “If you check the degradation rate, the first year will be 1% guaranteed which is half of what’s standard on the market for p-type,” says JinkoSolar’s Murgioni. “Then you have linear annual degradation, which for our TOPCon module is 0.4% rather than 0.55%.”

On top of that, TOPCon comes with a higher bifaciality factor than PERC, making for more generation from the rear side, and also a better temperature coefficient leading to stronger performance in high temperature environments. With these advantages, it is no accident that many of the early large-scale deployments of TOPCon modules are found in the Middle East region, where the harsh climate and bright, reflective conditions make for an even bigger advantage.

A further advantage noted by Murgioni is in low-light performance, allowing TOPCon modules to start generating earlier and go on later into the day. “If you run the LCOE model calculation, you’re going to see average energy gain of 1.5-3% compared to PERC modules,” he said.

Potential production

TOPCon is “ready” for mass production in the eyes of many, and the fact that it can be produced on upgraded PERC lines means the technology lends itself well to a quick ramp up. PERC’s lower investment cost, and the sheer size of the capacity that’s already running, however, means that it will likely retain considerable market share for the next few years at least.

For TOPCon to become even more competitive, there are plenty of routes open in research to lower production costs and raise efficiencies. Alongside settling the debate over the polysilicon deposition method, Fraunhofer ISE’s Hermle identifies reducing silver consumption with copper plating or other methods as a key topic for TOPCon development – while TOPCon uses less silver than HJT, it requires considerably more than PERC due to its use on both sides of the cell.

On the efficiency side, Hermle notes that ISE has already achieved 26% efficiency by moving the emitter layer to the rear of a cell, an approach it calls TOPCon rear emitter, or TOPCoRE. With this approach, the institute expects to be able to achieve cell efficiencies better than 27%. Describing the approach in an April 2021 paper in Nature Energy, scientists at the institute state that they see this as a potential candidate “expected to play an important role in the next decade, before multijunction solar cell concepts … dominate the PV mass production.”