Cu-plated contacts enable 18.4% conversion efficiency for large area solar cells

At the European Photovoltaic Solar Energy Conference EUPVSEC in Hamburg, Germany, IMEC presented a large-area solar with a conversion efficiency of 18.4%. Compared to the consortium’s standard i-PERC cell process, IMEC’s solar cell features a shallow emitter and advanced front metallization using copper plating. The results were obtained on large-area cells (125cm²), proving the industrial viability of the process.

IMEC’s i-PERC cell with shallow entter and Cu metallization.

The shallow emitter results in an enhanced blue response, and thus in a higher conversion efficiency than with a standard emitter. For the front contacts, a novel metallization stack is added which is applied to local openings in the antireflective coating. Dr. Joachim John, team manager at IMEC said the move to copper from from screen-printed silver pastes was critical to achieving grid partiy. “The way to grid parity does not work with the technology in use today,” he said. “Using copper instead of silver adds to the sustainability of solar cell production. IMEC was able to do this because it has extensive experience with copper plating on silicon”. A similar efficiency result was obtained with screen printed contacts, but the long-term sustainability and low-cost potential of Cu-based contacting solutions and the fact that this was a first result obtained without dedicated fine-tuning makes this result particularly encouraging.” John said that the industry needs to move to much finer lines in order to block less light from entering the cell. Screen printing is limited to 100 μm or so. Copper can easily achieve 50 μm lines.  “We think at IMEC that silver is not the way to go. This is not only because we have a lot of experience (with copper) but it’s cheaper and more sustainable,” he said. The disadvantage of copper is that it requires seed layers and diffusion bayers, but again experience with copper in semiconductor manufacturing makes this a well understood issue.

 

IMEC’s i–PERC concept, introduced in 2005, ia based on processing of very thin cells. PERC stands for `passivated emitter and rear cell.” The `i’ stands for `industrial’, and refers to the fact that, in contrast to the PERC concept, this process is based on industrially applicable techniques. The process relies on an efficient dielectric passivation of the rear surface and local contacts. Openings are made in the passivating layer and Al is printed over the whole rear surface. During the short thermal anneal, an alloy is formed locally, creating a small BSF region, which shields the minority carriers from the high recombination velocity at the contacts. The resulting structure at the rear has a low average recombination velocity and provides a good contact with the rear electrode.

 

Importantly, this is achieved by a process sequence that is fully compatible with the current practice for industrial solar cells (including screen printing and belt-furnace annealing). Because the stress induced by Al-alloying is only local in the i-PERC, in contrast with the extended stress field induced by a full Al-BSF, the bowing problem is totally eliminated, even for wafers as thin as 80µm.

 

Dr. Jef Poortmans, IMEC’s Photovoltaics Program Director, commented on the new PERC cells that employ Cu contacts: “These cells and the new metallization stack involved are a further successful step in IMEC’s target to develop ever more cost-effective, efficient crystalline Si solar cells – eventually targeting cells that are only 40µm thick with efficiencies above 20%.

 

John said that the team will initially focus on advancing solar cell technology, but that that new cell concepts will require new module concepts. “We are focused pretty much on the cell but the module is the technology demonstrator for the cell,” he said.

 

Another part of the IMEC program will look at shallow emitters which also hold the promise to increased efficiency. Today, the top emitter region is fairly thick, about 195 nm. The problem with this, says John, is that many carriers rapidly recombine before they reach the conductors. Shallow emitters would be either lower doped or have higher resistance.