Yingli Green Energy updated the financial community as to its long-term technology roadmap during its fifth annual investor day event during Solar Power International 2014, held in Las Vegas.
Brian Grenko, VP Technology at Yingli Green Energy Americas discussed both multicrystalline (multi c-Si) and monocrystalline (mono c-Si) technology roadmaps that remain on-track to meet commercial production efficiency targets in 2020 of 19% and 23% respectively.
In effect, both roadmaps have not changed from the previous year, when Yingli Green first updated its cell technology roadmap beyond 2015 to include expected cell efficiencies through 2020.
Then as now, Yingli Green expects to narrow the gap between lab results and actual volume production with only 1% difference for both multi and mono technologies by 2020.
All solar cell efficiency data points on an annual basis between lab and volume production remain unchanged as well as 2020 peak cell efficiency milestones as outlined at the same event in 2013.
However, details of how the company plans to meet its technology roadmap were briefly detailed.
Multi c-Si technology enhancements
Unsurprisingly, Grenko highlighted that higher quality wafers were key part of the overall cell efficiency drive within Yingli Green’s multi c-Si technology roadmap.
However, the benefits also transcended lower manufacturing costs, a key driver of the company for many years. As such the focus hasn’t changed much with continued focus of the optimisation of its DSS furnace casting process that it designed to continuously lower the defect densities.
Grenko noted that changes in the ingot crucible used by the company had led to increased ingot yield, lower defect density as well as lower costs. Modifications to the crucible geometry, further cycle time reductions and lower electricity costs were key contibutors.
The VP of technology also highlighted the use of high-purity coatings on the crucibles and wafer surface texturing reduced tail waste, while improving bulk minority carrier lifetime of the wafer. Lower waste and operating cost through increases in saw slurry recycling were also said to have contributed to lower wafering costs.
High-efficiency multi c-Si wafers were being produced that had significantly better grain size consistency, increasing cell efficiency by up to 3%, according to Grenko.
At the solar cell, Grenko highlighted stencil printing and double printing processes that have contributed to its technology roadmap.
The use of stencil printing had the benefit of reduced finger widths and a reduction in line-edge roughness that increased the active surface area of the cell, while decreasing the consumption of silver paste by up to 25%, subsequently lowering non-silicon production costs. Lower series resistance at the busbar interface was also noted as another advantage of stencil printing over conventional screen printing.
Grenko said that when double printing was used in combination with stencil printing, significantly higher cell efficiencies could be obtained by lower series resistance, with focus currently on improving the ohmic contact with the first print step, and improving conductivity with second print step.
Multi c-Si cells would also benefit from anti-reflective (AR) soldering ribbon that would increase light capture, boosting power while improving cell aesthetics. Grenko noted that manufacturing trials of the AR ribbon on Yingli’s standard 60-cell modules had provided around a 5W power gain overall.
Yingli Green expects its multi c-Si modules to benefit from the incremental cell efficiency gains through 2017, resulting in average module power outputs (60-cell modules) of 260W in 2015 and reaching close to 270 in 2017.
PANDA cell technology roadmap update
According to Grenko, cell efficiency gains remain the most significant potential cost reduction driver through 2020. Grenko reiterated that Yingli’s PANDA cell technology was introduced back in 2010 through a collaboration project with Amtech Systems and ECN. However, further technology advancements to PANDA were now being implemented.
Beginning in 2014, Yingli Green had incorporated ion implantation into its mono c-Si cell architecture.
Ion implantation has the ability to provide around 1% absolute cell efficiency gain with selective emitters, while reducing eliminating steps such as edge isolation and dopant glass removal, while providing the opportunity for highly consistent oxide passivation during implantation anneal at no additional cost.
The technology also potentially allows for further cell efficiency gains and lower costs by replacing traditional aluminium (Al) BSF process that could enable the use of ultra-thin and or less expensive wafers with higher defectivity as well as enabling the formation of boron doped emitters and phosphorus (P) doped front or back surface fields to enable n-type front and back junction cells.
“We see more disruptive technology to be unlocked than what we have shown here,” noted Grenko at the investor event.
Yingli Green is targeting PANDA cells using ion implant to deliver PANDA cell efficiencies of around 20.1% by year-end. The company had achieved average cell efficiencies of 19.7% in its production line.
Interestingly, Kingstone Semiconductor, a majority-owned subsidiary of Amtech Systems had previously developed its IonSolar system, which had been placed at ECN in 2013. However, Varian Semiconductor, now part of Applied Materials had also developed its ion implanter to solar cells several years ahead of Kingstone, which is being used at SunIva's fab in the US.
The move to MWT is expected in 2015 with Yingli Green highlighting cell efficiencies of around 20.5% and optimised ion implantation with MWT providing cells efficiencies of 20.8% in 2016.
Although Grenko did not go into any detail regarding the PANDA cell roadmap to 23% cell efficiencies in 2020, a chart presented highlighted that well known cell technology upgrades were clearly compatible with the Panda cell and gave an insight into future upgrades to mono c-Si cells over the next six years.
Yingli Green has had 600MW of integrated PANDA production in place since the second quarter of 2011 but has not yet ramped the lines to full capacity for PANDA only production.