Difficulty in making direct contact, limitation of Al content in paste composition, as well as compatibility with the LECO process are the major challenges for TBC cell p-Poly side contact formation
According to Guo, paste makers should optimize the glass frit and silver powder to improve p-Poly side contact
The optimization of p-Poly paste composition is required for forming good contact on the thinner p-Poly side, says Guo
The ever-evolving trends in PV cell architecture, whether based on p-type or n-type wafers, have driven the development of metallization paste aligned with key cell performance metrics, such as lower metal-semiconductor interface recombination current (J0,metal), reduced contact resistivity, higher Voc, FF, and efficiency gains. Back contact (BC) cell structure, shifting positive and negative metal contacts to the rear side of the cell, demands metallization pastes suitable for making good contacts of both polarities.
At the recent TaiyangNews Annual Virtual Conference – High-Efficiency Solar Technologies 2024 – Fangqing Guo, New Technology Development Director at DKEM, a leading global PV cell metallization paste provider, shared a brief overview of the challenges and solutions in the development of TBC cell metallization paste (see DKEM presentation here).
The advancement of BC cell technology from non-passivated contact to passivated contact for both polarities in TBC cells, the BC variant of TOPCon cell architecture, exhibits multiple metallization challenges, says Guo. Although TBC shares some similarities with TOPCon from a metallization perspective, such as the use of paste for n-Poly contact formation, the development of rear-side p-Poly contacts – unlike the front-side p+ emitter contact in TOPCon – presents multiple challenges for paste makers. Polysilicon thickness, doping concentration, surface structure (polished, textured, or micro-textured), and passivation layers of the TBC cell have raised varying requirements for J0,m and contact resistance balance in the paste, noted Guo. Additionally, the co-firing of the n-Poly and p-Poly areas of the TBC cell at the same temperature, unlike in TOPCon, introduces differences in the paste recipe.
Regarding the contact formation mechanism in TOPCon, Guo explained that the Ag-Al paste, used for the front-side p+ emitter contact, forms a direct contact through AgAl spikes and tunneling with silver crystallites in the glass layer, which reduces ohmic contact resistance. In contrast, the LEF paste forms direct contact by creating AgSi colloids and tunneling with silver crystallites in the glass layer.
For n-Poly contact formation, the metallization paste forms direct contact through silver precipitates and tunneling with silver crystallites in the glass layer. However, for TBC cells, particularly on the 300 nm thick p-Poly side, silver ions are difficult to reduce to form silver precipitates in the glass layer. Additionally, the use of aluminum in the paste is challenging due to its tendency to form large silver-aluminum spikes, which raise J0,m. Moreover, the LEF process has not yet been adopted for p-Poly contact formation, preventing the formation of direct contact. In contrast, the n-Poly contact formation mechanism is largely similar to that of TOPCon. Unlike the textured wafer surface of TOPCon, the polished rear surface of the TBC cell, which has a higher Voc, creates difficulties for the glass frit to adequately soak the surface. This results in random etching on the wafer surface and reduces the chances of forming silver crystallites and silver precipitation. These challenges are particularly prevalent on the p-Poly side, added Guo.
To overcome these metallization challenges on the p-Poly side of the TBC cell, Guo suggested that paste makers should optimize the glass frit and silver powder to improve contact. This can be achieved by optimizing the floating behavior of the glass frit to better soak the flat surface. Additionally, the optimized glass frit design should melt enough silver to form silver ions and precipitates, which can help avoid high J0,m, and increase contact resistance by controlling the size and number of silver precipitates.
Furthermore, the trend of thinning polysilicon layers in both the p-Poly (>200 nm) and n-Poly (250-300 nm) areas, required for cost optimization and higher bifaciality, faces challenges – particularly in the p-Poly area – due to increased contact resistance. Therefore, to reduce p-Poly thickness, fine-tuning the metallization paste composition is essential. For the thinner n-Poly layer, paste optimization involves balancing Voc and fill factor (FF).
Regarding the co-firing of n-Poly and p-Poly pastes at the same temperature, and the limitations of the p-Poly paste in forming a good contact, Guo suggested optimizing the p-Poly paste recipe and firing temperature, followed by optimizing the n-Poly paste for the firing process. Finger morphology, which directly impacts contact area and gridline resistance, can be controlled by adjusting silver powder size, sintering characteristics, and compatibility with the glass frit. This makes the optimization of the p-Poly paste’s organic and silver content key to achieving high-efficiency and low-cost TBC cells, as noted by Guo.
To reduce silver (Ag) consumption for cost-optimized TBC cell metallization, the industry is exploring multiple approaches, including fine-line printing, lower silver-content pastes, and base metal pastes.
DKEM’s paste solutions for TBC cell metallization include the DK95H series for n-Poly paste, the DK73K series for p-Poly paste, and the DK82B series for busbar paste.