A research team from South Korea has made significant strides in the recycling of silicon from end-of-life (EoL) photovoltaic (PV) modules, successfully converting it into silicon nitride (Si₃N₄). This high-value ceramic material is sought after in various industries, including automotive, aerospace, electronics, medical, energy, and manufacturing, due to its exceptional properties such as high strength, thermal stability, wear resistance, and electrical insulation.
Groundbreaking Research
According to Jin-Seok Lee, the corresponding author of the study, this represents the first instance of silicon recovered from actual EoL PV modules being transformed into Si₃N₄. The research team’s objective was not merely to recover silicon as a secondary raw material but to establish a viable pathway for repurposing recycled silicon into a higher-value application.
Future Directions
Looking ahead, Lee indicated that the team is focused on advancing from proof-of-concept research to a practical and scalable recycling and upcycling process. Collaborating with Wonkwang S&T, a Korean PV recycling company, they aim to develop mobile PV recycling technology. This innovation seeks to process EoL PV modules closer to their point of generation, which could potentially reduce transportation costs by around 30% and lower carbon emissions by more than 10% compared to traditional centralized recycling methods.
Methodology of Upcycling
The researchers began their process with a single end-of-life Suntech STP200-18/Ub module, which contained 54 polycrystalline silicon cells based on an aluminium back-surface field (Al-BSF) architecture. The initial steps involved removing the junction box and aluminium frame, followed by separating the glass from the ethylene-vinyl acetate (EVA)/cell/EVA/backsheet laminate using a hot knife. The laminate was then cut into cell-sized pieces, which were milled at varying speeds (400 rpm, 600 rpm, and 800 rpm) to evaluate the impact of milling speed on particle agglomeration and impurity removal.
To further refine the material, large backsheet fragments and residual EVA were removed using 3 mm and 0.5 mm sieves, respectively. The remaining organic materials were combusted in air at 600 °C for one hour.
Purification Process
Following the initial processing, the researchers employed particle-size analysis and scanning electron microscopy (SEM) imaging to assess agglomeration. The silicon underwent a two-stage purification process: first, it was treated with 36 wt% hydrochloric acid (HCl) for 20 minutes to eliminate aluminium, copper, tin, and lead; then, it was treated with 36 wt% nitric acid (HNO₃) for 30 minutes to dissolve silver. Both treatments were conducted with stirring and ultrasonication to enhance effectiveness.
To remove titanium dioxide (TiO₂), which is resistant to acid and originates from the backsheet, the researchers dispersed 10 grams of powder in one litre of water for 20 minutes. They allowed the mixture to settle for a period ranging from 5 to 20 minutes, ultimately determining that a 5-minute settling time was optimal for removing 800 mL of supernatant.
Nitridation and Characterization
After characterisation, the highest-purity powder, produced at 400 rpm, was selected for nitridation. Both purified and unpurified powders were ball-milled in ethanol for 20 hours to achieve an average particle size of approximately 1 µm. The nitridation process was conducted under a flow of 95% nitrogen and 5% hydrogen, first at 1,350 °C for one hour and then at 1,450 °C for 10 minutes.
The researchers employed X-ray diffraction (XRD) to calculate the conversion of silicon and to determine the proportions of the alpha and beta crystalline phases of silicon nitride (α-Si₃N₄ and β-Si₃N₄). Additionally, SEM was used to compare particle morphology, while inductively coupled plasma optical emission spectrometry (ICP-OES) measured the purity of the final ceramic product.
Findings and Implications
Lee noted that the degree of particle agglomeration during milling significantly influences the effectiveness of metallic impurity removal. By optimising milling conditions and combining stepwise acid etching with a straightforward sedimentation process, the team successfully controlled both metallic and ceramic impurities in the recovered silicon.
The results indicated that excessive particle agglomeration at 800 rpm led to a residual aluminium concentration of 4,290 ppm after HCl etching, in stark contrast to just 189 ppm at 400 rpm. This finding underscores the importance of optimising recycling processes rather than simply increasing milling intensity.
Moreover, the sedimentation process proved to be remarkably effective; within just 5 minutes, 71.4% of the TiO₂ impurity was removed while maintaining a silicon recovery rate of 92.3%. Ultimately, after managing these impurities, the recycled silicon achieved a purity level of 99.95%, with the resulting Si₃N₄ containing 93.1% α-Si₃N₄. In comparison, Si₃N₄ synthesised from recycled silicon without the additional purification process contained only 54.7% α-Si₃N₄. This highlights how impurities from waste PV modules can directly impact the characteristics of the final upcycled product.
Conclusion
The findings of this research are detailed in the study titled “Upcycling silicon recovered from photovoltaic waste into silicon nitride via the field-applicable control of metal and ceramic impurities,” published in Materials Today Sustainability. The collaborative effort involved researchers from the Korea Institute of Energy Research and Chungnam National University, marking a significant advancement in the sustainable recycling of PV materials.
Source: pv magazine Global
