WCPS: Ultratrace Analysis of Phosphorus, Boron and Other Impurities in Photovoltaic Silicon and Trichlorosilane by ICP-MS with High Energy Collision Cell

Importance of the Topic

Efficient solar-energy conversion relies on the ultrapure silicon used in photovoltaic (PV) devices. Trace levels of boron, phosphorus and other metallic contaminants can significantly impair cell performance. As demand grows for higher conversion efficiency, sensitive and reliable analytical methods are essential to monitor and control impurities at low parts-per-billion levels.

Objectives and Study Overview

This work aimed to develop and validate a robust procedure for ultratrace analysis of boron, phosphorus and additional elements in PV silicon (wafers, polysilicon and intermediates) and trichlorosilane (TCS). Key goals included lowering detection limits below 1 ppb, resolving isobaric interferences, and demonstrating sample transport and preparation practices that preserve sample integrity.

Methodology and Instrumentation

• Sample digestion: PV silicon was split into two fractions—one for boron and another for other elements—using a controlled HF/HNO₃ mixture. A small H₂SO₄ spike enhanced silicon removal in the non-boron fraction. Heating was carefully stopped before complete dryness to avoid boron loss.
  
• TCS treatment: Conducted under inert, dry atmosphere to convert SiHCl₃ to SiO₂, followed by HF digestion with glycerol addition to stabilize boron during heating.
  
• ICP-MS analysis: Agilent 7700s equipped with a high-energy collision cell. Helium at 2 mL/min with a 100 V kinetic energy discrimination (KED) provided optimal interference removal for 31P16O measurement, while boron was measured directly.
  
• Quality control: Matrix blanks, spiked recoveries at 5 ppb and container-comparison tests (glass vs stainless steel) evaluated accuracy and contamination risks.
  

Main Results and Discussion

• Silicon removal: Addition of H₂SO₄ reduced residual Si to <1 ppb, whereas its absence left 20–30 ppm of Si in solution.
• Recoveries: Boron recovery approached 99%, while phosphorus (measured as 31P16O) yielded 84% recovery; silver exhibited poor recovery and remains under investigation.
• Container effects: TCS samples transported in stainless steel showed severe metallic contamination (iron up to 180 ppb, chromium 22 ppb, nickel 18 ppb). Glass or plastic vials avoided these artefacts when labs are nearby.
• Detection limits: Method enabled reliable quantitation of key analytes below 1 ppb, meeting stringent PV-industry requirements.

Benefits and Practical Applications

• High-sensitivity screening of PV silicon feedstocks and intermediates for boron, phosphorus and metal contaminants.
• Improved process control during wafer fabrication and TCS production.
• Container selection guidelines to prevent sampling artefacts in ultratrace analyses.

Future Trends and Opportunities

• Integration of in-line ultratrace monitoring systems to enable real-time process adjustments.
• Expansion of collision-cell chemistries to further suppress polyatomic interferences in complex silicon matrices.
• Miniaturized sample-prep modules for on-site analysis at PV fabs and chemical plants.
• Application of high-resolution ICP-MS and novel plasma sources to push detection limits even lower.

Conclusion

This study presents a validated ICP-MS protocol with optimized sample preparation for trace impurity analysis in photovoltaic silicon and trichlorosilane. By addressing silicon matrix removal, interference management and sampling practices, detection of boron, phosphorus and other elements at sub-ppb levels is achievable, supporting quality assurance in solar-grade silicon production.

References

Junichi Takahashi, Noriyuki Yamada, Yasuyuki Shikamori. Ultratrace Analysis of Phosphorus, Boron and Other Impurities in Photovoltaic Silicon and Trichlorosilane by ICP-MS with High Energy Collision Cell. European Winter Conference on Plasma Spectrochemistry, Zaragoza 2011.