Analysis of photovoltaic grade silicon using triple quadrupole inductively coupled plasma mass spectrometry (ICP-MS)

Importance of the Topic

Photovoltaic (PV) technology is a cornerstone of renewable energy strategies due to its zero‐carbon emissions and use of abundant silicon. High‐purity, photovoltaic‐grade silicon must be characterized for bulk and trace element impurities—particularly dopants like phosphorus and boron—to ensure maximum cell performance, long‐term stability, and cost effectiveness. Reliable, sensitive, and high‐throughput analytical methods are essential for both research and production quality control of PV materials.

Study Objectives

This application note describes the development and validation of a robust analytical workflow using triple quadrupole inductively coupled plasma mass spectrometry (ICP-MS) to determine 42 elements in photovoltaic‐grade silicon. Key goals included:

• Achieving low detection limits and high linearity across a wide concentration range (bulk to ultra-trace).
• Demonstrating effective interference removal for critical analytes such as phosphorus and selenium in high‐silicon matrices.
• Proving method robustness for high‐throughput analysis up to 5,000 mg·L⁻¹ silicon.

Methodology and Instrumentation

Sample Preparation:

• Dissolution of silicon wafer in 50 % HF and 68 % HNO₃ in PFA vessels, followed by dilution to 100 mL.
• Manual 20-fold dilution of the digest into 2 % HNO₃ spiked with Y, Rh, and Ir internal standards.

Instrumentation Used:

• Thermo Scientific™ iCAP™ MTX triple quadrupole ICP-MS with argon gas dilution (AGD).
• Thermo Scientific™ iSC-65 Autosampler with Step Ahead functionality.
• PFA concentric nebulizer (400 µL·min⁻¹), PFA cyclonic spray chamber at 2.7 °C, PLUS torch with sapphire injector.
• Collision/reaction cell operated in He KED mode for general elements and O₂ reaction mode (TQ-O₂) for highly interfered isotopes (P, S, As, Se).
• Thermo Scientific™ Qtegra™ ISDS software with Reaction Finder and autotune routines.

Analytical Conditions:

• Plasma power: 1,550 W; AGD flow: 0.50 L·min⁻¹; nebulizer gas: 0.45 L·min⁻¹.
• Peristaltic pump speed: 25 rpm; Step Ahead uptake: 50 s; wash: 25 s; total sample time: 3 min 33 s.

Key Results and Discussion

Sensitivity and Linearity:

• All 42 elements achieved R² > 0.999 over calibration ranges up to 5 × 10⁶ µg·L⁻¹ Si.
• Instrument detection limits (IDLs) were sub-µg·L⁻¹ for most analytes; TQ-O₂ mode improved LODs for P and Se by >10-fold compared to He KED.

Interference Removal and Accuracy:

• Mass shift with O₂ in triple quadrupole mode effectively eliminated isobaric and polyatomic interferences (e.g., ³¹Si on ³¹P).
• Spike recoveries for 10 µg·L⁻¹ additions in silicon solutions (10–5 000 mg·L⁻¹) ranged 85–108 % for all elements.
• Internal standard recoveries remained within 87–106 % across varying Si concentrations, demonstrating robust suppression control.

Long-Term Performance:

• A 9-hour batch of 197 analyses (120 matrix samples, 15 standards, 24 QC checks) yielded continuing calibration verification recoveries of 88–124 %.
• Stable internal standard signals throughout the run confirmed system robustness and minimal matrix‐related drift.

Benefits and Practical Applications

• Simultaneous bulk and trace element determination in photovoltaic‐grade silicon with minimal sample preparation.
• High throughput enabled by Step Ahead autosampling and fast analysis times (<4 min per sample for 42 elements).
• Superior interference removal for critical dopants supports accurate quality control and R&D in PV manufacturing.
• Argon gas dilution (AGD) permits direct analysis of high‐silicon matrices, reducing the need for excessive sample dilution.

Future Trends and Opportunities

• Integration of discrete sampling valves to further shorten analysis cycles.
• Expansion of the method to other PV materials (e.g., thin‐film semiconductors) and emerging contaminants.
• Automation of sample digestion and on-line dilution to streamline high-throughput workflows.
• Deployment of advanced data analytics and machine learning for predictive QC and trend monitoring.

Conclusion

The triple quadrupole ICP-MS approach with combined He KED and O₂ reaction modes delivers a versatile, sensitive, and interference-free method for broad‐range elemental analysis in photovoltaic‐grade silicon. The workflow demonstrates excellent linearity, low detection limits, and robust performance in high-silicon matrices, enabling reliable quality control and process optimization in solar cell production.

Reference

• Thermo Fisher Scientific. Product Spotlight 44485: Thermo Scientific iCAP Qnova Series ICP-MS PLUS Torch for Improved ICP-MS Analysis of Challenging Samples.