Analysis of Ultratrace Impurities in High Silicon Matrix Samples by ICP-QQQ

Significance of the Topic

Ultratrace detection of metallic impurities in high-purity silicon matrices is critical for semiconductor manufacturing quality control. As device dimensions shrink to sub-10 nm and production volumes grow, the stringent purity requirements demand analytical techniques capable of ppt-level sensitivity and robust interference suppression. Reliable quantification of 38 trace elements ensures material integrity and supports process optimization in FABs.

Objectives and Study Overview

This study evaluates the performance of the Agilent 8900 ICP-QQQ equipped with the optional m-lens for the determination of 38 elements in high-silicon matrix samples. Two sample matrices (10 ppm and 100 ppm Si) were spiked at 50 ppt to assess sensitivity, accuracy, and stability over a continuous one-hour run. The goals include minimizing background equivalent concentrations, demonstrating drift control, and verifying spectral interference removal capabilities.

Methodology and Instrumentation

Sample preparation involved digestion of 9N purity poly-silicon in a 1:1 (w/w) HF/HNO3 acid mix to produce a 1000 ppm Si stock solution, which was diluted to 10 ppm and 100 ppm matrices. Calibration used aqueous standards matching the acid matrix and multi-element spike solutions. A 50 ppt multi-element spike was applied to each matrix level. MassHunter software controlled a multi-tune MS/MS acquisition, automatically switching cell gases (He, H2, NH3, O2) to optimize interference removal. Key operating parameters included 1550 W RF power, 0.8 L/min carrier gas, 0.4 L/min makeup gas, and 4 mL/min He cell flow.

Instrumentation Used

• Agilent 8900 ICP-QQQ with optional m-lens ion lens
• Inert sample introduction: 200 µL/min MicroFlow PFA nebulizer, PFA spray chamber, endcap, connector, and demountable torch with 2.5 mm sapphire injector
• Pt-tipped sampling cone and Ni-based skimmer cone for m-lens
• Inert Kit components for HF resistance
• ORS4 reaction cell gases: He, H2, NH3, O2

Main Results and Discussion

All 38 analytes in both 10 ppm and 100 ppm Si matrices exhibited BECs below 50 ng/L, with most under 5 ng/L. Spike recoveries ranged from 90 to 110% (except Fe at 112% in the 10 ppm matrix), demonstrating excellent accuracy. Precision was maintained with RSDs below 6% over eleven replicates during a one-hour run, indicating minimal signal drift and high plasma robustness. Automated cell gas switching effectively eliminated Si-based and other polyatomic interferences on critical isotopes such as Ti, Fe, Ni, Cu, and Ge.

Benefits and Practical Applications

• ppt-level detection of 38 trace elements in silicon matrices
• High matrix tolerance and minimal signal suppression under robust plasma conditions
• Automated interference removal via MS/MS with multi-tune gas switching
• Suitable for quality control of high-purity silicon and process chemicals in semiconductor fabs
• Supports advanced device nodes and stringent purity specifications

Future Trends and Potential Applications

As semiconductor devices scale to 5 nm and below, analytical demands will intensify toward even lower detection limits and faster turnaround times. Future developments may include integration of ICP-QQQ with automation and AI-driven data analysis for real-time process monitoring, expansion to emerging materials such as 2D semiconductors, and coupling with advanced sample handling techniques like vapor phase decomposition for wafer analysis.

Conclusion

The Agilent 8900 ICP-QQQ with m-lens demonstrates outstanding sensitivity, stability, and interference removal for ultratrace impurity analysis in high-silicon matrices. The method meets the evolving purity requirements of modern semiconductor manufacturing, offering reliable ppt-level quantification and robust performance over extended analytical runs.

References

• Agilent Technologies. Applications of ICP-MS: Measuring Inorganic Impurities in Semiconductor Manufacturing. Agilent publication 5991-9495EN.
• Bolea-Fernandez E, Balcaen L, Resano M, Vanhaecke F. Overcoming spectral overlap via inductively coupled plasma-tandem mass spectrometry (ICP-MS/MS): A tutorial review. J. Anal. At. Spectrom. 2017;32:1660–1679.
• McCurdy E, Woods G, Sugiyama N. Method Development with ICP-MS/MS: Tools and Techniques to Ensure Accurate Results in Reaction Mode. Spectroscopy. 2019;34(9):20–27.