Analysis of Metallic Impurities in Specialty Semiconductor Gases Using Gas Exchange Device (GED)-ICP-MS

Significance of the Topic

Analyzing metallic impurities and nanoparticles in specialty semiconductor gases is critical for ensuring the yield and performance of integrated circuits. Ultratrace metals and particles in etching and cleaning gases such as HF and Cl2 can compromise device quality. Conventional ICP-MS cannot directly handle corrosive gases due to plasma extinguishing. The Gas Exchange Device (GED) combined with an aerosol standard generator (MSAG) and ICP-QQQ provides a robust solution to quantify both dissolved gaseous metals and particulate contaminants at ultralow levels.

Objectives and Study Overview

This study aimed to evaluate GED-MSAG-ICP-QQQ for measuring total metal content and nanoparticles in semiconductor-grade HF and Cl2 gases. Key goals included:

• Quantitative detection of ultratrace gaseous metal impurities and particles
• Development of an automated workflow integrating GED, MSAG, and Agilent 8900 ICP-QQQ
• Comparison of direct gas analysis and single nanoparticle (spICP-MS) approaches

Methodology and Instrumentation

The experimental setup comprised an IAS GED_LAB membrane cell and an IAS Metal Standard Aerosol Generator (MSAG) coupled to an Agilent 8900 Triple Quadrupole ICP-MS. Key methodological elements:

• GED principle: sample gas flows inside a membrane where >99.99% is exchanged by Ar sweep gas; particles are retained and carried to the plasma
• MSAG dual-syringe system delivers blank and multi-element standard to a nebulizer at 0.3 L/min Ar, enabling standard addition calibration without plasma wetting
• ICP-QQQ operated in MS/MS mode with NH3/He and O2 reaction gases to remove interferences; single particle mode used 1 ms dwell for spICP-MS
• Direct gas analysis bypassed the membrane, introducing HF or Cl2 at low flow rates into the plasma for spectrum acquisition (1 s integration)

Main Results and Discussion

HF gas analysis:

• spICP-MS detected Fe, Ni, Cr, Mn, Na, K, Ca, Sn, W particles; filtration reduced particle counts
• Direct spectrum mode revealed elevated gaseous B, P, Cu, Ge, W; transient spikes highlighted the need to equilibrate gas cylinders
• Quantitative data: gaseous impurities in µg/kg, particulate concentrations in ng/kg

Cl2 gas analysis:

• spICP-MS showed Ca, Cr, Mn, Fe, Ni, Cu particles; P, Ge, As, Sn, Sb produced continuous background from volatile chlorides
• Direct analysis found high P, Fe, Cu, Ge, As, Sn, Sb levels; particulate levels were orders of magnitude lower

Benefits and Practical Applications

The GED-MSAG-ICP-QQQ method enables:

• Simultaneous quantitation of dissolved metal species and nanoparticles in corrosive semiconductor gases
• Automated standard addition calibration with low matrix effects
• Interference-free, high-sensitivity detection using MS/MS and reaction gases

Future Trends and Potential Applications

Potential developments include:

• Extension to other specialty gases (NH3, CO2, HCl, NF3, SiH2Cl2, SF6)
• Improved membrane materials for enhanced chemical compatibility and lower particle loss
• Integration of real-time inline monitoring for process control in semiconductor fabs
• Advancements in spICP-MS for detection of smaller nanoparticles and other emerging contaminants

Conclusion

The combined GED, MSAG, and Agilent 8900 ICP-QQQ approach delivers a comprehensive, automated solution for ultratrace analysis of metallic impurities and nanoparticles in corrosive semiconductor gases. This methodology meets the stringent requirements of modern IC manufacturing by offering high sensitivity, low background, and robust interference removal.

Instrumentation Used

• IAS GED_LAB electropolished stainless steel membrane cell
• IAS Metal Standard Aerosol Generator (MSAG) dual-syringe system
• Agilent 8900 Triple Quadrupole ICP-MS with HF-tolerant PFA inert torch, Pt injector, m-lens, Pt-tipped Ni skimmer cone
• Cleanroom-grade reagents, gas filters, and regulators

References

• Measuring Inorganic Impurities in Semiconductor Manufacturing, Agilent publication, 5991-9495EN
• Nishiguchi K., et al., Real-time multielement monitoring of airborne particulate matter using ICP-MS with gas converter, J. Anal. At. Spectrom., 2008, 23, 1125–1129
• Kovacs R., et al., Development of direct atmospheric sampling for LA-ICP-MS, J. Anal. At. Spectrom., 2010, 25, 142–147
• Suzuki Y., et al., Real-time monitoring of Pb in a single airborne nanoparticle, J. Anal. At. Spectrom., 2010, 25, 947–949
• von der Weiden S.-L., et al., Particle loss calculator for aerosol inlet systems, Atoms. Meas. Tech. Discuss., 2009, 2, 1099–1141
• Shimamura Y., Hsu D., Yamanaka M., Multielement Nanoparticle Analysis of Chemicals Using spICP-QQQ, Agilent publication, 5994-0987EN
• Hsu D., Shimamura Y., Liao B., et al., Analysis of Nanoparticles in Organic Reagents by 8900 ICP-QQQ in spICP-MS Mode, Agilent publication, 5994-1306EN