Measuring destruction efficiency of greenhouse gases released by semiconductor fabrication tools

FTIR gas analysis — Measuring destruction efficiency of greenhouse gases released by semiconductor fabrication tools (Application note AN54654)

Significance of the topic

Modern semiconductor manufacturing uses fluorinated gases in plasma processes that are potent greenhouse gases (GHGs). Accurate, repeatable quantification of how effectively abatement (destruction + dilution) systems remove these species is required for regulatory reporting (EPA, regional rules) and for responsible environmental management. A compact, accurate measurement workflow that reduces instrument complexity while providing multi-species coverage and sensitivity is therefore valuable for fabs, environmental compliance teams and process engineers.

Objectives and overview of the study

• Develop a simpler integrated measurement method to determine abatement device destruction efficiencies for fluorinated GHGs used or produced in semiconductor etch/CVD tools.
• Replace a bulky configuration (two FTIRs + mass spectrometer + krypton tracer) with a single FTIR coupled to an automated sample console while maintaining accuracy across large pre- vs post-abatement concentration ranges.
• Demonstrate capability to quantify multiple reactant and by-product species, and to identify unknown plasma by-products.

Methodology

• Key principle: determine both the dilution factor (DF = total volumetric flow out / total volumetric flow in) and the actual pre- and post-abatement concentrations of each target GHG, then compute destruction efficiency (DRE) from these values.
• Tracer-dilution approach: a 100% tracer gas (infrared-active) is injected and measured to derive volumetric flow rates without requiring a krypton tracer or a mass spectrometer. Separate tracer injections are made upstream (to determine TVFin) and downstream (to determine TVFout).
• Single-FTIR concentration range handling: the Automated Sample Console (ASC-10) is used to dilute high pre-abatement concentrations (typical 1:10 dilution) so the single FTIR can use the same non-saturating calibration bands for both pre- and post-abatement measurements, avoiding need for two FTIRs with different pathlengths.
• Sampling geometry: curved stainless-steel probes are positioned to promote turbulent mixing when tracer is injected and to avoid stratification, improving representativeness of sampled gas.
• Data processing: spectral deconvolution and subtraction (MAX-Acquisition Automation software) are used to quantify reactants, products and minor components. Unknown bands are investigated against databases (NIST/EPA) and by peak-matching tools.
• Calculation: DRE is computed from measured pre-abatement concentration Ci, post-abatement concentration Co, and measured dilution factor DF (equation equivalent to DRE = 1 - (Co / (Ci * DF)) or the form provided in the application note).

Used instrumentation

• Thermo Scientific MAX-iR FTIR Gas Analyzer with DTGS detector (spectral range 500–5000 cm-1; no liquid nitrogen required).
• Thermo Scientific Automated Sample Console (ASC-10) with integrated mass flow control for tracer and dilution management.
• MAX-Acquisition Automation software for spectral automation, component subtraction and unknown-feature analysis; database matching using NIST/EPA gas-phase library for identification of unknowns.

Main results and discussion

• System simplification: The combined MAX-iR + ASC-10 approach reduced the measurement system from two FTIRs plus a mass spectrometer to one FTIR + ASC-10 while delivering comparable performance for abatement efficiency assessment.
• Example measurement (C4F8): Pre-abatement measurement (after applying the 1:10 sampled dilution and scaling) yielded Ci ≈ 854.5 ppm. Post-abatement concentration measured directly was Co ≈ 7.92 ppm. Using CF4 as the tracer the dilution factor DF was 9.75. These values produced a calculated destruction efficiency of ~91% for that abatement unit under test.
• Destruction efficiency range: measured DREs varied substantially across species and conditions in tests — from >98% for some species (e.g., C4F6) down to as low as ~17% for CF4 — illustrating the importance of species-specific assessment.
• By-product identification: During plasma-on operation (etch mix C4F6/O2/Ar) the system quantified reactants and produced species including CO, CO2, HF (weak features 3800–4000 cm-1), COF2, CF4, SiF4 and C2F6. Two residual unidentified features (near 1170 and 1345 cm-1) were matched via database search and identified as tetrafluoroethylene (C2F4), a relevant etch product. This demonstrates the system’s capability to detect and identify process by-products beyond the primary GHGs.

Practical benefits and applications

• Regulatory compliance: Provides a faster, less cumbersome method to produce facility-dependent GHG emission audits required by regulatory authorities.
• Operational efficiency: Single-instrument workflow reduces lab footprint, instrument complexity and operational overhead while speeding measurement time to ~1–1.5 hours per abatement device.
• Process insight: Ability to monitor both feed gas composition and plasma-generated by-products supports process troubleshooting and optimization in real time.
• Flexibility: Multi-species IR detection plus spectral analysis tools enables quantification of both known fluorocarbon species and discovery/identification of unexpected products.

Future trends and possibilities

• Expanded spectral libraries and automated matching will further improve identification of low-level and previously unreported plasma by-products.
• Integration with fab process control systems could enable on-line or semi-continuous abatement performance monitoring and automated alerts when DRE falls below thresholds.
• Higher-sensitivity detectors or extended pathlength sampling accessories may expand dynamic range and lower limits of detection for difficult species such as CF4.
• Standardization of tracer-based single-FTIR measurement protocols could enable cross-facility comparisons and streamlined regulatory reporting.

Conclusion

The application note demonstrates a practical, streamlined method to measure destruction efficiencies of fluorinated greenhouse gases from semiconductor abatement devices using a single FTIR (MAX-iR) coupled to an automated sample console (ASC-10). By combining tracer-based flow/dilution measurements with controlled sampling dilution, the approach covers large dynamic concentration ranges, simplifies instrumentation, and preserves analytical performance. The system also provides useful process-level information about plasma by-products and can identify unknown species, making it valuable for regulatory compliance, process monitoring and emission-control optimization.

References

• Thermo Fisher Scientific, Application Note AN54654, FTIR gas analysis — Measuring destruction efficiency of greenhouse gases released by semiconductor fabrication tools. 2023.