Data-Driven Filter for Detector Oscillation Artifacts in Time-of-Flight Mass Spectrometry.

Area of Science:

• Analytical Chemistry
• Mass Spectrometry

Background:

• Time-of-flight mass spectrometry (TOF-MS) is crucial for analyzing complex mixtures, detecting thousands of compounds per sample.
• High-intensity ions can cause detector ringing, creating artifacts that mimic real signals and complicate data analysis.

Purpose of the Study:

• To introduce a data-driven filtering approach to identify and remove detector ringing artifacts in TOF-MS data.
• To establish a universal criterion for artifact detection independent of experimental conditions.

Main Methods:

• Developed a filtering workflow exploiting the time-domain origin of detector ringing.
• Utilized the relationship between ion flight time and $\sqrt{m/z}$ for artifact identification.
• Applied the filter to GC-, LC-, and ion mobility-TOF-MS datasets, and mass spectral libraries (MassBank, NIST DART-MS Forensics).

Main Results:

• Successfully identified and removed up to 5.3% of features caused by detector ringing across various platforms.
• Discovered that artifact spacings ($\Delta \sqrt{m/z}$) are instrument-specific and polarity-independent.
• Identified over 1000 mass spectra with ringing artifacts in spectral libraries, with clustering revealing platform-specific groupings.

Conclusions:

• The developed workflow effectively reduces false positives in TOF-MS analyses.
• Improved feature prioritization and compound identification by removing detector-derived artifacts.
• Demonstrated the potential for artifact spacing analysis in data forensics and instrument traceability.