Redefining Compound Characterization Via GC-MS with Cold EI

Reliable gas chromatography–mass spectrometry (GC-MS) compound characterization is typically constrained by volatility limitations, thermal decomposition at the GC injector and/or column, ion-source-induced degradation, and a weak or absent signal of molecular ions that are inherent to conventional GC–MS. Standard electron ionization (EI) often fragments analytes excessively, suppressing molecular ions and forcing reliance on orthogonal techniques (including expensive instrumentation), existing literature, data interpretation strategies, and reference materials. Furthermore, low-volatility and thermally sensitive species remain underrepresented in the field due to extended elution times and instability at conventional injector and column temperatures. Furthermore, compounds with OH and NH structures tend to degrade on the metallic surface of standard EI ion sources.

To address these bottlenecks, we employed a supersonic molecular beam interface enabling ionization of vibrationally cold molecules in a fly-through ion source configuration, commonly referred to as Cold EI. The approach was evaluated for its ability to preserve molecular integrity, accelerate chromatographic elution, and recover labile and polar compounds that elude reliable treatment via standard EI.

Various compounds such as hydrocarbons, pharmaceutical agents, large polyaromatic structures, and plasticizers were analyzed using both Cold and Standard EI configurations. Our Cold EI technique had consistently produced clearer and more reliable molecular ions, lowered elution temperatures by up to 100 °C, and reduced retention times by up to nearly an order of magnitude. These combined gains mitigate the need for orthogonal techniques and enable a fuller characterization from a single GC–MS experiment.

Rather than viewing GC–MS with Cold EI as an incremental upgrade, our results position it as a general-purpose ion stabilization platform that expands both the chemical space accessible to GC–MS and the confidence level of the obtained data.

Several comparative examples will be given as demonstrations