Molecular characterization of petroleum mixtures using multiple ionization modes and GC×GC-HRTOFMS

Introduction: Comprehensive two-dimensional gas chromatography (GC×GC) has become a method of choice for complex mixture characterization, especially in the petroleum industry. Indeed, the high resolving power of the 2D separations offers structured separation allowing pattern recognition and group type classification of sample composition. For years, the technique was relying on electron ionization (EI) fragmentogram and chromatographic-based identification due to practical limitations on the detector side, mainly the high acquisition frequency required. The development of high-speed high-resolution time-of-flight mass spectrometers (TOFMS) offers opportunities to go deeper in the sample characterization. To obtain the most of the HRMS dimension, there is a growing interest to combine EI with softer ionization techniques, which preserve the molecular ion.

Method: In this study, different base oil samples and standard mixtures were analyzed by GC×GC-HRTOMS. Three different soft-ionization techniques including photo ionization (PI), chemical ionization (CI), and field ionization (FI) were compared against EI to elucidate their relative capabilities to reveal different base oil hydrocarbon classes. Deeper investigations were also conducted on the PI fragmentation process for different chemical families. All the experiments were performed on a single system, a JEOL AccuTOF GCv 4G with modular ionization technologies. A low-polar (ZB-XLB-HT Inferno, 15.0 m, 0.25 mm ID, 0.1 μm, Phenomenex ) and polar (ZB-50HT, 2.0 m, 0.1 mm ID, 0.1 μm, Phenomenex) columns were used for base oil analysis for first and second dimension separation respectively.

Preliminary Data: Compared to EI (70 eV), capabilities and limitations of PI were tested using an authentic mixture of compounds of several chemical classes. Ionization energy exhibited by PI, equivalent to 10.8 eV, resulted in significant retention of molecular ion information; [M]+• for alkanes, ketones, FAMEs, aromatics, [M−H]+• for chloroalkanes, and [M−H2O]+• for alcohols. In addition, considering the potential of PI for hydrocarbons, base oils, complex mixtures of saturated and unsaturated hydrocarbons blended for finished lubricant formulations, were extensively evaluated. PI retained significant molecular ion (M+‧) information for a large number of isomeric species including branched-alkanes and saturated mono-cyclic hydrocarbons along with unique fragmentation patterns. However, for bi-/poly cyclic naphthenic and aromatic compounds, EI played upper hand by retaining molecular as well as fragment ions to identify the species, whereas PI exhibited mainly molecular ion signals. CI revealed selectivity towards different base oil groups, particularly for steranes, sulfur-containing thiophenes, and esters; yielding protonated molecular ions (M+H) + for unsaturated and hydride abstracted ions (M-H+) for saturated hydrocarbons. FI, as expected, generated intact molecular ions (M+‧) irrespective to the base oil chemical classes. It allowed elemental composition by TOFMS with a mass resolving power up to 8,000 (FWHM) and a mass accuracy of 1 mDa, leading to the calculation of heteroatomic content, and carbon number of the compounds. The qualitative and quantitative results presented herein offer a unique perspective into the detailed comparison of different ionization techniques corresponding to several chemical classes.

Novel aspect: The GC×GC field is ongoing a paradigm shift in which the MS dimension is even further contributing than before.