Characterization of deasphalted heavy fuel oil using APPI (+) FT-ICR mass spectrometry and NMR spectroscopy

Abstract

Asphaltenes are n-alkane insoluble compounds found in crude oils and heavy fuels (high and non-boiling petroleum fractions). Asphaltene molecular structure has not been fully elucidated, and their presence in fuels is a source of concern. They reduce combustion efficiency and are responsible for particulate matter emissions. Removing the asphaltenes, or deasphalting, is a way of upgrading the fuel to improve its quality. This study reports the removal of asphaltenes from heavy fuel oil (HFO) using a solvent extraction method, and the detailed molecular characterization of the deasphalted oil (DAO) using positive ion atmospheric pressure photo ionization Fourier transform-ion cyclotron resonance mass spectrometry (APPI (+) FT-ICR/MS) and 1H and 13C Nuclear magnetic resonance (NMR) spectroscopy. Approximately 8.2 mass % of asphaltenes were removed from HFO using n-heptane as solvent. This resulted in significant improvements in the HFO’s physical properties. The resulting DAO was five times less viscous and contained significantly less heavy metals (e.g., Ni and V). There was also a slight reduction in the sulfur content from 3.3 to 3.1 mass %. 52,753 and 46,315 ions with a mass to charge ratio (m/z) ranging from 154 to 1200 were detected in HFO and DAO samples, respectively, using APPI FT-ICR/MS. Amongst them, 6729 (HFO) and 6030 (DAO) ions were resolved into their underlying elemental compositions (C, H, O, N and S) and a unique chemical formula was assigned to each mass. The resolved masses were then further classified based on their molecular class and were analyzed as a function of double bond equivalent (DBE) vs carbon number. 1H and 13C NMR analyses of HFO and DAO were performed and the results indicate the total aromatic groups in HFO (1H 7.7 mol %, 13C 37.6 mol %) are more compared to DAO (1H 4.7 mol %, 13C 32.6 mol %). The average molecular parameters (AMPs) of HFO and DAO were also calculated from the 1H and 13C NMR spectra and compared. A surrogate molecule that visualizes the average molecular structure of the entire fuel was developed for both HFO and DAO using the data from the above analytical techniques. Understanding the molecular chemistry of these fuels provides valuable data to develop better desulfurization techniques for these sulfur laden fuels and help predict fuel properties using structure-property relationships.