We report here an examination of the mass spectrometric fragmentation behavior of molecular ions generated (and excited) by electron ionization (EI) from several asphaltene model compounds simulating both the island and archipelago structural models. This behavior is compared to that of protonated molecules generated from the same compounds by atmospheric pressure chemical ionization (APCI) and excited by collision-activated dissociation (CAD). The fragmentation behavior of the protonated molecules and molecular ions is surprisingly similar. Both types of ions yielded distinct fragmentation patterns for both types of model compounds. Ions derived from the island-type model compounds fragment predominantly by losing their alkyl chains (with either all carbons or all but one), one after another, which allows for the identification of the chain lengths and counting the number of chains. Increasing the length of the alkyl chains reduces the extent of spontaneous fragmentation occurring upon EI, likely because of more efficient cooling of the fragmenting ions via emission of infrared (IR) light made possible by the reduced fragmentation rates of the longer chains. Ions derived from the archipelago model compounds with ethylene bridges connecting two or three aromatic cores (without alkyl side chains) readily undergo cleavages in these bridges. Increasing the length of the alkyl chain between the aromatic cores reduces the extent of fragmentation caused by EI. Similarly, the addition of long external alkyl chains to archipelago model compounds with an ethylene bridging two aromatic cores greatly hinders fragmentation upon EI. When these molecules are protonated and subjected to high-energy CAD, they appear to fragment almost randomly but, nevertheless, indicating some preference for cleavages of the bonds in the chain connecting the aromatic cores. A comparison of these findings to the fragmentation patterns observed for protonated asphaltenes indicates that the asphaltene molecules studied are likely composed of many isomeric and isobaric molecules. Each may contain several aromatic rings and a distribution of mostly aliphatic alkyl chains (and possibly naphthenic rings) ranging in size from 1 to at least 14 carbons, several containing methyl branching at the α carbons. The results do not allow for the unambiguous differentiation between island- and archipelago-type structures, although they are in a better agreement with the island model.

Asphaltenes are a complex mixture of aromatic compounds that are present in petroleum and liquefied coal and have unknown molecular structures. Determination of the molecular structures of asphaltenes is greatly hindered by the extreme complexity of the mixtures. The average molecular weight of coal asphaltenes is known to be roughly half that of petroleum asphaltenes. In this study, tandem mass spectrometry experiments were used to compare the fragmentation patterns of molecular ions of the same mass-to-charge (m/z) value derived from coal and petroleum asphaltenes. The fragmentation pathways of the asphaltene ions were used to determine structural differences between the two types of asphaltenes. This type of analysis is new to the field of asphaltenes research and has not yet been utilized to compare coal and petroleum asphaltenes. Thus, this study was performed to both confirm its viability as well as further corroborate previous data gathered using different techniques. The results reported here demonstrate that the molecules in coal asphaltenes have shorter alkyl chains and larger aromatic cores than their petroleum-derived counterparts with the same molecular weight (MW). Furthermore, they have a higher aromatic-to-alkane carbon ratio, independent of the MW. However, the typical petroleum and coal asphaltenes (when considering the most abundant molecules in each) have a similarly sized aromatic core (∼8 rings), but very different total lengths of the alkyl side chains (∼4 for coal asphaltenes, ∼22 for petroleum asphaltenes).

Mass spectrometric methodology was developed for the determination and manipulation of the primary products of fast pyrolysis of carbohydrates. To determine the true primary pyrolysis products, a very fast heating pyroprobe was coupled to a linear quadrupole ion trap mass spectrometer through a custom-built adaptor. A home-built flow tube that simulates pyrolysis reactor conditions was used to examine the secondary reactions of the primary products. Depending on the experiment, the pyrolysis products were either evaporated and quenched or allowed to react for a period of time. The quenched products were ionized in an atmospheric pressure chemical ionization (APCI) source infused with one of two ionization reagents, chloroform or ammonium hydroxide, to aid in ionization. During APCI in negative ion mode, chloroform produces chloride anions that are known to readily add to carbohydrates with little bias and little to no fragmentation. On the other hand, in positive ion mode APCI, ammonium hydroxide forms ammonium adducts with carbohydrates with little to no fragmentation. The latter method ionizes compounds that are not readily ionized upon negative ion mode APCI, such as furan derivatives. Six model compounds were studied to verify the ability of the ionization methods to ionize known pyrolysis products: glycolaldehyde, hydroxyacetone, furfural, 5-hydroxymethylfurfural, levoglucosan, and cellobiosan. The method was then used to examine fast pyrolysis of cellobiose. The primary fast pyrolysis products were determined to consist of only a handful of compounds that quickly polymerize to form anhydro-oligosaccharides when allowed to react at high temperatures for an extended period of time.

In this work, six petroleum asphaltene samples of different geographical origins were studied using atmospheric pressure chemical ionization (APCI) in positive ion mode in a linear quadrupole ion trap mass spectrometer (LQIT). APCI doped with carbon disulfide reagent was selected as the ionization method as it has been previously demonstrated to generate stable molecular ions with no fragmentation for asphaltene molecules. The mass spectra measured using this approach revealed the apparent molecular weights (MWs) of the molecules in the asphaltene samples. The results show that petroleum asphaltenes from the American continent, Europe and China have similar apparent molecular weight distributions, ranging from 200 up to 1450 Da, with slightly different apparent average MWs ranging from 570 to 700 Da. Further, molecular ions with eight randomly selected mass-to-charge ratios (m/z) ranging from m/z 500 up to m/z 808 were isolated for each asphaltene sample and subjected to collisionally activated dissociation (CAD) at the same collision energy to examine their structures. The CAD mass spectra (MS 2 experiment) provided information on the maximum total number of carbons in the alkyl chains and the smallest possible size of the aromatic cores in the ionized molecules. Additionally, MS 3 experiments were performed to investigate the fragmentation patterns of the fragment ions generated in the MS 2 experiments. The results obtained support the island structural model for these asphaltenes. Moreover, molecules of greater MWs are shown to have more carbons in alkyl chains (ranging from 17 to 41) but the minimum core size is fairly constant.

Saturated and unsaturated, linear, branched, and cyclic hydrocarbons, as well as polyaromatic and heteroaromatic hydrocarbons, were successfully ionized by atmospheric pressure chemical ionization (APCI) using small hydrocarbons as reagents in a linear quadrupole ion trap (LQIT) mass spectrometer. Pentane was proved to be the best reagent among the hydrocarbon reagents studied. This ionization method generated different types of abundant ions (i.e., [M + H] , with little or no fragmentation. The radical cations can be differentiated from the even-electron ions by using dimethyl disulfide, thus facilitating molecular weight (MW) determination. While some steroids and lignin monomer model compounds, such as androsterone and 4-hydroxy-3-methoxybenzaldehyde, also formed abundant M +• and [M+H] + ions, this was not true for all of them. Analysis of two known mixtures as well as a base oil sample demonstrated that each component of the known mixtures could be observed and that a correct MW distribution was obtained for the base oil. The feasibility of using this ionization method on the chromatographic time scale was demonstrated by using high-performance liquid chromatography (HPLC) with hexane as the mobile phase (and APCI reagent) to separate an artificial mixture prior to mass spectrometric analysis.