(145a) Compositional Analysis of Opportunity Materials: Characterization of Heavy Crude Oil and Bitumen by Fourier Transform Ion Cyclotron Resonance Mass Spectrometry

Previous efforts in the detailed characterization of crude oils and bitumen, their associated production deposits and isolated subfractions have been hindered by the oil's inherent complexity. The use of conventional analytical techniques requires extensive fractionation procedures and even then, the complexity of the isolated fractions typically exceeds the capabilities of most commercially available analytical equipment. Furthermore, the lack of suitable standard compounds often precludes identification based on retention time comparisons. For example, the detailed characterization of naphthenic acids requires a pre-isolation step followed by chemical derivitization in order to make the acids gas chromatography (GC) amenable. However, the complexity often leads to coeluting peaks that hinders identification. The addition of a mass spectrometer may aid in the analysis; however, as with GC, coeluting peaks complicate the mass spectrum and often prevent individual component identification. The recent rapid development of FT-ICR mass spectrometry allows for the detailed compositional analysis of complex petroleum samples with little or no sample fractionation.1 Sample preparation is minimal, and simply involves dilution of the sample in a toluene:methanol prior to introduction. The attainable mass-resolving power allows baseline resolution of isobaric species (i.e. C3 vs. SH4, both with a nominal mass of 36 but an exact mass difference of 3.4 mDa) differing by as little as 1 mDa, and the high mass accuracy allows for individual component identification at the level of elemental composition assignment. The ability to assign elemental compositions by class, type, and carbon number to all peaks in the mass spectrum by FT-ICR affords three different levels of compositional information. First, the heteroatom content or class of all peaks is determined. For example, carboxylic acids are members of the class, O2. Second, for determination of the aromaticity range of a specific class, the type or number of rings plus double bonds (double bond equivalents, or DBE) is determined. Third, for every class and type, the carbon number distribution provides a measure of the degree of alkylation for every class and type of species identified. Based on its ability to provide class, type and carbon number distributions for all identified species, FT-ICR mass spectrometry is currently unmatched in its ability to provide detailed compositional information on petroleum derived materials. Here we present the latest results from the ESI FT-ICR MS analysis of heavy crude oils and their associated production deposits as well as Athabasca bitumen samples subjected to common upgrading and fractionation processes.

Experimental

Sample preparation for ESI FT-ICR MS. Crude oil and bitumen samples were each prepared by dissolving ~ 30 mg of material in 30 mL of toluene. A one mL aliquot was then diluted with 0.5 mL of methanol. Ten microliters of either acetic acid (positive ion ESI) or ammonium hydroxide (negative ion ESI) was added to 1.5 mL of the final solution to facilitate protonation or deprotonation in the electrospray ionization process.

Mass spectrometry. Mass analysis was carried out with a homebuilt FT-ICR mass spectrometer equipped with a 22 cm horizontal room-temperature bore 9.4 T magnet (Oxford Instruments America, Inc., Concord MA). Data were collected and processed by a modular ICR data acquisition system (MIDAS). Two hundred co-added time domain data sets were Hanning apodized, followed by a single zero-fill before fast Fourier transformation and magnitude calculation.

Data processing. Mass values for singly charged ions between 225-1000 Da with FT-ICR mass spectral magnitude greater than 5-sigma of baseline noise were converted from IUPAC mass to Kendrick mass as previously described.1 The Kendrick masses were then sorted based on Kendrick mass defect and nominal-z value and imported into an Excel spreadsheet. A molecular formula calculator program limited to molecular formulas consisting of up to 100 12C, 2 13C, 200 1H, 5 14N, 10 16O, 5 32S, and 1 34S assigned elemental compositions. Because members of a homologous series differ only by integer multiples of CH2, assignment of a single member of such a series usually sufficed to identify all higher-mass members of that series. Due to the wealth of information provided by FT-ICR MS analysis of petroleum derived materials, several data reduction methods are useful in quickly determining class, type, and carbon number differences between samples of interest. Kendrick, van Krevelen, and DBE vs. Carbon number plots are commonly used to summarize compositional information provided by FT-ICR MS analysis of petroleum derived materials. Information on all three may be found elsewhere.1-3

Results and Discussion

Detailed characterization of the acidic species in Athabasca bitumen identified the suite of acidic species present. The negative ion ESI FT-ICR MS analysis of the bitumen identified 13 abundant classes of interest. The DBE and carbon number variations in the naphthenic acid class were of special interest. However, a similar level of composition detail is available for all other classes identified in the analysis. Abundant acidic classes include N, NS, NOx, Ox and SOx. The most abundant species are the DBE=3 naphthenic acids that range from 25-40 carbon numbers. A successively decreasing abundance is shown for both increasing carbon numbers and DBE values. The presence of highly abundant SOx species was surprising, and was further investigated through distillation cuts.

The compositional information provided by the FT-ICR MS analysis of the bitumen sparked interest in the behavior of these species in common refining operations. The fate of these species in distillation was investigated through a simulated vacuum distillation performed at NCUT in Devon, Alberta, Canada. Fractions (8) were collected from ~300 ? 500+ °C and subsequently analyzed by FT-ICR MS. The TAN of the collected fractions increased as temperature increased for all but the 500+ °C distillation cut. However, the FT-ICR MS results showed a gradual decrease in the naphthenic acid (O2 class) profile. Interestingly, as distillation temperature increased the levels of multifunctional acidic (NO2, O3-4 and SxOx) species increased. Results from the 500+ °C cut showed a reduction in SO2 and O4 species compared to the 475-500 °C cut, suggesting that these species may be responsible for the decrease in TAN number for the highest boiling fraction. As expected, the carbon number and DBE values progressively increased as the distillation temperature increased. As distillation temperature increases the O2 species shifted from DBE 2-8 (C# 20-35) to DBE 2-12 (C# 25-45) and finally 2-14 (C# 30-50) in the highest temperature cut. All three samples share similar ?hot spots? in the images, indicating that there is one preferred core structure. The SO2 species follow a similar trend with observed shifts from DBE 2-8 (C# 15-35) to DBE 2-12 (C# 20-40) and finally 2-16 (C# 25-50) in the highest temperature cut. Interestingly, the SO2 class has a pair of ?hot spots? in the DBE vs. Carbon Number images, suggesting that there is a change in core structure as distillation temperature increases. Similar details will be presented for all other classes identified in the analysis. Thermal conversion of acidic species in Athabasca bitumen was carried out in a 1L autoclave at 300-400 °C and a residence time of 60 minutes. The reactions were carried out in the presence of slow flow of nitrogen to remove water formed in the condensation reactions of naphthenic acids. Besides small gaseous products and light end products collected in the condenser, major products (heavy ends) were collected from the reactor and submitted for different analyses. Under the reaction conditions employed the amount of gaseous products was negligible. At low temperatures, including 300 °C and 325 °C, the light ends were low in abundance, and only at higher temperatures of 375 and 400 °C were the light ends collected in the condenser at 3 and 6 wt% respectively. The analysis of these products showed a large amount of low DBE, low carbon number naphthenic acids. The mass spectra showed that there is little change in the species identified as the temperature increased from 300 °C to 375 °C. However, there was a significant decrease in oxygen content at 400 °C. The FT-ICR MS analysis of the thermally treated samples showed very little difference in the naphthenic acid (O2 class) distributions for the four samples, although the TAN of the heaviest fraction was substantially lower than for the other 3 lower temperatures. However, the FT-ICR MS analysis did show substantial differences in the O4 class for the different reaction temperatures. As the temperature increases, the number and relative abundance of the O4 species rapidly declines. The depletion of these species could account for the TAN reduction. Other studies presented include the compositional analysis of a heavy crude oil and its associated deposits and compositional variations in pressure drop and solvent drop asphaltenes. Examples and the implications of self aggregation will be presented for both bitumen and heavy crude oil.

1. Marshall, Alan G.; Rodgers, Ryan P., Accounts of Chemical Research (2004), 37(1), 53-59. 2. Wu, Zhigang; Rodgers, Ryan P.; Marshall, Alan G., Analytical Chemistry (2004), 76(9), 2511-2516. 3. Kim, Sunghwan; Kramer, Robert W.; Hatcher, Patrick G., Analytical Chemistry (2003), 75(20), 5336-5344.