Sequential Extraction of Petroleum Asphaltenes with Magnesium

Aug 31, 2015 - Department of Chemistry and Biochemistry, Baylor University, Waco, Texas 76798, United States. Energy ... Furthermore, both UV and MS d...
0 downloads 5 Views 465KB Size
Subscriber access provided by The University Library | University of Sheffield

Article

Sequential Extraction of Petroleum Asphaltenes with Magnesium Oxide: A Method to Reduce Complexity and Improve Heteroatom Identity Aaron K Zimmer, Christopher Becker, and C. Kevin Chambliss Energy Fuels, Just Accepted Manuscript • DOI: 10.1021/acs.energyfuels.5b01076 • Publication Date (Web): 31 Aug 2015 Downloaded from http://pubs.acs.org on September 7, 2015

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

Energy & Fuels is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 25

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

Sequential Extraction of Petroleum Asphaltenes with Magnesium Oxide: A Method to Reduce Complexity and Improve Heteroatom Identity Aaron K. Zimmer, Christopher Becker, C. Kevin Chambliss* Department of Chemistry and Biochemistry, Baylor University, Waco, TX, 76798, USA

AUTHOR EMAIL ADDRESSES: [email protected]; [email protected]; [email protected] RECEIVED DATE (to be automatically inserted after your manuscript is accepted if required according to the journal that you are submitting your paper to) CORRESPONDING AUTHOR FOOTNOTE: *One Bear Place #97348, Waco, TX 76798-7348 (Phone) 254-710-6849; (Fax) 254-710-4272; (Email) [email protected] ABSTRACT Sequential extraction of an asphaltene sample with magnesium oxide nanoparticles provides an approach to selectively remove molecules from a complex fraction. This method of extraction relies, in part, on adsorption preferences for differing heteroatoms to leave weakly-adsorbing sample constituents in a toluene solution. The extracted sample exhibits reduced complexity enabling more reliable identification of the remaining molecules in solution. Mass spectrometry (MS) data indicate a 1 ACS Paragon Plus Environment

Energy & Fuels

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 25

general bias for preferential removal of higher molecular weight species. This is supported by a shift in the average m/z ratio of the asphaltene distribution to lower m/z, as well as a decrease in the intensity observed for higher m/z ion signals, for higher extraction numbers. UV-Visible absorption data corroborate MS data to provide an appreciable visual means to quantify sample uptake after several sequential extraction steps with MgO. Furthermore, both UV and MS data indicate a point of diminishing returns, after which subsequent extraction with MgO nanoparticles results in limited adsorption of remaining asphaltene constituents. The remaining asphaltene constituents can then be treated with NiO nanoparticles in order to identify molecules containing pyridyl functional groups. Implementation of a more exhaustive MgO extraction, prior to treatment with NiO, resulted in an improved method for profiling pyridyl-containing structures in a complex asphaltene mixture relative to previous work.

KEYWORDS: magnesium oxide, MgO, nickel oxide, NiO, adsorption, nanoparticle, metal oxide, asphaltene, heteroatom distribution, pyridine, pyridyl carboxylic-acid, carboxyl, functional groups, crude oil, selective adsorption, complex mixture, heavy oil

1. INTRODUCTION Demand for petroleum fuels has increased at such a rate that oil industries have resorted to refining heavier, more problematic fractions of crude oils [1]. The heaviest and most polar fraction of crude oil contains a class of compounds known as asphaltenes. This class of compounds, defined largely by solubility characteristics [2], creates persistent problems in petroleum processing, upgrading, and transportation [3]. Bulk properties of asphaltenes have been known for some time [2, 4-23]. Broadly defined, asphaltenes may be characterized as a chemical class with a relatively constant hydrogen-to-carbon ratio (1.15  0.5%) and variable heteroatom composition [2, 4-8]. Mass 2 ACS Paragon Plus Environment

Page 3 of 25

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

spectrometry data supports an average molecular weight of 750  200 Daltons [24-27]. While it is generally known that heteroatoms persist in asphaltenes as common functional groups, more intimate knowledge of functional group distributions in a given sample could add significantly to a fundamental understanding of asphaltenes on a molecular level. Metal oxide nanoparticles are known to adsorb asphaltenes with high capacity and affinity [2830]. The propensity of asphaltenes to adsorb onto metal oxides has been correlated with acid-base properties of metal oxide surfaces [31] and to the presence of heteroatom functional groups in asphaltenes [28, 32]. In a more recent publication [33], model compounds were used to demonstrate that functional groups expected to exist in asphaltenes display varying affinities for a given metal oxide. Carboxyl, pyridyl, and phenolic groups were generally shown to be the most likely contributors to high adsorption affinity. However, the pyridyl group was found to have divergent adsorption behavior on magnesium oxide (MgO) and nickel oxide (NiO) (i.e., essentially no affinity for MgO but high affinity for NiO). In the same paper, this unique selectivity difference was exploited, in combination with mass spectrometry, to demonstrate proof-of-concept for a novel approach to profile pyridyl-containing structures in asphaltenes. The present study was designed to evaluate the effect of successive MgO extractions on the composition of dissolved asphaltene. An interesting trend favoring preferential adsorption of higher molecular weight species from toluene solution was observed. Additionally, successive extractions resulted in iterative reductions of mass spectral complexity within a narrow mass range, demonstrating that a complementary separation based on adsorption could leverage future efforts to elucidate asphaltene structures via mass spectrometry. Lastly, implementing a more exhaustive extraction of asphaltenes with MgO oxide, prior to treatment with NiO, substantially improved the ability to identify

3 ACS Paragon Plus Environment

Energy & Fuels

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 4 of 25

pyridyl-containing structures in a complex asphaltene mixture, relative to an earlier proof-of-concept demonstration [33].

2. EXPERIMENTAL SECTION 2.1 Nanomaterials, Solvent, and Asphaltene Sample. Nanoparticles: NiO (100 nm average particle size), and MgO (100 nm average particle size) were obtained from Nanostructured & Amorphous Material, Inc., TX, USA. Ultra Resi- Analyzed grade toluene was obtained from J.T. Baker. The asphaltene sample was from a San Andro region crude oil. Details pertaining to the obtainment and storage of this asphaltene sample are reported elsewhere [33]. Elemental (C, H, N, S) analyses of the sample were performed by Atlantic Microlab, Inc. (Norcross, GA). Oxygen analysis was performed by Galbraith Laboratories (Knoxville, TN). Elemental composition, reflecting the mass percentage determined for each element, was as follows: C – 80.1%, H – 8.0%, N – 1.2%, O – 1.5%, S – 7.8%. 2.2a Asphaltene adsorption protocol for ultraviolet visible (UV-Vis) spectroscopy. A sample of asphaltene (0.010 g) was dissolved in 10 mL of toluene. An aliquot of this solution was taken and diluted 10 fold with toluene for UV-visible analysis. Afterwards, a 9 mL aliquot of the asphaltene was treated with 0.090 g of MgO. This mixture was shaken for 30 minutes on a Barnstead/Lab-line MaxQ 4000 shaker and then centrifuged for 40 seconds at 3000 rpm. Previous work with model compounds demonstrated that adsorption equilibrium was achieved within 2-3 minutes [33], and 30 minutes is assumed to be a sufficient amount of time for equilibrium to be achieved in the present study. An aliquot of the centrifuged solution was diluted 10-fold with toluene prior to UV-visible analysis. The ratio of 10 mg of MgO nanoparticle to 1 mL of sample solution was used in all subsequent extraction steps. A total of 9 extractions were performed, with successive reductions of solution volume and mass of nanoparticle by 1 mL and 10 mg, respectively, each time. In an independent experiment, a 9 4 ACS Paragon Plus Environment

Page 5 of 25

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

mL asphaltene solution (1 mg/mL) was treated with 0.81 g of NiO. Note that the nanoparticle mass was scaled up in this experiment to promote, in a single extraction, a removal efficiency similar to what would be expected for 9 successive extractions if the nanoparticle mass-to-solution volume ratio were as described above. This asphaltene solution was also analyzed by UV-visible spectroscopy before and after extraction. UV-visible analysis was performed using an Agilent 8453 UV-visible spectroscopy system equipped with deuterium and tungsten lamps. Data were acquired from 280-1100 nm, using a quartz cuvette (pathlength = 1 cm) and a fixed wavelength of 410 nm was selected for analysis. A toluene blank was applied as a baseline. 2.2b Asphaltene adsorption protocol for mass spectrometry (MS). The extraction protocol for MS analysis was the same as described in section 2.2a except that the initial amount of asphaltene (0.010 g in 10 ml) was extracted with 0.100 g of MgO (the 10 mg of nanoparticle to 1 mL sample mass-volume ratio remained the same), the sample was not centrifuged, and the dilution prior to MS analysis was 1:100. Nine successive extractions with MgO were performed, followed by a 10th extraction with NiO. The toluene solution of asphaltenes was analyzed via direct infusion on an LTQ Orbitrap mass spectrometer (Thermo Scientific, Waltham, MA) equipped with a dual APPI/APCI ionization source. Mass spectrometry conditions are reported in a prior publication [33]. Accurate mass measurements were calibrated using 9,10-diphenylanthracene (m/z 330.1403) as an internal reference. 2.2c Fourier transform infrared (FTIR) analysis of asphaltenes. A sample of asphaltene (0.010 g) was dissolved in 10 mL of toluene. A 9-mL aliquot of this solution was extracted with 0.81 g of MgO. Aliquots of the original and extracted solutions were independently used to cast a thin film of dissolved asphaltenes on a KBr salt plate (i.e., drops of solution were transferred directly onto a KBr

5 ACS Paragon Plus Environment

Energy & Fuels

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 6 of 25

salt plate and the solvent was allowed to evaporate between additions). FTIR spectra were recorded on a Thermo Nicolet 380 FT-IR spectrometer (Thermo Scientific, Waltham, MA) at room temperature from 500-4000 cm-1 with a spectral resolution of 4 cm-1.

3. RESULTS AND DISCUSSION 3.1 Effect of Successive Extractions on Asphaltene Composition. Nine successive extractions of an asphaltene sample with MgO nanoparticles led to a substantial reduction of dissolved asphaltene, as evidenced by decreased intensity of the sample’s UV-visible absorption spectra (Figure 1A). The average decrease in mean absorbance at 400 nm observed for each successive extraction was 13  2% for the first 6 extractions, relative to the absorbance observed for an unextracted sample (Figure 1B). To the extent that absorbance at 400 nm is representative of bulk asphaltene, the noted average provides a rough estimate of the relative mass of asphaltene removed in each initial extraction step. Such an assumption is supported by similar analyses of absorbance data collected at 410 nm and 500 nm. In both cases, decreases in mean absorbance for successive extractions were essentially identical to those observed at 400 nm on a percentage basis (see Supplementary Figures S1A and S1B). A notable decrease in extraction efficiency was observed for extractions 7-9, as evidenced by the much smaller absorbance change observed between these experiments. These data suggest that absorbance had reached a limiting value after 9 extractions, beyond which further extractions with MgO would likely remove only a minimal amount of remaining asphaltene from toluene solution. Decreases in absorbance over 9 extractions indicate that summative treatments with MgO removed 92-96% of bulk asphaltene. Mass spectrometry analysis of an asphaltene solution that was subjected to successive MgO extractions revealed a general preference for removal of higher molecular weight species. This

6 ACS Paragon Plus Environment

Page 7 of 25

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

conclusion is qualitatively supported by mass spectra in Figure 2 which show a general shift in the distribution of dissolved asphaltenes to lower m/z with increasing extraction number. Observed decreases in the average molecular weight (i.e., m/z) and the half-width of the asphaltene distribution are shown in Table 1 as a function of extraction number. These data quantitatively support that larger molecules were removed in preference to smaller molecules present in the sample; minimally for the first 6-7 extractions. Beyond the 7th extraction, increased relative intensity of background ions made it difficult to conclusively identify m/z values corresponding to the bulk average or bulk end of the asphaltene distribution. Thus, it is possible (even likely based on the mass spectra shown in Figure 2) that preferential extraction of larger molecules continued to occur during extractions 7-9. A definitive, molecular level understanding of the extraction mechanism responsible for removal of asphaltenes by MgO is beyond the scope of the present study. However, it is important to recognize that a dynamic equilibrium involving asphaltene monomers (i.e., independently solvated molecules) and a polydisperse collection of nanoaggregates is likely to exist in toluene solution. Prior to treatment with MgO, the asphaltene concentration in toluene was 1 g/L. This concentration is roughly one order of magnitude higher than the critical nanoaggregate concentration (CNAC) reported for asphaltenes in this solvent (0.1 g/L) [35]. Several studies have attributed asphaltene adsorption onto metal oxides to the presence of heteroatom functional groups [28, 32, 33]. If the structural orientation of presumed nanoaggregates is such that heteroatoms are insulated from solution (e.g., if the structure is similar, in principle, to an inverted micelle), or if heteroatoms are otherwise involved in intermolecular interactions that stabilizes nanoaggregates, the ability of these heteroatoms to interact with a MgO surface would be greatly reduced relative to the affinity expected for heteroatoms presented in monomer structures. Thus, it seems unlikely that nanoaggregates would be directly adsorbed onto MgO. A more plausible mechanism for removal of asphaltenes from solution would likely involve

7 ACS Paragon Plus Environment

Energy & Fuels

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 8 of 25

adsorption of monomers, which in turn would progressively shift the noted equilibrium in a direction favoring dissolution of nanoaggregates. Such a mechanism is consistent with previous work demonstrating that only asphaltene monomers are adsorbed at an oil-water interface [36]. A study of model compounds recently demonstrated that carboxyl groups are most likely responsible for adsorption of asphaltenes onto 100 nm MgO nanoparticles [33]. The same study demonstrated that phenolic and pyrolic functionalities may also contribute to asphaltene adsorption, but affinities observed for both of these groups were significantly smaller than the affinity observed for a carboxyl group. In the present study, infrared absorption spectra were collected on cast films of dissolved asphaltene before and after extraction (Figure 3). Qualitative spectral labeling in Figure 3 was based on previous FTIR studies of asphaltenes in literature [37-39]. Absorption was universally weaker in spectra obtained for the extracted sample, consistent with >90% of dissolved asphaltene being removed. In this particular analysis, absorption intensity for the extracted sample was sufficiently low, relative to the instrumental detection limit, to preclude a determination of whether oxygen-containing species were present. Thus, IR data in Figure 3 are inconclusive with respect to preferential removal of oxygen-containing species during the extraction of asphaltenes. Interestingly, UV-visible absorbance data suggest that pyrolic groups sequestered in ring systems (e.g., metalloporphyrins in asphaltenes) were not efficiently extracted. The absorbance maximum near 410 nm in Figure 1A has been previously attributed to the porphyrin Soret band [40]. This spectral feature becomes more pronounced with increasing extraction number, suggesting that the molecule(s) responsible for this portion of the absorbance profile are being extracted at a slower rate than other constituents in asphaltenes, if at all. UV-visible spectra were also obtained for an asphaltene sample that was extracted with NiO instead of MgO. These spectra were qualitatively similar to those

8 ACS Paragon Plus Environment

Page 9 of 25

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

observed when MgO was used (see Supplementary Figure S2). This result suggests that porphyrins present in asphaltenes also exhibit relatively weak affinity for NiO. A speculative explanation for the observed size bias (i.e., higher molecular weight species showing preferential extraction) relies on initial adsorption of heteroatom-containing molecules, but also allows for alternative interactions at the particle-solution interface. Although a probabilistic argument may be made in favor of higher molecular weight species being more likely to contain a heteroatom group(s) that is strongly attracted to MgO (i.e., a carboxyl or phenolic group), the mass percentage of oxygen determined for the asphaltene sample studied in this work (1.47  0.1%) is likely too small to explain the sum total of asphaltene removed over nine extractions. Assuming that the mass of an asphaltene molecule is 750 daltons, the mass percentage of oxygen required for every molecule to contain one oxygen atom is 2.1%. Recent work has demonstrated that aromatic core structures, composed of fused ring systems, are adsorbed at oil-water interfaces [41, 42]. Thus, it is possible that interactions between polycyclic aromatic hydrocarbon (PAH) moieties and the MgO surface also contribute to asphaltene removal. This would, indeed, be consistent with higher molecular weight species being preferentially extracted. However, a preliminary attempt to remove a simple PAH (i.e., anthracene) from toluene solution suggests low potential for a strong interaction between fused aromatic ring systems and MgO (see Supplementary Figure S3). Aggregation of asphaltenes does not appear to occur at oil-water interfaces [36, 41, 42], but very little is known about molecular interactions of asphaltenes at mineral surfaces. An extraction mechanism involving initial adsorption of heteroatom-containing monomers and subsequent secondary interactions between adsorbed molecules and asphaltenes remaining in solution seems plausible based on the evidence presented here.

9 ACS Paragon Plus Environment

Energy & Fuels

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 10 of 25

3.2 Meaningful Reductions of Complexity in Asphaltene Mass Spectra. Successive extractions with MgO also resulted in substantial and meaningful reductions of the complexity observed in asphaltene mass spectra. Mass spectrometry is able to separately detect and identify an extensive number of molecules in a given sample. However, the complexity of asphaltenes is sufficient to challenge the resolving power of even the most sophisticated Fourier transform ion cyclotron resonance instruments [34]. The adsorption strategy reported here represents a complementary separation with the ability to decrease spectral complexity by selectively removing molecules that have the same nominal mass but different molecular formulas or structures prior to mass spectral analysis. Note that this type of selectivity is independent of (and in addition to) the general trend noted above, whereby extraction with MgO was shown to preferentially remove larger molecules present in asphaltenes. A representative example of this concept can be seen by viewing the removal of “shoulder” peaks in a narrow m/z range upon successive extractions with MgO. An expanded view of asphaltene mass spectra, corresponding to nominal m/z 488, is shown in Figure 4 for various extraction numbers (i.e., 0, 5, and 9). Each successive extraction resulted in a relative reduction of overall complexity and a narrowing of the remaining asphaltene peaks by means of selective compound removal. For example, comparison of Figure 4A with Figure 4B demonstrates that 5 extractions with MgO resulted in removal of the “shoulder” peak appearing at lower mass relative to the major peak at m/z 488.17. Removal of the compound(s) responsible for the “shoulder” resulted in an approximate 50 ppm reduction of baseline peak width for m/z 488.17. Note that in this particular instance, the “shoulder” was removed with essentially no change in spectral intensity of the larger peaks appearing in this m/z window. An analogous comparison of Figure 4B with Figure 4C demonstrates that extractions 6-9 resulted in removal of similar “shoulders” on the two largest peaks appearing at higher m/z in this mass window. In this case, removal of the shoulder was accompanied

10 ACS Paragon Plus Environment

Page 11 of 25

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

by a noticeable (yet relatively minor) change in spectral intensity. A comparison of Figure 4A with Figure 4C displays the overall reduction of sample complexity that was achieved by 9 successive extractions with MgO. A similar reduction of complexity could likely be achieved with fewer or perhaps a single extraction(s), if desired, by increasing the mass of MgO used relative to the volume of asphaltene solution being extracted.

3.3 Improved Ability to Identify Pyridyl-Containing Molecules in Asphaltenes via Selective Adsorption onto NiO. Recent work has shown that the discriminative adsorption behavior of asphaltenes onto MgO and NiO can be used in combination with mass spectrometry to enable identification of pyridyl-containing molecules at a particular m/z [33]. In this earlier proof-of-concept example, two successive extractions of an asphaltene solution with MgO were followed by a single extraction with NiO. Mass spectra were compared before and after each extraction, and m/z peaks that were not removed by MgO, yet were removed by NiO, were assigned to molecules containing a pyridyl functional group(s) in their structure(s). In the same study, model compounds were used to demonstrate that MgO displayed high affinity for an aromatic structure containing a carboxyl group but essentially no affinity for pyridine. In contrast, NiO displayed high affinity for both structural groups. In the present study, nine successive extractions of an asphaltene solution with MgO were followed by a single extraction with NiO. Representative mass spectra are shown in Figure 5, as a function of extraction number, for three nominal masses spanning the m/z range 460-490 Daltons. Arrows in Figure 5 designate m/z peaks that were relatively insensitive to (i.e., not removed by) extraction with MgO but that disappeared (essentially) after a single contact with NiO. Ions appearing at these m/z values almost certainly correspond to pyridyl-containing asphaltenes. In total, 205 peaks

11 ACS Paragon Plus Environment

Energy & Fuels

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 12 of 25

in the m/z range 310-699 Daltons exhibited similar extraction behavior (Table S1). Note that mass spectra corresponding to the 1st, 5th, and 9th extractions in Figure 5 not only confirm the insensitivity of some peaks to MgO, but also serve to reinforce the discussion of reduced mass spectral complexity presented in Section 3.2. With respect to providing insight on chemical structures present in asphaltenes, the experiment summarized in Figure 5 represents a substantial improvement over our previous work. In terms of profiling pyridyl-containing molecules, the earlier study [33] resulted in assignment of 45 m/z peaks. In the present study, which utilized the same asphaltene sample, the number of peaks assigned increased by more than a factor of four. Moreover, after nine extractions with MgO, one can be reasonably confident that most carboxyl-containing structures are unlikely to remain in solution. Thus, any oxygen appearing in elemental formulas for m/z peaks assigned to pyridyl-containing structures is more likely to be represented in an alternative oxygen-containing functional group (e.g., ketone or phenolic). This statement about oxygen may also be extended to elemental formulas determined for any peaks remaining in the mass spectrum after subsequent treatment with NiO. Remaining peaks (that were not present in the toluene blank) likely correspond to asphaltene molecules that exhibit low affinity for both MgO and NiO (e.g., molecules with ketone, phenolic, thiol, disulfide, and/or pyrolic functionality). A more exhaustive extraction with MgO is also expected to improve confidence in elemental formula assignments derived from accurate mass measurements. Definitive elemental formula assignments should not be pursued for asphaltenes without access to instrumentation that has higher resolving power than that used in this work. However, the comparative examples presented below are included to illustrate how one could come to erroneous conclusions about elemental composition of pyridyl-containing asphaltenes if only 2 MgO extractions were employed prior to treatment with NiO.

12 ACS Paragon Plus Environment

Page 13 of 25

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

Table 2 lists possible elemental formulas, within a mass accuracy threshold of 3 ppm, for peaks designated by arrows in the center panel of Figure 5 (i.e., for m/z 476.17, 476.26, and 476.35). Additional criteria imposed on formula assignments are given in the table’s footnote. Formulas listed on the left and right hand side of the table, respectively, were determined from mass measurements made after 9 and 2 successive extractions of an asphaltene solution with MgO. Given the differential selectivity imparted by successive MgO and NiO extractions, as well as the high adsorption affinity of NiO for pyridyl functional groups, it is unlikely that the structure(s) responsible for a given m/z peak in Table 2 do not contain nitrogen. Additionally, it is acknowledged that a single peak in the mass spectrum of asphaltenes may represent more than one compound, but heteroatoms are likely to be represented in similar structures (i.e., isomers). Focusing first on m/z 476.17 in Table 2, it can be observed that 9 successive extractions with MgO resulted in fewer elemental formula possibilities relative to the scenario utilizing 2 extractions. Only five formulas were common to both extraction scenarios, and their positions on the mass accuracy scale were reversed. That is, among formulas common to both lists, the formula with the smallest mass error after 9 extractions (#1) had the highest mass error after only 2 extractions (#11), and the formula with the highest mass error after 9 extractions (#8) had the smallest mass error after only 2 extractions (#2). Possible formulas for m/z 476.26 were more comparable between the two extraction scenarios. The top four candidates in terms of mass accuracy were common to both scenarios, and the formula with the smallest mass error was identical in both lists. In contrast, none of the possible formulas for m/z 476.35 were common between the two extraction scenarios. It was shown earlier that peak purity improves following a more exhaustive extraction with MgO (see Figures 4 and 5 for examples). Thus, mass measurements made after 9 extractions are generally expected to reflect superior accuracy relative to those made after only 2 extractions.

13 ACS Paragon Plus Environment

Energy & Fuels

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 14 of 25

4. Conclusions Successive extractions of an asphaltene solution with MgO resulted in preferential adsorption of higher molecular weight species. Thus, the approach presented here could potentially be utilized as a “pull down” strategy in future work. Given that the evolution of asphaltene structure as a function of molecular weight remains a mystery, such an effort could have a significant impact on the fundamental knowledge base related to asphaltenes. Additionally, it has been demonstrated that a more exhaustive extraction of asphaltenes with MgO, prior to treatment with NiO, greatly improves the ability to profile pyridyl-containing structures in a complex asphaltene mixture. Lastly, substantial reductions in mass spectral complexity afforded by a complementary separation of asphaltenes (with MgO and NiO) may provide an opportunity for meaningful collision induced dissociation (CID) experiments on select asphaltene ions using alternative instrumentation (e.g., FTICR-MS). Such a statement is supported by the observation of what appears to be a single peak in the expanded mass spectrum shown for nominal m/z 490 in Figure 5. However, it will be important to confirm the purity of this and similar peaks at higher resolving power before making a definitive conclusion to this end.

Acknowledgments This paper acknowledges the Baylor University Mass Spectrometry Center for instrument access and support provided during the course of this work. This work was funded in part by the Baylor University Faculty Reinvestment Program and the Department of Chemistry and Biochemistry.

Supporting Information Available This information is available free of charge via the Internet at http://pubs.acs.org/.

14 ACS Paragon Plus Environment

Page 15 of 25

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

References (1) Herod, A. A.; Bartle, K. D.; Morgan, T. J.; Kandiyoti, R. Chemical Reviews, 2012, 112, 3892-3923. (2) Speight, James G. The Chemistry and Technology of Petroleum, 4th ed.; CRC Press Taylor and Francis Group: Boca Raton, FL, 2007. (3) Buckley, J. S. Energy Fuels, 2012, 26, 4086-4090. (4) Girdler, R. B. Proc. Assoc. Asphalt Paving Technol., 1965, 34, 45-79. (5) Mitchell, D. L. and Speight, J. G. Fuel, 1973, 52, 149-152. (6) Long, R. B. Div. Petrol. Chem. Am. Chem. Soc. 1979 Preprints, 24(4), 891-900. (7) Long, R. B. The Chemistry of Asphaltenes. Bunger, J.W. and Li, N., eds. Advances in Chemistry Series No. 195. American Chemical Society, Washington, D.C. 1981. (8) Speight, J. G. Asphaltenes and Asphalts. I. Developments in Petroleum Science, 40. Yen, T. F. and Chilingarian, G. V. eds. 1994, Elsevier, Amsterdam. Chapter 2. (9) Mitra-Kirtley, S., Mullins, O. C., van Elp, J., George, S. J., Chen, J., and Cramer, S. P. J. Am. Chem. Soc., 1993, 115, 252-258. (10) Nicksic, S. W. and Jeffries-Harris, M. J. J. Inst. Petrol., 1968, 54, 107-111. (11) Petersen, J. C., Barbour, F. A., and Dorrence, S. M. Proc. Assoc. Asphalt Paving Technol.,1974, 43, 162-177. (12) Moschopedis, S. E. and Speight, J. G. Fuel, 1976a, 55, 187-192. (13) Moschopedis, S. E. and Speight, J. G. Fuel, 1976b, 55, 334-336. (14) Ritchie, R. G. S., Roche, R. S. and Steedman, W. Fuel, 1979, 58, 523-530. (15) Speight, J. G. and Moschopedis, S. E. Div. Petrol. Chem. Am. Chem. Soc. 1981b, Preprints, 26(4), 907-911. (16) Rose, K. D. and Francisco, M. A. Energy Fuels, 1987, 1, 233-239.

15 ACS Paragon Plus Environment

Energy & Fuels

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 16 of 25

(17) Clerc, R. J. and O’Neal, M. J. Anal. Chem., 1961, 33, 380-382. (18) Drushel, H. V. Div. Petrol. Chem. Am. Chem. Soc. 1970, Preprints, 15, C13. (19) Yen, T. F. Energy Sources, 1974, 1, 447-463. (20) Speight, J. G. and Pancirov, R. J. Liq. Fuels Technol., 1984, 2, 287-305. (21) Rose, K. D. and Francisco, M. A. J. Am. Chem. Soc., 1988, 110, 637-638. (22) Keleman, S. R., George, G. N., and Gorbaty, M. L. Fuel, 1990, 69, 939-944. (23) Mackie, I. D.; Di Labio, G. A. Energy & Fuels, 2010, 24, 6468-6475. (24) Becker C.; Qian, K.; Russell, D. H. Analytical Chemistry, 2008, 80, 8592-8597. (25) Mullins, O. C.; Martinez-Haya, B.; Marshall, A. G. Energy & Fuels, 2008, 22, 1765-1773. (26) Sabbah, H., Pomerantz, A. E., Wagner, M., Müllen, K., and Zare, R. N. Energy Fuels, 2012, 26, 3521-3526. (27) Pinkston, D. S., Duan, P., Gallardo, V. A., Habicht, S. C., Tan, X., Qian, K., Gray, M., Mullen, K., and Kenttämmaa, H. I., Energy Fuels, 2009, 23, 5564-5570. (28) Nassar, N. N.; Hassan, A.; Pereira-Almao, P. Energy & Fuels, 2011, 25, 1017-1023. (29) Cortes, F. B.; Mejia, J. M.; Ruiz, M. A.; Benjumea, P.; Riffel, D. B. Energy & Fuels, 2012, 26, 1725-1730. (30) Franco, C.; Patino, E.; Benjumea, P.; Ruiz, M. A.; Cortes, F. B. Fuel, 2013, 105, 408-414. (31) Hosseinpour, N.; Khodadadi, A. A.; Bahramian, A.; Mortazavi, Y. Langmuir, 2013, 29, 1413514146. (32) González, M. F.; Stull, C. S.; Lópex-Linares, F.; Pereira-Almao, P. Energy & Fuels, 2007, 21, 234241. (33) Zimmer, A. K.; Becker, C.; Chambliss, C. K. Energy & Fuels, 2013, 27, 4574-4580. (34) Xian, F.; Hendrickson, C. L.; Marshall, A. G. Analytical Chemistry, 2012, 84, 708-719.

16 ACS Paragon Plus Environment

Page 17 of 25

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

(35) Andreatta, G.; Bostrom N.; Mullins, O. C. Langmuir, 2005, 21, 2728-2736. (36) Rane, J. P.; Harbottle, D.; Pauchard, V.; Couzis, A.; Banerjee, S. Langmuir 2012, 28, 9986-9995. (37) Shayan, N. N.; Mirzayi, B. Energy & Fuels 2015, 29, 1397-1406. (38) Davarpanah, L.; Vahabzadeh, F.; Darmanaki, A. Oil Gas Sci. Technol. 2013, DOI: 10.2516/ogst/2012066. (39) Sharma, B. K.; Sharma, C. D.; Tyagi, O. S.; Bhagat, S. D.; Erhan, S. Z. Petrol. Sci. Technol. 2007, 25, 121-139. (40) Putnam, J. C.; Rowland, S. M.; Corilo, Y. E.; McKenna, A. M. Anal. Chem. 2014, 86, 1070810715. (41) Andrews, A. B.; McClelland, A.; Korkeila, O.; Demidov, A.; Krummel, A.; Mullins, O. C.; Chen, Z. Langmuir 2011, 27, 6049-6058. (42) Rane, J. P.; Pauchard, V.; Couzis, A.; Banerjee, S. Langmuir 2013, 29, 4750-4759.

17 ACS Paragon Plus Environment

Energy & Fuels

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 18 of 25

Table 1. Bulk qualities of the asphaltene distribution observed by mass spectrometry as a function of extraction number.

Extraction #a 0 1 2 3 4 5 6 7 8 9

a b c d e

Averageb (m/z) 540 530 520 490 480 490 420 440 410 440

Endc (m/z) 720 690 660 620 580 600 470 510 460 530

Half-widthd (m/z) 180 160 140 130 100 110 50 70 50 90

Ratioe 1.33 1.30 1.27 1.27 1.21 1.22 1.12 1.16 1.12 1.20

Indicates how many sequential extractions took place. Indicates the approximate center of the observed asphaltene distribution (i.e., average molecular weight, assuming singly-charged species). Indicates the approximate end of the observed asphaltene distribution at high m/z. Refers to the difference: End  Average. Refers to the ratio: End/Average.

18 ACS Paragon Plus Environment

Page 19 of 25

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

Table 2. Possible molecular formulas for pyridyl-containing structures in the m/z window 476.0-476.4 after 9 or 2 sequential extractions with MgO.a  

Data for 9 Extractions m/z

476.1678

 

#

Formula [M+H] +

1 2 3 4 5 6 7 8 9

C31 H26 O2 N S C23 H22 O5 N7 C16 H30 O2 N9 S3 C23 H30 O4 N3 S2 C17 H28 O8 N6 S C15 H26 O7 N9 S C24 H26 N7 S2 C25 H24 O6 N4 C29 H24 O N4 S

 

476.2572 

 

476.3484

 

DBE

Error (ppm)

19.5 16.5 6.5 10.5 7 7.5 15.5 16 20

-0.16 0.22 -0.23 1.21 -1.23 1.59 -1.60 -2.60 2.66

 

m/z

#

Formula [M+H] +

DBE

Error (ppm)

476.1692

1 2 3 4 5 6 7 8 9 10 11 12

C18 H32 O3 N6 S3 C25 H24 O6 N4 C18 H24 O4 N10 S C19 H30 O9 N3 S C24 H26 N7 S2 C26 H28 O N4 S2 C17 H28 O8 N6 S C26 H20 O2 N8 C27 H26 O7 N C16 H30 O2 N9 S3 C31 H26 O2 N S C20 H34 O4 N3 S3

6 16 12 6.5 15.5 15 7 21 15.5 6.5 19.5 5.5

-0.11 0.35 -1.10 -1.11 1.34 -1.48 1.71 -2.46 -2.48 2.71 2.78 -2.93

 

1  2  3 4 5 6

C31 H32 O N4  C17 H34 O7 N9  C25 H38 O4 N3 S C23 H36 O3 N6 S C33 H34 O2 N C18 H38 O2 N9 S2

18  5.5  8.5 9 17.5 4.5

0.29 -0.78 -1.16 1.66 -2.53 -2.60

476.2569

1 2 3 4 5

C31 H32 O N4 C23 H36 O3 N6 S C17 H34 O7 N9 C25 H38 O4 N3 S C29 H30 N7

18 9 5.5 8.5 18.5

-0.34 1.03 -1.41 -1.79 2.48

1 2

C27 H46 O4 N3 C28 H42 N7

6.5 11.5

0.25 -2.56

476.3512

1 2

C30 H44 O N4 C32 H46 O2 N

11 10.5

0.50 -2.32

 

 

 

  a

Data for 2 Extractions

 

 

Formula assignments assume singly-charged ions. Additional criteria imposed were as follows: DBE (double bond equivalence) ≥ 4; mass error  3ppm; C  40, H  80, N and O  10, and S  15.

19 ACS Paragon Plus Environment

Energy & Fuels

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 20 of 25

FIGURE CAPTIONS: FIGURE 1. Effect of sequential MgO extractions on the UV-Visible absorption properties of an asphaltene solution. (A) The top (black) curve corresponds to the authentic (unextracted) sample. The baseline is relative to a toluene blank (shown in red). (B) Mean absorption (n = 3) at 400 nm decreased linearly for the first 6-7 extraction cycles and leveled off thereafter, suggesting that additional extractions would result in minimal removal of compounds from toluene solution. Error bars represent one standard deviation.

FIGURE 2. Effect of sequential MgO extractions on mass spectra of asphaltenes, demonstrating preferential removal of higher molecular weight species. Data were collected in positive APCI/APPI mode over the mass range 250-700 m/z.

FIGURE 3. FTIR spectra of dissolved asphaltenes before (top) and after (bottom) extraction with MgO.

FIGURE 4. Effect of sequential MgO extractions on asphaltene mass spectra at nominal m/z 488, demonstrating a substantial reduction of overall sample complexity.

FIGURE 5. Effect of nine sequential MgO extractions, and a tenth extraction with NiO, on three nominal masses spanning the m/z range 460-490 Daltons. Arrows indicate peaks corresponding to pyridyl-containing structures (see text for details). From left to right, these peaks correspond to m/z 460.23, 476.17, 476.26, 476.35, 490.28, and 490.37.

20 ACS Paragon Plus Environment

Page 21 of 25

Figure 1

A Absorbance

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

B

21 ACS Paragon Plus Environment

Energy & Fuels

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 22 of 25

Figure 2

22 ACS Paragon Plus Environment

Page 23 of 25

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

Figure 3

Before Extraction

After Extraction

23 ACS Paragon Plus Environment

Energy & Fuels

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 24 of 25

Figure 4

24 ACS Paragon Plus Environment

Page 25 of 25

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

Figure 5

25 ACS Paragon Plus Environment