Article pubs.acs.org/EF
Heavy Petroleum Composition. 3. Asphaltene Aggregation Amy M. McKenna,† Lynda J. Donald,‡ Jade E. Fitzsimmons,† Priyanka Juyal,§ Victor Spicer,∥ Kenneth G. Standing,∥ Alan G. Marshall,*,†,⊥ and Ryan P. Rodgers*,†,⊥ †
National High Magnetic Field Laboratory, Florida State University, 1800 East Paul Dirac Drive, Tallahassee, Florida 32310-4005, United States ‡ Department of Chemistry, University of Manitoba, Winnipeg, Manitoba R3T 2N2, Canada § Energy Services Division, Nalco Company, 7705 Highway 90-A, Sugar Land, Texas 77478, United States ∥ Department of Physics and Astronomy, University of Manitoba, Winnipeg, Manitoba R3T 2N2, Canada ⊥ Department of Chemistry and Biochemistry, Florida State University, 95 Chieftain Way, Tallahassee, Florida 32306, United States ABSTRACT: Molecular characterization of asphaltenes by conventional analytical techniques is a challenge because of their compositional complexity, high heteroatom content, and asphaltene aggregate formation at low concentrations. Thus, most common characterization techniques rely on bulk properties or solution-phase behavior (solubility). Proposed over 20 years ago, the Boduszynski model proposes a continuous progression in petroleum composition (molecular weight, structure, and heteroatom content) as a function of the atmospheric equivalent boiling point. Although exhaustive detailed compositional analysis of petroleum distillates validates the continuum model, the available compositional data from asphaltene fractions supports the extension of the continuum model into the nondistillables only indirectly. Asphaltenes, defined by their insolubility in alkane solvents, accumulate in high-boiling fractions and form stable aggregate structures at low parts per billion (ppb) concentrations, far below the concentration required for most mass analyzers. Here, we present direct mass spectral detection of stable asphaltene aggregates at lower concentrations than previously published and observe the onset of asphaltene nanoaggregate formation by time-of-flight mass spectrometry (TOF−MS). We conclude that a fraction of asphaltenes must be present as nanoaggregates (not monomers) in all atmospheric pressure and laser-based ionization methods. Thus, those methods access a subset of the asphaltene continuum.
■
to 400 °C) petroleum fractions but still provides limited molecular-level detail for asphaltenes.19,20 Vapor pressure osmometry (VPO), diffusion measurement,21−24 gel permeation chromatography (GPC), and sizeexclusion chromatography (SEC) applied to asphaltenes have produced a wide range of average molecular weights, thought to be influenced by concentration-dependent self-association.1,22,25−29 Asphaltene molecular weight measurements are inconclusive because of asphaltene self-association at sub-parts per billion (ppb) concentrations.30−33 Thus, VPO and GPC measurements typically reflect the mass of asphaltene aggregates as well as asphaltene monomers.2,3 Fluorescence depolarization (FD) 34−36 yields an average asphaltene molecular weight (MW) of 450−850 g/mol but remains controversial.37−40 Light and small-angle X-ray scattering,8,41−48 ultracentrifugation,49,50 ultrasonic relaxation,30,51 ultrafiltration,52−55 calorimetry,56 low-frequency dielectric relaxation spectroscopy,57 and direct-current (DC) conductivity13,32,58,59 measurements report asphaltene aggregate formation at subppm concentrations. Asphaltene aggregates have been detected at high temperature in reservoir,60 and the results suggest in situ stable aggregate formation.42,61 The decades-long controversy about asphaltene molecular weight,62,63 structure,16,38,64,65 and the ability of conventional
INTRODUCTION
Oil recovery can induce asphaltene precipitation because of pressure, temperature, and oil-phase compositional changes, and resultant undesirable phase separation negatively impacts every facet of oil production.1−5 Asphaltene deposition has been linked to problems associated with well production and exploration,6 pipeline transport, land- and sea-based transportation, oil refining and processing,1,4,6,7 and emulsion stabilization.2,3,8−11 Asphaltene aggregation, precipitation, flocculation, and deposition mechanisms remain controversial.12−17 Molecular-level characterization of heavy petroleum is difficult because of its low volatility, high polarity, and compositional polydispersity. Asphaltenes are conventionally defined, not by their chemical functionality and structure, but operationally as those species that precipitate upon the addition of excess paraffinic solvent (e.g., n-pentane and n-heptane) but remain soluble in aromatic solvents (e.g., toluene and benzene).2,4,18 Analytical characterization of asphaltenes has until recently been limited to bulk properties. The most universally accepted asphaltene property is the bulk H/C ratio ≈ 1, with heteroatom content varying from a few percentages (by weight) to 10% sulfur for heavy oil and bitumen.4,5,7,18 The asphaltene fraction for any crude oil carries the lowest economic value, because they are comprised of the most aromatic molecules and heteroatomic species (nitrogen, sulfur, and oxygen). Advances in analytical technology can now address the wide polydispersity associated with high-boiling (up © 2013 American Chemical Society
Received: November 14, 2012 Revised: January 15, 2013 Published: January 16, 2013 1246
dx.doi.org/10.1021/ef3018578 | Energy Fuels 2013, 27, 1246−1256
Energy & Fuels
Article
observation of gas-phase asphaltene aggregates of average m/z ≈14 000, well above the upper mass limit for other mass analyzers. Our results suggest that all mass spectral techniques are limited to observation of a subset of asphaltenes that are not tightly bound in stable aggregates.
ionization techniques to efficiently ionize asphaltene monomers66 is a direct consequence of the tendency of asphaltenes to form stable aggregates at low concentrations.4,5 Most reports converge on the asphaltene molecular weight range of 500− 2000 Da, as summarized elsewhere.2,3,31,67−70 The first direct measurement of high-molecular-weight petroleum compounds was performed by Roussis et al., who reported bimodal asphaltene molecular weights of 300 < m/z < 20 000 obtained with a hybrid magnetic sector−time-of-flight (TOF) tandem mass spectrometer.71,72 They attributed peaks above m/z ∼3000 to monomers, even though in-source collisional activation significantly depleted the high-molecularweight components. On the basis of model compounds, they concluded that “the majority of the structures in the saturate, (naphtheno)aromatic, polar aromatic and a significant portion of asphaltenes are small molecules” but that most asphaltenes are high-molecular-weight species.71 Asphaltene mass analysis is further complicated by the need for compatible solvent systems. Asphaltenes are “completely” soluble in aromatic solvents, such as benzene and toluene, but modified solvent systems [typically required in electrospray ionization (ESI)] destabilize the asphaltene solution and result in flocculation. Thus, attempts to characterize asphaltenes by ESI have been largely unsuccessful. Moreover, methods that require evaporation of solvent droplets or films [ESI,73−78 atmospheric pressure photoionization (APPI),79−81 laser desorption (LD),82,83 laser-induced acoustic desorption (LIAD), 84,85 matrix-assisted laser desorption/ionization (MALDI),86 and atmospheric pressure chemical ionization (APCI)87] require analyte concentrations of ∼250−1000 μg/ mL, nearly an order of magnitude above the critical nanoaggregate concentration for asphaltenes (CNAC ≈ 50 μg/mL), and therefore, potentially discriminate against those asphaltenic components that preferentially form stable nanoaggregates. The Boduszynski continuum model dictates that the composition of the nondistillable fraction of crude oil is grounded in the low-boiling fraction and that asphaltenes represent an extension of those structural motifs.88−91 Validation of the Boduszynski model in parts 1 (10.1021/ ef100149n) and 2 (10.1021/ef1001502) of this series established the compositional continuum of petroleum and established compositional boundaries for distillable petroleum species (maltenes).88,89 Extension of the compositional trends of distillates to asphaltenes requires detailed molecular-level characterization of distillates as a function of the boiling point, molecular weight, and molecular structure, accessible only by ultrahigh-resolution mass spectrometry. Mass spectral analysis of asphaltenes above the CNAC should reveal nanoaggregates (m/z > 2000) and monomeric species (m/z < 2000). Time-of-flight mass spectrometry (TOF−MS)92,93 offers access to singly charged ions up to 20 kDa and is thus well-suited for detection of asphaltene aggregates. Conversely, the compositional polydispersity of asphaltenes is at least 10-fold higher than for the parent crude and therefore requires ultrahigh mass resolving power (m/ Δm50% > 500 000 at m/z 500) obtainable only by Fourier transform ion cyclotron resonance mass spectrometry (FT-ICR MS) for determination of elemental composition (CcHhNnOoSs) of lower mass species. Here, we perform both ultrahigh-resolution FT-ICR MS and high-molecular-weight TOF−MS from custom-built93 and commercial platforms. TOF−MS provides the first direct
■
EXPERIMENTAL SECTION
Sample Preparation. All solvents were high-performance liquid chromatography (HPLC)-grade and purchased from Sigma-Aldrich Chemical Co. (St. Louis, MO). Middle Eastern heavy crude oil vacuum residue (593+ °C) was supplied by General Electric Global Research (Niskayuna, NY) and fractionated according to IP 143/90. Briefly, 500 mL of n-heptane was added to the residue sample (10 g), refluxed for 1 h in a 1 L round-bottom flask, and stored in the dark (24 h). The solids (asphaltenes) were isolated by gravity filtration through Whatman (Kent, U.K.) 2V-grade filter paper. The n-heptane-soluble fraction (maltenes) was desolvated under dry nitrogen. Hot heptane added to the asphaltene residue completed the transfer of solids. The filter paper with asphaltenes was further refluxed with additional heptane at a rate of 3−5 solvent drops/min until the solution ran clear to remove any entrained species (2 h). The asphaltene fraction was desorbed from the filter paper with toluene, desolvated under dry nitrogen, and weighed. Stock solutions were prepared by dissolving ∼5 mg of material in 5 mL of toluene. A 2 mL aliquot was further diluted with 6 mL of toluene to yield 8 mL of analyte solution (250 μg/mL) for APPI FT-ICR mass spectral analysis for both fractions. Electrospray samples were diluted to final concentrations (from 5 ng/mL to 500 μg/mL) in 50:50 (v/v) toluene/methanol and further modified with 1% (by volume) formic acid to aid in protonation. Silver Complexation. Roussis and Proulx71 observed silver nitrate adducts formed from nonpolar species from heavy petroleum fractions, not detectable by conventional ESI. Silver adduction occurs by silver cation (Ag+) binding to potential electron sources through cation−π interaction, host−guest, or donor−acceptor chemistry. Silver trifluoromethane sulfonate (CF3SO3Ag, CAS number 2923-28-6) was purchased from Sigma-Aldrich (St. Louis, MO). The triflate anion (trifluoromethanesulfonate anion, CF3SO3−) is resonance-stabilized because of the dispersal of negative charge over three oxygen atoms and one sulfur atom. The triflate anion is a more facile leaving group than the nitrate anion because of the strong electron-withdrawing effect of the trifluoromethyl group. The exemplary stability and nonnucleophilicity of the triflate anion enhance the electron affinity of the silver cation and thus favor efficient ionic complex formation with aromatic structures.71 Complexation was achieved at 3:1 (w/w) silver triflate/crude oil in 1:1 (v/v) toluene/methanol, vortexed immediately prior to introduction into the ESI source. Instrumentation: APPI Source. The APPI source (ThermoFisher Scientific, San Jose, CA) was coupled to the first pumping stage of a custom-built FT-ICR mass spectrometer (see below) through a custom-built interface.80,94 A Hamilton gas-tight syringe (2.5 mL) and syringe pump were used to deliver the sample (50 μL/min) to the heated vaporizer region (325 °C) of the APPI source, where N2 sheath gas (50 psi) facilitates nebulization, while the auxiliary port remains plugged. Gas-phase molecules flow out of the heated vaporizer in a confined jet beneath a krypton vacuum ultraviolet gas discharge lamp (∼10−10.2 eV photons, 120 nm), where photoionization occurs. Toluene increases the ionization efficiency for nonpolar aromatic compounds by means of dopant-assisted APPI through charge exchange,95,96 and proton transfer80 between ionized toluene molecules and neutral analyte molecules occurs at atmospheric pressure.97 Protonated ions exhibit half-integer double bond equivalent values (DBE = c − h/2 + n/2 + 1, calculated from the molecular ion elemental composition, CcHhNnOoSs+) and may thus be distinguished from radical cations with integer DBE values. Instrumentation: 9.4 T FT-ICR MS. Asphaltene and maltene fractions isolated from Middle Eastern heavy crude oil were analyzed with a custom-built FT-ICR mass spectrometer98 equipped with a 9.4T horizontal 220 mm bore diameter superconducting solenoid magnet operated at room temperature and a modular ICR data station 1247
dx.doi.org/10.1021/ef3018578 | Energy Fuels 2013, 27, 1246−1256
Energy & Fuels
Article
Figure 1. Selected distributions of positive-ion ESI TOF mass spectra from the Manitoba TOF instrument for asphaltenes isolated from a Middle Eastern heavy crude oil, at each of six concentrations. The declustering voltage was 120 V for all spectra. The broadband mass spectra reveal the onset of nanoaggregate formation at ∼50 μg/mL, consistent with previous reports in the literature. The solvent system is 50:50 toluene/methanol. increase resolution for isobaric species as previously described.105 Absorption-mode spectral resolving power is higher by a factor of up to 2 than magnitude-mode resolving power. Frequency-to-m/z calibration and data analysis are subsequently performed as for magnitude-mode display. Mass Calibration and Data Analysis. ICR frequencies were converted to ion masses based on the quadrupolar trapping potential approximation.106,107 and internally calibrated with respect to a highly abundant homologous alkylation series, whose members differ in mass by integer multiples of 14.015 65 Da (mass of CH2 unit) confirmed by isotopic fine structure based on the “walking” calibration equation.75,108,109 Elemental compositions were confirmed by identification of the 34S isotopologue peak at 1.9958 Da (mass difference between 34 S and 32S) higher mass than the monoisotopic peak. Experimentally
(Predator) facilitating instrument control, data acquisition, and data analysis.99,100 Positive ions generated at atmospheric pressure were accumulated in an external linear octopole ion trap101 for 250−500 ms and transferred by radio frequency (rf)-only octopoles102 to a 10 cm diameter, 30 cm long open cylindrical Penning ion trap. Octopoles were operated at 2.0 MHz and 240 Vp−p amplitude. Broadband frequency sweep (“chirp”) dipolar excitation (70−700 kHz at 50 Hz/ μs sweep rate and 350 Vp−p amplitude) was followed by direct-mode image current detection to yield 8 Mword time-domain data. Digitized time-domain transients were co-added (200 acquisitions), Hanningapodized, and zero-filled once before fast Fourier transform and magnitude calculation.100,103 Instrumentation: Broadband Phase Correction. Because of higher mass spectral complexity at higher m/z, broadband phase correction104 was applied to the entire FT-ICR mass spectrum to 1248
dx.doi.org/10.1021/ef3018578 | Energy Fuels 2013, 27, 1246−1256
Energy & Fuels
Article
Figure 2. Positive-ion ESI spectrum obtained on the Manitoba TOF instrument for 500 μg/mL asphaltenes isolated from Middle Eastern heavy crude oil and complexed with silver triflate, with a declustering voltage of 160 V. Asphaltene monomer, dimer, and multimer distributions are observed, extending up to ∼22 kDa, reflecting asphaltene aggregate formation in whole crude oil at concentrations below those required for all mass analyzers. The solvent system is 50:50 toluene/methanol. measured masses were converted from the International Union of Pure and Applied Chemistry (IUPAC) mass scale to the Kendrick mass scale110 for identification of the homologous series for each heteroatom class (i.e., species with the same NnOoSs content, differing only by their degree of alkylation).111−113 Peak assignments were performed by Kendrick mass defect analysis, as previously described.75,76,112 For each elemental composition, CcHhNnOoSs, heteroatom class, type (DBE = the number of rings plus double bonds involving carbon),114 and carbon number, c, were tabulated for subsequent generation of heteroatom class relative abundance distributions and graphical images of DBE or H/C ratio versus carbon number. Instrumentation: Custom-Built TOF−MS. Positive-ion electrospray and silver cationization TOF mass spectra were acquired with a custom-built TOF mass spectrometer at the University of Manitoba (Winnipeg, Manitoba, Canada).93 A 16 kV accelerating voltage increases transmission and detection efficiency for high mass ions.92,93,115−121 Asphaltene samples were prepared as described above. A 25 μL Hamilton syringe was used to deliver samples from a 26-gauge metal capillary at 0.2 μL/min. Instrumentation: TOF−MS, Waters Corporation. TOF mass spectral characterization of the parent residue and isolated asphaltenes confirmed was performed with a Waters LCT Premier mass spectrometer (Waters Corporation, Beverly, MA). Data were acquired in continuum acquisition mode at a rate of 1 spectrum/s, under conventional ESI conditions, 2.5 kV needle potential, and a sample infusion rate of 1 μL/min. The only modifications were to the ion transmission optics, which were set to their highest possible voltages to maximize transmission of high m/z species. The accelerating voltage was set to 8 kV.
The concentration dependence of asphaltene aggregate formation is shown in Figure 1 for asphaltene concentrations (5, 50, and 500 ng/mL and 5, 50, and 250 μg/mL) bracketing the previously determined asphaltene CNAC (∼50 μg/mL). TOF mass spectra suggest that asphaltenes are largely unassociated or weakly associated (dimer/trimer) at low concentrations (left panels of Figure 1) but form prominent aggregates of 100-fold higher signal-to-noise ratio at 50 μg/mL. At higher concentration (250 μg/mL), the asphaltene aggregates yield a tailing Gaussian distribution centered at m/ z 7000 with a full width at half height of m/z 2500. The increase in the average molecular weight with increasing asphaltene concentration constitutes clinching evidence for non-covalent aggregation. The present results necessitate re-interpretation of prior mass spectral measurements of asphaltene molecular weight. The asphaltenic species of greatest interest (i.e., those that nanoaggregate) are not accurately represented in the lowmolecular-weight (monomeric) distribution measured by FTICR MS. Instead, we report a complex overlapping distribution of singly charged aggregates of variable composition that are present as monomers and, thus, are not detected by MS at m/z < 2000. It is unclear at this time what effect, if any, the desolvation process imparts on the aggregation of asphaltenes in the low-concentration (
