Correlations between Molecular Composition and Adsorption

Subscriber access provided by UNIV OF DURHAM

Article

Correlations between molecular composition and the adsorption, aggregation and emulsifying behavior of Petrophase 2017 asphaltenes and their TLC fractions Deisy Giraldo-Dávila, Martha L. Chacón-Patiño, Amy M McKenna, Cristian Blanco-Tirado, and Marianny Y. Combariza Energy Fuels, Just Accepted Manuscript • DOI: 10.1021/acs.energyfuels.7b02859 • Publication Date (Web): 14 Dec 2017 Downloaded from http://pubs.acs.org on December 15, 2017

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 41 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

Correlations

between

molecular

composition

and

the

adsorption,

aggregation and emulsifying behavior of Petrophase 2017 asphaltenes and their TLC fractions. Deisy Giraldo-Dávila,† Martha L. Chacón−Patiño,†‡ Amy M. McKenna,‡ Cristian Blanco−Tirado,† and Marianny Y. Combariza†* †Escuela de Química, Universidad Industrial de Santander, Bucaramanga, Santander 680002, Colombia. ‡ National High Magnetic Field Laboratory, Florida State University, 1800 East Paul Dirac Dr., Tallahassee, Florida, USA 3231-4005, USA.

■ ABSTRACT Asphaltenes are polydisperse, compositionally complex solubility fractions of crude oil, with a range of structural motifs that contain two primary structural classes: molecules with single aromatic cores (island) and bridged aromatic moieties (archipelago), each with varying degrees of alkyl substitution (e.g., varied-length alkyl units, multi-heteroatom functionalities responsible for aggregate formation through weak and strong intermolecular associations. Understanding the link between asphaltene molecular composition and problematic behavior, such as deposit and emulsion formation, is vital to improve crude oils up- and down-stream processes in petrochemical industries.

Some reports suggest asphaltene precipitation occurs primarily

through self-association of polar functional groups, and highlight direct correlations between aggregate size and emulsion stability. Here, we compare the molecular composition of whole and TLC-fractioned Petrophase 2017 asphaltenes characterizd by Fourier transorm ion cyclotron resonance mass spectrometry (FT-ICR MS), and correlate composition to NIR scattering measurements of HepTol (Heptane:toluene) dispersions. Asphaltene retention on SiO2 plates is dictated by molecular features such as aromaticity, heteroatom content, and degree of saturation (or content of methylene units). Thus, the most polar compunds are retained on SiO2, and consist of compounds that have a bimodal

ACS Paragon Plus Environment

1

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 41

distribution in compositional space (carbn number and aromaticity) with highly alkyated, high heteroatomic (N,O,S) compounds In contrast, the least retained compounds consist primarily highly aromatic, sulfur-containing species with short alkyl chains, which indicates a clear relationship between fraction polarity (retention on silica). Furthermore, a correlation between asphaltene aggregation and emulsion stability was observed: the fractions that were retained the most on the silica plate, the quicker it aggregates in HepTol, and the more stable its emulsion. Collectively, this suggests that not all asphaltenes are the same, and indicates the presence of a direct relationship between asphaltene polarity, aggregation tendency and emulsion stability. ■ INTRODUCTION Adequate and effective optimization of refinery processes requires in depth knowledge of the effects heavy oil fractions such as asphaltenes on crude oil deposit and emulsion formation.1,2 The immense compositional complexity of asphaltenes varies between crude oils fromdifferent reservoirs due to variations in source rock temperature, reservoir depth and pressure, geographic location and geologic conditions.3–6 Asphaltenes are operationally-defined as the fraction of crude oil insoluble in paraffinic solvents (e.g., n-heptane, n-pentane) but soluble in aromatic solvents (e.g., benzene, toluene),7–9 and arguably the most polydisperse and compositionally complex organic mixture known. However, the primary factor that determines macroscopic behavior is molecular composition and not solubility.10 For more than five decades (1950-2000), asphaltene structure was thought to be dominated by at least fifteen fused aromatic rings with alkyl substitution (the island model) and molecular weights up to 100,000 g/mol, in contrast to many reports.11,12 However, the island model is not consistent with several asphaltene properties, such as heterogeneous aggregation and the molecular composition of the pyrolysis products.11,13–15

These discrepancies brought the archipelago model, which

suggests that asphaltene molecules contain several aromatic cores linked by aliphatic moieties


2

Page 3 of 41 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

(bridges).11,16,17 Today, the hypothesis about the coexistence of island and archipelago motifs is gaining strength and attributes several asphaltene physicochemical properties that are inconsistent based on the island model to the presence of bridged archipelago-type structures..17,18 Current molecular research converges on asphaltenes as a complex, polydisperse heteroatom-rich hydrocarbon mixture comprised of island and archipelago structural motifs,17,19– 21

including bridged aromatic moieties,20–22 single aromatic cores, with short and long alkyl units

and multi-heteroatom functionalities accessible for weak and strong intermolecular associations. Intermolecular associations include acid−base interactions, hydrogen bonding interactions (carboxylic acids and basic nitrogen), coordination to vanadium or nickel metal complexes and basic nitrogen, wax interactions of paraffinic side chains, and π−π stacking between aromatic cores.11,23 The synergy between these interactions suggests that asphaltene adsorption and aggregation occurs through a complex hierarchical structural model.24,25Asphaltene properties include strong aggregation tendencies,11,20,26 adsorption on mineral surfaces (e.g., silica, quartz, and alumina),12,27 occlusion inside asphaltene macrostructures,28,29 porosity of asphaltene aggregates,29 formation of films at oil−water interfaces,30 interactions with resins and surfactants,31 and contribution to various upgrading problems.11,16,19,32 Asphaltene monomers self-associate to form stable aggregates (e.g., dimers and trimers) in toluene at low concentration (1% (by weight) asphaltenes remains critical in production facilities. The reason for this behavior is simple: heavy oils usually contain more aromatics and resins, crude oil fractions that increase asphaltene solubility and prevent asphaltene precipitation.31,36,37 Light crudes with low asphaltene content are enriched in saturates, compounds that decrease asphaltene stability and can induce asphaltene precipitation.38 Asphaltene aggregation and precipitation mechanisms rely heavily on the molecular composition of the parent crude. Therefore, effective solutions that minimize asphaltene aggregation depend on compositional characterization at the molecular level. In the early 2000s, petroleomics emerged as a research field aimed to predict crude oil behavior through detailed molecular-level characterization of its composition. Application of ultrahigh resolution mass spectrometry (Fourier transform ion cyclotron resonance mass spectrometry (FT-ICR MS)) to separate and resolve the tens of thousands of peaks in a single crude oil provides elemental composition assignment, with hundred of mass spectral peaks within less than half a nominal mass unit, within a matter of minutes.39 Ultrahigh resolving power (m/∆m50% > 750,000 at m/z 400, in which m50% is the full mass spectral peak width at halfmaximum peak height) is critical for heavy oils and asphaltenes characterization, only achievable by FT-ICR MS, which routinely achieves resolving powers sufficient to separate isobaric overlaps prevalent in heavy oil compounds (e.g., 3.4 mDa and 1.1 mDa).8,40–42 Elemental compositions can be converted to double bond equivalents (DBE), the number of rings plus double bonds within the molecules (DBE = C- h/2 +n/2 + 1, calculated from elemental


4

Page 5 of 41 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

composition CcHhNnOoSs), a measure of aromaticity.43 DBE is inversely proportional to the number of hydrogens per molecule, and each additional ring or double bond reduces the number of hydrogen atoms by 2. Because each atomic isotope has a unique mass defect – the difference between nominal (nearest integer) mass and exact mass each unique composition of atoms (elemental composition) can be uniquely identified by mass. Moreover, the molecular mass defect can be used to differentiate between compounds of the same heteroatom class (NnOoSs) but different hydrogen deficiency.44 Therefore, examination of a single nominal mass derived from the FT-ICR mass spectrum of an asphaltene sample highlights compounds with a wide range of DBE values.44 Recent advancements in asphaltene characterization elucidate the importance of meticulous sample preparation. Asphaltenes are commonly obtained by SARA (Saturated, Aromatic, Resins, Asphaltenes) fractionation, where crude oils are separated into two major fractions: maltenes (soluble in n-alkanes (saturates, aromatics, resins)) and asphaltenes (insoluble

in

n-alkanes).8,9

Isolation

of

asphaltenes

occurs

through

flocculation

of

nanoaggregates and subsequent precipitation, which can cause occlusion of particular compounds (i.e., resins, petroporphyrins, steranes, C27-C31 hopanes and small PAHs (polycyclic aromatic hydrocarbons)).18,28,29,36 Therefore, comprehensive molecular characterization of asphaltenes requires.36,45 Implementation of detailed purification methods for asphaltenes and the non-destructive extraction of occluded compounds is essential to advance in asphaltene petroleomics. Previously, we reported a modified version of the standard ASTM D6560-12 method for C7 asphaltene isolation and purification.36 Here, heptane is added dropwise to crude oil under sonication to encourage slow flocculation and minimize co-precipitants. The initial precipitant is the typical asphaltene – bright shiny and black; however, the sample is subsequently pulverized and rewashed with hot heptane to extract the occluded compounds.36 Recently, Rogel et al.46 highlighted that sequential asphaltene cleaning with hot heptane does


5

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 41

not guarantee total extraction of occluded compounds, in agreement with our previous results. However, the amount of maltenic co-precipitates and occluded compounds decreases with extensive washing compared to the ASTM standard. These occluded maltene compounds are known to interfere with mass spectral analysis of asphaltene samples.33,36,44 During the Petrophase 2016 conference (Elsinore, Denmark), a group of scientists proposed the use of this modified asphaltene isolation methodology to produce a standard sample that could be distributed globally to researchers to study asphaltene composition and behavior collectively. The results of these studies were presented in a special session of Petrophase 2017 (Le Havre, France). A Middle Eastern heavy crude oil sample, provided by Total, was used as the asphaltene source and shipped to our laboratory for C7 asphaltene isolation and purification. The resulting fraction was called “Petrophase 2017 asphaltenes”. In this work, Petrophase 2017 asphaltenes were fractionated by high-performance thin layer chromatography (TLC) and subsequently characterized by FT-ICR mass spectrometry to correlate asphaltene adsorption, aggregation and emulsion stabilization to aromaticity, heteroatom content and degree of alkylation. TLC fractionation of asphaltenes produced three TLC fractions determined by retention on silica gel. Subsequent characterization by positive-ion atmospheric pressure photoionization (APPI) FT-ICR MS reveals that the most retained species on TLC plates consist of highly saturated, polar asphaltenes enriched with heteroatomcontaining groups, and low DBE values, in contrast to medium and low-retained species. NIR radiation scattering measurements of HepTol dispersions of each fraction and the whole asphaltene sample identifies a direct correlation between asphaltene adsorption on TLC silica plates and aggregation: The more the fraction is retained on the silica the faster it precipitates. In addition, emulsion stability tests suggest that the most polar asphaltene fraction produces the most stable emulsions. These findings suggest a clear and distinct relationship between asphaltene polarity, aggregation and emulsion stability.


6

Page 7 of 41 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

■ EXPERIMENTAL METHODS Sample Preparation. All solvents (toluene (Tol), n−heptane (n−C7), n−hexane (n−C6), dichloromethane (DCM), tetrahydrofuran (THF), N-methylpyrrolidone (NMP), water (H2O) and methanol (MeOH)) were high-performance liquid chromatography (HPLC)-grade from SigmaAldrich (St. Louis, MO, USA). Thin layer chromatography SiO2 plates 60 sorbent (particle size 9.5 - 11.5 µm) were purchased from Merck Millipore (Darmstadt, Germany). Asphaltene samples. Asphaltenes were isolated from a middle eastern heavy crude oil, provided by TOTAL (hereafter referred to as “Petrophase 2017 asphaltenes”) per a modified ASTM D6560−12. Briefly, asphaltenes were isolated at a crude oil/n-heptane ratio of 1/40; i.e. 400 mL of n-C7 was added dropwise (1 mL/minute) to 10 g of crude oil under sonication (Branson Ultrasonics, Danbury, CT, 22 kHz, and 130 W) at 60 ◦C. The mixture was refluxed for 90 minutes at 95 ◦C and allowed to stand for 12 hours. Precipitated asphaltenes were collected by gravity filtration (Whatman filter paper grade 42) and washed with hot n-heptane in a Soxhlet extractor (average temperature at the thimble 60 ◦C) until the solvent was clear (~72-150 hours). Solid asphaltenes were recovered by redissolution in hot toluene (~98 ◦C) followed by solvent evaporation under N2 flow. Asphaltene purification. Asphaltenes were purified to remove occluded compounds.36 Approximately 1 g of washed n-C7 asphaltenes from the previous step was crushed to increase surface area and placed in a Soxhlet apparatus with (~300 mL) n−heptane. After five hours, the n−heptane was removed and replaced with clean n-heptane, and solid asphaltenes pulverized and extracted with clean n-heptane (5 h), and repeated twice for 20 h total wash time until no significant amount material was extracted from the asphaltene sample (>0.2 mg/200 mL heptane). Asphaltene fractionation by preparative Thin Layer Chromatography. The TLC fractionation method was reported previously.12 Briefly, 2 µL of purified asphaltene stock


7

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 41

solution (5000 ppm in toluene) was seeded on the TLC SiO2 plates. Sequential elution of purified asphaltenes based on a series of elutropic solvents occurs with n-hexane, toluene, and 90:10 v/v of DCM/ MeOH. Fractions were collected based on mobile phase composition toluene-eluted (T), dichloromethane/methanol eluted (DM), non-eluted (NE), whole asphaltenes (W), and whole asphaltenes adsorbed on silica and desorbed without fractionation (WD). The TLC runs were repeated to isolate enough material for aggregation and stability tests and FTICR MS characterization. Optimized conditions for asphaltene recovery from silica plates have been previously reported,12 where the stationary phase that contains adsorbed asphaltene subfractions was removed from the glass plate and placed in a Soxhlet extractor and extracted for 24 hours with THF/MeOH (95/5 v/v) prior to desolvation under dry nitrogen (N2) and weighed. Positive-ion Atmospheric Pressure Photoionization. Whole Petrophase 2017 asphaltenes (W), whole desorbed asphaltenes (WD) and (3) TLC fractions were dissolved in toluene (300 µg/mL) and infused at 50 µL/min into a Thermo-Fisher Ion MaxxTM APPI source (Thermo-Fisher Scientific, Inc., San Jose, CA). Nebulization occurred at ~320 °C with N2 sheath gas (50 psi) and auxiliary gas (32 mL/min) to minimize sample oxidation. Gas-phase neutral flow out of the heated vaporizer in a confined jet and photoionization is initiated by a krypton vacuum ultraviolet gas discharge lamp (Syagen Technology, Inc., Tustin, CA) (10 eV and 10.6 eV photons, 120 nm) where photoionization occurs. Toluene increases the ionization efficiency for nonpolar aromatic compounds by means of dopant-assisted APPI43,47 through charge exchange and proton-transfer reactions between ionized toluene molecules and neutral analyte molecules at atmospheric pressure.48 9.4 T FT-ICR MS. Asphaltenes and fractions were analyzed with a custom-built Fourier transform ion cyclotron resonance mass spectrometer equipped with a 9.4 T horizontal 220 mm bore diameter superconducting magnet operated at room temperature (Oxford Corp., Oxney Mead, U.K.), and a modular ICR data acquisition system (Predator) facilitated instrument


8

Page 9 of 41 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

control, data acquisition, and data analysis.49,50 Ions generated at atmospheric pressure were externally accumulated in an external linear octopole trap with tilted wire extraction electrodes for 200-600 ms.51

Helium gas introduced into the octopole collisionally cools ions before

transfer through RF-only quadrupoles (total length 127 cm) equipped with an auxiliary rfwaveform52 into a 7-segment open cylindrical cell.53 Chirp excitation (70-720 kHz at a sweep rate of 50 Hz/µs and amplitude Vp-p of 1.0 V) excited ions to a detectable cyclotron radius. 200 time-domain acquisitions were co-added, Hanning-apodized, and zero-filled once before Fourier transform and magnitude calculation. Frequency was converted to m/z by the quadrupolar electric trapping potential approximation.54 Each m/z spectrum was internally calibrated with respect to a highly abundant homologous alkylation whose members differ in mass by an additional methylene unit (14.01565 Da), and confirmed by isotopic fine structure based on the “walking” calibration.55 Elemental formula assignments and data visualization were performed with PetroOrg N-13.3 software.56 Asphaltene Aggregation Stability. Asphaltene aggregation tests were performed using mixtures of n-heptane and toluene (HepTol, 70:30 v/v). Toluene solutions of whole asphaltenes, WD asphaltenes (resulting from desorbing unfractionated asphaltenes from the SiO2 plates), and TLC subfractions (T, DM, and NE) were prepared at concentration of 300 ppm with (HepTol, 70:30 v/v), vortexed at 2500 rpm during 5 seconds and placed immediately in the Turbiscan Lab instrument (Formulaction, L'Union, France). The aggregation behavior of each sample, understood as variations in NIR radiation backscattering and transmission behavior of the sample, was followed for 90 minutes, and ran at least per triplicate. Emulsion Stability test. Water-in-oil emulsions were prepared using 10 mg of each sample (whole asphaltenes and WD, NE, DM and T fractions) dissolved in 8 mL of a mixture of n-heptane, DCM and toluene (Hep:Tol:DCM 1:1:1 v/v). These solutions were mixed with 8 mL of deionized water, then vortexed at 1000 rpm over the course of 10 minutes and placed immediately in the Turbiscan Lab instrument (Formulaction, L'Union, France). These


9

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 41

measurements were followed for 90 minutes. This method is similar to that used by Clingenpeel et al.30 ■ RESULTS AND DISCUSSION TLC Fractionation. Asphaltene compounds have different retention on TLC plates due to intermolecular interactions with the SiO2 surface and the mobile phase.12,27,57 Functional groups on the silica surface include polar silanols (-OH), which retain polar asphaltenes through hydrogen bonding, and hydrophobic siloxanes (-O-Si-O-) that have little interaction with polar compounds. The high polarity of the silica surface is due to the high silanol content in vicinal, geminal or single configurations, and from SiOH---H2O silanol complexes present on hydrated silica.58 Active groups on the SiO2 surface can interact strongly with functional groups found in petroleum, e.g., carbonyls, fused aromatic rings, basic nitrogen (pyridine, amines), and metal complexes (vanadyl and nickel porphyrins).27,59 Figure 1 shows a schematic of preparative thin-layer chromatography fractionation of Petrophase 2017 asphaltenes on silica plates with an elutropic series of mobile phases to elute low polarity asphaltenes with hexane (n-C6), alkyl aromatics with toluene (T), polar compounds with dichloromethane/methanol (DM) (9/1 v/v). The total elution process yielded three distinct bands of (1) non-eluted compounds (NE), (2) compounds eluted with toluene (T), and (3) compounds eluted with dichloromethane/methanol (DM), which were subsequently extracted from the silica plate as described in the experimental section. No bands were observed with the elution with n-C6.


10

Page 11 of 41 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 1. Schematics of Petrophase 2017 asphaltenes fractionation by TLC. Mass Recovery. The total recovery and retention factors Rf for each fraction are shown in Table 1. The retention factor of a compound (or a group of compounds) is defined as the ratio of the distance the compound migrates from the seeding place on the plate, to the distance the solvent front moves from the same point to the end of the plate.

Asphaltenes are defined by insolubility in paraffinic solvents; therefore, elution with n-C6 (eluent strength (ε0) 0) does not promote analyte migration on the SiO2 surface.57,60 The addition of toluene (ε0 0.29) promotes asphaltene migration (≈19.3 wt%), yielding the fraction with the largest retention factor: Rf=0.94. High Rf corresponds to low affinity for SiO2 and high compatibility for eluent. Subsequent elution with DCM/MeOH (ε0 0.47) yields a medium affinity fraction (≈20.7 wt%0 with Rf = 0.81). Finally, asphaltene fractionation results in non-eluted compounds (Rf=0) that adhere strongly to the stationary phase. Gravimetric yields in Table 1 indicate high polarity for the Petrophase 2017 asphaltene due to 46.7±0.94 wt% in the non-


11

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 41

eluted fraction, which reveals a high concentration of compounds that interact strongly with polar silanols on the SiO2 surface. The total percent recovery for all fractions was 86%, which is not surprising as several reports suggest that some asphaltenic compounds irreversibly adsorb on polar stationary phases such as silica.61,62 Table 1. Total recovery and retention factors for DM, T and NE fractions of Petrophase 2017 asphaltenes. Fraction Mass Recovery wt% Rf T 19.3±1.4 0.94 DM 20.7±2.7 0.81 NE 46.7±1.9 0.00

Aggregation in HepTol and emulsion stability of Petrophase 2017 asphaltenes, and their TLC fractions, using NIR backscattering measurements. Asphaltene aggregation and contribution to emulsion formation in solution are complex phenomena that occur at the molecular level. However, from a macroscopic view, the formation of asphaltene aggregates or stable emulsions are both accompanied by variations in solution optical properties due to variations in the amount or size of dispersed particles,.63,64 which can be effectively monitored via backscattering and transmission measurements. The Turbiscan Lab instrument, for instance, monitors changes in the intensity of backscattered or transmitted radiation (NIR, λ 850 nm) vs. sample height as a function of time and yields kinetic dispersion stability data. Backscattering is commonly applied to study petroleum dispersionsand opaque fractions of crude oil such as asphaltenes.65–67 The Instability Index (equation 1) is a convenient way to follow the dynamics of aggregation processes and emulsion stability. This index is a statistical parameter derived from the analysis of variations in the backscattering intensity of the sample, relative to the original. The index is calculated as follows:

∑& ' | !" # !"$% | (

(Equation 1)


12

Page 13 of 41 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

Where BS+ is the dimensionless backscattering value for each scan and n is the number of scans. High values of Instability Index indicate a high probability of phase separation, which translates into either a high tendency to aggregate in the case of HepTol dispersions of asphaltenes, or low emulsion stability in the case of w/o emulsions where asphaltenes may act as interfacially-active compounds. Stability plots of HepTol dispersions of Petrophase 2017 asphaltenes and their TLC fractions. Figure 2a shows Instability Index vs. time plots for HepTol (70:30 v/v) dispersions (300 ppm) of whole Petrophase 2017 asphaltenes, desorbed asphaltenes (WD), and TLC fractions (T, DM, and NE). It is important to highlight that stability, in the HepTol aggregation tests, refers to lack of precipitation; a high value of Instability Index indicate low stability of the HepTol dispersion and proneness to precipitate. Along these lines, whole asphaltenes (W, black trace) present maximum Instability Index of 56, while dispersions of whole desorbed WD asphaltenes (gray) exhibit a maximum value of 20, indicating a 2.8-fold increase in dispersion stabilization due to the removal active compounds from whole asphaltenes through desorption from the silica (WD). On the other hand, non-eluted compounds (NE, blue trace) show an Instability Index value of 41, toluene eluted compounds (T, cyan trace) instability index of 5 and DM eluted compounds (DM, dark cyan trace) a value of 4. The low instability index profiles for toluene (T) and DCM/MeOH (DM) fractions, when compared with the whole sample (W), reveal that the species with a high probability for phase separation, the most problematic asphaltenes, are irreversibly adsorbed on the SiO2 surface. Extractable species off the SiO2 surface found in TLC fractions highlight the enrichment of problematic species in the non-eluted fraction (NE). Whereas dispersions of T (cyan) and DM (dark cyan) fractions are more stable. Aggregation stability tests performed on TLC fractions correlate polarity (retention on silica) with aggregation tendency: the more a fraction is retained


13

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 41

on the silica, the faster it aggregates. This behavior is also evident in the optical images of Figure 2a. Clearly, fractions DM and T (far right) do not aggregate in HepTol (70:30) after 90 minutes; in contrast, samples W, WD, and fraction NE show aggregate formation and precipitation. Kilpatrick and co-workers have made significant contributions to the role asphaltene molecular composition plays in aggregation.68–70 Through fractionation of asphaltenes based on differential precipitation, they conclude that not all asphaltenes and subfractions exhibit strong aggregation trends or follow the same aggregation mechanisms .71 Based on H/C, O/C, and N/C atomic ratios, the authors concluded that aggregation of asphaltene fractions from an Argentinian crude oil was driven by aromatic associations (π-stacking, in accordance with the Yen model)72 and dispersion interactions. Conversely, asphaltenes from off-shore Californian crude oils seemed to aggregate through hydrogen bonding between polar functionalities.71 In addition,, Kilpatrick et al.73 performed redissolution tests of the differential precipitation subfractions from off-shore Californian and Argentinian crude oils, and report that the most unstable asphaltene fraction is less soluble in methylene chloride compared to whole asphaltenes. The least soluble asphaltene fractions require increased solvent temperature to achieve total redissolution, which suggests hydrogen bonding as the primary molecular interaction driving the aggregation.


14

Page 15 of 41 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 2. Stability tests for Petrophase 2017 asphaltenes (W, WD) and TLC fractions (NE, DM, and T using NIR scattering measurements and optical images for a) HepTol (70:30) solutions and b) w/o emulsions at time 0 and after 1.5 h. c) Optical microscopy of emulsion interfaces. We hypothesize that out of the TLC fractions from the Petrophase 2017 asphaltenes in the whole desorbed sample (WD), the most stable asphaltene compositions (DM and T fractions) play an important role stabilizing the most “problematic or reactive” asphaltenes (NE fraction). Thus, T and DM asphaltenes cooperatively solvate the most polar and problematic


15

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 41

asphaltenes (NE species) in the whole, desorbed, unfractionated asphaltene sample by preventing hydrogen bonding and strong polar interactions. This could explain the difference in maximum instability index (up to two times lower than NE dispersions) between whole desorbed asphaltenes and NE dispersions (Figure 2a), in agreement with Fogler et al.74–78 Fogler and coworkers fractioned asphaltenes from stable and unstable crude oils and organic solid deposits by polarity based on differential precipitation in DCM/pentane mixtures,75 and report unstable crude oils that are enriched with high-polarity asphaltenic fractions exhibit a strong tendency to flocculate and cause deposition problems, and conclude that high polarity asphaltenes play a crucial role in crude oil stability.75 Stability Plots for synthetic emulsions of Petrophase 2017 asphaltenes and their TLC fractions. Figure 2b shows the Instability Index vs. time plots for synthetic emulsions prepared with Petrophase 2017 asphaltenes, desorbed asphaltenes (WD), and TLC fractions (T, DM, and NE). Emulsion formation and stabilization occurs when emulsifiers (compounds with the ability to lower the oil/water interfacial tension) or stabilizers (compounds restricting interfacial interactions) are present in the mixture of two immiscible liquids. Additionally, emulsion breakdown occurs through phenomena involving flocculation, creaming, sedimentation, coalescence or phase inversion.79–83 Active emulsifiers or stabilizers promote the formation of stable emulsions by disrupting emulsion breakdown processes. It is important to highlight that stability is defined as little or no phase separation in the context of the emulsification tests; a low Instability Index indicates highly stable emulsions.

Whole Petrophase 2017 asphaltenes (W, black) form the most stable emulsions (with no phase separation after 1.5 h) noted by the lowest Instability Index of 6. Removal of surfaceactive surface species in WD (gray) fraction through irreversible adsorption on the SiO2 surface results in a 3-fold decrease in emulsion stability (compared to the whole asphaltene) and an


16

Page 17 of 41 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

increase in the Instability index to 15 (less stable emulsion). WD exhibit decreased content of NnOo and NnOoSs compound families. Reports about characterization of interfacial material suggest that these compound families are surface active and contribute to emulsion stabilization.84 Similarly, the NE (blue) fraction form stable emulsions with Instability Index of 13. On the other hand, fractions DM (dark cyan) and T (cyan), with Instability Indexes of 60 and 53 respectively, produce the least stable emulsions of all fractions. Emulsion stability tests of TLC fractions correlate polarity (retention on silica) with interfacially active behavior: the more a fraction is retained on the silica, the more stable the emulsion. Similarly, TLC fraction aggregation tendency and interfacial activity are connected the faster a fraction aggregates in HepTol, the more stable the emulsion. This behavior is also evident in the optical images of Figure 2b. Clearly, fractions DM and T (far right) do not form stable emulsions, and after 90 minutes they quickly undergo phase separation (flocculation followed by coalescence). In contrast, samples W, WD, and fraction NE show stable emulsion formation, indicating the presence of interfacially active compounds. Using small-angle neutron scattering (SANS) Kilpatrick et al.69,73 suggested not only that asphaltene precipitation is driven by self-association, but also that there is a direct correlation between asphaltene aggregate’s size and emulsion stability. In turn, aggregate formation and the formation of cohesive films at the o/w interface are both phenomena likely driven by polar interactions, such as hydrogen bonding, between heteroatoms. Our observations are in agreement with previous reports.85 Microscopy images of material found in the emulsion interfacial layer corroborate emulsion stability behavior, (Figure 2c), which shows the presence of small water droplets surrounded by interfacially active material in emulsions prepared with W, WD and NE fractions. Conversely, emulsions prepared with fractions DM and T do not exhibit stable water droplets in the interfacial boundary. Small and abundant water droplets in the interfacial boundary of the WD, NE, and W w/o emulsions suggests that these samples are enriched with polar functionalities capable of interacting with H2O.86


17

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 41

Molecular Characterization by APPI FT-ICR MS. Heteroatom Class Distribution. Petrophase 2017 asphaltenes versus whole desorbed (WD) asphaltenes. Figure 3 shows the heteroatom class distribution for all detected species of >0.25% relative abundance in the APPI FT-ICR mass spectra for the asphaltenes, desorbed asphaltenes (WD) and TLC fractions (NE, T, DM). The whole Petrophase 2017 asphaltene sample is depleted of HC (hydrocarbons without heteroatoms), monoheteroatomic N- and S-containing classes (i.e., S1, N1) and polyheteroatomic compounds (i.e. NnOo, NnSs and NnOoSs), while the whole desorbed fraction (WD) is enriched with hydrocarbons without heteroatoms and monoheteroatomic S1-, Oo- and N1- compounds. Heteroatom Class Distribution. Petrophase 2017 asphaltenes TLC fractions. Figure 3b shows the heteroatom class distribution for non-eluted (NE), toluene-eluted (T) and DCM/MeOH (DM)-eluted compounds. Oxygen- and nitrogen-containing species (N1O1, N1O2, N2O1, N1O1S1, N1O2S1, N1O1S2 and N2O1S1) are the most abundant classes in the NE fraction and indicate that more polar compounds remain adsorbed onto SiO2, in agreement with previous reports.61 The presence of these compounds seem to drive the strong aggregation tendency and emulsion stabilization effects of the NE fraction. DM-eluted compounds are enriched in vanadyl porphyrins (N4O1V1), followed by sulfur-containing classes (e.g., O1S1, S1, S2 and S2O1) and nitrogen classes (N1, N1O1, and N2) and oxygen classes (O1 and O2). Tolueneeluted (T) compounds exhibit high abundance of S-containing classes such as S1-S3, N1, and hydrocarbons without heteroatoms, similar to whole heavy crudes.87 These observations indicate that sulfur, highly concentrated in T fraction, does not play an important role on asphaltene adsorption on the polar SiO2 surface. Hence, S does not seem to participate in asphaltene aggregation or influence emulsion stability. More importantly, our results indicate that vanadyl-porphyrins, mainly concentrated in the DM fraction, do not extensively participate in


18

Page 19 of 41 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

emulsion stabilization and precipitation.88–90 However, it is important to highlight that our conclusions are drawn from one particular sample, the Petrophase 2017 asphaltenes.

Figure 3. a) Compound class distribution in whole Petrophase 2017 asphaltenes and WD fraction and b) Compound class distribution in TLC fractions (NE, DM, and T). Compositional Space of Petrophase 2017 asphaltenes, whole desorbed (WD) asphaltenes, and TLC fractions.


19

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 41

Representation of molecular formulas in isoabundance contour plots of double bond equivalents (DBE) versus carbon number, for a given heteroatom class, are useful to understand compositional differences between samples.16,17,36,40,44

Figure 4 shows that

compounds in Petrophase 2017 asphaltenes exhibit wide DBE distributions, starting at DBE ~4 and extending up to DBE ~33, and broader distributions of carbon numbers, which span between C# ~20-60. WD compounds cover a similar compositional space as the whole sample (same DBE range) but over a narrower carbon number range. However, ample evidence suggests molecular composition by HR-MS of whole (non-fractionated) petroleum-derived samples does not always reveal useful information. Several reports by Rodgers,17 Rowland,91 Rogel,92 Combariza36,93 and Wittrig94 indicate the need for sample fractionation to improve compositional space accessibility in mass spectrometric analysis of complex petroleum samples. Asphaltenes are not the exception; reports by McKenna and Rodgers33,44 indicate that the characterization of whole asphaltenes by MS is incomplete because of the ultra-complexity of the sample and the preferential aggregation of specific subfractions. McKenna reported that asphaltenic samples could present more than 50,000 unique elemental formulas and that the most aromatic species are always preferentially ionized in APPI.44

Thus, asphaltene

fractionation using chromatographic separations with subsequent analysis by ultrahigh resolution mass spectrometry is required to improve the molecular characterization of these complex samples. Figure 4 shows DBE vs carbon number plots for the hydrocarbon (HC), S1, O1 and O1S1for TLC fractions of Petrophase 2017 asphaltenes. The toluene fraction is enriched with species that correspond to compositional space previously reported for asphaltenes.44,92,95,96 Specifically, the fraction T exhibits a high relative abundance of compounds with DBE values between 15 and 30, grouped at the planar limit area. The average homologous series length of the T fraction is ~18, which indicates around 25 methylene units attached to the aromatic cores. In other words, toluene eluted compositions consist mainly of alkyl-aromatics enriched in S-


20

Page 21 of 41 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

functionalities with compound families such as S1, S2, and S3. Figure 5 indicate that this fraction exhibits abundant compounds with low H:C ratios (0.5

Correlations between Molecular Composition and Adsorption

Recommend Documents