•Stepwise structural characterization of polar compounds during hydrotreatment process.•Average hydrocarbon structure of feedstock and hydrotreating products by NMR.•The stable and unstable type of polar compounds during hydrotreatment.A continuous fixed-bed micro-reactor was adopted to conduct a hydrotreatment experiment on the atmospheric residue obtained from Saudi Arabia. The feedstock and hydrotreated liquid products were subjected to high-temperature gas chromatography to undertake simulated distillation analysis, and to gel permeation chromatography to carry out molecular size investigations. 1H NMR and element analysis were carried out to derive structural parameters. Positive- and negative-ion electrospray ionization Fourier transform-ion cyclotron resonance mass spectrometry was combined with other analytical means to explore the stepwise structural characterization of polar components (especially the nitrogen (N)-containing compounds) when passing through beds filled with different catalysts during the hydrotreatment process. According to the different numbers of nitrogen, oxygen, and sulfur atoms, N-containing compounds could be divided into Nx, NxOy, and NxSy class species. The results indicated that hydrotreatment process led to more concentrated distribution of hydrocarbon molecules in the products. The heteroatoms, which accounted for the largest proportion in the feedstock and liquid products, were N1 class species. During the hydrotreatment process, the relative abundance of N1, N2, N1O1, N1O2, and N1S1 class species changed significantly with the progress of hydrotreatment process, however the conversion regularity varied among heteroatomic compounds.

•A novel characterization of petroleum fraction based on GC/MS and HNMR was stated.•DMU model was applied based on new method and the prediction showed good results.•The optimum condition of PCA removal from furfural-extract oil was investigated.Solvent extraction is highly efficient process for removing the toxic polycyclic aromatic hydrocarbons (PaHs) from vacuum gas oil (VGO) to produce eco-friendly processing oil. In this work, we established a model which can predict both the PCA content and hydrocarbon composition of extraction product. Commercial furfural-extract oil (FEO) from vacuum gas oil was used to obtain liquid–liquid equilibrium data using solvent extraction under the condition of extraction temperature 323 K–353 K, solvent–oil ratio 1:1–4:1. The effect on the PaHs distribution and polycyclic aromatics (PCA) content was studied. A novel characterization was stated by dividing the oil into eight model-molecules based on the GC-FI TOF/MS and 1H NMR data. The construction of eight molecules including paraffin, naphthene, one to five ring aromatics and sulfur containing polycyclic aromatics was determined by hydrocarbon composition and the H atom distribution, and the UNIFAC groups could be calculated directly. The PCA content was also correlated well with hydrocarbon compositions. This model has advantage of using only one set of adjustable interaction parameter between furfural group and sulfur-containing group. The modeling results showed the calculated yields, hydrocarbon composition and PCA content fitted well with the experimental data, indicating that the model with new characterization method has some predictive ability on the solvent extraction process in this work.

Atmospheric residue (AR) from Saudi Arabia (SZAR) was subjected to supercritical fluid extraction fractionation (SFEF) and four extractable subfractions (SFEF-1–4) and an unextractable end-cut were obtained. SZAR and SFEF-1–4 were subjected to hydrotreatment (HDT) in a continuous mini fixed-bed reactor. Three commercial catalysts were used to remove the impurity in the feedstock. The composition and structural transformation of nitrogen (N)-containing compounds were investigated. Electrospray ionization (ESI) Fourier transform–ion cyclotron resonance mass spectrometry (FT-ICR MS) was used for molecular characterization of N-containing compounds of SZAR, extractable SFEF subfractions, and HDT products. Results showed that N1 class species with high aromaticity and/or low carbon number exhibited higher catalytic hydrogenation reactivity. N1 class species with lower aromaticity were removed regardless of chain length.

The supercritical fluid extraction and fractionation (SFEF) method has the capacity of separating heavy petroleum residue into fractions with different processabilities. Different from traditional heavy residue feedstock, slurry oil (SLO) is a heavy liquid product from the fluid catalytic cracking (FCC) process, and there is concern for its potential application in producing carbonaceous materials. The high-value utilization of FCC SLO depends upon the aromatic and contaminate contents in the feedstock. The present study aimed to extend the SFEF method to the FCC SLO system and investigate the property and molecular composition distribution of derived fractions. A FCC SLO was subject to the SFEF process and separated into 13 extract fractions with each 5 wt % yield and 1 unextracted end cut. Systematic bulk property measurements, including boiling point distribution, density, molecular weight, viscosity, and carbon residue, were performed on each fraction. Hydrocarbon molecules in different fractions were characterized in terms of hydrocarbon types and molecular composition by gas chromatography–mass spectrometry (GC–MS) and high-resolution mass spectrometry, respectively. After separation, allof the asphaltenes were removed from extracts. With an increasing extraction pressure, the derived fraction has a growing boiling point, aromaticity, and polarity. Hydrocarbon-type analysis shows that aromatics predominated in the feedstock and most of the fractions. The aromatic content gradually increases with the extracted pressure. The high-resolution MS result shows that SLO is a low-molecular-weight sample with a high aromaticity and short side chain. The measured aromaticity value and GC–MS-measured ring number show a similar increasing trend with the growing extracted pressure. The lower phenanthrene/pyrene ratio indicates that heavier SFEF cuts have potential to have good mesophase development performance.

Atmospheric residue from Saudi Arabia (SZAR) was subjected to supercritical fluid extraction fractionation (SFEF) experiments. Four extractable subfractions (SFEF 1–4) and an unextractable end-cut were obtained. SFEF 1–4 were subjected to a hydrotreating (HDT) test in a mini-tubular reactor packed with commercial hydrodemetallization (HDM) and hydrodesulfurization (HDS) catalysts in order to determine the structural composition transformation of sulfur-containing compounds. Experiments were undertaken at a temperature of 360 °C, a hydrogen pressure of 14.7 MPa, a liquid hour space velocity (LHSV) of 0.25 h–1, and a H2/feed ratio of 650:1 (m3/m3). The SFEF subfractions and their products were subjected to analysis of their sulfur and nitrogen contents, using an ANTEK 7000S analyzer, and their saturates, aromatics, resins, and asphaltenes (SARA) compositions were analyzed using open column liquid chromatography. The detailed molecular composition of sulfur heteroatom species was determined by methylation, followed by positive-ion electrospray ionization Fourier transform ion cyclotron resonance mass spectrometry (ESI FT-ICR MS). In ESI FT-ICR analysis, the sulfur-containing compounds were characterized in terms of class, type (double bond equivalence, DBE), carbon number distribution, and relative abundance. According to the different number of sulfur atoms, sulfur-containing compounds can be divided into S1 and S2 class species. The results show that, depending on different transformation mechanisms, S1 class species, which have less-aromatic cores, are easy to convert, no matter the side chain length. In addition, S2 class species with less-aromatic cores and/or less carbon numbers have better hydrogenation reactivities. Meanwhile, S2 class species with long and/or complex chains would be easy to cleavage into small ones under the reaction conditions. Compare to S1 class species, S2 species are more active. The potential transformation of sulfur compounds in HDT of subfractions of atmospheric residua were inferred from the DBE distribution and carbon number data by involving SFEF, methylation followed by FT-ICR MS for sulfur species of SFEF subfractions and their HDT products. The results may help understanding of the heavy oil sulfur removal sequence and catalyst/process optimization for heavy oil desulfurization and thereby assist the efficient production of clean transporation fuels.

Solvent deasphalting is a highly efficient process for removing asphaltene and metals from the petroleum residue; however, there are rare modeling studies about the interaction between metals and asphaltene. This work establishes the association models of metals (Ni and V) with asphaltene during solvent deasphalting. The models reveal that the distribution factors of metals are related to the solvent density, association enthalpy, and temperature. The association energy of metal–asphaltene is higher than that of metal–resin. The distribution factor of metals with asphaltene is lower than that without asphaltene. The tuned model could well-predict the distribution factors for both Ni and V. The models provide the thermodynamics and theoretical guide for removing the metals from the heavy oil during solvent deasphalting.

A novel methodology was extended for modeling the detailed composition of petroleum heavy vacuum resid fractions. The resid molecules were organized in terms of basic structural attributes: cores, intercore linkages, and side chains. The identities of the structural attributes were determined both from the extrapolation of chemical characteristics of light petroleum and the analysis of detailed mass spectrometric measurement of heavy resid fragmentation products. A building block library was constructed containing ∼600 attributes. The molecular composition was constructed by the combination of attributes, or building blocks, into discrete molecules. The quantitative abundance of each molecule was determined by the juxtaposition of a set of structural attribute probability density functions (PDFs) constraining pure hydrocarbon and heteroatom mixtures. Quantitative structure–property relationships (QSPRs) were applied to calculate the bulk properties of both the constructed molecules and the mixture. The adjustable parameters of the PDFs were determined using an optimization loop that employed an objective function that contained a term for each of the available analytical data points. The resulting optimal molecular compositions were in good agreement with the experimental structural information.

The hindered stepwise aggregation (HSA) model was used to elucidate the molecular aggregation in heavy petroleum fractions which were derived from supercritical fluid extraction fractionation (SFEF) of Venezuela Orinoco vacuum residue (VR). The SFEF fractions consisted of multiple extractable narrow fractions and a nonextractable end-cut. The SFEF fractions were diluted with toluene, and their number-average molecular weights (MWs) were determined using vapor pressure osmometry (VPO). The initial molecular association constants (K1) and aggregation hindrance factors (H) of the HSA model for each SFEF fraction were calculated from the VPO MWs at various SFEF solution concentrations. The results showed that the HSA model fit well with VPO MW data and the parameters of the HSA model are physically significant. The values of MW and K1 increased as the SFEF fraction became heavier. The SFEF end-cut had the highest K1 and lowest H value, in which the aggregates were 2 to 8 monomers. Except for the initial fraction, all the SFEF fractions formed aggregates at solution concentrations higher than 30 g/L. The value of K1 was dependent on the number of aromatic rings, whereas H is dependent on the size of aromatic ring and side-chain length. The VPO MWs of light SFEF fractions were in agreement with those determined from electrospray ionization (ESI) mass spectrometry (MS) or gel permission chromatography (GPC). The VPO MWs of the highly aggregated SFEF fractions were higher than those of ESI MS due to low ionization efficiency but were much lower than those of GPC.

•Systematic test of apodization on both simulated and actual complex mixture FT-ICR signals.•Effect of apodization on 13C isotope ratio is negligible.•Complex FT-ICR analysis should sacrifice dynamic range to increase resolving power.•Low dynamic window functions are recommended for less resolving power reduction while ensuring correct peak selection.Apodization is a generally applied signal processing method in FT-ICR MS (Fourier transform ion cyclotron resonance mass spectrometry) analysis for complex petroleum samples that is used to lower the spectral leakage and smoothen the spectrum line shape. Present study evaluates the effect of apodization on the FT-ICR MS results through a systematic examination on both simulated and actual heavy petroleum signals. The result shows that the resolving power, instead of the dynamic range, is the key factor for heavy petroleum FT-ICR MS analysis due to the severe peak overlapping. Window functions with high dynamic range will reduce the resolving power so that many of the species cannot be detected. The improvement on accuracy of apodization with peak intensity is negligible and all the window functions have similar intensity error. Low dynamic range window functions perform better for FT-ICR MS analysis since they have less resolving power reduction and the spectral leakage is eliminated to ensure correct peak selection. Sine-bell and low beta value Kaiser window functions are recommended for complex mixtures.

The building of a separation-reaction network for heavy feedstock requires an understanding of the reactivity for subfractions. This paper proposes a study on the coking reactivity for two heavy petroleum samples, vacuum residua (VR) from China Liaohe and Venezuela Orinoco crude oils, and their supercritical fluid extraction fractionation (SFEF) subfractions. The properties of feedstocks and their SFEF series were analyzed, including density, molecular weight, elemental content, Conradson carbon residue (CCR), SARA components (saturates, aromatics, resins and asphaltenes) and structure parameters by nuclear magnetic resonance (NMR). All the samples were subjected to a laboratory-scale batch reactor to investigate the coking reactivity. Coke yield increased as the SFEF subfraction became heavier; while the yields of naphtha, diesel, and gas oil all decreased. The results show that the classic linear correlation between coke yield and CCR are not appropriate for separated fractions of VR. Instead, a negative power function was observed. Furthermore, the coke yield shows a linear relationship with aromaticity for both SFEF series. To precisely estimate coking reactivity for feedstocks and subfractions, a new SARA component based prediction model was carried out, which reveals the contribution of each component.

Thermal cracking of Venezuela vacuum residue (VR) was carried out in a batch reactor system. The thermally cracked resid was diluted with toluene and was subject to antisolvent (SAS) separation using n-pentane at the liquid and supercritical state in a continuous SAS apparatus. The SAS-derived liquid and solid products were obtained and subjected to elemental and saturates/aromatics/resins/asphaltenes (SARA) analyses. The SAS liquid products are comprised predominantly of saturates and aromatics, whereas SAS solids consist of asphaltenes and toluene insolubles (TIs). The resin content of SAS liquid and solid products varies with the thermal cracking severity. As the thermal cracking severity increases, the yield of SAS liquid decreases. The saturate content in the SAS liquid increases at the expense of resin content. The TI content of SAS solids increases as the thermal severity increases. The sulfur, nitrogen, and Conradson carbon residue (CCR) contents of SAS liquid increase and the hydrogen-to-carbon (H/C) ratio and metal (nickel and vanadium) contents of SAS liquid decrease with thermal cracking severity. Physical characterization of SAS solids was conducted, indicating that the SAS solids consist of submicrometer amorphous granules when the SAS process was operated at low temperatures but form aggregates at high SAS process temperatures. As the resid concentration of SAS feed decreases, the quality of SAS liquid decreases, and the size of the SAS granules decreases. The results from this work showed that it is viable to use the SAS process to separate the heavy fraction from the thermally cracked resid.

Venezuela Orinoco heavy crude oil was fractionated into diesel, vacuum gas oil (VGO), vacuum residue (VR), and asphaltene fractions, which were subjected to molecular weight (MW) measurement by vapor pressure osmometry (VPO). The VPO is known to overestimate the average molecular weight (MW) of heavy hydrocarbons, because of molecular aggregation. This paper proposes a hindered stepwise aggregation (HSA) model to simulate the molecular aggregation and used the model to estimate the true MW of heavy petroleum fractions. A data regression procedure was developed to determine the model parameters, aggregation equilibrium constant, and aggregate distribution, using a fast simulated annealing (FSA) algorithm based on the VPO data. This data analysis method is self-tuned to fit the VPO data to the HSA models of various petroleum fractions using the optimized solution of the FSA algorithm. The results showed that the VPO data of heavy petroleum fractions at various solution concentrations were in good agreement with those predicted by the HSA model. The aggregation equilibrium constant and aggregate distribution data obtained from the HSA model suggested that various degrees of molecular aggregation occur in heavy petroleum fractions. The molecules of diesel and VGO were monomers, regardless of the solution concentration. The molecules of VR formed dimer aggregates at high solution concentrations; the number of dimer aggregates exceeded that of monomers as the solution concentration increased. The molecules of asphaltenes were polymer aggregates. The size of asphaltene polymer aggregates increased significantly with the solution concentration. The MW of asphaltenes determined by the HSA model was much lower than that by the conventional linear regression method.

Negative-ion electrospray ionization (ESI) enabled the direct mass spectrographic analysis of phenols, naphthenic acids, and neutral nitrogen compounds in petroleum fractions without prefractionation; however, ESI results provide few quantitative and structural information about the analytes: the composition of acidic compounds in heavy oil, such as distillate resid, is still unclear. In this study, extrography was used to fractionate oilsands bitumen-derived vacuum-topped bitumen (VTB) and its maltene and asphaltene fractions into multiple subfractions. The molecular compositions of acidic functional compounds in the VTB and its subfractions were analyzed by negative-ion ESI Fourier transform ion cyclotron resonance mass spectrometry (FT-ICR MS). Extrographic separation was a necessary step to isolate acidic compounds of various acidities and/or polarities in vacuum resid fractions to achieve a systematic analysis. The results showed that the O2 class species in VTB were highly condensed phenols and carboxylic acids. The maltene fraction contained most of the less condensed naphthenic acids, whereas the asphaltene fraction contained highly condensed carboxylic acids and phenolic compounds with a bouble-bond equivalent (DBE) higher than 6. The presence of acids had no significant impact on the yield of asphaltenes in n-C7 solvent precipitation. Acid-free asphaltene fractions, which account for more than 90 wt % of the asphaltenes, cannot be ionized by negative-ion ESI.

The interaction between vanadyl porphyrins and asphaltene in alkane solvents is important to increase the removal rate of metals during the solvent deasphalting process. Thus, the adsorption kinetics and thermodynamics of vanadyl etioporphyrins on a Canadian oil sands bitumen vacuum tower bottom (VTB) asphaltene in n-pentane were investigated. After adsorption, asphaltene was analyzed via transmission electron microscopy (TEM), Brunauer–Emmett–Teller (BET), and Fourier transform ion cyclotron resonance mass spectrometry (FT-ICR MS). A certain amount of vanadyl porphyrins was adsorbed by the VTB asphaltene. This adsorption process was affected by the asphaltene dosage (0.01 and 0.02 g), the concentration of the n-pentane solution containing vanadyl porphyrins (10 and 15 μg/mL), and the temperature (288, 293, and 298 K). The adsorption rate was initially distinctly high. However, this rate became much slower after around 300 min, until equilibrium was reached after 1800 min. A comparison of four kinetic models of the overall adsorption rate showed that the adsorption process can be well-described by a pseudo-first-order equation. Furthermore, the adsorption equilibrium fit the Freundlich isotherm. The ΔG° and ΔH° values of the adsorption process between vanadyl porphyrins and asphaltenes had been regressed at different temperatures. The absolute value of ΔG° was less than 20 kJ/mol, whereas that of ΔH° was greater than 40 kJ/mol.

A vacuum topped Canadian oilsands bitumen (VTB) was subjected to solvent precipitation and subsequently characterized by elemental analysis, gel permeation chromatograph (GPC), 1H-NMR spectroscopy and negative-ion electrospray ionization (ESI) Fourier transform ion cyclotron resonance mass spectrometry (FT-ICR MS). Effects of experimental conditions such as solvent types (n-C5, n-C6, and n-C7), solvent purity, and solvent washing time on asphaltenes yields, bulk composition, and molecular composition of detectable heteroatom compounds in ESI source were determined. Elemental nitrogen and sulfur were enriched in asphaltenes while elemental oxygen had comparable content in maltenes and asphaltenes. Molecular composition of asphaltenes varies with separation conditions. The N1 and O1 species identified by ESI FT-ICR MS were enriched in maltenes. The O2 species exhibited two different double bond equivalents (DBE) distributions and solubility in normal paraffin solvents, indicating two types of molecular structures. Multi oxygen atom containing compounds mainly detected in asphaltenes. Compound class distributions are similar for maltenes derived from n-C5, n-C6, and n-C7, as well as for asphaltenes. The cyclic paraffin impurities in normal paraffin solvents had a significant influence on asphaltenes yields and heteroatom molecular composition. A portion of neutral N1 species and acidic O2 species adsorbed on asphaltenes could be dissolved by increasing washing time. Cautions should be exercised when interpreting the properties and composition of asphaltenes obtained with different experimental conditions.

A novel technique was developed for characterization of saturated hydrocarbons. Linear alkanes were selectively oxidized to ketones by ruthenium ion catalyzed oxidation (RICO). Branched and cyclic alkanes were oxidized to alcohols and ketones. The ketones were then reduced to alcohols by lithium aluminum hydride (LiAlH4). The monohydric alcohols (O1) in the products obtained from the RICO and RICO-LiAlH4 reduction reactions were characterized using negative-ion electrospray ionization (ESI) Fourier transform ion cyclotron resonance mass spectrometry (FTICR MS) for identification of iso-paraffins, acyclic paraffins and cyclic paraffins. Various model saturated compounds were used to determine the RICO reaction and ionization selectivity. The results from the FTICR MS analysis on the petroleum distillates derived saturated fraction were in agreement with those from field ionization gas chromatography time-of-flight mass spectrometry (FI GC-TOF MS) analysis. The technique was also used to characterize a petroleum vacuum residue (VR) derived saturates. The results showed that the saturated molecules in the VR contained up to 11 cyclic rings, and the maximum carbon number was up to 92.

Granulation is an effectual method to deal with huge amounts of deoiled asphalt (DOA) produced in the residue selective asphaltene extraction (SELEX-Asp) process, and the softening point of DOA plays an important role in asphalt particle production. In this paper, the variation of the softening point under different operation conditions and its influence on asphalt granulation behavior were studied with various solvents and feedstocks. Moreover, properties and structures of five C7 asphaltenes were analyzed with 13C nuclear magnetic resonance (NMR) and X-ray diffraction (XRD) measurements to examine the reasons of asphalt softening point differences for various residues under the same operation conditions. Ordered degree index (OI) and softening point index (SI) were proposed to investigate the influence of asphalt ordered aromatic sheets on the softening point. High softening point asphalt could be obtained with lower temperatures and higher pressures of the extractor, a larger mass ratio of solvent to feed, and a heavier solvent. Asphalt tends to form particles with the increase of the softening point, and nice particles would be produced when the softening point is higher than 170 °C. OI of C7 asphaltene is an important factor affecting the asphalt softening point. A close correlation between SI and the softening point was established; that is, feedstocks with a high SI would have the ability to produce high softening point asphalt and possess a good asphalt granulation behavior. The quantity of ordered aromatic sheets in asphalt is the fundamental determinant of the softening point.

A Chinese crude oil was distilled into multiple narrow boiling fractions. The crude oil, 39 narrow distillate fractions (up to 560 °C), and atmospheric and vacuum residues were analyzed using negative electrospray ionization Fourier transform ion cyclotron resonance mass spectrometry (ESI FT-ICR MS). The heteroatoms, N1, N2, N1O1, N1O2, O1, and O2 class species, were identified and characterized by double-bond equivalent (DBE) values and carbon numbers. The composition of crude oil was correlated with increased boiling point. Most abundant O1 and O2 class species had DBE values and carbon numbers corresponding to biological skeleton structures, such as hopanoic acid, secohopanoic acid, and sterol. The distribution of acids and neutral nitrogen compounds in the various fractions were determined. At higher carbon numbers, the amount of the compounds and DBE values increased gradually with the boiling point for most oil fractions. The abundant N1 class species were centered at DBE values of 9, 12, 15, and 18. These were likely pyrrolic compounds with various numbers of aromatic rings. Species such as hopanoic acids and secohopanoic acids were highly abundant in fractions above 500 °C. Sterol-like compounds were enriched in the 460−500 °C fractions. These are likely the major species causing a high total acid number (TAN) in the crude oil.

The relationship between the structure and stability of the visbreaking product of Venezuela extra heavy oil has been studied. The number of resin layers adsorbed to asphaltene is proposed to describe the colloidal stability of Venezuela heavy oil and its visbreaking product. The results show that the colloidal stability of the visbreaking product is inferior to Venezuela heavy oil. The factors influencing the storage stability of the visbreaking product have been studied by analyzing the changes of kinematic viscosity, total acid number (TAN), weight loss, ratio of resins/asphaltenes, as well as free-radical content of visbreaking samples sealed and exposed to air or nitrogen for 90 days. It is found that the storage stability of the visbreaking product is influenced by oxidative condensation, evaporation, and non-oxygen condensation, among which oxidative condensation is the most significant factor. The free-radical reactions do not terminate after the thermal reaction, and the content of free radicals is reduced to different degrees for the storage cases. Antioxidants 2,6-di-tert-butyl-4-methylphenol and p,p-di-iso-octyl-diphenylamine were added to visbreaking products by 1 wt % to prevent oxidative condensation. When the TAN and content of free radicals of visbreaking samples with and without antioxidant are compared, it is demonstrated that the antioxidants play the role of generating stable free radicals, which favors the retention of stability.

Narrow fractions of Athabasca vacuum topped bitumen (VTB) were prepared by supercritical fluid extraction and fractionation (SFEF) and characterized. Phase equilibria of various narrow fractions in the propane system were determined using a high-pressure PVT unit, operating at 30−50 °C, 4−6 MPa, and solvent/oil (S/O) ratio of 1.5 to 5 (wt/wt). Solubility parameters were determined using the activity coefficient equation from Scathard−Hildebrand’s regular solution theory. The value of the solubility parameter of the VTB fraction was obtained by extrapolation of the Gaussian function at the S/O ratio approaching zero. It was found that the solubility parameter of the VTB fraction was not constant, varying from 14 to 16 MPa1/2. The solubility parameter exhibited a Gaussian distribution as a function of the S/O ratio, with a maximum at the S/O ratio of 3.25, and a parabolic function of pressure, with a maximum at 5 MPa. It decreased with temperature. As the VTB fraction becomes heavier, the solubility parameter of the VTB fraction and the maximum value of the solubility parameter increased. The solubility parameter of the VTB end-cut (primarily C5-insoluble asphaltenes) was obtained using a modified titration method by varying the composition of a pentane/toluene blend solvent. The correlation of the solubility parameter of the VTB fraction was derived as a function of various key feedstock properties. The measured solubility parameters of narrow VTB fractions were compared to predictions from five published models. The results showed that the published models overestimate solubility parameters of the VTB fractions. In some cases, the trend of predictions was inconsistent.