Molecular maltenes

Colloidal State. The principal outcome of many of the composition studies has been the delineation of the asphalt system as a colloidal system at ambient or normal service conditions. This particular concept was proposed in 1924 and described the system as an oil medium in which the asphaltene fraction was dispersed. The transition from a coUoid to a Newtonian Hquid is dependent on temperature, hardness, shear rate, chemical nature, etc. At normal service temperatures asphalt is viscoelastic, and viscous at higher temperatures. The disperse phase is a micelle composed of the molecular species that make up the asphaltenes and the higher molecular weight aromatic components of the petrolenes or the maltenes (ie, the nonasphaltene components). Complete peptization of the micelle seems probable if the system contains sufficient aromatic constituents, in relation to the concentration of asphaltenes, to allow the asphaltenes to remain in the dispersed phase. 

A significant portion of the sulfur- and nitrogen-containing species in crude oil can be found in heterocyclic form within the asphaltene, maltene, and resin compounds. Oxygen-containing heterocycles may also be present. Examples of high-molecular-weight aromatic, resinous, and polar compounds found in crude oil are provided in TABLE 3-1. 

Asphaltenes These are complex high-molecular-weight polycyclic aromatic compounds which may contain oxygen, sulfur, or nitrogen heteroatoms. They are found in crude oil and in certain heavy fuel oils in micellar form. Their dispersion throughout oil can be stabilized by asphaltene precursors called resins and maltenes. 

Boduszynski et al. (1980) also employed the more conventional separation procedure based on solubility properties (Corbett, 1969) to provide asphaltene and maltene samples from the 675°C+ residuum. Asphaltenes are isolated by precipitation in an alkane solvent, with further separation of maltenes by chromatography in solvents of increasing elution strength. The FIMS results in Fig. 9 illustrate, significantly, that asphaltenes are not necessarily the highest-molecular-weight components in residuum. Asphaltenes have, rather, a relatively low but broad distribution of molecular weights. 

Asphaltenes may contain both porphyrin and nonporphyrin metals, depending upon the origin of the crude oil. Yen et al. (1969) characterized the vanadium complexes in a petroleum asphaltene by mass spectroscopy, optical spectroscopy, and ESR. Porphyrins (Etio and DPEP), acid-resistant porphyrin macrocycles of increased aromaticity (Rhodo), and nonporphyrins with mixed donor complexes were identified. Baker (1966) and Baker et al. (1967) extracted porphyrins from Boscan crude oil asphaltenes and also found Etio and DPEP as the two major porphyrin series. These homologous series range in molecular weight by 7 to 18 methylene groups. Gallegos (1967) observed by mass spectroscopy that asphaltenes and maltenes from a Boscan crude oil had nearly identical porphyrins in terms of mass distribution. 

Percent Vanadium in Each Molecular Weight Category of Vanadyl Compounds in the Four Heavy Crude Petroleums and Their Asphaltenes, Maltenes, Asphaltene Polar Extracts, and Extracted Asphaltenes by 50/100/1000 A SEC-HPLC-GFAA Analysis"- ... 

Panzer t al. (5) extracted Athabasca tar sand in two steps, the first with compressed n-pentane (Tc = 570 K, Pc = 3.37 MPa) and the second with compressed benzene (Tc = 563 K, Pc = 4.92 MPa). At 533-563 K and 2.0-7.7 MPa, n-pentane extracted 95% of the maltenes and asphaltenes from the tar sand, whereas at atmospheric pressure only 75% was extracted. Further extraction with benzene at 633 K and 2.0 MPa removed the remaining higher molecular weight asphaltenes. This indicates that the chemical nature of the dense gas is important in some applications. 

Figure 12. SIR-GC/MS traces for the molecular ions of the hopane sulfides recorded for the whole sulfide concentrate from Athabasca maltene. (Reproduced from Ref. 23 with permission. Copyright 1986, Pergamon Press Ltd.)...

An early hypothesis of the physical structure of petroleum (52) indicated that asphaltenes are the centers of micelles formed by adsorption, or even by absorption of part of the maltenes, that is, resin material, onto the surfaces or into the interiors of the asphaltene particles. Thus, most of those substances with greater molecular weight and with the most pronounced aromatic nature are situated closest to the nucleus and are surrounded by lighter constituents of less aromatic nature. The transition of the intermicellular (dispersed or oil) phase is gradual and almost continuous. Continued attention to this aspect of asphaltene chemistry has led to the assumption that asphaltenes exist as clusters within the micelle. This arises mainly because of the tendency for asphaltenes to associate in dilute solution in solvents of low polarity and from possible misinterpretation of viscosity data (58, 64). The presence of asphaltene stacks in the solid phase, as deduced from x-ray diffraction patterns (68), also seemed to support the concept of the widespread existence of asphaltene clusters in the micelle. 

Asphalt Solubility in Normal Alkanes. The separation of an asphalt into two fractions—asphaltenes and maltenes—by precipitation with low molecular weight alkanes is a physical method of separation based on solubility. Figure 1 shows that pentane precipitates more asphalt components (17.0 wt %) than does heptane (10.6 wt %). It would be expected that when pentane asphaltenes are treated with heptane, the amount of material equal to 10.6 wt % (based on asphalt) would be precipitated. However, more—14.8 wt %—is precipitated. A similar, but even more pronounced, effect can be seen when heptane and decane treatments are compared (see Figure 1). 

The molecular size distributions and the size-distribution profiles for the nickel-, vanadium-, and sulfur-containing molecules in the asphaltenes and maltenes from six petroleum residua were determined using analytical and preparative scale gel permeation chromatography (GPC). The size distribution data were useful in understanding several aspects of residuum processing. A comparison of the molecular size distributions to the pore-size distribution of a small-pore desulfurization catalyst showed the importance of the catalyst pore size in efficient residuum desulfurization. In addition, differences between size distributions of the sulfur- and metal-containing molecules for the residua examined helped to explain reported variations in demetallation and desulfurization selectivities. Finally, the GPC technique also was used to monitor effects of both thermal and catalytic processing on the asphaltene size distributions. 

Preparative GPC. The preparative GPC work was performed on the experimental setup shown in Figure 1. Four 1-in. i.d. glass columns were packed with Styragel (Waters Associates) with 1 ft 104 A porosity, 2 ft 500 A porosity, and 1 ft 100 A porosity. The Styragel porosites were chosen to give good resolution for the entire range of molecular sizes found in residua. Two separate column systems were used—one for maltenes, the other for asphaltenes. 

This calibration does not assume that the n-alkanes and polystyrenes are typical of residual molecules. However, they do provide well-defined size standards in the elution time range of interest. No assumptions can be made concerning the shapes of the asphaltene or maltene molecules. Therefore, the GPC size calculated is defined as the critical molecular dimension, which determines if the asphaltene or maltene molecule will diffuse into the pores of the GPC packing. This size is assumed to be related to the size parameter that determines the molecular diffusion into hydrotreating catalyst pores. 

The sizes determined in this work are the apparent molecular sizes and not necessarily the sizes of the asphaltene and maltene molecules at process conditions. Association efforts for asphaltene molecules have been observed for both vapor-phase osmometry molecular weight and viscosity measurements (14, 15). The sizes reported here were measured at 0.1 wt % in tetrahydrofuran at room temperature. Other solvent systems (chloroform, 5% methanol-chloroform, and 10% trichlorobenzene-chloroform) gave similar size distributions. Under these conditions, association effects should be minimized but may still be present. At process conditions (650-850°F and 5-30% asphaltene concentration in a maltene solvent), the asphaltene sizes may be smaller. However, for this work the apparent sizes determined can be meaningfully correlated with catalyst pore size distributions to give reasonable explanations of the observed differences in asphaltene and maltene process-abilities (vide infra). In addition, the relative size distributions of the six residua are useful in explaining the different processing severities required for the various stocks. Therefore, the apparent sizes determined here have some physical significance and will be referred to just as sizes. 

Both asphaltene and maltene molecular size distributions were compared with the pore size distribution of a small pore desulfurization catalyst. Figure 4 shows the Kuwait maltene and asphaltene size distributions along with the catalyst pore size distribution. Most of the maltene molecules are small enough to diffuse into the catalytic pores. In contrast, the Kuwait asphaltenes have a... 

Size characterization measurements have provided useful information on the importance of the hydroprocessing catalyst pore size distribution and on the effects of visbreaking and hydroprocessing on the residua molecular size distributions. It is apparent that asphaltenes and maltenes are not unique entities, but instead have considerable overlap in their size distributions. A complete study of the effects of processing conditions would require consideration of all components of a residuum. 

Simulated Distillation Results. Important insight concerning the molecular-size distribution of asphaltenes vs. maltenes is gained by the simulated distillation data. The boiling point distribution curves are shown in Figure 3. These curves were drawn as follows. Quantitative simulated distillation data was obtained on the virgin bitumen and the maltenes. Direct information is obtained up to a nominal boiling point of 535°C shown by the vertical dashed line. The area under the curve for the nonvolatile portion is... 

Attempts to obtain a distillation curve directly on the asphaltenes fraction resulted in essentially no FID response. This result is attributed to the high free energy of association of these molecules. Once these species are allowed to associate, it is difficult to disassociate them. The results in Figure 3 suggest that volatile molecules are contained in the asphaltene fraction and that the overlap in molecular weight and volatility between asphaltenes and maltenes is substantial. 

The effect of these two distributions, namely the partitioning of species between the two phases, and the overlap of the two phases with respect to molecular weight and functionality implies that the separation between maltenes and asphaltenes is nondefinitive in terms of chemical functionalities and structures. While properties measurements commonly detect differences in the average properties, these differences are small when compared with the width of the distribution about that average. 

This concept of asphaltenes is useful in the interpretation of the present data, and conversely, the data support the concept. First, 13C NMR data show that the saturated hydrocarbon structure, which constitutes the majority of the carbon in the fractions, is virtually identical between the asphaltenes and the maltenes, within the limited sensitivity of 13C NMR. This factor is consistent with the argument that there is a partitioning between fractions and that the appearance of a particular species predominantly in the asphaltene fraction results because of a relatively higher aromaticity or the presence of polar heteroatoms for a specified molecular weight. It is important to recognize, from a processing standpoint, that only a minor weight percent of the fraction (or molecule) may be responsible for its classification as an asphaltene. 

It has been believed that coke is produced by the precipitation of large molecular hydrocarbons such as asphaltenes when their solubility in oil is lowered (13, 14). An increase in the conversion of vacuum residue increases the aromaticy of the asphaltenes and decreases the aromaticy of the maltenes (15). Consequently, the solubility of the asphaltenes in the maltenes decreases. Absi-Halabi et al. propose that absorption of asphaltenes on the acidic sites of an alumina support is a major cause of the initial rapid coke deactivation, while a decrease in asphaltene solubility causes the following steady coke build-up (14). This explain that an amount of coke increases from the entrance to the exit of the reactors as asphaltene solubility decreases and that an increase in the residue conversion increases an amount of coke in the reactor exit. 

Crude oil residues and bitumen are colloidal disperse systems. In these systems, high-molecular solid structure units (asphaltenes) are dispersed in an oily phase (maltenes) (see section 8.2). In industrial thermal cracking processes, these units precipitate as coke. Coke formation is caused by polycondensation reactions of aromatic cores of asphaltenes, which lose the paraffinic periphery. The main objective of a substantial portion of this chapter is to show how deep cracking of bitumen at low temperature can be achieved without coke formation (i.e., without polycondensation of asphaltenes). The main reactions of asphaltenes that lead to coke formation are described. Also described are ways to reduce the negative influence of these reactions on the process. 

In contrast to the results at 400°C, no reduction in asphaltene molecular weight was observed for residence times up to 40 minutes for reactions conducted at 425°C (see Fig. 9.2). This means that at higher temperatures, polycondensation reactions proceed faster than decomposition reactions. At any temperature, the determined molecular weight of asphaltenes shows that it reaches equilibrium as the reaction proceeds. This implies that at a longer residence time, the molecular weight of the asphaltene fraction will not increase any further because after achieving equilibrium molecular weight, they become less soluble in the maltenes. This leads to their flocculation from the maltenes fraction (Fig. 9.2) and, finally, to coke formation. 

---