Energy & Fuels 2003, 17, 127-134
127
Aggregates Structure Analysis of Petroleum Asphaltenes with Small-Angle Neutron Scattering Ryuzo Tanaka* Central Research Laboratories, Idemitsu Kosan Co., Ltd., 1280, Kamiizumi, Sodegaura 299-0293, Japan
Jerry E. Hunt, Randall E. Winans, and Pappannan Thiyagarajan Chemistry Division and Intense Pulsed Neutron Source Division, Argonne National Laboratory, 9700 South Cass Avenue, Argonne, Illinois 60439
Shinya Sato and Toshimasa Takanohashi Institute for Energy Utilization, National Institute of Advanced Industrial Science and Technology, 16-1, Onogawa, Tsukuba 305-8569, Japan Received February 13, 2002
The objective of this study is to examine changes in the structures of petroleum asphaltene aggregates in situ with small-angle neutron scattering (SANS). Asphaltenes were isolated from three different crude oils: Maya, Khafji, and Iranian Light. An aliquot of the 5 wt % asphaltene solution in deuterated Decalin, 1-methylnaphthalene, or quinoline was loaded in a special stainless steel cell for SANS measurements. SANS data measured at various temperatures from 25 to 350 °C showed various topological features different with asphaltene or solvent species. A fractal network was formed only with asphaltene of Maya in Decalin, and it remained even at 350 °C. In all of the solvents, asphaltenes aggregate in the form of a prolate ellipsoid with a high aspect ratio at 25 °C and got smaller with increasing temperature. That became a compact sphere with the size of around 25 Å in radius at 350 °C.
Introduction Asphaltenes, the most polar and heaviest compounds of oil, associate themselves in solution to form complex colloidal structures. Asphaltenes aggregating in the porous reservoir space present an impediment to oil recovery. The obstruction of oil-carrying pipe lines by asphaltenic crude oils is due to the formation and deposition of sediments in the pipelines. In refinery operations, processing of the heavier asphaltene-rich fraction also causes serious problems, many of them being again related to the presence of aggregates in the fraction. These difficulties have led to an important research effort aimed at understanding the colloidal properties of asphaltenes as a function of thermodynamic conditions (pressure, temperature, and oil composition). It is generally believed that asphaltenes retain such a structure in solution, the aliphatic/naphthenic cluster being solvate to some amount. This structure might be responsible for the micellar-like properties observed in asphaltene solutions. These entities associate, in turn, into structures of larger scale (aggregates) whose properties remain to a large extent unknown and are the subject of a number of hypotheses and specula* Corresponding author. Ryuzo Tanaka, Central Research Laboratories, Idemitsu Kosan Co., Ltd., 1280, Kamiizumi, Sodegaura, Chiba 299-0293, Japan. Phone: +81 (438) 75-4380. Fax: +81 (438) 75-7213. E-mail:
[email protected].
tions. The following issues are of basic scientific interest as well as for process conditions used in industry worldwide: (1) size and shape of asphaltene aggregates from various crude oils, (2) effects of solvent and temperature condition on size and shape of aggregates, and (3) relationship between molecular structure of asphaltenes and the size and shape of their aggregates. Many of the research efforts in this field were reviewed in some papers.1-3 Recently, some unique and sophisticated approaches for study of asphaltene aggregation were reported. Castillo et al. performed precipitation and adsorption experiments4 or nonlinear optical response measurements5 to characterize asphaltene aggregates. These results were used to improve the model for asphaltene colloids, which suggests that colloids consist of a wellpacked and insoluble asphaltene core, impervious to the solvent, and a loose-packed periphery which, by allowing (1) Sheu, E. Y. Energy Fuels 2002, 16, 74-82. (2) Sheu, E. Y.; Mullins, O. C. Asphaltenes, Fundamentals and Applications; Plenum Press: New York, 1995. (3) Yen, T. F. Symposium on Chemistry and Characterization of Asphalts, Divisions of Petroleum Chemistry and Fuel Chemistry, Aug 26-31, 1990; ACS 200th National Meeting; American Chemical Society: Washington, DC, 1990; pp 314-319. (4) Castillo, J.; Ferna´ndez, A.; Ranaudo, M. A.; Acevedo, S. Pet. Sci. Technol. 2001, 19, 75-106. (5) Castillo, J.; Hung, J.; Ferna´ndez, A.; Mujica, V. Fuel 2001, 80, 1239-1243.
10.1021/ef020019i CCC: $25.00 © 2003 American Chemical Society Published on Web 11/27/2002
128
Energy & Fuels, Vol. 17, No. 1, 2003
Tanaka et al.
Table 1. Properties of Asphaltenes yield on vacuum residue, wt % elemental analysis, wt % carbon hydrogen sulfur nitrogen oxygen H/C metals, ppm Ni V density, g/cm3 molecular weight
Maya
Khafji
Iranian Light
24.9
14.2
6.3
82.0 7.5 7.1 1.3 1.2 1.10
82.2 7.6 7.6 0.9 1.1 1.11
83.2 6.8 5.9 1.4 1.5 0.98
390 1800 1.1767 4000
200 550 1.1683 4000
390 1200 1.1669 2400
solvent penetration, keep the colloid in solution. C¸ imenogˇlu carried out magnetic resonance measurements of asphaltene solutions.6 The results indicated that solvent molecules can diffuse freely in and out of the asphaltene micelles making dipolar interaction and some solvent molecules can attach to the asphaltene particles for an extremely short time forming complexes. Sharma et al. observed high-resolution transmission electron microscopy images of stacking in asphaltenes and aromatic ring systems.7 (The paper is concerned more with molecular structure of asphaltenes than with their aggregation.) The micrographs of powder asphaltenes exhibit some very local order and long-range disorder, and these images illustrate that the ring systems often occur with two or three stacking together. Small-angle neutron scattering (SANS) is the one of the useful techniques for studying size and shape of the asphaltene aggregates.8-12 In this study, SANS measurements were performed for three asphaltenes from different crude oils: Maya (a Mexican crude oil), Khafji (Arabian heavy oil), and an Iranian light oil, in three different solvents at a wide range of temperatures (up to 350 °C). On the basis of our experience in SANS of asphaltenes from Maya crude oil,13 we expect that the comparison of these three asphaltenes will help us understand the relationship between asphaltene properties and aggregates structures. Experimental Section Sample Preparation and Analyses. The residua (>500 °C) was obtained by vacuum distillation of three crude oils. Asphaltenes were isolated by addition of a 20:1 excess of n-heptane to each of the residua at 25 °C. The suspension was stirred for 1 h at 100 °C in the autoclave. After cooling and standing at 25 °C overnight, the suspension was filtered. The precipitate was washed with n-heptane twice and dried. The yields of asphaltenes (precipitates) of Maya, Khafji, and Iranian Light are 24.9, 14.2, and 6.3 wt %, respectively. (6) C¸ imenogˇlu, M. A. Fuel 2001, 80, 2041-2047. (7) Sharma, A.; Groenzin, H.; Tomita, A.; Mullins, O. C. Energy Fuels 2002, 16, 490-496. (8) Ravey, J. C.; Ducouret, G.; Espinat, D. Fuel 1988, 67, 15601567. (9) Overfield, B.; Sheu, E. Y.; Liang, K. S.; Sinha, S. K. Fuel Sci. Technol. Int. 1989, 7, 611-624. (10) Sheu, E. Y. Phys. Rev. A 1992, 45, 2428-2438. (11) Sheu, E. Y.; Liang, K. S.; Sinha, S. K.; Overfield, R. E. J. Colloid Interface Sci. 1992, 153, 399-410. (12) Liu, Y. C.; Sheu, E. Y.; Chent, S. H.; Storm, D. A. Fuel 1995, 74, 1352-1356. (13) Thiyagarajan, P.; Hunt, J. E.; Winans, R. E.; Anderson, K. B.; Miler, J. T. Energy Fuels 1995, 9, 829-833.
Table 1 shows properties of the asphaltenes. Carbon and hydrogen contents were measured using a CHN-O-Rapid (Elementar); sulfur, nitrogen, and oxygen contents were measured using an AQS-6W sulfur tester (Tanaka Scientific Instrument), an ANTEK7000 (Antek), and a CHN-O-Rapid (Heraeus), respectively. Metals were determined by the induced coupled plasma (ICP) method using a SPS1500VR Plasma Spectrometer (Seiko Instruments). Densities were measured in conformity with JIS K 7112 using a DMA45 (Paar). Molecular weights were measured using an Automatic Molecular Weight Apparatus (Rigosha). Maya asphaltene (As-MY) is the heaviest, having the highest density and the largest molecular weight, and contains the most amount of metals among the three asphaltenes. Khafji asphaltene (As-KF) is medium heavy and contains the least amount of metals. The H/C atomic ratio of As-KF is the highest, and this could be interpreted as the lowest aromaticity. Iranian Light asphaltene (As-IL) is the lightest, having the lowest density and smallest molecular weight, but contains much nitrogen and metals. It also has the lowest H/C atomic ratio, meaning the highest aromaticity. These properties are considered to affect aggregation phenomena of asphaltene molecules. Small-Angle Neutron Scattering. SANS, which is conducted with neutrons, is one of the small-angle scattering techniques such as SAXS (X-ray) or SALS (light). In each of these techniques, radiation is elastically scattered by a sample and the resulting scattering pattern is analyzed to provide information about the size and shape of some component of the sample. The most fundamental difference between neutron and electromagnetic radiation is the mechanism by which the incident radiation interacts with matter. Light and X-rays are both scattered by the electrons surrounding atomic nuclei, but neutrons are scattered by the nucleus itself. From this single fact, the following important conclusions can be drawn.14 In the case of light or X-rays, the scattering cross-section of an atom increases proportionally to the number of electrons, so increases with increasing atomic number, Z. However, the strength of the neutron-nucleus interaction varies completely irregularly with Z; not even isotopes of the same element have the same neutron scattering cross-section. The most significant isotopic variation occurs when Z ) 1, hydrogen and deuterium. Thus neutrons can not only detect hydrogen isotopes, but also differentiate between them. This simple fact is very important for a SANS experiment. Asphaltenes and solvents are both hydrocarbon compounds, which themselves do not show clear contrast in SANS because they have no significant difference in element constitutions. The deuterated solvents were used for sample solutions to reveal the asphaltene aggregate outlines. The interaction of neutrons with matter is weak and the absorption of neutrons by most materials is correspondingly small. Neutron radiation is therefore very penetrating. For example, it would require X-ray with energies of some 105 eV to penetrate a sample and its container more than a millimeter or two thick. Neutrons, on the other hand, can be used to probe the samples contained inside complex pieces of apparatus (cryostats, furnaces, pressure cells, shear apparatus, etc.). In this study, stainless steel sample cells were used for highpressure and high-temperature experiments. Theory. The intensity of scattered neutrons is proportional both to the incident neutron intensity and differential crosssection. Because differential cross-sections contain all the information on the shape, size, and interactions of the scattering bodies in the sample, the objective of a SANS (14) King, S. M. Introduction to Small-Angle Neutron Scattering. http://www.isis.rl.ac.uk/largescale/loq/documents/sans.htm (accessed October 2000).
Aggregates Structure Analysis of Petroleum Asphaltenes
Energy & Fuels, Vol. 17, No. 1, 2003 129
Figure 1. Plots of log(I) vs log(q) for the 5 wt % asphaltenes in Decalin (Dec), 1-methylnaphthalene (1MN), and quinoline (Qui) at 25, 150, 300, and 350 °C. experiment is to determine the differential cross-section. The differential cross-section is given by
dΣ (q) ) φp(∆δ)2P(q)S(q) + Binc dΩ
(1)
where φp is the volume fraction of scattering bodies (given the subscript “p” for “particles”), (∆δ)2 is the square of the difference in neutron scattering length density (commonly called the contrast), P(q) is a function known as the form factor, S(q) is the interparticle structure factor, q is the scattering vector, and Binc is the incoherent background signal. The scattering vector is the modulus of the resultant between the incident, ki, and scattered, kf, wave vectors and is given by
q ) |q| ) |kf - ki| )
4π θ n sin λ 2
()
(2)
in neutron scattering n ∼ 1 and λ is wavelength of the neutron. q has dimensions of (length)-1, normally quoted in Å-1. The contrast is simply the difference in the neutron scattering length density, δ, between that part of the sample of interest, δp, and the surrounding medium or matrix, δm, all squared; (∆δ)2 ) (δp - δm)2. Clearly, if (∆δ)2 is zero, then eq 1 is zero
and there is no SANS. The form factor is a function that describes how (dΣ/dΩ)(q) is modulated by interference effects between radiation scattered by different parts of the same scattering body. The general form of P(q) is given by Van de Hulst’s equation5
P(q) )
∫
1 | V2p
Vp
0
exp[if(qR)] dVp|
(3)
where Vp is the volume of one scattering body and R is a shape parameter that might represent a length or a radius of gyration. The interparticle structure factor is a function that describes how (dΣ/dΩ)(q) is modulated by interference effects between radiation scattered by different scattering bodies. Consequently, it is dependent on the degree of local order in the sample, such as might arise in an interacting system. Experiment. An aliquot of the 5 wt % asphaltene solution in Decalin-d18, 1-methylnaphthalene-d10, or quinoline-d7 was loaded in a stainless steel cell constructed especially for SANS measurements. The wall thickness of 1 mm and a sample thickness of 3 mm make the total path length of 5 mm for neutrons. Only 90% of the volume of the cell was filled to allow for expansion at temperatures above the boiling point of
130
Energy & Fuels, Vol. 17, No. 1, 2003
Tanaka et al.
Figure 2. Plots of log (I) vs log(q) of KF/Dec data with fitting curves of ellipsoids (solid lines) and spheres (dotted lines). solvents (241, 113, and 190 °C for 1-methylnaphthalene-d10, quinoline-d7, and Decalin-d18, respectively). The net transmission for neutrons for the sample averaged over all wavelengths (1-14 Å) was 0.61. Small-angle neutron scattering was performed by placing the sample cell in a boron nitride furnace tube in the small-angle neutron diffractometer (SAND) at the Intense Pulsed Neutron Source (IPNS) at Argonne National Laboratory. The temperature of the furnace was measured using a type K thermocouple and maintained within 0.5 °C using a Micricon controller. At each temperature, the sample was equilibrated for about 30 min prior to SANS measurements. The temperature was increased at the rate of 5 °C/ min. The SAND instrument uses neutrons produced in pulses by spallation due to the deposition of 450 MeV protons on a depleted uranium target, followed by a solid methane moderator (22 K) yielding a wavelength range of 1-14 Å. Detection
of scattered neutrons was accomplished with a 128 × 128 array, 40 × 40 cm2 area sensitive, gas-filled proportional counter, and the wavelength of the scattered neutrons was determined by their times of flight. However, only data from a circular region with a radius of 10 cm from the center of the direct beam were used due to the restrictions placed by the limited radius of the cylindrical boron nitride furnace tube. Data were corrected for unit transmission of the sample, the scattering from the stainless steel cell, and incoherent scattering. The accessible q range using SAND is from 0.008 to 0.2 Å-1.
Results and Discussion Outline of SANS Spectra. Figure 1 shows the SANS spectra of 5 wt % solutions of As-MY, As-KF, and As-
Aggregates Structure Analysis of Petroleum Asphaltenes
Energy & Fuels, Vol. 17, No. 1, 2003 131
Table 2. Results of Data Fitting asphaltene Maya
Khafji
Iranian Light
temperature, °C
ra, Å
ab, Å
b c, Å
a.r.d
χ2sphere e
χ2ellipsoid f
c.r.g
Decalin
25 150 300 350
49.27 41.65 35.23 26.59
7892.50 74.81 57.89 38.64
18.32 25.87 22.74 14.17
430.81 2.89 2.55 2.73
56.72 4.09 0.96 1.00
3.77 0.82 0.58 1.05
15.05 4.99 1.66 0.95
1-methylnaphthalene
25 150 300 350
34.80 30.94 27.58 26.00
75.76 50.57 31.07 27.31
25.08 22.72 25.77 25.31
3.02 2.23 1.21 1.08
25.08 3.54 0.83 1.29
1.29 1.27 0.86 1.35
19.44 2.79 0.97 0.96
quinoline
25 150 300 350
41.80 32.65 28.23 25.34
83.79 51.06 38.53 35.33
25.31 24.09 22.43 19.83
3.31 2.12 1.72 1.78
6.87 1.59 1.40 1.62
1.34 1.03 1.38 1.66
5.13 1.54 1.01 0.98
Decalin
25 150 300 350
49.83 33.72 30.51 28.66
142.40 59.35 47.75 45.81
36.11 25.00 22.38 19.36
3.94 2.37 2.13 2.37
105.75 14.98 1.91 1.78
2.87 1.34 0.99 1.47
36.83 11.18 1.93 1.21
1-methylnaphthalene
25 150 300 350
32.89 29.80 26.25 21.68
63.28 51.65 43.05 24.52
24.23 21.33 17.08 20.18
2.61 2.42 2.52 1.22
20.06 3.43 0.98 1.24
1.87 0.81 0.78 1.31
10.73 4.23 1.26 0.95
quinoline
25 150 300 350
34.30 29.29 26.82 25.12
70.12 44.26 34.90 25.31
24.77 22.07 22.38 25.03
2.83 2.01 1.56 1.01
18.47 1.51 0.92 1.87
1.55 1.05 0.94 1.96
11.92 1.44 0.98 0.95
Decalin
25 150 300 350
45.62 34.58 28.76 22.63
135.65 50.25 28.44 23.45
32.16 25.26 29.21 22.19
4.22 1.99 0.97 1.06
116.22 1.73 1.94 0.75
1.93 1.43 1.86 0.79
60.22 1.21 1.04 0.95
1-methylnaphthalene
25 150 300 350
30.84 28.97 25.94 24.67
55.52 44.89 38.13 26.65
21.81 20.97 19.46 23.63
2.55 2.14 1.96 1.13
5.42 1.86 1.22 1.20
0.88 1.26 1.12 1.25
6.16 1.48 1.09 0.96
quinoline
25 150 300 350
31.17 28.31 35.57 25.39
70.34 54.21 42.12 40.20
22.62 20.65 18.76 19.04
3.11 2.63 2.25 2.11
23.81 6.19 1.59 1.00
1.51 0.71 1.00 0.76
15.77 8.72 1.59 1.32
solvent
a Radius of sphere. b Parallel semiaxis of ellipsoid. c Perpendicular semiaxis of ellipsoid. d Aspect ratio of ellipsoid (a/b). e Reduced chisquared of spherical data fitting. f Reduced chi-squared of ellipsoidal data fitting. g Ratio of reduced chi-squared of spherical and ellipsoidal 2 2 data fitting (χsphere /χellipsoid ).
IL measured at 25, 150, 300, and 350 °C. The solvents are Decalin (Dec), 1-methylnaphthalene (1MN), and quinoline (Qui), all of the solvents are fully deuterated. The data in the region of 0.02 < q < 0.12 Å-1 show the coherent scattering from asphaltene aggregates. With scattering data at this q region of coherent scattering, we can get topological information about aggregates with the size of ten to hundreds of angstroms. With the data at the lower q region, information about larger structures can be deduced. Data at a larger q region is incoherent scattering, which is determined only by composition of the solutions in the beam path of a neutron, in other words, by concentration of asphaltenes. Red open circles show the data measured at 25 °C. Blue squares, green triangles, and brown solid circles show the data measured at 150, 300, and 350 °C, respectively. Scattered neutron intensities are decreased with increasing temperature. The scattering intensity is proportional to the summation of cross sections of all aggregates. Therefore, the decrease of the scattering intensity means that the single molecules not contributing to the aggregate formation would increase with increasing temperature. As clear in the As-KF spectra, data plot curves turn over more under the q ) 0.3 Å-1 region due to the
interparticle interaction of the colloidal particle. As temperature increases, the particle size becomes smaller and the number of particles becomes larger, therefore the average distance of the particles become shorter and interaction become larger. To analyze this interparticle interaction quantitatively, nonlinear regression based on the detailed nature of the interaction is needed. It requires further study, because we do not have enough information about that so far. Shape and Size of Aggregates in Solvents. The size and shape of the scattering particle can be obtained by combining information from the Guinier analysis15 in the low q region, the power-law behavior, if in the low q region of log(I) vs log(q) plot, and nonlinear regression analysis using the expression for the form factor of an appropriate shape. Polydispersity analysis is also important to understand the distribution of the asphaltene aggregate’s size.10-12 In this work, a monodispersity model was applied as the first approximation to gain the representative values of the aggregate size for each asphaltene in solvents. It simplifies the comparison of the samples and understanding of the (15) Van de Hulst, H. C. Light Scattering by Small Particles; John Wiley: New York, 1957.
132
Energy & Fuels, Vol. 17, No. 1, 2003
Tanaka et al.
Figure 4. Parallel semiaxes of fitted ellipsoids at 25, 150, 300, and 350 °C: MY (upper), KF (middle), and IL (lower). Figure 3. Parallel and perpendicular semiaxes of fitted ellipsoids for three asphaltenes: MY (upper), KF (middle), and IL (lower).
relationship between the asphaltene properties and its aggregation phenomena. Topologically, the forms of asphaltene aggregates are summarized to the sphere, disk, and rod. In our analysis we used either the form factor for a sphere with a radius R (eq 4) 3 2
I(q) ) I(0)[(3 sin qR - qR cos qR)/(qR) ] + B
(4)
or an ellipsoid (eq 5), which covers a wide range of aspect ratio to obtain information on the size and morphology of the aggregates. In eq 4, the variables for nonlinear regression are I(0), R, and incoherent background, B. In the presence of interparticle interactions,
the measured scattering intensity is a function of the form factor, F(q), and the interparticle structure factor, S(q), i.e., I(q) ) S(q)|F(q)|2. The use of eqs 4 and 5 assumes that a q domain can be found where S(q) f 1. In general, such domains exist in the high q region. The form factor used for the prolate ellipsoid averaged over all the orientations in solution was
I(q) ) I(0)
π/2 [(3 sin X - X cos X)/X3]2 cos β dβ + B ∫β)0
(5) In eq 5, X ) qAs[cos2 β + (As/Bs)2 sin2 β]1/2, where As and Bs are the major and minor semi-axes of the prolate ellipsoid, and β is the orientational averaging angle. The variables in eq 5 for nonlinear regression are I(0), As,
Aggregates Structure Analysis of Petroleum Asphaltenes
Energy & Fuels, Vol. 17, No. 1, 2003 133
Figure 5. Plots of log(I) vs log(q) of MY/Dec data with fitting curves of power law slopes.
and Bs, and the incoherent background, B. As an example of the data fitting, Figure 2 shows the data and fitting curves of the Decalin solution of Khafji asphaltene (KF/Dec). Table 2 shows all of the fitting results. In the case of spherical fitting, the radius of the sphere decreases from 50 to 20 Å as the temperature is increased from 25 to 350 °C. On the other hand, in the case of ellipsoidal fitting, parallel semiaxes show more drastic change from ca. 100 Å (at 25 °C) to ca. 30 Å (at 350 °C), while perpendicular semiaxes do not change much around 20 Å at any temperature. Reduced chisquared (χ2) of ellipsoidal fitting are smaller than that of spherical fitting at 25 and 150 °C. At 25 °C, χ2 of spheres are very large, meaning that the fittings are
worse. At 300 and 350 °C, χ2 of the spherical fittings are similar to that of ellipsoidal fittings. From these χ2 features, it is supposed that the shape of asphaltene aggregates is ellipsoid at 25 and 150 °C, and it becomes spherical with increasing temperature. The change of parallel semiaxis and perpendicular semiaxis with temperature is shown in Figure 3. The straight line (to present data of spheres) is also shown in the figures for reference. The parallel semiaxis of a Decalin solution of Maya asphaltene (MY/Dec) at 25 °C is calculated as 7892 Å; it is far out of the range of the figure. MY/Dec makes a fractal network, which is a large aggregate, as mentioned later, and the data of MY/ Dec at 25 °C seem to be affected much by the fractal
134
Energy & Fuels, Vol. 17, No. 1, 2003
network. The plots of any samples move from upper right to lower left with temperature increasing as 25, 150, 300, and 350 °C. The changes of parallel semiaxes with temperature represent the change of the aggregate’s size, because perpendicular semiaxes do not change much with temperature. Figure 4 shows the change of the parallel semiaxis with temperature. The parallel semiaxes of all asphaltene aggregates become shorter to a great extent at temperatures between 25 °C and 150 °C. The parallel semiaxis of MY/Dec is longer than that of MY/1MN and MY/Qui at any temperature over 150 °C. On the other hand, the parallel semiaxes of As-KF and As-IL aggregates in Dec are the same as those in other solvents at high temperature. Fractal Network. As seen in Figure 1, only MY/Dec data show the scattering neutron intensity in the low q region (q < 0.01 Å-1), while other data have no intensity in that region. It is supposed that the neutron scattering in that region is due to a fractal network of asphaltene aggregates. The fractal network may be a larger secondary aggregate composed with associated primary ellipsoidal aggregates. The fractal dimensions of the aggregates are estimated by subtracting the exponent of the power law slopes from 6. The fitting of power law slopes is shown in Figure 5. The surface fractal dimensions at 25, 150, 300, and 350 °C are 2.72, 2.71, 2.46, and 2.42, respectively. The surface fractal dimension represents the smoothness of the aggregates; the dimension of a completely smooth surface is 2. The fractal dimension decreases a little with increasing temperature, meaning the aggregate surface become slightly smoother with increasing temperature. Only the combination of Maya asphaltene with Decalin (MY/Dec) shows the fractal network. MY/Dec keeps the fractal network even at 350 °C (thermal cracking starts at around this temperature). The high cokemaking tendency of Maya asphaltene may be related to the fractal network of its aggregates. In fact, only in the case of MY/Dec experiment does a very thin layer of sludges or coking materials remain inside the sample cell wall. Incoherent Scattering and Precipitation of Asphaltenes. As shown in Figure 1, incoherent scattering
Tanaka et al.
intensities (q > 0.12 Å-1) of MY/Dec and IL/Dec at 25 and 150 °C are smaller than those at 300 and 350 °C. Incoherent scattering intensities depend not on the aggregate’s structure but only on the contents of elements in the neutron beam pathway. Therefore, the depressions of incoherent scattering intensities at low temperature are interpreted as meaning that the asphaltene concentration in the neutron beam pathway is relatively lower. It seems to indicate that some portions of asphaltenes precipitate at low temperature in MY/Dec and IL/Dec. As-IL precipitates more than either As-MY or As-KF, which may be due to the high aromaticity of As-IL. Conclusion The shape and size of asphaltene aggregates in solvents were investigated with in-situ SANS experiments. The shapes of asphaltene aggregates deduced from the simple curve fittings are a prolate ellipsoid with high aspect ratio at low temperature and it becomes a sphere with increasing temperature. The size of asphaltene aggregates changes corresponding to their species, solvents, and temperature. Maya asphaltene aggregate is the largest among the three asphaltene aggregates in any solvents. Maya asphaltene makes a fractal network in Decalin and it persists even at 350 °C. The fractal network may be related to the high coking tendency of Maya asphaltene. Scattering data of Khafji asphaltene aggregates show the influence of interparticle interactions. Iranian Light asphaltene precipitates extensively in Decalin at lower temperature owing to its high aromaticity. Acknowledgment. This work was performed under the auspices of the New Energy Development Organization under sponsorship of the Ministry of Economy, Trade and Industry of Japan. This work has benefited from the use of the Intense Pulse Neutron Source at Argonne National Laboratory under the auspices of the U.S. Department of Energy, Office of Basic Energy Science, under Contract No. W-31-109-ENG-38. EF020019I
