160
Energy & Fuels 2000, 14, 160-163
Rapid Bulk Fractionation of Maltenes into Saturates, Aromatics, and Resins by Flash Chromatography Laı¨la Raki, J-F. Masson,* and Peter Collins Institute for Research in Construction, National Research Council of Canada, Ottawa, Ontario, Canada, K1A 0R6 Received May 24, 1999. Revised Manuscript Received August 7, 1999
Flash chromatography was used to fractionate the maltenic portion of bitumen. The method uses a dry chromatographic support rather than a chromatographic gel. The development of the method is described along with the effect of a number of parameters on the purity of the fractions, namely, the effect of bitumen concentration, settling time in n-heptane, filter size, support type, and mesh size. The method was used to fractionate the maltenes within 45 min to provide milligram to gram quantities of saturates, aromatics, and resins with greater than 83% purity, as determined by Iatroscan. The accessibility of such quantities of relatively pure bitumen fractions will allow for easier study of their composition and characteristics.
Introduction The physical, mechanical, and rheological behavior of bitumen in road and building construction is governed by its chemical composition.1 For this reason, considerable effort goes toward widening the knowledge of this composition and relating it to performance.1-5 Because bitumen is a complex mixture of molecules ranging from nonpolar saturated hydrocarbons to polar polynuclear aromatics, its composition is often conveniently reported in weight percent of saturates (S), aromatics (A), resins (R), and asphaltenes (As), collectively called SARAs.6-12 To fractionate bitumen, several techniques have been proposed, all of which provide either rapid or bulk fractionation. Classical low-pressure liquid chromatography (LC)13-15 allows for obtaining sizable quantities * Author to whom correspondence should be addressed. (1) Petersen, J. C.; Robertson, R. E.; Branthaver, J. F.; Harnsberger, P. M.; Duvall, J. J.; Kim, S. S.; Anderson, D. A.; Christiansen, D. W.; Bahia, H. U. Binder Characterization and Evaluation, Vol. 1; Strategic Highway Research Program, National Research Council, Washington, DC, 1994. (2) Chatergoon, L.; Whiting, R.; Smith, C. Analyst 1996, 121, 373376. (3) Petersen, J. C.; Plancher, H.; Ensley, E. K.; Venable, R. L.; Miyake, G. Transp. Res. Rec. 1982, 843, 95-104. (4) Lee, D. Y.; Huang, R. J. Appl. Spectrosc. 1973, 27 (6), 435-440. (5) Petersen, J. C. Transp. Res. Rec. 1984, 999, 13-30. (6) Philip, C. V.; Bullin, C. J.; Davison, R. R. Reprints from ACS Symposium on Analytical Chemistry of Heavy Oils/Resids, Dallas, April 1989, pp 311-315. (7) Wilt, B. K.; Welch, W. T.; Rankin, J. G. Energy Fuels 1998, 12, 1008-1012. (8) Huang, J.; Yuro, R.; Romeo, G. A., Jr. Fuel Sci. Technol. Int. 1995, 13 (9), 1121-1134. (9) Helm, R. V. Anal. Chem. 1969, 41 (10), 1342-1344. (10) Davis, T. C.; Petersen, J. C.; Haines, W. E. Anal. Chem. 1966, 38 (2), 241-243. (11) Green, J. B.; Hoff, J.; Woodward, P. W.; Stevens, L. L. Fuel 1984, 63, 1290-1301. (12) Chatergoon, L.; Whiting, R.; Smith, C. Analyst 1992, 117, 18691873. (13) Corbett, L. W. Anal. Chem. 1969, 41, 576-579. (14) Goodrich, J. L.; Goodrich, J. E.; Kari, W. J. Transp. Res. Rec. 1984, 1096, 146-167. (15) Standard Test Method for Separation of Asphalt into Four Fractions, ASTM Methods D 4124, 1991, 424-429.
of bitumen fractions but the method is rather tedious. In contrast, microgram quantities of bitumen can be fractionated rapidly with the Iatroscan,16 which combines thin-layer chromatography and flame ionization detection (TLC/FID),17 but the bitumen fractions cannot be characterized further because of the destructive nature of the analysis. Here we present a method that combines advantages of the Iatroscan and classical LC methods. It allows for the rapid fractionation of multigram quantities of the maltenic portion of bitumen into almost pure saturates, aromatics, and resins. The proposed method, termed flash chromatography,18 requires only 25% or less of the solvent used in classical LC. It uses a filtering funnel filled with dry chromatographic support rather than a long column of chromatographic gel. In this paper we describe the effect of a number of factors on the purity of the SARAs fractions obtained by flash chromatography, including settling time in n-heptane, concentration of bitumen in solvent, filter size, support type, and mesh size. It is shown with the Iatroscan that each fraction can be obtained in 83% purity or better. Experimental Section Materials. The bitumen, labeled ABA, was provided by the Strategic Highway Research Program (SHRP). The chromatography supports were Type F-20 alumina 72-180 µm, from Sigma Chemicals Co., St. Louis, MO and silica 63-200 µm and 15-40 µm from EM Sciences, Gibbstown, NJ. Prior to weighing and use, alumina and silica were dried overnight at 200 °C and under vacuum. HPLC grade n-heptane, toluene, and tetrahydrofuran (THF) were purchased from Aldrich Chemicals Co., Inc., Milwaukee, WI, and used as received. The (16) Wan, C. C.; Waters, T. H.; Wolever, R. D. Prepr. Pap.s Am. Chem. Soc., Div. Fuel. Chem. 1992, 37 (3), 1350-1359. (17) Ranny´, M. Thin-Layer Chromatography with Flame Ionization Detection; D. Reidel Publishing Company: Dordrecht, Holland, 1987. (18) Furniss, B. S.; Hannaford, A. J.; Smith, P. W. G.; Tatchell, A. R. Vogel’s Textbook of Practical Organic Chemistry, 5th ed.; Longman Scientific & Technical: Harlow Essex, England, 1989; pp 220-221.
10.1021/ef990100u CCC: $19.00 © 2000 American Chemical Society Published on Web 11/19/1999
Fractionation of Maltenes by Flash Chromatography
Energy & Fuels, Vol. 14, No. 1, 2000 161 Iatroscan. Scanning rate was 30 s/rod; hydrogen flow, 160 mL/ min; air flow, 2 L/min. The results are the averages from the analysis of 10 chromarods.
Results and Discussion
Figure 1. Setup for flash chromatography.
Figure 2. Flowchart for bitumen fractionation by flash chromatography. chromatography columns were sintered glass funnels with “C” porosity purchased from Ace Glass Incorporated, Vineland, NJ. The funnels had volumes of 15 mL (20 mm in diameter and 48 mm in height) and 186 mL (65 mm in diameter and 56 mm in height). Sample Preparation. Bitumen was mixed with heptane in concentrations of 1, 5, and 10% (w/v). The dark brown mixtures were stirred for 30-60 min until homogeneous and then left to settle 0.5-16 h. The heptane-insolubles were recovered in a Bu¨chner funnel over an Erlenmeyer flask linked to a water pump. The filtrate was retained for fractionation onto the dry column. Flash Chromatography. The funnel was filled with dry alumina or silica. The support was compacted with the suction of a water pump and the elution of some heptane (Figure 1). Once the column was dry, the maltenic solution was loaded evenly onto the surface of the support and allowed to run through the filter. This heptane eluate was retained for SARAs analysis with the Iatroscan. Fixed volumes of toluene and THF were sequentially run through the filter to obtain a third and fourth eluate that were also retained for SARAs analysis (Figure 2). Table 1 shows the various parameters studied in developing the fractionation method. Iatroscan.17 Two microliters of a solution were deposited near the bottom of a chromarod before successive elution in heptane (45 min), toluene (17 min), and tetrahydrofuran (5 min). After separation, the fractions were burned in the
Bitumen, the heptane-insolubles, and the various filtrates were characterized by Iatroscan to determine the purity at which the various SARAs fractions could be obtained by flash chromatography. The combination of flash chromatography and Iatroscan provided a twodimensional fractionation of bitumen (Figure 3). For convenience, we termed “iatrogramme” the signal obtained from the Iatroscan. The iatrogramme of the original bitumen showed four peaks corresponding to the four SARAs fractions (bottom iatrogramme, Figure 3). An exact correspondence between area and weight percent was assumed, although this is not strictly true,19 so that the composition of bitumen was calculated as 11, 16, 57, 16 wt % saturates, aromatics, resins, and asphaltenes, respectively. After removal of heptane-insolubles and elution in heptane, toluene, and THF, the various eluates contained only one or two fractions, depending on the experimental conditions as seen from the corresponding numbers of peaks in the iatrogramme. It was also interesting to note that the heptane precipitate, often referred to as the asphaltenes, was found to be a mixture of asphaltenes with resins when analyzed with the Iatroscan. The effect of settling time on the composition of the heptane eluate was tested when 1% (w/v) solutions were maintained at room temperature for 0.5 and 16 h (Table 2, conditions 1a and 1b). After 0.5 h of settling, the asphaltenes content in the heptane and THF eluates was high and moreover all eluates contained the four SARAs fractions. After 16 h of settling, a significant improvement in the purity of the heptane eluate was obtained (96%), but other eluates were mixtures of little purity. The concentration of the bitumen solution was then increased from 1 to 5% with the aim of obtaining greater quantities of the SARAs fractions. As a result of column saturation, however, the heptane filtrate became contaminated with resins and the purity of the saturates decreased (Table 2, condition 1b). The elution of resins is consistent with the saturation of the interaction sites on alumina by aromatics, which would prevent the interaction of resins with alumina. Similarly, the toluene and THF eluates were mixtures of SARAs. The funnel size was then increased from 15 to 186 mL, and the weight of alumina went from 10 to 100 g (Table 1, condition 2). It was not possible to keep the column width-to-height ratio constant, however, as the sintered glass funnels come in few sizes. With 50 mL of a 5% bitumen solution poured onto 100 g of alumina, the purity of the saturates eluate increased again, to 91%, but the other eluates remained mixtures with only 47-64% purity (Table 2, condition 2). Noteworthy was the almost complete absence of asphaltenes in the various eluates. It is probable that with the increase in the weight of alumina, the asphaltenes that did not precipitate remained adsorbed onto the column. (19) Fuhr, B. J.; Holloway, L. R.; Reichert, C. AOSTRA J. Res. 1985, 1 (4), 281-288.
162
Energy & Fuels, Vol. 14, No. 1, 2000
Raki et al.
Table 1. Experimental Conditions for Flash Chromatography of Maltenes
condition
funnel size (mL)
1
15
2
186
3
186
4
186
5
186
6
186
bitumen solution (weight bitumen (g)/ volume heptane (mL))
support (wt, g) Al2O3 (72-180 µm) Al2O3 (72-180 µm) SiO2 (63-200 µm) SiO2 (15-40 µm) SiO2 (15-40 µm) SiO2 (15-40 µm)
10
volume of eluent (mL) heptane
toluene
THF
0.1/10, 0.5/10
0
10
10
100
2.5/50
0
50
50
70
2.5/50
0
50
50
60
2.5/50
0
50
50
60
2.5/50, 5/50
100
150
150
60
5/50
100
100
150
Table 2. Effect of the Experimental Conditions on the Purity of the Fractions bitumen composition (%)b concentration condition (w/v %) eluate S A R As 1aa
1
1b
1 5
2
5
3
5
4
5
5
5 10
6
Figure 3. Representative iatrograms for the various eluates. The n-heptane precipitate was typically 55% resins and 45% asphaltenes after the 16 h settling of a 5% bitumen-heptane mixture.
In an attempt to obtain toluene and THF eluates of higher purity, silica was substituted for alumina. With similar experimental conditions, silica was not as effective as alumina (Table 2, Condition 3). The binding strength of silica toward polar compounds being lower than that of alumina, fewer of the weakly interacting polar saturates, aromatics, and resins remained adsorbed on silica and consequently all the collected eluates were mixtures. The asphaltenes content was also significantly higher than with alumina. The lower density of silica compared with alumina is also significant in this regard. In the same volume, 100 g of alumina or only 70 g of silica could be used. To compensate for the lower binding strength of silica, its surface area was increased by the use of a smaller particle size (15-40 µm). This translated into an increase in the concentration of binding sites and an increase in the purity of all the eluates. The heptane
10
heptane toluene THF heptane toluene THF heptane toluene THF heptane toluene THF heptane toluene THF heptane toluene THF heptane toluene THF heptane toluene THF heptane toluene THF
44 6 2 96 9 1 35 1 3 91 25 3 70 25 4 100 28 1 94 0 0 99 0 0 99 1 0
14 26 8 2 45 36 9 68 33 8 47 33 0 45 71 0 65 11 6 85 0 1 70 0 1 83 1
16 63 77 1 27 58 40 24 64 2 27 64 22 26 24 0 6 87 0 12 93 0 24 94 0 14 97
25 5 12 1 19 5 16 7 0 0 1 1 8 4 1 0 1 1 0 3 7 0 6 6 0 2 2
a Settling time of 0.5 h; in all other cases, the settling time was 16 h. b S ) saturates, A ) aromatics, R ) resins, As ) asphaltenes.
eluate was then 100% saturates, the toluene eluate, 65% aromatics with 28% saturates, and the THF eluate, 87% resins (Table 2, condition 4). Further improvements in the purity of the fractions were obtained with an increase in the volume of eluents (Table 1, condition 5). The elution of more heptane allowed for the collection of the saturates previously left in the column which before were only eluted with the aromatics. In these conditions, the heptane eluate was 94% saturates, the toluene eluate 85% aromatics, and the THF eluate 93% resins. To further increase the quantity of SARAs that could be obtained by flash chromatography, the concentration of the bitumen solution was increased from 5 to 10% (Table 1, condition 5). The heptane eluate again had a high concentration of saturates (99%), the toluene eluate was a mixture of aromatics and resin (70% and 24%, respectively), and the THF eluate was 94% resins. The purity of the toluene eluate was then readily improved by decreasing the volume of toluene from 150 to 100
Fractionation of Maltenes by Flash Chromatography
mL, (Table 2, condition 6). In the end, all fractions showed purity of 83-99%. Conclusion Flash chromatography was used to fractionate the maltenic portion of bitumen into saturates, aromatics, and resins as defined by the Iatroscan. In flash chromatography, a sintered glass funnel is filled with dry chromatographic support. The method is economical, it uses only 25% of the solvent required by classical liquid chromatography, and rapid. Fractionation is complete within 45 min. In adapting the method for bitumen, the effect of a number of parameters on the purity of the collected fractions was examined, namely, the effect of bitumen concentration, settling time in n-heptane, filter size, support type, and mesh size. The results indicate that the asphaltenes in the maltenic filtrate are detrimental to good resolution because the polar asphaltenes bind to the support active sites. A settling time of 16 h (overnight) of the asphaltenes in n-heptane is sufficient to provide a maltenic
Energy & Fuels, Vol. 14, No. 1, 2000 163
fraction that can be fractionated efficiently. In identical conditions alumina provides better separation than silica, but silica is to be preferred for the fractionation because it can be obtained in a finer particle size necessary for good resolution. In the end, bitumen (5 g) was successfully fractionated onto 60 g of 15-40 µm silica into saturates, aromatics, and resins with respective purity of 99, 83, and 97%. The availability of gram quantities of bitumen fractions in such purity will no doubt facilitate the study of their composition and properties, and in turn allow for better defining the structure-property relationship of the mother mixture, bitumen. Acknowledgment. The Natural Sciences and Engineering Research Council of Canada (NSERC) is gratefully acknowledged for granting a postdoctoral fellowship to Laı¨la Raki. The authors also thank the Strategic Highway Research Program (SHRP) for providing bitumen. EF990100U