778
Energy & Fuels 2004, 18, 778-788
The Calibration of Size Exclusion Chromatography Columns: Molecular Mass Distributions of Heavy Hydrocarbon Liquids Fatma Karaca,† Carlos A. Islas,‡ Marcos Millan, Mahtab Behrouzi, Trevor J. Morgan, Alan A. Herod, and R. Kandiyoti* Department of Chemical Engineering and Chemical Technology, Imperial College London, London SW7 2AZ, UK Received October 23, 2003
The aim of the study was to obtain calibration curves for a pair of size exclusion chromatography (SEC) columns operating with 1-methyl-2-pyrrolidinone (NMP) as eluent. The dependence of the calibrations on sample chemical structures has been examined. The calibrations have been compared with elution times of several sets of standards. The level of agreement between SEC and MALDI-mass spectrometry has been evaluated. Molecular mass distributions of several complex samples have been examined in terms of these calibrations. The polystyrene (PS) and poly(methyl methacrylate) (PMMA) calibration curves were close, while a set of polysaccharides (PSAC) and other oxygenates eluted much earlier. However, numerous other samples eluted closer to the PS-PMMA line. To a first approximation, deviations between the PSAC and PS-PMMA lines may be treated as an upper limit to errors arising from structure-dependent variations in this SEC system. Below 15 000 u, MMs of oxygenated samples could be estimated to within a factor of ∼2-2.5. Other structural features gave rise to smaller deviations. Good agreement was observed up to about m/z 3000, between SEC and MALDI and LD-MS. The techniques are independent, suggesting that up to this limit, SEC may be considered as a quantitative tool. The accuracy of the measurement is subject to greater uncertainty with increasing molecular mass. The often-made assumption that high-mass materials are composed of aggregates has been examined. Furthermore, evidence from several analytical techniques provides indications of entirely different structural makeup (e.g., nature of fragments in mass spectrometry; trace element concentration) between fractions with different apparent molecular massessas determined by SEC. It is possible that some molecules adopt 3-dimensional conformations and show up as larger than they really are. While the “aggregates” assumption did not explain our experimental observations, structures of material appearing under the excluded peak in SEC require further careful study.
Introduction Significant proportions of heavy hydrocarbon liquids escape detection in gas chromatography (GC) and GCmass spectrometry (GC-MS). The problem arises from the relatively low molecular mass ceilings of these techniques. GC-columns are normally limited to a little over 300 u for aromatic compounds and about 500-550 u for aliphatics. Columns able to reach 400 °C might extend these ranges by some 100 u. Similarly, probeMS can seldom show up coal- or petroleum-derived material much above 550-600 u. The list of samples for which these techniques provide only a partial picture is extensive and includes coal- and petroleum-derived asphaltenes, biomass tars and extracts, soil-derived organic materials, sediments, and sedimentary rocks.1,2 * Corresponding author. E-mail:
[email protected]. † Present address: Yildiz Teknik Universitesi, Kimya Muh. Bolumu, Davutpasa Cad. No. 127, 34210, Esenler, Istanbul, Turkey. ‡ Present address: CAI, Mexican Petroleum Institute, Mexico 07730 D. F., Mexico. (1) Poirier, N.; Derenne, S.; Rouzaud, J.-N.; Largeau, C.; Balesdent, J.; Maquet, J. Org. Geochem. 2000, 31, 813.
Considerable headway has been made in examining these samples by the parallel use of size exclusion chromatography (SEC) 3-10 and matrix-assisted laser desorption/ionization (MALDI) mass spectrometry; ref 11 presents an up-to-date overview of this work. Prior fractionation of these complex samples has usually served to enhance the resolution of SEC and MALDI(2) Hedges, J. I.; Eglinton, G.; Hatcher, P.; Kirchman, D. L.; Arnosti, C.; Derenne, S.; Evershed, R. P.; Ko¨gel-Knaber, I.; de Leeuw, J. W.; Littke, R.; Michaelis, W.; Rullko¨tter, J. Org. Geochem. 2000, 31, 945. (3) Handbook of Size Exclusion Chromatography, Chromatographic Science Series, Vol. 69; Wu, Chi-san, Ed.; Marcel Dekker: New York, 1995. (4) Lafleur, A. L.; Nakagawa, Y. Fuel 1989, 68, 741. (5) Herod, A. A.; Shearman, J.; Lazaro, M. J.; Johnson, B. R.; Bartle, K. D.; Kandiyoti, R. Energy Fuels 1998, 12, 174. (6) Herod, A. A.; Zhang, S. F.; Johnson, B. R.; Bartle, K. D.; Kandiyoti, R. Energy Fuels 1996, 10, 743. (7) Domin, M.; Moreea, R.; Lazaro, M. J.; Herod, A. A.; Kandiyoti, R. Rapid Commun. Mass Spectrom. 1997, 11, 638-645. (8) Lazaro, M. J.; Herod, A. A.; Cocksedge, M.; Domin, M.; Kandiyoti, R. Fuel 1997, 76, 1225-1233. (9) Vander Hayden, Y.; Popovici, S. T.; Schoenmakers, P. J. J. Chromatogr. A 2002, 957, 127-137. (10) Lazaro, M. J.; Islas, C. A.; Herod, A. A.; Kandiyoti, R. Energy Fuels 1999, 13, 1212.
10.1021/ef030178h CCC: $27.50 © 2004 American Chemical Society Published on Web 03/25/2004
Calibration of Size Exclusion Chromatography Columns
MS, as well as of techniques used for bulk structural characterization.12 The present paper describes recent development work on the calibration of SEC columns. Size exclusion chromatography relies on separating analytes in solution on the basis of their molecular sizes. Much early SEC work on coal-derived materials was carried out using tetrahydrofuran (THF) as eluent.13-15 The inadequate solvent power of THF eventually led to its abandonment in favor of NMP (1-methyl-2-pyrrolidinone). Problems encountered with THF included surface interactions between packing and sample molecules and partial precipitation of sample with attendant gradual increases in column back-pressure.6,16 NMP was first used as eluent in SEC by Lafleur and Nakagawa4 who attempted to establish a molecular mass calibration. NMP appears capable of dissolving coal extracts and coal tar pitch samples completely, but it is not miscible with aliphatic species. Polystyrene molecular mass standards up to 15 × 106 u and poly(methyl methacrylate) standards up to 1 × 106 u have been eluted in NMP with apparently minimum interference from surface effects.17 The latter may be diagnosed by delayed elution of samples, at times longer than the permeation limit of the column, i.e., longer than the time required for the elution of the smallest molecules such as benzene. In SEC, elution time data of samples are translated into MMs through a calibration curve. This curve is obtained by plotting the logarithms of known molecular masses of standards (log MM) against their elution times. Most commonly, these standards are synthetic materials with narrow molecular mass distributions, commonly a set of polystyrenes.9 One important drawback of using THF as eluent was the structure dependence of observed elution times: aliphatics eluted more quickly than alicyclics of similar molecular mass, which preceded aromatic material of corresponding MMs.13-15 As will be shown below, elution times in NMP are less dependent on chemical structure. In a recent calibration exercise, model compounds covering a MM-range to 1086 u have been used. The classes of compounds tested included polycyclic aromatics, azaarenes and other nitrogen-bearing compounds, several dyes, other polar compounds, and numerous oxygenated compounds. Mostly, the polystyrene-based calibration predicted the elution behavior of these compounds to within (1 min.10 The notable lack of agreement found in the case of fullerenes proved instructive. The ultimate aim in this type of work is to obtain a calibration, which provides reasonable estimates of (11) Zhuo, Y.; Herod, A. A.; Kandiyoti, R. The thermochemical reactions of middle rank coals. In Natural and Laboratory-Simulated Thermal Geochemical Processes; Ikan, R., Ed.; Kluwer Academic Publishers: Dordrecht, 2003; Chapter 3, pp 53-151. (12) Herod, A. A.; Lazaro, M. J.; Domin, M.; Islas, C. A.; Kandiyoti, R. Fuel 2000, 79, 323. (13) Bartle, K. D.; Taylor, N.; Mulligan, M. J.; Mills, D. G.; Gibson, C. Fuel 1983, 62, 1181. (14) Bartle, K. D.; Mills, D. G.; Mulligan, M. J.; Amaechina, I. O.; Taylor, N. Anal. Chem. 1986, 58, 2404. (15) Bartle, K. D.; Mulligan, M. J.; Taylor, N.; Martin, T. G.; Snape, C. E. Fuel 1984, 63, 1556. (16) Herod, A. A.; Bartle, K. D.; Carter, D. M.; Cocksedge, M. J.; Domin, M.; Kandiyoti, R. Rapid Commun. Mass Spectrom. 1995, 9, 1446. (17) Islas, C. A.; Suelves, I.; Morgan, T. J.; Herod, A. A.; Kandiyoti, R. Proc. 11th Int, Conf. Coal Sci., San Francisco, CA, 2001; ICCSPaper 215.
Energy & Fuels, Vol. 18, No. 3, 2004 779
molecular mass distributions of heavy hydrocarbon liquids, irrespective of their origins and of the structural features of their components. To the extent that such samples contain multiplicities of molecules with different and largely ill-defined structural features, it is important to minimize the level of structure dependence of elution times. The aim of the present study was to obtain calibration curves for two SEC columns, using NMP as eluent, and to test the level of dependence of the calibrations on chemical structures. The new calibration data cover a wider range of polystyrene (PS) standards compared to previous work and have been overlaid with elution times of a set of poly(methyl methacrylate) (PMMA) standards. The resulting PSPMMA calibration has been compared against elution times of several sets of standards: (i) a set of polysaccharides (PSAC), (ii) narrow polydispersity-known highmass polymers of various origins, (iii) some model compounds, and (iv) silica granules with graded particle sizes. The level of agreement between SEC and MALDI mass spectrometry has also been examined, using narrow SEC elution fractions of a coal tar pitch. Recorded elution times have been matched against MM values observed in MALDI-MS. The two techniques are independent and the level of agreement between them provides a yardstick with which to evaluate the results. Finally, calibrations arising from the present work have been used to discuss molecular mass distributions of three samples expected to contain high-mass molecules: the pyridine-insoluble fraction of a coal tar pitch, two petroleum asphaltene samples, and one sample of domestic soot.18 Experimental Section Analytical Size Exclusion Chromatography. Two 300 mm long, 7.5 mm i.d. polystyrene/polydivinylbenzene-packed columns (Polymer Laboratories, UK), labeled as Mixed-A (10 µm particles) and Mixed-D (5 µm particles) have been used. Operating conditions of the system have been previously described in detail.5-8,18,19 The Mixed-D column was operated at 80 °C with a Perkin-Elmer LC 250 isocratic pump, with eluent NMP pumped at 0.5 mL min-1. In addition, the Mixed-D column was used to fractionate the coal tar pitch into 0.5 min time-period fractions for examination by laser-desorption mass spectrometry. The larger porosity Mixed-A column was operated at room temperature with a flow rate similar to that of the Mixed-D column. SEC has the ability to be coupled on-line to a variety of detectors. A Perkin-Elmer LC 290 variable wavelength UVabsorbance detector and an Applied Biosystems diode array detector set at 280, 300, 350, and 370 nm were used routinely. An evaporative light scattering detector (Polymer Laboratories ELS 1000) has normally been used in series. This was particularly useful during calibration with PMMA standards and other work with samples with no or weak UV-absorption. On the other hand, UV-detection is useful for observing lowerboiling aromatics that normally evaporate with the eluent, during the ELS thermospray stage. Previous attempts to evaluate the utility of UV-fluorescence spectrometers have found no sensitivity to sample concentration above molecular masses of about 3000 u;20,21 also see below for further discus(18) Herod, A. A.; Lazaro, M. J.; Dubau, C.; Richaud, R.; Shearman, J.; Card, J.; Jones, A. R.; Kandiyoti, R. Energy Fuels 2000, 14, 1009. (19) Apicella, B.; Ciajolo, A.; Suelves, I.; Morgan, T. J.; Herod, A. A.; Kandiyoti, R. Comb. Sci. Technol. 2002, 174 (11-12), 345-359. (20) Li, C.-Z.; Wu, F.; Xu, B.; Kandiyoti, R. Fuel 1995, 74, 37-45. (21) Herod, A. A.; Zhang, S.-F.; Johnson, B. R.; Bartle, K. D.; Kandiyoti, R. Energy Fuels 1996, 10, 743-750.
780
Energy & Fuels, Vol. 18, No. 3, 2004
sion. Refractive index turns out to be less sensitive and requires high sample concentrations, which might cause the SEC column to overload. Preparative SEC. A Perkin-Elmer LC 250 isocratic pump was used to maintain an NMP flow rate of 2 mL min-1. The preparative SEC column (600 mm; 25 mm i.d. packed with polystyrene/polydivinylbenzene beads), used to fractionate the sample, was maintained at 85 °C and connected to a PerkinElmer LC 290 UV-absorbance detector at 450 nm. A single run of 100 min gave enough sample for each of the subsequent analyses. Two separate experiments have been carried out using this column. In the first, a series of fractions were collected over three-minute periods from just before the first excluded peak appeared. In the second experiment, narrower time-fractions corresponding to 10 s elution periods were obtained with three minute intervals between the fractions; this method 22 was used to reduce overlapping between successive elution fractions. MALDI Mass Spectrometry. A Fisons VG TOFSPEC mass spectrometer (VG Organic Manchester, UK) fitted with a UV-laser (337 nm) and a VAX 4000-base data system with OPUS software has been used.12,23,24 The spectrometer was used in the linear mode; the use of the reflectron was avoided in order to increase instrument sensitivity to high-mass materials. An accelerating voltage of 28 kV was used at maximum laser power. For the three-minute fractions, mercaptobenzothiazole (MBT) was used as matrix and added to the target in methanol solution after the fraction had been deposited in NMP and vacuum-dried since MBT evaporated in the vacuum oven under the conditions needed to remove NMP. During the series of runs with the 10 s interval fractions, sinapinic acid was premixed in NMP in a 4:1 matrix to sample solution volume ratio before deposition on the target. The target was then dried in a vacuum oven at 150 °C. The peak mass of each spectrum, Mp, was recorded and plotted against the elution time from SEC of the peak intensity of the fraction. The mass spectra were also evaluated by calculation of the standard deviation of the signal at the high mass limit of the spectrum.6 While only Mp values are considered in the present work, number and weight average masses were also calculated. One set of data was obtained using a Bruker Daltonics Reflex IV MALDI-TOF-MS, to examine the 0.5 min fractions from the analytical column by laser-desorption alone, with operation in the linear mode. Sample in NMP solution was added to the target and dried in a vacuum before analysis. The spectra were evaluated by the peak-intensity mass of the envelope of mass observed. Samples. Standard Polystyrenes (PS). Samples with Mp values of 1700, 3250, 7000, 11 600, 22 000, 275 000, 629 500, 1 950 000, 5 000 000 and 15 400 000 u were from Polymer Laboratories. Polydispersities of the standards were low: ∼1.04. Standard Poly(methyl methacrylate)s (PMMA). Samples with Mp values of 640, 2400, 4250, 7600, 20 200, 31 500, 50 000, 100 000, 120 000, 254 000, 772 000 and 1 520 000 u were from Polymer Laboratories. Polydispersities of the standards were similarly low: ∼1.04. Standard Polysaccharides (PSAC). Samples with Mp values 738, 5900, 11 800, 22 800, 47 300, 112 000, 212 000, 404 000, and 788 000 u were from Polymer Laboratories. Polydispersities were listed as ∼1.04 for all PSAC samples. Other Polymers. Other polymers were poly(ethylene oxide) (Mp 58 400, polydispersity 1.03) and poly(ethylene glycol) (Mp 4120, polydispersity 1.02) from Polymer Laboratories. Poly(N-vinylcarbazole (Mp 90 000)), polyvinylpyrrolidinones (Mp 3500 and 58 000), polyvinyl acetate (Mp 170 000), polyethylene adipate (Mp 10 000) from were from Acros Organics, supplied (22) Islas, C. A. PhD Thesis, University of London, 2001. (23) Domin, M.; Moreea, R.; Lazaro, M.-J.; Herod, A. A.; Kandiyoti, R. Rapid Commun. Mass Spectrom. 1997, 11, 1845-1852. (24) Lazaro, M.-J.; Herod, A. A.; Domin, M.; Zhuo, Y.; Islas, C. A.; Kandiyoti, R. Rapid Commun. Mass Spectrom. 1999, 13, 1401-1412.
Karaca et al.
Figure 1. Calibration data [log MM vs elution time] for the Mixed-A column.
Figure 2. Calibration data [log MM vs elution time] for the Mixed-D column. by Fisher Scientific of Loughborough, UK. Their polydispersities were not supplied but were expected to be greater than the polydispersity (1.04) of the PS and PMMA standards. The average molecular weights shown were obtained from SEC by the manufacturers. The polymer samples were dissolved in NMP by standing the mixture in an ultrasonic bath for up to 1 h, with further standing and manual agitation before injection. Three-Dimensional Standards. Three-dimensional standards included candle soot 18,19 of about 40 nm diameter, colloidal silica samples of diameters 22, 12, and 9 nm, and fullerene (a mixture of C70 and C60) of diameter ∼1 nm. Other Materials Used. The coal tar pitch has been used extensively in previous work;7,12,22 pyridine-insolubles were prepared by repeated washing of the pitch to leave the insoluble material, without recourse to chromatographic media. A vacuum residue from a sample of Forties crude petroleum has been described previously.25 A sample of soot from the domestic combustion of coal in an open fire was obtained using a standard procedure for operating the fire. A sample of a petroleum vacuum residue of unknown source was obtained from Shell, Amsterdam, and the asphaltenes were prepared as the heptane-insoluble fraction by repeated extraction.
Results and Discussion Calibration of SEC Columns. Figures 1 and 2 present calibration data [log MM vs elution time] for the Mixed-A and Mixed-D columns, respectively. Three sets of polymer standards with known MMs were used: (25) Suelves, I.; Islas, C. A.; Millan, M.; Galmes, C.; Carter, J. F.; Herod, A. A.; Kandiyoti, R. Fuel 2003, 82, 1-14.
Calibration of Size Exclusion Chromatography Columns
polystyrenes (PS), poly(methyl methacrylate)s (PMMA), and polysaccharides (PSAC), the latter with masses up to 788 000 u. The ELS detector was used when working with the PMMA and PSAC samples, which do not contain UV-absorbing chromophores. The linear part of the calibration plot corresponds to the molecular size range resolved by a particular SEC column. At the highmass end, the void volume represents the shortest possible elution time. Void volumes for both the polystyrene and the poly(methyl methacrylate) standards were found to be above 1.85 million u. Exclusion limits of the two columns are indicated by the onset of departure from linearity at the short elution time end of the line. This occurred at ∼14 min for the Mixed-A column, just short of the largest commercially available polystyrene standard of 15 million u (Figure 1) and ∼11 min for the Mixed-D column at a mass between ∼200 000 and 400 000 u in Figure 2.10,12,19 Parameters were determined for these calibration graphs on Mixed-A and Mixed-D columns in the equation log10 molecular mass ) A - B. elution times (x min) for the linear regions are as follow:
Mixed-A PS
y ) 14.15 - 0.5533x
Figure 3. Elution times of seven polymeric MM-standards vs the PS-PMMA-calibration line of the Mixed-A column: 1-polyvinyl acetate 170 000 u; 2-poly-N-vinylcarbazole 90 000 u; 3-poly(ethylene oxide) 58 400 u; 4-poly(vinylpyrrolidone) 58 000 u; 5-polyethyleneadipate 10 000 u; 6-poly(ethylene glycol) 4120 u; 7-poly(vinylpyrrolidone) 3 500 u.
R2 ) 0.9993
PMMA
y ) 14.539 - 0.5747x R2 ) 0.9984
PSAC
y ) 9.5362 - 0.3116x R2 ) 0.9976
Mixed-D PS
Energy & Fuels, Vol. 18, No. 3, 2004 781
y ) 9.4024 - 0.3677x R2 ) 0.997
PMMA
y ) 9.6756 - 0.3884x R2 ) 0.9951
PSAC
y ) 7.7816 - 0.2826x R2 ) 0.9744
Figures 1 and 2 also show that in both columns, the elution times of PS and PMMA standards up to masses of ∼1.5 million u were statistically nearly indistinguishable.17 In view of the structural differences between PS and PMMA, these findings are encouraging. However, the elution times of the highly oxygenated polysaccharide (PSAC) samples traced a different curve. To the extent that these differences are observed, the present column/eluent combination cannot be said to operate in a fashion independent of chemical structure. Considering Figure 1 in detail, closer agreement was observed between the three calibration lines at longer elution times (smaller MMs). The data were nearly indistinguishable at 19-20 min, but differed by a factor of ∼1.8 at 18 min [∼15.5 k u vs 8.5 k u] and by a factor of between 12 and 15 at 14.5 min [∼1.34 (PS) or 1.61 (PMMA) M u vs 0.104 (PSAC) M u]. To put these elution times into context, 18 min corresponded to the forward edge of the resolved peak in Figure 6a, while 14.5 min was on the low-mass tail of the excluded peak. We also need to put the differences between the PSPMMA and the PSAC lines into a broader context. Figure 3 shows the elution times of seven polymeric MM-standards plotted alongside the PS-PMMA-calibration line of the Mixed-A column. The interest in these samples arises because of their considerable structural differences from PSs, PMMAs, and PSACs. Of the five polymer standards, four eluted within half a minute of
Figure 4. (a) Elution times of eleven model compounds plotted alongside the polystyrene calibration line; identities of the compounds given in text. (b) Elution times of three sets of model compounds plotted alongside the PS-PMMA calibration line of the Mixed-D column.
the calibration line, despite polydispersities greater than those of the PSs and PMMAs. Only poly-N-vinylcarbazole (90 000 u) eluted ∼1.2 min early, indicating nearly an order of magnitude difference with the calibration line, between ∼100 000 and 1 million u. However, the oxygenates poly-ethylene -oxide and -glycol eluted early. These results show the necessity for caution in the levels of confidence attached to quantitative estimates of molecular masses of structurally less well-defined samples, when using the PS-PMMA calibration line. However, the structural differences between hydrocarbon liquids (our main focus) and the PSAC-polymers are quite marked. The smaller deviations in Figure 3
782
Energy & Fuels, Vol. 18, No. 3, 2004
Figure 5. Plots of log (Mp) vs elution time in the Mixed-D column, for three sets of fractions of pitch collected from preparative or analytical SEC, with measurement of Mp values for fractions by MALDI-MS or LD-MS.
Figure 6. (a) Size exclusion chromatograms of the pyridineinsoluble fraction of a coal tar pitch, obtained using the Mixed-A column. (b) Size exclusion chromatograms of the pyridine-insoluble fraction of a coal tar pitch, obtained using the Mixed-D column.
suggest that to a first approximation, we might consider deviations between the PSAC and PS-PMMA lines as an upper limit to levels of error arising from the structure-dependent variations in the elution behavior of these samples. The results also show that below 15 k u, even the MMs of PSACs may be estimated to within a factor of ∼2-2.5, and that confidence in quantitative determinations appear subject to increasing uncertainty with increasing MM. These low levels of precision of mass measurement bear no relation to the mass measurement achieved by working with GC-MS and other high-precision methods of analysis. However, many
Karaca et al.
materials lie outside the reach of these well-known techniques, and SEC is useful in attempting to estimate the ranges of MMs of these samples. Elution Times of Some Low-Mass Model Compounds. The mass analysis range quoted for the Mixed-A column by the manufacturers is between 1000 and 10 million u. This column would not normally be selected for resolving mixtures of relatively small molecules. It is necessary, however, to understand the elution behavior of small molecules in order to interpret signals observed at long elution times, during the characterization of complex mixtures. Figure 4a shows the elution times of eleven model compounds plotted alongside the polystyrene calibration line. These were the following: toluene [92 u; no. 1], benzene [78 u; no. 2], acetone [58 u; no. 3], 9-methyl anthracene [192 u; no. 4], fluoranthene [202 u; no. 5], 5,6-benzoquinoline [179 u; no. 6], perylene [252 u; no. 7], chrysene [228 u; no. 8], 2,3-dimethylindole [145 u; no. 9], rubrene [532 u; no. 10], and pyrogallol [126 u; no. 11]. In parallel with the behavior of the highly oxygenated PSAC samples, pyrogallol eluted earlier than expected. Many of the samples eluted at about 0.5 min longer than predicted by extrapolating the PScalibration line; acetone and anthracene eluted with slightly less than ∼1 min delay. The largest deviations of nearly 2 min late were observed for the smallest aromatic compoundss benzene and toluene. The latter effect appears related to greater diffusivities of smaller polycyclic aromatic (PCA) compounds within the column packing, due to their compact shapes. Pyrogallol eluted early as did the oxygenated polymers discussed above. Figure 4b presents analogous data for three sets of model compounds plotted alongside the PS-PMMA calibration line of the Mixed-D column. The full lists of PCA compounds (squares), oxygenated compounds (cross), and nitrogen-bearing compounds (triangles) have been given.10 The PCA standards (including cataannelated, pericondensed, nonplanar, and some alkyl-substituted species) were found to behave much as they did in the Mixed-A column, systematically eluting at longer times than expected from their molecular masses. As above, the effect appears related to high diffusivities due to the compact shapes of these molecules. In calculating average MMs of complex mixtures, the deviation from linearity of smaller PCA species needs to be taken into account. The nitrogen-bearing compounds (ranging from pyridine to the dye, alcian blue, of MM 1086 u) showed more symmetric scatter about the PS-PMMA calibration line than the PCA-group, while elution times of oxygenates (ranging from acetone to stearyl alcohol, MM 270 u) were more heavily weighted toward shorter elution times, reflecting the behavior of the PSAC standards and other oxygenates. Only the mixture of C60 and C70 fullerenes with their three-dimensional structures were found to show anomalous behavior, eluting near the exclusion limit of both the Mixed-A and the Mixed-D columns [cf. ref 10]. This constituted the greatest single departure from the PS-PMMA line and will be discussed below in some detail. MALDI-Mass Spectrometry of Successively Eluting SEC Fractions of a Coal Tar Pitch. Two sets of samples have been prepared using the same coal tar pitch as fractions exiting from a preparative scale SEC column.
Calibration of Size Exclusion Chromatography Columns
The first set consisted of successively recovered 3 min elution fractions; the second set was made up of 10 s fractions collected every 3 min. The elution times of these fractions were determined in the analytical Mixed-D column. A third set of fractions was prepared using 30 s fractions eluting from the same analytical Mixed-D column. MMs of all three sets of fractions were evaluated by mass at peak intensity (Mp) in the MALDImass spectra and LD-mass spectra. For fractions of narrow molecular dispersity, the mass at peak intensity, Mp, is expected to be closer to the number average mass, Mn. Mercaptobenzothiazole was used as matrix for the 3 min samples, and sinapinic acid for the 10 s fractions. The third set was examined by LD-MS (i.e., no matrix). Figure 5 presents plots of log (Mp) vs elution time in the Mixed-D column, for all three sets of fractions. The data were internally consistent and reproducible, despite accumulation of results at different times, from three separate SEC experiments, using two different matrixes, two different mass spectrometers, and several spectrometer operators. Later-than-expected elution times were observed for smaller molecular mass samples in all three sets, in parallel with shifts observed for known model compounds in Figure 4a and 4b. The MALDI-derived Mp values were observed to flatten out for samples eluting at shorter elution times, i.e., MMs above 3000 u; this is being investigated. Thus, for molecular masses of pitch fractions up to about m/z 3000, reasonably good agreement has been found between the PS-PMMA calibration of the Mixed-D column and Mp values arrived at by MALDI and LD-MS. We have already noted that (i) the largest deviations from the PS-PMMA line were found in the case of oxygenated samples, (ii) that even for these samples, MMs may be estimated to within a factor of ∼2-2.5 below about 15k u, and that (iii) other structural features (e.g., ring-embedded nitrogen) are likely to give rise to smaller errors. Put in this context, the agreement observed between findings from SEC and MALDI-MS up to about 3000 u in Figure 5 suggests that up to this limit and perhaps a little beyond, SEC as defined in this work may be considered as a quantitative tool, the accuracy of the measurement being subject to greater uncertainty with increasing molecular mass. Size Exclusion Chromatography of Heavy Hydrocarbon Liquids. Size exclusion chromatograms of three typical samples will be presented and discussed in terms of the calibration data outlined above. The Pyridine-Insoluble Fraction of a Coal Tar Pitch. Figures 6a and 6b present size exclusion chromatograms of the pyridine-insoluble fraction of a coal tar pitch, obtained using the Mixed-A and Mixed-D columns, respectively. Both chromatograms showed the characteristic two-peak behavior first observed by Lafleur and Nakagawa.4 In both sets of chromatograms, the early eluting peak represents material at and near the exclusion limit of the column. In the Mixed-A column this corresponds to ∼14.5 min, i.e. approximately 1.3 M u according to the polystyrene calibration. In the case of the Mixed-D column, the center of the excluded peak at ∼10 min corresponds to about 530 000 u according to the PS-calibration and 710 000 by the PMMA calibration. The levels of confidence that may be attached to these very large apparent molecular masses will be discussed below.
Energy & Fuels, Vol. 18, No. 3, 2004 783
Figure 7. SEC chromatograms of the heptane-insoluble (curve 1) and heptane-soluble (curve 2) fractions of the vacuum distillation residue of a Forties crude oil.
Comparing chromatograms from the two columns, slightly better resolution was obtained from the largerporosity Mixed-A column for material appearing in the excluded region. However, there was no visible shift of material from the excluded to the retained part, compared to the Mixed-D chromatogram; a minor peak at 16.5 min in Figure 5a provided the only difference between the performance of the two columns. A Petroleum Asphaltene. Figure 7 presents the chromatograms of the heptane-soluble and heptane-insoluble fractions of the vacuum distillation residue of a Forties crude oil. These chromatograms were obtained using the Mixed-D column, with the UV-detector set at 350 nm. They show a pronounced excluded-peak for the heptane-insoluble fraction centered at 9-10 min. The material resolved by the column gave a peak centered around 18 min. A much reduced excluded peak was observed for the heptane-soluble fraction. Similar data, not shown, have been generated for similar fractions of a variety of other vacuum residues. Vacuum residues and their heptane-insoluble fractions are known to contain more aromatic material than lighter fractions of the same crude oil. Clearly, only those parts of the petroleum-derived sample capable of absorbing UV light will show up in the chromatograms of Figure 7. In any case, alkanes do not appreciably dissolve in NMP. Calibration data have been obtained using alkane standards in a separate SEC system, using heptane as eluent. These data are being prepared for publication. Comparison of these asphaltenes with the pyridineinsoluble pitch fraction is revealing. First, petroleum asphaltenes are more amenable to analysis by pyrolysisGC-MS, which shows relatively small polycyclic aromatic (PCA) groups [compare data in refs 25 with those in ref 26]. There are also clear differences in solubility; heptane-insoluble fractions of petroleum vacuum residues dissolve completely, or nearly completely in pyridine. It may be noted, however, that the elution times of peaks in Figure 7 are quite close to those in Figure 6(a,b). The heptane-insoluble part constituted about 2.1% of the Forties vacuum residue (Figure 7a) and only 1.6% of Sample 2, whereas the pyridine-insoluble fraction of the pitch constituted about 15% of the whole sample. Prior fractionation has proved useful for observing the heptane-insolubles by SEC; the chromatograms of the original vacuum residues were quite similar to those of the heptane-soluble fractions. Clearly, these vacuum residues were prepared from relatively light crudes, where heptane-insolubles make up a modest proportion
784
Energy & Fuels, Vol. 18, No. 3, 2004
Karaca et al.
Table 1. Comparison of Weight-Average Molecular Mass Estimates by SEC and by Static Light Scattering
sample NMP mobiles of a low-temperature tara NMP mobiles of a coal tar pitcha candle soot
polydispersity
SEC signal begins at min
0.4-1.0
600 000
664
11
4.0-10.0
2.2 × 106
980
9
3.7 × 106
0.01-0.05
1.1 × 109
4300
6
1.7 × 109
concentration mg/mL
Mw by SEC u
( extrapolated)
a
Mw by static light scattering u 610 000
Heaviest fraction during the column chromatographic separation [e.g., ref 40].
Figure 8. (a) Domestic soot sample examined using the Mixed-A column. (b) Domestic soot sample examined using the Mixed-D column.
of the whole crude oil. However, the increasing production of heavier crude oils worldwide makes asphaltene processing an issue that is likely to affect virtually all refinery conversion processes. Recently, characterization data were reported on three crude oils with heptaneinsoluble contents of 1, 3.6, and 14.1%, respectively.27 Soot from a Domestic Chimney. A domestic soot sample has also been examined using the Mixed-A and Mixed-D columns, and chromatograms are shown in Figures 8a and 8b, respectively. The earliest eluting material was observed at 11 min on the Mixed-A and at 7 min on the Mixed-D column. As will be explained below, this appears to correspond to soot particle diameters of the order of about 40 nm. The molecular masses of these species are unknown, but in the Mixed-A column, the forward edge of the chromatogram shows elution at somewhat earlier than a polystyrene standard of mass 15 M u. Extrapolation of the polysty(26) Herod, A. A.; Islas, C. A.; Lazaro, M.-J.; Dubau, C.; Carter, J. F.; Kandiyoti, R. Rapid Commun. Mass Spectrom. 1999, 13, 201-210. (27) Ancheyta, J.; Centeno, G.; Trejo, F.; Marroquin, G.; Garcia, C. A.; Tenorio, E.; Torres, A. Energy Fuels 2002, 16 1121-1127.
rene calibration indicated a molecular mass for the 7 min peak as somewhat greater than 109 u.19 Table 1 shows that measurement by static light scattering gave 1.7 × 109 u for the same sample. While there appears to be broad agreement between the two techniques for this and two other heavy fractions [ref 22; cf. Table 1], the magnitudes of these numbers require scrutiny (see below). At the other end of the chromatogram, some of the lighter soot material corresponding to PCA hydrocarbons eluted near/after 22 min from the Mixed-A column. It may also be noted (Figure 8a,b) that the retained material (second peak) in the Mixed-A column constitutes a greater proportion of the whole sample than on the Mixed-D column. It appears that some of the material excluded from column porosity in the Mixed-D column crossed into the retained region of the largerporosity Mixed-A column. This shift from elution in the excluded zone in the Mixed-D column to the resolved zone in the Mixed-A column is observed more clearly in chromatograms recorded by ELS (data not shown); part of the sample appears to switch across the valley of low intensity. The data of Figure 9a,b, discussed next, suggest that these materials are well into the size ranges where the samples might be conforming to the behavior of 3-dimensional particles. Figure 9b shows that the material eluting at 10-11 min in the Mixed-D column corresponds to 3-D material in the vicinity of 1 nm diameter. Solid Silica Samples with Known Diameters. In discussing Figure 4b, we mentioned that a sample of mixed C60 and C70 fullerenes had eluted near the exclusion limit of the column. In view of their MMs of 720 and 840 u, these materials could be classed as the only massive outliers with respect to all the calibration curves described above. Equally, with diameters ∼1 nm, the shapes of the fullerenes make them the only species in the range of compounds tested, with as great a discrepancy between molecular mass and molecular size. Clearly, much of size exclusion chromatography depends on a broad correlation between molecular mass and molecular size. We also reported earlier 18,19 that soot samples could be fractionated by filtering out material caught by a 20 nm filter. When the material retained on the filter was redissolved in NMP, the earlier chromatograms could be reproduced. Clearly, the samples we have been putting through these SEC columns contain material with diameters in the tens of nanometers. These findings suggested that there might be a direct relationship between elution times and the sizes of more fully 3-dimensional objects in these columns. A set of colloidal silica samples (Nissan Chemical Industries Ltd. of
Calibration of Size Exclusion Chromatography Columns
Figure 9. (a) Plot of log (particle diameter, nm) vs elution times, in the Mixed-A column. (b) Plot of log (particle diameter, nm) vs elution times, in the Mixed-D column.
Houston, Texas) of diameters 22, 12, and 9 nm have been used to test the relationship between actual particle diameters and elution time. Figure 9(ba,b) show plots of log (particle diameter) vs elution times, in the Mixed-A and Mixed-D columns, respectively. Both plots appear to conform to linear behavior. The fourth datapoint pairs on the two diagrams belong to the mixture of C60 and C70 fullerenes, eluting at ∼13.5 min in the Mixed-A and ∼10.5 min in the Mixed-D columns. Figure 9a,b indicates that the sample of mixed fullerenes has eluted according to molecular size rather than molecular mass. Considering the fullerene data in a little more detail, in the Mixed-D column, the fullerene mixture gave two peaks, shown at ∼1 nm diameter; it is likely that the later-eluting peak corresponded to C60 and the earlier peak to C70, which in this calibration would have an indicated diameter of about 2.5 nm. On the Mixed-A column, two peaks were observed during initial runs but only one in the repeat run. All these data points have been shown in Figure 9a. We note that the possibility of reaction between NMP and fullerene in solution has been raised,28 and it is not clear what effect this would have on the elution time of fullerene. These data still leave unclear the relationship between diameter and molecular shape. The materials selected for this calibration were spherical. It is interesting to contemplate whether rodlike molecules of similar diameters would elute in similar fashion. In the absence of suitable standards, that possibility cannot (28) Yevlampieva, N. P.; Biryulin, Yu. F.; Melenevskaja, E. Yu.; Zgonnik, V. N.; Rjumtsev, E. I. Colloids Surf, A: Physicochemical and Engineering Aspects 2002, 209, 167-171.
Energy & Fuels, Vol. 18, No. 3, 2004 785
be investigated. However, the consistency of the fullerene data with the linear relationship of silica particles suggests, at least to a first approximation, that for samples with well-defined 3-D conformations, the effect of particle density on elution times may be negligible. The observed internal consistency of these data also addresses concerns that fullerenes might be eluting in the form of aggregated clusters. In fact, we might already have discounted that possibility in terms of the extreme levels of sample dilution, as calculated below. How to Interpret Results from Size Exclusion Chromatography? The use of NMP as eluent in SEC is principally due to Lafleur and Nakagawa,4 who were also the first to report the two-peak behavior observed, among others, in Figures 6-8 of this work. As already explained, the earlier of the two peaks corresponds to rather large apparent molecular masses. Starting with ref 4, a long line of investigators have interpreted such findings in terms of the aggregation of polar material, making the sample material appear much larger than its actual molecular size. Polar carbonaceous materials may well form aggregates at high concentrations and/ or when solvent power is low. In the present context, however, what is important to determine is whether aggregation may have distorted the measurement, under the conditions of size exclusion chromatography, and indeed, under conditions used in several other other measurement techniques. The Effect of Concentration on Sample Aggregation. Sheu and co-workers29 have measured surface tension as a function of asphaltene concentration in nitrobenzene and in pyridine. In both solutions (25 °C), discontinuities were observed in the surface tension as solute concentrations increased above about 0.05 wt % [equivalent to log (concentration) ≈ -1.3; concentration as wt %]. By analogy with the behavior of surfactant solutions, these authors considered the observed discontinuity as the “critical concentration” above which micelles would aggregate. While analogy does not constitute proof, the argument appears reasonable. It is interesting to compare this critical micelle concentration with an estimate of concentration in (the stronger solvent) NMP of sample under the excluded peak during SEC. Assuming we inject 20 µL of a 1% solution (∼0.2 × 10-3 g sample) and that 30-50% of the sample appears under the excluded peak, and that the peak elutes in about 3 min, we have
0.1 × 10-3 g sample in ∼1.5-2 mL (i.e., in ∼1.52.0 g) of NMP solvent The concentration of sample in the stream going through the detector may then be calculated as
∼0.5-0.66 × 10-4 g/g or 5-6.6 × 10-5 g/g So the concentration as weight percent turns out to be about 5-6.6 × 10-3 %. Sample concentrations in SEC appear to be 1 order of magnitude lower than the critical micelle concentrations of about 0.05%. The use of the much stronger solvent NMP [dissolves the 15% of pitch (29) Sheu, E. Y.; De Tar, M. M.; Storm, D. A. In Asphaltene particles in fossil fuel exploration, recovery, refining and production processes; Sharma, M. K., Yen, T. F., Eds.; Plenum Press: New York & London, 1994; p 118.
786
Energy & Fuels, Vol. 18, No. 3, 2004
undissolved in pyridine] must be taken as an additional factor impeding sample aggregation under these conditions. It is also worth noting that when fractions eluting at the exclusion limit of the Mixed-D column were reinjected, the much-diluted sample eluted at the same time as before.10 Any aggregates would have been expected to show some disaggregation and shift to longer elution times. Seeking Evidence for Aggregation or the Lack of It. It is difficult to prove the non-existence of a phenomenon. There is, nevertheless, much evidence pointing in the same direction, strongly suggesting that SEC and determinations by several other methods have been made under conditions where aggregation could not have been a predominant phenomenon. It is useful to review these points briefly, before attempting to look for explanations for the very large apparent molecular masses observed. a. Probe-MS and LD-MS Related Findings. Figure 6 in ref 10 presents probe-MS data of coal tar pitch fractions, successively eluted from the Mixed-D analytical column. Each fraction represented sample collected during 1 min of elution, between 8 and 24 min. The results obtained for the first 11 fractions (out of a total of 17 fractions) did not show the presence of any trace of material other than NMP and NMP oxidation fragments. By contrast, later eluting fractions showed light PCA materials. These data complement the absence of signal during heated probe-MS from the pyridine-insoluble fraction of a coal tar pitch, with the exception of traces of pyridine and a phthalate contaminant.30 The data clearly showed a lack of material observable below 500 u. In both cases (i.e., SEC-elution fractions and the pyridine-insolubles), sample evaporation would have been expected to occasion some dis-aggregation and the appearance of at least some small polar, possibly aromatic material. In another study,31 LD-mass spectrometry also showed a very weak signal for the early eluting fractions, although the longer eluting fractions gave a significant signal. While these observations cannot prove the absence of molecular aggregation, they show no evidence of any such aggregation, under conditions where some indications should have been observedsif the sample had been composed of aggregates. In the case of SEC fractions, what the mass spectroscopic data do give is a clear indication of entirely different structural makeup between the early and late eluting fractions. b. Pyrolysis-GC-MS Related Findings. Ref 26 shows pyrolysis-GC-mass spectra of three coal tar pitch fractions separated by planar chromatography. The three spectra did not show similar species. In fact, the structural properties of the very few compounds identified in the two heaviest fractions were entirely different from the lightest fraction, which mainly showed aromatics up to the detection limit of the technique. Analogous pyrolysis-GC-mass spectra from a coal extract32 and a low-temperature coal tar33 all showed (30) Begon, V.; Islas, C. A.; Lazaro, M. J.; Suelves, I.; Herod, A. A.; Dugwell, D. R.; Kandiyoti, R. Eur. J. Mass Spectrom. 2000, 6 39-48. (31) Islas, C. A.; Suelves, I.; Millan, M.; Apicella, B.; Herod, A. A.; Kandiyoti, R. J. Sep, Sci. 2003, 26, 1422-1428.
Karaca et al.
Figure 10. SEC chromatograms by UV-A and -F of a petroleum asphaltene; the heavy line is UV-fluorescence with excitation at 350 nm and emission at 450 nm while UV-A was at 300 nm.
differences in structural features between the smaller and larger mass material. Once again, the data show no evidence of aggregation, but clear indications of entirely different structural makeup between fractions with different apparent molecular massessas determined by SEC. c. UV-Fluorescence Spectroscopy. Numerous samples of heavy hydrocarbon liquids have been fractionated. When these fractions, showing sequential differences in molecular mass distributions by SEC, were examined by UV-fluorescence spectrometry, the resulting spectra were consistently distinct from one another. The samples characterized included a coal tar pitch12 and a Point of Ayr coal extract32 fractionated by planar chromatography, two petroleum residues25 fractionated by column chromatography, and a low-temperature coal tar fractionated by both column and planar chromatography.33 These data clearly show structural differences between separated fractions showing progressively increasing molecular mass distributionssas determined by SEC. It must be noted that UV-fluorescence is blind to coalderived material above the 3000 u level. This was found when a UV-fluorescence and a UV-absorption spectrometer were used as detectors in SEC. While the earliest signal by UV-absorption could vary according to the properties of the sample, the shortest elution times where the UV-fluorescence detector could show signal was static and corresponded to masses below 3000. The experiments have been repeated over a wide range of emission and excitation wavelength combinations, without improving detection by UV-fluorescence at shorter elution times20,21 (read greater molecular masses). Similar results on petroleum asphaltenes are being prepared for publication, and Figure 10 shows UV-A at 300 nm and UV-F (excitation at 350 nm, emission at 450 nm) of a petroleum heptane-insoluble asphaltene. d. Differences in Trace Element Contents of ColumnChromatography Fractions. Fundamentally different structural profiles may also be discerned by comparing trace element contents of fractions successively eluted by column chromatography. We reported34 on trace element contents of column-chromatography fractions (32) Islas, C. A.; Suelves, I.; Carter, J. F.; Herod, A. A.; Kandiyoti, R. Rapid. Commun. Mass Spectrom. 2000, 14, 1766-1782. (33) Islas, C. A.; Suelves, I.; Carter, J. F.; Li, W.; Herod, A. A.; Kandiyoti, R. Rapid Commun. Mass. Spectrom. 2002, 16, 774-784. (34) Herod, A. A.; George, A.; Islas, C. A.; Suelves, I.; Kandiyoti, R. Energy Fuels 2003, 17, 862-873.
Calibration of Size Exclusion Chromatography Columns
Energy & Fuels, Vol. 18, No. 3, 2004 787
of a coal tar pitch, a coal liquefaction extract, and a lowtemperature tar. The solvents used for sequential extraction were acetonitrile, pyridine, and NMP; mass balances were best if chromatographic media were avoided in the fractionation. In all three samples, the greater part of the trace elements analyzed have been found in fractions shown by SEC to contain the largest molecules. Structural data from this work and previous characterizations suggest that within larger molecules, increasingly large PCA ring systems are being held together by a variety of aliphatic and alicyclic bridging structures. In the absence of mineral matter or other solids, it is thought that the high trace element concentrations represented organic associations within these complex molecules. Within the framework of these data, agglomeration would only appear possible between molecules already shown to be of large-MM by SEC. Addition of Salts to Eluent NMP. The addition of salts to NMP, when used as eluent in SEC, has been claimed to help dissociate molecular aggregates. In one set of experiments, LiBr was added to the eluent NMP and found to shift the SEC chromatograms of coal-derived liquids to longer retention times (smaller apparent masses). These shifts have been attributed to the dissipation of noncovalent binding forces, causing disaggregation of polar clusters, which would otherwise have appeared at retention times appropriate to larger molecular masses.35,36 More recently, similar claims have been made for tetrabutylammonium acetate37 (TBAA). To examine some of these claims, experiments can be designed to effectively decouple effects due to polarity and to molecular mass. One obvious device is to use nonpolar samples. SEC chromatograms of a fullerene mixture, a naphthalene mesophase pitch and its fractions, separated by planar chromatography, all clearly showed significant signal under the excluded peak, inevitably due entirely to nonpolar material. However, when LiBr was added to the eluent NMP, precipitation of solute out of solution and shifts of chromatograms to longer retention times were observed.5 These shifts to smaller apparent molecular masses were clearly unrelated to sample polarity, since the samples being used were nonpolar. The addition of LiBr to NMP was also observed to cause a partial breakdown of the size exclusion mechanism. Chromatograms extended to well beyond the permeation limit of the column, showing similarities with results from experiments with eluents of insufficient solvent strength (e.g., THF, chloroform). In the types of columns presently in use, dosing LiBr or TBAA into NMP appears to give rise to interactions analogous to loss of solvent power for coal-derived solutes and the promotion of surface effects. These surface effects did not, however, alter elution times of the nonpolar polystyrene standards.5 It would seems straightforward to run a few blanks with known compounds such as coronene, to assess how such additives affect the functioning of SEC columns.
The effects of LiBr and TBAA addition to solvents in planar chromatography were of an expected nature. Both TBAA and LiBr appear to enhance solvent power and assist solvents, e.g., acetonitrile, to mobilize larger mass material. The latter effect simply appears to reflect the polarity of LiBr and of TBAA. There was no evidence that these two reagents were dissociating samples; quite to the contrary, they helped displace from the polar silica material with greater apparent MMs. This work is being prepared for publication. The Possible Three-Dimensional Nature of Material under the Excluded Peak. Data presented above do not support the idea that apparently large mass material, appearing under the excluded peak, might have been formed by the aggregation of smaller (probably polar) molecules. In 1995, we wrote:38 “There is indeed nothing surprising in the detection of such large molecular masses in coal-derived products by mass spectrometric techniques: peptidessthermally far more delicate molecules than most coal-derived speciesswith MMs up to 300 000 u have been identified in biopolymers by MALDI-MS.39” This still holds true. There is much work in both the fossil fuel field and elsewhere discussing large (say >3000 u) molecular mass material. However, the characteristic, twin-peak shapes of SEC chromatograms of fossil fuel-derived samples (e.g., Figures 6-8) deserve comment. A recognizable “valley” is observed between the peak for excluded material at short elution times and the peak for material resolved by column porosity. When a reasonably dilute sample is injected, signal levels between the two peaks usually decline to near-zero intensity. The signal level in the valley is nonzero when larger concentrations of sample are injected. In these cases, the front end (excluded zone) of the chromatogram begins to overload. These difficulties have been overcome by Islas,22 who recovered sample corresponding to the elution time of an overloaded “valley” and re-injected to find the sample redistribute itself between the excluded and retained peaks, with zero signal intensity in the valley. Furthermore, the proportion of a sample of pitch excluded in a Mixed-E column remained similar to the proportion of excluded material from the same sample in a larger porosity Mixed-D column. (Some shifting between the two peaks could be observed between Figure 8a and 8b above). There does appear to be a mass barrier, however, above which, sample quickly shifts forward to the excluded peak irrespective of column characteristics. It seems difficult to interpret these observations, and the following is offered as no more than speculation. In general, the effective molecular size (rather than the molecular mass) is the factor that determines the elution time of molecules. Size exclusion chromatography works well when molecular mass and molecular size correlate. However, we know of instances when molecular size is determined by the particular conformation of the molecule in a way that is not closely related to molecular mass. We have seen how fullerenes eluted in the excluded zone, at much earlier times than would be expected from their MMs. By analogy, changes of
(35) Takanohashi, T.; Iino, M.; Nakumara, K. Energy Fuels 1994, 8, 395. (36) Masuda, K.; Okuma, O.; Knaji, M.; Matsumara, T. Fuel 1996, 75, 1065. (37) Chen, C.; Iino, M. Fuel 2001, 80, 929.
(38) Herod, A. A.; Kandiyoti, R. Fuel 1995, 74, 784-786. (39) Hillenkamp, F.; Keras, M.; Beavis, R. C.; Chait, B. T. Anal. Chem. 1991, 63 (24), 1193A. (40) Islas, C. A.; Suelves, I.; Li, W.; Morgan, T. J.; Herod, A. A.; Kandiyoti, R. Fuel 2003, 82, 1813-1823.
788
Energy & Fuels, Vol. 18, No. 3, 2004
shape from predominantly planar conformations (when molecules appear under the retained peak), to a threedimensional shape for somewhat larger molecules may be the factor responsible for splitting the chromatogram into two zones separated by a valley. The folded-over molecules would also show up as very large in static light scattering. Thus, it is possible that already large molecules (e.g., of an asphaltene or pitch sample of ∼several 1000 u) appear as much larger mass moleculessdue to a change in shape, once the molecules get above a certain size. In the pyridine-insoluble fraction of a coal tar pitch, for both columns, the data of Figure 9a,b suggest that the diameters of the material under the excluded peak runs from about 5 nm down to less than 1 nm. In soots and other samples where the early eluting peak was observed at 6-7 min, an analogous step-change in structure may be responsible for the valley of low or zero intensity between the excluded material and the peak at the void volume. Soot particles were observed to consist of material with diameters greater than 20 nm up to about 50 nm. The use of spherical standards has suggested that similarly shaped materials might elute within the exclusion zone of either column, but proof of such shapes in coal- or oil-derived materials is lacking. While the broad brush of the “aggregates” label has not explained much of our experimental observations, the molecular masses, structures, and conformations of material appearing under the excluded peak in size exclusion chromatography require further careful study. Conclusions 1. The polystyrene (PS) and poly(methyl methacrylate) (PMMA) calibration curves were nearly indistinguishable. A set of polysaccharides (PSAC) eluted earlier and gave a separate calibration line. More generally, the largest deviations from the PS-PMMA line were found for oxygenated samples. By contrast, numerous polymers with quite different structural features gave smaller deviations from the PS-PMMA calibration. To a first approximation, it seems reasonable to consider deviations between the PSAC and PS-PMMA lines as an upper limit to errors arising from structure-dependent variations in this SEC system. 2. Good agreement was observed up to about m/z 3000, when elution times of pitch fractions in the Mixed-D column were plotted against peak mass values arrived at by MALDI and LD-MS, alongside the PSPMMA calibration curve. The two techniques are independent and the level of agreement between them provides a reliable yardstick with which to evaluate these measurements. 3. Using these calibrations, MMs of oxygenated samples may be estimated to within a factor of ∼22.5, below 15 000 u. Other structural features gave rise to smaller deviations. The agreement observed between findings from SEC and MALDI-MS up to about 3000 u suggests that up to this limit and perhaps a little beyond, SEC may be considered as a quantitative tool, the accuracy of the measurement being subject to greater uncertainty with increasing molecular mass. 4. The often-made assumption that apparently highmass materials in SEC reflect the aggregation of small polar molecules has been examined. Polar materials do
Karaca et al.
aggregate at high concentrations and/or when solvent power is low. The question that must be answered is whether aggregation may have distorted MMs under prevailing measurement conditions. Sample concentrations during SEC appear to be at least 1 order of magnitude smaller than observations of critical micelle concentrations and in a more powerful solvent (NMP) than those previously used. Evidence from several analytical techniques has also been reviewed. The data provide clear indications of entirely different structural makeup (e.g., aromaticity; nature of fragments in mass spectrometry; trace element concentration) between fractions with different apparent molecular massessas determined by SEC. 5. UV-fluorescence is blind to coal-derived material and petroleum asphaltenes above the 3000 u level. This is clearly observed when UV-fluorescence and UVabsorption spectrometers are used as dual detectors in SEC. The earliest signal by UV-absorption varies according to the properties of the sample, while the shortest elution time (largest apparent MM) where the UV-fluorescence detector showed signal remains static at elution times corresponding to masses below 3000. 6. Shorter than expected elution times of C60 and C70 fullerenes showed the greatest departure from the PSPMMA line. Fullerenes have disproportionately large sizes for their MMs and eluted according to molecular size rather than molecular mass. A linear relationship was found between log (diameter) of a set of colloidal silica samples and their elution times. The consistency of the fullerene data with this relationship suggests that samples with well-defined 3-D conformations elute with a different elution-time relationship than the more planar heavy hydrocarbon molecules. To a first approximation, this relationship appears independent of particle density. 7. The twin-peak shapes of SEC chromatograms of fossil fuel-derived samples have been reviewed. Known 3-dimensional materials (colloidal silica and fullerenes) with diameters > ∼1 nm have appeared in the excluded zone. It seems possible that already large molecules (e.g., of an asphaltene or pitch sample of say ∼ several 1000 u) may switch from planar to other conformations and thus appear as much larger mass molecules. However, proof of such shapes in coal- or petroleumderived materials is lacking. While the broad brush of the “aggregates” label has not explained much of our experimental observations, the molecular masses, structures, and conformations of material appearing under the excluded peak in size exclusion chromatography require further careful study. Acknowledgment. The authors express their gratitude for the sustained support of this work by British Coal Utilization Research Association (BCURA) and the UK Department of Trade and Industry (DTI), under Projects No. B44 and B53. The authors also thank Nissan Chemical Industries Ltd. of Houston, Texas, for the gift of the colloidal silica samples and Shell, Amsterdam, for the gift of petroleum vacuum residues. We also lthank the Scientific and Technical Research Organization of Turkey for a fellowship to F.K. Our special thanks go to Trevor Roberts of CPI for his encouragement, many helpful discussions, and for supplying the domestic soot sample. EF030178H
