Energy & Fuels 2004, 18, 987-994
987
Analysis of Petroleum Resins Using Electrospray Ionization Tandem Mass Spectrometry D. J. Porter and P. M. Mayer* Department of Chemistry, University of Ottawa, Ottawa, Canada K1N 6N5
M. Fingas Emergencies Science and Technology Division, Environmental Protection Service, Environment Canada 3439 River Road, Ottawa, Canada K1A 0H3 Received May 21, 2003. Revised Manuscript Received March 30, 2004
Electrospray ionization mass spectrometry (ESI-MS) is becoming a common method for the analysis of petroleum-based chemicals. Here, we have demonstrated the usefulness of this method for the analysis of polar resin fractions of crude oil, fuel oil, and diesel. The mass spectra of crude oil resins can be used to make comparisons between the constituents present in these oils. Tandem mass spectra of one crude oil sample identified a major constituent to be alkylated carbazoles. The analysis of resins from a diesel sample showed the presence of quinoline compounds, along with oxygen or sulfur heterocycles and alkyl carbazoles. The effect of weathering on the composition of petroleum resins was studied by comparing the average number molecular weights generated from mass spectra.
Introduction Crude oil is a complex mixture that is known to consist of small quantities of nitrogen-, sulfur-, and oxygen-containing compounds. Many of these polar chemicals can be found in the heavier asphaltene and resin fractions of oils. Use of the popular SARA fractionation procedure allows oils to be separated into fractions that contain saturates, aromatics, resins, and asphaltenes. This has enabled individual and chemical class identifications of the many constituents present in oils worldwide. The polar resin fraction of an oil is usually a dark brown to black ductile semisolid. Its complexity and physical properties make it difficult for any type of compound class separation, let alone individual chemical identification. Despite this difficulty, resins have been studied on numerous occasions and are reported to contain many different compounds, such as sulfoxides, sulfides, sulfones, amides, thiophenes, pyridines, quinolines, and carbazoles.1,2 The methods used for the analyses of resins have varied. There are several reports on the use of structural Fourier transform infrared (FTIR) spectroscopy3-6 and one that involves X-ray absorption near-edge structure (XANES) spectroscopy.2 Other methods include nuclear magnetic resonance (NMR),6,7 mass spectrometry (MS),6,8 and high-performance liquid chromatography-flame ionization detection (HPLC-FID).9 * Author to whom correspondence should be addressed. Telephone: 613-562-5800, ext 6038. Fax: 613-562-5170. E-mail address: pmayer@ science.uottawa.ca. (1) Rudzinski, W. E.; Aminabhavi, T. M. Energy Fuels 2000, 14, 464-475. (2) Mitra-Kirtley, S.; Mullins, O. C.; Ralston, C. Y.; Sellis, D.; Pareis, C. Appl. Spectrosc. 1998, 52, 1522-1525. (3) Boukir, A.; Aries, E.; Guiliano, M.; Asia, L.; Doumenq, P.; Mille, G. Chemosphere 2001, 43, 279-286. (4) Khadim, M.; Sarbar, M. J. Pet. Sci. Eng. 1999, 23, 213-221. (5) Christy, A.; Dahl, B.; Kvalheim, M. Fuel 1989, 69, 430-435. (6) Ali, M. F.; Bukhari, A.; Hasan, M. Fuel Sci. Technol. Int. 1989, 7, 1179-1208.
Electrospray ionization mass spectrometry (ESI-MS) is becoming a common method for the analysis of petroleum-based chemicals10-14 and oilfield additives.15,16 This is due to the ease with which one can couple HPLC to MS and conduct analyses of polar, ionic, and neutral compounds dissolved in simple solvents. An excellent review of the use of mass spectrometry in petrochemical analysis has recently been published.17 In the present study, we have used ESI-MS and electrospray ionization tandem mass spectrometry (ESI-MS/MS) for the analysis of the polar resin fraction from several crude oils, fuel oils, and diesel. Experimental Section Resin Samples. Polar resin samples prepared by SARA fractionation18 were obtained from the Emergencies Science and Technology Division of Environment Canada in Ottawa. The resins were extracted from oil samples of Alberta Sweet Mixed Blend (ASMB), Mars, Alaskan Northslope, Hibernia, Maui, Arabian Heavy, and Tchat. Resin samples from the petroleum products diesel, heavy fuel oil (HFO 6303), and Fuel Oil No. 5 were also obtained, along with weathered samples of all resins, except ASMB. The origin of each resin sample can be found in Table 1. (7) Khattab, S. A.; El-Azzaby, O. H.; Abo-lemon, F. S.; Roushdy, M. I. Egypt. J. Chem. 1978, 21, 421-430. (8) Buchanan, M. Anal. Chem. 1982, 54, 571-574. (9) Pearson, C.; Gharfeh, S. Anal. Chem. 1986, 58, 307-311. (10) Zhan, D.; Fenn, J. B. Int. J. Mass Spectrom. 2000, 194, 197208. (11) Roussis, S.; Proulx, R. Anal. Chem. 2002, 74, 1408-1414. (12) Roussis, S. G.; Fedora, J. W. Rapid Commun. Mass Spectrom. 2002, 16, 1295-1303. (13) Rodgers, R.; Hendrickson, C.; Emmett, M.; Marshall, A.; Greaney, M.; Qian, K. Can. J. Chem. 2001, 79, 546-551. (14) Qian, K.; Robbins, W.; Hughey, C.; Cooper, H.; Rodgers, R.; Marshall, A. Energy Fuels 2001, 15, 1505-1511. (15) Gough, M. A.; Langley, G. J. Rapid Commun. Mass Spectrom. 1999, 13, 227-236. (16) McCormack, P.; Jones, P.; Rowland, S. J. Rapid Commun. Mass Spectrom. 2002, 16, 705-712. (17) Marshall, A. G.; Rogers, R. P. Acc. Chem. Res. 2004, 37, 5359.
10.1021/ef0340099 CCC: $27.50 © 2004 American Chemical Society Published on Web 05/11/2004
988
Energy & Fuels, Vol. 18, No. 4, 2004
Porter et al.
Figure 1. Mass spectra of ASMB resin obtained using (a) a 50/50 acetonitrile/benzene mobile phase and (b) an acetone mobile phase. Table 1. List of Resins and Their Origins resin
origin
notes
Maui Tchatamba (Tchat) Alaskan Northslope (ANS) Arabian Heavy Hibernia Mars Alberta Sweet Mix Blend (ASMB) Fuel Oil No. 5
crude oil crude oil crude oil blend crude oil blend crude oil crude oil crude oil blend refined oil product
HFO 6303
refined oil product
diesel
refined oil product
from Shell Todd Oil Services Ltd., New Zealand from Marathon Oil, Gabon (Africa) drawn off the TAPS pipeline in Valdez, Alaska from Saudi Arabia; “Heavy” indicates a commercial (density) grade oil from Hibernia, offshore Newfoundland, Canada from Shell, offshore Gulf of Mexico, USA from ESSO, Alberta, Canada from the U.S. Department of the Interior, M. M. S., OHMSETT, New Jersey, USA; for ship fuel and power generation from Imperial Oil Ltd, Nova Scotia, Canada; Heavy Fuel Oil (HFO) for ship fuel and power generation; also known as Bunker C, Marine Boiler Fuel, and Land Bunker from a local retailer (Stinson’s Gas, Ontario, Canada); also known as “Summer Diesel”
Electrospray Ionization (ESI) Sample Preparation. Stock solutions of the resin in the range of 4-7 g resin/L were prepared using various solvents (discussed below). After the addition of solvent, the samples were shaken vigorously for 2 min and allowed to stand overnight. All were orange-brown in color, and some had small amounts of undissolved resin remaining on the sides of the vial. Subsamples with concentrations of ∼1 g resin/L were later prepared and were spiked with a conductivity additive as needed (0.4% acetic acid v/v). All subsamples formed clear yellow-orange solutions and seemed to have no undissolved matter. Experiments were also conducted with resins prepared in perdeuterated acetone obtained from C/D/N Isotopes, Inc. (Point-Claire, Quebec, Canada; 99% pure, 99.5% isotopic enrichment). For these experiments, the mobile phase was matched to that of the sample solvent. Electrospray Ionization Mass Spectrometry (ESIMS). All mass spectra were obtained using a Micromass Quattro-LC triple-quadrupole mass spectrometer that was fitted with a Z-Spray ion source and was running the MassLynx data acquisition and processing software. Typically, 5 µL of a resin sample was injected into a mobile phase of the same solvent composition at a flow rate of 40-50 µL/min. The inlet capillary voltage was normally held between +4.0 eV and +4.5 eV (positive-ion mode), and cone and extractor voltages were adjusted as needed to obtain a stable signal. The nitrogen desolvation gas flow rate was 100 L/h. Electrospray Ionization Tandem Mass Spectrometry (ESI-MS/MS). Collision-induced dissociation (CID) mass spectra19 were recorded using the triple-quadrupole operating in MS/MS mode. The first quadrupole selects the precursor (18) Draft Method for the Determination of the Hydrocarbon Group Constituents of Oil and Petroleum Products, Emergencies Science and Technology Division, Environment Canada, Ottawa, Canada, January 2002.
ion m/z of interest and transmits it to the rf-only hexapole. CID mass spectra were obtained using argon collision gas in the rf-only hexapole. The CID mass spectrum of a compound will be dependent on the experimental conditions used. For these experiments, cone, entrance, and exit voltages were optimized to produce the best spectra. A constant collision gas pressure of 2.5 (( 0.5) × 10-4 mbar was used for all experiments. The collision energy (laboratory frame) has a dramatic impact on the CID mass spectra. For the initial CID experiments, the collision energy was varied to determine the best setting for fragmentation. The collision energy dependence was studied of the CID mass spectra of ions obtained from the ASMB resin. The significance of the results will be discussed in more detail in the Results and Discussion section. The collision energies used in individual CID experiments are indicated in the text.
Results and Discussion Crude Oil Resins. The first task involved finding a solvent system that could adequately dissolve the polar resin compounds and could be used with an electrospray mass spectrometer. A solvent with a 50/50 composition of acetonitrile/benzene coupled with the same mobile phase worked well. Acetic acid (0.4% v/v) was added to the samples to increase the conductivity. Using this mobile phase, the mass spectrum from an ASMB sample was observed to contain numerous compounds in the range of m/z 20-600 with the highest peak at m/z 312 (Figure 1a). A distinct “hump” is noticeable in the range of m/z 200-600 with incremental peaks differing by two (19) Busch, K. L.; Glish, G. L.; McLuckey, S. A. Mass Spectrometry/ Mass Spectrometry; VCH Publishers: New York, 1988.
Analysis of Petroleum Resins Using ESI-MS/MS
Energy & Fuels, Vol. 18, No. 4, 2004 989
Figure 2. Mass spectra of crude oil resins: (a) Hibernia, (b) Mars, (c) Tchat, (d) Alaskan Northslope, (e) Arabian Heavy, (f) Maui, and (g) ASMB.
Figure 3. CID mass spectra of ASMB compound m/z 298 obtained with (a) acetone as the mobile phase and (b) acetone-d6 as the mobile phase. Spectra were obtained using a collision energy of 80 eV; the ordinate scale has been expanded to enhance fragment ion peaks.
mass units. This is consistent with a difference in saturation between compounds of neighboring m/z values. Also observed were periodic peaks that differed by 14 mass units; these peaks were attributed to the presence of an additional CH2 group on an alkyl chain. Zhan and Fenn10 encountered the same phenomena in their ESI spectra of crude oil. With the 2 mass unit difference, ∼150 compounds are present in the observed resin mass spectrum. Because all observed peaks are
of even mass, the unprotonated compounds are of odd mass and may contain an alkyl or heterocyclic N atom. Acetone was also used as both the sample solvent and the mobile phase, and acetic acid (0.4% v/v) was added to increase conductivity. The mass spectrum for an ASMB sample with this system was similar to that obtained in the acetonitrile/benzene mobile phase (see Figure 1b). Acetone proved to be efficient in extracting a similar amount of polar material from the resins and
990
Energy & Fuels, Vol. 18, No. 4, 2004
Porter et al.
Figure 4. CID mass spectra of ASMB sample ions with (a) m/z 268, (b) m/z 254, and (c) m/z 240 at a collision energy of 50 eV.
Figure 5. CID mass spectra at 25 eV (unless indicated otherwise) of protonated (a) 2,6-dimethylpyridine (m/z 108), (b) 2,4dimethylquinoline (m/z 158), (c) 2,3-dimethylindole (m/z 146), (d) 1,4,5,8,9-pentamethyl carbazole (m/z 238) at 50 eV lab frame collision energy, and (e) ASMB m/z 238 at 50 eV; the ordinate scales of panels a, b, and c have been expanded.
produced more-consistent spectra. On the basis of the results of these studies, it was decided that acetone would be used as both the sample solvent and the mobile phase for the remaining resins. The mass spectra obtained for the other six oil resins were similar to that for ASMB, with a main distribution
of peaks in the range of m/z 200-600 (Figure 2a-g). All spectra displayed the same pattern of incremental peaks, differing by 2 mass units, and distinct tall peaks 14 mass units apart. These spectra differed slightly by the presence of low-molecular-weight compounds in the m/z 75-200 range.
Analysis of Petroleum Resins Using ESI-MS/MS
Energy & Fuels, Vol. 18, No. 4, 2004 991
Figure 6. CID mass spectra of protonated carbazoles: (a) 1,4,5,8,9-pentamethyl- (m/z ) 238), (b) 9-ethyl- (m/z ) 196), and (c) 9-methyl- (m/z ) 182) at a collision energy of 50 eV.
Figure 7. Mass spectra of fuel oils: (a) HFO 6303 and (b) Fuel Oil No. 5.
The fragmentation of m/z 298 from the ASMB resin at 80 eV collision energy is presented in Figure 3a and for the same ion in acetone-d6 solvent in Figure 3b. A series of ions separated by 14 mass units is observed. This is similar to a hydrocarbon cracking pattern in the mass spectrometry of alkanes. The fragment peaks in the spectrum end at m/z 142, giving a total loss of 156 mass units from the main compound. The m/z 142 peak is probably the main ring structure of the compound and is very stable, because it is not susceptible to any further dissociation. It is presumed to contain a heteroatom, such as nitrogen, which enables it to be protonated and detected using ESI-MS. An investigation of the 2-mass-unit increment observed in all resin mass spectra was conducted by performing collision experiments on the m/z trio 294, 296, and 298. The fragmentation patterns should indicate whether the peaks are the result of increasing saturation in a ring or alkyl chain. If a double bond is present in an alkyl chain, it would be expected to exhibit a unique fragmentation pattern that differs from that of the next higher or lower m/z compound. However, if present in a ring system, the fragmentation pattern will be similar to that for the next higher or lower m/z
compound, with the fragment ion peaks being 2 mass units higher or lower. Indeed, the mass spectrum of m/z 296 was similar to that of m/z 294, except that the fragment ion peaks were 2 mass units higher. The same was observed for m/z 298 when compared to m/z 296. It can be concluded from this set of spectra that the 2-unit incremental difference is indeed the result of increasing saturation in a ring system and not in an alkyl chain. The fragmentation of the three ions with m/z 240, 254, and 268 from the ASMB resin was studied at a collision energy of 50 eV to determine if these compounds were related. The spectra of this grouping revealed a similar pattern (see Figure 4), with the fragment ion peaks from m/z 240 being present in the spectra for m/z 254, and the same relationship being observed between m/z 254 and m/z 268. Therefore, it is reasonable to conclude that this trio of ions is comprised of compounds from the same chemical class. After obtaining CID mass spectra for several resin compounds, a comparison was made with the spectra for some pyridines, indoles, benzoquinolines, quinolines, and carbazoles (Figure 5). The only class of compounds that yielded similar results were the carbazoles. Indeed, previous researchers have identified carbazoles in resins
992
Energy & Fuels, Vol. 18, No. 4, 2004
Porter et al.
Figure 8. Mass spectrum of a sample of “Summer Diesel”. Table 2. CID Mass Spectra (50 eV) for Substituted Carbazoles fragment (m/z)
neutral loss
relative intensity (%)
167 153 152 151 140 138 128 127 114
9-methyl-; Parent m/z 182 15 29 30 31 42 44 54 55 68
100 41 22 16 10 6 8 8 5
180 167 154 153 140 128 126 114 113
9-ethyl-; Parent m/z 196 16 30 42 43 56 68 70 82 83
80 100 16 18 5 7 4 8 6
166
9-phenyl-; Parent m/z 244 78
100
223 207 206 192 180 165
1,4,5,8,9-pentamethyl-; Parent m/z 238 15 100 31 28 32 16 46 4 58 2 73 1
131 130 129 77
1,2,3,4-tetrahydro-; Parent m/z 172 41 100 42 76 43 54 95 52
and other fossil fuels.6,20,21 Since the carbazoles were obtained predissolved in benzene, samples were prepared in a 50/50 acetonitrile/benzene mixture and run with the same mobile phase into the electrospray. The results are summarized in Table 2. The CID mass spectrum for protonated 1,4,5,8,9-pentamethyl carbazole (m/z 238) is very similar to that obtained for the ASMB resin peak at m/z 238 (see Figure 5) at various collision energies. Second, Figure 6 shows that the fragment ion peaks in the CID mass spectrum of 9-methyl carbazole are all present in the CID mass spectrum of 9-ethyl (20) Willsch, H.; Clegg, H.; Horsfield, B.; Radke, M.; Wilkes, H. Anal. Chem. 1997, 69, 4203-4209. (21) Ignatiadis, I.; Schmitter, J.; Arpino, P. J. Chromatogr. 1985, 324, 87-111.
Table 3. CID Fragment Peaks and Neutral Losses in Mass Spectra of Diesel Compounds at 30 eV fragment (m/z) 80 79 77 67 65 55 53 51 43 41 39 29 27 120 107 106 93 91 80 79 77 65 59 55 41 143 142 130 116 115 91 157 142 129 116 114 91 171 157 156 143 130 115 105 91 185 171 157 142
relative intensity (%) Parent m/z 95 2 5 4 9 2 35 6 2 3 14 5 11 3 Parent m/z 135 16 11 7 5 11 3 8 25 3 6 1 4 Parent m/z 158 32 12 4 8 26 19 Parent m/z 172 41 3 5 11 21 19 Parent m/z 186 67 10 9 3 5 1 3 1 Parent m/z 200 39 46 9 6
neutral loss 15 16 18 28 30 40 42 44 52 54 56 66 68 15 28 29 42 44 55 56 58 70 76 80 94 15 16 28 42 43 67 15 30 43 56 58 81 15 29 30 43 56 71 81 95 15 29 43 58
carbazole, which is 14 mass units larger. This same trend was observed in the fragmentation patterns of
Analysis of Petroleum Resins Using ESI-MS/MS
Energy & Fuels, Vol. 18, No. 4, 2004 993
Figure 9. CID mass spectra of protonated (a) 2,4-dimethylquinoline (m/z ) 158), (b) 2,7-dimethylquinoline (m/z ) 158), and (c) diesel (m/z ) 158). A collision energy of 25 eV was used for the quinolines, and a collision energy of 30 eV was used for the diesel compound.
resin compounds m/z 240, 254, and 268 (discussed previously). From these results, it is reasonable to conclude that the protonated resin constituents studied here are alkylated carbazoles. Fuel Oils. Two fuel oils were also studied. The mass spectra for these oils are presented in Figure 7. Immediately, one can see that the mass spectrum for Heavy Fuel Oil 6303 (HFO) is quite unique, with a narrow distribution of low-molecular-weight compounds encompassing the range m/z 100-400. Unlike the crude oil mass spectra, a low-intensity band of peaks continues from m/z 400 to m/z 1000, forming a slight hump between m/z 600 and m/z 900. The mass spectrum of Fuel Oil No. 5 could easily be mistaken for one of the crude oils, because it contains the same “hump” of peaks in the m/z 200-500 region, which slowly tapers off toward m/z 900. Both fuel oils contain the same ions in the m/z 100-400 region but show differences in the abundances of these ions. Although HFO seems to contain more higher-molecular-weight material, a determination of the average number molecular weight (Mn) for the m/z 220-750 regions of these oils revealed similar values: Mn ) 405 Da for HFO and Mn ) 404 Da for Fuel Oil No. 5. Diesel. The mass spectrum for a sample of “Summer Diesel” was observed to contain significantly fewer peaks than any of the other resins. All of these peaks occupied the m/z 60-340 range (Figure 8), with m/z 172 being the most abundant. The mass spectrum is less congested than those of the oil resins but has the same pattern of periodic peaks 14 mass units apart. CID experiments on several ions in the diesel sample are summarized in Table 3. CID experiments conducted on two odd m/z ionssm/z 95 and m/z 135, which are likely oxygen or sulfur compoundssshowed similar fragmentation patterns, indicating that they are structurally related. The other four substances are assumed to be nitrogencontaining aromatics, based on the even m/z for the protonated parent molecules. The CID mass spectrum of m/z 158 compared very well with the mass spectra for 2,7- and 2,4-dimethylquinoline (Figure 9). Therefore, this peak is likely to be a dimethylquinoline isomer or a mixture of isomers. The m/z 172 and m/z 186 ions
Table 4. Molecular Weight Statistics for Weathered and Unweathered Resins amount weathered (%)
average S2 29 718 ( 2346
0 14.12 29.76 43.45
Maui Resin 367 ( 2 370 ( 3 363 ( 7 361 ( 14
34 087 ( 590 35 768 ( 2343 33 522 ( 235 35 659 ( 2395
0 19.4 38.4
Tchat Resin 382 ( 5 382 ( 3 376 ( 8
33 527 ( 1071 33 537 ( 551 32 119 ( 1617
0
a
average Mna ASMB Resin 395 ( 21
0 10 22.5 30.5
Alaskan Northslope Resin 393 ( 12 393 ( 1 394 ( 6 393 ( 1
36 116 ( 4250 35 482 ( 1577 34 127 ( 800 33 784 ( 1089
0 8.59 16.07 23.5
Arabian Heavy Resin 408 ( 16 411 ( 4 416 ( 11 417 ( 6
29 948 ( 929 30 182 ( 383 29 963 ( 509 29 382 ( 762
0 10.13 20.8 35.56
Hibernia Resin 423 ( 15 416 ( 1 419 ( 3 410 ( 6
30 041 ( 2982 31 708 ( 818 31 116 ( 755 32 418 ( 270
0 8.42 17.21 26.15
Mars Resin 414 ( 1 411 ( 6 413 ( 4 415 ( 7
31 307 ( 965 30 757 ( 938 30 486 ( 1374 31 170 ( 1178
0 7.25
Fuel Oil No. 5 Resin 404 ( 4 409 ( 1
32 003 ( 1295 30 206 ( 2102
0 2.5
HFO 6303 Resin 405 ( 9 378 ( 30
29 928 ( 1826 32 665 ( 3576
0 7.18 14.20 21.95
Diesel Resin 209 ( 4 207 ( 2 211 ( 7 212 ( 4
6382 ( 480 6441 ( 544 6564 ( 120 6465 ( 122
Average number molecular weight.
display common fragmentation patterns that are not consistent with those of the carbazoles, pyridines, quinolines, or indoles. The m/z 200 ion has the same
994
Energy & Fuels, Vol. 18, No. 4, 2004
fragmentation pattern as some of the ASMB compounds presented previously and is likely a carbazole. Weathered Resins. The Emergencies Science and Technology Division of Environment Canada conducts weathering experiments to simulate the physical and chemical changes that occur when oil and its products are exposed to sunlight, wind, and water.22 The “percent weathered” value refers to the percentage evaporative mass loss over a 48 h period. To determine if weathered resin samples were significantly different from unweathered samples, the number average molecular weights (Mn values) were determined for the m/z 220-750 regions of the crude and fuel oils and the m/z 75-350 region for diesel. The selection of the m/z range was arbitrary but chosen to include all significant peaks in the mass spectra. This method was chosen because no significant differences were apparent through visual examination of the mass spectra for most resin samples. For each resin, the mass spectrum of three different samples was obtained and the results averaged (Table 4). The Mn values for all the unweathered crude oil resins were in the range of 367-423. For the two fuel oils, the unweathered Mn values were similar (404 and 405), as mentioned previously, whereas for diesel, a value of Mn ) 209 was obtained. With respect to weathering, we were unable to conclude whether this process resulted in any change in the constituents of the resins when the standard error for each was taken into account. There were two possible explanations for this observation: either (i) all the components of each resin degraded equally (resulting in no observable increase or decrease in Mn) or (ii) the components that were lost were not part of the polar constituents that were detectable by ESI-MS. For the fuel oils, an increase in Mn was only observed for the weathered Fuel Oil No. 5, and it is marginal. The breath of (22) Draft Method for the Simulated Weathering of Oil and Petroleum Products, Emergencies Science and Technology Division, Environment Canada, Ottawa, Canada, January 2002.
Porter et al.
distribution of each resin can be observed in the variance values in Table 4. As seen in the mass spectra, the crude and fuel oil Mn values have very large distributions, whereas a much smaller spread is observed for the diesel resins. Conclusions We have demonstrated the usefulness of electrospray ionization mass spectrometry (ESI-MS) for the analysis of resin fractions of crude oil, fuel oil, and diesel. It was possible to obtain mass spectra of the polar constituents of these resins and make comparisons between them. Tandem mass spectra of the individual compounds present in Alberta Sweet Mixed Blend (ASMB) crude oil indicated that a major component was alkylsubstituted carbazoles. The resins of commercial-grade fuel oils can be significantly different in their composition. The analysis of resins from a “Summer Diesel” showed the presence of quinoline compounds, along with oxygen- or sulfur-based chemicals, in addition to alkyl carbazoles. Finally, the effects of weathering on the composition of petroleum resins can be studied by comparing the average number molecular weight (Mn) values generated from the mass spectra. In the present study, increases and decreases in the MN values for several resins were observed; however, they were not significant when standard errors were taken into account. Acknowledgment. P.M.M. thanks the Natural Sciences and Engineering Research Council of Canada for its continued financial support. The authors thank the Canadian Foundation for Innovation (New Opportunities Fund), the Ontario Research and Development Challenge Fund (ORDCF-II), VWR-Canlab Canada, the Eastern Cereal and Oilseed Research Centre of Agriculture and Agri-Foods Canada, the Emergencies Science Division of Environment Canada, and the University of Ottawa for the funds to purchase the Quattro-LC. EF0340099