Anal. Chem. 2010, 82, 413–419
Enrichment, Resolution, and Identification of Nickel Porphyrins in Petroleum Asphaltene by Cyclograph Separation and Atmospheric Pressure Photoionization Fourier Transform Ion Cyclotron Resonance Mass Spectrometry Anal. Chem. 2010.82:413-419. Downloaded from pubs.acs.org by UNIV OF LOUISIANA AT LAFAYETTE on 01/23/19. For personal use only.
Kuangnan Qian,* Kathleen E. Edwards, Anthony S. Mennito, Clifford C. Walters, and J. Douglas Kushnerick ExxonMobil Research and Engineering Company, Annandale, New Jersey 08801 We report here the first high resolution mass spectrometric evidence of nickel porphyrins in petroleum. A petroleum asphaltene sample is fractionated by a silica-gel cyclograph. Nickel content is enriched by ∼3 fold in one of the cyclograph fractions. The fraction is subsequently analyzed by atmospheric pressure photoionization (APPI) Fourier transform ion cyclotron resonance mass spectrometry (FTICR MS) with an average mass resolving power of over 500 K (M/∆Mfwhm). Similar to vanadyl porphyrins, monocylcoalkano-type (presumed to be deocophylerythro-etioporphyrin DPEP) Ni porphyrins are found to be the most abundant family followed by etio, bicycloalkano-type, and rhodo-monocylcoalkanotype Ni porphyrins. A Z number ranging from -28 to -44 and a carbon number ranging from 26 to 41 were observed. A significant amount of nickel and vanadyl geoporphyrins are in more condensed tetrapyrrolic cores than just chlorophyll-derived DPEP- and etioporphyrins. Ni has a higher etio/DPEP ratio and rhodoetio/rhodo-DPEP ratio than does VO. V and Ni are two major metal elements in crude oil that are bound in the form of metalloporphyrins.1,2 These compounds are derived from chlorophyll, bacteriochlorophylls, hemes, and other tetrapyrrolic biochemicals that convert in the geosphere to deoxophylloerythroetioporphyrins (DPEP), etioporphyrins (etio), tetrahydrobenzoporphyrin, benzo- (rhodo-) etio, benzo-DPEP porphrins, and a host of other tetrapyrrolic structures.3,4 Petroporphyrins are commonly characterized by solvent extraction and/ or gas or liquid chromatographic separation followed by UV/vis * To whom correspondence should be addressed. E-mail: kuangnan.qian@ exxonmobil.com. (1) Filby, R. H.; Van Berkel, G. J. In Metal Complexes in Fossil Fuels, ACS Symposium Series 344; Filby, R. H., Branthaver, J. F., Eds.; American Chemical Society: Washington, DC, 1987; pp 2-39. (2) Baker, E. W.; Louda, J. W. In Biological Markers in the Sedimentary Record; John, R. B., Ed.; Elsevier: Amsterdam, 1986; pp 125-225. (3) Quirke, J. M. E. In Metal Complexes in Fossil Fuels, ACS Symposium Series 344; Filby, R. H., Branthaver, J. F., Eds.; American Chemical Society: Washington, DC, 1987; pp 308-331. (4) Callot, H. J.; Ocampo, R. In The Porphyrin Handbook; Kadish, K. M., Smith, K. M., Guilard, R., Eds.; Academic Press: New York, 2000; Vol. 1, pp 349398. 10.1021/ac902367n 2010 American Chemical Society Published on Web 12/08/2009
absorption spectroscopy and low-resolution mass spectrometry.3,5-8 The petroporphrins may be analyzed in their demetellated6,9,10 or metalated11-15 forms. We recently reported the direct observations of vanadyl (VO) porphyrins and sulfur-containing VO porphyrins in a petroleum asphaltene by atmospheric pressure photoionization (APPI) Fourier transform ion cyclotron resonance mass spectrometry (FTICR MS).16 Both elemental formula and isotope ratios are determined, matching theoretical predictions. VO porphyrins with a Z number ranging from -28 to 54 and carbon number ranging from C26 to C65 were reported. For the first time, sulfur-containing VO porphyrins were observed using the ultrahigh resolution capability. The observation of VO porphyrins were also reported by McKenna et al.17 using a similar technique. Curiously, both groups were unable to detect the presence of nickel porphyrins. In a subsequent examination of asphaltenes from a suite of oils containing a range of Ni and V content, we found that the VO porphyrins were easily detected by APPI-FTICR MS while no nickel porphyrins could be found.18 Nickel porphyrins are documented in sedimentary rock extracts using methods involving HPLC separation/purification, UV, NMR, resonance Raman spectroscopy, and low resolution MS (5) Eckardt, C. B.; Carter, J. F.; Maxwell, J. R. Energy Fuels 1990, 4, 741–747. (6) Johnson, J. V.; Britton, E. D.; Yost, R. A.; Quirke, J. M. E.; Cuesta, L. L. Anal. Chem. 1986, 58, 1325–1329. (7) Sundararaman, P. Anal. Chem. 1985, 57, 2204–2206. (8) Verne-Mismer, J.; Ocampo, R.; Bauder, C.; Callot, H. J.; Albrecht, P. Energy Fuels 1990, 4, 639–643. (9) Xu, H.; Que, G.; Yu, D.; Lu, J. R. Energy Fuels 2005, 19, 517–524. (10) Blum, W.; Eglinton, G. J. High Resolut. Chromatogr. 1989, 12, 621–623. (11) Ali, M. F.; Perzanowski, H.; Bukhari, A.; Al-Haji, A. A. Energy Fuels 1993, 7, 179–184. (12) Huseby, B.; Barth, T.; Ocampo, R. Org. Geochem. 1996, 25, 273–294. (13) Ocampo, R.; Bauder, C.; Callot, H. J.; Albrecht, P. Geochim. Cosmochim. Acta 1992, 56, 745–761. (14) Sundararaman, P.; Vestal, C. Org. Geochem. 1993, 20, 1099–1104. (15) Van Berkel, G. J.; Glish, G. L.; McLuckey, S. A.; Tuinman, A. A. Energy Fuels 1990, 4, 720–729. (16) Qian, K.; Mennito, A. S.; Edwards, K. E. F.; Dave, T. Rapid Commun. Mass Spectrom. 2008, 22, 2153–2160. (17) McKenna, A. M.; Purcell, J. M.; Rodgers, R. P.; Marshall, A. G. Energy Fuels 2009, 23, 2122–2128. (18) Qian, K.; Kelemen, S. R.; Sansone, M.; Walters, C. C. 24th International Meeting on Organic Geochemistry, Bremen, Germany, September 6-11,2009.
Analytical Chemistry, Vol. 82, No. 1, January 1, 2010
413
techniques.12,13,19-22 Ni porphyrins also were observed in a fractionated gilsonite bitumen extract using electrospray ionizationlow resolution mass spectroscopy.23 However, we have found no published mass spectrometric data of Ni porphyrins with proper isotopic ratios from produced petroleum. In many studies where Ni porphyrin distributions are reported for crude oil, their presence were inferred from the analysis of demetalated petroleum fractions9,10,24 Nevertheless, x-ray absorption fine structure (XAFS) spectroscopy has shown that nearly all of the nickel in petroleum exists associated with a tetrapyrrole environment25 and the nondetection by APPI-FTICR MS is unexpected. Nickel is less abundant than vanadium in most crude oils. However, assuming that the ratio of nickel to vanadyl porphyrins is directly proportional to the Ni/V of the total metal content, the more predominant nickel porphyrins should have been easily detected. Indeed, APPI-FTICR MS proved capable of detecting a host of extended aklylated and aromatized vanadyl porphyrins present in trace amounts that have not been seen in previous methods.16 The absence of nickel porphyrin spectra suggests that these compounds may have low ionization efficiency or may be masked by the presence of other compounds. An alternative hypothesis is that nickel porphyrins in crude oils are in a complex form that cannot be vaporized/ionized by APPI. Here, we report the successful detection of nickel porphyrins in a fractionated crude oil by APPI-FTICR MS. Our studies reveal the underlying causes of their nondetection in earlier analyses and highlight the need for sample enrichment even for ultrahigh resolution mass spectrometry. EXPERIMENTAL SECTION Sample Preparation. An onshore Californian oil was selected for study as it has a relatively high total nickel (71 ppm) and nitrogen content (7760 ppm). Total vanadium is 88 ppm, and sulfur is 1.49 wt %. A nickel-enriched sample was prepared. Volatile hydrocarbons were removed by distillation, and the vacuum distillation bottom was recovered (>1000 °F, 538 °C). Asphaltenes then were precipitated by dissolving 20 g of the vacuum resid into 200 mL of n-heptane. The solution was equilibrated overnight at ambient temperature. The insoluble material (n-C7 asphaltene) was then filtered through a medium porosity Buchner funnel and dried in a vacuum oven at 100 °C for several hours. The asphaltene accounts for ∼20 wt % of the sample. The elemental composition of the asphaltene is given in Table 1. Direct analysis of the sample by APPI-FTICR MS suggested the possible presence of nickel porphyrins. However, the signal-to-noise is low, and we could not make a firm conclusion. The asphaltene fraction was further fractionated using an Analtech Cyclograph I TLC System (Cole Parmer). One gram of (19) Junium, C. K.; Mawson, D. H.; Arthur, M. A.; Freeman, K. H.; Keely, B. J. Org. Geochem. 2008, 39, 1081–1087. (20) Kashiyama, Y.; Kitazato, H.; Ohkouchi, N. J. Chromatogr., A 2007, 1138, 73–83. (21) Boggess, J. M.; Czernuszewicz, R. S.; Lash, T. D. Org. Geochem. 2002, 33, 1111–1126. (22) Magi, E.; Ianni, C.; Rivaro, P.; Frache, R. J. Chromatogr., A 2001, 905, 141–149. (23) Van Berkel, G. J.; McLuckey, S. A.; Glish, G. L. Anal. Chem. 2002, 63, 1098–1109. (24) Xu, H.; Yu, D.; Que, G. Fuel 2005, 84, 647–652. (25) Miller, J. T.; Fisher, R. B.; van der Eerden, A. M. J.; Koningsberger, D. C. Energy Fuels 1999, 13, 719–727.
414
Analytical Chemistry, Vol. 82, No. 1, January 1, 2010
Table 1. Elemental Composition of Asphaltene S (wt %) N (wt %) Ni (ppm) V (ppm) C (wt %) H (wt %) H/C atomic
2.65 2.52 497 734 83.59 8.53 1.22
asphaltene was dissolved in 10 mL toluene and loaded onto a fresh, prescraped 8 mm silica rotor. The rotor speed was 1000 rpm, and the solvent flow rate was 8 mL/min. The first dark band was eluted with toluene until the eluent became clear. A second dark band was eluted first with a 20:80 methanol/toluene mixture and then with methanol until the eluent became clear. The final fraction was eluted with methanol spiked with 2% acetic acid. All samples were blown down under nitrogen and weighed. Metals content of each fraction was measured by ICPMS analysis. The first fraction was found to be enriched in nickel and was further analyzed by APPI-FTICR MS. The solvent sequence and fraction quantities are summarized in Table 2. APPI-FTICR MS Analysis. All experiments were conducted on a 12 T Bruker Apex Qe FTICR MS. About 10 mg of the nickelenriched asphaltene fraction was dissolved in 10 mL of toluene to form a ∼1000 ppm solution. The solution was introduced into the APPI source using a Cole-Parmer syringe pump with a 250 µL syringe. The flow rate was controlled at 120 µL/hour. The Syagen APPI source composed of a heated capillary needle and krypton UV lamp with ionization energy of 10.6 eV. Nitrogen was used for both nebulizing gas and drying gas. Nebulizing gas flow rate was set between 1 and 3 L/min while the drying gas flow rate was set between 2 and 7 L/min. The flow rates were adjusted to maximize the APPI FTICR MS signals. The nebulizing gas temperature was set at 450 °C to maximize the vaporization of asphaltene molecules. Toluene was used as both solvent and chemical ionization agent. Toluene was ionized by direct photoionization while asphaltene molecules (including metalloporphyrins) were ionized by charge transfer with toluene ions. The process generates primarily molecular ions. We did not observe any thermal chemistry in APPI possibly due to short residence time of the sample ions. In FTICR MS, the excited cyclotron motion of the ions is detected on receiver plates as a time domain signal that contains all the cyclotron frequencies that have been excited. Fourier transformation of the time domain signal results in the frequency domain signal that can be converted into a mass spectrum. In practical operations, the achievable mass resolution of FTICR MS is determined by a number of experimental parameters including mass range; data set size, which defines number of data points collected for the whole time-domain data; ICR cell pressure; and other acquisition parameters. In this work, the mass range was set at m/z 300 to 3000. The data set size is set to 4 Megawords. Ion accumulation time is 2 s. Data sets (1000) were coadded to generate the final spectrum. Bruker Data Analysis (DA) software is used to find the mass peak list with signal-to-noise ratio (S/N) greater than 6. The mass peak list is further analyzed for identification of asphaltene and Ni and VO porphyrin molecules. Mass Calibration. External mass calibration was performed using a blend of eight in-house synthesized aromatic compounds
Table 2. Cyclograph Separation of Asphaltene
Figure 1. APPI-FTICR MS responses of octaethyl vanadyl porphyrin and octaethyl nickel porphyrins across a wide concentration range (100 to 1150 ppm). Overall, nickel porphyrin is about 3 times less sensitive than that of vanadyl porphyrins.
covering a mass range from ∼350 to 1800 Da. In general, 2 ppm mass accuracy can be achieved with external calibration. Bruker DA molecular formula tool assisted in identifying a major homologous series. Internal calibration was then performed using the identified homologous series. On average, ∼0.2 ppm mass accuracy can be achieved with internal mass calibration. RESULTS AND DISCUSSION APPI-FTICR MS Sensitivity. One possible cause of the inability to detect Ni porphyrins in unfractionated asphaltenes is that APPI is less efficient in ionizing Ni porphryins versus VO pophyrins. Two model compounds, octaethyl-VO porphyrin (OEVOP) and octaethyl-Ni porphyrin (OENiP) were analyzed at concentrations ranging from 100 to 1200 ppm. The molecular ion signal of OEVOP and OENiP plotted against the concentration yield linear response curves (Figure 1). The quantification limit is around 100 ppm which translates into ∼10 ppm in metal content. The experiments show that OEVOP is about three times more sensitive than OENiP in APPI. The lower sensitivity of the Ni porphyrin is partially caused by the fact that Ni has two abundant isotopes, 58Ni and 60Ni. The latter has a relative abundance about 35% of 58Ni. The lower APPI ionization efficiency of the nickel porphyrins and their inherent lower abundance compared to the vanadyl porphyrins provide a simple explanation was to why the former compounds may not be detected by APPI-FTICR MS in nonenriched asphaltene samples. In low maturity and severely biodegraded crude oils, the total vanadium content may range from 400 to >1000 ppm, while the total nickel content rarely exceeds
Figure 2. Most probable interferences in high resolution identification of nickel porphyrins. (1) A protonation product of the NS2 molecule; (2) a parent ion of the O3S3 molecule, and (3) a 2 13C isotope of the sulfur molecule.
100 ppm. Even in isolated asphaltenes the Ni concentrations are typically less than 500 ppm. Considering that the Ni is complexed into a large number of tetrapyrrolic compounds, it is likely that APPI-FTICR MS sensitivity is not sufficient to detect Ni porphyrins directly. It is also known that FTICR MS has relatively limited dynamic range (typically less than 1 million) and weak signals can be buried easily in the noise. Mass Interferences. Another issue related to Ni porphyrin identification is mass interference. Vanadium has very unique mass defect, and there are almost no major petroleum components that overlap with VO porphyrins (assuming N, S, and O < 5 and V and Ni < 2, ∆M< 0.5 mDa). Within the same constraints, there are severe mass interferences for Ni porphyrins. Figure 2 shows three potential overlaps. The first one arises from a protonated molecule ion containing NS2, molecules that are relatively abundant in high-sulfur petroleum samples. It has been reported under certain conditions that petroleum molecules (especially nitrogen containing molecules) can be protonated in APPI.26 Mass resolving power (RP) required for differentiating the two species is over 1.2 million. The second interference is from O3S3 species. The mass difference is even smaller. RP of 1.9 million is required to resolve nickel porphyrin from this molecule. The closest interference is from a 13C2 isotope of S molecule, requiring 3.2 million RP to differentiate the two (26) Purcell, J. M.; Rodgers, R. P.; Hendrickson, C. L.; Marshall, A. G. J. Am. Soc. Mass Spectrom. 2007, 18, 1265–1273.
Analytical Chemistry, Vol. 82, No. 1, January 1, 2010
415
Table 3. ICPMS Analysis of Ni and V of Cyclograph Fractions
asphaltene feed fraction 1 fraction 2 fraction 3 sum % recovered
wt (g)
ppm Ni
ppm V
µg Ni
µg V
V/Ni
1.000 0.123 0.662 0.271 1.056 105.6
497 1357 442 152
734 206 823 432
497 167 293 41 501 101
734 25 545 117 687 94
1.48 0.15 1.86 2.83
Figure 3. Broadband mass spectra of cyclograph fraction 1. (a) Full mass range spectrum (350 to 1150 Da); (b) a detailed view of m/z 518 and C3158Ni DPEP; (c) a detailed view of m/z 520 and C3158Ni Etio; and (d) a detailed view of m/z 574 and C3558Ni DPEP.
species. Current high field FTICR cannot resolve these species in the broadband mode. Enrichment of Nickel. Clearly, our best chance to detect Ni porphyrins directly by APPI-FTICR MS is to select a low-sulfur, high-nickel crude. A vacuum residua asphaltene precipitated from a low maturity oil from onshore California was selected for study. (See Experimental Section and Table 1.) The nickel and vanadium content of the precipitated asphaltenes are 497 and 874 ppm, respectively. APPI-FTICR MS analysis suggested clearly the presence of VO porphyrins, but the masses corresponding to the nickel porphyrin homologues are weak and ambiguous. Consequently, a quick chromatographic separation of the asphaltenes was made to obtain a Ni-enriched fraction. Although silica gel chromatography has been successfully used to enrich oil and rock extract fractions in Ni porpnhyrins,27 there are some reports that these compounds degrade during separation.11 The cyclograph system employed in this work is a centrifugally accelerated device for performing preparative thin-layer chromatographic separations on silica-gel. It provides faster elution of petroleum components and, thus, minimizes potential sample degradation during the separation process. The unit is routinely used in our lab for the separation of lube base oil and additives. It can handle a relatively
large quantity of material and provide enough material for metal analysis. In this study, 1 g of the asphaltene is fractionated following the procedure outlined in the Experimental Section. Results shown in Table 2 show a mass balance of 105% for the separation. The over recovery may be attributed to residual solvent or solids washed out from the cyclograph. Results of ICPMS analysis of the three fractions are given in Table 3. We expect Ni porphyrins to be less polar than the VO porphyrin and, thus, should be enriched in the first fraction eluted with toluene. Indeed, fraction 1, accounting for only 12% of total weight is enriched in nickel, containing 1359 ppm or nearly tripled the original Ni concentration. APPI-FTICR MS Analysis. Figure 3 shows the full range mass spectrum and selected nominal masses of fraction 1 by APPIFTICR MS. Mass peaks range from 400 to 1100 Da and peaked around 550 Da. APPI of hydrocarbon molecules only produce singly charged molecular ions. Consequently, the value of mass to charge ratio (m/z) is equal to the monoisotopic molecular weight of the hydrocarbon species. Molecular weight (in Da) and m/z are used interchangeably in this paper. The mass distribution is similar to that of other reported asphaltenes analyzed by APPIFTICR MS28-30 and confirm the presence of low molecular weight
(27) Van Berkel, G. J.; Quinones, M. A.; Quirke, J. M. E. Energy Fuels 2002, 7, 411–419.
(28) Smith, D. F.; Klein, G. C.; Yen, A. T.; Squicciarini, M. P.; Rodgers, R. P.; Marshall, A. G. Energy Fuels 2008, 22, 3112–3117.
416
Analytical Chemistry, Vol. 82, No. 1, January 1, 2010
Figure 4. C31 Ni DPEP isotope distribution. Table 4. C31 DPEP Nickel Porphyrin Isotope Distribution accurate mass formula C31H32N458Ni C3013CH32N458Ni C31H32N460Ni C2913C2H32N458Ni C3013CH32N460Ni C31H32N462Ni
isotope ratio
neutral (M) ion (M+•) exp. (M+•) theo. 518.1980 519.2014 520.1935 520.2048 521.1968 522.1910
518.1975 519.2009 520.1930 520.2042 521.1963 522.1905
518.1975 519.2008 520.1929 520.2040 521.1963 522.1904
1.00 0.35 0.39 0.06 0.13 0.05
exp. 1.00 0.34 0.37 0.06 0.12 0.04
asphaltene molecules (
