Throwing Light on Petroleum: Simulated Exposure ... - ACS Publications

Dec 11, 2013 - equivalents (a count of the number of rings plus double bonds within the carbon framework). During the .... The profile for the class b...
0 downloads 6 Views 2MB Size
Article pubs.acs.org/ac

Throwing Light on Petroleum: Simulated Exposure of Crude Oil to Sunlight and Characterization Using Atmospheric Pressure Photoionization Fourier Transform Ion Cyclotron Resonance Mass Spectrometry Matthew T. Griffiths, Raffaello Da Campo, Peter B. O’Connor, and Mark P. Barrow* Department of Chemistry, University of Warwick, Coventry, CV4 7AL, United Kingdom S Supporting Information *

ABSTRACT: The change in profile of crude oil following a release into the environment is a topic of significant interest, and there is a need to develop analytical methodologies for understanding natural processes which affect related complex mixture profiles. One such process is the exposure to sunlight. In the following investigation, three oil samples were studied: one served as a control, a second was subjected to irradiation by an ultraviolet lamp, and a third sample was irradiated by a SoLux light source which closely models the solar emission profile. The usage of the SoLux light source represents a new method which enables a controlled experiment to mimic the effects of sunlight upon the sample. Atmospheric pressure photoionization was selected as the primary ionization method due to the ability to ionize a broad range of compounds, including low polarity components which could not be observed using electrospray ionization. During a test of sample preparation methods, the addition of a protic cosolvent to the sample solutions was shown to broaden the range of heteroatom-containing components observed. Following characterization, it was found that the polyaromatic hydrocarbons did not change in profile, while compounds containing a heteroatom exhibited a tendency to oxidize following photoirradiation. Sulfur-containing compounds with a low number of double bond equivalents were among the most reactive components of the complex mixture. The photooxidation of compounds in petroleum, following exposure to sunlight, is expected to have significance with regards to solubility and potential toxicity.

P

present in fuels.4,6,7 Oxygen-containing compounds can include naphthenic acids,9 which can cause corrosion10,11 and can be hazardous to aquatic organisms.12−16 Naphthenic acids are particularly well-known with regards to the oil sands process waters associated with the Athabasca oil sands industry.9,14,17−24 During the production and transport of petroleum, both by land and by sea, there can be occasional releases into the environment, and it is known that natural processes can affect the chemical composition of petroleum. These releases can result from: damage to oil tankers, damaged pipelines, natural seepage, and blowouts.25As supplies of unconventional oil supplies are increasingly being exploited to meet demand, oil exploration has moved to more challenging areas. This increases the risk of petroleum accidentally being released into the environment, with deepwater spills being much harder to cap than those from shallower waters.25 An example of this was the 2010 Deepwater Horizon incident, which led to the largest accidental release of oil into the sea, releasing

etroleum is a vital resource and is likely to remain so for the foreseeable future.1 The U.S. Energy Information Administration stated in 2012 that, during the previous year, petroleum was the basis of 93% of the U.S. transportation sector’s energy consumption and 40% of the U.S. industrial sector’s energy consumption,2 in addition to being the precursor for many materials, such as plastics, synthetic rubber, solvents, pesticides, fertilizers, medicines, dyes, waxes, lubricants, and more. Because of its finite availability, production of petroleum is currently in decline, while worldwide consumption continues to increase. As a result, it has therefore become increasingly necessary to turn to sources of lower quality petroleum. These sources can have a higher content of sulfurcontaining compounds, acidic components, and other hydrocarbons which incorporate heteroatoms and may present challenges for the refining process. While the majority of a crude oil consists of pure hydrocarbons, it is often the heteroatom-containing species that are of greatest interest. It is known that nitrogen-containing components can poison catalysts,3−5 which presents challenges during processing of the crude oil, such as hydrocracking. Similarly, sulfur-containing components can poison catalysts4,6−8 and also present a problem with regards to pollution of the environment when © XXXX American Chemical Society

Received: August 9, 2013 Accepted: November 26, 2013

A

dx.doi.org/10.1021/ac4025335 | Anal. Chem. XXXX, XXX, XXX−XXX

Analytical Chemistry

Article

containing compounds, by positive-ion ESI52,53 and acidic components, such as naphthenic acids,9,11,54 by negative-ion ESI.55,56 Less polar compounds, such as polyaromatic hydrocarbons (PAHs) and many sulfur-containing components, such as thiophene-based compounds, cannot be ionized using ESI, however. Work by Müller et al.8 demonstrated the ability to observe polycyclic aromatic sulfur heterocycles (PASHs) by derivatizing the compounds to form sulfonium salts, prior to analysis by ESI. A less widespread ionization method, but one which affords the ability to ionize a wide range of less polar molecules and without the need for derivatization, is atmospheric pressure photoionization (APPI).57−59 The applications of APPI to petroleum60−62and other complex mixtures, such as oil sands process waters (OSPW),21 have been demonstrated. The ability to observe thiophene-based compounds and polycyclic aromatic hydrocarbons, for example, results in a more comprehensive characterization of a complex mixture and is particularly useful when monitoring the effects of a particular process on a complex mixture as a whole. APPI can also generate ions via different pathways, resulting in the formation of both radical cations and protonated ions from the same sample. This fact, coupled with the ability to ionize a wider range of molecules, means that mass spectra resulting from the application of APPI can be significantly more complex, further necessitating the need for ultrahigh resolving power and mass accuracy.63 The focus of the following study has been to develop a method for studying the effects of solar radiation upon crude oil. The effects of solar radiation have been simulated using a SoLux halogen bulb, which closely mimics natural sunlight, and the results were compared with those obtained using UV light and a control. APPI FTICR mass spectrometry was selected as the primary method to study detailed compositional changes of a crude oil sample at a molecular level, with complementary data being provided by negative-ion and positive-ion ESI.

approximately 4.9 million barrels of crude oil into the Gulf of Mexico.26The fate of the crude oil and debris from the incident has been a subject of significant interest.27−29 Weathering of the oil was studied and an increase in oxygen content, including formation of carboxylic acids, was observed.28 Natural seepage can also release significant amounts of petroleum into the environment; it has been estimated that between 560 000 and 1 400 000 barrels of oil are released annually into the environment in the Gulf of Mexico through natural means.30 Once released into the environment, the composition of petroleum can be affected by various processes, including: photodegradation,31,32biodegradation,26,33,34 and weathering.35,36 Understanding how these factors can alter the composition of the oil is therefore important and it has been stated that the effects of photochemical processes upon crude oil are relatively less well characterized.34 There is interest in methods of analysis of petroleum-related compounds following a discharge into the environment37−40 and compositional changes due to photochemical reactions may be significant with regards to environmental aspects of oil spills,41−43 potentially also affecting the toxicity and solubility of the oil. D’Auria et al.44 investigated the photochemical degradation of a sample of crude oil by direct irradiation, in the presence of a catalyst (TiO2) and with a catalyst on zeolite. Experiments were performed for 100 h, with samples suspended on water and exposed to irradiation from a 125 W ultraviolet (UV) lamp. It was reported that using a catalyst was the most effective way of degrading the crude oil, while the process was slower when performed with zeolite, which was attributed to light scattering. Pesarini et al.45 reported that, following irradiation by sunlight, the asphaltene content in an Arabian light oil and a Brazilian intermediate oil increased after only 5 h. Characterization was performed using thermogravimetric analysis, Fourier transform infrared (FT-IR) spectroscopy, nuclear magnetic resonance (NMR), and fluorescence spectroscopy, determining that the asphaltenes became less fluorescent but more aromatic. Bobinger et al.46 investigated the photooxidation of eleven benzothiophenes, which can be found in crude oil. Irradiation experiments were performed using sunlight or a medium pressure mercury lamp, where light with a wavelength of under 300 nm was filtered out. Gas chromatography with atomic emission detection (GC-AED) and gas chromatography−mass spectrometry (GC-MS) were used for characterizing the reaction products. It was found that the main products formed were sulfobenzoic acids, which lowered the pH of the solution. Fathalla et al.47 investigated the products of polycyclic aromatic sulfur heterocycles that were formed following photodegradation and found a large variety of sulfonic acids, aliphatic and aromatic acids, and alcohols. Crude oil is potentially the most complex of naturally occurring mixtures.48 Molecular characterization of crude oils is therefore highly challenging and, because of the inherent ultrahigh resolving power and mass accuracy associated with the technique, Fourier transform ion cyclotron resonance mass spectrometry49−51 (FTICR MS) is uniquely well-suited to characterization of such complex mixtures. A variety of ionization methods may be coupled with a mass spectrometer, each with their own advantages and disadvantages. Electrospray ionization (ESI) is one of the most commonly used ionization methods and has been used in conjunction with FTICR mass spectrometry for the determination of polar compounds within petroleum. Early work in this area focused upon the characterization of basic components, such as nitrogen-



EXPERIMENTAL PROCEDURES In this study, the crude oil that was used for the photodegradation experiments was a light-sour crude oil which is a standard reference material obtained from the National Institute of Standards and Technology (NIST, Gaithersburg, Maryland, U.S.A.). For the photodegradation experiment, there were three separate samples; a control, one irradiated with a 4 W UV lamp with emission at a wavelength of 254 nm (UltraViolet Products Ltd., Cambridge, U.K.) and a sample irradiated with a SoLux bulb (Q35MR16/CG/41/10-Solux, Eiko Ltd., Shawnee, Kansas, U.S.A.) with a color temperature of 4700 K and output of 35 W. The SoLux lamp was selected due to the similarity of its profile with that of solar radiation (Supporting Information Figure S1), and the UV lamp was chosen to monitor the effects of exposure to UV radiation, which is outside of the profile of the SoLux lamp. The output of the SoLux bulb was measured to be 54 klux; in comparison with an Oriel Sol3A solar simulator (Newport Spectra-Physics Ltd., Didcot, U.K.), the SoLux bulb was measured to have an output that is equal to 42% of solar irradiation. To provide a control sample, 10 mL of the crude oil was placed in a sealed conical flask which was wrapped in aluminum foil to prevent exposure to light. For irradiation with the SoLux and UV light sources, 10 mL of the crude oil was placed on a 150 mm watchglass, which was then positioned at a height of 11 cm under the corresponding lamp. The irradiation was B

dx.doi.org/10.1021/ac4025335 | Anal. Chem. XXXX, XXX, XXX−XXX

Analytical Chemistry

Article

Figure 1. Example broadband positive-ion APPI FTICR mass spectra of the samples, acquired at the completion of the irradiation experiment (938 h). The data shown are for the control sample and samples exposed to irradiation from a UV source and a SoLux source.

allowed to proceed for up to a total of 938 h (approximately five and a half weeks), which was the maximum length of time available for the particular experiment. Following a comparison of different solvent systems (see Supporting Information and Figures S2 and S3), the samples were dissolved using a ratio of 80:20 propan-2-ol/toluene (0.1 mg/mL) prior to analysis using mass spectrometry, as the addition of propan-2-ol (a protic solvent) enhanced the observation of many protonated, heteroatom-containing compound classes. For the characterization of the samples, a 12 T solariX FTICR mass spectrometer (Bruker Daltonik GmbH, Bremen, Germany) was used, fitted with an APPI II ion source (Bruker Daltonik GmbH, Bremen, Germany). Experiments were also later performed using negative-ion and positive-ion ESI to provide complementary information, regarding the acidic and basic components of the samples, although the low polarity components would not be accessible; further details of the ESI experiments are contained within the Supporting Information. The instrument was operated using the positive-ion mode and broadband mass spectra were acquired using the range m/z 147.4−3000. Each mass spectrum was the result of 300 scans, where 4 megaword data sets were collected. Following acquisition, each data set was zero-filled once and sine-bell apodization was applied. A fast Fourier transform (FFT) was used to convert from the time domain to the frequency domain, from which the data could be converted to mass spectra. Magnitude-mode data were used, with an associated resolving power of approximately 500 000 (fwhm) at m/z 400 after apodization. The mass spectra were externally calibrated using arginine clusters and then internally calibrated using homologous series associated with hydrocarbons not

containing a heteroatom, because of their abundance and wide m/z range. The data were then analyzed using a combination of DataAnalysis 4.0 SP4 (Bruker Daltonik GmbH, Bremen, Germany), Composer 1.0.5 (Sierra Analytics, Modesto, California, U.S.A.), Excel (Microsoft Corporation, Redmond, Washington, U.S.A.), and Aabel 3.0.6 (Gigawiz Ltd. Co., Tulsa, Oklahoma, U.S.A.).



RESULTS AND DISCUSSION At the completion of the photodegradation experiment after 938 h, there were clear, physical differences between the samples: while the control sample had changed very little, the sample exposed to the UV source had become more viscous and the sample exposed to the SoLux source demonstrated the most significant physical change, more closely resembling a solid material than a liquid (see Supporting Information Figure S4). Figure 1 shows the mass spectra obtained for the three samples after the period of irradiation. The distribution of signals spanned a wide m/z range (approximately m/z 200− 1400), and although there was a subtle shift toward higher m/z following exposure to the two sources of light, further data analysis was required to determine the changes in composition. Following the assignments of the peaks within the mass spectrum, various methods can be employed to visualize the data. Assignments can be categorized in terms of their carbon number, number of double bond equivalents within the carbon framework, and their heteroatom content. The compound class is determined by the heteroatom content; for example, “S” or “S1” would denote a composition of CcHhS1 and “S2” would denote a composition of CcHhS2. Where a molecule does not incorporate any heteroatoms (i.e., CcHh), it is assigned to the C

dx.doi.org/10.1021/ac4025335 | Anal. Chem. XXXX, XXX, XXX−XXX

Analytical Chemistry

Article

“HC” class. Assignments that include the label “[H]” indicate that the signals were detected as protonated ions; assignments that do not include this label indicate the signal correlates with a radical ion. Figure 2 provides an overview of the three sample profiles. The contribution from the HC class was noticeably consistent

While bar charts of total contributions from different compound classes can provide a useful overview, additional insight can be obtained by considering other factors associated with the assignments, such as carbon number and double bond equivalents (a count of the number of rings plus double bonds within the carbon framework). During the data analysis, it was found that the carbon number range for the individual classes was relatively unchanged on this time scale, but that there were clear differences between the DBE ranges. When considering compound classes that were among the most abundant, it was again found that the HC class was relatively uninfluenced by the exposure to light, both in terms of carbon number and DBE, but the S1 class (radical ions), the second most abundant compound class, demonstrated noticeable changes. Example plots of DBE versus carbon number are shown in Figure 3. It

Figure 2. Positive-ion APPI data after completion of the irradiation experiment (938 h), showing a comparison of the relative contributions from each compound class. The HC class, which does not contain a heteroatom, was relatively unchanged after exposure to the different sources of light. There was a significant decrease, however, in the proportion of the S1 and N1 classes, with an accompanying increase in oxygen-containing classes.

and did not appear to be influenced by exposure to light. This may be expected for photostable polyaromatic hydrocarbons and may indicate that such compounds would not degrade as easily in the environment, as well as having potential as markers for individual oil sources. One of the most noticeable trends were that selected sulfur-containing (S1, S1[H], S2, and S2[H]) and nitrogen-containing (N1[H] and N1) classes displayed significant reductions, while selected oxygen-containing classes (O1[H], O2[H], OS, NO[H], and O2S[H]) increased. The large increase in abundance of oxygen-containing classes following irradiation, especially with the SoLux lamp, can be attributed to photoinduced oxidation of the sample. Similarly, an increase in oxygen-containing compounds was seen in previous studies that exposed the samples to light.41Some of the oxygen-containing components (including those which also incorporate other heteroatoms) may be more acidic28 and would also likely be more soluble and more bioavailable64 and, therefore, potentially toxic in a marine environment. The results demonstrate that the compounds containing a heteroatom more readily undergo oxidation in the presence of light. Such behavior can be partly rationalized by taking into account the presence of nonbonding electrons associated with heteroatoms such as nitrogen and sulfur. The availability of such electrons makes the compounds containing these heteroatoms more inclined to react with oxidizing agents and generally broadens their absorption profile toward longer wavelengths. The samples exposed to the SoLux light exhibited greater change than the sample exposed to UV light alone. This can be explained by the fact that the SoLux bulb has both a higher power (35 W) and a broader emission spectrum than the monochromatic UV lamp (4 W) that was used.

Figure 3. Plots of double bond equivalents versus carbon number for the S1 compound class radical ions. The positive-ion APPI data shown were acquired after 938 h for the control sample and samples irradiated by the UV and SoLux sources. D

dx.doi.org/10.1021/ac4025335 | Anal. Chem. XXXX, XXX, XXX−XXX

Analytical Chemistry

Article

Figure 4. Positive-ion APPI data after 938 h for each sample, showing contributions to the observed protonated ions and radical ions as a function of compound class and double bond equivalents.

rings), and therefore larger DBE values, possess a larger internal “vibrational bath” and are more likely to accommodate increases in energy by means of rapid vibrational energy redistribution, reducing the breaking of covalent bonds. It is also possible, however, that intermolecular vibrational energy transfer occurs, with potential for greater efficiency of heat dissipation into the surrounding environment from the large heteroaromatic compounds. Based upon the data shown in Figure 3 and following the line of reasoning above, it is also possible that in the case of small, aromatic chromophores, covalently linked by alkyl side chains, any aliphatic backbone merely acts as a spectator during any energy redistribution process. In order to fully understand highly complex data sets, such as the behavior of petroleum samples, there is of course an ongoing need to study photochemical phenomena on a small scale and to relate this to the macroscopic behavior of a bulk material. For the S1 radical ion class, it can be seen that there are more intense contributions at 6, 9, and 12 DBE. This is considered to be indicative of favored structures, hence providing structural insight. For formulas incorporating a single sulfur atom, a minimum DBE of 6 would correlate with alkylated

should be noted that protonated ions, as observed in positiveion mode, or deprotonated ions, as would be observed in negative-ion mode, have DBE values that are half integers because of the change in electron availability associated gain or loss of a proton. For radical ions, however, the DBE value will be an integer. In this manner, the ion type can be distinguished by a calculation of the DBE. When examining the results for the S1 radical ion class as shown in Figure 3, the carbon number range again does not appear to be significantly affected, but the molecules of lower DBE are preferentially reduced. Compositions of higher DBE were relatively unaffected, potentially due to the enhanced photostability of their highly conjugated πsystems. In the case of small molecules with isolated aromatic ring structures, such as pyrrole- and thiophene-based structures with low DBE values, it is known that many of these species exhibit a strong propensity to undergo extremely rapid excited electronic state quenching, converting this excess electronic energy into heat in the electronic ground state;65,66 this process has been referred to as “photostability.” These isolated molecules are also known to undergo rapid fragmentation and degradation, however.66−68 In contrast, larger structures with higher numbers of adjoining rings (particularly aromatic E

dx.doi.org/10.1021/ac4025335 | Anal. Chem. XXXX, XXX, XXX−XXX

Analytical Chemistry

Article

and started at 0.5 DBE (1 DBE for the neutral molecule), indicating basicity, the potential inclusion of a carbonyl group, and that the components are therefore most likely to be sulfoxides.71 Following exposure to light, the samples displayed trends toward higher DBE, with the sample exposed to the SoLux lamp exhibiting components at higher DBE values than the sample exposed to UV light. Compounds with a higher oxygen or sulfur content were preferentially observed by negative-ion ESI, however, indicating acidity. With the significant increase in oxygen-containing classes overall, following exposure to light sources, it is postulated that crude oil that is exposed to sunlight may become more soluble in water and it has previously been reported that there can be an increase in toxicity after photooxidation,41−43,64,72,73 which would be highly relevant in the event of a discharge of crude oil into an aquatic environment. When comparing the positive-ion APPI data (Figure 4) for samples exposed to UV and SoLux sources, the samples irradiated by the SoLux source exhibited a greater decrease of the low DBE sulfur-containing species. This may be interpreted as photooxidation progressing further when using the SoLux light source, compared to the usage of the UV light source. Similarly, the N1[H] class was reduced in total contribution and there was an associated increase in the NO[H] class. It has previously been determined that, when performing APPI experiments, pyridinic nitrogen-containing structures preferentially form protonated ions, while pyrrolic nitrogen-containing compounds, which are weakly acidic, preferentially form radical ions, and thus the ion type is potentially indicative of structure; similarly, pyrrolic compounds would be observed using negative-ion ESI and pyridinic compounds would be observed using positive-ion ESI.62 Positive-ion ESI experiments (Supporting Information Figures S6, S7, and S10) show that the basic nitrogen-containing compounds predominate, as is known for many crude oil samples, and the profiles begin at a DBE of 3.5 for the positive ions (4 for the neutral molecule), which would be consistent with the presence of a six-membered aromatic ring, such as that found within pyridinic structures. There was no indication of selective oxidation of particular homologous series of the basic nitrogen-containing components. For the positive-ion APPI data (Figure 4), the changes in profile for many radical ion species were less pronounced, potentially indicating that the molecules which preferentially ionize as radicals, rather than readily protonating, were more stable with respect to oxidation. The HC class, which formed the greatest contribution to the profile of the radical ion species in the APPI data, was the most stable of the compound classes, highlighting that such compounds may also be expected to be resistant to compositional change in the environment. For all three samples, there was a marked increase in the contributions to the HC radical ion class when moving from the range of DBE 1−3 to a DBE of 4 or higher; a minimum DBE of 4 is required to form a six-membered aromatic ring, which would enhance the thermodynamic stability of the molecule. As also shown in Figure 3, the S1 radical ion class, the second largest component of the overall complex mixture, exhibited the preferred DBE of 6 and 9, while the lower DBE range of the S1 class decreased proportionally more when moving from the control sample to the irradiated samples. As observed for the protonated ions, the N1 radical ion class decreased when moving from the control sample to the irradiated samples, and the same was true for the S2 radical ion class. Supporting

benzothiophenes and a minimum DBE of 9 would be consistent with alkylated dibenzothiophenes, for example; such compounds are well-known components of petroleum samples. Comparing the control sample and the SoLux sample, it can be seen that the low DBE range (DBE of 2−5 in this case, which can correlate with thiols, thioethers, and thiophenes) disappeared, but that the homologous series at a DBE of 6 and higher remained stable and this would correspond with the enhanced stability of the aforementioned benzothiophene and dibenzothiophene structures. This observation is consistent with recent work by Hegazi et al.69 where the effects of weathering upon sulfur-containing compounds in a crude oil were monitored. It is also worth noting that polycyclic aromatic sulfur heterocycles would not be detected unless using APPI or using ESI following derivatization of the specific components of interest.70 Indeed, the most abundant compound classes (HC, S1, S1[H], and HC[H]) are not accessible using ESI, but ESI data (Supporting Information Figure S5) can still play a complementary role in providing information about the likely structures of acidic and basic components. As the greatest correlation, for a range of compound classes, appeared to be with DBE, graphs of contributions as a function of compound class and DBE were plotted (Figure 4). The assignments associated with the protonated ions exhibited the most considerable change, with assignments associated with radical ions displaying less change following exposure to light. For the protonated species, moving from the control sample to the samples exposed to UV and SoLux light, there is a significant reduction in the S1[H] class, particularly at lower DBE (approximately DBE 0.5−7.5), and in the S2[H] class. Potentially favored DBE series for the S1[H] class were observed at DBE of 1.5−3.5 (2−4 for the neutral species), 5.5− 6.5 (6−7 for the neutral species), and 8.5−9.5 (9−10 for the neutral species). In addition to the decrease of particular compound classes, there was an accompanying increase in the O1[H], O2[H], OS[H], and O2S[H] classes. The increase in the O2[H] class was pronounced in the sample irradiated by the SoLux bulb, with species spanning a broad DBE range. Comparison with positive-ion and negative-ion ESI data (Supporting Information) revealed that the O2-containing class was acidic, due to the fact that it is primarily observed in negative-ion mode. The profile for the class began at a DBE of 1.5 for the negative ions (Supporting Information Figures S7 and S8; alternative representation shown in Supporting Information Figure S9) or 1 DBE for the neutral species, and combined with the acidic nature of the compounds, this would be consistent with the formation of naphthenic acids,9,56 which include a carboxylic acid group and hence require a minimum of 1 DBE. The number of O2-containing components and the number of higher DBE (or more highly hydrogen deficient) components increased following exposure to light, with the SoLux lamp having the most pronounced effect. At a DBE of 4.5 or higher for the negative ions (4 or higher for the neutral species), one or more six-membered aromatic rings may be incorporated. When considering the positive-ion APPI data (Figure 4) and examining the OS[H] class, the increase in relative abundance appears to occur within the low DBE range, where the S1[H] class decreased in relative abundance; this would correlate with oxidation of the S1[H] compound class without a significant change in double bond equivalents in the process. The ESI experiments highlighted that the OScontaining class was primarily observed in positive-ion mode F

dx.doi.org/10.1021/ac4025335 | Anal. Chem. XXXX, XXX, XXX−XXX

Analytical Chemistry

Article

them as a potential marker for the origins of a sample after exposure to the environment.

Information Figures S7 and S8 exhibit the favored homologous series (DBE of 9, 12, and 15), consistent with the presence of pyrrolic components, such as carbazole-related compounds, which are known to preferentially ionize by negative-ion ESI rather than positive-ion ESI. In contrast to the positive-ion ESI experiments, but consistent with a comparison between the N1[H] and N1 radical ion classes within the positive-ion APPI data, there was a proportionally greater decrease in the contribution from the N1 class in the negative-ion ESI data following exposure to light (Supporting Information Figures S6, S7, and S8), which would indicate that the pyrrolic nitrogencontaining compounds more readily underwent photooxidation than the pyridinic nitrogen-containing compounds. In parallel to the increase in oxygen-containing classes for the protonated species within the positive-ion APPI data, there was an increase in the O1 radical ion class. Overall, the results indicate that the PAHs within the system were resistant to modification, while heteroatom-containing compounds were likely to become oxidized following irradiation and that molecules of low DBE were most reactive.



ASSOCIATED CONTENT

S Supporting Information *

Additional material as described in the text. This material is available free of charge via the Internet at http://pubs.acs.org.



AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. Notes

The authors declare no competing financial interest.



ACKNOWLEDGMENTS



REFERENCES

The authors thank David Stranz (Sierra Analytics) for software development, Gareth Roberts (University of Warwick) for useful discussion, and Andrew Marsh and Jane Emmerson (both from the University of Warwick) for their contributions to the photoirradiation experiments. The authors also gratefully acknowledge the University of Warwick for providing funding.



CONCLUSIONS The usage of a SoLux lamp has been demonstrated to offer a method for studying the photoinduced reactions of petroleum in a controlled laboratory environment. From the irradiation of a light-sour crude oil with a SoLux halogen bulb and a UV lamp, it has been observed that heteroatom-containing compounds are more prone to undergo photooxidation than pure hydrocarbon compounds. The presence of oxygencontaining compounds increased following exposure to light, indicating that irradiation was promoting the oxidation of the more reactive components of the crude oil. On the time scale of approximately five weeks, the reactivities of the different components most closely correlated with compound class and DBE, rather than carbon number. Structures containing a lower number of double bond equivalents were typically more reactive. Structurally preferred DBE values for particular compound classes, such as DBE of 6 and higher for the S1 class (commonly associated with benzothiophenes), demonstrated a resistance toward change in composition, however. The choice of solvents employed also played an important role in determining the compounds that could be observed. The addition of a protic solvent increased the range of observable components, a factor which is important when characterizing the profile of a complex mixture such as a crude oil. Samples exposed to light from the SoLux lamp, which simulates the emission profile of sunlight, exhibited a greater change in composition than samples exposed to a UV light source. This may be attributed to the fact that the SoLux light source had a broader emission spectrum and a higher power output than the UV lamp. Following photooxidation, it is likely that many of the original compounds, which had a low polarity, would become more soluble in water and may also display an increase in acidity. Indeed, of the oxygen-containing compounds produced, including sulfoxides, a number of photooxidation products were determined to be naphthenic acids. This would represent a potential cause of increased toxicity of crude oil after exposure to a light source and therefore an additional danger to the aquatic environment in the event of an oil spillage. By contrast, the pure hydrocarbon components were observed to be highly stable; a DBE of 4 or higher was structurally preferred, which would correspond with polyaromatic hydrocarbons being less likely to degrade, highlighting

(1) Barrow, M. P. Biofuels 2010, 1, 651−655. (2) Annual Energy Review 2011; Energy Information Administration, Department of Energy: Washington, DC, 2012. (3) Hajji, A. A.; Muller, H.; Koseoglu, O. R. Oil Gas Sci. Technol. 2008, 63, 115−128. (4) Bae, E. J.; Na, J. G.; Chung, S. H.; Kim, H. S.; Kim, S. Energy Fuels 2010, 24, 2563−2569. (5) Rodgers, R. P.; McKenna, A. M. Anal. Chem. 2011, 83, 4665− 4687. (6) Guan, S.; Marshall, A. G.; Scheppele, S. E. Anal. Chem. 1996, 68, 46−71. (7) Rodgers, R. P.; White, F. M.; Hendrickson, C. L.; Marshall, A. G.; Andersen, K. V. Anal. Chem. 1998, 70, 4743−4750. (8) Müller, H.; Andersson, J. T.; Schrader, W. Anal. Chem. 2005, 77, 2536−2543. (9) Headley, J. V.; Peru, K. M.; Barrow, M. P. Mass Spectrom. Rev. 2009, 28, 121−134. (10) Turnbull, A.; Slavcheva, E.; Shone, B. Corrosion 1998, 54, 922− 930. (11) Slavcheva, E.; Shone, B.; Turnbull, A. Br. Corros. J. 1999, 34, 125−131. (12) Rockhold, W. AMA Arch. Ind. Health 1955, 12, 477−482. (13) Rogers, V. V.; Wickstrom, M.; Liber, K.; MacKinnon, M. D. Toxicol. Sci. 2002, 66, 347−355. (14) Headley, J. V.; McMartin, D. W. J. Environ. Sci. Health, Part A: Toxic/Hazard. Subst. Environ. Eng. 2004, A39, 1989−2010. (15) Barrow, M. P.; Headley, J. V.; Peru, K. M.; Derrick, P. J. J. Chromatogr., A 2004, 1058, 51−59. (16) Jones, D.; Scarlett, A. G.; West, C. E.; Rowland, S. J. Environ. Sci. Technol. 2011, 45, 9776−9782. (17) Bataineh, M.; Scott, A. C.; Fedorak, P. M.; Martin, J. W. Anal. Chem. 2006, 78, 8354−8361. (18) Smith, D. F.; Schaub, T. M.; Rahimi, P.; Teclemariam, A.; Rodgers, R. P.; Marshall, A. G. Energy Fuels 2007, 21, 1309−1316. (19) Headley, J. V.; Peru, K. M.; Armstrong, S. A.; Han, X. M.; Martin, J. W.; Mapolelo, M. M.; Smith, D. F.; Rogers, R. P.; Marshall, A. G. Rapid Commun. Mass Spectrom. 2009, 23, 515−522. (20) Da Campo, R.; Barrow, M. P.; Shepherd, A. G.; Salisbury, M.; Derrick, P. J. Energy Fuels 2009, 23, 5544−5549. (21) Barrow, M. P.; Witt, M.; Headley, J. V.; Peru, K. M. Anal. Chem. 2010, 82, 3727−3735. G

dx.doi.org/10.1021/ac4025335 | Anal. Chem. XXXX, XXX, XXX−XXX

Analytical Chemistry

Article

(22) Headley, J. V.; Barrow, M. P.; Peru, K. M.; Derrick, P. J. J. Environ. Sci. Health, Part A: Toxic/Hazard. Subst. Environ. Eng. 2011, 46, 844−854. (23) Rowland, S. J.; Scarlett, A. G.; Jones, D.; West, C. E.; Frank, R. A. Environ. Sci. Technol. 2011, 45, 3154−3159. (24) Headley, J. V.; Barrow, M. P.; Peru, K. M.; Fahlman, B.; Frank, R. A.; Bickerton, G.; McMaster, M. E.; Parrott, J.; Hewitt, L. M. Rapid Commun. Mass Spectrom. 2011, 25, 1899−1909. (25) Jernelov, A. AMBIO 2010, 39, 353−366. (26) Kostka, J. E.; Prakash, O.; Overholt, W. A.; Green, S. J.; Freyer, G.; Canion, A.; Delgardio, J.; Norton, N.; Hazen, T. C.; Huettel, M. Appl. Environ. Microbiol. 2011, 77, 7962−7974. (27) Reddy, C. M.; Arey, J. S.; Seewald, J. S.; Sylva, S. P.; Lemkau, K. L.; Nelson, R. K.; Carmichael, C. A.; McIntyre, C. P.; Fenwick, J.; Ventura, G. T.; Van Mooy, B. A.; Camilli, R. Proc. Natl. Acad. Sci. U.S.A. 2012, 109, 20229−20234. (28) Aeppli, C.; Carmichael, C. A.; Nelson, R. K.; Lemkau, K. L.; Graham, W. M.; Redmond, M. C.; Valentine, D. L.; Reddy, C. M. Environ. Sci. Technol. 2012, 46, 8799−8807. (29) Carmichael, C. A.; Arey, J. S.; Graham, W. M.; Linn, L. J.; Lemkau, K. L.; Nelson, R. K.; Reddy, C. M. Environ. Res. Lett. 2012, 7, 015301. (30) Schmidt Etkin, D. Analysis of U.S. Oil Spillage, API Publication 356; American Petroleum Institute: Washington, DC, 2009. (31) Payne, J. R.; Phillips, C. R. Environ. Sci. Technol. 1985, 19, 569− 579. (32) Nicodem, D. E.; Fernandes, M. C. Z.; Guedes, C. L. B.; Correa, R. J. Biogeochemistry 1997, 39, 121−138. (33) Wang, Z.; Fingas, M.; Blenkinsopp, S.; Sergy, G.; Landriault, M.; Sigouin, L.; Foght, J.; Semple, K.; Westlake, D. W. S. J. Chromatogr. A 1998, 809, 89−107. (34) Dutta, T. K.; Harayama, S. Environ. Sci. Technol. 2000, 34, 1500−1505. (35) Rodgers, R.; Blumer, E. N.; Freitas, M. A.; Marshall, A. G. Environ. Sci. Technol. 2000, 34, 1671−1678. (36) Bacon, M. M.; Electricwala, M.; Romero-Zeron, L. B. Petrol. Sci. Technol. 2011, 29, 349−357. (37) Mansuy, L.; Philp, R. P.; Allen, J. Environ. Sci. Technol. 1997, 31, 3417−3425. (38) Wang, Z.; Fingas, M.; Page, D. S. J. Chromatogr. A 1999, 843, 369−411. (39) Christensen, J. H.; Tomasi, G. J. Chromatogr. A 2007, 1169, 1− 22. (40) Avino, P.; Notardonato, I.; Cinelli, G.; Russo, M. V. Curr. Anal. Chem. 2009, 5, 339−346. (41) Maki, H.; Sasaki, T.; Harayama, S. Chemosphere 2001, 44, 1145−1151. (42) Lee, R. F. Spill. Sci. Technol. Bull. 2003, 8, 157−162. (43) Arfsten, D. P.; Schaeffer, D. J.; Mulveny, D. C. Ecotoxicol. Environ. Safe. 1996, 33, 1−24. (44) D’Auria, M.; Emanuele, L.; Racioppi, R.; Velluzzi, V. J. Hazard. Mater. 2009, 164, 32−38. (45) Pesarini, P. F.; de Souza, R. G. S.; Correa, R. J.; Nicodem, D. E.; de Lucas, N. C. J. Photochem. Photobiol. A 2010, 214, 48−53. (46) Bobinger, S.; Andersson, J. T. Environ. Sci. Technol. 2009, 43, 8119−8125. (47) Fathalla, E. M.; Andersson, J. T. Environ. Toxicol. Chem. 2011, 30, 2004−2012. (48) Head, I. M.; Jones, D. M.; Röling, W. F. Nat. Rev. Microbiol. 2006, 4, 173−182. (49) Amster, I. J. J. Mass Spectrom. 1996, 31, 1325−1337. (50) Marshall, A. G.; Hendrickson, C. L.; Jackson, G. S. Mass Spectrom. Rev. 1998, 17, 1−35. (51) Barrow, M. P.; Burkitt, W. I.; Derrick, P. J. Analyst 2005, 130, 18−28. (52) Qian, K.; Rodgers, R. P.; Hendrickson, C. L.; Emmett, M. R.; Marshall, A. G. Energy Fuels 2001, 15, 492−498. (53) Hughey, C. A.; Rodgers, R. P.; Marshall, A. G. Anal. Chem. 2002, 74, 4145−4149.

(54) Hsu, C. S.; Dechert, G. J.; Robbins, W. K.; Fukuda, E. K. Energy Fuels 2000, 14, 217−223. (55) Qian, K.; Robbins, W. K.; Hughey, C. A.; Cooper, H. J.; Rodgers, R. P.; Marshall, A. G. Energy Fuels 2001, 15, 1505−1511. (56) Barrow, M. P.; McDonnell, L. A.; Feng, X.; Walker, J.; Derrick, P. J. Anal. Chem. 2003, 75, 860−866. (57) Robb, D. B.; Covey, T. R.; Bruins, A. P. Anal. Chem. 2000, 72, 3653−3659. (58) Raffaelli, A.; Saba, A. Mass Spectrom. Rev. 2003, 22, 318−331. (59) Hanold, K. A.; Fischer, S. M.; Cormia, P. H.; Miller, C. E.; Syage, J. A. Anal. Chem. 2004, 76, 2842−2851. (60) Purcell, J. M.; Hendrickson, C. L.; Rodgers, R. P.; Marshall, A. G. Anal. Chem. 2006, 78, 5906−5912. (61) Purcell, J. M.; Juyal, P.; Kim, D. G.; Rodgers, R. P.; Hendrickson, C. L.; Marshall, A. G. Energy Fuels 2007, 21, 2869−2874. (62) Purcell, J. M.; Rodgers, R. P.; Hendrickson, C. L.; Marshall, A. G. J. Am. Soc. Mass Spectrom. 2007, 18, 1265−1273. (63) Marshall, A. G.; Rodgers, R. P. Proc. Natl. Acad. Sci. U.S.A. 2008, 105, 18090−18095. (64) Charrié-Duhaut, A.; Lemoine, S.; Adam, P.; Connan, J.; Albrecht, P. Org. Geochem. 2000, 31, 977−1003. (65) Sobolewski, L.; Domcke, W.; Dedonder-Lardeux, C.; Jouvet, C. Phys. Chem. Chem. Phys. 2002, 4, 1093−1100. (66) Ashfold, M. N.; King, G. A.; Murdock, D.; Nix, M. G.; Oliver, T. A.; Sage, A. G. Phys. Chem. Chem. Phys. 2010, 12, 1218−1238. (67) Yu, H.; Evans, N. L.; Stavros, V. G.; Ullrich, S. Phys. Chem. Chem. Phys. 2012, 14, 6266−6272. (68) Williams, C. A.; Roberts, G. M.; Yu, H.; Evans, N. L.; Ullrich, S.; Stavros, V. G. J. Phys. Chem. A 2011, 116, 2600−2609. (69) Hegazi, A. H.; Fathalla, E. M.; Panda, S. K.; Schrader, W.; Andersson, J. T. Chemosphere 2012, 89, 205−212. (70) Panda, S. K.; Andersson, J. T.; Schrader, W. Angew. Chem., Int. Ed. 2009, 48, 1788−1791. (71) Liu, P.; Xu, C.; Shi, Q.; Pan, N.; Zhang, Y.; Zhao, S.; Chung, K. H. Anal. Chem. 2010, 82, 6601−6606. (72) Duesterloh, S.; Short, J. W.; Barron, M. G. Environ. Sci. Technol. 2002, 36, 3953−3959. (73) Mearns, A. J.; Reish, D. J.; Oshida, P. S.; Ginn, T.; RempelHester, M. A. Water Environ. Res. 2011, 83, 1789−1852.

H

dx.doi.org/10.1021/ac4025335 | Anal. Chem. XXXX, XXX, XXX−XXX