Detailed characterization of polar compounds of residual oil in contaminated soil revealed by Fourier transform ion cyclotron resonance mass spectrometry

Chemosphere 85 (2011) 609–615

Contents lists available at ScienceDirect

Chemosphere journal homepage: www.elsevier.com/locate/chemosphere

Detailed characterization of polar compounds of residual oil in contaminated soil revealed by Fourier transform ion cyclotron resonance mass spectrometry Jian Wang, Xu Zhang, Guanghe Li ⇑ School of Environment, Tsinghua University, Beijing 100084, China State Key Joint Laboratory of Environmental Simulation & Pollution Control, Tsinghua University, Beijing 100084, China

a r t i c l e

i n f o

Article history: Received 24 March 2011 Received in revised form 23 June 2011 Accepted 26 June 2011 Available online 20 July 2011 Keywords: Bioremediation Ozonation Residual oil Fourier transform ion cyclotron resonance mass spectrometry Soil pollution

a b s t r a c t Effects of remediation technologies on polar compounds of crude oil in contaminated soils have not been well understood when compared to hydrocarbons. In this study, ultrahigh resolution Fourier transform ion cyclotron resonance mass spectrometry (FT-ICR MS) was used to characterize the changes in NSO polar compounds of crude oil and residual oil after long-term natural attenuation, biostimulation and subsequent ozonation following biostimulation of contaminated soils. N1 and O1 species, which were abundant in the crude oil, were selectively biodegraded, and species with higher double bond equivalent values and smaller carbon numbers appeared to be more resistant to microbial alteration. O2–O6 species were enriched by biodegradation and contained a large number of compounds with a high degree of unsaturation. Ozone could react with a variety of polar compounds in residual oil after biodegradation and showed high reactivity with polar species containing aromatic or multi-aliphatic rings, including the residual N1 and O1 species, naphthenic acids and unsaturated O3–O6 compounds. Fatty acids and O3–O8 species dominated by saturated alkyl compounds were resistant to ozonation or the primarily incomplete ozonation products. Principal component analysis of identified peaks in the FT-ICR MS spectra provided a comprehensive overview of the complex samples at the molecular level and the results were consistent with the detailed analysis. Taken together, these results showed the high complexity of polar compounds in residual oils after biodegradation or ozonation in contaminated soil and would contribute to a better understanding of bioremediation and ozonation processes. Ó 2011 Elsevier Ltd. All rights reserved.

1. Introduction Contamination of water and soil by crude oil during exploration, production, maintenance, transportation, storage, and accidental release has caused significant environmental impacts and presented substantial hazards to humans and the ecosystem. Crude oil is considered to be one of the most complex organic mixtures in the world (Marshall and Rodgers, 2003), consisting of normal and branched alkanes, cycloalkanes, monocyclic aromatic hydrocarbons and polycyclic aromatic hydrocarbons (PAHs), and polar compounds containing heteroatoms. Remediation technologies such as natural/enhanced biodegradation, phytoremediation and chemical oxidation have been developed for site cleanup in recent decades. Changes in hydrocarbons in response to different treatments have been well documented with the aid of GC/FID, GC/MS, and two-dimensional GC (GCXGC) coupled with FID or time-of-flight mass spectrometry (TOF-MS) ⇑ Corresponding author. Address: Sino-Italian Tsinghua Ecological & Energy Efficient Building, Tsinghua University, Beijing 100084, China. Tel./fax: 86 10 62773232. E-mail addresses: [email protected] (J. Wang), [email protected] (G. Li). 0045-6535/$ - see front matter Ó 2011 Elsevier Ltd. All rights reserved. doi:10.1016/j.chemosphere.2011.06.103

(Wang and Fingas, 1997; Skoczynnska et al., 2008). Moreover, evaluation of the removal efficiency of remediation technologies and risk assessment of contaminated/remediated sites are usually based on concerned hydrocarbons. When compared to hydrocarbons, the fate of polar compounds of crude oils in contaminated environments has been less studied. Although polar compounds account for a small portion of unweathered crude oils, they usually become enriched during degradation processes such as biodegradation and chemical oxidation as a combined result of selective removal, preservation, and formation. In some severely biodegraded oils, polar compounds may become the major compositions. Characterization of the polar compounds is important for the following reasons: (1) polar compounds altered during remediation processes can be used as indices of degradation and evidence of its pathway, and (2) some polar compositions are shown to be recalcitrant to biodegradation and toxic to the environment (Lundstedt et al., 2007; Whitby, 2010). Therefore, revealing the polar compounds in environmental samples can allow us to obtain more information regarding their fate and potential toxicity in the environment. With this information, the efficiency, limitation and/or side effects of different remediation technologies can be better evaluated.

610

J. Wang et al. / Chemosphere 85 (2011) 609–615

Although GC/MS and/or GCXGC/TOF-MS can be used to identify some of the polar compounds in crude oil, such as carbazoles and benzocarbazoles, alcohols, naphthenic acids and ketone/quinonesubstituted PAHs (Huang et al., 2003; Clemente and Fedorak, 2005; Rushdi et al., 2006; Skoczynnska et al., 2008; Layshock et al., 2010), it is not ideal to use these methods for large-scale characterization of the polar compounds in petroleum. This is due to the limitations of the methods, which include the availability of standard substances, low volatility and complexity of the polar compounds in petroleum, trace amounts of individual species and/or no universal procedure for sample preparation (Hughey et al., 2002). A relatively new analytical method, Fourier transform ion cyclotron resonance mass spectrometry (FT-ICR MS), affords ultra-high mass accuracy and resolving power, making it possible to identify individual molecular formulas in complex mixtures (Marshall and Rodgers, 2008) and to resolve thousands of compounds in samples without previous separation (Hughey et al., 2002; Shi et al., 2010a). Using this technique, Kim et al. (2005) characterized, in detail, the alteration of NSO compounds in crude oil based on biodegradation reservoir conditions and showed that the polar compounds were not as bioresistant as once believed. Two publications reported detailed compositional changes in crude oil in soil extracts subject to biodegradation in field (Hughey et al., 2007, 2008). These reports showed that the oxygen-only species increased and highly oxygenated compounds (up to O6 species) appeared after biodegradation when compared to the undegraded source oil. Additionally, these studies demonstrated that naphthenic acids could be indicators of crude oil biodegradation in soil. Lu et al. (2010) characterized residual oil treated with Fenton-like oxidation and found that cyclic naphthenic acids were removed by the Fenton reaction. Overall, these previous reports demonstrate that FT-ICR MS is a promising method to monitor the fate of polar compounds during different remediation processes, expanding our understanding of the resistance/reactivity of the compounds, possible reaction mechanisms and formation of new products. In this study, we investigated the changes in acidic and neutral NSO compounds in residual oil as a result of biodegradation and subsequent ozonation in soil formerly contaminated by crude oil. The crude oil contaminated soil underwent long-term biodegradation, and ozonation was used to further remove the residual components in the soil that were resistant to biodegradation. Our previous study demonstrated the feasibility of ozonation to remove recalcitrant residual oils from contaminated soil (Liang et al., 2009). However, the changes in the polar compounds in residual oil treated by ozonation have never been presented, even though ozonation of petroleum contaminated soils has been intensively studied. Detailed characterization of the chemical compositions using FT-ICR MS was presented in this study to reveal: (1) the resistant polar compositions that accumulated during biodegradation, (2) the removal of these resistant compositions by subsequent ozonation and (3) possible intermediates of ozonation. The results presented here may contribute to a better understanding of the remediation processes on a molecular level and facilitate the development of proper management strategies for petroleum contaminated sites.

2. Materials and methods 2.1. Contaminated soil treatments Soil contaminated with crude oil caused by oil spills during exploitation was collected from Daqing Oilfield, China in February 2004. The soil, which had a pH of 7.3, was classified as loamy sand (0.6% clay, 24.7% silt, and 74.7% sand). The soil was biodegraded separately by natural attenuation (NA) and biostimulation (BS) methods in the laboratory at room temperature (20–25 °C). During

biodegradation, the soil was turned over regularly to maintain aerobic conditions, and the moisture was kept at 10–25% (w/w). For BS, nutrients were added to the soil according to C:P:N:K = 100:10:1:1. The biodegradation began in May 2004 and was maintained for 250 d. Then the soil was stored at room temperature for more than 600 d without special care. NA- and BS-treated soils were collected and stored at 4 °C from January 2008 until further use. Ozonation was used to further remove the residual oil in BS-treated soil by laboratory-scale column experiments at room temperature in May 2008, as shown in Supplementary Material (SM), Fig. SM-1. Briefly, 100 g of the soil (with a moisture of 11%, sieved through a 1.25 mm mesh) was packed into a column (3 cm id), after which gaseous ozone (44 mg L 1) produced from oxygen (>99.5%) was introduced at a flow rate of 0.5 L min 1. Ozonation was conducted for 2, 6, and 15 h, separately. Crude oil collected from an oil well near the contaminated site was used as a reference because the oil in the original contaminated soil was not available. Thus, the crude oil and residual oils in NA, BS, and ozonated soils (2, 6, and 15 h; labeled as OZ1, OZ2, and OZ3, respectively) were analyzed in this study. 2.2. Oil sample preparation The crude oil was dissolved in chloroform and cleaned up using a glass column packed with anhydrous sodium sulfate to remove the water and undissolved solids. The residual oil in the soil samples was extracted by ultrasonic-aided Soxhlet extraction (Liang et al., 2009). The oil contents in the soils were determined by the gravimetric method and expressed as total solvent extractable matters (TSEM). The characteristics of the crude oil and residual oil (soil extracts) are shown in Table 1, and the detailed analytical methods can be found in SM. As expected, BS showed a higher degree of biodegradation than NA, resulting in lower TSEM, n-alkanes, PAHs and hopanes when compared to NA. Ozonation was effective at reducing the residual oil after BS, as an additional 11%, 24%, and 33% of TSEM were removed after 2, 6, and 15 h of ozonation, respectively. Biodegradation resulted in decreases of saturates and aromatics and increases of resins. Ozonation removed aromatics, saturates (including n-alkanes and biomarkers such as hopanes) and resins from the residual oil. The asphaltenes, a group of polar compositions that did not dissolve in n-hexane, did not change significantly in response to biodegradation, but increased dramatically in ozonated samples. 2.3. Negative-ion electrospray ionization(ESI) FT-ICR analysis Negative-ion ESI FT-ICR MS was used to analyze the polar NSO compounds in the crude oil and TSEMs. The crude oil and the TSEMs of soils (mixture of duplicate extracts for each soil sample) were dissolved in toluene to produce 10 mg mL 1 solutions. A 20 lL aliquot of the solution was diluted with 1 mL toluene/methanol (1:1, v/v) solution and then spiked with 10 lL 35% (w/w) NH4OH to facilitate deprotonation of acidic compounds in the ESI source. The analysis was conducted on an Apex-ultra FT-ICR MS (Bruker Daltonics, USA) equipped with a 9.4 T superconducting magnet. The operating conditions of the FT-ICR MS have been described elsewhere (Shi et al., 2010a,b). Peaks with m/z within 200–700 and a signal intensity greater than five times the standard deviation of the baseline noise (SNR > 5) in the FT-ICR mass spectra were exported to a spreadsheet. The elemental compositions were assigned solely based on mass measurement to ±1 ppm facilitated by Kendrick mass analysis (Marshall and Rodgers, 2003, 2008). The molecular formula assignment was limited to 100 12C, 200 1H, 3 14N, 9 16O, 2 32S, and 2 13C atoms. A homologous series, in which members differ by only nCH2, is described by its class (heteroatom contents) and number of rings plus double bonds. The latter parameter is

611

J. Wang et al. / Chemosphere 85 (2011) 609–615 Table 1 Characteristics of the crude oil and residual oils (soil extracts). TSEM (mg g

Crude oilf NA BS OZ1 OZ2 OZ3

1 a

Crude oil/TSEM fractionsb

)

– 36.9 ± 0.3g 27.0 ± 0.2 20.7 ± 0.6 13.7 ± 0.3 8.4 ± 0.6

Saturates

Aromatics

Resins

Asphaltenes

58.9 19.9 ± 0.2 9.1 ± 0.1 5.5 ± 0.4 3.0 ± 0.2 1.8 ± 0.3

22.2 10.0 ± 0.1 7.7 ± 1.1 6.0 ± 0.2 2.8 ± 0.1 1.8 ± 0.1

13.7 6.1 ± 0.0 9.8 ± 0.9 5.6 ± 0.6 4.0 ± 0.1 2.4 ± 0.3

5.2 1.0 ± 0.0 0.4 ± 0.0 3.6 ± 0.5 3.9 ± 0.3 2.4 ± 0.2

R n-alkanesc

R PAHsd

17a(H),21b(H)-hopanee

145.1 296 ± 1 91 ± 3 85 ± 5 57 ± 2 45 ± 4

2.7 12.7 ± 0.2 7.5 ± 0.1 4.8 ± 0.1 2.6 ± 0.1 1.0 ± 0.2

0.9 42 ± 2 30 ± 1 16 ± 4 8±1 6±1

a

TSEM: total solvent extractable matters; TSEM in the contaminated soil before treatment was 56.0 mg g 1. % for the crude oil and mg g 1 soil for the other samples. c Sum of the concentrations of n-C12 to n-C40 alkanes, mg g 1 for the crude oil and mg kg 1 soil for the other samples. d Sum of the concentrations of detectable 16 EPA’s priority PAHs, mg g 1 for the crude oil and mg kg 1 soil for the other samples. e mg g 1 for crude oil and mg kg 1 soil for the other samples. f Crude oil: the reference crude oil collected in Daqing oilfield, NA: soil treated by natural attenuation, BS: soil treated by biostimulation, OZ1: soil ozonated for 2 h after biostimulation, OZ2: soil ozonated for 6 h after biostimulation, OZ3: soil ozonated for 15 h after biostimulation. g Standard errors of the means (n = 2). b

the RA of O4–O8 species decreased when compared to OZ2, possibly due to the decomposition of these complex intermediates in response to longer ozonation.

expressed in double bond equivalent (DBE) calculated as DBE = c h/2 + n/2 + 1 for a compound with a molecular formula of CcHhNnOoSs. The relative abundance (RA), which is defined as the signal intensity (peak height) of a peak divided by the total intensity of all detected peaks, was used to characterize the change in corresponding compositions. It should be noted that different compounds may have different ionization efficiencies and that the RA of one compound is affected by its own transformation and that of other species. Therefore, the RA value cannot be treated as an accurate quantification. However, these data do generally reflect the elemental changes (Hughey et al., 2007).

3.2. Changes in N1 species The relative isoabundance plots of DBE versus carbon number for N1 species of the crude oil, NA and BS are shown in Fig. 2. In the crude oil, N1 species with a total RA of 52% showed wide ranges of DBEs (6–25) and carbon numbers (15–50), and were dominated by benzocarbazole-like (DBE = 12) and carbazole-like (DBE = 9) compounds, followed by compounds with DBE values of 10, 11, and 13. The RA of N1 species decreased to 5 and 1% in NA and BS, respectively. As the degree of biodegradation increased, dominant N1 species changed to higher DBE. In BS, compounds with DBE = 15 (represented by dibenzocarbazoles) were the most abundant species, followed by DBE = 12 and 13 species. As shown in Fig. 2, compounds with higher DBE and smaller carbon numbers accumulated after enhanced biodegradation (BS). Few N1 species with carbon numbers >40 were detected in BS. These results showed that N1 species in the crude oil were primarily compounds with multi-fused rings, and that compounds with higher aromacity and less alkylation show more resistance to microbial alteration, suggesting that the degrading enzymes might preferentially attack pyrrolic cores with less aromatic cores and alkyl side chains rather than the rings. These results were consistent with those of a previous study of the biodegradation of crude oil in reservoirs (Kim et al., 2005). Ozonation was effective at removing the residual N1 species in BS, and only trace amounts of N1 species could be detected, even in OZ1 (ozonated for 2 h).

3. Results and discussion 3.1. NSO compositions revealed by FT-ICR MS More than 10 000 peaks (SNR > 5) between m/z 200–700 were detected for each sample. More than 6600 peaks, accounting for up to 93% of the total intensity of all peaks, were assigned a chemical formula. The RAs of NSO classes for different samples are shown in Fig. 1. In the crude oil, N1 species (52%) were the predominant compounds, followed by O2 (16%), O1 (11%), and NO1–NO3 compounds. The RA of N1, O1 species decreased, while O2–O6 and NO1–NO3 species increased with increasing degrees of biodegradation. O2 species (58% and 59% for NA and BS separately) became dominant in the biodegraded samples. Ozonation of BS treated soil further decreased the RA of N1 and O1 species of the residual oil. Ozonation within 6 h decreased the RA of O2 species and increased the RA of O3–O8 species. In OZ2 (6 h of ozonation), O4 species became the dominant species. However, in OZ3 (15 h of ozonation),

Relative abundance (%)

Crude oil

60 50

NA

BS

OZ1

OZ2

OZ3

60

18

50

15

30

12

10

9 20

5

6 10

0

3 0

0 N1

NO1

NO2

NO3

O1

O2

O3

O4

O5

Fig. 1. Compound class distributions in different samples.

O6

O7

O8

612

J. Wang et al. / Chemosphere 85 (2011) 609–615

DBE

Crude oil

NA

BS

25

25

25

20

20

20

15

15

15

10

10

10 5

5

5 10

20

30

40

50

10

20

30

40

50

10

20

30

40

50

Carbon number

Carbon number

Carbon number

Fig. 2. Relative isoabundance plots of DBE versus carbon number for N1 species in the crude oil, NA and BS.

3.3. Changes in oxygen-containing species Biodegradation and chemical oxidation are always involved in the oxygenating process; thus the distributions, types, and abundance of oxygen-containing species would provide evidence of degradation pathways. 3.3.1. O1 species The isoabundance of O1 species as a function of DBE and carbon number of all samples are presented in Fig. 3. O1 species in the crude oil had a broad carbon number range of 14–45 and were dominated by compounds with DBE = 4, which likely corresponded to phenols or aromatic alcohols or sterols. Biodegradation resulted in a decrease in compounds with DBE = 4 and 5. Compounds with DBE = 3–17 showed normal distributions with peaks of DBE = 6 and 7 for NA and BS, respectively, which differed from the distribution in crude oil. Similar to the change trend observed in N1 species, O1 species with higher DBE values and carbon numbers also showed more resistance to biodegradation, and O1 species with carbon numbers >32 were completely removed by BS. Compounds with DBE = 1, which could be monocyclic alcohols, aldehydes or ketones, became more abundant in biodegraded residual oils, probably because of the conversion of hydrocarbons.

Crude oil

DBE

25

3.3.2. O2–O8 species O2 species in crude oils and biodegraded oils are always discussed because O2 species are abundant in oils and the composition of O2 species could be an important indicator of the degree of biodegradation (Kim et al., 2005; Hughey et al., 2007). Distributions of O2 species of different samples are shown in Fig. 4. Compounds with DBE = 1 and carbon number = 16 and 18, which corresponds to C16 and C18 fatty acids, were not considered because they usually act as contaminants during negative ESI analysis (Shi et al., 2010b). In the crude oil, O2 species were dominated by compounds with DBE = 1 and 2. As the degree of biodegradation increased, naphthenic acids with DBE = 3–7 were enriched. Moreover, biodegradation resulted in the appearance of O2 compounds

25

NA

25

20

20

20

15

15

15

10

10

10

5

5

5

0

0 10

20

30

40

20

30

40

25

50 OZ2

10

20

20

15

15

15

10

10

10

5

5

5

0 10

20

30

Carbon number

40

50

20

30

40

25

20

0

BS

0 10

50 OZ1

25

DBE

The residual O1 species in BS were further removed by ozonation, especially for the series of DBE = 5–13. As a result, compounds with DBE = 1 were dominant after ozonation. Furthermore, several O1 compounds with DBE P 16 formed after ozonation for 2 and 6 h. These compounds likely corresponded to ketone-substituted PAHs containing five or more aromatic rings. These compounds detected in OZ1 and OZ2 were removed during longer ozonation (15 h), suggesting that they were non-resistant intermediates of ozonation with oil residuals.

50 OZ3

0 10

20

30

Carbon number

40

50

10

20

30

Carbon number

Fig. 3. Relative isoabundance plots of DBE versus carbon number for O1 species in different samples.

40

50

613

J. Wang et al. / Chemosphere 85 (2011) 609–615

DBE

25

Crude oil

NA

25

20

20

20

15

15

15

10

10

10

5

5

5

0 15

20

25

30

35

25

40

45

10

15

20

25

30

35

25

OZ1

40

10

45

OZ2

20

20

15

15

15

10

10

10

5

5

5

15

20

25

30

35

Carbon number

40

45

20

25

30

35

40

45 OZ3

0

0 10

15

25

20

0

BS

0

0 10

DBE

25

10

15

20

25

30

35

Carbon number

40

45

10

15

20

25

30

35

40

45

Carbon number

Fig. 4. Relative isoabundance plots of DBE versus carbon number for O2 species in different samples. C16 and C18 fatty acids (carbon number = 16 and 18, DBE = 1) were not considered since they usually act as contaminants in the negative-ion ESI analysis.

with carbon numbers >40, which were not detected in the crude oil. Among the naphthenic acids in the BS, the most abundant series was DBE = 5, which likely corresponded to compounds with 4 aliphatic rings (secohopanoic acids). This was followed by the series of DBE = 4 and DBE = 6, which likely corresponded to tricyclic acids and hopanoic acids, respectively. The carbon number distributions of O2 species with DBE = 3, 4, 5, and 6 in BS are shown in Fig. SM-2. The entire series showed a wide range of carbon numbers with 2 or 3 peaks, i.e., C31, C24, and/or C15. Naphthenic acids are considered to originate from microbial alternation of cycloalkanes and sulfur-containing compounds (Hughey et al., 2008). The two- or three-peak carbon number distribution of naphthenic acids in BS was different from the single peak-distributions reported by other studies of the biodegradation of crude oil in reservoirs (Kim et al., 2005; Hughey et al., 2007), which could be attributed to the difference in microbial population involved in the processes. The RA of O2 species in the three ozonated soils also largely decreased when compared with BS. Ozonation decreased the abundance of naphthenic acids with DBE > 3 and fatty acids became the most abundant series. Studies have shown that ozone can remove the naphthenic acids from oil sands process water (OSPW), leading to decreased toxicity and acceleration of the subsequent biodegradation (Scott et al., 2008; Martin et al., 2010). In this study, the effectiveness of ozonation of naphthenic acids in the residual oil in soil was demonstrated. Different from the results shown by Scott et al. (2008), who found that total organic carbon (TOC) in OSPW did not change during 130 min of ozonation, we found a decrease of TSEM (by 68%) and TOC (by 45%) in the soil after 15 h of ozonation. This difference might be attributed to the presence of natural catalysts in the soil promoting mineralization during ozonation. O3–O8 species only accounted for 4% in the crude oil. The RA of O3+ species increased to 14–20% after biodegradation and 59–68% after ozonation, suggesting that both processes would produce a high diversity of oxygen-containing compounds. The isoabundances of O4 species as a function of DBE and carbon number of all the samples are shown in Fig. 5 as an example of how the different treatments changed the highly oxygenated compounds. In the

crude oil, only a few O4 species could be detected, and these showed an irregular distribution pattern. Biodegradation increased the abundance and number of O4 species. Notably, compounds with higher DBE (>10) and higher carbon numbers (>35) appeared in NA and BS, indicating a large amount of highly unsaturated compounds were enriched. Considering the disappearance of O1 species following biodegradation (Fig. 2), the enriched highly oxygenated species with high carbon numbers in NA and BS might be transformed from O1 species under aerobic conditions. When compared with NA, in which the O4 species were dominated by compounds with DBE = 2, the most abundant O4 compounds in BS were at DBE = 4 and 5. Ozonation showed higher reactivity toward compounds with higher DBEs, and this probably occurred through ring cleavage reactions. For example, O4 species with DBE > 14 were completely removed by 6 h of ozonation. The most abundant compounds were observed at DBE = 3 (DBE-O = 1), followed by DBE = 2 (DBE-O = 2) and 4 (DBE-O = 0) in the ozonated samples with a carbon number range of 10–35. These series showed normal-like distributions with carbon numbers that were quite different from the distributions of BS, indicating that these more saturated compounds were formed by ozonation. Similar changing trends were also found for O3 and O5–O8 compounds. The distribution plots of O3, O5 and O6 compounds can be found in Figs. SM-3–SM-5. These results showed that biodegradation would result in the accumulation of highly oxygenated compounds with aromatic/ unsaturated structures. Ozonation could react with a variety of NSO polar compounds of the residual oil in biodegraded soil and showed high reactivity toward these unsaturated compounds and produced more saturated highly oxygenated compounds, which was also clearly showed on the van Krevelen diagrams (Fig. SM6). O2–O8 species were the major products of ozonation, and all were dominated by compounds with DBE-O = 1. Although some of these highly oxygenated products of ozonation would be further decomposed by longer ozonation/higher ozone dosage, the removal would be slow. Furthermore, detailed analysis of O2–O8 species may allow us to develop diagnostic ratios for environmental forensics. An index based on the ratio of acyclic to cyclic naphthenic acids was used to estimate the degree of biodegradation

614

J. Wang et al. / Chemosphere 85 (2011) 609–615

DBE

20

20

Crude oil

15

15

15

10

10

10

5

5

5

0

0 10

15

20

25

30

35

20

DBE

20

NA

40

45

0 10

15

20

25

30

35

20

OZ1

40

10

45

15

15

10

10

10

5

5

5

0 10

15

20

25

30

35

Carbon number

40

45

15

20

25

30

35

40

20

OZ2

15

0

BS

45

OZ3

0 10

15

20

25

30

35

40

45

10

15

20

25

30

35

40

45

Carbon number

Carbon number

Fig. 5. Relative isoabundance plots of DBE versus carbon number for O4 species in different samples.

-0.02 -0.01 100 Crude oil

0.00

0.01

0.02 0.02

(a)

0.01

50

PC 2 (25.7%)

(Kim et al., 2005). The results of our study suggest that highly oxygenated species such as O4 species may also have the potential for use in environmental forensics because the abundance and distributions of these species were quite different under different treatment, allowing development of ratios to distinguish degrees of biodegradation as well as biological processes versus chemical processes. 3.4. Comprehensive analysis of NSO compounds revealed by FT-ICR MS

OZ2 OZ3 OZ1 0.00

0

-0.01

-50

NA BS

-100

-0.02

-100

-50

N1

0

O1

50

O2

-0.02 -0.01 100 Crude oil

O3

0.00

O4

100

O5+

0.01

0.02 0.02

(b)

50

PC 2 (25.7%)

Multivariate statistical approaches would provide a comprehensive interpretation and overview of the FT-ICR MS data (Hur et al., 2009; Sleighter et al., 2010). In this study, principal component analysis (PCA) was conducted using all assigned elemental formulas of the six samples. Biplots of the score and loading parameters according to the PCA results were constructed to show the compositional differences among samples, as shown in Fig. 6a, in which the score plot shows the distribution of the samples, and the loading plot presents the contribution of the variables to the variability of the samples. Based on the score plot, the samples could be separated into three groups by the PC coordinates. Group I contained the crude oil, group II contained TSEMs of NA and BS, and group III contained the three ozonated samples. As a strong oxidant, ozone changed the TSEM of BS dramatically in a short time (2 h). However, after longer time of ozonation, the polar compounds did not differ much from those of samples that were ozonated for 2 h, as shown in Fig. 6a. These findings suggest that ozonation turned into a slow phase after a rapid oxidation phase, which is similar to the results of studies conducted to evaluate the removal of total petroleum hydrocarbons or survival of microorganisms in soils (Ahn et al., 2005; Jung et al., 2005). The compound classes that regulated the distribution of the samples could be recognized by the loading plot (Fig. 6a). The crude oil was rich in N1 and O1 species, while NA and BS were rich in O2–O4 species and OZ1–OZ3 were rich in O3 and O4 species as well as O5+ (O5–O8) species. The loading plot of the single class compounds obtained by hiding other classes and adding structural information (DBE values) provided a closer look into the specific loading variables. For example, the biplots of the score and loading

OZ2

0.01

OZ3 0

OZ1

-50

0.00

-0.01

BS NA

-100 -100

-0.02 -50

0

50

100

PC 1 (34.3%) DBE 0

6

12

18

24

Fig. 6. Biplots of the scores and loadings for PCA of the six samples. The scores were obtained from PCA results using all of the compounds indentified by FT-ICR MS. The bottom and left axes show the scores and the top and right axes show the loadings. The loadings are displayed for (a) major compound classes, and (b) O4 compounds. O5+ refers to O5–O8 species.

of O4 compounds are shown in Fig. 6b. Although the TSEMs of soils treated by biodegradation and biodegradation + ozonation both contained abundant O4 compounds, the figure showed that TSEMs

J. Wang et al. / Chemosphere 85 (2011) 609–615

following ozonation were rich in compounds with less DBE. These results obtained by PCA were quite consistent with the detailed analysis shown above, suggesting that PCA could not only be used to group samples, but also to obtain general compositional information regarding the samples. 4. Conclusions FT-ICR MS is a powerful tool to reveal the changes in polar compounds of residual oil by sequenced bioremediation and ozonation in contaminated soil. Crude oil and residual oil both showed a high diversity of polar compounds, with more than 6600 polar compounds being identified in each sample. N1 and O1 species, which are abundant in crude oil, were selectively removed by biodegradation and species with higher DBE and lower carbon numbers were more resistant to biodegradation. Oxygenated compounds (O2–O6 species) increased after biodegradation and were dominated by compounds with high DBE. Ozonation effectively removed the residual N1 species and oxygen-containing compounds with unsaturated structures in the residual oil that were suspected to be resistant to microbial alternation in the biodegraded soil. Fatty acids and O3–O8 species dominated by compounds with lower DBE (high saturate carbon skeleton) were the main products of ozonation. PCA revealed a comprehensive overview of the complex samples at the molecular level, which was quite consistent with the detailed analysis. These results may provide a better understanding of the use of the two remediation techniques (bioremediation and ozonation) for treatment of crude oil contaminated soil. Both biodegradation and ozonation caused increased oxygen content in the residual oil, but these changes occurred through different pathways and resulted in different compositions. Therefore diagnostic ratios developed based on specific species (for example O4 species) may enable identification of treatments that the soil undergoes. Moreover, further study may be needed to evaluate the potential impacts of the highly oxygenated compounds generated by biodegradation and/or ozonation on the soil environment. Acknowledgments This study was supported by the National Natural Science Foundation of China (Nos. 40672162 and 40730738), Tsinghua University Initiative Scientific Research Program (No. 2009THZ02238), and the National Key Technology R&D Program of China (No. 2007BAC16B06). The authors thank Dongjuan Dai, Yan Zhou, and Shan Mu for their assistance in sample preparation. The FT-ICR MS data were collected at the Analysis Center, State Key Laboratory of Heavy Oil Processing, China University of Petroleum with the help of Dr. Quan Shi and Dr. Yahe Zhang. Appendix A. Supplementary material Supplementary data associated with this article can be found, in the online version, at doi:10.1016/j.chemosphere.2011.06.103. References Ahn, Y., Jung, H., Tatavarty, R., Choi, H., Yang, J.W., Kim, I.S., 2005. Monitoring of petroleum hydrocarbon degradative potential of indigenous microorganisms in ozonated soil. Biodegradation 16, 45–56.

615

Clemente, J.S., Fedorak, P.M., 2005. A review of the occurrence, analyses, toxicity, and biodegradation of naphthenic acids. Chemosphere 60, 585–600. Huang, H., Bowler, B.F.J., Zhang, Z., Oldenburg, T.B.P., Larter, S.R., 2003. Influence of biodegradation on carbazole and benzocarbazole distributions in oil columns from the Liaohe basin, NE China. Org. Geochem. 34, 951–969. Hughey, C.A., Rodgers, R.P., Marshall, A.G., Qian, K., Robbins, W.K., 2002. Identification of acidic NSO compounds in crude oils of different geochemical origins by negative ion electrospray Fourier transform ion cyclotron resonance mass spectrometry. Org. Geochem. 33, 743–759. Hughey, C.A., Galasso, S.A., Zumberge, J.E., 2007. Detailed compositional comparison of acidic NSO compounds in biodegraded reservoir and surface crude oils by negative ion electrospray Fourier transform ion cyclotron resonance mass spectrometry. Fuel 86, 758–768. Hughey, C.A., Minardi, C.S., Galasso-Roth, S.A., Paspalof, G.B., Mapolelo, M.M., Rodgers, R.P., Marshall, A.G., Ruderman, D.L., 2008. Naphthenic acids as indicators of crude oil biodegradation in soil, based on semi-quantitative electrospray ionization Fourier transform ion cyclotron resonance mass spectrometry. Rapid Commun. Mass Spectrom. 22, 3968–3976. Hur, M., Yeo, I., Park, E., Kim, Y.H., Yoo, J., Kim, E., No, M.-H., Koh, J., Kim, S., 2009. Combination of statistical methods and Fourier transform ion cyclotron resonance mass spectrometry for more comprehensive, molecular-level interpretations of petroleum samples. Anal. Chem. 82, 211–218. Jung, H., Ahn, Y., Choi, H., Kim, I.S., 2005. Effects of in-situ ozonation on indigenous microorganisms in diesel contaminated soil: survival and regrowth. Chemosphere 61, 923–932. Kim, S., Stanford, L.A., Rodgers, R.P., Marshall, A.G., Walters, C.C., Qian, K., Wenger, L.M., Mankiewicz, P., 2005. Microbial alteration of the acidic and neutral polar NSO compounds revealed by Fourier transform ion cyclotron resonance mass spectrometry. Org. Geochem. 36, 1117–1134. Layshock, J.A., Wilson, G., Anderson, K.A., 2010. Ketone and quinone-substituted polycyclic aromatic hydrocarbons in mussel tissue, sediment, urban dust, and diesel particulate matrices. Environ. Toxicol. Chem. 29, 2450–2460. Liang, Y., Van Nostrand, J.D., Wang, J., Zhang, X., Zhou, J., Li, G., 2009. Microarraybased functional gene analysis of soil microbial communities during ozonation and biodegradation of crude oil. Chemosphere 75, 193–199. Lu, M., Zhang, Z., Qiao, W., Guan, Y., Xiao, M., Peng, C., 2010. Removal of residual contaminants in petroleum-contaminated soil by Fenton-like oxidation. J. Hazard. Mater. 179, 604–611. Lundstedt, S., White, P.A., Lemieux, C.L., Lynes, K.D., Lambert, I.B., Öberg, L., Haglund, P., Tysklind, M., 2007. Sources, fate, and toxic hazards of oxygenated polycyclic aromatic hydrocarbons (PAHs) at PAH-contaminated sites. Ambio 36, 475–485. Marshall, A.G., Rodgers, R.P., 2003. Petroleomics: the next grand challenge for chemical analysis. Accounts Chem. Res. 37, 53–59. Marshall, A.G., Rodgers, R.P., 2008. Petroleomics: chemistry of the underworld. Proc. Natl. Acad. Sci. USA 105, 18090–18095. Martin, J.W., Barri, T., Han, X., Fedorak, P.M., El-Din, M.G., Perez, L., Scott, A.C., Jiang, J.T., 2010. Ozonation of oil sands process-affected water accelerates microbial bioremediation. Environ. Sci. Technol. 44, 8350–8356. Rushdi, A.I., DouAbul, A.A., Mohammed, S.S., Simoneit, B.R.T., 2006. Compositions and sources of extractable organic matter in Mesopotamian marshland surface sediments of Iraq: II. Polar compounds. Environ. Geol. 50, 1171–1181. Scott, A.C., Zubot, W., MacKinnon, M.D., Smith, D.W., Fedorak, P.M., 2008. Ozonation of oil sands process water removes naphthenic acids and toxicity. Chemosphere 71, 156–160. Shi, Q., Hou, D., Chung, K.H., Xu, C., Zhao, S., Zhang, Y., 2010a. Characterization of heteroatom compounds in a crude oil and its saturates, aromatics, resins, and asphaltenes (SARA) and non-basic nitrogen fractions analyzed by negative-ion electrospray ionization Fourier transform ion cyclotron resonance mass spectrometry. Energy Fuel 24, 2545–2553. Shi, Q., Zhao, S., Xu, Z., Chung, K.H., Zhang, Y., Xu, C., 2010b. Distribution of acids and neutral nitrogen compounds in a Chinese crude oil and its fractions: characterized by negative-ion electrospray ionization Fourier transform ion cyclotron resonance mass spectrometry. Energy Fuel 24, 4005–4011. Skoczynnska, E., Korytaar, P., de Boer, J., 2008. Maximizing chromatographic information from environmental extracts by GCXGC-ToF-MS. Environ. Sci. Technol. 42, 6611–6618. Sleighter, R.L., Liu, Z., Xue, J., Hatcher, P.G., 2010. Multivariate statistical approaches for the characterization of dissolved organic matter analyzed by ultrahigh resolution mass spectrometry. Environ. Sci. Technol. 44, 7576–7582. Wang, Z., Fingas, M., 1997. Developments in the analysis of petroleum hydrocarbons in oils, petroleum products and oil-spill-related environmental samples by gas chromatography. J. Chromatogr. A 774, 51–78. Whitby, C., 2010. Microbial naphthenic acid degradation. In: Allen, I.L., Sima, S., Geoffrey, M.G. (Eds.), Advances in Applied Microbiology. Academic Press, San Diego, pp. 93–125.