Quantification of synthetic-based drilling mud olefins in crude oil and oiled sediment by liquid column silver nitrate and gas chromatography

Abstract Synthetic-based drilling muds (SBMs) are complex mixtures of man-made fluids used during the drilling of oil and gas wells. SBM-derived chemicals can enter the environment through failed wells and routine or poor disposal practices, where they can persist and thereby warrant measurement. SBMs are commonly formulated with linear and methyl branched α- and internal-olefins mostly in the C14 to C20 carbon range, which are not native to crude oils. Thus, SBM-derived olefins can provide a basis to recognize the impact of these drilling wastes in the environment. However, the presence of abundant native hydrocarbons in oils and sediments can hinder the detection of trace level SBM-derived olefins by conventional organic sample preparation and analytical methods. Silver ion chromatography using silver nitrate (AgNO3) impregnated silica gel can serve to physically separate olefins from saturated aliphatic hydrocarbons native in crude oils, which can subsequently be analyzed and measured by conventional one-dimensional gas chromatography-flame ionization detection (GC-FID). In this study, SBM-derived olefins are measured in crude oils from the Deepwater Horizon oil spill and in their laboratory mixtures to a detection limit of approximately 5000 µg/g (0.5 wt%). In oiled sediment, SBM-derived olefins were reliably detected at concentrations as low as 1 µg/g-dry. An application of this method is demonstrated through analysis of crude oils and oil-contaminated seafloor sediment from the Taylor Energy oil spill site in the northern Gulf of Mexico where SBM was historically used and discharged.


Introduction
Synthetic-based drilling muds (SBMs) are man-made fluids commonly used in rotary drilling of offshore oil and gas wells to provide lubricity, clean and condition the hole, and counterbalance formation pressure (Neff et al., 2000).SBMs are generally comprised of 30 to 90% volume (20-50 wt%) of synthetic (man-made) organic compounds, which act as lubricants, that are dispersed in a salt brine to form an emulsion, along with other ingredients including emulsifiers, barite, clays, lignite, or lime.The most common synthetic organic compounds in SBMs are monounsaturated acyclic hydrocarbons, commonly referred to as olefins (C n H 2n ).
The olefins in SBMs can occur in a variety of forms, viz., linear a-olefins (LAOs), internal olefins (IOs) or, less commonly, poly a-olefins (PAOs).LAOs are formed by the catalytic oligomerization of ethylene (C 2 H 4 ) to produce olefins containing a double bond in the a-position and, owing to the ethylene monomer, are dominated by compounds containing an even number of carbons mostly between 14 (C 14 H 28 ) and 20 (C 20 H 40 ) (American Chemistry Council, 2006).IOs are formed by the catalytic isomerization of LAOs that shifts the double bond from the a-position to various internal positions along the carbon chain, thereby forming numerous IO isomers.Additionally, some mostly methyl-branched olefin isomers (MBOs) are formed as by-products of LAO and IO synthesis (American Chemistry Council, 2006).Following blending and distillation to produce the desired physical properties most commercial SBMs contain a complex mixture of LAO and IO isomers, with smaller amounts of MBOs, typically dominated by C 16 and C 18 olefin isomers.
Although not widely studied, olefins are found naturally in many crude oils (Curiale and Frolov, 1998).Naturally occurring olefins in crude oils are referred to as native olefins herein, in order to distinguish them from SBM-derived synthetic olefins.Existing studies that utilized thin-layer chromatography (TLC) indicate that the concentrations of native olefins in crude oils range from less than 3000 mg/g (0.3 wt%) to more than 100,000 mg/g (10 wt%; Frolov and Smirnov, 1994;Frolov, 1995;Frolov et al., 1996) wherein their presence is predominantly attributed to radiolytic dehydrogenation of saturated hydrocarbons over geologic time (Frolov and Smirnov, 1994;Frolov et al., 1998).These same studies have shown that molecular distributions of native olefins in crude oil mimic those of the native saturated hydrocarbons, and thus are distinct from the concentrated "clusters" of synthetic, even-carbon olefin isomers found in SBM.SBMderived olefins can contaminate crude oil samples collected from newly drilled wells and alter its reservoir fluid properties, which are critical metrics in assessing the well's overall economics and reservoir development (MacMillan et al., 1997).Therefore, recognizing and quantifying SBM-derived olefins in crude oils obtained from newly drilled exploration or development wells is often necessary.
In addition, SBMs' use has increased owing to their technical benefits in the drilling of difficult deepwater, deviated or horizontal wells, coupled with the economic advantages of discharging SBM-impacted rock cuttings at the drill site (Neff, 2000).Although SBM is not typically discharged in bulk to the sea, drill cuttings containing traces of adhered SBM are permitted to be discharged during active drilling.Thus, recognizing the presence of SBM-derived olefins in marine sediments and measuring their concentration is also important for environmental impact studies of offshore drilling activities.Multiple seafloor studies have assessed the impact of these routine discharges on marine sediments, which have collectively shown that high concentrations (>10,000 mg/g-dry) of SBMderived olefins mostly occurred within an area less than 0.2 km 2 around the drill site, but sometimes can spread further depending upon ocean depth and currents (Neff et al., 2000(Neff et al., , 2005;;Melton et al., 2003;Continental Shelf Associates, 2004a, 2004b).
Catastrophic discharges of large volumes of SBM also can occur during well blowouts, such as had occurred during the Deepwater Horizon disaster when more than 2200 barrels (bbl) of SBM present within the failed Macondo well and drill string was instantly lost to the deep sea (BP, 2010).Multiple unsuccessful attempts to plug ("kill") the leaking Macondo well employed nearly 30,000 bbl of SBM, some unknown volume of which was also lost into the deep sea (DOE, 2014).Subsequent study of the region's seafloor showed sediments within $6.5 km 2 area around the failed well contained mostly between 1000 and 10,000 mg/g-dry of SBM-derived olefins (avg.3270 mg/ g-dry; max 28,654 mg/g-dry; Stout and Payne, 2017).
Recognizing the presence and measuring the concentrations of SBM-derived olefins in crude oils or marine sediments is conventionally achieved using one-dimensional gas chromatography with flame ionization detection (GC-FID) of whole oils or sediment extracts (MacMillan et al., 1997;Gonzalpour et al., 1999;Stout and Payne, 2017).In this approach the GC-FID chromatograms of oils or hydrocarbons extracted from sediments are qualitatively assessed for the presence of "clusters" of even-carbon olefin isomers characteristic of SBM, which if present can be integrated and quantified separately from any native (non-SBM) hydrocarbons in crude oil or sediment samples.Owing to the comparable GC-FID response factors of SBM-derived olefins and the other hydrocarbons in crude oil this method allows for the calculation of SBM-derived olefins' absolute concentrations in oil or sediment through a conventional calibration used to measure total petroleum hydrocarbons or individual alkanes (e.g., US EPA Method 8015D).At high(er) relative concentrations of SBM-derived olefins, wherein their presence is obvious based upon prominent olefin "clusters" visible within the chromatogram and readily integrated separately from native hydrocarbons, conventional one-dimensional GC-FID can satisfactorily establish the relative and absolute concentrations of SBM-derived olefins in an oil or sediment sample since any contribution of coeluting native hydrocarbons is minimized (Stout and Payne, 2017).(An example of a pure SBM 1-dimensional chromatogram is shown in Figure S1.)However, at low(er) relative concentrations of SBMderived olefins, olefin "clusters" may not be visually evident within an oil or sediment sample's chromatogram, and their presence may go unnoticed or, if evident, their integration be confounded by the coelution with the native hydrocarbons present.
Taking advantage of polarity differences, comprehensive two-dimensional gas chromatography with flame ionization detection (GC x GC-FID) is capable of separating SBM-derived olefins from native hydrocarbons in oil and thereby avoid the confounding issues of low(er) relative olefins concentration or coelution (Reddy et al., 2007;Aeppli et al., 2013).However, unlike conventional one-dimensional GC-FID (e.g., US EPA Method 8015D), GC x GC-FID is not yet a validated method within US EPA SW-846 and requires both time-consuming and complex data interpretation.Consequently, GC x GC is often prohibitively expensive or logistically impractical for use in commercial environmental laboratories.
Alternatively, confounding issues of one-dimensional GC-FID can be mitigated by utilizing a conventional silver nitrate (AgNO 3 ) liquid chromatography for the separation and purification of olefins and other unsaturated compounds (Mander and Williams, 2016).Silica-gel impregnated with AgNO 3 is an effective stationary phase absorbent in liquid column chromatography owing to the silver acting as a p-acceptor to the carbon-carbon double bond p-donor within alkenes (Lienne et al., 1987).As such, liquid column AgNO 3 chromatography, which requires only a modest modification to routine liquid column silica-gel cleanup (US EPA Method 3630C), provides a reliable and practical means to physically separate SBM-derived olefins from any native hydrocarbons in a crude oil and sediment.Following liquid column AgNO 3 cleanup, conventional (and widely commercially-available) one-dimensional GC-FID (e.g., US EPA Method 8015D) can then be used to recognize and quantify SBM-derived olefins in oil or sediment samples, even at low(er) relative and absolute concentrations.Reddy et al. (2007) demonstrated the potential for analyzing low concentrations of SBM-derived olefins in four crude oils using AgNO 3 liquid column chromatography, where a high correlation (r 2 ¼0.98) between the relative concentrations (wt% of total) of SBM-derived olefins determined by GC x GC-FID of whole oils and by GC-FID of AgNO 3 -purified olefin fractions was obtained.This study (Reddy et al., 2007) did not employ internal or external standards and thereby did not report absolute concentrations, only relative concentrations of SBMderived olefins by GC-FID and GC x GC-FID.
In the present study, we have used commerciallyavailable AgNO 3 impregnated silica-gel to analyze crude oil blended in the laboratory with varying but low levels of SBM in order to demonstrate the efficacy of the column cleanup method and report absolute concentrations of SBM-derived olefins in crude oil at low(er) concentrations using conventional, one-dimensional GC-FID.We further demonstrate application of the method in a small suite of environmental samplessix crude oils and one oiled seafloor sedimentcollected from an oil spill site in the northern Gulf of Mexico where SBM was historically used/discharged to the seafloor.

Samples
Fresh crude oils Three fresh crude oil samples containing varying amounts of SBM obtained during the response efforts following the Deepwater Horizon disaster were evaluated herein.Fresh crude oil free of SBM was collected on May 21, 2010 on the Discoverer Enterprise drillship from the riser insertion tube, which was receiving oil directly from the rig's broken riser tube near the sea floor.Detailed chemical characterization of this fresh oil is available elsewhere (Stout et al., 2016).The two additional samples of fresh crude oil were collected August 14, 2010 from storage tanks on the Q4000 recovery ship that had been used during the well kill efforts.These two samples contained what is qualitatively described, based upon GC-FID chromatograms, as containing "low" and "high" relative abundances of SBM (quantified herein; see below).Additional details regarding the crude oil samples are found in Supporting Information (Table S1).

Laboratory mixtures
All three fresh crude oils were evaluated individually.In addition, the samples containing no SBM and a low relative abundance of SBM were mixed in varying proportions (mass basis) that increasingly diluted the amount of SBM in the mixed oils.Specifically, aliquots of the fresh crude oil containing a low relative abundance of SBM was mixed on a mass basis with fresh oil containing no SBM in proportions of 20:80, 10:90, and 5:95, referred to herein as Mixture A, Mixture B, and Mixture C, respectively.Analysis of these three laboratory mixtures provided a basis upon which to establish a de facto detection limit of 5000 mg/g (0.5 wt%) for the presence of SBM-derived olefins in crude oil (see below).Additional details regarding the crude oil mixtures are found in Supporting Information (Table S1).

Environmental samples
Six crude oil samples and one sediment sample collected from an oil spill site in the northern Gulf of Mexico were studied.Specifically, the Taylor Energy Company site located in Mississippi Canyon Block 20 (MC20) has been under investigation since the collapse of an oil production platform, and rupture of (up to) 28 production wells, following Hurricane Ivan in September 2004 (Mason et al., 2019;Bryant et al., 2020).Despite the completion of nine intervention wells, crude oil continues to leak from the seafloor, predominantly from a location above the platform's buried ($50 m) conductor bundle terminus.Prior to installation of a subsea oil collection/containment system in April 2019, which now captures most of the leaking oil, persistent oil sheens/slicks occurred at the area's sea surface (e.g., Sun et al., 2018).Seafloor sediments in the area are variably contaminated with oil, and in some locations SBM derived (at least in part) from discharges during the drilling of the intervention wells (Bryant et al., 2020).The presence/absence of SBM in the oil actively exiting the MC20 site's seafloor (and in sediments) is relevant to the source(s) of the oil, i.e., actively leaking wells (no SBM) versus residual oil and SBM in sediments dating back to the original accident and intervention well drilling (Bryant et al., 2020).
Three oils were collected at the Site in September 2018 (i.e., prior to installation of the subsea containment system) by remote operated vehicle (ROV) as they rose through the water column (Mason et al., 2019).These samples were collected separately on ROV dives over three days and are referred to herein as Water Column A, B, and C. Three more oils were collected in late April 2019 during offloading of the oil captured within the subsea containment systemabout three weeks after its installation.Finally, the sediment sample studied was obtained from the 0-10 cm portion of box core collected in September 2018 near the location where most oil continues to leak from the seafloor.Additional details regarding the environmental samples are found in Supporting Information (Table S1).

Column chromatography
Oil and sediment samples first were homogenized by sonication or mixing.The oil samples ($0.10 g) were diluted in 10 mL in analytical grade dichloromethane (DCM), concentrated to 1 mL and spiked with surrogate internal standards (SIS), o-terphenyl and n-tetracosane-d 50 .The sediment sample was spiked with SIS and dried with anhydrous Na 2 SO 4 and seriallyextracted (3-times; 6, 2, and 0.5 hrs) using 100 mL analytical grade DCM on a shaker table.The three extracts for the sediment sample were combined, filtered through glass wool, dried with anhydrous Na 2 SO 4 , and concentrated to 1 mL using Kuderna Danish apparatus and nitrogen blow-down and treated with activated copper to remove elemental sulphur.
The oil and sediment extracts were solvent exchanged into analytical grade n-hexane, concentrated to 4 mL (as above), and subsequently processed via the column chromatography scheme depicted in Figure 1 with the objective of obtaining the olefins (alkenes) fraction (F12) of each sample.Briefly, this was achieved by charging $4 mL of the solvent exchanged oil or sediment extract onto a glass column (30.5 Â 10.5 mm ID) containing $5 g of fully-activated silica gel (CAS 112926-00-8) with 100-200 mesh, Davisil grade 923 (Sigma-Aldrich; Cat.No. 214477).The sample's aliphatic fraction (F1) was collected first by eluting the column with 18 mL of analytical grade n-pentane, and subsequently with 25 mL of DCM to elute the sample's aromatic fraction (F2; Figure 1).Each sample's F1 fraction was concentrated to 1 mL and split, reserving one split for GC-FID analysis.The second F1 split was loaded onto a glass column (30.5 Â 10.5 mm ID) charged with $5 g of fully-activated silver nitrate (AgNO 3 ) impregnated silica gel (CAS 7761-88-8) containing $10 wt% AgNO 3 on þ230 mesh (Sigma-Aldrich; Cat.No. 248762).Alkanes were first eluted with 18 mL of analytical grade n-pentane (F11) followed by the alkenes eluted with 25 mL of analytical grade DCM (F12).The F1, F11, and F12 fractions were concentrated to 1.0, 1.0, and 0.25 mL, respectively (as above) and spiked with recovery internal standards (RIS; 5a-androstane) prior to GC-FID analysis.

Gas chromatography: flame ionization detection
GC-FID analysis of each sample's whole (unfractionated) oil/extract and corresponding F12 fractions (Figure 1) was performed via EPA Method 8015D using an Agilent 6890 gas chromatograph equipped with a Restek Rtx-5 (60 m Â 0.25 mm ID, 0.25 mm film) fused silica capillary column.Samples were injected (1 mL, pulsed splitless) into the GC programmed from 40 C (1 min) and ramped at 6 C/min to 315 C (30 min) using H 2 ($1 mL/min) as the carrier gas.The F1 and F11 fractions of the two crude oils containing relatively low and high abundances of SBM (only) were also analysed by GC-FID to monitor the effectiveness of the separation scheme.The F2 fractions were not analyzed as part of this study.
GC-FID analysis of each sample's whole oil/extract and F12 fraction was used to determine the total concentration of GC-amenable hydrocarbons eluting between n-C 9 and n-C 44 (exclusive of internal standards and after blank subtraction).Whole oil/extract concentrations are reported as total petroleum hydrocarbons (TPH) and F12 concentrations are reported as TPH F12 .In addition, the concentration of SBMderived olefins was determined through integration of olefin-clusters within the n-C 15 to n-C 20 range of the F12 chromatograms and reported separately as TPH SBM .TPH SBM was determined through integration only in those samples whose F12 chromatograms showed the presence of olefin-clusters characteristic of SBM.All TPH, TPH F12 , and TPH SBM concentrations measured in the oil and sediment samples are reported in mg/g and mg/g-dry, respectively.
Absolute concentrations of TPH, TPH F12 , and TPH SBM, as well as recovery of the surrogate internal standard, n-tetracosane-d 50 , in the oil and sediment samples were calculated by the method of internal standards using the recovery internal standard, 5aandrostane.Prior to sample analysis a five-point calibration was performed to demonstrate the linear range of the analysis.The calibration solution was composed of selected alkanes within the n-C 9 to n-C 40 range.Analyte concentrations in the standard solutions ranged from 1 to 200 ng/mL.All calibration solution compounds were used to generate an average RF (0.960) for TPH, TPH F12 , and TPH SBM calculations.Use of an RF derived from alkanes in the calibration solution for quantification of the alkenes in TPH 12 and TPH SBM is acceptable given these compound groups' comparable responses on GC-FID (e.g., Aeppli et al., 2013).Calibration check standards representative of the mid-level of the initial calibration and the instrument blank were analysed every 10 samples.The check standard's response was compared versus the average RF of the respective analytes contained in the initial calibration.All authentic samples and quality control samples, the latter of which included oil and sediment method (solvent) blanks, were bracketed by passing mid-check standards.

Results and discussion
The concentrations of TPH, TPH F12 , and TPH SBM obtained for the three (3) fresh crude oils, three (3) laboratory mixtures, and seven (7) environmental samples studied are given in Table 1.

Fresh crude oils
GC-FID chromatograms for the three fresh crude oils studied are shown in Figure 2. TPH concentrations in these oils ranged from 629,000 to 719,000 mg/g (Table 1), with the balance(s) of their masses associated with hydrocarbons outside of the n-C 9 and n-C 44 range or non-chromatographable material.These whole oil chromatograms demonstrate both the ability and the limitation of 1-dimensinoal GC-FID in recognizing the presence of SBM-derived olefins in fresh crude oils, which depends strongly upon the abundance of SBM-derived olefins present within the oil.The chromatogram for the fresh oil containing a high relative abundance of SBM-derived olefins exhibits two obvious clusters of peaks typical of SBM-derived olefins that can be easily recognized among the native hydrocarbons in the oil, which in this case are dominated by a homologous series of resolved n-alkanes (Figure 2A).Owing to SBM-derived olefins' production via ethylene (C 2 H 4 ) polymerization these olefin clusters evident are overwhelmingly comprised of even-carbon C 16 and C 18 olefins and contain little to no C 15 , C 17 , or C 19 olefins.Any C 14 or C 20 olefin clusters are not obvious, which likely reflects their removal during distillation of the specific blend of SBM present.The resolved and partially resolved peaks within each cluster correspond to an individual a-olefin (a) and numerous internal olefin (IO) and methyl-branched olefin (MB) isomers (Figure 2A, inset), as is typical of olefin-based SBMs (Neff et al., 2000).Assuming the presence of all possible molecular configurations, there are seven IO and seven MB isomers possible in the C 16 cluster and eight IO and eight MB isomers possible in the C 18 cluster.Thus, including the a-olefins there are 15 and 17 olefin isomers possible within the C 16 and C 18 clusters, respectively (Figure 2A, inset).Based upon GC-FID retention time comparison for selected linear a-olefin standards (1-hexadecene and 1-octadecene) the linear a-olefin in each cluster is identified as eluting just ahead of the corresponding carbon numbered n-alkanes (Figure 2B, inset).Other peak identities are based upon previous studies (Aeppli et al. 2013) in which GC x GC time of flight mass spectrometry (TOF/MS) confirmed the elution/ identity of the linear a-olefins and multiple IO and MB isomers within each cluster.Three prominent but unidentified IO isomers are labelled a, b and c herein (Figure 2B inset).See Table S2 for an inventory of all peak labels used in figures herein.
The presence of SBM-derived olefins in the chromatogram of the crude oil containing only a low relative abundance of SBM is still evident (Figure 2B), although clearly less obvious than in the crude oil containing a high relative abundance (Figure 2A).Owing to the low relative abundance of SBM, the C 16 and C 18 olefin clusters each now clearly include prominent n-alkane peaks (n-C 16 and n-C 18 ) and a phytane peak within the C 18 cluster (Figure 2B, inset) that are native to the fresh oil (Figure 2C).This co-elution of native hydrocarbons in crude oil with SBM-derived olefins exemplifies the difficulty of quantifying the latter via GC-FID of the whole (unfractionated) oils.As expected, no SBM-derived olefin clusters are visible in the fresh crude oil containing no SBM (Figure 2C).Considering these three whole oils' chromatograms (Figure 2) it is not difficult to imagine that a very low relative abundance of SBM in a crude oil may go unnoticed, or at least its presence be equivocal, if only GC-FID chromatograms of whole oils are available.
This problem can be overcome through fractionation using liquid column silver ion chromatography (Figure 1).For example, Figure 3 shows the GC-FID chromatograms for the F1 (aliphatic), F11 (alkane), and F12 (alkene) fractions obtained from the crude oil containing a low relative abundance of SBM.The F1 fraction (Figure 3A) clearly contains the prominent C 16 nd C 18 SBM-derived olefin clusters that were present in the whole (unfractionated) crude oil, as well as the prominent and native n-alkanes and acyclic isoprenoids (e.g., pristane and phytane; Figure 2B).Chromatograms of the F11 and F12 fractions, however, demonstrate the effective separation of the native n-alkanes and acyclic isoprenoids (and other alkanes) into the F11 fraction (Figure 3B) and SBM-derived olefins into the F12 fraction (Figure 3C) that is achieved by the liquid column AgNO 3 impregnated silica gel fractionation step (Figure 1).In fact, only after isolation of the F12 fraction are the presence of small C 20 peaks corresponding to the SBM-derived olefin cluster in this sample evident (Figure 3C,inset).The low relative abundance of SBM-derived olefins observed qualitatively in this whole oil's GC-FID chromatogram (Figure 2B) could be quantified through integration of the olefin clusters in the F12 chromatogram (Figure 3C), which determined that TPH SBM was 36,900 mg/g (Table 1), or approximately 3.7 wt% of the whole oil and 5.9 wt% of the TPH.Similarly, the fresh crude oil containing a high relative abundance of SBM-derived olefins (Figure 2A) had a TPH SBM of 195,000 mg/g (Table 1), or approximately 19.5 wt% of the whole oil and 27 wt% of the TPH.As anticipated, there were no SBM-derived olefins detectable in the F12 fraction of the fresh crude oil containing no SBM (Table 1).
An interesting consequence of the F12 fractions' analyses is revealed in the concentrations of native olefins in each of the fresh oils, which can be estimated by the difference between TPH F12 (total olefin concentrations) and TPH SBM (SBM-derived olefin   S2 for peak label identities.S2 for peak label identities. concentration; Table 1).Specifically, the native olefin concentration in all three fresh crude oils studied, each representing the same Macondo crude oil from the Deepwater Horizon disaster, narrowly ranged from 13,300 to 14,000 mg/g, or approximately 1.3 to 1.4 wt% of the oil (Table 1).This concentration range is comparable to that reported via TLC for other oils worldwide (Frolov and Smirnov, 1994), which tends to further substantiate the effectiveness of the AgNO 3 chromatography approach described herein.The F12 chromatograms reveal that these native olefins overwhelmingly occur within a small but broad unresolved complex mixture (UCM) in the F12 chromatogram (not visible in Figure 3C).There is some caution warranted in assuming the TPH F12 fraction contains purely olefins (alkenes) as it is possible that some monoaromatic hydrocarbons retained in the F1 fraction may subsequently breakthrough into the F12 fraction (Figure 1).

Laboratory mixtures
As noted above, the concentration of SBM-derived olefins in the fresh crude oil sample containing relatively low abundance of SBM-derived olefins, as judged from its gas chromatogram (Figure 2B), were subsequently fractionated and quantified within its F12 fraction to be 36,900 mg/g (Table 1), or about 3.7 wt% of the whole oil.This sample was mixed with higher amounts of the fresh crude oil containing no SBM-derived olefins (80:20, 90:10, and 95:05; mass basis) in order to dilute the SBM-derived olefins in the resulting mixtures, and thereby allow some assessment of the fractionation method's ability to detect and measure SBM-derived olefins at increasingly lower concentrations in crude oil.The various TPH concentrations determined from this exercise are given in Table 1 and the whole oil and F12 fraction GC-FID chromatograms for Mixture B (80:20) are shown in Figure 4. Chromatograms for the other mixtures are provided in the Supporting Information (Figures S2 and S3).
The Mixture B whole oil gas chromatogram is dominated by n-alkanes and acyclic isoprenoid peaks and does not exhibit any recognizable SBM-derived olefins (Figure 4A, inset).However, following fractionation, the F12 fraction's chromatogram of Mixture B is found to be dominated by the C 16 and C 18 clusters typical of SBM-derived olefins (Figure 4B).Even a minor amount of C 20 olefin isomers are revealed (Figure 4B,inset).These SBM-derived olefins clusters in the F12 fraction of Mixture B (Figure 4B) exhibit the same isomer profiles as were present in the parent crude oil containing a relatively low abundance of SBM-derived olefins (Figure 3C).Some additional peaks are present in the F12 fraction of Mixture B, including some trace n-alkanes, that indicate some breakthrough of the F11 fraction into the F12 fraction likely occurred.(This is also evident in the both Mixture A's and C's F12 fractions (Figures S1 and S2, respectively, suggesting additional mass of AgNO 3impregnated silica gel may be appropriate.)This breakthrough, however, did not inhibit the ability to recognize the presence of SBM-derived olefins in the C 15 to C 20 range of each of the mixtures.
The TPH SBM concentrations measured in Mixtures A, B, and C (11,800, 5570, and 4620 mg/g, respectively; Table 1) are in each case higher than would be calculated from the amounts of the two parent oils used in the mixtures.For example, Mixture B should contain only $3690 mg/g of TPH SBM (i.e., 10% of 36,900 mg/g in the fresh crude oil with low SBM), yet was measured to contain 5570 mg/g of TPH SBM through integration of the F12 chromatogram's C 15 to C 20 range (Table 1).This difference testifies to the difficulty of accurately measuring SBM-derived olefins at low concentrations (although the breakthrough of some alkanes noted above also likely contributed to the measured mass of SBM olefins).This issue is most obvious within Mixture C (95:05) wherein the TPH SBM concentration measured (4620 mg/g; Table 1) is $2.5-times higher than would be calculated ($1850 mg/g, or 5% of 36,900 mg/g).This disparity argues that a de facto detection limit for SBM-derived olefins in crude oil via one-dimensional GC-FID of the oil's F12 fraction is on the order of 5000 mg/g, or 0.5 wt%.SBM-derived olefins may be detected in crude oil at lower concentrations although careful integration of only clearly identified olefin clusters must be performed.Although the authors acknowledge this simple mixing exercise is not a robust quantitative assessment of the method's SBM-derived olefin detection limit, it serves to demonstrate its improved capability over an assessment of these compounds' presence/absence or concentration without F12 fractionation.

Oils and sediment: Taylor energy site
The presence of SBM-derived olefins in some sediments and surfaced oil sheens historically collected at the Taylor Energy oil spill site (Mississippi Canyon, Block 20 or MC20) has been reported based upon one-dimensional GC-FID and GC x GC-FID (Bryant et al., 2020).Olefins detected in some surface oil sheens indicates that at least some of the oil that exits the seafloor at the spill site was likely derived from sediments contaminated with SBM during the drilling of intervention wells following the platform's toppling (Bryant et al., 2020).However, most surface sheens from the site do not appear to contain SBM-derived olefins, at least not at concentrations visible when the whole (unfractionated) sheens are analysed using onedimensional GC-FID (Overton and Reddy, 2017).Yet oil continues to exit the seafloor in one particular area, i.e., within an erosional pit, at a persistent high rate (Mason et al., 2019;Couvillion Group, LLC, 2021), which indicates that oil continues to actively leak from damaged wells somewhere below the pit.Because crude oil from an actively leaking well(s) would not contain SBM-derived olefin, analysis of six oils collected above the erosional pit and one sediment collected near but outside the erosional pit (Table S-1) using the AgNO 3 fractionation method (described herein) provides an opportunity to further assess whether or not the crude oil escaping the seafloor at the site's erosional pit contains SBM-derived olefins at a concentration not obvious from GC-FID of the whole (unfractionated) oils.
Figure 5 shows the GC-FID chromatogram for one of the whole (unfractionated) oils collected from the site's subsea containment system installed over the erosional pit.Unlike the fresh (unweathered) crude oil from the Macondo well (Figure 2), the crude oil from the Taylor Energy site is biodegraded as evidenced by the absence of n-alkanes and the presence of a prominent UCM (Figure 5A), both well-established features of biodegraded crude oil (Frysinger et al., 2003).Peaks  S2 for peak label identities.
attributable to SBM-derived olefins are not visually evident in the whole oil chromatogram (Figure 5A), but it is uncertain if their presence at some low level is lost within the biodegraded oil's UCM.However, following AgNO 3 fractionation their absence is confirmed in the chromatogram of the oil's F12 fraction (Figure 5B).In fact, all six crude oils collected from the Taylor Energy site were determined to contain no SBMderived olefins (Table 1).The six oils did, however, contain 14,300 to 32,400 mg/g of native olefins (1.4 to 3.2 wt%; Table 1).These concentrations are, on average (23,300 ± 6570 mg/g), higher than were observed in the three Deepwater Horizon oils studied (13,600 ± 378 mg/ g; Table 1), which likely is attributable to these two oil groups' different geologic origins in the northern Gulf of Mexico.The absence of detectable SBM-derived olefins in the six oils collected from the Taylor Energy site, including three oils from the site's subsea containment system installed over the erosional pit and three oils captured rising through the water column by ROV (Table 1), argues that these oils are overwhelming or exclusively emanating from actively leaking wells, and are not being exhumed from sediments containing crude oil contaminated with SBM.This conclusion is supported by the results obtained for the seafloor sediment sample collected from the Taylor Energy site seafloor, but outside of the erosional pit. Figure 6A shows that the sediment's whole (unfractionated) extract does not appear to contain any SBMderived olefins.However, isolation and GC-FID analysis of the extract's F12 fraction clearly shows the sediment contains SBM-derived olefins (Figure 6B) at a concentration of 0.978 mg/g-dry (Table 1).This result confirms that although sediments at the site may contain low concentrations of SBM-derived olefins (Figure 6B)and that some surface sheens may contain SBMderived olefins (Bryant et al., 2020) most of the oil exiting the seafloor does not (Table 1; Figure 5B).

Conclusions
Offshore oil well drill or oil spill sites can be impacted by both crude oil and by chemicals present in synthetic-based drilling muds (SBMs) discharged during the original drilling of the wells, or (as occurred during the Deepwater Horizon event) during attempts to "kill" a leaking well.Native hydrocarbons in crude oil can prevent the detection and confound the  S2 for peak label identities.
quantification of (non-native) SBM-derived olefins, especially when the latter are present in low concentrations relative to crude oil.In this study the efficacy of silica gel impregnated with AgNO 3 to physically isolate olefins from other (aromatic and saturated aliphatic) hydrocarbons in crude oil and oil-impacted sediment prior to analysis using conventional (one-dimensional) GC-FID is demonstrated.The presence of any SBM-derived olefins can then be readily recognized, owing to their distinctive C 14 -C 20 even carbon clusters, and their concentrations measured separately from any native olefins in crude oil.This method provides a practical alternative to comprehensive (two-dimensional) GC x GC-FID chromatography for SBM-derived olefin quantification, which presently has a limited commercial availability.
Even when present at relatively low concentrations, SBM-derived olefins can be readily isolated by open column chromatography (Figure 1) and then quantified by GC-FID.This study of laboratory mixtures showed that a de facto detection limit of SBM-derived olefins in oil on the order of 5000 mg/g, or 0.5 wt% can be achieved.SBM-derived olefins in oiled seafloor sediment were easily detected at a concentration of 1 mg/g-dry.
Although the liquid column AgNO 3 fractionation method demonstrated herein was directed at quantifying SBM-derived olefins, the method's application for studying native olefins in crude and refined petroleum warrants further study.

Figure 1 .
Figure 1.Flowchart showing the open column chromatography scheme for isolating SBM-derived olefins within the F12 alkenes fraction in oil (or sediment extract).Subsequent GC-FID analysis of the F12 fraction was used to recognize and quantify the absolute concentration of SBM-derived olefins in oil (or sediment extract) samples.DCM: dichloromethane.

Figure 2 .
Figure 2. GC-FID chromatograms of unfractionated fresh crude oils containing (A) a high relative abundance of SBM-derived olefins, (B) a low relative abundance of SBM-derived olefins, and (C) no SBM-derived olefins.Insets show expanded view of the C 15 to C 20 range in each.See TableS2for peak label identities.

Figure 3 .
Figure 3. GC-FID chromatograms of crude oil containing a low relative abundance of SBM-derived olefins demonstrating the effectiveness of liquid column chromatographic separation and isolation of SBM-derived alkenes.(A) F1 aliphatic fraction, (B) F11 alkane fraction, and (C) F12 alkene fraction (per Figure 1).Insets show expanded view of the C 15 to C 20 range in each.See TableS2for peak label identities.

Figure 4 .
Figure 4. GC-FID chromatograms of (A) the whole (unfractionated) oil of Mixture B in which the SBM-derived olefins known to be present are not visibly evident and (B) the F12 fraction of the Mixture B oil revealing the presence of SBM-derived olefin clusters.Insets show expanded view of the C 15 to C 20 range in each.See TableS2for peak label identities.

Figure 5 .
Figure 5. GC-FID chromatograms of (A) whole (unfractionated) weathered crude oil from the subsea containment system at the Taylor Energy site (Containment oil B) in which the SBM-derived olefins present are not visibly evident and (B) F12 fraction of the same oil showing the absence of SBM-derived olefins.Insets show expanded view of the C 15 to C 20 range in each.See TableS2for peak label identities.

Table 1 .
Concentration of TPH, TPH F12 , and TPH SBM in the samples studied.(oil wt.) and sediment concentrations in mg/g (dry wt.).a TPH F12 minus TPH SBM .b Proportions refer to weight percent (fresh crude oil w/ low SBM:fresh crude oil w/ no SBM).nd: SBM-derived olefin clusters not detected within F12 chromatograms.