Overview: Dr. Darrell Sparks at Mississippi State University was awarded an RFP-II grant at $1,122,299 to conduct the RFP-II project titled, "Characterizing the Composition and Biogeochemical Behavior of Dispersants and Their Transformation Products in Gulf of Mexico Coastal Ecosystems". The project consisted of 3 institutions (Mississippi State University, Duke University, Stony Brook University), 2 principal investigators (Sparks, Armbrust--Armbrust was lead-PI for the first 2 months of project), 4 co-PIs (Armbrust, Brownawell, Ferguson, McElroy), 4 graduate students (Adelwale, Dasgupta, Getzinger, Huang) and participation by several research technicians and graduate students.
The use of chemical dispersants is a key emergency response option available to mitigate impacts of large oil spills on sensitive coastal or surface water ecosystems and resources. Long controversial, the relative merits of dispersant use came under intense scrutiny when approximately 1.84 million gallons of Corexit® series dispersants were used during the Deepwater Horizon (DH) oil spill that occurred in the Gulf of Mexico (GOM) in 2010. Despite recent and intense investigation, many questions remain concerning the hazards posed by largescale dispersant application in marine oil spills. Importantly, the transformation, fate, and transport of crude oil components released to the marine environment through accidental spills has been well studied over many years, but the fate of dispersants applied to mitigate spills of oil in the sea is largely unknown. The proposed research addresses toxicity, sorption, and occurrence in seafood of dispersant components and transformation products, questions that have not been well addressed in the past. This proposal addresses GoMRI-II Project Theme #2: "Chemical evolution and biological degradation of the petroleum/dispersant systems and subsequent interaction with coastal, open-ocean, and deepwater ecosystems". To date, there has been insufficient effort investigating microbial degradation of dispersants, their toxicity and their interaction with particulate matter. Comprehensive examination of these processes will improve the understanding of dispersant fate and potential effects in the GOM, and will help to drive development of safer and more effective dispersants which will ultimately lead to protection of the GOM ecosystem. The over-arching goals of the proposed research are to improve the knowledge base on dispersant behavior in the coastal marine environment and to increase understanding of the processes that controlled dispersant fate and biological exposure in the GOM following the DH oil spill. The tools developed and information gained from this project will be critical for more informed comparative assessments of the safety of different dispersants based on their specific ingredients. The specific aims of the research are to:
- Characterize the persistence of the complex mixture of components in Corexit® 9500 and a range of other surfactants that are presently used in alternative commercial formulations approved for use in the GOM, and elucidate resulting persistent degradation products; this will be conducted in aerobic incubations with GOM seawater; through coordination with the Tulane University dispersant GoMRI consortium, similar tests with more promising surfactant/dispersants developed during the project period will also be conducted.
- Identify toxic transformation products of dispersants produced under aerobic conditions in seawater under Aim 1; transformation products will be isolated and characterized with novel toxicity identification and evaluation (TIE) approaches on degraded dispersant mixtures.
- Assess the particle reactivity of dispersant components to evaluate phase partitioning that controls bioavailability and possible loss to sediment reservoirs in the GOM; sorption of dispersant chemicals, and identified transformation products (from #1 & #2) to surficial GOM sediments and suspended particles will be determined.
- Develop tools and evaluate the migration of potentially toxic dispersant transformation products into marine organisms used for human consumption through incorporation of results from (#1 and #2) into improved methods for determining dispersant component residues and Itransformation products in seafood from the GOM in freshly collected and archived samples.
The work represents a tightly integrated collaboration between three laboratories, The State Chemical Laboratory at Mississippi State University, Duke University, and Stony Brook University. Each partner will contribute to important aspects of the research. This research team is highly complementary, bringing to bear the best expertise available to address the research goals. Key to the overall effort is the complimentary advanced analytical (e.g. HPLC-MS) capabilities and expertise at all three institutions, as well as previous experience in analysis of dispersants and surfactant components. Toxicity screening studies will help identify important dispersant components and metabolites.
Research Highlights: Dr. Sparks' research, which included 2 outreach products and activities, resulted in 4 peer-reviewed publications and 7 datasets being submitting to the GoMRI Information and Data Cooperative (GRIIDC), which are/will be available to the public. Dr. Sparks engaged 4 graduate students over the award period. Significant outcomes of his research according to GoMRI Research Theme 2 are highlighted below.
The primary tasks can be categorized as follows:
- Task 1: Assessment of the toxicity of dispersants, their components, and breakdown products
- Task 2: Assessment of phase partitioning and sorption behavior of dispersants, their components, and breakdown products
- Task 3: Characterization of dispersants, their components, and breakdown products
Detailed and complementary methods were developed at Duke (Ferguson) and Stony Brook (Brownawell) Universities and at Mississippi State University (Sparks) to determine dispersant and dispersant residues with three very different LC-MS platforms. With diverse applications and approaches, and different abilities to resolve Corexit components, our work has supported method development in a way not possible had we been working in isolation. One not previously appreciated finding has been the very large number of Tween components that have very similar masses (but can be separated either be chromatography or high resolution mass spectrometry), and other structurally different homologous series that have identical elemental formulas, often very difficult to separate chromatographically. We are nearly ready to publish a comprehensive analysis of Corexit components that greatly extends what was already known about Tween and SPAN composition based on less detailed or comprehensive approaches. We have similarly identified some of the major polyethoxylated metabolites of Tweens, and we have seen that they are relatively persistent in comparison to their parent compounds. Much work remains to be done to identify other Corexit ingredient metabolites and determine their relative persistence and potential toxicity.
a. Results for Task 1: Assessment of the toxicity of dispersants, their components, and breakdown products
Toxicity assessment using Sheepshead minnow embryos and larvae found Corexit 9500 and 0537 to have limited acute (48hr) toxicity with LC50s in the range of 60 mg/L, similar to what others have reported in a variety of aquatic organisms. Comparison of the LC50 of 7 Corexit components including solvents, anionic and ionic dispersants demonstrated only DOSS was significantly more toxic than the Corexit mixtures with 96hr LC50s of 8 mg/L in larvae. Of potentially greater importance is the genotoxicity of Corexit and some of its components. DOSS and some of the solvents lead to significant increases in DNA damage (measured using the comet assay) at concentrations as low as 2 mg/L. Further studies at lower concentrations are planned. These studies demonstrate that some persistent components of Corexit have enhanced toxicity and genotoxicity. The potential for interactive effects with other co-occurring contaminants and oil needs to be assessed, as well as the mechanisms by which these compounds are exerting their toxic effects.
Highlights of toxicity assessment activities in Year 2 included the first comprehensive analysis of Corexit and Corexit components in one species. Larvae were found to be more sensitive than embryos to Corexit 9500 with LC 50s of 56 vs 162 mg/L for larvae and embryos respectively. These results indicated that early life stage sheepshead minnow were among the more sensitive organisms tested, particularly when compared to other fish species. Of the Corexit components tested, only DOSS showed appreciable acute toxicity (LC50 larvae of 7 mg/L), indicating that it is likely the primary contributor to the acute toxicity of the Corexit mixture. Further work evaluating the genoxicity of Corexit and its major components using the Comet assay indicated that both Corexit 9500 and 9527 were only weakly genotoxic, but that DOSS and two of the solvents found in Corexit (DPGBE, Petroleum distillates and 2-butoxyethanol) could significantly increase DNA damage at much lower concentrations with lowest observable effects concentrations sometimes as low as 0.5 mg/L. The ramifications of this result are being further evaluated. DOSS and DPnB were found to be persistent in our exposure systems, thereby raising the possibility of more sustained exposures to these compounds, particularly in benthic environments in environmental settings. This indicates that while environmental levels of Corexit and DOSS are generally much lower than those found to be toxic here, further work on their environmental persistence and effects are warranted. Experiments evaluating the combined effect of hypoxia on exposure to Corexit alone, and WAF and CEWAFs made from Macondo Surrogate Oil demonstrated that hypoxia significantly enhanced toxicity in most cases. Interestingly, although WAFs generated at 1g/L caused no significant increase in mortality in sheepshead embryos and had very low total petroleum hydrocarbons concentrations (<10 mg/L), they significantly induced cytochrome P450 1A (EROD) activity in exposed larvae generated more of a response than CEWAFs (despite much higher TPH levels), indicating that something not captured in our fluorescent TPH measurement is contributing significantly to the effects observed. EROD activity was reduced by hypoxia, while DNA damage measured by the Comet assay was increased by hypoxia, indicating the possibility of a mechanistic link between CYP1A activity and hypoxic response. The enhancement of toxicity observed under hypoxic exposures indicates that oil release with or without dispersant use in hypoxic waters or sediments possess a greater risk to aquatic organisms.
b. Results for Task 2: Assessment of phase partitioning and sorption behavior of dispersants, their components, and breakdown products
As expected, the sorption of nonionic surfactants is appreciably greater than that of anionic DOSS, and the sorption coefficients measured are low enough to indicate that DOSS would not be appreciably scavenged from deep GOM water columns by sinking particles, consistent with hypothesized near conservative behavior in the deep water plume during and following the Deepwater Horizon spill. The low level measurements of DOSS in sediments near and away from the Macondo well (in the low ng/g range) represent a very small fraction of the water column inventories that were found in sediment repositories. Our results will provide more insight on the levels that could accumulate in coastal or shelf sediments should Corexit be used closer to shore, where particle fluxes are greater and higher water column concentrations are found. More particle reactive Span and Tween components are present at much lower levels in Corexit dispersants. Ongoing studies will better inform whether these components are likely to accumulate in sedimentary environments at higher levels than the more soluble and intrinsically persistent DOSS.
Our studies on effects of sediment properties and DOSS concentrations (isotherms conducted over a range of concentrations of up to nearly 5 orders of magnitude and overlapping with levels determined in the GoM) are perhaps the most important to those interested in the fate of Corexit components during and after the DwH spill. The results from sorption isotherm studies point to an excellent relationship between sorption coefficients and sediment organic matter and that isotherms on 3 sediments completed so far (partial isotherms on 3 others) indicate a surprising degree of isotherm linearity. This implies a near constant sorption coefficient for between dissolved DOSS and suspended or sinking organic matter in the water column. Based on these observations of organic matter normalized sorption coefficients, we have results that constrain the interpretations of the mechanism that led to the deposition of and preservation of relatively high levels of DOSS (2-9 ppm) in sediments in three core tops reported by White and coworkers this past year. Based on surface water and deep plume concentrations that have now been reported, our sorption coefficients, and the conclusion that water column DOSS measured was largely dissolved, we conclude that it is very unlikely that the mechanism of delivery of DOSS to sediments with ppm concentrations could be caused by scavenging of DOSS from the dissolved phase by falling particles appears (consistent with the sentiment expressed by White) as it would require too high of an organic matter flux through contaminated areas of the water column. This conclusion assumes that oil coatings or oil-particle aggregates do not have much higher partition coefficients. That similar scavenging by oil phases may also not have been important is supported by preliminary work we have done on partitioning of DOSS between seawater and surrogate Macondo oil, work we are currently following up on. A large amount of effort on sorption was spent on determining the effects of salinity on sorption of DOSS with 3 sediments, the effect of pH (one sediment), and following up on those results with detailed studies separating out the effects of important components of sea salt (namely Na, Ca., and Mg), to help with our interpretations of mechanisms.
c. Results for Task 3: Characterization of dispersants, their components, and breakdown products
Our analytical efforts have yielded significant advances in the understanding of Corexit 9500 component identification. Specifically, we have identified and are in the process of cataloguing all of the monomeric and polymeric surfactant components of this mixture using high resolution mass spectrometry. We routinely utilize an LTQ-Orbitrap mass spectrometer to achieve < 2 ppm mass accuracy and >100,000 resolution in this analysis, and this capability has allowed us to resolve many nearly isobaric components in the mixture. Further, we have also utilized next-generation Orbitrap technology (in cooperation with ThermoFisher Co.) to achieve < 1 ppm mass accuracy and 450,000 resolution for Corexit 9500 component analysis. These latter experiments have provided unambiguous molecular formula confirmation for more than 1,000 individual molecular species. Our results to-date for degradation of Corexit 9500 dispersant components in seawater indicate a wide variation in the persistence of individual molecular species in this mixture, and much of this variability seems to be associated with the degree of polyethoxylation of the nonionic surfactant species in the mixture. Our ongoing work is centered on identifying and quantifying the individual degradation products of these mixtures and more precisely defining the degradation rates for important species within Corexit 9500.
One of the most significant results has been the advance in the understanding of Corexit 9500 component identification that we have made. Our analytical methods have enabled the identification and characterization of most monomeric and polymeric surfactant components of Corexit 9500 using high-resolution mass spectrometry. This is enabled by our application of both ultra-high resolution (100,000-250,000) mass spectrometry with very high mass accuracy (< 1 ppm), as well as the development and application of multidimensional UHPLC separation methods to nonionic surfactant analysis. These experiments have enabled unambiguous molecular formula confirmation for more than 1,000 individual molecular species. Brownawell and Ferguson are preparing a manuscript detailing these results for publication in Year 3 of the project. Results of Corexit 9500 dispersant component biodegradation experiments in seawater show that individual surfactants within the dispersant degrade at highly variable rates, and much of this variability seems to be associated with the degree of polyethoxylation of the nonionic surfactant species in the mixture. The ionic surfactant component, dioctylsulfosuccinate seems to be the most persistent component of Corexit 9500, although our results indicate that this compound is ultimately degradable over month time scales. Ongoing work centers on identifying and quantifying the individual degradation products of these mixtures and more precisely defining the degradation rates for important species within Corexit 9500.
The most significant accomplishment was the full development of a comprehensive LC x LC two-dimensional UHPLC method for separation and analysis of nonionic surfactants at trace levels in seawater. This method allows for the first time complete separation of both ethoxymers and hydrophobes in highly complex mixtures of oligomeric nonionic surfactants. Application of the method to dispersant analysis in seawater has resulted in increased sensitivity and more complete understanding of the fate of these surfactant components in the marine environment. The method will be published as a manuscript during the final year of the project.
Results of Corexit 9500 dispersant component biodegradation experiments in seawater have illustrated that the individual nonionic surfactants within the dispersant degrade at highly variable rates, and much of this variability is associated with the degree of polyethoxylation of the nonionic surfactant species in the mixture. The ionic surfactant component, dioctylsulfosuccinate seems to be the most persistent component of Corexit 9500, although our results indicate that this compound is ultimately degradable over month time scales. The Ferguson laboratory's ongoing work seeks to elucidate the kinetics of these biotransformation processes under different nutrient conditions in seawater, and to elucidate definitively the mechanisms by which dispersant transformation products are formed under both biotic and abiotic conditions.