Monitoring polycyclic aromatic hydrocarbon (PAH) attenuation in Arctic waters using fluorescence spectroscopy

Highlights

• Five high molecular weight PAHs proved potentially useful as degradation model.

• PAHs exhibited degradation/dissipation under simulated Arctic conditions.

• EEM fluorescence spectroscopy provides valuable method for monitoring concentrations.

• A three component PARAFAC model achieved 98.8% fit.

• Chemical species modulation was observed using PARAFAC.

Abstract

As oil exploration in the Arctic grows, the risk of crude oil exposure to the environment through spills and leakage increases. Polycyclic aromatic hydrocarbons (PAHs) are a toxic component of crude oil that are highly insoluble and persist in the environment. Much is known about PAH degradation through abiotic and biotic factors and remediation strategies in temperate climates; however, little is known about the degradation of these compounds in the Arctic where cold temperatures and sea ice predominate and remediation strategies differ greatly. In this study, excitation-emission matrix (EEM) fluorescence spectroscopy was used along with parallel factor analysis (PARAFAC) to analyze concentrations of PAHs, associated hydroxylated metabolites, and microbial biomass (as based on the protein-like indicators: tryptophan and tyrosine) in surrogate solutions to develop a correlation between PAH biodegradation and native microbial growth. EEMs generated from solutions of 16 EPA-listed priority pollutant PAHs, metabolites, tryptophan, and tyrosine were characterized. Based on maximum emission wavelength peak intensity (EM_λmax), PAHs were found to best categorically group, in an effort to determine which PAHs would serve as effective indicators in comparison to bioindicators (microbial fluorescence-absorbing proteins, smaller labile PAHs, and hydroxyl-PAHs), when EEMs were divided into two regions at EM = 400 nm for all excitation wavelengths, establishing a line of division within the matrix to minimize spectral overlap between indicator groups. Five high molecular weight PAHs (benzo(b)fluoranthene, benzo(k)fluoranthene, fluoranthene, benzo(ghi)perylene, and indeno(1,2,3-cd)pyrene) exhibited peak fluorescence intensities above EM = 400 nm allowing them to serve as PAH indicators while bioindicators presented near or below the line of division. A microcosm batch-incubation experiment, consisting of two PAH treatment groups and a control, demonstrated degradation/dissipation of the high molecular weight PAHs (p < 0.05). A half-life of 128 d was derived for the PAH group treated with a solution of the five high molecular weight PAHs (C₀ = 512 μg/L) and 31 d for a treatment of only fluoranthene (C₀ = 96 μg/L). A three component PARAFAC model containing incubation samples and aqueous standards accounted for 98.8% variance. The combination of EEM fluorescence spectroscopy and multivariate analysis provides a valuable method for modeling degradation studies and monitoring PAH concentrations and microbial growth under Arctic conditions.

Introduction

Petroleum remains the world's leading energy source fueling the need for exploration of untapped reserves. The U.S. Geological Survey (2008) estimated that approximately 13% of the world's undiscovered oil lies in the Arctic, a region that has become more accessible due to climate change. The risk of crude oil exposure to the environment by means of leakage and accidental spills rises as offshore drilling, production, and transportation increases. Current response and remediation strategies include booms for containment, skimmers for collection, the application of chemical dispersants, oil herding surfactants, and in situ burning (ISB) (American Petroleum Institute, 2016). Many of these strategies have proven highly effective in temperate climates; however, the Arctic's unique environment poses challenges including cold temperatures, drift ice, and remoteness, decreasing response time and requiring alternative strategies from conventional clean-up methods (Aggarwal et al., 2017, Bullock et al., 2017).

Under cold Arctic conditions, a typical remediation strategy involves the application of oil herding surfactants around the spill in order to contract the oil, thickening the slick. With a thickness of ≥ 2 mm (minimum 3 mm for weathered oil), ignition of an oil slick and successful burning can be achieved (Buist et al., 2011, Aggarwal et al., 2017). Mid-scale laboratory experiments conducted at the Fire Training Grounds in Prudhoe Bay, Alaska during 2006 demonstrated a direct correlation between oil volume and burn efficiency under brash ice and slush ice conditions while field experiments performed in 2008 near Svalbard in the Barents Sea achieved 90% consumption of 603 L of Heidrun crude oil herded by a U.S. Navy cold-water (USN) herder (Buist et al., 2011). Despite large reductions in oil volume, residues varying in chemical composition based on crude oil type and burn efficiency remain. Chemical analysis of the residues suggests that burning may increase concentrations of high molecular weight polycyclic aromatic hydrocarbons (PAHs). Several factors including thickness of the slick, initial oil density, incomplete burning, and residue temperature may cause the residues to sink, introducing PAHs into the aquatic environment (Fritt-Rasmussen et al., 2015).

The majority of hydrocarbons within crude oil can be degraded by microorganisms under aerobic and anaerobic conditions (Atlas and Hazen, 2011). The rates and extent of biodegradation are affected by several factors including temperature, nutrient availability, and native microbial communities. Psychrophilic and psychrotolerant microbes having the ability to break down organic contaminants have been discovered in Arctic sea ice and seawater. Previous studies have identified degraders in the families of Pseudoalteromonas, Psuedomonas, Psychrobacter, and Shewanella (Dong et al., 2015). In addition to the discovery of Arctic microorganisms, microcosm experiments have shown degradation of PAHs occur at temperatures ranging from − 1.8–5.5 °C (Brakstad et al., 2008). Naphthalene exhibited half-lives from 1.5–1.7 d, phenanthrene from 2.1–4.0 d, and fluorene from 2.1–4.4 d (Siron et al., 1995). Despite these findings, the dynamics of PAH biodegradation by native microbes in Arctic waters and potential for use as a remediation tool is still, in large, unknown.

PAHs are of great concern to human health due to their carcinogenic, teratogenic, and mutagenic properties (Dong et al., 2015). Comprised of two or more fused benzene rings, varying in linear arrangements, these compounds are highly lipophilic. As molecular weight increases, aqueous solubility decreases (Kanaly and Harayama, 2000). PAHs are known to persist in aquatic marine environments (Sinaei and Mashinchian, 2014), which increases the potential to enter marine organisms through absorption or trophic uptake. The stronger the hydrophobic properties of the individual compounds, the greater the accrual within fat cells leading to potential bioaccumulation and biomagnification (Kanaly and Harayama, 2000, Pena et al., 2015). The lack of solubility that causes PAHs to be stored in adipose tissue prompts the compounds to be metabolized into more soluble forms in an effort to excrete them from the body. This process occurs through the cytochrome P450 monooxygenase system, which may convert some of the PAHs to highly toxic metabolites (Pena et al., 2015).

Due to high toxicity at relatively low concentrations, 16 PAHs (Fig. 1) have been recognized by the U.S. Environmental Protection Agency (EPA) as priority pollutants. Of the 16 PAHs, fluoranthene and pyrene, commonly found in most crude oils and ISB residues (Fritt-Rasmussen et al., 2013, Fritt-Rasmussen et al., 2015, Pampanin and Sydnes, 2013, Stout and Payne, 2016), are frequently used as models for PAH degradation (Holman et al., 2002, Kanaly and Harayama, 2000). Monitoring the metabolites of these two compounds (Fig. 1) may serve as an indicator of microbial degradation.

The aromatic ring structures within the PAHs and their hydroxylated metabolites possess naturally fluorescing properties. Fluorescence spectroscopy may be used to analyze the concentrations of these compounds to determine degradation and metabolite production (Nahorniak and Booksh, 2006, Pena et al., 2015, Zhuo et al., 2015). Traditional fluorescence techniques, which generate a single line fluorescence spectrum by scanning emission wavelengths at a fixed excitation wavelength or excitation wavelengths at a fixed emission wavelength, lack selectivity due to broad fluorescence peaks. To generate an excitation-emission matrix (EEM), individual scans covering the desired region are combined. This process requires time and may lead to an inaccurate representation of dynamic solutions (Nahorniak and Booksh, 2006); however, recent technological developments that utilize multi-dimensional EEMs, derived from EEM fluorescence spectroscopy, scan excitation and emission ranges simultaneously, reducing broad spectra, and increasing selectivity (Lu et al., 2013, Nahorniak and Booksh, 2006).

EEMs, each consisting of three-way data, form a data cube when combined for a series of samples. Decomposition of the data cube into three loading matrices, generating a trilinear model that minimizes the sum of squares of the residuals (E_ijk), may be achieved using the parallel factor analysis (PARAFAC) algorithm:

$$X_{\mathit{ijk}}=\sum _{f=1}^F{A}_{\mathit{if}}{B}_{\mathit{jf}}{C}_{\mathit{kf}}+E_{\mathit{ijk}}$$

where X_ijk is the fluorescence intensity produced at the emission wavelength j and excitation wavelength k for sample i based on the number of fluorescing components F used to build the PARAFAC model (Stedmon and Bro, 2008). Each component consists of one score vector, C, that represents concentration and two loading vectors, A and B, allowing for differentiation between fluorescing analytes and interferents and classification of such analytes (Alostraz et al., 2008, Jiji et al., 1999, Lu et al., 2013). Previous studies have analyzed crude oil and other petroleum products by means of EEM fluorescence spectroscopy and PARAFAC analysis to establish selectivity and determine concentration (Alostraz et al., 2008, Christensen et al., 2005). Yang et al. used PARAFAC to categorize individual PAHs with overlapping spectral bands and variation in fluorescence intensity, in addition to predicting concentrations of PAHs (phenanthrene, pyrene, anthracene, fluorene, acenaphthene, and fluoranthene) added to river and reservoir water (Yang et al., 2016).

Fluorescence spectroscopy is widely used to analyze microbial biomass due to the naturally fluorescing aromatic and indole rings found in amino acid residues (Fig. 2) within the cell (Ghisaidoobe and Chung, 2014). These compounds fluoresce in a different region from large PAHs of 4 or more rings. Dubnik et al. (2010) noted protein-like compounds (tryptophan and tyrosine) exhibited emission maxima at 330/320 nm; whereas, Pena et al. (2015) observed emission maxima of benzo(b)fluoranthene, commonly found in ISB residues (Fritt-Rasmussen et al., 2013), at 440 nm.

In this study, 16 PAHs were examined and grouped according to their spectral overlap properties to determine which compound classes resolve sufficiently in an attempt to provide a differentiable PARAFAC model and which representative PAHs may be useful as a model for degradation studies. Further, the findings of the spectral properties of these surrogate solutions were compared to the results of a microcosm experiment consisting of seawater spiked with PAHs. Using PARAFAC, microbial growth and PAH degradation correlation was evaluated. To our knowledge, there are no known studies having made such attempts to identify potential regions of spectral overlap to classify the absorbance regions, which can be grouped for future PAH degradation studies, or any effort to attempt to link PAH biodegradation with microbial growth directly and simultaneously.