Alberta’s oil sands produce 1.9 million barrels of oil each day and contributes to 55% to all Canadian crude production (1). This industrial activity poses a potential threat to the regional environment, with concerns over potential water contamination from toxic chemicals. LCGC spoke to Jonathan W. Martin, an associate professor from the Division of Analytical & Environmental Toxicology at University of Alberta (Alberta, Canada) about the analytical challenges he faces and solves in this research.
Q. How did you enter the field of environmental toxicology? What are the main objectives of your research group?
A: In high school my favourite classes were chemistry and biology, and by University I realized that toxicology was the perfect blend of both fields. Over the course of my undergraduate program in Toxicology (University of Guelph, Ontario, Canada) I was most inspired by a course in Analytical Environmental Toxicology. To me, this was like detective work with the aim of finding and convicting the guilty chemical culprits. If you were sloppy with your detective work, your evidence would never stand up in court! I was hooked, and spent the following four years doing a PhD in the area of environmental chemistry and trace environmental analysis. It just so happened that, during my PhD, I discovered several new chemical culprits (fluorinated ones) that today continue to occupy my research team at the University of Alberta. These included long-chain perfluorinated carboxylates in Arctic mammal tissues, which we confirmed by HPLC-QTOF, and semi-volatile perfluorinated acid precursors in ambient air samples by GC–MS, using positive and negative chemical ionization.
Overall, the main objectives of my research program are to examine the sources, fate, and effects of environmental organic contaminants. The analytical methods we develop depend on the particular question, on the environmental matrix, and on the properties of the analytes. The ultimate method of choice is almost always chromatography paired to mass spectrometry. On the chromatography side we use GC, HPLC, comprehensive 2D-HPLC, and most recently we have grown very fond of supercritical fluid chromatography (SFC). On the mass spectrometry side we use everything from low-resolution triple quadrupoles, through high-resolution QTOF, and ultrahigh resolution orbitraps. The burgeoning oil sands industry located ~300 km north of our campus near Fort McMurray, has created some unique opportunities to apply all of the above to exotic chemical mixtures being released to air and water.
Q. What is the importance of oil sands and why do they pose an environmental threat?
A: The oil sands industry has been a major driver of national prosperity in Canada, but over the last five years the associated environmental issues have attracted increasing national and international scrutiny. One reason for this attention, and for my work in the area, are the conclusions from two expert panels that historic environmental monitoring around the industry has not been adequate. The issues are vast and include water and air pollution, land disturbance, and greenhouse gas emissions, but I’m most interested in the water issues because all the major industrial processes use large volumes of water to extract the oil (more correctly termed bitumen) from clay and sand particles in the ore.
For water, part of the problem is simple math. Each barrel of bitumen requires at least 1 barrel of fresh water for production, and current production rates approach 2 million barrels per day. The resulting process-affected water is toxic, owing to a complex and mostly unresolved mixture of organic compounds, and so-far not a drop of water has been remediated. As a first step, we are presently applying effects-directed analysis to identify which group(s) of chemicals in this water imparts most of the toxicity. Current volumes of this water are close to 1 billion cubic meters, and people are growing concerned that this water is leaking into natural streams and rivers in the area. The toxic dissolved organic compounds are also quite persistent, meaning that the water will remain toxic for many decades if left to sit – which is the only strategy being used.
At the bench-scale, my collaborators and I have shown that ozonation and advanced oxidation work very well to detoxify the water, and to degrade the highly persistent organic acids, but this has not yet been tested in the real world. Currently the only policy around this water is “no release”, meaning the water must be stored in large man-made reservoirs, which I don’t believe is sustainable.
Q. How are you applying analytical chemistry techniques to quantitate the effects of the oil sands industry on the local environment?
A: The toxic dissolved organic compounds in oil sands process-affected water are a super complex mixture of acidic and polar substances, including ‘naphthenic acids’ – a curious and confusing term, since nobody seems to agree on how to define it, based either on empirical formula or structure. Nobody knows for sure, but I estimate that a typical process-affected water sample contains more than 1 million distinct chemicals, and to date only a couple dozen of these have been identified. To begin to unravel this mixture, our approach has been to use various forms of chromatography (such as HPLC, 2D-HPLC, and SFC) paired to “high” and “ultrahigh” resolution mass spectrometry (2,3). Switching between positive and chemical ionization modes, or between different sources, reveals complexity that is similar to the complexity faced in petroleomics.
To monitor the surrounding natural rivers, sensitivity is another challenge because any seepage of industrial water can be quickly diluted. The Athabasca River is huge, discharging hundreds of cubic meters of water per second at peak flow, so without low detection limits the ability to detect any pollution is lost. To this end we have developed in-line solid-phase extraction (SPE) methodology, whereby 5 mL of natural water can be directly injected and thousands of bitumen-derived substances can be separated and detected in a single analysis (4).
Q. What is unique about the in-line solid phase extraction method that you and your team developed? What advantages does it offer over existing methods?
A: The in-line SPE method we developed is not unique, but the application of it to comprehensive profiling of complex environmental samples is rare. Today the most common water sample preparation method for oil sands monitoring is liquid-liquid extraction into a hydrophobic solvent, such as dichloromethane. More contemporary extraction methods, like SPE, have not gained traction in this field, I think because no commercial phases have been identified that seem to work well for all of the thousands of analytes present in a typical sample. In our hands, manual SPE does not work. Furthermore, because we are using full-scan non-targeted analysis, it became evident to us that most solvents and manual SPE cartridges were “dirty” and contributed a variable background signal to each sample that made profiling difficult. The in-line SPE method (using a C18 SPE with 12 μm particle diameter, 2 cm x 2.1 mm)recovered all the analytes of interest quantitatively (>90%) and did not affect the sample profile for thousands of analytes, as compared to direct analysis without SPE. The analyte recoveries were also more reproducible, there was less background contamination, and most importantly we save lots of time and do not produce so much waste solvents.
Q. What are you working on next?
A: Over the next three years we will be conducting large surveys of natural water around oil sands development to monitor for evidence of leaking tailings ponds. The challenge we face is that most, if not all, of the same organic compounds we see in tailings water are also present naturally in rivers and streams in the region. This is because the rivers have carved into the same bitumen deposits that are being mined, and natural groundwater that has been in contact with bitumen flows freely into the Athabasca River and its tributaries. We will try to use the inherent complexity of the organic compounds in each sample to our advantage, as the above analytical methods will be used to “fingerprint” the various anthropogenic and natural sources.
Q. Do you have any useful tips or advice for other analysts starting out in this field?
A: I believe that highly sensitive non-targeted analytical methods hold a great deal of promise to accelerate contaminant discovery in environmental analytical chemistry. The pairing of liquid chromatographic separations with ultrahigh resolution mass spectrometry allows unknown molecules to be separated and identified with reasonable accuracy at the empirical formula level. With subsequent MSn experiments, as is possible with Orbitrap technology, the structure of unknowns can then be fully or partially elucidated. Unfortunately this is also an expensive approach for someone that is just starting out, but the price gap between low-resolution and ultrahigh-resolution MS instruments appears to be narrowing.
Alberta Government, Alberta’s Oil Sands, http://www.oilsands.alberta.ca/resource.html
A.S. Pereira, S. Bhattacharjee, J.W. Martin, Environ. Sci. Technol. 47, 5504−5513 (2013).
M.S. Ross, A. dos Santos Pereira, J. Fennell, M. Davies, J. Johnson, L. Sliva, J.W. Martin, Environ. Sci. Technol. 46, 12796–805 (2012).
A.S. Pereira, and J.W. Martin, On-Line Solid Phase Extraction – HPLC – Orbitrap Mass Spectrometry for Screening and Quantifying Targeted and Non-Targeted Analytes in Oil Sands Process-Affected Water and Natural Waters in the Athabasca Oil Sands Region. OSRIN Report No. TR-45. 33 pp. http://hdl.handle.net/10402/era.37793 (2014).