Entering the Second Dimension

LCGC Europe spoke to Alina Muscalu from the Organic Contaminants Section at Ontario Ministry of the Environment, Conservation and Parks (MECP) in Toronto, Canada, about the practical advances in environmental analysis using comprehensive two-dimensional gas chromatography (GC×GC) for targeted and nontargeted analysis, the role of GC×GC in routine analysis, and the benefits of using a micro-electron capture detector (μECD) for environmental analysis.

Q. The number of publications citing the use of comprehensive two-dimensional gas chromatography (GC×GC) has significantly increased in recent years. What technological advances have led to the technique being more commonly used?

A: The analysis of persistent organic pollutants (POPs) in environmental matrices is very challenging because of the large number of compounds with different chemical and physical properties that are typically present in the sample at concentrations ranging from ultratrace to percentage levels. As the legislative requirements for environmental testing continue to grow more challenging, there is a continuous need to develop analytical methods that allow fast and reliable detection, with better sensitivity and resolution for complex matrices.

The use of GC×GC hyphenated to universal or selective detectors for environmental analysis proved to have an outstanding potential for the development of multiresidue methods. The latest advances in mass spectrometry (MS) technology made the combination of GC×GC hyphenated with different MS systems, such as quadrupole-, high-resolution (HR)- time-of-flight (TOF)-, triple quadrupole (QqQ)-, and QTOF-MS, possible, adding another analytical dimension for separation, and making the technique more accessible.

Various GC×GC systems are now available from several vendors and operate using two primary techniques, either flow modulation or thermal modulation. Considerable development in modulation technology has occurred over the last 20 years, making this technique more accessible (1). To help with data interpretation, new software tools, such as Kendrick mass defect plots or automatic peak identification algorithms, have been developed and applied to nontarget analysis for both identification and quantitation. Data processing software packages are now available from more vendors than before, developed to ease the integration and quantitation processes when handling the large amount of data generated by GC×GC.

Q. What benefits does comprehensive two-dimensional chromatography offer separation scientists involved in environmental analysis?

A: The introduction of GC×GC provided a welcome relief to environmental testing because of the significant increase in the separating power and peak capacity offered by the technique, as well as the increased sensitivity. The selectivity is increased by applying two independent separation mechanisms to a sample in a single analysis, resulting in improved resolution of target compounds from structurally similar compounds and matrix interferences. The sample preparation steps can typically be simplified when using GC×GC, for example, fractionation of the extracts is no longer needed because multiple classes of contaminants can be detected in a single analysis. The increased resolving power of GC×GC and the generation of structured chromatograms allows more economical detectors such as electron capture detectors (ECDs) to be used routinely with increased confidence. For routine analysis, this technique results in savings in instrument time and overall speed of the analysis, and a reduction of chemicals used and costs.

As the number of chemicals present in the samples is continuously growing, the use of GC×GC offers more information on the composition of different groups of xenobiotics and an in-depth characterization of different components in a complex mixture. In addition, the enhanced chromatographic resolution of GC×GC enables scientists to provide fingerprinting information in environmental forensics studies to identify minor compounds that would normally be missed with the use of a one dimensional (1D-GC) technique, and to improve data accuracy, for example, by eliminating overestimation (2). As well as the target compound analysis, GC×GC coupled with TOF-MS, HRMS, MS, or other suitable detectors is able to detect and identify compounds that might be environmentally relevant but are not routinely analyzed, such as benzotriazoles (2) and halogenated benzotriazoles (3), which are both classified as emerging contaminants. GC×GC has proved to be of significant benefit owing to both its quantitation and screening capabilities. Thus, the development of nontargeted, multiresidue methods combined with statistical data interpretation to identify and prioritize emerging contaminants of concern in different matrices was also possible with this technique. In my view, there is no single analytical technique or method that can separate and detect all the contaminants present in the samples; therefore, to fully characterize environmental samples, GC×GC should be combined with complementary techniques, such as liquid chromatography (LC) or other hyphenated techniques.

Q. Can you elaborate on the practical application of comprehensive two-dimensional gas chromatography for target compound analysis in important areas of environmental analysis?A: GC×GC has been successfully used for qualitative determination and accurate quantitation of target compounds in different environmental matrices. In a review of GC×GC published recently, we highlighted the advances and the advantages of using this technique for targeted and nontargeted analysis in complex environmental matrices (4). The paper summarized important applications over the past 10 years, providing both a critical review and a short description of the applications: sample preparation, instrumentation, data interpretation approaches, the different types of matrices tested, and different classes of contaminants that have been analyzed, as well as applications in quality assurance (QA) and quality control (QC) for routine analyses.

Target compound analyses require method optimization QA/QC procedures to be implemented to ensure that the compounds are quantitatively extracted from the matrix; the background interferences can be effectively separated from the targets and the data produced are accurate.

GC×GC was applied successfully in many studies for the detection of contaminants, such as pesticides, pharmaceuticals, disinfection by-products, hydrocarbons, and synthetic musks, in water, biota, air, and solid samples, particularly in drinking water samples where concentration limits are set by regulations (4). For example, extracts from various water samples (river, wastewater, and groundwater) were analyzed for synthetic musks by GC×GC–TOF-MS and statistical tools were used to assess the similarities between different water sources. Finding these compounds in water sources used for drinking water was important and the researchers concluded that the bioaccumulation factors should not be neglected, even though these compounds were present at low concentrations (5).

In a different example, GC×GC was used in biodegradation studies where the changes in distribution and abundance of the target compounds were monitored using GC×GC and GC, which helped to evaluate the efficiency of the bioremediation processes (6). GC×GC–TOF-MS (7) and GC×GC–micro-electron capture detector (µECD) (8) were used as a less expensive alternative to GC–HRMS for the qualitative and quantitative analyses of polychlorinated dioxins and furans in different environmental matrices, as well as for forensic investigations. Many practical applications are already published, proving once more that using a technique as powerful as GC×GC for the determination of historical and emerging contaminants of concern (routinely or non-routinely monitored) in environmental samples has significant benefits owing to both its quantitation and screening capabilities.

Q. Where is this technique being used for nontargeted screening in environmental analysis?

A: With thousands of chemicals used daily, the need for multiresidue methods is increasing. There is always the need to identify potential new contaminants present in the environment that might have biological effects while also having sufficient sensitive methods to quantitate the target compounds. Monitoring these hazardous chemicals in environmental samples is difficult because the samples are very complex and sufficiently sensitive validated methods that are applicable under routine laboratory conditions are often not available.

Screening multiresidue approaches were implemented to prioritize potentially persistent, bioaccumulating, and toxic contaminants, to provide information related to the presence of nontargets in environmental forensic investigations and preliminary risk assessment of organic contaminants in different environmental matrices.

Different approaches were used for GC×GC analyses; for example, nontargeted analysis was first used to detect the analytes and then selected environmentally relevant contaminants were analyzed by target approaches. GC×GC analyses provided more information on the presence and content of different groups of contaminants, provided additional selectivity for more accurate detection, and helped with compound class visualization and identification of minor compounds in complex environmental samples.

Modern applications in the environmental field emphasized the advances for nontargeted analysis in complex matrices and helped in contamination source tracking, risk assessment, or improving different treatment processes. For example, the data collected from GC×GC screening of organic contaminants in surface water, soil, and sediment collected from a known contaminated industrial site were used to predict the toxicity of the compounds found and to provide an initial assessment of the potential environmental impact (9). In another example, by using both targeted and nontargeted GC×GC approaches for the analysis of fish oil samples, multiple classes of persistent organic pollutants were detected along with other untargeted chemicals. This was an important finding as detecting these contaminants in dietary fish oil suggested that conventional treatment processes of fish oil were ineffective in removal of the heavier organic contaminants (10). Along with monitoring target analytes, GC×GC coupled with TOF-MS, HRMS, or other MS detectors is more and more required in environmental analysis to detect and identify compounds that might be environmentally significant but are not routinely analyzed.

Q. You developed a method using GC×GC with a μECD detector. Can you tell us more about this project?

A: Since 2010, we have been using GC×GC–µECD for routine analysis of polychlorinated biphenyls, organochlorine pesticides and herbicides, and trifluralin in different matrices: solid samples (sediments, soil, sludge) (11–12), biota (fish, clams, and mussels) (13), water (effluent, ground water, surface water, and drinking water), and passive samplers (14). Quality control and quality assurance procedures-implemented as outlined in the international standard (ISO 17025) requirements-showed that the method performs very well. 

Our laboratory has implemented three fully validated GC×GC–µECD methods-two of them being accredited to ISO17025 standard-which are successfully applied to “real‑life” samples.

The GC×GC methods also participate in proficiency testing programmes provided by external organizations to satisfy accreditation requirements. In addition to GC×GC–µECD, we use GC×GC–TOF-MS as a complementary technique to further confirm the target compounds or identify and quantify the “unknown” halogenated organic contaminants. One of the challenges when developing the method was selecting the best column combination to achieve the within- and between-class separation of the 118 target compounds, as well as their separation from the matrix interferences. The current column setup includes a nonpolar column coupled to a shape-selective column and a transfer line that connects the 2D column from the secondary oven to the detector. As the main oven was set at a lower temperature than the secondary oven, this narrower transfer line helped to preserve the separation. Wraparound was also observed in the 2D chromatograms; however, it did not affect the separation nor the quantitation of the targets. Data processing can be difficult at times because the matrix effects resulted in peak shifting in the second dimension plan. To ease the data interpretation process, retention time reference compounds specific to each class of target contaminants were added to account for any shifts. Manual integration is still required and data interpretation could be more time-consuming when negative peaks are observed as a result of the high hydrocarbon levels in the sample.

Q. What was novel about your approach and what were your main findings?

A: By implementing these methods for routine work, one GC×GC–µECD instrument replaced four different GC–ECD systems and one GC–MS system that were using seven column phases to analyze multiple fractions of each sample extract. The method uses the separation power of GC×GC while reducing the sample preparation procedures because no fractionation of extracts is required prior to instrumental analysis, which resulted in significant savings in time and analysis costs, as well as savings in labour, equipment, bench space, and reduced solvent requirements, yielding improved health and safety conditions and definitive enhancement of data quality.

Additionally, by using a technique such as GC×GC–µECD for routine analysis, the identification of other nontargeted halogenated contaminants present in the environmental samples became possible. The GC×GC–µECD method combined with GC×GC–TOF-MS for further characterization proved to have many benefits, including identification of nontargeted compounds and the ability to re-direct the samples to validated methods for the identified compounds when available, for example, polybrominated diphenyl ethers, polychlorinated naphthalenes; tracking down the source of contamination based on Aroclor matching; identifying the source of DDT contamination when dicofol was detected; or identifying short‑chain chlorinated paraffins (SCCPs) as a ubiquitous contaminant class in sediment samples.

The advantage of using GC×GC–µECD for routine environmental analysis is its screening power: 2D chromatograms of samples analyzed routinely for other halogenated compounds, such as PCBs, could be sorted and data further processed when SCCPs were present. Based on these findings, we have further developed a quantitative method for the analysis of SCCPs in sediment samples by using GC×GC–µECD (15).

The method was successfully applied to both “real-life” samples and inter-laboratory study samples, performing very well. A few drawbacks of this procedure should be noted: SCCPs could not be fully separated from medium‑chain chlorinated paraffins (MCCPs) (quantitation was not possible if MCCPs were present) and data processing was very time-consuming because of software limitations.

Q. The technique has a reputation for being very complicated to use. Could the technique be used for routine analysis in the environmental sector, for example, in QC or QA?

A: The advantages of using GC×GC over one-dimensional GC for routine analysis has been proved and published over the years. In our laboratory, we have successfully implemented and used this technique for over eight years, proving again that routine analysis is possible.

Nevertheless, the technique is not commonly implemented for routine analysis. In my opinion, GC×GC is just an extension of 1D GC. Users who understand GC should have very few problems switching to GC×GC. In practice, when using standardized methods, there is no difference between using a standard GC system versus a GC×GC system: the analyst injects the sample in the same way, and then the analysis begins. Having said that, there are some differences, especially when one wants to develop their own methods. Since two columns are used simultaneously, there are numerous combinations that could be tested. Also, because the columns are connected in a single train, many parameters are interrelated. Understanding those relationships is crucial for successful method development.

Finally, data interpretation is more complicated. These features might seem intimidating to many potential users, and there is some reluctance to adopting GC×GC as a routine technique in the environmental sector. In my opinion, the only solution to this is continuous education. GC×GC courses are offered at several conferences, such as ISCC and the GC×GC symposium. In addition, most manufacturers offer user training for their instruments. At the end of the day, it is the job of the early adopters to demonstrate the advantages of GC×GC in a convincing enough fashion to sway the opinions of potential users who may be interested in the technique but are hesitant. We are not there yet, but I believe this day will come.

Based on my experience for the last eight years in analyzing environmental samples routinely, this technique is very robust and easy to use once the method is in place. Strategies for handling the large amount of data are published and are not very difficult to implement for routine analysis.

References

1. P.Q. Tranchida, J. Chromatogr. A1536, 2–5 (2018).
2. E. Jover, V. Matamoros, and J.M. Bayona, J. Chromatogr. A1216, 4013–4019 (2009).
3. S. Prebihalo, A. Brockman, J. Cochran, and F.L. Dorman, J. Chromatogr. A1419, 109–115 (2015).
4. A.M. Muscalu and T. Gorecki, TrAC, Trends in Analytical Chemistry106, 225–245 (2018).
5. D. Relić, A. Popović, D. Đorđević, and J. Čáslavský, Environ. Earth Sci.76, 122 (2017).
6. V.P. Beškoski, S. Miletić, M. Ilić, G. Gojgić-Cvijović, P. Papić, N. Marić, T. Šolević-Knudsen, B.S. Jovančićević, T. Nakano, and M.M. Vrvić, Clean Soil Air Water 45(2), 1600023.
7. J. de Vos, P. Gorst-Allman, and E. Rohwer, J. Chromatogr. A1218, 3282–3290 (2011).
8. C. Rimayi, L. Chimuka, D. Odusanya, J. de Boer, and J. Weiss, Chemosphere145, 314–321 (2016).
9. P. Bastos and P. Haglund, J. Soils Sediments12, 1079–1088 (2012).
10. E. Hoh, S.J. Lehotay, K.C. Pangallo, K. Mastovska, H.L. Ngo, C.M. Reddy, and W. Vetter, J. Agric. Food Chem.57, 2653–2660 (2009).
11. A. Muscalu, E. Reiner, S. Liss, T. Chen, G. Ladwig, and D. Morse, Anal. Bioanal. Chem.401, 2403–2413 (2011).
12. A.M. Muscalu, The Determination of Polychlorinated Biphenyl Congeners (PCBc), Organochlorines (OCs) and Chlorobenzenes (CBs) in Solid Samples by GC×GC-μECD. Method E3487 (Ministry of the Environment, Conservation and Parks, Laboratory Services Branch, Organic Contaminants Section, Toronto, Ontario, Canada, 2010).
13. A.M. Muscalu, The Determination of Polychlorinated Biphenyls (PCBs) in Biota Samples by GC×GC-μECD. Method E3485 (Ministry of the Environment, Conservation and Parks, Laboratory Services Branch, Organic Contaminants Section, Toronto, Ontario, Canada, 2017).
14. A.M. Muscalu, The Determination of Polychlorinated Biphenyl Congeners (PCBc), Organochlorines (OCs), Chlorobenzenes (CBs) and Trifluralin in Water Samples by GC×GC-μECD. Method E3488 (Ministry of the Environment, Conservation and Parks, Laboratory Services Branch, Organic Contaminants Section, Toronto, Ontario, Canada, 2015).
15. A.M. Muscalu, D. Morse, E.J. Reiner, and T. Górecki, Anal. Bioanal. Chem.409, 2065–2074 (2016).

Alina Muscalu is a development scientist with the Organic Contaminants Section at Ontario Ministry of the Environment, Conservation and Parks (MECP) in Toronto, Canada. She joined MECP in 2003 to perform routine analyses and research for the determination of the halogenated organic pollutants in different environmental matrices. Alina achieved a bachelor’s degree in biochemistry and chemistry in 1997 from University of Bucharest, Romania, a master’s degree in environmental applied science and management in 2009 from Ryerson University, Toronto, Canada, and is currently a PhD candidate in analytical chemistry (Prof. Górecki’s group) at the University of Waterloo, Canada.