Entering Another Dimension: The Benefits of Multi‑2D LC×LC in Food Analysis

Lidia Montero and Miguel Herrero from the Institute of Food Science Research in Spain discuss the evolving role of comprehensive two-dimensional liquid chromatography (LC×LC) and multi-second dimension LC×LC (multi-²D LC×LC) in food analysis.

You recently published a paper called Multi‑²D LC×LC as a Novel and Powerful Implement for the Maximum Separation of Complex Samples (1). Firstly, can you define multi‑²D LC×LC and what benefits this approach offers compared to traditional 2D-LC techniques?

The main difference between multi-²D LC×LC and LC×LC is that the former provides superior separation power due to the use of two ²D columns with orthogonal separation characteristics. The configuration of these two ²D columns is performed by connecting an additional automatic switching valve (two-position six-port) that is connected in series after the modulator. This allows the automatic selection of one of the two available ²D columns during the analysis according to the chemical properties of the analytes eluted from the ¹D column.

The selection of the two ²D columns must provide complementary information, in a way that compounds that cannot be separated in the ¹D column nor in one of the two ²D columns are separated in the other ²D column.

Therefore, multi-²D LC×LC is a perfect tool to enhance the separation power of potentially correlated LC×LC couplings such as the coupling of two reversed-phase modes in both dimensions (RP×RP), where typically highly polar compounds cannot be separated in both dimensions; in this case, the coupling of an additional more polar ²D column, for example, hydrophilic-interaction chromatography (HILIC) can resolve the separation of these polar compounds by sending the first modulations of the analysis (where the high polar compounds are coeluted in the ¹D) to this ²D-HILIC column. Then, when compounds with lower polarity are eluted from the ¹D, the modulations of these peaks are directed to the ²D RP column for their complete separation.

In other words, multi-²D LC×LC allows tuning the ²D separation according to the chemical properties of the compounds present in the sample.

What types of complex samples is multi‑²D LC×LC suitable for?

In our group, we have focused our attention on the need for more powerful separation techniques to develop analytical methods able to provide the maximum compositional information for the evaluation of food and natural bioactive compounds, as well as in food quality and authenticity applications. Food matrices are clearly highly complex, formed by different families of compounds with a wide range of physicochemical properties.

The idea of multi-²D LC×LC arose after the effort of several years of optimization of conventional LC×LC methods for the analysis of complex food samples—in particular, the secondary metabolites of plant-based foods like the phenolic profiles of grape seeds, apples, berries, macroalgae or food byproducts, or secondary metabolite profiles involving different families, such as licorice (triterpene saponins and phenolic compounds) or industrial hemp (cannabinoids and phenolic compounds). So, we realized that on one hand, by coupling two non-correlated separation mechanisms in the ¹D and ²D separations, we achieved high orthogonal separation of the compounds, but at the cost of experiencing a loss of separation as a result of solvent mismatch between dimensions. On the other hand, using two correlated separation mechanisms (RP×RP, for instance), although we could achieve a relatively high orthogonality degree by applying shifting gradients, there were always compounds that could not be separated. Hence, with the development of multi-²D LC×LC, the two mentioned issues were solved.

Therefore, we can strongly recommend this multidimensional technique for profiling and chemically characterizing food samples where the main objective is to obtain the maximum information about their composition.

What were the main challenges you encountered and how did you overcome them?

The challenges that we have found during the development of multi-²D LC×LC methods were essentially the same as in conventional LC×LC, although with the great advantage that in multi-²D LC×LC, we do not have to fight with solvent incompatibility issues or the lack of separation affinity of compounds in the ²D, because these problems are solved by selecting the appropriate ²D column in each case.

2D-LC is often regarded as a complicated technique. Is this really the case? And if so, is multi-²D LC even more complicated?

Multi-²D LC×LC methods involve the same concerns as LC×LC in terms of instrumentation, 2D-LC knowledge, and method development. However, we have to consider that at the beginning of the multi-²D LC×LC method development, we will need to twice optimize LC×LC, once for each ²D column in order to establish the possibilities for each ²D column in the multi-²D setup.

Another thing that we have to consider is that currently, with the setup proposed in our recent paper (1) that uses just one binary pump for performing the ²D separation, the method is limited to the use of the same mobile phases for both ²D columns.

In our application using PFP × HILIC/C18 this was not a problem because both HILIC and C18 multi-²D gradients were performed using water and acetonitrile as mobile phases, just in the opposite elution order. However, if ²D columns that need different mobile phase setups are coupled, a ²D quaternary pump could be used or, in the worst case, an additional ²D pump could be installed to perform individual gradients in each ²D column.

So, in summary, we can conclude that multi-²D LC may be regarded as complex as 2D-LC, although not necessarily more.

In what other areas of food analysis could this technique be used?

Our next step is to take advantage of this powerful technique for maximizing the information about the chemical composition of high-quality foods, to evaluate their authenticity and be able to perform the evaluation of potential food fraud in cases of foods like those with differentiated quality, including protected designation of origin (PDO) or geographical indication (GI), as well as for the evaluation of potential food frauds in organic foods.

Could this technique be useful in other areas outside food analysis?

Absolutely! Multi-²D LC×LC is a versatile technique that can be applied to every field where LC×LC has shown its power to solve separation challenges that conventional 1D-LC methods cannot address, such as applications in pharma (analysis of antibodies), clinical (analysis of biological complex samples such as urine, blood, or tissues), or environmental (pesticides, polycyclic aromatic hydrocarbons [PAHs]) analyses, among many others.

Is traditional 2D-LC being used more often in routine food analysis? Are there any technology drivers that will enable 2D-LC to be used more often in the future in food analysis?

2D-LC is indeed attracting great interest in the LC community. A good sign of this is the growing number of analytical applications that are using 2D-LC methods, and the presence and organization of several sessions dedicated to multidimensional LC in worldwide recognized conferences, such as the last HPLC 2023 in Dusseldorf, Germany. We enjoyed three sessions exclusively focused on multidimensional LC, together with a roundtable that brought together some of the greatest 2D-LC experts. However, in our opinion, we are far from using 2D-LC as a routine alternative, not only due to the need for more expensive instrumentation, but also due to the complex method development, problems with detection limits, and limitations in the data treatment of the 2D data matrix that is generated.

Currently, a big effort is being made to develop computer-aided method developments for LC×LC, to predict and provide information about LC×LC parameters such as column selection, optimal modulation periods, column dimensions (length, particle size, and internal diameter), or gradient programs, column temperature, and mobile phases to reduce the currently time-consuming method development (2–8).

Concerning 2D data treatment, currently, the lack of software dedicated to 2D data treatment is a bottleneck that limits or hampers the potential of comprehensive two-dimensional chromatography (both LC×LC and GC×GC) in trending analytical fields, like -omics. So, the situation is that we have a very powerful analytical technique that provides a huge amount of data but we cannot easily use and process the information generated. Although some articles have dealt with this problem and report important advances in this field (9,10), in the future, the development of computational tools that allow taking advantage of all the powerful information provided by multidimensional LC will increase the use of multidimensional techniques. Besides, automatic data treatment will allow the possibility of coupling even more powerful setups, such as the coupling of LC×LC to ion mobility–mass spectrometry (IM-MS), providing additional separation dimensions.

What is the future of 2D-LC?

Recently, there have been great theoretical LC×LC advances and developments, and we have also seen many innovations in the setup configuration, like multi-²D LC×LC or new modulation strategies, showing that LC×LC is a versatile technique that is constantly evolving and being improved. Although more effort is needed to address the limits mentioned previously, it is time to apply all this knowledge that we are accumulating to include LC×LC as part of our research projects and industrial applications, more than treating LC×LC as an individual technique in the developing stage.

Is there anyone who has inspired you for this research project?

We would like to express sincere thanks to the Applied Analytical Department from the University of Duisburg-Essen, where the multi-²D LC×LC configuration was developed, in particular, special thanks to Prof. Oliver Schmitz for his magnificent support and his always positive attitude towards new analytical developments.

References

(1) Montero, L.; Ayala-Cabrera, J. F.; Bristy, F. F.; Schmitz, O. J. Multi-2D LC×LC as a Novel and Powerful Implement for the Maximum Separation of Complex Samples. Anal. Chem. 2023, 95 (6), 3398–3405. DOI: 10.1021/acs.analchem.2c04870

(2) Haidar Ahmad, I. A.; Makey, D. M.; Wang, H.; et al. In Silico Multifactorial Modeling for Streamlined Development and Optimization of Two-Dimensional Liquid Chromatography. Anal. Chem. 2021, 93 (33), 11532–11539. DOI: 10.1021/acs.analchem.1c01970

(3) Vivó-Truyols, G.; van der Wal, S. J.; Schoenmakers, P. J. Comprehensive Study on the Optimization of Online Two-Dimensional Liquid Chromatographic Systems Considering Losses in Theoretical Peak Capacity in First- and Second Dimensions: A Pareto-Optimality Approach. Anal. Chem. 2010, 82 (20), 8525–8536. DOI: 10.1021/ac101420f

(4) Pirok, B. W. J.; Pous-Torres, S.; Ortiz-Bolsico, C.; Vivó-Truyols, G; Schoenmakers, P. J. Program for the Interpretive Optimization of Two-Dimensional Resolution. J. Chromatogr. A. 2016, 1450, 29–37. DOI: 10.1016/j.chroma.2016.04.061

(5) Andrighetto, L. M.; Stevenson, P. G.; Pearson, J. R.; Henderson, L. C; Conlan, X. A. DryLab® Optimised Two-Dimensional High Perfomance Liquid Chromatography for Differentiation of Ephedrine and Pseudoephedrine-Based Methamphetamine Samples. Forensic Sci. Int. 2014, 244, 302–305. DOI: 10.1016/j.forsciint.2014.09.018

(6) Andrighetto, L. M.; Burns, N. K.; Stevenson, P. G.; et al. In-Silico Optimisation of Two-Dimensional High Performance Liquid Chromatography for the Determination of Australian Methamphetamine Seizure Samples. Forensic Sci. Int. 2016, 266, 511–516. DOI: 10.1016/j.forsciint.2016.07.016

(7) Lindsey, R. K.; Eggimann, B. L.; Stoll, D. R.; Carr, P. W.; Schure, M. R.; Siepmann, J. I. Column Selection for Comprehensive Two-Dimensional Liquid Chromatography using the Hydrophobic Subtraction Model. J. Chromatogr. A. 2019, 1589, 47–55. DOI: 10.1016/j.chroma.2018.09.018

(8) Snyder, L. R.; Dolan, J. W; Carr, P. W. The Hydrophobic-Subtraction Model of Reversed-Phase Column Selectivity. J. Chromatogr. A 2004, 1060 (1–2), 77–116. DOI: 10.1016/j.chroma.2004.08.121

(9) Navarro-Reig, M.; Jaumot, J.; Tauler, R. An Untargeted Lipidomic Strategy Combining Comprehensive Two-Dimensional Liquid Chromatography and Chemometric Analysis. J. Chromatogr. A 2018, 1568, 80–90. DOI: 10.1016/j.chroma.2018.07.017

(10) Montero, L.; Meckelmann, S. W.; Kim, H.; Ayala- Cabrera, J. F.; Schmitz, O. J. Differentiation of Industrial Hemp Strains by Their Cannabinoid and Phenolic Compounds Using LC×LC–HRMS. Anal. Bioanal. Chem. 2018, 414, 5445–5459. DOI: 10.1007/s00216-022-03925-8

Miguel Herrero, PhD, is a tenured researcher at the Institute of Food Science Research (CIAL) belonging to the Spanish National Research Council (CSIC), in Madrid (Spain). He holds a PhD in Food Science and Technology from the University Autonoma of Madrid (2006) and carried out a two-year postdoc stage at the University of Messina (Italy). His main research interests are the study and characterization of new functional ingredients including the development of new advanced extraction and analytical methods to obtain and characterize interesting food-related compounds. Particularly, on the development of new methods and applications using LC×LC coupled to mass spectrometry.

Lidia Montero, PhD is a post-doctoral researcher under the “Ramón y Cajal” program at the Institute of Food Science Research (CIAL) belonging to the Spanish National Research Council (CSIC). From 2018 to 2022 she was a postdoctoral researcher in the Applied Analytical Chemistry Department of the University of Duisburg‐Essen in Germany. Her main research interests have been the development of 2D-LC methods and setups for food quality, authenticity, and bioactivity analysis as well as the application of emergent extraction techniques like sub‐and supercritical extraction processes to obtain food ingredients of high‐added value from food by‐products.