CO2-based analytical methods are attractive for greener analysis, but fully integrated online multidimensional systems remain complex to implement and often require in-house adaptations, as no commercial instruments are currently available. Integrating multiple techniques such as SFE and SFC or SFC and SFC, requires special interfaces that can collect and transfer effluents at the junction between each technique. Connecting two techniques is already a challenge, but the complexity increases further as more techniques are coupled. While such an approach can save time and support an integrated “all-in-one” workflow, it also requires expertise in each technique as well as in coupling technologies. Without doubt, the most challenging part is thus interfacing. These methods are indeed very sensitive to solvent composition, transfer conditions, and injection effects. This can rapidly become problematic when targeting compounds with a broad polarity range, since their analysis require different solvent compositions. This makes it difficult to combine extraction and separation, or even two separation techniques, in one workflow, without sacrificing separation quality. Within the reporting period, the project produced useful results that go beyond a simple trial-and-error approach and help prepare the next stages of development.
First, a reliable SFC-MS method was developed for separating and identifying compounds with a wide polarity range. A major outcome was the detailed optimization of both MS ionization and chromatographic conditions, which helped identify practical pitfalls and key method-development rules for SFC-MS analysis of complex mixtures with a broad polarity range. In particular, the work showed that co-solvent composition and residence time during MRM optimization can strongly affect not only signal intensity but also peak shape, highlighting the importance of using conditions close to real SFC elution conditions.
An important result from the green chemistry side of the project was that ethanol, used as an alternative to methanol, gave comparable SFC-MS performance to methanol. Although methanol remained slightly better overall under the tested conditions, the differences were rather limited and compound-dependent, supporting ethanol as a realistic option in SFC-MS.
Preliminary offline SFE tests, including a first feasibility test on a real bamboo sample, gave a good starting point for the next stages of the project. Additional work is still needed to build a final method, especially because only part of the compound set was optimized at this stage, but this work already highlighted important practical issues for broad-polarity applications, especially contamination/carry-over and the need for appropriate pre- and post-cleaning procedures. Taken together, these results do not yet correspond to the final integrated online SFE-SFCxSFC-MS platform initially planned, but they provide important technical knowledge and a solid basis for future development. In particular, they reduce uncertainty for the next steps, especially for transfer optimization, interface design, and preserving separation quality during coupling. Further progress will require additional work on SFE method development, online transfer/interface optimization, and feasibility testing on the final integrated setup. Wider use of this approach would also benefit from comparison with more conventional step-by-step workflows to better evaluate its performance and added value.
