Greener SFC Method Detects Furocoumarins Fast

Coumarin and a related group of compounds called furocoumarins are natural substances found in plants like parsley, celery, fennel, and especially Citrus fruits (mainly in the peel). Coumarin has a sweet smell and is commonly used as a fragrance ingredient in perfumes and cosmetics. But both coumarin and furocoumarins have a downside: coumarin can cause skin irritation and allergic reactions, and can harm the liver if ingested, while furocoumarins can make skin extremely sensitive to sunlight, sometimes causing rashes or burns when skin exposed to these compounds is then subjected to ultraviolet (UV) light. Because of these risks, regulators in the European Union and the fragrance industry have set strict limits on how much of these compounds can be used in cosmetics, skincare, and sun-related products.

Given these safety concerns, it is important to have reliable ways to measure exactly how much coumarin and furocoumarins are present in products. Traditionally, scientists have used high-performance liquid chromatography (HPLC) paired with light-based detectors to do this. This technique works well and is accurate, but it tends to be slow and often requires using solvents that pose risk for the environment or worker’s safety. A more advanced version that pairs the same HPLC with mass spectrometry (MS) offers better accuracy and can measure many compounds at once, even in tiny amounts, but it is costly, energy-intensive, and has a bigger environmental footprint.

That's where the greener alternative of supercritical fluid chromatography (SFC) comes in. This technique uses far less harsh solvents, runs faster, and still delivers results on par with the older methods. Coumarin and furocoumarins are a great match for this approach, but only a handful of studies have explored using SFC for these compounds, and the methods that do exist tend to take longer than they need to.

With that gap in mind, a joint study conducted by the University of Messina and Chromaleont S.r.l. (Italy), together with Merck Life Science and Shimadzu Europa set out to build a faster, more environmentally friendly method using SFC to identify and measure sixteen furocoumarins that are subject to safety regulations, using Citrus essential oils as their test samples. They also included coumarin itself, since it is the chemical "parent" that furocoumarins are built from, and because doing so means the method could later be used to test cosmetic products too. To make the method as green as possible, they tested several bio-renewable or greener solvents (including ethanol, methanol, isopropanol, and dimethyl carbonate) to see which worked best. They also studied how each compound behaved during the process, to better understand and fine-tune the separation. Finally, they thoroughly tested the method and applied it to real Citrus essential oils from bergamot, bitter orange, grapefruit, and lemon.

LCGC International spoke to Cristian Reale, lead author of the paper resulting from this study,¹ about their work.

You chose supercritical fluid chromatography-photodiode array detection (SFC-PDA) over SFC-tandem mass spectrometry (MS/MS) despite legal limits being described as "relatively high." What sensitivity trade-offs did that involve, and how did you confirm your limits of quantification cleared thresholds like the 15-ppm IFRA limit for bergapten?

When choosing photodiode array detection over tandem mass spectrometry, the primary trade-off is sensitivity, other than selectivity. However, in the context of regulatory compliance for Citrus essential oils, MS/MS sensitivity is often far beyond what is practically required, while adding significant operational costs, instrument complexity, and higher environmental impact. We confirmed that our method easily clears required regulatory thresholds by validating it under the Eurachem guidelines. For bergapten, our method achieved a limit of quantification (LoQ) of approximately 0.13 ppm. Considering that the IFRA safety threshold for bergapten in leave-on cosmetic products is 15 ppm, our method’s LoQ is more than 100 times lower than the required limit.

You screened ethanol, methanol, isopropanol, and dimethyl carbonate as bio-based modifiers. What separation or peak-shape differences did you see between them, and what ultimately drove your final choice?

During modifier screening, we observed clear differences in retention. Retention increased in the order of methanol, ethanol, isopropanol, and dimethyl carbonate (DMC), suggesting decreasing elution strength. This directly impacted run times, delayed from 3 minutes with methanol to 25 minutes with DMC. While methanol and ethanol maintained consistent elution orders, isopropanol and DMC induced notable selectivity shifts. DMC, being aprotic and less polar, drastically increased the retention of dihydroxylated and epoxy furocoumarins. Byakangelicin even failed to be eluted under the used isocratic DMC conditions. Bio-based methanol was our final choice because it provided good balance between resolving power and elution strength, enabling the elution of all target compounds in less than four minutes.

Prior SFC methods for these compounds range from 7.5 to 15 minutes. How did your run time compare, and what specific optimizations (modifier gradient, flow rate, back-pressure) helped you improve on these benchmarks?

Our method reduced the total separation time from 7.5-15 minutes to 3.5 minutes. This speed was achieved through a combination of strategic parameters. First, adopting isocratic elution eliminated the additional run time for column re-equilibration required by gradient methods. Second, taking advantage of the low viscosity of supercritical carbon dioxide, we increased the flow rate to 4.0 mL min^-1 across two serially coupled PFP columns. Finally, backpressure setting was chosen to compromise between peak shape and separation.

Earlier high-performance liquid chromatography-ultraviolet (HPLC-UV) methods relied on toxic additives like acids or tetrahydrofuran to improve peak shape. Did your SFC-PDA method need additives to resolve critical pairs, and how did you keep those choices aligned with green chemistry principles?

Unlike earlier HPLC methods relying on hazardous additives like tetrahydrofuran or organic acids to address peak tailing, our SFC-PDA method avoided toxic modifiers entirely. Instead, we incorporated a small amount of water, 2.5% into the bio-methanol, providing multiple benefits: improved peak symmetry, slightly reduced retention, and enhanced selectivity for the most retained analytes. From a green chemistry standpoint, the method used a mobile phase composed of recycled CO₂ and only 6% bio-based methanol/water, bringing total organic solvent consumption down to just 840 microliters per run. This eliminated hazardous waste generation and earned the method a top Whiteness score of 88 on the WECA scale.

Coumarin doesn't naturally occur in Citrus essential oils but was included anyway. Did its retention behavior differ meaningfully from the furocoumarins, and did that require any gradient compromises to keep it well-resolved?

Including coumarin in our study ensured versatility for future applications in finished cosmetic products, where coumarin is a regulated fragrance allergen added to fix the fragrance. It is a simple benzo-alpha-pyrone that lacks both the fused furan ring and the bulky side-chain substituents present in furocoumarins. Consequently, it is less polar and sterically hindered, causing it to elute very early on the PFP stationary phase. Accordingly, it did not require gradient compromises or adjustments of the chromatographic parameters. It was fully resolved under the exact same fast, isocratic conditions used for essential oil screening.

Citrus oils are complex matrices with many co-extracted compounds. What matrix interferences came up during development, and how did you confirm PDA detection was specific enough to avoid misidentifying co-eluting compounds?

During method development, handling potential co-elutions was critical for accurate quantification, even more not relying on the high resolving power of MS/MS.We addressed this primarily through chromatographic efficiency. By serially coupling two 15-cm PFP columns, we boosted the theoretical plate count to over 44,000 and increased the peak capacity to 80. This afforded baseline resolution for critical pairs that typically co-elute in SFC, such as bergapten and isoimperatorin, or epoxybergamottin and byakangelicin. To confirm PDA specificity, we evaluated UV spectra from 200 to 400 nm against certified reference standards and literature profiles, while also analyzing distilled lemon oil as blank sample,to confirm the absence of baseline interferences at 310 nm.

What did your retention studies reveal about how furan-ring versus coumarin-ring oxygen positioning affects retention under supercritical CO₂ conditions, and how did that inform your column and modifier choice?

During method scouting, we tested a bare silica column. However, most analytes co-eluted within the first two minutes because this stationary phase relies almost exclusively on strong polar silanol interactions. In contrast, the pentafluorophenyl (PFP) column provided a better peak distribution across the chromatographic window. The superior selectivity of PFP phase could be explained considering its chemistry, as the fluorinated aromatic ring offers a combination of both polar and non-polar interaction mechanisms, including π-π interactions, dipole-dipole forces, and charge transfer. Through subsequent retention studies, we gained a deeper understanding of the interaction between furocoumarins with differing substituents and the PFP stationary phase. In particular, 8-substituted furocoumarins were less retained as their oxygen atoms are all oriented on the same side of the molecular framework, creating a net dipole moment and allowing stronger hydrogen bonding with the protic mobile phase. Differently, the oxygen atoms in 5-substituted furocoumarins are distributed on opposite sides of the nucleus, leading to partial cancellation of the dipole moment and likely imposing steric constraints that limit local interactions with the solvent, thereby enhancing retention.

Some SFC methods in the literature require 2.5–3 minutes of column re-equilibration per run. How did you handle reconditioning, and what was the impact on overall throughput compared to HPLC-UV and HPLC-MS/MS?

In our workflow, since separations are performed under isocratic conditions with 6% bio-based methanol/water in CO₂, column reconditioning is eliminated.This drastically improves the analytical throughput. With a total run time of 3.5 minutes and zero re-equilibration downtime, we can analyse approximately 17 samples per hour. Compared to conventional HPLC-PDA or SFC-MS/MS methods that often require more than 10 to 15 minutes per sample including reconditioning, our protocol nearly triples sample throughput while cutting operating costs down to roughly 5 cents per injection.

You applied the method to essential oils from bergamot, bitter orange, grapefruit, and lemon. Did any source show distinct furocoumarin profiles or concentrations that required adjusting your calibration range?

When applying the method to real samples, we observed distinct chemical profiles across the different Citrus species. Bergamot oil presented massive concentrations of bergamottin and bergapten, up to 1,800 and 12,000 mg L^-1, respectively. Lemon oil displayed the most complex profile, with up to ten different furocoumarins detected. In contrast, bitter orange and grapefruit oils contained much lower overall levels, dominated by epoxybergamottin.To accommodate these vast concentration differences without exceeding the linear response of the detector, we simply used different dilution ratios for the different essential oils analyzed. Additionally, we used a weighted linear regression model with a 1/y² factor, which effectively corrected for heteroscedasticity and ensured accurate quantification across the wide calibration range investigated.

Given MS/MS's superior selectivity but higher cost and environmental footprint, where do you see PDA-based SFC fitting into routine QC workflows, and are there cases where you'd still recommend confirmatory MS/MS?

In routine quality control workflows, PDA-based SFC is the ideal first-line screening tool. It offers low operational complexity, minimal maintenance, and an exceptionally small environmental footprint. For routine screening of raw material and to verify compliance with IFRA limits in essential oils, SFC-PDA delivers all the analytical rigor required at a fraction of the cost.That said, confirmatory MS/MS remains essential in specific scenarios. I would still recommend MS/MS when analyzing heavily diluted cosmetic formulations where target analytes approach trace levels below 0.01 mg/L, as well as when dealing with heavily adulterated matrices (where severe peak overlap may alter UV spectra).

Reference

1. Reale, C.; Cafeo, G.; Satira, A. et al. High-Throughput Analysis of Target Furocoumarins in Citrus Essential Oils by Supercritical Fluid Chromatography with Bio-Based Solvents. Green Anal Chem 2026, 18,100366. DOI: 10.1016/j.greeac.2026.100366