Quantification of coumarins, furocoumarins and polymethoxyflavones in hydroalcoholic fragrances by supercritical fluid chromatography‐tandem mass spectrometry

ABSTRACT Citrus essential oils, thanks to their pleasant aroma, are certainly the most used ingredients in the formulation of hydroalcoholic fragrances. The non-volatile fraction of Citrus essential oil is composed for 10–20% of coumarins, furocoumarins and polymethoxyflavones. It is well known that furocoumarins induce photosensitization and have potential carcinogenic and mutagenic effects. It follows that furocoumarins levels in cosmetics product are constantly monitored by opinions and regulations issued by the International Fragrance Association. The aim of this research article was to quantify coumarins, furocoumarins and polymethoxyflavones in thirty commercial hydroalcoholic fragrances using supercritical fluid chromatography in combination with triple quadrupole mass spectrometry technique (SFC-QqQ-MS). According to author’s knowledge, this is the first report on the determination of oxygen heterocyclic compounds in hydroalcoholic fragrances by means of SFC-QqQ-MS technique.


Introduction
The cosmetics industry, despite not providing essential products for human health, plays an important role in the global market.In particular, it is able to influence social life on a global level (1).There are many products commercialized by cosmetics industry, such as fragrances, creams, lotions, make up and so on.All of these products are composed of a variety of ingredient containing aroma, like discreet scents, floral freshness, sea smells and so on.
Citrus essential oils, thanks to their pleasant aroma, are certainly among the ingredients used in the formulation of fragrances (2).Cold-pressed Citrus essential oils are mainly constituted of volatile fraction (about 90%) and the remaining part of the non-volatile one.Oxygen heterocyclic compounds (OHCs), in particular coumarins, furocoumarins and polymethoxyflavones, represent the 10-20% of the non-volatile fraction (3).It is well known that, after exposure to UVA rays, furocoumarins can represent a risk to human health.In 2009, the European Parliament provided a Regulation (No. 1223/2009) on cosmetic products, in which the phototoxicity of furocoumarins is highlighted (4).For the protection of consumers' health, the International Fragrance Association (IFRA), within the amendments 48 and 49, released regulations on the use of essential oils in cosmetics with restriction limits of furocoumarins in the finished products (5,6).It is, therefore, necessary to quantify the presence of these molecules even at low concentration levels.For this reason, IFRA suggested an analytical method for the quantitative determination of fifteen furocoumarins (heraclenin, xanthotoxin, byakangelicol, psoralen, byakangelicin, 8-geranyloxypsoralen, bergapten, phellopterin, bergamottin, epoxybergamottin, isoimperatorin, imperatorin, isopimpinellin, oxypeucedanin and oxypeucedanin hydrate) by liquid chromatography coupled to a photodiode array detector (7).
A molecule largely used in fragrances as ingredient and fixative agent is coumarin (2 H-1-benzopyran-2-one) (8).According to Nelson et al (9).coumarin is in the top ten allergens causing allergic contact dermatitis.Therefore, considering that each person uses an impressive number of skin-care products daily and that these cosmetic are often flavoured with fragrances, a limitation in the use of coumarin was established by IFRA (10).Vocanson et al (11,12).reported that irritant properties of coumarin are due to the purity of molecule used.Authors asserted that the pure molecule is well tolerated, on the other hand the dermal sensitization and systemic toxicity can be attributed to coumarin impurities.In fact, they continue saying that irritant properties can be shown at a concentration of both 8% and in less than 0.3%.
In 2020 IFRA, recognizing allergic contact dermatitis of coumarin (amendment 49), limited the addition of coumarin in cosmetic products (10).In this amendment the leave-on cosmetics must not exceed in the coumarin content as follows: 0.0024% in baby creams, from 0.035 to 0.38% in lips and creams/lotions, and 1.5 % in fragrances.
Over the past thirty years, OHCs were analyzed in Citrus essential oils, foods, drugs, cosmetics, tobacco products, and toys using different analytical methods (3,(13)(14)(15).It can, however, be said that liquid chromatography coupled to a photodiode array detector is the most popular analytical technique used to determine OHCs (16).Recently, an innovative analytical approach for quantification of OHCs was the use of supercritical fluid chromatography coupled to triple quadrupole mass spectrometry detector (SFC-QqQ-MS) (17).Twenty-eight OHCs were quantified in thirteen Citrus cold-pressed essential oils in less than 8 min validating a green quality control method.
From a literature survey on the determination of coumarins and furocoumarins in cosmetics, several analytical techniques were employed for the quantification of OHCs in fragrances, creams, lotions, shampoos and so on (13).Yang and co-workers (18) proposed the use of a fluorescent sensor to determine coumarins in cosmetics; liquid chromatography coupled to mass spectrometry was employed to quantify furocoumarins (19) and coumarins (20) in different types of cosmetics.All these applications included a sample pre-treatment step before the analysis.On the other hand, OHCs determination in hydroalcoholic fragrances was carried out using both liquid (21) and gas (8) chromatography without any sample pretreatment.
SFC technique was already employed to quantify sunscreens (22), waxes (23), caffeine and UV-filters (24), provitamin B5 (25), and polyglycerol esters (26) in cosmetics.Recently, Li and co-workers proposed an SFC method to quantify eight furocoumarins in four cosmetics after sample extraction (27).To the best of author's knowledge, SFC was not used to investigate the OHCs content in hydroalcoholic fragrances without any sample pretreatment.The aim of this research was to quantify nine coumarins, fifteen furocoumarins and four polymethoxyflavones in thirty commercial hydroalcoholic fragrances using SFC-QqQ-MS instrument.According to author's knowledge, this is the first report on the determination of twenty-eight OHCs in hydroalcoholic fragrances by means of SFC-QqQ-MS technique.

Materials and samples
Thirty men's and women's commercial hydroalcoholic fragrances (eaux de parfum and eaux de toilette) are listed in Table 1.All the samples were directly injected without any pre-treatment.Each fragrance was analysed in triplicate.All samples were purchased in a local store.Between the samples analysed, twenty-three hydroalcoholic fragrances were flavoured with Citrus ingredients, according to top, heart and base notes description.
LC-MS grade 2-propanol and methanol (MeOH) were purchased from Merck Life Science and, 4.8 grade carbon dioxide (CO 2 ) was supplied by Rivoira.

SFC-QqQ-MS instrumentation
The SFC-QqQ-MS analyses were performed on a Shimadzu Nexera-UC system (Shimadzu, Kyoto, Japan).Instrumentation is composed by an LC-30ADSF CO 2 pump, two LC-20ADXR dual plunger parallel-flow pumps, a DGU-20A 5 R degasser, a SFC-30A backpressure regulator, a CBM-20A communication bus module, a CTO-20AC column oven, a SIL-30AC autosampler, and a LCMS-8050 triple quadrupole mass spectrometer equipped with an atmospheric pressure chemical ionization (APCI) source (the entire SFC flow was directed into the MS).The entire system was controlled by the Lab Solution ver.5.80.

SFC-QqQ-MS method
Separation of OHCs was performed according to a method previously validated (17).Briefly, analyses were carried out on an Ascentis Express F5 (150 × 2.1 mm, 2.7 µm) column.Mobile phase was composed of CO 2 (solvent A) and MeOH (solvent B).Analyses were carried out under gradient conditions: from 0 to 10 min increasing from 2% to 10% of B. Flow rate was 1.0 mL min −1 ; make-up pump solvent and flow rate were MeOH and 1.0 mL min −1, respectively.The injection volume was 2 µL.The oven temperature and the BPR (back pressure regulator) were set at 40°C and 120 bar, respectively.
The analyses were performed with the following mass parameters: DL temperature, 250°C, heat block temperature 200°C; interface temperature, 350°C; nebulizing gas flow (N 2 ) 3 L min −1 ; and drying gas flow (N 2 ) 5 L min −1 ; event time 0.024 s for each event; acquisition mode: multiple reaction monitoring (MRM); and acquisition time: 10 min for all targets.An automatic method, as included function of the software LabSolutions ver.5.80 and based on set retention times and Q transitions, was used for peak integration.
To quantify OHCs content in the thirty fragrance samples, calibration curves have been constructed using a mix of all the standards prepared from the stock solutions of each coumarin, furocoumarin and polymethoxyflavone at about 1000 mg L −1 concentration level.Method validation was carried out according to the method previously developed and validated by Arigò et al (17).Briefly, ten different concentrations of each standard material, in the range between 0.001 and 3 mg L −1, were analyzed five consecutive times by SFC-QqQ-MS (validation parameters are reported in Tables S1-2).

Results and discussions
The identification and quantification of coumarins, furocoumarins and polymethoxyflavones in thirty men's and women's commercial hydroalcoholic fragrances (eaux de parfum and eaux de toilette) were attained by means of SFC-QqQ-MS system.The identification of compounds was confirmed by comparison with standard materials.
The SFC-QqQ-MS method validated in our previous report (17) was used to quantify OHCs in the samples of our interest.The environmental-friendly analytical approach allowed a fast separation of twenty-eight OHCs in less than 8 min, with a low consumption of MeOH (10.5 mL per analysis).Moreover, thanks to the low limit of quantification (LoQ, in the range between 0.0015 and 0.1536 mg L −1 ), the quantification of OHCs was possible even when they were at trace level.All the hydroalcoholic fragrances were analysed without any sample pretreatment, eliminating errors resulting from sample preparation (sample loss and contamination), and increasing analytical precision.This aspect is of fundamental importance when an analytical method is developed in order to quantify molecules in trace level.
Tables 2-4 report the quantitative values (mgL −1 ± standard deviation) of coumarins, furocoumarins and polymethoxyflavones in the sample analyzed.As can be seen from Tables 2-4, OHCs in most of the samples analysed were contained in concentration under the detection and/or quantification limits.Table 4 shows that polymethoxyflavones were the most abundant OHCs class in the thirty men's and women's commercial hydroalcoholic fragrances analysed.Polymethoxyflavones total concentration ranged from LoQ to 19.05 mg L −1 .While, as reported in Tables (2 and 3), total furocoumarins and coumarins concentration did not exceed 5.49 and 5.37 mg L −1, respectively.
Between furocoumarins, bergamottin and 5-MOP were detected in most of the analysed samples.On the other hand, cnidilin, oxypeucedanin and byacangelicol were quantified in only one sample; cnidicin, psoralen and 8-geranyloxy-psoralen were reported in two samples.As it can be seen in Table 3, coumarin and citropten were quantified in most of the hydroalcoholic fragrances analysed.While epoxyaurapten was present in only one sample.Regarding polymethoxyflavones content, nobiletin and tangeretin were the most abundant molecules between OHC classes.From Tables 1-3, it can be seen that some hydroalcoholic fragrances (samples 8, 16, 17, 24, 25 and 29) do not contain any of the OHC classes investigated.These data agreed with the description of top, heart and base notes of the perfumes.In fact, all these samples do not contain Citrus ingredients.Only one sample (sample 10) between the non-Citrus flavoured hydroalcoholic fragrances showed the presence of sinensetin and nobiletin.This result suggests that the quantified polymethoxyflavones may come from an ingredient of a different nature.
The simultaneous presence of all the analysed OHCs classes in the hydroalcoholic fragrances analysed was detected in three samples (samples 13, 26 and 27).This finding could suggest the use of a mixture of cold-pressed Citrus essential oil or a single Citrus essential oil that normally contains the three classes (e.g.bergamot, grapefruit or bitter orange) (3). Figure 1 reports the SFC-QqQ-MS chromatograms of samples 13, 26 and 27.Epoxyaurapten, in sample 26, suggested the presence of grapefruit oil (3).Samples 13 and 27 showed the presence of 8-geranyloxy-psoralen, a furocoumarin present in cold-pressed lime and lemon essential oil.Epoxybergamottin content suggested the use of grapefruit or bitter orange essential oils.Through these associations, it is therefore possible to hypothesize which type of oil was used as ingredient.These associations are in good agreement with the description of top, heart and base notes of the perfumes.As reported by manufacturing company, in all these three perfumes bergamot, lemon and grapefruit were used as ingredient.
Quantitative data on furocoumarin content were in good agreement with a published research article on the quantification of 5-MOP, 8-MOP, bergamottin, epoxybergamottin, byakangelicol, isopimpinellin and oxypeucedanin in hydroalcoholic fragrances determination by LC-HRMS (21).
Table 5 reports the percentage composition of 5-MOP and coumarin in the nineteen hydroalcoholic fragrances that have shown their presence.From the bergapten content reported in Table 5, it is possible to see how all of the thirty commercial hydroalcoholic fragrances analysed showed concentration below 0.000087% for 5-MOP.These data completely fall within the IFRA restrictions (6) as analyte concentrations are far below the desired threshold of 0.0015% as reported in the amendment.As it can be seen from Table 4, the coumarin content in all of the commercial hydroalcoholic fragrances analysed ranged from 0.000003 to 0.00012%.These data are below the limit reported in the IFRA amendment (10), and with previous published results on coumarin content in fragrance products (8).
Considering furocoumarins and coumarin content, it can be asserted that all the thirty men's and women's commercial hydroalcoholic fragrances analysed in this research article are safe for human health.Thanks to the SFC-QqQ-MS method employed, it was possible to  characterize the OHCs content in all the samples in less than 8 min and MeOH to low concentration level (10.5 mL per analysis).This innovative analytical method can be applied for the analysis of OHCs in other types of cosmetics.

Conclusion
The environmental-friendly SFC-QqQ-MS analytical approach described in this work allowed a fast separation of twenty-eight OHCs in thirty men's and women's commercial hydroalcoholic fragrances.Characterization was achieved in less than 8 min, with a low consumption of MeOH.All the hydroalcoholic fragrances were analyzed without any sample pretreatment, eliminating errors resulting from sample preparation.
In accordance with both IFRA and European Parliament regulations, the SFC-QqQ/MS method guaranteed the quantification of each furocoumarins and coumarin under the concentration limits proposed.Coumarin was found in most of the sample analysed, because it is normally employed in cosmetic industry to fix aroma.Polymethoxyflavones were quantified in most of the samples suggesting the use of cold-pressed grapefruit, mandarin and sweet orange essential oils as ingredient.None of the analysed commercial hydroalcoholic fragrances showed furocoumarins concentration level higher than the limit set by IFRA regulation.*IFRA restriction for 5-MOP is 0.0015% (6); § According to IFRA regulation (10) coumarin content in fragrances must not exceed 1.5 %.