A Sustainable CE-DAD Screening Method for Multi-Class Polyphenol Profiling in Rosehip-Based Herbal Tea and Supplement

TL;DR

This paper introduces a sustainable and efficient method for analyzing polyphenols in rosehip-based products, enabling accurate quality control and health benefit evaluation.

Contribution

A green CE-DAD method for multi-class polyphenol profiling in rosehip products is developed and validated.

Findings

The CE-DAD method showed excellent linearity (R2 > 0.99) and reproducibility (RSD 0.8–1.6%).

Eight target polyphenols were identified and quantified in real samples.

Herbal tea had higher antioxidant activity compared to the food supplement in DPPH and ABTS assays.

Abstract

Rosehip-based products are rich in polyphenols with recognized health benefits, making accurate characterization essential for quality control and functional evaluation. Conventional analytical approaches for polyphenol determination are often time-consuming, costly, and environmentally demanding. In this study, a sustainable analytical method based on capillary zone electrophoresis coupled with diode array detection (CE–DAD) was developed as a green and accessible screening method for polyphenol analysis in rosehip-based products. Twelve polyphenolic compounds belonging to different classes (stilbenes, flavonols, flavanols, flavanones, and flavones) were used to optimize the electrophoretic conditions, including the buffer pH, voltage, and electrolyte concentration. Herbal tea and supplement samples were analyzed before and after a simple cartridge-based clean-up step to reduce matrix interferences. The method enabled simultaneous profiling of multiple polyphenol classes in a single CE–DAD run, showing excellent linearity (R2 > 0.99), run to run reproducibility (RSD 0.8–1.6%), and sensitivity (LOD 0.4–1.4 μg/mL; LOQ 0.9–4.7 μg/mL). Eight target polyphenols were identified and quantified in real samples. Polyphenol profiling was complemented by DPPH and ABTS antioxidant assays, allowing a functional interpretation of compositional data and a case-based comparison between different product formulations. Specifically, the herbal tea showed the values of a 13.8 mg Trolox/g sample (80.5% DPPH inhibition) and 15.3 mg Trolox/g sample (98.5% ABTS inhibition), whereas the food supplement presented a 7.4 mg Trolox/g sample (34.2% DPPH inhibition) and 7.4 mg Trolox/g sample (54.1% ABTS inhibition). Method sustainability and applicability were also evaluated using the Blue Applicability Grade Index (BAGI), confirming a low environmental footprint.

Figures

Full text

1. Introduction

Profiling phytochemicals in plants, foods, beverages, and supplements is essential to assess their quality and health properties, with implications for sustainable health and therapeutic use [1,2]. Compounds like polyphenols, glucosinolates, and carotenoids are studied for their antioxidant, anti-inflammatory, and protective effects against chronic diseases [3,4], and they also influence color, aroma, and taste. Polyphenols comprise a large and heterogeneous group of secondary metabolites, including phenolic acids, flavonoids, lignans, and stilbenes, with flavonoids representing the most abundant and structurally diverse subclass [5]. As reported in the literature, polyphenols are widely present in foods and beverages [6,7], and in plants such as Rosa canina, commonly known as rosehip, which contains high levels of phenolic acids and flavonoids [8,9,10,11].

Rosehip extracts are used in teas and supplements for their antioxidant and anti-inflammatory properties [12,13,14]. Polyphenol solubility is strongly influenced by molecular polarity and interactions with the sample matrix, which can promote the formation of insoluble complexes; therefore, extraction protocols must be carefully tailored to each polyphenol subclass and specific matrix. Typically, analytical workflows involve homogenization followed by liquid–liquid or solid-phase extraction [15,16] and subsequent chromatographic analysis, most commonly HPLC–UV or HPLC–MS/MS [17,18]. However, mass spectrometric techniques are not always accessible or necessary for routine or preliminary analyses. Over the past two decades, capillary electrophoresis (CE) has emerged as a greener alternative or complementary technique to HPLC, owing to its high efficiency, rapid analysis, and reduced solvent consumption, making it well-suited for phytochemical analysis [19,20]. Although flavonoid determination in food matrices is well-established, capillary zone electrophoresis (CZE) remains underused for multi-class polyphenol profiling in nutraceutical matrices such as herbal teas and food supplements, with limited scientific evidence available for their determination in these products [21,22]. This limited adoption is primarily due to the inherent analytical challenges of resolving structurally diverse polyphenols in complex matrices, the relatively lower sensitivity of UV–Vis detection compared to mass spectrometry, and the requirement for careful optimization of separation parameters (e.g., buffer composition, pH, and applied voltage) to achieve reproducible migration times for multiple compound classes. Previous studies have mainly focused on specific compound groups, using the optimization of CE separations coupled with mass spectrometric detection [23], or on targeted analyses of flavonoids in tea samples [24,25]. A comprehensive review published in 2023 summarized the application of CE in phytochemical analysis [20]; however, it primarily addressed different compound classes and sample types. Capillary electrophoresis is a well-consolidated technique for the separation of anthocyanidins, owing to their cationic nature under strongly acidic conditions and the high separation efficiency of the method [26].

On the other hand, capillary electrophoresis is also a well-established and versatile technique for the separation and analysis of phenolic acids, taking advantage of their ionization under alkaline conditions [27].

In line with our previous research on flavonoids in natural matrices [25], this work focuses on the separation of twelve aglycone and glycoside polyphenols from five classes—stilbenes (resveratrol, astringin, piceatannol), flavonols (quercetin, rutin, isoquercitrin, quercitrin, kaempferol), flavanols (catechin, epicatechin), flavanones (hesperetin), and flavones (luteolin)—using capillary electrophoresis with diode array detection (CE–DAD) as a practical alternative for routine analysis where mass spectrometric detection is not required.

The analytes selected for this study were chosen based on their expected abundance in the analyzed matrices. Anthocyanidins and phenolic acids that were extensively studied in the literature [26,27] were not included here because their pH and distinct UV–Vis absorption require dedicated extraction and separation conditions, which would complicate simultaneous CE–DAD analysis of the selected polyphenol classes.

Hence, the aim of this study was to develop and validate a green, simple, rapid, and sufficiently sensitive CE–DAD method for the simultaneous analysis of five polyphenol classes, which have not previously been analyzed together using this technique. The inclusion of both glycosides and aglycones is crucial, since glycosylation modulates polyphenol occurrence, stability, bioactivity, and nutritional relevance, providing a more complete and physiologically meaningful profile of their functional properties in nutraceutical products. The optimized method, combined with an eco-friendly sample clean-up strategy that aimed to reduce matrix interferences, was applied to rosehip-based herbal teas and food supplements obtained from retail sources. These products were selected as representative nutraceuticals that are commonly available to consumers, providing a practical context to demonstrate the method’s applicability to both infusion-based and solid matrices. Eight of the twelve target compounds were successfully identified and quantified, with catechin/epicatechin and quercetin derivatives predominating. The analytical results were related to the total flavonoid content (TFC) and were compared with the literature data, supporting CE–DAD as a reliable first-level approach for polyphenol profiling. Given the importance of polyphenol profiling to functional properties such as antioxidant capacity in nutraceutical products, ABTS and DPPH assays demonstrated the high antioxidant potential of the analyzed matrices, supporting the functional relevance of the identified polyphenolic profile. Beyond analytical performance, modern method development increasingly emphasizes sustainability. Green Analytical Chemistry (GAC) principles promote the reduction in hazardous reagents, waste generation, and, if possible, energy consumption. Among the available assessment tools, the BAGI combines environmental impact with practical applicability, making it particularly suitable for evaluating the proposed CE method [28]. BAGI was applied to provide a comprehensive evaluation of both the greenness and applicability of the developed approach.

The novelty of the present study lies in its integrated approach, rather than in any single analytical element. Specifically, it combines multi-class CE–DAD analysis of both aglycones and glycosides with the investigation of complex nutraceutical matrices, eco-friendly sample preparation, and sustainability assessment using the BAGI metric. This comprehensive strategy enables rapid and environmentally sustainable profiling of polyphenols for the preliminary screening of rosehip-based products, with a potential extension to other nutraceutical matrices pending further validation.

2. Materials and Methods

2.1. Chemical and Regents

The solvents, reagents, standards and material used were of high purity.

Methanol (CH_3_OH, MeOH) (Carlo Erba Reagents, Val-de-Reuil, France).Acetonitrile (AcN) (Romil-UpSTM Ultra Purity Solvents, London, UK).Formic acid (HCOOH) (Carlo Erba Reagents, France).Water Milli-Q (Millipore Corporation, Burlington, MA, USA).Hydrochloric acid 37% (HCl_aq_) (Carlo Erba Reagents, France).Sodium Hydroxide (NaOH) (Carlo Erba Reagents, France).Phosphate buffer solution 50 mM pH 7.0 (Agilent technologies, Santa Clara, CA, USA), Sodium ammonium bicarbonate (NH_4_HCO_3_) (Sigma-Aldrich Co., Saint Louis, MO, USA) solution 50 mM pH 8.5, and Sodium Tetraborate solution pH = 9.3 100 mM (Agilent Technologies) were used as background electrolytes (BGE).Sodium nitrite (NaNO_2_) (Carlo Erba Reagents, France).Aluminum trichloride (AlCl_3_) (Carlo Erba Reagents, France).

Different SPE cartridges, STRATA, C_8_ (55 μm, 70 Å, 100 mg/mL), C_18_-E (55 μm, 70 Å, 100 mg/mL), XL (100 μm, 30 mg/mL), and Phenomenex^®^ (Castel Maggiore, Bologna, Italy), were compared for the best recovery. A vacuum manifold, Air Cadet^®^ (Vernon Hills, IL, USA), was used for the elution of cartridges, and it consists of a glass container which collection tubes are placed inside and in which a mild vacuum is generated.

Filtration was performed by 0.45 µm filters (Millex HV, Millipore, MA, USA), and 0.22 µm filters Millex^®^-CV 0.22 µm from Phenomenex^®^ (Castel Maggiore, Bologna, Italy).

The classes, names, acronyms, structures, molecular weights and company names of the 12 investigated polyphenolic compounds are listed in Supplementary Material Table S1. Investigated standard solutions, filtered before use, were prepared at concentrations of 1 mg/mL, each dissolving the pure compounds in methanol stored at −20 °C in the dark. The working solutions were prepared daily by diluting the primary solution with a 10 mM buffer solution, according to the desired concentration (0.4–50 μg/mL), and storing them at 4 °C in amber glass vials.

2.2. Instruments

The capillary electrophoresis system coupled to a diode array detector (DAD) Agilent 7100 (Agilent Scientific Instruments, Santa Clara, CA, USA) is equipped with a bare fused silica capillary (Agilent Scientific Instruments Santa Clara, CA, USA) with a 50 μm internal diameter and 50 cm length, working at 25 °C. The distance between the anode, where the samples are injected, and the cathode, where the detector is placed, is 40 cm.

The sample introduction is performed by hydrodynamic injection, applying a difference in pressure between the capillary ends (50 mpsi pressure for 6 s). The injected sample volume is few nL. The instrument is managed by the specific Agilent Chem Station software (open lab LTS01.11), which allows us to adjust all the operating method parameters for the analysis and to process the data.

2.3. Samples

The selected samples are intended as real-matrix case studies, rather than statistically representative specimens. In particular, the supplement and the rosehip herbal tea, in the form of a tea bag, were purchased from a local herbalist shop/market in [Rome, Italy]. The herbal tea consisted of dried rosehip fruits (Rosa canina L.), cut into small pieces and slightly crushed, as declared by the retailer. All samples were purchased within their stated shelf-life and stored according to manufacturer’s instructions prior to analysis. No specific brand names were indicated on the product labels, and expiration dates were verified prior to use.

Each infusion or supplement sample was prepared in triplicate, and each preparation was processed and injected in triplicate.

As reported by the Ministerial Guidelines (LGM), “supplements are food products intended to supplement the common diet and which constitute a concentrated source of nutrients, such as vitamins and minerals, or other substances having a nutritional or physiological effect, in particular (but not exclusively) extracts of plant origin in pre-dosed forms” (Directive 2002/46/EC) [29]. Rosehip is used in many foodstuffs, beverages and food supplements as a remedy for the treatment of various conditions such as chronic pain, inflammation, flu, colds, and skin care [30]. In this study, a food supplement in the form of tablets was chosen. This product is claimed as a valid aid in regaining natural balance after a period of debilitation for its restorative and supportive action. The herbal tea was chosen, taking into account that these beverages are widely consumed for their bioactive properties, which include being antioxidant and anti-inflammatory, with protective effects against chronic diseases and proven benefits for human health.

2.4. Sample Pretreatment

2.4.1. Herbal Tea

The herbal tea was prepared by the manufacturer’s instructions to simulate the common domestic preparation: 300 mL of water were heated at 100 °C and the herbal tea bag (7 g typical commercial size) was subsequently infused for five min. A 10 mL aliquot of herbal tea was taken and filtered using a 0.45 μm filter.

2.4.2. Food Supplement

Ten capsules were accurately weighed and ground using a mortar. The powder quantity equivalent to one capsule (corresponding to 500 mg whole rosehip) was weighed and dissolved in distilled hot water in a 25 mL volumetric flask. Sonication was performed for 10 min in an ultrasonic bath. The supernatant was carefully separated from the residue, centrifuged at 4000 rpm for 5 min, and filtered through a 0.45 μm syringe filter. Lastly, the solution was diluted with a 10 mM tetraborate buffer solution to achieve a desired concentration within the working range. These steps were optimized to obtain a clear solution that was suitable for CE–DAD analysis and to ensure reproducibility.

2.5. Sample Preparation

The proposed SPE strategy is designed to simplify matrix removal, rather than to maximize analyte enrichment, in line with a green and routine-oriented analytical philosophy.

The sample clean-up procedure involved a solid phase extraction step, using a green interference removal/trapping strategy using STRATA C18-E cartridge, according to the scheme in Figure 1. After conditioning (1 mL of MeOH) and washing the cartridge with alkaline water (1 mL pH= 9.3), the samples (1 mL), treated as above (Section 2.4), are loaded onto the cartridge after alkalinization with BGE buffer 1:1 solution. The analytes are released, passing through the cartridge without interacting, whereas the interferences are retained by the cartridge. The eluate is then evaporated by Rotavapor (Büchi R-200, BUCHI Labortechnik AG—Flawil, Switzerland), 40 °C and 150 mbar and the final residue is ultimately dissolved in 50 μL of the background electrolyte solution before the CE injection. This resulted in a fixed pre-concentration factor being applied consistently across all samples.

2.6. Operating Conditions

Capillary electrophoresis separations were performed under optimized conditions. Detailed capillary maintenance and conditioning procedures are provided in the Supplementary Material (Section S2.6).

To optimize the separation for the investigated compounds, several experiments were performed under different conditions, like the type of buffer, pH, ionic strength (concentration of BGE) and applied potential. After preliminary tests using sodium phosphate buffer (monobasic/dibasic Na_2_HPO_4_/NaH_2_PO_4_, 50 mM at pH 7) and ammonium bicarbonate (NH_4_HCO_3_ 50 mM at pH 8.5), sodium tetraborate buffer (Na_2_B_4_O_7_) was chosen as the background electrolyte (BGE) at different concentrations. We evaluated the electropherograms obtained by applying potentials ranging from 12 to 30 kV. Once the best potential was set (20 kV), the buffer concentration was varied according to Table S2. The best conditions were achieved at a buffer concentration of 50 mM at pH 9.3, along with better reproducibility and resolution.

2.7. Data Quality

Linearity was evaluated by the R^2^ of the calibration curve in the experimental range LOQ μg/mL to 50 μg/mL.

The detection limit, LOD, is defined as the minimum quantity of analyte that the method can discriminate compared to a blank with a certain probability of not having a false negative; the limit of quantification, LOQ, is the smallest quantity of analyte that can be determined with a certain degree of precision.

Repeatability is the measurement of precision under the same operating conditions and over a small interval of time (intra-day), while reproducibility is the measurement of precision under the same operating conditions, but over a larger interval of time (inter-day). Statically, precision is defined as the percentage relative standard deviation (RSD) and calculated as the ratio between the standard deviation and the average value of the measurements. To evaluate the intra-day and inter-day repeatability of the analytical method developed in this study, in the first case, the analyses were repeated three times on the same day, whereas in the second case, they were repeated three times on three different days, according to the following formula:

$$[eqn]$$

The recovery of the analytical method allows us to evaluate whether losses have occurred during the sample preparation procedure. In this study, recoveries of all analytes of interest at a known concentration were calculated. For this purpose, three rosehip-based samples were spiked with a multistandard solution of the analytes (upstream) of the analytical procedure. Contextually, a fourth sample was spiked with the same multistandard solution (downstream) of the analytical procedure, before the injection. By comparing the data obtained from the upstream and downstream spiked extracts, the average percentage recoveries can be calculated for each analyte, according to the following formula:

$$[eqn]$$

To evaluate the matrix effect, the calibration curves of the analytes of interest were built in solvent, as well as in the downstream spiked matrix (subtracted by the endogenous analyte contribution) at known increasing concentrations. The evaluation of the matrix effect was obtained from the percentage ratio of the slopes of the two calibration curves, according to the following formula:

$$[eqn]$$

Ratios equal to 100% suggest the absence of matrix effect; ratios higher or lower than 100% indicate, respectively, a positive matrix effect with signal enhancement or a negative matrix effect with signal suppression. In general, for ratios between 90% and 110%, a negligible matrix effect is assumed.

2.8. Total Flavonoids Content (TFC)

The total flavonoids content (TFC) was determined using the aluminum chloride (AlCl_3_) colorimetric assay, following the modified method already described [31]. Briefly, 400 μL of pre-treated sample (Section 2.2) was sequentially mixed with 1.2 mL of deionized water and 95 μL of 5% NaNO_2_ solution. After 5 min of incubation, 95 μL of 10% AlCl_3_ solution was added. Following an additional 5 min, 640 μL of 1 M NaOH was added, and the mixture was vortexed for 1 min.

The resulting solution was allowed to stand in the dark for 15 min, after which the absorbance was measured at 510 nm, using a UV–Vis spectrophotometer. TFC values were calculated by linear interpolation from a quercetin calibration curve (100–5000 μg/mL) and expressed as mg quercetin equivalents per gram of sample.

2.9. Total Antioxidant Capacity

2.9.1. DPPH Radical Scavenging Activity

Antioxidant activity was assessed using a modified 2,2-diphenyl-1-picrylhydrazyl (DPPH) radical scavenging assay, adapted from Chan et al. [32]. A total of 2925 μL of a 0.1 mM DPPH solution (prepared by dissolving 2 mg of DPPH in 50 mL of ethanol) was combined with 75 μL of the pre-treated sample (Section 2.4). The absorbance was immediately recorded at 517 nm (time 0), and the mixture was then incubated in the dark with gentle agitation for 30 min at 25 °C. The absorbance was subsequently measured again at the same wavelength.

The percentage of DPPH radical inhibition was determined using Equation (1):

$$[eqn]$$

The DPPH scavenging activity was quantified by interpolation against a Trolox calibration curve (0.25–1.25 mM) and expressed as milligrams of Trolox equivalents per gram of sample (mg TE/g sample).

2.9.2. ABTS Radical Cation Decolorization Assay

The antioxidant capacity was further evaluated by the ABTS radical cation decolorization method. The ABTS•^+^ radical was generated by reacting 7 mM ABTS solution with 2.45 mM sodium persulfate and allowing the mixture to stand in the dark for 12–16 h at room temperature. For the assay, 1980 μL of the ABTS•^+^ solution was mixed with 20 μL of the pre-treated sample (Section 2.4). The absorbance was measured at 734 nm at 0 and again after 6 min of reaction.

The percentage of ABTS radical inhibition was calculated according to Equation (2):

$$[eqn]$$

Results were obtained by interpolation from a Trolox standard curve (0.25–1.25 mM) and expressed as milligrams of Trolox equivalents per gram of sample (mg TE/g sample).

2.10. Blue Applicability Grade Index

The BAGI is a new analytical “blueness” metric tool for evaluating the practicality of an analytical method, which was published by Manousi et al. in 2023 [33]. The overall result of assessing a method using BAGI is an asteroid-shaped pictogram with a number in the center. The hue of the pictogram reflects the compliance of the method with ten designated criteria, such as the type of analysis, the number of analytes simultaneously determined, the number of samples processed concurrently, the sample preparation, the number of samples analyzed per hour, the type of reagents and materials used, the requirement for preconcentration, the degree of automation, and the sample amount. The central number indicates the overall applicability score of the method, ranging from 25 to 100. A score of 100 corresponds to excellent performance, while 25 indicates the poorest applicability. Methods scoring at least 60 points are considered practical. Additionally, BAGI accounts for the field of application to mitigate bias and set realistic expectations. For example, it differentiates between bioanalytical methods, which typically involve low sample and reagent volumes, and methods for food or environmental samples, where sample quantities can be increased to meet legislative and sensitivity requirements [34].

3. Results and Discussion

3.1. Method Optimization

Since it is known that many parameters, such as pH, buffer concentration, BGE composition, eventual organic additive, voltage, temperature, and injection conditions might affect the electrophoretic separation, the first step of this work was focused on finding the best experimental conditions as a good compromise between the highest resolution and the shortest time of analysis. Polyphenolic compounds are weak organic acids (pka > 7–7.5). At pH > pKa they are totally/partially dissociated as anions, migrating naturally in an electric field toward the anode. In the presence of electro-osmotic flow, due to a basic buffer (pH > 7) in an uncoated fused-silica capillary, the anions invert their migration, and it is possible to detect them with a cathodic UV detector.

3.1.1. Effect of pH

As a starting point, we compared different buffers, since one of the most important parameters for capillary electrophoresis separation is the pH and the BGE nature. As is already known, by adjusting the pH, it is possible to generate and modulate the electro-osmotic flow (EOF), which is responsible for analyte migration. In addition, the pH of the buffer is crucial, as it governs the dissociation of polyphenols into their ionic forms, enhancing their mobility under an electric field. Consequently, the pH of the BGE buffers was varied between pH 7.0 and 9.3, testing different BGEs: phosphate, carbonate, and borate, at the same concentration (50 mM) and voltage (20 kV). The best results, in terms of peak separation, were obtained at pH 9.3, using the borate buffer [33,34]. The improved resolution is mainly attributed to the high pH, which enhances the EOF and modulates the migration of polyphenols. Additionally, borate can reversibly interact with cis-diol groups in some polyphenols, slightly altering their effective charge and mobility, which may further contribute to peak separation [34,35].

3.1.2. Effect of the Applied Voltage

The voltage was tested between 15 and 30 kV. Analyses performed at a low voltage (<20 kV) lead to zone broadening and longer run time. At a higher voltage (>25 kV), the current generated in the capillary increases up to 80 μA and does not guarantee good repeatability. The best compromise was 20 kV, in terms of satisfactory resolution and short time of analysis.

3.1.3. BGE Concentration

The variation in the buffer concentration influences both the electro-osmotic flow, increasing or reducing it, and the separation of the peaks. An increase in BGE concentration leads to a decrease in the ζ potential and consequently, a reduction in the electro-osmotic flow and an increase in the time of analysis. It is therefore necessary to evaluate and choose the optimal buffer concentration to obtain a good compromise between the separation efficiency and analysis times. For this reason, a standard mixture solution in borate buffer at a different molarity was analyzed (20–60 mM). The graph in Figure 2 summarizes the effect of the variation in BGE (borate buffer) concentration at 20 kV, 25 °C and pH 9.3. The best solution was a buffer concentration of 50 mM with (pH 9.3 at 20 kV), chosen to evaluate the peak broadening, resolution and time of analysis. Under no circumstances does catechin separate from epicatechin, but their combined quantification is adequate for nutritional and screening purposes. This may represent a limitation of the method, even though it is sufficient for general assessment.

3.1.4. Injection Condition

A peculiarity of capillary electrophoresis is that it is not possible to know what volume of the sample enters the capillary (order of nLs). The electropherograms obtained demonstrated that increasing the pressure and injection time has the advantage of injecting a greater quantity of the sample, obtaining more intense peaks. On the other hand, an excessive increase in the pressure could cause the sample to leak out of the vial, resulting in deposits on the electrode and clogging of the capillaries or overloading of the peaks. In our analysis, 50 psi for 6 s was chosen as the optimal injection condition.

3.1.5. DAD Detection

The instrumentation was equipped with a UV-DAD (diode array detector), which allowed us to obtain UV spectra in a specific wavelength range and in real time for all the detected compounds. The analytes of interest showed specific absorption bands in the wavelength range between 200 and 400 nm. Specific detection wavelengths, corresponding to the maximum absorption peaks of the compounds of interest (220, 270, 320, 350 and 380 nm) were used to selectively discriminate the presence of the compounds.

Additionally, the peak purity function proved useful in cases of doubt or potential peak co-elutions, ensuring reliable qualitative analyses.

To enhance the sensitivity and flexibility of the analysis, in cases where the compounds did not exhibit direct UV absorption, a measurable decrease (negative peak) in the UV signal was observed. This decrease, interpreted as a baseline shift in the electropherogram, is directly correlated with the analyte concentration.

As an example, Figure 3 illustrates the CE separation of the analytes under the definitive operating conditions (Table 1) at wavelengths of 220 nm and 380 nm, respectively, highlighting the potential of indirect detection in cases of uncertainty. In this approach, compounds that do not absorb at the detection wavelength displace the background electrolyte or buffer, producing a measurable change in the UV signal, which can appear as a negative peak. This negative peak is directly correlated with the analyte concentration and allows for qualitative and quantitative detection, even for compounds with weak or no native absorption at the chosen wavelength. The mechanism is thus based on the differential absorbance of the background versus the analyte, a well-established approach in capillary electrophoresis with diode array detection.

At 220 nm, all the compounds absorbed, and at 380 nm, rutin, luteolin, quercitrin, and isoquercitrin produced a positive signal, whereas astringin, resveratrol, catechin/epicatechin, hesperetin, piceatannol, kaempferol, and quercetin showed a negative signal. The electropherograms display peaks belonging to different classes of compounds, which are well-resolved, sharp and separated (except for catechin and epicatechin) in less than 15 min with a run-to-run reproducibility (RSD) ranging between 0.8 and 1.6%. In agreement with other studies, it can be assumed that factors such as the position of substituents, charge, and degree of ionization can influence the migration factor [7].

3.2. Validation Parameters

Table 2 shows the validation data parameters of this study.

The LODs were low enough to allow for the detection of all the compounds. Due to the negligible matrix effect, after SPE filtration, for the detected compounds, the calibration curves were built in water solutions in a range suited to real samples or diluted ones. The linearity was assessed by R^2^, always >0.99. The repeatability in terms of migration times of the neutral marker was 0.8% (run-to-run) and 1.6% (day-to-day). Intra-day and inter-day repeatability, referring to peak areas, were always <8% and 18%, respectively. The matrix effect was evaluated for all analytes; by comparing the calibration curves in matrix and in solvent, a ratio of the slopes between 90% and 110% was observed for each flavonoid, indicating a negligible matrix effect.

3.3. Polyphenolic Profiling in Rosehip-Based Samples After Pretreatment and Clean-Up

Due to the complexity and interferences of the matrix, the electropherograms of the samples, simply pre-treated as in Section 2.4, were particularly challenging and difficult to interpret.

The intrinsic difficulty of interpreting electrophoretic traces has prompted the development of a simple and effective sample purification strategy (Section 2.5). To this end, we tested several solid phase extraction (SPE) cartridges in reversed phase mode, Strata C_18_-E, Strata XL and C_8_. An interference removal strategy was adopted, where the analytes of interest are released by passing through the cartridge, while matrix interferences are trapped by the adsorbent. This sample treatment is in line with a sustainable ecological methodology, as it dramatically reduced the use of organic solvents by relying solely on water as the medium. The final choice of the cartridge was a compromise between efficient removal of interferences and the highest recovery of the target analytes.

Furthermore, cartridges with the smallest possible amount of adsorbent were used, aligning with the goal of reducing material waste. These results pointed out the Strata C_18_ cartridge as the most efficient.

Table 3 summarizes the results for recovery, with an average recovery of 81%, which is higher than that tested with the other cartridges, which is typically around 65–70%. The recovery was evaluated in accordance with Section 2.7.

The advantages and usefulness of the SPE procedure can be assessed by comparing the electropherograms of the unpurified and purified sample (Figure 4A,B).

The electropherogram in Figure 4B clearly shows a drastic background reduction between mins 7 and 12. Furthermore, the elimination of interferences allowed for more peaks to be highlighted. Identification of the compounds in real samples was based on four criteria: (1) comparison of migration times, (2) indirect detection, (3) the enrichment method, and (4) comparison of UV absorption spectra. All these data combined allowed us to achieve a reliable identification. In addition, the resolution achieved was analogous to that usually obtained with a classical HPLC separation by UV detection [7,11]. The results demonstrate that the combined green filtration followed by a CZE analysis approach can identify and quantify a wide range of polyphenols, with detection limits below 1 µg/mL. The analytical separation showed excellent reproducibility and superior resolution compared to traditional chromatographic methods, making CZE a highly competitive option.

The application of SPE in a “not bind but elute” mode significantly improved sample purity and analytical sensitivity while drastically reducing solvent consumption. This approach, which retains interference and allows the target analytes to pass through and be directly eluted, demonstrated a clear advantage over the exclusive use of CZE without sample preparation, particularly in complex matrices.

The quantification of the flavonoids was performed using solvent calibration curves (due to the absence of the matrix effect, after the SPE filtration step) obtained by the external standardization, and the relative results are reported in Table 4. The table reports the total flavonoid content (TFC, w/w) determined by a spectrophotometric method and the sum of the flavonoids investigated by CE–DAD, with their contributions expressed as percentages to the TFC. The TFC value, normalized to quercetin equivalents, is taken as 100%, and the reported percentages (approximately 57% and 66%) indicate the relative weight-based contribution of the investigated flavonoids to the total flavonoid content of the samples. Although the spectrophotometric method provides only an approximate estimation of the total flavonoids, these percentages offer a useful indication of how much of the TFC is accounted for by the flavonoids quantified in this study.

These percentage ratios highlight that the investigated flavonoids represent more than half of the total flavonoid content, as shown in Figure 5.

According to this table, the trend of the two types of samples investigated (herbal tea and food supplement) is the same: the most abundant compounds were the flavanols catechin/epicatechin, followed by the flavonols quercetin, quercetrin, isoquercitrin and rutin. Flavanones and flavones were the least abundant, whereas stilbenes were detected at the LOD level. It is important to underline that the herbal tea bag is usually filled with different parts of the plant, such as flowers, fruits, leaves, stems, and roots. Even if the geographical origin of rosehip in this study is unknown, the data are in good agreement with other investigations into different kinds of rosehip. Some of these compounds such as rutin, quercetin and quercetin-3-O-glucoside, catechin, quercetrin, and isoquercitrin have been found by other authors [8,9,13,30]. Another paper [12] also extended the investigation to other phytochemical compounds in crude extract and fractions from Rosa canina, including phenolic acids (e.g., methyl gallate), anthocyanins (e.g., cyanidin-3-O-glucoside), tannins, lignans, pectins, carotenoids, fatty acids (e.g., oleic and palmitic acids), and organic acids such as ascorbic, malic, and citric acids, reflecting its rich and complex phytochemical profile.

In more detail, a quantitative comparison with the literature shows that individual flavonoid concentrations in rosehip infusions vary depending on the extraction and processing conditions. In particular, catechin/epicatechin can reach levels up to ~37,960 µg/g, while quercetin ranges ~334–460 µg/g and rutin 0–46.4 µg/g, whereas quercitrin and isoquercitrin are consistently detected but not always quantified [36]. The concentrations measured in the present study are generally consistent with, or exceed, the ranges reported in the literature.

The observed variability in flavonoid concentrations is consistent with previous studies and may depend on environmental conditions, harvesting time, genetic factors, plant part composition, infusion parameters (temperature, duration, water ratio), and the analytical methodology [37]. Determining glycosylated forms of polyphenols alongside aglycones enhances the analytical and scientific rigor of polyphenol studies, contributing to fields such as food science, nutraceuticals, pharmacology, and health research.

3.4. Nutritional Speculations and Antioxidant Activity

Despite the limited number of samples investigated, the analysis of the data may reveal some peculiarities regarding the nutritional value of rosehip herbal tea versus the food supplement.

From a nutritional point of view (see Table 3), the consumption of 300 mL of rosehip herbal tea corresponds to an intake of more than 60 mg of these flavonoids, whereas with 3–6 tablets/days of food supplement, about 11–22 mg of these flavonoids are introduced, although the bioavailability data for flavonoids in food supplements are not currently available, leaving uncertainty about the actual absorption and utilization by the body. These values might be underestimated, as additional polyphenols, such as anthocyanins, and phenolic acids could also contribute to the total intake, but they were not measured.

In addition to flavonoid quantification, the DPPH and ABTS antioxidant activity assays of both products were also evaluated (see Table 5). The use of both DPPH and ABTS assays provides complementary information on the antioxidant capacity of the samples, as the two methods are based on different reaction mechanisms. While the DPPH assay mainly reflects the ability of antioxidants to donate hydrogen atoms or electrons to a stable radical, the ABTS assay is applicable to a wider range of antioxidant compounds, including both hydrophilic and lipophilic molecules.

The results of the DPPH and ABTS assays showed that the rosehip herbal tea exhibited a markedly higher antioxidant potential compared to the food supplement, in line with its greater polyphenol content. Specifically, the herbal tea showed values of 13.8 mg Trolox/g sample (80.5% DPPH inhibition) and 15.3 mg Trolox/g sample (98.5% ABTS inhibition), whereas the food supplement presented 7.4 mg Trolox/g sample (34.2% DPPH inhibition) and 7.4 mg Trolox/g sample (54.1% ABTS inhibition). This evidence, if extended to a larger representative number of samples, would support the nutritional relevance of rosehip infusion, not only in terms of flavonoid intake, but also regarding its immediate antioxidant capacity, which may provide additional short-term protective effects. Based on the flavonoid profile measured (Table 4), the higher antioxidant activity of rosehip herbal tea compared to the food supplement is consistent with its greater content of catechin/epicatechin, quercetin, quercitrin, isoquercitrin, and rutin. Similar flavonoid constituents have been identified in Rosa canina extracts in previous studies, where higher levels of compounds such as (+)-catechin, (−)-epicatechin, rutin, and quercitrin were associated with the stronger antioxidant activity measured by DPPH and ABTS assays. This consistency suggests that these flavonoids may substantially contribute to the overall radical scavenging capacity of rosehip-based products. However, while these compounds are likely major contributors, it is not possible within the present study to attribute the observed antioxidant activity to individual flavonoids or to generalize this finding beyond the two samples analyzed. To further explore these differences in a transparent and quantitative manner, a descriptive comparison of individual flavonoid levels between the two products was performed and reported in the Supplementary Materials (Table S3). Given the very limited number of samples (two), no inferential statistical tests could be meaningfully applied. Therefore, the changes in flavonoid content were evaluated in terms of direction and magnitude and compared with the corresponding trends observed in DPPH and ABTS antioxidant assays. This exploratory approach does not imply statistical significance, but provides a structured framework to identify flavonoids that most closely follow the overall antioxidant pattern, guiding future studies with larger sample sizes.

In this context, the differences in antioxidant capacity observed in the present study should be interpreted as reflecting the combined effect of the overall polyphenol composition, rather than the action of individual compounds [38]. This interpretation is consistent with a comprehensive review by [39], which highlighted that the antioxidant activity of rosehip extracts arises from the synergistic action of multiple bioactive constituents, including polyphenols. Accordingly, while the descriptive analysis reported here and in the Supplementary Materials suggests that certain flavonoids may follow the same directional trend as antioxidant activity, it does not allow for attribution of causality. Recent findings by Peña et al. (2023) further support the relevance of polyphenolic profiles, showing that rosehip (Rosa canina L.) extracts with higher concentrations of flavonols and catechin exhibit significantly greater antioxidant activity, including DPPH radical scavenging [40]. Together, these observations reinforce the role of polyphenols in determining antioxidant potential, while underscoring the need for future studies with larger sample sizes and targeted analytical approaches to identify the specific bioactive compounds responsible for the observed effects.

3.5. Greenness Evaluation: BAGI Metric

BAGI was applied to systematically assess the practicality and environmental sustainability of the proposed CE-DAD analytical protocol through the evaluation of ten predefined criteria. These parameters describe both the analytical performance and the operational and environmental impact of the method, allowing for an objective comparison with conventional chromatographic approaches.

With respect to the type of analysis, the CE-DAD method was classified as quantitative and confirmatory, positively contributing to the overall score. The method was also suitable for multi-element analysis within a single run, enhancing its analytical applicability. CE-DAD relies on simple and widely available instrumentation, representing an advantage in terms of cost and accessibility.

Several BAGI parameters highlighted the intrinsic greenness of the method. In particular, a very small sample amount (<100 µL) was required, resulting in one of the highest scores among the evaluated criteria. Similarly, the reagents and materials parameter received a favorable evaluation, as the method employs limited volumes of aqueous background electrolytes and avoids extensive use of organic solvents that are typical of HPLC-based techniques, reducing solvent consumption, chemical waste, and environmental burden.

Moderate scores were assigned to the number of samples analyzed per hour (2–4 samples h^−1^) and the degree of automation, reflecting the semi-automated nature of the protocol. Lower scores were attributed to sample preparation complexity and the lack of parallel processing. In addition, the requirement of a preconcentration step slightly penalized the operational simplicity.

Overall, high scores for low sample and reagent consumption, simple instrumentation, and quantitative capability resulted in a global BAGI score of 67.5 (Figure 6), confirming that CE-DAD achieves a balanced compromise between analytical performance, practicality, and environmental sustainability.

Compared to HPLC-MS/MS and HPLC-UV methods, CE-DAD offers significant advantages in environmental friendliness and operational cost, requiring smaller volumes of reagents and samples while being faster and simpler to implement. Minh et al. [41] reported a BAGI score of 65 for an HPLC–DAD method applied to herbal mixtures. Based on our CE-DAD method’s lower solvent and sample consumption, shorter analysis time, and simpler operation, a theoretical BAGI assessment indicates a slightly higher score (~68–70), confirming its improved environmental friendliness and operational efficiency compared to conventional HPLC.

Table 6 summarizes the comparison among these techniques, highlighting CE-DAD’s strengths in low solvent consumption, minimal sample preparation, low instrumental and operating costs, and short analysis time, while maintaining adequate sensitivity, selectivity, and quantification capability. These data confirm CE-DAD as a practical, cost-effective, and greener alternative for routine screening of phytochemicals in nutraceutical products. Although this study is exploratory and based on a single sample, it establishes a solid foundation for future applications of CE-DAD as a rapid and environmentally friendly screening technique.

4. Conclusions

Profiling bioactive compounds is essential for the development of functional foods, beverages, and dietary supplements that promote health, ensure safety, and support informed consumer choices. In this context, understanding the phytochemical composition of herbal teas and supplements is crucial for quality control, efficacy assessment, and the formulation of targeted, health-oriented products.

The goal of the present paper is not to present the ultimate method for all polyphenols and all matrices, but to demonstrate that CE–DAD can effectively function as a green, rapid, and cost-effective screening tool for this type of product.

In this work, a green, fast, simple, and cost-effective analytical method based on capillary zone electrophoresis coupled with diode array detection was developed and optimized for the separation of twelve polyphenolic compounds belonging to different chemical classes. By optimizing the background electrolyte composition, pH, buffer concentration, and applied voltage, effective separation was achieved in less than 15 min. This performance compares favorably with previously reported CE-based approaches for polyphenol analysis.

Following a simple and eco-sustainable sample pretreatment and purification procedure, the validated method was successfully applied to real samples of rosehip-based herbal tea and food supplements, reflecting realistic consumption scenarios and enabling the evaluation of both natural and processed matrices. Despite the limited sample introduction (a few nL), eight flavonoids were successfully identified and quantified, while additional compounds were detected at trace levels. Despite the absence of mass spectrometric detection, the most qualitative reliable technique, clear identification was achieved through migration time comparison, UV spectral matching, enrichment experiments, indirect absorption, and peak purity analysis. The obtained separations were competitive with those achieved by HPLC, even in complex matrices.

In the analyzed samples, a single serving of rosehip herbal tea showed a higher immediate flavonoid content than the recommended daily dose of the selected food supplement, while the latter may be associated with cumulative exposure upon prolonged consumption. Antioxidant activity assays (DPPH and ABTS) consistently indicated higher antioxidant potential for the herbal tea sample, highlighting the functional relevance of the infusion within the limits of this case-based comparison.

Although the quantified compounds represent only a fraction of the total polyphenol content, the method shows strong potential for extension to additional phytochemicals, such as benzoic and cinnamic acids. The results were consistent with the literature data and provided indicative nutritional information. Of course, these considerations are purely speculative, since the present work represents a preliminary, method-oriented study, aimed at demonstrating the applicability of the analytical approach to different commercial matrices, rather than at providing nutritional, functional, or clinical validation. Future studies involving a larger number of products and batches will be required to support any nutritional or functional interpretation.

Rather than serving as an alternative to chromatographic techniques, the proposed CE–DAD method is designed as a first-level, environmentally friendly screening approach for polyphenol profiling in nutraceutical products. Although exploratory in nature and limited to only one matrix, this study demonstrates that CZE–DAD allows rapid, multi-class characterization of bioactive compounds in complex food and nutraceutical matrices with minimal sample preparation. The suitability of the method is further supported by its Blue Applicability Grade Index (BAGI) score of 67.5, indicating a low environmental impact, high practical applicability, reduced solvent use, limited waste generation, and the absence of toxic reagents. Overall, the integration of polyphenol profiling with antioxidant evaluation highlights the potential of CZE–DAD to deliver relevant insights into both the compositional and functional attributes of nutraceutical products. Future studies should expand the number and diversity of polyphenols analyzed, including additional chemical classes and individual compounds, and should involve a larger set of samples from both similar and different product types.