Recovery of Phenolic Compounds From Babassu Mesocarp Flour Using Pressurized Liquid Extraction

Abstract

Babassu mesocarp flour (BMF) is a starchy coproduct obtained during the processing of babassu fruit and represents a potential source of bioactive compounds that remain underexplored. This work aimed to evaluate the recovery of phenolics from BMF by comparing the pressurized liquid extraction (PLE) with conventional maceration, in terms of extraction yield, total phenolic content (TPC), and antioxidant capacity (AC), under the hypothesis that PLE could provide comparable or higher values for these parameters in a shorter extraction time. Extractions were performed with an ethanol–water mixture (75:25, v/v). PLE was conducted in three static cycles of 5 min each at temperatures ranging from 55 to 95 °C, a pressure of 10 MPa, and a volumetric flow rate of 5.0 mL/min. For maceration, the sample (1:10, w/v) was kept at room temperature for 24 h. All extracts were analyzed for global yields, TPC, antioxidant capacity (DPPH, FRAP, and ORAC assays), and their phenolic profile was investigated by UHPLC-ESI-HRMS/MS. PLE at 95 °C provided a higher TPC (25.82 ± 1.05 mg GAE/g) and higher AC (DPPH = 130.54 ± 0.16 µmol TE/g, FRAP = 152.99 ± 3.4 µmol TE/g, ORAC = 158.82 ± 13.66 µmol TE/g) than maceration, while reducing the extraction time from 24 h to 15 min. The main phenolics detected were flavonoids and tannins. Based on these results, PLE can be considered a more time-efficient strategy for recovering phenolic bioactives from BMF.

Introduction

Babassu (Attalea speciosa) is an economically important palm crop from the Arecaceae family, native to a transition zone between Brazilian Caatinga, Cerrado, and Amazon [1]. It is naturally found in the northern and northeastern regions of the country, especially in the states of Maranhão, Piauí, and Tocantins. The babassu palm produces drupe fruits that grow in dense bunches, commonly referred to as coconuts. In 2024, a total of 25,572 tons of seeds were collected, with the state of Maranhão producing the largest amount [2]. It is worth noting that this species holds substantial socioeconomic and cultural importance for traditional communities. For instance, the collection, management, and breaking of babassu coconuts are activities predominantly carried out by women known as “coconut breakers”. In this context, these elements highlight the relevance of babassu as a significant resource that contributes to sustainable livelihood strategies and regional bioeconomic resilience.

The babassu fruit is composed of four parts: the epicarp (11%), mesocarp (23%), endocarp (59%), and seeds (7%) [3]. The seeds are the source of babassu oil, which is the highest value-added product obtained from fruit processing. The mesocarp, in turn, is considered a coproduct of oil extraction and undergoes drying and milling processes to produce babassu mesocarp flour (BMF), a starchy (> 60%) ingredient applied in food formulations [4, 5]. The distinctive coloration of BMF suggests the presence of phenolic compounds and pigments, reinforcing the potential of this coproduct for developing innovative food and health-related applications, as well as strengthening value chains based on native resources. These practices align with the growing interest in sustainability and the circular economy, particularly with the objectives outlined in the Sustainable Development Goals (SDGs), which emphasize the importance of valorization and the use of natural resources [6].

A few studies have addressed the recovery of phenolic compounds from BMF [7, 8]. These components are relevant to human health due to their nutritional properties and are also of interest to the pharmaceutical industry for the development of therapeutic products and to the food industry for formulating natural colorings, antioxidants, and flavor enhancers [9]. In this context, solid-liquid extraction is widely used to extract these bioactive compounds from natural matrices. However, conventional methods often involve organic solvents that raise safety and environmental concerns and require long extraction times and high solvent consumption, which can result in lower selectivity and limited yields. This scenario highlights the need to explore emerging green extraction techniques that provide multiple benefits, including reduced solvent usage, improved time and energy efficiency, and more sustainable processing [10]. Therefore, adopting more advanced extraction approaches may increase the range of the technological applications of BMF extracts.

Pressurized liquid extraction (PLE) is an emerging technology that employs solvents at high pressure and temperature close to or above their boiling point [11]. PLE using ethanol and water may represent a sustainable approach for obtaining valuable compounds from natural sources and food coproducts, since both solvents are generally recognized as safe (GRAS) and enable the production of extracts enriched in polar target compounds, including phenolic compounds with significant antioxidant activity [12]. Compared with conventional extraction techniques, PLE involves shorter extraction times and reduced solvent consumption, generally resulting in higher extraction efficiency [11]. Higher temperatures, except in the case of thermosensitive compounds, tend to favor greater extraction yields, as they may enhance mass transfer and extraction rates through improved matrix solubilization, partial disruption of solute–matrix interactions, decreased solvent viscosity, and reduced surface tension [13, 14]. Thus, the positive effect of high temperatures in PLE has been demonstrated by previous investigations as an alternative to improve the recovery of phenolic compounds [9–11, 15]. However, studies investigating the PLE of BMF are scarce, and the promising potential of its extracts highlights the need to explore sustainable extraction methodologies. In this context, this study aimed to apply PLE to obtain phenolic compounds from BMF and to compare its performance with a conventional extraction method (maceration). By investigating extraction yields, total phenolic content, antioxidant activity, and phenolic profile, the study is based on the hypothesis that PLE may be a promising approach for valorizing this underutilized native resource.

Materials and Methods

Chemicals

The Total Starch and Total Dietary Fiber Assay Kits were purchased from Megazyme International (Wicklow, Ireland). Gallic acid (CAS 149-91-7), Trolox (6-hydroxy-2,5,7,8-tetramethylchromane-2-carboxylic acid; CAS 53188-07-1), 2,2-diphenyl-1-picrylhydrazil (DPPH; CAS 1898-66-4), 2,2′-azobis(2-methylamidinopropane)-dihydrochloride (AAPH; CAS 2997-92-4), 2,4,6-tris(2-pyridyl)-s-triazine (TPTZ; CAS 3682-35-7), fluorescein (CAS 2321-07-5), sodium carbonate (CAS 497-19-8), and Folin–Ciocalteu reagent were purchased from Sigma–Aldrich (São Paulo, Brazil). All other chemical reagents used were of analytical grade.

Plant Material

BMF was provided by Cooperativa de Produtores Agroextrativistas de Esperantinópolis (Coopaesp), located in Maranhão, Brazil (4°52′6″S latitude, 44°40′19″W longitude). Briefly, the procedure for obtaining the flour involved collecting and cleaning the fruits that had fallen to the ground, breaking the babassu coconuts with a manual breaker, and removing the seeds. The babassu mesocarp was subsequently obtained, dried, subjected to particle size reduction, sieving, packed into plastic bags, and stored until further use.

Proximate Composition

BMF was evaluated according to the methodologies established by the American Association of Cereal Chemists [16]. Moisture content was performed by desiccation in an oven with air circulation and recycling at 105 °C until constant weight (method 44–15.02); total lipid content by Soxhlet extraction (method 30–25.01); ash content by muffle incineration (method 08–01.01); proteins (total nitrogen content) by the Kjeldahl method with a conversion factor of 6.25 (method 46–13.01); total dietary fiber by the enzymatic-gravimetric method (method 32-07.01), and total starch by the enzymatic method (method 76-13.01).

Pressurized Liquid Extraction

PLE was performed according to the methodology described by Cornelio-Santiago et al. [17], with modifications. Approximately 3 g of BMF were placed in a 50 mL stainless steel column, forming a fixed bed. The PLE was carried out in three batches with a static period of 5 min each. The temperatures of 55, 75, and 95 °C were tested, using an ethanol-water mixture (75%, v/v) as solvent. The solvent-to-feed ratio (S/F) was 25 mL/g. The solvent was fed through an HPLC pump (Jasco, PU-2087, Japan) and subsequently heated to the designated process temperature using an electric heating system. The preheated solvent was then introduced into the extraction column, where it permeated the sample matrix and solubilized the extractable constituents. The resulting extract exited the column and was collected after depressurization through a micrometer valve. Pressure and volumetric flow rate were kept constant at 10.0 ± 0.5 MPa and 5.0 mL/min, respectively. The final volumes of the extracts obtained by PLE were 59 mL at 55 °C, 49 mL at 75 °C, and 86 mL at 95 °C. The collected extracts were stored in amber vessels at − 18 °C for further analysis.

Maceration

To perform maceration (MAC), approximately 3 g of BMF were homogenized with 30 mL of a 75% ethanol–water solution (1:10, w/v) and maintained at room temperature (25 ± 2 °C) for 24 h, protected from light. After that, the solution was centrifuged (Omega, Laborline, Brazil) for 15 min at 3200 rpm, and the extract was collected. The samples were stored in amber vessels at − 18 °C for further analysis.

Global Yield (X₀)

The global yields of PLE and MAC were calculated as the mass ratio between dry extract (m_ext) and the sample subjected to extraction (F), according to Eq. 1. m_ext was determined by drying a 5 mL aliquot at 65 °C in an air-circulation oven (Marconi, MA-184, Brazil) until a constant weight. m_ext was calculated by multiplying the solids content obtained (g/mL) by the volume (mL) of extract recovered from the extraction.

1

Total Phenolic Compounds

The total phenolic content (TPC) was determined using the Folin–Ciocalteu method described by Singleton, Orthofer, and Lamuela-Raventós [18], with minor modifications. Briefly, an aliquot of the appropriately diluted extract (0.5 mL) was transferred to a test tube and mixed with 2.5 mL of 10% Folin–Ciocalteu reagent, followed by the addition of 2 mL of 7.5% sodium carbonate. The reaction was allowed to proceed in the dark at 45 °C for 45 min. Absorbance was measured at 760 nm using a spectrophotometer (Cirrus 80, Femto, Brazil). TPC was quantified using gallic acid as the calibration standard (10–100 µg/mL) and expressed as milligrams of gallic acid equivalents per gram of dry BMF (mg GAE/g).

Antioxidant Capacity

The antioxidant capacity (AC) of the extracts obtained by PLE and MAC was assessed using the DPPH free-radical scavenging assay, oxygen radical absorbance capacity (ORAC), and ferric reducing antioxidant power (FRAP) methods.

DPPH was determined according to the methodology of Brand-Williams et al [19]. Briefly, 800 µL of the diluted extract was mixed with 4 mL of DPPH solution (0.004%, w/v). Samples were incubated in the dark at room temperature for 30 min. Subsequently, the absorbance was measured at 517 nm using a spectrophotometer (Cirrus 80, Femto, Brazil). Trolox solutions were used as a standard using a calibration curve (5–50 µg/mL), and the results were expressed as µmol of Trolox equivalent (TE)/g of sample.

The FRAP method was conducted as reported by Benzie and Strain [20], with some modifications. The FRAP solution was prepared using a 0.3 M acetate buffer (pH 3.6), 10 mM TPTZ dissolved in 40 mM HCl solution, and 20 mM FeCl₃. A 25 µL aliquot of the diluted extract was added to 175 µL of FRAP solution and kept in the dark at 30 min at 37 °C. Subsequently, the absorbance was measured at 595 nm in a microplate reader (FLUOstar Omega BMG LABTECH GmbH, Ortenberg, Germany). The Trolox standard curve (0.01–0.06 mg/mL) was plotted, and the results were expressed as µmol of TE/g of sample.

The ORAC assay was performed following the methodology described by Dávalos et al [21], with adaptations. The extracts were diluted in a potassium phosphate buffer (pH 7.4) (PB). In the dark, 25 µL of sample/standard/blank (PB), 150 µL of fluorescein in PB, and 25 µL of AAPH (108 mg/mL) were added to black 96-well microplates. Fluorescence was measured for 101 min using a microplate reader (FLUOstar Omega BMG LABTECH GmbH, Ortenberg, Germany) at 37 °C, with a 485 nm excitation filter and a 520 nm emission filter. The Trolox standard curve (5–25 µg/mL) was plotted, and the results were expressed as µmol of TE/g of sample.

Identification of Phenolic Compounds By UHPLC-ESI-HRMS/MS

Qualitative analyses of phenolic compounds were performed using an ultra-high-performance liquid chromatography (UHPLC) system (Ultimate 3000, Thermo Scientific) coupled to a high-resolution Orbitrap mass spectrometer (Exploris 120, Thermo Scientific) with a heated electrospray ionization (H-ESI) source operated in tandem mass spectrometry (MS/MS) mode. Specifically, the extracts were filtered through a 0.22 μm PVDF filter and 10 µL of diluted sample was injected into the system. The chromatographic separation was performed on a Kinetex C18 column (3.0 × 100 mm, Phenomenex Inc., Torrance, CA, USA) operated at 40 °C. The mobile phases consisted of water with 0.01% formic acid (A) and methanol (B) for the positive ionization mode, and water with 0.01% ammonium hydroxide (A) and methanol (B) for the negative ionization mode. The gradient elution was set from 10 to 90% B over 0–15 min, with both ionization modes run at a flow rate of 0.30 mL/min. Ionization voltages were set at 3,500 V for positive and 2,500 V for negative ion modes, respectively. The sheath and auxiliary gas flow rates were set to 50 and 10 arbitrary units (a.u.), respectively. The ion transfer temperature was 325 °C. Resolutions were 60,000 for MS and 15,000 for MS/MS, with a scan range of 100–1000 m/z.

Data processing was performed using the Compound Discoverer and FreeStyle software packages (Thermo Fisher Scientific). The MS data, MS/MS fragmentation profiles, and molecular formula were compared with the literature data and some databases, such as MassLists, ChemSpider, and mzCloud, to annotate the phytochemicals analyzed from the extracts. A maximum mass error of 5 ppm was accepted.

Statistical Analysis

All analyses for proximate composition, X₀, TPC, and AC were carried out in triplicate (n = 3), and the results were expressed as means ± standard deviations. Differences were evaluated by analysis of variance (ANOVA), followed by Tukey’s test at a 5% significance level, using R-Studio (Version 9.2.191126, RStudio Inc., MA, USA).

Results and Discussion

Proximate Composition of Babassu Mesocarp Flour

The proximate composition of BMF is summarized in Table 1. The sample exhibited a moisture content considered adequate for safe storage. Additionally, starch was the predominant fraction, which may influence technological processes, emphasizing the importance of determining its content. Moreover, BMF presented dietary fibers (Table 1) and minor components (< 2 g/100 g), such as protein, lipid, and ash. Previous studies have demonstrated that this amylaceous raw material exhibits significant variations in its chemical composition: starch (65.63–84.57%), moisture (10.53–12.89%), proteins (1.39–1.77%), lipids (0.17–1.51%), fibers (11.11–16.48%), and ashes (1.04–1.59%) [4–5, 22]. It is important to note that the chemical composition of BMF is influenced by several factors, including its origin, edaphoclimatic conditions, species, harvesting period, and the processes used for flour production, among others [1].

Table 1.

Proximate composition of babassu mesocarp flour

Components	BMF
Moisture (g/100 g, w.b.)	10.57 ± 0.31
Protein (g/100 g, d.b.)	1.35 ± 0.00
Lipid (g/100 g, d.b.)	0.81 ± 0.14
Ash (g/100 g, d.b.)	1.09 ± 0.04
Total Starch (g/100 g, d.b.)	84.64 ± 1.90
Total Dietary Fiber (g/100 g, d.b.)	7.38 ± 0.20

Results are expressed as mean ± standard deviation (n = 3). BMF babassu mesocarp flour, w.b. wet basis, d.b. dry basis.

Global Yields

The X₀ values obtained by PLE and MAC are shown in Fig. 1. There are significant differences (p < 0.05) among the extraction methods employed in this work. The highest X₀ values were obtained by PLE at 95 °C (5.36%) and 75 °C (5.2%), which showed no statistical difference (p > 0.05), followed by PLE at 55 °C (4.25%) and MAC (3.39%). In general, high temperatures weaken van der Waals, hydrogen, and dipole-dipole interactions between target compounds and the vegetable matrix, thereby lowering the energy required for desorption [15, 23]. Additionally, elevated temperatures reduce solvent viscosity and surface tension, improving solvent penetration into the matrix and accelerating extract dissolution. As a result, mass transfer is enhanced, leading to higher extraction yields [24].

Fig. 1.

Global yields of the extracts obtained from babassu mesocarp flour through PLE and maceration

Total Phenolic Content

Figure 2 presents the TPC of the extracts, showing that PLE at 95 °C resulted in significantly higher values (p < 0.05) than all other conditions. PLE at 95 °C (25.82 ± 1.05 mg GAE/g) increased the phenolic extraction in about 1.3, 1.34, and 1.47 times compared to PLE at 75 °C (19.79 mg ± 0.51 GAE/g), PLE at 55 °C (19.23 ± 0.15 mg GAE/g), and MAC (17.53 ± 0.69 mg GAE/g), respectively, indicating that high temperature is effective in phenolic extraction. No significant differences were observed between PLE at 55 °C and MAC (p > 0.05), indicating that PLE achieves equivalent TPC recovery in a shorter extraction time. Regarding PLE at 75 °C and 55 °C, no statistical differences (p > 0.05) were reported, which might be related to the fact that, within this temperature range, the phenolic compounds in free form or weakly bound to the matrix were already recovered at 55 °C. As a result, the temperature increased to 75 °C, which may suggest the occurrence of partial starch gelatinization, increasing viscosity of the medium, and promoting the encapsulation of phenolic compounds within the formed gel, which, in turn, did not result in a notable enhancement in the extraction of additional phenolics [25]. In contrast, PLE at 95 °C resulted in a significant increase in TPC. According to Wu et al [25], higher extraction temperatures promote the disruption of the starch granular structure, facilitating the release of phenolic compounds that were previously entrapped or strongly bound within the matrix. Furthermore, higher thermal energy enhances molecular mobility, improving compound diffusion and solvent penetration [23]. Therefore, it is hypothesized that the combined effect of structural breakdown of the starch-based matrix and increased thermal energy may contribute to the more efficient extraction of phenolic compounds in PLE at 95 °C.

Fig. 2.

Total phenolic content of the extracts obtained from babassu mesocarp flour through PLE and maceration

Lima et al [8] recovered bioactive compounds from BMF using a hydroethanolic extract, which was subsequently diluted in methanol and subjected to liquid-liquid partition with solvents of increasing polarity, including ethyl acetate. They reported TPC of 536.50 mg GAE/g in the ethyl acetate fraction and 686.18 mg GAE/g in the hydromethanolic fraction. Despite the high recovery, ethyl acetate and methanol are considered toxic when inhaled, ingested, or absorbed through the skin [26]. Although studies on BMF are limited, the TPC values obtained in the present work are higher than those reported for other babassu fractions, such as the oil (0.0213–0.8724 mg GAE/g of oil), as reported by Souza et al. [27]. While direct comparisons are influenced by differences in matrix composition and extraction conditions, these findings highlight the importance of temperature as a key factor in enhancing phenolic release.

Antioxidant Capacity

The AC of the PLE and MAC extracts obtained from BMF and determined by the DPPH, ORAC, and FRAP methods are shown in Fig. 3. Both methods showed the superiority of PLE over MAC. In the DPPH assay, AC decreased in the order PLE at 95 °C (130.54 ± 0.16 µmol TE/g) > PLE at 55 °C (96.86 ± 0.58 µmol TE/g) > PLE at 75 °C (83.05 ± 0.33 µmol TE/g) > MAC (54.5 ± 0.24 µmol TE/g). A similar pattern was observed for the FRAP test, with values of 152.99 ± 3.4, 117.88 ± 1.34, 45.2 ± 1.5, and 25.98 ± 1.3 µmol TE/g for the same extraction conditions. Interestingly, the higher AC observed in PLE at 55 °C compared to 75 °C in both DPPH and FRAP procedures may be associated with changes resulting from the starch gelatinization phenomenon, as previously mentioned. This process may have led to the formation of a polymeric matrix capable of encapsulating or interacting with phenolic compounds, limiting their extractability, as described by Grisales-Mejía et al. [28] in avocado seed. This hypothesis is also supported by the thermal properties of babassu mesocarp starch, with an onset gelatinization temperature (T_o) of ~ 66 °C and a peak gelatinization temperature (T_p) of ~ 73 °C, indicating that partial gelatinization may occur within this temperature range [29]. Then, this intermediate temperature may induce partial swelling of the starch-rich babassu matrix, facilitating the release of bound or high-molecular-weight phenolic compounds, which could exhibit lower AC. These findings highlight the complex interplay between temperature, starch matrix changes, and the stability of bioactive compounds during extraction processes.

Fig. 3.

Antioxidant capacity of the extracts obtained from babassu mesocarp flour through PLE and maceration

The ORAC assay (Fig. 3) showed that the AC obtained by PLE at 95 °C (158.82 ± 13.66 µmol TE/g) and 75 °C (145.82 ± 1.99 µmol TE/g) were statistically similar (p > 0.05), as were those from PLE at 75 °C and 55 °C (133.04 ± 3.63 µmol TE/g). Maceration exhibited the lowest ORAC value (96.44 ± 5.47 µmol TE/g), which was significantly lower than the other treatments (p < 0.05). However, in comparison with previously reported values for babassu oil (0.156–3.97 µmol TE/g of oil) [27], the present study revealed higher AC values for BMF. It is possible that higher TPC does not directly translate into higher AC, since the selectivity of extraction can affect which compounds are present in the extract. In other words, not all phenolic compounds recovered from a matrix may contribute directly to antioxidant activity. In addition to the structural and functional differences among various phenolic compounds, it is important to consider that many of those are thermally unstable and may be susceptible to degradation when exposed to extended extraction times under high temperatures [30].

It is also important to highlight that each method measures a distinct mechanism of action. The DPPH assay is based on the direct scavenging of DPPH• organic radicals by antioxidant compounds [14]. The FRAP technique evaluates the AC by measuring the reduction of Fe³⁺ to Fe²⁺ at low pH [31]. It is important to note that colorimetric methods may not be as sensitive and selective. For instance, the FRAP method may lead to an overestimation of AC, as the reduction of Fe³⁺ can also be facilitated by the presence of transition metals with lower redox potential, not exclusively by electron donation from organic antioxidant compounds [14]. In contrast, ORAC values are considered biologically relevant indicators of antioxidant activity. The ORAC test evaluates the AC by assessing the inhibition of the oxidation of peroxyl radicals, which are predominant in lipid oxidation within biological systems and food matrices. The assay measures the oxidative degradation of fluorescent molecules after the addition of free radical generators (e.g., azo-initiators), with antioxidants acting to protect against this degradation [31].

Lima et al. [7] applied ultrasound-assisted extraction (UAE) as a green technique to recover bioactive compounds from BMF and reported FRAP values ranging from 31.67 to 40.37 µmol TE/g in the ethanolic extract. In contrast, PLE in the current study resulted in higher FRAP values, while also offering additional advantages over UAE by eliminating the need for extract filtration, allowing easy scale-up to volumes over 100 L, and facilitating automation [32]. Based on these results, BMF stands as a viable alternative to extract phenolic constituents, offering potential for food and non-food applications.

UHPLC-ESI-HRMS/MS Identification of Phenolic Compounds

Phenolic compounds were tentatively identified based on their exact masses, MS/MS spectra, and molecular formulas. A total of nine compounds were identified across the treatments, with a predominance of flavonoids and tannins. The outcomes of qualitative analysis of BMF extracts using UHPLC-ESI-HRMS/MS are shown in Table 2. All identified compounds exhibited low mass errors, indicating the accuracy of the measured exact masses and molecular formulas.

Table 2.

UHPLC-ESI-HRMS/MS identification of phenolic compounds from babassu mesocarp flour

Proposed compounds	RT (min)	Molecular formula	Theoric mass	a[M-H] − / b[M-H] +	Maceration	PLE at 55 °C	PLE at 75 °C	PLE at 95 °C	MS/MS ion fragments	Mass error (ppm)
Gentisic acid	1.32	C7H6O4	154.0266	153.0197a	✓		✓	✓	153, 109	0.9
Vitexin 2”-O-p-coumarate	1.53	C30H26O12	578.1424	579.1496b	✓	✓	✓	✓	163, 127	−0.18
B-type procyanidin dimer	1.53	C30H26O12	578.1424	577.1345a/ 579.1496b	✓	✓	✓	✓	407, 289, 125	−0.87
A-type procyanidin dimer	1.54	C30H24O12	576.1267	577.1341b	✓	✓	✓	✓	289	−0.14
3,4,2’,4’,6’-Pentahydroxydihydrochalcone	5.91	C15H14O6	290.0790	291.0864b	✓	✓	✓	✓	227, 123	0.17
2,6,3’,4’-Tetrahydroxy-2-benzylcoumaranone	7.55	C15H12O6	288.0633	289.0707b	✓	✓	✓	✓	163, 135, 123	0.05
Catechin/Epicatechin	7.70	C15H14O6	290.0790	289.0718a/ 291.0863b	✓	✓	✓	✓	245, 205 139	−0.18
Quercetin 3-methyl ether 3’-glucoside	10.32	C22H22O12	478.1111	477.103a	✓		✓		477, 314, 227	−1.72
Propyl gallate	13.17	C10H12O5	212.0684	153.0197a	✓		✓	✓	153, 125, 109	−0.55

R_T retention time, [M-H]⁻ negative mode ionization, [M-H]⁺ positive mode ionization, PLE pressurized liquid extraction.

Among these phenolics were procyanidins, a class of condensed tannins that can be formed through the condensation of catechin and epicatechin monomers. Two procyanidin dimers were detected in all samples, including the A-type procyanidin dimer and the B-type procyanidin dimer (Table 2). Those procyanidins were reported in the BMF extracts by Lima et al. [8]. It is worth mentioning that procyanidins possess high radical scavenging activity and a strong capacity to decrease oxidative stress associated with neurodegeneration. Furthermore, these compounds have demonstrated anticancer, antidiabetic, anti-obesity, and cardioprotective activities [33]. Then, it is expected that this study will contribute to the development of BMF phenolic compounds as natural antioxidants in the future.

Catechin and epicatechin, which are isomers characterized by the molecular ion [M − H−289]⁻ or [M − H−291]⁺, found in the extracts of all treatments. A previous study conducted on BMF extracts also confirmed the presence of these compounds [34]. Due to the phenolic hydroxyl groups, catechin is reported to exhibit a wide range of biological activities, including neuroprotective, cardioprotective, anti-atherosclerotic, antitumor, anti-inflammatory, and anti-aging effects [35].

Other annotated compounds were detected in all extracts, including 2,6,3’,4’-tetrahydroxy-2-benzylcoumaranone (a polyhydroxylated phenolic derivative), 3,4,2’,4’,6’-pentahydroxydihydrochalcone (a dihydrochalcone belonging to the flavonoid subclass), and vitexin 2’’-O-p-coumarate (an acylated flavone glycoside). Moreover, gentisic acid (a hydroxybenzoic acid derivative) and propyl gallate (a gallate ester derived from a phenolic acid) were tentatively identified in some treatments, suggesting that processing conditions may influence their presence. The compound annotated as quercetin 3-methyl ether 3’-glucoside (an O-methylated flavanol glycoside) was detected in the present study. It is worth noting that a quercetin glucoside derivative has previously been reported by Lima [8], suggesting the possible occurrence of flavanols derived from quercetin in this matrix.

Notably, the temperature applied during thermal processing plays a crucial role in determining the thermal stability of phenolic compounds. It is hypothesized that, during PLE, the abundant starch matrix in BMF may interact with phenolic compounds through hydrogen bonding or glycosidic linkages, potentially influencing their release and stability under high-temperature conditions. Although high temperatures may promote partial degradation, oxidation, or structural transformation of thermolabile phenolics, extraction at 95 °C in this study yielded the highest total phenolic content and antioxidant activity, suggesting that the extensive disruption of the starch matrix enhances the release of both bound and thermally stable bioactive compounds. It is important to note that individual flavonoids were not quantified, and further studies are necessary to identify the specific molecules responsible for the observed antioxidant effects and to elucidate their interactions with the starch structure better.

Conclusions

The application of high pressure enables extraction at significantly higher temperatures, resulting in increased yields and antioxidant activities compared to conventional MAC. Regarding TPC, the hydroalcoholic extracts obtained at 95 °C exhibited the highest concentration. The extracts were mainly composed of flavonoids and tannins, which were tentatively identified in this study. Therefore, PLE appears to be a promising approach for the recovery of bioactive compounds from BMF. However, future efforts should focus on optimizing extraction conditions, accurately quantifying the target compounds from this underutilized crop, investigating their degradation kinetics, and evaluating their potential biological and technological applications.

Author Contributions

LAB: Conceptualization, Methodology, Investigation, Formal analysis, Data curation, Writing – original draft, Writing – review & editing. RPFM: Conceptualization, Methodology, Formal analysis, Data curation, Writing – review & editing. LDB: Conceptualization, Methodology, Formal analysis, Data curation, Writing – review & editing. JM: Conceptualization, Resources. LWH: Conceptualization, Resources. KKR: Conceptualization, Resources. PE: Conceptualization, Investigation, Project administration, Resources, Supervision, Writing – review & editing. All authors read and approved the final version of the manuscript.

Funding

The Article Processing Charge (APC) for the publication of this research was funded by the Coordenação de Aperfeiçoamento de Pessoal de Nível Superior - Brasil (CAPES) (ROR identifier: 00x0ma614). This research was supported by the Brazilian National Council for Scientific and Technological Development (CNPq) through a scholarship awarded to L. A. Borges (grant n°. 140510/2023–0). This study was also financed by Mombora Indústria e Comércio de Alimentos Ltda, Brazil.

Data Availability

The authors declare that the data supporting the findings of this study are available within the paper.

Declarations

Competing interests

The authors declare no competing interests.