Polyphenol Extraction Using Microwaves and Pressured Liquid Condition from Cupuaçu Seed By-Product: Optimization and Comparative Study

Abstract

Cupuaçu (Theobroma grandiflorum Schum.) is an Amazonian fruit rich in antioxidants potential due to its phenolic compounds. Its seeds produce cupulate, a chocolate-like product, and the residue remains rich in biocompounds. This study aimed to optimize and compare microwave-assisted (MAE) and high pressure/temperature extraction (HPTE) of polyphenolic and flavonoid compounds from cupuaçu seed by-products. A Box-Behnken 3³ factorial design assessed the influence of extraction time, solid/liquid ratio, and ethanol concentration. Responses included total polyphenol content, total flavonoid content, and anti-radical power via the ABTS (2,2-azino-bis-3-ethylbenzothiazoline-6-sulphonic acid) method. In optimum conditions, MAE had the highest total polyphenol content and HPTE for flavonoids. Optimal conditions for both methods were 65% (m/v) ethanol concentration, 45 min extraction time, and 0.03 g mL⁻¹ solid/liquid ratio. MAE and HPTE achieved greater recovery of (−)-epicatechin and (−)-epigallocatechin-3-gallate compared to percolation. Both techniques proved viable for recovering polyphenols from cupuaçu seed by-products, in less time and with less solvent. Thus, cupuaçu by-products can be a sustainable and functional food source, with MAE and HPTE providing high yields and quality extracts rich in polyphenols.

Keywords:
Theobroma grandiflorum ; antioxidant activity; microwave-assisted extraction; high pressure/temperature extraction; waste valorization

Introduction

Theobroma grandiflorum Schum. (cupuaçu) is a native species of tropical rainforests, originating from the Eastern Amazon in Brazil, and can be found in primary vegetation areas, measuring between 15 and 20 meters in height.¹ The economic interest in cupuaçu arises from its aromatic pulp and its appreciated acidity levels among consumers. The pulp of the fruit is extensively used in both industrial and homemade culinary specialties in the Amazon region.¹, ² Due to the various applications of cupuaçu and its ease of cultivation in tropical climates, there has been increased investment in its cultivation, driven by growing international market.³ Consequently, some companies are investing in the management and production of cupuaçu for economic purposes, leading to an increase in waste from the agro-industry of this fruit, which is rich in antioxidant compounds.

Recently, there has been a significant discussion in academia regarding the importance of reuse of agro-industrial biocompounds from fruits with high economic potential, such as cupuaçu (Theobroma grandiflorum Schum.). It is estimated that this Amazonian fruit, whose seeds are used in the production of cupulate, a product similar to chocolate, yields 27 thousand tons in the state of Pará, covering 8.5 thousand hectares with an average yield of 3 t ha⁻¹. This production generates approximately 30,000 tons of cupuaçu fruit husk annually in Pará.⁴, ⁵

Yang et al.² and Pugliese et al.⁶ found that cupuaçu seeds contain high levels of phenolic compounds, which are responsible for high antioxidant activity in vitro. These authors identified nine known flavonoids with antioxidant activity in the seeds, namely the flavan-3-ols (+)-catechin and (−)-epicatechin, the flavones isoscutellarein-8-O-β-D-glucuronide, isoscutellarein-8-O-β-D-glucuronide-6”-methyl ester, and hypolaetin-8-O-β-glucuronide, and the flavonols, quercetin, quercetin-3-O-β-D-glucuronide, quercetin-3-O-β-D-glucuronide-6”-methyl ester, and kaempferol, as well as two new sulfated flavones, theograndin I and theograndin II.² The analysis also showed that both (−)-epicatechin and (+)-catechin were present in low quantities as terminal units in proanthocyanidin oligomers, with the former being six times more abundant than the latter.⁶ These studies highlight the importance of reusing industrial by-products, as they still contain significant amounts of bioactive compounds.

Several studies⁴, ⁵, ⁷, ⁸, ⁹ have been conducted to combine different waste recovery processes with sustainable techniques by applying green chemistry principles. Green chemistry aims to utilize innovative techniques that are less harmful to the environment, like extraction techniques capable of reducing extraction time, solvent consumption, and preserving thermolabile constituents. Among these techniques, microwave-assisted extraction (MAE) stands out as one of the most promising methods to assist the solvent extraction of chemical compounds from food and its by-products, due to its shorter extraction time compared to conventional techniques such as maceration and percolation.¹⁰, ¹¹, ¹², ¹³ Additionally, due to its low cost, the reduced extraction time, and ease of equipment installation, it has been meeting the concept of eco-friendliness.¹⁴, ¹⁵

Microwave-assisted extraction consists of solvent extraction under irradiation with microwaves, which are electromagnetic radiation, with frequency varying in the range from 0.3 to 300 GHz, able to interact with polar compounds and produce bulk heating via ionic conduction or rotational dipole.¹⁶ Two frequencies are often used in chemical research, namely 2.45 GHz in laboratory equipment and 0.91 GHz in industrial equipment.¹⁷ To carry out an ideal setup based on MAE, it is crucial to estimate the average response to radiation (permittivity), whose knowledge allows to characterize the process in terms of heating rate, penetration depth, temperature distribution, etc., in addition to helping to understand the phenomena that occur during the process.¹⁸ To successfully apply MAE to the recovery of the bioactive compounds from plant raw materials, some factors that influence the successful application of MAE include the dissipation factor,¹⁹ which is responsible for promoting cell disruption and release of its constituents,²⁰ the solubility of the target compounds in the solvent, their partition coefficient, and mainly their diffusion.²¹

High pressure/temperature extraction (HPTE) is another promising technique that facilitates solvent extraction by combining high pressure and temperature. This technique increases extraction efficiency by decreasing solvent viscosity, breaking solute-matrix interactions, and forcing solvents into matrices that would not be reached under atmospheric conditions. This technique has already been successfully applied using olive pomace as raw material.²²

In general, the unconventional extraction techniques such as MAE and HPTE are considered more environmentally friendly than conventional methods, minimizing solvent use, toxicity, waste production, carbon dioxide (CO₂) emissions, and energy consumption while increasing polyphenol extraction efficiency.²³, ²⁴, ²⁵ Therefore, green extraction based on the design of processes that reduce energy consumption allows the use of alternative, renewable solvents and that guarantees safe and quality products.⁵, ²⁶

The most crucial aspect of plant extraction is the cellular structure, as valuable compounds such as essential oils and polyphenols are stored in the central vacuoles of guard and epidermal cells, as well as in the subepidermal cells of leaves, hindering extraction.¹⁹ Other factors that influence the extraction of bioactives include their chemical bonds and the presence of other compounds in the plant structure. For instance, polyphenols are seldom found in free form, as many are covalently bound to the plant cell wall, while others are present in waxes or on the outer surfaces of plant organs and are linked by glycosides.²⁷ In summary, the complexity of extracting active compounds from plants due to its cellular structure underscores the importance of advanced extraction techniques, such as MAE and HPTE, in overcoming these challenges. This understanding is crucial for the development of effective methods for extracting bioactive compounds, such as polyphenols, which have significant potential in various industrial and medicinal applications.

This study aims to compare the performance of MAE and HPTE for polyphenols and flavonoids extraction from the cupuaçu seed by-product, as well as evaluate the anti-radical power of the extracts. Factors such as ethanol concentration, extraction time, and solid/liquid ratio were investigated using a factorial experimental design approach, and using a conventional percolation extraction as a control.

Experimental

Preparation and processing of cupuaçu seed by-product

The cupuaçu seeds by-product (CSB) was obtained after pressing fruit to recover the crude oil. This material was donated by the Natura Cosmetical company (Benevides, PA, Brazil) and transported to the Cosmetics Laboratory of the Federal University of Pará, where it was kept at 8 °C in refrigerator (CRA34DBANA, Consul®, Joinville, SC, Brazil). The CSB was dried in an oven with air circulation at 40 ± 2 °C and then pulverized in a crusher up to a mean particle size of 710 µm.

Chemicals and reagents

To perform the colorimetric and chromatographic assays, the following standards and reagents were used: gallic acid (Fluka Chemika, Buchs, Switzerland), quercetin, (−)-epicatechin, (+)-catechin, 2,2’-azinobis(3-ethylbenzothiazoline-6-sulfonic acid) diammonium salt (ABTS^+•), (±)-6 hydroxy-2,5,7,8-tetramethyl-chromane-2-carboxylic acid (Trolox), quercetin 3-glucoside, p-coumaric acid, (−)-epigallocatechin-3-gallate, protocatechuic acid, Folin-Ciocalteu phenol reagent, methanol, acetonitrile, acetic acid, sodium persulfate (Sigma-Aldrich, St. Louis, MO, USA).

Conventional extraction by percolation

Conventional extraction by percolation (CE) was used as control for comparison of the extraction processes, consisted of solvent extraction with a 70% (m/v) of ethanol solution²⁸ to obtain the crude extract, the whole process took five days. The resulting solution was concentrated in a rotary evaporator (Laborota 4000, Heidolph, Schwabach, Germany), under low pressure and controlled temperature (40 ± 5 °C). The CE extract obtained in this way had a 78.85 ± 0.50% dry mass content.²⁹

Microwave assisted extraction

MAE was carried out in 100 mL closed polytetrafluorethylene (PTFE) vessels using a multimodal microwave laboratory oven (Ethos E, Milestone, Sorisole, BG, Italy) equipped with a fiber optic temperature sensor and an automatic temperature control. Samples were placed in vessels with 20 mL of ethanolic solution. Prior to extraction, nitrogen was bubbled into the vessel to provide an inert atmosphere useful for preventing degradation reactions. The extraction took place at a temperature of 120 °C, and then the extract was filtered through Whatman No. 2 filter paper (Sigma-Aldrich, St. Louis, MO, USA).

High pressure/temperature extraction

HPTE extraction was performed in a stirred reactor (model 4560, PARR Instrument Company, Moline, IL, USA), which contains appropriate valves to allow the introduction and removal of gases into/from the reaction chamber. To avoid oxidation of phenolics during the experiments, all tests were performed under a nitrogen atmosphere. After hermetically sealing the reactor, the extraction process was performed as described by Aliakbarian et al.²² The HPTE extraction was conducted at 12 bar at the same temperature (120 °C) applied for the extraction by MAE. The extract was filtered using Whatman No. 2 filter paper (Sigma-Aldrich, St. Louis, MO, USA).

Determination of total polyphenols content

The total polyphenols content (TPC) of extracts obtained by different processes (MAE, HPTE and CE) was determined by the Folin-Ciocalteu assay as described by Aliakbarian et al.,²² with some adaptations. For this, a UV-Vis spectrometer (PerkinElmer Lambda 20, Beaconsfield, United Kingdom) was used at 725 nm wavelength. The TPC was expressed in milligrams of gallic acid equivalents per gram of dry mass (mg GAE g⁻¹ DM). The calibration curve was constructed with an ethanolic solution of the gallic acid standard in the concentration range of 6.25-100.0 g L⁻¹.

Determination of total flavonoids content

The total flavonoids content (TFC) of the same extracts was determined by the colorimetric method described by Jemai et al.,³⁰ and adapted by Aliakbarian et al.,²² and expressed in milligrams of catechin equivalents per gram of dry mass (mg CE g⁻¹ DM). The absorbance of the extracts was determined at 510 nm, using the same spectrophotometer mentioned above. The calibration curve was constructed with a methanolic solution of the catechin standard in the concentration range of 0.01-0.50 g L⁻¹.

Determination of anti-radical power

The anti-radical power (ARP) of the same extracts was determined by the ABTS^+• method, according to the procedure described by Re et al.,³¹ with adaptations. The cationic radical ABTS^+• was prepared by reaction of 7.0 mM of ABTS^+• aqueous solution with 140 mM of potassium persulfate, and the mixture was kept under refrigeration for 16 h in the absence of light. At the time of analysis, the ABTS^+• solution was diluted and distilled. After diluiting 50 μL aliquots of the different extracts, 1.0 mL of the ABTS^+• solution was added, and the absorbance of samples was read at 734 nm after 2 min of reaction. The ARP of the extracts was calculated using a Trolox standard curve as Trolox equivalent antioxidant capacity (TE) and expressed in µg TE L⁻¹.

HPLC determination of phenolic compounds

The content of the main polyphenols contained in the optimized extracts (MAE, HPTE) and CE by percolation were also quantified by high performance liquid chromatography (HPLC). For this purpose, standard solutions of gallic acid, protocatechuic acid, (−)-epigallocatechin-3-gallate, (−)-epicatechin, p-coumaric acid and quercetin were prepared (0.5 mg mL⁻¹), while that of quercetin 3-glucoside was prepared at a concentration of 0.1 mg mL⁻¹. Before the analyses, standard solutions were diluted (1:2) with methanol, while the extracts, whose concentration was 100 mg mL⁻¹, were filtered through membranes with 0.22 μm-pore diameter (Millipore, Bedford, MA, USA). Analyses were performed-using a HPLC system, (model 1100, Agilent, Palo Alto, CA, USA), coupled with a photodiode array detector (PAD) (model 1260 Infinity, Agilent, Palo Alto, CA, USA), and equipped with a C18 reverse phase Eclipse plus column (4.5 × 250 mm) packed with 5 μm diameter particles (Agilent). Samples (20 µL) were analyzed at a constant flow rate of 1.0 mL min⁻¹ and a column temperature of 30 °C. The mobile phase was a gradient system, with mobile phase A (methanol/acetonitrile, 1:1) and B (1% acetic acid in water), varying the mobile phase B, for 0-5 min to 100%, 10 min to 95%, 30 min to 70%, 40 min to 60%, 45 min to 52%, 55 min to 30% 60-65 min to 0%. Chromatographic peaks of these analytes were detected at 280 nm for all phenolics.²⁹ Then, they were confirmed by comparing their retention time and PAD-UV spectra with those of reference standards. Quantification was carried out by integration of peaks using the external standard method.

Experimental design and statistical analysis

In order to maximize the performance of extraction by unconventional techniques, namely MAE and HPTE, a Box-Behnken 3³-factorial design was applied. Three levels were selected for each independent variable (−1, 0, +1), namely extraction time (30, 60 and 90 min), solid/liquid ratio (0.1, 0.05 and 0.03 g mL⁻¹) and ethanol concentration (65, 80, 95%), resulting in 15 experimental runs, including three replicates of the central point. Three responses were evaluated: namely TPC, TFC, and ARP, determined as antioxidant activity by the ABTS^+• method, whose experimental values adjusted to a second-order polynomial function.

The bioactive compounds and anti-radical power were determined in triplicate, and the results expressed as mean ± standard deviation. Statistical analyses were performed using Statistica 12 (StatSoft, Inc., Tulsa, USA).³² The regression was performed by least squares algorithm based on the second-order polynomial model coefficients (equation 1). Linear, quadratic, and interaction effects of the three independent variables were considered for the response variables.

(1) ${\text{y}}_{\text{k}}={\text{β}}_0+{\displaystyle {\sum }_{\text{i}=1}^{\text{k}}{\text{β}}_{\text{i}}{\text{x}}_{\text{i}}}+{\displaystyle {\sum }_{\text{i}=1}^{\text{k}}{\text{β}}_{\text{ii}}{\text{x}}_{\text{i}}^2}+{\displaystyle {\sum }_{\text{i}=1}^{\text{k}}{\displaystyle {\sum }_{\text{j}=\text{i}=1}^{\text{k}}{\text{β}}_{\text{ij}}{\text{x}}_{\text{i}}{\text{x}}_{\text{j}}}}$

where y_k (k = 1, 2 and 3) are the predicted (theoretical) values of responses, β₀ is the intercept coefficient, β_i are the coefficients of linear terms, β_ii are those of quadratic terms, and β_ij those of linear interaction terms, while x_i and x_j are the actual (experimental) values of the three independent variables (x₁, x₂ and x₃ for solvent concentration, extraction time and solid/liquid ratio, respectively). The data fitness for generated models were evaluated by factorial variance analysis (ANOVA) considering a 5% level of significance (p ≤ 0.05) and the determination coefficient (R²). The statistical significant polynomial models were used to predict the optimal responses (TPC, TFC and ARP). The model of each response was expressed in terms of real variables.

Multiple response optimization was performed by the desirability function. Theses optimized conditions were validated for the maximum extraction of bioactive compounds (TPC and TFC) and antioxidant activity (ABTS^+•) based on the values obtained. The experimental values were compared with predicted values based on coefficient of variation (CV%) to evaluate the precision and repeatability. The comparison between the methods were performed in optimum conditions of phenolics extraction yield by a variance analysis (two-way analysis of variance-ANOVA), followed by Tukey’s post-test, both tests with a 5% significance level.

Results and Discussion

The recovery of bioactive compounds presents in fruits, including phenolic compounds, strongly depends on the type of solvent and extraction conditions (such as concentration, pH, temperature, time, etc.) that directly impact their chemical structure, concentration and antioxidant activity.¹⁸, ³³, ³⁴, ³⁵ Therefore, to select the optimal conditions for extracting the bioactive compounds from CSB, two emergent extraction techniques, namely MAE and HPTE, were employed. For this purpose, experiments were conducted following a Box-Behnken 3³-factorial design, with ethanol concentration (ethanol, %), extraction time (time, min) and solid/liquid ratio (S/L, g mL⁻¹) chosen as the independent variables, and the TPC, TFC, and ARP, selected as the responses.

Effects of the independent variables

Table 1 shows the experimental design matrix with real values of independent variables and responses for both MAE and HPTE, while Table 2 lists the regression coefficients of intercept, linear, quadratic and interaction terms of the second order polynomial models determined by the least square’s technique. None of the quadratic effects (ethanol²; time² and S/L²) were statistically sigificant (p < 0.05), indicating a first-order (linear) relationship between independent variables and responses.

Table 1
3³-Box-Behnken experimental matrix with real values of variables for both microwave-assisted extraction (MAE) and high pressure/temperature extraction (HPTE)

Run	Ethanol / %	time / min	S/L ratio / (g mL−1)	MAE			HPTE
				TPC / (mg GAE g−1 DM)	TFC / (mg CE g−1 DM)	ARP / (µg TE L−1)	TPC / (mg GAE g−1 DM)	TFC / (mg CE g−1 DM)	ARP / (µg TE L−1)
1	65	90	0.05	93.02 ± 1.42	19.88 ± 3.20	23.30 ± 2.39	53.43 ± 1.95	30.54 ± 4.23	31.27 ± 0.72
2	95	60	0.10	11.91 ± 0.40	5.77 ± 0.93	18.69 ± 1.41	15.67 ± 0.25	25.56 ± 2.04	22.20 ± 1.09
3	65	60	0.03	120.63 ± 0.68	20.90 ± 2.11	13.98 ± 0.82	48.85 ± 2.27	25.78 ± 0.30	25.50 ± 5.08
4	80	60	0.05	31.52 ± 1.09	16.21 ± 2.59	13.98 ± 0.03	39.13 ± 4.34	21.87 ± 2.03	30.28 ± 2.00
5	95	30	0.05	11.53 ± 1.13	9.33 ± 1.89	6.81 ± 0.73	18.47 ± 0.35	11.02 ± 0.33	15.23 ± 0.23
6	80	60	0.05	31.40 ± 0.21	13.36 ± 0.84	13.80 ± 1.61	33.62 ± 3.54	18.92 ± 0.58	21.51 ± 1.59
7	80	90	0.10	26.38 ± 1.06	12.89 ± 0.82	23.04 ± 1.05	27.41 ± 1.56	16.87 ± 0.77	29.23 ± 1.44
8	80	60	0.05	29.83 ± 2.11	12.14 ± 1.08	14.63 ± 0.08	32.47 ± 0.14	22.06 ± 1.14	18.41 ± 1.17
9	95	90	0.05	14.31 ± 1.55	5.86 ± 0.27	8.06 ± 0.08	13.21 ± 0.35	28.95 ± 5.76	13.58 ± 0.43
10	80	30	0.03	31.29 ± 2.56	16.77 ± 0.20	11.18 ± 1.69	39.65 ± 2.41	19.24 ± 1.52	18.09 ± 0.66
11	65	60	0.10	42.96 ± 2.91	13.31 ± 1.14	29.50 ± 0.48	44.16 ± 1.96	18.57 ± 2.65	55.72 ± 3.36
12	65	30	0.05	58.36 ± 1.67	19.62 ± 0.54	20.54 ± 0.67	44.64 ± 1.96	21.87 ± 2.45	27.80 ± 1.11
13	95	60	0.03	13.96 ± 0.99	13.42 ± 0.53	5.29 ± 0.06	21.63 ± 3.33	32.31 ± 1.57	11.10 ± 1.15
14	80	90	0.03	32.84 ± 1.08	16.35 ± 1.61	11.06 ± 0.32	44.50 ± 5.11	36.15 ± 3.71	19.11 ± 0.86
15	80	30	0.10	31.46 ± 1.37	10.98 ± 0.84	21.15 ± 0.80	26.61 ± 2.92	19.11 ± 1.49	28.62 ± 2.11

• S/L ratio: solid/liquid ratio; TPC: total phenolic content; TFC: total flavonoids content; ARP: anti-radical power according to the ABTS^+• method; GAE: gallic acid equivalents; CE: catechin equivalents; DM: dry mass; TE: Trolox equivalent antioxidant capacity. Results are expressed as mean ± standard deviation.

Table 2
Regression coefficients of the second-order polynomial models for the microwave-assisted extraction (MAE) and high-pressure/temperature extraction (HPTE)

Effect	MAE			HPTE
	TPC / (mg GAE g−1 DM)	TFC / (mg CE g−1 DM)	ARP / (µg TE L−1)	TPC / (mg GAE g−1 DM)	TFC / (mg CE g−1 DM)	ARP / (µg TE L−1)
Average	30.92a	13.90a	14.14a	35.08a	20.95	23.40
Linear
Ethanol	−32.91a	−4.92a	−7.06a	−15.26a	0.13	−9.77a
time	4.24	−0.22	0.72a	1.15	5.16a	0.43
S/L	10.75	3.06a	−5.36a	5.10a	4.17	−7.75a
Interaction
Ethanol time	−7.97	−0.93	−0.38a	−3.51	2.32	−1.28
Ethanol S/L	−18.90a	0.015	2.53	0.32	−0.11	4.78
time S/L	1.65	−0.59	−0.50	1.01	4.79	0.10
Quadratic
Ethanol2	15.13	−0.57	−0.60	−2.30	2.43	1.72
time2	−1.74	0.33	1.14	−0.33	−0.28	−3.15
S/L2	1.32	0.0084	1.33	−0.20	2.18	3.51

• ^a
  Statistically significant according to the Student’s t-test (p ≤ 0.05); S/L ratio: solid/liquid ratio; TPC: total polyphenol content of cupuaçu seed by-product; TFC: total flavonoid content of cupuaçu seed by-product; ARP: anti-radical power according to the ABTS^+• method; GAE: gallic acid equivalents; CE: catechin equivalents; DM: dry mass; TE: Trolox equivalent antioxidant capacity.

The linear effect of ethanol was negative and statistically significant for all responses, except for TFC extracted by HPTE (Table 2; Figure 1). Such a rather uniform behavior indicates that the increase in ethanol proportion in the hydroalcoholic solvent reduced the extraction yield of bioactive compounds by both MAE and HPTE. This effect was the highest for TPC by MAE and the lowest, although not statistically significant, for TFC by HPTE (Table 2). It has been reported¹⁸ in this respect that when the microwave absorption capacity of the plant material is higher than that of the solvent, the extraction yield may change significantly. Ethanol had the most significant effect on the three responses using both MAE and HPTE, which means that, regardless of the extraction method, this independent variable had the greatest influence on the extraction of polyphenols, flavonoids (except for HPTE) and antioxidant activity under the investigated conditions. In particular, this effect on TPC using MAE was twice that observed using HPTE, and a qualitatively similar behavior was observed for TFC and ARP (Figure 1).

Figure 1
Pareto diagram of total polyphenol content (TPC) and total flavonoid content (TFC) showing which extraction parameters (ethanol, extraction time and solid/liquid ratio) influence the content of bioactive compounds in the cupuaçu seed by-product obtained by microwave-assisted extraction (MAE) and extraction at high pressure and temperature (HPTE).

The linear effect of the extraction time was positive and statistically significant only for TFC extracted by HPTE under the conditions investigated in the present study (Table 2; Figure 1d), which means that any attempt to improve the extraction performance by acting on this variable should be specific for the desired bioactive.

The S/L ratio linear effect was statistically significant and positive for TFC using MAE and for TPC using HPTE, but negative for ARP using both (Table 2). While this effect on TFC extracted by MAE was the second most significant one after that of ethanol, it was not statistically significant on TFC extracted by HPTE (p > 0.05), and exactly the opposite occurred with TPC. Such inverse and direct proportional relationships of S/L ratio with ARP on the one hand and with TFC and TPC on the other, leave no space to act on this independent variable to enhance the process yield unless the goal is to maximize only one of these responses.

Among the interaction effects, only the one between ethanol and S/L ratio was statistically significant for TPC obtained by MAE (Figure 1a), being the second most significant one among all. The inverse proportional relationship evidenced by the negative value of its coefficient indicates that TPC recovery may be improved using microwave-assisted extraction by reducing ethanol or S/L ratio (Table 2).

In summary, ethanol was the independent variable with the greatest influence on the three responses using HPTE and mainly MAE, showing inverse proportional relationships with them under the conditions tested in this study (Figure 1). On the other hand, the effect of the extraction time was statistically significant only for TFC extracted by HPTE, with a direct proportional relationship (Figure 1d). The positive effect of the S/L ratio on TPC and the negative one on ARP, which was stronger using HPTE rather than MAE, corroborates the observation that the increase in the proportion of solids usually increases the recovery of polyphenols (Table 2). The interaction effect between ethanol and S/L ratio was significant for TPC obtained by MAE, indicating an inverse relationship; therefore, we can consider as the best results those obtained using the lowest levels of ethanol concentration in the solvent and of solid proportion.

Statistical evaluation by analysis of variance

Table 3 lists the results of the analysis of variance (ANOVA) applied to the regression of TPC, TFC and ARP data obtained using both MAE and HPTE. The second-order polynomial models were evaluated considering all parameters (effects), while the residues were evaluated fractionally as pure error and lack of fit.

Table 3
Analysis of variance (ANOVA) applied to models proposed for microwave-assisted extraction (MAE) and high pressure/temperature extraction (HPTE) of cupuaçu by-product polyphenols

	MAE
	TPC	TFC	ARP
F-value	7.56	5.71	19.57
p-value	0.0192a	0.0347a	0.0022a
Lack of fit (p-value)	0.0029a	0.4467	0.0311a
R2	0.9315	0.9114	0.9724
	HPTE
	TPC	TFC	ARP
F-value	13.67	2.50	4.77
p-value	0.0051a	0.1630	0.0501a
Lack of fit (p-value)	0.4009	0.0819	0.5929
R2	0.9609	0.8180	0.8956

• ^a
  Statistically significant (p ≤ 0.05); R²: determination coefficient; TPC: total polyphenol content of cupuaçu seed by-product; TFC: total flavonoid content of cupuaçu seed by-product; ARP: anti-radical power according to the ABTS^+• method.

All three response model regressions were statistically significant (p ≤ 0.05) using MAE (Table 3). Although the TFC model using HPTE was not statistically significant in term of lack of fit, it was able to explain the variance of the experimental data and could be used for prediction. In fact, the lack of fit indicates that the pure error (reproducibility error) did not interfere with the model regression variance, suggesting that this variance was solely to deviations between predicted and experimental values. While evaluating HPTE by ANOVA, the TPC and ARP responses were also statistically significant (p ≤ 0.05) (Table 3).

Response surface prediction

Considering only the statistically significant regression models with non-significant lack of fit (p > 0.05) (TFC for MAE and TPC and ARP for HPTE) (Table 3), response surface graphs were generated to visualize the prediction of the polynomial models.

It can be seen in Figure 2 that the use of MAE and HPTE led to a similar trend in the TPC yield, with the highest values of this response being obtained at the extraction time of approximately 40 min and the lowest ethanol concentration in the hydroalcoholic solvent. This suggests a fundamental role of water in the extraction of phenolic compounds from CSB. These results are consistent with those of other studies where MAE was used to recover phenolics from different raw materials.³⁶ To provide only a few examples, the hydroalcoholic mixture (water/EtOH, 50%) was the best extraction solvent allowing the highest polyphenol extraction yield from coffee grounds,¹³ pomegranate (Punica granatum L.) peels,³⁷ and onion solid waste.³⁸

Figure 2
Surface plot of total polyphenol content (TPC) and total flavonoid content (TFC) as a function of ethanol concentration/time extraction influence the content in the cupuaçu seed by-product obtained by microwave-assisted extraction (MAE) and extraction at high pressure and temperature (HPTE).

It is well known that the presence of water, either added or naturally present in the sample, enhances the absorption capacity of microwaves and facilitates its heating due to the increase in the polarity of the extraction solvent. Water can also favor matrix swelling and/or interactions among analytes, making them more readily available for extraction.³⁹ Although ethanol-water mixtures allow the extraction not only of polar compounds but also of weak polar compounds and even non-polar compounds at relatively high ethanol concentrations, the solvent can dehydrate plant cells, hindering the extraction of polyphenols and the diffusion of the plant matrix into the liquid, resulting in reduced extraction yields.¹¹, ⁴⁰

The reduction of ethanol concentration in TPC yield was also observed using HPTE showing better recovery of bioactive compounds (Figure 2c). Machado et al.⁴¹ and Alexandre et al.⁴² reported that the use of 50 and 56% ethanol concentrations in pressurized liquid extraction maximized phenolics extraction from blackberry (Rubus fruticosus L.) residue and from pomegranate peels, respectively. On the contrary, Barrales et al.⁴³ found, using the same extraction process, that ethanol concentration had no significant effect on phenolics extraction from citrus by-products.

Although the extraction time did not show statistical significance for almost all response, except for TFC using HPTE (Figure 1d), it was maintained in the models to avoid a reduction in the determination coefficients and, thus to preserve their hierarchy. On the other hand, Ben Hamissa et al.⁴⁴ found that time was statistically significant for TPC and TFC extraction from Agave americana L. leaves by HPTE.

Figure 2 also illustrates the response surfaces of TFC extracts as a function of ethanol concentration/time during extraction of total flavonoids content prepared by MAE and HPTE. One can see that, differently from TPC, MAE and HPTE led to different TFC trends, and that MAE led to similar TPC and TFC trends. On the other hand, in HPTE, the extraction time had the strongest positive effect on TFC yield, followed by S/L ratio and its interaction term (Figure 1d). Alexandre et al.⁴² found that ethanol concentration (quadratic and linear terms) had the highest positive effect on flavonoids extraction by HPTE from pomegranate peels, while it was observed also a positive effect of ethanol concentration and S/L ratio on flavonoids extraction from propolis under high pressure.⁴⁵

As shown in Figure 3, which depicts the ARP response surfaces for extracts produced by both MAE and HPTE, the lower the ethanol concentration in the solvent, the higher this response, while the extraction time did not exert any significant influence (Figure 3c), consistently with the results of Table 2.

Figure 3
(a)-(c) Pareto diagram of anti-radical power showing which extraction parameters (ethanol, extraction time and solid/liquid ratio) influence the content of bioactive compounds in the cupuaçu seed by-product obtained by microwave-assisted extraction (MAE) and extraction at high pressure and temperature (HPTE), (b)-(d) surface plot of anti-radical power as a function of ethanol concentration/time extraction.

Optimization and experimental validation

Based on the criteria imposed to select the optimal extraction conditions, i.e., highest extraction efficiency, lowest ethanol-in-water concentration, the optimal conditions predicted by the desirability function for CSB extraction were 65% (m/v) ethanol, extraction time of 45 min, and S/L ratio of 0.03 g mL⁻¹, with a desirability value of 0.783, using MAE, and 65% (m/v) ethanol, extraction time of 42.18 min, and S/L ratio of 0.03 g mL⁻¹, with a desirability of 0.679, using HPTE. Correlation of the predicted values with the experimental ones resulted in determination coefficients (R²) of 99.87% for MAE and 96.65% for HPTE (Table 4).

Table 4
Experimental validation of the optimum conditions of microwave-assisted extraction (MAE) and high pressure/temperature extraction (HPTE)

Response	Predicted value	Experimental value	Difference (CV) / %
	MAE
TPC / (mg GAE g−1 DM)	120.63	103.17 ± 0.2	0.18
TFC / (mg CE g−1 DM)	21.68	21.88 ± 1.5	6.88
ARP / (µg TE L−1)	16.89	22.70 ± 0.7	3.53
R2			99.87
		HPTE
TPC / (mg GAE g−1 DM)	57.31	55.45 ± 0.3	0.53
TFC / (mg CE g−1 DM)	36.15	32.74 ± 1.2	3.48
ARP / (µg TE L−1)	25.06	30.37 ± 0.7	2.52
R2			96.65

• CV: coefficient of variation; TPC: total polyphenol content of cupuaçu seed by-product; TFC: total flavonoid content of cupuaçu seed by-product; ARP: anti-radical power according to the ABTS+ method; R²: determination coefficient.

Kaderides et al.³⁷ used MAE to extract phenolics from pomegranate peels, found optimum operating conditions of 50% ethanol and solid ratio of 0.02 g mL⁻¹. Using the same process to extract phenolics, and optimum solid/solvent ratio (0.04 g mL⁻¹) was reported by Mellinas et al.⁴⁶ for cocoa bean shell (Theobroma cacao), while the ethanol concentration was very close to that predicted in the present work (66.2%) was reported for tomatoes peels.⁴⁷ On the other hand, Pineiro et al.⁴⁸ found as optimum conditions for grape cultivars a solid/solvent ratio of 0.30 g mL⁻¹, using 50% methanol in water (pH 2) at 70 °C, and an extraction time of only 3 min. These results as a whole confirm that the optimal conditions of MAE strongly depend on the starting raw material.

The total yield of TPC obtained under optimized conditions with MAE (103.17 ± 0.2 mg GAE g⁻¹ DM) was 86% higher than that obtained with HPTE, while the yield of TFC (21.88 ± 1.5 mg CE g⁻¹ DM) and ARP (22.70 ± 0.7 µg TE L⁻¹) were 33 and 25% lower, respectively (Table 4). Comparing these values with those previously obtained by Costa et al.²⁹ for the same by-product obtained by percolation (CE) using 70% (m/v) ethanol (TPC = 16.9 ± 1.8 mg GAE g⁻¹ DM, TFC = 5.92 ± 3.42 mg CE g⁻¹ DM, ARP = 74.89 ± 5.5 µg TE L⁻¹), it is evident that both proposed non-conventional extraction processes can effectively extract both total polyphenols and flavonoids, with less solvent expenditure and in shorter time.

The high performance of MAE, especially in terms of TPC yield, can be related to the relative contribution of energy conversion mechanisms, i.e., dipole rotation and ionic conduction.¹⁷ For small molecules like water and some other solvents like ethanol, dipole rotation decreases, while ionic conduction increases with increasing temperature; therefore, as samples containing both ions and other compounds are heated, the contribution of dipole rotation dominates at the beginning, while that of ionic conduction becomes prevalent at the temperature selected for extraction (120 °C in this case). Therefore, the relative contribution of these two heating mechanisms depends on the mobility and concentration of ions in the samples and on their relaxation time.¹⁷, ³⁹

Surprisingly, the positive effect of high temperature on extraction, expected by increasing the diffusion rate and solubility of the analytes,⁴⁹ when combined with high pressure in the HPTE process, increased the TPC yield compared to CE, exalting the TFC yield and ARP. Nonetheless, HPTE was successful in the extraction of TPC from olive pomace.²², ⁵⁰ These results taken together highlight the use of HPTE as an alternative efficient technique to recover of phenolic compounds and flavonoids from solid residues from plant matrices, including the CSB (Table 4), but also suggest to perform specific preliminary lab-scale test for each specific residue to select the most effective extraction technique to recover the desired class of bioactives.

In summary, although both MAE and HPTE proved to be efficient techniques for the extraction of phenolic compounds from the cupuaçu seed by-product, MAE ensured the highest TPC yields consistently with its ability to absorb energy under the experimental conditions adopted in this study. However, the best results are expected when the plant material is heated selectively, that is, when the best dielectric properties of the solvent are ensured. In other words, by properly correlating the dielectric constant of the plant material with that of the solvent, the former may reach higher temperatures than the solvent, consequently resulting in the rupture of cell membrane and the release of bioactive components.³⁴, ⁵¹ Overall, MAE proved to be more attractive than HPTE for extracting bioactive compounds from CSB, due to greater solvent penetration power and mass transfer, reduced solvent volume, increased overall yield of bioactive compounds, and lower process costs.⁵²

HPLC determination of phenolic compounds

HPLC was used to identify and quantify the main phenolic compounds present in the CSB extracts prepared under optimized conditions by MAE, HPTE and CE, with results gathered in Table 5.

Table 5
Phenolic compounds of cupuaçu seed by-product extracted by microwave-assisted extraction (MAE), high pressure/temperature extraction (HPTE), and percolation extraction (CE), quantified by high performance liquid chromatography (HPLC)

	Compound / (g L−1)
	MAE	HPTE	CE
Gallic acid	0.02 ± 0.01a	0.04 ± 0.02a	0.02 ± 0.01a
Quercetin	0.46 ± 0.3a	0.49 ± 0.17a	0.04 ± 0.00b
Quercetin 3-glucoside	0.16 ± 0.07a	0.24 ± 0.12a	0.09 ± 0.05b
Protocatechuic acid	0.06 ± 0.01a	0.08 ± 0.02a	0.15 ± 0.04b
p-Coumaric acid	0.04 ± 0.03a	0.05 ± 0.01a	0.00 ± 0.00a
(−)-Epicatechin	0.29 ± 0.23a	0.43 ± 0.04a	0.08 ± 0.03b
(−)-Epigallocatechin-3-gallate	2.39 ± 1.39a	3.12 ± 0.84a	0.04 ± 0.00b

• Means within a row with different letters are significantly different by Tukey’s post hoc test (p ≤ 0.05).

With the exception of protocatechuic acid, both unconventional extraction techniques were able to extract higher levels of polyphenols from the CSB compared to CE by percolation. Quercetin, (−)-epicatechin and (−)-epigallocatechin-3-gallate obtained by extraction with both HPTE and MAE are the majors compounds in extracts when compared to CE. Likewise, a greater extraction recovery of (−)-epicatechin (48%) and (−)-epigallocatechin-3-gallate (30%) was obtained by HPTE compared to MAE, and about 3.5 and 59.0 times higher compared to CE, respectively (Table 5).

All applied processes allowed the extraction of these substances present in the by-product, regardless of the type of extraction process applied, except for p-coumaric acid, which was not detected after exhaustive extraction by percolation. Thus, the mechanisms used in the extraction processes interfered with the extraction of polyphenols from the CSB, when identified and quantified by type of active substance by HPLC.

The total concentration of phenolic compounds determined by HPLC is 3.67% lower than the sum of concentrations of the individual phenolic compounds expressed as gallic acid equivalents (mg GA g⁻¹ DM). This difference may have been due not only to the lack of measurement by HPLC of some minor phenolic compounds present in the by-product extracts but also to the presence of some non-phenolic antioxidant compounds interfering with the electron transfer reaction with the Folin-Ciocalteu reagent.²² Furthermore, as different phenolic compounds have very different reaction stoichiometries, the latter quantification method may have caused an overestimation of the total phenolic content of the cupuaçu seed by-product.⁵³

Conclusions

Microwave-assisted (MAE) and high pressure/temperature extractions (HPTE) are already known as effective techniques to extract polyphenols compared to conventional extraction by percolation to recover polyphenols from vegetable by-products. Our study shows that there are differences for phenolics profile extraction between MAE and HPTE. In optimum conditions, MAE had a major total phenolics extraction yield compared to HPTE. By other side, HPTE had major total flavonoids extraction yield. Moreover, the central composite design (CCD) study showed different relations between the input variables and responses for each method, suggesting that the main variables and effects are different for each process; and each method affects differently the phenolics interaction inside the cupuaçu by-product matrix. Considering that the mild conditions of these processes can preserve the biological properties of bioactive compounds, its use promises to be of great significance for the exploitation of the cupuaçu seed by-product. The experimental results of the design and the regression method were successful in optimizing them simultaneously. Therefore, both MAE and HPTE extractions show themselves as viable options for the extraction of polyphenols in cupuaçu seed by-product.

Acknowledgments

The authors would like to thank the Brazilian Coordination for the Improvement of Higher Education Personnel (CAPES) (grant No. 001) for the financial support as sandwich PhD fellowship of Russany S. da Costa, process No. 99999.003074/2015-03. The authors would also like to thank PROPESP-UFPA for the financial resources provided for article publication.

References

• ¹
  Carvalho, J. E. U.; Müller, C. H.; Alves, R. M.; Nazaré, R. F. R.; Comunicado Técnico 115, Cupuaçuzeiro, 1^st ed.; Embrapa Amazônia Oriental: Belém, 2004. [Link] accessed in July 2024
  » Link
• ²
  Yang, H.; Protiva, P.; Cui, B.; Ma, C.; Baggett, S.; Hequet, V.; Mori, S.; Weinstein, I. B.; Kennelly, E. J.; J. Nat. Prod. 2003, 66, 1501. [Crossref]
  » Crossref
• ³
  Cuenca, M. A. G.; Nazário, C. C.; Importância Econômica e Evolução da Cultura do Cacau no Brasil e na Região dos Tabuleiros Costeiros da Bahia entre 1990 e 2022, 1^st ed.; Embrapa Tabuleiros Costeiros: Aracaju, 2004. [Link] accessed in July 2024
  » Link
• ⁴
  Verma, N.; Himshweta, In Green Sustainable Process for Chemical and Environmental Engineering and Science; Altalhi, T.; Inamuddin, eds.; Elsevier: London, 2023, ch. 15. [Crossref]
  » Crossref
• ⁵
  Tlais, A. Z. A.; Fiorino, G. M.; Polo, A.; Filannino, P.; Di Cagno, R.; Molecules 2020, 25, 2987. [Crossref]
  » Crossref
• ⁶
  Pugliese, A. G.; Tomas-Barberan, F. A.; Truchado, P.; Genovese, M. I.; J. Agric. Food Chem. 2013, 61, 2720. [Crossref]
  » Crossref
• ⁷
  Casazza, A. A.; Aliakbarian, B.; Mantegna, S.; Cravotto, G.; Perego, P.; J. Food Eng. 2010, 100, 50. [Crossref]
  » Crossref
• ⁸
  Chemat, F.; Abert Vian, M.; Fabiano-Tixier, A. S.; Nutrizio, M.; Režek Jambrak, A.; Munekata, P. E. S.; Lorenzo, J. M.; Barba, F. J.; Binello, A.; Cravotto, G.; Green Chem. 2020, 22, 2325. [Crossref]
  » Crossref
• ⁹
  Silva, R.; Brand, A. L.; Tinoco, N.; Freitas, S.; Rezende, C.; J. Braz.. Chem. Soc. 2024, 55, 3. [Crossref]
  » Crossref
• ¹⁰
  Bakić, M. T.; Pedisić, S.; Zorić, Z.; Dragović-Uzelac, V.; Grassino, A. N.; Acta Chim. Slov. 2019, 66, 367. [Crossref]
  » Crossref
• ¹¹
  Feki, F.; Klisurova, D.; Masmoudi, M. A.; Choura, S.; Denev, P.; Trendafilova, A.; Chamkha, M.; Sayadi, S.; Food Chem. 2021, 556, 129670. [Crossref]
  » Crossref
• ¹²
  Moreira, M. M.; Barroso, M. F.; Boeykens, A.; Withouck, H.; Morais, S.; Delerue-Matos, C.; Ind. Crops Prod. 2017, 104, 210. [Crossref]
  » Crossref
• ¹³
  Pettinato, M.; Casazza, A. A.; Perego, P.; FoodBioprod. Process. 2019, 114, 227. [Crossref]
  » Crossref
• ¹⁴
  Ekezie, F. G. C.; Sun, D. W.; Cheng, J. H.; Trends Food Sci. Technot. 2017, 67, 160. [Crossref]
  » Crossref
• ¹⁵
  Felkai-Haddache, L.; Remini, H.; Dulong, V.; Mamou-Belhabib, K.; Picton, L.; Madani, K.; Rihouey, C.; Food Bioprocess Technot. 2016, 9, 481. [Crossref]
  » Crossref
• ¹⁶
  Périno-Issartier, S.; Huma, Z.; Abert-Vian, M.; Chemat, F.; Food Bioprocess Technot. 2011, 4, 1020. [Crossref]
  » Crossref
• ¹⁷
  Vinatoru, M.; Mason, T. J.; Calinescu, I.; TrAC, Trends Anat. Chem. 2017, 97, 159. [Crossref]
  » Crossref
• ¹⁸
  Álvarez, A.; Fayos-Fernández, J.; Monzó-Cabrera, J.; Cocero, M. J.; Mato, R. B.; J. Food Eng. 2017, 197, 98. [Crossref]
  » Crossref
• ¹⁹
  Díaz-Ortiz, A.; Prieto, P.; De La Hoz, A.; Account, P.; Chem. Rec. 2019, 19, 85. [Crossref]
  » Crossref
• ²⁰
  Chan, C. H.; Yeoh, H. K.; Yusoff, R.; Ngoh, G. C.; J. Food Eng. 2016, 188, 98. [Crossref]
  » Crossref
• ²¹
  Saleh, I. A.; Vinatoru, M.; Mason, T. J.; Abdel-Azim, N. S.; Aboutabl, E. A.; Hammouda, F. M.; Uttrason. Sonochem. 2016, 31, 330. [Crossref]
  » Crossref
• ²²
  Aliakbarian, B.; Casazza, A. A.; Perego, P.; Food Chem. 2011, 128, 704. [Crossref]
  » Crossref
• ²³
  Boukroufa, M.; Boutekedjiret, C.; Petigny, L.; Rakotomanomana, N.; Chemat, F.; Uttrason. Sonochem. 2015, 24, 72. [Crossref]
  » Crossref
• ²⁴
  Chen, J.; Liu, M.; Wang, Q.; Du, H.; Zhang, L.; Motecutes2016, 21, 1383. [Crossref]
  » Crossref
• ²⁵
  Saifullah, M.; Mccullum, R.; Van Vuong, Q.; Processes 2021, 9, 2212. [Crossref]
  » Crossref
• ²⁶
  Moro, K. I. B.; Bender, A. B. B.; da Silva, L. P.; Penna, N. G.; Food Bioprocess Technot. 2021, 14, 1407. [Crossref]
  » Crossref
• ²⁷
  Ismail-Suhaimy, N. W.; Abd Gani, S. S.; Zaidan, U. H.; Halmi, M. I. E.; Bawon, P.; Ptants 2021, 10, 682. [Crossref]
  » Crossref
• ²⁸
  Farmacopeia dos Estados Unidos do Brasit, 1^st ed.; Editora Indústria Gráfica Siqueira S.A.: Rio de Janeiro, 1959.
• ²⁹
  da Costa, R. S.; dos Santos, O. V.; Lannes, S. C. S.; Casazza, A. A.; Aliakbarian, B.; Perego, P.; Ribeiro-Costa, R. M.; Converti, A.; Silva Júnior, J. O. C.; Food Sci. Technot. 2020, 40, 401. [Crossref]
  » Crossref
• ³⁰
  Jemai, H.; Bouaziz, M.; Sayadi, S.; J. Agric. Food Chem. 2009, 57, 2961. [Crossref]
  » Crossref
• ³¹
  Re, R.; Pellegrini, N.; Proteggente, A.; Pannala, A.; Yang, M.; Rice-Evans, C.; Free Radicat Biot. Med. 1999, 26, 1231. [Crossref]
  » Crossref
• ³²
  Statistica® version 12; StatSoft Power Solutions, Inc., Tulsa, OK, USA, 2012.
• ³³
  Ignat, I.; Volf, I.; Popa, V. I.; Food Chem. 2011, 126, 1821. [Crossref]
  » Crossref
• ³⁴
  Lee, C. S.; Binner, E.; Winkworth-Smith, C.; John, R.; Gomes, R.; Robinson, J.; Chem. Eng. Sci. 2016, 149, 97. [Crossref]
  » Crossref
• ³⁵
  Wang, W.; Jung, J.; Tomasino, E.; Zhao, Y.; LWT--Food Sci. Technot. 2016, 72, 229. [Crossref]
  » Crossref
• ³⁶
  Keskin Çavdar, H.; Bilgin, H.; Fadiloğlu, S.; Yilmaz, F. M.; Eur. J. Lipid Sci. Technot. 2023, 125, 2200089. [Crossref]
  » Crossref
• ³⁷
  Kaderides, K.; Papaoikonomou, L.; Serafim, M.; Goula, A. M.; Chem. Eng. Process. 2019, 137, 1. [Crossref]
  » Crossref
• ³⁸
  Imeneo, V.; de Bruno, A.; Piscopo, A.; Romeo, R.; Poiana, M.; Sustainabitity 2022, 14, 4387. [Crossref]
  » Crossref
• ³⁹
  de Castro, M. D. L.; Castillo-Peinado, L. S. In Innovative Food Processing Technotogies, 1^st ed.; Kai Knoerzer, P. J.; Smithers, G., eds.; Elsevier: London, 2016, p. 57. [Crossref]
  » Crossref
• ⁴⁰
  Sant’Anna, V.; Brandelli, A.; Marczak, L. D. F.; Tessaro, I. C.; Sep. Purif. Technot. 2012, 100, 82. [Crossref]
  » Crossref
• ⁴¹
  Machado, A. P. D. F.; Pasquel-Reátegui, J. L.; Barbero, G. F.; Martínez, J.; Food Res. Int. 2015, 77, 675. [Crossref]
  » Crossref
• ⁴²
  Alexandre, E. M. C.; Araújo, P.; Duarte, M. F.; de Freitas, V.; Pintado, M.; Saraiva, J. A.; Food Bioprocess Technot. 2017, 10, 886. [Crossref]
  » Crossref
• ⁴³
  Barrales, F. M.; Silveira, P.; Barbosa, P. P. M.; Ruviaro, A. R.; Paulino, B. N.; Pastore, G. M.; Macedo, G. A.; Martinez, J.; Food Bioprod. Process. 2018, 112, 9. [Crossref]
  » Crossref
• ⁴⁴
  Ben Hamissa, A. M.; Seffen, M.; Aliakbarian, B.; Casazza, A. A.; Perego, P.; Converti, A.; Food Bioprod. Process. 2012, 90, 17. [Crossref]
  » Crossref
• ⁴⁵
  Shouqin, Z.; Jun, X.; Changzheng, W.; J. Chem. Technot. Biotechnot. 2005, 80, 50. [Crossref]
  » Crossref
• ⁴⁶
  Mellinas, A. C.; Jimenez, A.; Garrigós, M. C.; LWT--Food Sci. Technot. 2020, 127, 109361. [Crossref]
  » Crossref
• ⁴⁷
  Li, H.; Deng, Z.; Wu, T.; Liu, R.; Loewen, S.; Tsao, R.; Food Chem. 2012, 130, 928. [Crossref]
  » Crossref
• ⁴⁸
  Pineiro, Z.; Aliaño-González, M. J.; González-de-Peredo, A. V.; Palma, M.; de Andrés, M. T.; Eur. Food Res. Technot. 2022, 248, 1883. [Crossref]
  » Crossref
• ⁴⁹
  Ju, Z. Y.; Howard, L. R.; J. Agric. Food Chem. 2003, 51, 5207. [Crossref]
  » Crossref
• ⁵⁰
  Alu’datt, M. H.; Rababah, T.; Alhamad, M. N.; Gammoh, S.; Ereifej, K.; Kubow, S.; Alli, I.; Food Hydrocottoids 2016, 61, 119. [Crossref]
  » Crossref
• ⁵¹
  Li, Y.; Fabiano-Tixier, A. S.; Vian, M. A.; Chemat, F.; TrAC, Trends Anat. Chem. 2013, 47, 1. [Crossref]
  » Crossref
• ⁵²
  Nawirska-Olszańska, A.; Stępień, B.; Biesiada, A.; LWT--Food Sci. Technot. 2017, 77, 276. [Crossref]
  » Crossref
• ⁵³
  Hrncirik, K.; Fritsche, S.; Eur. J. Lipid Sci. Technot. 2004, 106, 540. [Crossref]
  » Crossref

Edited by

Editor handled this article: Hector Henrique F. Koolen (Associate)

Publication Dates

• Publication in this collection
  23 Aug 2024
• Date of issue
  2025

History

• Received
  08 Apr 2024
• Accepted
  02 Aug 2024