Characterization of dark chocolates based on polyphenolic profiles and antioxidant activity

Abstract

Dark chocolates were characterized according to geographical origin, cocoa variety, and cocoa content using the methylxanthine and polyphenolic composition and antioxidant activity as the data. The main study objective was to uncover sample patterns and identify possible markers of quality, variety, or origin to deal with authentication or fraud detection issues. In the study, a set of 26 dark chocolates from different varieties (e.g., Criollo, Forastero, and Trinitario) harvested in Africa, America, and Asia was analyzed. The optimized sample treatment consisted of defatting the chocolate (1 g of sample with 5 mL of cyclohexane for 15 min, three times) and then extracting the analytes by sonication with methanol/water 60:40 (v:v) for 15 min. The filtered extracts were analyzed by reversed‐phase high‐performance liquid chromatography with UV and spectrophotometric methods (Folin–Ciocalteu, ferric reducing antioxidant power, and aluminum methods) to determine individual phenolics and overall indexes of antioxidant and flavonoid content. Results from this chocolate set indicated that American samples are richer than African counterparts in alkaloids and phenolics (e.g., 1.7 vs. 1.1 mg g⁻¹ caffeine and 14.5 vs. 12.5 mg g⁻¹ total flavanols, respectively). Regarding cocoa varieties, Criollo cocoa was richer in bioactive compounds and antioxidant capacity (e.g., 16, 15, and 12 mg g⁻¹ total flavanols for Criollo, Forastero, and Trinitario, respectively). These results indicate that the analytes resulted in potential descriptors of varietal or geographical attributes.

1. INTRODUCTION

The popularity of dark chocolate has increased considerably in recent years, not only because it is deemed a delicacy but also because of its remarkable healthy properties, among which the antioxidant, anti‐inflammatory, antimicrobial, anticancer, and cardioprotective activities stand out (Febrianto et al., 2022; Gil et al., 2021; Ludovici et al., 2017; Samanta et al., 2022; Tan et al., 2021). These beneficial effects are mainly related to its high polyphenol content coming from cocoa as its main ingredient (Cinar et al., 2021).

The cocoa tree (Theobroma cacao) naturally comes from the Amazonian and Orinoco basins, although it is now grown in all tropical areas of the planet. Currently, Africa is the largest cocoa producer in the world, although the one from Central and South America is much more appreciated for its quality. Africa accounts for ca. 70% of the total production of cocoa beans, with the Ivory Coast representing 40% of the global. Asia produces the 17%, and the remaining 13% corresponds to America (Beg et al., 2017).

The type of cocoa is an essential feature, highlighting the principal varieties Criollo, Forastero, and Trinitario. In general, the Criollo variety is the most appreciated for its exceptional organoleptic characteristics, but it is the least productive (Aprotosoaie et al., 2016; Tuenter et al., 2020; Ullrich et al., 2023). In Ecuador, it is known as Nacional cocoa, with subvarieties grown in specific regions that are highly valued by the most select palates (Colonges et al., 2021). In contrast, Forastero cocoa provides the highest agricultural yield and is the least sensitive to diseases, but its flavor and aroma are less appreciated (Tuenter et al., 2020). Halfway between both varieties, the Trinitario cocoa is a hybrid, which combines the good productivity with quite pleasant organoleptic characteristics (Colonges et al., 2021; Rottiers et al., 2019).

Beyond the cultivar type, the cocoa processing, including fermentation, drying, grinding, roasting, and conching, is essential to obtain products of excellence (Barrientos et al., 2019; Febrianto & Zhu, 2019; Toker et al., 2019). Owing to the complexity of this issue, robust and effective analytical methods are essential to assess the cocoa and chocolate quality (Sentellas & Saurina, 2023).

The authentication and detection of possible adulterations can be carried out through a profiling approach based on some natural components as a source of information (Hernandez & Granados, 2021; Quelal‐Vasconez et al., 2020), including phenolic compounds (Agudelo et al., 2022; Cambrai et al., 2017) together with methylxanthines (Nascimento et al., 2020; Samaniego et al., 2020; Febrianto & Zhu, 2019). Amino acids and biogenic amines have also been used to characterize cocoa or chocolate (Tran et al., 2015). The elemental composition reflects the type of soil in which the plant is grown, thus providing information on geographical origin (Vanderschueren et al., 2019).

This study aims to assess the significance of geographical and varietal factors in the content of alkaloids, phenolic compounds, and the antioxidant capacity of chocolate samples. To date, studies published are limited in scope, and conclusions may be contradictory due to the multifactorial dependence of the chemical composition. Here, chocolates from select companies have been analyzed to limit the variability associated with the production process. It should be noted that several published studies deal with cocoa, whereas our focus on chocolate entails additional challenges such as the influence of the manufacturing process on bioactive contents. To obtain robust and reliable data, extraction conditions have been carefully optimized and previously validated HPLC‐UV‐FLD and spectroscopic methods have been applied (Vidal‐Casanella et al., 2022). In this more controlled scenario, we will try to establish differences in methylxanthines and polyphenols among classes. As advantages over other published studies, the sample treatment by sonication and the chromatographic method are simple and fast and provide remarkable descriptions without needing to profile many compounds. The studies cited in the previous paragraphs have not found specific markers of the attributes examined but evidence of up‐ or down‐regulated compounds. We believe that in the absence of claimed specificity, compositional differences (i.e., the cross‐selectivity) among attributes can be relevant in classifying or authenticating samples from multivariate data. As another novel aspect of the study, we combined the information provided by individual compounds with the global content of phenolic compounds, flavonoids, and antioxidant capacity to enrich the description. Furthermore, these last parameters have substantiated the healthy qualities of the chocolates.

2. MATERIALS AND METHODS

2.1. Chemicals and solutions

General reagents and solvents used for the sample treatment and the preparation of the mobile phase were as follows: hydrochloric acid (37% w/w), acetonitrile (ACN) (super gradient grade), and methanol (super gradient grade) from PanReac; formic acid (≥95% w/w) and cyclohexane (≥99.9%, analytical reagent grade) from Sigma‐Aldrich; dimethyl sulfoxide (DMSO > 99% w/w) from Merck; purified water using an Elix 3 coupled to a Milli‐Q system. Chemicals for the spectrophotometric reactions were FeCl₃, sodium acetate, and NaOH from Merck; Folin–Ciocalteu (FC) reagent, NaNO₂, Al(NO₃)₃·9H₂O, and Na₂CO₃ from PanReac; 2,4,6‐Tris(2‐pyridyl)‐s‐triazine (TPTZ) from Alfa Aesar.

Analyte standards were from the following suppliers: Caffeine, catechin, epicatechin, and gallic acid were from Sigma‐Aldrich; procyanidin A2 and procyanidin C1 were from PhytoLab; procyanidin B2 (≥98%) from Chengdu Biopurify Phytochemicals LTD; and theobromine from TRC.

Stock solutions of each standard were prepared in methanol by weighing accurately the solid substance to obtain a standard concentration of ca. 1000 mg L⁻¹. Stock solutions were placed in amber glass vials and stored at 4°C.

2.2. Samples

A set of 26 dark chocolates available in supermarkets and specialized shops was analyzed in this work. The characteristics of these samples regarding cocoa origin, variety, and content are given in Table 1. Cocoa for the production of chocolates was from Africa (e.g., Ghana, Madagascar, Tanzania, Cameroun, or Ivory Coast), South America (Ecuador, Peru, or Venezuela), Central America (Mexico, Nicaragua, the Dominican Republic, Haiti, or Cuba), and Asia (Papua New Guinea and the Philippines). Regarding variety, Criollo (and Nacional), Forastero, and Trinitario samples were available. The cocoa content varied from 64% to 100%, most of them being in the range of 70%–85%.

TABLE 1.

Summary of the analyzed chocolate samples and their characteristics of geographical origin, variety, and cocoa content.

Sample	Origin	Variety	Cocoa (%)
Suchard BIO	Dominican Republic and Peru	Trinitario	70
Museu Xocolata 01	Ghana	Forastero	70
Museu Xocolata 02	Madagascar	Trinitario	70
Museu Xocolata 03	El Salvador	Criollo and Trinitario	68
Museu Xocolata 04	Tanzania	Trinitario	70
Museu Xocolata 05	Dominican Republic	Trinitario	70
Museu Xocolata 06	Papua Nueva Guinea	Trinitario	70
Sampaka America 1	Mexico	Criollo	73
Sampaka America 2	Ecuador	Forastero	72
Sampaka America 3	Venezuela	Criollo	73
Sampaka America 4	Cuba	Criollo	75
Sampaka América 5	Santo Domingo	Trinitario	74
Sampaka América 6	Peru	Criollo	84
Sampaka Africa 1	Ghana	Forastero	72
Sampaka Africa 2	Sao Tome	Forastero	79
Sampaka Africa 3	Cameroun	Trinitario	79
Sampaka Africa 4	Ivory Coast	Forastero	72
Sampaka Africa 5	Madagascar	Trinitario	73
Sampaka Africa 6	Tanzania	Criollo	79
Blanxart 1	Dominican Republic	Not declared	82
Blanxart 2	Peru	Criollo	100
Blanxart 3	Philippines	Trinitario	71
Blanxart 4	Peru	Criollo	77
Blanxart 5	Ecuador	Trinitario	95
Blanxart 6	Congo	Forastero	82
Blanxart 7	Nicaragua	Trinitario	85

2.3. Sample treatment

The chocolate bars were grated and kept in plastic containers at room temperature for later processing. The grated product was defatted to facilitate the recovery of polyphenols and alkaloids from the samples (adapted from Fayeulle et al. (2019) and do Carmo Brito et al. (2017)) as follows: A sample of 1 g was treated with 5 mL of cyclohexane in a 50‐mL polyethylene Falcon tube by ultrasound‐assisted extraction (UAE). The mixture was stirred in a Genius 3 Vortex mixer (IKA) until no sediment was observed, sonicated for 15 min in a Branson 5510 ultrasonic bath (Branson Ultrasonic Corporation), and centrifuged for 5 min at 4500 rpm with a Rotanta 460 RS centrifuge (Hettich). The overfloating liquid containing triglycerides and other lipophilic compounds soluble in cyclohexane was withdrawn. This process was repeated three times to ensure the quantitative removal of lipids. The defatted residue was left to dry overnight.

Phenolics and methylxanthines were recovered from the defatted chocolate samples by UAE with a mixture of methanol and 0.1% formic acid aqueous solution (60:40 v/v). Briefly, 0.2 g of defatted product was mixed with 4 mL of the extraction solvent in a 15‐mL polyethylene Falcon tube, homogenized with a vortex, sonicated for 15 min at room temperature, and centrifuged for 5 min at 4500 rpm. The supernatant solution was collected with a syringe, filtered using a 0.22 µm nylon membrane (Merck Millipore), and kept in the freezer until analysis. Every sample was extracted in triplicate.

A quality control (QC) was prepared by mixing 100 µL of each extract; the QC was used to evaluate the reproducibility of the chromatographic data and soundness of the chemometric results.

2.4. Sample analysis

2.4.1. Chromatographic method

The chromatographic equipment consisted of an Agilent 1100 HPLC system composed of vacuum degasser (G1322A), binary pump (G1312A), autosampler (G1367A), UV–visible diode array detector (G1315B), and fluorescence (FLD) detector (G1321A). The Agilent ChemStation was used for instrument control and data acquisition. The analyte separation in reversed‐phase mode used a Kinetex C18 column (150 × 4.6 mm² of 100 Å pore size and 2.6 µm particle size) from Phenomenex with an elution gradient based on 0.1% formic acid aqueous solution and ACN. The gradient program was as follows: 0–20 min, 18%–48% ACN (linear); 20–22 min, 48%–90% ACN (linear); 22–24 min, 90% ACN; 24–24.2 min, 90%–18% ACN (linear); and 24.2–29 min, 18% ACN. The flow rate was 0.7 mL min⁻¹, and the injection volume was 5 µL. For sample characterization, UV detection was at 280, 325, and 370 nm, and the FLD at 280 and 330 nm as the excitation and emission wavelengths, respectively. However, UV detection at 280 nm was selected for quantitation purposes.

Standard solutions prepared in water/methanol (1:1 v:v) at a concentration range of 0.1–100 mg L⁻¹ were injected at the beginning and at the end of the study to obtain the calibration models. The samples were analyzed randomly to minimize the influence of spurious factors, and the QC was injected every 10 samples until the end of the series.

2.4.2. Folin–Ciocalteu method

The total content of reducing compounds, often associated with the total content of polyphenols (TPC), was determined using the FC spectrophotometric method (Alcalde et al., 2019) using an 8453 UV‐Vis Spectrophotometer (Agilent,) with QS quartz high‐performance cuvettes (10 mm optical path) from Hellma Analytics. The FC procedure consisted of mixing 250 µL of the FC reagent with 1 mL of water. After 8 min, 75 µL of 7.5% (w:v) sodium carbonate aqueous solution and 50 µL of chocolate extracts were added. The mixture was diluted to 5 mL with water, allowed to react for 2 h, and, then, the absorbance was measured at 765 nm. Gallic acid standard solutions in ACN/water (50/50, v/v) with concentrations between 0.5 and 20 mg L⁻¹ were used for calibration. The results were expressed in gallic acid equivalents (mg gallic acid g⁻¹ sample).

2.4.3. Total flavonoid content (TFC) method

The total flavonoid content (TFC) was determined by aluminum complexation mixing 200 µL of fourfold diluted chocolate extract (or standard solution) with 150 µL of 5% NaNO₂ and 1.5 mL of water. After 6 min, 750 µL of 2% Al(NO₃)₃ were added, and the mixture was allowed to react for 6 min at room temperature (Fernández‐Estellé et al., 2023). Then, 1 mL of 1 M NaOH and water were added up to a final volume of 5 mL. The reaction was developed for 15 min, and the absorbance was then measured at 510 nm. Catechin standard solutions in the range 100 to 1000 mg L⁻¹ were prepared in ACN/water (50/50, v/v). Results were expressed in catechin equivalents (mg catechin g⁻¹ sample).

2.4.4. Ferric reducing antioxidant power (FRAP) method

Antioxidant activity was estimated by the ferric reducing antioxidant power (FRAP) as described by Fernández‐Estellé et al. (2023). The FRAP reagent solution was prepared by mixing 20 mmol L⁻¹ FeCl₃, 10 mmol L⁻¹ TPTZ (in 50 mmol L⁻¹ HCl), and 50 mmol L⁻¹ formic acid in a ratio of 1:2:10 (v:v:v). A volume of 600 µL FRAP reagent and 50 µL of 10‐fold diluted chocolate extract in methanol/water (60:40) (or the standard solution) were mixed and diluted to a final volume of 5 mL with Milli‐Q water. The absorbance was measured at 595 nm after 5 min of reaction with the Agilent 8453 UV‐Vis Spectrophotometer. Trolox standards for calibration were prepared in ACN/water (50/50, v/v) in a concentration range of 0.5–5 mg L⁻¹. The results were expressed as Trolox equivalents (TE, mg Trolox g⁻¹ sample).

2.5. Data analysis

The exploratory and statistical study of the influence of the experimental factors on the analyte extraction was carried out with Microsoft Excel using boxplots, statistical tests, and ANOVA; in any case, the significance level was 0.05.

3. RESULTS AND DISCUSSION

Based on previous cases from our working group, in which phenolic composition resulted in an excellent type of descriptor to characterize agri‐food samples, for example, for wines, spices, or nutraceutical products (Izquierdo‐Llopart & Saurina, 2019; Núñez et al., 2020; Pardo‐Mates et al., 2017; Serrano‐Lourido et al., 2012), we reasonably think that they could be used as sample biomarkers of chocolate features, such as origin, variety, or cocoa content. Thus, in this paper, we want to evaluate the role of phenolic compounds, together with theobromine and caffeine, as potential descriptors of quality attributes of chocolates to address their comprehensive characterization. Additionally, information on more general features related to these compounds is also used, such as spectrophotometric indexes of total phenolic content (TPC by FC method), TFC (TFC by the aluminum complexation method), and antioxidant capacity (FRAP method).

To facilitate the extraction of the compounds of interest and minimize the presence of non‐polar components in the extracts, the chocolates were first defatted (see details in Section 2). Although the content of phenolic compounds and methylxanthines, expressed in mg g⁻¹ chocolate, determined in the extracts of untreated and defatted chocolates, were equivalent, the variability is greater when the samples have not been defatted. Under these circumstances, it was concluded that the defatting process was recommendable for dealing with the chocolate authentication as the precision was improved. In addition, the percentages of extractable fat matter were here determined as additional outcomes providing information about quality, with values ranging from 30% to 50% depending on the cocoa content and fat additives. The lower fat percentages corresponded, in general, to chocolate with cocoa content greater than 90%, whereas the higher were for those samples with less than 70% cocoa.

In this paper, previously established and validated HPLC methods were adapted to separate the principal phenolic compounds occurring in chocolate, considering also theobromine and caffeine as the main cocoa bioactive components. As a result, methylxanthines (theobromine and caffeine) and the characteristic phenolic compounds (catechin, procyanidin B2, epicatechin, and procyanidin C1) were fully resolved from each other (Figure 1). Limits of detection and quantitation ranged from 0.002 mg g⁻¹ defatted chocolate (for caffeine) to 0.015 mg g⁻¹ defatted chocolate (for theobromine) and from 0.007 mg g⁻¹ defatted chocolate to 0.052 mg g⁻¹ defatted chocolate, respectively. The precision of the method was satisfactory with RSD lower than 4% for all the target compounds at high concentrations and lower than 10% at concentrations close to the quantitation limit. These data demonstrated the suitability of the used method to analyze the samples under study. For detailed information about the method performance, see Table S1.

FIGURE 1.

Chromatogram recorded at 280 nm of a defatted chocolate extract using the method described in Section 2.4. Peak assignment: 1, gallic acid; 2, theobromine; 3, catechin; 4, caffeine; 5, procyanidin B2; 6, epicatechin; 7, procyanidin C1.

3.1. Extraction procedure

The optimization of the extraction of phenolic compounds from defatted chocolate by an ultrasound‐assisted procedure addressed the study of the experimental variables considered to be most relevant, highlighting the solvent composition, extraction time, and temperature. Extracts were analyzed by high‐performance liquid chromatography with UV (HPLC–UV) to estimate the concentration of the most abundant analytes, including theobromine, caffeine, and phenolics (e.g., epicatechin, procyanidins, and various phenolic acids). Each condition was assayed in triplicate (except for some preliminary studies to select the extraction medium) using a representative chocolate sample (Suchard BIO 70% cocoa, Trinitario from the Dominican Republic and Peru) as the model.

The preliminary evaluation of the solvent explored the extraction performance of various pure solvents with different polarities (chloroform, acetone, ACN, methanol, DMSO, and water) and hydro‐organic mixtures (ACN/water 1/1 v:v, methanol/water 1/1 v:v, and DMSO/water 1/1 v:v). In this case, 0.2 g of sample and 4 mL of solvent were mixed and sonicated for 30 min at room temperature. The results, detailed in Figure 2, showed that pure solvents provided more modest recoveries, whereas hydro‐organic mixtures performed much better. Among them, the results of the ACN/water and MeOH/water mixtures are completely equivalent, with the second combination finally being selected due to its lower toxicity and cost.

FIGURE 2.

Study of the influence of the solvent type on the extraction of methylxanthines and phenolic compounds from the defatted chocolate. ACN, acetonitrile; DMSO, dimethyl sulfoxide.

Once methanol was selected as the cosolvent for the hydro‐organic mixture, an optimization of the optimal percentage in the extraction medium was performed using ca. 0.1 g of defatted sample and 2 mL of solvent. Four MeOH/water mixtures were investigated with MeOH content of 20%, 40%, 60%, and 80 v/v (each condition assayed in triplicate). As can be seen in Figure 3, the results indicated that extraction yield increased slightly from 20% to 40% MeOH, to then remain practically constant from 40% to 80%, except in the case of gallic acid. As gallic acid is the most polar compound of the series, its extraction using a lower percentage of methanol is favored. In any case, although the differences between 20% and 40% are generally statistically significant, the improvement in the recovery is small, ca. 10% for, for example, theobromine. However, increasing the percentage of methanol up to 80% did not enhance the extraction efficiency. Summarizing a percentage of MeOH between 40% and 60% seems to be adequate to carry out this extraction. As a compromise for all types of compounds, a MeOH percentage of 60% was selected.

FIGURE 3.

Study of the influence of the methanol percentage on the extraction of methylxanthines and phenolic compounds from the defatted chocolate. Error bars indicate the standard deviation. Equal letters for each compound mean no significant differences (p > 0.05).

The addition of a given (low) concentration of an organic acid, such as acetic or formic acid, to the extraction solvent has been sometimes proposed to improve the extraction efficiency of phenolic acids and flavonoids. The acid medium maintains the protonated (neutral) phenolic species, which can be extracted more easily with organic solvents. Furthermore, the chemical and microbiological stability of acid extracts is superior since basic or neutral media can favor hydrolysis reactions or the development of microorganisms. Here, the effect of the formic acid concentration on the analyte recovery was evaluated using MeOH/water 40:60 v:v as the solvent. Experiments were carried out at four levels of formic acid (0%, 1%, 2%, and 4%). The results, depicted in Figure 4, show that the influence of the formic acid content on the extraction of the analytes is scarce. The changes in the extract concentration are not significant for compounds such as gallic acid or catechin. For others, such as theobromine, epicatechin, or caffeine, slightly higher concentrations were attained at formic acid concentrations of 1%–2%. In this sense, an acidic extracting solvent was recommended because of the quantitative impact on the yield and the increased sample stability. An extraction solvent consisting of MeOH/water 40:60 v:v with 1% formic acid was eventually selected.

FIGURE 4.

Study of the influence of the formic acid percentage on the extraction of methylxanthines and phenolic compounds from the defatted chocolate. Error bars indicate the standard deviation. Equal letters for each compound mean no significant differences (p > 0.05).

The influence of the sonication time (10, 30, 60, and 120 min) on the extraction yield was also assessed from the individual phenolic concentrations as the data. The results showed a slight increase in the extraction rate over time (see Figure 5), which is statistically significant in most cases from 10 to 30 min but not beyond. Although the maximum values were reached after 60 min, 30 min of sonication was selected as a compromise between response and process time. It should be mentioned that the concentration values obtained practically no longer improve compared to the aforementioned studies, thus suggesting that the working conditions allow a quantitative extraction of the analytes. Once the UAE time was set, the potential effect of the bath temperature on the extraction was evaluated, comparing the composition of the extracts obtained at 20 and 50°C. No significant differences were found in the extraction yield either for global contents or for individual compounds. Therefore, for simplicity, room temperature sonication was selected for subsequent studies.

FIGURE 5.

Study of the influence of the sonication time on the extraction of methylxanthines and phenolic compounds from the defatted chocolate. Error bars indicate the standard deviation. Equal letters for each compound mean no significant differences (p > 0.05).

3.2. Characterization of chocolate samples

After optimizing the extraction conditions, the extracts resulting from a chocolate set of Criollo (with the National variety from Ecuador), Trinitario, and Forastero samples from American and African origin were analyzed by HPLC–UV and spectroscopic methods as detailed in Section 2. Various target compounds, including theobromine, caffeine, epicatechin, catechin, and procyanidins, were quantified to evaluate whether their distribution depends on the continental origin or the botanical variety of cocoa. The total polyphenol and flavonoid contents were estimated using gallic acid and catechin equivalents, respectively. The antioxidant power was also estimated using TEs.

Apart from the geographical origin and botanical variety, the percentage of cocoa influences the bioactive compound content. Obviously, the higher the percentage of cocoa mass in the chocolates, the higher the concentration of bioactives in the final product. Hence, a significant correlation between the cocoa percentage and theobromine (p = 1.7E − 04), catechin (p = 4.2E − 06), and caffeine (p = 2.4E − 05) is observed (Figure S1). However, the correlation of other components such as procyanidin B2 and epicatechin was not noticed for the set of samples studied (p > 0.05). This is probably because other variable(s) (botanic variety or geographical origin, among others) may have a major influence on their prevalence. Similarly, the spectroscopic indexes showed poor correlation with cocoa content, except for TFC (p = 2.7E − 03).

Focusing on the influence of geographical origin and botanical variety, all the data are given in Tables S2 and S3, and the most illustrative results are in Figure 6 using boxplot representations. As a global result, regardless of varieties and regions, TFC values are in the range of 10–26 mg g CE, very similar to those reported by Jaćimović et al. (2022) for dark chocolates. In general, theobromine contents are between 18 and 35 mg g⁻¹ and caffeine between 0.8 and 3.0 mg g⁻¹, values of the same order of magnitude as those obtained for Serbian dark chocolates (Todorovic et al., 2015), although some of the samples analyzed here show slightly higher concentrations. Catechin and epicatechin concentrations are also consistent with these same studies.

FIGURE 6.

Boxplots of the concentrations of representative compounds based on geographical origins or cocoa varieties: (a1) theobromine versus origin; (a2) caffeine versus origin; (a3) epicatechin versus origin; (a4) total flavanol content versus origin; (b1) theobromine versus variety; (b2) caffeine versus variety; (b3) epicatechin versus variety; (b4) total flavanol content versus variety. Sample assignment: AF, Africa; AM, America; C, Criollo; F, Forastero; T, Trinitario. Equal letters mean no significant differences (p > 0.05).

The comparison of samples of American or African origin suggests that the first ones are richer in bioactive components. Without being selective descriptors, it can be seen that, on average, the values of theobromine, epicatechin, or TFCs are ca. 20% higher in the American samples (e.g., an average of 14.5 and 12.5 mg g⁻¹ TFC in American and African chocolates, respectively). The differences are even more notable for caffeine, greater than 50%, with mean values of 1.7 and 1.1 mg g⁻¹ caffeine for American and African continents, respectively. Although these results are not generalizable, as they are based on a small number of samples, it does seem that there may be differences in the compositional profiles depending on the origin. For the other compounds and antioxidant indices, the trends, although similar, are not so marked because the dispersion of the values is greater.

Regarding botanical varieties, the differences are not as marked as in the issue of origin, perhaps because this factor is quantitatively more important and masks the conditions in this regard. One of the clearest patterns is observed with the global flavonoid content, higher for Criollo, intermediate for Forastero, and lower for Trinitario. This trend is also seen for epicatechin and is less defined for the other flavanols. On the contrary, theobromine content seems to be unrelated to the varieties, whereas caffeine is slightly more abundant in Criollo and Trinitario cocoa. When the variability associated with the origin is eliminated, that is, when the American and African samples are compared separately, despite the small number of samples, it is concluded that, indeed, the Criollo variety is richer in phenolic compounds and slightly higher in alkaloids.

Despite the scarce literature on the matter, some works have been published attempting to find patterns dealing with the content of bioactive or antioxidant compounds with geographical or varietal attributes of cocoa or chocolate. For instance, Agudelo et al. (2022) determined the monomeric and oligomeric flavanol species to investigate their relationship with the clones, origin, or harvest year. Principal component analysis (PCA) revealed the strong impact of the year factor on the chemical composition (Agudelo et al., 2022). On the contrary, compositional differences associated with crop varieties or geographical areas were not significant, so they could not establish biomarkers of these attributes. Other authors also proposed polyphenols for studying botanical varieties or origins (e.g., South America, Caribbean, and Africa) (Cambrai et al., 2017), although without concluding on potential discriminant markers. In another study, Nascimento et al. (2020) used theobromine, caffeine, epicatechin, and catechin concentrations for grouping artisanal and fine chocolates using PCA and hierarchical cluster analysis, pointing out that higher quality chocolates seem to contain higher levels of bioactives. In another case, methylxanthines and flavanols were quantified in cocoa samples of different genotypes (Febrianto & Zhu, 2019). Statistical studies, boxplots, and PCA showed the differences among crops, revealing those that may be healthier or organoleptically more interesting. These same compounds were explored in another study to characterize cocoa beans from Amazonia and the Pacific coast of Ecuador, showing that Amazonian samples were richer in alkaloids, whereas no conclusive observations could be drawn on flavanol levels (Samaniego et al., 2020).

4. CONCLUSIONS

This study explores the quality of dark chocolate based on the compositional profiles of the major alkaloids and flavanols as well as overall antioxidant or phenolic levels. Some analytes (e.g., theobromine, caffeine, epicatechin, and procyanidins), not being specific, are overexpressed in some cases, thus resulting in sound descriptors of some sample attributes. Besides, the antioxidant capacity provides complementary information dealing with the claimed healthy and superfood characteristics of dark chocolate. Although the study design covers a small number of samples, the analyte concentrations depend on geographical and varietal factors, in addition to the cocoa content. Chocolates made with American cocoa are richer in bioactive compounds; likewise, Criollo cocoa also appears to contain higher levels of phenolics. Despite the multifactorial nature of the analyte composition, this exploratory study can be the basis for addressing further cocoa or chocolate authentication studies. Furthermore, combining data from individual compounds and generic phenolic indices or antioxidant capacity for multivariate analysis may provide more comprehensive information on the sample behavior, both at the level of descriptors and to infer potential beneficial qualities for health.

AUTHOR CONTRIBUTIONS

Tamara Parada: Investigation; methodology. Pablo Pardo: Investigation; methodology. Javier Saurina: Conceptualization; funding acquisition; supervision; formal analysis; writing—review and editing; writing—original draft. Sonia Sentellas: Conceptualization; writing—original draft; writing—review and editing; formal analysis; supervision.

CONFLICT OF INTEREST STATEMENT

The authors declare no conflicts of interest.

Supporting information

Table S1 Figure of merits (at 280 nm) of the chromatographic method used for sample analysis.

Table S2 Total polyphenolic content (TPC), total flavonoid content (TFC), and antioxidant capacity (FRAP) of the set of samples under study.

Table S3 Bioactive contents in the samples under study.

Figure S1 Correlation of cocoa percentage with (a) theobromine, (b) catechin, and (c) caffeine content.

Parada, T. , Pardo, P. , Saurina, J. , & Sentellas, S. (2024). Characterization of dark chocolates based on polyphenolic profiles and antioxidant activity. Journal of Food Science, 89, 8857–8867. 10.1111/1750-3841.17451