The plant material used in this study consisted of dried cocoa beans (marketable cocoa) collected from six cocoa-growing zones of Côte d’Ivoire: Abengourou (east), Dabou (south), Niablé (central-west), Man (west), San-Pédro (south-west), and Yamoussoukro (centre).

In each zone, three samples were collected from three producers located in different localities. In total, 18 samples were collected. Each sample was ground separately using a multifunctional grinder (QE-100, Zhejiang Yili Tool Co., Ltd., China) and then sieved through a 125 µm mesh. The resulting powder was stored under refrigeration until subsequent analysis. Cocoa powder was prepared following the method of [15].

Secondary metabolites were extracted according to [16]. Dried extracts were obtained by macerating 5 g of cocoa powder in 100 mL of ethanol/water (70:30, v/v) for 24 h in the dark under continuous magnetic stirring (1 g powder per 10 mL solvent). Extraction was repeated three times on the same plant residue (marc). The combined extracts were filtered and the solvent removed by oven evaporation at 40˚C. Extraction yield was calculated as follows:

where M₁ is the mass of the plant extract (g) and M₀ is the mass of the plant material used (g).

Total polyphenol content was determined by the Folin-Ciocalteu method [17]. Dried extracts, dissolved in ethanol/water (70:30, v/v) to a final concentration of 5 mg/mL, were mixed (30 µL) with 2.5 mL of Folin-Ciocalteu reagent (diluted 1:10) and incubated in the dark at room temperature (30˚C ± 2˚C) for 2 min. Sodium carbonate solution (2 mL, 7.5% Na₂CO₃) was then added, and the mixture incubated in a water bath at 50˚C for 15 min before rapid cooling. Absorbance was read at 765 nm against a distilled-water blank. Gallic acid was used as the reference standard, selected for its stability, well-characterized pharmacological activity, and representativeness of phenolic compounds generally [18] [19]. Content was calculated from a calibration curve constructed with gallic acid solutions (0 - 1.4 mg/mL Y = 0.2 X + 0.0012 ; R² = 1) and expressed, from triplicate measurements, as milligrams of gallic acid equivalents per gram of cocoa powder (mg GAE/g).

Total flavonoid content was determined colorimetrically [16] [20]. Dried extracts (5 mg/mL in ethanol/water, 70:30, v/v; 2.5 mL) were mixed with 0.75 mL of 5% sodium nitrite (NaNO₂) for 5 min, followed by 0.75 mL of 10% aluminium chloride (AlCl₃) and a 6 min dark incubation. Sodium hydroxide solution (5 mL, 1 N) was then added and the final volume adjusted to 25 mL with distilled water.Absorbance was read at 510 nm against a distilled-water blank. Catechin was used as the reference standard, and content was calculated from a calibration curve (0 - 0.9 g/L; Y = 0.543 X − 0.0379 , R² = 0.992) and expressed, from Y = 2 X + 0.0012 triplicate measurements, as milligrams of catechin equivalents per gram of cocoa powder (mg CE/g).

Total tannin content was measured by a colorimetric assay adapted from [21]. Extract (100 µL, 5 mg/mL) was combined with 7.5 mL of distilled water and 0.5 mL of Folin-Ciocalteu reagent, followed by 1 mL of 35% sodium carbonate (Na₂CO₃) solution and dilution to 10 mL with distilled water. The mixture was left to react at room temperature (25˚C - 30˚C) for 30 min. Absorbance was read at 734 nm against a distilled-water blank. Tannic acid was used as the standard (calibration range 0 - 1000 µg/mL; Y = 2 X − 0.0012 , R² = 1); results are expressed as milligrams of tannic acid equivalents per gram of cocoa powder (mg TAE/g).

Antioxidant activity was assessed using the ABTS·+ (2,2’-azino-bis(3-ethylbenzothiazoline-6-sulfonic acid)) radical-scavenging assay, following [22] with minor modifications. The ABTS·+ stock solution was prepared by combining 8 mmol/L ABTS solution (87.7 mg in 20 mL distilled water) with 3 mmol/L potassium persulfate solution (16.2 mg K₂S₂O₈ in 20 mL distilled water) and incubating in the dark at room temperature (30˚C ± 2˚C) for 16 h. A working solution was freshly prepared before each assay by diluting the stock solution in methanol to an absorbance of 0.70 ± 0.02 at 734 nm. Cocoa extract (100 µL, 2.5 mg/mL) was added to 3.9 mL of the ABTS·+ working solution, vortexed, and incubated in the dark at room temperature for 6 min. Absorbance was read at 734 nm against a methanol blank. Antioxidant activity was expressed as the percentage inhibition of the ABTS·+ radical relative to a Trolox calibration curve, with results additionally convertible to micromoles of Trolox equivalents (µmol TE) per gram of sample.

Differences in total polyphenol, flavonoid, and tannin content, and in antioxidant activity, among production zones were tested by one-way ANOVA (factor “production zone”, six levels, n = 3 replicates per level, N = 18 observations), with statistical significance set at p < 0.05.

Mean total polyphenol concentrations in cocoa powder extracts from the six production zones, determined from the gallic acid calibration curve, are presented in Table 1 and Figure 1, Figure 2. Dabou cocoa contained the highest total polyphenol content (33.06 ± 0.55 mg GAE/g), followed by intermediate levels in Séguéla (27.78 ± 0.75 mg GAE/g), Man (25.69 ± 0.85 mg GAE/g), Yamoussoukro (23.20 ± 0.45 mg GAE/g), and Abengourou (21.61 ± 0.11 mg GAE/g); San-Pédro recorded the lowest content (10.87 ± 0.35 mg GAE/g). Zones ranked by increasing polyphenol concentration were therefore: San-Pédro < Abengourou < Yamoussoukro < Man < Séguéla < Dabou. One-way ANOVA (p < 0.05) confirmed that these inter-zone differences were statistically significant, identifying Dabou beans as the richest in total polyphenols and San-Pédro beans as the poorest.

Table 1. Total polyphenol content of Ivorian cocoa powder by production zone.

Total flavonoid content, determined from the catechin calibration curve, is presented in Table 2 and Figure 3, Figure 4. Content was highest in cocoa from Dabou (5.96 ± 0.12 mg CE/g) and Séguéla (4.58 ± 0.10 mg CE/g), with intermediate levels in Man (3.97 ± 0.04 mg CE/g), Abengourou (3.73 ± 0.08 mg CE/g), Yamoussoukro (3.07 ± 0.03 mg CE/g), and San-Pédro (3.03 ± 0.02 mg CE/g). Zones ranked by decreasing flavonoid content were: Dabou > Séguéla > Man > Abengourou > Yamoussoukro ≈ San-Pédro, indicating that Dabou and Séguéla beans were the richest in

Table 2. Total flavonoid content of Ivorian cocoa powder by production zone.

total flavonoids while San-Pédro beans contained the lowest amounts. One-way ANOVA did not detect statistically significant differences among zones (p > 0.05), although the trend approached significance (p = 0.085).

Mean total tannin concentrations, determined from the tannic acid calibration curve, are presented in Table 3 and Figure 5, Figure 6. Content was highest in Dabou cocoa (3.45 ± 0.15 mg TAE/g), with intermediate levels recorded in Séguéla (2.66 ± 0.10 mg TAE/g), Yamoussoukro (2.53 ± 0.05 mg TAE/g), Abengourou

Table 3. Total tannin content of Ivorian cocoa powder by production zone.

(2.11 ± 0.09 mg TAE/g), and Man (2.03 ± 0.07 mg TAE/g); San-Pédro contained the lowest tannin content (1.76 ± 0.15 mg TAE/g). Zones ranked by decreasing tannin content were: Dabou > Séguéla > Yamoussoukro > Abengourou > Man > San-Pédro, confirming that Dabou beans possessed the highest tannin concentration and San-Pédro beans the lowest. Despite this apparent variability, one-way ANOVA revealed no statistically significant difference in tannin content among the six production zones (p > 0.05).

Antioxidant activity, measured as ABTS·+ inhibition, is presented in Table 4 and Figure 7, Figure 8. Activity was highest in cocoa from Dabou (67.98 ± 0.46%) and Séguéla (57.92 ± 0.59%), followed by comparable but slightly lower activity in Man (52.52 ± 0.51%), Yamoussoukro (50.60 ± 0.60%), and Abengourou (49.62 ± 0.42%); San-Pédro showed the lowest activity (29.82 ± 0.48%). Zones ranked by increasing antioxidant activity were: San-Pédro < Abengourou < Yamoussoukro < Man < Séguéla < Dabou, mirroring the ranking observed for total polyphenol content. Inhibition percentages ranged overall from 29.82% to 67.98%, indicating substantial variability in antioxidant potential according to bean origin. One-way ANOVA (p < 0.05) confirmed that these inter-zone differences were statistically significant.

Table 4. Antioxidant activity (ABTS·+ inhibition) of Ivorian cocoa powder by production zone.

The total polyphenol contents recorded in this study are lower than several previously reported values for cocoa beans. Borja Fajardo etal. [23] found concentrations of 44.51 - 95.41 mg GAE/g in Colombian cocoa, Samaniego et al. [24] reported 42.75 - 71.66 mg GAE/g in Ecuadorian beans, and Sanchez-Capa et al. [25] recorded 57.23 - 79.18 mg GAE/g for National Arriba cocoa and 48.46 - 55.54 mg GAE/g for Trinitario cocoa. Several factors may account for the comparatively lower values observed here. Drying is a critical post-harvest operation that strongly influences the final quality of cocoa beans and their derived products [26]; five drying techniques are commonly used in the cocoa sector - sun, solar, oven, microwave, and freeze-drying [27]—each of which alters the physical structure of the beans and promotes the degradation of polyphenols [28] [29]. Oxidative degradation converts these compounds first into quinones and subsequently into melanins under the action of polyphenol oxidase [28] [30], an effect compounded by fermentation [25]. Regional differences in vegetation and agroecological conditions may further influence polyphenol biosynthesis. Conversely, several other studies have reported markedly lower polyphenol contents than those obtained here: Febrianto and Zhu [31] reported 0.501 - 0.917 mg GAE/g, Mihai etal. [32] 2.29 - 2.97 mg GAE/g, and Cortez et al. [33] values as low as 0.19 mg GAE/g, while de Barros Kobi et al. [34] reported 0.257 mg GAE/g and Deus et al. [29] 0.65 - 1.074 mg GAE/g. By contrast, Nieto-Figueroa et al. [35] (24.8 - 27.3 mg GAE/g) and Alean et al. [28] (33.30 mg GAE/g) reported values close to those of the present study, as did Akafou et al. [36] who, although working on cocoa pod husk rather than beans, measured higher polyphenol contents in yellow pods (35.50 ± 0.50 mg GAE/g dry extract) than in green pods (25.02 ± 0.01 mg GAE/g dry extract) - a range broadly comparable to our own results.

Flavonoid content varied by production zone and was markedly lower than in several previous reports. Loureiro et al. [37] recorded an epicatechin content of 43.27 mg/g in Forastero cocoa from the Amazon, close to the catechin content of 43.6 mg/g described by Febrianto and Zhu [38]. Schlüter et al. [39] found unfermented Trinitario cocoa to contain 38.9 mg/g (defatted dry matter) of epicatechin and a lower but still substantial 1.67 mg/g of catechin. Akafou et al. [36] similarly reported high total flavonoid contents in cocoa pod husk, both yellow (65.33 ± 0.33 mg epicatechin equivalents/g dry extract) and green (65.06 ± 0.04 mg epicatechin equivalents/g dry extract). De Taeye et al. [40], working with two Trinitario clones, detected the anthocyanins cyanidin-3-arabinoside and cyanidin-3-galactoside at 2.850 - 3.112 mg/g - higher than the levels found in fermented beans - while N’goran Haddad et al. [41] reported values ranging from 4.37 to 6.47 mg/g dry matter, closer to those obtained in the present study. This variability likely reflects differences in geographic origin, post-harvest processing, and analytical methodology; farming system, notably shading conditions, has also been shown to influence epicatechin and catechin content [34].

Mean total tannin content in the present study ranged from 2.03 ± 0.07 to 3.45 ± 0.15 mg TAE/g dry extract. These values are substantially lower than those reported by Akafou et al. [36] for unfermented green and yellow cocoa pod husks (23.98 ± 0.02 and 12.23 ± 0.35 mg TAE/g dry extract, respectively) and by Ibero-Baraibar et al. [42], who reported a total tannin content of 240 mg TAE/g dry matter, but higher than the 0.908 mg/g dry matter reported by Gomez et al. [43]. This discrepancy may reflect differences in sample type and processing, particularly the absence of fermentation in some of the comparator studies.

The antioxidant activity values obtained here indicate that cocoa beans from Dabou and Séguéla possess strong antioxidant properties, consistent with the 47.2% - 85.8% inhibition range reported by Ali Faisal et al. [44] and the 68.09% inhibition reported by Racine et al. [45]. The strong correlation observed between antioxidant activity and both polyphenol and flavonoid content in the present dataset supports the established contribution of these compounds to the free-radical-scavenging capacity of cocoa [7], and underlines total polyphenol content as the principal driver of regional differences in antioxidant potential among the cocoa powders studied.