The biological material used in this study consisted of treated Citrus sinensis zest and Moringa oleifera leaves.

Weighing was carried out on a weighing balance (Shimadzu, Japan) and a precision balance (OSI, M-220 D). A binatone blender (BLG 555, China) was also used to grind the samples, which were then sieved using commercial sieves (mesh ≤ 100 µm). Humidity was determined using an oven (memmert UL 30, Germany). Wavelengths were read using a spectrophotometer (Jenway 7315, UK).

The samples were collected in the city of Abidjan, in the south of Côte d’Ivoire. Citrus sinensis peels were collected in the communes of Adjamé (the biggest citrus market, particularly for oranges), Cocody (a wealthy commune) and Yopougon (the biggest commune, middle class).

Next, 10 kg of fresh Moringa oleifera leaves were collected from gardeners in the commune of Bingerville. After collection, the samples were transported to the laboratory for further study. Samples were collected in triplicate.

The collected samples were sorted, soaked in distilled water for 5 min and then drained. They were then withered at 50˚C for 6 h, oxidized for 4 h and dried at 50˚C to constant mass. Finally, each species was ground.

The powders obtained from each species were mixed using a composite central plane. In this way, formulations with different proportions were determined. This experimental design took into account two variables (quantity of Citrus sinensis zest and quantity of Moringa oleifera leaves). The combination of the 2 variables led to 11 formulations with 4 factorial trials (FT), 4 star trials (ST) and 3 center trials (CT). The various composite powders, the reference tea and their codes are shown in Table 1 [11] .

EF: factorial essay; EE: star essay; EC: essay at the centre; ET1. sample control.

Herbal teas were prepared according to [12] . In addition, 100 ml of boiling water was poured over 2 g of each composite powder, then filtered after 5 min.

Total polyphenol content was determined using the Folin-Ciocalteu method [13] . Absorbance was measured at a wavelength of 725 nm. Gallic acid was used as a standard. Results were expressed as milligram equivalent of gallic acid per 100 g of dry matter (mg EAG/100g MS) and milligram equivalent of gallic acid per liter of herbal tea (mg EAG/L).

Flavonoid content was determined according to [14] . To 0.5 mL infusion were added 0.5 mL distilled water, 0.5 mL 10% (w/v) aluminum chloride, 0.5 mL 1M sodium acetate and 2 mL distilled water. After reaction in the dark for 30 min, optical density was read at 415 nm. Results were expressed in milligram equivalents of quercetin per 100 g of dry matter and in milligram equivalents of quercetin per liter of herbal tea.

1 mL extract was diluted in 1.5 mL pH 1.0 buffer and 1.5 mL pH 4.5 buffer respectively. The absorbance (A) of these two solutions was read at λmax = 520 nm and λmax = 700 nm respectively [15] . Total anthocyanin (TA) content was expressed in milliequivalents of cyanidin 3-glycoside per 100 g of dry matter and in milliequivalents of cyanidin 3-glycoside per liter of herbal tea.

Two methods were used to determine the antioxydant activity of herbal teas: DPPH (1,1-diphenyl-2-picrylhydrazyl), et FRAP (Ferric reducing antioxydant power).

1) DPPH

The free radical scavenging activity of herbal teas was determined using the DPPH (1,1,diphenyl-2-picrylhydrazyl) free radical reduction method described by [16] . A 3 mM DPPH solution was prepared. 0.5 mL of herbal tea was added to 1.5 mL of DPPH solution. After incubation for 30 min in the dark, absorbance was measured at 517 nm using the DPPH solution as control. Assays were performed in triplicate. Results were expressed in equivalent micromoles of ascorbic acid per gram of dry matter and in equivalent micromoles of ascorbic acid per 100 mL of herbal tea.

2) FRAP

The reducing power of an extract is associated with its antioxydant power. The iron activity of our extracts was determined according to the method described by [16] . based on the chemical reduction of Fe³⁺ present in the K₃Fe(CN)₆ complex to Fe²⁺. A 1 ml volume of infusion was mixed with 2.5 ml of 0.2 M phosphate buffer (pH = 6.6) and 2.5 ml of 1% K₃Fe(CN)₆ solution. Incubate at 50˚C for 20 minutes, then cool to room temperature. Next, 2.5 ml of 10% trichloroacetic acid was added to stop the reaction. 2.5 ml of the extract was added to 2.5 ml of distilled water, and 500 µl of 0.1% (FeCl₃, 6H₂O) solution was added to the mixture.

Absorbances were read against a blank at 700 nm using a spectrophotometer. Trials were carried out in triplicate. Results were expressed in micromoles of ascorbic acid equivalent per gram of dry matter (µmol EAA/gMS) and in micromoles of ascorbic acid equivalent per 100 millilitres of herbal tea (µmol EAA/100mL).

3) Composite antioxydant index

The antioxydant composite index (ACI) was determined according to [4] and calculated as follows:

$\text{ACI}=\frac{\text{sample}\text{score}}{\text{best}\text{score}}\ast 100$

Table 2 shows the polyphenol, total flavonoid and anthocyanin contents of the samples. Polyphenol contents range from 645.41 to 2961.38 mg EAG/100g MS, i.e. from 129.08 to 592.28 mg EAG/L infusion. These levels differed statistically (p ≤ 0.05) between the different samples. The control sample (ET) made up of 100% Camellia sinensis had the highest polyphenol content (2961.38 mg EAG/100 g), while (EE05) made up of 62.5; 37.5% Citrus sinensis and Moringa oleifera respectively had the lowest total polyphenol content (645.41 mg EAG/100g). Infusions of EC09, EC10, EC11 have identical total polyphenol contents (p > 0.05). These contents are estimated at 734.82 mg EAG/100g or 146.99 mg EAG/L infusion. Moreover, these have the same composition: 70% Citrus sinensis and 30% Moringa oleifera.

Results are means ± S.D. (n = 3); values of the same column. followed by the same letter (a-f) are not statistically different as measured by Duncan’s test (p > 0.05). EF: factorial essay; EE: star essay; EC: essay at the centre; ET1. sample control; EQ: quercetin equivalent; EAG: gallic acid equivalent ET. sample control; Composite powder with respective percentages of Citrus sinensis and Moringa oleifera: EF01. 77.89/22.11; EF02. 88.14/15.85; EF03. 55.85/44.14; EF04. 65.59/34.4; EE05. 62.5/37.5; EE06. 75/25; EE07. 87.5/12.5; EE08. 58.33/41.66; EC09. 70/30; EC10. 70/30; EC11. 70/30.

The total flavonoid content of the composite herbal teas ranged from 291.83 to 383.93 mg EAG/100g, representing 58.37 to 76.80 mg EQ/L infusion. The EF02 trial (88.14% Citrus sinensis and 15.86% Moringa oleifera) contained the lowest flavonoid content, while EC10 (70% Citrus sinensis and 30% Moringa oleifera) contained more flavonoids. However, the sample made up of 100% Camellia sinensis (ET) has the highest content (p ≤ 0.05).

The anthocyanin content of the various samples ranged from 0 to 0.04 mg/100mL infusion, or 0 to 0.24 mg/100g dry matter. Among the infusions, EF01 (77.89% Citrus sinensis/22.11% Moringa oleifera) has the highest anthocyanin content, while EF04 (65.59% Citrus ssinensis/4.4% Moringa oleifera) and EC09 (70% Citrus sinensis/30% Moringa oleifera) have the lowest. EF02, EF03, EF04, EE05, EE06, EE08, EC09, EC10 and EC11 have identical anthocyanin contents (p > 0.05).

Antioxydant capacity

The antioxydant activity of the infusions was determined using two tests (DPPH and FRAP). The values obtained are shown in Table 3 . The control sample (ET) made from 100% Camellia sinensis showed the highest free radical scavenging activity (170.10 ± 0.17 µmol EAA/g MS), i.e. 340.21 ± 0.33 µmol EAA/100mL infusion, as well as the highest reducing capacity (501.59 ± 2.2 µmol EAA/g DW). The composite herbal tea from EF01 (77.89% Citrus sinensis and 22.11% Moringa oleifera) has an antiradical activity close to that of the control (ET). This activity is estimated at 149.32 µmol EAA/g MS or 298.65 µmol EAA/100mL herbal tea. By contrast, EE07 (87.5% Citrus sinensis; 12.5% Moringa oleifera) and EE06 (75% Citrus sinensis; 25% Moringa oleifera) have the lowest free radical scavenging activity, estimated at 204.44 and 122.5 µmol EAA/100mL respectively, at the same 20 mg/mL concentration.

Results are means ± S.D. (n = 3); values of the same column. followed by the same letter (a-h) are not statistically different as measured by Duncan’s test. EF: factorial essay; EE: star essay; EC: essay at the centre; ET. sample control; Composite powder with respective percentages of Citrus sinensis and Moringa oleifera: EF01. 77.89/22.11; EF02. 88.14/15.85; EF03. 55.85/44.14; EF04. 65.59/34.4; EE05. 62.5/37.5; EE06. 75/25; EE07. 87.5/12.5; EE08. 58.33/41.66; EC09. 70/30; EC10. 70/30; EC11. 70/30.

FRAP test values range from 166.92 µmol EAA/100mL herbal tea, equivalent to 83.46 µmol EAA/g MS (EE08: 58.33% Citrus sinensis 41.66% Moringa oleifera) to 220.63 µmol EAA/100mL, equivalent to 110.31 µmol EAA/g MS (EF03: 55.85% Citrus sinensis; 44.14% Moringa oleifera). Herbal teas EC09, EC10, EC11 made from 70% Citrus sinensis/30% Moringa oleifera show the same reducing activity (p > 0.05). Also, EF01 (77.89% Citrus sinensis; 22.11% Moringa oleifera), EF02 (88.14% Citrus sinensis; 22.11% Moringa oleifera), EE05 (62.5% Citrus sinensis; 37.5% Moringa oleifera) are not statistically different (p > 0.05).

Antioxydant Composite Index

Table 4 shows the antioxydant composite index (ACI) of the different infusions. Indeed, the reference sample ET (Camellia sinensis) has the highest index. The ICA of the composite herbal teas is ranked in ascending order as follows: EF01 > EF03 > EC10 > EC11 > EC09 > EF04 > EE05 > EF02 > EE08 > EE07 > EE06. Herbal tea EF01 (77.89% Citrus sinensis/22.11% Moringa oleifera) has the highest composite antioxydant index, followed by EF03 (55.85% Citrus sinensis/44.14% Moringa oleifera) and EC10 (70% Citrus sinensis/30% Moringa oleifera). These indices are 53.2, 50.9 and 50.6 respectively. Furthermore, EE08 (58.33% Citrus sinensis/41.66% Moringa oleifera); EE07 (87.5% Citrus sinensis/12.5% Moringa oleifera); EE06 (75% Citrus sinensis/25% Moringa oleifera) present the lowest ICA. They are estimated at 44.8, 38.4 and 26.9 respectively.

EF: factorial essay; EE: star essay; EC: essay at the centre; ET1. sample control ET. sample control; Composite powder with respective percentages of Citrus sinensis and Moringa oleifera: EF01. 77.89/22.11; EF02. 88.14/15.85; EF03. 55.85/44.14; EF04. 65.59/34.4; EE05. 62.5/37.5; EE06. 75/25; EE07. 87.5/12.5; EE08. 58.33/41.66; EC09. 70/30; EC10. 70/30; EC11. 70/30.

Phenolic compounds are secondary metabolites synthesized by plants for protection against excessive ultraviolet radiation, various physical aggressions or pathogens.

Polyphenolic composition

The total polyphenol contents of the samples are different from those obtained by authors [17] . These authors obtained contents ranging from 24.77 ± 2.02 to 252.65 ± 4.74 mg EAG/g MS through a study carried out on 30 teas from Camellia sinensis consumed in China. These differences are due to the species and production technology, as oxidation reduces the total polyphenol content. In addition, polyphenoloxidases transform flavanols into theaflavins and thearubigins, thus reducing polyphenol content [18] .

In addition, the herbal teas studied have similar contents to herbal teas from Chromolaena odorata [19] . These contain on average 10.61 to 16.02 mg EAG/100mL infusion. Furthermore, authors estimated the total polyphenol content of “Rooibos tea” infusions in South Africa at 25.78 ± 1.12 g EAG/100g [20] . For the same species, obtained 35.07 ± 3.44 mg EAG/100g and 25.05 ± 2.84 mg EAG/100g respectively in methanolic and aqueous extracts (infusion) [21] .

Our herbal teas are therefore more concentrated in polyphenols than herbal teas available in these countries. The heterogeneity of the different contents between our herbal teas is partly due to their composition. In addition, the temperature of the water used for infusion, as well as the oxidation time during the production process, has an influence on polyphenol content. Polyphenols are recognized for their health benefits. In fact, they reduce the risk of diseases linked to oxidative stress [22] [23] . In addition, polyphenols are responsible for certain organoleptic properties such as astringency and bitterness in herbal teas [24] .

However, certain phenolic compounds such as tannins may reduce the digestibility of food by binding and precipitating dietary carbohydrates, proteins and digestive enzymes. Thus, they are considered antinutrients [25] . In addition, malnourished people may drink our herbal teas moderately, but over-nourished/obese people may drink more. Polyphenols have several health benefits, such as cardiovascular protection and anti-cancer effects, which are associated with their strong antioxydant power. In addition, authors, obtained average total flavonoid contents of 76.82 mg EQ/100g DM in Moringa oleifera leaves grown in Côte d’Ivoire [26] [27] .

A study showed that Citrus sinensis orange peel contained 190 ± 0.09 mg EQ/100 g MS [28] . These contents are low compared with those obtained in our composite herbal teas made from Citrus sinensis zest and Moringa oleifera leaves. Thus, the combination of the two species would be beneficial in terms of contribution of polyphenolic compounds. The ET infusion (100% Camellia sinensis) has a higher total polyphenol and flavonoid content than the various composite teas formulated (EF01, EF02, EF03, EF04, EE05, EE06, EE07, EE08, EC09, EC10, EC11). However, EC09, EC10, EC11, all made up of 70% Citrus sinensis and 30% Moringa oleifera, have contents closer to the ET control (100% Camellia sinensis). In fact, production technology and conditions could have an influence on polyphenol content, as Camellia sinensis (ET) tea sold on the market has undergone a well-mastered transformation process.

In a comparative study of several Brazilian teas, [29] obtained flavonoid contents estimated at 34.09 ± 3.28 mg CTE/L for “Pimpinella anisum tea”. These values are lower than those obtained in our composite teas. The latter determined the flavonoid content of “Camellia sinensis tea” equivalent to 179.88 ± 32.41 mg CTE/L, higher than our values in composite herbal teas and the ET control consisting of 100% Camellia sinensis estimated at 91.26 ± 5.86 mg EQ/L. This difference may be due to the growing area, the tea production technology and the method used for the assay, as our results were determined using a quercetin calibration line.

It is estimated that the average human intake of flavonoids is between 25 mg/day and 1 g/day [30] . Consumption of our herbal teas could therefore cover requirements. Several studies have shown that a diet rich in flavonoids can have beneficial effects on health. Thanks to their ability to inhibit LDL oxidation, flavonoids have significant cardioprotective effects [31] . Another study showed that a high intake of flavonoids can reduce mortality from coronary heart disease (CHD) and lower the risk of CHD by 38% in post-menopausal women .

Antioxydant capacity

Several methods can be used to determine the antioxydant activity of extracts: DPPH test, FRAP, ABTS, ORAC. Moreover, the antioxydant activity of extracts cannot be reasonably validated by a single method due to the complex nature of phytochemical compounds and their interactions, hence the importance of using multiple test systems [33] [34] . Thus, in the present study two tests commonly used to assess the antioxydant activity of extracts were employed, namely the DPPH (dpph radical scavenging) and FRAP (Fe³⁺ to Fe²⁺ reduction) assays. These were combined to comprehensively assess the antioxydant activities of composite herbal tea infusions.

The values obtained from the DPPH and FRAP tests show that the infusions have a high level of both antiradical and iron-reducing activity. The antiradical activity of the composite herbal teas, ranging from 122.5 to 298.65 µmol EAA/100mL, and the reducing power (166.92 to 220.63 µmol EAA/100mL) are higher than the averages obtained by Konan et al. (2014) in juices consumed in Côte d’Ivoire. The values for the various composite herbal teas are lower than those for tea (Camellia sinensis).

In terms of anti-free radical activity, the classification is as follows: ET > EF01 > EF04 > EC10 > EC11 > EC09 > EF03 > EE05 > EE08 > EF02 > EE07 > EE06. FRAP test values indicate the ability of herbal teas to reduce ferric ions (Fe III) to ferrous ions (Fe II). So, in addition to their antiradical activity, the composite herbal teas formulated have a strong reducing power. Among the composite teas formulated, EF03 (55.85% Citrus sinensis/44.14% Moringa oleifera) has the highest reducing capacity. It is estimated at 220.63 mg AAE/100mL. The reducing capacity of the herbal teas is ranked in ascending order as follows: ET > EF03 > EC11 > EC09 > EC10 > EF04 > EF02 > EF01 > EE05 > EE06 > EE07 > EE08.

Thus, these herbal teas with their high anti-free radical activity could be beneficial for smokers [35] , as antioxydant deficiency is more pronounced in these individuals due to the attacks generated by free radicals in smoke. Elderly (increased oxidative stress with age) and overweight subjects, who constitute a group at risk of antioxydant deficiency [36] , could also benefit from these herbal teas to reduce their deficiency. In addition, for subjects with a high level of potentially prooxydant iron (ferric ions), these herbal teas could play an antioxydant role by reducing these ferric ions to ferrous ions.

The composite antioxydant index (CIA) of herbal teas is ranked in ascending order as follows: ET > EF01 > EF03 > EC10 > EC11 > EC09 > EF04 > EE05 > EF02 > EE08 > EE07 > EE06. In general, there is a positive correlation between total polyphenol content and composite antioxydant index. Among the composite teas formulated, EF01 (77.89% Citrus sinensis/22.11% Moringa oleifera) has the highest index and the highest total polyphenol content. However, EE06 (75% Citrus sinensis/25% Moringa oleifera) has the lowest ICA, while EE05 has the lowest total polyphenol content. These results could be explained by a difference in the composition of the herbal teas. In addition, some polyphenols have no free radical scavenging or ferric ion reducing activity [37] .

Principal Component Analysis

The quality parameters studied are correlated with 5 factors. However, the F1 factor, with an eigenvalue greater than 1, is used for PCA. It accounts for 62.97% of total variability. Nevertheless, the second factor, with an eigenvalue of 0.97 and total variability of 19.55%, was combined with the first factor for PCA representation. The factor (F1) with an eigenvalue of 3.14 is predominantly formed by all the characteristics studied. These are positively correlated. Parameters and individuals are projected onto the plane formed by factors 1 and 2, which account for 82.52% of total variability ( Figure 1 & Figure 2 ). It divides the individuals into 3 groups. Group 1 is essentially made up of individuals superimposed on traits positively correlated with the F1 factor. They are characterized by high anthocyanin, polyphenol, flavonoid, DPPH and FRAP values.

This is the ET sample. The second group contains samples with lower values than those in the first group. Samples in the third group are characterized by low values for anthocyanins, polyphenols, flavonoids, DPPH and FRAP.

Hierarchical ascending classification (HAC) using the Euclidean distance method confirms the variability observed in PCA. Truncation of the dendrogram at an aggregation Euclidean distance of 8 reveals three classes ( Figure 2 ). Individuals in the first class are distinguished by higher anthocyanin, polyphenol, flavonoid, DPPH and FRAP values than the other samples analyzed. The second class represents the intermediate class, distinguished by anthocyanin, polyphenol, flavonoid, DPPH and FRAP values lower than those of the first class, but higher than those of the third class. The third class includes samples with low anthocyanin, polyphenol, flavonoid, DPPH and FRAP values.