Optimization of Antioxidant Synergy in a Polyherbal Combination by Experimental Design

Culinary herbs and spices are known to be good sources of natural antioxidants. Although the antioxidant effects of individual culinary herbs and spices are widely reported, little is known about their effects when used in combination. The current study was therefore undertaken to compare the antioxidant effects of crude extracts and essential oils of some common culinary herbs and spices in various combinations. The antioxidant interactions of 1:1 combinations of the most active individual extracts and essential oils were investigated as well as the optimization of various ratios using the design of experiments (DoE) approach. The 2,2-diphenyl-1-picrylhydrazyl (DPPH), 2,2-azinobis (3-ethylbenzothiazoline-6-sulphonic acid) (ABTS), and ferric reducing antioxidant power (FRAP) assays were used to determine the antioxidant activity, and MODDE 9.1® software (Umetrics AB, Umea, Sweden) was used to determine the DoE. The results revealed synergism for the following combinations: Mentha piperita with Thymus vulgaris methanol extract (ΣFIC = 0.32 and ΣFIC = 0.15 using the DPPH and FRAP assays, respectively); Rosmarinus officinalis with Syzygium aromaticum methanol extract (ΣFIC = 0.47 using the FRAP assay); T. vulgaris with Zingiber officinalis methanol extracts (ΣFIC = 0.19 using the ABTS assay); and R. officinalis with Z. officinalis dichloromethane extract (ΣFIC = 0.22 using the ABTS assay). The DoE produced a statistically significant (R2 = 0.905 and Q2 = 0.710) model that was able to predict extract combinations with high antioxidant activities, as validated experimentally. The antioxidant activities of the crude extracts from a selection of culinary herbs and spices were improved when in combination, hence creating an innovative opportunity for the future development of supplements for optimum health.


Introduction
Culinary herbs and spices are an important part of human nutrition, and they have been used for centuries not only as flavoring agents but as food preservatives and for health benefits. Each spice or herb contains many bioactive compounds such as flavonoids, phenolics, sulfur-containing compounds, tannins, alkaloids, vitamins, and essential oils. Some of these compounds are responsible for their reported antioxidant activities [1,2]. Antioxidants are free radical scavengers and thus inhibit lipid peroxidation and other freeradical-mediated processes, protecting the human body from several diseases attributed to the reactions of free radicals [3][4][5][6]. Furthermore, antioxidants are added to food to prevent deterioration through oxidative processes [7,8]. Natural antioxidants, such as those present in culinary herbs and spices, have been investigated as alternatives to synthetic counterparts due to the concerns of potential carcinogenicity and other adverse effects [9,10]. In recent years, various studies have reported on the antioxidant activities of culinary herbs and spices [1,4,8,[11][12][13][14][15][16].
Spices and herbs such as Syzygium aromaticum (clove), Mentha piperita (peppermint), Cinnamomum zeylanicum (cinnamon), Origanum vulgare (oregano), Thymus vulgaris (thyme), Salvia officinalis (sage), and Rosmarinus officinalis (rosemary), to name a few, were reported to be strong antioxidants due to their high levels of phenolic compounds [5,[17][18][19]. Spices and herbs are often added as blended mixtures in culinary preparations. Furthermore, many traditional healing modalities such as the Indian system of medicine (Ayurveda) and Chinese and African traditional medicines rely on combinations of highly complex herbal mixtures to achieve an enhanced effect. Many scientific reports, however, astutely focus on the bioactivity of individual extracts of culinary herbs and spices [11,12,20,21], but their combined effects are not well-documented. The antioxidant interactions of combinations of essential oils from spices and herbs [22,23] and natural antioxidant compounds with synthetic antioxidants [24] have been reported. It could therefore be of great interest to investigate the possible antioxidant interactions occurring when spices are combined. This approach may increase their antioxidant effect at sufficiently low concentrations by taking advantage of their possible synergistic or additive effects. To study the interactions between various spices, herbal extracts, and essential oils requires many experiments to determine which combination ratios provide the optimum antioxidant effect. A design of experiments (DOE) approach was thus utilized to identify and optimize the synergy potential of the extracts, using a limited number of experimental assays, thus saving time and resources.
Several in vitro assays are used to evaluate the antioxidant activities of spices and herb extracts and essential oils, namely, the 2,2-diphenyl-1-picrylhydrazyl (DPPH), oxygen radical absorbance capacity (ORAC), ferric reducing antioxidant power (FRAP), 2,2-azinobis (3-ethylbenzothiazoline-6-sulphonic acid) (ABTS), and microsomal lipid peroxidation (MLP) assays. The present study was therefore designed to evaluate the antioxidant activity of a selection of spices, herb crude extracts, and essential oils by comparatively evaluating the results of three antioxidant assays (DPPH, ABTS, and FRAP) as well as optimizing the synergy potential of the crude extracts and essential oils by investigating the various combinations using the DoE approach.

Antioxidant Activity of Individual Crude Extracts and Essential Oils
The data on the comparative analysis of the antioxidant activities of crude extracts and essential oils are presented in Table 1. The total antioxidant activity measured by the DPPH assay, presented as half-maximal effective concentration (EC 50 ), ranged from 5.48 to 497.10 µg/mL for methanol extracts, 7.66 to 1340.00 µg/mL for the water extracts, 17.85 to 807.80 µg/mL for dichloromethane extracts, and 2.55 to 861.50 µg/mL for the essential oils. The antioxidant properties of a spice or herb differed depending on the solvent used for the extraction. Based on the results in Table 1, C. zeylanicum methanol and water extracts demonstrated noteworthy DPPH radical scavenging activities with EC 50 values of 5.48 and 7.66 µg/mL, respectively. Furthermore, the recorded activity was better than that of the positive control, ascorbic acid (EC 50 = 10.25 µg/mL). Gupta et al. [25] and Mansour et al. [26] in their studies also reported high DPPH radical scavenging activity for C. zeylanicum methanol extracts, though the values were expressed as % DPPH inhibition. Kim et al. [11] reported C. zeylanicum water extract to possess a higher DPPH radical scavenging activity compared to other spice extracts. However, the EC 50 value was higher (0.254 mg/mL) compared to that obtained in this study. The least active extract was C. sativum water extract (1340.00 µg/mL). With regards to the essential oils, noteworthy activity was displayed by S. aromaticum oil (EC 50 = 2.55 µg/mL), followed by M. piperita and C. zeylanicum oils with EC 50 values of 6.88 and 7.17 µg/mL, respectively. Other oils displayed moderate to low activities, with T. vulgaris essential oil displaying the least antioxidant activity with EC 50 = 861.50 µg/mL. The antioxidant trend for the ABTS assay was different from the DPPH assay, with the total antioxidant activity ranging from EC 50 values of 6.06 to 69.19 µg/mL for methanol extracts, 5.79 to 145.90 µg/mL for water extracts, 3.09 to 258.40 µg/mL for dichloromethane extracts, and 5.81 to 1397 µg/mL for the essential oils. The S. aromaticum and Z. officinalis dichloromethane extracts displayed notable activities with EC 50 values of 3.09 and 4.15 µg/mL, respectively.
The FRAP assay demonstrated a higher variation compared to the DPPH and ABTS assays, with EC 50 values ranging from 33.45 to 11,498.00 µg/mL for the methanol extracts and 42.96 to 1143.00 µg/mL for the water extracts. The EC 50 values for the dichloromethane extracts ranged from 36.26 to 81,083.00 µg/mL, while essential oils ranged from 5.96 to 794.80 µg/mL. The order of active methanol extracts according to the FRAP assay, in comparison to ascorbic acid was R. officinalis > C. zeylanicum > M. officinalis > T. vulgaris > M. piperita > S. officinalis > S. aromaticum > C. citratus > ascorbic acid. The C. zeylanicum, R. officinalis, and S. aromaticum methanol and dichloromethane extracts showed promising FRAP reducing activities (EC 50 values = 36.44, 33.45, and 58.47 µg/mL for the methanol extracts and 36.26, 58.92, and 40.95 µg/mL for the dichloromethane extracts, respectively). Furthermore, the C. zeylanicum and R. officinalis water extracts also demonstrated notable effects (EC 50 = 42.96 and 45.86 µg/mL, respectively) compared to ascorbic acid (90.59), while M. officinalis and M. piperita methanol extracts were active at EC 50 = 39.49 and 53.40 µg/mL, respectively. Other extracts had moderate, low, or no activity with reference to ascorbic acid. In their study, Chan et al. [27] reported the notable FRAP reducing ability of C. zeylanicum water extracts. However, the EC 50 value was higher (0.24 mg/mL) compared to the value obtained in the current study.
The results on the antioxidant activity of spices or herbal extracts and essential oils demonstrated stronger to very weak correlations among the three assays used in the study ( Table 2). A weaker correlation existed between the DPPH and FRAP assays for the water (r = 0.42) and methanol (r = 0.37) extracts. This is contrary to the findings by Ulewicz-Magulska and Wesolowski [19], who reported a stronger relationship between the two assays (r = 0.94). Meanwhile, very weak correlations existed for the dichloromethane extracts (r = 0.13) and essential oils (r = 0.22). A higher correlation (r = 0.71) was shown between the DPPH and ABTS assays for the water extracts. Furthermore, a weak correlation existed between the ABTS and FRAP assays for all the extracts and essential oils. No correlation existed between the DPPH and ABTS assays for methanol extracts.

Interactive Studies
For all the assays, the antioxidant activity of the extracts was compared to that of the positive control. All the extracts and essential oils displaying EC 50 values closer to the positive control and demonstrating notable antioxidant activity in the three assays were selected for antioxidant interaction studies. Extracts (12) and essential oils (2) were combined (1:1 v/v) to evaluate the interactive antioxidant activity using the DPPH, ABTS, and FRAP assays., The data results of the comparative analysis of the antioxidant interactions from the different combinations of the extracts and essential oils are presented in Table 3. The results show that 3.80% of the combinations were synergistic in the DPPH assay, and 7.70% were synergistic in the FRAP and ABTS assays, while 50% of the extracts were additive using the DPPH and FRAP assays, and 19.20% were additive with the ABTS assay. Other combinations were indifferent with ABTS (69.20%), and 38.40% were indifferent with the DPPH and FRAP assays. Antagonism was shown for 11.50% of the combinations with DPPH assay and 3.82% for the FRAP assay. None of the combinations were antagonistic -with the ABTS assay. Combining T. vulgaris and Z. officinalis methanol extracts yielded synergy (ΣFIC = 0.19) in the ABTS assay, while the combination was indifferent using the DPPH and FRAP assays. Synergy was also observed when combining M. piperita and T. vulgaris methanol extracts using the FRAP (ΣFIC = 0.15) and DPPH assays (ΣFIC = 0.32). Meanwhile these combinations were indifferent in the ABTS assay. Other combinations that were synergistic include R. officinalis with S. aromaticum methanol extracts at ΣFIC = 0.47 following the FRAP method. However, this combination was additive in the DPPH and ABTS assays. This clearly demonstrates that variability exists between the studied methods, and including different assays, as in this study, provides a better overall assessment of efficacy. Similar observations were made when comparing the assays, whereby M. officinalis combined with S. aromaticum methanol extracts and M. piperita with R. officinalis methanol extracts displayed additive effects. Meanwhile, the combinations of R. officinalis with either C. zeylanicum or Z. officinalis methanol extracts were indifferent in the three assays. The combinations of M. piperita methanol extract with either C. zeylanicum or S. aromaticum extracts were antagonistic using the DPPH method. A study by Mansour et al. [26] reported synergistic antioxidant effects from a combination of C. zeylanicum and S. aromaticum methanol extracts with the DPPH assay. Similarly, Purkait et al. [23] reported synergy from combining the two oils using the DPPH assay. However, the combination was indifferent in this study. Saeed et al. [28] also reported synergy from combining C. zeylanicum with S. aromaticum water extracts using the DPPH assay. Meanwhile, an additive effect was observed in the current study. The difference between their study and the results of this study is that they used the combination index (CI) to evaluate the antioxidant interactions, while in the current study, the ΣFIC was calculated. Furthermore, in their study, synergy was interpreted as any value < 1, while in this study, a more stringent criterion (ΣFIC ≤ 0.5 interpreted as synergy) was used. Overall, the results from the interactive study demonstrated that selected crude extracts from culinary herbs and spices can be combined to achieve synergistic or additive antioxidant effects and thus enhance efficacy.

Design of Experiments (DOE) Data Analysis and Model Verification
Three extracts (M. piperita, T. vulgaris, and Z. officinalis) were selected for DOE studies based on a synergistic outcome from the combination/interaction studies. This was con-ducted in order to determine the optimum ratio at which a formulation of the three extracts can be combined to obtain the highest antioxidant effect. Twelve experimental runs with the combinations indicated in Table 4 were obtained from the design, and the EC 50 values obtained for each combination using the DPPH, ABTS, and FRAP assays were imported into MODDE ® 9.0 (Umetrics AB, Umea, Sweden. The data in Table 4 were modeled, and the replication plot, histogram, summary of fit, coefficient plot, residual N-plot, observed versus predicted plot, and the response contour plots were obtained and used to assess the suitability of the PLS model for predictions.

Summary of Fit Plot
The linear generated model was fitted against the data, and the response is shown in the summary of the fit plot in Figure 1, which provides the information on the strength and robustness of the model. The R2 value (0.91) signified a low variation in the response and a strong fit between the data and the model. Meanwhile, the Q2 value of 0.71 (ideally > 0.5) demonstrated the high predictive power of the model. Furthermore, the model demonstrated a strong validity of 0.44, which is greater than the 0.25 that is required for good models. The model reproducibility of 0.97 is far greater than the requisite value of 0.50, indicating good model design and low error ( Figure 1).

Coefficient Plot
The center-scaled coefficients of each term in the model were used to estimate the significance of the factors to the desired response (EC 50 ). It was determined that M. piperita, Z. officinalis, and T. vulgaris methanol extracts and the interactions between M. piperita and Z. officinalis were important factors to the model outcome, whereas the interactions between T. vulgaris with either M. piperita or Z. officinalis were found to be non-significant factors ( Figure 2). Typically, large regression coefficients represent factors with a large contribution to the model response, such as M. piperita extract, while regressions with a positive number denote a positive contribution towards the response, such as Z. officinalis extract, and a negative number denotes a negative response.

Response Contour Plot
The response contour plot allows for the identification of ratios of the combinations that demonstrate the best and worst overall antioxidant effects ( Figure 3). The generated model predicted various ratios at which spice/herb extracts can be combined to produce an optimal antioxidant effect. Extract combinations in the red region were predicted to produce high EC 50 values and, thus, low antioxidant activity. These combinations generally comprise higher proportions of Z. officinalis and T. vulgaris, and low M. piperita. On the opposite end of the color spectrum, combinations in the blue region were predicted to produce lower EC 50 values and hence higher antioxidant activity. In the blue region, predicted EC 50 values are given with ratios of combinations with high M. piperita, low T. vulgaris, and a wider range of Z. officinalis content. The optimizer function was then used to determine the best ratio at which the extracts should be combined to obtain the optimum synergistic antioxidant effects. Five combinations in the blue region of the response contour plot were selected, as displayed in Table 4, to validate the model predictions. The DPPH, ABTS, and FRAP methods were used for the model design. However, the FRAP method was chosen, as it produced the best model predictions compared to the DPPH and ABTS methods. In this study, a statistically significant model was produced, which provided the best combination ratios of the extracts, using limited time and resources. The optimizer showed the optimal antioxidant activity was predicted with a mixture of M. piperita (55.00%), T. vulgaris (44.00%), and Z. officinalis (1.00 %) at EC 50 = 39.59 µg/mL using the FRAP method. The results from the validation experiments confirmed the reliability and fitness of the model because the correlation coefficient (r = 0.7594), calculated using Microsoft Excel ® 2019, [29] between the experimental and predicted EC 50 values for all the combinations (Table 5).

Materials and Methods
Seventeen commonly used culinary herbs and spices with documented antioxidant activities (Table 6) were selected for this study. Spices were purchased from Warren Chemical Specialties (Pty) Ltd. (Johannesburg, South Africa). The identification was based on the supplier product labelling, as the products were obtained commercially in powder form for all the spices. The materials were kept in a cool and dry place before extraction. Extracts of different polarities (water, methanol, and dichloromethane) were prepared for each of the herbs and spice samples by macerating the preparations in the solvents at a 1:10 solvent ratio, followed by shaking in the dark for 24 h at room temperature using a mechanical shaker. The mixtures were filtered through Whatman No.1 filter paper. The filtrates were then evaporated to dryness under vacuum at 40 • C in a vacuum evaporator (H50-500, Magna Analytical, Labtech, South Africa). The 17 essential oils were purchased from Prana Monde (Belgium). All commercial essential oils were accompanied by a certificate of analysis. The chromatographic profiling was performed in-house and the marker compounds were identified for each species. The stock solutions of the extracts (1 mg/mL) were prepared by dissolving a known amount of the sample in either methanol, water, or dichloromethane. The solutions were stored at 4 • C until use. The working concentrations (500, 250, 125, 62.5, 31.25, 15.62, 7.81, and 3.91 µg/mL) were prepared from the stock solutions by diluting with the correct volume of methanol. The positive control, ascorbic acid, was prepared in methanol. The individual extracts and essential oils ( Table 6) were screened for antioxidant activity using the DPPH, ABTS, and FRAP assays. The three assays were employed to provide broader information on the antioxidant activity of the tested extracts and essential oils. Measurements were obtained in triplicate for each sample in each assay. The EC 50 values were calculated for the control and samples, representing the antioxidant capacity in the sample necessary for 50% of the maximal antioxidant effect. In all three assays, the EC 50 values were determined using GraphPad Prism software, version 5.0 (GraphPad software Inc., San Diego, CA, USA).

The 2,2-Diphenyl-1-picrylhydrazyl (DPPH) Radical Scavenging Assay
The DPPH radical scavenging effects of the extracts and essential oils were estimated according to the method of Brand-Williams et al. [50], with some modifications. Briefly, a DPPH (Sigma-Aldrich, Germany) solution (0.1 Mm w/v) was prepared in methanol, and 100 µL of the DPPH solution was mixed with 100 µL of the sample in each of the wells of a 96-well microtiter plate. For the DPPH and methanol controls, a volume of 200 µL of DPPH (96 µM in HPLC-grade methanol (Merck, South Africa)) and 200 µL of HPLC-grade methanol were added to the corresponding wells, respectively. Ascorbic acid (22.5 µg/mL) was used as a positive control. The plates were then shaken at 960 rpm for 2 min and incubated in the dark at 27 • C for 30 min. After incubation, the absorbance was read at a single wavelength of 517 nm using a microplate reader (SpectraMax M2 Multimode Microplate Reader, Molecular Devices Inc., USA) linked to a computer with SoftMax ® Pro version 6.5.1 for data acquisition and analysis. The inhibition of the DPPH radical by the active samples was determined by calculating the DPPH free radical scavenging activity percentage according to Equation (1).
where A = absorbance at 517 nm; A control = average absorbance of DPPH − average absorbance of methanol; A methanol = average absorbance obtained in the wells containing methanol; A test = absorbance obtained in the wells containing DPPH and the test sample.

The 2,2 -Azino-Bis (3-Ethylbenzothiazoline-6-Sulphonic Acid) (ABTS) Cation Radical Scavenging Assay
The spices and herbs were tested for their ABTS radical scavenging activity according to the method of Re et al. [51], with some modifications. Two stock solutions of 7 mM ABTS (Sigma-Aldrich, Burlington, MA, USA) in double-distilled water and 2.45 mM potassium persulphate (Merck, Lethabong, South Africa) were mixed. The mixture was incubated in the dark for 12-16 h at room temperature and used as a working solution. The solution was adjusted with cold ethanol to obtain an absorbance of 0.700 (±0.02) at 732 nm using a microplate reader (Spectra Max M2, Molecular Devices Inc., Silicon Valley, CA, USA). A 100 µL volume of ABTS+ solution was added to the microtiter plate wells containing 100 µL of crude extracts or essential oils. After 30 min of incubation, the percentage of decolorization of ABTS+ at 734 nm was calculated for each concentration relative to the blank, according to Equation (2). Ascorbic acid (22.5 µg/mL) was used as a positive control. % Decolorization = 100 × [(A control ) − (A test + A methanol )]/A control (2) where A = absorbance at 734 nm; A control = average absorbance of ABST + − average absorbance of methanol; A methanol = average absorbance obtained in the wells containing methanol; A test = average absorbance obtained in the wells containing ABST + and the test sample.

Ferric Iron-Reducing Antioxidant (FRAP) Power Assay
The ferric iron-reducing antioxidant (FRAP) assay is based on the ability of the antioxidant to reduce Fe 3+ to Fe 2+ in the presence of 2,4,6-Tris(2-pyridyl)-1,3,5-triazine (TPTZ), forming an intense blue Fe 2+ -TPTZ [52]. The FRAP reagent was freshly prepared before each experiment by mixing 300 mM acetate buffer (pH 3.6), 20 mM ferric chloride in distilled water, and 10 mM TPTZ (Sigma-Aldrich, Switzerland, in 40 mM HCl) in a ratio of 10:1:1. The FRAP solution (100 µL) was mixed with 100 µL of the test samples and standard solution in each of the wells of a 96-well microtiter plate. Following the same procedure, a blank test containing methanol instead of the extract was used as a negative control, while ascorbic acid at 22.5 µg/mL served as the positive control. The reaction mixtures were then incubated in the dark for 30 min, and the reduction of the Fe 3+ -TPTZ complex to a colored Fe 2+ -TPTZ complex by the extracts and essential oils was monitored by measuring the absorbance after 4 min of incubation at 593 nm using a microplate reader (SpectraMax M2, Molecular Devices Inc., USA). The FRAP value of each sample was calculated using Equation (3) and was expressed as µg/mL ascorbic acid. where A = absorbance at 593 nm.

Statistical Analysis
A one-way analysis of variance (ANOVA) was performed on each plant extract, and a post hoc Tukey's multiple comparison test was conducted to calculate the differences between extracts of the same plant. The data from the three antioxidant assays (DPPH, ABTS, and FRAP) were correlated by calculating the Pearson's correlation coefficient (r) using Microsoft Excel ® 2019.

Fractional Inhibitory Concentration (FIC)
The sum of the fractional inhibitory concentration index (ΣFIC) was used to measure interactions from different 1:1 combinations of herbs, spices extracts, and essential oils with promising antioxidant effects when tested using the DPPH, FRAP, and ABTS assays. Samples of A and B (50 µL) for the combinations were plated out in each of the wells, which were marked accordingly, followed by the addition of an antioxidant reagent (100 µL) to final concentrations of 3.1 to 500 µg/mL. The ΣFICs for each of the combinations were calculated using Equation (4).
where (a) is the EC 50 of one spice extract or essential oil in the combination and (b) is the EC 50 of the other extract or essential oil. The ΣFICs for each combination were interpreted as synergy where the ΣFICs were less than or equal to 0.5 and as additive effects when the ΣFICs were greater than 0.5 but less than or equal to 1.0. For indifference, the ΣFICs were greater than 1.0 but less than or equal to 4.0, and for antagonism, the ΣFICs were greater than 4.0 [53]. For all antioxidant assays, positive and negative controls were included, with a known antioxidant, ascorbic acid (22.5 µg/mL), used as a positive control. Methanol, which was used as a diluent to dissolve the test samples, was used as a negative control.

Experimental Design Using the DOE Model to Determine Effective Antioxidant Combinations
The DoE model was prepared and evaluated using MODDE ® version 9.0 (Umetrics AB, Umea, Sweden). The models were fitted with partial least squares (PLS) and were adjusted by removing non-significant terms. To determine which input parameters resulted in the desired outcomes, screening experiments were carried out using a fractional factorial design. Both the independent and dependent variables were fitted to a linear model. This required 12 experimental runs with three center points. All the experiments were completely randomized by the software to reduce bias and experimental errors. After the experimental runs were completed, the results were analyzed using MODDE ® version 9.0, and the dataset gave a close-to-normal distribution. Hence, it did not require a logarithmic transformation. A mixture design worksheet for the chosen extracts was produced, and the modeling of responses was generated to confirm the best model fit. A prediction contour plot was generated, which showed the average prediction (point estimate prediction) for every possible combination of the tested extracts. The predictions with desired EC 50 values were selected for the model validation. These model predictions were then verified by carrying out additional laboratory experiments and comparing the results to the predicted values.

Conclusions
The results of this study demonstrated the antioxidant variability between crude extracts and essential oils from culinary herbs and spices using different in vitro methods (DPPH, ABTS, and FRAP assays). The methanol extracts demonstrated better radical scavenging effects compared to the other tested extracts, based on the DPPH and ABTS assays. The C. zeylanicum methanol and water extracts were found to be more effective DPPH radical scavengers compared to the standard antioxidant, ascorbic acid. The most superior ABTS radical scavengers were the S. aromaticum and Z. officinalis dichloromethane extracts, which had greater activity than ascorbic acid. The most effective extracts in reducing ferric iron were the R. officinalis, C. zeylanicum, M. officinalis, T. vulgaris, M. piperita, and S. aromaticum methanol extracts. Apium graveolens C. zeylanicum, Mentha piperita, and S. aromaticum essential oils were found to be effective antioxidants. The results from the different assays could not be correlated, except for the relationship between the DPPH and FRAP assays for the water extract, which showed a high positive correlation.
The 1:1 combination displayed synergistic, additive, indifferent, and antagonistic effects. Strong synergism was shown by combining T. vulgaris with Z. officinalis (methanol extract) using the ABTS assay, R. officinalis with S. aromaticum (methanol extracts) using the FRAP assay, and from combining T. vulgaris and M. piperita methanol extracts using the DPPH and FRAP assays. Using the DoE, a model that could easily predict the ratios at which the spice extracts can be combined to achieve the highest antioxidant effect was produced. The use of extracts in various combinations resulted in an optimum antioxidant effect, even at lower concentrations compared to the single extracts due to synergistic interactions.