A Procedure for Assessment of the Reducing Capacity of Plants-Derived Beverages Based on the Formation of the FeII / 2 , 2 ’-Bipyridine Complex

An alternative spectrophotometric ferric reducing activity power (FRAP) method for quantification of total reducing capacity (TRC) was developed. The method is based on the reduction of Fe to Fe by antioxidant compounds containing 2,2’-bipyridine (bipy) in aqueous solution. Absorbance values recorded at 521 nm, characteristic of the Fe(bipy)3 complex formed, were used to determine the TRC of some plants-derived beverages. For the teas samples, the TRC values obtained with the proposed method and cupric reducing antioxidant capacity (CUPRAC) reagent had an excellent agreement (adjusted correlation coefficient (r) = 0.951). Concerning herbs samples, the TRC values obtained with the proposed FRAP method correlated very well with values obtained using the 2,2’-azino-bis-3-ethylbenzothiazoline-6-sulfonic acid (ABTS) method (adjusted r = 0.975). It can be inferred from these results that other beverages derived from plants (e.g., beers, wines, and fruits juices) could also be analyzed with this new FRAP assay. In addition, the reducing capacity of 21 phenolic derivatives was determined with the proposed method in order to elucidate their structure-reactivity relationship. As expected, the phenolic derivative structure changes greatly the TRC values obtained with this proposed FRAP assay.


Introduction
Originally, the acronym for ferric reducing activity power (FRAP) was employed to designate the ferric reducing ability of plasma, an assay designed to measure the antioxidant power of this biological sample.3][4][5] Based on this more comprehensive definition, other methods using the reduction reaction of Fe III in solution containing different complexing agents for Fe II have also been developed.
Phenanthroline and batho-phenanthroline, both chelating agents that form stable and colored complexes with Fe II at pH 4.6 (acetate buffer), were utilized in a thorough study dealing with the quantification of reducing capacity of many mixtures of standard polyphenols. 6owever, these iron complexes, as far as we know, have not yet been applied to determine the reducing capacity in any sample of plant origin.
On the other hand, the reduction of Fe III in acid solution (1.0 mol L -1 HCl) containing 3-(2-pyridyl)-5,6-bis(4-phenylsulfonic acid)-1,2,4-triazine (ferrozine) was utilized to quantify the total reducing capacity (TRC) of teas leading to sound results. 7ecently, a comprehensive study based on the reduction of Fe III in aqueous solution (pH 8.0, tris(hydroxymethyl) aminomethane (tris) buffer) containing the 3-hydroxy-4-nitroso-2,7-naphthalenedisulfonic acid was carried out.Several antioxidant agents were evaluated with this alternative FRAP assay before it was effectively used to determine the reducing capacity of aqueous extracts of many medicinal plants. 8ll these FRAP assays have in common the use of an electron-transfer reaction between the antioxidants (present in the samples) and the oxidant agent (Fe III /complexes), both in the same solution, but whose order of added reagents may be quite different.
It is well known that Fe II forms with 3-fold excess of the organic bidentate ligand 2,2'-bipyridine (bipy; Figure 1) a very stable chelate Fe(bipy) 3 2+ (log β 3 = 17.2 at 25 o C). 9 This aqueous orange-red complex shows a maximum absorption at 521 nm (ε 521 nm = 7.5 × 10 3 L cm -1 mol -1 ) 10 and has been commonly used for direct determination of total iron content in different type of samples after reduction of Fe III by addition of a suitable reducing agent. 11Consequently, if Fe III and bipy (1:3 ratio) are in excess when compared to the reducing agent, it is possible to determine indirectly this own reducing agent based on the formation of the Fe(bipy) 3 2+ complex.
Based on those findings, the reduction reaction of Fe(bipy) 3 3+ to Fe(bipy) 3 2+ complex was used in this present work to determine the reducing capacity of several standard AOs (mostly phenolic acids and flavonoids).This detailed study elucidates which is the oxidation ability of Fe(bipy) 3 3+ complex towards these AOs under the same experimental conditions used in the aforementioned study developed to the quantification of polyphenol content (pH 4.6; acetate buffer). 12dditionally, this same redox reaction was used to develop a spectrophotometric FRAP method to quantify the reducing capacity of aqueous extracts of twelve Brazilian medicinal plants.The plants analyzed have been used as a food source, for their healing properties (utilized in folk medicine) and in religious rituals.Besides, this reaction was also employed to quantify the TRC of twelve teas found in the local market and largely consumed by the population.
For comparison purposes, TRC values obtained with the suggested method were compared with two wellestablished methods.Regarding teas samples the TRC values obtained were compared with the cupric reducing antioxidant capacity (CUPRAC) method which is based on reduction of Cu II to Cu I in the presence of neocuproine. 13For medicinal plants the TRC values were checked out with the method based on the extinction of the free radical derived from the 2,2'-azino-bis-3-ethylbenzothiazoline-6-sulfonic acid diammonium salt (ABTS). 14inally, the TRC values of the two groups of samples (teas and medicinal plants) obtained with the proposed method were also compared with the total polyphenolic content values obtained with the Folin-Ciocalteu reagent (FCR), as recommended by Brazilian Pharmacopoeia. 15

Apparatus
All spectrophotometric measurements were made in an HPUV 8453 (Agilent, USA) spectrophotometer using a 1.00 cm glass cell.

Materials
Reverse osmosis water (Quimis Q842-210, Brazil) was used to prepare the analytical-grade chemicals and in all sample dilutions (except when another solvent is indicated).

Preparation of tea samples
A procedure previously described was used for the preparation of tea samples. 19Briefly, 300 mg of commercial tea were transferred to a 100.0 mL beaker containing 50 mL of water and kept in water bath (65 o C, 30 min).After cooling, this solution was transferred to a 100.0 mL volumetric flask, completed with water and then filtered.When necessary a 5-fold dilution was used, transferring 5.0 mL of this solution to a 25.0 mL volumetric flask.
These aqueous samples were used for the total polyphenolic content (TPC) determination, with FCR, and for the TRC quantification with CUPRAC and the proposed method.

Aqueous samples
The extraction procedure described in previous studies 8,19 was used to prepare the aqueous extracts of medicinal herbs.These aqueous extracts were used for two quantifications: the TPC, with FCR, and the TRC, with the proposed method.

Samples in organic solvents
The extracts in methanol/acetone mixture were obtained as described elsewhere and used to quantify the antioxidant capacity with the ABTS •+ method. 14,20C determination with the FCR The multiple standard addition method used for the TPC quantification was already described. 8,12,16For teas, a 0.0188 mg mL -1 GA standard solution was used to express the TPC (mg GA g -1 dry material) as used in another study. 19For the aqueous extracts of medicinal herbs, a 0.0189 mg mL -1 PA standard solution was used to express the TPC (g PA 100 g -1 dry material) as recommended by Brazilian Pharmacopoeia. 15ducing capacity quantification with proposed method Calibration graph with a standard antioxidant (AA) In eight 5.0 mL volumetric flasks the following reactants were added: 0.50 mL of 4.90 mg mL -1 Fe 2 (SO 4 ) 3 solution, different volumes (0.2 to 0.9 mL) of a 0.0352 mg mL -1 AA standard solution, 0.50 mL of acetate buffer (HAc/KAc; pH 4.6) solution and 1.0 mL of 2.58 mg mL -1 bipy solution.The AA final concentration (C AA ) was (1.41-6.34)× 10 -3 mg mL -1 .Absorbance measurements at 521 nm (A 521 nm ) were recorded using a mixture containing 0.490 mg mL -1 Fe 2 (SO 4 ) 3 and 0.515 mg mL -1 bipy in the same acetate buffer solution as reference solution (blank reagent).A calibration graph (A 521 nm vs. C AA , where C AA is in mg mL -1 ) obtained is described by the equation A 521 nm = a + b C AA .
Calibration graphs with some phenolic compounds (PC) Calibration graphs with standard PC were performed like the one made with AA standard solution.For each PC analyzed, at least three calibration graphs (A 521 nm vs. [PC], where [PC] is the concentration of PC in mol L -1 ) were obtained.In these studies, only straight lines were considered, originating from the calibration graphs that showed very good linearity (adjusted correlation coefficient (r 2 ) ≥ 0.99) described by the equation A 521 nm = a + b [PC].These calibration graphs need to be obtained in mol L -1 in order to comply with the definition of the reducing capacity of standard AOs. 7,8,21libration graphs with samples (teas and aqueous extracts of medicinal herbs) In five 5.0 mL volumetric flasks were added: 0.50 mL of 4.90 mg mL -1 Fe 2 (SO 4 ) 3 solution, 100 to 1000 µL (depending on the kind of sample) of 3.0 mg mL -1 aqueous extracts of teas or herbs (both obtained with dry material), 0.50 mL of acetate buffer solution (pH 4.6) and 1.0 mL of 2.58 mg mL -1 bipy solution.A 521 nm were recorded using the same blank reagent above described.A calibration graph (A 521 nm vs. C DM , where C DM is the dry material (DM) concentration in mg mL -1 ) obtained is described by the equation A 521 nm = a' + b' C DM .

Calculation of reducing capacity of standard PC
The reducing capacity of each PC investigated was expressed as ascorbic acid equivalent capacity (AA EC ), defined as the concentration in 10 -3 mol L -1 of AA standard solution which presented a reducing capacity value equivalent to a 1.0 × 10 -3 mol L -1 of PC solution under the same experimental conditions. 7,8,21lculation of reducing capacity in samples (aqueous extracts of teas and medicinal herbs) The equation A 521 nm = a + b C AA is applied to calculate the A 521 nm value corresponding to a 1.0 mg mL -1 AA standard solution.This A 521 nm value is replaced in the equation A 521 nm = a' + b' C DM providing the concentration (mg mL -1 ) of the solution analyzed (and the corresponding DM mass), which is equivalent to the TRC of a 1.0 mg mL -1 AA.The TRC values obtained (corrected to 5-fold dilution when necessary) were expressed as g DM g -1 AA and can be more easily calculated using the equation 1: (1) where a, b, a' and b' are the coefficients of the straight line equations above described, fd is the dilution factor and m DM is the mass (in grams) of dry material.

Determination of the reducing capacity of tea with CUPRAC reagent
This method, based on the reduction of Cu II to Cu I in solution containing neocuproine (pH 7.0), was performed as described elsewhere. 13,19termination of the TRC of herbal extracts using the ABTS •+ The preparation of ABTS •+ solution and the procedure used here were carried out as previously described. 14he antioxidant capacity values were expressed in µM Trolox g -1 dry material.

Results and Discussion
Bipy is partially protonated in aqueous solutions in pH < 4.0 (pK a1 = -0.2;pK a2 = 4.3), 9-11 and ferric hydroxo complexes (e.g., FeOH 2+ and Fe(OH) 2 + ) may be present in unbuffered solutions in pH > 3.5. 22Thus, in the present study, the pH was maintained at 4.6 with acetate buffer solution, which has also been used with the same reduction reaction of Fe III to Fe II in a solution containing bipy, in a recently proposed method for the quantification of the polyphenolic content in medicinal plants. 12he above considerations support the experimental conditions adopted in this proposed method (0.490 mg mL -1 Fe 2 (SO 4 ) 3 , 0.515 mg mL -1 bipy as final concentration at pH 4.6 kept with acetate buffer).The procedure described here can be performed in few minutes and it might be adapted for flow injection analysis, though it is not the purpose of the present study.
AA was chosen and used as the standard antioxidant to express the reducing capacity due to its fast reaction (ca. 10 min), low cost and being biologically active.A typical calibration graph (A 521 nm vs. C AA ) obtained from the absorption spectra (Figure 2) leads to a straight line described by the equation y = -0.0217+ 115x (n = 8; adjusted r 2 = 0.997) for a linear range from (1.41 to 6.34) × 10 -3 mg mL -1 AA (Figure 2, inset).The angular coefficient, defined as apparent absorptivity (at 521 nm) for AA, was 115 ± 4 mL cm -1 mg -1 for 20 calibration curves (relative standard deviation (RSD) = 3.3%).
Trolox (a water-soluble compound analogous to vitamin E) also reduces Fe III to Fe II in solution containing bipy, but has a current cost about 35 times greater (considering a pack of 25 g) and almost half of the capacity reduction value of AA (AA EC value of 0.79, Table 1).

The reducing capacity of polyphenolic compounds
In Table 1, there is a basic structure of phenol that helps in the interpretation of AA EC obtained for the antioxidant compounds investigated.
Tannic acid has the highest AA EC value (7.07), which is due to the highest number of hydroxyl groups (HG).
GA (pK a1 = 4.4) and its isomer 2,3,4-THB (pK a1 = 3.0) have the -COOH group partially deprotonated under these experimental conditions (pH 4.6). 9In 2,3,4-THB the -COOH group is in vicinal position to the three HG, but in GA the -COOH group is symmetrically opposed to the three HG.This seems to be the reason that GA (AA EC 2.76) has a reducing capacity value 2 times higher than 2,3,4-THB (AA EC 1.39).In fact, it has been pointed out that the less acidic the phenol the easier its oxidation. 5n addition, (-)-epigallocatechin gallate (an ester of GA with epigallocatechin) has an AA EC value (2.44) about 10% lower than the GA, despite having eight HG.This shows that it is not only the number of HG that influences the reducing capacity of polyphenols, but also their acidity and the position of HG in the benzene ring. 6,23mong benzenotriols isomers the 1,2,4-benzenetriol (AA EC 2.74) is a stronger reducing agent than PA (AA EC 2.14).It is well known that the HGs in ortho position increase the reducing capacity, but the presence of an HG in the C2 position of PA makes it a weaker reducing agent than 1,2,4-benzenetriol (HG in C1 and C3 positions).In phloroglucinol the 3 HGs are proportionally distributed in the benzene, which strongly decreases the AA EC value to 0.07.All these observations are in agreement with theoretical information. 24,25egarding benzenediols, the AA EC values follow the order: o-pyrocatechol (1.03) > hydroquinone (0.82) > resorcinol (0.01).It shows that reducing capacity is higher with second HG in ortho position and that the para position provides an electron donation more easily than meta position. 25These results are in accordance with previous experimental findings 8,21 and also with theoretical information, which points out that oxidation of phenols to quinones seems to be easier if two HG are in ortho or para positions in the benzene ring. 24,25henol (pK a = 9.8), 9 with only one HG, is the weakest reducing agent evaluated with the lowest AA EC value (0.001), confirming that HG participates actively in the reduction reaction of Fe III to Fe II in solution containing bipy.
The introduction of radicals (other than -OH) in the aromatic ring modifies significantly the AA EC value when compared to phenol.For instance, addition of -CH=CH-COOH group (which happens to be an electron-releasing radical) into C3 position originates p-coumaric acid (pK a1 = 4.64), 9 which has an AA EC value (0.10) 100 times higher than phenol.Adding a donating group -OCH 3 in C1 position of p-coumaric acid forms ferulic acid (pK a1 = 3.60) 9 that presents an AA EC value (0.66) about 600 times higher than phenol.Another -OCH 3 group added at C5 position of ferulic acid gives the sinapic acid (pK a1 = 4.58) 9 that presents an AA EC value (0.99) about 1000 times higher than phenol.These three Absorption spectra of (a) 0.490 mg mL -1 Fe 2 (SO 4 ) 3 , 0.515 mg mL -1 bipy at pH 4.6 with acetate buffer solution; (b) to (i) (1.41, 2.11, 2.82, 3.52, 4.23, 4.93, 5.64 and 6.34) × 10 -3 mg mL -1 ascorbic acid (AA) + (a), respectively, using water as reference solution.Inset: calibration curve for AA using the A 521 nm of the Fe(bipy) 3 2+ complex (b = 1.0 cm).monohydroxylated phenols have a -COOH group and the AA EC values do not seem to be strictly connected to the acidity conditions of proposed method (pH 4.6).In fact, in sinapic acid the presence of the two -OCH 3 adjacent to the HG favors significantly its reduction capacity, which is in agreement with theory. 24,25affeic acid (pK a1 = 3.0) 9 is a dihydroxylated cinnamic acid derivative, has one more -OH group than p-coumaric acid, being both -OH in opposite position to the -CH=CH-COOH group.The presence of this second -OH group (in ortho position) increases ten times the reducing capacity value (AA EC 1.07) with respect to p-coumaric acid.In this case, the number of HG contributes more strongly to the AA EC value than the acidity of the phenol derivative. 8,23anillin (a phenolic aldehyde with an -OCH 3 in C1 position) is the main component of the vanilla seed extract.Vanillic acid (pK a = 4.45) 9 is an oxidized form of vanillin.Although partially dissociated in these experimental conditions, vanillic acid has an AA EC value (0.25) about 9-fold higher than vanillin (AA EC 0.03), probably due to the proton dissociation.
The number of -OH group also plays an important role in the AA EC values of flavonoids, another class of antioxidant compounds that exhibit great reducing capacity. 26,27Two flavonoids with the same framework (quercetin and rutin) were analyzed using the proposed method.Quercetin has an AA EC value (3.24) 2.5 times greater than rutin (AA EC 1.29), which can be attributed to the replacement of an -OH group in quercetin by a disaccharide group in rutin, in agreement with the theory. 6en though the reaction of a polyphenol standard solution may not reproduce the analytical response of complex matrices (like extracts of medicinal herbs or teas), the results present in Table 1 are useful for assessment of the reactivity of single polyphenol.As expected, these data revealed that type, number, and position of a given chemical radical (mainly -OH groups) attached to the benzene ring change the reducing capacity values obtained with the proposed method.
Eventually, as the linear range of most of the phenolic acid derivatives analyzed (Table 1) is between (1-500) × 10 -6 mol L -1 (with exception of resorcinol, phenol and vanillin), the procedure presented here can be used in more diluted samples.The results might be expressed in another standard compound instead of AA.In this context, a polyphenol with a high AA EC value, but with a more affordable cost (e.g., GA or quercetin), could be used.

The reducing capacity of teas samples
Table 2 shows the AA EC results for twelve teas.The TRC values obtained with the proposed method (Fe(bipy) 3 2+ complex) and CUPRAC reagent had an excellent agreement (adjusted r 2 = 0.951).This shows that despite the different values of the conditional reduction potential of the Fe III /Fe II couple in solution containing bipy (1.08 V vs. normal hydrogen electrode (NHE)) 11,12 and Cu II /Cu I in neocuproine medium (0.635 V vs. NHE), 18,19,28 both seem to oxidize (at least proportionally) the compounds present in the tea samples.Since it was observed a good correlation between the TPC quantified with the Folin-Ciocalteu reagent and the TRC values obtained with CUPRAC reagent (adjusted r 2 = 0.811) and with the proposed method (adjusted r 2 = 0.816), it can be assumed that in these samples the agents responsible for the reducing capacity are polyphenols.

The reducing capacity of aqueous extracts of medicinal herbs
TRC values obtained with both assays (proposed and ABTS •+ methods) showed a very good agreement (adjusted r 2 = 0.975), indicating that both methods can be used to quantify the reducing capacity of herbs.The proposed method has the advantage of being conducted in aqueous solution, unlike ABTS •+ method, which uses organic solvents such as acetone and methanol.These two procedures do not present much difference in the completion time of the reaction, although the ABTS •+ solution requires at least 16 h of previous preparation. 14,29egarding the price of reagents (ABTS is currently about 14 times more expensive than bipy) the proposed method is more attractive from an economic point of view.Besides, the suggested method is conducted in aqueous medium and the ligand used (bipy) can be recycled, making it environmentally attractive. 30n addition, good correlations between the TPC quantified with the Folin-Ciocalteu reagent and the reducing capacity values obtained with the ABTS •+ method (adjusted r 2 = 0.792) and the proposed method (adjusted r 2 = 0.835) were found, showing that the polyphenols present in these herbs should be responsible for this reducing capacity (Table 3).
Additionally, the results of the TRC obtained with the Fe(bipy) 3 2+ complex for teas and herbs suggested that the proposed method can also be used to quantify the reducing capacity of other samples derived from plants that are rich in polyphenolic compounds (e.g., fruits and fruit juices).
As other methods based on the reduction of metal ion M n+ to M (n -1)+ developed to quantify the TRC (in a solution containing a complexing agent for M (n -1)+ ), the assay suggested here does not require a lag phase type of measurement.In this context, the values of TRC can also be used to express the total antioxidant capacity.

Conclusions
The method suggested here for quantifying the total reduction capacity of teas and herbs is simple, fast, reliable and easy to perform.The good results obtained allow us to infer that the proposed procedure can also be used to quantify the reducing capacity of other samples of plant origin (e.g., fruit juices, beers and wines).This study also revealed that the reducing capacity of polyphenolic compounds with the proposed method depends on their chemical structure (mainly the presence and position of hydroxyl groups).
Both the equipment (spectrophotometer) and the reagents (iron(III) sulfate, 2,2'-bipyridine and acetate buffer) used in the proposed method are not expensive, so they can be adopted by laboratories performing routine analyses.Furthermore, as this method is conducted in aqueous medium and the ligand (2,2'-bipyridine) can be recycled, it becomes environmentally attractive.

A
Procedure for Assessment of the Reducing Capacity of Plants-Derived Beverages Based on the Formation of the Fe II /2,2'-Bipyridine Complex Rafaela L. Sacchi, a Waila E. L. Santana, a Cecilia V. Nunez b and Horacio D. Moya * ,a

Table 1 .
Parameters of the linear regression of the calibration graphs (A 521 nm = a + b [PC]) and reducing capacity values (AA EC ) of some polyphenolic compounds obtained with the proposed method a Trolox and ascorbic acid were included in this table for comparison purposes.FM: formula mass; HG: hydroxyl group in benzene ring; HGP: position of the hydroxyl group on the benzene ring; LR: linear range; n: number of points of the calibration graphs; a, b and r 2 : linear, angular and correlation coefficients of the calibration graphs, respectively; AA EC : reducing capacity expressed as ascorbic acid equivalent; 2,3,4-THB: 2,3,4-trihydroxybenzoic acid.