A Computational Study on the Acidity Dependence of Radical

Jul 7, 2010 - Quantitative studies on structure–ORAC relationships of anthocyanins from eggplant and radish using 3D- ... Food & Function 2012 3, 79...
0 downloads 0 Views 198KB Size
Download PDF
9706

J. Phys. Chem. B 2010, 114, 9706–9712

A Computational Study on the Acidity Dependence of Radical-Scavenging Mechanisms of Anthocyanidins Laura Este´vez, Nicola´s Otero, and Ricardo A. Mosquera* Departamento de Quı´mica Fı´sica, Facultade de Quı´mica, UniVersidade de Vigo, Lagoas-Marcosende s/n 36310-Vigo, Galicia, Spain ReceiVed: May 6, 2010; ReVised Manuscript ReceiVed: June 14, 2010

On the basis of quantum chemical calculations, the radical-scavenging property attributed to anthocyanidins was analyzed considering three mechanisms: hydrogen atom transfer (HAT), stepwise electron-transfer-protontransfer (ET-PT), and sequential proton loss electron transfer (SPLET). We found that the activity of anthocyanidins and the mechanism through which they react are pH-dependent, because the diverse colorful forms in which anthocyanidins may exist in prototropic equilibria (cationic, neutral, anionic) are susceptible to experience each of the mechanisms proposed. According to redox parameters calculated, we can conclude that HAT is always the most favored of the generally accepted mechanisms to scavenge reactive oxygen species (ROS) by the three colored forms. Nevertheless, only neutral and anionic forms are found to be able to scavenge 1,1-diphenyl-2-picrylhydrazyl (DPPH · ) radical through HAT and SPLET mechanisms from a thermodynamical point of view, whereas ET-PT is only feasible for anions. Sequential proton loss hydrogen atom transfer (SPL-HAT) is proposed as the only pathway for the reaction between anthocyanidin cations and the DPPH · radical. It should be viable according to our quantum mechanical calculations and even competitive with typical HAT, ET-PT, and SPLET. Introduction Antioxidant compounds in food play an important role as a health-protecting factor. Most of the antioxidant compounds in a typical diet are derived from plant sources and belong to various classes of compounds with a wide variety of physical and chemical properties. The main characteristic of an antioxidant is its ability to trap free radicals. Antioxidant compounds such as polyphenols and flavonoids scavenge free radicals such as peroxide and hydroperoxide and thus inhibit the oxidative mechanism that lead to degenerative diseases.1,2 Anthocyanins, a parent class of flavonoids, are a group of naturally occurring phenolic antioxidants found in fruits and vegetables.3 They have been shown to efficiently trap lipid peroxyl radicals3a,4 and other reactive oxygen species (ROS) such as superoxides.5 Three pathways are commonly discussed to explain the mechanism through which phenolic antioxidants (ArOH) scavenge radicals: H-atom transfer (HAT, eq 1); stepwise electrontransfer-proton-transfer (ET-PT, eq 2) and sequential proton loss electron transfer (SPLET, eq 3).

ArOH + R · f ArO · +RH

(1)

ArOH + R · f ArOH · + + R- f ArO · +RH

(2)

ArOH f ArO- · ;ArO- + R · f R- + ArO · ; R- + H+ f RH (3) The first pathway (eq 1) corresponds to the homolytic dissociation of an O-H bond.6 As it does not involve charge separation, it has been considered the dominating mechanism in nonpolar solvents.7 Feasibility of this mechanism depends * Corresponding author. E-mail: [email protected].

on the balance of two bond dissociation enthalpies (BDEs), that of the O-H group in ArOH and that of bond established by the radical (R · ) and abstracted hydrogen to yield RH. Pathways 2 and 3 should be favored in polar solvents,8 and their feasibility is mainly measured by the ionization potential (IP) of the antioxidant (eq 2) or anions derived after proton loss (eq 3).9 On one hand, the measurement of antiradical activity of polyphenols antioxidants is performed along the literature by using different methods based on their reaction with diverse radicals in different solvents. It is usually found that the same antioxidant may exhibit distinct behaviors in scavenging different radicals and/or in different environments.7–9 Antiradical activities should therefore only be compared when the measurements are made in the same conditions. All of these features result in frequent debates on the action mechanisms of antioxidants.10 In the specific case of anthocyanins and anthocyanidins, the elucidation of the mechanism of action as radical scavenger and antioxidant has one extra complication because of the series of reversible structural transformations exhibited by anthocyanins upon changing the pH of the medium, which includes PT between colored forms.3b On the other hand, accurate estimation of BDE from theoretical calculations is a challenging task, since high levels of calculations are necessary for taking into account the effect of both dynamical and nondynamical part of electron correlation. High level ab initio calculations are in most of the cases prohibitive for large size molecules. Moreover, it was corroborated that MP2 and MP4 methods significantly overestimate the absolute O-H BDE value.11 In contrast, computational studies hitherto carried out on O-H BDEs of phenolic compounds make use of density functional theory (DFT)-based methods.11–15 In addition, Wright et al.,16 based on the gas-phase BDEs and IPs calculated by DFT methods, have concluded that, in most hydrogen-transfer reactions of phenolic antioxidants, HAT will be dominant rather than ET. More recent papers,

10.1021/jp1041266  2010 American Chemical Society Published on Web 07/07/2010

Radical-Scavenging Mechanisms of Anthocyanidins

J. Phys. Chem. B, Vol. 114, No. 29, 2010 9707

∆H2 ) H(ArOH+ · ) + H(R-)-H(ArOH) + H(R · ) ∆H3 ) H(ArO · ) + H(R-)-H(ArO-) + H(R · ) Readers should bear in mind that our results and conclusion can be applied to systems in which concentrations of ArOH are sufficiently low to ensure that there is no ArOH selfassociation via H-bonding or weaker interactions. Figure 1. Basic chemical structure of anthocyanidins. The labeling scheme for the three rings (A, B, C) and atom numbering are also shown.

supported by B3LYP calculations17 and experimental data,9 report that HAT and SPLET are more feasible than ET-PT. Also, Martinez18 makes use of DFT to calculate IP and EA for diverse compounds including anthocyanidins. Despite their wide use, DFT methods are known to be affected by additional sources of error.19 Among them, an error not shared by Hartree-Fock theory arises from the spurious interaction of an electron with itself, usually known as selfinteraction error (SIE). Recently, the M05 family of functionals, developed by Truhlar and co-workers,20 has been optimized in such a way to ameliorate some of the problems caused by spurious self-exchange, and, furthermore, the authors claim these functionals are completely free of self-correlation error. In this way, the accuracy of the values of thermodynamic parameters (BDE, IP, EA, etc,) obtained from B3LYP calculations are analyzed by comparing with those obtained from M05 in order to evaluate how large this error can become with B3LYP and if it could be a serious problem in its application to systems and ∆Hi values such as those studied here. Furthermore, parameters obtained from both DFT functionals will be also compared with those obtained from coupled cluster (CC) calculations (see below). Overall, this work attempts to be helpful to shed more light on this front by estimating in which way (i) solvent, (ii) the acid-base form in which anthocyanidins are present, and (iii) the radical itself could alter the radical scavenging power of anthocyanidins and/or the pathway through they react. These tasks are discussed on the basis of redox properties obtained by quantum-chemical calculations. Solvent effects were taken into account with the polarized continuum model (PCM)21 to model protic (water and methanol) and nonprotic (cyclohexane) media. Protopropic equilibria in anthocyanidins were taken into account considering that ArOH (eq 1-3) represents cationic, neutral, or anionic forms of each natural anthocyanidin (Figure 1): pelargonidin (Pg), cyanidin (Cy), delphinidin (Dp), peonidin (Pn), petunidin (Pt), and malvinidin (Mv), so that, each one of those forms is susceptible to react with radicals through the mechanisms proposed (Scheme 1). Different radical species, R · , were considered keeping in mind that the relative importance of HAT or ET-PT is not only determined by microenvironmental features (lipid phase, aqueous phase) but also governed by the characteristics of the scavenged radical.22 In this way, we computed redox properties for 1,1-diphenyl-2-picrylhydrazyl radical (DPPH · ),23,24 methoxyl (MeO · ) and methyl peroxy (MeO2 · ) radicals. Thus, we calculated the electron affinity (EA) for these radicals to obtain enthalpy differences involved in pathways 2 (∆H2) and in the second step of pathway 3 (∆H3). Also, we computed their H atom affinities (HAAs) to obtain ∆H1 (pathway 1).

∆H1 ) H(ArO · ) + H(RH)-H(ArOH) + H(R · )

Computational Details First of all, we have checked the reliability of the parameters obtained from B3LYP hybrid density functional25 calculations by comparing with those obtained from CCSD(T)/6-31++(d,p)// CCSD/6-31++(d,p) calculations on simpler models of ArOH as phenol, p-hydroxybenzyl cation and catechol reacting through HAT, SPLET, or ET-PT mechanisms with ROS or H2N2H · radical, which was selected as the simplest model for DPPH radical. We have also performed those calculations by using the improved M05-2X functional (see below). For anthocyanidins, geometry optimization of each ArO and ArOH+ radical was performed starting from the optimized structure of the parent molecule (ArOH),26 after removing an electron from the molecule or the H atom from any O-H group, with the unrestricted UB3LYP and the restricted B3LYP, respectively, by using the 6-31++G(d,p) basis set. Concerning the DPPH · radical, the initial geometry was obtained from three-dimensional X-ray diffraction data.27 It was optimized at the same computational level and confirmed as a true minimum by vibrational analysis. The same was done for the related species, such as the closed shell species DPPHH, and the corresponding anion formed by accepting an electron, DPPH-, to obtain HAA and EA, respectively. The effect of different solvents was checked for cyanidin. Thus, BDEs and IPs were calculated in water, methanol, and cyclohexane using the B3LYP/6-311++G(2d,2p)//B3LYP/631++G(d,p) level and at the corresponding unrestricted one. Corresponding EAs and HAAs of ROS and DPPH were also calculated. It has to be stressed that larger basis sets did not improve the accuracy (see below). In all cases,28 gas phase IP and O-H BDE for all the hydroxyl groups were calculated for the most stable cation, neutral, and anion of each anthocyanidin, including zero point vibrational energy (ZPVE) and thermal correction to the enthalpy of the final and initial states. The IP was calculated as H(ArOH+ · ) H(ArOH) (eq 2) or as H(ArO · ) - H(ArO-) (eq 3), and the O-H-BDE was calculated as H(ArO · ) + H(H · ) - H(ArOH). We also calculated the gas phase O-H proton dissociation enthalpies (OH-PDEs), involved in the SPLET mechanism and also in the second step of the ET-PT mechanism. O-H PDE is defined as H(ArO-) + H(H+) - H(ArOH). For DPPH · , EA and HAA were obtained, respectively, as H(R-)DPPH- H(R · )DPPH · and H(RH)DPPHH - H(R · )DPPH · - H(H · ). It was also necessary to obtain the proton affinity (PA) for the DPPH · and DPPH- species. Geometries for alkoxyl (MeO · ) and peroxyl (MeO2 · ) radical species were fully optimized at the same level in order to obtain the corresponding EA, HAA, and PA parameters. Solvent (water, methanol, and cyclohexane) effects were taken into consideration by employing the self-consistent reaction field (SCRF) method with PCM. Thermodynamical parameters in solution were calculated in terms of ∆Gsol, meaning that cavitation, dispersion, and repulsion energies are also included.26c,28 All these quantum chemical calculations were accomplished using the Gaussian 03 software.29

9708

J. Phys. Chem. B, Vol. 114, No. 29, 2010

Este´vez et al.

SCHEME 1: Schematic Representation of the Mechanisms Proposed for the Radical Scavenge by Anthocyanidinsa

a The scheme shows how the three colored acid/base forms (cation, AH3+, neutral, AH2, anion AH-) of anthocyanidins could react through any of them.

TABLE 1: BDE and IP Values (in kcal mol-1) for Simpler Phenols as well as HAA and EA Values (in kcal mol-1) for ROS and for the H2N2H Radical Selected As the Simplest Model of DPPHa C6H5OH CCSD(T) M05-2X B3LYP a

CH2C6H4OH

MeO ·

C 6 H6 O2

MeO2 ·

H2 N2 H ·

BDE

IP

BDE

IP

BDE

IP

HAA

EA

HAA

EA

HAA

EA

90.31 93.93 89.27

189.52 193.73 190.67

115.21 118.28 114.57

328.81 331.46 329.73

81.88 84.93 80.04

181.60 185.81 182.07

-106.44 -109.21 -106.80

-18.98 -30.89 -33.59

-87.04 -88.04 -86.54

-7.69 -21.56 -24.51

-77.76 -76.32 -78.03

12.31 9.23 6.52

Values are obtained from DFT (B3LYP and M05-2X) and CCSD(T) calculations using the 6-31++G(d,p) basis set.

Results and Discussion Evaluation of the Method of Choice. Truhlar and coworkers30 have proposed a simpler criterion, called B1, for characterizing the multireference character of a bond, and therefore one should choose M05 or the M05-2X depending on the B1 value.20b B1 for the O-H bond was obtained, and only single-reference behavior was encountered (B1 < 10 kJ mol-1). In that case, one can obtain higher quantitative accuracy by switching to M05-2X. Thus, we have employed the M052X functional to obtain alternative values for redox properties. Table 1 collects the BDE, IP, HAA, and EA values obtained from DFT (B3LYP and M05-2X) calculations and those obtained from CC for simpler selected models (CC calculations are not feasible for anthocyanidins). It can be observed that no significant differences are found between data provided by CCSD(T) and DFT for almost all studied thermodynamical parameters. Particularly, BDE values obtained from CCSD(T) and B3LYP agree within 1 kcal mol-1, if we exclude the O-H BDE of C6H6O2. The same good agreement was found for HAA values and also for IPs. In contrast, larger discrepancies are found for EA values due to the smaller molecular size of the anions (CH3O-, CH3O2- and H2N2H-) that leads to larger SIE, as it has been discussed elsewhere.18 Despite the deviations in EA values with both methods, equivalent conclusions can be extracted from DFT and CC methods about the balance between HAT and ET mechanisms because IP values are always large enough to make ET not feasible. Therefore, data obtained from CCSD(T) and DFT offer the same information for isodesmic reactions, even when the chemical bond broken differs from that formed. We can conclude that there is no reason not to use DFT for the tasks discussed in this work. Finally, we have also noticed that M052X displays, in general, larger deviations than B3LYP calculations compared to CCSD(T). Hereafter, all results and discussion about it are referred to as DFT-B3LYP calculations.

TABLE 2: The lowest O-H BDEa Values (in kcal mol-1), Obtained from the Most Stable Conformer and Tautomer of Each Form (Cation, AH3+, Neutral, AH2, and Anion, AH-) for the Anthocyanidins Studied Here in the Gas Phase and in Aqueous Solution Simulated with PCM (in Italics) at the B3LYP/6-31++G(d,p) Level Pg Cy Dp Pn Pt Mv

AH3+ f AH2+ ·

AH2 f AH ·

AH- f A- ·

81.09/ 89.94 81.21/ 88.26b 75.49b/ 83.62b 80.54/ 89.66b 80.17b/ 85.82b 78.38/ 88.41b

68.78/ 69.10/ 69.08/ 69.01/ 70.21/ 67.57/

60.72/ 75.99 60.70/ 75.69 61.31/ 75.12 60.70b/ 75.88 61.41/ 75.49 61.83b/ 74.92

81.45 81.57 80.44 81.54 81.00 80.21

a The lowest O-H BDE values is almost always at the 3-OH group. b Indicating that the lowest O-H BDE values is at the 4′-OH group.

HAT Pathway. In order to estimate the contribution of the HAT mechanism in anthocyanidins, BDEs were calculated for each OH group of the six anthocyanidins (Figure 1) in each of their three colored acid/base forms (cationic, neutral, and anionic). In fact, anthocyanidins can be regarded as a source of n H atoms (as many as OH groups) that will convert radicals into the corresponding closed-shell molecule. Therefore, all OH groups in anthocyanidins could contribute in the radical scanvenging process through HAT. Here, we only report the lowest value for each anthocyanidin in the three-acid/base species (Table 2) to discuss the feasibility of HAT. In cations, the most stable radicals are those formed by H-abstraction in the 4′OH group (4′-O · ) for anthocyanidins containing a catechol moiety in the B-ring (Cy, Dp, and Pt), whereas the 3-OH abstraction leads to the most stable radical (3-O · ) when this moiety is not present. Nevertheless, there is also a significant contribution of the 3-O · radical in the former case. For neutral anthocyanidins, the most populated tautomers are usually those formed by deprotonating the hydroxyl at positions 7 or 5, named, respectively, as N7 and N5. In those

Radical-Scavenging Mechanisms of Anthocyanidins

J. Phys. Chem. B, Vol. 114, No. 29, 2010 9709

Figure 2. Radical species, formed after H-atom abstraction, from the most populated rotamers of cations (C), for the most populated neutral and anionic tautomers (higher than 20%) (N and A, respectively), obtained from PCM (in water) calculations.

TABLE 3: ∆H1 Values (in kcal mol-1) Obtained at the B3LYP/6-31++G(d,p) Level in Aqueous Solution Modeled with PCM cation Pg Cy Dp Pn Pt Mv

neutral

anion

DPPH ·

MeO ·

MeO2 ·

DPPH ·

MeO ·

MeO2 ·

DPPH ·

MeO ·

MeO2 ·

6.14 4.46 -0.18 5.86 2.02 4.61

-21.96 -23.64 -28.28 -22.24 -26.08 -23.49

-2.93 -4.61 -9.25 -3.21 -7.05 -4.46

-2.35 -2.23 -3.36 -2.26 -2.8 -3.59

-30.45 -30.33 -31.46 -30.36 -30.9 -31.69

-11.42 -11.3 -12.43 -11.33 -11.87 -12.66

-7.81 -8.11 -8.68 -7.92 -8.31 -8.88

-35.91 -36.21 -36.78 -36.02 -36.41 -36.98

-16.88 -17.18 -17.75 -16.99 -17.38 -17.95

cases, the 37-O · and 35-O · are the lowest energy radicals if we exclude Cy and Dp, where 34′-O · radicals are the most favored (N4′ being in these cases the most stable tautomer of neutral forms),26 showing that the 3O-H BDE is the lowest one for all neutral anthocyanidins. For anionic species, the most populated tautomers are those resulting from the deprotonation of 7OH and 4′ OH (named A74′) or 5OH and 4′OH (A54′).27 We found two contributing radicals for anions, 374′-O · and 354′-O · (with 3O-H BDE being the lowest one again). PCM calculations keep this description for anions and neutral forms, so that the most stable radical is still formed by homolytic dissociation of the 3OH group, but in cations, the 4′-O · radical is found to be the most stable one independently of the substitution pattern mentioned above excluding Pg (Figure 2 and Table 2). The statement above agrees with previous studies in which it is postulated that the maximal radical scavenging activity for flavonoids is achieved if there is a free hydroxyl group in ring C (3-OH).31 For all neutral forms and anions, the general rule is that the 3-O · radical is always the most favored one (obtained through the lowest BDE) independently of the substitution pattern. We also notice BDE values decrease when pH increases. This agrees with the increase in electron delocalization promoted by deprotonation. It stabilizes the radical anions more than the corresponding closed-shell anion. Hence, deprotonation lowers the BDE values of the remaining OH groups. According to eq 1, the lower the OH-BDE, the more active the antioxidant. Comparing the three colored species, anions should be the most active form of anthocyanidins for scavenging radicals through the HAT mechanism, both in gas as in solution phases (Table 2). They are followed by the neutral and cationic forms. Although it is usually considered that radicals react with phenols via hydrogen abstraction, there is controversy when the radical is DPPH · .24 Thus, when the HAA value of the DPPH · radical (-75.30 kcal mol-1, in gas phase) is considered, we obtain that HAT is forbidden for cations (Table 2). In contrast, neutrals and anions give rise to negative enthalpies (absolute

TABLE 4: IP Values (in kcal mol-1) Obtained for the Most Stable Conformer and Tautomer of Each Form (Cation, AH3+, Neutral, AH2, and Anion, AH-) for the Anthocyanidins Studied in the Gas-Phase and in Aqueous Solution Simulated with PCM (in Italics) at the B3LYP/ 6-31++G(d,p) Level Pg Cy Dp Pn Pt Mv

AH3+ f AH32+ ·

AH2 f AH2+ ·

AH- f AH ·

246.54/ 242.65/ 240.94/ 239.26/ 235.74/ 232.58/

154.22/ 152.31/ 150.21/ 152.64/ 152.90/ 151.63/

74.29/ 73.60/ 73.28/ 62.75/ 72.07/ 62.71/

136.07 134.45 133.31 137.15 133.57 135.99

117.63 114.94 112.46 116.81 114.72 114.14

106.77 105.26 103.65 108.37 106.12 105.59

A2- f A- · -13.23/ 90.05 -11.9/ 89.91 5.62/ 96.17 -9.56/ 89.08 -12.81/ 88.83 -13.0/ 87.61

values are higher than 5 kcal mol-1 in the gas phase) pointing to the fact that the HAT mechanism is allowed for them. Considering the values obtained in aqueous solution from PCM calculations (HAA in aqueous solution for DPPH · is -83.80 kcal mol-1), we can conclude (Table 3) that cations remain unreactive for HAT except Dp, which shows a slightly negative enthalpy balance, whereas neutral and anionic forms are still good radical scavengers, but displaying smaller ∆H values than in gas phase. The HAA values obtained for ROS in gas phase (and in solution) (-97.92 (-111.90) kcal mol-1 and -80.39 (-92.87) kcal mol-1, respectively, for MeO · and MeO2 · ) indicate that the HAT mechanism is permitted to scavenge those radicals at any pH in aqueous solution, as their absolute value exceeds the lowest OH-BDEs of any of the three acid/base forms of anthocyanidins (Table 3). ET-PT. The feasibility of this mechanism depends on its first step (ET from anthocyanidin to the scavenged radical), as the second step (the corresponding pt, see Scheme 1) is always accompanied by negative enthalpies (Table S1, Supporting Information). To evaluate the ET ability of anthocyanidins, we have calculated the IPs (Table 4) for the most stable tautomer/ conformer of each acid/base form of every anthocyanidin.26 It can be inferred from the IP values of the series that, in general, the substitution of a hydroxyl group by a methoxyl one decreases the IP value. As expected, cations have the highest IP values, followed by neutral forms and anions. It is also noticeable that

9710

J. Phys. Chem. B, Vol. 114, No. 29, 2010

Este´vez et al.

TABLE 5: ∆H2 and ∆H3 Values (in kcal mol-1) Obtained at the B3LYP/6-31++G(d,p) Level in Aqueous Solution Modeled with PCM cation Pg Cy Dp Pn Pt Mv

neutral

anion

dianion

DPPH ·

MeO ·

MeO2 ·

DPPH ·

MeO ·

MeO2 ·

DPPH ·

MeO ·

MeO2 ·

DPPH ·

MeO ·

MeO2 ·

28.04 26.42 25.28 29.12 25.54 27.96

32.39 30.77 29.63 33.47 29.89 32.31

39.28 37.66 36.52 40.36 36.78 39.2

9.6 6.91 4.43 8.78 6.69 6.11

13.95 11.26 8.78 13.13 11.04 10.46

20.84 18.15 15.67 20.02 17.93 17.35

-1.26 -2.77 -4.38 0.34 -1.91 -2.44

3.09 1.58 -0.03 4.69 2.44 1.91

9.98 8.47 6.86 11.58 9.33 8.8

-17.98 -18.12 -11.86 -18.95 -19.2 -20.42

-13.63 -13.77 -7.51 -14.6 -14.85 -16.07

-6.74 -6.88 -0.62 -7.71 -7.96 -9.18

the influence of the solvent on IP values, and consequently on ∆H2, is rather strong. So, it is observed that the IP of cations decrease more (exceed 100 kcal mol-1) than that of neutrals (below 42 kcal mol-1) because the larger the charge of the species, the larger the stabilizing effect of the solvent. In the same vein, PCM increases the IP values of anions (between 30 and 40 kcal mol-1) and dianions (between 80 and 110 kcal mol-1) because of the uncharged species (or less charged for dianions) resulting now from electron loss. IPs computed in aqueous solution for the three forms of each antioxidant are on the order of those obtained for phenolic antioxidants such as catechin and quercetin17a or for the model of vitamin E17b and their anions, respectively. It is noticeable that IPs of cations (in aqueous solution) are on the same order (135.09 kcal mol-1) of those of uncharged catechin and quercetin, IPs of neutrals are between the values of neutral and anionic catechin, and IPs of anthocyanidins anions are of the same order as that of catechin anion. Calculations in the gas phase clearly show that the second mechanism is endothermic for cations and neutral species. Thus, summation of the EA of DPPH · in the gas phase (-78.10 kcal mol-1) with the IPs of the cationic and neutral forms (higher than 232.58 and 150.21 kcal/mol, respectively) indicates that ET from cationic or neutral forms of the antioxidant to DPPH · radical should be thermodynamically forbidden in the gas phase. The situation does not change when we take into account solvation effects (Table 5), as the absolute value for the EA of DPPH · (-108.03 kcal mol-1) is still lower than the IPs of those forms of anthocyanidins. A different trend is shown by the anions, whose IPs (Table 4) are lower than the absolute value of the EA of DPPH · radical in both phases, indicating that the first step of ET-PT mechanism is permitted. Feasibility of ET from dianions will be commented in the SPLET section, as it is only involved in the second step of that mechanism (Scheme 1). The combination of the IPs of the anthocyanidins with the EAs of the MeO · (-33.88 and -103.68 kcal mol-1, respectively, in the gas phase and in aqueous solution) and MeO2 · (-25.71 and -96.79 kcal mol-1, respectively, in the gas phase and in aqueous solution) radicals is always endothermic, indicating that the first step of ET-PT is always forbidden thermodynamically, even for anions (Table 5). Overall, ET should be only viable for anions in both phases with DPPH · , whereas it should be never allowed for any of the remaining radicals here studied. SPLET. Litwinienko and Ingold9 proposed that phenolic anions play a crucial role in scavenging DPPH · radicals through the SPLET mechanism (eq 3) because of their extremely high electron-donating ability. Anthocyanidins, like other polyphenols, can be partially ionized losing a proton and transforming into the corresponding conjugated base (Scheme 1) in solvents able to accept protons. Assuming the first step of this mechanism is always permitted to some extent (the lower the PDE, the larger the extent; Table

S2, Supporting Information), we estimate the global feasibility of this mechanism by comparing IPs of the conjugated base (Table 4) with the EA of the radical. Thus, SPLET should only be permitted for neutral (and anion) through the corresponding anion (and dianion) where ET to DPPH · is thermodynamically feasible (IPs for anions and dianions plus EA of the radicals is exoenergetic as shown in Table 5). For ROS, the SPLET mechanism is forbidden for neutral forms, but viable for anions (Table 5). Sequential Proton Loss Hydrogen Atom Transfer (SPLHAT). In spite of the controversy with regard to the mechanism through which anthocyanidins scavenge DPPH · , we have found, according to our calculations, that neither HAT nor ET nor even SPLET are permitted for cations. Nonetheless, experimental assays carried out in a methanolic solution (acidic condition), where it should be expected that the dominant presence of the cationic form of anthocyanidins, give rise to neutralized DPPHH. Looking for an alternative mechanism that could explain the antiradical activity of anthocyanidins cations we notice that, to the best of our knowledge, no work considers the possibility of the HAT mechanism after deprotonation. Scheme 1 shows that, after deprotonation, ET is not the only process that the conjugated base may experience. While subsequent PL (double proton loss from the majoritary form at a given pH) can be considered negligible, there is no reason to discard HAT as a second step as in the case of SPLET. This mechanism, which could be named SPL-HAT, is (according to data presented in Tables 2 and 3) always viable, so the three acid/base forms of anthocyanidins would contribute in scavenging DPPH · and the other radicals. Moreover, SPL-HAT (Table 3) is always energetically favored over SPLET (Table 5). Thus, according to our data, SPL-HAT is the only allowed mechanism for DPPH · scavenging by cationic anthocyanidins, and it may be competitive (or even preferred) to HAT for neutrals and anions depending on the gift the solvent has to accept the proton lost in its first step. This mechanism could also be used to explain why there is not a good direct relation between lower O-H BDE values and higher antiradical activity, for instance, in the DPPH · assay carried out by Borkowski et al.23f Thus, according to SPL-HAT, one should consider first which anthocyanidin cation displays the lowest O-H PDE (Supporting Information) to give rise to the neutral species and therefore the lowest O-H BDE value of each of them (Table 2). Taking into account the SPL-HAT mechanism, our calculated results and experimental are in good agreement. Effects of the Medium. Looking for a different behavior in the redox properties of anthocyanidins and, hence, in the preference for one or another mechanism when reactions are carried out in different solvents, we have performed PCMDFT(B3LYP/6-311++G(2d,2p)//B3LYP/6-31++G(d,p)) calculations on the three species acid/base of cyanidin and on the corresponding radicals to obtain BDEs and IPs. We have

-0.79 137.78 -10.60 74.84 -18.78 30.52 -21.01 120.58 -30.81 65.87 -38.99 21.55 -3.74 43.06 -11.12 21.54 -17.42 9.95 -23.27 25.04 -30.65 14.89 -36.95 3.30 5.70 30.59 -1.68 9.07 -7.98 -2.52

6.05 96.44 -3.76 41.74 -11.93 -2.58

MeO2 · MeO · DPPH · MeO2 · MeO ·

∆Hi(PCM-B3LYP/ 6-311++G(2d,2p)) ∆Hi(PCM B3LYP/ 6-311++G(2d,2p))

DPPH ·

cyclohexane methanol

J. Phys. Chem. B, Vol. 114, No. 29, 2010 9711 selected methanol and cyclohexane to model experimental conditions, as experimental assays are usually carried out in methanol or in bulk oil. In the same way, HAAs and EAs were computed for radical species. First of all, no significant differences are found in the thermodynamical parameters values when they were obtained by using larger basis sets (Table 6). Comparing O-H BDE values obtained in the three solvents, we can observe that the smallest one is always obtained in cyclohexane and the highest one in water. Contrary, IP values in water decreased with regard to those in methanol and cyclohexane. For the radical species, ROS and DPPH · , the pattern is quite different, so the smallest values for both EAs and HAA are displayed in cyclohexane and the highest in water. We can also observe how protic solvents favor ET and how HAT is not so influenced by the medium. Overall, although there is slight difference in the redox properties values depending on the medium, they are not enough to cause substantial changes in the preferences for one or another mechanism. Thus, cations do not react with DPPH · in any of the three solvents tested by HAT, ET, or SPLET but they do by SPL-HAT.

-3.87 47.18 -10.74 24.43 -17.27 14.06 -23.36 32.99 -30.23 12.83 -36.76 2.52 0.60 217.19 -11.07 127.08 -21.12 47.75 6.04 165.10 -5.63 74.98 -15.68 -4.35 0.85 216.68 -11.29 126.34 -19.72 47.74

-18.03 205.15 -29.70 115.03 -39.75 35.70

5.38 28.35 -1.49 8.19 -8.02 -2.12

MeO2 · MeO · DPPH · MeO2 · MeO · DPPH · MeO2 ·

∆Hi(PCM-B3LYP/ 6-311++G(2d,2p)) ∆Hi(B3LYP/ 6-311++G(2d,2p))

water

Conclusions

-17.39 205.46 -29.53 115.13 -37.96 36.52 5.94 164.34 -6.21 74.00 -14.63 -4.60 anion

neutral

cation

HAT ET HAT ET HAT ET

MeO · DPPH · cyanidin

∆Hi (B3LYP/ 6-31++G(d,p))

gas

TABLE 6: ∆H1, ∆H2, and ∆H3 Values (in kcal mol-1) for the Radical (MeO · , MeO2 · , and DPPH · ) Scavenging by Cyanidin: Cation, Neutral, and Anion in Aqueous Solution, Methanol, and Cyclohexane Modeled with PCM

Radical-Scavenging Mechanisms of Anthocyanidins

First of all, in spite of the “spurious” error that plagues DFT by construction, we observe that DFT methods (B3LYP or M05-2X) provide accurate BDEs and HAA values in this series of compounds. Thus, we have obtained globally equivalent results (for mechanisms) for simpler models using both DFT and CCSD levels. B3LYP/6-31++G(d,p) calculations establish that, in most hydrogen-transfer reactions of anthocyanidins, the HAT mechanism (eq 1) is more favored than SPLET and ET-PT for radical scavenging by anthocyanidins. The HAA value of the radicals studied here exceeds the OH-BDEs of anthocyanidins in any of the three acid/basic species, with the exception of scavenging DPPH · radical by cations. The latter case could be explained by sequential combination of PL and HAT. Thus, SPL-HAT seems to be the unique thermodynamically allowed mechanism for DPPH · scavenging by cations. It may be competitive with (or even preferred to) one-step HAT (eq 1) for neutrals and anions depending on the gift the solvent has to accept of the proton lost in its first step. Finally, no different reactivity was found when different solvents were used in the calculation of BDEs, IPs, HAAs, and EAs. As it is known that the HAT mechanism is less affected by the solvent than the ET one. We can confirm the experimental observation looking at the BDE values obtained in different solvents, which differ by less than 5 kcal mol-1, while differences among IP values are higher than 15 kcal mol-1. Acknowledgment. Authors thank “Centro de Supercomputacio´n de Galicia” (CESGA) for free access to its computational facilities and the Galician Government for funding this research through project INCITE09E1R3141091ES. Supporting Information Available: O-H proton dissociation energies for the most stable conformer and tautomer of each form of the anthocyanidins and species formed after electron transfer, as well as diverse energy parameters concerning oxidative pathways for cyanidin. This material is available free of charge via the Internet at http://pubs.acs.org. References and Notes (1) (a) Hertog, M. G. L.; Feskens, E. J. M.; Hollman, P. C. H.; Katan, M. B.; Kromhout, D. Lancet 1993, 342, 1007. (b) Renaud, S.; de Lorgeril,

9712

J. Phys. Chem. B, Vol. 114, No. 29, 2010

M. Lancet 1992, 339, 1523. (c) Leake, D. S. In Phytochemistry of Fruit and Vegetables; Toma´s-Barbera´n, F. A., Robins, R. J., Eds.; Clarendon Press: Oxford, 1997; p 287. (2) (a) Vinson, J. A.; Dabbagh, Y. A.; Serry, M. M.; Jang, J. J. Agric. Food Chem. 1995, 43, 2800. (b) Vinson, J. A.; Jang, J.; Dabbagh, Y. A.; Serry, M. M.; Cai, S. J. Agric. Food Chem. 1995, 43, 2798. (c) da Silva, E. L.; Piskula, M. K.; Yamamoto, N.; Moon, J.-H.; Terao, J. FEBS Lett. 1998, 430, 405. (d) de Whalley, C. V.; Rankin, S. M.; Hoult, J. R. S.; Jessup, W. D.; Leake, S. Biochem. Pharmacol. 1990, 39, 1743. (e) Miura, S.; Watanabe, J.; Sano, M.; Tomita, T.; Osawa, T.; Hara, Y.; Tomita, I. Biol. Pharm. Bull. 1995, 18, 1. (3) (a) The FlaVonoids, AdVances in Research since 1986; Harborne, J. B., Ed.; Chapman and Hall: London, 1994; (b) Haslam, E. Practical Polyphenolics; Cambridge University Press: Cambridge, 1998; (c) Bohm, B. A. Introduction to FlaVonoids; Harwood Academic Publishers: Amsterdam, 1998. (4) (a) Lapidot, T.; Harel, S.; Akiri, B.; Granit, R.; Kanner, J. J. Agric. Food Chem. 1999, 47, 67. (b) Satue´-Gracia, M. T.; Heinonen, M.; Frankel, E. N. J. Agric. Food Chem. 1997, 45, 3362. (c) Tsuda, T.; Watanabe, M.; Ohshima, K.; Norinobu, S.; Choi, S.-W.; Kawakishi, S.; Osawa, T. J. Agric. Food Chem. 1994, 42, 2407. (d) Tamura, H.; Yamagami, A. J. Agric. Food Chem. 1994, 42, 1612. (e) Tsuda, T.; Ohshima, K.; Kawakishi, S.; Osawa, T. J. Agric. Food Chem. 1994, 42, 248. (5) Yamasaki, H.; Uefuji, H.; Sakihama, Y. Arch. Biochem. Biophys. 1996, 332, 183. (6) (a) van Acker, S. A. B. E.; Koymans, L. M. H.; Bast, A. Free Radical Biol. Med. 1993, 15, 311. (b) Migliavacca, E.; Carrupt, P. A.; Testa, B. HelV. Chim. Acta 1997, 80, 1613. (c) Zhang, H. Y. J. Am. Oil Chem. Soc. 1998, 75, 1705. (d) Zhang, H. Y.; Sun, Y. M.; Zhang, G. Q.; Chen, D. Z. Quant. Struct.-Act. Relat. 2000, 19, 375. (7) (a) Burton, G. W.; Le Page, Y.; Gabe, E. J.; Ingold, K. U. J. Am. Chem. Soc. 1980, 102, 7791. (b) Burton, G. W.; Doba, T.; Gabe, E. J.; Hughes, L.; Lee, F. L.; Prasad, L.; Ingold, K. U. J. Am. Chem. Soc. 1985, 107, 7053. (c) de Heer, M. I.; Mulder, P.; Korth, H.-G.; Ingold, K. U.; Lusztyk, J.-N. J. Am. Chem. Soc. 2000, 122, 2355. (d) Foti, M. C.; Johnson, E. R.; Vinqvist, M. R.; Wright, J. S.; Barclay, L. R. C.; Ingold, K. U. J. Org. Chem. 2002, 67, 5190. (8) (a) Jovanovic, S. V.; Steenken, S.; Tosic, M.; Marjanovic, B.; Simic, M. G. J. Am. Chem. Soc. 1994, 116, 4846. (b) Jovanovic, S. V.; Steenken, S.; Hara, Y.; Simic, M. G. J. Chem. Soc., Perkin Trans. 1996, 2497. (c) Nakanishi, I.; Miyazaki, K.; Shimada, T.; Ohkubo, K.; Urano, S.; Ikota, N.; Ozawa, T.; Fukuzumi, S.; Fukuhara, K. J. Phys. Chem. A 2002, 106, 11123. (9) (a) Litwinienko, G.; Ingold, K. U. J. Org. Chem. 2003, 68, 3433. (b) Litwinienko, G.; Ingold, K. U. J. Org. Chem. 2004, 69, 5888. (10) Zhang, H.-Y.; Sun, Y.-M.; Wang, X.-L. Chem.sEur. J. 2003, 9, 502. (11) Brinck, T.; Haeberlein, M.; Jonsson, M. J. Am. Chem. Soc. 1997, 119, 4239. (12) Wu, Y.-D.; Lai, D. K. W. J. Org. Chem. 1996, 61, 7904. (13) Wright, J. S.; Carpenter, D. J.; McKay, D. J.; Ingold, K. U. J. Am. Chem. Soc. 1997, 119, 4245. (14) Himo, F.; Eriksson, L. A.; Blomberg, M. R. A.; Siegbahn, P. E. M. Int. J. Quantum Chem. 2000, 76, 714. (15) Chandra, A. K.; Uchimaru, T. Int. J. Mol. Sci. 2002, 3, 407. (16) Wright, J. S.; Johnson, E. R.; DiLabio, G. A. J. Am. Chem. Soc. 2001, 123, 1173. (17) (a) Wang, L.-F.; Zhang, H.-Y. Bioorg. Chem. 2005, 33, 108. (b) Zhang, H.-Y.; Ji, H.-F. New J. Chem. 2006, 30, 503. (18) Martı´nez, A. J. Phys. Chem. B 2009, 113, 4915.

Este´vez et al. (19) (a) Johansson, A. J.; Blomberg, M. R. A.; Siegbahn, P. E. M. J. Chem. Phys. 2008, 129, 154301, and references therein. (b) RienstraKiracofe, J. C.; Tschumper, G. S.; Schaefer, H. F., III; Nandi, S.; Ellison, G. B. Chem. ReV. 2002, 102, 231. (20) (a) Zhao, Y.; Schultz, N. E.; Truhlar, D. G. J. Chem. Phys. 2005, 123, 161103. (b) Zhao, Y.; Schultz, N. E.; Truhlar, D. G. J. Chem. Theory Comput. 2006, 2, 364. (21) Tomasi, J.; Mennuci, B.; Cammi, R. Chem. ReV. 2005, 105, 2999. ´ lvarez-Diduk, R.; Ramı´rez-Silva, M. T.; Alarco´n(22) Galano, A.; A ´ Angeles, G.; Rojas-Herna´ndez, A. Chem. Phys. 2009, 363, 13. (23) (a) Valgimigli, L.; Banks, J. T.; Ingold, K. U.; Lusztyk, J. J. Am. Chem. Soc. 1995, 117, 9966. (b) Dangles, O.; Fargeix, G.; Dufour, C. J. Chem. Soc., Perkin Trans. 2 1999, 1387. (c) Dangles, O.; Fargeix, G.; Dufour, C. J. Chem. Soc., Perkin Trans. 2 2000, 1653. (d) Fukumoto, L. R.; Mazza, G. J. Agric. Food Chem. 2000, 48, 3597. (e) Ka¨hko¨nen, M. P.; Heinonen, M. J. Agric. Food Chem. 2003, 51, 628. (f) Borkowski, T.; Szymusiak, H.; Gliszczyn´ska-S´wiglo, A.; Rietjens, I. M. C. M.; Tyrakowska, B. J. Agric. Food Chem. 2005, 53, 5526. (24) Carreras, A.; Esparbe, I.; Brillas, E.; Rius, J.; Torres, J. L.; Julia, L. J. Org. Chem. 2009, 74, 2368. The studies to measure the antioxidant activity of polyphenols using DPPH · as a persistent free-radical sensor do not provide clear information of the mechanism involved, and both mechanisms, HAT and ET, have been proposed as alternative reduction paths. (25) (a) Becke, A. D. J. Chem. Phys. 1993, 98, 5648. (b) Lee, C.; Yang, G.; Parr, R. G. Phys. ReV. 1988, 37, 785. (26) (a) Este´vez, L.; Mosquera, R. A. J. Phys. Chem. A 2007, 111, 11100. (b) Este´vez, L.; Mosquera, R. A. J. Phys. Chem. A 2008, 112, 10614. (c) Este´vez, L.; Mosquera, R. A. J. Phys. Chem. A 2009, 113, 9908. (27) Williams, D. E. J. Am. Chem. Soc. 1967, 89, 4280. (28) IPs, BDEs, HAAs, or EAs values obtained from calculations in solution modeled with PCM are given in terms of ∆∆Gsol instead of ∆H as in the gas phase. We have noticed that there are not significant changes in relative values or in global results of the mechanisms when thermal and other corrections obtained from gas phase calculations are included. Therefore, along the paper we have called indistinctly ∆Hi in the gas phase or in solution to make the reading easier. (29) ; Frisch, M. J. G.; Trucks, W.; Schlegel, H. B.; Scuseria, G. E.; Robb, M. A.; Cheeseman, J. R.; Montgomery, J. A., Jr.; Vreven, T.; Kudin, K. N.; Burant, J. C.; Millam, J. M.; Iyengar, S. S.; Tomasi, J.; Barone, V.; Mennucci, B.; Cossi, M.; Scalmani, G.; Rega, N.; Petersson, G. A.; Nakatsuji, H.; Hada, M.; Ehara, M.; Toyota, K.; Fukuda, R.; Hasegawa, J.; Ishida, M.; Nakajima, T.; Honda, Y.; Kitao, O.; Nakai, H.; Klene, M.; Li, X.; Knox, J. E.; Hratchian, H. P.; Cross, J. B.; Bakken, V.; Adamo, C.; Jaramillo, J.; Gomperts, R.; Stratmann, R. E.; Yazyev, O.; Austin, A. J.; Cammi, R.; Pomelli, C.; Ochterski, J. W.; Ayala, P. Y.; Morokuma, K.; Voth, G. A.; Salvador, P.; Dannenberg, J. J.; Zakrzewski, V. G.; Dapprich, S.; Daniels, A. D.; Strain, M. C.; Farkas, O.; Malick, D. K.; Rabuck, A. D.; Raghavachari, K.; Foresman, J. B.; Ortiz, J. V.; Cui, Q.; Baboul, A. G.; Clifford, S.; Cioslowski, J.; Stefanov, B. B.; Liu, G.; Liashenko, A.; Piskorz, P.; Komaromi, I.; Martin, R. L.; Fox, D. J.; Keith, T.; Al-Laham, M. A.; Peng, C. Y.; Nanayakkara, A.; Challacombe, M.; Gill, P. M. W.; Johnson, B.; Chen, W.; Wong, M. W.; Gonzalez, C. Pople, J. A. Gaussian 03, revision E.01,; Gaussian, Inc.: Wallingford, CT, 2004. (30) Schultz, N.; Zhao, Y.; Truhlar, D. G. J. Phys. Chem. 2005, 109, 11127. (31) Rice-Evans, C. A.; Miller, N. J.; Paganga, G. Free Radical Biol. Med. 1996, 20, 933. (b) Bors, W.; Heller, W.; Michel, C.; Saran, M. Methods Enzymol. 1990, 21, 703.

JP1041266