DFT study on the quenching mechanism of singlet oxygen by lycopene

Abstract

Lycopene is a kind of natural food ingredient that is one of the strongest antioxidants in nature at present. However, the mechanism of scavenging singlet oxygen is still unclear which has restricted the extensive application of lycopene in many fields. Density functional theory has been applied to optimize the configurations of the ground and excited states of lycopene and oxygen respectively. The quenching mechanism of singlet oxygen by lycopene is analyzed based on the calculations at the B3LYP/6-311+(d,p) level. The excited state oxygen in the singlet turns into the ground state in the triplet and gives out energy, while at the same time lycopene in the ground state absorbs the energy, changing into excited lycopene. The quenching process involves exchange of two electrons of different spin coming from lycopene and singlet oxygen. The quenching is spontaneous according to the calculated energy and Gibbs free energy change. Regarding energy level and the construction of frontier orbitals, electron exchange occurs simultaneously with energy transfer and change of spin. In addition, it also can be concluded that oxygen and lycopene can pass electric charge to each other in the meantime, however, it is dominated by one or the other in different stages. Overall, the close energy levels and good symmetry matching of the reactants’ frontier orbitals, lack of electron spin-flip in the quenching system, and the negative Gibbs free energy difference, all make the quenching process very easy.

---

1. Introduction

Since their first discovery by Moses Gomberg in 1900, free radicals have gained more and more attention.¹ Free radicals can be atoms, molecules or ions that possess unpaired valence electrons. They are made up of reactive oxygen species (ROS) and reactive nitrogen species (RNS).² ROS and RNS are a double-edged sword, of which ROS is the most important class. With more opportunities for humans to be exposed to complex physicochemical conditions and pathophysiological states, the disadvantages of ROS have outweighed the benefits. ROS comprises singlet oxygen (SO, ¹O₂) and a range of oxidising free radicals.³ SO plays a key role in bio-oxidation processes. It can perturb or even lead to the destruction of cells or organisms, resulting in some adverse effects on well-functioning systems.⁴ SO is generated in liposomes by photosensitization with two photosensitizers; the SO at the membrane surface is generated by water-soluble Rose bengal (RB), and that in the inner region of membranes by lipid soluble pyrenedodecanoic acid (PDA).⁵ SO is the high energy form of oxygen which is in its first excited singlet state.⁶ Once produced, it will react with a wide range of molecules.⁴ Therefore, SO and its quenching remain at the forefront of scientific research.

Fortunately, nature has bestowed upon us a wealth of protective antioxidants, such as superoxide dismutase, catalase, glutathione peroxidases and reductase, vitamin E and carotenoids.^7,8 Among them, carotenoids are a common antioxidant and possess higher capacity to eliminate SO. Many researchers have shown that the singlet-oxygen-quenching ability of lycopene is ten times greater than that of vitamin E.^9,10 Lycopene (lyco), a kind of red carotenoid widely distributed in fruits and vegetables, is the most effective SO quencher in comparison with other carotenoids.^11–13 Moreover, it is its strong antioxidant capacity that induces an inhibitory effect on low-density lipoprotein oxidation and HMG-CoA reductase activity, impacts cholesterol biosynthesis, and lowers blood triglyceride and cholesterol levels.^14,15 Lyco also has anti-amnesic effects and is anti-cancerous.¹⁶ In addition, it is the most reactive and efficient quencher of SO in vitro.^17,18

Lyco, C₄₀H₅₆, has a long straight-chain structure with 11 consecutive conjugated double bonds and exhibits a certain symmetry around the central double bond.¹⁹ The main isomer forms in vivo are all-trans, 5-cis, 9-cis, 13-cis and 15-cis.²⁰ All-trans lyco makes up the vast majority, up to 80–97%, and is the predominant isomer.^21,22 Exposure to light leads to no conspicuous change to total and all-trans lyco, but a significant decrease of cis isomers has been observed.²³ Since the quenching of SO by lyco proceeds in vivo, the processing, digestion and absorption of it are particularly important. Heated food contains more lyco, which includes all-trans and cis isomers, for heating can destroy the food matrix and release more lyco from it.^23,24 At the same time, heating could promote the isomerization of lyco, however, cis forms are unstable and quickly degrade, with only 1.7% to 10.1% left.^23,25,26 Hence there is only a small increase in cis-isomers after processing.²⁷ Food which is ingested from the digestive tract then is absorbed through the small intestine, and ends up within storage in the liver.²⁴ Compared to processed foodstuffs, in serum the isomerization ratio of lyco from all-trans to cis is considerably large.²⁸ In serum, the proportion of all-trans lyco drops to 41% of the total lyco, 5-cis achieves 28%, 13-cis and 15-cis together reach 12%, and other isomers of lyco are 16% on average.²⁸ On the whole, cis lyco is numerically superior, while all-trans lyco still is the most common isomer in vivo in the case of a single isomer. Moreover, the cis isomers display a lower capacity to quench SO compared to that of the all-trans isomer because of the much lower resonance effect of the cis isomer.^29,30 Hence, the paper mainly studies all-trans lyco. All of the lyco in this paper refers to all-trans lyco and its structure is shown in Fig. 1. The scavenger mechanism is just as the formula SO + lyco → TO + *lyco (the asterisk * represents the excited state) depicts. The excited state oxygen in the singlet turns into the ground state in the triplet and gives out energy, while at the same time lyco in the ground state absorbs the energy, changing into excited lyco. Excited lyco can dissipate its energy in the form of heat to the surrounding environment, and return to the ground state.²¹ This physical quenching process makes up more than 99.5% of the overall quenching of SO by lyco, so the paper mainly studies the physical quenching.^31,32

Fig. 1 The molecular structure of all-trans lycopene.

Until now, there has been much research on the quenching function of lyco using experimental methodology all over the world. Nevertheless, there are few studies on the mechanism of lyco quenching SO at the molecular level by theoretical calculation. It is only mentioned that the energy transfer process mediates the exchange of electrons.³³ However, a specific quenching mechanism has not been clear and distinct up to now.^10,34 Studies concerning the quenching mechanism by lyco are very limited and thus should be encouraged so as to improve the understanding of the quenching process. Study requirements of the quenching mechanism are harsh as the related reactants are special and uncontrollable and the quenching process is considerably fast. Therefore, computational methods in quantum chemistry to analyze the mechanism of quenching SO by lyco are the best choice and are more convenient and direct. This paper will start with analysis of the possibility of quenching from an ultraviolet-visible spectrum and the Gibbs free energy for the macro aspects. Then we will focus largely on expanding the analysis of the quenching mechanism to the active sites of the scavenging SO, the frontier orbitals and the energy transfer process. This study will definitely ensure that the quenching mechanism of SO by lyco is understood more deeply. Also the performed work would certainly provide the theoretical basis and support for the further application of lyco in food, chemistry, the pharmaceutical industry, and medical and other fields. At the same time, it can guide the direction for the synthesis and development of other new drugs.

2. Methodology

The computation was performed with the density functional theory (DFT) and QCISD approaches implemented in the Gaussian 09 programs.³⁵ Using the QCISD approach, SO and triplet oxygen (TO) were added to the scan type to calculate the electrical potential curves. Geometry optimizations and frequency calculations have been carried out using the B3LYP functional in conjunction with the 6-311+G(d,p) basis set.³⁶ Fukuzawa has investigated singlet oxygen quenching in a lipid membrane environment using either water-soluble or lipid-soluble singlet oxygen, which showed little difference among the carotenoids.³⁷ Furthermore, lyco can be randomly dispersed in the lipid membrane and could increase the fluidity of the membrane. Thus the optimizations in this paper are all under vacuum conditions. The B3LYP density functional was applied to all of the substances that include the ground and excited states of lyco, SO, TO, the reactants system and the products system in this study. On this basis, the structures of the prior state where the ground state of lyco and SO conduct energy transmission at different locations are also optimized. The corresponding structures after the energy delivery, likewise, are optimized. In addition, we also attempt to optimize the state of the system of the ground state of lyco and SO and the system of the excited state of lyco and TO at different distances to elucidate the energy exchange process. Furthermore, the ultraviolet absorption spectrum of the lyco was calculated with the time-dependent density functional theory (TDDFT) method.

3. Results and discussion

3.1 Proving that the quenching can occur from the macro aspect

UV-VIS and potential energy curve. For the physical quenching of SO by lyco, singlet oxygen releases energy, turning into triplet oxygen, and at the same time, the energy is absorbed by the ground state of lyco which changes into the excited state. Here, the spin multiplicity of the lyco’s ground state is singlet, and the excited state’s is triplet. In other words, the spin states of the two reactants are changed when energy transfer occurs. The transition is a process in which leaping changes of the quantum mechanics system take place. No matter whether in terms of physics or chemistry, it must follow the transition rule. The transition rule formally constrains the possible transitions of a system from one quantum state to another, such as the transition that the initial and final states of the spin states make, so a difference is forbidden just as the transition of singlet oxygen and lyco is. Actually, the forbidden transition can occur, but the probability of the transition is very low.

It is known that the band which is forbidden by the transition rule is generally weak. When the ε value of the absorption band is between 10⁴ and 10⁶ or the oscillator strength value is more than 0.1, the transition is completely allowed. This paper calculates the absorption spectra of lycopene and triplet oxygen with the TDDFT method at the B3LYP 6-311+G(d,p) level. The results for lycopene show that the maximum absorption wavelength is 594 nm in the gas phase (Hernandez-Marin mentioned in his article that the maximum absorption wavelength is 590 nm in the gas phase),³⁶ the oscillator strength is 4.4734 and the ε_max value of it is 181 177.84. Therefore, although the spin multiplicity changes during the transition of lycopene, the transition of it can still happen. For triplet oxygen, the results are that the maximum absorption wavelength is 142 nm, the oscillator strength is 0.2047 and the ε_max value of it is 8351.36 which is close to 10⁴. As a result, the transition between the SO and TO can also occur. Additionally, for singlet and triplet oxygen, we obtain the potential energy curve as well with the QCISD method at the 6-311+G(d) level. The potential energy curve shown in Fig. 2. It is evident that the curves of singlet oxygen and triplet oxygen have a cross which shows once more that the transition between the SO and TO can occur.

Fig. 2 The potential energy curves of SO and TO.

The analysis of the Gibbs free energy and energy. Gibbs free energy is a very important thermodynamics function in chemical thermodynamics. The Gibbs free energy difference between the final and initial states could be the criterion for the direction and limitation of the quenching process. The energy between the initial and final states of the substances is also illustrated. Table 1 lists the Gibbs free energy at room temperature (298.15 K) and the energy based on the structure optimizations and frequency calculations with the B3LYP/6-311+G(d,p) method.

Table 1 The free energy and energy of the substances (Hartree)

Compound	SO	Lyco	TO	*Lyco
G^298.15 K	−150.32419	−1557.53821	−150.38666	−1557.52206
E	−150.30896	−1558.28560	−150.37042	−1558.26243

---

The calculated results illustrate that the Gibbs free energy difference at 298.15 K is −29.07 kcal mol⁻¹, which is negative, so the quenching process can be spontaneous at 298.15 K in the macro.

In addition, the energy difference between the reactants and the products is a positive value, telling us that the reaction is exothermic. Considering that the quenching is just an energy transfer process, the quenching requires no additional energy and it can occur spontaneously and smoothly. It is obvious that there exists an energy difference between the SO and TO, and the ground and excited states of lyco. The energy difference between the SO and TO, 38.57 kcal mol⁻¹, is larger than the energy difference between the ground and excited lyco, 14.54 kcal mol⁻¹. That is to say, the energy which SO releases is greater than the energy that the ground state of lyco absorbs. Consequently, the quenching can happen.

3.2 Speculating the mechanism at the micro level

The determination of the electronic exchange of energy transfer. The physical quenching of singlet oxygen by lycopene is essentially an excited state energy transfer process in which the energy is transferred from the donor to the acceptor. The mechanism includes a radiation mechanism and a non-radiation mechanism. The process could be represented by the expression D* + A → A* + D. When D* and A are far apart, and the distance between them is more than λ, the light that is emitted by D* can be absorbed directly by A. Here is the radiation mechanism. This mechanism requires that there must be a large spectral overlap of the donor emission and the acceptor absorption spectra. From the data in the second paragraph of Section 3.1 of this article, it is clear that lycopene and singlet oxygen do not meet this requirement. If the distance between D* and A is small, the non-radiation mechanism is adopted. It is made up of a resonance energy transition mechanism and an electronic exchange of energy transfer mechanism. Among these, the resonance mechanism belongs to the long-range energy transfer process, and the distance between donor and acceptor is always from several to tens of nm. Crucially, it requires that ΔE(D* − D) = ΔE(A − A*). It is evident that ΔE(SO − TO) ≠ ΔE(lyco − lyco*). Hence, the mechanism of physical quenching of singlet oxygen by lycopene is not the resonance energy transition mechanism but the electronic exchange of energy transfer mechanism.^21,38 For the latter, it is a short-range energy transfer process and it needs the distance between them to be less than 1 nm.

The sites and process of the quenching. Since the hydrophilicity of lyco is extremely poor and it lies in the membrane, it occupies a position deep within the hydrophobic core of the cell membrane lipid bilayers and its hydrocarbon chains parallel to the membrane surface.³⁹ As the singlet oxygen is generated at the membrane surface or in the membrane inner region, SO could approach lyco by diffusion and then energy transfer between oxygen and lycopene happens. Namely, they have the following process: SO + lyco → SO⋯lyco → TO⋯*lyco → TO + *lyco. The SO⋯lyco represents the state before energy transfer at a shorter range compared to the state of SO + lyco and the TO⋯*lyco correspondingly refers to the state after energy transfer.

In this paper we attempt to optimize different sites where SO is close to lyco to determine the relative location of SO and lycopene in SO⋯lyco system. Finally, 12 relative locations are obtained and then the corresponding TO⋯*lyco structures are also optimized. The sites are shown in Fig. 3.

Fig. 3 The sites in which singlet oxygen approaches lycopene.

3.3 Description of the mechanism from a static point of view

Frontier molecular orbital analysis. The energy transfer process begins with the reactants both in their singlet states (SO and lyco). During this process the spin antiparallel electrons exchange along with excitation energy of the singlet oxygen transferring to lycopene. This results in the products, TO and *lyco, both in their triplet states, as Fig. 4 shows. The energy level of the SO’s LUMO is −5.34 eV and the HOMO energy level of lycopene is −4.63 eV. The HOMO energy level of lycopene is higher than the LUMO energy level of singlet oxygen and their difference is 0.72 eV. This means the energy levels of the SO’s LUMO and the lyco’s HOMO are very close and can react with each other. Although the spin states of the reactants have been both changed in this process, there is no spin-flip. That is to say, if the electron transferred from lycopene to SO is spin-up, the other electron transferred from SO to lycopene is spin-down as shown in Fig. 4. It seems to be one of the important causes that the capacity of quenching singlet oxygen by lycopene is higher than that for other antioxidants.

Fig. 4 Schematic representation of the electron exchange process (based on the relative size of the molecular orbital energy level).

The conditions and methods of quenching often depend on the symmetry of the frontier molecular orbital. The effective occurrence of a reaction depends on the symmetry matching between the HOMO structure of one reactant molecule and the LUMO structure of the other reactant molecule. The electronegativity of the singlet oxygen is stronger than that of the lycopene, therefore the singlet oxygen attracts the electric charge of lycopene, namely the lycopene transfers electric charge to the singlet oxygen during the approaching process. Molecular orbital pictures of lyco and SO are shown in Fig. 5. Just as we can see, the symmetry of them matches well and there is the largest overlap in this case, as supported by Fig. 6, which is beneficial to the electron transfer. In addition, it is evident and intuitive that the SO and lyco take two different ways to exchange electron and transfer energy from Fig. 6. Either way, however, the orbital symmetry of them always matches well, as seen in the HOMO pictures of the reactants system. This indirectly illustrates the necessity of the orbital symmetry and at the same time this picture also explains the real existence of electron transfer between the two reactants.

Fig. 5 Molecular orbital picture of lycopene and singlet oxygen.

Fig. 6 The HOMO structure of the reactants system: (A) when the distance between SO and lycopene is 3.2 angstroms; (B) when the distance between oxygen and lycopene is 2.15 angstroms (the range here is the shortest distance between the oxygen molecule and the lycopene molecule).

There are many directions in which the singlet oxygen approaches the lycopene molecule such as from the top of the lycopene, below the lycopene, or at the same flat surface as the lycopene, which are shown in Fig. 3. We select a start direction in which the singlet oxygen approaches the lycopene from the same flat surface to represent the approaching process. It appeared that the singlet oxygen and lycopene were at the same surface when the distance was 3.8 angstroms and then when they were pulled close to the distance of 3.5 angstroms the singlet oxygen was located above the lycopene surface. This suggests that the singlet oxygen automatically adjusts the approaching location to make the frontier orbitals of the two molecules match.

3.4 Explaining the process of quenching from a dynamic perspective

The general process of the quenching. The lycopene and singlet oxygen (excited state) both have no single electrons, so the spin multiplicity of the reactants system is singlet. The two products, excited lycopene and triplet oxygen (ground state), both have a pair of spin parallel electrons and the spin states of the two pairs of spin parallel electrons are opposite. Therefore, the total number of spin state electrons is not changed in the system. While the Gaussian program has some limitations by which it is unable to set the spin states as mentioned above, we have no choice but to set the spin multiplicity of the products system as quintuple. That is to say, we postulate that the two products both have a pair of spin parallel electrons which are in the same direction. Any one site at which the singlet oxygen approaches lycopene has been selected and the distance between the singlet oxygen and the lycopene was pulled from 6 angstroms to 2.1 angstroms to optimize the system’s configuration and calculate the vibration to correct the energy of the system. During the process, when the configuration of the system is such that the range is 2.1 angstroms, there has arisen a slight deformation, as shown in Fig. 7. From Fig. 2, we can see that the bond length of the oxygen molecule must reach up to 2 angstroms if the spin state of oxygen changes. In fact, the bond length of the oxygen molecule is 1.245 angstroms when the distance between the singlet oxygen and the lycopene is 2.1 angstroms. Therefore, the quenching of singlet oxygen to triplet oxygen by lycopene does occur through the spin inversion of oxygen itself, as shown in Fig. 2, via the intersection of the potential energy curves of the SO and TO, but by means of electron exchange between lycopene and SO.

Fig. 7 The lycopene and oxygen system in which the distance is 2.1 angstroms.

The electric charge density and the spin density. The oxygen’s electric charge density and spin density maps of the energy transfer process are shown in Fig. 8. For the aggregate variation trend of the oxygen’s electric charge density during the energy transfer process, there is an obvious valley and peak which demonstrates the process of electron exchange. It also can be seen that the spin density of the oxygen changes from 0 to 2. This indicates that the number of single electrons on oxygen changes from zero to two, that is to say, singlet oxygen has turned into triplet oxygen. Nonetheless, the total spin quantum number in this system does not change during this process.

Fig. 8 The oxygen’s electric charge and spin density maps of energy transfer process.

For charge density, during the approaching process, a valley appears. This shows that there is some electric charge transferred from the lycopene molecule to the oxygen molecule. It can be seen that the values of charge density are very small. This phenomenon shows that there is also a small amount of charge transferred to the lycopene molecule from the oxygen molecule. However, the electric charge transferred from lycopene to oxygen is slightly larger than the charge transferred from oxygen to lycopene. Thus the externally displayed value of charge density is a result of the charge densities cancelling each other out. However, the value of spin density is not small compared to that of charge density because the transferred charges coming from SO and lycopene are in opposite spin states respectively. As the oxygen molecule moves away from the lycopene molecule gradually, a peak appears. This shows that the charge is transferred principally from oxygen to lycopene because the conjugated effect of lycopene on oxygen outweighs the oxygen’s electronic absorption effect. This indirectly accounts for the reason that the scavenging ability of all-trans lycopene is superior to that of the cis isomers as the resonance effect of the cis isomer is lower than that of the all-trans.^29,30 The redistribution of spin density occurs along with the process of charge exchange.

The energy (zero point energy corrected) and the Gibbs free energy (corrected). From Fig. 9 it can be seen that the general trend in the change of the energy and Gibbs free energy is roughly the same. The process by which two molecules at a distance of infinity are put in a system is not automatic, so the Gibbs free energy increases at the beginning. However, during the process of approaching, the energy of the system decreases gradually and a minimum value occurs when the range is 3.2 angstroms. Therefore, this process is automatic and the Gibbs free energy also has a tendency to decrease. As the distance becomes smaller, the energy increases sharply and reaches a maximum value when the distance is 2.15 angstroms and there is the same trend between the Gibbs free energy and the energy. This elucidates that much charge in this process exchanges fast and frequently as mentioned above, so the rapid change of energy can occur. Then when the two molecules are moving away, the energy is −1707.790168 a.u. as the range is 4.3 angstroms and −1707.790164 a.u. when the two products are set to an infinite distance. There is an upgrade of energy between them due to the attractive force.

Fig. 9 Energy diagram of the system of oxygen and lycopene within the energy transfer process: (A) zero-point corrected and (B) Gibbs free energy corrected.

On the whole for both the energy and Gibbs free energy there is a decrease, which shows that the energy of singlet oxygen has been transferred to lycopene successfully during the process.

4. Conclusions

Lyco is an important antioxidant used widely in the health industry. However, some antioxidant mechanisms, especially the mechanism of quenching SO, are not clear yet. This has restricted many further applications of lyco. As a result of the quenching process, the excited state oxygen in the singlet turns into the ground state in the triplet and gives out energy, while at the same time lyco in the ground state absorbs the energy, changing into excited lyco. In this work, the mechanism of lyco quenching of SO was analyzed based on DFT calculations at the B3LYP/6-311+G(d,p) level.

Regarding the thermodynamic properties, the Gibbs free energy difference is negative and the total energy is reduced during the quenching process. Clearly the quenching of SO by lyco is spontaneous. From a microcosmic perspective, we have analyzed the change of spin states with energy level and construction of the frontier orbitals which explains electron exchange. The symmetry of SO’s LUMO and lyco’s HOMO matches well, furthermore, the energy levels of them are very close. This is beneficial to the quenching process involving electron exchange and spin state change. Furthermore, in the quenching process there is no spin-flip and lyco scavenges SO via exchanging two electrons of opposite spin. In addition, emphasis is put on the charge and spin density of oxygen to depict the electron exchange process. Among them, the oxygen and lyco can pass electric charge to each other at the same time. However, on the whole, it is dominated by one or the other in different stages, so the value of charge density is very small during the process. However, the value of spin density is not small compared to that of charge density which explains once more that the quenching process is an exchange of two electrons of opposite spin. In summary, it is the close electronic energy levels, good matching conditions of symmetry of the frontier molecular orbitals, and lack of electron spin-flip in the quenching system, along with the negative Gibbs free energy difference, that jointly make the quenching perform effectively.

Theoretical study on quenching of SO by lyco is a quite appropriate method. However, it has certain limitations like the setup of spin multiplicity, and more experiments are needed to test and verify it. Only if researchers persistently figure out the mechanisms of scavenging of SO and even the other free radicals, can the true value of lyco be obtained under accurate assessment, and the applied range of lyco can be extended in the mean time.

Acknowledgements

The authors are thankful for financial support from the National Natural Science Foundation of China (Grant No. 20972042) and computational support from the High Performance Computing Center of Henan Normal University.

Notes and references

1. Tidwell, T. Thomas, Abstracts of Papers of the American Chemical Society, 2014, vol. 248.
2. W. Droge, Physiol. Rev., 2002, 82(1), 47–95.
3. F. Boehm, R. Edge and T. G. Truscott, Acta Biochim. Pol., 2012, 59(1), 27–30.
4. P. R. Ogilby, Chem. Soc. Rev., 2010, 39(8), 3181–3209.
5. K. Fukuzawa, Y. Inokami, A. Tokumura, J. Terao and A. Suzuki, Lipids, 1998, 33(8), 751–756.
6. M. J. Davies, Biochem. Biophys. Res. Commun., 2003, 305(3), 761–770.
7. T. P. A. Devasagayam, J. C. Tilak, K. K. Boloor, K. S. Sane, S. S. Ghaskadbi and R. D. Lele, J. Assoc. Physicians India, 2004, 52, 794–804.
8. W.-H. Guo, C.-Y. Tu and C.-H. Hu, J. Phys. Chem. B, 2008, 112(38), 12158–12167.
9. J. H. Weisburger, Experimental biology and medicine, Maywood, N.J., 2002, vol. 227, ch. 10, pp. 924–927.
10. H. Tatsuzawa, T. Maruyama, N. Misawa, K. Fujimori and M. Nakano, FEBS Lett., 2000, 484(3), 280–284.
11. D. Liu, J. Shi, A. C. Ibarra, Y. Kakuda and S. J. Xue, LWT--Food Sci. Technol., 2008, 41(7), 1344–1349.
12. A. El-Agamey, G. M. Lowe, D. J. McGarvey, A. Mortensen, D. M. Phillip, T. G. Truscott and A. J. Young, Arch. Biochem. Biophys., 2004, 430(1), 37–48.
13. S. Srivastava and A. K. Srivastava, J. Food Sci. Technol., 2015, 52(1), 41–53.
14. A. V. Rao, Experimental biology and medicine, Maywood, N.J., 2002, vol. 227, ch. 10, pp. 908–913.
15. D. Heber and Q.-Y. Lu, Experimental biology and medicine, Maywood, N.J., 2002, vol. 227, ch. 10, pp. 920–923.
16. R. Bala, D. Khanna, S. Mehan and S. Kalra, RSC Adv., 2015, 5(89), 72881–72892.
17. P. Di Mascio, S. Kaiser and H. Sies, Arch. Biochem. Biophys., 1989, 274(2), 532–538.
18. P. Di Mascio, H. Wefers, H. P. Do-Thi, M. V. Lafleur and H. Sies, Biochim. Biophys. Acta, 1989, 1007(2), 151–157.
19. G. Britton, FASEB J., 1995, 9(15), 1551–1558.
20. M. G. Ferruzzi, M. L. Nguyen, L. C. Sander, C. L. Rock and S. J. Schwartz, J. Chromatogr. B: Biomed. Sci. Appl., 2001, 760(2), 289–299.
21. A. Martinez, C. M. Stinco and A. J. Melendez-Martinez, J. Phys. Chem. B, 2014, 118(33), 9819–9825.
22. L. Mueller, P. Goupy, K. Froehlich, O. Dangles, C. Caris-Veyrat and V. Boehm, J. Agric. Food Chem., 2011, 59(9), 4504–4511.
23. J. Shi, Y. Dai, Y. Kakuda, G. Mittal and S. J. Xue, Food Control, 2008, 19(5), 514–520.
24. T. W. M. Boileau, A. C. Boileau and J. W. Erdman Jr, Experimentalbiology and medicine, Maywood, N.J., 2002, vol. 227, ch. 10, pp. 914–919.
25. L. Arab and S. Steck, Am. J. Clin. Nutr., 2000, 71(6), 1691S–5S , discussion 1696S–7S.
26. P. M. Bramley, Phytochemistry, 2000, 54(3), 233–236.
27. A. C. Boileau, N. R. Merchen, K. Wasson, C. A. Atkinson and J. W. Erdman Jr, J. Nutr., 1999, 129(6), 1176–1181.
28. W. Bretzel, J. Schierle, I. Biihler, N. Faccin, D. Hess, K. Steiner and W. Schiiep, Food Chem., 1997, 59, 459–465.
29. M. T. Lee and B. H. Chen, Chromatographia, 2001, 54, 613–617.
30. E. Rodrigues, L. R. B. Mariutti, R. C. Chiste and A. Z. Mercadante, Food Chem., 2012, 135(3), 2103–2111.
31. W. Stahl and H. Sies, Arch. Biochem. Biophys., 1996, 336(1), 1–9.
32. W. Stahl and H. Sies, Mol. Aspects Med., 2003, 24(6), 345–351.
33. C.-P. Hsu, Acc. Chem. Res., 2009, 42(4), 509–518.
34. M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria, M. A. Robb, J. R. Cheeseman, G. Scalmani, V. Barone, B. Mennucci, G. A. Petersson, H. Nakatsuji, M. Caricato, X. Li, H. P. Hratchian, A. F. Izmaylov, J. Bloino, G. Zheng, J. L. Sonnenberg, M. Hada, M. Ehara, K. Toyota, R. Fukuda, J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda, O. Kitao, H. Nakai, T. Vreven, J. A. Montgomery Jr, J. E. Peralta, F. Ogliaro, M. Bearpark, J. J. Heyd, E. Brothers, K. N. Kudin, V. N. Staroverov, T. Keith, R. Kobayashi, J. Normand, K. Raghavachari, A. Rendell, J. C. Burant, S. S. Iyengar, J. Tomasi, M. Cossi, N. Rega, J. M. Millam, M. Klene, J. E. Knox, J. B. Cross, V. Bakken, C. Adamo, J. Jaramillo, R. Gomperts, R. E. Stratmann, O. Yazyev, A. J. Austin, R. Cammi, C. Pomelli, J. W. Ochterski, R. L. Martin, K. Morokuma, V. G. Zakrzewski, G. A. Voth, P. Salvador, J. J. Dannenberg, S. Dapprich, A. D. Daniels, O. Farkas, J. B. Foresman, J. V. Ortiz, J. Cioslowski, and D. J. Fox, Gaussian 09, Revision D. 01, Gaussian Inc, Wallingford CT, 2013.
35. L. Chen and J.-Q. Sun, Comm. Nonlinear Sci. Numer. Simulat., 2016, 41, 1–10.
36. E. Hernandez-Marin, A. Galano and A. Martinez, J. Phys. Chem. B, 2013, 117(15), 4050–4061.
37. A. Cantrell, D. J. McGarvey, T. G. Truscott, F. Rancan and F. Bohm, Arch. Biochem. Biophys., 2003, 412(1), 47–54.
38. A. Farmilo and F. Wilkinson, Photochem. Photobiol., 1973, 18(6), 447–450.
39. A. J. Young and G. M. Lowe, Arch. Biochem. Biophys., 2001, 385(1), 20–27.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c6ra19639j

---