Theoretical Study of the Peripheral Disulfide Bridge Substituent Effects

Central Laboratory, College of Chemistry, Nankai University, Tianjin 300071, People's Republic of China, and Key Laboratory of Bioactive Materials, Mi...
1 downloads 0 Views 1MB Size
1008

J. Phys. Chem. A 2010, 114, 1008–1016

Theoretical Study of the Peripheral Disulfide Bridge Substituent Effects on the Antioxidant Properties of Naphthyridine Diol Derivatives Ao Yu,*,† Jian Wang,† Xiaosong Xue,† and Yongjian Wang*,‡ Central Laboratory, College of Chemistry, Nankai UniVersity, Tianjin 300071, People’s Republic of China, and Key Laboratory of BioactiVe Materials, Ministry of Education, College of Life Sciences, Nankai UniVersity, Tianjin 300071, People’s Republic of China ReceiVed: September 8, 2009; ReVised Manuscript ReceiVed: NoVember 13, 2009

The effect of a peripheral disulfide bridge substituent on the phenolic O-H bond dissociation energy (BDE) and the ionization potential (IP) of naphthyridine diol has been studied by density functional theory (DFT) calculation. Compared with naphthalene diol, the substituent of a peripheral disulfide bridge group is very efficient in reducing the BDE, whereas the insertion of nitrogen atoms into the naphthalenic ring only slightly changes the BDE of O-H bond but dramatically enhances the IP. It is similar with the stereoelectronic effect of the heterocyclic ring for the well-known R-tocopherol antioxidant and leads to a highly delocalized spin distribution. With the incorporation of these two aspects, a potential antioxidant is expected to be more active and more stable than R-tocopherol. CHART 1

1. Introduction Free radical chain oxidation is one of the most ubiquitous phenomena in nature. Some processes of oxidation reactions may result in the deterioration of foods and materials as well as various degenerative diseases.1-4 Consequently, to explore potential antioxidants which can efficiently retard or prevent the radical chain oxidation reactions is always a very interesting research object.5-9 By far, natural and synthetic phenols are the most abundant and widely used antioxidants. When acting as radical inhibitors in autoxidation systems, they can easily transfer the H-atom to peroxyl or alkoxyl radicals in the degradation reactions for the low O-H bond dissociation energy (BDE) (eq 1).

ArOH + ROO• f ArO• + ROOH

(1)

Therefore, the lower the BDE(O-H) value is, the easier the dissociation of the O-H bond will be, and also the easier the phenols react with the radicals. Thus, the BDE of O-H bond is one of the most important parameters to start with while designing a potential antioxidant. Generally, electron-donating (ED) groups at the ortho- and para-positions relative to the phenolic O-H can weaken the O-H bond and thus enhance the reactivity of the O-H group to peroxyl radicals.10-17 This strategy has been proven to be the primary guideline for rational design of the novel and more effective antioxidants. Nevertheless, the efforts to increase the transferring rate of H-atom to peroxyl radical by ED substituent have not been thoroughly successful. The reason is that the substitution of phenols with strong ED groups decreases the BDE(O-H) values and also the ionization potentials (IPs) of molecules at the same time which in turn renders the direct reaction with oxygen through electron transfer of phenolic compounds (eq 2). For instance, because of the sensitivity to oxygen in the air, 1,8-napthalene * Corresponding authors: Ao Yu ([email protected]) and Yongjian Wang ([email protected]). † Central Laboratory, College of Chemistry. ‡ Key Laboratory of Bioactive Materials, Ministry of Education, College of Life Sciences.

diol (1, Chart 1) was found to be useless as an antioxidant thought it was predicted to have a lower BDE(O-H) than R-tocopherol (R-TOH, 2, the best-known example of a phenolic antioxidant, Chart 1) by approximately 7 kcal/mol.18 So, the ionization potential of a molecule is assumed to be another crucial index for designing a novel potential antioxidant. Besides these, the other criteria, including solubility, bioavailability, and nontoxicity, must also be considered when designing an effective and safe antioxidant.19,20 However, considering the activity of a potential antioxidant, BDE(O-H) and IP, which directly determine if a phenolic compound could act as a radical inhibitor, are the two most significant parameters.

ArOH + O2 f [ArOH]•+ + O2•-

(2)

The rapid development of quantum chemistry and computation methodologies allows the reliable BDEs and IPs of compounds to be calculated with equivalent or greater accuracy than those obtained from experiments. Theoretical calculations, therefore, could be used as a cogent tool for predicting the relationship between the structure and activity of a compound and also for designing novel potential antioxidants. To our pleasure, the strategy not only remarkably shortens the designing period but also reduces the blindness. Up to now, there are several successful examples in rational interpretation of

10.1021/jp908658z  2010 American Chemical Society Published on Web 12/11/2009

Antioxidant Properties of Naphthyridine Diol Derivatives CHART 2

J. Phys. Chem. A, Vol. 114, No. 2, 2010 1009 CHART 3

CHART 4

structure-activity relation (SAR) of some natural antioxidants21-23 and design of novel antioxidants14-18,24 using the powerful and economical quantum chemical methods. Presently, in order to explore a potential antioxidant with more activity than R-TOH and more inert to oxygen, Pratt and co-workers14,16,17,25 have made great progress in investigating the effects of the incorporation of a nitrogen atom into phenolic ring on BDEs and IPs using both theoretical and experimental methods (4-6, Chart 2). The results show that the insertion of a nitrogen atom into the phenolic ring just slightly changes the BDE of the O-H bond but dramatically enhances the IP of the target molecule which could make the target molecule stable in the air. Subsequently, substituent 3-pyridinols and 5-pyrimidinols, such as compounds 7-10, were examined to be the excellent peroxyl radical scavengers in the process of autoxidation with the similar BDE but very high IP compared with corresponding no N-substitution compounds in organic solution. Unfortunately, although 88-fold more reactive toward peroxyl radical than R-TOH, compound 10 decomposes significantly in air (ca. 30% in 6 h). Therefore, the lower BDE and much higher IP than R-TOH are two key factors to develop the novel antioxidants. Additionally, up to now, most of the above rationally designed antioxidants are the monophenolic compounds based on the R-TOH template, and the naphthols, especially naphthalene diols, catch little attention.18,26 Here, the 1,8-naphthalene diol and 2,7-naphthalene diol were selected as the targets to investigate the effects of the nitrogen atom insertion and hence to develop a new model system for antioxidant design. As an essential property of an antioxidant, it is well-known that antioxidant-derived radicals should not propagate the chain oxidation of a substance. However, Bowry and co-workers27,28 reported in the early 1990s that even R-TOH, the most famous bioactive antioxidant, could mediate the peroxidation of LDL lipids in the absence of coantioxidants, such as ascorbate, and render it a prooxidant. So, it is important to find a new group which not only can provide a low BDE but can reduce the activity of an antioxidant-derived radical to substance. The ortho-substitution of bulky groups, such as tert-butyl, is regarded as the universal means to maintain the kinetic and thermodynamic stability of phenoxyl radicals.29,30 Nevertheless, this method may ultimately be counterproductive under certain conditions. BO-653 (3, Chart 1), another potential synthetic antioxidant with two tert-butyls in the ortho-position of phenolic O-H designed by Niki and co-workers,19 has low H-atom abstraction reactivity although its phenoxyl radical is more stable than that of R-TOH due to the steric hindrance of two bulky ortho groups.7 Accordingly, the strategy described as purely

electronic effects in origin should be taken into account here. It is well-known that delocalization of spin density of the phenoxyl radical is an efficient method to reduce its reactivity.31,32 In general, CH3O-, CH3S-, and (CH3)2N- at the othro or para position of phenolic hydroxyl could fully stabilize the phenoxyl radicals for the contribution of the p lone-pair electron conjugating with the aromatic ring plane. Interestingly, since the late 1970s, Haddon and co-workers33-35 had reported the substituent peripheral disulfide bridge group could be effective in the delocalization of spin density and concomitant stabilization of radicals (Chart 3), and they reported in 2007 the first radical, 11, based on a single phenalenyl unit which was stabilized against σ-dimerization in the solid state by electronic effects rather than by steric bulky substituents. Their studies also revealed that the disulfide may not be the reactive center under general conditions. To the best of our knowledge, the disulfide bond usually exists in many proteins (i.e., lysozyme and immunoglobulin). Therefore, we suggested the peripheral disulfide bridge in novel potential antioxidants may not cause additional toxicity. Encouraged by these results, we introduced a disulfide bridge group into our present system. As part of the continuing investigation for designing effective antioxidants using DFT calculation, first, we systematically investigated the effects of the incorporation of nitrogen atoms into the naphtholic ring and the substituent of a peripheral disulfide bridge group on the BDE(O-H) and IP values (Chart 4). Then we compared the effects of two other chalcogen atoms, O and Se, and the familiar strong electron-donating groups on the BDE and IP with the first step result and, in turn, developed an optimum new strategy for antioxidant design. Finally, three original compounds with potential antioxidant activity were predicted. 2. Computational Methodology All of the theoretical calculations here were carried out using the GAUSSIAN 0336 packages mounted on the NK-Star Supercomputer. The geometries of the neutral molecules were fully optimized by the B3LYP method in conjunction with the 6-311G(d,p) basis set, and the UB3LYP procedure was used for the geometry optimization of the corresponding radicals. Frequency was computed on these geometries at the same level

1010

J. Phys. Chem. A, Vol. 114, No. 2, 2010

Yu et al.

SCHEME 1

to verify that they are real minima on the potential-energy surface without any imaginary frequency. The zero point energy (ZPE) and thermal corrections were also calculated using the same method (unscaled). For the molecules with more than one possibly stable conformation, the one with the lowest electronic energy was singled out for the following calculations except for special illustration. Single point energy calculations were then carried out with an extended basis set of 6-311++G(2df,2p). The B3LYP and restricted open shell formalism ROB3LYP with the large basis set were applied for the neutral molecules and open shell radicals at the (U)B3LYP/6-311G(d,p) optimized geometry. The enthalpy of a species at 298.15 K was obtained using the equation

H(298.15 K) ) E0 + ZPE + Htrans + Hrot + Hvib + RT (3) The homolytic O-H bond dissociation enthalpy [BDE(O-H)] of each molecule (X-A-OH) was calculated as the enthalpy change in gas phase at 298.15 K of the following equation (eq 4), where the enthalpy of the H atom was estimated by taking its exact electronic energy of -0.5 hartree.

X-A-OH(g) f X-A-O•(g) + H•(g)

(4)

The gas-phase adiabatic ionization potential (IP) of X-A-OH is defined as the amount of energy required to remove an electron from a molecule at 0 K. It was computed as the energy change of the equation

X-A-OH(g) f X-A-OH•+H(g) + e-(g)

(5)

For both the neutral molecule and its radical cation, the electronic energies and the unscaled zero point vibrational energies were taken at the (U)B3LYP/6-311G(d,p) level for IP calculation. The distribution of spin density of radicals was also calculated at the UB3LYP/6-311G(d,p) level. 3. Results and Discussion 3.1. Bond Dissociation Energies and Ionization Potentials of Compounds 14-17. The model (RO)B3LYP/6311++G(2df,2p)//(U)B3LYP/6-311G(d,p) was chosen to

estimate the BDE and IP values which were first proposed by DiLabio and co-workers37 and had been demonstrated to be an optimum model for obtaining more reliable BDE values of various types of bonds, including X-H (X ) C, O, S, and P).38-42 The compounds investigated in this research are shown in Chart 4. The calculated values for compounds 14-17 are listed in Scheme 1. The results show that, compared with the parent molecule, the insertion of nitrogen atoms at the meta-positions of the naphtholic ring relative to hydroxyl increases the BDE(O-H) slightly by 2.18 kcal/mol and the substituent of a peripheral disulfide bridge group decreases it dramatically by 13.17 kcal/ mol; with the two effects combined together, the apparent BDE value decreases 8.50 kcal/mol (Scheme 1). To explore how the insertion of nitrogen atoms and substitution of a peripheral disulfide bridge group affect the BDE, the ground-state effect (GE), radical effect (RE), and total effect (TE ) RE - GE) were calculated based on the reaction enthalpies of the isodesmic reactions shown in Scheme 2. These calculation methods have been used successfully to investigate the substituent effect on the BDE values.25,43-45 Here, the GE elucidates how the ground-state stability of neutral molecule changes with the introduction of a substituent heteroatom or group, the RE provides the same effect on the stability of radical, and the GE and RE together determine the tendency of the BDE value to change. As indicated in Scheme 3, both the meta-position N-atom replacement and peripheral disulfide bridge substitution cause a destabilizing effect on the neutral molecule (with a negative GE), but an inverse effect exists for the corresponding radical, that is, the meta-position N-atom replacement causes a destabilizing effect (RE ) -3.82 kcal/mol) while the peripheral disulfide bridge substitution stabilizes the radical to a much larger extent (RE ) 9.99 kcal/mol). Moreover, we found the value changes for the GE and RE of the combination of the two effects are -4.64 and 3.86 kcal/mol, respectively. It should be noted that the total GE or RE cannot be attributed to the simple linear addition of the two separate effects but exhibits a cutdown effect (viz., for GE, 4.64 kcal/mol