Correlation among Singlet-Oxygen Quenching ... - ACS Publications

Subscriber access provided by LAURENTIAN UNIV

Article

Correlation Among Singlet-Oxygen Quenching, Free-Radical Scavenging, and Excited-State Intramolecular-Proton-Transfer Activities in Hydroxyflavones, Anthocyanidins, and 1-Hydroxyanthraquinones Shin-ichi Nagaoka, Yuki Bandoh, Umpei Nagashima, and Keishi Ohara J. Phys. Chem. A, Just Accepted Manuscript • DOI: 10.1021/acs.jpca.7b07869 • Publication Date (Web): 29 Sep 2017 Downloaded from http://pubs.acs.org on October 5, 2017

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

The Journal of Physical Chemistry A is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 46

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

Correlation

among

Singlet-Oxygen

Scavenging,

and

Excited-State

Activities

in

Hydroxyflavones,

Quenching,

Free-Radical

Intramolecular-Proton-Transfer Anthocyanidins,

and

1-Hydroxyanthraquinones Shin-ichi Nagaoka,*,† Yuki Bandoh,† Umpei Nagashima,‡ and Keishi Ohara†

†

Department of Chemistry, Faculty of Science and Graduate School of Science and

Engineering, Ehime University, Matsuyama 790-8577, Japan ‡

Foundation for Computational Science, 7-1-28 Minatojima-minami-machi, Chuo-ku,

Kobe 650-0047, Japan

ABSTRACT:

(1O2)

Singlet-oxygen

quenching,

free-radical

scavenging,

and

excited-state intramolecular proton-transfer (ESIPT) activities of hydroxyflavones, anthocyanidins, and 1-hydroxyanthraquinones have been studied by means of laser, stopped-flow, and steady-state spectroscopies.

In hydroxyflavones and anthocyanidins,

the 1O2 quenching activity positively correlates to the free-radical scavenging activity. The reason for this correlation can be understood by considering that an early step of each reaction involves electron transfer from the unfused phenyl ring (B-ring), which is singly-bonded to the bicyclic chromen or chromenylium moiety (A- and C-rings). Substitution of an electron-donating OH group at B-ring enhances the electron transfer leading to activation of the

1

O2 quenching and free-radical scavenging.

In

3-hydroxyflavones, the OH substitution at B-ring reduces the activity of ESIPT within C-ring, which can be explained in terms of the nodal-plane model.

As a result, the 1O2

quenching and free-radical scavenging activities negatively correlate to the ESIPT activity.

A catechol structure at B-ring is another factor that enhances the free-radical

scavenging

in

hydroxyflavones.

In

contrast

to

1

ACS Paragon Plus Environment

these

hydroxyflavones,


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

1-hydroxyanthraquinones having an electron-donating OH substituent adjacent to the O–H---O=C moiety susceptible to ESIPT do not show a simple correlation between their 1O2 quenching and ESIPT activities, because the OH substitution modulates these reactions.

2


Page 2 of 46

Page 3 of 46

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


1. INTRODUCTION Belonging to very fundamental processes in chemistry and biochemistry, the antioxidant reaction1 and excited-state intramolecular proton transfer (ESIPT)2–5 are in the focus of interest for a lot of researchers.

Those are also associated with the

maintenance of life and/or significant applications in several fields. reaction1 includes singlet-oxygen (1O2,

The antioxidant

1

∆g state) quenching6 and free-radical

scavenging,7,8 between whose activities a positive correlation is seen in the case of vitamin E derivatives.9 In the 1∆g state of 1O2, two paired electrons occupy the same πg orbital.

On the other hand, a main class of UV absorbers achieves UV protection by

using ESIPT,10 which can dissipate harmful energy of UV radiation generating 1O2 and free-radicals.1

Like this, ESIPT and creation/annihilation of 1O2 and free-radicals are

frequently related to one another. In previous papers,11–13 Nagaoka and co-workers studied ESIPT and

1

O2

quenching activities of some intramolecularly hydrogen-bonded anthraquinone derivatives, and found that their ESIPT activity positively correlated to their 1O2 quenching activity.12,13

The reason for this correlation could be understood by

considering ESIPT-induced distortion of their ground-state potential surface and their encounter complex formation with 1O2.13

Their free-radical scavenging activity was

found to be negligible.11 There remain, however, some important questions to be investigated further. First, is the positive correlation seen between the ESIPT and 1O2 quenching activities in the anthraquinone derivatives12,13 generally found in various intramolecularly hydrogen-bonded molecules?

If not, especially if a negative correlation is revealed in

some molecules, an investigation is needed to clarify the reason for the difference from the positive correlation found in the anthraquinone derivatives mentioned above.12,13 Second, is the free-radical scavenging activity always negligible in various intramolecularly hydrogen-bonded molecules?

If some molecules scavenge a

3



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 4 of 46

free-radical in contrast to the anthraquinone derivatives,11 it is desirable to examine whether or not the scavenging activity positively correlates to the 1O2 quenching activity as in the case of vitamin E derivatives.9

Third, in the situation that the

intramolecular hydrogen bond is removed from the molecules concerned and ESIPT is absent, what becomes of the 1O2 quenching and free-radical scavenging activities? Fourth, are those activity phenomena consistent with the nodal-plane model?

Nagaoka

and co-workers have explained ESIPTs of various molecules, including the above-mentioned anthraquinone derivatives, in terms of the nodal-plane model,14 which is a simple but informative tool for prediction of ESIPT properties.4

Many chemists

have cited the nodal-plane model and recognized usefulness of the explanation even in recent years.2–4

The nodal-plane model is applicable not only to ESIPT but also to

photorearrangement of o-vinyl diaryl ether,15 photoisomerization of bacteriorhodopsin,16 boron coordination,17 and so on.18 In the present work, to answer the above-mentioned questions, we have studied the 1

O2 quenching, free-radical scavenging, and ESIPT activities of hydroxyflavones (HFs),

anthocyanidins (ACs), and 1-hydroxyanthraquinones (1HAQs) by means of laser, stopped-flow, and steady-state spectroscopies.

HFs and ACs (Chart 1) belong to

flavonoids, which are a group of natural substances with variable phenolic structures, are found in various foods,19 and have miscellaneous favorable biochemical and antioxidant effects associated with various diseases such as cancer, Alzheimer’s disease, atherosclerosis, and so on.1,19

In HFs and ACs, an unfused phenyl ring (B-ring) is

singly-bonded to a bicyclic chromen or chromenylium moiety (A- and C-rings). 3-Hydroxyflavones (3HFs, R3 = OH in HFs shown in Chart 1) are well-studied molecules undergoing ESIPT.3,5

ACs are a class of flavonoids lacking

intramolecularly hydrogen-bonded C=O on HFs, and are pigments responsible for colors in plants, flowers, and fruits.19

1HAQs are a large group of natural quinones,

and a wide body of literature has demonstrated that 1HAQs possess a broad range of 4


Page 5 of 46

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


bioactivities such as anticancer, anti-inflammatory, and antioxidant activities.20

As

reported previously,12,13 the anthraquinone derivatives having a substituent spatially apart from the O–H---O=C or N–H---O=C moiety susceptible to ESIPT show a simple positive correlation between the 1O2 quenching and ESIPT activities.

In the present

work, we have studied 1HAQs having a OH substituent adjacent to the O–H---O=C moiety (Chart 1), which may modulate ESIPT together with the 1O2 quenching and spoil the simple correlation between their activities. Furthermore, we have also interpreted ESIPTs of 3HFs and 1HAQs on the basis of the nodal-plane model,14 and have found that the nodal-plane model for 3HFs does not contradict principally with a previous approach based on a zwitterionic tautomer.3,5

2. EXPERIMENTAL SECTION 2.1. Sample Preparation.

HFs, ACs, 1HAQs (Chart 1), rose bengal, and ethanol

used for our measurements were commercially available and used without further purification.

3-Hydroxyflavone (3-HF), 3,4’-dihydroxyflavone (3,4’-DHF), purpurin

(1,2,4-THAQ), and rose bengal were obtained from Tokyo Chemical Industry. 3,3’,4’-Trihydroxyflavone (3,3’,4’-THF) and 3’,4’-dihydroxyflavone (3’,4’-DHF) were purchased from INDOFINE Chemical Company.

Pelargonidin chloride (4’-AC) and

delphinidin chloride (3’,4’,5’-AC) were supplied from ChromaDex. Cyanidin chloride (3’,4’-AC), alizarin (1,2-DHAQ), anthragallol (anthracene brown, 1,2,3-THAQ), and ethanol (EtOH) were obtained from PhytoLab, Acros Organics, Santa Cruz Biotechnology,

and

Wako

Pure

Chemical

Industries,

respectively.

2,6-di-t-butyl-4-(4-methoxyphenyl)phenoxyl (ArO•, Chart 1) was prepared as reported previously.21

2.2. 1O2 Quenching. temperature

The 1O2 quenching (reaction 1) was measured at room

with a Hamamatsu C7990-01 near-infrared fluorescence-lifetime 5



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 6 of 46

measurement system, in which the second harmonic of a Nd:YAG laser (532 nm) was used as the excitation light. 1

kQ → 3O2 + quenching product

O2 + AOH

(1)

where kQ and AOH denote the second-order rate-constant and an antioxidant reagent, respectively.

The setup and experimental procedures were described in detail in

previous papers.11–13

A brief explanation of the experimental protocols for the 1O2

quenching is also given in Supporting Information.

2.3. Free-Radical Scavenging.

Kinetic data of the free-radical scavenging

reactions (reaction 2) in EtOH at 25 °C were obtained by using a Unisoku RSP-1000-3F stopped-flow spectrophotometer or a Shimadzu UV–vis spectrophotometer (UV -1800). ArO• + AOH where

ks,

ks → ArOH + AO•

ArOH,

and

AO•

(2) denote

the

second-order

rate-constant,

2,6-di-t-butyl-4-(4-methoxyphenyl)phenol, and a phenoxyl radical produced from AOH, respectively.

The setup and experimental procedures were described in detail in

previous papers.11,22 with nitrogen gas.

Before all the measurements, the sample solutions were bubbled Equal volumes of EtOH solutions of an AOH and ArO• were mixed

under pseudo-first-order conditions ([AOH] >> [ArO•], where the brackets [ ] indicate molar concentration). The [AOH] value was chosen so that [ArO•] would not largely decrease from the initial concentration within the short period of time (10–20 ms) required to completely mix AOH with ArO• and to make the solution homogeneous. A change in absorption spectrum measured during the reaction between ArO• and 3,3’,4’-THF (AOH = 3,3’,4’-THF in reaction 2) is shown in Figure 1.

Although ArO•

in EtOH is stable in the absence of 3,3’,4’-THF, when an EtOH solution containing excess 3,3’,4’-THF is added to the solution, the ArO• absorption peak disappears immediately.

This observation indicates that 3,3’,4’-THF scavenges ArO•. 6


Figure S3

Page 7 of 46

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


shows the absorbance decays of ArO• at 578 nm during the ArO• scavenging by 3,3’,4’-THF.

Each decay is well-characterized as a single-exponential decay and the

decay rate (kad) increases as [3,3’,4’-THF] increases. The ks value of the ArO• scavenging by 3,3’,4’-THF can be determined by measuring the kad value of ArO• in the solution.

According to ref 22, the kad value is

given by kad = kad0 + ks[3,3’,4’-THF]

(3)

where kad0 stands for the first-order rate-constant for the natural decay of ArO• in EtOH, and is much less than ks[3,3’,4’-THF] under our experimental conditions. shows the dependence of the kad value on prepared [3,3’,4’-THF].

Figure S4

Since [3,3’,4’-THF]

was nearly constant during the ArO• scavenging under the pseudo-first-order conditions ([3,3’,4’-THF] >> [ArO•]), the ks value was evaluated from the slope of the kad versus [3,3’,4’-THF] plot.

The ks values of HFs, ACs, and 1HAQs were also obtained in

similar ways.

2.4. ESIPT.

The absorption spectra of 3HFs and 1HAQs were measured in

EtOH at room temperature with a Shimadzu UV mini-1240 spectrophotometer. fluorescence

spectra

were

spectrofluorophotometer.

obtained

in

EtOH

with

a

Shimadzu

The

RF-5000

The setup and experimental procedures were described in

detail in previous papers.11–13

Before all the measurements, the sample solutions were

bubbled with nitrogen gas.

3. RESULTS The second-order rate-constants of the 1O2 quenching and ArO• scavenging (kQ and ks, respectively) of HFs, ACs, and 1HAQs (Chart 1) are listed in Table 1. Antioxidant capacities22,23 (see Supporting Information) estimated for these molecules are given in Table S1. Some of the kQ and ks values given in Table 1 are comparable 7



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 8 of 46

to, or somewhat less than, those of α-tocopherol (vitamin E),11,24 which is well-known as an efficient 1O2 quencher and free-radical scavenger.25 Figures 2 and 3 respectively show the absorption and fluorescence spectra of 3HFs and 1HAQs in EtOH, some of which exhibit dual fluorescence in short and long wavelength regions (SWE and LWE).

Table 1 also indicates comparison in

fluorescence-intensity between SWE and LWE in 3HFs and 1HAQs together with the red-shift from the absorption peak (Stokes-shift) for LWE.

Currently available data of

the quantum yields of SWE and LWE (and/or their sum) are given in Table S1. The absorption and fluorescence spectra of 3-HF here obtained agree well with those reported previously.26

The kQ value of 3’4’-AC is similar to those of cyanidin

3-glucoside, its 3-rutinoside, and its 3-galactoside measured in 1% HCl methanol solution.27

The kQ value, absorption spectrum, and fluorescence spectrum of

1,2-DHAQ are close to those obtained previously,28,29 and the spectra are consistent with a very recent computational result.30

The ks value is not negligible in some 3HFs

in contrast to those in some intramolecularly hydrogen-bonded anthraquinone derivatives studied previously.11

4. DISCUSSION 4.1. ESIPT in 3HFs. (Figure 2a).

3-HF shows LWE with an unusually great Stokes-shift

This observation can be explained in terms of ESIPT (Figure 4):2–5

photoexcitation of the normal form (stable S0-state species, S0(N)) produces the S1(N) state, in which ESIPT along the O–H---O=C intramolecular hydrogen-bond to produce the S1(T) state takes place rapidly, stabilizing the S1 state.

The S1(T) state decays to

the S0(T) state through LWE or non-radiative transition.

Then the reverse proton

transfer takes place in the S0(T) state and the S0(N) species is regenerated.

The

proton-transferred form shows LWE at a lower energy, and a significant amount of the absorbed

photon

energy

is

dissipated

as

8


heat

during

the

Page 9 of 46

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


S0(N)→S1(N)→S1(T)→S0(T)→S0(N) cycle.

The greatly Stokes-shifted LWE is thus

an indication of ESIPT, which causes distortion and displacement of the potential surfaces of the S0 and S1 states. Scheme 1 indicates a previous approach to ESIPT mechanism of 3HFs.3,5

LWE

of 3-HF (Figure 2a) was proposed to originate from an excited state of a zwitterionic tautomer (S1(T) state), which could be stabilized by two resonant structures shown in Scheme 1a.

SWE seen in Figure 2a was explained to come from an excited state of an

intermolecularly-hydrogen-bonded normal form (Chart 2).26

LWE of 3HFs with an

electron-donating group bonded to B-ring (Figures 2b and c) was also proposed to originate from a zwitterionic tautomer (S1(T) state, Scheme 1b) similar to that of 3-HF (Scheme 1a), whereas its SWE was described to come from a normal form with electron-transfer character (S1(N) state, Scheme 1b).5

However, the zwitterionic

tautomer employed in this previous approach may be rather unstable owing to Coulomb attraction between the nearby + and – charges within the single molecule, as suggested previously.14 On the other hand, there is an alternative approach based on the nodal-plane model,2,14 by using which ESIPT mechanism of 3-HF was previously explained.18 Furthermore, the nodal-plane model was also applied to ESIPT of a flavonoid in another previous paper.31 influence on ESIPT.

In the nodal-plane model, the presence of A-ring has a great Figure 5a shows the nodal patterns of the wave functions in the

S0 and S1 states of 3-HF.

As in the case of a particle in a two-dimensional rectangular

potential box,32 the wave function in the S0 state has no nodal plane, whereas the S1 wave function has one nodal plane.

In ref 18, we regarded the intramolecularly

proton-transferred form as a structure including double-bond delocalization among C8a, O1, and C2 (right-hand side structure of Scheme 2), but it would resonate the structure given in Figure 5a (left-hand side structure of Scheme 2).

9



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 10 of 46

In the S0 state of 3-HF, π electrons are distributed mainly over the three rings, and are used to form double bonds.

In the S1 state (1(π,π*) state) whose wave function has

one nodal plane perpendicular to the molecular plane, the π electrons are distributed except on the nodal plane.

Accordingly, one can write at least two double bonds along

the nodal plane in the S1 state, and one double bond is across the nodal plane from the other, as in the excited states of benzene (Chart 3).

As shown in Figure 5a, when the

two double bonds are C5=C6 and C7=C8 in 3-HF, lone π electrons are seemingly localized at C4a and C8a atoms.

If ESIPT yielding the proton-transferred form takes

place in the S1 state, the two lone π electrons can be significantly delocalized (curved blue arrow in Figure 5a) owing to formation of C4=C4a bond and C2–C8a bond (or double-bond delocalization among C8a, O1, and C2, Scheme 2).

The lone π electrons

facilitate the rearrangement of the bonds to produce the proton-transferred form.

The

S1(T) state then becomes low in energy owing to the delocalization, and the S1 species is susceptible to the intramolecular proton transfer because of the favorable nodal pattern. In contrast to the S1(T) state, the S0 state has the wave function with no nodal plane in 3-HF.

Accordingly, the proton-transferred form is not preferred in the S0 state

(Figure 4) because C4=C4a and C2–C8a bonds cannot be formed: if these bonds were formed, C4a and C8a atoms would become pentavalent.

ESIPT causes distortion and

displacement of the potential surfaces of the S0 and S1 states (Figure 4), and the greatly Stokes-shifted LWE is observed (Table 1).

Like this, the nodal-plane model for 3-HF

(Figure 5a) does not contradict principally with the previous approach (Scheme 1), except that the S1(T) species based on the nodal-plane model (Scheme 2) does not suffer from the above-mentioned destabilization due to the intramolecular Coulomb attraction. As the number of OH group bonded to B-ring increases in 3HFs, the relative intensity and Stokes-shift of LWE decrease (Table 1 and Figures 2a–c).

The

nodal-plane model is again consistent with this substituent effect on ESIPT in 3HFs. In 3,4’-DHF, the delocalization of the lone π electron on C8a in the S1 state (curved blue 10


Page 11 of 46

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


arrow in Figure 5b) is hindered by electron-donating property of the OH group bonded at the 4’-position (curved red arrow).

Coulomb repulsion between the electron

delocalized through ESIPT and the electron donated by the OH group destabilizes the S1(T) state.

As a result, the S1(T) state of 3,4’-DHF is less stable than that of 3-HF,

and ESIPT is not so activated.

Accordingly, the intensity of LWE is less than that of

SWE in 3,4’-DHF (Figure 2b and SWE>LWE in Table 1) in contrast to the case of 3-HF (SWE>LWE in Table 1) can also be explained similarly in terms of the nodal-plane model.

4.2. Correlation between Antioxidant Activities in Flavonoids (HFs and ACs). Figure 6 shows a plot between the

1

O2 quenching-rate (kQ) and the ArO•

scavenging-rate (ks) for flavonoids (HFs and ACs) in EtOH. correlation is seen between these two antioxidant activities.

A positive and linear

The plots for both of HFs

and ACs conform to a common straight line, although ACs are a class of flavonoids lacking intramolecularly hydrogen-bonded C4=O4 on HFs. As indicated in Figure 6 and Table 1, the ks and kQ values of ACs are much greater than those of HFs (HFs 4’-hydroxyflavone) and between 3,3’,4’-THF and 3’4’-DHF (3,3’,4’-THF > 3’4’-DHF) can again be seen in ks.

Thus,

factor-2 enhances not only the 1O2 quenching but also the ArO• scavenging in 3HFs. The ESIPT-induced distortion of the ground-state potential surface may also enhance the ArO• scavenging activity as well as the 1O2 quenching activity. As mentioned above, besides the large/small relations in the kQ values of all the HFs, the relations in the ks values, 3,4’-DHF > 4’-hydroxyflavone and 3,3’,4’-THF > 3’4’-DHF (Table 1), are also consistent with the idea of factor-2.

Nevertheless, the ks

relation between 3,4’-DHF and 3’,4’-DHF, both of which have two electron-donating OH groups, is not consistent with the idea of factor-2.

In fact, the ks value of

3’,4’-DHF without the O3–H3---O4=C4 intramolecular hydrogen bonding (17 M-1s-1) is no less than five times as great as that of 3,4’-DHF with it (3.4 M-1s-1).

The reason for

this will be explained in Section 4.4.

4.3. Correlation between ESIPT and Antioxidant Activities in 3HFs.

Figures

7 and S5 show correlations between ESIPT (LWE) and antioxidant activities (kQ and ks) in 3HFs.

As the relative intensity of LWE decreases, the kQ and ks values increase

(Figure 7); that is, as ESIPT becomes less active, the 1O2 quenching and ArO• scavenging activities are more enhanced, and a negative correlation is found between ESIPT and these antioxidant activities in 3HFs.

A similar negative correlation can also

be found between the antioxidant activities and the Stokes-shift of LWE that decreases with the decrease in ESIPT susceptibility (Figure S5). In contrast to this, a positive correlation was previously found between the LWE 14


Page 15 of 46

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


intensity and kQ value in some intramolecularly hydrogen-bonded anthraquinone derivatives.12,13

In these anthraquinone derivatives, the reason for the positive

correlation could be understood by considering the ESIPT-induced distortion of their ground-state potential surface and their encounter complex formation with 1O2 (factor-2 mentioned in Section 4.2).13

Then, the ESIPT susceptibility increases the kQ value and

also the ks value in some cases. In addition to factor-2, the electron transfer being accelerated by the electron-donating OH group bonded to B-ring (factor-1) also enhances the kQ and ks values in 3HFs, as noted in Section 4.2.

Then, the electron-donating property hinders

ESIPT as explained by using the nodal-plane model in Section 4.1 (Figure 5b). Accordingly, factor-1 hinders ESIPT, which induces factor-2, although both of factors-1 and -2 increase the antioxidant activities in 3HFs as shown in Section 4.2.

If

factor-1 has a greater influence on the antioxidant activities than factor-2, a negative correlation will be found between the ESIPT and antioxidant activities.

To the

contrary, if factor-1 had a less influence on the antioxidant activities than factor-2, a positive correlation would be found.

In reality, since negative correlations are seen in

Figures 7 and S5, factor-1 is thought to enhance the antioxidant activities more than factor-2 in 3HFs.

4.4. Another Factor Enhancing Free-Radical Scavenging in Flavonoids. There would be another factor that enhances the free-radical scavenging (reaction 5) in flavonoids as reported previously.41,42

For example (one can see other examples in

Supporting Information), in 3’,4’-DHF, the ortho arrangement of the OH groups bonded to B-ring (catechol structure) leads to formation of an O–H---O–H intramolecular hydrogen-bond (Scheme 3a).

In the catechol-type B-ring, the H atom of the hydroxy

group written in green in Scheme 3a is susceptible to hydrogen donation, because the intramolecular hydrogen bonding to the hydroxy O-atom weakens the O–H bond 15



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

written in green.

Page 16 of 46

In fact, the O–H---O–H intramolecular hydrogen-bonding seemingly

makes the O-atom pseudo-trivalent (C–O, O–H, and the intramolecular hydrogen bond), although it is inherently divalent.

As a result, the O–H bond-order decreases, the bond

is weakened, the hydroxy H-atom is easily pushed out of B-ring, and the hydrogen transfer to ArO• easily takes place from the catechol-type B-ring in 3’4’-DHF.

Hence,

AO• radical (reaction 2) produced at the catechol-type B-ring of 3’4’-DHF is more stable than 4’-phenoxyl AO• radical formed at B-ring of 4’-hydroxyflavone, and the ks value of 3’,4’-DHF becomes greater than that of 4’-hydroxyflavone (Table 1). Although the numbers of the electron-donating OH group bonded to 3’,4’-DHF and 3,4’-DHF are the same as each other and factor-2 enhances the ks value of 3,4’-DHF, the ks value of 3’,4’-DHF is no less than five times as great as that of 3,4’-DHF (Section 4.2).

The reason for this can be explained similarly by the presence

and absence of the catechol-type B-ring in 3’,4’-DHF and 3,4’-DHF, respectively.

The

catechol-type B-ring of 3’,4’-DHF enhances the ArO• scavenging. In contrast to the above-mentioned HFs, 3-HF has only one OH group not at B-ring but at the 3-position of C-ring.

H3 hydrogen transfer to ArO• is not active in

3-HF, because it requires breaking of O3–H3---O4=C4 intramolecular hydrogen-bond.43 As a result, the ks value of 3-HF is negligible (Table 1), as well as the previously-reported results of some anthraquinone derivatives having an intramolecular hydrogen bond between O–H and C=O.11

The reason why the ks value is not

negligible in 3HFs having an OH group bonded to B-ring is that the hydrogen transfer to ArO• takes place from the OH group bonded to B-ring being spatially apart from the O3–H3---O4=C4 intramolecular hydrogen-bond.

4.5. 1HAQs Having OH Group at 2 Position.

1-HAQ shows ESIPT-induced

LWE (Figure 3a), which is not proposed to originate from a zwitterionic tautomer but can be explained in terms of the nodal-plane model.44

Figure 8a shows the nodal

16


Page 17 of 46

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


patterns of the wave functions in the S0 and S1 states of 1-HAQ.

If ESIPT yielding the

proton-transferred form takes place in the S1 state, two lone π electrons on C1 and C9a atoms can be significantly delocalized (curved blue arrow in Figure 8a) owing to formation of C1=O1 and C9=C9a bonds.

The S1(T) state is then stabilized in

comparison with the S1(N) state, ESIPT is activated, and LWE dominates over SWE that comes from the S1(N)→S0(N) transition (Figure 4). In Figures 3a–c, SWE and LWE of 1HAQs are not clearly resolved from each other in contrast to those of 3HFs.

However, the Stokes-shift of LWE decreases in

order of 1-HAQ > 1,2-DHAQ > 1,2,3-THAQ (Table 1), and the decrease is an indication of decreased ESIPT susceptibility as noted in Section 4.1.

Accordingly,

ESIPT is likely to be suppressed in the order of 1-HAQ > 1,2-DHAQ > 1,2,3-THAQ. Large/small relation between SWE and LWE intensities in 1HAQs can be expected as listed in Table 1. The nodal-plane model is again consistent with this substituent effect on ESIPT in 1HAQs.

In ESIPT of 1,2-DHAQ, the lone π electron is still left on C1 atom owing to

electron-donation from the OH group bonded at the 2-position (curved red arrow in Figure 8b); that is, the delocalization of the lone π electron on C1 in the S1 state (curved blue arrow) is hindered by the electron-donating property of the OH group. the S1(T) state is not much stabilized, and ESIPT is not so activated.

As a result,

Accordingly, the

Stokes-shift of LWE in 1,2-DHAQ is less than that in 1-HAQ (1,2-DHAQ < 1-HAQ, Figure 3 and Table 1).

The substituent effect on the LWE Stokes-shift (1,2,3-THAQ

Correlation among Singlet-Oxygen Quenching ... - ACS Publications

Recommend Documents